2023-04-20 03:28:45

by Feng Zhou

[permalink] [raw]
Subject: [PATCH bpf-next v2 0/2] Access variable length array relaxed for integer type

From: Feng Zhou <[email protected]>

Add support for integer type of accessing variable length array.
Add a selftest to check it.

Feng Zhou (2):
bpf: support access variable length array of integer type
selftests/bpf: Add test to access integer type of variable array

Changelog:
v1->v2: Addressed comments from Alexei Starovoitov
- Add one more use case.
Details in here:
https://lore.kernel.org/bpf/[email protected]/

kernel/bpf/btf.c | 8 +++++---
.../selftests/bpf/bpf_testmod/bpf_testmod.c | 20 +++++++++++++++++++
.../bpf/prog_tests/access_variable_array.c | 16 +++++++++++++++
.../selftests/bpf/prog_tests/tracing_struct.c | 2 ++
.../bpf/progs/test_access_variable_array.c | 19 ++++++++++++++++++
.../selftests/bpf/progs/tracing_struct.c | 13 ++++++++++++
6 files changed, 75 insertions(+), 3 deletions(-)
create mode 100644 tools/testing/selftests/bpf/prog_tests/access_variable_array.c
create mode 100644 tools/testing/selftests/bpf/progs/test_access_variable_array.c

--
2.20.1


2023-04-20 03:30:13

by Feng Zhou

[permalink] [raw]
Subject: [PATCH bpf-next v2 1/2] bpf: support access variable length array of integer type

From: Feng Zhou <[email protected]>

After this commit:
bpf: Support variable length array in tracing programs (9c5f8a1008a1)
Trace programs can access variable length array, but for structure
type. This patch adds support for integer type.

Example:
Hook load_balance
struct sched_domain {
...
unsigned long span[];
}

The access: sd->span[0].

Co-developed-by: Chengming Zhou <[email protected]>
Signed-off-by: Chengming Zhou <[email protected]>
Signed-off-by: Feng Zhou <[email protected]>
---
kernel/bpf/btf.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c
index 027f9f8a3551..a0887ee44e89 100644
--- a/kernel/bpf/btf.c
+++ b/kernel/bpf/btf.c
@@ -6157,11 +6157,13 @@ static int btf_struct_walk(struct bpf_verifier_log *log, const struct btf *btf,
if (off < moff)
goto error;

- /* Only allow structure for now, can be relaxed for
- * other types later.
- */
+ /* allow structure and integer */
t = btf_type_skip_modifiers(btf, array_elem->type,
NULL);
+
+ if (btf_type_is_int(t))
+ return WALK_SCALAR;
+
if (!btf_type_is_struct(t))
goto error;

--
2.20.1

2023-04-20 03:32:55

by Feng Zhou

[permalink] [raw]
Subject: [PATCH bpf-next v2 2/2] selftests/bpf: Add test to access integer type of variable array

From: Feng Zhou <[email protected]>

Add prog test for accessing integer type of variable array in tracing
program.
In addition, hook load_balance function to access sd->span[0], only
to confirm whether the load is successful. Because there is no direct
way to trigger load_balance call.

Co-developed-by: Chengming Zhou <[email protected]>
Signed-off-by: Chengming Zhou <[email protected]>
Signed-off-by: Feng Zhou <[email protected]>
---
.../selftests/bpf/bpf_testmod/bpf_testmod.c | 20 +++++++++++++++++++
.../bpf/prog_tests/access_variable_array.c | 16 +++++++++++++++
.../selftests/bpf/prog_tests/tracing_struct.c | 2 ++
.../bpf/progs/test_access_variable_array.c | 19 ++++++++++++++++++
.../selftests/bpf/progs/tracing_struct.c | 13 ++++++++++++
5 files changed, 70 insertions(+)
create mode 100644 tools/testing/selftests/bpf/prog_tests/access_variable_array.c
create mode 100644 tools/testing/selftests/bpf/progs/test_access_variable_array.c

diff --git a/tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c b/tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c
index fe847ebfb731..52785ba671e6 100644
--- a/tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c
+++ b/tools/testing/selftests/bpf/bpf_testmod/bpf_testmod.c
@@ -28,6 +28,11 @@ struct bpf_testmod_struct_arg_2 {
long b;
};

+struct bpf_testmod_struct_arg_3 {
+ int a;
+ int b[];
+};
+
__diag_push();
__diag_ignore_all("-Wmissing-prototypes",
"Global functions as their definitions will be in bpf_testmod.ko BTF");
@@ -63,6 +68,12 @@ bpf_testmod_test_struct_arg_5(void) {
return bpf_testmod_test_struct_arg_result;
}

+noinline int
+bpf_testmod_test_struct_arg_6(struct bpf_testmod_struct_arg_3 *a) {
+ bpf_testmod_test_struct_arg_result = a->b[0];
+ return bpf_testmod_test_struct_arg_result;
+}
+
__bpf_kfunc void
bpf_testmod_test_mod_kfunc(int i)
{
@@ -195,6 +206,7 @@ bpf_testmod_test_read(struct file *file, struct kobject *kobj,
};
struct bpf_testmod_struct_arg_1 struct_arg1 = {10};
struct bpf_testmod_struct_arg_2 struct_arg2 = {2, 3};
+ struct bpf_testmod_struct_arg_3 *struct_arg3;
int i = 1;

while (bpf_testmod_return_ptr(i))
@@ -206,6 +218,14 @@ bpf_testmod_test_read(struct file *file, struct kobject *kobj,
(void)bpf_testmod_test_struct_arg_4(struct_arg1, 1, 2, 3, struct_arg2);
(void)bpf_testmod_test_struct_arg_5();

