Re: [PATCH] ublk: skip blk_mq_tag_to_rq() bounds check

[Caleb Sander Mateos]

On Fri, Apr 11, 2025 at 12:56 AM Ming Lei <ming.lei@xxxxxxxxxx> wrote:
>
> On Tue, Apr 08, 2025 at 08:49:54PM -0600, Caleb Sander Mateos wrote:
> > The ublk driver calls blk_mq_tag_to_rq() in several places.
> > blk_mq_tag_to_rq() tolerates an invalid tag for the tagset, checking it
> > against the number of tags and returning NULL if it is out of bounds.
> > But all the calls from the ublk driver have already verified the tag
> > against the ublk queue's queue depth. In ublk_commit_completion(),
> > ublk_handle_need_get_data(), and case UBLK_IO_COMMIT_AND_FETCH_REQ, the
> > tag has already been checked in __ublk_ch_uring_cmd(). In
> > ublk_abort_queue(), the loop bounds the tag by the queue depth. In
> > __ublk_check_and_get_req(), the tag has already been checked in
> > __ublk_ch_uring_cmd(), in the case of ublk_register_io_buf(), or in
> > ublk_check_and_get_req().
> >
> > So just index the tagset's rqs array directly in the ublk driver.
> > Convert the tags to unsigned, as blk_mq_tag_to_rq() does.
>
> If blk_mq_tag_to_rq() turns out to be not efficient enough, we can kill it
> in fast path by storing it in ublk_io and sharing space with 'struct io_uring_cmd *',
> since the two's lifetime isn't overlapped basically.

I agree it would be nice to just store a pointer from in struct
ublk_io to its current struct request. I guess we would set it in
ubq_complete_io_cmd() and clear it in ublk_commit_completion()
(matching when UBLK_IO_FLAG_OWNED_BY_SRV is set), as well as in
ublk_timeout() for UBLK_F_UNPRIVILEGED_DEV?

I'm not sure it is possible to overlap the fields, though. When using
UBLK_U_IO_NEED_GET_DATA, the cmd field is overwritten with the a
pointer to the UBLK_U_IO_NEED_GET_DATA command, but the req would need
to be recorded earlier upon completion of the
UBLK_U_IO_(COMMIT_AND_)FETCH_REQ command. Would you be okay with 2
separate fields?

Best,
Caleb