Merge branch 'io_uring-zerocopy-send' of git://git.kernel.org/pub/scm/linux/kernel/git/kuba/linux

Pavel Begunkov says:

====================
io_uring zerocopy send

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 UDP with an optimised version of the selftest (see [1]), which
sends a bunch of requests, waits for completions and repeats. "+ 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 also 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.

For TCP on localhost (with hacks enabling localhost zerocopy) and including
additional overhead for receive:

IO size | non-zc | zc
1200 | 4174 | 4148
4096 | 7597 | 11228

Using a real NIC 1200 bytes, zc is worse than non-zc ~5-10%, maybe the
omitted optimisations will somewhat help, should look better for 4000,
but couldn't test properly because of setup problems.

Links:

liburing (benchmark + tests):
[1] https://github.com/isilence/liburing/tree/zc_v4

kernel repo:
[2] https://github.com/isilence/linux/tree/zc_v4

RFC v1:
[3] https://lore.kernel.org/io-uring/[e-mail]/

RFC v2:
https://lore.kernel.org/io-uring/[e-mail]/

Net patches based:
[e-mail]:isilence/linux.git zc_v4-net-base
or
https://github.com/isilence/linux/tree/zc_v4-net-base

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.

====================

Link: https://lore.kernel.org/r/[e-mail]
Signed-off-by: Jakub Kicinski <[e-mail]>

include/linux/skbuff.h

···
 	 * charged to the kernel memory.
 	 */
 	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)
+#define SKBFL_ALL_ZEROCOPY	(SKBFL_ZEROCOPY_FRAG | SKBFL_PURE_ZEROCOPY | \
+				 SKBFL_DONT_ORPHAN | SKBFL_MANAGED_FRAG_REFS)
 
 /*
  * The callback notifies userspace to release buffers when skb DMA is done in
···
 void msg_zerocopy_callback(struct sk_buff *skb, struct ubuf_info *uarg,
 			   bool success);
 
-int __zerocopy_sg_from_iter(struct sock *sk, struct sk_buff *skb,
-			    struct iov_iter *from, size_t length);
+int __zerocopy_sg_from_iter(struct msghdr *msg, struct sock *sk,
+			    struct sk_buff *skb, struct iov_iter *from,
+			    size_t length);
 
 static inline int skb_zerocopy_iter_dgram(struct sk_buff *skb,
 					  struct msghdr *msg, int len)
 {
-	return __zerocopy_sg_from_iter(skb->sk, skb, &msg->msg_iter, len);
+	return __zerocopy_sg_from_iter(msg, skb->sk, skb, &msg->msg_iter, len);
 }
 
 int skb_zerocopy_iter_stream(struct sock *sk, struct sk_buff *skb,
···
 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,
···
 
 		skb_shinfo(skb)->flags &= ~SKBFL_ALL_ZEROCOPY;
 	}
+}
+
+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)
···
 	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_len_add - adds a number to len fields of skb
  * @skb: buffer to add len to
···
 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;
···
 {
 	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);
 }
···
  */
 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);
 }
 
 /**

include/linux/socket.h

···
 struct pid;
 struct cred;
 struct socket;
+struct sock;
+struct sk_buff;
 
 #define __sockaddr_check_size(size)	\
 	BUILD_BUG_ON(((size) > sizeof(struct __kernel_sockaddr_storage)))
···
 	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;
+	int (*sg_from_iter)(struct sock *sk, struct sk_buff *skb,
+			    struct iov_iter *from, size_t length);
 };
 
 struct user_msghdr {

net/compat.c

···
 		return -EMSGSIZE;
 
 	kmsg->msg_iocb = NULL;
+	kmsg->msg_ubuf = NULL;
 	*ptr = msg.msg_iov;
 	*len = msg.msg_iovlen;
 	return 0;

net/core/datagram.c

···
 }
 EXPORT_SYMBOL(skb_copy_datagram_from_iter);
 
-int __zerocopy_sg_from_iter(struct sock *sk, struct sk_buff *skb,
-			    struct iov_iter *from, size_t length)
+int __zerocopy_sg_from_iter(struct msghdr *msg, struct sock *sk,
+			    struct sk_buff *skb, struct iov_iter *from,
+			    size_t length)
 {
-	int frag = skb_shinfo(skb)->nr_frags;
+	int frag;
+
+	if (msg && msg->sg_from_iter)
+		return msg->sg_from_iter(sk, skb, from, length);
+
+	frag = skb_shinfo(skb)->nr_frags;
 
 	while (length && iov_iter_count(from)) {
 		struct page *pages[MAX_SKB_FRAGS];
···
 	if (skb_copy_datagram_from_iter(skb, 0, from, copy))
 		return -EFAULT;
 
-	return __zerocopy_sg_from_iter(NULL, skb, from, ~0U);
+	return __zerocopy_sg_from_iter(NULL, NULL, skb, from, ~0U);
 }
 EXPORT_SYMBOL(zerocopy_sg_from_iter);
 

net/core/skbuff.c

···
 			      &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);
 
···
  */
 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);
 
···
 	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);
 
···
 	if (uarg) {
 		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
···
 	if (orig_uarg && uarg != orig_uarg)
 		return -EEXIST;
 
-	err = __zerocopy_sg_from_iter(sk, skb, &msg->msg_iter, len);
+	err = __zerocopy_sg_from_iter(msg, sk, skb, &msg->msg_iter, len);
 	if (err == -EFAULT || (err == -EMSGSIZE && skb->len == orig_len)) {
 		struct sock *save_sk = skb->sk;
 
···
 	return skb->len - orig_len;
 }
 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)
···
 	BUG_ON(nhead < 0);
 
 	BUG_ON(skb_shared(skb));
+
+	skb_zcopy_downgrade_managed(skb);
 
 	size = SKB_DATA_ALIGN(size);
 
···
 	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. */
···
 	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 {

net/ipv4/ip_output.c

···
 	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;
···
 	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;
···
 	    (!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;
-		} 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);
+			}
 		}
 	}
 
···
 				 (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;
···
 				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;

net/ipv4/tcp.c

···
 
 	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) &&
···
 
 			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)

net/ipv6/ip6_output.c

···
 	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;
···
 	    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;
-		} 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);
+			}
 		}
 	}
 
···
 				 (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;
 
···
 				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;

net/socket.c

···
 	msg.msg_control = NULL;
 	msg.msg_controllen = 0;
 	msg.msg_namelen = 0;
+	msg.msg_ubuf = NULL;
 	if (addr) {
 		err = move_addr_to_kernel(addr, addr_len, &address);
 		if (err < 0)
···
 		return -EMSGSIZE;
 
 	kmsg->msg_iocb = NULL;
+	kmsg->msg_ubuf = NULL;
 	*uiov = msg.msg_iov;
 	*nsegs = msg.msg_iovlen;
 	return 0;