lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Tue, 12 May 2020 15:23:18 -0700
From:   Andrii Nakryiko <andrii.nakryiko@...il.com>
To:     Yonghong Song <yhs@...com>
Cc:     Andrii Nakryiko <andriin@...com>, bpf <bpf@...r.kernel.org>,
        Martin KaFai Lau <kafai@...com>,
        Networking <netdev@...r.kernel.org>,
        Alexei Starovoitov <ast@...com>,
        Daniel Borkmann <daniel@...earbox.net>,
        Kernel Team <kernel-team@...com>,
        Alexei Starovoitov <ast@...nel.org>
Subject: Re: [PATCH bpf-next 4/8] bpf: add WARN_ONCE if bpf_seq_read show()
 return a positive number

On Tue, May 12, 2020 at 8:54 AM Yonghong Song <yhs@...com> wrote:
>
> In seq_read() implementation, a positive integer return value
> of seq_ops->show() indicates that the current object seq_file
> buffer is discarded and next object should be checked.
> bpf_seq_read() implemented in a similar way if show()
> returns a positive integer value.
>
> But for bpf_seq_read(), show() didn't return positive integer for
> all currently supported targets. Let us add a WARN_ONCE for
> such cases so we can get an alert when things are changed.
>
> Suggested-by: Alexei Starovoitov <ast@...nel.org>
> Signed-off-by: Yonghong Song <yhs@...com>
> ---
>  kernel/bpf/bpf_iter.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/kernel/bpf/bpf_iter.c b/kernel/bpf/bpf_iter.c
> index 0a45a6cdfabd..b0c8b3bdf3b0 100644
> --- a/kernel/bpf/bpf_iter.c
> +++ b/kernel/bpf/bpf_iter.c
> @@ -120,6 +120,7 @@ static ssize_t bpf_seq_read(struct file *file, char __user *buf, size_t size,
>
>         err = seq->op->show(seq, p);
>         if (err > 0) {
> +               WARN_ONCE(1, "seq_ops->show() returns %d\n", err);

This makes it look like it's a bug or non-safe, honestly. I'd drop the
warning altogether, but if not, probably leaving a comment explaining
why we added WARN_ONCE here and that it's ok to remove it would be
good.

>                 /* object is skipped, decrease seq_num, so next
>                  * valid object can reuse the same seq_num.
>                  */
> @@ -156,6 +157,7 @@ static ssize_t bpf_seq_read(struct file *file, char __user *buf, size_t size,
>
>                 err = seq->op->show(seq, p);
>                 if (err > 0) {
> +                       WARN_ONCE(1, "seq_ops->show() returns %d\n", err);
>                         bpf_iter_dec_seq_num(seq);
>                         seq->count = offs;
>                 } else if (err < 0 || seq_has_overflowed(seq)) {
> --
> 2.24.1
>

Powered by blists - more mailing lists