[PATCH bpf-next v5 0/4] Add prog_kfunc feature probe

[PATCH bpf-next v5 0/4] Add prog_kfunc feature probe

[Tao Chen]

More and more kfunc functions are being added to the kernel.
Different prog types have different restrictions when using kfunc.
Therefore, prog_kfunc probe is added to check whether it is supported,
and the use of this api will be added to bpftool later.

Change list:
- v4 -> v5:
  - use fd_array on stack
  - declare the scope of use of btf_fd
- v4
  https://lore.kernel.org/bpf/20250206051557.27913-1-chen.dylane@gmail.com/
  
- v3 -> v4:
  - add fd_array init for kfunc in mod btf
  - add test case for kfunc in mod btf
  - refactor common part as prog load type check for
    libbpf_probe_bpf_{helper,kfunc}
- v3
  https://lore.kernel.org/bpf/20250124144411.13468-1-chen.dylane@gmail.com

- v2 -> v3:
  - rename parameter off with btf_fd
  - extract the common part for libbpf_probe_bpf_{helper,kfunc}
- v2
  https://lore.kernel.org/bpf/20250123170555.291896-1-chen.dylane@gmail.com

- v1 -> v2:
  - check unsupported prog type like probe_bpf_helper
  - add off parameter for module btf
  - check verifier info when kfunc id invalid
- v1
  https://lore.kernel.org/bpf/20250122171359.232791-1-chen.dylane@gmail.com

Tao Chen (4):
  libbpf: Extract prog load type check from libbpf_probe_bpf_helper
  libbpf: Init fd_array when prog probe load
  libbpf: Add libbpf_probe_bpf_kfunc API
  selftests/bpf: Add libbpf_probe_bpf_kfunc API selftests

 tools/lib/bpf/libbpf.h                        |  19 ++-
 tools/lib/bpf/libbpf.map                      |   1 +
 tools/lib/bpf/libbpf_probes.c                 |  84 +++++++++++--
 .../selftests/bpf/prog_tests/libbpf_probes.c  | 118 ++++++++++++++++++
 4 files changed, 208 insertions(+), 14 deletions(-)

-- 
2.43.0

[PATCH bpf-next v5 1/4] libbpf: Extract prog load type check from libbpf_probe_bpf_helper

[Tao Chen]

Extract prog load type check part from libbpf_probe_bpf_helper
suggested by Andrii, which will be used in both
libbpf_probe_bpf_{helper, kfunc}.

Signed-off-by: Tao Chen <chen.dylane@gmail.com>
---
 tools/lib/bpf/libbpf_probes.c | 29 ++++++++++++++++++-----------
 1 file changed, 18 insertions(+), 11 deletions(-)

diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c
index 9dfbe7750f56..aeb4fd97d801 100644
--- a/tools/lib/bpf/libbpf_probes.c
+++ b/tools/lib/bpf/libbpf_probes.c
@@ -413,6 +413,23 @@ int libbpf_probe_bpf_map_type(enum bpf_map_type map_type, const void *opts)
 	return libbpf_err(ret);
 }
 
+static bool can_probe_prog_type(enum bpf_prog_type prog_type)
+{
+	/* we can't successfully load all prog types to check for BPF helper
+	 * and kfunc support.
+	 */
+	switch (prog_type) {
+	case BPF_PROG_TYPE_TRACING:
+	case BPF_PROG_TYPE_EXT:
+	case BPF_PROG_TYPE_LSM:
+	case BPF_PROG_TYPE_STRUCT_OPS:
+		return false;
+	default:
+		break;
+	}
+	return true;
+}
+
 int libbpf_probe_bpf_helper(enum bpf_prog_type prog_type, enum bpf_func_id helper_id,
 			    const void *opts)
 {
@@ -427,18 +444,8 @@ int libbpf_probe_bpf_helper(enum bpf_prog_type prog_type, enum bpf_func_id helpe
 	if (opts)
 		return libbpf_err(-EINVAL);
 
-	/* we can't successfully load all prog types to check for BPF helper
-	 * support, so bail out with -EOPNOTSUPP error
-	 */
-	switch (prog_type) {
-	case BPF_PROG_TYPE_TRACING:
-	case BPF_PROG_TYPE_EXT:
-	case BPF_PROG_TYPE_LSM:
-	case BPF_PROG_TYPE_STRUCT_OPS:
+	if (!can_probe_prog_type(prog_type))
 		return -EOPNOTSUPP;
-	default:
-		break;
-	}
 
 	buf[0] = '\0';
 	ret = probe_prog_load(prog_type, insns, insn_cnt, buf, sizeof(buf));
-- 
2.43.0

[PATCH bpf-next v5 2/4] libbpf: Init fd_array when prog probe load

[Tao Chen]

fd_array used to store module btf fd, which will
be used for kfunc probe in module btf.

