platform/adaptation/renesas_rcar/renesas_kernel.git
11 years agoxfs: remove two unused macro definitions in xfs_linux.h
Zhi Yong Wu [Wed, 7 Aug 2013 10:11:06 +0000 (10:11 +0000)]
xfs: remove two unused macro definitions in xfs_linux.h

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix the comment of xfs_btree_get_iroot()
Zhi Yong Wu [Wed, 7 Aug 2013 10:11:05 +0000 (10:11 +0000)]
xfs: fix the comment of xfs_btree_get_iroot()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix the comment of xfs_iroot_realloc()
Zhi Yong Wu [Wed, 7 Aug 2013 10:11:04 +0000 (10:11 +0000)]
xfs: fix the comment of xfs_iroot_realloc()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: remove one blank line in xfs_btree_make_block_unfull()
Zhi Yong Wu [Wed, 7 Aug 2013 10:11:03 +0000 (10:11 +0000)]
xfs: remove one blank line in xfs_btree_make_block_unfull()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix the comment of xlog_write_setup_copy()
Zhi Yong Wu [Wed, 7 Aug 2013 10:11:02 +0000 (10:11 +0000)]
xfs: fix the comment of xlog_write_setup_copy()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix the comment of xfs_mod_incore_sb_unlocked()
Zhi Yong Wu [Wed, 7 Aug 2013 10:11:01 +0000 (10:11 +0000)]
xfs: fix the comment of xfs_mod_incore_sb_unlocked()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix the comment of xfs_btree_lookup()
Zhi Yong Wu [Wed, 7 Aug 2013 10:11:00 +0000 (10:11 +0000)]
xfs: fix the comment of xfs_btree_lookup()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix the comment of xfs_buf_free()
Zhi Yong Wu [Wed, 7 Aug 2013 10:10:59 +0000 (10:10 +0000)]
xfs: fix the comment of xfs_buf_free()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix the comment of xfs_check_sizes()
Zhi Yong Wu [Wed, 7 Aug 2013 10:10:58 +0000 (10:10 +0000)]
xfs: fix the comment of xfs_check_sizes()

Signed-off-by: Zhi Yong Wu <wuzhy@linux.vnet.ibm.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: use reference counts to free clean buffer items
Dave Chinner [Mon, 12 Aug 2013 10:50:11 +0000 (20:50 +1000)]
xfs: use reference counts to free clean buffer items

When a transaction is cancelled and the buffer log item is clean in
the transaction, the buffer log item is unconditionally freed. If
the log item is in the AIL, however, this leads to a use after free
condition as the item still has other users.

In this case, xfs_buf_item_relse() should only be called on clean
buffer items if the reference count has dropped to zero. This
ensures only the last user frees the item.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoenable building user namespace with xfs
Dwight Engen [Thu, 15 Aug 2013 18:08:04 +0000 (14:08 -0400)]
enable building user namespace with xfs

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: add capability check to free eofblocks ioctl
Dwight Engen [Thu, 15 Aug 2013 18:08:03 +0000 (14:08 -0400)]
xfs: add capability check to free eofblocks ioctl

Check for CAP_SYS_ADMIN since the caller can truncate preallocated
blocks from files they do not own nor have write access to. A more
fine grained access check was considered: require the caller to
specify their own uid/gid and to use inode_permission to check for
write, but this would not catch the case of an inode not reachable
via path traversal from the callers mount namespace.

Add check for read-only filesystem to free eofblocks ioctl.

Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: create internal eofblocks structure with kuid_t types
Dwight Engen [Thu, 15 Aug 2013 18:08:02 +0000 (14:08 -0400)]
xfs: create internal eofblocks structure with kuid_t types

Have eofblocks ioctl convert uid_t to kuid_t into internal structure.
Update internal filter matching to compare ids with kuid_t types.

Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: convert kuid_t to/from uid_t for internal structures
Dwight Engen [Thu, 15 Aug 2013 18:08:01 +0000 (14:08 -0400)]
xfs: convert kuid_t to/from uid_t for internal structures

Use uint32 from init_user_ns for xfs internal uid/gid
representation in xfs_icdinode, xfs_dqid_t.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: ioctl check for capabilities in the current user namespace
Dwight Engen [Thu, 15 Aug 2013 18:08:00 +0000 (14:08 -0400)]
xfs: ioctl check for capabilities in the current user namespace

Use inode_capable() to check if SUID|SGID bits should be cleared to match
similar check in inode_change_ok().

The check for CAP_LINUX_IMMUTABLE was not modified since all other file
systems also check against init_user_ns rather than current_user_ns.

Only allow changing of projid from init_user_ns.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: convert kuid_t to/from uid_t in ACLs
Dwight Engen [Thu, 15 Aug 2013 18:07:59 +0000 (14:07 -0400)]
xfs: convert kuid_t to/from uid_t in ACLs

Change permission check for setting ACL to use inode_owner_or_capable()
which will additionally allow a CAP_FOWNER user in a user namespace to
be able to set an ACL on an inode covered by the user namespace mapping.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: create wrappers for converting kuid_t to/from uid_t
Dwight Engen [Thu, 15 Aug 2013 18:07:58 +0000 (14:07 -0400)]
xfs: create wrappers for converting kuid_t to/from uid_t

Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Gao feng <gaofeng@cn.fujitsu.com>
Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split the CIL lock
Dave Chinner [Mon, 12 Aug 2013 10:50:08 +0000 (20:50 +1000)]
xfs: split the CIL lock

The xc_cil_lock is used for two purposes - to protect the CIL
itself, and to protect the push/commit state and lists. These are
two logically separate structures and operations, so can have their
own locks. This means that pushing on the CIL and the commit wait
ordering won't contend for a lock with other transactions that are
completing concurrently. As the CIL insertion is the hottest path
throught eh CIL, this is a big win.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Combine CIL insert and prepare passes
Dave Chinner [Mon, 12 Aug 2013 10:50:07 +0000 (20:50 +1000)]
xfs: Combine CIL insert and prepare passes

Now that all the log item preparation and formatting is done under
the CIL lock, we can get rid of the intermediate log vector chain
used to track items to be inserted into the CIL.

We can already find all the items to be committed from the
transaction handle, so as long as we attach the log vectors to the
item before we insert the items into the CIL, we don't need to
create a log vector chain to pass around.

This means we can move all the item insertion code into and optimise
it into a pair of simple passes across all the items in the
transaction. The first pass does the formatting and accounting, the
second inserts them all into the CIL.

We keep this two pass split so that we can separate the CIL
insertion - which must be done under the CIL spinlock - from the
formatting. We could insert each item into the CIL with a single
pass, but that massively increases the number of times we have to
grab the CIL spinlock. It is much more efficient (and hence
scalable) to do a batch operation and insert all objects in a single
lock grab.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: avoid CIL allocation during insert
Dave Chinner [Mon, 12 Aug 2013 10:50:06 +0000 (20:50 +1000)]
xfs: avoid CIL allocation during insert

Now that we have the size of the log vector that has been allocated,
we can determine if we need to allocate a new log vector for
formatting and insertion. We only need to allocate a new vector if
it won't fit into the existing buffer.

However, we need to hold the CIL context lock while we do this so
that we can't race with a push draining the currently queued log
vectors. It is safe to do this as long as we do GFP_NOFS allocation
to avoid avoid memory allocation recursing into the filesystem.
Hence we can safely overwrite the existing log vector on the CIL if
it is large enough to hold all the dirty regions of the current
item.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Reduce allocations during CIL insertion
Dave Chinner [Mon, 12 Aug 2013 10:50:05 +0000 (20:50 +1000)]
xfs: Reduce allocations during CIL insertion

Now that we have the size of the object before the formatting pass
is called, we can allocation the log vector and it's buffer in a
single allocation rather than two separate allocations.

Store the size of the allocated buffer in the log vector so that
we potentially avoid allocation for future modifications of the
object.

While touching this code, remove the IOP_FORMAT definition.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: return log item size in IOP_SIZE
Dave Chinner [Mon, 12 Aug 2013 10:50:04 +0000 (20:50 +1000)]
xfs: return log item size in IOP_SIZE

To begin optimising the CIL commit process, we need to have IOP_SIZE
return both the number of vectors and the size of the data pointed
to by the vectors. This enables us to calculate the size ofthe
memory allocation needed before the formatting step and reduces the
number of memory allocations per item by one.

