Re: [PATCH] bpf: fix stackmap overflow check in __bpf_get_stackid()

[Yonghong Song <yonghong.song@linux.dev>]

On 7/29/25 2:19 AM, Arnaud Lecomte wrote:
> Syzkaller reported a KASAN slab-out-of-bounds write in __bpf_get_stackid()
> when copying stack trace data. The issue occurs when the perf trace
>   contains more stack entries than the stack map bucket can hold,
>   leading to an out-of-bounds write in the bucket's data array.
> For build_id mode, we use sizeof(struct bpf_stack_build_id)
>   to determine capacity, and for normal mode we use sizeof(u64).
>
> Reported-by: syzbot+c9b724fbb41cf2538b7b@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/bug?extid=c9b724fbb41cf2538b7b
> Tested-by: syzbot+c9b724fbb41cf2538b7b@syzkaller.appspotmail.com
> Signed-off-by: Arnaud Lecomte <contact@arnaud-lcm.com>
> ---
>   kernel/bpf/stackmap.c | 12 +++++++++++-
>   1 file changed, 11 insertions(+), 1 deletion(-)
>
> diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c
> index 3615c06b7dfa..0f9f6e4b6fe9 100644
> --- a/kernel/bpf/stackmap.c
> +++ b/kernel/bpf/stackmap.c
> @@ -230,7 +230,7 @@ static long __bpf_get_stackid(struct bpf_map *map,
>   	struct bpf_stack_map *smap = container_of(map, struct bpf_stack_map, map);
>   	struct stack_map_bucket *bucket, *new_bucket, *old_bucket;
>   	u32 skip = flags & BPF_F_SKIP_FIELD_MASK;
> -	u32 hash, id, trace_nr, trace_len, i;
> +	u32 hash, id, trace_nr, trace_len, i, max_depth;
>   	bool user = flags & BPF_F_USER_STACK;
>   	u64 *ips;
>   	bool hash_matches;
> @@ -241,6 +241,16 @@ static long __bpf_get_stackid(struct bpf_map *map,
>   
>   	trace_nr = trace->nr - skip;
>   	trace_len = trace_nr * sizeof(u64);
> +
> +	/* Clamp the trace to max allowed depth */
> +	if (stack_map_use_build_id(map))
> +		max_depth = smap->map.value_size / sizeof(struct bpf_stack_build_id);
> +	else
> +		max_depth = smap->map.value_size / sizeof(u64);

Replace the above with
	max_depth = smap->map.value_size / stack_map_data_size(map)
?

> +
> +	if (trace_nr > max_depth)
> +		trace_nr = max_depth;
> +
>   	ips = trace->ip + skip;
>   	hash = jhash2((u32 *)ips, trace_len / sizeof(u32), 0);
>   	id = hash & (smap->n_buckets - 1);