korg/linux
0
0
Fork 0
mirror of https://mirrors.bfsu.edu.cn/git/linux.git synced 2024-11-15 08:14:15 +08:00
Code Issues Packages Projects Releases Wiki Activity
0fe2b18ddc
linux/io_uring/notif.c

133 lines
3.2 KiB
C
Raw Normal View History

io_uring: add zc notification infrastructure Add internal part of send zerocopy notifications. There are two main structures, the first one is struct io_notif, which carries inside struct ubuf_info and maps 1:1 to it. io_uring will be binding a number of zerocopy send requests to it and ask to complete (aka flush) it. When flushed and all attached requests and skbs complete, it'll generate one and only one CQE. There are intended to be passed into the network layer as struct msghdr::msg_ubuf. The second concept is notification slots. The userspace will be able to register an array of slots and subsequently addressing them by the index in the array. Slots are independent of each other. Each slot can have only one notifier at a time (called active notifier) but many notifiers during the lifetime. When active, a notifier not going to post any completion but the userspace can attach requests to it by specifying the corresponding slot while issueing send zc requests. Eventually, the userspace will want to "flush" the notifier losing any way to attach new requests to it, however it can use the next atomatically added notifier of this slot or of any other slot. When the network layer is done with all enqueued skbs attached to a notifier and doesn't need the specified in them user data, the flushed notifier will post a CQE. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3ecf54c31a85762bf679b0a432c9f43ecf7e61cc.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:38 +08:00
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/file.h>
#include <linux/slab.h>
#include <linux/net.h>
#include <linux/io_uring.h>
#include "io_uring.h"
#include "notif.h"
io_uring: add rsrc referencing for notifiers In preparation to zerocopy sends with fixed buffers make notifiers to reference the rsrc node to protect the used fixed buffers. We can't just grab it for a send request as notifiers can likely outlive requests that used it. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3cd7a01d26837945b6982fa9cf15a63230f2ed4f.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:41 +08:00
#include "rsrc.h"
io_uring: add zc notification infrastructure Add internal part of send zerocopy notifications. There are two main structures, the first one is struct io_notif, which carries inside struct ubuf_info and maps 1:1 to it. io_uring will be binding a number of zerocopy send requests to it and ask to complete (aka flush) it. When flushed and all attached requests and skbs complete, it'll generate one and only one CQE. There are intended to be passed into the network layer as struct msghdr::msg_ubuf. The second concept is notification slots. The userspace will be able to register an array of slots and subsequently addressing them by the index in the array. Slots are independent of each other. Each slot can have only one notifier at a time (called active notifier) but many notifiers during the lifetime. When active, a notifier not going to post any completion but the userspace can attach requests to it by specifying the corresponding slot while issueing send zc requests. Eventually, the userspace will want to "flush" the notifier losing any way to attach new requests to it, however it can use the next atomatically added notifier of this slot or of any other slot. When the network layer is done with all enqueued skbs attached to a notifier and doesn't need the specified in them user data, the flushed notifier will post a CQE. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3ecf54c31a85762bf679b0a432c9f43ecf7e61cc.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:38 +08:00
io_uring/notif: implement notification stacking The network stack allows only one ubuf_info per skb, and unlike MSG_ZEROCOPY, each io_uring zerocopy send will carry a separate ubuf_info. That means that send requests can't reuse a previosly allocated skb and need to get one more or more of new ones. That's fine for large sends, but otherwise it would spam the stack with lots of skbs carrying just a little data each. To help with that implement linking notification (i.e. an io_uring wrapper around ubuf_info) into a list. Each is refcounted by skbs and the stack as usual. additionally all non head entries keep a reference to the head, which they put down when their refcount hits 0. When the head have no more users, it'll efficiently put all notifications in a batch. As mentioned previously about ->io_link_skb, the callback implementation always allows to bind to an skb without a ubuf_info. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/bf1e7f9b72f9ecc99999fdc0d2cded5eea87fd0b.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:42 +08:00
