we used to need to clean it in RCU callback freeing an inode;
in 3.2 that requirement went away.  Unfortunately, it hadn't
been reflected in Documentation/filesystems/porting.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

we'll initialize it in inode_init_always() when we allocate that
object again.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

don't rely on proc_mounts->m being the first field; container_of()
is there for purpose.  No need to bother with ->private, while
we are at it - the same container_of will do nicely.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

it's enough to set ->mnt_ns of internal vfsmounts to something
distinct from all struct mnt_namespace out there; then we can
just use the check for ->mnt_ns != NULL in the fast path of
mntput_no_expire()
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

READ is 0, so the result of the bit-and operation is 0.  Rewrite with == as
done elsewhere in the same file.

This problem was found using Coccinelle (http://coccinelle.lip6.fr/).
Signed-off-by: Julia Lawall <julia@diku.dk>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

This patch makes affs stop using the VFS '->write_super()' method along with
the 's_dirt' superblock flag, because they are on their way out.

The whole "superblock write-out" VFS infrastructure is served by the
'sync_supers()' kernel thread, which wakes up every 5 (by default) seconds and
writes out all dirty superblocks using the '->write_super()' call-back.  But the
problem with this thread is that it wastes power by waking up the system every
5 seconds, even if there are no diry superblocks, or there are no client
file-systems which would need this (e.g., btrfs does not use
'->write_super()'). So we want to kill it completely and thus, we need to make
file-systems to stop using the '->write_super()' VFS service, and then remove
it together with the kernel thread.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

Add an 'sb' VFS superblock back-reference to the 'struct affs_sb_info' data
structure - we will need to find the VFS superblock from a 'struct
affs_sb_info' object in the next patch, so this change is jut a preparation.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

The VFS's 'lock_super()' and 'unlock_super()' calls are deprecated and unwanted
and just wait for a brave knight who'd kill them. This patch makes AFFS stop
using them and use the buffer-head's own lock instead.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

AFFS wants to serialize the superblock (the root block in AFFS terms) updates
and uses 'lock_super()/unlock_super()' for these purposes. This patch pushes the
locking down to the 'affs_commit_super()' from the callers.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

We do not need to write out the superblock from '->remount_fs()' because
VFS has already called '->sync_fs()' by this time and the superblock has
already been written out. Thus, remove the 'affs_write_super()'
infocation from 'affs_remount()'.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

We do not need to write out the superblock from '->put_super()' because VFS has
already called '->sync_fs()' by this time and the superblock has already been
written out. Thus, remove the 'affs_commit_super()' infocation from
'affs_put_super()'.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

AFFS stores values '1' and '2' in 'bm_flags', and I fail to see any logic when
it prefers one or another. AFFS writes '1' only from '->put_super()', while
'->sync_fs()' and '->write_super()' store value '2'.  So on the first glance,
it looks like we want to have '1' if we unmount.  However, this does not really
happen in these cases:
  1. superblock is written via 'write_super()' then we unmount;
  2. we re-mount R/O, then unmount.
which are quite typical.

I could not find good documentation describing this field, except of one random
piece of documentation in the internet which says that -1 means that the root
block is valid, which is not consistent with what we have in the Linux AFFS
driver.

Jan Kara commented on this: "I have some vague recollection that on Amiga
boolean was usually encoded as: 0 == false, ~0 == -1 == true. But it has been
ages..."

Thus, my conclusion is that value of '1' is as good as value of '2' and we can
just always use '2'. An Jan Kara suggested to go further: "generally bm_flags
handling looks strange. If they are 0, we mount fs read only and thus cannot
change them.  If they are != 0, we write 2 there. So IMHO if you just removed
bm_flags setting, nothing will really happen."

So this patch removes the bm_flags setting completely. This makes the "clean"
argument of the 'affs_commit_super()' function unneeded, so it is also removed.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

Pull use-after-free RAID1 bugfix from NeilBrown.

