netdev
[Top] [All Lists]

Re: PATCH: sundance #2

To: Jeff Garzik <jgarzik@xxxxxxxxxxxxxxxx>
Subject: Re: PATCH: sundance #2
From: Donald Becker <becker@xxxxxxxxx>
Date: Thu, 19 Sep 2002 13:29:31 -0400 (EDT)
Cc: netdev@xxxxxxxxxxx, Jason Lunz <lunz@xxxxxxxxxxxx>, Richard Gooch <rgooch@xxxxxxxxxxxxxxx>, "Patrick R. McManus" <mcmanus@xxxxxxxxxxxx>, Linux Kernel Mailing List <linux-kernel@xxxxxxxxxxxxxxx>, <edward_peng@xxxxxxxxxxxx>
In-reply-to: <3D8A056C.2000605@mandrakesoft.com>
Sender: netdev-bounce@xxxxxxxxxxx
On Thu, 19 Sep 2002, Jeff Garzik wrote:

> Thanks to Donald for his comments.  This patch addresses the first of 
> his two emails.
> 
> This patch is _cumulative_ with the last one I sent (sundance 1.04), so 
> do not discard that one.
> 
> Again additional testing is appreciated.  Keep the feedback coming, 
> there will be more sundance bugfixes (patch #3, #4, etc.)

+       - If no phy is found, fail to load that board
+       - Always start phy id scan at id 1 to avoid problems (Donald Becker)
+       - Autodetect where mii_preable_required is needed,
+       default to not needed.  (Donald Becker)
...
+
+/* Set iff a MII transceiver on any interface requires mdio preamble.
+   This only set with older tranceivers, so the extra
+   code size of a per-interface flag is not worthwhile. */
+static int mii_preamble_required = 0;

You can get rid of this as a module option, and make it a per-interface
setting. 
The transceiver on the Kendin chip requires this (rather old-fashioned)
access method, while none of the previous Sundance-based boards with
external transceivers did.

I added it as a module parameter as a back-up over-ride, but I'm certain
that the automatic detection works.

This is a module parameter because I recently had a bad experience
with a specific 3Com 3c905B chip rev.  It claimed to not need
transceiver preamble, but would not work without it.

>                               Theory of Operation

Whoever changed the transmit path should update the TOO.  

-       {"Sundance Technology Alta", {0x020113F0, 0xffffffff,},
-        PCI_IOTYPE, 128, CanHaveMII},
+       {"D-Link DFE-550TX FAST Ethernet Adapter"},
+       {"D-Link DFE-550FX 100Mbps Fiber-optics Adapter"},

Yeah, you should probably throw away the rest of the changes.
You are probably going to want to keep the drv_flags field.  I know
that all of the current chips have the same flag (CanHaveMII), but...



-- 
Donald Becker                           becker@xxxxxxxxx
Scyld Computing Corporation             http://www.scyld.com
410 Severn Ave. Suite 210               Second Generation Beowulf Clusters
Annapolis MD 21403                      410-990-9993


<Prev in Thread] Current Thread [Next in Thread>