From: Menglong Dong <menglong.dong@linux.dev>
To: Alexei Starovoitov <alexei.starovoitov@gmail.com>,
Menglong Dong <menglong8.dong@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>,
Jiri Olsa <jolsa@kernel.org>, bpf <bpf@vger.kernel.org>,
Menglong Dong <dongml2@chinatelecom.cn>,
"H. Peter Anvin" <hpa@zytor.com>,
Martin KaFai Lau <martin.lau@linux.dev>,
Eduard Zingerman <eddyz87@gmail.com>, Song Liu <song@kernel.org>,
Yonghong Song <yonghong.song@linux.dev>,
John Fastabend <john.fastabend@gmail.com>,
KP Singh <kpsingh@kernel.org>,
Stanislav Fomichev <sdf@fomichev.me>, Hao Luo <haoluo@google.com>,
LKML <linux-kernel@vger.kernel.org>,
Network Development <netdev@vger.kernel.org>
Subject: Re: [PATCH bpf-next v2 02/18] x86,bpf: add bpf_global_caller for global trampoline
Date: Tue, 15 Jul 2025 16:36:57 +0800 [thread overview]
Message-ID: <45f4d349-7b08-45d3-9bec-3ab75217f9b6@linux.dev> (raw)
In-Reply-To: <CAADnVQKP1-gdmq1xkogFeRM6o3j2zf0Q8Atz=aCEkB0PkVx++A@mail.gmail.com>
On 7/15/25 10:25, Alexei Starovoitov wrote:
> On Thu, Jul 3, 2025 at 5:17 AM Menglong Dong <menglong8.dong@gmail.com> wrote:
>> +static __always_inline void
>> +do_origin_call(unsigned long *args, unsigned long *ip, int nr_args)
>> +{
>> + /* Following code will be optimized by the compiler, as nr_args
>> + * is a const, and there will be no condition here.
>> + */
>> + if (nr_args == 0) {
>> + asm volatile(
>> + RESTORE_ORIGIN_0 CALL_NOSPEC "\n"
>> + "movq %%rax, %0\n"
>> + : "=m"(args[nr_args]), ASM_CALL_CONSTRAINT
>> + : [args]"r"(args), [thunk_target]"r"(*ip)
>> + :
>> + );
>> + } else if (nr_args == 1) {
>> + asm volatile(
>> + RESTORE_ORIGIN_1 CALL_NOSPEC "\n"
>> + "movq %%rax, %0\n"
>> + : "=m"(args[nr_args]), ASM_CALL_CONSTRAINT
>> + : [args]"r"(args), [thunk_target]"r"(*ip)
>> + : "rdi"
>> + );
>> + } else if (nr_args == 2) {
>> + asm volatile(
>> + RESTORE_ORIGIN_2 CALL_NOSPEC "\n"
>> + "movq %%rax, %0\n"
>> + : "=m"(args[nr_args]), ASM_CALL_CONSTRAINT
>> + : [args]"r"(args), [thunk_target]"r"(*ip)
>> + : "rdi", "rsi"
>> + );
>> + } else if (nr_args == 3) {
>> + asm volatile(
>> + RESTORE_ORIGIN_3 CALL_NOSPEC "\n"
>> + "movq %%rax, %0\n"
>> + : "=m"(args[nr_args]), ASM_CALL_CONSTRAINT
>> + : [args]"r"(args), [thunk_target]"r"(*ip)
>> + : "rdi", "rsi", "rdx"
>> + );
>> + } else if (nr_args == 4) {
>> + asm volatile(
>> + RESTORE_ORIGIN_4 CALL_NOSPEC "\n"
>> + "movq %%rax, %0\n"
>> + : "=m"(args[nr_args]), ASM_CALL_CONSTRAINT
>> + : [args]"r"(args), [thunk_target]"r"(*ip)
>> + : "rdi", "rsi", "rdx", "rcx"
>> + );
>> + } else if (nr_args == 5) {
>> + asm volatile(
>> + RESTORE_ORIGIN_5 CALL_NOSPEC "\n"
>> + "movq %%rax, %0\n"
>> + : "=m"(args[nr_args]), ASM_CALL_CONSTRAINT
>> + : [args]"r"(args), [thunk_target]"r"(*ip)
>> + : "rdi", "rsi", "rdx", "rcx", "r8"
>> + );
>> + } else if (nr_args == 6) {
>> + asm volatile(
>> + RESTORE_ORIGIN_6 CALL_NOSPEC "\n"
>> + "movq %%rax, %0\n"
>> + : "=m"(args[nr_args]), ASM_CALL_CONSTRAINT
>> + : [args]"r"(args), [thunk_target]"r"(*ip)
>> + : "rdi", "rsi", "rdx", "rcx", "r8", "r9"
>> + );
>> + }
>> +}
> What is the performance difference between 0-6 variants?