Signed-off-by: Tao Chen <chen.dylane@gmail.com>
---
 tools/lib/bpf/libbpf_probes.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c
index aeb4fd97d801..e142130cb83c 100644
--- a/tools/lib/bpf/libbpf_probes.c
+++ b/tools/lib/bpf/libbpf_probes.c
@@ -102,12 +102,15 @@ __u32 get_kernel_version(void)
 
 static int probe_prog_load(enum bpf_prog_type prog_type,
 			   const struct bpf_insn *insns, size_t insns_cnt,
+			   int *fd_array, size_t fd_array_cnt,
 			   char *log_buf, size_t log_buf_sz)
 {
 	LIBBPF_OPTS(bpf_prog_load_opts, opts,
 		.log_buf = log_buf,
 		.log_size = log_buf_sz,
 		.log_level = log_buf ? 1 : 0,
+		.fd_array = fd_array,
+		.fd_array_cnt = fd_array_cnt,
 	);
 	int fd, err, exp_err = 0;
 	const char *exp_msg = NULL;
@@ -214,7 +217,7 @@ int libbpf_probe_bpf_prog_type(enum bpf_prog_type prog_type, const void *opts)
 	if (opts)
 		return libbpf_err(-EINVAL);
 
-	ret = probe_prog_load(prog_type, insns, insn_cnt, NULL, 0);
+	ret = probe_prog_load(prog_type, insns, insn_cnt, NULL, 0, NULL, 0);
 	return libbpf_err(ret);
 }
 
@@ -448,7 +451,7 @@ int libbpf_probe_bpf_helper(enum bpf_prog_type prog_type, enum bpf_func_id helpe
 		return -EOPNOTSUPP;
 
 	buf[0] = '\0';
-	ret = probe_prog_load(prog_type, insns, insn_cnt, buf, sizeof(buf));
+	ret = probe_prog_load(prog_type, insns, insn_cnt, NULL, 0, buf, sizeof(buf));
 	if (ret < 0)
 		return libbpf_err(ret);
 
-- 
2.43.0

[PATCH bpf-next v5 3/4] libbpf: Add libbpf_probe_bpf_kfunc API

[Tao Chen]

Similarly to libbpf_probe_bpf_helper, the libbpf_probe_bpf_kfunc
used to test the availability of the different eBPF kfuncs on the
current system.

Cc: Tao Chen <dylane.chen@didiglobal.com>
Signed-off-by: Tao Chen <chen.dylane@gmail.com>
---
 tools/lib/bpf/libbpf.h        | 19 +++++++++++++-
 tools/lib/bpf/libbpf.map      |  1 +
 tools/lib/bpf/libbpf_probes.c | 48 +++++++++++++++++++++++++++++++++++
 3 files changed, 67 insertions(+), 1 deletion(-)

diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 3020ee45303a..e796e38cf255 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -1680,7 +1680,24 @@ LIBBPF_API int libbpf_probe_bpf_map_type(enum bpf_map_type map_type, const void
  */
 LIBBPF_API int libbpf_probe_bpf_helper(enum bpf_prog_type prog_type,
 				       enum bpf_func_id helper_id, const void *opts);
-
+/**
+ * @brief **libbpf_probe_bpf_kfunc()** detects if host kernel supports the
+ * use of a given BPF kfunc from specified BPF program type.
+ * @param prog_type BPF program type used to check the support of BPF kfunc
+ * @param kfunc_id The btf ID of BPF kfunc to check support for
+ * @param btf_fd The module BTF FD, if kfunc is defined in kernel module,
+ * btf_fd is used to point to module's BTF, which is >= 0, and -1 means kfunc
+ * defined in vmlinux.
+ * @param opts reserved for future extensibility, should be NULL
+ * @return 1, if given combination of program type and kfunc is supported; 0,
+ * if the combination is not supported; negative error code if feature
+ * detection for provided input arguments failed or can't be performed
+ *
+ * Make sure the process has required set of CAP_* permissions (or runs as
+ * root) when performing feature checking.
+ */
+LIBBPF_API int libbpf_probe_bpf_kfunc(enum bpf_prog_type prog_type,
+				      int kfunc_id, int btf_fd, const void *opts);
 /**
  * @brief **libbpf_num_possible_cpus()** is a helper function to get the
  * number of possible CPUs that the host kernel supports and expects.
diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
index b5a838de6f47..3bbfe13aeb6a 100644
--- a/tools/lib/bpf/libbpf.map
+++ b/tools/lib/bpf/libbpf.map
@@ -438,4 +438,5 @@ LIBBPF_1.6.0 {
 		bpf_linker__new_fd;
 		btf__add_decl_attr;
 		btf__add_type_attr;
+		libbpf_probe_bpf_kfunc;
 } LIBBPF_1.5.0;
diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c
index e142130cb83c..53f1196394bf 100644
--- a/tools/lib/bpf/libbpf_probes.c
+++ b/tools/lib/bpf/libbpf_probes.c
@@ -433,6 +433,54 @@ static bool can_probe_prog_type(enum bpf_prog_type prog_type)
 	return true;
 }
 
+int libbpf_probe_bpf_kfunc(enum bpf_prog_type prog_type, int kfunc_id, int btf_fd,
+			   const void *opts)
+{
+	struct bpf_insn insns[] = {
+		BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, BPF_PSEUDO_KFUNC_CALL, 1, kfunc_id),
+		BPF_EXIT_INSN(),
+	};
+	const size_t insn_cnt = ARRAY_SIZE(insns);
+	char buf[4096];
+	int fd_array[2] = {-1};
+	int ret;
+
+	if (opts)
+		return libbpf_err(-EINVAL);
+
+	if (!can_probe_prog_type(prog_type))
+		return -EOPNOTSUPP;
+
+	if (btf_fd >= 0) {
+		fd_array[1] = btf_fd;
+	} else if (btf_fd == -1) {
+		/* insn.off = 0, means vmlinux btf */
+		insns[0].off = 0;
+	} else {
+		return libbpf_err(-EINVAL);
+	}
+
+	buf[0] = '\0';
+	ret = probe_prog_load(prog_type, insns, insn_cnt, btf_fd >= 0 ? fd_array : NULL,
+			      0, buf, sizeof(buf));
+	if (ret < 0)
+		return libbpf_err(ret);
+
+	/* If BPF verifier recognizes BPF kfunc but it's not supported for
+	 * given BPF program type, it will emit "calling kernel function
+	 * bpf_cpumask_create is not allowed", if the kfunc id is invalid,
+	 * it will emit "kernel btf_id 4294967295 is not a function". If btf fd
+	 * invalid in module btf, it will emit "invalid module BTF fd specified" or
+	 * "negative offset disallowed for kernel module function call"
+	 */
+	if (ret == 0 && (strstr(buf, "not allowed") || strstr(buf, "not a function") ||
+			(strstr(buf, "invalid module BTF fd")) ||
+			(strstr(buf, "negative offset disallowed"))))
+		return 0;
+
+	return 1; /* assume supported */
+}
+
 int libbpf_probe_bpf_helper(enum bpf_prog_type prog_type, enum bpf_func_id helper_id,
 			    const void *opts)
 {
-- 
2.43.0

[PATCH bpf-next v5 4/4] selftests/bpf: Add libbpf_probe_bpf_kfunc API selftests

[Tao Chen]

Add selftests for prog_kfunc feature probing.

