Sage Weil [Tue, 24 May 2011 20:06:04 +0000 (13:06 -0700)]
vfs: dentry_unhash immediately prior to rmdir
This presumes that there is no reason to unhash a dentry if we fail because
it is a mountpoint or the LSM check fails, and that the LSM checks do not
depend on the dentry being unhashed.
Signed-off-by: Sage Weil <sage@newdream.net> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Jan Kara [Mon, 23 May 2011 22:23:35 +0000 (00:23 +0200)]
vfs: Block mmapped writes while the fs is frozen
We should not allow file modification via mmap while the filesystem is
frozen. So block in block_page_mkwrite() while the filesystem is frozen.
We cannot do the blocking wait in __block_page_mkwrite() since e.g. ext4
will want to call that function with transaction started in some cases
and that would deadlock. But we can at least do the non-blocking reliable
check in __block_page_mkwrite() which is the hardest part anyway.
We have to check for frozen filesystem with the page marked dirty and under
page lock with which we then return from ->page_mkwrite(). Only that way we
cannot race with writeback done by freezing code - either we mark the page
dirty after the writeback has started, see freezing in progress and block, or
writeback will wait for our page lock which is released only when the fault is
done and then writeback will writeout and writeprotect the page again.
Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Jan Kara [Mon, 23 May 2011 22:23:34 +0000 (00:23 +0200)]
vfs: Create __block_page_mkwrite() helper passing error values back
Create __block_page_mkwrite() helper which does all what block_page_mkwrite()
does except that it passes back errors from __block_write_begin /
block_commit_write calls.
Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Roman Borisov [Wed, 25 May 2011 23:26:48 +0000 (16:26 -0700)]
fs/namespace.c: bound mount propagation fix
This issue was discovered by users of busybox. And the bug is actual for
busybox users, I don't know how it affects others. Apparently, mount is
called with and without MS_SILENT, and this affects mount() behaviour.
But MS_SILENT is only supposed to affect kernel logging verbosity.
The following script was run in an empty test directory:
mkdir -p mount.dir mount.shared1 mount.shared2
touch mount.dir/a mount.dir/b
mount -vv --bind mount.shared1 mount.shared1
mount -vv --make-rshared mount.shared1
mount -vv --bind mount.shared2 mount.shared2
mount -vv --make-rshared mount.shared2
mount -vv --bind mount.shared2 mount.shared1
mount -vv --bind mount.dir mount.shared2
ls -R mount.dir mount.shared1 mount.shared2
umount mount.dir mount.shared1 mount.shared2 2>/dev/null
umount mount.dir mount.shared1 mount.shared2 2>/dev/null
umount mount.dir mount.shared1 mount.shared2 2>/dev/null
rm -f mount.dir/a mount.dir/b mount.dir/c
rmdir mount.dir mount.shared1 mount.shared2
mount -vv was used to show the mount() call arguments and result.
Output shows that flag argument has 0x00008000 = MS_SILENT bit:
mount: mount('mount.shared1','mount.shared1','(null)',0x00009000,'(null)'):0
mount: mount('','mount.shared1','',0x0010c000,''):0
mount: mount('mount.shared2','mount.shared2','(null)',0x00009000,'(null)'):0
mount: mount('','mount.shared2','',0x0010c000,''):0
mount: mount('mount.shared2','mount.shared1','(null)',0x00009000,'(null)'):0
mount: mount('mount.dir','mount.shared2','(null)',0x00009000,'(null)'):0
mount.dir:
a
b
mount.shared1:
mount.shared2:
a
b
After adding --loud option to remove MS_SILENT bit from just one mount cmd:
The analysis shows that MS_SILENT flag which is ON by default in any
busybox-> mount operations cames to flags_to_propagation_type function and
causes the error return while is_power_of_2 checking because the function
expects only one bit set. This doesn't allow to do busybox->mount with
any --make-[r]shared, --make-[r]private etc options.
Moreover, the recently added flags_to_propagation_type() function doesn't
allow us to do such operations as --make-[r]private --make-[r]shared etc.
when MS_SILENT is on. The idea or clearing the MS_SILENT flag came from
to Denys Vlasenko.
Signed-off-by: Roman Borisov <ext-roman.borisov@nokia.com> Reported-by: Denys Vlasenko <vda.linux@googlemail.com> Cc: Chuck Ebbert <cebbert@redhat.com> Cc: Alexander Shishkin <virtuoso@slind.org> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Jonas Gorski [Tue, 24 May 2011 18:12:08 +0000 (20:12 +0200)]
exportfs: reallow building as a module
Commit 990d6c2d7aee921e3bce22b2d6a750fd552262be ("vfs: Add name to file
handle conversion support") changed EXPORTFS to be a bool.
This was needed for earlier revisions of the original patch, but the actual
commit put the code needing it into its own file that only gets compiled
when FHANDLE is selected which in turn selects EXPORTFS.
So EXPORTFS can be safely compiled as a module when not selecting FHANDLE.
