Implement support for mixed sampled render targets

src/gpu/effects/GrPorterDuffXferProcessor.cpp

Index: src/gpu/effects/GrPorterDuffXferProcessor.cpp
diff --git a/src/gpu/effects/GrPorterDuffXferProcessor.cpp b/src/gpu/effects/GrPorterDuffXferProcessor.cpp
index 4453c8935d84ec84c190b7be33f0737c31539371..d9c458d9a6baee99a3df844ba2139c87af5fa5b0 100644
--- a/src/gpu/effects/GrPorterDuffXferProcessor.cpp
+++ b/src/gpu/effects/GrPorterDuffXferProcessor.cpp
@@ -402,7 +402,8 @@ PorterDuffXferProcessor::internalGetOptimizations(const GrProcOptInfo& colorPOI,
     // has to read at all. If not, we'll disable blending.
     if (!hasCoverage) {
         if (dstCoeffIsZero) {
-            if (kOne_GrBlendCoeff == fSrcBlend) {
+            if (kOne_GrBlendCoeff == fSrcBlend &&
+                !colorPOI.needsCoverageModulation()) {

[bsalomon, 2015/03/31 16:37:33]

I know needCoverageModulation mirrors the extension language but in this context
I find it confusing. We use "coverage" to mean any fractional pixel coverage no
matter how it is implemented (stencil, fragment shader calculation) and we
always use it to "modulate" somewhere in the pipeline (glBlend or in the frag
shader).

Could we use something a little more descriptive?

applyStencilCoverageModulation? modulateByStencilCoverageFraction?

[vbuzinov, 2015/04/02 13:10:37]

Done.

                 // if there is no coverage and coeffs are (1,0) then we
                 // won't need to read the dst at all, it gets replaced by src
                 fDstBlend = kZero_GrBlendCoeff;