 ./test_progs -t libbpf_probe_kfuncs
 #153     libbpf_probe_kfuncs:OK
 Summary: 1/0 PASSED, 0 SKIPPED, 0 FAILED

Signed-off-by: Tao Chen <chen.dylane@gmail.com>
---
 .../selftests/bpf/prog_tests/libbpf_probes.c  | 118 ++++++++++++++++++
 1 file changed, 118 insertions(+)

diff --git a/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c b/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
index 4ed46ed58a7b..fc4c9dc2cbed 100644
--- a/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
+++ b/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
@@ -126,3 +126,121 @@ void test_libbpf_probe_helpers(void)
 		ASSERT_EQ(res, d->supported, buf);
 	}
 }
+
+static int module_btf_fd(char *module)
+{
+	int fd, err;
+	__u32 id = 0, len;
+	struct bpf_btf_info info;
+	char name[64];
+
+	while (true) {
+		err = bpf_btf_get_next_id(id, &id);
+		if (err && (errno == ENOENT || errno == EPERM))
+			return 0;
+		if (err) {
+			err = -errno;
+			return err;
+		}
+		fd = bpf_btf_get_fd_by_id(id);
+		if (fd < 0) {
+			if (errno == ENOENT)
+				continue;
+			err = -errno;
+			return err;
+		}
+		len = sizeof(info);
+		memset(&info, 0, sizeof(info));
+		info.name = ptr_to_u64(name);
+		info.name_len = sizeof(name);
+		err = bpf_btf_get_info_by_fd(fd, &info, &len);
+		if (err) {
+			err = -errno;
+			goto err_out;
+		}
+		/* find target module btf */
+		if (!strcmp(name, module))
+			break;
+
+		close(fd);
+	}
+
+	return fd;
+err_out:
+	close(fd);
+	return err;
+}
+
+void test_libbpf_probe_kfuncs(void)
+{
+	int ret, kfunc_id, fd;
+	char *kfunc = "bpf_cpumask_create";
+	struct btf *vmlinux_btf = NULL;
+	struct btf *module_btf = NULL;
+
+	vmlinux_btf = btf__parse("/sys/kernel/btf/vmlinux", NULL);
+	if (!ASSERT_OK_PTR(vmlinux_btf, "btf_parse"))
+		return;
+
+	kfunc_id = btf__find_by_name_kind(vmlinux_btf, kfunc, BTF_KIND_FUNC);
+	if (!ASSERT_GT(kfunc_id, 0, kfunc))
+		goto cleanup;
+
+	/* prog BPF_PROG_TYPE_SYSCALL supports kfunc bpf_cpumask_create */
+	ret = libbpf_probe_bpf_kfunc(BPF_PROG_TYPE_SYSCALL, kfunc_id, -1, NULL);
+	if (!ASSERT_EQ(ret, 1, "kfunc in vmlinux support"))
+		goto cleanup;
+
+	/* prog BPF_PROG_TYPE_KPROBE does not support kfunc bpf_cpumask_create */
+	ret = libbpf_probe_bpf_kfunc(BPF_PROG_TYPE_KPROBE, kfunc_id, -1, NULL);
+	if (!ASSERT_EQ(ret, 0, "kfunc in vmlinux not suuport"))
+		goto cleanup;
+
+	ret = libbpf_probe_bpf_kfunc(BPF_PROG_TYPE_KPROBE, -1, -1, NULL);
+	if (!ASSERT_EQ(ret, 0, "invalid kfunc id:-1"))
+		goto cleanup;
+
+	ret = libbpf_probe_bpf_kfunc(100000, kfunc_id, -1, NULL);
+	if (!ASSERT_ERR(ret, "invalid prog type:100000"))
+		goto cleanup;
+
+	if (!env.has_testmod)
+		goto cleanup;
+
+	module_btf = btf__load_module_btf("bpf_testmod", vmlinux_btf);
+	if (!ASSERT_OK_PTR(module_btf, "load module BTF"))
+		goto cleanup;
+
+	kfunc_id = btf__find_by_name(module_btf, "bpf_kfunc_call_test1");
+	if (!ASSERT_GT(kfunc_id, 0, "func not found"))
+		goto cleanup;
+
+	fd = module_btf_fd("bpf_testmod");
+	if (!ASSERT_GT(fd, 0, "module btf fd"))
+		goto cleanup;
+
+	/* prog BPF_PROG_TYPE_SYSCALL supports kfunc bpf_kfunc_call_test1 in bpf_testmod */
+	ret = libbpf_probe_bpf_kfunc(BPF_PROG_TYPE_SYSCALL, kfunc_id, fd, NULL);
+	if (!ASSERT_EQ(ret, 1, "kfunc in module btf support"))
+		goto cleanup_fd;
+
+	/* prog BPF_PROG_TYPE_KPROBE does not support kfunc bpf_kfunc_call_test1
+	 * in bpf_testmod
+	 */
+	ret = libbpf_probe_bpf_kfunc(BPF_PROG_TYPE_KPROBE, kfunc_id, fd, NULL);
+	if (!ASSERT_EQ(ret, 0, "kfunc in module btf not support"))
+		goto cleanup_fd;
+
+	ret = libbpf_probe_bpf_kfunc(BPF_PROG_TYPE_SYSCALL, -1, fd, NULL);
+	if (!ASSERT_EQ(ret, 0, "invalid kfunc id in module btf"))
+		goto cleanup_fd;
+
+	ret = libbpf_probe_bpf_kfunc(BPF_PROG_TYPE_SYSCALL, kfunc_id, 100, NULL);
+	ASSERT_EQ(ret, 0, "invalid btf fd in module btf");
+
+cleanup_fd:
+	close(fd);
+cleanup:
+	btf__free(vmlinux_btf);
+	btf__free(module_btf);
+}
-- 
2.43.0

