diff options
author | msarett <msarett@google.com> | 2016-06-03 08:25:21 -0700 |
---|---|---|
committer | Commit bot <commit-bot@chromium.org> | 2016-06-03 08:25:22 -0700 |
commit | 2cee902847a940c4bab56b42dabbb3721be3f9ac (patch) | |
tree | 528c0fdd2139c95f5edf741db7480d977df4f497 /include/codec | |
parent | 577967c155fe3e2ce8251b053da60bd0d8fb32f4 (diff) |
Add color correction benchmark - with comparison to qcms
--colorImages is empty by default so this won't affect the runtime of
the bots. To run locally, use --colorImages <path> and
--benchType skcolorcodec.
Two takeaways so far:
Color correction is (currently) slower than jpeg decodes.
Our reference solution is slower than qcms.
TBR=reed@google.com
BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2035793002
Review-Url: https://codereview.chromium.org/2035793002
Diffstat (limited to 'include/codec')
-rw-r--r-- | include/codec/SkCodec.h | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/include/codec/SkCodec.h b/include/codec/SkCodec.h index f931f2bb12..d2e101dea4 100644 --- a/include/codec/SkCodec.h +++ b/include/codec/SkCodec.h @@ -27,6 +27,7 @@ namespace DM { class CodecSrc; class ColorCodecSrc; } +class ColorCodecBench; /** * Abstraction layer directly on top of an image codec. @@ -776,8 +777,10 @@ private: virtual SkSampler* getSampler(bool /*createIfNecessary*/) { return nullptr; } // For testing with qcms - // FIXME: Remove this when we are done comparing with qcms. + // FIXME: Remove these when we are done comparing with qcms. friend class DM::ColorCodecSrc; + friend class ColorCodecBench; + friend class DM::CodecSrc; // for fillIncompleteImage friend class SkSampledCodec; friend class SkIcoCodec; |