diff options
| author |  Yuqian Li <liyuqian@google.com> | 2017-09-15 17:39:09 -0400 |
|---|---|---|
| committer |  Skia Commit-Bot <skia-commit-bot@chromium.org> | 2017-09-18 18:10:03 +0000 |
| commit | 6d1aaca8276ff4ae2e10870f7e2c3222907cc4aa (patch) | |
| tree | a38b2a56507ec0b3906322bf842905ebfc88e7b2 /src/core/SkBlitter.cpp | |
| parent | e6c0dee067f6cdb84c98aab3bd9270c24b995084 (diff) | |
Call blitFatAntiRect to avoid overhead in MaskAdditiveBlitter
This results in 25% (720ns vs 560ns) speedup for
path_fill_small_rect bench in 8888 config. Some skps have a lot of stroked
horizontal/vertical lines (e.g., bar charts) so this improvement could
have a great impact there. For example, cereal converts Microsoft word docx
to PNGs on server and the sample docx has a big bar chart. That inspired
this improvement.
Bug: skia:
Change-Id: Icf96c966edf87427b3d1f53da09a49930eda2ac1
Reviewed-on: https://skia-review.googlesource.com/46584
Commit-Queue: Yuqian Li <liyuqian@google.com>
Reviewed-by: Mike Klein <mtklein@google.com>
Diffstat (limited to 'src/core/SkBlitter.cpp')
| -rw-r--r-- | src/core/SkBlitter.cpp | 5 |
1 files changed, 2 insertions, 3 deletions
diff --git a/src/core/SkBlitter.cpp b/src/core/SkBlitter.cpp index 5eb5dd80d0..8059f6ed57 100644 --- a/src/core/SkBlitter.cpp +++ b/src/core/SkBlitter.cpp @@ -45,9 +45,8 @@ inline static SkAlpha ScalarToAlpha(SkScalar a) { return (SkAlpha)(a * 255); } -void SkBlitter::blitFatAntiRect(const SkRect& rect) { - SkIRect bounds = rect.roundOut(); - SkASSERT(bounds.width() >= 3 && bounds.height() >= 3); +void SkBlitter::blitFatAntiRect(const SkRect& rect, const SkIRect& bounds) { + SkASSERT(rect.roundOut() == bounds && bounds.width() >= 3 && bounds.height() >= 3); int runSize = bounds.width() + 1; // +1 so we can set runs[bounds.width()] = 0 void* storage = this->allocBlitMemory(runSize * (sizeof(int16_t) + sizeof(SkAlpha))); |