From: Eduard Zingerman <eddyz87@gmail.com>
To: bpf@vger.kernel.org, ast@kernel.org
Cc: andrii@kernel.org, daniel@iogearbox.net, kernel-team@fb.com,
yhs@fb.com, memxor@gmail.com, ecree.xilinx@gmail.com,
Eduard Zingerman <eddyz87@gmail.com>
Subject: [PATCH bpf-next 7/7] selftests/bpf: test case for relaxed prunning of active_lock.id
Date: Fri, 9 Dec 2022 15:57:33 +0200 [thread overview]
Message-ID: <20221209135733.28851-8-eddyz87@gmail.com> (raw)
In-Reply-To: <20221209135733.28851-1-eddyz87@gmail.com>
Check that verifier.c:states_equal() uses check_ids() to match
consistent active_lock/map_value configurations. This allows to prune
states with active spin locks even if numerical values of
active_lock ids do not match across compared states.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
.../selftests/bpf/verifier/spin_lock.c | 75 +++++++++++++++++++
1 file changed, 75 insertions(+)
diff --git a/tools/testing/selftests/bpf/verifier/spin_lock.c b/tools/testing/selftests/bpf/verifier/spin_lock.c
index 0a8dcfc37fc6..eaf114f07e2e 100644
--- a/tools/testing/selftests/bpf/verifier/spin_lock.c
+++ b/tools/testing/selftests/bpf/verifier/spin_lock.c
@@ -370,3 +370,78 @@
.prog_type = BPF_PROG_TYPE_SCHED_CLS,
.flags = BPF_F_TEST_STATE_FREQ,
},
+/* Make sure that regsafe() compares ids for spin lock records using
+ * check_ids():
+ * 1: r9 = map_lookup_elem(...) ; r9.id == 1
+ * 2: r8 = map_lookup_elem(...) ; r8.id == 2
+ * 3: r7 = ktime_get_ns()
+ * 4: r6 = ktime_get_ns()
+ * 5: if r6 > r7 goto <9>
+ * 6: spin_lock(r8)
+ * 7: r9 = r8
+ * 8: goto <10>
+ * 9: spin_lock(r9)
+ * 10: spin_unlock(r9) ; r9.id == 1 || r9.id == 2 and lock is active,
+ * ; second visit to (10) should be considered safe
+ * ; if check_ids() is used.
+ * 11: exit(0)
+ */
+{
+ "spin_lock: regsafe() check_ids() similar id mappings",
+ .insns = {
+ BPF_ST_MEM(BPF_W, BPF_REG_10, -4, 0),
+ /* r9 = map_lookup_elem(...) */
+ BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+ BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -4),
+ BPF_LD_MAP_FD(BPF_REG_1,
+ 0),
+ BPF_EMIT_CALL(BPF_FUNC_map_lookup_elem),
+ BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 24),
+ BPF_MOV64_REG(BPF_REG_9, BPF_REG_0),
+ /* r8 = map_lookup_elem(...) */
+ BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+ BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -4),
+ BPF_LD_MAP_FD(BPF_REG_1,
+ 0),
+ BPF_EMIT_CALL(BPF_FUNC_map_lookup_elem),
+ BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 18),
+ BPF_MOV64_REG(BPF_REG_8, BPF_REG_0),
+ /* r7 = ktime_get_ns() */
+ BPF_EMIT_CALL(BPF_FUNC_ktime_get_ns),
+ BPF_MOV64_REG(BPF_REG_7, BPF_REG_0),
+ /* r6 = ktime_get_ns() */
+ BPF_EMIT_CALL(BPF_FUNC_ktime_get_ns),
+ BPF_MOV64_REG(BPF_REG_6, BPF_REG_0),
