Re: should we add SH4 code to mmeye locore.S?

On Sun, 1 Oct 2023 at 08:09, Andrius V <[email protected]> wrote:
>
> On Sun, Oct 1, 2023 at 5:32=E2=80=AFAM Izumi Tsutsui <[email protected]=
e.jp> wrote:
> >
> > I guess SH4 MMEYE kernels were tested only without cache enabled.
> > (7750R has 8KB VIPT cache and we need to handle annoying alias issue
> >  especially in port specific device drivers and bus_dma(9) functions)
> >
> > So I'd say "leave it as is unless someone actually have test it."
>
> Thank you for clarification.
>
> >
> > ---
> > so it's not quite SH4 specific, but mostly used only for SH4 cache
> > in primitive asm code.
> >
> > ---
>
> I guess my assumption was based that those two lines are missing in
> mmeye version.
>
> Based on your response, I will leave the code untouched. Thank you.

Would it make sense to add a comment to that effect to the code -
including a specific note asking for anyone with appropriate hardware
to test it?

David