[dpdk-dev] Why DPDK is not using compressed TRIE for LPM6?

[Vladimir Medvedkin]

Hi all,

Currently I'm working on compressed trie implementation for LPM4 lib. I
think it could be adapted for LPM6.

2017-05-29 12:30 GMT+03:00 Bruce Richardson <bruce.richardson at intel.com>:

> On Sat, May 27, 2017 at 12:34:57AM +0530, Atul Shree wrote:
> > Hello All,
> >
> > I was doing some experiments related to LPM6 look up and I have added 20K
> > entries in the table. By looking at the rte_lpm6_lookup() code I found an
> > opportunity to compress the TRIE and there is a significant improvement
> > after compression.
> >
>
> Although I'm maintainer for LPM library, I'm not the original author of
> the LPM6 code. However, I'll give my thoughts here. Adding Cristian D. on
> CC as he was involved in the original implementation, IIRC.
>
> > Here are my questions:
> > Q1: Why DPDK is not doing the compression?
>
> It's probably not a deliberate omission, more likely that nobody has
> done it.
>
> > Q2. In the worst case the table will behave like an uncompressed TRIE and
> > in other cases, there is a scope of improvement. Is it worth doing?
> >
>
> If there is improvement in the normal case, with the worst-case perf
> being no worse, it sounds like it may be worth doing. Feel free to
> submit patches for evaluation on the list.
>
> Regards,
> /Bruce
>
>


-- 
Regards,
Vladimir