[U-Boot] [PATCH 0/3] enable CONFIG_NET_MULTI for LAN91C96

[U-Boot] [PATCH 0/3] enable CONFIG_NET_MULTI for LAN91C96

[Nishanth Menon]

Based on the discussion[1] here is the patchset

Patchset is based on [2], and tested on SDP3430.
Other platforms would be prefered to be
build/tested/fixed as required

Nishanth Menon (3):
  NET: LAN91C96 CONFIG_NET_MULTIify
  TI OMAP3: SDP3430 FIX NET_MULTI Warning
  LAN91C96: Enable NET_MULTI LAN driver

 board/apollon/apollon.c         |   12 +-
 board/ti/sdp3430/sdp.c          |   12 +-
 drivers/net/Makefile            |    2 +-
 drivers/net/lan91c96.c          |  452 ++++++++++++++++-----------------------
 drivers/net/lan91c96.h          |  110 +++++-----
 include/configs/B2.h            |    2 +-
 include/configs/apollon.h       |    3 +-
 include/configs/assabet.h       |    3 +-
 include/configs/gcplus.h        |    3 +-
 include/configs/lubbock.h       |    3 +-
 include/configs/omap1510inn.h   |    3 +-
 include/configs/omap1610h2.h    |    3 +-
 include/configs/omap1610inn.h   |    3 +-
 include/configs/omap2420h4.h    |    3 +-
 include/configs/omap3_sdp3430.h |    3 +-
 include/configs/omap5912osk.h   |    3 +-
 include/configs/omap730p2.h     |    3 +-
 include/configs/pleb2.h         |    6 +-
 include/netdev.h                |    1 +
 19 files changed, 277 insertions(+), 353 deletions(-)

Regards,
Nishanth Menon

Ref:
[1] http://www.nabble.com/-U-Boot--NET%3A-SDP3430%3A-trouble-with-shifting-from-LAN9C916-to-SMC91XX-driver-td25884700.html
[2] http://git.denx.de/cgi-bin/gitweb.cgi?p=u-boot/u-boot-ti.git

[U-Boot] [PATCH 1/3] NET: LAN91C96 CONFIG_NET_MULTIify

[Nishanth Menon]

Make the lan91c96 driver capable of CONFIG_NET_MULTI
to be clean for the new arch, add a a lil detect function
Most of the formatting change was done to keep checkpatch
silent, but a few functions and #if 0ed code which
does not make sense for NET_MULTI have been removed

Now, use the lan91c96_initialize() function to init the driver

Signed-off-by: Nishanth Menon <nm@ti.com>
Cc: Andrea Scian <andrea.scian@dave-tech.it>
Cc: Ben Warren <biggerbadderben@gmail.com>
Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
Cc: George G. Davis <gdavis@mvista.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Nishant Kamat <nskamat@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Rishi Bhattacharya <rishi@ti.com>
Cc: Sandeep Paulraj <s-paulraj@ti.com>
---
 drivers/net/Makefile   |    2 +-
 drivers/net/lan91c96.c |  452 +++++++++++++++++++----------------------------
 drivers/net/lan91c96.h |  110 ++++++-------
 include/netdev.h       |    1 +
 4 files changed, 236 insertions(+), 329 deletions(-)

diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index be5c484..a4ab447 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -44,7 +44,7 @@ COBJS-$(CONFIG_GRETH) += greth.o
 COBJS-$(CONFIG_INCA_IP_SWITCH) += inca-ip_sw.o
 COBJS-$(CONFIG_KIRKWOOD_EGIGA) += kirkwood_egiga.o
 COBJS-$(CONFIG_DRIVER_KS8695ETH) += ks8695eth.o
-COBJS-$(CONFIG_DRIVER_LAN91C96) += lan91c96.o
+COBJS-$(CONFIG_LAN91C96) += lan91c96.o
 COBJS-$(CONFIG_MACB) += macb.o
 COBJS-$(CONFIG_MCFFEC) += mcffec.o mcfmii.o
 COBJS-$(CONFIG_MPC5xxx_FEC) += mpc5xxx_fec.o
diff --git a/drivers/net/lan91c96.c b/drivers/net/lan91c96.c
index 65565bc..90e4002 100644
--- a/drivers/net/lan91c96.c
+++ b/drivers/net/lan91c96.c
@@ -60,6 +60,7 @@
 
 #include <common.h>
 #include <command.h>
+#include <malloc.h>
 #include "lan91c96.h"
 #include <net.h>
 
@@ -108,11 +109,7 @@
  *
  *------------------------------------------------------------------------
  */
-#define CARDNAME "LAN91C96"
-
-#define SMC_BASE_ADDRESS CONFIG_LAN91C96_BASE
-
-#define SMC_DEV_NAME "LAN91C96"
+#define DRIVER_NAME "LAN91C96"
 #define SMC_ALLOC_MAX_TRY 5
 #define SMC_TX_TIMEOUT 30
 
@@ -124,64 +121,12 @@
 #undef USE_32_BIT
 #endif
 
-/*-----------------------------------------------------------------
- *
- *  The driver can be entered at any of the following entry points.
- *
- *-----------------------------------------------------------------
- */
-
-extern int eth_init (bd_t * bd);
-extern void eth_halt (void);
-extern int eth_rx (void);
-extern int eth_send (volatile void *packet, int length);
-#if 0
-static int smc_hw_init (void);
-#endif
-
-/*
- * This is called by  register_netdev().  It is responsible for
- * checking the portlist for the SMC9000 series chipset.  If it finds
- * one, then it will initialize the device, find the hardware information,
- * and sets up the appropriate device parameters.
- * NOTE: Interrupts are *OFF* when this procedure is called.
- *
- * NB:This shouldn't be static since it is referred to externally.
- */
-int smc_init (void);
-
-/*
- * This is called by  unregister_netdev().  It is responsible for
- * cleaning up before the driver is finally unregistered and discarded.
- */
-void smc_destructor (void);
-
-/*
- * The kernel calls this function when someone wants to use the device,
- * typically 'ifconfig ethX up'.
- */
-static int smc_open (bd_t *bd);
-
-
-/*
- * This is called by the kernel in response to 'ifconfig ethX down'.  It
- * is responsible for cleaning up everything that the open routine
- * does, and maybe putting the card into a powerdown state.
- */
-static int smc_close (void);
-
-/*
- * This is a separate procedure to handle the receipt of a packet, to
- * leave the interrupt code looking slightly cleaner
- */
-static int smc_rcv (void);
-
 /* See if a MAC address is defined in the current environment. If so use it. If not
  . print a warning and set the environment and other globals with the default.
  . If an EEPROM is present it really should be consulted.
 */
-int smc_get_ethaddr(bd_t *bd);
-int get_rom_mac(unsigned char *v_rom_mac);
+static int smc_get_ethaddr(bd_t *bd, struct eth_device *dev);
+static int get_rom_mac(struct eth_device *dev, unsigned char *v_rom_mac);
 
 /* ------------------------------------------------------------
  * Internal routines
@@ -195,7 +140,7 @@ static unsigned char smc_mac_addr[] = { 0xc0, 0x00, 0x00, 0x1b, 0x62, 0x9c };
  * the default mac address.
  */
 
-void smc_set_mac_addr (const unsigned char *addr)
+static void smc_set_mac_addr(const unsigned char *addr)
 {
 	int i;
 
@@ -204,45 +149,21 @@ void smc_set_mac_addr (const unsigned char *addr)
 	}
 }
 
-/*
- * smc_get_macaddr is no longer used. If you want to override the default
- * mac address, call smc_get_mac_addr as a part of the board initialisation.
- */
-
-#if 0
-void smc_get_macaddr (byte * addr)
-{
-	/* MAC ADDRESS AT FLASHBLOCK 1 / OFFSET 0x10 */
-	unsigned char *dnp1110_mac = (unsigned char *) (0xE8000000 + 0x20010);
-	int i;
-
-
-	for (i = 0; i < 6; i++) {
-		addr[0] = *(dnp1110_mac + 0);
-		addr[1] = *(dnp1110_mac + 1);
-		addr[2] = *(dnp1110_mac + 2);
-		addr[3] = *(dnp1110_mac + 3);
-		addr[4] = *(dnp1110_mac + 4);
-		addr[5] = *(dnp1110_mac + 5);
-	}
-}
-#endif /* 0 */
-
 /***********************************************
  * Show available memory                       *
  ***********************************************/
-void dump_memory_info (void)
+void dump_memory_info(struct eth_device *dev)
 {
 	word mem_info;
 	word old_bank;
 
