not the memory, but somewhat conceding to a hardware problem

[Robert Schweikert]

In my ongoing quest to get to the bottom of my compile problems with gcc
I have now made the following changes to my system:

- I flashed the BIOS upgrading to AMI BIOS 2.00.02
           was hoping I would be able to turn the secondary cache of
with this but, but I still can't.

- I installed a Tekram DC-390U3W SCSI card and a Quantum Atlas V SCSI
drive
           the rather slow speed of IDE was getting on my nerves.

- Upgraded to kernel 2.4-test5

As a quick repeat a short listing of my system:
- Dual PII 450MHz running on Tyan S1832DL motherboard with Intel BX440
chipset. 128MB PC100 memory, Fujitsu IDE drive (currently not active)

What I have done to get closer to the problem:

I transferred my system (RH6.2 with some upgrades) to the new SCSI
drive. Then I wiped out the / and /usr partitions of the old IDE drive
and installed RH6.1 onto the clean partitions. RH6.1 compiler is
gcc-2.91.66. I changed my bios setting back to boot from the IDE drive
where I now have the "clean install" RH6.1 distro. This boots fine and
as desired I am not running X at this point, to take another variable
out of the equation. Now I copied the tarball of gcc-2.95.2 to the new
/usr/local on the IDE drive unpacked it and started my quest. If all
would have worked it would have been nice, I would have had someone to
yell at right away. But well, life is not that easy.

So anyway, things move along nicely for a while, then I the following
crap "hda lost interrupt" and "stuck on TLB IPI wait (CPU#0)". Ok these
messages obviously come from the hardware and hopefully someone here
knows what this means and whom I need to yell at, Intel or Tyan.
Needless to say I would like to fix this problem and any help is
appreciated.

To continue the story. I know went back to my system I have installed on
the new SCSI drive. This is now RH6.2 with an upgrade to gcc-2.95 and I
figure well I'll give this another try. Now I am also running X (GNOME
with Icewm). So I kick of a build. After a while my CPU load drops to
zero (could this be the "stuck on..." problem?), so I killed the process
as I have seen the system hang up before when I saw this phenomenon. The
I restart and again everything is chucking along for quite a while. Then
in stage2 (never made it this far before) I get the following message:

stage2/xgcc -Bstage2/ -B/usr/local/i686-pc-linux-gnu/bin/ -c
-DIN_GCC     -O2 -g -O2  -DHAVE_CONFIG_H    -I. -I. -I./config
-I./../include final.c
/tmp/ccKOwRgr.s: Assembler messages:
/tmp/ccKOwRgr.s:8496: Internal error!
Assertion failure in cvt_frag_to_fill at write.c line 540.
Please report this bug.
make[2]: *** [final.o] Error 1
make[2]: Leaving directory `/usr/local/gcc-2.95.2/gcc'
make[1]: *** [bootstrap] Error 2
make[1]: Leaving directory `/usr/local/gcc-2.95.2/gcc'
make: *** [bootstrap] Error 2

Any help is appreciated.

Thanks,
Robert

--
Robert Schweikert                      MAY THE SOURCE BE WITH YOU
rjschwei@mindspring.com                         LINUX