While there, kill the IOP_SIZE macro.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs:free bp in xlog_find_tail() error path
Eric Sandeen [Thu, 1 Aug 2013 01:33:47 +0000 (20:33 -0500)]
xfs:free bp in xlog_find_tail() error path

xlog_find_tail() currently leaks a bp on one error path.

There is no error target, so manually free the bp before
returning the error.

Found by Coverity.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: free bp in xlog_find_zeroed() error path
Eric Sandeen [Thu, 1 Aug 2013 01:32:30 +0000 (20:32 -0500)]
xfs: free bp in xlog_find_zeroed() error path

xlog_find_zeroed() currently leaks a bp on one error path.

Using the bp_err: target resolves this.

Found by Coverity.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: avoid double-free in xfs_attr_node_addname
Eric Sandeen [Thu, 1 Aug 2013 01:18:54 +0000 (20:18 -0500)]
xfs: avoid double-free in xfs_attr_node_addname

xfs_attr_node_addname()'s error handling tests whether it
should free "state" in the out: error handling label:

out:
        if (state)
                xfs_da_state_free(state);

but an earlier free doesn't set state to NULL afterwards; this
could lead to a double free.  Fix it by setting state to NULL
after it's freed.

This was found by Coverity.

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: call roundup_64() to calculate the min_logblks
Jie Liu [Tue, 13 Aug 2013 05:32:00 +0000 (13:32 +0800)]
xfs: call roundup_64() to calculate the min_logblks

Replace roundup() with roundup_64() as we calculate min_logblks
with 64-bit divisions.  Hence, call roundup() will cause the
following error while compiling a 32-bit kernel:

fs/built-in.o: In function `xfs_log_calc_minimum_size':
fs/xfs/xfs_log_rlimit.c:140: undefined reference to `__udivdi3'

Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Cc: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Validate log space at mount time
Jie Liu [Mon, 12 Aug 2013 10:50:03 +0000 (20:50 +1000)]
xfs: Validate log space at mount time

Validate log space during log mount stage, the underlying function
will drop a warning message via syslog in critical level if the log
space is too small or too large.

[ dchinner: For CRC enable filesystems, abort the mounting of the
filesystem as mkfs should never make a log too small for the given
filesystem configuration. ]

[ dchinner: make a note of the fact that the log size limits in
block counts are in units of filesystem blocks, not basic blocks. ]

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Add xfs_log_rlimit.c
Jie Liu [Mon, 12 Aug 2013 10:50:02 +0000 (20:50 +1000)]
xfs: Add xfs_log_rlimit.c

Add source files for xfs_log_rlimit.c The new file is used for log
size calculations and validation shared with userspace.

[dchinner: xfs_log_calc_max_attrsetm_res() does not modify the
tr_attrsetm reservation, just calculates the maximum. ]

[dchinner: rework loop in xfs_log_get_max_trans_res() ]

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Refactor xfs_ticket_alloc() to extract a new helper
Jie Liu [Mon, 12 Aug 2013 10:50:01 +0000 (20:50 +1000)]
xfs: Refactor xfs_ticket_alloc() to extract a new helper

Refactor xlog_ticket_alloc() to extract a new helper, i.e.
xfs_log_calc_unit_res().

This helper would be used to calculate the total log reservation
size by adding extra log operation/transation headers for a new
log ticket.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Get rid of all XFS_XXX_LOG_RES() macro
Jie Liu [Mon, 12 Aug 2013 10:50:00 +0000 (20:50 +1000)]
xfs: Get rid of all XFS_XXX_LOG_RES() macro

Get rid of all XFS_XXX_LOG_RES() macros since they are obsoleted now.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: refactor xfs_trans_reserve() interface
Jie Liu [Mon, 12 Aug 2013 10:49:59 +0000 (20:49 +1000)]
xfs: refactor xfs_trans_reserve() interface

With the new xfs_trans_res structure has been introduced, the log
reservation size, log count as well as log flags are pre-initialized
at mount time.  So it's time to refine xfs_trans_reserve() interface
to be more neat.

Also, introduce a new helper M_RES() to return a pointer to the
mp->m_resv structure to simplify the input.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Make writeid transaction use tr_writeid
Jie Liu [Mon, 12 Aug 2013 10:49:58 +0000 (20:49 +1000)]
xfs: Make writeid transaction use tr_writeid

tr_writeid is defined at mp->m_resv structure, however, it does not
really being used when it should be..

This patch changes it to tr_writeid to fetch the correct log
reservation size.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Introduce tr_fsyncts to m_reservation
Jie Liu [Mon, 12 Aug 2013 10:49:57 +0000 (20:49 +1000)]
xfs: Introduce tr_fsyncts to m_reservation

A preparation step.

For now fsync_ts transaction use the pre-calculated log reservation
size of tr_swrite.  This patch introduce a new item tr_fsyncts to
mp->m_reservations structure so that we can fetch the log
reservation value for it in a same manner to others.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Introduce a new structure to hold transaction reservation items
Jie Liu [Mon, 12 Aug 2013 10:49:56 +0000 (20:49 +1000)]
xfs: Introduce a new structure to hold transaction reservation items

Introduce a new structure xfs_trans_res to hold transaction
reservation item info per log ticket.

We also need to improve xfs_trans_resv_calc() by initializing the
log count as well as log flags for permanent log reservation.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: make struct xfs_perag kernel only
Dave Chinner [Mon, 12 Aug 2013 10:49:55 +0000 (20:49 +1000)]
xfs: make struct xfs_perag kernel only

The struct xfs_perag has many kernel-only definitions in it,
requiring a __KERNEL__ guard so userspace can use it to. Move it to
xfs_mount.h so that it it kernel-only, and let userspace redefine
it's own version of the structure containing only what it needs.
This gets rid of another __KERNEL__ check in the XFS header files.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: move kernel specific type definitions to xfs.h
Dave Chinner [Mon, 12 Aug 2013 10:49:54 +0000 (20:49 +1000)]
xfs: move kernel specific type definitions to xfs.h

xfs_types.h is shared with userspace, so having kernel specific
types defined in it is problematic. Move all the kernel specific
defines to xfs_linux.h so we can remove the __KERNEL__ guards from
xfs_types.h.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: xfs_filestreams.h doesn't need __KERNEL__
Dave Chinner [Mon, 12 Aug 2013 10:49:53 +0000 (20:49 +1000)]
xfs: xfs_filestreams.h doesn't need __KERNEL__

Because it is only used within the kernel.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: remove __KERNEL__ check from xfs_dir2_leaf.c
Dave Chinner [Mon, 12 Aug 2013 10:49:52 +0000 (20:49 +1000)]
xfs: remove __KERNEL__ check from xfs_dir2_leaf.c

It's actually an ifndef section, which means it is only included in
userspace. however, it's deep within the libxfs code, so it's
unlikely that the condition checked in userspace can actually occur
(search an empty leaf) through the libxfs interfaces. i.e. if it can
happen in usrspace, it can happen in the kernel, so remove it from
userspace too....

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: remove __KERNEL__ from debug code
Dave Chinner [Mon, 12 Aug 2013 10:49:51 +0000 (20:49 +1000)]
xfs: remove __KERNEL__ from debug code

There is no reason the remaining kernel-only debug code needs to
remain kernel-only. Kill the __KERNEL__ part of the defines, and let
userspace handle the debug code appropriately.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: kill __KERNEL__ check for debug code in allocation code
Dave Chinner [Mon, 12 Aug 2013 10:49:50 +0000 (20:49 +1000)]
xfs: kill __KERNEL__ check for debug code in allocation code

Userspace running debug builds is relatively rare, so there's need
to special case the allocation algorithm code coverage debug switch.
As it is, userspace defines random numbers to 0, so  invert the
logic of the switch so it is effectively a no-op in userspace.
This kills another couple of __KERNEL__ users.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: don't special case shared superblock mounts
Dave Chinner [Mon, 12 Aug 2013 10:49:49 +0000 (20:49 +1000)]
xfs: don't special case shared superblock mounts

Neither kernel or userspace support shared read-only mounts, so
don't bother special casing the support check to be different
between kernel and userspace. The same check can be used as neither
like it...

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: consolidate extent swap code
Dave Chinner [Mon, 12 Aug 2013 10:49:48 +0000 (20:49 +1000)]
xfs: consolidate extent swap code

So we don't need xfs_dfrag.h in userspace anymore, move the extent
swap ioctl structure definition to xfs_fs.h where most of the other
ioctl structure definitions are.

