ImageBitmap should report its memory cost
Created attachment 345690 [details]
Comment on attachment 345690 [details]
This should be testable somehow, right?
(In reply to Simon Fraser (smfr) from comment #2)
> Comment on attachment 345690 [details]
> This should be testable somehow, right?
I don't know, I can't find a comparable test for e.g. HTMLCanvasElement.
The changes here primarily increase the reported memory cost of ImageBitmap objects, bringing down the delay in which GC is triggered. This helps destroying stale ImageBitmaps and underlying pixel data. Any pointers about how to test this are welcome.
Created attachment 442145 [details]
Taking the liberty to taking the bug, feel free to comment if you disagree.
For the test I could only find internals.memoryInfo() but didn't feel confident it was working since none of the tests use it..
I can add the test but I uploaded the patch if it's useful for bug 232139
Committed r284772 (243481@main): <https://commits.webkit.org/243481@main>
All reviewed patches have been landed. Closing bug and clearing flags on attachment 442145 [details].