Re: [PATCH 5/5] io_uring: batch completion in prior_task_list

[Hao Xu <haoxu@linux.alibaba.com>]

Hi,

在 2021/12/8 上午5:16, Pavel Begunkov 写道:
> On 12/7/21 21:01, Pavel Begunkov wrote:
>> On 12/7/21 09:39, Hao Xu wrote:
>>> In previous patches, we have already gathered some tw with
>>> io_req_task_complete() as callback in prior_task_list, let's complete
>>> them in batch while we cannot grab uring lock. In this way, we batch
>>> the req_complete_post path.
> [...]
>>> +        if (likely(*uring_locked))
>>> +            req->io_task_work.func(req, uring_locked);
>>> +        else
>>> +            __io_req_complete_post(req, req->result, 
>>> io_put_kbuf(req));
>>
>> I think there is the same issue as last time, first iteration of 
>> tctx_task_work()
>> sets ctx but doesn't get uring_lock. Then you go here, find a request 
>> with the
>> same ctx and end up here with locking.
>
> Maybe something like below on top? Totally untested. We basically always
> want *uring_locked != *compl_locked, so we don't even need to to store
> both vars.
>
Thanks, Pavel. I Forgot to think about when req->ctx == ctx, tested the 
fix here

and it looks good.


Regards,
Hao