- Linux-stable-mirror - lists.linaro.org

[Linux-stable-mirror] Patch "usb: Add device quirk for Logitech HD Pro Webcam C925e" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usb: Add device quirk for Logitech HD Pro Webcam C925e to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usb-add-device-quirk-for-logitech-hd-pro-webcam-c925e.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 7f038d256c723dd390d2fca942919573995f4cfd Mon Sep 17 00:00:00 2001 From: Dmitry Fleytman Dmitry Fleytman <dmitry.fleytman(a)gmail.com> Date: Tue, 19 Dec 2017 06:02:04 +0200 Subject: usb: Add device quirk for Logitech HD Pro Webcam C925e From: Dmitry Fleytman Dmitry Fleytman <dmitry.fleytman(a)gmail.com> commit 7f038d256c723dd390d2fca942919573995f4cfd upstream. Commit e0429362ab15 ("usb: Add device quirk for Logitech HD Pro Webcams C920 and C930e") introduced quirk to workaround an issue with some Logitech webcams. There is one more model that has the same issue - C925e, so applying the same quirk as well. See aforementioned commit message for detailed explanation of the problem. Signed-off-by: Dmitry Fleytman <dmitry.fleytman(a)gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/core/quirks.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/usb/core/quirks.c +++ b/drivers/usb/core/quirks.c @@ -53,10 +53,11 @@ static const struct usb_device_id usb_qu /* Microsoft LifeCam-VX700 v2.0 */ { USB_DEVICE(0x045e, 0x0770), .driver_info = USB_QUIRK_RESET_RESUME }, - /* Logitech HD Pro Webcams C920, C920-C and C930e */ + /* Logitech HD Pro Webcams C920, C920-C, C925e and C930e */ { USB_DEVICE(0x046d, 0x082d), .driver_info = USB_QUIRK_DELAY_INIT }, { USB_DEVICE(0x046d, 0x0841), .driver_info = USB_QUIRK_DELAY_INIT }, { USB_DEVICE(0x046d, 0x0843), .driver_info = USB_QUIRK_DELAY_INIT }, + { USB_DEVICE(0x046d, 0x085b), .driver_info = USB_QUIRK_DELAY_INIT }, /* Logitech ConferenceCam CC3000e */ { USB_DEVICE(0x046d, 0x0847), .driver_info = USB_QUIRK_DELAY_INIT }, Patches currently in stable-queue which might be from dmitry.fleytman(a)gmail.com are queue-3.18/usb-add-device-quirk-for-logitech-hd-pro-webcam-c925e.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] FAILED: patch "[PATCH] phy: Add Tegra XUSB pad controller support" failed to apply to 4.9-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.9-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 53d2a715c24034ee4017f3c15c82bb4a53a07da5 Mon Sep 17 00:00:00 2001 From: Thierry Reding <treding(a)nvidia.com> Date: Wed, 11 Nov 2015 18:24:21 +0100 Subject: [PATCH] phy: Add Tegra XUSB pad controller support Add a new driver for the XUSB pad controller found on NVIDIA Tegra SoCs. This hardware block used to be exposed as a pin controller, but it turns out that this isn't a good fit. The new driver and DT binding much more accurately describe the hardware and are more flexible in supporting new SoC generations. Acked-by: Kishon Vijay Abraham I <kishon(a)ti.com> Signed-off-by: Thierry Reding <treding(a)nvidia.com> diff --git a/drivers/phy/Kconfig b/drivers/phy/Kconfig index 26566db09de0..27e5f6ee9a2a 100644 --- a/drivers/phy/Kconfig +++ b/drivers/phy/Kconfig @@ -421,4 +421,6 @@ config PHY_CYGNUS_PCIE Enable this to support the Broadcom Cygnus PCIe PHY. If unsure, say N. +source "drivers/phy/tegra/Kconfig" + endmenu diff --git a/drivers/phy/Makefile b/drivers/phy/Makefile index 24596a96a887..d4f06e69fd9a 100644 --- a/drivers/phy/Makefile +++ b/drivers/phy/Makefile @@ -52,3 +52,5 @@ obj-$(CONFIG_PHY_TUSB1210) += phy-tusb1210.o obj-$(CONFIG_PHY_BRCMSTB_SATA) += phy-brcmstb-sata.o obj-$(CONFIG_PHY_PISTACHIO_USB) += phy-pistachio-usb.o obj-$(CONFIG_PHY_CYGNUS_PCIE) += phy-bcm-cygnus-pcie.o + +obj-$(CONFIG_ARCH_TEGRA) += tegra/ diff --git a/drivers/phy/tegra/Kconfig b/drivers/phy/tegra/Kconfig new file mode 100644 index 000000000000..a3b1de953fb7 --- /dev/null +++ b/drivers/phy/tegra/Kconfig @@ -0,0 +1,8 @@ +config PHY_TEGRA_XUSB + tristate "NVIDIA Tegra XUSB pad controller driver" + depends on ARCH_TEGRA + help + Choose this option if you have an NVIDIA Tegra SoC. + + To compile this driver as a module, choose M here: the module will + be called phy-tegra-xusb. diff --git a/drivers/phy/tegra/Makefile b/drivers/phy/tegra/Makefile new file mode 100644 index 000000000000..31150b4337cd --- /dev/null +++ b/drivers/phy/tegra/Makefile @@ -0,0 +1,5 @@ +obj-$(CONFIG_PHY_TEGRA_XUSB) += phy-tegra-xusb.o + +phy-tegra-xusb-y += xusb.o +phy-tegra-xusb-$(CONFIG_ARCH_TEGRA_124_SOC) += xusb-tegra124.o +phy-tegra-xusb-$(CONFIG_ARCH_TEGRA_132_SOC) += xusb-tegra124.o diff --git a/drivers/phy/tegra/xusb-tegra124.c b/drivers/phy/tegra/xusb-tegra124.c new file mode 100644 index 000000000000..119957249a51 --- /dev/null +++ b/drivers/phy/tegra/xusb-tegra124.c @@ -0,0 +1,1752 @@ +/* + * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#include <linux/delay.h> +#include <linux/io.h> +#include <linux/mailbox_client.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/phy/phy.h> +#include <linux/platform_device.h> +#include <linux/regulator/consumer.h> +#include <linux/reset.h> +#include <linux/slab.h> + +#include <soc/tegra/fuse.h> + +#include "xusb.h" + +#define FUSE_SKU_CALIB_HS_CURR_LEVEL_PADX_SHIFT(x) ((x) ? 15 : 0) +#define FUSE_SKU_CALIB_HS_CURR_LEVEL_PAD_MASK 0x3f +#define FUSE_SKU_CALIB_HS_IREF_CAP_SHIFT 13 +#define FUSE_SKU_CALIB_HS_IREF_CAP_MASK 0x3 +#define FUSE_SKU_CALIB_HS_SQUELCH_LEVEL_SHIFT 11 +#define FUSE_SKU_CALIB_HS_SQUELCH_LEVEL_MASK 0x3 +#define FUSE_SKU_CALIB_HS_TERM_RANGE_ADJ_SHIFT 7 +#define FUSE_SKU_CALIB_HS_TERM_RANGE_ADJ_MASK 0xf + +#define XUSB_PADCTL_USB2_PORT_CAP 0x008 +#define XUSB_PADCTL_USB2_PORT_CAP_PORTX_CAP_SHIFT(x) ((x) * 4) +#define XUSB_PADCTL_USB2_PORT_CAP_PORT_CAP_MASK 0x3 +#define XUSB_PADCTL_USB2_PORT_CAP_DISABLED 0x0 +#define XUSB_PADCTL_USB2_PORT_CAP_HOST 0x1 +#define XUSB_PADCTL_USB2_PORT_CAP_DEVICE 0x2 +#define XUSB_PADCTL_USB2_PORT_CAP_OTG 0x3 + +#define XUSB_PADCTL_SS_PORT_MAP 0x014 +#define XUSB_PADCTL_SS_PORT_MAP_PORTX_INTERNAL(x) (1 << (((x) * 4) + 3)) +#define XUSB_PADCTL_SS_PORT_MAP_PORTX_MAP_SHIFT(x) ((x) * 4) +#define XUSB_PADCTL_SS_PORT_MAP_PORTX_MAP_MASK(x) (0x7 << ((x) * 4)) +#define XUSB_PADCTL_SS_PORT_MAP_PORTX_MAP(x, v) (((v) & 0x7) << ((x) * 4)) +#define XUSB_PADCTL_SS_PORT_MAP_PORT_MAP_MASK 0x7 + +#define XUSB_PADCTL_ELPG_PROGRAM 0x01c +#define XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_VCORE_DOWN (1 << 26) +#define XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_CLAMP_EN_EARLY (1 << 25) +#define XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_CLAMP_EN (1 << 24) +#define XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_VCORE_DOWN(x) (1 << (18 + (x) * 4)) +#define XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_CLAMP_EN_EARLY(x) \ + (1 << (17 + (x) * 4)) +#define XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_CLAMP_EN(x) (1 << (16 + (x) * 4)) + +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL1 0x040 +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL1_PLL0_LOCKDET (1 << 19) +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL1_REFCLK_SEL_MASK (0xf << 12) +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL1_PLL_RST (1 << 1) + +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL2 0x044 +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL2_REFCLKBUF_EN (1 << 6) +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL2_TXCLKREF_EN (1 << 5) +#define XUSB_PADCTL_IOPHY_PLL_P0_CTL2_TXCLKREF_SEL (1 << 4) + +#define XUSB_PADCTL_IOPHY_USB3_PADX_CTL2(x) (0x058 + (x) * 4) +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_CDR_CNTL_SHIFT 24 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_CDR_CNTL_MASK 0xff +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_CDR_CNTL_VAL 0x24 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_SHIFT 16 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_MASK 0x3f +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_SHIFT 8 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_MASK 0x3f +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_SHIFT 8 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_MASK 0xffff +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_VAL 0xf070 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_WANDER_SHIFT 4 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_WANDER_MASK 0xf +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_WANDER_VAL 0xf + +#define XUSB_PADCTL_IOPHY_USB3_PADX_CTL4(x) (0x068 + (x) * 4) +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_SHIFT 24 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_MASK 0x1f +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_SHIFT 16 +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_MASK 0x7f +#define XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_VAL 0x002008ee + +#define XUSB_PADCTL_IOPHY_MISC_PAD_PX_CTL2(x) ((x) < 2 ? 0x078 + (x) * 4 : \ + 0x0f8 + (x) * 4) +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL2_SPARE_IN_SHIFT 28 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL2_SPARE_IN_MASK 0x3 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL2_SPARE_IN_VAL 0x1 + +#define XUSB_PADCTL_IOPHY_MISC_PAD_PX_CTL5(x) ((x) < 2 ? 0x090 + (x) * 4 : \ + 0x11c + (x) * 4) +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL5_RX_QEYE_EN (1 << 8) + +#define XUSB_PADCTL_IOPHY_MISC_PAD_PX_CTL6(x) ((x) < 2 ? 0x098 + (x) * 4 : \ + 0x128 + (x) * 4) +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SHIFT 24 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_G_Z_MASK 0x3f +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_TAP_MASK 0x1f +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_AMP_MASK 0x7f +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT 16 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_MASK 0xff +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_G_Z 0x21 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_TAP 0x32 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_AMP 0x33 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_CTLE_Z 0x48 +#define XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_LATCH_G_Z 0xa1 + +#define XUSB_PADCTL_USB2_OTG_PADX_CTL0(x) (0x0a0 + (x) * 4) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_PD_ZI (1 << 21) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_PD2 (1 << 20) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_PD (1 << 19) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_LS_RSLEW_SHIFT 14 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_LS_RSLEW_MASK 0x3 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_LS_RSLEW_VAL(x) ((x) ? 0x0 : 0x3) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_SLEW_SHIFT 6 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_SLEW_MASK 0x3f +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_SLEW_VAL 0x0e +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_CURR_LEVEL_SHIFT 0 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_CURR_LEVEL_MASK 0x3f + +#define XUSB_PADCTL_USB2_OTG_PADX_CTL1(x) (0x0ac + (x) * 4) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL1_HS_IREF_CAP_SHIFT 9 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL1_HS_IREF_CAP_MASK 0x3 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL1_TERM_RANGE_ADJ_SHIFT 3 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL1_TERM_RANGE_ADJ_MASK 0x7 +#define XUSB_PADCTL_USB2_OTG_PAD_CTL1_PD_DR (1 << 2) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL1_PD_DISC_FORCE_POWERUP (1 << 1) +#define XUSB_PADCTL_USB2_OTG_PAD_CTL1_PD_CHRP_FORCE_POWERUP (1 << 0) + +#define XUSB_PADCTL_USB2_BIAS_PAD_CTL0 0x0b8 +#define XUSB_PADCTL_USB2_BIAS_PAD_CTL0_PD (1 << 12) +#define XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_DISCON_LEVEL_SHIFT 2 +#define XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_DISCON_LEVEL_MASK 0x7 +#define XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_DISCON_LEVEL_VAL 0x5 +#define XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_SQUELCH_LEVEL_SHIFT 0 +#define XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_SQUELCH_LEVEL_MASK 0x3 + +#define XUSB_PADCTL_HSIC_PADX_CTL0(x) (0x0c0 + (x) * 4) +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWN_SHIFT 12 +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWN_MASK 0x7 +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWP_SHIFT 8 +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWP_MASK 0x7 +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEN_SHIFT 4 +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEN_MASK 0x7 +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEP_SHIFT 0 +#define XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEP_MASK 0x7 + +#define XUSB_PADCTL_HSIC_PADX_CTL1(x) (0x0c8 + (x) * 4) +#define XUSB_PADCTL_HSIC_PAD_CTL1_RPU_STROBE (1 << 10) +#define XUSB_PADCTL_HSIC_PAD_CTL1_RPU_DATA (1 << 9) +#define XUSB_PADCTL_HSIC_PAD_CTL1_RPD_STROBE (1 << 8) +#define XUSB_PADCTL_HSIC_PAD_CTL1_RPD_DATA (1 << 7) +#define XUSB_PADCTL_HSIC_PAD_CTL1_PD_ZI (1 << 5) +#define XUSB_PADCTL_HSIC_PAD_CTL1_PD_RX (1 << 4) +#define XUSB_PADCTL_HSIC_PAD_CTL1_PD_TRX (1 << 3) +#define XUSB_PADCTL_HSIC_PAD_CTL1_PD_TX (1 << 2) +#define XUSB_PADCTL_HSIC_PAD_CTL1_AUTO_TERM_EN (1 << 0) + +#define XUSB_PADCTL_HSIC_PADX_CTL2(x) (0x0d0 + (x) * 4) +#define XUSB_PADCTL_HSIC_PAD_CTL2_RX_STROBE_TRIM_SHIFT 4 +#define XUSB_PADCTL_HSIC_PAD_CTL2_RX_STROBE_TRIM_MASK 0x7 +#define XUSB_PADCTL_HSIC_PAD_CTL2_RX_DATA_TRIM_SHIFT 0 +#define XUSB_PADCTL_HSIC_PAD_CTL2_RX_DATA_TRIM_MASK 0x7 + +#define XUSB_PADCTL_HSIC_STRB_TRIM_CONTROL 0x0e0 +#define XUSB_PADCTL_HSIC_STRB_TRIM_CONTROL_STRB_TRIM_MASK 0x1f + +#define XUSB_PADCTL_USB3_PAD_MUX 0x134 +#define XUSB_PADCTL_USB3_PAD_MUX_PCIE_IDDQ_DISABLE(x) (1 << (1 + (x))) +#define XUSB_PADCTL_USB3_PAD_MUX_SATA_IDDQ_DISABLE(x) (1 << (6 + (x))) + +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1 0x138 +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL1_LOCKDET (1 << 27) +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL1_MODE (1 << 24) +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL0_REFCLK_NDIV_SHIFT 20 +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL0_REFCLK_NDIV_MASK 0x3 +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_PWR_OVRD (1 << 3) +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_RST (1 << 1) +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_IDDQ (1 << 0) + +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2 0x13c +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL1_CP_CNTL_SHIFT 20 +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL1_CP_CNTL_MASK 0xf +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL0_CP_CNTL_SHIFT 16 +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL0_CP_CNTL_MASK 0xf +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_TCLKOUT_EN (1 << 12) +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_TXCLKREF_SEL (1 << 4) +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_XDIGCLK_SEL_SHIFT 0 +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL2_XDIGCLK_SEL_MASK 0x7 + +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL3 0x140 +#define XUSB_PADCTL_IOPHY_PLL_S0_CTL3_RCAL_BYPASS (1 << 7) + +#define XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1 0x148 +#define XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1_IDDQ_OVRD (1 << 1) +#define XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1_IDDQ (1 << 0) + +#define XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL2 0x14c + +#define XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL5 0x158 + +#define XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL6 0x15c + +struct tegra124_xusb_fuse_calibration { + u32 hs_curr_level[3]; + u32 hs_iref_cap; + u32 hs_term_range_adj; + u32 hs_squelch_level; +}; + +struct tegra124_xusb_padctl { + struct tegra_xusb_padctl base; + + struct tegra124_xusb_fuse_calibration fuse; +}; + +static inline struct tegra124_xusb_padctl * +to_tegra124_xusb_padctl(struct tegra_xusb_padctl *padctl) +{ + return container_of(padctl, struct tegra124_xusb_padctl, base); +} + +static int tegra124_xusb_padctl_enable(struct tegra_xusb_padctl *padctl) +{ + u32 value; + + mutex_lock(&padctl->lock); + + if (padctl->enable++ > 0) + goto out; + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value &= ~XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_CLAMP_EN; + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(100, 200); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value &= ~XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_CLAMP_EN_EARLY; + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(100, 200); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value &= ~XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_VCORE_DOWN; + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + +out: + mutex_unlock(&padctl->lock); + return 0; +} + +static int tegra124_xusb_padctl_disable(struct tegra_xusb_padctl *padctl) +{ + u32 value; + + mutex_lock(&padctl->lock); + + if (WARN_ON(padctl->enable == 0)) + goto out; + + if (--padctl->enable > 0) + goto out; + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value |= XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_VCORE_DOWN; + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(100, 200); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value |= XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_CLAMP_EN_EARLY; + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(100, 200); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value |= XUSB_PADCTL_ELPG_PROGRAM_AUX_MUX_LP0_CLAMP_EN; + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + +out: + mutex_unlock(&padctl->lock); + return 0; +} + +static int tegra124_usb3_save_context(struct tegra_xusb_padctl *padctl, + unsigned int index) +{ + struct tegra_xusb_usb3_port *port; + struct tegra_xusb_lane *lane; + u32 value, offset; + + port = tegra_xusb_find_usb3_port(padctl, index); + if (!port) + return -ENODEV; + + port->context_saved = true; + lane = port->base.lane; + + if (lane->pad == padctl->pcie) + offset = XUSB_PADCTL_IOPHY_MISC_PAD_PX_CTL6(lane->index); + else + offset = XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL6; + + value = padctl_readl(padctl, offset); + value &= ~(XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_MASK << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT); + value |= XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_TAP << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT; + padctl_writel(padctl, value, offset); + + value = padctl_readl(padctl, offset) >> + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SHIFT; + port->tap1 = value & XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_TAP_MASK; + + value = padctl_readl(padctl, offset); + value &= ~(XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_MASK << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT); + value |= XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_AMP << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT; + padctl_writel(padctl, value, offset); + + value = padctl_readl(padctl, offset) >> + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SHIFT; + port->amp = value & XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_AMP_MASK; + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_USB3_PADX_CTL4(index)); + value &= ~((XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_SHIFT)); + value |= (port->tap1 << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_SHIFT) | + (port->amp << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_SHIFT); + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_USB3_PADX_CTL4(index)); + + value = padctl_readl(padctl, offset); + value &= ~(XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_MASK << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT); + value |= XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_LATCH_G_Z << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT; + padctl_writel(padctl, value, offset); + + value = padctl_readl(padctl, offset); + value &= ~(XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_MASK << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT); + value |= XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_G_Z << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT; + padctl_writel(padctl, value, offset); + + value = padctl_readl(padctl, offset) >> + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SHIFT; + port->ctle_g = value & + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_G_Z_MASK; + + value = padctl_readl(padctl, offset); + value &= ~(XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_MASK << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT); + value |= XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_CTLE_Z << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SEL_SHIFT; + padctl_writel(padctl, value, offset); + + value = padctl_readl(padctl, offset) >> + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_SHIFT; + port->ctle_z = value & + XUSB_PADCTL_IOPHY_MISC_PAD_CTL6_MISC_OUT_G_Z_MASK; + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_USB3_PADX_CTL2(index)); + value &= ~((XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_SHIFT)); + value |= (port->ctle_g << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_SHIFT) | + (port->ctle_z << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_SHIFT); + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_USB3_PADX_CTL2(index)); + + return 0; +} + +static int tegra124_hsic_set_idle(struct tegra_xusb_padctl *padctl, + unsigned int index, bool idle) +{ + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + + if (idle) + value |= XUSB_PADCTL_HSIC_PAD_CTL1_RPD_DATA | + XUSB_PADCTL_HSIC_PAD_CTL1_RPU_STROBE; + else + value &= ~(XUSB_PADCTL_HSIC_PAD_CTL1_RPD_DATA | + XUSB_PADCTL_HSIC_PAD_CTL1_RPU_STROBE); + + padctl_writel(padctl, value, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + + return 0; +} + +#define TEGRA124_LANE(_name, _offset, _shift, _mask, _type) \ + { \ + .name = _name, \ + .offset = _offset, \ + .shift = _shift, \ + .mask = _mask, \ + .num_funcs = ARRAY_SIZE(tegra124_##_type##_functions), \ + .funcs = tegra124_##_type##_functions, \ + } + +static const char * const tegra124_usb2_functions[] = { + "snps", + "xusb", + "uart", +}; + +static const struct tegra_xusb_lane_soc tegra124_usb2_lanes[] = { + TEGRA124_LANE("usb2-0", 0x004, 0, 0x3, usb2), + TEGRA124_LANE("usb2-1", 0x004, 2, 0x3, usb2), + TEGRA124_LANE("usb2-2", 0x004, 4, 0x3, usb2), +}; + +static struct tegra_xusb_lane * +tegra124_usb2_lane_probe(struct tegra_xusb_pad *pad, struct device_node *np, + unsigned int index) +{ + struct tegra_xusb_usb2_lane *usb2; + int err; + + usb2 = kzalloc(sizeof(*usb2), GFP_KERNEL); + if (!usb2) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&usb2->base.list); + usb2->base.soc = &pad->soc->lanes[index]; + usb2->base.index = index; + usb2->base.pad = pad; + usb2->base.np = np; + + err = tegra_xusb_lane_parse_dt(&usb2->base, np); + if (err < 0) { + kfree(usb2); + return ERR_PTR(err); + } + + return &usb2->base; +} + +static void tegra124_usb2_lane_remove(struct tegra_xusb_lane *lane) +{ + struct tegra_xusb_usb2_lane *usb2 = to_usb2_lane(lane); + + kfree(usb2); +} + +static const struct tegra_xusb_lane_ops tegra124_usb2_lane_ops = { + .probe = tegra124_usb2_lane_probe, + .remove = tegra124_usb2_lane_remove, +}; + +static int tegra124_usb2_phy_init(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_enable(lane->pad->padctl); +} + +static int tegra124_usb2_phy_exit(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_disable(lane->pad->padctl); +} + +static int tegra124_usb2_phy_power_on(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_usb2_lane *usb2 = to_usb2_lane(lane); + struct tegra_xusb_usb2_pad *pad = to_usb2_pad(lane->pad); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + struct tegra124_xusb_padctl *priv; + struct tegra_xusb_usb2_port *port; + unsigned int index = lane->index; + u32 value; + int err; + + port = tegra_xusb_find_usb2_port(padctl, index); + if (!port) { + dev_err(&phy->dev, "no port found for USB2 lane %u\n", index); + return -ENODEV; + } + + priv = to_tegra124_xusb_padctl(padctl); + + value = padctl_readl(padctl, XUSB_PADCTL_USB2_BIAS_PAD_CTL0); + value &= ~((XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_SQUELCH_LEVEL_MASK << + XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_SQUELCH_LEVEL_SHIFT) | + (XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_DISCON_LEVEL_MASK << + XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_DISCON_LEVEL_SHIFT)); + value |= (priv->fuse.hs_squelch_level << + XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_SQUELCH_LEVEL_SHIFT) | + (XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_DISCON_LEVEL_VAL << + XUSB_PADCTL_USB2_BIAS_PAD_CTL0_HS_DISCON_LEVEL_SHIFT); + padctl_writel(padctl, value, XUSB_PADCTL_USB2_BIAS_PAD_CTL0); + + value = padctl_readl(padctl, XUSB_PADCTL_USB2_PORT_CAP); + value &= ~(XUSB_PADCTL_USB2_PORT_CAP_PORT_CAP_MASK << + XUSB_PADCTL_USB2_PORT_CAP_PORTX_CAP_SHIFT(index)); + value |= XUSB_PADCTL_USB2_PORT_CAP_HOST << + XUSB_PADCTL_USB2_PORT_CAP_PORTX_CAP_SHIFT(index); + padctl_writel(padctl, value, XUSB_PADCTL_USB2_PORT_CAP); + + value = padctl_readl(padctl, XUSB_PADCTL_USB2_OTG_PADX_CTL0(index)); + value &= ~((XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_CURR_LEVEL_MASK << + XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_CURR_LEVEL_SHIFT) | + (XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_SLEW_MASK << + XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_SLEW_SHIFT) | + (XUSB_PADCTL_USB2_OTG_PAD_CTL0_LS_RSLEW_MASK << + XUSB_PADCTL_USB2_OTG_PAD_CTL0_LS_RSLEW_SHIFT) | + XUSB_PADCTL_USB2_OTG_PAD_CTL0_PD | + XUSB_PADCTL_USB2_OTG_PAD_CTL0_PD2 | + XUSB_PADCTL_USB2_OTG_PAD_CTL0_PD_ZI); + value |= (priv->fuse.hs_curr_level[index] + + usb2->hs_curr_level_offset) << + XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_CURR_LEVEL_SHIFT; + value |= XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_SLEW_VAL << + XUSB_PADCTL_USB2_OTG_PAD_CTL0_HS_SLEW_SHIFT; + value |= XUSB_PADCTL_USB2_OTG_PAD_CTL0_LS_RSLEW_VAL(index) << + XUSB_PADCTL_USB2_OTG_PAD_CTL0_LS_RSLEW_SHIFT; + padctl_writel(padctl, value, XUSB_PADCTL_USB2_OTG_PADX_CTL0(index)); + + value = padctl_readl(padctl, XUSB_PADCTL_USB2_OTG_PADX_CTL1(index)); + value &= ~((XUSB_PADCTL_USB2_OTG_PAD_CTL1_TERM_RANGE_ADJ_MASK << + XUSB_PADCTL_USB2_OTG_PAD_CTL1_TERM_RANGE_ADJ_SHIFT) | + (XUSB_PADCTL_USB2_OTG_PAD_CTL1_HS_IREF_CAP_MASK << + XUSB_PADCTL_USB2_OTG_PAD_CTL1_HS_IREF_CAP_SHIFT) | + XUSB_PADCTL_USB2_OTG_PAD_CTL1_PD_DR | + XUSB_PADCTL_USB2_OTG_PAD_CTL1_PD_CHRP_FORCE_POWERUP | + XUSB_PADCTL_USB2_OTG_PAD_CTL1_PD_DISC_FORCE_POWERUP); + value |= (priv->fuse.hs_term_range_adj << + XUSB_PADCTL_USB2_OTG_PAD_CTL1_TERM_RANGE_ADJ_SHIFT) | + (priv->fuse.hs_iref_cap << + XUSB_PADCTL_USB2_OTG_PAD_CTL1_HS_IREF_CAP_SHIFT); + padctl_writel(padctl, value, XUSB_PADCTL_USB2_OTG_PADX_CTL1(index)); + + err = regulator_enable(port->supply); + if (err) + return err; + + mutex_lock(&pad->lock); + + if (pad->enable++ > 0) + goto out; + + value = padctl_readl(padctl, XUSB_PADCTL_USB2_BIAS_PAD_CTL0); + value &= ~XUSB_PADCTL_USB2_BIAS_PAD_CTL0_PD; + padctl_writel(padctl, value, XUSB_PADCTL_USB2_BIAS_PAD_CTL0); + +out: + mutex_unlock(&pad->lock); + return 0; +} + +static int tegra124_usb2_phy_power_off(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_usb2_pad *pad = to_usb2_pad(lane->pad); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + struct tegra_xusb_usb2_port *port; + u32 value; + + port = tegra_xusb_find_usb2_port(padctl, lane->index); + if (!port) { + dev_err(&phy->dev, "no port found for USB2 lane %u\n", + lane->index); + return -ENODEV; + } + + mutex_lock(&pad->lock); + + if (WARN_ON(pad->enable == 0)) + goto out; + + if (--pad->enable > 0) + goto out; + + value = padctl_readl(padctl, XUSB_PADCTL_USB2_BIAS_PAD_CTL0); + value |= XUSB_PADCTL_USB2_BIAS_PAD_CTL0_PD; + padctl_writel(padctl, value, XUSB_PADCTL_USB2_BIAS_PAD_CTL0); + +out: + regulator_disable(port->supply); + mutex_unlock(&pad->lock); + return 0; +} + +static const struct phy_ops tegra124_usb2_phy_ops = { + .init = tegra124_usb2_phy_init, + .exit = tegra124_usb2_phy_exit, + .power_on = tegra124_usb2_phy_power_on, + .power_off = tegra124_usb2_phy_power_off, + .owner = THIS_MODULE, +}; + +static struct tegra_xusb_pad * +tegra124_usb2_pad_probe(struct tegra_xusb_padctl *padctl, + const struct tegra_xusb_pad_soc *soc, + struct device_node *np) +{ + struct tegra_xusb_usb2_pad *usb2; + struct tegra_xusb_pad *pad; + int err; + + usb2 = kzalloc(sizeof(*usb2), GFP_KERNEL); + if (!usb2) + return ERR_PTR(-ENOMEM); + + mutex_init(&usb2->lock); + + pad = &usb2->base; + pad->ops = &tegra124_usb2_lane_ops; + pad->soc = soc; + + err = tegra_xusb_pad_init(pad, padctl, np); + if (err < 0) { + kfree(usb2); + goto out; + } + + err = tegra_xusb_pad_register(pad, &tegra124_usb2_phy_ops); + if (err < 0) + goto unregister; + + dev_set_drvdata(&pad->dev, pad); + + return pad; + +unregister: + device_unregister(&pad->dev); +out: + return ERR_PTR(err); +} + +static void tegra124_usb2_pad_remove(struct tegra_xusb_pad *pad) +{ + struct tegra_xusb_usb2_pad *usb2 = to_usb2_pad(pad); + + kfree(usb2); +} + +static const struct tegra_xusb_pad_ops tegra124_usb2_ops = { + .probe = tegra124_usb2_pad_probe, + .remove = tegra124_usb2_pad_remove, +}; + +static const struct tegra_xusb_pad_soc tegra124_usb2_pad = { + .name = "usb2", + .num_lanes = ARRAY_SIZE(tegra124_usb2_lanes), + .lanes = tegra124_usb2_lanes, + .ops = &tegra124_usb2_ops, +}; + +static const char * const tegra124_ulpi_functions[] = { + "snps", + "xusb", +}; + +static const struct tegra_xusb_lane_soc tegra124_ulpi_lanes[] = { + TEGRA124_LANE("ulpi-0", 0x004, 12, 0x1, ulpi), +}; + +static struct tegra_xusb_lane * +tegra124_ulpi_lane_probe(struct tegra_xusb_pad *pad, struct device_node *np, + unsigned int index) +{ + struct tegra_xusb_ulpi_lane *ulpi; + int err; + + ulpi = kzalloc(sizeof(*ulpi), GFP_KERNEL); + if (!ulpi) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&ulpi->base.list); + ulpi->base.soc = &pad->soc->lanes[index]; + ulpi->base.index = index; + ulpi->base.pad = pad; + ulpi->base.np = np; + + err = tegra_xusb_lane_parse_dt(&ulpi->base, np); + if (err < 0) { + kfree(ulpi); + return ERR_PTR(err); + } + + return &ulpi->base; +} + +static void tegra124_ulpi_lane_remove(struct tegra_xusb_lane *lane) +{ + struct tegra_xusb_ulpi_lane *ulpi = to_ulpi_lane(lane); + + kfree(ulpi); +} + +static const struct tegra_xusb_lane_ops tegra124_ulpi_lane_ops = { + .probe = tegra124_ulpi_lane_probe, + .remove = tegra124_ulpi_lane_remove, +}; + +static int tegra124_ulpi_phy_init(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_enable(lane->pad->padctl); +} + +static int tegra124_ulpi_phy_exit(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_disable(lane->pad->padctl); +} + +static int tegra124_ulpi_phy_power_on(struct phy *phy) +{ + return 0; +} + +static int tegra124_ulpi_phy_power_off(struct phy *phy) +{ + return 0; +} + +static const struct phy_ops tegra124_ulpi_phy_ops = { + .init = tegra124_ulpi_phy_init, + .exit = tegra124_ulpi_phy_exit, + .power_on = tegra124_ulpi_phy_power_on, + .power_off = tegra124_ulpi_phy_power_off, + .owner = THIS_MODULE, +}; + +static struct tegra_xusb_pad * +tegra124_ulpi_pad_probe(struct tegra_xusb_padctl *padctl, + const struct tegra_xusb_pad_soc *soc, + struct device_node *np) +{ + struct tegra_xusb_ulpi_pad *ulpi; + struct tegra_xusb_pad *pad; + int err; + + ulpi = kzalloc(sizeof(*ulpi), GFP_KERNEL); + if (!ulpi) + return ERR_PTR(-ENOMEM); + + pad = &ulpi->base; + pad->ops = &tegra124_ulpi_lane_ops; + pad->soc = soc; + + err = tegra_xusb_pad_init(pad, padctl, np); + if (err < 0) { + kfree(ulpi); + goto out; + } + + err = tegra_xusb_pad_register(pad, &tegra124_ulpi_phy_ops); + if (err < 0) + goto unregister; + + dev_set_drvdata(&pad->dev, pad); + + return pad; + +unregister: + device_unregister(&pad->dev); +out: + return ERR_PTR(err); +} + +static void tegra124_ulpi_pad_remove(struct tegra_xusb_pad *pad) +{ + struct tegra_xusb_ulpi_pad *ulpi = to_ulpi_pad(pad); + + kfree(ulpi); +} + +static const struct tegra_xusb_pad_ops tegra124_ulpi_ops = { + .probe = tegra124_ulpi_pad_probe, + .remove = tegra124_ulpi_pad_remove, +}; + +static const struct tegra_xusb_pad_soc tegra124_ulpi_pad = { + .name = "ulpi", + .num_lanes = ARRAY_SIZE(tegra124_ulpi_lanes), + .lanes = tegra124_ulpi_lanes, + .ops = &tegra124_ulpi_ops, +}; + +static const char * const tegra124_hsic_functions[] = { + "snps", + "xusb", +}; + +static const struct tegra_xusb_lane_soc tegra124_hsic_lanes[] = { + TEGRA124_LANE("hsic-0", 0x004, 14, 0x1, hsic), + TEGRA124_LANE("hsic-1", 0x004, 15, 0x1, hsic), +}; + +static struct tegra_xusb_lane * +tegra124_hsic_lane_probe(struct tegra_xusb_pad *pad, struct device_node *np, + unsigned int index) +{ + struct tegra_xusb_hsic_lane *hsic; + int err; + + hsic = kzalloc(sizeof(*hsic), GFP_KERNEL); + if (!hsic) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&hsic->base.list); + hsic->base.soc = &pad->soc->lanes[index]; + hsic->base.index = index; + hsic->base.pad = pad; + hsic->base.np = np; + + err = tegra_xusb_lane_parse_dt(&hsic->base, np); + if (err < 0) { + kfree(hsic); + return ERR_PTR(err); + } + + return &hsic->base; +} + +static void tegra124_hsic_lane_remove(struct tegra_xusb_lane *lane) +{ + struct tegra_xusb_hsic_lane *hsic = to_hsic_lane(lane); + + kfree(hsic); +} + +static const struct tegra_xusb_lane_ops tegra124_hsic_lane_ops = { + .probe = tegra124_hsic_lane_probe, + .remove = tegra124_hsic_lane_remove, +}; + +static int tegra124_hsic_phy_init(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_enable(lane->pad->padctl); +} + +static int tegra124_hsic_phy_exit(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_disable(lane->pad->padctl); +} + +static int tegra124_hsic_phy_power_on(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_hsic_lane *hsic = to_hsic_lane(lane); + struct tegra_xusb_hsic_pad *pad = to_hsic_pad(lane->pad); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + unsigned int index = lane->index; + u32 value; + int err; + + err = regulator_enable(pad->supply); + if (err) + return err; + + padctl_writel(padctl, hsic->strobe_trim, + XUSB_PADCTL_HSIC_STRB_TRIM_CONTROL); + + value = padctl_readl(padctl, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + + if (hsic->auto_term) + value |= XUSB_PADCTL_HSIC_PAD_CTL1_AUTO_TERM_EN; + else + value &= ~XUSB_PADCTL_HSIC_PAD_CTL1_AUTO_TERM_EN; + + padctl_writel(padctl, value, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + + value = padctl_readl(padctl, XUSB_PADCTL_HSIC_PADX_CTL0(index)); + value &= ~((XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEN_MASK << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEN_SHIFT) | + (XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEP_MASK << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEP_SHIFT) | + (XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWN_MASK << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWN_SHIFT) | + (XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWP_MASK << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWP_SHIFT)); + value |= (hsic->tx_rtune_n << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEN_SHIFT) | + (hsic->tx_rtune_p << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RTUNEP_SHIFT) | + (hsic->tx_rslew_n << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWN_SHIFT) | + (hsic->tx_rslew_p << + XUSB_PADCTL_HSIC_PAD_CTL0_TX_RSLEWP_SHIFT); + padctl_writel(padctl, value, XUSB_PADCTL_HSIC_PADX_CTL0(index)); + + value = padctl_readl(padctl, XUSB_PADCTL_HSIC_PADX_CTL2(index)); + value &= ~((XUSB_PADCTL_HSIC_PAD_CTL2_RX_STROBE_TRIM_MASK << + XUSB_PADCTL_HSIC_PAD_CTL2_RX_STROBE_TRIM_SHIFT) | + (XUSB_PADCTL_HSIC_PAD_CTL2_RX_DATA_TRIM_MASK << + XUSB_PADCTL_HSIC_PAD_CTL2_RX_DATA_TRIM_SHIFT)); + value |= (hsic->rx_strobe_trim << + XUSB_PADCTL_HSIC_PAD_CTL2_RX_STROBE_TRIM_SHIFT) | + (hsic->rx_data_trim << + XUSB_PADCTL_HSIC_PAD_CTL2_RX_DATA_TRIM_SHIFT); + padctl_writel(padctl, value, XUSB_PADCTL_HSIC_PADX_CTL2(index)); + + value = padctl_readl(padctl, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + value &= ~(XUSB_PADCTL_HSIC_PAD_CTL1_RPD_STROBE | + XUSB_PADCTL_HSIC_PAD_CTL1_RPU_DATA | + XUSB_PADCTL_HSIC_PAD_CTL1_PD_RX | + XUSB_PADCTL_HSIC_PAD_CTL1_PD_ZI | + XUSB_PADCTL_HSIC_PAD_CTL1_PD_TRX | + XUSB_PADCTL_HSIC_PAD_CTL1_PD_TX); + value |= XUSB_PADCTL_HSIC_PAD_CTL1_RPD_DATA | + XUSB_PADCTL_HSIC_PAD_CTL1_RPU_STROBE; + padctl_writel(padctl, value, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + + return 0; +} + +static int tegra124_hsic_phy_power_off(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_hsic_pad *pad = to_hsic_pad(lane->pad); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + unsigned int index = lane->index; + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + value |= XUSB_PADCTL_HSIC_PAD_CTL1_PD_RX | + XUSB_PADCTL_HSIC_PAD_CTL1_PD_ZI | + XUSB_PADCTL_HSIC_PAD_CTL1_PD_TRX | + XUSB_PADCTL_HSIC_PAD_CTL1_PD_TX; + padctl_writel(padctl, value, XUSB_PADCTL_HSIC_PADX_CTL1(index)); + + regulator_disable(pad->supply); + + return 0; +} + +static const struct phy_ops tegra124_hsic_phy_ops = { + .init = tegra124_hsic_phy_init, + .exit = tegra124_hsic_phy_exit, + .power_on = tegra124_hsic_phy_power_on, + .power_off = tegra124_hsic_phy_power_off, + .owner = THIS_MODULE, +}; + +static struct tegra_xusb_pad * +tegra124_hsic_pad_probe(struct tegra_xusb_padctl *padctl, + const struct tegra_xusb_pad_soc *soc, + struct device_node *np) +{ + struct tegra_xusb_hsic_pad *hsic; + struct tegra_xusb_pad *pad; + int err; + + hsic = kzalloc(sizeof(*hsic), GFP_KERNEL); + if (!hsic) + return ERR_PTR(-ENOMEM); + + pad = &hsic->base; + pad->ops = &tegra124_hsic_lane_ops; + pad->soc = soc; + + err = tegra_xusb_pad_init(pad, padctl, np); + if (err < 0) { + kfree(hsic); + goto out; + } + + err = tegra_xusb_pad_register(pad, &tegra124_hsic_phy_ops); + if (err < 0) + goto unregister; + + dev_set_drvdata(&pad->dev, pad); + + return pad; + +unregister: + device_unregister(&pad->dev); +out: + return ERR_PTR(err); +} + +static void tegra124_hsic_pad_remove(struct tegra_xusb_pad *pad) +{ + struct tegra_xusb_hsic_pad *hsic = to_hsic_pad(pad); + + kfree(hsic); +} + +static const struct tegra_xusb_pad_ops tegra124_hsic_ops = { + .probe = tegra124_hsic_pad_probe, + .remove = tegra124_hsic_pad_remove, +}; + +static const struct tegra_xusb_pad_soc tegra124_hsic_pad = { + .name = "hsic", + .num_lanes = ARRAY_SIZE(tegra124_hsic_lanes), + .lanes = tegra124_hsic_lanes, + .ops = &tegra124_hsic_ops, +}; + +static const char * const tegra124_pcie_functions[] = { + "pcie", + "usb3-ss", + "sata", +}; + +static const struct tegra_xusb_lane_soc tegra124_pcie_lanes[] = { + TEGRA124_LANE("pcie-0", 0x134, 16, 0x3, pcie), + TEGRA124_LANE("pcie-1", 0x134, 18, 0x3, pcie), + TEGRA124_LANE("pcie-2", 0x134, 20, 0x3, pcie), + TEGRA124_LANE("pcie-3", 0x134, 22, 0x3, pcie), + TEGRA124_LANE("pcie-4", 0x134, 24, 0x3, pcie), +}; + +static struct tegra_xusb_lane * +tegra124_pcie_lane_probe(struct tegra_xusb_pad *pad, struct device_node *np, + unsigned int index) +{ + struct tegra_xusb_pcie_lane *pcie; + int err; + + pcie = kzalloc(sizeof(*pcie), GFP_KERNEL); + if (!pcie) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&pcie->base.list); + pcie->base.soc = &pad->soc->lanes[index]; + pcie->base.index = index; + pcie->base.pad = pad; + pcie->base.np = np; + + err = tegra_xusb_lane_parse_dt(&pcie->base, np); + if (err < 0) { + kfree(pcie); + return ERR_PTR(err); + } + + return &pcie->base; +} + +static void tegra124_pcie_lane_remove(struct tegra_xusb_lane *lane) +{ + struct tegra_xusb_pcie_lane *pcie = to_pcie_lane(lane); + + kfree(pcie); +} + +static const struct tegra_xusb_lane_ops tegra124_pcie_lane_ops = { + .probe = tegra124_pcie_lane_probe, + .remove = tegra124_pcie_lane_remove, +}; + +static int tegra124_pcie_phy_init(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_enable(lane->pad->padctl); +} + +static int tegra124_pcie_phy_exit(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_disable(lane->pad->padctl); +} + +static int tegra124_pcie_phy_power_on(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + unsigned long timeout; + int err = -ETIMEDOUT; + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_P0_CTL1); + value &= ~XUSB_PADCTL_IOPHY_PLL_P0_CTL1_REFCLK_SEL_MASK; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_P0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_P0_CTL2); + value |= XUSB_PADCTL_IOPHY_PLL_P0_CTL2_REFCLKBUF_EN | + XUSB_PADCTL_IOPHY_PLL_P0_CTL2_TXCLKREF_EN | + XUSB_PADCTL_IOPHY_PLL_P0_CTL2_TXCLKREF_SEL; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_P0_CTL2); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_P0_CTL1); + value |= XUSB_PADCTL_IOPHY_PLL_P0_CTL1_PLL_RST; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_P0_CTL1); + + timeout = jiffies + msecs_to_jiffies(50); + + while (time_before(jiffies, timeout)) { + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_P0_CTL1); + if (value & XUSB_PADCTL_IOPHY_PLL_P0_CTL1_PLL0_LOCKDET) { + err = 0; + break; + } + + usleep_range(100, 200); + } + + value = padctl_readl(padctl, XUSB_PADCTL_USB3_PAD_MUX); + value |= XUSB_PADCTL_USB3_PAD_MUX_PCIE_IDDQ_DISABLE(lane->index); + padctl_writel(padctl, value, XUSB_PADCTL_USB3_PAD_MUX); + + return err; +} + +static int tegra124_pcie_phy_power_off(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_USB3_PAD_MUX); + value &= ~XUSB_PADCTL_USB3_PAD_MUX_PCIE_IDDQ_DISABLE(lane->index); + padctl_writel(padctl, value, XUSB_PADCTL_USB3_PAD_MUX); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_P0_CTL1); + value &= ~XUSB_PADCTL_IOPHY_PLL_P0_CTL1_PLL_RST; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_P0_CTL1); + + return 0; +} + +static const struct phy_ops tegra124_pcie_phy_ops = { + .init = tegra124_pcie_phy_init, + .exit = tegra124_pcie_phy_exit, + .power_on = tegra124_pcie_phy_power_on, + .power_off = tegra124_pcie_phy_power_off, + .owner = THIS_MODULE, +}; + +static struct tegra_xusb_pad * +tegra124_pcie_pad_probe(struct tegra_xusb_padctl *padctl, + const struct tegra_xusb_pad_soc *soc, + struct device_node *np) +{ + struct tegra_xusb_pcie_pad *pcie; + struct tegra_xusb_pad *pad; + int err; + + pcie = kzalloc(sizeof(*pcie), GFP_KERNEL); + if (!pcie) + return ERR_PTR(-ENOMEM); + + pad = &pcie->base; + pad->ops = &tegra124_pcie_lane_ops; + pad->soc = soc; + + err = tegra_xusb_pad_init(pad, padctl, np); + if (err < 0) { + kfree(pcie); + goto out; + } + + err = tegra_xusb_pad_register(pad, &tegra124_pcie_phy_ops); + if (err < 0) + goto unregister; + + dev_set_drvdata(&pad->dev, pad); + + return pad; + +unregister: + device_unregister(&pad->dev); +out: + return ERR_PTR(err); +} + +static void tegra124_pcie_pad_remove(struct tegra_xusb_pad *pad) +{ + struct tegra_xusb_pcie_pad *pcie = to_pcie_pad(pad); + + kfree(pcie); +} + +static const struct tegra_xusb_pad_ops tegra124_pcie_ops = { + .probe = tegra124_pcie_pad_probe, + .remove = tegra124_pcie_pad_remove, +}; + +static const struct tegra_xusb_pad_soc tegra124_pcie_pad = { + .name = "pcie", + .num_lanes = ARRAY_SIZE(tegra124_pcie_lanes), + .lanes = tegra124_pcie_lanes, + .ops = &tegra124_pcie_ops, +}; + +static const struct tegra_xusb_lane_soc tegra124_sata_lanes[] = { + TEGRA124_LANE("sata-0", 0x134, 26, 0x3, pcie), +}; + +static struct tegra_xusb_lane * +tegra124_sata_lane_probe(struct tegra_xusb_pad *pad, struct device_node *np, + unsigned int index) +{ + struct tegra_xusb_sata_lane *sata; + int err; + + sata = kzalloc(sizeof(*sata), GFP_KERNEL); + if (!sata) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&sata->base.list); + sata->base.soc = &pad->soc->lanes[index]; + sata->base.index = index; + sata->base.pad = pad; + sata->base.np = np; + + err = tegra_xusb_lane_parse_dt(&sata->base, np); + if (err < 0) { + kfree(sata); + return ERR_PTR(err); + } + + return &sata->base; +} + +static void tegra124_sata_lane_remove(struct tegra_xusb_lane *lane) +{ + struct tegra_xusb_sata_lane *sata = to_sata_lane(lane); + + kfree(sata); +} + +static const struct tegra_xusb_lane_ops tegra124_sata_lane_ops = { + .probe = tegra124_sata_lane_probe, + .remove = tegra124_sata_lane_remove, +}; + +static int tegra124_sata_phy_init(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_enable(lane->pad->padctl); +} + +static int tegra124_sata_phy_exit(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + return tegra124_xusb_padctl_disable(lane->pad->padctl); +} + +static int tegra124_sata_phy_power_on(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + unsigned long timeout; + int err = -ETIMEDOUT; + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1); + value &= ~XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1_IDDQ_OVRD; + value &= ~XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1_IDDQ; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + value &= ~XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_PWR_OVRD; + value &= ~XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_IDDQ; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + value |= XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL1_MODE; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + value |= XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_RST; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + + timeout = jiffies + msecs_to_jiffies(50); + + while (time_before(jiffies, timeout)) { + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + if (value & XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL1_LOCKDET) { + err = 0; + break; + } + + usleep_range(100, 200); + } + + value = padctl_readl(padctl, XUSB_PADCTL_USB3_PAD_MUX); + value |= XUSB_PADCTL_USB3_PAD_MUX_SATA_IDDQ_DISABLE(lane->index); + padctl_writel(padctl, value, XUSB_PADCTL_USB3_PAD_MUX); + + return err; +} + +static int tegra124_sata_phy_power_off(struct phy *phy) +{ + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_USB3_PAD_MUX); + value &= ~XUSB_PADCTL_USB3_PAD_MUX_SATA_IDDQ_DISABLE(lane->index); + padctl_writel(padctl, value, XUSB_PADCTL_USB3_PAD_MUX); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + value &= ~XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_RST; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + value &= ~XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL1_MODE; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + value |= XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_PWR_OVRD; + value |= XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL_IDDQ; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1); + value |= ~XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1_IDDQ_OVRD; + value |= ~XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1_IDDQ; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL1); + + return 0; +} + +static const struct phy_ops tegra124_sata_phy_ops = { + .init = tegra124_sata_phy_init, + .exit = tegra124_sata_phy_exit, + .power_on = tegra124_sata_phy_power_on, + .power_off = tegra124_sata_phy_power_off, + .owner = THIS_MODULE, +}; + +static struct tegra_xusb_pad * +tegra124_sata_pad_probe(struct tegra_xusb_padctl *padctl, + const struct tegra_xusb_pad_soc *soc, + struct device_node *np) +{ + struct tegra_xusb_sata_pad *sata; + struct tegra_xusb_pad *pad; + int err; + + sata = kzalloc(sizeof(*sata), GFP_KERNEL); + if (!sata) + return ERR_PTR(-ENOMEM); + + pad = &sata->base; + pad->ops = &tegra124_sata_lane_ops; + pad->soc = soc; + + err = tegra_xusb_pad_init(pad, padctl, np); + if (err < 0) { + kfree(sata); + goto out; + } + + err = tegra_xusb_pad_register(pad, &tegra124_sata_phy_ops); + if (err < 0) + goto unregister; + + dev_set_drvdata(&pad->dev, pad); + + return pad; + +unregister: + device_unregister(&pad->dev); +out: + return ERR_PTR(err); +} + +static void tegra124_sata_pad_remove(struct tegra_xusb_pad *pad) +{ + struct tegra_xusb_sata_pad *sata = to_sata_pad(pad); + + kfree(sata); +} + +static const struct tegra_xusb_pad_ops tegra124_sata_ops = { + .probe = tegra124_sata_pad_probe, + .remove = tegra124_sata_pad_remove, +}; + +static const struct tegra_xusb_pad_soc tegra124_sata_pad = { + .name = "sata", + .num_lanes = ARRAY_SIZE(tegra124_sata_lanes), + .lanes = tegra124_sata_lanes, + .ops = &tegra124_sata_ops, +}; + +static const struct tegra_xusb_pad_soc *tegra124_pads[] = { + &tegra124_usb2_pad, + &tegra124_ulpi_pad, + &tegra124_hsic_pad, + &tegra124_pcie_pad, + &tegra124_sata_pad, +}; + +static int tegra124_usb2_port_enable(struct tegra_xusb_port *port) +{ + return 0; +} + +static void tegra124_usb2_port_disable(struct tegra_xusb_port *port) +{ +} + +static struct tegra_xusb_lane * +tegra124_usb2_port_map(struct tegra_xusb_port *port) +{ + return tegra_xusb_find_lane(port->padctl, "usb2", port->index); +} + +static const struct tegra_xusb_port_ops tegra124_usb2_port_ops = { + .enable = tegra124_usb2_port_enable, + .disable = tegra124_usb2_port_disable, + .map = tegra124_usb2_port_map, +}; + +static int tegra124_ulpi_port_enable(struct tegra_xusb_port *port) +{ + return 0; +} + +static void tegra124_ulpi_port_disable(struct tegra_xusb_port *port) +{ +} + +static struct tegra_xusb_lane * +tegra124_ulpi_port_map(struct tegra_xusb_port *port) +{ + return tegra_xusb_find_lane(port->padctl, "ulpi", port->index); +} + +static const struct tegra_xusb_port_ops tegra124_ulpi_port_ops = { + .enable = tegra124_ulpi_port_enable, + .disable = tegra124_ulpi_port_disable, + .map = tegra124_ulpi_port_map, +}; + +static int tegra124_hsic_port_enable(struct tegra_xusb_port *port) +{ + return 0; +} + +static void tegra124_hsic_port_disable(struct tegra_xusb_port *port) +{ +} + +static struct tegra_xusb_lane * +tegra124_hsic_port_map(struct tegra_xusb_port *port) +{ + return tegra_xusb_find_lane(port->padctl, "hsic", port->index); +} + +static const struct tegra_xusb_port_ops tegra124_hsic_port_ops = { + .enable = tegra124_hsic_port_enable, + .disable = tegra124_hsic_port_disable, + .map = tegra124_hsic_port_map, +}; + +static int tegra124_usb3_port_enable(struct tegra_xusb_port *port) +{ + struct tegra_xusb_usb3_port *usb3 = to_usb3_port(port); + struct tegra_xusb_padctl *padctl = port->padctl; + struct tegra_xusb_lane *lane = usb3->base.lane; + unsigned int index = port->index, offset; + int ret = 0; + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_SS_PORT_MAP); + + if (!usb3->internal) + value &= ~XUSB_PADCTL_SS_PORT_MAP_PORTX_INTERNAL(index); + else + value |= XUSB_PADCTL_SS_PORT_MAP_PORTX_INTERNAL(index); + + value &= ~XUSB_PADCTL_SS_PORT_MAP_PORTX_MAP_MASK(index); + value |= XUSB_PADCTL_SS_PORT_MAP_PORTX_MAP(index, usb3->port); + padctl_writel(padctl, value, XUSB_PADCTL_SS_PORT_MAP); + + /* + * TODO: move this code into the PCIe/SATA PHY ->power_on() callbacks + * and conditionalize based on mux function? This seems to work, but + * might not be the exact proper sequence. + */ + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_USB3_PADX_CTL2(index)); + value &= ~((XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_WANDER_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_WANDER_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_CDR_CNTL_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_CDR_CNTL_SHIFT)); + value |= (XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_WANDER_VAL << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_WANDER_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_CDR_CNTL_VAL << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_CDR_CNTL_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_VAL << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_SHIFT); + + if (usb3->context_saved) { + value &= ~((XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_SHIFT)); + value |= (usb3->ctle_g << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_G_SHIFT) | + (usb3->ctle_z << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL2_RX_EQ_Z_SHIFT); + } + + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_USB3_PADX_CTL2(index)); + + value = XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_VAL; + + if (usb3->context_saved) { + value &= ~((XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_SHIFT) | + (XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_MASK << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_SHIFT)); + value |= (usb3->tap1 << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_TAP_SHIFT) | + (usb3->amp << + XUSB_PADCTL_IOPHY_USB3_PAD_CTL4_DFE_CNTL_AMP_SHIFT); + } + + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_USB3_PADX_CTL4(index)); + + if (lane->pad == padctl->pcie) + offset = XUSB_PADCTL_IOPHY_MISC_PAD_PX_CTL2(lane->index); + else + offset = XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL2; + + value = padctl_readl(padctl, offset); + value &= ~(XUSB_PADCTL_IOPHY_MISC_PAD_CTL2_SPARE_IN_MASK << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL2_SPARE_IN_SHIFT); + value |= XUSB_PADCTL_IOPHY_MISC_PAD_CTL2_SPARE_IN_VAL << + XUSB_PADCTL_IOPHY_MISC_PAD_CTL2_SPARE_IN_SHIFT; + padctl_writel(padctl, value, offset); + + if (lane->pad == padctl->pcie) + offset = XUSB_PADCTL_IOPHY_MISC_PAD_PX_CTL5(lane->index); + else + offset = XUSB_PADCTL_IOPHY_MISC_PAD_S0_CTL5; + + value = padctl_readl(padctl, offset); + value |= XUSB_PADCTL_IOPHY_MISC_PAD_CTL5_RX_QEYE_EN; + padctl_writel(padctl, value, offset); + + /* Enable SATA PHY when SATA lane is used */ + if (lane->pad == padctl->sata) { + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + value &= ~(XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL0_REFCLK_NDIV_MASK << + XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL0_REFCLK_NDIV_SHIFT); + value |= 0x2 << + XUSB_PADCTL_IOPHY_PLL_S0_CTL1_PLL0_REFCLK_NDIV_SHIFT; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL1); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL2); + value &= ~((XUSB_PADCTL_IOPHY_PLL_S0_CTL2_XDIGCLK_SEL_MASK << + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_XDIGCLK_SEL_SHIFT) | + (XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL1_CP_CNTL_MASK << + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL1_CP_CNTL_SHIFT) | + (XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL0_CP_CNTL_MASK << + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL0_CP_CNTL_SHIFT) | + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_TCLKOUT_EN); + value |= (0x7 << + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_XDIGCLK_SEL_SHIFT) | + (0x8 << + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL1_CP_CNTL_SHIFT) | + (0x8 << + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_PLL0_CP_CNTL_SHIFT) | + XUSB_PADCTL_IOPHY_PLL_S0_CTL2_TXCLKREF_SEL; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL2); + + value = padctl_readl(padctl, XUSB_PADCTL_IOPHY_PLL_S0_CTL3); + value &= ~XUSB_PADCTL_IOPHY_PLL_S0_CTL3_RCAL_BYPASS; + padctl_writel(padctl, value, XUSB_PADCTL_IOPHY_PLL_S0_CTL3); + } + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value &= ~XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_VCORE_DOWN(index); + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(100, 200); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value &= ~XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_CLAMP_EN_EARLY(index); + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(100, 200); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value &= ~XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_CLAMP_EN(index); + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + return ret; +} + +static void tegra124_usb3_port_disable(struct tegra_xusb_port *port) +{ + struct tegra_xusb_padctl *padctl = port->padctl; + u32 value; + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value |= XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_CLAMP_EN_EARLY(port->index); + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(100, 200); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value |= XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_CLAMP_EN(port->index); + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + usleep_range(250, 350); + + value = padctl_readl(padctl, XUSB_PADCTL_ELPG_PROGRAM); + value |= XUSB_PADCTL_ELPG_PROGRAM_SSPX_ELPG_VCORE_DOWN(port->index); + padctl_writel(padctl, value, XUSB_PADCTL_ELPG_PROGRAM); + + value = padctl_readl(padctl, XUSB_PADCTL_SS_PORT_MAP); + value &= ~XUSB_PADCTL_SS_PORT_MAP_PORTX_MAP_MASK(port->index); + value |= XUSB_PADCTL_SS_PORT_MAP_PORTX_MAP(port->index, 0x7); + padctl_writel(padctl, value, XUSB_PADCTL_SS_PORT_MAP); +} + +static const struct tegra_xusb_lane_map tegra124_usb3_map[] = { + { 0, "pcie", 0 }, + { 1, "pcie", 1 }, + { 1, "sata", 0 }, + { 0, NULL, 0 }, +}; + +static struct tegra_xusb_lane * +tegra124_usb3_port_map(struct tegra_xusb_port *port) +{ + return tegra_xusb_port_find_lane(port, tegra124_usb3_map, "usb3-ss"); +} + +static const struct tegra_xusb_port_ops tegra124_usb3_port_ops = { + .enable = tegra124_usb3_port_enable, + .disable = tegra124_usb3_port_disable, + .map = tegra124_usb3_port_map, +}; + +static int +tegra124_xusb_read_fuse_calibration(struct tegra124_xusb_fuse_calibration *fuse) +{ + unsigned int i; + int err; + u32 value; + + err = tegra_fuse_readl(TEGRA_FUSE_SKU_CALIB_0, &value); + if (err < 0) + return err; + + for (i = 0; i < ARRAY_SIZE(fuse->hs_curr_level); i++) { + fuse->hs_curr_level[i] = + (value >> FUSE_SKU_CALIB_HS_CURR_LEVEL_PADX_SHIFT(i)) & + FUSE_SKU_CALIB_HS_CURR_LEVEL_PAD_MASK; + } + fuse->hs_iref_cap = + (value >> FUSE_SKU_CALIB_HS_IREF_CAP_SHIFT) & + FUSE_SKU_CALIB_HS_IREF_CAP_MASK; + fuse->hs_term_range_adj = + (value >> FUSE_SKU_CALIB_HS_TERM_RANGE_ADJ_SHIFT) & + FUSE_SKU_CALIB_HS_TERM_RANGE_ADJ_MASK; + fuse->hs_squelch_level = + (value >> FUSE_SKU_CALIB_HS_SQUELCH_LEVEL_SHIFT) & + FUSE_SKU_CALIB_HS_SQUELCH_LEVEL_MASK; + + return 0; +} + +static struct tegra_xusb_padctl * +tegra124_xusb_padctl_probe(struct device *dev, + const struct tegra_xusb_padctl_soc *soc) +{ + struct tegra124_xusb_padctl *padctl; + int err; + + padctl = devm_kzalloc(dev, sizeof(*padctl), GFP_KERNEL); + if (!padctl) + return ERR_PTR(-ENOMEM); + + padctl->base.dev = dev; + padctl->base.soc = soc; + + err = tegra124_xusb_read_fuse_calibration(&padctl->fuse); + if (err < 0) + return ERR_PTR(err); + + return &padctl->base; +} + +static void tegra124_xusb_padctl_remove(struct tegra_xusb_padctl *padctl) +{ +} + +static const struct tegra_xusb_padctl_ops tegra124_xusb_padctl_ops = { + .probe = tegra124_xusb_padctl_probe, + .remove = tegra124_xusb_padctl_remove, + .usb3_save_context = tegra124_usb3_save_context, + .hsic_set_idle = tegra124_hsic_set_idle, +}; + +const struct tegra_xusb_padctl_soc tegra124_xusb_padctl_soc = { + .num_pads = ARRAY_SIZE(tegra124_pads), + .pads = tegra124_pads, + .ports = { + .usb2 = { + .ops = &tegra124_usb2_port_ops, + .count = 3, + }, + .ulpi = { + .ops = &tegra124_ulpi_port_ops, + .count = 1, + }, + .hsic = { + .ops = &tegra124_hsic_port_ops, + .count = 2, + }, + .usb3 = { + .ops = &tegra124_usb3_port_ops, + .count = 2, + }, + }, + .ops = &tegra124_xusb_padctl_ops, +}; +EXPORT_SYMBOL_GPL(tegra124_xusb_padctl_soc); + +MODULE_AUTHOR("Thierry Reding <treding(a)nvidia.com>"); +MODULE_DESCRIPTION("NVIDIA Tegra 124 XUSB Pad Controller driver"); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/phy/tegra/xusb.c b/drivers/phy/tegra/xusb.c new file mode 100644 index 000000000000..ec83dfdbc206 --- /dev/null +++ b/drivers/phy/tegra/xusb.c @@ -0,0 +1,1021 @@ +/* + * Copyright (c) 2014-2015, NVIDIA CORPORATION. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#include <linux/delay.h> +#include <linux/io.h> +#include <linux/mailbox_client.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/of_device.h> +#include <linux/phy/phy.h> +#include <linux/platform_device.h> +#include <linux/regulator/consumer.h> +#include <linux/reset.h> +#include <linux/slab.h> +#include <linux/workqueue.h> + +#include <soc/tegra/fuse.h> + +#include "xusb.h" + +static struct phy *tegra_xusb_pad_of_xlate(struct device *dev, + struct of_phandle_args *args) +{ + struct tegra_xusb_pad *pad = dev_get_drvdata(dev); + struct phy *phy = NULL; + unsigned int i; + + if (args->args_count != 0) + return ERR_PTR(-EINVAL); + + for (i = 0; i < pad->soc->num_lanes; i++) { + if (!pad->lanes[i]) + continue; + + if (pad->lanes[i]->dev.of_node == args->np) { + phy = pad->lanes[i]; + break; + } + } + + if (phy == NULL) + phy = ERR_PTR(-ENODEV); + + return phy; +} + +static const struct of_device_id tegra_xusb_padctl_of_match[] = { +#if defined(CONFIG_ARCH_TEGRA_124_SOC) || defined(CONFIG_ARCH_TEGRA_132_SOC) + { + .compatible = "nvidia,tegra124-xusb-padctl", + .data = &tegra124_xusb_padctl_soc, + }, +#endif +#if defined(CONFIG_ARCH_TEGRA_210_SOC) + { + .compatible = "nvidia,tegra210-xusb-padctl", + .data = &tegra210_xusb_padctl_soc, + }, +#endif + { } +}; +MODULE_DEVICE_TABLE(of, tegra_xusb_padctl_of_match); + +static struct device_node * +tegra_xusb_find_pad_node(struct tegra_xusb_padctl *padctl, const char *name) +{ + /* + * of_find_node_by_name() drops a reference, so make sure to grab one. + */ + struct device_node *np = of_node_get(padctl->dev->of_node); + + np = of_find_node_by_name(np, "pads"); + if (np) + np = of_find_node_by_name(np, name); + + return np; +} + +static struct device_node * +tegra_xusb_pad_find_phy_node(struct tegra_xusb_pad *pad, unsigned int index) +{ + /* + * of_find_node_by_name() drops a reference, so make sure to grab one. + */ + struct device_node *np = of_node_get(pad->dev.of_node); + + np = of_find_node_by_name(np, "lanes"); + if (!np) + return NULL; + + return of_find_node_by_name(np, pad->soc->lanes[index].name); +} + +int tegra_xusb_lane_lookup_function(struct tegra_xusb_lane *lane, + const char *function) +{ + unsigned int i; + + for (i = 0; i < lane->soc->num_funcs; i++) + if (strcmp(function, lane->soc->funcs[i]) == 0) + return i; + + return -EINVAL; +} + +int tegra_xusb_lane_parse_dt(struct tegra_xusb_lane *lane, + struct device_node *np) +{ + struct device *dev = &lane->pad->dev; + const char *function; + int err; + + err = of_property_read_string(np, "nvidia,function", &function); + if (err < 0) + return err; + + err = tegra_xusb_lane_lookup_function(lane, function); + if (err < 0) { + dev_err(dev, "invalid function \"%s\" for lane \"%s\"\n", + function, np->name); + return err; + } + + lane->function = err; + + return 0; +} + +static void tegra_xusb_lane_destroy(struct phy *phy) +{ + if (phy) { + struct tegra_xusb_lane *lane = phy_get_drvdata(phy); + + lane->pad->ops->remove(lane); + phy_destroy(phy); + } +} + +static void tegra_xusb_pad_release(struct device *dev) +{ + struct tegra_xusb_pad *pad = to_tegra_xusb_pad(dev); + + pad->soc->ops->remove(pad); +} + +static struct device_type tegra_xusb_pad_type = { + .release = tegra_xusb_pad_release, +}; + +int tegra_xusb_pad_init(struct tegra_xusb_pad *pad, + struct tegra_xusb_padctl *padctl, + struct device_node *np) +{ + int err; + + device_initialize(&pad->dev); + INIT_LIST_HEAD(&pad->list); + pad->dev.parent = padctl->dev; + pad->dev.type = &tegra_xusb_pad_type; + pad->dev.of_node = np; + pad->padctl = padctl; + + err = dev_set_name(&pad->dev, "%s", pad->soc->name); + if (err < 0) + goto unregister; + + err = device_add(&pad->dev); + if (err < 0) + goto unregister; + + return 0; + +unregister: + device_unregister(&pad->dev); + return err; +} + +int tegra_xusb_pad_register(struct tegra_xusb_pad *pad, + const struct phy_ops *ops) +{ + struct device_node *children; + struct phy *lane; + unsigned int i; + int err; + + children = of_find_node_by_name(pad->dev.of_node, "lanes"); + if (!children) + return -ENODEV; + + pad->lanes = devm_kcalloc(&pad->dev, pad->soc->num_lanes, sizeof(lane), + GFP_KERNEL); + if (!pad->lanes) { + of_node_put(children); + return -ENOMEM; + } + + for (i = 0; i < pad->soc->num_lanes; i++) { + struct device_node *np = tegra_xusb_pad_find_phy_node(pad, i); + struct tegra_xusb_lane *lane; + + /* skip disabled lanes */ + if (!np || !of_device_is_available(np)) { + of_node_put(np); + continue; + } + + pad->lanes[i] = phy_create(&pad->dev, np, ops); + if (IS_ERR(pad->lanes[i])) { + err = PTR_ERR(pad->lanes[i]); + of_node_put(np); + goto remove; + } + + lane = pad->ops->probe(pad, np, i); + if (IS_ERR(lane)) { + phy_destroy(pad->lanes[i]); + err = PTR_ERR(lane); + goto remove; + } + + list_add_tail(&lane->list, &pad->padctl->lanes); + phy_set_drvdata(pad->lanes[i], lane); + } + + pad->provider = of_phy_provider_register_full(&pad->dev, children, + tegra_xusb_pad_of_xlate); + if (IS_ERR(pad->provider)) { + err = PTR_ERR(pad->provider); + goto remove; + } + + return 0; + +remove: + while (i--) + tegra_xusb_lane_destroy(pad->lanes[i]); + + of_node_put(children); + + return err; +} + +void tegra_xusb_pad_unregister(struct tegra_xusb_pad *pad) +{ + unsigned int i = pad->soc->num_lanes; + + of_phy_provider_unregister(pad->provider); + + while (i--) + tegra_xusb_lane_destroy(pad->lanes[i]); + + device_unregister(&pad->dev); +} + +static struct tegra_xusb_pad * +tegra_xusb_pad_create(struct tegra_xusb_padctl *padctl, + const struct tegra_xusb_pad_soc *soc) +{ + struct tegra_xusb_pad *pad; + struct device_node *np; + int err; + + np = tegra_xusb_find_pad_node(padctl, soc->name); + if (!np || !of_device_is_available(np)) + return NULL; + + pad = soc->ops->probe(padctl, soc, np); + if (IS_ERR(pad)) { + err = PTR_ERR(pad); + dev_err(padctl->dev, "failed to create pad %s: %d\n", + soc->name, err); + return ERR_PTR(err); + } + + /* XXX move this into ->probe() to avoid string comparison */ + if (strcmp(soc->name, "pcie") == 0) + padctl->pcie = pad; + + if (strcmp(soc->name, "sata") == 0) + padctl->sata = pad; + + if (strcmp(soc->name, "usb2") == 0) + padctl->usb2 = pad; + + if (strcmp(soc->name, "ulpi") == 0) + padctl->ulpi = pad; + + if (strcmp(soc->name, "hsic") == 0) + padctl->hsic = pad; + + return pad; +} + +static void __tegra_xusb_remove_pads(struct tegra_xusb_padctl *padctl) +{ + struct tegra_xusb_pad *pad, *tmp; + + list_for_each_entry_safe_reverse(pad, tmp, &padctl->pads, list) { + list_del(&pad->list); + tegra_xusb_pad_unregister(pad); + } +} + +static void tegra_xusb_remove_pads(struct tegra_xusb_padctl *padctl) +{ + mutex_lock(&padctl->lock); + __tegra_xusb_remove_pads(padctl); + mutex_unlock(&padctl->lock); +} + +static void tegra_xusb_lane_program(struct tegra_xusb_lane *lane) +{ + struct tegra_xusb_padctl *padctl = lane->pad->padctl; + const struct tegra_xusb_lane_soc *soc = lane->soc; + u32 value; + + /* choose function */ + value = padctl_readl(padctl, soc->offset); + value &= ~(soc->mask << soc->shift); + value |= lane->function << soc->shift; + padctl_writel(padctl, value, soc->offset); +} + +static void tegra_xusb_pad_program(struct tegra_xusb_pad *pad) +{ + unsigned int i; + + for (i = 0; i < pad->soc->num_lanes; i++) { + struct tegra_xusb_lane *lane; + + if (pad->lanes[i]) { + lane = phy_get_drvdata(pad->lanes[i]); + tegra_xusb_lane_program(lane); + } + } +} + +static int tegra_xusb_setup_pads(struct tegra_xusb_padctl *padctl) +{ + struct tegra_xusb_pad *pad; + unsigned int i; + + mutex_lock(&padctl->lock); + + for (i = 0; i < padctl->soc->num_pads; i++) { + const struct tegra_xusb_pad_soc *soc = padctl->soc->pads[i]; + int err; + + pad = tegra_xusb_pad_create(padctl, soc); + if (IS_ERR(pad)) { + err = PTR_ERR(pad); + dev_err(padctl->dev, "failed to create pad %s: %d\n", + soc->name, err); + __tegra_xusb_remove_pads(padctl); + mutex_unlock(&padctl->lock); + return err; + } + + if (!pad) + continue; + + list_add_tail(&pad->list, &padctl->pads); + } + + list_for_each_entry(pad, &padctl->pads, list) + tegra_xusb_pad_program(pad); + + mutex_unlock(&padctl->lock); + return 0; +} + +static bool tegra_xusb_lane_check(struct tegra_xusb_lane *lane, + const char *function) +{ + const char *func = lane->soc->funcs[lane->function]; + + return strcmp(function, func) == 0; +} + +struct tegra_xusb_lane *tegra_xusb_find_lane(struct tegra_xusb_padctl *padctl, + const char *type, + unsigned int index) +{ + struct tegra_xusb_lane *lane, *hit = ERR_PTR(-ENODEV); + char *name; + + name = kasprintf(GFP_KERNEL, "%s-%u", type, index); + if (!name) + return ERR_PTR(-ENOMEM); + + list_for_each_entry(lane, &padctl->lanes, list) { + if (strcmp(lane->soc->name, name) == 0) { + hit = lane; + break; + } + } + + kfree(name); + return hit; +} + +struct tegra_xusb_lane * +tegra_xusb_port_find_lane(struct tegra_xusb_port *port, + const struct tegra_xusb_lane_map *map, + const char *function) +{ + struct tegra_xusb_lane *lane, *match = ERR_PTR(-ENODEV); + + for (map = map; map->type; map++) { + if (port->index != map->port) + continue; + + lane = tegra_xusb_find_lane(port->padctl, map->type, + map->index); + if (IS_ERR(lane)) + continue; + + if (!tegra_xusb_lane_check(lane, function)) + continue; + + if (!IS_ERR(match)) + dev_err(&port->dev, "conflicting match: %s-%u / %s\n", + map->type, map->index, match->soc->name); + else + match = lane; + } + + return match; +} + +static struct device_node * +tegra_xusb_find_port_node(struct tegra_xusb_padctl *padctl, const char *type, + unsigned int index) +{ + /* + * of_find_node_by_name() drops a reference, so make sure to grab one. + */ + struct device_node *np = of_node_get(padctl->dev->of_node); + + np = of_find_node_by_name(np, "ports"); + if (np) { + char *name; + + name = kasprintf(GFP_KERNEL, "%s-%u", type, index); + np = of_find_node_by_name(np, name); + kfree(name); + } + + return np; +} + +struct tegra_xusb_port * +tegra_xusb_find_port(struct tegra_xusb_padctl *padctl, const char *type, + unsigned int index) +{ + struct tegra_xusb_port *port; + struct device_node *np; + + np = tegra_xusb_find_port_node(padctl, type, index); + if (!np) + return NULL; + + list_for_each_entry(port, &padctl->ports, list) { + if (np == port->dev.of_node) { + of_node_put(np); + return port; + } + } + + of_node_put(np); + + return NULL; +} + +struct tegra_xusb_usb2_port * +tegra_xusb_find_usb2_port(struct tegra_xusb_padctl *padctl, unsigned int index) +{ + struct tegra_xusb_port *port; + + port = tegra_xusb_find_port(padctl, "usb2", index); + if (port) + return to_usb2_port(port); + + return NULL; +} + +struct tegra_xusb_usb3_port * +tegra_xusb_find_usb3_port(struct tegra_xusb_padctl *padctl, unsigned int index) +{ + struct tegra_xusb_port *port; + + port = tegra_xusb_find_port(padctl, "usb3", index); + if (port) + return to_usb3_port(port); + + return NULL; +} + +static void tegra_xusb_port_release(struct device *dev) +{ +} + +static struct device_type tegra_xusb_port_type = { + .release = tegra_xusb_port_release, +}; + +static int tegra_xusb_port_init(struct tegra_xusb_port *port, + struct tegra_xusb_padctl *padctl, + struct device_node *np, + const char *name, + unsigned int index) +{ + int err; + + INIT_LIST_HEAD(&port->list); + port->padctl = padctl; + port->index = index; + + device_initialize(&port->dev); + port->dev.type = &tegra_xusb_port_type; + port->dev.of_node = of_node_get(np); + port->dev.parent = padctl->dev; + + err = dev_set_name(&port->dev, "%s-%u", name, index); + if (err < 0) + goto unregister; + + err = device_add(&port->dev); + if (err < 0) + goto unregister; + + return 0; + +unregister: + device_unregister(&port->dev); + return err; +} + +static void tegra_xusb_port_unregister(struct tegra_xusb_port *port) +{ + device_unregister(&port->dev); +} + +static int tegra_xusb_usb2_port_parse_dt(struct tegra_xusb_usb2_port *usb2) +{ + struct tegra_xusb_port *port = &usb2->base; + struct device_node *np = port->dev.of_node; + + usb2->internal = of_property_read_bool(np, "nvidia,internal"); + + usb2->supply = devm_regulator_get(&port->dev, "vbus"); + if (IS_ERR(usb2->supply)) + return PTR_ERR(usb2->supply); + + return 0; +} + +static int tegra_xusb_add_usb2_port(struct tegra_xusb_padctl *padctl, + unsigned int index) +{ + struct tegra_xusb_usb2_port *usb2; + struct device_node *np; + int err = 0; + + /* + * USB2 ports don't require additional properties, but if the port is + * marked as disabled there is no reason to register it. + */ + np = tegra_xusb_find_port_node(padctl, "usb2", index); + if (!np || !of_device_is_available(np)) + goto out; + + usb2 = devm_kzalloc(padctl->dev, sizeof(*usb2), GFP_KERNEL); + if (!usb2) { + err = -ENOMEM; + goto out; + } + + err = tegra_xusb_port_init(&usb2->base, padctl, np, "usb2", index); + if (err < 0) + goto out; + + usb2->base.ops = padctl->soc->ports.usb2.ops; + + usb2->base.lane = usb2->base.ops->map(&usb2->base); + if (IS_ERR(usb2->base.lane)) { + err = PTR_ERR(usb2->base.lane); + goto out; + } + + err = tegra_xusb_usb2_port_parse_dt(usb2); + if (err < 0) { + tegra_xusb_port_unregister(&usb2->base); + goto out; + } + + list_add_tail(&usb2->base.list, &padctl->ports); + +out: + of_node_put(np); + return err; +} + +static int tegra_xusb_ulpi_port_parse_dt(struct tegra_xusb_ulpi_port *ulpi) +{ + struct tegra_xusb_port *port = &ulpi->base; + struct device_node *np = port->dev.of_node; + + ulpi->internal = of_property_read_bool(np, "nvidia,internal"); + + return 0; +} + +static int tegra_xusb_add_ulpi_port(struct tegra_xusb_padctl *padctl, + unsigned int index) +{ + struct tegra_xusb_ulpi_port *ulpi; + struct device_node *np; + int err = 0; + + np = tegra_xusb_find_port_node(padctl, "ulpi", index); + if (!np || !of_device_is_available(np)) + goto out; + + ulpi = devm_kzalloc(padctl->dev, sizeof(*ulpi), GFP_KERNEL); + if (!ulpi) { + err = -ENOMEM; + goto out; + } + + err = tegra_xusb_port_init(&ulpi->base, padctl, np, "ulpi", index); + if (err < 0) + goto out; + + ulpi->base.ops = padctl->soc->ports.ulpi.ops; + + ulpi->base.lane = ulpi->base.ops->map(&ulpi->base); + if (IS_ERR(ulpi->base.lane)) { + err = PTR_ERR(ulpi->base.lane); + goto out; + } + + err = tegra_xusb_ulpi_port_parse_dt(ulpi); + if (err < 0) { + tegra_xusb_port_unregister(&ulpi->base); + goto out; + } + + list_add_tail(&ulpi->base.list, &padctl->ports); + +out: + of_node_put(np); + return err; +} + +static int tegra_xusb_hsic_port_parse_dt(struct tegra_xusb_hsic_port *hsic) +{ + /* XXX */ + return 0; +} + +static int tegra_xusb_add_hsic_port(struct tegra_xusb_padctl *padctl, + unsigned int index) +{ + struct tegra_xusb_hsic_port *hsic; + struct device_node *np; + int err = 0; + + np = tegra_xusb_find_port_node(padctl, "hsic", index); + if (!np || !of_device_is_available(np)) + goto out; + + hsic = devm_kzalloc(padctl->dev, sizeof(*hsic), GFP_KERNEL); + if (!hsic) { + err = -ENOMEM; + goto out; + } + + err = tegra_xusb_port_init(&hsic->base, padctl, np, "hsic", index); + if (err < 0) + goto out; + + hsic->base.ops = padctl->soc->ports.hsic.ops; + + hsic->base.lane = hsic->base.ops->map(&hsic->base); + if (IS_ERR(hsic->base.lane)) { + err = PTR_ERR(hsic->base.lane); + goto out; + } + + err = tegra_xusb_hsic_port_parse_dt(hsic); + if (err < 0) { + tegra_xusb_port_unregister(&hsic->base); + goto out; + } + + list_add_tail(&hsic->base.list, &padctl->ports); + +out: + of_node_put(np); + return err; +} + +static int tegra_xusb_usb3_port_parse_dt(struct tegra_xusb_usb3_port *usb3) +{ + struct tegra_xusb_port *port = &usb3->base; + struct device_node *np = port->dev.of_node; + u32 value; + int err; + + err = of_property_read_u32(np, "nvidia,usb2-companion", &value); + if (err < 0) { + dev_err(&port->dev, "failed to read port: %d\n", err); + return err; + } + + usb3->port = value; + + usb3->internal = of_property_read_bool(np, "nvidia,internal"); + + usb3->supply = devm_regulator_get(&port->dev, "vbus"); + if (IS_ERR(usb3->supply)) + return PTR_ERR(usb3->supply); + + return 0; +} + +static int tegra_xusb_add_usb3_port(struct tegra_xusb_padctl *padctl, + unsigned int index) +{ + struct tegra_xusb_usb3_port *usb3; + struct device_node *np; + int err = 0; + + /* + * If there is no supplemental configuration in the device tree the + * port is unusable. But it is valid to configure only a single port, + * hence return 0 instead of an error to allow ports to be optional. + */ + np = tegra_xusb_find_port_node(padctl, "usb3", index); + if (!np || !of_device_is_available(np)) + goto out; + + usb3 = devm_kzalloc(padctl->dev, sizeof(*usb3), GFP_KERNEL); + if (!usb3) { + err = -ENOMEM; + goto out; + } + + err = tegra_xusb_port_init(&usb3->base, padctl, np, "usb3", index); + if (err < 0) + goto out; + + usb3->base.ops = padctl->soc->ports.usb3.ops; + + usb3->base.lane = usb3->base.ops->map(&usb3->base); + if (IS_ERR(usb3->base.lane)) { + err = PTR_ERR(usb3->base.lane); + goto out; + } + + err = tegra_xusb_usb3_port_parse_dt(usb3); + if (err < 0) { + tegra_xusb_port_unregister(&usb3->base); + goto out; + } + + list_add_tail(&usb3->base.list, &padctl->ports); + +out: + of_node_put(np); + return err; +} + +static void __tegra_xusb_remove_ports(struct tegra_xusb_padctl *padctl) +{ + struct tegra_xusb_port *port, *tmp; + + list_for_each_entry_safe_reverse(port, tmp, &padctl->ports, list) { + list_del(&port->list); + tegra_xusb_port_unregister(port); + } +} + +static int tegra_xusb_setup_ports(struct tegra_xusb_padctl *padctl) +{ + struct tegra_xusb_port *port; + unsigned int i; + int err = 0; + + mutex_lock(&padctl->lock); + + for (i = 0; i < padctl->soc->ports.usb2.count; i++) { + err = tegra_xusb_add_usb2_port(padctl, i); + if (err < 0) + goto remove_ports; + } + + for (i = 0; i < padctl->soc->ports.ulpi.count; i++) { + err = tegra_xusb_add_ulpi_port(padctl, i); + if (err < 0) + goto remove_ports; + } + + for (i = 0; i < padctl->soc->ports.hsic.count; i++) { + err = tegra_xusb_add_hsic_port(padctl, i); + if (err < 0) + goto remove_ports; + } + + for (i = 0; i < padctl->soc->ports.usb3.count; i++) { + err = tegra_xusb_add_usb3_port(padctl, i); + if (err < 0) + goto remove_ports; + } + + list_for_each_entry(port, &padctl->ports, list) { + err = port->ops->enable(port); + if (err < 0) + dev_err(padctl->dev, "failed to enable port %s: %d\n", + dev_name(&port->dev), err); + } + + goto unlock; + +remove_ports: + __tegra_xusb_remove_ports(padctl); +unlock: + mutex_unlock(&padctl->lock); + return err; +} + +static void tegra_xusb_remove_ports(struct tegra_xusb_padctl *padctl) +{ + mutex_lock(&padctl->lock); + __tegra_xusb_remove_ports(padctl); + mutex_unlock(&padctl->lock); +} + +static int tegra_xusb_padctl_probe(struct platform_device *pdev) +{ + struct device_node *np = of_node_get(pdev->dev.of_node); + const struct tegra_xusb_padctl_soc *soc; + struct tegra_xusb_padctl *padctl; + const struct of_device_id *match; + struct resource *res; + int err; + + /* for backwards compatibility with old device trees */ + np = of_find_node_by_name(np, "pads"); + if (!np) { + dev_warn(&pdev->dev, "deprecated DT, using legacy driver\n"); + return tegra_xusb_padctl_legacy_probe(pdev); + } + + of_node_put(np); + + match = of_match_node(tegra_xusb_padctl_of_match, pdev->dev.of_node); + soc = match->data; + + padctl = soc->ops->probe(&pdev->dev, soc); + if (IS_ERR(padctl)) + return PTR_ERR(padctl); + + platform_set_drvdata(pdev, padctl); + INIT_LIST_HEAD(&padctl->ports); + INIT_LIST_HEAD(&padctl->lanes); + INIT_LIST_HEAD(&padctl->pads); + mutex_init(&padctl->lock); + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + padctl->regs = devm_ioremap_resource(&pdev->dev, res); + if (IS_ERR(padctl->regs)) { + err = PTR_ERR(padctl->regs); + goto remove; + } + + padctl->rst = devm_reset_control_get(&pdev->dev, NULL); + if (IS_ERR(padctl->rst)) { + err = PTR_ERR(padctl->rst); + goto remove; + } + + err = reset_control_deassert(padctl->rst); + if (err < 0) + goto remove; + + err = tegra_xusb_setup_pads(padctl); + if (err < 0) { + dev_err(&pdev->dev, "failed to setup pads: %d\n", err); + goto reset; + } + + err = tegra_xusb_setup_ports(padctl); + if (err) { + dev_err(&pdev->dev, "failed to setup XUSB ports: %d\n", err); + goto remove_pads; + } + + return 0; + +remove_pads: + tegra_xusb_remove_pads(padctl); +reset: + reset_control_assert(padctl->rst); +remove: + soc->ops->remove(padctl); + return err; +} + +static int tegra_xusb_padctl_remove(struct platform_device *pdev) +{ + struct tegra_xusb_padctl *padctl = platform_get_drvdata(pdev); + int err; + + tegra_xusb_remove_ports(padctl); + tegra_xusb_remove_pads(padctl); + + err = reset_control_assert(padctl->rst); + if (err < 0) + dev_err(&pdev->dev, "failed to assert reset: %d\n", err); + + padctl->soc->ops->remove(padctl); + + return err; +} + +static struct platform_driver tegra_xusb_padctl_driver = { + .driver = { + .name = "tegra-xusb-padctl", + .of_match_table = tegra_xusb_padctl_of_match, + }, + .probe = tegra_xusb_padctl_probe, + .remove = tegra_xusb_padctl_remove, +}; +module_platform_driver(tegra_xusb_padctl_driver); + +struct tegra_xusb_padctl *tegra_xusb_padctl_get(struct device *dev) +{ + struct tegra_xusb_padctl *padctl; + struct platform_device *pdev; + struct device_node *np; + + np = of_parse_phandle(dev->of_node, "nvidia,xusb-padctl", 0); + if (!np) + return ERR_PTR(-EINVAL); + + /* + * This is slightly ugly. A better implementation would be to keep a + * registry of pad controllers, but since there will almost certainly + * only ever be one per SoC that would be a little overkill. + */ + pdev = of_find_device_by_node(np); + if (!pdev) { + of_node_put(np); + return ERR_PTR(-ENODEV); + } + + of_node_put(np); + + padctl = platform_get_drvdata(pdev); + if (!padctl) { + put_device(&pdev->dev); + return ERR_PTR(-EPROBE_DEFER); + } + + return padctl; +} +EXPORT_SYMBOL_GPL(tegra_xusb_padctl_get); + +void tegra_xusb_padctl_put(struct tegra_xusb_padctl *padctl) +{ + if (padctl) + put_device(padctl->dev); +} +EXPORT_SYMBOL_GPL(tegra_xusb_padctl_put); + +int tegra_xusb_padctl_usb3_save_context(struct tegra_xusb_padctl *padctl, + unsigned int port) +{ + if (padctl->soc->ops->usb3_save_context) + return padctl->soc->ops->usb3_save_context(padctl, port); + + return -ENOSYS; +} +EXPORT_SYMBOL_GPL(tegra_xusb_padctl_usb3_save_context); + +int tegra_xusb_padctl_hsic_set_idle(struct tegra_xusb_padctl *padctl, + unsigned int port, bool idle) +{ + if (padctl->soc->ops->hsic_set_idle) + return padctl->soc->ops->hsic_set_idle(padctl, port, idle); + + return -ENOSYS; +} +EXPORT_SYMBOL_GPL(tegra_xusb_padctl_hsic_set_idle); + +int tegra_xusb_padctl_usb3_set_lfps_detect(struct tegra_xusb_padctl *padctl, + unsigned int port, bool enable) +{ + if (padctl->soc->ops->usb3_set_lfps_detect) + return padctl->soc->ops->usb3_set_lfps_detect(padctl, port, + enable); + + return -ENOSYS; +} +EXPORT_SYMBOL_GPL(tegra_xusb_padctl_usb3_set_lfps_detect); + +MODULE_AUTHOR("Thierry Reding <treding(a)nvidia.com>"); +MODULE_DESCRIPTION("Tegra XUSB Pad Controller driver"); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/phy/tegra/xusb.h b/drivers/phy/tegra/xusb.h new file mode 100644 index 000000000000..b49dbc36efa3 --- /dev/null +++ b/drivers/phy/tegra/xusb.h @@ -0,0 +1,421 @@ +/* + * Copyright (c) 2014-2015, NVIDIA CORPORATION. All rights reserved. + * Copyright (c) 2015, Google Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#ifndef __PHY_TEGRA_XUSB_H +#define __PHY_TEGRA_XUSB_H + +#include <linux/io.h> +#include <linux/mutex.h> +#include <linux/workqueue.h> + +/* legacy entry points for backwards-compatibility */ +int tegra_xusb_padctl_legacy_probe(struct platform_device *pdev); +int tegra_xusb_padctl_legacy_remove(struct platform_device *pdev); + +struct phy; +struct phy_provider; +struct platform_device; +struct regulator; + +/* + * lanes + */ +struct tegra_xusb_lane_soc { + const char *name; + + unsigned int offset; + unsigned int shift; + unsigned int mask; + + const char * const *funcs; + unsigned int num_funcs; +}; + +struct tegra_xusb_lane { + const struct tegra_xusb_lane_soc *soc; + struct tegra_xusb_pad *pad; + struct device_node *np; + struct list_head list; + unsigned int function; + unsigned int index; +}; + +int tegra_xusb_lane_parse_dt(struct tegra_xusb_lane *lane, + struct device_node *np); + +struct tegra_xusb_usb2_lane { + struct tegra_xusb_lane base; + + u32 hs_curr_level_offset; +}; + +static inline struct tegra_xusb_usb2_lane * +to_usb2_lane(struct tegra_xusb_lane *lane) +{ + return container_of(lane, struct tegra_xusb_usb2_lane, base); +} + +struct tegra_xusb_ulpi_lane { + struct tegra_xusb_lane base; +}; + +static inline struct tegra_xusb_ulpi_lane * +to_ulpi_lane(struct tegra_xusb_lane *lane) +{ + return container_of(lane, struct tegra_xusb_ulpi_lane, base); +} + +struct tegra_xusb_hsic_lane { + struct tegra_xusb_lane base; + + u32 strobe_trim; + u32 rx_strobe_trim; + u32 rx_data_trim; + u32 tx_rtune_n; + u32 tx_rtune_p; + u32 tx_rslew_n; + u32 tx_rslew_p; + bool auto_term; +}; + +static inline struct tegra_xusb_hsic_lane * +to_hsic_lane(struct tegra_xusb_lane *lane) +{ + return container_of(lane, struct tegra_xusb_hsic_lane, base); +} + +struct tegra_xusb_pcie_lane { + struct tegra_xusb_lane base; +}; + +static inline struct tegra_xusb_pcie_lane * +to_pcie_lane(struct tegra_xusb_lane *lane) +{ + return container_of(lane, struct tegra_xusb_pcie_lane, base); +} + +struct tegra_xusb_sata_lane { + struct tegra_xusb_lane base; +}; + +static inline struct tegra_xusb_sata_lane * +to_sata_lane(struct tegra_xusb_lane *lane) +{ + return container_of(lane, struct tegra_xusb_sata_lane, base); +} + +struct tegra_xusb_lane_ops { + struct tegra_xusb_lane *(*probe)(struct tegra_xusb_pad *pad, + struct device_node *np, + unsigned int index); + void (*remove)(struct tegra_xusb_lane *lane); +}; + +/* + * pads + */ +struct tegra_xusb_pad_soc; +struct tegra_xusb_padctl; + +struct tegra_xusb_pad_ops { + struct tegra_xusb_pad *(*probe)(struct tegra_xusb_padctl *padctl, + const struct tegra_xusb_pad_soc *soc, + struct device_node *np); + void (*remove)(struct tegra_xusb_pad *pad); +}; + +struct tegra_xusb_pad_soc { + const char *name; + + const struct tegra_xusb_lane_soc *lanes; + unsigned int num_lanes; + + const struct tegra_xusb_pad_ops *ops; +}; + +struct tegra_xusb_pad { + const struct tegra_xusb_pad_soc *soc; + struct tegra_xusb_padctl *padctl; + struct phy_provider *provider; + struct phy **lanes; + struct device dev; + + const struct tegra_xusb_lane_ops *ops; + + struct list_head list; +}; + +static inline struct tegra_xusb_pad *to_tegra_xusb_pad(struct device *dev) +{ + return container_of(dev, struct tegra_xusb_pad, dev); +} + +int tegra_xusb_pad_init(struct tegra_xusb_pad *pad, + struct tegra_xusb_padctl *padctl, + struct device_node *np); +int tegra_xusb_pad_register(struct tegra_xusb_pad *pad, + const struct phy_ops *ops); +void tegra_xusb_pad_unregister(struct tegra_xusb_pad *pad); + +struct tegra_xusb_usb2_pad { + struct tegra_xusb_pad base; + + struct clk *clk; + unsigned int enable; + struct mutex lock; +}; + +static inline struct tegra_xusb_usb2_pad * +to_usb2_pad(struct tegra_xusb_pad *pad) +{ + return container_of(pad, struct tegra_xusb_usb2_pad, base); +} + +struct tegra_xusb_ulpi_pad { + struct tegra_xusb_pad base; +}; + +static inline struct tegra_xusb_ulpi_pad * +to_ulpi_pad(struct tegra_xusb_pad *pad) +{ + return container_of(pad, struct tegra_xusb_ulpi_pad, base); +} + +struct tegra_xusb_hsic_pad { + struct tegra_xusb_pad base; + + struct regulator *supply; + struct clk *clk; +}; + +static inline struct tegra_xusb_hsic_pad * +to_hsic_pad(struct tegra_xusb_pad *pad) +{ + return container_of(pad, struct tegra_xusb_hsic_pad, base); +} + +struct tegra_xusb_pcie_pad { + struct tegra_xusb_pad base; + + struct reset_control *rst; + struct clk *pll; + + unsigned int enable; +}; + +static inline struct tegra_xusb_pcie_pad * +to_pcie_pad(struct tegra_xusb_pad *pad) +{ + return container_of(pad, struct tegra_xusb_pcie_pad, base); +} + +struct tegra_xusb_sata_pad { + struct tegra_xusb_pad base; + + struct reset_control *rst; + struct clk *pll; + + unsigned int enable; +}; + +static inline struct tegra_xusb_sata_pad * +to_sata_pad(struct tegra_xusb_pad *pad) +{ + return container_of(pad, struct tegra_xusb_sata_pad, base); +} + +/* + * ports + */ +struct tegra_xusb_port_ops; + +struct tegra_xusb_port { + struct tegra_xusb_padctl *padctl; + struct tegra_xusb_lane *lane; + unsigned int index; + + struct list_head list; + struct device dev; + + const struct tegra_xusb_port_ops *ops; +}; + +struct tegra_xusb_lane_map { + unsigned int port; + const char *type; + unsigned int index; + const char *func; +}; + +struct tegra_xusb_lane * +tegra_xusb_port_find_lane(struct tegra_xusb_port *port, + const struct tegra_xusb_lane_map *map, + const char *function); + +struct tegra_xusb_port * +tegra_xusb_find_port(struct tegra_xusb_padctl *padctl, const char *type, + unsigned int index); + +struct tegra_xusb_usb2_port { + struct tegra_xusb_port base; + + struct regulator *supply; + bool internal; +}; + +static inline struct tegra_xusb_usb2_port * +to_usb2_port(struct tegra_xusb_port *port) +{ + return container_of(port, struct tegra_xusb_usb2_port, base); +} + +struct tegra_xusb_usb2_port * +tegra_xusb_find_usb2_port(struct tegra_xusb_padctl *padctl, + unsigned int index); + +struct tegra_xusb_ulpi_port { + struct tegra_xusb_port base; + + struct regulator *supply; + bool internal; +}; + +static inline struct tegra_xusb_ulpi_port * +to_ulpi_port(struct tegra_xusb_port *port) +{ + return container_of(port, struct tegra_xusb_ulpi_port, base); +} + +struct tegra_xusb_hsic_port { + struct tegra_xusb_port base; +}; + +static inline struct tegra_xusb_hsic_port * +to_hsic_port(struct tegra_xusb_port *port) +{ + return container_of(port, struct tegra_xusb_hsic_port, base); +} + +struct tegra_xusb_usb3_port { + struct tegra_xusb_port base; + struct regulator *supply; + bool context_saved; + unsigned int port; + bool internal; + + u32 tap1; + u32 amp; + u32 ctle_z; + u32 ctle_g; +}; + +static inline struct tegra_xusb_usb3_port * +to_usb3_port(struct tegra_xusb_port *port) +{ + return container_of(port, struct tegra_xusb_usb3_port, base); +} + +struct tegra_xusb_usb3_port * +tegra_xusb_find_usb3_port(struct tegra_xusb_padctl *padctl, + unsigned int index); + +struct tegra_xusb_port_ops { + int (*enable)(struct tegra_xusb_port *port); + void (*disable)(struct tegra_xusb_port *port); + struct tegra_xusb_lane *(*map)(struct tegra_xusb_port *port); +}; + +/* + * pad controller + */ +struct tegra_xusb_padctl_soc; + +struct tegra_xusb_padctl_ops { + struct tegra_xusb_padctl * + (*probe)(struct device *dev, + const struct tegra_xusb_padctl_soc *soc); + void (*remove)(struct tegra_xusb_padctl *padctl); + + int (*usb3_save_context)(struct tegra_xusb_padctl *padctl, + unsigned int index); + int (*hsic_set_idle)(struct tegra_xusb_padctl *padctl, + unsigned int index, bool idle); + int (*usb3_set_lfps_detect)(struct tegra_xusb_padctl *padctl, + unsigned int index, bool enable); +}; + +struct tegra_xusb_padctl_soc { + const struct tegra_xusb_pad_soc * const *pads; + unsigned int num_pads; + + struct { + struct { + const struct tegra_xusb_port_ops *ops; + unsigned int count; + } usb2, ulpi, hsic, usb3; + } ports; + + const struct tegra_xusb_padctl_ops *ops; +}; + +struct tegra_xusb_padctl { + struct device *dev; + void __iomem *regs; + struct mutex lock; + struct reset_control *rst; + + const struct tegra_xusb_padctl_soc *soc; + + struct tegra_xusb_pad *pcie; + struct tegra_xusb_pad *sata; + struct tegra_xusb_pad *ulpi; + struct tegra_xusb_pad *usb2; + struct tegra_xusb_pad *hsic; + + struct list_head ports; + struct list_head lanes; + struct list_head pads; + + unsigned int enable; + + struct clk *clk; +}; + +static inline void padctl_writel(struct tegra_xusb_padctl *padctl, u32 value, + unsigned long offset) +{ + dev_dbg(padctl->dev, "%08lx < %08x\n", offset, value); + writel(value, padctl->regs + offset); +} + +static inline u32 padctl_readl(struct tegra_xusb_padctl *padctl, + unsigned long offset) +{ + u32 value = readl(padctl->regs + offset); + dev_dbg(padctl->dev, "%08lx > %08x\n", offset, value); + return value; +} + +struct tegra_xusb_lane *tegra_xusb_find_lane(struct tegra_xusb_padctl *padctl, + const char *name, + unsigned int index); + +#if defined(CONFIG_ARCH_TEGRA_124_SOC) || defined(CONFIG_ARCH_TEGRA_132_SOC) +extern const struct tegra_xusb_padctl_soc tegra124_xusb_padctl_soc; +#endif +#if defined(CONFIG_ARCH_TEGRA_210_SOC) +extern const struct tegra_xusb_padctl_soc tegra210_xusb_padctl_soc; +#endif + +#endif /* __PHY_TEGRA_XUSB_H */ diff --git a/drivers/pinctrl/tegra/pinctrl-tegra-xusb.c b/drivers/pinctrl/tegra/pinctrl-tegra-xusb.c index 2f06029c9405..946cda3fee35 100644 --- a/drivers/pinctrl/tegra/pinctrl-tegra-xusb.c +++ b/drivers/pinctrl/tegra/pinctrl-tegra-xusb.c @@ -873,7 +873,7 @@ static const struct of_device_id tegra_xusb_padctl_of_match[] = { }; MODULE_DEVICE_TABLE(of, tegra_xusb_padctl_of_match); -static int tegra_xusb_padctl_probe(struct platform_device *pdev) +int tegra_xusb_padctl_legacy_probe(struct platform_device *pdev) { struct tegra_xusb_padctl *padctl; const struct of_device_id *match; @@ -955,8 +955,9 @@ static int tegra_xusb_padctl_probe(struct platform_device *pdev) reset_control_assert(padctl->rst); return err; } +EXPORT_SYMBOL_GPL(tegra_xusb_padctl_legacy_probe); -static int tegra_xusb_padctl_remove(struct platform_device *pdev) +int tegra_xusb_padctl_legacy_remove(struct platform_device *pdev) { struct tegra_xusb_padctl *padctl = platform_get_drvdata(pdev); int err; @@ -969,17 +970,4 @@ static int tegra_xusb_padctl_remove(struct platform_device *pdev) return err; } - -static struct platform_driver tegra_xusb_padctl_driver = { - .driver = { - .name = "tegra-xusb-padctl", - .of_match_table = tegra_xusb_padctl_of_match, - }, - .probe = tegra_xusb_padctl_probe, - .remove = tegra_xusb_padctl_remove, -}; -module_platform_driver(tegra_xusb_padctl_driver); - -MODULE_AUTHOR("Thierry Reding <treding(a)nvidia.com>"); -MODULE_DESCRIPTION("Tegra 124 XUSB Pad Control driver"); -MODULE_LICENSE("GPL v2"); +EXPORT_SYMBOL_GPL(tegra_xusb_padctl_legacy_remove); diff --git a/include/linux/phy/tegra/xusb.h b/include/linux/phy/tegra/xusb.h new file mode 100644 index 000000000000..8e1a57a78d9f --- /dev/null +++ b/include/linux/phy/tegra/xusb.h @@ -0,0 +1,30 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#ifndef PHY_TEGRA_XUSB_H +#define PHY_TEGRA_XUSB_H + +struct tegra_xusb_padctl; +struct device; + +struct tegra_xusb_padctl *tegra_xusb_padctl_get(struct device *dev); +void tegra_xusb_padctl_put(struct tegra_xusb_padctl *padctl); + +int tegra_xusb_padctl_usb3_save_context(struct tegra_xusb_padctl *padctl, + unsigned int port); +int tegra_xusb_padctl_hsic_set_idle(struct tegra_xusb_padctl *padctl, + unsigned int port, bool idle); +int tegra_xusb_padctl_usb3_set_lfps_detect(struct tegra_xusb_padctl *padctl, + unsigned int port, bool enable); + +#endif /* PHY_TEGRA_XUSB_H */

