On Sun, 24 May 1998, Klaus Weber wrote:

: setenv CFLAGS "-O6 -fomit-frame-pointer -finline-functions -ffast-math
: -funroll-loops -fforce-mem -fforce-addr" 

That's quite a bit of optimization choices, as well as -O6 (which is
meaningless, as the highest level that does anything is -O3 :).  It's a
known fact that many optimization "tuning" options do break occasionally on
gcc 2.7.2.x.

I'll move this PR to "suspended", which means "get back to us if you use
egcs and it still causes the same problem".  It's highly unlikely that this
optimization breakage will be fixed in the 1.3-series gcc (based on gcc
2.7.2.2).

-- 
-- Todd Vierling (Personal tv@pobox.com; Bus. todd_vierling@xn.xerox.com)