openwrt/staging/blogic.git
9 years agoext4: fix growing of tiny filesystems
Jan Kara [Sun, 3 May 2015 03:58:32 +0000 (23:58 -0400)]
ext4: fix growing of tiny filesystems

The estimate of necessary transaction credits in ext4_flex_group_add()
is too pessimistic. It reserves credit for sb, resize inode, and resize
inode dindirect block for each group added in a flex group although they
are always the same block and thus it is enough to account them only
once. Also the number of modified GDT block is overestimated since we
fit EXT4_DESC_PER_BLOCK(sb) descriptors in one block.

Make the estimation more precise. That reduces number of requested
credits enough that we can grow 20 MB filesystem (which has 1 MB
journal, 79 reserved GDT blocks, and flex group size 16 by default).

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
9 years agoext4: move check under lock scope to close a race.
Davide Italiano [Sun, 3 May 2015 03:21:15 +0000 (23:21 -0400)]
ext4: move check under lock scope to close a race.

fallocate() checks that the file is extent-based and returns
EOPNOTSUPP in case is not. Other tasks can convert from and to
indirect and extent so it's safe to check only after grabbing
the inode mutex.

Signed-off-by: Davide Italiano <dccitaliano@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
9 years agoext4: fix data corruption caused by unwritten and delayed extents
Lukas Czerner [Sun, 3 May 2015 01:36:55 +0000 (21:36 -0400)]
ext4: fix data corruption caused by unwritten and delayed extents

Currently it is possible to lose whole file system block worth of data
when we hit the specific interaction with unwritten and delayed extents
in status extent tree.

The problem is that when we insert delayed extent into extent status
tree the only way to get rid of it is when we write out delayed buffer.
However there is a limitation in the extent status tree implementation
so that when inserting unwritten extent should there be even a single
delayed block the whole unwritten extent would be marked as delayed.

At this point, there is no way to get rid of the delayed extents,
because there are no delayed buffers to write out. So when a we write
into said unwritten extent we will convert it to written, but it still
remains delayed.

When we try to write into that block later ext4_da_map_blocks() will set
the buffer new and delayed and map it to invalid block which causes
the rest of the block to be zeroed loosing already written data.

For now we can fix this by simply not allowing to set delayed status on
written extent in the extent status tree. Also add WARN_ON() to make
sure that we notice if this happens in the future.

This problem can be easily reproduced by running the following xfs_io.

xfs_io -f -c "pwrite -S 0xaa 4096 2048" \
          -c "falloc 0 131072" \
          -c "pwrite -S 0xbb 65536 2048" \
          -c "fsync" /mnt/test/fff

echo 3 > /proc/sys/vm/drop_caches
xfs_io -c "pwrite -S 0xdd 67584 2048" /mnt/test/fff

This can be theoretically also reproduced by at random by running fsx,
but it's not very reliable, though on machines with bigger page size
(like ppc) this can be seen more often (especially xfstest generic/127)

Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@vger.kernel.org
9 years agoext4 crypto: remove duplicated encryption mode definitions
Chanho Park [Sat, 2 May 2015 14:29:22 +0000 (10:29 -0400)]
ext4 crypto: remove duplicated encryption mode definitions

This patch removes duplicated encryption modes which were already in
ext4.h. They were duplicated from commit 3edc18d and commit f542fb.

Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Michael Halcrow <mhalcrow@google.com>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Signed-off-by: Chanho Park <chanho61.park@samsung.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: do not select from EXT4_FS_ENCRYPTION
Herbert Xu [Sat, 2 May 2015 14:29:19 +0000 (10:29 -0400)]
ext4 crypto: do not select from EXT4_FS_ENCRYPTION

This patch adds a tristate EXT4_ENCRYPTION to do the selections
for EXT4_FS_ENCRYPTION because selecting from a bool causes all
the selected options to be built-in, even if EXT4 itself is a
module.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: add padding to filenames before encrypting
Theodore Ts'o [Fri, 1 May 2015 20:56:50 +0000 (16:56 -0400)]
ext4 crypto: add padding to filenames before encrypting

This obscures the length of the filenames, to decrease the amount of
information leakage.  By default, we pad the filenames to the next 4
byte boundaries.  This costs nothing, since the directory entries are
aligned to 4 byte boundaries anyway.  Filenames can also be padded to
8, 16, or 32 bytes, which will consume more directory space.

Change-Id: Ibb7a0fb76d2c48e2061240a709358ff40b14f322
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: simplify and speed up filename encryption
Theodore Ts'o [Fri, 1 May 2015 20:56:45 +0000 (16:56 -0400)]
ext4 crypto: simplify and speed up filename encryption

Avoid using SHA-1 when calculating the user-visible filename when the
encryption key is available, and avoid decrypting lots of filenames
when searching for a directory entry in a directory block.

Change-Id: If4655f144784978ba0305b597bfa1c8d7bb69e63
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: enable encryption feature flag
Theodore Ts'o [Thu, 16 Apr 2015 05:56:00 +0000 (01:56 -0400)]
ext4 crypto: enable encryption feature flag

Also add the test dummy encryption mode flag so we can more easily
test the encryption patches using xfstests.

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: add symlink encryption
Theodore Ts'o [Thu, 16 Apr 2015 05:55:00 +0000 (01:55 -0400)]
ext4 crypto: add symlink encryption

Signed-off-by: Uday Savagaonkar <savagaon@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: enable filename encryption
Michael Halcrow [Sun, 12 Apr 2015 05:09:05 +0000 (01:09 -0400)]
ext4 crypto: enable filename encryption

Signed-off-by: Uday Savagaonkar <savagaon@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: filename encryption modifications
Michael Halcrow [Sun, 12 Apr 2015 05:09:03 +0000 (01:09 -0400)]
ext4 crypto: filename encryption modifications

Modifies htree_dirblock_to_tree, dx_make_map, ext4_match search_dir,
and ext4_find_dest_de to support fname crypto.  Filename encryption
feature is not yet enabled at this patch.

Signed-off-by: Uday Savagaonkar <savagaon@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: partial update to namei.c for fname crypto
Michael Halcrow [Sun, 12 Apr 2015 05:07:01 +0000 (01:07 -0400)]
ext4 crypto: partial update to namei.c for fname crypto

Modifies dx_show_leaf and dx_probe to support fname encryption.
Filename encryption not yet enabled.

Signed-off-by: Uday Savagaonkar <savagaon@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: insert encrypted filenames into a leaf directory block
Michael Halcrow [Sun, 12 Apr 2015 04:56:28 +0000 (00:56 -0400)]
ext4 crypto: insert encrypted filenames into a leaf directory block

Signed-off-by: Uday Savagaonkar <savagaon@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: teach ext4_htree_store_dirent() to store decrypted filenames
Theodore Ts'o [Sun, 12 Apr 2015 04:56:26 +0000 (00:56 -0400)]
ext4 crypto: teach ext4_htree_store_dirent() to store decrypted filenames

For encrypted directories, we need to pass in a separate parameter for
the decrypted filename, since the directory entry contains the
encrypted filename.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: filename encryption facilities
Michael Halcrow [Sun, 12 Apr 2015 04:56:17 +0000 (00:56 -0400)]
ext4 crypto: filename encryption facilities

Signed-off-by: Uday Savagaonkar <savagaon@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: implement the ext4 decryption read path
Michael Halcrow [Sun, 12 Apr 2015 04:56:10 +0000 (00:56 -0400)]
ext4 crypto: implement the ext4 decryption read path

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: implement the ext4 encryption write path
Michael Halcrow [Sun, 12 Apr 2015 04:55:10 +0000 (00:55 -0400)]
ext4 crypto: implement the ext4 encryption write path

Pulls block_write_begin() into fs/ext4/inode.c because it might need
to do a low-level read of the existing data, in which case we need to
decrypt it.

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: inherit encryption policies on inode and directory create
Michael Halcrow [Sun, 12 Apr 2015 04:55:09 +0000 (00:55 -0400)]
ext4 crypto: inherit encryption policies on inode and directory create

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: enforce context consistency
Theodore Ts'o [Sun, 12 Apr 2015 04:55:08 +0000 (00:55 -0400)]
ext4 crypto: enforce context consistency

Enforce the following inheritance policy:

1) An unencrypted directory may contain encrypted or unencrypted files
or directories.

2) All files or directories in a directory must be protected using the
same key as their containing directory.

As a result, assuming the following setup:

mke2fs -t ext4 -Fq -O encrypt /dev/vdc
mount -t ext4 /dev/vdc /vdc
mkdir /vdc/a /vdc/b /vdc/c
echo foo | e4crypt add_key /vdc/a
echo bar | e4crypt add_key /vdc/b
for i in a b c ; do cp /etc/motd /vdc/$i/motd-$i ; done

Then we will see the following results:

cd /vdc
mv a b # will fail; /vdc/a and /vdc/b have different keys
mv b/motd-b a # will fail, see above
ln a/motd-a b # will fail, see above
mv c a      # will fail; all inodes in an encrypted directory
       # must be encrypted
ln c/motd-c b # will fail, see above
mv a/motd-a c # will succeed
mv c/motd-a a # will succeed

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: add encryption key management facilities
Michael Halcrow [Sun, 12 Apr 2015 04:55:06 +0000 (00:55 -0400)]
ext4 crypto: add encryption key management facilities

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Ildar Muslukhov <muslukhovi@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: add ext4 encryption facilities
Michael Halcrow [Sun, 12 Apr 2015 04:43:56 +0000 (00:43 -0400)]
ext4 crypto: add ext4 encryption facilities

On encrypt, we will re-assign the buffer_heads to point to a bounce
page rather than the control_page (which is the original page to write
that contains the plaintext). The block I/O occurs against the bounce
page.  On write completion, we re-assign the buffer_heads to the
original plaintext page.

