io_uring/cmd_net: fix too strict requirement on ioctl
authorAsbjørn Sloth Tønnesen <ast@fiberby.net>
Mon, 16 Feb 2026 10:27:18 +0000 (10:27 +0000)
committerJens Axboe <axboe@kernel.dk>
Mon, 16 Feb 2026 15:08:01 +0000 (08:08 -0700)
commit600b665b903733bd60334e86031b157cc823ee55
tree9c8452c00c7030377552c9cd2a6943fe79147471
parent56112578c71213a10c995a56835bddb5e9ab1ed0
io_uring/cmd_net: fix too strict requirement on ioctl

Attempting SOCKET_URING_OP_SETSOCKOPT on an AF_NETLINK socket resulted
in an -EOPNOTSUPP, as AF_NETLINK doesn't have an ioctl in its struct
proto, but only in struct proto_ops.

Prior to the blamed commit, io_uring_cmd_sock() only had two cmd_op
operations, both requiring ioctl, thus the check was warranted.

Since then, 4 new cmd_op operations have been added, none of which
depend on ioctl. This patch moves the ioctl check, so it only applies
to the original operations.

AFAICT, the ioctl requirement was unintentional, and it wasn't
visible in the blamed patch within 3 lines of context.

Cc: stable@vger.kernel.org
Fixes: a5d2f99aff6b ("io_uring/cmd: Introduce SOCKET_URING_OP_GETSOCKOPT")
Signed-off-by: Asbjørn Sloth Tønnesen <ast@fiberby.net>
Reviewed-by: Gabriel Krisman Bertazi <krisman@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_uring/cmd_net.c