[PATCH net-next v3 00/25] io_uring zerocopy send

[PATCH net-next v3 00/25] io_uring zerocopy send

[Pavel Begunkov]

The patchset implements io_uring zerocopy send. It works with both registered
and normal buffers, mixing is allowed but not recommended. Apart from usual
request completions, just as with MSG_ZEROCOPY, io_uring separately notifies
the userspace when buffers are freed and can be reused (see API design below),
which is delivered into io_uring's Completion Queue. Those "buffer-free"
notifications are not necessarily per request, but the userspace has control
over it and should explicitly attaching a number of requests to a single
notification. The series also adds some internal optimisations when used with
registered buffers like removing page referencing.

From the kernel networking perspective there are two main changes. The first
one is passing ubuf_info into the network layer from io_uring (inside of an
in kernel struct msghdr). This allows extra optimisations, e.g. ubuf_info
caching on the io_uring side, but also helps to avoid cross-referencing
and synchronisation problems. The second part is an optional optimisation
removing page referencing for requests with registered buffers.

Benchmarking with an optimised version of the selftest (see [1]), which in a
loop sends a bunch of requests and then waits for their completions. "+ flush"
column posts one additional "buffer-free" notification per request, and
just "zc" doesn't post buffer notifications at all.

NIC (requests / second):
IO size | non-zc    | zc             | zc + flush
4000    | 495134    | 606420 (+22%)  | 558971 (+12%)
1500    | 551808    | 577116 (+4.5%) | 565803 (+2.5%)
1000    | 584677    | 592088 (+1.2%) | 560885 (-4%)
600     | 596292    | 598550 (+0.4%) | 555366 (-6.7%)

dummy (requests / second):
IO size | non-zc    | zc             | zc + flush
8000    | 1299916   | 2396600 (+84%) | 2224219 (+71%)
4000    | 1869230   | 2344146 (+25%) | 2170069 (+16%)
1200    | 2071617   | 2361960 (+14%) | 2203052 (+6%)
600     | 2106794   | 2381527 (+13%) | 2195295 (+4%)

Previously it also brought a massive performance speedup compared to the
msg_zerocopy tool (see [3]), which is probably not super interesting.

There is an additional bunch of refcounting optimisations that was omitted from
the series for simplicity and as they don't change the picture drastically,
they will be sent as follow up, as well as flushing optimisations closing the
performance gap b/w two last columns.

Note: the series is based on net-next + for-5.20/io_uring, but as vanilla
net-next fails for me the repo (see [2]) is on top of for-5.20/io_uring.

Links:

  liburing (benchmark + some tests):
  [1] https://github.com/isilence/liburing/tree/zc_v3

  kernel repo:
  [2] https://github.com/isilence/linux/tree/zc_v3

  RFC v1:
  [3] https://lore.kernel.org/io-uring/cover.1638282789.git.asml.silence@gmail.com/

  RFC v2:
  https://lore.kernel.org/io-uring/cover.1640029579.git.asml.silence@gmail.com/

API design overview:

  The series introduces an io_uring concept of notifactors. From the userspace
  perspective it's an entity to which it can bind one or more requests and then
  requesting to flush it. Flushing a notifier makes it impossible to attach new
  requests to it, and instructs the notifier to post a completion once all
  requests attached to it are completed and the kernel doesn't need the buffers
  anymore.

  Notifications are stored in notification slots, which should be registered as
  an array in io_uring. Each slot stores only one notifier at any particular
  moment. Flushing removes it from the slot and the slot automatically replaces
  it with a new notifier. All operations with notifiers are done by specifying
  an index of a slot it's currently in.

  When registering a notification the userspace specifies a u64 tag for each
  slot, which will be copied in notification completion entries as
  cqe::user_data. cqe::res is 0 and cqe::flags is equal to wrap around u32
  sequence number counting notifiers of a slot.

Changelog:

  RFC v2 -> v3:
    mem accounting for non-registered buffers
    allow mixing registered and normal requests per notifier
    notification flushing via IORING_OP_RSRC_UPDATE
    TCP support
    fix buffer indexing
    fix io-wq ->uring_lock locking
    fix bugs when mixing with MSG_ZEROCOPY
    fix managed refs bugs in skbuff.c

  RFC -> RFC v2:
    remove additional overhead for non-zc from skb_release_data()
    avoid msg propagation, hide extra bits of non-zc overhead
    task_work based "buffer free" notifications
    improve io_uring's notification refcounting
    added 5/19, (no pfmemalloc tracking)
    added 8/19 and 9/19 preventing small copies with zc
    misc small changes

Pavel Begunkov (25):
  ipv4: avoid partial copy for zc
  ipv6: avoid partial copy for zc
  skbuff: add SKBFL_DONT_ORPHAN flag
  skbuff: carry external ubuf_info in msghdr
  net: bvec specific path in zerocopy_sg_from_iter
  net: optimise bvec-based zc page referencing
  net: don't track pfmemalloc for managed frags
  skbuff: don't mix ubuf_info of different types
  ipv4/udp: support zc with managed data
  ipv6/udp: support zc with managed data
  tcp: support zc with managed data
  io_uring: add zc notification infrastructure
  io_uring: export task put
  io_uring: cache struct io_notif
  io_uring: complete notifiers in tw
  io_uring: add notification slot registration
  io_uring: wire send zc request type
  io_uring: account locked pages for non-fixed zc
  io_uring: allow to pass addr into sendzc
  io_uring: add rsrc referencing for notifiers
  io_uring: sendzc with fixed buffers
  io_uring: flush notifiers after sendzc
  io_uring: rename IORING_OP_FILES_UPDATE
  io_uring: add zc notification flush requests
  selftests/io_uring: test zerocopy send

 include/linux/io_uring_types.h                |  37 ++
 include/linux/skbuff.h                        |  59 +-
 include/linux/socket.h                        |   7 +
 include/uapi/linux/io_uring.h                 |  43 +-
 io_uring/Makefile                             |   2 +-
 io_uring/io_uring.c                           |  40 +-
 io_uring/io_uring.h                           |  21 +
 io_uring/net.c                                | 134 ++++
 io_uring/net.h                                |   4 +
 io_uring/notif.c                              | 215 +++++++
 io_uring/notif.h                              |  87 +++
 io_uring/opdef.c                              |  24 +-
 io_uring/rsrc.c                               |  55 +-
 io_uring/rsrc.h                               |  16 +-
 io_uring/tctx.h                               |  26 -
 net/compat.c                                  |   2 +
 net/core/datagram.c                           |  53 +-
 net/core/skbuff.c                             |  35 +-
 net/ipv4/ip_output.c                          |  63 +-
 net/ipv4/tcp.c                                |  52 +-
 net/ipv6/ip6_output.c                         |  62 +-
 net/socket.c                                  |   6 +
 tools/testing/selftests/net/Makefile          |   1 +
 .../selftests/net/io_uring_zerocopy_tx.c      | 605 ++++++++++++++++++
 .../selftests/net/io_uring_zerocopy_tx.sh     | 131 ++++
 25 files changed, 1652 insertions(+), 128 deletions(-)
 create mode 100644 io_uring/notif.c
 create mode 100644 io_uring/notif.h
 create mode 100644 tools/testing/selftests/net/io_uring_zerocopy_tx.c
 create mode 100755 tools/testing/selftests/net/io_uring_zerocopy_tx.sh

-- 
2.36.1

[PATCH net-next v3 01/25] ipv4: avoid partial copy for zc

[Pavel Begunkov]

Even when zerocopy transmission is requested and possible,
__ip_append_data() will still copy a small chunk of data just because it
allocated some extra linear space (e.g. 148 bytes). It wastes CPU cycles
on copy and iter manipulations and also misalignes potentially aligned
data. Avoid such coies. And as a bonus we can allocate smaller skb.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 net/ipv4/ip_output.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 00b4bf26fd93..581d1e233260 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -969,7 +969,6 @@ static int __ip_append_data(struct sock *sk,
 	struct inet_sock *inet = inet_sk(sk);
 	struct ubuf_info *uarg = NULL;
 	struct sk_buff *skb;
-
 	struct ip_options *opt = cork->opt;
 	int hh_len;
 	int exthdrlen;
@@ -977,6 +976,7 @@ static int __ip_append_data(struct sock *sk,
 	int copy;
 	int err;
 	int offset = 0;
+	bool zc = false;
 	unsigned int maxfraglen, fragheaderlen, maxnonfragsize;
 	int csummode = CHECKSUM_NONE;
 	struct rtable *rt = (struct rtable *)cork->dst;
@@ -1025,6 +1025,7 @@ static int __ip_append_data(struct sock *sk,
 		if (rt->dst.dev->features & NETIF_F_SG &&
 		    csummode == CHECKSUM_PARTIAL) {
 			paged = true;
+			zc = true;
 		} else {
 			uarg->zerocopy = 0;
 			skb_zcopy_set(skb, uarg, &extra_uref);
@@ -1091,9 +1092,12 @@ static int __ip_append_data(struct sock *sk,
 				 (fraglen + alloc_extra < SKB_MAX_ALLOC ||
 				  !(rt->dst.dev->features & NETIF_F_SG)))
 				alloclen = fraglen;
-			else {
+			else if (!zc) {
 				alloclen = min_t(int, fraglen, MAX_HEADER);
 				pagedlen = fraglen - alloclen;
+			} else {
+				alloclen = fragheaderlen + transhdrlen;
+				pagedlen = datalen - transhdrlen;
 			}
 
 			alloclen += alloc_extra;
-- 
2.36.1

[PATCH net-next v3 02/25] ipv6: avoid partial copy for zc

[Pavel Begunkov]

Even when zerocopy transmission is requested and possible,
__ip_append_data() will still copy a small chunk of data just because it
allocated some extra linear space (e.g. 128 bytes). It wastes CPU cycles
on copy and iter manipulations and also misalignes potentially aligned
data. Avoid such coies. And as a bonus we can allocate smaller skb.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 net/ipv6/ip6_output.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index 77e3f5970ce4..fc74ce3ed8cc 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -1464,6 +1464,7 @@ static int __ip6_append_data(struct sock *sk,
 	int copy;
 	int err;
 	int offset = 0;
+	bool zc = false;
 	u32 tskey = 0;
 	struct rt6_info *rt = (struct rt6_info *)cork->dst;
 	struct ipv6_txoptions *opt = v6_cork->opt;
@@ -1549,6 +1550,7 @@ static int __ip6_append_data(struct sock *sk,
 		if (rt->dst.dev->features & NETIF_F_SG &&
 		    csummode == CHECKSUM_PARTIAL) {
 			paged = true;
+			zc = true;
 		} else {
 			uarg->zerocopy = 0;
 			skb_zcopy_set(skb, uarg, &extra_uref);
@@ -1630,9 +1632,12 @@ static int __ip6_append_data(struct sock *sk,
 				 (fraglen + alloc_extra < SKB_MAX_ALLOC ||
 				  !(rt->dst.dev->features & NETIF_F_SG)))
 				alloclen = fraglen;
-			else {
+			else if (!zc) {
 				alloclen = min_t(int, fraglen, MAX_HEADER);
 				pagedlen = fraglen - alloclen;
+			} else {
+				alloclen = fragheaderlen + transhdrlen;
+				pagedlen = datalen - transhdrlen;
 			}
 			alloclen += alloc_extra;
 
-- 
2.36.1

[PATCH net-next v3 03/25] skbuff: add SKBFL_DONT_ORPHAN flag

[Pavel Begunkov]

We don't want to list every single ubuf_info callback in
skb_orphan_frags(), add a flag controlling the behaviour.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/linux/skbuff.h | 8 +++++---
 net/core/skbuff.c      | 2 +-
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index d3d10556f0fa..8e12b3b9ad6c 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -686,10 +686,13 @@ enum {
 	 * charged to the kernel memory.
 	 */
 	SKBFL_PURE_ZEROCOPY = BIT(2),
+
+	SKBFL_DONT_ORPHAN = BIT(3),
 };
 
 #define SKBFL_ZEROCOPY_FRAG	(SKBFL_ZEROCOPY_ENABLE | SKBFL_SHARED_FRAG)
-#define SKBFL_ALL_ZEROCOPY	(SKBFL_ZEROCOPY_FRAG | SKBFL_PURE_ZEROCOPY)
+#define SKBFL_ALL_ZEROCOPY	(SKBFL_ZEROCOPY_FRAG | SKBFL_PURE_ZEROCOPY | \
+				 SKBFL_DONT_ORPHAN)
 
 /*
  * The callback notifies userspace to release buffers when skb DMA is done in
@@ -3182,8 +3185,7 @@ static inline int skb_orphan_frags(struct sk_buff *skb, gfp_t gfp_mask)
 {
 	if (likely(!skb_zcopy(skb)))
 		return 0;
-	if (!skb_zcopy_is_nouarg(skb) &&
-	    skb_uarg(skb)->callback == msg_zerocopy_callback)
+	if (skb_shinfo(skb)->flags & SKBFL_DONT_ORPHAN)
 		return 0;
 	return skb_copy_ubufs(skb, gfp_mask);
 }
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 5b3559cb1d82..5b35791064d1 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -1193,7 +1193,7 @@ static struct ubuf_info *msg_zerocopy_alloc(struct sock *sk, size_t size)
 	uarg->len = 1;
 	uarg->bytelen = size;
 	uarg->zerocopy = 1;
-	uarg->flags = SKBFL_ZEROCOPY_FRAG;
+	uarg->flags = SKBFL_ZEROCOPY_FRAG | SKBFL_DONT_ORPHAN;
 	refcount_set(&uarg->refcnt, 1);
 	sock_hold(sk);
 
-- 
2.36.1

[PATCH net-next v3 04/25] skbuff: carry external ubuf_info in msghdr

[Pavel Begunkov]

Make possible for network in-kernel callers like io_uring to pass in a
custom ubuf_info by setting it in a new field of struct msghdr.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/linux/socket.h | 7 +++++++
 io_uring/net.c         | 4 ++++
 net/compat.c           | 2 ++
 net/socket.c           | 6 ++++++
 4 files changed, 19 insertions(+)

diff --git a/include/linux/socket.h b/include/linux/socket.h
index 17311ad9f9af..ba84ee614d5a 100644
--- a/include/linux/socket.h
+++ b/include/linux/socket.h
@@ -66,9 +66,16 @@ struct msghdr {
 	};
 	bool		msg_control_is_user : 1;
 	bool		msg_get_inq : 1;/* return INQ after receive */
+	/*
+	 * The data pages are pinned and won't be released before ->msg_ubuf
+	 * is released. ->msg_iter should point to a bvec and ->msg_ubuf has
+	 * to be non-NULL.
+	 */
+	bool		msg_managed_data : 1;
 	unsigned int	msg_flags;	/* flags on received message */
 	__kernel_size_t	msg_controllen;	/* ancillary data buffer length */
 	struct kiocb	*msg_iocb;	/* ptr to iocb for async requests */
+	struct ubuf_info *msg_ubuf;
 };
 
 struct user_msghdr {
diff --git a/io_uring/net.c b/io_uring/net.c
index 19a805c3814c..d95c88d83f9f 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -255,6 +255,8 @@ int io_send(struct io_kiocb *req, unsigned int issue_flags)
 	msg.msg_control = NULL;
 	msg.msg_controllen = 0;
 	msg.msg_namelen = 0;
+	msg.msg_ubuf = NULL;
+	msg.msg_managed_data = false;
 
 	flags = sr->msg_flags;
 	if (issue_flags & IO_URING_F_NONBLOCK)
@@ -525,6 +527,8 @@ int io_recv(struct io_kiocb *req, unsigned int issue_flags)
 	msg.msg_flags = 0;
 	msg.msg_controllen = 0;
 	msg.msg_iocb = NULL;
+	msg.msg_ubuf = NULL;
+	msg.msg_managed_data = false;
 
 	flags = sr->msg_flags;
 	if (force_nonblock)
diff --git a/net/compat.c b/net/compat.c
index 210fc3b4d0d8..435846fa85e0 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -80,6 +80,8 @@ int __get_compat_msghdr(struct msghdr *kmsg,
 		return -EMSGSIZE;
 
 	kmsg->msg_iocb = NULL;
+	kmsg->msg_ubuf = NULL;
+	kmsg->msg_managed_data = false;
 	*ptr = msg.msg_iov;
 	*len = msg.msg_iovlen;
 	return 0;
diff --git a/net/socket.c b/net/socket.c
index 2bc8773d9dc5..0963a02b1472 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -2106,6 +2106,8 @@ int __sys_sendto(int fd, void __user *buff, size_t len, unsigned int flags,
 	msg.msg_control = NULL;
 	msg.msg_controllen = 0;
 	msg.msg_namelen = 0;
+	msg.msg_ubuf = NULL;
+	msg.msg_managed_data = false;
 	if (addr) {
 		err = move_addr_to_kernel(addr, addr_len, &address);
 		if (err < 0)
@@ -2171,6 +2173,8 @@ int __sys_recvfrom(int fd, void __user *ubuf, size_t size, unsigned int flags,
 	msg.msg_namelen = 0;
 	msg.msg_iocb = NULL;
 	msg.msg_flags = 0;
+	msg.msg_ubuf = NULL;
+	msg.msg_managed_data = false;
 	if (sock->file->f_flags & O_NONBLOCK)
 		flags |= MSG_DONTWAIT;
 	err = sock_recvmsg(sock, &msg, flags);
@@ -2409,6 +2413,8 @@ int __copy_msghdr_from_user(struct msghdr *kmsg,
 		return -EMSGSIZE;
 
 	kmsg->msg_iocb = NULL;
+	kmsg->msg_ubuf = NULL;
+	kmsg->msg_managed_data = false;
 	*uiov = msg.msg_iov;
 	*nsegs = msg.msg_iovlen;
 	return 0;
-- 
2.36.1

[PATCH net-next v3 05/25] net: bvec specific path in zerocopy_sg_from_iter

[Pavel Begunkov]

Add an bvec specialised and optimised path in zerocopy_sg_from_iter.
It'll be used later for {get,put}_page() optimisations.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 net/core/datagram.c | 47 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/net/core/datagram.c b/net/core/datagram.c
index 50f4faeea76c..5237cb533bb4 100644
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -613,11 +613,58 @@ int skb_copy_datagram_from_iter(struct sk_buff *skb, int offset,
 }
 EXPORT_SYMBOL(skb_copy_datagram_from_iter);
 
