http://bugs.winehq.org/show_bug.cgi?id=4806
------- Additional Comments From Speeddymon@gmail.com 2006-25-03 13:23 ------- I honestly don't think that that patch will get committed to wine.
The way wine is running with that line of code is that it returns an error if the file it is trying to read is smaller than the header of that same file. The fact that wine is picking up a header that is bigger than the size of the file being read astonishes me, as it is not possible.
What that means is that we need to figure out where wine is really pulling the number it is filling header_size with and fix that.. If that number is correct, then we need a test case to show that that line of code really is the problem, and not that it just exposes a problem somewhere else. It should be obvious by my last statement, but if the test case causes the same error under windows, then we know we are in conformance, and the problem lies somewhere else (wherever wine is getting header_size from)..
I forwarded this bug to wine-devel, so hopefully someone with more experience than myself can offer some input.