| Age | Commit message (Collapse) | Author |
|---|
|
In preparation for adding multiple ifqs, add a helper returning a region
for mmaping zcrx refill queue. For now it's trivial and returns the same
ctx global ->zcrx_region.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/d9006e2ef8cd5e5b337c2ba2228ede8409eb60f2.1745141261.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
sqe->file_index and sqe->zcrx_ifq_idx are aliased, so even though
io_recvzc_prep() has all necessary handling and checking for
->zcrx_ifq_idx, it's already rejected a couple of lines above.
It's not a real problem, however, as we can only have one ifq at the
moment, which index is always zero, and it returns correct error
codes to the user.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/b51acedd326d1a87bf53e33303e6fc87661a3e3f.1745141261.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We'll need io_zcrx_iov_page at the top to keep offset calculations
closer together, move it there.
Reviewed-by: David Wei <dw@davidwei.uk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/575617033a8b84a5985c7eb760b7121efdbe7e56.1745141261.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Several lines below we already initialise the freelist, don't do it
twice.
Reviewed-by: David Wei <dw@davidwei.uk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/71b07c4e00d1db65899d1ed8d603d961fe7d1c28.1745141261.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
WARN_ON_ONCE() checking imu for NULL in io_import_fixed() is in the hot
path and too protective, it's time to get rid of it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/3782c01e311dbd474bca45aefaf79a2f2822fafb.1745083025.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We don't need special handling for the first page in
io_coalesce_buffer(), move it inside the loop.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Anuj Gupta <anuj20.g@samsung.com>
Link: https://lore.kernel.org/r/ad698cddc1eadb3d92a7515e95bb13f79420323d.1745083025.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We want to have a full folio to be left pinned but with only one
reference, for that we "unpin" all but the first page with
unpin_user_pages(), which can be confusing. There is a new helper to
achieve that called unpin_user_folio(), so use that.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Anuj Gupta <anuj20.g@samsung.com>
Link: https://lore.kernel.org/r/e0b2be8f9ea68f6b351ec3bb046f04f437f68491.1745083025.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
There are two helpers here, one assigns and increments the node ref
count, and the other is simply a wrapper around that for the buffer node
handling.
The buffer node assignment benefits from checking and setting
REQ_F_BUF_NODE together, otherwise stalls have been observed on setting
that flag later in the process. Hence re-do it so that it's set when
checked, and cleared in case of (unlikely) failure. With that, the
buffer node helper can go, and then drop the generic
io_req_assign_rsrc_node() helper as well as there's only a single user
of it left at that point.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
This works just like pipe2(2), except it also supports fixed file
descriptors. Used in a similar fashion as for other fd instantiating
opcodes (like accept, socket, open, etc), where sqe->file_slot is set
appropriately if two direct descriptors are desired rather than a set
of normal file descriptors.
sqe->addr must be set to a pointer to an array of 2 integers, which
is where the fixed/normal file descriptors are copied to.
sqe->pipe_flags contains flags, same as what is allowed for pipe2(2).
Future expansion of per-op private flags can go in sqe->ioprio,
like we do for other opcodes that take both a "syscall" flag set and
an io_uring opcode specific flag set.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Pass buffer group id into the rest of helpers via struct buf_sel_arg
and remove all reassignments of req->buf_index back to bgid. Now, it
only stores buffer indexes, and the group is provided by callers.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/3ea9fa08113ecb4d9224b943e7806e80a324bdf9.1743437358.git.asml.silence@gmail.com
Link: https://lore.kernel.org/io-uring/0c01d76ff12986c2f48614db8610caff8f78c869.1743500909.git.asml.silence@gmail.com/
[axboe: fold in patch from second link]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The current situation with buffer group id juggling is not ideal.
req->buf_index first stores the bgid, then it's overwritten by a buffer
id, and then it can get restored back no recycling / etc. It's not so
easy to control, and it's not handled consistently across request types
with receive requests saving and restoring the bgid it by hand.
