13 Apr, 2010
1 commit
-
Removing sufficiently large files would create aliases for a large
number of segments. This in turn results in a large number of journal
entries and an overflow of s_je_array.Cheap fix is to add a BUG_ON, turning memory corruption into something
annoying, but less dangerous. Real fix is to count the number of
affected segments and prevent the problem completely.Signed-off-by: Joern Engel
31 Mar, 2010
1 commit
-
All callers are long gone.
Signed-off-by: Joern Engel
30 Mar, 2010
2 commits
-
If the device contains on old logfs image and the journal is moved to
segment that have never been used by the current logfs and not all
journal segments are erased before the next mount, the old content can
confuse mount code. To prevent this, always erase the new journal
segments.Signed-off-by: Joern Engel
-
Fixes a GC livelock.
Signed-off-by: Joern Engel
29 Mar, 2010
1 commit
-
do_logfs_journal_wl_pass() must call freeseg(), thereby clear
PagePrivate on all pages of the current journal segment.Signed-off-by: Joern Engel
28 Mar, 2010
2 commits
-
A comment in the old code read:
/* The math in this function can surely use some love */And indeed it did. In the case that area->a_used_bytes is exactly
4096 bytes below segment size it fell apart. pad_wbuf is now split
into two helpers that are significantly less complicated.Signed-off-by: Joern Engel
-
The comment was correct, so make the code match the comment. As the
new comment indicates, we might be able to do a little less work. But
for the current -rc series let's keep it simple and just fix the bug.Signed-off-by: Joern Engel
27 Mar, 2010
10 commits
-
Found by Al Viro.
Signed-off-by: Joern Engel
-
rootdir was already allocated, so we must iput it again.
Found by Al Viro.Signed-off-by: Joern Engel
-
If the first superblock is wrong and the second gets written, there
will still be a mismatch on next mount. Write both to make sure they
match.Signed-off-by: Joern Engel
-
Apparently filldir can sleep, which forbids kmap_atomic.
Signed-off-by: Joern Engel
-
Signed-off-by: Joern Engel
-
Intel SSDs have a limit of 0xffff as queue_max_hw_sectors(q). Such a
limit may make sense from a hardware pov, but it causes bio_alloc() to
return NULL.Signed-off-by: Joern Engel
-
logfs_recover_sb() needs it open.
Signed-off-by: Joern Engel
-
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2:
nilfs2: fix imperfect completion wait in nilfs_wait_on_logs
nilfs2: fix hang-up of cleaner after log writer returned with error
nilfs2: fix duplicate call to nilfs_segctor_cancel_freev -
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
Restore LOOKUP_DIRECTORY hint handling in final lookup on open() -
Lose want_dir argument, while we are at it - since now
nd->flags & LOOKUP_DIRECTORY is equivalent to it.Signed-off-by: Al Viro
26 Mar, 2010
1 commit
-
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
ext4: Fixed inode allocator to correctly track a flex_bg's used_dirs
ext4: Don't use delayed allocation by default when used instead of ext3
ext4: Fix spelling of CONTIG_FS_EXT3 to CONFIG_FS_EXT3
ext4: Fix estimate of # of blocks needed to write indirect-mapped files
25 Mar, 2010
12 commits
-
* 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6:
NFS: don't try to decode GETATTR if DELEGRETURN returned error
sunrpc: handle allocation errors from __rpc_lookup_create()
SUNRPC: Fix the return value of rpc_run_bc_task()
SUNRPC: Fix a use after free bug with the NFSv4.1 backchannel
SUNRPC: Fix a potential memory leak in auth_gss
NFS: Prevent another deadlock in nfs_release_page() -
Sparse complained about this missing spin_unlock()
Signed-off-by: Dan Carpenter
Signed-off-by: David Howells
Signed-off-by: Linus Torvalds -
do_sync_read/write() should set kiocb.ki_nbytes to be consistent with
do_sync_readv_writev().Signed-off-by: David Howells
Signed-off-by: Linus Torvalds -
Fix an incorrect for-loop in elf_core_vma_data_size(). The advance-pointer
statement lacks an assignment:CC fs/binfmt_elf_fdpic.o
fs/binfmt_elf_fdpic.c: In function 'elf_core_vma_data_size':
fs/binfmt_elf_fdpic.c:1593: warning: statement with no effectSigned-off-by: David Howells
Signed-off-by: Linus Torvalds -
Smaller size than a minimum blocksize can't be used, after all it's
handled like 0 size.For extended partition itself, this makes sure to use bigger size than one
logical sector size at least.Signed-off-by: OGAWA Hirofumi
Cc: Daniel Taylor
Cc: "H. Peter Anvin"
Cc:
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
In order to use disks larger than 2TiB on Windows XP, it is necessary to
use 4096-byte logical sectors in an MBR.Although the kernel storage and functions called from msdos.c used
"sector_t" internally, msdos.c still used u32 variables, which results in
the ability to handle XP-compatible large disks.This patch changes the internal variables to "sector_t".