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH 23/26] stating: ccree: fix allocation of void sized buf

by Gilad Ben-Yossef

We were allocating buffers using sizeof(*struct->field) where field was type void. Fix it by having a local variable with the real type. Cc: stable(a)vger.kernel.org Signed-off-by: Gilad Ben-Yossef <gilad(a)benyossef.com> --- drivers/staging/ccree/cc_debugfs.c | 3 ++- drivers/staging/ccree/ssi_ivgen.c | 9 ++++----- drivers/staging/ccree/ssi_sram_mgr.c | 9 ++++++--- 3 files changed, 12 insertions(+), 9 deletions(-) diff --git a/drivers/staging/ccree/cc_debugfs.c b/drivers/staging/ccree/cc_debugfs.c index 62bf011..140c04b 100644 --- a/drivers/staging/ccree/cc_debugfs.c +++ b/drivers/staging/ccree/cc_debugfs.c @@ -6,6 +6,7 @@ #include <linux/stringify.h> #include "ssi_driver.h" #include "cc_crypto_ctx.h" +#include "cc_debugfs.h" struct cc_debugfs_ctx { struct dentry *dir; @@ -24,7 +25,7 @@ struct cc_debugfs_ctx { */ static struct dentry *cc_debugfs_dir; -struct debugfs_reg32 debug_regs[] = { +static struct debugfs_reg32 debug_regs[] = { CC_DEBUG_REG(HOST_SIGNATURE), CC_DEBUG_REG(HOST_IRR), CC_DEBUG_REG(HOST_POWER_DOWN_EN), diff --git a/drivers/staging/ccree/ssi_ivgen.c b/drivers/staging/ccree/ssi_ivgen.c index 8051fa1..44acded 100644 --- a/drivers/staging/ccree/ssi_ivgen.c +++ b/drivers/staging/ccree/ssi_ivgen.c @@ -175,13 +175,10 @@ int cc_ivgen_init(struct cc_drvdata *drvdata) int rc; /* Allocate "this" context */ - drvdata->ivgen_handle = kzalloc(sizeof(*drvdata->ivgen_handle), - GFP_KERNEL); - if (!drvdata->ivgen_handle) + ivgen_ctx = kzalloc(sizeof(*ivgen_ctx), GFP_KERNEL); + if (!ivgen_ctx) return -ENOMEM; - ivgen_ctx = drvdata->ivgen_handle; - /* Allocate pool's header for initial enc. key/IV */ ivgen_ctx->pool_meta = dma_alloc_coherent(device, CC_IVPOOL_META_SIZE, &ivgen_ctx->pool_meta_dma, @@ -200,6 +197,8 @@ int cc_ivgen_init(struct cc_drvdata *drvdata) goto out; } + drvdata->ivgen_handle = ivgen_ctx; + return cc_init_iv_sram(drvdata); out: diff --git a/drivers/staging/ccree/ssi_sram_mgr.c b/drivers/staging/ccree/ssi_sram_mgr.c index 0613aaf..5bf37bf 100644 --- a/drivers/staging/ccree/ssi_sram_mgr.c +++ b/drivers/staging/ccree/ssi_sram_mgr.c @@ -32,13 +32,16 @@ void cc_sram_mgr_fini(struct cc_drvdata *drvdata) */ int cc_sram_mgr_init(struct cc_drvdata *drvdata) { + struct cc_sram_ctx *ctx; + /* Allocate "this" context */ - drvdata->sram_mgr_handle = kzalloc(sizeof(*drvdata->sram_mgr_handle), - GFP_KERNEL); + ctx = kzalloc(sizeof(*ctx), GFP_KERNEL); - if (!drvdata->sram_mgr_handle) + if (!ctx) return -ENOMEM; + drvdata->sram_mgr_handle = ctx; + return 0; } -- 2.7.4

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH 14/26] staging: ccree: handle end of sg list gracefully

by Gilad Ben-Yossef

If we are asked for number of entries of an offset bigger than the sg list we should not crash. Cc: stable(a)vger.kernel.org Signed-off-by: Gilad Ben-Yossef <gilad(a)benyossef.com> --- drivers/staging/ccree/ssi_buffer_mgr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/staging/ccree/ssi_buffer_mgr.c b/drivers/staging/ccree/ssi_buffer_mgr.c index 33b9709..cb3e235 100644 --- a/drivers/staging/ccree/ssi_buffer_mgr.c +++ b/drivers/staging/ccree/ssi_buffer_mgr.c @@ -94,7 +94,7 @@ static unsigned int cc_get_sgl_nents(struct device *dev, { unsigned int nents = 0; - while (nbytes) { + while (nbytes && sg_list) { if (sg_list->length) { nents++; /* get the number of bytes in the last entry */ -- 2.7.4

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH 13/26] staging: ccree: check DMA pool buf !NULL before free

by Gilad Ben-Yossef

If we ran out of DMA pool buffers, we get into the unmap code path with a NULL before. Deal with this by checking the virtual mapping is not NULL. Cc: stable(a)vger.kernel.org Signed-off-by: Gilad Ben-Yossef <gilad(a)benyossef.com> --- drivers/staging/ccree/ssi_buffer_mgr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/staging/ccree/ssi_buffer_mgr.c b/drivers/staging/ccree/ssi_buffer_mgr.c index 795f403..33b9709 100644 --- a/drivers/staging/ccree/ssi_buffer_mgr.c +++ b/drivers/staging/ccree/ssi_buffer_mgr.c @@ -465,7 +465,8 @@ void cc_unmap_blkcipher_request(struct device *dev, void *ctx, DMA_TO_DEVICE); } /* Release pool */ - if (req_ctx->dma_buf_type == CC_DMA_BUF_MLLI) { + if (req_ctx->dma_buf_type == CC_DMA_BUF_MLLI && + req_ctx->mlli_params.mlli_virt_addr) { dma_pool_free(req_ctx->mlli_params.curr_pool, req_ctx->mlli_params.mlli_virt_addr, req_ctx->mlli_params.mlli_dma_addr); -- 2.7.4

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH 11/26] stating: ccree: revert "staging: ccree: fix leak of import() after init()"

by Gilad Ben-Yossef

This reverts commit c5f39d07860c ("staging: ccree: fix leak of import() after init()") and commit aece09024414 ("staging: ccree: Uninitialized return in ssi_ahash_import()"). This is the wrong solution and ends up relying on uninitialized memory, although it was not obvious to me at the time. Cc: stable(a)vger.kernel.org Signed-off-by: Gilad Ben-Yossef <gilad(a)benyossef.com> --- drivers/staging/ccree/ssi_hash.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/drivers/staging/ccree/ssi_hash.c b/drivers/staging/ccree/ssi_hash.c index cc11344..9d71b33 100644 --- a/drivers/staging/ccree/ssi_hash.c +++ b/drivers/staging/ccree/ssi_hash.c @@ -1673,7 +1673,7 @@ static int cc_hash_import(struct ahash_request *req, const void *in) struct device *dev = drvdata_to_dev(ctx->drvdata); struct ahash_req_ctx *state = ahash_request_ctx(req); u32 tmp; - int rc = 0; + int rc; memcpy(&tmp, in, sizeof(u32)); if (tmp != CC_EXPORT_MAGIC) { @@ -1682,12 +1682,9 @@ static int cc_hash_import(struct ahash_request *req, const void *in) } in += sizeof(u32); - /* call init() to allocate bufs if the user hasn't */ - if (!state->digest_buff) { - rc = cc_hash_init(req); - if (rc) - goto out; - } + rc = cc_hash_init(req); + if (rc) + goto out; dma_sync_single_for_cpu(dev, state->digest_buff_dma_addr, ctx->inter_digestsize, DMA_BIDIRECTIONAL); -- 2.7.4

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: stub: stop printing kernel pointer addresses in messages" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: stub: stop printing kernel pointer addresses in messages to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 248a22044366f588d46754c54dfe29ffe4f8b4df Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Mon, 18 Dec 2017 17:23:37 -0700 Subject: usbip: stub: stop printing kernel pointer addresses in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 248a22044366f588d46754c54dfe29ffe4f8b4df upstream. Remove and/or change debug, info. and error messages to not print kernel pointer addresses. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/stub_main.c | 5 +++-- drivers/usb/usbip/stub_rx.c | 7 ++----- drivers/usb/usbip/stub_tx.c | 6 +++--- 3 files changed, 8 insertions(+), 10 deletions(-) --- a/drivers/usb/usbip/stub_main.c +++ b/drivers/usb/usbip/stub_main.c @@ -252,11 +252,12 @@ void stub_device_cleanup_urbs(struct stu struct stub_priv *priv; struct urb *urb; - dev_dbg(&sdev->udev->dev, "free sdev %p\n", sdev); + dev_dbg(&sdev->udev->dev, "Stub device cleaning up urbs\n"); while ((priv = stub_priv_pop(sdev))) { urb = priv->urb; - dev_dbg(&sdev->udev->dev, "free urb %p\n", urb); + dev_dbg(&sdev->udev->dev, "free urb seqnum %lu\n", + priv->seqnum); usb_kill_urb(urb); kmem_cache_free(stub_priv_cache, priv); --- a/drivers/usb/usbip/stub_rx.c +++ b/drivers/usb/usbip/stub_rx.c @@ -225,9 +225,6 @@ static int stub_recv_cmd_unlink(struct s if (priv->seqnum != pdu->u.cmd_unlink.seqnum) continue; - dev_info(&priv->urb->dev->dev, "unlink urb %p\n", - priv->urb); - /* * This matched urb is not completed yet (i.e., be in * flight in usb hcd hardware/driver). Now we are @@ -266,8 +263,8 @@ static int stub_recv_cmd_unlink(struct s ret = usb_unlink_urb(priv->urb); if (ret != -EINPROGRESS) dev_err(&priv->urb->dev->dev, - "failed to unlink a urb %p, ret %d\n", - priv->urb, ret); + "failed to unlink a urb # %lu, ret %d\n", + priv->seqnum, ret); return 0; } --- a/drivers/usb/usbip/stub_tx.c +++ b/drivers/usb/usbip/stub_tx.c @@ -102,7 +102,7 @@ void stub_complete(struct urb *urb) /* link a urb to the queue of tx. */ spin_lock_irqsave(&sdev->priv_lock, flags); if (sdev->ud.tcp_socket == NULL) { - usbip_dbg_stub_tx("ignore urb for closed connection %p", urb); + usbip_dbg_stub_tx("ignore urb for closed connection\n"); /* It will be freed in stub_device_cleanup_urbs(). */ } else if (priv->unlinking) { stub_enqueue_ret_unlink(sdev, priv->seqnum, urb->status); @@ -204,8 +204,8 @@ static int stub_send_ret_submit(struct s /* 1. setup usbip_header */ setup_ret_submit_pdu(&pdu_header, urb); - usbip_dbg_stub_tx("setup txdata seqnum: %d urb: %p\n", - pdu_header.base.seqnum, urb); + usbip_dbg_stub_tx("setup txdata seqnum: %d\n", + pdu_header.base.seqnum); usbip_header_correct_endian(&pdu_header, 1); iov[iovnum].iov_base = &pdu_header; Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.9/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.9/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.9/usbip-prevent-leaking-socket-pointer-address-in-messages.patch queue-4.9/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: vhci: stop printing kernel pointer addresses in messages" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: vhci: stop printing kernel pointer addresses in messages to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 8272d099d05f7ab2776cf56a2ab9f9443be18907 Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Mon, 18 Dec 2017 17:24:22 -0700 Subject: usbip: vhci: stop printing kernel pointer addresses in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 8272d099d05f7ab2776cf56a2ab9f9443be18907 upstream. Remove and/or change debug, info. and error messages to not print kernel pointer addresses. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/vhci_hcd.c | 10 ---------- drivers/usb/usbip/vhci_rx.c | 25 ++++++++++++------------- drivers/usb/usbip/vhci_tx.c | 3 ++- 3 files changed, 14 insertions(+), 24 deletions(-) --- a/drivers/usb/usbip/vhci_hcd.c +++ b/drivers/usb/usbip/vhci_hcd.c @@ -506,9 +506,6 @@ static int vhci_urb_enqueue(struct usb_h struct vhci_device *vdev; unsigned long flags; - usbip_dbg_vhci_hc("enter, usb_hcd %p urb %p mem_flags %d\n", - hcd, urb, mem_flags); - if (portnum > VHCI_HC_PORTS) { pr_err("invalid port number %d\n", portnum); return -ENODEV; @@ -671,8 +668,6 @@ static int vhci_urb_dequeue(struct usb_h struct vhci_device *vdev; unsigned long flags; - pr_info("dequeue a urb %p\n", urb); - spin_lock_irqsave(&vhci->lock, flags); priv = urb->hcpriv; @@ -700,7 +695,6 @@ static int vhci_urb_dequeue(struct usb_h /* tcp connection is closed */ spin_lock(&vdev->priv_lock); - pr_info("device %p seems to be disconnected\n", vdev); list_del(&priv->list); kfree(priv); urb->hcpriv = NULL; @@ -712,8 +706,6 @@ static int vhci_urb_dequeue(struct usb_h * vhci_rx will receive RET_UNLINK and give back the URB. * Otherwise, we give back it here. */ - pr_info("gives back urb %p\n", urb); - usb_hcd_unlink_urb_from_ep(hcd, urb); spin_unlock_irqrestore(&vhci->lock, flags); @@ -741,8 +733,6 @@ static int vhci_urb_dequeue(struct usb_h unlink->unlink_seqnum = priv->seqnum; - pr_info("device %p seems to be still connected\n", vdev); - /* send cmd_unlink and try to cancel the pending URB in the * peer */ list_add_tail(&unlink->list, &vdev->unlink_tx); --- a/drivers/usb/usbip/vhci_rx.c +++ b/drivers/usb/usbip/vhci_rx.c @@ -37,24 +37,23 @@ struct urb *pickup_urb_and_free_priv(str urb = priv->urb; status = urb->status; - usbip_dbg_vhci_rx("find urb %p vurb %p seqnum %u\n", - urb, priv, seqnum); + usbip_dbg_vhci_rx("find urb seqnum %u\n", seqnum); switch (status) { case -ENOENT: /* fall through */ case -ECONNRESET: - dev_info(&urb->dev->dev, - "urb %p was unlinked %ssynchronuously.\n", urb, - status == -ENOENT ? "" : "a"); + dev_dbg(&urb->dev->dev, + "urb seq# %u was unlinked %ssynchronuously\n", + seqnum, status == -ENOENT ? "" : "a"); break; case -EINPROGRESS: /* no info output */ break; default: - dev_info(&urb->dev->dev, - "urb %p may be in a error, status %d\n", urb, - status); + dev_dbg(&urb->dev->dev, + "urb seq# %u may be in a error, status %d\n", + seqnum, status); } list_del(&priv->list); @@ -80,9 +79,9 @@ static void vhci_recv_ret_submit(struct spin_unlock_irqrestore(&vdev->priv_lock, flags); if (!urb) { - pr_err("cannot find a urb of seqnum %u\n", pdu->base.seqnum); - pr_info("max seqnum %d\n", - atomic_read(&vhci->seqnum)); + pr_err("cannot find a urb of seqnum %u max seqnum %d\n", + pdu->base.seqnum, + atomic_read(&vhci_hcd->seqnum)); usbip_event_add(ud, VDEV_EVENT_ERROR_TCP); return; } @@ -104,7 +103,7 @@ static void vhci_recv_ret_submit(struct if (usbip_dbg_flag_vhci_rx) usbip_dump_urb(urb); - usbip_dbg_vhci_rx("now giveback urb %p\n", urb); + usbip_dbg_vhci_rx("now giveback urb %u\n", pdu->base.seqnum); spin_lock_irqsave(&vhci->lock, flags); usb_hcd_unlink_urb_from_ep(vhci_to_hcd(vhci), urb); @@ -170,7 +169,7 @@ static void vhci_recv_ret_unlink(struct pr_info("the urb (seqnum %d) was already given back\n", pdu->base.seqnum); } else { - usbip_dbg_vhci_rx("now giveback urb %p\n", urb); + usbip_dbg_vhci_rx("now giveback urb %d\n", pdu->base.seqnum); /* If unlink is successful, status is -ECONNRESET */ urb->status = pdu->u.ret_unlink.status; --- a/drivers/usb/usbip/vhci_tx.c +++ b/drivers/usb/usbip/vhci_tx.c @@ -83,7 +83,8 @@ static int vhci_send_cmd_submit(struct v memset(&msg, 0, sizeof(msg)); memset(&iov, 0, sizeof(iov)); - usbip_dbg_vhci_tx("setup txdata urb %p\n", urb); + usbip_dbg_vhci_tx("setup txdata urb seqnum %lu\n", + priv->seqnum); /* 1. setup usbip_header */ setup_cmd_submit_pdu(&pdu_header, urb); Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.9/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.9/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.9/usbip-prevent-leaking-socket-pointer-address-in-messages.patch queue-4.9/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: prevent leaking socket pointer address in messages" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: prevent leaking socket pointer address in messages to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-prevent-leaking-socket-pointer-address-in-messages.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 90120d15f4c397272aaf41077960a157fc4212bf Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Fri, 15 Dec 2017 10:50:09 -0700 Subject: usbip: prevent leaking socket pointer address in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 90120d15f4c397272aaf41077960a157fc4212bf upstream. usbip driver is leaking socket pointer address in messages. Remove the messages that aren't useful and print sockfd in the ones that are useful for debugging. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/stub_dev.c | 3 +-- drivers/usb/usbip/usbip_common.c | 14 ++++---------- drivers/usb/usbip/vhci_hcd.c | 2 +- 3 files changed, 6 insertions(+), 13 deletions(-) --- a/drivers/usb/usbip/stub_dev.c +++ b/drivers/usb/usbip/stub_dev.c @@ -163,8 +163,7 @@ static void stub_shutdown_connection(str * step 1? */ if (ud->tcp_socket) { - dev_dbg(&sdev->udev->dev, "shutdown tcp_socket %p\n", - ud->tcp_socket); + dev_dbg(&sdev->udev->dev, "shutdown sockfd %d\n", ud->sockfd); kernel_sock_shutdown(ud->tcp_socket, SHUT_RDWR); } --- a/drivers/usb/usbip/usbip_common.c +++ b/drivers/usb/usbip/usbip_common.c @@ -335,13 +335,10 @@ int usbip_recv(struct socket *sock, void char *bp = buf; int osize = size; - usbip_dbg_xmit("enter\n"); - - if (!sock || !buf || !size) { - pr_err("invalid arg, sock %p buff %p size %d\n", sock, buf, - size); + if (!sock || !buf || !size) return -EINVAL; - } + + usbip_dbg_xmit("enter\n"); do { sock->sk->sk_allocation = GFP_NOIO; @@ -354,11 +351,8 @@ int usbip_recv(struct socket *sock, void msg.msg_flags = MSG_NOSIGNAL; result = kernel_recvmsg(sock, &msg, &iov, 1, size, MSG_WAITALL); - if (result <= 0) { - pr_debug("receive sock %p buf %p size %u ret %d total %d\n", - sock, buf, size, result, total); + if (result <= 0) goto err; - } size -= result; buf += result; --- a/drivers/usb/usbip/vhci_hcd.c +++ b/drivers/usb/usbip/vhci_hcd.c @@ -823,7 +823,7 @@ static void vhci_shutdown_connection(str /* need this? see stub_dev.c */ if (ud->tcp_socket) { - pr_debug("shutdown tcp_socket %p\n", ud->tcp_socket); + pr_debug("shutdown tcp_socket %d\n", ud->sockfd); kernel_sock_shutdown(ud->tcp_socket, SHUT_RDWR); } Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.9/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.9/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.9/usbip-prevent-leaking-socket-pointer-address-in-messages.patch queue-4.9/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: fix usbip bind writing random string after command in match_busid" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: fix usbip bind writing random string after command in match_busid to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 544c4605acc5ae4afe7dd5914147947db182f2fb Mon Sep 17 00:00:00 2001 From: Juan Zea <juan.zea(a)qindel.com> Date: Fri, 15 Dec 2017 10:21:20 +0100 Subject: usbip: fix usbip bind writing random string after command in match_busid From: Juan Zea <juan.zea(a)qindel.com> commit 544c4605acc5ae4afe7dd5914147947db182f2fb upstream. usbip bind writes commands followed by random string when writing to match_busid attribute in sysfs, caused by using full variable size instead of string length. Signed-off-by: Juan Zea <juan.zea(a)qindel.com> Acked-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- tools/usb/usbip/src/utils.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) --- a/tools/usb/usbip/src/utils.c +++ b/tools/usb/usbip/src/utils.c @@ -30,6 +30,7 @@ int modify_match_busid(char *busid, int char command[SYSFS_BUS_ID_SIZE + 4]; char match_busid_attr_path[SYSFS_PATH_MAX]; int rc; + int cmd_size; snprintf(match_busid_attr_path, sizeof(match_busid_attr_path), "%s/%s/%s/%s/%s/%s", SYSFS_MNT_PATH, SYSFS_BUS_NAME, @@ -37,12 +38,14 @@ int modify_match_busid(char *busid, int attr_name); if (add) - snprintf(command, SYSFS_BUS_ID_SIZE + 4, "add %s", busid); + cmd_size = snprintf(command, SYSFS_BUS_ID_SIZE + 4, "add %s", + busid); else - snprintf(command, SYSFS_BUS_ID_SIZE + 4, "del %s", busid); + cmd_size = snprintf(command, SYSFS_BUS_ID_SIZE + 4, "del %s", + busid); rc = write_sysfs_attribute(match_busid_attr_path, command, - sizeof(command)); + cmd_size); if (rc < 0) { dbg("failed to write match_busid: %s", strerror(errno)); return -1; Patches currently in stable-queue which might be from juan.zea(a)qindel.com are queue-4.9/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: vhci: stop printing kernel pointer addresses in messages" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: vhci: stop printing kernel pointer addresses in messages to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 8272d099d05f7ab2776cf56a2ab9f9443be18907 Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Mon, 18 Dec 2017 17:24:22 -0700 Subject: usbip: vhci: stop printing kernel pointer addresses in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 8272d099d05f7ab2776cf56a2ab9f9443be18907 upstream. Remove and/or change debug, info. and error messages to not print kernel pointer addresses. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/vhci_hcd.c | 10 ---------- drivers/usb/usbip/vhci_rx.c | 23 +++++++++++------------ drivers/usb/usbip/vhci_tx.c | 3 ++- 3 files changed, 13 insertions(+), 23 deletions(-) --- a/drivers/usb/usbip/vhci_hcd.c +++ b/drivers/usb/usbip/vhci_hcd.c @@ -467,9 +467,6 @@ static int vhci_urb_enqueue(struct usb_h int ret = 0; struct vhci_device *vdev; - usbip_dbg_vhci_hc("enter, usb_hcd %p urb %p mem_flags %d\n", - hcd, urb, mem_flags); - /* patch to usb_sg_init() is in 2.5.60 */ BUG_ON(!urb->transfer_buffer && urb->transfer_buffer_length); @@ -627,8 +624,6 @@ static int vhci_urb_dequeue(struct usb_h struct vhci_priv *priv; struct vhci_device *vdev; - pr_info("dequeue a urb %p\n", urb); - spin_lock(&the_controller->lock); priv = urb->hcpriv; @@ -656,7 +651,6 @@ static int vhci_urb_dequeue(struct usb_h /* tcp connection is closed */ spin_lock(&vdev->priv_lock); - pr_info("device %p seems to be disconnected\n", vdev); list_del(&priv->list); kfree(priv); urb->hcpriv = NULL; @@ -668,8 +662,6 @@ static int vhci_urb_dequeue(struct usb_h * vhci_rx will receive RET_UNLINK and give back the URB. * Otherwise, we give back it here. */ - pr_info("gives back urb %p\n", urb); - usb_hcd_unlink_urb_from_ep(hcd, urb); spin_unlock(&the_controller->lock); @@ -698,8 +690,6 @@ static int vhci_urb_dequeue(struct usb_h unlink->unlink_seqnum = priv->seqnum; - pr_info("device %p seems to be still connected\n", vdev); - /* send cmd_unlink and try to cancel the pending URB in the * peer */ list_add_tail(&unlink->list, &vdev->unlink_tx); --- a/drivers/usb/usbip/vhci_rx.c +++ b/drivers/usb/usbip/vhci_rx.c @@ -37,24 +37,23 @@ struct urb *pickup_urb_and_free_priv(str urb = priv->urb; status = urb->status; - usbip_dbg_vhci_rx("find urb %p vurb %p seqnum %u\n", - urb, priv, seqnum); + usbip_dbg_vhci_rx("find urb seqnum %u\n", seqnum); switch (status) { case -ENOENT: /* fall through */ case -ECONNRESET: - dev_info(&urb->dev->dev, - "urb %p was unlinked %ssynchronuously.\n", urb, - status == -ENOENT ? "" : "a"); + dev_dbg(&urb->dev->dev, + "urb seq# %u was unlinked %ssynchronuously\n", + seqnum, status == -ENOENT ? "" : "a"); break; case -EINPROGRESS: /* no info output */ break; default: - dev_info(&urb->dev->dev, - "urb %p may be in a error, status %d\n", urb, - status); + dev_dbg(&urb->dev->dev, + "urb seq# %u may be in a error, status %d\n", + seqnum, status); } list_del(&priv->list); @@ -78,8 +77,8 @@ static void vhci_recv_ret_submit(struct spin_unlock(&vdev->priv_lock); if (!urb) { - pr_err("cannot find a urb of seqnum %u\n", pdu->base.seqnum); - pr_info("max seqnum %d\n", + pr_err("cannot find a urb of seqnum %u max seqnum %d\n", + pdu->base.seqnum, atomic_read(&the_controller->seqnum)); usbip_event_add(ud, VDEV_EVENT_ERROR_TCP); return; @@ -102,7 +101,7 @@ static void vhci_recv_ret_submit(struct if (usbip_dbg_flag_vhci_rx) usbip_dump_urb(urb); - usbip_dbg_vhci_rx("now giveback urb %p\n", urb); + usbip_dbg_vhci_rx("now giveback urb %u\n", pdu->base.seqnum); spin_lock(&the_controller->lock); usb_hcd_unlink_urb_from_ep(vhci_to_hcd(the_controller), urb); @@ -165,7 +164,7 @@ static void vhci_recv_ret_unlink(struct pr_info("the urb (seqnum %d) was already given back\n", pdu->base.seqnum); } else { - usbip_dbg_vhci_rx("now giveback urb %p\n", urb); + usbip_dbg_vhci_rx("now giveback urb %d\n", pdu->base.seqnum); /* If unlink is successful, status is -ECONNRESET */ urb->status = pdu->u.ret_unlink.status; --- a/drivers/usb/usbip/vhci_tx.c +++ b/drivers/usb/usbip/vhci_tx.c @@ -82,7 +82,8 @@ static int vhci_send_cmd_submit(struct v memset(&msg, 0, sizeof(msg)); memset(&iov, 0, sizeof(iov)); - usbip_dbg_vhci_tx("setup txdata urb %p\n", urb); + usbip_dbg_vhci_tx("setup txdata urb seqnum %lu\n", + priv->seqnum); /* 1. setup usbip_header */ setup_cmd_submit_pdu(&pdu_header, urb); Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.4/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.4/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.4/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: stub: stop printing kernel pointer addresses in messages" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: stub: stop printing kernel pointer addresses in messages to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 248a22044366f588d46754c54dfe29ffe4f8b4df Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Mon, 18 Dec 2017 17:23:37 -0700 Subject: usbip: stub: stop printing kernel pointer addresses in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 248a22044366f588d46754c54dfe29ffe4f8b4df upstream. Remove and/or change debug, info. and error messages to not print kernel pointer addresses. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/stub_main.c | 5 +++-- drivers/usb/usbip/stub_rx.c | 7 ++----- drivers/usb/usbip/stub_tx.c | 4 ++-- 3 files changed, 7 insertions(+), 9 deletions(-) --- a/drivers/usb/usbip/stub_main.c +++ b/drivers/usb/usbip/stub_main.c @@ -252,11 +252,12 @@ void stub_device_cleanup_urbs(struct stu struct stub_priv *priv; struct urb *urb; - dev_dbg(&sdev->udev->dev, "free sdev %p\n", sdev); + dev_dbg(&sdev->udev->dev, "Stub device cleaning up urbs\n"); while ((priv = stub_priv_pop(sdev))) { urb = priv->urb; - dev_dbg(&sdev->udev->dev, "free urb %p\n", urb); + dev_dbg(&sdev->udev->dev, "free urb seqnum %lu\n", + priv->seqnum); usb_kill_urb(urb); kmem_cache_free(stub_priv_cache, priv); --- a/drivers/usb/usbip/stub_rx.c +++ b/drivers/usb/usbip/stub_rx.c @@ -230,9 +230,6 @@ static int stub_recv_cmd_unlink(struct s if (priv->seqnum != pdu->u.cmd_unlink.seqnum) continue; - dev_info(&priv->urb->dev->dev, "unlink urb %p\n", - priv->urb); - /* * This matched urb is not completed yet (i.e., be in * flight in usb hcd hardware/driver). Now we are @@ -271,8 +268,8 @@ static int stub_recv_cmd_unlink(struct s ret = usb_unlink_urb(priv->urb); if (ret != -EINPROGRESS) dev_err(&priv->urb->dev->dev, - "failed to unlink a urb %p, ret %d\n", - priv->urb, ret); + "failed to unlink a urb # %lu, ret %d\n", + priv->seqnum, ret); return 0; } --- a/drivers/usb/usbip/stub_tx.c +++ b/drivers/usb/usbip/stub_tx.c @@ -201,8 +201,8 @@ static int stub_send_ret_submit(struct s /* 1. setup usbip_header */ setup_ret_submit_pdu(&pdu_header, urb); - usbip_dbg_stub_tx("setup txdata seqnum: %d urb: %p\n", - pdu_header.base.seqnum, urb); + usbip_dbg_stub_tx("setup txdata seqnum: %d\n", + pdu_header.base.seqnum); usbip_header_correct_endian(&pdu_header, 1); iov[iovnum].iov_base = &pdu_header; Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.4/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.4/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.4/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: fix usbip bind writing random string after command in match_busid" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: fix usbip bind writing random string after command in match_busid to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 544c4605acc5ae4afe7dd5914147947db182f2fb Mon Sep 17 00:00:00 2001 From: Juan Zea <juan.zea(a)qindel.com> Date: Fri, 15 Dec 2017 10:21:20 +0100 Subject: usbip: fix usbip bind writing random string after command in match_busid From: Juan Zea <juan.zea(a)qindel.com> commit 544c4605acc5ae4afe7dd5914147947db182f2fb upstream. usbip bind writes commands followed by random string when writing to match_busid attribute in sysfs, caused by using full variable size instead of string length. Signed-off-by: Juan Zea <juan.zea(a)qindel.com> Acked-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- tools/usb/usbip/src/utils.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) --- a/tools/usb/usbip/src/utils.c +++ b/tools/usb/usbip/src/utils.c @@ -30,6 +30,7 @@ int modify_match_busid(char *busid, int char command[SYSFS_BUS_ID_SIZE + 4]; char match_busid_attr_path[SYSFS_PATH_MAX]; int rc; + int cmd_size; snprintf(match_busid_attr_path, sizeof(match_busid_attr_path), "%s/%s/%s/%s/%s/%s", SYSFS_MNT_PATH, SYSFS_BUS_NAME, @@ -37,12 +38,14 @@ int modify_match_busid(char *busid, int attr_name); if (add) - snprintf(command, SYSFS_BUS_ID_SIZE + 4, "add %s", busid); + cmd_size = snprintf(command, SYSFS_BUS_ID_SIZE + 4, "add %s", + busid); else - snprintf(command, SYSFS_BUS_ID_SIZE + 4, "del %s", busid); + cmd_size = snprintf(command, SYSFS_BUS_ID_SIZE + 4, "del %s", + busid); rc = write_sysfs_attribute(match_busid_attr_path, command, - sizeof(command)); + cmd_size); if (rc < 0) { dbg("failed to write match_busid: %s", strerror(errno)); return -1; Patches currently in stable-queue which might be from juan.zea(a)qindel.com are queue-4.4/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: vhci: stop printing kernel pointer addresses in messages" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: vhci: stop printing kernel pointer addresses in messages to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 8272d099d05f7ab2776cf56a2ab9f9443be18907 Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Mon, 18 Dec 2017 17:24:22 -0700 Subject: usbip: vhci: stop printing kernel pointer addresses in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 8272d099d05f7ab2776cf56a2ab9f9443be18907 upstream. Remove and/or change debug, info. and error messages to not print kernel pointer addresses. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/vhci_hcd.c | 10 ---------- drivers/usb/usbip/vhci_rx.c | 23 +++++++++++------------ drivers/usb/usbip/vhci_tx.c | 3 ++- 3 files changed, 13 insertions(+), 23 deletions(-) --- a/drivers/usb/usbip/vhci_hcd.c +++ b/drivers/usb/usbip/vhci_hcd.c @@ -670,9 +670,6 @@ static int vhci_urb_enqueue(struct usb_h struct vhci_device *vdev; unsigned long flags; - usbip_dbg_vhci_hc("enter, usb_hcd %p urb %p mem_flags %d\n", - hcd, urb, mem_flags); - if (portnum > VHCI_HC_PORTS) { pr_err("invalid port number %d\n", portnum); return -ENODEV; @@ -836,8 +833,6 @@ static int vhci_urb_dequeue(struct usb_h struct vhci_device *vdev; unsigned long flags; - pr_info("dequeue a urb %p\n", urb); - spin_lock_irqsave(&vhci->lock, flags); priv = urb->hcpriv; @@ -865,7 +860,6 @@ static int vhci_urb_dequeue(struct usb_h /* tcp connection is closed */ spin_lock(&vdev->priv_lock); - pr_info("device %p seems to be disconnected\n", vdev); list_del(&priv->list); kfree(priv); urb->hcpriv = NULL; @@ -877,8 +871,6 @@ static int vhci_urb_dequeue(struct usb_h * vhci_rx will receive RET_UNLINK and give back the URB. * Otherwise, we give back it here. */ - pr_info("gives back urb %p\n", urb); - usb_hcd_unlink_urb_from_ep(hcd, urb); spin_unlock_irqrestore(&vhci->lock, flags); @@ -906,8 +898,6 @@ static int vhci_urb_dequeue(struct usb_h unlink->unlink_seqnum = priv->seqnum; - pr_info("device %p seems to be still connected\n", vdev); - /* send cmd_unlink and try to cancel the pending URB in the * peer */ list_add_tail(&unlink->list, &vdev->unlink_tx); --- a/drivers/usb/usbip/vhci_rx.c +++ b/drivers/usb/usbip/vhci_rx.c @@ -37,24 +37,23 @@ struct urb *pickup_urb_and_free_priv(str urb = priv->urb; status = urb->status; - usbip_dbg_vhci_rx("find urb %p vurb %p seqnum %u\n", - urb, priv, seqnum); + usbip_dbg_vhci_rx("find urb seqnum %u\n", seqnum); switch (status) { case -ENOENT: /* fall through */ case -ECONNRESET: - dev_info(&urb->dev->dev, - "urb %p was unlinked %ssynchronuously.\n", urb, - status == -ENOENT ? "" : "a"); + dev_dbg(&urb->dev->dev, + "urb seq# %u was unlinked %ssynchronuously\n", + seqnum, status == -ENOENT ? "" : "a"); break; case -EINPROGRESS: /* no info output */ break; default: - dev_info(&urb->dev->dev, - "urb %p may be in a error, status %d\n", urb, - status); + dev_dbg(&urb->dev->dev, + "urb seq# %u may be in a error, status %d\n", + seqnum, status); } list_del(&priv->list); @@ -81,8 +80,8 @@ static void vhci_recv_ret_submit(struct spin_unlock_irqrestore(&vdev->priv_lock, flags); if (!urb) { - pr_err("cannot find a urb of seqnum %u\n", pdu->base.seqnum); - pr_info("max seqnum %d\n", + pr_err("cannot find a urb of seqnum %u max seqnum %d\n", + pdu->base.seqnum, atomic_read(&vhci_hcd->seqnum)); usbip_event_add(ud, VDEV_EVENT_ERROR_TCP); return; @@ -105,7 +104,7 @@ static void vhci_recv_ret_submit(struct if (usbip_dbg_flag_vhci_rx) usbip_dump_urb(urb); - usbip_dbg_vhci_rx("now giveback urb %p\n", urb); + usbip_dbg_vhci_rx("now giveback urb %u\n", pdu->base.seqnum); spin_lock_irqsave(&vhci->lock, flags); usb_hcd_unlink_urb_from_ep(vhci_hcd_to_hcd(vhci_hcd), urb); @@ -172,7 +171,7 @@ static void vhci_recv_ret_unlink(struct pr_info("the urb (seqnum %d) was already given back\n", pdu->base.seqnum); } else { - usbip_dbg_vhci_rx("now giveback urb %p\n", urb); + usbip_dbg_vhci_rx("now giveback urb %d\n", pdu->base.seqnum); /* If unlink is successful, status is -ECONNRESET */ urb->status = pdu->u.ret_unlink.status; --- a/drivers/usb/usbip/vhci_tx.c +++ b/drivers/usb/usbip/vhci_tx.c @@ -83,7 +83,8 @@ static int vhci_send_cmd_submit(struct v memset(&msg, 0, sizeof(msg)); memset(&iov, 0, sizeof(iov)); - usbip_dbg_vhci_tx("setup txdata urb %p\n", urb); + usbip_dbg_vhci_tx("setup txdata urb seqnum %lu\n", + priv->seqnum); /* 1. setup usbip_header */ setup_cmd_submit_pdu(&pdu_header, urb); Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.14/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.14/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.14/usbip-prevent-leaking-socket-pointer-address-in-messages.patch queue-4.14/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: stub: stop printing kernel pointer addresses in messages" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: stub: stop printing kernel pointer addresses in messages to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 248a22044366f588d46754c54dfe29ffe4f8b4df Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Mon, 18 Dec 2017 17:23:37 -0700 Subject: usbip: stub: stop printing kernel pointer addresses in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 248a22044366f588d46754c54dfe29ffe4f8b4df upstream. Remove and/or change debug, info. and error messages to not print kernel pointer addresses. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/stub_main.c | 5 +++-- drivers/usb/usbip/stub_rx.c | 7 ++----- drivers/usb/usbip/stub_tx.c | 6 +++--- 3 files changed, 8 insertions(+), 10 deletions(-) --- a/drivers/usb/usbip/stub_main.c +++ b/drivers/usb/usbip/stub_main.c @@ -251,11 +251,12 @@ void stub_device_cleanup_urbs(struct stu struct stub_priv *priv; struct urb *urb; - dev_dbg(&sdev->udev->dev, "free sdev %p\n", sdev); + dev_dbg(&sdev->udev->dev, "Stub device cleaning up urbs\n"); while ((priv = stub_priv_pop(sdev))) { urb = priv->urb; - dev_dbg(&sdev->udev->dev, "free urb %p\n", urb); + dev_dbg(&sdev->udev->dev, "free urb seqnum %lu\n", + priv->seqnum); usb_kill_urb(urb); kmem_cache_free(stub_priv_cache, priv); --- a/drivers/usb/usbip/stub_rx.c +++ b/drivers/usb/usbip/stub_rx.c @@ -225,9 +225,6 @@ static int stub_recv_cmd_unlink(struct s if (priv->seqnum != pdu->u.cmd_unlink.seqnum) continue; - dev_info(&priv->urb->dev->dev, "unlink urb %p\n", - priv->urb); - /* * This matched urb is not completed yet (i.e., be in * flight in usb hcd hardware/driver). Now we are @@ -266,8 +263,8 @@ static int stub_recv_cmd_unlink(struct s ret = usb_unlink_urb(priv->urb); if (ret != -EINPROGRESS) dev_err(&priv->urb->dev->dev, - "failed to unlink a urb %p, ret %d\n", - priv->urb, ret); + "failed to unlink a urb # %lu, ret %d\n", + priv->seqnum, ret); return 0; } --- a/drivers/usb/usbip/stub_tx.c +++ b/drivers/usb/usbip/stub_tx.c @@ -102,7 +102,7 @@ void stub_complete(struct urb *urb) /* link a urb to the queue of tx. */ spin_lock_irqsave(&sdev->priv_lock, flags); if (sdev->ud.tcp_socket == NULL) { - usbip_dbg_stub_tx("ignore urb for closed connection %p", urb); + usbip_dbg_stub_tx("ignore urb for closed connection\n"); /* It will be freed in stub_device_cleanup_urbs(). */ } else if (priv->unlinking) { stub_enqueue_ret_unlink(sdev, priv->seqnum, urb->status); @@ -204,8 +204,8 @@ static int stub_send_ret_submit(struct s /* 1. setup usbip_header */ setup_ret_submit_pdu(&pdu_header, urb); - usbip_dbg_stub_tx("setup txdata seqnum: %d urb: %p\n", - pdu_header.base.seqnum, urb); + usbip_dbg_stub_tx("setup txdata seqnum: %d\n", + pdu_header.base.seqnum); usbip_header_correct_endian(&pdu_header, 1); iov[iovnum].iov_base = &pdu_header; Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.14/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.14/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.14/usbip-prevent-leaking-socket-pointer-address-in-messages.patch queue-4.14/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: prevent leaking socket pointer address in messages" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: prevent leaking socket pointer address in messages to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-prevent-leaking-socket-pointer-address-in-messages.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 90120d15f4c397272aaf41077960a157fc4212bf Mon Sep 17 00:00:00 2001 From: Shuah Khan <shuahkh(a)osg.samsung.com> Date: Fri, 15 Dec 2017 10:50:09 -0700 Subject: usbip: prevent leaking socket pointer address in messages From: Shuah Khan <shuahkh(a)osg.samsung.com> commit 90120d15f4c397272aaf41077960a157fc4212bf upstream. usbip driver is leaking socket pointer address in messages. Remove the messages that aren't useful and print sockfd in the ones that are useful for debugging. Signed-off-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/usb/usbip/stub_dev.c | 3 +-- drivers/usb/usbip/usbip_common.c | 16 +++++----------- drivers/usb/usbip/vhci_hcd.c | 2 +- 3 files changed, 7 insertions(+), 14 deletions(-) --- a/drivers/usb/usbip/stub_dev.c +++ b/drivers/usb/usbip/stub_dev.c @@ -163,8 +163,7 @@ static void stub_shutdown_connection(str * step 1? */ if (ud->tcp_socket) { - dev_dbg(&sdev->udev->dev, "shutdown tcp_socket %p\n", - ud->tcp_socket); + dev_dbg(&sdev->udev->dev, "shutdown sockfd %d\n", ud->sockfd); kernel_sock_shutdown(ud->tcp_socket, SHUT_RDWR); } --- a/drivers/usb/usbip/usbip_common.c +++ b/drivers/usb/usbip/usbip_common.c @@ -331,26 +331,20 @@ int usbip_recv(struct socket *sock, void struct msghdr msg = {.msg_flags = MSG_NOSIGNAL}; int total = 0; + if (!sock || !buf || !size) + return -EINVAL; + iov_iter_kvec(&msg.msg_iter, READ|ITER_KVEC, &iov, 1, size); usbip_dbg_xmit("enter\n"); - if (!sock || !buf || !size) { - pr_err("invalid arg, sock %p buff %p size %d\n", sock, buf, - size); - return -EINVAL; - } - do { - int sz = msg_data_left(&msg); + msg_data_left(&msg); sock->sk->sk_allocation = GFP_NOIO; result = sock_recvmsg(sock, &msg, MSG_WAITALL); - if (result <= 0) { - pr_debug("receive sock %p buf %p size %u ret %d total %d\n", - sock, buf + total, sz, result, total); + if (result <= 0) goto err; - } total += result; } while (msg_data_left(&msg)); --- a/drivers/usb/usbip/vhci_hcd.c +++ b/drivers/usb/usbip/vhci_hcd.c @@ -989,7 +989,7 @@ static void vhci_shutdown_connection(str /* need this? see stub_dev.c */ if (ud->tcp_socket) { - pr_debug("shutdown tcp_socket %p\n", ud->tcp_socket); + pr_debug("shutdown tcp_socket %d\n", ud->sockfd); kernel_sock_shutdown(ud->tcp_socket, SHUT_RDWR); } Patches currently in stable-queue which might be from shuahkh(a)osg.samsung.com are queue-4.14/usbip-vhci-stop-printing-kernel-pointer-addresses-in-messages.patch queue-4.14/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch queue-4.14/usbip-prevent-leaking-socket-pointer-address-in-messages.patch queue-4.14/usbip-stub-stop-printing-kernel-pointer-addresses-in-messages.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "usbip: fix usbip bind writing random string after command in match_busid" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usbip: fix usbip bind writing random string after command in match_busid to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 544c4605acc5ae4afe7dd5914147947db182f2fb Mon Sep 17 00:00:00 2001 From: Juan Zea <juan.zea(a)qindel.com> Date: Fri, 15 Dec 2017 10:21:20 +0100 Subject: usbip: fix usbip bind writing random string after command in match_busid From: Juan Zea <juan.zea(a)qindel.com> commit 544c4605acc5ae4afe7dd5914147947db182f2fb upstream. usbip bind writes commands followed by random string when writing to match_busid attribute in sysfs, caused by using full variable size instead of string length. Signed-off-by: Juan Zea <juan.zea(a)qindel.com> Acked-by: Shuah Khan <shuahkh(a)osg.samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- tools/usb/usbip/src/utils.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) --- a/tools/usb/usbip/src/utils.c +++ b/tools/usb/usbip/src/utils.c @@ -30,6 +30,7 @@ int modify_match_busid(char *busid, int char command[SYSFS_BUS_ID_SIZE + 4]; char match_busid_attr_path[SYSFS_PATH_MAX]; int rc; + int cmd_size; snprintf(match_busid_attr_path, sizeof(match_busid_attr_path), "%s/%s/%s/%s/%s/%s", SYSFS_MNT_PATH, SYSFS_BUS_NAME, @@ -37,12 +38,14 @@ int modify_match_busid(char *busid, int attr_name); if (add) - snprintf(command, SYSFS_BUS_ID_SIZE + 4, "add %s", busid); + cmd_size = snprintf(command, SYSFS_BUS_ID_SIZE + 4, "add %s", + busid); else - snprintf(command, SYSFS_BUS_ID_SIZE + 4, "del %s", busid); + cmd_size = snprintf(command, SYSFS_BUS_ID_SIZE + 4, "del %s", + busid); rc = write_sysfs_attribute(match_busid_attr_path, command, - sizeof(command)); + cmd_size); if (rc < 0) { dbg("failed to write match_busid: %s", strerror(errno)); return -1; Patches currently in stable-queue which might be from juan.zea(a)qindel.com are queue-4.14/usbip-fix-usbip-bind-writing-random-string-after-command-in-match_busid.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH] Sparc

by David Miller

Please queue up the following sparc64 bug fix for 4.14.x -stable. Thank you.

7 years, 8 months

2
1
0 0

[Linux-stable-mirror] Patch "[PATCH] sparc64: repair calling incorrect hweight function from stubs" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled [PATCH] sparc64: repair calling incorrect hweight function from stubs to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sparc64-repair-calling-incorrect-hweight-function-from-stubs.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Jan 1 11:05:39 CET 2018 From: Jan Engelhardt <jengelh(a)inai.de> Date: Mon, 25 Dec 2017 03:43:53 +0100 Subject: [PATCH] sparc64: repair calling incorrect hweight function from stubs From: Jan Engelhardt <jengelh(a)inai.de> [ Upstream commit 59585b4be9ae4dc6506551709bdcd6f5210b8a01 ] Commit v4.12-rc4-1-g9289ea7f952b introduced a mistake that made the 64-bit hweight stub call the 16-bit hweight function. Fixes: 9289ea7f952b ("sparc64: Use indirect calls in hamming weight stubs") Signed-off-by: Jan Engelhardt <jengelh(a)inai.de> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/sparc/lib/hweight.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/arch/sparc/lib/hweight.S +++ b/arch/sparc/lib/hweight.S @@ -44,8 +44,8 @@ EXPORT_SYMBOL(__arch_hweight32) .previous ENTRY(__arch_hweight64) - sethi %hi(__sw_hweight16), %g1 - jmpl %g1 + %lo(__sw_hweight16), %g0 + sethi %hi(__sw_hweight64), %g1 + jmpl %g1 + %lo(__sw_hweight64), %g0 nop ENDPROC(__arch_hweight64) EXPORT_SYMBOL(__arch_hweight64) Patches currently in stable-queue which might be from jengelh(a)inai.de are queue-4.14/sparc64-repair-calling-incorrect-hweight-function-from-stubs.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH] usb: quirks: Add reset-resume quirk for QCA6174 Rome Bluetooth

by Leif Liddy

This is a rework of reverted commit fd865802c66bc451dc515ed89360f84376ce1a56 The issue is that some QCA Rome bluetooth controllers stop functioning upon resume from suspend. These devices seem to be losing power during suspend. This patch will enable reset_resume in usb core (instead of btusb) and will target the specific device 0x0cf3:0xe300 Signed-off-by: Leif Liddy <leif.linux(a)gmail.com> --- drivers/usb/core/quirks.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/usb/core/quirks.c b/drivers/usb/core/quirks.c index a10b346b9777..d94e6658d054 100644 --- a/drivers/usb/core/quirks.c +++ b/drivers/usb/core/quirks.c @@ -248,6 +248,9 @@ static const struct usb_device_id usb_quirk_list[] = { /* INTEL VALUE SSD */ { USB_DEVICE(0x8086, 0xf1a5), .driver_info = USB_QUIRK_RESET_RESUME }, + /* QCA Rome Bluetooth in QCA6174 wireless module */ + { USB_DEVICE(0x0cf3, 0xe300), .driver_info = USB_QUIRK_RESET_RESUME }, + { } /* terminating entry must be last */ }; -- 2.14.3

7 years, 8 months

3
2
0 0

Re: [Linux-stable-mirror] How to report regression in stable tree?

by Ian Pilcher

On 12/30/2017 04:41 PM, Ben Hutchings wrote: > For i915, I think you should either: > - open a bug at bugs.freedesktop.org (under DRI, DRM/Intel) > - mail intel-gfx(a)lists.freedesktop.org and cc stable Done. https://bugs.freedesktop.org/show_bug.cgi?id=104425 -- ======================================================================== Ian Pilcher arequipeno(a)gmail.com -------- "I grew up before Mark Zuckerberg invented friendship" -------- ========================================================================

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH 2/2] x86/dumpstack: Print registers for first stack frame

by Josh Poimboeuf

In the stack dump code, if the frame after the starting pt_regs is also a regs frame, the registers don't get printed. Fix that. Reported-by: Andy Lutomirski <luto(a)amacapital.net> Tested-by: Alexander Tsoy <alexander(a)tsoy.me> Fixes: 3b3fa11bc700 ("x86/dumpstack: Print any pt_regs found on the stack") Cc: stable(a)vger.kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe(a)redhat.com> --- arch/x86/kernel/dumpstack.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index e4c2c73bd99e..0c2f29e8a5a2 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -115,6 +115,7 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, unwind_start(&state, task, regs, stack); stack = stack ? : get_stack_pointer(task, regs); + regs = unwind_get_entry_regs(&state, &partial); /* * Iterate through the stacks, starting with the current stack pointer. @@ -132,7 +133,7 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, * - hardirq stack * - entry stack */ - for (regs = NULL; stack; stack = PTR_ALIGN(stack_info.next_sp, sizeof(long))) { + for ( ; stack; stack = PTR_ALIGN(stack_info.next_sp, sizeof(long))) { const char *stack_name; if (get_stack_info(stack, task, &stack_info, &visit_mask)) { -- 2.13.6

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCH 1/2] x86/dumpstack: Fix partial register dumps

by Josh Poimboeuf

The show_regs_safe() logic is wrong. When there's an iret stack frame, it prints the entire pt_regs -- most of which is random stack data -- instead of just the five registers at the end. show_regs_safe() is also poorly named: the on_stack() checks aren't for safety. Rename the function to show_regs_if_on_stack() and add a comment to explain why the checks are needed. These issues were introduced with the "partial register dump" feature of the following commit: b02fcf9ba121 ("x86/unwinder: Handle stack overflows more gracefully") That patch had gone through a few iterations of development, and the above issues were artifacts from a previous iteration of the patch where 'regs' pointed directly to the iret frame rather than to the (partially empty) pt_regs. Tested-by: Alexander Tsoy <alexander(a)tsoy.me> Fixes: b02fcf9ba121 ("x86/unwinder: Handle stack overflows more gracefully") Cc: stable(a)vger.kernel.org Signed-off-by: Josh Poimboeuf <jpoimboe(a)redhat.com> --- arch/x86/include/asm/unwind.h | 17 +++++++++++++---- arch/x86/kernel/dumpstack.c | 28 ++++++++++++++++++++-------- arch/x86/kernel/stacktrace.c | 2 +- 3 files changed, 34 insertions(+), 13 deletions(-) diff --git a/arch/x86/include/asm/unwind.h b/arch/x86/include/asm/unwind.h index c1688c2d0a12..1f86e1b0a5cd 100644 --- a/arch/x86/include/asm/unwind.h +++ b/arch/x86/include/asm/unwind.h @@ -56,18 +56,27 @@ void unwind_start(struct unwind_state *state, struct task_struct *task, #if defined(CONFIG_UNWINDER_ORC) || defined(CONFIG_UNWINDER_FRAME_POINTER) /* - * WARNING: The entire pt_regs may not be safe to dereference. In some cases, - * only the iret frame registers are accessible. Use with caution! + * If 'partial' returns true, only the iret frame registers are valid. */ -static inline struct pt_regs *unwind_get_entry_regs(struct unwind_state *state) +static inline struct pt_regs *unwind_get_entry_regs(struct unwind_state *state, + bool *partial) { if (unwind_done(state)) return NULL; + if (partial) { +#ifdef CONFIG_UNWINDER_ORC + *partial = !state->full_regs; +#else + *partial = false; +#endif + } + return state->regs; } #else -static inline struct pt_regs *unwind_get_entry_regs(struct unwind_state *state) +static inline struct pt_regs *unwind_get_entry_regs(struct unwind_state *state, + bool *partial) { return NULL; } diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index 36b17e0febe8..e4c2c73bd99e 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -76,12 +76,23 @@ void show_iret_regs(struct pt_regs *regs) regs->sp, regs->flags); } -static void show_regs_safe(struct stack_info *info, struct pt_regs *regs) +static void show_regs_if_on_stack(struct stack_info *info, struct pt_regs *regs, + bool partial) { - if (on_stack(info, regs, sizeof(*regs))) + /* + * These on_stack() checks aren't strictly necessary: the unwind code + * has already validated the 'regs' pointer. The checks are done for + * ordering reasons: if the registers are on the next stack, we don't + * want to print them out yet. Otherwise they'll be shown as part of + * the wrong stack. Later, when show_trace_log_lvl() switches to the + * next stack, this function will be called again with the same regs so + * they can be printed in the right context. + */ + if (!partial && on_stack(info, regs, sizeof(*regs))) { __show_regs(regs, 0); - else if (on_stack(info, (void *)regs + IRET_FRAME_OFFSET, - IRET_FRAME_SIZE)) { + + } else if (partial && on_stack(info, (void *)regs + IRET_FRAME_OFFSET, + IRET_FRAME_SIZE)) { /* * When an interrupt or exception occurs in entry code, the * full pt_regs might not have been saved yet. In that case @@ -98,6 +109,7 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, struct stack_info stack_info = {0}; unsigned long visit_mask = 0; int graph_idx = 0; + bool partial; printk("%sCall Trace:\n", log_lvl); @@ -140,7 +152,7 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, printk("%s <%s>\n", log_lvl, stack_name); if (regs) - show_regs_safe(&stack_info, regs); + show_regs_if_on_stack(&stack_info, regs, partial); /* * Scan the stack, printing any text addresses we find. At the @@ -164,7 +176,7 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, /* * Don't print regs->ip again if it was already printed - * by show_regs_safe() below. + * by show_regs_if_on_stack(). */ if (regs && stack == &regs->ip) goto next; @@ -199,9 +211,9 @@ void show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs, unwind_next_frame(&state); /* if the frame has entry regs, print them */ - regs = unwind_get_entry_regs(&state); + regs = unwind_get_entry_regs(&state, &partial); if (regs) - show_regs_safe(&stack_info, regs); + show_regs_if_on_stack(&stack_info, regs, partial); } if (stack_name) diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c index 77835bc021c7..7dd0d2a0d142 100644 --- a/arch/x86/kernel/stacktrace.c +++ b/arch/x86/kernel/stacktrace.c @@ -102,7 +102,7 @@ __save_stack_trace_reliable(struct stack_trace *trace, for (unwind_start(&state, task, NULL, NULL); !unwind_done(&state); unwind_next_frame(&state)) { - regs = unwind_get_entry_regs(&state); + regs = unwind_get_entry_regs(&state, NULL); if (regs) { /* * Kernel mode registers on the stack indicate an -- 2.13.6

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sync objtool's copy of x86-opcode-map.txt" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sync objtool's copy of x86-opcode-map.txt to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sync-objtool-s-copy-of-x86-opcode-map.txt.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 14:13:42 CET 2017 Date: Sun, 31 Dec 2017 14:13:42 +0100 To: Greg KH <gregkh(a)linuxfoundation.org> From: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Subject: sync objtool's copy of x86-opcode-map.txt When building objtool, we get the warning: warning: objtool: x86 instruction decoder differs from kernel That's due to commit 2816c0455cea088f07a210f8a00701a82a78aa9c which was commit 12a78d43de767eaf8fb272facb7a7b6f2dc6a9df upstream that modified arch/x86/lib/x86-opcode-map.txt without also updating the objtool copy. The objtool copy was updated in a much larger patch upstream, but we don't need all of that here, so just update the single file. If this gets too annoying, I'll just end up doing what we did for 4.14 and backport the whole series to keep this from happening again, but as this seems to be rare in the 4.9-stable series, this single patch should be fine. Cc: Masami Hiramatsu <mhiramat(a)kernel.org> Cc: H. Peter Anvin <hpa(a)zytor.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: Ingo Molnar <mingo(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- tools/objtool/arch/x86/insn/x86-opcode-map.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/tools/objtool/arch/x86/insn/x86-opcode-map.txt +++ b/tools/objtool/arch/x86/insn/x86-opcode-map.txt @@ -896,7 +896,7 @@ EndTable GrpTable: Grp3_1 0: TEST Eb,Ib -1: +1: TEST Eb,Ib 2: NOT Eb 3: NEG Eb 4: MUL AL,Eb Patches currently in stable-queue which might be from gregkh(a)linuxfoundation.org are queue-4.9/asoc-tlv320aic31xx-fix-gpio1-register-definition.patch queue-4.9/net-mlx5-fix-rate-limit-packet-pacing-naming-and-struct.patch queue-4.9/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.9/asoc-fsl_ssi-ac-97-ops-need-regmap-clock-and-cleaning-up-on-failure.patch queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/asoc-twl4030-fix-child-node-lookup.patch queue-4.9/x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch queue-4.9/asoc-wm_adsp-fix-validation-of-firmware-and-coeff-lengths.patch queue-4.9/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.9/adding-missing-rcu_read_unlock-in-ipxip6_rcv.patch queue-4.9/net-reevalulate-autoflowlabel-setting-after-sysctl-setting.patch queue-4.9/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.9/alsa-hda-drop-useless-warn_on.patch queue-4.9/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.9/net-qmi_wwan-add-sierra-em7565-1199-9091.patch queue-4.9/net-bridge-fix-early-call-to-br_stp_change_bridge_id-and-plug-newlink-leaks.patch queue-4.9/ptr_ring-add-barriers.patch queue-4.9/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.9/x86-mm-enable-cr4.pcide-on-supported-systems.patch queue-4.9/net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch queue-4.9/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.9/net-fec-unmap-the-xmit-buffer-that-are-not-transferred-by-dma.patch queue-4.9/netlink-add-netns-check-on-taps.patch queue-4.9/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.9/sync-objtool-s-copy-of-x86-opcode-map.txt.patch queue-4.9/alsa-hda-fix-headset-mic-detection-issue-on-a-dell-machine.patch queue-4.9/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.9/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.9/net-phy-micrel-ksz9031-reconfigure-autoneg-after-phy-autoneg-workaround.patch queue-4.9/tracing-fix-crash-when-it-fails-to-alloc-ring-buffer.patch queue-4.9/tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch queue-4.9/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.9/tracing-remove-extra-zeroing-out-of-the-ring-buffer-page.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/net-fec-allow-reception-of-frames-bigger-than-1522-bytes.patch queue-4.9/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.9/kbuild-add-fno-stack-check-to-kernel-build-options.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/ipv4-fix-use-after-free-when-flushing-fib-tables.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/rds-check-cmsg_len-before-dereferencing-cmsg_data.patch queue-4.9/net-mlx5-fix-error-flow-in-create_qp-command.patch queue-4.9/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.9/ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch queue-4.9/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch queue-4.9/x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch queue-4.9/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch queue-4.9/tcp_bbr-record-full-bw-reached-decision-in-new-full_bw_reached-bit.patch queue-4.9/iw_cxgb4-only-validate-the-msn-for-successful-completions.patch queue-4.9/asoc-da7218-fix-fix-child-node-lookup.patch queue-4.9/sctp-replace-use-of-sockets_allocated-with-specified-macro.patch queue-4.9/ipv6-mcast-better-catch-silly-mtu-values.patch queue-4.9/net-mlx5e-fix-features-check-of-ipv6-traffic.patch queue-4.9/tracing-fix-possible-double-free-on-failure-of-allocating-trace-buffer.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp md5sig: Use skb's saddr when replying to an incoming segment" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp md5sig: Use skb's saddr when replying to an incoming segment to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Christoph Paasch <cpaasch(a)apple.com> Date: Mon, 11 Dec 2017 00:05:46 -0800 Subject: tcp md5sig: Use skb's saddr when replying to an incoming segment From: Christoph Paasch <cpaasch(a)apple.com> [ Upstream commit 30791ac41927ebd3e75486f9504b6d2280463bf0 ] The MD5-key that belongs to a connection is identified by the peer's IP-address. When we are in tcp_v4(6)_reqsk_send_ack(), we are replying to an incoming segment from tcp_check_req() that failed the seq-number checks. Thus, to find the correct key, we need to use the skb's saddr and not the daddr. This bug seems to have been there since quite a while, but probably got unnoticed because the consequences are not catastrophic. We will call tcp_v4_reqsk_send_ack only to send a challenge-ACK back to the peer, thus the connection doesn't really fail. Fixes: 9501f9722922 ("tcp md5sig: Let the caller pass appropriate key for tcp_v{4,6}_do_calc_md5_hash().") Signed-off-by: Christoph Paasch <cpaasch(a)apple.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_ipv4.c | 2 +- net/ipv6/tcp_ipv6.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -823,7 +823,7 @@ static void tcp_v4_reqsk_send_ack(const tcp_time_stamp, req->ts_recent, 0, - tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->daddr, + tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->saddr, AF_INET), inet_rsk(req)->no_srccheck ? IP_REPLY_ARG_NOSRCCHECK : 0, ip_hdr(skb)->tos); --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -949,7 +949,7 @@ static void tcp_v6_reqsk_send_ack(const tcp_rsk(req)->rcv_nxt, req->rsk_rcv_wnd >> inet_rsk(req)->rcv_wscale, tcp_time_stamp, req->ts_recent, sk->sk_bound_dev_if, - tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr), + tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->saddr), 0, 0); } Patches currently in stable-queue which might be from cpaasch(a)apple.com are queue-4.4/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tg3: Fix rx hang on MTU change with 5717/5719" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tg3: Fix rx hang on MTU change with 5717/5719 to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Brian King <brking(a)linux.vnet.ibm.com> Date: Fri, 15 Dec 2017 15:21:50 -0600 Subject: tg3: Fix rx hang on MTU change with 5717/5719 From: Brian King <brking(a)linux.vnet.ibm.com> [ Upstream commit 748a240c589824e9121befb1cba5341c319885bc ] This fixes a hang issue seen when changing the MTU size from 1500 MTU to 9000 MTU on both 5717 and 5719 chips. In discussion with Broadcom, they've indicated that these chipsets have the same phy as the 57766 chipset, so the same workarounds apply. This has been tested by IBM on both Power 8 and Power 9 systems as well as by Broadcom on x86 hardware and has been confirmed to resolve the hang issue. Signed-off-by: Brian King <brking(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/broadcom/tg3.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -14228,7 +14228,9 @@ static int tg3_change_mtu(struct net_dev /* Reset PHY, otherwise the read DMA engine will be in a mode that * breaks all requests to 256 bytes. */ - if (tg3_asic_rev(tp) == ASIC_REV_57766) + if (tg3_asic_rev(tp) == ASIC_REV_57766 || + tg3_asic_rev(tp) == ASIC_REV_5717 || + tg3_asic_rev(tp) == ASIC_REV_5719) reset_phy = true; err = tg3_restart_hw(tp, reset_phy); Patches currently in stable-queue which might be from brking(a)linux.vnet.ibm.com are queue-4.4/tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sctp: Replace use of sockets_allocated with specified macro." has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: Replace use of sockets_allocated with specified macro. to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-replace-use-of-sockets_allocated-with-specified-macro.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> Date: Fri, 22 Dec 2017 10:15:20 -0800 Subject: sctp: Replace use of sockets_allocated with specified macro. From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> [ Upstream commit 8cb38a602478e9f806571f6920b0a3298aabf042 ] The patch(180d8cd942ce) replaces all uses of struct sock fields' memory_pressure, memory_allocated, sockets_allocated, and sysctl_mem to accessor macros. But the sockets_allocated field of sctp sock is not replaced at all. Then replace it now for unifying the code. Fixes: 180d8cd942ce ("foundations of per-cgroup memory pressure controlling.") Cc: Glauber Costa <glommer(a)parallels.com> Signed-off-by: Tonghao Zhang <zhangtonghao(a)didichuxing.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/socket.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4153,7 +4153,7 @@ static int sctp_init_sock(struct sock *s SCTP_DBG_OBJCNT_INC(sock); local_bh_disable(); - percpu_counter_inc(&sctp_sockets_allocated); + sk_sockets_allocated_inc(sk); sock_prot_inuse_add(net, sk->sk_prot, 1); /* Nothing can fail after this block, otherwise @@ -4197,7 +4197,7 @@ static void sctp_destroy_sock(struct soc } sctp_endpoint_free(sp->ep); local_bh_disable(); - percpu_counter_dec(&sctp_sockets_allocated); + sk_sockets_allocated_dec(sk); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); local_bh_enable(); } Patches currently in stable-queue which might be from xiangxia.m.yue(a)gmail.com are queue-4.4/sctp-replace-use-of-sockets_allocated-with-specified-macro.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sock: free skb in skb_complete_tx_timestamp on error" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sock: free skb in skb_complete_tx_timestamp on error to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Willem de Bruijn <willemb(a)google.com> Date: Wed, 13 Dec 2017 14:41:06 -0500 Subject: sock: free skb in skb_complete_tx_timestamp on error From: Willem de Bruijn <willemb(a)google.com> [ Upstream commit 35b99dffc3f710cafceee6c8c6ac6a98eb2cb4bf ] skb_complete_tx_timestamp must ingest the skb it is passed. Call kfree_skb if the skb cannot be enqueued. Fixes: b245be1f4db1 ("net-timestamp: no-payload only sysctl") Fixes: 9ac25fc06375 ("net: fix socket refcounting in skb_complete_tx_timestamp()") Reported-by: Richard Cochran <richardcochran(a)gmail.com> Signed-off-by: Willem de Bruijn <willemb(a)google.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/skbuff.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -3676,7 +3676,7 @@ void skb_complete_tx_timestamp(struct sk struct sock *sk = skb->sk; if (!skb_may_tx_timestamp(sk, false)) - return; + goto err; /* Take a reference to prevent skb_orphan() from freeing the socket, * but only if the socket refcount is not zero. @@ -3685,7 +3685,11 @@ void skb_complete_tx_timestamp(struct sk *skb_hwtstamps(skb) = *hwtstamps; __skb_complete_tx_timestamp(skb, sk, SCM_TSTAMP_SND); sock_put(sk); + return; } + +err: + kfree_skb(skb); } EXPORT_SYMBOL_GPL(skb_complete_tx_timestamp); Patches currently in stable-queue which might be from willemb(a)google.com are queue-4.4/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: reevalulate autoflowlabel setting after sysctl setting" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: reevalulate autoflowlabel setting after sysctl setting to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-reevalulate-autoflowlabel-setting-after-sysctl-setting.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Shaohua Li <shli(a)fb.com> Date: Wed, 20 Dec 2017 12:10:21 -0800 Subject: net: reevalulate autoflowlabel setting after sysctl setting From: Shaohua Li <shli(a)fb.com> [ Upstream commit 513674b5a2c9c7a67501506419da5c3c77ac6f08 ] sysctl.ip6.auto_flowlabels is default 1. In our hosts, we set it to 2. If sockopt doesn't set autoflowlabel, outcome packets from the hosts are supposed to not include flowlabel. This is true for normal packet, but not for reset packet. The reason is ipv6_pinfo.autoflowlabel is set in sock creation. Later if we change sysctl.ip6.auto_flowlabels, the ipv6_pinfo.autoflowlabel isn't changed, so the sock will keep the old behavior in terms of auto flowlabel. Reset packet is suffering from this problem, because reset packet is sent from a special control socket, which is created at boot time. Since sysctl.ipv6.auto_flowlabels is 1 by default, the control socket will always have its ipv6_pinfo.autoflowlabel set, even after user set sysctl.ipv6.auto_flowlabels to 1, so reset packset will always have flowlabel. Normal sock created before sysctl setting suffers from the same issue. We can't even turn off autoflowlabel unless we kill all socks in the hosts. To fix this, if IPV6_AUTOFLOWLABEL sockopt is used, we use the autoflowlabel setting from user, otherwise we always call ip6_default_np_autolabel() which has the new settings of sysctl. Note, this changes behavior a little bit. Before commit 42240901f7c4 (ipv6: Implement different admin modes for automatic flow labels), the autoflowlabel behavior of a sock isn't sticky, eg, if sysctl changes, existing connection will change autoflowlabel behavior. After that commit, autoflowlabel behavior is sticky in the whole life of the sock. With this patch, the behavior isn't sticky again. Cc: Martin KaFai Lau <kafai(a)fb.com> Cc: Eric Dumazet <eric.dumazet(a)gmail.com> Cc: Tom Herbert <tom(a)quantonium.net> Signed-off-by: Shaohua Li <shli(a)fb.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/ipv6.h | 3 ++- net/ipv6/af_inet6.c | 1 - net/ipv6/ip6_output.c | 12 ++++++++++-- net/ipv6/ipv6_sockglue.c | 1 + 4 files changed, 13 insertions(+), 4 deletions(-) --- a/include/linux/ipv6.h +++ b/include/linux/ipv6.h @@ -215,7 +215,8 @@ struct ipv6_pinfo { * 100: prefer care-of address */ dontfrag:1, - autoflowlabel:1; + autoflowlabel:1, + autoflowlabel_set:1; __u8 min_hopcount; __u8 tclass; __be32 rcv_flowinfo; --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -200,7 +200,6 @@ lookup_protocol: np->mcast_hops = IPV6_DEFAULT_MCASTHOPS; np->mc_loop = 1; np->pmtudisc = IPV6_PMTUDISC_WANT; - np->autoflowlabel = ip6_default_np_autolabel(sock_net(sk)); sk->sk_ipv6only = net->ipv6.sysctl.bindv6only; /* Init the ipv4 part of the socket since we can have sockets --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -148,6 +148,14 @@ int ip6_output(struct net *net, struct s !(IP6CB(skb)->flags & IP6SKB_REROUTED)); } +static bool ip6_autoflowlabel(struct net *net, const struct ipv6_pinfo *np) +{ + if (!np->autoflowlabel_set) + return ip6_default_np_autolabel(net); + else + return np->autoflowlabel; +} + /* * xmit an sk_buff (used by TCP, SCTP and DCCP) * Note : socket lock is not held for SYNACK packets, but might be modified @@ -211,7 +219,7 @@ int ip6_xmit(const struct sock *sk, stru hlimit = ip6_dst_hoplimit(dst); ip6_flow_hdr(hdr, tclass, ip6_make_flowlabel(net, skb, fl6->flowlabel, - np->autoflowlabel, fl6)); + ip6_autoflowlabel(net, np), fl6)); hdr->payload_len = htons(seg_len); hdr->nexthdr = proto; @@ -1675,7 +1683,7 @@ struct sk_buff *__ip6_make_skb(struct so ip6_flow_hdr(hdr, v6_cork->tclass, ip6_make_flowlabel(net, skb, fl6->flowlabel, - np->autoflowlabel, fl6)); + ip6_autoflowlabel(net, np), fl6)); hdr->hop_limit = v6_cork->hop_limit; hdr->nexthdr = proto; hdr->saddr = fl6->saddr; --- a/net/ipv6/ipv6_sockglue.c +++ b/net/ipv6/ipv6_sockglue.c @@ -872,6 +872,7 @@ pref_skip_coa: break; case IPV6_AUTOFLOWLABEL: np->autoflowlabel = valbool; + np->autoflowlabel_set = 1; retv = 0; break; } Patches currently in stable-queue which might be from shli(a)fb.com are queue-4.4/net-reevalulate-autoflowlabel-setting-after-sysctl-setting.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "netlink: Add netns check on taps" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled netlink: Add netns check on taps to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: netlink-add-netns-check-on-taps.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Wed, 6 Dec 2017 12:12:27 -0800 Subject: netlink: Add netns check on taps From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit 93c647643b48f0131f02e45da3bd367d80443291 ] Currently, a nlmon link inside a child namespace can observe systemwide netlink activity. Filter the traffic so that nlmon can only sniff netlink messages from its own netns. Test case: vpnns -- bash -c "ip link add nlmon0 type nlmon; \ ip link set nlmon0 up; \ tcpdump -i nlmon0 -q -w /tmp/nlmon.pcap -U" & sudo ip xfrm state add src 10.1.1.1 dst 10.1.1.2 proto esp \ spi 0x1 mode transport \ auth sha1 0x6162633132330000000000000000000000000000 \ enc aes 0x00000000000000000000000000000000 grep --binary abc123 /tmp/nlmon.pcap Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/netlink/af_netlink.c | 3 +++ 1 file changed, 3 insertions(+) --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -261,6 +261,9 @@ static int __netlink_deliver_tap_skb(str struct sock *sk = skb->sk; int ret = -ENOMEM; + if (!net_eq(dev_net(dev), sock_net(sk))) + return 0; + dev_hold(dev); if (is_vmalloc_addr(skb->head)) Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-4.4/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.4/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: qmi_wwan: add Sierra EM7565 1199:9091" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: qmi_wwan: add Sierra EM7565 1199:9091 to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-qmi_wwan-add-sierra-em7565-1199-9091.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Sebastian Sjoholm <ssjoholm(a)mac.com> Date: Mon, 11 Dec 2017 21:51:14 +0100 Subject: net: qmi_wwan: add Sierra EM7565 1199:9091 From: Sebastian Sjoholm <ssjoholm(a)mac.com> [ Upstream commit aceef61ee56898cfa7b6960fb60b9326c3860441 ] Sierra Wireless EM7565 is an Qualcomm MDM9x50 based M.2 modem. The USB id is added to qmi_wwan.c to allow QMI communication with the EM7565. Signed-off-by: Sebastian Sjoholm <ssjoholm(a)mac.com> Acked-by: Bjørn Mork <bjorn(a)mork.no> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -737,6 +737,7 @@ static const struct usb_device_id produc {QMI_FIXED_INTF(0x1199, 0x9079, 10)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 8)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 10)}, /* Sierra Wireless EM74xx */ + {QMI_FIXED_INTF(0x1199, 0x9091, 8)}, /* Sierra Wireless EM7565 */ {QMI_FIXED_INTF(0x1bbb, 0x011e, 4)}, /* Telekom Speedstick LTE II (Alcatel One Touch L100V LTE) */ {QMI_FIXED_INTF(0x1bbb, 0x0203, 2)}, /* Alcatel L800MA */ {QMI_FIXED_INTF(0x2357, 0x0201, 4)}, /* TP-LINK HSUPA Modem MA180 */ Patches currently in stable-queue which might be from ssjoholm(a)mac.com are queue-4.4/net-qmi_wwan-add-sierra-em7565-1199-9091.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Date: Wed, 6 Dec 2017 15:23:23 +0100 Subject: net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> [ Upstream commit 589bf32f09852041fbd3b7ce1a9e703f95c230ba ] add appropriate calls to clk_disable_unprepare() by jumping to out_mdio in case orion_mdio_probe() returns -EPROBE_DEFER. Found by Linux Driver Verification project (linuxtesting.org). Fixes: 3d604da1e954 ("net: mvmdio: get and enable optional clock") Signed-off-by: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/marvell/mvmdio.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/marvell/mvmdio.c +++ b/drivers/net/ethernet/marvell/mvmdio.c @@ -241,7 +241,8 @@ static int orion_mdio_probe(struct platf dev->regs + MVMDIO_ERR_INT_MASK); } else if (dev->err_interrupt == -EPROBE_DEFER) { - return -EPROBE_DEFER; + ret = -EPROBE_DEFER; + goto out_mdio; } mutex_init(&dev->lock); Patches currently in stable-queue which might be from Tobias.Jordan(a)elektrobit.com are queue-4.4/net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-phy-micrel-ksz9031-reconfigure-autoneg-after-phy-autoneg-workaround.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Grygorii Strashko <grygorii.strashko(a)ti.com> Date: Wed, 20 Dec 2017 18:45:10 -0600 Subject: net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround From: Grygorii Strashko <grygorii.strashko(a)ti.com> [ Upstream commit c1a8d0a3accf64a014d605e6806ce05d1c17adf1 ] Under some circumstances driver will perform PHY reset in ksz9031_read_status() to fix autoneg failure case (idle error count = 0xFF). When this happens ksz9031 will not detect link status change any more when connecting to Netgear 1G switch (link can be recovered sometimes by restarting netdevice "ifconfig down up"). Reproduced with TI am572x board equipped with ksz9031 PHY while connecting to Netgear 1G switch. Fix the issue by reconfiguring autonegotiation after PHY reset in ksz9031_read_status(). Fixes: d2fd719bcb0e ("net/phy: micrel: Add workaround for bad autoneg") Signed-off-by: Grygorii Strashko <grygorii.strashko(a)ti.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/phy/micrel.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -541,6 +541,7 @@ static int ksz9031_read_status(struct ph phydev->link = 0; if (phydev->drv->config_intr && phy_interrupt_is_valid(phydev)) phydev->drv->config_intr(phydev); + return genphy_config_aneg(phydev); } return 0; Patches currently in stable-queue which might be from grygorii.strashko(a)ti.com are queue-4.4/net-phy-micrel-ksz9031-reconfigure-autoneg-after-phy-autoneg-workaround.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: ipv4: fix for a race condition in raw_sendmsg" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: ipv4: fix for a race condition in raw_sendmsg to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Mohamed Ghannam <simo.ghannam(a)gmail.com> Date: Sun, 10 Dec 2017 03:50:58 +0000 Subject: net: ipv4: fix for a race condition in raw_sendmsg From: Mohamed Ghannam <simo.ghannam(a)gmail.com> [ Upstream commit 8f659a03a0ba9289b9aeb9b4470e6fb263d6f483 ] inet->hdrincl is racy, and could lead to uninitialized stack pointer usage, so its value should be read only once. Fixes: c008ba5bdc9f ("ipv4: Avoid reading user iov twice after raw_probe_proto_opt") Signed-off-by: Mohamed Ghannam <simo.ghannam(a)gmail.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/raw.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -500,11 +500,16 @@ static int raw_sendmsg(struct sock *sk, int err; struct ip_options_data opt_copy; struct raw_frag_vec rfv; + int hdrincl; err = -EMSGSIZE; if (len > 0xFFFF) goto out; + /* hdrincl should be READ_ONCE(inet->hdrincl) + * but READ_ONCE() doesn't work with bit fields + */ + hdrincl = inet->hdrincl; /* * Check the flags. */ @@ -579,7 +584,7 @@ static int raw_sendmsg(struct sock *sk, /* Linux does not mangle headers on raw sockets, * so that IP options + IP_HDRINCL is non-sense. */ - if (inet->hdrincl) + if (hdrincl) goto done; if (ipc.opt->opt.srr) { if (!daddr) @@ -601,9 +606,9 @@ static int raw_sendmsg(struct sock *sk, flowi4_init_output(&fl4, ipc.oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE, - inet->hdrincl ? IPPROTO_RAW : sk->sk_protocol, + hdrincl ? IPPROTO_RAW : sk->sk_protocol, inet_sk_flowi_flags(sk) | - (inet->hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), + (hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), daddr, saddr, 0, 0); if (!saddr && ipc.oif) { @@ -612,7 +617,7 @@ static int raw_sendmsg(struct sock *sk, goto done; } - if (!inet->hdrincl) { + if (!hdrincl) { rfv.msg = msg; rfv.hlen = 0; @@ -637,7 +642,7 @@ static int raw_sendmsg(struct sock *sk, goto do_confirm; back_from_confirm: - if (inet->hdrincl) + if (hdrincl) err = raw_send_hdrinc(sk, &fl4, msg, len, &rt, msg->msg_flags); Patches currently in stable-queue which might be from simo.ghannam(a)gmail.com are queue-4.4/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: igmp: Use correct source address on IGMPv3 reports" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: igmp: Use correct source address on IGMPv3 reports to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-igmp-use-correct-source-address-on-igmpv3-reports.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Mon, 11 Dec 2017 11:13:45 -0800 Subject: net: igmp: Use correct source address on IGMPv3 reports From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit a46182b00290839fa3fa159d54fd3237bd8669f0 ] Closing a multicast socket after the final IPv4 address is deleted from an interface can generate a membership report that uses the source IP from a different interface. The following test script, run from an isolated netns, reproduces the issue: #!/bin/bash ip link add dummy0 type dummy ip link add dummy1 type dummy ip link set dummy0 up ip link set dummy1 up ip addr add 10.1.1.1/24 dev dummy0 ip addr add 192.168.99.99/24 dev dummy1 tcpdump -U -i dummy0 & socat EXEC:"sleep 2" \ UDP4-DATAGRAM:239.101.1.68:8889,ip-add-membership=239.0.1.68:10.1.1.1 & sleep 1 ip addr del 10.1.1.1/24 dev dummy0 sleep 5 kill %tcpdump RFC 3376 specifies that the report must be sent with a valid IP source address from the destination subnet, or from address 0.0.0.0. Add an extra check to make sure this is the case. Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/igmp.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -89,6 +89,7 @@ #include <linux/rtnetlink.h> #include <linux/times.h> #include <linux/pkt_sched.h> +#include <linux/byteorder/generic.h> #include <net/net_namespace.h> #include <net/arp.h> @@ -327,6 +328,23 @@ igmp_scount(struct ip_mc_list *pmc, int return scount; } +/* source address selection per RFC 3376 section 4.2.13 */ +static __be32 igmpv3_get_srcaddr(struct net_device *dev, + const struct flowi4 *fl4) +{ + struct in_device *in_dev = __in_dev_get_rcu(dev); + + if (!in_dev) + return htonl(INADDR_ANY); + + for_ifa(in_dev) { + if (inet_ifa_match(fl4->saddr, ifa)) + return fl4->saddr; + } endfor_ifa(in_dev); + + return htonl(INADDR_ANY); +} + static struct sk_buff *igmpv3_newpack(struct net_device *dev, unsigned int mtu) { struct sk_buff *skb; @@ -374,7 +392,7 @@ static struct sk_buff *igmpv3_newpack(st pip->frag_off = htons(IP_DF); pip->ttl = 1; pip->daddr = fl4.daddr; - pip->saddr = fl4.saddr; + pip->saddr = igmpv3_get_srcaddr(dev, &fl4); pip->protocol = IPPROTO_IGMP; pip->tot_len = 0; /* filled in later */ ip_select_ident(net, skb, NULL); Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-4.4/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.4/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: Fix double free and memory corruption in get_net_ns_by_id()" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: Fix double free and memory corruption in get_net_ns_by_id() to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: "Eric W. Biederman" <ebiederm(a)xmission.com> Date: Tue, 19 Dec 2017 11:27:56 -0600 Subject: net: Fix double free and memory corruption in get_net_ns_by_id() From: "Eric W. Biederman" <ebiederm(a)xmission.com> [ Upstream commit 21b5944350052d2583e82dd59b19a9ba94a007f0 ] (I can trivially verify that that idr_remove in cleanup_net happens after the network namespace count has dropped to zero --EWB) Function get_net_ns_by_id() does not check for net::count after it has found a peer in netns_ids idr. It may dereference a peer, after its count has already been finaly decremented. This leads to double free and memory corruption: put_net(peer) rtnl_lock() atomic_dec_and_test(&peer->count) [count=0] ... __put_net(peer) get_net_ns_by_id(net, id) spin_lock(&cleanup_list_lock) list_add(&net->cleanup_list, &cleanup_list) spin_unlock(&cleanup_list_lock) queue_work() peer = idr_find(&net->netns_ids, id) | get_net(peer) [count=1] | ... | (use after final put) v ... cleanup_net() ... spin_lock(&cleanup_list_lock) ... list_replace_init(&cleanup_list, ..) ... spin_unlock(&cleanup_list_lock) ... ... ... ... put_net(peer) ... atomic_dec_and_test(&peer->count) [count=0] ... spin_lock(&cleanup_list_lock) ... list_add(&net->cleanup_list, &cleanup_list) ... spin_unlock(&cleanup_list_lock) ... queue_work() ... rtnl_unlock() rtnl_lock() ... for_each_net(tmp) { ... id = __peernet2id(tmp, peer) ... spin_lock_irq(&tmp->nsid_lock) ... idr_remove(&tmp->netns_ids, id) ... ... ... net_drop_ns() ... net_free(peer) ... } ... | v cleanup_net() ... (Second free of peer) Also, put_net() on the right cpu may reorder with left's cpu list_replace_init(&cleanup_list, ..), and then cleanup_list will be corrupted. Since cleanup_net() is executed in worker thread, while put_net(peer) can happen everywhere, there should be enough time for concurrent get_net_ns_by_id() to pick the peer up, and the race does not seem to be unlikely. The patch fixes the problem in standard way. (Also, there is possible problem in peernet2id_alloc(), which requires check for net::count under nsid_lock and maybe_get_net(peer), but in current stable kernel it's used under rtnl_lock() and it has to be safe. Openswitch begun to use peernet2id_alloc(), and possibly it should be fixed too. While this is not in stable kernel yet, so I'll send a separate message to netdev@ later). Cc: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Signed-off-by: Kirill Tkhai <ktkhai(a)virtuozzo.com> Fixes: 0c7aecd4bde4 "netns: add rtnl cmd to add and get peer netns ids" Reviewed-by: Andrey Ryabinin <aryabinin(a)virtuozzo.com> Reviewed-by: "Eric W. Biederman" <ebiederm(a)xmission.com> Signed-off-by: Eric W. Biederman <ebiederm(a)xmission.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Acked-by: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/net_namespace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -261,7 +261,7 @@ struct net *get_net_ns_by_id(struct net spin_lock_irqsave(&net->nsid_lock, flags); peer = idr_find(&net->netns_ids, id); if (peer) - get_net(peer); + peer = maybe_get_net(peer); spin_unlock_irqrestore(&net->nsid_lock, flags); rcu_read_unlock(); Patches currently in stable-queue which might be from ebiederm(a)xmission.com are queue-4.4/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-bridge-fix-early-call-to-br_stp_change_bridge_id-and-plug-newlink-leaks.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> Date: Mon, 18 Dec 2017 17:35:09 +0200 Subject: net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks From: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> [ Upstream commit 84aeb437ab98a2bce3d4b2111c79723aedfceb33 ] The early call to br_stp_change_bridge_id in bridge's newlink can cause a memory leak if an error occurs during the newlink because the fdb entries are not cleaned up if a different lladdr was specified, also another minor issue is that it generates fdb notifications with ifindex = 0. Another unrelated memory leak is the bridge sysfs entries which get added on NETDEV_REGISTER event, but are not cleaned up in the newlink error path. To remove this special case the call to br_stp_change_bridge_id is done after netdev register and we cleanup the bridge on changelink error via br_dev_delete to plug all leaks. This patch makes netlink bridge destruction on newlink error the same as dellink and ioctl del which is necessary since at that point we have a fully initialized bridge device. To reproduce the issue: $ ip l add br0 address 00:11:22:33:44:55 type bridge group_fwd_mask 1 RTNETLINK answers: Invalid argument $ rmmod bridge [ 1822.142525] ============================================================================= [ 1822.143640] BUG bridge_fdb_cache (Tainted: G O ): Objects remaining in bridge_fdb_cache on __kmem_cache_shutdown() [ 1822.144821] ----------------------------------------------------------------------------- [ 1822.145990] Disabling lock debugging due to kernel taint [ 1822.146732] INFO: Slab 0x0000000092a844b2 objects=32 used=2 fp=0x00000000fef011b0 flags=0x1ffff8000000100 [ 1822.147700] CPU: 2 PID: 13584 Comm: rmmod Tainted: G B O 4.15.0-rc2+ #87 [ 1822.148578] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_083030-gandalf 04/01/2014 [ 1822.150008] Call Trace: [ 1822.150510] dump_stack+0x78/0xa9 [ 1822.151156] slab_err+0xb1/0xd3 [ 1822.151834] ? __kmalloc+0x1bb/0x1ce [ 1822.152546] __kmem_cache_shutdown+0x151/0x28b [ 1822.153395] shutdown_cache+0x13/0x144 [ 1822.154126] kmem_cache_destroy+0x1c0/0x1fb [ 1822.154669] SyS_delete_module+0x194/0x244 [ 1822.155199] ? trace_hardirqs_on_thunk+0x1a/0x1c [ 1822.155773] entry_SYSCALL_64_fastpath+0x23/0x9a [ 1822.156343] RIP: 0033:0x7f929bd38b17 [ 1822.156859] RSP: 002b:00007ffd160e9a98 EFLAGS: 00000202 ORIG_RAX: 00000000000000b0 [ 1822.157728] RAX: ffffffffffffffda RBX: 00005578316ba090 RCX: 00007f929bd38b17 [ 1822.158422] RDX: 00007f929bd9ec60 RSI: 0000000000000800 RDI: 00005578316ba0f0 [ 1822.159114] RBP: 0000000000000003 R08: 00007f929bff5f20 R09: 00007ffd160e8a11 [ 1822.159808] R10: 00007ffd160e9860 R11: 0000000000000202 R12: 00007ffd160e8a80 [ 1822.160513] R13: 0000000000000000 R14: 0000000000000000 R15: 00005578316ba090 [ 1822.161278] INFO: Object 0x000000007645de29 @offset=0 [ 1822.161666] INFO: Object 0x00000000d5df2ab5 @offset=128 Fixes: 30313a3d5794 ("bridge: Handle IFLA_ADDRESS correctly when creating bridge device") Fixes: 5b8d5429daa0 ("bridge: netlink: register netdevice before executing changelink") Signed-off-by: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/bridge/br_netlink.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) --- a/net/bridge/br_netlink.c +++ b/net/bridge/br_netlink.c @@ -1067,19 +1067,20 @@ static int br_dev_newlink(struct net *sr struct net_bridge *br = netdev_priv(dev); int err; + err = register_netdevice(dev); + if (err) + return err; + if (tb[IFLA_ADDRESS]) { spin_lock_bh(&br->lock); br_stp_change_bridge_id(br, nla_data(tb[IFLA_ADDRESS])); spin_unlock_bh(&br->lock); } - err = register_netdevice(dev); - if (err) - return err; - err = br_changelink(dev, tb, data); if (err) - unregister_netdevice(dev); + br_dev_delete(dev, NULL); + return err; } Patches currently in stable-queue which might be from nikolay(a)cumulusnetworks.com are queue-4.4/net-bridge-fix-early-call-to-br_stp_change_bridge_id-and-plug-newlink-leaks.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv6: mcast: better catch silly mtu values" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv6: mcast: better catch silly mtu values to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv6-mcast-better-catch-silly-mtu-values.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:03:38 -0800 Subject: ipv6: mcast: better catch silly mtu values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b9b312a7a451e9c098921856e7cfbc201120e1a7 ] syzkaller reported crashes in IPv6 stack [1] Xin Long found that lo MTU was set to silly values. IPv6 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in mld code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv6 minimal MTU. [1] skbuff: skb_over_panic: text:0000000010b86b8d len:196 put:20 head:000000003b477e60 data:000000000e85441e tail:0xd4 end:0xc0 dev:lo ------------[ cut here ]------------ kernel BUG at net/core/skbuff.c:104! invalid opcode: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.15.0-rc2-mm1+ #39 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:skb_panic+0x15c/0x1f0 net/core/skbuff.c:100 RSP: 0018:ffff8801db307508 EFLAGS: 00010286 RAX: 0000000000000082 RBX: ffff8801c517e840 RCX: 0000000000000000 RDX: 0000000000000082 RSI: 1ffff1003b660e61 RDI: ffffed003b660e95 RBP: ffff8801db307570 R08: 1ffff1003b660e23 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffffffff85bd4020 R13: ffffffff84754ed2 R14: 0000000000000014 R15: ffff8801c4e26540 FS: 0000000000000000(0000) GS:ffff8801db300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000463610 CR3: 00000001c6698000 CR4: 00000000001406e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <IRQ> skb_over_panic net/core/skbuff.c:109 [inline] skb_put+0x181/0x1c0 net/core/skbuff.c:1694 add_grhead.isra.24+0x42/0x3b0 net/ipv6/mcast.c:1695 add_grec+0xa55/0x1060 net/ipv6/mcast.c:1817 mld_send_cr net/ipv6/mcast.c:1903 [inline] mld_ifc_timer_expire+0x4d2/0x770 net/ipv6/mcast.c:2448 call_timer_fn+0x23b/0x840 kernel/time/timer.c:1320 expire_timers kernel/time/timer.c:1357 [inline] __run_timers+0x7e1/0xb60 kernel/time/timer.c:1660 run_timer_softirq+0x4c/0xb0 kernel/time/timer.c:1686 __do_softirq+0x29d/0xbb2 kernel/softirq.c:285 invoke_softirq kernel/softirq.c:365 [inline] irq_exit+0x1d3/0x210 kernel/softirq.c:405 exiting_irq arch/x86/include/asm/apic.h:540 [inline] smp_apic_timer_interrupt+0x16b/0x700 arch/x86/kernel/apic/apic.c:1052 apic_timer_interrupt+0xa9/0xb0 arch/x86/entry/entry_64.S:920 Signed-off-by: Eric Dumazet <edumazet(a)google.com> Reported-by: syzbot <syzkaller(a)googlegroups.com> Tested-by: Xin Long <lucien.xin(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/mcast.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -1668,16 +1668,16 @@ static int grec_size(struct ifmcaddr6 *p } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ifmcaddr6 *pmc, - int type, struct mld2_grec **ppgr) + int type, struct mld2_grec **ppgr, unsigned int mtu) { - struct net_device *dev = pmc->idev->dev; struct mld2_report *pmr; struct mld2_grec *pgr; - if (!skb) - skb = mld_newpack(pmc->idev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = mld_newpack(pmc->idev, mtu); + if (!skb) + return NULL; + } pgr = (struct mld2_grec *)skb_put(skb, sizeof(struct mld2_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -1700,10 +1700,15 @@ static struct sk_buff *add_grec(struct s struct mld2_grec *pgr = NULL; struct ip6_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->mca_flags & MAF_NOREPORT) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV6_MIN_MTU) + return skb; + isquery = type == MLD2_MODE_IS_INCLUDE || type == MLD2_MODE_IS_EXCLUDE; truncate = type == MLD2_MODE_IS_EXCLUDE || @@ -1724,7 +1729,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); } } first = 1; @@ -1751,12 +1756,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -1790,7 +1795,7 @@ empty_source: mld_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.4/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.4/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.4/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.4/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.4/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.4/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: igmp: guard against silly MTU values" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: igmp: guard against silly MTU values to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-igmp-guard-against-silly-mtu-values.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:17:39 -0800 Subject: ipv4: igmp: guard against silly MTU values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b5476022bbada3764609368f03329ca287528dc8 ] IPv4 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in igmp code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv4 minimal MTU. This patch adds missing IPV4_MIN_MTU define, to not abuse ETH_MIN_MTU anymore. Signed-off-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/net/ip.h | 2 ++ net/ipv4/devinet.c | 2 +- net/ipv4/igmp.c | 24 +++++++++++++++--------- net/ipv4/ip_tunnel.c | 4 ++-- 4 files changed, 20 insertions(+), 12 deletions(-) --- a/include/net/ip.h +++ b/include/net/ip.h @@ -33,6 +33,8 @@ #include <net/flow.h> #include <net/flow_dissector.h> +#define IPV4_MIN_MTU 68 /* RFC 791 */ + struct sock; struct inet_skb_parm { --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -1358,7 +1358,7 @@ skip: static bool inetdev_valid_mtu(unsigned int mtu) { - return mtu >= 68; + return mtu >= IPV4_MIN_MTU; } static void inetdev_send_gratuitous_arp(struct net_device *dev, --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -410,16 +410,17 @@ static int grec_size(struct ip_mc_list * } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ip_mc_list *pmc, - int type, struct igmpv3_grec **ppgr) + int type, struct igmpv3_grec **ppgr, unsigned int mtu) { struct net_device *dev = pmc->interface->dev; struct igmpv3_report *pih; struct igmpv3_grec *pgr; - if (!skb) - skb = igmpv3_newpack(dev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = igmpv3_newpack(dev, mtu); + if (!skb) + return NULL; + } pgr = (struct igmpv3_grec *)skb_put(skb, sizeof(struct igmpv3_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -441,12 +442,17 @@ static struct sk_buff *add_grec(struct s struct igmpv3_grec *pgr = NULL; struct ip_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->multiaddr == IGMP_ALL_HOSTS) return skb; if (ipv4_is_local_multicast(pmc->multiaddr) && !sysctl_igmp_llm_reports) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV4_MIN_MTU) + return skb; + isquery = type == IGMPV3_MODE_IS_INCLUDE || type == IGMPV3_MODE_IS_EXCLUDE; truncate = type == IGMPV3_MODE_IS_EXCLUDE || @@ -467,7 +473,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); } } first = 1; @@ -494,12 +500,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -533,7 +539,7 @@ empty_source: igmpv3_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@ -400,8 +400,8 @@ static int ip_tunnel_bind_dev(struct net dev->needed_headroom = t_hlen + hlen; mtu -= (dev->hard_header_len + t_hlen); - if (mtu < 68) - mtu = 68; + if (mtu < IPV4_MIN_MTU) + mtu = IPV4_MIN_MTU; return mtu; } Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.4/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.4/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.4/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.4/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.4/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.4/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: Fix use-after-free when flushing FIB tables" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: Fix use-after-free when flushing FIB tables to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-fix-use-after-free-when-flushing-fib-tables.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:20:35 CET 2017 From: Ido Schimmel <idosch(a)mellanox.com> Date: Wed, 20 Dec 2017 19:34:19 +0200 Subject: ipv4: Fix use-after-free when flushing FIB tables From: Ido Schimmel <idosch(a)mellanox.com> [ Upstream commit b4681c2829e24943aadd1a7bb3a30d41d0a20050 ] Since commit 0ddcf43d5d4a ("ipv4: FIB Local/MAIN table collapse") the local table uses the same trie allocated for the main table when custom rules are not in use. When a net namespace is dismantled, the main table is flushed and freed (via an RCU callback) before the local table. In case the callback is invoked before the local table is iterated, a use-after-free can occur. Fix this by iterating over the FIB tables in reverse order, so that the main table is always freed after the local table. v3: Reworded comment according to Alex's suggestion. v2: Add a comment to make the fix more explicit per Dave's and Alex's feedback. Fixes: 0ddcf43d5d4a ("ipv4: FIB Local/MAIN table collapse") Signed-off-by: Ido Schimmel <idosch(a)mellanox.com> Reported-by: Fengguang Wu <fengguang.wu(a)intel.com> Acked-by: Alexander Duyck <alexander.h.duyck(a)intel.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/fib_frontend.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -1252,7 +1252,7 @@ fail: static void ip_fib_net_exit(struct net *net) { - unsigned int i; + int i; rtnl_lock(); #ifdef CONFIG_IP_MULTIPLE_TABLES @@ -1260,7 +1260,12 @@ static void ip_fib_net_exit(struct net * RCU_INIT_POINTER(net->ipv4.fib_main, NULL); RCU_INIT_POINTER(net->ipv4.fib_default, NULL); #endif - for (i = 0; i < FIB_TABLE_HASHSZ; i++) { + /* Destroy the tables in reverse order to guarantee that the + * local table, ID 255, is destroyed before the main table, ID + * 254. This is necessary as the local table may contain + * references to data contained in the main table. + */ + for (i = FIB_TABLE_HASHSZ - 1; i >= 0; i--) { struct hlist_head *head = &net->ipv4.fib_table_hash[i]; struct hlist_node *tmp; struct fib_table *tb; Patches currently in stable-queue which might be from idosch(a)mellanox.com are queue-4.4/ipv4-fix-use-after-free-when-flushing-fib-tables.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tg3: Fix rx hang on MTU change with 5717/5719" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tg3: Fix rx hang on MTU change with 5717/5719 to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Brian King <brking(a)linux.vnet.ibm.com> Date: Fri, 15 Dec 2017 15:21:50 -0600 Subject: tg3: Fix rx hang on MTU change with 5717/5719 From: Brian King <brking(a)linux.vnet.ibm.com> [ Upstream commit 748a240c589824e9121befb1cba5341c319885bc ] This fixes a hang issue seen when changing the MTU size from 1500 MTU to 9000 MTU on both 5717 and 5719 chips. In discussion with Broadcom, they've indicated that these chipsets have the same phy as the 57766 chipset, so the same workarounds apply. This has been tested by IBM on both Power 8 and Power 9 systems as well as by Broadcom on x86 hardware and has been confirmed to resolve the hang issue. Signed-off-by: Brian King <brking(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/broadcom/tg3.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -14202,7 +14202,9 @@ static int tg3_change_mtu(struct net_dev /* Reset PHY, otherwise the read DMA engine will be in a mode that * breaks all requests to 256 bytes. */ - if (tg3_asic_rev(tp) == ASIC_REV_57766) + if (tg3_asic_rev(tp) == ASIC_REV_57766 || + tg3_asic_rev(tp) == ASIC_REV_5717 || + tg3_asic_rev(tp) == ASIC_REV_5719) reset_phy = true; err = tg3_restart_hw(tp, reset_phy); Patches currently in stable-queue which might be from brking(a)linux.vnet.ibm.com are queue-3.18/tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp md5sig: Use skb's saddr when replying to an incoming segment" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp md5sig: Use skb's saddr when replying to an incoming segment to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Christoph Paasch <cpaasch(a)apple.com> Date: Mon, 11 Dec 2017 00:05:46 -0800 Subject: tcp md5sig: Use skb's saddr when replying to an incoming segment From: Christoph Paasch <cpaasch(a)apple.com> [ Upstream commit 30791ac41927ebd3e75486f9504b6d2280463bf0 ] The MD5-key that belongs to a connection is identified by the peer's IP-address. When we are in tcp_v4(6)_reqsk_send_ack(), we are replying to an incoming segment from tcp_check_req() that failed the seq-number checks. Thus, to find the correct key, we need to use the skb's saddr and not the daddr. This bug seems to have been there since quite a while, but probably got unnoticed because the consequences are not catastrophic. We will call tcp_v4_reqsk_send_ack only to send a challenge-ACK back to the peer, thus the connection doesn't really fail. Fixes: 9501f9722922 ("tcp md5sig: Let the caller pass appropriate key for tcp_v{4,6}_do_calc_md5_hash().") Signed-off-by: Christoph Paasch <cpaasch(a)apple.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_ipv4.c | 2 +- net/ipv6/tcp_ipv6.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -810,7 +810,7 @@ static void tcp_v4_reqsk_send_ack(struct tcp_time_stamp, req->ts_recent, 0, - tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->daddr, + tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->saddr, AF_INET), inet_rsk(req)->no_srccheck ? IP_REPLY_ARG_NOSRCCHECK : 0, ip_hdr(skb)->tos); --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1001,7 +1001,7 @@ static void tcp_v6_reqsk_send_ack(struct tcp_rsk(req)->snt_isn + 1 : tcp_sk(sk)->snd_nxt, tcp_rsk(req)->rcv_nxt, req->rcv_wnd, tcp_time_stamp, req->ts_recent, sk->sk_bound_dev_if, - tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr), + tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->saddr), 0, 0); } Patches currently in stable-queue which might be from cpaasch(a)apple.com are queue-3.18/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sctp: Replace use of sockets_allocated with specified macro." has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: Replace use of sockets_allocated with specified macro. to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-replace-use-of-sockets_allocated-with-specified-macro.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> Date: Fri, 22 Dec 2017 10:15:20 -0800 Subject: sctp: Replace use of sockets_allocated with specified macro. From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> [ Upstream commit 8cb38a602478e9f806571f6920b0a3298aabf042 ] The patch(180d8cd942ce) replaces all uses of struct sock fields' memory_pressure, memory_allocated, sockets_allocated, and sysctl_mem to accessor macros. But the sockets_allocated field of sctp sock is not replaced at all. Then replace it now for unifying the code. Fixes: 180d8cd942ce ("foundations of per-cgroup memory pressure controlling.") Cc: Glauber Costa <glommer(a)parallels.com> Signed-off-by: Tonghao Zhang <zhangtonghao(a)didichuxing.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/socket.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4166,7 +4166,7 @@ static int sctp_init_sock(struct sock *s SCTP_DBG_OBJCNT_INC(sock); local_bh_disable(); - percpu_counter_inc(&sctp_sockets_allocated); + sk_sockets_allocated_inc(sk); sock_prot_inuse_add(net, sk->sk_prot, 1); /* Nothing can fail after this block, otherwise @@ -4210,7 +4210,7 @@ static void sctp_destroy_sock(struct soc } sctp_endpoint_free(sp->ep); local_bh_disable(); - percpu_counter_dec(&sctp_sockets_allocated); + sk_sockets_allocated_dec(sk); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); local_bh_enable(); } Patches currently in stable-queue which might be from xiangxia.m.yue(a)gmail.com are queue-3.18/sctp-replace-use-of-sockets_allocated-with-specified-macro.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "netlink: Add netns check on taps" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled netlink: Add netns check on taps to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: netlink-add-netns-check-on-taps.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Wed, 6 Dec 2017 12:12:27 -0800 Subject: netlink: Add netns check on taps From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit 93c647643b48f0131f02e45da3bd367d80443291 ] Currently, a nlmon link inside a child namespace can observe systemwide netlink activity. Filter the traffic so that nlmon can only sniff netlink messages from its own netns. Test case: vpnns -- bash -c "ip link add nlmon0 type nlmon; \ ip link set nlmon0 up; \ tcpdump -i nlmon0 -q -w /tmp/nlmon.pcap -U" & sudo ip xfrm state add src 10.1.1.1 dst 10.1.1.2 proto esp \ spi 0x1 mode transport \ auth sha1 0x6162633132330000000000000000000000000000 \ enc aes 0x00000000000000000000000000000000 grep --binary abc123 /tmp/nlmon.pcap Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/netlink/af_netlink.c | 3 +++ 1 file changed, 3 insertions(+) --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -270,6 +270,9 @@ static int __netlink_deliver_tap_skb(str struct sock *sk = skb->sk; int ret = -ENOMEM; + if (!net_eq(dev_net(dev), sock_net(sk))) + return 0; + dev_hold(dev); if (is_vmalloc_addr(skb->head)) Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-3.18/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-3.18/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Date: Wed, 6 Dec 2017 15:23:23 +0100 Subject: net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> [ Upstream commit 589bf32f09852041fbd3b7ce1a9e703f95c230ba ] add appropriate calls to clk_disable_unprepare() by jumping to out_mdio in case orion_mdio_probe() returns -EPROBE_DEFER. Found by Linux Driver Verification project (linuxtesting.org). Fixes: 3d604da1e954 ("net: mvmdio: get and enable optional clock") Signed-off-by: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/marvell/mvmdio.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/marvell/mvmdio.c +++ b/drivers/net/ethernet/marvell/mvmdio.c @@ -241,7 +241,8 @@ static int orion_mdio_probe(struct platf dev->regs + MVMDIO_ERR_INT_MASK); } else if (dev->err_interrupt == -EPROBE_DEFER) { - return -EPROBE_DEFER; + ret = -EPROBE_DEFER; + goto out_mdio; } mutex_init(&dev->lock); Patches currently in stable-queue which might be from Tobias.Jordan(a)elektrobit.com are queue-3.18/net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: qmi_wwan: add Sierra EM7565 1199:9091" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: qmi_wwan: add Sierra EM7565 1199:9091 to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-qmi_wwan-add-sierra-em7565-1199-9091.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Sebastian Sjoholm <ssjoholm(a)mac.com> Date: Mon, 11 Dec 2017 21:51:14 +0100 Subject: net: qmi_wwan: add Sierra EM7565 1199:9091 From: Sebastian Sjoholm <ssjoholm(a)mac.com> [ Upstream commit aceef61ee56898cfa7b6960fb60b9326c3860441 ] Sierra Wireless EM7565 is an Qualcomm MDM9x50 based M.2 modem. The USB id is added to qmi_wwan.c to allow QMI communication with the EM7565. Signed-off-by: Sebastian Sjoholm <ssjoholm(a)mac.com> Acked-by: Bjørn Mork <bjorn(a)mork.no> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -778,6 +778,7 @@ static const struct usb_device_id produc {QMI_FIXED_INTF(0x1199, 0x9079, 10)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 8)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 10)}, /* Sierra Wireless EM74xx */ + {QMI_FIXED_INTF(0x1199, 0x9091, 8)}, /* Sierra Wireless EM7565 */ {QMI_FIXED_INTF(0x1bbb, 0x011e, 4)}, /* Telekom Speedstick LTE II (Alcatel One Touch L100V LTE) */ {QMI_FIXED_INTF(0x1bbb, 0x0203, 2)}, /* Alcatel L800MA */ {QMI_FIXED_INTF(0x2357, 0x0201, 4)}, /* TP-LINK HSUPA Modem MA180 */ Patches currently in stable-queue which might be from ssjoholm(a)mac.com are queue-3.18/net-qmi_wwan-add-sierra-em7565-1199-9091.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: igmp: Use correct source address on IGMPv3 reports" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: igmp: Use correct source address on IGMPv3 reports to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-igmp-use-correct-source-address-on-igmpv3-reports.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Mon, 11 Dec 2017 11:13:45 -0800 Subject: net: igmp: Use correct source address on IGMPv3 reports From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit a46182b00290839fa3fa159d54fd3237bd8669f0 ] Closing a multicast socket after the final IPv4 address is deleted from an interface can generate a membership report that uses the source IP from a different interface. The following test script, run from an isolated netns, reproduces the issue: #!/bin/bash ip link add dummy0 type dummy ip link add dummy1 type dummy ip link set dummy0 up ip link set dummy1 up ip addr add 10.1.1.1/24 dev dummy0 ip addr add 192.168.99.99/24 dev dummy1 tcpdump -U -i dummy0 & socat EXEC:"sleep 2" \ UDP4-DATAGRAM:239.101.1.68:8889,ip-add-membership=239.0.1.68:10.1.1.1 & sleep 1 ip addr del 10.1.1.1/24 dev dummy0 sleep 5 kill %tcpdump RFC 3376 specifies that the report must be sent with a valid IP source address from the destination subnet, or from address 0.0.0.0. Add an extra check to make sure this is the case. Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/igmp.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -89,6 +89,7 @@ #include <linux/rtnetlink.h> #include <linux/times.h> #include <linux/pkt_sched.h> +#include <linux/byteorder/generic.h> #include <net/net_namespace.h> #include <net/arp.h> @@ -318,6 +319,23 @@ igmp_scount(struct ip_mc_list *pmc, int return scount; } +/* source address selection per RFC 3376 section 4.2.13 */ +static __be32 igmpv3_get_srcaddr(struct net_device *dev, + const struct flowi4 *fl4) +{ + struct in_device *in_dev = __in_dev_get_rcu(dev); + + if (!in_dev) + return htonl(INADDR_ANY); + + for_ifa(in_dev) { + if (inet_ifa_match(fl4->saddr, ifa)) + return fl4->saddr; + } endfor_ifa(in_dev); + + return htonl(INADDR_ANY); +} + static struct sk_buff *igmpv3_newpack(struct net_device *dev, unsigned int mtu) { struct sk_buff *skb; @@ -366,7 +384,7 @@ static struct sk_buff *igmpv3_newpack(st pip->frag_off = htons(IP_DF); pip->ttl = 1; pip->daddr = fl4.daddr; - pip->saddr = fl4.saddr; + pip->saddr = igmpv3_get_srcaddr(dev, &fl4); pip->protocol = IPPROTO_IGMP; pip->tot_len = 0; /* filled in later */ ip_select_ident(skb, NULL); Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-3.18/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-3.18/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: ipv4: fix for a race condition in raw_sendmsg" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: ipv4: fix for a race condition in raw_sendmsg to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Mohamed Ghannam <simo.ghannam(a)gmail.com> Date: Sun, 10 Dec 2017 03:50:58 +0000 Subject: net: ipv4: fix for a race condition in raw_sendmsg From: Mohamed Ghannam <simo.ghannam(a)gmail.com> [ Upstream commit 8f659a03a0ba9289b9aeb9b4470e6fb263d6f483 ] inet->hdrincl is racy, and could lead to uninitialized stack pointer usage, so its value should be read only once. Fixes: c008ba5bdc9f ("ipv4: Avoid reading user iov twice after raw_probe_proto_opt") Signed-off-by: Mohamed Ghannam <simo.ghannam(a)gmail.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/raw.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -483,11 +483,16 @@ static int raw_sendmsg(struct kiocb *ioc u8 tos; int err; struct ip_options_data opt_copy; + int hdrincl; err = -EMSGSIZE; if (len > 0xFFFF) goto out; + /* hdrincl should be READ_ONCE(inet->hdrincl) + * but READ_ONCE() doesn't work with bit fields + */ + hdrincl = inet->hdrincl; /* * Check the flags. */ @@ -560,7 +565,7 @@ static int raw_sendmsg(struct kiocb *ioc /* Linux does not mangle headers on raw sockets, * so that IP options + IP_HDRINCL is non-sense. */ - if (inet->hdrincl) + if (hdrincl) goto done; if (ipc.opt->opt.srr) { if (!daddr) @@ -582,12 +587,12 @@ static int raw_sendmsg(struct kiocb *ioc flowi4_init_output(&fl4, ipc.oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE, - inet->hdrincl ? IPPROTO_RAW : sk->sk_protocol, + hdrincl ? IPPROTO_RAW : sk->sk_protocol, inet_sk_flowi_flags(sk) | - (inet->hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), + (hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), daddr, saddr, 0, 0); - if (!inet->hdrincl) { + if (!hdrincl) { err = raw_probe_proto_opt(&fl4, msg); if (err) goto done; @@ -609,7 +614,7 @@ static int raw_sendmsg(struct kiocb *ioc goto do_confirm; back_from_confirm: - if (inet->hdrincl) + if (hdrincl) err = raw_send_hdrinc(sk, &fl4, msg->msg_iov, len, &rt, msg->msg_flags); Patches currently in stable-queue which might be from simo.ghannam(a)gmail.com are queue-3.18/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv6: mcast: better catch silly mtu values" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv6: mcast: better catch silly mtu values to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv6-mcast-better-catch-silly-mtu-values.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:03:38 -0800 Subject: ipv6: mcast: better catch silly mtu values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b9b312a7a451e9c098921856e7cfbc201120e1a7 ] syzkaller reported crashes in IPv6 stack [1] Xin Long found that lo MTU was set to silly values. IPv6 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in mld code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv6 minimal MTU. [1] skbuff: skb_over_panic: text:0000000010b86b8d len:196 put:20 head:000000003b477e60 data:000000000e85441e tail:0xd4 end:0xc0 dev:lo ------------[ cut here ]------------ kernel BUG at net/core/skbuff.c:104! invalid opcode: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.15.0-rc2-mm1+ #39 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:skb_panic+0x15c/0x1f0 net/core/skbuff.c:100 RSP: 0018:ffff8801db307508 EFLAGS: 00010286 RAX: 0000000000000082 RBX: ffff8801c517e840 RCX: 0000000000000000 RDX: 0000000000000082 RSI: 1ffff1003b660e61 RDI: ffffed003b660e95 RBP: ffff8801db307570 R08: 1ffff1003b660e23 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffffffff85bd4020 R13: ffffffff84754ed2 R14: 0000000000000014 R15: ffff8801c4e26540 FS: 0000000000000000(0000) GS:ffff8801db300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000463610 CR3: 00000001c6698000 CR4: 00000000001406e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <IRQ> skb_over_panic net/core/skbuff.c:109 [inline] skb_put+0x181/0x1c0 net/core/skbuff.c:1694 add_grhead.isra.24+0x42/0x3b0 net/ipv6/mcast.c:1695 add_grec+0xa55/0x1060 net/ipv6/mcast.c:1817 mld_send_cr net/ipv6/mcast.c:1903 [inline] mld_ifc_timer_expire+0x4d2/0x770 net/ipv6/mcast.c:2448 call_timer_fn+0x23b/0x840 kernel/time/timer.c:1320 expire_timers kernel/time/timer.c:1357 [inline] __run_timers+0x7e1/0xb60 kernel/time/timer.c:1660 run_timer_softirq+0x4c/0xb0 kernel/time/timer.c:1686 __do_softirq+0x29d/0xbb2 kernel/softirq.c:285 invoke_softirq kernel/softirq.c:365 [inline] irq_exit+0x1d3/0x210 kernel/softirq.c:405 exiting_irq arch/x86/include/asm/apic.h:540 [inline] smp_apic_timer_interrupt+0x16b/0x700 arch/x86/kernel/apic/apic.c:1052 apic_timer_interrupt+0xa9/0xb0 arch/x86/entry/entry_64.S:920 Signed-off-by: Eric Dumazet <edumazet(a)google.com> Reported-by: syzbot <syzkaller(a)googlegroups.com> Tested-by: Xin Long <lucien.xin(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/mcast.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -1670,16 +1670,16 @@ static int grec_size(struct ifmcaddr6 *p } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ifmcaddr6 *pmc, - int type, struct mld2_grec **ppgr) + int type, struct mld2_grec **ppgr, unsigned int mtu) { - struct net_device *dev = pmc->idev->dev; struct mld2_report *pmr; struct mld2_grec *pgr; - if (!skb) - skb = mld_newpack(pmc->idev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = mld_newpack(pmc->idev, mtu); + if (!skb) + return NULL; + } pgr = (struct mld2_grec *)skb_put(skb, sizeof(struct mld2_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -1702,10 +1702,15 @@ static struct sk_buff *add_grec(struct s struct mld2_grec *pgr = NULL; struct ip6_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->mca_flags & MAF_NOREPORT) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV6_MIN_MTU) + return skb; + isquery = type == MLD2_MODE_IS_INCLUDE || type == MLD2_MODE_IS_EXCLUDE; truncate = type == MLD2_MODE_IS_EXCLUDE || @@ -1726,7 +1731,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); } } first = 1; @@ -1753,12 +1758,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -1792,7 +1797,7 @@ empty_source: mld_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) Patches currently in stable-queue which might be from edumazet(a)google.com are queue-3.18/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-3.18/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-3.18/ipv4-igmp-guard-against-silly-mtu-values.patch queue-3.18/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: igmp: guard against silly MTU values" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: igmp: guard against silly MTU values to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-igmp-guard-against-silly-mtu-values.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:31:10 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:17:39 -0800 Subject: ipv4: igmp: guard against silly MTU values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b5476022bbada3764609368f03329ca287528dc8 ] IPv4 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in igmp code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv4 minimal MTU. This patch adds missing IPV4_MIN_MTU define, to not abuse ETH_MIN_MTU anymore. Signed-off-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/net/ip.h | 2 ++ net/ipv4/devinet.c | 2 +- net/ipv4/igmp.c | 24 +++++++++++++++--------- net/ipv4/ip_tunnel.c | 4 ++-- 4 files changed, 20 insertions(+), 12 deletions(-) --- a/include/net/ip.h +++ b/include/net/ip.h @@ -33,6 +33,8 @@ #include <net/flow.h> #include <net/flow_keys.h> +#define IPV4_MIN_MTU 68 /* RFC 791 */ + struct sock; struct inet_skb_parm { --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -1328,7 +1328,7 @@ skip: static bool inetdev_valid_mtu(unsigned int mtu) { - return mtu >= 68; + return mtu >= IPV4_MIN_MTU; } static void inetdev_send_gratuitous_arp(struct net_device *dev, --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -402,16 +402,17 @@ static int grec_size(struct ip_mc_list * } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ip_mc_list *pmc, - int type, struct igmpv3_grec **ppgr) + int type, struct igmpv3_grec **ppgr, unsigned int mtu) { struct net_device *dev = pmc->interface->dev; struct igmpv3_report *pih; struct igmpv3_grec *pgr; - if (!skb) - skb = igmpv3_newpack(dev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = igmpv3_newpack(dev, mtu); + if (!skb) + return NULL; + } pgr = (struct igmpv3_grec *)skb_put(skb, sizeof(struct igmpv3_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -433,10 +434,15 @@ static struct sk_buff *add_grec(struct s struct igmpv3_grec *pgr = NULL; struct ip_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->multiaddr == IGMP_ALL_HOSTS) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV4_MIN_MTU) + return skb; + isquery = type == IGMPV3_MODE_IS_INCLUDE || type == IGMPV3_MODE_IS_EXCLUDE; truncate = type == IGMPV3_MODE_IS_EXCLUDE || @@ -457,7 +463,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); } } first = 1; @@ -484,12 +490,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -523,7 +529,7 @@ empty_source: igmpv3_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@ -395,8 +395,8 @@ static int ip_tunnel_bind_dev(struct net dev->needed_headroom = t_hlen + hlen; mtu -= (dev->hard_header_len + t_hlen); - if (mtu < 68) - mtu = 68; + if (mtu < IPV4_MIN_MTU) + mtu = IPV4_MIN_MTU; return mtu; } Patches currently in stable-queue which might be from edumazet(a)google.com are queue-3.18/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-3.18/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-3.18/ipv4-igmp-guard-against-silly-mtu-values.patch queue-3.18/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tg3: Fix rx hang on MTU change with 5717/5719" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tg3: Fix rx hang on MTU change with 5717/5719 to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Brian King <brking(a)linux.vnet.ibm.com> Date: Fri, 15 Dec 2017 15:21:50 -0600 Subject: tg3: Fix rx hang on MTU change with 5717/5719 From: Brian King <brking(a)linux.vnet.ibm.com> [ Upstream commit 748a240c589824e9121befb1cba5341c319885bc ] This fixes a hang issue seen when changing the MTU size from 1500 MTU to 9000 MTU on both 5717 and 5719 chips. In discussion with Broadcom, they've indicated that these chipsets have the same phy as the 57766 chipset, so the same workarounds apply. This has been tested by IBM on both Power 8 and Power 9 systems as well as by Broadcom on x86 hardware and has been confirmed to resolve the hang issue. Signed-off-by: Brian King <brking(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/broadcom/tg3.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -14226,7 +14226,9 @@ static int tg3_change_mtu(struct net_dev /* Reset PHY, otherwise the read DMA engine will be in a mode that * breaks all requests to 256 bytes. */ - if (tg3_asic_rev(tp) == ASIC_REV_57766) + if (tg3_asic_rev(tp) == ASIC_REV_57766 || + tg3_asic_rev(tp) == ASIC_REV_5717 || + tg3_asic_rev(tp) == ASIC_REV_5719) reset_phy = true; err = tg3_restart_hw(tp, reset_phy); Patches currently in stable-queue which might be from brking(a)linux.vnet.ibm.com are queue-4.9/tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp_bbr: record "full bw reached" decision in new full_bw_reached bit" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp_bbr: record "full bw reached" decision in new full_bw_reached bit to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp_bbr-record-full-bw-reached-decision-in-new-full_bw_reached-bit.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Neal Cardwell <ncardwell(a)google.com> Date: Thu, 7 Dec 2017 12:43:30 -0500 Subject: tcp_bbr: record "full bw reached" decision in new full_bw_reached bit From: Neal Cardwell <ncardwell(a)google.com> [ Upstream commit c589e69b508d29ed8e644dfecda453f71c02ec27 ] This commit records the "full bw reached" decision in a new full_bw_reached bit. This is a pure refactor that does not change the current behavior, but enables subsequent fixes and improvements. In particular, this enables simple and clean fixes because the full_bw and full_bw_cnt can be unconditionally zeroed without worrying about forgetting that we estimated we filled the pipe in Startup. And it enables future improvements because multiple code paths can be used for estimating that we filled the pipe in Startup; any new code paths only need to set this bit when they think the pipe is full. Note that this fix intentionally reduces the width of the full_bw_cnt counter, since we have never used the most significant bit. Signed-off-by: Neal Cardwell <ncardwell(a)google.com> Reviewed-by: Yuchung Cheng <ycheng(a)google.com> Acked-by: Soheil Hassas Yeganeh <soheil(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_bbr.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) --- a/net/ipv4/tcp_bbr.c +++ b/net/ipv4/tcp_bbr.c @@ -81,7 +81,8 @@ struct bbr { u32 lt_last_lost; /* LT intvl start: tp->lost */ u32 pacing_gain:10, /* current gain for setting pacing rate */ cwnd_gain:10, /* current gain for setting cwnd */ - full_bw_cnt:3, /* number of rounds without large bw gains */ + full_bw_reached:1, /* reached full bw in Startup? */ + full_bw_cnt:2, /* number of rounds without large bw gains */ cycle_idx:3, /* current index in pacing_gain cycle array */ has_seen_rtt:1, /* have we seen an RTT sample yet? */ unused_b:5; @@ -151,7 +152,7 @@ static bool bbr_full_bw_reached(const st { const struct bbr *bbr = inet_csk_ca(sk); - return bbr->full_bw_cnt >= bbr_full_bw_cnt; + return bbr->full_bw_reached; } /* Return the windowed max recent bandwidth sample, in pkts/uS << BW_SCALE. */ @@ -688,6 +689,7 @@ static void bbr_check_full_bw_reached(st return; } ++bbr->full_bw_cnt; + bbr->full_bw_reached = bbr->full_bw_cnt >= bbr_full_bw_cnt; } /* If pipe is probably full, drain the queue and then enter steady-state. */ @@ -821,6 +823,7 @@ static void bbr_init(struct sock *sk) bbr->restore_cwnd = 0; bbr->round_start = 0; bbr->idle_restart = 0; + bbr->full_bw_reached = 0; bbr->full_bw = 0; bbr->full_bw_cnt = 0; bbr->cycle_mstamp.v64 = 0; Patches currently in stable-queue which might be from ncardwell(a)google.com are queue-4.9/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.9/tcp_bbr-record-full-bw-reached-decision-in-new-full_bw_reached-bit.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp md5sig: Use skb's saddr when replying to an incoming segment" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp md5sig: Use skb's saddr when replying to an incoming segment to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Christoph Paasch <cpaasch(a)apple.com> Date: Mon, 11 Dec 2017 00:05:46 -0800 Subject: tcp md5sig: Use skb's saddr when replying to an incoming segment From: Christoph Paasch <cpaasch(a)apple.com> [ Upstream commit 30791ac41927ebd3e75486f9504b6d2280463bf0 ] The MD5-key that belongs to a connection is identified by the peer's IP-address. When we are in tcp_v4(6)_reqsk_send_ack(), we are replying to an incoming segment from tcp_check_req() that failed the seq-number checks. Thus, to find the correct key, we need to use the skb's saddr and not the daddr. This bug seems to have been there since quite a while, but probably got unnoticed because the consequences are not catastrophic. We will call tcp_v4_reqsk_send_ack only to send a challenge-ACK back to the peer, thus the connection doesn't really fail. Fixes: 9501f9722922 ("tcp md5sig: Let the caller pass appropriate key for tcp_v{4,6}_do_calc_md5_hash().") Signed-off-by: Christoph Paasch <cpaasch(a)apple.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_ipv4.c | 2 +- net/ipv6/tcp_ipv6.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -828,7 +828,7 @@ static void tcp_v4_reqsk_send_ack(const tcp_time_stamp, req->ts_recent, 0, - tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->daddr, + tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->saddr, AF_INET), inet_rsk(req)->no_srccheck ? IP_REPLY_ARG_NOSRCCHECK : 0, ip_hdr(skb)->tos); --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -962,7 +962,7 @@ static void tcp_v6_reqsk_send_ack(const tcp_rsk(req)->rcv_nxt, req->rsk_rcv_wnd >> inet_rsk(req)->rcv_wscale, tcp_time_stamp, req->ts_recent, sk->sk_bound_dev_if, - tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr), + tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->saddr), 0, 0); } Patches currently in stable-queue which might be from cpaasch(a)apple.com are queue-4.9/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp: invalidate rate samples during SACK reneging" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp: invalidate rate samples during SACK reneging to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-invalidate-rate-samples-during-sack-reneging.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Yousuk Seung <ysseung(a)google.com> Date: Thu, 7 Dec 2017 13:41:34 -0800 Subject: tcp: invalidate rate samples during SACK reneging From: Yousuk Seung <ysseung(a)google.com> [ Upstream commit d4761754b4fb2ef8d9a1e9d121c4bec84e1fe292 ] Mark tcp_sock during a SACK reneging event and invalidate rate samples while marked. Such rate samples may overestimate bw by including packets that were SACKed before reneging. < ack 6001 win 10000 sack 7001:38001 < ack 7001 win 0 sack 8001:38001 // Reneg detected > seq 7001:8001 // RTO, SACK cleared. < ack 38001 win 10000 In above example the rate sample taken after the last ack will count 7001-38001 as delivered while the actual delivery rate likely could be much lower i.e. 7001-8001. This patch adds a new field tcp_sock.sack_reneg and marks it when we declare SACK reneging and entering TCP_CA_Loss, and unmarks it after the last rate sample was taken before moving back to TCP_CA_Open. This patch also invalidates rate samples taken while tcp_sock.is_sack_reneg is set. Fixes: b9f64820fb22 ("tcp: track data delivery rate for a TCP connection") Signed-off-by: Yousuk Seung <ysseung(a)google.com> Signed-off-by: Neal Cardwell <ncardwell(a)google.com> Signed-off-by: Yuchung Cheng <ycheng(a)google.com> Acked-by: Soheil Hassas Yeganeh <soheil(a)google.com> Acked-by: Eric Dumazet <edumazet(a)google.com> Acked-by: Priyaranjan Jha <priyarjha(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/tcp.h | 3 ++- include/net/tcp.h | 2 +- net/ipv4/tcp.c | 1 + net/ipv4/tcp_input.c | 10 ++++++++-- net/ipv4/tcp_rate.c | 10 +++++++--- 5 files changed, 19 insertions(+), 7 deletions(-) --- a/include/linux/tcp.h +++ b/include/linux/tcp.h @@ -219,7 +219,8 @@ struct tcp_sock { } rack; u16 advmss; /* Advertised MSS */ u8 rate_app_limited:1, /* rate_{delivered,interval_us} limited? */ - unused:7; + is_sack_reneg:1, /* in recovery from loss with SACK reneg? */ + unused:6; u8 nonagle : 4,/* Disable Nagle algorithm? */ thin_lto : 1,/* Use linear timeouts for thin streams */ thin_dupack : 1,/* Fast retransmit on first dupack */ --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1001,7 +1001,7 @@ void tcp_rate_skb_sent(struct sock *sk, void tcp_rate_skb_delivered(struct sock *sk, struct sk_buff *skb, struct rate_sample *rs); void tcp_rate_gen(struct sock *sk, u32 delivered, u32 lost, - struct skb_mstamp *now, struct rate_sample *rs); + bool is_sack_reneg, struct skb_mstamp *now, struct rate_sample *rs); void tcp_rate_check_app_limited(struct sock *sk); /* These functions determine how the current flow behaves in respect of SACK --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2297,6 +2297,7 @@ int tcp_disconnect(struct sock *sk, int tp->snd_cwnd_cnt = 0; tp->window_clamp = 0; tcp_set_ca_state(sk, TCP_CA_Open); + tp->is_sack_reneg = 0; tcp_clear_retrans(tp); inet_csk_delack_init(sk); /* Initialize rcv_mss to TCP_MIN_MSS to avoid division by 0 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1966,6 +1966,8 @@ void tcp_enter_loss(struct sock *sk) NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSACKRENEGING); tp->sacked_out = 0; tp->fackets_out = 0; + /* Mark SACK reneging until we recover from this loss event. */ + tp->is_sack_reneg = 1; } tcp_clear_all_retrans_hints(tp); @@ -2463,6 +2465,7 @@ static bool tcp_try_undo_recovery(struct return true; } tcp_set_ca_state(sk, TCP_CA_Open); + tp->is_sack_reneg = 0; return false; } @@ -2494,8 +2497,10 @@ static bool tcp_try_undo_loss(struct soc NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSPURIOUSRTOS); inet_csk(sk)->icsk_retransmits = 0; - if (frto_undo || tcp_is_sack(tp)) + if (frto_undo || tcp_is_sack(tp)) { tcp_set_ca_state(sk, TCP_CA_Open); + tp->is_sack_reneg = 0; + } return true; } return false; @@ -3589,6 +3594,7 @@ static int tcp_ack(struct sock *sk, cons struct tcp_sacktag_state sack_state; struct rate_sample rs = { .prior_delivered = 0 }; u32 prior_snd_una = tp->snd_una; + bool is_sack_reneg = tp->is_sack_reneg; u32 ack_seq = TCP_SKB_CB(skb)->seq; u32 ack = TCP_SKB_CB(skb)->ack_seq; bool is_dupack = false; @@ -3711,7 +3717,7 @@ static int tcp_ack(struct sock *sk, cons tcp_schedule_loss_probe(sk); delivered = tp->delivered - delivered; /* freshly ACKed or SACKed */ lost = tp->lost - lost; /* freshly marked lost */ - tcp_rate_gen(sk, delivered, lost, &now, &rs); + tcp_rate_gen(sk, delivered, lost, is_sack_reneg, &now, &rs); tcp_cong_control(sk, ack, delivered, flag, &rs); tcp_xmit_recovery(sk, rexmit); return 1; --- a/net/ipv4/tcp_rate.c +++ b/net/ipv4/tcp_rate.c @@ -106,7 +106,7 @@ void tcp_rate_skb_delivered(struct sock /* Update the connection delivery information and generate a rate sample. */ void tcp_rate_gen(struct sock *sk, u32 delivered, u32 lost, - struct skb_mstamp *now, struct rate_sample *rs) + bool is_sack_reneg, struct skb_mstamp *now, struct rate_sample *rs) { struct tcp_sock *tp = tcp_sk(sk); u32 snd_us, ack_us; @@ -124,8 +124,12 @@ void tcp_rate_gen(struct sock *sk, u32 d rs->acked_sacked = delivered; /* freshly ACKed or SACKed */ rs->losses = lost; /* freshly marked lost */ - /* Return an invalid sample if no timing information is available. */ - if (!rs->prior_mstamp.v64) { + /* Return an invalid sample if no timing information is available or + * in recovery from loss with SACK reneging. Rate samples taken during + * a SACK reneging event may overestimate bw by including packets that + * were SACKed before the reneg. + */ + if (!rs->prior_mstamp.v64 || is_sack_reneg) { rs->delivered = -1; rs->interval_us = -1; return; Patches currently in stable-queue which might be from ysseung(a)google.com are queue-4.9/tcp-invalidate-rate-samples-during-sack-reneging.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sock: free skb in skb_complete_tx_timestamp on error" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sock: free skb in skb_complete_tx_timestamp on error to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Willem de Bruijn <willemb(a)google.com> Date: Wed, 13 Dec 2017 14:41:06 -0500 Subject: sock: free skb in skb_complete_tx_timestamp on error From: Willem de Bruijn <willemb(a)google.com> [ Upstream commit 35b99dffc3f710cafceee6c8c6ac6a98eb2cb4bf ] skb_complete_tx_timestamp must ingest the skb it is passed. Call kfree_skb if the skb cannot be enqueued. Fixes: b245be1f4db1 ("net-timestamp: no-payload only sysctl") Fixes: 9ac25fc06375 ("net: fix socket refcounting in skb_complete_tx_timestamp()") Reported-by: Richard Cochran <richardcochran(a)gmail.com> Signed-off-by: Willem de Bruijn <willemb(a)google.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/skbuff.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -3823,7 +3823,7 @@ void skb_complete_tx_timestamp(struct sk struct sock *sk = skb->sk; if (!skb_may_tx_timestamp(sk, false)) - return; + goto err; /* Take a reference to prevent skb_orphan() from freeing the socket, * but only if the socket refcount is not zero. @@ -3832,7 +3832,11 @@ void skb_complete_tx_timestamp(struct sk *skb_hwtstamps(skb) = *hwtstamps; __skb_complete_tx_timestamp(skb, sk, SCM_TSTAMP_SND); sock_put(sk); + return; } + +err: + kfree_skb(skb); } EXPORT_SYMBOL_GPL(skb_complete_tx_timestamp); Patches currently in stable-queue which might be from willemb(a)google.com are queue-4.9/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sctp: Replace use of sockets_allocated with specified macro." has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: Replace use of sockets_allocated with specified macro. to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-replace-use-of-sockets_allocated-with-specified-macro.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> Date: Fri, 22 Dec 2017 10:15:20 -0800 Subject: sctp: Replace use of sockets_allocated with specified macro. From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> [ Upstream commit 8cb38a602478e9f806571f6920b0a3298aabf042 ] The patch(180d8cd942ce) replaces all uses of struct sock fields' memory_pressure, memory_allocated, sockets_allocated, and sysctl_mem to accessor macros. But the sockets_allocated field of sctp sock is not replaced at all. Then replace it now for unifying the code. Fixes: 180d8cd942ce ("foundations of per-cgroup memory pressure controlling.") Cc: Glauber Costa <glommer(a)parallels.com> Signed-off-by: Tonghao Zhang <zhangtonghao(a)didichuxing.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/socket.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4246,7 +4246,7 @@ static int sctp_init_sock(struct sock *s SCTP_DBG_OBJCNT_INC(sock); local_bh_disable(); - percpu_counter_inc(&sctp_sockets_allocated); + sk_sockets_allocated_inc(sk); sock_prot_inuse_add(net, sk->sk_prot, 1); /* Nothing can fail after this block, otherwise @@ -4290,7 +4290,7 @@ static void sctp_destroy_sock(struct soc } sctp_endpoint_free(sp->ep); local_bh_disable(); - percpu_counter_dec(&sctp_sockets_allocated); + sk_sockets_allocated_dec(sk); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); local_bh_enable(); } Patches currently in stable-queue which might be from xiangxia.m.yue(a)gmail.com are queue-4.9/sctp-replace-use-of-sockets_allocated-with-specified-macro.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: update takeover IPs after configuration change" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: update takeover IPs after configuration change to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-update-takeover-ips-after-configuration-change.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:32 +0100 Subject: s390/qeth: update takeover IPs after configuration change From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit 02f510f326501470348a5df341e8232c3497bbbb ] Any modification to the takeover IP-ranges requires that we re-evaluate which IP addresses are takeover-eligible. Otherwise we might do takeover for some addresses when we no longer should, or vice-versa. Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_core.h | 4 +- drivers/s390/net/qeth_core_main.c | 4 +- drivers/s390/net/qeth_l3.h | 2 - drivers/s390/net/qeth_l3_main.c | 31 ++++++++++++++++-- drivers/s390/net/qeth_l3_sys.c | 63 ++++++++++++++++++++------------------ 5 files changed, 67 insertions(+), 37 deletions(-) --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -577,8 +577,8 @@ enum qeth_cq { struct qeth_ipato { bool enabled; - int invert4; - int invert6; + bool invert4; + bool invert6; struct list_head entries; }; --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -1476,8 +1476,8 @@ static int qeth_setup_card(struct qeth_c /* IP address takeover */ INIT_LIST_HEAD(&card->ipato.entries); card->ipato.enabled = false; - card->ipato.invert4 = 0; - card->ipato.invert6 = 0; + card->ipato.invert4 = false; + card->ipato.invert6 = false; /* init QDIO stuff */ qeth_init_qdio_info(card); INIT_DELAYED_WORK(&card->buffer_reclaim_work, qeth_buffer_reclaim_work); --- a/drivers/s390/net/qeth_l3.h +++ b/drivers/s390/net/qeth_l3.h @@ -80,7 +80,7 @@ void qeth_l3_del_vipa(struct qeth_card * int qeth_l3_add_rxip(struct qeth_card *, enum qeth_prot_versions, const u8 *); void qeth_l3_del_rxip(struct qeth_card *card, enum qeth_prot_versions, const u8 *); -int qeth_l3_is_addr_covered_by_ipato(struct qeth_card *, struct qeth_ipaddr *); +void qeth_l3_update_ipato(struct qeth_card *card); struct qeth_ipaddr *qeth_l3_get_addr_buffer(enum qeth_prot_versions); int qeth_l3_add_ip(struct qeth_card *, struct qeth_ipaddr *); int qeth_l3_delete_ip(struct qeth_card *, struct qeth_ipaddr *); --- a/drivers/s390/net/qeth_l3_main.c +++ b/drivers/s390/net/qeth_l3_main.c @@ -168,8 +168,8 @@ static void qeth_l3_convert_addr_to_bits } } -int qeth_l3_is_addr_covered_by_ipato(struct qeth_card *card, - struct qeth_ipaddr *addr) +static bool qeth_l3_is_addr_covered_by_ipato(struct qeth_card *card, + struct qeth_ipaddr *addr) { struct qeth_ipato_entry *ipatoe; u8 addr_bits[128] = {0, }; @@ -608,6 +608,27 @@ int qeth_l3_setrouting_v6(struct qeth_ca /* * IP address takeover related functions */ + +/** + * qeth_l3_update_ipato() - Update 'takeover' property, for all NORMAL IPs. + * + * Caller must hold ip_lock. + */ +void qeth_l3_update_ipato(struct qeth_card *card) +{ + struct qeth_ipaddr *addr; + unsigned int i; + + hash_for_each(card->ip_htable, i, addr, hnode) { + if (addr->type != QETH_IP_TYPE_NORMAL) + continue; + if (qeth_l3_is_addr_covered_by_ipato(card, addr)) + addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; + else + addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; + } +} + static void qeth_l3_clear_ipato_list(struct qeth_card *card) { struct qeth_ipato_entry *ipatoe, *tmp; @@ -619,6 +640,7 @@ static void qeth_l3_clear_ipato_list(str kfree(ipatoe); } + qeth_l3_update_ipato(card); spin_unlock_bh(&card->ip_lock); } @@ -643,8 +665,10 @@ int qeth_l3_add_ipato_entry(struct qeth_ } } - if (!rc) + if (!rc) { list_add_tail(&new->entry, &card->ipato.entries); + qeth_l3_update_ipato(card); + } spin_unlock_bh(&card->ip_lock); @@ -667,6 +691,7 @@ void qeth_l3_del_ipato_entry(struct qeth (proto == QETH_PROT_IPV4)? 4:16) && (ipatoe->mask_bits == mask_bits)) { list_del(&ipatoe->entry); + qeth_l3_update_ipato(card); kfree(ipatoe); } } --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -372,9 +372,8 @@ static ssize_t qeth_l3_dev_ipato_enable_ struct device_attribute *attr, const char *buf, size_t count) { struct qeth_card *card = dev_get_drvdata(dev); - struct qeth_ipaddr *addr; - int i, rc = 0; bool enable; + int rc = 0; if (!card) return -EINVAL; @@ -393,20 +392,12 @@ static ssize_t qeth_l3_dev_ipato_enable_ goto out; } - if (card->ipato.enabled == enable) - goto out; - card->ipato.enabled = enable; - - spin_lock_bh(&card->ip_lock); - hash_for_each(card->ip_htable, i, addr, hnode) { - if (addr->type != QETH_IP_TYPE_NORMAL) - continue; - if (!enable) - addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; - else if (qeth_l3_is_addr_covered_by_ipato(card, addr)) - addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; + if (card->ipato.enabled != enable) { + card->ipato.enabled = enable; + spin_lock_bh(&card->ip_lock); + qeth_l3_update_ipato(card); + spin_unlock_bh(&card->ip_lock); } - spin_unlock_bh(&card->ip_lock); out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; @@ -432,20 +423,27 @@ static ssize_t qeth_l3_dev_ipato_invert4 const char *buf, size_t count) { struct qeth_card *card = dev_get_drvdata(dev); + bool invert; int rc = 0; if (!card) return -EINVAL; mutex_lock(&card->conf_mutex); - if (sysfs_streq(buf, "toggle")) - card->ipato.invert4 = (card->ipato.invert4)? 0 : 1; - else if (sysfs_streq(buf, "1")) - card->ipato.invert4 = 1; - else if (sysfs_streq(buf, "0")) - card->ipato.invert4 = 0; - else + if (sysfs_streq(buf, "toggle")) { + invert = !card->ipato.invert4; + } else if (kstrtobool(buf, &invert)) { rc = -EINVAL; + goto out; + } + + if (card->ipato.invert4 != invert) { + card->ipato.invert4 = invert; + spin_lock_bh(&card->ip_lock); + qeth_l3_update_ipato(card); + spin_unlock_bh(&card->ip_lock); + } +out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; } @@ -611,20 +609,27 @@ static ssize_t qeth_l3_dev_ipato_invert6 struct device_attribute *attr, const char *buf, size_t count) { struct qeth_card *card = dev_get_drvdata(dev); + bool invert; int rc = 0; if (!card) return -EINVAL; mutex_lock(&card->conf_mutex); - if (sysfs_streq(buf, "toggle")) - card->ipato.invert6 = (card->ipato.invert6)? 0 : 1; - else if (sysfs_streq(buf, "1")) - card->ipato.invert6 = 1; - else if (sysfs_streq(buf, "0")) - card->ipato.invert6 = 0; - else + if (sysfs_streq(buf, "toggle")) { + invert = !card->ipato.invert6; + } else if (kstrtobool(buf, &invert)) { rc = -EINVAL; + goto out; + } + + if (card->ipato.invert6 != invert) { + card->ipato.invert6 = invert; + spin_lock_bh(&card->ip_lock); + qeth_l3_update_ipato(card); + spin_unlock_bh(&card->ip_lock); + } +out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; } Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.9/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.9/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.9/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.9/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: don't apply takeover changes to RXIP" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: don't apply takeover changes to RXIP to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-don-t-apply-takeover-changes-to-rxip.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:30 +0100 Subject: s390/qeth: don't apply takeover changes to RXIP From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit b22d73d6689fd902a66c08ebe71ab2f3b351e22f ] When takeover is switched off, current code clears the 'TAKEOVER' flag on all IPs. But the flag is also used for RXIP addresses, and those should not be affected by the takeover mode. Fix the behaviour by consistenly applying takover logic to NORMAL addresses only. Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_l3_main.c | 5 +++-- drivers/s390/net/qeth_l3_sys.c | 5 +++-- 2 files changed, 6 insertions(+), 4 deletions(-) --- a/drivers/s390/net/qeth_l3_main.c +++ b/drivers/s390/net/qeth_l3_main.c @@ -178,6 +178,8 @@ int qeth_l3_is_addr_covered_by_ipato(str if (!card->ipato.enabled) return 0; + if (addr->type != QETH_IP_TYPE_NORMAL) + return 0; qeth_l3_convert_addr_to_bits((u8 *) &addr->u, addr_bits, (addr->proto == QETH_PROT_IPV4)? 4:16); @@ -293,8 +295,7 @@ int qeth_l3_add_ip(struct qeth_card *car memcpy(addr, tmp_addr, sizeof(struct qeth_ipaddr)); addr->ref_counter = 1; - if (addr->type == QETH_IP_TYPE_NORMAL && - qeth_l3_is_addr_covered_by_ipato(card, addr)) { + if (qeth_l3_is_addr_covered_by_ipato(card, addr)) { QETH_CARD_TEXT(card, 2, "tkovaddr"); addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; } --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -398,10 +398,11 @@ static ssize_t qeth_l3_dev_ipato_enable_ card->ipato.enabled = enable; hash_for_each(card->ip_htable, i, addr, hnode) { + if (addr->type != QETH_IP_TYPE_NORMAL) + continue; if (!enable) addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; - else if (addr->type == QETH_IP_TYPE_NORMAL && - qeth_l3_is_addr_covered_by_ipato(card, addr)) + else if (qeth_l3_is_addr_covered_by_ipato(card, addr)) addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; } out: Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.9/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.9/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.9/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.9/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: lock IP table while applying takeover changes" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: lock IP table while applying takeover changes to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-lock-ip-table-while-applying-takeover-changes.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:31 +0100 Subject: s390/qeth: lock IP table while applying takeover changes From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit 8a03a3692b100d84785ee7a834e9215e304c9e00 ] Modifying the flags of an IP addr object needs to be protected against eg. concurrent removal of the same object from the IP table. Fixes: 5f78e29ceebf ("qeth: optimize IP handling in rx_mode callback") Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_l3_sys.c | 2 ++ 1 file changed, 2 insertions(+) --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -397,6 +397,7 @@ static ssize_t qeth_l3_dev_ipato_enable_ goto out; card->ipato.enabled = enable; + spin_lock_bh(&card->ip_lock); hash_for_each(card->ip_htable, i, addr, hnode) { if (addr->type != QETH_IP_TYPE_NORMAL) continue; @@ -405,6 +406,7 @@ static ssize_t qeth_l3_dev_ipato_enable_ else if (qeth_l3_is_addr_covered_by_ipato(card, addr)) addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; } + spin_unlock_bh(&card->ip_lock); out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.9/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.9/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.9/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.9/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: apply takeover changes when mode is toggled" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: apply takeover changes when mode is toggled to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:29 +0100 Subject: s390/qeth: apply takeover changes when mode is toggled From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit 7fbd9493f0eeae8cef58300505a9ef5c8fce6313 ] Just as for an explicit enable/disable, toggling the takeover mode also requires that the IP addresses get updated. Otherwise all IPs that were added to the table before the mode-toggle, get registered with the old settings. Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_core.h | 2 +- drivers/s390/net/qeth_core_main.c | 2 +- drivers/s390/net/qeth_l3_sys.c | 35 +++++++++++++++++------------------ 3 files changed, 19 insertions(+), 20 deletions(-) --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -576,7 +576,7 @@ enum qeth_cq { }; struct qeth_ipato { - int enabled; + bool enabled; int invert4; int invert6; struct list_head entries; --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -1475,7 +1475,7 @@ static int qeth_setup_card(struct qeth_c qeth_set_intial_options(card); /* IP address takeover */ INIT_LIST_HEAD(&card->ipato.entries); - card->ipato.enabled = 0; + card->ipato.enabled = false; card->ipato.invert4 = 0; card->ipato.invert6 = 0; /* init QDIO stuff */ --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -374,6 +374,7 @@ static ssize_t qeth_l3_dev_ipato_enable_ struct qeth_card *card = dev_get_drvdata(dev); struct qeth_ipaddr *addr; int i, rc = 0; + bool enable; if (!card) return -EINVAL; @@ -386,25 +387,23 @@ static ssize_t qeth_l3_dev_ipato_enable_ } if (sysfs_streq(buf, "toggle")) { - card->ipato.enabled = (card->ipato.enabled)? 0 : 1; - } else if (sysfs_streq(buf, "1")) { - card->ipato.enabled = 1; - hash_for_each(card->ip_htable, i, addr, hnode) { - if ((addr->type == QETH_IP_TYPE_NORMAL) && - qeth_l3_is_addr_covered_by_ipato(card, addr)) - addr->set_flags |= - QETH_IPA_SETIP_TAKEOVER_FLAG; - } - } else if (sysfs_streq(buf, "0")) { - card->ipato.enabled = 0; - hash_for_each(card->ip_htable, i, addr, hnode) { - if (addr->set_flags & - QETH_IPA_SETIP_TAKEOVER_FLAG) - addr->set_flags &= - ~QETH_IPA_SETIP_TAKEOVER_FLAG; - } - } else + enable = !card->ipato.enabled; + } else if (kstrtobool(buf, &enable)) { rc = -EINVAL; + goto out; + } + + if (card->ipato.enabled == enable) + goto out; + card->ipato.enabled = enable; + + hash_for_each(card->ip_htable, i, addr, hnode) { + if (!enable) + addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; + else if (addr->type == QETH_IP_TYPE_NORMAL && + qeth_l3_is_addr_covered_by_ipato(card, addr)) + addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; + } out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.9/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.9/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.9/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.9/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "RDS: Check cmsg_len before dereferencing CMSG_DATA" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled RDS: Check cmsg_len before dereferencing CMSG_DATA to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: rds-check-cmsg_len-before-dereferencing-cmsg_data.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Avinash Repaka <avinash.repaka(a)oracle.com> Date: Thu, 21 Dec 2017 20:17:04 -0800 Subject: RDS: Check cmsg_len before dereferencing CMSG_DATA From: Avinash Repaka <avinash.repaka(a)oracle.com> [ Upstream commit 14e138a86f6347c6199f610576d2e11c03bec5f0 ] RDS currently doesn't check if the length of the control message is large enough to hold the required data, before dereferencing the control message data. This results in following crash: BUG: KASAN: stack-out-of-bounds in rds_rdma_bytes net/rds/send.c:1013 [inline] BUG: KASAN: stack-out-of-bounds in rds_sendmsg+0x1f02/0x1f90 net/rds/send.c:1066 Read of size 8 at addr ffff8801c928fb70 by task syzkaller455006/3157 CPU: 0 PID: 3157 Comm: syzkaller455006 Not tainted 4.15.0-rc3+ #161 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x194/0x257 lib/dump_stack.c:53 print_address_description+0x73/0x250 mm/kasan/report.c:252 kasan_report_error mm/kasan/report.c:351 [inline] kasan_report+0x25b/0x340 mm/kasan/report.c:409 __asan_report_load8_noabort+0x14/0x20 mm/kasan/report.c:430 rds_rdma_bytes net/rds/send.c:1013 [inline] rds_sendmsg+0x1f02/0x1f90 net/rds/send.c:1066 sock_sendmsg_nosec net/socket.c:628 [inline] sock_sendmsg+0xca/0x110 net/socket.c:638 ___sys_sendmsg+0x320/0x8b0 net/socket.c:2018 __sys_sendmmsg+0x1ee/0x620 net/socket.c:2108 SYSC_sendmmsg net/socket.c:2139 [inline] SyS_sendmmsg+0x35/0x60 net/socket.c:2134 entry_SYSCALL_64_fastpath+0x1f/0x96 RIP: 0033:0x43fe49 RSP: 002b:00007fffbe244ad8 EFLAGS: 00000217 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000004002c8 RCX: 000000000043fe49 RDX: 0000000000000001 RSI: 000000002020c000 RDI: 0000000000000003 RBP: 00000000006ca018 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000217 R12: 00000000004017b0 R13: 0000000000401840 R14: 0000000000000000 R15: 0000000000000000 To fix this, we verify that the cmsg_len is large enough to hold the data to be read, before proceeding further. Reported-by: syzbot <syzkaller-bugs(a)googlegroups.com> Signed-off-by: Avinash Repaka <avinash.repaka(a)oracle.com> Acked-by: Santosh Shilimkar <santosh.shilimkar(a)oracle.com> Reviewed-by: Yuval Shaia <yuval.shaia(a)oracle.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/rds/send.c | 3 +++ 1 file changed, 3 insertions(+) --- a/net/rds/send.c +++ b/net/rds/send.c @@ -1006,6 +1006,9 @@ static int rds_rdma_bytes(struct msghdr continue; if (cmsg->cmsg_type == RDS_CMSG_RDMA_ARGS) { + if (cmsg->cmsg_len < + CMSG_LEN(sizeof(struct rds_rdma_args))) + return -EINVAL; args = CMSG_DATA(cmsg); *rdma_bytes += args->remote_vec.bytes; } Patches currently in stable-queue which might be from avinash.repaka(a)oracle.com are queue-4.9/rds-check-cmsg_len-before-dereferencing-cmsg_data.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ptr_ring: add barriers" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ptr_ring: add barriers to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ptr_ring-add-barriers.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: "Michael S. Tsirkin" <mst(a)redhat.com> Date: Tue, 5 Dec 2017 21:29:37 +0200 Subject: ptr_ring: add barriers From: "Michael S. Tsirkin" <mst(a)redhat.com> [ Upstream commit a8ceb5dbfde1092b466936bca0ff3be127ecf38e ] Users of ptr_ring expect that it's safe to give the data structure a pointer and have it be available to consumers, but that actually requires an smb_wmb or a stronger barrier. In absence of such barriers and on architectures that reorder writes, consumer might read an un=initialized value from an skb pointer stored in the skb array. This was observed causing crashes. To fix, add memory barriers. The barrier we use is a wmb, the assumption being that producers do not need to read the value so we do not need to order these reads. Reported-by: George Cherian <george.cherian(a)cavium.com> Suggested-by: Jason Wang <jasowang(a)redhat.com> Signed-off-by: Michael S. Tsirkin <mst(a)redhat.com> Acked-by: Jason Wang <jasowang(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/ptr_ring.h | 9 +++++++++ 1 file changed, 9 insertions(+) --- a/include/linux/ptr_ring.h +++ b/include/linux/ptr_ring.h @@ -99,12 +99,18 @@ static inline bool ptr_ring_full_bh(stru /* Note: callers invoking this in a loop must use a compiler barrier, * for example cpu_relax(). Callers must hold producer_lock. + * Callers are responsible for making sure pointer that is being queued + * points to a valid data. */ static inline int __ptr_ring_produce(struct ptr_ring *r, void *ptr) { if (unlikely(!r->size) || r->queue[r->producer]) return -ENOSPC; + /* Make sure the pointer we are storing points to a valid data. */ + /* Pairs with smp_read_barrier_depends in __ptr_ring_consume. */ + smp_wmb(); + r->queue[r->producer++] = ptr; if (unlikely(r->producer >= r->size)) r->producer = 0; @@ -244,6 +250,9 @@ static inline void *__ptr_ring_consume(s if (ptr) __ptr_ring_discard_one(r); + /* Make sure anyone accessing data through the pointer is up to date. */ + /* Pairs with smp_wmb in __ptr_ring_produce. */ + smp_read_barrier_depends(); return ptr; } Patches currently in stable-queue which might be from mst(a)redhat.com are queue-4.9/ptr_ring-add-barriers.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "netlink: Add netns check on taps" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled netlink: Add netns check on taps to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: netlink-add-netns-check-on-taps.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Wed, 6 Dec 2017 12:12:27 -0800 Subject: netlink: Add netns check on taps From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit 93c647643b48f0131f02e45da3bd367d80443291 ] Currently, a nlmon link inside a child namespace can observe systemwide netlink activity. Filter the traffic so that nlmon can only sniff netlink messages from its own netns. Test case: vpnns -- bash -c "ip link add nlmon0 type nlmon; \ ip link set nlmon0 up; \ tcpdump -i nlmon0 -q -w /tmp/nlmon.pcap -U" & sudo ip xfrm state add src 10.1.1.1 dst 10.1.1.2 proto esp \ spi 0x1 mode transport \ auth sha1 0x6162633132330000000000000000000000000000 \ enc aes 0x00000000000000000000000000000000 grep --binary abc123 /tmp/nlmon.pcap Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/netlink/af_netlink.c | 3 +++ 1 file changed, 3 insertions(+) --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -261,6 +261,9 @@ static int __netlink_deliver_tap_skb(str struct sock *sk = skb->sk; int ret = -ENOMEM; + if (!net_eq(dev_net(dev), sock_net(sk))) + return 0; + dev_hold(dev); if (is_vmalloc_addr(skb->head)) Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-4.9/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.9/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: reevalulate autoflowlabel setting after sysctl setting" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: reevalulate autoflowlabel setting after sysctl setting to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-reevalulate-autoflowlabel-setting-after-sysctl-setting.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Shaohua Li <shli(a)fb.com> Date: Wed, 20 Dec 2017 12:10:21 -0800 Subject: net: reevalulate autoflowlabel setting after sysctl setting From: Shaohua Li <shli(a)fb.com> [ Upstream commit 513674b5a2c9c7a67501506419da5c3c77ac6f08 ] sysctl.ip6.auto_flowlabels is default 1. In our hosts, we set it to 2. If sockopt doesn't set autoflowlabel, outcome packets from the hosts are supposed to not include flowlabel. This is true for normal packet, but not for reset packet. The reason is ipv6_pinfo.autoflowlabel is set in sock creation. Later if we change sysctl.ip6.auto_flowlabels, the ipv6_pinfo.autoflowlabel isn't changed, so the sock will keep the old behavior in terms of auto flowlabel. Reset packet is suffering from this problem, because reset packet is sent from a special control socket, which is created at boot time. Since sysctl.ipv6.auto_flowlabels is 1 by default, the control socket will always have its ipv6_pinfo.autoflowlabel set, even after user set sysctl.ipv6.auto_flowlabels to 1, so reset packset will always have flowlabel. Normal sock created before sysctl setting suffers from the same issue. We can't even turn off autoflowlabel unless we kill all socks in the hosts. To fix this, if IPV6_AUTOFLOWLABEL sockopt is used, we use the autoflowlabel setting from user, otherwise we always call ip6_default_np_autolabel() which has the new settings of sysctl. Note, this changes behavior a little bit. Before commit 42240901f7c4 (ipv6: Implement different admin modes for automatic flow labels), the autoflowlabel behavior of a sock isn't sticky, eg, if sysctl changes, existing connection will change autoflowlabel behavior. After that commit, autoflowlabel behavior is sticky in the whole life of the sock. With this patch, the behavior isn't sticky again. Cc: Martin KaFai Lau <kafai(a)fb.com> Cc: Eric Dumazet <eric.dumazet(a)gmail.com> Cc: Tom Herbert <tom(a)quantonium.net> Signed-off-by: Shaohua Li <shli(a)fb.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/ipv6.h | 3 ++- net/ipv6/af_inet6.c | 1 - net/ipv6/ip6_output.c | 12 ++++++++++-- net/ipv6/ipv6_sockglue.c | 1 + 4 files changed, 13 insertions(+), 4 deletions(-) --- a/include/linux/ipv6.h +++ b/include/linux/ipv6.h @@ -246,7 +246,8 @@ struct ipv6_pinfo { * 100: prefer care-of address */ dontfrag:1, - autoflowlabel:1; + autoflowlabel:1, + autoflowlabel_set:1; __u8 min_hopcount; __u8 tclass; __be32 rcv_flowinfo; --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -209,7 +209,6 @@ lookup_protocol: np->mcast_hops = IPV6_DEFAULT_MCASTHOPS; np->mc_loop = 1; np->pmtudisc = IPV6_PMTUDISC_WANT; - np->autoflowlabel = ip6_default_np_autolabel(sock_net(sk)); sk->sk_ipv6only = net->ipv6.sysctl.bindv6only; /* Init the ipv4 part of the socket since we can have sockets --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -156,6 +156,14 @@ int ip6_output(struct net *net, struct s !(IP6CB(skb)->flags & IP6SKB_REROUTED)); } +static bool ip6_autoflowlabel(struct net *net, const struct ipv6_pinfo *np) +{ + if (!np->autoflowlabel_set) + return ip6_default_np_autolabel(net); + else + return np->autoflowlabel; +} + /* * xmit an sk_buff (used by TCP, SCTP and DCCP) * Note : socket lock is not held for SYNACK packets, but might be modified @@ -219,7 +227,7 @@ int ip6_xmit(const struct sock *sk, stru hlimit = ip6_dst_hoplimit(dst); ip6_flow_hdr(hdr, tclass, ip6_make_flowlabel(net, skb, fl6->flowlabel, - np->autoflowlabel, fl6)); + ip6_autoflowlabel(net, np), fl6)); hdr->payload_len = htons(seg_len); hdr->nexthdr = proto; @@ -1691,7 +1699,7 @@ struct sk_buff *__ip6_make_skb(struct so ip6_flow_hdr(hdr, v6_cork->tclass, ip6_make_flowlabel(net, skb, fl6->flowlabel, - np->autoflowlabel, fl6)); + ip6_autoflowlabel(net, np), fl6)); hdr->hop_limit = v6_cork->hop_limit; hdr->nexthdr = proto; hdr->saddr = fl6->saddr; --- a/net/ipv6/ipv6_sockglue.c +++ b/net/ipv6/ipv6_sockglue.c @@ -874,6 +874,7 @@ pref_skip_coa: break; case IPV6_AUTOFLOWLABEL: np->autoflowlabel = valbool; + np->autoflowlabel_set = 1; retv = 0; break; } Patches currently in stable-queue which might be from shli(a)fb.com are queue-4.9/net-reevalulate-autoflowlabel-setting-after-sysctl-setting.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: qmi_wwan: add Sierra EM7565 1199:9091" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: qmi_wwan: add Sierra EM7565 1199:9091 to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-qmi_wwan-add-sierra-em7565-1199-9091.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Sebastian Sjoholm <ssjoholm(a)mac.com> Date: Mon, 11 Dec 2017 21:51:14 +0100 Subject: net: qmi_wwan: add Sierra EM7565 1199:9091 From: Sebastian Sjoholm <ssjoholm(a)mac.com> [ Upstream commit aceef61ee56898cfa7b6960fb60b9326c3860441 ] Sierra Wireless EM7565 is an Qualcomm MDM9x50 based M.2 modem. The USB id is added to qmi_wwan.c to allow QMI communication with the EM7565. Signed-off-by: Sebastian Sjoholm <ssjoholm(a)mac.com> Acked-by: Bjørn Mork <bjorn(a)mork.no> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -907,6 +907,7 @@ static const struct usb_device_id produc {QMI_FIXED_INTF(0x1199, 0x9079, 10)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 8)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 10)}, /* Sierra Wireless EM74xx */ + {QMI_FIXED_INTF(0x1199, 0x9091, 8)}, /* Sierra Wireless EM7565 */ {QMI_FIXED_INTF(0x1bbb, 0x011e, 4)}, /* Telekom Speedstick LTE II (Alcatel One Touch L100V LTE) */ {QMI_FIXED_INTF(0x1bbb, 0x0203, 2)}, /* Alcatel L800MA */ {QMI_FIXED_INTF(0x2357, 0x0201, 4)}, /* TP-LINK HSUPA Modem MA180 */ Patches currently in stable-queue which might be from ssjoholm(a)mac.com are queue-4.9/net-qmi_wwan-add-sierra-em7565-1199-9091.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-phy-micrel-ksz9031-reconfigure-autoneg-after-phy-autoneg-workaround.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Grygorii Strashko <grygorii.strashko(a)ti.com> Date: Wed, 20 Dec 2017 18:45:10 -0600 Subject: net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround From: Grygorii Strashko <grygorii.strashko(a)ti.com> [ Upstream commit c1a8d0a3accf64a014d605e6806ce05d1c17adf1 ] Under some circumstances driver will perform PHY reset in ksz9031_read_status() to fix autoneg failure case (idle error count = 0xFF). When this happens ksz9031 will not detect link status change any more when connecting to Netgear 1G switch (link can be recovered sometimes by restarting netdevice "ifconfig down up"). Reproduced with TI am572x board equipped with ksz9031 PHY while connecting to Netgear 1G switch. Fix the issue by reconfiguring autonegotiation after PHY reset in ksz9031_read_status(). Fixes: d2fd719bcb0e ("net/phy: micrel: Add workaround for bad autoneg") Signed-off-by: Grygorii Strashko <grygorii.strashko(a)ti.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/phy/micrel.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -624,6 +624,7 @@ static int ksz9031_read_status(struct ph phydev->link = 0; if (phydev->drv->config_intr && phy_interrupt_is_valid(phydev)) phydev->drv->config_intr(phydev); + return genphy_config_aneg(phydev); } return 0; Patches currently in stable-queue which might be from grygorii.strashko(a)ti.com are queue-4.9/net-phy-micrel-ksz9031-reconfigure-autoneg-after-phy-autoneg-workaround.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Date: Wed, 6 Dec 2017 15:23:23 +0100 Subject: net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> [ Upstream commit 589bf32f09852041fbd3b7ce1a9e703f95c230ba ] add appropriate calls to clk_disable_unprepare() by jumping to out_mdio in case orion_mdio_probe() returns -EPROBE_DEFER. Found by Linux Driver Verification project (linuxtesting.org). Fixes: 3d604da1e954 ("net: mvmdio: get and enable optional clock") Signed-off-by: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/marvell/mvmdio.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/marvell/mvmdio.c +++ b/drivers/net/ethernet/marvell/mvmdio.c @@ -232,7 +232,8 @@ static int orion_mdio_probe(struct platf dev->regs + MVMDIO_ERR_INT_MASK); } else if (dev->err_interrupt == -EPROBE_DEFER) { - return -EPROBE_DEFER; + ret = -EPROBE_DEFER; + goto out_mdio; } mutex_init(&dev->lock); Patches currently in stable-queue which might be from Tobias.Jordan(a)elektrobit.com are queue-4.9/net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Prevent possible races in VXLAN control flow" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Prevent possible races in VXLAN control flow to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Mon, 4 Dec 2017 09:57:43 +0200 Subject: net/mlx5e: Prevent possible races in VXLAN control flow From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 0c1cc8b2215f5122ca614b5adca60346018758c3 ] When calling add/remove VXLAN port, a lock must be held in order to prevent race scenarios when more than one add/remove happens at the same time. Fix by holding our state_lock (mutex) as done by all other parts of the driver. Note that the spinlock protecting the radix-tree is still needed in order to synchronize radix-tree access from softirq context. Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/vxlan.c | 4 ++++ 1 file changed, 4 insertions(+) --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c @@ -88,6 +88,7 @@ static void mlx5e_vxlan_add_port(struct struct mlx5e_vxlan *vxlan; int err; + mutex_lock(&priv->state_lock); vxlan = mlx5e_vxlan_lookup_port(priv, port); if (vxlan) { atomic_inc(&vxlan->refcount); @@ -117,6 +118,7 @@ err_free: err_delete_port: mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); free_work: + mutex_unlock(&priv->state_lock); kfree(vxlan_work); } @@ -130,6 +132,7 @@ static void mlx5e_vxlan_del_port(struct struct mlx5e_vxlan *vxlan; bool remove = false; + mutex_lock(&priv->state_lock); spin_lock_bh(&vxlan_db->lock); vxlan = radix_tree_lookup(&vxlan_db->tree, port); if (!vxlan) @@ -147,6 +150,7 @@ out_unlock: mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); kfree(vxlan); } + mutex_unlock(&priv->state_lock); kfree(vxlan_work); } Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.9/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.9/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.9/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.9/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Fix possible deadlock of VXLAN lock" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Fix possible deadlock of VXLAN lock to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Thu, 23 Nov 2017 13:52:28 +0200 Subject: net/mlx5e: Fix possible deadlock of VXLAN lock From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 6323514116404cc651df1b7fffa1311ddf8ce647 ] mlx5e_vxlan_lookup_port is called both from mlx5e_add_vxlan_port (user context) and mlx5e_features_check (softirq), but the lock acquired does not disable bottom half and might result in deadlock. Fix it by simply replacing spin_lock() with spin_lock_bh(). While at it, replace all unnecessary spin_lock_irq() to spin_lock_bh(). lockdep's WARNING: inconsistent lock state [ 654.028136] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. [ 654.028229] swapper/5/0 [HC0[0]:SC1[9]:HE1:SE0] takes: [ 654.028321] (&(&vxlan_db->lock)->rlock){+.?.}, at: [<ffffffffa06e7f0e>] mlx5e_vxlan_lookup_port+0x1e/0x50 [mlx5_core] [ 654.028528] {SOFTIRQ-ON-W} state was registered at: [ 654.028607] _raw_spin_lock+0x3c/0x70 [ 654.028689] mlx5e_vxlan_lookup_port+0x1e/0x50 [mlx5_core] [ 654.028794] mlx5e_vxlan_add_port+0x2e/0x120 [mlx5_core] [ 654.028878] process_one_work+0x1e9/0x640 [ 654.028942] worker_thread+0x4a/0x3f0 [ 654.029002] kthread+0x141/0x180 [ 654.029056] ret_from_fork+0x24/0x30 [ 654.029114] irq event stamp: 579088 [ 654.029174] hardirqs last enabled at (579088): [<ffffffff818f475a>] ip6_finish_output2+0x49a/0x8c0 [ 654.029309] hardirqs last disabled at (579087): [<ffffffff818f470e>] ip6_finish_output2+0x44e/0x8c0 [ 654.029446] softirqs last enabled at (579030): [<ffffffff810b3b3d>] irq_enter+0x6d/0x80 [ 654.029567] softirqs last disabled at (579031): [<ffffffff810b3c05>] irq_exit+0xb5/0xc0 [ 654.029684] other info that might help us debug this: [ 654.029781] Possible unsafe locking scenario: [ 654.029868] CPU0 [ 654.029908] ---- [ 654.029947] lock(&(&vxlan_db->lock)->rlock); [ 654.030045] <Interrupt> [ 654.030090] lock(&(&vxlan_db->lock)->rlock); [ 654.030162] *** DEADLOCK *** Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/vxlan.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c @@ -71,9 +71,9 @@ struct mlx5e_vxlan *mlx5e_vxlan_lookup_p struct mlx5e_vxlan_db *vxlan_db = &priv->vxlan; struct mlx5e_vxlan *vxlan; - spin_lock(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); vxlan = radix_tree_lookup(&vxlan_db->tree, port); - spin_unlock(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); return vxlan; } @@ -100,9 +100,9 @@ static void mlx5e_vxlan_add_port(struct vxlan->udp_port = port; - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); err = radix_tree_insert(&vxlan_db->tree, vxlan->udp_port, vxlan); - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); if (err) goto err_free; @@ -121,9 +121,9 @@ static void __mlx5e_vxlan_core_del_port( struct mlx5e_vxlan_db *vxlan_db = &priv->vxlan; struct mlx5e_vxlan *vxlan; - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); vxlan = radix_tree_delete(&vxlan_db->tree, port); - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); if (!vxlan) return; @@ -171,12 +171,12 @@ void mlx5e_vxlan_cleanup(struct mlx5e_pr struct mlx5e_vxlan *vxlan; unsigned int port = 0; - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); while (radix_tree_gang_lookup(&vxlan_db->tree, (void **)&vxlan, port, 1)) { port = vxlan->udp_port; - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); __mlx5e_vxlan_core_del_port(priv, (u16)port); - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); } - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); } Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.9/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.9/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.9/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.9/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Fix features check of IPv6 traffic" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Fix features check of IPv6 traffic to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-fix-features-check-of-ipv6-traffic.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Tue, 21 Nov 2017 17:49:36 +0200 Subject: net/mlx5e: Fix features check of IPv6 traffic From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 2989ad1ec03021ee6d2193c35414f1d970a243de ] The assumption that the next header field contains the transport protocol is wrong for IPv6 packets with extension headers. Instead, we should look the inner-most next header field in the buffer. This will fix TSO offload for tunnels over IPv6 with extension headers. Performance testing: 19.25x improvement, cool! Measuring bandwidth of 16 threads TCP traffic over IPv6 GRE tap. CPU: Intel(R) Xeon(R) CPU E5-2660 v2 @ 2.20GHz NIC: Mellanox Technologies MT28800 Family [ConnectX-5 Ex] TSO: Enabled Before: 4,926.24 Mbps Now : 94,827.91 Mbps Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -3038,6 +3038,7 @@ static netdev_features_t mlx5e_vxlan_fea struct sk_buff *skb, netdev_features_t features) { + unsigned int offset = 0; struct udphdr *udph; u16 proto; u16 port = 0; @@ -3047,7 +3048,7 @@ static netdev_features_t mlx5e_vxlan_fea proto = ip_hdr(skb)->protocol; break; case htons(ETH_P_IPV6): - proto = ipv6_hdr(skb)->nexthdr; + proto = ipv6_find_hdr(skb, &offset, -1, NULL, NULL); break; default: goto out; Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.9/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.9/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.9/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.9/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Add refcount to VXLAN structure" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Add refcount to VXLAN structure to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-add-refcount-to-vxlan-structure.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Sun, 3 Dec 2017 13:58:50 +0200 Subject: net/mlx5e: Add refcount to VXLAN structure From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 23f4cc2cd9ed92570647220aca60d0197d8c1fa9 ] A refcount mechanism must be implemented in order to prevent unwanted scenarios such as: - Open an IPv4 VXLAN interface - Open an IPv6 VXLAN interface (different socket) - Remove one of the interfaces With current implementation, the UDP port will be removed from our VXLAN database and turn off the offloads for the other interface, which is still active. The reference count mechanism will only allow UDP port removals once all consumers are gone. Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/vxlan.c | 50 ++++++++++++------------ drivers/net/ethernet/mellanox/mlx5/core/vxlan.h | 1 2 files changed, 28 insertions(+), 23 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c @@ -88,8 +88,11 @@ static void mlx5e_vxlan_add_port(struct struct mlx5e_vxlan *vxlan; int err; - if (mlx5e_vxlan_lookup_port(priv, port)) + vxlan = mlx5e_vxlan_lookup_port(priv, port); + if (vxlan) { + atomic_inc(&vxlan->refcount); goto free_work; + } if (mlx5e_vxlan_core_add_port_cmd(priv->mdev, port)) goto free_work; @@ -99,6 +102,7 @@ static void mlx5e_vxlan_add_port(struct goto err_delete_port; vxlan->udp_port = port; + atomic_set(&vxlan->refcount, 1); spin_lock_bh(&vxlan_db->lock); err = radix_tree_insert(&vxlan_db->tree, vxlan->udp_port, vxlan); @@ -116,32 +120,33 @@ free_work: kfree(vxlan_work); } -static void __mlx5e_vxlan_core_del_port(struct mlx5e_priv *priv, u16 port) +static void mlx5e_vxlan_del_port(struct work_struct *work) { + struct mlx5e_vxlan_work *vxlan_work = + container_of(work, struct mlx5e_vxlan_work, work); + struct mlx5e_priv *priv = vxlan_work->priv; struct mlx5e_vxlan_db *vxlan_db = &priv->vxlan; + u16 port = vxlan_work->port; struct mlx5e_vxlan *vxlan; + bool remove = false; spin_lock_bh(&vxlan_db->lock); - vxlan = radix_tree_delete(&vxlan_db->tree, port); - spin_unlock_bh(&vxlan_db->lock); - + vxlan = radix_tree_lookup(&vxlan_db->tree, port); if (!vxlan) - return; - - mlx5e_vxlan_core_del_port_cmd(priv->mdev, vxlan->udp_port); - - kfree(vxlan); -} + goto out_unlock; -static void mlx5e_vxlan_del_port(struct work_struct *work) -{ - struct mlx5e_vxlan_work *vxlan_work = - container_of(work, struct mlx5e_vxlan_work, work); - struct mlx5e_priv *priv = vxlan_work->priv; - u16 port = vxlan_work->port; + if (atomic_dec_and_test(&vxlan->refcount)) { + radix_tree_delete(&vxlan_db->tree, port); + remove = true; + } - __mlx5e_vxlan_core_del_port(priv, port); +out_unlock: + spin_unlock_bh(&vxlan_db->lock); + if (remove) { + mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); + kfree(vxlan); + } kfree(vxlan_work); } @@ -171,12 +176,11 @@ void mlx5e_vxlan_cleanup(struct mlx5e_pr struct mlx5e_vxlan *vxlan; unsigned int port = 0; - spin_lock_bh(&vxlan_db->lock); + /* Lockless since we are the only radix-tree consumers, wq is disabled */ while (radix_tree_gang_lookup(&vxlan_db->tree, (void **)&vxlan, port, 1)) { port = vxlan->udp_port; - spin_unlock_bh(&vxlan_db->lock); - __mlx5e_vxlan_core_del_port(priv, (u16)port); - spin_lock_bh(&vxlan_db->lock); + radix_tree_delete(&vxlan_db->tree, port); + mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); + kfree(vxlan); } - spin_unlock_bh(&vxlan_db->lock); } --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.h @@ -36,6 +36,7 @@ #include "en.h" struct mlx5e_vxlan { + atomic_t refcount; u16 udp_port; }; Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.9/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.9/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.9/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.9/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5: Fix rate limit packet pacing naming and struct" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5: Fix rate limit packet pacing naming and struct to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5-fix-rate-limit-packet-pacing-naming-and-struct.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Eran Ben Elisha <eranbe(a)mellanox.com> Date: Mon, 13 Nov 2017 10:11:27 +0200 Subject: net/mlx5: Fix rate limit packet pacing naming and struct From: Eran Ben Elisha <eranbe(a)mellanox.com> [ Upstream commit 37e92a9d4fe38dc3e7308913575983a6a088c8d4 ] In mlx5_ifc, struct size was not complete, and thus driver was sending garbage after the last defined field. Fixed it by adding reserved field to complete the struct size. In addition, rename all set_rate_limit to set_pp_rate_limit to be compliant with the Firmware <-> Driver definition. Fixes: 7486216b3a0b ("{net,IB}/mlx5: mlx5_ifc updates") Fixes: 1466cc5b23d1 ("net/mlx5: Rate limit tables support") Signed-off-by: Eran Ben Elisha <eranbe(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 4 ++-- drivers/net/ethernet/mellanox/mlx5/core/rl.c | 22 +++++++++++----------- include/linux/mlx5/mlx5_ifc.h | 8 +++++--- 3 files changed, 18 insertions(+), 16 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -367,7 +367,7 @@ static int mlx5_internal_err_ret_value(s case MLX5_CMD_OP_QUERY_VPORT_COUNTER: case MLX5_CMD_OP_ALLOC_Q_COUNTER: case MLX5_CMD_OP_QUERY_Q_COUNTER: - case MLX5_CMD_OP_SET_RATE_LIMIT: + case MLX5_CMD_OP_SET_PP_RATE_LIMIT: case MLX5_CMD_OP_QUERY_RATE_LIMIT: case MLX5_CMD_OP_ALLOC_PD: case MLX5_CMD_OP_ALLOC_UAR: @@ -502,7 +502,7 @@ const char *mlx5_command_str(int command MLX5_COMMAND_STR_CASE(ALLOC_Q_COUNTER); MLX5_COMMAND_STR_CASE(DEALLOC_Q_COUNTER); MLX5_COMMAND_STR_CASE(QUERY_Q_COUNTER); - MLX5_COMMAND_STR_CASE(SET_RATE_LIMIT); + MLX5_COMMAND_STR_CASE(SET_PP_RATE_LIMIT); MLX5_COMMAND_STR_CASE(QUERY_RATE_LIMIT); MLX5_COMMAND_STR_CASE(ALLOC_PD); MLX5_COMMAND_STR_CASE(DEALLOC_PD); --- a/drivers/net/ethernet/mellanox/mlx5/core/rl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/rl.c @@ -60,16 +60,16 @@ static struct mlx5_rl_entry *find_rl_ent return ret_entry; } -static int mlx5_set_rate_limit_cmd(struct mlx5_core_dev *dev, +static int mlx5_set_pp_rate_limit_cmd(struct mlx5_core_dev *dev, u32 rate, u16 index) { - u32 in[MLX5_ST_SZ_DW(set_rate_limit_in)] = {0}; - u32 out[MLX5_ST_SZ_DW(set_rate_limit_out)] = {0}; + u32 in[MLX5_ST_SZ_DW(set_pp_rate_limit_in)] = {0}; + u32 out[MLX5_ST_SZ_DW(set_pp_rate_limit_out)] = {0}; - MLX5_SET(set_rate_limit_in, in, opcode, - MLX5_CMD_OP_SET_RATE_LIMIT); - MLX5_SET(set_rate_limit_in, in, rate_limit_index, index); - MLX5_SET(set_rate_limit_in, in, rate_limit, rate); + MLX5_SET(set_pp_rate_limit_in, in, opcode, + MLX5_CMD_OP_SET_PP_RATE_LIMIT); + MLX5_SET(set_pp_rate_limit_in, in, rate_limit_index, index); + MLX5_SET(set_pp_rate_limit_in, in, rate_limit, rate); return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); } @@ -108,7 +108,7 @@ int mlx5_rl_add_rate(struct mlx5_core_de entry->refcount++; } else { /* new rate limit */ - err = mlx5_set_rate_limit_cmd(dev, rate, entry->index); + err = mlx5_set_pp_rate_limit_cmd(dev, rate, entry->index); if (err) { mlx5_core_err(dev, "Failed configuring rate: %u (%d)\n", rate, err); @@ -144,7 +144,7 @@ void mlx5_rl_remove_rate(struct mlx5_cor entry->refcount--; if (!entry->refcount) { /* need to remove rate */ - mlx5_set_rate_limit_cmd(dev, 0, entry->index); + mlx5_set_pp_rate_limit_cmd(dev, 0, entry->index); entry->rate = 0; } @@ -197,8 +197,8 @@ void mlx5_cleanup_rl_table(struct mlx5_c /* Clear all configured rates */ for (i = 0; i < table->max_size; i++) if (table->rl_entry[i].rate) - mlx5_set_rate_limit_cmd(dev, 0, - table->rl_entry[i].index); + mlx5_set_pp_rate_limit_cmd(dev, 0, + table->rl_entry[i].index); kfree(dev->priv.rl_table.rl_entry); } --- a/include/linux/mlx5/mlx5_ifc.h +++ b/include/linux/mlx5/mlx5_ifc.h @@ -143,7 +143,7 @@ enum { MLX5_CMD_OP_ALLOC_Q_COUNTER = 0x771, MLX5_CMD_OP_DEALLOC_Q_COUNTER = 0x772, MLX5_CMD_OP_QUERY_Q_COUNTER = 0x773, - MLX5_CMD_OP_SET_RATE_LIMIT = 0x780, + MLX5_CMD_OP_SET_PP_RATE_LIMIT = 0x780, MLX5_CMD_OP_QUERY_RATE_LIMIT = 0x781, MLX5_CMD_OP_ALLOC_PD = 0x800, MLX5_CMD_OP_DEALLOC_PD = 0x801, @@ -6689,7 +6689,7 @@ struct mlx5_ifc_add_vxlan_udp_dport_in_b u8 vxlan_udp_port[0x10]; }; -struct mlx5_ifc_set_rate_limit_out_bits { +struct mlx5_ifc_set_pp_rate_limit_out_bits { u8 status[0x8]; u8 reserved_at_8[0x18]; @@ -6698,7 +6698,7 @@ struct mlx5_ifc_set_rate_limit_out_bits u8 reserved_at_40[0x40]; }; -struct mlx5_ifc_set_rate_limit_in_bits { +struct mlx5_ifc_set_pp_rate_limit_in_bits { u8 opcode[0x10]; u8 reserved_at_10[0x10]; @@ -6711,6 +6711,8 @@ struct mlx5_ifc_set_rate_limit_in_bits { u8 reserved_at_60[0x20]; u8 rate_limit[0x20]; + + u8 reserved_at_a0[0x160]; }; struct mlx5_ifc_access_register_out_bits { Patches currently in stable-queue which might be from eranbe(a)mellanox.com are queue-4.9/net-mlx5-fix-rate-limit-packet-pacing-naming-and-struct.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5: Fix error flow in CREATE_QP command" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5: Fix error flow in CREATE_QP command to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5-fix-error-flow-in-create_qp-command.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Moni Shoua <monis(a)mellanox.com> Date: Mon, 4 Dec 2017 08:59:25 +0200 Subject: net/mlx5: Fix error flow in CREATE_QP command From: Moni Shoua <monis(a)mellanox.com> [ Upstream commit dbff26e44dc3ec4de6578733b054a0114652a764 ] In error flow, when DESTROY_QP command should be executed, the wrong mailbox was set with data, not the one that is written to hardware, Fix that. Fixes: 09a7d9eca1a6 '{net,IB}/mlx5: QP/XRCD commands via mlx5 ifc' Signed-off-by: Moni Shoua <monis(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/qp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/qp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/qp.c @@ -303,8 +303,8 @@ int mlx5_core_create_qp(struct mlx5_core err_cmd: memset(din, 0, sizeof(din)); memset(dout, 0, sizeof(dout)); - MLX5_SET(destroy_qp_in, in, opcode, MLX5_CMD_OP_DESTROY_QP); - MLX5_SET(destroy_qp_in, in, qpn, qp->qpn); + MLX5_SET(destroy_qp_in, din, opcode, MLX5_CMD_OP_DESTROY_QP); + MLX5_SET(destroy_qp_in, din, qpn, qp->qpn); mlx5_cmd_exec(dev, din, sizeof(din), dout, sizeof(dout)); return err; } Patches currently in stable-queue which might be from monis(a)mellanox.com are queue-4.9/net-mlx5-fix-error-flow-in-create_qp-command.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: ipv4: fix for a race condition in raw_sendmsg" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: ipv4: fix for a race condition in raw_sendmsg to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Mohamed Ghannam <simo.ghannam(a)gmail.com> Date: Sun, 10 Dec 2017 03:50:58 +0000 Subject: net: ipv4: fix for a race condition in raw_sendmsg From: Mohamed Ghannam <simo.ghannam(a)gmail.com> [ Upstream commit 8f659a03a0ba9289b9aeb9b4470e6fb263d6f483 ] inet->hdrincl is racy, and could lead to uninitialized stack pointer usage, so its value should be read only once. Fixes: c008ba5bdc9f ("ipv4: Avoid reading user iov twice after raw_probe_proto_opt") Signed-off-by: Mohamed Ghannam <simo.ghannam(a)gmail.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/raw.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -502,11 +502,16 @@ static int raw_sendmsg(struct sock *sk, int err; struct ip_options_data opt_copy; struct raw_frag_vec rfv; + int hdrincl; err = -EMSGSIZE; if (len > 0xFFFF) goto out; + /* hdrincl should be READ_ONCE(inet->hdrincl) + * but READ_ONCE() doesn't work with bit fields + */ + hdrincl = inet->hdrincl; /* * Check the flags. */ @@ -582,7 +587,7 @@ static int raw_sendmsg(struct sock *sk, /* Linux does not mangle headers on raw sockets, * so that IP options + IP_HDRINCL is non-sense. */ - if (inet->hdrincl) + if (hdrincl) goto done; if (ipc.opt->opt.srr) { if (!daddr) @@ -604,12 +609,12 @@ static int raw_sendmsg(struct sock *sk, flowi4_init_output(&fl4, ipc.oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE, - inet->hdrincl ? IPPROTO_RAW : sk->sk_protocol, + hdrincl ? IPPROTO_RAW : sk->sk_protocol, inet_sk_flowi_flags(sk) | - (inet->hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), + (hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), daddr, saddr, 0, 0); - if (!inet->hdrincl) { + if (!hdrincl) { rfv.msg = msg; rfv.hlen = 0; @@ -634,7 +639,7 @@ static int raw_sendmsg(struct sock *sk, goto do_confirm; back_from_confirm: - if (inet->hdrincl) + if (hdrincl) err = raw_send_hdrinc(sk, &fl4, msg, len, &rt, msg->msg_flags, &ipc.sockc); Patches currently in stable-queue which might be from simo.ghannam(a)gmail.com are queue-4.9/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: igmp: Use correct source address on IGMPv3 reports" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: igmp: Use correct source address on IGMPv3 reports to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-igmp-use-correct-source-address-on-igmpv3-reports.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Mon, 11 Dec 2017 11:13:45 -0800 Subject: net: igmp: Use correct source address on IGMPv3 reports From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit a46182b00290839fa3fa159d54fd3237bd8669f0 ] Closing a multicast socket after the final IPv4 address is deleted from an interface can generate a membership report that uses the source IP from a different interface. The following test script, run from an isolated netns, reproduces the issue: #!