+ struct_arg3 = kmalloc((sizeof(struct bpf_testmod_struct_arg_3) +
+ sizeof(int)), GFP_KERNEL);
+ if (struct_arg3 != NULL) {
+ struct_arg3->b[0] = 1;
+ (void)bpf_testmod_test_struct_arg_6(struct_arg3);
+ kfree(struct_arg3);
+ }
+
/* This is always true. Use the check to make sure the compiler
* doesn't remove bpf_testmod_loop_test.
*/
diff --git a/tools/testing/selftests/bpf/prog_tests/access_variable_array.c b/tools/testing/selftests/bpf/prog_tests/access_variable_array.c
new file mode 100644
index 000000000000..08131782437c
--- /dev/null
+++ b/tools/testing/selftests/bpf/prog_tests/access_variable_array.c
@@ -0,0 +1,16 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2022 Bytedance */
+
+#include <test_progs.h>
+#include "test_access_variable_array.skel.h"
+
+void test_access_variable_array(void)
+{
+ struct test_access_variable_array *skel;
+
+ skel = test_access_variable_array__open_and_load();
+ if (!ASSERT_OK_PTR(skel, "test_access_variable_array__open_and_load"))
+ return;
+
+ test_access_variable_array__destroy(skel);
+}
diff --git a/tools/testing/selftests/bpf/prog_tests/tracing_struct.c b/tools/testing/selftests/bpf/prog_tests/tracing_struct.c
index 48dc9472e160..1c75a32186d6 100644
--- a/tools/testing/selftests/bpf/prog_tests/tracing_struct.c
+++ b/tools/testing/selftests/bpf/prog_tests/tracing_struct.c
@@ -53,6 +53,8 @@ static void test_fentry(void)

ASSERT_EQ(skel->bss->t5_ret, 1, "t5 ret");

+ ASSERT_EQ(skel->bss->t6, 1, "t6 ret");
+
tracing_struct__detach(skel);
destroy_skel:
tracing_struct__destroy(skel);
diff --git a/tools/testing/selftests/bpf/progs/test_access_variable_array.c b/tools/testing/selftests/bpf/progs/test_access_variable_array.c
new file mode 100644
index 000000000000..808c49b79889
--- /dev/null
+++ b/tools/testing/selftests/bpf/progs/test_access_variable_array.c
@@ -0,0 +1,19 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2023 Bytedance */
+
+#include "vmlinux.h"
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+
+unsigned long span = 0;
+
+SEC("fentry/load_balance")
+int BPF_PROG(fentry_fentry, int this_cpu, struct rq *this_rq,
+ struct sched_domain *sd)
+{
+ span = sd->span[0];
+
+ return 0;
+}
+
+char _license[] SEC("license") = "GPL";
diff --git a/tools/testing/selftests/bpf/progs/tracing_struct.c b/tools/testing/selftests/bpf/progs/tracing_struct.c
index e718f0ebee7d..c435a3a8328a 100644
--- a/tools/testing/selftests/bpf/progs/tracing_struct.c
+++ b/tools/testing/selftests/bpf/progs/tracing_struct.c
@@ -13,12 +13,18 @@ struct bpf_testmod_struct_arg_2 {
long b;
};

+struct bpf_testmod_struct_arg_3 {
+ int a;
+ int b[];
+};
+
long t1_a_a, t1_a_b, t1_b, t1_c, t1_ret, t1_nregs;
__u64 t1_reg0, t1_reg1, t1_reg2, t1_reg3;
long t2_a, t2_b_a, t2_b_b, t2_c, t2_ret;
long t3_a, t3_b, t3_c_a, t3_c_b, t3_ret;
long t4_a_a, t4_b, t4_c, t4_d, t4_e_a, t4_e_b, t4_ret;
long t5_ret;
+int t6;

SEC("fentry/bpf_testmod_test_struct_arg_1")
int BPF_PROG2(test_struct_arg_1, struct bpf_testmod_struct_arg_2, a, int, b, int, c)
@@ -117,4 +123,11 @@ int BPF_PROG2(test_struct_arg_10, int, ret)
return 0;
}

+SEC("fentry/bpf_testmod_test_struct_arg_6")
+int BPF_PROG2(test_struct_arg_11, struct bpf_testmod_struct_arg_3 *, a)
+{
+ t6 = a->b[0];
+ return 0;
+}
+
char _license[] SEC("license") = "GPL";
--
2.20.1

2023-04-20 04:49:51

by patchwork-bot+netdevbpf

[permalink] [raw]
Subject: Re: [PATCH bpf-next v2 0/2] Access variable length array relaxed for integer type

Hello:

This series was applied to bpf/bpf-next.git (master)
by Alexei Starovoitov <[email protected]>:

On Thu, 20 Apr 2023 11:27:33 +0800 you wrote:
> From: Feng Zhou <[email protected]>
>
> Add support for integer type of accessing variable length array.
> Add a selftest to check it.
>
> Feng Zhou (2):
> bpf: support access variable length array of integer type
> selftests/bpf: Add test to access integer type of variable array
>
> [...]

Here is the summary with links:
- [bpf-next,v2,1/2] bpf: support access variable length array of integer type
https://git.kernel.org/bpf/bpf-next/c/2569c7b8726f
- [bpf-next,v2,2/2] selftests/bpf: Add test to access integer type of variable array
https://git.kernel.org/bpf/bpf-next/c/5ff54dedf35b

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html