static const struct ubuf_info_ops io_ubuf_ops;
io_uring/notif: simplify io_notif_flush() io_notif_flush() is partially duplicating io_tx_ubuf_complete(), so instead of duplicating it, make the flush call io_tx_ubuf_complete. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/19e41652c16718b946a5c80d2ad409df7682e47e.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:41 +08:00
static void io_notif_tw_complete(struct io_kiocb *notif, struct io_tw_state *ts)
io_uring: add zc notification infrastructure Add internal part of send zerocopy notifications. There are two main structures, the first one is struct io_notif, which carries inside struct ubuf_info and maps 1:1 to it. io_uring will be binding a number of zerocopy send requests to it and ask to complete (aka flush) it. When flushed and all attached requests and skbs complete, it'll generate one and only one CQE. There are intended to be passed into the network layer as struct msghdr::msg_ubuf. The second concept is notification slots. The userspace will be able to register an array of slots and subsequently addressing them by the index in the array. Slots are independent of each other. Each slot can have only one notifier at a time (called active notifier) but many notifiers during the lifetime. When active, a notifier not going to post any completion but the userspace can attach requests to it by specifying the corresponding slot while issueing send zc requests. Eventually, the userspace will want to "flush" the notifier losing any way to attach new requests to it, however it can use the next atomatically added notifier of this slot or of any other slot. When the network layer is done with all enqueued skbs attached to a notifier and doesn't need the specified in them user data, the flushed notifier will post a CQE. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3ecf54c31a85762bf679b0a432c9f43ecf7e61cc.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:38 +08:00
{
io_uring: notification completion optimisation We want to use all optimisations that we have for io_uring requests like completion batching, memory caching and more but for zc notifications. Fortunately, notification perfectly fit the request model so we can overlay them onto struct io_kiocb and use all the infratructure. Most of the fields of struct io_notif natively fits into io_kiocb, so we replace struct io_notif with struct io_kiocb carrying struct io_notif_data in the cmd cache line. Then we adapt io_alloc_notif() to use io_alloc_req()/io_alloc_req_refill(), and kill leftovers of hand coded caching. __io_notif_complete_tw() is converted to use io_uring's tw infra. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/9e010125175e80baf51f0ca63bdc7cc6a4a9fa56.1658913593.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-27 17:30:41 +08:00
struct io_notif_data *nd = io_notif_to_data(notif);
io_uring: add zc notification infrastructure Add internal part of send zerocopy notifications. There are two main structures, the first one is struct io_notif, which carries inside struct ubuf_info and maps 1:1 to it. io_uring will be binding a number of zerocopy send requests to it and ask to complete (aka flush) it. When flushed and all attached requests and skbs complete, it'll generate one and only one CQE. There are intended to be passed into the network layer as struct msghdr::msg_ubuf. The second concept is notification slots. The userspace will be able to register an array of slots and subsequently addressing them by the index in the array. Slots are independent of each other. Each slot can have only one notifier at a time (called active notifier) but many notifiers during the lifetime. When active, a notifier not going to post any completion but the userspace can attach requests to it by specifying the corresponding slot while issueing send zc requests. Eventually, the userspace will want to "flush" the notifier losing any way to attach new requests to it, however it can use the next atomatically added notifier of this slot or of any other slot. When the network layer is done with all enqueued skbs attached to a notifier and doesn't need the specified in them user data, the flushed notifier will post a CQE. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3ecf54c31a85762bf679b0a432c9f43ecf7e61cc.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:38 +08:00
