Commit 4a378d8
Al Viro
gfs2: be careful with inode refresh
1) gfs2_dinode_in() should *not* touch ->i_rdev on live inodes; even
"zero and immediately reread the same value from dinode" is broken -
have it overlap with ->release() of char device and you can get all
kinds of bogus behaviour.
2) mismatch on inode type on live inodes should be treated as fs
corruption rather than blindly setting ->i_mode.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>1 parent 60606ec commit 4a378d8
1 file changed
Lines changed: 14 additions & 8 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
394 | 394 | | |
395 | 395 | | |
396 | 396 | | |
| 397 | + | |
| 398 | + | |
397 | 399 | | |
398 | 400 | | |
399 | 401 | | |
| 402 | + | |
| 403 | + | |
400 | 404 | | |
401 | | - | |
402 | | - | |
403 | | - | |
404 | | - | |
405 | | - | |
406 | | - | |
407 | | - | |
408 | | - | |
| 405 | + | |
| 406 | + | |
| 407 | + | |
| 408 | + | |
| 409 | + | |
| 410 | + | |
| 411 | + | |
| 412 | + | |
| 413 | + | |
| 414 | + | |
409 | 415 | | |
410 | 416 | | |
411 | 417 | | |
| |||
0 commit comments