bpf, iter: Fix the condition on p when calling stop.
authorHao Luo <haoluo@google.com>
Fri, 5 Aug 2022 21:48:16 +0000 (14:48 -0700)
committerAlexei Starovoitov <ast@kernel.org>
Tue, 9 Aug 2022 16:12:00 +0000 (09:12 -0700)
In bpf_seq_read, seq->op->next() could return an ERR and jump to
the label stop. However, the existing code in stop does not handle
the case when p (returned from next()) is an ERR. Adds the handling
of ERR of p by converting p into an error and jumping to done.

Because all the current implementations do not have a case that
returns ERR from next(), so this patch doesn't have behavior changes
right now.

Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Hao Luo <haoluo@google.com>
Link: https://lore.kernel.org/r/20220805214821.1058337-4-haoluo@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
kernel/bpf/bpf_iter.c

index 2726a59..4b112aa 100644 (file)
@@ -197,6 +197,11 @@ static ssize_t bpf_seq_read(struct file *file, char __user *buf, size_t size,
        }
 stop:
        offs = seq->count;
+       if (IS_ERR(p)) {
+               seq->op->stop(seq, NULL);
+               err = PTR_ERR(p);
+               goto done;
+       }
        /* bpf program called if !p */
        seq->op->stop(seq, p);
        if (!p) {