io_uring/notif: implement notification stacking The network stack allows only one ubuf_info per skb, and unlike MSG_ZEROCOPY, each io_uring zerocopy send will carry a separate ubuf_info. That means that send requests can't reuse a previosly allocated skb and need to get one more or more of new ones. That's fine for large sends, but otherwise it would spam the stack with lots of skbs carrying just a little data each. To help with that implement linking notification (i.e. an io_uring wrapper around ubuf_info) into a list. Each is refcounted by skbs and the stack as usual. additionally all non head entries keep a reference to the head, which they put down when their refcount hits 0. When the head have no more users, it'll efficiently put all notifications in a batch. As mentioned previously about ->io_link_skb, the callback implementation always allows to bind to an skb without a ubuf_info. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/bf1e7f9b72f9ecc99999fdc0d2cded5eea87fd0b.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:42 +08:00
do {
notif = cmd_to_io_kiocb(nd);
io_uring/net: move mm accounting to a slower path We can also move mm accounting to the extended callbacks. It removes a few cycles from the hot path including skipping one function call and setting io_req_task_complete as a callback directly. For user backed I/O it shouldn't make any difference taking into considering atomic mm accounting and page pinning. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/1062f270273ad11c1b7b45ec59a6a317533d5e64.1667557923.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-11-04 18:59:46 +08:00
io_uring/notif: implement notification stacking The network stack allows only one ubuf_info per skb, and unlike MSG_ZEROCOPY, each io_uring zerocopy send will carry a separate ubuf_info. That means that send requests can't reuse a previosly allocated skb and need to get one more or more of new ones. That's fine for large sends, but otherwise it would spam the stack with lots of skbs carrying just a little data each. To help with that implement linking notification (i.e. an io_uring wrapper around ubuf_info) into a list. Each is refcounted by skbs and the stack as usual. additionally all non head entries keep a reference to the head, which they put down when their refcount hits 0. When the head have no more users, it'll efficiently put all notifications in a batch. As mentioned previously about ->io_link_skb, the callback implementation always allows to bind to an skb without a ubuf_info. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/bf1e7f9b72f9ecc99999fdc0d2cded5eea87fd0b.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:42 +08:00
lockdep_assert(refcount_read(&nd->uarg.refcnt) == 0);
if (unlikely(nd->zc_report) && (nd->zc_copied || !nd->zc_used))
notif->cqe.res |= IORING_NOTIF_USAGE_ZC_COPIED;
if (nd->account_pages && notif->ctx->user) {
__io_unaccount_mem(notif->ctx->user, nd->account_pages);
nd->account_pages = 0;
}
nd = nd->next;
io_req_task_complete(notif, ts);
} while (nd);
io_uring: move zc reporting from the hot path Add custom tw and notif callbacks on top of usual bits also handling zc reporting. That moves it from the hot path. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/40de4a6409042478e1f35adc4912e23226cb1b5c.1667557923.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-11-04 18:59:45 +08:00
}
io_uring/notif: simplify io_notif_flush() io_notif_flush() is partially duplicating io_tx_ubuf_complete(), so instead of duplicating it, make the flush call io_tx_ubuf_complete. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/19e41652c16718b946a5c80d2ad409df7682e47e.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:41 +08:00
void io_tx_ubuf_complete(struct sk_buff *skb, struct ubuf_info *uarg,
bool success)
