Commit 84882cf7 authored Nov 01, 2021 by Jakub Kicinski

Revert "net: avoid double accounting for pure zerocopy skbs"



This reverts commit f1a456f8.

  WARNING: CPU: 1 PID: 6819 at net/core/skbuff.c:5429 skb_try_coalesce+0x78b/0x7e0
  CPU: 1 PID: 6819 Comm: xxxxxxx Kdump: loaded Tainted: G S                5.15.0-04194-gd852503f7711 #16
  RIP: 0010:skb_try_coalesce+0x78b/0x7e0
  Code: e8 2a bf 41 ff 44 8b b3 bc 00 00 00 48 8b 7c 24 30 e8 19 c0 41 ff 44 89 f0 48 03 83 c0 00 00 00 48 89 44 24 40 e9 47 fb ff ff <0f> 0b e9 ca fc ff ff 4c 8d 70 ff 48 83 c0 07 48 89 44 24 38 e9 61
  RSP: 0018:ffff88881f449688 EFLAGS: 00010282
  RAX: 00000000fffffe96 RBX: ffff8881566e4460 RCX: ffffffff82079f7e
  RDX: 0000000000000003 RSI: dffffc0000000000 RDI: ffff8881566e47b0
  RBP: ffff8881566e46e0 R08: ffffed102619235d R09: ffffed102619235d
  R10: ffff888130c91ae3 R11: ffffed102619235c R12: ffff88881f4498a0
  R13: 0000000000000056 R14: 0000000000000009 R15: ffff888130c91ac0
  FS:  00007fec2cbb9700(0000) GS:ffff88881f440000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007fec1b060d80 CR3: 00000003acf94005 CR4: 00000000003706e0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  Call Trace:
   <IRQ>
   tcp_try_coalesce+0xeb/0x290
   ? tcp_parse_options+0x610/0x610
   ? mark_held_locks+0x79/0xa0
   tcp_queue_rcv+0x69/0x2f0
   tcp_rcv_established+0xa49/0xd40
   ? tcp_data_queue+0x18a0/0x18a0
   tcp_v6_do_rcv+0x1c9/0x880
   ? rt6_mtu_change_route+0x100/0x100
   tcp_v6_rcv+0x1624/0x1830

Signed-off-by: Jakub Kicinski <kuba@kernel.org>

parent 8a33dcc2

include/linux/skbuff.h

+1 −18

Original line number	Diff line number	Diff line
		@@ -454,15 +454,9 @@ enum {
		* all frags to avoid possible bad checksum
		*/
		SKBFL_SHARED_FRAG = BIT(1),

		/* segment contains only zerocopy data and should not be
		* charged to the kernel memory.
		*/
		SKBFL_PURE_ZEROCOPY = BIT(2),
		};

		#define SKBFL_ZEROCOPY_FRAG (SKBFL_ZEROCOPY_ENABLE \| SKBFL_SHARED_FRAG)
		#define SKBFL_ALL_ZEROCOPY (SKBFL_ZEROCOPY_FRAG \| SKBFL_PURE_ZEROCOPY)

		/*
		* The callback notifies userspace to release buffers when skb DMA is done in
		@@ -1470,17 +1464,6 @@ static inline struct ubuf_info skb_zcopy(struct sk_buff skb)
		return is_zcopy ? skb_uarg(skb) : NULL;
		}

		static inline bool skb_zcopy_pure(const struct sk_buff *skb)
		{
		return skb_shinfo(skb)->flags & SKBFL_PURE_ZEROCOPY;
		}

		static inline bool skb_pure_zcopy_same(const struct sk_buff *skb1,
		const struct sk_buff *skb2)
		{
		return skb_zcopy_pure(skb1) == skb_zcopy_pure(skb2);
		}

