bpf: reuse tc bpf prologue for sk skb progs
authorDaniel Borkmann <daniel@iogearbox.net>
Thu, 17 Aug 2017 15:22:37 +0000 (17:22 +0200)
committerDavid S. Miller <davem@davemloft.net>
Thu, 17 Aug 2017 17:25:19 +0000 (10:25 -0700)
Given both program types are effecitvely doing the same in the
prologue, just reuse the one that we had for tc and only adapt
to the corresponding drop verdict value. That way, we don't need
to have the duplicate from 8a31db561566 ("bpf: add access to sock
fields and pkt data from sk_skb programs") to maintain.

Reported-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/core/filter.c

index ea3ca34d0bf48bf398c0150b66de7f0cd9f1459c..0f4df86d936a1c767b7cf00ea51b870ac626a17e 100644 (file)
@@ -3455,8 +3455,8 @@ static bool sock_filter_is_valid_access(int off, int size,
        return true;
 }
 
-static int tc_cls_act_prologue(struct bpf_insn *insn_buf, bool direct_write,
-                              const struct bpf_prog *prog)
+static int bpf_unclone_prologue(struct bpf_insn *insn_buf, bool direct_write,
+                               const struct bpf_prog *prog, int drop_verdict)
 {
        struct bpf_insn *insn = insn_buf;
 
@@ -3483,7 +3483,7 @@ static int tc_cls_act_prologue(struct bpf_insn *insn_buf, bool direct_write,
         * return TC_ACT_SHOT;
         */
        *insn++ = BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 2);
-       *insn++ = BPF_ALU32_IMM(BPF_MOV, BPF_REG_0, TC_ACT_SHOT);
+       *insn++ = BPF_ALU32_IMM(BPF_MOV, BPF_REG_0, drop_verdict);
        *insn++ = BPF_EXIT_INSN();
 
        /* restore: */
@@ -3494,6 +3494,12 @@ static int tc_cls_act_prologue(struct bpf_insn *insn_buf, bool direct_write,
        return insn - insn_buf;
 }
 
+static int tc_cls_act_prologue(struct bpf_insn *insn_buf, bool direct_write,
+                              const struct bpf_prog *prog)
+{
+       return bpf_unclone_prologue(insn_buf, direct_write, prog, TC_ACT_SHOT);
+}
+
 static bool tc_cls_act_is_valid_access(int off, int size,
                                       enum bpf_access_type type,
                                       struct bpf_insn_access_aux *info)
@@ -3600,40 +3606,7 @@ static bool sock_ops_is_valid_access(int off, int size,
 static int sk_skb_prologue(struct bpf_insn *insn_buf, bool direct_write,
                           const struct bpf_prog *prog)
 {
-       struct bpf_insn *insn = insn_buf;
-
-       if (!direct_write)
-               return 0;
-
-       /* if (!skb->cloned)
-        *       goto start;
-        *
-        * (Fast-path, otherwise approximation that we might be
-        *  a clone, do the rest in helper.)
-        */
-       *insn++ = BPF_LDX_MEM(BPF_B, BPF_REG_6, BPF_REG_1, CLONED_OFFSET());
-       *insn++ = BPF_ALU32_IMM(BPF_AND, BPF_REG_6, CLONED_MASK);
-       *insn++ = BPF_JMP_IMM(BPF_JEQ, BPF_REG_6, 0, 7);
-
-       /* ret = bpf_skb_pull_data(skb, 0); */
-       *insn++ = BPF_MOV64_REG(BPF_REG_6, BPF_REG_1);
-       *insn++ = BPF_ALU64_REG(BPF_XOR, BPF_REG_2, BPF_REG_2);
-       *insn++ = BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
-                              BPF_FUNC_skb_pull_data);
-       /* if (!ret)
-        *      goto restore;
-        * return SK_DROP;
-        */
-       *insn++ = BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 2);
-       *insn++ = BPF_ALU32_IMM(BPF_MOV, BPF_REG_0, SK_DROP);
-       *insn++ = BPF_EXIT_INSN();
-
-       /* restore: */
-       *insn++ = BPF_MOV64_REG(BPF_REG_1, BPF_REG_6);
-       /* start: */
-       *insn++ = prog->insnsi[0];
-
-       return insn - insn_buf;
+       return bpf_unclone_prologue(insn_buf, direct_write, prog, SK_DROP);
 }
 
 static bool sk_skb_is_valid_access(int off, int size,