netdev - Re: [PATCH bpf-next 3/5] libbpf: Add bpf

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]

Message-ID: <CAEf4BzbakwDBO1w-R3dAVOiQ5DQ9QKwwTMO6N2YMkUtq_Jsdfw@mail.gmail.com>
Date:   Mon, 23 Aug 2021 21:54:55 -0700
From:   Andrii Nakryiko <andrii.nakryiko@...il.com>
To:     Dave Marchevsky <davemarchevsky@...com>
Cc:     bpf <bpf@...r.kernel.org>, Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Andrii Nakryiko <andrii@...nel.org>,
        Yonghong Song <yhs@...com>,
        Florent Revest <revest@...omium.org>,
        Networking <netdev@...r.kernel.org>,
        "open list:KERNEL SELFTEST FRAMEWORK" 
        <linux-kselftest@...r.kernel.org>
Subject: Re: [PATCH bpf-next 3/5] libbpf: Add bpf_vprintk convenience macro

On Fri, Aug 20, 2021 at 7:59 PM Dave Marchevsky <davemarchevsky@...com> wrote:
>
> bpf_vprintk functions similarly to BPF_SEQ_PRINTF and BPF_SNPRINTF
> macros elsewhere in the file - it allows use of bpf_trace_vprintk
> without manual conversion of varargs to u64 array.
>
> Like the bpf_printk macro, bpf_vprintk is meant to be the main interface
> to the bpf_trace_vprintk helper and thus is uncapitalized.
>
> Signed-off-by: Dave Marchevsky <davemarchevsky@...com>
> ---
>  tools/lib/bpf/bpf_helpers.h | 18 ++++++++++++++++++
>  1 file changed, 18 insertions(+)
>
> diff --git a/tools/lib/bpf/bpf_helpers.h b/tools/lib/bpf/bpf_helpers.h
> index b9987c3efa3c..43c8115956c3 100644
> --- a/tools/lib/bpf/bpf_helpers.h
> +++ b/tools/lib/bpf/bpf_helpers.h
> @@ -224,4 +224,22 @@ enum libbpf_tristate {
>                      ___param, sizeof(___param));               \
>  })
>
> +/*
> + * bpf_vprintk wraps the bpf_trace_printk helper with variadic arguments
> + * instead of an array of u64.
> + */
> +#define bpf_vprintk(fmt, args...)                              \

Given BPF_SNPRINTF and BPF_SEQ_PRINTF, should we call this one in
all-caps as well?

> +({                                                             \
> +       static const char ___fmt[] = fmt;                       \
> +       unsigned long long ___param[___bpf_narg(args)];         \

I wonder how hard would it be to still use bpf_trace_printk() if the
number of input arguments is less than 3? That way you could use
BPF_PRINTF() everywhere, even on old kernels (you just need to
remember to use < 3 arguments). WDYT? It might be more challenging
than it seems, given it's hard to do conditionals inside #defines :(

> +                                                               \
> +       _Pragma("GCC diagnostic push")                          \
> +       _Pragma("GCC diagnostic ignored \"-Wint-conversion\"")  \
> +       ___bpf_fill(___param, args);                            \
> +       _Pragma("GCC diagnostic pop")                           \
> +                                                               \
> +       bpf_trace_vprintk(___fmt, sizeof(___fmt),               \
> +                    ___param, sizeof(___param));               \
> +})
> +
>  #endif
> --
> 2.30.2
>