+static int __zerocopy_sg_from_bvec(struct sock *sk, struct sk_buff *skb,
+				   struct iov_iter *from, size_t length)
+{
+	int frag = skb_shinfo(skb)->nr_frags;
+	int ret = 0;
+	struct bvec_iter bi;
+	ssize_t copied = 0;
+	unsigned long truesize = 0;
+
+	bi.bi_size = min(from->count, length);
+	bi.bi_bvec_done = from->iov_offset;
+	bi.bi_idx = 0;
+
+	while (bi.bi_size && frag < MAX_SKB_FRAGS) {
+		struct bio_vec v = mp_bvec_iter_bvec(from->bvec, bi);
+
+		copied += v.bv_len;
+		truesize += PAGE_ALIGN(v.bv_len + v.bv_offset);
+		get_page(v.bv_page);
+		skb_fill_page_desc(skb, frag++, v.bv_page, v.bv_offset, v.bv_len);
+		bvec_iter_advance_single(from->bvec, &bi, v.bv_len);
+	}
+	if (bi.bi_size)
+		ret = -EMSGSIZE;
+
+	from->bvec += bi.bi_idx;
+	from->nr_segs -= bi.bi_idx;
+	from->count = bi.bi_size;
+	from->iov_offset = bi.bi_bvec_done;
+
+	skb->data_len += copied;
+	skb->len += copied;
+	skb->truesize += truesize;
+
+	if (sk && sk->sk_type == SOCK_STREAM) {
+		sk_wmem_queued_add(sk, truesize);
+		if (!skb_zcopy_pure(skb))
+			sk_mem_charge(sk, truesize);
+	} else {
+		refcount_add(truesize, &skb->sk->sk_wmem_alloc);
+	}
+	return ret;
+}
+
 int __zerocopy_sg_from_iter(struct sock *sk, struct sk_buff *skb,
 			    struct iov_iter *from, size_t length)
 {
 	int frag = skb_shinfo(skb)->nr_frags;
 
+	if (iov_iter_is_bvec(from))
+		return __zerocopy_sg_from_bvec(sk, skb, from, length);
+
 	while (length && iov_iter_count(from)) {
 		struct page *pages[MAX_SKB_FRAGS];
 		struct page *last_head = NULL;
-- 
2.36.1

[PATCH net-next v3 06/25] net: optimise bvec-based zc page referencing

[Pavel Begunkov]

Some users like io_uring can pass a bvec iterator to send and also can
implement page pinning more efficiently. Add a ->msg_managed_data toogle
in msghdr. When set, data pages are "managed" by upper layers, i.e.
refcounted and pinned by the caller and will live at least until
->msg_ubuf is released. msghdr has to have non-NULL ->msg_ubuf and
->msg_iter should point to a bvec.

Protocols supporting the feature will propagate it by setting
SKBFL_MANAGED_FRAG_REFS, which means that the skb doesn't hold refs to
its frag pages and only rely on ubuf_info lifetime gurantees. It should
only be used with zerocopy skbs with ubuf_info set.

It's allowed to convert skbs from managed to normal by calling
skb_zcopy_downgrade_managed(). The function will take all needed
page references and clear the flag.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/linux/skbuff.h | 25 +++++++++++++++++++++++--
 net/core/datagram.c    |  7 ++++---
 net/core/skbuff.c      | 29 +++++++++++++++++++++++++++--
 3 files changed, 54 insertions(+), 7 deletions(-)

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 8e12b3b9ad6c..712168c21736 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -688,11 +688,16 @@ enum {
 	SKBFL_PURE_ZEROCOPY = BIT(2),
 
 	SKBFL_DONT_ORPHAN = BIT(3),
+
+	/* page references are managed by the ubuf_info, so it's safe to
+	 * use frags only up until ubuf_info is released
+	 */
+	SKBFL_MANAGED_FRAG_REFS = BIT(4),
 };
 
 #define SKBFL_ZEROCOPY_FRAG	(SKBFL_ZEROCOPY_ENABLE | SKBFL_SHARED_FRAG)
 #define SKBFL_ALL_ZEROCOPY	(SKBFL_ZEROCOPY_FRAG | SKBFL_PURE_ZEROCOPY | \
-				 SKBFL_DONT_ORPHAN)
+				 SKBFL_DONT_ORPHAN | SKBFL_MANAGED_FRAG_REFS)
 
 /*
  * The callback notifies userspace to release buffers when skb DMA is done in
@@ -1809,6 +1814,11 @@ static inline bool skb_zcopy_pure(const struct sk_buff *skb)
 	return skb_shinfo(skb)->flags & SKBFL_PURE_ZEROCOPY;
 }
 
+static inline bool skb_zcopy_managed(const struct sk_buff *skb)
+{
+	return skb_shinfo(skb)->flags & SKBFL_MANAGED_FRAG_REFS;
+}
+
 static inline bool skb_pure_zcopy_same(const struct sk_buff *skb1,
 				       const struct sk_buff *skb2)
 {
@@ -1883,6 +1893,14 @@ static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy_success)
 	}
 }
 
+void __skb_zcopy_downgrade_managed(struct sk_buff *skb);
+
+static inline void skb_zcopy_downgrade_managed(struct sk_buff *skb)
+{
+	if (unlikely(skb_zcopy_managed(skb)))
+		__skb_zcopy_downgrade_managed(skb);
+}
+
 static inline void skb_mark_not_on_list(struct sk_buff *skb)
 {
 	skb->next = NULL;
@@ -3498,7 +3516,10 @@ static inline void __skb_frag_unref(skb_frag_t *frag, bool recycle)
  */
 static inline void skb_frag_unref(struct sk_buff *skb, int f)
 {
-	__skb_frag_unref(&skb_shinfo(skb)->frags[f], skb->pp_recycle);
+	struct skb_shared_info *shinfo = skb_shinfo(skb);
+
+	if (!skb_zcopy_managed(skb))
+		__skb_frag_unref(&shinfo->frags[f], skb->pp_recycle);
 }
 
 /**
diff --git a/net/core/datagram.c b/net/core/datagram.c
index 5237cb533bb4..a93c05156f56 100644
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -631,7 +631,6 @@ static int __zerocopy_sg_from_bvec(struct sock *sk, struct sk_buff *skb,
 
 		copied += v.bv_len;
 		truesize += PAGE_ALIGN(v.bv_len + v.bv_offset);
-		get_page(v.bv_page);
 		skb_fill_page_desc(skb, frag++, v.bv_page, v.bv_offset, v.bv_len);
 		bvec_iter_advance_single(from->bvec, &bi, v.bv_len);
 	}
@@ -660,11 +659,13 @@ static int __zerocopy_sg_from_bvec(struct sock *sk, struct sk_buff *skb,
 int __zerocopy_sg_from_iter(struct sock *sk, struct sk_buff *skb,
 			    struct iov_iter *from, size_t length)
 {
-	int frag = skb_shinfo(skb)->nr_frags;
+	int frag;
 
-	if (iov_iter_is_bvec(from))
+	if (skb_zcopy_managed(skb))
 		return __zerocopy_sg_from_bvec(sk, skb, from, length);
 
+	frag = skb_shinfo(skb)->nr_frags;
+
 	while (length && iov_iter_count(from)) {
 		struct page *pages[MAX_SKB_FRAGS];
 		struct page *last_head = NULL;
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 5b35791064d1..71870def129c 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -666,11 +666,18 @@ static void skb_release_data(struct sk_buff *skb)
 			      &shinfo->dataref))
 		goto exit;
 
-	skb_zcopy_clear(skb, true);
+	if (skb_zcopy(skb)) {
+		bool skip_unref = shinfo->flags & SKBFL_MANAGED_FRAG_REFS;
+
+		skb_zcopy_clear(skb, true);
+		if (skip_unref)
+			goto free_head;
+	}
 
 	for (i = 0; i < shinfo->nr_frags; i++)
 		__skb_frag_unref(&shinfo->frags[i], skb->pp_recycle);
 
+free_head:
 	if (shinfo->frag_list)
 		kfree_skb_list(shinfo->frag_list);
 
@@ -895,7 +902,10 @@ EXPORT_SYMBOL(skb_dump);
  */
 void skb_tx_error(struct sk_buff *skb)
 {
-	skb_zcopy_clear(skb, true);
+	if (skb) {
+		skb_zcopy_downgrade_managed(skb);
+		skb_zcopy_clear(skb, true);
+	}
 }
 EXPORT_SYMBOL(skb_tx_error);
 
@@ -1371,6 +1381,16 @@ int skb_zerocopy_iter_stream(struct sock *sk, struct sk_buff *skb,
 }
 EXPORT_SYMBOL_GPL(skb_zerocopy_iter_stream);
 
+void __skb_zcopy_downgrade_managed(struct sk_buff *skb)
+{
+	int i;
+
+	skb_shinfo(skb)->flags &= ~SKBFL_MANAGED_FRAG_REFS;
+	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
+		skb_frag_ref(skb, i);
+}
+EXPORT_SYMBOL_GPL(__skb_zcopy_downgrade_managed);
+
 static int skb_zerocopy_clone(struct sk_buff *nskb, struct sk_buff *orig,
 			      gfp_t gfp_mask)
 {
@@ -1688,6 +1708,8 @@ int pskb_expand_head(struct sk_buff *skb, int nhead, int ntail,
 
 	BUG_ON(skb_shared(skb));
 
+	skb_zcopy_downgrade_managed(skb);
+
 	size = SKB_DATA_ALIGN(size);
 
 	if (skb_pfmemalloc(skb))
@@ -3484,6 +3506,8 @@ void skb_split(struct sk_buff *skb, struct sk_buff *skb1, const u32 len)
 	int pos = skb_headlen(skb);
 	const int zc_flags = SKBFL_SHARED_FRAG | SKBFL_PURE_ZEROCOPY;
 
+	skb_zcopy_downgrade_managed(skb);
+
 	skb_shinfo(skb1)->flags |= skb_shinfo(skb)->flags & zc_flags;
 	skb_zerocopy_clone(skb1, skb, 0);
 	if (len < pos)	/* Split line is inside header. */
@@ -3837,6 +3861,7 @@ int skb_append_pagefrags(struct sk_buff *skb, struct page *page,
 	if (skb_can_coalesce(skb, i, page, offset)) {
 		skb_frag_size_add(&skb_shinfo(skb)->frags[i - 1], size);
 	} else if (i < MAX_SKB_FRAGS) {
+		skb_zcopy_downgrade_managed(skb);
 		get_page(page);
 		skb_fill_page_desc(skb, i, page, offset, size);
 	} else {
-- 
2.36.1

[PATCH net-next v3 07/25] net: don't track pfmemalloc for managed frags

[Pavel Begunkov]

Managed pages contain pinned userspace pages and controlled by upper
layers, there is no need in tracking skb->pfmemalloc for them.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/linux/skbuff.h | 28 +++++++++++++++++-----------
 net/core/datagram.c    |  7 +++++--
 2 files changed, 22 insertions(+), 13 deletions(-)

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 712168c21736..2d5badd4b9ff 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -2549,6 +2549,22 @@ static inline unsigned int skb_pagelen(const struct sk_buff *skb)
 	return skb_headlen(skb) + __skb_pagelen(skb);
 }
 
+static inline void __skb_fill_page_desc_noacc(struct skb_shared_info *shinfo,
+					      int i, struct page *page,
+					      int off, int size)
+{
+	skb_frag_t *frag = &shinfo->frags[i];
+
+	/*
+	 * Propagate page pfmemalloc to the skb if we can. The problem is
+	 * that not all callers have unique ownership of the page but rely
+	 * on page_is_pfmemalloc doing the right thing(tm).
+	 */
+	frag->bv_page		  = page;
+	frag->bv_offset		  = off;
+	skb_frag_size_set(frag, size);
+}
+
 /**
  * __skb_fill_page_desc - initialise a paged fragment in an skb
  * @skb: buffer containing fragment to be initialised
@@ -2565,17 +2581,7 @@ static inline unsigned int skb_pagelen(const struct sk_buff *skb)
 static inline void __skb_fill_page_desc(struct sk_buff *skb, int i,
 					struct page *page, int off, int size)
 {
-	skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
-
-	/*
-	 * Propagate page pfmemalloc to the skb if we can. The problem is
-	 * that not all callers have unique ownership of the page but rely
-	 * on page_is_pfmemalloc doing the right thing(tm).
-	 */
-	frag->bv_page		  = page;
-	frag->bv_offset		  = off;
-	skb_frag_size_set(frag, size);
-
+	__skb_fill_page_desc_noacc(skb_shinfo(skb), i, page, off, size);
 	page = compound_head(page);
 	if (page_is_pfmemalloc(page))
 		skb->pfmemalloc	= true;
diff --git a/net/core/datagram.c b/net/core/datagram.c
index a93c05156f56..3c913a6342ad 100644
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -616,7 +616,8 @@ EXPORT_SYMBOL(skb_copy_datagram_from_iter);
 static int __zerocopy_sg_from_bvec(struct sock *sk, struct sk_buff *skb,
 				   struct iov_iter *from, size_t length)
 {
-	int frag = skb_shinfo(skb)->nr_frags;
+	struct skb_shared_info *shinfo = skb_shinfo(skb);
+	int frag = shinfo->nr_frags;
 	int ret = 0;
 	struct bvec_iter bi;
 	ssize_t copied = 0;
@@ -631,12 +632,14 @@ static int __zerocopy_sg_from_bvec(struct sock *sk, struct sk_buff *skb,
 
 		copied += v.bv_len;
 		truesize += PAGE_ALIGN(v.bv_len + v.bv_offset);
-		skb_fill_page_desc(skb, frag++, v.bv_page, v.bv_offset, v.bv_len);
+		__skb_fill_page_desc_noacc(shinfo, frag++, v.bv_page,
+					   v.bv_offset, v.bv_len);
 		bvec_iter_advance_single(from->bvec, &bi, v.bv_len);
 	}
 	if (bi.bi_size)
 		ret = -EMSGSIZE;
 
+	shinfo->nr_frags = frag;
 	from->bvec += bi.bi_idx;
 	from->nr_segs -= bi.bi_idx;
 	from->count = bi.bi_size;
-- 
2.36.1

[PATCH net-next v3 08/25] skbuff: don't mix ubuf_info of different types

[Pavel Begunkov]

We should not append MSG_ZEROCOPY requests to skbuff with non
MSG_ZEROCOPY ubuf_info, they are not compatible.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 net/core/skbuff.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 71870def129c..7e6fcb3cd817 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -1222,6 +1222,10 @@ struct ubuf_info *msg_zerocopy_realloc(struct sock *sk, size_t size,
 		const u32 byte_limit = 1 << 19;		/* limit to a few TSO */
 		u32 bytelen, next;
 
+		/* there might be non MSG_ZEROCOPY users */
+		if (uarg->callback != msg_zerocopy_callback)
+			return NULL;
+
 		/* realloc only when socket is locked (TCP, UDP cork),
 		 * so uarg->len and sk_zckey access is serialized
 		 */
-- 
2.36.1

[PATCH net-next v3 09/25] ipv4/udp: support zc with managed data

[Pavel Begunkov]

Teach ipv4/udp about managed data. Make it recognise and use
msg->msg_ubuf, and also set/propagate SKBFL_MANAGED_FRAG_REFS
down to skb_zerocopy_iter_dgram().

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 net/ipv4/ip_output.c | 57 +++++++++++++++++++++++++++++++++-----------
 1 file changed, 43 insertions(+), 14 deletions(-)

diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 581d1e233260..3fd1bf675598 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -1017,18 +1017,35 @@ static int __ip_append_data(struct sock *sk,
 	    (!exthdrlen || (rt->dst.dev->features & NETIF_F_HW_ESP_TX_CSUM)))
 		csummode = CHECKSUM_PARTIAL;
 
-	if (flags & MSG_ZEROCOPY && length && sock_flag(sk, SOCK_ZEROCOPY)) {
-		uarg = msg_zerocopy_realloc(sk, length, skb_zcopy(skb));
-		if (!uarg)
-			return -ENOBUFS;
-		extra_uref = !skb_zcopy(skb);	/* only ref on new uarg */
-		if (rt->dst.dev->features & NETIF_F_SG &&
-		    csummode == CHECKSUM_PARTIAL) {
-			paged = true;
-			zc = true;
-		} else {
-			uarg->zerocopy = 0;
-			skb_zcopy_set(skb, uarg, &extra_uref);
+	if ((flags & MSG_ZEROCOPY) && length) {
+		struct msghdr *msg = from;
+
+		if (getfrag == ip_generic_getfrag && msg->msg_ubuf) {
+			if (skb_zcopy(skb) && msg->msg_ubuf != skb_zcopy(skb))
+				return -EINVAL;
+
+			/* Leave uarg NULL if can't zerocopy, callers should
+			 * be able to handle it.
+			 */
+			if ((rt->dst.dev->features & NETIF_F_SG) &&
+			    csummode == CHECKSUM_PARTIAL) {
+				paged = true;
+				zc = true;
+				uarg = msg->msg_ubuf;
+			}
+		} else if (sock_flag(sk, SOCK_ZEROCOPY)) {
+			uarg = msg_zerocopy_realloc(sk, length, skb_zcopy(skb));
+			if (!uarg)
+				return -ENOBUFS;
+			extra_uref = !skb_zcopy(skb);	/* only ref on new uarg */
+			if (rt->dst.dev->features & NETIF_F_SG &&
+			    csummode == CHECKSUM_PARTIAL) {
+				paged = true;
+				zc = true;
+			} else {
+				uarg->zerocopy = 0;
+				skb_zcopy_set(skb, uarg, &extra_uref);
+			}
 		}
 	}
 
@@ -1192,13 +1209,14 @@ static int __ip_append_data(struct sock *sk,
 				err = -EFAULT;
 				goto error;
 			}
-		} else if (!uarg || !uarg->zerocopy) {
+		} else if (!zc) {
 			int i = skb_shinfo(skb)->nr_frags;
 
 			err = -ENOMEM;
 			if (!sk_page_frag_refill(sk, pfrag))
 				goto error;
 
+			skb_zcopy_downgrade_managed(skb);
 			if (!skb_can_coalesce(skb, i, pfrag->page,
 					      pfrag->offset)) {
 				err = -EMSGSIZE;
@@ -1223,7 +1241,18 @@ static int __ip_append_data(struct sock *sk,
 			skb->truesize += copy;
 			wmem_alloc_delta += copy;
 		} else {
-			err = skb_zerocopy_iter_dgram(skb, from, copy);
+			struct msghdr *msg = from;
+
+			if (!skb_shinfo(skb)->nr_frags) {
+				if (msg->msg_managed_data)
+					skb_shinfo(skb)->flags |= SKBFL_MANAGED_FRAG_REFS;
+			} else {
+				/* appending, don't mix managed and unmanaged */
+				if (!msg->msg_managed_data)
+					skb_zcopy_downgrade_managed(skb);
+			}
+
+			err = skb_zerocopy_iter_dgram(skb, msg, copy);
 			if (err < 0)
 				goto error;
 		}
-- 
2.36.1

[PATCH net-next v3 10/25] ipv6/udp: support zc with managed data

[Pavel Begunkov]

Just as with ipv4/udp make ipv6/udp to take advantage of managed data
and propagate SKBFL_MANAGED_FRAG_REFS to skb_zerocopy_iter_dgram().

