xfs: fixes for 3.12-rc2
Get series with b4
b4 am [email protected]Versions newest first
- v1 Latest3 patches 0 repliesNew
Patches in v1 No applies recorded yet
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/3 | xfs: don't try to mark uncached buffers stale on error.fs/ | +15 | 0 replies | |
| 2/3 | xfs: lock the AIL before removing the buffer itemfs/ | +1 | 0 replies | |
| 3/3 | xfs: asserting lock not held during freeing not validfs/ | +4 | 0 replies |
Reviews trailers per patch
No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.
Cover letter [PATCH 0/3] xfs: fixes for 3.12-rc2
Hi folks,
Three patches for 3.12-rc2. The first has been posted previously, and prevents an assert failure during a shutdown while zeroing space with an uncached buffer.
The second patch is to fix a regression introduced in the 3.12 merge by commit 46f9d2e ("xfs: aborted buf items can be in the AIL").
The third is fixing a thinko made a long time ago when adding RCU freeing to the struct xfs_inode, but has only been exposed by the recent merge of the per-node LRUs and shrinkers allowing concurrent reclaim and lookup races to be hit easily. The RCU traversals of the inode radix trees take the lock on the inode to determine if it is being freed or not, and so that can race with freeing of the inode. Hence we can't assert that the lock is not held when freeing the inode.
These all fix regressions triggered by xfstests....
Cheers,
Dave.
_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs