[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CACLgkEb+5OU+op+FvrrqiA1mgsp7NbA=KB_dCa532R6AL2c3Kw@mail.gmail.com>
Date: Tue, 20 May 2025 21:14:03 +0530
From: Krishna Kumar <krikku@...il.com>
To: Ahmed Zaki <ahmed.zaki@...el.com>
Cc: netdev@...r.kernel.org, davem@...emloft.net, anthony.l.nguyen@...el.com,
przemyslaw.kitszel@...el.com, edumazet@...gle.com,
intel-wired-lan@...ts.osuosl.org, andrew+netdev@...n.ch, kuba@...nel.org,
pabeni@...hat.com, sridhar.samudrala@...el.com, krishna.ku@...pkart.com
Subject: Re: [PATCH] net: ice: Perform accurate aRFS flow match
On Tue, May 20, 2025 at 8:18 PM Ahmed Zaki <ahmed.zaki@...el.com> wrote:
> Are these numbers with the patch applied? Can we get a w/o and with patch?
The numbers are for the original driver vs the new driver. For
purposes of calculating
them, I had sysctls in the code (removed for submission) in the
original and the new
driver at each of the locations given above (a simple
atomic64_inc(&kk_arfs_add), etc).
> A table might be better to visualize, also may be drop the
> "rps_flow_cnt=1024*" case. I think it is enough to show min and max ones.
I will re-send the patch after adding this table (and Simon's feedback).
> Also, please add instructions on how to get these values, so that
> validation team may be able to replicate.
I have a large set of scripts that measure each of these parameters.
If you wish, I can
send you the set of scripts separately.
Regards,
- Krishna
Powered by blists - more mailing lists