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 net/ipv6/ip6_output.c | 57 ++++++++++++++++++++++++++++++++-----------
 1 file changed, 43 insertions(+), 14 deletions(-)

diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index fc74ce3ed8cc..34eb3b5da5e2 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -1542,18 +1542,35 @@ static int __ip6_append_data(struct sock *sk,
 	    rt->dst.dev->features & (NETIF_F_IPV6_CSUM | NETIF_F_HW_CSUM))
 		csummode = CHECKSUM_PARTIAL;
 
-	if (flags & MSG_ZEROCOPY && length && sock_flag(sk, SOCK_ZEROCOPY)) {
-		uarg = msg_zerocopy_realloc(sk, length, skb_zcopy(skb));
-		if (!uarg)
-			return -ENOBUFS;
-		extra_uref = !skb_zcopy(skb);	/* only ref on new uarg */
-		if (rt->dst.dev->features & NETIF_F_SG &&
-		    csummode == CHECKSUM_PARTIAL) {
-			paged = true;
-			zc = true;
-		} else {
-			uarg->zerocopy = 0;
-			skb_zcopy_set(skb, uarg, &extra_uref);
+	if ((flags & MSG_ZEROCOPY) && length) {
+		struct msghdr *msg = from;
+
+		if (getfrag == ip_generic_getfrag && msg->msg_ubuf) {
+			if (skb_zcopy(skb) && msg->msg_ubuf != skb_zcopy(skb))
+				return -EINVAL;
+
+			/* Leave uarg NULL if can't zerocopy, callers should
+			 * be able to handle it.
+			 */
+			if ((rt->dst.dev->features & NETIF_F_SG) &&
+			    csummode == CHECKSUM_PARTIAL) {
+				paged = true;
+				zc = true;
+				uarg = msg->msg_ubuf;
+			}
+		} else if (sock_flag(sk, SOCK_ZEROCOPY)) {
+			uarg = msg_zerocopy_realloc(sk, length, skb_zcopy(skb));
+			if (!uarg)
+				return -ENOBUFS;
+			extra_uref = !skb_zcopy(skb);	/* only ref on new uarg */
+			if (rt->dst.dev->features & NETIF_F_SG &&
+			    csummode == CHECKSUM_PARTIAL) {
+				paged = true;
+				zc = true;
+			} else {
+				uarg->zerocopy = 0;
+				skb_zcopy_set(skb, uarg, &extra_uref);
+			}
 		}
 	}
 
@@ -1747,13 +1764,14 @@ static int __ip6_append_data(struct sock *sk,
 				err = -EFAULT;
 				goto error;
 			}
-		} else if (!uarg || !uarg->zerocopy) {
+		} else if (!zc) {
 			int i = skb_shinfo(skb)->nr_frags;
 
 			err = -ENOMEM;
 			if (!sk_page_frag_refill(sk, pfrag))
 				goto error;
 
+			skb_zcopy_downgrade_managed(skb);
 			if (!skb_can_coalesce(skb, i, pfrag->page,
 					      pfrag->offset)) {
 				err = -EMSGSIZE;
@@ -1778,7 +1796,18 @@ static int __ip6_append_data(struct sock *sk,
 			skb->truesize += copy;
 			wmem_alloc_delta += copy;
 		} else {
-			err = skb_zerocopy_iter_dgram(skb, from, copy);
+			struct msghdr *msg = from;
+
+			if (!skb_shinfo(skb)->nr_frags) {
+				if (msg->msg_managed_data)
+					skb_shinfo(skb)->flags |= SKBFL_MANAGED_FRAG_REFS;
+			} else {
+				/* appending, don't mix managed and unmanaged */
+				if (!msg->msg_managed_data)
+					skb_zcopy_downgrade_managed(skb);
+			}
+
+			err = skb_zerocopy_iter_dgram(skb, msg, copy);
 			if (err < 0)
 				goto error;
 		}
-- 
2.36.1

[PATCH net-next v3 11/25] tcp: support zc with managed data

[Pavel Begunkov]

Also make tcp to use managed data and propagate SKBFL_MANAGED_FRAG_REFS
to optimise frag pages referencing.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 net/ipv4/tcp.c | 52 ++++++++++++++++++++++++++++++++------------------
 1 file changed, 33 insertions(+), 19 deletions(-)

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 390eb3dc53bd..05e2f6271f65 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1223,17 +1223,23 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 
 	flags = msg->msg_flags;
 
-	if (flags & MSG_ZEROCOPY && size && sock_flag(sk, SOCK_ZEROCOPY)) {
+	if ((flags & MSG_ZEROCOPY) && size) {
 		skb = tcp_write_queue_tail(sk);
-		uarg = msg_zerocopy_realloc(sk, size, skb_zcopy(skb));
-		if (!uarg) {
-			err = -ENOBUFS;
-			goto out_err;
-		}
 
-		zc = sk->sk_route_caps & NETIF_F_SG;
-		if (!zc)
-			uarg->zerocopy = 0;
+		if (msg->msg_ubuf) {
+			uarg = msg->msg_ubuf;
+			net_zcopy_get(uarg);
+			zc = sk->sk_route_caps & NETIF_F_SG;
+		} else if (sock_flag(sk, SOCK_ZEROCOPY)) {
+			uarg = msg_zerocopy_realloc(sk, size, skb_zcopy(skb));
+			if (!uarg) {
+				err = -ENOBUFS;
+				goto out_err;
+			}
+			zc = sk->sk_route_caps & NETIF_F_SG;
+			if (!zc)
+				uarg->zerocopy = 0;
+		}
 	}
 
 	if (unlikely(flags & MSG_FASTOPEN || inet_sk(sk)->defer_connect) &&
@@ -1356,9 +1362,11 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 
 			copy = min_t(int, copy, pfrag->size - pfrag->offset);
 
-			if (tcp_downgrade_zcopy_pure(sk, skb))
-				goto wait_for_space;
-
+			if (unlikely(skb_zcopy_pure(skb) || skb_zcopy_managed(skb))) {
+				if (tcp_downgrade_zcopy_pure(sk, skb))
+					goto wait_for_space;
+				skb_zcopy_downgrade_managed(skb);
+			}
 			copy = tcp_wmem_schedule(sk, copy);
 			if (!copy)
 				goto wait_for_space;
@@ -1381,15 +1389,21 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 			pfrag->offset += copy;
 		} else {
 			/* First append to a fragless skb builds initial
-			 * pure zerocopy skb
+			 * zerocopy skb
 			 */
-			if (!skb->len)
+			if (!skb->len) {
+				if (msg->msg_managed_data)
+					skb_shinfo(skb)->flags |= SKBFL_MANAGED_FRAG_REFS;
 				skb_shinfo(skb)->flags |= SKBFL_PURE_ZEROCOPY;
-
-			if (!skb_zcopy_pure(skb)) {
-				copy = tcp_wmem_schedule(sk, copy);
-				if (!copy)
-					goto wait_for_space;
+			} else {
+				/* appending, don't mix managed and unmanaged */
+				if (!msg->msg_managed_data)
+					skb_zcopy_downgrade_managed(skb);
+				if (!skb_zcopy_pure(skb)) {
+					copy = tcp_wmem_schedule(sk, copy);
+					if (!copy)
+						goto wait_for_space;
+				}
 			}
 
 			err = skb_zerocopy_iter_stream(sk, skb, msg, copy, uarg);
-- 
2.36.1

[PATCH net-next v3 12/25] io_uring: add zc notification infrastructure

[Pavel Begunkov]

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>
---
 include/linux/io_uring_types.h |   5 ++
 io_uring/Makefile              |   2 +-
 io_uring/io_uring.c            |   6 +-
 io_uring/io_uring.h            |   1 +
 io_uring/notif.c               | 102 +++++++++++++++++++++++++++++++++
 io_uring/notif.h               |  64 +++++++++++++++++++++
 6 files changed, 177 insertions(+), 3 deletions(-)
 create mode 100644 io_uring/notif.c
 create mode 100644 io_uring/notif.h

diff --git a/include/linux/io_uring_types.h b/include/linux/io_uring_types.h
index 3ca8f363f504..a64eb2558e04 100644
--- a/include/linux/io_uring_types.h
+++ b/include/linux/io_uring_types.h
@@ -33,6 +33,9 @@ struct io_file_table {
 	unsigned int alloc_hint;
 };
 
+struct io_notif;
+struct io_notif_slot;
+
 struct io_hash_bucket {
 	spinlock_t		lock;
 	struct hlist_head	list;
@@ -207,6 +210,8 @@ struct io_ring_ctx {
 		unsigned		nr_user_files;
 		unsigned		nr_user_bufs;
 		struct io_mapped_ubuf	**user_bufs;
+		struct io_notif_slot	*notif_slots;
+		unsigned		nr_notif_slots;
 
 		struct io_submit_state	submit_state;
 
diff --git a/io_uring/Makefile b/io_uring/Makefile
index 466639c289be..8cc8e5387a75 100644
--- a/io_uring/Makefile
+++ b/io_uring/Makefile
@@ -7,5 +7,5 @@ obj-$(CONFIG_IO_URING)		+= io_uring.o xattr.o nop.o fs.o splice.o \
 					openclose.o uring_cmd.o epoll.o \
 					statx.o net.o msg_ring.o timeout.o \
 					sqpoll.o fdinfo.o tctx.o poll.o \
-					cancel.o kbuf.o rsrc.o rw.o opdef.o
+					cancel.o kbuf.o rsrc.o rw.o opdef.o notif.o
 obj-$(CONFIG_IO_WQ)		+= io-wq.o
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 070ee9ec9ee7..eff4adca1813 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -89,6 +89,7 @@
 #include "kbuf.h"
 #include "rsrc.h"
 #include "cancel.h"
+#include "notif.h"
 
 #include "timeout.h"
 #include "poll.h"
@@ -735,8 +736,7 @@ struct io_uring_cqe *__io_get_cqe(struct io_ring_ctx *ctx)
 	return &rings->cqes[off];
 }
 
-static bool io_fill_cqe_aux(struct io_ring_ctx *ctx,
-			    u64 user_data, s32 res, u32 cflags)
+bool io_fill_cqe_aux(struct io_ring_ctx *ctx, u64 user_data, s32 res, u32 cflags)
 {
 	struct io_uring_cqe *cqe;
 
@@ -2498,6 +2498,7 @@ static __cold void io_ring_ctx_free(struct io_ring_ctx *ctx)
 	}
 #endif
 	WARN_ON_ONCE(!list_empty(&ctx->ltimeout_list));
+	WARN_ON_ONCE(ctx->notif_slots || ctx->nr_notif_slots);
 
 	io_mem_free(ctx->rings);
 	io_mem_free(ctx->sq_sqes);
@@ -2674,6 +2675,7 @@ static __cold void io_ring_ctx_wait_and_kill(struct io_ring_ctx *ctx)
 		io_unregister_personality(ctx, index);
 	if (ctx->rings)
 		io_poll_remove_all(ctx, NULL, true);
+	io_notif_unregister(ctx);
 	mutex_unlock(&ctx->uring_lock);
 
 	/* failed during ring init, it couldn't have issued any requests */
diff --git a/io_uring/io_uring.h b/io_uring/io_uring.h
index f77e4a5403e4..7b7b63503c02 100644
--- a/io_uring/io_uring.h
+++ b/io_uring/io_uring.h
@@ -24,6 +24,7 @@ void io_req_complete_failed(struct io_kiocb *req, s32 res);
 void __io_req_complete(struct io_kiocb *req, unsigned issue_flags);
 void io_req_complete_post(struct io_kiocb *req);
 void __io_req_complete_post(struct io_kiocb *req);
+bool io_fill_cqe_aux(struct io_ring_ctx *ctx, u64 user_data, s32 res, u32 cflags);
 bool io_post_aux_cqe(struct io_ring_ctx *ctx, u64 user_data, s32 res, u32 cflags);
 void __io_commit_cqring_flush(struct io_ring_ctx *ctx);
 
diff --git a/io_uring/notif.c b/io_uring/notif.c
new file mode 100644
index 000000000000..e9e0c5566c4a
--- /dev/null
+++ b/io_uring/notif.c
@@ -0,0 +1,102 @@
+#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"
+
+static void __io_notif_complete_tw(struct callback_head *cb)
+{
+	struct io_notif *notif = container_of(cb, struct io_notif, task_work);
+	struct io_ring_ctx *ctx = notif->ctx;
+
+	io_cq_lock(ctx);
+	io_fill_cqe_aux(ctx, notif->tag, 0, notif->seq);
+	io_cq_unlock_post(ctx);
+
+	percpu_ref_put(&ctx->refs);
+	kfree(notif);
+}
+
+static inline void io_notif_complete(struct io_notif *notif)
+{
+	__io_notif_complete_tw(&notif->task_work);
+}
+
+static void io_notif_complete_wq(struct work_struct *work)
+{
+	struct io_notif *notif = container_of(work, struct io_notif, commit_work);
+
+	io_notif_complete(notif);
+}
+
+static void io_uring_tx_zerocopy_callback(struct sk_buff *skb,
+					  struct ubuf_info *uarg,
+					  bool success)
+{
+	struct io_notif *notif = container_of(uarg, struct io_notif, uarg);
+
+	if (!refcount_dec_and_test(&uarg->refcnt))
+		return;
+	INIT_WORK(&notif->commit_work, io_notif_complete_wq);
+	queue_work(system_unbound_wq, &notif->commit_work);
+}
+
+struct io_notif *io_alloc_notif(struct io_ring_ctx *ctx,
+				struct io_notif_slot *slot)
+	__must_hold(&ctx->uring_lock)
+{
+	struct io_notif *notif;
+
+	notif = kzalloc(sizeof(*notif), GFP_ATOMIC | __GFP_ACCOUNT);
+	if (!notif)
+		return NULL;
+
+	notif->seq = slot->seq++;
+	notif->tag = slot->tag;
+	notif->ctx = ctx;
+	notif->uarg.flags = SKBFL_ZEROCOPY_FRAG | SKBFL_DONT_ORPHAN;
+	notif->uarg.callback = io_uring_tx_zerocopy_callback;
+	/* master ref owned by io_notif_slot, will be dropped on flush */
+	refcount_set(&notif->uarg.refcnt, 1);
+	percpu_ref_get(&ctx->refs);
+	return notif;
+}
+
+static void io_notif_slot_flush(struct io_notif_slot *slot)
+	__must_hold(&ctx->uring_lock)
+{
+	struct io_notif *notif = slot->notif;
+
+	slot->notif = NULL;
+
+	if (WARN_ON_ONCE(in_interrupt()))
+		return;
+	/* drop slot's master ref */
+	if (refcount_dec_and_test(&notif->uarg.refcnt))
+		io_notif_complete(notif);
+}
+
+__cold int io_notif_unregister(struct io_ring_ctx *ctx)
+	__must_hold(&ctx->uring_lock)
+{
+	int i;
+
+	if (!ctx->notif_slots)
+		return -ENXIO;
+
+	for (i = 0; i < ctx->nr_notif_slots; i++) {
+		struct io_notif_slot *slot = &ctx->notif_slots[i];
+
+		if (slot->notif)
+			io_notif_slot_flush(slot);
+	}
+
+	kvfree(ctx->notif_slots);
+	ctx->notif_slots = NULL;
+	ctx->nr_notif_slots = 0;
+	return 0;
+}
\ No newline at end of file
diff --git a/io_uring/notif.h b/io_uring/notif.h
new file mode 100644
index 000000000000..3d7a1d242e17
--- /dev/null
+++ b/io_uring/notif.h
@@ -0,0 +1,64 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/net.h>
+#include <linux/uio.h>
+#include <net/sock.h>
+#include <linux/nospec.h>
+
+struct io_notif {
+	struct ubuf_info	uarg;
+	struct io_ring_ctx	*ctx;
+
+	/* cqe->user_data, io_notif_slot::tag if not overridden */
+	u64			tag;
+	/* see struct io_notif_slot::seq */
+	u32			seq;
+
+	union {
+		struct callback_head	task_work;
+		struct work_struct	commit_work;
+	};
+};
+
+struct io_notif_slot {
+	/*
+	 * Current/active notifier. A slot holds only one active notifier at a
+	 * time and keeps one reference to it. Flush releases the reference and
+	 * lazily replaces it with a new notifier.
+	 */
+	struct io_notif		*notif;
+
+	/*
+	 * Default ->user_data for this slot notifiers CQEs
+	 */
+	u64			tag;
+	/*
+	 * Notifiers of a slot live in generations, we create a new notifier
+	 * only after flushing the previous one. Track the sequential number
+	 * for all notifiers and copy it into notifiers's cqe->cflags
+	 */
+	u32			seq;
+};
+
+int io_notif_unregister(struct io_ring_ctx *ctx);
+
+struct io_notif *io_alloc_notif(struct io_ring_ctx *ctx,
+				struct io_notif_slot *slot);
+
+static inline struct io_notif *io_get_notif(struct io_ring_ctx *ctx,
+					    struct io_notif_slot *slot)
+{
+	if (!slot->notif)
+		slot->notif = io_alloc_notif(ctx, slot);
+	return slot->notif;
+}
+
+static inline struct io_notif_slot *io_get_notif_slot(struct io_ring_ctx *ctx,
+						      int idx)
+	__must_hold(&ctx->uring_lock)
+{
+	if (idx >= ctx->nr_notif_slots)
+		return NULL;
+	idx = array_index_nospec(idx, ctx->nr_notif_slots);
+	return &ctx->notif_slots[idx];
+}
-- 
2.36.1

