* [PATCH v2 bpf-next 1/4] bpf: consider types listed in reg2btf_ids as trusted
2023-07-19 9:29 [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types Anton Protopopov
@ 2023-07-19 9:29 ` Anton Protopopov
2023-07-19 9:29 ` [PATCH v2 bpf-next 2/4] bpf: consider CONST_PTR_TO_MAP as trusted pointer to struct bpf_map Anton Protopopov
` (3 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: Anton Protopopov @ 2023-07-19 9:29 UTC (permalink / raw)
To: Martin KaFai Lau, Alexei Starovoitov, Daniel Borkmann,
Andrii Nakryiko, Song Liu, Yonghong Song, John Fastabend,
KP Singh, Stanislav Fomichev, Hao Luo, Jiri Olsa, Mykola Lysenko,
Shuah Khan, Hou Tao, Joe Stringer, Anton Protopopov, bpf,
linux-kernel, linux-kselftest
The reg2btf_ids array contains a list of types for which we can (and need)
to find a corresponding static BTF id. All the types in the list can be
considered as trusted for purposes of kfuncs.
Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
---
kernel/bpf/verifier.c | 21 ++++++++++++---------
1 file changed, 12 insertions(+), 9 deletions(-)
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 0b9da95331d7..05123feab378 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -5413,12 +5413,24 @@ static bool is_flow_key_reg(struct bpf_verifier_env *env, int regno)
return reg->type == PTR_TO_FLOW_KEYS;
}
+static u32 *reg2btf_ids[__BPF_REG_TYPE_MAX] = {
+#ifdef CONFIG_NET
+ [PTR_TO_SOCKET] = &btf_sock_ids[BTF_SOCK_TYPE_SOCK],
+ [PTR_TO_SOCK_COMMON] = &btf_sock_ids[BTF_SOCK_TYPE_SOCK_COMMON],
+ [PTR_TO_TCP_SOCK] = &btf_sock_ids[BTF_SOCK_TYPE_TCP],
+#endif
+};
+
static bool is_trusted_reg(const struct bpf_reg_state *reg)
{
/* A referenced register is always trusted. */
if (reg->ref_obj_id)
return true;
+ /* Types listed in the reg2btf_ids are always trusted */
+ if (reg2btf_ids[base_type(reg->type)])
+ return true;
+
/* If a register is not referenced, it is trusted if it has the
* MEM_ALLOC or PTR_TRUSTED type modifiers, and no others. Some of the
* other type modifiers may be safe, but we elect to take an opt-in
@@ -10052,15 +10064,6 @@ static bool __btf_type_is_scalar_struct(struct bpf_verifier_env *env,
return true;
}
-
-static u32 *reg2btf_ids[__BPF_REG_TYPE_MAX] = {
-#ifdef CONFIG_NET
- [PTR_TO_SOCKET] = &btf_sock_ids[BTF_SOCK_TYPE_SOCK],
- [PTR_TO_SOCK_COMMON] = &btf_sock_ids[BTF_SOCK_TYPE_SOCK_COMMON],
- [PTR_TO_TCP_SOCK] = &btf_sock_ids[BTF_SOCK_TYPE_TCP],
-#endif
-};
-
enum kfunc_ptr_arg_type {
KF_ARG_PTR_TO_CTX,
KF_ARG_PTR_TO_ALLOC_BTF_ID, /* Allocated object */
--
2.34.1
^ permalink raw reply related [flat|nested] 6+ messages in thread* [PATCH v2 bpf-next 2/4] bpf: consider CONST_PTR_TO_MAP as trusted pointer to struct bpf_map
2023-07-19 9:29 [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types Anton Protopopov
2023-07-19 9:29 ` [PATCH v2 bpf-next 1/4] bpf: consider types listed in reg2btf_ids as trusted Anton Protopopov
@ 2023-07-19 9:29 ` Anton Protopopov
2023-07-19 9:29 ` [PATCH v2 bpf-next 3/4] bpf: make an argument const in the bpf_map_sum_elem_count kfunc Anton Protopopov
` (2 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: Anton Protopopov @ 2023-07-19 9:29 UTC (permalink / raw)
To: Martin KaFai Lau, Alexei Starovoitov, Daniel Borkmann,
Andrii Nakryiko, Song Liu, Yonghong Song, John Fastabend,
KP Singh, Stanislav Fomichev, Hao Luo, Jiri Olsa, Mykola Lysenko,
Shuah Khan, Hou Tao, Joe Stringer, Anton Protopopov, bpf,
linux-kernel, linux-kselftest
Add the BTF id of struct bpf_map to the reg2btf_ids array. This makes the
values of the CONST_PTR_TO_MAP type to be considered as trusted by kfuncs.
This, in turn, allows users to execute trusted kfuncs which accept `struct
bpf_map *` arguments from non-tracing programs.
While exporting the btf_bpf_map_id variable, save some bytes by defining
it as BTF_ID_LIST_GLOBAL_SINGLE (which is u32[1]) and not as BTF_ID_LIST
(which is u32[64]).
Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
---
include/linux/btf_ids.h | 1 +
kernel/bpf/map_iter.c | 3 +--
kernel/bpf/verifier.c | 1 +
3 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/include/linux/btf_ids.h b/include/linux/btf_ids.h
index 00950cc03bff..a3462a9b8e18 100644
--- a/include/linux/btf_ids.h
+++ b/include/linux/btf_ids.h
@@ -267,5 +267,6 @@ MAX_BTF_TRACING_TYPE,
extern u32 btf_tracing_ids[];
extern u32 bpf_cgroup_btf_id[];
extern u32 bpf_local_storage_map_btf_id[];
+extern u32 btf_bpf_map_id[];
#endif
diff --git a/kernel/bpf/map_iter.c b/kernel/bpf/map_iter.c
index d06d3b7150e5..b67996147895 100644
--- a/kernel/bpf/map_iter.c
+++ b/kernel/bpf/map_iter.c
@@ -78,8 +78,7 @@ static const struct seq_operations bpf_map_seq_ops = {
.show = bpf_map_seq_show,
};
-BTF_ID_LIST(btf_bpf_map_id)
-BTF_ID(struct, bpf_map)
+BTF_ID_LIST_GLOBAL_SINGLE(btf_bpf_map_id, struct, bpf_map)
static const struct bpf_iter_seq_info bpf_map_seq_info = {
.seq_ops = &bpf_map_seq_ops,
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 05123feab378..803b91135ca0 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -5419,6 +5419,7 @@ static u32 *reg2btf_ids[__BPF_REG_TYPE_MAX] = {
[PTR_TO_SOCK_COMMON] = &btf_sock_ids[BTF_SOCK_TYPE_SOCK_COMMON],
[PTR_TO_TCP_SOCK] = &btf_sock_ids[BTF_SOCK_TYPE_TCP],
#endif
+ [CONST_PTR_TO_MAP] = btf_bpf_map_id,
};
static bool is_trusted_reg(const struct bpf_reg_state *reg)
--
2.34.1
^ permalink raw reply related [flat|nested] 6+ messages in thread* [PATCH v2 bpf-next 3/4] bpf: make an argument const in the bpf_map_sum_elem_count kfunc
2023-07-19 9:29 [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types Anton Protopopov
2023-07-19 9:29 ` [PATCH v2 bpf-next 1/4] bpf: consider types listed in reg2btf_ids as trusted Anton Protopopov
2023-07-19 9:29 ` [PATCH v2 bpf-next 2/4] bpf: consider CONST_PTR_TO_MAP as trusted pointer to struct bpf_map Anton Protopopov
@ 2023-07-19 9:29 ` Anton Protopopov
2023-07-19 9:29 ` [PATCH v2 bpf-next 4/4] bpf: allow any program to use " Anton Protopopov
2023-07-19 17:00 ` [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types patchwork-bot+netdevbpf
4 siblings, 0 replies; 6+ messages in thread
From: Anton Protopopov @ 2023-07-19 9:29 UTC (permalink / raw)
To: Martin KaFai Lau, Alexei Starovoitov, Daniel Borkmann,
Andrii Nakryiko, Song Liu, Yonghong Song, John Fastabend,
KP Singh, Stanislav Fomichev, Hao Luo, Jiri Olsa, Mykola Lysenko,
Shuah Khan, Hou Tao, Joe Stringer, Anton Protopopov, bpf,
linux-kernel, linux-kselftest
We use the map pointer only to read the counter values, no locking
involved, so mark the argument as const.
Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
---
kernel/bpf/map_iter.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/bpf/map_iter.c b/kernel/bpf/map_iter.c
index b67996147895..011adb41858e 100644
--- a/kernel/bpf/map_iter.c
+++ b/kernel/bpf/map_iter.c
@@ -197,7 +197,7 @@ __diag_push();
__diag_ignore_all("-Wmissing-prototypes",
"Global functions as their definitions will be in vmlinux BTF");
-__bpf_kfunc s64 bpf_map_sum_elem_count(struct bpf_map *map)
+__bpf_kfunc s64 bpf_map_sum_elem_count(const struct bpf_map *map)
{
s64 *pcount;
s64 ret = 0;
--
2.34.1
^ permalink raw reply related [flat|nested] 6+ messages in thread* [PATCH v2 bpf-next 4/4] bpf: allow any program to use the bpf_map_sum_elem_count kfunc
2023-07-19 9:29 [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types Anton Protopopov
` (2 preceding siblings ...)
2023-07-19 9:29 ` [PATCH v2 bpf-next 3/4] bpf: make an argument const in the bpf_map_sum_elem_count kfunc Anton Protopopov
@ 2023-07-19 9:29 ` Anton Protopopov
2023-07-19 17:00 ` [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types patchwork-bot+netdevbpf
4 siblings, 0 replies; 6+ messages in thread
From: Anton Protopopov @ 2023-07-19 9:29 UTC (permalink / raw)
To: Martin KaFai Lau, Alexei Starovoitov, Daniel Borkmann,
Andrii Nakryiko, Song Liu, Yonghong Song, John Fastabend,
KP Singh, Stanislav Fomichev, Hao Luo, Jiri Olsa, Mykola Lysenko,
Shuah Khan, Hou Tao, Joe Stringer, Anton Protopopov, bpf,
linux-kernel, linux-kselftest
Register the bpf_map_sum_elem_count func for all programs, and update the
map_ptr subtest of the test_progs test to test the new functionality.
