31 May, 2007
2 commits
-
It should be pass "newsize" to vmtruncate function to modify the
inode->i_size, while the old size is passed to vmtruncate.This bug was caught by LTP truncate test case on Blackfin platform.
After it was fixed, the LTP truncate test case passed.Signed-off-by: Bryan Wu
Cc: David Howells
Signed-off-by: Linus Torvalds -
The current code is leaking a reference to dreq->kref when the calls to
nfs_direct_read_schedule() and nfs_direct_write_schedule() return an
error.
This patch moves the call to kref_put() from nfs_direct_wait() back into
nfs_direct_read() and nfs_direct_write() (which are the functions that
actually took the reference in the first place) fixing the leak.Thanks to Denis V. Lunev for spotting the bug and proposing the original
fix.Acked-by: Denis V. Lunev
Acked-by: Chuck Lever
Signed-off-by: Trond Myklebust
29 May, 2007
1 commit
-
The recent fix for preventing NULL files from being left around does not
update the file size corectly in all cases. The missing case is a write
extending the file that does not need to allocate a block.In that case we used a read mapping of the extent which forced the use of
the read I/O completion handler instead of the write I/O completion
handle. Hence the file size was not updated on I/O completion.SGI-PV: 965068
SGI-Modid: xfs-linux-melb:xfs-kern:28657aSigned-off-by: David Chinner
Signed-off-by: Nathan Scott
Signed-off-by: Tim Shimmin
26 May, 2007
6 commits
-
Why is it that since the 2f1a2ccb9c0de632ab07193becf5f7121794f6ae console
UTF-8 fixes went into 2.6.22-rc1, the PowerMac G5 shows only inverse video
question marks for the text on tty2-6? whereas tty1 is fine, and so is x86.No fault of that patch: by removing the old fallback behaviour, it reveals
that 32-bit setfont running on 64-bit kernels has only really worked on
the current console, the rest getting faked by that inadequate fallback.Bring the compat do_unimap_ioctl into line with the main one: PIO_UNIMAP
and GIO_UNIMAP apply to the specified tty, not redirected to fg_console.
Use the same checks, and most particularly, remember to check access_ok:
con_set_unimap and con_get_unimap are using __get_user and __put_user.And the compat vt_check should ask for the same capability as the main
one, CAP_SYS_TTY_CONFIG rather than CAP_SYS_ADMIN. Added in vt_ioctl's
vc_cons_allocated check for safety, though failure may well be impossible.Signed-off-by: Hugh Dickins
Signed-off-by: Linus Torvalds -
Signed-off-by: Christoph Hellwig
Signed-off-by: Mark Fasheh -
We weren't cleaning up our inode reference on error in
ocfs2_reserve_local_alloc_bits(). Add a check for error return and iput() if
need be. Move the code to set the alloc context inode info to the end of the
function so we don't have any possibility of passing back a bad pointer.Signed-off-by: Mark Fasheh
-
Use zero_user_page() instead of open-coding it.
Signed-off-by: Nate Diller
Signed-off-by: Andrew Morton
Signed-off-by: Mark Fasheh -
We weren't calling this before, but since ocfs2 handles the entire truncate
operation, we should.Signed-off-by: Mark Fasheh
-
Similarly to the page lock / cluster lock inversion in ocfs2_readpage, we
can deadlock on ip_alloc_sem. We can down_read_trylock() instead and just
return AOP_TRUNCATED_PAGE if the operation fails.Signed-off-by: Mark Fasheh
25 May, 2007
1 commit
-
* 'fixes' of git://git.linux-nfs.org/pub/linux/nfs-2.6:
NFS: Fix nfs_direct_dirty_pages()
NFS: Fix handful of compiler warnings in direct.c
NFS: Avoid a deadlock situation on write
24 May, 2007
11 commits
-
We only need to dirty the pages that were actually read in.
