openwrt/staging/blogic.git
12 years agoext4: fix undefined behavior in ext4_fill_flex_info()
Xi Wang [Tue, 10 Jan 2012 16:51:10 +0000 (11:51 -0500)]
ext4: fix undefined behavior in ext4_fill_flex_info()

Commit 503358ae01b70ce6909d19dd01287093f6b6271c ("ext4: avoid divide by
zero when trying to mount a corrupted file system") fixes CVE-2009-4307
by performing a sanity check on s_log_groups_per_flex, since it can be
set to a bogus value by an attacker.

sbi->s_log_groups_per_flex = sbi->s_es->s_log_groups_per_flex;
groups_per_flex = 1 << sbi->s_log_groups_per_flex;

if (groups_per_flex < 2) { ... }

This patch fixes two potential issues in the previous commit.

1) The sanity check might only work on architectures like PowerPC.
On x86, 5 bits are used for the shifting amount.  That means, given a
large s_log_groups_per_flex value like 36, groups_per_flex = 1 << 36
is essentially 1 << 4 = 16, rather than 0.  This will bypass the check,
leaving s_log_groups_per_flex and groups_per_flex inconsistent.

2) The sanity check relies on undefined behavior, i.e., oversized shift.
A standard-confirming C compiler could rewrite the check in unexpected
ways.  Consider the following equivalent form, assuming groups_per_flex
is unsigned for simplicity.

groups_per_flex = 1 << sbi->s_log_groups_per_flex;
if (groups_per_flex == 0 || groups_per_flex == 1) {

We compile the code snippet using Clang 3.0 and GCC 4.6.  Clang will
completely optimize away the check groups_per_flex == 0, leaving the
patched code as vulnerable as the original.  GCC keeps the check, but
there is no guarantee that future versions will do the same.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
12 years agoext4: make more symbols static
Eric Sandeen [Thu, 5 Jan 2012 03:33:28 +0000 (22:33 -0500)]
ext4: make more symbols static

A couple more functions can reasonably be made static if desired.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: make local symbol ext4_initxattrs static
Djalal Harouni [Thu, 5 Jan 2012 03:32:12 +0000 (22:32 -0500)]
ext4: make local symbol ext4_initxattrs static

The ext4_initxattrs symbol is used only in this file, so it should be
declared static.

Signed-off-by: Djalal Harouni <tixxdz@opendz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agojbd2: fix hung processes in jbd2_journal_lock_updates()
Jan Kara [Thu, 5 Jan 2012 03:03:11 +0000 (22:03 -0500)]
jbd2: fix hung processes in jbd2_journal_lock_updates()

Toshiyuki Okajima found out that when running

for ((i=0; i < 100000; i++)); do
        if ((i%2 == 0)); then
                chattr +j /mnt/file
        else
                chattr -j /mnt/file
        fi
        echo "0" >> /mnt/file
done

process sometimes hangs indefinitely in jbd2_journal_lock_updates().

Toshiyuki identified that the following race happens:

jbd2_journal_lock_updates()            |jbd2_journal_stop()
---------------------------------------+---------------------------------------
 write_lock(&journal->j_state_lock)    |    .
 ++journal->j_barrier_count            |    .
 spin_lock(&tran->t_handle_lock)       |    .
 atomic_read(&tran->t_updates) //not 0 |
                                       | atomic_dec_and_test(&tran->t_updates)
                                       |    // t_updates = 0
                                       | wake_up(&journal->j_wait_updates)
 prepare_to_wait()                     |    // no process is woken up.
 spin_unlock(&tran->t_handle_lock)     |
 write_unlock(&journal->j_state_lock)  |
 schedule() // never return            |

We fix the problem by first calling prepare_to_wait() and only after that
checking t_updates in jbd2_journal_lock_updates().

Reported-and-analyzed-by: Toshiyuki Okajima <toshi.okajima@jp.fujitsu.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: reserve new feature flag codepoints
Theodore Ts'o [Thu, 5 Jan 2012 03:01:53 +0000 (22:01 -0500)]
ext4: reserve new feature flag codepoints

Reserve the ext4 features flags EXT4_FEATURE_RO_COMPAT_METADATA_CSUM,
EXT4_FEATURE_INCOMPAT_INLINEDATA, and EXT4_FEATURE_INCOMPAT_LARGEDIR.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: Report max_batch_time option correctly
Ben Hutchings [Thu, 5 Jan 2012 02:22:51 +0000 (21:22 -0500)]
ext4: Report max_batch_time option correctly

Currently the value reported for max_batch_time is really the
value of min_batch_time.

Reported-by: Russell Coker <russell@coker.com.au>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
12 years agoext4: add missing ext4_resize_end on error paths
Djalal Harouni [Wed, 4 Jan 2012 22:09:52 +0000 (17:09 -0500)]
ext4: add missing ext4_resize_end on error paths

Online resize ioctls 'EXT4_IOC_GROUP_EXTEND' and 'EXT4_IOC_GROUP_ADD'
call ext4_resize_begin() to check permissions and to set the
EXT4_RESIZING bit lock, they do their work and they must finish with
ext4_resize_end() which calls clear_bit_unlock() to unlock and to
avoid -EBUSY errors for the next resize operations.

This patch adds the missing ext4_resize_end() calls on error paths.

Patch tested.

Cc: stable@vger.kernel.org
Signed-off-by: Djalal Harouni <tixxdz@opendz.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: let ext4_group_add() use common code
Yongqiang Yang [Wed, 4 Jan 2012 22:09:50 +0000 (17:09 -0500)]
ext4: let ext4_group_add() use common code

This patch lets ext4_group_add() call ext4_flex_group_add().

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: let ext4_group_extend() use common code
Yongqiang Yang [Wed, 4 Jan 2012 22:09:48 +0000 (17:09 -0500)]
ext4: let ext4_group_extend() use common code

ext4_group_extend_no_check() is moved out from ext4_group_extend(),
this patch lets ext4_group_extend() call ext4_group_extentd_no_check()
instead.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add new online resize interface
Yongqiang Yang [Wed, 4 Jan 2012 22:09:44 +0000 (17:09 -0500)]
ext4: add new online resize interface

This patch adds new online resize interface, whose input argument is a
64-bit integer indicating how many blocks there are in the resized fs.

In new resize impelmentation, all work like allocating group tables
are done by kernel side, so the new resize interface can support
flex_bg feature and prepares ground for suppoting resize with features
like bigalloc and exclude bitmap. Besides these, user-space tools just
passes in the new number of blocks.

We delay initializing the bitmaps and inode tables of added groups if
possible and add multi groups (a flex groups) each time, so new resize
is very fast like mkfs.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a new function which adds a flex group to a fs
Yongqiang Yang [Wed, 4 Jan 2012 04:44:38 +0000 (23:44 -0500)]
ext4: add a new function which adds a flex group to a fs

This patch adds a new function named ext4_flex_group_add() which adds a
flex group to a fs.  The function is used by 64bit-resize interface.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a new function which allocates bitmaps and inode tables
Yongqiang Yang [Wed, 4 Jan 2012 04:44:38 +0000 (23:44 -0500)]
ext4: add a new function which allocates bitmaps and inode tables

This patch adds a new function named ext4_allocates_group_table()
which allocates block bitmaps, inode bitmaps and inode tables for a
flex groups and is used by resize code.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: pass verify_reserved_gdb() the number of group decriptors
Yongqiang Yang [Wed, 4 Jan 2012 04:43:39 +0000 (23:43 -0500)]
ext4: pass verify_reserved_gdb() the number of group decriptors

The 64bit resizer adds a flex group each time, so verify_reserved_gdb
can not use s_groups_count directly, it should use the number of group
decriptors before the added group.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a function which updates the super block during online resizing
Yongqiang Yang [Wed, 4 Jan 2012 04:41:39 +0000 (23:41 -0500)]
ext4: add a function which updates the super block during online resizing

This patch adds a function named ext4_update_super() which updates
super block so the newly created block groups are visible to the file
system.  This code is copied from ext4_group_add().