io_uring: add zc notification infrastructure Add internal part of send zerocopy notifications. There are two main structures, the first one is struct io_notif, which carries inside struct ubuf_info and maps 1:1 to it. io_uring will be binding a number of zerocopy send requests to it and ask to complete (aka flush) it. When flushed and all attached requests and skbs complete, it'll generate one and only one CQE. There are intended to be passed into the network layer as struct msghdr::msg_ubuf. The second concept is notification slots. The userspace will be able to register an array of slots and subsequently addressing them by the index in the array. Slots are independent of each other. Each slot can have only one notifier at a time (called active notifier) but many notifiers during the lifetime. When active, a notifier not going to post any completion but the userspace can attach requests to it by specifying the corresponding slot while issueing send zc requests. Eventually, the userspace will want to "flush" the notifier losing any way to attach new requests to it, however it can use the next atomatically added notifier of this slot or of any other slot. When the network layer is done with all enqueued skbs attached to a notifier and doesn't need the specified in them user data, the flushed notifier will post a CQE. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3ecf54c31a85762bf679b0a432c9f43ecf7e61cc.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:38 +08:00
{
io_uring: notification completion optimisation We want to use all optimisations that we have for io_uring requests like completion batching, memory caching and more but for zc notifications. Fortunately, notification perfectly fit the request model so we can overlay them onto struct io_kiocb and use all the infratructure. Most of the fields of struct io_notif natively fits into io_kiocb, so we replace struct io_notif with struct io_kiocb carrying struct io_notif_data in the cmd cache line. Then we adapt io_alloc_notif() to use io_alloc_req()/io_alloc_req_refill(), and kill leftovers of hand coded caching. __io_notif_complete_tw() is converted to use io_uring's tw infra. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/9e010125175e80baf51f0ca63bdc7cc6a4a9fa56.1658913593.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-27 17:30:41 +08:00
struct io_notif_data *nd = container_of(uarg, struct io_notif_data, uarg);
struct io_kiocb *notif = cmd_to_io_kiocb(nd);
io_uring/notif: disable LAZY_WAKE for linked notifs Notifications may now be linked and thus a single tw can post multiple CQEs, it's not safe to use LAZY_WAKE with them. Disable LAZY_WAKE for now, if that'd prove to be a problem we can count them and pass the expected number of CQEs into __io_req_task_work_add(). Fixes: 6fe4220912d19 ("io_uring/notif: implement notification stacking") Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/0a5accdb7d2d0d27ebec14f8106e14e0192fae17.1714488419.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-30 23:42:31 +08:00
unsigned tw_flags;
io_uring: cache struct io_notif kmalloc'ing struct io_notif is too expensive when done frequently, cache them as many other resources in io_uring. Keep two list, the first one is from where we're getting notifiers, it's protected by ->uring_lock. The second is protected by ->completion_lock, to which we queue released notifiers. Then we splice one list into another when needed. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/9dec18f7fcbab9f4bd40b96e5ae158b119945230.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:39 +08:00
io_uring/net: introduce IORING_SEND_ZC_REPORT_USAGE flag It might be useful for applications to detect if a zero copy transfer with SEND[MSG]_ZC was actually possible or not. The application can fallback to plain SEND[MSG] in order to avoid the overhead of two cqes per request. Or it can generate a log message that could indicate to an administrator that no zero copy was possible and could explain degraded performance. Cc: stable@vger.kernel.org # 6.1 Link: https://lore.kernel.org/io-uring/fb6a7599-8a9b-15e5-9b64-6cd9d01c6ff4@gmail.com/T/#m2b0d9df94ce43b0e69e6c089bdff0ce6babbdfaa Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/8945b01756d902f5d5b0667f20b957ad3f742e5e.1666895626.git.metze@samba.org Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-10-28 02:34:45 +08:00
if (nd->zc_report) {
if (success && !nd->zc_used && skb)
WRITE_ONCE(nd->zc_used, true);
else if (!success && !nd->zc_copied)
WRITE_ONCE(nd->zc_copied, true);
}
io_uring/net: merge ubuf sendzc callbacks Splitting io_tx_ubuf_callback_ext from io_tx_ubuf_callback is a pre mature optimisation that doesn't give us much. Merge the functions into one and reclaim some simplicity back. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/d44d68f6f7add33a0dcf0b7fd7b73c2dc543604f.1712534031.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 07:54:55 +08:00
io_uring/notif: refactor io_tx_ubuf_complete() Flip the dec_and_test "if", that makes the function extension easier in the future. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/43939e2b04dff03bff5d7227c98afedf951227b3.1713185320.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-15 20:50:11 +08:00
