Re: [PATCH for-6.1? v2 6/7] mirror: Check job_is_cancelled() earlier

From: Max Reitz <mreitz@redhat.com>
To: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>,
	qemu-block@nongnu.org
Cc: Kevin Wolf <kwolf@redhat.com>, John Snow <jsnow@redhat.com>,
	qemu-devel@nongnu.org
Subject: Re: [PATCH for-6.1? v2 6/7] mirror: Check job_is_cancelled() earlier
Date: Tue, 27 Jul 2021 17:40:45 +0200	[thread overview]
Message-ID: <3771f388-52e0-50e3-29cb-2d173d314cfc@redhat.com> (raw)
In-Reply-To: <be900ee5-f9c4-cee6-e3c6-66aeb0a0d89f@virtuozzo.com>

On 27.07.21 15:13, Vladimir Sementsov-Ogievskiy wrote:
> 26.07.2021 17:46, Max Reitz wrote:
>> We must check whether the job is force-cancelled early in our main loop,
>> most importantly before any `continue` statement.  For example, we used
>> to have `continue`s before our current checking location that are
>> triggered by `mirror_flush()` failing.  So, if `mirror_flush()` kept
>> failing, force-cancelling the job would not terminate it.
>>
>> A job being force-cancelled should be treated the same as the job having
>> failed, so put the check in the same place where we check `s->ret < 0`.
>>
>> Buglink: https://gitlab.com/qemu-project/qemu/-/issues/462
>> Signed-off-by: Max Reitz <mreitz@redhat.com>
>> ---
>>   block/mirror.c | 7 +------
>>   1 file changed, 1 insertion(+), 6 deletions(-)
>>
>> diff --git a/block/mirror.c b/block/mirror.c
>> index 72e02fa34e..46d1a1e5a2 100644
>> --- a/block/mirror.c
>> +++ b/block/mirror.c
>> @@ -993,7 +993,7 @@ static int coroutine_fn mirror_run(Job *job, 
>> Error **errp)
>>               mirror_wait_for_any_operation(s, true);
>>           }
>>   -        if (s->ret < 0) {
>> +        if (s->ret < 0 || job_is_cancelled(&s->common.job)) {
>>               ret = s->ret;
>>               goto immediate_exit;
>>           }
>> @@ -1078,8 +1078,6 @@ static int coroutine_fn mirror_run(Job *job, 
>> Error **errp)
>>               break;
>>           }
>>   -        ret = 0;
>> -
>
> That's just a cleanup, that statement is useless pre-patch, yes?

I think it was intended for if we left this loop via the 
job_is_cancelled() condition below.  Since it’s removed, this statement 
seems meaningless, so I removed it along with the `break`.

Max

>
>>           if (job_is_ready(&s->common.job) && !should_complete) {
>>               delay_ns = (s->in_flight == 0 &&
>>                           cnt == 0 ? BLOCK_JOB_SLICE_TIME : 0);
>> @@ -1087,9 +1085,6 @@ static int coroutine_fn mirror_run(Job *job, 
>> Error **errp)
>>           trace_mirror_before_sleep(s, cnt, 
>> job_is_ready(&s->common.job),
>>                                     delay_ns);
>>           job_sleep_ns(&s->common.job, delay_ns);
>> -        if (job_is_cancelled(&s->common.job)) {
>> -            break;
>> -        }
>>           s->last_pause_ns = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
>>       }
>>
>
> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
>