| Age | Commit message (Collapse) | Author |
|---|
|
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging-2.6
* 'staging-next' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging-2.6: (510 commits)
staging: speakup: fix failure handling
staging: usbip: remove double giveback of URB
Staging: batman-adv: Remove batman-adv from staging
Staging: hv: Use only one txf buffer per channel and kmalloc/GFP_KERNEL on initialize
staging: hv: remove unneeded osd_schedule_callback
staging: hv: convert channel_mgmt.c to not call osd_schedule_callback
staging: hv: convert vmbus_on_msg_dpc to not call osd_schedule_callback
staging: brcm80211: Fix WL_<type> logging macros
Staging: IIO: DDS: AD9833 / AD9834 driver
Staging: IIO: dds.h convenience macros
Staging: IIO: Direct digital synthesis abi documentation
staging: brcm80211: Convert ETHER_TYPE_802_1X to ETH_P_PAE
staging: brcm80211: Remove unused ETHER_TYPE_<foo> #defines
staging: brcm80211: Remove ETHER_HDR_LEN, use ETH_HLEN
staging: brcm80211: Convert ETHER_ADDR_LEN to ETH_ALEN
staging: brcm80211: Convert ETHER_IS<FOO> to is_<foo>_ether_addr
staging: brcm80211: Remove unused ether_<foo> #defines and struct
staging: brcm80211: Convert ETHER_IS_MULTI to is_multicast_ether_addr
staging: brcm80211: Remove unused #defines ETHER_<foo>_LOCALADDR
Staging: comedi: Fix checkpatch.pl issues in file s526.c
...
Fix up trivial conflict in drivers/video/udlfb.c
|
|
And a test for it:
[acme@felicio linux]$ perf test
1: vmlinux symtab matches kallsyms: Ok
2: detect open syscall event: Ok
3: detect open syscall event on all cpus: Ok
[acme@felicio linux]$
Translating C the test does:
1. generates different number of open syscalls on each CPU
by using sched_setaffinity
2. Verifies that the expected number of events is generated
on each CPU
It works as expected.
LKML-Reference: <new-submission>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
|
|
The size field should not be set until after the data is successfully
copied in.
Signed-off-by: Dan Rosenberg <drosenberg@vsecurity.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mjg59/platform-drivers-x86
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mjg59/platform-drivers-x86: (36 commits)
sony-laptop: support new hotkeys on the P, Z and EC series
platform/x86: Consistently select LEDS Kconfig options
sony-laptop: fix sparse non-ANSI function warning
intel_ips: fix sparse non-ANSI function warning
Support KHLB2 in the compal laptop driver
acer-wmi: Enabled Acer Launch Manager mode
[PATCH] intel_pmic_gpio: modify EOI handling following change of kernel irq subsystem
ACPI Thinkpad: We must always call va_end() after va_start() but do not do so in thinkpad_acpi.c::acpi_evalf()
acer-wmi: Initialize wlan/bluetooth/wwan rfkill software block state
acer-wmi: Detect the WiFi/Bluetooth/3G devices available
acer-wmi: Add 3G rfkill sysfs file
acer-wmi: Add acer wmi hotkey events support
platform/x86: Kconfig: Replace select by depends on ACPI_WMI
ideapad: pass ideapad_priv as argument (part 2)
ideapad: pass ideapad_priv as argument (part 1)
ideapad: add markups, unify comments and return result when init
ideapad: add hotkey support
ideapad: let camera power control entry under platform driver
ideapad: add platform driver for ideapad
fujitsu-laptop: fix compiler warning on pnp_ids
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k:
m68knommu: Need to check __get_user()/__put_user() result
m68knommu: signal.c __user annotations
m68knommu: Equivalent of "m68k: handle new gcc's"
m68knommu: f_pcr has been gone since headers' merge
m68knommu: Don't lose state if sigframe setup fails
m68knommu: Handle multiple pending signals
m68knommu: Switch to saner sigsuspend
m68knommu: Don't bother with SA_ONESHOT
m68k: Check __get_user()/__put_user() return value
m68k: Missing syscall_trace() on sigreturn
m68k: Fix stack mangling logics in sigreturn
m68k: If we fail to set sigframe up, just leave regs alone...
m68k: Don't lose state if sigframe setup fails
m68k: Simplify the singlestepping handling in signals
m68k: Switch to saner sigsuspend()
m68k: Resetting sa_handler in local copy of k_sigaction is pointless
m68k/sun3: Kill pte_unmap() warnings
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
[IA64] Avoid array overflow if there are too many cpus in SRAT table
[IA64] Remove unlikely from cpu_is_offline
[IA64] irq_ia64, use set_irq_chip
[IA64] perfmon: Change vmalloc to vzalloc and drop memset.
