Simon Burge <simonb@wasabisystems.com> writes:
> A disassembly of Ultrix's sbrk.o show exactly the same code (well, ok,
> the addu and li are switched) as our sbrk.S...

heh.

OK, so, then, if we fix it (and maybe we should) the compat code would
have to be conditionalized several ways.


> .../libc/sys/arch/... ? :-)

yeah, yeah, i'm context switching really hard today.


cgd