bpf-next: Prevent out of bound buffer write in __bpf_get_stack

[PATCH] bpf-next: Prevent out of bound buffer write in __bpf_get_stack

Posted by Arnaud Lecomte 1 day, 23 hours ago

Syzkaller reported a KASAN slab-out-of-bounds write in __bpf_get_stack()
during stack trace copying.

The issue occurs when: the callchain entry (stored as a per-cpu variable)
grow between collection and buffer copy, causing it to exceed the initially
calculated buffer size based on max_depth.

The callchain collection intentionally avoids locking for performance
reasons, but this creates a window where concurrent modifications can
occur during the copy operation.

To prevent this from happening, we clamp the trace len to the max
depth initially calculated with the buffer size and the size of
a trace.

Reported-by: syzbot+d1b7fa1092def3628bd7@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/691231dc.a70a0220.22f260.0101.GAE@google.com/T/
Fixes: e17d62fedd10 ("bpf: Refactor stack map trace depth calculation into helper function")
Tested-by: syzbot+d1b7fa1092def3628bd7@syzkaller.appspotmail.com
Cc: Brahmajit Das <listout@listout.xyz>
Signed-off-by: Arnaud Lecomte <contact@arnaud-lcm.com>
---
Thanks Brahmajit Das for the initial fix he proposed that I tweaked
with the correct justification and a better implementation in my
opinion.
---
 kernel/bpf/stackmap.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c
index da3d328f5c15..e56752a9a891 100644
--- a/kernel/bpf/stackmap.c
+++ b/kernel/bpf/stackmap.c
@@ -465,7 +465,6 @@ static long __bpf_get_stack(struct pt_regs *regs, struct task_struct *task,
 
 	if (trace_in) {
 		trace = trace_in;
-		trace->nr = min_t(u32, trace->nr, max_depth);
 	} else if (kernel && task) {
 		trace = get_callchain_entry_for_task(task, max_depth);
 	} else {
@@ -479,7 +478,8 @@ static long __bpf_get_stack(struct pt_regs *regs, struct task_struct *task,
 		goto err_fault;
 	}
 
-	trace_nr = trace->nr - skip;
+	trace_nr = min(trace->nr, max_depth);
+	trace_nr = trace_nr - skip;
 	copy_len = trace_nr * elem_size;
 
 	ips = trace->ip + skip;
-- 
2.43.0

Re: [PATCH] bpf-next: Prevent out of bound buffer write in __bpf_get_stack

Posted by Andrii Nakryiko 19 hours ago

On Sun, Jan 4, 2026 at 12:52 PM Arnaud Lecomte <contact@arnaud-lcm.com> wrote:
>
> Syzkaller reported a KASAN slab-out-of-bounds write in __bpf_get_stack()
> during stack trace copying.
>
> The issue occurs when: the callchain entry (stored as a per-cpu variable)
> grow between collection and buffer copy, causing it to exceed the initially
> calculated buffer size based on max_depth.
>
> The callchain collection intentionally avoids locking for performance
> reasons, but this creates a window where concurrent modifications can
> occur during the copy operation.
>
> To prevent this from happening, we clamp the trace len to the max
> depth initially calculated with the buffer size and the size of
> a trace.
>
> Reported-by: syzbot+d1b7fa1092def3628bd7@syzkaller.appspotmail.com
> Closes: https://lore.kernel.org/all/691231dc.a70a0220.22f260.0101.GAE@google.com/T/
> Fixes: e17d62fedd10 ("bpf: Refactor stack map trace depth calculation into helper function")
> Tested-by: syzbot+d1b7fa1092def3628bd7@syzkaller.appspotmail.com
> Cc: Brahmajit Das <listout@listout.xyz>
> Signed-off-by: Arnaud Lecomte <contact@arnaud-lcm.com>
> ---
> Thanks Brahmajit Das for the initial fix he proposed that I tweaked
> with the correct justification and a better implementation in my
> opinion.
> ---
>  kernel/bpf/stackmap.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c
> index da3d328f5c15..e56752a9a891 100644
> --- a/kernel/bpf/stackmap.c
> +++ b/kernel/bpf/stackmap.c
> @@ -465,7 +465,6 @@ static long __bpf_get_stack(struct pt_regs *regs, struct task_struct *task,
>
>         if (trace_in) {
>                 trace = trace_in;
> -               trace->nr = min_t(u32, trace->nr, max_depth);
>         } else if (kernel && task) {
>                 trace = get_callchain_entry_for_task(task, max_depth);
>         } else {
> @@ -479,7 +478,8 @@ static long __bpf_get_stack(struct pt_regs *regs, struct task_struct *task,
>                 goto err_fault;
>         }
>
> -       trace_nr = trace->nr - skip;
> +       trace_nr = min(trace->nr, max_depth);

there is `trace->nr < skip` check right above, should it be moved here
and done against adjusted trace_nr (but before we subtract skip, of
course)?

> +       trace_nr = trace_nr - skip;
>         copy_len = trace_nr * elem_size;
>
>         ips = trace->ip + skip;
> --
> 2.43.0
>

Re: [PATCH] bpf-next: Prevent out of bound buffer write in __bpf_get_stack

Posted by Yonghong Song 1 day, 14 hours ago


On 1/4/26 12:52 PM, Arnaud Lecomte wrote:
> Syzkaller reported a KASAN slab-out-of-bounds write in __bpf_get_stack()
> during stack trace copying.
>
> The issue occurs when: the callchain entry (stored as a per-cpu variable)
> grow between collection and buffer copy, causing it to exceed the initially
> calculated buffer size based on max_depth.
>
> The callchain collection intentionally avoids locking for performance
> reasons, but this creates a window where concurrent modifications can
> occur during the copy operation.
>
> To prevent this from happening, we clamp the trace len to the max
> depth initially calculated with the buffer size and the size of
> a trace.
>
> Reported-by: syzbot+d1b7fa1092def3628bd7@syzkaller.appspotmail.com
> Closes: https://lore.kernel.org/all/691231dc.a70a0220.22f260.0101.GAE@google.com/T/
> Fixes: e17d62fedd10 ("bpf: Refactor stack map trace depth calculation into helper function")
> Tested-by: syzbot+d1b7fa1092def3628bd7@syzkaller.appspotmail.com
> Cc: Brahmajit Das <listout@listout.xyz>
> Signed-off-by: Arnaud Lecomte <contact@arnaud-lcm.com>

LGTM.

Acked-by: Yonghong Song <yonghong.song@linux.dev>