June 2018 - Linux-stable-mirror

patch "usb: dwc2: fix the incorrect bitmaps for the ports of multi_tt hub" added to usb-linus

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled usb: dwc2: fix the incorrect bitmaps for the ports of multi_tt hub to my usb git tree which can be found at git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb.git in the usb-linus branch. The patch will show up in the next release of the linux-next tree (usually sometime within the next 24 hours during the week.) The patch will hopefully also be merged in Linus's tree for the next -rc kernel release. If you have any questions about this process, please let me know. >From 8760675932ddb614e83702117d36ea644050c609 Mon Sep 17 00:00:00 2001 From: William Wu <william.wu(a)rock-chips.com> Date: Mon, 21 May 2018 18:12:00 +0800 Subject: usb: dwc2: fix the incorrect bitmaps for the ports of multi_tt hub The dwc2_get_ls_map() use ttport to reference into the bitmap if we're on a multi_tt hub. But the bitmaps index from 0 to (hub->maxchild - 1), while the ttport index from 1 to hub->maxchild. This will cause invalid memory access when the number of ttport is hub->maxchild. Without this patch, I can easily meet a Kernel panic issue if connect a low-speed USB mouse with the max port of FE2.1 multi-tt hub (1a40:0201) on rk3288 platform. Fixes: 9f9f09b048f5 ("usb: dwc2: host: Totally redo the microframe scheduler") Cc: <stable(a)vger.kernel.org> Reviewed-by: Douglas Anderson <dianders(a)chromium.org> Acked-by: Minas Harutyunyan hminas(a)synopsys.com> Signed-off-by: William Wu <william.wu(a)rock-chips.com> Signed-off-by: Felipe Balbi <felipe.balbi(a)linux.intel.com> --- drivers/usb/dwc2/hcd_queue.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/dwc2/hcd_queue.c b/drivers/usb/dwc2/hcd_queue.c index d7c3d6c776d8..9c55d1addba9 100644 --- a/drivers/usb/dwc2/hcd_queue.c +++ b/drivers/usb/dwc2/hcd_queue.c @@ -383,7 +383,7 @@ static unsigned long *dwc2_get_ls_map(struct dwc2_hsotg *hsotg, /* Get the map and adjust if this is a multi_tt hub */ map = qh->dwc_tt->periodic_bitmaps; if (qh->dwc_tt->usb_tt->multi) - map += DWC2_ELEMENTS_PER_LS_BITMAP * qh->ttport; + map += DWC2_ELEMENTS_PER_LS_BITMAP * (qh->ttport - 1); return map; } -- 2.18.0

6 years, 7 months

1
0
0 0

[PATCH V2] clk: ti: omap3: Fix am3517 mdio clock alias

by Adam Ford

A previous patch removed OMAP clock aliases that were perceived to be unnecessary. Unfortunately, it broke the ethernet on the am3517-evm. This patch enables the MDIO clock and EMAC clock. Fixes: 0ed266d7ae5e ("clk: ti: omap3: cleanup unnecessary clock aliases") Cc: stable(a)vger.kernel.org #4.16+ Signed-off-by: Adam Ford <aford173(a)gmail.com> --- V2: Instead of modifying clk-3xxx.c, this patch modifie the device tree to work like the original patch intended. diff --git a/arch/arm/boot/dts/am3517.dtsi b/arch/arm/boot/dts/am3517.dtsi index ca294914bbb1..4b6062b631b1 100644 --- a/arch/arm/boot/dts/am3517.dtsi +++ b/arch/arm/boot/dts/am3517.dtsi @@ -39,6 +39,8 @@ ti,davinci-ctrl-ram-size = <0x2000>; ti,davinci-rmii-en = /bits/ 8 <1>; local-mac-address = [ 00 00 00 00 00 00 ]; + clocks = <&emac_ick>; + clock-names = "ick"; }; davinci_mdio: ethernet@5c030000 { @@ -49,6 +51,8 @@ bus_freq = <1000000>; #address-cells = <1>; #size-cells = <0>; + clocks = <&emac_fck>; + clock-names = "fck"; }; uart4: serial@4809e000 { -- 2.17.1

