i915 deadlock on laptop undock

Daniel Vetter daniel at ffwll.ch
Tue Oct 8 22:22:25 CEST 2013


On Tue, Oct 8, 2013 at 6:08 PM, Roland Dreier <roland at kernel.org> wrote:
> On Mon, Oct 7, 2013 at 11:10 PM, Daniel Vetter <daniel at ffwll.ch> wrote:
>> Can you please boot with drm.debug=0xe, reproduce the issue and then
>> attach the dmesg. The additional debug spam should shed some light on how
>> we managed to get into this peculiar situation ...
>
> Sure, here it is.

As suspected a DP screen, and we disconnect right before a modeset
sequence. The unexpected thing here is that the modeset santizer kicks
in and wreaks utter havoc due to a lid event. Now we shouldn't hang
the entire box fundamentally when we unplug the DP cable, and that's a
bug in the DP code. It's also quite some work. But we also should be a
bit less enthusiastic with wreaking havoc in the lid noifier, and I
think that can be fixed quickly.

I'll reply with some patch.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch


More information about the dri-devel mailing list