Re: [PATCH] build: riscv is not a valid -march value

["Stanisław Kardach" <kda@semihalf.com>]

On Wed, Nov 22, 2023 at 5:41 PM David Marchand
<david.marchand@redhat.com> wrote:
>
> On Wed, Nov 22, 2023 at 5:17 PM Bruce Richardson
> <bruce.richardson@intel.com> wrote:
> >
> > On Wed, Nov 22, 2023 at 05:02:56PM +0100, David Marchand wrote:
> > > On Tue, Nov 21, 2023 at 5:49 PM <christian.ehrhardt@canonical.com> wrote:
> > > >
> > > > From: Christian Ehrhardt <christian.ehrhardt@canonical.com>
> > > >
> > > > If building riscv natively with -Dplatform=generic config/meson.build
> > > > will select cpu_instruction_set=riscv.
> > > >
> > > > That was fine because config/riscv/meson.build did override it to valid
> > > > values later, but since b7676fcccab4 ("config: verify machine arch
> > > > flag") it will break the build as it tries to test -march=riscv which
> > > > is not a value value.
> > > >
> > > > The generic setting used in most cases is rv64gc, set this here
> > > > as well.
> > > >
> > > > Fixes: b7676fcccab4 ("config: verify machine arch flag")
> > > > Fixes: f22e705ebf12 ("eal/riscv: support RISC-V architecture")
> > > >
> > > > Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
> > > > ---
> > > >  config/meson.build | 2 +-
> > > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > diff --git a/config/meson.build b/config/meson.build
> > > > index d732154731..a9ccd56deb 100644
> > > > --- a/config/meson.build
> > > > +++ b/config/meson.build
> > > > @@ -152,7 +152,7 @@ if cpu_instruction_set == 'generic'
> > > >      elif host_machine.cpu_family().startswith('ppc')
> > > >          cpu_instruction_set = 'power8'
> > > >      elif host_machine.cpu_family().startswith('riscv')
> > > > -        cpu_instruction_set = 'riscv'
> > > > +        cpu_instruction_set = 'rv64gc'
> > >
> > > Copying more people.
> > >
> > > This fix is probably the best, so close to the release.
> > >
> >
> > Agreed
>
> I took this patch as is, for now.
Sorry for reviving an old thread, I was on a rather long OoO, hence I
did not answer.
Thank you for taking this patch.
>
> >
> > >
> > > However, I think a more complete fix would be to set this here to generic.
> > > And do the march validation in config/riscv/meson.build in a similar
> > > fashion to ARM.
> > >
> > > Or maybe the validation added in b7676fcccab4 ("config: verify machine
> > > arch flag") should be moved after subdir(arch_subdir).
> > > Bruce, opinion?
> > >
> >
> > Probably the first of these two is best, to do the march validation in the
> > riscv-specific file. However, I've no strong opinions either way.
>
> Stanislaw, could you look at doing some enhancement on this topic?
> And, in any case, what we lack is a CI for RISC V.
It seems that there is not much traction for RISC-V DPDK yet. StarFive
seems to be focused more on the platform side of things and therefore
I don't have any server-grade HW to really run CI on.
>
>
> --
> David Marchand
>


--
Best Regards,
Stanisław Kardach