Skip to contentSkip to conversation

bpf: Remove migrate_disable in kprobe_multi_link_prog_run

PATCH Superseded Tao Chen 2 participants 4 messages Aug 5, 2025 – Aug 14, 2025

Subsystems (4)

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Tao Chen

author
#0 lore raw

PATCH v2 1/1 bpf: Remove migrate_disable in kprobe_multi_link_prog_run

bpf program should run under migration disabled, kprobe_multi_link_prog_run called all the way from graph tracer, which disables preemption in function_graph_enter_regs, as Jiri and Yonghong suggested, there is no need to use migrate_disable. As a result, some overhead maybe will be reduced.

Fixes 0dcac2725406 ("bpf: Add multi kprobe link")
---
 kernel/trace/bpf_trace.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

Change list:
v1 -> v2:
- s/called the way/called all the way/.(Jiri)
v1: https://lore.kernel.org/bpf/[email protected]

kernel/trace/bpf_trace.c +7 -2
index 3ae52978cae..5701791e3cb 100644
@@ -2734,14 +2734,19 @@ kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link,
 		goto out;
 	}
 
-	migrate_disable();
+	/*
+	 * bpf program should run under migration disabled, kprobe_multi_link_prog_run
+	 * called all the way from graph tracer, which disables preemption in
+	 * function_graph_enter_regs, so there is no need to use migrate_disable.
+	 * Accessing the above percpu data bpf_prog_active is also safe for the same
+	 * reason.
+	 */
 	rcu_read_lock();
 	regs = ftrace_partial_regs(fregs, bpf_kprobe_multi_pt_regs_ptr());
 	old_run_ctx = bpf_set_run_ctx(&run_ctx.session_ctx.run_ctx);
 	err = bpf_prog_run(link->link.prog, regs);
 	bpf_reset_run_ctx(old_run_ctx);
 	rcu_read_unlock();
-	migrate_enable();
 
  out:
 	__this_cpu_dec(bpf_prog_active);
Signature · 1 line
2.48.1
#1 lore raw

kernel/trace/bpf_trace.c kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link,

Quoted 25 lines · Tao Chen

bpf program should run under migration disabled, kprobe_multi_link_prog_run called all the way from graph tracer, which disables preemption in function_graph_enter_regs, as Jiri and Yonghong suggested, there is no need to use migrate_disable. As a result, some overhead maybe will be reduced.

Fixes 0dcac2725406 ("bpf: Add multi kprobe link")
---
 kernel/trace/bpf_trace.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

Change list:
v1 -> v2:
- s/called the way/called all the way/.(Jiri)
v1: https://lore.kernel.org/bpf/[email protected]

kernel/trace/bpf_trace.c +0 -0
index 3ae52978cae..5701791e3cb 100644
@@ -2734,14 +2734,19 @@ kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link,

even though bpf_prog_run() eventually calls cant_migrate(), we should add it before that __this_cpu_inc_return() call as well, because that one is relying on that non-migration independently from bpf_prog_run()

kernel/trace/bpf_trace.c kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link, · line 2743

Quoted 11 lines
kernel/trace/bpf_trace.c +7 -1
                goto out;
        }
 
-       migrate_disable();
+       /*
+        * bpf program should run under migration disabled, kprobe_multi_link_prog_run
+        * called all the way from graph tracer, which disables preemption in
+        * function_graph_enter_regs, so there is no need to use migrate_disable.
+        * Accessing the above percpu data bpf_prog_active is also safe for the same
+        * reason.
+        */

let's shorten this a bit to something like:

/* graph tracer framework ensures we won't migrate */ cant_migrate();

all the other stuff in the comment can become outdated way too easily and/or is sort of general BPF implementation knowledge

pw-bot: cr

Quoted 13 lines
kernel/trace/bpf_trace.c +0 -2
        rcu_read_lock();
        regs = ftrace_partial_regs(fregs, bpf_kprobe_multi_pt_regs_ptr());
        old_run_ctx = bpf_set_run_ctx(&run_ctx.session_ctx.run_ctx);
        err = bpf_prog_run(link->link.prog, regs);
        bpf_reset_run_ctx(old_run_ctx);
        rcu_read_unlock();
-       migrate_enable();
 
  out:
        __this_cpu_dec(bpf_prog_active);
--

2.48.1

author
#2 lore raw
Quoted 21 lines · Andrii Nakryiko
Quoted 20 lines · Tao Chen

bpf program should run under migration disabled, kprobe_multi_link_prog_run called all the way from graph tracer, which disables preemption in function_graph_enter_regs, as Jiri and Yonghong suggested, there is no need to use migrate_disable. As a result, some overhead maybe will be reduced.

