From: Andrii Nakryiko <andrii@kernel.org>
To: <bpf@vger.kernel.org>, <ast@kernel.org>, <daniel@iogearbox.net>,
<martin.lau@kernel.org>
Cc: <andrii@kernel.org>, <kernel-team@meta.com>,
Eduard Zingerman <eddyz87@gmail.com>,
Hao Sun <sunhao.th@gmail.com>
Subject: [PATCH v2 bpf 1/3] bpf: handle ldimm64 properly in check_cfg()
Date: Thu, 9 Nov 2023 16:26:36 -0800 [thread overview]
Message-ID: <20231110002638.4168352-2-andrii@kernel.org> (raw)
In-Reply-To: <20231110002638.4168352-1-andrii@kernel.org>
ldimm64 instructions are 16-byte long, and so have to be handled
appropriately in check_cfg(), just like the rest of BPF verifier does.
This has implications in three places:
- when determining next instruction for non-jump instructions;
- when determining next instruction for callback address ldimm64
instructions (in visit_func_call_insn());
- when checking for unreachable instructions, where second half of
ldimm64 is expected to be unreachable;
We take this also as an opportunity to report jump into the middle of
ldimm64. And adjust few test_verifier tests accordingly.
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Reported-by: Hao Sun <sunhao.th@gmail.com>
Fixes: 475fb78fbf48 ("bpf: verifier (add branch/goto checks)")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
include/linux/bpf.h | 8 ++++--
kernel/bpf/verifier.c | 27 ++++++++++++++-----
.../testing/selftests/bpf/verifier/ld_imm64.c | 8 +++---
3 files changed, 30 insertions(+), 13 deletions(-)
diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index b4825d3cdb29..35bff17396c0 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -909,10 +909,14 @@ bpf_ctx_record_field_size(struct bpf_insn_access_aux *aux, u32 size)
aux->ctx_field_size = size;
}
+static bool bpf_is_ldimm64(const struct bpf_insn *insn)
+{
+ return insn->code == (BPF_LD | BPF_IMM | BPF_DW);
+}
+
static inline bool bpf_pseudo_func(const struct bpf_insn *insn)
{
- return insn->code == (BPF_LD | BPF_IMM | BPF_DW) &&
- insn->src_reg == BPF_PSEUDO_FUNC;
+ return bpf_is_ldimm64(insn) && insn->src_reg == BPF_PSEUDO_FUNC;
}
struct bpf_prog_ops {
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index bd1c42eb540f..b87715b364fd 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -15439,15 +15439,16 @@ static int visit_func_call_insn(int t, struct bpf_insn *insns,
struct bpf_verifier_env *env,
bool visit_callee)
{
- int ret;
+ int ret, insn_sz;
- ret = push_insn(t, t + 1, FALLTHROUGH, env, false);
+ insn_sz = bpf_is_ldimm64(&insns[t]) ? 2 : 1;
+ ret = push_insn(t, t + insn_sz, FALLTHROUGH, env, false);
if (ret)
return ret;
- mark_prune_point(env, t + 1);
+ mark_prune_point(env, t + insn_sz);
/* when we exit from subprog, we need to record non-linear history */
- mark_jmp_point(env, t + 1);
+ mark_jmp_point(env, t + insn_sz);
if (visit_callee) {
mark_prune_point(env, t);
@@ -15469,15 +15470,17 @@ static int visit_func_call_insn(int t, struct bpf_insn *insns,
static int visit_insn(int t, struct bpf_verifier_env *env)
{
struct bpf_insn *insns = env->prog->insnsi, *insn = &insns[t];
- int ret, off;
+ int ret, off, insn_sz;
if (bpf_pseudo_func(insn))
return visit_func_call_insn(t, insns, env, true);
/* All non-branch instructions have a single fall-through edge. */
if (BPF_CLASS(insn->code) != BPF_JMP &&
- BPF_CLASS(insn->code) != BPF_JMP32)
- return push_insn(t, t + 1, FALLTHROUGH, env, false);
+ BPF_CLASS(insn->code) != BPF_JMP32) {
+ insn_sz = bpf_is_ldimm64(insn) ? 2 : 1;
+ return push_insn(t, t + insn_sz, FALLTHROUGH, env, false);
+ }
switch (BPF_OP(insn->code)) {
case BPF_EXIT:
@@ -15607,11 +15610,21 @@ static int check_cfg(struct bpf_verifier_env *env)
}
for (i = 0; i < insn_cnt; i++) {
+ struct bpf_insn *insn = &env->prog->insnsi[i];
+
if (insn_state[i] != EXPLORED) {
verbose(env, "unreachable insn %d\n", i);
ret = -EINVAL;
goto err_free;
}
+ if (bpf_is_ldimm64(insn)) {
+ if (insn_state[i + 1] != 0) {
+ verbose(env, "jump into the middle of ldimm64 insn %d\n", i);
+ ret = -EINVAL;
+ goto err_free;
+ }
+ i++; /* skip second half of ldimm64 */
+ }
}
ret = 0; /* cfg looks good */
diff --git a/tools/testing/selftests/bpf/verifier/ld_imm64.c b/tools/testing/selftests/bpf/verifier/ld_imm64.c
index f9297900cea6..78f19c255f20 100644
--- a/tools/testing/selftests/bpf/verifier/ld_imm64.c
+++ b/tools/testing/selftests/bpf/verifier/ld_imm64.c
@@ -9,8 +9,8 @@
BPF_MOV64_IMM(BPF_REG_0, 2),
BPF_EXIT_INSN(),
},
- .errstr = "invalid BPF_LD_IMM insn",
- .errstr_unpriv = "R1 pointer comparison",
+ .errstr = "jump into the middle of ldimm64 insn 1",
+ .errstr_unpriv = "jump into the middle of ldimm64 insn 1",
.result = REJECT,
},
{
@@ -23,8 +23,8 @@
BPF_LD_IMM64(BPF_REG_0, 1),
BPF_EXIT_INSN(),
},
- .errstr = "invalid BPF_LD_IMM insn",
- .errstr_unpriv = "R1 pointer comparison",
+ .errstr = "jump into the middle of ldimm64 insn 1",
+ .errstr_unpriv = "jump into the middle of ldimm64 insn 1",
.result = REJECT,
},
{
--
2.34.1
next prev parent reply other threads:[~2023-11-10 0:26 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-10 0:26 [PATCH v2 bpf 0/3] BPF control flow graph and precision backtrack fixes Andrii Nakryiko
2023-11-10 0:26 ` Andrii Nakryiko [this message]
2023-11-10 0:26 ` [PATCH v2 bpf 2/3] bpf: fix precision backtracking instruction iteration Andrii Nakryiko
2023-11-10 0:26 ` [PATCH v2 bpf 3/3] selftests/bpf: add edge case backtracking logic test Andrii Nakryiko
2023-11-10 1:34 ` Alexei Starovoitov
2023-11-10 3:43 ` Andrii Nakryiko
2023-11-10 4:05 ` Andrii Nakryiko
2023-11-10 4:14 ` Alexei Starovoitov
2023-11-10 4:48 ` Andrii Nakryiko
2023-11-10 5:06 ` Andrii Nakryiko
2023-11-10 5:50 ` [PATCH v2 bpf 0/3] BPF control flow graph and precision backtrack fixes 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=20231110002638.4168352-2-andrii@kernel.org \
--to=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=eddyz87@gmail.com \
--cc=kernel-team@meta.com \
--cc=martin.lau@kernel.org \
--cc=sunhao.th@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox