Re: [PATCH 2/3] x86/HVM: re-work hvm_wait_for_io() a little

[Jan Beulich <jbeulich@xxxxxxxx>]

On 15.07.2020 14:47, Paul Durrant wrote:
>> From: Jan Beulich <jbeulich@xxxxxxxx>
>> Sent: 15 July 2020 13:04
>>
>> @@ -139,20 +132,24 @@ static bool hvm_wait_for_io(struct hvm_i
>>              p->state = STATE_IOREQ_NONE;
>>              data = p->data;
>>              break;
>> +
> 
> Possibly mention the style fix-up in the comment comment.

Done.

>>          case STATE_IOREQ_READY:  /* IOREQ_{READY,INPROCESS} -> IORESP_READY 
>> */
>>          case STATE_IOREQ_INPROCESS:
>>              wait_on_xen_event_channel(sv->ioreq_evtchn,
>>                                        ({ state = p->state;
>>                                           smp_rmb();
>>                                           state != prev_state; }));
>> -            goto recheck;
>> +            continue;
>> +
> 
> You could just break out of the switch now, I guess.

I can't because of (see below).

> Anyway...
> 
> Reviewed-by: Paul Durrant <paul@xxxxxxx>

Thanks.

>>          default:
>>              gdprintk(XENLOG_ERR, "Weird HVM iorequest state %u\n", state);
>>              sv->pending = false;
>>              domain_crash(sv->vcpu->domain);
>>              return false; /* bail */
>>          }
>> -    } while ( false );
>> +
>> +        break;
>> +    }

This "break" requires the use of "continue" inside the switch().

Jan