+ /* if r6 > r7 goto +5 ; no new information about the state is derived from
+ * ; this check, thus produced verifier states differ
+ * ; only in 'insn_idx'
+ * spin_lock(r8)
+ * r9 = r8
+ * goto unlock
+ */
+ BPF_JMP_REG(BPF_JGT, BPF_REG_6, BPF_REG_7, 5),
+ BPF_MOV64_REG(BPF_REG_1, BPF_REG_8),
+ BPF_ALU64_IMM(BPF_ADD, BPF_REG_1, 4),
+ BPF_EMIT_CALL(BPF_FUNC_spin_lock),
+ BPF_MOV64_REG(BPF_REG_9, BPF_REG_8),
+ BPF_JMP_A(3),
+ /* spin_lock(r9) */
+ BPF_MOV64_REG(BPF_REG_1, BPF_REG_9),
+ BPF_ALU64_IMM(BPF_ADD, BPF_REG_1, 4),
+ BPF_EMIT_CALL(BPF_FUNC_spin_lock),
+ /* spin_unlock(r9) */
+ BPF_MOV64_REG(BPF_REG_1, BPF_REG_9),
+ BPF_ALU64_IMM(BPF_ADD, BPF_REG_1, 4),
+ BPF_EMIT_CALL(BPF_FUNC_spin_unlock),
+ /* exit(0) */
+ BPF_MOV64_IMM(BPF_REG_0, 0),
+ BPF_EXIT_INSN(),
+ },
+ .fixup_map_spin_lock = { 3, 10 },
+ .result = VERBOSE_ACCEPT,
+ .errstr = "28: safe",
+ .result_unpriv = REJECT,
+ .errstr_unpriv = "",
+ .prog_type = BPF_PROG_TYPE_CGROUP_SKB,
+ .flags = BPF_F_TEST_STATE_FREQ,
+},
--
2.34.1
next prev parent reply other threads:[~2022-12-09 13:59 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-12-09 13:57 [PATCH bpf-next 0/7] stricter register ID checking in regsafe() Eduard Zingerman
2022-12-09 13:57 ` [PATCH bpf-next 1/7] bpf: regsafe() must not skip check_ids() Eduard Zingerman
2022-12-14 0:35 ` Andrii Nakryiko
2022-12-14 13:25 ` Eduard Zingerman
2022-12-14 19:37 ` Andrii Nakryiko
2022-12-09 13:57 ` [PATCH bpf-next 2/7] selftests/bpf: test cases for regsafe() bug skipping check_id() Eduard Zingerman
2022-12-09 13:57 ` [PATCH bpf-next 3/7] bpf: states_equal() must build idmap for all function frames Eduard Zingerman
2022-12-14 0:35 ` Andrii Nakryiko
2022-12-14 15:33 ` Eduard Zingerman
2022-12-14 17:24 ` Andrii Nakryiko
2022-12-09 13:57 ` [PATCH bpf-next 4/7] selftests/bpf: verify states_equal() maintains idmap across all frames Eduard Zingerman
2022-12-14 0:35 ` Andrii Nakryiko
2022-12-14 16:38 ` Eduard Zingerman
2022-12-14 17:10 ` Andrii Nakryiko
2022-12-09 13:57 ` [PATCH bpf-next 5/7] bpf: use check_ids() for active_lock comparison Eduard Zingerman
2022-12-09 13:57 ` [PATCH bpf-next 6/7] selftests/bpf: Add pruning test case for bpf_spin_lock Eduard Zingerman
2022-12-10 21:45 ` Alexei Starovoitov
2022-12-09 13:57 ` Eduard Zingerman [this message]
2022-12-10 21:50 ` [PATCH bpf-next 0/7] stricter register ID checking in regsafe() patchwork-bot+netdevbpf
2022-12-14 0:34 ` Andrii Nakryiko
2022-12-14 16:28 ` Eduard Zingerman
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=20221209135733.28851-8-eddyz87@gmail.com \
--to=eddyz87@gmail.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=ecree.xilinx@gmail.com \
--cc=kernel-team@fb.com \
--cc=memxor@gmail.com \
--cc=yhs@fb.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