From: Eduard Zingerman <eddyz87@gmail.com>
To: bpf@vger.kernel.org, ast@kernel.org
Cc: andrii@kernel.org, daniel@iogearbox.net, martin.lau@linux.dev,
kernel-team@fb.com, yonghong.song@linux.dev, tj@kernel.org,
Eduard Zingerman <eddyz87@gmail.com>
Subject: [PATCH bpf-next v2 4/5] bpf: use register liveness information for func_states_equal
Date: Mon, 3 Mar 2025 23:42:38 -0800 [thread overview]
Message-ID: <20250304074239.2328752-5-eddyz87@gmail.com> (raw)
In-Reply-To: <20250304074239.2328752-1-eddyz87@gmail.com>
Liveness analysis DFA computes a set of registers live before each
instruction. Leverage this information to skip comparison of dead
registers in func_states_equal(). This helps with convergance of
iterator processing loops, as bpf_reg_state->live marks can't be used
when loops are processed.
This has certain performance impact for selftests, here is a veristat
listing using `-f "insns_pct>5" -f "!insns<200"`
selftests:
File Program States (A) States (B) States (DIFF)
-------------------- ----------------------------- ---------- ---------- --------------
arena_htab.bpf.o arena_htab_llvm 37 35 -2 (-5.41%)
arena_htab_asm.bpf.o arena_htab_asm 37 33 -4 (-10.81%)
arena_list.bpf.o arena_list_add 37 22 -15 (-40.54%)
dynptr_success.bpf.o test_dynptr_copy 22 16 -6 (-27.27%)
dynptr_success.bpf.o test_dynptr_copy_xdp 68 58 -10 (-14.71%)
iters.bpf.o checkpoint_states_deletion 918 40 -878 (-95.64%)
iters.bpf.o clean_live_states 136 66 -70 (-51.47%)
iters.bpf.o iter_nested_deeply_iters 43 37 -6 (-13.95%)
iters.bpf.o iter_nested_iters 72 62 -10 (-13.89%)
iters.bpf.o iter_pass_iter_ptr_to_subprog 30 26 -4 (-13.33%)
iters.bpf.o iter_subprog_iters 68 59 -9 (-13.24%)
iters.bpf.o loop_state_deps2 35 32 -3 (-8.57%)
iters_css.bpf.o iter_css_for_each 32 29 -3 (-9.38%)
pyperf600_iter.bpf.o on_event 286 192 -94 (-32.87%)
Total progs: 3577
Old success: 2060
New success: 2060
States diff min: -95.64%
States diff max: 0.00%
-100 .. -90 %: 1
-55 .. -45 %: 3
-45 .. -35 %: 2
-35 .. -25 %: 5
-20 .. -10 %: 12
-10 .. 0 %: 6
sched_ext:
File Program States (A) States (B) States (DIFF)
----------------- ---------------------- ---------- ---------- ---------------
bpf.bpf.o lavd_dispatch 8950 7065 -1885 (-21.06%)
bpf.bpf.o lavd_init 516 480 -36 (-6.98%)
bpf.bpf.o layered_dispatch 662 501 -161 (-24.32%)
bpf.bpf.o layered_dump 298 237 -61 (-20.47%)
bpf.bpf.o layered_init 523 423 -100 (-19.12%)
bpf.bpf.o layered_init_task 24 22 -2 (-8.33%)
bpf.bpf.o layered_runnable 151 125 -26 (-17.22%)
bpf.bpf.o p2dq_dispatch 66 53 -13 (-19.70%)
bpf.bpf.o p2dq_init 170 142 -28 (-16.47%)
bpf.bpf.o refresh_layer_cpumasks 120 78 -42 (-35.00%)
bpf.bpf.o rustland_init 37 34 -3 (-8.11%)
bpf.bpf.o rustland_init 37 34 -3 (-8.11%)
bpf.bpf.o rusty_select_cpu 125 108 -17 (-13.60%)
scx_central.bpf.o central_dispatch 59 43 -16 (-27.12%)
scx_central.bpf.o central_init 39 28 -11 (-28.21%)
scx_nest.bpf.o nest_init 58 51 -7 (-12.07%)
scx_pair.bpf.o pair_dispatch 142 111 -31 (-21.83%)
