[Intel-gfx] [PATCH] drm/i915: Only dump GPU state on set-wedged if interesting

[Chris Wilson]

Quoting Mika Kuoppala (2019-01-15 12:50:50)
> Chris Wilson <chris at chris-wilson.co.uk> writes:
> 
> > As we may frequently mark the device as wedged to flush requests off it
> > during the normal course of events, quite often we have a large state
> > dump that is of no interest. Don't bother dumping it all if the engines
> > are all idle.
> >
> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> > Cc: Mika Kuoppala <mika.kuoppala at linux.intel.com>
> > ---
> >  drivers/gpu/drm/i915/i915_gem.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> > index 90c167f71345..80264cb9ca7f 100644
> > --- a/drivers/gpu/drm/i915/i915_gem.c
> > +++ b/drivers/gpu/drm/i915/i915_gem.c
> > @@ -3192,7 +3192,7 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
> >  
> >       GEM_TRACE("start\n");
> >  
> > -     if (GEM_SHOW_DEBUG()) {
> > +     if (GEM_SHOW_DEBUG() && !intel_engines_are_idle(i915)) {
> 
> This atleast keeps the concurrency on reports at bay
> as wedged driver has only idle engines.

Nah, we are calling i915_gem_set_wedged() concurrently and get
interleaved reports for some errors :)

This is just making CI logging tidier by removing noise.
-Chris