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com> Acked-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Al Viro [Fri, 25 Mar 2011 15:00:12 +0000 (11:00 -0400)]
merge handle_reval_dot and nameidata_drop_rcu_last
new helper: complete_walk(). Done on successful completion
of walk, drops out of RCU mode, does d_revalidate of final
result if that hadn't been done already.
handle_reval_dot() and nameidata_drop_rcu_last() subsumed into
that one; callers converted to use of complete_walk().
Merge branch 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
perf, x86, nmi: Move LVT un-masking into irq handlers
perf events, x86: Work around the Nehalem AAJ80 erratum
perf, x86: Fix BTS condition
ftrace: Build without frame pointers on Microblaze
Merge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86: ce4100: Configure IOAPIC pins for USB and SATA to level type
x86: devicetree: Configure IOAPIC pin only once
x86, setup: When probing memory with e801, use ax/bx as a pair
Merge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging
* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
hwmon: (lm85) Fix error paths in probe function
hwmon: (lm85) Add missing list terminators
hwmon: (adm1021) Clarify documentation regarding Xeon processors
hwmon: (lm90) Fix update interval information in driver documentation
hwmon: (lm90) Add support for ADT7461A and NCT1008
Merge branch 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6
* 'omap-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6:
OMAP3+: voltage: remove initial voltage
OMAP4: Intialize IVA Device in addition to DSP device.
omap: rx51: mark reserved memory earlier
OMAP3: l3: fix for "irq 10: nobody cared" message
arm: omap2: enable smc instruction for sleep34xx
OMAP2/3: hwmod: fix gpio-reset timeouts seen during bootup.
OMAP3: PM: Do not rely on ROM code to restore CM_AUTOIDLE_PLL.AUTO_PERIPH_DPLL
OMAP2+: PM: Fix the saving of CM_AUTOIDLE_PLL register on scratchpad area
OMAP4: clock data: Change DSS clock aliases
OMAP2+: hwmod data: Fix wrong dma_system end address
Recent Xeon processor thermal sensors are supported by the coretemp
driver and not the adm1021 driver. Only one old generation of Xeon
processors (the first Netburst ones) are supported by the adm1021
driver.
Reported-by: Darren Hart <dvhart@linux.intel.com> Signed-off-by: Jean Delvare <khali@linux-fr.org> Acked-by: Guenter Roeck <guenter.roeck@ericsson.com>
hwmon: (lm90) Add support for ADT7461A and NCT1008
This patch adds support for ADT7461A and NCT1008 to the lm90 driver.
Both chips have identical functionality and report the same manufacturing ID
and device ID values.
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com> Signed-off-by: Jean Delvare <khali@linux-fr.org>
Thomas Gleixner [Thu, 28 Apr 2011 22:02:00 +0000 (00:02 +0200)]
hrtimer: Initialize CLOCK_ID to HRTIMER_BASE table statically
Sedat and Bruno reported RCU stalls which turned out to be caused by
the following;
sched_init() calls init_rt_bandwidth() which calls hrtimer_init()
_BEFORE_ hrtimers_init() is called. While not entirely correct this
worked because hrtimer_init() only accessed statically initialized
data (hrtimer_bases.clock_base[CLOCK_MONOTONIC])
Commit e06383db9 (hrtimers: extend hrtimer base code to handle more
then 2 clockids) added an indirection to the hrtimer_bases.clock_base
lookup to avoid gap handling in the hot path. The table which is used
for the translataion from CLOCK_ID to HRTIMER_BASE index is
initialized at runtime in hrtimers_init(). So the early call of the
scheduler code translates CLOCK_MONOTONIC to HRTIMER_BASE_REALTIME.
Thus the rt_bandwith timer ends up on CLOCK_REALTIME. If the timer is
armed and the wall clock time is set (e.g. ntpdate in the early boot
process - which also gives the problem deterministic behaviour
i.e. magic recovery after N hours), then the timer ends up with an
expiry time far into the future. That breaks the RT throttler
mechanism as rt runtime is accumulated and never cleared, so the rt
throttler detects a false cpu hog condition and blocks all RT tasks
until the timer finally expires. That in turn stalls the RCU thread of
TINYRCU which leads to an huge amount of RCU callbacks piling up.
Make the translation table statically initialized, so we are back to
the status of <= 2.6.39.
