Move GrSurfaceProxy::MakeLazy to GrProxyProvider::createLazy

This is pretty much a straight up move of the GrSurfaceProxy code with some plumbing to get the ProxyProvider in the right places.

Change-Id: I63cecb242dada503f97dbd1c0ce7ede75323100d
Reviewed-on: https://skia-review.googlesource.com/94200
Reviewed-by: Greg Daniel <egdaniel@google.com>
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>

1 files changed, 3 insertions, 2 deletions

diff --git a/src/gpu/GrReducedClip.h b/src/gpu/GrReducedClip.h
index f47b09ab64..c3838d8802 100644
--- a/src/gpu/GrReducedClip.h
+++ b/src/gpu/GrReducedClip.h
@@ -97,8 +97,9 @@ public:
      * the render target context, surface allocations, and even switching render targets (pre MDB)
      * may cause flushes or otherwise change which opList the actual draw is going into.
      */
-    std::unique_ptr<GrFragmentProcessor> finishAndDetachAnalyticFPs(uint32_t opListID, int rtWidth,
-                                                                    int rtHeight);
+    std::unique_ptr<GrFragmentProcessor> finishAndDetachAnalyticFPs(GrProxyProvider*,
+                                                                    uint32_t opListID,
+                                                                    int rtWidth, int rtHeight);
 
 private:
     void walkStack(const SkClipStack&, const SkRect& queryBounds);