scx_qmap.bpf.o qmap_dispatch 174 141 -33 (-18.97%)
scx_qmap.bpf.o qmap_init 768 654 -114 (-14.84%)
Total progs: 216
Old success: 186
New success: 186
States diff min: -35.00%
States diff max: 0.00%
-35 .. -25 %: 3
-25 .. -20 %: 6
-20 .. -15 %: 6
-15 .. -5 %: 7
-5 .. 0 %: 6
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
kernel/bpf/verifier.c | 14 ++++++++++----
1 file changed, 10 insertions(+), 4 deletions(-)
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 09298e0e4b73..9685b283224a 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -18500,15 +18500,17 @@ static bool refsafe(struct bpf_verifier_state *old, struct bpf_verifier_state *c
* the current state will reach 'bpf_exit' instruction safely
*/
static bool func_states_equal(struct bpf_verifier_env *env, struct bpf_func_state *old,
- struct bpf_func_state *cur, enum exact_level exact)
+ struct bpf_func_state *cur, u32 insn_idx, enum exact_level exact)
{
- int i;
+ u16 live_regs = env->insn_aux_data[insn_idx].live_regs_before;
+ u16 i;
if (old->callback_depth > cur->callback_depth)
return false;
for (i = 0; i < MAX_BPF_REG; i++)
- if (!regsafe(env, &old->regs[i], &cur->regs[i],
+ if (((1 << i) & live_regs) &&
+ !regsafe(env, &old->regs[i], &cur->regs[i],
&env->idmap_scratch, exact))
return false;
@@ -18529,6 +18531,7 @@ static bool states_equal(struct bpf_verifier_env *env,
struct bpf_verifier_state *cur,
enum exact_level exact)
{
+ u32 insn_idx;
int i;
if (old->curframe != cur->curframe)
@@ -18552,9 +18555,12 @@ static bool states_equal(struct bpf_verifier_env *env,
* and all frame states need to be equivalent
*/
for (i = 0; i <= old->curframe; i++) {
+ insn_idx = i == old->curframe
+ ? env->insn_idx
+ : old->frame[i + 1]->callsite;
if (old->frame[i]->callsite != cur->frame[i]->callsite)
return false;
- if (!func_states_equal(env, old->frame[i], cur->frame[i], exact))
+ if (!func_states_equal(env, old->frame[i], cur->frame[i], insn_idx, exact))
return false;
}
return true;
--
2.48.1
next prev parent reply other threads:[~2025-03-04 7:43 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-03-04 7:42 [PATCH bpf-next v2 0/5] bpf: simple DFA-based live registers analysis Eduard Zingerman
2025-03-04 7:42 ` [PATCH bpf-next v2 1/5] bpf: jmp_offset() and verbose_insn() utility functions Eduard Zingerman
2025-03-04 7:42 ` [PATCH bpf-next v2 2/5] bpf: get_call_summary() utility function Eduard Zingerman
2025-03-04 7:42 ` [PATCH bpf-next v2 3/5] bpf: simple DFA-based live registers analysis Eduard Zingerman
2025-03-04 17:00 ` Alexei Starovoitov
2025-03-04 17:55 ` Eduard Zingerman
2025-03-04 7:42 ` Eduard Zingerman [this message]
2025-03-04 7:42 ` [PATCH bpf-next v2 5/5] selftests/bpf: test cases for compute_live_registers() 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=20250304074239.2328752-5-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=kernel-team@fb.com \
--cc=martin.lau@linux.dev \
--cc=tj@kernel.org \
--cc=yonghong.song@linux.dev \
/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.