Reported-and-tested-by: Sedat Dilek <sedat.dilek@gmail.com> Reported-by: Bruno Prémont <bonbons@linux-vserver.org> Cc: John stultz <johnstul@us.ibm.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/%3Calpine.LFD.2.02.1104282353140.3005%40ionos%3E Reviewed-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Merge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6
* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
drm/i915: restore only the mode of this driver on lastclose (v2)
drm/radeon/kms: add info query for tile pipes
drm/radeon/kms: add missing safe regs for 6xx/7xx
drm: select FRAMEBUFFER_CONSOLE_PRIMARY if we have FRAMEBUFFER_CONSOLE
Merge branch 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6
* 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6:
nfs: don't lose MS_SYNCHRONOUS on remount of noac mount
NFS: Return meaningful status from decode_secinfo()
NFSv4: Ensure we request the ordinary fileid when doing readdirplus
NFSv4: Ensure that clientid and session establishment can time out
SUNRPC: Allow RPC calls to return ETIMEDOUT instead of EIO
NFSv4.1: Don't loop forever in nfs4_proc_create_session
NFSv4: Handle NFS4ERR_WRONGSEC outside of nfs4_handle_exception()
NFSv4.1: Don't update sequence number if rpc_task is not sent
NFSv4.1: Ensure state manager thread dies on last umount
SUNRPC: Fix the SUNRPC Kerberos V RPCSEC_GSS module dependencies
NFS: Use correct variable for page bounds checking
NFS: don't negotiate when user specifies sec flavor
NFS: Attempt mount with default sec flavor first
NFS: flav_array honors NFS_MAX_SECFLAVORS
NFS: Fix infinite loop in gss_create_upcall()
Don't mark_inode_dirty_sync() while holding lock
NFS: Get rid of pointless test in nfs_commit_done
NFS: Remove unused argument from nfs_find_best_sec()
NFS: Eliminate duplicate call to nfs_mark_request_dirty
NFS: Remove dead code from nfs_fs_mount()
mm: check if PTE is already allocated during page fault
With transparent hugepage support, handle_mm_fault() has to be careful
that a normal PMD has been established before handling a PTE fault. To
achieve this, it used __pte_alloc() directly instead of pte_alloc_map as
pte_alloc_map is unsafe to run against a huge PMD. pte_offset_map() is
called once it is known the PMD is safe.
pte_alloc_map() is smart enough to check if a PTE is already present
before calling __pte_alloc but this check was lost. As a consequence,
PTEs may be allocated unnecessarily and the page table lock taken. Thi
useless PTE does get cleaned up but it's a performance hit which is
visible in page_test from aim9.
This patch simply re-adds the check normally done by pte_alloc_map to
check if the PTE needs to be allocated before taking the page table lock.
The effect is noticable in page_test from aim9.
(While this affects 2.6.38, it is a performance rather than a functional
bug and normally outside the rules -stable. While the big performance
differences are to a microbench, the difference in fork and exec
performance may be significant enough that -stable wants to consider the
patch)
Reported-by: Raz Ben Yehuda <raziebe@gmail.com> Signed-off-by: Mel Gorman <mgorman@suse.de> Reviewed-by: Rik van Riel <riel@redhat.com> Reviewed-by: Andrea Arcangeli <aarcange@redhat.com> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: <stable@kernel.org> [2.6.38.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
kernel/watchdog.c: disable nmi perf event in the error path of enabling watchdog
In corner cases where softlockup watchdog is not setup successfully, the
relevant nmi perf event for hardlockup watchdog could be disabled, then
the status of the underlying hardware remains unchanged.
Also, if the kthread doesn't start then the hrtimer won't run and the
hardlockup detector will falsely fire.
Signed-off-by: Hillf Danton <dhillf@gmail.com> Signed-off-by: Don Zickus <dzickus@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Reported-by: Simon Danner <danner.simon@gmail.com> Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: Jeff Dike <jdike@addtoit.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Richard Weinberger <richard@nod.at> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This adds support for 64 bit atomic operations on 32 bit UML systems. XFS
needs them since 2.6.38.
$ make ARCH=um SUBARCH=i386
...
LD .tmp_vmlinux1
fs/built-in.o: In function `xlog_regrant_reserve_log_space':
xfs_log.c:(.text+0xd8584): undefined reference to `atomic64_read_386'
xfs_log.c:(.text+0xd85ac): undefined reference to `cmpxchg8b_emu'
...
PTE pages eat up memory just like anything else, but we do not account for
them in any way in the OOM scores. They are also _guaranteed_ to get
freed up when a process is OOM killed, while RSS is not.
Kukjin Kim [Wed, 27 Apr 2011 22:26:49 +0000 (15:26 -0700)]
MAINTAINERS: add EXYNOS ARM architectures
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com> Cc: Ben Dooks <ben-linux@fluff.org> Cc: Russell King <rmk@arm.linux.org.uk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
memcg: update documentation to describe usage_in_bytes
Since 569b846d ("memcg: coalesce uncharge during unmap/truncate"), we do
batched (delayed) uncharge at truncation/unmap. And since cdec2e42(memcg:
coalesce charging via percpu storage), we have percpu cache for
res_counter.
These changes improved performance of memory cgroup very much, but made
res_counter->usage usually have a bigger value than the actual value of
memory usage. So, *.usage_in_bytes, which show res_counter->usage, are
not desirable for precise values of memory(and swap) usage anymore.
Instead of removing these files completely(because we cannot know
res_counter->usage without them), this patch updates the meaning of those
files.
Andrea Arcangeli [Wed, 27 Apr 2011 22:26:45 +0000 (15:26 -0700)]
mm: thp: fix /dev/zero MAP_PRIVATE and vm_flags cleanups
The huge_memory.c THP page fault was allowed to run if vm_ops was null
(which would succeed for /dev/zero MAP_PRIVATE, as the f_op->mmap wouldn't
setup a special vma->vm_ops and it would fallback to regular anonymous
memory) but other THP logics weren't fully activated for vmas with vm_file
not NULL (/dev/zero has a not NULL vma->vm_file).
