[Intel-gfx] [PATCH] drm/i915: Restore GT coarse power gating workaround

[Imre Deak]

On Thu, Nov 14, 2019 at 03:43:07PM +0000, Chris Wilson wrote:
> Quoting Imre Deak (2019-11-14 15:26:21)
> > The workaround to disable coarse power gating is still needed on SKL
> > GT3/GT4 machines and since the RC6 context corruption was discovered by
> > the hardware team also on all GEN9 machines. Restore applying the
> > workaround.
> > 
> > Fixes: c113236718e8 ("drm/i915: Extract GT render sleep (rc6) management")
> > Testcase: igt/intel_gt_pm_late_selftests/live_rc6_ctx
> > Cc: Chris Wilson <chris at chris-wilson.co.uk>
> > Cc: Andi Shyti <andi.shyti at intel.com>
> > Signed-off-by: Imre Deak <imre.deak at intel.com>
> > ---
> >  drivers/gpu/drm/i915/gt/intel_rc6.c | 6 ++++--
> >  1 file changed, 4 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/gt/intel_rc6.c b/drivers/gpu/drm/i915/gt/intel_rc6.c
> > index f7c0baeb3793..de2d302e8d55 100644
> > --- a/drivers/gpu/drm/i915/gt/intel_rc6.c
> > +++ b/drivers/gpu/drm/i915/gt/intel_rc6.c
> > @@ -175,8 +175,10 @@ static void gen9_rc6_enable(struct intel_rc6 *rc6)
> >             GEN6_RC_CTL_RC6_ENABLE |
> >             rc6_mode);
> >  
> > -       set(uncore, GEN9_PG_ENABLE,
> > -           GEN9_RENDER_PG_ENABLE | GEN9_MEDIA_PG_ENABLE);
> > +       /* WaRsDisableCoarsePowerGating:skl,cnl - Render/Media PG need to be disabled with RC6. */
> > +       if (!NEEDS_WaRsDisableCoarsePowerGating(rc6_to_i915(rc6)))
> > +               set(uncore, GEN9_PG_ENABLE,
> > +                   GEN9_RENDER_PG_ENABLE | GEN9_MEDIA_PG_ENABLE);
> 
> Eeek, I don't even recall noticing it got dropped.
> 
> Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk>
> 
> It might be more sensible as if (0) since gen9 is forever barred.

Hm, right.

The only trace of this issue I found now in BSpec is WA#0916
(Index/14091). That one has a different WA (disable GAFS clock gating),
so not sure if it's the same issue. I'll try to find the one the above
was added for.

Someone may also want to ignore the chance for RC6 CTX corruption and
(choosing better powersaving instead), so we may still want the above WA
to be optional (I'd need to add then a kconfig option for the CTX
corruption WA).

So for now I'd just keep the above form.

> -Chris