linux-kernel - Re: [PATCH bpf V7 1/7] selftests/bpf: xdp_hw_metadata default disable 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]

Date:   Tue, 11 Apr 2023 15:42:13 -0700
From:   Stanislav Fomichev <sdf@...gle.com>
To:     Jesper Dangaard Brouer <brouer@...hat.com>
Cc:     bpf@...r.kernel.org,
        Toke Høiland-Jørgensen <toke@...hat.com>,
        netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
        martin.lau@...nel.org, ast@...nel.org, daniel@...earbox.net,
        alexandr.lobakin@...el.com, larysa.zaremba@...el.com,
        xdp-hints@...-project.net, anthony.l.nguyen@...el.com,
        yoong.siang.song@...el.com, boon.leong.ong@...el.com,
        intel-wired-lan@...ts.osuosl.org, pabeni@...hat.com,
        jesse.brandeburg@...el.com, kuba@...nel.org, edumazet@...gle.com,
        john.fastabend@...il.com, hawk@...nel.org, davem@...emloft.net,
        tariqt@...dia.com, saeedm@...dia.com, leon@...nel.org,
        linux-rdma@...r.kernel.org
Subject: Re: [PATCH bpf V7 1/7] selftests/bpf: xdp_hw_metadata default disable bpf_printk

On Sat, Apr 8, 2023 at 12:24 PM Jesper Dangaard Brouer
<brouer@...hat.com> wrote:
>
> The tool xdp_hw_metadata can be used by driver developers
> implementing XDP-hints kfuncs.  The tool transfers the
> XDP-hints via metadata information to an AF_XDP userspace
> process. When everything works the bpf_printk calls are
> unncesssary.  Thus, disable bpf_printk by default, but
> make it easy to reenable for driver developers to use
> when debugging their driver implementation.
>
> This also converts bpf_printk "forwarding UDP:9091 to AF_XDP"
> into a code comment.  The bpf_printk's that are important
> to the driver developers is when bpf_xdp_adjust_meta fails.
> The likely mistake from driver developers is expected to
> be that they didn't implement XDP metadata adjust support.
>
> Signed-off-by: Jesper Dangaard Brouer <brouer@...hat.com>
> ---
>  .../testing/selftests/bpf/progs/xdp_hw_metadata.c  |   16 ++++++++++++++--
>  1 file changed, 14 insertions(+), 2 deletions(-)
>
> diff --git a/tools/testing/selftests/bpf/progs/xdp_hw_metadata.c b/tools/testing/selftests/bpf/progs/xdp_hw_metadata.c
> index 4c55b4d79d3d..980eb60d8e5b 100644
> --- a/tools/testing/selftests/bpf/progs/xdp_hw_metadata.c
> +++ b/tools/testing/selftests/bpf/progs/xdp_hw_metadata.c
> @@ -5,6 +5,19 @@
>  #include <bpf/bpf_helpers.h>
>  #include <bpf/bpf_endian.h>
>
> +/* Per default below bpf_printk() calls are disabled.  Can be
> + * reenabled manually for convenience by XDP-hints driver developer,
> + * when troublshooting the drivers kfuncs implementation details.
> + *
> + * Remember BPF-prog bpf_printk info output can be access via:
> + *  /sys/kernel/debug/tracing/trace_pipe
> + */
> +//#define DEBUG        1
> +#ifndef DEBUG
> +#undef  bpf_printk
> +#define bpf_printk(fmt, ...) ({})
> +#endif

Are you planning to eventually do somethike similar to what I've
mentioned in [0]? If not, should I try to send a patch?

0: https://lore.kernel.org/netdev/CAKH8qBupRYEg+SPMTMb4h532GESG7P1QdaFJ-+zrbARVN9xrdA@mail.gmail.com/

> +
>  struct {
>         __uint(type, BPF_MAP_TYPE_XSKMAP);
>         __uint(max_entries, 256);
> @@ -49,11 +62,10 @@ int rx(struct xdp_md *ctx)
>         if (!udp)
>                 return XDP_PASS;
>
> +       /* Forwarding UDP:9091 to AF_XDP */
>         if (udp->dest != bpf_htons(9091))
>                 return XDP_PASS;
>
> -       bpf_printk("forwarding UDP:9091 to AF_XDP");
> -
>         ret = bpf_xdp_adjust_meta(ctx, -(int)sizeof(struct xdp_meta));
>         if (ret != 0) {
>                 bpf_printk("bpf_xdp_adjust_meta returned %d", ret);
>
>