On Mon, 2009-01-05 at 14:07 +0100, Gregoire Favre wrote:
> On Sun, Jan 04, 2009 at 02:28:24PM -0500, Andy Walls wrote:
>
> Hello,
>
> thank for your complete answer :
> http://www.linuxtv.org/pipermail/linux-dvb/2009-January/031264.html
>
> And for the one about "Kernel oops loading cx88 drivers when two
> WinTV-HVR4000 cards present" :
> http://www.linuxtv.org/pipermail/linux-dvb/2009-January/031230.html
Sure. I'm actually a sucker for looking at oops dumps. They're like
simple little puzzles waiting to be solved. Unfortunately, once I know
the "answer", I rarely follow through with the final solution.
> > I hope that helps someone. I'm not and expert on the cx88 modules and
> > their inter-relationships.
>
> For a user point of view : what has to be done ?
I you run across the oops often, then the suspected race condition in
the function I mentioned needs to be fixed. That may be as simple as
this lame patch:
diff -r 76c0ec8ab927 linux/drivers/media/video/cx88/cx88-mpeg.c
--- a/linux/drivers/media/video/cx88/cx88-mpeg.c Sun Jan 04 19:51:17
2009 -0500
+++ b/linux/drivers/media/video/cx88/cx88-mpeg.c Mon Jan 05 19:44:17
2009 -0500
@@ -831,6 +831,9 @@ static int __devinit cx8802_probe(struct
if (err != 0)
goto fail_free;
+ /* Maintain a reference so cx88-video can query the 8802 device. */
+ core->dvbdev = dev;
+
INIT_LIST_HEAD(&dev->drvlist);
list_add_tail(&dev->devlist,&cx8802_devlist);
@@ -851,20 +854,19 @@ static int __devinit cx8802_probe(struct
__func__);
videobuf_dvb_dealloc_frontends(&dev->frontends);
err = -ENOMEM;
+ /* FIXME - need to pull dev off cx8802_devlist*/
goto fail_free;
}
}
}
#endif
- /* Maintain a reference so cx88-video can query the 8802 device. */
- core->dvbdev = dev;
-
/* now autoload cx88-dvb or cx88-blackbird */
request_modules(dev);
return 0;
fail_free:
+ /* FIXME - shouldn't we pull dev off the cx8802_devlist - oops */
kfree(dev);
fail_core:
cx88_core_put(core,pci_dev);
> Maybe all this are related of the tuning problem I got with my system ?
Not if you don't run into the oops regularly. You either hit this race
condition when initializing the devices, or you don't and everything is
"normal".
Regards,
Andy
> Thanks.
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