/bin/bash ip link add dummy0 type dummy ip link add dummy1 type dummy ip link set dummy0 up ip link set dummy1 up ip addr add 10.1.1.1/24 dev dummy0 ip addr add 192.168.99.99/24 dev dummy1 tcpdump -U -i dummy0 & socat EXEC:"sleep 2" \ UDP4-DATAGRAM:239.101.1.68:8889,ip-add-membership=239.0.1.68:10.1.1.1 & sleep 1 ip addr del 10.1.1.1/24 dev dummy0 sleep 5 kill %tcpdump RFC 3376 specifies that the report must be sent with a valid IP source address from the destination subnet, or from address 0.0.0.0. Add an extra check to make sure this is the case. Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/igmp.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -89,6 +89,7 @@ #include <linux/rtnetlink.h> #include <linux/times.h> #include <linux/pkt_sched.h> +#include <linux/byteorder/generic.h> #include <net/net_namespace.h> #include <net/arp.h> @@ -321,6 +322,23 @@ igmp_scount(struct ip_mc_list *pmc, int return scount; } +/* source address selection per RFC 3376 section 4.2.13 */ +static __be32 igmpv3_get_srcaddr(struct net_device *dev, + const struct flowi4 *fl4) +{ + struct in_device *in_dev = __in_dev_get_rcu(dev); + + if (!in_dev) + return htonl(INADDR_ANY); + + for_ifa(in_dev) { + if (inet_ifa_match(fl4->saddr, ifa)) + return fl4->saddr; + } endfor_ifa(in_dev); + + return htonl(INADDR_ANY); +} + static struct sk_buff *igmpv3_newpack(struct net_device *dev, unsigned int mtu) { struct sk_buff *skb; @@ -368,7 +386,7 @@ static struct sk_buff *igmpv3_newpack(st pip->frag_off = htons(IP_DF); pip->ttl = 1; pip->daddr = fl4.daddr; - pip->saddr = fl4.saddr; + pip->saddr = igmpv3_get_srcaddr(dev, &fl4); pip->protocol = IPPROTO_IGMP; pip->tot_len = 0; /* filled in later */ ip_select_ident(net, skb, NULL); Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-4.9/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.9/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: Fix double free and memory corruption in get_net_ns_by_id()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: Fix double free and memory corruption in get_net_ns_by_id() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: "Eric W. Biederman" <ebiederm(a)xmission.com> Date: Tue, 19 Dec 2017 11:27:56 -0600 Subject: net: Fix double free and memory corruption in get_net_ns_by_id() From: "Eric W. Biederman" <ebiederm(a)xmission.com> [ Upstream commit 21b5944350052d2583e82dd59b19a9ba94a007f0 ] (I can trivially verify that that idr_remove in cleanup_net happens after the network namespace count has dropped to zero --EWB) Function get_net_ns_by_id() does not check for net::count after it has found a peer in netns_ids idr. It may dereference a peer, after its count has already been finaly decremented. This leads to double free and memory corruption: put_net(peer) rtnl_lock() atomic_dec_and_test(&peer->count) [count=0] ... __put_net(peer) get_net_ns_by_id(net, id) spin_lock(&cleanup_list_lock) list_add(&net->cleanup_list, &cleanup_list) spin_unlock(&cleanup_list_lock) queue_work() peer = idr_find(&net->netns_ids, id) | get_net(peer) [count=1] | ... | (use after final put) v ... cleanup_net() ... spin_lock(&cleanup_list_lock) ... list_replace_init(&cleanup_list, ..) ... spin_unlock(&cleanup_list_lock) ... ... ... ... put_net(peer) ... atomic_dec_and_test(&peer->count) [count=0] ... spin_lock(&cleanup_list_lock) ... list_add(&net->cleanup_list, &cleanup_list) ... spin_unlock(&cleanup_list_lock) ... queue_work() ... rtnl_unlock() rtnl_lock() ... for_each_net(tmp) { ... id = __peernet2id(tmp, peer) ... spin_lock_irq(&tmp->nsid_lock) ... idr_remove(&tmp->netns_ids, id) ... ... ... net_drop_ns() ... net_free(peer) ... } ... | v cleanup_net() ... (Second free of peer) Also, put_net() on the right cpu may reorder with left's cpu list_replace_init(&cleanup_list, ..), and then cleanup_list will be corrupted. Since cleanup_net() is executed in worker thread, while put_net(peer) can happen everywhere, there should be enough time for concurrent get_net_ns_by_id() to pick the peer up, and the race does not seem to be unlikely. The patch fixes the problem in standard way. (Also, there is possible problem in peernet2id_alloc(), which requires check for net::count under nsid_lock and maybe_get_net(peer), but in current stable kernel it's used under rtnl_lock() and it has to be safe. Openswitch begun to use peernet2id_alloc(), and possibly it should be fixed too. While this is not in stable kernel yet, so I'll send a separate message to netdev@ later). Cc: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Signed-off-by: Kirill Tkhai <ktkhai(a)virtuozzo.com> Fixes: 0c7aecd4bde4 "netns: add rtnl cmd to add and get peer netns ids" Reviewed-by: Andrey Ryabinin <aryabinin(a)virtuozzo.com> Reviewed-by: "Eric W. Biederman" <ebiederm(a)xmission.com> Signed-off-by: Eric W. Biederman <ebiederm(a)xmission.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Acked-by: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/net_namespace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -263,7 +263,7 @@ struct net *get_net_ns_by_id(struct net spin_lock_irqsave(&net->nsid_lock, flags); peer = idr_find(&net->netns_ids, id); if (peer) - get_net(peer); + peer = maybe_get_net(peer); spin_unlock_irqrestore(&net->nsid_lock, flags); rcu_read_unlock(); Patches currently in stable-queue which might be from ebiederm(a)xmission.com are queue-4.9/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: fec: unmap the xmit buffer that are not transferred by DMA" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: fec: unmap the xmit buffer that are not transferred by DMA to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-fec-unmap-the-xmit-buffer-that-are-not-transferred-by-dma.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Fugang Duan <fugang.duan(a)nxp.com> Date: Fri, 22 Dec 2017 17:12:09 +0800 Subject: net: fec: unmap the xmit buffer that are not transferred by DMA From: Fugang Duan <fugang.duan(a)nxp.com> [ Upstream commit 178e5f57a8d8f8fc5799a624b96fc31ef9a29ffa ] The enet IP only support 32 bit, it will use swiotlb buffer to do dma mapping when xmit buffer DMA memory address is bigger than 4G in i.MX platform. After stress suspend/resume test, it will print out: log: [12826.352864] fec 5b040000.ethernet: swiotlb buffer is full (sz: 191 bytes) [12826.359676] DMA: Out of SW-IOMMU space for 191 bytes at device 5b040000.ethernet [12826.367110] fec 5b040000.ethernet eth0: Tx DMA memory map failed The issue is that the ready xmit buffers that are dma mapped but DMA still don't copy them into fifo, once MAC restart, these DMA buffers are not unmapped. So it should check the dma mapping buffer and unmap them. Signed-off-by: Fugang Duan <fugang.duan(a)nxp.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/freescale/fec_main.c | 6 ++++++ 1 file changed, 6 insertions(+) --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -813,6 +813,12 @@ static void fec_enet_bd_init(struct net_ for (i = 0; i < txq->bd.ring_size; i++) { /* Initialize the BD for every fragment in the page. */ bdp->cbd_sc = cpu_to_fec16(0); + if (bdp->cbd_bufaddr && + !IS_TSO_HEADER(txq, fec32_to_cpu(bdp->cbd_bufaddr))) + dma_unmap_single(&fep->pdev->dev, + fec32_to_cpu(bdp->cbd_bufaddr), + fec16_to_cpu(bdp->cbd_datlen), + DMA_TO_DEVICE); if (txq->tx_skbuff[i]) { dev_kfree_skb_any(txq->tx_skbuff[i]); txq->tx_skbuff[i] = NULL; Patches currently in stable-queue which might be from fugang.duan(a)nxp.com are queue-4.9/net-fec-unmap-the-xmit-buffer-that-are-not-transferred-by-dma.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: fec: Allow reception of frames bigger than 1522 bytes" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: fec: Allow reception of frames bigger than 1522 bytes to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-fec-allow-reception-of-frames-bigger-than-1522-bytes.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Andrew Lunn <andrew(a)lunn.ch> Date: Sun, 30 Jul 2017 19:36:05 +0200 Subject: net: fec: Allow reception of frames bigger than 1522 bytes From: Andrew Lunn <andrew(a)lunn.ch> [ Upstream commit fbbeefdd21049fcf9437c809da3828b210577f36 ] The FEC Receive Control Register has a 14 bit field indicating the longest frame that may be received. It is being set to 1522. Frames longer than this are discarded, but counted as being in error. When using DSA, frames from the switch has an additional header, either 4 or 8 bytes if a Marvell switch is used. Thus a full MTU frame of 1522 bytes received by the switch on a port becomes 1530 bytes when passed to the host via the FEC interface. Change the maximum receive size to 2048 - 64, where 64 is the maximum rx_alignment applied on the receive buffer for AVB capable FEC cores. Use this value also for the maximum receive buffer size. The driver is already allocating a receive SKB of 2048 bytes, so this change should not have any significant effects. Tested on imx51, imx6, vf610. Signed-off-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/freescale/fec_main.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -172,10 +172,12 @@ MODULE_PARM_DESC(macaddr, "FEC Ethernet #endif /* CONFIG_M5272 */ /* The FEC stores dest/src/type/vlan, data, and checksum for receive packets. + * + * 2048 byte skbufs are allocated. However, alignment requirements + * varies between FEC variants. Worst case is 64, so round down by 64. */ -#define PKT_MAXBUF_SIZE 1522 +#define PKT_MAXBUF_SIZE (round_down(2048 - 64, 64)) #define PKT_MINBUF_SIZE 64 -#define PKT_MAXBLR_SIZE 1536 /* FEC receive acceleration */ #define FEC_RACC_IPDIS (1 << 1) @@ -853,7 +855,7 @@ static void fec_enet_enable_ring(struct for (i = 0; i < fep->num_rx_queues; i++) { rxq = fep->rx_queue[i]; writel(rxq->bd.dma, fep->hwp + FEC_R_DES_START(i)); - writel(PKT_MAXBLR_SIZE, fep->hwp + FEC_R_BUFF_SIZE(i)); + writel(PKT_MAXBUF_SIZE, fep->hwp + FEC_R_BUFF_SIZE(i)); /* enable DMA1/2 */ if (i) Patches currently in stable-queue which might be from andrew(a)lunn.ch are queue-4.9/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.9/net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch queue-4.9/net-fec-allow-reception-of-frames-bigger-than-1522-bytes.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-bridge-fix-early-call-to-br_stp_change_bridge_id-and-plug-newlink-leaks.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> Date: Mon, 18 Dec 2017 17:35:09 +0200 Subject: net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks From: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> [ Upstream commit 84aeb437ab98a2bce3d4b2111c79723aedfceb33 ] The early call to br_stp_change_bridge_id in bridge's newlink can cause a memory leak if an error occurs during the newlink because the fdb entries are not cleaned up if a different lladdr was specified, also another minor issue is that it generates fdb notifications with ifindex = 0. Another unrelated memory leak is the bridge sysfs entries which get added on NETDEV_REGISTER event, but are not cleaned up in the newlink error path. To remove this special case the call to br_stp_change_bridge_id is done after netdev register and we cleanup the bridge on changelink error via br_dev_delete to plug all leaks. This patch makes netlink bridge destruction on newlink error the same as dellink and ioctl del which is necessary since at that point we have a fully initialized bridge device. To reproduce the issue: $ ip l add br0 address 00:11:22:33:44:55 type bridge group_fwd_mask 1 RTNETLINK answers: Invalid argument $ rmmod bridge [ 1822.142525] ============================================================================= [ 1822.143640] BUG bridge_fdb_cache (Tainted: G O ): Objects remaining in bridge_fdb_cache on __kmem_cache_shutdown() [ 1822.144821] ----------------------------------------------------------------------------- [ 1822.145990] Disabling lock debugging due to kernel taint [ 1822.146732] INFO: Slab 0x0000000092a844b2 objects=32 used=2 fp=0x00000000fef011b0 flags=0x1ffff8000000100 [ 1822.147700] CPU: 2 PID: 13584 Comm: rmmod Tainted: G B O 4.15.0-rc2+ #87 [ 1822.148578] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_083030-gandalf 04/01/2014 [ 1822.150008] Call Trace: [ 1822.150510] dump_stack+0x78/0xa9 [ 1822.151156] slab_err+0xb1/0xd3 [ 1822.151834] ? __kmalloc+0x1bb/0x1ce [ 1822.152546] __kmem_cache_shutdown+0x151/0x28b [ 1822.153395] shutdown_cache+0x13/0x144 [ 1822.154126] kmem_cache_destroy+0x1c0/0x1fb [ 1822.154669] SyS_delete_module+0x194/0x244 [ 1822.155199] ? trace_hardirqs_on_thunk+0x1a/0x1c [ 1822.155773] entry_SYSCALL_64_fastpath+0x23/0x9a [ 1822.156343] RIP: 0033:0x7f929bd38b17 [ 1822.156859] RSP: 002b:00007ffd160e9a98 EFLAGS: 00000202 ORIG_RAX: 00000000000000b0 [ 1822.157728] RAX: ffffffffffffffda RBX: 00005578316ba090 RCX: 00007f929bd38b17 [ 1822.158422] RDX: 00007f929bd9ec60 RSI: 0000000000000800 RDI: 00005578316ba0f0 [ 1822.159114] RBP: 0000000000000003 R08: 00007f929bff5f20 R09: 00007ffd160e8a11 [ 1822.159808] R10: 00007ffd160e9860 R11: 0000000000000202 R12: 00007ffd160e8a80 [ 1822.160513] R13: 0000000000000000 R14: 0000000000000000 R15: 00005578316ba090 [ 1822.161278] INFO: Object 0x000000007645de29 @offset=0 [ 1822.161666] INFO: Object 0x00000000d5df2ab5 @offset=128 Fixes: 30313a3d5794 ("bridge: Handle IFLA_ADDRESS correctly when creating bridge device") Fixes: 5b8d5429daa0 ("bridge: netlink: register netdevice before executing changelink") Signed-off-by: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/bridge/br_netlink.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) --- a/net/bridge/br_netlink.c +++ b/net/bridge/br_netlink.c @@ -1092,19 +1092,20 @@ static int br_dev_newlink(struct net *sr struct net_bridge *br = netdev_priv(dev); int err; + err = register_netdevice(dev); + if (err) + return err; + if (tb[IFLA_ADDRESS]) { spin_lock_bh(&br->lock); br_stp_change_bridge_id(br, nla_data(tb[IFLA_ADDRESS])); spin_unlock_bh(&br->lock); } - err = register_netdevice(dev); - if (err) - return err; - err = br_changelink(dev, tb, data); if (err) - unregister_netdevice(dev); + br_dev_delete(dev, NULL); + return err; } Patches currently in stable-queue which might be from nikolay(a)cumulusnetworks.com are queue-4.9/net-bridge-fix-early-call-to-br_stp_change_bridge_id-and-plug-newlink-leaks.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv6: mcast: better catch silly mtu values" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv6: mcast: better catch silly mtu values to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv6-mcast-better-catch-silly-mtu-values.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:03:38 -0800 Subject: ipv6: mcast: better catch silly mtu values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b9b312a7a451e9c098921856e7cfbc201120e1a7 ] syzkaller reported crashes in IPv6 stack [1] Xin Long found that lo MTU was set to silly values. IPv6 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in mld code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv6 minimal MTU. [1] skbuff: skb_over_panic: text:0000000010b86b8d len:196 put:20 head:000000003b477e60 data:000000000e85441e tail:0xd4 end:0xc0 dev:lo ------------[ cut here ]------------ kernel BUG at net/core/skbuff.c:104! invalid opcode: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.15.0-rc2-mm1+ #39 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:skb_panic+0x15c/0x1f0 net/core/skbuff.c:100 RSP: 0018:ffff8801db307508 EFLAGS: 00010286 RAX: 0000000000000082 RBX: ffff8801c517e840 RCX: 0000000000000000 RDX: 0000000000000082 RSI: 1ffff1003b660e61 RDI: ffffed003b660e95 RBP: ffff8801db307570 R08: 1ffff1003b660e23 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffffffff85bd4020 R13: ffffffff84754ed2 R14: 0000000000000014 R15: ffff8801c4e26540 FS: 0000000000000000(0000) GS:ffff8801db300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000463610 CR3: 00000001c6698000 CR4: 00000000001406e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <IRQ> skb_over_panic net/core/skbuff.c:109 [inline] skb_put+0x181/0x1c0 net/core/skbuff.c:1694 add_grhead.isra.24+0x42/0x3b0 net/ipv6/mcast.c:1695 add_grec+0xa55/0x1060 net/ipv6/mcast.c:1817 mld_send_cr net/ipv6/mcast.c:1903 [inline] mld_ifc_timer_expire+0x4d2/0x770 net/ipv6/mcast.c:2448 call_timer_fn+0x23b/0x840 kernel/time/timer.c:1320 expire_timers kernel/time/timer.c:1357 [inline] __run_timers+0x7e1/0xb60 kernel/time/timer.c:1660 run_timer_softirq+0x4c/0xb0 kernel/time/timer.c:1686 __do_softirq+0x29d/0xbb2 kernel/softirq.c:285 invoke_softirq kernel/softirq.c:365 [inline] irq_exit+0x1d3/0x210 kernel/softirq.c:405 exiting_irq arch/x86/include/asm/apic.h:540 [inline] smp_apic_timer_interrupt+0x16b/0x700 arch/x86/kernel/apic/apic.c:1052 apic_timer_interrupt+0xa9/0xb0 arch/x86/entry/entry_64.S:920 Signed-off-by: Eric Dumazet <edumazet(a)google.com> Reported-by: syzbot <syzkaller(a)googlegroups.com> Tested-by: Xin Long <lucien.xin(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/mcast.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -1682,16 +1682,16 @@ static int grec_size(struct ifmcaddr6 *p } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ifmcaddr6 *pmc, - int type, struct mld2_grec **ppgr) + int type, struct mld2_grec **ppgr, unsigned int mtu) { - struct net_device *dev = pmc->idev->dev; struct mld2_report *pmr; struct mld2_grec *pgr; - if (!skb) - skb = mld_newpack(pmc->idev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = mld_newpack(pmc->idev, mtu); + if (!skb) + return NULL; + } pgr = (struct mld2_grec *)skb_put(skb, sizeof(struct mld2_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -1714,10 +1714,15 @@ static struct sk_buff *add_grec(struct s struct mld2_grec *pgr = NULL; struct ip6_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->mca_flags & MAF_NOREPORT) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV6_MIN_MTU) + return skb; + isquery = type == MLD2_MODE_IS_INCLUDE || type == MLD2_MODE_IS_EXCLUDE; truncate = type == MLD2_MODE_IS_EXCLUDE || @@ -1738,7 +1743,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); } } first = 1; @@ -1774,12 +1779,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -1814,7 +1819,7 @@ empty_source: mld_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.9/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.9/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.9/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.9/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.9/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.9/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.9/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: igmp: guard against silly MTU values" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: igmp: guard against silly MTU values to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-igmp-guard-against-silly-mtu-values.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:17:39 -0800 Subject: ipv4: igmp: guard against silly MTU values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b5476022bbada3764609368f03329ca287528dc8 ] IPv4 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in igmp code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv4 minimal MTU. This patch adds missing IPV4_MIN_MTU define, to not abuse ETH_MIN_MTU anymore. Signed-off-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/net/ip.h | 2 ++ net/ipv4/devinet.c | 2 +- net/ipv4/igmp.c | 24 +++++++++++++++--------- net/ipv4/ip_tunnel.c | 4 ++-- 4 files changed, 20 insertions(+), 12 deletions(-) --- a/include/net/ip.h +++ b/include/net/ip.h @@ -33,6 +33,8 @@ #include <net/flow.h> #include <net/flow_dissector.h> +#define IPV4_MIN_MTU 68 /* RFC 791 */ + struct sock; struct inet_skb_parm { --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -1380,7 +1380,7 @@ skip: static bool inetdev_valid_mtu(unsigned int mtu) { - return mtu >= 68; + return mtu >= IPV4_MIN_MTU; } static void inetdev_send_gratuitous_arp(struct net_device *dev, --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -404,16 +404,17 @@ static int grec_size(struct ip_mc_list * } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ip_mc_list *pmc, - int type, struct igmpv3_grec **ppgr) + int type, struct igmpv3_grec **ppgr, unsigned int mtu) { struct net_device *dev = pmc->interface->dev; struct igmpv3_report *pih; struct igmpv3_grec *pgr; - if (!skb) - skb = igmpv3_newpack(dev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = igmpv3_newpack(dev, mtu); + if (!skb) + return NULL; + } pgr = (struct igmpv3_grec *)skb_put(skb, sizeof(struct igmpv3_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -436,12 +437,17 @@ static struct sk_buff *add_grec(struct s struct igmpv3_grec *pgr = NULL; struct ip_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->multiaddr == IGMP_ALL_HOSTS) return skb; if (ipv4_is_local_multicast(pmc->multiaddr) && !net->ipv4.sysctl_igmp_llm_reports) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV4_MIN_MTU) + return skb; + isquery = type == IGMPV3_MODE_IS_INCLUDE || type == IGMPV3_MODE_IS_EXCLUDE; truncate = type == IGMPV3_MODE_IS_EXCLUDE || @@ -462,7 +468,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); } } first = 1; @@ -498,12 +504,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -538,7 +544,7 @@ empty_source: igmpv3_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@ -346,8 +346,8 @@ static int ip_tunnel_bind_dev(struct net dev->needed_headroom = t_hlen + hlen; mtu -= (dev->hard_header_len + t_hlen); - if (mtu < 68) - mtu = 68; + if (mtu < IPV4_MIN_MTU) + mtu = IPV4_MIN_MTU; return mtu; } Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.9/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.9/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.9/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.9/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.9/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.9/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.9/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: Fix use-after-free when flushing FIB tables" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: Fix use-after-free when flushing FIB tables to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-fix-use-after-free-when-flushing-fib-tables.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: Ido Schimmel <idosch(a)mellanox.com> Date: Wed, 20 Dec 2017 19:34:19 +0200 Subject: ipv4: Fix use-after-free when flushing FIB tables From: Ido Schimmel <idosch(a)mellanox.com> [ Upstream commit b4681c2829e24943aadd1a7bb3a30d41d0a20050 ] Since commit 0ddcf43d5d4a ("ipv4: FIB Local/MAIN table collapse") the local table uses the same trie allocated for the main table when custom rules are not in use. When a net namespace is dismantled, the main table is flushed and freed (via an RCU callback) before the local table. In case the callback is invoked before the local table is iterated, a use-after-free can occur. Fix this by iterating over the FIB tables in reverse order, so that the main table is always freed after the local table. v3: Reworded comment according to Alex's suggestion. v2: Add a comment to make the fix more explicit per Dave's and Alex's feedback. Fixes: 0ddcf43d5d4a ("ipv4: FIB Local/MAIN table collapse") Signed-off-by: Ido Schimmel <idosch(a)mellanox.com> Reported-by: Fengguang Wu <fengguang.wu(a)intel.com> Acked-by: Alexander Duyck <alexander.h.duyck(a)intel.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/fib_frontend.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -1253,14 +1253,19 @@ fail: static void ip_fib_net_exit(struct net *net) { - unsigned int i; + int i; rtnl_lock(); #ifdef CONFIG_IP_MULTIPLE_TABLES RCU_INIT_POINTER(net->ipv4.fib_main, NULL); RCU_INIT_POINTER(net->ipv4.fib_default, NULL); #endif - for (i = 0; i < FIB_TABLE_HASHSZ; i++) { + /* Destroy the tables in reverse order to guarantee that the + * local table, ID 255, is destroyed before the main table, ID + * 254. This is necessary as the local table may contain + * references to data contained in the main table. + */ + for (i = FIB_TABLE_HASHSZ - 1; i >= 0; i--) { struct hlist_head *head = &net->ipv4.fib_table_hash[i]; struct hlist_node *tmp; struct fib_table *tb; Patches currently in stable-queue which might be from idosch(a)mellanox.com are queue-4.9/ipv4-fix-use-after-free-when-flushing-fib-tables.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "adding missing rcu_read_unlock in ipxip6_rcv" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled adding missing rcu_read_unlock in ipxip6_rcv to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: adding-missing-rcu_read_unlock-in-ipxip6_rcv.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:13:15 CET 2017 From: "Nikita V. Shirokov" <tehnerd(a)fb.com> Date: Wed, 6 Dec 2017 17:15:43 -0800 Subject: adding missing rcu_read_unlock in ipxip6_rcv From: "Nikita V. Shirokov" <tehnerd(a)fb.com> [ Upstream commit 74c4b656c3d92ec4c824ea1a4afd726b7b6568c8 ] commit 8d79266bc48c ("ip6_tunnel: add collect_md mode to IPv6 tunnels") introduced new exit point in ipxip6_rcv. however rcu_read_unlock is missing there. this diff is fixing this v1->v2: instead of doing rcu_read_unlock in place, we are going to "drop" section (to prevent skb leakage) Fixes: 8d79266bc48c ("ip6_tunnel: add collect_md mode to IPv6 tunnels") Signed-off-by: Nikita V. Shirokov <tehnerd(a)fb.com> Acked-by: Alexei Starovoitov <ast(a)kernel.org> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/ip6_tunnel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -911,7 +911,7 @@ static int ipxip6_rcv(struct sk_buff *sk if (t->parms.collect_md) { tun_dst = ipv6_tun_rx_dst(skb, 0, 0, 0); if (!tun_dst) - return 0; + goto drop; } ret = __ip6_tnl_rcv(t, skb, tpi, tun_dst, dscp_ecn_decapsulate, log_ecn_error); Patches currently in stable-queue which might be from tehnerd(a)fb.com are queue-4.9/adding-missing-rcu_read_unlock-in-ipxip6_rcv.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "vxlan: restore dev->mtu setting based on lower device" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled vxlan: restore dev->mtu setting based on lower device to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: vxlan-restore-dev-mtu-setting-based-on-lower-device.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Alexey Kodanev <alexey.kodanev(a)oracle.com> Date: Thu, 14 Dec 2017 20:20:00 +0300 Subject: vxlan: restore dev->mtu setting based on lower device From: Alexey Kodanev <alexey.kodanev(a)oracle.com> [ Upstream commit f870c1ff65a6d1f3a083f277280802ee09a5b44d ] Stefano Brivio says: Commit a985343ba906 ("vxlan: refactor verification and application of configuration") introduced a change in the behaviour of initial MTU setting: earlier, the MTU for a link created on top of a given lower device, without an initial MTU specification, was set to the MTU of the lower device minus headroom as a result of this path in vxlan_dev_configure(): if (!conf->mtu) dev->mtu = lowerdev->mtu - (use_ipv6 ? VXLAN6_HEADROOM : VXLAN_HEADROOM); which is now gone. Now, the initial MTU, in absence of a configured value, is simply set by ether_setup() to ETH_DATA_LEN (1500 bytes). This breaks userspace expectations in case the MTU of the lower device is higher than 1500 bytes minus headroom. This patch restores the previous behaviour on newlink operation. Since max_mtu can be negative and we update dev->mtu directly, also check it for valid minimum. Reported-by: Junhan Yan <juyan(a)redhat.com> Fixes: a985343ba906 ("vxlan: refactor verification and application of configuration") Signed-off-by: Alexey Kodanev <alexey.kodanev(a)oracle.com> Acked-by: Stefano Brivio <sbrivio(a)redhat.com> Signed-off-by: Stefano Brivio <sbrivio(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/vxlan.c | 5 +++++ 1 file changed, 5 insertions(+) --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@ -3105,6 +3105,11 @@ static void vxlan_config_apply(struct ne max_mtu = lowerdev->mtu - (use_ipv6 ? VXLAN6_HEADROOM : VXLAN_HEADROOM); + if (max_mtu < ETH_MIN_MTU) + max_mtu = ETH_MIN_MTU; + + if (!changelink && !conf->mtu) + dev->mtu = max_mtu; } if (dev->mtu > max_mtu) Patches currently in stable-queue which might be from alexey.kodanev(a)oracle.com are queue-4.14/vxlan-restore-dev-mtu-setting-based-on-lower-device.patch queue-4.14/ip6_gre-fix-device-features-for-ioctl-setup.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tipc: fix hanging poll() for stream sockets" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tipc: fix hanging poll() for stream sockets to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tipc-fix-hanging-poll-for-stream-sockets.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Parthasarathy Bhuvaragan <parthasarathy.bhuvaragan(a)gmail.com> Date: Thu, 28 Dec 2017 12:03:06 +0100 Subject: tipc: fix hanging poll() for stream sockets From: Parthasarathy Bhuvaragan <parthasarathy.bhuvaragan(a)gmail.com> [ Upstream commit 517d7c79bdb39864e617960504bdc1aa560c75c6 ] In commit 42b531de17d2f6 ("tipc: Fix missing connection request handling"), we replaced unconditional wakeup() with condtional wakeup for clients with flags POLLIN | POLLRDNORM | POLLRDBAND. This breaks the applications which do a connect followed by poll with POLLOUT flag. These applications are not woken when the connection is ESTABLISHED and hence sleep forever. In this commit, we fix it by including the POLLOUT event for sockets in TIPC_CONNECTING state. Fixes: 42b531de17d2f6 ("tipc: Fix missing connection request handling") Acked-by: Jon Maloy <jon.maloy(a)ericsson.com> Signed-off-by: Parthasarathy Bhuvaragan <parthasarathy.bhuvaragan(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/tipc/socket.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -709,11 +709,11 @@ static unsigned int tipc_poll(struct fil switch (sk->sk_state) { case TIPC_ESTABLISHED: + case TIPC_CONNECTING: if (!tsk->cong_link_cnt && !tsk_conn_cong(tsk)) mask |= POLLOUT; /* fall thru' */ case TIPC_LISTEN: - case TIPC_CONNECTING: if (!skb_queue_empty(&sk->sk_receive_queue)) mask |= (POLLIN | POLLRDNORM); break; Patches currently in stable-queue which might be from parthasarathy.bhuvaragan(a)gmail.com are queue-4.14/tipc-fix-hanging-poll-for-stream-sockets.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tg3: Fix rx hang on MTU change with 5717/5719" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tg3: Fix rx hang on MTU change with 5717/5719 to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Brian King <brking(a)linux.vnet.ibm.com> Date: Fri, 15 Dec 2017 15:21:50 -0600 Subject: tg3: Fix rx hang on MTU change with 5717/5719 From: Brian King <brking(a)linux.vnet.ibm.com> [ Upstream commit 748a240c589824e9121befb1cba5341c319885bc ] This fixes a hang issue seen when changing the MTU size from 1500 MTU to 9000 MTU on both 5717 and 5719 chips. In discussion with Broadcom, they've indicated that these chipsets have the same phy as the 57766 chipset, so the same workarounds apply. This has been tested by IBM on both Power 8 and Power 9 systems as well as by Broadcom on x86 hardware and has been confirmed to resolve the hang issue. Signed-off-by: Brian King <brking(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/broadcom/tg3.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -14227,7 +14227,9 @@ static int tg3_change_mtu(struct net_dev /* Reset PHY, otherwise the read DMA engine will be in a mode that * breaks all requests to 256 bytes. */ - if (tg3_asic_rev(tp) == ASIC_REV_57766) + if (tg3_asic_rev(tp) == ASIC_REV_57766 || + tg3_asic_rev(tp) == ASIC_REV_5717 || + tg3_asic_rev(tp) == ASIC_REV_5719) reset_phy = true; err = tg3_restart_hw(tp, reset_phy); Patches currently in stable-queue which might be from brking(a)linux.vnet.ibm.com are queue-4.14/tg3-fix-rx-hang-on-mtu-change-with-5717-5719.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp_bbr: reset long-term bandwidth sampling on loss recovery undo" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp_bbr: reset long-term bandwidth sampling on loss recovery undo to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp_bbr-reset-long-term-bandwidth-sampling-on-loss-recovery-undo.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Neal Cardwell <ncardwell(a)google.com> Date: Thu, 7 Dec 2017 12:43:32 -0500 Subject: tcp_bbr: reset long-term bandwidth sampling on loss recovery undo From: Neal Cardwell <ncardwell(a)google.com> [ Upstream commit 600647d467c6d04b3954b41a6ee1795b5ae00550 ] Fix BBR so that upon notification of a loss recovery undo BBR resets long-term bandwidth sampling. Under high reordering, reordering events can be interpreted as loss. If the reordering and spurious loss estimates are high enough, this can cause BBR to spuriously estimate that we are seeing loss rates high enough to trigger long-term bandwidth estimation. To avoid that problem, this commit resets long-term bandwidth sampling on loss recovery undo events. Signed-off-by: Neal Cardwell <ncardwell(a)google.com> Reviewed-by: Yuchung Cheng <ycheng(a)google.com> Acked-by: Soheil Hassas Yeganeh <soheil(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_bbr.c | 1 + 1 file changed, 1 insertion(+) --- a/net/ipv4/tcp_bbr.c +++ b/net/ipv4/tcp_bbr.c @@ -878,6 +878,7 @@ static u32 bbr_undo_cwnd(struct sock *sk bbr->full_bw = 0; /* spurious slow-down; reset full pipe detection */ bbr->full_bw_cnt = 0; + bbr_reset_lt_bw_sampling(sk); return tcp_sk(sk)->snd_cwnd; } Patches currently in stable-queue which might be from ncardwell(a)google.com are queue-4.14/tcp-refresh-tcp_mstamp-from-timers-callbacks.patch queue-4.14/tcp_bbr-reset-full-pipe-detection-on-loss-recovery-undo.patch queue-4.14/tcp_bbr-reset-long-term-bandwidth-sampling-on-loss-recovery-undo.patch queue-4.14/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.14/tcp-fix-potential-underestimation-on-rcv_rtt.patch queue-4.14/tcp_bbr-record-full-bw-reached-decision-in-new-full_bw_reached-bit.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp_bbr: reset full pipe detection on loss recovery undo" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp_bbr: reset full pipe detection on loss recovery undo to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp_bbr-reset-full-pipe-detection-on-loss-recovery-undo.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Neal Cardwell <ncardwell(a)google.com> Date: Thu, 7 Dec 2017 12:43:31 -0500 Subject: tcp_bbr: reset full pipe detection on loss recovery undo From: Neal Cardwell <ncardwell(a)google.com> [ Upstream commit 2f6c498e4f15d27852c04ed46d804a39137ba364 ] Fix BBR so that upon notification of a loss recovery undo BBR resets the full pipe detection (STARTUP exit) state machine. Under high reordering, reordering events can be interpreted as loss. If the reordering and spurious loss estimates are high enough, this could previously cause BBR to spuriously estimate that the pipe is full. Since spurious loss recovery means that our overall sending will have slowed down spuriously, this commit gives a flow more time to probe robustly for bandwidth and decide the pipe is really full. Signed-off-by: Neal Cardwell <ncardwell(a)google.com> Reviewed-by: Yuchung Cheng <ycheng(a)google.com> Acked-by: Soheil Hassas Yeganeh <soheil(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_bbr.c | 4 ++++ 1 file changed, 4 insertions(+) --- a/net/ipv4/tcp_bbr.c +++ b/net/ipv4/tcp_bbr.c @@ -874,6 +874,10 @@ static u32 bbr_sndbuf_expand(struct sock */ static u32 bbr_undo_cwnd(struct sock *sk) { + struct bbr *bbr = inet_csk_ca(sk); + + bbr->full_bw = 0; /* spurious slow-down; reset full pipe detection */ + bbr->full_bw_cnt = 0; return tcp_sk(sk)->snd_cwnd; } Patches currently in stable-queue which might be from ncardwell(a)google.com are queue-4.14/tcp-refresh-tcp_mstamp-from-timers-callbacks.patch queue-4.14/tcp_bbr-reset-full-pipe-detection-on-loss-recovery-undo.patch queue-4.14/tcp_bbr-reset-long-term-bandwidth-sampling-on-loss-recovery-undo.patch queue-4.14/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.14/tcp-fix-potential-underestimation-on-rcv_rtt.patch queue-4.14/tcp_bbr-record-full-bw-reached-decision-in-new-full_bw_reached-bit.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp_bbr: record "full bw reached" decision in new full_bw_reached bit" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp_bbr: record "full bw reached" decision in new full_bw_reached bit to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp_bbr-record-full-bw-reached-decision-in-new-full_bw_reached-bit.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Neal Cardwell <ncardwell(a)google.com> Date: Thu, 7 Dec 2017 12:43:30 -0500 Subject: tcp_bbr: record "full bw reached" decision in new full_bw_reached bit From: Neal Cardwell <ncardwell(a)google.com> [ Upstream commit c589e69b508d29ed8e644dfecda453f71c02ec27 ] This commit records the "full bw reached" decision in a new full_bw_reached bit. This is a pure refactor that does not change the current behavior, but enables subsequent fixes and improvements. In particular, this enables simple and clean fixes because the full_bw and full_bw_cnt can be unconditionally zeroed without worrying about forgetting that we estimated we filled the pipe in Startup. And it enables future improvements because multiple code paths can be used for estimating that we filled the pipe in Startup; any new code paths only need to set this bit when they think the pipe is full. Note that this fix intentionally reduces the width of the full_bw_cnt counter, since we have never used the most significant bit. Signed-off-by: Neal Cardwell <ncardwell(a)google.com> Reviewed-by: Yuchung Cheng <ycheng(a)google.com> Acked-by: Soheil Hassas Yeganeh <soheil(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_bbr.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) --- a/net/ipv4/tcp_bbr.c +++ b/net/ipv4/tcp_bbr.c @@ -110,7 +110,8 @@ struct bbr { u32 lt_last_lost; /* LT intvl start: tp->lost */ u32 pacing_gain:10, /* current gain for setting pacing rate */ cwnd_gain:10, /* current gain for setting cwnd */ - full_bw_cnt:3, /* number of rounds without large bw gains */ + full_bw_reached:1, /* reached full bw in Startup? */ + full_bw_cnt:2, /* number of rounds without large bw gains */ cycle_idx:3, /* current index in pacing_gain cycle array */ has_seen_rtt:1, /* have we seen an RTT sample yet? */ unused_b:5; @@ -180,7 +181,7 @@ static bool bbr_full_bw_reached(const st { const struct bbr *bbr = inet_csk_ca(sk); - return bbr->full_bw_cnt >= bbr_full_bw_cnt; + return bbr->full_bw_reached; } /* Return the windowed max recent bandwidth sample, in pkts/uS << BW_SCALE. */ @@ -717,6 +718,7 @@ static void bbr_check_full_bw_reached(st return; } ++bbr->full_bw_cnt; + bbr->full_bw_reached = bbr->full_bw_cnt >= bbr_full_bw_cnt; } /* If pipe is probably full, drain the queue and then enter steady-state. */ @@ -850,6 +852,7 @@ static void bbr_init(struct sock *sk) bbr->restore_cwnd = 0; bbr->round_start = 0; bbr->idle_restart = 0; + bbr->full_bw_reached = 0; bbr->full_bw = 0; bbr->full_bw_cnt = 0; bbr->cycle_mstamp = 0; Patches currently in stable-queue which might be from ncardwell(a)google.com are queue-4.14/tcp-refresh-tcp_mstamp-from-timers-callbacks.patch queue-4.14/tcp_bbr-reset-full-pipe-detection-on-loss-recovery-undo.patch queue-4.14/tcp_bbr-reset-long-term-bandwidth-sampling-on-loss-recovery-undo.patch queue-4.14/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.14/tcp-fix-potential-underestimation-on-rcv_rtt.patch queue-4.14/tcp_bbr-record-full-bw-reached-decision-in-new-full_bw_reached-bit.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp: refresh tcp_mstamp from timers callbacks" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp: refresh tcp_mstamp from timers callbacks to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-refresh-tcp_mstamp-from-timers-callbacks.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Tue, 12 Dec 2017 18:22:52 -0800 Subject: tcp: refresh tcp_mstamp from timers callbacks From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit 4688eb7cf3ae2c2721d1dacff5c1384cba47d176 ] Only the retransmit timer currently refreshes tcp_mstamp We should do the same for delayed acks and keepalives. Even if RFC 7323 does not request it, this is consistent to what linux did in the past, when TS values were based on jiffies. Fixes: 385e20706fac ("tcp: use tp->tcp_mstamp in output path") Signed-off-by: Eric Dumazet <edumazet(a)google.com> Cc: Soheil Hassas Yeganeh <soheil(a)google.com> Cc: Mike Maloney <maloney(a)google.com> Cc: Neal Cardwell <ncardwell(a)google.com> Acked-by: Neal Cardwell <ncardwell(a)google.com> Acked-by: Soheil Hassas Yeganeh <soheil(a)google.com> Acked-by: Mike Maloney <maloney(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_timer.c | 2 ++ 1 file changed, 2 insertions(+) --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -264,6 +264,7 @@ void tcp_delack_timer_handler(struct soc icsk->icsk_ack.pingpong = 0; icsk->icsk_ack.ato = TCP_ATO_MIN; } + tcp_mstamp_refresh(tcp_sk(sk)); tcp_send_ack(sk); __NET_INC_STATS(sock_net(sk), LINUX_MIB_DELAYEDACKS); } @@ -627,6 +628,7 @@ static void tcp_keepalive_timer (unsigne goto out; } + tcp_mstamp_refresh(tp); if (sk->sk_state == TCP_FIN_WAIT2 && sock_flag(sk, SOCK_DEAD)) { if (tp->linger2 >= 0) { const int tmo = tcp_fin_time(sk) - TCP_TIMEWAIT_LEN; Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.14/tcp-refresh-tcp_mstamp-from-timers-callbacks.patch queue-4.14/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.14/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.14/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.14/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.14/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.14/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.14/tcp-fix-potential-underestimation-on-rcv_rtt.patch queue-4.14/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp md5sig: Use skb's saddr when replying to an incoming segment" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp md5sig: Use skb's saddr when replying to an incoming segment to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Christoph Paasch <cpaasch(a)apple.com> Date: Mon, 11 Dec 2017 00:05:46 -0800 Subject: tcp md5sig: Use skb's saddr when replying to an incoming segment From: Christoph Paasch <cpaasch(a)apple.com> [ Upstream commit 30791ac41927ebd3e75486f9504b6d2280463bf0 ] The MD5-key that belongs to a connection is identified by the peer's IP-address. When we are in tcp_v4(6)_reqsk_send_ack(), we are replying to an incoming segment from tcp_check_req() that failed the seq-number checks. Thus, to find the correct key, we need to use the skb's saddr and not the daddr. This bug seems to have been there since quite a while, but probably got unnoticed because the consequences are not catastrophic. We will call tcp_v4_reqsk_send_ack only to send a challenge-ACK back to the peer, thus the connection doesn't really fail. Fixes: 9501f9722922 ("tcp md5sig: Let the caller pass appropriate key for tcp_v{4,6}_do_calc_md5_hash().") Signed-off-by: Christoph Paasch <cpaasch(a)apple.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_ipv4.c | 2 +- net/ipv6/tcp_ipv6.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -844,7 +844,7 @@ static void tcp_v4_reqsk_send_ack(const tcp_time_stamp_raw() + tcp_rsk(req)->ts_off, req->ts_recent, 0, - tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->daddr, + tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->saddr, AF_INET), inet_rsk(req)->no_srccheck ? IP_REPLY_ARG_NOSRCCHECK : 0, ip_hdr(skb)->tos); --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -988,7 +988,7 @@ static void tcp_v6_reqsk_send_ack(const req->rsk_rcv_wnd >> inet_rsk(req)->rcv_wscale, tcp_time_stamp_raw() + tcp_rsk(req)->ts_off, req->ts_recent, sk->sk_bound_dev_if, - tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr), + tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->saddr), 0, 0); } Patches currently in stable-queue which might be from cpaasch(a)apple.com are queue-4.14/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp: invalidate rate samples during SACK reneging" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp: invalidate rate samples during SACK reneging to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-invalidate-rate-samples-during-sack-reneging.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Yousuk Seung <ysseung(a)google.com> Date: Thu, 7 Dec 2017 13:41:34 -0800 Subject: tcp: invalidate rate samples during SACK reneging From: Yousuk Seung <ysseung(a)google.com> [ Upstream commit d4761754b4fb2ef8d9a1e9d121c4bec84e1fe292 ] Mark tcp_sock during a SACK reneging event and invalidate rate samples while marked. Such rate samples may overestimate bw by including packets that were SACKed before reneging. < ack 6001 win 10000 sack 7001:38001 < ack 7001 win 0 sack 8001:38001 // Reneg detected > seq 7001:8001 // RTO, SACK cleared. < ack 38001 win 10000 In above example the rate sample taken after the last ack will count 7001-38001 as delivered while the actual delivery rate likely could be much lower i.e. 7001-8001. This patch adds a new field tcp_sock.sack_reneg and marks it when we declare SACK reneging and entering TCP_CA_Loss, and unmarks it after the last rate sample was taken before moving back to TCP_CA_Open. This patch also invalidates rate samples taken while tcp_sock.is_sack_reneg is set. Fixes: b9f64820fb22 ("tcp: track data delivery rate for a TCP connection") Signed-off-by: Yousuk Seung <ysseung(a)google.com> Signed-off-by: Neal Cardwell <ncardwell(a)google.com> Signed-off-by: Yuchung Cheng <ycheng(a)google.com> Acked-by: Soheil Hassas Yeganeh <soheil(a)google.com> Acked-by: Eric Dumazet <edumazet(a)google.com> Acked-by: Priyaranjan Jha <priyarjha(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/tcp.h | 3 ++- include/net/tcp.h | 2 +- net/ipv4/tcp.c | 1 + net/ipv4/tcp_input.c | 10 ++++++++-- net/ipv4/tcp_rate.c | 10 +++++++--- 5 files changed, 19 insertions(+), 7 deletions(-) --- a/include/linux/tcp.h +++ b/include/linux/tcp.h @@ -214,7 +214,8 @@ struct tcp_sock { u8 chrono_type:2, /* current chronograph type */ rate_app_limited:1, /* rate_{delivered,interval_us} limited? */ fastopen_connect:1, /* FASTOPEN_CONNECT sockopt */ - unused:4; + is_sack_reneg:1, /* in recovery from loss with SACK reneg? */ + unused:3; u8 nonagle : 4,/* Disable Nagle algorithm? */ thin_lto : 1,/* Use linear timeouts for thin streams */ unused1 : 1, --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1085,7 +1085,7 @@ void tcp_rate_skb_sent(struct sock *sk, void tcp_rate_skb_delivered(struct sock *sk, struct sk_buff *skb, struct rate_sample *rs); void tcp_rate_gen(struct sock *sk, u32 delivered, u32 lost, - struct rate_sample *rs); + bool is_sack_reneg, struct rate_sample *rs); void tcp_rate_check_app_limited(struct sock *sk); /* These functions determine how the current flow behaves in respect of SACK --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -2356,6 +2356,7 @@ int tcp_disconnect(struct sock *sk, int tp->snd_cwnd_cnt = 0; tp->window_clamp = 0; tcp_set_ca_state(sk, TCP_CA_Open); + tp->is_sack_reneg = 0; tcp_clear_retrans(tp); inet_csk_delack_init(sk); /* Initialize rcv_mss to TCP_MIN_MSS to avoid division by 0 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1975,6 +1975,8 @@ void tcp_enter_loss(struct sock *sk) NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSACKRENEGING); tp->sacked_out = 0; tp->fackets_out = 0; + /* Mark SACK reneging until we recover from this loss event. */ + tp->is_sack_reneg = 1; } tcp_clear_all_retrans_hints(tp); @@ -2428,6 +2430,7 @@ static bool tcp_try_undo_recovery(struct return true; } tcp_set_ca_state(sk, TCP_CA_Open); + tp->is_sack_reneg = 0; return false; } @@ -2459,8 +2462,10 @@ static bool tcp_try_undo_loss(struct soc NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSPURIOUSRTOS); inet_csk(sk)->icsk_retransmits = 0; - if (frto_undo || tcp_is_sack(tp)) + if (frto_undo || tcp_is_sack(tp)) { tcp_set_ca_state(sk, TCP_CA_Open); + tp->is_sack_reneg = 0; + } return true; } return false; @@ -3551,6 +3556,7 @@ static int tcp_ack(struct sock *sk, cons struct tcp_sacktag_state sack_state; struct rate_sample rs = { .prior_delivered = 0 }; u32 prior_snd_una = tp->snd_una; + bool is_sack_reneg = tp->is_sack_reneg; u32 ack_seq = TCP_SKB_CB(skb)->seq; u32 ack = TCP_SKB_CB(skb)->ack_seq; bool is_dupack = false; @@ -3666,7 +3672,7 @@ static int tcp_ack(struct sock *sk, cons delivered = tp->delivered - delivered; /* freshly ACKed or SACKed */ lost = tp->lost - lost; /* freshly marked lost */ - tcp_rate_gen(sk, delivered, lost, sack_state.rate); + tcp_rate_gen(sk, delivered, lost, is_sack_reneg, sack_state.rate); tcp_cong_control(sk, ack, delivered, flag, sack_state.rate); tcp_xmit_recovery(sk, rexmit); return 1; --- a/net/ipv4/tcp_rate.c +++ b/net/ipv4/tcp_rate.c @@ -106,7 +106,7 @@ void tcp_rate_skb_delivered(struct sock /* Update the connection delivery information and generate a rate sample. */ void tcp_rate_gen(struct sock *sk, u32 delivered, u32 lost, - struct rate_sample *rs) + bool is_sack_reneg, struct rate_sample *rs) { struct tcp_sock *tp = tcp_sk(sk); u32 snd_us, ack_us; @@ -124,8 +124,12 @@ void tcp_rate_gen(struct sock *sk, u32 d rs->acked_sacked = delivered; /* freshly ACKed or SACKed */ rs->losses = lost; /* freshly marked lost */ - /* Return an invalid sample if no timing information is available. */ - if (!rs->prior_mstamp) { + /* Return an invalid sample if no timing information is available or + * in recovery from loss with SACK reneging. Rate samples taken during + * a SACK reneging event may overestimate bw by including packets that + * were SACKed before the reneg. + */ + if (!rs->prior_mstamp || is_sack_reneg) { rs->delivered = -1; rs->interval_us = -1; return; Patches currently in stable-queue which might be from ysseung(a)google.com are queue-4.14/tcp-invalidate-rate-samples-during-sack-reneging.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "tcp: fix potential underestimation on rcv_rtt" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled tcp: fix potential underestimation on rcv_rtt to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: tcp-fix-potential-underestimation-on-rcv_rtt.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Wei Wang <weiwan(a)google.com> Date: Tue, 12 Dec 2017 16:28:58 -0800 Subject: tcp: fix potential underestimation on rcv_rtt From: Wei Wang <weiwan(a)google.com> [ Upstream commit 9ee11bd03cb1a5c3ca33c2bb70e7ed325f68890f ] When ms timestamp is used, current logic uses 1us in tcp_rcv_rtt_update() when the real rcv_rtt is within 1 - 999us. This could cause rcv_rtt underestimation. Fix it by always using a min value of 1ms if ms timestamp is used. Fixes: 645f4c6f2ebd ("tcp: switch rcv_rtt_est and rcvq_space to high resolution timestamps") Signed-off-by: Wei Wang <weiwan(a)google.com> Signed-off-by: Eric Dumazet <edumazet(a)google.com> Acked-by: Neal Cardwell <ncardwell(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/tcp_input.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -521,9 +521,6 @@ static void tcp_rcv_rtt_update(struct tc u32 new_sample = tp->rcv_rtt_est.rtt_us; long m = sample; - if (m == 0) - m = 1; - if (new_sample != 0) { /* If we sample in larger samples in the non-timestamp * case, we could grossly overestimate the RTT especially @@ -560,6 +557,8 @@ static inline void tcp_rcv_rtt_measure(s if (before(tp->rcv_nxt, tp->rcv_rtt_est.seq)) return; delta_us = tcp_stamp_us_delta(tp->tcp_mstamp, tp->rcv_rtt_est.time); + if (!delta_us) + delta_us = 1; tcp_rcv_rtt_update(tp, delta_us, 1); new_measure: @@ -576,8 +575,11 @@ static inline void tcp_rcv_rtt_measure_t (TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq >= inet_csk(sk)->icsk_ack.rcv_mss)) { u32 delta = tcp_time_stamp(tp) - tp->rx_opt.rcv_tsecr; - u32 delta_us = delta * (USEC_PER_SEC / TCP_TS_HZ); + u32 delta_us; + if (!delta) + delta = 1; + delta_us = delta * (USEC_PER_SEC / TCP_TS_HZ); tcp_rcv_rtt_update(tp, delta_us, 0); } } Patches currently in stable-queue which might be from weiwan(a)google.com are queue-4.14/tcp-fix-potential-underestimation-on-rcv_rtt.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sock: free skb in skb_complete_tx_timestamp on error" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sock: free skb in skb_complete_tx_timestamp on error to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Willem de Bruijn <willemb(a)google.com> Date: Wed, 13 Dec 2017 14:41:06 -0500 Subject: sock: free skb in skb_complete_tx_timestamp on error From: Willem de Bruijn <willemb(a)google.com> [ Upstream commit 35b99dffc3f710cafceee6c8c6ac6a98eb2cb4bf ] skb_complete_tx_timestamp must ingest the skb it is passed. Call kfree_skb if the skb cannot be enqueued. Fixes: b245be1f4db1 ("net-timestamp: no-payload only sysctl") Fixes: 9ac25fc06375 ("net: fix socket refcounting in skb_complete_tx_timestamp()") Reported-by: Richard Cochran <richardcochran(a)gmail.com> Signed-off-by: Willem de Bruijn <willemb(a)google.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/skbuff.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -4296,7 +4296,7 @@ void skb_complete_tx_timestamp(struct sk struct sock *sk = skb->sk; if (!skb_may_tx_timestamp(sk, false)) - return; + goto err; /* Take a reference to prevent skb_orphan() from freeing the socket, * but only if the socket refcount is not zero. @@ -4305,7 +4305,11 @@ void skb_complete_tx_timestamp(struct sk *skb_hwtstamps(skb) = *hwtstamps; __skb_complete_tx_timestamp(skb, sk, SCM_TSTAMP_SND, false); sock_put(sk); + return; } + +err: + kfree_skb(skb); } EXPORT_SYMBOL_GPL(skb_complete_tx_timestamp); Patches currently in stable-queue which might be from willemb(a)google.com are queue-4.14/skbuff-skb_copy_ubufs-must-release-uarg-even-without-user-frags.patch queue-4.14/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.14/skbuff-orphan-frags-before-zerocopy-clone.patch queue-4.14/skbuff-in-skb_copy_ubufs-unclone-before-releasing-zerocopy.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "skbuff: skb_copy_ubufs must release uarg even without user frags" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled skbuff: skb_copy_ubufs must release uarg even without user frags to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: skbuff-skb_copy_ubufs-must-release-uarg-even-without-user-frags.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Willem de Bruijn <willemb(a)google.com> Date: Wed, 20 Dec 2017 17:37:50 -0500 Subject: skbuff: skb_copy_ubufs must release uarg even without user frags From: Willem de Bruijn <willemb(a)google.com> [ Upstream commit b90ddd568792bcb0054eaf0f61785c8f80c3bd1c ] skb_copy_ubufs creates a private copy of frags[] to release its hold on user frags, then calls uarg->callback to notify the owner. Call uarg->callback even when no frags exist. This edge case can happen when zerocopy_sg_from_iter finds enough room in skb_headlen to copy all the data. Fixes: 3ece782693c4 ("sock: skb_copy_ubufs support for compound pages") Signed-off-by: Willem de Bruijn <willemb(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/skbuff.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -1182,7 +1182,7 @@ int skb_copy_ubufs(struct sk_buff *skb, u32 d_off; if (!num_frags) - return 0; + goto release; if (skb_shared(skb) || skb_unclone(skb, gfp_mask)) return -EINVAL; @@ -1242,6 +1242,7 @@ int skb_copy_ubufs(struct sk_buff *skb, __skb_fill_page_desc(skb, new_frags - 1, head, 0, d_off); skb_shinfo(skb)->nr_frags = new_frags; +release: skb_zcopy_clear(skb, false); return 0; } Patches currently in stable-queue which might be from willemb(a)google.com are queue-4.14/skbuff-skb_copy_ubufs-must-release-uarg-even-without-user-frags.patch queue-4.14/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.14/skbuff-orphan-frags-before-zerocopy-clone.patch queue-4.14/skbuff-in-skb_copy_ubufs-unclone-before-releasing-zerocopy.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "skbuff: orphan frags before zerocopy clone" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled skbuff: orphan frags before zerocopy clone to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: skbuff-orphan-frags-before-zerocopy-clone.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Willem de Bruijn <willemb(a)google.com> Date: Wed, 20 Dec 2017 17:37:49 -0500 Subject: skbuff: orphan frags before zerocopy clone From: Willem de Bruijn <willemb(a)google.com> [ Upstream commit 268b790679422a89e9ab0685d9f291edae780c98 ] Call skb_zerocopy_clone after skb_orphan_frags, to avoid duplicate calls to skb_uarg(skb)->callback for the same data. skb_zerocopy_clone associates skb_shinfo(skb)->uarg from frag_skb with each segment. This is only safe for uargs that do refcounting, which is those that pass skb_orphan_frags without dropping their shared frags. For others, skb_orphan_frags drops the user frags and sets the uarg to NULL, after which sock_zerocopy_clone has no effect. Qemu hangs were reported due to duplicate vhost_net_zerocopy_callback calls for the same data causing the vhost_net_ubuf_ref_>refcount to drop below zero. Link: http://lkml.kernel.org/r/<CAF=yD-LWyCD4Y0aJ9O0e_CHLR+3JOeKicRRTEVCPxgw4XOcqGQ(a)mail.gmail.com> Fixes: 1f8b977ab32d ("sock: enable MSG_ZEROCOPY") Reported-by: Andreas Hartmann <andihartmann(a)01019freenet.de> Reported-by: David Hill <dhill(a)redhat.com> Signed-off-by: Willem de Bruijn <willemb(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/skbuff.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -3657,8 +3657,6 @@ normal: skb_shinfo(nskb)->tx_flags |= skb_shinfo(head_skb)->tx_flags & SKBTX_SHARED_FRAG; - if (skb_zerocopy_clone(nskb, head_skb, GFP_ATOMIC)) - goto err; while (pos < offset + len) { if (i >= nfrags) { @@ -3684,6 +3682,8 @@ normal: if (unlikely(skb_orphan_frags(frag_skb, GFP_ATOMIC))) goto err; + if (skb_zerocopy_clone(nskb, frag_skb, GFP_ATOMIC)) + goto err; *nskb_frag = *frag; __skb_frag_ref(nskb_frag); Patches currently in stable-queue which might be from willemb(a)google.com are queue-4.14/skbuff-skb_copy_ubufs-must-release-uarg-even-without-user-frags.patch queue-4.14/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.14/skbuff-orphan-frags-before-zerocopy-clone.patch queue-4.14/skbuff-in-skb_copy_ubufs-unclone-before-releasing-zerocopy.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "skbuff: in skb_copy_ubufs unclone before releasing zerocopy" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled skbuff: in skb_copy_ubufs unclone before releasing zerocopy to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: skbuff-in-skb_copy_ubufs-unclone-before-releasing-zerocopy.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Willem de Bruijn <willemb(a)google.com> Date: Thu, 28 Dec 2017 12:38:13 -0500 Subject: skbuff: in skb_copy_ubufs unclone before releasing zerocopy From: Willem de Bruijn <willemb(a)google.com> skb_copy_ubufs must unclone before it is safe to modify its skb_shared_info with skb_zcopy_clear. Commit b90ddd568792 ("skbuff: skb_copy_ubufs must release uarg even without user frags") ensures that all skbs release their zerocopy state, even those without frags. But I forgot an edge case where such an skb arrives that is cloned. The stack does not build such packets. Vhost/tun skbs have their frags orphaned before cloning. TCP skbs only attach zerocopy state when a frag is added. But if TCP packets can be trimmed or linearized, this might occur. Tracing the code I found no instance so far (e.g., skb_linearize ends up calling skb_zcopy_clear if !skb->data_len). Still, it is non-obvious that no path exists. And it is fragile to rely on this. Fixes: b90ddd568792 ("skbuff: skb_copy_ubufs must release uarg even without user frags") Signed-off-by: Willem de Bruijn <willemb(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/skbuff.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -1181,12 +1181,12 @@ int skb_copy_ubufs(struct sk_buff *skb, int i, new_frags; u32 d_off; - if (!num_frags) - goto release; - if (skb_shared(skb) || skb_unclone(skb, gfp_mask)) return -EINVAL; + if (!num_frags) + goto release; + new_frags = (__skb_pagelen(skb) + PAGE_SIZE - 1) >> PAGE_SHIFT; for (i = 0; i < new_frags; i++) { page = alloc_page(gfp_mask); Patches currently in stable-queue which might be from willemb(a)google.com are queue-4.14/skbuff-skb_copy_ubufs-must-release-uarg-even-without-user-frags.patch queue-4.14/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.14/skbuff-orphan-frags-before-zerocopy-clone.patch queue-4.14/skbuff-in-skb_copy_ubufs-unclone-before-releasing-zerocopy.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sfc: pass valid pointers from efx_enqueue_unwind" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sfc: pass valid pointers from efx_enqueue_unwind to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sfc-pass-valid-pointers-from-efx_enqueue_unwind.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Bert Kenward <bkenward(a)solarflare.com> Date: Thu, 7 Dec 2017 17:18:58 +0000 Subject: sfc: pass valid pointers from efx_enqueue_unwind From: Bert Kenward <bkenward(a)solarflare.com> [ Upstream commit d4a7a8893d4cdbc89d79ac4aa704bf8d4b67b368 ] The bytes_compl and pkts_compl pointers passed to efx_dequeue_buffers cannot be NULL. Add a paranoid warning to check this condition and fix the one case where they were NULL. efx_enqueue_unwind() is called very rarely, during error handling. Without this fix it would fail with a NULL pointer dereference in efx_dequeue_buffer, with efx_enqueue_skb in the call stack. Fixes: e9117e5099ea ("sfc: Firmware-Assisted TSO version 2") Reported-by: Jarod Wilson <jarod(a)redhat.com> Signed-off-by: Bert Kenward <bkenward(a)solarflare.com> Tested-by: Jarod Wilson <jarod(a)redhat.com> Acked-by: Jarod Wilson <jarod(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/sfc/tx.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/sfc/tx.c +++ b/drivers/net/ethernet/sfc/tx.c @@ -77,6 +77,7 @@ static void efx_dequeue_buffer(struct ef } if (buffer->flags & EFX_TX_BUF_SKB) { + EFX_WARN_ON_PARANOID(!pkts_compl || !bytes_compl); (*pkts_compl)++; (*bytes_compl) += buffer->skb->len; dev_consume_skb_any((struct sk_buff *)buffer->skb); @@ -426,12 +427,14 @@ static int efx_tx_map_data(struct efx_tx static void efx_enqueue_unwind(struct efx_tx_queue *tx_queue) { struct efx_tx_buffer *buffer; + unsigned int bytes_compl = 0; + unsigned int pkts_compl = 0; /* Work backwards until we hit the original insert pointer value */ while (tx_queue->insert_count != tx_queue->write_count) { --tx_queue->insert_count; buffer = __efx_tx_queue_get_insert_buffer(tx_queue); - efx_dequeue_buffer(tx_queue, buffer, NULL, NULL); + efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); } } Patches currently in stable-queue which might be from bkenward(a)solarflare.com are queue-4.14/sfc-pass-valid-pointers-from-efx_enqueue_unwind.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sctp: Replace use of sockets_allocated with specified macro." has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: Replace use of sockets_allocated with specified macro. to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-replace-use-of-sockets_allocated-with-specified-macro.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> Date: Fri, 22 Dec 2017 10:15:20 -0800 Subject: sctp: Replace use of sockets_allocated with specified macro. From: Tonghao Zhang <xiangxia.m.yue(a)gmail.com> [ Upstream commit 8cb38a602478e9f806571f6920b0a3298aabf042 ] The patch(180d8cd942ce) replaces all uses of struct sock fields' memory_pressure, memory_allocated, sockets_allocated, and sysctl_mem to accessor macros. But the sockets_allocated field of sctp sock is not replaced at all. Then replace it now for unifying the code. Fixes: 180d8cd942ce ("foundations of per-cgroup memory pressure controlling.") Cc: Glauber Costa <glommer(a)parallels.com> Signed-off-by: Tonghao Zhang <zhangtonghao(a)didichuxing.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/socket.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4413,7 +4413,7 @@ static int sctp_init_sock(struct sock *s SCTP_DBG_OBJCNT_INC(sock); local_bh_disable(); - percpu_counter_inc(&sctp_sockets_allocated); + sk_sockets_allocated_inc(sk); sock_prot_inuse_add(net, sk->sk_prot, 1); /* Nothing can fail after this block, otherwise @@ -4457,7 +4457,7 @@ static void sctp_destroy_sock(struct soc } sctp_endpoint_free(sp->ep); local_bh_disable(); - percpu_counter_dec(&sctp_sockets_allocated); + sk_sockets_allocated_dec(sk); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); local_bh_enable(); } Patches currently in stable-queue which might be from xiangxia.m.yue(a)gmail.com are queue-4.14/sctp-replace-use-of-sockets_allocated-with-specified-macro.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "sctp: make sure stream nums can match optlen in sctp_setsockopt_reset_streams" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled sctp: make sure stream nums can match optlen in sctp_setsockopt_reset_streams to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: sctp-make-sure-stream-nums-can-match-optlen-in-sctp_setsockopt_reset_streams.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Xin Long <lucien.xin(a)gmail.com> Date: Sun, 10 Dec 2017 15:40:51 +0800 Subject: sctp: make sure stream nums can match optlen in sctp_setsockopt_reset_streams From: Xin Long <lucien.xin(a)gmail.com> [ Upstream commit 2342b8d95bcae5946e1b9b8d58645f37500ef2e7 ] Now in sctp_setsockopt_reset_streams, it only does the check optlen < sizeof(*params) for optlen. But it's not enough, as params->srs_number_streams should also match optlen. If the streams in params->srs_stream_list are less than stream nums in params->srs_number_streams, later when dereferencing the stream list, it could cause a slab-out-of-bounds crash, as reported by syzbot. This patch is to fix it by also checking the stream numbers in sctp_setsockopt_reset_streams to make sure at least it's not greater than the streams in the list. Fixes: 7f9d68ac944e ("sctp: implement sender-side procedures for SSN Reset Request Parameter") Reported-by: Dmitry Vyukov <dvyukov(a)google.com> Signed-off-by: Xin Long <lucien.xin(a)gmail.com> Acked-by: Marcelo Ricardo Leitner <marcelo.leitner(a)gmail.com> Acked-by: Neil Horman <nhorman(a)tuxdriver.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sctp/socket.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -3874,13 +3874,17 @@ static int sctp_setsockopt_reset_streams struct sctp_association *asoc; int retval = -EINVAL; - if (optlen < sizeof(struct sctp_reset_streams)) + if (optlen < sizeof(*params)) return -EINVAL; params = memdup_user(optval, optlen); if (IS_ERR(params)) return PTR_ERR(params); + if (params->srs_number_streams * sizeof(__u16) > + optlen - sizeof(*params)) + goto out; + asoc = sctp_id2assoc(sk, params->srs_assoc_id); if (!asoc) goto out; Patches currently in stable-queue which might be from lucien.xin(a)gmail.com are queue-4.14/sctp-make-sure-stream-nums-can-match-optlen-in-sctp_setsockopt_reset_streams.patch queue-4.14/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: update takeover IPs after configuration change" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: update takeover IPs after configuration change to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-update-takeover-ips-after-configuration-change.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:32 +0100 Subject: s390/qeth: update takeover IPs after configuration change From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit 02f510f326501470348a5df341e8232c3497bbbb ] Any modification to the takeover IP-ranges requires that we re-evaluate which IP addresses are takeover-eligible. Otherwise we might do takeover for some addresses when we no longer should, or vice-versa. Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_core.h | 4 +- drivers/s390/net/qeth_core_main.c | 4 +- drivers/s390/net/qeth_l3.h | 2 - drivers/s390/net/qeth_l3_main.c | 31 ++++++++++++++++-- drivers/s390/net/qeth_l3_sys.c | 63 ++++++++++++++++++++------------------ 5 files changed, 67 insertions(+), 37 deletions(-) --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -565,8 +565,8 @@ enum qeth_cq { struct qeth_ipato { bool enabled; - int invert4; - int invert6; + bool invert4; + bool invert6; struct list_head entries; }; --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -1480,8 +1480,8 @@ static int qeth_setup_card(struct qeth_c /* IP address takeover */ INIT_LIST_HEAD(&card->ipato.entries); card->ipato.enabled = false; - card->ipato.invert4 = 0; - card->ipato.invert6 = 0; + card->ipato.invert4 = false; + card->ipato.invert6 = false; /* init QDIO stuff */ qeth_init_qdio_info(card); INIT_DELAYED_WORK(&card->buffer_reclaim_work, qeth_buffer_reclaim_work); --- a/drivers/s390/net/qeth_l3.h +++ b/drivers/s390/net/qeth_l3.h @@ -82,7 +82,7 @@ void qeth_l3_del_vipa(struct qeth_card * int qeth_l3_add_rxip(struct qeth_card *, enum qeth_prot_versions, const u8 *); void qeth_l3_del_rxip(struct qeth_card *card, enum qeth_prot_versions, const u8 *); -int qeth_l3_is_addr_covered_by_ipato(struct qeth_card *, struct qeth_ipaddr *); +void qeth_l3_update_ipato(struct qeth_card *card); struct qeth_ipaddr *qeth_l3_get_addr_buffer(enum qeth_prot_versions); int qeth_l3_add_ip(struct qeth_card *, struct qeth_ipaddr *); int qeth_l3_delete_ip(struct qeth_card *, struct qeth_ipaddr *); --- a/drivers/s390/net/qeth_l3_main.c +++ b/drivers/s390/net/qeth_l3_main.c @@ -163,8 +163,8 @@ static void qeth_l3_convert_addr_to_bits } } -int qeth_l3_is_addr_covered_by_ipato(struct qeth_card *card, - struct qeth_ipaddr *addr) +static bool qeth_l3_is_addr_covered_by_ipato(struct qeth_card *card, + struct qeth_ipaddr *addr) { struct qeth_ipato_entry *ipatoe; u8 addr_bits[128] = {0, }; @@ -605,6 +605,27 @@ int qeth_l3_setrouting_v6(struct qeth_ca /* * IP address takeover related functions */ + +/** + * qeth_l3_update_ipato() - Update 'takeover' property, for all NORMAL IPs. + * + * Caller must hold ip_lock. + */ +void qeth_l3_update_ipato(struct qeth_card *card) +{ + struct qeth_ipaddr *addr; + unsigned int i; + + hash_for_each(card->ip_htable, i, addr, hnode) { + if (addr->type != QETH_IP_TYPE_NORMAL) + continue; + if (qeth_l3_is_addr_covered_by_ipato(card, addr)) + addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; + else + addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; + } +} + static void qeth_l3_clear_ipato_list(struct qeth_card *card) { struct qeth_ipato_entry *ipatoe, *tmp; @@ -616,6 +637,7 @@ static void qeth_l3_clear_ipato_list(str kfree(ipatoe); } + qeth_l3_update_ipato(card); spin_unlock_bh(&card->ip_lock); } @@ -640,8 +662,10 @@ int qeth_l3_add_ipato_entry(struct qeth_ } } - if (!rc) + if (!rc) { list_add_tail(&new->entry, &card->ipato.entries); + qeth_l3_update_ipato(card); + } spin_unlock_bh(&card->ip_lock); @@ -664,6 +688,7 @@ void qeth_l3_del_ipato_entry(struct qeth (proto == QETH_PROT_IPV4)? 4:16) && (ipatoe->mask_bits == mask_bits)) { list_del(&ipatoe->entry); + qeth_l3_update_ipato(card); kfree(ipatoe); } } --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -370,9 +370,8 @@ static ssize_t qeth_l3_dev_ipato_enable_ struct device_attribute *attr, const char *buf, size_t count) { struct qeth_card *card = dev_get_drvdata(dev); - struct qeth_ipaddr *addr; - int i, rc = 0; bool enable; + int rc = 0; if (!card) return -EINVAL; @@ -391,20 +390,12 @@ static ssize_t qeth_l3_dev_ipato_enable_ goto out; } - if (card->ipato.enabled == enable) - goto out; - card->ipato.enabled = enable; - - spin_lock_bh(&card->ip_lock); - hash_for_each(card->ip_htable, i, addr, hnode) { - if (addr->type != QETH_IP_TYPE_NORMAL) - continue; - if (!enable) - addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; - else if (qeth_l3_is_addr_covered_by_ipato(card, addr)) - addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; + if (card->ipato.enabled != enable) { + card->ipato.enabled = enable; + spin_lock_bh(&card->ip_lock); + qeth_l3_update_ipato(card); + spin_unlock_bh(&card->ip_lock); } - spin_unlock_bh(&card->ip_lock); out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; @@ -430,20 +421,27 @@ static ssize_t qeth_l3_dev_ipato_invert4 const char *buf, size_t count) { struct qeth_card *card = dev_get_drvdata(dev); + bool invert; int rc = 0; if (!card) return -EINVAL; mutex_lock(&card->conf_mutex); - if (sysfs_streq(buf, "toggle")) - card->ipato.invert4 = (card->ipato.invert4)? 0 : 1; - else if (sysfs_streq(buf, "1")) - card->ipato.invert4 = 1; - else if (sysfs_streq(buf, "0")) - card->ipato.invert4 = 0; - else + if (sysfs_streq(buf, "toggle")) { + invert = !card->ipato.invert4; + } else if (kstrtobool(buf, &invert)) { rc = -EINVAL; + goto out; + } + + if (card->ipato.invert4 != invert) { + card->ipato.invert4 = invert; + spin_lock_bh(&card->ip_lock); + qeth_l3_update_ipato(card); + spin_unlock_bh(&card->ip_lock); + } +out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; } @@ -609,20 +607,27 @@ static ssize_t qeth_l3_dev_ipato_invert6 struct device_attribute *attr, const char *buf, size_t count) { struct qeth_card *card = dev_get_drvdata(dev); + bool invert; int rc = 0; if (!card) return -EINVAL; mutex_lock(&card->conf_mutex); - if (sysfs_streq(buf, "toggle")) - card->ipato.invert6 = (card->ipato.invert6)? 0 : 1; - else if (sysfs_streq(buf, "1")) - card->ipato.invert6 = 1; - else if (sysfs_streq(buf, "0")) - card->ipato.invert6 = 0; - else + if (sysfs_streq(buf, "toggle")) { + invert = !card->ipato.invert6; + } else if (kstrtobool(buf, &invert)) { rc = -EINVAL; + goto out; + } + + if (card->ipato.invert6 != invert) { + card->ipato.invert6 = invert; + spin_lock_bh(&card->ip_lock); + qeth_l3_update_ipato(card); + spin_unlock_bh(&card->ip_lock); + } +out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; } Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.14/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.14/s390-qeth-fix-error-handling-in-checksum-cmd-callback.patch queue-4.14/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.14/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.14/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: lock IP table while applying takeover changes" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: lock IP table while applying takeover changes to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-lock-ip-table-while-applying-takeover-changes.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:31 +0100 Subject: s390/qeth: lock IP table while applying takeover changes From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit 8a03a3692b100d84785ee7a834e9215e304c9e00 ] Modifying the flags of an IP addr object needs to be protected against eg. concurrent removal of the same object from the IP table. Fixes: 5f78e29ceebf ("qeth: optimize IP handling in rx_mode callback") Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_l3_sys.c | 2 ++ 1 file changed, 2 insertions(+) --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -395,6 +395,7 @@ static ssize_t qeth_l3_dev_ipato_enable_ goto out; card->ipato.enabled = enable; + spin_lock_bh(&card->ip_lock); hash_for_each(card->ip_htable, i, addr, hnode) { if (addr->type != QETH_IP_TYPE_NORMAL) continue; @@ -403,6 +404,7 @@ static ssize_t qeth_l3_dev_ipato_enable_ else if (qeth_l3_is_addr_covered_by_ipato(card, addr)) addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; } + spin_unlock_bh(&card->ip_lock); out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.14/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.14/s390-qeth-fix-error-handling-in-checksum-cmd-callback.patch queue-4.14/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.14/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.14/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: fix error handling in checksum cmd callback" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: fix error handling in checksum cmd callback to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-fix-error-handling-in-checksum-cmd-callback.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 20 Dec 2017 18:07:18 +0100 Subject: s390/qeth: fix error handling in checksum cmd callback From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit ad3cbf61332914711e5f506972b1dc9af8d62146 ] Make sure to check both return code fields before processing the response. Otherwise we risk operating on invalid data. Fixes: c9475369bd2b ("s390/qeth: rework RX/TX checksum offload") Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_core_main.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -5445,6 +5445,13 @@ out: } EXPORT_SYMBOL_GPL(qeth_poll); +static int qeth_setassparms_inspect_rc(struct qeth_ipa_cmd *cmd) +{ + if (!cmd->hdr.return_code) + cmd->hdr.return_code = cmd->data.setassparms.hdr.return_code; + return cmd->hdr.return_code; +} + int qeth_setassparms_cb(struct qeth_card *card, struct qeth_reply *reply, unsigned long data) { @@ -6304,7 +6311,7 @@ static int qeth_ipa_checksum_run_cmd_cb( (struct qeth_checksum_cmd *)reply->param; QETH_CARD_TEXT(card, 4, "chkdoccb"); - if (cmd->hdr.return_code) + if (qeth_setassparms_inspect_rc(cmd)) return 0; memset(chksum_cb, 0, sizeof(*chksum_cb)); Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.14/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.14/s390-qeth-fix-error-handling-in-checksum-cmd-callback.patch queue-4.14/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.14/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.14/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: don't apply takeover changes to RXIP" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: don't apply takeover changes to RXIP to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-don-t-apply-takeover-changes-to-rxip.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:30 +0100 Subject: s390/qeth: don't apply takeover changes to RXIP From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit b22d73d6689fd902a66c08ebe71ab2f3b351e22f ] When takeover is switched off, current code clears the 'TAKEOVER' flag on all IPs. But the flag is also used for RXIP addresses, and those should not be affected by the takeover mode. Fix the behaviour by consistenly applying takover logic to NORMAL addresses only. Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_l3_main.c | 5 +++-- drivers/s390/net/qeth_l3_sys.c | 5 +++-- 2 files changed, 6 insertions(+), 4 deletions(-) --- a/drivers/s390/net/qeth_l3_main.c +++ b/drivers/s390/net/qeth_l3_main.c @@ -173,6 +173,8 @@ int qeth_l3_is_addr_covered_by_ipato(str if (!card->ipato.enabled) return 0; + if (addr->type != QETH_IP_TYPE_NORMAL) + return 0; qeth_l3_convert_addr_to_bits((u8 *) &addr->u, addr_bits, (addr->proto == QETH_PROT_IPV4)? 4:16); @@ -289,8 +291,7 @@ int qeth_l3_add_ip(struct qeth_card *car memcpy(addr, tmp_addr, sizeof(struct qeth_ipaddr)); addr->ref_counter = 1; - if (addr->type == QETH_IP_TYPE_NORMAL && - qeth_l3_is_addr_covered_by_ipato(card, addr)) { + if (qeth_l3_is_addr_covered_by_ipato(card, addr)) { QETH_CARD_TEXT(card, 2, "tkovaddr"); addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; } --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -396,10 +396,11 @@ static ssize_t qeth_l3_dev_ipato_enable_ card->ipato.enabled = enable; hash_for_each(card->ip_htable, i, addr, hnode) { + if (addr->type != QETH_IP_TYPE_NORMAL) + continue; if (!enable) addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; - else if (addr->type == QETH_IP_TYPE_NORMAL && - qeth_l3_is_addr_covered_by_ipato(card, addr)) + else if (qeth_l3_is_addr_covered_by_ipato(card, addr)) addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; } out: Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.14/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.14/s390-qeth-fix-error-handling-in-checksum-cmd-callback.patch queue-4.14/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.14/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.14/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "s390/qeth: apply takeover changes when mode is toggled" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled s390/qeth: apply takeover changes when mode is toggled to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Date: Wed, 13 Dec 2017 18:56:29 +0100 Subject: s390/qeth: apply takeover changes when mode is toggled From: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> [ Upstream commit 7fbd9493f0eeae8cef58300505a9ef5c8fce6313 ] Just as for an explicit enable/disable, toggling the takeover mode also requires that the IP addresses get updated. Otherwise all IPs that were added to the table before the mode-toggle, get registered with the old settings. Signed-off-by: Julian Wiedmann <jwi(a)linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/s390/net/qeth_core.h | 2 +- drivers/s390/net/qeth_core_main.c | 2 +- drivers/s390/net/qeth_l3_sys.c | 35 +++++++++++++++++------------------ 3 files changed, 19 insertions(+), 20 deletions(-) --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -564,7 +564,7 @@ enum qeth_cq { }; struct qeth_ipato { - int enabled; + bool enabled; int invert4; int invert6; struct list_head entries; --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -1479,7 +1479,7 @@ static int qeth_setup_card(struct qeth_c qeth_set_intial_options(card); /* IP address takeover */ INIT_LIST_HEAD(&card->ipato.entries); - card->ipato.enabled = 0; + card->ipato.enabled = false; card->ipato.invert4 = 0; card->ipato.invert6 = 0; /* init QDIO stuff */ --- a/drivers/s390/net/qeth_l3_sys.c +++ b/drivers/s390/net/qeth_l3_sys.c @@ -372,6 +372,7 @@ static ssize_t qeth_l3_dev_ipato_enable_ struct qeth_card *card = dev_get_drvdata(dev); struct qeth_ipaddr *addr; int i, rc = 0; + bool enable; if (!card) return -EINVAL; @@ -384,25 +385,23 @@ static ssize_t qeth_l3_dev_ipato_enable_ } if (sysfs_streq(buf, "toggle")) { - card->ipato.enabled = (card->ipato.enabled)? 0 : 1; - } else if (sysfs_streq(buf, "1")) { - card->ipato.enabled = 1; - hash_for_each(card->ip_htable, i, addr, hnode) { - if ((addr->type == QETH_IP_TYPE_NORMAL) && - qeth_l3_is_addr_covered_by_ipato(card, addr)) - addr->set_flags |= - QETH_IPA_SETIP_TAKEOVER_FLAG; - } - } else if (sysfs_streq(buf, "0")) { - card->ipato.enabled = 0; - hash_for_each(card->ip_htable, i, addr, hnode) { - if (addr->set_flags & - QETH_IPA_SETIP_TAKEOVER_FLAG) - addr->set_flags &= - ~QETH_IPA_SETIP_TAKEOVER_FLAG; - } - } else + enable = !card->ipato.enabled; + } else if (kstrtobool(buf, &enable)) { rc = -EINVAL; + goto out; + } + + if (card->ipato.enabled == enable) + goto out; + card->ipato.enabled = enable; + + hash_for_each(card->ip_htable, i, addr, hnode) { + if (!enable) + addr->set_flags &= ~QETH_IPA_SETIP_TAKEOVER_FLAG; + else if (addr->type == QETH_IP_TYPE_NORMAL && + qeth_l3_is_addr_covered_by_ipato(card, addr)) + addr->set_flags |= QETH_IPA_SETIP_TAKEOVER_FLAG; + } out: mutex_unlock(&card->conf_mutex); return rc ? rc : count; Patches currently in stable-queue which might be from jwi(a)linux.vnet.ibm.com are queue-4.14/s390-qeth-lock-ip-table-while-applying-takeover-changes.patch queue-4.14/s390-qeth-fix-error-handling-in-checksum-cmd-callback.patch queue-4.14/s390-qeth-apply-takeover-changes-when-mode-is-toggled.patch queue-4.14/s390-qeth-update-takeover-ips-after-configuration-change.patch queue-4.14/s390-qeth-don-t-apply-takeover-changes-to-rxip.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "RDS: Check cmsg_len before dereferencing CMSG_DATA" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled RDS: Check cmsg_len before dereferencing CMSG_DATA to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: rds-check-cmsg_len-before-dereferencing-cmsg_data.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Avinash Repaka <avinash.repaka(a)oracle.com> Date: Thu, 21 Dec 2017 20:17:04 -0800 Subject: RDS: Check cmsg_len before dereferencing CMSG_DATA From: Avinash Repaka <avinash.repaka(a)oracle.com> [ Upstream commit 14e138a86f6347c6199f610576d2e11c03bec5f0 ] RDS currently doesn't check if the length of the control message is large enough to hold the required data, before dereferencing the control message data. This results in following crash: BUG: KASAN: stack-out-of-bounds in rds_rdma_bytes net/rds/send.c:1013 [inline] BUG: KASAN: stack-out-of-bounds in rds_sendmsg+0x1f02/0x1f90 net/rds/send.c:1066 Read of size 8 at addr ffff8801c928fb70 by task syzkaller455006/3157 CPU: 0 PID: 3157 Comm: syzkaller455006 Not tainted 4.15.0-rc3+ #161 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:17 [inline] dump_stack+0x194/0x257 lib/dump_stack.c:53 print_address_description+0x73/0x250 mm/kasan/report.c:252 kasan_report_error mm/kasan/report.c:351 [inline] kasan_report+0x25b/0x340 mm/kasan/report.c:409 __asan_report_load8_noabort+0x14/0x20 mm/kasan/report.c:430 rds_rdma_bytes net/rds/send.c:1013 [inline] rds_sendmsg+0x1f02/0x1f90 net/rds/send.c:1066 sock_sendmsg_nosec net/socket.c:628 [inline] sock_sendmsg+0xca/0x110 net/socket.c:638 ___sys_sendmsg+0x320/0x8b0 net/socket.c:2018 __sys_sendmmsg+0x1ee/0x620 net/socket.c:2108 SYSC_sendmmsg net/socket.c:2139 [inline] SyS_sendmmsg+0x35/0x60 net/socket.c:2134 entry_SYSCALL_64_fastpath+0x1f/0x96 RIP: 0033:0x43fe49 RSP: 002b:00007fffbe244ad8 EFLAGS: 00000217 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00000000004002c8 RCX: 000000000043fe49 RDX: 0000000000000001 RSI: 000000002020c000 RDI: 0000000000000003 RBP: 00000000006ca018 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000217 R12: 00000000004017b0 R13: 0000000000401840 R14: 0000000000000000 R15: 0000000000000000 To fix this, we verify that the cmsg_len is large enough to hold the data to be read, before proceeding further. Reported-by: syzbot <syzkaller-bugs(a)googlegroups.com> Signed-off-by: Avinash Repaka <avinash.repaka(a)oracle.com> Acked-by: Santosh Shilimkar <santosh.shilimkar(a)oracle.com> Reviewed-by: Yuval Shaia <yuval.shaia(a)oracle.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/rds/send.c | 3 +++ 1 file changed, 3 insertions(+) --- a/net/rds/send.c +++ b/net/rds/send.c @@ -1009,6 +1009,9 @@ static int rds_rdma_bytes(struct msghdr continue; if (cmsg->cmsg_type == RDS_CMSG_RDMA_ARGS) { + if (cmsg->cmsg_len < + CMSG_LEN(sizeof(struct rds_rdma_args))) + return -EINVAL; args = CMSG_DATA(cmsg); *rdma_bytes += args->remote_vec.bytes; } Patches currently in stable-queue which might be from avinash.repaka(a)oracle.com are queue-4.14/rds-check-cmsg_len-before-dereferencing-cmsg_data.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ptr_ring: add barriers" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ptr_ring: add barriers to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ptr_ring-add-barriers.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: "Michael S. Tsirkin" <mst(a)redhat.com> Date: Tue, 5 Dec 2017 21:29:37 +0200 Subject: ptr_ring: add barriers From: "Michael S. Tsirkin" <mst(a)redhat.com> [ Upstream commit a8ceb5dbfde1092b466936bca0ff3be127ecf38e ] Users of ptr_ring expect that it's safe to give the data structure a pointer and have it be available to consumers, but that actually requires an smb_wmb or a stronger barrier. In absence of such barriers and on architectures that reorder writes, consumer might read an un=initialized value from an skb pointer stored in the skb array. This was observed causing crashes. To fix, add memory barriers. The barrier we use is a wmb, the assumption being that producers do not need to read the value so we do not need to order these reads. Reported-by: George Cherian <george.cherian(a)cavium.com> Suggested-by: Jason Wang <jasowang(a)redhat.com> Signed-off-by: Michael S. Tsirkin <mst(a)redhat.com> Acked-by: Jason Wang <jasowang(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/ptr_ring.h | 9 +++++++++ 1 file changed, 9 insertions(+) --- a/include/linux/ptr_ring.h +++ b/include/linux/ptr_ring.h @@ -101,12 +101,18 @@ static inline bool ptr_ring_full_bh(stru /* Note: callers invoking this in a loop must use a compiler barrier, * for example cpu_relax(). Callers must hold producer_lock. + * Callers are responsible for making sure pointer that is being queued + * points to a valid data. */ static inline int __ptr_ring_produce(struct ptr_ring *r, void *ptr) { if (unlikely(!r->size) || r->queue[r->producer]) return -ENOSPC; + /* Make sure the pointer we are storing points to a valid data. */ + /* Pairs with smp_read_barrier_depends in __ptr_ring_consume. */ + smp_wmb(); + r->queue[r->producer++] = ptr; if (unlikely(r->producer >= r->size)) r->producer = 0; @@ -275,6 +281,9 @@ static inline void *__ptr_ring_consume(s if (ptr) __ptr_ring_discard_one(r); + /* Make sure anyone accessing data through the pointer is up to date. */ + /* Pairs with smp_wmb in __ptr_ring_produce. */ + smp_read_barrier_depends(); return ptr; } Patches currently in stable-queue which might be from mst(a)redhat.com are queue-4.14/ptr_ring-add-barriers.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "phylink: ensure the PHY interface mode is appropriately set" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled phylink: ensure the PHY interface mode is appropriately set to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: phylink-ensure-the-phy-interface-mode-is-appropriately-set.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Russell King <rmk+kernel(a)armlinux.org.uk> Date: Wed, 20 Dec 2017 23:21:28 +0000 Subject: phylink: ensure the PHY interface mode is appropriately set From: Russell King <rmk+kernel(a)armlinux.org.uk> [ Upstream commit 182088aa3c6c7f7c20a2c1dcc9ded4a3fc631f38 ] When setting the ethtool settings, ensure that the validated PHY interface mode is propagated to the current link settings, so that 2500BaseX can be selected. Fixes: 9525ae83959b ("phylink: add phylink infrastructure") Signed-off-by: Russell King <rmk+kernel(a)armlinux.org.uk> Reviewed-by: Florian Fainelli <f.fainelli(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/phy/phylink.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -948,6 +948,7 @@ int phylink_ethtool_ksettings_set(struct mutex_lock(&pl->state_mutex); /* Configure the MAC to match the new settings */ linkmode_copy(pl->link_config.advertising, our_kset.link_modes.advertising); + pl->link_config.interface = config.interface; pl->link_config.speed = our_kset.base.speed; pl->link_config.duplex = our_kset.base.duplex; pl->link_config.an_enabled = our_kset.base.autoneg != AUTONEG_DISABLE; Patches currently in stable-queue which might be from rmk+kernel(a)armlinux.org.uk are queue-4.14/phylink-ensure-the-phy-interface-mode-is-appropriately-set.patch queue-4.14/phylink-ensure-an-is-enabled.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "phylink: ensure AN is enabled" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled phylink: ensure AN is enabled to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: phylink-ensure-an-is-enabled.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Russell King <rmk+kernel(a)armlinux.org.uk> Date: Wed, 20 Dec 2017 23:21:34 +0000 Subject: phylink: ensure AN is enabled From: Russell King <rmk+kernel(a)armlinux.org.uk> [ Upstream commit 74ee0e8c1bf9925c59cc8f1c65c29adf6e4cf603 ] Ensure that we mark AN as enabled at boot time, rather than leaving it disabled. This is noticable if your SFP module is fiber, and it supports faster speeds than 1G with 2.5G support in place. Fixes: 9525ae83959b ("phylink: add phylink infrastructure") Signed-off-by: Russell King <rmk+kernel(a)armlinux.org.uk> Reviewed-by: Florian Fainelli <f.fainelli(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/phy/phylink.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -525,6 +525,7 @@ struct phylink *phylink_create(struct ne pl->link_config.pause = MLO_PAUSE_AN; pl->link_config.speed = SPEED_UNKNOWN; pl->link_config.duplex = DUPLEX_UNKNOWN; + pl->link_config.an_enabled = true; pl->ops = ops; __set_bit(PHYLINK_DISABLE_STOPPED, &pl->phylink_disable_state); Patches currently in stable-queue which might be from rmk+kernel(a)armlinux.org.uk are queue-4.14/phylink-ensure-the-phy-interface-mode-is-appropriately-set.patch queue-4.14/phylink-ensure-an-is-enabled.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "openvswitch: Fix pop_vlan action for double tagged frames" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled openvswitch: Fix pop_vlan action for double tagged frames to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: openvswitch-fix-pop_vlan-action-for-double-tagged-frames.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Eric Garver <e(a)erig.me> Date: Wed, 20 Dec 2017 15:09:22 -0500 Subject: openvswitch: Fix pop_vlan action for double tagged frames From: Eric Garver <e(a)erig.me> [ Upstream commit c48e74736fccf25fb32bb015426359e1c2016e3b ] skb_vlan_pop() expects skb->protocol to be a valid TPID for double tagged frames. So set skb->protocol to the TPID and let skb_vlan_pop() shift the true ethertype into position for us. Fixes: 5108bbaddc37 ("openvswitch: add processing of L3 packets") Signed-off-by: Eric Garver <e(a)erig.me> Reviewed-by: Jiri Benc <jbenc(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/openvswitch/flow.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) --- a/net/openvswitch/flow.c +++ b/net/openvswitch/flow.c @@ -532,6 +532,7 @@ static int key_extract(struct sk_buff *s return -EINVAL; skb_reset_network_header(skb); + key->eth.type = skb->protocol; } else { eth = eth_hdr(skb); ether_addr_copy(key->eth.src, eth->h_source); @@ -545,15 +546,23 @@ static int key_extract(struct sk_buff *s if (unlikely(parse_vlan(skb, key))) return -ENOMEM; - skb->protocol = parse_ethertype(skb); - if (unlikely(skb->protocol == htons(0))) + key->eth.type = parse_ethertype(skb); + if (unlikely(key->eth.type == htons(0))) return -ENOMEM; + /* Multiple tagged packets need to retain TPID to satisfy + * skb_vlan_pop(), which will later shift the ethertype into + * skb->protocol. + */ + if (key->eth.cvlan.tci & htons(VLAN_TAG_PRESENT)) + skb->protocol = key->eth.cvlan.tpid; + else + skb->protocol = key->eth.type; + skb_reset_network_header(skb); __skb_push(skb, skb->data - skb_mac_header(skb)); } skb_reset_mac_len(skb); - key->eth.type = skb->protocol; /* Network layer. */ if (key->eth.type == htons(ETH_P_IP)) { Patches currently in stable-queue which might be from e(a)erig.me are queue-4.14/openvswitch-fix-pop_vlan-action-for-double-tagged-frames.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "netlink: Add netns check on taps" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled netlink: Add netns check on taps to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: netlink-add-netns-check-on-taps.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Wed, 6 Dec 2017 12:12:27 -0800 Subject: netlink: Add netns check on taps From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit 93c647643b48f0131f02e45da3bd367d80443291 ] Currently, a nlmon link inside a child namespace can observe systemwide netlink activity. Filter the traffic so that nlmon can only sniff netlink messages from its own netns. Test case: vpnns -- bash -c "ip link add nlmon0 type nlmon; \ ip link set nlmon0 up; \ tcpdump -i nlmon0 -q -w /tmp/nlmon.pcap -U" & sudo ip xfrm state add src 10.1.1.1 dst 10.1.1.2 proto esp \ spi 0x1 mode transport \ auth sha1 0x6162633132330000000000000000000000000000 \ enc aes 0x00000000000000000000000000000000 grep --binary abc123 /tmp/nlmon.pcap Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/netlink/af_netlink.c | 3 +++ 1 file changed, 3 insertions(+) --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -254,6 +254,9 @@ static int __netlink_deliver_tap_skb(str struct sock *sk = skb->sk; int ret = -ENOMEM; + if (!net_eq(dev_net(dev), sock_net(sk))) + return 0; + dev_hold(dev); if (is_vmalloc_addr(skb->head)) Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-4.14/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.14/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: sched: fix static key imbalance in case of ingress/clsact_init error" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: sched: fix static key imbalance in case of ingress/clsact_init error to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-sched-fix-static-key-imbalance-in-case-of-ingress-clsact_init-error.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Jiri Pirko <jiri(a)mellanox.com> Date: Fri, 15 Dec 2017 12:40:13 +0100 Subject: net: sched: fix static key imbalance in case of ingress/clsact_init error From: Jiri Pirko <jiri(a)mellanox.com> [ Upstream commit b59e6979a86384e68b0ab6ffeab11f0034fba82d ] Move static key increments to the beginning of the init function so they pair 1:1 with decrements in ingress/clsact_destroy, which is called in case ingress/clsact_init fails. Fixes: 6529eaba33f0 ("net: sched: introduce tcf block infractructure") Signed-off-by: Jiri Pirko <jiri(a)mellanox.com> Acked-by: Cong Wang <xiyou.wangcong(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/sched/sch_ingress.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) --- a/net/sched/sch_ingress.c +++ b/net/sched/sch_ingress.c @@ -59,11 +59,12 @@ static int ingress_init(struct Qdisc *sc struct net_device *dev = qdisc_dev(sch); int err; + net_inc_ingress_queue(); + err = tcf_block_get(&q->block, &dev->ingress_cl_list); if (err) return err; - net_inc_ingress_queue(); sch->flags |= TCQ_F_CPUSTATS; return 0; @@ -153,6 +154,9 @@ static int clsact_init(struct Qdisc *sch struct net_device *dev = qdisc_dev(sch); int err; + net_inc_ingress_queue(); + net_inc_egress_queue(); + err = tcf_block_get(&q->ingress_block, &dev->ingress_cl_list); if (err) return err; @@ -161,9 +165,6 @@ static int clsact_init(struct Qdisc *sch if (err) return err; - net_inc_ingress_queue(); - net_inc_egress_queue(); - sch->flags |= TCQ_F_CPUSTATS; return 0; Patches currently in stable-queue which might be from jiri(a)mellanox.com are queue-4.14/net-sched-fix-static-key-imbalance-in-case-of-ingress-clsact_init-error.patch queue-4.14/mlxsw-spectrum-disable-mac-learning-for-ovs-port.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: reevalulate autoflowlabel setting after sysctl setting" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: reevalulate autoflowlabel setting after sysctl setting to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-reevalulate-autoflowlabel-setting-after-sysctl-setting.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Shaohua Li <shli(a)fb.com> Date: Wed, 20 Dec 2017 12:10:21 -0800 Subject: net: reevalulate autoflowlabel setting after sysctl setting From: Shaohua Li <shli(a)fb.com> [ Upstream commit 513674b5a2c9c7a67501506419da5c3c77ac6f08 ] sysctl.ip6.auto_flowlabels is default 1. In our hosts, we set it to 2. If sockopt doesn't set autoflowlabel, outcome packets from the hosts are supposed to not include flowlabel. This is true for normal packet, but not for reset packet. The reason is ipv6_pinfo.autoflowlabel is set in sock creation. Later if we change sysctl.ip6.auto_flowlabels, the ipv6_pinfo.autoflowlabel isn't changed, so the sock will keep the old behavior in terms of auto flowlabel. Reset packet is suffering from this problem, because reset packet is sent from a special control socket, which is created at boot time. Since sysctl.ipv6.auto_flowlabels is 1 by default, the control socket will always have its ipv6_pinfo.autoflowlabel set, even after user set sysctl.ipv6.auto_flowlabels to 1, so reset packset will always have flowlabel. Normal sock created before sysctl setting suffers from the same issue. We can't even turn off autoflowlabel unless we kill all socks in the hosts. To fix this, if IPV6_AUTOFLOWLABEL sockopt is used, we use the autoflowlabel setting from user, otherwise we always call ip6_default_np_autolabel() which has the new settings of sysctl. Note, this changes behavior a little bit. Before commit 42240901f7c4 (ipv6: Implement different admin modes for automatic flow labels), the autoflowlabel behavior of a sock isn't sticky, eg, if sysctl changes, existing connection will change autoflowlabel behavior. After that commit, autoflowlabel behavior is sticky in the whole life of the sock. With this patch, the behavior isn't sticky again. Cc: Martin KaFai Lau <kafai(a)fb.com> Cc: Eric Dumazet <eric.dumazet(a)gmail.com> Cc: Tom Herbert <tom(a)quantonium.net> Signed-off-by: Shaohua Li <shli(a)fb.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/ipv6.h | 3 ++- net/ipv6/af_inet6.c | 1 - net/ipv6/ip6_output.c | 12 ++++++++++-- net/ipv6/ipv6_sockglue.c | 1 + 4 files changed, 13 insertions(+), 4 deletions(-) --- a/include/linux/ipv6.h +++ b/include/linux/ipv6.h @@ -272,7 +272,8 @@ struct ipv6_pinfo { * 100: prefer care-of address */ dontfrag:1, - autoflowlabel:1; + autoflowlabel:1, + autoflowlabel_set:1; __u8 min_hopcount; __u8 tclass; __be32 rcv_flowinfo; --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -210,7 +210,6 @@ lookup_protocol: np->mcast_hops = IPV6_DEFAULT_MCASTHOPS; np->mc_loop = 1; np->pmtudisc = IPV6_PMTUDISC_WANT; - np->autoflowlabel = ip6_default_np_autolabel(net); np->repflow = net->ipv6.sysctl.flowlabel_reflect; sk->sk_ipv6only = net->ipv6.sysctl.bindv6only; --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -166,6 +166,14 @@ int ip6_output(struct net *net, struct s !(IP6CB(skb)->flags & IP6SKB_REROUTED)); } +static bool ip6_autoflowlabel(struct net *net, const struct ipv6_pinfo *np) +{ + if (!np->autoflowlabel_set) + return ip6_default_np_autolabel(net); + else + return np->autoflowlabel; +} + /* * xmit an sk_buff (used by TCP, SCTP and DCCP) * Note : socket lock is not held for SYNACK packets, but might be modified @@ -230,7 +238,7 @@ int ip6_xmit(const struct sock *sk, stru hlimit = ip6_dst_hoplimit(dst); ip6_flow_hdr(hdr, tclass, ip6_make_flowlabel(net, skb, fl6->flowlabel, - np->autoflowlabel, fl6)); + ip6_autoflowlabel(net, np), fl6)); hdr->payload_len = htons(seg_len); hdr->nexthdr = proto; @@ -1626,7 +1634,7 @@ struct sk_buff *__ip6_make_skb(struct so ip6_flow_hdr(hdr, v6_cork->tclass, ip6_make_flowlabel(net, skb, fl6->flowlabel, - np->autoflowlabel, fl6)); + ip6_autoflowlabel(net, np), fl6)); hdr->hop_limit = v6_cork->hop_limit; hdr->nexthdr = proto; hdr->saddr = fl6->saddr; --- a/net/ipv6/ipv6_sockglue.c +++ b/net/ipv6/ipv6_sockglue.c @@ -878,6 +878,7 @@ pref_skip_coa: break; case IPV6_AUTOFLOWLABEL: np->autoflowlabel = valbool; + np->autoflowlabel_set = 1; retv = 0; break; case IPV6_RECVFRAGSIZE: Patches currently in stable-queue which might be from shli(a)fb.com are queue-4.14/net-reevalulate-autoflowlabel-setting-after-sysctl-setting.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: qmi_wwan: add Sierra EM7565 1199:9091" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: qmi_wwan: add Sierra EM7565 1199:9091 to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-qmi_wwan-add-sierra-em7565-1199-9091.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Sebastian Sjoholm <ssjoholm(a)mac.com> Date: Mon, 11 Dec 2017 21:51:14 +0100 Subject: net: qmi_wwan: add Sierra EM7565 1199:9091 From: Sebastian Sjoholm <ssjoholm(a)mac.com> [ Upstream commit aceef61ee56898cfa7b6960fb60b9326c3860441 ] Sierra Wireless EM7565 is an Qualcomm MDM9x50 based M.2 modem. The USB id is added to qmi_wwan.c to allow QMI communication with the EM7565. Signed-off-by: Sebastian Sjoholm <ssjoholm(a)mac.com> Acked-by: Bjørn Mork <bjorn(a)mork.no> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -1204,6 +1204,7 @@ static const struct usb_device_id produc {QMI_FIXED_INTF(0x1199, 0x9079, 10)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 8)}, /* Sierra Wireless EM74xx */ {QMI_FIXED_INTF(0x1199, 0x907b, 10)}, /* Sierra Wireless EM74xx */ + {QMI_FIXED_INTF(0x1199, 0x9091, 8)}, /* Sierra Wireless EM7565 */ {QMI_FIXED_INTF(0x1bbb, 0x011e, 4)}, /* Telekom Speedstick LTE II (Alcatel One Touch L100V LTE) */ {QMI_FIXED_INTF(0x1bbb, 0x0203, 2)}, /* Alcatel L800MA */ {QMI_FIXED_INTF(0x2357, 0x0201, 4)}, /* TP-LINK HSUPA Modem MA180 */ Patches currently in stable-queue which might be from ssjoholm(a)mac.com are queue-4.14/net-qmi_wwan-add-sierra-em7565-1199-9091.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-phy-micrel-ksz9031-reconfigure-autoneg-after-phy-autoneg-workaround.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Grygorii Strashko <grygorii.strashko(a)ti.com> Date: Wed, 20 Dec 2017 18:45:10 -0600 Subject: net: phy: micrel: ksz9031: reconfigure autoneg after phy autoneg workaround From: Grygorii Strashko <grygorii.strashko(a)ti.com> [ Upstream commit c1a8d0a3accf64a014d605e6806ce05d1c17adf1 ] Under some circumstances driver will perform PHY reset in ksz9031_read_status() to fix autoneg failure case (idle error count = 0xFF). When this happens ksz9031 will not detect link status change any more when connecting to Netgear 1G switch (link can be recovered sometimes by restarting netdevice "ifconfig down up"). Reproduced with TI am572x board equipped with ksz9031 PHY while connecting to Netgear 1G switch. Fix the issue by reconfiguring autonegotiation after PHY reset in ksz9031_read_status(). Fixes: d2fd719bcb0e ("net/phy: micrel: Add workaround for bad autoneg") Signed-off-by: Grygorii Strashko <grygorii.strashko(a)ti.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/phy/micrel.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -622,6 +622,7 @@ static int ksz9031_read_status(struct ph phydev->link = 0; if (phydev->drv->config_intr && phy_interrupt_is_valid(phydev)) phydev->drv->config_intr(phydev); + return genphy_config_aneg(phydev); } return 0; Patches currently in stable-queue which might be from grygorii.strashko(a)ti.com are queue-4.14/net-phy-micrel-ksz9031-reconfigure-autoneg-after-phy-autoneg-workaround.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: phy: marvell: Limit 88m1101 autoneg errata to 88E1145 as well." has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: phy: marvell: Limit 88m1101 autoneg errata to 88E1145 as well. to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-phy-marvell-limit-88m1101-autoneg-errata-to-88e1145-as-well.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Zhao Qiang <qiang.zhao(a)nxp.com> Date: Mon, 18 Dec 2017 10:26:43 +0800 Subject: net: phy: marvell: Limit 88m1101 autoneg errata to 88E1145 as well. From: Zhao Qiang <qiang.zhao(a)nxp.com> [ Upstream commit c505873eaece2b4aefd07d339dc7e1400e0235ac ] 88E1145 also need this autoneg errata. Fixes: f2899788353c ("net: phy: marvell: Limit errata to 88m1101") Signed-off-by: Zhao Qiang <qiang.zhao(a)nxp.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/phy/marvell.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/net/phy/marvell.c +++ b/drivers/net/phy/marvell.c @@ -2069,7 +2069,7 @@ static struct phy_driver marvell_drivers .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, .config_init = &m88e1145_config_init, - .config_aneg = &marvell_config_aneg, + .config_aneg = &m88e1101_config_aneg, .read_status = &genphy_read_status, .ack_interrupt = &marvell_ack_interrupt, .config_intr = &marvell_config_intr, Patches currently in stable-queue which might be from qiang.zhao(a)nxp.com are queue-4.14/net-phy-marvell-limit-88m1101-autoneg-errata-to-88e1145-as-well.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Date: Wed, 6 Dec 2017 15:23:23 +0100 Subject: net: mvmdio: disable/unprepare clocks in EPROBE_DEFER case From: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> [ Upstream commit 589bf32f09852041fbd3b7ce1a9e703f95c230ba ] add appropriate calls to clk_disable_unprepare() by jumping to out_mdio in case orion_mdio_probe() returns -EPROBE_DEFER. Found by Linux Driver Verification project (linuxtesting.org). Fixes: 3d604da1e954 ("net: mvmdio: get and enable optional clock") Signed-off-by: Tobias Jordan <Tobias.Jordan(a)elektrobit.com> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/marvell/mvmdio.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/marvell/mvmdio.c +++ b/drivers/net/ethernet/marvell/mvmdio.c @@ -344,7 +344,8 @@ static int orion_mdio_probe(struct platf dev->regs + MVMDIO_ERR_INT_MASK); } else if (dev->err_interrupt == -EPROBE_DEFER) { - return -EPROBE_DEFER; + ret = -EPROBE_DEFER; + goto out_mdio; } if (pdev->dev.of_node) Patches currently in stable-queue which might be from Tobias.Jordan(a)elektrobit.com are queue-4.14/net-mvmdio-disable-unprepare-clocks-in-eprobe_defer-case.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Prevent possible races in VXLAN control flow" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Prevent possible races in VXLAN control flow to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Mon, 4 Dec 2017 09:57:43 +0200 Subject: net/mlx5e: Prevent possible races in VXLAN control flow From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 0c1cc8b2215f5122ca614b5adca60346018758c3 ] When calling add/remove VXLAN port, a lock must be held in order to prevent race scenarios when more than one add/remove happens at the same time. Fix by holding our state_lock (mutex) as done by all other parts of the driver. Note that the spinlock protecting the radix-tree is still needed in order to synchronize radix-tree access from softirq context. Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/vxlan.c | 4 ++++ 1 file changed, 4 insertions(+) --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c @@ -88,6 +88,7 @@ static void mlx5e_vxlan_add_port(struct struct mlx5e_vxlan *vxlan; int err; + mutex_lock(&priv->state_lock); vxlan = mlx5e_vxlan_lookup_port(priv, port); if (vxlan) { atomic_inc(&vxlan->refcount); @@ -117,6 +118,7 @@ err_free: err_delete_port: mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); free_work: + mutex_unlock(&priv->state_lock); kfree(vxlan_work); } @@ -130,6 +132,7 @@ static void mlx5e_vxlan_del_port(struct struct mlx5e_vxlan *vxlan; bool remove = false; + mutex_lock(&priv->state_lock); spin_lock_bh(&vxlan_db->lock); vxlan = radix_tree_lookup(&vxlan_db->tree, port); if (!vxlan) @@ -147,6 +150,7 @@ out_unlock: mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); kfree(vxlan); } + mutex_unlock(&priv->state_lock); kfree(vxlan_work); } Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.14/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.14/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.14/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.14/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Fix possible deadlock of VXLAN lock" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Fix possible deadlock of VXLAN lock to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Thu, 23 Nov 2017 13:52:28 +0200 Subject: net/mlx5e: Fix possible deadlock of VXLAN lock From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 6323514116404cc651df1b7fffa1311ddf8ce647 ] mlx5e_vxlan_lookup_port is called both from mlx5e_add_vxlan_port (user context) and mlx5e_features_check (softirq), but the lock acquired does not disable bottom half and might result in deadlock. Fix it by simply replacing spin_lock() with spin_lock_bh(). While at it, replace all unnecessary spin_lock_irq() to spin_lock_bh(). lockdep's WARNING: inconsistent lock state [ 654.028136] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. [ 654.028229] swapper/5/0 [HC0[0]:SC1[9]:HE1:SE0] takes: [ 654.028321] (&(&vxlan_db->lock)->rlock){+.?.}, at: [<ffffffffa06e7f0e>] mlx5e_vxlan_lookup_port+0x1e/0x50 [mlx5_core] [ 654.028528] {SOFTIRQ-ON-W} state was registered at: [ 654.028607] _raw_spin_lock+0x3c/0x70 [ 654.028689] mlx5e_vxlan_lookup_port+0x1e/0x50 [mlx5_core] [ 654.028794] mlx5e_vxlan_add_port+0x2e/0x120 [mlx5_core] [ 654.028878] process_one_work+0x1e9/0x640 [ 654.028942] worker_thread+0x4a/0x3f0 [ 654.029002] kthread+0x141/0x180 [ 654.029056] ret_from_fork+0x24/0x30 [ 654.029114] irq event stamp: 579088 [ 654.029174] hardirqs last enabled at (579088): [<ffffffff818f475a>] ip6_finish_output2+0x49a/0x8c0 [ 654.029309] hardirqs last disabled at (579087): [<ffffffff818f470e>] ip6_finish_output2+0x44e/0x8c0 [ 654.029446] softirqs last enabled at (579030): [<ffffffff810b3b3d>] irq_enter+0x6d/0x80 [ 654.029567] softirqs last disabled at (579031): [<ffffffff810b3c05>] irq_exit+0xb5/0xc0 [ 654.029684] other info that might help us debug this: [ 654.029781] Possible unsafe locking scenario: [ 654.029868] CPU0 [ 654.029908] ---- [ 654.029947] lock(&(&vxlan_db->lock)->rlock); [ 654.030045] <Interrupt> [ 654.030090] lock(&(&vxlan_db->lock)->rlock); [ 654.030162] *** DEADLOCK *** Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/vxlan.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c @@ -71,9 +71,9 @@ struct mlx5e_vxlan *mlx5e_vxlan_lookup_p struct mlx5e_vxlan_db *vxlan_db = &priv->vxlan; struct mlx5e_vxlan *vxlan; - spin_lock(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); vxlan = radix_tree_lookup(&vxlan_db->tree, port); - spin_unlock(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); return vxlan; } @@ -100,9 +100,9 @@ static void mlx5e_vxlan_add_port(struct vxlan->udp_port = port; - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); err = radix_tree_insert(&vxlan_db->tree, vxlan->udp_port, vxlan); - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); if (err) goto err_free; @@ -121,9 +121,9 @@ static void __mlx5e_vxlan_core_del_port( struct mlx5e_vxlan_db *vxlan_db = &priv->vxlan; struct mlx5e_vxlan *vxlan; - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); vxlan = radix_tree_delete(&vxlan_db->tree, port); - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); if (!vxlan) return; @@ -171,12 +171,12 @@ void mlx5e_vxlan_cleanup(struct mlx5e_pr struct mlx5e_vxlan *vxlan; unsigned int port = 0; - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); while (radix_tree_gang_lookup(&vxlan_db->tree, (void **)&vxlan, port, 1)) { port = vxlan->udp_port; - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); __mlx5e_vxlan_core_del_port(priv, (u16)port); - spin_lock_irq(&vxlan_db->lock); + spin_lock_bh(&vxlan_db->lock); } - spin_unlock_irq(&vxlan_db->lock); + spin_unlock_bh(&vxlan_db->lock); } Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.14/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.14/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.14/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.14/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Fix features check of IPv6 traffic" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Fix features check of IPv6 traffic to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-fix-features-check-of-ipv6-traffic.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Tue, 21 Nov 2017 17:49:36 +0200 Subject: net/mlx5e: Fix features check of IPv6 traffic From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 2989ad1ec03021ee6d2193c35414f1d970a243de ] The assumption that the next header field contains the transport protocol is wrong for IPv6 packets with extension headers. Instead, we should look the inner-most next header field in the buffer. This will fix TSO offload for tunnels over IPv6 with extension headers. Performance testing: 19.25x improvement, cool! Measuring bandwidth of 16 threads TCP traffic over IPv6 GRE tap. CPU: Intel(R) Xeon(R) CPU E5-2660 v2 @ 2.20GHz NIC: Mellanox Technologies MT28800 Family [ConnectX-5 Ex] TSO: Enabled Before: 4,926.24 Mbps Now : 94,827.91 Mbps Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -3554,6 +3554,7 @@ static netdev_features_t mlx5e_tunnel_fe struct sk_buff *skb, netdev_features_t features) { + unsigned int offset = 0; struct udphdr *udph; u8 proto; u16 port; @@ -3563,7 +3564,7 @@ static netdev_features_t mlx5e_tunnel_fe proto = ip_hdr(skb)->protocol; break; case htons(ETH_P_IPV6): - proto = ipv6_hdr(skb)->nexthdr; + proto = ipv6_find_hdr(skb, &offset, -1, NULL, NULL); break; default: goto out; Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.14/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.14/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.14/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.14/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5e: Add refcount to VXLAN structure" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5e: Add refcount to VXLAN structure to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5e-add-refcount-to-vxlan-structure.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Gal Pressman <galp(a)mellanox.com> Date: Sun, 3 Dec 2017 13:58:50 +0200 Subject: net/mlx5e: Add refcount to VXLAN structure From: Gal Pressman <galp(a)mellanox.com> [ Upstream commit 23f4cc2cd9ed92570647220aca60d0197d8c1fa9 ] A refcount mechanism must be implemented in order to prevent unwanted scenarios such as: - Open an IPv4 VXLAN interface - Open an IPv6 VXLAN interface (different socket) - Remove one of the interfaces With current implementation, the UDP port will be removed from our VXLAN database and turn off the offloads for the other interface, which is still active. The reference count mechanism will only allow UDP port removals once all consumers are gone. Fixes: b3f63c3d5e2c ("net/mlx5e: Add netdev support for VXLAN tunneling") Signed-off-by: Gal Pressman <galp(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/vxlan.c | 50 ++++++++++++------------ drivers/net/ethernet/mellanox/mlx5/core/vxlan.h | 1 2 files changed, 28 insertions(+), 23 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.c @@ -88,8 +88,11 @@ static void mlx5e_vxlan_add_port(struct struct mlx5e_vxlan *vxlan; int err; - if (mlx5e_vxlan_lookup_port(priv, port)) + vxlan = mlx5e_vxlan_lookup_port(priv, port); + if (vxlan) { + atomic_inc(&vxlan->refcount); goto free_work; + } if (mlx5e_vxlan_core_add_port_cmd(priv->mdev, port)) goto free_work; @@ -99,6 +102,7 @@ static void mlx5e_vxlan_add_port(struct goto err_delete_port; vxlan->udp_port = port; + atomic_set(&vxlan->refcount, 1); spin_lock_bh(&vxlan_db->lock); err = radix_tree_insert(&vxlan_db->tree, vxlan->udp_port, vxlan); @@ -116,32 +120,33 @@ free_work: kfree(vxlan_work); } -static void __mlx5e_vxlan_core_del_port(struct mlx5e_priv *priv, u16 port) +static void mlx5e_vxlan_del_port(struct work_struct *work) { + struct mlx5e_vxlan_work *vxlan_work = + container_of(work, struct mlx5e_vxlan_work, work); + struct mlx5e_priv *priv = vxlan_work->priv; struct mlx5e_vxlan_db *vxlan_db = &priv->vxlan; + u16 port = vxlan_work->port; struct mlx5e_vxlan *vxlan; + bool remove = false; spin_lock_bh(&vxlan_db->lock); - vxlan = radix_tree_delete(&vxlan_db->tree, port); - spin_unlock_bh(&vxlan_db->lock); - + vxlan = radix_tree_lookup(&vxlan_db->tree, port); if (!vxlan) - return; - - mlx5e_vxlan_core_del_port_cmd(priv->mdev, vxlan->udp_port); - - kfree(vxlan); -} + goto out_unlock; -static void mlx5e_vxlan_del_port(struct work_struct *work) -{ - struct mlx5e_vxlan_work *vxlan_work = - container_of(work, struct mlx5e_vxlan_work, work); - struct mlx5e_priv *priv = vxlan_work->priv; - u16 port = vxlan_work->port; + if (atomic_dec_and_test(&vxlan->refcount)) { + radix_tree_delete(&vxlan_db->tree, port); + remove = true; + } - __mlx5e_vxlan_core_del_port(priv, port); +out_unlock: + spin_unlock_bh(&vxlan_db->lock); + if (remove) { + mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); + kfree(vxlan); + } kfree(vxlan_work); } @@ -171,12 +176,11 @@ void mlx5e_vxlan_cleanup(struct mlx5e_pr struct mlx5e_vxlan *vxlan; unsigned int port = 0; - spin_lock_bh(&vxlan_db->lock); + /* Lockless since we are the only radix-tree consumers, wq is disabled */ while (radix_tree_gang_lookup(&vxlan_db->tree, (void **)&vxlan, port, 1)) { port = vxlan->udp_port; - spin_unlock_bh(&vxlan_db->lock); - __mlx5e_vxlan_core_del_port(priv, (u16)port); - spin_lock_bh(&vxlan_db->lock); + radix_tree_delete(&vxlan_db->tree, port); + mlx5e_vxlan_core_del_port_cmd(priv->mdev, port); + kfree(vxlan); } - spin_unlock_bh(&vxlan_db->lock); } --- a/drivers/net/ethernet/mellanox/mlx5/core/vxlan.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/vxlan.h @@ -36,6 +36,7 @@ #include "en.h" struct mlx5e_vxlan { + atomic_t refcount; u16 udp_port; }; Patches currently in stable-queue which might be from galp(a)mellanox.com are queue-4.14/net-mlx5e-fix-possible-deadlock-of-vxlan-lock.patch queue-4.14/net-mlx5e-prevent-possible-races-in-vxlan-control-flow.patch queue-4.14/net-mlx5e-add-refcount-to-vxlan-structure.patch queue-4.14/net-mlx5e-fix-features-check-of-ipv6-traffic.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5: FPGA, return -EINVAL if size is zero" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5: FPGA, return -EINVAL if size is zero to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5-fpga-return-einval-if-size-is-zero.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Kamal Heib <kamalh(a)mellanox.com> Date: Sun, 29 Oct 2017 04:03:37 +0200 Subject: net/mlx5: FPGA, return -EINVAL if size is zero From: Kamal Heib <kamalh(a)mellanox.com> [ Upstream commit bae115a2bb479142605726e6aa130f43f50e801a ] Currently, if a size of zero is passed to mlx5_fpga_mem_{read|write}_i2c() the "err" return value will not be initialized, which triggers gcc warnings: [..]/mlx5/core/fpga/sdk.c:87 mlx5_fpga_mem_read_i2c() error: uninitialized symbol 'err'. [..]/mlx5/core/fpga/sdk.c:115 mlx5_fpga_mem_write_i2c() error: uninitialized symbol 'err'. fix that. Fixes: a9956d35d199 ('net/mlx5: FPGA, Add SBU infrastructure') Signed-off-by: Kamal Heib <kamalh(a)mellanox.com> Reviewed-by: Yevgeny Kliteynik <kliteyn(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/fpga/sdk.c | 6 ++++++ 1 file changed, 6 insertions(+) --- a/drivers/net/ethernet/mellanox/mlx5/core/fpga/sdk.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fpga/sdk.c @@ -66,6 +66,9 @@ static int mlx5_fpga_mem_read_i2c(struct u8 actual_size; int err; + if (!size) + return -EINVAL; + if (!fdev->mdev) return -ENOTCONN; @@ -95,6 +98,9 @@ static int mlx5_fpga_mem_write_i2c(struc u8 actual_size; int err; + if (!size) + return -EINVAL; + if (!fdev->mdev) return -ENOTCONN; Patches currently in stable-queue which might be from kamalh(a)mellanox.com are queue-4.14/net-mlx5-fpga-return-einval-if-size-is-zero.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5: Fix rate limit packet pacing naming and struct" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5: Fix rate limit packet pacing naming and struct to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5-fix-rate-limit-packet-pacing-naming-and-struct.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Eran Ben Elisha <eranbe(a)mellanox.com> Date: Mon, 13 Nov 2017 10:11:27 +0200 Subject: net/mlx5: Fix rate limit packet pacing naming and struct From: Eran Ben Elisha <eranbe(a)mellanox.com> [ Upstream commit 37e92a9d4fe38dc3e7308913575983a6a088c8d4 ] In mlx5_ifc, struct size was not complete, and thus driver was sending garbage after the last defined field. Fixed it by adding reserved field to complete the struct size. In addition, rename all set_rate_limit to set_pp_rate_limit to be compliant with the Firmware <-> Driver definition. Fixes: 7486216b3a0b ("{net,IB}/mlx5: mlx5_ifc updates") Fixes: 1466cc5b23d1 ("net/mlx5: Rate limit tables support") Signed-off-by: Eran Ben Elisha <eranbe(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 4 ++-- drivers/net/ethernet/mellanox/mlx5/core/rl.c | 22 +++++++++++----------- include/linux/mlx5/mlx5_ifc.h | 8 +++++--- 3 files changed, 18 insertions(+), 16 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c @@ -362,7 +362,7 @@ static int mlx5_internal_err_ret_value(s case MLX5_CMD_OP_QUERY_VPORT_COUNTER: case MLX5_CMD_OP_ALLOC_Q_COUNTER: case MLX5_CMD_OP_QUERY_Q_COUNTER: - case MLX5_CMD_OP_SET_RATE_LIMIT: + case MLX5_CMD_OP_SET_PP_RATE_LIMIT: case MLX5_CMD_OP_QUERY_RATE_LIMIT: case MLX5_CMD_OP_CREATE_SCHEDULING_ELEMENT: case MLX5_CMD_OP_QUERY_SCHEDULING_ELEMENT: @@ -505,7 +505,7 @@ const char *mlx5_command_str(int command MLX5_COMMAND_STR_CASE(ALLOC_Q_COUNTER); MLX5_COMMAND_STR_CASE(DEALLOC_Q_COUNTER); MLX5_COMMAND_STR_CASE(QUERY_Q_COUNTER); - MLX5_COMMAND_STR_CASE(SET_RATE_LIMIT); + MLX5_COMMAND_STR_CASE(SET_PP_RATE_LIMIT); MLX5_COMMAND_STR_CASE(QUERY_RATE_LIMIT); MLX5_COMMAND_STR_CASE(CREATE_SCHEDULING_ELEMENT); MLX5_COMMAND_STR_CASE(DESTROY_SCHEDULING_ELEMENT); --- a/drivers/net/ethernet/mellanox/mlx5/core/rl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/rl.c @@ -125,16 +125,16 @@ static struct mlx5_rl_entry *find_rl_ent return ret_entry; } -static int mlx5_set_rate_limit_cmd(struct mlx5_core_dev *dev, +static int mlx5_set_pp_rate_limit_cmd(struct mlx5_core_dev *dev, u32 rate, u16 index) { - u32 in[MLX5_ST_SZ_DW(set_rate_limit_in)] = {0}; - u32 out[MLX5_ST_SZ_DW(set_rate_limit_out)] = {0}; + u32 in[MLX5_ST_SZ_DW(set_pp_rate_limit_in)] = {0}; + u32 out[MLX5_ST_SZ_DW(set_pp_rate_limit_out)] = {0}; - MLX5_SET(set_rate_limit_in, in, opcode, - MLX5_CMD_OP_SET_RATE_LIMIT); - MLX5_SET(set_rate_limit_in, in, rate_limit_index, index); - MLX5_SET(set_rate_limit_in, in, rate_limit, rate); + MLX5_SET(set_pp_rate_limit_in, in, opcode, + MLX5_CMD_OP_SET_PP_RATE_LIMIT); + MLX5_SET(set_pp_rate_limit_in, in, rate_limit_index, index); + MLX5_SET(set_pp_rate_limit_in, in, rate_limit, rate); return mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); } @@ -173,7 +173,7 @@ int mlx5_rl_add_rate(struct mlx5_core_de entry->refcount++; } else { /* new rate limit */ - err = mlx5_set_rate_limit_cmd(dev, rate, entry->index); + err = mlx5_set_pp_rate_limit_cmd(dev, rate, entry->index); if (err) { mlx5_core_err(dev, "Failed configuring rate: %u (%d)\n", rate, err); @@ -209,7 +209,7 @@ void mlx5_rl_remove_rate(struct mlx5_cor entry->refcount--; if (!entry->refcount) { /* need to remove rate */ - mlx5_set_rate_limit_cmd(dev, 0, entry->index); + mlx5_set_pp_rate_limit_cmd(dev, 0, entry->index); entry->rate = 0; } @@ -262,8 +262,8 @@ void mlx5_cleanup_rl_table(struct mlx5_c /* Clear all configured rates */ for (i = 0; i < table->max_size; i++) if (table->rl_entry[i].rate) - mlx5_set_rate_limit_cmd(dev, 0, - table->rl_entry[i].index); + mlx5_set_pp_rate_limit_cmd(dev, 0, + table->rl_entry[i].index); kfree(dev->priv.rl_table.rl_entry); } --- a/include/linux/mlx5/mlx5_ifc.h +++ b/include/linux/mlx5/mlx5_ifc.h @@ -147,7 +147,7 @@ enum { MLX5_CMD_OP_ALLOC_Q_COUNTER = 0x771, MLX5_CMD_OP_DEALLOC_Q_COUNTER = 0x772, MLX5_CMD_OP_QUERY_Q_COUNTER = 0x773, - MLX5_CMD_OP_SET_RATE_LIMIT = 0x780, + MLX5_CMD_OP_SET_PP_RATE_LIMIT = 0x780, MLX5_CMD_OP_QUERY_RATE_LIMIT = 0x781, MLX5_CMD_OP_CREATE_SCHEDULING_ELEMENT = 0x782, MLX5_CMD_OP_DESTROY_SCHEDULING_ELEMENT = 0x783, @@ -7233,7 +7233,7 @@ struct mlx5_ifc_add_vxlan_udp_dport_in_b u8 vxlan_udp_port[0x10]; }; -struct mlx5_ifc_set_rate_limit_out_bits { +struct mlx5_ifc_set_pp_rate_limit_out_bits { u8 status[0x8]; u8 reserved_at_8[0x18]; @@ -7242,7 +7242,7 @@ struct mlx5_ifc_set_rate_limit_out_bits u8 reserved_at_40[0x40]; }; -struct mlx5_ifc_set_rate_limit_in_bits { +struct mlx5_ifc_set_pp_rate_limit_in_bits { u8 opcode[0x10]; u8 reserved_at_10[0x10]; @@ -7255,6 +7255,8 @@ struct mlx5_ifc_set_rate_limit_in_bits { u8 reserved_at_60[0x20]; u8 rate_limit[0x20]; + + u8 reserved_at_a0[0x160]; }; struct mlx5_ifc_access_register_out_bits { Patches currently in stable-queue which might be from eranbe(a)mellanox.com are queue-4.14/net-mlx5-fix-rate-limit-packet-pacing-naming-and-struct.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net/mlx5: Fix error flow in CREATE_QP command" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net/mlx5: Fix error flow in CREATE_QP command to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-mlx5-fix-error-flow-in-create_qp-command.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Moni Shoua <monis(a)mellanox.com> Date: Mon, 4 Dec 2017 08:59:25 +0200 Subject: net/mlx5: Fix error flow in CREATE_QP command From: Moni Shoua <monis(a)mellanox.com> [ Upstream commit dbff26e44dc3ec4de6578733b054a0114652a764 ] In error flow, when DESTROY_QP command should be executed, the wrong mailbox was set with data, not the one that is written to hardware, Fix that. Fixes: 09a7d9eca1a6 '{net,IB}/mlx5: QP/XRCD commands via mlx5 ifc' Signed-off-by: Moni Shoua <monis(a)mellanox.com> Signed-off-by: Saeed Mahameed <saeedm(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlx5/core/qp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/drivers/net/ethernet/mellanox/mlx5/core/qp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/qp.c @@ -213,8 +213,8 @@ int mlx5_core_create_qp(struct mlx5_core err_cmd: memset(din, 0, sizeof(din)); memset(dout, 0, sizeof(dout)); - MLX5_SET(destroy_qp_in, in, opcode, MLX5_CMD_OP_DESTROY_QP); - MLX5_SET(destroy_qp_in, in, qpn, qp->qpn); + MLX5_SET(destroy_qp_in, din, opcode, MLX5_CMD_OP_DESTROY_QP); + MLX5_SET(destroy_qp_in, din, qpn, qp->qpn); mlx5_cmd_exec(dev, din, sizeof(din), dout, sizeof(dout)); return err; } Patches currently in stable-queue which might be from monis(a)mellanox.com are queue-4.14/ib-core-verify-that-qp-is-security-enabled-in-create-and-destroy.patch queue-4.14/ib-uverbs-fix-command-checking-as-part-of-ib_uverbs_ex_modify_qp.patch queue-4.14/net-mlx5-fix-error-flow-in-create_qp-command.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: ipv4: fix for a race condition in raw_sendmsg" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: ipv4: fix for a race condition in raw_sendmsg to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Mohamed Ghannam <simo.ghannam(a)gmail.com> Date: Sun, 10 Dec 2017 03:50:58 +0000 Subject: net: ipv4: fix for a race condition in raw_sendmsg From: Mohamed Ghannam <simo.ghannam(a)gmail.com> [ Upstream commit 8f659a03a0ba9289b9aeb9b4470e6fb263d6f483 ] inet->hdrincl is racy, and could lead to uninitialized stack pointer usage, so its value should be read only once. Fixes: c008ba5bdc9f ("ipv4: Avoid reading user iov twice after raw_probe_proto_opt") Signed-off-by: Mohamed Ghannam <simo.ghannam(a)gmail.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/raw.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -513,11 +513,16 @@ static int raw_sendmsg(struct sock *sk, int err; struct ip_options_data opt_copy; struct raw_frag_vec rfv; + int hdrincl; err = -EMSGSIZE; if (len > 0xFFFF) goto out; + /* hdrincl should be READ_ONCE(inet->hdrincl) + * but READ_ONCE() doesn't work with bit fields + */ + hdrincl = inet->hdrincl; /* * Check the flags. */ @@ -593,7 +598,7 @@ static int raw_sendmsg(struct sock *sk, /* Linux does not mangle headers on raw sockets, * so that IP options + IP_HDRINCL is non-sense. */ - if (inet->hdrincl) + if (hdrincl) goto done; if (ipc.opt->opt.srr) { if (!daddr) @@ -615,12 +620,12 @@ static int raw_sendmsg(struct sock *sk, flowi4_init_output(&fl4, ipc.oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE, - inet->hdrincl ? IPPROTO_RAW : sk->sk_protocol, + hdrincl ? IPPROTO_RAW : sk->sk_protocol, inet_sk_flowi_flags(sk) | - (inet->hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), + (hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), daddr, saddr, 0, 0, sk->sk_uid); - if (!inet->hdrincl) { + if (!hdrincl) { rfv.msg = msg; rfv.hlen = 0; @@ -645,7 +650,7 @@ static int raw_sendmsg(struct sock *sk, goto do_confirm; back_from_confirm: - if (inet->hdrincl) + if (hdrincl) err = raw_send_hdrinc(sk, &fl4, msg, len, &rt, msg->msg_flags, &ipc.sockc); Patches currently in stable-queue which might be from simo.ghannam(a)gmail.com are queue-4.14/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: igmp: Use correct source address on IGMPv3 reports" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: igmp: Use correct source address on IGMPv3 reports to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-igmp-use-correct-source-address-on-igmpv3-reports.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Kevin Cernekee <cernekee(a)chromium.org> Date: Mon, 11 Dec 2017 11:13:45 -0800 Subject: net: igmp: Use correct source address on IGMPv3 reports From: Kevin Cernekee <cernekee(a)chromium.org> [ Upstream commit a46182b00290839fa3fa159d54fd3237bd8669f0 ] Closing a multicast socket after the final IPv4 address is deleted from an interface can generate a membership report that uses the source IP from a different interface. The following test script, run from an isolated netns, reproduces the issue: #!/bin/bash ip link add dummy0 type dummy ip link add dummy1 type dummy ip link set dummy0 up ip link set dummy1 up ip addr add 10.1.1.1/24 dev dummy0 ip addr add 192.168.99.99/24 dev dummy1 tcpdump -U -i dummy0 & socat EXEC:"sleep 2" \ UDP4-DATAGRAM:239.101.1.68:8889,ip-add-membership=239.0.1.68:10.1.1.1 & sleep 1 ip addr del 10.1.1.1/24 dev dummy0 sleep 5 kill %tcpdump RFC 3376 specifies that the report must be sent with a valid IP source address from the destination subnet, or from address 0.0.0.0. Add an extra check to make sure this is the case. Signed-off-by: Kevin Cernekee <cernekee(a)chromium.org> Reviewed-by: Andrew Lunn <andrew(a)lunn.ch> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/igmp.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -89,6 +89,7 @@ #include <linux/rtnetlink.h> #include <linux/times.h> #include <linux/pkt_sched.h> +#include <linux/byteorder/generic.h> #include <net/net_namespace.h> #include <net/arp.h> @@ -321,6 +322,23 @@ igmp_scount(struct ip_mc_list *pmc, int return scount; } +/* source address selection per RFC 3376 section 4.2.13 */ +static __be32 igmpv3_get_srcaddr(struct net_device *dev, + const struct flowi4 *fl4) +{ + struct in_device *in_dev = __in_dev_get_rcu(dev); + + if (!in_dev) + return htonl(INADDR_ANY); + + for_ifa(in_dev) { + if (inet_ifa_match(fl4->saddr, ifa)) + return fl4->saddr; + } endfor_ifa(in_dev); + + return htonl(INADDR_ANY); +} + static struct sk_buff *igmpv3_newpack(struct net_device *dev, unsigned int mtu) { struct sk_buff *skb; @@ -368,7 +386,7 @@ static struct sk_buff *igmpv3_newpack(st pip->frag_off = htons(IP_DF); pip->ttl = 1; pip->daddr = fl4.daddr; - pip->saddr = fl4.saddr; + pip->saddr = igmpv3_get_srcaddr(dev, &fl4); pip->protocol = IPPROTO_IGMP; pip->tot_len = 0; /* filled in later */ ip_select_ident(net, skb, NULL); Patches currently in stable-queue which might be from cernekee(a)chromium.org are queue-4.14/net-igmp-use-correct-source-address-on-igmpv3-reports.patch queue-4.14/netlink-add-netns-check-on-taps.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] [PATCHES] Networking

