Matthieu Regnauld
2018-Nov-05 20:17 UTC
[Vorbis-dev] Decode ogg file while downloading: what are the good practices?
I haven't managed yet to get any stacktrace, but I'll work on it (that's the problem using C/C++ code in an Android app...). Also, it's important to notice that I manage to make ov_read() crash, *even when the audio file is fully downloaded!* What I do to reproduce the bug is to call ov_pcm_seek() (without any reload of the file) many times in a row, very quickly (like 6-7 times per seconds, during a few seconds). So far, any idea how to fix that? Thanks. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/vorbis-dev/attachments/20181105/2f9f474f/attachment.html>
Ralph Giles
2018-Nov-05 22:49 UTC
[Vorbis-dev] Decode ogg file while downloading: what are the good practices?
On 2018-11-05 12:17 PM, Matthieu Regnauld wrote:> What I do to reproduce the bug is to call ov_pcm_seek() (without any > reload of the file) many times in a row, very quickly (like 6-7 times > per seconds, during a few seconds).It certainly shouldn't crash, or return variant data. Hard to tell more without a stack trace. Are you making all your calls from the same thread? The VorbisFile struct isn't re-entrant. -r
Matthieu Regnauld
2018-Nov-06 18:20 UTC
[Vorbis-dev] Decode ogg file while downloading: what are the good practices?
I'm making all my calls from the same thread, yes. Also, I'll try to reproduce the bug in a very basic C/C++ project, without using Android, so I can get a stacktrace (the only way to make vorbisfile work in Android was to compile it as a static library - as .a file - so I think that's why I can't get any stacktrace so far). If there is any working similar project, I'm interested. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/vorbis-dev/attachments/20181106/e1fba118/attachment.html>