[U-Boot] [PATCH v2] Program net device MAC addresses after initializing

[U-Boot] [PATCH v2] Program net device MAC addresses after initializing

[Ben Warren]

Add a new function to the eth_device struct for programming a network
controller's hardware address.

After all network devices have been initialized and the proper MAC address for
each has been determined, make a device driver call to program the address
into the device.  Only device instances with valid unicast addresses will be
programmed.

Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
Acked-by: Heiko Schocher <hs@denx.de>
Acked-by: Prafulla Wadaskar <prafulla@marvell.com>
Acked-by: Detlev Zundel <dzu@denx.de>
---

v1->v2 Add documentation & environment variable for overriding new behavior
ACKs mentioned above are for v1

 README                 |    5 +++++
 doc/README.drivers.eth |    7 ++++++-
 doc/README.enetaddr    |   10 ++++++----
 include/net.h          |    1 +
 net/eth.c              |   13 +++++++++++++
 5 files changed, 31 insertions(+), 5 deletions(-)

diff --git a/README b/README
index cdd81d4..2f92bd2 100644
--- a/README
+++ b/README
@@ -3303,6 +3303,11 @@ o If both the SROM and the environment contain a MAC address, and the
 o If neither SROM nor the environment contain a MAC address, an error
   is raised.
 
+If Ethernet drivers implement the 'write_hwaddr' function, valid MAC addresses
+will be programmed into hardware as part of the initialization process.  This
+may be skipped by setting the appropriate 'ethmacskip' environment variable.
+The naming convention is as follows:
+"eth0macskip" (=>eth0), "eth1macskip" (=>eth1) etc.
 
 Image Formats:
 ==============
diff --git a/doc/README.drivers.eth b/doc/README.drivers.eth
index d0c3571..eb83038 100644
--- a/doc/README.drivers.eth
+++ b/doc/README.drivers.eth
@@ -70,6 +70,7 @@ int ape_register(bd_t *bis, int iobase)
 	dev->halt = ape_halt;
 	dev->send = ape_send;
 	dev->recv = ape_recv;
+	dev->write_hwaddr = ape_write_hwaddr;
 
 	eth_register(dev);
 
@@ -102,11 +103,12 @@ not checking its state or doing random probing.
  -----------
 
 Now that we've registered with the ethernet layer, we can start getting some
-real work done.  You will need four functions:
+real work done.  You will need five functions:
 	int ape_init(struct eth_device *dev, bd_t *bis);
 	int ape_send(struct eth_device *dev, volatile void *packet, int length);
 	int ape_recv(struct eth_device *dev);
 	int ape_halt(struct eth_device *dev);
+	int ape_write_hwaddr(struct eth_device *dev);
 
 The init function checks the hardware (probing/identifying) and gets it ready
 for send/recv operations.  You often do things here such as resetting the MAC
@@ -150,6 +152,9 @@ The halt function should turn off / disable the hardware and place it back in
 its reset state.  It can be called at any time (before any call to the related
 init function), so make sure it can handle this sort of thing.
 
+The write_hwaddr function should program the MAC address stored in dev->enetaddr
+into the Ethernet controller.
+
 So the call graph at this stage would look something like:
 some net operation (ping / tftp / whatever...)
 	eth_init()
diff --git a/doc/README.enetaddr b/doc/README.enetaddr
index 94d800a..53a0c25 100644
--- a/doc/README.enetaddr
+++ b/doc/README.enetaddr
@@ -33,11 +33,13 @@ Correct flow of setting up the MAC address (summarized):
 1. Read from hardware in initialize() function
 2. Read from environment in net/eth.c after initialize()
 3. Give priority to the value in the environment if a conflict
-4. Program hardware in the device's init() function.
+4. Program the address into hardware if the following conditions are met:
+	a) The relevant driver has a 'write_addr' function
+	b) The user hasn't set an 'eth%dmacskip' environment varialbe
+	c) The address is valid (unicast, not all-zeros)
 
-If somebody wants to subvert the design philosophy, this can be done
-in the board-specific board_eth_init() function by calling eth_init()
-after all the NICs have been registered.
+Previous behavior had the MAC address always being programmed into hardware
+in the device's init() function.
 
 -------
  Usage
diff --git a/include/net.h b/include/net.h
index 3f6a5d1..a180881 100644
--- a/include/net.h
+++ b/include/net.h
@@ -105,6 +105,7 @@ struct eth_device {
 #ifdef CONFIG_MCAST_TFTP
 	int (*mcast) (struct eth_device*, u32 ip, u8 set);
 #endif
+	int  (*write_hwaddr) (struct eth_device*);
 	struct eth_device *next;
 	void *priv;
 };
diff --git a/net/eth.c b/net/eth.c
index aff6987..7364699 100644
--- a/net/eth.c
+++ b/net/eth.c
@@ -60,6 +60,14 @@ int eth_getenv_enetaddr_by_index(int index, uchar *enetaddr)
 	return eth_getenv_enetaddr(enetvar, enetaddr);
 }
 
+static int eth_mac_skip(int index)
+{
+	char enetvar[15];
+	char *skip_state;
+	sprintf(enetvar, "eth%dmacskip");
+	return ((skip_state = getenv(enetvar)) != NULL);
+}
+ 
 #ifdef CONFIG_NET_MULTI
 
 /*
@@ -242,6 +250,11 @@ int eth_initialize(bd_t *bis)
 
 				memcpy(dev->enetaddr, env_enetaddr, 6);
 			}
+			if (dev->write_hwaddr &&
+				!eth_mac_skip(eth_number) &&
+				is_valid_ether_addr(dev->enetaddr)) {
+				dev->write_hwaddr(dev);
+			}
 
 			eth_number++;
 			dev = dev->next;
-- 
1.6.0.4

[U-Boot] [PATCH v2] Program net device MAC addresses after initializing

[Wolfgang Denk]

Dear Ben Warren,

In message <1272261030-9419-1-git-send-email-biggerbadderben@gmail.com> you wrote:
> Add a new function to the eth_device struct for programming a network
> controller's hardware address.
> 
> After all network devices have been initialized and the proper MAC address for
> each has been determined, make a device driver call to program the address
> into the device.  Only device instances with valid unicast addresses will be
> programmed.

Thanks.

> +If Ethernet drivers implement the 'write_hwaddr' function, valid MAC addresses
> +will be programmed into hardware as part of the initialization process.  This
> +may be skipped by setting the appropriate 'ethmacskip' environment variable.
> +The naming convention is as follows:
> +"eth0macskip" (=>eth0), "eth1macskip" (=>eth1) etc.

Would it not be more consistent to use "ethmacskip" instead of
"eth0macskip" similar to using "ethaddr" (instead of "eth0addr") ?

The majority of boards still has only a single Ethernet interface
anyway.

Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
F u cn rd ths u cnt spl wrth a dm!

[U-Boot] [PATCH v2] Program net device MAC addresses after initializing

[Prafulla Wadaskar]

 

> -----Original Message-----
> From: u-boot-bounces at lists.denx.de 
> [mailto:u-boot-bounces at lists.denx.de] On Behalf Of Wolfgang Denk
> Sent: Monday, April 26, 2010 5:22 PM
> To: Ben Warren
> Cc: u-boot at lists.denx.de
> Subject: Re: [U-Boot] [PATCH v2] Program net device MAC 
> addresses after initializing
> 
> Dear Ben Warren,
...snip...
> > +If Ethernet drivers implement the 'write_hwaddr' function, 
> valid MAC addresses
> > +will be programmed into hardware as part of the 
> initialization process.  This
> > +may be skipped by setting the appropriate 'ethmacskip' 
> environment variable.
> > +The naming convention is as follows:
> > +"eth0macskip" (=>eth0), "eth1macskip" (=>eth1) etc.
> 
> Would it not be more consistent to use "ethmacskip" instead of
> "eth0macskip" similar to using "ethaddr" (instead of "eth0addr") ?
> 
> The majority of boards still has only a single Ethernet interface
> anyway.

Kirkwood has two egiga controllers and some of boards are using both, like OpenRD, GuruPlug.
Guruplug has two ports- eth0 and eth1, and setting eth1macskip I can only program eth0 mac address.

Regards..
Prafulla . .

[U-Boot] [PATCH v2] Program net device MAC addresses after initializing

[Ben Warren]

Hi Wolfgang,

On 4/26/2010 4:52 AM, Wolfgang Denk wrote:
> Dear Ben Warren,
>
> In message<1272261030-9419-1-git-send-email-biggerbadderben@gmail.com>  you wrote:
>    
>> Add a new function to the eth_device struct for programming a network
>> controller's hardware address.
>>
>> After all network devices have been initialized and the proper MAC address for
>> each has been determined, make a device driver call to program the address
>> into the device.  Only device instances with valid unicast addresses will be
>> programmed.
>>      
> Thanks.
>
>    
>> +If Ethernet drivers implement the 'write_hwaddr' function, valid MAC addresses
>> +will be programmed into hardware as part of the initialization process.  This
>> +may be skipped by setting the appropriate 'ethmacskip' environment variable.
>> +The naming convention is as follows:
>> +"eth0macskip" (=>eth0), "eth1macskip" (=>eth1) etc.
>>      
> Would it not be more consistent to use "ethmacskip" instead of
> "eth0macskip" similar to using "ethaddr" (instead of "eth0addr") ?
>
>    
I guess it depends which consistency we're going for :)  We can be 
consistent relative to an existing variable that is itself inconsistent 
for historical reasons, or we can be consistent in usage of the new 
variable.  I don't have a strong opinion either way, so will change this.
> The majority of boards still has only a single Ethernet interface
> anyway.
>
> Best regards,
>
> Wolfgang Denk
>
>    
regards,
Ben