The function will be used by new resize implementation.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a function which sets up a block group descriptors of a flex bg
Yongqiang Yang [Wed, 4 Jan 2012 04:37:31 +0000 (23:37 -0500)]
ext4: add a function which sets up a block group descriptors of a flex bg

This patch adds a function named ext4_setup_new_descs which sets up the
block group descriptors of a flex bg.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a function which sets up group blocks of a flex bg
Yongqiang Yang [Wed, 4 Jan 2012 04:32:52 +0000 (23:32 -0500)]
ext4: add a function which sets up group blocks of a flex bg

This patch adds a function named setup_new_flex_group_blocks() which
sets up group blocks of a flex bg.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a structure which will be used by 64bit-resize interface
Yongqiang Yang [Wed, 4 Jan 2012 04:22:50 +0000 (23:22 -0500)]
ext4: add a structure which will be used by 64bit-resize interface

This patch adds a structure which will be used by 64bit-resize interface.
Two functions which allocate and destroy the structure respectively are
added.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a function which adds a new group descriptors to a fs
Yongqiang Yang [Wed, 4 Jan 2012 04:20:50 +0000 (23:20 -0500)]
ext4: add a function which adds a new group descriptors to a fs

This patch adds a function named ext4_add_new_descs() which adds one
or more new group descriptors to a fs and whose code is copied from
ext4_group_add().

The function will be used by new resize implementation.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add a function which extends a group without checking parameters
Yongqiang Yang [Wed, 4 Jan 2012 04:18:50 +0000 (23:18 -0500)]
ext4: add a function which extends a group without checking parameters

This patch added a function named ext4_group_extend_no_check() whose code
is copied from ext4_group_extend().  ext4_group_extend_no_check() assumes
the parameter is valid and has been checked by caller.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: use proper little-endian bitops
Akinobu Mita [Thu, 29 Dec 2011 01:32:07 +0000 (20:32 -0500)]
ext4: use proper little-endian bitops

ext4_{set,clear}_bit() is defined as __test_and_{set,clear}_bit_le() for
ext4.  Only two ext4_{set,clear}_bit() calls check the return value.  The
rest of calls ignore the return value and they can be replaced with
__{set,clear}_bit_le().

This changes ext4_{set,clear}_bit() from __test_and_{set,clear}_bit_le()
to __{set,clear}_bit_le() and introduces ext4_test_and_{set,clear}_bit()
for the two places where old bit needs to be returned.

This ext4_{set,clear}_bit() change is considered safe, because if someone
uses these macros without noticing the change, new ext4_{set,clear}_bit
don't have return value and causes compiler errors where the return value
is used.

This also removes unused ext4_find_first_zero_bit().

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: remove no longer used functions in inode.c
Zheng Liu [Thu, 29 Dec 2011 01:25:40 +0000 (20:25 -0500)]
ext4: remove no longer used functions in inode.c

The functions ext4_block_truncate_page() and ext4_block_zero_page_range()
are no longer used, so remove them.

Signed-off-by: Zheng Liu <wenqing.lz@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: avoid counting the number of free inodes twice in find_group_orlov()
Theodore Ts'o [Thu, 29 Dec 2011 01:25:13 +0000 (20:25 -0500)]
ext4: avoid counting the number of free inodes twice in find_group_orlov()

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add missing spaces to debugging printk's
Zheng Liu [Thu, 29 Dec 2011 00:00:25 +0000 (19:00 -0500)]
ext4: add missing spaces to debugging printk's

Fix ext4_debug format in ext4_ext_handle_uninitialized_extents() and
ext4_end_io_dio().

Signed-off-by: Zheng Liu <wenqing.lz@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agojbd2: clear revoked flag on buffers before a new transaction started
Yongqiang Yang [Wed, 28 Dec 2011 22:46:46 +0000 (17:46 -0500)]
jbd2: clear revoked flag on buffers before a new transaction started

Currently, we clear revoked flag only when a block is reused.  However,
this can tigger a false journal error.  Consider a situation when a block
is used as a meta block and is deleted(revoked) in ordered mode, then the
block is allocated as a data block to a file.  At this moment, user changes
the file's journal mode from ordered to journaled and truncates the file.
The block will be considered re-revoked by journal because it has revoked
flag still pending from the last transaction and an assertion triggers.

We fix the problem by keeping the revoked status more uptodate - we clear
revoked flag when switching revoke tables to reflect there is no revoked
buffers in current transaction any more.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: flush journal when switching from data=journal mode
Yongqiang Yang [Wed, 28 Dec 2011 18:55:51 +0000 (13:55 -0500)]
ext4: flush journal when switching from data=journal mode

It's necessary to flush the journal when switching away from
data=journal mode.  This is because there are no revoke records when
data blocks are journalled, but revoke records are required in the
other journal modes.

However, it is not necessary to flush the journal when switching into
data=journal mode, and flushing the journal is expensive.  So let's
avoid it in that case.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: allocate delalloc blocks before changing journal mode
Yongqiang Yang [Wed, 28 Dec 2011 17:02:13 +0000 (12:02 -0500)]
ext4: allocate delalloc blocks before changing journal mode

delalloc blocks should be allocated before changing journal mode,
otherwise they can not be allocated and even more truncate on
delalloc blocks could triggre BUG by flushing delalloc buffers.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: remove unneeded file_remove_suid() from ext4_ioctl()
Theodore Ts'o [Wed, 21 Dec 2011 19:14:31 +0000 (14:14 -0500)]
ext4: remove unneeded file_remove_suid() from ext4_ioctl()

In the code to support EXT4_IOC_MOVE_EXT, ext4_ioctl calls
file_remove_suid() after the call to ext4_move_extents() if any
extents has been moved.  There are at least three things wrong with
this.  First, file_remove_suid() should be called with i_mutex down,
which is not here.  Second, it should be called before the donor file
has been modified, to avoid a potential race condition.  Third, and
most importantly, it's pointless, because ext4_file_extents() already
checks if the donor file has the setuid or setgid bit set, and will
return an error in that case.  So the first two objections don't
really matter, since file_remove_suid() will never need to modify the
inode in any case.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: optimize ext4_find_delalloc_range() in nodelalloc mode
Robin Dong [Mon, 19 Dec 2011 04:05:43 +0000 (23:05 -0500)]
ext4: optimize ext4_find_delalloc_range() in nodelalloc mode

We found performance regression when using bigalloc with "nodelalloc"
(1MB cluster size):

1. mke2fs -C 1048576 -O ^has_journal,bigalloc /dev/sda
2. mount -o nodelalloc /dev/sda /test/
3. time dd if=/dev/zero of=/test/io bs=1048576 count=1024

The "dd" will cost about 2 seconds to finish, but if we mke2fs without
"bigalloc", "dd" will only cost less than 1 second.

The reason is: when using ext4 with "nodelalloc", it will call
ext4_find_delalloc_cluster() nearly everytime it call
ext4_ext_map_blocks(), and ext4_find_delalloc_range() will also scan
all pages in cluster because no buffer is "delayed".  A cluster has
256 pages (1MB cluster), so it will scan 256 * 256k pags when creating
a 1G file. That severely hurts the performance.

Therefore, we return immediately from ext4_find_delalloc_range() in
nodelalloc mode, since by definition there can't be any delalloc
pages.

Signed-off-by: Robin Dong <sanbai@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: remove unused local variable
Curt Wohlgemuth [Sun, 18 Dec 2011 22:39:02 +0000 (17:39 -0500)]
ext4: remove unused local variable

In get_implied_cluster_alloc(), rr_cluster_end was being
defined and set, but was never used.  Removed this.

Signed-off-by: Curt Wohlgemuth <curtw@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: fix error handling on inode bitmap corruption
Jan Kara [Sun, 18 Dec 2011 22:37:02 +0000 (17:37 -0500)]
ext4: fix error handling on inode bitmap corruption

When insert_inode_locked() fails in ext4_new_inode() it most likely means inode
bitmap got corrupted and we allocated again inode which is already in use. Also
doing unlock_new_inode() during error recovery is wrong since the inode does
not have I_NEW set. Fix the problem by jumping to fail: (instead of fail_drop:)
which declares filesystem error and does not call unlock_new_inode().

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: add missing space to ext4_msg output in ext4_fill_super()
Zheng Liu [Sun, 18 Dec 2011 21:13:58 +0000 (16:13 -0500)]
ext4: add missing space to ext4_msg output in ext4_fill_super()

Signed-off-by: Zheng Liu <wenqing.lz@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoext4: do not reference pa_inode from group_pa
Yongqiang Yang [Sun, 18 Dec 2011 20:49:54 +0000 (15:49 -0500)]
ext4: do not reference pa_inode from group_pa

pa_inode in group_pa is set NULL in ext4_mb_new_group_pa, so
pa_inode should be not referenced.

Reported-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
12 years agoLinux 3.2-rc6
Linus Torvalds [Sat, 17 Dec 2011 02:36:26 +0000 (18:36 -0800)]
Linux 3.2-rc6

12 years agoRevert "drm/i915: fix infinite recursion on unbind due to ilk vt-d w/a"
Linus Torvalds [Fri, 16 Dec 2011 20:58:39 +0000 (12:58 -0800)]
Revert "drm/i915: fix infinite recursion on unbind due to ilk vt-d w/a"

This reverts commit eb1711bb94991e93669c5a1b5f84f11be2d51ea1.