[PATCH net-next v3 13/25] io_uring: export task put

[Pavel Begunkov]

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/linux/io_uring_types.h | 25 +++++++++++++++++++++++++
 io_uring/io_uring.c            | 11 +----------
 io_uring/io_uring.h            | 10 ++++++++++
 io_uring/tctx.h                | 26 --------------------------
 4 files changed, 36 insertions(+), 36 deletions(-)

diff --git a/include/linux/io_uring_types.h b/include/linux/io_uring_types.h
index a64eb2558e04..26a1504ad24c 100644
--- a/include/linux/io_uring_types.h
+++ b/include/linux/io_uring_types.h
@@ -4,6 +4,7 @@
 #include <linux/blkdev.h>
 #include <linux/task_work.h>
 #include <linux/bitmap.h>
+#include <linux/llist.h>
 #include <uapi/linux/io_uring.h>
 
 struct io_wq_work_node {
@@ -46,6 +47,30 @@ struct io_hash_table {
 	unsigned		hash_bits;
 };
 
+/*
+ * Arbitrary limit, can be raised if need be
+ */
+#define IO_RINGFD_REG_MAX 16
+
+struct io_uring_task {
+	/* submission side */
+	int				cached_refs;
+	const struct io_ring_ctx 	*last;
+	struct io_wq			*io_wq;
+	struct file			*registered_rings[IO_RINGFD_REG_MAX];
+
+	struct xarray			xa;
+	struct wait_queue_head		wait;
+	atomic_t			in_idle;
+	atomic_t			inflight_tracked;
+	struct percpu_counter		inflight;
+
+	struct { /* task_work */
+		struct llist_head	task_list;
+		struct callback_head	task_work;
+	} ____cacheline_aligned_in_smp;
+};
+
 struct io_uring {
 	u32 head ____cacheline_aligned_in_smp;
 	u32 tail ____cacheline_aligned_in_smp;
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index eff4adca1813..5fbbdcad14fa 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -603,7 +603,7 @@ static bool io_cqring_overflow_flush(struct io_ring_ctx *ctx)
 	return ret;
 }
 
-static void __io_put_task(struct task_struct *task, int nr)
+void __io_put_task(struct task_struct *task, int nr)
 {
 	struct io_uring_task *tctx = task->io_uring;
 
@@ -613,15 +613,6 @@ static void __io_put_task(struct task_struct *task, int nr)
 	put_task_struct_many(task, nr);
 }
 
-/* must to be called somewhat shortly after putting a request */
-static inline void io_put_task(struct task_struct *task, int nr)
-{
-	if (likely(task == current))
-		task->io_uring->cached_refs += nr;
-	else
-		__io_put_task(task, nr);
-}
-
 static void io_task_refs_refill(struct io_uring_task *tctx)
 {
 	unsigned int refill = -tctx->cached_refs + IO_TCTX_REFS_CACHE_NR;
diff --git a/io_uring/io_uring.h b/io_uring/io_uring.h
index 7b7b63503c02..e978654d1b14 100644
--- a/io_uring/io_uring.h
+++ b/io_uring/io_uring.h
@@ -59,6 +59,7 @@ void io_wq_submit_work(struct io_wq_work *work);
 
 void io_free_req(struct io_kiocb *req);
 void io_queue_next(struct io_kiocb *req);
+void __io_put_task(struct task_struct *task, int nr);
 
 bool io_match_task_safe(struct io_kiocb *head, struct task_struct *task,
 			bool cancel_all);
@@ -244,4 +245,13 @@ static inline void io_commit_cqring_flush(struct io_ring_ctx *ctx)
 		__io_commit_cqring_flush(ctx);
 }
 
+/* must to be called somewhat shortly after putting a request */
+static inline void io_put_task(struct task_struct *task, int nr)
+{
+	if (likely(task == current))
+		task->io_uring->cached_refs += nr;
+	else
+		__io_put_task(task, nr);
+}
+
 #endif
diff --git a/io_uring/tctx.h b/io_uring/tctx.h
index 8a33ff6e5d91..25974beed4d6 100644
--- a/io_uring/tctx.h
+++ b/io_uring/tctx.h
@@ -1,31 +1,5 @@
 // SPDX-License-Identifier: GPL-2.0
 
-#include <linux/llist.h>
-
-/*
- * Arbitrary limit, can be raised if need be
- */
-#define IO_RINGFD_REG_MAX 16
-
-struct io_uring_task {
-	/* submission side */
-	int				cached_refs;
-	const struct io_ring_ctx 	*last;
-	struct io_wq			*io_wq;
-	struct file			*registered_rings[IO_RINGFD_REG_MAX];
-
-	struct xarray			xa;
-	struct wait_queue_head		wait;
-	atomic_t			in_idle;
-	atomic_t			inflight_tracked;
-	struct percpu_counter		inflight;
-
-	struct { /* task_work */
-		struct llist_head	task_list;
-		struct callback_head	task_work;
-	} ____cacheline_aligned_in_smp;
-};
-
 struct io_tctx_node {
 	struct list_head	ctx_node;
 	struct task_struct	*task;
-- 
2.36.1

[PATCH net-next v3 14/25] io_uring: cache struct io_notif

[Pavel Begunkov]

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>
---
 include/linux/io_uring_types.h |  7 +++++
 io_uring/io_uring.c            |  3 ++
 io_uring/notif.c               | 57 +++++++++++++++++++++++++++++-----
 io_uring/notif.h               |  5 +++
 4 files changed, 65 insertions(+), 7 deletions(-)

diff --git a/include/linux/io_uring_types.h b/include/linux/io_uring_types.h
index 26a1504ad24c..0a3fdaa368a4 100644
--- a/include/linux/io_uring_types.h
+++ b/include/linux/io_uring_types.h
@@ -244,6 +244,9 @@ struct io_ring_ctx {
 		struct xarray		io_bl_xa;
 		struct list_head	io_buffers_cache;
 
+		/* struct io_notif cache, protected by uring_lock */
+		struct list_head	notif_list;
+
 		struct io_hash_table	cancel_table_locked;
 		struct list_head	cq_overflow_list;
 		struct list_head	apoll_cache;
@@ -255,6 +258,10 @@ struct io_ring_ctx {
 	struct io_wq_work_list	locked_free_list;
 	unsigned int		locked_free_nr;
 
+	/* struct io_notif cache protected by completion_lock */
+	struct list_head	notif_list_locked;
+	unsigned int		notif_locked_nr;
+
 	const struct cred	*sq_creds;	/* cred used for __io_sq_thread() */
 	struct io_sq_data	*sq_data;	/* if using sq thread polling */
 
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 5fbbdcad14fa..6054e71e6ade 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -318,6 +318,8 @@ static __cold struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
 	INIT_WQ_LIST(&ctx->locked_free_list);
 	INIT_DELAYED_WORK(&ctx->fallback_work, io_fallback_req_func);
 	INIT_WQ_LIST(&ctx->submit_state.compl_reqs);
+	INIT_LIST_HEAD(&ctx->notif_list);
+	INIT_LIST_HEAD(&ctx->notif_list_locked);
 	return ctx;
 err:
 	kfree(ctx->dummy_ubuf);
@@ -2491,6 +2493,7 @@ static __cold void io_ring_ctx_free(struct io_ring_ctx *ctx)
 	WARN_ON_ONCE(!list_empty(&ctx->ltimeout_list));
 	WARN_ON_ONCE(ctx->notif_slots || ctx->nr_notif_slots);
 
+	io_notif_cache_purge(ctx);
 	io_mem_free(ctx->rings);
 	io_mem_free(ctx->sq_sqes);
 
diff --git a/io_uring/notif.c b/io_uring/notif.c
index e9e0c5566c4a..ffbd5ce03c36 100644
--- a/io_uring/notif.c
+++ b/io_uring/notif.c
@@ -15,10 +15,12 @@ static void __io_notif_complete_tw(struct callback_head *cb)
 
 	io_cq_lock(ctx);
 	io_fill_cqe_aux(ctx, notif->tag, 0, notif->seq);
+
+	list_add(&notif->cache_node, &ctx->notif_list_locked);
+	ctx->notif_locked_nr++;
 	io_cq_unlock_post(ctx);
 
 	percpu_ref_put(&ctx->refs);
-	kfree(notif);
 }
 
 static inline void io_notif_complete(struct io_notif *notif)
@@ -45,21 +47,62 @@ static void io_uring_tx_zerocopy_callback(struct sk_buff *skb,
 	queue_work(system_unbound_wq, &notif->commit_work);
 }
 
+static void io_notif_splice_cached(struct io_ring_ctx *ctx)
+	__must_hold(&ctx->uring_lock)
+{
+	spin_lock(&ctx->completion_lock);
+	list_splice_init(&ctx->notif_list_locked, &ctx->notif_list);
+	ctx->notif_locked_nr = 0;
+	spin_unlock(&ctx->completion_lock);
+}
+
+void io_notif_cache_purge(struct io_ring_ctx *ctx)
+	__must_hold(&ctx->uring_lock)
+{
+	io_notif_splice_cached(ctx);
+
+	while (!list_empty(&ctx->notif_list)) {
+		struct io_notif *notif = list_first_entry(&ctx->notif_list,
+						struct io_notif, cache_node);
+
+		list_del(&notif->cache_node);
+		kfree(notif);
+	}
+}
+
+static inline bool io_notif_has_cached(struct io_ring_ctx *ctx)
+	__must_hold(&ctx->uring_lock)
+{
+	if (likely(!list_empty(&ctx->notif_list)))
+		return true;
+	if (data_race(READ_ONCE(ctx->notif_locked_nr) <= IO_NOTIF_SPLICE_BATCH))
+		return false;
+	io_notif_splice_cached(ctx);
+	return !list_empty(&ctx->notif_list);
+}
+
 struct io_notif *io_alloc_notif(struct io_ring_ctx *ctx,
 				struct io_notif_slot *slot)
 	__must_hold(&ctx->uring_lock)
 {
 	struct io_notif *notif;
 
-	notif = kzalloc(sizeof(*notif), GFP_ATOMIC | __GFP_ACCOUNT);
-	if (!notif)
-		return NULL;
+	if (likely(io_notif_has_cached(ctx))) {
+		notif = list_first_entry(&ctx->notif_list,
+					 struct io_notif, cache_node);
+		list_del(&notif->cache_node);
+	} else {
+		notif = kzalloc(sizeof(*notif), GFP_ATOMIC | __GFP_ACCOUNT);
+		if (!notif)
+			return NULL;
+		/* pre-initialise some fields */
+		notif->ctx = ctx;
+		notif->uarg.flags = SKBFL_ZEROCOPY_FRAG | SKBFL_DONT_ORPHAN;
+		notif->uarg.callback = io_uring_tx_zerocopy_callback;
+	}
 
 	notif->seq = slot->seq++;
 	notif->tag = slot->tag;
-	notif->ctx = ctx;
-	notif->uarg.flags = SKBFL_ZEROCOPY_FRAG | SKBFL_DONT_ORPHAN;
-	notif->uarg.callback = io_uring_tx_zerocopy_callback;
 	/* master ref owned by io_notif_slot, will be dropped on flush */
 	refcount_set(&notif->uarg.refcnt, 1);
 	percpu_ref_get(&ctx->refs);
diff --git a/io_uring/notif.h b/io_uring/notif.h
index 3d7a1d242e17..b23c9c0515bb 100644
--- a/io_uring/notif.h
+++ b/io_uring/notif.h
@@ -5,6 +5,8 @@
 #include <net/sock.h>
 #include <linux/nospec.h>
 
+#define IO_NOTIF_SPLICE_BATCH	32
+
 struct io_notif {
 	struct ubuf_info	uarg;
 	struct io_ring_ctx	*ctx;
@@ -13,6 +15,8 @@ struct io_notif {
 	u64			tag;
 	/* see struct io_notif_slot::seq */
 	u32			seq;
+	/* hook into ctx->notif_list and ctx->notif_list_locked */
+	struct list_head	cache_node;
 
 	union {
 		struct callback_head	task_work;
@@ -41,6 +45,7 @@ struct io_notif_slot {
 };
 
 int io_notif_unregister(struct io_ring_ctx *ctx);
+void io_notif_cache_purge(struct io_ring_ctx *ctx);
 
 struct io_notif *io_alloc_notif(struct io_ring_ctx *ctx,
 				struct io_notif_slot *slot);
-- 
2.36.1

[PATCH net-next v3 15/25] io_uring: complete notifiers in tw

[Pavel Begunkov]

We need a task context to post CQEs but using wq is too expensive.
Try to complete notifiers using task_work and fall back to wq if fails.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 io_uring/notif.c | 22 +++++++++++++++++++---
 io_uring/notif.h |  3 +++
 2 files changed, 22 insertions(+), 3 deletions(-)

diff --git a/io_uring/notif.c b/io_uring/notif.c
index ffbd5ce03c36..f795e820de56 100644
--- a/io_uring/notif.c
+++ b/io_uring/notif.c
@@ -13,6 +13,11 @@ static void __io_notif_complete_tw(struct callback_head *cb)
 	struct io_notif *notif = container_of(cb, struct io_notif, task_work);
 	struct io_ring_ctx *ctx = notif->ctx;
 
+	if (likely(notif->task)) {
+		io_put_task(notif->task, 1);
+		notif->task = NULL;
+	}
+
 	io_cq_lock(ctx);
 	io_fill_cqe_aux(ctx, notif->tag, 0, notif->seq);
 
@@ -43,6 +48,14 @@ static void io_uring_tx_zerocopy_callback(struct sk_buff *skb,
 
 	if (!refcount_dec_and_test(&uarg->refcnt))
 		return;
+
+	if (likely(notif->task)) {
+		init_task_work(&notif->task_work, __io_notif_complete_tw);
+		if (likely(!task_work_add(notif->task, &notif->task_work,
+					  TWA_SIGNAL)))
+		return;
+	}
+
 	INIT_WORK(&notif->commit_work, io_notif_complete_wq);
 	queue_work(system_unbound_wq, &notif->commit_work);
 }
@@ -134,12 +147,15 @@ __cold int io_notif_unregister(struct io_ring_ctx *ctx)
 	for (i = 0; i < ctx->nr_notif_slots; i++) {
 		struct io_notif_slot *slot = &ctx->notif_slots[i];
 
-		if (slot->notif)
-			io_notif_slot_flush(slot);
+		if (!slot->notif)
+			continue;
+		if (WARN_ON_ONCE(slot->notif->task))
+			slot->notif->task = NULL;
+		io_notif_slot_flush(slot);
 	}
 
 	kvfree(ctx->notif_slots);
 	ctx->notif_slots = NULL;
 	ctx->nr_notif_slots = 0;
 	return 0;
-}
\ No newline at end of file
+}
diff --git a/io_uring/notif.h b/io_uring/notif.h
index b23c9c0515bb..23ca7620fff9 100644
--- a/io_uring/notif.h
+++ b/io_uring/notif.h
@@ -11,6 +11,9 @@ struct io_notif {
 	struct ubuf_info	uarg;
 	struct io_ring_ctx	*ctx;
 
+	/* complete via tw if ->task is non-NULL, fallback to wq otherwise */
+	struct task_struct	*task;
+
 	/* cqe->user_data, io_notif_slot::tag if not overridden */
 	u64			tag;
 	/* see struct io_notif_slot::seq */
-- 
2.36.1

[PATCH net-next v3 16/25] io_uring: add notification slot registration

[Pavel Begunkov]

Let the userspace to register and unregister notification slots.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/uapi/linux/io_uring.h | 17 ++++++++++++++
 io_uring/io_uring.c           |  9 ++++++++
 io_uring/notif.c              | 43 +++++++++++++++++++++++++++++++++++
 io_uring/notif.h              |  3 +++
 4 files changed, 72 insertions(+)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 09e7c3b13d2d..9b7ea3e1018f 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -429,6 +429,10 @@ enum {
 	/* sync cancelation API */
 	IORING_REGISTER_SYNC_CANCEL		= 24,
 
+	/* zerocopy notification API */
+	IORING_REGISTER_NOTIFIERS		= 25,
+	IORING_UNREGISTER_NOTIFIERS		= 26,
+
 	/* this goes last */
 	IORING_REGISTER_LAST
 };
@@ -475,6 +479,19 @@ struct io_uring_rsrc_update2 {
 	__u32 resv2;
 };
 
+struct io_uring_notification_slot {
+	__u64 tag;
+	__u64 resv[3];
+};
+
+struct io_uring_notification_register {
+	__u32 nr_slots;
+	__u32 resv;
+	__u64 resv2;
+	__u64 data;
+	__u64 resv3;
+};
+
 /* Skip updating fd indexes set to this value in the fd table */
 #define IORING_REGISTER_FILES_SKIP	(-2)
 
