bpf: sockmap, strparser, and tls are reusing qdisc_skb_cb and colliding

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: 604326b41a ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: Jussi Maki <joamaki@gmail.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20211103204736.248403-5-john.fastabend@gmail.com
This commit is contained in:
John Fastabend 2021-11-03 13:47:35 -07:00 committed by Daniel Borkmann
parent c5d2177a72
commit e0dc3b93bd
3 changed files with 38 additions and 10 deletions

View File

@ -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 */

View File

@ -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);

View File

@ -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 */