So that nsSVGMaskFrame::GetMaskForMaskedFrame does not need to create a new A8 surface again.
In nsSVGMaskFrame::GetMaskForMaskedFrame 1. The size of maskDT has to be matched with the A8 surface, aParams.maskDT, that we pass int. So that we can map maskDT pixels onto aParams.maskDT. Doing it will create a bigger RGBA surface and hurt performance. This is a reason that we probably should stop working on this bug. 2. ComputesRGBLuminanceMask_NEON is not handle yet. 3. Benchmark data need(https://pastebin.mozilla.org/8915569)
No obvious perf improvement
Status: NEW → RESOLVED
Last Resolved: 2 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.