On decrypt, we will attach a read completion callback to the bio
struct. This read completion will decrypt the read contents in-place
prior to setting the page up-to-date.

The current encryption mode, AES-256-XTS, lacks cryptographic
integrity. AES-256-GCM is in-plan, but we will need to devise a
mechanism for handling the integrity data.

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Ildar Muslukhov <ildarm@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: add encryption policy and password salt support
Michael Halcrow [Sat, 11 Apr 2015 11:48:01 +0000 (07:48 -0400)]
ext4 crypto: add encryption policy and password salt support

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Ildar Muslukhov <muslukhovi@gmail.com>
9 years agoext4 crypto: add encryption xattr support
Michael Halcrow [Sat, 11 Apr 2015 11:47:00 +0000 (07:47 -0400)]
ext4 crypto: add encryption xattr support

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: export ext4_empty_dir()
Michael Halcrow [Sat, 11 Apr 2015 11:46:49 +0000 (07:46 -0400)]
ext4 crypto: export ext4_empty_dir()

Required for future encryption xattr changes.

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: add ext4 encryption Kconfig
Theodore Ts'o [Sat, 11 Apr 2015 11:46:47 +0000 (07:46 -0400)]
ext4 crypto: add ext4 encryption Kconfig

Signed-off-by: Michael Halcrow <mhalcrow@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: reserve codepoints used by the ext4 encryption feature
Theodore Ts'o [Sat, 11 Apr 2015 11:44:12 +0000 (07:44 -0400)]
ext4 crypto: reserve codepoints used by the ext4 encryption feature

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4 crypto: add ext4_mpage_readpages()
Theodore Ts'o [Wed, 8 Apr 2015 04:00:32 +0000 (00:00 -0400)]
ext4 crypto: add ext4_mpage_readpages()

This takes code from fs/mpage.c and optimizes it for ext4.  Its
primary reason is to allow us to more easily add encryption to ext4's
read path in an efficient manner.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: make fsync to sync parent dir in no-journal for real this time
Lukas Czerner [Fri, 3 Apr 2015 14:46:58 +0000 (10:46 -0400)]
ext4: make fsync to sync parent dir in no-journal for real this time

Previously commit 14ece1028b3ed53ffec1b1213ffc6acaf79ad77c added a
support for for syncing parent directory of newly created inodes to
make sure that the inode is not lost after a power failure in
no-journal mode.

However this does not work in majority of cases, namely:
 - if the directory has inline data
 - if the directory is already indexed
 - if the directory already has at least one block and:
- the new entry fits into it
- or we've successfully converted it to indexed

So in those cases we might lose the inode entirely even after fsync in
the no-journal mode. This also includes ext2 default mode obviously.

I've noticed this while running xfstest generic/321 and even though the
test should fail (we need to run fsck after a crash in no-journal mode)
I could not find a newly created entries even when if it was fsynced
before.

Fix this by adjusting the ext4_add_entry() successful exit paths to set
the inode EXT4_STATE_NEWENTRY so that fsync has the chance to fsync the
parent directory as well.

Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Frank Mayhar <fmayhar@google.com>
Cc: stable@vger.kernel.org
9 years agoext4: don't release reserved space for previously allocated cluster
Eric Whitney [Fri, 3 Apr 2015 04:17:31 +0000 (00:17 -0400)]
ext4: don't release reserved space for previously allocated cluster

When xfstests' auto group is run on a bigalloc filesystem with a
4.0-rc3 kernel, e2fsck failures and kernel warnings occur for some
tests. e2fsck reports incorrect iblocks values, and the warnings
indicate that the space reserved for delayed allocation is being
overdrawn at allocation time.

Some of these errors occur because the reserved space is incorrectly
decreased by one cluster when ext4_ext_map_blocks satisfies an
allocation request by mapping an unused portion of a previously
allocated cluster.  Because a cluster's worth of reserved space was
already released when it was first allocated, it should not be released
again.

This patch appears to correct the e2fsck failure reported for
generic/232 and the kernel warnings produced by ext4/001, generic/009,
and generic/033.  Failures and warnings for some other tests remain to
be addressed.

Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: fix loss of delalloc extent info in ext4_zero_range()
Eric Whitney [Fri, 3 Apr 2015 04:13:42 +0000 (00:13 -0400)]
ext4: fix loss of delalloc extent info in ext4_zero_range()

In ext4_zero_range(), removing a file's entire block range from the
extent status tree removes all records of that file's delalloc extents.
The delalloc accounting code uses this information, and its loss can
then lead to accounting errors and kernel warnings at writeback time and
subsequent file system damage.  This is most noticeable on bigalloc
file systems where code in ext4_ext_map_blocks() handles cases where
delalloc extents share clusters with a newly allocated extent.

Because we're not deleting a block range and are correctly updating the
status of its associated extent, there is no need to remove anything
from the extent status tree.

When this patch is combined with an unrelated bug fix for
ext4_zero_range(), kernel warnings and e2fsck errors reported during
xfstests runs on bigalloc filesystems are greatly reduced without
introducing regressions on other xfstests-bld test scenarios.

Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: allocate entire range in zero range
Lukas Czerner [Fri, 3 Apr 2015 04:09:13 +0000 (00:09 -0400)]
ext4: allocate entire range in zero range

Currently there is a bug in zero range code which causes zero range
calls to only allocate block aligned portion of the range, while
ignoring the rest in some cases.

In some cases, namely if the end of the range is past i_size, we do
attempt to preallocate the last nonaligned block. However this might
cause kernel to BUG() in some carefully designed zero range requests
on setups where page size > block size.

Fix this problem by first preallocating the entire range, including
the nonaligned edges and converting the written extents to unwritten
in the next step. This approach will also give us the advantage of
having the range to be as linearly contiguous as possible.

Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: remove unnecessary lock/unlock of i_block_reservation_lock
Maurizio Lombardi [Fri, 3 Apr 2015 04:02:53 +0000 (00:02 -0400)]
ext4: remove unnecessary lock/unlock of i_block_reservation_lock

This is a leftover of commit 71d4f7d032149b935a26eb3ff85c6c837f3714e1

Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Lukas Czerner <lczerner@redhat.com>
9 years agoext4: remove block_device_ejected
Christoph Hellwig [Fri, 3 Apr 2015 03:56:32 +0000 (23:56 -0400)]
ext4: remove block_device_ejected

bdi->dev now never goes away, so this function became useless.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: remove useless condition in if statement.
Wei Yuan [Fri, 3 Apr 2015 03:50:48 +0000 (23:50 -0400)]
ext4: remove useless condition in if statement.

In this if statement, the previous condition is useless, the later one
has covered it.

Signed-off-by: Weiyuan <weiyuan.wei@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Lukas Czerner <lczerner@redhat.com>
9 years agoext4: remove unused header files
Sheng Yong [Fri, 3 Apr 2015 03:47:42 +0000 (23:47 -0400)]
ext4: remove unused header files

Remove unused header files and header files which are included in
ext4.h.

Signed-off-by: Sheng Yong <shengyong1@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: fix comments in ext4_can_extents_be_merged()
Xiaoguang Wang [Thu, 2 Apr 2015 20:53:11 +0000 (16:53 -0400)]
ext4: fix comments in ext4_can_extents_be_merged()

Since commit a9b8241594add, we are allowed to merge unwritten extents,
so here these comments are wrong, remove it.

Signed-off-by: Xiaoguang Wang <wangxg.fnst@cn.fujitsu.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: fix transposition typo in format string
Rasmus Villemoes [Thu, 2 Apr 2015 20:42:43 +0000 (16:42 -0400)]
ext4: fix transposition typo in format string

According to C99, %*.s means the same as %*.0s, in other words, print as
many spaces as the field width argument says and effectively ignore the
string argument. That is certainly not what was meant here. The kernel's
printf implementation, however, treats it as if the . was not there,
i.e. as %*s. I don't know if de->name is nul-terminated or not, but in
any case I'm guessing the intention was to use de->name_len as precision
instead of field width.

[ Note: this is debugging code which is commented out, so this is not
  security issue; a developer would have to explicitly enable
  INLINE_DIR_DEBUG before this would be an issue. ]

Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
9 years agoext4: fix bh leak on error paths in ext4_rename() and ext4_cross_rename()
Konstantin Khlebnikov [Thu, 2 Apr 2015 20:32:15 +0000 (16:32 -0400)]
ext4: fix bh leak on error paths in ext4_rename() and ext4_cross_rename()

Release references to buffer-heads if ext4_journal_start() fails.

Fixes: 5b61de757535 ("ext4: start handle at least possible moment when renaming files")
Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
9 years agofs: add dirtytime_expire_seconds sysctl
Theodore Ts'o [Tue, 17 Mar 2015 16:23:32 +0000 (12:23 -0400)]
fs: add dirtytime_expire_seconds sysctl

Add a tuning knob so we can adjust the dirtytime expiration timeout,
which is very useful for testing lazytime.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
9 years agofs: make sure the timestamps for lazytime inodes eventually get written
Theodore Ts'o [Tue, 17 Mar 2015 16:23:19 +0000 (12:23 -0400)]
fs: make sure the timestamps for lazytime inodes eventually get written

Jan Kara pointed out that if there is an inode which is constantly
getting dirtied with I_DIRTY_PAGES, an inode with an updated timestamp
will never be written since inode->dirtied_when is constantly getting
updated.  We fix this by adding an extra field to the inode,
dirtied_time_when, so inodes with a stale dirtytime can get detected
and handled.

