28 Mar, 2007
7 commits
-
The user can generate console output if they cause do_mmap() to fail
during sys_io_setup(). This was seen in a regression test that does
exactly that by spinning calling mmap() until it gets -ENOMEM before
calling io_setup().We don't need this printk at all, just remove it.
Signed-off-by: Zach Brown
Signed-off-by: Linus Torvalds -
* 'for-linus' of git://git.kernel.dk/data/git/linux-2.6-block:
Export __splice_from_pipe()
2/2 splice: dont readpage
1/2 splice: dont steal
make elv_register() output atomic
block: blk_max_pfn is somtimes wrong -
Without attached patch against current -git I get following with
!PROC_SYSCTL (with EMBEDDED and PROC_FS set):CC init/version.o
LD init/built-in.o
LD vmlinux
fs/built-in.o: In function `do_proc_sys_lookup':
proc_sysctl.c:(.text+0x26583): undefined reference to `sysctl_head_next'
fs/built-in.o: In function `proc_sys_revalidate':
proc_sysctl.c:(.text+0x265bb): undefined reference to `sysctl_head_finish'
fs/built-in.o: In function `proc_sys_readdir':
proc_sysctl.c:(.text+0x26720): undefined reference to `sysctl_head_next'
proc_sysctl.c:(.text+0x267d8): undefined reference to `sysctl_head_finish'
proc_sysctl.c:(.text+0x268e7): undefined reference to `sysctl_head_next'
proc_sysctl.c:(.text+0x26910): undefined reference to `sysctl_head_finish'
fs/built-in.o: In function `proc_sys_write':
proc_sysctl.c:(.text+0x2695d): undefined reference to `sysctl_perm'
proc_sysctl.c:(.text+0x2699c): undefined reference to `sysctl_head_finish'
fs/built-in.o: In function `proc_sys_read':
proc_sysctl.c:(.text+0x269e9): undefined reference to `sysctl_perm'
proc_sysctl.c:(.text+0x26a25): undefined reference to `sysctl_head_finish'
fs/built-in.o: In function `proc_sys_permission':
proc_sysctl.c:(.text+0x26ad1): undefined reference to `sysctl_perm'
proc_sysctl.c:(.text+0x26adb): undefined reference to `sysctl_head_finish'
fs/built-in.o: In function `proc_sys_lookup':
proc_sysctl.c:(.text+0x26b39): undefined reference to `sysctl_head_finish'
make: *** [vmlinux] Virhe 1All those functions are in fs/proc/proc_sysctl.c, which has no CONFIG_
#define's in it, so the patch makes the compilation of that file to depend
on CONFIG_PROC_SYSCTL (the simplest choice).Acked-by: "Eric W. Biederman"
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This cancel_delayed_work call is called from a function that is only called
from a piece of code that immediate follows a cancel and destruction of the
workqueue, so it's clearly a mistake.Cc: Oleg Nesterov
Signed-off-by: "J. Bruce Fields"
Signed-off-by: Neil Brown
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The reused clientid here is a more of a problem for the client than the
server, and the client can report the problem itself if it's serious.Signed-off-by: "J. Bruce Fields"
Signed-off-by: Neil Brown
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
A regression introduced in the last set of acl patches removed the
INHERIT_ONLY flag from aces derived from the posix acl. Fix.Signed-off-by: "J. Bruce Fields"
Signed-off-by: Neil Brown
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
->readdir passes lofft_t offsets (used as nfs cookies) to
nfs3svc_encode_entry{,_plus}, but when they pass it on to encode_entry it
becomes an 'off_t', which isn't good.So filesystems that returned 64bit offsets would lose.
Signed-off-by: Neil Brown
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
27 Mar, 2007
4 commits
-
Ocfs2 wants to implement it's own splice write actor so that it can better
manage cluster / page locks. This lets us re-use the rest of splice write
while only providing our own code where it's actually important.Signed-off-by: Mark Fasheh
Signed-off-by: Jens Axboe -
Splice does not need to readpage to bring the page uptodate before writing
to it, because prepare_write will take care of that for us.Splice is also wrong to SetPageUptodate before the page is actually uptodate.
