Commit 1c8007b0769d37aa5fcb343b383b0af89ade2f71
1 parent
3da3f872aa
Exists in
master
and in
6 other branches
jfs: flush journal completely before releasing metadata inodes
This fixes a race during unmount. We need to not only make sure that the journal is completely written, but that the metadata changes make it to disk before releasing ipimap and ipbmap. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
Showing 1 changed file with 2 additions and 2 deletions Side-by-side Diff
fs/jfs/jfs_umount.c
... | ... | @@ -68,7 +68,7 @@ |
68 | 68 | /* |
69 | 69 | * Wait for outstanding transactions to be written to log: |
70 | 70 | */ |
71 | - jfs_flush_journal(log, 1); | |
71 | + jfs_flush_journal(log, 2); | |
72 | 72 | |
73 | 73 | /* |
74 | 74 | * close fileset inode allocation map (aka fileset inode) |
... | ... | @@ -146,7 +146,7 @@ |
146 | 146 | * |
147 | 147 | * remove file system from log active file system list. |
148 | 148 | */ |
149 | - jfs_flush_journal(log, 1); | |
149 | + jfs_flush_journal(log, 2); | |
150 | 150 | |
151 | 151 | /* |
152 | 152 | * Make sure all metadata makes it to disk |