Also convert nfs_direct_dirty_pages() to call set_page_dirty() instead of
set_page_dirty_lock(). A call to lock_page() is unacceptable in an rpciod
callback function.Signed-off-by: Trond Myklebust
-
This patch fixes a couple of signage issues that were causing an Oops
when running the LTP diotest4 test. get_user_pages() returns a signed
error, hence we need to be careful when comparing with the unsigned
number of pages from data->npages.Signed-off-by: Chuck Lever
Signed-off-by: Trond Myklebust -
When processes are allowed to attempt to lock a non-contiguous range of nfs
write requests, it is possible for generic_writepages to 'wrap round' the
address space, and call writepage() on a request that is already locked by
the same process.We avoid the deadlock by checking if the page index is contiguous with the
list of nfs write requests that is already held in our
nfs_pageio_descriptor prior to attempting to lock a new request.Signed-off-by: Trond Myklebust
-
Delay writing 0's out in eCryptfs after a seek past the end of the file
until data is actually written.http://www.opengroup.org/onlinepubs/009695399/functions/lseek.html
``The lseek() function shall not, by itself, extend the size of a
file.''Without this fix, applications that lseek() past the end of the file without
writing will experience unexpected behavior.Signed-off-by: Michael Halcrow
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Gathering signals in bulk enables server applications to drain a signal
queue (almost full of realtime signals) more efficiently by reducing the
syscall and file look-up overhead.Very similar to the sigtimedwait4() call described by Niels Provos, Chuck
Lever, and Stephen Tweedie in a paper entitled "Analyzing the Overload
Behavior of a Simple Web Server". The paper lists more details and
advantages.Signed-off-by: Davi E. M. Arnaut
Acked-by: Davide Libenzi
Cc: Oleg Nesterov
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
When inode is dropped (no more references) delete it from cache.
There's not much point in keeping it cached, when a new lookup will refresh
the attributes anyway.Signed-off-by: Miklos Szeredi
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
This fixes O_APPEND in direct IO mode. Also checks writes against file size
limits, notably rlimits.Reported by Greg Bruno.
Signed-off-by: Miklos Szeredi
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
We don't allow loading ELF shared library from noexec points so the
same should apply to sys_uselib aswell.Signed-off-by: Christoph Hellwig
Cc: Ulrich Drepper
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
In stree.c, MIN_KEY is declared const. The extern declaration in dir.c
doesn't match...Signed-off-by: David Woodhouse
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Optimize select by a using stack space for small fd sets.
core_sys_select() already has this optimization. This is for compat
version.Signed-off-by: Badari Pulavarty
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
The wrong lookup flag was tested in ->create() causing havoc (error or
Oops) when a regular file was created with mknod() in a fuse filesystem.Thanks to J. Cameijo Cerdeira for the report.
Kernels 2.6.18 onward are affected. Please apply to -stable as well.
Signed-off-by: Miklos Szeredi
Cc:
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
23 May, 2007
1 commit
-
* master.kernel.org:/pub/scm/linux/kernel/git/lethal/sh-2.6:
sh: Fix dreamcast build for IRQ changes.
sh: Fix clock multiplier on SH7722.
sh: Wire up kdump crash kernel exec in die().
sh: sr.bl toggling around idle sleep.
sh: disable genrtc support.
fs: Kill sh dependency for binfmt_flat.
sh: Disable psw support for R7785RP.
sh: Fix page size alignment in __copy_user_page().
sh: Fix up various compile warnings for SE boards.
sh: Wire up signalfd/timerfd/eventfd syscalls.
sh: revert addition of page fault notifiers
spelling fixes: arch/sh/
input: hp680_ts compile fixes.
sh: landisk: Header cleanups.
sh: landisk: rtc-rs5c313 support.
sh: Kill off pmb slab cache destructor.
sh: Fix up psw build rules for r7780rp.
sh: Shut up compiler warnings in __do_page_fault().
22 May, 2007
4 commits
-
This from a "tested" patch...
Signed-off-by: Jeff Garzik
Cc: Anton Altaparmakov
Signed-off-by: Linus Torvalds -
This fixes the LDM driver so that it works with Windows Vista dynamic
disks which are subtly different to Windows 2000/XP ones.The patch was needed to get a Vista formatted dynamic disk to be
recognized and parsed successfully.Thanks go to Chris Teachworth for the report and testing.
Cc: Richard Russon
Signed-off-by: Anton Altaparmakov
Signed-off-by: Linus Torvalds -
First thing mm.h does is including sched.h solely for can_do_mlock() inline
function which has "current" dereference inside. By dealing with can_do_mlock()
mm.h can be detached from sched.h which is good. See below, why.This patch
a) removes unconditional inclusion of sched.h from mm.h
b) makes can_do_mlock() normal function in mm/mlock.c
c) exports can_do_mlock() to not break compilation
d) adds sched.h inclusions back to files that were getting it indirectly.
e) adds less bloated headers to some files (asm/signal.h, jiffies.h) that were
getting them indirectlyNet result is:
a) mm.h users would get less code to open, read, preprocess, parse, ... if
they don't need sched.h
b) sched.h stops being dependency for significant number of files:
on x86_64 allmodconfig touching sched.h results in recompile of 4083 files,
after patch it's only 3744 (-8.3%).Cross-compile tested on
all arm defconfigs, all mips defconfigs, all powerpc defconfigs,
alpha alpha-up
arm
i386 i386-up i386-defconfig i386-allnoconfig
ia64 ia64-up
m68k
mips
parisc parisc-up
powerpc powerpc-up
s390 s390-up
sparc sparc-up
sparc64 sparc64-up
um-x86_64
x86_64 x86_64-up x86_64-defconfig x86_64-allnoconfigas well as my two usual configs.
