User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a4pre) Gecko/20070329 Minefield/3.0a4pre Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a4pre) Gecko/20070329 Minefield/3.0a4pre Several calls to DrawSolidBorderSegment have a PRBool as the third arguments, when it takes appUnitsPerPixel (which is incorrectly named twipsPerPixel). Patch in a second. Reproducible: Always
Created attachment 260079 [details] [diff] [review] Simple fix. Simple fix. Unless someone was trying to be really creative... But then there should be a comment, and PR_TRUE/PR_FALSE shouldn't be used...
This code is in the midst of being rewritten, but we'll take this!
It seems like this bug ought to have exhibited itself in actual testcases.
(In reply to comment #4) > It seems like this bug ought to have exhibited itself in actual testcases. > I think this would have only resulted in the code taking a slow path. I don't fully understand it, but it looked like it was just testing to see if the segment was smaller than a pixel and if it was, just draw a block. The incorrect parameters would have caused those tests to fail.