if (!refcount_dec_and_test(&uarg->refcnt))
return;
io_uring/notif: implement notification stacking The network stack allows only one ubuf_info per skb, and unlike MSG_ZEROCOPY, each io_uring zerocopy send will carry a separate ubuf_info. That means that send requests can't reuse a previosly allocated skb and need to get one more or more of new ones. That's fine for large sends, but otherwise it would spam the stack with lots of skbs carrying just a little data each. To help with that implement linking notification (i.e. an io_uring wrapper around ubuf_info) into a list. Each is refcounted by skbs and the stack as usual. additionally all non head entries keep a reference to the head, which they put down when their refcount hits 0. When the head have no more users, it'll efficiently put all notifications in a batch. As mentioned previously about ->io_link_skb, the callback implementation always allows to bind to an skb without a ubuf_info. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/bf1e7f9b72f9ecc99999fdc0d2cded5eea87fd0b.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:42 +08:00
if (nd->head != nd) {
io_tx_ubuf_complete(skb, &nd->head->uarg, success);
return;
}
io_uring/notif: disable LAZY_WAKE for linked notifs Notifications may now be linked and thus a single tw can post multiple CQEs, it's not safe to use LAZY_WAKE with them. Disable LAZY_WAKE for now, if that'd prove to be a problem we can count them and pass the expected number of CQEs into __io_req_task_work_add(). Fixes: 6fe4220912d19 ("io_uring/notif: implement notification stacking") Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/0a5accdb7d2d0d27ebec14f8106e14e0192fae17.1714488419.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-30 23:42:31 +08:00
tw_flags = nd->next ? 0 : IOU_F_TWQ_LAZY_WAKE;
io_uring/notif: refactor io_tx_ubuf_complete() Flip the dec_and_test "if", that makes the function extension easier in the future. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/43939e2b04dff03bff5d7227c98afedf951227b3.1713185320.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-15 20:50:11 +08:00
notif->io_task_work.func = io_notif_tw_complete;
io_uring/notif: disable LAZY_WAKE for linked notifs Notifications may now be linked and thus a single tw can post multiple CQEs, it's not safe to use LAZY_WAKE with them. Disable LAZY_WAKE for now, if that'd prove to be a problem we can count them and pass the expected number of CQEs into __io_req_task_work_add(). Fixes: 6fe4220912d19 ("io_uring/notif: implement notification stacking") Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/0a5accdb7d2d0d27ebec14f8106e14e0192fae17.1714488419.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-30 23:42:31 +08:00
__io_req_task_work_add(notif, tw_flags);
io_uring: cache struct io_notif kmalloc'ing struct io_notif is too expensive when done frequently, cache them as many other resources in io_uring. Keep two list, the first one is from where we're getting notifiers, it's protected by ->uring_lock. The second is protected by ->completion_lock, to which we queue released notifiers. Then we splice one list into another when needed. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/9dec18f7fcbab9f4bd40b96e5ae158b119945230.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:39 +08:00
}
io_uring/notif: implement notification stacking The network stack allows only one ubuf_info per skb, and unlike MSG_ZEROCOPY, each io_uring zerocopy send will carry a separate ubuf_info. That means that send requests can't reuse a previosly allocated skb and need to get one more or more of new ones. That's fine for large sends, but otherwise it would spam the stack with lots of skbs carrying just a little data each. To help with that implement linking notification (i.e. an io_uring wrapper around ubuf_info) into a list. Each is refcounted by skbs and the stack as usual. additionally all non head entries keep a reference to the head, which they put down when their refcount hits 0. When the head have no more users, it'll efficiently put all notifications in a batch. As mentioned previously about ->io_link_skb, the callback implementation always allows to bind to an skb without a ubuf_info. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/bf1e7f9b72f9ecc99999fdc0d2cded5eea87fd0b.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:42 +08:00
static int io_link_skb(struct sk_buff *skb, struct ubuf_info *uarg)
{
struct io_notif_data *nd, *prev_nd;
struct io_kiocb *prev_notif, *notif;