This results in the old uninitialised memory leak. This gets fixed as a
matter of course when removing the readpage logic.Signed-off-by: Nick Piggin
Signed-off-by: Jens Axboe -
Stealing pages with splice is problematic because we cannot just insert
an uptodate page into the pagecache and hope the filesystem can take care
of it later.We also cannot just ClearPageUptodate, then hope prepare_write does not
write anything into the page, because I don't think prepare_write gives
that guarantee.Remove support for SPLICE_F_MOVE for now. If we really want to bring it
back, we might be able to do so with a the new filesystem buffered write
aops APIs I'm working on. If we really don't want to bring it back, then
we should decide that sooner rather than later, and remove the flag and
all the stealing infrastructure before anybody starts using it.Signed-off-by: Nick Piggin
Signed-off-by: Jens Axboe -
This patch makes te needlessly global struct v9fs_cached_file_operations
static.Signed-off-by: Adrian Bunk
Signed-off-by: Eric Van Hensbergen
24 Mar, 2007
2 commits
-
A little mistake in 8a2bfdcbfa441d8b0e5cb9c9a7f45f77f80da465 is making all
transactions synchronous, which reduces ext3 performance to comical levels.Cc: Mingming Cao
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Fix the /proc/pid/stat representation of executable boundaries. It should
show the bounds of the executable, but instead shows the bounds of the
loader.Before the patch is applied, the bug can be seen by examining, say, inetd:
# ps | grep inetd
610 root 0 S /usr/sbin/inetd -i
# cat /proc/610/maps
c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so
c3180000-c31dede4 r-xs 00000000 00:0b 14582179 /lib/libuClibc-0.9.28.so
c328c000-c328ea00 rw-p 00008000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so
c3290000-c329b6c0 rw-p 00000000 00:00 0
c32a0000-c32c0000 rwxp 00000000 00:00 0
c32d4000-c32d8000 rw-p 00000000 00:00 0
c3394000-c3398000 rw-p 00000000 00:00 0
c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd
c3470000-c34748f8 rw-p 00004000 00:0b 16384612 /usr/sbin/inetd
c34cc000-c34d0000 rw-p 00000000 00:00 0
c34d4000-c34d8000 rw-p 00000000 00:00 0
c34d8000-c34dc000 rw-p 00000000 00:00 0
# cat /proc/610/stat
610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 8 0 0 19 0 1 0 94392000718
950272 0 4294967295 3233480704 3233523592 3274440352 3274439976
3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0The code boundaries are 3233480704 to 3233523592, which are:
(gdb) p/x 3233480704
$1 = 0xc0bb0000
(gdb) p/x 3233523592
$2 = 0xc0bba788Which corresponds to this line in the maps file:
c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so
Which is wrong. After the patch is applied, the maps file is pretty much
identical (there's some minor shuffling of the location of some of the
anonymous VMAs), but the stat file is now:# cat /proc/610/stat
610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 7 0 0 18 0 1 0 94392000722
950272 0 4294967295 3276111872 3276141668 3274440352 3274439976
3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0The code boundaries are then 3276111872 to 3276141668, which are:
(gdb) p/x 3276111872
$1 = 0xc3458000
(gdb) p/x 3276141668
$2 = 0xc345f464And these correspond to this line in the maps file instead:
c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd
Which is now correct.
Signed-off-by: David Howells
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
23 Mar, 2007
3 commits
-
* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6:
[CIFS] Allow reset of file to ATTR_NORMAL when archive bit not set
[CIFS] Do not negotiate new POSIX_PATH_OPERATIONS_CAP yet
[CIFS] reset mode when client notices that ATTR_READONLY is no longer set -
Since freezable workqueues are broken in 2.6.21-rc
(cf. http://marc.theaimsgroup.com/?l=linux-kernel&m=116855740612755,
http://marc.theaimsgroup.com/?l=linux-kernel&m=117261312523921&w=2)
it's better to change the only user of them, which is XFS, to use "normal"
nonfreezable workqueues.Signed-off-by: Rafael J. Wysocki
Cc: Pavel Machek
Cc: David Chinner
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
When a file had a dos attribute of 0x1 (readonly - but dos attribute
of archive was not set) - doing chmod 0777 or equivalent would
try to set a dos attribute of 0 (which some servers ignore)
rather than ATTR_NORMAL (0x20) which most servers accept.
