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

Re: [PATCH] kbuild: Handle builtin dtb files containing hyphens

by Masahiro Yamada

2018-03-08 16:09 GMT+09:00 James Hogan <jhogan(a)kernel.org>: > On Wed, Mar 07, 2018 at 03:19:11PM -0800, Frank Rowand wrote: >> On 03/07/18 12:25, James Hogan wrote: >> > On Wed, Mar 07, 2018 at 12:11:41PM -0800, Frank Rowand wrote: >> >> On 03/07/18 06:06, James Hogan wrote: >> >>> Quite a lot of dts files have hyphens, but its only a problem on MIPS >> >>> where such files can be built into the kernel. For example when >> >>> CONFIG_DT_NETGEAR_CVG834G=y, or on BMIPS kernels when the dtbs target is >> >>> used (in the latter case it admitedly shouldn't really build all the >> >>> dtb.o files, but thats a separate issue). > >> > I'll keep the paragraph about MIPS and the example configuration though, >> > as I think its important information to reproduce the problem, and to >> > justify why it wouldn't be appropriate to just rename the files (which >> > was my first reaction). >> >> Other than the part that says "its only a problem on MIPS". That is >> pedantically correct because no other architecture (that I am aware >> of, not that I searched) currently has a devicetree source file name >> with a hyphen in it, where that file is compiled into the kernel as >> an asm file. But it is potentially a problem on any architecture >> to it is misleading to label it as MIPS only. > > Okay I'll reword to make it clearer and do a v2. > > Thanks > James The code looks good. If you send v2, I can shortly apply it to the fixes branch. If possible, I want to send a PR this weekend. -- Best Regards Masahiro Yamada

7 years, 3 months

2
2
0 0

[PATCH] ubi: Reject MLC NAND

by Richard Weinberger

While UBI and UBIFS seem to work at first sight with MLC NAND, you will most likely lose all your data upon a power-cut or due to read/write disturb. In order to protect users from bad surprises, refuse to attach to MLC NAND. Cc: stable(a)vger.kernel.org Signed-off-by: Richard Weinberger <richard(a)nod.at> --- drivers/mtd/ubi/build.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c index e941395de3ae..753494e042d5 100644 --- a/drivers/mtd/ubi/build.c +++ b/drivers/mtd/ubi/build.c @@ -854,6 +854,17 @@ int ubi_attach_mtd_dev(struct mtd_info *mtd, int ubi_num, return -EINVAL; } + /* + * Both UBI and UBIFS have been designed for SLC NAND and NOR flashes. + * MLC NAND is different and needs special care, otherwise UBI or UBIFS + * will die soon and you will lose all your data. + */ + if (mtd->type == MTD_MLCNANDFLASH) { + pr_err("ubi: refuse attaching mtd%d - MLC NAND is not supported\n", + mtd->index); + return -EINVAL; + } + if (ubi_num == UBI_DEV_NUM_AUTO) { /* Search for an empty slot in the @ubi_devices array */ for (ubi_num = 0; ubi_num < UBI_MAX_DEVICES; ubi_num++) -- 2.13.6

7 years, 3 months

4
6
0 0

[RFC] mmc: block: fix updating ext_csd caches on ioctl call

by Bastian Stender

PARTITION_CONFIG is cached in mmc_card->ext_csd.part_config and the currently active partition in mmc_blk_data->part_curr. These caches do not always reflect changes if the ioctl call modifies the PARTITION_CONFIG registers, e.g. by changing BOOT_PARTITION_ENABLE. Write the PARTITION_CONFIG value extracted from the ioctl call to the cache and update the currently active partition accordingly. This ensures that the user space cannot change the values behind the kernel's back. The next call to mmc_blk_part_switch() will operate on the data set by the ioctl and reflect the changes appropriately. Signed-off-by: Bastian Stender <bst(a)pengutronix.de> Signed-off-by: Jan Luebbe <jlu(a)pengutronix.de> Cc: stable(a)vger.kernel.org --- drivers/mmc/core/block.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/drivers/mmc/core/block.c b/drivers/mmc/core/block.c index 20135a5de748..2cfb963d9f37 100644 --- a/drivers/mmc/core/block.c +++ b/drivers/mmc/core/block.c @@ -72,6 +72,7 @@ MODULE_ALIAS("mmc:block"); #define MMC_BLK_TIMEOUT_MS (10 * 1000) #define MMC_SANITIZE_REQ_TIMEOUT 240000 #define MMC_EXTRACT_INDEX_FROM_ARG(x) ((x & 0x00FF0000) >> 16) +#define MMC_EXTRACT_VALUE_FROM_ARG(x) ((x & 0x0000FF00) >> 8) #define mmc_req_rel_wr(req) ((req->cmd_flags & REQ_FUA) && \ (rq_data_dir(req) == WRITE)) @@ -586,6 +587,24 @@ static int __mmc_blk_ioctl_cmd(struct mmc_card *card, struct mmc_blk_data *md, return data.error; } + /* + * Make sure the cache of the PARTITION_CONFIG register and + * PARTITION_ACCESS bits is updated in case the ioctl ext_csd write + * changed it successfully. + */ + if ((MMC_EXTRACT_INDEX_FROM_ARG(cmd.arg) == EXT_CSD_PART_CONFIG) && + (cmd.opcode == MMC_SWITCH)) { + struct mmc_blk_data *main_md = dev_get_drvdata(&card->dev); + u8 value = MMC_EXTRACT_VALUE_FROM_ARG(cmd.arg); + + /* + * Update cache so the next mmc_blk_part_switch call operates + * on up-to-date data. + */ + card->ext_csd.part_config = value; + main_md->part_curr = value & EXT_CSD_PART_CONFIG_ACC_MASK; + } + /* * According to the SD specs, some commands require a delay after * issuing the command. -- 2.16.1

7 years, 3 months

1
1
0 0

patch "usb: dwc3: Fix lock-up on ID change during system suspend/resume" 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: dwc3: Fix lock-up on ID change during system suspend/resume 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 084a804e01205bcd74cd0849bc72cb5c88f8e648 Mon Sep 17 00:00:00 2001 From: Roger Quadros <rogerq(a)ti.com> Date: Tue, 27 Feb 2018 12:41:41 +0200 Subject: usb: dwc3: Fix lock-up on ID change during system suspend/resume To reproduce the lock up do the following - connect otg host adapter and a USB device to the dual-role port so that it is in host mode. - suspend to mem. - disconnect otg adapter. - resume the system. If we call dwc3_host_exit() before tasks are thawed xhci_plat_remove() seems to lock up at the second usb_remove_hcd() call. To work around this we queue the _dwc3_set_mode() work on the system_freezable_wq. Fixes: 41ce1456e1db ("usb: dwc3: core: make dwc3_set_mode() work properly") Cc: <stable(a)vger.kernel.org> # v4.12+ Suggested-by: Manu Gautam <mgautam(a)codeaurora.org> Signed-off-by: Roger Quadros <rogerq(a)ti.com> Signed-off-by: Felipe Balbi <felipe.balbi(a)linux.intel.com> --- drivers/usb/dwc3/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/dwc3/core.c b/drivers/usb/dwc3/core.c index f1d838a4acd6..e94bf91cc58a 100644 --- a/drivers/usb/dwc3/core.c +++ b/drivers/usb/dwc3/core.c @@ -175,7 +175,7 @@ void dwc3_set_mode(struct dwc3 *dwc, u32 mode) dwc->desired_dr_role = mode; spin_unlock_irqrestore(&dwc->lock, flags); - queue_work(system_power_efficient_wq, &dwc->drd_work); + queue_work(system_freezable_wq, &dwc->drd_work); } u32 dwc3_core_fifo_space(struct dwc3_ep *dep, u8 type) -- 2.16.2

