While we're allocating ram for a new transaction, we drop our spinlock.
When we get the lock back, we do check to see if a transaction started
while we slept, but we don't check to make sure it isn't blocked
because a commit has already started.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
In case we were able to map less than we wanted (length < PAGE_SIZE
clause is true) btrfs_bio is still allocated and we have to free it.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The scrub readahead branch brought in a new error handling hook,
but it was leaking extent_buffer references.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The new ioctls to follow backrefs are not clean for 32/64 bit
compat. This reworks them for u64s everywhere. They are brand new, so
there are no problems with changing the interface now.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
We all keep getting those stupid warnings from use_block_rsv when running
stress.sh, and it's because the delayed insertion stuff is being stupid. It's
not the delayed insertion stuffs fault, it's all just stupid. When marking an
inode dirty for oh say updating the time on it, we just do a
btrfs_join_transaction, which doesn't reserve any space. This is stupid because
we're going to have to have space reserve to make this change, but we do it
because it's fast because chances are we're going to call it over and over again
and it doesn't matter. Well thanks to the delayed insertion stuff this is
mostly the case, so we do actually need to make this reservation. So if
trans->bytes_reserved is 0 then try to do a normal reservation. If not return
ENOSPC which will make the btrfs_dirty_inode start a proper transaction which
will let it do the whole ENOSPC dance and reserve enough space for the delayed
insertion to steal the reservation from the transaction.
The other stupid thing we do is not reserve space for the inode when writing to
the thing. Usually this is ok since we have to update the time so we'd have
already done all this work before we get to the endio stuff, so it doesn't
matter. But this is stupid because we could write the data after the
transaction commits where we changed the mtime of the inode so we have to cow
all the way down to the inode anyway. This used to be masked by the delalloc
reservation stuff, but because we delay the update it doesn't get masked in this
case. So again the delayed insertion stuff bites us in the ass. So if our
trans->block_rsv is delalloc, just steal the reservation from the delalloc
reserve. Hopefully this won't bite us in the ass, but I've said that before.
With this patch stress.sh no longer spits out those stupid warnings (famous last
words). Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Failure testing was tripping up over stale PageError bits in
metadata pages. If we have an io error on a block, and later on
end up reusing it, nobody ever clears PageError on those pages.
During commit, we'll find PageError and think we had trouble writing
the block, which will lead to aborts and other problems.
This changes clean_tree_block and the btrfs writepage code to
clear the PageError bit. In both cases we're either completely
done with the page or the page has good stuff and the error bit
is no longer valid.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Because of the overcommit stuff I had to make it so that we committed the
transaction all the time in reserve_metadata_bytes in case we had overcommitted
because of delayed items. This was because previously we had no way of knowing
how much space was reserved for delayed items. Now that we have the
delayed_block_rsv we can check it to see if committing the transaction would get
us anywhere. This patch breaks out the committing logic into a helper function
that will check to see if committing the transaction would free enough space for
us to get anything done. With this patch xfstests 83 goes from taking 445
seconds to taking 28 seconds on my box. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
I've been hitting warnings in use_block_rsv when running the delayed insertion
stuff. It's because we will readjust global block rsv based on what is in use,
which means we could end up discarding reservations that are for the delayed
insertion stuff. So instead create a seperate block rsv for the delayed
insertion stuff. This will also make it easier to debug problems with the
delayed insertion reservations since we will know that only the delayed
insertion code touches this block_rsv. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This takes some of the free space in the btrfs super block
to record information about most of the roots in the last four
commits.
It also adds a -o recovery to use the root history log when
we're not able to read the tree of tree roots, the extent
tree root, the device tree root or the csum root.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
fs_info has now ~9kb, more than fits into one page. This will cause
mount failure when memory is too fragmented. Top space consumers are
super block structures super_copy and super_for_commit, ~2.8kb each.
Allocate them dynamically. fs_info will be ~3.5kb. (measured on x86_64)
Add a wrapper for freeing fs_info and all of it's dynamically allocated
members.
Signed-off-by: David Sterba <dsterba@suse.cz>
We no longer use the orphan block rsv for holding the reservation for truncating
the inode, so instead use the global block rsv and check to make sure it has
enough space for us to truncate the space. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
I fixed a problem where we weren't reserving space for an orphan item when we
had to fallback to using the global reserve for an unlink, but I introduced
another problem. I was migrating the bytes from the transaction reserve to the
global reserve and then releasing from the global reserve in
btrfs_end_transaction(). The problem with this is that a migrate will jack up
the size for the destination, but leave the size alone for the source, with the
idea that you can do a release normally on the source and it all washes out, and
then you can do a release again on the destination and it works out right. My
way was skipping the release on the trans_block_rsv which still had the jacked
up size from our original reservation. So instead release manually from the
global reserve if this transaction was using it, and then set the
trans->block_rsv back to the trans_block_rsv so that btrfs_end_transaction
cleans everything up properly. With this patch xfstest 83 doesn't emit warnings
about leaking space. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
write_cache_pages tries to build up a large bio to stuff down the pipe.