Now that we don't need separate files for extent swapping, separate
the basic file descriptor checking code to xfs_ioctl.c, and the code
that does the extent swap operation to xfs_bmap_util.c.  This
cleanly separates the user interface code from the physical
mechanism used to do the extent swap.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: consolidate xfs_utils.c
Dave Chinner [Mon, 12 Aug 2013 10:49:47 +0000 (20:49 +1000)]
xfs: consolidate xfs_utils.c

There are a few small helper functions in xfs_util, all related to
xfs_inode modifications. Move them all to xfs_inode.c so all
xfs_inode operations are consiolidated in the one place.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: consolidate xfs_rename.c
Dave Chinner [Mon, 12 Aug 2013 10:49:46 +0000 (20:49 +1000)]
xfs: consolidate xfs_rename.c

Move the rename code to xfs_inode.c to continue consolidating
all the kernel xfs_inode operations in the one place.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: kill xfs_vnodeops.[ch]
Dave Chinner [Mon, 12 Aug 2013 10:49:45 +0000 (20:49 +1000)]
xfs: kill xfs_vnodeops.[ch]

Now we have xfs_inode.c for holding kernel-only XFS inode
operations, move all the inode operations from xfs_vnodeops.c to
this new file as it holds another set of kernel-only inode
operations. The name of this file traces back to the days of Irix
and it's vnodes which we don't have anymore.

Essentially this move consolidates the inode locking functions
and a bunch of XFS inode operations into the one file. Eventually
the high level functions will be merged into the VFS interface
functions in xfs_iops.c.

This leaves only internal preallocation, EOF block manipulation and
hole punching functions in vnodeops.c. Move these to xfs_bmap_util.c
where we are already consolidating various in-kernel physical extent
manipulation and querying functions.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix issues that cause userspace warnings
Dave Chinner [Mon, 12 Aug 2013 10:49:44 +0000 (20:49 +1000)]
xfs: fix issues that cause userspace warnings

Some of the code shared with userspace causes compilation warnings
from things turned off in the kernel code, such as differences in
variable signedness. Fix those issues.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: minor cleanups
Dave Chinner [Mon, 12 Aug 2013 10:49:43 +0000 (20:49 +1000)]
xfs: minor cleanups

These come from syncing the shared userspace and kernel code. Small
whitespace and trivial cleanups.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: create xfs_bmap_util.[ch]
Dave Chinner [Mon, 12 Aug 2013 10:49:42 +0000 (20:49 +1000)]
xfs: create xfs_bmap_util.[ch]

There is a bunch of code in xfs_bmap.c that is kernel specific and
not shared with userspace. To minimise the difference between the
kernel and userspace code, shift this unshared code to
xfs_bmap_util.c, and the declarations to xfs_bmap_util.h.

The biggest issue here is xfs_bmap_finish() - userspace has it's own
definition of this function, and so we need to move it out of
xfs_bmap.[ch]. This means several other files need to include
xfs_bmap_util.h as well.

It also introduces and interesting dance for the stack switching
code in xfs_bmapi_allocate(). The stack switching/workqueue code is
actually moved to xfs_bmap_util.c, so that userspace can simply use
a #define in a header file to connect the dots without needing to
know about the stack switch code at all.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: introduce xfs_sb.c for sharing with libxfs
Dave Chinner [Mon, 12 Aug 2013 10:49:41 +0000 (20:49 +1000)]
xfs: introduce xfs_sb.c for sharing with libxfs

xfs_mount.c is shared with userspace, but the only functions that
are shared are to do with physical superblock manipulations. This
means that less than 25% of the xfs_mount.c code is actually shared
with userspace. Move all the superblock functions to xfs_sb.c and
share that instead with libxfs.

Note that this will leave all the in-core transaction related
superblock counter modifications in xfs_mount.c as none of that is
shared with userspace. With a few more small changes, xfs_mount.h
won't need to be shared with userspace anymore, either.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out the remote symlink handling
Dave Chinner [Mon, 12 Aug 2013 10:49:40 +0000 (20:49 +1000)]
xfs: split out the remote symlink handling

The remote symlink format definition and manipulation needs to be
shared with userspace, but the in-kernel interfaces do not. Split
the remote symlink format handling out into xfs_symlink_remote.[ch]
fo it can easily be shared with userspace.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out attribute fork truncation code into separate file
Dave Chinner [Mon, 12 Aug 2013 10:49:39 +0000 (20:49 +1000)]
xfs: split out attribute fork truncation code into separate file

The attribute inactivation code is not used by userspace, so like
the attribute listing, split it out into a separate file to minimise
the differences between the filesystem shared with libxfs in
userspace.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out attribute listing code into separate file
Dave Chinner [Mon, 12 Aug 2013 10:49:38 +0000 (20:49 +1000)]
xfs: split out attribute listing code into separate file

The attribute listing code is not used by userspace, so like the
directory readdir code, split it out into a separate file to
minimise the differences between the filesystem shared with libxfs
in userspace.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: reshuffle dir2 definitions around for userspace
Dave Chinner [Mon, 12 Aug 2013 10:49:37 +0000 (20:49 +1000)]
xfs: reshuffle dir2 definitions around for userspace

Many of the definitions within xfs_dir2_priv.h are needed in
userspace outside libxfs. Definitions within xfs_dir2_priv.h are
wholly contained within libxfs, so we need to shuffle some of the
definitions around to keep consistency across files shared between
user and kernel space.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: move getdents code into it's own file
Dave Chinner [Mon, 12 Aug 2013 10:49:36 +0000 (20:49 +1000)]
xfs: move getdents code into it's own file

The directory readdir code is not used by userspace, but it is
intermingled with files that are shared with userspace. This makes
it difficult to compare the differences between the userspac eand
kernel files are the userspace files don't have the getdents code in
them. Move all the kernel getdents code to a separate file to bring
the shared content between userspace and kernel files closer
together.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: introduce xfs_inode_buf.c for inode buffer operations
Dave Chinner [Mon, 12 Aug 2013 10:49:35 +0000 (20:49 +1000)]
xfs: introduce xfs_inode_buf.c for inode buffer operations

The only thing remaining in xfs_inode.[ch] are the operations that
read, write or verify physical inodes in their underlying buffers.
Move all this code to xfs_inode_buf.[ch] and so we can stop sharing
xfs_inode.[ch] with userspace.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: move unrelated definitions out of xfs_inode.h
Dave Chinner [Mon, 12 Aug 2013 10:49:34 +0000 (20:49 +1000)]
xfs: move unrelated definitions out of xfs_inode.h

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: move inode fork definitions to a new header file
Dave Chinner [Mon, 12 Aug 2013 10:49:33 +0000 (20:49 +1000)]
xfs: move inode fork definitions to a new header file

The inode fork definitions are a combination of on-disk format
definition and in-memory tracking and manipulation. They are both
shared with userspace, so move them all into their own file so
sharing is easy to do and track.  This removes all inode fork
related information from xfs_inode.h.

Do the same for the all the C code that currently resides in
xfs_inode.c for the same reason.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out transaction reservation code
Dave Chinner [Mon, 12 Aug 2013 10:49:32 +0000 (20:49 +1000)]
xfs: split out transaction reservation code

The transaction reservation size calculations is used by both kernel
and userspace, but most of the transaction code in xfs_trans.c is
kernel specific. Split all the transaction reservation code out into
it's own files to make sharing with userspace simpler. This just
leaves kernel-only definitions in xfs_trans.h, so it doesn't need to
be shared with userspace anymore, either.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: sync minor header differences needed by userspace.
Dave Chinner [Mon, 12 Aug 2013 10:49:31 +0000 (20:49 +1000)]
xfs: sync minor header differences needed by userspace.