by David Miller

Thought you'd make it into 2018 without some more networking bug fixes? Think again! :-) Please queue up the following networking bug fixes for 4.9.x and 4.14.x -stable, respectively. Thanks!

7 years, 8 months

2
1
0 0

[Linux-stable-mirror] Patch "net: Fix double free and memory corruption in get_net_ns_by_id()" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: Fix double free and memory corruption in get_net_ns_by_id() to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: "Eric W. Biederman" <ebiederm(a)xmission.com> Date: Tue, 19 Dec 2017 11:27:56 -0600 Subject: net: Fix double free and memory corruption in get_net_ns_by_id() From: "Eric W. Biederman" <ebiederm(a)xmission.com> [ Upstream commit 21b5944350052d2583e82dd59b19a9ba94a007f0 ] (I can trivially verify that that idr_remove in cleanup_net happens after the network namespace count has dropped to zero --EWB) Function get_net_ns_by_id() does not check for net::count after it has found a peer in netns_ids idr. It may dereference a peer, after its count has already been finaly decremented. This leads to double free and memory corruption: put_net(peer) rtnl_lock() atomic_dec_and_test(&peer->count) [count=0] ... __put_net(peer) get_net_ns_by_id(net, id) spin_lock(&cleanup_list_lock) list_add(&net->cleanup_list, &cleanup_list) spin_unlock(&cleanup_list_lock) queue_work() peer = idr_find(&net->netns_ids, id) | get_net(peer) [count=1] | ... | (use after final put) v ... cleanup_net() ... spin_lock(&cleanup_list_lock) ... list_replace_init(&cleanup_list, ..) ... spin_unlock(&cleanup_list_lock) ... ... ... ... put_net(peer) ... atomic_dec_and_test(&peer->count) [count=0] ... spin_lock(&cleanup_list_lock) ... list_add(&net->cleanup_list, &cleanup_list) ... spin_unlock(&cleanup_list_lock) ... queue_work() ... rtnl_unlock() rtnl_lock() ... for_each_net(tmp) { ... id = __peernet2id(tmp, peer) ... spin_lock_irq(&tmp->nsid_lock) ... idr_remove(&tmp->netns_ids, id) ... ... ... net_drop_ns() ... net_free(peer) ... } ... | v cleanup_net() ... (Second free of peer) Also, put_net() on the right cpu may reorder with left's cpu list_replace_init(&cleanup_list, ..), and then cleanup_list will be corrupted. Since cleanup_net() is executed in worker thread, while put_net(peer) can happen everywhere, there should be enough time for concurrent get_net_ns_by_id() to pick the peer up, and the race does not seem to be unlikely. The patch fixes the problem in standard way. (Also, there is possible problem in peernet2id_alloc(), which requires check for net::count under nsid_lock and maybe_get_net(peer), but in current stable kernel it's used under rtnl_lock() and it has to be safe. Openswitch begun to use peernet2id_alloc(), and possibly it should be fixed too. While this is not in stable kernel yet, so I'll send a separate message to netdev@ later). Cc: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Signed-off-by: Kirill Tkhai <ktkhai(a)virtuozzo.com> Fixes: 0c7aecd4bde4 "netns: add rtnl cmd to add and get peer netns ids" Reviewed-by: Andrey Ryabinin <aryabinin(a)virtuozzo.com> Reviewed-by: "Eric W. Biederman" <ebiederm(a)xmission.com> Signed-off-by: Eric W. Biederman <ebiederm(a)xmission.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Acked-by: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/core/net_namespace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -266,7 +266,7 @@ struct net *get_net_ns_by_id(struct net spin_lock_bh(&net->nsid_lock); peer = idr_find(&net->netns_ids, id); if (peer) - get_net(peer); + peer = maybe_get_net(peer); spin_unlock_bh(&net->nsid_lock); rcu_read_unlock(); Patches currently in stable-queue which might be from ebiederm(a)xmission.com are queue-4.14/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: fec: unmap the xmit buffer that are not transferred by DMA" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: fec: unmap the xmit buffer that are not transferred by DMA to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-fec-unmap-the-xmit-buffer-that-are-not-transferred-by-dma.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Fugang Duan <fugang.duan(a)nxp.com> Date: Fri, 22 Dec 2017 17:12:09 +0800 Subject: net: fec: unmap the xmit buffer that are not transferred by DMA From: Fugang Duan <fugang.duan(a)nxp.com> [ Upstream commit 178e5f57a8d8f8fc5799a624b96fc31ef9a29ffa ] The enet IP only support 32 bit, it will use swiotlb buffer to do dma mapping when xmit buffer DMA memory address is bigger than 4G in i.MX platform. After stress suspend/resume test, it will print out: log: [12826.352864] fec 5b040000.ethernet: swiotlb buffer is full (sz: 191 bytes) [12826.359676] DMA: Out of SW-IOMMU space for 191 bytes at device 5b040000.ethernet [12826.367110] fec 5b040000.ethernet eth0: Tx DMA memory map failed The issue is that the ready xmit buffers that are dma mapped but DMA still don't copy them into fifo, once MAC restart, these DMA buffers are not unmapped. So it should check the dma mapping buffer and unmap them. Signed-off-by: Fugang Duan <fugang.duan(a)nxp.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/freescale/fec_main.c | 6 ++++++ 1 file changed, 6 insertions(+) --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -818,6 +818,12 @@ static void fec_enet_bd_init(struct net_ for (i = 0; i < txq->bd.ring_size; i++) { /* Initialize the BD for every fragment in the page. */ bdp->cbd_sc = cpu_to_fec16(0); + if (bdp->cbd_bufaddr && + !IS_TSO_HEADER(txq, fec32_to_cpu(bdp->cbd_bufaddr))) + dma_unmap_single(&fep->pdev->dev, + fec32_to_cpu(bdp->cbd_bufaddr), + fec16_to_cpu(bdp->cbd_datlen), + DMA_TO_DEVICE); if (txq->tx_skbuff[i]) { dev_kfree_skb_any(txq->tx_skbuff[i]); txq->tx_skbuff[i] = NULL; Patches currently in stable-queue which might be from fugang.duan(a)nxp.com are queue-4.14/net-fec-unmap-the-xmit-buffer-that-are-not-transferred-by-dma.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: dsa: bcm_sf2: Clear IDDQ_GLOBAL_PWR bit for PHY" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: dsa: bcm_sf2: Clear IDDQ_GLOBAL_PWR bit for PHY to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-dsa-bcm_sf2-clear-iddq_global_pwr-bit-for-phy.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Florian Fainelli <f.fainelli(a)gmail.com> Date: Tue, 21 Nov 2017 17:37:46 -0800 Subject: net: dsa: bcm_sf2: Clear IDDQ_GLOBAL_PWR bit for PHY From: Florian Fainelli <f.fainelli(a)gmail.com> [ Upstream commit 4b52d010113e11006a389f2a8315167ede9e0b10 ] The PHY on BCM7278 has an additional bit that needs to be cleared: IDDQ_GLOBAL_PWR, without doing this, the PHY remains stuck in reset out of suspend/resume cycles. Fixes: 0fe9933804eb ("net: dsa: bcm_sf2: Add support for BCM7278 integrated switch") Signed-off-by: Florian Fainelli <f.fainelli(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/dsa/bcm_sf2.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/net/dsa/bcm_sf2.c +++ b/drivers/net/dsa/bcm_sf2.c @@ -167,7 +167,7 @@ static void bcm_sf2_gphy_enable_set(stru reg = reg_readl(priv, REG_SPHY_CNTRL); if (enable) { reg |= PHY_RESET; - reg &= ~(EXT_PWR_DOWN | IDDQ_BIAS | CK25_DIS); + reg &= ~(EXT_PWR_DOWN | IDDQ_BIAS | IDDQ_GLOBAL_PWR | CK25_DIS); reg_writel(priv, reg, REG_SPHY_CNTRL); udelay(21); reg = reg_readl(priv, REG_SPHY_CNTRL); Patches currently in stable-queue which might be from f.fainelli(a)gmail.com are queue-4.14/phylink-ensure-the-phy-interface-mode-is-appropriately-set.patch queue-4.14/net-dsa-bcm_sf2-clear-iddq_global_pwr-bit-for-phy.patch queue-4.14/phylink-ensure-an-is-enabled.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: net-bridge-fix-early-call-to-br_stp_change_bridge_id-and-plug-newlink-leaks.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> Date: Mon, 18 Dec 2017 17:35:09 +0200 Subject: net: bridge: fix early call to br_stp_change_bridge_id and plug newlink leaks From: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> [ Upstream commit 84aeb437ab98a2bce3d4b2111c79723aedfceb33 ] The early call to br_stp_change_bridge_id in bridge's newlink can cause a memory leak if an error occurs during the newlink because the fdb entries are not cleaned up if a different lladdr was specified, also another minor issue is that it generates fdb notifications with ifindex = 0. Another unrelated memory leak is the bridge sysfs entries which get added on NETDEV_REGISTER event, but are not cleaned up in the newlink error path. To remove this special case the call to br_stp_change_bridge_id is done after netdev register and we cleanup the bridge on changelink error via br_dev_delete to plug all leaks. This patch makes netlink bridge destruction on newlink error the same as dellink and ioctl del which is necessary since at that point we have a fully initialized bridge device. To reproduce the issue: $ ip l add br0 address 00:11:22:33:44:55 type bridge group_fwd_mask 1 RTNETLINK answers: Invalid argument $ rmmod bridge [ 1822.142525] ============================================================================= [ 1822.143640] BUG bridge_fdb_cache (Tainted: G O ): Objects remaining in bridge_fdb_cache on __kmem_cache_shutdown() [ 1822.144821] ----------------------------------------------------------------------------- [ 1822.145990] Disabling lock debugging due to kernel taint [ 1822.146732] INFO: Slab 0x0000000092a844b2 objects=32 used=2 fp=0x00000000fef011b0 flags=0x1ffff8000000100 [ 1822.147700] CPU: 2 PID: 13584 Comm: rmmod Tainted: G B O 4.15.0-rc2+ #87 [ 1822.148578] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.7.5-20140531_083030-gandalf 04/01/2014 [ 1822.150008] Call Trace: [ 1822.150510] dump_stack+0x78/0xa9 [ 1822.151156] slab_err+0xb1/0xd3 [ 1822.151834] ? __kmalloc+0x1bb/0x1ce [ 1822.152546] __kmem_cache_shutdown+0x151/0x28b [ 1822.153395] shutdown_cache+0x13/0x144 [ 1822.154126] kmem_cache_destroy+0x1c0/0x1fb [ 1822.154669] SyS_delete_module+0x194/0x244 [ 1822.155199] ? trace_hardirqs_on_thunk+0x1a/0x1c [ 1822.155773] entry_SYSCALL_64_fastpath+0x23/0x9a [ 1822.156343] RIP: 0033:0x7f929bd38b17 [ 1822.156859] RSP: 002b:00007ffd160e9a98 EFLAGS: 00000202 ORIG_RAX: 00000000000000b0 [ 1822.157728] RAX: ffffffffffffffda RBX: 00005578316ba090 RCX: 00007f929bd38b17 [ 1822.158422] RDX: 00007f929bd9ec60 RSI: 0000000000000800 RDI: 00005578316ba0f0 [ 1822.159114] RBP: 0000000000000003 R08: 00007f929bff5f20 R09: 00007ffd160e8a11 [ 1822.159808] R10: 00007ffd160e9860 R11: 0000000000000202 R12: 00007ffd160e8a80 [ 1822.160513] R13: 0000000000000000 R14: 0000000000000000 R15: 00005578316ba090 [ 1822.161278] INFO: Object 0x000000007645de29 @offset=0 [ 1822.161666] INFO: Object 0x00000000d5df2ab5 @offset=128 Fixes: 30313a3d5794 ("bridge: Handle IFLA_ADDRESS correctly when creating bridge device") Fixes: 5b8d5429daa0 ("bridge: netlink: register netdevice before executing changelink") Signed-off-by: Nikolay Aleksandrov <nikolay(a)cumulusnetworks.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/bridge/br_netlink.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) --- a/net/bridge/br_netlink.c +++ b/net/bridge/br_netlink.c @@ -1223,19 +1223,20 @@ static int br_dev_newlink(struct net *sr struct net_bridge *br = netdev_priv(dev); int err; + err = register_netdevice(dev); + if (err) + return err; + if (tb[IFLA_ADDRESS]) { spin_lock_bh(&br->lock); br_stp_change_bridge_id(br, nla_data(tb[IFLA_ADDRESS])); spin_unlock_bh(&br->lock); } - err = register_netdevice(dev); - if (err) - return err; - err = br_changelink(dev, tb, data, extack); if (err) - unregister_netdevice(dev); + br_dev_delete(dev, NULL); + return err; } Patches currently in stable-queue which might be from nikolay(a)cumulusnetworks.com are queue-4.14/net-bridge-fix-early-call-to-br_stp_change_bridge_id-and-plug-newlink-leaks.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "mlxsw: spectrum: Disable MAC learning for ovs port" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled mlxsw: spectrum: Disable MAC learning for ovs port to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: mlxsw-spectrum-disable-mac-learning-for-ovs-port.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Yuval Mintz <yuvalm(a)mellanox.com> Date: Fri, 15 Dec 2017 08:44:21 +0100 Subject: mlxsw: spectrum: Disable MAC learning for ovs port From: Yuval Mintz <yuvalm(a)mellanox.com> [ Upstream commit fccff0862838908d21eaf956d57e09c6c189f7c5 ] Learning is currently enabled for ports which are OVS slaves - even though OVS doesn't need this indication. Since we're not associating a fid with the port, HW would continuously notify driver of learned [& aged] MACs which would be logged as errors. Fixes: 2b94e58df58c ("mlxsw: spectrum: Allow ports to work under OVS master") Signed-off-by: Yuval Mintz <yuvalm(a)mellanox.com> Reviewed-by: Ido Schimmel <idosch(a)mellanox.com> Signed-off-by: Jiri Pirko <jiri(a)mellanox.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -4164,6 +4164,7 @@ static int mlxsw_sp_port_stp_set(struct static int mlxsw_sp_port_ovs_join(struct mlxsw_sp_port *mlxsw_sp_port) { + u16 vid = 1; int err; err = mlxsw_sp_port_vp_mode_set(mlxsw_sp_port, true); @@ -4176,8 +4177,19 @@ static int mlxsw_sp_port_ovs_join(struct true, false); if (err) goto err_port_vlan_set; + + for (; vid <= VLAN_N_VID - 1; vid++) { + err = mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, + vid, false); + if (err) + goto err_vid_learning_set; + } + return 0; +err_vid_learning_set: + for (vid--; vid >= 1; vid--) + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, vid, true); err_port_vlan_set: mlxsw_sp_port_stp_set(mlxsw_sp_port, false); err_port_stp_set: @@ -4187,6 +4199,12 @@ err_port_stp_set: static void mlxsw_sp_port_ovs_leave(struct mlxsw_sp_port *mlxsw_sp_port) { + u16 vid; + + for (vid = VLAN_N_VID - 1; vid >= 1; vid--) + mlxsw_sp_port_vid_learning_set(mlxsw_sp_port, + vid, true); + mlxsw_sp_port_vlan_set(mlxsw_sp_port, 2, VLAN_N_VID - 1, false, false); mlxsw_sp_port_stp_set(mlxsw_sp_port, false); Patches currently in stable-queue which might be from yuvalm(a)mellanox.com are queue-4.14/mlxsw-spectrum-disable-mac-learning-for-ovs-port.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv6: set all.accept_dad to 0 by default" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv6: set all.accept_dad to 0 by default to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv6-set-all.accept_dad-to-0-by-default.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Date: Tue, 14 Nov 2017 14:21:32 +0100 Subject: ipv6: set all.accept_dad to 0 by default From: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> [ Upstream commit 094009531612246d9e13f9e0c3ae2205d7f63a0a ] With commits 35e015e1f577 and a2d3f3e33853, the global 'accept_dad' flag is also taken into account (default value is 1). If either global or per-interface flag is non-zero, DAD will be enabled on a given interface. This is not backward compatible: before those patches, the user could disable DAD just by setting the per-interface flag to 0. Now, the user instead needs to set both flags to 0 to actually disable DAD. Restore the previous behaviour by setting the default for the global 'accept_dad' flag to 0. This way, DAD is still enabled by default, as per-interface flags are set to 1 on device creation, but setting them to 0 is enough to disable DAD on a given interface. - Before 35e015e1f57a7 and a2d3f3e33853: global per-interface DAD enabled [default] 1 1 yes X 0 no X 1 yes - After 35e015e1f577 and a2d3f3e33853: global per-interface DAD enabled [default] 1 1 yes 0 0 no 0 1 yes 1 0 yes - After this fix: global per-interface DAD enabled 1 1 yes 0 0 no [default] 0 1 yes 1 0 yes Fixes: 35e015e1f577 ("ipv6: fix net.ipv6.conf.all interface DAD handlers") Fixes: a2d3f3e33853 ("ipv6: fix net.ipv6.conf.all.accept_dad behaviour for real") CC: Stefano Brivio <sbrivio(a)redhat.com> CC: Matteo Croce <mcroce(a)redhat.com> CC: Erik Kline <ek(a)google.com> Signed-off-by: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Acked-by: Stefano Brivio <sbrivio(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/addrconf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -231,7 +231,7 @@ static struct ipv6_devconf ipv6_devconf .proxy_ndp = 0, .accept_source_route = 0, /* we do not accept RH0 by default. */ .disable_ipv6 = 0, - .accept_dad = 1, + .accept_dad = 0, .suppress_frag_ndisc = 1, .accept_ra_mtu = 1, .stable_secret = { Patches currently in stable-queue which might be from nicolas.dichtel(a)6wind.com are queue-4.14/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.14/ipv6-set-all.accept_dad-to-0-by-default.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv6: mcast: better catch silly mtu values" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv6: mcast: better catch silly mtu values to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv6-mcast-better-catch-silly-mtu-values.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:03:38 -0800 Subject: ipv6: mcast: better catch silly mtu values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b9b312a7a451e9c098921856e7cfbc201120e1a7 ] syzkaller reported crashes in IPv6 stack [1] Xin Long found that lo MTU was set to silly values. IPv6 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in mld code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv6 minimal MTU. [1] skbuff: skb_over_panic: text:0000000010b86b8d len:196 put:20 head:000000003b477e60 data:000000000e85441e tail:0xd4 end:0xc0 dev:lo ------------[ cut here ]------------ kernel BUG at net/core/skbuff.c:104! invalid opcode: 0000 [#1] SMP KASAN Dumping ftrace buffer: (ftrace buffer empty) Modules linked in: CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.15.0-rc2-mm1+ #39 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 RIP: 0010:skb_panic+0x15c/0x1f0 net/core/skbuff.c:100 RSP: 0018:ffff8801db307508 EFLAGS: 00010286 RAX: 0000000000000082 RBX: ffff8801c517e840 RCX: 0000000000000000 RDX: 0000000000000082 RSI: 1ffff1003b660e61 RDI: ffffed003b660e95 RBP: ffff8801db307570 R08: 1ffff1003b660e23 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffffffff85bd4020 R13: ffffffff84754ed2 R14: 0000000000000014 R15: ffff8801c4e26540 FS: 0000000000000000(0000) GS:ffff8801db300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000463610 CR3: 00000001c6698000 CR4: 00000000001406e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <IRQ> skb_over_panic net/core/skbuff.c:109 [inline] skb_put+0x181/0x1c0 net/core/skbuff.c:1694 add_grhead.isra.24+0x42/0x3b0 net/ipv6/mcast.c:1695 add_grec+0xa55/0x1060 net/ipv6/mcast.c:1817 mld_send_cr net/ipv6/mcast.c:1903 [inline] mld_ifc_timer_expire+0x4d2/0x770 net/ipv6/mcast.c:2448 call_timer_fn+0x23b/0x840 kernel/time/timer.c:1320 expire_timers kernel/time/timer.c:1357 [inline] __run_timers+0x7e1/0xb60 kernel/time/timer.c:1660 run_timer_softirq+0x4c/0xb0 kernel/time/timer.c:1686 __do_softirq+0x29d/0xbb2 kernel/softirq.c:285 invoke_softirq kernel/softirq.c:365 [inline] irq_exit+0x1d3/0x210 kernel/softirq.c:405 exiting_irq arch/x86/include/asm/apic.h:540 [inline] smp_apic_timer_interrupt+0x16b/0x700 arch/x86/kernel/apic/apic.c:1052 apic_timer_interrupt+0xa9/0xb0 arch/x86/entry/entry_64.S:920 Signed-off-by: Eric Dumazet <edumazet(a)google.com> Reported-by: syzbot <syzkaller(a)googlegroups.com> Tested-by: Xin Long <lucien.xin(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/mcast.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) --- a/net/ipv6/mcast.c +++ b/net/ipv6/mcast.c @@ -1682,16 +1682,16 @@ static int grec_size(struct ifmcaddr6 *p } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ifmcaddr6 *pmc, - int type, struct mld2_grec **ppgr) + int type, struct mld2_grec **ppgr, unsigned int mtu) { - struct net_device *dev = pmc->idev->dev; struct mld2_report *pmr; struct mld2_grec *pgr; - if (!skb) - skb = mld_newpack(pmc->idev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = mld_newpack(pmc->idev, mtu); + if (!skb) + return NULL; + } pgr = skb_put(skb, sizeof(struct mld2_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -1714,10 +1714,15 @@ static struct sk_buff *add_grec(struct s struct mld2_grec *pgr = NULL; struct ip6_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->mca_flags & MAF_NOREPORT) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV6_MIN_MTU) + return skb; + isquery = type == MLD2_MODE_IS_INCLUDE || type == MLD2_MODE_IS_EXCLUDE; truncate = type == MLD2_MODE_IS_EXCLUDE || @@ -1738,7 +1743,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); } } first = 1; @@ -1774,12 +1779,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) mld_sendpack(skb); - skb = mld_newpack(idev, dev->mtu); + skb = mld_newpack(idev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -1814,7 +1819,7 @@ empty_source: mld_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.14/tcp-refresh-tcp_mstamp-from-timers-callbacks.patch queue-4.14/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.14/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.14/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.14/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.14/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.14/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.14/tcp-fix-potential-underestimation-on-rcv_rtt.patch queue-4.14/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv6: Honor specified parameters in fibmatch lookup" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv6: Honor specified parameters in fibmatch lookup to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv6-honor-specified-parameters-in-fibmatch-lookup.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Ido Schimmel <idosch(a)mellanox.com> Date: Wed, 20 Dec 2017 12:28:25 +0200 Subject: ipv6: Honor specified parameters in fibmatch lookup From: Ido Schimmel <idosch(a)mellanox.com> [ Upstream commit 58acfd714e6b02e8617448b431c2b64a2f1f0792 ] Currently, parameters such as oif and source address are not taken into account during fibmatch lookup. Example (IPv4 for reference) before patch: $ ip -4 route show 192.0.2.0/24 dev dummy0 proto kernel scope link src 192.0.2.1 198.51.100.0/24 dev dummy1 proto kernel scope link src 198.51.100.1 $ ip -6 route show 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium 2001:db8:2::/64 dev dummy1 proto kernel metric 256 pref medium fe80::/64 dev dummy0 proto kernel metric 256 pref medium fe80::/64 dev dummy1 proto kernel metric 256 pref medium $ ip -4 route get fibmatch 192.0.2.2 oif dummy0 192.0.2.0/24 dev dummy0 proto kernel scope link src 192.0.2.1 $ ip -4 route get fibmatch 192.0.2.2 oif dummy1 RTNETLINK answers: No route to host $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy0 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy1 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium After: $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy0 2001:db8:1::/64 dev dummy0 proto kernel metric 256 pref medium $ ip -6 route get fibmatch 2001:db8:1::2 oif dummy1 RTNETLINK answers: Network is unreachable The problem stems from the fact that the necessary route lookup flags are not set based on these parameters. Instead of duplicating the same logic for fibmatch, we can simply resolve the original route from its copy and dump it instead. Fixes: 18c3a61c4264 ("net: ipv6: RTM_GETROUTE: return matched fib result when requested") Signed-off-by: Ido Schimmel <idosch(a)mellanox.com> Acked-by: David Ahern <dsahern(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/route.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -3700,19 +3700,13 @@ static int inet6_rtm_getroute(struct sk_ if (!ipv6_addr_any(&fl6.saddr)) flags |= RT6_LOOKUP_F_HAS_SADDR; - if (!fibmatch) - dst = ip6_route_input_lookup(net, dev, &fl6, flags); - else - dst = ip6_route_lookup(net, &fl6, 0); + dst = ip6_route_input_lookup(net, dev, &fl6, flags); rcu_read_unlock(); } else { fl6.flowi6_oif = oif; - if (!fibmatch) - dst = ip6_route_output(net, NULL, &fl6); - else - dst = ip6_route_lookup(net, &fl6, 0); + dst = ip6_route_output(net, NULL, &fl6); } @@ -3729,6 +3723,15 @@ static int inet6_rtm_getroute(struct sk_ goto errout; } + if (fibmatch && rt->dst.from) { + struct rt6_info *ort = container_of(rt->dst.from, + struct rt6_info, dst); + + dst_hold(&ort->dst); + ip6_rt_put(rt); + rt = ort; + } + skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL); if (!skb) { ip6_rt_put(rt); Patches currently in stable-queue which might be from idosch(a)mellanox.com are queue-4.14/ipv6-honor-specified-parameters-in-fibmatch-lookup.patch queue-4.14/mlxsw-spectrum-disable-mac-learning-for-ovs-port.patch queue-4.14/ipv4-fix-use-after-free-when-flushing-fib-tables.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: igmp: guard against silly MTU values" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: igmp: guard against silly MTU values to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-igmp-guard-against-silly-mtu-values.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Eric Dumazet <edumazet(a)google.com> Date: Mon, 11 Dec 2017 07:17:39 -0800 Subject: ipv4: igmp: guard against silly MTU values From: Eric Dumazet <edumazet(a)google.com> [ Upstream commit b5476022bbada3764609368f03329ca287528dc8 ] IPv4 stack reacts to changes to small MTU, by disabling itself under RTNL. But there is a window where threads not using RTNL can see a wrong device mtu. This can lead to surprises, in igmp code where it is assumed the mtu is suitable. Fix this by reading device mtu once and checking IPv4 minimal MTU. This patch adds missing IPV4_MIN_MTU define, to not abuse ETH_MIN_MTU anymore. Signed-off-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/net/ip.h | 1 + net/ipv4/devinet.c | 2 +- net/ipv4/igmp.c | 24 +++++++++++++++--------- net/ipv4/ip_tunnel.c | 4 ++-- 4 files changed, 19 insertions(+), 12 deletions(-) --- a/include/net/ip.h +++ b/include/net/ip.h @@ -34,6 +34,7 @@ #include <net/flow_dissector.h> #define IPV4_MAX_PMTU 65535U /* RFC 2675, Section 5.1 */ +#define IPV4_MIN_MTU 68 /* RFC 791 */ struct sock; --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -1420,7 +1420,7 @@ skip: static bool inetdev_valid_mtu(unsigned int mtu) { - return mtu >= 68; + return mtu >= IPV4_MIN_MTU; } static void inetdev_send_gratuitous_arp(struct net_device *dev, --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -404,16 +404,17 @@ static int grec_size(struct ip_mc_list * } static struct sk_buff *add_grhead(struct sk_buff *skb, struct ip_mc_list *pmc, - int type, struct igmpv3_grec **ppgr) + int type, struct igmpv3_grec **ppgr, unsigned int mtu) { struct net_device *dev = pmc->interface->dev; struct igmpv3_report *pih; struct igmpv3_grec *pgr; - if (!skb) - skb = igmpv3_newpack(dev, dev->mtu); - if (!skb) - return NULL; + if (!skb) { + skb = igmpv3_newpack(dev, mtu); + if (!skb) + return NULL; + } pgr = skb_put(skb, sizeof(struct igmpv3_grec)); pgr->grec_type = type; pgr->grec_auxwords = 0; @@ -436,12 +437,17 @@ static struct sk_buff *add_grec(struct s struct igmpv3_grec *pgr = NULL; struct ip_sf_list *psf, *psf_next, *psf_prev, **psf_list; int scount, stotal, first, isquery, truncate; + unsigned int mtu; if (pmc->multiaddr == IGMP_ALL_HOSTS) return skb; if (ipv4_is_local_multicast(pmc->multiaddr) && !net->ipv4.sysctl_igmp_llm_reports) return skb; + mtu = READ_ONCE(dev->mtu); + if (mtu < IPV4_MIN_MTU) + return skb; + isquery = type == IGMPV3_MODE_IS_INCLUDE || type == IGMPV3_MODE_IS_EXCLUDE; truncate = type == IGMPV3_MODE_IS_EXCLUDE || @@ -462,7 +468,7 @@ static struct sk_buff *add_grec(struct s AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) { if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); } } first = 1; @@ -498,12 +504,12 @@ static struct sk_buff *add_grec(struct s pgr->grec_nsrcs = htons(scount); if (skb) igmpv3_sendpack(skb); - skb = igmpv3_newpack(dev, dev->mtu); + skb = igmpv3_newpack(dev, mtu); first = 1; scount = 0; } if (first) { - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); first = 0; } if (!skb) @@ -538,7 +544,7 @@ empty_source: igmpv3_sendpack(skb); skb = NULL; /* add_grhead will get a new one */ } - skb = add_grhead(skb, pmc, type, &pgr); + skb = add_grhead(skb, pmc, type, &pgr, mtu); } } if (pgr) --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@ -349,8 +349,8 @@ static int ip_tunnel_bind_dev(struct net dev->needed_headroom = t_hlen + hlen; mtu -= (dev->hard_header_len + t_hlen); - if (mtu < 68) - mtu = 68; + if (mtu < IPV4_MIN_MTU) + mtu = IPV4_MIN_MTU; return mtu; } Patches currently in stable-queue which might be from edumazet(a)google.com are queue-4.14/tcp-refresh-tcp_mstamp-from-timers-callbacks.patch queue-4.14/net-fix-double-free-and-memory-corruption-in-get_net_ns_by_id.patch queue-4.14/sock-free-skb-in-skb_complete_tx_timestamp-on-error.patch queue-4.14/tcp-invalidate-rate-samples-during-sack-reneging.patch queue-4.14/tcp-md5sig-use-skb-s-saddr-when-replying-to-an-incoming-segment.patch queue-4.14/net-ipv4-fix-for-a-race-condition-in-raw_sendmsg.patch queue-4.14/ipv4-igmp-guard-against-silly-mtu-values.patch queue-4.14/tcp-fix-potential-underestimation-on-rcv_rtt.patch queue-4.14/ipv6-mcast-better-catch-silly-mtu-values.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: Fix use-after-free when flushing FIB tables" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: Fix use-after-free when flushing FIB tables to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-fix-use-after-free-when-flushing-fib-tables.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Ido Schimmel <idosch(a)mellanox.com> Date: Wed, 20 Dec 2017 19:34:19 +0200 Subject: ipv4: Fix use-after-free when flushing FIB tables From: Ido Schimmel <idosch(a)mellanox.com> [ Upstream commit b4681c2829e24943aadd1a7bb3a30d41d0a20050 ] Since commit 0ddcf43d5d4a ("ipv4: FIB Local/MAIN table collapse") the local table uses the same trie allocated for the main table when custom rules are not in use. When a net namespace is dismantled, the main table is flushed and freed (via an RCU callback) before the local table. In case the callback is invoked before the local table is iterated, a use-after-free can occur. Fix this by iterating over the FIB tables in reverse order, so that the main table is always freed after the local table. v3: Reworded comment according to Alex's suggestion. v2: Add a comment to make the fix more explicit per Dave's and Alex's feedback. Fixes: 0ddcf43d5d4a ("ipv4: FIB Local/MAIN table collapse") Signed-off-by: Ido Schimmel <idosch(a)mellanox.com> Reported-by: Fengguang Wu <fengguang.wu(a)intel.com> Acked-by: Alexander Duyck <alexander.h.duyck(a)intel.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/fib_frontend.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -1274,14 +1274,19 @@ err_table_hash_alloc: static void ip_fib_net_exit(struct net *net) { - unsigned int i; + int i; rtnl_lock(); #ifdef CONFIG_IP_MULTIPLE_TABLES RCU_INIT_POINTER(net->ipv4.fib_main, NULL); RCU_INIT_POINTER(net->ipv4.fib_default, NULL); #endif - for (i = 0; i < FIB_TABLE_HASHSZ; i++) { + /* Destroy the tables in reverse order to guarantee that the + * local table, ID 255, is destroyed before the main table, ID + * 254. This is necessary as the local table may contain + * references to data contained in the main table. + */ + for (i = FIB_TABLE_HASHSZ - 1; i >= 0; i--) { struct hlist_head *head = &net->ipv4.fib_table_hash[i]; struct hlist_node *tmp; struct fib_table *tb; Patches currently in stable-queue which might be from idosch(a)mellanox.com are queue-4.14/ipv6-honor-specified-parameters-in-fibmatch-lookup.patch queue-4.14/mlxsw-spectrum-disable-mac-learning-for-ovs-port.patch queue-4.14/ipv4-fix-use-after-free-when-flushing-fib-tables.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ipv4: fib: Fix metrics match when deleting a route" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ipv4: fib: Fix metrics match when deleting a route to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ipv4-fib-fix-metrics-match-when-deleting-a-route.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Phil Sutter <phil(a)nwl.cc> Date: Tue, 19 Dec 2017 15:17:13 +0100 Subject: ipv4: fib: Fix metrics match when deleting a route From: Phil Sutter <phil(a)nwl.cc> [ Upstream commit d03a45572efa068fa64db211d6d45222660e76c5 ] The recently added fib_metrics_match() causes a regression for routes with both RTAX_FEATURES and RTAX_CC_ALGO if the latter has TCP_CONG_NEEDS_ECN flag set: | # ip link add d0 type dummy | # ip link set d0 up | # ip route add 172.29.29.0/24 dev d0 features ecn congctl dctcp | # ip route del 172.29.29.0/24 dev d0 features ecn congctl dctcp | RTNETLINK answers: No such process During route insertion, fib_convert_metrics() detects that the given CC algo requires ECN and hence sets DST_FEATURE_ECN_CA bit in RTAX_FEATURES. During route deletion though, fib_metrics_match() compares stored RTAX_FEATURES value with that from userspace (which obviously has no knowledge about DST_FEATURE_ECN_CA) and fails. Fixes: 5f9ae3d9e7e4a ("ipv4: do metrics match when looking up and deleting a route") Signed-off-by: Phil Sutter <phil(a)nwl.cc> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv4/fib_semantics.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -706,7 +706,7 @@ bool fib_metrics_match(struct fib_config nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) { int type = nla_type(nla); - u32 val; + u32 fi_val, val; if (!type) continue; @@ -723,7 +723,11 @@ bool fib_metrics_match(struct fib_config val = nla_get_u32(nla); } - if (fi->fib_metrics->metrics[type - 1] != val) + fi_val = fi->fib_metrics->metrics[type - 1]; + if (type == RTAX_FEATURES) + fi_val &= ~DST_FEATURE_ECN_CA; + + if (fi_val != val) return false; } Patches currently in stable-queue which might be from phil(a)nwl.cc are queue-4.14/ipv4-fib-fix-metrics-match-when-deleting-a-route.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ip6_gre: fix device features for ioctl setup" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ip6_gre: fix device features for ioctl setup to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ip6_gre-fix-device-features-for-ioctl-setup.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Alexey Kodanev <alexey.kodanev(a)oracle.com> Date: Wed, 20 Dec 2017 19:36:03 +0300 Subject: ip6_gre: fix device features for ioctl setup From: Alexey Kodanev <alexey.kodanev(a)oracle.com> [ Upstream commit e5a9336adb317db55eb3fe8200856096f3c71109 ] When ip6gre is created using ioctl, its features, such as scatter-gather, GSO and tx-checksumming will be turned off: # ip -f inet6 tunnel add gre6 mode ip6gre remote fd00::1 # ethtool -k gre6 (truncated output) tx-checksumming: off scatter-gather: off tcp-segmentation-offload: off generic-segmentation-offload: off [requested on] But when netlink is used, they will be enabled: # ip link add gre6 type ip6gre remote fd00::1 # ethtool -k gre6 (truncated output) tx-checksumming: on scatter-gather: on tcp-segmentation-offload: on generic-segmentation-offload: on This results in a loss of performance when gre6 is created via ioctl. The issue was found with LTP/gre tests. Fix it by moving the setup of device features to a separate function and invoke it with ndo_init callback because both netlink and ioctl will eventually call it via register_netdevice(): register_netdevice() - ndo_init() callback -> ip6gre_tunnel_init() or ip6gre_tap_init() - ip6gre_tunnel_init_common() - ip6gre_tnl_init_features() The moved code also contains two minor style fixes: * removed needless tab from GRE6_FEATURES on NETIF_F_HIGHDMA line. * fixed the issue reported by checkpatch: "Unnecessary parentheses around 'nt->encap.type == TUNNEL_ENCAP_NONE'" Fixes: ac4eb009e477 ("ip6gre: Add support for basic offloads offloads excluding GSO") Signed-off-by: Alexey Kodanev <alexey.kodanev(a)oracle.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/ip6_gre.c | 57 +++++++++++++++++++++++++++++------------------------ 1 file changed, 32 insertions(+), 25 deletions(-) --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -1020,6 +1020,36 @@ static void ip6gre_tunnel_setup(struct n eth_random_addr(dev->perm_addr); } +#define GRE6_FEATURES (NETIF_F_SG | \ + NETIF_F_FRAGLIST | \ + NETIF_F_HIGHDMA | \ + NETIF_F_HW_CSUM) + +static void ip6gre_tnl_init_features(struct net_device *dev) +{ + struct ip6_tnl *nt = netdev_priv(dev); + + dev->features |= GRE6_FEATURES; + dev->hw_features |= GRE6_FEATURES; + + if (!(nt->parms.o_flags & TUNNEL_SEQ)) { + /* TCP offload with GRE SEQ is not supported, nor + * can we support 2 levels of outer headers requiring + * an update. + */ + if (!(nt->parms.o_flags & TUNNEL_CSUM) || + nt->encap.type == TUNNEL_ENCAP_NONE) { + dev->features |= NETIF_F_GSO_SOFTWARE; + dev->hw_features |= NETIF_F_GSO_SOFTWARE; + } + + /* Can use a lockless transmit, unless we generate + * output sequences + */ + dev->features |= NETIF_F_LLTX; + } +} + static int ip6gre_tunnel_init_common(struct net_device *dev) { struct ip6_tnl *tunnel; @@ -1054,6 +1084,8 @@ static int ip6gre_tunnel_init_common(str if (!(tunnel->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT)) dev->mtu -= 8; + ip6gre_tnl_init_features(dev); + return 0; } @@ -1302,11 +1334,6 @@ static const struct net_device_ops ip6gr .ndo_get_iflink = ip6_tnl_get_iflink, }; -#define GRE6_FEATURES (NETIF_F_SG | \ - NETIF_F_FRAGLIST | \ - NETIF_F_HIGHDMA | \ - NETIF_F_HW_CSUM) - static void ip6gre_tap_setup(struct net_device *dev) { @@ -1386,26 +1413,6 @@ static int ip6gre_newlink(struct net *sr nt->net = dev_net(dev); ip6gre_tnl_link_config(nt, !tb[IFLA_MTU]); - dev->features |= GRE6_FEATURES; - dev->hw_features |= GRE6_FEATURES; - - if (!(nt->parms.o_flags & TUNNEL_SEQ)) { - /* TCP offload with GRE SEQ is not supported, nor - * can we support 2 levels of outer headers requiring - * an update. - */ - if (!(nt->parms.o_flags & TUNNEL_CSUM) || - (nt->encap.type == TUNNEL_ENCAP_NONE)) { - dev->features |= NETIF_F_GSO_SOFTWARE; - dev->hw_features |= NETIF_F_GSO_SOFTWARE; - } - - /* Can use a lockless transmit, unless we generate - * output sequences - */ - dev->features |= NETIF_F_LLTX; - } - err = register_netdevice(dev); if (err) goto out; Patches currently in stable-queue which might be from alexey.kodanev(a)oracle.com are queue-4.14/vxlan-restore-dev-mtu-setting-based-on-lower-device.patch queue-4.14/ip6_gre-fix-device-features-for-ioctl-setup.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "bnxt_en: Fix sources of spurious netpoll warnings" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled bnxt_en: Fix sources of spurious netpoll warnings to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: bnxt_en-fix-sources-of-spurious-netpoll-warnings.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: Calvin Owens <calvinowens(a)fb.com> Date: Fri, 8 Dec 2017 09:05:26 -0800 Subject: bnxt_en: Fix sources of spurious netpoll warnings From: Calvin Owens <calvinowens(a)fb.com> [ Upstream commit 2edbdb3159d6f6bd3a9b6e7f789f2b879699a519 ] After applying 2270bc5da3497945 ("bnxt_en: Fix netpoll handling") and 903649e718f80da2 ("bnxt_en: Improve -ENOMEM logic in NAPI poll loop."), we still see the following WARN fire: ------------[ cut here ]------------ WARNING: CPU: 0 PID: 1875170 at net/core/netpoll.c:165 netpoll_poll_dev+0x15a/0x160 bnxt_poll+0x0/0xd0 exceeded budget in poll <snip> Call Trace: [<ffffffff814be5cd>] dump_stack+0x4d/0x70 [<ffffffff8107e013>] __warn+0xd3/0xf0 [<ffffffff8107e07f>] warn_slowpath_fmt+0x4f/0x60 [<ffffffff8179519a>] netpoll_poll_dev+0x15a/0x160 [<ffffffff81795f38>] netpoll_send_skb_on_dev+0x168/0x250 [<ffffffff817962fc>] netpoll_send_udp+0x2dc/0x440 [<ffffffff815fa9be>] write_ext_msg+0x20e/0x250 [<ffffffff810c8125>] call_console_drivers.constprop.23+0xa5/0x110 [<ffffffff810c9549>] console_unlock+0x339/0x5b0 [<ffffffff810c9a88>] vprintk_emit+0x2c8/0x450 [<ffffffff810c9d5f>] vprintk_default+0x1f/0x30 [<ffffffff81173df5>] printk+0x48/0x50 [<ffffffffa0197713>] edac_raw_mc_handle_error+0x563/0x5c0 [edac_core] [<ffffffffa0197b9b>] edac_mc_handle_error+0x42b/0x6e0 [edac_core] [<ffffffffa01c3a60>] sbridge_mce_output_error+0x410/0x10d0 [sb_edac] [<ffffffffa01c47cc>] sbridge_check_error+0xac/0x130 [sb_edac] [<ffffffffa0197f3c>] edac_mc_workq_function+0x3c/0x90 [edac_core] [<ffffffff81095f8b>] process_one_work+0x19b/0x480 [<ffffffff810967ca>] worker_thread+0x6a/0x520 [<ffffffff8109c7c4>] kthread+0xe4/0x100 [<ffffffff81884c52>] ret_from_fork+0x22/0x40 This happens because we increment rx_pkts on -ENOMEM and -EIO, resulting in rx_pkts > 0. Fix this by only bumping rx_pkts if we were actually given a non-zero budget. Signed-off-by: Calvin Owens <calvinowens(a)fb.com> Acked-by: Michael Chan <michael.chan(a)broadcom.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -1875,7 +1875,7 @@ static int bnxt_poll_work(struct bnxt *b * here forever if we consistently cannot allocate * buffers. */ - else if (rc == -ENOMEM) + else if (rc == -ENOMEM && budget) rx_pkts++; else if (rc == -EBUSY) /* partial completion */ break; @@ -1961,7 +1961,7 @@ static int bnxt_poll_nitroa0(struct napi cpu_to_le32(RX_CMPL_ERRORS_CRC_ERROR); rc = bnxt_rx_pkt(bp, bnapi, &raw_cons, &event); - if (likely(rc == -EIO)) + if (likely(rc == -EIO) && budget) rx_pkts++; else if (rc == -EBUSY) /* partial completion */ break; Patches currently in stable-queue which might be from calvinowens(a)fb.com are queue-4.14/bnxt_en-fix-sources-of-spurious-netpoll-warnings.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "adding missing rcu_read_unlock in ipxip6_rcv" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled adding missing rcu_read_unlock in ipxip6_rcv to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: adding-missing-rcu_read_unlock-in-ipxip6_rcv.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Sun Dec 31 11:12:48 CET 2017 From: "Nikita V. Shirokov" <tehnerd(a)fb.com> Date: Wed, 6 Dec 2017 17:15:43 -0800 Subject: adding missing rcu_read_unlock in ipxip6_rcv From: "Nikita V. Shirokov" <tehnerd(a)fb.com> [ Upstream commit 74c4b656c3d92ec4c824ea1a4afd726b7b6568c8 ] commit 8d79266bc48c ("ip6_tunnel: add collect_md mode to IPv6 tunnels") introduced new exit point in ipxip6_rcv. however rcu_read_unlock is missing there. this diff is fixing this v1->v2: instead of doing rcu_read_unlock in place, we are going to "drop" section (to prevent skb leakage) Fixes: 8d79266bc48c ("ip6_tunnel: add collect_md mode to IPv6 tunnels") Signed-off-by: Nikita V. Shirokov <tehnerd(a)fb.com> Acked-by: Alexei Starovoitov <ast(a)kernel.org> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- net/ipv6/ip6_tunnel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -912,7 +912,7 @@ static int ipxip6_rcv(struct sk_buff *sk if (t->parms.collect_md) { tun_dst = ipv6_tun_rx_dst(skb, 0, 0, 0); if (!tun_dst) - return 0; + goto drop; } ret = __ip6_tnl_rcv(t, skb, tpi, tun_dst, dscp_ecn_decapsulate, log_ecn_error); Patches currently in stable-queue which might be from tehnerd(a)fb.com are queue-4.14/adding-missing-rcu_read_unlock-in-ipxip6_rcv.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Enable CR4.PCIDE on supported systems" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Enable CR4.PCIDE on supported systems to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-enable-cr4.pcide-on-supported-systems.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 660da7c9228f685b2ebe664f9fd69aaddcc420b5 Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Thu, 29 Jun 2017 08:53:21 -0700 Subject: x86/mm: Enable CR4.PCIDE on supported systems From: Andy Lutomirski <luto(a)kernel.org> commit 660da7c9228f685b2ebe664f9fd69aaddcc420b5 upstream. We can use PCID if the CPU has PCID and PGE and we're not on Xen. By itself, this has no effect. A followup patch will start using PCID. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Reviewed-by: Nadav Amit <nadav.amit(a)gmail.com> Reviewed-by: Boris Ostrovsky <boris.ostrovsky(a)oracle.com> Reviewed-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Juergen Gross <jgross(a)suse.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/6327ecd907b32f79d5aa0d466f04503bbec5df88.149875120… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/tlbflush.h | 8 ++++++++ arch/x86/kernel/cpu/common.c | 22 ++++++++++++++++++++++ arch/x86/xen/enlighten.c | 6 ++++++ 3 files changed, 36 insertions(+) --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -191,6 +191,14 @@ static inline void __flush_tlb_all(void) __flush_tlb_global(); else __flush_tlb(); + + /* + * Note: if we somehow had PCID but not PGE, then this wouldn't work -- + * we'd end up flushing kernel translations for the current ASID but + * we might fail to flush kernel translations for other cached ASIDs. + * + * To avoid this issue, we force PCID off if PGE is off. + */ } static inline void __flush_tlb_one(unsigned long addr) --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -324,6 +324,25 @@ static __always_inline void setup_smap(s } } +static void setup_pcid(struct cpuinfo_x86 *c) +{ + if (cpu_has(c, X86_FEATURE_PCID)) { + if (cpu_has(c, X86_FEATURE_PGE)) { + cr4_set_bits(X86_CR4_PCIDE); + } else { + /* + * flush_tlb_all(), as currently implemented, won't + * work if PCID is on but PGE is not. Since that + * combination doesn't exist on real hardware, there's + * no reason to try to fully support it, but it's + * polite to avoid corrupting data if we're on + * an improperly configured VM. + */ + clear_cpu_cap(c, X86_FEATURE_PCID); + } + } +} + /* * Protection Keys are not available in 32-bit mode. */ @@ -1082,6 +1101,9 @@ static void identify_cpu(struct cpuinfo_ setup_smep(c); setup_smap(c); + /* Set up PCID */ + setup_pcid(c); + /* * The vendor-specific functions might have changed features. * Now we do "generic changes." --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -444,6 +444,12 @@ static void __init xen_init_cpuid_mask(v ~((1 << X86_FEATURE_MTRR) | /* disable MTRR */ (1 << X86_FEATURE_ACC)); /* thermal monitoring */ + /* + * Xen PV would need some work to support PCID: CR3 handling as well + * as xen_flush_tlb_others() would need updating. + */ + cpuid_leaf1_ecx_mask &= ~(1 << X86_FEATURE_PCID); /* disable PCID */ + if (!xen_initial_domain()) cpuid_leaf1_edx_mask &= ~((1 << X86_FEATURE_ACPI)); /* disable ACPI */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch queue-4.9/x86-mm-enable-cr4.pcide-on-supported-systems.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch queue-4.9/x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm/64: Fix reboot interaction with CR4.PCIDE" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm/64: Fix reboot interaction with CR4.PCIDE to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 924c6b900cfdf376b07bccfd80e62b21914f8a5a Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sun, 8 Oct 2017 21:53:05 -0700 Subject: x86/mm/64: Fix reboot interaction with CR4.PCIDE From: Andy Lutomirski <luto(a)kernel.org> commit 924c6b900cfdf376b07bccfd80e62b21914f8a5a upstream. Trying to reboot via real mode fails with PCID on: long mode cannot be exited while CR4.PCIDE is set. (No, I have no idea why, but the SDM and actual CPUs are in agreement here.) The result is a GPF and a hang instead of a reboot. I didn't catch this in testing because neither my computer nor my VM reboots this way. I can trigger it with reboot=bios, though. Fixes: 660da7c9228f ("x86/mm: Enable CR4.PCIDE on supported systems") Reported-and-tested-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Borislav Petkov <bp(a)alien8.de> Link: https://lkml.kernel.org/r/f1e7d965998018450a7a70c2823873686a8b21c0.15075247… Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/kernel/reboot.c | 4 ++++ 1 file changed, 4 insertions(+) --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -106,6 +106,10 @@ void __noreturn machine_real_restart(uns load_cr3(initial_page_table); #else write_cr3(real_mode_header->trampoline_pgd); + + /* Exiting long mode will fail if CR4.PCIDE is set. */ + if (static_cpu_has(X86_FEATURE_PCID)) + cr4_clear_bits(X86_CR4_PCIDE); #endif /* Jump to the identity-mapped low memory code */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch queue-4.9/x86-mm-enable-cr4.pcide-on-supported-systems.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch queue-4.9/x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Add the 'nopcid' boot option to turn off PCID" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Add the 'nopcid' boot option to turn off PCID to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 0790c9aad84901ca1bdc14746175549c8b5da215 Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Thu, 29 Jun 2017 08:53:20 -0700 Subject: x86/mm: Add the 'nopcid' boot option to turn off PCID From: Andy Lutomirski <luto(a)kernel.org> commit 0790c9aad84901ca1bdc14746175549c8b5da215 upstream. The parameter is only present on x86_64 systems to save a few bytes, as PCID is always disabled on x86_32. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Reviewed-by: Nadav Amit <nadav.amit(a)gmail.com> Reviewed-by: Borislav Petkov <bp(a)suse.de> Reviewed-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/8bbb2e65bcd249a5f18bfb8128b4689f08ac2b60.149875120… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- Documentation/kernel-parameters.txt | 2 ++ arch/x86/kernel/cpu/common.c | 18 ++++++++++++++++++ 2 files changed, 20 insertions(+) --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -2795,6 +2795,8 @@ bytes respectively. Such letter suffixes nopat [X86] Disable PAT (page attribute table extension of pagetables) support. + nopcid [X86-64] Disable the PCID cpu feature. + norandmaps Don't use address space randomization. Equivalent to echo 0 > /proc/sys/kernel/randomize_va_space --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -163,6 +163,24 @@ static int __init x86_mpx_setup(char *s) } __setup("nompx", x86_mpx_setup); +#ifdef CONFIG_X86_64 +static int __init x86_pcid_setup(char *s) +{ + /* require an exact match without trailing characters */ + if (strlen(s)) + return 0; + + /* do not emit a message if the feature is not present */ + if (!boot_cpu_has(X86_FEATURE_PCID)) + return 1; + + setup_clear_cpu_cap(X86_FEATURE_PCID); + pr_info("nopcid: PCID feature disabled\n"); + return 1; +} +__setup("nopcid", x86_pcid_setup); +#endif + static int __init x86_noinvpcid_setup(char *s) { /* noinvpcid doesn't accept parameters */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch queue-4.9/x86-mm-enable-cr4.pcide-on-supported-systems.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch queue-4.9/x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Enable CR4.PCIDE on supported systems" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Enable CR4.PCIDE on supported systems to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-enable-cr4.pcide-on-supported-systems.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 660da7c9228f685b2ebe664f9fd69aaddcc420b5 Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Thu, 29 Jun 2017 08:53:21 -0700 Subject: x86/mm: Enable CR4.PCIDE on supported systems From: Andy Lutomirski <luto(a)kernel.org> commit 660da7c9228f685b2ebe664f9fd69aaddcc420b5 upstream. We can use PCID if the CPU has PCID and PGE and we're not on Xen. By itself, this has no effect. A followup patch will start using PCID. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Reviewed-by: Nadav Amit <nadav.amit(a)gmail.com> Reviewed-by: Boris Ostrovsky <boris.ostrovsky(a)oracle.com> Reviewed-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Juergen Gross <jgross(a)suse.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/6327ecd907b32f79d5aa0d466f04503bbec5df88.149875120… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/tlbflush.h | 8 ++++++++ arch/x86/kernel/cpu/common.c | 22 ++++++++++++++++++++++ arch/x86/xen/enlighten.c | 6 ++++++ 3 files changed, 36 insertions(+) --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -190,6 +190,14 @@ static inline void __flush_tlb_all(void) __flush_tlb_global(); else __flush_tlb(); + + /* + * Note: if we somehow had PCID but not PGE, then this wouldn't work -- + * we'd end up flushing kernel translations for the current ASID but + * we might fail to flush kernel translations for other cached ASIDs. + * + * To avoid this issue, we force PCID off if PGE is off. + */ } static inline void __flush_tlb_one(unsigned long addr) --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -321,6 +321,25 @@ static __always_inline void setup_smap(s } } +static void setup_pcid(struct cpuinfo_x86 *c) +{ + if (cpu_has(c, X86_FEATURE_PCID)) { + if (cpu_has(c, X86_FEATURE_PGE)) { + cr4_set_bits(X86_CR4_PCIDE); + } else { + /* + * flush_tlb_all(), as currently implemented, won't + * work if PCID is on but PGE is not. Since that + * combination doesn't exist on real hardware, there's + * no reason to try to fully support it, but it's + * polite to avoid corrupting data if we're on + * an improperly configured VM. + */ + clear_cpu_cap(c, X86_FEATURE_PCID); + } + } +} + /* * Some CPU features depend on higher CPUID levels, which may not always * be available due to CPUID level capping or broken virtualization @@ -952,6 +971,9 @@ static void identify_cpu(struct cpuinfo_ setup_smep(c); setup_smap(c); + /* Set up PCID */ + setup_pcid(c); + /* * The vendor-specific functions might have changed features. * Now we do "generic changes." --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -433,6 +433,12 @@ static void __init xen_init_cpuid_mask(v ~((1 << X86_FEATURE_MTRR) | /* disable MTRR */ (1 << X86_FEATURE_ACC)); /* thermal monitoring */ + /* + * Xen PV would need some work to support PCID: CR3 handling as well + * as xen_flush_tlb_others() would need updating. + */ + cpuid_leaf1_ecx_mask &= ~(1 << X86_FEATURE_PCID); /* disable PCID */ + if (!xen_initial_domain()) cpuid_leaf1_edx_mask &= ~((1 << X86_FEATURE_ACPI)); /* disable ACPI */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch queue-4.4/x86-mm-enable-cr4.pcide-on-supported-systems.patch queue-4.4/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.4/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.4/x86-mm-disable-pcid-on-32-bit-kernels.patch queue-4.4/x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm/64: Fix reboot interaction with CR4.PCIDE" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm/64: Fix reboot interaction with CR4.PCIDE to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 924c6b900cfdf376b07bccfd80e62b21914f8a5a Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sun, 8 Oct 2017 21:53:05 -0700 Subject: x86/mm/64: Fix reboot interaction with CR4.PCIDE From: Andy Lutomirski <luto(a)kernel.org> commit 924c6b900cfdf376b07bccfd80e62b21914f8a5a upstream. Trying to reboot via real mode fails with PCID on: long mode cannot be exited while CR4.PCIDE is set. (No, I have no idea why, but the SDM and actual CPUs are in agreement here.) The result is a GPF and a hang instead of a reboot. I didn't catch this in testing because neither my computer nor my VM reboots this way. I can trigger it with reboot=bios, though. Fixes: 660da7c9228f ("x86/mm: Enable CR4.PCIDE on supported systems") Reported-and-tested-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Borislav Petkov <bp(a)alien8.de> Link: https://lkml.kernel.org/r/f1e7d965998018450a7a70c2823873686a8b21c0.15075247… Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/kernel/reboot.c | 4 ++++ 1 file changed, 4 insertions(+) --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -93,6 +93,10 @@ void __noreturn machine_real_restart(uns load_cr3(initial_page_table); #else write_cr3(real_mode_header->trampoline_pgd); + + /* Exiting long mode will fail if CR4.PCIDE is set. */ + if (static_cpu_has(X86_FEATURE_PCID)) + cr4_clear_bits(X86_CR4_PCIDE); #endif /* Jump to the identity-mapped low memory code */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch queue-4.4/x86-mm-enable-cr4.pcide-on-supported-systems.patch queue-4.4/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.4/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.4/x86-mm-disable-pcid-on-32-bit-kernels.patch queue-4.4/x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Add the 'nopcid' boot option to turn off PCID" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Add the 'nopcid' boot option to turn off PCID to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 0790c9aad84901ca1bdc14746175549c8b5da215 Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Thu, 29 Jun 2017 08:53:20 -0700 Subject: x86/mm: Add the 'nopcid' boot option to turn off PCID From: Andy Lutomirski <luto(a)kernel.org> commit 0790c9aad84901ca1bdc14746175549c8b5da215 upstream. The parameter is only present on x86_64 systems to save a few bytes, as PCID is always disabled on x86_32. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Reviewed-by: Nadav Amit <nadav.amit(a)gmail.com> Reviewed-by: Borislav Petkov <bp(a)suse.de> Reviewed-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/8bbb2e65bcd249a5f18bfb8128b4689f08ac2b60.149875120… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- Documentation/kernel-parameters.txt | 2 ++ arch/x86/kernel/cpu/common.c | 18 ++++++++++++++++++ 2 files changed, 20 insertions(+) --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -2555,6 +2555,8 @@ bytes respectively. Such letter suffixes nopat [X86] Disable PAT (page attribute table extension of pagetables) support. + nopcid [X86-64] Disable the PCID cpu feature. + norandmaps Don't use address space randomization. Equivalent to echo 0 > /proc/sys/kernel/randomize_va_space --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -162,6 +162,24 @@ static int __init x86_mpx_setup(char *s) } __setup("nompx", x86_mpx_setup); +#ifdef CONFIG_X86_64 +static int __init x86_pcid_setup(char *s) +{ + /* require an exact match without trailing characters */ + if (strlen(s)) + return 0; + + /* do not emit a message if the feature is not present */ + if (!boot_cpu_has(X86_FEATURE_PCID)) + return 1; + + setup_clear_cpu_cap(X86_FEATURE_PCID); + pr_info("nopcid: PCID feature disabled\n"); + return 1; +} +__setup("nopcid", x86_pcid_setup); +#endif + static int __init x86_noinvpcid_setup(char *s) { /* noinvpcid doesn't accept parameters */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-add-the-nopcid-boot-option-to-turn-off-pcid.patch queue-4.4/x86-mm-enable-cr4.pcide-on-supported-systems.patch queue-4.4/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.4/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.4/x86-mm-disable-pcid-on-32-bit-kernels.patch queue-4.4/x86-mm-64-fix-reboot-interaction-with-cr4.pcide.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "kbuild: add '-fno-stack-check' to kernel build options" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled kbuild: add '-fno-stack-check' to kernel build options to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: kbuild-add-fno-stack-check-to-kernel-build-options.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 3ce120b16cc548472f80cf8644f90eda958cf1b6 Mon Sep 17 00:00:00 2001 From: Linus Torvalds <torvalds(a)linux-foundation.org> Date: Fri, 29 Dec 2017 17:34:43 -0800 Subject: kbuild: add '-fno-stack-check' to kernel build options MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Linus Torvalds <torvalds(a)linux-foundation.org> commit 3ce120b16cc548472f80cf8644f90eda958cf1b6 upstream. It appears that hardened gentoo enables "-fstack-check" by default for gcc. That doesn't work _at_all_ for the kernel, because the kernel stack doesn't act like a user stack at all: it's much smaller, and it doesn't auto-expand on use. So the extra "probe one page below the stack" code generated by -fstack-check just breaks the kernel in horrible ways, causing infinite double faults etc. [ I have to say, that the particular code gcc generates looks very stupid even for user space where it works, but that's a separate issue. ] Reported-and-tested-by: Alexander Tsoy <alexander(a)tsoy.me> Reported-and-tested-by: Toralf Förster <toralf.foerster(a)gmx.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Jiri Kosina <jikos(a)kernel.org> Cc: Andy Lutomirski <luto(a)amacapital.net> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- Makefile | 3 +++ 1 file changed, 3 insertions(+) --- a/Makefile +++ b/Makefile @@ -788,6 +788,9 @@ KBUILD_CFLAGS += $(call cc-disable-warni # disable invalid "can't wrap" optimizations for signed / pointers KBUILD_CFLAGS += $(call cc-option,-fno-strict-overflow) +# Make sure -fstack-check isn't enabled (like gentoo apparently did) +KBUILD_CFLAGS += $(call cc-option,-fno-stack-check,) + # conserve stack if available KBUILD_CFLAGS += $(call cc-option,-fconserve-stack) Patches currently in stable-queue which might be from torvalds(a)linux-foundation.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/kbuild-add-fno-stack-check-to-kernel-build-options.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "kbuild: add '-fno-stack-check' to kernel build options" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled kbuild: add '-fno-stack-check' to kernel build options to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: kbuild-add-fno-stack-check-to-kernel-build-options.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 3ce120b16cc548472f80cf8644f90eda958cf1b6 Mon Sep 17 00:00:00 2001 From: Linus Torvalds <torvalds(a)linux-foundation.org> Date: Fri, 29 Dec 2017 17:34:43 -0800 Subject: kbuild: add '-fno-stack-check' to kernel build options MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Linus Torvalds <torvalds(a)linux-foundation.org> commit 3ce120b16cc548472f80cf8644f90eda958cf1b6 upstream. It appears that hardened gentoo enables "-fstack-check" by default for gcc. That doesn't work _at_all_ for the kernel, because the kernel stack doesn't act like a user stack at all: it's much smaller, and it doesn't auto-expand on use. So the extra "probe one page below the stack" code generated by -fstack-check just breaks the kernel in horrible ways, causing infinite double faults etc. [ I have to say, that the particular code gcc generates looks very stupid even for user space where it works, but that's a separate issue. ] Reported-and-tested-by: Alexander Tsoy <alexander(a)tsoy.me> Reported-and-tested-by: Toralf Förster <toralf.foerster(a)gmx.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Jiri Kosina <jikos(a)kernel.org> Cc: Andy Lutomirski <luto(a)amacapital.net> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- Makefile | 3 +++ 1 file changed, 3 insertions(+) --- a/Makefile +++ b/Makefile @@ -782,6 +782,9 @@ KBUILD_CFLAGS += $(call cc-disable-warni # disable invalid "can't wrap" optimizations for signed / pointers KBUILD_CFLAGS += $(call cc-option,-fno-strict-overflow) +# Make sure -fstack-check isn't enabled (like gentoo apparently did) +KBUILD_CFLAGS += $(call cc-option,-fno-stack-check,) + # conserve stack if available KBUILD_CFLAGS += $(call cc-option,-fconserve-stack) Patches currently in stable-queue which might be from torvalds(a)linux-foundation.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.4/kbuild-add-fno-stack-check-to-kernel-build-options.patch queue-4.4/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.4/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "kbuild: add '-fno-stack-check' to kernel build options" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled kbuild: add '-fno-stack-check' to kernel build options to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: kbuild-add-fno-stack-check-to-kernel-build-options.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 3ce120b16cc548472f80cf8644f90eda958cf1b6 Mon Sep 17 00:00:00 2001 From: Linus Torvalds <torvalds(a)linux-foundation.org> Date: Fri, 29 Dec 2017 17:34:43 -0800 Subject: kbuild: add '-fno-stack-check' to kernel build options MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Linus Torvalds <torvalds(a)linux-foundation.org> commit 3ce120b16cc548472f80cf8644f90eda958cf1b6 upstream. It appears that hardened gentoo enables "-fstack-check" by default for gcc. That doesn't work _at_all_ for the kernel, because the kernel stack doesn't act like a user stack at all: it's much smaller, and it doesn't auto-expand on use. So the extra "probe one page below the stack" code generated by -fstack-check just breaks the kernel in horrible ways, causing infinite double faults etc. [ I have to say, that the particular code gcc generates looks very stupid even for user space where it works, but that's a separate issue. ] Reported-and-tested-by: Alexander Tsoy <alexander(a)tsoy.me> Reported-and-tested-by: Toralf Förster <toralf.foerster(a)gmx.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Jiri Kosina <jikos(a)kernel.org> Cc: Andy Lutomirski <luto(a)amacapital.net> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- Makefile | 3 +++ 1 file changed, 3 insertions(+) --- a/Makefile +++ b/Makefile @@ -802,6 +802,9 @@ KBUILD_CFLAGS += $(call cc-disable-warni # disable invalid "can't wrap" optimizations for signed / pointers KBUILD_CFLAGS += $(call cc-option,-fno-strict-overflow) +# Make sure -fstack-check isn't enabled (like gentoo apparently did) +KBUILD_CFLAGS += $(call cc-option,-fno-stack-check,) + # conserve stack if available KBUILD_CFLAGS += $(call cc-option,-fconserve-stack) Patches currently in stable-queue which might be from torvalds(a)linux-foundation.org are queue-4.14/x86-mm-pti-allow-nx-poison-to-be-set-in-p4d-pgd.patch queue-4.14/x86-dumpstack-indicate-in-oops-whether-pti-is-configured-and-enabled.patch queue-4.14/x86-mm-clarify-the-whole-asid-kernel-pcid-user-pcid-naming.patch queue-4.14/x86-mm-abstract-switching-cr3.patch queue-4.14/x86-mm-optimize-restore_cr3.patch queue-4.14/x86-mm-pti-force-entry-through-trampoline-when-pti-active.patch queue-4.14/x86-entry-align-entry-text-section-to-pmd-boundary.patch queue-4.14/x86-mm-64-make-a-full-pgd-entry-size-hole-in-the-memory-map.patch queue-4.14/x86-cpu_entry_area-add-debugstore-entries-to-cpu_entry_area.patch queue-4.14/x86-mm-pti-share-entry-text-pmd.patch queue-4.14/x86-ldt-make-the-ldt-mapping-ro.patch queue-4.14/x86-mm-dump_pagetables-check-user-space-page-table-for-wx-pages.patch queue-4.14/x86-pti-map-the-vsyscall-page-if-needed.patch queue-4.14/x86-mm-pti-disable-global-pages-if-page_table_isolation-y.patch queue-4.14/x86-mm-dump_pagetables-allow-dumping-current-pagetables.patch queue-4.14/x86-mm-pti-add-infrastructure-for-page-table-isolation.patch queue-4.14/x86-cpufeatures-add-x86_bug_cpu_insecure.patch queue-4.14/x86-mm-use-fix-pcid-to-optimize-user-kernel-switches.patch queue-4.14/x86-mm-use-invpcid-for-__native_flush_tlb_single.patch queue-4.14/x86-mm-pti-prepare-the-x86-entry-assembly-code-for-entry-exit-cr3-switching.patch queue-4.14/x86-mm-allow-flushing-for-future-asid-switches.patch queue-4.14/x86-mm-pti-add-kconfig.patch queue-4.14/x86-mm-pti-add-mapping-helper-functions.patch queue-4.14/x86-mm-pti-map-espfix-into-user-space.patch queue-4.14/kbuild-add-fno-stack-check-to-kernel-build-options.patch queue-4.14/x86-mm-pti-share-cpu_entry_area-with-user-space-page-tables.patch queue-4.14/x86-mm-pti-add-functions-to-clone-kernel-pmds.patch queue-4.14/x86-pti-add-the-pti-cmdline-option-and-documentation.patch queue-4.14/x86-events-intel-ds-map-debug-buffers-in-cpu_entry_area.patch queue-4.14/x86-mm-pti-allocate-a-separate-user-pgd.patch queue-4.14/x86-pti-put-the-ldt-in-its-own-pgd-if-pti-is-on.patch queue-4.14/x86-mm-pti-populate-user-pgd.patch queue-4.14/x86-mm-dump_pagetables-add-page-table-directory-to-the-debugfs-vfs-hierarchy.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "kbuild: add '-fno-stack-check' to kernel build options" has been added to the 3.18-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled kbuild: add '-fno-stack-check' to kernel build options to the 3.18-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: kbuild-add-fno-stack-check-to-kernel-build-options.patch and it can be found in the queue-3.18 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 3ce120b16cc548472f80cf8644f90eda958cf1b6 Mon Sep 17 00:00:00 2001 From: Linus Torvalds <torvalds(a)linux-foundation.org> Date: Fri, 29 Dec 2017 17:34:43 -0800 Subject: kbuild: add '-fno-stack-check' to kernel build options MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Linus Torvalds <torvalds(a)linux-foundation.org> commit 3ce120b16cc548472f80cf8644f90eda958cf1b6 upstream. It appears that hardened gentoo enables "-fstack-check" by default for gcc. That doesn't work _at_all_ for the kernel, because the kernel stack doesn't act like a user stack at all: it's much smaller, and it doesn't auto-expand on use. So the extra "probe one page below the stack" code generated by -fstack-check just breaks the kernel in horrible ways, causing infinite double faults etc. [ I have to say, that the particular code gcc generates looks very stupid even for user space where it works, but that's a separate issue. ] Reported-and-tested-by: Alexander Tsoy <alexander(a)tsoy.me> Reported-and-tested-by: Toralf Förster <toralf.foerster(a)gmx.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Jiri Kosina <jikos(a)kernel.org> Cc: Andy Lutomirski <luto(a)amacapital.net> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- Makefile | 3 +++ 1 file changed, 3 insertions(+) --- a/Makefile +++ b/Makefile @@ -762,6 +762,9 @@ KBUILD_CFLAGS += $(call cc-disable-warni # disable invalid "can't wrap" optimizations for signed / pointers KBUILD_CFLAGS += $(call cc-option,-fno-strict-overflow) +# Make sure -fstack-check isn't enabled (like gentoo apparently did) +KBUILD_CFLAGS += $(call cc-option,-fno-stack-check,) + # conserve stack if available KBUILD_CFLAGS += $(call cc-option,-fconserve-stack) Patches currently in stable-queue which might be from torvalds(a)linux-foundation.org are queue-3.18/kbuild-add-fno-stack-check-to-kernel-build-options.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] How to report regression in stable tree?

