Commit 7ebb9315 (NFS: use secinfo when crossing mountpoints) introduces
a regression when decoding an NFSv4 readdir entry that sets the
rdattr_error field.
By treating the resulting value as if it is a decoding error, the current
code may cause us to skip valid readdir entries.
Reported-by: Andy Adamson <andros@netapp.com>
Cc: stable@kernel.org [2.6.39]
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
AFS: Use i_generation not i_version for the vnode uniquifier
AFS: Set s_id in the superblock to the volume name
vfs: Fix data corruption after failed write in __block_write_begin()
afs: afs_fill_page reads too much, or wrong data
VFS: Fix vfsmount overput on simultaneous automount
fix wrong iput on d_inode introduced by e6bc45d65d
Delay struct net freeing while there's a sysfs instance refering to it
afs: fix sget() races, close leak on umount
ubifs: fix sget races
ubifs: split allocation of ubifs_info into a separate function
fix leak in proc_set_super()
There's no reason not to support cache flushing on external log devices.
The only thing this really requires is flushing the data device first
both in fsync and log commits. A side effect is that we also have to
remove the barrier write test during mount, which has been superflous
since the new FLUSH+FUA code anyway. Also use the chance to flush the
RT subvolume write cache before the fsync commit, which is required
for correct semantics.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>
Store the AFS vnode uniquifier in the i_generation field, not the i_version
field of the inode struct. i_version can then be given the AFS data version
number.
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Set s_id in the superblock to the name of the AFS volume that this superblock
corresponds to.
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
I've got a report of a file corruption from fsxlinux on ext3. The important
operations to the page were:
mapwrite to a hole
partial write to the page
read - found the page zeroed from the end of the normal write
The culprit seems to be that if get_block() fails in __block_write_begin()
(e.g. transient ENOSPC in ext3), the function does ClearPageUptodate(page).
Thus when we retry the write, the logic in __block_write_begin() thinks zeroing
of the page is needed and overwrites old data. In fact, I don't see why we
should ever need to zero the uptodate bit here - either the page was uptodate
when we entered __block_write_begin() and it should stay so when we leave it,
or it was not uptodate and noone had right to set it uptodate during
__block_write_begin() so it remains !uptodate when we leave as well. So just
remove clearing of the bit.
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
afs_fill_page should read the page that is about to be written but
the current implementation has a number of issues. If we aren't
extending the file we always read PAGE_CACHE_SIZE at offset 0. If we
are extending the file we try to read the entire file.
Change afs_fill_page to read PAGE_CACHE_SIZE at the right offset,
clamped to i_size.
While here, avoid calling afs_fill_page when we are doing a
PAGE_CACHE_SIZE write.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
[Kudos to dhowells for tracking that crap down]
If two processes attempt to cause automounting on the same mountpoint at the
same time, the vfsmount holding the mountpoint will be left with one too few
references on it, causing a BUG when the kernel tries to clean up.
The problem is that lock_mount() drops the caller's reference to the
mountpoint's vfsmount in the case where it finds something already mounted on
the mountpoint as it transits to the mounted filesystem and replaces path->mnt
with the new mountpoint vfsmount.
During a pathwalk, however, we don't take a reference on the vfsmount if it is
the same as the one in the nameidata struct, but do_add_mount() doesn't know
this.
The fix is to make sure we have a ref on the vfsmount of the mountpoint before
calling do_add_mount(). However, if lock_mount() doesn't transit, we're then
left with an extra ref on the mountpoint vfsmount which needs releasing.
We can handle that in follow_managed() by not making assumptions about what
we can and what we cannot get from lookup_mnt() as the current code does.
The callers of follow_managed() expect that reference to path->mnt will be
grabbed iff path->mnt has been changed. follow_managed() and follow_automount()
keep track of whether such reference has been grabbed and assume that it'll
happen in those and only those cases that'll have us return with changed
path->mnt. That assumption is almost correct - it breaks in case of
racing automounts and in even harder to hit race between following a mountpoint
and a couple of mount --move. The thing is, we don't need to make that
assumption at all - after the end of loop in follow_manage() we can check
if path->mnt has ended up unchanged and do mntput() if needed.
The BUG can be reproduced with the following test program:
#include <stdio.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <unistd.h>
#include <sys/wait.h>
int main(int argc, char **argv)
{
int pid, ws;
struct stat buf;
pid = fork();
stat(argv[1], &buf);
if (pid > 0) wait(&ws);
return 0;
}
and the following procedure:
(1) Mount an NFS volume that on the server has something else mounted on a
subdirectory. For instance, I can mount / from my server:
mount warthog:/ /mnt -t nfs4 -r
On the server /data has another filesystem mounted on it, so NFS will see
a change in FSID as it walks down the path, and will mark /mnt/data as
being a mountpoint. This will cause the automount code to be triggered.
!!! Do not look inside the mounted fs at this point !!!
(2) Run the above program on a file within the submount to generate two
simultaneous automount requests:
/tmp/forkstat /mnt/data/testfile
(3) Unmount the automounted submount:
umount /mnt/data
(4) Unmount the original mount:
umount /mnt
At this point the kernel should throw a BUG with something like the
following:
BUG: Dentry ffff880032e3c5c0{i=2,n=} still in use (1) [unmount of nfs4 0:12]
Note that the bug appears on the root dentry of the original mount, not the
mountpoint and not the submount because sys_umount() hasn't got to its final
mntput_no_expire() yet, but this isn't so obvious from the call trace:
[<ffffffff8117cd82>] shrink_dcache_for_umount+0x69/0x82
[<ffffffff8116160e>] generic_shutdown_super+0x37/0x15b
[<ffffffffa00fae56>] ? nfs_super_return_all_delegations+0x2e/0x1b1 [nfs]
[<ffffffff811617f3>] kill_anon_super+0x1d/0x7e
[<ffffffffa00d0be1>] nfs4_kill_super+0x60/0xb6 [nfs]
[<ffffffff81161c17>] deactivate_locked_super+0x34/0x83
[<ffffffff811629ff>] deactivate_super+0x6f/0x7b
[<ffffffff81186261>] mntput_no_expire+0x18d/0x199
[<ffffffff811862a8>] mntput+0x3b/0x44
[<ffffffff81186d87>] release_mounts+0xa2/0xbf
[<ffffffff811876af>] sys_umount+0x47a/0x4ba
[<ffffffff8109e1ca>] ? trace_hardirqs_on_caller+0x1fd/0x22f
[<ffffffff816ea86b>] system_call_fastpath+0x16/0x1b
as do_umount() is inlined. However, you can see release_mounts() in there.
Note also that it may be necessary to have multiple CPU cores to be able to
trigger this bug.
Tested-by: Jeff Layton <jlayton@redhat.com>
Tested-by: Ian Kent <raven@themaw.net>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Git bisection shows that commit e6bc45d65d causes
BUG_ONs under high I/O load:
kernel BUG at fs/inode.c:1368!
[ 2862.501007] Call Trace:
[ 2862.501007] [<ffffffff811691d8>] d_kill+0xf8/0x140
[ 2862.501007] [<ffffffff81169c19>] dput+0xc9/0x190
[ 2862.501007] [<ffffffff8115577f>] fput+0x15f/0x210
[ 2862.501007] [<ffffffff81152171>] filp_close+0x61/0x90
[ 2862.501007] [<ffffffff81152251>] sys_close+0xb1/0x110
[ 2862.501007] [<ffffffff814c14fb>] system_call_fastpath+0x16/0x1b
A reliable way to reproduce this bug is:
Login to KDE, run 'rsnapshot sync', and apt-get install openjdk-6-jdk,
and apt-get remove openjdk-6-jdk.
The buggy part of the patch is this:
struct inode *inode = NULL;
.....
- if (nd.last.name[nd.last.len])
- goto slashes;
inode = dentry->d_inode;
- if (inode)
- ihold(inode);
+ if (nd.last.name[nd.last.len] || !inode)
+ goto slashes;
+ ihold(inode)
...
if (inode)
iput(inode); /* truncate the inode here */
If nd.last.name[nd.last.len] is nonzero (and thus goto slashes branch is taken),
and dentry->d_inode is non-NULL, then this code now does an additional iput on
the inode, which is wrong.
Fix this by only setting the inode variable if nd.last.name[nd.last.len] is 0.
Reference: https://lkml.org/lkml/2011/6/15/50
Reported-by: Norbert Preining <preining@logic.at>
Reported-by: Török Edwin <edwintorok@gmail.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Török Edwin <edwintorok@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This reverts commit 7f81c8890c.
It turns out that it's not actually a build-time check on x86-64 UML,
which does some seriously crazy stuff with VM_STACK_FLAGS.
The VM_STACK_FLAGS define depends on the arch-supplied
VM_STACK_DEFAULT_FLAGS value, and on x86-64 UML we have
arch/um/sys-x86_64/shared/sysdep/vm-flags.h:
#define VM_STACK_DEFAULT_FLAGS \
(test_thread_flag(TIF_IA32) ? vm_stack_flags32 : vm_stack_flags)
#define VM_STACK_DEFAULT_FLAGS vm_stack_flags
(yes, seriously: two different #define's for that thing, with the first
one being inside an "#ifdef TIF_IA32")
It's possible that it is UML that should just be fixed in this area, but
for now let's just undo the (very small) optimization.
Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Acked-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Richard Weinberger <richard@nod.at>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit a8bef8ff6e ("mm: migration: avoid race between shift_arg_pages()
and rmap_walk() during migration by not migrating temporary stacks")
introduced a BUG_ON() to ensure that VM_STACK_FLAGS and
VM_STACK_INCOMPLETE_SETUP do not overlap. The check is a compile time
one, so BUILD_BUG_ON is more appropriate.
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Don't call iput with the inode half setup to be a namespace filedescriptor.
Instead rearrange the code so that we don't initialize ei->ns_ops until
after I ns_ops->get succeeds, preventing us from invoking ns_ops->put
when ns_ops->get failed.
Reported-by: Ingo Saitz <Ingo.Saitz@stud.uni-hannover.de>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
We can lockup if we try to allow new writers join the transaction and we have
flushoncommit set or have a pending snapshot. This is because we set
no_trans_join and then loop around and try to wait for ordered extents again.
The problem is the ordered endio stuff needs to join the transaction, which it
can't do because no_trans_join is set. So instead wait until after this loop to
set no_trans_join and then make sure to wait for num_writers == 1 in case
anybody got started in between us exiting the loop and setting no_trans_join.
This could easily be reproduced by mounting -o flushoncommit and running xfstest
13. It cannot be reproduced with this patch. Thanks,
Reported-by: Jim Schutt <jaschut@sandia.gov>
Signed-off-by: Josef Bacik <josef@redhat.com>
Currently there is nothing protecting the pending_snapshots list on the
transaction. We only hold the directory mutex that we are snapshotting and a
read lock on the subvol_sem, so we could race with somebody else creating a
snapshot in a different directory and end up with list corruption. So protect
this list with the trans_lock. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
The delayed ref patch accidently removed the btrfs_free_path in
btrfs_unlink_subvol, this puts it back and means we don't leak a path. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
mark_matching_lsegs_invalid could put the last ref to the layout, so
the get_layout_hdr needs to be called first.
Signed-off-by: Fred Isaman <iisaman@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
We always get a reference on the layout header and we rely on
nfs4_layoutreturn_release to put it. If we hit an allocation error
before starting the rpc proc we bail out early without dereferncing
the layout header properly.
Signed-off-by: Benny Halevy <benny@tonian.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
(d)printks should use %zd for ssize_t arguments not %ld, otherwise they might
get a warning. I see the following with MN10300.
fs/nfs/objlayout/objlayout.c: In function 'objlayout_read_done':
fs/nfs/objlayout/objlayout.c:294: warning: format '%ld' expects type 'long int', but argument 3 has type 'ssize_t'
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Trond Myklebust <Trond.Myklebust@netapp.com>
cc: linux-nfs@vger.kernel.org
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The break condition to skip out of the loop got broken when cmp_layout
was change. Essentially, we want to stop looking once we know no layout
on the remainder of the list can match the first byte of the looked-up
range.
Reported-by: Peng Tao <peng_tao@emc.com>
Signed-off-by: Benny Halevy <benny@tonian.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
_pnfs_return_layout had the following problems:
- it did not call pnfs_free_lseg_list on all paths
- it unintentionally did a forgetful return when there was no outstanding io
- it raced with concurrent LAYOUTGETS
Signed-off-by: Fred Isaman <iisaman@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Commit 28331a46d8 "Ensure we request the
ordinary fileid when doing readdirplus"
changed the meaning of NFS_ATTR_FATTR_FILEID which used to be set when
FATTR4_WORD1_MOUNTED_ON_FILED was requested.
Allow nfs_fhget to succeed with only a mounted on fileid when crossing
a mountpoint or a referral.
Ask for the fileid of the absent file system if mounted_on_fileid is not
supported.
Signed-off-by: Andy Adamson <andros@netapp.com>
cc:stable@kernel.org [2.6.39]
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
When we add something to the global device id cache, we need to bump the
reference count, so that the cache itself holds a reference.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
We don't support header padding yet so better off ditching it
Reported-by: Sid Moore <learnmost@gmail.com>
Signed-off-by: Benny Halevy <benny@tonian.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
nfs_update_inode will update isize if there is no queued pages. For pNFS,
layoutcommit is supposed to change file size on server, the same effect as queued
pages. nfs_update_inode may be called when dirty pages are written back (nfsi->npages==0)
but layoutcommit is not sent, and it will change client file size according to server
file size. Then client ends up losing what it just writes back in pNFS path.
So we should skip updating client file size if file needs layoutcommit.
Signed-off-by: Peng Tao <peng_tao@emc.com>
Cc: stable@kernel.org [2.6.39]
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
If the NLM daemon is killed on the NFS server, we can currently end up
hanging forever on an 'unlock' request, instead of aborting. Basically,
if the rpcbind request fails, or the server keeps returning garbage, we
really want to quit instead of retrying.
Tested-by: Vasily Averin <vvs@sw.ru>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@kernel.org
Unmounting a pnfs filesystem hangs using filelayout and possibly others.
This fixes the use of the rcu protected node by making use of a new 'tmpnode'
for the temporary purge list. Also, the spinlock shouldn't be held when calling
synchronize_rcu().
