Re: [PATCH 03/24] aspeed/i2c: Fix DMA address mask

[Cédric Le Goater]

On 4/7/21 11:22 PM, Philippe Mathieu-Daudé wrote:
> Hi Cédric,
> 
> On 4/7/21 7:16 PM, Cédric Le Goater wrote:
>> The RAM memory region is now used for DMAs accesses instead of the
>> memory address space region. Mask off the top bits of the DMA address
>> to reflect this change.
>>
>> Cc: Philippe Mathieu-Daudé <f4bug@amsat.org>
>> Signed-off-by: Cédric Le Goater <clg@kaod.org>
>> ---
>>  hw/i2c/aspeed_i2c.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/hw/i2c/aspeed_i2c.c b/hw/i2c/aspeed_i2c.c
>> index 518a3f5c6f9d..e7133528899f 100644
>> --- a/hw/i2c/aspeed_i2c.c
>> +++ b/hw/i2c/aspeed_i2c.c
>> @@ -601,7 +601,7 @@ static void aspeed_i2c_bus_write(void *opaque, hwaddr 
>> offset,
>>              break;
>>          }
>>  
>> -        bus->dma_addr = value & 0xfffffffc;
>> +        bus->dma_addr = value & 0x3ffffffc;
> 
> This field is migrated (aspeed_i2c_bus_vmstate).

yes.
> Does the first patch "aspeed/smc: Use the RAM memory region for DMAs"
> break the migration?

You are right it does. Maintaining migration compatibility is overkill 
for this machine, but I should mention the first patch is breaking it.

Thanks,

C.