Does not affect servers which support the CIFS Unix Extensions.Acked-by: Prasad Potluri
Acked-by: Shirish Pargaonkar
Signed-off-by: Steve French
17 Mar, 2007
11 commits
-
This bug was seen on ppc64, but it could have occurred on any
architecture with a page size of 64k or above. The problem is that in
fs/binfmt_elf.c:randomize_stack_top() randomizes the stack to within
0x7ff pages. On 4k page machines, this is 8MB; on 64k page boxes, this
is 128MB.The problem is that the new binary layout (selected in
arch_pick_mmap_layout) places the mapping segment 128MB or the stack
rlimit away from the top of the process memory, whichever is larger. If
you chose an rlimit of less than 128MB (most defaults are in the 8Mb
range) then you can end up having your entire stack randomized away.The fix is to make randomize_stack_top() only steal at most 8MB, which this
patch does. However, I have to point out that even with this, your stack
rlimit might not be exactly what you get if it's > 128MB, because you're
still losing the random offset of up to 8MB.The true fix should be to leave an explicit gap for the randomization plus
a buffer when determining mmap_base, but that would involve fixing all the
architectures.Cc: Arjan van de Ven
Cc: Ingo Molnar
Cc: Andi Kleen
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Remove the misleading "Presently only useful on the IA-64 platform" text
from the EFI partition Kconfig.EFI partitions are also used by Apple on their Intel-based machines and
thus you need EFI partition support if you (for example) want to attach
such a machine in target disk mode.Signed-off-by: Johannes Berg
Acked-by: Matt Domsch
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Looks like we need a check in nfs_getattr() for a regular file. It makes
no sense to call nfs_sync_mapping_range() on anything else. I think that
should fix your problem: it will stop the NFS client from interfering
with dirty pages on that inode's mapping.Signed-off-by: Trond Myklebust
Acked-by: Olof Johansson
Cc:
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The current NFS client congestion logic is severly broken, it marks the
backing device congested during each nfs_writepages() call but doesn't
mirror this in nfs_writepage() which makes for deadlocks. Also it
implements its own waitqueue.Replace this by a more regular congestion implementation that puts a cap on
the number of active writeback pages and uses the bdi congestion waitqueue.Also always use an interruptible wait since it makes sense to be able to
SIGKILL the process even for mounts without 'intr'.Signed-off-by: Peter Zijlstra
Acked-by: Trond Myklebust
Cc: Christoph Lameter
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The only error code which comes from the partition checkers is -1, when
they finds an EIO. As per the discussion, ENOMEM values were ignored,
as they might scare the users.So, with the current code, we end up returning -1 and not EIO for the
ioctl() calls. Which doesn't give any clue to the user of what went
wrong.Signed-off-by: Suzuki K P
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
smbfs allocates rq_trans2buffer to handle server's multi transaction2 response
messages. As struct smb_request may be reused, rq_trans2buffer is freed
before each new request. However if last servers's response is not multi but
single trans2 message then new rq_trans2buffer is not allocated but last
smb_rput still tries to free it again.To prevent this issue rq_trans2buffer pointer should be set to NULL after
kfree.Signed-off-by: Vasily Averin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
ecryptfs_d_release() first dereferences a pointer (via
ecryptfs_dentry_to_lower()) and then afterwards checks to see if the
pointer it just dereferenced is NULL (via ecryptfs_dentry_to_private()).This patch moves all of the work done on the dereferenced pointer inside a
block governed by the condition that the pointer is non-NULL.Signed-off-by: Michael Halcrow
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
During modification of code to support UFS2 writing, the case with
"three indirect" blocks in truncate path was missed, this patch fixes
this situation.Signed-off-by: Evgeniy Dushistov
Acked-by: Christoph Hellwig
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This patch fix behaviour in such test scenario:
lseek(fd, BIG_OFFSET)
write(fd, buf, sizeof(buf))
truncate(BIG_OFFSET)
truncate(BIG_OFFSET + sizeof(buf))
read(fd, buf...)Because of if file big enough(BIG_OFFSET) we start allocate space by block,
ordinary block size > page size, so we should zeroize the rest of block in
truncate(except last framgnet, about which VFS should care), to not get
garbage, when we extend file.Also patch corrects conversion from pointer to block to physical block number,
this helps in case of not common used UFS types.And add to debug output inode number.
Signed-off-by: Evgeniy Dushistov
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This fixes "change blocks numbers on the fly" in case when "prepare
write page" is in the call chain, in this case some buffers may be not
uptodate and not mapped, we should care to map them and load from disk.This patch was tested with:
- ufs regressions simple tests
- fsx-linux
- ltp(20060306)
- untar and build kernelSigned-off-by: Evgeniy Dushistov
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This patch corrects work with time in UFS2 case.
