aboutsummaryrefslogtreecommitdiffhomepage
path: root/include
diff options
context:
space:
mode:
authorGravatar reed <reed@google.com>2015-03-23 08:15:14 -0700
committerGravatar Commit bot <commit-bot@chromium.org>2015-03-23 08:15:14 -0700
commit8010632f2408ce9d840c1dcd26a8be0e4967233e (patch)
treed6b302cb048fee561cc401ed9f9340a4ba20bf87 /include
parent9c328187d9dee33736b77dc14dfb59529d948bb1 (diff)
remove meaningless matrix benches, add mapPts() and add new benches
mapPts definitely faster than mapPoints (identity and perspective same speed). Up to 3x for large values of N. cloned from https://codereview.chromium.org/1031443002/ BUG=skia: Review URL: https://codereview.chromium.org/1030653002
Diffstat (limited to 'include')
-rw-r--r--include/core/SkMatrix.h2
1 files changed, 2 insertions, 0 deletions
diff --git a/include/core/SkMatrix.h b/include/core/SkMatrix.h
index b2ee3c61bf..d00beda373 100644
--- a/include/core/SkMatrix.h
+++ b/include/core/SkMatrix.h
@@ -414,6 +414,8 @@ public:
this->mapPoints(pts, pts, count);
}
+ void mapPts(SkPoint dst[], const SkPoint src[], int count) const;
+
/** Like mapPoints but with custom byte stride between the points. Stride
* should be a multiple of sizeof(SkScalar).
*/