BPF Archive on lore.kernel.org
 help / color / Atom feed
From: Cong Wang <xiyou.wangcong@gmail.com>
To: netdev@vger.kernel.org
Cc: bpf@vger.kernel.org, duanxiongchun@bytedance.com,
	wangdongdong.6@bytedance.com, jiang.wang@bytedance.com,
	Cong Wang <cong.wang@bytedance.com>,
	Jakub Sitnicki <jakub@cloudflare.com>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Lorenz Bauer <lmb@cloudflare.com>,
	John Fastabend <john.fastabend@gmail.com>
Subject: [Patch bpf-next v6 3/8] bpf: compute data_end dynamically with JIT code
Date: Fri, 19 Feb 2021 21:29:19 -0800
Message-ID: <20210220052924.106599-4-xiyou.wangcong@gmail.com> (raw)
In-Reply-To: <20210220052924.106599-1-xiyou.wangcong@gmail.com>

From: Cong Wang <cong.wang@bytedance.com>

Currently, we compute ->data_end with a compile-time constant
offset of skb. But as Jakub pointed out, we can actually compute
it in eBPF JIT code at run-time, so that we can competely get
rid of ->data_end. This is similar to skb_shinfo(skb) computation
in bpf_convert_shinfo_access().

Suggested-by: Jakub Sitnicki <jakub@cloudflare.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Lorenz Bauer <lmb@cloudflare.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Cong Wang <cong.wang@bytedance.com>
---
 include/net/tcp.h |  6 ------
 net/core/filter.c | 48 +++++++++++++++++++++++++++--------------------
 net/core/skmsg.c  |  1 -
 3 files changed, 28 insertions(+), 27 deletions(-)

diff --git a/include/net/tcp.h b/include/net/tcp.h
index c00e125dcfb9..947ef5da6867 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -886,18 +886,12 @@ struct tcp_skb_cb {
 		struct {
 			__u32 flags;
 			struct sock *sk_redir;
-			void *data_end;
 		} bpf;
 	};
 };
 
 #define TCP_SKB_CB(__skb)	((struct tcp_skb_cb *)&((__skb)->cb[0]))
 
-static inline void bpf_compute_data_end_sk_skb(struct sk_buff *skb)
-{
-	TCP_SKB_CB(skb)->bpf.data_end = skb->data + skb_headlen(skb);
-}
-
 static inline bool tcp_skb_bpf_ingress(const struct sk_buff *skb)
 {
 	return TCP_SKB_CB(skb)->bpf.flags & BPF_F_INGRESS;
diff --git a/net/core/filter.c b/net/core/filter.c
index adfdad234674..13bcf248ee7b 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -1863,10 +1863,7 @@ static const struct bpf_func_proto bpf_sk_fullsock_proto = {
 static inline int sk_skb_try_make_writable(struct sk_buff *skb,
 					   unsigned int write_len)
 {
-	int err = __bpf_try_make_writable(skb, write_len);
-
-	bpf_compute_data_end_sk_skb(skb);
-	return err;
+	return __bpf_try_make_writable(skb, write_len);
 }
 
 BPF_CALL_2(sk_skb_pull_data, struct sk_buff *, skb, u32, len)
@@ -3577,7 +3574,6 @@ BPF_CALL_4(sk_skb_adjust_room, struct sk_buff *, skb, s32, len_diff,
 			return -ENOMEM;
 		__skb_pull(skb, len_diff_abs);
 	}
-	bpf_compute_data_end_sk_skb(skb);
 	if (tls_sw_has_ctx_rx(skb->sk)) {
 		struct strp_msg *rxm = strp_msg(skb);
 
@@ -3742,10 +3738,7 @@ static const struct bpf_func_proto bpf_skb_change_tail_proto = {
 BPF_CALL_3(sk_skb_change_tail, struct sk_buff *, skb, u32, new_len,
 	   u64, flags)
 {
-	int ret = __bpf_skb_change_tail(skb, new_len, flags);
-
-	bpf_compute_data_end_sk_skb(skb);
-	return ret;
+	return __bpf_skb_change_tail(skb, new_len, flags);
 }
 
 static const struct bpf_func_proto sk_skb_change_tail_proto = {
@@ -3808,10 +3801,7 @@ static const struct bpf_func_proto bpf_skb_change_head_proto = {
 BPF_CALL_3(sk_skb_change_head, struct sk_buff *, skb, u32, head_room,
 	   u64, flags)
 {
-	int ret = __bpf_skb_change_head(skb, head_room, flags);
-
-	bpf_compute_data_end_sk_skb(skb);
-	return ret;
+	return __bpf_skb_change_head(skb, head_room, flags);
 }
 
 static const struct bpf_func_proto sk_skb_change_head_proto = {
@@ -9655,22 +9645,40 @@ static u32 sock_ops_convert_ctx_access(enum bpf_access_type type,
 	return insn - insn_buf;
 }
 
