netdev - Re: [PATCH bpf-next] tools/resolve_btfids: Warn when having multiple IDs for single type

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 linux-cve-announce 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, 5 Jan 2021 13:15:39 -0800
From:   Andrii Nakryiko <andrii.nakryiko@...il.com>
To:     Jiri Olsa <jolsa@...nel.org>
Cc:     Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Andrii Nakryiko <andriin@...com>,
        Qais Yousef <qais.yousef@....com>,
        Networking <netdev@...r.kernel.org>, bpf <bpf@...r.kernel.org>,
        Martin KaFai Lau <kafai@...com>,
        Song Liu <songliubraving@...com>, Yonghong Song <yhs@...com>,
        John Fastabend <john.fastabend@...il.com>,
        KP Singh <kpsingh@...omium.org>
Subject: Re: [PATCH bpf-next] tools/resolve_btfids: Warn when having multiple
 IDs for single type

On Tue, Jan 5, 2021 at 7:41 AM Jiri Olsa <jolsa@...nel.org> wrote:
>
> The kernel image can contain multiple types (structs/unions)
> with the same name. This causes distinct type hierarchies in
> BTF data and makes resolve_btfids fail with error like:
>
>   BTFIDS  vmlinux
> FAILED unresolved symbol udp6_sock
>
> as reported by Qais Yousef [1].
>
> This change adds warning when multiple types of the same name
> are detected:
>
>   BTFIDS  vmlinux
> WARN: multiple IDs found for 'file' (526, 113351)
> WARN: multiple IDs found for 'sk_buff' (2744, 113958)
>
> We keep the lower ID for the given type instance and let the
> build continue.
>
> [1] https://lore.kernel.org/lkml/20201229151352.6hzmjvu3qh6p2qgg@e107158-lin/
> Reported-by: Qais Yousef <qais.yousef@....com>
> Signed-off-by: Jiri Olsa <jolsa@...nel.org>
> ---

see comments below, but otherwise lgtm

Acked-by: Andrii Nakryiko <andrii@...nel.org>

>  tools/bpf/resolve_btfids/main.c | 11 +++++++++--
>  1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/tools/bpf/resolve_btfids/main.c b/tools/bpf/resolve_btfids/main.c
> index e3ea569ee125..36a3b1024cdc 100644
> --- a/tools/bpf/resolve_btfids/main.c
> +++ b/tools/bpf/resolve_btfids/main.c
> @@ -139,6 +139,8 @@ int eprintf(int level, int var, const char *fmt, ...)
>  #define pr_debug2(fmt, ...) pr_debugN(2, pr_fmt(fmt), ##__VA_ARGS__)
>  #define pr_err(fmt, ...) \
>         eprintf(0, verbose, pr_fmt(fmt), ##__VA_ARGS__)
> +#define pr_info(fmt, ...) \
> +       eprintf(0, verbose, pr_fmt(fmt), ##__VA_ARGS__)

how is it different from pr_err? Did you forget to update verboseness
levels or it's intentional?

>
>  static bool is_btf_id(const char *name)
>  {
> @@ -526,8 +528,13 @@ static int symbols_resolve(struct object *obj)
>
>                 id = btf_id__find(root, str);
>                 if (id) {
> -                       id->id = type_id;
> -                       (*nr)--;
> +                       if (id->id) {
> +                               pr_info("WARN: multiple IDs found for '%s' (%d, %d)\n",
> +                                       str, id->id, type_id);
> +                       } else {
> +                               id->id = type_id;
> +                               (*nr)--;

btw, there is a nasty shadowing of nr variable, which is used both for
the for() loop condition (as int) and as `int *` inside the loop body.
It's better to rename inner (or outer) nr, it's extremely confusing as
is.

> +                       }
>                 }
>         }
>
> --
> 2.26.2
>