In addition, if we have a dirtytime inode caused by an atime update,
and there is no write activity on the file system, we need to have a
secondary system to make sure these inodes get written out.  We do
this by setting up a second delayed work structure which wakes up the
CPU much more rarely compared to writeback_expire_centisecs.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
9 years agoLinux 4.0-rc2
Linus Torvalds [Tue, 3 Mar 2015 17:04:59 +0000 (09:04 -0800)]
Linux 4.0-rc2

9 years agodrm/i915: Fix modeset state confusion in the load detect code
Daniel Vetter [Tue, 3 Mar 2015 16:31:21 +0000 (17:31 +0100)]
drm/i915: Fix modeset state confusion in the load detect code

This is a tricky story of the new atomic state handling and the legacy
code fighting over each another. The bug at hand is an underrun of the
framebuffer reference with subsequent hilarity caused by the load
detect code. Which is peculiar since the the exact same code works
fine as the implementation of the legacy setcrtc ioctl.

Let's look at the ingredients:

- Currently our code is a crazy mix of legacy modeset interfaces to
  set the parameters and half-baked atomic state tracking underneath.
  While this transition is going we're using the transitional plane
  helpers to update the atomic side (drm_plane_helper_disable/update
  and friends), i.e. plane->state->fb. Since the state structure owns
  the fb those functions take care of that themselves.

  The legacy state (specifically crtc->primary->fb) is still managed
  by the old code (and mostly by the drm core), with the fb reference
  counting done by callers (core drm for the ioctl or the i915 load
  detect code). The relevant commit is

  commit ea2c67bb4affa84080c616920f3899f123786e56
  Author: Matt Roper <matthew.d.roper@intel.com>
  Date:   Tue Dec 23 10:41:52 2014 -0800

      drm/i915: Move to atomic plane helpers (v9)

- drm_plane_helper_disable has special code to handle multiple calls
  in a row - it checks plane->crtc == NULL and bails out. This is to
  match the proper atomic implementation which needs the crtc to get
  at the implied locking context atomic updates always need. See

  commit acf24a395c5a9290189b080383564437101d411c
  Author: Daniel Vetter <daniel.vetter@ffwll.ch>
  Date:   Tue Jul 29 15:33:05 2014 +0200

      drm/plane-helper: transitional atomic plane helpers

- The universal plane code split out the implicit primary plane from
  the CRTC into it's own full-blown drm_plane object. As part of that
  the setcrtc ioctl (which updated both the crtc mode and primary
  plane) learned to set crtc->primary->crtc on modeset to make sure
  the plane->crtc assignments statate up to date in

  commit e13161af80c185ecd8dc4641d0f5df58f9e3e0af
  Author: Matt Roper <matthew.d.roper@intel.com>
  Date:   Tue Apr 1 15:22:38 2014 -0700

      drm: Add drm_crtc_init_with_planes() (v2)

  Unfortunately we've forgotten to update the load detect code. Which
  wasn't a problem since the load detect modeset is temporary and
  always undone before we drop the locks.

- Finally there is a organically grown history (i.e. don't ask) around
  who sets the legacy plane->fb for the various driver entry points.
  Originally updating that was the drivers duty, but for almost all
  places we've moved that (plus updating the refcounts) into the core.
  Again the exception is the load detect code.

Taking all together the following happens:
- The load detect code doesn't set crtc->primary->crtc. This is only
  really an issue on crtcs never before used or when userspace
  explicitly disabled the primary plane.

- The plane helper glue code short-circuits because of that and leaves
  a non-NULL fb behind in plane->state->fb and plane->fb. The state
  fb isn't a real problem (it's properly refcounted on its own), it's
  just the canary.

- Load detect code drops the reference for that fb, but doesn't set
  plane->fb = NULL. This is ok since it's still living in that old
  world where drivers had to clear the pointer but the core/callers
  handled the refcounting.

- On the next modeset the drm core notices plane->fb and takes care of
  refcounting it properly by doing another unref. This drops the
  refcount to zero, leaving state->plane now pointing at freed memory.

- intel_plane_duplicate_state still assume it owns a reference to that
  very state->fb and bad things start to happen.

Fix this all by applying the same duct-tape as for the legacy setcrtc
ioctl code and set crtc->primary->crtc properly.

Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Paul Bolle <pebolle@tiscali.nl>
Cc: Rob Clark <robdclark@gmail.com>
Cc: Paulo Zanoni <przanoni@gmail.com>
Cc: Sean Paul <seanpaul@chromium.org>
Cc: Matt Roper <matthew.d.roper@intel.com>
Reported-and-tested-by: Linus Torvalds <torvalds@linux-foundation.org>
Reported-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agoMerge tag 'gpio-v4.0-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux...
Linus Torvalds [Mon, 2 Mar 2015 22:13:39 +0000 (14:13 -0800)]
Merge tag 'gpio-v4.0-2' of git://git./linux/kernel/git/linusw/linux-gpio

Pull GPIO fixes from Linus Walleij:
 "Two GPIO fixes:

   - Fix a translation problem in of_get_named_gpiod_flags()

   - Fix a long standing container_of() mistake in the TPS65912 driver"

* tag 'gpio-v4.0-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio:
  gpio: tps65912: fix wrong container_of arguments
  gpiolib: of: allow of_gpiochip_find_and_xlate to find more than one chip per node

9 years agoMerge branch 'fixes-for-4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 2 Mar 2015 22:08:10 +0000 (14:08 -0800)]
Merge branch 'fixes-for-4.0-rc2' of git://git./linux/kernel/git/evalenti/linux-soc-thermal

Pull thermal management fixes from Eduardo Valentin:
 "Specifics:

   - Several fixes in tmon tool.

   - Fixes in intel int340x for _ART and _TRT tables.

   - Add id for Avoton SoC into powerclamp driver.

   - Fixes in RCAR thermal driver to remove race conditions and fix fail
     path

   - Fixes in TI thermal driver: removal of unnecessary code and build
     fix if !CONFIG_PM_SLEEP

   - Cleanups in exynos thermal driver

   - Add stubs for include/linux/thermal.h.  Now drivers using thermal
     calls but that also work without CONFIG_THERMAL will be able to
     compile for systems that don't care about thermal.

  Note: I am sending this pull on Rui's behalf while he fixes issues in
  his Linux box"

* 'fixes-for-4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/evalenti/linux-soc-thermal:
  thermal: int340x_thermal: Ignore missing _ART, _TRT tables
  thermal/intel_powerclamp: add id for Avoton SoC
  tools/thermal: tmon: silence 'set but not used' warnings
  tools/thermal: tmon: use pkg-config to determine library dependencies
  tools/thermal: tmon: support cross-compiling
  tools/thermal: tmon: add .gitignore
  tools/thermal: tmon: fixup tui windowing calculations
  tools/thermal: tmon: tui: don't hard-code dialog window size assumptions
  tools/thermal: tmon: add min/max macros
  tools/thermal: tmon: add --target-temp parameter
  thermal: exynos: Clean-up code to use oneline entry for exynos compatible table
  thermal: rcar: Make error and remove paths symmetrical with init
  thermal: rcar: Fix race condition between init and interrupt
  thermal: Introduce dummy functions when thermal is not defined
  ti-soc-thermal: Delete an unnecessary check before the function call "cpufreq_cooling_unregister"
  thermal: ti-soc-thermal: bandgap: Fix build warning if !CONFIG_PM_SLEEP

9 years agoMerge tag 'md/4.0-fixes' of git://neil.brown.name/md
Linus Torvalds [Mon, 2 Mar 2015 22:03:27 +0000 (14:03 -0800)]
Merge tag 'md/4.0-fixes' of git://neil.brown.name/md

Pull md fixes from Neil Brown:
 "Three md fixes:

   - fix a read-balance problem that was reported 2 years ago, but that
     I never noticed the report :-(

   - fix for rare RAID6 problem causing incorrect bitmap updates when
     two devices fail.

   - add __ATTR_PREALLOC annotation now that it is possible"

* tag 'md/4.0-fixes' of git://neil.brown.name/md:
  md: mark some attributes as pre-alloc
  raid5: check faulty flag for array status during recovery.
  md/raid1: fix read balance when a drive is write-mostly.