struct ubuf_info *prev_uarg = skb_zcopy(skb);
nd = container_of(uarg, struct io_notif_data, uarg);
notif = cmd_to_io_kiocb(nd);
if (!prev_uarg) {
net_zcopy_get(&nd->uarg);
skb_zcopy_init(skb, &nd->uarg);
return 0;
}
/* handle it separately as we can't link a notif to itself */
if (unlikely(prev_uarg == &nd->uarg))
return 0;
/* we can't join two links together, just request a fresh skb */
if (unlikely(nd->head != nd || nd->next))
return -EEXIST;
/* don't mix zc providers */
if (unlikely(prev_uarg->ops != &io_ubuf_ops))
return -EEXIST;
prev_nd = container_of(prev_uarg, struct io_notif_data, uarg);
prev_notif = cmd_to_io_kiocb(nd);
/* make sure all noifications can be finished in the same task_work */
if (unlikely(notif->ctx != prev_notif->ctx ||
notif->task != prev_notif->task))
return -EEXIST;
nd->head = prev_nd->head;
nd->next = prev_nd->next;
prev_nd->next = nd;
net_zcopy_get(&nd->head->uarg);
return 0;
}
net: extend ubuf_info callback to ops structure We'll need to associate additional callbacks with ubuf_info, introduce a structure holding ubuf_info callbacks. Apart from a more smarter io_uring notification management introduced in next patches, it can be used to generalise msg_zerocopy_put_abort() and also store ->sg_from_iter, which is currently passed in struct msghdr. Reviewed-by: Jens Axboe <axboe@kernel.dk> Reviewed-by: David Ahern <dsahern@kernel.org> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Reviewed-by: Willem de Bruijn <willemb@google.com> Link: https://lore.kernel.org/all/a62015541de49c0e2a8a0377a1d5d0a5aeb07016.1713369317.git.asml.silence@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2024-04-19 19:08:39 +08:00
static const struct ubuf_info_ops io_ubuf_ops = {
.complete = io_tx_ubuf_complete,
io_uring/notif: implement notification stacking The network stack allows only one ubuf_info per skb, and unlike MSG_ZEROCOPY, each io_uring zerocopy send will carry a separate ubuf_info. That means that send requests can't reuse a previosly allocated skb and need to get one more or more of new ones. That's fine for large sends, but otherwise it would spam the stack with lots of skbs carrying just a little data each. To help with that implement linking notification (i.e. an io_uring wrapper around ubuf_info) into a list. Each is refcounted by skbs and the stack as usual. additionally all non head entries keep a reference to the head, which they put down when their refcount hits 0. When the head have no more users, it'll efficiently put all notifications in a batch. As mentioned previously about ->io_link_skb, the callback implementation always allows to bind to an skb without a ubuf_info. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/bf1e7f9b72f9ecc99999fdc0d2cded5eea87fd0b.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:42 +08:00
.link_skb = io_link_skb,
net: extend ubuf_info callback to ops structure We'll need to associate additional callbacks with ubuf_info, introduce a structure holding ubuf_info callbacks. Apart from a more smarter io_uring notification management introduced in next patches, it can be used to generalise msg_zerocopy_put_abort() and also store ->sg_from_iter, which is currently passed in struct msghdr. Reviewed-by: Jens Axboe <axboe@kernel.dk> Reviewed-by: David Ahern <dsahern@kernel.org> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Reviewed-by: Willem de Bruijn <willemb@google.com> Link: https://lore.kernel.org/all/a62015541de49c0e2a8a0377a1d5d0a5aeb07016.1713369317.git.asml.silence@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2024-04-19 19:08:39 +08:00
};
io_uring/net: simplify zerocopy send user API Following user feedback, this patch simplifies zerocopy send API. One of the main complaints is that the current API is difficult with the userspace managing notification slots, and then send retries with error handling make it even worse. Instead of keeping notification slots change it to the per-request notifications model, which posts both completion and notification CQEs for each request when any data has been sent, and only one CQE if it fails. All notification CQEs will have IORING_CQE_F_NOTIF set and IORING_CQE_F_MORE in completion CQEs indicates whether to wait a notification or not. IOSQE_CQE_SKIP_SUCCESS is disallowed with zerocopy sends for now. This is less flexible, but greatly simplifies the user API and also the kernel implementation. We reuse notif helpers in this patch, but in the future there won't be need for keeping two requests. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/95287640ab98fc9417370afb16e310677c63e6ce.1662027856.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-09-01 18:54:04 +08:00
