io_uring/eventfd: ensure io_eventfd_signal() defers another RCU period

io_eventfd_do_signal() is invoked from an RCU callback, but when
dropping the reference to the io_ev_fd, it calls io_eventfd_fr

io_uring/eventfd: ensure io_eventfd_signal() defers another RCU period

io_eventfd_do_signal() is invoked from an RCU callback, but when
dropping the reference to the io_ev_fd, it calls io_eventfd_free()
directly if the refcount drops to zero. This isn't correct, as any
potential freeing of the io_ev_fd should be deferred another RCU grace
period.

Just call io_eventfd_put() rather than open-code the dec-and-test and
free, which will correctly defer it another RCU grace period.

Fixes: 21a091b970cd ("io_uring: signal registered eventfd to process deferred task work")
Reported-by: Jann Horn <[email protected]>
Cc: [email protected]
Tested-by: Li Zetao <[email protected]>
Reviewed-by: Li Zetao<[email protected]>
Reviewed-by: Prasanna Kumar T S M <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: move ctx->evfd_last_cq_tail into io_ev_fd

Everything else about the io_uring eventfd support is nicely kept
private to that code, except the cached_cq_tail tracking. With
everythin

io_uring/eventfd: move ctx->evfd_last_cq_tail into io_ev_fd

Everything else about the io_uring eventfd support is nicely kept
private to that code, except the cached_cq_tail tracking. With
everything else in place, move io_eventfd_flush_signal() to using
the ev_fd grab+release helpers, which then enables the direct use of
io_ev_fd for this tracking too.

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: abstract out ev_fd grab + release helpers

In preparation for needing the ev_fd grabbing (and releasing) from
another path, abstract out two helpers for that.

Link: https://lore.ke

io_uring/eventfd: abstract out ev_fd grab + release helpers

In preparation for needing the ev_fd grabbing (and releasing) from
another path, abstract out two helpers for that.

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: move trigger check into a helper

It's a bit hard to read what guards the triggering, move it into a
helper and add a comment explaining it too. This additionally moves
the ev_fd ==

io_uring/eventfd: move trigger check into a helper

It's a bit hard to read what guards the triggering, move it into a
helper and add a comment explaining it too. This additionally moves
the ev_fd == NULL check in there as well.

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: move actual signaling part into separate helper

In preparation for using this from multiple spots, move the signaling
into a helper.

Link: https://lore.kernel.org/r/20240921080307

io_uring/eventfd: move actual signaling part into separate helper

In preparation for using this from multiple spots, move the signaling
into a helper.

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: check for the need to async notifier earlier

It's not necessary to do this post grabbing a reference. With that, we
can drop the out goto path as well.

Link: https://lore.kernel.o

io_uring/eventfd: check for the need to async notifier earlier

It's not necessary to do this post grabbing a reference. With that, we
can drop the out goto path as well.

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: abstract out ev_fd put helper

We call this in two spot, have a helper for it. In preparation for
extending this part.

Link: https://lore.kernel.org/r/20240921080307.185186-2-axboe

io_uring/eventfd: abstract out ev_fd put helper

We call this in two spot, have a helper for it. In preparation for
extending this part.

Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: move refs to refcount_t

atomic_t for the struct io_ev_fd references and there are no issues with
it. While the ref getting and putting for the eventfd code is somewhat
performance

io_uring/eventfd: move refs to refcount_t

atomic_t for the struct io_ev_fd references and there are no issues with
it. While the ref getting and putting for the eventfd code is somewhat
performance critical for cases where eventfd signaling is used (news
flash, you should not...), it probably doesn't warrant using an atomic_t
for this. Let's just move to it to refcount_t to get the added
protection of over/underflows.

Link: https://lore.kernel.org/lkml/[email protected]/
Reported-by: kernel test robot <[email protected]>
Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring: add new line after variable declaration

Fixes checkpatch warning

Signed-off-by: Anuj Gupta <[email protected]>
Link: https://lore.kernel.org/r/20240902062134.136387-2-anuj20.g@samsung.

io_uring: add new line after variable declaration

Fixes checkpatch warning

Signed-off-by: Anuj Gupta <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>

show more ...

io_uring/eventfd: move eventfd handling to separate file

This is pretty nicely abstracted already, but let's move it to a separate
file rather than have it in the main io_uring file. With that, we c

io_uring/eventfd: move eventfd handling to separate file

This is pretty nicely abstracted already, but let's move it to a separate
file rather than have it in the main io_uring file. With that, we can
also move the io_ev_fd struct and enum out of global scope.

Signed-off-by: Jens Axboe <[email protected]>

show more ...