-	old_bank = SMC_inw (LAN91C96_BANK_SELECT) & 0xF;
+	old_bank = SMC_inw(dev, LAN91C96_BANK_SELECT) & 0xF;
 
-	SMC_SELECT_BANK (0);
-	mem_info = SMC_inw (LAN91C96_MIR);
+	SMC_SELECT_BANK(dev, 0);
+	mem_info = SMC_inw(dev, LAN91C96_MIR);
 	PRINTK2 ("Memory: %4d available\n", (mem_info >> 8) * 2048);
 
-	SMC_SELECT_BANK (old_bank);
+	SMC_SELECT_BANK(dev, old_bank);
 }
 
 /*
@@ -252,28 +173,15 @@ void dump_memory_info (void)
 static void print_packet (byte *, int);
 #endif
 
-/* #define tx_done(dev) 1 */
-
-
-/* this does a soft reset on the device */
-static void smc_reset (void);
-
-/* Enable Interrupts, Receive, and Transmit */
-static void smc_enable (void);
-
-/* this puts the device in an inactive state */
-static void smc_shutdown (void);
-
-
-static int poll4int (byte mask, int timeout)
+static int poll4int (struct eth_device *dev, byte mask, int timeout)
 {
 	int tmo = get_timer (0) + timeout * CONFIG_SYS_HZ;
 	int is_timeout = 0;
-	word old_bank = SMC_inw (LAN91C96_BANK_SELECT);
+	word old_bank = SMC_inw(dev, LAN91C96_BANK_SELECT);
 
 	PRINTK2 ("Polling...\n");
-	SMC_SELECT_BANK (2);
-	while ((SMC_inw (LAN91C96_INT_STATS) & mask) == 0) {
+	SMC_SELECT_BANK(dev, 2);
+	while ((SMC_inw(dev, LAN91C96_INT_STATS) & mask) == 0) {
 		if (get_timer (0) >= tmo) {
 			is_timeout = 1;
 			break;
@@ -281,7 +189,7 @@ static int poll4int (byte mask, int timeout)
 	}
 
 	/* restore old bank selection */
-	SMC_SELECT_BANK (old_bank);
+	SMC_SELECT_BANK(dev, old_bank);
 
 	if (is_timeout)
 		return 1;
@@ -290,7 +198,7 @@ static int poll4int (byte mask, int timeout)
 }
 
 /*
- * Function: smc_reset( void )
+ * Function: smc_reset
  * Purpose:
  *	This sets the SMC91111 chip to its normal state, hopefully from whatever
  *	mess that any other DOS driver has put it in.
@@ -306,28 +214,28 @@ static int poll4int (byte mask, int timeout)
  *	5.  clear all interrupts
  *
 */
-static void smc_reset (void)
+static void smc_reset(struct eth_device *dev)
 {
-	PRINTK2 ("%s:smc_reset\n", SMC_DEV_NAME);
+	PRINTK2("%s:smc_reset\n", dev->name);
 
 	/* This resets the registers mostly to defaults, but doesn't
 	   affect EEPROM.  That seems unnecessary */
-	SMC_SELECT_BANK (0);
-	SMC_outw (LAN91C96_RCR_SOFT_RST, LAN91C96_RCR);
+	SMC_SELECT_BANK(dev, 0);
+	SMC_outw(dev, LAN91C96_RCR_SOFT_RST, LAN91C96_RCR);
 
 	udelay (10);
 
 	/* Disable transmit and receive functionality */
-	SMC_outw (0, LAN91C96_RCR);
-	SMC_outw (0, LAN91C96_TCR);
+	SMC_outw(dev, 0, LAN91C96_RCR);
+	SMC_outw(dev, 0, LAN91C96_TCR);
 
 	/* set the control register */
-	SMC_SELECT_BANK (1);
-	SMC_outw (SMC_inw (LAN91C96_CONTROL) | LAN91C96_CTR_BIT_8,
+	SMC_SELECT_BANK(dev, 1);
+	SMC_outw(dev, SMC_inw(dev, LAN91C96_CONTROL) | LAN91C96_CTR_BIT_8,
 			  LAN91C96_CONTROL);
 
 	/* Disable all interrupts */
-	SMC_outb (0, LAN91C96_INT_MASK);
+	SMC_outb(dev, 0, LAN91C96_INT_MASK);
 }
 
 /*
@@ -338,24 +246,24 @@ static void smc_reset (void)
  *	2.  Enable the transmitter
  *	3.  Enable the receiver
 */
-static void smc_enable ()
+static void smc_enable(struct eth_device *dev)
 {
-	PRINTK2 ("%s:smc_enable\n", SMC_DEV_NAME);
-	SMC_SELECT_BANK (0);
+	PRINTK2("%s:smc_enable\n", dev->name);
+	SMC_SELECT_BANK(dev, 0);
 
 	/* Initialize the Memory Configuration Register. See page
 	   49 of the LAN91C96 data sheet for details. */
-	SMC_outw (LAN91C96_MCR_TRANSMIT_PAGES, LAN91C96_MCR);
+	SMC_outw(dev, LAN91C96_MCR_TRANSMIT_PAGES, LAN91C96_MCR);
 
 	/* Initialize the Transmit Control Register */
-	SMC_outw (LAN91C96_TCR_TXENA, LAN91C96_TCR);
+	SMC_outw(dev, LAN91C96_TCR_TXENA, LAN91C96_TCR);
 	/* Initialize the Receive Control Register
 	 * FIXME:
 	 * The promiscuous bit set because I could not receive ARP reply
 	 * packets from the server when I send a ARP request. It only works
 	 * when I set the promiscuous bit
 	 */
-	SMC_outw (LAN91C96_RCR_RXEN | LAN91C96_RCR_PRMS, LAN91C96_RCR);
+	SMC_outw(dev, LAN91C96_RCR_RXEN | LAN91C96_RCR_PRMS, LAN91C96_RCR);
 }
 
 /*
@@ -372,18 +280,18 @@ static void smc_enable ()
  *	the manual says that it will wake up in response to any I/O requests
  *	in the register space.   Empirical results do not show this working.
  */
-static void smc_shutdown ()
+static void smc_shutdown(struct eth_device *dev)
 {
-	PRINTK2 (CARDNAME ":smc_shutdown\n");
+	PRINTK2("%s:smc_shutdown\n", dev->name);
 
 	/* no more interrupts for me */
-	SMC_SELECT_BANK (2);
-	SMC_outb (0, LAN91C96_INT_MASK);
+	SMC_SELECT_BANK(dev, 2);
+	SMC_outb(dev, 0, LAN91C96_INT_MASK);
 
 	/* and tell the card to stay away from that nasty outside world */
-	SMC_SELECT_BANK (0);
-	SMC_outb (0, LAN91C96_RCR);
-	SMC_outb (0, LAN91C96_TCR);
+	SMC_SELECT_BANK(dev, 0);
+	SMC_outb(dev, 0, LAN91C96_RCR);
+	SMC_outb(dev, 0, LAN91C96_TCR);
 }
 
 
@@ -405,7 +313,8 @@ static void smc_shutdown ()
  *	Enable the transmit interrupt, so I know if it failed
  *	Free the kernel data if I actually sent it.
  */
-static int smc_send_packet (volatile void *packet, int packet_length)
+static int smc_send_packet(struct eth_device *dev, volatile void *packet,
+		int packet_length)
 {
 	byte packet_no;
 	unsigned long ioaddr;
@@ -417,7 +326,7 @@ static int smc_send_packet (volatile void *packet, int packet_length)
 	byte status;
 
 
-	PRINTK3 ("%s:smc_hardware_send_packet\n", SMC_DEV_NAME);
+	PRINTK3("%s:smc_hardware_send_packet\n", dev->name);
 
 	length = ETH_ZLEN < packet_length ? packet_length : ETH_ZLEN;
 
@@ -437,30 +346,31 @@ static int smc_send_packet (volatile void *packet, int packet_length)
 	numPages >>= 8;				/* Divide by 256 */
 
 	if (numPages > 7) {
-		printf ("%s: Far too big packet error. \n", SMC_DEV_NAME);
+		printf("%s: Far too big packet error. \n", dev->name);
 		return 0;
 	}
 
