bootstrap failure on mainline for mmix

[Hans-Peter Nilsson]

On Thu, 25 Apr 2002, Neil Booth wrote:
> Hans-Peter Nilsson wrote:-
>
> > The SEGV is function_attribute_inlinable_p trying to dereference
> > a NULL pointer:
> > =>    for (i = 0; targetm.attribute_table[i].name != NULL; i++)
>
> Oops, I thought attribs.c was the only user.  That function is
> very verbose for what it does, so I took the opportunity to
> simplify it whilst I was there.  Does this patch work for you?

FWIW, it looks like that should do it.  Sorry, no machine cycles
to spare right now, will try later.

You might want to try mmix-knuth-mmixware or mips-elf as per
<URL:http://gcc.gnu.org/simtest-howto.html>.  The regression
tester now tests (but doesn't report to suspects) mips-elf
failures (see
<URL:http://gcc.gnu.org/ml/gcc-regression/2002-04/>), and mips
too doesn't have machine attributes, so it's quite safe to
assume it's the same failure it reported to that list.

brgds, H-P