Re: [PATCH bpf] bpf, test_run: fix &xdp_frame misplacement for LIVE_FRAMES

From: Alexander Lobakin
Date: Fri Feb 10 2023 - 08:20:13 EST


From: Alexander Lobakin <alexandr.lobakin@xxxxxxxxx>
Date: Fri, 10 Feb 2023 13:31:28 +0100

> From: Toke Høiland-Jørgensen <toke@xxxxxxxxxx>
> Date: Thu, 09 Feb 2023 21:58:07 +0100
>
>> Alexander Lobakin <alexandr.lobakin@xxxxxxxxx> writes:

[...]

>> Hmm, IIRC my reasoning was that both those cache lines will be touched
>> by the code in xdp_test_run_batch(), so it wouldn't matter? But if
>> there's a performance benefit I don't mind adding an explicit alignment
>> annotation, certainly!
>
> Let me retest both ways and will see. I saw some huge CPU loads on
> reading xdpf in ice_xdp_xmit(), so that was my first thought.

No visible difference in perf and CPU load... Ok, aligning isn't worth it.

>
>>
>>> (but in bpf-next probably)
>>
>> Yeah...
>>
>> -Toke
>>
Olek