On 23/04/18 10:11, Carl Eugen Hoyos wrote: > 2018-03-14 7:31 GMT+01:00, Martin Vignali <martin.vignali at gmail.com>: > >> In that case we can let the test using "none" >> compression (bypass the snappy part) > > These tests are also broken, please fix or > remove them: > https://buildd.debian.org/status/fetch.php?pkg=ffmpeg&arch=i386&ver=7%3A4.0-1&stamp=1524444218&raw=0 > ("Error 1")
I've had a brief look at this error (and a similar error on s390x) and it looks like a float rounding issue in some of the functions in libavc/texturedspenc.c. The output from the hap encoder is only different by a few bits. i386 fails because it promotes floats to long double when evaluating (at least on Debian which has SSE disabled), and s390x fails because it promotes floats to doubles. I think they're the only architectures which promote floats (although some have not built yet). I'll probably just ignore these tests for now since I'm not sure what the best solution is. James _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