7 years, 3 months

1
0
0 0

[PATCH 1/4] xhci: fix endpoint context tracer output

by Mathias Nyman

Fix incorrent values showed for max Primary stream and Linear stream array (LSA) values in the endpoint context decoder. Fixes: 19a7d0d65c4a ("usb: host: xhci: add Slot and EP Context tracers") Cc: <stable(a)vger.kernel.org> # v4.12+ Signed-off-by: Mathias Nyman <mathias.nyman(a)linux.intel.com> --- drivers/usb/host/xhci.h | 22 ++++++++++++---------- 1 file changed, 12 insertions(+), 10 deletions(-) diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h index e4d7d3d..d20e57b 100644 --- a/drivers/usb/host/xhci.h +++ b/drivers/usb/host/xhci.h @@ -718,11 +718,12 @@ struct xhci_ep_ctx { /* bits 10:14 are Max Primary Streams */ /* bit 15 is Linear Stream Array */ /* Interval - period between requests to an endpoint - 125u increments. */ -#define EP_INTERVAL(p) (((p) & 0xff) << 16) -#define EP_INTERVAL_TO_UFRAMES(p) (1 << (((p) >> 16) & 0xff)) -#define CTX_TO_EP_INTERVAL(p) (((p) >> 16) & 0xff) -#define EP_MAXPSTREAMS_MASK (0x1f << 10) -#define EP_MAXPSTREAMS(p) (((p) << 10) & EP_MAXPSTREAMS_MASK) +#define EP_INTERVAL(p) (((p) & 0xff) << 16) +#define EP_INTERVAL_TO_UFRAMES(p) (1 << (((p) >> 16) & 0xff)) +#define CTX_TO_EP_INTERVAL(p) (((p) >> 16) & 0xff) +#define EP_MAXPSTREAMS_MASK (0x1f << 10) +#define EP_MAXPSTREAMS(p) (((p) << 10) & EP_MAXPSTREAMS_MASK) +#define CTX_TO_EP_MAXPSTREAMS(p) (((p) & EP_MAXPSTREAMS_MASK) >> 10) /* Endpoint is set up with a Linear Stream Array (vs. Secondary Stream Array) */ #define EP_HAS_LSA (1 << 15) /* hosts with LEC=1 use bits 31:24 as ESIT high bits. */ @@ -2549,21 +2550,22 @@ static inline const char *xhci_decode_ep_context(u32 info, u32 info2, u64 deq, u8 burst; u8 cerr; u8 mult; - u8 lsa; - u8 hid; + + bool lsa; + bool hid; esit = CTX_TO_MAX_ESIT_PAYLOAD_HI(info) << 16 | CTX_TO_MAX_ESIT_PAYLOAD(tx_info); ep_state = info & EP_STATE_MASK; - max_pstr = info & EP_MAXPSTREAMS_MASK; + max_pstr = CTX_TO_EP_MAXPSTREAMS(info); interval = CTX_TO_EP_INTERVAL(info); mult = CTX_TO_EP_MULT(info) + 1; - lsa = info & EP_HAS_LSA; + lsa = !!(info & EP_HAS_LSA); cerr = (info2 & (3 << 1)) >> 1; ep_type = CTX_TO_EP_TYPE(info2); - hid = info2 & (1 << 7); + hid = !!(info2 & (1 << 7)); burst = CTX_TO_MAX_BURST(info2); maxp = MAX_PACKET_DECODED(info2); -- 2.7.4

7 years, 3 months

1
0
0 0

Re: 4.14.18 -> 4.14.24 - almost all guests hanged

by Greg KH

On Wed, Mar 07, 2018 at 09:29:10PM +0100, Nikola Ciprich wrote: > Hi, > > > > > I'd like to report that when upgrading our cluster from 4.14.18 to > > > > 4.14.24-rc1 (with live guests migration), almost none of guests survived.. > > > What's your hardware setup, intel with IBPB enabled microcode? > > Intel(R) Xeon(R) CPU E5-2650 0 @ 2.00GHz > > > > therefore I suppose no IBPB (at least meltdown checker reports so) > > > > > > > Does guests hang right after live migration? > > yes, just tried it. > > > > > > > > > > Are you able to reproduce the problem, does it work with latest upstream? > > yup, so I'm able to reproduce quickly. I'll revert the cluster to 4.14.18 now, > > but setup test system just afterwards, so and test the patch you've proposed. > > > > > > > > Not sure it helps, but following patch is missing in 4.14.24 > > > > > > commit 37b95951c58fdf08dc10afa9d02066ed9f176fb5 upstream. > > > > > > kvm_valid_sregs() should use X86_CR0_PG and X86_CR4_PAE to check bit > > > status rather than X86_CR0_PG_BIT and X86_CR4_PAE_BIT. This patch is > > > to fix it. > > > > > > Fixes: f29810335965a(KVM/x86: Check input paging mode when cs.l is set) > > > Reported-by: Jeremi Piotrowski <jeremi.piotrowski(a)gmail.com> > > > Cc: Paolo Bonzini <pbonzini(a)redhat.com> > > > Cc: Radim Krčmář <rkrcmar(a)redhat.com> > > > Signed-off-by: Tianyu Lan <Tianyu.Lan(a)microsoft.com> > > > Signed-off-by: Radim Krčmář <rkrcmar(a)redhat.com> > > > > I'll test and report. > > so indeed, this one on top of 4.14.24-rc1 fixes the migration for me. > Greg, could you queue this one up please? As was already pointed out, this is already queued up to be in the next release. thanks, greg k-h

7 years, 3 months

1
0
0 0

Patch "leds: do not overflow sysfs buffer in led_trigger_show" 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 leds: do not overflow sysfs buffer in led_trigger_show 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: leds-do-not-overflow-sysfs-buffer-in-led_trigger_show.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 3b9b95363c45365d606ad4bbba16acca75fdf6d3 Mon Sep 17 00:00:00 2001 From: Nathan Sullivan <nathan.sullivan(a)ni.com> Date: Mon, 15 Aug 2016 17:20:14 -0500 Subject: leds: do not overflow sysfs buffer in led_trigger_show From: Nathan Sullivan <nathan.sullivan(a)ni.com> commit 3b9b95363c45365d606ad4bbba16acca75fdf6d3 upstream. Per the documentation, use scnprintf instead of sprintf to ensure there is never more than PAGE_SIZE bytes of trigger names put into the buffer. Signed-off-by: Nathan Sullivan <nathan.sullivan(a)ni.com> Signed-off-by: Zach Brown <zach.brown(a)ni.com> Signed-off-by: Jacek Anaszewski <j.anaszewski(a)samsung.com> Cc: Willy Tarreau <w(a)1wt.eu> Cc: Vlastimil Babka <vbabka(a)suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/leds/led-triggers.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) --- a/drivers/leds/led-triggers.c +++ b/drivers/leds/led-triggers.c @@ -88,21 +88,23 @@ ssize_t led_trigger_show(struct device * down_read(&led_cdev->trigger_lock); if (!led_cdev->trigger) - len += sprintf(buf+len, "[none] "); + len += scnprintf(buf+len, PAGE_SIZE - len, "[none] "); else - len += sprintf(buf+len, "none "); + len += scnprintf(buf+len, PAGE_SIZE - len, "none "); list_for_each_entry(trig, &trigger_list, next_trig) { if (led_cdev->trigger && !strcmp(led_cdev->trigger->name, trig->name)) - len += sprintf(buf+len, "[%s] ", trig->name); + len += scnprintf(buf+len, PAGE_SIZE - len, "[%s] ", + trig->name); else - len += sprintf(buf+len, "%s ", trig->name); + len += scnprintf(buf+len, PAGE_SIZE - len, "%s ", + trig->name); } up_read(&led_cdev->trigger_lock); up_read(&triggers_list_lock); - len += sprintf(len+buf, "\n"); + len += scnprintf(len+buf, PAGE_SIZE - len, "\n"); return len; } EXPORT_SYMBOL_GPL(led_trigger_show); Patches currently in stable-queue which might be from nathan.sullivan(a)ni.com are queue-4.4/leds-do-not-overflow-sysfs-buffer-in-led_trigger_show.patch