by Ian Pilcher

I just discovered that commit 7042c2b9a19e74972f5783ab3b7ef98ebdee293f in 4.14.4 completely breaks the external displays attached to my Dell TB16 dock. How/where should I report this? (Does the stable tree use the regular kernel Bugzilla?) Thanks! -- ======================================================================== Ian Pilcher arequipeno(a)gmail.com -------- "I grew up before Mark Zuckerberg invented friendship" -------- ========================================================================

7 years, 8 months

2
1
0 0

[Linux-stable-mirror] [PATCH] af_key: fix buffer overread in parse_exthdrs()

by Eric Biggers

From: Eric Biggers <ebiggers(a)google.com> If a message sent to a PF_KEY socket ended with an incomplete extension header (fewer than 4 bytes remaining), then parse_exthdrs() read past the end of the message, into uninitialized memory. Fix it by returning -EINVAL in this case. Reproducer: #include <linux/pfkeyv2.h> #include <sys/socket.h> #include <unistd.h> int main() { int sock = socket(PF_KEY, SOCK_RAW, PF_KEY_V2); char buf[17] = { 0 }; struct sadb_msg *msg = (void *)buf; msg->sadb_msg_version = PF_KEY_V2; msg->sadb_msg_type = SADB_DELETE; msg->sadb_msg_len = 2; write(sock, buf, 17); } Cc: stable(a)vger.kernel.org Signed-off-by: Eric Biggers <ebiggers(a)google.com> --- net/key/af_key.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/key/af_key.c b/net/key/af_key.c index 596499cc8b2f..d40861a048fe 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -516,6 +516,9 @@ static int parse_exthdrs(struct sk_buff *skb, const struct sadb_msg *hdr, void * uint16_t ext_type; int ext_len; + if (len < sizeof(*ehdr)) + return -EINVAL; + ext_len = ehdr->sadb_ext_len; ext_len *= sizeof(uint64_t); ext_type = ehdr->sadb_ext_type; -- 2.15.1

7 years, 8 months

2
1
0 0

[Linux-stable-mirror] [PATCH] af_key: fix buffer overread in verify_address_len()

by Eric Biggers

From: Eric Biggers <ebiggers(a)google.com> If a message sent to a PF_KEY socket ended with one of the extensions that takes a 'struct sadb_address' but there were not enough bytes remaining in the message for the ->sa_family member of the 'struct sockaddr' which is supposed to follow, then verify_address_len() read past the end of the message, into uninitialized memory. Fix it by returning -EINVAL in this case. This bug was found using syzkaller with KMSAN. Reproducer: #include <linux/pfkeyv2.h> #include <sys/socket.h> #include <unistd.h> int main() { int sock = socket(PF_KEY, SOCK_RAW, PF_KEY_V2); char buf[24] = { 0 }; struct sadb_msg *msg = (void *)buf; struct sadb_address *addr = (void *)(msg + 1); msg->sadb_msg_version = PF_KEY_V2; msg->sadb_msg_type = SADB_DELETE; msg->sadb_msg_len = 3; addr->sadb_address_len = 1; addr->sadb_address_exttype = SADB_EXT_ADDRESS_SRC; write(sock, buf, 24); } Reported-by: Alexander Potapenko <glider(a)google.com> Cc: stable(a)vger.kernel.org Signed-off-by: Eric Biggers <ebiggers(a)google.com> --- net/key/af_key.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/net/key/af_key.c b/net/key/af_key.c index 3dffb892d52c..596499cc8b2f 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -401,6 +401,11 @@ static int verify_address_len(const void *p) #endif int len; + if (sp->sadb_address_len < + DIV_ROUND_UP(sizeof(*sp) + offsetofend(typeof(*addr), sa_family), + sizeof(uint64_t))) + return -EINVAL; + switch (addr->sa_family) { case AF_INET: len = DIV_ROUND_UP(sizeof(*sp) + sizeof(*sin), sizeof(uint64_t)); -- 2.15.1

7 years, 8 months

2
1
0 0

[Linux-stable-mirror] 4.14.9 doesn't boot (regression)

by Toralf Förster

I can confirm now, that that kernel breaks both a desktop (an ThinkPad T440s i5) and a headless server (i3930) setup. For the server the attached .config works fine but switching from CONFIG_GENERIC_CPU to CONFIG_MCORE2 legt them hang at boot w/op any messages. Similar picture at the desktop. Both are stable Gentoo Linux hardened systems. This issue seems to exist in mainline too, probably visible with d120cd749 (stable) and 9aaefe7b59 (upstream). -- Toralf PGP C4EACDDE 0076E94E

7 years, 8 months

9
31
0 0

[Linux-stable-mirror] v4.4.y-stable-queue went backwards

by Guenter Roeck

Hi Greg, v4.4.y-stable-queue reference used to be v4.4.108-15-gbb44baf, now it is v4.4.108. Did something get lost, or did you drop the pending patches ? Guenter

7 years, 8 months

2
1
0 0

[Linux-stable-mirror] Patch "x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP code" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP code to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ce4a4e565f5264909a18c733b864c3f74467f69e Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sun, 28 May 2017 10:00:14 -0700 Subject: x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP code From: Andy Lutomirski <luto(a)kernel.org> commit ce4a4e565f5264909a18c733b864c3f74467f69e upstream. The UP asm/tlbflush.h generates somewhat nicer code than the SMP version. Aside from that, it's fallen quite a bit behind the SMP code: - flush_tlb_mm_range() didn't flush individual pages if the range was small. - The lazy TLB code was much weaker. This usually wouldn't matter, but, if a kernel thread flushed its lazy "active_mm" more than once (due to reclaim or similar), it wouldn't be unlazied and would instead pointlessly flush repeatedly. - Tracepoints were missing. Aside from that, simply having the UP code around was a maintanence burden, since it means that any change to the TLB flush code had to make sure not to break it. Simplify everything by deleting the UP code. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bpetkov(a)suse.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <nadav.amit(a)gmail.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: linux-mm(a)kvack.org Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/Kconfig | 2 arch/x86/include/asm/hardirq.h | 2 arch/x86/include/asm/mmu.h | 6 -- arch/x86/include/asm/mmu_context.h | 2 arch/x86/include/asm/tlbflush.h | 78 ------------------------------------- arch/x86/mm/init.c | 2 arch/x86/mm/tlb.c | 17 -------- 7 files changed, 5 insertions(+), 104 deletions(-) --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -45,7 +45,7 @@ config X86 select ARCH_USE_CMPXCHG_LOCKREF if X86_64 select ARCH_USE_QUEUED_RWLOCKS select ARCH_USE_QUEUED_SPINLOCKS - select ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH if SMP + select ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH select ARCH_WANTS_DYNAMIC_TASK_STRUCT select ARCH_WANT_FRAME_POINTERS select ARCH_WANT_IPC_PARSE_VERSION if X86_32 --- a/arch/x86/include/asm/hardirq.h +++ b/arch/x86/include/asm/hardirq.h @@ -22,8 +22,8 @@ typedef struct { #ifdef CONFIG_SMP unsigned int irq_resched_count; unsigned int irq_call_count; - unsigned int irq_tlb_count; #endif + unsigned int irq_tlb_count; #ifdef CONFIG_X86_THERMAL_VECTOR unsigned int irq_thermal_count; #endif --- a/arch/x86/include/asm/mmu.h +++ b/arch/x86/include/asm/mmu.h @@ -33,12 +33,6 @@ typedef struct { #endif } mm_context_t; -#ifdef CONFIG_SMP void leave_mm(int cpu); -#else -static inline void leave_mm(int cpu) -{ -} -#endif #endif /* _ASM_X86_MMU_H */ --- a/arch/x86/include/asm/mmu_context.h +++ b/arch/x86/include/asm/mmu_context.h @@ -99,10 +99,8 @@ static inline void load_mm_ldt(struct mm static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk) { -#ifdef CONFIG_SMP if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) this_cpu_write(cpu_tlbstate.state, TLBSTATE_LAZY); -#endif } static inline int init_new_context(struct task_struct *tsk, --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -7,6 +7,7 @@ #include <asm/processor.h> #include <asm/cpufeature.h> #include <asm/special_insns.h> +#include <asm/smp.h> static inline void __invpcid(unsigned long pcid, unsigned long addr, unsigned long type) @@ -65,10 +66,8 @@ static inline void invpcid_flush_all_non #endif struct tlb_state { -#ifdef CONFIG_SMP struct mm_struct *active_mm; int state; -#endif /* * Access to this CR4 shadow and to H/W CR4 is protected by @@ -216,79 +215,6 @@ static inline void __flush_tlb_one(unsig * and page-granular flushes are available only on i486 and up. */ -#ifndef CONFIG_SMP - -/* "_up" is for UniProcessor. - * - * This is a helper for other header functions. *Not* intended to be called - * directly. All global TLB flushes need to either call this, or to bump the - * vm statistics themselves. - */ -static inline void __flush_tlb_up(void) -{ - count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); - __flush_tlb(); -} - -static inline void flush_tlb_all(void) -{ - count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); - __flush_tlb_all(); -} - -static inline void local_flush_tlb(void) -{ - __flush_tlb_up(); -} - -static inline void flush_tlb_mm(struct mm_struct *mm) -{ - if (mm == current->active_mm) - __flush_tlb_up(); -} - -static inline void flush_tlb_page(struct vm_area_struct *vma, - unsigned long addr) -{ - if (vma->vm_mm == current->active_mm) - __flush_tlb_one(addr); -} - -static inline void flush_tlb_range(struct vm_area_struct *vma, - unsigned long start, unsigned long end) -{ - if (vma->vm_mm == current->active_mm) - __flush_tlb_up(); -} - -static inline void flush_tlb_mm_range(struct mm_struct *mm, - unsigned long start, unsigned long end, unsigned long vmflag) -{ - if (mm == current->active_mm) - __flush_tlb_up(); -} - -static inline void native_flush_tlb_others(const struct cpumask *cpumask, - struct mm_struct *mm, - unsigned long start, - unsigned long end) -{ -} - -static inline void reset_lazy_tlbstate(void) -{ -} - -static inline void flush_tlb_kernel_range(unsigned long start, - unsigned long end) -{ - flush_tlb_all(); -} - -#else /* SMP */ - -#include <asm/smp.h> - #define local_flush_tlb() __flush_tlb() #define flush_tlb_mm(mm) flush_tlb_mm_range(mm, 0UL, TLB_FLUSH_ALL, 0UL) @@ -319,8 +245,6 @@ static inline void reset_lazy_tlbstate(v this_cpu_write(cpu_tlbstate.active_mm, &init_mm); } -#endif /* SMP */ - #ifndef CONFIG_PARAVIRT #define flush_tlb_others(mask, mm, start, end) \ native_flush_tlb_others(mask, mm, start, end) --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -764,10 +764,8 @@ void __init zone_sizes_init(void) } DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate) = { -#ifdef CONFIG_SMP .active_mm = &init_mm, .state = 0, -#endif .cr4 = ~0UL, /* fail hard if we screw up cr4 shadow initialization */ }; EXPORT_SYMBOL_GPL(cpu_tlbstate); --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -15,7 +15,7 @@ #include <linux/debugfs.h> /* - * Smarter SMP flushing macros. + * TLB flushing, formerly SMP-only * c/o Linus Torvalds. * * These mean you can really definitely utterly forget about @@ -28,8 +28,6 @@ * Implement flush IPI by CALL_FUNCTION_VECTOR, Alex Shi */ -#ifdef CONFIG_SMP - struct flush_tlb_info { struct mm_struct *flush_mm; unsigned long flush_start; @@ -59,8 +57,6 @@ void leave_mm(int cpu) } EXPORT_SYMBOL_GPL(leave_mm); -#endif /* CONFIG_SMP */ - void switch_mm(struct mm_struct *prev, struct mm_struct *next, struct task_struct *tsk) { @@ -91,10 +87,8 @@ void switch_mm_irqs_off(struct mm_struct set_pgd(pgd, init_mm.pgd[stack_pgd_index]); } -#ifdef CONFIG_SMP this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK); this_cpu_write(cpu_tlbstate.active_mm, next); -#endif cpumask_set_cpu(cpu, mm_cpumask(next)); @@ -152,9 +146,7 @@ void switch_mm_irqs_off(struct mm_struct if (unlikely(prev->context.ldt != next->context.ldt)) load_mm_ldt(next); #endif - } -#ifdef CONFIG_SMP - else { + } else { this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK); BUG_ON(this_cpu_read(cpu_tlbstate.active_mm) != next); @@ -181,11 +173,8 @@ void switch_mm_irqs_off(struct mm_struct load_mm_ldt(next); } } -#endif } -#ifdef CONFIG_SMP - /* * The flush IPI assumes that a thread switch happens in this order: * [cpu0: the cpu that switches] @@ -438,5 +427,3 @@ static int __init create_tlb_single_page return 0; } late_initcall(create_tlb_single_page_flush_ceiling); - -#endif /* CONFIG_SMP */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Reimplement flush_tlb_page() using flush_tlb_mm_range()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Reimplement flush_tlb_page() using flush_tlb_mm_range() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ca6c99c0794875c6d1db6e22f246699691ab7e6b Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Mon, 22 May 2017 15:30:01 -0700 Subject: x86/mm: Reimplement flush_tlb_page() using flush_tlb_mm_range() From: Andy Lutomirski <luto(a)kernel.org> commit ca6c99c0794875c6d1db6e22f246699691ab7e6b upstream. flush_tlb_page() was very similar to flush_tlb_mm_range() except that it had a couple of issues: - It was missing an smp_mb() in the case where current->active_mm != mm. (This is a longstanding bug reported by Nadav Amit) - It was missing tracepoints and vm counter updates. The only reason that I can see for keeping it at as a separate function is that it could avoid a few branches that flush_tlb_mm_range() needs to decide to flush just one page. This hardly seems worthwhile. If we decide we want to get rid of those branches again, a better way would be to introduce an __flush_tlb_mm_range() helper and make both flush_tlb_page() and flush_tlb_mm_range() use it. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Acked-by: Kees Cook <keescook(a)chromium.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bpetkov(a)suse.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <nadav.amit(a)gmail.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/3cc3847cf888d8907577569b8bac3f01992ef8f9.149549206… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/tlbflush.h | 6 +++++- arch/x86/mm/tlb.c | 27 --------------------------- 2 files changed, 5 insertions(+), 28 deletions(-) --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -297,11 +297,15 @@ static inline void flush_tlb_kernel_rang flush_tlb_mm_range(vma->vm_mm, start, end, vma->vm_flags) extern void flush_tlb_all(void); -extern void flush_tlb_page(struct vm_area_struct *, unsigned long); extern void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start, unsigned long end, unsigned long vmflag); extern void flush_tlb_kernel_range(unsigned long start, unsigned long end); +static inline void flush_tlb_page(struct vm_area_struct *vma, unsigned long a) +{ + flush_tlb_mm_range(vma->vm_mm, a, a + PAGE_SIZE, VM_NONE); +} + void native_flush_tlb_others(const struct cpumask *cpumask, struct mm_struct *mm, unsigned long start, unsigned long end); --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -354,33 +354,6 @@ out: preempt_enable(); } -void flush_tlb_page(struct vm_area_struct *vma, unsigned long start) -{ - struct mm_struct *mm = vma->vm_mm; - - preempt_disable(); - - if (current->active_mm == mm) { - if (current->mm) { - /* - * Implicit full barrier (INVLPG) that synchronizes - * with switch_mm. - */ - __flush_tlb_one(start); - } else { - leave_mm(smp_processor_id()); - - /* Synchronize with switch_mm. */ - smp_mb(); - } - } - - if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids) - flush_tlb_others(mm_cpumask(mm), mm, start, start + PAGE_SIZE); - - preempt_enable(); -} - static void do_flush_tlb_all(void *info) { count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED); Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Disable PCID on 32-bit kernels" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Disable PCID on 32-bit kernels to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-disable-pcid-on-32-bit-kernels.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From cba4671af7550e008f7a7835f06df0763825bf3e Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Thu, 29 Jun 2017 08:53:19 -0700 Subject: x86/mm: Disable PCID on 32-bit kernels From: Andy Lutomirski <luto(a)kernel.org> commit cba4671af7550e008f7a7835f06df0763825bf3e upstream. 32-bit kernels on new hardware will see PCID in CPUID, but PCID can only be used in 64-bit mode. Rather than making all PCID code conditional, just disable the feature on 32-bit builds. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Reviewed-by: Nadav Amit <nadav.amit(a)gmail.com> Reviewed-by: Borislav Petkov <bp(a)suse.de> Reviewed-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/2e391769192a4d31b808410c383c6bf0734bc6ea.149875120… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/disabled-features.h | 4 +++- arch/x86/kernel/cpu/bugs.c | 8 ++++++++ 2 files changed, 11 insertions(+), 1 deletion(-) --- a/arch/x86/include/asm/disabled-features.h +++ b/arch/x86/include/asm/disabled-features.h @@ -21,11 +21,13 @@ # define DISABLE_K6_MTRR (1<<(X86_FEATURE_K6_MTRR & 31)) # define DISABLE_CYRIX_ARR (1<<(X86_FEATURE_CYRIX_ARR & 31)) # define DISABLE_CENTAUR_MCR (1<<(X86_FEATURE_CENTAUR_MCR & 31)) +# define DISABLE_PCID 0 #else # define DISABLE_VME 0 # define DISABLE_K6_MTRR 0 # define DISABLE_CYRIX_ARR 0 # define DISABLE_CENTAUR_MCR 0 +# define DISABLE_PCID (1<<(X86_FEATURE_PCID & 31)) #endif /* CONFIG_X86_64 */ #ifdef CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS @@ -43,7 +45,7 @@ #define DISABLED_MASK1 0 #define DISABLED_MASK2 0 #define DISABLED_MASK3 (DISABLE_CYRIX_ARR|DISABLE_CENTAUR_MCR|DISABLE_K6_MTRR) -#define DISABLED_MASK4 0 +#define DISABLED_MASK4 (DISABLE_PCID) #define DISABLED_MASK5 0 #define DISABLED_MASK6 0 #define DISABLED_MASK7 0 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -19,6 +19,14 @@ void __init check_bugs(void) { +#ifdef CONFIG_X86_32 + /* + * Regardless of whether PCID is enumerated, the SDM says + * that it can't be enabled in 32-bit mode. + */ + setup_clear_cpu_cap(X86_FEATURE_PCID); +#endif + identify_boot_cpu(); #ifndef CONFIG_SMP pr_info("CPU: "); Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.9/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.9/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP code" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP code to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ce4a4e565f5264909a18c733b864c3f74467f69e Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sun, 28 May 2017 10:00:14 -0700 Subject: x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP code From: Andy Lutomirski <luto(a)kernel.org> commit ce4a4e565f5264909a18c733b864c3f74467f69e upstream. The UP asm/tlbflush.h generates somewhat nicer code than the SMP version. Aside from that, it's fallen quite a bit behind the SMP code: - flush_tlb_mm_range() didn't flush individual pages if the range was small. - The lazy TLB code was much weaker. This usually wouldn't matter, but, if a kernel thread flushed its lazy "active_mm" more than once (due to reclaim or similar), it wouldn't be unlazied and would instead pointlessly flush repeatedly. - Tracepoints were missing. Aside from that, simply having the UP code around was a maintanence burden, since it means that any change to the TLB flush code had to make sure not to break it. Simplify everything by deleting the UP code. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bpetkov(a)suse.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <nadav.amit(a)gmail.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: linux-mm(a)kvack.org Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/Kconfig | 2 arch/x86/include/asm/hardirq.h | 2 arch/x86/include/asm/mmu.h | 6 -- arch/x86/include/asm/mmu_context.h | 2 arch/x86/include/asm/tlbflush.h | 78 ------------------------------------- arch/x86/mm/init.c | 2 arch/x86/mm/tlb.c | 17 -------- 7 files changed, 5 insertions(+), 104 deletions(-) --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -42,7 +42,7 @@ config X86 select ARCH_USE_CMPXCHG_LOCKREF if X86_64 select ARCH_USE_QUEUED_RWLOCKS select ARCH_USE_QUEUED_SPINLOCKS - select ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH if SMP + select ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH select ARCH_WANTS_DYNAMIC_TASK_STRUCT select ARCH_WANT_FRAME_POINTERS select ARCH_WANT_IPC_PARSE_VERSION if X86_32 --- a/arch/x86/include/asm/hardirq.h +++ b/arch/x86/include/asm/hardirq.h @@ -22,8 +22,8 @@ typedef struct { #ifdef CONFIG_SMP unsigned int irq_resched_count; unsigned int irq_call_count; - unsigned int irq_tlb_count; #endif + unsigned int irq_tlb_count; #ifdef CONFIG_X86_THERMAL_VECTOR unsigned int irq_thermal_count; #endif --- a/arch/x86/include/asm/mmu.h +++ b/arch/x86/include/asm/mmu.h @@ -24,12 +24,6 @@ typedef struct { atomic_t perf_rdpmc_allowed; /* nonzero if rdpmc is allowed */ } mm_context_t; -#ifdef CONFIG_SMP void leave_mm(int cpu); -#else -static inline void leave_mm(int cpu) -{ -} -#endif #endif /* _ASM_X86_MMU_H */ --- a/arch/x86/include/asm/mmu_context.h +++ b/arch/x86/include/asm/mmu_context.h @@ -98,10 +98,8 @@ static inline void load_mm_ldt(struct mm static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk) { -#ifdef CONFIG_SMP if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) this_cpu_write(cpu_tlbstate.state, TLBSTATE_LAZY); -#endif } extern void switch_mm(struct mm_struct *prev, struct mm_struct *next, --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -6,6 +6,7 @@ #include <asm/processor.h> #include <asm/special_insns.h> +#include <asm/smp.h> static inline void __invpcid(unsigned long pcid, unsigned long addr, unsigned long type) @@ -64,10 +65,8 @@ static inline void invpcid_flush_all_non #endif struct tlb_state { -#ifdef CONFIG_SMP struct mm_struct *active_mm; int state; -#endif /* * Access to this CR4 shadow and to H/W CR4 is protected by @@ -215,79 +214,6 @@ static inline void __flush_tlb_one(unsig * and page-granular flushes are available only on i486 and up. */ -#ifndef CONFIG_SMP - -/* "_up" is for UniProcessor. - * - * This is a helper for other header functions. *Not* intended to be called - * directly. All global TLB flushes need to either call this, or to bump the - * vm statistics themselves. - */ -static inline void __flush_tlb_up(void) -{ - count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); - __flush_tlb(); -} - -static inline void flush_tlb_all(void) -{ - count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); - __flush_tlb_all(); -} - -static inline void local_flush_tlb(void) -{ - __flush_tlb_up(); -} - -static inline void flush_tlb_mm(struct mm_struct *mm) -{ - if (mm == current->active_mm) - __flush_tlb_up(); -} - -static inline void flush_tlb_page(struct vm_area_struct *vma, - unsigned long addr) -{ - if (vma->vm_mm == current->active_mm) - __flush_tlb_one(addr); -} - -static inline void flush_tlb_range(struct vm_area_struct *vma, - unsigned long start, unsigned long end) -{ - if (vma->vm_mm == current->active_mm) - __flush_tlb_up(); -} - -static inline void flush_tlb_mm_range(struct mm_struct *mm, - unsigned long start, unsigned long end, unsigned long vmflag) -{ - if (mm == current->active_mm) - __flush_tlb_up(); -} - -static inline void native_flush_tlb_others(const struct cpumask *cpumask, - struct mm_struct *mm, - unsigned long start, - unsigned long end) -{ -} - -static inline void reset_lazy_tlbstate(void) -{ -} - -static inline void flush_tlb_kernel_range(unsigned long start, - unsigned long end) -{ - flush_tlb_all(); -} - -#else /* SMP */ - -#include <asm/smp.h> - #define local_flush_tlb() __flush_tlb() #define flush_tlb_mm(mm) flush_tlb_mm_range(mm, 0UL, TLB_FLUSH_ALL, 0UL) @@ -318,8 +244,6 @@ static inline void reset_lazy_tlbstate(v this_cpu_write(cpu_tlbstate.active_mm, &init_mm); } -#endif /* SMP */ - #ifndef CONFIG_PARAVIRT #define flush_tlb_others(mask, mm, start, end) \ native_flush_tlb_others(mask, mm, start, end) --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -753,10 +753,8 @@ void __init zone_sizes_init(void) } DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate) = { -#ifdef CONFIG_SMP .active_mm = &init_mm, .state = 0, -#endif .cr4 = ~0UL, /* fail hard if we screw up cr4 shadow initialization */ }; EXPORT_SYMBOL_GPL(cpu_tlbstate); --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -15,7 +15,7 @@ #include <linux/debugfs.h> /* - * Smarter SMP flushing macros. + * TLB flushing, formerly SMP-only * c/o Linus Torvalds. * * These mean you can really definitely utterly forget about @@ -28,8 +28,6 @@ * Implement flush IPI by CALL_FUNCTION_VECTOR, Alex Shi */ -#ifdef CONFIG_SMP - struct flush_tlb_info { struct mm_struct *flush_mm; unsigned long flush_start; @@ -59,8 +57,6 @@ void leave_mm(int cpu) } EXPORT_SYMBOL_GPL(leave_mm); -#endif /* CONFIG_SMP */ - void switch_mm(struct mm_struct *prev, struct mm_struct *next, struct task_struct *tsk) { @@ -77,10 +73,8 @@ void switch_mm_irqs_off(struct mm_struct unsigned cpu = smp_processor_id(); if (likely(prev != next)) { -#ifdef CONFIG_SMP this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK); this_cpu_write(cpu_tlbstate.active_mm, next); -#endif cpumask_set_cpu(cpu, mm_cpumask(next)); /* @@ -137,9 +131,7 @@ void switch_mm_irqs_off(struct mm_struct if (unlikely(prev->context.ldt != next->context.ldt)) load_mm_ldt(next); #endif - } -#ifdef CONFIG_SMP - else { + } else { this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK); BUG_ON(this_cpu_read(cpu_tlbstate.active_mm) != next); @@ -166,11 +158,8 @@ void switch_mm_irqs_off(struct mm_struct load_mm_ldt(next); } } -#endif } -#ifdef CONFIG_SMP - /* * The flush IPI assumes that a thread switch happens in this order: * [cpu0: the cpu that switches] @@ -423,5 +412,3 @@ static int __init create_tlb_single_page return 0; } late_initcall(create_tlb_single_page_flush_ceiling); - -#endif /* CONFIG_SMP */ Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.4/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.4/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Reimplement flush_tlb_page() using flush_tlb_mm_range()" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Reimplement flush_tlb_page() using flush_tlb_mm_range() to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ca6c99c0794875c6d1db6e22f246699691ab7e6b Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Mon, 22 May 2017 15:30:01 -0700 Subject: x86/mm: Reimplement flush_tlb_page() using flush_tlb_mm_range() From: Andy Lutomirski <luto(a)kernel.org> commit ca6c99c0794875c6d1db6e22f246699691ab7e6b upstream. flush_tlb_page() was very similar to flush_tlb_mm_range() except that it had a couple of issues: - It was missing an smp_mb() in the case where current->active_mm != mm. (This is a longstanding bug reported by Nadav Amit) - It was missing tracepoints and vm counter updates. The only reason that I can see for keeping it at as a separate function is that it could avoid a few branches that flush_tlb_mm_range() needs to decide to flush just one page. This hardly seems worthwhile. If we decide we want to get rid of those branches again, a better way would be to introduce an __flush_tlb_mm_range() helper and make both flush_tlb_page() and flush_tlb_mm_range() use it. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Acked-by: Kees Cook <keescook(a)chromium.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bpetkov(a)suse.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <nadav.amit(a)gmail.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/3cc3847cf888d8907577569b8bac3f01992ef8f9.149549206… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/tlbflush.h | 6 +++++- arch/x86/mm/tlb.c | 27 --------------------------- 2 files changed, 5 insertions(+), 28 deletions(-) --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -296,11 +296,15 @@ static inline void flush_tlb_kernel_rang flush_tlb_mm_range(vma->vm_mm, start, end, vma->vm_flags) extern void flush_tlb_all(void); -extern void flush_tlb_page(struct vm_area_struct *, unsigned long); extern void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start, unsigned long end, unsigned long vmflag); extern void flush_tlb_kernel_range(unsigned long start, unsigned long end); +static inline void flush_tlb_page(struct vm_area_struct *vma, unsigned long a) +{ + flush_tlb_mm_range(vma->vm_mm, a, a + PAGE_SIZE, VM_NONE); +} + void native_flush_tlb_others(const struct cpumask *cpumask, struct mm_struct *mm, unsigned long start, unsigned long end); --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -339,33 +339,6 @@ out: preempt_enable(); } -void flush_tlb_page(struct vm_area_struct *vma, unsigned long start) -{ - struct mm_struct *mm = vma->vm_mm; - - preempt_disable(); - - if (current->active_mm == mm) { - if (current->mm) { - /* - * Implicit full barrier (INVLPG) that synchronizes - * with switch_mm. - */ - __flush_tlb_one(start); - } else { - leave_mm(smp_processor_id()); - - /* Synchronize with switch_mm. */ - smp_mb(); - } - } - - if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids) - flush_tlb_others(mm_cpumask(mm), mm, start, start + PAGE_SIZE); - - preempt_enable(); -} - static void do_flush_tlb_all(void *info) { count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED); Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.4/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.4/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Disable PCID on 32-bit kernels" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Disable PCID on 32-bit kernels to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-disable-pcid-on-32-bit-kernels.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From cba4671af7550e008f7a7835f06df0763825bf3e Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Thu, 29 Jun 2017 08:53:19 -0700 Subject: x86/mm: Disable PCID on 32-bit kernels From: Andy Lutomirski <luto(a)kernel.org> commit cba4671af7550e008f7a7835f06df0763825bf3e upstream. 32-bit kernels on new hardware will see PCID in CPUID, but PCID can only be used in 64-bit mode. Rather than making all PCID code conditional, just disable the feature on 32-bit builds. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Reviewed-by: Nadav Amit <nadav.amit(a)gmail.com> Reviewed-by: Borislav Petkov <bp(a)suse.de> Reviewed-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Arjan van de Ven <arjan(a)linux.intel.com> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Mel Gorman <mgorman(a)suse.de> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: linux-mm(a)kvack.org Link: http://lkml.kernel.org/r/2e391769192a4d31b808410c383c6bf0734bc6ea.149875120… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/disabled-features.h | 4 +++- arch/x86/kernel/cpu/bugs.c | 8 ++++++++ 2 files changed, 11 insertions(+), 1 deletion(-) --- a/arch/x86/include/asm/disabled-features.h +++ b/arch/x86/include/asm/disabled-features.h @@ -21,11 +21,13 @@ # define DISABLE_K6_MTRR (1<<(X86_FEATURE_K6_MTRR & 31)) # define DISABLE_CYRIX_ARR (1<<(X86_FEATURE_CYRIX_ARR & 31)) # define DISABLE_CENTAUR_MCR (1<<(X86_FEATURE_CENTAUR_MCR & 31)) +# define DISABLE_PCID 0 #else # define DISABLE_VME 0 # define DISABLE_K6_MTRR 0 # define DISABLE_CYRIX_ARR 0 # define DISABLE_CENTAUR_MCR 0 +# define DISABLE_PCID (1<<(X86_FEATURE_PCID & 31)) #endif /* CONFIG_X86_64 */ /* @@ -35,7 +37,7 @@ #define DISABLED_MASK1 0 #define DISABLED_MASK2 0 #define DISABLED_MASK3 (DISABLE_CYRIX_ARR|DISABLE_CENTAUR_MCR|DISABLE_K6_MTRR) -#define DISABLED_MASK4 0 +#define DISABLED_MASK4 (DISABLE_PCID) #define DISABLED_MASK5 0 #define DISABLED_MASK6 0 #define DISABLED_MASK7 0 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -19,6 +19,14 @@ void __init check_bugs(void) { +#ifdef CONFIG_X86_32 + /* + * Regardless of whether PCID is enumerated, the SDM says + * that it can't be enabled in 32-bit mode. + */ + setup_clear_cpu_cap(X86_FEATURE_PCID); +#endif + identify_boot_cpu(); #ifndef CONFIG_SMP pr_info("CPU: "); Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-remove-the-up-asm-tlbflush.h-code-always-use-the-formerly-smp-code.patch queue-4.4/x86-mm-reimplement-flush_tlb_page-using-flush_tlb_mm_range.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch queue-4.4/x86-mm-disable-pcid-on-32-bit-kernels.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "block: don't let passthrough IO go into .make_request_fn()" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled block: don't let passthrough IO go into .make_request_fn() to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: block-don-t-let-passthrough-io-go-into-.make_request_fn.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 14cb0dc6479dc5ebc63b3a459a5d89a2f1b39fed Mon Sep 17 00:00:00 2001 From: Ming Lei <ming.lei(a)redhat.com> Date: Mon, 18 Dec 2017 15:40:43 +0800 Subject: block: don't let passthrough IO go into .make_request_fn() From: Ming Lei <ming.lei(a)redhat.com> commit 14cb0dc6479dc5ebc63b3a459a5d89a2f1b39fed upstream. Commit a8821f3f3("block: Improvements to bounce-buffer handling") tries to make sure that the bio to .make_request_fn won't exceed BIO_MAX_PAGES, but ignores that passthrough I/O can use blk_queue_bounce() too. Especially, passthrough IO may not be sector-aligned, and the check of 'sectors < bio_sectors(*bio_orig)' inside __blk_queue_bounce() may become true even though the max bvec number doesn't exceed BIO_MAX_PAGES, then cause the bio splitted, and the original passthrough bio is submited to generic_make_request(). This patch fixes this issue by checking if the bio is passthrough IO, and use bio_kmalloc() to allocate the cloned passthrough bio. Cc: NeilBrown <neilb(a)suse.com> Fixes: a8821f3f3("block: Improvements to bounce-buffer handling") Tested-by: Michele Ballabio <barra_cuda(a)katamail.com> Signed-off-by: Ming Lei <ming.lei(a)redhat.com> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- block/bounce.c | 6 ++++-- include/linux/blkdev.h | 21 +++++++++++++++++++-- 2 files changed, 23 insertions(+), 4 deletions(-) --- a/block/bounce.c +++ b/block/bounce.c @@ -200,6 +200,7 @@ static void __blk_queue_bounce(struct re unsigned i = 0; bool bounce = false; int sectors = 0; + bool passthrough = bio_is_passthrough(*bio_orig); bio_for_each_segment(from, *bio_orig, iter) { if (i++ < BIO_MAX_PAGES) @@ -210,13 +211,14 @@ static void __blk_queue_bounce(struct re if (!bounce) return; - if (sectors < bio_sectors(*bio_orig)) { + if (!passthrough && sectors < bio_sectors(*bio_orig)) { bio = bio_split(*bio_orig, sectors, GFP_NOIO, bounce_bio_split); bio_chain(bio, *bio_orig); generic_make_request(*bio_orig); *bio_orig = bio; } - bio = bio_clone_bioset(*bio_orig, GFP_NOIO, bounce_bio_set); + bio = bio_clone_bioset(*bio_orig, GFP_NOIO, passthrough ? NULL : + bounce_bio_set); bio_for_each_segment_all(to, bio, i) { struct page *page = to->bv_page; --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -241,14 +241,24 @@ struct request { struct request *next_rq; }; +static inline bool blk_op_is_scsi(unsigned int op) +{ + return op == REQ_OP_SCSI_IN || op == REQ_OP_SCSI_OUT; +} + +static inline bool blk_op_is_private(unsigned int op) +{ + return op == REQ_OP_DRV_IN || op == REQ_OP_DRV_OUT; +} + static inline bool blk_rq_is_scsi(struct request *rq) { - return req_op(rq) == REQ_OP_SCSI_IN || req_op(rq) == REQ_OP_SCSI_OUT; + return blk_op_is_scsi(req_op(rq)); } static inline bool blk_rq_is_private(struct request *rq) { - return req_op(rq) == REQ_OP_DRV_IN || req_op(rq) == REQ_OP_DRV_OUT; + return blk_op_is_private(req_op(rq)); } static inline bool blk_rq_is_passthrough(struct request *rq) @@ -256,6 +266,13 @@ static inline bool blk_rq_is_passthrough return blk_rq_is_scsi(rq) || blk_rq_is_private(rq); } +static inline bool bio_is_passthrough(struct bio *bio) +{ + unsigned op = bio_op(bio); + + return blk_op_is_scsi(op) || blk_op_is_private(op); +} + static inline unsigned short req_get_ioprio(struct request *req) { return req->ioprio; Patches currently in stable-queue which might be from ming.lei(a)redhat.com are queue-4.14/block-don-t-let-passthrough-io-go-into-.make_request_fn.patch queue-4.14/block-fix-blk_rq_append_bio.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "block: fix blk_rq_append_bio" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled block: fix blk_rq_append_bio to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: block-fix-blk_rq_append_bio.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 0abc2a10389f0c9070f76ca906c7382788036b93 Mon Sep 17 00:00:00 2001 From: Jens Axboe <axboe(a)kernel.dk> Date: Mon, 18 Dec 2017 15:40:44 +0800 Subject: block: fix blk_rq_append_bio From: Jens Axboe <axboe(a)kernel.dk> commit 0abc2a10389f0c9070f76ca906c7382788036b93 upstream. Commit caa4b02476e3(blk-map: call blk_queue_bounce from blk_rq_append_bio) moves blk_queue_bounce() into blk_rq_append_bio(), but don't consider the fact that the bounced bio becomes invisible to caller since the parameter type is 'struct bio *'. Make it a pointer to a pointer to a bio, so the caller sees the right bio also after a bounce. Fixes: caa4b02476e3 ("blk-map: call blk_queue_bounce from blk_rq_append_bio") Cc: Christoph Hellwig <hch(a)lst.de> Reported-by: Michele Ballabio <barra_cuda(a)katamail.com> (handling failure of blk_rq_append_bio(), only call bio_get() after blk_rq_append_bio() returns OK) Tested-by: Michele Ballabio <barra_cuda(a)katamail.com> Signed-off-by: Ming Lei <ming.lei(a)redhat.com> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- block/blk-map.c | 38 +++++++++++++++++++++---------------- drivers/scsi/osd/osd_initiator.c | 4 ++- drivers/target/target_core_pscsi.c | 4 +-- include/linux/blkdev.h | 2 - 4 files changed, 28 insertions(+), 20 deletions(-) --- a/block/blk-map.c +++ b/block/blk-map.c @@ -12,22 +12,29 @@ #include "blk.h" /* - * Append a bio to a passthrough request. Only works can be merged into - * the request based on the driver constraints. + * Append a bio to a passthrough request. Only works if the bio can be merged + * into the request based on the driver constraints. */ -int blk_rq_append_bio(struct request *rq, struct bio *bio) +int blk_rq_append_bio(struct request *rq, struct bio **bio) { - blk_queue_bounce(rq->q, &bio); + struct bio *orig_bio = *bio; + + blk_queue_bounce(rq->q, bio); if (!rq->bio) { - blk_rq_bio_prep(rq->q, rq, bio); + blk_rq_bio_prep(rq->q, rq, *bio); } else { - if (!ll_back_merge_fn(rq->q, rq, bio)) + if (!ll_back_merge_fn(rq->q, rq, *bio)) { + if (orig_bio != *bio) { + bio_put(*bio); + *bio = orig_bio; + } return -EINVAL; + } - rq->biotail->bi_next = bio; - rq->biotail = bio; - rq->__data_len += bio->bi_iter.bi_size; + rq->biotail->bi_next = *bio; + rq->biotail = *bio; + rq->__data_len += (*bio)->bi_iter.bi_size; } return 0; @@ -80,14 +87,12 @@ static int __blk_rq_map_user_iov(struct * We link the bounce buffer in and could have to traverse it * later so we have to get a ref to prevent it from being freed */ - ret = blk_rq_append_bio(rq, bio); - bio_get(bio); + ret = blk_rq_append_bio(rq, &bio); if (ret) { - bio_endio(bio); __blk_rq_unmap_user(orig_bio); - bio_put(bio); return ret; } + bio_get(bio); return 0; } @@ -220,7 +225,7 @@ int blk_rq_map_kern(struct request_queue int reading = rq_data_dir(rq) == READ; unsigned long addr = (unsigned long) kbuf; int do_copy = 0; - struct bio *bio; + struct bio *bio, *orig_bio; int ret; if (len > (queue_max_hw_sectors(q) << 9)) @@ -243,10 +248,11 @@ int blk_rq_map_kern(struct request_queue if (do_copy) rq->rq_flags |= RQF_COPY_USER; - ret = blk_rq_append_bio(rq, bio); + orig_bio = bio; + ret = blk_rq_append_bio(rq, &bio); if (unlikely(ret)) { /* request is too big */ - bio_put(bio); + bio_put(orig_bio); return ret; } --- a/drivers/scsi/osd/osd_initiator.c +++ b/drivers/scsi/osd/osd_initiator.c @@ -1576,7 +1576,9 @@ static struct request *_make_request(str return req; for_each_bio(bio) { - ret = blk_rq_append_bio(req, bio); + struct bio *bounce_bio = bio; + + ret = blk_rq_append_bio(req, &bounce_bio); if (ret) return ERR_PTR(ret); } --- a/drivers/target/target_core_pscsi.c +++ b/drivers/target/target_core_pscsi.c @@ -920,7 +920,7 @@ pscsi_map_sg(struct se_cmd *cmd, struct " %d i: %d bio: %p, allocating another" " bio\n", bio->bi_vcnt, i, bio); - rc = blk_rq_append_bio(req, bio); + rc = blk_rq_append_bio(req, &bio); if (rc) { pr_err("pSCSI: failed to append bio\n"); goto fail; @@ -938,7 +938,7 @@ pscsi_map_sg(struct se_cmd *cmd, struct } if (bio) { - rc = blk_rq_append_bio(req, bio); + rc = blk_rq_append_bio(req, &bio); if (rc) { pr_err("pSCSI: failed to append bio\n"); goto fail; --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -952,7 +952,7 @@ extern int blk_rq_prep_clone(struct requ extern void blk_rq_unprep_clone(struct request *rq); extern blk_status_t blk_insert_cloned_request(struct request_queue *q, struct request *rq); -extern int blk_rq_append_bio(struct request *rq, struct bio *bio); +extern int blk_rq_append_bio(struct request *rq, struct bio **bio); extern void blk_delay_queue(struct request_queue *, unsigned long); extern void blk_queue_split(struct request_queue *, struct bio **); extern void blk_recount_segments(struct request_queue *, struct bio *); Patches currently in stable-queue which might be from axboe(a)kernel.dk are queue-4.14/block-don-t-let-passthrough-io-go-into-.make_request_fn.patch queue-4.14/block-fix-blk_rq_append_bio.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] block: two patches for 4.14.x

by Michele Ballabio

Hi, there are two commits in master that should be cherry-picked for 4.14.x: commit 0abc2a10389f0c9070f76ca906c7382788036b93 block: fix blk_rq_append_bio commit 14cb0dc6479dc5ebc63b3a459a5d89a2f1b39fed block: don't let passthrough IO go into .make_request_fn() These avoid oopses, especially on x86-32. Thanks, Michele Ballabio