+/* data_end = skb->data + skb_headlen() */
+static struct bpf_insn *bpf_convert_data_end_access(const struct bpf_insn *si,
+						    struct bpf_insn *insn)
+{
+	/* si->dst_reg = skb->data */
+	*insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF(struct sk_buff, data),
+			      si->dst_reg, si->src_reg,
+			      offsetof(struct sk_buff, data));
+	/* AX = skb->len */
+	*insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF(struct sk_buff, len),
+			      BPF_REG_AX, si->src_reg,
+			      offsetof(struct sk_buff, len));
+	/* si->dst_reg = skb->data + skb->len */
+	*insn++ = BPF_ALU64_REG(BPF_ADD, si->dst_reg, BPF_REG_AX);
+	/* AX = skb->data_len */
+	*insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF(struct sk_buff, data_len),
+			      BPF_REG_AX, si->src_reg,
+			      offsetof(struct sk_buff, data_len));
+	/* si->dst_reg = skb->data + skb->len - skb->data_len */
+	*insn++ = BPF_ALU64_REG(BPF_SUB, si->dst_reg, BPF_REG_AX);
+
+	return insn;
+}
+
 static u32 sk_skb_convert_ctx_access(enum bpf_access_type type,
 				     const struct bpf_insn *si,
 				     struct bpf_insn *insn_buf,
 				     struct bpf_prog *prog, u32 *target_size)
 {
 	struct bpf_insn *insn = insn_buf;
-	int off;
 
 	switch (si->off) {
 	case offsetof(struct __sk_buff, data_end):
-		off  = si->off;
-		off -= offsetof(struct __sk_buff, data_end);
-		off += offsetof(struct sk_buff, cb);
-		off += offsetof(struct tcp_skb_cb, bpf.data_end);
-		*insn++ = BPF_LDX_MEM(BPF_SIZEOF(void *), si->dst_reg,
-				      si->src_reg, off);
+		insn = bpf_convert_data_end_access(si, insn);
 		break;
 	default:
 		return bpf_convert_ctx_access(type, si, insn_buf, prog,
diff --git a/net/core/skmsg.c b/net/core/skmsg.c
index 7f400d044cda..2d8bbb3fd87c 100644
--- a/net/core/skmsg.c
+++ b/net/core/skmsg.c
@@ -744,7 +744,6 @@ EXPORT_SYMBOL_GPL(sk_psock_msg_verdict);
 static int sk_psock_bpf_run(struct sk_psock *psock, struct bpf_prog *prog,
 			    struct sk_buff *skb)
 {
-	bpf_compute_data_end_sk_skb(skb);
 	return bpf_prog_run_pin_on_cpu(prog, skb);
 }
 
-- 
2.25.1


  parent reply index

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-20  5:29 [Patch bpf-next v6 0/8] sock_map: clean up and refactor code for BPF_SK_SKB_VERDICT Cong Wang
2021-02-20  5:29 ` [Patch bpf-next v6 1/8] bpf: clean up sockmap related Kconfigs Cong Wang
2021-02-22  8:51   ` Jakub Sitnicki
2021-02-22 23:23     ` Cong Wang
2021-02-20  5:29 ` [Patch bpf-next v6 2/8] skmsg: get rid of struct sk_psock_parser Cong Wang
2021-02-20  5:29 ` Cong Wang [this message]
2021-02-20  5:29 ` [Patch bpf-next v6 4/8] skmsg: move sk_redir from TCP_SKB_CB to skb Cong Wang
2021-02-22 12:20   ` Jakub Sitnicki
2021-02-22 19:27     ` Cong Wang
2021-02-23 17:52       ` Jakub Sitnicki
2021-02-23 18:04         ` Cong Wang
2021-02-23 18:36           ` Jakub Sitnicki
2021-02-20  5:29 ` [Patch bpf-next v6 5/8] sock_map: rename skb_parser and skb_verdict Cong Wang
2021-02-22 12:28   ` Jakub Sitnicki
2021-02-22 19:32     ` Cong Wang
2021-02-20  5:29 ` [Patch bpf-next v6 6/8] sock_map: make sock_map_prog_update() static Cong Wang
2021-02-22 12:29   ` Jakub Sitnicki
2021-02-20  5:29 ` [Patch bpf-next v6 7/8] skmsg: make __sk_psock_purge_ingress_msg() static Cong Wang
2021-02-22 12:30   ` Jakub Sitnicki
2021-02-20  5:29 ` [Patch bpf-next v6 8/8] skmsg: get rid of sk_psock_bpf_run() Cong Wang
2021-02-22 12:31   ` Jakub Sitnicki
2021-02-22 12:32 ` [Patch bpf-next v6 0/8] sock_map: clean up and refactor code for BPF_SK_SKB_VERDICT Jakub Sitnicki

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210220052924.106599-4-xiyou.wangcong@gmail.com \
    --to=xiyou.wangcong@gmail.com \
    --cc=bpf@vger.kernel.org \
    --cc=cong.wang@bytedance.com \
    --cc=daniel@iogearbox.net \
    --cc=duanxiongchun@bytedance.com \
    --cc=jakub@cloudflare.com \
    --cc=jiang.wang@bytedance.com \
    --cc=john.fastabend@gmail.com \
    --cc=lmb@cloudflare.com \
    --cc=netdev@vger.kernel.org \
    --cc=wangdongdong.6@bytedance.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

BPF Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/bpf/0 bpf/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 bpf bpf/ https://lore.kernel.org/bpf \
		bpf@vger.kernel.org
	public-inbox-index bpf

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.bpf


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git