Message ID | 20170807164842.26802-1-chris@chris-wilson.co.uk (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Chris Wilson <chris@chris-wilson.co.uk> writes: > In the null state batch, we try to load the CC_STATE_POINTERS, but pass > in a pointer to invalid state for the color-calc and depth-stencil > state. In the rendercopy batch this was imported from, the 1024 offset > used is known to be 64bytes of zeroes. Tweaking the > gen6_null_state_batch to point those two offset at known zeroes however > was not sufficient. > > The effect of loading depth-stencil and color-calc causes the golden > renderstate batch to take 10s (yes, ten seconds) on my snb gt1 (i5-2500). > > It would be nice if we could drop this incomplete "golden render state" > entirely. > Agreed. But until then, lets do changes through igt/tools/null_state_gen. Patch posted. -Mika > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> > Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> > --- > drivers/gpu/drm/i915/intel_renderstate_gen6.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/gpu/drm/i915/intel_renderstate_gen6.c b/drivers/gpu/drm/i915/intel_renderstate_gen6.c > index 11c8e7b3dd7c..46cfc405ca30 100644 > --- a/drivers/gpu/drm/i915/intel_renderstate_gen6.c > +++ b/drivers/gpu/drm/i915/intel_renderstate_gen6.c > @@ -107,8 +107,8 @@ static const u32 gen6_null_state_batch[] = { > 0x00000000, > 0x780e0002, > 0x00000441, > - 0x00000401, > - 0x00000401, > + 0x00000400, > + 0x00000400, > 0x78021002, > 0x00000000, > 0x00000000, > -- > 2.13.3
diff --git a/drivers/gpu/drm/i915/intel_renderstate_gen6.c b/drivers/gpu/drm/i915/intel_renderstate_gen6.c index 11c8e7b3dd7c..46cfc405ca30 100644 --- a/drivers/gpu/drm/i915/intel_renderstate_gen6.c +++ b/drivers/gpu/drm/i915/intel_renderstate_gen6.c @@ -107,8 +107,8 @@ static const u32 gen6_null_state_batch[] = { 0x00000000, 0x780e0002, 0x00000441, - 0x00000401, - 0x00000401, + 0x00000400, + 0x00000400, 0x78021002, 0x00000000, 0x00000000,
In the null state batch, we try to load the CC_STATE_POINTERS, but pass in a pointer to invalid state for the color-calc and depth-stencil state. In the rendercopy batch this was imported from, the 1024 offset used is known to be 64bytes of zeroes. Tweaking the gen6_null_state_batch to point those two offset at known zeroes however was not sufficient. The effect of loading depth-stencil and color-calc causes the golden renderstate batch to take 10s (yes, ten seconds) on my snb gt1 (i5-2500). It would be nice if we could drop this incomplete "golden render state" entirely. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> --- drivers/gpu/drm/i915/intel_renderstate_gen6.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)