Little things like exported functions, __KERNEL__ protections, and
so on that ensure user and kernel shared headers are identical.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: introduce xfs_quota_defs.h
Dave Chinner [Mon, 12 Aug 2013 10:49:30 +0000 (20:49 +1000)]
xfs: introduce xfs_quota_defs.h

There are a lot of quota flag definitions that are shared by user
and kernel space. Move them all to xfs_quota_defs.h so we can
unshare xfs_quota.h and remove the __KERNEL__ regions from it.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: introduce xfs_rtalloc_defs.h
Dave Chinner [Mon, 12 Aug 2013 10:49:29 +0000 (20:49 +1000)]
xfs: introduce xfs_rtalloc_defs.h

There are quite a few realtime device definitions shared with
userspace. Move them from xfs_rtalloc.h to xfs_rt_alloc_defs.h
so we don't need to share xfs_rtalloc.h with userspace anymore.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out on-disk transaction definitions
Dave Chinner [Mon, 12 Aug 2013 10:49:28 +0000 (20:49 +1000)]
xfs: split out on-disk transaction definitions

There's a bunch of definitions in xfs_trans.h that define on-disk
formats - transaction headers that get written into the log, log
item type definitions, etc. Split out everything into a separate
file so that all which remains in xfs_trans.h are kernel only
definitions.

Also, remove the duplicate magic number definitions for
XFS_TRANS_MAGIC...

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: separate icreate log format definitions from xfs_icreate_item.h
Dave Chinner [Mon, 12 Aug 2013 10:49:27 +0000 (20:49 +1000)]
xfs: separate icreate log format definitions from xfs_icreate_item.h

The on disk log format definitions for the icreate log item are
intertwined with the kernel-only in-memory log item definitions.
Separate the log format definitions out into their own header file
so they can easily be shared with userspace.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: separate dquot on disk format definitions out of xfs_quota.h
Dave Chinner [Mon, 12 Aug 2013 10:49:26 +0000 (20:49 +1000)]
xfs: separate dquot on disk format definitions out of xfs_quota.h

The on disk format definitions of the on-disk dquot, log formats and
quota off log formats are all intertwined with other definitions for
quotas. Separate them out into their own header file so they can
easily be shared with userspace.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out EFI/EFD log item format definition
Dave Chinner [Mon, 12 Aug 2013 10:49:25 +0000 (20:49 +1000)]
xfs: split out EFI/EFD log item format definition

The EFI/EFD item format definitions are shared with userspace. Split
the out of header files that contain kernel only defintions to make
it simple to shared them.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out buf log item format definitions
Dave Chinner [Mon, 12 Aug 2013 10:49:24 +0000 (20:49 +1000)]
xfs: split out buf log item format definitions

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: split out inode log item format definition
Dave Chinner [Mon, 12 Aug 2013 10:49:23 +0000 (20:49 +1000)]
xfs: split out inode log item format definition

The log item format definitions are shared with userspace. Split
them out of header files that contain kernel only defintions to make
it simple to shared them.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: separate out log format definitions
Dave Chinner [Mon, 12 Aug 2013 10:49:22 +0000 (20:49 +1000)]
xfs: separate out log format definitions

The on-disk format definitions for the log are spread randoms
through a couple of header files. Consolidate it all in a single
file that can be shared easily with userspace. This means that
xfs_log.h and xfs_log_priv.h no longer need to be shared with
userspace.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: WQ_NON_REENTRANT is meaningless and going away
Tejun Heo [Tue, 30 Jul 2013 12:40:26 +0000 (08:40 -0400)]
xfs: WQ_NON_REENTRANT is meaningless and going away

dbf2576e37 ("workqueue: make all workqueues non-reentrant") made
WQ_NON_REENTRANT no-op and the flag is going away.  Remove its usages.

This patch doesn't introduce any behavior changes.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Ben Myers <bpm@sgi.com>
Cc: Alex Elder <elder@kernel.org>
Cc: xfs@oss.sgi.com
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: di_flushiter considered harmful
Dave Chinner [Wed, 24 Jul 2013 05:47:30 +0000 (15:47 +1000)]
xfs: di_flushiter considered harmful

When we made all inode updates transactional, we no longer needed
the log recovery detection for inodes being newer on disk than the
transaction being replayed - it was redundant as replay of the log
would always result in the latest version of the inode would be on
disk. It was redundant, but left in place because it wasn't
considered to be a problem.

However, with the new "don't read inodes on create" optimisation,
flushiter has come back to bite us. Essentially, the optimisation
made always initialises flushiter to zero in the create transaction,
and so if we then crash and run recovery and the inode already on
disk has a non-zero flushiter it will skip recovery of that inode.
As a result, log recovery does the wrong thing and we end up with a
corrupt filesystem.

Because we have to support old kernel to new kernel upgrades, we
can't just get rid of the flushiter support in log recovery as we
might be upgrading from a kernel that doesn't have fully transactional
inode updates.  Unfortunately, for v4 superblocks there is no way to
guarantee that log recovery knows about this fact.

We cannot add a new inode format flag to say it's a "special inode
create" because it won't be understood by older kernels and so
recovery could do the wrong thing on downgrade. We cannot specially
detect the combination of zero mode/non-zero flushiter on disk to
non-zero mode, zero flushiter in the log item during recovery
because wrapping of the flushiter can result in false detection.

Hence that makes this "don't use flushiter" optimisation limited to
a disk format that guarantees that we don't need it. And that means
the only fix here is to limit the "no read IO on create"
optimisation to version 5 superblocks....

Reported-by: Markus Trippelsdorf <markus@trippelsdorf.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Start using pquotaino from the superblock.
Chandra Seetharaman [Fri, 19 Jul 2013 22:36:02 +0000 (17:36 -0500)]
xfs: Start using pquotaino from the superblock.

Start using pquotino and define a macro to check if the
superblock has pquotino.

Keep backward compatibilty by alowing mount of older superblock
with no separate pquota inode.

Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Initialize all quota inodes to be NULLFSINO
Chandra Seetharaman [Fri, 19 Jul 2013 22:32:55 +0000 (17:32 -0500)]
xfs: Initialize all quota inodes to be NULLFSINO

mkfs doesn't initialize the quota inodes to NULLFSINO as it does for the
other internal inodes. This leads to two in-core values (0 and NULLFSINO)
to be checked against, to make sure if a quota inode is valid.

Solve that problem by initializing the in-core values of all quotaino
values to NULLFSINO if they are 0 in the disk.

Note that these values are not written back to on-disk superblock unless
some quota is enabled on the filesystem. Even in that case sb_pquotino is
written to disk only if the on-disk superblock supports pquotino

Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: Fix a deadlock in xfs_log_commit_cil() code path
Chandra Seetharaman [Fri, 19 Jul 2013 22:31:38 +0000 (17:31 -0500)]
xfs: Fix a deadlock in xfs_log_commit_cil() code path

While testing and rearranging pquota/gquota code, I stumbled
on a xfs_shutdown() during a mount. But the mount just hung.

Debugged and found that there is a deadlock involving
&log->l_cilp->xc_ctx_lock.

It is in a code path where &log->l_cilp->xc_ctx_lock is first
acquired in read mode and some levels down the same semaphore
is being acquired in write mode causing a deadlock.

This is the stack:
xfs_log_commit_cil -> acquires &log->l_cilp->xc_ctx_lock in read mode
  xlog_print_tic_res
    xfs_force_shutdown
      xfs_log_force_umount
        xlog_cil_force
          xlog_cil_force_lsn
            xlog_cil_push_foreground
              xlog_cil_push - tries to acquire same semaphore in write mode

This patch fixes the deadlock by changing the reason code for
xfs_force_shutdown in xlog_print_tic_res() to SHUTDOWN_LOG_IO_ERROR.

SHUTDOWN_LOG_IO_ERROR is the right reason code to be set since
we are in the log path.

Thanks to Dave for suggesting this solution.

Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoxfs: fix assertion failure in xfs_vm_write_failed()
Jie Liu [Tue, 16 Jul 2013 05:11:16 +0000 (13:11 +0800)]
xfs: fix assertion failure in xfs_vm_write_failed()

In xfs_vm_write_failed(), we evaluate the block_offset of pos with
PAGE_MASK which is an unsigned long.  That is fine on 64-bit platforms
regardless of whether the request pos is 32-bit or 64-bit.  However, on
32-bit platforms the value is 0xfffff000 and so the high 32 bits in it
will be masked off with (pos & PAGE_MASK) for a 64-bit pos.

As a result, the evaluated block_offset is incorrect which will cause
this failure ASSERT(block_offset + from == pos); and potentially pass
the wrong block to xfs_vm_kill_delalloc_range().

