Re: showstopper race condition in sync() ??? [2.1.119|120]

Stephen C. Tweedie (sct@redhat.com)
Wed, 16 Sep 1998 21:01:47 +0100


Hi,

On Wed, 16 Sep 1998 08:44:38 +0200 (MET DST), Cyrille Chepelov
<chepelov@rip.ens-cachan.fr> said:

> Sorry, it does uses rename (see the attached strace). But the problem is
> not here : it is just after the rename() call, during the sync()
> call.

The activity just prior to the sync is not significant: the previous
filesystem calls are just writing back to the buffer cache, and all of
those changes don't hit disk until the sync().

Could you find the "wchan" of the stalled process and look it up in
System.map? The wchan is displayed in "ps -l" output, and will be
looked up automatically if the appropriate System.map is in /boot
(either as /boot/System.map-2.?.??? (fill in the blanks) or
/boot/System.map).

--Stephen

-
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/