Re: [PATCH 11/15] rx: move BIOS load from MCU to board

[Philippe Mathieu-Daudé]

On 10/26/20 6:34 PM, Paolo Bonzini wrote:
> On 26/10/20 18:24, Alex Bennée wrote:
> > Paolo Bonzini <pbonzini@redhat.com> writes:
> >
> > > The ROM loader state is global and not part of the MCU, and the
> > > BIOS is in machine->firmware.  So just like the kernel case,
> > > load it in the board.
> > >
> > > Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
> > > Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> > > ---
> > >   hw/rx/rx-gdbsim.c | 7 +++++++
> > >   hw/rx/rx62n.c     | 9 ---------
> > >   2 files changed, 7 insertions(+), 9 deletions(-)
> > >
> > > diff --git a/hw/rx/rx-gdbsim.c b/hw/rx/rx-gdbsim.c
> > > index 417ec0564b..040006c1c5 100644
> > > --- a/hw/rx/rx-gdbsim.c
> > > +++ b/hw/rx/rx-gdbsim.c
> > > @@ -142,6 +142,13 @@ static void rx_gdbsim_init(MachineState *machine)
> > >               /* Set dtb address to R1 */
> > >               RX_CPU(first_cpu)->env.regs[1] = SDRAM_BASE + dtb_offset;
> > >           }
> > > +    } else {
> > > +        if (machine->firmware) {
> > > +            rom_add_file_fixed(machine->firmware, RX62N_CFLASH_BASE, 0);
> > > +        } else if (!qtest_enabled()) {
> > > +            error_report("No bios or kernel specified");
> > > +            exit(1);
> > > +        }
> > >       }
> > >   }
> > >   
> > > diff --git a/hw/rx/rx62n.c b/hw/rx/rx62n.c
> > > index 6eb4eea700..17ec73fc7b 100644
> > > --- a/hw/rx/rx62n.c
> > > +++ b/hw/rx/rx62n.c
> > > @@ -245,15 +245,6 @@ static void rx62n_realize(DeviceState *dev, Error **errp)
> > >                              rxc->rom_flash_size, &error_abort);
> > >       memory_region_add_subregion(s->sysmem, RX62N_CFLASH_BASE, &s->c_flash);
> > >   
> > > -    if (!s->kernel) {
> > > -        if (bios_name) {
> > > -            rom_add_file_fixed(bios_name, RX62N_CFLASH_BASE, 0);

The acceptance test fails. There is some odd reset order problem,
eventually related to this discussion "CPU reset handler priority":
https://www.mail-archive.com/qemu-devel@nongnu.org/msg686362.html

> > > -        }  else if (!qtest_enabled()) {
> > > -            error_report("No bios or kernel specified");
> > > -            exit(1);
> > > -        }
> > > -    }
> > > -
> >
> > I'm confused because on the face of it these are two different models.
> > I'll defer to the domain expert on this one.

Yoshinori started to clean that here:
https://www.mail-archive.com/qemu-devel@nongnu.org/msg734135.html

> rx62n is the SoC, rx-gdbsim.c instead includes the rx62n7 and rx62n8
> machines.
>
> Paolo