[U-Boot] [PATCH 2/2] usb: eth: smsc95xx: Add EEPROM access support

[Marek Vasut]

On Tuesday, October 07, 2014 at 03:35:40 PM, Alban Bedel wrote:
> On Tue, 7 Oct 2014 14:22:09 +0200
> 
> Pavel Machek <pavel at denx.de> wrote:
> > On Tue 2014-10-07 11:19:37, Alban Bedel wrote:
> > > Use the new ethernet eeprom API to allow the user to read/write the
> > > EEPROM.
> > > 
> > > Change-Id: I21233b6ee805a75bd8a03ca12e22c41421b7629c
> > > Signed-off-by: Alban Bedel <alban.bedel at avionic-design.de>
> > > ---
> > > 
> > >  drivers/usb/eth/smsc95xx.c | 199
> > >  +++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 192
> > >  insertions(+), 7 deletions(-)
> > > 
> > > diff --git a/drivers/usb/eth/smsc95xx.c b/drivers/usb/eth/smsc95xx.c
> > > index 6bca34d..eb29565 100644
> > > --- a/drivers/usb/eth/smsc95xx.c
> > > +++ b/drivers/usb/eth/smsc95xx.c
> > > @@ -59,6 +59,8 @@
> > > 
> > >  #define E2P_CMD				0x30
> > >  #define E2P_CMD_BUSY_			0x80000000
> > > 
> > > +#define E2P_CMD_EWEN_			0x20000000
> > > +#define E2P_CMD_WRITE_			0x30000000
> > > 
> > >  #define E2P_CMD_READ_			0x00000000
> > >  #define E2P_CMD_TIMEOUT_		0x00000400
> > >  #define E2P_CMD_LOADED_			0x00000200
> > > 
> > > @@ -146,6 +148,131 @@ struct smsc95xx_private {
> > > 
> > >  	int have_hwaddr;  /* 1 if we have a hardware MAC address */
> > >  
> > >  };
> > > 
> > > +#ifdef CONFIG_CMD_ETH_EEPROM
> > 
> > Is this layout common for all machines using this driver? If not, is
> > it worth comment which machine is it?
> 
> These are supposed to be the same values as would be used by the
> controller when the eeprom is not programmed. So they are not machine
> specific, however it is specific to the LAN9514, looking at the
> datasheet for the LAN9500 the eeprom layout is different.
> 
> Then I'll submit a new version that use per chip defaults, as well
> as allowing the board file to override the default data.
> 
> > > +static u8 eeprom_defaults[] = {
> > > +	/* 0x00 */
> > > +	0xA5,		/* Signature */
> > > +	0xFF, 0xFF,	/* MAC bytes 0-1 */
> > > +	0xFF, 0xFF,	/* MAC bytes 2-3 */
> > > +	0xFF, 0xFF,	/* MAC bytes 4-5 */
> > 
> > Normally, we use all zeros for unset...?
> 
> That's what the controller use when the EEPROM hasn't been programmed,
> but I'll change it to all 0.

Well isn't it better to use these settings as defaults then ?

Best regards,
Marek Vasut