> Sorry for the previous post, but tot_ener has to be initialized each time > L3psycho_anal since it is not declared static. > So forget the memset, and declare tot_ener as > > FLOAT8 tot_ener[4]={0,0,0,0}; > > Under MSVC this will give not FPU problems anymore (at least when the > EM_OVERFLOW,_EM_UNDERFLOW and the _EM_ZERODIVIDE bits are cleared. There are > still some problems in the lame_encode function when the _EM_INVALID is > cleared as well. > > > Albert > > http://www.cdex.n3.net/ > So we cant blame this on a gcc compiler bug :-( I think _EM_INVALID means operating on unintialized data? There must be some kind of memory allocation/access problem that is trashing variables in an unpreditcable way, introduced between 3.61 and 3.62. I just discovered the -D date option for CVS update, so tonight I will try to find out exactly when the problem appeared. Mark -- MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
- Re: [MP3 ENCODER] using lame363 for shuttle tra... Robert Hegemann
- Re: [MP3 ENCODER] using lame363 for shuttle transmi... Albert Faber
- Re: [MP3 ENCODER] using lame363 for shuttle tra... Robert Hegemann
- [MP3 ENCODER] xing vs. fhg (vbr) Ampex
- RE: [MP3 ENCODER] using lame363 for shuttle transmission Mathew Hendry
- RE: [MP3 ENCODER] using lame363 for shuttle transmi... Takehiro Tominaga
- Re: [MP3 ENCODER] using lame363 for shuttle transmi... Mark Taylor
- Re: [MP3 ENCODER] using lame363 for shuttle transmi... Albert Faber
- Re: [MP3 ENCODER] using lame363 for shuttle tra... Albert Faber
- [MP3 ENCODER] mp3 vs. wma Mark Taylor
- [MP3 ENCODER] mp3 vs. wma Ampex
- Re: [MP3 ENCODER] mp3 vs. wma Mark Taylor
- Re: [MP3 ENCODER] mp3 vs. wma Scott Manley
- Re: [MP3 ENCODER] mp3 vs. wma Mark Stephens
- RE: [MP3 ENCODER] using lame363 for shuttle transmission Mathew Hendry
- [MP3 ENCODER] crc check Robin
- Re: [MP3 ENCODER] crc check Mark Taylor
- Re: [MP3 ENCODER] using lame363 for shuttle transmission Mark Taylor
- Re: [MP3 ENCODER] using lame363 for shuttle transmi... Stefan Bellon