 	/* now, try to allocate the memory */
 
-	SMC_SELECT_BANK (2);
-	SMC_outw (LAN91C96_MMUCR_ALLOC_TX | numPages, LAN91C96_MMU);
+	SMC_SELECT_BANK(dev, 2);
+	SMC_outw(dev, LAN91C96_MMUCR_ALLOC_TX | numPages, LAN91C96_MMU);
 
   again:
 	try++;
 	time_out = MEMORY_WAIT_TIME;
 	do {
-		status = SMC_inb (LAN91C96_INT_STATS);
+		status = SMC_inb(dev, LAN91C96_INT_STATS);
 		if (status & LAN91C96_IST_ALLOC_INT) {
 
-			SMC_outb (LAN91C96_IST_ALLOC_INT, LAN91C96_INT_STATS);
+			SMC_outb(dev, LAN91C96_IST_ALLOC_INT,
+					LAN91C96_INT_STATS);
 			break;
 		}
 	} while (--time_out);
 
 	if (!time_out) {
 		PRINTK2 ("%s: memory allocation, try %d failed ...\n",
-				 SMC_DEV_NAME, try);
+				 dev->name, try);
 		if (try < SMC_ALLOC_MAX_TRY)
 			goto again;
 		else
@@ -468,30 +378,30 @@ static int smc_send_packet (volatile void *packet, int packet_length)
 	}
 
 	PRINTK2 ("%s: memory allocation, try %d succeeded ...\n",
-			 SMC_DEV_NAME, try);
+			 dev->name, try);
 
 	/* I can send the packet now.. */
 
-	ioaddr = SMC_BASE_ADDRESS;
+	ioaddr = dev->iobase;
 
 	buf = (byte *) packet;
 
 	/* If I get here, I _know_ there is a packet slot waiting for me */
-	packet_no = SMC_inb (LAN91C96_ARR);
+	packet_no = SMC_inb(dev, LAN91C96_ARR);
 	if (packet_no & LAN91C96_ARR_FAILED) {
 		/* or isn't there?  BAD CHIP! */
-		printf ("%s: Memory allocation failed. \n", SMC_DEV_NAME);
+		printf("%s: Memory allocation failed. \n", dev->name);
 		return 0;
 	}
 
 	/* we have a packet address, so tell the card to use it */
-	SMC_outb (packet_no, LAN91C96_PNR);
+	SMC_outb(dev, packet_no, LAN91C96_PNR);
 
 	/* point to the beginning of the packet */
-	SMC_outw (LAN91C96_PTR_AUTO_INCR, LAN91C96_POINTER);
+	SMC_outw(dev, LAN91C96_PTR_AUTO_INCR, LAN91C96_POINTER);
 
-	PRINTK3 ("%s: Trying to xmit packet of length %x\n",
-			 SMC_DEV_NAME, length);
+	PRINTK3("%s: Trying to xmit packet of length %x\n",
+			 dev->name, length);
 
 #if SMC_DEBUG > 2
 	printf ("Transmitting Packet\n");
@@ -501,11 +411,11 @@ static int smc_send_packet (volatile void *packet, int packet_length)
 	/* send the packet length ( +6 for status, length and ctl byte )
 	   and the status word ( set to zeros ) */
 #ifdef USE_32_BIT
-	SMC_outl ((length + 6) << 16, LAN91C96_DATA_HIGH);
+	SMC_outl(dev, (length + 6) << 16, LAN91C96_DATA_HIGH);
 #else
-	SMC_outw (0, LAN91C96_DATA_HIGH);
+	SMC_outw(dev, 0, LAN91C96_DATA_HIGH);
 	/* send the packet length ( +6 for status words, length, and ctl */
-	SMC_outw ((length + 6), LAN91C96_DATA_HIGH);
+	SMC_outw(dev, (length + 6), LAN91C96_DATA_HIGH);
 #endif /* USE_32_BIT */
 
 	/* send the actual data
@@ -516,54 +426,52 @@ static int smc_send_packet (volatile void *packet, int packet_length)
 	 * almost as much time as is saved?
 	 */
 #ifdef USE_32_BIT
-	SMC_outsl (LAN91C96_DATA_HIGH, buf, length >> 2);
+	SMC_outsl(dev, LAN91C96_DATA_HIGH, buf, length >> 2);
 	if (length & 0x2)
-		SMC_outw (*((word *) (buf + (length & 0xFFFFFFFC))),
+		SMC_outw(dev, *((word *) (buf + (length & 0xFFFFFFFC))),
 				  LAN91C96_DATA_HIGH);
 #else
-	SMC_outsw (LAN91C96_DATA_HIGH, buf, (length) >> 1);
+	SMC_outsw(dev, LAN91C96_DATA_HIGH, buf, (length) >> 1);
 #endif /* USE_32_BIT */
 
 	/* Send the last byte, if there is one.   */
 	if ((length & 1) == 0) {
-		SMC_outw (0, LAN91C96_DATA_HIGH);
+		SMC_outw(dev, 0, LAN91C96_DATA_HIGH);
 	} else {
-		SMC_outw (buf[length - 1] | 0x2000, LAN91C96_DATA_HIGH);
+		SMC_outw(dev, buf[length - 1] | 0x2000, LAN91C96_DATA_HIGH);
 	}
 
 	/* and let the chipset deal with it */
-	SMC_outw (LAN91C96_MMUCR_ENQUEUE, LAN91C96_MMU);
+	SMC_outw(dev, LAN91C96_MMUCR_ENQUEUE, LAN91C96_MMU);
 
 	/* poll for TX INT */
-	if (poll4int (LAN91C96_MSK_TX_INT, SMC_TX_TIMEOUT)) {
+	if (poll4int (dev, LAN91C96_MSK_TX_INT, SMC_TX_TIMEOUT)) {
 		/* sending failed */
-		PRINTK2 ("%s: TX timeout, sending failed...\n", SMC_DEV_NAME);
+		PRINTK2("%s: TX timeout, sending failed...\n", dev->name);
 
 		/* release packet */
-		SMC_outw (LAN91C96_MMUCR_RELEASE_TX, LAN91C96_MMU);
+		SMC_outw(dev, LAN91C96_MMUCR_RELEASE_TX, LAN91C96_MMU);
 
 		/* wait for MMU getting ready (low) */
-		while (SMC_inw (LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY) {
+		while (SMC_inw(dev, LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY)
 			udelay (10);
-		}
 
-		PRINTK2 ("MMU ready\n");
+		PRINTK2("MMU ready\n");
 
 
 		return 0;
 	} else {
 		/* ack. int */
-		SMC_outw (LAN91C96_IST_TX_INT, LAN91C96_INT_STATS);
+		SMC_outw(dev, LAN91C96_IST_TX_INT, LAN91C96_INT_STATS);
 
-		PRINTK2 ("%s: Sent packet of length %d \n", SMC_DEV_NAME, length);
+		PRINTK2("%s: Sent packet of length %d \n", dev->name, length);
 
 		/* release packet */
-		SMC_outw (LAN91C96_MMUCR_RELEASE_TX, LAN91C96_MMU);
+		SMC_outw(dev, LAN91C96_MMUCR_RELEASE_TX, LAN91C96_MMU);
 
 		/* wait for MMU getting ready (low) */
-		while (SMC_inw (LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY) {
+		while (SMC_inw(dev, LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY)
 			udelay (10);
-		}
 
 		PRINTK2 ("MMU ready\n");
 	}
@@ -571,20 +479,6 @@ static int smc_send_packet (volatile void *packet, int packet_length)
 	return length;
 }
 
-/*-------------------------------------------------------------------------
- * smc_destructor( struct net_device * dev )
- *   Input parameters:
- *	dev, pointer to the device structure
- *
- *   Output:
- *	None.
- *--------------------------------------------------------------------------
- */
-void smc_destructor ()
-{
-	PRINTK2 (CARDNAME ":smc_destructor\n");
-}
-
 
 /*
  * Open and Initialize the board
@@ -592,20 +486,20 @@ void smc_destructor ()
  * Set up everything, reset the card, etc ..
  *
  */
-static int smc_open (bd_t *bd)
+static int smc_open(bd_t *bd, struct eth_device *dev)
 {
 	int i, err;			/* used to set hw ethernet address */
 
-	PRINTK2 ("%s:smc_open\n", SMC_DEV_NAME);
+	PRINTK2("%s:smc_open\n", dev->name);
 
 	/* reset the hardware */
 
-	smc_reset ();
-	smc_enable ();
-
-	SMC_SELECT_BANK (1);
+	smc_reset(dev);
+	smc_enable(dev);
 
-	err = smc_get_ethaddr (bd);	/* set smc_mac_addr, and sync it with u-boot globals */
+	SMC_SELECT_BANK(dev, 1);
+	/* set smc_mac_addr, and sync it with u-boot globals */
+	err = smc_get_ethaddr(bd, dev);
 	if (err < 0)
 		return -1;
 #ifdef USE_32_BIT
@@ -614,11 +508,11 @@ static int smc_open (bd_t *bd)
 
 		address = smc_mac_addr[i + 1] << 8;
 		address |= smc_mac_addr[i];
-		SMC_outw (address, LAN91C96_IA0 + i);
+		SMC_outw(dev, address, LAN91C96_IA0 + i);
 	}
 #else
 	for (i = 0; i < 6; i++)
-		SMC_outb (smc_mac_addr[i], LAN91C96_IA0 + i);
+		SMC_outb(dev, smc_mac_addr[i], LAN91C96_IA0 + i);
 #endif
 	return 0;
 }
@@ -635,7 +529,7 @@ static int smc_open (bd_t *bd)
  * o otherwise, read in the packet
  *-------------------------------------------------------------
  */
-static int smc_rcv ()
+static int smc_rcv(struct eth_device *dev)
 {
 	int packet_number;
 	word status;
@@ -647,26 +541,26 @@ static int smc_rcv ()
 #endif
 
 
-	SMC_SELECT_BANK (2);
-	packet_number = SMC_inw (LAN91C96_FIFO);
+	SMC_SELECT_BANK(dev, 2);
+	packet_number = SMC_inw(dev, LAN91C96_FIFO);
 
 	if (packet_number & LAN91C96_FIFO_RXEMPTY) {
 		return 0;
 	}
 
-	PRINTK3 ("%s:smc_rcv\n", SMC_DEV_NAME);
+	PRINTK3("%s:smc_rcv\n", dev->name);
 	/*  start reading from the start of the packet */
-	SMC_outw (LAN91C96_PTR_READ | LAN91C96_PTR_RCV |
+	SMC_outw(dev, LAN91C96_PTR_READ | LAN91C96_PTR_RCV |
 			  LAN91C96_PTR_AUTO_INCR, LAN91C96_POINTER);
 
 	/* First two words are status and packet_length */
 #ifdef USE_32_BIT
-	stat_len = SMC_inl (LAN91C96_DATA_HIGH);
+	stat_len = SMC_inl(dev, LAN91C96_DATA_HIGH);
 	status = stat_len & 0xffff;
 	packet_length = stat_len >> 16;
 #else
-	status = SMC_inw (LAN91C96_DATA_HIGH);
-	packet_length = SMC_inw (LAN91C96_DATA_HIGH);
+	status = SMC_inw(dev, LAN91C96_DATA_HIGH);
+	packet_length = SMC_inw(dev, LAN91C96_DATA_HIGH);
 #endif
 
 	packet_length &= 0x07ff;	/* mask off top bits */
@@ -690,13 +584,14 @@ static int smc_rcv ()
 		   to send the DWORDs or the bytes first, or some
 		   mixture.  A mixture might improve already slow PIO
 		   performance  */
-		SMC_insl (LAN91C96_DATA_HIGH, NetRxPackets[0], packet_length >> 2);
+		SMC_insl(dev, LAN91C96_DATA_HIGH, NetRxPackets[0],
+				packet_length >> 2);
 		/* read the left over bytes */
 		if (packet_length & 3) {
 			int i;
 
 			byte *tail = (byte *) (NetRxPackets[0] + (packet_length & ~3));
-			dword leftover = SMC_inl (LAN91C96_DATA_HIGH);
+			dword leftover = SMC_inl(dev, LAN91C96_DATA_HIGH);
 
 			for (i = 0; i < (packet_length & 3); i++)
 				*tail++ = (byte) (leftover >> (8 * i)) & 0xff;
@@ -704,13 +599,14 @@ static int smc_rcv ()
 #else
 		PRINTK3 (" Reading %d words and %d byte(s) \n",
 				 (packet_length >> 1), packet_length & 1);
-		SMC_insw (LAN91C96_DATA_HIGH, NetRxPackets[0], packet_length >> 1);
+		SMC_insw(dev, LAN91C96_DATA_HIGH, NetRxPackets[0],
+				packet_length >> 1);
 
 #endif /* USE_32_BIT */
 
 #if	SMC_DEBUG > 2
 		printf ("Receiving Packet\n");
-		print_packet (NetRxPackets[0], packet_length);
+		print_packet((byte *)NetRxPackets[0], packet_length);
 #endif
 	} else {
 		/* error ... */
@@ -718,13 +614,13 @@ static int smc_rcv ()
 		is_error = 1;
 	}
 
-	while (SMC_inw (LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY)
+	while (SMC_inw(dev, LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY)
 		udelay (1);		/* Wait until not busy */
 
 	/*  error or good, tell the card to get rid of this packet */
-	SMC_outw (LAN91C96_MMUCR_RELEASE_RX, LAN91C96_MMU);
+	SMC_outw(dev, LAN91C96_MMUCR_RELEASE_RX, LAN91C96_MMU);
 
-	while (SMC_inw (LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY)
+	while (SMC_inw(dev, LAN91C96_MMU) & LAN91C96_MMUCR_NO_BUSY)
 		udelay (1);		/* Wait until not busy */
 
 	if (!is_error) {
@@ -745,18 +641,18 @@ static int smc_rcv ()
  * an 'ifconfig ethX down'
  *
  -----------------------------------------------------*/
-static int smc_close ()
+static int smc_close(struct eth_device *dev)
 {
-	PRINTK2 ("%s:smc_close\n", SMC_DEV_NAME);
+	PRINTK2("%s:smc_close\n", dev->name);
 
 	/* clear everything */
-	smc_shutdown ();
+	smc_shutdown(dev);
 
 	return 0;
 }
 
 #if SMC_DEBUG > 2
-static void print_packet (byte * buf, int length)
+static void print_packet(byte *buf, int length)
 {
 #if 0
 	int i;
@@ -792,86 +688,40 @@ static void print_packet (byte * buf, int length)
 }
 #endif /* SMC_DEBUG > 2 */
 
-int eth_init (bd_t * bd)
-{
-	return (smc_open(bd));
-}
-
-void eth_halt ()
+static int  lan91c96_init(struct eth_device *dev, bd_t *bd)
 {
-	smc_close ();
+	return smc_open(bd, dev);
 }
 
-int eth_rx ()
+static void lan91c96_halt(struct eth_device *dev)
 {
-	return smc_rcv ();
+	smc_close(dev);
 }
 
-int eth_send (volatile void *packet, int length)
+static int lan91c96_recv(struct eth_device *dev)
 {
-	return smc_send_packet (packet, length);
+	return smc_rcv(dev);
 }
 
-
-#if 0
-/*-------------------------------------------------------------------------
- * smc_hw_init()
- *
- *   Function:
- *      Reset and enable the device, check if the I/O space location
- *      is correct
- *
- *   Input parameters:
- *      None
- *
- *   Output:
- *	0 --> success
- *	1 --> error
- *--------------------------------------------------------------------------
- */
-static int smc_hw_init ()
+static int lan91c96_send(struct eth_device *dev, volatile void *packet,
+		int length)
 {
-	unsigned short status_test;
-
-	/* The attribute register of the LAN91C96 is located at address
-	   0x0e000000 on the lubbock platform */
-	volatile unsigned *attaddr = (unsigned *) (0x0e000000);
-
-	/* first reset, then enable the device. Sequence is critical */
-	attaddr[LAN91C96_ECOR] |= LAN91C96_ECOR_SRESET;
-	udelay (100);
-	attaddr[LAN91C96_ECOR] &= ~LAN91C96_ECOR_SRESET;
-	attaddr[LAN91C96_ECOR] |= LAN91C96_ECOR_ENABLE;
-
-	/* force 16-bit mode */
-	attaddr[LAN91C96_ECSR] &= ~LAN91C96_ECSR_IOIS8;
-	udelay (100);
-
-	/* check if the I/O address is correct, the upper byte of the
-	   bank select register should read 0x33 */
-
-	status_test = SMC_inw (LAN91C96_BANK_SELECT);
-	if ((status_test & 0xFF00) != 0x3300) {
-		printf ("Failed to initialize ethernetchip\n");
-		return 1;
-	}
-	return 0;
+	return smc_send_packet(dev, packet, length);
 }
-#endif /* 0 */
 
-/* smc_get_ethaddr (bd_t * bd)
+/* smc_get_ethaddr
  *
  * This checks both the environment and the ROM for an ethernet address. If
  * found, the environment takes precedence.
  */
 
