Re: [Qemu-devel] [PATCH] arm: Stub out NRF51 TWI magnetometer/accelerometer detection

[Stefan Hajnoczi <stefanha@gmail.com>]

On Tue, Jan 8, 2019 at 11:57 AM Peter Maydell <peter.maydell@linaro.org> wrote:
> On Sat, 5 Jan 2019 at 15:02, Stefan Hajnoczi <stefanha@redhat.com> wrote:
> >
> > From: Steffen Görtz <contrib@steffen-goertz.de>
> >
> > Recent microbit firmwares panic if the TWI magnetometer/accelerometer
> > devices are not detected during startup.  We don't implement TWI (I2C)
> > so let's stub out these devices just to let the firmware boot.
> >
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> > Based-on: <20190103091119.9367-1-stefanha@redhat.com>
> > ---
> > Steffen: Please post your Signed-off-by.  I did some minor cleanups so
> > your patch can be merged.  Thanks!
> >
> >  include/hw/arm/nrf51.h     |  1 +
> >  include/hw/arm/nrf51_soc.h |  1 +
> >  hw/arm/nrf51_soc.c         | 62 ++++++++++++++++++++++++++++++++++++++
> >  3 files changed, 64 insertions(+)
>
> >
> > +/* Two-Wire Interface (TWI) is not implemented but the microbit-dal firmware
> > + * panics if the TWI accelerometer and magnetometer WHO_AM_I registers cannot
> > + * be read.  Stub out this read sequence so microbit-dal starts up.
> > + */
> > +static uint32_t twi_read_sequence[] = {0x5A, 0x5A, 0x40};
> > +static uint32_t twi_regs[0x1000 / 4];
> > +
> > +static uint64_t twi_read(void *opaque, hwaddr addr, unsigned int size)
> > +{
> > +    static int i;
> > +    uint64_t data = 0x00;
> > +
> > +    switch (addr) {
> > +    case NRF51_TWI_EVENT_STOPPED:
> > +        data = 0x01;
> > +        break;
> > +    case NRF51_TWI_EVENT_RXDREADY:
> > +        data = 0x01;
> > +        break;
> > +    case NRF51_TWI_EVENT_TXDSENT:
> > +        data = 0x01;
> > +        break;
> > +    case NRF51_TWI_REG_RXD:
> > +        data = twi_read_sequence[i];
> > +        if (i < ARRAY_SIZE(twi_read_sequence)) {
> > +            i++;
> > +        }
> > +        break;
> > +    default:
> > +        data = twi_regs[addr / 4];
> > +        break;
> > +    }
> > +
> > +    qemu_log_mask(LOG_UNIMP, "%s: 0x%" HWADDR_PRIx " [%u] = %" PRIx32 "\n",
> > +                  __func__, addr, size, (uint32_t)data);
> > +
> > +
> > +    return data;
> > +}
> > +
> > +static void twi_write(void *opaque, hwaddr addr, uint64_t data,
> > +                      unsigned int size)
> > +{
> > +    qemu_log_mask(LOG_UNIMP, "%s: 0x%" HWADDR_PRIx " <- 0x%" PRIx64 " [%u]\n",
> > +                  __func__, addr, data, size);
> > +    twi_regs[addr / 4] = data;
> > +}
> > +
> > +static const MemoryRegionOps twi_ops = {
> > +    .read = twi_read,
> > +    .write = twi_write
> > +};
>
> Can you put this in its own device in its own source file,
> please? This is too much device-specific code to have
> in the SoC's source file.

Sure, will send v2.  It should also be hooked up by the microbit
machine type instead of being hardcoded into the nRF51 SoC.

Stefan