Re: 2.6.9-rc3-mm2

[Hugh Dickins]

On Tue, 5 Oct 2004, Hugh Dickins wrote:
> On Tue, 5 Oct 2004, Peter Zijlstra wrote:
> > On Mon, 2004-10-04 at 19:13, Pasi Savolainen wrote:
> > > make: *** wait: No child processes.  Stop.
> > Just a simple mee too!
> 
> I've been suffering the occasional leftover zombie from multiple
> kernel builds precisely since the preempt-smp.patch went in; been
> hunting it unsuccessfully in spare moments, yesterday noticed that
> bug, today realize it's probably what I've been hunting - I'm
> about to start my own tests again, can't be sure until tomorrow.

Yes, that _raw_read_trylock patch has fixed my zombie problem: since
2.6.9-rc2-mm1 my repeated kernel builds would hang after several hours,
with a zombie (usually cc1) on one cpu and its waiting parent (usually
gcc) on another, with the puzzle of why the parent had never got woken -
because do_wait's read_lock(&tasklist_lock) went ahead even while
exit_notify held the write lock.

Hugh

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/