On Thu, Jul 22, 2021 at 09:03:45AM -0700, Lucas De Marchi wrote:
> On Thu, Jul 22, 2021 at 08:20:34AM +0100, Christoph Hellwig wrote:
> > On Thu, Jul 22, 2021 at 01:55:23AM -0400, Lucas De Marchi wrote:
> > > humn... PORT_F. KBL doesn't have PORT_F. We decided to keep the handling
> > > of DISPLAY_VER == 10 and DISPLAY_VER == 9 together and trust the VBT,
> > > but when the VBT is not present, DDI F will get added unconditio.
> > > 
> > > maybe best thing to do is to split that
> > 
> > With this and the previous patch the stack traces are gone.
> 
> maybe not the perfect fix though... can you share how you're starting
> qemu so I can try to reproduce it? I'd like to see a log with
> drm.debug=0xe param.
> 
> In theory 5a9d38b20a5a ("drm/i915/display: hide workaround for broken vbt in intel_bios.c")
> should filter out the PORT_F in KBL.

Here is my qemu script that reproduces the issue: