22/11/2023 10:37, fengchengwen:
> On 2023/11/22 16:28, Robin Jarry wrote:
> > Stephen Hemminger, Nov 21, 2023 at 18:27:
> >> On Tue, 21 Nov 2023 22:51:19 +0530
> >> Jerin Jacob wrote:
> >> > IMO, That may be too much of non-productive work to save a couple of
> >> > lines(seed array) of duplicate
On 2023/11/22 16:28, Robin Jarry wrote:
> Stephen Hemminger, Nov 21, 2023 at 18:27:
>> On Tue, 21 Nov 2023 22:51:19 +0530
>> Jerin Jacob wrote:
>> > IMO, That may be too much of non-productive work to save a couple of
>> > lines(seed array) of duplicate code between c and python.
>>
>> But the too
Stephen Hemminger, Nov 21, 2023 at 18:27:
On Tue, 21 Nov 2023 22:51:19 +0530
Jerin Jacob wrote:
> IMO, That may be too much of non-productive work to save a couple of
> lines(seed array) of duplicate code between c and python.
But the tool is incomplete right now, it doesn't cover all the drive
On Tue, 21 Nov 2023 22:51:19 +0530
Jerin Jacob wrote:
> On Tue, Nov 21, 2023 at 10:34 PM Stephen Hemminger
> wrote:
> >
> > On Tue, 21 Nov 2023 22:21:54 +0530
> > Jerin Jacob wrote:
> >
> > > > The tool should not need to have driver specific tables like this.
> > > > DPDK is already riddled
On Tue, Nov 21, 2023 at 10:34 PM Stephen Hemminger
wrote:
>
> On Tue, 21 Nov 2023 22:21:54 +0530
> Jerin Jacob wrote:
>
> > > The tool should not need to have driver specific tables like this.
> > > DPDK is already riddled with enough driver specific quirks..
> > >
> > > That shows a flaw in the
On Tue, 21 Nov 2023 22:21:54 +0530
Jerin Jacob wrote:
> > The tool should not need to have driver specific tables like this.
> > DPDK is already riddled with enough driver specific quirks..
> >
> > That shows a flaw in the rss design, which should have been fixed.
>
> Every NIC's implements st
On Tue, Nov 21, 2023 at 10:08 PM Stephen Hemminger
wrote:
>
> On Mon, 23 Oct 2023 10:07:11 +0200
> Robin Jarry wrote:
>
> > +class DriverInfo:
> > +def __init__(self, key: bytes = None, reta_size: int = None):
> > +self.__key = key
> > +self.__reta_size = reta_size
> > +
> > +
21/11/2023 17:38, Stephen Hemminger:
> The tool should not need to have driver specific tables like this.
> DPDK is already riddled with enough driver specific quirks..
>
> That shows a flaw in the rss design, which should have been fixed.
> I should have seen this in the earlier versions, would h
On Mon, 23 Oct 2023 10:07:11 +0200
Robin Jarry wrote:
> +class DriverInfo:
> +def __init__(self, key: bytes = None, reta_size: int = None):
> +self.__key = key
> +self.__reta_size = reta_size
> +
> +def rss_key(self) -> bytes:
> +return self.__key
> +
> +def re
> -Original Message-
> From: Robin Jarry
> Sent: Monday, October 23, 2023 1:37 PM
> To: dev@dpdk.org
> Cc: Sunil Kumar Kori ; tho...@monjalon.net;
> jerinjac...@gmail.com; Robin Jarry
> Subject: [EXT] [PATCH 1/2] usertools/rss: add driver abstraction
[EXT] Re: [PATCH 1/2] usertools/rss: add driver abstractions
>
> External Email
>
> --
> Hi, Jerin, Sunil Komar,
>
> Could you have a look at this series and report if it brings the needed
> abstractions for integrating CNXK?
>
> Thanks!
Hi, Jerin, Sunil Komar,
Could you have a look at this series and report if it brings the needed
abstractions for integrating CNXK?
Thanks!
The default RETA size is not the same for all drivers. In some drivers
(mlx5), the RETA size may also be dependent on the number of RX queues.
Introduce a new DriverInfo abstraction for known keys. Define a simple
API to expose the RSS key and RETA size (based on the number of RX
queues).
Use tha
13 matches
Mail list logo
