[xdp-hints] Re: [PATCH RFC bpf-next 32/52] bpf, cpumap: switch to GRO from netif_receive_skb_list()

[Lorenzo Bianconi <lorenzo@kernel.org>]

[-- Attachment #1: Type: text/plain, Size: 5583 bytes --]

> From: Alexander Lobakin <aleksander.lobakin@intel.com>
> Date: Thu, 8 Aug 2024 13:57:00 +0200
> 
> > From: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
> > Date: Thu, 8 Aug 2024 06:54:06 +0200
> > 
> >>> Hi Alexander,
> >>>
> >>> On Tue, Jun 28, 2022, at 12:47 PM, Alexander Lobakin wrote:
> >>>> cpumap has its own BH context based on kthread. It has a sane batch
> >>>> size of 8 frames per one cycle.
> >>>> GRO can be used on its own, adjust cpumap calls to the
> >>>> upper stack to use GRO API instead of netif_receive_skb_list() which
> >>>> processes skbs by batches, but doesn't involve GRO layer at all.
> >>>> It is most beneficial when a NIC which frame come from is XDP
> >>>> generic metadata-enabled, but in plenty of tests GRO performs better
> >>>> than listed receiving even given that it has to calculate full frame
> >>>> checksums on CPU.
> >>>> As GRO passes the skbs to the upper stack in the batches of
> >>>> @gro_normal_batch, i.e. 8 by default, and @skb->dev point to the
> >>>> device where the frame comes from, it is enough to disable GRO
> >>>> netdev feature on it to completely restore the original behaviour:
> >>>> untouched frames will be being bulked and passed to the upper stack
> >>>> by 8, as it was with netif_receive_skb_list().
> >>>>
> >>>> Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
> >>>> ---
> >>>>  kernel/bpf/cpumap.c | 43 ++++++++++++++++++++++++++++++++++++++-----
> >>>>  1 file changed, 38 insertions(+), 5 deletions(-)
> >>>>
> >>>
> >>> AFAICT the cpumap + GRO is a good standalone improvement. I think
> >>> cpumap is still missing this.
> > 
> > The only concern for having GRO in cpumap without metadata from the NIC
> > descriptor was that when the checksum status is missing, GRO calculates
> > the checksum on CPU, which is not really fast.
> > But I remember sometimes GRO was faster despite that.
> > 
> >>>
> >>> I have a production use case for this now. We want to do some intelligent
> >>> RX steering and I think GRO would help over list-ified receive in some cases.
> >>> We would prefer steer in HW (and thus get existing GRO support) but not all
> >>> our NICs support it. So we need a software fallback.
> >>>
> >>> Are you still interested in merging the cpumap + GRO patches?
> > 
> > For sure I can revive this part. I was planning to get back to this
> > branch and pick patches which were not related to XDP hints and send
> > them separately.
> > 
> >>
> >> Hi Daniel and Alex,
> >>
> >> Recently I worked on a PoC to add GRO support to cpumap codebase:
> >> - https://github.com/LorenzoBianconi/bpf-next/commit/a4b8264d5000ecf016da5a2dd9ac302deaf38b3e
> >>   Here I added GRO support to cpumap through gro-cells.
> >> - https://github.com/LorenzoBianconi/bpf-next/commit/da6cb32a4674aa72401c7414c9a8a0775ef41a55
> >>   Here I added GRO support to cpumap trough napi-threaded APIs (with a some
> >>   changes to them).
> > 
> > Hmm, when I was testing it, adding a whole NAPI to cpumap was sorta
> > overkill, that's why I separated GRO structure from &napi_struct.
> > 
> > Let me maybe find some free time, I would then test all 3 solutions
> > (mine, gro_cells, threaded NAPI) and pick/send the best?
> > 
> >>
> >> Please note I have not run any performance tests so far, just verified it does
> >> not crash (I was planning to resume this work soon). Please let me know if it
> >> works for you.
> 
> I did tests on both threaded NAPI for cpumap and my old implementation
> with a traffic generator and I have the following (in Kpps):
> 
>             direct Rx    direct GRO    cpumap    cpumap GRO
> baseline    2900         5800          2700      2700 (N/A)
> threaded                               2300      4000
> old GRO                                2300      4000

cool, very nice improvement

> 
> IOW,
> 
> 1. There are no differences in perf between Lorenzo's threaded NAPI
>    GRO implementation and my old implementation, but Lorenzo's is also
>    a very nice cleanup as it switches cpumap to threaded NAPI completely
>    and the final diffstat even removes more lines than adds, while mine
>    adds a bunch of lines and refactors a couple hundred, so I'd go with
>    his variant.
> 
> 2. After switching to NAPI, the performance without GRO decreases (2.3
>    Mpps vs 2.7 Mpps), but after enabling GRO the perf increases hugely
>    (4 Mpps vs 2.7 Mpps) even though the CPU needs to compute checksums
>    manually.
> 
> Note that the code is not polished to the top and I also have a good
> improvement for allocating skb heads from the percpu NAPI cache in my
> old tree which I'm planning to add to the series, so the final
> improvement will be even bigger.
> 
> + after we find how to pass checksum hint to cpumap, it will be yet
> another big improvement for GRO (current code won't benefit from
> this at all)
> 
> To Lorenzo:
> 
> Would it be fine if I prepare a series containing your patch for
> threaded NAPI for cpumap (I'd polish it and break into 2 or 3) +
> skb allocation optimization and send it OR you wanted to send this
> on your own? I'm fine with either, in the first case, everything
> would land within one series with the respective credits; in case
> of the latter, I'd need to send a followup :)

Sure, I am fine to send my codebase into a bigger series.
Thanks a lot for testing :)

Regards,
Lorenzo

> 
> >>
> >> Regards,
> >> Lorenzo
> >>
> >>>
> >>> Thanks,
> >>> Daniel
> 
> Thanks,
> Olek

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]