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, puranjay@kernel.org,
jose.marchesi@oracle.com, Eduard Zingerman <eddyz87@gmail.com>
Subject: [RFC bpf-next v2 3/9] bpf, x86, riscv, arm: no_caller_saved_registers for bpf_get_smp_processor_id()
Date: Thu, 4 Jul 2024 03:23:55 -0700 [thread overview]
Message-ID: <20240704102402.1644916-4-eddyz87@gmail.com> (raw)
In-Reply-To: <20240704102402.1644916-1-eddyz87@gmail.com>
The function bpf_get_smp_processor_id() is processed in a different
way, depending on the arch:
- on x86 verifier replaces call to bpf_get_smp_processor_id() with a
sequence of instructions that modify only r0;
- on riscv64 jit replaces call to bpf_get_smp_processor_id() with a
sequence of instructions that modify only r0;
- on arm64 jit replaces call to bpf_get_smp_processor_id() with a
sequence of instructions that modify only r0 and tmp registers.
These rewrites satisfy attribute no_caller_saved_registers contract.
Allow rewrite of no_caller_saved_registers patterns for
bpf_get_smp_processor_id() in order to use this function as a canary
for no_caller_saved_registers tests.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
kernel/bpf/helpers.c | 1 +
kernel/bpf/verifier.c | 11 +++++++++--
2 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
index 229396172026..26863b162a88 100644
--- a/kernel/bpf/helpers.c
+++ b/kernel/bpf/helpers.c
@@ -158,6 +158,7 @@ const struct bpf_func_proto bpf_get_smp_processor_id_proto = {
.func = bpf_get_smp_processor_id,
.gpl_only = false,
.ret_type = RET_INTEGER,
+ .allow_nocsr = true,
};
BPF_CALL_0(bpf_get_numa_node_id)
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index d16a249b59ad..99115c552e3b 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -16029,7 +16029,14 @@ static u8 get_helper_reg_mask(const struct bpf_func_proto *fn)
*/
static bool verifier_inlines_helper_call(struct bpf_verifier_env *env, s32 imm)
{
- return false;
+ switch (imm) {
+#ifdef CONFIG_X86_64
+ case BPF_FUNC_get_smp_processor_id:
+ return env->prog->jit_requested && bpf_jit_supports_percpu_insn();
+#endif
+ default:
+ return false;
+ }
}
/* If 'insn' is a call that follows no_caller_saved_registers contract
@@ -20703,7 +20710,7 @@ static int do_misc_fixups(struct bpf_verifier_env *env)
#ifdef CONFIG_X86_64
/* Implement bpf_get_smp_processor_id() inline. */
if (insn->imm == BPF_FUNC_get_smp_processor_id &&
- prog->jit_requested && bpf_jit_supports_percpu_insn()) {
+ verifier_inlines_helper_call(env, insn->imm)) {
/* BPF_FUNC_get_smp_processor_id inlining is an
* optimization, so if pcpu_hot.cpu_number is ever
* changed in some incompatible and hard to support
--
2.45.2
next prev parent reply other threads:[~2024-07-04 10:24 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-07-04 10:23 [RFC bpf-next v2 0/9] no_caller_saved_registers attribute for helper calls Eduard Zingerman
2024-07-04 10:23 ` [RFC bpf-next v2 1/9] bpf: add a get_helper_proto() utility function Eduard Zingerman
2024-07-09 23:42 ` Andrii Nakryiko
2024-07-10 0:26 ` Eduard Zingerman
2024-07-04 10:23 ` [RFC bpf-next v2 2/9] bpf: no_caller_saved_registers attribute for helper calls Eduard Zingerman
2024-07-09 23:42 ` Andrii Nakryiko
2024-07-10 3:00 ` Eduard Zingerman
2024-07-10 6:01 ` Andrii Nakryiko
2024-07-10 7:57 ` Eduard Zingerman
2024-07-10 15:36 ` Andrii Nakryiko
2024-07-10 16:15 ` Eduard Zingerman
2024-07-10 17:50 ` Andrii Nakryiko
2024-07-10 18:40 ` Eduard Zingerman
2024-07-10 18:49 ` Andrii Nakryiko
2024-07-10 19:03 ` Eduard Zingerman
2024-07-10 19:16 ` Andrii Nakryiko
2024-07-10 19:07 ` Alexei Starovoitov
2024-07-10 19:17 ` Andrii Nakryiko
2024-07-10 19:01 ` Alexei Starovoitov
2024-07-10 9:46 ` Eduard Zingerman
2024-07-10 15:23 ` Andrii Nakryiko
2024-07-10 1:09 ` Alexei Starovoitov
2024-07-10 3:06 ` Eduard Zingerman
2024-07-04 10:23 ` Eduard Zingerman [this message]
2024-07-04 10:23 ` [RFC bpf-next v2 4/9] selftests/bpf: extract utility function for BPF disassembly Eduard Zingerman
2024-07-09 23:46 ` Andrii Nakryiko
2024-07-04 10:23 ` [RFC bpf-next v2 5/9] selftests/bpf: no need to track next_match_pos in struct test_loader Eduard Zingerman
2024-07-04 10:23 ` [RFC bpf-next v2 6/9] selftests/bpf: extract test_loader->expect_msgs as a data structure Eduard Zingerman
2024-07-04 10:23 ` [RFC bpf-next v2 7/9] selftests/bpf: allow checking xlated programs in verifier_* tests Eduard Zingerman
2024-07-04 10:24 ` [RFC bpf-next v2 8/9] selftests/bpf: __arch_* macro to limit test cases to specific archs Eduard Zingerman
2024-07-09 23:50 ` Andrii Nakryiko
2024-07-04 10:24 ` [RFC bpf-next v2 9/9] selftests/bpf: test no_caller_saved_registers spill/fill removal Eduard Zingerman
2024-07-08 11:44 ` [RFC bpf-next v2 0/9] no_caller_saved_registers attribute for helper calls Puranjay Mohan
2024-07-08 17:29 ` Eduard Zingerman
2024-07-10 1:18 ` Alexei Starovoitov
2024-07-10 3:35 ` 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=20240704102402.1644916-4-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=jose.marchesi@oracle.com \
--cc=kernel-team@fb.com \
--cc=martin.lau@linux.dev \
--cc=puranjay@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