diff options
Diffstat (limited to 'Documentation/filesystems/Locking')
-rw-r--r-- | Documentation/filesystems/Locking | 5 |
1 files changed, 3 insertions, 2 deletions
diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index af1608070cd5..61c98f03baa1 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking | |||
@@ -429,8 +429,9 @@ check_flags: no | |||
429 | implementations. If your fs is not using generic_file_llseek, you | 429 | implementations. If your fs is not using generic_file_llseek, you |
430 | need to acquire and release the appropriate locks in your ->llseek(). | 430 | need to acquire and release the appropriate locks in your ->llseek(). |
431 | For many filesystems, it is probably safe to acquire the inode | 431 | For many filesystems, it is probably safe to acquire the inode |
432 | mutex. Note some filesystems (i.e. remote ones) provide no | 432 | mutex or just to use i_size_read() instead. |
433 | protection for i_size so you will need to use the BKL. | 433 | Note: this does not protect the file->f_pos against concurrent modifications |
434 | since this is something the userspace has to take care about. | ||
434 | 435 | ||
435 | Note: ext2_release() was *the* source of contention on fs-intensive | 436 | Note: ext2_release() was *the* source of contention on fs-intensive |
436 | loads and dropping BKL on ->release() helps to get rid of that (we still | 437 | loads and dropping BKL on ->release() helps to get rid of that (we still |