linux-stable.git/fs, branch v3.16.78

btrfs: alloc_chunk: fix more DUP stripe size handling

2019-11-22T15:57:33+00:00

commit baf92114c7e6dd6124aa3d506e4bc4b694da3bc3 upstream.

Commit 92e222df7b "btrfs: alloc_chunk: fix DUP stripe size handling"
fixed calculating the stripe_size for a new DUP chunk.

However, the same calculation reappears a bit later, and that one was
not changed yet. The resulting bug that is exposed is that the newly
allocated device extents ('stripes') can have a few MiB overlap with the
next thing stored after them, which is another device extent or the end
of the disk.

The scenario in which this can happen is:
* The block device for the filesystem is less than 10GiB in size.
* The amount of contiguous free unallocated disk space chosen to use for
  chunk allocation is 20% of the total device size, or a few MiB more or
  less.

An example:
- The filesystem device is 7880MiB (max_chunk_size gets set to 788MiB)
- There's 1578MiB unallocated raw disk space left in one contiguous
  piece.

In this case stripe_size is first calculated as 789MiB, (half of
1578MiB).

Since 789MiB (stripe_size * data_stripes) > 788MiB (max_chunk_size), we
enter the if block. Now stripe_size value is immediately overwritten
while calculating an adjusted value based on max_chunk_size, which ends
up as 788MiB.

Next, the value is rounded up to a 16MiB boundary, 800MiB, which is
actually more than the value we had before. However, the last comparison
fails to detect this, because it's comparing the value with the total
amount of free space, which is about twice the size of stripe_size.

In the example above, this means that the resulting raw disk space being
allocated is 1600MiB, while only a gap of 1578MiB has been found. The
second device extent object for this DUP chunk will overlap for 22MiB
with whatever comes next.

The underlying problem here is that the stripe_size is reused all the
time for different things. So, when entering the code in the if block,
stripe_size is immediately overwritten with something else. If later we
decide we want to have the previous value back, then the logic to
compute it was copy pasted in again.

With this change, the value in stripe_size is not unnecessarily
destroyed, so the duplicated calculation is not needed any more.

Signed-off-by: Hans van Kranenburg 
Signed-off-by: David Sterba 
Signed-off-by: Ben Hutchings

btrfs: volumes: Cleanup stripe size calculation

2019-11-22T15:57:32+00:00

commit 793ff2c88c6397b3531c08cc4f920619b56a9def upstream.

Cleanup the following things:
1) open coded SZ_16M round up
2) use min() to replace open-coded size comparison
3) code style

Signed-off-by: Qu Wenruo 
Reviewed-by: Nikolay Borisov 
Reviewed-by: Gu Jinxiang 
[ reformat comment ]
Signed-off-by: David Sterba 
[bwh: Backported to 3.16 as dependency of commit baf92114c7
 "btrfs: alloc_chunk: fix more DUP stripe size handling":
 - Add #include  for definition of SZ_16M]
Signed-off-by: Ben Hutchings 
Cc: Hans van Kranenburg

btrfs: partially apply b8b93addde

2019-11-22T15:57:32+00:00

Extracted from commit b8b93addde "btrfs: cleanup 64bit/32bit divs,
provably bounded values", to allow commits 793ff2c88c6 "btrfs:
volumes: Cleanup stripe size calculation" and baf92114c7 "btrfs:
alloc_chunk: fix more DUP stripe size handling" to apply cleanly.

[bwh: Add patch description]
Signed-off-by: Ben Hutchings

CIFS: Fix use after free of file info structures

2019-11-22T15:57:31+00:00

commit 1a67c415965752879e2e9fad407bc44fc7f25f23 upstream.

Currently the code assumes that if a file info entry belongs
to lists of open file handles of an inode and a tcon then
it has non-zero reference. The recent changes broke that
assumption when putting the last reference of the file info.
There may be a situation when a file is being deleted but
nothing prevents another thread to reference it again
and start using it. This happens because we do not hold
the inode list lock while checking the number of references
of the file info structure. Fix this by doing the proper
locking when doing the check.

Fixes: 487317c99477d ("cifs: add spinlock for the openFileList to cifsInodeInfo")
Fixes: cb248819d209d ("cifs: use cifsInodeInfo->open_file_lock while iterating to avoid a panic")
Reviewed-by: Ronnie Sahlberg 
Signed-off-by: Pavel Shilovsky 
Signed-off-by: Steve French 
Signed-off-by: Ben Hutchings

cifs: use cifsInodeInfo->open_file_lock while iterating to avoid a panic

2019-11-22T15:57:31+00:00

commit cb248819d209d113e45fed459773991518e8e80b upstream.