9 years agoMerge tag 'metag-fixes-v4.0-1' of git://git.kernel.org/pub/scm/linux/kernel/git/jhoga...
Linus Torvalds [Mon, 2 Mar 2015 22:02:17 +0000 (14:02 -0800)]
Merge tag 'metag-fixes-v4.0-1' of git://git./linux/kernel/git/jhogan/metag

Pull arch/metag fix from James Hogan:
 "This is just a single patch to fix the KSTK_EIP() and KSTK_ESP()
  macros for metag which have always been erronously returning the PC
  and stack pointer of the task's kernel context rather than from its
  user context saved at entry from userland into the kernel, which
  affects the contents of /proc/<pid>/maps and /proc/<pid>/stat"

* tag 'metag-fixes-v4.0-1' of git://git.kernel.org/pub/scm/linux/kernel/git/jhogan/metag:
  metag: Fix KSTK_EIP() and KSTK_ESP() macros

9 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 1 Mar 2015 20:22:44 +0000 (12:22 -0800)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Ingo Molnar:
 "A CR4-shadow 32-bit init fix, plus two typo fixes"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86: Init per-cpu shadow copy of CR4 on 32-bit CPUs too
  x86/platform/intel-mid: Fix trivial printk message typo in intel_mid_arch_setup()
  x86/cpu/intel: Fix trivial typo in intel_tlb_table[]

9 years agoMerge branch 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 1 Mar 2015 20:00:25 +0000 (12:00 -0800)]
Merge branch 'timers-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull timer fixes from Ingo Molnar:
 "Three clockevents/clocksource driver fixes"

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  clocksource: pxa: Fix section mismatch
  clocksource: mtk: Fix race conditions in probe code
  clockevents: asm9260: Fix compilation error with sparc/sparc64 allyesconfig

9 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 1 Mar 2015 19:56:13 +0000 (11:56 -0800)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf fixes from Ingo Molnar:
 "Two kprobes fixes and a handful of tooling fixes"

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf tools: Make sparc64 arch point to sparc
  perf symbols: Define EM_AARCH64 for older OSes
  perf top: Fix SIGBUS on sparc64
  perf tools: Fix probing for PERF_FLAG_FD_CLOEXEC flag
  perf tools: Fix pthread_attr_setaffinity_np build error
  perf tools: Define _GNU_SOURCE on pthread_attr_setaffinity_np feature check
  perf bench: Fix order of arguments to memcpy_alloc_mem
  kprobes/x86: Check for invalid ftrace location in __recover_probed_insn()
  kprobes/x86: Use 5-byte NOP when the code might be modified by ftrace

9 years agoMerge branch 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 1 Mar 2015 19:27:04 +0000 (11:27 -0800)]
Merge branch 'locking-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull locking fix from Ingo Molnar:
 "An rtmutex deadlock path fixlet"

* 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  locking/rtmutex: Set state back to running on error

9 years agoMerge tag 'perf-urgent-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git...
Ingo Molnar [Sun, 1 Mar 2015 16:41:42 +0000 (17:41 +0100)]
Merge tag 'perf-urgent-for-mingo' of git://git./linux/kernel/git/acme/linux into perf/urgent

Pull perf/urgent fixes from Arnaldo Carvalho de Melo:

  - pthread_attr_setaffinity_np() feature detection build fixes (Adrian Hunter, Josh Boyer)

  - Fix probing for PERF_FLAG_FD_CLOEXEC flag (Adrian Hunter)

  - Fix order of arguments to memcpy_alloc_mem in 'perf bench' (Bruce Merry)

  - Sparc64 and Aarch64 build and segfault fixes (David Ahern)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
9 years agolocking/rtmutex: Set state back to running on error
Sebastian Andrzej Siewior [Fri, 27 Feb 2015 16:57:09 +0000 (17:57 +0100)]
locking/rtmutex: Set state back to running on error

The "usual" path is:

 - rt_mutex_slowlock()
 - set_current_state()
 - task_blocks_on_rt_mutex() (ret 0)
 - __rt_mutex_slowlock()
   - sleep or not but do return with __set_current_state(TASK_RUNNING)
 - back to caller.

In the early error case where task_blocks_on_rt_mutex() return
-EDEADLK we never change the task's state back to RUNNING. I
assume this is intended. Without this change after ww_mutex
using rt_mutex the selftest passes but later I get plenty of:

  | bad: scheduling from the idle thread!

backtraces.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Mike Galbraith <umgwanakikbuti@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: afffc6c1805d ("locking/rtmutex: Optimize setting task running after being blocked")
Link: http://lkml.kernel.org/r/1425056229-22326-4-git-send-email-bigeasy@linutronix.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
9 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Sat, 28 Feb 2015 18:36:48 +0000 (10:36 -0800)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

Pull drm fixes from Dave Airlie:
 "Just general fixes: radeon, i915, atmel, tegra, amdkfd and one core
  fix"

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (28 commits)
  drm: atmel-hlcdc: remove clock polarity from crtc driver
  drm/radeon: only enable DP audio if the monitor supports it
  drm/radeon: fix atom aux payload size check for writes (v2)
  drm/radeon: fix 1 RB harvest config setup for TN/RL
  drm/radeon: enable SRBM timeout interrupt on EG/NI
  drm/radeon: enable SRBM timeout interrupt on SI
  drm/radeon: enable SRBM timeout interrupt on CIK v2
  drm/radeon: dump full IB if we hit a packet error
  drm/radeon: disable mclk switching with 120hz+ monitors
  drm/radeon: use drm_mode_vrefresh() rather than mode->vrefresh
  drm/radeon: enable native backlight control on old macs
  drm/i915: Fix frontbuffer false positve.
  drm/i915: Align initial plane backing objects correctly
  drm/i915: avoid processing spurious/shared interrupts in low-power states
  drm/i915: Check obj->vma_list under the struct_mutex
  drm/i915: Fix a use after free, and unbalanced refcounting
  drm: atmel-hlcdc: remove useless pm_runtime_put_sync in probe
  drm: atmel-hlcdc: reset layer A2Q and UPDATE bits when disabling it
  drm: Fix deadlock due to getconnector locking changes
  drm/i915: Dell Chromebook 11 has PWM backlight
  ...

9 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-block
Linus Torvalds [Sat, 28 Feb 2015 18:21:57 +0000 (10:21 -0800)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-block

Pull block layer fixes from Jens Axboe:
 "Two smaller fixes for this cycle:

   - A fixup from Keith so that NVMe compiles without BLK_INTEGRITY,
     basically just moving the code around appropriately.

   - A fixup for shm, fixing an oops in shmem_mapping() for mapping with
     no inode.  From Sasha"

[ The shmem fix doesn't look block-layer-related, but fixes a bug that
  happened due to the backing_dev_info removal..  - Linus ]

* 'for-linus' of git://git.kernel.dk/linux-block:
  mm: shmem: check for mapping owner before dereferencing
  NVMe: Fix for BLK_DEV_INTEGRITY not set

9 years agoMerge tag 'xfs-for-linus-4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sat, 28 Feb 2015 18:06:33 +0000 (10:06 -0800)]
Merge tag 'xfs-for-linus-4.0-rc2' of git://git./linux/kernel/git/dgc/linux-xfs

Pull xfs fixes from Dave Chinner:
 "These are fixes for regressions/bugs introduced in the 4.0 merge cycle
  and problems discovered during the merge window that need to be pushed
  back to stable kernels ASAP.

  This contains:
   - ensure quota type is reset in on-disk dquots
   - fix missing partial EOF block data flush on truncate extension
   - fix transaction leak in error handling for new pnfs block layout
     support
   - add missing target_ip check to RENAME_EXCHANGE"

* tag 'xfs-for-linus-4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs:
  xfs: cancel failed transaction in xfs_fs_commit_blocks()
  xfs: Ensure we have target_ip for RENAME_EXCHANGE
  xfs: ensure truncate forces zeroed blocks to disk
  xfs: Fix quota type in quota structures when reusing quota file

9 years agoMerge branch 'akpm' (patches from Andrew)
Linus Torvalds [Sat, 28 Feb 2015 17:58:03 +0000 (09:58 -0800)]
Merge branch 'akpm' (patches from Andrew)

Merge misc fixes from Andrew Morton:
 "13 fixes"

* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
  mm: add missing __PAGETABLE_{PUD,PMD}_FOLDED defines
  mm: page_alloc: revert inadvertent !__GFP_FS retry behavior change
  kernel/sys.c: fix UNAME26 for 4.0
  mm: memcontrol: use "max" instead of "infinity" in control knobs
  zram: use proper type to update max_used_pages
  drivers/rtc/rtc-ds1685.c: fix conditional in ds1685_rtc_sysfs_time_regs_{show,store}
  nilfs2: fix potential memory overrun on inode
  scripts/gdb: add empty package initialization script
  rtc: ds1685: remove superfluous checks for out-of-range u8 values
  rtc: ds1685: fix ds1685_rtc_alarm_irq_enable build error
  memcg: fix low limit calculation
  mm/nommu: fix memory leak
  ocfs2: update web page + git tree in documentation

9 years agomm: add missing __PAGETABLE_{PUD,PMD}_FOLDED defines
Kirill A. Shutemov [Fri, 27 Feb 2015 23:52:12 +0000 (15:52 -0800)]
mm: add missing __PAGETABLE_{PUD,PMD}_FOLDED defines

Core mm expects __PAGETABLE_{PUD,PMD}_FOLDED to be defined if these page
table levels folded.  Usually, these defines are provided by
<asm-generic/pgtable-nopmd.h> and <asm-generic/pgtable-nopud.h>.

But some architectures fold page table levels in a custom way.  They
need to define these macros themself.  This patch adds missing defines.

The patch fixes mm->nr_pmds underflow and eliminates dead __pmd_alloc()
and __pud_alloc() on architectures without these page table levels.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Aaro Koskinen <aaro.koskinen@iki.fi>
Cc: David Howells <dhowells@redhat.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Helge Deller <deller@gmx.de>
Cc: "James E.J. Bottomley" <jejb@parisc-linux.org>
Cc: Koichi Yasutake <yasutake.koichi@jp.panasonic.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agomm: page_alloc: revert inadvertent !__GFP_FS retry behavior change
Johannes Weiner [Fri, 27 Feb 2015 23:52:09 +0000 (15:52 -0800)]
mm: page_alloc: revert inadvertent !__GFP_FS retry behavior change

Historically, !__GFP_FS allocations were not allowed to invoke the OOM
killer once reclaim had failed, but nevertheless kept looping in the
allocator.

