Hi, with current lame version transcode segfaults. a backtrace told me the SIGSEGV was caused in libmp3lame.so. current lame cvs version works with no problems. the version reads: LAME version 3.96 (alpha 2, Jan 31 2004 19:30:10) (http://www.mp3dev.org/) pls consider upgrading in portage. thanks
do you know what the bug was? We'd prefer not to release a cvs snapshot version, but if you could provide us with more info (like a link to an article describing the fix on the lame dev list) we could include the patch.
there is no entry on the mailing list (not-so-high-traffic) about it,at least i didn't found one. I looked a bit at the comments of new CVS commits,but i had also no success in finding the right fix (bigger changes fixing this?). sorry. perhaps it's better to wait for the next release and have this as `known bug' when CVS is not prefered.
You might want to compile lame with debugging symbols enabled (CFLAGS has -g and FEATURES has nostrip). You can then run it in gdb and provide a backtrace to help focus the search on the cvs ChangeLog: $ gdb transcode (gdb) run blah SEGFAULT at someplace.c:1234 (gdb) bt #2 ... #1 ... #0 ... In addition to the bt, can you please give the output of 'emerge info'?
also, please run 'emerge info' and provide that information here...
i had to recompile transcode and lame (versions in portage). now it doesn't segfault any more..(but it only produces crap,but thats known say ppl in transcode irc channel)...perhaps some gcc error that wasn't triggered in the lame cvs version any more..or i'm too stupid to reproduce the bug. sorry for wasting your time.
no worries... if you end up figuring out how to reproduce it, just reopen... Thanks.