7 years, 3 months

1
0
0 0

Patch "leds: do not overflow sysfs buffer in led_trigger_show" 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 leds: do not overflow sysfs buffer in led_trigger_show 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: leds-do-not-overflow-sysfs-buffer-in-led_trigger_show.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 3b9b95363c45365d606ad4bbba16acca75fdf6d3 Mon Sep 17 00:00:00 2001 From: Nathan Sullivan <nathan.sullivan(a)ni.com> Date: Mon, 15 Aug 2016 17:20:14 -0500 Subject: leds: do not overflow sysfs buffer in led_trigger_show From: Nathan Sullivan <nathan.sullivan(a)ni.com> commit 3b9b95363c45365d606ad4bbba16acca75fdf6d3 upstream. Per the documentation, use scnprintf instead of sprintf to ensure there is never more than PAGE_SIZE bytes of trigger names put into the buffer. Signed-off-by: Nathan Sullivan <nathan.sullivan(a)ni.com> Signed-off-by: Zach Brown <zach.brown(a)ni.com> Signed-off-by: Jacek Anaszewski <j.anaszewski(a)samsung.com> Cc: Willy Tarreau <w(a)1wt.eu> Cc: Vlastimil Babka <vbabka(a)suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/leds/led-triggers.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) --- a/drivers/leds/led-triggers.c +++ b/drivers/leds/led-triggers.c @@ -78,21 +78,23 @@ ssize_t led_trigger_show(struct device * down_read(&led_cdev->trigger_lock); if (!led_cdev->trigger) - len += sprintf(buf+len, "[none] "); + len += scnprintf(buf+len, PAGE_SIZE - len, "[none] "); else - len += sprintf(buf+len, "none "); + len += scnprintf(buf+len, PAGE_SIZE - len, "none "); list_for_each_entry(trig, &trigger_list, next_trig) { if (led_cdev->trigger && !strcmp(led_cdev->trigger->name, trig->name)) - len += sprintf(buf+len, "[%s] ", trig->name); + len += scnprintf(buf+len, PAGE_SIZE - len, "[%s] ", + trig->name); else - len += sprintf(buf+len, "%s ", trig->name); + len += scnprintf(buf+len, PAGE_SIZE - len, "%s ", + trig->name); } up_read(&led_cdev->trigger_lock); up_read(&triggers_list_lock); - len += sprintf(len+buf, "\n"); + len += scnprintf(len+buf, PAGE_SIZE - len, "\n"); return len; } EXPORT_SYMBOL_GPL(led_trigger_show); Patches currently in stable-queue which might be from nathan.sullivan(a)ni.com are queue-3.18/leds-do-not-overflow-sysfs-buffer-in-led_trigger_show.patch

7 years, 3 months

1
0
0 0

Re: [PATCH 00/45] arm Spectre fix backport review for LTS 4.9

by Pali Rohár

On Thursday 08 March 2018 13:27:15 Pavel Machek wrote: > Hi! > > > Resent without non-upstream patches. > > > > This backport patchset fixed the spectre issue, it's original branch: > > https://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git/log/?h=kpti > > A few dependency or fixingpatches are also picked up, if they are necessary > > and no functional changes. > > > > No bug found from kernelci.org and lkft testing. It also could be gotten from: > > > > git://git.linaro.org/kernel/linux-linaro-stable.git v4.9-spectre-upstream-only > > > > Comments are appreciated! > > Not entirely related to this patched, but... I have few older ARM > boards here, and Nokia N9000 I really care about. > > AFAICT Meltdown is arm64 only? IIRC ARMv7 is not affected by meltdown. > Spectre affects the older boards, too, right? Was there any work done > on that? cpuinfo says "ARMv7" for N900. I remember that I saw some spectre patches for ARMv7 on LKML. In general for ARMv7 it is problematic as mitigation needs to change IBE bit which is not possible on OMAP HS devices. But for Nokia N900 there is special code which do it via smc instruction (function rx51_secure_update_aux_cr(), see also nokia_n900_legacy_init()). -- Pali Rohár pali.rohar(a)gmail.com

7 years, 3 months

3
2
0 0

[PATCH resend] module: propagate error in modules_open()

by Leon Yu

otherwise kernel can oops later in seq_release() due to dereferencing null file->private_data which is only set if seq_open() succeeds. BUG: unable to handle kernel NULL pointer dereference at 0000000000000000 IP: seq_release+0xc/0x30 Call Trace: close_pdeo+0x37/0xd0 proc_reg_release+0x5d/0x60 __fput+0x9d/0x1d0 ____fput+0x9/0x10 task_work_run+0x75/0x90 do_exit+0x252/0xa00 do_group_exit+0x36/0xb0 SyS_exit_group+0xf/0x10 Fixes: 516fb7f2e73d ("/proc/module: use the same logic as /proc/kallsyms for address exposure") Cc: Jessica Yu <jeyu(a)kernel.org> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: stable(a)vger.kernel.org # 4.15+ Signed-off-by: Leon Yu <chianglungyu(a)gmail.com> --- kernel/module.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/module.c b/kernel/module.c index ad2d420024f6..e42764acedb4 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -4228,7 +4228,7 @@ static int modules_open(struct inode *inode, struct file *file) m->private = kallsyms_show_value() ? NULL : (void *)8ul; } - return 0; + return err; } static const struct file_operations proc_modules_operations = { -- 2.16.2

7 years, 3 months

2
1
0 0

[PATCH] staging: android: ashmem: Remove deadlock

by Paul Lawrence

Regression introduced in commit ce8a3a9e76d0193e2e8d74a06d275b3c324ca652 ("staging: android: ashmem: Fix a race condition in pin ioctls") causing deadlock. No need to hold ashmem_mutex while copying from user Stacks are: ashmem_mmap+0x53/0x400 drivers/staging/android/ashmem.c:379 mmap_region+0x7dd/0xfd0 mm/mmap.c:1694 do_mmap+0x57b/0xbe0 mm/mmap.c:1473 and lock_acquire+0x12e/0x410 kernel/locking/lockdep.c:3756 __might_fault+0x14a/0x1d0 mm/memory.c:4014 copy_from_user arch/x86/include/asm/uaccess.h:705 [inline] ashmem_pin_unpin drivers/staging/android/ashmem.c:719 [inline] Signed-off-by: Paul Lawrence <paullawrence(a)google.com> Cc: <stable(a)vger.kernel.org> # 4.9.x Cc: <stable(a)vger.kernel.org> # 4.4.x Cc: <stable(a)vger.kernel.org> # 3.18.x: ce8a3a9e76d01 Cc: <stable(a)vger.kernel.org> # 3.18.x Cc: Ben Hutchings <ben(a)decadent.org.uk> --- drivers/staging/android/ashmem.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/staging/android/ashmem.c b/drivers/staging/android/ashmem.c index 6dbba5aff191..8c55706c2cfa 100644 --- a/drivers/staging/android/ashmem.c +++ b/drivers/staging/android/ashmem.c @@ -702,16 +702,14 @@ static int ashmem_pin_unpin(struct ashmem_area *asma, unsigned long cmd, size_t pgstart, pgend; int ret = -EINVAL; + if (unlikely(copy_from_user(&pin, p, sizeof(pin)))) + return -EFAULT; + mutex_lock(&ashmem_mutex); if (unlikely(!asma->file)) goto out_unlock; - if (unlikely(copy_from_user(&pin, p, sizeof(pin)))) { - ret = -EFAULT; - goto out_unlock; - } - /* per custom, you can pass zero for len to mean "everything onward" */ if (!pin.len) pin.len = PAGE_ALIGN(asma->size) - pin.offset; -- 2.16.2.395.g2e18187dfd-goog