6 years, 7 months

2
1
0 0

[PATCH] x86/efi: Fix efi_call_phys_epilog() with CONFIG_X86_5LEVEL=y

by Kirill A. Shutemov

Open-coded page table entry checks don't work correctly when we fold the page table level at runtime. pgd_present() on 4-level paging machine always returns true, but open-coded version of the check may return false-negative result and we silently skip the rest of the loop body in efi_call_phys_epilog(). Replace open-coded checks with proper helpers. Signed-off-by: Kirill A. Shutemov <kirill.shutemov(a)linux.intel.com> Fixes: 94133e46a0f5 ("x86/efi: Correct EFI identity mapping under 'efi=old_map' when KASLR is enabled") Cc: stable(a)vger.kernel.org # 4.12+ Cc: Baoquan He <bhe(a)redhat.com> Cc: Matt Fleming <matt(a)codeblueprint.co.uk> Cc: Andrey Ryabinin <aryabinin(a)virtuozzo.com> --- arch/x86/platform/efi/efi_64.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/platform/efi/efi_64.c b/arch/x86/platform/efi/efi_64.c index e01f7ceb9e7a..77873ce700ae 100644 --- a/arch/x86/platform/efi/efi_64.c +++ b/arch/x86/platform/efi/efi_64.c @@ -166,14 +166,14 @@ void __init efi_call_phys_epilog(pgd_t *save_pgd) pgd = pgd_offset_k(pgd_idx * PGDIR_SIZE); set_pgd(pgd_offset_k(pgd_idx * PGDIR_SIZE), save_pgd[pgd_idx]); - if (!(pgd_val(*pgd) & _PAGE_PRESENT)) + if (!pgd_present(*pgd)) continue; for (i = 0; i < PTRS_PER_P4D; i++) { p4d = p4d_offset(pgd, pgd_idx * PGDIR_SIZE + i * P4D_SIZE); - if (!(p4d_val(*p4d) & _PAGE_PRESENT)) + if (!p4d_present(*p4d)) continue; pud = (pud_t *)p4d_page_vaddr(*p4d); -- 2.18.0

6 years, 7 months

1
0
0 0

[PATCH 6/8] mpt3sas: Fix _transport_smp_handler() error path

by Bart Van Assche

This patch avoids that smatch complains about a double unlock on ioc->transport_cmds.mutex. Fixes: 651a01364994 ("scsi: scsi_transport_sas: switch to bsg-lib for SMP passthrough") Signed-off-by: Bart Van Assche <bart.vanassche(a)wdc.com> Cc: Christoph Hellwig <hch(a)lst.de> Cc: Sathya Prakash <sathya.prakash(a)broadcom.com> Cc: Chaitra P B <chaitra.basappa(a)broadcom.com> Cc: Suganath Prabu Subramani <suganath-prabu.subramani(a)broadcom.com> Cc: stable(a)vger.kernel.org --- drivers/scsi/mpt3sas/mpt3sas_transport.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/mpt3sas/mpt3sas_transport.c b/drivers/scsi/mpt3sas/mpt3sas_transport.c index 77a8ec401461..53224d97f6c1 100644 --- a/drivers/scsi/mpt3sas/mpt3sas_transport.c +++ b/drivers/scsi/mpt3sas/mpt3sas_transport.c @@ -1938,12 +1938,12 @@ _transport_smp_handler(struct bsg_job *job, struct Scsi_Host *shost, pr_info(MPT3SAS_FMT "%s: host reset in progress!\n", __func__, ioc->name); rc = -EFAULT; - goto out; + goto job_done; } rc = mutex_lock_interruptible(&ioc->transport_cmds.mutex); if (rc) - goto out; + goto job_done; if (ioc->transport_cmds.status != MPT3_CMD_NOT_USED) { pr_err(MPT3SAS_FMT "%s: transport_cmds in use\n", ioc->name, @@ -2068,6 +2068,7 @@ _transport_smp_handler(struct bsg_job *job, struct Scsi_Host *shost, out: ioc->transport_cmds.status = MPT3_CMD_NOT_USED; mutex_unlock(&ioc->transport_cmds.mutex); +job_done: bsg_job_done(job, rc, reslen); } -- 2.17.0