In this case, we can get a kernel panic if CONFIG_XFS_DEBUG is enabled:

XFS: Assertion failed: block_offset + from == pos, file: fs/xfs/xfs_aops.c, line: 1504

------------[ cut here ]------------
 kernel BUG at fs/xfs/xfs_message.c:100!
 invalid opcode: 0000 [#1] SMP
 ........
 Pid: 4057, comm: mkfs.xfs Tainted: G           O 3.9.0-rc2 #1
 EIP: 0060:[<f94a7e8b>] EFLAGS: 00010282 CPU: 0
 EIP is at assfail+0x2b/0x30 [xfs]
 EAX: 00000056 EBX: f6ef28a0 ECX: 00000007 EDX: f57d22a4
 ESI: 1c2fb000 EDI: 00000000 EBP: ea6b5d30 ESP: ea6b5d1c
 DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
 CR0: 8005003b CR2: 094f3ff4 CR3: 2bcb4000 CR4: 000006f0
 DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000
 DR6: ffff0ff0 DR7: 00000400
 Process mkfs.xfs (pid: 4057, ti=ea6b4000 task=ea5799e0 task.ti=ea6b4000)
 Stack:
 00000000 f9525c48 f951fa80 f951f96b 000005e4 ea6b5d7c f9494b34 c19b0ea2
 00000066 f3d6c620 c19b0ea2 00000000 e9a91458 00001000 00000000 00000000
 00000000 c15c7e89 00000000 1c2fb000 00000000 00000000 1c2fb000 00000080
 Call Trace:
 [<f9494b34>] xfs_vm_write_failed+0x74/0x1b0 [xfs]
 [<c15c7e89>] ? printk+0x4d/0x4f
 [<f9494d7d>] xfs_vm_write_begin+0x10d/0x170 [xfs]
 [<c110a34c>] generic_file_buffered_write+0xdc/0x210
 [<f949b669>] xfs_file_buffered_aio_write+0xf9/0x190 [xfs]
 [<f949b7f3>] xfs_file_aio_write+0xf3/0x160 [xfs]
 [<c115e504>] do_sync_write+0x94/0xd0
 [<c115ed1f>] vfs_write+0x8f/0x160
 [<c115e470>] ? wait_on_retry_sync_kiocb+0x50/0x50
 [<c115f017>] sys_write+0x47/0x80
 [<c15d860d>] sysenter_do_call+0x12/0x28
 .............
 EIP: [<f94a7e8b>] assfail+0x2b/0x30 [xfs] SS:ESP 0068:ea6b5d1c
 ---[ end trace cdd9af4f4ecab42f ]---
 Kernel panic - not syncing: Fatal exception

In order to avoid this, we can evaluate the block_offset of the start
of the page by using shifts rather than masks the mismatch problem.

Thanks Dave Chinner for help finding and fixing this bug.

Reported-by: Michael L. Semon <mlsemon35@gmail.com>
Reviewed-by: Dave Chinner <david@fromorbit.com>
Reviewed-by: Mark Tinguely <tinguely@sgi.com>
Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
11 years agoLinux 3.11-rc1
Linus Torvalds [Sun, 14 Jul 2013 22:18:27 +0000 (15:18 -0700)]
Linux 3.11-rc1

11 years agoMerge branch 'slab/for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/penber...
Linus Torvalds [Sun, 14 Jul 2013 22:14:29 +0000 (15:14 -0700)]
Merge branch 'slab/for-linus' of git://git./linux/kernel/git/penberg/linux

Pull slab update from Pekka Enberg:
 "Highlights:

  - Fix for boot-time problems on some architectures due to
    init_lock_keys() not respecting kmalloc_caches boundaries
    (Christoph Lameter)

  - CONFIG_SLUB_CPU_PARTIAL requested by RT folks (Joonsoo Kim)

  - Fix for excessive slab freelist draining (Wanpeng Li)

  - SLUB and SLOB cleanups and fixes (various people)"

I ended up editing the branch, and this avoids two commits at the end
that were immediately reverted, and I instead just applied the oneliner
fix in between myself.

* 'slab/for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/linux
  slub: Check for page NULL before doing the node_match check
  mm/slab: Give s_next and s_stop slab-specific names
  slob: Check for NULL pointer before calling ctor()
  slub: Make cpu partial slab support configurable
  slab: add kmalloc() to kernel API documentation
  slab: fix init_lock_keys
  slob: use DIV_ROUND_UP where possible
  slub: do not put a slab to cpu partial list when cpu_partial is 0
  mm/slub: Use node_nr_slabs and node_nr_objs in get_slabinfo
  mm/slub: Drop unnecessary nr_partials
  mm/slab: Fix /proc/slabinfo unwriteable for slab
  mm/slab: Sharing s_next and s_stop between slab and slub
  mm/slab: Fix drain freelist excessively
  slob: Rework #ifdeffery in slab.h
  mm, slab: moved kmem_cache_alloc_node comment to correct place

11 years agoslub: Check for page NULL before doing the node_match check
Steven Rostedt [Thu, 17 Jan 2013 18:10:58 +0000 (13:10 -0500)]
slub: Check for page NULL before doing the node_match check

In the -rt kernel (mrg), we hit the following dump:

BUG: unable to handle kernel NULL pointer dereference at           (null)
IP: [<ffffffff811573f1>] kmem_cache_alloc_node+0x51/0x180
PGD a2d39067 PUD b1641067 PMD 0
Oops: 0000 [#1] PREEMPT SMP
Modules linked in: sunrpc cpufreq_ondemand ipv6 tg3 joydev sg serio_raw pcspkr k8temp amd64_edac_mod edac_core i2c_piix4 e100 mii shpchp ext4 mbcache jbd2 sd_mod crc_t10dif sr_mod cdrom sata_svw ata_generic pata_acpi pata_serverworks radeon ttm drm_kms_helper drm hwmon i2c_algo_bit i2c_core dm_mirror dm_region_hash dm_log dm_mod
CPU 3
Pid: 20878, comm: hackbench Not tainted 3.6.11-rt25.14.el6rt.x86_64 #1 empty empty/Tyan Transport GT24-B3992
RIP: 0010:[<ffffffff811573f1>]  [<ffffffff811573f1>] kmem_cache_alloc_node+0x51/0x180
RSP: 0018:ffff8800a9b17d70  EFLAGS: 00010213
RAX: 0000000000000000 RBX: 0000000001200011 RCX: ffff8800a06d8000
RDX: 0000000004d92a03 RSI: 00000000000000d0 RDI: ffff88013b805500
RBP: ffff8800a9b17dc0 R08: ffff88023fd14d10 R09: ffffffff81041cbd
R10: 00007f4e3f06e9d0 R11: 0000000000000246 R12: ffff88013b805500
R13: ffff8801ff46af40 R14: 0000000000000001 R15: 0000000000000000
FS:  00007f4e3f06e700(0000) GS:ffff88023fd00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000000 CR3: 00000000a2d3a000 CR4: 00000000000007e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process hackbench (pid: 20878, threadinfo ffff8800a9b16000, task ffff8800a06d8000)
Stack:
 ffff8800a9b17da0 ffffffff81202e08 ffff8800a9b17de0 000000d001200011
 0000000001200011 0000000001200011 0000000000000000 0000000000000000
 00007f4e3f06e9d0 0000000000000000 ffff8800a9b17e60 ffffffff81041cbd
Call Trace:
 [<ffffffff81202e08>] ? current_has_perm+0x68/0x80
 [<ffffffff81041cbd>] copy_process+0xdd/0x15b0
 [<ffffffff810a2125>] ? rt_up_read+0x25/0x30
 [<ffffffff8104369a>] do_fork+0x5a/0x360
 [<ffffffff8107c66b>] ? migrate_enable+0xeb/0x220
 [<ffffffff8100b068>] sys_clone+0x28/0x30
 [<ffffffff81527423>] stub_clone+0x13/0x20
 [<ffffffff81527152>] ? system_call_fastpath+0x16/0x1b
Code: 89 fc 89 75 cc 41 89 d6 4d 8b 04 24 65 4c 03 04 25 48 ae 00 00 49 8b 50 08 4d 8b 28 49 8b 40 10 4d 85 ed 74 12 41 83 fe ff 74 27 <48> 8b 00 48 c1 e8 3a 41 39 c6 74 1b 8b 75 cc 4c 89 c9 44 89 f2
RIP  [<ffffffff811573f1>] kmem_cache_alloc_node+0x51/0x180
 RSP <ffff8800a9b17d70>
CR2: 0000000000000000
---[ end trace 0000000000000002 ]---

Now, this uses SLUB pretty much unmodified, but as it is the -rt kernel
with CONFIG_PREEMPT_RT set, spinlocks are mutexes, although they do
disable migration. But the SLUB code is relatively lockless, and the
spin_locks there are raw_spin_locks (not converted to mutexes), thus I
believe this bug can happen in mainline without -rt features. The -rt
patch is just good at triggering mainline bugs ;-)

Anyway, looking at where this crashed, it seems that the page variable
can be NULL when passed to the node_match() function (which does not
check if it is NULL). When this happens we get the above panic.