struct io_kiocb *io_alloc_notif(struct io_ring_ctx *ctx)
io_uring: add zc notification infrastructure Add internal part of send zerocopy notifications. There are two main structures, the first one is struct io_notif, which carries inside struct ubuf_info and maps 1:1 to it. io_uring will be binding a number of zerocopy send requests to it and ask to complete (aka flush) it. When flushed and all attached requests and skbs complete, it'll generate one and only one CQE. There are intended to be passed into the network layer as struct msghdr::msg_ubuf. The second concept is notification slots. The userspace will be able to register an array of slots and subsequently addressing them by the index in the array. Slots are independent of each other. Each slot can have only one notifier at a time (called active notifier) but many notifiers during the lifetime. When active, a notifier not going to post any completion but the userspace can attach requests to it by specifying the corresponding slot while issueing send zc requests. Eventually, the userspace will want to "flush" the notifier losing any way to attach new requests to it, however it can use the next atomatically added notifier of this slot or of any other slot. When the network layer is done with all enqueued skbs attached to a notifier and doesn't need the specified in them user data, the flushed notifier will post a CQE. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3ecf54c31a85762bf679b0a432c9f43ecf7e61cc.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:38 +08:00
__must_hold(&ctx->uring_lock)
{
io_uring: notification completion optimisation We want to use all optimisations that we have for io_uring requests like completion batching, memory caching and more but for zc notifications. Fortunately, notification perfectly fit the request model so we can overlay them onto struct io_kiocb and use all the infratructure. Most of the fields of struct io_notif natively fits into io_kiocb, so we replace struct io_notif with struct io_kiocb carrying struct io_notif_data in the cmd cache line. Then we adapt io_alloc_notif() to use io_alloc_req()/io_alloc_req_refill(), and kill leftovers of hand coded caching. __io_notif_complete_tw() is converted to use io_uring's tw infra. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/9e010125175e80baf51f0ca63bdc7cc6a4a9fa56.1658913593.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-27 17:30:41 +08:00
struct io_kiocb *notif;
struct io_notif_data *nd;
io_uring: refactor req allocation Follow the io_get_sqe pattern returning the result via a pointer and hide request cache refill inside io_alloc_req(). Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/8c37c2e8a3cb5e4cd6a8ae3b91371227a92708a6.1674484266.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2023-01-23 22:37:16 +08:00
if (unlikely(!io_alloc_req(ctx, &notif)))
io_uring: notification completion optimisation We want to use all optimisations that we have for io_uring requests like completion batching, memory caching and more but for zc notifications. Fortunately, notification perfectly fit the request model so we can overlay them onto struct io_kiocb and use all the infratructure. Most of the fields of struct io_notif natively fits into io_kiocb, so we replace struct io_notif with struct io_kiocb carrying struct io_notif_data in the cmd cache line. Then we adapt io_alloc_notif() to use io_alloc_req()/io_alloc_req_refill(), and kill leftovers of hand coded caching. __io_notif_complete_tw() is converted to use io_uring's tw infra. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/9e010125175e80baf51f0ca63bdc7cc6a4a9fa56.1658913593.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-27 17:30:41 +08:00
return NULL;
notif->opcode = IORING_OP_NOP;
notif->flags = 0;
notif->file = NULL;
notif->task = current;
io_get_task_refs(1);
notif->rsrc_node = NULL;
nd = io_notif_to_data(notif);
io_uring/net: merge ubuf sendzc callbacks Splitting io_tx_ubuf_callback_ext from io_tx_ubuf_callback is a pre mature optimisation that doesn't give us much. Merge the functions into one and reclaim some simplicity back. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/d44d68f6f7add33a0dcf0b7fd7b73c2dc543604f.1712534031.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-08 07:54:55 +08:00
nd->zc_report = false;
nd->account_pages = 0;
