[PATCH bpf v3] bpf: fix NULL pointer dereference in bpf_task_from_vpid()

[PATCH bpf v3] bpf: fix NULL pointer dereference in bpf_task_from_vpid()

[Sechang Lim]

bpf_task_from_vpid() looks up a task in the pid namespace of the
current task, via find_task_by_vpid():

  find_task_by_vpid(vpid)
    find_task_by_pid_ns(vpid, task_active_pid_ns(current))
      find_pid_ns(nr, ns) -> idr_find(&ns->idr, nr)

cgroup_skb programs run in softirq, which may interrupt a task that is
itself in do_exit(). Once that task has passed
exit_notify() -> release_task() -> __unhash_process(), its thread_pid is
cleared, so task_active_pid_ns(current) returns NULL and find_pid_ns()
dereferences &NULL->idr:

  BUG: kernel NULL pointer dereference, address: 0000000000000050
  RIP: 0010:idr_find+0x11/0x30 lib/idr.c:176
  Call Trace:
   <IRQ>
   find_pid_ns kernel/pid.c:370 [inline]
   find_task_by_pid_ns+0x3b/0xe0 kernel/pid.c:485
   bpf_task_from_vpid+0x5b/0x200 kernel/bpf/helpers.c:2916
   bpf_prog_run_array_cg+0x17e/0x530 kernel/bpf/cgroup.c:81
   __cgroup_bpf_run_filter_skb+0x12b/0x250 kernel/bpf/cgroup.c:1612
   sk_filter_trim_cap+0x1dc/0x4c0 net/core/filter.c:148
   tcp_v4_rcv+0x18d1/0x2200 net/ipv4/tcp_ipv4.c:2223
   </IRQ>
   <TASK>
   do_exit+0xa63/0x1270 kernel/exit.c:1010
   get_signal+0x141c/0x1530 kernel/signal.c:3037

Return NULL when bpf_task_from_vpid() runs in interrupt
context, or when current has no pid namespace.

Acked-by: Yonghong Song <yonghong.song@linux.dev>
Fixes: 675c3596ff32 ("bpf: Add bpf_task_from_vpid() kfunc")
Signed-off-by: Sechang Lim <rhkrqnwk98@gmail.com>
---
v3:
 - Also handle current with no pid namespace 

v2:
 - Reject calls from interrupt context (Yonghong Song)
 - https://lore.kernel.org/bpf/20260605200501.1619406-1-rhkrqnwk98@gmail.com/

v1:
 - https://lore.kernel.org/bpf/20260603204206.773482-1-rhkrqnwk98@gmail.com/

 kernel/bpf/helpers.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
index b5314c9fed3c..226c31ccb5d6 100644
--- a/kernel/bpf/helpers.c
+++ b/kernel/bpf/helpers.c
@@ -2912,7 +2912,14 @@ __bpf_kfunc struct task_struct *bpf_task_from_vpid(s32 vpid)
 {
 	struct task_struct *p;
 
+	if (in_interrupt())
+		return NULL;
+
 	rcu_read_lock();
+	if (!task_active_pid_ns(current)) {
+		rcu_read_unlock();
+		return NULL;
+	}
 	p = find_task_by_vpid(vpid);
 	if (p)
 		p = bpf_task_acquire(p);
-- 
2.43.0

Re: [PATCH bpf v3] bpf: fix NULL pointer dereference in bpf_task_from_vpid()

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The use of the deprecated `in_interrupt()` macro causes `bpf_task_from_vpid()` to incorrectly fail and return `NULL` in valid task contexts when Bottom Halves (BH) are disabled.
--

commit 17040face7147d85a8cf3090be3539991c26df30
Author: Sechang Lim <rhkrqnwk98@gmail.com>

bpf: fix NULL pointer dereference in bpf_task_from_vpid()

This commit addresses a NULL pointer dereference in bpf_task_from_vpid()
by adding checks to return NULL when running in interrupt context or when
the current task has no active pid namespace.

> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> index b5314c9fed3cf..226c31ccb5d61 100644
> --- a/kernel/bpf/helpers.c
> +++ b/kernel/bpf/helpers.c
> @@ -2912,7 +2912,14 @@ __bpf_kfunc struct task_struct *bpf_task_from_vpid(s32 vpid)
>  {
>  	struct task_struct *p;
>  
> +	if (in_interrupt())
> +		return NULL;

[Severity: Medium]
Does the use of in_interrupt() here incorrectly reject valid task contexts
where bottom halves (BH) are disabled?

Since in_interrupt() evaluates to irq_count(), it evaluates to true not only
in hardware and softirq contexts, but also in normal process contexts if
local_bh_disable() or spin_lock_bh() is active. This might cause valid BPF
programs running in BH-disabled task contexts to unexpectedly receive NULL
from bpf_task_from_vpid().

Would it be better to use !in_task() here to accurately filter out non-task
asynchronous contexts without penalizing BH-disabled task contexts?

> +
>  	rcu_read_lock();
> +	if (!task_active_pid_ns(current)) {
> +		rcu_read_unlock();
> +		return NULL;
> +	}
>  	p = find_task_by_vpid(vpid);

[ ... ]

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260606091941.1803115-1-rhkrqnwk98@gmail.com?part=1

Re: [PATCH bpf v3] bpf: fix NULL pointer dereference in bpf_task_from_vpid()

[bot+bpf-ci]

[-- Attachment #1: Type: text/plain, Size: 1470 bytes --]

> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> index b5314c9fed3c..226c31ccb5d6 100644
> --- a/kernel/bpf/helpers.c
> +++ b/kernel/bpf/helpers.c
> @@ -2912,7 +2912,14 @@ __bpf_kfunc struct task_struct *bpf_task_from_vpid(s32 vpid)
>  {
>  	struct task_struct *p;
>
> +	if (in_interrupt())
> +		return NULL;

This isn't a bug, but a review comment on v3 about this in_interrupt()
check does not appear to have been answered.

The comment noted that in_interrupt() evaluates to irq_count(), so it
returns true not only in hardware and softirq context but also in normal
task context whenever bottom halves are disabled, such as inside
local_bh_disable() or spin_lock_bh().

Would this cause bpf_task_from_vpid() to return NULL for a valid BPF
program running in a BH-disabled task context, where current still has a
valid pid namespace?

Would using !in_task() instead filter out the asynchronous contexts
without rejecting BH-disabled task contexts?

  https://lore.kernel.org/bpf/20260606091941.1803115-1-rhkrqnwk98@gmail.com/

> +
>  	rcu_read_lock();
> +	if (!task_active_pid_ns(current)) {
> +		rcu_read_unlock();
> +		return NULL;
> +	}
>  	p = find_task_by_vpid(vpid);
>  	if (p)
>  		p = bpf_task_acquire(p);


---
AI reviewed your patch. Please fix the bug or email reply why it's not a bug.
See: https://github.com/kernel-patches/vmtest/blob/master/ci/claude/README.md

CI run summary: https://github.com/kernel-patches/bpf/actions/runs/27058795186

Re: [PATCH bpf v3] bpf: fix NULL pointer dereference in bpf_task_from_vpid()

[Leon Hwang]

On 2026/6/6 17:19, Sechang Lim wrote:
[...]
> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> index b5314c9fed3c..226c31ccb5d6 100644
> --- a/kernel/bpf/helpers.c
> +++ b/kernel/bpf/helpers.c
> @@ -2912,7 +2912,14 @@ __bpf_kfunc struct task_struct *bpf_task_from_vpid(s32 vpid)
>  {
>  	struct task_struct *p;
>  
> +	if (in_interrupt())
> +		return NULL;
> +
>  	rcu_read_lock();

Better to use guard(rcu)() here, and drop the rcu_read_unlock().

Thanks,
Leon

> +	if (!task_active_pid_ns(current)) {
> +		rcu_read_unlock();
> +		return NULL;
> +	}
>  	p = find_task_by_vpid(vpid);
>  	if (p)
>  		p = bpf_task_acquire(p);