> I would think save/restore of regs shouldn't be that expensive.
> bpf trampoline saves only what's necessary because it can do
> this micro optimization, but for this one, I think, doing
> _one_ global trampoline that covers all cases will simplify the code
> a lot, but please benchmark the difference to understand
> the trade-off.
According to my benchmark, it has ~5% overhead to save/restore
*5* variants when compared with *0* variant. The save/restore of regs
is fast, but it still need 12 insn, which can produce ~6% overhead.
I think the performance is more import and we should keep this logic.
Should we? If you think the do_origin_call() is not simple enough, we
can recover all the 6 regs from the stack directly for the origin call,
which won't
introduce too much overhead, and keep the save/restore logic.
What do you think?
>
> The major simplification will be due to skipping nr_args.
> There won't be a need to do btf model and count the args.
> Just do one trampoline for them all.
>
> Also funcs with 7+ arguments need to be thought through
> from the start.
In the current version, the attachment will be rejected if any functions
have
7+ arguments.
> I think it's ok trade-off if we allow global trampoline
> to be safe to attach to a function with 7+ args (and
> it will not mess with the stack), but bpf prog can only
> access up to 6 args. The kfuncs to access arg 7 might be
> more complex and slower. It's ok trade off.
It's OK for fentry-multi, but we can't allow fexit-multi and
modify_return-multi
to be attached to the function with 7+ args, as we need to do the origin
call, and we can't recover the arguments in the stack for the origin
call for now.
So we can allow the functions with 7+ args to be attached as long as the
accessed
arguments are all in regs for fentry-multi. And I think we need one more
patch to
do the "all accessed arguments are in regs" checking, so maybe we can
put it in
the next series? As current series is a little complex :/
Anyway, I'll have a try to see if we can add this part in this series :)
>
>> +
>> +static __always_inline notrace void
>> +run_tramp_prog(struct kfunc_md_tramp_prog *tramp_prog,
>> + struct bpf_tramp_run_ctx *run_ctx, unsigned long *args)
>> +{
>> + struct bpf_prog *prog;
>> + u64 start_time;
>> +
>> + while (tramp_prog) {
>> + prog = tramp_prog->prog;
>> + run_ctx->bpf_cookie = tramp_prog->cookie;
>> + start_time = bpf_gtramp_enter(prog, run_ctx);
>> +
>> + if (likely(start_time)) {
>> + asm volatile(
>> + CALL_NOSPEC "\n"
>> + : : [thunk_target]"r"(prog->bpf_func), [args]"D"(args)
>> + );
> Why this cannot be "call *(prog->bpf_func)" ?
Do you mean "prog->bpf_func(args, NULL);"? In my previous testing, this
cause
bad performance, and I see others do the indirect call in this way. And
I just do
the benchmark again, it seems the performance is not affected in this
way anymore.
So I think I can replace it with "prog->bpf_func(args, NULL);" in the
next version.
