People have been reporting ENOSPC crashes in finish_ordered_io. This is because
we try to steal from the delalloc block rsv to satisfy a reservation to update
the inode. The problem with this is we don't explicitly save space for updating
the inode when doing delalloc. This is kind of a problem and we've gotten away
with this because way back when we just stole from the delalloc reserve without
any questions, and this worked out fine because generally speaking the leaf had
been modified either by the mtime update when we did the original write or
because we just updated the leaf when we inserted the file extent item, only on
rare occasions had the leaf not actually been modified, and that was still ok
because we'd just use a block or two out of the over-reservation that is
delalloc.
Then came the delayed inode stuff. This is amazing, except it wants a full
reservation for updating the inode since it may do it at some point down the
road after we've written the blocks and we have to recow everything again. This
worked out because the delayed inode stuff just stole from the global reserve,
that is until recently when I changed that because it caused other problems.
So here we are, we're doing everything right and being screwed for it. So take
an extra reservation for the inode at delalloc reservation time and carry it
through the life of the delalloc reservation. If we need it we can steal it in
the delayed inode stuff. If we have already stolen it try and do a normal
metadata reservation. If that fails try to steal from the delalloc reservation.
If _that_ fails we'll get a WARN_ON() so I can start thinking of a better way to
solve this and in the meantime we'll steal from the global reserve.
With this patch I ran xfstests 13 in a loop for a couple of hours and didn't see
any problems.
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Otherwise we get compile errors like this:
ERROR: "led_classdev_unregister" [drivers/platform/x86/dell-laptop.ko] undefined!
ERROR: "led_classdev_register" [drivers/platform/x86/dell-laptop.ko] undefined!
make[1]: *** [__modpost] Error 1
make: *** [modules] Error 2
when the dell-laptop support is enabled without the necessary LED
support being enabled.
Reported-by: Alessandro Suardi <alessandro.suardi@gmail.com>
Acked-by: Matthew Garrett <mjg59@srcf.ucam.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This file uses core functions like module_init() and module_exit()
and so it explicitly needs to include the module.h header.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
If we fail to reserve space in the transaction during truncate, we can
error out with a NULL trans handle. The cleanup code needs an extra
check to make sure we aren't trying to use the bad handle.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Alek Du reported that the code erroneously applies time to jiffies
conversions twice to the t1 and t2 values. In normal use on a modem link
this cases no visible problem but on a slower link it will break as with
HZ=1000 as is typical we are running t1/t2 ten times too fast.
Alek's original patch removed the conversion from the timer setting but we
in fact have to be more careful as the contents of t1/t2 are visible via
the device API and we thus need to correct the constants.
Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
CONFIG_MSM_SCM uses the smc instruction, which with some
toolchains requires a ".arch_extension" directive.
Cc: David Brown <davidb@codeaurora.org>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: David Brown <davidb@codeaurora.org>
In the recent usb-audio driver, the initialization of volume ranges
may be delayed when the device doesn't respond well at the probing time.
But the volume quirks for certain devices are applied only in
mixer_ctl_feature_info() thus only at the very first probe and will be
missing when the volume range is initialized later.
This patch moves the volume quirk code to be always called from the
volume-range extraction (get_min_max()), so that the quirks are properly
applied in the later init time.
Reported-and-tested-by: Alexey Fisher <bug-track@fisher-privat.net>
Cc: <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Ensure ioend->io_error gets propagated back to e.g. AIO completions.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
The log item ops aren't nessecarily the biggest exploit vector, but marking
them const is easy enough. Also remove the unused xfs_item_ops_t typedef
while we're at it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
Fixes a possible memory corruption when the link is larger than
MAXPATHLEN and XFS_DEBUG is not enabled. This also remove the
S_ISLNK assert, since the inode mode is checked previously in
xfs_readlink_by_handle() and via VFS.
