diff options
author | Jim Schutt <jaschut@sandia.gov> | 2013-05-15 14:03:35 -0400 |
---|---|---|
committer | Alex Elder <elder@inktank.com> | 2013-05-17 13:45:48 -0400 |
commit | 39be95e9c8c0b5668c9f8806ffe29bf9f4bc0f40 (patch) | |
tree | 580f3e90a2175e7c9173cce615696c9972df4982 /fs/ceph/mds_client.c | |
parent | c420276a532a10ef59849adc2681f45306166b89 (diff) |
ceph: ceph_pagelist_append might sleep while atomic
Ceph's encode_caps_cb() worked hard to not call __page_cache_alloc()
while holding a lock, but it's spoiled because ceph_pagelist_addpage()
always calls kmap(), which might sleep. Here's the result:
[13439.295457] ceph: mds0 reconnect start
[13439.300572] BUG: sleeping function called from invalid context at include/linux/highmem.h:58
[13439.309243] in_atomic(): 1, irqs_disabled(): 0, pid: 12059, name: kworker/1:1
. . .
[13439.376225] Call Trace:
[13439.378757] [<ffffffff81076f4c>] __might_sleep+0xfc/0x110
[13439.384353] [<ffffffffa03f4ce0>] ceph_pagelist_append+0x120/0x1b0 [libceph]
[13439.391491] [<ffffffffa0448fe9>] ceph_encode_locks+0x89/0x190 [ceph]
[13439.398035] [<ffffffff814ee849>] ? _raw_spin_lock+0x49/0x50
[13439.403775] [<ffffffff811cadf5>] ? lock_flocks+0x15/0x20
[13439.409277] [<ffffffffa045e2af>] encode_caps_cb+0x41f/0x4a0 [ceph]
[13439.415622] [<ffffffff81196748>] ? igrab+0x28/0x70
[13439.420610] [<ffffffffa045e9f8>] ? iterate_session_caps+0xe8/0x250 [ceph]
[13439.427584] [<ffffffffa045ea25>] iterate_session_caps+0x115/0x250 [ceph]
[13439.434499] [<ffffffffa045de90>] ? set_request_path_attr+0x2d0/0x2d0 [ceph]
[13439.441646] [<ffffffffa0462888>] send_mds_reconnect+0x238/0x450 [ceph]
[13439.448363] [<ffffffffa0464542>] ? ceph_mdsmap_decode+0x5e2/0x770 [ceph]
[13439.455250] [<ffffffffa0462e42>] check_new_map+0x352/0x500 [ceph]
[13439.461534] [<ffffffffa04631ad>] ceph_mdsc_handle_map+0x1bd/0x260 [ceph]
[13439.468432] [<ffffffff814ebc7e>] ? mutex_unlock+0xe/0x10
[13439.473934] [<ffffffffa043c612>] extra_mon_dispatch+0x22/0x30 [ceph]
[13439.480464] [<ffffffffa03f6c2c>] dispatch+0xbc/0x110 [libceph]
[13439.486492] [<ffffffffa03eec3d>] process_message+0x1ad/0x1d0 [libceph]
[13439.493190] [<ffffffffa03f1498>] ? read_partial_message+0x3e8/0x520 [libceph]
. . .
[13439.587132] ceph: mds0 reconnect success
[13490.720032] ceph: mds0 caps stale
[13501.235257] ceph: mds0 recovery completed
[13501.300419] ceph: mds0 caps renewed
Fix it up by encoding locks into a buffer first, and when the number
of encoded locks is stable, copy that into a ceph_pagelist.
[elder@inktank.com: abbreviated the stack info a bit.]
Cc: stable@vger.kernel.org # 3.4+
Signed-off-by: Jim Schutt <jaschut@sandia.gov>
Reviewed-by: Alex Elder <elder@inktank.com>
Diffstat (limited to 'fs/ceph/mds_client.c')
-rw-r--r-- | fs/ceph/mds_client.c | 65 |
1 files changed, 35 insertions, 30 deletions
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index d9ca15255477..4d2920304be8 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c | |||
@@ -2478,39 +2478,44 @@ static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap, | |||
2478 | 2478 | ||
2479 | if (recon_state->flock) { | 2479 | if (recon_state->flock) { |
2480 | int num_fcntl_locks, num_flock_locks; | 2480 | int num_fcntl_locks, num_flock_locks; |
2481 | struct ceph_pagelist_cursor trunc_point; | 2481 | struct ceph_filelock *flocks; |
2482 | 2482 | ||
2483 | ceph_pagelist_set_cursor(pagelist, &trunc_point); | 2483 | encode_again: |
2484 | do { | 2484 | lock_flocks(); |
2485 | lock_flocks(); | 2485 | ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks); |
2486 | ceph_count_locks(inode, &num_fcntl_locks, | 2486 | unlock_flocks(); |
2487 | &num_flock_locks); | 2487 | flocks = kmalloc((num_fcntl_locks+num_flock_locks) * |
2488 | rec.v2.flock_len = cpu_to_le32(2*sizeof(u32) + | 2488 | sizeof(struct ceph_filelock), GFP_NOFS); |
2489 | (num_fcntl_locks+num_flock_locks) * | 2489 | if (!flocks) { |
2490 | sizeof(struct ceph_filelock)); | 2490 | err = -ENOMEM; |
2491 | unlock_flocks(); | 2491 | goto out_free; |
2492 | 2492 | } | |
2493 | /* pre-alloc pagelist */ | 2493 | lock_flocks(); |
2494 | ceph_pagelist_truncate(pagelist, &trunc_point); | 2494 | err = ceph_encode_locks_to_buffer(inode, flocks, |
2495 | err = ceph_pagelist_append(pagelist, &rec, reclen); | 2495 | num_fcntl_locks, |
2496 | if (!err) | 2496 | num_flock_locks); |
2497 | err = ceph_pagelist_reserve(pagelist, | 2497 | unlock_flocks(); |
2498 | rec.v2.flock_len); | 2498 | if (err) { |
2499 | 2499 | kfree(flocks); | |
2500 | /* encode locks */ | 2500 | if (err == -ENOSPC) |
2501 | if (!err) { | 2501 | goto encode_again; |
2502 | lock_flocks(); | 2502 | goto out_free; |
2503 | err = ceph_encode_locks(inode, | 2503 | } |
2504 | pagelist, | 2504 | /* |
2505 | num_fcntl_locks, | 2505 | * number of encoded locks is stable, so copy to pagelist |
2506 | num_flock_locks); | 2506 | */ |
2507 | unlock_flocks(); | 2507 | rec.v2.flock_len = cpu_to_le32(2*sizeof(u32) + |
2508 | } | 2508 | (num_fcntl_locks+num_flock_locks) * |
2509 | } while (err == -ENOSPC); | 2509 | sizeof(struct ceph_filelock)); |
2510 | err = ceph_pagelist_append(pagelist, &rec, reclen); | ||
2511 | if (!err) | ||
2512 | err = ceph_locks_to_pagelist(flocks, pagelist, | ||
2513 | num_fcntl_locks, | ||
2514 | num_flock_locks); | ||
2515 | kfree(flocks); | ||
2510 | } else { | 2516 | } else { |
2511 | err = ceph_pagelist_append(pagelist, &rec, reclen); | 2517 | err = ceph_pagelist_append(pagelist, &rec, reclen); |
2512 | } | 2518 | } |
2513 | |||
2514 | out_free: | 2519 | out_free: |
2515 | kfree(path); | 2520 | kfree(path); |
2516 | out_dput: | 2521 | out_dput: |