If fsession exists, we will use the bit (1 << BPF_TRAMP_M_IS_RETURN) in
ctx[-1] to store the "is_return" flag.
The logic of bpf_session_is_return() for fsession is implemented in the
verifier by inline following code:
bool bpf_session_is_return(void *ctx)
{
return !!(((u64 *)ctx)[-1] & (1 << BPF_TRAMP_M_IS_RETURN));
}
Signed-off-by: Menglong Dong <dongml2@chinatelecom.cn>
Co-developed-by: Leon Hwang <leon.hwang@linux.dev>
Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
---
v9:
- remove the definition of bpf_fsession_is_return()
v7:
- reuse the kfunc bpf_session_is_return() instead of introduce new kfunc
v4:
- split out the bpf_fsession_cookie() to another patch
v3:
- merge the bpf_tracing_is_exit and bpf_fsession_cookie into a single
patch
v2:
- store the session flags after return value, instead of before nr_args
- inline the bpf_tracing_is_exit, as Jiri suggested
---
include/linux/bpf.h | 3 +++
kernel/bpf/verifier.c | 15 ++++++++++++++-
kernel/trace/bpf_trace.c | 28 +++++++++++++++++-----------
3 files changed, 34 insertions(+), 12 deletions(-)
diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index 41228b0add52..2640ec2157e1 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -1229,6 +1229,9 @@ enum {
#endif
};
+#define BPF_TRAMP_M_NR_ARGS 0
+#define BPF_TRAMP_M_IS_RETURN 8
+
struct bpf_tramp_links {
struct bpf_tramp_link *links[BPF_MAX_TRAMP_LINKS];
int nr_links;
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index bfff3f84fd91..1b0292a03186 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -12374,6 +12374,7 @@ enum special_kfunc_type {
KF_bpf_arena_alloc_pages,
KF_bpf_arena_free_pages,
KF_bpf_arena_reserve_pages,
+ KF_bpf_session_is_return,
};
BTF_ID_LIST(special_kfunc_list)
@@ -12451,6 +12452,7 @@ BTF_ID(func, bpf_task_work_schedule_resume_impl)
BTF_ID(func, bpf_arena_alloc_pages)
BTF_ID(func, bpf_arena_free_pages)
BTF_ID(func, bpf_arena_reserve_pages)
+BTF_ID(func, bpf_session_is_return)
static bool is_task_work_add_kfunc(u32 func_id)
{
@@ -12505,7 +12507,8 @@ get_kfunc_ptr_arg_type(struct bpf_verifier_env *env,
struct bpf_reg_state *reg = ®s[regno];
bool arg_mem_size = false;
- if (meta->func_id == special_kfunc_list[KF_bpf_cast_to_kern_ctx])
+ if (meta->func_id == special_kfunc_list[KF_bpf_cast_to_kern_ctx] ||
+ meta->func_id == special_kfunc_list[KF_bpf_session_is_return])
return KF_ARG_PTR_TO_CTX;
if (argno + 1 < nargs &&
@@ -22558,6 +22561,16 @@ static int fixup_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn,
desc->func_id == special_kfunc_list[KF_bpf_rdonly_cast]) {
insn_buf[0] = BPF_MOV64_REG(BPF_REG_0, BPF_REG_1);
*cnt = 1;
+ } else if (desc->func_id == special_kfunc_list[KF_bpf_session_is_return] &&
+ env->prog->expected_attach_type == BPF_TRACE_FSESSION) {
+ /* implement and inline the bpf_session_is_return() for
+ * fsession, and the logic is:
+ * return !!(((u64 *)ctx)[-1] & (1 << BPF_TRAMP_M_IS_RETURN))
+ */
+ insn_buf[0] = BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_1, -8);
+ insn_buf[1] = BPF_ALU64_IMM(BPF_RSH, BPF_REG_0, BPF_TRAMP_M_IS_RETURN);
+ insn_buf[2] = BPF_ALU64_IMM(BPF_AND, BPF_REG_0, 1);
+ *cnt = 3;
}
if (env->insn_aux_data[insn_idx].arg_prog) {
diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
index 297dcafb2c55..1fe508d451b7 100644
--- a/kernel/trace/bpf_trace.c
+++ b/kernel/trace/bpf_trace.c
@@ -3334,34 +3334,40 @@ __bpf_kfunc __u64 *bpf_session_cookie(void *ctx)
__bpf_kfunc_end_defs();
-BTF_KFUNCS_START(kprobe_multi_kfunc_set_ids)