Signed-off-by: Alexey Dobriyan
Signed-off-by: Linus Torvalds -
The bug was introduced by 01f2705daf5a36208e69d7cf95db9c330f843af6.
It misses to convert the first argument, it should be "new_page".This became a cause of fatfs corruption.
Cc: Nate Diller
Signed-off-by: OGAWA Hirofumi
Signed-off-by: Linus Torvalds
21 May, 2007
1 commit
-
Not really sure where this bogosity came from, but there's certainly
nothing special about sh that lets us use flat files with the MMU on.Kill the dependency, and leave it as !MMU, like it is for all of the
other nommu-wielding ports.Signed-off-by: Paul Mundt
19 May, 2007
2 commits
-
The timerfd was using the unlocked waitqueue operations, but it was
using a different lock, so poll_wait() would race with it.This makes timerfd directly use the waitqueue lock.
Signed-off-by: Davide Libenzi
Signed-off-by: Linus Torvalds -
The eventfd was using the unlocked waitqueue operations, but it was
using a different lock, so poll_wait() would race with it.This makes eventfd directly use the waitqueue lock.
Signed-off-by: Davide Libenzi
Signed-off-by: Linus Torvalds
17 May, 2007
10 commits
-
grow_dev_page() simply passes GFP_NOFS to find_or_create_page. This means
the allocation of radix tree nodes is done with GFP_NOFS and the allocation
of a new page is done using GFP_NOFS.The mapping has a flags field that contains the necessary allocation flags
for the page cache allocation. These need to be consulted in order to get
DMA and HIGHMEM allocations etc right. And yes a blockdev could be
allowing Highmem allocations if its a ramdisk.Cc: Hugh Dickins
Signed-off-by: Christoph Lameter
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
i_mutex on quota files is special. Unlike i_mutexes for other inodes it is
acquired under dqonoff_mutex. Tell lockdep about this lock ranking. Also
comment and code in quota_sync_sb() seem to be bogus (as i_mutex for quota
file can be acquired under dqonoff_mutex). Move truncate_inode_pages()
call under dqonoff_mutex and save some problems with races...Signed-off-by: Jan Kara
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Use zero_user_page() instead of open-coding it.
Signed-off-by: Nate Diller
Cc: Michael Halcrow
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Make sysctl/kernel/core_pattern and fs/exec.c agree on maximum core
filename size and change it to 128, so that extensive patterns such as
'/local/cores/%e-%h-%s-%t-%p.core' won't result in truncated filename
generation.Signed-off-by: Dan Aloni
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Just thought this is easier to read.
Acked-by: Davide Libenzi
Signed-off-by: Heiko Carstens
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
SLAB_CTOR_CONSTRUCTOR is always specified. No point in checking it.
Signed-off-by: Christoph Lameter
Cc: David Howells
Cc: Jens Axboe
Cc: Steven French
Cc: Michael Halcrow
Cc: OGAWA Hirofumi
Cc: Miklos Szeredi
Cc: Steven Whitehouse
Cc: Roman Zippel
Cc: David Woodhouse
Cc: Dave Kleikamp
Cc: Trond Myklebust
Cc: "J. Bruce Fields"
Cc: Anton Altaparmakov
Cc: Mark Fasheh
Cc: Paul Mackerras
Cc: Christoph Hellwig
Cc: Jan Kara
Cc: David Chinner
Cc: "David S. Miller"
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
afs_prepare_write() should not mark a page up to date if it only partially
fills it in, in expectation of the caller filling in the rest prior to calling
commit_write(). commit_write(), however, should mark the page up to date.Signed-off-by: David Howells
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Fix AFS to write back dirty on unmounting. This didn't happen because
afs_super_ops.drop_inode was pointing to generic_delete_inode. Now this
pointer is left set to NULL so that the default behaviour occurs instead.Signed-off-by: David Howells
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
16 May, 2007
1 commit