Daniel said: "In the near future, WD will be releasing products that need
this patch".[hirofumi@mail.parknet.co.jp: tweaks and fix]
Signed-off-by: Daniel Taylor
Signed-off-by: OGAWA Hirofumi
Cc: "H. Peter Anvin"
Cc:
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
"m" is never NULL here. We need a different test for the end of list
condition.Signed-off-by: Dan Carpenter
Acked-by: KAMEZAWA Hiroyuki
Acked-by: WANG Cong
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The reiserfs journal behaves inconsistently when determining whether to
allow a mount of a read-only device.This is due to the use of the continue_replay variable to short circuit
the journal scanning. If it's set, it's assumed that there are
transactions to replay, but there may not be. If it's unset, it's assumed
that there aren't any, and that may not be the case either.I've observed two failure cases:
1) Where a clean file system on a read-only device refuses to mount
2) Where a clean file system on a read-only device passes the
optimization and then tries writing the journal header to update
the latest mount id.The former is easily observable by using a freshly created file system on
a read-only loopback device.This patch moves the check into journal_read_transaction, where it can
bail out before it's about to replay a transaction. That way it can go
through and skip transactions where appropriate, yet still refuse to mount
a file system with outstanding transactions.Signed-off-by: Jeff Mahoney
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Commit 57fe60df ("reiserfs: add atomic addition of selinux attributes
during inode creation") contains a bug that will cause it to oops when
mounting a file system that didn't previously contain extended attributes
on a system using security.* xattrs.The issue is that while creating the privroot during mount
reiserfs_security_init calls reiserfs_xattr_jcreate_nblocks which
dereferences the xattr root. The xattr root doesn't exist, so we get an
oops.Addresses http://bugzilla.kernel.org/show_bug.cgi?id=15309
Signed-off-by: Jeff Mahoney
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
fs/binfmt_aout.c: In function `aout_core_dump':
fs/binfmt_aout.c:125: warning: passing argument 2 of `dump_write' makes pointer from integer without a cast
include/linux/coredump.h:12: note: expected `const void *' but argument is of type `long unsigned int'
fs/binfmt_aout.c:132: warning: passing argument 2 of `dump_write' makes pointer from integer without a cast
include/linux/coredump.h:12: note: expected `const void *' but argument is of type `long unsigned int'due to dump_write() expecting a user void *. Fold casts into the
START_DATA/START_STACK macros and shut up the warnings.Signed-off-by: Borislav Petkov
Cc: Daisuke HATAYAMA
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
When ext4 driver is used to mount a filesystem instead of the ext3 file
system driver (through CONFIG_EXT4_USE_FOR_EXT23), do not enable delayed
allocation by default since some ext3 users and application writers have
developed unfortunate expectations about the safety of writing files on
systems subject to sudden and violent death without using fsync().Signed-off-by: Jan Kara
Signed-off-by: "Theodore Ts'o" -
Oops. (Blush.)
Thanks to Sedat Dilek for pointing this out.
Signed-off-by: "Theodore Ts'o"
24 Mar, 2010
2 commits
-
When used_dirs was introduced for the flex_groups struct, it looks
like the accounting was not put into place properly, in some places
manipulating free_inodes rather than used_dirs.Signed-off-by: Eric Sandeen
Signed-off-by: "Theodore Ts'o" -
nilfs_wait_on_logs has a potential to slip out before completion of
all bio requests when it met an error. This synchronization fault may
cause unexpected results, for instance, violative access to freed
segment buffers from an end-bio callback routine.This fixes the issue by ensuring that nilfs_wait_on_logs waits all
given logs.Signed-off-by: Ryusuke Konishi
23 Mar, 2010
2 commits
-
According to the report from Andreas Beckmann (Message-ID:
), nilfs in 2.6.33 kernel got stuck
after a disk full error.This turned out to be a regression by log writer updates merged at
kernel 2.6.33. nilfs_segctor_abort_construction, which is a cleanup
function for erroneous cases, was skipping writeback completion for
some logs.This fixes the bug and would resolve the hang issue.