Signed-off-by: Weston Andros Adamson <dros@netapp.com>
Signed-off-by: Andy Adamson <andros@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
->mknod() should return negative on errors and PTR_ERR() gives
already negative value...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>
Currently processes waiting with poll on cancelable timerfd timers are
not woken up when the timers are canceled. When the system time is set
the clock_was_set() function calls timerfd_clock_was_set() to cancel
and wake up processes waiting on potential cancelable timerfd
timers. However the wake up currently has no effect because in the
case of timerfd_read it is dependent on ctx->ticks not being
0. timerfd_poll also requires ctx->ticks being non zero. As a
consequence processes waiting on cancelable timers only get woken up
when the timers expire. This patch fixes this by incrementing
ctx->ticks before calling wake_up.
Signed-off-by: Max Asbock <masbock@linux.vnet.ibm.com>
Cc: kay.sievers@vrfy.org
Cc: virtuoso@slind.org
Cc: johnstul <johnstul@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/1307985512.4710.41.camel@w-amax.beaverton.ibm.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
We were iterating across stripe boundaries properly, but not moving the
write buffer pointer forward. This caused us to rewrite the same data
after the break. Fix by adjusting the data pointer forward, and
recalculating the io and buffer alignment after the break.
Signed-off-by: Sage Weil <sage@newdream.net>
Long ago (in commit 00e485b0), I added some code to handle share-level
passwords in CIFSTCon. That code ignored the fact that it's legit to
pass in a NULL tcon pointer when connecting to the IPC$ share on the
server.
This wasn't really a problem until recently as we only called CIFSTCon
this way when the server returned -EREMOTE. With the introduction of
commit c1508ca2 however, it gets called this way on every mount, causing
an oops when share-level security is in effect.
Fix this by simply treating a NULL tcon pointer as if user-level
security were in effect. I'm not aware of any servers that protect the
IPC$ share with a specific password anyway. Also, add a comment to the
top of CIFSTCon to ensure that we don't make the same mistake again.
Cc: <stable@kernel.org>
Reported-by: Martijn Uffing <mp3project@sarijopen.student.utwente.nl>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
It's possible for the following set of events to happen:
cifsd calls cifs_reconnect which reconnects the socket. A userspace
process then calls cifs_negotiate_protocol to handle the NEGOTIATE and
gets a reply. But, while processing the reply, cifsd calls
cifs_reconnect again. Eventually the GlobalMid_Lock is dropped and the
reply from the earlier NEGOTIATE completes and the tcpStatus is set to
CifsGood. cifs_reconnect then goes through and closes the socket and sets the
pointer to zero, but because the status is now CifsGood, the new socket
is not created and cifs_reconnect exits with the socket pointer set to
NULL.
Fix this by only setting the tcpStatus to CifsGood if the tcpStatus is
CifsNeedNegotiate, and by making sure that generic_ip_connect is always
called at least once in cifs_reconnect.
Note that this is not a perfect fix for this issue. It's still possible
that the NEGOTIATE reply is handled after the socket has been closed and
reconnected. In that case, the socket state will look correct but it no
NEGOTIATE was performed on it be for the wrong socket. In that situation
though the server should just shut down the socket on the next attempted
send, rather than causing the oops that occurs today.
Cc: <stable@kernel.org> # .38.x: fd88ce9: [CIFS] cifs: clarify the meaning of tcpStatus == CifsGood
Reported-and-Tested-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
cifs_sb_master_tlink was declared as inline, but without a definition.
Remove the declaration and move the definition up.
Signed-off-by: Pavel Shilovsky <piastryyy@gmail.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
jbd2_journal_remove_journal_head() can oops when trying to access
journal_head returned by bh2jh(). This is caused for example by the
following race:
TASK1 TASK2
jbd2_journal_commit_transaction()
...
processing t_forget list
__jbd2_journal_refile_buffer(jh);
if (!jh->b_transaction) {
jbd_unlock_bh_state(bh);
jbd2_journal_try_to_free_buffers()
jbd2_journal_grab_journal_head(bh)
jbd_lock_bh_state(bh)
__journal_try_to_free_buffer()
jbd2_journal_put_journal_head(jh)
jbd2_journal_remove_journal_head(bh);
jbd2_journal_put_journal_head() in TASK2 sees that b_jcount == 0 and
buffer is not part of any transaction and thus frees journal_head
before TASK1 gets to doing so. Note that even buffer_head can be
released by try_to_free_buffers() after
jbd2_journal_put_journal_head() which adds even larger opportunity for
oops (but I didn't see this happen in reality).
Fix the problem by making transactions hold their own journal_head
reference (in b_jcount). That way we don't have to remove journal_head
explicitely via jbd2_journal_remove_journal_head() and instead just
remove journal_head when b_jcount drops to zero. The result of this is
that [__]jbd2_journal_refile_buffer(),
[__]jbd2_journal_unfile_buffer(), and
__jdb2_journal_remove_checkpoint() can free journal_head which needs
modification of a few callers. Also we have to be careful because once
journal_head is removed, buffer_head might be freed as well. So we
have to get our own buffer_head reference where it matters.
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
ceph: unwind canceled flock state
ceph: fix ENOENT logic in striped_read
ceph: fix short sync reads from the OSD
ceph: fix sync vs canceled write
ceph: use ihold when we already have an inode ref
6b4517a791 (block: implement bd_claiming and claiming block)
introduced claiming block to support O_EXCL blkdev opens properly.
bd_start_claiming() looks up the part 0 bdev and starts claiming
block. The function assumed that there is only one part 0 bdev and
always used bdget_disk(disk, 0) to look it up; unfortunately, this
isn't true for some drivers (floppy) which use multiple block devices
to denote different operating parameters for the same physical device.
There can be multiple part 0 bdev's for the same device number.
This incorrect assumption caused the wrong bdev to be used during
claiming leading to unbalanced bd_holders as reported in the following
bug.
https://bugzilla.kernel.org/show_bug.cgi?id=28522
This patch updates bd_start_claiming() such that it uses the bdev
specified as argument if its partno is zero.
Note that this means that different bdev's can be used for the same
device and O_EXCL check can be effectively bypassed. It has always
been broken that way and floppy is fortunately on its way out. Leave
that breakage alone.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Alex Villacis Lasso <avillaci@ceibo.fiec.espol.edu.ec>
Tested-by: Alex Villacis Lasso <avillaci@ceibo.fiec.espol.edu.ec>
Cc: stable@kernel.org # >= v2.6.36
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
credits isn't a parameter for jbd2_journal_get_write_access and
jbd2_journal_get_undo_access. So remove the corresponding comments.
Acked-by: Jan Kara <jack@suse.cz>
Cc: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Tao Ma <boyu.mt@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
* new refcount in struct net, controlling actual freeing of the memory
* new method in kobj_ns_type_operations (->drop_ns())
* ->current_ns() semantics change - it's supposed to be followed by
corresponding ->drop_ns(). For struct net in case of CONFIG_NET_NS it bumps
the new refcount; net_drop_ns() decrements it and calls net_free() if the
last reference has been dropped. Method renamed to ->grab_current_ns().
* old net_free() callers call net_drop_ns() instead.
* sysfs_exit_ns() is gone, along with a large part of callchain
leading to it; now that the references stored in ->ns[...] stay valid we
do not need to hunt them down and replace them with NULL. That fixes
problems in sysfs_lookup() and sysfs_readdir(), along with getting rid
of sb->s_instances abuse.