* tag 'md-3.5-fixes' of git://neil.brown.name/md:
  md/raid1: fix use-after-free bug in RAID1 data-check code.

Pull the leap second fixes from Thomas Gleixner:
 "It's a rather large series, but well discussed, refined and reviewed.
  It got a massive testing by John, Prarit and tip.

  In theory we could split it into two parts.  The first two patches

    f55a6faa: hrtimer: Provide clock_was_set_delayed()
    4873fa07: timekeeping: Fix leapsecond triggered load spike issue

  are merely preventing the stuff loops forever issues, which people
  have observed.

  But there is no point in delaying the other 4 commits which achieve
  full correctness into 3.6 as they are tagged for stable anyway.  And I
  rather prefer to have the full fixes merged in bulk than a "prevent
  the observable wreckage and deal with the hidden fallout later"
  approach."

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  hrtimer: Update hrtimer base offsets each hrtimer_interrupt
  timekeeping: Provide hrtimer update function
  hrtimers: Move lock held region in hrtimer_interrupt()
  timekeeping: Maintain ktime_t based offsets for hrtimers
  timekeeping: Fix leapsecond triggered load spike issue
  hrtimer: Provide clock_was_set_delayed()

If a seccomp filter program is installed, older static binaries and
distributions with older libc implementations (glibc 2.13 and earlier)
that rely on vsyscall use will be terminated regardless of the filter
program policy when executing time, gettimeofday, or getcpu.  This is
only the case when vsyscall emulation is in use (vsyscall=emulate is the
default).

This patch emulates system call entry inside a vsyscall=emulate by
populating regs->ax and regs->orig_ax with the system call number prior
to calling into seccomp such that all seccomp-dependencies function
normally.  Additionally, system call return behavior is emulated in line
with other vsyscall entrypoints for the trace/trap cases.

[ v2: fixed ip and sp on SECCOMP_RET_TRAP/TRACE (thanks to luto@mit.edu) ]
Reported-and-tested-by: Owen Kibel <qmewlo@gmail.com>
Signed-off-by: Will Drewry <wad@chromium.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Please pull one hwmon subsystem fix from Jean Delvare.

* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jdelvare/staging:
  hwmon: (it87) Preserve configuration register bits on init

Pull NFS client bugfixes from Trond Myklebust:
 - Fix an NFSv4 mount regression
 - Fix O_DIRECT list manipulation snafus

* tag 'nfs-for-3.5-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  NFSv4: Fix an NFSv4 mount regression
  NFS: Fix list manipulation snafus in fs/nfs/direct.c

This can be trivially triggered from userspace by passing in something unexpected.

    kernel BUG at fs/locks.c:1468!
    invalid opcode: 0000 [#1] SMP
    RIP: 0010:generic_setlease+0xc2/0x100
    Call Trace:
      __vfs_setlease+0x35/0x40
      fcntl_setlease+0x76/0x150
      sys_fcntl+0x1c6/0x810
      system_call_fastpath+0x1a/0x1f
Signed-off-by: Dave Jones <davej@redhat.com>
Cc: stable@kernel.org # 3.2+
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull input layer fixes from Dmitry Torokhov:
 "The changes are limited to adding new VID/PID combinations to drivers
  to enable support for new versions of hardware, most notably hardware
  found in new MacBook Pro Retina boxes."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: xpad - add Andamiro Pump It Up pad
  Input: xpad - add signature for Razer Onza Tournament Edition
  Input: xpad - handle all variations of Mad Catz Beat Pad
  Input: bcm5974 - Add support for 2012 MacBook Pro Retina
  HID: add support for 2012 MacBook Pro Retina

Pull media fixes from Mauro Carvalho Chehab:
 - Some regression fixes at the audio part for devices with
   cx23885/cx25840
 - A DMA corruption fix at cx231xx
 - two fixes at the winbond IR driver
 - Several fixes for the EXYNOS media driver (s5p)
 - two fixes at the OMAP3 preview driver
 - one fix at the dvb core failure path
 - an include missing (slab.h) at smiapp-core causing compilation
   breakage
 - em28xx was not loading the IR driver driver anymore.

* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (31 commits)
  [media] Revert "[media] V4L: JPEG class documentation corrections"
  [media] s5p-fimc: Add missing FIMC-LITE file operations locking
  [media] omap3isp: preview: Fix contrast and brightness handling
  [media] omap3isp: preview: Fix output size computation depending on input format
  [media] winbond-cir: Initialise timeout, driver_type and allowed_protos
  [media] winbond-cir: Fix txandrx module info
  [media] cx23885: Silence unknown command warnings
  [media] cx23885: add support for HVR-1255 analog (cx23888 variant)
  [media] cx23885: make analog support work for HVR_1250 (cx23885 variant)
  [media] cx25840: fix vsrc/hsrc usage on cx23888 designs
  [media] cx25840: fix regression in HVR-1800 analog audio
  [media] cx25840: fix regression in analog support hue/saturation controls
  [media] cx25840: fix regression in HVR-1800 analog support
  [media] s5p-mfc: Fixed setup of custom controls in decoder and encoder
  [media] cx231xx: don't DMA to random addresses
  [media] em28xx: fix em28xx-rc load
  [media] dvb-core: Release semaphore on error path dvb_register_device()
  [media] s5p-fimc: Stop media entity pipeline if fimc_pipeline_validate fails
  [media] s5p-fimc: Fix compiler warning in fimc-lite.c
  [media] s5p-fimc: media_entity_pipeline_start() may fail
  ...

Pull MMC fixes from Chris Ball:
 - Revert a patch that made failing to select power class fatal;
   it turns out that it fails non-fatally on Tegra boards.
   Regression against 3.5-rc1.
 - Add the IRQF_ONESHOT flag to the cd-gpio driver, which turned
   into a regression in 3.5-rc1 when IRQF_ONESHOT became required
   for threaded IRQs with no handler.

* tag 'mmc-fixes-for-3.5-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/cjb/mmc:
  mmc: cd-gpio: pass IRQF_ONESHOT to request_threaded_irq()
  mmc: core: Revert "skip card initialization if power class selection fails"

Pull late MTD fixes from David Woodhouse:
 - fix 'sparse warning fix' regression which totally breaks MXC NAND
 - fix GPMI NAND regression when used with UBI
 - update/correct sysfs documentation for new 'bitflip_threshold' field
 - fix nandsim build failure

* tag 'for-linus-20120712' of git://git.infradead.org/linux-mtd:
  mtd: nandsim: don't open code a do_div helper
  mtd: ABI documentation: clarification of bitflip_threshold
  mtd: gpmi-nand: fix read page when reading to vmalloced area
  mtd: mxc_nand: use 32bit copy functions

Pull MFD Fixes from Samuel Ortiz:
 - Three Palmas fixes, One of them being a build error fix.
 - Two mc13xx fixes.  One for fixing an SPI regmap configuration and
   another one for working around an i.Mx hardware bug.
 - One omap-usb regression fix.
 - One twl6040 build breakage fix.
 - One file deletion (ab5500-core.h) that was overlooked during the last
   merge window.

* tag 'mfd-for-linus-3.5' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-2.6:
  mfd: Add missing hunk to change palmas irq to clear on read
  mfd: Fix palmas regulator pdata missing
  mfd: USB: Fix the omap-usb EHCI ULPI PHY reset fix issues.
  mfd: Update twl6040 Kconfig to avoid build breakage
  mfd: Delete ab5500-core.h
  mfd: mc13xxx workaround SPI hardware bug on i.Mx
  mfd: Fix mc13xxx SPI regmap
  mfd: Add terminating entry for i2c_device_id palmas table

Pull SuperH fixes from Paul Mundt.