6 years, 7 months

2
1
0 0

[PATCH] clk: ti: omap3: Fix am3517 mdio clock alias

by Adam Ford

A previous patch removed OMAP clock aliases that were perceived to be unnecessary. Unfortunately, it broke the ethernet on the am3517-evm. This patch renables the MDIO clock. Fixes: 0ed266d7ae5e ("clk: ti: omap3: cleanup unnecessary clock aliases") Cc: stable(a)vger.kernel.org #4.16+ Signed-off-by: Adam Ford <aford173(a)gmail.com> diff --git a/drivers/clk/ti/clk-3xxx.c b/drivers/clk/ti/clk-3xxx.c index 8aa5f5793835..f84734bd0aa7 100644 --- a/drivers/clk/ti/clk-3xxx.c +++ b/drivers/clk/ti/clk-3xxx.c @@ -254,6 +254,8 @@ static struct ti_dt_clk omap36xx_am35xx_omap3430es2plus_clks[] = { }; static struct ti_dt_clk am35xx_clks[] = { + DT_CLK("davinci_emac.0", NULL, "emac_ick"), + DT_CLK("davinci_mdio.0", NULL, "emac_fck"), DT_CLK(NULL, "hsotgusb_ick", "hsotgusb_ick_am35xx"), DT_CLK(NULL, "hsotgusb_fck", "hsotgusb_fck_am35xx"), DT_CLK(NULL, "uart4_ick", "uart4_ick_am35xx"), -- 2.17.1