Note that struct net *shutdown* logics has not changed - net_cleanup()
is called exactly when it used to be called. The only thing postponed by
having a sysfs instance refering to that struct net is actual freeing of
memory occupied by struct net.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
* set ->s_fs_info in set() callback passed to sget()
* allocate the thing and set it up enough for afs_test_super() before
making it visible
* have it freed in ->kill_sb() (current tree simply leaks it)
* have ->put_super() leave ->s_fs_info->volume alone; it's too early for
dropping it; do that from ->kill_sb() after having called kill_anon_super().
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
* allocate ubifs_info in ->mount(), fill it enough for sb_test() and
set ->s_fs_info to it in set() callback passed to sget().
* do *not* free it in ->put_super(); do that in ->kill_sb() after we'd
done kill_anon_super().
* don't free it in ubifs_fill_super() either - deactivate_locked_super()
done by caller when ubifs_fill_super() returns an error will take care
of that sucker.
* get rid of kludge with passing ubi to ubifs_fill_super() in ->s_fs_info;
we only need it in alloc_ubifs_info(), so ubifs_fill_super() will need
only ubifs_info. Which it will find in ->s_fs_info just fine, no need to
reassign anything...
As the result, sb_test() becomes safe to apply to all superblocks that
can be found by sget() (and a kludge with temporary use of ->s_fs_info
to store a pointer to very different structure goes away).
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
Btrfs: use join_transaction in btrfs_evict_inode()
Btrfs - use %pU to print fsid
Btrfs: fix extent state leak on failed nodatasum reads
btrfs: fix unlocked access of delalloc_inodes
Btrfs: avoid stack bloat in btrfs_ioctl_fs_info()
btrfs: remove 64bit alignment padding to allow extent_buffer to fit into one fewer cacheline
Btrfs: clear current->journal_info on async transaction commit
Btrfs: make sure to recheck for bitmaps in clusters
btrfs: remove unneeded includes from scrub.c
btrfs: reinitialize scrub workers
btrfs: scrub: errors in tree enumeration
Btrfs: don't map extent buffer if path->skip_locking is set
Btrfs: unlock the trans lock properly
Btrfs: don't map extent buffer if path->skip_locking is set
Btrfs: fix duplicate checking logic
Btrfs: fix the allocator loop logic
Btrfs: fix bitmap regression
Btrfs: don't commit the transaction if we dont have enough pinned bytes
Btrfs: noinline the cluster searching functions
Btrfs: cache bitmaps when searching for a cluster
The WARN_ON() in start_transaction() was triggered while balancing.
The cause is btrfs_relocate_chunk() started a transaction and
then called iput() on the inode that stores free space cache,
and iput() called btrfs_start_transaction() again.
Reported-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Reviewed-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Checkpoint generation interval of nilfs goes wrong after user has
changed the interval parameter with nilfs-tune tool.
segctord starting. Construction interval = 5 seconds,
CP frequency < 30 seconds
segctord starting. Construction interval = 0 seconds,
CP frequency < 30 seconds
This turned out to be caused by a trivial bug in initialization code
of log writer. This will fix it.
Reported-by: Andrea Gelmini <andrea.gelmini@gmail.com>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
nilfs_btree_delete function does not terminate part of virtual block
addresses when shrinking the last remaining child node into the root
node. The missing address termination causes that dead btree node
blocks persist and chip away free disk space.
This fixes the leak bug on the btree node deletion.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
nilfs_btree_delete function wrongly terminates virtual block address
of the btree node held by its parent at index 0. When concatenating
the index-0 node with its right sibling node, nilfs_btree_delete
terminates the block address of index-0 node instead of the right
sibling node which should be deleted.
This bug not only wears disk space in the long run, but also causes
file system corruption. This will fix it.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Get rid of FIXME comment. Uuids from dmesg are now the same as uuids
given by btrfs-progs.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
When encountering an EIO while reading from a nodatasum extent, we
insert an error record into the inode's failure tree.
btrfs_readpage_end_io_hook returns early for nodatasum inodes. We'd
better clear the failure tree in that case, otherwise the kernel
complains about
BUG extent_state: Objects remaining on kmem_cache_close()
on rmmod.
Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
list_splice_init will make delalloc_inodes empty, but without a spinlock
around, this may produce corrupted list head, accessed in many placess,
The race window is very tight and nobody seems to have hit it so far.
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The size of struct btrfs_ioctl_fs_info_args is as big as 1KB, so
don't declare the variable on stack.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Reviewed-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Reorder extent_buffer to remove 8 bytes of alignment padding on 64 bit
builds. This shrinks its size to 128 bytes allowing it to fit into one
fewer cache lines and allows more objects per slab in its kmem_cache.
slabinfo extent_buffer reports :-
before:-
Sizes (bytes) Slabs
----------------------------------
Object : 136 Total : 123
SlabObj: 136 Full : 121
SlabSiz: 4096 Partial: 0
Loss : 0 CpuSlab: 2
Align : 8 Objects: 30
after :-
Object : 128 Total : 4
SlabObj: 128 Full : 2
SlabSiz: 4096 Partial: 0
Loss : 0 CpuSlab: 2
Align : 8 Objects: 32
Signed-off-by: Richard Kennedy <richard@rsk.demon.co.uk>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Normally current->jouranl_info is cleared by commit_transaction. For an
async snap or subvol creation, though, it runs in a work queue. Clear
it in btrfs_commit_transaction_async() to avoid leaking a non-NULL
journal_info when we return to userspace. When the actual commit runs in
the other thread it won't care that it's current->journal_info is already
NULL.
Signed-off-by: Sage Weil <sage@newdream.net>
Tested-by: Jim Schutt <jaschut@sandia.gov>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Josef recently changed the free extent cache to look in
the block group cluster for any bitmaps before trying to
add a new bitmap for the same offset. This avoids BUG_ON()s due
covering duplicate ranges.
But it didn't go quite far enough. A given free range might span
between one or more bitmaps or free space entries. The code has
looping to cover this, but it doesn't check for clustered bitmaps
every time.
This shuffles our gotos to check for a bitmap in the cluster
for every new bitmap entry we try to add.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Scrub starts the workers each time a scrub starts and stops them after it
finished. This patch adds an initialization for the workers before each
start, otherwise the workers behave strangely.
Signed-off-by: Arne Jansen <sensille@gmx.net>
due to the semantics of btrfs_search_slot the path can point to an
invalid slot when ret > 0. This condition went unnoticed, which in
turn could have led to an incomplete scrubbing.
Signed-off-by: Arne Jansen <sensille@gmx.net>
Arne's scrub stuff exposed a problem with mapping the extent buffer in
reada_for_search. He searches the commit root with multiple threads and with
skip_locking set, so we can race and overwrite node->map_token since node isn't
locked. So fix this so that we only map the extent buffer if we don't already
have a map_token and skip_locking isn't set. Without this patch scrub would
panic almost immediately, with the patch it doesn't panic anymore. Thanks,
Reported-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Josef Bacik <josef@redhat.com>
Unconditionally changing the address limit to USER_DS and not restoring
it to its old value in the error path is wrong because it prevents us
using kernel memory on repeated calls to this function. This, in fact,
breaks the fallback of hard coded paths to the init program from being
ever successful if the first candidate fails to load.