* tag 'sh-for-linus' of git://github.com/pmundt/linux-sh:
  SH: Convert out[bwl] macros to inline functions
  sh: Fix up se7721 GPIOLIB=y build warnings.

Pull a couple of KVM fixes from Avi Kivity:
 "One is an adjustment for an irq layer change that affected device
  assignment, the other a one-liner ppc fix."

* git://git.kernel.org/pub/scm/virt/kvm/kvm:
  powerpc/kvm: Fix "PR" KVM implementation of H_CEDE
  KVM: Fix device assignment threaded irq handler

Commit 080399aa ("block: don't mark buffers beyond end of disk as
mapped") exposed a bug in __getblk_slow that causes mount to hang as it
loops infinitely waiting for a buffer that lies beyond the end of the
disk to become uptodate.

The problem was initially reported by Torsten Hilbrich here:

    https://lkml.org/lkml/2012/6/18/54

and also reported independently here:

    http://www.sysresccd.org/forums/viewtopic.php?f=13&t=4511

and then Richard W.M.  Jones and Marcos Mello noted a few separate
bugzillas also associated with the same issue.  This patch has been
confirmed to fix:

    https://bugzilla.redhat.com/show_bug.cgi?id=835019

The main problem is here, in __getblk_slow:

        for (;;) {
                struct buffer_head * bh;
                int ret;

                bh = __find_get_block(bdev, block, size);
                if (bh)
                        return bh;

                ret = grow_buffers(bdev, block, size);
                if (ret < 0)
                        return NULL;
                if (ret == 0)
                        free_more_memory();
        }

__find_get_block does not find the block, since it will not be marked as
mapped, and so grow_buffers is called to fill in the buffers for the
associated page.  I believe the for (;;) loop is there primarily to
retry in the case of memory pressure keeping grow_buffers from
succeeding.  However, we also continue to loop for other cases, like the
block lying beond the end of the disk.  So, the fix I came up with is to
only loop when grow_buffers fails due to memory allocation issues
(return value of 0).

The attached patch was tested by myself, Torsten, and Rich, and was
found to resolve the problem in call cases.
Signed-off-by: Jeff Moyer <jmoyer@redhat.com>
Reported-and-Tested-by: Torsten Hilbrich <torsten.hilbrich@secunet.com>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Josh Boyer <jwboyer@redhat.com>
Cc: Stable <stable@vger.kernel.org>  # 3.0+
[ Jens is on vacation, taking this directly  - Linus ]
--
Stable Notes: this patch requires backport to 3.0, 3.2 and 3.3.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

I couldn't find the vendor ID in any of the online databases, but this
mat has a Pump It Up logo on the top side of the controller compartment,
and a disclaimer stating that Andamiro will not be liable on the bottom.
Signed-off-by: Yuri Khan <yurivkhan@gmail.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

We were accidentally losing one bit in the configuration register on
device initialization. It was reported to freeze one specific system
right away. Properly preserve all bits we don't explicitly want to
change in order to prevent that.
Reported-by: Stevie Trujillo <stevie.trujillo@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>

The macros just called BUG(), but that results in unused variable
warnings all over the place, like in the IPMI driver.  The build
regression emails were annoying me, so here's the fix.  I have
not even compile tested this, but it's rather obvious.

[ port type mangled to unsigned long ]
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>

Pull fbdev fixes from Florian Tobias Schandinat:
 "Two fixes for OMAPDSS by Tomi Valkeinen:
   - one to avoid warnings when runtime PM is not enabled
   - one workaround to dependancy issues during suspend/resume"

* tag 'fbdev-fixes-for-3.5-2' of git://github.com/schandinat/linux-2.6:
  OMAPDSS: fix warnings if CONFIG_PM_RUNTIME=n
  OMAPDSS: Use PM notifiers for system suspend

Merge random patches from Andrew Morton.