Re: [PATCH bpf-next v5 4/4] selftests/bpf: Add libbpf_probe_bpf_kfunc API selftests

[Jiri Olsa]

On Mon, Feb 10, 2025 at 01:59:45PM +0800, Tao Chen wrote:
> Add selftests for prog_kfunc feature probing.
> 
>  ./test_progs -t libbpf_probe_kfuncs
>  #153     libbpf_probe_kfuncs:OK
>  Summary: 1/0 PASSED, 0 SKIPPED, 0 FAILED
> 
> Signed-off-by: Tao Chen <chen.dylane@gmail.com>
> ---
>  .../selftests/bpf/prog_tests/libbpf_probes.c  | 118 ++++++++++++++++++
>  1 file changed, 118 insertions(+)
> 
> diff --git a/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c b/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
> index 4ed46ed58a7b..fc4c9dc2cbed 100644
> --- a/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
> +++ b/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
> @@ -126,3 +126,121 @@ void test_libbpf_probe_helpers(void)
>  		ASSERT_EQ(res, d->supported, buf);
>  	}
>  }
> +
> +static int module_btf_fd(char *module)
> +{
> +	int fd, err;
> +	__u32 id = 0, len;
> +	struct bpf_btf_info info;
> +	char name[64];
> +
> +	while (true) {
> +		err = bpf_btf_get_next_id(id, &id);
> +		if (err && (errno == ENOENT || errno == EPERM))
> +			return 0;
> +		if (err) {
> +			err = -errno;
> +			return err;
> +		}
> +		fd = bpf_btf_get_fd_by_id(id);
> +		if (fd < 0) {
> +			if (errno == ENOENT)
> +				continue;
> +			err = -errno;
> +			return err;

nit, return -errno ?

jirka

> +		}
> +		len = sizeof(info);
> +		memset(&info, 0, sizeof(info));
> +		info.name = ptr_to_u64(name);
> +		info.name_len = sizeof(name);
> +		err = bpf_btf_get_info_by_fd(fd, &info, &len);
> +		if (err) {
> +			err = -errno;
> +			goto err_out;
> +		}
> +		/* find target module btf */
> +		if (!strcmp(name, module))
> +			break;
> +
> +		close(fd);
> +	}
> +
> +	return fd;
> +err_out:
> +	close(fd);
> +	return err;
> +}

SNIP

Re: [PATCH bpf-next v5 3/4] libbpf: Add libbpf_probe_bpf_kfunc API

[Jiri Olsa]

On Mon, Feb 10, 2025 at 01:59:44PM +0800, Tao Chen wrote:

SNIP

> diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c
> index e142130cb83c..53f1196394bf 100644
> --- a/tools/lib/bpf/libbpf_probes.c
> +++ b/tools/lib/bpf/libbpf_probes.c
> @@ -433,6 +433,54 @@ static bool can_probe_prog_type(enum bpf_prog_type prog_type)
>  	return true;
>  }
>  
> +int libbpf_probe_bpf_kfunc(enum bpf_prog_type prog_type, int kfunc_id, int btf_fd,
> +			   const void *opts)
> +{
> +	struct bpf_insn insns[] = {
> +		BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, BPF_PSEUDO_KFUNC_CALL, 1, kfunc_id),
> +		BPF_EXIT_INSN(),
> +	};
> +	const size_t insn_cnt = ARRAY_SIZE(insns);
> +	char buf[4096];
> +	int fd_array[2] = {-1};
> +	int ret;
> +
> +	if (opts)
> +		return libbpf_err(-EINVAL);
> +
> +	if (!can_probe_prog_type(prog_type))
> +		return -EOPNOTSUPP;
> +
> +	if (btf_fd >= 0) {
> +		fd_array[1] = btf_fd;
> +	} else if (btf_fd == -1) {
> +		/* insn.off = 0, means vmlinux btf */
> +		insns[0].off = 0;
> +	} else {
> +		return libbpf_err(-EINVAL);
> +	}
> +
> +	buf[0] = '\0';
> +	ret = probe_prog_load(prog_type, insns, insn_cnt, btf_fd >= 0 ? fd_array : NULL,
> +			      0, buf, sizeof(buf));

hum, you pass fd_array_cnt as 0, which IIUC will work properly

but I guess then we don't need to have fd_array_cnt argument in
probe_prog_load if all callers pass 0 ?

jirka

> +	if (ret < 0)
> +		return libbpf_err(ret);
> +
> +	/* If BPF verifier recognizes BPF kfunc but it's not supported for
> +	 * given BPF program type, it will emit "calling kernel function
> +	 * bpf_cpumask_create is not allowed", if the kfunc id is invalid,
> +	 * it will emit "kernel btf_id 4294967295 is not a function". If btf fd
> +	 * invalid in module btf, it will emit "invalid module BTF fd specified" or
> +	 * "negative offset disallowed for kernel module function call"
> +	 */
> +	if (ret == 0 && (strstr(buf, "not allowed") || strstr(buf, "not a function") ||
> +			(strstr(buf, "invalid module BTF fd")) ||
> +			(strstr(buf, "negative offset disallowed"))))
> +		return 0;
> +
> +	return 1; /* assume supported */
> +}
> +
>  int libbpf_probe_bpf_helper(enum bpf_prog_type prog_type, enum bpf_func_id helper_id,
>  			    const void *opts)
>  {
> -- 
> 2.43.0
> 

Re: [PATCH bpf-next v5 4/4] selftests/bpf: Add libbpf_probe_bpf_kfunc API selftests

[Tao Chen]

在 2025/2/10 21:47, Jiri Olsa 写道:
> On Mon, Feb 10, 2025 at 01:59:45PM +0800, Tao Chen wrote:
>> Add selftests for prog_kfunc feature probing.
>>
>>   ./test_progs -t libbpf_probe_kfuncs
>>   #153     libbpf_probe_kfuncs:OK
>>   Summary: 1/0 PASSED, 0 SKIPPED, 0 FAILED
>>
>> Signed-off-by: Tao Chen <chen.dylane@gmail.com>
>> ---
>>   .../selftests/bpf/prog_tests/libbpf_probes.c  | 118 ++++++++++++++++++
>>   1 file changed, 118 insertions(+)
>>
>> diff --git a/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c b/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
>> index 4ed46ed58a7b..fc4c9dc2cbed 100644
>> --- a/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
>> +++ b/tools/testing/selftests/bpf/prog_tests/libbpf_probes.c
>> @@ -126,3 +126,121 @@ void test_libbpf_probe_helpers(void)
>>   		ASSERT_EQ(res, d->supported, buf);
>>   	}
>>   }
>> +
>> +static int module_btf_fd(char *module)
>> +{
>> +	int fd, err;
>> +	__u32 id = 0, len;
>> +	struct bpf_btf_info info;
>> +	char name[64];
>> +
>> +	while (true) {
>> +		err = bpf_btf_get_next_id(id, &id);
>> +		if (err && (errno == ENOENT || errno == EPERM))
>> +			return 0;
>> +		if (err) {
>> +			err = -errno;
>> +			return err;
>> +		}
>> +		fd = bpf_btf_get_fd_by_id(id);
>> +		if (fd < 0) {
>> +			if (errno == ENOENT)
>> +				continue;
>> +			err = -errno;
>> +			return err;
> 
> nit, return -errno ?
> 
> jirka
> 

Ack.

