Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Breno Leitao <leitao@debian.org>
To:     asml.silence@gmail.com, axboe@kernel.dk
Cc:     io-uring@vger.kernel.org, netdev@vger.kernel.org,
        davem@davemloft.net, kuba@kernel.org, edumazet@google.com,
        pabeni@redhat.com, linux-kernel@vger.kernel.org
Subject: [RFC PATCH 0/3] io_uring: Initial support for {s,g}etsockopt commands
Date:   Wed, 19 Jul 2023 03:27:34 -0700
Message-Id: <20230719102737.2513246-1-leitao@debian.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

This patchset adds support for SOL_SOCKET level getsockopt and
setsockopt in io_uring, using the command op. SOL_SOCKET seems to be the
most common level parameter for get/setsockopt(2).

This implementation benefits from the work done to leverage sockptr_t in
SOL_SOCKET path.

For getsockopt command, the optlen field is not a userspace
pointers, but an absolute value, so this is slightly different from
getsockopt(2) behaviour. The updated value is returned in cqe->res.

If this approach is good enough, I am planning to extend the support for
those "levels" that have already implemented sockptr_t support

This patch was tested with a new test[1] in liburing.
This patch depends on "io_uring: Add io_uring command support for sockets"[2]

[1] Link: https://github.com/leitao/liburing/blob/getsock/test/socket-getsetsock-cmd.c
[2] Link: https://lore.kernel.org/all/20230627134424.2784797-1-leitao@debian.org/

Breno Leitao (3):
  net: expose sock_use_custom_sol_socket
  io_uring/cmd: Add support for getsockopt command
  io_uring/cmd: Add support for set_sockopt

 include/linux/net.h           |  5 ++++
 include/uapi/linux/io_uring.h |  8 ++++++
 io_uring/uring_cmd.c          | 51 +++++++++++++++++++++++++++++++++++
 net/socket.c                  |  5 ----
 4 files changed, 64 insertions(+), 5 deletions(-)

-- 
2.34.1