With this patch applied switching to USER_DS is delayed until the point
of no return is reached which makes it possible to have a multi-arch
rootfs with one arch specific init binary for each of the (hard coded)
probed paths.
Since the address limit is already set to USER_DS when start_thread()
will be invoked, this redundancy can be safely removed.
Signed-off-by: Mathias Krause <minipli@googlemail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In btrfs_wait_for_commit if we came upon a transaction that had committed we
just exited, but that's bad since we are holding the trans_lock. So break
instead so that the lock is dropped. Thanks,
Reported-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Josef Bacik <josef@redhat.com>
Arne's scrub stuff exposed a problem with mapping the extent buffer in
reada_for_search. He searches the commit root with multiple threads and with
skip_locking set, so we can race and overwrite node->map_token since node isn't
locked. So fix this so that we only map the extent buffer if we don't already
have a map_token and skip_locking isn't set. Without this patch scrub would
panic almost immediately, with the patch it doesn't panic anymore. Thanks,
Reported-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Josef Bacik <josef@redhat.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6:
cifs: trivial: add space in fsc error message
cifs: silence printk when establishing first session on socket
CIFS ACL support needs CONFIG_KEYS, so depend on it
possible memory corruption in cifs_parse_mount_options()
cifs: make CIFS depend on CRYPTO_ECB
cifs: fix the kernel release version in the default security warning message
When merging my code into the integration test the second check for duplicate
entries got screwed up. This patch fixes it by dropping ret2 and just using ret
for the return value, and checking if we got an error before adding the bitmap
to the local list. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
I was testing with empty_cluster = 0 to try and reproduce a problem and kept
hitting early enospc panics. This was because our loop logic was a little
confused. So this is what I did
1) Make the loop variable the ultimate decider on wether we should loop again
isntead of checking to see if we had an uncached bg, empty size or empty
cluster.
2) Increment loop before checking to see what we are on to make the loop
definitions make more sense.
3) If we are on the chunk alloc loop don't set empty_size/empty_cluster to 0
unless we didn't actually allocate a chunk. If we did allocate a chunk we
should be able to easily setup a new cluster so clearing
empty_size/empty_cluster makes us less efficient.
This kept me from hitting panics while trying to reproduce the other problem.
Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
In cleaning up the clustering code I accidently introduced a regression by
adding bitmap entries to the cluster rb tree. The problem is if we've maxed out
the number of bitmaps we can have for the block group we can only add free space
to the bitmaps, but since the bitmap is on the cluster we can't find it and we
try to create another one. This would result in a panic because the total
bitmaps was bigger than the max bitmaps that were allowed. This patch fixes
this by checking to see if we have a cluster, and then looking at the cluster rb
tree to see if it has a bitmap entry and if it does and that space belongs to
that bitmap, go ahead and add it to that bitmap.
I could hit this panic every time with an fs_mark test within a couple of
minutes. With this patch I no longer hit the panic and fs_mark goes to
completion. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
I noticed when running an enospc test that we would get stuck committing the
transaction in check_data_space even though we truly didn't have enough space.
So check to see if bytes_pinned is bigger than num_bytes, if it's not don't
commit the transaction. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
When profiling the find cluster code it's hard to tell where we are spending our
time because the bitmap and non-bitmap functions get inlined by the compiler, so
make that not happen. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
If we are looking for a cluster in a particularly sparse or fragmented block
group, we will do a lot of looping through the free space tree looking for
various things, and if we need to look at bitmaps we will endup doing the whole
dance twice. So instead add the bitmap entries to a temporary list so if we
have to do the bitmap search we can just look through the list of entries we've
found quickly instead of having to loop through the entire tree again. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
If we request a lock and then abort (e.g., ^C), we need to send a matching
unlock request to the MDS to unwind our lock attempt to avoid indefinitely
blocking other clients.
Reported-by: Brian Chrisman <brchrisman@gmail.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Getting ENOENT is equivalent to reading 0 bytes. Make that correction
before setting up the hit_stripe and was_short flags.
Fixes the following case:
dd if=/dev/zero of=/mnt/fs_depot/dd3 bs=1 seek=1048576 count=0
dd if=/mnt/fs_depot/dd3 of=/root/ddout1 skip=8 bs=500 count=2 iflag=direct
Reported-by: Henry C Chang <henry.cy.chang@gmail.com>
Signed-off-by: Sage Weil <sage@newdream.net>
If we get a short read from the OSD because the object is small, we need to
zero the remainder of the buffer. For O_DIRECT reads, the attempted range
is not trimmed to i_size by the VFS, so we were actually looping
indefinitely.
Fix by trimming by i_size, and the unconditionally zeroing the trailing
range.
Reported-by: Jeff Wu <cpwu@tnsoft.com.cn>
Signed-off-by: Sage Weil <sage@newdream.net>
We should use ihold whenever we already have a stable inode ref, even
when we aren't holding i_lock. This avoids adding new and unnecessary
locking dependencies.
Signed-off-by: Sage Weil <sage@newdream.net>
* 'for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
vfs: make unlink() and rmdir() return ENOENT in preference to EROFS
lmLogOpen() broken failure exit
usb: remove bad dput after dentry_unhash
more conservative S_NOSEC handling
If user space attempts to remove a non-existent file or directory, and
the file system is mounted read-only, return ENOENT instead of EROFS.
Either error code is arguably valid/correct, but ENOENT is a more
specific error message.
Reported-by: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Callers of lmLogOpen() expect it to return -E... on failure exits, which
is what it returns, except for the case of blkdev_get_by_dev() failure.
It that case lmLogOpen() return the error with the wrong sign...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com>
When signing is enabled, the first session that's established on a
socket will cause a printk like this to pop:
CIFS VFS: Unexpected SMB signature
This is because the key exchange hasn't happened yet, so the signature
field is bogus. Don't try to check the signature on the socket until the
first session has been established. Also, eliminate the specific check
for SMB_COM_NEGOTIATE since this check covers that case too.
Cc: stable@kernel.org
Cc: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
fix for commit 4795bb37ef, nfsd: break
lease on unlink, link, and rename
if the LINK operation breaks a delegation, it returns NFS4ERR_NOENT
(which is not a valid error in rfc 5661) instead of NFS4ERR_DELAY.
the return value of nfsd_break_lease() in nfsd_link() must be
converted from host_err to err
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
Cc: stable@kernel.org
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
nfsd V4 support uses crypto interfaces, so select CRYPTO
to fix build errors in 2.6.39:
ERROR: "crypto_destroy_tfm" [fs/nfsd/nfsd.ko] undefined!
ERROR: "crypto_alloc_base" [fs/nfsd/nfsd.ko] undefined!
Reported-by: Wakko Warner <wakko@animx.eu.org>
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: stable@kernel.org
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Commit b0b0c0a26e "nfsd: add proc file listing kernel's gss_krb5
enctypes" added an nunnecessary dependency of nfsd on the auth_rpcgss
module.
It's a little ad hoc, but since the only piece of information nfsd needs
from rpcsec_gss_krb5 is a single static string, one solution is just to
share it with an include file.
Cc: stable@kernel.org
Reported-by: Michael Guntsche <mike@it-loops.com>
Cc: Kevin Coffman <kwc@citi.umich.edu>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Build fails if CONFIG_KEYS is not selected.
Signed-off-by: Darren Salt <linux@youmustbejoking.demon.co.uk>
Reviewed-by: Shirish Pargaonkar <shirishp@us.ibm.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
While creating fixed tracepoints for ext3, basically by porting them
from ext4, I found a lot of useless retyping, wrong type usage, useless
variable passing and other inconsistencies in the ext4 fixed tracepoint
code.
This patch cleans the fixed tracepoint code for ext4 and also simplify
some of them.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Currently we are not marking the extent as the last one
(FIEMAP_EXTENT_LAST) if there is a hole at the end of the file. This is
because we just do not check for it right now and continue searching for
next extent. But at the point we hit the hole at the end of the file, it
is too late.
This commit adds check for the allocated block in subsequent extent and
if there is no more extents (block = EXT_MAX_BLOCKS) just flag the
current one as the last one.
This behaviour has been spotted unintentionally by 252 xfstest, when the
test hangs out, because of wrong loop condition. However on other
filesystems (like xfs) it will exit anyway, because we notice the last
extent flag and exit.
With this patch xfstest 252 does not hang anymore, ext4 fiemap
implementation still reports bad extent type in some cases, however
this seems to be different issue.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Kazuya Mio reported that he was able to hit BUG_ON(next == lblock)
in ext4_ext_put_gap_in_cache() while creating a sparse file in extent
format and fill the tail of file up to its end. We will hit the BUG_ON
when we write the last block (2^32-1) into the sparse file.
The root cause of the problem lies in the fact that we specifically set
s_maxbytes so that block at s_maxbytes fit into on-disk extent format,
which is 32 bit long. However, we are not storing start and end block
number, but rather start block number and length in blocks. It means
that in order to cover extent from 0 to EXT_MAX_BLOCK we need
EXT_MAX_BLOCK+1 to fit into len (because we counting block 0 as well) -
and it does not.
The only way to fix it without changing the meaning of the struct
ext4_extent members is, as Kazuya Mio suggested, to lower s_maxbytes
by one fs block so we can cover the whole extent we can get by the
on-disk extent format.
Also in many places EXT_MAX_BLOCK is used as length instead of maximum
logical block number as the name suggests, it is all a bit messy. So
this commit renames it to EXT_MAX_BLOCKS and change its usage in some
places to actually be maximum number of blocks in the extent.
The bug which this commit fixes can be reproduced as follows:
dd if=/dev/zero of=/mnt/mp1/file bs=<blocksize> count=1 seek=$((2**32-2))
sync
dd if=/dev/zero of=/mnt/mp1/file bs=<blocksize> count=1 seek=$((2**32-1))
Reported-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
metadata is not parameter of ext4_free_blocks() any more.
Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable: (25 commits)
btrfs: fix uninitialized variable warning
btrfs: add helper for fs_info->closing
Btrfs: add mount -o inode_cache
btrfs: scrub: add explicit plugging
btrfs: use btrfs_ino to access inode number
Btrfs: don't save the inode cache if we are deleting this root
btrfs: false BUG_ON when degraded
Btrfs: don't save the inode cache in non-FS roots
Btrfs: make sure we don't overflow the free space cache crc page
Btrfs: fix uninit variable in the delayed inode code
btrfs: scrub: don't reuse bios and pages
Btrfs: leave spinning on lookup and map the leaf
Btrfs: check for duplicate entries in the free space cache
Btrfs: don't try to allocate from a block group that doesn't have enough space
Btrfs: don't always do readahead
Btrfs: try not to sleep as much when doing slow caching
Btrfs: kill BTRFS_I(inode)->block_group
Btrfs: don't look at the extent buffer level 3 times in a row
Btrfs: map the node block when looking for readahead targets
Btrfs: set range_start to the right start in count_range_bits
...
With Linus' tree, today's linux-next build (powercp ppc64_defconfig)
produced this warning:
fs/btrfs/delayed-inode.c: In function 'btrfs_delayed_update_inode':
fs/btrfs/delayed-inode.c:1598:6: warning: 'ret' may be used
uninitialized in this function
Introduced by commit 16cdcec736 ("btrfs: implement delayed inode items
operation").
This fixes a bug in btrfs_update_inode(): if the returned value from
btrfs_delayed_update_inode is a nonzero garbage, inode stat data are not
updated and several call paths may hit a BUG_ON or fail with strange
code.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: David Sterba <dsterba@suse.cz>
This makes the inode map cache default to off until we
fix the overflow problem when the free space crcs don't fit
inside a single page.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
With the removal of the implicit plugging scrub ends up doing more and
smaller I/O than necessary. This patch adds explicit plugging per chunk.
Signed-off-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
commit 4cb5300bc ("Btrfs: add mount -o auto_defrag") accesses inode
number directly while it should use the helper with the new inode
number allocator.
Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
With xfstest 254 I can panic the box every time with the inode number caching
stuff on. This is because we clean the inodes out when we delete the subvolume,
but then we write out the inode cache which adds an inode to the subvolume inode
tree, and then when it gets evicted again the root gets added back on the dead
roots list and is deleted again, so we have a double free. To stop this from
happening just return 0 if refs is 0 (and we're not the tree root since tree
root always has refs of 0). With this fix 254 no longer panics. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Tested-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
In degraded mode the struct btrfs_device of missing devs don't have
device->name set. A kstrdup of NULL correctly returns NULL. Don't
BUG in this case.
Signed-off-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This adds extra checks to make sure the inode map we are caching really
belongs to a FS root instead of a special relocation tree. It
prevents crashes during balancing operations.
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The free space cache uses only one page for crcs right now,
which means we can't have a cache file bigger than the
crcs we can fit in the first page. This adds a check to
enforce that restriction.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The current scrub implementation reuses bios and pages as often as possible,
allocating them only on start and releasing them when finished. This leads
to more problems with the block layer than it's worth. The elevator gets
confused when there are more pages added to the bio than bi_size suggests.
This patch completely rips out the reuse of bios and pages and allocates
them freshly for each submit.
Signed-off-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Chris Maosn <chris.mason@oracle.com>
* 'for-linus' of git://git.kernel.dk/linux-block:
block: Use hlist_entry() for io_context.cic_list.first
cfq-iosched: Remove bogus check in queue_fail path
xen/blkback: potential null dereference in error handling
xen/blkback: don't call vbd_size() if bd_disk is NULL
block: blkdev_get() should access ->bd_disk only after success
CFQ: Fix typo and remove unnecessary semicolon
block: remove unwanted semicolons
Revert "block: Remove extra discard_alignment from hd_struct."
nbd: adjust 'max_part' according to part_shift
nbd: limit module parameters to a sane value
nbd: pass MSG_* flags to kernel_recvmsg()
block: improve the bio_add_page() and bio_add_pc_page() descriptions
Caching "we have already removed suid/caps" was overenthusiastic as merged.
On network filesystems we might have had suid/caps set on another client,
silently picked by this client on revalidate, all of that *without* clearing
the S_NOSEC flag.
AFAICS, the only reasonably sane way to deal with that is
* new superblock flag; unless set, S_NOSEC is not going to be set.
* local block filesystems set it in their ->mount() (more accurately,
mount_bdev() does, so does btrfs ->mount(), users of mount_bdev() other than
local block ones clear it)
* if any network filesystem (or a cluster one) wants to use S_NOSEC,
it'll need to set MS_NOSEC in sb->s_flags *AND* take care to clear S_NOSEC when
inode attribute changes are picked from other clients.
