[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <8e9bf420-0405-fd71-826d-7924528f2d09@kernel.dk>
Date: Fri, 6 May 2022 15:29:26 -0600
From: Jens Axboe <axboe@...nel.dk>
To: Hao Xu <haoxu.linux@...il.com>, io-uring@...r.kernel.org
Cc: Pavel Begunkov <asml.silence@...il.com>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 5/5] io_uring: implement multishot mode for accept
On 5/6/22 2:50 PM, Jens Axboe wrote:
> On 5/6/22 1:01 AM, Hao Xu wrote:
>> @@ -5748,8 +5758,12 @@ static int io_accept(struct io_kiocb *req, unsigned int issue_flags)
>> if (!fixed)
>> put_unused_fd(fd);
>> ret = PTR_ERR(file);
>> - if (ret == -EAGAIN && force_nonblock)
>> - return -EAGAIN;
>> + if (ret == -EAGAIN && force_nonblock) {
>> + if ((req->flags & REQ_F_APOLL_MULTI_POLLED) ==
>> + REQ_F_APOLL_MULTI_POLLED)
>> + ret = 0;
>> + return ret;
>
> FWIW, this
>
> if ((req->flags & REQ_F_APOLL_MULTI_POLLED) == REQ_F_APOLL_MULTI_POLLED)
>
> is identical to
>
> if (req->flags & REQ_F_APOLL_MULTI_POLLED)
>
> but I suspect this used to check more flags (??), because as it stands
> it seems a bit nonsensical.
Looking deeper, it is indeed a mask and not a single flag! So the check
looks fine.
--
Jens Axboe
Powered by blists - more mailing lists