6 years, 7 months

3
3
0 0

[PATCH 4.4 00/37] 4.4.136-stable review

by Greg Kroah-Hartman

This is the start of the stable review cycle for the 4.4.136 release. There are 37 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know. Responses should be made by Thu Jun 7 17:00:49 UTC 2018. Anything received after that time might be too late. The whole patch series can be found in one patch at: https://www.kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.4.136-rc… or in the git tree and branch at: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.4.y and the diffstat can be found below. thanks, greg k-h ------------- Pseudo-Shortlog of commits: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Linux 4.4.136-rc1 David S. Miller <davem(a)davemloft.net> sparc64: Fix build warnings with gcc 7. David S. Miller <davem(a)davemloft.net> sparc64: Don't clibber fixed registers in __multi4. David S. Miller <davem(a)davemloft.net> sparc64: Add __multi3 for gcc 7.x and later. Hugh Dickins <hughd(a)google.com> mm: fix the NULL mapping case in __isolate_lru_page() Al Viro <viro(a)zeniv.linux.org.uk> fix io_destroy()/aio_complete() race Arnd Bergmann <arnd(a)arndb.de> Kbuild: change CC_OPTIMIZE_FOR_SIZE definition Ondrej Zary <linux(a)rainbow-software.org> drm/i915: Disable LVDS on Radiant P845 Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> hwtracing: stm: fix build error on some arches Alexander Shishkin <alexander.shishkin(a)linux.intel.com> stm class: Use vmalloc for the master map Bart Van Assche <bart.vanassche(a)wdc.com> scsi: scsi_transport_srp: Fix shost to rport translation Maciej W. Rozycki <macro(a)mips.com> MIPS: prctl: Disallow FRE without FR with PR_SET_FP_MODE requests Maciej W. Rozycki <macro(a)mips.com> MIPS: ptrace: Fix PTRACE_PEEKUSR requests for 64-bit FGRs Martin Kelly <mkelly(a)xevo.com> iio:kfifo_buf: check for uint overflow Yoshihiro Shimoda <yoshihiro.shimoda.uh(a)renesas.com> dmaengine: usb-dmac: fix endless loop in usb_dmac_chan_terminate_all() Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: revoke START request early Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: check master irqs before slave irqs Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: don't issue stop when HW does it automatically Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: init new messages in irq Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: refactor setup of a msg Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: remove spinlock Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: remove unused IOERROR state Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: rework hw init Wolfram Sang <wsa+renesas(a)sang-engineering.com> i2c: rcar: make sure clocks are on when doing clock calculation Eric Dumazet <edumazet(a)google.com> tcp: avoid integer overflows in tcp_rcv_space_adjust() Arnd Bergmann <arnd(a)arndb.de> irda: fix overly long udelay() Colin Ian King <colin.king(a)canonical.com> ASoC: Intel: sst: remove redundant variable dma_dev_name Matthias Kaehlcke <mka(a)chromium.org> rtlwifi: rtl8192cu: Remove variable self-assignment in rf.c Eric Biggers <ebiggers(a)google.com> cfg80211: further limit wiphy names to 64 bytes Sachin Grover <sgrover(a)codeaurora.org> selinux: KASAN: slab-out-of-bounds in xattr_getsecurity Steven Rostedt (VMware) <rostedt(a)goodmis.org> tracing: Fix crash when freeing instances with event triggers Benjamin Tissoires <benjamin.tissoires(a)redhat.com> Input: elan_i2c_smbus - fix corrupted stack Mimi Zohar <zohar(a)linux.vnet.ibm.com> Revert "ima: limit file hash setting by user to fix and log modes" Brian Foster <bfoster(a)redhat.com> xfs: detect agfl count corruption and reset agfl Yoshinori Sato <ysato(a)users.sourceforge.jp> sh: New gcc support Geert Uytterhoeven <geert(a)linux-m68k.org> USB: serial: cp210x: use tcflag_t to fix incompatible pointer type Michael Neuling <mikey(a)neuling.org> powerpc/64s: Clear PCR on boot Will Deacon <will.deacon(a)arm.com> arm64: lse: Add early clobbers to some input/output asm operands ------------- Diffstat: Makefile | 4 +- arch/arm64/include/asm/atomic_lse.h | 14 +- arch/mips/kernel/process.c | 4 + arch/mips/kernel/ptrace.c | 2 +- arch/mips/kernel/ptrace32.c | 2 +- arch/powerpc/kernel/cpu_setup_power.S | 4 + arch/sh/kernel/sh_ksyms_32.c | 3 + arch/sh/lib/ashlsi3.S | 35 +++- arch/sh/lib/ashrsi3.S | 33 +++- arch/sh/lib/lshrsi3.S | 34 +++- arch/sparc/kernel/ds.c | 2 +- arch/sparc/lib/Makefile | 1 + arch/sparc/lib/multi3.S | 35 ++++ drivers/dma/sh/usb-dmac.c | 4 +- drivers/gpu/drm/i915/intel_lvds.c | 8 + drivers/hwtracing/stm/core.c | 7 +- drivers/i2c/busses/i2c-rcar.c | 202 +++++++++------------ drivers/iio/buffer/kfifo_buf.c | 7 + drivers/input/mouse/elan_i2c_smbus.c | 22 +-- drivers/net/irda/w83977af_ir.c | 4 +- .../net/wireless/realtek/rtlwifi/rtl8192cu/rf.c | 3 - drivers/scsi/scsi_transport_srp.c | 22 ++- drivers/usb/serial/cp210x.c | 9 +- fs/aio.c | 3 +- fs/xfs/libxfs/xfs_alloc.c | 94 ++++++++++ fs/xfs/xfs_mount.h | 1 + fs/xfs/xfs_trace.h | 9 +- include/linux/tcp.h | 2 +- include/uapi/linux/nl80211.h | 2 +- init/Kconfig | 13 ++ kernel/trace/trace_events_trigger.c | 5 +- mm/vmscan.c | 2 +- net/ipv4/tcp_input.c | 10 +- security/integrity/ima/ima_appraise.c | 8 +- security/selinux/ss/services.c | 2 +- sound/soc/intel/common/sst-firmware.c | 2 - 36 files changed, 410 insertions(+), 204 deletions(-)

