gcc keeps receiving signal 6 (SIGABRT)

[James Ponder <james_(_at_)_oaktree_(_dot_)_co_(_dot_)_uk>]

Hi,

I've just got a new machine with an AMD processor (K6-2) 333Mhz, it works
fine, except that gcc dies with segvs (sometimes) and SIGABRTs (all the time)
for no apparent reason.

I installed OpenBSD 2.3, no problems, I grabbed 'src' from cvs no problems,
and then I tried to compile a kernel, which I managed, but only after
restarting the make about 10 times, as gcc kept dying with an error telling
me it had received signal 6.  Very annoying.  I assumed it might be a kernel
problem so I just continued until I had a new kernel... however, building
all the userland utils resulting in about a million of these errors too.

I've checked the motherboard settings, they SEEM to be fine (according to
the manual anyway), but since this is the first time I've used one of these
K6-2 processors I thought I'd check to see if anyone recognises the above
as something known?  I don't even know whether to blame the motherboard,
the processor or anything.  I would of hoped for more crashing and burning
if there was a hardware fault, rather than an abort() signal ;)

I now have OpenBSD 2.4-current installed on it, but rather than pretend
I never had the problems above, I'd like to know *why* ;)  Day to day
usage of the machine seem unaffected.

The computer is an AMD K6-2 333Mhz, Triton TXPro MB, 32MB SDRAM (10ns),
NE2000 ISA card, IDE HD, AWE64 SB, floppy.  The MB is configured for
95Mhz bus operation with a 3.5 multipler, 2.2V CPU, 3.3V SDRAM.

A bit of praise - I installed mpg123 and it worked first time with the
sound card, I didn't even have to configure it :-)


Best wishes, James

-=- James Ponder -=- james_(_at_)_oaktree_(_dot_)_co_(_dot_)_uk -=- http://www.oaktree.co.uk/ -=-