* Merge emailed patches from Andrew Morton <akpm@linux-foundation.org>: (32 commits)
  memblock: free allocated memblock_reserved_regions later
  mm: sparse: fix usemap allocation above node descriptor section
  mm: sparse: fix section usemap placement calculation
  xtensa: fix incorrect memset
  shmem: cleanup shmem_add_to_page_cache
  shmem: fix negative rss in memcg memory.stat
  tmpfs: revert SEEK_DATA and SEEK_HOLE
  drivers/rtc/rtc-twl.c: fix threaded IRQ to use IRQF_ONESHOT
  fat: fix non-atomic NFS i_pos read
  MAINTAINERS: add OMAP CPUfreq driver to OMAP Power Management section
  sgi-xp: nested calls to spin_lock_irqsave()
  fs: ramfs: file-nommu: add SetPageUptodate()
  drivers/rtc/rtc-mxc.c: fix irq enabled interrupts warning
  mm/memory_hotplug.c: release memory resources if hotadd_new_pgdat() fails
  h8300/uaccess: add mising __clear_user()
  h8300/uaccess: remove assignment to __gu_val in unhandled case of get_user()
  h8300/time: add missing #include <asm/irq_regs.h>
  h8300/signal: fix typo "statis"
  h8300/pgtable: add missing #include <asm-generic/pgtable.h>
  drivers/rtc/rtc-ab8500.c: ensure correct probing of the AB8500 RTC when Device Tree is enabled
  ...

memblock_free_reserved_regions() calls memblock_free(), but
memblock_free() would double reserved.regions too, so we could free the
old range for reserved.regions.

Also tj said there is another bug which could be related to this.

| I don't think we're saving any noticeable
| amount by doing this "free - give it to page allocator - reserve
| again" dancing.  We should just allocate regions aligned to page
| boundaries and free them later when memblock is no longer in use.

in that case, when DEBUG_PAGEALLOC, will get panic:

     memblock_free: [0x0000102febc080-0x0000102febf080] memblock_free_reserved_regions+0x37/0x39
  BUG: unable to handle kernel paging request at ffff88102febd948
  IP: [<ffffffff836a5774>] __next_free_mem_range+0x9b/0x155
  PGD 4826063 PUD cf67a067 PMD cf7fa067 PTE 800000102febd160
  Oops: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
  CPU 0
  Pid: 0, comm: swapper Not tainted 3.5.0-rc2-next-20120614-sasha #447
  RIP: 0010:[<ffffffff836a5774>]  [<ffffffff836a5774>] __next_free_mem_range+0x9b/0x155

See the discussion at https://lkml.org/lkml/2012/6/13/469

So try to allocate with PAGE_SIZE alignment and free it later.
Reported-by: Sasha Levin <levinsasha928@gmail.com>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

After commit f5bf18fa ("bootmem/sparsemem: remove limit constraint
in alloc_bootmem_section"), usemap allocations may easily be placed
outside the optimal section that holds the node descriptor, even if
there is space available in that section.  This results in unnecessary
hotplug dependencies that need to have the node unplugged before the
section holding the usemap.

The reason is that the bootmem allocator doesn't guarantee a linear
search starting from the passed allocation goal but may start out at a
much higher address absent an upper limit.

Fix this by trying the allocation with the limit at the section end,
then retry without if that fails.  This keeps the fix from f5bf18fa
of not panicking if the allocation does not fit in the section, but
still makes sure to try to stay within the section at first.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: <stable@vger.kernel.org>	[3.3.x, 3.4.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Commit 238305bb ("mm: remove sparsemem allocation details from the
bootmem allocator") introduced a bug in the allocation goal calculation
that put section usemaps not in the same section as the node
descriptors, creating unnecessary hotplug dependencies between them:

  node 0 must be removed before remove section 16399
  node 1 must be removed before remove section 16399
  node 2 must be removed before remove section 16399
  node 3 must be removed before remove section 16399
  node 4 must be removed before remove section 16399
  node 5 must be removed before remove section 16399
  node 6 must be removed before remove section 16399

The reason is that it applies PAGE_SECTION_MASK to the physical address
of the node descriptor when finding a suitable place to put the usemap,
when this mask is actually intended to be used with PFNs.  Because the
PFN mask is wider, the target address will point beyond the wanted
section holding the node descriptor and the node must be offlined before
the section holding the usemap can go.

