PNGImageDEcoder: Handle interlace buffer allocation failure
Created attachment 139617 [details] Patch
Comment on attachment 139617 [details] Patch Do we have an image we can test this with? It seems if we keep maintaining a PNG decoder (which we really *don't* want to do as a project), we will need to come up with a testing harness and a corpus of a bunch of broken png files. :(
No, no test image sorry. Agree we need to come up with a better test suite.
So it's impractical to test this with our current technology, but the plan is to move towards using pngsuite for testing our decoder, which should catch issues like this one. http://www.schaik.com/pngsuite
Comment on attachment 139617 [details] Patch OK.
Comment on attachment 139617 [details] Patch Clearing flags on attachment: 139617 Committed r115784: <http://trac.webkit.org/changeset/115784>
All reviewed patches have been landed. Closing bug.
(In reply to comment #4) > So it's impractical to test this with our current technology, but the plan is to move towards using pngsuite for testing our decoder, which should catch issues like this one. http://www.schaik.com/pngsuite "may help catch issues like this one", filed bug 85460
Closed bug 85460 (too much broken style bot noise). Filed bug 85567.