[Intel-gfx] [PATCH 1/3] drm/i915: Capture dmc firmware information before reset

[Michel Thierry]

On 2/24/2017 11:34 AM, Chris Wilson wrote:
> On Fri, Feb 24, 2017 at 11:25:29AM -0800, Michel Thierry wrote:
>> On 2/24/2017 11:22 AM, Chris Wilson wrote:
>>> On Fri, Feb 24, 2017 at 11:05:28AM -0800, Michel Thierry wrote:
>>>> The firmware may change between the hang and cat /sys/class/drm/card0/error
>>>>
>>>> Cc: Chris Wilson <chris at chris-wilson.co.uk>
>>>> Signed-off-by: Michel Thierry <michel.thierry at intel.com>
>>>
>>> Looks sane...
>>>
>>>> ---
>>>> drivers/gpu/drm/i915/i915_drv.h       |  4 ++++
>>>> drivers/gpu/drm/i915/i915_gpu_error.c | 22 ++++++++++++++++------
>>>> 2 files changed, 20 insertions(+), 6 deletions(-)
>>>>
>>>> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
>>>> index a74b87b1b5a9..3b0cff05f5f7 100644
>>>> --- a/drivers/gpu/drm/i915/i915_drv.h
>>>> +++ b/drivers/gpu/drm/i915/i915_drv.h
>>>> @@ -953,6 +953,10 @@ struct i915_gpu_state {
>>>> 	u32 gab_ctl;
>>>> 	u32 gfx_mode;
>>>>
>>>> +	/* Firmware load state */
>>>> +	u32 dmc_loaded;
>>>
>>> Can dmc_version be zero? i.e. Does !0 => dmc_loaded?
>>>
>>
>> Nothing would stop them to build a v0.0 firmware, but in this case,
>> we can save one variable.
>
> On the other guc thread, the suggestion is to use 0 as an invalid
> version...
> -Chris
>
ok, let's go with that, I'll remove *_loaded.