It's not an earth-shattering hole (anybody that can set suid on another client
will almost certainly be able to write to the file before doing that anyway),
but it's a bug that needs fixing.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
When CONFIG_CRYPTO_ECB is not set, trying to mount a CIFS share with NTLM
security resulted in mount failure with the following error:
"CIFS VFS: could not allocate des crypto API"
Seems like a leftover from commit 43988d7.
Signed-off-by: Suresh Jayaraman <sjayaraman@suse.de>
CC: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
When ntlm security mechanim is used, the message that warns about the upgrade
to ntlmv2 got the kernel release version wrong (Blame it on Linus :). Fix it.
Signed-off-by: Suresh Jayaraman <sjayaraman@suse.de>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
The free space fixup is currently initiated during mount after the call to
ubifs_write_master() which results in a write to PEBs; this has been observed
with the patch 'assert no fixup when writing a node' applied:
Move the free space fixup on mount to before the calls to
ubifs_recover_inl_heads() and ubifs_write_master(). This results in no
assertions with the previously mentioned patch applied.
Artem: tweaked the patch a bit
Signed-off-by: Ben Gardiner <bengardiner@nanometrics>
Reviewed-by: Matthew L. Creech <mlcreech@gmail.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
The current 'mount_ubifs()' implementation does not initialize the LPT until the
the master node is marked dirty. Move the LPT initialization to before marking
the master node dirty. This is a preparation for the next patch which will move
the free-space-fixup check to before marking the master node dirty, because we
have to fix-up the free space before doing any writes.
Artem: massaged the patch and commit message.
Signed-off-by: Ben Gardiner <bengardiner@nanometrics.ca>
Reviewed-by: Matthew L. Creech <mlcreech@gmail.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
The current free space fixup can result in some writing to the UBI volume
when the space_fixup flag is set.
To catch instances where UBIFS is writing to the NAND while the space_fixup
flag is set, add an assert to ubifs_write_node().
Artem: tweaked the patch, added similar assertion to the write buffer
write path.
Signed-off-by: Ben Gardiner <bengardiner@nanometrics.ca>
Reviewed-by: Matthew L. Creech <mlcreech@gmail.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
UBIFS maintains per-filesystem and global clean znode counters
('c->clean_zn_cnt' and 'ubifs_clean_zn_cnt'). It is important to maintain
correct values there since the shrinker relies on 'ubifs_clean_zn_cnt'.
However, in case of failures during commit the counters were corrupted. E.g.,
if a failure happens in the middle of 'write_index()', then some nodes in the
commit list ('c->cnext') are marked as clean, and some are marked as dirty. And
the 'ubifs_destroy_tnc_subtree()' frees does not retrun correct count, and we
end up with non-zero 'c->clean_zn_cnt' when unmounting. This means that if we
have 2 file-sytem and one of them fails, and we unmount it,
'ubifs_clean_zn_cnt' stays incorrect and confuses the shrinker.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
UBIFS leaks memory on error path in 'ubifs_jnl_update()' in case of write
failure because it forgets to free the 'struct ubifs_dent_node *dent' object.
Although the object is small, the alignment can make it large - e.g., 2KiB
if the min. I/O unit is 2KiB.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: stable@kernel.org
Sometimes VM asks the shrinker to return amount of objects it can shrink,
and we return the ubifs_clean_zn_cnt in that case. However, it is possible
that this counter is negative for a short period of time, due to the way
UBIFS TNC code updates it. And I can observe the following warnings sometimes:
shrink_slab: ubifs_shrinker+0x0/0x2b7 [ubifs] negative objects to delete nr=-8541616642706119788
This patch makes sure UBIFS never returns negative count of objects.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: stable@kernel.org
Unfortunately, the recovery fix d1606a59b6be4ea392eabd40d1250aa1eeb19efb
(UBIFS: fix extremely rare mount failure) broke recovery. This commit make
UBIFS drop the last min. I/O unit in all journal heads, but this is needed only
for the GC head. And this does not work for non-GC heads. For example, if
suppose we have min. I/O units A and B, and A contains a valid node X, which
was fsynced, and then a group of nodes Y which spans the rest of A and B. In
this case we'll drop not only Y, but also X, which is obviously incorrect.
This patch fixes the issue and additionally makes recovery to drop last min.
I/O unit only for the GC head, and leave things as they have been for ages for
the other heads - this is safer.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Instead of passing "grouped" parameter to 'ubifs_recover_leb()' which tells
whether the nodes are grouped in the LEB to recover, pass the journal head
number and let 'ubifs_recover_leb()' look at the journal head's 'grouped' flag.
This patch is a preparation to a further fix where we'll need to know the
journal head number for other purposes.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Journal heads are different in a way how UBIFS writes nodes there. All normal
journal heads receive grouped nodes, while the GC journal heads receives
ungrouped nodes. This patch adds a 'grouped' flag to 'struct ubifs_jhead' which
describes this property.
This patch is a preparation to a further recovery fix.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Commit ab51afe05273741f72383529ef488aa1ea598ec6 was a good clean-up, but
it introduced a regression - now UBIFS prints scary error messages during
recovery on all corrupted nodes, even though the corruptions are expected
(due to a power cut). This patch fixes the issue.
Additionally fix a typo in a commentary introduced by the same commit.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
d4dc210f69 (block: don't block events on excl write for non-optical
devices) added dereferencing of bdev->bd_disk to test
GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE; however, bdev->bd_disk can be
%NULL if open failed which can lead to an oops.
Test the flag after testing open was successful, not before.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: David Miller <davem@davemloft.net>
Tested-by: David Miller <davem@davemloft.net>
Cc: stable@kernel.org
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
The lockdep warning below detects a possible A->B/B->A locking
dependency of mm->mmap_sem and dcookie_mutex. The order in
sync_buffer() is mm->mmap_sem/dcookie_mutex, while in
sys_lookup_dcookie() it is vice versa.
Fixing it in sys_lookup_dcookie() by unlocking dcookie_mutex before
copy_to_user().