Commit 9879de7373fc ("mm: page_alloc: embed OOM killing naturally into
allocation slowpath"), which should have been a simple cleanup patch,
accidentally changed the behavior to aborting the allocation at that
point.  This creates problems with filesystem callers (?) that currently
rely on the allocator waiting for other tasks to intervene.

Revert the behavior as it shouldn't have been changed as part of a
cleanup patch.

Fixes: 9879de7373fc ("mm: page_alloc: embed OOM killing naturally into allocation slowpath")
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Michal Hocko <mhocko@suse.cz>
Reported-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Dave Chinner <david@fromorbit.com>
Acked-by: David Rientjes <rientjes@google.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: <stable@vger.kernel.org> [3.19.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agokernel/sys.c: fix UNAME26 for 4.0
Jon DeVree [Fri, 27 Feb 2015 23:52:07 +0000 (15:52 -0800)]
kernel/sys.c: fix UNAME26 for 4.0

There's a uname workaround for broken userspace which can't handle kernel
versions of 3.x.  Update it for 4.x.

Signed-off-by: Jon DeVree <nuxi@vault24.org>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agomm: memcontrol: use "max" instead of "infinity" in control knobs
Johannes Weiner [Fri, 27 Feb 2015 23:52:04 +0000 (15:52 -0800)]
mm: memcontrol: use "max" instead of "infinity" in control knobs

The memcg control knobs indicate the highest possible value using the
symbolic name "infinity", which is long and awkward to type.

Switch to the string "max", which is just as descriptive but shorter and
sweeter.

This changes a user interface, so do it before the release and before
the development flag is dropped from the default hierarchy.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agozram: use proper type to update max_used_pages
Joonsoo Kim [Fri, 27 Feb 2015 23:52:01 +0000 (15:52 -0800)]
zram: use proper type to update max_used_pages

max_used_pages is defined as atomic_long_t so we need to use unsigned
long to keep temporary value for it rather than int which is smaller
than unsigned long in a 64 bit system.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agodrivers/rtc/rtc-ds1685.c: fix conditional in ds1685_rtc_sysfs_time_regs_{show,store}
Joshua Kinard [Fri, 27 Feb 2015 23:51:59 +0000 (15:51 -0800)]
drivers/rtc/rtc-ds1685.c: fix conditional in ds1685_rtc_sysfs_time_regs_{show,store}

Fix a conditional statement checking for NULL in both
ds1685_rtc_sysfs_time_regs_show and ds1685_rtc_sysfs_time_regs_store
that was using a logical AND when it should be using a logical OR so
that we fail out of the function properly if the condition ever
evaluates to true.

Fixes: aaaf5fbf56f1 ("rtc: add driver for DS1685 family of real time clocks")
Signed-off-by: Joshua Kinard <kumba@gentoo.org>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agonilfs2: fix potential memory overrun on inode
Ryusuke Konishi [Fri, 27 Feb 2015 23:51:56 +0000 (15:51 -0800)]
nilfs2: fix potential memory overrun on inode

Each inode of nilfs2 stores a root node of a b-tree, and it turned out to
have a memory overrun issue:

Each b-tree node of nilfs2 stores a set of key-value pairs and the number
of them (in "bn_nchildren" member of nilfs_btree_node struct), as well as
a few other "bn_*" members.

Since the value of "bn_nchildren" is used for operations on the key-values
within the b-tree node, it can cause memory access overrun if a large
number is incorrectly set to "bn_nchildren".

For instance, nilfs_btree_node_lookup() function determines the range of
binary search with it, and too large "bn_nchildren" leads
nilfs_btree_node_get_key() in that function to overrun.

As for intermediate b-tree nodes, this is prevented by a sanity check
performed when each node is read from a drive, however, no sanity check
has been done for root nodes stored in inodes.

This patch fixes the issue by adding missing sanity check against b-tree
root nodes so that it's called when on-memory inodes are read from ifile,
inode metadata file.

Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agoscripts/gdb: add empty package initialization script
Jan Kiszka [Fri, 27 Feb 2015 23:51:53 +0000 (15:51 -0800)]
scripts/gdb: add empty package initialization script

This got lost during the initial merge process: Python requires an
__init__.py script, even if empty, in order to accept a directory as
package.  Add it, this time as a non-empty file.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agortc: ds1685: remove superfluous checks for out-of-range u8 values
Geert Uytterhoeven [Fri, 27 Feb 2015 23:51:51 +0000 (15:51 -0800)]
rtc: ds1685: remove superfluous checks for out-of-range u8 values

drivers/rtc/rtc-ds1685.c: In function `ds1685_rtc_read_alarm':
drivers/rtc/rtc-ds1685.c:402: warning: comparison is always true due to limited range of data type
drivers/rtc/rtc-ds1685.c:409: warning: comparison is always true due to limited range of data type
drivers/rtc/rtc-ds1685.c:416: warning: comparison is always true due to limited range of data type
drivers/rtc/rtc-ds1685.c: In function `ds1685_rtc_set_alarm':
drivers/rtc/rtc-ds1685.c:475: warning: comparison is always true due to limited range of data type
drivers/rtc/rtc-ds1685.c:478: warning: comparison is always true due to limited range of data type
drivers/rtc/rtc-ds1685.c:481: warning: comparison is always true due to limited range of data type

u8 cannot contain a value larger than 0xff, hence drop the checks.
Wrapping the checks in unlikely() indicated some sense of humor, though ;-)

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Joshua Kinard <kumba@gentoo.org>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agortc: ds1685: fix ds1685_rtc_alarm_irq_enable build error
Arnd Bergmann [Fri, 27 Feb 2015 23:51:48 +0000 (15:51 -0800)]
rtc: ds1685: fix ds1685_rtc_alarm_irq_enable build error

The newly added ds1685 driver causes a build error when enabled without
CONFIG_RTC_INTF_DEV:

  drivers/rtc/rtc-ds1685.c:919:22: error: 'ds1685_rtc_alarm_irq_enable' undeclared here (not in a function)
    .alarm_irq_enable = ds1685_rtc_alarm_irq_enable,

Apparently the driver was incorrectly changed to reflect the interface
change from 16380c153a69c ("RTC: Convert rtc drivers to use the
alarm_irq_enable method"), which removed the respective #ifdef from all
other rtc drivers.

This does the same change that was merged for the other drivers before and
removes the #ifdef, allowing the interrupts to be enabled through the
in-kernel rtc interface independent of the existence of /dev/rtc.

Fixes: aaaf5fbf56f ("rtc: add driver for DS1685 family of real time clocks")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Joshua Kinard <kumba@gentoo.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agomemcg: fix low limit calculation
Michal Hocko [Fri, 27 Feb 2015 23:51:46 +0000 (15:51 -0800)]
memcg: fix low limit calculation

A memcg is considered low limited even when the current usage is equal to
the low limit.  This leads to interesting side effects e.g.
groups/hierarchies with no memory accounted are considered protected and
so the reclaim will emit MEMCG_LOW event when encountering them.

Another and much bigger issue was reported by Joonsoo Kim.  He has hit a
NULL ptr dereference with the legacy cgroup API which even doesn't have
low limit exposed.  The limit is 0 by default but the initial check fails
for memcg with 0 consumption and parent_mem_cgroup() would return NULL if
use_hierarchy is 0 and so page_counter_read would try to dereference NULL.

I suppose that the current implementation is just an overlook because the
documentation in Documentation/cgroups/unified-hierarchy.txt says:

  "The memory.low boundary on the other hand is a top-down allocated
  reserve.  A cgroup enjoys reclaim protection when it and all its
  ancestors are below their low boundaries"

Fix the usage and the low limit comparision in mem_cgroup_low accordingly.

Fixes: 241994ed8649 (mm: memcontrol: default hierarchy interface for memory)
Reported-by: Joonsoo Kim <js1304@gmail.com>
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agomm/nommu: fix memory leak
Joonsoo Kim [Fri, 27 Feb 2015 23:51:43 +0000 (15:51 -0800)]
mm/nommu: fix memory leak

Maxime reported the following memory leak regression due to commit
dbc8358c7237 ("mm/nommu: use alloc_pages_exact() rather than its own
implementation").

On v3.19, I am facing a memory leak.  Each time I run a command one page
is lost.  Here an example with busybox's free command:

  / # free
               total       used       free     shared    buffers     cached
  Mem:          7928       1972       5956          0          0        492
  -/+ buffers/cache:       1480       6448
  / # free
               total       used       free     shared    buffers     cached
  Mem:          7928       1976       5952          0          0        492
  -/+ buffers/cache:       1484       6444
  / # free
               total       used       free     shared    buffers     cached
  Mem:          7928       1980       5948          0          0        492
  -/+ buffers/cache:       1488       6440
  / # free
               total       used       free     shared    buffers     cached
  Mem:          7928       1984       5944          0          0        492
  -/+ buffers/cache:       1492       6436
  / # free
               total       used       free     shared    buffers     cached
  Mem:          7928       1988       5940          0          0        492
  -/+ buffers/cache:       1496       6432

At some point, the system fails to sastisfy 256KB allocations:

  free: page allocation failure: order:6, mode:0xd0
  CPU: 0 PID: 67 Comm: free Not tainted 3.19.0-05389-gacf2cf1-dirty #64
  Hardware name: STM32 (Device Tree Support)
    show_stack+0xb/0xc
    warn_alloc_failed+0x97/0xbc
    __alloc_pages_nodemask+0x295/0x35c
    __get_free_pages+0xb/0x24
    alloc_pages_exact+0x19/0x24
    do_mmap_pgoff+0x423/0x658
    vm_mmap_pgoff+0x3f/0x4e
    load_flat_file+0x20d/0x4f8
    load_flat_binary+0x3f/0x26c
    search_binary_handler+0x51/0xe4
    do_execveat_common+0x271/0x35c
    do_execve+0x19/0x1c
    ret_fast_syscall+0x1/0x4a
  Mem-info:
  Normal per-cpu:
  CPU    0: hi:    0, btch:   1 usd:   0
  active_anon:0 inactive_anon:0 isolated_anon:0
   active_file:0 inactive_file:0 isolated_file:0
   unevictable:123 dirty:0 writeback:0 unstable:0
   free:1515 slab_reclaimable:17 slab_unreclaimable:139
   mapped:0 shmem:0 pagetables:0 bounce:0
   free_cma:0
  Normal free:6060kB min:352kB low:440kB high:528kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:492kB isolated(anon):0ks
  lowmem_reserve[]: 0 0
  Normal: 23*4kB (U) 22*8kB (U) 24*16kB (U) 23*32kB (U) 23*64kB (U) 23*128kB (U) 1*256kB (U) 0*512kB 0*1024kB 0*2048kB 0*4096kB = 6060kB
  123 total pagecache pages
  2048 pages of RAM
  1538 free pages
  66 reserved pages
  109 slab pages
  -46 pages shared
  0 pages swap cached
  nommu: Allocation of length 221184 from process 67 (free) failed
  Normal per-cpu:
  CPU    0: hi:    0, btch:   1 usd:   0
  active_anon:0 inactive_anon:0 isolated_anon:0
   active_file:0 inactive_file:0 isolated_file:0
   unevictable:123 dirty:0 writeback:0 unstable:0
   free:1515 slab_reclaimable:17 slab_unreclaimable:139
   mapped:0 shmem:0 pagetables:0 bounce:0
   free_cma:0
  Normal free:6060kB min:352kB low:440kB high:528kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:492kB isolated(anon):0ks
  lowmem_reserve[]: 0 0
  Normal: 23*4kB (U) 22*8kB (U) 24*16kB (U) 23*32kB (U) 23*64kB (U) 23*128kB (U) 1*256kB (U) 0*512kB 0*1024kB 0*2048kB 0*4096kB = 6060kB
  123 total pagecache pages
  Unable to allocate RAM for process text/data, errno 12 SEGV

This problem happens because we allocate ordered page through
__get_free_pages() in do_mmap_private() in some cases and we try to free
individual pages rather than ordered page in free_page_series().  In
this case, freeing pages whose refcount is not 0 won't be freed to the
page allocator so memory leak happens.

To fix the problem, this patch changes __get_free_pages() to
alloc_pages_exact() since alloc_pages_exact() returns
physically-contiguous pages but each pages are refcounted.

Fixes: dbc8358c7237 ("mm/nommu: use alloc_pages_exact() rather than its own implementation").
Reported-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Tested-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: <stable@vger.kernel.org> [3.19]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agoocfs2: update web page + git tree in documentation
Mark Fasheh [Fri, 27 Feb 2015 23:51:40 +0000 (15:51 -0800)]
ocfs2: update web page + git tree in documentation

We (the Ocfs2 project) recently moved the location of our ocfs2-tools
git tree and project web page.  The pertinent discussion can be seen
here:

  https://oss.oracle.com/pipermail/ocfs2-devel/2015-February/010579.html

The following patch updates the Ocfs2 documentation in MAINTAINERS,
ocfs2.txt, and dlmfs.txt.  I added our new official web page, changed
the location of our tools git tree and removed the link to Joel's
ancient kernel git tree - Andrew has handled our patches for a while
now.

Signed-off-by: Mark Fasheh <mfasheh@suse.de>
Cc: Joel Becker <jlbec@evilplan.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
9 years agox86: Init per-cpu shadow copy of CR4 on 32-bit CPUs too
Steven Rostedt [Fri, 27 Feb 2015 19:50:19 +0000 (14:50 -0500)]
x86: Init per-cpu shadow copy of CR4 on 32-bit CPUs too

Commit:

   1e02ce4cccdc ("x86: Store a per-cpu shadow copy of CR4")

added a shadow CR4 such that reads and writes that do not
modify the CR4 execute much faster than always reading the
register itself.

The change modified cpu_init() in common.c, so that the
shadow CR4 gets initialized before anything uses it.

Unfortunately, there's two cpu_init()s in common.c. There's
one for 64-bit and one for 32-bit. The commit only added
the shadow init to the 64-bit path, but the 32-bit path
needs the init too.

Link: http://lkml.kernel.org/r/20150227125208.71c36402@gandalf.local.home
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Andy Lutomirski <luto@amacapital.net>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20150227145019.2bdd4354@gandalf.local.home
Signed-off-by: Ingo Molnar <mingo@kernel.org>
9 years agoMerge branch 'linus' into x86/urgent, to merge dependent patch
Ingo Molnar [Sat, 28 Feb 2015 07:03:10 +0000 (08:03 +0100)]
Merge branch 'linus' into x86/urgent, to merge dependent patch

Signed-off-by: Ingo Molnar <mingo@kernel.org>
9 years agoMerge branch 'tmon-fixes' of .git into next
Zhang Rui [Sat, 28 Feb 2015 06:07:03 +0000 (14:07 +0800)]
Merge branch 'tmon-fixes' of .git into next

9 years agothermal: int340x_thermal: Ignore missing _ART, _TRT tables
Srinivas Pandruvada [Thu, 26 Feb 2015 02:31:50 +0000 (18:31 -0800)]
thermal: int340x_thermal: Ignore missing _ART, _TRT tables

It is possible that _ART/_TRT tables are missing or have errors.
Ignore those failures, as INT3400 thermal zone is still required
for _OSC or mode switch.

Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agothermal/intel_powerclamp: add id for Avoton SoC
Miguel Bernal Marin [Thu, 19 Feb 2015 18:40:58 +0000 (12:40 -0600)]
thermal/intel_powerclamp: add id for Avoton SoC

Enable Intel Powerclamp driver on Atom* Processor C2000 Product
Family for Microservers (Avoton). Avoton - SoCs for micro-servers
has package C-states which can be used for idle injection.

Reported-by: Jose Navarro <jose.navarro@intel.com>
Suggested-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Tested-by: Jose Carlos Venegas Munoz <jos.c.venegas.munoz@intel.com>
Signed-off-by: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: silence 'set but not used' warnings
Brian Norris [Wed, 18 Feb 2015 02:18:36 +0000 (18:18 -0800)]
tools/thermal: tmon: silence 'set but not used' warnings

gcc complains about the 'cols' variable being unused. This is
unavoidable, given the ncurses getmaxyx() macro-based API, which wants
to assign to a variable directly, even when we're not going to use it.

Warning:

    gcc -O1 -Wall -Wshadow -W -Wformat -Wimplicit-function-declaration -Wimplicit-int -fstack-protector -D VERSION=\"1.0\"   -c -o tui.o tui.c
    tui.c: In function ‘show_dialogue’:
    tui.c:288:12: warning: variable ‘cols’ set but not used [-Wunused-but-set-variable]
      int rows, cols;
                ^

So, add a hack to get rid of that warning.

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: use pkg-config to determine library dependencies
Brian Norris [Wed, 18 Feb 2015 02:18:35 +0000 (18:18 -0800)]
tools/thermal: tmon: use pkg-config to determine library dependencies

Some distros (e.g., Arch Linux) don't package the tinfo library
separately from ncurses, so don't unconditionally include it. Instead,
use pkg-config.

The $(STATIC) ugliness is to handle the reported build case from commit
6b533269fb25 ("tools/thermal: tmon: fix compilation errors when building
statically"), where a developer wants to be able to build with:

  make LDFLAGS=-static

which requires an additional pkg-config flag.

Finally, support a lowest common denominator fallback (-lpanel
-lncurses) for build systems that don't have pkg-config entries for
ncurses.

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: support cross-compiling
Brian Norris [Wed, 18 Feb 2015 02:18:34 +0000 (18:18 -0800)]
tools/thermal: tmon: support cross-compiling

We might want to prepare CFLAGS outside of this Makefile, so don't
overwrite its initial value.

Then, support $(CROSS_COMPILE), so we can use a cross-compile toolchain.

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: add .gitignore
Brian Norris [Wed, 18 Feb 2015 02:18:33 +0000 (18:18 -0800)]
tools/thermal: tmon: add .gitignore

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: fixup tui windowing calculations
Brian Norris [Wed, 18 Feb 2015 02:18:32 +0000 (18:18 -0800)]
tools/thermal: tmon: fixup tui windowing calculations

The number of rows in the dialog vary according to the number of cooling
devices. However, some of the windowing computations were assuming a
fixed number of rows. This computation is OK when we have between 4 and
9 cooling devices (and they wrap to the next column), but with fewer
devices, we end up printing off the end of the window.

This unifies the row computation into a single function and uses that
throughout the TUI code. This also accounts for increasing the number of
rows when there are more than 9 total cooling devices.

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: tui: don't hard-code dialog window size assumptions
Brian Norris [Wed, 18 Feb 2015 02:18:31 +0000 (18:18 -0800)]
tools/thermal: tmon: tui: don't hard-code dialog window size assumptions

We can use the ncurses API to get the number of rows.

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: add min/max macros
Brian Norris [Wed, 18 Feb 2015 02:18:30 +0000 (18:18 -0800)]
tools/thermal: tmon: add min/max macros

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agotools/thermal: tmon: add --target-temp parameter
Brian Norris [Wed, 18 Feb 2015 02:18:29 +0000 (18:18 -0800)]
tools/thermal: tmon: add --target-temp parameter

If we launch in daemon mode (--daemon), we don't have the ncurses UI,
but we might want to set the target temperature still. For example,
someone might stick the following in their boot script:

  tmon --control intel_powerclamp --target-temp 90 --log --daemon

This would turn on CPU idle injection when we're around 90 degrees
celsius, and would log temperature and throttling info to
/var/tmp/tmon.log.

Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
9 years agoMerge tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm...
Linus Torvalds [Sat, 28 Feb 2015 00:18:33 +0000 (16:18 -0800)]
Merge tag 'fixes-for-linus' of git://git./linux/kernel/git/arm/arm-soc

Pull ARM SoC fixes from Arnd Bergmann:
 "The arm-soc bug fixes this time around are mostly for the omap
  platform, coming from a pull request from Tony Lindgren and are almost
  entirely fixing dts files.

  The other two changes enable support for the shmobile platform in
  generic armv7 kernels and change some properties in the ARM64
  reference board dts files"

* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:
  ARM: multi_v7_defconfig: Enable shmobile platforms
  arm64: Add L2 cache topology to ARM Ltd boards/models
  ARM: dts: am335x-bone*: usb0 is hardwired for peripheral
  ARM: dts: dra7x-evm: beagle-x15: Fix USB Host
  ARM: omap2plus_defconfig: Fix SATA boot
  ARM: omap2plus_defconfig: Enable OMAP NAND BCH driver
  ARM: dts: dra7: Correct the dma controller's property names
  ARM: dts: omap5: Correct the dma controller's property names
  ARM: dts: omap4: Correct the dma controller's property names
  ARM: dts: omap3: Correct the dma controller's property names
  ARM: dts: omap2: Correct the dma controller's property names
  ARM: dts: am437x-idk: fix sleep pinctrl state
  ARM: omap2plus_defconfig: enable TPS62362 regulator
  ARM: dts: am437x-idk: fix TPS62362 i2c bus
  ARM: dts: n900: Fix offset for smc91x ethernet
  ARM: dts: n900: fix i2c bus numbering
  ARM: dts: Fix USB dts configuration for dm816x
  ARM: dts: OMAP5: Fix SATA PHY node
  ARM: dts: DRA7: Fix SATA PHY node

9 years agoMerge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Linus Torvalds [Sat, 28 Feb 2015 00:09:37 +0000 (16:09 -0800)]
Merge tag 'arm64-fixes' of git://git./linux/kernel/git/arm64/linux

Pull arm64 fixes from Catalin Marinas:
 "Various arm64 fixes:
   - ftrace branch generation fix
   - branch instruction encoding fix
   - include files, guards and unused prototypes clean-up
   - minor VDSO ABI fix (clock_getres)
   - PSCI functions moved to .S to avoid compilation error with gcc 5
   - pte_modify fix to not ignore the mapping type
   - crypto: AES interleaved increased to 4x (for performance reasons)
   - text patching fix for modules
   - swiotlb increased back to 64MB
   - copy_siginfo_to_user32() fix for big endian"

* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  arm64: cpuidle: add asm/proc-fns.h inclusion
  arm64: compat Fix siginfo_t -> compat_siginfo_t conversion on big endian
  arm64: Increase the swiotlb buffer size 64MB
  arm64: Fix text patching logic when using fixmap
  arm64: crypto: increase AES interleave to 4x
  arm64: enable PTE type bit in the mask for pte_modify
  arm64: mm: remove unused functions and variable protoypes
  arm64: psci: move psci firmware calls out of line
  arm64: vdso: minor ABI fix for clock_getres
  arm64: guard asm/assembler.h against multiple inclusions
  arm64: insn: fix compare-and-branch encodings
  arm64: ftrace: fix ftrace_modify_graph_caller for branch replace

9 years agoMerge tag 'renesas-sh-drivers-for-v4.0' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 28 Feb 2015 00:08:45 +0000 (16:08 -0800)]
Merge tag 'renesas-sh-drivers-for-v4.0' of git://git./linux/kernel/git/horms/renesas

Pull SH driver fix from Simon Horman:
 "Disable PM runtime for multi-platform r8a7740 with genpd"

* tag 'renesas-sh-drivers-for-v4.0' of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas:
  drivers: sh: Disable PM runtime for multi-platform r8a7740 with genpd

9 years agoarm64: cpuidle: add asm/proc-fns.h inclusion
Lorenzo Pieralisi [Fri, 27 Feb 2015 17:54:31 +0000 (17:54 +0000)]
arm64: cpuidle: add asm/proc-fns.h inclusion

ARM64 CPUidle driver requires the cpu_do_idle function so that it can
be used to enter the shallowest idle state, and it is declared in
asm/proc-fns.h.

The current ARM64 CPUidle driver does not include asm/proc-fns.h
explicitly and it has so far relied on implicit inclusion from other
header files.

Owing to some header dependencies reshuffling this currently triggers
build failures when CONFIG_ARM64_64K_PAGES=y:

drivers/cpuidle/cpuidle-arm64.c: In function "arm64_enter_idle_state"
drivers/cpuidle/cpuidle-arm64.c:42:3: error: implicit declaration of
function "cpu_do_idle" [-Werror=implicit-function-declaration]
   cpu_do_idle();
   ^

This patch adds the explicit inclusion of the asm/proc-fns.h header file
in the arm64 asm/cpuidle.h header file, so that the build breakage is fixed
and the required header inclusion is added to the appropriate arch back-end
CPUidle header, already included by the CPUidle arm64 driver, where
CPUidle arch related function declarations belong.

Reported-by: Laura Abbott <lauraa@codeaurora.org>
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Tested-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoarm64: compat Fix siginfo_t -> compat_siginfo_t conversion on big endian
Catalin Marinas [Mon, 23 Feb 2015 15:13:40 +0000 (15:13 +0000)]
arm64: compat Fix siginfo_t -> compat_siginfo_t conversion on big endian

The native (64-bit) sigval_t union contains sival_int (32-bit) and
sival_ptr (64-bit). When a compat application invokes a syscall that
takes a sigval_t value (as part of a larger structure, e.g.
compat_sys_mq_notify, compat_sys_timer_create), the compat_sigval_t
union is converted to the native sigval_t with sival_int overlapping
with either the least or the most significant half of sival_ptr,
depending on endianness. When the corresponding signal is delivered to a
compat application, on big endian the current (compat_uptr_t)sival_ptr
cast always returns 0 since sival_int corresponds to the top part of
sival_ptr. This patch fixes copy_siginfo_to_user32() so that sival_int
is copied to the compat_siginfo_t structure.

Cc: <stable@vger.kernel.org>
Reported-by: Bamvor Jian Zhang <bamvor.zhangjian@huawei.com>
Tested-by: Bamvor Jian Zhang <bamvor.zhangjian@huawei.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoarm64: Increase the swiotlb buffer size 64MB
Catalin Marinas [Thu, 5 Feb 2015 18:01:53 +0000 (18:01 +0000)]
arm64: Increase the swiotlb buffer size 64MB

With commit 3690951fc6d4 (arm64: Use swiotlb late initialisation), the
swiotlb buffer size is limited to MAX_ORDER_NR_PAGES. However, there are
platforms with 32-bit only devices that require bounce buffering via
swiotlb. This patch changes the swiotlb initialisation to an early 64MB
memblock allocation. In order to get the swiotlb buffer correctly
allocated (via memblock_virt_alloc_low_nopanic), this patch also defines
ARCH_LOW_ADDRESS_LIMIT to the maximum physical address capable of 32-bit
DMA.

Reported-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Tested-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoMerge branch 'drm-atmel-hlcdc-fixes' of git://github.com/bbrezillon/linux-at91 into...
Dave Airlie [Fri, 27 Feb 2015 00:31:40 +0000 (10:31 +1000)]
Merge branch 'drm-atmel-hlcdc-fixes' of git://github.com/bbrezillon/linux-at91 into drm-fixes

minor atmel hclcdc fixes.

* 'drm-atmel-hlcdc-fixes' of git://github.com/bbrezillon/linux-at91:
  drm: atmel-hlcdc: remove clock polarity from crtc driver
  drm: atmel-hlcdc: remove useless pm_runtime_put_sync in probe
  drm: atmel-hlcdc: reset layer A2Q and UPDATE bits when disabling it

9 years agoMerge tag 'drm-intel-fixes-2015-02-26' of git://anongit.freedesktop.org/drm-intel...
Dave Airlie [Fri, 27 Feb 2015 00:30:07 +0000 (10:30 +1000)]
Merge tag 'drm-intel-fixes-2015-02-26' of git://anongit.freedesktop.org/drm-intel into drm-fixes

First batch of fixes for v4.0-rc, plenty of cc: stable material.

* tag 'drm-intel-fixes-2015-02-26' of git://anongit.freedesktop.org/drm-intel:
  drm/i915: Fix frontbuffer false positve.
  drm/i915: Align initial plane backing objects correctly
  drm/i915: avoid processing spurious/shared interrupts in low-power states
  drm/i915: Check obj->vma_list under the struct_mutex
  drm/i915: Fix a use after free, and unbalanced refcounting
  drm/i915: Dell Chromebook 11 has PWM backlight
  drm/i915/skl: handle all pixel formats in skylake_update_primary_plane()
  drm/i915/bdw: PCI IDs ending in 0xb are ULT.