1) According to UFS2 disk layout modification/access and so on "time"
should be hold in two variables one 64bit for seconds and another 32bit for
nanoseconds,at now for some unknown reason we suppose that "inode time" holds in
three variables 32bit for seconds, 32bit for milliseconds and 32bit for
nanoseconds.2) We set amount of nanoseconds in "VFS inode" to 0 during read, instead of
getting values from "on disk inode"(this should close
http://bugzilla.kernel.org/show_bug.cgi?id=7991).Signed-off-by: Evgeniy Dushistov
Cc: Bjoern Jacke
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
16 Mar, 2007
3 commits
-
Samba server now expects that clients which send the new
POSIX_PATH_OPERATIONS_CAP send all opens with this new
SMB - and expects that clients that could send the new
posix open/create but don't as indicating that they really
want Windows semantics on that handle (which allows Samba
to support clients which want to support both types of
behaviors on different handles on the same mount)We will put this capability back in the SetFSInfo
negotiation with servers like Samba when the
new POSIXCreate (create/open/mkdir) code is finished.Signed-off-by: Steve French
-
This patch (as869) reinstates the mutual exclusion between sysfs
attribute method calls and attribute unregistration. The
previously-reported deadlocks have been fixed, and this exclusion is
by far the simplest way to avoid races during driver unbinding.The check for orphaned read-buffers has been moved down slightly, so
that the remainder of a partially-read buffer will still be available
to userspace even after the attribute has been unregistered.Signed-off-by: Alan Stern
Cc: Hugh Dickins
Cc: Cornelia Huck
Cc: Oliver Neukum
Signed-off-by: Linus Torvalds -
This patch (as868) adds a helper routine for device drivers that need
to set up a callback to perform some action in a different process's
context. This is intended for use by attribute methods that want to
unregister themselves or their parent device. Attribute method calls
are mutually exclusive with unregistration, so such actions cannot be
taken directly.Two attribute methods are converted to use the new helper routine: one
for SCSI device deletion and one for System/390 ccwgroup devices.Signed-off-by: Alan Stern
Cc: Hugh Dickins
Cc: Cornelia Huck
Cc: Oliver Neukum
Signed-off-by: Linus Torvalds
15 Mar, 2007
10 commits
-
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/mfasheh/ocfs2:
ocfs2_dlm: Add missing locks in dlm_empty_lockres
ocfs2_dlm: Missing get/put lockres in dlm_run_purge_lockres
configfs: add missing mutex_unlock()
ocfs2: add some missing address space callbacks
ocfs2: Concurrent access of o2hb_region->hr_task was not locked
ocfs2: Proper cleanup in case of error in ocfs2_register_hb_callbacks() -
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
not needed and actually breaks build on frv, while we are at it
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
have it return the buffer it had allocated
Acked-by: Stephen Smalley
Acked-by: James Morris
Signed-off-by: Al Viro
Signed-off-by: Linus Torvalds -
__dlm_lockres_unused() expects the caller to take the lockres spinlock.
Signed-off-by: Sunil Mushran
Signed-off-by: Mark Fasheh -
In some circumstances, this was causing us to reference freed memory.
Signed-off-by: Sunil Mushran
Signed-off-by: Mark Fasheh -
d_alloc() failure in configfs_register_subsystem() would fail to unlock
the mutex taken above. Reorganize the exit path to ensure the unlock
happens.Reported-by: Akinobu Mita
Signed-off-by: Joel Becker
Signed-off-by: Mark Fasheh -
Under load, OCFS2 would crash in invalidate_inode_pages2_range() because
invalidate_complete_page2() was unable to invalidate a page. It would
appear that JBD is holding on to the page. ext3 has a specific
->releasepage() handler to cover this case.Steal ext3's ->releasepage(), ->invalidatepage(), and ->migratepage(), as
they appear completely appropriate for OCFS2.Signed-off-by: Joel Becker
Signed-off-by: Mark Fasheh -
This means that a build-up and a teardown could race which would result in a
double-kthread_stop().Protect the setting and clearing of hr_task with o2hb_live_lock, as it's not
a common thing and not performance critical.Signed-off-by: Joel Becker
Signed-off-by: Mark Fasheh -
If ocfs2_register_hb_callbacks() succeeds on its first callback but fails
its second, it doesn't release the first on the way out. Fix that.While we're at it, o2hb_unregister_callback() never returns anything but
0, so let's make it void.Signed-off-by: Joel Becker
Signed-off-by: Mark Fasheh