oprofiled/4432 is trying to acquire lock:
(&mm->mmap_sem){++++++}, at: [<ffffffff810b444b>] might_fault+0x53/0xa3
but task is already holding lock:
(dcookie_mutex){+.+.+.}, at: [<ffffffff81124d28>] sys_lookup_dcookie+0x45/0x149
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (dcookie_mutex){+.+.+.}:
[<ffffffff8106557f>] lock_acquire+0xf8/0x11e
[<ffffffff814634f0>] mutex_lock_nested+0x63/0x309
[<ffffffff81124e5c>] get_dcookie+0x30/0x144
[<ffffffffa0000fba>] sync_buffer+0x196/0x3ec [oprofile]
[<ffffffffa0001226>] task_exit_notify+0x16/0x1a [oprofile]
[<ffffffff81467b96>] notifier_call_chain+0x37/0x63
[<ffffffff8105803d>] __blocking_notifier_call_chain+0x50/0x67
[<ffffffff81058068>] blocking_notifier_call_chain+0x14/0x16
[<ffffffff8105a718>] profile_task_exit+0x1a/0x1c
[<ffffffff81039e8f>] do_exit+0x2a/0x6fc
[<ffffffff8103a5e4>] do_group_exit+0x83/0xae
[<ffffffff8103a626>] sys_exit_group+0x17/0x1b
[<ffffffff8146ad4b>] system_call_fastpath+0x16/0x1b
-> #0 (&mm->mmap_sem){++++++}:
[<ffffffff81064dfb>] __lock_acquire+0x1085/0x1711
[<ffffffff8106557f>] lock_acquire+0xf8/0x11e
[<ffffffff810b4478>] might_fault+0x80/0xa3
[<ffffffff81124de7>] sys_lookup_dcookie+0x104/0x149
[<ffffffff8146ad4b>] system_call_fastpath+0x16/0x1b
other info that might help us debug this:
1 lock held by oprofiled/4432:
#0: (dcookie_mutex){+.+.+.}, at: [<ffffffff81124d28>] sys_lookup_dcookie+0x45/0x149
stack backtrace:
Pid: 4432, comm: oprofiled Not tainted 2.6.39-00008-ge5a450d #9
Call Trace:
[<ffffffff81063193>] print_circular_bug+0xae/0xbc
[<ffffffff81064dfb>] __lock_acquire+0x1085/0x1711
[<ffffffff8102ef13>] ? get_parent_ip+0x11/0x42
[<ffffffff810b444b>] ? might_fault+0x53/0xa3
[<ffffffff8106557f>] lock_acquire+0xf8/0x11e
[<ffffffff810b444b>] ? might_fault+0x53/0xa3
[<ffffffff810d7d54>] ? path_put+0x22/0x27
[<ffffffff810b4478>] might_fault+0x80/0xa3
[<ffffffff810b444b>] ? might_fault+0x53/0xa3
[<ffffffff81124de7>] sys_lookup_dcookie+0x104/0x149
[<ffffffff8146ad4b>] system_call_fastpath+0x16/0x1b
References: https://bugzilla.kernel.org/show_bug.cgi?id=13809
Cc: <stable@kernel.org> # .27+
Signed-off-by: Robert Richter <robert.richter@amd.com>
The dentry_unhash push-down series missed that shink_dcache_parent needs to
be called prior to rmdir or dir rename to clear DCACHE_REFERENCED and
allow efficient dentry reclaim.
Reported-by: Dave Chinner <david@fromorbit.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
It was not a good idea to start dereferencing disk->queue from
the fs sysfs strategy for displaying discard alignment. We ran
into first a NULL pointer deref, and after fixing that we sometimes
see unvalid disk->queue pointer values.
Since discard is the only one of the bunch actually looking into
the queue, just revert the change.
This reverts commit 23ceb5b771.
Conflicts:
fs/partitions/check.c
Now that ecryptfs_lookup_interpose() is no longer using
ecryptfs_header_cache_2 to read in metadata, the kmem_cache can be
removed and the ecryptfs_header_cache_1 kmem_cache can be renamed to
ecryptfs_header_cache.
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
ecryptfs_lookup_interpose() has turned into spaghetti code over the
years. This is an effort to clean it up.
- Shorten overly descriptive variable names such as ecryptfs_dentry
- Simplify gotos and error paths
- Create helper function for reading plaintext i_size from metadata
It also includes an optimization when reading i_size from the metadata.
A complete page-sized kmem_cache_alloc() was being done to read in 16
bytes of metadata. The buffer for that is now statically declared.
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Instead of having the calling functions translate the true/false return
code to either 0 or -EINVAL, have contains_ecryptfs_marker() return 0 or
-EINVAL so that the calling functions can just reuse the return code.
Also, rename the function to ecryptfs_validate_marker() to avoid callers
mistakenly thinking that it returns true/false codes.
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Only unlock and d_add() new inodes after the plaintext inode size has
been read from the lower filesystem. This fixes a race condition that
was sometimes seen during a multi-job kernel build in an eCryptfs mount.
https://bugzilla.kernel.org/show_bug.cgi?id=36002
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Reported-by: David <david@unsolicited.net>
Tested-by: David <david@unsolicited.net>
* 'for-2.6.40' of git://linux-nfs.org/~bfields/linux: (22 commits)
nfsd: make local functions static
NFSD: Remove unused variable from nfsd4_decode_bind_conn_to_session()
NFSD: Check status from nfsd4_map_bcts_dir()
NFSD: Remove setting unused variable in nfsd_vfs_read()
nfsd41: error out on repeated RECLAIM_COMPLETE
nfsd41: compare request's opcnt with session's maxops at nfsd4_sequence
nfsd v4.1 lOCKT clientid field must be ignored
nfsd41: add flag checking for create_session
nfsd41: make sure nfs server process OPEN with EXCLUSIVE4_1 correctly
nfsd4: fix wrongsec handling for PUTFH + op cases
nfsd4: make fh_verify responsibility of nfsd_lookup_dentry caller
nfsd4: introduce OPDESC helper
nfsd4: allow fh_verify caller to skip pseudoflavor checks
nfsd: distinguish functions of NFSD_MAY_* flags
svcrpc: complete svsk processing on cb receive failure
svcrpc: take advantage of tcp autotuning
SUNRPC: Don't wait for full record to receive tcp data
svcrpc: copy cb reply instead of pages
svcrpc: close connection if client sends short packet
svcrpc: note network-order types in svc_process_calldir
...
* 'nfs-for-2.6.40' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6:
SUNRPC: Support for RPC over AF_LOCAL transports
SUNRPC: Remove obsolete comment
SUNRPC: Use AF_LOCAL for rpcbind upcalls
SUNRPC: Clean up use of curly braces in switch cases
NFS: Revert NFSROOT default mount options
SUNRPC: Rename xs_encode_tcp_fragment_header()
nfs,rcu: convert call_rcu(nfs_free_delegation_callback) to kfree_rcu()
nfs41: Correct offset for LAYOUTCOMMIT
NFS: nfs_update_inode: print current and new inode size in debug output
NFSv4.1: Fix the handling of NFS4ERR_SEQ_MISORDERED errors
NFSv4: Handle expired stateids when the lease is still valid
SUNRPC: Deal with the lack of a SYN_SENT sk->sk_state_change callback...
Commit 1495f230fa ("vmscan: change shrinker API by passing
shrink_control struct") changed the API of ->shrink(), but missed ubifs
and cifs instances.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Implement pg_test vector to test for max IO sizes. We calculate
a max_io_size member only once, and cache it in lseg so to not
do so on every page insert.
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
[simplify logic]
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
By default, unless pnfs is used coalesce pages until pg_bsize
(rsize or wsize) is reached.
pnfs layout drivers define their own pg_test methods that use
pnfs_generic_pg_test and need to define their own I/O size
limits (e.g. based on the file stripe size).
[Move a check from nfs_pageio_do_add_request to nfs_generic_pg_test]
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Use common code for pnfs_pageio_init_{read,write} and use
a common generic pg_test function.
Note that this function always assumes the the layout driver's
pg_test method is implemented.
[Fix BUG]
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
* Define API for io-engines to report delta_space_used in IOs
* Encode the osd-layout specific information of the layoutcommit
XDR buffer.
Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Add a layout driver method to encode the layout type specific
opaque part of layout commit in-line in the xdr stream.
Currently, the pnfs-objects layout driver uses it to encode metadata hints
to the MDS and the blocks layout driver to commit provisionally allocated
extents to the file.
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
