From: Jesper Dangaard Brouer <brouer@redhat.com>
To: Tariq Toukan <tariqt@mellanox.com>
Cc: brouer@redhat.com, "David S. Miller" <davem@davemloft.net>,
netdev@vger.kernel.org, Eran Ben Elisha <eranbe@mellanox.com>,
Saeed Mahameed <saeedm@mellanox.com>,
Rana Shahout <ranas@mellanox.com>
Subject: Re: [PATCH net-next 8/8] net/mlx5e: XDP TX xmit more
Date: Tue, 20 Sep 2016 09:46:29 +0200 [thread overview]
Message-ID: <20160920094629.12e78a85@redhat.com> (raw)
In-Reply-To: <1474293539-2595-9-git-send-email-tariqt@mellanox.com>
On Mon, 19 Sep 2016 16:58:59 +0300
Tariq Toukan <tariqt@mellanox.com> wrote:
> From: Saeed Mahameed <saeedm@mellanox.com>
>
> Previously we rang XDP SQ doorbell on every forwarded XDP packet.
>
> Here we introduce a xmit more like mechanism that will queue up more
> than one packet into SQ (up to RX napi budget) w/o notifying the hardware.
>
> Once RX napi budget is consumed and we exit napi RX loop, we will
> flush (doorbell) all XDP looped packets in case there are such.
I've already raised strong concerns with this approach on the RFC
patchset. Of not really taking advantage of RX bulking.
Please do not ignore this!
If you can promise, that we/you will also try to other approach I'm
suggesting, then I'm fine with this patch.
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
Author of http://www.iptv-analyzer.org
LinkedIn: http://www.linkedin.com/in/brouer
next prev parent reply other threads:[~2016-09-20 7:46 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-19 13:58 [PATCH net-next 0/8] mlx5e XDP support Tariq Toukan
2016-09-19 13:58 ` [PATCH net-next 1/8] net/mlx5e: Build RX SKB on demand Tariq Toukan
2016-09-19 13:58 ` [PATCH net-next 2/8] net/mlx5e: Union RQ RX info per RQ type Tariq Toukan
2016-09-19 13:58 ` [PATCH net-next 3/8] net/mlx5e: Slightly reduce hardware LRO size Tariq Toukan
2016-09-19 13:58 ` [PATCH net-next 4/8] net/mlx5e: Dynamic RQ type infrastructure Tariq Toukan
2016-09-19 13:58 ` [PATCH net-next 5/8] net/mlx5e: XDP fast RX drop bpf programs support Tariq Toukan
2016-09-19 13:58 ` [PATCH net-next 6/8] net/mlx5e: Have a clear separation between different SQ types Tariq Toukan
2016-09-19 13:58 ` [PATCH net-next 7/8] net/mlx5e: XDP TX forwarding support Tariq Toukan
2016-09-20 8:29 ` Jesper Dangaard Brouer
2016-09-20 11:33 ` Jesper Dangaard Brouer
2016-09-20 12:53 ` Tariq Toukan
2016-09-20 15:40 ` Alexei Starovoitov
2016-09-20 15:51 ` Tom Herbert
2016-09-20 15:58 ` Eric Dumazet
2016-09-20 16:06 ` Jesper Dangaard Brouer
2016-09-20 16:13 ` Eric Dumazet
2016-09-20 16:27 ` Jesper Dangaard Brouer
2016-09-20 16:45 ` Alexei Starovoitov
2016-09-20 16:58 ` Thomas Graf
2016-09-20 17:39 ` Eric Dumazet
2016-09-20 18:59 ` Jesper Dangaard Brouer
2016-09-20 19:21 ` Tom Herbert
2016-09-20 20:47 ` Jesper Dangaard Brouer
2016-09-20 21:04 ` Jesper Dangaard Brouer
2016-09-20 16:00 ` Alexei Starovoitov
2016-09-20 15:57 ` Jesper Dangaard Brouer
2016-09-21 7:57 ` Tariq Toukan
2016-09-21 8:16 ` Jesper Dangaard Brouer
2016-09-19 13:58 ` [PATCH net-next 8/8] net/mlx5e: XDP TX xmit more Tariq Toukan
2016-09-20 7:46 ` Jesper Dangaard Brouer [this message]
2016-09-20 8:19 ` Tariq Toukan
2016-09-20 9:26 ` Jesper Dangaard Brouer
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160920094629.12e78a85@redhat.com \
--to=brouer@redhat.com \
--cc=davem@davemloft.net \
--cc=eranbe@mellanox.com \
--cc=netdev@vger.kernel.org \
--cc=ranas@mellanox.com \
--cc=saeedm@mellanox.com \
--cc=tariqt@mellanox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).