It blows up the i915 seqno tracking, resulting in the

BUG_ON(seqno == 0);

in i915_wait_request() triggering, which will cause lock-ups.

See for example
  https://bugs.launchpad.net/ubuntu/+source/linux/+bug/903010
  https://lkml.org/lkml/2011/12/14/395

Reported-requested-and-tested-by: Dirk Hohndel <dirk@hohndel.org>
Reported-by: Richard Eames <Richard.Eames@flinders.edu.au>
Reported-by: Rocko Requin <rockorequin@hotmail.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Airlie <airlied@redhat.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Keith Packard <keithp@keithp.com>
Cc: Eric Anholt <eric@anholt.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc
Linus Torvalds [Fri, 16 Dec 2011 20:26:26 +0000 (12:26 -0800)]
Merge git://git./linux/kernel/git/davem/sparc

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc32: Be less strict in matching %lo part of relocation.
  sbus: convert drivers/sbus/char/* to use module_platform_driver()
  bbc_i2c: Remove unneeded err variable
  sparc: Use kmemdup rather than duplicating its implementation

12 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Fri, 16 Dec 2011 20:17:32 +0000 (12:17 -0800)]
Merge git://git./linux/kernel/git/davem/net

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net:
  ipv6: Check dest prefix length on original route not copied one in rt6_alloc_cow().
  sch_gred: should not use GFP_KERNEL while holding a spinlock
  ipip, sit: copy parms.name after register_netdevice
  ipv6: Fix for adding multicast route for loopback device automatically.
  ssb: fix init regression with SoCs
  rtl8192{ce,cu,de,se}: avoid problems because of possible ERFOFF -> ERFSLEEP transition
  mac80211: fix another race in aggregation start
  fsl_pq_mdio: Clean up tbi address configuration
  ppp: fix pptp double release_sock in pptp_bind()
  net/fec: fix the use of pdev->id
  ath9k: fix check for antenna diversity support
  batman-adv: delete global entry in case of roaming
  batman-adv: in case of roaming mark the client with TT_CLIENT_ROAM
  Bluetooth: Correct version check in hci_setup
  btusb: fix a memory leak in btusb_send_frame()
  Bluetooth: bnep: Fix module reference
  Bluetooth: cmtp: Fix module reference
  Bluetooth: btmrvl: support Marvell Bluetooth device SD8797

12 years agoMerge branches 'for-linus' and 'for-linus-3.2' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Fri, 16 Dec 2011 20:15:50 +0000 (12:15 -0800)]
Merge branches 'for-linus' and 'for-linus-3.2' of git://git./linux/kernel/git/mason/linux-btrfs

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: unplug every once and a while
  Btrfs: deal with NULL srv_rsv in the delalloc inode reservation code
  Btrfs: only set cache_generation if we setup the block group
  Btrfs: don't panic if orphan item already exists
  Btrfs: fix leaked space in truncate
  Btrfs: fix how we do delalloc reservations and how we free reservations on error
  Btrfs: deal with enospc from dirtying inodes properly
  Btrfs: fix num_workers_starting bug and other bugs in async thread
  BTRFS: Establish i_ops before calling d_instantiate
  Btrfs: add a cond_resched() into the worker loop
  Btrfs: fix ctime update of on-disk inode
  btrfs: keep orphans for subvolume deletion
  Btrfs: fix inaccurate available space on raid0 profile
  Btrfs: fix wrong disk space information of the files
  Btrfs: fix wrong i_size when truncating a file to a larger size
  Btrfs: fix btrfs_end_bio to deal with write errors to a single mirror

* 'for-linus-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  btrfs: lower the dirty balance poll interval

12 years agoMerge branch 'drm-intel-fixes' of git://people.freedesktop.org/~keithp/linux
Linus Torvalds [Fri, 16 Dec 2011 19:27:56 +0000 (11:27 -0800)]
Merge branch 'drm-intel-fixes' of git://people.freedesktop.org/~keithp/linux

* 'drm-intel-fixes' of git://people.freedesktop.org/~keithp/linux:
  drm/i915/dp: Dither down to 6bpc if it makes the mode fit
  drm/i915: enable semaphores on per-device defaults
  drm/i915: don't set unpin_work if vblank_get fails
  drm/i915: By default, enable RC6 on IVB and SNB when reasonable
  iommu: Export intel_iommu_enabled to signal when iommu is in use
  drm/i915/sdvo: Include LVDS panels for the IS_DIGITAL check
  drm/i915: prevent division by zero when asking for chipset power
  drm/i915: add PCH info to i915_capabilities
  drm/i915: set the right SDVO transcoder for CPT
  drm/i915: no-lvds quirk for ASUS AT5NM10T-I
  drm/i915: Treat pre-gen4 backlight duty cycle value consistently
  drm/i915: Hook up Ivybridge eDP
  drm/i915: add multi-threaded forcewake support

12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Linus Torvalds [Fri, 16 Dec 2011 19:14:42 +0000 (11:14 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: HDA: Use LPIB Position fix for Intel SCH Poulsbo
  ALSA: hda: fix mute led polarity for HP laptops with buggy BIOS
  ALSA: HDA: Set position fix to LPIB for an Atom/Poulsbo based device
  ASoC: Fix hx4700 error handling to free gpios if snd_soc_register_card fails
  ASoC: WM8958: correctly show firmware magic on mismatch
  ASoC: mxs: Add appropriate MODULE_ALIAS()
  ASoC: mxs: Add missing MODULE_LICENSE("GPL")
  ASoC: Fix WM8996 24.576MHz clock operation
  ASoC: Include linux/module.h for smdk2443_wm9710
  ASoC: Fix a typo in jive_wm8750
  ASoC: Fix build dependency for SND_SOC_JZ4740_CODEC
  ASoC: Include linux/io.h for jz4740 codec

12 years agoMerge branch 'gpio/merge' of git://git.secretlab.ca/git/linux-2.6
Linus Torvalds [Fri, 16 Dec 2011 19:13:49 +0000 (11:13 -0800)]
Merge branch 'gpio/merge' of git://git.secretlab.ca/git/linux-2.6

* 'gpio/merge' of git://git.secretlab.ca/git/linux-2.6:
  gpio: Fix DA9052 GPIO build errors.
  gpio: mpc8xxx: don't allow input-only pins to be output for MPC5121
  gpio-ml-ioh: Add the irq_disable/irq_enable hooks for ml-ioh irq chip
  gpio-ml-ioh: fix a bug in the interrupt handler
  gpio: pl061: drop extra check for NULL platform_data

12 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc
Linus Torvalds [Fri, 16 Dec 2011 18:07:50 +0000 (10:07 -0800)]
Merge branch 'fixes' of git://git./linux/kernel/git/arm/arm-soc

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
  ARM: OMAP: rx51: fix USB
  ARM: OMAP: mcbsp: Fix possible memory corruption
  arm/imx: fix power button on imx51 babbage board
  ARM: imx: fix cpufreq build errors
  ARM: mx5: add __initconst for fec pdata
  MXC PWM: should active during DOZE/WAIT/DBG mode
  ARM: EXYNOS: Fix build error without CONFIG_LOCAL_TIMERS
  ARM: EXYNOS: Fix for stall in case of cpu hotplug or sleep
  ARM: S5PV210: Set 1000ns as PWM backlight period on SMDKV210
  ARM: SAMSUNG: remove duplicated header include

12 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-block
Linus Torvalds [Fri, 16 Dec 2011 18:05:14 +0000 (10:05 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block

* 'for-linus' of git://git.kernel.dk/linux-block:
  block: don't kick empty queue in blk_drain_queue()
  block/swim3: Locking fixes
  loop: Fix discard_alignment default setting
  cfq-iosched: fix cfq_cic_link() race confition
  cfq-iosched: free cic_index if blkio_alloc_blkg_stats fails
  cciss: fix flush cache transfer length
  cciss: Add IRQF_SHARED back in for the non-MSI(X) interrupt handler
  loop: fix loop block driver discard and encryption comment
  block: initialize request_queue's numa node during

12 years agobtrfs: lower the dirty balance poll interval
Wu Fengguang [Fri, 16 Dec 2011 17:32:57 +0000 (12:32 -0500)]
btrfs: lower the dirty balance poll interval

Tests show that the original large intervals can easily make the dirty
limit exceeded on 100 concurrent dd's. So adapt to as large as the
next check point selected by the dirty throttling algorithm.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agodrm/i915/dp: Dither down to 6bpc if it makes the mode fit
Adam Jackson [Tue, 13 Dec 2011 23:41:00 +0000 (15:41 -0800)]
drm/i915/dp: Dither down to 6bpc if it makes the mode fit

Some active adaptors (VGA usually) only have two lanes at 2.7GHz.
That's a maximum pixel clock of 144MHz at 8bpc, but 192MHz at 6bpc.