7 years, 8 months

2
1
0 0

[Linux-stable-mirror] Patch "x86/mm: Remove flush_tlb() and flush_tlb_current_task()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Remove flush_tlb() and flush_tlb_current_task() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 29961b59a51f8c6838a26a45e871a7ed6771809b Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sat, 22 Apr 2017 00:01:20 -0700 Subject: x86/mm: Remove flush_tlb() and flush_tlb_current_task() From: Andy Lutomirski <luto(a)kernel.org> commit 29961b59a51f8c6838a26a45e871a7ed6771809b upstream. I was trying to figure out what how flush_tlb_current_task() would possibly work correctly if current->mm != current->active_mm, but I realized I could spare myself the effort: it has no callers except the unused flush_tlb() macro. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Brian Gerst <brgerst(a)gmail.com> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Denys Vlasenko <dvlasenk(a)redhat.com> Cc: H. Peter Anvin <hpa(a)zytor.com> Cc: Josh Poimboeuf <jpoimboe(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/e52d64c11690f85e9f1d69d7b48cc2269cd2e94b.149284437… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/tlbflush.h | 9 --------- arch/x86/mm/tlb.c | 17 ----------------- 2 files changed, 26 deletions(-) --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -205,7 +205,6 @@ static inline void __flush_tlb_one(unsig /* * TLB flushing: * - * - flush_tlb() flushes the current mm struct TLBs * - flush_tlb_all() flushes all processes TLBs * - flush_tlb_mm(mm) flushes the specified mm context TLB's * - flush_tlb_page(vma, vmaddr) flushes one page @@ -237,11 +236,6 @@ static inline void flush_tlb_all(void) __flush_tlb_all(); } -static inline void flush_tlb(void) -{ - __flush_tlb_up(); -} - static inline void local_flush_tlb(void) { __flush_tlb_up(); @@ -303,14 +297,11 @@ static inline void flush_tlb_kernel_rang flush_tlb_mm_range(vma->vm_mm, start, end, vma->vm_flags) extern void flush_tlb_all(void); -extern void flush_tlb_current_task(void); extern void flush_tlb_page(struct vm_area_struct *, unsigned long); extern void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start, unsigned long end, unsigned long vmflag); extern void flush_tlb_kernel_range(unsigned long start, unsigned long end); -#define flush_tlb() flush_tlb_current_task() - void native_flush_tlb_others(const struct cpumask *cpumask, struct mm_struct *mm, unsigned long start, unsigned long end); --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -287,23 +287,6 @@ void native_flush_tlb_others(const struc smp_call_function_many(cpumask, flush_tlb_func, &info, 1); } -void flush_tlb_current_task(void) -{ - struct mm_struct *mm = current->mm; - - preempt_disable(); - - count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); - - /* This is an implicit full barrier that synchronizes with switch_mm. */ - local_flush_tlb(); - - trace_tlb_flush(TLB_LOCAL_SHOOTDOWN, TLB_FLUSH_ALL); - if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids) - flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL); - preempt_enable(); -} - /* * See Documentation/x86/tlb.txt for details. We choose 33 * because it is large enough to cover the vast majority (at Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/vm86/32: Switch to flush_tlb_mm_range() in mark_screen_rdonly()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/vm86/32: Switch to flush_tlb_mm_range() in mark_screen_rdonly() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 9ccee2373f0658f234727700e619df097ba57023 Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sat, 22 Apr 2017 00:01:19 -0700 Subject: x86/vm86/32: Switch to flush_tlb_mm_range() in mark_screen_rdonly() From: Andy Lutomirski <luto(a)kernel.org> commit 9ccee2373f0658f234727700e619df097ba57023 upstream. mark_screen_rdonly() is the last remaining caller of flush_tlb(). flush_tlb_mm_range() is potentially faster and isn't obsolete. Compile-tested only because I don't know whether software that uses this mechanism even exists. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Brian Gerst <brgerst(a)gmail.com> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Denys Vlasenko <dvlasenk(a)redhat.com> Cc: H. Peter Anvin <hpa(a)zytor.com> Cc: Josh Poimboeuf <jpoimboe(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Sasha Levin <sasha.levin(a)oracle.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/791a644076fc3577ba7f7b7cafd643cc089baa7d.149284437… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/kernel/vm86_32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -191,7 +191,7 @@ static void mark_screen_rdonly(struct mm pte_unmap_unlock(pte, ptl); out: up_write(&mm->mmap_sem); - flush_tlb(); + flush_tlb_mm_range(mm, 0xA0000, 0xA0000 + 32*PAGE_SIZE, 0UL); } Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Make flush_tlb_mm_range() more predictable" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Make flush_tlb_mm_range() more predictable to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-make-flush_tlb_mm_range-more-predictable.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ce27374fabf553153c3f53efcaa9bfab9216bd8c Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sat, 22 Apr 2017 00:01:21 -0700 Subject: x86/mm: Make flush_tlb_mm_range() more predictable From: Andy Lutomirski <luto(a)kernel.org> commit ce27374fabf553153c3f53efcaa9bfab9216bd8c upstream. I'm about to rewrite the function almost completely, but first I want to get a functional change out of the way. Currently, if flush_tlb_mm_range() does not flush the local TLB at all, it will never do individual page flushes on remote CPUs. This seems to be an accident, and preserving it will be awkward. Let's change it first so that any regressions in the rewrite will be easier to bisect and so that the rewrite can attempt to change no visible behavior at all. The fix is simple: we can simply avoid short-circuiting the calculation of base_pages_to_flush. As a side effect, this also eliminates a potential corner case: if tlb_single_page_flush_ceiling == TLB_FLUSH_ALL, flush_tlb_mm_range() could have ended up flushing the entire address space one page at a time. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Acked-by: Dave Hansen <dave.hansen(a)intel.com> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Brian Gerst <brgerst(a)gmail.com> Cc: Denys Vlasenko <dvlasenk(a)redhat.com> Cc: H. Peter Anvin <hpa(a)zytor.com> Cc: Josh Poimboeuf <jpoimboe(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/4b29b771d9975aad7154c314534fec235618175a.149284437… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/mm/tlb.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -307,6 +307,12 @@ void flush_tlb_mm_range(struct mm_struct unsigned long base_pages_to_flush = TLB_FLUSH_ALL; preempt_disable(); + + if ((end != TLB_FLUSH_ALL) && !(vmflag & VM_HUGETLB)) + base_pages_to_flush = (end - start) >> PAGE_SHIFT; + if (base_pages_to_flush > tlb_single_page_flush_ceiling) + base_pages_to_flush = TLB_FLUSH_ALL; + if (current->active_mm != mm) { /* Synchronize with switch_mm. */ smp_mb(); @@ -323,15 +329,11 @@ void flush_tlb_mm_range(struct mm_struct goto out; } - if ((end != TLB_FLUSH_ALL) && !(vmflag & VM_HUGETLB)) - base_pages_to_flush = (end - start) >> PAGE_SHIFT; - /* * Both branches below are implicit full barriers (MOV to CR or * INVLPG) that synchronize with switch_mm. */ - if (base_pages_to_flush > tlb_single_page_flush_ceiling) { - base_pages_to_flush = TLB_FLUSH_ALL; + if (base_pages_to_flush == TLB_FLUSH_ALL) { count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); local_flush_tlb(); } else { Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.9/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.9/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.9/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/vm86/32: Switch to flush_tlb_mm_range() in mark_screen_rdonly()" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/vm86/32: Switch to flush_tlb_mm_range() in mark_screen_rdonly() to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 9ccee2373f0658f234727700e619df097ba57023 Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sat, 22 Apr 2017 00:01:19 -0700 Subject: x86/vm86/32: Switch to flush_tlb_mm_range() in mark_screen_rdonly() From: Andy Lutomirski <luto(a)kernel.org> commit 9ccee2373f0658f234727700e619df097ba57023 upstream. mark_screen_rdonly() is the last remaining caller of flush_tlb(). flush_tlb_mm_range() is potentially faster and isn't obsolete. Compile-tested only because I don't know whether software that uses this mechanism even exists. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Brian Gerst <brgerst(a)gmail.com> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Denys Vlasenko <dvlasenk(a)redhat.com> Cc: H. Peter Anvin <hpa(a)zytor.com> Cc: Josh Poimboeuf <jpoimboe(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Sasha Levin <sasha.levin(a)oracle.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/791a644076fc3577ba7f7b7cafd643cc089baa7d.149284437… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/kernel/vm86_32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -187,7 +187,7 @@ static void mark_screen_rdonly(struct mm pte_unmap_unlock(pte, ptl); out: up_write(&mm->mmap_sem); - flush_tlb(); + flush_tlb_mm_range(mm, 0xA0000, 0xA0000 + 32*PAGE_SIZE, 0UL); } Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Make flush_tlb_mm_range() more predictable" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Make flush_tlb_mm_range() more predictable to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-make-flush_tlb_mm_range-more-predictable.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ce27374fabf553153c3f53efcaa9bfab9216bd8c Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sat, 22 Apr 2017 00:01:21 -0700 Subject: x86/mm: Make flush_tlb_mm_range() more predictable From: Andy Lutomirski <luto(a)kernel.org> commit ce27374fabf553153c3f53efcaa9bfab9216bd8c upstream. I'm about to rewrite the function almost completely, but first I want to get a functional change out of the way. Currently, if flush_tlb_mm_range() does not flush the local TLB at all, it will never do individual page flushes on remote CPUs. This seems to be an accident, and preserving it will be awkward. Let's change it first so that any regressions in the rewrite will be easier to bisect and so that the rewrite can attempt to change no visible behavior at all. The fix is simple: we can simply avoid short-circuiting the calculation of base_pages_to_flush. As a side effect, this also eliminates a potential corner case: if tlb_single_page_flush_ceiling == TLB_FLUSH_ALL, flush_tlb_mm_range() could have ended up flushing the entire address space one page at a time. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Acked-by: Dave Hansen <dave.hansen(a)intel.com> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Brian Gerst <brgerst(a)gmail.com> Cc: Denys Vlasenko <dvlasenk(a)redhat.com> Cc: H. Peter Anvin <hpa(a)zytor.com> Cc: Josh Poimboeuf <jpoimboe(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/4b29b771d9975aad7154c314534fec235618175a.149284437… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/mm/tlb.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -292,6 +292,12 @@ void flush_tlb_mm_range(struct mm_struct unsigned long base_pages_to_flush = TLB_FLUSH_ALL; preempt_disable(); + + if ((end != TLB_FLUSH_ALL) && !(vmflag & VM_HUGETLB)) + base_pages_to_flush = (end - start) >> PAGE_SHIFT; + if (base_pages_to_flush > tlb_single_page_flush_ceiling) + base_pages_to_flush = TLB_FLUSH_ALL; + if (current->active_mm != mm) { /* Synchronize with switch_mm. */ smp_mb(); @@ -308,15 +314,11 @@ void flush_tlb_mm_range(struct mm_struct goto out; } - if ((end != TLB_FLUSH_ALL) && !(vmflag & VM_HUGETLB)) - base_pages_to_flush = (end - start) >> PAGE_SHIFT; - /* * Both branches below are implicit full barriers (MOV to CR or * INVLPG) that synchronize with switch_mm. */ - if (base_pages_to_flush > tlb_single_page_flush_ceiling) { - base_pages_to_flush = TLB_FLUSH_ALL; + if (base_pages_to_flush == TLB_FLUSH_ALL) { count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); local_flush_tlb(); } else { Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "x86/mm: Remove flush_tlb() and flush_tlb_current_task()" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled x86/mm: Remove flush_tlb() and flush_tlb_current_task() to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 29961b59a51f8c6838a26a45e871a7ed6771809b Mon Sep 17 00:00:00 2001 From: Andy Lutomirski <luto(a)kernel.org> Date: Sat, 22 Apr 2017 00:01:20 -0700 Subject: x86/mm: Remove flush_tlb() and flush_tlb_current_task() From: Andy Lutomirski <luto(a)kernel.org> commit 29961b59a51f8c6838a26a45e871a7ed6771809b upstream. I was trying to figure out what how flush_tlb_current_task() would possibly work correctly if current->mm != current->active_mm, but I realized I could spare myself the effort: it has no callers except the unused flush_tlb() macro. Signed-off-by: Andy Lutomirski <luto(a)kernel.org> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Brian Gerst <brgerst(a)gmail.com> Cc: Dave Hansen <dave.hansen(a)intel.com> Cc: Denys Vlasenko <dvlasenk(a)redhat.com> Cc: H. Peter Anvin <hpa(a)zytor.com> Cc: Josh Poimboeuf <jpoimboe(a)redhat.com> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Nadav Amit <namit(a)vmware.com> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Rik van Riel <riel(a)redhat.com> Cc: Thomas Gleixner <tglx(a)linutronix.de> Link: http://lkml.kernel.org/r/e52d64c11690f85e9f1d69d7b48cc2269cd2e94b.149284437… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Cc: Hugh Dickins <hughd(a)google.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- arch/x86/include/asm/tlbflush.h | 9 --------- arch/x86/mm/tlb.c | 17 ----------------- 2 files changed, 26 deletions(-) --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -204,7 +204,6 @@ static inline void __flush_tlb_one(unsig /* * TLB flushing: * - * - flush_tlb() flushes the current mm struct TLBs * - flush_tlb_all() flushes all processes TLBs * - flush_tlb_mm(mm) flushes the specified mm context TLB's * - flush_tlb_page(vma, vmaddr) flushes one page @@ -236,11 +235,6 @@ static inline void flush_tlb_all(void) __flush_tlb_all(); } -static inline void flush_tlb(void) -{ - __flush_tlb_up(); -} - static inline void local_flush_tlb(void) { __flush_tlb_up(); @@ -302,14 +296,11 @@ static inline void flush_tlb_kernel_rang flush_tlb_mm_range(vma->vm_mm, start, end, vma->vm_flags) extern void flush_tlb_all(void); -extern void flush_tlb_current_task(void); extern void flush_tlb_page(struct vm_area_struct *, unsigned long); extern void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start, unsigned long end, unsigned long vmflag); extern void flush_tlb_kernel_range(unsigned long start, unsigned long end); -#define flush_tlb() flush_tlb_current_task() - void native_flush_tlb_others(const struct cpumask *cpumask, struct mm_struct *mm, unsigned long start, unsigned long end); --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -272,23 +272,6 @@ void native_flush_tlb_others(const struc smp_call_function_many(cpumask, flush_tlb_func, &info, 1); } -void flush_tlb_current_task(void) -{ - struct mm_struct *mm = current->mm; - - preempt_disable(); - - count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); - - /* This is an implicit full barrier that synchronizes with switch_mm. */ - local_flush_tlb(); - - trace_tlb_flush(TLB_LOCAL_SHOOTDOWN, TLB_FLUSH_ALL); - if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids) - flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL); - preempt_enable(); -} - /* * See Documentation/x86/tlb.txt for details. We choose 33 * because it is large enough to cover the vast majority (at Patches currently in stable-queue which might be from luto(a)kernel.org are queue-4.4/x86-vm86-32-switch-to-flush_tlb_mm_range-in-mark_screen_rdonly.patch queue-4.4/x86-mm-make-flush_tlb_mm_range-more-predictable.patch queue-4.4/x86-mm-remove-flush_tlb-and-flush_tlb_current_task.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Build failures in v3.18.y-stable-queue

by Guenter Roeck

Building arm:allmodconfig (and others) ... failed -------------- Error log: /opt/buildbot/slave/stable-queue-3.18/build/drivers/mfd/cros_ec_spi.c:26:8: error: expected '=', ',', ';', 'asm' or '__attribute__' before '->' token The file (and the patch changing it) is messed up. Guenter

7 years, 8 months

2
2
0 0

Re: [Linux-stable-mirror] stable-rc build: 0 warnings 7 failures (stable-rc/v3.18.90-13-g444eb56)

by Arnd Bergmann

On Thu, Dec 28, 2017 at 8:38 AM, Olof's autobuilder <build(a)lixom.net> wrote: > Here are the build results from automated periodic testing. > > The tree being built was stable-rc, found at: > > URL: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git > > Branch: linux-3.18.y > > Topmost commits: > 444eb56 Linux 3.18.91-rc1 > d50e6c4 net: mvneta: clear interface link status on port disable > 52e4bc8 powerpc/perf: Dereference BHRB entries safely > > Build logs (stderr only) can be found at the following link (experimental): > > http://arm-soc.lixom.net/buildlogs/stable-rc/v3.18.90-13-g444eb56/ > > > Runtime: 37m 53s > > Passed: 125 > Failed: 7 > > Warnings: 0 > > Section mismatches: 0 > > ------------------------------------------------------------------------------- > > Failed defconfigs: > arm64.allmodconfig > arm.allmodconfig > arm.exynos_defconfig > arm.multi_lpae_defconfig > arm.multi_v7_defconfig > arm.multi_v7_nosmp_defconfig > arm.tegra_defconfig > > ------------------------------------------------------------------------------- > > Errors: > > arm64.allmodconfig: > /work/build/batch/drivers/mfd/cros_ec_spi.c:26:8: error: expected '=', ',', ';', 'asm' or '__attribute__' before '->' token > > arm.allmodconfig: > /work/build/batch/drivers/mfd/cros_ec_spi.c:26:8: error: expected '=', ',', ';', 'asm' or '__attribute__' before '->' token This commit evidently got messed up during backporting: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git/… Arnd

7 years, 8 months

2
2
0 0

[Linux-stable-mirror] Patch "ring-buffer: Mask out the info bits when returning buffer page length" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ring-buffer: Mask out the info bits when returning buffer page length to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 45d8b80c2ac5d21cd1e2954431fb676bc2b1e099 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" <rostedt(a)goodmis.org> Date: Fri, 22 Dec 2017 20:32:35 -0500 Subject: ring-buffer: Mask out the info bits when returning buffer page length From: Steven Rostedt (VMware) <rostedt(a)goodmis.org> commit 45d8b80c2ac5d21cd1e2954431fb676bc2b1e099 upstream. Two info bits were added to the "commit" part of the ring buffer data page when returned to be consumed. This was to inform the user space readers that events have been missed, and that the count may be stored at the end of the page. What wasn't handled, was the splice code that actually called a function to return the length of the data in order to zero out the rest of the page before sending it up to user space. These data bits were returned with the length making the value negative, and that negative value was not checked. It was compared to PAGE_SIZE, and only used if the size was less than PAGE_SIZE. Luckily PAGE_SIZE is unsigned long which made the compare an unsigned compare, meaning the negative size value did not end up causing a large portion of memory to be randomly zeroed out. Fixes: 66a8cb95ed040 ("ring-buffer: Add place holder recording of dropped events") Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- kernel/trace/ring_buffer.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -280,6 +280,8 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_data /* Missed count stored at end */ #define RB_MISSED_STORED (1 << 30) +#define RB_MISSED_FLAGS (RB_MISSED_EVENTS|RB_MISSED_STORED) + struct buffer_data_page { u64 time_stamp; /* page time stamp */ local_t commit; /* write committed index */ @@ -331,7 +333,9 @@ static void rb_init_page(struct buffer_d */ size_t ring_buffer_page_len(void *page) { - return local_read(&((struct buffer_data_page *)page)->commit) + struct buffer_data_page *bpage = page; + + return (local_read(&bpage->commit) & ~RB_MISSED_FLAGS) + BUF_PAGE_HDR_SIZE; } Patches currently in stable-queue which might be from rostedt(a)goodmis.org are queue-4.9/tracing-fix-crash-when-it-fails-to-alloc-ring-buffer.patch queue-4.9/tracing-remove-extra-zeroing-out-of-the-ring-buffer-page.patch queue-4.9/ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch queue-4.9/tracing-fix-possible-double-free-on-failure-of-allocating-trace-buffer.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: wm_adsp: Fix validation of firmware and coeff lengths" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: wm_adsp: Fix validation of firmware and coeff lengths to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-wm_adsp-fix-validation-of-firmware-and-coeff-lengths.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 50dd2ea8ef67a1617e0c0658bcbec4b9fb03b936 Mon Sep 17 00:00:00 2001 From: Ben Hutchings <ben.hutchings(a)codethink.co.uk> Date: Fri, 8 Dec 2017 16:15:20 +0000 Subject: ASoC: wm_adsp: Fix validation of firmware and coeff lengths From: Ben Hutchings <ben.hutchings(a)codethink.co.uk> commit 50dd2ea8ef67a1617e0c0658bcbec4b9fb03b936 upstream. The checks for whether another region/block header could be present are subtracting the size from the current offset. Obviously we should instead subtract the offset from the size. The checks for whether the region/block data fit in the file are adding the data size to the current offset and header size, without checking for integer overflow. Rearrange these so that overflow is impossible. Signed-off-by: Ben Hutchings <ben.hutchings(a)codethink.co.uk> Acked-by: Charles Keepax <ckeepax(a)opensource.cirrus.com> Tested-by: Charles Keepax <ckeepax(a)opensource.cirrus.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/wm_adsp.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -1465,7 +1465,7 @@ static int wm_adsp_load(struct wm_adsp * le64_to_cpu(footer->timestamp)); while (pos < firmware->size && - pos - firmware->size > sizeof(*region)) { + sizeof(*region) < firmware->size - pos) { region = (void *)&(firmware->data[pos]); region_name = "Unknown"; reg = 0; @@ -1526,8 +1526,8 @@ static int wm_adsp_load(struct wm_adsp * regions, le32_to_cpu(region->len), offset, region_name); - if ((pos + le32_to_cpu(region->len) + sizeof(*region)) > - firmware->size) { + if (le32_to_cpu(region->len) > + firmware->size - pos - sizeof(*region)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, regions, region_name, @@ -1992,7 +1992,7 @@ static int wm_adsp_load_coeff(struct wm_ blocks = 0; while (pos < firmware->size && - pos - firmware->size > sizeof(*blk)) { + sizeof(*blk) < firmware->size - pos) { blk = (void *)(&firmware->data[pos]); type = le16_to_cpu(blk->type); @@ -2066,8 +2066,8 @@ static int wm_adsp_load_coeff(struct wm_ } if (reg) { - if ((pos + le32_to_cpu(blk->len) + sizeof(*blk)) > - firmware->size) { + if (le32_to_cpu(blk->len) > + firmware->size - pos - sizeof(*blk)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, blocks, region_name, Patches currently in stable-queue which might be from ben.hutchings(a)codethink.co.uk are queue-4.9/asoc-wm_adsp-fix-validation-of-firmware-and-coeff-lengths.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "iw_cxgb4: Only validate the MSN for successful completions" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled iw_cxgb4: Only validate the MSN for successful completions to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: iw_cxgb4-only-validate-the-msn-for-successful-completions.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From f55688c45442bc863f40ad678c638785b26cdce6 Mon Sep 17 00:00:00 2001 From: Steve Wise <swise(a)opengridcomputing.com> Date: Mon, 18 Dec 2017 13:10:00 -0800 Subject: iw_cxgb4: Only validate the MSN for successful completions From: Steve Wise <swise(a)opengridcomputing.com> commit f55688c45442bc863f40ad678c638785b26cdce6 upstream. If the RECV CQE is in error, ignore the MSN check. This was causing recvs that were flushed into the sw cq to be completed with the wrong status (BAD_MSN instead of FLUSHED). Signed-off-by: Steve Wise <swise(a)opengridcomputing.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/cxgb4/cq.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) --- a/drivers/infiniband/hw/cxgb4/cq.c +++ b/drivers/infiniband/hw/cxgb4/cq.c @@ -575,10 +575,10 @@ static int poll_cq(struct t4_wq *wq, str ret = -EAGAIN; goto skip_cqe; } - if (unlikely((CQE_WRID_MSN(hw_cqe) != (wq->rq.msn)))) { + if (unlikely(!CQE_STATUS(hw_cqe) && + CQE_WRID_MSN(hw_cqe) != wq->rq.msn)) { t4_set_wq_in_error(wq); - hw_cqe->header |= htonl(CQE_STATUS_V(T4_ERR_MSN)); - goto proc_cqe; + hw_cqe->header |= cpu_to_be32(CQE_STATUS_V(T4_ERR_MSN)); } goto proc_cqe; } Patches currently in stable-queue which might be from swise(a)opengridcomputing.com are queue-4.9/iw_cxgb4-only-validate-the-msn-for-successful-completions.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: twl4030: fix child-node lookup" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: twl4030: fix child-node lookup to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-twl4030-fix-child-node-lookup.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 15f8c5f2415bfac73f33a14bcd83422bcbfb5298 Mon Sep 17 00:00:00 2001 From: Johan Hovold <johan(a)kernel.org> Date: Mon, 13 Nov 2017 12:12:56 +0100 Subject: ASoC: twl4030: fix child-node lookup From: Johan Hovold <johan(a)kernel.org> commit 15f8c5f2415bfac73f33a14bcd83422bcbfb5298 upstream. Fix child-node lookup during probe, which ended up searching the whole device tree depth-first starting at the parent rather than just matching on its children. To make things worse, the parent codec node was also prematurely freed, while the child node was leaked. Fixes: 2d6d649a2e0f ("ASoC: twl4030: Support for DT booted kernel") Signed-off-by: Johan Hovold <johan(a)kernel.org> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/twl4030.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/sound/soc/codecs/twl4030.c +++ b/sound/soc/codecs/twl4030.c @@ -232,7 +232,7 @@ static struct twl4030_codec_data *twl403 struct twl4030_codec_data *pdata = dev_get_platdata(codec->dev); struct device_node *twl4030_codec_node = NULL; - twl4030_codec_node = of_find_node_by_name(codec->dev->parent->of_node, + twl4030_codec_node = of_get_child_by_name(codec->dev->parent->of_node, "codec"); if (!pdata && twl4030_codec_node) { @@ -241,9 +241,11 @@ static struct twl4030_codec_data *twl403 GFP_KERNEL); if (!pdata) { dev_err(codec->dev, "Can not allocate memory\n"); + of_node_put(twl4030_codec_node); return NULL; } twl4030_setup_pdata_of(pdata, twl4030_codec_node); + of_node_put(twl4030_codec_node); } return pdata; Patches currently in stable-queue which might be from johan(a)kernel.org are queue-4.9/asoc-twl4030-fix-child-node-lookup.patch queue-4.9/asoc-da7218-fix-fix-child-node-lookup.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: tlv320aic31xx: Fix GPIO1 register definition" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: tlv320aic31xx: Fix GPIO1 register definition to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-tlv320aic31xx-fix-gpio1-register-definition.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 737e0b7b67bdfe24090fab2852044bb283282fc5 Mon Sep 17 00:00:00 2001 From: "Andrew F. Davis" <afd(a)ti.com> Date: Wed, 29 Nov 2017 15:32:46 -0600 Subject: ASoC: tlv320aic31xx: Fix GPIO1 register definition From: Andrew F. Davis <afd(a)ti.com> commit 737e0b7b67bdfe24090fab2852044bb283282fc5 upstream. GPIO1 control register is number 51, fix this here. Fixes: bafcbfe429eb ("ASoC: tlv320aic31xx: Make the register values human readable") Signed-off-by: Andrew F. Davis <afd(a)ti.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/tlv320aic31xx.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/sound/soc/codecs/tlv320aic31xx.h +++ b/sound/soc/codecs/tlv320aic31xx.h @@ -115,7 +115,7 @@ struct aic31xx_pdata { /* INT2 interrupt control */ #define AIC31XX_INT2CTRL AIC31XX_REG(0, 49) /* GPIO1 control */ -#define AIC31XX_GPIO1 AIC31XX_REG(0, 50) +#define AIC31XX_GPIO1 AIC31XX_REG(0, 51) #define AIC31XX_DACPRB AIC31XX_REG(0, 60) /* ADC Instruction Set Register */ Patches currently in stable-queue which might be from afd(a)ti.com are queue-4.9/asoc-tlv320aic31xx-fix-gpio1-register-definition.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: fsl_ssi: AC'97 ops need regmap, clock and cleaning up on failure" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: fsl_ssi: AC'97 ops need regmap, clock and cleaning up on failure to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-fsl_ssi-ac-97-ops-need-regmap-clock-and-cleaning-up-on-failure.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 695b78b548d8a26288f041e907ff17758df9e1d5 Mon Sep 17 00:00:00 2001 From: "Maciej S. Szmigiero" <mail(a)maciej.szmigiero.name> Date: Mon, 20 Nov 2017 23:14:55 +0100 Subject: ASoC: fsl_ssi: AC'97 ops need regmap, clock and cleaning up on failure From: Maciej S. Szmigiero <mail(a)maciej.szmigiero.name> commit 695b78b548d8a26288f041e907ff17758df9e1d5 upstream. AC'97 ops (register read / write) need SSI regmap and clock, so they have to be set after them. We also need to set these ops back to NULL if we fail the probe. Signed-off-by: Maciej S. Szmigiero <mail(a)maciej.szmigiero.name> Acked-by: Nicolin Chen <nicoleotsuka(a)gmail.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/fsl/fsl_ssi.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) --- a/sound/soc/fsl/fsl_ssi.c +++ b/sound/soc/fsl/fsl_ssi.c @@ -1467,12 +1467,6 @@ static int fsl_ssi_probe(struct platform sizeof(fsl_ssi_ac97_dai)); fsl_ac97_data = ssi_private; - - ret = snd_soc_set_ac97_ops_of_reset(&fsl_ssi_ac97_ops, pdev); - if (ret) { - dev_err(&pdev->dev, "could not set AC'97 ops\n"); - return ret; - } } else { /* Initialize this copy of the CPU DAI driver structure */ memcpy(&ssi_private->cpu_dai_drv, &fsl_ssi_dai_template, @@ -1583,6 +1577,14 @@ static int fsl_ssi_probe(struct platform return ret; } + if (fsl_ssi_is_ac97(ssi_private)) { + ret = snd_soc_set_ac97_ops_of_reset(&fsl_ssi_ac97_ops, pdev); + if (ret) { + dev_err(&pdev->dev, "could not set AC'97 ops\n"); + goto error_ac97_ops; + } + } + ret = devm_snd_soc_register_component(&pdev->dev, &fsl_ssi_component, &ssi_private->cpu_dai_drv, 1); if (ret) { @@ -1666,6 +1668,10 @@ error_sound_card: fsl_ssi_debugfs_remove(&ssi_private->dbg_stats); error_asoc_register: + if (fsl_ssi_is_ac97(ssi_private)) + snd_soc_set_ac97_ops(NULL); + +error_ac97_ops: if (ssi_private->soc->imx) fsl_ssi_imx_clean(pdev, ssi_private); Patches currently in stable-queue which might be from mail(a)maciej.szmigiero.name are queue-4.9/asoc-fsl_ssi-ac-97-ops-need-regmap-clock-and-cleaning-up-on-failure.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: da7218: fix fix child-node lookup" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: da7218: fix fix child-node lookup to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-da7218-fix-fix-child-node-lookup.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From bc6476d6c1edcb9b97621b5131bd169aa81f27db Mon Sep 17 00:00:00 2001 From: Johan Hovold <johan(a)kernel.org> Date: Mon, 13 Nov 2017 12:12:55 +0100 Subject: ASoC: da7218: fix fix child-node lookup From: Johan Hovold <johan(a)kernel.org> commit bc6476d6c1edcb9b97621b5131bd169aa81f27db upstream. Fix child-node lookup during probe, which ended up searching the whole device tree depth-first starting at the parent rather than just matching on its children. To make things worse, the parent codec node was also prematurely freed. Fixes: 4d50934abd22 ("ASoC: da7218: Add da7218 codec driver") Signed-off-by: Johan Hovold <johan(a)kernel.org> Acked-by: Adam Thomson <Adam.Thomson.Opensource(a)diasemi.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/da7218.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/sound/soc/codecs/da7218.c +++ b/sound/soc/codecs/da7218.c @@ -2519,7 +2519,7 @@ static struct da7218_pdata *da7218_of_to } if (da7218->dev_id == DA7218_DEV_ID) { - hpldet_np = of_find_node_by_name(np, "da7218_hpldet"); + hpldet_np = of_get_child_by_name(np, "da7218_hpldet"); if (!hpldet_np) return pdata; Patches currently in stable-queue which might be from johan(a)kernel.org are queue-4.9/asoc-twl4030-fix-child-node-lookup.patch queue-4.9/asoc-da7218-fix-fix-child-node-lookup.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ALSA: hda - fix headset mic detection issue on a Dell machine" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ALSA: hda - fix headset mic detection issue on a Dell machine to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: alsa-hda-fix-headset-mic-detection-issue-on-a-dell-machine.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 285d5ddcffafa5d5e68c586f4c9eaa8b24a2897d Mon Sep 17 00:00:00 2001 From: Hui Wang <hui.wang(a)canonical.com> Date: Fri, 22 Dec 2017 11:17:45 +0800 Subject: ALSA: hda - fix headset mic detection issue on a Dell machine From: Hui Wang <hui.wang(a)canonical.com> commit 285d5ddcffafa5d5e68c586f4c9eaa8b24a2897d upstream. It has the codec alc256, and add its pin definition to pin quirk table to let it apply ALC255_FIXUP_DELL1_MIC_NO_PRESENCE. Signed-off-by: Hui Wang <hui.wang(a)canonical.com> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/pci/hda/patch_realtek.c | 5 +++++ 1 file changed, 5 insertions(+) --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -5972,6 +5972,11 @@ static const struct snd_hda_pin_quirk al {0x1b, 0x01011020}, {0x21, 0x02211010}), SND_HDA_PIN_QUIRK(0x10ec0256, 0x1028, "Dell", ALC255_FIXUP_DELL1_MIC_NO_PRESENCE, + {0x12, 0x90a60130}, + {0x14, 0x90170110}, + {0x1b, 0x01011020}, + {0x21, 0x0221101f}), + SND_HDA_PIN_QUIRK(0x10ec0256, 0x1028, "Dell", ALC255_FIXUP_DELL1_MIC_NO_PRESENCE, {0x12, 0x90a60160}, {0x14, 0x90170120}, {0x21, 0x02211030}), Patches currently in stable-queue which might be from hui.wang(a)canonical.com are queue-4.9/alsa-hda-fix-headset-mic-detection-issue-on-a-dell-machine.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ALSA: hda: Drop useless WARN_ON()" has been added to the 4.9-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ALSA: hda: Drop useless WARN_ON() to the 4.9-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: alsa-hda-drop-useless-warn_on.patch and it can be found in the queue-4.9 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From a36c2638380c0a4676647a1f553b70b20d3ebce1 Mon Sep 17 00:00:00 2001 From: Takashi Iwai <tiwai(a)suse.de> Date: Fri, 22 Dec 2017 10:45:07 +0100 Subject: ALSA: hda: Drop useless WARN_ON() From: Takashi Iwai <tiwai(a)suse.de> commit a36c2638380c0a4676647a1f553b70b20d3ebce1 upstream. Since the commit 97cc2ed27e5a ("ALSA: hda - Fix yet another i915 pointer leftover in error path") cleared hdac_acomp pointer, the WARN_ON() non-NULL check in snd_hdac_i915_register_notifier() may give a false-positive warning, as the function gets called no matter whether the component is registered or not. For fixing it, let's get rid of the spurious WARN_ON(). Fixes: 97cc2ed27e5a ("ALSA: hda - Fix yet another i915 pointer leftover in error path") Reported-by: Kouta Okamoto <kouta.okamoto(a)toshiba.co.jp> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/hda/hdac_i915.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/sound/hda/hdac_i915.c +++ b/sound/hda/hdac_i915.c @@ -319,7 +319,7 @@ static int hdac_component_master_match(s */ int snd_hdac_i915_register_notifier(const struct i915_audio_component_audio_ops *aops) { - if (WARN_ON(!hdac_acomp)) + if (!hdac_acomp) return -ENODEV; hdac_acomp->audio_ops = aops; Patches currently in stable-queue which might be from tiwai(a)suse.de are queue-4.9/alsa-hda-drop-useless-warn_on.patch queue-4.9/alsa-hda-fix-headset-mic-detection-issue-on-a-dell-machine.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ring-buffer: Mask out the info bits when returning buffer page length" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ring-buffer: Mask out the info bits when returning buffer page length to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 45d8b80c2ac5d21cd1e2954431fb676bc2b1e099 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" <rostedt(a)goodmis.org> Date: Fri, 22 Dec 2017 20:32:35 -0500 Subject: ring-buffer: Mask out the info bits when returning buffer page length From: Steven Rostedt (VMware) <rostedt(a)goodmis.org> commit 45d8b80c2ac5d21cd1e2954431fb676bc2b1e099 upstream. Two info bits were added to the "commit" part of the ring buffer data page when returned to be consumed. This was to inform the user space readers that events have been missed, and that the count may be stored at the end of the page. What wasn't handled, was the splice code that actually called a function to return the length of the data in order to zero out the rest of the page before sending it up to user space. These data bits were returned with the length making the value negative, and that negative value was not checked. It was compared to PAGE_SIZE, and only used if the size was less than PAGE_SIZE. Luckily PAGE_SIZE is unsigned long which made the compare an unsigned compare, meaning the negative size value did not end up causing a large portion of memory to be randomly zeroed out. Fixes: 66a8cb95ed040 ("ring-buffer: Add place holder recording of dropped events") Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- kernel/trace/ring_buffer.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -280,6 +280,8 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_data /* Missed count stored at end */ #define RB_MISSED_STORED (1 << 30) +#define RB_MISSED_FLAGS (RB_MISSED_EVENTS|RB_MISSED_STORED) + struct buffer_data_page { u64 time_stamp; /* page time stamp */ local_t commit; /* write committed index */ @@ -331,7 +333,9 @@ static void rb_init_page(struct buffer_d */ size_t ring_buffer_page_len(void *page) { - return local_read(&((struct buffer_data_page *)page)->commit) + struct buffer_data_page *bpage = page; + + return (local_read(&bpage->commit) & ~RB_MISSED_FLAGS) + BUF_PAGE_HDR_SIZE; } Patches currently in stable-queue which might be from rostedt(a)goodmis.org are queue-4.4/tracing-fix-crash-when-it-fails-to-alloc-ring-buffer.patch queue-4.4/tracing-remove-extra-zeroing-out-of-the-ring-buffer-page.patch queue-4.4/ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch queue-4.4/tracing-fix-possible-double-free-on-failure-of-allocating-trace-buffer.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: twl4030: fix child-node lookup" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: twl4030: fix child-node lookup to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-twl4030-fix-child-node-lookup.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 15f8c5f2415bfac73f33a14bcd83422bcbfb5298 Mon Sep 17 00:00:00 2001 From: Johan Hovold <johan(a)kernel.org> Date: Mon, 13 Nov 2017 12:12:56 +0100 Subject: ASoC: twl4030: fix child-node lookup From: Johan Hovold <johan(a)kernel.org> commit 15f8c5f2415bfac73f33a14bcd83422bcbfb5298 upstream. Fix child-node lookup during probe, which ended up searching the whole device tree depth-first starting at the parent rather than just matching on its children. To make things worse, the parent codec node was also prematurely freed, while the child node was leaked. Fixes: 2d6d649a2e0f ("ASoC: twl4030: Support for DT booted kernel") Signed-off-by: Johan Hovold <johan(a)kernel.org> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/twl4030.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/sound/soc/codecs/twl4030.c +++ b/sound/soc/codecs/twl4030.c @@ -232,7 +232,7 @@ static struct twl4030_codec_data *twl403 struct twl4030_codec_data *pdata = dev_get_platdata(codec->dev); struct device_node *twl4030_codec_node = NULL; - twl4030_codec_node = of_find_node_by_name(codec->dev->parent->of_node, + twl4030_codec_node = of_get_child_by_name(codec->dev->parent->of_node, "codec"); if (!pdata && twl4030_codec_node) { @@ -241,9 +241,11 @@ static struct twl4030_codec_data *twl403 GFP_KERNEL); if (!pdata) { dev_err(codec->dev, "Can not allocate memory\n"); + of_node_put(twl4030_codec_node); return NULL; } twl4030_setup_pdata_of(pdata, twl4030_codec_node); + of_node_put(twl4030_codec_node); } return pdata; Patches currently in stable-queue which might be from johan(a)kernel.org are queue-4.4/mfd-twl4030-audio-fix-sibling-node-lookup.patch queue-4.4/asoc-twl4030-fix-child-node-lookup.patch queue-4.4/mfd-twl6040-fix-child-node-lookup.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "iw_cxgb4: Only validate the MSN for successful completions" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled iw_cxgb4: Only validate the MSN for successful completions to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: iw_cxgb4-only-validate-the-msn-for-successful-completions.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From f55688c45442bc863f40ad678c638785b26cdce6 Mon Sep 17 00:00:00 2001 From: Steve Wise <swise(a)opengridcomputing.com> Date: Mon, 18 Dec 2017 13:10:00 -0800 Subject: iw_cxgb4: Only validate the MSN for successful completions From: Steve Wise <swise(a)opengridcomputing.com> commit f55688c45442bc863f40ad678c638785b26cdce6 upstream. If the RECV CQE is in error, ignore the MSN check. This was causing recvs that were flushed into the sw cq to be completed with the wrong status (BAD_MSN instead of FLUSHED). Signed-off-by: Steve Wise <swise(a)opengridcomputing.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/cxgb4/cq.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) --- a/drivers/infiniband/hw/cxgb4/cq.c +++ b/drivers/infiniband/hw/cxgb4/cq.c @@ -579,10 +579,10 @@ static int poll_cq(struct t4_wq *wq, str ret = -EAGAIN; goto skip_cqe; } - if (unlikely((CQE_WRID_MSN(hw_cqe) != (wq->rq.msn)))) { + if (unlikely(!CQE_STATUS(hw_cqe) && + CQE_WRID_MSN(hw_cqe) != wq->rq.msn)) { t4_set_wq_in_error(wq); - hw_cqe->header |= htonl(CQE_STATUS_V(T4_ERR_MSN)); - goto proc_cqe; + hw_cqe->header |= cpu_to_be32(CQE_STATUS_V(T4_ERR_MSN)); } goto proc_cqe; } Patches currently in stable-queue which might be from swise(a)opengridcomputing.com are queue-4.4/iw_cxgb4-only-validate-the-msn-for-successful-completions.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: fsl_ssi: AC'97 ops need regmap, clock and cleaning up on failure" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: fsl_ssi: AC'97 ops need regmap, clock and cleaning up on failure to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-fsl_ssi-ac-97-ops-need-regmap-clock-and-cleaning-up-on-failure.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 695b78b548d8a26288f041e907ff17758df9e1d5 Mon Sep 17 00:00:00 2001 From: "Maciej S. Szmigiero" <mail(a)maciej.szmigiero.name> Date: Mon, 20 Nov 2017 23:14:55 +0100 Subject: ASoC: fsl_ssi: AC'97 ops need regmap, clock and cleaning up on failure From: Maciej S. Szmigiero <mail(a)maciej.szmigiero.name> commit 695b78b548d8a26288f041e907ff17758df9e1d5 upstream. AC'97 ops (register read / write) need SSI regmap and clock, so they have to be set after them. We also need to set these ops back to NULL if we fail the probe. Signed-off-by: Maciej S. Szmigiero <mail(a)maciej.szmigiero.name> Acked-by: Nicolin Chen <nicoleotsuka(a)gmail.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/fsl/fsl_ssi.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) --- a/sound/soc/fsl/fsl_ssi.c +++ b/sound/soc/fsl/fsl_ssi.c @@ -1408,12 +1408,6 @@ static int fsl_ssi_probe(struct platform sizeof(fsl_ssi_ac97_dai)); fsl_ac97_data = ssi_private; - - ret = snd_soc_set_ac97_ops_of_reset(&fsl_ssi_ac97_ops, pdev); - if (ret) { - dev_err(&pdev->dev, "could not set AC'97 ops\n"); - return ret; - } } else { /* Initialize this copy of the CPU DAI driver structure */ memcpy(&ssi_private->cpu_dai_drv, &fsl_ssi_dai_template, @@ -1473,6 +1467,14 @@ static int fsl_ssi_probe(struct platform return ret; } + if (fsl_ssi_is_ac97(ssi_private)) { + ret = snd_soc_set_ac97_ops_of_reset(&fsl_ssi_ac97_ops, pdev); + if (ret) { + dev_err(&pdev->dev, "could not set AC'97 ops\n"); + goto error_ac97_ops; + } + } + ret = devm_snd_soc_register_component(&pdev->dev, &fsl_ssi_component, &ssi_private->cpu_dai_drv, 1); if (ret) { @@ -1556,6 +1558,10 @@ error_sound_card: fsl_ssi_debugfs_remove(&ssi_private->dbg_stats); error_asoc_register: + if (fsl_ssi_is_ac97(ssi_private)) + snd_soc_set_ac97_ops(NULL); + +error_ac97_ops: if (ssi_private->soc->imx) fsl_ssi_imx_clean(pdev, ssi_private); Patches currently in stable-queue which might be from mail(a)maciej.szmigiero.name are queue-4.4/asoc-fsl_ssi-ac-97-ops-need-regmap-clock-and-cleaning-up-on-failure.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ALSA: hda: Drop useless WARN_ON()" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ALSA: hda: Drop useless WARN_ON() to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: alsa-hda-drop-useless-warn_on.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From a36c2638380c0a4676647a1f553b70b20d3ebce1 Mon Sep 17 00:00:00 2001 From: Takashi Iwai <tiwai(a)suse.de> Date: Fri, 22 Dec 2017 10:45:07 +0100 Subject: ALSA: hda: Drop useless WARN_ON() From: Takashi Iwai <tiwai(a)suse.de> commit a36c2638380c0a4676647a1f553b70b20d3ebce1 upstream. Since the commit 97cc2ed27e5a ("ALSA: hda - Fix yet another i915 pointer leftover in error path") cleared hdac_acomp pointer, the WARN_ON() non-NULL check in snd_hdac_i915_register_notifier() may give a false-positive warning, as the function gets called no matter whether the component is registered or not. For fixing it, let's get rid of the spurious WARN_ON(). Fixes: 97cc2ed27e5a ("ALSA: hda - Fix yet another i915 pointer leftover in error path") Reported-by: Kouta Okamoto <kouta.okamoto(a)toshiba.co.jp> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/hda/hdac_i915.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/sound/hda/hdac_i915.c +++ b/sound/hda/hdac_i915.c @@ -183,7 +183,7 @@ static int hdac_component_master_match(s */ int snd_hdac_i915_register_notifier(const struct i915_audio_component_audio_ops *aops) { - if (WARN_ON(!hdac_acomp)) + if (!hdac_acomp) return -ENODEV; hdac_acomp->audio_ops = aops; Patches currently in stable-queue which might be from tiwai(a)suse.de are queue-4.4/alsa-usb-audio-fix-the-missing-ctl-name-suffix-at-parsing-su.patch queue-4.4/alsa-hda-drop-useless-warn_on.patch queue-4.4/alsa-rawmidi-avoid-racy-info-ioctl-via-ctl-device.patch queue-4.4/alsa-hda-fix-headset-mic-detection-issue-on-a-dell-machine.patch queue-4.4/acpi-apei-erst-fix-missing-error-handling-in-erst_reader.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ALSA: hda - fix headset mic detection issue on a Dell machine" has been added to the 4.4-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ALSA: hda - fix headset mic detection issue on a Dell machine to the 4.4-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: alsa-hda-fix-headset-mic-detection-issue-on-a-dell-machine.patch and it can be found in the queue-4.4 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 285d5ddcffafa5d5e68c586f4c9eaa8b24a2897d Mon Sep 17 00:00:00 2001 From: Hui Wang <hui.wang(a)canonical.com> Date: Fri, 22 Dec 2017 11:17:45 +0800 Subject: ALSA: hda - fix headset mic detection issue on a Dell machine From: Hui Wang <hui.wang(a)canonical.com> commit 285d5ddcffafa5d5e68c586f4c9eaa8b24a2897d upstream. It has the codec alc256, and add its pin definition to pin quirk table to let it apply ALC255_FIXUP_DELL1_MIC_NO_PRESENCE. Signed-off-by: Hui Wang <hui.wang(a)canonical.com> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/pci/hda/patch_realtek.c | 5 +++++ 1 file changed, 5 insertions(+) --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -5954,6 +5954,11 @@ static const struct snd_hda_pin_quirk al {0x1b, 0x01011020}, {0x21, 0x02211010}), SND_HDA_PIN_QUIRK(0x10ec0256, 0x1028, "Dell", ALC255_FIXUP_DELL1_MIC_NO_PRESENCE, + {0x12, 0x90a60130}, + {0x14, 0x90170110}, + {0x1b, 0x01011020}, + {0x21, 0x0221101f}), + SND_HDA_PIN_QUIRK(0x10ec0256, 0x1028, "Dell", ALC255_FIXUP_DELL1_MIC_NO_PRESENCE, {0x12, 0x90a60160}, {0x14, 0x90170120}, {0x21, 0x02211030}), Patches currently in stable-queue which might be from hui.wang(a)canonical.com are queue-4.4/alsa-hda-fix-headset-mic-detection-issue-on-a-dell-machine.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ring-buffer: Mask out the info bits when returning buffer page length" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ring-buffer: Mask out the info bits when returning buffer page length to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 45d8b80c2ac5d21cd1e2954431fb676bc2b1e099 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" <rostedt(a)goodmis.org> Date: Fri, 22 Dec 2017 20:32:35 -0500 Subject: ring-buffer: Mask out the info bits when returning buffer page length From: Steven Rostedt (VMware) <rostedt(a)goodmis.org> commit 45d8b80c2ac5d21cd1e2954431fb676bc2b1e099 upstream. Two info bits were added to the "commit" part of the ring buffer data page when returned to be consumed. This was to inform the user space readers that events have been missed, and that the count may be stored at the end of the page. What wasn't handled, was the splice code that actually called a function to return the length of the data in order to zero out the rest of the page before sending it up to user space. These data bits were returned with the length making the value negative, and that negative value was not checked. It was compared to PAGE_SIZE, and only used if the size was less than PAGE_SIZE. Luckily PAGE_SIZE is unsigned long which made the compare an unsigned compare, meaning the negative size value did not end up causing a large portion of memory to be randomly zeroed out. Fixes: 66a8cb95ed040 ("ring-buffer: Add place holder recording of dropped events") Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- kernel/trace/ring_buffer.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -281,6 +281,8 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_data /* Missed count stored at end */ #define RB_MISSED_STORED (1 << 30) +#define RB_MISSED_FLAGS (RB_MISSED_EVENTS|RB_MISSED_STORED) + struct buffer_data_page { u64 time_stamp; /* page time stamp */ local_t commit; /* write committed index */ @@ -332,7 +334,9 @@ static void rb_init_page(struct buffer_d */ size_t ring_buffer_page_len(void *page) { - return local_read(&((struct buffer_data_page *)page)->commit) + struct buffer_data_page *bpage = page; + + return (local_read(&bpage->commit) & ~RB_MISSED_FLAGS) + BUF_PAGE_HDR_SIZE; } Patches currently in stable-queue which might be from rostedt(a)goodmis.org are queue-4.14/ring-buffer-do-no-reuse-reader-page-if-still-in-use.patch queue-4.14/tracing-fix-crash-when-it-fails-to-alloc-ring-buffer.patch queue-4.14/tracing-remove-extra-zeroing-out-of-the-ring-buffer-page.patch queue-4.14/ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch queue-4.14/tracing-fix-possible-double-free-on-failure-of-allocating-trace-buffer.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "iw_cxgb4: Only validate the MSN for successful completions" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled iw_cxgb4: Only validate the MSN for successful completions to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: iw_cxgb4-only-validate-the-msn-for-successful-completions.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From f55688c45442bc863f40ad678c638785b26cdce6 Mon Sep 17 00:00:00 2001 From: Steve Wise <swise(a)opengridcomputing.com> Date: Mon, 18 Dec 2017 13:10:00 -0800 Subject: iw_cxgb4: Only validate the MSN for successful completions From: Steve Wise <swise(a)opengridcomputing.com> commit f55688c45442bc863f40ad678c638785b26cdce6 upstream. If the RECV CQE is in error, ignore the MSN check. This was causing recvs that were flushed into the sw cq to be completed with the wrong status (BAD_MSN instead of FLUSHED). Signed-off-by: Steve Wise <swise(a)opengridcomputing.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/cxgb4/cq.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) --- a/drivers/infiniband/hw/cxgb4/cq.c +++ b/drivers/infiniband/hw/cxgb4/cq.c @@ -586,10 +586,10 @@ static int poll_cq(struct t4_wq *wq, str ret = -EAGAIN; goto skip_cqe; } - if (unlikely((CQE_WRID_MSN(hw_cqe) != (wq->rq.msn)))) { + if (unlikely(!CQE_STATUS(hw_cqe) && + CQE_WRID_MSN(hw_cqe) != wq->rq.msn)) { t4_set_wq_in_error(wq); - hw_cqe->header |= htonl(CQE_STATUS_V(T4_ERR_MSN)); - goto proc_cqe; + hw_cqe->header |= cpu_to_be32(CQE_STATUS_V(T4_ERR_MSN)); } goto proc_cqe; } Patches currently in stable-queue which might be from swise(a)opengridcomputing.com are queue-4.14/iw_cxgb4-only-validate-the-msn-for-successful-completions.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ring-buffer: Do no reuse reader page if still in use" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ring-buffer: Do no reuse reader page if still in use to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ring-buffer-do-no-reuse-reader-page-if-still-in-use.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ae415fa4c5248a8cf4faabd5a3c20576cb1ad607 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" <rostedt(a)goodmis.org> Date: Fri, 22 Dec 2017 21:19:29 -0500 Subject: ring-buffer: Do no reuse reader page if still in use From: Steven Rostedt (VMware) <rostedt(a)goodmis.org> commit ae415fa4c5248a8cf4faabd5a3c20576cb1ad607 upstream. To free the reader page that is allocated with ring_buffer_alloc_read_page(), ring_buffer_free_read_page() must be called. For faster performance, this page can be reused by the ring buffer to avoid having to free and allocate new pages. The issue arises when the page is used with a splice pipe into the networking code. The networking code may up the page counter for the page, and keep it active while sending it is queued to go to the network. The incrementing of the page ref does not prevent it from being reused in the ring buffer, and this can cause the page that is being sent out to the network to be modified before it is sent by reading new data. Add a check to the page ref counter, and only reuse the page if it is not being used anywhere else. Fixes: 73a757e63114d ("ring-buffer: Return reader page back into existing ring buffer") Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- kernel/trace/ring_buffer.c | 6 ++++++ 1 file changed, 6 insertions(+) --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -4443,8 +4443,13 @@ void ring_buffer_free_read_page(struct r { struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu]; struct buffer_data_page *bpage = data; + struct page *page = virt_to_page(bpage); unsigned long flags; + /* If the page is still in use someplace else, we can't reuse it */ + if (page_ref_count(page) > 1) + goto out; + local_irq_save(flags); arch_spin_lock(&cpu_buffer->lock); @@ -4456,6 +4461,7 @@ void ring_buffer_free_read_page(struct r arch_spin_unlock(&cpu_buffer->lock); local_irq_restore(flags); + out: free_page((unsigned long)bpage); } EXPORT_SYMBOL_GPL(ring_buffer_free_read_page); Patches currently in stable-queue which might be from rostedt(a)goodmis.org are queue-4.14/ring-buffer-do-no-reuse-reader-page-if-still-in-use.patch queue-4.14/tracing-fix-crash-when-it-fails-to-alloc-ring-buffer.patch queue-4.14/tracing-remove-extra-zeroing-out-of-the-ring-buffer-page.patch queue-4.14/ring-buffer-mask-out-the-info-bits-when-returning-buffer-page-length.patch queue-4.14/tracing-fix-possible-double-free-on-failure-of-allocating-trace-buffer.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "IB/mlx5: Serialize access to the VMA list" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled IB/mlx5: Serialize access to the VMA list to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ib-mlx5-serialize-access-to-the-vma-list.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From ad9a3668a434faca1339789ed2f043d679199309 Mon Sep 17 00:00:00 2001 From: Majd Dibbiny <majd(a)mellanox.com> Date: Sun, 24 Dec 2017 13:54:56 +0200 Subject: IB/mlx5: Serialize access to the VMA list From: Majd Dibbiny <majd(a)mellanox.com> commit ad9a3668a434faca1339789ed2f043d679199309 upstream. User-space applications can do mmap and munmap directly at any time. Since the VMA list is not protected with a mutex, concurrent accesses to the VMA list from the mmap and munmap can cause data corruption. Add a mutex around the list. Fixes: 7c2344c3bbf9 ("IB/mlx5: Implements disassociate_ucontext API") Reviewed-by: Yishai Hadas <yishaih(a)mellanox.com> Signed-off-by: Majd Dibbiny <majd(a)mellanox.com> Signed-off-by: Leon Romanovsky <leon(a)kernel.org> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/mlx5/main.c | 8 ++++++++ drivers/infiniband/hw/mlx5/mlx5_ib.h | 4 ++++ 2 files changed, 12 insertions(+) --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -1415,6 +1415,7 @@ static struct ib_ucontext *mlx5_ib_alloc } INIT_LIST_HEAD(&context->vma_private_list); + mutex_init(&context->vma_private_list_mutex); INIT_LIST_HEAD(&context->db_page_list); mutex_init(&context->db_page_mutex); @@ -1576,7 +1577,9 @@ static void mlx5_ib_vma_close(struct vm * mlx5_ib_disassociate_ucontext(). */ mlx5_ib_vma_priv_data->vma = NULL; + mutex_lock(mlx5_ib_vma_priv_data->vma_private_list_mutex); list_del(&mlx5_ib_vma_priv_data->list); + mutex_unlock(mlx5_ib_vma_priv_data->vma_private_list_mutex); kfree(mlx5_ib_vma_priv_data); } @@ -1596,10 +1599,13 @@ static int mlx5_ib_set_vma_data(struct v return -ENOMEM; vma_prv->vma = vma; + vma_prv->vma_private_list_mutex = &ctx->vma_private_list_mutex; vma->vm_private_data = vma_prv; vma->vm_ops = &mlx5_ib_vm_ops; + mutex_lock(&ctx->vma_private_list_mutex); list_add(&vma_prv->list, vma_head); + mutex_unlock(&ctx->vma_private_list_mutex); return 0; } @@ -1642,6 +1648,7 @@ static void mlx5_ib_disassociate_ucontex * mlx5_ib_vma_close. */ down_write(&owning_mm->mmap_sem); + mutex_lock(&context->vma_private_list_mutex); list_for_each_entry_safe(vma_private, n, &context->vma_private_list, list) { vma = vma_private->vma; @@ -1656,6 +1663,7 @@ static void mlx5_ib_disassociate_ucontex list_del(&vma_private->list); kfree(vma_private); } + mutex_unlock(&context->vma_private_list_mutex); up_write(&owning_mm->mmap_sem); mmput(owning_mm); put_task_struct(owning_process); --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -115,6 +115,8 @@ enum { struct mlx5_ib_vma_private_data { struct list_head list; struct vm_area_struct *vma; + /* protect vma_private_list add/del */ + struct mutex *vma_private_list_mutex; }; struct mlx5_ib_ucontext { @@ -129,6 +131,8 @@ struct mlx5_ib_ucontext { /* Transport Domain number */ u32 tdn; struct list_head vma_private_list; + /* protect vma_private_list add/del */ + struct mutex vma_private_list_mutex; unsigned long upd_xlt_page; /* protect ODP/KSM */ Patches currently in stable-queue which might be from majd(a)mellanox.com are queue-4.14/ib-mlx5-serialize-access-to-the-vma-list.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "IB/uverbs: Fix command checking as part of ib_uverbs_ex_modify_qp()" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled IB/uverbs: Fix command checking as part of ib_uverbs_ex_modify_qp() to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ib-uverbs-fix-command-checking-as-part-of-ib_uverbs_ex_modify_qp.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 05d14e7b0c138cb07ba30e464f47b39434f3fdef Mon Sep 17 00:00:00 2001 From: Moni Shoua <monis(a)mellanox.com> Date: Sun, 24 Dec 2017 13:54:57 +0200 Subject: IB/uverbs: Fix command checking as part of ib_uverbs_ex_modify_qp() From: Moni Shoua <monis(a)mellanox.com> commit 05d14e7b0c138cb07ba30e464f47b39434f3fdef upstream. If the input command length is larger than the kernel supports an error should be returned in case the unsupported bytes are not cleared, instead of the other way aroudn. This matches what all other callers of ib_is_udata_cleared do and will avoid user ABI problems in the future. Fixes: 189aba99e700 ("IB/uverbs: Extend modify_qp and support packet pacing") Reviewed-by: Yishai Hadas <yishaih(a)mellanox.com> Signed-off-by: Moni Shoua <monis(a)mellanox.com> Signed-off-by: Leon Romanovsky <leon(a)kernel.org> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/core/uverbs_cmd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -2085,8 +2085,8 @@ int ib_uverbs_ex_modify_qp(struct ib_uve return -EOPNOTSUPP; if (ucore->inlen > sizeof(cmd)) { - if (ib_is_udata_cleared(ucore, sizeof(cmd), - ucore->inlen - sizeof(cmd))) + if (!ib_is_udata_cleared(ucore, sizeof(cmd), + ucore->inlen - sizeof(cmd))) return -EOPNOTSUPP; } Patches currently in stable-queue which might be from monis(a)mellanox.com are queue-4.14/ib-core-verify-that-qp-is-security-enabled-in-create-and-destroy.patch queue-4.14/ib-uverbs-fix-command-checking-as-part-of-ib_uverbs_ex_modify_qp.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "IB/core: Verify that QP is security enabled in create and destroy" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled IB/core: Verify that QP is security enabled in create and destroy to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ib-core-verify-that-qp-is-security-enabled-in-create-and-destroy.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 4a50881bbac309e6f0684816a180bc3c14e1485d Mon Sep 17 00:00:00 2001 From: Moni Shoua <monis(a)mellanox.com> Date: Sun, 24 Dec 2017 13:54:58 +0200 Subject: IB/core: Verify that QP is security enabled in create and destroy From: Moni Shoua <monis(a)mellanox.com> commit 4a50881bbac309e6f0684816a180bc3c14e1485d upstream. The XRC target QP create flow sets up qp_sec only if there is an IB link with LSM security enabled. However, several other related uAPI entry points blindly follow the qp_sec NULL pointer, resulting in a possible oops. Check for NULL before using qp_sec. Fixes: d291f1a65232 ("IB/core: Enforce PKey security on QPs") Reviewed-by: Daniel Jurgens <danielj(a)mellanox.com> Signed-off-by: Moni Shoua <monis(a)mellanox.com> Signed-off-by: Leon Romanovsky <leon(a)kernel.org> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/core/security.c | 3 +++ drivers/infiniband/core/verbs.c | 3 ++- 2 files changed, 5 insertions(+), 1 deletion(-) --- a/drivers/infiniband/core/security.c +++ b/drivers/infiniband/core/security.c @@ -386,6 +386,9 @@ int ib_open_shared_qp_security(struct ib if (ret) return ret; + if (!qp->qp_sec) + return 0; + mutex_lock(&real_qp->qp_sec->mutex); ret = check_qp_port_pkey_settings(real_qp->qp_sec->ports_pkeys, qp->qp_sec); --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -1400,7 +1400,8 @@ int ib_close_qp(struct ib_qp *qp) spin_unlock_irqrestore(&real_qp->device->event_handler_lock, flags); atomic_dec(&real_qp->usecnt); - ib_close_shared_qp_security(qp->qp_sec); + if (qp->qp_sec) + ib_close_shared_qp_security(qp->qp_sec); kfree(qp); return 0; Patches currently in stable-queue which might be from monis(a)mellanox.com are queue-4.14/ib-core-verify-that-qp-is-security-enabled-in-create-and-destroy.patch queue-4.14/ib-uverbs-fix-command-checking-as-part-of-ib_uverbs_ex_modify_qp.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "IB/hfi: Only read capability registers if the capability exists" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled IB/hfi: Only read capability registers if the capability exists to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: ib-hfi-only-read-capability-registers-if-the-capability-exists.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 4c009af473b2026caaa26107e34d7cc68dad7756 Mon Sep 17 00:00:00 2001 From: "Michael J. Ruhl" <michael.j.ruhl(a)intel.com> Date: Fri, 22 Dec 2017 08:47:20 -0800 Subject: IB/hfi: Only read capability registers if the capability exists From: Michael J. Ruhl <michael.j.ruhl(a)intel.com> commit 4c009af473b2026caaa26107e34d7cc68dad7756 upstream. During driver init, various registers are saved to allow restoration after an FLR or gen3 bump. Some of these registers are not available in some circumstances (i.e. Virtual machines). This bug makes the driver unusable when the PCI device is passed into a VM, it fails during probe. Delete unnecessary register read/write, and only access register if the capability exists. Fixes: a618b7e40af2 ("IB/hfi1: Move saving PCI values to a separate function") Reviewed-by: Mike Marciniszyn <mike.marciniszyn(a)intel.com> Signed-off-by: Michael J. Ruhl <michael.j.ruhl(a)intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro(a)intel.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/hfi1/hfi.h | 1 - drivers/infiniband/hw/hfi1/pcie.c | 30 ++++++++++++------------------ 2 files changed, 12 insertions(+), 19 deletions(-) --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1129,7 +1129,6 @@ struct hfi1_devdata { u16 pcie_lnkctl; u16 pcie_devctl2; u32 pci_msix0; - u32 pci_lnkctl3; u32 pci_tph2; /* --- a/drivers/infiniband/hw/hfi1/pcie.c +++ b/drivers/infiniband/hw/hfi1/pcie.c @@ -411,15 +411,12 @@ int restore_pci_variables(struct hfi1_de if (ret) goto error; - ret = pci_write_config_dword(dd->pcidev, PCIE_CFG_SPCIE1, - dd->pci_lnkctl3); - if (ret) - goto error; - - ret = pci_write_config_dword(dd->pcidev, PCIE_CFG_TPH2, dd->pci_tph2); - if (ret) - goto error; - + if (pci_find_ext_capability(dd->pcidev, PCI_EXT_CAP_ID_TPH)) { + ret = pci_write_config_dword(dd->pcidev, PCIE_CFG_TPH2, + dd->pci_tph2); + if (ret) + goto error; + } return 0; error: @@ -469,15 +466,12 @@ int save_pci_variables(struct hfi1_devda if (ret) goto error; - ret = pci_read_config_dword(dd->pcidev, PCIE_CFG_SPCIE1, - &dd->pci_lnkctl3); - if (ret) - goto error; - - ret = pci_read_config_dword(dd->pcidev, PCIE_CFG_TPH2, &dd->pci_tph2); - if (ret) - goto error; - + if (pci_find_ext_capability(dd->pcidev, PCI_EXT_CAP_ID_TPH)) { + ret = pci_read_config_dword(dd->pcidev, PCIE_CFG_TPH2, + &dd->pci_tph2); + if (ret) + goto error; + } return 0; error: Patches currently in stable-queue which might be from michael.j.ruhl(a)intel.com are queue-4.14/ib-hfi-only-read-capability-registers-if-the-capability-exists.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "gpio: fix "gpio-line-names" property retrieval" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled gpio: fix "gpio-line-names" property retrieval to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: gpio-fix-gpio-line-names-property-retrieval.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 822703354774ec935169cbbc8d503236bcb54fda Mon Sep 17 00:00:00 2001 From: Christophe Leroy <christophe.leroy(a)c-s.fr> Date: Fri, 15 Dec 2017 15:02:33 +0100 Subject: gpio: fix "gpio-line-names" property retrieval From: Christophe Leroy <christophe.leroy(a)c-s.fr> commit 822703354774ec935169cbbc8d503236bcb54fda upstream. Following commit 9427ecbed46cc ("gpio: Rework of_gpiochip_set_names() to use device property accessors"), "gpio-line-names" DT property is not retrieved anymore when chip->parent is not set by the driver. This is due to OF based property reads having been replaced by device based property reads. This patch fixes that by making use of fwnode_property_read_string_array() instead of device_property_read_string_array() and handing over either of_fwnode_handle(chip->of_node) or dev_fwnode(chip->parent) to that function. Fixes: 9427ecbed46cc ("gpio: Rework of_gpiochip_set_names() to use device property accessors") Signed-off-by: Christophe Leroy <christophe.leroy(a)c-s.fr> Acked-by: Mika Westerberg <mika.westerberg(a)linux.intel.com> Signed-off-by: Linus Walleij <linus.walleij(a)linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpio/gpiolib-acpi.c | 2 +- drivers/gpio/gpiolib-devprop.c | 17 +++++++---------- drivers/gpio/gpiolib-of.c | 3 ++- drivers/gpio/gpiolib.h | 3 ++- 4 files changed, 12 insertions(+), 13 deletions(-) --- a/drivers/gpio/gpiolib-acpi.c +++ b/drivers/gpio/gpiolib-acpi.c @@ -1074,7 +1074,7 @@ void acpi_gpiochip_add(struct gpio_chip } if (!chip->names) - devprop_gpiochip_set_names(chip); + devprop_gpiochip_set_names(chip, dev_fwnode(chip->parent)); acpi_gpiochip_request_regions(acpi_gpio); acpi_gpiochip_scan_gpios(acpi_gpio); --- a/drivers/gpio/gpiolib-devprop.c +++ b/drivers/gpio/gpiolib-devprop.c @@ -19,30 +19,27 @@ /** * devprop_gpiochip_set_names - Set GPIO line names using device properties * @chip: GPIO chip whose lines should be named, if possible + * @fwnode: Property Node containing the gpio-line-names property * * Looks for device property "gpio-line-names" and if it exists assigns * GPIO line names for the chip. The memory allocated for the assigned * names belong to the underlying firmware node and should not be released * by the caller. */ -void devprop_gpiochip_set_names(struct gpio_chip *chip) +void devprop_gpiochip_set_names(struct gpio_chip *chip, + const struct fwnode_handle *fwnode) { struct gpio_device *gdev = chip->gpiodev; const char **names; int ret, i; - if (!chip->parent) { - dev_warn(&gdev->dev, "GPIO chip parent is NULL\n"); - return; - } - - ret = device_property_read_string_array(chip->parent, "gpio-line-names", + ret = fwnode_property_read_string_array(fwnode, "gpio-line-names", NULL, 0); if (ret < 0) return; if (ret != gdev->ngpio) { - dev_warn(chip->parent, + dev_warn(&gdev->dev, "names %d do not match number of GPIOs %d\n", ret, gdev->ngpio); return; @@ -52,10 +49,10 @@ void devprop_gpiochip_set_names(struct g if (!names) return; - ret = device_property_read_string_array(chip->parent, "gpio-line-names", + ret = fwnode_property_read_string_array(fwnode, "gpio-line-names", names, gdev->ngpio); if (ret < 0) { - dev_warn(chip->parent, "failed to read GPIO line names\n"); + dev_warn(&gdev->dev, "failed to read GPIO line names\n"); kfree(names); return; } --- a/drivers/gpio/gpiolib-of.c +++ b/drivers/gpio/gpiolib-of.c @@ -493,7 +493,8 @@ int of_gpiochip_add(struct gpio_chip *ch /* If the chip defines names itself, these take precedence */ if (!chip->names) - devprop_gpiochip_set_names(chip); + devprop_gpiochip_set_names(chip, + of_fwnode_handle(chip->of_node)); of_node_get(chip->of_node); --- a/drivers/gpio/gpiolib.h +++ b/drivers/gpio/gpiolib.h @@ -224,7 +224,8 @@ static inline int gpio_chip_hwgpio(const return desc - &desc->gdev->descs[0]; } -void devprop_gpiochip_set_names(struct gpio_chip *chip); +void devprop_gpiochip_set_names(struct gpio_chip *chip, + const struct fwnode_handle *fwnode); /* With descriptor prefix */ Patches currently in stable-queue which might be from christophe.leroy(a)c-s.fr are queue-4.14/gpio-fix-gpio-line-names-property-retrieval.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "cpufreq: schedutil: Use idle_calls counter of the remote CPU" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled cpufreq: schedutil: Use idle_calls counter of the remote CPU to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: cpufreq-schedutil-use-idle_calls-counter-of-the-remote-cpu.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 466a2b42d67644447a1765276259a3ea5531ddff Mon Sep 17 00:00:00 2001 From: Joel Fernandes <joelaf(a)google.com> Date: Thu, 21 Dec 2017 02:22:45 +0100 Subject: cpufreq: schedutil: Use idle_calls counter of the remote CPU From: Joel Fernandes <joelaf(a)google.com> commit 466a2b42d67644447a1765276259a3ea5531ddff upstream. Since the recent remote cpufreq callback work, its possible that a cpufreq update is triggered from a remote CPU. For single policies however, the current code uses the local CPU when trying to determine if the remote sg_cpu entered idle or is busy. This is incorrect. To remedy this, compare with the nohz tick idle_calls counter of the remote CPU. Fixes: 674e75411fc2 (sched: cpufreq: Allow remote cpufreq callbacks) Acked-by: Viresh Kumar <viresh.kumar(a)linaro.org> Acked-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Signed-off-by: Joel Fernandes <joelaf(a)google.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- include/linux/tick.h | 1 + kernel/sched/cpufreq_schedutil.c | 2 +- kernel/time/tick-sched.c | 13 +++++++++++++ 3 files changed, 15 insertions(+), 1 deletion(-) --- a/include/linux/tick.h +++ b/include/linux/tick.h @@ -119,6 +119,7 @@ extern void tick_nohz_idle_exit(void); extern void tick_nohz_irq_exit(void); extern ktime_t tick_nohz_get_sleep_length(void); extern unsigned long tick_nohz_get_idle_calls(void); +extern unsigned long tick_nohz_get_idle_calls_cpu(int cpu); extern u64 get_cpu_idle_time_us(int cpu, u64 *last_update_time); extern u64 get_cpu_iowait_time_us(int cpu, u64 *last_update_time); #else /* !CONFIG_NO_HZ_COMMON */ --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -244,7 +244,7 @@ static void sugov_iowait_boost(struct su #ifdef CONFIG_NO_HZ_COMMON static bool sugov_cpu_is_busy(struct sugov_cpu *sg_cpu) { - unsigned long idle_calls = tick_nohz_get_idle_calls(); + unsigned long idle_calls = tick_nohz_get_idle_calls_cpu(sg_cpu->cpu); bool ret = idle_calls == sg_cpu->saved_idle_calls; sg_cpu->saved_idle_calls = idle_calls; --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -1010,6 +1010,19 @@ ktime_t tick_nohz_get_sleep_length(void) } /** + * tick_nohz_get_idle_calls_cpu - return the current idle calls counter value + * for a particular CPU. + * + * Called from the schedutil frequency scaling governor in scheduler context. + */ +unsigned long tick_nohz_get_idle_calls_cpu(int cpu) +{ + struct tick_sched *ts = tick_get_tick_sched(cpu); + + return ts->idle_calls; +} + +/** * tick_nohz_get_idle_calls - return the current idle calls counter value * * Called from the schedutil frequency scaling governor in scheduler context. Patches currently in stable-queue which might be from joelaf(a)google.com are queue-4.14/cpufreq-schedutil-use-idle_calls-counter-of-the-remote-cpu.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: wm_adsp: Fix validation of firmware and coeff lengths" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: wm_adsp: Fix validation of firmware and coeff lengths to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-wm_adsp-fix-validation-of-firmware-and-coeff-lengths.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 50dd2ea8ef67a1617e0c0658bcbec4b9fb03b936 Mon Sep 17 00:00:00 2001 From: Ben Hutchings <ben.hutchings(a)codethink.co.uk> Date: Fri, 8 Dec 2017 16:15:20 +0000 Subject: ASoC: wm_adsp: Fix validation of firmware and coeff lengths From: Ben Hutchings <ben.hutchings(a)codethink.co.uk> commit 50dd2ea8ef67a1617e0c0658bcbec4b9fb03b936 upstream. The checks for whether another region/block header could be present are subtracting the size from the current offset. Obviously we should instead subtract the offset from the size. The checks for whether the region/block data fit in the file are adding the data size to the current offset and header size, without checking for integer overflow. Rearrange these so that overflow is impossible. Signed-off-by: Ben Hutchings <ben.hutchings(a)codethink.co.uk> Acked-by: Charles Keepax <ckeepax(a)opensource.cirrus.com> Tested-by: Charles Keepax <ckeepax(a)opensource.cirrus.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/wm_adsp.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) --- a/sound/soc/codecs/wm_adsp.c +++ b/sound/soc/codecs/wm_adsp.c @@ -1733,7 +1733,7 @@ static int wm_adsp_load(struct wm_adsp * le64_to_cpu(footer->timestamp)); while (pos < firmware->size && - pos - firmware->size > sizeof(*region)) { + sizeof(*region) < firmware->size - pos) { region = (void *)&(firmware->data[pos]); region_name = "Unknown"; reg = 0; @@ -1782,8 +1782,8 @@ static int wm_adsp_load(struct wm_adsp * regions, le32_to_cpu(region->len), offset, region_name); - if ((pos + le32_to_cpu(region->len) + sizeof(*region)) > - firmware->size) { + if (le32_to_cpu(region->len) > + firmware->size - pos - sizeof(*region)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, regions, region_name, @@ -2253,7 +2253,7 @@ static int wm_adsp_load_coeff(struct wm_ blocks = 0; while (pos < firmware->size && - pos - firmware->size > sizeof(*blk)) { + sizeof(*blk) < firmware->size - pos) { blk = (void *)(&firmware->data[pos]); type = le16_to_cpu(blk->type); @@ -2327,8 +2327,8 @@ static int wm_adsp_load_coeff(struct wm_ } if (reg) { - if ((pos + le32_to_cpu(blk->len) + sizeof(*blk)) > - firmware->size) { + if (le32_to_cpu(blk->len) > + firmware->size - pos - sizeof(*blk)) { adsp_err(dsp, "%s.%d: %s region len %d bytes exceeds file length %zu\n", file, blocks, region_name, Patches currently in stable-queue which might be from ben.hutchings(a)codethink.co.uk are queue-4.14/asoc-wm_adsp-fix-validation-of-firmware-and-coeff-lengths.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: twl4030: fix child-node lookup" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: twl4030: fix child-node lookup to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-twl4030-fix-child-node-lookup.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 15f8c5f2415bfac73f33a14bcd83422bcbfb5298 Mon Sep 17 00:00:00 2001 From: Johan Hovold <johan(a)kernel.org> Date: Mon, 13 Nov 2017 12:12:56 +0100 Subject: ASoC: twl4030: fix child-node lookup From: Johan Hovold <johan(a)kernel.org> commit 15f8c5f2415bfac73f33a14bcd83422bcbfb5298 upstream. Fix child-node lookup during probe, which ended up searching the whole device tree depth-first starting at the parent rather than just matching on its children. To make things worse, the parent codec node was also prematurely freed, while the child node was leaked. Fixes: 2d6d649a2e0f ("ASoC: twl4030: Support for DT booted kernel") Signed-off-by: Johan Hovold <johan(a)kernel.org> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/twl4030.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/sound/soc/codecs/twl4030.c +++ b/sound/soc/codecs/twl4030.c @@ -232,7 +232,7 @@ static struct twl4030_codec_data *twl403 struct twl4030_codec_data *pdata = dev_get_platdata(codec->dev); struct device_node *twl4030_codec_node = NULL; - twl4030_codec_node = of_find_node_by_name(codec->dev->parent->of_node, + twl4030_codec_node = of_get_child_by_name(codec->dev->parent->of_node, "codec"); if (!pdata && twl4030_codec_node) { @@ -241,9 +241,11 @@ static struct twl4030_codec_data *twl403 GFP_KERNEL); if (!pdata) { dev_err(codec->dev, "Can not allocate memory\n"); + of_node_put(twl4030_codec_node); return NULL; } twl4030_setup_pdata_of(pdata, twl4030_codec_node); + of_node_put(twl4030_codec_node); } return pdata; Patches currently in stable-queue which might be from johan(a)kernel.org are queue-4.14/asoc-twl4030-fix-child-node-lookup.patch queue-4.14/asoc-da7218-fix-fix-child-node-lookup.patch

7 years, 8 months

1
0
0 0

[Linux-stable-mirror] Patch "ASoC: da7218: fix fix child-node lookup" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled ASoC: da7218: fix fix child-node lookup to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: asoc-da7218-fix-fix-child-node-lookup.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From bc6476d6c1edcb9b97621b5131bd169aa81f27db Mon Sep 17 00:00:00 2001 From: Johan Hovold <johan(a)kernel.org> Date: Mon, 13 Nov 2017 12:12:55 +0100 Subject: ASoC: da7218: fix fix child-node lookup From: Johan Hovold <johan(a)kernel.org> commit bc6476d6c1edcb9b97621b5131bd169aa81f27db upstream. Fix child-node lookup during probe, which ended up searching the whole device tree depth-first starting at the parent rather than just matching on its children. To make things worse, the parent codec node was also prematurely freed. Fixes: 4d50934abd22 ("ASoC: da7218: Add da7218 codec driver") Signed-off-by: Johan Hovold <johan(a)kernel.org> Acked-by: Adam Thomson <Adam.Thomson.Opensource(a)diasemi.com> Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- sound/soc/codecs/da7218.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/sound/soc/codecs/da7218.c +++ b/sound/soc/codecs/da7218.c @@ -2520,7 +2520,7 @@ static struct da7218_pdata *da7218_of_to } if (da7218->dev_id == DA7218_DEV_ID) { - hpldet_np = of_find_node_by_name(np, "da7218_hpldet"); + hpldet_np = of_get_child_by_name(np, "da7218_hpldet"); if (!hpldet_np) return pdata; Patches currently in stable-queue which might be from johan(a)kernel.org are queue-4.14/asoc-twl4030-fix-child-node-lookup.patch queue-4.14/asoc-da7218-fix-fix-child-node-lookup.patch

7 years, 8 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror