In the __cgroup_bpf_query() function, it is possible to dereference
the null pointer in the line id = prog->aux->id; since there is no
check for a non-zero value of the variable prog.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Fixes: af6eea57437a ("bpf: Implement bpf_link-based cgroup BPF program attachment")
Cc: [email protected]
Signed-off-by: Mikhail Lobanov <[email protected]>
---
kernel/bpf/cgroup.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
index 491d20038cbe..7f2db96f0c6a 100644
--- a/kernel/bpf/cgroup.c
+++ b/kernel/bpf/cgroup.c
@@ -1092,6 +1092,8 @@ static int __cgroup_bpf_query(struct cgroup *cgrp, const union bpf_attr *attr,
i = 0;
hlist_for_each_entry(pl, progs, node) {
prog = prog_list_prog(pl);
+ if (!prog_list_prog(pl))
+ continue;
id = prog->aux->id;
if (copy_to_user(prog_ids + i, &id, sizeof(id)))
return -EFAULT;
--
2.43.0
On 4/6/24 8:14 AM, Mikhail Lobanov wrote:
> In the __cgroup_bpf_query() function, it is possible to dereference
> the null pointer in the line id = prog->aux->id; since there is no
> check for a non-zero value of the variable prog.
>
> Found by Linux Verification Center (linuxtesting.org) with SVACE.
>
> Fixes: af6eea57437a ("bpf: Implement bpf_link-based cgroup BPF program attachment")
> Cc: [email protected]
> Signed-off-by: Mikhail Lobanov <[email protected]>
> ---
> kernel/bpf/cgroup.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
> index 491d20038cbe..7f2db96f0c6a 100644
> --- a/kernel/bpf/cgroup.c
> +++ b/kernel/bpf/cgroup.c
> @@ -1092,6 +1092,8 @@ static int __cgroup_bpf_query(struct cgroup *cgrp, const union bpf_attr *attr,
> i = 0;
> hlist_for_each_entry(pl, progs, node) {
> prog = prog_list_prog(pl);
> + if (!prog_list_prog(pl))
prog cannot be null. It is under cgroup_lock().
> + continue;
> id = prog->aux->id;
> if (copy_to_user(prog_ids + i, &id, sizeof(id)))
> return -EFAULT;