All of lore.kernel.org
 help / color / mirror / Atom feed
From: John Fastabend <john.fastabend@gmail.com>
To: bpf@vger.kernel.org, netdev@vger.kernel.org
Cc: daniel@iogearbox.net, joamaki@gmail.com,
	xiyou.wangcong@gmail.com, jakub@cloudflare.com,
	john.fastabend@gmail.com
Subject: [PATCH bpf v2 4/5] bpf: sockmap, strparser, and tls are reusing qdisc_skb_cb and colliding
Date: Wed,  3 Nov 2021 13:47:35 -0700	[thread overview]
Message-ID: <20211103204736.248403-5-john.fastabend@gmail.com> (raw)
In-Reply-To: <20211103204736.248403-1-john.fastabend@gmail.com>

Strparser is reusing the qdisc_skb_cb struct to stash the skb message
handling progress, e.g. offset and length of the skb. First this is
poorly named and inherits a struct from qdisc that doesn't reflect the
actual usage of cb[] at this layer.

But, more importantly strparser is using the following to access its
metadata.

(struct _strp_msg *)((void *)skb->cb + offsetof(struct qdisc_skb_cb, data))

Where _strp_msg is defined as,

 struct _strp_msg {
        struct strp_msg            strp;                 /*     0     8 */
        int                        accum_len;            /*     8     4 */

        /* size: 12, cachelines: 1, members: 2 */
        /* last cacheline: 12 bytes */
 };

So we use 12 bytes of ->data[] in struct. However in BPF code running
parser and verdict the user has read capabilities into the data[]
array as well. Its not too problematic, but we should not be
exposing internal state to BPF program. If its really needed then we can
use the probe_read() APIs which allow reading kernel memory. And I don't
believe cb[] layer poses any API breakage by moving this around because
programs can't depend on cb[] across layers.

In order to fix another issue with a ctx rewrite we need to stash a temp
variable somewhere. To make this work cleanly this patch builds a cb
struct for sk_skb types called sk_skb_cb struct. Then we can use this
consistently in the strparser, sockmap space. Additionally we can
start allowing ->cb[] write access after this.

Fixes: 604326b41a6fb ("bpf, sockmap: convert to generic sk_msg interface"
Tested-by: Jussi Maki <joamaki@gmail.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 include/net/strparser.h   | 16 +++++++++++++++-
 net/core/filter.c         | 22 ++++++++++++++++++++++
 net/strparser/strparser.c | 10 +---------
 3 files changed, 38 insertions(+), 10 deletions(-)

diff --git a/include/net/strparser.h b/include/net/strparser.h
index 1d20b98493a1..bec1439bd3be 100644
--- a/include/net/strparser.h
+++ b/include/net/strparser.h
@@ -54,10 +54,24 @@ struct strp_msg {
 	int offset;
 };
 
+struct _strp_msg {
+	/* Internal cb structure. struct strp_msg must be first for passing
+	 * to upper layer.
+	 */
+	struct strp_msg strp;
+	int accum_len;
+};
+
+struct sk_skb_cb {
+#define SK_SKB_CB_PRIV_LEN 20
+	unsigned char data[SK_SKB_CB_PRIV_LEN];
+	struct _strp_msg strp;
+};
+
 static inline struct strp_msg *strp_msg(struct sk_buff *skb)
 {
 	return (struct strp_msg *)((void *)skb->cb +
-		offsetof(struct qdisc_skb_cb, data));
+		offsetof(struct sk_skb_cb, strp));
 }
 
 /* Structure for an attached lower socket */
diff --git a/net/core/filter.c b/net/core/filter.c
index a68418268e92..c3936d0724b8 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -9782,11 +9782,33 @@ static u32 sk_skb_convert_ctx_access(enum bpf_access_type type,
 				     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):
 		insn = bpf_convert_data_end_access(si, insn);
 		break;
+	case offsetof(struct __sk_buff, cb[0]) ...
+	     offsetofend(struct __sk_buff, cb[4]) - 1:
+		BUILD_BUG_ON(sizeof_field(struct sk_skb_cb, data) < 20);
+		BUILD_BUG_ON((offsetof(struct sk_buff, cb) +
+			      offsetof(struct sk_skb_cb, data)) %
+			     sizeof(__u64));
+
+		prog->cb_access = 1;
+		off  = si->off;
+		off -= offsetof(struct __sk_buff, cb[0]);
+		off += offsetof(struct sk_buff, cb);
+		off += offsetof(struct sk_skb_cb, data);
+		if (type == BPF_WRITE)
+			*insn++ = BPF_STX_MEM(BPF_SIZE(si->code), si->dst_reg,
+					      si->src_reg, off);
+		else
+			*insn++ = BPF_LDX_MEM(BPF_SIZE(si->code), si->dst_reg,
+					      si->src_reg, off);
+		break;
+
+
 	default:
 		return bpf_convert_ctx_access(type, si, insn_buf, prog,
 					      target_size);
diff --git a/net/strparser/strparser.c b/net/strparser/strparser.c
index 9c0343568d2a..1a72c67afed5 100644
--- a/net/strparser/strparser.c
+++ b/net/strparser/strparser.c
@@ -27,18 +27,10 @@
 
 static struct workqueue_struct *strp_wq;
 
-struct _strp_msg {
-	/* Internal cb structure. struct strp_msg must be first for passing
-	 * to upper layer.
-	 */
-	struct strp_msg strp;
-	int accum_len;
-};
-
 static inline struct _strp_msg *_strp_msg(struct sk_buff *skb)
 {
 	return (struct _strp_msg *)((void *)skb->cb +
-		offsetof(struct qdisc_skb_cb, data));
+		offsetof(struct sk_skb_cb, strp));
 }
 
 /* Lower lock held */
-- 
2.33.0


  parent reply	other threads:[~2021-11-03 20:48 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-03 20:47 [PATCH bpf v2 0/5] bpf, sockmap: fixes stress testing and regression John Fastabend
2021-11-03 20:47 ` [PATCH bpf v2 1/5] bpf, sockmap: Use stricter sk state checks in sk_lookup_assign John Fastabend
2021-11-08 10:10   ` Jakub Sitnicki
2021-11-03 20:47 ` [PATCH bpf v2 2/5] bpf, sockmap: Remove unhash handler for BPF sockmap usage John Fastabend
2021-11-03 20:47 ` [PATCH bpf v2 3/5] bpf, sockmap: Fix race in ingress receive verdict with redirect to self John Fastabend
2021-11-03 20:47 ` John Fastabend [this message]
2021-11-03 20:47 ` [PATCH bpf v2 5/5] bpf, sockmap: sk_skb data_end access incorrect when src_reg = dst_reg John Fastabend
2021-11-08 10:16 ` [PATCH bpf v2 0/5] bpf, sockmap: fixes stress testing and regression Jakub Sitnicki
2021-11-09  0:10 ` patchwork-bot+netdevbpf

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=20211103204736.248403-5-john.fastabend@gmail.com \
    --to=john.fastabend@gmail.com \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=jakub@cloudflare.com \
    --cc=joamaki@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=xiyou.wangcong@gmail.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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.