Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH v2 0/4] open/accept directly into io_uring fixed file
 table
To:     Pavel Begunkov <asml.silence@gmail.com>,
        Stefan Metzmacher <metze@samba.org>, io-uring@vger.kernel.org,
        Josh Triplett <josh@joshtriplett.org>
Cc:     "David S . Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>, linux-kernel@vger.kernel.org,
        netdev@vger.kernel.org
References: <cover.1628871893.git.asml.silence@gmail.com>
 <17841c48-093e-af1c-c7c9-aa00859eb1b9@samba.org>
 <78e2d63a-5d3a-6334-8177-11646d4ec261@gmail.com>
From:   Jens Axboe <axboe@kernel.dk>
Message-ID: <ab450460-1faa-dea1-f8a3-894a42461597@kernel.dk>
Date:   Tue, 17 Aug 2021 08:57:13 -0600
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <78e2d63a-5d3a-6334-8177-11646d4ec261@gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 8/17/21 3:33 AM, Pavel Begunkov wrote:
> On 8/16/21 4:45 PM, Stefan Metzmacher wrote:
>> Hi Pavel,
>>
>>> The behaviour is controlled by setting sqe->file_index, where 0 implies
>>> the old behaviour. If non-zero value is specified, then it will behave
>>> as described and place the file into a fixed file slot
>>> sqe->file_index - 1. A file table should be already created, the slot
>>> should be valid and empty, otherwise the operation will fail.
>>>
>>> Note 1: we can't use IOSQE_FIXED_FILE to switch between modes, because
>>> accept takes a file, and it already uses the flag with a different
>>> meaning.
>>
>> Would it be hard to support IOSQE_FIXED_FILE for the dirfd of openat*, renameat, unlinkat, statx?
>> (And mkdirat, linkat, symlinkat when they arrive)
>> renameat and linkat might be trickier as they take two dirfds, but it
>> would make the feature more complete and useful.
> 
> Good idea. There is nothing blocking on the io_uring side, but
> the fs part may get ugly, e.g. too intrusive. We definitely need
> to take a look

Indeed, the io_uring side is trivial, but the VFS interface would
require a lot of man handling... That's why I didn't add support for
fixed files originally.

-- 
Jens Axboe