Messages in this thread | | | Date | Mon, 08 Jun 1998 11:26:59 +0200 (MET DST) | From | Pierre Etchemaite <> | Subject | non-critical problems with 2.0.34 |
| |
Suspecting some disk problems in a swap partition (this IBM udma disk worked hard as system disk under 2.0.33 until it developed a bad spot under an inode table...), I disabled it and run a badblocks -w. Quickly,
... kernel: VFS: brelse: Trying to free free buffer ... last message repeated 6 times
appeared in the logs and the badblocks process get stuck in __wait_on_buffer WCHAN (unkillable, of course, so the load is always >= 1). Beside that, the system looks fine.
That could be the result of the disk flakyness, but no disk error message appeared in the logs. Could it be a buffer cache/page cache problem 'cause the partition was used for swap previously ?
Other minor problems I get from time to time are :
- closing apps in xterm (say, closing xhifs), sometimes produces ... kernel: Warning: dev (04:c2) tty->count(1) != #fd's(2) in do_tty_hang This one has already been reported I think, but it's still not uncommon (I had to try only 4 times to reproduce it).
- under heavy load, ... kernel: EXT2-fs error (device 16:03): ext2_read_inode: bad inode number: 0 already happened 3 times. It looks harmless too, but the filesystem is marked erroneous and checked during next boot. It did not show under final .34 yet, but I guess it's a matter of time.
Again, nothing I can't live with, but worth reporting to make 2.0.x kernels even more bugfree.
-- Tired of (your) money and marketing driven computing ? Support open standards, and use Linux now! http://www.linux.org Homepage http://www.insat.com/~arnoud/petchemaite/
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu
| |