Re: [PATCH v5 bpf-next 2/4] bpf: Support setting variable-length tunnel options

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Aug 23, 2022 at 9:41 PM Shmulik Ladkani
<shmulik@xxxxxxxxxxxxxxxx> wrote:
>
> Existing 'bpf_skb_set_tunnel_opt' allows setting tunnel options given
> an option buffer (ARG_PTR_TO_MEM) and the compile-time fixed buffer
> size (ARG_CONST_SIZE).
>
> However, in certain cases we wish to set tunnel options of dynamic
> length.
>
> For example, we have an ebpf program that gets geneve options on
> incoming packets, stores them into a map (using a key representing
> the incoming flow), and later needs to assign *same* options to
> reply packets (belonging to same flow).
>
> This is currently imposssible without knowing sender's exact geneve
> options length, which unfortunately is dymamic.
>
> Introduce 'bpf_skb_set_tunnel_opt_dynptr'.
>
> This is a variant of 'bpf_skb_set_tunnel_opt' which gets a bpf dynamic
> pointer (ARG_PTR_TO_DYNPTR) parameter 'ptr' whose data points to the
> options buffer, and 'len', the byte length of options data caller wishes
> to copy into ip_tunnnel_info.
> 'len' must never exceed the dynptr's internal size, o/w EINVAL is
> returned.
>
> Signed-off-by: Shmulik Ladkani <shmulik.ladkani@xxxxxxxxx>
> ---
> v3: Avoid 'inline' for the __bpf_skb_set_tunopt helper function
> v4: change API to be based on bpf_dynptr, suggested by John Fastabend <john.fastabend@xxxxxxxxx>
> ---
>  include/uapi/linux/bpf.h       | 12 ++++++++++++
>  net/core/filter.c              | 36 ++++++++++++++++++++++++++++++++--
>  tools/include/uapi/linux/bpf.h | 12 ++++++++++++
>  3 files changed, 58 insertions(+), 2 deletions(-)
>
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 644600dbb114..c7b313e30635 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -5367,6 +5367,17 @@ union bpf_attr {
>   *     Return
>   *             Current *ktime*.
>   *
> + * long bpf_skb_set_tunnel_opt_dynptr(struct sk_buff *skb, struct bpf_dynptr *opt, u32 len)

why can't we rely on dynptr's len instead of specifying extra one
here? dynptr is a range of memory, so just specify that you take that
entire range?

And then we'll have (or partially already have) generic dynptr helpers
to adjust internal dynptr offset and len.

> + *     Description
> + *             Set tunnel options metadata for the packet associated to *skb*
> + *             to the variable length *len* bytes of option data pointed to
> + *             by the *opt* dynptr.
> + *
> + *             See also the description of the **bpf_skb_get_tunnel_opt**\ ()
> + *             helper for additional information.
> + *     Return
> + *             0 on success, or a negative error in case of failure.
> + *
>   */

[...]



[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux