Re: [hatari-devel] FPU detection, who is wrong?

[ Thread Index | Date Index | More Archives ]

 > This null state frame is generated until the first nonconditional
 > floatingpoint instruction is executed (conditionals include FNOP, FBcc,
 > FDBcc, FScc, and FTRAPcc).

Uh, oh. I wasn't aware that FNOP is treated as "conditonal". So Hatari
is correct, and this is another bug in the detection code.

FNOP instruction is not "true" NOP instruction, it uses same opcode bit pattern as (useless) conditional branch variant: FBF ("branch if false is true").

Mail converted by MHonArc 2.6.19+