It's a prep patch that adds a buffer group id argument to
io_buffer_select(). The caller will be responsible for stashing a copy
somewhere and passing it into the function.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/a210d6427cc3f4f42271a6853274cd5a50e56820.1743437358.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Move REQ_F_IMPORT_BUFFER to the common send setup. Currently, the only
user is send zc, but we'll want for normal sends to support that in the
future.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/18b74edfc61d8a1b6c9fed3b78a9276fe80f8ced.1743437358.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Prep code is interested whether it's a selected buffer request, not
whether a buffer has already been selected like what
io_do_buffer_select() returns. Check for REQ_F_BUFFER_SELECT directly.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/4488a029ac698554bebf732263fe19d7734affa6.1743437358.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
struct io_wq stores do_work and free_work function pointers which are
called on each work item. But these function pointers are always set to
io_wq_submit_work and io_wq_free_work, respectively. So remove these
function pointers and just call the functions directly.
Signed-off-by: Caleb Sander Mateos <csander@purestorage.com>
Link: https://lore.kernel.org/r/20250329161527.3281314-1-csander@purestorage.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
There is a problem with page pools not dma-unmapping immediately when
the device is going down, and delaying it until the page pool is
destroyed, which is not allowed (see links). That just got fixed for
normal page pools, and we need to address memory providers as well.
Unmap pages in the memory provider uninstall callback, and protect it
with a new lock. There is also a gap between when a dma mapping is
created and the mp is installed, so if the device is killed in between,
io_uring would be holding on to dma mappings to a dead device with no
one to call ->uninstall. Move it to page pool init and rely on
->is_mapped to make sure it's only done once.
Link: https://lore.kernel.org/lkml/8067f204-1380-4d37-8ffd-007fc6f26738@kernel.org/T/
Link: https://lore.kernel.org/all/20250409-page-pool-track-dma-v9-0-6a9ef2e0cba8@redhat.com/
Fixes: 34a3e60821ab9 ("io_uring/zcrx: implement zerocopy receive pp memory provider")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/ef9b7db249b14f6e0b570a1bb77ff177389f881c.1744965853.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
kbuf imports have the front offset adjusted and segments removed, but
the tail segments are still included in the segment count that gets
passed in the iov_iter. As the segments aren't necessarily all the
same size, move importing to a separate helper and iterate the
mapped length to get an exact count.
Reviewed-by: Nitesh Shetty <nj.shetty@samsung.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Sending exact nr_segs, avoids bio split check and processing in
block layer, which takes around 5%[1] of overall CPU utilization.
In our setup, we see overall improvement of IOPS from 7.15M to 7.65M [2]
and 5% less CPU utilization.
[1]
3.52% io_uring [kernel.kallsyms] [k] bio_split_rw_at
1.42% io_uring [kernel.kallsyms] [k] bio_split_rw
0.62% io_uring [kernel.kallsyms] [k] bio_submit_split
[2]
sudo taskset -c 0,1 ./t/io_uring -b512 -d128 -c32 -s32 -p1 -F1 -B1 -n2
-r4 /dev/nvme0n1 /dev/nvme1n1
Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>
[Pavel: fixed for kbuf, rebased and reworked on top of cleanups]
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/7a1a49a8d053bd617c244291d63dbfbc07afde36.1744882081.git.asml.silence@gmail.com
[axboe: fold in fix factoring in buf reg offset]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_import_fixed is a mess. Even though we know the final len of the
iterator, we still assign offset + len and do some magic after to
correct for that.
Do offset calculation first and finalise it with iov_iter_bvec at the
end.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/2d5107fed24f8b23245ef2ede9a5a7f7c426df61.1744882081.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Kernel registered buffers are special because segments are not uniform
in size, and we have a bunch of optimisations based on that uniformity
for normal buffers. Handle kbuf separately, it'll be cleaner this way.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/4e9e5990b0ab5aee723c0be5cd9b5bcf810375f9.1744882081.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Don't optimise for requests with offset=0. Large registered buffers are
the preference and hence the user is likely to pass an offset, and the
adjustments are not expensive and will be made even cheaper in following
patches.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/1c2beb20470ee3c886a363d4d8340d3790db19f3.1744882081.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
It'll likely change how page pools store memory providers, so in
preparation for that, keep accesses in one place in io_uring by
introducing a helper.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/3522eb8fa9b4e21bcf32e7e9ae656c616b282210.1744722526.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
IORING_OP_RECV_ZC requests take a zcrx object id via sqe::zcrx_ifq_idx,
which binds it to the corresponding if / queue. However, we don't return
that id back to the user. It's fine as currently there can be only one
zcrx and the user assumes that its id should be 0, but as we'll need
multiple zcrx objects in the future let's explicitly pass it back on
registration.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/8714667d370651962f7d1a169032e5f02682a73e.1744722517.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
This isn't fixing a real issue, but there's also zero point in going
through group and buffer setup, when the buffers are going to be
rejected once attempted to get used.