But if it needs to wait for a page lock, it needs to make sure and send
down any pending writes so we don't deadlock with anyone who has the
page lock and is waiting for writeback of things inside the bio.
Dave Sterba triggered this as a deadlock between the autodefrag code and
the extent write_cache_pages
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The tree log had two important bugs that could cause corruptions after a
crash. Sometimes we were allowing tree log blocks to be reused after
the tree log was committed but before the transaction commit was done.
This allowed a future metadata write to overwrite the tree log data. It
is fixed by adding a new variant of freeing reserved extents that always
pins them. Credit goes to Stefan Behrens and Arne Jansen for many many
hours spent tracking this bug down.
During tree log replay, we do a pass through the tree log and pin all
the extents we find. This makes sure the replay code won't go in and
use any of those blocks for new allocations during replay. The problem
is the free space cache isn't honoring these pinned extents. So the
allocator can end up handing them out, leading to all kinds of problems
during replay.
The fix here is to force any free space cache to load while we pin the
extents, and then to make sure we remove the pinned extents from the
free space rbtree.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Reported-by: Stefan Behrens <sbehrens@giantdisaster.de>
btrfs_remove_free_space needs to make sure to set ret back to a
valid return value after setting it to EAGAIN, otherwise we return
it to the callers.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
When we're doing log commits, we try to wait for more writers to come in
and make the commit bigger. This helps improve performance on rotating
disks, but on SSDs it adds latencies.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The arch/arm/mach-exynos4 directory (CONFIG_ARCH_EXYNOS4) has
made for plaforms based on EXYNOS4 SoCs. But since upcoming
Samsung's SoCs such as EXYNOS5 (ARM Cortex A15) can reuse most
codes in current mach-exynos4, one mach-exynos directory will
be used for them.
This patch changes to CONFIG_ARCH_EXYNOS (arch/arm/mach-exynos)
but keeps original CONFIG_ARCH_EXYNOS4 in mach-exynos/Kconfig to
avoid changing in driver side.
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
MCT recently gained per cpu interrupts, and missed the fact that
ARM has moved to a genirq based implementation.
This patch converts the driver to the new API.
Boot tested on Origen.
Cc: Kukjin Kim <kgene.kim@samsung.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
PWM timers use pclk("timers" clk) as parent clk. If this pclk is the
disabled state when PWM driver is probed, then it causes wrong read and
write operation about registers of PWM.
Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
Fix following build error.
arch/arm/plat-samsung/dev-backlight.c: In function 'samsung_bl_set':
arch/arm/plat-samsung/dev-backlight.c:145: error: implicit declaration of function 'kfree'
Signed-off-by: Jonghwan Choi <jhbird.choi@gmail.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
Quiet the sparse noise:
warning: symbol 'create_fs_client' was not declared. Should it be static?
warning: symbol 'destroy_fs_client' was not declared. Should it be static?
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Sage Weil <sage@newdream.net>
ceph-devel@vger.kernel.org
Signed-off-by: Sage Weil <sage@newdream.net>
Quiet the following sparse noise:
warning: symbol 'get_nonsnap_parent' was not declared. Should it be static?
warning: symbol 'done_closing_sessions' was not declared. Should it be static?
Local functions don't need external visability. Make them static.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Sage Weil <sage@newdream.net>
Signed-off-by: Sage Weil <sage@newdream.net>
We used to use a flag on the directory inode to track whether the dcache
contents for a directory were a complete cached copy. Switch to a dentry
flag CEPH_D_COMPLETE that is safely updated by ->d_prune().
Signed-off-by: Sage Weil <sage@newdream.net>
* 'next/devel2' of git://git.linaro.org/people/arnd/arm-soc: (30 commits)
ARM: mmp: register internal sram bank
ARM: mmp: register audio sram bank
ARM: mmp: add sram allocator
gpio/samsung: Complain loudly if we don't know the SoC
ARM: S3C64XX: Fix SoC identification for S3C64xx devices
ARM: S3C2443: Remove redundant s3c_register_clocks call for init_clocks
ARM: S3C24XX: Add devname for hsmmc1 pclk
ARM: S3C24XX: use clk_get_rate to init fclk in common_setup_clocks
ARM: S3C2443: Accommodate cpufreq frequency scheme in armdiv
ARM: S3C2443: handle unset armdiv values gracefully
ARM: S3C2443: Add get_rate operation for clk_armdiv
ARM: S3C2416: Add comment describing the armdiv/armclk
ARM: S3C2443: Move clk_arm and clk_armdiv to common code
ARM: S3C24XX: Add infrastructure to transmit armdiv to common code
ARM: S3C2416: Add armdiv_mask constant
ARM: EXYNOS4: Add support for M-5MOLS camera on Nuri board
ARM: EXYNOS4: Enable MFC on ORIGEN
ARM: SAMSUNG: Add support s3c2416-adc for S3C2416/S3C2450
ARM: SAMSUNG: Add support s3c2443-adc for S3C2443
ARM: SAMSUNG: Allow overriding of adc device name for S3C24XX
...
* 'next/cleanup3' of git://git.linaro.org/people/arnd/arm-soc: (79 commits)
ARM: SAMSUNG: Move fimc plat. device from board files to plat-samsung
ARM: SAMSUNG: Cleanup resources by using macro
ARM: SAMSUNG: Cleanup plat-samsung/devs.c and devs.h
ARM: S5P: To merge devs.c files to one devs.c
ARM: S3C64XX: To merge devs.c files to one devs.c
ARM: S3C24XX: To merge s3c24xx devs.c files to one devs.c
ARM: S5P64X0: Add Power Management support
ARM: S5P: Make the sleep code common for S5P series SoCs
ARM: S5P: Make the common S5P PM code conditionally compile
ARM: SAMSUNG: Move S5P header files to plat-samsung
ARM: SAMSUNG: Move S3C24XX header files to plat-samsung
ARM: SAMSUNG: Moving each SoC support header files
ARM: SAMSUNG: Consolidate plat/pll.h
ARM: SAMSUNG: Consolidate plat/pwm-clock.h
ARM: SAMSUNG: Cleanup mach/clkdev.h
ARM: SAMSUNG: remove sdhci default configuration setup platform helper
ARM: EXYNOS4: Add FIMC device on SMDKV310 board
ARM: EXYNOS4: Add header file protection macros
ARM: EXYNOS4: Add usb ehci device to the SMDKV310
ARM: S3C2443: Add hsspi-clock from pclk and rename S3C2443 hsspi sclk
...
Fix up conflicts in
- arch/arm/mach-exynos4/{Kconfig,clock.c}
ARM_CPU_SUSPEND, various random device tables (gah!)
- drivers/gpio/Makefile
sa1100 gpio added, samsung gpio drivers merged
* git://www.linux-watchdog.org/linux-watchdog:
watchdog: Convert wm831x driver to watchdog core
watchdog: s3c2410: convert to use the watchdog framework
Documentation: watchdog: add guide how to convert drivers to new framework
watchdog: iTCO_wdt.c - problems with newer hardware due to SMI clearing
watchdog: Add WDIOC_GETTIMELEFT ioctl support to w83627 watchdog driver
watchdog: irq: Remove IRQF_DISABLED
watchdog: Octeon: Mark octeon_wdt interrupt as IRQF_NO_THREAD
watchdog: sc520_wdt: Remove unnecessary cast.
* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (45 commits)
[SCSI] Fix block queue and elevator memory leak in scsi_alloc_sdev
[SCSI] scsi_dh_alua: Fix the time inteval for alua rtpg commands
[SCSI] scsi_transport_iscsi: Fix documentation os parameter
[SCSI] mv_sas: OCZ RevoDrive3 & zDrive R4 support
[SCSI] libfc: improve flogi retries to avoid lport stuck
[SCSI] libfc: avoid exchanges collision during lport reset
[SCSI] libfc: fix checking FC_TYPE_BLS
[SCSI] edd: Treat "XPRS" host bus type the same as "PCI"
[SCSI] isci: overriding max_concurr_spinup oem parameter by max(oem, user)
[SCSI] isci: revert bcn filtering
[SCSI] isci: Fix hard reset timeout conditions.
[SCSI] isci: No need to manage the pending reset bit on pending requests.
[SCSI] isci: Remove redundant isci_request.ttype field.
[SCSI] isci: Fix task management for SMP, SATA and on dev remove.
[SCSI] isci: No task_done callbacks in error handler paths.
[SCSI] isci: Handle task request timeouts correctly.
[SCSI] isci: Fix tag leak in tasks and terminated requests.
[SCSI] isci: Immediately fail I/O to removed devices.
[SCSI] isci: Lookup device references through requests in completions.
[SCSI] ipr: add definitions for additional adapter
...
The includes in the pxa2xx_cm_x2xx PCMCIA driver are rather random;
the driver doesn't require anything from these headers except the
cpu_is_xxx() macros which come from another include.
The concern is that it's getting these definitions via mach/system.h,
which is supposed to only be included by arch/arm/kernel/process.c.
