unable to run current CVS [_IO_*/__flockfile, __pthread_waitfor_restart_signal]

[Rob Browning]

Josh Sled <jsled@asynchronous.org> writes:

> #0  0x400f2966 in __sigsuspend (set=0xbfffe380)
>     at ../sysdeps/unix/sysv/linux/sigsuspend.c:45
> #1  0x40d84d61 in __pthread_wait_for_restart_signal (self=0x40d8e460)
>     at pthread.c:969
> #2  0x40d86aa3 in __pthread_alt_lock (lock=0x81d8048, self=0x0) at restart.h:34
> #3  0x40d82fb6 in __pthread_mutex_lock (mutex=0x81d8038) at mutex.c:120
> #4  0x40d85d8b in __flockfile (stream=0x81d7fa0) at lockfile.c:39
> #5  0x4013c9c5 in _IO_link_in (fp=0x81d7fa0) at genops.c:108
> #6  0x4013c5f9 in _IO_new_file_init (fp=0x81d7fa0) at fileops.c:150
> #7  0x401322af in _IO_new_fopen (
>     filename=0xbfffe590 "/usr/lib/python1.5/site.pyc", mode=0x409780d2 "rb")
>     at iofopen.c:63
>
> ... though frames above #4 change in ea. occurance [frames 0..4 are
> consistent].
>
> I noticed it first with attempting to load the modular libraries; going
> from libtool1.4b to libtool-1.4.2 seems to have fixed this.

So libtool 1.4.2 fixed the problem?

> Then, I was getting it on attempting to load
> /usr/share/pixmaps/gnome-term-tiger.png [gnome_window_icon_init -> ... ->
> gdk_imlib_load_image] ... removing this file sidesteps that error,
> which then lets me see the above backtrace, occuring on attempting to
> load site.pyc [for Guppi init].
>
> Any clues/help?

Interesting.  I guess you never really tracked this down, but I'm
wondering if it might be some library version mismatch, in particular
if say some of the various libraries involved might be linked against
different copies of libltdl (for example) -- since that can happen
with the current (broken) libtool libltdl convenience arrangements,
and somehow that's causing the trouble...

dunno.

-- 
Rob Browning
rlb @defaultvalue.org, @linuxdevel.com, and @debian.org
Previously @cs.utexas.edu
GPG=1C58 8B2C FB5E 3F64 EA5C  64AE 78FE E5FE F0CB A0AD