Re: [RFC bpf-next v1 0/7] Add kfunc bpf_xdp_pull_data

[Gal Pressman <gal@nvidia.com>]

On 26/08/2025 1:41, Jakub Kicinski wrote:
> On Mon, 25 Aug 2025 12:39:11 -0700 Amery Hung wrote:
>> Hi all,
>>
>> This patchset introduces a new kfunc bpf_xdp_pull_data() to allow
>> pulling nonlinear xdp data. This may be useful when a driver places
>> headers in fragments. When an xdp program would like to keep parsing
>> packet headers using direct packet access, it can call
>> bpf_xdp_pull_data() to make the header available in the linear data
>> area. The kfunc can also be used to decapsulate the header in the
>> nonlinear data, as currently there is no easy way to do this.
>>
>> This patchset also tries to fix an issue in the mlx5e driver. The driver
>> curretly assumes the packet layout to be unchanged after xdp program
>> runs and may generate packet with corrupted data or trigger kernel warning
>> if xdp programs calls layout-changing kfunc such as bpf_xdp_adjust_tail(),
>> bpf_xdp_adjust_head() or bpf_xdp_pull_data() introduced in this set.
>>
>> Tested with the added bpf selftest using bpf test_run and also on
>> mlx5e with the tools/testing/selftests/drivers/net/xdp.py. mlx5e with
>> striding RQ will produce xdp_buff with empty linear data.
>> xdp.test_xdp_native_pass_mb would fail to parse the header before this
>> patchset.
>>
>> Grateful for any feedback (especially the driver part).
> 
> CC: Gal, this is the correct way to resolve the XDP not having headers
> in the first frag for mlx5.

Thanks for the heads up Jakub, CC'd Dragos and Nimrod to review.