Fixes Asus UX31 panel being black at startup due to no valid modes since
dc22ee6fc18ce0f15424e753e8473c306ece95c1.

v2: Rebased to current code, resulting in the fix applying to EDP panels as
    well.  Also changed from spatio-temporal to just spatial dithering on
    pre-ironlake, to be conssitent (and less visual flicker)

Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
Tested-by: Eric Anholt <eric@anholt.net>
Tested-by: Dirk Hohndel <hohndel@infradead.org>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915: enable semaphores on per-device defaults
Eugeni Dodonov [Sat, 10 Dec 2011 01:16:37 +0000 (17:16 -0800)]
drm/i915: enable semaphores on per-device defaults

This adds a default setting for semaphores parameter, and enables
semaphores by default on IVB.

For now, as semaphores interaction with VTd causes random issues on
SNB, we do not enable them by default. But they can still be enabled
via the semaphores=1 kernel parameter.

v2: enables semaphores on SNB when IO remapping is disabled, with base
on Keith Packard patch.

CC: Daniel Vetter <daniel.vetter@ffwll.ch>
CC: Ben Widawsky <ben@bwidawsk.net>
CC: Keith Packard <keithp@keithp.com>
CC: Jesse Barnes <jbarnes@virtuousgeek.org>
CC: Chris Wilson <chris@chris-wilson.co.uk>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42696
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=40564
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41353
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38862
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915: don't set unpin_work if vblank_get fails
Jesse Barnes [Mon, 29 Aug 2011 16:45:28 +0000 (09:45 -0700)]
drm/i915: don't set unpin_work if vblank_get fails

This fixes a race where we may try to finish a page flip and decrement
the refcount even if our vblank_get failed and we ended up with a
spurious flip pending interrupt.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=34211.

Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915: By default, enable RC6 on IVB and SNB when reasonable
Keith Packard [Thu, 17 Nov 2011 06:24:52 +0000 (22:24 -0800)]
drm/i915: By default, enable RC6 on IVB and SNB when reasonable

RC6 should always work on IVB, and should work on SNB whenever IO
remapping is disabled. RC6 never works on Ironlake. Make the default
value for the parameter follow these guidelines. Setting the value
to either 0 or 1 will force the specified behavior.

Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38567
Cc: Ted Phelps <phelps@gnusto.com>
Cc: Peter <pab1612@gmail.com>
Cc: Lukas Hejtmanek <xhejtman@fi.muni.cz>
Cc: Andrew Lutomirski <luto@mit.edu>
12 years agoiommu: Export intel_iommu_enabled to signal when iommu is in use
Eugeni Dodonov [Wed, 23 Nov 2011 18:42:14 +0000 (16:42 -0200)]
iommu: Export intel_iommu_enabled to signal when iommu is in use

In i915 driver, we do not enable either rc6 or semaphores on SNB when dmar
is enabled. The new 'intel_iommu_enabled' variable signals when the
iommu code is in operation.

Cc: Ted Phelps <phelps@gnusto.com>
Cc: Peter <pab1612@gmail.com>
Cc: Lukas Hejtmanek <xhejtman@fi.muni.cz>
Cc: Andrew Lutomirski <luto@mit.edu>
CC: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Eugeni Dodonov <eugeni.dodonov@intel.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915/sdvo: Include LVDS panels for the IS_DIGITAL check
Chris Wilson [Mon, 20 Jun 2011 13:45:50 +0000 (14:45 +0100)]
drm/i915/sdvo: Include LVDS panels for the IS_DIGITAL check

We were checking whether the supplied edid matched the connector it was
read from. We do this in case a DDC read returns an EDID for another
device on a multifunction or otherwise interesting card. However, we
failed to include LVDS as a digital device and so rejecting an otherwise
valid EDID.

Fixes the detection of the secondary SDVO LVDS panel on the Libretto
W105.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39216
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915: prevent division by zero when asking for chipset power
Eugeni Dodonov [Thu, 10 Nov 2011 15:55:15 +0000 (13:55 -0200)]
drm/i915: prevent division by zero when asking for chipset power

This prevents an in-kernel division by zero which happens when we are
asking for i915_chipset_val too quickly, or within a race condition
between the power monitoring thread and userspace accesses via debugfs.

The issue can be reproduced easily via the following command:
while ``; do cat /sys/kernel/debug/dri/0/i915_emon_status; done

This is particularly dangerous because it can be triggered by
a non-privileged user by just reading the debugfs entry.

This issue was also found independently by Konstantin Belousov
<kostikbel@gmail.com>, who proposed a similar patch.

Reported-by: Konstantin Belousov <kostikbel@gmail.com>
Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Acked-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org>
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915: add PCH info to i915_capabilities
Paulo Zanoni [Fri, 14 Oct 2011 21:17:41 +0000 (18:17 -0300)]
drm/i915: add PCH info to i915_capabilities

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915: set the right SDVO transcoder for CPT
Paulo Zanoni [Fri, 14 Oct 2011 21:16:22 +0000 (18:16 -0300)]
drm/i915: set the right SDVO transcoder for CPT

v2: add a CPT-specific macro, make code cleaner
v3: fix commit message

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41272
Cc: stable@kernel.org
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agodrm/i915: no-lvds quirk for ASUS AT5NM10T-I
Adam Jackson [Mon, 28 Nov 2011 17:22:56 +0000 (12:22 -0500)]
drm/i915: no-lvds quirk for ASUS AT5NM10T-I

https://bugzilla.redhat.com/show_bug.cgi?id=750006

Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Keith Packard <keithp@keithp.com>
12 years agoMerge branch 'samsung-fixes-3' of git://git.kernel.org/pub/scm/linux/kernel/git/kgene...
Olof Johansson [Fri, 16 Dec 2011 07:34:17 +0000 (23:34 -0800)]
Merge branch 'samsung-fixes-3' of git://git./linux/kernel/git/kgene/linux-samsung into fixes

12 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux...
Olof Johansson [Fri, 16 Dec 2011 04:58:37 +0000 (20:58 -0800)]
Merge branch 'fixes' of git://git./linux/kernel/git/tmlind/linux-omap into fixes

12 years agoMerge branch 'linux-next' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad...
Linus Torvalds [Thu, 15 Dec 2011 22:16:47 +0000 (14:16 -0800)]
Merge branch 'linux-next' of git://git./linux/kernel/git/konrad/ibft

* 'linux-next' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/ibft:
  ibft: Fix finding IBFT ACPI table on UEFI

12 years agoMerge branch 'staging-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Thu, 15 Dec 2011 22:15:01 +0000 (14:15 -0800)]
Merge branch 'staging-linus' of git://git./linux/kernel/git/gregkh/staging

* 'staging-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging:
  staging: r8712u: Add new USB ID
  staging: tidspbridge: request dmtimer clocks on init
  staging: tidspbridge: include module.h by default

12 years agoMerge branch 'usb-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Linus Torvalds [Thu, 15 Dec 2011 22:14:41 +0000 (14:14 -0800)]
Merge branch 'usb-linus' of git://git./linux/kernel/git/gregkh/usb

* 'usb-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
  USB: option: Removing one bogus and adding some new Huawei combinations
  USB: cdc-acm: add IDs for Motorola H24 HSPA USB module.
  usb: option: Add Huawei E398 controlling interfaces
  usb: renesas_usbhs: add hcd->has_tt for low/full speed
  usb: renesas_usbhs: typofix: irq_dtch control DTCHE
  usb: gadget: storage: release superspeed descriptors.

12 years agoARM: unwinder: fix bisection to find origin in .idx section
Uwe Kleine-König [Thu, 15 Dec 2011 20:47:56 +0000 (21:47 +0100)]
ARM: unwinder: fix bisection to find origin in .idx section

The bisection implemented in unwind_find_origin() stopped to early.  If
there is only a single entry left to check the original code just took
the end point as origin which might be wrong.

This was introduced in commit de66a979012d ("ARM: 7187/1: fix unwinding
for XIP kernels").

Reported-and-tested-by: Nick Bowler <nbowler@elliptictech.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
12 years agoBtrfs: unplug every once and a while
Chris Mason [Thu, 15 Dec 2011 20:38:41 +0000 (15:38 -0500)]
Btrfs: unplug every once and a while

The btrfs io submission threads can build up massive plug lists.  This
keeps things more reasonable so we don't hand over huge dumps of IO at
once.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoblock: don't kick empty queue in blk_drain_queue()
Tejun Heo [Thu, 15 Dec 2011 19:03:04 +0000 (20:03 +0100)]
block: don't kick empty queue in blk_drain_queue()

While probing, fd sets up queue, probes hardware and tears down the
queue if probing fails.  In the process, blk_drain_queue() kicks the
queue which failed to finish initialization and fd is unhappy about
that.

  floppy0: no floppy controllers found
  ------------[ cut here ]------------
  WARNING: at drivers/block/floppy.c:2929 do_fd_request+0xbf/0xd0()
  Hardware name: To Be Filled By O.E.M.
  VFS: do_fd_request called on non-open device
  Modules linked in:
  Pid: 1, comm: swapper Not tainted 3.2.0-rc4-00077-g5983fe2 #2
  Call Trace:
   [<ffffffff81039a6a>] warn_slowpath_common+0x7a/0xb0
   [<ffffffff81039b41>] warn_slowpath_fmt+0x41/0x50
   [<ffffffff813d657f>] do_fd_request+0xbf/0xd0
   [<ffffffff81322b95>] blk_drain_queue+0x65/0x80
   [<ffffffff81322c93>] blk_cleanup_queue+0xe3/0x1a0
   [<ffffffff818a809d>] floppy_init+0xdeb/0xe28
   [<ffffffff818a72b2>] ? daring+0x6b/0x6b
   [<ffffffff810002af>] do_one_initcall+0x3f/0x170
   [<ffffffff81884b34>] kernel_init+0x9d/0x11e
   [<ffffffff810317c2>] ? schedule_tail+0x22/0xa0
   [<ffffffff815dbb14>] kernel_thread_helper+0x4/0x10
   [<ffffffff81884a97>] ? start_kernel+0x2be/0x2be
   [<ffffffff815dbb10>] ? gs_change+0xb/0xb

Avoid it by making blk_drain_queue() kick queue iff dispatch queue has
something on it.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Ralf Hildebrandt <Ralf.Hildebrandt@charite.de>
Reported-by: Wu Fengguang <fengguang.wu@intel.com>
Tested-by: Sergei Trofimovich <slyich@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
12 years agoMerge branch 'stable/for-linus-fixes-3.2' of git://git.kernel.org/pub/scm/linux/kerne...
Linus Torvalds [Thu, 15 Dec 2011 18:52:40 +0000 (10:52 -0800)]
Merge branch 'stable/for-linus-fixes-3.2' of git://git./linux/kernel/git/konrad/xen

* 'stable/for-linus-fixes-3.2' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xen/swiotlb: Use page alignment for early buffer allocation.
  xen: only limit memory map to maximum reservation for domain 0.

12 years agoMerge branch 'for-chris' of http://git.kernel.org/pub/scm/linux/kernel/git/josef...
Chris Mason [Thu, 15 Dec 2011 18:43:49 +0000 (13:43 -0500)]
Merge branch 'for-chris' of git./linux/kernel/git/josef/btrfs-work into integration

Conflicts:
fs/btrfs/inode.c

Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoBtrfs: deal with NULL srv_rsv in the delalloc inode reservation code
Chris Mason [Thu, 15 Dec 2011 18:36:29 +0000 (13:36 -0500)]
Btrfs: deal with NULL srv_rsv in the delalloc inode reservation code

btrfs_update_inode is sometimes called with a null reservation.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agogpio: Fix DA9052 GPIO build errors.
Ashish Jangam [Thu, 15 Dec 2011 09:25:46 +0000 (14:55 +0530)]
gpio: Fix DA9052 GPIO build errors.

This patch is functionally tested on Samsung SMDKV6410.

Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: David Dajun Chen <dchen@diasemi.com>
Signed-off-by: Ashish Jangam <ashish.jangam@kpitcummins.com>
[grant.likely: don't create an unnecessary header file]
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
12 years agoxen/swiotlb: Use page alignment for early buffer allocation.
Konrad Rzeszutek Wilk [Thu, 15 Dec 2011 16:28:46 +0000 (11:28 -0500)]
xen/swiotlb: Use page alignment for early buffer allocation.

This fixes an odd bug found on a Dell PowerEdge 1850/0RC130
(BIOS A05 01/09/2006) where all of the modules doing pci_set_dma_mask
would fail with:

ata_piix 0000:00:1f.1: enabling device (0005 -> 0007)
ata_piix 0000:00:1f.1: can't derive routing for PCI INT A
ata_piix 0000:00:1f.1: BMDMA: failed to set dma mask, falling back to PIO

The issue was the Xen-SWIOTLB was allocated such as that the end of
buffer was stradling a page (and also above 4GB). The fix was
spotted by Kalev Leonid  which was to piggyback on git commit
e79f86b2ef9c0a8c47225217c1018b7d3d90101c "swiotlb: Use page alignment
for early buffer allocation" which:

We could call free_bootmem_late() if swiotlb is not used, and
it will shrink to page alignment.

So alloc them with page alignment at first, to avoid lose two pages

And doing that fixes the outstanding issue.

CC: stable@kernel.org
Suggested-by: "Kalev, Leonid" <Leonid.Kalev@ca.com>
Reported-and-Tested-by: "Taylor, Neal E" <Neal.Taylor@ca.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
12 years agoxen: only limit memory map to maximum reservation for domain 0.
Ian Campbell [Wed, 14 Dec 2011 12:16:08 +0000 (12:16 +0000)]
xen: only limit memory map to maximum reservation for domain 0.

d312ae878b6a "xen: use maximum reservation to limit amount of usable RAM"
clamped the total amount of RAM to the current maximum reservation. This is
correct for dom0 but is not correct for guest domains. In order to boot a guest
"pre-ballooned" (e.g. with memory=1G but maxmem=2G) in order to allow for
future memory expansion the guest must derive max_pfn from the e820 provided by
the toolstack and not the current maximum reservation (which can reflect only
the current maximum, not the guest lifetime max). The existing algorithm
already behaves this correctly if we do not artificially limit the maximum
number of pages for the guest case.

For a guest booted with maxmem=512, memory=128 this results in:
 [    0.000000] BIOS-provided physical RAM map:
 [    0.000000]  Xen: 0000000000000000 - 00000000000a0000 (usable)
 [    0.000000]  Xen: 00000000000a0000 - 0000000000100000 (reserved)
-[    0.000000]  Xen: 0000000000100000 - 0000000008100000 (usable)
-[    0.000000]  Xen: 0000000008100000 - 0000000020800000 (unusable)
+[    0.000000]  Xen: 0000000000100000 - 0000000020800000 (usable)
...
 [    0.000000] NX (Execute Disable) protection: active
 [    0.000000] DMI not present or invalid.
 [    0.000000] e820 update range: 0000000000000000 - 0000000000010000 (usable) ==> (reserved)
 [    0.000000] e820 remove range: 00000000000a0000 - 0000000000100000 (usable)
-[    0.000000] last_pfn = 0x8100 max_arch_pfn = 0x1000000
+[    0.000000] last_pfn = 0x20800 max_arch_pfn = 0x1000000
 [    0.000000] initial memory mapped : 0 - 027ff000
 [    0.000000] Base memory trampoline at [c009f000] 9f000 size 4096
-[    0.000000] init_memory_mapping: 0000000000000000-0000000008100000
-[    0.000000]  0000000000 - 0008100000 page 4k
-[    0.000000] kernel direct mapping tables up to 8100000 @ 27bb000-27ff000
+[    0.000000] init_memory_mapping: 0000000000000000-0000000020800000
+[    0.000000]  0000000000 - 0020800000 page 4k
+[    0.000000] kernel direct mapping tables up to 20800000 @ 26f8000-27ff000
 [    0.000000] xen: setting RW the range 27e8000 - 27ff000
 [    0.000000] 0MB HIGHMEM available.
-[    0.000000] 129MB LOWMEM available.
-[    0.000000]   mapped low ram: 0 - 08100000
-[    0.000000]   low ram: 0 - 08100000
+[    0.000000] 520MB LOWMEM available.
+[    0.000000]   mapped low ram: 0 - 20800000
+[    0.000000]   low ram: 0 - 20800000

With this change "xl mem-set <domain> 512M" will successfully increase the
guest RAM (by reducing the balloon).