Updated to address concerns raised by Ben Hutchings about the loose
attention paid to 32- vs 64-bit values, and the lack of handling a
potentially negative pathlen value:
- Changed type of "pathlen" to be xfs_fsize_t, to match that of
ip->i_d.di_size
- Added checking for a negative pathlen to the too-long pathlen
test, and generalized the message that gets reported in that case
to reflect the change
As a result, if a negative pathlen were encountered, this function
would return EFSCORRUPTED (and would fail an assertion for a debug
build)--just as would a too-long pathlen.
Signed-off-by: Alex Elder <aelder@sgi.com>
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Fix compile failure in drivers/iommu/omap-iommu-debug.c
because of missing module.h include.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
The merge commit "526b264 Merge branch 'imx/cleanup' into imx/devel"
left a duplicated mx28_add_mxs_mmc() call, which causes the problem
below during boot.
kobject_add_internal failed for mxs-mmc.1 with -EEXIST, don't try
to register things with the same name in the same directory.
The patch removes this leftover and also change mmc0 adding to align
with mmc1.
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Otherwise the fh changing the master control won't get the inactive state
change event for the slave controls.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Setting sev->fh to NULL causes problems for the del op added in the next
patch of this series, since this op needs a way to get to its own data
structures, and typically this will be done by using container_of on an
embedded v4l2_fh struct.
The reason the original code is setting sev->fh to NULL is to signal
to users of the event framework that the unsubscription has happened,
but since their is no shared lock between the event framework and users
of it, this is inherently racy, and it also turns out to be unnecessary
as long as both the event framework and the user of the framework do their
own locking properly and the user guarantees that it holds no references
to the subcribed_event structure after its del operation has been called.
This is best explained by looking at the only code currently checking for
sev->fh being set to NULL on unsubscribe, which is the v4l2-ctrls.c send_event
function. Here is the relevant code from v4l2-ctrls: send_event():
if (sev->fh && (sev->fh != fh ||
(sev->flags & V4L2_EVENT_SUB_FL_ALLOW_FEEDBACK)))
v4l2_event_queue_fh(sev->fh, &ev);
Now lets say that v4l2_event_unsubscribe and v4l2-ctrls: send_event() race
on the same sev, then the following could happens:
1) send_event checks sev->fh, finds it is not NULL
<thread switch>
2) v4l2_event_unsubscribe sets sev->fh NULL
3) v4l2_event_unsubscribe calls v4l2_ctrls del_event function, this blocks
as the thread calling send_event holds the ctrl_lock
<thread switch>
4) send_event calls v4l2_event_queue_fh(sev->fh, &ev) which not is equivalent
to calling: v4l2_event_queue_fh(NULL, &ev)
5) oops, NULL pointer deref.
Now again without setting sev->fh to NULL in v4l2_event_unsubscribe and
without the (now senseless since always true) sev->fh != NULL check in
1) send_event is about to call v4l2_event_queue_fh(sev->fh, &ev)
<thread switch>
2) v4l2_event_unsubscribe removes sev->list from the fh->subscribed list
<thread switch>
3) send_event calls v4l2_event_queue_fh(sev->fh, &ev)
4) v4l2_event_queue_fh blocks on the fh_lock spinlock
<thread switch>
5) v4l2_event_unsubscribe unlocks the fh_lock spinlock
6) v4l2_event_unsubscribe calls v4l2_ctrls del_event function, this blocks
as the thread calling send_event holds the ctrl_lock
<thread switch>
8) v4l2_event_queue_fh takes the fh_lock
7) v4l2_event_queue_fh calls v4l2_event_subscribed, does not find it since
sev->list has been removed from fh->subscribed already -> does nothing
9) v4l2_event_queue_fh releases the fh_lock
10) the caller of send_event releases the ctrl lock (mutex)
<thread switch>
11) v4l2_ctrls del_event takes the ctrl lock
12) v4l2_ctrls del_event removes sev->node from the ev_subs list
13) v4l2_ctrls del_event releases the ctrl lock
14) v4l2_event_unsubscribe frees the sev, to which no references are being
held anymore
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The kev pointers inside the pending events queue (the available queue) of the
fh point to data inside the sev, unsubscribing frees the sev, thus making these
pointers point to freed memory!