+BTF_KFUNCS_START(session_kfunc_set_ids)
BTF_ID_FLAGS(func, bpf_session_is_return)
BTF_ID_FLAGS(func, bpf_session_cookie)
-BTF_KFUNCS_END(kprobe_multi_kfunc_set_ids)
+BTF_KFUNCS_END(session_kfunc_set_ids)
-static int bpf_kprobe_multi_filter(const struct bpf_prog *prog, u32 kfunc_id)
+static int bpf_session_filter(const struct bpf_prog *prog, u32 kfunc_id)
{
- if (!btf_id_set8_contains(&kprobe_multi_kfunc_set_ids, kfunc_id))
+ if (!btf_id_set8_contains(&session_kfunc_set_ids, kfunc_id))
return 0;
- if (!is_kprobe_session(prog) && !is_uprobe_session(prog))
+ if (!is_kprobe_session(prog) && !is_uprobe_session(prog) &&
+ prog->expected_attach_type != BPF_TRACE_FSESSION)
return -EACCES;
return 0;
}
-static const struct btf_kfunc_id_set bpf_kprobe_multi_kfunc_set = {
+static const struct btf_kfunc_id_set bpf_session_kfunc_set = {
.owner = THIS_MODULE,
- .set = &kprobe_multi_kfunc_set_ids,
- .filter = bpf_kprobe_multi_filter,
+ .set = &session_kfunc_set_ids,
+ .filter = bpf_session_filter,
};
-static int __init bpf_kprobe_multi_kfuncs_init(void)
+static int __init bpf_trace_kfuncs_init(void)
{
- return register_btf_kfunc_id_set(BPF_PROG_TYPE_KPROBE, &bpf_kprobe_multi_kfunc_set);
+ int err = 0;
+
+ err = err ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_KPROBE, &bpf_session_kfunc_set);
+ err = err ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_TRACING, &bpf_session_kfunc_set);
+
+ return err;
}
-late_initcall(bpf_kprobe_multi_kfuncs_init);
+late_initcall(bpf_trace_kfuncs_init);
typedef int (*copy_fn_t)(void *dst, const void *src, u32 size, struct task_struct *tsk);
--
2.52.0On Sat, Jan 10, 2026 at 6:12 AM Menglong Dong <menglong8.dong@gmail.com> wrote:
>
> If fsession exists, we will use the bit (1 << BPF_TRAMP_M_IS_RETURN) in
> ctx[-1] to store the "is_return" flag.
>
> The logic of bpf_session_is_return() for fsession is implemented in the
> verifier by inline following code:
>
> bool bpf_session_is_return(void *ctx)
> {
> return !!(((u64 *)ctx)[-1] & (1 << BPF_TRAMP_M_IS_RETURN));
this look unnecessarily scary :) !! part is unnecessary because
non-zero integer will be converted to proper true(1)/false(0) by
compiler. But I'd just rewrite it in arguably slightly simpler form
that lays itself to assembly more directly:
return ((u64 *)ctx[-1] >> BPF_TRAMP_M_IS_RETURN) & 1;
> }
>
> Signed-off-by: Menglong Dong <dongml2@chinatelecom.cn>
> Co-developed-by: Leon Hwang <leon.hwang@linux.dev>
> Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
> ---
> v9:
> - remove the definition of bpf_fsession_is_return()
>
> v7:
> - reuse the kfunc bpf_session_is_return() instead of introduce new kfunc
>
> v4:
> - split out the bpf_fsession_cookie() to another patch
>
> v3:
> - merge the bpf_tracing_is_exit and bpf_fsession_cookie into a single
> patch
>
> v2:
> - store the session flags after return value, instead of before nr_args
> - inline the bpf_tracing_is_exit, as Jiri suggested
> ---
> include/linux/bpf.h | 3 +++
> kernel/bpf/verifier.c | 15 ++++++++++++++-
> kernel/trace/bpf_trace.c | 28 +++++++++++++++++-----------
> 3 files changed, 34 insertions(+), 12 deletions(-)
>
> diff --git a/include/linux/bpf.h b/include/linux/bpf.h
> index 41228b0add52..2640ec2157e1 100644
> --- a/include/linux/bpf.h
> +++ b/include/linux/bpf.h
> @@ -1229,6 +1229,9 @@ enum {
> #endif
> };
>
> +#define BPF_TRAMP_M_NR_ARGS 0
> +#define BPF_TRAMP_M_IS_RETURN 8
nit: What does "M" stand for? Macro? Mask? Menglong? ;) Some new
convention, why?
> +
> struct bpf_tramp_links {
> struct bpf_tramp_link *links[BPF_MAX_TRAMP_LINKS];
> int nr_links;
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index bfff3f84fd91..1b0292a03186 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -12374,6 +12374,7 @@ enum special_kfunc_type {
> KF_bpf_arena_alloc_pages,
> KF_bpf_arena_free_pages,
> KF_bpf_arena_reserve_pages,
> + KF_bpf_session_is_return,
> };
>
> BTF_ID_LIST(special_kfunc_list)
> @@ -12451,6 +12452,7 @@ BTF_ID(func, bpf_task_work_schedule_resume_impl)
> BTF_ID(func, bpf_arena_alloc_pages)
> BTF_ID(func, bpf_arena_free_pages)
> BTF_ID(func, bpf_arena_reserve_pages)
> +BTF_ID(func, bpf_session_is_return)
>
> static bool is_task_work_add_kfunc(u32 func_id)
> {
> @@ -12505,7 +12507,8 @@ get_kfunc_ptr_arg_type(struct bpf_verifier_env *env,
> struct bpf_reg_state *reg = ®s[regno];
> bool arg_mem_size = false;
>
> - if (meta->func_id == special_kfunc_list[KF_bpf_cast_to_kern_ctx])
> + if (meta->func_id == special_kfunc_list[KF_bpf_cast_to_kern_ctx] ||
> + meta->func_id == special_kfunc_list[KF_bpf_session_is_return])
> return KF_ARG_PTR_TO_CTX;
>
> if (argno + 1 < nargs &&
> @@ -22558,6 +22561,16 @@ static int fixup_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn,
> desc->func_id == special_kfunc_list[KF_bpf_rdonly_cast]) {
> insn_buf[0] = BPF_MOV64_REG(BPF_REG_0, BPF_REG_1);
> *cnt = 1;
> + } else if (desc->func_id == special_kfunc_list[KF_bpf_session_is_return] &&
> + env->prog->expected_attach_type == BPF_TRACE_FSESSION) {
> + /* implement and inline the bpf_session_is_return() for
nit: comment style
> + * fsession, and the logic is:
> + * return !!(((u64 *)ctx)[-1] & (1 << BPF_TRAMP_M_IS_RETURN))
> + */
> + insn_buf[0] = BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_1, -8);
> + insn_buf[1] = BPF_ALU64_IMM(BPF_RSH, BPF_REG_0, BPF_TRAMP_M_IS_RETURN);
> + insn_buf[2] = BPF_ALU64_IMM(BPF_AND, BPF_REG_0, 1);
lol, your assembly is simpler than that C expression above, let's keep
C close to what you actually are doing in assembler
> + *cnt = 3;
> }
>
> if (env->insn_aux_data[insn_idx].arg_prog) {
> diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
> index 297dcafb2c55..1fe508d451b7 100644
> --- a/kernel/trace/bpf_trace.c
> +++ b/kernel/trace/bpf_trace.c
> @@ -3334,34 +3334,40 @@ __bpf_kfunc __u64 *bpf_session_cookie(void *ctx)
>
> __bpf_kfunc_end_defs();
>
> -BTF_KFUNCS_START(kprobe_multi_kfunc_set_ids)
> +BTF_KFUNCS_START(session_kfunc_set_ids)
> BTF_ID_FLAGS(func, bpf_session_is_return)
> BTF_ID_FLAGS(func, bpf_session_cookie)
> -BTF_KFUNCS_END(kprobe_multi_kfunc_set_ids)
> +BTF_KFUNCS_END(session_kfunc_set_ids)
>
> -static int bpf_kprobe_multi_filter(const struct bpf_prog *prog, u32 kfunc_id)
> +static int bpf_session_filter(const struct bpf_prog *prog, u32 kfunc_id)
> {
> - if (!btf_id_set8_contains(&kprobe_multi_kfunc_set_ids, kfunc_id))
> + if (!btf_id_set8_contains(&session_kfunc_set_ids, kfunc_id))
> return 0;
>
> - if (!is_kprobe_session(prog) && !is_uprobe_session(prog))
> + if (!is_kprobe_session(prog) && !is_uprobe_session(prog) &&
> + prog->expected_attach_type != BPF_TRACE_FSESSION)
check both expected_attach_type *and* prog_type, please (and I think
it would be good to check prog type for kprobe_session and
uprobe_session as well, because now it's not guaranteed that program
will be of BPF_PROG_TYPE_KPROBE
> return -EACCES;
>
> return 0;
> }
>
> -static const struct btf_kfunc_id_set bpf_kprobe_multi_kfunc_set = {
> +static const struct btf_kfunc_id_set bpf_session_kfunc_set = {
> .owner = THIS_MODULE,
> - .set = &kprobe_multi_kfunc_set_ids,
> - .filter = bpf_kprobe_multi_filter,
> + .set = &session_kfunc_set_ids,
> + .filter = bpf_session_filter,
> };
>
> -static int __init bpf_kprobe_multi_kfuncs_init(void)
> +static int __init bpf_trace_kfuncs_init(void)
> {
> - return register_btf_kfunc_id_set(BPF_PROG_TYPE_KPROBE, &bpf_kprobe_multi_kfunc_set);
> + int err = 0;
> +
> + err = err ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_KPROBE, &bpf_session_kfunc_set);
> + err = err ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_TRACING, &bpf_session_kfunc_set);
> +
> + return err;
> }
>
> -late_initcall(bpf_kprobe_multi_kfuncs_init);
> +late_initcall(bpf_trace_kfuncs_init);
>
> typedef int (*copy_fn_t)(void *dst, const void *src, u32 size, struct task_struct *tsk);
>
> --
> 2.52.0
>
© 2016 - 2026 Red Hat, Inc.