There is no change for dom0.

Reported-and-Tested-by: George Shuklin <george.shuklin@gmail.com>
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: stable@kernel.org
Reviewed-by: David Vrabel <david.vrabel@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
12 years agoBtrfs: only set cache_generation if we setup the block group
Josef Bacik [Tue, 13 Dec 2011 21:04:54 +0000 (16:04 -0500)]
Btrfs: only set cache_generation if we setup the block group

A user reported a problem booting into a new kernel with the old format inodes.
He was panicing in cow_file_range while writing out the inode cache.  This is
because if the block group is not cached we'll just skip writing out the cache,
however if it gets dirtied again in the same transaction and it finished caching
we'd go ahead and write it out, but since we set cache_generation to the transid
we think we've already truncated it and will just carry on, running into
cow_file_range and blowing up.  We need to make sure we only set
cache_generation if we've done the truncate.  The user tested this patch and
verified that the panic no longer occured.  Thanks,

Reported-and-Tested-by: Klaus Bitto <klaus.bitto@gmail.com>
Signed-off-by: Josef Bacik <josef@redhat.com>
12 years agoBtrfs: don't panic if orphan item already exists
Josef Bacik [Tue, 13 Dec 2011 17:55:58 +0000 (12:55 -0500)]
Btrfs: don't panic if orphan item already exists

I've been hitting this BUG_ON() in btrfs_orphan_add when running xfstest 269 in
a loop.  This is because we will add an orphan item, do the truncate, the
truncate will fail for whatever reason (*cough*ENOSPC*cough*) and then we're
left with an orphan item still in the fs.  Then we come back later to do another
truncate and it blows up because we already have an orphan item.  This is ok so
just fix the BUG_ON() to only BUG() if ret is not EEXIST.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
12 years agoBtrfs: fix leaked space in truncate
Josef Bacik [Fri, 9 Dec 2011 18:26:22 +0000 (13:26 -0500)]
Btrfs: fix leaked space in truncate

We were occasionaly leaking space when running xfstest 269.  This is because if
we failed to start the transaction in the truncate loop we'd just goto out, but
we need to break so that the inode is removed from the orphan list and the space
is properly freed.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
12 years agoBtrfs: fix how we do delalloc reservations and how we free reservations on error
Josef Bacik [Fri, 9 Dec 2011 16:18:51 +0000 (11:18 -0500)]
Btrfs: fix how we do delalloc reservations and how we free reservations on error

Running xfstests 269 with some tracing my scripts kept spitting out errors about
releasing bytes that we didn't actually have reserved.  This took me down a huge
rabbit hole and it turns out the way we deal with reserved_extents is wrong,
we need to only be setting it if the reservation succeeds, otherwise the free()
method will come in and unreserve space that isn't actually reserved yet, which
can lead to other warnings and such.  The math was all working out right in the
end, but it caused all sorts of other issues in addition to making my scripts
yell and scream and generally make it impossible for me to track down the
original issue I was looking for.  The other problem is with our error handling
in the reservation code.  There are two cases that we need to deal with

1) We raced with free.  In this case free won't free anything because csum_bytes
is modified before we dro the lock in our reservation path, so free rightly
doesn't release any space because the reservation code may be depending on that
reservation.  However if we fail, we need the reservation side to do the free at
that point since that space is no longer in use.  So as it stands the code was
doing this fine and it worked out, except in case #2

2) We don't race with free.  Nobody comes in and changes anything, and our
reservation fails.  In this case we didn't reserve anything anyway and we just
need to clean up csum_bytes but not free anything.  So we keep track of
csum_bytes before we drop the lock and if it hasn't changed we know we can just
decrement csum_bytes and carry on.

Because of the case where we can race with free()'s since we have to drop our
spin_lock to do the reservation, I'm going to serialize all reservations with
the i_mutex.  We already get this for free in the heavy use paths, truncate and
file write all hold the i_mutex, just needed to add it to page_mkwrite and
various ioctl/balance things.  With this patch my space leak scripts no longer
scream bloody murder.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
12 years agoBtrfs: deal with enospc from dirtying inodes properly
Josef Bacik [Wed, 30 Nov 2011 15:45:38 +0000 (10:45 -0500)]
Btrfs: deal with enospc from dirtying inodes properly

Now that we're properly keeping track of delayed inode space we've been getting
a lot of warnings out of btrfs_dirty_inode() when running xfstest 83.  This is
because a bunch of people call mark_inode_dirty, which is void so we can't
return ENOSPC.  This needs to be fixed in a few areas

1) file_update_time - this updates the mtime and such when writing to a file,
which will call mark_inode_dirty.  So copy file_update_time into btrfs so we can
call btrfs_dirty_inode directly and return an error if we get one appropriately.

2) fix symlinks to use btrfs_setattr for ->setattr.  For some reason we weren't
setting ->setattr for symlinks, even though we should have been.  This catches
one of the cases where we were getting errors in mark_inode_dirty.

3) Fix btrfs_setattr and btrfs_setsize to call btrfs_dirty_inode directly
instead of mark_inode_dirty.  This lets us return errors properly for truncate
and chown/anything related to setattr.

4) Add a new btrfs_fs_dirty_inode which will just call btrfs_dirty_inode and
print an error if we have one.  The only remaining user we can't control for
this is touch_atime(), but we don't really want to keep people from walking
down the tree if we don't have space to save the atime update, so just complain
but don't worry about it.

With this patch xfstests 83 complains a handful of times instead of hundreds of
times.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
12 years agoBtrfs: fix num_workers_starting bug and other bugs in async thread
Josef Bacik [Fri, 18 Nov 2011 19:37:27 +0000 (14:37 -0500)]
Btrfs: fix num_workers_starting bug and other bugs in async thread

Al pointed out we have some random problems with the way we account for
num_workers_starting in the async thread stuff.  First of all we need to make
sure to decrement num_workers_starting if we fail to start the worker, so make
__btrfs_start_workers do this.  Also fix __btrfs_start_workers so that it
doesn't call btrfs_stop_workers(), there is no point in stopping everybody if we
failed to create a worker.  Also check_pending_worker_creates needs to call
__btrfs_start_work in it's work function since it already increments
num_workers_starting.

People only start one worker at a time, so get rid of the num_workers argument
everywhere, and make btrfs_queue_worker a void since it will always succeed.
Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
12 years agoBTRFS: Establish i_ops before calling d_instantiate
Casey Schaufler [Thu, 15 Dec 2011 15:09:07 +0000 (10:09 -0500)]
BTRFS: Establish i_ops before calling d_instantiate

The Smack LSM hook for security_d_instantiate checks
the inode's i_op->getxattr value to determine if the
containing filesystem supports extended attributes.
The BTRFS filesystem sets the inode's i_op value only
after it has instantiated the inode. This results in
Smack incorrectly giving new BTRFS inodes attributes
from the filesystem defaults on the assumption that
values can't be stored on the filesystem. This patch
moves the assignment of inode operation vectors ahead
of the calls to d_instantiate, letting Smack know that
the filesystem supports extended attributes. There
should be no impact on the performance or behavior of
BTRFS.

Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoBtrfs: add a cond_resched() into the worker loop
Chris Mason [Thu, 15 Dec 2011 14:29:43 +0000 (09:29 -0500)]
Btrfs: add a cond_resched() into the worker loop

If we have a constant stream of end_io completions or crc work,
we can hit softlockup messages from the async helper threads.  This
adds a cond_resched() into the loop to avoid them.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoBtrfs: fix ctime update of on-disk inode
Li Zefan [Thu, 15 Dec 2011 01:12:02 +0000 (20:12 -0500)]
Btrfs: fix ctime update of on-disk inode

To reproduce the bug:

    # touch /mnt/tmp
    # stat /mnt/tmp | grep Change
    Change: 2011-12-09 09:32:23.412105981 +0800
    # chattr +i /mnt/tmp
    # stat /mnt/tmp | grep Change
    Change: 2011-12-09 09:32:43.198105295 +0800
    # umount /mnt
    # mount /dev/loop1 /mnt
    # stat /mnt/tmp | grep Change
    Change: 2011-12-09 09:32:23.412105981 +0800

We should update ctime of in-memory inode before calling
btrfs_update_inode().

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agobtrfs: keep orphans for subvolume deletion
Arne Jansen [Thu, 15 Dec 2011 01:12:02 +0000 (20:12 -0500)]
btrfs: keep orphans for subvolume deletion

Since we have the free space caches, btrfs_orphan_cleanup also runs for
the tree_root. Unfortunately this also cleans up the orphans used to mark
subvol deletions in progress.

