Hi Andreas, On 2015-10-24 21:44, Andreas Cadhalpun wrote: > I've discussed this with ffmpeg upstream developers and it turns out > that fixing the assembler code was feasible, after all. > It is now fixed upstream in commit 52f84d8 [1]. > Actually, it had been reported before in above trac ticket, but > couldn't be reproduced. Thanks again for providing the reproducer! > > So this will be fixed with the ffmpeg 2.8.2 upload, but that might > take a while.
That is great news. Thank you once again for taking action so quickly, for the great analysis and for talking to the upstream developers! Best regards, Michael
