Messages in this thread | | | Date | Sat, 5 Jun 1999 14:23:22 -0700 (PDT) | From | Tim Smith <> | Subject | file locking starvation? |
| |
[I posted about this a couple of weeks ago, when this list was being very slow. I never saw it appear, so am assuming it got lost]
Consider this situation:
1. Process A locks a file, using fcntl(..., F_SETLKW, ... ).
2. Process B tries to lock the file the same way, and so goes to sleep waiting for the lock.
3. Process A releases the lock, but then immediately tries to reaquire it.
On Linux (2.2.9 is the only one I tested), process A gets the lock back in step #3. Anyone else think that step #3 should put A to sleep and let B get the lock? An application that wants that behavior can get it by throwing in a sleep(0) between A's releasing the lock and attempting to reaquire it, but is it supposed to have to do that?
--Tim Smith
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |