| Message ID | 57337E39.40105@unsolicited.net (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 11/05/16 19:47, David R wrote: > Hi > > Please consider applying the attached patch (or something like it) to > V4L2, and whatever is appropriate to the mainstream kernel. Without this > my media server crashes and burns at boot. > > See https://lkml.org/lkml/2016/5/7/88 for more details > > Thanks > David > I see the offending patch was reverted earlier today. My box is fine with my (more simple) alternative, but your call. David -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi David, Em Wed, 11 May 2016 20:05:22 +0100 David R <david@unsolicited.net> escreveu: > On 11/05/16 19:47, David R wrote: > > Hi > > > > Please consider applying the attached patch (or something like it) to > > V4L2, and whatever is appropriate to the mainstream kernel. Without this > > my media server crashes and burns at boot. > > > > See https://lkml.org/lkml/2016/5/7/88 for more details > > > > Thanks > > David > > > I see the offending patch was reverted earlier today. My box is fine > with my (more simple) alternative, but your call. Yes, I noticed the bug earlier today, while testing a DVB device. As this affects 2 stable releases plus the upcoming Kernel 4.6, I decided to just revert it for now, while we don't solve the issue. Your patch looks good. So, eventually it will be merged on a new version of this patch, after we test it properly. Thanks! Mauro -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi Mauro and David, On Wed, May 11, 2016 at 05:38:23PM -0300, Mauro Carvalho Chehab wrote: > Hi David, > > Em Wed, 11 May 2016 20:05:22 +0100 > David R <david@unsolicited.net> escreveu: > > > On 11/05/16 19:47, David R wrote: > > > Hi > > > > > > Please consider applying the attached patch (or something like it) to > > > V4L2, and whatever is appropriate to the mainstream kernel. Without this > > > my media server crashes and burns at boot. > > > > > > See https://lkml.org/lkml/2016/5/7/88 for more details > > > > > > Thanks > > > David > > > > > I see the offending patch was reverted earlier today. My box is fine > > with my (more simple) alternative, but your call. > > Yes, I noticed the bug earlier today, while testing a DVB device. > As this affects 2 stable releases plus the upcoming Kernel 4.6, > I decided to just revert it for now, while we don't solve the > issue. > > Your patch looks good. So, eventually it will be merged on a new > version of this patch, after we test it properly. Indeed vb2_core_dqbuf() is called with pb == NULL in file I/O, which was unfortunately missed until now. The proposed fix looks good to me.
--- drivers/media/v4l2-core/videobuf2-core.c.old 2016-05-11 18:44:50.917083559 +0100 +++ drivers/media/v4l2-core/videobuf2-core.c 2016-05-11 18:45:10.136184837 +0100 @@ -1665,7 +1665,7 @@ static int __vb2_get_done_vb(struct vb2_ * Only remove the buffer from done_list if v4l2_buffer can handle all * the planes. */ - ret = call_bufop(q, verify_planes_array, *vb, pb); + ret = pb ? call_bufop(q, verify_planes_array, *vb, pb) : 0; if (!ret) list_del(&(*vb)->done_entry); spin_unlock_irqrestore(&q->done_lock, flags);