As page is only used in slab_alloc() to check if the node matches, if
it's NULL I'm assuming that we can say it doesn't and call the
__slab_alloc() code. Is this a correct assumption?

Acked-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
11 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Sun, 14 Jul 2013 18:42:26 +0000 (11:42 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/viro/vfs

Pull more vfs stuff from Al Viro:
 "O_TMPFILE ABI changes, Oleg's fput() series, misc cleanups, including
  making simple_lookup() usable for filesystems with non-NULL s_d_op,
  which allows us to get rid of quite a bit of ugliness"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  sunrpc: now we can just set ->s_d_op
  cgroup: we can use simple_lookup() now
  efivarfs: we can use simple_lookup() now
  make simple_lookup() usable for filesystems that set ->s_d_op
  configfs: don't open-code d_alloc_name()
  __rpc_lookup_create_exclusive: pass string instead of qstr
  rpc_create_*_dir: don't bother with qstr
  llist: llist_add() can use llist_add_batch()
  llist: fix/simplify llist_add() and llist_add_batch()
  fput: turn "list_head delayed_fput_list" into llist_head
  fs/file_table.c:fput(): add comment
  Safer ABI for O_TMPFILE

11 years agosunrpc: now we can just set ->s_d_op
Al Viro [Sun, 14 Jul 2013 13:55:39 +0000 (17:55 +0400)]
sunrpc: now we can just set ->s_d_op

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
11 years agocgroup: we can use simple_lookup() now
Al Viro [Sun, 14 Jul 2013 13:50:23 +0000 (17:50 +0400)]
cgroup: we can use simple_lookup() now

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
11 years agoefivarfs: we can use simple_lookup() now
Al Viro [Sun, 14 Jul 2013 13:48:35 +0000 (17:48 +0400)]
efivarfs: we can use simple_lookup() now

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
11 years agomake simple_lookup() usable for filesystems that set ->s_d_op
Al Viro [Sun, 14 Jul 2013 13:43:25 +0000 (17:43 +0400)]
make simple_lookup() usable for filesystems that set ->s_d_op

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
11 years agoconfigfs: don't open-code d_alloc_name()
Al Viro [Sun, 14 Jul 2013 13:16:52 +0000 (17:16 +0400)]
configfs: don't open-code d_alloc_name()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
11 years ago__rpc_lookup_create_exclusive: pass string instead of qstr
Al Viro [Sun, 14 Jul 2013 13:09:57 +0000 (17:09 +0400)]
__rpc_lookup_create_exclusive: pass string instead of qstr

... and use d_hash_and_lookup() instead of open-coding it, for fsck sake...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
11 years agorpc_create_*_dir: don't bother with qstr
Al Viro [Sun, 14 Jul 2013 12:43:54 +0000 (16:43 +0400)]
rpc_create_*_dir: don't bother with qstr

just pass the name

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
11 years agoMerge branch 'for_linus' of git://cavan.codon.org.uk/platform-drivers-x86
Linus Torvalds [Sun, 14 Jul 2013 01:08:23 +0000 (18:08 -0700)]
Merge branch 'for_linus' of git://cavan.codon.org.uk/platform-drivers-x86

Pull x86 platform driver updates from Matthew Garrett:
 "Nothing overly exciting here - a couple of new drivers that don't do a
  great deal, along with some miscellaneous fixes and a couple of small
  feature enablement patches"

* 'for_linus' of git://cavan.codon.org.uk/platform-drivers-x86:
  x86 platform drivers: fix gpio leak
  toshiba_acpi: Add dependency on SERIO_I8042
  asus-nb-wmi: set wapf=4 for ASUSTeK COMPUTER INC. 1015E/U
  Add trivial driver to disable Intel Smart Connect
  Add support driver for Intel Rapid Start Technology
  hp-wmi: add supports for POST code error
  asus-wmi: control wlan-led only if wapf == 4
  drivers/platform/x86/intel_ips: Convert to module_pci_driver
  asus-nb-wmi: ignore ALS notification key code
  asus-wmi: append newline to messages
  x86: asus-laptop: fix invalid point access
  x86: msi-laptop: fix memleak
  amilo-rfkill: Add dependency on SERIO_I8042
  dell-laptop: fix error return code in dell_init()
  hp-wmi: Enable hotkeys on some systems

11 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
Linus Torvalds [Sun, 14 Jul 2013 01:05:13 +0000 (18:05 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/dtor/input

Pull second round of input updates from Dmitry Torokhov:
 "An update to Elantech driver to support hardware v7, fix to the new
  cyttsp4 driver to use proper addressing, ads7846 device tree support
  and nspire-keypad got a small cleanup."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: nspire-keypad - replace magic offset with define
  Input: elantech - fix for newer hardware versions (v7)
  Input: cyttsp4 - use 16bit address for I2C/SPI communication
  Input: ads7846 - add device tree bindings
  Input: ads7846 - make sure we do not change platform data

11 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Sun, 14 Jul 2013 00:42:22 +0000 (17:42 -0700)]
Merge git://git./linux/kernel/git/davem/net

Pull networking fixes from David Miller:
 "Just a bunch of small fixes and tidy ups:

   1) Finish the "busy_poll" renames, from Eliezer Tamir.

   2) Fix RCU stalls in IFB driver, from Ding Tianhong.

   3) Linearize buffers properly in tun/macvtap zerocopy code.

   4) Don't crash on rmmod in vxlan, from Pravin B Shelar.

   5) Spinlock used before init in alx driver, from Maarten Lankhorst.

   6) A sparse warning fix in bnx2x broke TSO checksums, fix from Dmitry
      Kravkov.

   7) Dummy and ifb driver load failure paths can oops, fixes from Tan
      Xiaojun and Ding Tianhong.

   8) Correct MTU calculations in IP tunnels, from Alexander Duyck.

   9) Account all TCP retransmits in SNMP stats properly, from Yuchung
      Cheng.

  10) atl1e and via-rhine do not handle DMA mapping failures properly,
      from Neil Horman.

  11) Various equal-cost multipath route fixes in ipv6 from Hannes
      Frederic Sowa"

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (36 commits)
  ipv6: only static routes qualify for equal cost multipathing
  via-rhine: fix dma mapping errors
  atl1e: fix dma mapping warnings
  tcp: account all retransmit failures
  usb/net/r815x: fix cast to restricted __le32
  usb/net/r8152: fix integer overflow in expression
  net: access page->private by using page_private
  net: strict_strtoul is obsolete, use kstrtoul instead
  drivers/net/ieee802154: don't use devm_pinctrl_get_select_default() in probe
  drivers/net/ethernet/cadence: don't use devm_pinctrl_get_select_default() in probe
  drivers/net/can/c_can: don't use devm_pinctrl_get_select_default() in probe
  net/usb: add relative mii functions for r815x
  net/tipc: use %*phC to dump small buffers in hex form
  qlcnic: Adding Maintainers.
  gre: Fix MTU sizing check for gretap tunnels
  pkt_sched: sch_qfq: remove forward declaration of qfq_update_agg_ts
  pkt_sched: sch_qfq: improve efficiency of make_eligible
  gso: Update tunnel segmentation to support Tx checksum offload
  inet: fix spacing in assignment
  ifb: fix oops when loading the ifb failed
  ...

11 years agoMerge tag 'scsi-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb...
Linus Torvalds [Sun, 14 Jul 2013 00:41:21 +0000 (17:41 -0700)]
Merge tag 'scsi-for-linus' of git://git./linux/kernel/git/jejb/scsi

Pull final round of SCSI updates from James Bottomley:
 "This is the remaining set of SCSI patches for the merge window.  It's
  mostly driver updates (scsi_debug, qla2xxx, storvsc, mp3sas).  There
  are also several bug fixes in fcoe, libfc, and megaraid_sas.  We also
  have a couple of core changes to try to make device destruction more
  deterministic"

* tag 'scsi-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (46 commits)
  [SCSI] scsi constants: command, sense key + additional sense strings
  fcoe: Reduce number of sparse warnings
  fcoe: Stop fc_rport_priv structure leak
  libfcoe: Fix meaningless log statement
  libfc: Differentiate echange timer cancellation debug statements
  libfc: Remove extra space in fc_exch_timer_cancel definition
  fcoe: fix the link error status block sparse warnings
  fcoe: Fix smatch warning in fcoe_fdmi_info function
  libfc: Reject PLOGI from nodes with incompatible role
  [SCSI] enable destruction of blocked devices which fail LUN scanning
  [SCSI] Fix race between starved list and device removal
  [SCSI] megaraid_sas: fix a bug for 64 bit arches
  [SCSI] scsi_debug: reduce duplication between prot_verify_read and prot_verify_write
  [SCSI] scsi_debug: simplify offset calculation for dif_storep
  [SCSI] scsi_debug: invalidate protection info for unmapped region
  [SCSI] scsi_debug: fix NULL pointer dereference with parameters dif=0 dix=1
  [SCSI] scsi_debug: fix incorrectly nested kmap_atomic()
  [SCSI] scsi_debug: fix invalid address passed to kunmap_atomic()
  [SCSI] mpt3sas: Bump driver version to v02.100.00.00
  [SCSI] mpt3sas: when async scanning is enabled then while scanning, devices are removed but their transport layer entries are not removed
  ...

11 years agoMerge branch 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 13 Jul 2013 22:37:57 +0000 (15:37 -0700)]
Merge branch 'sched-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull scheduler fix from Thomas Gleixner:
 "Fix a potential deadlock versus hrtimers"

* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: Fix HRTICK

11 years agoMerge branch 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 13 Jul 2013 22:37:30 +0000 (15:37 -0700)]
Merge branch 'irq-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull irq updates from Thomas Gleixner:
 - core fix for missing round up in the generic irq chip implementation
 - new irq chip for MOXA SoCs
 - a few fixes and cleanups in the irqchip drivers

* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  irqchip: Add support for MOXA ART SoCs
  genirq: generic chip: Use DIV_ROUND_UP to calculate numchips
  irqchip: nvic: Fix wrong num_ct argument for irq_alloc_domain_generic_chips()
  irqchip: sun4i: Staticize sun4i_irq_ack()
  irqchip: vt8500: Staticize local symbols

11 years agoMerge branch 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 13 Jul 2013 22:36:09 +0000 (15:36 -0700)]
Merge branch 'timers-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull timer updates from Thomas Gleixner:
 - watchdog fixes for full dynticks
 - improved debug output for full dynticks
 - remove an obsolete full dynticks check
 - two ARM SoC clocksource drivers for sharing across SoCs
 - tick broadcast fix for CPU hotplug

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  tick: broadcast: Check broadcast mode on CPU hotplug
  clocksource: arm_global_timer: Add ARM global timer support
  clocksource: Add Marvell Orion SoC timer
  nohz: Remove obsolete check for full dynticks CPUs to be RCU nocbs
  watchdog: Boot-disable by default on full dynticks
  watchdog: Rename confusing state variable
  watchdog: Register / unregister watchdog kthreads on sysctl control
  nohz: Warn if the machine can not perform nohz_full

11 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 13 Jul 2013 22:35:47 +0000 (15:35 -0700)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf fixes from Thomas Gleixner:
 - fix for do_div() abuse on x86
 - locking fix in perf core
 - a pile of (build) fixes and cleanups in perf tools

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (27 commits)
  perf/x86: Fix incorrect use of do_div() in NMI warning
  perf: Fix perf_lock_task_context() vs RCU
  perf: Remove WARN_ON_ONCE() check in __perf_event_enable() for valid scenario
  perf: Clone child context from parent context pmu
  perf script: Fix broken include in Context.xs
  perf tools: Fix -ldw/-lelf link test when static linking
  perf tools: Revert regression in configuration of Python support
  perf tools: Fix perf version generation
  perf stat: Fix per-socket output bug for uncore events
  perf symbols: Fix vdso list searching
  perf evsel: Fix missing increment in sample parsing
  perf tools: Update symbol_conf.nr_events when processing attribute events
  perf tools: Fix new_term() missing free on error path
  perf tools: Fix parse_events_terms() segfault on error path
  perf evsel: Fix count parameter to read call in event_format__new
  perf tools: fix a typo of a Power7 event name
  perf tools: Fix -x/--exclude-other option for report command
  perf evlist: Enhance perf_evlist__start_workload()
  perf record: Remove -f/--force option
  perf record: Remove -A/--append option
  ...

11 years agoMerge branch 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sat, 13 Jul 2013 22:35:12 +0000 (15:35 -0700)]
Merge branch 'core-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull core locking updates from Thomas Gleixner:
 "Header cleanup as requested by Linus"

(This is the "don't include support for ww_mutex in a header file that
everybody wants, when almost nobody wants the ww part" change)

* 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  mutex: Move ww_mutex definitions to ww_mutex.h

11 years agoMerge tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm...
Linus Torvalds [Sat, 13 Jul 2013 22:00:26 +0000 (15:00 -0700)]
Merge tag 'fixes-for-linus' of git://git./linux/kernel/git/arm/arm-soc

Pull ARM SoC fixes from Olof Johansson:
 "This is our first set of fixes from arm-soc for 3.11.
   - A handful of build and warning fixes from Arnd
   - A collection of OMAP fixes
   - defconfig updates to make the default configs more useful for real
     use (and testing) out of the box on hardware

  And a couple of other small fixes.  Some of these have been recently
  applied but it's normally how we deal with fixes, with less bake time
  in -next needed"

* tag 'fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (32 commits)
  arm: multi_v7_defconfig: Tweaks for omap and sunxi
  arm: multi_v7_defconfig: add i.MX options and NFS root
  ARM: omap2: add select of TI_PRIV_EDMA
  ARM: exynos: select PM_GENERIC_DOMAINS only when used
  ARM: ixp4xx: avoid circular header dependency
  ARM: OMAP: omap_common_late_init may be unused
  ARM: sti: move DEBUG_STI_UART into alphabetical order
  ARM: OMAP: build mach-omap code only if needed
  ARM: zynq: use DT_MACHINE_START
  ARM: omap5: omap5 has SCU and TWD
  ARM: OMAP2+: omap2plus_defconfig: Enable appended DTB support
  ARM: OMAP2+: Enable TI_EDMA in omap2plus_defconfig
  ARM: OMAP2+: omap2plus_defconfig: enable DRA752 thermal support by default
  ARM: OMAP2+: omap2plus_defconfig: enable TI bandgap driver
  ARM: OMAP2+: devices: remove duplicated include from devices.c
  ARM: OMAP3: igep0020: Set DSS pins in correct mux mode.
  ARM: OMAP2+: N900: enable N900-specific drivers even if device tree is enabled
  ARM: OMAP2+: Cocci spatch "ptr_ret.spatch"
  ARM: OMAP2+: Remove obsolete Makefile line
  ARM: OMAP5: Enable Cortex A15 errata 798181
  ...

11 years agoMerge branch 'fixes' of git://git.linaro.org/people/rmk/linux-arm
Linus Torvalds [Sat, 13 Jul 2013 21:58:36 +0000 (14:58 -0700)]
Merge branch 'fixes' of git://git.linaro.org/people/rmk/linux-arm

Pull ARM fixes from Russell King:
 "A few fixes for ARM, mostly just one liners with the exception of the
  missing section specification.  We decided not to rely on .previous to
  fix this but to explicitly state the section we want the code to be
  in."

* 'fixes' of git://git.linaro.org/people/rmk/linux-arm:
  ARM: 7778/1: smp_twd: twd_update_frequency need be run on all online CPUs
  ARM: 7782/1: Kconfig: Let ARM_ERRATA_364296 not depend on CONFIG_SMP
  ARM: mm: fix boot on SA1110 Assabet
  ARM: 7781/1: mmu: Add debug_ll_io_init() mappings to early mappings
  ARM: 7780/1: add missing linker section markup to head-common.S

11 years agoMerge branch 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus
Linus Torvalds [Sat, 13 Jul 2013 21:52:21 +0000 (14:52 -0700)]
Merge branch 'upstream' of git://git.linux-mips.org/ralf/upstream-linus

Pull MIPS updates from Ralf Baechle:
 "MIPS updates:

   - All the things that didn't make 3.10.
   - Removes the Windriver PPMC platform.  Nobody will miss it.
   - Remove a workaround from kernel/irq/irqdomain.c which was there
     exclusivly for MIPS.  Patch by Grant Likely.
   - More small improvments for the SEAD 3 platform
   - Improvments on the BMIPS / SMP support for the BCM63xx series.
   - Various cleanups of dead leftovers.
   - Platform support for the Cavium Octeon-based EdgeRouter Lite.

