zcrx tries to detach ifq / terminate page pools when the io_uring ctx
owning it is being destroyed. There will be multiple io_uring instances
attached to it in the future, so add a separate counter to track the
users. Note, refs can't be reused for this purpose as it only used to
prevent zcrx and rings destruction, and also used by page pools to keep
it alive.
Signed-off-by: David Wei <dw@davidwei.uk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It'll be annoying and take enough of boilerplate code to implement
new zcrx features as separate io_uring register opcode. Introduce
IORING_REGISTER_ZCRX_CTRL that will multiplex such calls to zcrx.
Note, there are no real users of the opcode in this patch.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Merge 6.18-rc io_uring fixes, as certain coming changes depend on some
of these.
* io_uring-6.18:
io_uring/rsrc: don't use blk_rq_nr_phys_segments() as number of bvecs
io_uring/query: return number of available queries
io_uring/rw: ensure allocated iovec gets cleared for early failure
io_uring: fix regbuf vector size truncation
io_uring: fix types for region size calulation
io_uring/zcrx: remove sync refill uapi
io_uring: fix buffer auto-commit for multishot uring_cmd
io_uring: correct __must_hold annotation in io_install_fixed_file
io_uring zcrx: add MAINTAINERS entry
io_uring: Fix code indentation error
io_uring/sqpoll: be smarter on when to update the stime usage
io_uring/sqpoll: switch away from getrusage() for CPU accounting
io_uring: fix incorrect unlikely() usage in io_waitid_prep()
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a refcount to struct io_zcrx_ifq to reverse the refcounting
relationship i.e. rings now reference ifqs instead. As a result of this,
remove ctx->refs that an ifq holds on a ring via the page pool memory
provider.
This ref ifq->refs is held by internal users of an ifq, namely rings and
the page pool memory provider associated with an ifq. This is needed to
keep the ifq around until the page pool is destroyed.
Since ifqs now no longer hold refs to ring ctx, there isn't a need to
split the cleanup of ifqs into two: io_shutdown_zcrx_ifqs() in
io_ring_exit_work() while waiting for ctx->refs to drop to 0, and
io_unregister_zcrx_ifqs() after. Remove io_shutdown_zcrx_ifqs().
Signed-off-by: David Wei <dw@davidwei.uk>
Co-developed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
In preparation for removing ifq->ctx and making ifq lifetime independent
of ring ctx, add user_struct and mm_struct to io_zcrx_ifq.
In the ifq cleanup path, these are the only fields used from the main
ring ctx to do accounting. Taking a copy in the ifq allows ifq->ctx to
be removed later, including the ctx->refs held by the ifq.
Signed-off-by: David Wei <dw@davidwei.uk>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There is a better way to handle the problem IORING_REGISTER_ZCRX_REFILL
solves. The uapi can also be slightly adjusted to accommodate future
extensions. Remove the feature for now, it'll be reworked for the next
release.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Returning buffers via a ring is performant and convenient, but it
becomes a problem when/if the user misconfigured the ring size and it
becomes full. Add a synchronous way to return buffers back to the page
pool via a new register opcode. It's supposed to be a reliable slow
path for refilling.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Remove ifq->lock and reuse pp_lock to protect the netdev pointer.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
In preparation for reusing the lock for other purposes, rename it to
"pp_lock". As before, it can be taken deeper inside the networking stack
by page pool, and so the syscall io_uring must avoid holding it while
doing queue reconfiguration or anything that can result in immediate pp
init/destruction.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of using PAGE_SIZE for the niov size add a niov_shift field to
ifq, and patch up all important places. Copy fallback still assumes
PAGE_SIZE, so it'll be wasting some memory for now.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Set struct io_zcrx_mem::sgt for umem areas as well to simplify looking
up the current sg table.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It was removed from uapi, so now it's always 0 and can be removed
together with offset handling in io_populate_area_dma().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Refill queue region is a part of zcrx and should stay in struct
io_zcrx_ifq. We can't have multiple queues without it, so move it there.
As a result there is no context global zcrx region anymore, and the
region is looked up together with its ifq. To protect a concurrent mmap
from seeing an inconsistent region we were protecting changes to
->zcrx_region with mmap_lock, but now it protect the publishing of the
ifq.
Reviewed-by: David Wei <dw@davidwei.uk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/24f1a728fc03d0166f16d099575457e10d9d90f2.1745141261.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Refill queue lock and other bits are only used from the allocation path
on the rx softirq side, but it shares the cache line with other fields
like ctx that are used also in the "syscall" path, which causes cache
bouncing when softirq runs on a different CPU.
Separate them into different cache lines. The first one now contains
constant fields used by both contextx, followed by a line responsible
for refill queue data.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/6d1f598e27d623c07fc49d6baee13089a9b1216c.1743848241.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently multishot recvzc requests have no read limit and will remain
active so as long as the socket remains open. But, there are sometimes a
need to do a fixed length read e.g. peeking at some data in the socket.
Add a length limit to recvzc requests `len`. A value of 0 means no limit
which is the previous behaviour. A positive value N specifies how many
bytes to read from the socket.
Data will still be posted in aux completions, as before. This could be
split across multiple frags. But the primary recvzc request will now
complete once N bytes have been read. The completion of the recvzc
request will have res and cflags both set to 0.
Signed-off-by: David Wei <dw@davidwei.uk>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/20250224041319.2389785-2-dw@davidwei.uk
[axboe: fixup io_zcrx_recv() for !CONFIG_NET]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add io_uring opcode OP_RECV_ZC for doing zero copy reads out of a
socket. Only the connection should be land on the specific rx queue set
up for zero copy, and the socket must be handled by the io_uring
instance that the rx queue was registered for zero copy with. That's
because neither net_iovs / buffers from our queue can be read by outside
applications, nor zero copy is possible if traffic for the zero copy
connection goes to another queue. This coordination is outside of the
scope of this patch series. Also, any traffic directed to the zero copy
enabled queue is immediately visible to the application, which is why
CAP_NET_ADMIN is required at the registration step.
Of course, no data is actually read out of the socket, it has already
been copied by the netdev into userspace memory via DMA. OP_RECV_ZC
reads skbs out of the socket and checks that its frags are indeed
net_iovs that belong to io_uring. A cqe is queued for each one of these
frags.
Recall that each cqe is a big cqe, with the top half being an
io_uring_zcrx_cqe. The cqe res field contains the len or error. The
lower IORING_ZCRX_AREA_SHIFT bits of the struct io_uring_zcrx_cqe::off
field contain the offset relative to the start of the zero copy area.
The upper part of the off field is trivially zero, and will be used
to carry the area id.
For now, there is no limit as to how much work each OP_RECV_ZC request
does. It will attempt to drain a socket of all available data. This
request always operates in multishot mode.
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: David Wei <dw@davidwei.uk>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20250215000947.789731-7-dw@davidwei.uk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Implement a page pool memory provider for io_uring to receieve in a
zero copy fashion. For that, the provider allocates user pages wrapped
around into struct net_iovs, that are stored in a previously registered
struct net_iov_area.
Unlike the traditional receive, that frees pages and returns them back
to the page pool right after data was copied to the user, e.g. inside
recv(2), we extend the lifetime until the user space confirms that it's
done processing the data. That's done by taking a net_iov reference.
When the user is done with the buffer, it must return it back to the
kernel by posting an entry into the refill ring, which is usually polled
off the io_uring memory provider callback in the page pool's netmem
allocation path.
There is also a separate set of per net_iov "user" references accounting
whether a buffer is currently given to the user (including possible
fragmentation).
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Mina Almasry <almasrymina@google.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: David Wei <dw@davidwei.uk>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20250215000947.789731-5-dw@davidwei.uk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add io_zcrx_area that represents a region of userspace memory that is
used for zero copy. During ifq registration, userspace passes in the
uaddr and len of userspace memory, which is then pinned by the kernel.
Each net_iov is mapped to one of these pages.
The freelist is a spinlock protected list that keeps track of all the
net_iovs/pages that aren't used.
For now, there is only one area per ifq and area registration happens
implicitly as part of ifq registration. There is no API for
adding/removing areas yet. The struct for area registration is there for
future extensibility once we support multiple areas and TCP devmem.
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: David Wei <dw@davidwei.uk>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20250215000947.789731-3-dw@davidwei.uk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a new object called an interface queue (ifq) that represents a net
rx queue that has been configured for zero copy. Each ifq is registered
using a new registration opcode IORING_REGISTER_ZCRX_IFQ.
The refill queue is allocated by the kernel and mapped by userspace
using a new offset IORING_OFF_RQ_RING, in a similar fashion to the main
SQ/CQ. It is used by userspace to return buffers that it is done with,
which will then be re-used by the netdev again.
The main CQ ring is used to notify userspace of received data by using
the upper 16 bytes of a big CQE as a new struct io_uring_zcrx_cqe. Each
entry contains the offset + len to the data.
For now, each io_uring instance only has a single ifq.
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: David Wei <dw@davidwei.uk>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20250215000947.789731-2-dw@davidwei.uk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