7 years, 3 months

3
2
0 0

Re: 4.14.18 -> 4.14.24 - almost all guests hanged

by 王金浦

2018-03-07 21:29 GMT+01:00 Nikola Ciprich <nikola.ciprich(a)linuxbox.cz>: > Hi, > >> > > I'd like to report that when upgrading our cluster from 4.14.18 to >> > > 4.14.24-rc1 (with live guests migration), almost none of guests survived.. >> > What's your hardware setup, intel with IBPB enabled microcode? >> Intel(R) Xeon(R) CPU E5-2650 0 @ 2.00GHz >> >> therefore I suppose no IBPB (at least meltdown checker reports so) >> >> >> > Does guests hang right after live migration? >> yes, just tried it. >> >> >> > >> > Are you able to reproduce the problem, does it work with latest upstream? >> yup, so I'm able to reproduce quickly. I'll revert the cluster to 4.14.18 now, >> but setup test system just afterwards, so and test the patch you've proposed. >> >> > >> > Not sure it helps, but following patch is missing in 4.14.24 >> > >> > commit 37b95951c58fdf08dc10afa9d02066ed9f176fb5 upstream. >> > >> > kvm_valid_sregs() should use X86_CR0_PG and X86_CR4_PAE to check bit >> > status rather than X86_CR0_PG_BIT and X86_CR4_PAE_BIT. This patch is >> > to fix it. >> > >> > Fixes: f29810335965a(KVM/x86: Check input paging mode when cs.l is set) >> > Reported-by: Jeremi Piotrowski <jeremi.piotrowski(a)gmail.com> >> > Cc: Paolo Bonzini <pbonzini(a)redhat.com> >> > Cc: Radim Krčmář <rkrcmar(a)redhat.com> >> > Signed-off-by: Tianyu Lan <Tianyu.Lan(a)microsoft.com> >> > Signed-off-by: Radim Krčmář <rkrcmar(a)redhat.com> >> >> I'll test and report. > > so indeed, this one on top of 4.14.24-rc1 fixes the migration for me. > Greg, could you queue this one up please? > > Jack, thanks for the hint! > BR > nik Hi Nik, Thanks for testing and let we know, the patch is already queued in 4.14.25-rc1 by Greg. There're some other KVM fixes and performance enhancement. Regards, Jack > > > >> >> n. >> >> >> > >> > Regards, >> > Jack >> > > >> > > I noticed that most of them got stuck in "paused" state without >> > > possibility to resume (virsh just reported guest cannot be continued and >> > > needs to be rebooted). >> > > >> > > in dmesg, lots of following messages appeared: >> > > >> > > [ 116.593508] device vnet0 entered promiscuous mode >> > > [ 124.143532] *** Guest State *** >> > > [ 124.143594] CR0: actual=0x0000000000000030, shadow=0x0000000060000010, gh_mask=fffffffffffffff7 >> > > [ 124.143668] CR4: actual=0x0000000000002050, shadow=0x0000000000000000, gh_mask=ffffffffffffe871 >> > > [ 124.143871] CR3 = 0x00000000feffc000 >> > > [ 124.143984] RSP = 0xffffffff82003e98 RIP = 0xffffffff816df002 >> > > [ 124.144102] RFLAGS=0x00000246 DR7 = 0x0000000000000400 >> > > [ 124.144221] Sysenter RSP=0000000000000000 CS:RIP=0000:0000000000000000 >> > > [ 124.144341] CS: sel=0xf000, attr=0x0009b, limit=0x0000ffff, base=0x00000000ffff0000 >> > > [ 124.144516] DS: sel=0x0000, attr=0x00093, limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.144692] SS: sel=0x0000, attr=0x00093, limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.144907] ES: sel=0x0000, attr=0x00093, limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.145089] FS: sel=0x0000, attr=0x00093, limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.145272] GS: sel=0x0000, attr=0x00093, limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.145447] GDTR: limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.145626] LDTR: sel=0x0000, attr=0x00082, limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.145814] IDTR: limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.145995] TR: sel=0x0000, attr=0x0008b, limit=0x0000ffff, base=0x0000000000000000 >> > > [ 124.146173] EFER = 0x0000000000000000 PAT = 0x0007040600070406 >> > > [ 124.146292] DebugCtl = 0x0000000000000000 DebugExceptions = 0x0000000000000000 >> > > [ 124.146466] Interruptibility = 00000000 ActivityState = 00000000 >> > > [ 124.146579] *** Host State *** >> > > [ 124.146687] RIP = 0xffffffffa046a817 RSP = 0xffffc900200a7cb8 >> > > [ 124.146832] CS=0010 SS=0018 DS=0000 ES=0000 FS=0000 GS=0000 TR=0040 >> > > [ 124.146961] FSBase=00007fe82eff7700 GSBase=ffff881fffb40000 TRBase=fffffe00000df000 >> > > [ 124.147144] GDTBase=fffffe00000dd000 IDTBase=fffffe0000000000 >> > > [ 124.147262] CR0=0000000080050033 CR3=0000001f5b8fe004 CR4=00000000000626e0 >> > > [ 124.147381] Sysenter RSP=fffffe00000de200 CS:RIP=0010:ffffffff81801f60 >> > > [ 124.147499] EFER = 0x0000000000000d01 PAT = 0x0407050600070106 >> > > [ 124.147614] *** Control State *** >> > > [ 124.147734] PinBased=0000007f CPUBased=96a1e9fa SecondaryExec=000004f2 >> > > [ 124.147849] EntryControls=0000d1ff ExitControls=002fefff >> > > [ 124.147965] ExceptionBitmap=00060042 PFECmask=00000000 PFECmatch=00000000 >> > > [ 124.148085] VMEntry: intr_info=80000081 errcode=00000000 ilen=00000000 >> > > [ 124.148201] VMExit: intr_info=00000000 errcode=00000000 ilen=00000000 >> > > [ 124.148318] reason=80000021 qualification=0000000000000000 >> > > [ 124.148432] IDTVectoring: info=00000000 errcode=00000000 >> > > [ 124.148545] TSC Offset = 0xffed7296fb06bc34 >> > > [ 124.148655] TPR Threshold = 0x00 >> > > [ 124.148770] EPT pointer = 0x0000001f1a0af01e >> > > [ 124.148882] PLE Gap=00000080 Window=00001000 >> > > [ 124.148995] Virtual processor ID = 0x0001 >> > > >> > > (never seen anything like that) >> > > >> > > I haven't yet went through all patches between those two versions, so don't >> > > have any suspicion yet.. If anyone recognizes this as known problem, please >> > > let me know.. >> > > >> > > I'm going to try whether I'm able to reproduce the problem. >> > > >> > > BR >> > > >> > > nik >> > >> >> -- >> ------------------------------------- >> Ing. Nikola CIPRICH >> LinuxBox.cz, s.r.o. >> 28.rijna 168, 709 00 Ostrava >> >> tel.: +420 591 166 214 >> fax: +420 596 621 273 >> mobil: +420 777 093 799 >> www.linuxbox.cz >> >> mobil servis: +420 737 238 656 >> email servis: servis(a)linuxbox.cz >> ------------------------------------- >> > > -- > ------------------------------------- > Ing. Nikola CIPRICH > LinuxBox.cz, s.r.o. > 28. rijna 168, 709 00 Ostrava > > tel.: +420 591 166 214 > fax: +420 596 621 273 > mobil: +420 777 093 799 > > www.linuxbox.cz > > mobil servis: +420 737 238 656 > email servis: servis(a)linuxbox.cz > -------------------------------------

7 years, 3 months

1
0
0 0

[stable-4.14 0/2] 2 kvm fix for 4.14

by Jack Wang

From: Jack Wang <jinpu.wang(a)profitbricks.com> Hi Greg, I noticed 2 fixes for kvm are missing in your queue-4.14, both are bugfix, can be cherry pick cleanly. The patch from Tianyu should close bug below, also included in 3.16 https://bugzilla.kernel.org/show_bug.cgi?id=198991 Eric Biggers (1): KVM/x86: remove WARN_ON() for when vm_munmap() fails Tianyu Lan (1): KVM/x86: Fix wrong macro references of X86_CR0_PG_BIT and X86_CR4_PAE_BIT in kvm_valid_sregs() arch/x86/kvm/x86.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) -- 2.7.4