6 years, 7 months

7
46
0 0

[PATCH] ALSA: hda/realtek - Fix the problem of two front mics on more machines

by Hui Wang

We have 3 more Lenovo machines, they all have 2 front mics on them, so they need the fixup to change the location for one of two mics. Among these 3 Lenovo machines, one of them has the same pin cfg as the machine with subid 0x17aa3138, so use the pin cfg table to apply fixup for them. The rest machines don't share the same pin cfg, so far use the subid to apply fixup for them. Fixes: a3dafb2200bf ("ALSA: hda/realtek - adjust the location of one mic") Cc: <stable(a)vger.kernel.org> Signed-off-by: Hui Wang <hui.wang(a)canonical.com> --- sound/pci/hda/patch_realtek.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 4834847..418f27b 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -6594,8 +6594,9 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x30bb, "ThinkCentre AIO", ALC233_FIXUP_LENOVO_LINE2_MIC_HOTKEY), SND_PCI_QUIRK(0x17aa, 0x30e2, "ThinkCentre AIO", ALC233_FIXUP_LENOVO_LINE2_MIC_HOTKEY), SND_PCI_QUIRK(0x17aa, 0x310c, "ThinkCentre Station", ALC294_FIXUP_LENOVO_MIC_LOCATION), + SND_PCI_QUIRK(0x17aa, 0x312a, "ThinkCentre Station", ALC294_FIXUP_LENOVO_MIC_LOCATION), SND_PCI_QUIRK(0x17aa, 0x312f, "ThinkCentre Station", ALC294_FIXUP_LENOVO_MIC_LOCATION), - SND_PCI_QUIRK(0x17aa, 0x3138, "ThinkCentre Station", ALC294_FIXUP_LENOVO_MIC_LOCATION), + SND_PCI_QUIRK(0x17aa, 0x3136, "ThinkCentre Station", ALC294_FIXUP_LENOVO_MIC_LOCATION), SND_PCI_QUIRK(0x17aa, 0x313c, "ThinkCentre Station", ALC294_FIXUP_LENOVO_MIC_LOCATION), SND_PCI_QUIRK(0x17aa, 0x3902, "Lenovo E50-80", ALC269_FIXUP_DMIC_THINKPAD_ACPI), SND_PCI_QUIRK(0x17aa, 0x3977, "IdeaPad S210", ALC283_FIXUP_INT_MIC), @@ -6773,6 +6774,12 @@ static const struct snd_hda_pin_quirk alc269_pin_fixup_tbl[] = { {0x14, 0x90170110}, {0x19, 0x02a11030}, {0x21, 0x02211020}), + SND_HDA_PIN_QUIRK(0x10ec0235, 0x17aa, "Lenovo", ALC294_FIXUP_LENOVO_MIC_LOCATION, + {0x14, 0x90170110}, + {0x19, 0x02a11030}, + {0x1a, 0x02a11040}, + {0x1b, 0x01014020}, + {0x21, 0x0221101f}), SND_HDA_PIN_QUIRK(0x10ec0236, 0x1028, "Dell", ALC255_FIXUP_DELL1_MIC_NO_PRESENCE, {0x12, 0x90a60140}, {0x14, 0x90170110}, -- 2.7.4