-int smc_get_ethaddr (bd_t * bd)
+static int smc_get_ethaddr(bd_t *bd, struct eth_device *dev)
 {
 	uchar v_mac[6];
 
 	if (!eth_getenv_enetaddr("ethaddr", v_mac)) {
 		/* get ROM mac value if any */
-		if (!get_rom_mac(v_mac)) {
+		if (!get_rom_mac(dev, v_mac)) {
 			printf("\n*** ERROR: ethaddr is NOT set !!\n");
 			return -1;
 		}
@@ -888,7 +738,7 @@ int smc_get_ethaddr (bd_t * bd)
  * Note, this has omly been tested for the OMAP730 P2.
  */
 
-int get_rom_mac (unsigned char *v_rom_mac)
+static int get_rom_mac(struct eth_device *dev, unsigned char *v_rom_mac)
 {
 #ifdef HARDCODE_MAC	/* used for testing or to supress run time warnings */
 	char hw_mac_addr[] = { 0x02, 0x80, 0xad, 0x20, 0x31, 0xb8 };
@@ -897,11 +747,75 @@ int get_rom_mac (unsigned char *v_rom_mac)
 	return (1);
 #else
 	int i;
-	SMC_SELECT_BANK (1);
+	SMC_SELECT_BANK(dev, 1);
 	for (i=0; i<6; i++)
 	{
-		v_rom_mac[i] = SMC_inb (LAN91C96_IA0 + i);
+		v_rom_mac[i] = SMC_inb(dev, LAN91C96_IA0 + i);
 	}
 	return (1);
 #endif
 }
+
+/* Structure to detect the device IDs */
+struct id_type {
+	u8 id;
+	char *name;
+};
+static struct id_type supported_chips[] = {
+	{0, ""}, /* Dummy entry to prevent id check failure */
+	{9, "LAN91C110"},
+	{8, "LAN91C100FD"},
+	{7, "LAN91C100"},
+	{5, "LAN91C95"},
+	{4, "LAN91C94/LAN91C96"},
+	{3, "LAN91C90/LAN91C92"},
+};
+/* lan91c96_detect_chip
+ * See:
+ * http://www.embeddedsys.com/subpages/resources/images/documents/LAN91C96_datasheet.pdf
+ * page 71 - that is the closest we get to detect this device
+ */
+static int lan91c96_detect_chip(struct eth_device *dev)
+{
+	u8 chip_id;
+	int r;
+	SMC_SELECT_BANK(dev, 3);
+	chip_id = SMC_inw(dev, 0xA) & LAN91C96_REV_REVID;
+	SMC_SELECT_BANK(dev, 0);
+	for (r = 0; r < sizeof(supported_chips) / sizeof(struct id_type); r++)
+		if (chip_id == supported_chips[r].id)
+			return r;
+	return 0;
+}
+
+int lan91c96_initialize(u8 dev_num, int base_addr)
+{
+	struct eth_device *dev;
+	int r = 0;
+
+	dev = malloc(sizeof(*dev));
+	if (!dev) {
+		free(dev);
+		return 0;
+	}
+	memset(dev, 0, sizeof(*dev));
+
+	dev->iobase = base_addr;
+
+	/* Try to detect chip. Will fail if not present. */
+	r = lan91c96_detect_chip(dev);
+	if (!r) {
+		free(dev);
+		return 0;
+	}
+	get_rom_mac(dev, dev->enetaddr);
+
+	dev->init = lan91c96_init;
+	dev->halt = lan91c96_halt;
+	dev->send = lan91c96_send;
+	dev->recv = lan91c96_recv;
+	sprintf(dev->name, "%s-%hu", supported_chips[r].name, dev_num);
+
+	eth_register(dev);
+	return 0;
+}
diff --git a/drivers/net/lan91c96.h b/drivers/net/lan91c96.h
index 5beddda..6fbb0e3 100644
--- a/drivers/net/lan91c96.h
+++ b/drivers/net/lan91c96.h
@@ -46,14 +46,6 @@
 #include <asm/io.h>
 #include <config.h>
 
-/*
- * This function may be called by the board specific initialisation code
- * in order to override the default mac address.
- */
-
-void smc_set_mac_addr(const unsigned char *addr);
-
-
 /* I want some simple types */
 
 typedef unsigned char			byte;
@@ -86,66 +78,71 @@ typedef unsigned long int		dword;
 #define	SMC_IO_SHIFT	0
 #endif
 
-#define	SMCREG(r)	(SMC_BASE_ADDRESS+((r)<<SMC_IO_SHIFT))
+#define	SMCREG(edev, r)	((edev)->iobase+((r)<<SMC_IO_SHIFT))
 
-#define	SMC_inl(r)	(*((volatile dword *)SMCREG(r)))
-#define	SMC_inw(r)	(*((volatile word *)SMCREG(r)))
-#define SMC_inb(p) ({ \
+#define	SMC_inl(edev, r)	(*((volatile dword *)SMCREG(edev, r)))
+#define	SMC_inw(edev, r)	(*((volatile word *)SMCREG(edev, r)))
+#define SMC_inb(edev, p) ({ \
 	unsigned int __p = p; \
-	unsigned int __v = SMC_inw(__p & ~1); \
+	unsigned int __v = SMC_inw(edev, __p & ~1); \
 	if (__p & 1) __v >>= 8; \
 	else __v &= 0xff; \
 	__v; })
 
-#define	SMC_outl(d,r)	(*((volatile dword *)SMCREG(r)) = d)
-#define	SMC_outw(d,r)	(*((volatile word *)SMCREG(r)) = d)
-#define	SMC_outb(d,r)	({	word __d = (byte)(d);  \
-				word __w = SMC_inw((r)&~1);  \
+#define	SMC_outl(edev, d, r)	(*((volatile dword *)SMCREG(edev, r)) = d)
+#define	SMC_outw(edev, d, r)	(*((volatile word *)SMCREG(edev, r)) = d)
+#define	SMC_outb(edev, d, r)	({	word __d = (byte)(d);  \
+				word __w = SMC_inw(edev, (r)&~1);  \
 				__w &= ((r)&1) ? 0x00FF : 0xFF00;  \
 				__w |= ((r)&1) ? __d<<8 : __d;  \
-				SMC_outw(__w,(r)&~1);  \
+				SMC_outw(edev, __w, (r)&~1);  \
 			})
 
-#define SMC_outsl(r,b,l)	({	int __i; \
+#define SMC_outsl(edev, r, b, l)	({	int __i; \
 					dword *__b2; \
 					__b2 = (dword *) b; \
 					for (__i = 0; __i < l; __i++) { \
-					    SMC_outl( *(__b2 + __i), r ); \
+						SMC_outl(edev, *(__b2 + __i),\
+							r); \
 					} \
 				})
 
-#define SMC_outsw(r,b,l)	({	int __i; \
+#define SMC_outsw(edev, r, b, l)	({	int __i; \
 					word *__b2; \
 					__b2 = (word *) b; \
 					for (__i = 0; __i < l; __i++) { \
-					    SMC_outw( *(__b2 + __i), r ); \
+						SMC_outw(edev, *(__b2 + __i),\
+							r); \
 					} \
 				})
 
-#define SMC_insl(r,b,l)		({	int __i ;  \
+#define SMC_insl(edev, r, b, l)		({	int __i ;  \
 					dword *__b2;  \
 					__b2 = (dword *) b;  \
 					for (__i = 0; __i < l; __i++) {  \
-					  *(__b2 + __i) = SMC_inl(r);  \
-					  SMC_inl(0);  \
+						*(__b2 + __i) = SMC_inl(edev,\
+							r);  \
+						SMC_inl(edev, 0);  \
 					};  \
 				})
 
-#define SMC_insw(r,b,l)		({	int __i ;  \
+#define SMC_insw(edev, r, b, l)		({	int __i ;  \
 					word *__b2;  \
 					__b2 = (word *) b;  \
 					for (__i = 0; __i < l; __i++) {  \
-					  *(__b2 + __i) = SMC_inw(r);  \
-					  SMC_inw(0);  \
+						*(__b2 + __i) = SMC_inw(edev,\
+							r);  \
+						SMC_inw(edev, 0);  \
 					};  \
 				})
 
-#define SMC_insb(r,b,l)		({	int __i ;  \
+#define SMC_insb(edev, r, b, l)		({	int __i ;  \
 					byte *__b2;  \
 					__b2 = (byte *) b;  \
 					for (__i = 0; __i < l; __i++) {  \
-					  *(__b2 + __i) = SMC_inb(r);  \
-					  SMC_inb(0);  \
+						*(__b2 + __i) = SMC_inb(edev,\
+							r);  \
+						SMC_inb(edev, 0);  \
 					};  \
 				})
 
@@ -155,40 +152,35 @@ typedef unsigned long int		dword;
  * We have only 16 Bit PCMCIA access on Socket 0
  */
 
-#define	SMC_inw(r)	(*((volatile word *)(SMC_BASE_ADDRESS+(r))))
-#define  SMC_inb(r)	(((r)&1) ? SMC_inw((r)&~1)>>8 : SMC_inw(r)&0xFF)
+#define	SMC_inw(edev, r)	(*((volatile word *)((edev)->iobase+(r))))
+#define  SMC_inb(edev, r)	(((r)&1) ? SMC_inw(edev, (r)&~1)>>8 :\
+					SMC_inw(edev, r)&0xFF)
 
-#define	SMC_outw(d,r)	(*((volatile word *)(SMC_BASE_ADDRESS+(r))) = d)
-#define	SMC_outb(d,r)	({	word __d = (byte)(d);  \
-				word __w = SMC_inw((r)&~1);  \
+#define	SMC_outw(edev, d, r)	(*((volatile word *)((edev)->iobase+(r))) = d)
+#define	SMC_outb(edev, d, r)	({	word __d = (byte)(d);  \
+				word __w = SMC_inw(edev, (r)&~1);  \
 				__w &= ((r)&1) ? 0x00FF : 0xFF00;  \
 				__w |= ((r)&1) ? __d<<8 : __d;  \
-				SMC_outw(__w,(r)&~1);  \
+				SMC_outw(edev, __w, (r)&~1);  \
 			})
-#if 0
-#define	SMC_outsw(r,b,l)	outsw(SMC_BASE_ADDRESS+(r), (b), (l))
-#else
-#define SMC_outsw(r,b,l)	({	int __i; \
+#define SMC_outsw(edev, r, b, l)	({	int __i; \
 					word *__b2; \
 					__b2 = (word *) b; \
 					for (__i = 0; __i < l; __i++) { \
-					    SMC_outw( *(__b2 + __i), r); \
+						SMC_outw(edev, *(__b2 + __i),\
+							r); \
 					} \
 				})
-#endif
 
-#if 0
-#define	SMC_insw(r,b,l)	insw(SMC_BASE_ADDRESS+(r), (b), (l))
-#else
-#define SMC_insw(r,b,l)	({	int __i ;  \
+#define SMC_insw(edev, r, b, l)	({	int __i ;  \
 					word *__b2;  \
 					__b2 = (word *) b;  \
 					for (__i = 0; __i < l; __i++) {  \
-					  *(__b2 + __i) = SMC_inw(r);  \
-					  SMC_inw(0);  \
+						*(__b2 + __i) = SMC_inw(edev,\
+							r);  \
+						SMC_inw(edev, 0);  \
 					};  \
 				})
-#endif
 
 #endif
 
@@ -608,25 +600,25 @@ typedef unsigned long int		dword;
 
 /* select a register bank, 0 to 3  */
 
-#define SMC_SELECT_BANK(x)  { SMC_outw( x, LAN91C96_BANK_SELECT ); }
+#define SMC_SELECT_BANK(edev, x)  { SMC_outw(edev, x, LAN91C96_BANK_SELECT); }
 
 /* this enables an interrupt in the interrupt mask register */
-#define SMC_ENABLE_INT(x) {\
+#define SMC_ENABLE_INT(edev, x) {\
 		unsigned char mask;\
-		SMC_SELECT_BANK(2);\
-		mask = SMC_inb( LAN91C96_INT_MASK );\
+		SMC_SELECT_BANK(edev, 2);\
+		mask = SMC_inb(edev, LAN91C96_INT_MASK);\
 		mask |= (x);\
-		SMC_outb( mask, LAN91C96_INT_MASK ); \
+		SMC_outb(edev, mask, LAN91C96_INT_MASK); \
 }
 
 /* this disables an interrupt from the interrupt mask register */
 
-#define SMC_DISABLE_INT(x) {\
+#define SMC_DISABLE_INT(edev, x) {\
 		unsigned char mask;\
-		SMC_SELECT_BANK(2);\
-		mask = SMC_inb( LAN91C96_INT_MASK );\
+		SMC_SELECT_BANK(edev, 2);\
+		mask = SMC_inb(edev, LAN91C96_INT_MASK);\
 		mask &= ~(x);\
-		SMC_outb( mask, LAN91C96_INT_MASK ); \
+		SMC_outb(edev, mask, LAN91C96_INT_MASK); \
 }
 
 /*----------------------------------------------------------------------
diff --git a/include/netdev.h b/include/netdev.h
index a50ec67..9baccd4 100644
--- a/include/netdev.h
+++ b/include/netdev.h
@@ -57,6 +57,7 @@ int greth_initialize(bd_t *bis);
 void gt6426x_eth_initialize(bd_t *bis);
 int inca_switch_initialize(bd_t *bis);
 int kirkwood_egiga_initialize(bd_t *bis);
+int lan91c96_initialize(u8 dev_num, int base_addr);
 int macb_eth_initialize(int id, void *regs, unsigned int phy_addr);
 int mcdmafec_initialize(bd_t *bis);
 int mcffec_initialize(bd_t *bis);
-- 
1.6.3.3

[U-Boot] [PATCH 2/3] TI OMAP3: SDP3430 FIX NET_MULTI Warning

[Nishanth Menon]

Enable the NET MULTI option and remove build warning

Tested: SDP3430

Signed-off-by: Nishanth Menon <nm@ti.com>
Cc: Ben Warren <biggerbadderben@gmail.com>
Cc: Sandeep Paulraj <s-paulraj@ti.com>
---
 board/ti/sdp3430/sdp.c          |   12 +++++++-----
 include/configs/omap3_sdp3430.h |    3 ++-
 2 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/board/ti/sdp3430/sdp.c b/board/ti/sdp3430/sdp.c
index 40cf26f..0d8e20d 100644
--- a/board/ti/sdp3430/sdp.c
+++ b/board/ti/sdp3430/sdp.c
@@ -22,6 +22,7 @@
  * MA 02111-1307 USA
  */
 #include <common.h>
+#include <netdev.h>
 #include <twl4030.h>
 #include <asm/io.h>
 #include <asm/arch/mux.h>
@@ -125,12 +126,13 @@ int board_init(void)
 #define ETH_CONTROL_REG		(CONFIG_LAN91C96_BASE + 0x30b)
 
 /**
- * @brief ether_init Take the Ethernet controller out of reset and wait
+ * @brief board_eth_init Take the Ethernet controller out of reset and wait
  * for the EEPROM load to complete.
  */
-static void ether_init(void)
+int board_eth_init(bd_t *bis)
 {
-#ifdef CONFIG_DRIVER_LAN91C96
+	int rc = 0;
+#ifdef CONFIG_LAN91C96
 	int cnt = 20;
 
 	writew(0x0, LAN_RESET_REGISTER);
@@ -155,10 +157,11 @@ static void ether_init(void)
 
 	writeb(readb(ETH_CONTROL_REG) & ~0x1, ETH_CONTROL_REG);
 	udelay(1000);
+	rc = lan91c96_initialize(0, CONFIG_LAN91C96_BASE);
 reset_err_out:
-	return;
 
 #endif
+	return rc;
 }
 
 /**
@@ -187,7 +190,6 @@ int misc_init_r(void)
 	 *   VSIM  - off (init, variable) for MMC1.DAT[3..7], SIM
 	 *   VPLL2 - 1.8V
 	 */
-	ether_init();
 
 	return 0;
 }
diff --git a/include/configs/omap3_sdp3430.h b/include/configs/omap3_sdp3430.h
index 229dc5e..7401671 100644
--- a/include/configs/omap3_sdp3430.h
+++ b/include/configs/omap3_sdp3430.h
@@ -200,7 +200,8 @@
  */
 #if defined(CONFIG_CMD_NET)
 
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE	DEBUG_BASE
 #define CONFIG_LAN91C96_EXT_PHY
 
-- 
1.6.3.3

[U-Boot] [PATCH 3/3] LAN91C96: Enable NET_MULTI LAN driver

[Nishanth Menon]

This modification is NOT tested on any of the
platforms modified as I dont have them. please
help by testing+building+fixing