[IA64] eliminate race condition in smp_flush_tlb_mm
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson
* 'for-torvalds' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson:
ux500: allow 5500 and 8500 to be built together
ux500: modem_irq is only for 5500
ux500: dynamic SOC detection
ux500: rename MOP board Kconfig
ux500: remove build-time changing macros
|
|
* 'msm-smp' of git://codeaurora.org/quic/kernel/davidb/linux-msm:
msm: add SMP support for msm
msm: hotplug: support cpu hotplug on msm
msm: timer: SMP timer support for msm
msm: scm-boot: Support for setting cold/warm boot addresses
msm: Secure Channel Manager (SCM) support
|
|
For the VFs, the Mailbox Data "registers" are actually backed by
T4's "MA" interface rather than PL Registers (as is the case for
the PFs). Because these are in different coherency domains, the
write to the VF's PL-register-backed Mailbox Control can race in
front of the writes to the MA-backed VF Mailbox Data "registers".
So we need to do a read-back on at least one byte of the VF Mailbox
Data registers before doing the write to the VF Mailbox Control
register.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
o Updating module parameter after driver load is not supported
except auto_fw_reset parameter. Changing these parameter after
driver load, can have weird result.
o Update driver version to 5.0.15.
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IRQ diag test was getting executed only when both register test
and link test passed. The test should get executed if ETH_TEST_FL_OFFLINE
flag is set.
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Driver is reading flash fw version from defined address, this address
may be invalid. Indeed Driver should read address for fw version through
flash layout table. Flash layout table has defined region and address for
fw version address should be read from fw image region.
Driver has check for old firmware, this bug can cause driver load fail.
This patch will try to read fw version from flash image region, if that fails,
read from defined address.
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Dan Rosenberg pointed out that there were some signed comparison bugs
in the phonet protocol.
http://marc.info/?l=full-disclosure&m=129424528425330&w=2
The problem is that we check for array overflows but "protocol" is
signed and we don't check for array underflows. If you have already
have CAP_SYS_ADMIN then you could use the bugs to get root, or someone
could cause an oops by mistake.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Acked-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
With writeback caches, corrupted RX packets will be sent up the stack
without any error markings.
Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We don't use this local "Mac" data anywhere (since we rely on the
netdev's storage), so punt it.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
No need for this to be exported since it is only used in this driver.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use netdev_* and pr_* helper funcs for output rather than printk.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Helper functions for I2C and SMBus transactions don't modify the
i2c_client that is passed to them, so it can be marked const.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
|
|
The I2C specification explicitly describes both SDA and SCL as
bidirectional lines. An I2C master with a read-only SCL is thus not
compliant. If a slow slave stretches the clock, errors will happen,
so the bus can't be considered as reliable.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
|
|
Use a function pointer to decide whether to call i2c_add_adapter or
i2c_add_numbered_adapter. This makes the code more compact than the
current strategy of having the common code in a separate function.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
|
|
Add an i2c mux driver providing access to i2c bus segments using a
hardware MUX sitting on a master bus and controlled through gpio pins.
E.G. something like:
---------- ---------- Bus segment 1 - - - - -
| | SCL/SDA | |-------------- | |
| |------------| |
| | | | Bus segment 2 | |
| Linux | GPIO 1..N | MUX |--------------- Devices
| |------------| | | |
| | | | Bus segment M
| | | |---------------| |
---------- ---------- - - - - -
SCL/SDA of the master I2C bus is multiplexed to bus segment 1..M
according to the settings of the GPIO pins 1..N.
Signed-off-by: Peter Korsgaard <peter.korsgaard@barco.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
|
|
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
|
|
Commit 5a0e3ad6af8660be21ca98a971cd00f331318c05 added direct inclusion
of <linux/slab.h> to those source files that appeared to need it, but
somehow missed this. On most architectures <linux/slab.h> is still
indirectly included, but there are exceptions such as alpha.
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
|
|
Put the variables accessed together in the hot-path into common
cachelines, and separate them by RW vs RO to avoid false dirtying.
We keep a local copy of the lkey and rkey in the target to avoid
traversing pointers (and associated cache lines) to find them.
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: David Dillow <dillowda@ornl.gov>
|
|
We don't need protection against the SCSI stack, so use our own lock to
allow parallel progress on separate CPUs.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
[ broken out and small cleanups by David Dillow ]
Signed-off-by: David Dillow <dillowda@ornl.gov>
|
|
We only need the lock to cover list and credit manipulations, so push
those into srp_remove_req() and update the call chains.
We reorder the request removal and command completion in
srp_process_rsp() to avoid the SCSI mid-layer sending another command
before we've released our request and added any credits returned by the
target. This prevents us from returning HOST_BUSY unneccesarily.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
[ broken out, small cleanups, and modified to avoid potential extraneous
HOST_BUSY returns by David Dillow ]
Signed-off-by: David Dillow <dillowda@ornl.gov>
|
|
We only need locks to protect our lists and number of credits available.
By pre-consuming the credit for the request, we can reduce our lock
coverage to just those areas. If we don't actually send the request,
we'll need to put the credit back into the pool.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
[ broken out and small cleanups by David Dillow ]
Signed-off-by: David Dillow <dillowda@ornl.gov>
|
|
We use req->scmnd != NULL to indicate an active request, so there's no
need to keep a separate list for them. We can afford the array iteration
during error handling, and dropping it gives us one less item that needs
lock protection.
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
[ broken out and small cleanups by David Dillow ]
Signed-off-by: David Dillow <dillowda@ornl.gov>
|
|
fix the failure handling in kobjects and the main function so that we
release the virtual keyboard if we exit due to another failure.
Signed-off-by: William Hubbs <w.d.hubbs@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
|
|
In the vhci_urb_dequeue() function the TCP connection is checked twice.
Each time when the TCP connection is closed the URB is unlinked and given
back. Remove the second attempt of unlinking and giving back of the URB completely.
This patch fixes the bug described at https://bugzilla.kernel.org/show_bug.cgi?id=24872 .
Signed-off-by: Márton Németh <nm127@freemail.hu>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
|
|
Conflicts:
fs/nfs/nfs2xdr.c
fs/nfs/nfs3xdr.c
fs/nfs/nfs4xdr.c
|
|
vm_map_ram() is not available on NOMMU platforms, and causes trouble
on incoherrent architectures such as ARM when we access the page data
through both the direct and the virtual mapping.
The alternative is to use the direct mapping to access page data
for the case when we are not crossing a page boundary, but to copy
the data into a linear scratch buffer when we are accessing data
that spans page boundaries.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Tested-by: Marc Kleine-Budde <mkl@pengutronix.de>
Cc: stable@kernel.org [2.6.37]
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6: (30 commits)
MAINTAINERS: Add tomoyo-dev-en ML.
SELinux: define permissions for DCB netlink messages
encrypted-keys: style and other cleanup
encrypted-keys: verify datablob size before converting to binary
trusted-keys: kzalloc and other cleanup
trusted-keys: additional TSS return code and other error handling
syslog: check cap_syslog when dmesg_restrict
Smack: Transmute labels on specified directories
selinux: cache sidtab_context_to_sid results
SELinux: do not compute transition labels on mountpoint labeled filesystems
This patch adds a new security attribute to Smack called SMACK64EXEC. It defines label that is used while task is running.
SELinux: merge policydb_index_classes and policydb_index_others
selinux: convert part of the sym_val_to_name array to use flex_array
selinux: convert type_val_to_struct to flex_array
flex_array: fix flex_array_put_ptr macro to be valid C
SELinux: do not set automatic i_ino in selinuxfs
selinux: rework security_netlbl_secattr_to_sid
SELinux: standardize return code handling in selinuxfs.c
SELinux: standardize return code handling in selinuxfs.c
SELinux: standardize return code handling in policydb.c
...
|
|
Using "iptables -L" with a lot of rules have a too big BH latency.
Jesper mentioned ~6 ms and worried of frame drops.
Switch to a per_cpu seqlock scheme, so that taking a snapshot of
counters doesnt need to block BH (for this cpu, but also other cpus).
This adds two increments on seqlock sequence per ipt_do_table() call,
its a reasonable cost for allowing "iptables -L" not block BH
processing.
Reported-by: Jesper Dangaard Brouer <hawk@comx.dk>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Patrick McHardy <kaber@trash.net>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Acked-by: Jesper Dangaard Brouer <hawk@comx.dk>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This is driver for EasyPoint AS5011 2 axis joystick chip. This chip is
plugged on an I2C bus.
Tested on ARM processor (i.MX27).
Signed-off-by: Fabien Marteau <fabien.marteau@armadeus.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
|
|
EXT2_XATTR_DEBUG
When I enable EXT2_XATTR_DEBUG in fs/ext2/xattr.c I get a build error stating
the following:
CC fs/ext2/xattr.o
fs/ext2/xattr.c: In function 'ext2_xattr_cache_insert':
fs/ext2/xattr.c:841: error: dereferencing pointer to incomplete type
fs/ext2/xattr.c:846: error: dereferencing pointer to incomplete type
make[2]: *** [fs/ext2/xattr.o] Error 1
make[1]: *** [fs/ext2] Error 2
make: *** [fs] Error 2
These lines reference ext2_xattr_cache->c_entry_count which is defined
in struct mb_cache. struct mb_cache is currently only defined in fs/mbcache.c.
Moving struct mb_cache definition to include/linux/mbcache.h to resolve the
issue.
Signed-off-by: Josh Hunt <johunt@akamai.com>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
IS_ERR() already implies unlikely(), so it can be omitted here.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
IS_ERR() already implies unlikely(), so it can be omitted here.
Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
The addition of 64k block capability in the rec_len_from_disk
and rec_len_to_disk functions added a bit of math overhead which
slows down file create workloads needlessly when the architecture
cannot even support 64k blocks, thanks to page size limits.
Similar changes already exist in the ext4 codebase.
The directory entry checking can also be optimized a bit
by sprinkling in some unlikely() conditions to move the
error handling out of line.
bonnie++ sequential file creates on a 512MB ramdisk speeds up
from about 77,000/s to about 82,000/s, about a 6% improvement.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
The addition of 64k block capability in the rec_len_from_disk
and rec_len_to_disk functions added a bit of math overhead which
slows down file create workloads needlessly when the architecture
cannot even support 64k blocks, thanks to page size limits.
The directory entry checking can also be optimized a bit
by sprinkling in some unlikely() conditions to move the
error handling out of line.
bonnie++ sequential file creates on a 512MB ramdisk speeds up
from about 2200/s to about 2500/s, about a 14% improvement.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
Check return value of ext3_journal_get_write_acccess() and
ext3_journal_dirty_metadata().
Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
Check return value of ext3_journal_get_write_access() and
ext3_journal_dirty_metadata().
Signed-off-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
fallout
Use %pV in __quota_error so a single printk can not be
interleaved with other logging messages.
Add __attribute__((format (printf, 3, 4))) so format
and arguments can be verified by compiler.
Make sure printk formats and arguments match.
Block # needed a pointer dereference.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
The ioctl takes fstrim_range structure (defined in include/linux/fs.h) as an
argument specifying a range of filesystem to trim and the minimum size of an
continguous extent to trim. After the FITRIM is done, the number of bytes
passed from the filesystem down the block stack to the device for potential
discard is stored in fstrim_range.len. This number is a maximum discard amount
from the storage device's perspective, because FITRIM called repeatedly will
keep sending the same sectors for discard. fstrim_range.len will report the
same potential discard bytes each time, but only sectors which had been written
to between the discards would actually be discarded by the storage device.
Further, the kernel block layer reserves the right to adjust the discard ranges
to fit raid stripe geometry, non-trim capable devices in a LVM setup, etc.
These reductions would not be reflected in fstrim_range.len.
Thus fstrim_range.len can give the user better insight on how much storage
space has potentially been released for wear-leveling, but it needs to be one
of only one criteria the userspace tools take into account when trying to
optimize calls to FITRIM.