Cc: stable@vger.kernel.org
Reported-by: syzbot+58928048fd1416f1457c@syzkaller.appspotmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
A preparation patch that separates the number of pages / folios from
the number of niovs. They will not match in the future to support huge
pages, improved dma mapping and/or larger chunk sizes.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/0780ac966ee84200385737f45bb0f2ada052392b.1743848231.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>
|
|
Buffer / file table registration is all or nothing, if it fails all
resources we might have partially registered are dropped and the table
is killed. If that happens, it doesn't make sense to post any rsrc tag
CQEs. That would be confusing to the application, which should not need
to handle that case.
Cc: stable@vger.kernel.org
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Fixes: 7029acd8a9503 ("io_uring/rsrc: get rid of per-ring io_rsrc_node list")
Link: https://lore.kernel.org/r/c514446a8dcb0197cddd5d4ba8f6511da081cf1f.1743777957.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_uring always switches requests to atomic refcounting for iowq
execution before there is any parallilism by setting REQ_F_REFCOUNT,
and the flag is not cleared until the request completes. That should be
fine as long as the compiler doesn't make up a non existing value for
the flags, however KCSAN still complains when the request owner changes
oter flag bits:
BUG: KCSAN: data-race in io_req_task_cancel / io_wq_free_work
...
read to 0xffff888117207448 of 8 bytes by task 3871 on cpu 0:
req_ref_put_and_test io_uring/refs.h:22 [inline]
Skip REQ_F_REFCOUNT checks for iowq, we know it's set.
Reported-by: syzbot+903a2ad71fb3f1e47cf5@syzkaller.appspotmail.com
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/d880bc27fb8c3209b54641be4ff6ac02b0e5789a.1743679736.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_uring has supported fixed kernel buffer via io_buffer_register_bvec()
and io_buffer_unregister_bvec().
The vectored fixed buffer has been ready, so it is natural to support
fixed kernel buffer, one use case is ublk.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20250325135155.935398-4-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Add helper of validate_fixed_range() for validating fixed buffer
range.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Caleb Sander Mateos <csander@purestorage.com>
Link: https://lore.kernel.org/r/20250325135155.935398-2-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
When readlen is set for a recvzc request, tcp_read_sock() will call
io_zcrx_recv_skb() one final time with len == desc->count == 0. This is
caused by the !desc->count check happening too late. The offset + 1 !=
skb->len happens earlier and causes the while loop to continue.
Fix this in io_zcrx_recv_skb() instead of tcp_read_sock(). Return early
if len is 0 i.e. the read is done.
Fixes: 6699ec9a23f8 ("io_uring/zcrx: add a read limit to recvzc requests")
Signed-off-by: David Wei <dw@davidwei.uk>
Link: https://lore.kernel.org/r/20250401195355.1613813-1-dw@davidwei.uk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
With registered buffers we set up iterators in helpers like
io_import_fixed(), and there is no need for a import_ubuf() before that.
It was fine as we used real pointers for offset calculation, but that's
not the case anymore since introduction of ublk kernel buffers.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/9b2de1a50844f848f62c8de609b494971033a6b9.1743437358.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We're relying on callers to verify the IO size, do it inside of
io_import_fixed() instead. It's safer, easier to deal with, and more
consistent as now it's done close to the iter init site.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/f9c2c75ec4d356a0c61289073f68d98e8a9db190.1743446271.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Add a local variable for the sqe pointer to avoid repetition.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/8dbac0f9acda2d3842534eeb7ce10d9276b021ae.1743357108.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
There is now an io_uring private part of cmd async_data, move saved sqe
into it. Drivers are accessing it via struct io_uring_cmd::cmd.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/ecbe078dd57acefdbc4366d083327086c0879378.1743357121.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Reflect in the kconfig that zcrx requires io_uring compiled.