>> +		}
>> +		len = sizeof(info);
>> +		memset(&info, 0, sizeof(info));
>> +		info.name = ptr_to_u64(name);
>> +		info.name_len = sizeof(name);
>> +		err = bpf_btf_get_info_by_fd(fd, &info, &len);
>> +		if (err) {
>> +			err = -errno;
>> +			goto err_out;
>> +		}
>> +		/* find target module btf */
>> +		if (!strcmp(name, module))
>> +			break;
>> +
>> +		close(fd);
>> +	}
>> +
>> +	return fd;
>> +err_out:
>> +	close(fd);
>> +	return err;
>> +}
> 
> SNIP


-- 
Best Regards
Dylane Chen

Re: [PATCH bpf-next v5 3/4] libbpf: Add libbpf_probe_bpf_kfunc API

[Tao Chen]

在 2025/2/10 21:47, Jiri Olsa 写道:
> On Mon, Feb 10, 2025 at 01:59:44PM +0800, Tao Chen wrote:
> 
> SNIP
> 
>> diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c
>> index e142130cb83c..53f1196394bf 100644
>> --- a/tools/lib/bpf/libbpf_probes.c
>> +++ b/tools/lib/bpf/libbpf_probes.c
>> @@ -433,6 +433,54 @@ static bool can_probe_prog_type(enum bpf_prog_type prog_type)
>>   	return true;
>>   }
>>   
>> +int libbpf_probe_bpf_kfunc(enum bpf_prog_type prog_type, int kfunc_id, int btf_fd,
>> +			   const void *opts)
>> +{
>> +	struct bpf_insn insns[] = {
>> +		BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, BPF_PSEUDO_KFUNC_CALL, 1, kfunc_id),
>> +		BPF_EXIT_INSN(),
>> +	};
>> +	const size_t insn_cnt = ARRAY_SIZE(insns);
>> +	char buf[4096];
>> +	int fd_array[2] = {-1};
>> +	int ret;
>> +
>> +	if (opts)
>> +		return libbpf_err(-EINVAL);
>> +
>> +	if (!can_probe_prog_type(prog_type))
>> +		return -EOPNOTSUPP;
>> +
>> +	if (btf_fd >= 0) {
>> +		fd_array[1] = btf_fd;
>> +	} else if (btf_fd == -1) {
>> +		/* insn.off = 0, means vmlinux btf */
>> +		insns[0].off = 0;
>> +	} else {
>> +		return libbpf_err(-EINVAL);
>> +	}
>> +
>> +	buf[0] = '\0';
>> +	ret = probe_prog_load(prog_type, insns, insn_cnt, btf_fd >= 0 ? fd_array : NULL,
>> +			      0, buf, sizeof(buf));
> 
> hum, you pass fd_array_cnt as 0, which IIUC will work properly
> 
> but I guess then we don't need to have fd_array_cnt argument in
> probe_prog_load if all callers pass 0 ?
> 
> jirka
> 

Hi, jiri
In fact, 0 is indeed used everywhere here. I was just thinking about 
whether we might need it in the future. Anyway, it seems better to 
remove it. I'll make the modifications in the next version.

>> +	if (ret < 0)
>> +		return libbpf_err(ret);
>> +
>> +	/* If BPF verifier recognizes BPF kfunc but it's not supported for
>> +	 * given BPF program type, it will emit "calling kernel function
>> +	 * bpf_cpumask_create is not allowed", if the kfunc id is invalid,
>> +	 * it will emit "kernel btf_id 4294967295 is not a function". If btf fd
>> +	 * invalid in module btf, it will emit "invalid module BTF fd specified" or
>> +	 * "negative offset disallowed for kernel module function call"
>> +	 */
>> +	if (ret == 0 && (strstr(buf, "not allowed") || strstr(buf, "not a function") ||
>> +			(strstr(buf, "invalid module BTF fd")) ||
>> +			(strstr(buf, "negative offset disallowed"))))
>> +		return 0;
>> +
>> +	return 1; /* assume supported */
>> +}
>> +
>>   int libbpf_probe_bpf_helper(enum bpf_prog_type prog_type, enum bpf_func_id helper_id,
>>   			    const void *opts)
>>   {
>> -- 
>> 2.43.0
>>


-- 
Best Regards
Dylane Chen