This patch fixes these dangling pointers in the available queue by removing
all matching pending events on unsubscription.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
MFC uses the multi-plane API, but it reported single-plane
when querying capabilities.
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Queued list was not reset on REQBUFS(0) call. This caused to enqueue a
freed buffer to the driver.
Reported-by: Angela Wan <angela.j.wan@gmail.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
v4l2_planes[plane].length field was not initialized for userptr buffers.
This patch fixes this issue.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
CC: Pawel Osciak <pawel@osciak.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
__buffer_in_use() might be called for empty/uninitialized buffer in the
following scenario: REQBUF(n, USER_PTR), QUERYBUF(). This patch fixes
kernel ops in such case.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
CC: Pawel Osciak <pawel@osciak.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
I've been seeing memory leaks on my system in the form of large
(300-400MB) GEM objects created by now-dead processes laying around
clogging up memory. I usually notice when it gets to about 1.2GB of
them. Hopefully this clears up the issue, but I just found this bug
by inspection.
Signed-off-by: Eric Anholt <eric@anholt.net>
Cc: stable@kernel.org
Signed-off-by: Keith Packard <keithp@keithp.com>
Kernels with no iommu support cannot ever need the Ironlake
work-around, so never enable it in that case.
Might be better to completely remove the work-around from the kernel
in this case?
Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
The semi-colon is a typo here and it makes the if statement
unconditional.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
Fix below build error:
CC drivers/hwspinlock/u8500_hsem.o
drivers/hwspinlock/u8500_hsem.c: In function 'u8500_hsem_probe':
drivers/hwspinlock/u8500_hsem.c:113: error: label 'free_state' used but not defined
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
Fix below build warning:
CC arch/arm/mach-omap2/hwspinlock.o
In file included from arch/arm/mach-omap2/hwspinlock.c:22:
include/linux/hwspinlock.h: In function '__hwspin_unlock':
include/linux/hwspinlock.h:121: warning: 'return' with a value, in function returning void
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
When a Realtek codec has a matrix-style capture-source selection, we
need to scan all connections instead of only imux items. Otherwise some
input might be kept unmuted. Although the corresponding input must be
dead so there should be no input from it, it's still safer to mute the
route completely.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The recently merged 228cf79376 looks a bit hackish while it is not.
The change was quite simple. In a virtualized environment the
patch unhacks old kludge introduced for old broken AC97 hardware.
This patch adds proper comment to "unkludge" code.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Konstantin Ozerkov <kozerkov@parallels.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
All updates that occur under STRIPE_ACTIVE should be globally visible
when STRIPE_ACTIVE clears. test_and_set_bit() implies a barrier, but
clear_bit() does not.
This is suitable for 3.1-stable.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
When the number of failed devices exceeds the allowed number
we must abort any active parity operations (checks or updates) as they
are no longer meaningful, and can lead to a BUG_ON in
handle_parity_checks6.
This bug was introduce by commit 6c0069c0ae
in 2.6.29.
Reported-by: Manish Katiyar <mkatiyar@gmail.com>
Tested-by: Manish Katiyar <mkatiyar@gmail.com>
Acked-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
When running with HV KVM and CBE config options enabled, I get
build failures like the following:
arch/powerpc/kernel/head_64.o: In function `cbe_system_error_hv':
(.text+0x1228): undefined reference to `do_kvm_0x1202'
arch/powerpc/kernel/head_64.o: In function `cbe_maintenance_hv':
(.text+0x1628): undefined reference to `do_kvm_0x1602'
arch/powerpc/kernel/head_64.o: In function `cbe_thermal_hv':
(.text+0x1828): undefined reference to `do_kvm_0x1802'
This is because we jump to a KVM handler when HV is enabled, but we
only generate the handler with PR KVM mode.
Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
The lv1_gpu_attribute hcall takes three, not five input
arguments. Adjust the lv1 hcall table and all calls.
Signed-off-by: Geoff Levand <geoff@infradead.org>
CC: Takashi Iwai <tiwai@suse.de>
Acked-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Since commit [e58aa3d2: genirq: Run irq handlers with interrupts disabled],
We run all interrupt handlers with interrupts disabled
and we even check and yell when an interrupt handler
returns with interrupts enabled (see commit [b738a50a:
genirq: Warn when handler enables interrupts]).
