[Intel-gfx] [PATCH 1/5] drm/i915: Add control flags to i915_handle_error()

[Michel Thierry]

On 3/19/2018 5:44 PM, Chris Wilson wrote:
> Quoting Michel Thierry (2018-03-20 00:39:35)
>> On 3/19/2018 5:18 PM, Chris Wilson wrote:
>>> Not all callers want the GPU error to handled in the same way, so expose
>>> a control parameter. In the first instance, some callers do not want the
>>> heavyweight error capture so add a bit to request the state to be
>>> captured and saved.
>>>
>>> v2: Pass msg down to i915_reset/i915_reset_engine so that we include the
>>> reason for the reset in the dev_notice(), superseding the earlier option
>>> to not print that notice.
>>>
>>> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
>>> Cc: Jeff McGee <jeff.mcgee at intel.com>
>>> Cc: Mika Kuoppala <mika.kuoppala at intel.com>
>>> Cc: Michel Thierry <michel.thierry at intel.com>
>>> ---
>>>    drivers/gpu/drm/i915/i915_debugfs.c              |  4 +--
>>>    drivers/gpu/drm/i915/i915_drv.c                  | 17 +++++------
>>>    drivers/gpu/drm/i915/i915_drv.h                  | 10 +++---
>>>    drivers/gpu/drm/i915/i915_irq.c                  | 39 +++++++++++++-----------
>>>    drivers/gpu/drm/i915/intel_hangcheck.c           | 13 ++++----
>>>    drivers/gpu/drm/i915/selftests/intel_hangcheck.c | 13 +++-----
>>>    6 files changed, 48 insertions(+), 48 deletions(-)
>>>
>> ...
>>> diff --git a/drivers/gpu/drm/i915/intel_hangcheck.c b/drivers/gpu/drm/i915/intel_hangcheck.c
>>> index 42e45ae87393..fd0ffb8328d0 100644
>>> --- a/drivers/gpu/drm/i915/intel_hangcheck.c
>>> +++ b/drivers/gpu/drm/i915/intel_hangcheck.c
>>> @@ -246,9 +246,8 @@ engine_stuck(struct intel_engine_cs *engine, u64 acthd)
>>>         */
>>>        tmp = I915_READ_CTL(engine);
>>>        if (tmp & RING_WAIT) {
>>> -             i915_handle_error(dev_priv, 0,
>>> -                               "Kicking stuck wait on %s",
>>> -                               engine->name);
>>> +             i915_handle_error(dev_priv, BIT(engine->id), 0,
>>> +                               "stuck wait on %s", engine->name);
>> Before we were not resetting anything here, is this change on purpose?
>> (if it is, it's worth adding it to the commit msg since it's changing
>> behavior).
>>
>>>                I915_WRITE_CTL(engine, tmp);
>>>                return ENGINE_WAIT_KICK;
>>>        } > @@ -258,8 +257,8 @@ engine_stuck(struct intel_engine_cs *engine, u64
>> acthd)
>>>                default:
>>>                        return ENGINE_DEAD;
>>>                case 1:
>>> -                     i915_handle_error(dev_priv, 0,
>>> -                                       "Kicking stuck semaphore on %s",
>>> +                     i915_handle_error(dev_priv, ALL_ENGINES, 0,
>> Same here,
> 
> Both are functionally no-op changes, as they are only for !per-engine
> platforms (unless someone manages to send just the wrong type of garbage
> to the GPU). I just thought it interesting to document that wait-event
> needs a local kick and the wait-sema needs to kick the other engines.
i915_handle_error has this before full reset:

	if (!engine_mask)
		goto out;

No reset at all was happening before.