>
>> + }
>> +
>> + bpf_gtramp_exit(prog, start_time, run_ctx);
>> + tramp_prog = tramp_prog->next;
>> + }
>> +}
>> +
>> +static __always_inline notrace int
>> +bpf_global_caller_run(unsigned long *args, unsigned long *ip, int nr_args)
> Pls share top 10 from "perf report" while running the bench.
> I'm curious about what's hot.
> Last time I benchmarked fentry/fexit migrate_disable/enable were
> one the hottest functions. I suspect it's the case here as well.
You are right, the migrate_disable/enable are the hottest functions in
both bpf trampoline and global trampoline. Following is the perf top
for fentry-multi:
36.36% bpf_prog_2dcccf652aac1793_bench_trigger_fentry_multi [k]
bpf_prog_2dcccf652aac1793_bench_trigger_fentry_multi 20.54% [kernel] [k]
migrate_enable 19.35% [kernel] [k] bpf_global_caller_5_run 6.52%
[kernel] [k] bpf_global_caller_5 3.58% libc.so.6 [.] syscall 2.88%
[kernel] [k] entry_SYSCALL_64 1.50% [kernel] [k] memchr_inv 1.39%
[kernel] [k] fput 1.04% [kernel] [k] migrate_disable 0.91% [kernel] [k]
_copy_to_user
And I also did the testing for fentry:
54.63% bpf_prog_2dcccf652aac1793_bench_trigger_fentry [k]
bpf_prog_2dcccf652aac1793_bench_trigger_fentry
10.43% [kernel] [k] migrate_enable
10.07% bpf_trampoline_6442517037 [k] bpf_trampoline_6442517037
8.06% [kernel] [k] __bpf_prog_exit_recur 4.11% libc.so.6 [.] syscall
2.15% [kernel] [k] entry_SYSCALL_64 1.48% [kernel] [k] memchr_inv 1.32%
[kernel] [k] fput 1.16% [kernel] [k] _copy_to_user 0.73% [kernel] [k]
bpf_prog_test_run_raw_tp
The migrate_enable/disable are used to do the recursive checking,
and I even wanted to perform recursive checks in the same way as
ftrace to eliminate this overhead :/
>
next prev parent reply other threads:[~2025-07-15 8:38 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20250703121521.1874196-1-dongml2@chinatelecom.cn>
2025-07-03 12:15 ` [PATCH bpf-next v2 01/18] bpf: add function hash table for tracing-multi Menglong Dong
2025-07-04 16:07 ` kernel test robot
2025-07-15 1:55 ` Alexei Starovoitov
2025-07-15 2:37 ` Menglong Dong
2025-07-15 2:49 ` Alexei Starovoitov
2025-07-15 3:13 ` Menglong Dong
2025-07-15 9:06 ` Menglong Dong
2025-07-15 16:22 ` Alexei Starovoitov
2025-07-03 12:15 ` [PATCH bpf-next v2 02/18] x86,bpf: add bpf_global_caller for global trampoline Menglong Dong
2025-07-15 2:25 ` Alexei Starovoitov
2025-07-15 8:36 ` Menglong Dong [this message]
2025-07-15 9:30 ` Menglong Dong
2025-07-16 16:56 ` Inlining migrate_disable/enable. Was: " Alexei Starovoitov
2025-07-16 18:24 ` Peter Zijlstra
2025-07-16 22:35 ` Alexei Starovoitov
2025-07-16 22:49 ` Steven Rostedt
2025-07-16 22:50 ` Steven Rostedt
2025-07-28 9:20 ` Menglong Dong
2025-07-31 16:15 ` Alexei Starovoitov
2025-08-01 1:42 ` Menglong Dong
2025-08-06 8:44 ` Menglong Dong
2025-08-08 0:58 ` Alexei Starovoitov
2025-08-08 5:48 ` Menglong Dong
2025-08-08 6:32 ` Menglong Dong
2025-08-08 15:47 ` Alexei Starovoitov
2025-07-15 16:35 ` Alexei Starovoitov
2025-07-16 13:05 ` Menglong Dong
2025-07-17 0:59 ` multi-fentry proposal. Was: " Alexei Starovoitov
2025-07-17 1:50 ` Menglong Dong
2025-07-17 2:13 ` Alexei Starovoitov
2025-07-17 2:37 ` Menglong Dong