6 years, 7 months

2
1
0 0

FAILED: patch "[PATCH] media: rc: ensure input/lirc device can be opened after" failed to apply to 4.16-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.16-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 d7832cd2a3c87eb6ae1e802c88b6fc56c5823f6d Mon Sep 17 00:00:00 2001 From: Sean Young <sean(a)mess.org> Date: Thu, 24 May 2018 05:47:17 -0400 Subject: [PATCH] media: rc: ensure input/lirc device can be opened after register Since commit cb84343fced1 ("media: lirc: do not call close() or open() on unregistered devices") rc_open() will return -ENODEV if rcdev->registered is false. Ensure this is set before we register the input device and the lirc device, else we have a short window where the neither the lirc or input device can be opened. Fixes: cb84343fced1 ("media: lirc: do not call close() or open() on unregistered devices") Cc: stable(a)vger.kernel.org # v4.16+ Signed-off-by: Sean Young <sean(a)mess.org> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung(a)kernel.org> diff --git a/drivers/media/rc/rc-main.c b/drivers/media/rc/rc-main.c index b7071bde670a..2e222d9ee01f 100644 --- a/drivers/media/rc/rc-main.c +++ b/drivers/media/rc/rc-main.c @@ -1862,6 +1862,8 @@ int rc_register_device(struct rc_dev *dev) dev->device_name ?: "Unspecified device", path ?: "N/A"); kfree(path); + dev->registered = true; + if (dev->driver_type != RC_DRIVER_IR_RAW_TX) { rc = rc_setup_rx_device(dev); if (rc) @@ -1881,8 +1883,6 @@ int rc_register_device(struct rc_dev *dev) goto out_lirc; } - dev->registered = true; - dev_dbg(&dev->dev, "Registered rc%u (driver: %s)\n", dev->minor, dev->driver_name ? dev->driver_name : "unknown");

6 years, 7 months

2
1
0 0

[PATCHES] Networking

by David Miller

Please queue up the following networking bug fixes for 4.16 and 4.17 -stable, respectively.

