Re: R5000SC issues and fix

On Fri, 5 Dec 2025 at 17:49, Jason Thorpe <[email protected]> wrote:
>
>
>
> > On Dec 5, 2025, at 4:43=E2=80=AFPM, Adrian Chadd <[email protected]> w=
rote:
> > Izumi, do you still have an R5000 Indy and/or other devices? I'm not fa=
miliar
> > (yet?) with the R5000 cache stuff, and I can't imagine it never worked!
>
> The R5000 cache stuff worked at one point!  I wrote all that code origina=
lly on an Algorithmics P-5064 with an R5K module and it worked great.

It's possible that it's something shady with this machine. I put in a
different set of RAM - and only 32MB instead of 256MB - and it
ran for a while longer but it panicked whilst I wasn't looking.

>
> I also have a Cobalt Qube 2 and while I haven=E2=80=99t fired it up terri=
bly recently, it does have an RM5200 and worked with pre-netbsd-10-current:
>
> https://dmesgd.nycbug.org/dmesgd?do=3Dview&id=3D5404
>
> I guess I ought to boot it up again.  No secondary cache on that machine,=
 tho.

I can try this with the secondary cache disabled entirely and see how
things behave.
I do encourage more MIPS stuff though, I want to work my way forward
through to mips24k/mips34k/mips74k and then the SMP MIPS chips. :-)



-adrian

>
> -- thorpej
>