Revert "bpf: remove unnecessary rcu_read_{lock,unlock}() in multi-uprobe attach logic"
Subsystems (4)
PATCH 1/1 Revert "bpf: remove unnecessary rcu_read_{lock,unlock}() in multi-uprobe attach logic"
From: Di Shen <[email protected]>
This reverts commit 4a8f635a60540888dab3804992e86410360339c8.
Althought get_pid_task() internally already calls rcu_read_lock() and rcu_read_unlock(), the find_vpid() was not.
The documentation for find_vpid() clearly states:
"Must be called with the tasklist_lock or rcu_read_lock() held."
Add proper rcu_read_lock/unlock() to protect the find_vpid().
--- kernel/trace/bpf_trace.c | 2 ++ 1 file changed, 2 insertions(+)
kernel/trace/bpf_trace.c +2 -0
@@ -3417,7 +3417,9 @@ int bpf_uprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *pr }if (pid) { +rcu_read_lock (); task =get_pid_task (find_vpid (pid),PIDTYPE_TGID ); +rcu_read_unlock ();if (!task) { err = -ESRCH ; goto error_path_put;
Signature · 1 line
2.25.1
Quoted 25 lines · Xuewen Yan
From: Di Shen <[email protected]>
This reverts commit
4a8f635a60540888dab3804992e86410360339c8.Althought
get_pid_task()internally already callsrcu_read_lock()andrcu_read_unlock(), thefind_vpid()was not.The documentation for
find_vpid()clearly states:"Must be called with the
tasklist_lockorrcu_read_lock()held."Add proper rcu_read_lock/unlock() to protect the
find_vpid().Reported-by Xuewen Yan <[email protected]>Signed-off-by Di Shen <[email protected]>--- kernel/trace/bpf_trace.c | 2 ++ 1 file changed, 2 insertions(+)kernel/trace/bpf_trace.c +0
-0
kernel/trace/bpf_trace.c int bpf_uprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *pr · line 3425
Quoted 9 lines
kernel/trace/bpf_trace.c +2
-0@@ -3417,7 +3417,9 @@ int bpf_uprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *pr }if (pid) { +rcu_read_lock (); task =get_pid_task (find_vpid (pid),PIDTYPE_TGID ); +rcu_read_unlock ();if (!task) { err = -ESRCH ; goto error_path_put;
hmm. indeed.
Jiri ?
Andrii Nakryiko
Acked-byQuoted 29 lines · Alexei Starovoitov
Quoted 27 lines · Xuewen Yan
From: Di Shen <[email protected]>
This reverts commit
4a8f635a60540888dab3804992e86410360339c8.Althought
get_pid_task()internally already callsrcu_read_lock()andrcu_read_unlock(), thefind_vpid()was not.The documentation for
find_vpid()clearly states:"Must be called with the
tasklist_lockorrcu_read_lock()held."Add proper rcu_read_lock/unlock() to protect the
find_vpid().Reported-by Xuewen Yan <[email protected]>Signed-off-by Di Shen <[email protected]>--- kernel/trace/bpf_trace.c | 2 ++ 1 file changed, 2 insertions(+)kernel/trace/bpf_trace.c +0
-0@@ -3417,7 +3417,9 @@ int bpf_uprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *pr }
Quoted 10 lines
kernel/trace/bpf_trace.c int bpf_uprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *pr · line 3425
Quoted 7 lines
kernel/trace/bpf_trace.c +2
-0if (pid) { +rcu_read_lock (); task =get_pid_task (find_vpid (pid),PIDTYPE_TGID ); +rcu_read_unlock ();if (!task) { err = -ESRCH ; goto error_path_put;hmm. indeed.
yep, my bad, missed find_vpid() restrictions. revert LGTM
Quoted 1 line
Jiri ?
This patch was applied to bpf/bpf-next.git (master) by Alexei Starovoitov <[email protected]>:
Hello:
This patch was applied to bpf/bpf-next.git (master) by Alexei Starovoitov <[email protected]>:
Quoted 8 lines · On Tue, 20 May 2025 13:49:43 +0800 you wrote:
From: Di Shen <[email protected]>
This reverts commit
4a8f635a60540888dab3804992e86410360339c8.Althought
get_pid_task()internally already callsrcu_read_lock()andrcu_read_unlock(), thefind_vpid()was not.[...]
Here is the summary with links:
- Revert "bpf: remove unnecessary rcu_read_{lock,unlock}() in multi-uprobe attach logic"
https://git.kernel.org/bpf/bpf-next/c/4e2e6841ff76
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html