If this is caught in the unit tests ~1/30 times, then it's happening despite the window positionings and view positionings being the same. There's multiple layers of indirection in there (two context types, four libraries) all totally closed source. Tracking it down feels like it would take way too much effort and I'm swamped. If you have some spare time...
Avi On Tue, Sep 22, 2009 at 11:56 AM, Erik Kay <[email protected]> wrote: > I'd suspect an alignment / positioning bug for what you're seeing. > Often rect fill algorithms have several paths with different loop > unrolling tricks based on the size and position of the rect. I agree > with Evan that it may be worth tracking this down a bit more. Even if > it's not our bug, we need to find a way to avoid the memory stomping. > I'm nervous about adding this to the upstream suppression list. I > think that's OK to do for memory leaks, or for memory errors where > it's been demonstrated that the result of the error is benign (like > the UMRs in parts of Microsoft's STL implementation), but it doesn't > seem like this fits into that case. > > Erik > > > On Mon, Sep 21, 2009 at 1:15 PM, Avi Drissman <[email protected]> wrote: > > I have no evidence to confirm/deny that. Even so it deserves an > upstreaming. > > I'll look at it but why would it show up 1/30 times? > > > > Avi > > > > On Mon, Sep 21, 2009 at 4:07 PM, Evan Martin <[email protected]> wrote: > >> > >> Could it possibly be related to passing a zero-sized rect in somewhere? > >> > >> On Mon, Sep 21, 2009 at 12:27 PM, Avi Drissman <[email protected]> wrote: > >> > crbug.com/18189 > >> > crbug.com/18539 > >> > > >> > I got the first because it involved the status bubble; I got the > second > >> > because I got the first. > >> > > >> > NSRectFill(). Deep down that ends up in sseCGSFill8by1, which looks > like > >> > it > >> > sometimes scribbles off the end of some buffer. I have no idea what we > >> > could > >> > be doing wrong to cause it nor what we could be doing to affect it at > >> > all. I > >> > want to just dup one to the other and mark both as > >> > CANNOTFIXBADAPPLECODE^WWontFix. Any objections? > >> > > >> > Avi > >> > > >> > > > >> > > > > > > > > > > > > --~--~---------~--~----~------------~-------~--~----~ Chromium Developers mailing list: [email protected] View archives, change email options, or unsubscribe: http://groups.google.com/group/chromium-dev -~----------~----~----~----~------~----~------~--~---
