1
0
Fork 0
alistair23-linux/drivers
Mike Snitzer f84cb8a46a dm mpath: disable WRITE SAME if it fails
Workaround the SCSI layer's problematic WRITE SAME heuristics by
disabling WRITE SAME in the DM multipath device's queue_limits if an
underlying device disabled it.

The WRITE SAME heuristics, with both the original commit 5db44863b6
("[SCSI] sd: Implement support for WRITE SAME") and the updated commit
66c28f971 ("[SCSI] sd: Update WRITE SAME heuristics"), default to enabling
WRITE SAME(10) even without successfully determining it is supported.
After the first failed WRITE SAME the SCSI layer will disable WRITE SAME
for the device (by setting sdkp->device->no_write_same which results in
'max_write_same_sectors' in device's queue_limits to be set to 0).

When a device is stacked ontop of such a SCSI device any changes to that
SCSI device's queue_limits do not automatically propagate up the stack.
As such, a DM multipath device will not have its WRITE SAME support
disabled.  This causes the block layer to continue to issue WRITE SAME
requests to the mpath device which causes paths to fail and (if mpath IO
isn't configured to queue when no paths are available) it will result in
actual IO errors to the upper layers.

This fix doesn't help configurations that have additional devices
stacked ontop of the mpath device (e.g. LVM created linear DM devices
ontop).  A proper fix that restacks all the queue_limits from the bottom
of the device stack up will need to be explored if SCSI will continue to
use this model of optimistically allowing op codes and then disabling
them after they fail for the first time.

Before this patch:

EXT4-fs (dm-6): mounted filesystem with ordered data mode. Opts: (null)
device-mapper: multipath: XXX snitm debugging: got -EREMOTEIO (-121)
device-mapper: multipath: XXX snitm debugging: failing WRITE SAME IO with error=-121
end_request: critical target error, dev dm-6, sector 528
dm-6: WRITE SAME failed. Manually zeroing.
device-mapper: multipath: Failing path 8:112.
end_request: I/O error, dev dm-6, sector 4616
dm-6: WRITE SAME failed. Manually zeroing.
end_request: I/O error, dev dm-6, sector 4616
end_request: I/O error, dev dm-6, sector 5640
end_request: I/O error, dev dm-6, sector 6664
end_request: I/O error, dev dm-6, sector 7688
end_request: I/O error, dev dm-6, sector 524288
Buffer I/O error on device dm-6, logical block 65536
lost page write due to I/O error on dm-6
JBD2: Error -5 detected when updating journal superblock for dm-6-8.
end_request: I/O error, dev dm-6, sector 524296
Aborting journal on device dm-6-8.
end_request: I/O error, dev dm-6, sector 524288
Buffer I/O error on device dm-6, logical block 65536
lost page write due to I/O error on dm-6
JBD2: Error -5 detected when updating journal superblock for dm-6-8.

# cat /sys/block/sdh/queue/write_same_max_bytes
0
# cat /sys/block/dm-6/queue/write_same_max_bytes
33553920

After this patch:

EXT4-fs (dm-6): mounted filesystem with ordered data mode. Opts: (null)
device-mapper: multipath: XXX snitm debugging: got -EREMOTEIO (-121)
device-mapper: multipath: XXX snitm debugging: WRITE SAME I/O failed with error=-121
end_request: critical target error, dev dm-6, sector 528
dm-6: WRITE SAME failed. Manually zeroing.

# cat /sys/block/sdh/queue/write_same_max_bytes
0
# cat /sys/block/dm-6/queue/write_same_max_bytes
0

It should be noted that WRITE SAME support wasn't enabled in DM
multipath until v3.10.

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: stable@vger.kernel.org # 3.10+
2013-09-20 10:36:34 -04:00
..
accessibility
acpi ACPI and power management fixes for 3.12-rc1 2013-09-12 11:22:45 -07:00
amba
ata Merge branch 'for-3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/libata 2013-09-03 18:19:53 -07:00
atm atm: he: print MAC via %pM 2013-09-04 14:41:55 -04:00
auxdisplay
base thp: account anon transparent huge pages into NR_ANON_PAGES 2013-09-12 15:38:03 -07:00
bcma
block Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
bluetooth
bus ARM: SoC platform changes for 3.12 2013-09-06 13:30:06 -07:00
cdrom
char Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
clk The common clk framework changes for 3.12 are dominated by clock driver 2013-09-09 15:49:04 -07:00
clocksource Merge branch 'timers/core' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 2013-09-16 16:10:26 -04:00
connector
cpufreq Merge branch 'pm-cpufreq' 2013-09-12 13:04:11 +02:00
cpuidle ACPI and power management fixes for 3.12-rc1 2013-09-12 11:22:45 -07:00
crypto Merge git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux 2013-09-07 12:53:35 +10:00
dca
devfreq
dio
dma Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
edac Merge git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile 2013-09-06 11:14:33 -07:00
eisa
extcon Driver core patches for 3.12-rc1 2013-09-03 11:37:15 -07:00
firewire firewire: ohci: Fix deadlock at bus reset 2013-08-29 22:35:05 +02:00
firmware firmware/dmi_scan: drop OOM messages 2013-09-11 15:58:51 -07:00
fmc
gpio Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
gpu Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2013-09-12 15:01:38 -07:00
hid Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
hsi
hv Drivers: hv: vmbus: Do not attempt to negoatiate a new version prematurely 2013-08-30 12:08:04 -07:00
hwmon hwmon: (k10temp) remove unnecessary pci_set_drvdata() 2013-09-12 08:17:40 -07:00
hwspinlock
i2c Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
ide ide: sgiioc4: Staticize ioc4_ide_attach_one() 2013-09-05 15:21:30 -04:00
idle
iio Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
infiniband Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
input Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input 2013-09-15 07:13:39 -04:00
iommu IOMMU Updates for Linux v3.12 2013-09-12 11:29:26 -07:00
ipack
irqchip ARM: SoC late changes for v3.12 2013-09-09 16:35:29 -07:00
isdn Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next 2013-09-05 14:54:29 -07:00
leds Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/linux-leds 2013-09-12 11:35:33 -07:00
lguest lguest: fix guest kernel stack overflow when TF bit set. 2013-09-06 08:09:27 +09:30
macintosh
mailbox
md dm mpath: disable WRITE SAME if it fails 2013-09-20 10:36:34 -04:00
media Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
memory ARM: SoC platform changes for 3.12 2013-09-06 13:30:06 -07:00
memstick memstick: add support for legacy memorysticks 2013-09-11 15:59:35 -07:00
message
mfd Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
misc Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
mmc Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
mtd Just a single fastmap fix plus a regression fix. 2013-09-16 15:37:52 -04:00
net Merge branch 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus 2013-09-15 17:45:52 -04:00
nfc
ntb ntb: clean up unnecessary MSI/MSI-X capability find 2013-09-05 11:08:01 -07:00
nubus
of Merge branch 'for-v3.12-fix' of git://git.linaro.org/people/mszyprowski/linux-dma-mapping 2013-09-11 19:11:59 -07:00
oprofile oprofile: get rid of pointless forward declarations of struct super_block 2013-09-03 22:52:48 -04:00
parisc PCI changes for the v3.12 merge window: 2013-09-03 16:24:35 -07:00
parport drivers: parport: Kconfig: exclude h8300 for PARPORT_PC 2013-08-30 12:08:04 -07:00
pci Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
pcmcia
pinctrl PTR_RET() is a weird name, and led to some confusing usage. We ended 2013-09-04 17:31:11 -07:00
platform Merge branch 'for_linus' of git://cavan.codon.org.uk/platform-drivers-x86 2013-09-15 17:42:59 -04:00
pnp pnp: change pnp bus pm_ops to invoke pnp driver dev_pm_ops if specified 2013-09-11 15:58:15 -07:00
power Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
pps Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
ps3
ptp
pwm ARM: SoC cleanups for 3.12 2013-09-06 13:21:16 -07:00
rapidio
regulator PTR_RET() is a weird name, and led to some confusing usage. We ended 2013-09-04 17:31:11 -07:00
remoteproc
reset
rpmsg
rtc rtc: simplify devm_request_mem_region/devm_ioremap 2013-09-11 15:58:59 -07:00
s390 s390/vmcore: use vmcore for zfcpdump 2013-09-11 15:59:15 -07:00
sbus
scsi SCSI misc on 20130915 2013-09-15 17:41:30 -04:00
sfi
sh
sn
spi Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
ssb
staging Merge git://git.kvack.org/~bcrl/aio-next 2013-09-13 10:55:58 -07:00
target Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending 2013-09-12 16:11:45 -07:00
tc
thermal drivers: thermal: add check when unregistering cpu cooling 2013-09-03 09:10:34 -04:00
tty Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
uio Driver core patches for 3.12-rc1 2013-09-03 11:37:15 -07:00
usb Merge git://git.kvack.org/~bcrl/aio-next 2013-09-13 10:55:58 -07:00
uwb
vfio vfio-pci: PCI hot reset interface 2013-09-04 11:28:04 -06:00
vhost Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending 2013-09-12 16:11:45 -07:00
video Bye, bye, WfW flag 2013-09-11 19:55:12 -07:00
virt
virtio virtio_pci: pm: Use CONFIG_PM_SLEEP instead of CONFIG_PM 2013-09-09 10:02:53 +09:30
vlynq
vme vme: vme_ca91cx42.c: fix to pass correct device identity to free_irq() 2013-08-27 21:49:26 -07:00
w1 Remove GENERIC_HARDIRQ config option 2013-09-13 15:09:52 +02:00
watchdog Merge git://www.linux-watchdog.org/linux-watchdog 2013-09-12 08:45:53 -07:00
xen Bug-fixes: 2013-09-12 11:28:24 -07:00
zorro
Kconfig
Makefile