2025-07-16 14:40 ` Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 03/18] ftrace: factor out ftrace_direct_update from register_ftrace_direct Menglong Dong
2025-07-05 2:41 ` kernel test robot
2025-07-03 12:15 ` [PATCH bpf-next v2 04/18] ftrace: add reset_ftrace_direct_ips Menglong Dong
2025-07-03 15:30 ` Steven Rostedt
2025-07-04 1:54 ` Menglong Dong
2025-07-07 18:52 ` Steven Rostedt
2025-07-08 1:26 ` Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 05/18] bpf: introduce bpf_gtramp_link Menglong Dong
2025-07-04 7:00 ` kernel test robot
2025-07-04 7:52 ` kernel test robot
2025-07-03 12:15 ` [PATCH bpf-next v2 06/18] bpf: tracing: add support to record and check the accessed args Menglong Dong
2025-07-14 22:07 ` Andrii Nakryiko
2025-07-14 23:45 ` Menglong Dong
2025-07-15 17:11 ` Andrii Nakryiko
2025-07-16 12:50 ` Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 07/18] bpf: refactor the modules_array to ptr_array Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 08/18] bpf: verifier: add btf to the function args of bpf_check_attach_target Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 09/18] bpf: verifier: move btf_id_deny to bpf_check_attach_target Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 10/18] x86,bpf: factor out arch_bpf_get_regs_nr Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 11/18] bpf: tracing: add multi-link support Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 12/18] libbpf: don't free btf if tracing_multi progs existing Menglong Dong
2025-07-14 22:07 ` Andrii Nakryiko
2025-07-15 1:15 ` Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 13/18] libbpf: support tracing_multi Menglong Dong
2025-07-14 22:07 ` Andrii Nakryiko
2025-07-15 1:58 ` Menglong Dong
2025-07-15 17:20 ` Andrii Nakryiko
2025-07-16 12:43 ` Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 14/18] libbpf: add btf type hash lookup support Menglong Dong
2025-07-14 22:07 ` Andrii Nakryiko
2025-07-15 4:40 ` Menglong Dong
2025-07-15 17:20 ` Andrii Nakryiko
2025-07-16 11:53 ` Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 15/18] libbpf: add skip_invalid and attach_tracing for tracing_multi Menglong Dong
2025-07-14 22:07 ` Andrii Nakryiko
2025-07-15 5:48 ` Menglong Dong
2025-07-15 17:23 ` Andrii Nakryiko
2025-07-16 11:46 ` Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 16/18] selftests/bpf: move get_ksyms and get_addrs to trace_helpers.c Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 17/18] selftests/bpf: add basic testcases for tracing_multi Menglong Dong
2025-07-03 12:15 ` [PATCH bpf-next v2 18/18] selftests/bpf: add bench tests " Menglong Dong
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=45f4d349-7b08-45d3-9bec-3ab75217f9b6@linux.dev \
--to=menglong.dong@linux.dev \
--cc=alexei.starovoitov@gmail.com \
--cc=bpf@vger.kernel.org \
--cc=dongml2@chinatelecom.cn \
--cc=eddyz87@gmail.com \
--cc=haoluo@google.com \
--cc=hpa@zytor.com \
--cc=john.fastabend@gmail.com \
--cc=jolsa@kernel.org \
--cc=kpsingh@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=martin.lau@linux.dev \
--cc=menglong8.dong@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=rostedt@goodmis.org \
--cc=sdf@fomichev.me \
--cc=song@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).