I'm sure I'm using the wrong terminology, but we have a number of OOM crashes in AppendUTF8toUTF16 that are not due to the bug in that code, but the fact that somebody calls it with "unreasonable" requests to convert strings that are too large. It is possible that some of the callers can do something smarter, so is it worth basing the crash reports on the callers of AppendUTF8toUTF16, rather than AppendUTF8toUTF16 itself. Do that skip one magic that we do.
We should add it to the prefix list then so that the next frame is being added to the signature. Are there bugs reported on crashes that will be affected by this?
Summary: Should skip AppendUTF8toUTF16 in reports → Add AppendUTF8toUTF16 to prefix skiplist
There's a good handful of these string helpers: https://dxr.mozilla.org/mozilla-central/source/xpcom/string/nsReadableUtils.h#35 Maybe we could do something like this? (Lossy)?(Copy|Append).*UTF.*
(In reply to Robert Kaiser (:email@example.com) from comment #1) > > Are there bugs reported on crashes that will be affected by this? I filed this while looking at bug 1033090, and then there is also bug 996303 (and an oldie bug 511135.)
I found some others too: NS_LossyConvertUTF16toASCII, NS_ConvertUTF8toUTF16::NS_ConvertUTF8toUTF16, etc. So how about... (NS_)?(Lossy)?(Copy|Append|Convert).*UTF.* Lars, could you add something like that to the list please?
QA Contact: lars
Assignee: nobody → lars
QA Contact: lars
Commit pushed to master at https://github.com/mozilla/socorro https://github.com/mozilla/socorro/commit/a260655bb8b292a7b303205a1b640ed3c5633679 Merge pull request #2846 from twobraids/utf8-utf16-wtf32 Fixes Bug 1130960 - added new RE to prefix list for UTF signatures
Status: NEW → RESOLVED
Last Resolved: 4 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.