Re: Hang on suspend

[Julian Blake Kongslie]

On Sun, 2008-02-24 at 23:35 +0100, Rafael J. Wysocki wrote:
> [Please don't drop the CC list from replies.]

Yeah, sorry, I accidentally whacked the wrong reply button.

> On Sunday, 24 of February 2008, Julian Blake Kongslie wrote:
> > On Sun, 2008-02-24 at 12:23 +0100, Rafael J. Wysocki wrote:
> > > Can you please check what's at the address platform_suspend_late+0x19?
> > 
> > 0xc0253ca9 is in platform_suspend_late (drivers/base/platform.c:579).
> > 574	{
> > 575		struct platform_driver *drv = to_platform_driver(dev->driver);
> > 576		struct platform_device *pdev;
> > 577		int ret = 0;
> > 578	
> > 579		pdev = container_of(dev, struct platform_device, dev);
> > 580		if (dev->driver && drv->suspend_late)
> > 581			ret = drv->suspend_late(pdev, mesg);
> > 582	
> > 583		return ret;
> 
> That's strange, it looks like the container_of() accessing beyond the
> structure.
> 
> Is this 100% reproducible?

Yes, it is. I've had this hang on suspend since at least some time in
early 2.6.24-rc, although because I have to transcribe the oops manually
I haven't had enough incentive to get it reported until now.
Unfortunately, I just don't need to suspend very often, so it may have
been introduced some time before then.

If you don't have any other ideas, I can try to bisect and find the
change that introduced it, but it'll probably take me a day or two; I
have other things I need to use my computer for, so it's a bit difficult
to do the constant-reboot-and-test cycle.

> Rafael

Thanks,

-- 
-Julian Blake Kongslie
<jblake@xxxxxxxxxxxxxxx>

If this is a mailing list, please CC me on replies.
vim: set ft=text :

Attachment: signature.asc
Description: This is a digitally signed message part