		static inline void net_zcopy_get(struct ubuf_info *uarg)
		{
		refcount_inc(&uarg->refcnt);
		@@ -1545,7 +1528,7 @@ static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy_success)
		if (!skb_zcopy_is_nouarg(skb))
		uarg->callback(skb, uarg, zerocopy_success);

		skb_shinfo(skb)->flags &= ~SKBFL_ALL_ZEROCOPY;
		skb_shinfo(skb)->flags &= ~SKBFL_ZEROCOPY_FRAG;
		}
		}

include/net/tcp.h

+2 −6

Original line number	Diff line number	Diff line
		@@ -293,10 +293,7 @@ static inline bool tcp_out_of_memory(struct sock *sk)
		static inline void tcp_wmem_free_skb(struct sock sk, struct sk_buff skb)
		{
		sk_wmem_queued_add(sk, -skb->truesize);
		if (!skb_zcopy_pure(skb))
		sk_mem_uncharge(sk, skb->truesize);
		else
		sk_mem_uncharge(sk, SKB_TRUESIZE(MAX_TCP_HEADER));
		__kfree_skb(skb);
		}

		@@ -977,8 +974,7 @@ static inline bool tcp_skb_can_collapse(const struct sk_buff *to,
		const struct sk_buff *from)
		{
		return likely(tcp_skb_can_collapse_to(to) &&
		mptcp_skb_can_collapse(to, from) &&
		skb_pure_zcopy_same(to, from));
		mptcp_skb_can_collapse(to, from));
		}

		/* Events passed to congestion control interface */

net/core/datagram.c

+1 −2

Original line number	Diff line number	Diff line
		@@ -646,7 +646,6 @@ int __zerocopy_sg_from_iter(struct sock sk, struct sk_buff skb,
		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);

net/core/skbuff.c

+1 −2

Original line number	Diff line number	Diff line
		@@ -3433,9 +3433,8 @@ static inline void skb_split_no_header(struct sk_buff *skb,
		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_shinfo(skb1)->flags \|= skb_shinfo(skb)->flags & zc_flags;
		skb_shinfo(skb1)->flags \|= skb_shinfo(skb)->flags & SKBFL_SHARED_FRAG;
		skb_zerocopy_clone(skb1, skb, 0);
		if (len < pos) /* Split line is inside header. */
		skb_split_inside_header(skb, skb1, len, pos);

net/ipv4/tcp.c

+2 −20

Original line number	Diff line number	Diff line
		@@ -863,7 +863,6 @@ struct sk_buff tcp_stream_alloc_skb(struct sock sk, int size, gfp_t gfp,
		if (likely(skb)) {
		bool mem_scheduled;

		skb->truesize = SKB_TRUESIZE(size + MAX_TCP_HEADER);
		if (force_schedule) {
		mem_scheduled = true;
		sk_forced_mem_schedule(sk, skb->truesize);
		@@ -1320,15 +1319,6 @@ int tcp_sendmsg_locked(struct sock sk, struct msghdr msg, size_t size)

		copy = min_t(int, copy, pfrag->size - pfrag->offset);

		/* skb changing from pure zc to mixed, must charge zc */
		if (unlikely(skb_zcopy_pure(skb))) {
		if (!sk_wmem_schedule(sk, skb->data_len))
		goto wait_for_space;

		sk_mem_charge(sk, skb->data_len);
		skb_shinfo(skb)->flags &= ~SKBFL_PURE_ZEROCOPY;
		}

		if (!sk_wmem_schedule(sk, copy))
		goto wait_for_space;

		@@ -1349,16 +1339,8 @@ 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
		*/
		if (!skb->len)
		skb_shinfo(skb)->flags \|= SKBFL_PURE_ZEROCOPY;

		if (!skb_zcopy_pure(skb)) {
		if (!sk_wmem_schedule(sk, copy))
		goto wait_for_space;
		}

		err = skb_zerocopy_iter_stream(sk, skb, msg, copy, uarg);
		if (err == -EMSGSIZE \|\| err == -EEXIST) {