Bug 1048141 Inquiry

[Seth Forshee]

On Tue, Oct 16, 2012 at 09:54:24AM -0400, Christopher M. Penalver wrote:
> * Please note this message has been sent to Ubuntu Bug Control mailing
> list already at
> https://lists.launchpad.net/ubuntu-bugcontrol/msg03774.html .
> 
> Dear Ubuntu Kernel Team / Ubuntu Bug Control:
> 
> Regarding https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1048141 ,
> I am interested to know if this is a "problem of a problem" bug.
> 
> I say this because in order for the Original Reporter to get the
> computer working to begin with, the OR has to boot with kernel
> parameter:
> acpi=noirq
> 
> I have questions about this bug:
> + Could acpi=noirq cause the new bug of not being able to boot?

The way I read the bug acpi=noirq *fixes* the problem of not being able
to boot. This is consistent with other reports I found in various
places.

Unless by "not being able to boot" you actually mean "boots to a blank
screen." I don't think this is related to acpi=noirq, as the OR points
out that his screen works with the 3.5.3 mainline build. The kernel in
use when the bug is reported is 3.5.0-14.15, which includes 3.5.3, so
it's likely that the problem is caused by one of the patches we're
carrying on 3.5.x. And it's likely the patch is a backport from 3.6
since the OR reports those mainline builds are also affected.

> + If acpi=noirq is causing this problem, wouldn't we want the bug?! of
> having to use acpi=noirq solved first, instead of troubleshooting
> collateral damage problems?

Of course it would be nice to have a solution to the acpi=noirq, but as
I said above I don't think this is a case of collateral damage.

Seth