Fix this by extending the mask to address width before use.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Addresses: https://bugzilla.kernel.org/show_bug.cgi?id=43871

Reported-by: <dcb314@hotmail.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

shmem_add_to_page_cache() has three callsites, but only one of them wants
the radix_tree_preload() (an exceptional entry guarantees that the radix
tree node is present in the other cases), and only that site can achieve
mem_cgroup_uncharge_cache_page() (PageSwapCache makes it a no-op in the
other cases).  We did it this way originally to reflect
add_to_page_cache_locked(); but it's confusing now, so move the radix_tree
preloading and mem_cgroup uncharging to that one caller.
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

When adding the page_private checks before calling shmem_replace_page(), I
did realize that there is a further race, but thought it too unlikely to
need a hurried fix.

But independently I've been chasing why a mem cgroup's memory.stat
sometimes shows negative rss after all tasks have gone: I expected it to
be a stats gathering bug, but actually it's shmem swapping's fault.

It's an old surprise, that when you lock_page(lookup_swap_cache(swap)),
the page may have been removed from swapcache before getting the lock; or
it may have been freed and reused and be back in swapcache; and it can
even be using the same swap location as before (page_private same).

The swapoff case is already secure against this (swap cannot be reused
until the whole area has been swapped off, and a new swapped on); and
shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for
the expected radix_tree entry - but a little too late.

By that time, we might have already decided to shmem_replace_page(): I
don't know of a problem from that, but I'd feel more at ease not to do so
spuriously.  And we have already done mem_cgroup_cache_charge(), on
perhaps the wrong mem cgroup: and this charge is not then undone on the
error path, because PageSwapCache ends up preventing that.

It's this last case which causes the occasional negative rss in
memory.stat: the page is charged here as cache, but (sometimes) found to
be anon when eventually it's uncharged - and in between, it's an
undeserved charge on the wrong memcg.

Fix this by adding an earlier check on the radix_tree entry: it's
inelegant to descend the tree twice, but swapping is not the fast path,
and a better solution would need a pair (try+commit) of memcg calls, and a
rework of shmem_replace_page() to keep out of the swapcache.

We can use the added shmem_confirm_swap() function to replace the
find_get_page+page_cache_release we were already doing on the error path.
And add a comment on that -EEXIST: it seems a peculiar errno to be using,
but originates from its use in radix_tree_insert().

[It can be surprising to see positive rss left in a memcg's memory.stat
after all tasks have gone, since it is supposed to count anonymous but not
shmem.  Aside from sharing anon pages via fork with a task in some other
memcg, it often happens after swapping: because a swap page can't be freed
while under writeback, nor while locked.  So it's not an error, and these
residual pages are easily freed once pressure demands.]
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Revert 4fb5ef08 ("tmpfs: support SEEK_DATA and SEEK_HOLE").  I believe
it's correct, and it's been nice to have from rc1 to rc6; but as the
original commit said:

I don't know who actually uses SEEK_DATA or SEEK_HOLE, and whether it
would be of any use to them on tmpfs.  This code adds 92 lines and 752
bytes on x86_64 - is that bloat or worthwhile?

Nobody asked for it, so I conclude that it's bloat: let's revert tmpfs to
the dumb generic support for v3.5.  We can always reinstate it later if
useful, and anyone needing it in a hurry can just get it out of git.
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Josef Bacik <josef@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Andreas Dilger <adilger@dilger.ca>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Marco Stornelli <marco.stornelli@gmail.com>
Cc: Jeff liu <jeff.liu@oracle.com>
Cc: Chris Mason <chris.mason@fusionio.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>