io_uring/notif: implement notification stacking The network stack allows only one ubuf_info per skb, and unlike MSG_ZEROCOPY, each io_uring zerocopy send will carry a separate ubuf_info. That means that send requests can't reuse a previosly allocated skb and need to get one more or more of new ones. That's fine for large sends, but otherwise it would spam the stack with lots of skbs carrying just a little data each. To help with that implement linking notification (i.e. an io_uring wrapper around ubuf_info) into a list. Each is refcounted by skbs and the stack as usual. additionally all non head entries keep a reference to the head, which they put down when their refcount hits 0. When the head have no more users, it'll efficiently put all notifications in a batch. As mentioned previously about ->io_link_skb, the callback implementation always allows to bind to an skb without a ubuf_info. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/bf1e7f9b72f9ecc99999fdc0d2cded5eea87fd0b.1713369317.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2024-04-19 19:08:42 +08:00
nd->next = NULL;
nd->head = nd;
io_uring/notif: add constant for ubuf_info flags Add a constant IO_NOTIF_UBUF_FLAGS for struct ubuf_info flags that notifications use. That should minimise merge conflicts for planned changes touching both io_uring and net at the same time. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2023-04-15 21:20:08 +08:00
nd->uarg.flags = IO_NOTIF_UBUF_FLAGS;
net: extend ubuf_info callback to ops structure We'll need to associate additional callbacks with ubuf_info, introduce a structure holding ubuf_info callbacks. Apart from a more smarter io_uring notification management introduced in next patches, it can be used to generalise msg_zerocopy_put_abort() and also store ->sg_from_iter, which is currently passed in struct msghdr. Reviewed-by: Jens Axboe <axboe@kernel.dk> Reviewed-by: David Ahern <dsahern@kernel.org> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Reviewed-by: Willem de Bruijn <willemb@google.com> Link: https://lore.kernel.org/all/a62015541de49c0e2a8a0377a1d5d0a5aeb07016.1713369317.git.asml.silence@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2024-04-19 19:08:39 +08:00
nd->uarg.ops = &io_ubuf_ops;
io_uring: notification completion optimisation We want to use all optimisations that we have for io_uring requests like completion batching, memory caching and more but for zc notifications. Fortunately, notification perfectly fit the request model so we can overlay them onto struct io_kiocb and use all the infratructure. Most of the fields of struct io_notif natively fits into io_kiocb, so we replace struct io_notif with struct io_kiocb carrying struct io_notif_data in the cmd cache line. Then we adapt io_alloc_notif() to use io_alloc_req()/io_alloc_req_refill(), and kill leftovers of hand coded caching. __io_notif_complete_tw() is converted to use io_uring's tw infra. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/9e010125175e80baf51f0ca63bdc7cc6a4a9fa56.1658913593.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-27 17:30:41 +08:00
refcount_set(&nd->uarg.refcnt, 1);
io_uring: add zc notification infrastructure Add internal part of send zerocopy notifications. There are two main structures, the first one is struct io_notif, which carries inside struct ubuf_info and maps 1:1 to it. io_uring will be binding a number of zerocopy send requests to it and ask to complete (aka flush) it. When flushed and all attached requests and skbs complete, it'll generate one and only one CQE. There are intended to be passed into the network layer as struct msghdr::msg_ubuf. The second concept is notification slots. The userspace will be able to register an array of slots and subsequently addressing them by the index in the array. Slots are independent of each other. Each slot can have only one notifier at a time (called active notifier) but many notifiers during the lifetime. When active, a notifier not going to post any completion but the userspace can attach requests to it by specifying the corresponding slot while issueing send zc requests. Eventually, the userspace will want to "flush" the notifier losing any way to attach new requests to it, however it can use the next atomatically added notifier of this slot or of any other slot. When the network layer is done with all enqueued skbs attached to a notifier and doesn't need the specified in them user data, the flushed notifier will post a CQE. Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/3ecf54c31a85762bf679b0a432c9f43ecf7e61cc.1657643355.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-07-13 04:52:38 +08:00
return notif;
}
Reference in New Issue Copy Permalink