Definite Pentium Pro Problems...

[Roger W. Brown]

On 24-Feb-98 Jeffrey A Law wrote:
> It looks like you used a compiler built from the mainline sources to
> bootstrap the release branch; if indeed the mainline sources are
> broken, they could mis-compile the release branch and lead to this
> kind of error.
> 
> I did a "configure; make bootstrap" from the release branch on my
> 686-linux box last night along with "make check".

  I tried again:

  568  export PATH=/usr/local/bin:/usr/X11/bin:/usr/bin:/bin:.
  569  rm -r /usr/local/EGCS/*
  570  make clean
  571  rm config.cache
  572  gcc --verbose
            Reading specs from
             /usr/lib/gcc-lib/i686-Bregor-linux/2.7.2.3.f.1/specs
             gcc version 2.7.2.3.f.1
  573  ./configure --prefix=/usr/local/EGCS --enable-shared
        --enable-threads --disable-haifa --target=i686-Bregor-linux
        --build=i686-Bregor-linux --host=i686-Bregor-linux
  574  make bootstrap
  577  make install
  579  cd ./libraries/libf2c/libF77
  581  /usr/local/EGCS/bin/gcc -c -DSkip_f2c_Undefs -I../ -g -O2
       -march=pentiumpro --verbose r_tanh.c

       gcc: Internal compiler error: program cc1 got fatal signal 6


  Sure, the verbose output includes:

  (i686-Bregor-linux) compiled by GNU C version egcs-2.91.09
  980221 (gcc-2.8.0 release), but does that not come from the
  stage1 and stage2 re-builds ?

  This may all be due to a fault installed on my system, and this is not
  an appropriate forum to discuss system hygiene; yet several others report
  similar findings.  I'll keep looking.


  Regards,

           Roger Brown

  
---------------------------------------------
bregor@anusf.anu.edu.au   +61 2 6249-2046   (02) 6249-2046
Date: 24-Feb-98            Time: 18:15:20

Australian National University Supercomputer Facility
=====================================================