linux-kernel - Re: [PATCH 2.6.33 1/3] net: Micrel KSZ8841/2 PCI Ethernet driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <4B51E64B.6010706@openwrt.org>
Date:	Sat, 16 Jan 2010 17:16:11 +0100
From:	Felix Fietkau <nbd@...nwrt.org>
To:	Michał Mirosław <mirqus@...il.com>
CC:	"Ha, Tristram" <Tristram.Ha@...rel.com>,
	Dave Miller <davem@...emloft.net>, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org, Michael Buesch <mb@...sch.de>
Subject: Re: [PATCH 2.6.33 1/3] net: Micrel KSZ8841/2 PCI Ethernet driver

On 2010-01-16 5:00 PM, Michał Mirosław wrote:
> Hello,
> 
> 2010/1/16 Ha, Tristram <Tristram.Ha@...rel.com>:
>> This is a new network driver for Micrel KSZ8841/KSZ8842 PCI Ethernet chips.  The same driver can run both chips at the same time.  It supports IPv4 TCP hardware checksumming and so can use scatter/gather transmission.
>>
>> The KSZ8842 has a switch with lots of hardware configurations.  The driver uses the proc system to allow users to configure the switch.  If this is not desired the whole thing can be removed by not calling the init_proc() function.
>>
>> The KSZ8842 switch has 2 ports.  Some users like to take direct control of those ports.  So KSZ8842 has a multiple devices mode in which the driver creates another network device so that users can specify which port to send packets.  This mode is enabled by passing the "multi_dev=1" parameter to the driver during loading.
> [...]
> 
> I briefly looked over the datasheet for your device and it looks like
> it has the same basic design idea as a "ROBO" switch from Broadcom
> that is used in bcm53xx SoCs. This can be seen as a 3-port switch with
> one port fixed to be link to CPU (a special kind of PHY).
> 
> If my assumptions are correct, then I thing it would be better to have
> a driver that presents this device as two: one - an ethernet device,
> and second - a switch management interface like the one used in
> OpenWRT.
> 
> Here are some links for reference:
> http://downloads.openwrt.org/kamikaze/8.09.2/kamikaze_8.09.2_source.tar.bz2;
> subdir: package/switch/src/
>   -> switch management driver for ADMTEK Adm6996 and Broadcom BCM5325E/536x
Don't look at that ;)
The code you're referring to is just a simple hack of mine that is a
few years old now. It's going to be removed as soon as the drivers have
been ported to a proper API.

I've been working on a switch configuration API based on netlink, which 
is being used on some of the other platforms in OpenWrt.

You can find the sources here:
https://dev.openwrt.org/export/19173/trunk/target/linux/generic-2.6/files/drivers/net/phy/swconfig.c
https://dev.openwrt.org/export/19173/trunk/target/linux/generic-2.6/files/include/linux/switch.h

It's meant to be used with drivers that hook into the PHY abstraction layer.
You can find an example of such a driver here:
https://dev.openwrt.org/export/19173/trunk/target/linux/generic-2.6/files/drivers/net/phy/rtl8306.c

I'll do the final cleanups and submit it for review once I have time to do so.

- Felix
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/