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 6054e71e6ade..3b885d65e569 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -3862,6 +3862,15 @@ static int __io_uring_register(struct io_ring_ctx *ctx, unsigned opcode,
 			break;
 		ret = io_sync_cancel(ctx, arg);
 		break;
+	case IORING_REGISTER_NOTIFIERS:
+		ret = io_notif_register(ctx, arg, nr_args);
+		break;
+	case IORING_UNREGISTER_NOTIFIERS:
+		ret = -EINVAL;
+		if (arg || nr_args)
+			break;
+		ret = io_notif_unregister(ctx);
+		break;
 	default:
 		ret = -EINVAL;
 		break;
diff --git a/io_uring/notif.c b/io_uring/notif.c
index f795e820de56..2e9329f97d2c 100644
--- a/io_uring/notif.c
+++ b/io_uring/notif.c
@@ -157,5 +157,48 @@ __cold int io_notif_unregister(struct io_ring_ctx *ctx)
 	kvfree(ctx->notif_slots);
 	ctx->notif_slots = NULL;
 	ctx->nr_notif_slots = 0;
+	io_notif_cache_purge(ctx);
+	return 0;
+}
+
+__cold int io_notif_register(struct io_ring_ctx *ctx,
+			     void __user *arg, unsigned int size)
+	__must_hold(&ctx->uring_lock)
+{
+	struct io_uring_notification_slot __user *slots;
+	struct io_uring_notification_slot slot;
+	struct io_uring_notification_register reg;
+	unsigned i;
+
+	if (ctx->nr_notif_slots)
+		return -EBUSY;
+	if (size != sizeof(reg))
+		return -EINVAL;
+	if (copy_from_user(&reg, arg, sizeof(reg)))
+		return -EFAULT;
+	if (!reg.nr_slots || reg.nr_slots > IORING_MAX_NOTIF_SLOTS)
+		return -EINVAL;
+	if (reg.resv || reg.resv2 || reg.resv3)
+		return -EINVAL;
+
+	slots = u64_to_user_ptr(reg.data);
+	ctx->notif_slots = kvcalloc(reg.nr_slots, sizeof(ctx->notif_slots[0]),
+				GFP_KERNEL_ACCOUNT);
+	if (!ctx->notif_slots)
+		return -ENOMEM;
+
+	for (i = 0; i < reg.nr_slots; i++, ctx->nr_notif_slots++) {
+		struct io_notif_slot *notif_slot = &ctx->notif_slots[i];
+
+		if (copy_from_user(&slot, &slots[i], sizeof(slot))) {
+			io_notif_unregister(ctx);
+			return -EFAULT;
+		}
+		if (slot.resv[0] | slot.resv[1] | slot.resv[2]) {
+			io_notif_unregister(ctx);
+			return -EINVAL;
+		}
+		notif_slot->tag = slot.tag;
+	}
 	return 0;
 }
diff --git a/io_uring/notif.h b/io_uring/notif.h
index 23ca7620fff9..6dde39c6afbe 100644
--- a/io_uring/notif.h
+++ b/io_uring/notif.h
@@ -6,6 +6,7 @@
 #include <linux/nospec.h>
 
 #define IO_NOTIF_SPLICE_BATCH	32
+#define IORING_MAX_NOTIF_SLOTS (1U << 10)
 
 struct io_notif {
 	struct ubuf_info	uarg;
@@ -47,6 +48,8 @@ struct io_notif_slot {
 	u32			seq;
 };
 
+int io_notif_register(struct io_ring_ctx *ctx,
+		      void __user *arg, unsigned int size);
 int io_notif_unregister(struct io_ring_ctx *ctx);
 void io_notif_cache_purge(struct io_ring_ctx *ctx);
 
-- 
2.36.1

[PATCH net-next v3 17/25] io_uring: wire send zc request type

[Pavel Begunkov]

Add a new io_uring opcode IORING_OP_SENDZC. The main distinction from
IORING_OP_SEND is that the user should specify a notification slot
index in sqe::notification_idx and the buffers are safe to reuse only
when the used notification is flushed and completes.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/uapi/linux/io_uring.h |  5 +++
 io_uring/net.c                | 84 +++++++++++++++++++++++++++++++++++
 io_uring/net.h                |  4 ++
 io_uring/opdef.c              | 15 +++++++
 4 files changed, 108 insertions(+)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 9b7ea3e1018f..0e1e179cec1d 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -62,6 +62,10 @@ struct io_uring_sqe {
 	union {
 		__s32	splice_fd_in;
 		__u32	file_index;
+		struct {
+			__u16	notification_idx;
+			__u16	__pad;
+		} __attribute__((packed));
 	};
 	union {
 		struct {
@@ -193,6 +197,7 @@ enum io_uring_op {
 	IORING_OP_GETXATTR,
 	IORING_OP_SOCKET,
 	IORING_OP_URING_CMD,
+	IORING_OP_SENDZC,
 
 	/* this goes last, obviously */
 	IORING_OP_LAST,
diff --git a/io_uring/net.c b/io_uring/net.c
index d95c88d83f9f..ef492f1360c8 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -13,6 +13,7 @@
 #include "io_uring.h"
 #include "kbuf.h"
 #include "net.h"
+#include "notif.h"
 
 #if defined(CONFIG_NET)
 struct io_shutdown {
@@ -58,6 +59,14 @@ struct io_sr_msg {
 	unsigned int			flags;
 };
 
+struct io_sendzc {
+	struct file			*file;
+	void __user			*buf;
+	size_t				len;
+	u16				slot_idx;
+	int				msg_flags;
+};
+
 #define IO_APOLL_MULTI_POLLED (REQ_F_APOLL_MULTISHOT | REQ_F_POLLED)
 
 int io_shutdown_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
@@ -770,4 +779,79 @@ int io_connect(struct io_kiocb *req, unsigned int issue_flags)
 	io_req_set_res(req, ret, 0);
 	return IOU_OK;
 }
+
+int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
+{
+	struct io_sendzc *zc = io_kiocb_to_cmd(req);
+
+	if (READ_ONCE(sqe->ioprio) || READ_ONCE(sqe->addr2) || READ_ONCE(sqe->__pad2[0]))
+		return -EINVAL;
+
+	zc->buf = u64_to_user_ptr(READ_ONCE(sqe->addr));
+	zc->len = READ_ONCE(sqe->len);
+	zc->msg_flags = READ_ONCE(sqe->msg_flags) | MSG_NOSIGNAL;
+	zc->slot_idx = READ_ONCE(sqe->notification_idx);
+	if (zc->msg_flags & MSG_DONTWAIT)
+		req->flags |= REQ_F_NOWAIT;
+#ifdef CONFIG_COMPAT
+	if (req->ctx->compat)
+		zc->msg_flags |= MSG_CMSG_COMPAT;
+#endif
+	return 0;
+}
+
+int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
+{
+	struct io_ring_ctx *ctx = req->ctx;
+	struct io_sendzc *zc = io_kiocb_to_cmd(req);
+	struct io_notif_slot *notif_slot;
+	struct io_notif *notif;
+	struct msghdr msg;
+	struct iovec iov;
+	struct socket *sock;
+	unsigned msg_flags;
+	int ret, min_ret = 0;
+
+	if (issue_flags & IO_URING_F_UNLOCKED)
+		return -EAGAIN;
+	sock = sock_from_file(req->file);
+	if (unlikely(!sock))
+		return -ENOTSOCK;
+
+	notif_slot = io_get_notif_slot(ctx, zc->slot_idx);
+	if (!notif_slot)
+		return -EINVAL;
+	notif = io_get_notif(ctx, notif_slot);
+	if (!notif)
+		return -ENOMEM;
+
+	msg.msg_name = NULL;
+	msg.msg_control = NULL;
+	msg.msg_controllen = 0;
+	msg.msg_namelen = 0;
+	msg.msg_managed_data = 0;
+
+	ret = import_single_range(WRITE, zc->buf, zc->len, &iov, &msg.msg_iter);
+	if (unlikely(ret))
+		return ret;
+
+	msg_flags = zc->msg_flags | MSG_ZEROCOPY;
+	if (issue_flags & IO_URING_F_NONBLOCK)
+		msg_flags |= MSG_DONTWAIT;
+	if (msg_flags & MSG_WAITALL)
+		min_ret = iov_iter_count(&msg.msg_iter);
+
+	msg.msg_flags = msg_flags;
+	msg.msg_ubuf = &notif->uarg;
+	ret = sock_sendmsg(sock, &msg);
+
+	if (unlikely(ret < min_ret)) {
+		if (ret == -EAGAIN && (issue_flags & IO_URING_F_NONBLOCK))
+			return -EAGAIN;
+		return ret == -ERESTARTSYS ? -EINTR : ret;
+	}
+
+	io_req_set_res(req, ret, 0);
+	return IOU_OK;
+}
 #endif
diff --git a/io_uring/net.h b/io_uring/net.h
index 81d71d164770..1dba8befebb3 100644
--- a/io_uring/net.h
+++ b/io_uring/net.h
@@ -40,4 +40,8 @@ int io_socket(struct io_kiocb *req, unsigned int issue_flags);
 int io_connect_prep_async(struct io_kiocb *req);
 int io_connect_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe);
 int io_connect(struct io_kiocb *req, unsigned int issue_flags);
+
+int io_sendzc(struct io_kiocb *req, unsigned int issue_flags);
+int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe);
+
 #endif
diff --git a/io_uring/opdef.c b/io_uring/opdef.c
index 0be00db9e31c..91d425b43174 100644
--- a/io_uring/opdef.c
+++ b/io_uring/opdef.c
@@ -466,6 +466,21 @@ const struct io_op_def io_op_defs[] = {
 		.issue			= io_uring_cmd,
 		.prep_async		= io_uring_cmd_prep_async,
 	},
+	[IORING_OP_SENDZC] = {
+		.name			= "SENDZC",
+		.needs_file		= 1,
+		.unbound_nonreg_file	= 1,
+		.pollout		= 1,
+		.audit_skip		= 1,
+		.ioprio			= 1,
+#if defined(CONFIG_NET)
+		.prep			= io_sendzc_prep,
+		.issue			= io_sendzc,
+#else
+		.prep			= io_eopnotsupp_prep,
+#endif
+
+	},
 };
 
 const char *io_uring_get_opcode(u8 opcode)
-- 
2.36.1

[PATCH net-next v3 18/25] io_uring: account locked pages for non-fixed zc

[Pavel Begunkov]

Fixed buffers are RLIMIT_MEMLOCK accounted, however it doesn't cover iovec
based zerocopy sends. Do the accounting on the io_uring side.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 io_uring/net.c   | 1 +
 io_uring/notif.c | 6 ++++++
 2 files changed, 7 insertions(+)

diff --git a/io_uring/net.c b/io_uring/net.c
index ef492f1360c8..d5b00e07e72b 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -834,6 +834,7 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
 	ret = import_single_range(WRITE, zc->buf, zc->len, &iov, &msg.msg_iter);
 	if (unlikely(ret))
 		return ret;
+	mm_account_pinned_pages(&notif->uarg.mmp, zc->len);
 
 	msg_flags = zc->msg_flags | MSG_ZEROCOPY;
 	if (issue_flags & IO_URING_F_NONBLOCK)
diff --git a/io_uring/notif.c b/io_uring/notif.c
index 2e9329f97d2c..0a03d04c010b 100644
--- a/io_uring/notif.c
+++ b/io_uring/notif.c
@@ -12,7 +12,13 @@ static void __io_notif_complete_tw(struct callback_head *cb)
 {
 	struct io_notif *notif = container_of(cb, struct io_notif, task_work);
 	struct io_ring_ctx *ctx = notif->ctx;
+	struct mmpin *mmp = &notif->uarg.mmp;
 
+	if (mmp->user) {
+		atomic_long_sub(mmp->num_pg, &mmp->user->locked_vm);
+		free_uid(mmp->user);
+		mmp->user = NULL;
+	}
 	if (likely(notif->task)) {
 		io_put_task(notif->task, 1);
 		notif->task = NULL;
-- 
2.36.1

[PATCH net-next v3 19/25] io_uring: allow to pass addr into sendzc

[Pavel Begunkov]

Allow to specify an address to zerocopy sends making it more like
sendto(2).

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/uapi/linux/io_uring.h |  2 +-
 io_uring/net.c                | 17 ++++++++++++++++-
 2 files changed, 17 insertions(+), 2 deletions(-)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 0e1e179cec1d..abb8a9502f6e 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -64,7 +64,7 @@ struct io_uring_sqe {
 		__u32	file_index;
 		struct {
 			__u16	notification_idx;
-			__u16	__pad;
+			__u16	addr_len;
 		} __attribute__((packed));
 	};
 	union {
diff --git a/io_uring/net.c b/io_uring/net.c
index d5b00e07e72b..e63dda89c222 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -65,6 +65,8 @@ struct io_sendzc {
 	size_t				len;
 	u16				slot_idx;
 	int				msg_flags;
+	int				addr_len;
+	void __user			*addr;
 };
 
 #define IO_APOLL_MULTI_POLLED (REQ_F_APOLL_MULTISHOT | REQ_F_POLLED)
@@ -784,7 +786,7 @@ int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 {
 	struct io_sendzc *zc = io_kiocb_to_cmd(req);
 
-	if (READ_ONCE(sqe->ioprio) || READ_ONCE(sqe->addr2) || READ_ONCE(sqe->__pad2[0]))
+	if (READ_ONCE(sqe->ioprio) || READ_ONCE(sqe->__pad2[0]))
 		return -EINVAL;
 
 	zc->buf = u64_to_user_ptr(READ_ONCE(sqe->addr));
@@ -793,6 +795,10 @@ int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	zc->slot_idx = READ_ONCE(sqe->notification_idx);
 	if (zc->msg_flags & MSG_DONTWAIT)
 		req->flags |= REQ_F_NOWAIT;
+
+	zc->addr = u64_to_user_ptr(READ_ONCE(sqe->addr2));
+	zc->addr_len = READ_ONCE(sqe->addr_len);
+
 #ifdef CONFIG_COMPAT
 	if (req->ctx->compat)
 		zc->msg_flags |= MSG_CMSG_COMPAT;
@@ -802,6 +808,7 @@ int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 
 int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
 {
+	struct sockaddr_storage address;
 	struct io_ring_ctx *ctx = req->ctx;
 	struct io_sendzc *zc = io_kiocb_to_cmd(req);
 	struct io_notif_slot *notif_slot;
@@ -836,6 +843,14 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
 		return ret;
 	mm_account_pinned_pages(&notif->uarg.mmp, zc->len);
 
+	if (zc->addr) {
+		ret = move_addr_to_kernel(zc->addr, zc->addr_len, &address);
+		if (unlikely(ret < 0))
+			return ret;
+		msg.msg_name = (struct sockaddr *)&address;
+		msg.msg_namelen = zc->addr_len;
+	}
+
 	msg_flags = zc->msg_flags | MSG_ZEROCOPY;
 	if (issue_flags & IO_URING_F_NONBLOCK)
 		msg_flags |= MSG_DONTWAIT;
-- 
2.36.1

[PATCH net-next v3 20/25] io_uring: add rsrc referencing for notifiers

[Pavel Begunkov]

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>
---
 io_uring/notif.c |  5 +++++
 io_uring/notif.h |  1 +
 io_uring/rsrc.h  | 12 +++++++++---
 3 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/io_uring/notif.c b/io_uring/notif.c
index 0a03d04c010b..a53acdda9ec0 100644
--- a/io_uring/notif.c
+++ b/io_uring/notif.c
@@ -7,10 +7,12 @@
 
 #include "io_uring.h"
 #include "notif.h"
+#include "rsrc.h"
 
 static void __io_notif_complete_tw(struct callback_head *cb)
 {
 	struct io_notif *notif = container_of(cb, struct io_notif, task_work);
+	struct io_rsrc_node *rsrc_node = notif->rsrc_node;
 	struct io_ring_ctx *ctx = notif->ctx;
 	struct mmpin *mmp = &notif->uarg.mmp;
 
@@ -31,6 +33,7 @@ static void __io_notif_complete_tw(struct callback_head *cb)
 	ctx->notif_locked_nr++;
 	io_cq_unlock_post(ctx);
 
+	io_rsrc_put_node(rsrc_node, 1);
 	percpu_ref_put(&ctx->refs);
 }
 
@@ -125,6 +128,8 @@ struct io_notif *io_alloc_notif(struct io_ring_ctx *ctx,
 	/* master ref owned by io_notif_slot, will be dropped on flush */
 	refcount_set(&notif->uarg.refcnt, 1);
 	percpu_ref_get(&ctx->refs);
+	notif->rsrc_node = ctx->rsrc_node;
+	io_charge_rsrc_node(ctx);
 	return notif;
 }
 
diff --git a/io_uring/notif.h b/io_uring/notif.h
index 6dde39c6afbe..00efe164bdc4 100644
--- a/io_uring/notif.h
+++ b/io_uring/notif.h
@@ -11,6 +11,7 @@
 struct io_notif {
 	struct ubuf_info	uarg;
 	struct io_ring_ctx	*ctx;
+	struct io_rsrc_node	*rsrc_node;
 
 	/* complete via tw if ->task is non-NULL, fallback to wq otherwise */
 	struct task_struct	*task;
diff --git a/io_uring/rsrc.h b/io_uring/rsrc.h
index 87f58315b247..af342fd239d0 100644
--- a/io_uring/rsrc.h
+++ b/io_uring/rsrc.h
@@ -135,6 +135,13 @@ static inline void io_req_put_rsrc_locked(struct io_kiocb *req,
 	}
 }
 
+static inline void io_charge_rsrc_node(struct io_ring_ctx *ctx)
+{
+	ctx->rsrc_cached_refs--;
+	if (unlikely(ctx->rsrc_cached_refs < 0))
+		io_rsrc_refs_refill(ctx);
+}
+
 static inline void io_req_set_rsrc_node(struct io_kiocb *req,
 					struct io_ring_ctx *ctx,
 					unsigned int issue_flags)
@@ -144,9 +151,8 @@ static inline void io_req_set_rsrc_node(struct io_kiocb *req,
 
 		if (!(issue_flags & IO_URING_F_UNLOCKED)) {
 			lockdep_assert_held(&ctx->uring_lock);
-			ctx->rsrc_cached_refs--;
-			if (unlikely(ctx->rsrc_cached_refs < 0))
-				io_rsrc_refs_refill(ctx);
+
+			io_charge_rsrc_node(ctx);
 		} else {
 			percpu_ref_get(&req->rsrc_node->refs);
 		}
-- 
2.36.1

[PATCH net-next v3 21/25] io_uring: sendzc with fixed buffers

[Pavel Begunkov]

