Re: CVS commit: src/sys/arch/m68k/m68k

[Nat Sloss <nathanialsloss%yahoo.com.au@localhost>]

On Mon, 7 Apr 2025 17:44:09 Tetsuya Isaki wrote:
> At Sat, 5 Apr 2025 22:01:37 +1100,
> 
> Nat Sloss wrote:
> > Following advice and following the spirit of my intentions adding nop in
> > the switch i added nops before "rte" in locore.s in sys/mac68k as
> > opposed to m68k/swicher.
> 
> Are you sure?
> 
> First, errata E4 said the workaround is "nop-before-f-line",
> Somehow, what you wrote was "nop-before-rts" (in the kernel).
> I've also pointed it out.
> 
> Next you claimed that as if these rts instructions returned the
> flow to the userland directly...  (Of course not)
> So I've just pointed out that the rts is not what you imagined.
> In the first place, "nop-before-rts" is already your fiction.
> 
> 
> If I believe this E4 text, no kernel patch should be necessary
> for this issue, as Martin alreadly said.
> ---
> Tetsuya Isaki <isaki%pastel-flower.jp@localhost / isaki%NetBSD.org@localhost>


nop - f* instructions are not atomic....what will happen if nop is executed 
and then the processor is interrupted (by say pressing a keystroke on the 
keyboard)  then after that interrupt has run the f* instruction is executed.

How should this be handled?

Nat