From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Warren Date: Tue, 08 Jan 2008 11:42:02 -0500 Subject: [U-Boot-Users] Possible TSEC Ethernet driver patch In-Reply-To: <36D7B34A3A79F84F82FA0C154F299F25065C2E0D@E03MVX1-UKDY.domain1.systemhost.net> References: <36D7B34A3A79F84F82FA0C154F299F25065C2E0D@E03MVX1-UKDY.domain1.systemhost.net> Message-ID: <4783A7DA.10904@gmail.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de michael.firth at bt.com wrote: > While debugging a board recently I found that the MDIO (mii) command > support in the TSEC Ethernet driver is somewhat unhelpful. > > Currently, even though there is a comment in the code that "For now, > only TSEC1 (index 0) has access to the PHYs, so all of the entries have > '0'", all MDIO commands are processed by searching for a TSEC instance > that has the requested MDIO address associated with it, and then using > that instance to run the command, even though, because of the > aforementioned comment, all instances process MDIO commands through the > same port. > > This means that it is only possible to communicate with MDIO addresses > that have a TSEC instance associated with them, even though the hardware > is capable of accessing any address. It also means that there is a list > search that isn't needed. > > I have patched the 1.3.1 U-Boot code to remove this search, and to > interrogate the requested PHY directly. This means that it is possible > to directly access all 32 PHY addresses. > > Is this a change that would be more generally useful to the U-Boot > community, and, if so, how should I submit a more general patch for > this? > > Why don't you post what you have, clearly label it as 'RFC' and we'll have a look. In my spare time (very spare indeed) I'm trying to decouple PHYs from MACs, but time is hard to find and meanwhile things need to work. regards, Ben