[PATCH v7 00/25] Building a better rte_flow parser

[Burakov, Anatoly]

On 10/14/2026 10:50 PM, Thomas Monjalon wrote:
> 13/10/2026 19:18, Bruce Richardson:
>> Patchset applied to dpdk-next-net-intel.
>>
>> As well as the arm build issue, I also fixed a number of checkpatch issues,
>> since there were a large number on this set. Most I kept unfixed, since I
>> felt the exceptions were reasonable, but please check what was applied in
>> case some issue was introduced with the checkpatch changes I made on apply.
> 
> Unfortunately it fails compiling on my FreeBSD,
> because of an internal API included in a header included in testpmd.
> 
> You can make it fail reliably with Clang on Linux with this patch:
> https://patches.dpdk.org/project/dpdk/patch/20261014201627.571496-1-thomas@monjalon.net/
> 
> And you can fix the problem with this patch:
> https://patches.dpdk.org/project/dpdk/patch/20261014203435.583064-1-thomas@monjalon.net/
> 
> Please review them.
> 
> In the meantime I cannot pull this series in main,
> except the first 2 patches for ethdev API.

Done and done!

I wonder if some of that could've been avoided if the CI was better at 
applying patches on top of other trees instead of always trying main. I 
remember trying to do it with VFIO patchset, and I don't think I 
succeeded in getting it to apply/build on top of next-net even though I 
did the recheck requests and everything. I couldn't really get this 
patchset building through CI so some Arm compile issues have slipped by 
me as Bruce mentioned.

-- 
Thanks,
Anatoly