RE: [PATCH net-next] net: increase LL_MAX_HEADER if HYPERV_NET is enabled
From: KY Srinivasan <kys@microsoft.com>
Date: 2015-11-03 18:09:54
-----Original Message----- From: David Miller [mailto:davem@davemloft.net] Sent: Tuesday, November 3, 2015 7:33 AM To: KY Srinivasan <kys@microsoft.com> Cc: eric.dumazet@gmail.com; Haiyang Zhang <haiyangz@microsoft.com>; edumazet@google.com; netdev@vger.kernel.org Subject: Re: [PATCH net-next] net: increase LL_MAX_HEADER if HYPERV_NET is enabled From: KY Srinivasan <kys@microsoft.com> Date: Tue, 3 Nov 2015 07:59:36 +0000quoted
I have implemented the scheme we had discussed a few weeks ago. In this new implementation our driver is NOT requesting addition headroom - rndis header and the per packet state is being maintained outside of the skb. What I am seeing is that when I have LL_MAX_HEADER set to 220 bytes, even though our driver is not using the additional head room, I see about a 10% boost in the peak performance (about 34 Gbps on a 40Gbps interface). However, when I set the LL_MAX_HEADER value to the current default, the peak performance drops back to what we currently have (around 31 Gbps). In both these cases, there is no reallocation of skb since no additional headroom is being requested and yet there is a significant difference in performance. I trying to figure out why this is the case, your insights will be greatly appreciated.It probably has something to do with cache line or data alignment.
Thanks David. I too am leaning towards the same conclusion. Since I am not using any part of skb, it looks like the alignment issue is solved by having a larger LL_MAX_HEADER (220 bytes to be specific). I will do some bare metal testing on the setup we have and report back. Regards, K. Y