Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Wed, 21 Aug 2019 02:01:55 +0200
From:   Andrew Lunn <andrew@lunn.ch>
To:     Sabrina Dubroca <sd@queasysnail.net>
Cc:     Antoine Tenart <antoine.tenart@bootlin.com>,
        Igor Russkikh <Igor.Russkikh@aquantia.com>,
        "davem@davemloft.net" <davem@davemloft.net>,
        "f.fainelli@gmail.com" <f.fainelli@gmail.com>,
        "hkallweit1@gmail.com" <hkallweit1@gmail.com>,
        "netdev@vger.kernel.org" <netdev@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "thomas.petazzoni@bootlin.com" <thomas.petazzoni@bootlin.com>,
        "alexandre.belloni@bootlin.com" <alexandre.belloni@bootlin.com>,
        "allan.nielsen@microchip.com" <allan.nielsen@microchip.com>,
        "camelia.groza@nxp.com" <camelia.groza@nxp.com>,
        Simon Edelhaus <Simon.Edelhaus@aquantia.com>,
        Pavel Belous <Pavel.Belous@aquantia.com>
Subject: Re: [PATCH net-next v2 6/9] net: macsec: hardware offloading
 infrastructure
Message-ID: <20190821000155.GA4285@lunn.ch>
References: <20190808140600.21477-1-antoine.tenart@bootlin.com>
 <20190808140600.21477-7-antoine.tenart@bootlin.com>
 <e96fa4ae-1f2c-c1be-b2d8-060217d8e151@aquantia.com>
 <20190813085817.GA3200@kwain>
 <20190813131706.GE15047@lunn.ch>
 <2e3c2307-d414-a531-26cb-064e05fa01fc@aquantia.com>
 <20190816132959.GC8697@bistromath.localdomain>
 <20190820100140.GA3292@kwain>
 <20190820144119.GA28714@bistromath.localdomain>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20190820144119.GA28714@bistromath.localdomain>
User-Agent: Mutt/1.5.23 (2014-03-12)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

> If you look at IPsec offloading, the networking stack builds up the
> ESP header, and passes the unencrypted data down to the driver. I'm
> wondering if the same would be possible with MACsec offloading: the
> macsec virtual interface adds the header (and maybe a dummy ICV), and
> then the HW does the encryption. In case of HW that needs to add the
> sectag itself, the driver would first strip the headers that the stack
> created. On receive, the driver would recreate a sectag and the macsec
> interface would just skip all verification (decrypt, PN).

Hi Sabrina

I assume the software implementation cannot make use of TSO or GSO,
letting the hardware segment a big buffer up into Ethernet frames?
When using hardware MACSEC, is it possible to enable these? By the
time the frames have reach the PHY GSO has been done. So it sees a
stream of frames it needs to encode/decode.

But if you are suggesting the extra headers are added by the virtual
interface, i don't think GSO can be used? My guess would be, we get a
performance boost from using hardware MAC sec, but there will also be
a performance boost if GSO can be enabled when it was disabled before.

      Andrew