The usage is allowed as long as the pointer to the map is trusted (when
using tracing programs) or is a const pointer to map, as in the following
example:
struct {
__uint(type, BPF_MAP_TYPE_HASH);
...
} hash SEC(".maps");
...
static inline int some_bpf_prog(void)
{
struct bpf_map *map = (struct bpf_map *)&hash;
__s64 count;
count = bpf_map_sum_elem_count(map);
...
}
Signed-off-by: Anton Protopopov <aspsk@isovalent.com>
---
kernel/bpf/map_iter.c | 2 +-
tools/testing/selftests/bpf/progs/map_ptr_kern.c | 5 +++++
2 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/kernel/bpf/map_iter.c b/kernel/bpf/map_iter.c
index 011adb41858e..6fc9dae9edc8 100644
--- a/kernel/bpf/map_iter.c
+++ b/kernel/bpf/map_iter.c
@@ -226,6 +226,6 @@ static const struct btf_kfunc_id_set bpf_map_iter_kfunc_set = {
static int init_subsystem(void)
{
- return register_btf_kfunc_id_set(BPF_PROG_TYPE_TRACING, &bpf_map_iter_kfunc_set);
+ return register_btf_kfunc_id_set(BPF_PROG_TYPE_UNSPEC, &bpf_map_iter_kfunc_set);
}
late_initcall(init_subsystem);
diff --git a/tools/testing/selftests/bpf/progs/map_ptr_kern.c b/tools/testing/selftests/bpf/progs/map_ptr_kern.c
index db388f593d0a..3325da17ec81 100644
--- a/tools/testing/selftests/bpf/progs/map_ptr_kern.c
+++ b/tools/testing/selftests/bpf/progs/map_ptr_kern.c
@@ -103,6 +103,8 @@ struct {
__type(value, __u32);
} m_hash SEC(".maps");
+__s64 bpf_map_sum_elem_count(struct bpf_map *map) __ksym;
+
static inline int check_hash(void)
{
struct bpf_htab *hash = (struct bpf_htab *)&m_hash;
@@ -115,6 +117,8 @@ static inline int check_hash(void)
VERIFY(hash->elem_size == 64);
VERIFY(hash->count.counter == 0);
+ VERIFY(bpf_map_sum_elem_count(map) == 0);
+
for (i = 0; i < HALF_ENTRIES; ++i) {
const __u32 key = i;
const __u32 val = 1;
@@ -123,6 +127,7 @@ static inline int check_hash(void)
return 0;
}
VERIFY(hash->count.counter == HALF_ENTRIES);
+ VERIFY(bpf_map_sum_elem_count(map) == HALF_ENTRIES);
return 1;
}
--
2.34.1
^ permalink raw reply related [flat|nested] 6+ messages in thread* Re: [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types
2023-07-19 9:29 [PATCH v2 bpf-next 0/4] allow bpf_map_sum_elem_count for all program types Anton Protopopov
` (3 preceding siblings ...)
2023-07-19 9:29 ` [PATCH v2 bpf-next 4/4] bpf: allow any program to use " Anton Protopopov
@ 2023-07-19 17:00 ` patchwork-bot+netdevbpf
4 siblings, 0 replies; 6+ messages in thread
From: patchwork-bot+netdevbpf @ 2023-07-19 17:00 UTC (permalink / raw)
To: Anton Protopopov
Cc: martin.lau, ast, daniel, andrii, song, yhs, john.fastabend,
kpsingh, sdf, haoluo, jolsa, mykolal, shuah, houtao1, joe, bpf,
linux-kernel, linux-kselftest
Hello:
This series was applied to bpf/bpf-next.git (master)
by Alexei Starovoitov <ast@kernel.org>:
On Wed, 19 Jul 2023 09:29:48 +0000 you wrote:
> This series is a follow up to the recent change [1] which added
> per-cpu insert/delete statistics for maps. The bpf_map_sum_elem_count
> kfunc presented in the original series was only available to tracing
> programs, so let's make it available to all.
>
> The first patch makes types listed in the reg2btf_ids[] array to be
> considered trusted by kfuncs.
>
> [...]
Here is the summary with links:
- [v2,bpf-next,1/4] bpf: consider types listed in reg2btf_ids as trusted
https://git.kernel.org/bpf/bpf-next/c/831deb2976de
- [v2,bpf-next,2/4] bpf: consider CONST_PTR_TO_MAP as trusted pointer to struct bpf_map
https://git.kernel.org/bpf/bpf-next/c/5ba190c29cf9
- [v2,bpf-next,3/4] bpf: make an argument const in the bpf_map_sum_elem_count kfunc
https://git.kernel.org/bpf/bpf-next/c/9c29804961c1
- [v2,bpf-next,4/4] bpf: allow any program to use the bpf_map_sum_elem_count kfunc
https://git.kernel.org/bpf/bpf-next/c/72829b1c1f16
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 6+ messages in thread