Re: [Qemu-devel] [PATCH v4 2/2] Add support for Zipit Z2 machine

[Vasily Khoruzhick]

On Friday 24 June 2011 17:45:06 Peter Maydell wrote:
> On 17 June 2011 11:04, Vasily Khoruzhick <address@hidden> wrote:
> > Zipit Z2 is small PXA270 based handheld.
> > 
> > Signed-off-by: Vasily Khoruzhick <address@hidden>
> 
> These patches are affected by the bug in current qemu master
> which breaks cpu_physical_memory_map() so I haven't tested them yet.
> 
> Some minor nitpicks (nearly there now):
> > +            if (z->cur_reg == 0x22 && val == 0x0000) {
> > +                z->enabled = 1;
> > +                printf("%s: LCD enabled\n", __func__);
> > +            } else if (z->cur_reg == 0x10 && val == 0x0000) {
> > +                z->enabled = 0;
> > +                printf("%s: LCD disabled\n", __func__);
> > +            }
> 
> Drop or use DPRINTF for these printfs, please.

Hm, I'd like to keep them to see then software enables or disables LCD.

> > +            break;
> > +        default:
> > +            fprintf(stderr, "%s: unknown command!\n", __func__);
> 
> Ditto.

Ok

> > +static VMStateDescription vmstate_zipit_lcd_state = {
> > +    .name = "zipit-lcd",
> > +    .version_id = 1,
> > +    .minimum_version_id = 1,
> > +    .minimum_version_id_old = 1,
> > +    .fields = (VMStateField[]) {
> > +        VMSTATE_INT32(enabled, ZipitLCD),
> > +        VMSTATE_END_OF_LIST(),
> > +    }
> > +};
> 
> This is missing fields for selected, buf[] and cur_reg.
> 
> > +    if (s->len++ > 2) {
> > +        fprintf(stderr, "%s: message too long (%i bytes)\n",
> > +            __func__, s->len);
> > +        return 1;
> > +    }
> 
> DPRINTF.

Ok

> > +    case I2C_START_RECV:
> > +        if (s->len != 1) {
> > +            fprintf(stderr, "%s: short message!?\n", __func__);
> > +        }
> > +        break;
> 
> Ditto.

Ok

> > +static VMStateDescription vmstate_aer915_state = {
> > +    .name = "aer915",
> > +    .version_id = 1,
> > +    .minimum_version_id = 1,
> > +    .minimum_version_id_old = 1,
> > +    .fields = (VMStateField[]) {
> > +        VMSTATE_END_OF_LIST(),
> > +    }
> > +};
> 
> Missing fields for len and buf[].
> 
> Looks ok otherwise. (Patch 1 looks ok too.)
> 
> Have you tried vmload/vmsave, by the way? (I don't know if all the
> devices the pxa2xx uses have save/load support implemented, it
> would be interesting to check if you haven't already.)

Nope, how to try vmload/vmsave?

> -- PMM

Thanks for review.

Regards
Vasily