"Juan J. Quintela" wrote:
>
> >>>>> "matt" == Matt Yourst <yourst@mit.edu> writes:
>
> matt> Well, pre8 is still broken here. It seems I'm having trouble using su in an X
> matt> session now (strangely, it works fine from the console even with X running.) A
> matt> kdb traceback reveals that the kernel gets deadlocked around
> matt> invalidate_inode_pages (and stalls again whenever anything else refers to the
> matt> locked-up su process.)
>
> Could you send the traces, please.
>
> Later, Juan.
>
I was mistaken about the problem being in invalidate_inode_pages; I meant
truncate_inode_pages instead (but this is irrelevant now, as a more telling
example is listed below.) Here's a kdb traceback of the su process caught in the
act of deadlocking. After I finally captured the data below, it appears that
reiserfs might possibly be implicated in this. However, I was running reiserfs
(the 3.6.5 release for kernel 2.3.99-pre6) well before upgrading to pre7 (when
the problem appeared.) Before jumping to conclusions though, I'll try to test
this with ext2 as well. Still, I'm betting that something in the Linux page
cache code is at fault, given the recent problems with this.
KDB stack trace of pid 713 (su):
c0113741 schedule+0x261
c01090a1 __down+0x61
c0109204 __down_failed+0x8
c010a72d error_code+0x2d
c01516f1 reiserfs_writepage+0x11 (but just calls calls block_write_full_page
like ext2 does)
c0123665 filemap_write_page+0x15
c01238fa filemap_sync+0x25a
c01239d6 filemap_unmap+0x16
c0121012 do_munmap+0x172
c01210ea sys_munmap+0x2a
c010a61c system_call
I realize this data might not be very helpful, so I'll post again if and when I
get any farther on tracking this down.
- Matt Yourst
-------------------------------------------------------------
Matt T. Yourst Massachusetts Institute of Technology
yourst@mit.edu 617.225.7690
513 French House - 476 Memorial Drive - Cambridge, MA 02136
-------------------------------------------------------------
-
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/
This archive was generated by hypermail 2b29 : Mon May 15 2000 - 21:00:23 EST