  Two large KVM patchsets didn't make it for this pull request because
  their respective authors are vacationing"

* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus: (124 commits)
  MIPS: Kconfig: Add missing MODULES dependency to VPE_LOADER
  MIPS: BCM63xx: CLK: Add dummy clk_{set,round}_rate() functions
  MIPS: SEAD3: Disable L2 cache on SEAD-3.
  MIPS: BCM63xx: Enable second core SMP on BCM6328 if available
  MIPS: BCM63xx: Add SMP support to prom.c
  MIPS: define write{b,w,l,q}_relaxed
  MIPS: Expose missing pci_io{map,unmap} declarations
  MIPS: Malta: Update GCMP detection.
  Revert "MIPS: make CAC_ADDR and UNCAC_ADDR account for PHYS_OFFSET"
  MIPS: APSP: Remove <asm/kspd.h>
  SSB: Kconfig: Amend SSB_EMBEDDED dependencies
  MIPS: microMIPS: Fix improper definition of ISA exception bit.
  MIPS: Don't try to decode microMIPS branch instructions where they cannot exist.
  MIPS: Declare emulate_load_store_microMIPS as a static function.
  MIPS: Fix typos and cleanup comment
  MIPS: Cleanup indentation and whitespace
  MIPS: BMIPS: support booting from physical CPU other than 0
  MIPS: Only set cpu_has_mmips if SYS_SUPPORTS_MICROMIPS
  MIPS: GIC: Fix gic_set_affinity infinite loop
  MIPS: Don't save/restore OCTEON wide multiplier state on syscalls.
  ...

11 years agoMerge git://www.linux-watchdog.org/linux-watchdog
Linus Torvalds [Sat, 13 Jul 2013 20:42:54 +0000 (13:42 -0700)]
Merge git://www.linux-watchdog.org/linux-watchdog

Pull watchdog updates from Wim Van Sebroeck:
 - lots of devm_ conversions and cleanup
 - platform_set_drvdata cleanups
 - s3c2410: dev_err/dev_info + dev_pm_ops
 - watchdog_core: don't try to stop device if not running fix
 - wdrtas: use print_hex_dump
 - xilinx cleanups
 - orion_wdt fixes
 - softdog cleanup
 - hpwdt: check on UEFI bits
 - deletion of mpcore_wdt driver
 - addition of broadcom BCM2835 watchdog timer driver
 - addition of MEN A21 watcdog devices

* git://www.linux-watchdog.org/linux-watchdog: (38 commits)
  watchdog: hpwdt: Add check for UEFI bits
  watchdog: softdog: remove replaceable ping operation
  watchdog: New watchdog driver for MEN A21 watchdogs
  Watchdog: fix clearing of the watchdog interrupt
  Watchdog: allow orion_wdt to be built for Dove
  watchdog: Add Broadcom BCM2835 watchdog timer driver
  watchdog: delete mpcore_wdt driver
  watchdog: xilinx: Setup the origin compatible string
  watchdog: xilinx: Fix driver header
  watchdog: wdrtas: don't use custom version of print_hex_dump
  watchdog: core: don't try to stop device if not running
  watchdog: jz4740: Pass device to clk_get
  watchdog: twl4030: Remove redundant platform_set_drvdata()
  watchdog: mpcore: Remove redundant platform_set_drvdata()
  watchdog: da9055: use platform_{get,set}_drvdata()
  watchdog: da9052: use platform_{get,set}_drvdata()
  watchdog: cpwd: use platform_{get,set}_drvdata()
  watchdog: s3c2410_wdt: convert s3c2410wdt to dev_pm_ops
  watchdog: s3c2410_wdt: use dev_err()/dev_info() instead of pr_err()/pr_info()
  watchdog: wm831x: use platform_{get,set}_drvdata()
  ...

11 years agoMerge tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland...
Linus Torvalds [Sat, 13 Jul 2013 19:57:21 +0000 (12:57 -0700)]
Merge tag 'rdma-for-linus' of git://git./linux/kernel/git/roland/infiniband

Pull InfiniBand/RDMA changes from Roland Dreier:
 - AF_IB (native IB addressing) for CMA from Sean Hefty
 - new mlx5 driver for Mellanox Connect-IB adapters (including post
   merge request fixes)
 - SRP fixes from Bart Van Assche (including fix to first merge request)
 - qib HW driver updates
 - resurrection of ocrdma HW driver development
 - uverbs conversion to create fds with O_CLOEXEC set
 - other small changes and fixes

* tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (66 commits)
  mlx5: Return -EFAULT instead of -EPERM
  IB/qib: Log all SDMA errors unconditionally
  IB/qib: Fix module-level leak
  mlx5_core: Adjust hca_cap.uar_page_sz to conform to Connect-IB spec
  IB/srp: Let srp_abort() return FAST_IO_FAIL if TL offline
  IB/uverbs: Use get_unused_fd_flags(O_CLOEXEC) instead of get_unused_fd()
  mlx5_core: Fixes for sparse warnings
  IB/mlx5: Make profile[] static in main.c
  mlx5: Fix parameter type of health_handler_t
  mlx5: Add driver for Mellanox Connect-IB adapters
  IB/core: Add reserved values to enums for low-level driver use
  IB/srp: Bump driver version and release date
  IB/srp: Make HCA completion vector configurable
  IB/srp: Maintain a single connection per I_T nexus
  IB/srp: Fail I/O fast if target offline
  IB/srp: Skip host settle delay
  IB/srp: Avoid skipping srp_reset_host() after a transport error
  IB/srp: Fix remove_one crash due to resource exhaustion
  IB/qib: New transmitter tunning settings for Dell 1.1 backplane
  IB/core: Fix error return code in add_port()
  ...

11 years agoMerge branch 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab...
Linus Torvalds [Sat, 13 Jul 2013 19:09:57 +0000 (12:09 -0700)]
Merge branch 'v4l_for_linus' of git://git./linux/kernel/git/mchehab/linux-media

Pull media updates from Mauro Carvalho Chehab:
 "This series contain:
   - new i2c video drivers: ml86v7667 (video decoder),
                            ths8200 (video encoder)
   - a new video driver for EasyCap cards based on Fushicai USBTV007
   - Improved support for OF and embedded systems, with V4L2 async
     initialization and a better support for clocks
   - API cleanups on the ioctls used by the v4l2 debug tool
   - Lots of cleanups
   - As usual, several driver improvements and new cards additions
   - Revert two changesets that change the minimal symbol rate for
     stv0399, as request by Manu
   - Update MAINTAINERS and other files to point to my new e-mail"

* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (378 commits)
  MAINTAINERS & ABI: Update to point to my new email
  [media] stb0899: restore minimal rate to 5Mbauds
  [media] exynos4-is: Correct colorspace handling at FIMC-LITE
  [media] exynos4-is: Set valid initial format on FIMC.n subdevs
  [media] exynos4-is: Set valid initial format on FIMC-IS-ISP subdev pads
  [media] exynos4-is: Fix format propagation on FIMC-IS-ISP subdev
  [media] exynos4-is: Set valid initial format at FIMC-LITE
  [media] exynos4-is: Fix format propagation on FIMC-LITE.n subdevs
  [media] MAINTAINERS: Update S5P/Exynos FIMC driver entry
  [media] Documentation: Update driver's directory in video4linux/fimc.txt
  [media] exynos4-is: Change fimc-is firmware file names
  [media] exynos4-is: Add support for Exynos5250 MIPI-CSIS
  [media] exynos4-is: Add Exynos5250 SoC support to fimc-lite driver
  [media] exynos4-is: Drop drvdata handling in fimc-lite for non-dt platforms
  [media] media: i2c: tvp514x: remove manual setting of subdev name
  [media] media: i2c: tvp7002: remove manual setting of subdev name
  [media] mem2mem: set missing v4l2_dev pointer
  [media] wl128x: add missing struct v4l2_device
  [media] tvp514x: Fix init seqeunce
  [media] saa7134: Fix sparse warnings by adding __user annotation
  ...