RESOLVED FIXED 157779
BitmapImage::destroyDecodedDataIfNecessary() should only count frames with image data
https://bugs.webkit.org/show_bug.cgi?id=157779
Summary BitmapImage::destroyDecodedDataIfNecessary() should only count frames with im...
Simon Fraser (smfr)
Reported 2016-05-16 18:42:52 PDT
BitmapImage::destroyDecodedDataIfNecessary() should only count frames with image data
Attachments
Patch (3.44 KB, patch)
2016-05-16 18:51 PDT, Simon Fraser (smfr)
thorton: review+
Simon Fraser (smfr)
Comment 1 2016-05-16 18:51:00 PDT
Said Abou-Hallawa
Comment 2 2016-05-16 19:05:08 PDT
Comment on attachment 279080 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=279080&action=review Unofficial r=me > Source/WebCore/platform/graphics/BitmapImage.h:90 > + unsigned usedFrameBytes() const { return m_image ? m_frameBytes : 0; } Do we need to call it usedFrameBytes()? Should not it be just named frameBytes()? I think it supposed to return the bytes of the image stored in this frame. So if the image is nil the frameBytes should be zero.
Simon Fraser (smfr)
Comment 3 2016-05-16 19:07:09 PDT
(In reply to comment #2) > Comment on attachment 279080 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=279080&action=review > > Unofficial r=me > > > Source/WebCore/platform/graphics/BitmapImage.h:90 > > + unsigned usedFrameBytes() const { return m_image ? m_frameBytes : 0; } > > Do we need to call it usedFrameBytes()? Should not it be just named > frameBytes()? I think it supposed to return the bytes of the image stored in > this frame. So if the image is nil the frameBytes should be zero. I think it needs to be disambiguate from height*width*4 which we get from metadata. In reality, all callers but the one I changed check m_image too.
Simon Fraser (smfr)
Comment 4 2016-05-17 14:28:19 PDT
Note You need to log in before you can comment on or make changes to this bug.