Hi, On Tue, Mar 1, 2016 at 10:26 AM, Wan-Teh Chang <wtc-at-google....@ffmpeg.org > wrote:
> I'd like to reiterate that this is difficult for many good > programmers. I think you just hit the nail right on the head :). I think my personal problem is that I don't quite understand what the actual problem is (your wiki page on memory ordering was quite helpful here, thank you!), and how the various release/acquire/memorybarrier/atomic functions help in solving it (I still find this mind-bogglingly complex, I'll slowly read through the documents you sent, but may take me a bit). The fundamental problem being that it's difficult to understand, combined with that we can't really grasp the severity of the issue (will files fail to decode? will ffmpeg hang/crash? or is tsan just complaining for no reason?), means that it's usually incredibly hard to do a good technical review of these patches, so review may be slow'ish, sadly... I'm trying my best but I'll make mistakes along the way. Ronald _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