Commit 487317c99477 ("cifs: add spinlock for the openFileList to
cifsInodeInfo") added cifsInodeInfo->open_file_lock spin_lock to protect
the openFileList, but missed a few places where cifs_inode->openFileList
was enumerated.  Change these remaining tcon->open_file_lock to
cifsInodeInfo->open_file_lock to avoid panic in is_size_safe_to_change.

[17313.245641] RIP: 0010:is_size_safe_to_change+0x57/0xb0 [cifs]
[17313.245645] Code: 68 40 48 89 ef e8 19 67 b7 f1 48 8b 43 40 48 8d 4b 40 48 8d 50 f0 48 39 c1 75 0f eb 47 48 8b 42 10 48 8d 50 f0 48 39 c1 74 3a <8b> 80 88 00 00 00 83 c0 01 a8 02 74 e6 48 89 ef c6 07 00 0f 1f 40
[17313.245649] RSP: 0018:ffff94ae1baefa30 EFLAGS: 00010202
[17313.245654] RAX: dead000000000100 RBX: ffff88dc72243300 RCX: ffff88dc72243340
[17313.245657] RDX: dead0000000000f0 RSI: 00000000098f7940 RDI: ffff88dd3102f040
[17313.245659] RBP: ffff88dd3102f040 R08: 0000000000000000 R09: ffff94ae1baefc40
[17313.245661] R10: ffffcdc8bb1c4e80 R11: ffffcdc8b50adb08 R12: 00000000098f7940
[17313.245663] R13: ffff88dc72243300 R14: ffff88dbc8f19600 R15: ffff88dc72243428
[17313.245667] FS:  00007fb145485700(0000) GS:ffff88dd3e000000(0000) knlGS:0000000000000000
[17313.245670] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[17313.245672] CR2: 0000026bb46c6000 CR3: 0000004edb110003 CR4: 00000000007606e0
[17313.245753] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[17313.245756] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[17313.245759] PKRU: 55555554
[17313.245761] Call Trace:
[17313.245803]  cifs_fattr_to_inode+0x16b/0x580 [cifs]
[17313.245838]  cifs_get_inode_info+0x35c/0xa60 [cifs]
[17313.245852]  ? kmem_cache_alloc_trace+0x151/0x1d0
[17313.245885]  cifs_open+0x38f/0x990 [cifs]
[17313.245921]  ? cifs_revalidate_dentry_attr+0x3e/0x350 [cifs]
[17313.245953]  ? cifsFileInfo_get+0x30/0x30 [cifs]
[17313.245960]  ? do_dentry_open+0x132/0x330
[17313.245963]  do_dentry_open+0x132/0x330
[17313.245969]  path_openat+0x573/0x14d0
[17313.245974]  do_filp_open+0x93/0x100
[17313.245979]  ? __check_object_size+0xa3/0x181
[17313.245986]  ? audit_alloc_name+0x7e/0xd0
[17313.245992]  do_sys_open+0x184/0x220
[17313.245999]  do_syscall_64+0x5b/0x1b0

Fixes: 487317c99477 ("cifs: add spinlock for the openFileList to cifsInodeInfo")

Signed-off-by: Dave Wysochanski 
Reviewed-by: Ronnie Sahlberg 
Signed-off-by: Steve French 
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings 
Cc: Pavel Shilovskiy

smb3: send CAP_DFS capability during session setup

2019-11-22T15:57:13+00:00

commit 8d33096a460d5b9bd13300f01615df5bb454db10 upstream.

We had a report of a server which did not do a DFS referral
because the session setup Capabilities field was set to 0
(unlike negotiate protocol where we set CAP_DFS).  Better to
send it session setup in the capabilities as well (this also
more closely matches Windows client behavior).

Signed-off-by: Steve French 
Reviewed-off-by: Ronnie Sahlberg 
Reviewed-by: Pavel Shilovsky 
Signed-off-by: Ben Hutchings

SMB3: Fix deadlock in validate negotiate hits reconnect

2019-11-22T15:57:12+00:00

commit e99c63e4d86d3a94818693147b469fa70de6f945 upstream.

