Re: Crtl-C -> panic

[David Bateman]

Ok, I often have slight differences between my own code and the CVS, due
to the fact of developing on two machines and corporate firewalls preventing
me from easily keeping in sync. So to test the idea that the problem is
mine, I 

1) got a clean version of the octave CVS (updated a couple of hours ago)
2) ran autogen.sh
3) ran ./configure --enable-shared --disable-static --prefix=/opt/octave-2.1.57 
-enable-rpath=-Wl,-rpath -Wl,/opt/octave-2.1.57/lib/octave-2.1.57 
--with-blas=/usr/lib/libblas.so --with-lapack=/usr/lib/liblapack.so

   the blas and lapack are generic version that are later replaced at 
   runtime by the appropriate ATLAS version.

4) make; rm -fr /opt/octave-2.1.57; make install
5) export LD_LIBRARY_PATH=/usr/lib/atlas/PIIISSE1
5) ldd /opt/octave-2.1.57/bin/octave
        liboctinterp.so => /opt/octave-2.1.57/lib/octave-2.1.57/liboctinterp.so 
(0x40015000)
        liboctave.so => /opt/octave-2.1.57/lib/octave-2.1.57/liboctave.so 
(0x40542000)
        libcruft.so => /opt/octave-2.1.57/lib/octave-2.1.57/libcruft.so 
(0x409a7000)
        liblapack.so.3 => /usr/lib/atlas/PIIISSE1/liblapack.so.3 (0x409f6000)
        libblas.so.3 => /usr/lib/atlas/PIIISSE1/libblas.so.3 (0x40f5b000)
        libfftw3.so.3 => /usr/lib/libfftw3.so.3 (0x4126b000)
        libreadline.so.4.3 => /lib/libreadline.so.4.3 (0x41303000)
        libncurses.so.5 => /lib/libncurses.so.5 (0x41330000)
        libdl.so.2 => /lib/libdl.so.2 (0x41370000)
        libhdf5.so.0 => /usr/lib/libhdf5.so.0 (0x41374000)
        libz.so.1 => /lib/libz.so.1 (0x41450000)
        libg2c.so.0 => /usr/lib/libg2c.so.0 (0x4145e000)
        libm.so.6 => /lib/libm.so.6 (0x4147c000)
        libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x4149f000)
        libstdc++.so.5 => /usr/lib/libstdc++.so.5 (0x414a8000)
        libc.so.6 => /lib/libc.so.6 (0x41562000)
        libpthread.so.0 => /lib/libpthread.so.0 (0x4168d000)
        libgpm.so.1 => /usr/lib/libgpm.so.1 (0x416dd000)
        /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
6) /opt/octave-2.1.57/bin/octave
7) Ctrl-C
octave:1> panic: Aborted -- stopping myself...
attempting to save variables to `octave-core'...
save to `octave-core' complete
Abort

So the problem seems to be there in the current CVS, and is not due to
older versions of the dynmaic libraries hanging around.

8) gdb /opt/octave-2.1.57/bin/octave
   (gdb) handle SIGINT stop print pass
   (gdb) run
   octave:1> Ctrl-C
   (gdb) cont
Continuing.

Program received signal SIGINT, Interrupt.
0x4158a53c in sigprocmask () from /lib/libc.so.6
(gdb) cont
Continuing.

Program received signal SIGABRT, Aborted.
0x4158a571 in kill () from /lib/libc.so.6
(gdb) cont
Continuing.
panic: Aborted -- stopping myself...
attempting to save variables to `octave-core'...
save to `octave-core' complete

Program received signal SIGABRT, Aborted.
0x4158a571 in kill () from /lib/libc.so.6

Not sure what is happening here and I can't single step through the
code in gdb to get more info....

I also then tried to build octave-forge with this CVS and got no where,
firstly the configure script failed to correctly determine some option
(due to mkoctfile failing) and the build failed on the compile of the
first oct-file. This seems to have something to do with the introduction
of the int/uint types....




According to John W. Eaton <address@hidden> (on 06/17/04):
> On 16-Jun-2004, Teemu Ikonen <address@hidden> wrote:
> 
> | On 16/06/04 17:44, David Bateman wrote:
> | > I had the habit of hitting Ctrl-C to remove the current text from the
> | > commandline of octave. However, in the CVS version I currently have
> | > this causes the following
> | > 
> | > octave:1> panic: Aborted -- stopping myself...
> | > attempting to save variables to `octave-core'...
> | > save to `octave-core' complete
> | > Abort
> | 
> | I can repeat this for a few days old CVS checkout.
> 
> I'm unable to reproduce this using a current build.
> 
> Did you build your CVS version with --enable-shared?  Is it possible
> that it is linking to a set of shared libraries from a previously
> installed version?
> 
> jwe

-- 
David Bateman                                address@hidden
Motorola CRM                                 +33 1 69 35 48 04 (Ph) 
Parc Les Algorithmes, Commune de St Aubin    +33 1 69 35 77 01 (Fax) 
91193 Gif-Sur-Yvette FRANCE

The information contained in this communication has been classified as: 

[x] General Business Information 
[ ] Motorola Internal Use Only 
[ ] Motorola Confidential Proprietary