Change mapRectScaleTranslate to pass args/ret by value

src/core/SkMatrix.cpp

Index: src/core/SkMatrix.cpp
diff --git a/src/core/SkMatrix.cpp b/src/core/SkMatrix.cpp
index 0fd802087fc0c5f52e6da2be56edbfb1b9fcc442..f9c8c9d07e78f0ebf4f43ffc183ba27cef2fb5bb 100644
--- a/src/core/SkMatrix.cpp
+++ b/src/core/SkMatrix.cpp
@@ -1097,8 +1097,7 @@ void SkMatrix::mapVectors(SkPoint dst[], const SkPoint src[], int count) const {
     }
 }
 
-void SkMatrix::mapRectScaleTranslate(SkRect* dst, const SkRect& src) const {
-    SkASSERT(dst);
+SkRect SkMatrix::mapRectScaleTranslate(SkRect src) const {
     SkASSERT(this->isScaleTranslate());
     
     SkScalar sx = fMat[kMScaleX];
@@ -1107,22 +1106,25 @@ void SkMatrix::mapRectScaleTranslate(SkRect* dst, const SkRect& src) const {
     SkScalar ty = fMat[kMTransY];
     Sk4f scale(sx, sy, sx, sy);
     Sk4f trans(tx, ty, tx, ty);
-
+    
     Sk4f ltrb = Sk4f::Load(&src.fLeft) * scale + trans;
     // need to sort so we're not inverted
     Sk4f rblt(ltrb[2], ltrb[3], ltrb[0], ltrb[1]);
     Sk4f min = Sk4f::Min(ltrb, rblt);
     Sk4f max = Sk4f::Max(ltrb, rblt);
-    // We can extract either pair [0,1] or [2,3] from min and max and be correct, but on
-    // ARM this sequence generates the fastest (a single instruction).
-    Sk4f(min[2], min[3], max[0], max[1]).store(&dst->fLeft);
+    // We can extract either pair [0,1] or [2,3] from min and max and be correct.
+    // However, the current ABI for returning multiple floats is to use only 2 slots in each
+    // vector register. Thus we take [0..1] from min and max, as that perfectly matches the ABI.
+    SkRect dst;
+    Sk4f(min[0], min[1], max[0], max[1]).store(&dst.fLeft);
+    return dst;
 }
 
 bool SkMatrix::mapRect(SkRect* dst, const SkRect& src) const {
     SkASSERT(dst);
 
     if (this->isScaleTranslate()) {
-        this->mapRectScaleTranslate(dst, src);
+        *dst = this->mapRectScaleTranslate(src);
         return true;
     } else {
         SkPoint quad[4];