So this removes the vm_file checks so that /dev/zero also can safely use
THP (the other albeit safer approach to fix this bug would have been to
prevent the THP initial page fault to run if vm_file was set).
After removing the vm_file checks, this also makes huge_memory.c stricter
in khugepaged for the DEBUG_VM=y case. It doesn't replace the vm_file
check with a is_pfn_mapping check (but it keeps checking for VM_PFNMAP
under VM_BUG_ON) because for a is_cow_mapping() mapping VM_PFNMAP should
only be allowed to exist before the first page fault, and in turn when
vma->anon_vma is null (so preventing khugepaged registration). So I tend
to think the previous comment saying if vm_file was set, VM_PFNMAP might
have been set and we could still be registered in khugepaged (despite
anon_vma was not NULL to be registered in khugepaged) was too paranoid.
The is_linear_pfn_mapping check is also I think superfluous (as described
by comment) but under DEBUG_VM it is safe to stay.
Andrew Morton [Wed, 27 Apr 2011 22:26:41 +0000 (15:26 -0700)]
vfs: avoid large kmalloc()s for the fdtable
Azurit reports large increases in system time after 2.6.36 when running
Apache. It was bisected down to a892e2d7dcdfa6c76e6 ("vfs: use kmalloc()
to allocate fdmem if possible").
That patch caused the vfs to use kmalloc() for very large allocations and
this is causing excessive work (and presumably excessive reclaim) within
the page allocator.
Fix it by falling back to vmalloc() earlier - when the allocation attempt
would have been considered "costly" by reclaim.
We use io_apic_setup_irq_pin() in order to configure pin's interrupt
number polarity and type. This is done on every irq_create_of_mapping()
which happens for instance during pci enable calls. Level typed
interrupts are masked by default, edge are unmasked.
On the first ->xlate() call the level interrupt is configured and
masked. The driver calls request_irq() and the line is unmasked. Lets
assume the interrupt line is shared with another device and we call
pci_enable_device() for this device. The ->xlate() configures the pin
again and it is masked. request_irq() does not unmask the line because
it _is_ already unmasked according to its internal state. So the
interrupt will never be unmasked again.
This patch is based on an earlier work by Torben Hohn and solves the
problem by configuring the pin only once. Since all devices must agree
on the same type and polarity there is no point in configuring the pin
more than once.
[ tglx: Split out the ce4100 part into a separate patch ]
Merge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/parisc-2.6
* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/parisc-2.6:
[PARISC] slub: fix panic with DISCONTIGMEM
[PARISC] set memory ranges in N_NORMAL_MEMORY when onlined
Merge branch 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6
* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6: (42 commits)
[media] media: vb2: correct queue initialization order
[media] media: vb2: fix incorrect v4l2_buffer->flags handling
[media] s5p-fimc: Add support for the buffer timestamps and sequence
[media] s5p-fimc: Fix bytesperline and plane payload setup
[media] s5p-fimc: Do not allow changing format after REQBUFS
[media] s5p-fimc: Fix FIMC3 pixel limits on Exynos4
[media] tda18271: update tda18271c2_rf_cal as per NXP's rev.04 datasheet
[media] tda18271: update tda18271_rf_band as per NXP's rev.04 datasheet
[media] tda18271: fix bad calculation of main post divider byte
[media] tda18271: prog_cal and prog_tab variables should be s32, not u8
[media] tda18271: fix calculation bug in tda18271_rf_tracking_filters_init
[media] omap3isp: queue: Don't corrupt buf->npages when get_user_pages() fails
[media] v4l: Don't register media entities for subdev device nodes
[media] omap3isp: Don't increment node entity use count when poweron fails
[media] omap3isp: lane shifter support
[media] omap3isp: ccdc: support Y10/12, 8-bit bayer fmts
[media] media: add missing 8-bit bayer formats and Y12
[media] v4l: add V4L2_PIX_FMT_Y12 format
cx23885: Fix stv0367 Kconfig dependency
[media] omap3isp: Use isp xclk defines
...
Fix up trivial conflict (spelink errurs) in drivers/media/video/omap3isp/isp.c
Jeff Layton [Wed, 27 Apr 2011 15:49:09 +0000 (11:49 -0400)]
nfs: don't lose MS_SYNCHRONOUS on remount of noac mount
On a remount, the VFS layer will clear the MS_SYNCHRONOUS bit on the
assumption that the flags on the mount syscall will have it set if the
remounted fs is supposed to keep it.
In the case of "noac" though, MS_SYNCHRONOUS is implied. A remount of
such a mount will lose the MS_SYNCHRONOUS flag since "sync" isn't part
of the mount options.
Reported-by: Max Matveev <makc@redhat.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Cc: stable@kernel.org Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
NFS: Return meaningful status from decode_secinfo()
When compiling, I was getting this warning:
fs/nfs/nfs4xdr.c: In function ‘decode_secinfo’:
fs/nfs/nfs4xdr.c:4839:6: warning: variable ‘status’ set but not used
[-Wunused-but-set-variable]
We were unconditionally returning 0 as long as there wasn't an error
coming out of xdr_inline_decode(). We probably want to check the error
status coming out of decode_op_hdr() and decode_secinfo_gss(), rather
than assuming that everything is OK all the time.
NFSv4: Ensure we request the ordinary fileid when doing readdirplus
When readdir() returns a directory entry for the root of a mounted
filesystem, Linux follows the old convention of returning the inode
number of the covered directory (despite newer versions of POSIX declaring
that this is a bug).
To ensure this continues to work, the NFSv4 readdir implementation requests
the 'mounted-on-fileid' from the server.
However, readdirplus also needs to instantiate an inode for this entry, and
for that, we also need to request the real fileid as per this patch.
Don Zickus [Wed, 27 Apr 2011 10:32:33 +0000 (06:32 -0400)]
perf, x86, nmi: Move LVT un-masking into irq handlers
It was noticed that P4 machines were generating double NMIs for
each perf event. These extra NMIs lead to 'Dazed and confused'
messages on the screen.
I tracked this down to a P4 quirk that said the overflow bit had
to be cleared before re-enabling the apic LVT mask. My first
attempt was to move the un-masking inside the perf nmi handler
from before the chipset NMI handler to after.
This broke Nehalem boxes that seem to like the unmasking before
the counters themselves are re-enabled.
In order to keep this change simple for 2.6.39, I decided to
just simply move the apic LVT un-masking to the beginning of all
the chipset NMI handlers, with the exception of Pentium4's to
fix the double NMI issue.
Later on we can move the un-masking to later in the handlers to
save a number of 'extra' NMIs on those particular chipsets.
I tested this change on a P4 machine, an AMD machine, a Nehalem
box, and a core2quad box. 'perf top' worked correctly along
with various other small 'perf record' runs. Anything high
stress breaks all the machines but that is a different problem.
Thanks to various people for testing different versions of this
patch.
Michael Schmitz [Tue, 26 Apr 2011 02:51:53 +0000 (14:51 +1200)]
m68k/mm: Set all online nodes in N_NORMAL_MEMORY
For m68k, N_NORMAL_MEMORY represents all nodes that have present memory
since it does not support HIGHMEM. This patch sets the bit at the time
node_present_pages has been set by free_area_init_node.
At the time the node is brought online, the node state would have to be
done unconditionally since information about present memory has not yet
been recorded.
If N_NORMAL_MEMORY is not accurate, slub may encounter errors since it
uses this nodemask to setup per-cache kmem_cache_node data structures.
This pach is an alternative to the one proposed by David Rientjes
<rientjes@google.com> attempting to set node state immediately when
bringing the node online.
Dave Airlie [Thu, 21 Apr 2011 21:18:32 +0000 (22:18 +0100)]
drm/i915: restore only the mode of this driver on lastclose (v2)
i915 calls the panic handler function on last close to reset the modes,
however this is a really bad idea for multi-gpu machines, esp shareable
gpus machines. So add a new entry point for the driver to just restore
its own fbcon mode.
v2: move code into fb helper, fix panic code to block mode change on
powered off GPUs.
[airlied: this hits drm core and I wrote it and it was reviewed on intel-gfx
so really I signed it off twice ;-).] Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Dave Airlie <airlied@redhat.com>
Randy Dunlap [Tue, 26 Apr 2011 19:33:21 +0000 (12:33 -0700)]
init/Kconfig: fix EXPERT menu list
The EXPERT menu list was recently broken by the insertion of a
kconfig symbol (EMBEDDED) at the beginning of the EXPERT list of
kconfig items. Broken by:
commit 6a108a14fa356ef607be308b68337939e56ea94e
Author: David Rientjes <rientjes@google.com>
Date: Thu Jan 20 14:44:16 2011 -0800
kconfig: rename CONFIG_EMBEDDED to CONFIG_EXPERT
Restore the EXPERT menu list -- don't inject a symbol (EMBEDDED)
that does not depend on EXPERT into the list.
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: David Rientjes <rientjes@google.com> Cc: Peter Foley <pefoley2@verizon.net> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge branch 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6
* 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6:
[S390] kvm-390: Let kernel exit SIE instruction on work
[S390] dasd: check sense type in device change handler
[S390] pfault: fix token handling
[S390] qdio: reset error states immediately
[S390] fix page table walk for changing page attributes
[S390] prng: prevent access beyond end of stack
[S390] dasd: fix race between open and offline
perf events, x86: Work around the Nehalem AAJ80 erratum
On Nehalem CPUs the retired branch-misses event can be completely bogus,
when there are no branch-misses occuring. When there are a lot of branch
misses then the count is pretty accurate. Still, this leaves us with an
event that over-counts a lot.
Detect this erratum and work it around by using BR_MISP_EXEC.ANY events.
These will also count speculated branches but still it's a lot more
precise in practice than the architectural event.
* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
Btrfs: cleanup error handling in inode.c
Btrfs: put the right bio if we have an error
Btrfs: free bitmaps properly when evicting the cache
Btrfs: Free free_space item properly in btrfs_trim_block_group()
btrfs: add missing spin_unlock to a rare exit path
Btrfs: check return value of kmalloc()
btrfs: fix wrong allocating flag when reading page
Btrfs: fix missing mutex_unlock in btrfs_del_dir_entries_in_log()
Borislav Petkov [Wed, 30 Mar 2011 13:42:10 +0000 (15:42 +0200)]
amd64_edac: Erratum #637 workaround
F15h CPUs may report a non-DRAM address when reporting an error address
belonging to a CC6 state save area. Add a workaround to detect this
condition and compute the actual DRAM address of the error as documented
in the Revision Guide for AMD Family 15h Models 00h-0Fh Processors.
Borislav Petkov [Mon, 21 Mar 2011 19:45:06 +0000 (20:45 +0100)]
amd64_edac: Factor in CC6 save area
F15h and later use a portion of DRAM as a CC6 storage area. BIOS
programs D18F1x[17C:140,7C:40] DRAM Base/Limit accordingly by
subtracting the storage area from the DRAM limit setting. However, in
order for edac to consider that part of DRAM too, we need to include it
into the per-node range.
This warning was wrongfully added for a normal condition - intlvsel
actually selects the destination node when node interleaving is enabled
and it is not a mismatch. For a detailed example, see section 2.8.10.2
"Node Interleaving" in F10h BKDG.
Peter Zijlstra [Tue, 26 Apr 2011 11:24:33 +0000 (13:24 +0200)]
perf, x86: Fix BTS condition
Currently the x86 backend incorrectly assumes that any BRANCH_INSN
with sample_period==1 is a BTS request. This is not true when we do
frequency driven profiling such as 'perf record -e branches'.
Solves this error:
$ perf record -e branches ./array
Error: sys_perf_event_open() syscall returned with 95 (Operation not supported).
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Reported-by: Ingo Molnar <mingo@elte.hu> Cc: "Metzger, Markus T" <markus.t.metzger@intel.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Link: http://lkml.kernel.org/n/tip-rd2y4ct71hjawzz6fpvsy9hg@git.kernel.org Signed-off-by: Ingo Molnar <mingo@elte.hu>
Blindly setting 1.2V in the initial structure may not even match the
default voltages stored in the voltage table which are supported for
the domain. For example, OMAP3430 core domain does not use 1.2V and
ends up generating a warning on the first transition.
Further, since omap2_set_init_voltage is called as part of the pm
framework's initialization sequence to configure the voltage required
for the current OPP, the call does(and has to) setup the system
voltage(curr_volt as a result) using the right mechanisms appropriate
for the system at that point of time. This also overrides
initialization we are currently doing in voltage.c making it
redundant. So, remove the wrong and redundant initialization.
Signed-off-by: Nishanth Menon <nm@ti.com> Signed-off-by: Kevin Hilman <khilman@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
OMAP4: Intialize IVA Device in addition to DSP device.
OMAP4 has two different Devices IVA and DSP. DSP is bound
with IVA for DVFS. The registration of IVA dev in API
'omap2_init_processor_devices' was missing. Init dev for
'iva_dev' is added.
This also fixes the following error seen during boot as
omap2_set_init_voltage can now find the iva device
omap2_set_init_voltage: Invalid parameters!
omap2_set_init_voltage: Unable to put vdd_iva to its init voltage
Felipe Contreras [Tue, 26 Apr 2011 09:45:28 +0000 (02:45 -0700)]
omap: rx51: mark reserved memory earlier
So that omap_vram_set_sdram_vram() is called before
omap_vram_reserve_sdram_memblock().
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com> Acked-by: Tomi Valkeinen <tomi.valkeinen@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
ACPI / PM: Avoid infinite recurrence while registering power resources
There is at least one BIOS with a DSDT containing a power resource
object with a _PR0 entry pointing back to that power resource. In
consequence, while registering that power resource
acpi_bus_get_power_flags() sees that it depends on itself and tries
to register it again, which leads to an infinitely deep recurrence.
This problem was introduced by commit bf325f9538d8c89312be305b9779e
(ACPI / PM: Register power resource devices as soon as they are
needed).
To fix this problem use the observation that power resources cannot
be power manageable and prevent acpi_bus_get_power_flags() from
being called for power resource objects.
References: https://bugzilla.kernel.org/show_bug.cgi?id=31872 Reported-and-tested-by: Pascal Dormeau <pdormeau@free.fr> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Acked-by: Len Brown <lenb@kernel.org> Cc: stable@kernel.org
PM / Wakeup: Fix initialization of wakeup-related device sysfs files
It turns out that some PCI devices are only found to be
wakeup-capable during registration, in which case, when
device_set_wakeup_capable() is called, device_is_registered() already
returns 'true' for the given device, but dpm_sysfs_add() hasn't been
called for it yet. This leads to situations in which the device's
power.can_wakeup flag is not set as requested because of failing
wakeup_sysfs_add() and its wakeup-related sysfs files are not
created, although they should be present. This is a post-2.6.38
regression introduced by commit cb8f51bdadb7969139c2e39c2defd4cde98c1
(PM: Do not create wakeup sysfs files for devices that cannot wake
up).
