Re: [Qemu-devel] [PATCH 2/6] hw/sd/milkymist-memcard: split realize() out of SysBusDevice init()

[Michael Walle <michael@walle.cc>]

Hi Philippe,

Am 2018-01-09 19:16, schrieb Philippe Mathieu-Daudé:
> Hi Michael,
> 
>> Am 2018-01-03 17:23, schrieb Philippe Mathieu-Daudé:
>>> 
>>> Create the SDCard in the realize() function.
>>> 
>>> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
>>> ---
>>>  hw/sd/milkymist-memcard.c | 24 +++++++++++++++---------
>>>  1 file changed, 15 insertions(+), 9 deletions(-)
>>> 
>>> diff --git a/hw/sd/milkymist-memcard.c b/hw/sd/milkymist-memcard.c
>>> index 5de3e00e2f..5df3a0f815 100644
>>> --- a/hw/sd/milkymist-memcard.c
>>> +++ b/hw/sd/milkymist-memcard.c
>>> @@ -255,24 +255,29 @@ static void milkymist_memcard_reset(DeviceState 
>>> *d)
>>>  static int milkymist_memcard_init(SysBusDevice *dev)
>>>  {
>>>      MilkymistMemcardState *s = MILKYMIST_MEMCARD(dev);
>>> -    DriveInfo *dinfo;
>>> +
>>> +    memory_region_init_io(&s->regs_region, OBJECT(s), 
>>> &memcard_mmio_ops,
>>> s,
>>> +            "milkymist-memcard", R_MAX * 4);
>>> +    sysbus_init_mmio(dev, &s->regs_region);
>>> +
>>> +    return 0;
>>> +}
>> 
>> 
>> Creating the device (milkymist_memcard_create()) fails with an 
>> assertion:
>>   sysbus_mmio_map_common: Assertion `n >= 0 && n < dev->num_mmio' 
>> failed.
> 
> Thanks for trying this.
> 
> Odd 'make check-qtest-lm32' didn't catch this...
> 
>> because dev->num_mmio is still 0. Seems that sysbus_init_mmio() hasn't 
>> been
>> called yet. Do we have to move the milkymist_memcard_init() to
>> .instance_init?
> 
> Your patch seems correct :)
> 
> Do you have some milky one prebuilt image that I can use for testing?

You can use this one:
   http://milkymist.walle.cc/updates/2012-03-01/flickernoise

http://milkymist.walle.cc/README.qemu

-michael