Currently if a subvol deletion gets interrupted twice by umount/mount, the
deletion will not be continued and the space permanently lost, though it
would be possible to write a tool to recover those lost subvol deletions.
This patch checks if the orphan belongs to a subvol (dead root) and skips
the deletion.

Signed-off-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoBtrfs: fix inaccurate available space on raid0 profile
Miao Xie [Thu, 15 Dec 2011 01:12:02 +0000 (20:12 -0500)]
Btrfs: fix inaccurate available space on raid0 profile

When we use raid0 as the data profile, df command may show us a very
inaccurate value of the available space, which may be much less than the
real one. It may make the users puzzled. Fix it by changing the calculation
of the available space, and making it be more similar to a fake chunk
allocation.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoBtrfs: fix wrong disk space information of the files
Miao Xie [Thu, 15 Dec 2011 01:12:02 +0000 (20:12 -0500)]
Btrfs: fix wrong disk space information of the files

Btrfsck report errors after the 83th case of xfstests was run, The error
number is 400, it means the used disk space of the file is wrong.

The reason of this bug is that:
The file truncation may fail when the space of the file system is not enough,
and leave some file extents, whose offset are beyond the end of the files.
When we want to expand those files, we will drop those file extents, and
put in dummy file extents, and then we should update the i-node. But btrfs
forgets to do it.

This patch adds the forgotten i-node update.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoBtrfs: fix wrong i_size when truncating a file to a larger size
Miao Xie [Thu, 15 Dec 2011 01:12:01 +0000 (20:12 -0500)]
Btrfs: fix wrong i_size when truncating a file to a larger size

Btrfsck report error 100 after the 83th case of xfstests was run, it means
the i_size of the file is wrong.

The reason of this bug is that:
Btrfs increased i_size of the file at the beginning, but it failed to expand
the file, and failed to update the i_size to the old size because there is no
enough space in the file system, so we found a wrong i_size.

This patch fixes this bug by updating the i_size just when we pass the file
expanding and get enough space to update i-node.

Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
12 years agoMerge branch 'fix/asoc' into for-linus
Takashi Iwai [Thu, 15 Dec 2011 11:56:04 +0000 (12:56 +0100)]
Merge branch 'fix/asoc' into for-linus

12 years agoALSA: HDA: Use LPIB Position fix for Intel SCH Poulsbo
David Henningsson [Wed, 14 Dec 2011 07:52:30 +0000 (15:52 +0800)]
ALSA: HDA: Use LPIB Position fix for Intel SCH Poulsbo

Several people with this chipset have reported inconsistent/sloppy
values for position reporting when the DMA position buffer is used,
and that setting position_fix=1 have fixed their problems.

BugLink: https://bugs.launchpad.net/bugs/825709
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
12 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Thu, 15 Dec 2011 03:45:40 +0000 (19:45 -0800)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/radeon/kms: add some new pci ids

12 years agoMerge tag 'tytso-for-linus-20111214' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 15 Dec 2011 02:25:58 +0000 (18:25 -0800)]
Merge tag 'tytso-for-linus-20111214' of git://git./linux/kernel/git/tytso/ext4

* tag 'tytso-for-linus-20111214' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: handle EOF correctly in ext4_bio_write_page()
  ext4: remove a wrong BUG_ON in ext4_ext_convert_to_initialized
  ext4: correctly handle pages w/o buffers in ext4_discard_partial_buffers()
  ext4: avoid potential hang in mpage_submit_io() when blocksize < pagesize
  ext4: avoid hangs in ext4_da_should_update_i_disksize()
  ext4: display the correct mount option in /proc/mounts for [no]init_itable
  ext4: Fix crash due to getting bogus eh_depth value on big-endian systems
  ext4: fix ext4_end_io_dio() racing against fsync()

.. using the new signed tag merge of git that now verifies the gpg
signature automatically.  Yay.  The branchname was just 'dev', which is
prettier.  I'll tell Ted to use nicer tag names for future cases.

12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi...
Linus Torvalds [Thu, 15 Dec 2011 02:23:35 +0000 (18:23 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mszeredi/fuse

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
  fuse: llseek fix race
  fuse: fix llseek bug
  fuse: fix fuse_retrieve

12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Thu, 15 Dec 2011 02:22:55 +0000 (18:22 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/viro/vfs

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  fs/ncpfs: fix error paths and goto statements in ncp_fill_super()
  configfs: register_filesystem() called too early
  fuse: register_filesystem() called too early
  ubifs: too early register_filesystem()
  ... and the same kind of leak for mqueue
  procfs: fix a vfsmount longterm reference leak

12 years agosparc32: Be less strict in matching %lo part of relocation.
David S. Miller [Wed, 14 Dec 2011 18:05:22 +0000 (10:05 -0800)]
sparc32: Be less strict in matching %lo part of relocation.

The "(insn & 0x01800000) != 0x01800000" test matches 'restore'
but that is a legitimate place to see the %lo() part of a 32-bit
symbol relocation, particularly in tail calls.

Signed-off-by: David S. Miller <davem@davemloft.net>
Tested-by: Sergei Trofimovich <slyfox@gentoo.org>
12 years agodrm/radeon/kms: add some new pci ids
Alex Deucher [Mon, 12 Dec 2011 14:23:48 +0000 (09:23 -0500)]
drm/radeon/kms: add some new pci ids

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=43739

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
12 years agofs/ncpfs: fix error paths and goto statements in ncp_fill_super()
Djalal Harouni [Tue, 13 Dec 2011 01:47:29 +0000 (02:47 +0100)]
fs/ncpfs: fix error paths and goto statements in ncp_fill_super()

The label 'out_bdi' should be followed by bdi_destroy() instead of
fput() which should be after the 'out_fput' label.

If bdi_setup_and_register() fails then jump to the 'out_fput' label
instead of the 'out_bdi' one.

If fget(data.info_fd) fails then jump to the previously fixed 'out_bdi'
label to call bdi_destroy() otherwise the bdi object will not be
destroyed.

Compile tested only.

Signed-off-by: Djalal Harouni <tixxdz@opendz.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
12 years agoext4: handle EOF correctly in ext4_bio_write_page()
Yongqiang Yang [Wed, 14 Dec 2011 03:29:12 +0000 (22:29 -0500)]
ext4: handle EOF correctly in ext4_bio_write_page()

We need to zero out part of a page which beyond EOF before setting uptodate,
otherwise, mapread or write will see non-zero data beyond EOF.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
12 years agoext4: remove a wrong BUG_ON in ext4_ext_convert_to_initialized
Yongqiang Yang [Wed, 14 Dec 2011 03:13:42 +0000 (22:13 -0500)]
ext4: remove a wrong BUG_ON in ext4_ext_convert_to_initialized

If a file is fallocated on a hole, map->m_lblk + map->m_len may be greater
than ee_block + ee_len.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
12 years agoext4: correctly handle pages w/o buffers in ext4_discard_partial_buffers()
Yongqiang Yang [Wed, 14 Dec 2011 03:05:05 +0000 (22:05 -0500)]
ext4: correctly handle pages w/o buffers in ext4_discard_partial_buffers()

If a page has been read into memory and never been written, it has no
buffers, but we should handle the page in truncate or punch hole.

VFS code of writing operations has handled holes correctly, so this
patch removes the code handling holes in writing operations.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
12 years agoext4: avoid potential hang in mpage_submit_io() when blocksize < pagesize
Yongqiang Yang [Wed, 14 Dec 2011 02:51:55 +0000 (21:51 -0500)]
ext4: avoid potential hang in mpage_submit_io() when blocksize < pagesize

If there is an unwritten but clean buffer in a page and there is a
dirty buffer after the buffer, then mpage_submit_io does not write the
dirty buffer out.  As a result, da_writepages loops forever.

This patch fixes the problem by checking dirty flag.

Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
12 years agoext4: avoid hangs in ext4_da_should_update_i_disksize()
Andrea Arcangeli [Wed, 14 Dec 2011 02:41:15 +0000 (21:41 -0500)]
ext4: avoid hangs in ext4_da_should_update_i_disksize()

If the pte mapping in generic_perform_write() is unmapped between
iov_iter_fault_in_readable() and iov_iter_copy_from_user_atomic(), the
"copied" parameter to ->end_write can be zero. ext4 couldn't cope with
it with delayed allocations enabled. This skips the i_disksize
enlargement logic if copied is zero and no new data was appeneded to
the inode.

 gdb> bt
 #0  0xffffffff811afe80 in ext4_da_should_update_i_disksize (file=0xffff88003f606a80, mapping=0xffff88001d3824e0, pos=0x1\
 08000, len=0x1000, copied=0x0, page=0xffffea0000d792e8, fsdata=0x0) at fs/ext4/inode.c:2467
 #1  ext4_da_write_end (file=0xffff88003f606a80, mapping=0xffff88001d3824e0, pos=0x108000, len=0x1000, copied=0x0, page=0\
 xffffea0000d792e8, fsdata=0x0) at fs/ext4/inode.c:2512
 #2  0xffffffff810d97f1 in generic_perform_write (iocb=<value optimized out>, iov=<value optimized out>, nr_segs=<value o\
 ptimized out>, pos=0x108000, ppos=0xffff88001e26be40, count=<value optimized out>, written=0x0) at mm/filemap.c:2440
 #3  generic_file_buffered_write (iocb=<value optimized out>, iov=<value optimized out>, nr_segs=<value optimized out>, p\
 os=0x108000, ppos=0xffff88001e26be40, count=<value optimized out>, written=0x0) at mm/filemap.c:2482
 #4  0xffffffff810db5d1 in __generic_file_aio_write (iocb=0xffff88001e26bde8, iov=0xffff88001e26bec8, nr_segs=0x1, ppos=0\
 xffff88001e26be40) at mm/filemap.c:2600
 #5  0xffffffff810db853 in generic_file_aio_write (iocb=0xffff88001e26bde8, iov=0xffff88001e26bec8, nr_segs=<value optimi\
 zed out>, pos=<value optimized out>) at mm/filemap.c:2632
 #6  0xffffffff811a71aa in ext4_file_write (iocb=0xffff88001e26bde8, iov=0xffff88001e26bec8, nr_segs=0x1, pos=0x108000) a\
 t fs/ext4/file.c:136
 #7  0xffffffff811375aa in do_sync_write (filp=0xffff88003f606a80, buf=<value optimized out>, len=<value optimized out>, \
 ppos=0xffff88001e26bf48) at fs/read_write.c:406
 #8  0xffffffff81137e56 in vfs_write (file=0xffff88003f606a80, buf=0x1ec2960 <Address 0x1ec2960 out of bounds>, count=0x4\
 000, pos=0xffff88001e26bf48) at fs/read_write.c:435
 #9  0xffffffff8113816c in sys_write (fd=<value optimized out>, buf=0x1ec2960 <Address 0x1ec2960 out of bounds>, count=0x\
 4000) at fs/read_write.c:487
 #10 <signal handler called>
 #11 0x00007f120077a390 in __brk_reservation_fn_dmi_alloc__ ()
 #12 0x0000000000000000 in ?? ()
 gdb> print offset
 $22 = 0xffffffffffffffff
 gdb> print idx
 $23 = 0xffffffff
 gdb> print inode->i_blkbits
 $24 = 0xc
 gdb> up
 #1  ext4_da_write_end (file=0xffff88003f606a80, mapping=0xffff88001d3824e0, pos=0x108000, len=0x1000, copied=0x0, page=0\
 xffffea0000d792e8, fsdata=0x0) at fs/ext4/inode.c:2512
 2512                    if (ext4_da_should_update_i_disksize(page, end)) {
 gdb> print start
 $25 = 0x0
 gdb> print end
 $26 = 0xffffffffffffffff
 gdb> print pos
 $27 = 0x108000
 gdb> print new_i_size
 $28 = 0x108000
 gdb> print ((struct ext4_inode_info *)((char *)inode-((int)(&((struct ext4_inode_info *)0)->vfs_inode))))->i_disksize
 $29 = 0xd9000
 gdb> down
 2467            for (i = 0; i < idx; i++)
 gdb> print i
 $30 = 0xd44acbee

This is 100% reproducible with some autonuma development code tuned in
a very aggressive manner (not normal way even for knumad) which does
"exotic" changes to the ptes. It wouldn't normally trigger but I don't
see why it can't happen normally if the page is added to swap cache in
between the two faults leading to "copied" being zero (which then
hangs in ext4). So it should be fixed. Especially possible with lumpy
reclaim (albeit disabled if compaction is enabled) as that would
ignore the young bits in the ptes.

Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
12 years agostaging: r8712u: Add new USB ID
Larry Finger [Sun, 11 Dec 2011 16:27:54 +0000 (10:27 -0600)]
staging: r8712u: Add new USB ID

Add USB ID for Sitecom WLA-2000 v1.001 WLAN.

Reported-and-tested-by: Roland Gruber <post@rolandgruber.de>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: tidspbridge: request dmtimer clocks on init
Omar Ramirez Luna [Fri, 18 Nov 2011 22:18:54 +0000 (16:18 -0600)]
staging: tidspbridge: request dmtimer clocks on init

Given that dm timer framework doesn't support request of clocks
by soft | hard irqs because some recent changes, tidspbridge needs
to request its clocks on init and enable/disable them on demand.

This was first seen on 3.2-rc1.

Signed-off-by: Omar Ramirez Luna <omar.ramirez@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agostaging: tidspbridge: include module.h by default
Omar Ramirez Luna [Tue, 15 Nov 2011 02:16:57 +0000 (20:16 -0600)]
staging: tidspbridge: include module.h by default

Fixes compilation break when compiled as part of the kernel:

drivers/staging/tidspbridge/rmgr/drv_interface.c:134: error: expected declaration specifiers or '...' before string constant
drivers/staging/tidspbridge/rmgr/drv_interface.c:134: warning: data definition has no type or storage class
drivers/staging/tidspbridge/rmgr/drv_interface.c:134: warning: type defaults to 'int' in declaration of 'MODULE_AUTHOR'
drivers/staging/tidspbridge/rmgr/drv_interface.c:134: warning: function declaration isn't a prototype
drivers/staging/tidspbridge/rmgr/drv_interface.c:135: error: expected declaration specifiers or '...' before string constant
drivers/staging/tidspbridge/rmgr/drv_interface.c:135: warning: data definition has no type or storage class
drivers/staging/tidspbridge/rmgr/drv_interface.c:135: warning: type defaults to 'int' in declaration of 'MODULE_LICENSE'
drivers/staging/tidspbridge/rmgr/drv_interface.c:135: warning: function declaration isn't a prototype
drivers/staging/tidspbridge/rmgr/drv_interface.c:136: error: expected declaration specifiers or '...' before string constant
drivers/staging/tidspbridge/rmgr/drv_interface.c:136: warning: data definition has no type or storage class
drivers/staging/tidspbridge/rmgr/drv_interface.c:136: warning: type defaults to 'int' in declaration of 'MODULE_VERSION'
drivers/staging/tidspbridge/rmgr/drv_interface.c:136: warning: function declaration isn't a prototype
drivers/staging/tidspbridge/rmgr/drv_interface.c: In function 'omap34_xx_bridge_probe':
drivers/staging/tidspbridge/rmgr/drv_interface.c:359: error: 'THIS_MODULE' undeclared (first use in this function)
drivers/staging/tidspbridge/rmgr/drv_interface.c:359: error: (Each undeclared identifier is reported only once
drivers/staging/tidspbridge/rmgr/drv_interface.c:359: error: for each function it appears in.)

Signed-off-by: Omar Ramirez Luna <omar.ramirez@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
12 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Tue, 13 Dec 2011 23:02:31 +0000 (15:02 -0800)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  Revert "x86, efi: Calling __pa() with an ioremap()ed address is invalid"
  x86, efi: Make efi_call_phys_{prelog,epilog} CONFIG_RELOCATABLE-aware

12 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph...
Linus Torvalds [Tue, 13 Dec 2011 22:59:42 +0000 (14:59 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/sage/ceph-client

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
  ceph: add missing spin_unlock at ceph_mdsc_build_path()
  ceph: fix SEEK_CUR, SEEK_SET regression
  crush: fix mapping calculation when force argument doesn't exist
  ceph: use i_ceph_lock instead of i_lock
  rbd: remove buggy rollback functionality
  rbd: return an error when an invalid header is read
  ceph: fix rasize reporting by ceph_show_options

12 years agoMerge branch 'writeback-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Tue, 13 Dec 2011 22:58:56 +0000 (14:58 -0800)]
Merge branch 'writeback-for-linus' of git://git./linux/kernel/git/wfg/linux

* 'writeback-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/wfg/linux:
  writeback: set max_pause to lowest value on zero bdi_dirty
  writeback: permit through good bdi even when global dirty exceeded
  writeback: comment on the bdi dirty threshold
  fs: Make write(2) interruptible by a fatal signal
  writeback: Fix issue on make htmldocs