To work around this problem initialize the device's power.entry
field to an empty list head and make device_set_wakeup_capable()
check if it is still empty before attempting to add the devices
wakeup-related sysfs files with wakeup_sysfs_add(). Namely, if
power.entry is still empty at this point, device_pm_add() hasn't been
called yet for the device and its wakeup-related files will be
created later, so device_set_wakeup_capable() doesn't have to create
them.
Reported-and-tested-by: Tino Keitel <tino.keitel@tikei.de> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
If an error occurs in the L3 on any other initiator than MPU,
the interrupt goes unhandled given that the 'base' register
was calculated with the initialized err_source value (which
coincidentally points to MPU) and not with the actual source
of the error.
Removed parenthesis that are not needed for the touched lines.
Signed-off-by: Omar Ramirez Luna <omar.ramirez@ti.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ecryptfs/ecryptfs-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ecryptfs/ecryptfs-2.6:
eCryptfs: Flush dirty pages in setattr
eCryptfs: Handle failed metadata read in lookup
eCryptfs: Add reference counting to lower files
eCryptfs: dput dentries returned from dget_parent
eCryptfs: Remove extra d_delete in ecryptfs_rmdir
Eric Paris [Mon, 25 Apr 2011 20:26:29 +0000 (16:26 -0400)]
SELINUX: Make selinux cache VFS RCU walks safe
Now that the security modules can decide whether they support the
dcache RCU walk or not it's possible to make selinux a bit more
RCU friendly. The SELinux AVC and security server access decision
code is RCU safe. A specific piece of the LSM audit code may not
be RCU safe.
This patch makes the VFS RCU walk retry if it would hit the non RCU
safe chunk of code. It will normally just work under RCU. This is
done simply by passing the VFS RCU state as a flag down into the
avc_audit() code and returning ECHILD there if it would have an issue.
Based-on-patch-by: Andi Kleen <ak@linux.intel.com> Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that the whole dcache_hash_bucket crap is gone, go all the way and
also remove the weird locking layering violations for locking the hash
buckets. Add hlist_bl_lock/unlock helpers to move the locking into the
list abstraction instead of requiring each caller to open code it.
After all allowing for the bit locks is the whole point of these helpers
over the plain hlist variant.
Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
bit_spinlock: don't play preemption games inside the busy loop
When we are waiting for the bit-lock to be released, and are looping
over the 'cpu_relax()' should not be doing anything else - otherwise we
miss the point of trying to do the whole 'cpu_relax()'.
Do the preemption enable/disable around the loop, rather than inside of
it.
Noticed when I was looking at the code generation for the dcache
__d_drop usage, and the code just looked very odd.
After 57db4e8d73ef2b5e94a3f412108dff2576670a8a changed eCryptfs to
write-back caching, eCryptfs page writeback updates the lower inode
times due to the use of vfs_write() on the lower file.
To preserve inode metadata changes, such as 'cp -p' does with
utimensat(), we need to flush all dirty pages early in
ecryptfs_setattr() so that the user-updated lower inode metadata isn't
clobbered later in writeback.
Tyler Hicks [Tue, 15 Mar 2011 19:54:00 +0000 (14:54 -0500)]
eCryptfs: Handle failed metadata read in lookup
When failing to read the lower file's crypto metadata during a lookup,
eCryptfs must continue on without throwing an error. For example, there
may be a plaintext file in the lower mount point that the user wants to
delete through the eCryptfs mount.
If an error is encountered while reading the metadata in lookup(), the
eCryptfs inode's size could be incorrect. We must be sure to reread the
plaintext inode size from the metadata when performing an open() or
setattr(). The metadata is already being read in those paths, so this
adds minimal performance overhead.
This patch introduces a flag which will track whether or not the
plaintext inode size has been read so that an incorrect i_size can be
fixed in the open() or setattr() paths.
Josef Bacik [Mon, 25 Apr 2011 23:43:52 +0000 (19:43 -0400)]
Btrfs: free bitmaps properly when evicting the cache
If our space cache is wrong, we do the right thing and free up everything that
we loaded, however we don't reset the total_bitmaps counter or the thresholds or
anything. So in btrfs_remove_free_space_cache make sure to call free_bitmap()
if it's a bitmap, this will keep us from panicing when we check to make sure we
don't have too many bitmaps. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com> Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs: fix wrong allocating flag when reading page
the space cache use extent_readpages() to read free space information,
so we can not use GFP_KERNEL flag to allocate memory, or it may lead
to deadlock.
For any given lower inode, eCryptfs keeps only one lower file open and
multiplexes all eCryptfs file operations through that lower file. The
lower file was considered "persistent" and stayed open from the first
lookup through the lifetime of the inode.
This patch keeps the notion of a single, per-inode lower file, but adds
reference counting around the lower file so that it is closed when not
currently in use. If the reference count is at 0 when an operation (such
as open, create, etc.) needs to use the lower file, a new lower file is
opened. Since the file is no longer persistent, all references to the
term persistent file are changed to lower file.
Locking is added around the sections of code that opens the lower file
and assign the pointer in the inode info, as well as the code the fputs
the lower file when all eCryptfs users are done with it.
This patch is needed to fix issues, when mounted on top of the NFSv3
client, where the lower file is left silly renamed until the eCryptfs
inode is destroyed.
Call dput on the dentries previously returned by dget_parent() in
ecryptfs_rename(). This is needed for supported eCryptfs mounts on top
of the NFSv3 client.
vfs_rmdir() already calls d_delete() on the lower dentry. That was being
duplicated in ecryptfs_rmdir() and caused a NULL pointer dereference
when NFSv3 was the lower filesystem.
H. Peter Anvin [Mon, 25 Apr 2011 21:52:37 +0000 (14:52 -0700)]
x86, setup: When probing memory with e801, use ax/bx as a pair
When we use BIOS function e801 to probe memory, we should use ax/bx
(or cx/dx) as a pair, not mix and match. This was a typo during the
translation from assembly code, and breaks at least one set of
machines in the field (which return cx = dx = 0).
NFSv4: Ensure that clientid and session establishment can time out
The following patch ensures that we do not get permanently trapped in
the RPC layer when trying to establish a new client id or session.
This again ensures that the state manager can finish in a timely
fashion when the last filesystem to reference the nfs_client exits.
SUNRPC: Allow RPC calls to return ETIMEDOUT instead of EIO
On occasion, it is useful for the NFS layer to distinguish between
soft timeouts and other EIO errors due to (say) encoding errors,
or authentication errors.
The following patch ensures that the default behaviour of the RPC
layer remains to return EIO on soft timeouts (until we have
audited all the callers).
NFSv4.1: Don't loop forever in nfs4_proc_create_session
If a server for some reason keeps sending NFS4ERR_DELAY errors, we can end
up looping forever inside nfs4_proc_create_session, and so the usual
mechanisms for detecting if the nfs_client is dead don't work.
Fix this by ensuring that we loop inside the nfs4_state_manager thread
instead.
Dan Rosenberg [Tue, 5 Apr 2011 17:27:31 +0000 (13:27 -0400)]
[SCSI] pmcraid: reject negative request size
There's a code path in pmcraid that can be reached via device ioctl that
causes all sorts of ugliness, including heap corruption or triggering
the OOM killer due to consecutive allocation of large numbers of pages.
Not especially relevant from a security perspective, since users must
have CAP_SYS_ADMIN to open the character device.
First, the user can call pmcraid_chr_ioctl() with a type
PMCRAID_PASSTHROUGH_IOCTL. A pmcraid_passthrough_ioctl_buffer
is copied in, and the request_size variable is set to
buffer->ioarcb.data_transfer_length, which is an arbitrary 32-bit signed
value provided by the user.
If a negative value is provided here, bad things can happen. For
example, pmcraid_build_passthrough_ioadls() is called with this
request_size, which immediately calls pmcraid_alloc_sglist() with a
negative size. The resulting math on allocating a scatter list can
result in an overflow in the kzalloc() call (if num_elem is 0, the
sglist will be smaller than expected), or if num_elem is unexpectedly
large the subsequent loop will call alloc_pages() repeatedly, a high
number of pages will be allocated and the OOM killer might be invoked.
Prevent this value from being negative in pmcraid_ioctl_passthrough().
Signed-off-by: Dan Rosenberg <drosenberg@vsecurity.com> Cc: stable@kernel.org Cc: Anil Ravindranath <anil_ravindranath@pmc-sierra.com> Signed-off-by: James Bottomley <James.Bottomley@suse.de>
James Bottomley [Fri, 22 Apr 2011 15:39:59 +0000 (10:39 -0500)]
[SCSI] put stricter guards on queue dead checks
SCSI uses request_queue->queuedata == NULL as a signal that the queue
is dying. We set this state in the sdev release function. However,
this allows a small window where we release the last reference but
haven't quite got to this stage yet and so something will try to take
a reference in scsi_request_fn and oops. It's very rare, but we had a
report here, so we're pushing this as a bug fix
The actual fix is to set request_queue->queuedata to NULL in
scsi_remove_device() before we drop the reference. This causes
correct automatic rejects from scsi_request_fn as people who hold
additional references try to submit work and prevents anything from
getting a new reference to the sdev that way.
Cc: stable@kernel.org Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Mike Snitzer [Fri, 8 Apr 2011 19:05:36 +0000 (15:05 -0400)]
[SCSI] scsi_dh: fix reference counting in scsi_dh_activate error path
Commit db422318cbca55168cf965f655471dbf8be82433 ([SCSI] scsi_dh:
propagate SCSI device deletion) introduced a regression where the device
reference is not dropped prior to scsi_dh_activate's early return from
the error path.
Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@kernel.org # 2.6.38 Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <James.Bottomley@suse.de>
projects / mirror_ubuntu-eoan-kernel.git / log