As this header is scheduled for cleanup (and elimination) keeping
the status quo will cause build errors. So lets fix properly and
independent of the future work.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
A mismerge in 43872fa (Merge branch 'depends/rmk/gpio' into next/fixes)
causes these build errors:
In file included from arch/arm/mach-pxa/include/mach/gpio.h:30,
from arch/arm/include/asm/gpio.h:6,
from include/linux/gpio.h:31,
from arch/arm/mach-pxa/generic.c:20:
arch/arm/mach-pxa/include/mach/gpio-pxa.h: In function ■__gpio_is_occupied■:
arch/arm/mach-pxa/include/mach/gpio-pxa.h:121: error: invalid operands to binary >> (have ■void *■ and ■unsigned int■)
arch/arm/mach-pxa/include/mach/gpio-pxa.h:122: error: invalid operands to binary & (have ■void *■ and ■int■)
arch/arm/mach-pxa/include/mach/gpio-pxa.h:129: error: invalid operands to binary & (have ■void *■ and ■int■)
So fix them.
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Fairly large code churn but not much doing with that and the overall
result is a definite win.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Make this driver a user of the watchdog framework and remove now
centrally handled parts. Tested on a mini2440.
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Acked-by: Kukjin Kim <kgene.kim@samsung.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Redhat Bugzilla: Bug 727875 - TCO_EN bit is disabled by TCO driver
Jiri Slaby: 28d41f53f broke temperature sensors on a ICH10 chipset
The iTCO_wdt driver disables the SMI. This breaks good working of newer hardware.
The disabling of the SMI by the TCO logic dates back from the i810-tco driver
from Nils Faerber (around 28 July 2000). The reason for this was that some BIOSes
install handlers reset or disable the watchdog timer instead of resetting the system.
The trick to fix this was to disable the SMI (by clearing the SMI_TCO_EN bit of the
SMI_EN register) to prevent this from happening.
This however has strange effects on newer hardware. So we are in a situation that
a fix for broken old hardware affects newer hardware.
The correct solution is to make this fix an option (with the new module parameter:
turn_SMI_watchdog_clear_off) so that the default behaviour is the unfixed version.
the next patch will be to move this in the start and stop functions of the driver
and to add a new module parameter for the global_smi_en bit and to get rid of the
vendor_support code.
This fix can have an effect on old (typical ICH & ICH2 chipsets) motherboards that
have a broken BIOS implementation concerning TCO logic. In these case the module
parameter turn_SMI_watchdog_clear_off=1 will need to be added.
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Add WDIOC_GETTIMELEFT ioctl allowing you to check how much time is left
on the watchdog counter before a reset occurs.
Signed-off-by: Greg Lee <glee [at] swspec.com>
Signed-off-by: Padraig Brady <P@draigbrady.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Signed-off-by: Andrew Morton <akpm@google.com>
This flag is a NOOP and can be removed now.
Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
This is to exclude it from force threading to allow RT patch set to work.
The watchdog timers are per-CPU and the addresses of register that reset
the timer are calculated based on the current CPU. Therefore we cannot
allow it to run on a thread on a different CPU. Also we only do a
single register write, which is much faster than scheduling a handler
thread.
And while on this line remove IRQF_DISABLED as this flag is a NOP.
Signed-off-by: Venkat Subbiah<venkat.subbiah@cavium.com>
Acked-by: David Daney<david.daney@cavium.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
When breaking apart the DaVinci GPIO files I accidentally
marked it non-complex while it is indeed complex.
Reported-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
This reverts commit 4bd0fe1c78.
This implementation can introduce a problem and 'ARM: SMP: fix
per cpu timer setup before the cpu is marked online' patch can
solve the cpu_oneline vs. cpu_active problem so that should be
reverted.
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Fix:
WARNING: vmlinux.o(.text+0x1a820): Section mismatch in reference from the function eseries_register_clks() to the function .init.text:clkdev_add_table()
The function eseries_register_clks() references
the function __init clkdev_add_table().
This is often because eseries_register_clks lacks a __init
annotation or the annotation of clkdev_add_table is wrong.
by adding the __init annotation to eseries_register_clks() - this
function is only called from other __init-marked functions. While
we're here, mark it static as it's only called from within eseries.c.
Acked-by: Eric Miao <eric.y.miao@gmail.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Fix:
ERROR: "pxa2xx_configure_sockets" [drivers/pcmcia/pxa2xx_lubbock_cs.ko] undefined!
by exporting the required function.
Acked-by: Eric Miao <eric.y.miao@gmail.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The mark of conflict should be removed.
This happened at the commit fba9569924 ("Merge branch 'next' of
git://git.infradead.org/users/vkoul/slave-dma")
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
[ I always check the conflict resolution with "git diff" before I add
the result, but I clearly missed that this time, and didn't notice the
second conflict in that file after having fixed the first one.
Oops, my bad. - Linus ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
During the re-organization of Ethernet drivers, the MAINTAINERS
entry for etherh got dropped accidentally.
CC: Russell King <linux@arm.linux.org.uk>
CC: Joe Perches <joe@perches.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