Fixes 0dcac2725406 ("bpf: Add multi kprobe link")
---
  kernel/trace/bpf_trace.c | 9 +++++++--
  1 file changed, 7 insertions(+), 2 deletions(-)

Change list:
v1 -> v2:
- s/called the way/called all the way/.(Jiri)
v1: https://lore.kernel.org/bpf/[email protected]

Quoted 10 lines

kernel/trace/bpf_trace.c kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link,

Quoted 5 lines
kernel/trace/bpf_trace.c +0 -0
index 3ae52978cae..5701791e3cb 100644
@@ -2734,14 +2734,19 @@ kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link,

even though bpf_prog_run() eventually calls cant_migrate(), we should add it before that __this_cpu_inc_return() call as well, because that one is relying on that non-migration independently from bpf_prog_run()

Hi Andrii,

There is __this_cpu_preempt_check in __this_cpu_inc_return, and the judgment criteria are similar to cant_migrate, and I'm not sure if it is enough.

Quoted 12 lines

kernel/trace/bpf_trace.c kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link, · line 2743

Quoted 11 lines
kernel/trace/bpf_trace.c +7 -1
                 goto out;
         }
 
-       migrate_disable();
+       /*
+        * bpf program should run under migration disabled, kprobe_multi_link_prog_run
+        * called all the way from graph tracer, which disables preemption in
+        * function_graph_enter_regs, so there is no need to use migrate_disable.
+        * Accessing the above percpu data bpf_prog_active is also safe for the same
+        * reason.
+        */
Quoted 3 lines

let's shorten this a bit to something like:

/* graph tracer framework ensures we won't migrate */

will change it in v3.

Quoted 21 lines

cant_migrate();

all the other stuff in the comment can become outdated way too easily and/or is sort of general BPF implementation knowledge

pw-bot: cr

Quoted 13 lines
kernel/trace/bpf_trace.c +0 -2
         rcu_read_lock();
         regs = ftrace_partial_regs(fregs, bpf_kprobe_multi_pt_regs_ptr());
         old_run_ctx = bpf_set_run_ctx(&run_ctx.session_ctx.run_ctx);
         err = bpf_prog_run(link->link.prog, regs);
         bpf_reset_run_ctx(old_run_ctx);
         rcu_read_unlock();
-       migrate_enable();
 
   out:
         __this_cpu_dec(bpf_prog_active);
--

2.48.1

Signature · 2 lines
Best Regards
Tao Chen
author
#3 lore raw
In reply to Andrii Nakryiko
Quoted 21 lines · Andrii Nakryiko
Quoted 20 lines · Tao Chen

bpf program should run under migration disabled, kprobe_multi_link_prog_run called all the way from graph tracer, which disables preemption in function_graph_enter_regs, as Jiri and Yonghong suggested, there is no need to use migrate_disable. As a result, some overhead maybe will be reduced.

Fixes 0dcac2725406 ("bpf: Add multi kprobe link")
---
  kernel/trace/bpf_trace.c | 9 +++++++--
  1 file changed, 7 insertions(+), 2 deletions(-)

Change list:
v1 -> v2:
- s/called the way/called all the way/.(Jiri)
v1: https://lore.kernel.org/bpf/[email protected]

Quoted 10 lines

kernel/trace/bpf_trace.c kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link,

Quoted 5 lines
kernel/trace/bpf_trace.c +0 -0
index 3ae52978cae..5701791e3cb 100644
@@ -2734,14 +2734,19 @@ kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link,

even though bpf_prog_run() eventually calls cant_migrate(), we should add it before that __this_cpu_inc_return() call as well, because that one is relying on that non-migration independently from bpf_prog_run()

maybe cant_sleep() is better like trace_call_bpf, cant_sleep dose not check migration_disabled again, which is done in __this_cpu_preempt_check. I will add it in v3.

Quoted 36 lines

kernel/trace/bpf_trace.c kprobe_multi_link_prog_run(struct bpf_kprobe_multi_link *link, · line 2743

Quoted 11 lines
kernel/trace/bpf_trace.c +7 -1
                 goto out;
         }
 
-       migrate_disable();
+       /*
+        * bpf program should run under migration disabled, kprobe_multi_link_prog_run
+        * called all the way from graph tracer, which disables preemption in
+        * function_graph_enter_regs, so there is no need to use migrate_disable.
+        * Accessing the above percpu data bpf_prog_active is also safe for the same
+        * reason.
+        */

let's shorten this a bit to something like:

/* graph tracer framework ensures we won't migrate */ cant_migrate();

all the other stuff in the comment can become outdated way too easily and/or is sort of general BPF implementation knowledge

pw-bot: cr

Quoted 13 lines
kernel/trace/bpf_trace.c +0 -2
         rcu_read_lock();
         regs = ftrace_partial_regs(fregs, bpf_kprobe_multi_pt_regs_ptr());
         old_run_ctx = bpf_set_run_ctx(&run_ctx.session_ctx.run_ctx);
         err = bpf_prog_run(link->link.prog, regs);
         bpf_reset_run_ctx(old_run_ctx);
         rcu_read_unlock();
-       migrate_enable();
 
   out:
         __this_cpu_dec(bpf_prog_active);
--

2.48.1

Signature · 2 lines
Best Regards
Tao Chen