7 years, 3 months

3
4
0 0

[PATCH] kbuild: Handle builtin dtb files containing hyphens

by James Hogan

On dtb files which contain hyphens, the dt_S_dtb command to build the dtb.S files (which allow DTB files to be built into the kernel) results in errors like the following: bcm3368-netgear-cvg834g.dtb.S: Assembler messages: bcm3368-netgear-cvg834g.dtb.S:5: Error: : no such section bcm3368-netgear-cvg834g.dtb.S:5: Error: junk at end of line, first unrecognized character is `-' bcm3368-netgear-cvg834g.dtb.S:6: Error: unrecognized opcode `__dtb_bcm3368-netgear-cvg834g_begin:' bcm3368-netgear-cvg834g.dtb.S:8: Error: unrecognized opcode `__dtb_bcm3368-netgear-cvg834g_end:' bcm3368-netgear-cvg834g.dtb.S:9: Error: : no such section bcm3368-netgear-cvg834g.dtb.S:9: Error: junk at end of line, first unrecognized character is `-' This is due to the hyphen being used in symbol names. Replace all hyphens with underscores in the dt_S_dtb command to avoid this problem. Quite a lot of dts files have hyphens, but its only a problem on MIPS where such files can be built into the kernel. For example when CONFIG_DT_NETGEAR_CVG834G=y, or on BMIPS kernels when the dtbs target is used (in the latter case it admitedly shouldn't really build all the dtb.o files, but thats a separate issue). Fixes: 695835511f96 ("MIPS: BMIPS: rename bcm96358nb4ser to bcm6358-neufbox4-sercom") Signed-off-by: James Hogan <jhogan(a)kernel.org> Cc: Rob Herring <robh+dt(a)kernel.org> Cc: Frank Rowand <frowand.list(a)gmail.com> Cc: Masahiro Yamada <yamada.masahiro(a)socionext.com> Cc: Michal Marek <michal.lkml(a)markovi.net> Cc: Ralf Baechle <ralf(a)linux-mips.org> Cc: Florian Fainelli <f.fainelli(a)gmail.com> Cc: Kevin Cernekee <cernekee(a)gmail.com> Cc: devicetree(a)vger.kernel.org Cc: linux-kbuild(a)vger.kernel.org Cc: linux-mips(a)linux-mips.org Cc: <stable(a)vger.kernel.org> # 4.9+ --- scripts/Makefile.lib | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index 5589bae34af6..a6f538b31ad6 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -297,11 +297,11 @@ cmd_dt_S_dtb= \ echo '\#include <asm-generic/vmlinux.lds.h>'; \ echo '.section .dtb.init.rodata,"a"'; \ echo '.balign STRUCT_ALIGNMENT'; \ - echo '.global __dtb_$(*F)_begin'; \ - echo '__dtb_$(*F)_begin:'; \ + echo '.global __dtb_$(subst -,_,$(*F))_begin'; \ + echo '__dtb_$(subst -,_,$(*F))_begin:'; \ echo '.incbin "$<" '; \ - echo '__dtb_$(*F)_end:'; \ - echo '.global __dtb_$(*F)_end'; \ + echo '__dtb_$(subst -,_,$(*F))_end:'; \ + echo '.global __dtb_$(subst -,_,$(*F))_end'; \ echo '.balign STRUCT_ALIGNMENT'; \ ) > $@ -- 2.13.6

7 years, 3 months

2
2
0 0

[PATCH AUTOSEL for 3.18 01/53] usb: gadget: dummy_hcd: Fix wrong power status bit clear/reset in dummy_hub_control()

by Sasha Levin

From: Yuyang Du <yuyang.du(a)intel.com> [ Upstream commit 9f20dfb44d03745d0d3cef2ffb3abf8d8024fa61 ] This fixes the commit: 1cd8fd2887e1 ("usb: gadget: dummy_hcd: add SuperSpeed support"). In the case of ClearPortFeature and USB_PORT_FEAT_POWER, simply clear the right bit regardless of what the wValue is. Acked-by: Alan Stern <stern(a)rowland.harvard.edu> Signed-off-by: Yuyang Du <yuyang.du(a)intel.com> Signed-off-by: Felipe Balbi <felipe.balbi(a)linux.intel.com> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> --- drivers/usb/gadget/udc/dummy_hcd.c | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-) diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c index 783715a9e399..215441690d32 100644 --- a/drivers/usb/gadget/udc/dummy_hcd.c +++ b/drivers/usb/gadget/udc/dummy_hcd.c @@ -2001,16 +2001,13 @@ static int dummy_hub_control( } break; case USB_PORT_FEAT_POWER: - if (hcd->speed == HCD_USB3) { - if (dum_hcd->port_status & USB_PORT_STAT_POWER) - dev_dbg(dummy_dev(dum_hcd), - "power-off\n"); - } else - if (dum_hcd->port_status & - USB_SS_PORT_STAT_POWER) - dev_dbg(dummy_dev(dum_hcd), - "power-off\n"); - /* FALLS THROUGH */ + dev_dbg(dummy_dev(dum_hcd), "power-off\n"); + if (hcd->speed == HCD_USB3) + dum_hcd->port_status &= ~USB_SS_PORT_STAT_POWER; + else + dum_hcd->port_status &= ~USB_PORT_STAT_POWER; + set_link_state(dum_hcd); + break; default: dum_hcd->port_status &= ~(1 << wValue); set_link_state(dum_hcd); @@ -2181,14 +2178,13 @@ static int dummy_hub_control( if ((dum_hcd->port_status & USB_SS_PORT_STAT_POWER) != 0) { dum_hcd->port_status |= (1 << wValue); - set_link_state(dum_hcd); } } else if ((dum_hcd->port_status & USB_PORT_STAT_POWER) != 0) { dum_hcd->port_status |= (1 << wValue); - set_link_state(dum_hcd); } + set_link_state(dum_hcd); } break; case GetPortErrorCount: -- 2.14.1

7 years, 3 months

1
52
0 0

[PATCH AUTOSEL for 4.4 001/101] usb: dwc2: Make sure we disconnect the gadget state

by Sasha Levin

From: John Stultz <john.stultz(a)linaro.org> [ Upstream commit dad3f793f20fbb5c0c342f0f5a0bdf69a4d76089 ] I had seen some odd behavior with HiKey's usb-gadget interface that I finally seemed to have chased down. Basically every other time I plugged in the OTG port, the gadget interface would properly initialize. The other times, I'd get a big WARN_ON in dwc2_hsotg_init_fifo() about the fifo_map not being clear. Ends up if we don't disconnect the gadget state, the fifo-map doesn't get cleared properly, which causes WARN_ON messages and also results in the device not properly being setup as a gadget every other time the OTG port is connected. So this patch adds a call to dwc2_hsotg_disconnect() in the reset path so the state is properly cleared. With it, the gadget interface initializes properly on every plug in. Cc: Wei Xu <xuwei5(a)hisilicon.com> Cc: Guodong Xu <guodong.xu(a)linaro.org> Cc: Amit Pundir <amit.pundir(a)linaro.org> Cc: Rob Herring <robh+dt(a)kernel.org> Cc: John Youn <johnyoun(a)synopsys.com> Cc: Douglas Anderson <dianders(a)chromium.org> Cc: Chen Yu <chenyu56(a)huawei.com> Cc: Felipe Balbi <felipe.balbi(a)linux.intel.com> Cc: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Cc: linux-usb(a)vger.kernel.org Acked-by: John Youn <johnyoun(a)synopsys.com> Signed-off-by: John Stultz <john.stultz(a)linaro.org> Signed-off-by: Felipe Balbi <felipe.balbi(a)linux.intel.com> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> --- drivers/usb/dwc2/hcd.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/usb/dwc2/hcd.c b/drivers/usb/dwc2/hcd.c index 571c21727ff9..88bd950665fa 100644 --- a/drivers/usb/dwc2/hcd.c +++ b/drivers/usb/dwc2/hcd.c @@ -1385,6 +1385,7 @@ static void dwc2_conn_id_status_change(struct work_struct *work) dwc2_core_init(hsotg, false, -1); dwc2_enable_global_interrupts(hsotg); spin_lock_irqsave(&hsotg->lock, flags); + dwc2_hsotg_disconnect(hsotg); dwc2_hsotg_core_init_disconnected(hsotg, false); spin_unlock_irqrestore(&hsotg->lock, flags); dwc2_hsotg_core_connect(hsotg); -- 2.14.1

7 years, 3 months

1
100
0 0

Patch "platform/x86: dell-laptop: fix kbd_get_state's request value" has been added to the 4.15-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled platform/x86: dell-laptop: fix kbd_get_state's request value to the 4.15-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: platform-x86-dell-laptop-fix-kbd_get_state-s-request-value.patch and it can be found in the queue-4.15 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 eca39e7f0cdb9bde4003a29149fa695e876c6f73 Mon Sep 17 00:00:00 2001 From: Laszlo Toth <laszlth(a)gmail.com> Date: Tue, 13 Feb 2018 21:43:43 +0100 Subject: platform/x86: dell-laptop: fix kbd_get_state's request value MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Laszlo Toth <laszlth(a)gmail.com> commit eca39e7f0cdb9bde4003a29149fa695e876c6f73 upstream. Commit 9862b43624a5 ("platform/x86: dell-laptop: Allocate buffer on heap rather than globally") broke one request, changed it back to the original value. Tested on a Dell E6540, backlight came back. Fixes: 9862b43624a5 ("platform/x86: dell-laptop: Allocate buffer on heap rather than globally") Signed-off-by: Laszlo Toth <laszlth(a)gmail.com> Reviewed-by: Pali Rohár <pali.rohar(a)gmail.com> Reviewed-by: Mario Limonciello <mario.limonciello(a)dell.com> Signed-off-by: Andy Shevchenko <andriy.shevchenko(a)linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/platform/x86/dell-laptop.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/platform/x86/dell-laptop.c +++ b/drivers/platform/x86/dell-laptop.c @@ -1254,7 +1254,7 @@ static int kbd_get_state(struct kbd_stat struct calling_interface_buffer buffer; int ret; - dell_fill_request(&buffer, 0, 0, 0, 0); + dell_fill_request(&buffer, 0x1, 0, 0, 0); ret = dell_send_request(&buffer, CLASS_KBD_BACKLIGHT, SELECT_KBD_BACKLIGHT); if (ret) Patches currently in stable-queue which might be from laszlth(a)gmail.com are queue-4.15/platform-x86-dell-laptop-fix-kbd_get_state-s-request-value.patch