Reported-by: Andreas Beckmann
Signed-off-by: Ryusuke Konishi
Tested-by: Ryusuke Konishi
Cc: stable [2.6.33.x] -
It seems clear from the surrounding code that xpermits is allowed to be
NULL here.Signed-off-by: Dan Carpenter
Signed-off-by: David Howells
Signed-off-by: Linus Torvalds
22 Mar, 2010
2 commits
-
The reply parsing code attempts to decode the GETATTR response even if
the DELEGRETURN portion of the compound returned an error. The GETATTR
response won't actually exist if that's the case and we're asking the
parser to read past the end of the response.This bug is fairly benign. The parser catches this without reading past
the end of the response and decode_getfattr returns -EIO. Earlier
kernels however had decode_op_hdr using the READ_BUF macro, and this
bug would make this printk pop any time the client got an error from
a delegreturn:kernel: decode_op_hdr: reply buffer overflowed in line XXXX
More recent kernels seem to have replaced this printk with a dprintk.
Signed-off-by: Jeff Layton
Signed-off-by: Trond Myklebust -
Andreas Beckmann gave me a report that nilfs logged the following
warnings when it got a disk full:nilfs_sufile_do_cancel_free: segment 0 must be clean
nilfs_sufile_do_cancel_free: segment 1 must be cleanThese arise from a duplicate call to nilfs_segctor_cancel_freev in an
error path of log writer. This will fix the issue.Reported-by: Andreas Beckmann
Signed-off-by: Ryusuke Konishi
20 Mar, 2010
3 commits
-
We should not attempt to free the page if __GFP_FS is not set. Otherwise we
can deadlock as perhttp://bugzilla.kernel.org/show_bug.cgi?id=15578
Signed-off-by: Trond Myklebust
Cc: stable@kernel.org -
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client: (205 commits)
ceph: update for write_inode API change
ceph: reset osd after relevant messages timed out
ceph: fix flush_dirty_caps race with caps migration
ceph: include migrating caps in issued set
ceph: fix osdmap decoding when pools include (removed) snaps
ceph: return EBADF if waiting for caps on closed file
ceph: set osd request message front length correctly
ceph: reset front len on return to msgpool; BUG on mismatched front iov
ceph: fix snaptrace decoding on cap migration between mds
ceph: use single osd op reply msg
ceph: reset bits on connection close
ceph: remove bogus mds forward warning
ceph: remove fragile __map_osds optimization
ceph: fix connection fault STANDBY check
ceph: invalidate_authorizer without con->mutex held
ceph: don't clobber write return value when using O_SYNC
ceph: fix client_request_forward decoding
ceph: drop messages on unregistered mds sessions; cleanup
ceph: fix comments, locking in destroy_inode
ceph: move dereference after NULL test
...Fix trivial conflicts in Documentation/ioctl/ioctl-number.txt
-
* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6:
cifs: trivial white space
[CIFS] checkpatch cleanup
cifs: add cifs_revalidate_file
cifs: add a CIFSSMBUnixQFileInfo function
cifs: add a CIFSSMBQFileInfo function
cifs: overhaul cifs_revalidate and rename to cifs_revalidate_dentry
19 Mar, 2010
1 commit
-
* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable: (30 commits)
Btrfs: fix the inode ref searches done by btrfs_search_path_in_tree
Btrfs: allow treeid==0 in the inode lookup ioctl
Btrfs: return keys for large items to the search ioctl
Btrfs: fix key checks and advance in the search ioctl
Btrfs: buffer results in the space_info ioctl
Btrfs: use __u64 types in ioctl.h
Btrfs: fix search_ioctl key advance
Btrfs: fix gfp flags masking in the compression code
Btrfs: don't look at bio flags after submit_bio
btrfs: using btrfs_stack_device_id() get devid
btrfs: use memparse
Btrfs: add a "df" ioctl for btrfs
Btrfs: cache the extent state everywhere we possibly can V2
Btrfs: cache ordered extent when completing io
Btrfs: cache extent state in find_delalloc_range
Btrfs: change the ordered tree to use a spinlock instead of a mutex
Btrfs: finish read pages in the order they are submitted
btrfs: fix btrfs_mkdir goto for no free objectids
Btrfs: flush data on snapshot creation
Btrfs: make df be a little bit more understandable
...