Re: [Intel-gfx] linux-next: Tree for Jul 25 [ call-trace: drm | drm-intel related? ]

[Sedat Dilek <sedat.dilek@gmail.com>]

On Fri, Jul 26, 2013 at 11:02 AM, Sedat Dilek <sedat.dilek@gmail.com> wrote:
> On Fri, Jul 26, 2013 at 10:52 AM, Daniel Vetter <daniel.vetter@ffwll.ch> wrote:
>> On Fri, Jul 26, 2013 at 10:50 AM, Chris Wilson <chris@chris-wilson.co.uk> wrote:
>>> On Fri, Jul 26, 2013 at 10:27:03AM +0200, Sedat Dilek wrote:
>>>> On Fri, Jul 26, 2013 at 10:25 AM, Sedat Dilek <sedat.dilek@gmail.com> wrote:
>>>> > ...
>>>> > ... but does not start as well, so it seems to be a kernel-issue as
>>>> > assumed (2nd confirmation).
>>>> >
>>>> > X.log attached.
>>>> >
>>>>
>>>> Now, really w/ promised attachment.
>>>
>>> Yes, same failure (GTT mmaps) but at a later point, and UXA has no
>>> fallback plan.
>>
>> I'm running igt on my machines here to prep a new -next test cycle,
>> and gtt mmaps seem to fail across the board :( Currently I'd wager
>> that the vma offset manager is the culprit, since I've only recently
>> pulled that stuff in from drm-next.
>
> So, I resetted next-20130726 to the merge-commit of
> drm-intel/for-linux-next, which is as expected bad.
>
> First, I looked w/ a Grrr at i915_gem_stolen.c with a focus on
> "drm/i915: Use Graphics Base of Stolen Memory on all gen3+", but this
> commit is also in next-20130725.
>
> Yeah, might be the stuff comes from the merge including vma stuff.
>
> - Sedat -
>
> [1] http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/commit/drivers/gpu/drm/i915/i915_gem_stolen.c?id=4867509d9269228f69273312298fea11f414ae54
>

I have compared next-20130725 VS. next-20130726:

$ head -2313 next-20130725-VS-next-20130726.diff | grep ^+ | grep i915
+      drm/i915: Colocate all GT access routines in the same file
+      drm/i915: Use a private interface for register access within GT
+      drm/i915: Use the common register access functions for NOTRACE variants
+      drm/i915: Squash gen lookup through multiple indirections
inside GT access
+      drm/i915: Convert the register access tracepoint to be conditional
+      drm/i915: fix the racy object accounting
+      drm/i915: dvo_ch7xxx: fix vsync polarity setting
+      drm/i915: initialize gt_lock early with other spin locks
+      Revert "ACPI / video / i915: No ACPI backlight if firmware
expects Windows 8"
+ drivers/gpu/drm/i915/Makefile                      |    2 +
+ drivers/gpu/drm/i915/dvo_ch7xxx.c                  |    2 +-
+ drivers/gpu/drm/i915/i915_debugfs.c                |  836 +--
+ drivers/gpu/drm/i915/i915_dma.c                    |   91 +-
+ drivers/gpu/drm/i915/i915_drv.c                    |  292 +-
+ drivers/gpu/drm/i915/i915_drv.h                    |  413 +-
+ drivers/gpu/drm/i915/i915_gem.c                    |  411 +-
+ drivers/gpu/drm/i915/i915_irq.c                    | 1347 ++---
+ drivers/gpu/drm/i915/i915_trace.h                  |   16 +-
+ drivers/gpu/drm/i915/intel_display.c               | 1065 ++--
+ drivers/gpu/drm/i915/intel_drv.h                   |   36 +-
+ drivers/gpu/drm/i915/intel_pm.c                    |  431 +-
+ drivers/gpu/drm/i915/intel_uncore.c                |  595 +++

Any hints which of the commits could be culprit?

- Sedat -

> - Sedat -
>
>> -Daniel
>> --
>> Daniel Vetter
>> Software Engineer, Intel Corporation
>> +41 (0) 79 365 57 48 - http://blog.ffwll.ch