RESOLVED DUPLICATE of bug 146147102011
[TEXMAP] Check imageBuffer Creation failure when updateBackingStore
https://bugs.webkit.org/show_bug.cgi?id=102011
Summary [TEXMAP] Check imageBuffer Creation failure when updateBackingStore
ys.suh
Reported 2012-11-12 16:57:35 PST
[TEXMAP] Check imageBuffer Creation failure when updateBackingStore
Attachments
Patch (1.76 KB, patch)
2012-11-12 17:11 PST, ys.suh
noam: review-
ys.suh
Comment 1 2012-11-12 17:11:03 PST
Noam Rosenthal
Comment 2 2012-11-12 17:28:37 PST
Comment on attachment 173763 [details] Patch The right way to fix it is to tile the dirtyRects before uploading them. Your fix will create bugs along the way, like "Why does this layer not appear correctly on the screen" :)
ys.suh
Comment 3 2012-11-12 18:00:42 PST
(In reply to comment #2) > (From update of attachment 173763 [details]) > The right way to fix it is to tile the dirtyRects before uploading them. > Your fix will create bugs along the way, like "Why does this layer not appear correctly on the screen" :) I totally agree with you. But in my opinion, not appearing layer would be better than crashing. How about clipping the layer to some size just before calling setCompositedBounds in RenderLayerBacking::updateCompositedBounds?
Noam Rosenthal
Comment 4 2012-11-13 14:15:59 PST
(In reply to comment #3) > (In reply to comment #2) > > (From update of attachment 173763 [details] [details]) > > The right way to fix it is to tile the dirtyRects before uploading them. > > Your fix will create bugs along the way, like "Why does this layer not appear correctly on the screen" :) > > I totally agree with you. But in my opinion, not appearing layer would be better than crashing. Maybe for a production branch - in webkit.org trunk we hold a higher standard - meaning a patch has to take things forward, not replace a bug with one symptom (crash) with a bug with another symptom (layers not appearing). > > How about clipping the layer to some size just before calling setCompositedBounds in RenderLayerBacking::updateCompositedBounds? It's another hack. I will only accepts patches that take this code forward - workarounds only in extreme cases where the "good" solution is not trivial.
Fujii Hironori
Comment 5 2022-06-21 20:12:21 PDT
*** This bug has been marked as a duplicate of bug 146147 ***
Note You need to log in before you can comment on or make changes to this bug.