From: Donglin Peng <dolinux.peng@gmail.com>
To: ast@kernel.org, andrii.nakryiko@gmail.com
Cc: eddyz87@gmail.com, zhangxiaoqin@xiaomi.com,
ihor.solodrai@linux.dev, linux-kernel@vger.kernel.org,
bpf@vger.kernel.org, pengdonglin <pengdonglin@xiaomi.com>,
Alan Maguire <alan.maguire@oracle.com>
Subject: [PATCH bpf-next v9 08/10] bpf: Skip anonymous types in type lookup for performance
Date: Mon, 8 Dec 2025 14:23:51 +0800 [thread overview]
Message-ID: <20251208062353.1702672-9-dolinux.peng@gmail.com> (raw)
In-Reply-To: <20251208062353.1702672-1-dolinux.peng@gmail.com>
From: pengdonglin <pengdonglin@xiaomi.com>
Currently, vmlinux and kernel module BTFs are unconditionally
sorted during the build phase, with named types placed at the
end. Thus, anonymous types should be skipped when starting the
search. In my vmlinux BTF, the number of anonymous types is
61,747, which means the loop count can be reduced by 61,747.
Cc: Eduard Zingerman <eddyz87@gmail.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Cc: Alan Maguire <alan.maguire@oracle.com>
Cc: Ihor Solodrai <ihor.solodrai@linux.dev>
Cc: Xiaoqin Zhang <zhangxiaoqin@xiaomi.com>
Signed-off-by: pengdonglin <pengdonglin@xiaomi.com>
---
include/linux/btf.h | 1 +
kernel/bpf/btf.c | 15 ++++++++++-----
kernel/bpf/verifier.c | 7 +------
3 files changed, 12 insertions(+), 11 deletions(-)
diff --git a/include/linux/btf.h b/include/linux/btf.h
index f06976ffb63f..2d28f2b22ae5 100644
--- a/include/linux/btf.h
+++ b/include/linux/btf.h
@@ -220,6 +220,7 @@ bool btf_is_module(const struct btf *btf);
bool btf_is_vmlinux(const struct btf *btf);
struct module *btf_try_get_module(const struct btf *btf);
u32 btf_nr_types(const struct btf *btf);
+u32 btf_sorted_start_id(const struct btf *btf);
struct btf *btf_base_btf(const struct btf *btf);
bool btf_type_is_i32(const struct btf_type *t);
bool btf_type_is_i64(const struct btf_type *t);
diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c
index 925cb524f3a8..5f4f51b0acf4 100644
--- a/kernel/bpf/btf.c
+++ b/kernel/bpf/btf.c
@@ -550,6 +550,11 @@ u32 btf_nr_types(const struct btf *btf)
return total;
}
+u32 btf_sorted_start_id(const struct btf *btf)
+{
+ return btf->sorted_start_id ?: (btf->start_id ?: 1);
+}
+
/*
* Assuming that types are sorted by name in ascending order.
*/
@@ -3540,9 +3545,9 @@ const char *btf_find_decl_tag_value(const struct btf *btf, const struct btf_type
{
const char *value = NULL;
const struct btf_type *t;
- int len, id;
+ int len, id = btf->sorted_start_id > 0 ? btf->sorted_start_id - 1 : 0;
- id = btf_find_next_decl_tag(btf, pt, comp_idx, tag_key, 0);
+ id = btf_find_next_decl_tag(btf, pt, comp_idx, tag_key, id);
if (id < 0)
return ERR_PTR(id);
@@ -7859,7 +7864,7 @@ int btf_prepare_func_args(struct bpf_verifier_env *env, int subprog)
*/
for (i = 0; i < nargs; i++) {
u32 tags = 0;
- int id = 0;
+ int id = btf->sorted_start_id > 0 ? btf->sorted_start_id - 1 : 0;
/* 'arg:<tag>' decl_tag takes precedence over derivation of
* register type from BTF type itself
@@ -9340,7 +9345,7 @@ bpf_core_find_cands(struct bpf_core_ctx *ctx, u32 local_type_id)
}
/* Attempt to find target candidates in vmlinux BTF first */
- cands = bpf_core_add_cands(cands, main_btf, 1);
+ cands = bpf_core_add_cands(cands, main_btf, main_btf->sorted_start_id);
if (IS_ERR(cands))
return ERR_CAST(cands);
@@ -9372,7 +9377,7 @@ bpf_core_find_cands(struct bpf_core_ctx *ctx, u32 local_type_id)
*/
btf_get(mod_btf);
spin_unlock_bh(&btf_idr_lock);
- cands = bpf_core_add_cands(cands, mod_btf, btf_nr_types(main_btf));
+ cands = bpf_core_add_cands(cands, mod_btf, mod_btf->sorted_start_id);
btf_put(mod_btf);
if (IS_ERR(cands))
return ERR_CAST(cands);
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index f0ca69f888fa..2ae87075db6a 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -20655,12 +20655,7 @@ static int find_btf_percpu_datasec(struct btf *btf)
* types to look at only module's own BTF types.
*/
n = btf_nr_types(btf);
- if (btf_is_module(btf))
- i = btf_nr_types(btf_vmlinux);
- else
- i = 1;
-
- for(; i < n; i++) {
+ for (i = btf_sorted_start_id(btf); i < n; i++) {
t = btf_type_by_id(btf, i);
if (BTF_INFO_KIND(t->info) != BTF_KIND_DATASEC)
continue;
--
2.34.1
next prev parent reply other threads:[~2025-12-08 6:24 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-12-08 6:23 [PATCH bpf-next v9 00/10] Improve the performance of BTF type lookups with binary search Donglin Peng
2025-12-08 6:23 ` [PATCH bpf-next v9 01/10] libbpf: Add BTF permutation support for type reordering Donglin Peng
2025-12-16 22:00 ` Eduard Zingerman
2025-12-16 22:47 ` Andrii Nakryiko
2025-12-17 3:30 ` Donglin Peng
2025-12-08 6:23 ` [PATCH bpf-next v9 02/10] selftests/bpf: Add test cases for btf__permute functionality Donglin Peng
2025-12-16 22:00 ` Eduard Zingerman
2025-12-08 6:23 ` [PATCH bpf-next v9 03/10] tools/resolve_btfids: Support BTF sorting feature Donglin Peng
2025-12-16 22:13 ` Eduard Zingerman
2025-12-08 6:23 ` [PATCH bpf-next v9 04/10] libbpf: Optimize type lookup with binary search for sorted BTF Donglin Peng
2025-12-16 23:38 ` Eduard Zingerman
2025-12-16 23:43 ` Eduard Zingerman
2025-12-17 2:46 ` Donglin Peng
2025-12-17 2:32 ` Donglin Peng
2025-12-17 2:34 ` Donglin Peng
2025-12-17 2:57 ` Eduard Zingerman
2025-12-08 6:23 ` [PATCH bpf-next v9 05/10] libbpf: Verify BTF Sorting Donglin Peng
2025-12-17 0:32 ` Eduard Zingerman
2025-12-17 3:19 ` Donglin Peng
2025-12-08 6:23 ` [PATCH bpf-next v9 06/10] btf: Optimize type lookup with binary search Donglin Peng
2025-12-08 6:23 ` [PATCH bpf-next v9 07/10] btf: Verify BTF Sorting Donglin Peng
2025-12-09 3:21 ` Donglin Peng
2025-12-17 0:41 ` Eduard Zingerman
2025-12-17 0:48 ` Eduard Zingerman
2025-12-17 3:26 ` Donglin Peng
2025-12-08 6:23 ` Donglin Peng [this message]
2025-12-08 6:36 ` [PATCH bpf-next v9 08/10] bpf: Skip anonymous types in type lookup for performance Donglin Peng
2025-12-17 6:55 ` Eduard Zingerman
2025-12-17 9:21 ` Donglin Peng
2025-12-17 17:29 ` Eduard Zingerman
2025-12-18 9:16 ` Donglin Peng
2025-12-08 6:23 ` [PATCH bpf-next v9 09/10] bpf: Optimize the performance of find_bpffs_btf_enums Donglin Peng
2025-12-17 6:58 ` Eduard Zingerman
2025-12-08 6:23 ` [PATCH bpf-next v9 10/10] libbpf: Optimize the performance of determine_ptr_size Donglin Peng
2025-12-17 7:06 ` Eduard Zingerman
2025-12-17 8:38 ` Donglin Peng
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=20251208062353.1702672-9-dolinux.peng@gmail.com \
--to=dolinux.peng@gmail.com \
--cc=alan.maguire@oracle.com \
--cc=andrii.nakryiko@gmail.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=eddyz87@gmail.com \
--cc=ihor.solodrai@linux.dev \
--cc=linux-kernel@vger.kernel.org \
--cc=pengdonglin@xiaomi.com \
--cc=zhangxiaoqin@xiaomi.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