From: Erez Shitrit <erezsh-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
To: Jason Gunthorpe
<jgunthorpe-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
Cc: Erez Shitrit <erezsh-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>,
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>,
"linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org"
<linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>,
valex-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org,
Leon Romanovsky <leonro-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Subject: Re: [RFC for accelerated IPoIB 00/27] Enhanced mode for IPoIB driver
Date: Thu, 2 Mar 2017 21:17:16 +0200 [thread overview]
Message-ID: <CAAk-MO-YvUZq8Y-CiU6+O8CTYOV1cuodSmTc1XMzrPhWmWFS-Q@mail.gmail.com> (raw)
In-Reply-To: <20170301182833.GD14791-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
On Wed, Mar 1, 2017 at 8:28 PM, Jason Gunthorpe
<jgunthorpe-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org> wrote:
> On Wed, Mar 01, 2017 at 04:02:08PM +0200, Erez Shitrit wrote:
>> However, New features are exisiting in common network interfaces that
>> are very hard to implement in IPoIB interfaces while it uses the RDMA
>> layer, examples include TSS and RSS, tunneling offloads, and XDP.
>> Rather than continuously porting IP network interface developments into
>> the RDMA stack, we propose adding an abstract network data-path interfaces
>> to RDMA devices.
>
> I should also say that we have discussed this for a while and I am
> broadly in favor of this approach of delegating IPoIB skb transmit
> directly to the low level driver.
>
> There are too many optimizations that do not fit sanely into verbs,
> and the entire overhead of building a verbs wr and posting it then
> translating that to a driver wr is unnecessary layering.
>
> Longer term, it might be quite nice if the part verbs specific part of
> ipoib was cleanly split out into the 'default' implementation of your
> get_ipoib_accel_ops.
That what we tried to do here (hopefully..) we tried to keep the
separating of the control from data-path and verbs.
The accelerating functions are only for data-path or HW resources handling.
>
> Jason
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2017-03-02 19:17 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-03-01 14:02 [RFC for accelerated IPoIB 00/27] Enhanced mode for IPoIB driver Erez Shitrit
[not found] ` <1488376954-8346-1-git-send-email-erezsh-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
2017-03-01 14:02 ` [RFC for accelerated IPoIB 01/26] IB/ipoib: Separate control and data related initializations Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 02/26] IB/ipoib: separate control from HW operation on ipoib_open/stop ndo Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 03/26] IB/ipoib: Rename qpn to dqpn in ipoib_send and post_send functions Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 04/26] IB/verb: Add ipoib_options struct and API Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 05/26] IB/ipoib: Support ipoib acceleration options callbacks Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 06/26] IB/ipoib: Add context to ipoib to be used in acceleration layer Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 07/26] hw/mlx5: Add New bit to check over QP creation Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 08/26] linux/mlx5/mlx5_ifc.h: Add underlay_qpn field to PRM objects Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 09/26] net/mlx5e: Refactor EN code to support IB link Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 10/26] net/mlx5e: Creating and Destroying flow-steering tables for " Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 11/26] net/mlx5e: Support netdevice creation for IB link type Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 12/26] net/mlx5e: Refactor attach_netdev API Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 13/26] net/mlx5e: Use underlay_qpn in tis creation Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 14/26] net/mlx5e: Export resource creation function to be used in IB link Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 15/26] net/mlx5: Enable flow-steering for " Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 16/26] net/mlx5e: Enhanced flow table creation to support ETH and IB links Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 17/26] net/mlx5e: Change cleanup API in order to enable IB link Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 18/26] net/mlx5e: Change mlx5e_open_locked and mlx5e_close_locked api Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 19/26] net/mlx5e: Export open/close api for IB link Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 20/26] include/linux/mlx5: Add mlx5_wqe_eth_pad and enhanced-ipoib-qp-mode Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 21/26] net/mlx5e: Refactor TX send flow Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 22/26] net/mlx5e: Export send function for IB link type Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 23/26] net/mlx5e: New function pointer for build_rx_skb is Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 24/26] net/mlx5e: Change the function that checks the packet type Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 25/26] net/mlx5e: Add support for build_rx_skb for packet from IB type Erez Shitrit
2017-03-01 14:02 ` [RFC for accelerated IPoIB 26/26] mlx5_ib: skeleton for mlx5_ib to support ipoib_ops Erez Shitrit
2017-03-01 18:20 ` [RFC for accelerated IPoIB 00/27] Enhanced mode for IPoIB driver Jason Gunthorpe
[not found] ` <20170301182039.GC14791-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2017-03-02 19:13 ` Erez Shitrit
[not found] ` <CAAk-MO9tAHioaSXv8MPu=Kf3QSxjuQhAt6vYRVCzuNriXkm-+A-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2017-03-02 20:06 ` Jason Gunthorpe
[not found] ` <20170302200619.GA17530-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2017-03-03 0:14 ` Vishwanathapura, Niranjana
2017-03-01 18:28 ` Jason Gunthorpe
[not found] ` <20170301182833.GD14791-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
2017-03-02 19:17 ` Erez Shitrit [this message]
2017-03-02 20:30 ` ira.weiny
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=CAAk-MO-YvUZq8Y-CiU6+O8CTYOV1cuodSmTc1XMzrPhWmWFS-Q@mail.gmail.com \
--to=erezsh-ldsdmyg8hgv8yrgs2mwiifqbs+8scbdb@public.gmane.org \
--cc=dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=erezsh-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org \
--cc=jgunthorpe-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org \
--cc=leonro-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org \
--cc=linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=valex-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.