Reland "Stack-allocate pipelines for GrMeshDrawOp"

This is a reland of dfe5000a5ff2cabb50ddc139882dc1f3005fa429

Original change's description:
> Stack-allocate pipelines for GrMeshDrawOp
>
> Stack-allocates the pipelines in onExecute. This saves us from having
> to store the pipelines on the heap, as well as delaying the need to
> detach processors until onExecute. The delay is an improvement because
> it allows us to keep visiting proxies after onPrepare. (Previously,
> they were moved out of GrProcessorSet and into a pipeline during
> onPrepare, so visiting proxies was impossible after that point.)
>
> Bug: skia:8731
> Change-Id: Idc05063fb0dfbfed42b434e429fa5a497097bdae
> Reviewed-on: https://skia-review.googlesource.com/c/193368
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Robert Phillips <robertphillips@google.com>

TBR=robertphillips@google.com

Bug: skia:8731
Change-Id: I32def1a35bb0593470fa672691a9e697dc6d9680
Reviewed-on: https://skia-review.googlesource.com/c/195261
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
diff --git a/src/gpu/ops/GrAAConvexPathRenderer.cpp b/src/gpu/ops/GrAAConvexPathRenderer.cpp
index cb9d364..1e44b87 100644
--- a/src/gpu/ops/GrAAConvexPathRenderer.cpp
+++ b/src/gpu/ops/GrAAConvexPathRenderer.cpp
@@ -713,7 +713,6 @@
 
 private:
     void onPrepareDraws(Target* target) override {
-        auto pipe = fHelper.makePipeline(target);
         int instanceCount = fPaths.count();
 
         SkMatrix invert;
@@ -794,11 +793,14 @@
                 firstIndex += draw.fIndexCnt;
                 firstVertex += draw.fVertexCnt;
             }
-            target->draw(quadProcessor, pipe.fPipeline, pipe.fFixedDynamicState, nullptr, meshes,
-                         draws.count());
+            target->recordDraw(quadProcessor, meshes, draws.count());
         }
     }
 
+    void onExecute(GrOpFlushState* flushState, const SkRect& chainBounds) override {
+        fHelper.executeDrawsAndUploads(this, flushState, chainBounds);
+    }
+
     CombineResult onCombineIfPossible(GrOp* t, const GrCaps& caps) override {
         AAConvexPathOp* that = t->cast<AAConvexPathOp>();
         if (!fHelper.isCompatible(that->fHelper, caps, this->bounds(), that->bounds())) {