Allow zerocopy sends to use fixed buffers. There is an optimisation for
this case, the network layer don't need to reference the pages, see
SKBFL_MANAGED_FRAG_REFS, so io_uring have to ensure validity of fixed
buffers until the notifier is released.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/uapi/linux/io_uring.h |  7 ++++++
 io_uring/net.c                | 40 +++++++++++++++++++++++++++++------
 2 files changed, 41 insertions(+), 6 deletions(-)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index abb8a9502f6e..2509e6184bc7 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -272,6 +272,13 @@ enum io_uring_op {
  */
 #define IORING_ACCEPT_MULTISHOT	(1U << 0)
 
+/*
+ * IORING_OP_SENDZC flags
+ */
+enum {
+	IORING_SENDZC_FIXED_BUF		= (1U << 0),
+};
+
 /*
  * IO completion data structure (Completion Queue Entry)
  */
diff --git a/io_uring/net.c b/io_uring/net.c
index e63dda89c222..3dfe07749b04 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -14,6 +14,7 @@
 #include "kbuf.h"
 #include "net.h"
 #include "notif.h"
+#include "rsrc.h"
 
 #if defined(CONFIG_NET)
 struct io_shutdown {
@@ -65,6 +66,7 @@ struct io_sendzc {
 	size_t				len;
 	u16				slot_idx;
 	int				msg_flags;
+	unsigned			zc_flags;
 	int				addr_len;
 	void __user			*addr;
 };
@@ -782,11 +784,14 @@ int io_connect(struct io_kiocb *req, unsigned int issue_flags)
 	return IOU_OK;
 }
 
+#define IO_SENDZC_VALID_FLAGS IORING_SENDZC_FIXED_BUF
+
 int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 {
 	struct io_sendzc *zc = io_kiocb_to_cmd(req);
+	struct io_ring_ctx *ctx = req->ctx;
 
-	if (READ_ONCE(sqe->ioprio) || READ_ONCE(sqe->__pad2[0]))
+	if (READ_ONCE(sqe->__pad2[0]))
 		return -EINVAL;
 
 	zc->buf = u64_to_user_ptr(READ_ONCE(sqe->addr));
@@ -799,6 +804,20 @@ int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	zc->addr = u64_to_user_ptr(READ_ONCE(sqe->addr2));
 	zc->addr_len = READ_ONCE(sqe->addr_len);
 
+	zc->zc_flags = READ_ONCE(sqe->ioprio);
+	if (zc->zc_flags & ~IO_SENDZC_VALID_FLAGS)
+		return -EINVAL;
+
+	if (zc->zc_flags & IORING_SENDZC_FIXED_BUF) {
+		unsigned idx = READ_ONCE(sqe->buf_index);
+
+		if (unlikely(idx >= ctx->nr_user_bufs))
+			return -EFAULT;
+		idx = array_index_nospec(idx, ctx->nr_user_bufs);
+		req->imu = READ_ONCE(ctx->user_bufs[idx]);
+		io_req_set_rsrc_node(req, ctx, 0);
+	}
+
 #ifdef CONFIG_COMPAT
 	if (req->ctx->compat)
 		zc->msg_flags |= MSG_CMSG_COMPAT;
@@ -836,12 +855,21 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
 	msg.msg_control = NULL;
 	msg.msg_controllen = 0;
 	msg.msg_namelen = 0;
-	msg.msg_managed_data = 0;
+	msg.msg_managed_data = 1;
 
-	ret = import_single_range(WRITE, zc->buf, zc->len, &iov, &msg.msg_iter);
-	if (unlikely(ret))
-		return ret;
-	mm_account_pinned_pages(&notif->uarg.mmp, zc->len);
+	if (zc->zc_flags & IORING_SENDZC_FIXED_BUF) {
+		ret = io_import_fixed(WRITE, &msg.msg_iter, req->imu,
+					(u64)zc->buf, zc->len);
+		if (unlikely(ret))
+				return ret;
+	} else {
+		msg.msg_managed_data = 0;
+		ret = import_single_range(WRITE, zc->buf, zc->len, &iov,
+					  &msg.msg_iter);
+		if (unlikely(ret))
+			return ret;
+		mm_account_pinned_pages(&notif->uarg.mmp, zc->len);
+	}
 
 	if (zc->addr) {
 		ret = move_addr_to_kernel(zc->addr, zc->addr_len, &address);
-- 
2.36.1

[PATCH net-next v3 22/25] io_uring: flush notifiers after sendzc

[Pavel Begunkov]

Allow to flush notifiers as a part of sendzc request by setting
IORING_SENDZC_FLUSH flag. When the sendzc request succeedes it will
flush the used [active] notifier.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/uapi/linux/io_uring.h |  1 +
 io_uring/io_uring.c           | 11 +----------
 io_uring/io_uring.h           | 10 ++++++++++
 io_uring/net.c                |  4 +++-
 io_uring/notif.c              |  2 +-
 io_uring/notif.h              | 11 +++++++++++
 6 files changed, 27 insertions(+), 12 deletions(-)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 2509e6184bc7..2fd4e39a14d3 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -277,6 +277,7 @@ enum io_uring_op {
  */
 enum {
 	IORING_SENDZC_FIXED_BUF		= (1U << 0),
+	IORING_SENDZC_FLUSH		= (1U << 1),
 };
 
 /*
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 3b885d65e569..8f4152f01989 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -615,7 +615,7 @@ void __io_put_task(struct task_struct *task, int nr)
 	put_task_struct_many(task, nr);
 }
 
-static void io_task_refs_refill(struct io_uring_task *tctx)
+void io_task_refs_refill(struct io_uring_task *tctx)
 {
 	unsigned int refill = -tctx->cached_refs + IO_TCTX_REFS_CACHE_NR;
 
@@ -624,15 +624,6 @@ static void io_task_refs_refill(struct io_uring_task *tctx)
 	tctx->cached_refs += refill;
 }
 
-static inline void io_get_task_refs(int nr)
-{
-	struct io_uring_task *tctx = current->io_uring;
-
-	tctx->cached_refs -= nr;
-	if (unlikely(tctx->cached_refs < 0))
-		io_task_refs_refill(tctx);
-}
-
 static __cold void io_uring_drop_tctx_refs(struct task_struct *task)
 {
 	struct io_uring_task *tctx = task->io_uring;
diff --git a/io_uring/io_uring.h b/io_uring/io_uring.h
index e978654d1b14..cf154e9c8e28 100644
--- a/io_uring/io_uring.h
+++ b/io_uring/io_uring.h
@@ -60,6 +60,7 @@ void io_wq_submit_work(struct io_wq_work *work);
 void io_free_req(struct io_kiocb *req);
 void io_queue_next(struct io_kiocb *req);
 void __io_put_task(struct task_struct *task, int nr);
+void io_task_refs_refill(struct io_uring_task *tctx);
 
 bool io_match_task_safe(struct io_kiocb *head, struct task_struct *task,
 			bool cancel_all);
@@ -254,4 +255,13 @@ static inline void io_put_task(struct task_struct *task, int nr)
 		__io_put_task(task, nr);
 }
 
+static inline void io_get_task_refs(int nr)
+{
+	struct io_uring_task *tctx = current->io_uring;
+
+	tctx->cached_refs -= nr;
+	if (unlikely(tctx->cached_refs < 0))
+		io_task_refs_refill(tctx);
+}
+
 #endif
diff --git a/io_uring/net.c b/io_uring/net.c
index 3dfe07749b04..3cd75d69fe70 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -784,7 +784,7 @@ int io_connect(struct io_kiocb *req, unsigned int issue_flags)
 	return IOU_OK;
 }
 
-#define IO_SENDZC_VALID_FLAGS IORING_SENDZC_FIXED_BUF
+#define IO_SENDZC_VALID_FLAGS (IORING_SENDZC_FIXED_BUF|IORING_SENDZC_FLUSH)
 
 int io_sendzc_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 {
@@ -895,6 +895,8 @@ int io_sendzc(struct io_kiocb *req, unsigned int issue_flags)
 		return ret == -ERESTARTSYS ? -EINTR : ret;
 	}
 
+	if (zc->zc_flags & IORING_SENDZC_FLUSH)
+		io_notif_slot_flush_submit(notif_slot, 0);
 	io_req_set_res(req, ret, 0);
 	return IOU_OK;
 }
diff --git a/io_uring/notif.c b/io_uring/notif.c
index a53acdda9ec0..847535d34c65 100644
--- a/io_uring/notif.c
+++ b/io_uring/notif.c
@@ -133,7 +133,7 @@ struct io_notif *io_alloc_notif(struct io_ring_ctx *ctx,
 	return notif;
 }
 
-static void io_notif_slot_flush(struct io_notif_slot *slot)
+void io_notif_slot_flush(struct io_notif_slot *slot)
 	__must_hold(&ctx->uring_lock)
 {
 	struct io_notif *notif = slot->notif;
diff --git a/io_uring/notif.h b/io_uring/notif.h
index 00efe164bdc4..6cd73d7b965b 100644
--- a/io_uring/notif.h
+++ b/io_uring/notif.h
@@ -54,6 +54,7 @@ int io_notif_register(struct io_ring_ctx *ctx,
 int io_notif_unregister(struct io_ring_ctx *ctx);
 void io_notif_cache_purge(struct io_ring_ctx *ctx);
 
+void io_notif_slot_flush(struct io_notif_slot *slot);
 struct io_notif *io_alloc_notif(struct io_ring_ctx *ctx,
 				struct io_notif_slot *slot);
 
@@ -74,3 +75,13 @@ static inline struct io_notif_slot *io_get_notif_slot(struct io_ring_ctx *ctx,
 	idx = array_index_nospec(idx, ctx->nr_notif_slots);
 	return &ctx->notif_slots[idx];
 }
+
+static inline void io_notif_slot_flush_submit(struct io_notif_slot *slot,
+					      unsigned int issue_flags)
+{
+	if (!(issue_flags & IO_URING_F_UNLOCKED)) {
+		slot->notif->task = current;
+		io_get_task_refs(1);
+	}
+	io_notif_slot_flush(slot);
+}
-- 
2.36.1

[PATCH net-next v3 23/25] io_uring: rename IORING_OP_FILES_UPDATE

[Pavel Begunkov]

IORING_OP_FILES_UPDATE will be a more generic opcode serving different
resource types, rename it into IORING_OP_RSRC_UPDATE and add subtype
handling.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/uapi/linux/io_uring.h | 12 +++++++++++-
 io_uring/opdef.c              |  9 +++++----
 io_uring/rsrc.c               | 17 +++++++++++++++--
 io_uring/rsrc.h               |  4 ++--
 4 files changed, 33 insertions(+), 9 deletions(-)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 2fd4e39a14d3..e62e61ceb494 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -170,7 +170,8 @@ enum io_uring_op {
 	IORING_OP_FALLOCATE,
 	IORING_OP_OPENAT,
 	IORING_OP_CLOSE,
-	IORING_OP_FILES_UPDATE,
+	IORING_OP_RSRC_UPDATE,
+	IORING_OP_FILES_UPDATE = IORING_OP_RSRC_UPDATE,
 	IORING_OP_STATX,
 	IORING_OP_READ,
 	IORING_OP_WRITE,
@@ -219,6 +220,7 @@ enum io_uring_op {
 #define IORING_TIMEOUT_ETIME_SUCCESS	(1U << 5)
 #define IORING_TIMEOUT_CLOCK_MASK	(IORING_TIMEOUT_BOOTTIME | IORING_TIMEOUT_REALTIME)
 #define IORING_TIMEOUT_UPDATE_MASK	(IORING_TIMEOUT_UPDATE | IORING_LINK_TIMEOUT_UPDATE)
+
 /*
  * sqe->splice_flags
  * extends splice(2) flags
@@ -272,6 +274,14 @@ enum io_uring_op {
  */
 #define IORING_ACCEPT_MULTISHOT	(1U << 0)
 
+
+/*
+ * IORING_OP_RSRC_UPDATE flags
+ */
+enum {
+	IORING_RSRC_UPDATE_FILES,
+};
+
 /*
  * IORING_OP_SENDZC flags
  */
diff --git a/io_uring/opdef.c b/io_uring/opdef.c
index 91d425b43174..431b73e9b378 100644
--- a/io_uring/opdef.c
+++ b/io_uring/opdef.c
@@ -244,12 +244,13 @@ const struct io_op_def io_op_defs[] = {
 		.prep			= io_close_prep,
 		.issue			= io_close,
 	},
-	[IORING_OP_FILES_UPDATE] = {
+	[IORING_OP_RSRC_UPDATE] = {
 		.audit_skip		= 1,
 		.iopoll			= 1,
-		.name			= "FILES_UPDATE",
-		.prep			= io_files_update_prep,
-		.issue			= io_files_update,
+		.name			= "RSRC_UPDATE",
+		.prep			= io_rsrc_update_prep,
+		.issue			= io_rsrc_update,
+		.ioprio			= 1,
 	},
 	[IORING_OP_STATX] = {
 		.audit_skip		= 1,
diff --git a/io_uring/rsrc.c b/io_uring/rsrc.c
index 3a2a5ef263f0..0c3f95f24cef 100644
--- a/io_uring/rsrc.c
+++ b/io_uring/rsrc.c
@@ -21,6 +21,7 @@ struct io_rsrc_update {
 	u64				arg;
 	u32				nr_args;
 	u32				offset;
+	int				type;
 };
 
 static int io_sqe_buffer_register(struct io_ring_ctx *ctx, struct iovec *iov,
@@ -658,7 +659,7 @@ __cold int io_register_rsrc(struct io_ring_ctx *ctx, void __user *arg,
 	return -EINVAL;
 }
 
-int io_files_update_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
+int io_rsrc_update_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 {
 	struct io_rsrc_update *up = io_kiocb_to_cmd(req);
 
@@ -672,6 +673,7 @@ int io_files_update_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	if (!up->nr_args)
 		return -EINVAL;
 	up->arg = READ_ONCE(sqe->addr);
+	up->type = READ_ONCE(sqe->ioprio);
 	return 0;
 }
 
@@ -711,7 +713,7 @@ static int io_files_update_with_index_alloc(struct io_kiocb *req,
 	return ret;
 }
 
-int io_files_update(struct io_kiocb *req, unsigned int issue_flags)
+static int io_files_update(struct io_kiocb *req, unsigned int issue_flags)
 {
 	struct io_rsrc_update *up = io_kiocb_to_cmd(req);
 	struct io_ring_ctx *ctx = req->ctx;
@@ -740,6 +742,17 @@ int io_files_update(struct io_kiocb *req, unsigned int issue_flags)
 	return IOU_OK;
 }
 
+int io_rsrc_update(struct io_kiocb *req, unsigned int issue_flags)
+{
+	struct io_rsrc_update *up = io_kiocb_to_cmd(req);
+
+	switch (up->type) {
+	case IORING_RSRC_UPDATE_FILES:
+		return io_files_update(req, issue_flags);
+	}
+	return -EINVAL;
+}
+
 int io_queue_rsrc_removal(struct io_rsrc_data *data, unsigned idx,
 			  struct io_rsrc_node *node, void *rsrc)
 {
diff --git a/io_uring/rsrc.h b/io_uring/rsrc.h
index af342fd239d0..21813a23215f 100644
--- a/io_uring/rsrc.h
+++ b/io_uring/rsrc.h
@@ -167,6 +167,6 @@ static inline u64 *io_get_tag_slot(struct io_rsrc_data *data, unsigned int idx)
 	return &data->tags[table_idx][off];
 }
 
-int io_files_update(struct io_kiocb *req, unsigned int issue_flags);
-int io_files_update_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe);
+int io_rsrc_update(struct io_kiocb *req, unsigned int issue_flags);
+int io_rsrc_update_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe);
 #endif
-- 
2.36.1

[PATCH net-next v3 24/25] io_uring: add zc notification flush requests

[Pavel Begunkov]

Overlay notification control onto IORING_OP_RSRC_UPDATE (former
IORING_OP_FILES_UPDATE). It allows to flush a range of zc notifications
from slots with indexes [sqe->off, sqe->off+sqe->len). If sqe->arg is
not zero, it also copies sqe->arg as a new tag for all flushed
notifications.

Note, it doesn't flush a notification of a slot if there was no requests
attached to it (since last flush or registration).

