linux-toradex.git/fs/ceph, branch v3.12.29

ceph: Avoid data inconsistency due to d-cache aliasing in readpage()

2014-01-09T20:25:11+00:00

commit 56f91aad69444d650237295f68c195b74d888d95 upstream.

If the length of data to be read in readpage() is exactly
PAGE_CACHE_SIZE, the original code does not flush d-cache
for data consistency after finishing reading. This patches fixes
this.

Signed-off-by: Li Wang 
Signed-off-by: Sage Weil 
Signed-off-by: Greg Kroah-Hartman

ceph: allocate non-zero page to fscache in readpage()

2014-01-09T20:25:07+00:00

commit ff638b7df5a9264024a6448bdfde2b2bf5d1994a upstream.

ceph_osdc_readpages() returns number of bytes read, currently,
the code only allocate full-zero page into fscache, this patch
fixes this.

Signed-off-by: Li Wang 
Reviewed-by: Milosz Tanski 
Reviewed-by: Sage Weil 
Signed-off-by: Greg Kroah-Hartman

ceph: wake up 'safe' waiters when unregistering request

2014-01-09T20:25:07+00:00

commit fc55d2c9448b34218ca58733a6f51fbede09575b upstream.

We also need to wake up 'safe' waiters if error occurs or request
aborted. Otherwise sync(2)/fsync(2) may hang forever.

Signed-off-by: Yan, Zheng 
Signed-off-by: Sage Weil 
Signed-off-by: Greg Kroah-Hartman

ceph: cleanup aborted requests when re-sending requests.

2014-01-09T20:25:07+00:00

commit eb1b8af33c2e42a9a57fc0a7588f4a7b255d2e79 upstream.

Aborted requests usually get cleared when the reply is received.
If MDS crashes, no reply will be received. So we need to cleanup
aborted requests when re-sending requests.

Signed-off-by: Yan, Zheng 
Reviewed-by: Greg Farnum 
Signed-off-by: Sage Weil 
Signed-off-by: Greg Kroah-Hartman

ceph: hung on ceph fscache invalidate in some cases

2014-01-09T20:25:06+00:00

commit ffc79664d15841025d90afdd902c4112ffe168d6 upstream.

In some cases I'm on my ceph client cluster I'm seeing hunk kernel tasks in
the invalidate page code path. This is due to the fact that we don't check if
the page is marked as cache before calling fscache_wait_on_page_write().

This is the log from the hang

INFO: task XXXXXX:12034 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 ...
Call Trace:
[] schedule+0x29/0x70
[] __fscache_wait_on_page_write+0x6d/0xb0 [fscache]
[] ? add_wait_queue+0x60/0x60
[] ceph_invalidate_fscache_page+0x29/0x50 [ceph]
[] ceph_invalidatepage+0x70/0x190 [ceph]
[] ? delete_from_page_cache+0x5f/0x70
[] truncate_inode_page+0x8b/0x90
[] truncate_inode_pages_range.part.12+0x13d/0x620
[] truncate_inode_pages_range+0x4d/0x60
[] truncate_inode_pages+0x15/0x20
[] evict+0x1a6/0x1b0
[] iput+0x103/0x190
 ...

Signed-off-by: Milosz Tanski 
Reviewed-by: Sage Weil 
Signed-off-by: Greg Kroah-Hartman

ceph: use d_invalidate() to invalidate aliases

2013-09-06T19:55:29+00:00

d_invalidate() is the standard VFS method to invalidate dentry.
compare to d_delete(), it also try shrinking children dentries.

Signed-off-by: Yan, Zheng 
Reviewed-by: Sage Weil

ceph: remove ceph_lookup_inode()

2013-09-06T19:55:09+00:00

commit 6f60f889 (ceph: fix freeing inode vs removing session caps race)
introduced ceph_lookup_inode(). But there is already a ceph_find_inode()
which provides similar function. So remove ceph_lookup_inode(), use
ceph_find_inode() instead.

Signed-off-by: Yan, Zheng 
Reviewed-by: Alex Elder 
Reviewed-by: Sage Weil

ceph: trivial buildbot warnings fix

2013-09-06T16:50:12+00:00

The linux-next build bot found a three of warnings, this addresses all of them.

 * non-ANSI function declaration of function 'ceph_fscache_register' and
   'ceph_fscache_unregister'
 * symbol 'ceph_cache_netfs' was not declared, now it's extern in the header.
 * warning: "pr_fmt" redefined

Signed-off-by: Milosz Tanski

ceph: Do not do invalidate if the filesystem is mounted nofsc

2013-09-06T16:50:12+00:00

Previously we would always try to enqueue work even if the filesystem is not
mounted with fscache enabled (or the file has no cookie). In the case of the
filesystem mouned nofsc (but with fscache compiled in) this would lead to a
crash.

Signed-off-by: Milosz Tanski

ceph: page still marked private_2

2013-09-06T16:50:12+00:00

Previous patch that allowed us to cleanup most of the issues with pages marked
as private_2 when calling ceph_readpages. However, there seams to be a case in
the error case clean up in start read that still trigers this from time to
time. I've only seen this one a couple times.

BUG: Bad page state in process petabucket  pfn:335b82
page:ffffea000cd6e080 count:0 mapcount:0 mapping:          (null) index:0x0
page flags: 0x200000000001000(private_2)
Call Trace:
 [] dump_stack+0x46/0x58
 [] bad_page+0xc7/0x120
 [] free_pages_prepare+0x10e/0x120
 [] free_hot_cold_page+0x40/0x160
 [] __put_single_page+0x27/0x30
 [] put_page+0x25/0x40
 [] ceph_readpages+0x2e9/0x6f0 [ceph]
 [] __do_page_cache_readahead+0x1af/0x260

Signed-off-by: Milosz Tanski 
Signed-off-by: Sage Weil