Re: [PATCH v2 7/8] mips: allow compiling out CONFIG_MIPS_ITU

[Paolo Bonzini]

On Wed, Feb 7, 2024 at 8:12 PM Philippe Mathieu-Daudé <philmd@linaro.org> wrote:
>
> Hi Paolo,
>
> On 7/2/24 12:14, Paolo Bonzini wrote:
> > itc_reconfigure() is referenced from TCG, provide a stub if needed.
> > This makes it possible to build a QEMU binary that only includes
> > boards without a CPS device (only Malta and Boston create one).
> >
> > Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> > ---
> >   hw/mips/mips_itu-stub.c | 26 ++++++++++++++++++++++++++
> >   hw/mips/meson.build     |  1 +
> >   2 files changed, 27 insertions(+)
> >   create mode 100644 hw/mips/mips_itu-stub.c
> >
> > diff --git a/hw/mips/mips_itu-stub.c b/hw/mips/mips_itu-stub.c
> > new file mode 100644
> > index 00000000000..4cc82b8461f
> > --- /dev/null
> > +++ b/hw/mips/mips_itu-stub.c
> > @@ -0,0 +1,26 @@
> > +/*
> > + * Inter-Thread Communication Unit emulation.
> > + *
> > + * Copyright (c) 2016 Imagination Technologies
> > + *
> > + * This library is free software; you can redistribute it and/or
> > + * modify it under the terms of the GNU Lesser General Public
> > + * License as published by the Free Software Foundation; either
> > + * version 2.1 of the License, or (at your option) any later version.
> > + *
> > + * This library is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> > + * Lesser General Public License for more details.
> > + *
> > + * You should have received a copy of the GNU Lesser General Public
> > + * License along with this library; if not, see 
> > <http://www.gnu.org/licenses/>.
>
> This is your code addition, so "Inter-Thread Communication Unit stubs" /
> RH / SPDX GPLv2-or-later.
>
> > + */
> > +
> > +#include "qemu/osdep.h"
> > +#include "hw/misc/mips_itu.h"
> > +
> > +void itc_reconfigure(MIPSITUState *tag)
> > +{
> > +    abort();
>
> As Zoltan suggested, g_assert_not_reached(). Indeed this path
> can't be reached without ITU, TCG won't emit anything and will
> call instead:
>
>      qemu_log_mask(LOG_UNIMP, "mthc0 %s (reg %d sel %d)\n",
>                    register_name, reg, sel);
>
> I'm reluctant to add stubs, but since it helps you (hoping we
> can figure a clean way to split architectural access to hw/ from
> tcg/ one day):

Not sure how it's reached anyway, because the saar field of
DisasContext is never written...

Paolo