Fixes: 6f377873cb239 ("io_uring/zcrx: add interface queue and refill queue")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/8047135a344e79dbd04ee36a7a69cc260aabc2ca.1743356260.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We're caching some of file related request flags in a tricky way, put
a build check to make sure flags don't get reshuffled.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/9877577b83c25dd78224a8274f799187e7ec7639.1743407551.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Unlike earlier versions, io_msg_remote_post() creates a valid request
with a proper context, so don't pass a context to
io_req_task_work_add_remote() explicitly but derive it from the request.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/721f51cf34996d98b48f0bfd24ad40aa2730167e.1743190078.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
It's risky to have msg request opcode set to garbage, so at least
initialise it to nop. Later we might want to add a user inaccessible
opcode for such cases.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/9afe650fcb348414a4529d89f52eb8969ba06efd.1743190078.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_double_lock_ctx() doesn't lock both rings. Rename it to prevent any
future confusion.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/9e5defa000efd9b0f5e169cbb6bad4994d46ec5c.1743190078.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_send_setup() already sets up the iterator for IORING_OP_SEND_ZC, we
don't need repeating that at issue time. Move it all together with mem
accounting at prep time, which is more consistent with how the non-zc
version does that.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/eb54f007c493ad9f4ca89aa8e715baf30d83fb88.1743202294.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
In preparation to the next patch, set ->sg_from_iter callback at request
prep time.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/5fe2972701df3bacdb3d760bce195fa640bee201.1743202294.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We have multiple branches at prep for send vs sendmsg handling, put them
together so that the variant handling is more localised.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/33abf666d9ded74cba4da2f0d9fe58e88520dffe.1743202294.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_sendmsg_zc_setup() duplicates parts of io_sendmsg_setup(), and the
only difference between them is that the former support vectored
registered buffers with nothing zerocopy specific. Merge them together,
we want regular sendmsg to eventually support fixed buffers either way.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/7e5ec40f9dc93355399dc6fa0cbc8b31f0b20ac5.1743202294.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Save an instruction / trip to the cache and assign some of sendzc flags
together.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/c519d6f406776c3be3ef988a8339a88e45d1ffd9.1743202294.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Get rid of io_net_vec_assign() by open coding it into its only caller.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/19191c34b5cfe1161f7eeefa6e785418ea9ad56d.1743202294.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_sendmsg_setup() is trivial and io_sendmsg_copy_hdr() doesn't add
any good abstraction, open code one into another.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/565318ce585665e88053663eeee5178d2c15692f.1743202294.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Account pinned pages for IORING_OP_SENDMSG_ZC, just as we for
IORING_OP_SEND_ZC and net/ does for MSG_ZEROCOPY.
Fixes: 493108d95f146 ("io_uring/net: zerocopy sendmsg")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/4f00f67ca6ac8e8ed62343ae92b5816b1e0c9c4b.1743086313.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Pull more io_uring updates from Jens Axboe:
"Final separate updates for io_uring.
This started out as a series of cleanups improvements and improvements
for registered buffers, but as the last series of the io_uring changes
for 6.15, it also collected a few fixes for the other branches on top:
- Add support for vectored fixed/registered buffers.
Previously only single segments have been supported for commands,
now vectored variants are supported as well. This series includes
networking and file read/write support.
- Small series unifying return codes across multi and single shot.
- Small series cleaning up registerd buffer importing.
- Adding support for vectored registered buffers for uring_cmd.
- Fix for io-wq handling of command reissue.
- Various little fixes and tweaks"
* tag 'for-6.15/io_uring-reg-vec-20250327' of git://git.kernel.dk/linux: (25 commits)
io_uring/net: fix io_req_post_cqe abuse by send bundle
io_uring/net: use REQ_F_IMPORT_BUFFER for send_zc
io_uring: move min_events sanitisation
io_uring: rename "min" arg in io_iopoll_check()
io_uring: open code __io_post_aux_cqe()
io_uring: defer iowq cqe overflow via task_work
io_uring: fix retry handling off iowq
io_uring/net: only import send_zc buffer once
io_uring/cmd: introduce io_uring_cmd_import_fixed_vec
io_uring/cmd: add iovec cache for commands
io_uring/cmd: don't expose entire cmd async data
io_uring: rename the data cmd cache
io_uring: rely on io_prep_reg_vec for iovec placement
io_uring: introduce io_prep_reg_iovec()
io_uring: unify STOP_MULTISHOT with IOU_OK
io_uring: return -EAGAIN to continue multishot
io_uring: cap cached iovec/bvec size
io_uring/net: implement vectored reg bufs for zctx
io_uring/net: convert to struct iou_vec
io_uring/net: pull vec alloc out of msghdr import
...
|
|
Pull io_uring epoll support from Jens Axboe:
"This adds support for reading epoll events via io_uring.
While this may seem counter-intuitive (and/or productive), the
reasoning here is that quite a few existing epoll event loops can
easily do a partial conversion to a completion based model, but are
still stuck with one (or few) event types that remain readiness based.
For that case, they then need to add the io_uring fd to the epoll
context, and continue to rely on epoll_wait(2) for waiting on events.
This misses out on the finer grained waiting that io_uring can do, to
reduce context switches and wait for multiple events in one batch
reliably.
With adding support for reaping epoll events via io_uring, the whole
legacy readiness based event types can still be reaped via epoll, with
the overall waiting in the loop be driven by io_uring"
* tag 'for-6.15/io_uring-epoll-wait-20250325' of git://git.kernel.dk/linux:
io_uring/epoll: add support for IORING_OP_EPOLL_WAIT
io_uring/epoll: remove CONFIG_EPOLL guards
|
|
Pull io_uring zero-copy receive support from Jens Axboe:
"This adds support for zero-copy receive with io_uring, enabling fast
bulk receive of data directly into application memory, rather than
needing to copy the data out of kernel memory.
While this version only supports host memory as that was the initial
target, other memory types are planned as well, with notably GPU
memory coming next.
This work depends on some networking components which were queued up
on the networking side, but have now landed in your tree.
This is the work of Pavel Begunkov and David Wei. From the v14 posting:
'We configure a page pool that a driver uses to fill a hw rx queue
to hand out user pages instead of kernel pages. Any data that ends
up hitting this hw rx queue will thus be dma'd into userspace
memory directly, without needing to be bounced through kernel
memory. 'Reading' data out of a socket instead becomes a
_notification_ mechanism, where the kernel tells userspace where
the data is. The overall approach is similar to the devmem TCP
proposal
This relies on hw header/data split, flow steering and RSS to
ensure packet headers remain in kernel memory and only desired
flows hit a hw rx queue configured for zero copy. Configuring this
is outside of the scope of this patchset.
We share netdev core infra with devmem TCP. The main difference is
that io_uring is used for the uAPI and the lifetime of all objects
are bound to an io_uring instance. Data is 'read' using a new
io_uring request type. When done, data is returned via a new shared
refill queue. A zero copy page pool refills a hw rx queue from this
refill queue directly. Of course, the lifetime of these data
buffers are managed by io_uring rather than the networking stack,
with different refcounting rules.
This patchset is the first step adding basic zero copy support. We
will extend this iteratively with new features e.g. dynamically
allocated zero copy areas, THP support, dmabuf support, improved
copy fallback, general optimisations and more'
In a local setup, I was able to saturate a 200G link with a single CPU
core, and at netdev conf 0x19 earlier this month, Jamal reported
188Gbit of bandwidth using a single core (no HT, including soft-irq).
Safe to say the efficiency is there, as bigger links would be needed
to find the per-core limit, and it's considerably more efficient and
faster than the existing devmem solution"
* tag 'for-6.15/io_uring-rx-zc-20250325' of git://git.kernel.dk/linux:
io_uring/zcrx: add selftest case for recvzc with read limit
io_uring/zcrx: add a read limit to recvzc requests
io_uring: add missing IORING_MAP_OFF_ZCRX_REGION in io_uring_mmap
io_uring: Rename KConfig to Kconfig
io_uring/zcrx: fix leaks on failed registration
io_uring/zcrx: recheck ifq on shutdown
io_uring/zcrx: add selftest
net: add documentation for io_uring zcrx
io_uring/zcrx: add copy fallback
io_uring/zcrx: throttle receive requests
io_uring/zcrx: set pp memory provider for an rx queue
io_uring/zcrx: add io_recvzc request
io_uring/zcrx: dma-map area for the device
io_uring/zcrx: implement zerocopy receive pp memory provider
io_uring/zcrx: grab a net device
io_uring/zcrx: add io_zcrx_area
io_uring/zcrx: add interface queue and refill queue
|