9 years agoMerge branch 'drm-fixes-4.0' of git://people.freedesktop.org/~agd5f/linux into drm...
Dave Airlie [Fri, 27 Feb 2015 00:29:14 +0000 (10:29 +1000)]
Merge branch 'drm-fixes-4.0' of git://people.freedesktop.org/~agd5f/linux into drm-fixes

misc radeon fixes.

* 'drm-fixes-4.0' of git://people.freedesktop.org/~agd5f/linux:
  drm/radeon: only enable DP audio if the monitor supports it
  drm/radeon: fix atom aux payload size check for writes (v2)
  drm/radeon: fix 1 RB harvest config setup for TN/RL
  drm/radeon: enable SRBM timeout interrupt on EG/NI
  drm/radeon: enable SRBM timeout interrupt on SI
  drm/radeon: enable SRBM timeout interrupt on CIK v2
  drm/radeon: dump full IB if we hit a packet error
  drm/radeon: disable mclk switching with 120hz+ monitors
  drm/radeon: use drm_mode_vrefresh() rather than mode->vrefresh
  drm/radeon: enable native backlight control on old macs

9 years agoMerge tag 'hwmon-for-linus-v4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 26 Feb 2015 23:23:02 +0000 (15:23 -0800)]
Merge tag 'hwmon-for-linus-v4.0-rc2' of git://git./linux/kernel/git/groeck/linux-staging

Pull hwmon fix from Guenter Roeck:
 "Add missing return value check to ads7828 driver"

* tag 'hwmon-for-linus-v4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (ads7828) Check return value of devm_regmap_init_i2c

9 years agoMerge tag 'sound-4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai...
Linus Torvalds [Thu, 26 Feb 2015 18:55:19 +0000 (10:55 -0800)]
Merge tag 'sound-4.0-rc2' of git://git./linux/kernel/git/tiwai/sound

Pull sound fixes from Takashi Iwai:
 "Most of changes in this pull request are about the fixes of crash of
  FireWire drivers at hot-unplugging.  In addition, there are a few
  HD-audio fixes (removal of wrong static, a pin quirk for an ASUS mobo,
  a regression fix for runtime PM on Panther Point) and a long-standing
  (but fairly minor) bug of PCM core"

* tag 'sound-4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: hda - Disable runtime PM for Panther Point again
  ALSA: hda: controller code - do not export static functions
  ALSA: pcm: Don't leave PREPARED state after draining
  ALSA: fireworks/bebob/dice/oxfw: make it possible to shutdown safely
  ALSA: fireworks/bebob/dice/oxfw: allow stream destructor after releasing runtime
  ALSA: firewire-lib: remove reference counting
  ALSA: fireworks/bebob/dice/oxfw: add reference-counting for FireWire unit
  ALSA: hda - Add pin configs for ASUS mobo with IDT 92HD73XX codec
  ALSA: firewire-lib: fix an unexpected byte sequence for micro sign

9 years agoarm64: Fix text patching logic when using fixmap
Marc Zyngier [Tue, 24 Feb 2015 16:30:21 +0000 (16:30 +0000)]
arm64: Fix text patching logic when using fixmap

Patch 2f896d586610 ("arm64: use fixmap for text patching") changed
the way we patch the kernel text, using a fixmap when the kernel or
modules are flagged as read only.

Unfortunately, a flaw in the logic makes it fall over when patching
modules without CONFIG_DEBUG_SET_MODULE_RONX enabled:

[...]
[   32.032636] Call trace:
[   32.032716] [<fffffe00003da0dc>] __copy_to_user+0x2c/0x60
[   32.032837] [<fffffe0000099f08>] __aarch64_insn_write+0x94/0xf8
[   32.033027] [<fffffe000009a0a0>] aarch64_insn_patch_text_nosync+0x18/0x58
[   32.033200] [<fffffe000009c3ec>] ftrace_modify_code+0x58/0x84
[   32.033363] [<fffffe000009c4e4>] ftrace_make_nop+0x3c/0x58
[   32.033532] [<fffffe0000164420>] ftrace_process_locs+0x3d0/0x5c8
[   32.033709] [<fffffe00001661cc>] ftrace_module_init+0x28/0x34
[   32.033882] [<fffffe0000135148>] load_module+0xbb8/0xfc4
[   32.034044] [<fffffe0000135714>] SyS_finit_module+0x94/0xc4
[...]

This is triggered by the use of virt_to_page() on a module address,
which ends to pointing to Nowhereland if you're lucky, or corrupt
your precious data if not.

This patch fixes the logic by mimicking what is done on arm:
- If we're patching a module and CONFIG_DEBUG_SET_MODULE_RONX is set,
  use vmalloc_to_page().
- If we're patching the kernel and CONFIG_DEBUG_RODATA is set,
  use virt_to_page().
- Otherwise, use the provided address, as we can write to it directly.

Tested on 4.0-rc1 as a KVM guest.

Reported-by: Richard W.M. Jones <rjones@redhat.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Laura Abbott <lauraa@codeaurora.org>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
Cc: Will Deacon <will.deacon@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux...
Linus Torvalds [Thu, 26 Feb 2015 18:34:24 +0000 (10:34 -0800)]
Merge branch 'for-linus' of git://git./linux/kernel/git/mason/linux-btrfs

Pull btrfs fix from Chris Mason:
 "I'm still testing more fixes, but I wanted to get out the fix for the
  btrfs raid5/6 memory corruption I mentioned in my merge window pull"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: fix allocation size calculations in alloc_btrfs_bio

9 years agoarm64: crypto: increase AES interleave to 4x
Ard Biesheuvel [Thu, 19 Feb 2015 17:25:16 +0000 (17:25 +0000)]
arm64: crypto: increase AES interleave to 4x

This patch increases the interleave factor for parallel AES modes
to 4x. This improves performance on Cortex-A57 by ~35%. This is
due to the 3-cycle latency of AES instructions on the A57's
relatively deep pipeline (compared to Cortex-A53 where the AES
instruction latency is only 2 cycles).

At the same time, disable inline expansion of the core AES functions,
as the performance benefit of this feature is negligible.

  Measured on AMD Seattle (using tcrypt.ko mode=500 sec=1):

  Baseline (2x interleave, inline expansion)
  ------------------------------------------
  testing speed of async cbc(aes) (cbc-aes-ce) decryption
  test 4 (128 bit key, 8192 byte blocks): 95545 operations in 1 seconds
  test 14 (256 bit key, 8192 byte blocks): 68496 operations in 1 seconds

  This patch (4x interleave, no inline expansion)
  -----------------------------------------------
  testing speed of async cbc(aes) (cbc-aes-ce) decryption
  test 4 (128 bit key, 8192 byte blocks): 124735 operations in 1 seconds
  test 14 (256 bit key, 8192 byte blocks): 92328 operations in 1 seconds

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoarm64: enable PTE type bit in the mask for pte_modify
Feng Kan [Tue, 24 Feb 2015 23:40:21 +0000 (15:40 -0800)]
arm64: enable PTE type bit in the mask for pte_modify

Caught during Trinity testing. The pte_modify does not allow
modification for PTE type bit. This cause the test to hang
the system. It is found that the PTE can't transit from an
inaccessible page (b00) to a valid page (b11) because the mask
does not allow it. This happens when a big block of mmaped
memory is set the PROT_NONE, then the a small piece is broken
off and set to PROT_WRITE | PROT_READ cause a huge page split.

Signed-off-by: Feng Kan <fkan@apm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoarm64: mm: remove unused functions and variable protoypes
Yingjoe Chen [Wed, 25 Feb 2015 02:47:45 +0000 (10:47 +0800)]
arm64: mm: remove unused functions and variable protoypes

The functions __cpu_flush_user_tlb_range and __cpu_flush_kern_tlb_range
were removed in commit fa48e6f780 'arm64: mm: Optimise tlb flush logic
where we have >4K granule'. Global variable cpu_tlb was never used in
arm64.

Remove them.

Signed-off-by: Yingjoe Chen <yingjoe.chen@mediatek.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoarm64: psci: move psci firmware calls out of line
Will Deacon [Wed, 25 Feb 2015 12:10:35 +0000 (12:10 +0000)]
arm64: psci: move psci firmware calls out of line

An arm64 allmodconfig fails to build with GCC 5 due to __asmeq
assertions in the PSCI firmware calling code firing due to mcount
preambles breaking our assumptions about register allocation of function
arguments:

  /tmp/ccDqJsJ6.s: Assembler messages:
  /tmp/ccDqJsJ6.s:60: Error: .err encountered
  /tmp/ccDqJsJ6.s:61: Error: .err encountered
  /tmp/ccDqJsJ6.s:62: Error: .err encountered
  /tmp/ccDqJsJ6.s:99: Error: .err encountered
  /tmp/ccDqJsJ6.s:100: Error: .err encountered
  /tmp/ccDqJsJ6.s:101: Error: .err encountered

This patch fixes the issue by moving the PSCI calls out-of-line into
their own assembly files, which are safe from the compiler's meddling
fingers.

Reported-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
9 years agoarm64: vdso: minor ABI fix for clock_getres
Nathan Lynch [Tue, 24 Feb 2015 23:21:07 +0000 (17:21 -0600)]
arm64: vdso: minor ABI fix for clock_getres

The vdso implementation of clock_getres currently returns 0 (success)
whenever a null timespec is provided by the caller, regardless of the
clock id supplied.

This behavior is incorrect.  It should fall back to syscall when an
unrecognized clock id is passed, even when the timespec argument is
null.  This ensures that clock_getres always returns an error for
invalid clock ids.

Signed-off-by: Nathan Lynch <nathan_lynch@mentor.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>