Currently we skip SMB2_TREE_CONNECT command when checking during
reconnect because Tree Connect happens when establishing
an SMB session. For SMB 3.0 protocol version the code also calls
validate negotiate which results in SMB2_IOCL command being sent
over the wire. This may deadlock on trying to acquire a mutex when
checking for reconnect. Fix this by skipping SMB2_IOCL command
when doing the reconnect check.

Signed-off-by: Pavel Shilovsky 
Signed-off-by: Steve French 
Reviewed-by: Ronnie Sahlberg 
Signed-off-by: Ben Hutchings

NFSv4: Fix a potential sleep while atomic in nfs4_do_reclaim()

2019-11-22T15:57:11+00:00

commit c77e22834ae9a11891cb613bd9a551be1b94f2bc upstream.

John Hubbard reports seeing the following stack trace:

nfs4_do_reclaim
   rcu_read_lock /* we are now in_atomic() and must not sleep */
       nfs4_purge_state_owners
           nfs4_free_state_owner
               nfs4_destroy_seqid_counter
                   rpc_destroy_wait_queue
                       cancel_delayed_work_sync
                           __cancel_work_timer
                               __flush_work
                                   start_flush_work
                                       might_sleep:
                                        (kernel/workqueue.c:2975: BUG)

The solution is to separate out the freeing of the state owners
from nfs4_purge_state_owners(), and perform that outside the atomic
context.

Reported-by: John Hubbard 
Fixes: 0aaaf5c424c7f ("NFS: Cache state owners after files are closed")
Signed-off-by: Trond Myklebust 
Signed-off-by: Ben Hutchings

sched/fair: Don't free p->numa_faults with concurrent readers

2019-11-22T15:57:07+00:00

commit 16d51a590a8ce3befb1308e0e7ab77f3b661af33 upstream.

When going through execve(), zero out the NUMA fault statistics instead of
freeing them.

During execve, the task is reachable through procfs and the scheduler. A
concurrent /proc/*/sched reader can read data from a freed ->numa_faults
allocation (confirmed by KASAN) and write it back to userspace.
I believe that it would also be possible for a use-after-free read to occur
through a race between a NUMA fault and execve(): task_numa_fault() can
lead to task_numa_compare(), which invokes task_weight() on the currently
running task of a different CPU.

Another way to fix this would be to make ->numa_faults RCU-managed or add
extra locking, but it seems easier to wipe the NUMA fault statistics on
execve.

Signed-off-by: Jann Horn 
Signed-off-by: Peter Zijlstra (Intel) 
Cc: Linus Torvalds 
Cc: Peter Zijlstra 
Cc: Petr Mladek 
Cc: Sergey Senozhatsky 
Cc: Thomas Gleixner 
Cc: Will Deacon 
Fixes: 82727018b0d3 ("sched/numa: Call task_numa_free() from do_execve()")
Link: https://lkml.kernel.org/r/20190716152047.14424-1-jannh@google.com
Signed-off-by: Ingo Molnar 
[bwh: Backported to 3.16: adjust filename, context]
Signed-off-by: Ben Hutchings

coda: pass the host file in vma->vm_file on mmap

2019-10-31T22:15:04+00:00

commit 7fa0a1da3dadfd9216df7745a1331fdaa0940d1c upstream.

Patch series "Coda updates".

The following patch series is a collection of various fixes for Coda,
most of which were collected from linux-fsdevel or linux-kernel but
which have as yet not found their way upstream.

This patch (of 22):

Various file systems expect that vma->vm_file points at their own file
handle, several use file_inode(vma->vm_file) to get at their inode or
use vma->vm_file->private_data.  However the way Coda wrapped mmap on a
host file broke this assumption, vm_file was still pointing at the Coda
file and the host file systems would scribble over Coda's inode and
private file data.

This patch fixes the incorrect expectation and wraps vm_ops->open and
vm_ops->close to allow Coda to track when the vm_area_struct is
destroyed so we still release the reference on the Coda file handle at
the right time.

Link: http://lkml.kernel.org/r/0e850c6e59c0b147dc2dcd51a3af004c948c3697.1558117389.git.jaharkes@cs.cmu.edu
Signed-off-by: Jan Harkes 
Cc: Arnd Bergmann 
Cc: Colin Ian King 
Cc: Dan Carpenter 
Cc: David Howells 
Cc: Fabian Frederick 
Cc: Mikko Rapeli 
Cc: Sam Protsenko 
Cc: Yann Droneaud 
Cc: Zhouyang Jia 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
[bwh: Backported to 3.16: Keep calling file_operations::mmap directly]
Signed-off-by: Ben Hutchings