[kvm-unit-tests PATCH 2/2] riscv: Add kvmtool support

[Andrew Jones]

On Mon, Jul 07, 2025 at 08:28:30AM -0700, Jesse Taube wrote:
> On Fri, Jul 4, 2025 at 8:13 AM Andrew Jones <andrew.jones at linux.dev> wrote:
> >
> > arm/arm64 supports running tests with kvmtool as a first class citizen.
> > Most the code to do that is in the common scripts, so just add the riscv
> > specific bits needed to allow riscv to use kvmtool as a first class
> > citizen too.
> >
> > Signed-off-by: Andrew Jones <andrew.jones at linux.dev>
> 
> Reviewed-by: Jesse Taube <jesse at rivosinc.com>

Thanks!

...
> >  else
> > -    if [ "$arch" != "arm64" ] && [ "$arch" != "arm" ]; then
> > +    if [ "$arch" != "arm" ] && [ "$arch" != "arm64" ] &&
> > +       [ "$arch" != "riscv32" ] && [ "$arch" != "riscv64" ]; then
> 
> Are there plans to add i386 and x86_64 too?

I don't have plans myself, at least not right now, but I think it'd be
good to do it, so it'd be great if somebody would write the patch and
do the testing.

...
> > +       command="$(timeout_cmd) $kvmtool run"
> > +       if ( [ "$HOST" = "riscv64" ] && [ "$ARCH" = "riscv32" ] ) ||
> > +          ( [ "$HOST" = "riscv32" ] && [ "$ARCH" = "riscv64" ] ); then
> 
> Shouldn't there be a check like this on arm for when  [ "$HOST" =
> "arm" ] && [ "$ARCH" = "aarch64" ]?

arm kvm has been long deprecated and removed from Linux so we should
actually remove any references to HOST==arm throughout the framework
too, since it can't happen (any more).

Thanks,
drew