Thanks to Greg Freemyer <greg.freemyer@gmail.com> for better commit message.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
Walk through allocation groups and trim all free extents. It can be
invoked through FITRIM ioctl on the file system. The main idea is to
provide a way to trim the whole file system if needed, since some SSD's
may suffer from performance loss after the whole device was filled (it
does not mean that fs is full!).
It search for free extents in allocation groups specified by Byte range
start -> start+len. When the free extent is within this range, blocks are
marked as used and then trimmed. Afterwards these blocks are marked as
free in per-group bitmap.
[JK: Fixed up error handling and trimming of a single group]
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: Jan Kara <jack@suse.cz>
|
|
Since check_eofblocks_fl() only uses the m_lblk portion of the map
structure, we may as well pass that directly, rather than passing the
entire map, which IMHO obfuscates what parameters check_eofblocks_fl()
cares about. Not a big deal, but seems tidier and less confusing, to
me.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
|
|
Commit 40389687 moved a call to ext4_forget() out of
ext4_free_branches and let ext4_free_blocks() handle calling
bforget(). But that change unfortunately did not replace the call to
ext4_forget() with brelse(), which was needed to drop the in-use count
of the indirect block's buffer head, which lead to a memory leak when
deleting files that used indirect blocks. Fix this.
Thanks to Hugh Dickins for pointing this out.
Cc: stable@kernel.org
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
|
|
This function was never implemented, except for a BUG_ON which was
tripping when ext4 is run without a journal. The problem is that
although the comment asserts that "truncate (which is the only way to
free block) discards all preallocations", ext4_free_blocks() is also
called in various error recovery paths when blocks have been
allocated, but for various reasons, we were not able to use those data
blocks (for example, because we ran out of memory while trying to
manipulate the extent tree, or some other similar situation).
In addition to the fact that this function isn't implemented except
for the incorrect BUG_ON, the single caller of this function,
ext4_free_blocks(), doesn't use it all if the journal is enabled.
So remove the (stub) function entirely for now. If we decide it's
better to add it back, it's only going to be useful with a relatively
large number of code changes anyway.
Google-Bug-Id: 3236408
Cc: Jiaying Zhang <jiayingz@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
|
|
Ted first found the bug when running 2.6.36 kernel with dioread_nolock
mount option that xfstests #13 complained about wrong file size during fsck.
However, the bug exists in the older kernels as well although it is
somehow harder to trigger.
The problem is that ext4_end_io_work() can happen after we have truncated an
inode to a smaller size. Then when ext4_end_io_work() calls
ext4_convert_unwritten_extents(), we may reallocate some blocks that have
been truncated, so the inode size becomes inconsistent with the allocated
blocks.
The following patch flushes the i_completed_io_list during truncate to reduce
the risk that some pending end_io requests are executed later and convert
already truncated blocks to initialized.
Note that although the fix helps reduce the problem a lot there may still
be a race window between vmtruncate() and ext4_end_io_work(). The fundamental
problem is that if vmtruncate() is called without either i_mutex or i_alloc_sem
held, it can race with an ongoing write request so that the io_end request is
processed later when the corresponding blocks have been truncated.
Ted and I have discussed the problem offline and we saw a few ways to fix
the race completely:
a) We guarantee that i_mutex lock and i_alloc_sem write lock are both hold
whenever vmtruncate() is called. The i_mutex lock prevents any new write
requests from entering writeback and the i_alloc_sem prevents the race
from ext4_page_mkwrite(). Currently we hold both locks if vmtruncate()
is called from do_truncate(), which is probably the most common case.
However, there are places where we may call vmtruncate() without holding
either i_mutex or i_alloc_sem. I would like to ask for other people's
opinions on what locks are expected to be held before calling vmtruncate().
There seems a disagreement among the callers of that function.
b) We change the ext4 write path so that we change the extent tree to contain
the newly allocated blocks and update i_size both at the same time --- when
the write of the data blocks is completed.
c) We add some additional locking to synchronize vmtruncate() and
ext4_end_io_work(). This approach may have performance implications so we
need to be careful.
All of the above proposals may require more substantial changes, so
we may consider to take the following patch as a bandaid.
Signed-off-by: Jiaying Zhang <jiayingz@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
|