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 include/uapi/linux/io_uring.h |  1 +
 io_uring/rsrc.c               | 38 +++++++++++++++++++++++++++++++++++
 2 files changed, 39 insertions(+)

diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index e62e61ceb494..eeb0fbee19cb 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -280,6 +280,7 @@ enum io_uring_op {
  */
 enum {
 	IORING_RSRC_UPDATE_FILES,
+	IORING_RSRC_UPDATE_NOTIF,
 };
 
 /*
diff --git a/io_uring/rsrc.c b/io_uring/rsrc.c
index 0c3f95f24cef..af58d58dd21b 100644
--- a/io_uring/rsrc.c
+++ b/io_uring/rsrc.c
@@ -15,6 +15,7 @@
 #include "io_uring.h"
 #include "openclose.h"
 #include "rsrc.h"
+#include "notif.h"
 
 struct io_rsrc_update {
 	struct file			*file;
@@ -742,6 +743,41 @@ static int io_files_update(struct io_kiocb *req, unsigned int issue_flags)
 	return IOU_OK;
 }
 
+static int io_notif_update(struct io_kiocb *req, unsigned int issue_flags)
+{
+	struct io_rsrc_update *up = io_kiocb_to_cmd(req);
+	struct io_ring_ctx *ctx = req->ctx;
+	unsigned len = up->nr_args;
+	unsigned idx_end, idx = up->offset;
+	int ret = 0;
+
+	io_ring_submit_lock(ctx, issue_flags);
+	if (unlikely(check_add_overflow(idx, len, &idx_end))) {
+		ret = -EOVERFLOW;
+		goto out;
+	}
+	if (unlikely(idx_end > ctx->nr_notif_slots)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	for (; idx < idx_end; idx++) {
+		struct io_notif_slot *slot = &ctx->notif_slots[idx];
+
+		if (!slot->notif)
+			continue;
+		if (up->arg)
+			slot->tag = up->arg;
+		io_notif_slot_flush_submit(slot, issue_flags);
+	}
+out:
+	io_ring_submit_unlock(ctx, issue_flags);
+	if (ret < 0)
+		req_set_fail(req);
+	io_req_set_res(req, ret, 0);
+	return IOU_OK;
+}
+
 int io_rsrc_update(struct io_kiocb *req, unsigned int issue_flags)
 {
 	struct io_rsrc_update *up = io_kiocb_to_cmd(req);
@@ -749,6 +785,8 @@ int io_rsrc_update(struct io_kiocb *req, unsigned int issue_flags)
 	switch (up->type) {
 	case IORING_RSRC_UPDATE_FILES:
 		return io_files_update(req, issue_flags);
+	case IORING_RSRC_UPDATE_NOTIF:
+		return io_notif_update(req, issue_flags);
 	}
 	return -EINVAL;
 }
-- 
2.36.1

[PATCH net-next v3 25/25] selftests/io_uring: test zerocopy send

[Pavel Begunkov]

Add selftests for io_uring zerocopy sends and io_uring's notification
infrastructure. It's largely influenced by msg_zerocopy and uses it on
the receive side.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 tools/testing/selftests/net/Makefile          |   1 +
 .../selftests/net/io_uring_zerocopy_tx.c      | 605 ++++++++++++++++++
 .../selftests/net/io_uring_zerocopy_tx.sh     | 131 ++++
 3 files changed, 737 insertions(+)
 create mode 100644 tools/testing/selftests/net/io_uring_zerocopy_tx.c
 create mode 100755 tools/testing/selftests/net/io_uring_zerocopy_tx.sh

diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile
index 7ea54af55490..51261483744e 100644
--- a/tools/testing/selftests/net/Makefile
+++ b/tools/testing/selftests/net/Makefile
@@ -59,6 +59,7 @@ TEST_GEN_FILES += toeplitz
 TEST_GEN_FILES += cmsg_sender
 TEST_GEN_FILES += stress_reuseport_listen
 TEST_PROGS += test_vxlan_vnifiltering.sh
+TEST_GEN_FILES += io_uring_zerocopy_tx
 
 TEST_FILES := settings
 
diff --git a/tools/testing/selftests/net/io_uring_zerocopy_tx.c b/tools/testing/selftests/net/io_uring_zerocopy_tx.c
new file mode 100644
index 000000000000..899ddc84f8a9
--- /dev/null
+++ b/tools/testing/selftests/net/io_uring_zerocopy_tx.c
@@ -0,0 +1,605 @@
+/* SPDX-License-Identifier: MIT */
+/* based on linux-kernel/tools/testing/selftests/net/msg_zerocopy.c */
+#include <assert.h>
+#include <errno.h>
+#include <error.h>
+#include <fcntl.h>
+#include <limits.h>
+#include <stdbool.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+
+#include <arpa/inet.h>
+#include <linux/errqueue.h>
+#include <linux/if_packet.h>
+#include <linux/io_uring.h>
+#include <linux/ipv6.h>
+#include <linux/socket.h>
+#include <linux/sockios.h>
+#include <net/ethernet.h>
+#include <net/if.h>
+#include <netinet/in.h>
+#include <netinet/ip.h>
+#include <netinet/ip6.h>
+#include <netinet/tcp.h>
+#include <netinet/udp.h>
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+#include <sys/resource.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/time.h>
+#include <sys/types.h>
+#include <sys/un.h>
+#include <sys/wait.h>
+
+#define NOTIF_TAG 0xfffffffULL
+#define NONZC_TAG 0
+#define ZC_TAG 1
+
+enum {
+	MODE_NONZC	= 0,
+	MODE_ZC		= 1,
+	MODE_ZC_FIXED	= 2,
+	MODE_MIXED	= 3,
+};
+
+static bool cfg_flush		= false;
+static bool cfg_cork		= false;
+static int  cfg_mode		= MODE_ZC_FIXED;
+static int  cfg_nr_reqs		= 8;
+static int  cfg_family		= PF_UNSPEC;
+static int  cfg_payload_len;
+static int  cfg_port		= 8000;
+static int  cfg_runtime_ms	= 4200;
+
+static socklen_t cfg_alen;
+static struct sockaddr_storage cfg_dst_addr;
+
+static char payload[IP_MAXPACKET] __attribute__((aligned(4096)));
+
+struct io_sq_ring {
+	unsigned *head;
+	unsigned *tail;
+	unsigned *ring_mask;
+	unsigned *ring_entries;
+	unsigned *flags;
+	unsigned *array;
+};
+
+struct io_cq_ring {
+	unsigned *head;
+	unsigned *tail;
+	unsigned *ring_mask;
+	unsigned *ring_entries;
+	struct io_uring_cqe *cqes;
+};
+
+struct io_uring_sq {
+	unsigned *khead;
+	unsigned *ktail;
+	unsigned *kring_mask;
+	unsigned *kring_entries;
+	unsigned *kflags;
+	unsigned *kdropped;
+	unsigned *array;
+	struct io_uring_sqe *sqes;
+
+	unsigned sqe_head;
+	unsigned sqe_tail;
+
+	size_t ring_sz;
+};
+
+struct io_uring_cq {
+	unsigned *khead;
+	unsigned *ktail;
+	unsigned *kring_mask;
+	unsigned *kring_entries;
+	unsigned *koverflow;
+	struct io_uring_cqe *cqes;
+
+	size_t ring_sz;
+};
+
+struct io_uring {
+	struct io_uring_sq sq;
+	struct io_uring_cq cq;
+	int ring_fd;
+};
+
+#ifdef __alpha__
+# ifndef __NR_io_uring_setup
+#  define __NR_io_uring_setup		535
+# endif
+# ifndef __NR_io_uring_enter
+#  define __NR_io_uring_enter		536
+# endif
+# ifndef __NR_io_uring_register
+#  define __NR_io_uring_register	537
+# endif
+#else /* !__alpha__ */
+# ifndef __NR_io_uring_setup
+#  define __NR_io_uring_setup		425
+# endif
+# ifndef __NR_io_uring_enter
+#  define __NR_io_uring_enter		426
+# endif
+# ifndef __NR_io_uring_register
+#  define __NR_io_uring_register	427
+# endif
+#endif
+
+#if defined(__x86_64) || defined(__i386__)
+#define read_barrier()	__asm__ __volatile__("":::"memory")
+#define write_barrier()	__asm__ __volatile__("":::"memory")
+#else
+
+#define read_barrier()	__sync_synchronize()
+#define write_barrier()	__sync_synchronize()
+#endif
+
+static int io_uring_setup(unsigned int entries, struct io_uring_params *p)
+{
+	return syscall(__NR_io_uring_setup, entries, p);
+}
+
+static int io_uring_enter(int fd, unsigned int to_submit,
+			  unsigned int min_complete,
+			  unsigned int flags, sigset_t *sig)
+{
+	return syscall(__NR_io_uring_enter, fd, to_submit, min_complete,
+			flags, sig, _NSIG / 8);
+}
+
+static int io_uring_register_buffers(struct io_uring *ring,
+				     const struct iovec *iovecs,
+				     unsigned nr_iovecs)
+{
+	int ret;
+
+	ret = syscall(__NR_io_uring_register, ring->ring_fd,
+		      IORING_REGISTER_BUFFERS, iovecs, nr_iovecs);
+	return (ret < 0) ? -errno : ret;
+}
+
+static int io_uring_register_notifications(struct io_uring *ring,
+					   unsigned nr,
+					   struct io_uring_notification_slot *slots)
+{
+	int ret;
+	struct io_uring_notification_register r = {
+		.nr_slots = nr,
+		.data = (unsigned long)slots,
+	};
+
+	ret = syscall(__NR_io_uring_register, ring->ring_fd,
+		      IORING_REGISTER_NOTIFIERS, &r, sizeof(r));
+	return (ret < 0) ? -errno : ret;
+}
+
+static int io_uring_mmap(int fd, struct io_uring_params *p,
+			 struct io_uring_sq *sq, struct io_uring_cq *cq)
+{
+	size_t size;
+	void *ptr;
+	int ret;
+
+	sq->ring_sz = p->sq_off.array + p->sq_entries * sizeof(unsigned);
+	ptr = mmap(0, sq->ring_sz, PROT_READ | PROT_WRITE,
+		   MAP_SHARED | MAP_POPULATE, fd, IORING_OFF_SQ_RING);
+	if (ptr == MAP_FAILED)
+		return -errno;
+	sq->khead = ptr + p->sq_off.head;
+	sq->ktail = ptr + p->sq_off.tail;
+	sq->kring_mask = ptr + p->sq_off.ring_mask;
+	sq->kring_entries = ptr + p->sq_off.ring_entries;
+	sq->kflags = ptr + p->sq_off.flags;
+	sq->kdropped = ptr + p->sq_off.dropped;
+	sq->array = ptr + p->sq_off.array;
+
+	size = p->sq_entries * sizeof(struct io_uring_sqe);
+	sq->sqes = mmap(0, size, PROT_READ | PROT_WRITE,
+			MAP_SHARED | MAP_POPULATE, fd, IORING_OFF_SQES);
+	if (sq->sqes == MAP_FAILED) {
+		ret = -errno;
+err:
+		munmap(sq->khead, sq->ring_sz);
+		return ret;
+	}
+
+	cq->ring_sz = p->cq_off.cqes + p->cq_entries * sizeof(struct io_uring_cqe);
+	ptr = mmap(0, cq->ring_sz, PROT_READ | PROT_WRITE,
+			MAP_SHARED | MAP_POPULATE, fd, IORING_OFF_CQ_RING);
+	if (ptr == MAP_FAILED) {
+		ret = -errno;
+		munmap(sq->sqes, p->sq_entries * sizeof(struct io_uring_sqe));
+		goto err;
+	}
+	cq->khead = ptr + p->cq_off.head;
+	cq->ktail = ptr + p->cq_off.tail;
+	cq->kring_mask = ptr + p->cq_off.ring_mask;
+	cq->kring_entries = ptr + p->cq_off.ring_entries;
+	cq->koverflow = ptr + p->cq_off.overflow;
+	cq->cqes = ptr + p->cq_off.cqes;
+	return 0;
+}
+
+static int io_uring_queue_init(unsigned entries, struct io_uring *ring,
+			       unsigned flags)
+{
+	struct io_uring_params p;
+	int fd, ret;
+
+	memset(ring, 0, sizeof(*ring));
+	memset(&p, 0, sizeof(p));
+	p.flags = flags;
+
+	fd = io_uring_setup(entries, &p);
+	if (fd < 0)
+		return fd;
+	ret = io_uring_mmap(fd, &p, &ring->sq, &ring->cq);
+	if (!ret)
+		ring->ring_fd = fd;
+	else
+		close(fd);
+	return ret;
+}
+
+static int io_uring_submit(struct io_uring *ring)
+{
+	struct io_uring_sq *sq = &ring->sq;
+	const unsigned mask = *sq->kring_mask;
+	unsigned ktail, submitted, to_submit;
+	int ret;
+
+	read_barrier();
+	if (*sq->khead != *sq->ktail) {
+		submitted = *sq->kring_entries;
+		goto submit;
+	}
+	if (sq->sqe_head == sq->sqe_tail)
+		return 0;
+
+	ktail = *sq->ktail;
+	to_submit = sq->sqe_tail - sq->sqe_head;
+	for (submitted = 0; submitted < to_submit; submitted++) {
+		read_barrier();
+		sq->array[ktail++ & mask] = sq->sqe_head++ & mask;
+	}
+	if (!submitted)
+		return 0;
+
+	if (*sq->ktail != ktail) {
+		write_barrier();
+		*sq->ktail = ktail;
+		write_barrier();
+	}
+submit:
+	ret = io_uring_enter(ring->ring_fd, submitted, 0,
+				IORING_ENTER_GETEVENTS, NULL);
+	return ret < 0 ? -errno : ret;
+}
+
+static inline void io_uring_prep_send(struct io_uring_sqe *sqe, int sockfd,
+				      const void *buf, size_t len, int flags)
+{
+	memset(sqe, 0, sizeof(*sqe));
+	sqe->opcode = (__u8) IORING_OP_SEND;
+	sqe->fd = sockfd;
+	sqe->addr = (unsigned long) buf;
+	sqe->len = len;
+	sqe->msg_flags = (__u32) flags;
+}
+
+static inline void io_uring_prep_sendzc(struct io_uring_sqe *sqe, int sockfd,
+				        const void *buf, size_t len, int flags,
+				        unsigned slot_idx, unsigned zc_flags)
+{
+	io_uring_prep_send(sqe, sockfd, buf, len, flags);
+	sqe->opcode = (__u8) IORING_OP_SENDZC;
+	sqe->notification_idx = slot_idx;
+	sqe->ioprio = zc_flags;
+}
+
+static struct io_uring_sqe *io_uring_get_sqe(struct io_uring *ring)
+{
+	struct io_uring_sq *sq = &ring->sq;
+
+	if (sq->sqe_tail + 1 - sq->sqe_head > *sq->kring_entries)
+		return NULL;
+	return &sq->sqes[sq->sqe_tail++ & *sq->kring_mask];
+}
+
+static int io_uring_wait_cqe(struct io_uring *ring, struct io_uring_cqe **cqe_ptr)
+{
+	struct io_uring_cq *cq = &ring->cq;
+	const unsigned mask = *cq->kring_mask;
+	unsigned head = *cq->khead;
+	int ret;
+
+	*cqe_ptr = NULL;
+	do {
+		read_barrier();
+		if (head != *cq->ktail) {
+			*cqe_ptr = &cq->cqes[head & mask];
+			break;
+		}
+		ret = io_uring_enter(ring->ring_fd, 0, 1,
+					IORING_ENTER_GETEVENTS, NULL);
+		if (ret < 0)
+			return -errno;
+	} while (1);
+
+	return 0;
+}
+
+static inline void io_uring_cqe_seen(struct io_uring *ring)
+{
+	*(&ring->cq)->khead += 1;
+	write_barrier();
+}
+
+static unsigned long gettimeofday_ms(void)
+{
+	struct timeval tv;
+
+	gettimeofday(&tv, NULL);
+	return (tv.tv_sec * 1000) + (tv.tv_usec / 1000);
+}
+
+static void do_setsockopt(int fd, int level, int optname, int val)
+{
+	if (setsockopt(fd, level, optname, &val, sizeof(val)))
+		error(1, errno, "setsockopt %d.%d: %d", level, optname, val);
+}
+
+static int do_setup_tx(int domain, int type, int protocol)
+{
+	int fd;
+
+	fd = socket(domain, type, protocol);
+	if (fd == -1)
+		error(1, errno, "socket t");
+
+	do_setsockopt(fd, SOL_SOCKET, SO_SNDBUF, 1 << 21);
+
+	if (connect(fd, (void *) &cfg_dst_addr, cfg_alen))
+		error(1, errno, "connect");
+	return fd;
+}
+
+static void do_tx(int domain, int type, int protocol)
+{
+	struct io_uring_notification_slot b[1] = {{.tag = NOTIF_TAG}};
+	struct io_uring_sqe *sqe;
+	struct io_uring_cqe *cqe;
+	unsigned long packets = 0, bytes = 0;
+	struct io_uring ring;
+	struct iovec iov;
+	uint64_t tstop;
+	int i, fd, ret;
+	int compl_cqes = 0;
+
+	fd = do_setup_tx(domain, type, protocol);
+
+	ret = io_uring_queue_init(512, &ring, 0);
+	if (ret)
+		error(1, ret, "io_uring: queue init");
+
+	ret = io_uring_register_notifications(&ring, 1, b);
+	if (ret)
+		error(1, ret, "io_uring: tx ctx registration");
+
+	iov.iov_base = payload;
+	iov.iov_len = cfg_payload_len;
+
+	ret = io_uring_register_buffers(&ring, &iov, 1);
+	if (ret)
+		error(1, ret, "io_uring: buffer registration");
+
+	tstop = gettimeofday_ms() + cfg_runtime_ms;
+	do {
+		if (cfg_cork)
+			do_setsockopt(fd, IPPROTO_UDP, UDP_CORK, 1);
+
+		for (i = 0; i < cfg_nr_reqs; i++) {
+			unsigned zc_flags = 0;
+			unsigned buf_idx = 0;
+			unsigned slot_idx = 0;
+			unsigned mode = cfg_mode;
+			unsigned msg_flags = 0;
+
+			if (cfg_mode == MODE_MIXED)
+				mode = rand() % 3;
+
+			sqe = io_uring_get_sqe(&ring);
+
+			if (mode == MODE_NONZC) {
+				io_uring_prep_send(sqe, fd, payload,
+						   cfg_payload_len, msg_flags);
+				sqe->user_data = NONZC_TAG;
+			} else {
+				if (cfg_flush) {
+					zc_flags |= IORING_SENDZC_FLUSH;
+					compl_cqes++;
+				}
+				io_uring_prep_sendzc(sqe, fd, payload,
+						     cfg_payload_len,
+						     msg_flags, slot_idx, zc_flags);
+				if (mode == MODE_ZC_FIXED) {
+					sqe->ioprio |= IORING_SENDZC_FIXED_BUF;
+					sqe->buf_index = buf_idx;
+				}
+				sqe->user_data = ZC_TAG;
+			}
+		}
+
+		ret = io_uring_submit(&ring);
+		if (ret != cfg_nr_reqs)
+			error(1, ret, "submit");
+
+		for (i = 0; i < cfg_nr_reqs; i++) {
+			ret = io_uring_wait_cqe(&ring, &cqe);
+			if (ret)
+				error(1, ret, "wait cqe");
+
+			if (cqe->user_data == NOTIF_TAG) {
+				compl_cqes--;
+				i--;
+			} else if (cqe->user_data != NONZC_TAG &&
+				   cqe->user_data != ZC_TAG) {
+				error(1, cqe->res, "invalid user_data");
+			} else if (cqe->res <= 0 && cqe->res != -EAGAIN) {
+				error(1, cqe->res, "send failed");
+			} else {
+				if (cqe->res > 0) {
+					packets++;
+					bytes += cqe->res;
+				}
+				/* failed requests don't flush */
+				if (cfg_flush &&
+				    cqe->res <= 0 &&
+				    cqe->user_data == ZC_TAG)
+					compl_cqes--;
+			}
+			io_uring_cqe_seen(&ring);
+		}
+		if (cfg_cork)
+			do_setsockopt(fd, IPPROTO_UDP, UDP_CORK, 0);
+	} while (gettimeofday_ms() < tstop);
+
+	if (close(fd))
+		error(1, errno, "close");
+
+	fprintf(stderr, "tx=%lu (MB=%lu), tx/s=%lu (MB/s=%lu)\n",
+			packets, bytes >> 20,
+			packets / (cfg_runtime_ms / 1000),
+			(bytes >> 20) / (cfg_runtime_ms / 1000));
+
+	while (compl_cqes) {
+		ret = io_uring_wait_cqe(&ring, &cqe);
+		if (ret)
+			error(1, ret, "wait cqe");
+		io_uring_cqe_seen(&ring);
+		compl_cqes--;
+	}
+}
+
+static void do_test(int domain, int type, int protocol)
+{
+	int i;
+
+	for (i = 0; i < IP_MAXPACKET; i++)
+		payload[i] = 'a' + (i % 26);
+	do_tx(domain, type, protocol);
+}
+
+static void usage(const char *filepath)
+{
+	error(1, 0, "Usage: %s [-f] [-n<N>] [-z0] [-s<payload size>] "
+		    "(-4|-6) [-t<time s>] -D<dst_ip> udp", filepath);
+}
+
+static void parse_opts(int argc, char **argv)
+{
+	const int max_payload_len = sizeof(payload) -
+				    sizeof(struct ipv6hdr) -
+				    sizeof(struct tcphdr) -
+				    40 /* max tcp options */;
+	struct sockaddr_in6 *addr6 = (void *) &cfg_dst_addr;
+	struct sockaddr_in *addr4 = (void *) &cfg_dst_addr;
+	char *daddr = NULL;
+	int c;
+
+	if (argc <= 1)
+		usage(argv[0]);
+	cfg_payload_len = max_payload_len;
+
+	while ((c = getopt(argc, argv, "46D:p:s:t:n:fc:m:")) != -1) {
+		switch (c) {
+		case '4':
+			if (cfg_family != PF_UNSPEC)
+				error(1, 0, "Pass one of -4 or -6");
+			cfg_family = PF_INET;
+			cfg_alen = sizeof(struct sockaddr_in);
+			break;
+		case '6':
+			if (cfg_family != PF_UNSPEC)
+				error(1, 0, "Pass one of -4 or -6");
+			cfg_family = PF_INET6;
+			cfg_alen = sizeof(struct sockaddr_in6);
+			break;
+		case 'D':
+			daddr = optarg;
+			break;
+		case 'p':
+			cfg_port = strtoul(optarg, NULL, 0);
+			break;
+		case 's':
+			cfg_payload_len = strtoul(optarg, NULL, 0);
+			break;
+		case 't':
+			cfg_runtime_ms = 200 + strtoul(optarg, NULL, 10) * 1000;
+			break;
+		case 'n':
+			cfg_nr_reqs = strtoul(optarg, NULL, 0);
+			break;
+		case 'f':
+			cfg_flush = 1;
+			break;
+		case 'c':
+			cfg_cork = strtol(optarg, NULL, 0);
+			break;
+		case 'm':
+			cfg_mode = strtol(optarg, NULL, 0);
+			break;
+		}
+	}
+
+	switch (cfg_family) {
+	case PF_INET:
+		memset(addr4, 0, sizeof(*addr4));
+		addr4->sin_family = AF_INET;
+		addr4->sin_port = htons(cfg_port);
+		if (daddr &&
+		    inet_pton(AF_INET, daddr, &(addr4->sin_addr)) != 1)
+			error(1, 0, "ipv4 parse error: %s", daddr);
+		break;
+	case PF_INET6:
+		memset(addr6, 0, sizeof(*addr6));
+		addr6->sin6_family = AF_INET6;
+		addr6->sin6_port = htons(cfg_port);
+		if (daddr &&
+		    inet_pton(AF_INET6, daddr, &(addr6->sin6_addr)) != 1)
+			error(1, 0, "ipv6 parse error: %s", daddr);
+		break;
+	default:
+		error(1, 0, "illegal domain");
+	}
+
+	if (cfg_payload_len > max_payload_len)
+		error(1, 0, "-s: payload exceeds max (%d)", max_payload_len);
+	if (cfg_mode == MODE_NONZC && cfg_flush)
+		error(1, 0, "-f: only zerocopy modes support notifications");
+	if (optind != argc - 1)
+		usage(argv[0]);
+}
+
+int main(int argc, char **argv)
+{
+	const char *cfg_test = argv[argc - 1];
+
+	parse_opts(argc, argv);
+
+	if (!strcmp(cfg_test, "tcp"))
+		do_test(cfg_family, SOCK_STREAM, 0);
+	else if (!strcmp(cfg_test, "udp"))
+		do_test(cfg_family, SOCK_DGRAM, 0);
+	else
+		error(1, 0, "unknown cfg_test %s", cfg_test);
+	return 0;
+}
diff --git a/tools/testing/selftests/net/io_uring_zerocopy_tx.sh b/tools/testing/selftests/net/io_uring_zerocopy_tx.sh
new file mode 100755
index 000000000000..6a65e4437640
--- /dev/null
+++ b/tools/testing/selftests/net/io_uring_zerocopy_tx.sh
@@ -0,0 +1,131 @@
+#!/bin/bash
+#
+# Send data between two processes across namespaces
+# Run twice: once without and once with zerocopy
+
+set -e
+
+readonly DEV="veth0"
+readonly DEV_MTU=65535
+readonly BIN_TX="./io_uring_zerocopy_tx"
+readonly BIN_RX="./msg_zerocopy"
+
+readonly RAND="$(mktemp -u XXXXXX)"
+readonly NSPREFIX="ns-${RAND}"
+readonly NS1="${NSPREFIX}1"
+readonly NS2="${NSPREFIX}2"
+
+readonly SADDR4='192.168.1.1'
+readonly DADDR4='192.168.1.2'
+readonly SADDR6='fd::1'
+readonly DADDR6='fd::2'
+
+readonly path_sysctl_mem="net.core.optmem_max"
+
+# No arguments: automated test
+if [[ "$#" -eq "0" ]]; then
+	IPs=( "4" "6" )
+	protocols=( "tcp" "udp" )
+
+	for IP in "${IPs[@]}"; do
+		for proto in "${protocols[@]}"; do
+			for mode in $(seq 1 3); do
+				$0 "$IP" "$proto" -m "$mode" -t 1 -n 32
+				$0 "$IP" "$proto" -m "$mode" -t 1 -n 32 -f
+				$0 "$IP" "$proto" -m "$mode" -t 1 -n 32 -c -f
+			done
+		done
+	done
+
+	echo "OK. All tests passed"
+	exit 0
+fi
+
+# Argument parsing
+if [[ "$#" -lt "2" ]]; then
+	echo "Usage: $0 [4|6] [tcp|udp|raw|raw_hdrincl|packet|packet_dgram] <args>"
+	exit 1
+fi
+
+readonly IP="$1"
+shift
+readonly TXMODE="$1"
+shift
+readonly EXTRA_ARGS="$@"
+
+# Argument parsing: configure addresses
+if [[ "${IP}" == "4" ]]; then
+	readonly SADDR="${SADDR4}"
+	readonly DADDR="${DADDR4}"
+elif [[ "${IP}" == "6" ]]; then
+	readonly SADDR="${SADDR6}"
+	readonly DADDR="${DADDR6}"
+else
+	echo "Invalid IP version ${IP}"
+	exit 1
+fi
+
+# Argument parsing: select receive mode
+#
+# This differs from send mode for
+# - packet:	use raw recv, because packet receives skb clones
+# - raw_hdrinc: use raw recv, because hdrincl is a tx-only option
+case "${TXMODE}" in
+'packet' | 'packet_dgram' | 'raw_hdrincl')
+	RXMODE='raw'
+	;;
+*)
+	RXMODE="${TXMODE}"
+	;;
+esac
+
+# Start of state changes: install cleanup handler
+save_sysctl_mem="$(sysctl -n ${path_sysctl_mem})"
+
+cleanup() {
+	ip netns del "${NS2}"
+	ip netns del "${NS1}"
+	sysctl -w -q "${path_sysctl_mem}=${save_sysctl_mem}"
+}
+
+trap cleanup EXIT
+
+# Configure system settings
+sysctl -w -q "${path_sysctl_mem}=1000000"
+
+# Create virtual ethernet pair between network namespaces
+ip netns add "${NS1}"
+ip netns add "${NS2}"
+
+ip link add "${DEV}" mtu "${DEV_MTU}" netns "${NS1}" type veth \
+  peer name "${DEV}" mtu "${DEV_MTU}" netns "${NS2}"
+
+# Bring the devices up
+ip -netns "${NS1}" link set "${DEV}" up
+ip -netns "${NS2}" link set "${DEV}" up
+
+# Set fixed MAC addresses on the devices
+ip -netns "${NS1}" link set dev "${DEV}" address 02:02:02:02:02:02
+ip -netns "${NS2}" link set dev "${DEV}" address 06:06:06:06:06:06
+
+# Add fixed IP addresses to the devices
+ip -netns "${NS1}" addr add 192.168.1.1/24 dev "${DEV}"
+ip -netns "${NS2}" addr add 192.168.1.2/24 dev "${DEV}"
+ip -netns "${NS1}" addr add       fd::1/64 dev "${DEV}" nodad
+ip -netns "${NS2}" addr add       fd::2/64 dev "${DEV}" nodad
+
+# Optionally disable sg or csum offload to test edge cases
+# ip netns exec "${NS1}" ethtool -K "${DEV}" sg off
+
+do_test() {
+	local readonly ARGS="$1"
+
+	echo "ipv${IP} ${TXMODE} ${ARGS}"
+	ip netns exec "${NS2}" "${BIN_RX}" "-${IP}" -t 2 -C 2 -S "${SADDR}" -D "${DADDR}" -r "${RXMODE}" &
+	sleep 0.2
+	ip netns exec "${NS1}" "${BIN_TX}" "-${IP}" -t 1 -D "${DADDR}" ${ARGS} "${TXMODE}"
+	wait
+}
+
+do_test "${EXTRA_ARGS}"
+echo ok
-- 
2.36.1

