Subject: Re: [ecasound] floating point exception
From: Kai Vehmanen (k_AT_eca.cx)
Date: Thu Mar 20 2003 - 23:29:08 EET
On Thu, 20 Mar 2003, raoul.megelas wrote:
> to say that mlockall and munlockall are not yet implemented
> this is true. but what i forgot, is this question was answerd in the last year.
> So i apologize. In fact i use to build ecasound by issuing a good
> configure, gmake and install, and i never look at the config.h to undefine
> anything...
Hmm, does the configure process fail...? :o It should go through fine even
though mlockall is not found (although you will get the "can't lock
memory" warnings on runtime).
> now about the -y:100 without defining a chainsetup, i have the same bihaviourr
> but i cann't say anythin about this because i have a big problem as follows:
> ecasound -c -i:somefile.wav -o:/dev/dsp
> floating point exception handling.
> here's the gdb loading and backtrace
Hmm, what architecture (x86, i64, etc?) and what compiler version? I'd
suggest tryign compiling without -ffast-math (see configure.in:70).
> Reading symbols from /usr/libexec/ld-elf.so.1...done.
> Loaded symbols for /usr/libexec/ld-elf.so.1
> #0 0x2834f197 in _thread_stack_alloc () from /usr/lib/libc_r.so.5
> (gdb) backtrace
>
> #0 0x2834f197 in _thread_stack_alloc () from /usr/lib/libc_r.so.5
> #1 0x2834dff5 in pthread_create () from /usr/lib/libc_r.so.5
> #2 0x080ccbed in ecasound_setup_signals (state=0x820a020) at ecasound.cpp:400
> #3 0x080cb93a in main (argc=4, argv=0xbfbff7ac) at ecasound.cpp:93
> #4 0x080cb865 in _start ()
By the way, this just provides info for one thread. You need to switch
to the offending thread with "thread x", where x is the thread id. You can
get a list of different threads with "info threads". Once in the right
thread, then "bt" as usual. If you don't know which is the offending one,
then it's best to back-trace all of them.
-- http://www.eca.cx Audio software for Linux!
This archive was generated by hypermail 2b28 : Thu Mar 20 2003 - 23:21:01 EET