6 years, 7 months

2
2
0 0

FAILED: patch "[PATCH] blk-mq: avoid starving tag allocation after allocating" failed to apply to 4.17-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 4.17-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 e6fc46498784e799d3eb95d83079180e413c4e7d Mon Sep 17 00:00:00 2001 From: Ming Lei <ming.lei(a)redhat.com> Date: Thu, 24 May 2018 11:00:39 -0600 Subject: [PATCH] blk-mq: avoid starving tag allocation after allocating process migrates When the allocation process is scheduled back and the mapped hw queue is changed, fake one extra wake up on previous queue for compensating wake up miss, so other allocations on the previous queue won't be starved. This patch fixes one request allocation hang issue, which can be triggered easily in case of very low nr_request. The race is as follows: 1) 2 hw queues, nr_requests are 2, and wake_batch is one 2) there are 3 waiters on hw queue 0 3) two in-flight requests in hw queue 0 are completed, and only two waiters of 3 are waken up because of wake_batch, but both the two waiters can be scheduled to another CPU and cause to switch to hw queue 1 4) then the 3rd waiter will wait for ever, since no in-flight request is in hw queue 0 any more. 5) this patch fixes it by the fake wakeup when waiter is scheduled to another hw queue Cc: <stable(a)vger.kernel.org> Reviewed-by: Omar Sandoval <osandov(a)fb.com> Signed-off-by: Ming Lei <ming.lei(a)redhat.com> Modified commit message to make it clearer, and make it apply on top of the 4.18 branch. Signed-off-by: Jens Axboe <axboe(a)kernel.dk> diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c index 336dde07b230..a4e58fc28a06 100644 --- a/block/blk-mq-tag.c +++ b/block/blk-mq-tag.c @@ -134,6 +134,8 @@ unsigned int blk_mq_get_tag(struct blk_mq_alloc_data *data) ws = bt_wait_ptr(bt, data->hctx); drop_ctx = data->ctx == NULL; do { + struct sbitmap_queue *bt_prev; + /* * We're out of tags on this hardware queue, kick any * pending IO submits before going to sleep waiting for @@ -159,6 +161,7 @@ unsigned int blk_mq_get_tag(struct blk_mq_alloc_data *data) if (data->ctx) blk_mq_put_ctx(data->ctx); + bt_prev = bt; io_schedule(); data->ctx = blk_mq_get_ctx(data->q); @@ -170,6 +173,15 @@ unsigned int blk_mq_get_tag(struct blk_mq_alloc_data *data) bt = &tags->bitmap_tags; finish_wait(&ws->wait, &wait); + + /* + * If destination hw queue is changed, fake wake up on + * previous queue for compensating the wake up miss, so + * other allocations on previous queue won't be starved. + */ + if (bt != bt_prev) + sbitmap_queue_wake_up(bt_prev); + ws = bt_wait_ptr(bt, data->hctx); } while (1); diff --git a/include/linux/sbitmap.h b/include/linux/sbitmap.h index 0c4a9c242dd7..e6539536dea9 100644 --- a/include/linux/sbitmap.h +++ b/include/linux/sbitmap.h @@ -512,6 +512,13 @@ static inline struct sbq_wait_state *sbq_wait_ptr(struct sbitmap_queue *sbq, */ void sbitmap_queue_wake_all(struct sbitmap_queue *sbq); +/** + * sbitmap_queue_wake_up() - Wake up some of waiters in one waitqueue + * on a &struct sbitmap_queue. + * @sbq: Bitmap queue to wake up. + */ +void sbitmap_queue_wake_up(struct sbitmap_queue *sbq); + /** * sbitmap_queue_show() - Dump &struct sbitmap_queue information to a &struct * seq_file. diff --git a/lib/sbitmap.c b/lib/sbitmap.c index e6d7d610778d..6fdc6267f4a8 100644 --- a/lib/sbitmap.c +++ b/lib/sbitmap.c @@ -352,8 +352,9 @@ static void sbitmap_queue_update_wake_batch(struct sbitmap_queue *sbq, if (sbq->wake_batch != wake_batch) { WRITE_ONCE(sbq->wake_batch, wake_batch); /* - * Pairs with the memory barrier in sbq_wake_up() to ensure that - * the batch size is updated before the wait counts. + * Pairs with the memory barrier in sbitmap_queue_wake_up() + * to ensure that the batch size is updated before the wait + * counts. */ smp_mb__before_atomic(); for (i = 0; i < SBQ_WAIT_QUEUES; i++) @@ -463,15 +464,6 @@ static bool __sbq_wake_up(struct sbitmap_queue *sbq) unsigned int wake_batch; int wait_cnt; - /* - * Pairs with the memory barrier in set_current_state() to ensure the - * proper ordering of clear_bit()/waitqueue_active() in the waker and - * test_and_set_bit_lock()/prepare_to_wait()/finish_wait() in the - * waiter. See the comment on waitqueue_active(). This is __after_atomic - * because we just did clear_bit_unlock() in the caller. - */ - smp_mb__after_atomic(); - ws = sbq_wake_ptr(sbq); if (!ws) return false; @@ -507,17 +499,26 @@ static bool __sbq_wake_up(struct sbitmap_queue *sbq) return false; } -static void sbq_wake_up(struct sbitmap_queue *sbq) +void sbitmap_queue_wake_up(struct sbitmap_queue *sbq) { while (__sbq_wake_up(sbq)) ; } +EXPORT_SYMBOL_GPL(sbitmap_queue_wake_up); void sbitmap_queue_clear(struct sbitmap_queue *sbq, unsigned int nr, unsigned int cpu) { sbitmap_clear_bit_unlock(&sbq->sb, nr); - sbq_wake_up(sbq); + /* + * Pairs with the memory barrier in set_current_state() to ensure the + * proper ordering of clear_bit_unlock()/waitqueue_active() in the waker + * and test_and_set_bit_lock()/prepare_to_wait()/finish_wait() in the + * waiter. See the comment on waitqueue_active(). + */ + smp_mb__after_atomic(); + sbitmap_queue_wake_up(sbq); + if (likely(!sbq->round_robin && nr < sbq->sb.depth)) *per_cpu_ptr(sbq->alloc_hint, cpu) = nr; } @@ -529,7 +530,7 @@ void sbitmap_queue_wake_all(struct sbitmap_queue *sbq) /* * Pairs with the memory barrier in set_current_state() like in - * sbq_wake_up(). + * sbitmap_queue_wake_up(). */ smp_mb(); wake_index = atomic_read(&sbq->wake_index);

6 years, 7 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror June 2018