On Tue, 14 Feb 1995 10:57:48 +0000 (GMT), Nigel J Kettlewell said: > I seem to remember that RETI re-enables the interrupts - when an interrupt > occurs, they are disabled (I think) which is done by setting (resetting?) > the interrupt flip-flops, and RETI restores them to whatever they were > before.
RETN does this. When an NMI occurs, iff1 is cleared (disbling interrupts) and when RETN is executed, iff2 is copied into iff1 (restoring them to the previous status). When a maskable interrupt occurs, both iff1 and iff2 are cleared. I seem to remember doing an experiment which showed that RETI did not re-enable the interrupts. imc