7 years, 3 months

1
0
0 0

Re: [PATCH] ubi: Reject MLC NAND

by Pavel Machek

On Wed 2018-03-07 22:11:13, David Woodhouse wrote: > > > On Wed, 2018-03-07 at 14:08 -0800, Steve deRosier wrote: > > > > To clarify one thing: the reason for this is MLC has actually never > > been supported, nor worked properly. The fact that it kinda worked was > > incidental and the cause of major problems for people due to that not > > being clear. This patch only makes it explicit and avoids people > > mistakenly trying to use UBIFS on MLC flash and risking their data and > > products. To me, that's what's important. > > > > This is an important patch, even if all it does is keep people from > > loosing data. It also changes the conversation from "I have a > > corrupted UBIFS device, BTW it's on MLC..." to "What can we do to get > > UBIFS to work on MLC". Well, for -stable I'd suggest printk(KERN_ALERT ...) but keep the system running. > This is a bug fix. > > UBI on MLC never worked. It was a bug that we ever permitted it. This > is now fixed. Yeah, well, so lets say I have a working hardware (maybe using read-only UBI on MLC), update to next stable kernel, and now kernel refuses to see the partition. I'll certainly not consider this patch a bug fix. Removing support for hardware that "only works by mistake" may be good idea, but maybe it is slightly too surprising for a -stable. Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

7 years, 3 months

3
2
0 0

Re: [PATCH] kbuild: Handle builtin dtb files containing hyphens

by Frank Rowand

On 03/07/18 12:25, James Hogan wrote: > On Wed, Mar 07, 2018 at 12:11:41PM -0800, Frank Rowand wrote: >> I initially misread the patch description (and imagined an entirely >> different problem). >> >> >> On 03/07/18 06:06, James Hogan wrote: >>> On dtb files which contain hyphens, the dt_S_dtb command to build the> dtb.S files (which allow DTB files to be built into the kernel) results> in errors like the following:> > bcm3368-netgear-cvg834g.dtb.S: Assembler messages:> bcm3368-netgear-cvg834g.dtb.S:5: Error: : no such section> bcm3368-netgear-cvg834g.dtb.S:5: Error: junk at end of line, first unrecognized character is `-'> bcm3368-netgear-cvg834g.dtb.S:6: Error: unrecognized opcode `__dtb_bcm3368-netgear-cvg834g_begin:'> bcm3368-netgear-cvg834g.dtb.S:8: Error: unrecognized opcode `__dtb_bcm3368-netgear-cvg834g_end:'> bcm3368-netgear-cvg834g.dtb.S:9: Error: : no such section> bcm3368-netgear-cvg834g.dtb.S:9: Error: junk at end of line, first unrecognized character is `-' >> Please replace the following section: >> >>> This is due to the hyphen being used in symbol names. Replace all >>> hyphens >>> with underscores in the dt_S_dtb command to avoid this problem. >>> >>> Quite a lot of dts files have hyphens, but its only a problem on MIPS >>> where such files can be built into the kernel. For example when >>> CONFIG_DT_NETGEAR_CVG834G=y, or on BMIPS kernels when the dtbs target is >>> used (in the latter case it admitedly shouldn't really build all the >>> dtb.o files, but thats a separate issue). >> >> with: >> >> cmd_dt_S_dtb constructs the assembly source to incorporate a devicetree >> FDT (that is, the .dtb file) as binary data in the kernel image. >> This assembly source contains labels before and after the binary data. >> The label names incorporate the file name of the corresponding .dtb >> file. Hyphens are not legal characters in labels, so transform all >> hyphens from the file name to underscores when constructing the labels. > > Thanks, that is clearer. > > I'll keep the paragraph about MIPS and the example configuration though, > as I think its important information to reproduce the problem, and to > justify why it wouldn't be appropriate to just rename the files (which > was my first reaction). Other than the part that says "its only a problem on MIPS". That is pedantically correct because no other architecture (that I am aware of, not that I searched) currently has a devicetree source file name with a hyphen in it, where that file is compiled into the kernel as an asm file. But it is potentially a problem on any architecture to it is misleading to label it as MIPS only. > >> Reviewed-by: Frank Rowand <frowand.list(a)gmail.com> > > Thanks > James >

7 years, 3 months

1
0
0 0

Re: [PATCH] ubi: Reject MLC NAND

by Boris Brezillon

On Wed, 7 Mar 2018 22:43:42 +0100 Pavel Machek <pavel(a)ucw.cz> wrote: > On Wed 2018-03-07 09:01:16, Richard Weinberger wrote: > > Pavel, > > > > Am Mittwoch, 7. März 2018, 00:18:05 CET schrieb Pavel Machek: > > > On Sat 2018-03-03 11:45:54, Richard Weinberger wrote: > > > > While UBI and UBIFS seem to work at first sight with MLC NAND, you will > > > > most likely lose all your data upon a power-cut or due to read/write > > > > disturb. > > > > In order to protect users from bad surprises, refuse to attach to MLC > > > > NAND. > > > > > > > > Cc: stable(a)vger.kernel.org > > > > > > That sounds like _really_ bad idea for stable. All it does is it > > > removes support for hardware that somehow works. > > > > MLC is not supported and does not work. Full stop. > > If someone manages to get it somehow work, either with hardware or software > > hacks they are on their own. > > Having it in stable is the only chance we have to get it into vendor > > kernels. > > Can you show how it meets the stable kernel criteria? They are > documented in tree. This should not be in stable. > > And I'd like to see changelog improved. Real reason MLC is not > supported is upper/lower page parts on MLC. And real fix to work with > bigger pages in UBI. Come on! Don't you think this would have been fixed already if it was that easy?! Have you looked at an MLC datasheet to see how paired pages are combined? If you had you would now that paired pages are almost all the time not contiguous, thus preventing the trick you're suggesting here. Please document yourself before doing such presumptuous statements (you can have a look at these slides if you want some details about why this is not so simple [1]). I'm definitely not saying supporting MLC NANDs in Linux is impossible, and if you're interested in working on this topic I'd be happy to help. But please don't block this patch without understanding what supporting MLC NANDs implies. [1]https://events.static.linuxfound.org/sites/events/files/slides/ubi-mlc.pdf -- Boris Brezillon, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering https://bootlin.com

7 years, 3 months

1
0
0 0

Re: [PATCH] ubi: Reject MLC NAND

by Steve deRosier

Hi Pavel, On Wed, Mar 7, 2018 at 1:43 PM, Pavel Machek <pavel(a)ucw.cz> wrote: > On Wed 2018-03-07 09:01:16, Richard Weinberger wrote: >> Pavel, >> >> Am Mittwoch, 7. März 2018, 00:18:05 CET schrieb Pavel Machek: >> > On Sat 2018-03-03 11:45:54, Richard Weinberger wrote: >> > > While UBI and UBIFS seem to work at first sight with MLC NAND, you will >> > > most likely lose all your data upon a power-cut or due to read/write >> > > disturb. >> > > In order to protect users from bad surprises, refuse to attach to MLC >> > > NAND. >> > > >> > > Cc: stable(a)vger.kernel.org >> > >> > That sounds like _really_ bad idea for stable. All it does is it >> > removes support for hardware that somehow works. >> >> MLC is not supported and does not work. Full stop. >> If someone manages to get it somehow work, either with hardware or software >> hacks they are on their own. >> Having it in stable is the only chance we have to get it into vendor >> kernels. > > Can you show how it meets the stable kernel criteria? They are > documented in tree. This should not be in stable. > > And I'd like to see changelog improved. Real reason MLC is not > supported is upper/lower page parts on MLC. And real fix to work with > bigger pages in UBI. > To clarify one thing: the reason for this is MLC has actually never been supported, nor worked properly. The fact that it kinda worked was incidental and the cause of major problems for people due to that not being clear. This patch only makes it explicit and avoids people mistakenly trying to use UBIFS on MLC flash and risking their data and products. To me, that's what's important. This is an important patch, even if all it does is keep people from loosing data. It also changes the conversation from "I have a corrupted UBIFS device, BTW it's on MLC..." to "What can we do to get UBIFS to work on MLC". I don't know what the stable criteria is with re: to this patch. But what I do know is if it doesn't go back into the various stables, there will be manufacturers who will continue to try to use UBIFS on MLC in ignorance for the next several years until the current stable kernels EOL, despite there being a known patch that would make it immediately obvious they shouldn't. Thanks, - Steve