So now this flag is a NOOP and can be removed.
Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Acked-by: Geoff Levand <geoff@infradead.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Since commit [e58aa3d2: genirq: Run irq handlers with interrupts disabled],
We run all interrupt handlers with interrupts disabled
and we even check and yell when an interrupt handler
returns with interrupts enabled (see commit [b738a50a:
genirq: Warn when handler enables interrupts]).
So now this flag is a NOOP and can be removed.
Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Geoff Levand <geoff@infradead.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
This patch adds support for numa topology on powernv platforms running
OPAL formware. It checks for the type of platform at run time and
sets the affinity form correctly so that NUMA topology can be discovered
correctly.
Signed-off-by: Dipankar Sarma <dipankar@in.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
We've resisted adding System RAM to /proc/iomem because it is
the wrong place for it. Unfortunately we continue to find tools
that rely on this behaviour so give up and add it in.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Add HV mode KVM to Book3 server 64bit defconfigs as a module.
Doesn't add much to the size:
text data bss dec hex filename
8244109 4686767 994000 13924876 d47a0c vmlinux.vanilla
8256092 4691607 994128 13941827 d4bc43 vmlinux.kvm
This should enable more testing of this configuration.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
These files had implicit dependencies on modular support
which now show up as build failures with the module cleanup
work merged to mainline.
Reported-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Include module.h to fix below build error:
CC drivers/hwspinlock/u8500_hsem.o
drivers/hwspinlock/u8500_hsem.c:177: error: 'THIS_MODULE' undeclared here (not in a function)
[...]
drivers/hwspinlock/u8500_hsem.c:196: warning: type defaults to 'int' in declaration of 'MODULE_AUTHOR'
drivers/hwspinlock/u8500_hsem.c:196: warning: function declaration isn't a prototype
make[2]: *** [drivers/hwspinlock/u8500_hsem.o] Error 1
make[1]: *** [drivers/hwspinlock] Error 2
make: *** [drivers] Error 2
Signed-off-by: Axel Lin <axel.lin@gmail.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
module.h was previously implicitly included through mtd/mtd.h.
Fixes the following build failure after the module.h cleanup:
CC drivers/mtd/maps/bcm963xx-flash.o
drivers/mtd/maps/bcm963xx-flash.c: In function 'bcm963xx_probe':
drivers/mtd/maps/bcm963xx-flash.c:208:29: error: 'THIS_MODULE' undeclared (first use in this function)
[...]
drivers/mtd/maps/bcm963xx-flash.c:276:1: warning: type defaults to 'int' in declaration of 'MODULE_AUTHOR'
drivers/mtd/maps/bcm963xx-flash.c:276:15: warning: function declaration isn't a prototype
make[7]: *** [drivers/mtd/maps/bcm963xx-flash.o] Error 1
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
.. with new name. Because nothing says "really solid kernel release"
like naming it after an extinct animal that just happened to be in the
news lately.
Mountpoint crossing is similar to following procfs symlinks - we do
not get ->d_revalidate() called for dentry we have arrived at, with
unpleasant consequences for NFS4.
Simple way to reproduce the problem in mainline:
cat >/tmp/a.c <<'EOF'
#include <unistd.h>
#include <fcntl.h>
#include <stdio.h>
main()
{
struct flock fl = {.l_type = F_RDLCK, .l_whence = SEEK_SET, .l_len = 1};
if (fcntl(0, F_SETLK, &fl))
perror("setlk");
}
EOF
cc /tmp/a.c -o /tmp/test
then on nfs4:
mount --bind file1 file2
/tmp/test < file1 # ok
/tmp/test < file2 # spews "setlk: No locks available"...
What happens is the missing call of ->d_revalidate() after mountpoint
crossing and that's where NFS4 would issue OPEN request to server.
The fix is simple - treat mountpoint crossing the same way we deal with
following procfs-style symlinks. I.e. set LOOKUP_JUMPED...
Cc: stable@kernel.org
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds