Re: [PATCH] Relay : fix cpu offline problem

From: Mathieu Desnoyers
Date: Tue Dec 09 2008 - 20:54:47 EST


Ah, great, I did not see it passing by. Thanks !

Mathieu

* Lai Jiangshan (laijs@xxxxxxxxxxxxxx) wrote:
> Hi, Mathieu Desnoyers
>
> I have fixed this problem in mainline.
> http://lkml.org/lkml/2008/11/14/74
>
> Lai
>
> Mathieu Desnoyers wrote:
> > Here is a fix for how relay handles cpu hotplug in the allocation error path. If
> > a cpu goes down while allocating, it may result in a buffer never being freed.
> >
> > Lai Jiangshan originally identified this problem in the LTTng fork of
> > relay (ltt-relay-alloc). Here is the fix which applies to the original
> > kernel/relay.c.
> >
> > Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxx>
> > CC: Lai Jiangshan <laijs@xxxxxxxxxxxxxx>
> > CC: Tom Zanussi <zanussi@xxxxxxxxxxx>
> > ---
> > kernel/relay.c | 7 +++----
> > 1 file changed, 3 insertions(+), 4 deletions(-)
> >
> > Index: linux-2.6-lttng/kernel/relay.c
> > ===================================================================
> > --- linux-2.6-lttng.orig/kernel/relay.c 2008-12-09 10:58:45.000000000 -0500
> > +++ linux-2.6-lttng/kernel/relay.c 2008-12-09 11:09:58.000000000 -0500
> > @@ -611,10 +611,9 @@ struct rchan *relay_open(const char *bas
> > return chan;
> >
> > free_bufs:
> > - for_each_online_cpu(i) {
> > - if (!chan->buf[i])
> > - break;
> > - relay_close_buf(chan->buf[i]);
> > + for_each_possible_cpu(i) {
> > + if (chan->buf[i])
> > + relay_close_buf(chan->buf[i]);
> > }
> >
> > kref_put(&chan->kref, relay_destroy_channel);
>
>

--
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68
--
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/