7 years, 3 months

1
0
0 0

Re: [PATCH] staging: android: ashmem: Remove deadlock

by Greg Kroah-Hartman

On Wed, Mar 07, 2018 at 02:02:13PM -0800, Paul Lawrence wrote: > Great! We need to make sure this gets backported to 4.4 and 4.9, and to > 3.18 with the original dependency, please. That will happen when it lands in Linus's tree, which should be later this week if all goes well. thanks, greg k-h

7 years, 3 months

1
0
0 0

[PATCH 4.4,4.9] x86/apic/vector: Handle legacy irq data correctly

by Ben Hutchings

From: Thomas Gleixner <tglx(a)linutronix.de> The backport of upstream commit 45d55e7bac40 ("x86/apic/vector: Fix off by one in error path") missed to fixup the legacy interrupt data which is not longer available upstream. Handle legacy irq data correctly by clearing the legacy storage to prevent use after free. Fixes: 7fd133539289 ("x86/apic/vector: Fix off by one in error path") - 4.4.y Fixes: c557481a9491 ("x86/apic/vector: Fix off by one in error path") - 4.9.y Reported-by: Ben Hutchings <ben.hutchings(a)codethink.co.uk> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: Ben Hutchings <ben.hutchings(a)codethink.co.uk> --- arch/x86/kernel/apic/vector.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/arch/x86/kernel/apic/vector.c b/arch/x86/kernel/apic/vector.c index a41e523536a2..592e260ba05b 100644 --- a/arch/x86/kernel/apic/vector.c +++ b/arch/x86/kernel/apic/vector.c @@ -91,8 +91,12 @@ out_data: return NULL; } -static void free_apic_chip_data(struct apic_chip_data *data) +static void free_apic_chip_data(unsigned int virq, struct apic_chip_data *data) { +#ifdef CONFIG_X86_IO_APIC + if (virq < nr_legacy_irqs()) + legacy_irq_data[virq] = NULL; +#endif if (data) { free_cpumask_var(data->domain); free_cpumask_var(data->old_domain); @@ -316,11 +320,7 @@ static void x86_vector_free_irqs(struct irq_domain *domain, apic_data = irq_data->chip_data; irq_domain_reset_irq_data(irq_data); raw_spin_unlock_irqrestore(&vector_lock, flags); - free_apic_chip_data(apic_data); -#ifdef CONFIG_X86_IO_APIC - if (virq + i < nr_legacy_irqs()) - legacy_irq_data[virq + i] = NULL; -#endif + free_apic_chip_data(virq + i, apic_data); } } } @@ -361,7 +361,7 @@ static int x86_vector_alloc_irqs(struct irq_domain *domain, unsigned int virq, err = assign_irq_vector_policy(virq + i, node, data, info); if (err) { irq_data->chip_data = NULL; - free_apic_chip_data(data); + free_apic_chip_data(virq + i, data); goto error; } } -- 2.15.0.rc0

7 years, 3 months

2
2
0 0

Patch "nvme-rdma: don't suppress send 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 nvme-rdma: don't suppress send 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: nvme-rdma-don-t-suppress-send-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 b4b591c87f2b0f4ebaf3a68d4f13873b241aa584 Mon Sep 17 00:00:00 2001 From: Sagi Grimberg <sagi(a)grimberg.me> Date: Thu, 23 Nov 2017 17:35:21 +0200 Subject: nvme-rdma: don't suppress send completions From: Sagi Grimberg <sagi(a)grimberg.me> commit b4b591c87f2b0f4ebaf3a68d4f13873b241aa584 upstream. The entire completions suppress mechanism is currently broken because the HCA might retry a send operation (due to dropped ack) after the nvme transaction has completed. In order to handle this, we signal all send completions and introduce a separate done handler for async events as they will be handled differently (as they don't include in-capsule data by definition). Signed-off-by: Sagi Grimberg <sagi(a)grimberg.me> Reviewed-by: Max Gurtovoy <maxg(a)mellanox.com> Signed-off-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/nvme/host/rdma.c | 54 ++++++++++++----------------------------------- 1 file changed, 14 insertions(+), 40 deletions(-) --- a/drivers/nvme/host/rdma.c +++ b/drivers/nvme/host/rdma.c @@ -88,7 +88,6 @@ enum nvme_rdma_queue_flags { struct nvme_rdma_queue { struct nvme_rdma_qe *rsp_ring; - atomic_t sig_count; int queue_size; size_t cmnd_capsule_len; struct nvme_rdma_ctrl *ctrl; @@ -521,7 +520,6 @@ static int nvme_rdma_alloc_queue(struct queue->cmnd_capsule_len = sizeof(struct nvme_command); queue->queue_size = queue_size; - atomic_set(&queue->sig_count, 0); queue->cm_id = rdma_create_id(&init_net, nvme_rdma_cm_handler, queue, RDMA_PS_TCP, IB_QPT_RC); @@ -1232,21 +1230,9 @@ static void nvme_rdma_send_done(struct i nvme_end_request(rq, req->status, req->result); } -/* - * We want to signal completion at least every queue depth/2. This returns the - * largest power of two that is not above half of (queue size + 1) to optimize - * (avoid divisions). - */ -static inline bool nvme_rdma_queue_sig_limit(struct nvme_rdma_queue *queue) -{ - int limit = 1 << ilog2((queue->queue_size + 1) / 2); - - return (atomic_inc_return(&queue->sig_count) & (limit - 1)) == 0; -} - static int nvme_rdma_post_send(struct nvme_rdma_queue *queue, struct nvme_rdma_qe *qe, struct ib_sge *sge, u32 num_sge, - struct ib_send_wr *first, bool flush) + struct ib_send_wr *first) { struct ib_send_wr wr, *bad_wr; int ret; @@ -1255,31 +1241,12 @@ static int nvme_rdma_post_send(struct nv sge->length = sizeof(struct nvme_command), sge->lkey = queue->device->pd->local_dma_lkey; - qe->cqe.done = nvme_rdma_send_done; - wr.next = NULL; wr.wr_cqe = &qe->cqe; wr.sg_list = sge; wr.num_sge = num_sge; wr.opcode = IB_WR_SEND; - wr.send_flags = 0; - - /* - * Unsignalled send completions are another giant desaster in the - * IB Verbs spec: If we don't regularly post signalled sends - * the send queue will fill up and only a QP reset will rescue us. - * Would have been way to obvious to handle this in hardware or - * at least the RDMA stack.. - * - * Always signal the flushes. The magic request used for the flush - * sequencer is not allocated in our driver's tagset and it's - * triggered to be freed by blk_cleanup_queue(). So we need to - * always mark it as signaled to ensure that the "wr_cqe", which is - * embedded in request's payload, is not freed when __ib_process_cq() - * calls wr_cqe->done(). - */ - if (nvme_rdma_queue_sig_limit(queue) || flush) - wr.send_flags |= IB_SEND_SIGNALED; + wr.send_flags = IB_SEND_SIGNALED; if (first) first->next = &wr; @@ -1329,6 +1296,12 @@ static struct blk_mq_tags *nvme_rdma_tag return queue->ctrl->tag_set.tags[queue_idx - 1]; } +static void nvme_rdma_async_done(struct ib_cq *cq, struct ib_wc *wc) +{ + if (unlikely(wc->status != IB_WC_SUCCESS)) + nvme_rdma_wr_error(cq, wc, "ASYNC"); +} + static void nvme_rdma_submit_async_event(struct nvme_ctrl *arg, int aer_idx) { struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(arg); @@ -1350,10 +1323,12 @@ static void nvme_rdma_submit_async_event cmd->common.flags |= NVME_CMD_SGL_METABUF; nvme_rdma_set_sg_null(cmd); + sqe->cqe.done = nvme_rdma_async_done; + ib_dma_sync_single_for_device(dev, sqe->dma, sizeof(*cmd), DMA_TO_DEVICE); - ret = nvme_rdma_post_send(queue, sqe, &sge, 1, NULL, false); + ret = nvme_rdma_post_send(queue, sqe, &sge, 1, NULL); WARN_ON_ONCE(ret); } @@ -1639,7 +1614,6 @@ static blk_status_t nvme_rdma_queue_rq(s struct nvme_rdma_request *req = blk_mq_rq_to_pdu(rq); struct nvme_rdma_qe *sqe = &req->sqe; struct nvme_command *c = sqe->data; - bool flush = false; struct ib_device *dev; blk_status_t ret; int err; @@ -1668,13 +1642,13 @@ static blk_status_t nvme_rdma_queue_rq(s goto err; } + sqe->cqe.done = nvme_rdma_send_done; + ib_dma_sync_single_for_device(dev, sqe->dma, sizeof(struct nvme_command), DMA_TO_DEVICE); - if (req_op(rq) == REQ_OP_FLUSH) - flush = true; err = nvme_rdma_post_send(queue, sqe, req->sge, req->num_sge, - req->mr->need_inval ? &req->reg_wr.wr : NULL, flush); + req->mr->need_inval ? &req->reg_wr.wr : NULL); if (unlikely(err)) { nvme_rdma_unmap_data(queue, rq); goto err; Patches currently in stable-queue which might be from sagi(a)grimberg.me are queue-4.14/nvme-rdma-don-t-suppress-send-completions.patch

7 years, 3 months

1
0
0 0

[PATCH stable 4.14.y] nvme-rdma: don't suppress send completions

by Sagi Grimberg

The entire completions suppress mechanism is currently broken because the HCA might retry a send operation (due to dropped ack) after the nvme transaction has completed. In order to handle this, we signal all send completions and introduce a separate done handler for async events as they will be handled differently (as they don't include in-capsule data by definition). Cc: <stable(a)vger.kernel.org> # v4.14+ Signed-off-by: Sagi Grimberg <sagi(a)grimberg.me> Reviewed-by: Max Gurtovoy <maxg(a)mellanox.com> Signed-off-by: Christoph Hellwig <hch(a)lst.de> --- drivers/nvme/host/rdma.c | 54 +++++++++++++----------------------------------- 1 file changed, 14 insertions(+), 40 deletions(-) diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c index 33d4431c2b4b..93a082e0bdd4 100644 --- a/drivers/nvme/host/rdma.c +++ b/drivers/nvme/host/rdma.c @@ -88,7 +88,6 @@ enum nvme_rdma_queue_flags { struct nvme_rdma_queue { struct nvme_rdma_qe *rsp_ring; - atomic_t sig_count; int queue_size; size_t cmnd_capsule_len; struct nvme_rdma_ctrl *ctrl; @@ -521,7 +520,6 @@ static int nvme_rdma_alloc_queue(struct nvme_rdma_ctrl *ctrl, queue->cmnd_capsule_len = sizeof(struct nvme_command); queue->queue_size = queue_size; - atomic_set(&queue->sig_count, 0); queue->cm_id = rdma_create_id(&init_net, nvme_rdma_cm_handler, queue, RDMA_PS_TCP, IB_QPT_RC); @@ -1232,21 +1230,9 @@ static void nvme_rdma_send_done(struct ib_cq *cq, struct ib_wc *wc) nvme_end_request(rq, req->status, req->result); } -/* - * We want to signal completion at least every queue depth/2. This returns the - * largest power of two that is not above half of (queue size + 1) to optimize - * (avoid divisions). - */ -static inline bool nvme_rdma_queue_sig_limit(struct nvme_rdma_queue *queue) -{ - int limit = 1 << ilog2((queue->queue_size + 1) / 2); - - return (atomic_inc_return(&queue->sig_count) & (limit - 1)) == 0; -} - static int nvme_rdma_post_send(struct nvme_rdma_queue *queue, struct nvme_rdma_qe *qe, struct ib_sge *sge, u32 num_sge, - struct ib_send_wr *first, bool flush) + struct ib_send_wr *first) { struct ib_send_wr wr, *bad_wr; int ret; @@ -1255,31 +1241,12 @@ static int nvme_rdma_post_send(struct nvme_rdma_queue *queue, sge->length = sizeof(struct nvme_command), sge->lkey = queue->device->pd->local_dma_lkey; - qe->cqe.done = nvme_rdma_send_done; - wr.next = NULL; wr.wr_cqe = &qe->cqe; wr.sg_list = sge; wr.num_sge = num_sge; wr.opcode = IB_WR_SEND; - wr.send_flags = 0; - - /* - * Unsignalled send completions are another giant desaster in the - * IB Verbs spec: If we don't regularly post signalled sends - * the send queue will fill up and only a QP reset will rescue us. - * Would have been way to obvious to handle this in hardware or - * at least the RDMA stack.. - * - * Always signal the flushes. The magic request used for the flush - * sequencer is not allocated in our driver's tagset and it's - * triggered to be freed by blk_cleanup_queue(). So we need to - * always mark it as signaled to ensure that the "wr_cqe", which is - * embedded in request's payload, is not freed when __ib_process_cq() - * calls wr_cqe->done(). - */ - if (nvme_rdma_queue_sig_limit(queue) || flush) - wr.send_flags |= IB_SEND_SIGNALED; + wr.send_flags = IB_SEND_SIGNALED; if (first) first->next = &wr; @@ -1329,6 +1296,12 @@ static struct blk_mq_tags *nvme_rdma_tagset(struct nvme_rdma_queue *queue) return queue->ctrl->tag_set.tags[queue_idx - 1]; } +static void nvme_rdma_async_done(struct ib_cq *cq, struct ib_wc *wc) +{ + if (unlikely(wc->status != IB_WC_SUCCESS)) + nvme_rdma_wr_error(cq, wc, "ASYNC"); +} + static void nvme_rdma_submit_async_event(struct nvme_ctrl *arg, int aer_idx) { struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(arg); @@ -1350,10 +1323,12 @@ static void nvme_rdma_submit_async_event(struct nvme_ctrl *arg, int aer_idx) cmd->common.flags |= NVME_CMD_SGL_METABUF; nvme_rdma_set_sg_null(cmd); + sqe->cqe.done = nvme_rdma_async_done; + ib_dma_sync_single_for_device(dev, sqe->dma, sizeof(*cmd), DMA_TO_DEVICE); - ret = nvme_rdma_post_send(queue, sqe, &sge, 1, NULL, false); + ret = nvme_rdma_post_send(queue, sqe, &sge, 1, NULL); WARN_ON_ONCE(ret); } @@ -1639,7 +1614,6 @@ static blk_status_t nvme_rdma_queue_rq(struct blk_mq_hw_ctx *hctx, struct nvme_rdma_request *req = blk_mq_rq_to_pdu(rq); struct nvme_rdma_qe *sqe = &req->sqe; struct nvme_command *c = sqe->data; - bool flush = false; struct ib_device *dev; blk_status_t ret; int err; @@ -1668,13 +1642,13 @@ static blk_status_t nvme_rdma_queue_rq(struct blk_mq_hw_ctx *hctx, goto err; } + sqe->cqe.done = nvme_rdma_send_done; + ib_dma_sync_single_for_device(dev, sqe->dma, sizeof(struct nvme_command), DMA_TO_DEVICE); - if (req_op(rq) == REQ_OP_FLUSH) - flush = true; err = nvme_rdma_post_send(queue, sqe, req->sge, req->num_sge, - req->mr->need_inval ? &req->reg_wr.wr : NULL, flush); + req->mr->need_inval ? &req->reg_wr.wr : NULL); if (unlikely(err)) { nvme_rdma_unmap_data(queue, rq); goto err; -- 2.14.1

7 years, 3 months

3
3
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror