On 17/06/2016 11:08, Amit Shah wrote:
> On (Fri) 17 Jun 2016 [10:19:17], Paolo Bonzini wrote:
>> The implementation of SATN/STOP is completely busted.  The idea
>> would be that the next DMA read is for a SCSI message and after
>> that the adapter would transition to the command phase.
>>
>> The recent fix to SATN/STOP broke migration, which is one more
>> reason to drop SATN/STOP handling completely.  It is only used
>> in practice to send 3-byte messages (target number + tag type
>> + tag number) for tagged command queuing on adapters that lack
>> the SATN3 command, and we do not advertise support for TCQ.
>>
>> Signed-off-by: Paolo Bonzini <[email protected]>
> 
> [...]
> 
>> @@ -585,9 +544,12 @@ const VMStateDescription vmstate_esp = {
>>          VMSTATE_BUFFER(ti_buf, ESPState),
>>          VMSTATE_UINT32(status, ESPState),
>>          VMSTATE_UINT32(dma, ESPState),
>> -        VMSTATE_BUFFER(cmdbuf, ESPState),
>> -        VMSTATE_UINT32(cmdlen, ESPState),
>> -        VMSTATE_UINT32(do_cmd, ESPState),
>> +        /* Used to be cmdbuf, cmdlen, do_cmd, but the implementation
>> +         * of "Select with ATN and stop" was totally busted.
>> +         */
>> +        VMSTATE_UNUSED(16),
>> +        VMSTATE_UNUSED(4),
>> +        VMSTATE_UNUSED(1),
> 
> Why 1?

Because I thought it was a bool.  My mistake.

Paolo

Reply via email to