Re: [PATCH net-next v3 00/25] io_uring zerocopy send

[Pavel Begunkov]

On 7/5/22 16:01, Pavel Begunkov wrote:

NOTE: This is not be picked directly due to cross-subsystem merge problems.
After finding a consensus and getting necessary acks, I'll work out merging
with Jakub and Jens.


> The patchset implements io_uring zerocopy send. It works with both registered
> and normal buffers, mixing is allowed but not recommended. Apart from usual
> request completions, just as with MSG_ZEROCOPY, io_uring separately notifies
> the userspace when buffers are freed and can be reused (see API design below),
> which is delivered into io_uring's Completion Queue. Those "buffer-free"
> notifications are not necessarily per request, but the userspace has control
> over it and should explicitly attaching a number of requests to a single
> notification. The series also adds some internal optimisations when used with
> registered buffers like removing page referencing.
> 
>  From the kernel networking perspective there are two main changes. The first
> one is passing ubuf_info into the network layer from io_uring (inside of an
> in kernel struct msghdr). This allows extra optimisations, e.g. ubuf_info
> caching on the io_uring side, but also helps to avoid cross-referencing
> and synchronisation problems. The second part is an optional optimisation
> removing page referencing for requests with registered buffers.
> 
> Benchmarking with an optimised version of the selftest (see [1]), which in a
> loop sends a bunch of requests and then waits for their completions. "+ flush"
> column posts one additional "buffer-free" notification per request, and
> just "zc" doesn't post buffer notifications at all.
> 
> NIC (requests / second):
> IO size | non-zc    | zc             | zc + flush
> 4000    | 495134    | 606420 (+22%)  | 558971 (+12%)
> 1500    | 551808    | 577116 (+4.5%) | 565803 (+2.5%)
> 1000    | 584677    | 592088 (+1.2%) | 560885 (-4%)
> 600     | 596292    | 598550 (+0.4%) | 555366 (-6.7%)
> 
> dummy (requests / second):
> IO size | non-zc    | zc             | zc + flush
> 8000    | 1299916   | 2396600 (+84%) | 2224219 (+71%)
> 4000    | 1869230   | 2344146 (+25%) | 2170069 (+16%)
> 1200    | 2071617   | 2361960 (+14%) | 2203052 (+6%)
> 600     | 2106794   | 2381527 (+13%) | 2195295 (+4%)
> 
> Previously it also brought a massive performance speedup compared to the
> msg_zerocopy tool (see [3]), which is probably not super interesting.
> 
> There is an additional bunch of refcounting optimisations that was omitted from
> the series for simplicity and as they don't change the picture drastically,
> they will be sent as follow up, as well as flushing optimisations closing the
> performance gap b/w two last columns.
> 
> Note: the series is based on net-next + for-5.20/io_uring, but as vanilla
> net-next fails for me the repo (see [2]) is on top of for-5.20/io_uring.
> 
> Links:
> 
>    liburing (benchmark + some tests):
>    [1] https://github.com/isilence/liburing/tree/zc_v3
> 
>    kernel repo:
>    [2] https://github.com/isilence/linux/tree/zc_v3
> 
>    RFC v1:
>    [3] https://lore.kernel.org/io-uring/cover.1638282789.git.asml.silence@gmail.com/
> 
>    RFC v2:
>    https://lore.kernel.org/io-uring/cover.1640029579.git.asml.silence@gmail.com/
> 
> API design overview:
> 
>    The series introduces an io_uring concept of notifactors. From the userspace
>    perspective it's an entity to which it can bind one or more requests and then
>    requesting to flush it. Flushing a notifier makes it impossible to attach new
>    requests to it, and instructs the notifier to post a completion once all
>    requests attached to it are completed and the kernel doesn't need the buffers
>    anymore.
> 
>    Notifications are stored in notification slots, which should be registered as
>    an array in io_uring. Each slot stores only one notifier at any particular
>    moment. Flushing removes it from the slot and the slot automatically replaces
>    it with a new notifier. All operations with notifiers are done by specifying
>    an index of a slot it's currently in.
> 
>    When registering a notification the userspace specifies a u64 tag for each
>    slot, which will be copied in notification completion entries as
>    cqe::user_data. cqe::res is 0 and cqe::flags is equal to wrap around u32
>    sequence number counting notifiers of a slot.
> 
> Changelog:
> 
>    RFC v2 -> v3:
>      mem accounting for non-registered buffers
>      allow mixing registered and normal requests per notifier
>      notification flushing via IORING_OP_RSRC_UPDATE
>      TCP support
>      fix buffer indexing
>      fix io-wq ->uring_lock locking
>      fix bugs when mixing with MSG_ZEROCOPY
>      fix managed refs bugs in skbuff.c
> 
>    RFC -> RFC v2:
>      remove additional overhead for non-zc from skb_release_data()
>      avoid msg propagation, hide extra bits of non-zc overhead
>      task_work based "buffer free" notifications
>      improve io_uring's notification refcounting
>      added 5/19, (no pfmemalloc tracking)
>      added 8/19 and 9/19 preventing small copies with zc
>      misc small changes
> 
> Pavel Begunkov (25):
>    ipv4: avoid partial copy for zc
>    ipv6: avoid partial copy for zc
>    skbuff: add SKBFL_DONT_ORPHAN flag
>    skbuff: carry external ubuf_info in msghdr
>    net: bvec specific path in zerocopy_sg_from_iter
>    net: optimise bvec-based zc page referencing
>    net: don't track pfmemalloc for managed frags
>    skbuff: don't mix ubuf_info of different types
>    ipv4/udp: support zc with managed data
>    ipv6/udp: support zc with managed data
>    tcp: support zc with managed data
>    io_uring: add zc notification infrastructure
>    io_uring: export task put
>    io_uring: cache struct io_notif
>    io_uring: complete notifiers in tw
>    io_uring: add notification slot registration
>    io_uring: wire send zc request type
>    io_uring: account locked pages for non-fixed zc
>    io_uring: allow to pass addr into sendzc
>    io_uring: add rsrc referencing for notifiers
>    io_uring: sendzc with fixed buffers
>    io_uring: flush notifiers after sendzc
>    io_uring: rename IORING_OP_FILES_UPDATE
>    io_uring: add zc notification flush requests
>    selftests/io_uring: test zerocopy send
> 
>   include/linux/io_uring_types.h                |  37 ++
>   include/linux/skbuff.h                        |  59 +-
>   include/linux/socket.h                        |   7 +
>   include/uapi/linux/io_uring.h                 |  43 +-
>   io_uring/Makefile                             |   2 +-
>   io_uring/io_uring.c                           |  40 +-
>   io_uring/io_uring.h                           |  21 +
>   io_uring/net.c                                | 134 ++++
>   io_uring/net.h                                |   4 +
>   io_uring/notif.c                              | 215 +++++++
>   io_uring/notif.h                              |  87 +++
>   io_uring/opdef.c                              |  24 +-
>   io_uring/rsrc.c                               |  55 +-
>   io_uring/rsrc.h                               |  16 +-
>   io_uring/tctx.h                               |  26 -
>   net/compat.c                                  |   2 +
>   net/core/datagram.c                           |  53 +-
>   net/core/skbuff.c                             |  35 +-
>   net/ipv4/ip_output.c                          |  63 +-
>   net/ipv4/tcp.c                                |  52 +-
>   net/ipv6/ip6_output.c                         |  62 +-
>   net/socket.c                                  |   6 +
>   tools/testing/selftests/net/Makefile          |   1 +
>   .../selftests/net/io_uring_zerocopy_tx.c      | 605 ++++++++++++++++++
>   .../selftests/net/io_uring_zerocopy_tx.sh     | 131 ++++
>   25 files changed, 1652 insertions(+), 128 deletions(-)
>   create mode 100644 io_uring/notif.c
>   create mode 100644 io_uring/notif.h
>   create mode 100644 tools/testing/selftests/net/io_uring_zerocopy_tx.c
>   create mode 100755 tools/testing/selftests/net/io_uring_zerocopy_tx.sh
> 

-- 
Pavel Begunkov