Signed-off-by: Nishanth Menon <nm@ti.com>
Cc: Andrea Scian <andrea.scian@dave-tech.it>
Cc: Ben Warren <biggerbadderben@gmail.com>
Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
Cc: George G. Davis <gdavis@mvista.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Nishant Kamat <nskamat@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Rishi Bhattacharya <rishi@ti.com>
Cc: Sandeep Paulraj <s-paulraj@ti.com>
---
 board/apollon/apollon.c       |   12 +++++++-----
 include/configs/B2.h          |    2 +-
 include/configs/apollon.h     |    3 ++-
 include/configs/assabet.h     |    3 ++-
 include/configs/gcplus.h      |    3 ++-
 include/configs/lubbock.h     |    3 ++-
 include/configs/omap1510inn.h |    3 ++-
 include/configs/omap1610h2.h  |    3 ++-
 include/configs/omap1610inn.h |    3 ++-
 include/configs/omap2420h4.h  |    3 ++-
 include/configs/omap5912osk.h |    3 ++-
 include/configs/omap730p2.h   |    3 ++-
 include/configs/pleb2.h       |    6 ++++--
 13 files changed, 32 insertions(+), 18 deletions(-)

diff --git a/board/apollon/apollon.c b/board/apollon/apollon.c
index 8964eba..b93e880 100644
--- a/board/apollon/apollon.c
+++ b/board/apollon/apollon.c
@@ -24,6 +24,7 @@
  * MA 02111-1307 USA
  */
 #include <common.h>
+#include <netdev.h>
 #include <asm/arch/omap2420.h>
 #include <asm/io.h>
 #include <asm/arch/bits.h>
@@ -138,13 +139,14 @@ void wait_for_command_complete(unsigned int wd_base)
 }
 
 /*******************************************************************
- * Routine:ether_init
+ * Routine:board_eth_init
  * Description: take the Ethernet controller out of reset and wait
  *		   for the EEPROM load to complete.
  ******************************************************************/
-void ether_init(void)
+int board_eth_init(bd_t *bis)
 {
-#ifdef CONFIG_DRIVER_LAN91C96
+	int rc = 0;
+#ifdef CONFIG_LAN91C96
 	int cnt = 20;
 
 	__raw_writeb(0x03, OMAP2420_CTRL_BASE + 0x0f2);	/*protect->gpio74 */
@@ -171,10 +173,10 @@ void ether_init(void)
 
 	mask_config_reg(ETH_CONTROL_REG, 0x01);
 	udelay(1000);
-
+	rc = lan91c96_initialize(0, CONFIG_LAN91C96_BASE);
 eth_reset_err_out:
-	return;
 #endif
+	return rc;
 }
 
 /**********************************************
diff --git a/include/configs/B2.h b/include/configs/B2.h
index e5439f3..d7806e9 100644
--- a/include/configs/B2.h
+++ b/include/configs/B2.h
@@ -58,7 +58,7 @@
 /*
  * Hardware drivers
  */
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE		0x04000300 /* base address         */
 #define CONFIG_SMC_USE_32_BIT
 #undef  CONFIG_SHOW_ACTIVITY
diff --git a/include/configs/apollon.h b/include/configs/apollon.h
index 575f60e..be0dc04 100644
--- a/include/configs/apollon.h
+++ b/include/configs/apollon.h
@@ -87,7 +87,8 @@
 /*
  * SMC91c96 Etherent
  */
-#define	CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define	CONFIG_LAN91C96
 #define	CONFIG_LAN91C96_BASE	(APOLLON_CS1_BASE+0x300)
 #define	CONFIG_LAN91C96_EXT_PHY
 
diff --git a/include/configs/assabet.h b/include/configs/assabet.h
index 8c5b84c..d17d4bd 100644
--- a/include/configs/assabet.h
+++ b/include/configs/assabet.h
@@ -53,7 +53,8 @@
 /*
  * Hardware drivers
  */
-#define CONFIG_DRIVER_LAN91C96	/* we have an SMC9194 on-board */
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96	/* we have an SMC9194 on-board */
 #define CONFIG_LAN91C96_BASE	0x18000000
 
 /*
diff --git a/include/configs/gcplus.h b/include/configs/gcplus.h
index 85db4f5..41294b9 100644
--- a/include/configs/gcplus.h
+++ b/include/configs/gcplus.h
@@ -66,7 +66,8 @@
 /*
  * Hardware drivers
  */
-#define CONFIG_DRIVER_LAN91C96	/* we have an SMC9194 on-board */
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96	/* we have an SMC9194 on-board */
 #define CONFIG_LAN91C96_BASE	0x100e0000
 
 /*
diff --git a/include/configs/lubbock.h b/include/configs/lubbock.h
index 43913ca..0a69210 100644
--- a/include/configs/lubbock.h
+++ b/include/configs/lubbock.h
@@ -58,7 +58,8 @@
 /*
  * Hardware drivers
  */
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE 0x0C000000
 
 /*
diff --git a/include/configs/omap1510inn.h b/include/configs/omap1510inn.h
index 8408209..b0ebafd 100644
--- a/include/configs/omap1510inn.h
+++ b/include/configs/omap1510inn.h
@@ -60,7 +60,8 @@
 #define CONFIG_SMC9196_BASE 0x08000300
 #define CONFIG_SMC9196_EXT_PHY
 */
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE 0x08000300
 #define CONFIG_LAN91C96_EXT_PHY
 
diff --git a/include/configs/omap1610h2.h b/include/configs/omap1610h2.h
index 42e0198..0bbb5b3 100644
--- a/include/configs/omap1610h2.h
+++ b/include/configs/omap1610h2.h
@@ -57,7 +57,8 @@
 /*
  * Hardware drivers
  */
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE 0x04000300
 #define CONFIG_LAN91C96_EXT_PHY
 
diff --git a/include/configs/omap1610inn.h b/include/configs/omap1610inn.h
index 22c873e..832dd42 100644
--- a/include/configs/omap1610inn.h
+++ b/include/configs/omap1610inn.h
@@ -58,7 +58,8 @@
  */
 /*
 */
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE 0x04000300
 #define CONFIG_LAN91C96_EXT_PHY
 
diff --git a/include/configs/omap2420h4.h b/include/configs/omap2420h4.h
index 9c18842..6ab4438 100644
--- a/include/configs/omap2420h4.h
+++ b/include/configs/omap2420h4.h
@@ -81,7 +81,8 @@
 /*
  * SMC91c96 Etherent
  */
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE     (H4_CS1_BASE+0x300)
 #define CONFIG_LAN91C96_EXT_PHY
 
diff --git a/include/configs/omap5912osk.h b/include/configs/omap5912osk.h
index d0ce9dc..bc660e3 100644
--- a/include/configs/omap5912osk.h
+++ b/include/configs/omap5912osk.h
@@ -61,7 +61,8 @@
  */
 /*
 */
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE 0x04800300
 #define CONFIG_LAN91C96_EXT_PHY
 
diff --git a/include/configs/omap730p2.h b/include/configs/omap730p2.h
index 32a9b23..a6a8a02 100644
--- a/include/configs/omap730p2.h
+++ b/include/configs/omap730p2.h
@@ -65,7 +65,8 @@
  * Hardware drivers
  */
 
-#define CONFIG_DRIVER_LAN91C96
+#define CONFIG_NET_MULTI
+#define CONFIG_LAN91C96
 #define CONFIG_LAN91C96_BASE	   0x04000300
 #define CONFIG_LAN91C96_EXT_PHY
 
diff --git a/include/configs/pleb2.h b/include/configs/pleb2.h
index 635ef71..9e69411 100644
--- a/include/configs/pleb2.h
+++ b/include/configs/pleb2.h
@@ -56,8 +56,10 @@
  */
 
 /* None - PLEB 2 doesn't have any of this.
-   #define CONFIG_DRIVER_LAN91C96
-   #define CONFIG_LAN91C96_BASE 0x0C000000 */
+	#define CONFIG_NET_MULTI
+	#define CONFIG_LAN91C96
+	#define CONFIG_LAN91C96_BASE 0x0C000000
+ */
 
 /*
  * select serial console configuration
-- 
1.6.3.3

[U-Boot] [PATCH 2/3] TI OMAP3: SDP3430 FIX NET_MULTI Warning

[Dirk Behme]

Nishanth Menon wrote:
> Enable the NET MULTI option and remove build warning
> 
> Tested: SDP3430

Great! Thanks!

I'll keep my fingers crossed that it will work on the other boards, 
too, and will get an ack.

Best regards

Dirk

[U-Boot] [PATCH 0/3] enable CONFIG_NET_MULTI for LAN91C96

[Ben Warren]

Hi Nishanth,

Nishanth Menon wrote:
> Based on the discussion[1] here is the patchset
>
> Patchset is based on [2], and tested on SDP3430.
> Other platforms would be prefered to be
> build/tested/fixed as required
>
> Nishanth Menon (3):
>   NET: LAN91C96 CONFIG_NET_MULTIify
>   TI OMAP3: SDP3430 FIX NET_MULTI Warning
>   LAN91C96: Enable NET_MULTI LAN driver
>
>   
On first pass review, this looks great!  There are a few things I'd like 
to do to improve the driver, but it has nothing to do with converting to 
NET_MULTI.

I'm going to fire up a net-next branch this weekend and will try to 
apply it there.

regards,
Ben

[U-Boot] [PATCH 0/3] enable CONFIG_NET_MULTI for LAN91C96

[Menon, Nishanth]

> -----Original Message-----
> From: Ben Warren [mailto:biggerbadderben at gmail.com]
> Sent: Friday, October 16, 2009 11:30 AM
> Nishanth Menon wrote:
> > Based on the discussion[1] here is the patchset
> >
> > Patchset is based on [2], and tested on SDP3430.
> > Other platforms would be prefered to be
> > build/tested/fixed as required
> >
> > Nishanth Menon (3):
> >   NET: LAN91C96 CONFIG_NET_MULTIify
> >   TI OMAP3: SDP3430 FIX NET_MULTI Warning
> >   LAN91C96: Enable NET_MULTI LAN driver
> >
> >
> On first pass review, this looks great!  There are a few things I'd like
> to do to improve the driver, but it has nothing to do with converting to
> NET_MULTI.
> 
> I'm going to fire up a net-next branch this weekend and will try to
> apply it there.
Gee thanks..

Having very less knowledge on lan91c96 itself, I can help be the guinea pig 
for testing on SDP3430 at least :)

Regards,
Nishanth Menon

[U-Boot] [PATCH 3/3] LAN91C96: Enable NET_MULTI LAN driver

[Tom]

Nishanth Menon wrote:
> This modification is NOT tested on any of the
> platforms modified as I dont have them. please
> help by testing+building+fixing

On the testing front..
Regression testing on MAKEALL shows

#warning Ethernet driver .. error killing for
assabet, gcplus, lubbock, omap1510inn, omap1610h2,
omap2420h4, omap3_sdp3430, omap5912osk and omap730p2

There is a regression on B2.

et/libnet.a(net.o): In function `NetSendPacket':
u-boot-ti/net/net.c:623: undefined reference to `eth_send'
	net/libnet.a(net.o): In function `NetReceive':
u-boot-ti/net/net.c:1433: undefined reference to `eth_send'

Likely because you missed adding
#define CONFIG_NET_MULTI
to the config file.

apollon, does not configure. Not your fault
but testing on it is going to be hard ;)
Here is the error :

System not configured - see README
make: *** [all] Error 1

No new additions to apollon should be made until the
base cause of the apollon failure is fixed.

So
1. Fix B2.
2. Remove apollon changes
3. Wait for runtime testing.

I am not in favor of including this patch in the 2009.11
as we will likely apply the patch after waiting a reasonable
amount of time on the runtime testing.  Which is not the
same as knowing we aren't breaking the old boards.

Tom

> 
> Signed-off-by: Nishanth Menon <nm@ti.com>
> Cc: Andrea Scian <andrea.scian@dave-tech.it>
> Cc: Ben Warren <biggerbadderben@gmail.com>
> Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
> Cc: George G. Davis <gdavis@mvista.com>
> Cc: Kyungmin Park <kyungmin.park@samsung.com>
> Cc: Nishant Kamat <nskamat@ti.com>
> Cc: Richard Woodruff <r-woodruff2@ti.com>
> Cc: Rishi Bhattacharya <rishi@ti.com>
> Cc: Sandeep Paulraj <s-paulraj@ti.com>
> ---
>  board/apollon/apollon.c       |   12 +++++++-----
>  include/configs/B2.h          |    2 +-
>  include/configs/apollon.h     |    3 ++-
>  include/configs/assabet.h     |    3 ++-
>  include/configs/gcplus.h      |    3 ++-
>  include/configs/lubbock.h     |    3 ++-
>  include/configs/omap1510inn.h |    3 ++-
>  include/configs/omap1610h2.h  |    3 ++-
>  include/configs/omap1610inn.h |    3 ++-
>  include/configs/omap2420h4.h  |    3 ++-
>  include/configs/omap5912osk.h |    3 ++-
>  include/configs/omap730p2.h   |    3 ++-
>  include/configs/pleb2.h       |    6 ++++--
>  13 files changed, 32 insertions(+), 18 deletions(-)
> 
> diff --git a/board/apollon/apollon.c b/board/apollon/apollon.c
> index 8964eba..b93e880 100644
> --- a/board/apollon/apollon.c
> +++ b/board/apollon/apollon.c
> @@ -24,6 +24,7 @@
>   * MA 02111-1307 USA
>   */
>  #include <common.h>
> +#include <netdev.h>
>  #include <asm/arch/omap2420.h>
>  #include <asm/io.h>
>  #include <asm/arch/bits.h>
> @@ -138,13 +139,14 @@ void wait_for_command_complete(unsigned int wd_base)
>  }
>  
>  /*******************************************************************
> - * Routine:ether_init
> + * Routine:board_eth_init
>   * Description: take the Ethernet controller out of reset and wait
>   *		   for the EEPROM load to complete.
>   ******************************************************************/
> -void ether_init(void)
> +int board_eth_init(bd_t *bis)
>  {
> -#ifdef CONFIG_DRIVER_LAN91C96
> +	int rc = 0;
> +#ifdef CONFIG_LAN91C96
>  	int cnt = 20;
>  
>  	__raw_writeb(0x03, OMAP2420_CTRL_BASE + 0x0f2);	/*protect->gpio74 */
> @@ -171,10 +173,10 @@ void ether_init(void)
>  
>  	mask_config_reg(ETH_CONTROL_REG, 0x01);
>  	udelay(1000);
> -
> +	rc = lan91c96_initialize(0, CONFIG_LAN91C96_BASE);
>  eth_reset_err_out:
> -	return;
>  #endif
> +	return rc;
>  }
>  
>  /**********************************************
> diff --git a/include/configs/B2.h b/include/configs/B2.h
> index e5439f3..d7806e9 100644
> --- a/include/configs/B2.h
> +++ b/include/configs/B2.h
> @@ -58,7 +58,7 @@
>  /*
>   * Hardware drivers
>   */
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE		0x04000300 /* base address         */
>  #define CONFIG_SMC_USE_32_BIT
>  #undef  CONFIG_SHOW_ACTIVITY
> diff --git a/include/configs/apollon.h b/include/configs/apollon.h
> index 575f60e..be0dc04 100644
> --- a/include/configs/apollon.h
> +++ b/include/configs/apollon.h
> @@ -87,7 +87,8 @@
>  /*
>   * SMC91c96 Etherent
>   */
> -#define	CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define	CONFIG_LAN91C96
>  #define	CONFIG_LAN91C96_BASE	(APOLLON_CS1_BASE+0x300)
>  #define	CONFIG_LAN91C96_EXT_PHY
>  
> diff --git a/include/configs/assabet.h b/include/configs/assabet.h
> index 8c5b84c..d17d4bd 100644
> --- a/include/configs/assabet.h
> +++ b/include/configs/assabet.h
> @@ -53,7 +53,8 @@
>  /*
>   * Hardware drivers
>   */
> -#define CONFIG_DRIVER_LAN91C96	/* we have an SMC9194 on-board */
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96	/* we have an SMC9194 on-board */
>  #define CONFIG_LAN91C96_BASE	0x18000000
>  
>  /*
> diff --git a/include/configs/gcplus.h b/include/configs/gcplus.h
> index 85db4f5..41294b9 100644
> --- a/include/configs/gcplus.h
> +++ b/include/configs/gcplus.h
> @@ -66,7 +66,8 @@
>  /*
>   * Hardware drivers
>   */
> -#define CONFIG_DRIVER_LAN91C96	/* we have an SMC9194 on-board */
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96	/* we have an SMC9194 on-board */
>  #define CONFIG_LAN91C96_BASE	0x100e0000
>  
>  /*
> diff --git a/include/configs/lubbock.h b/include/configs/lubbock.h
> index 43913ca..0a69210 100644
> --- a/include/configs/lubbock.h
> +++ b/include/configs/lubbock.h
> @@ -58,7 +58,8 @@
>  /*
>   * Hardware drivers
>   */
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE 0x0C000000
>  
>  /*
> diff --git a/include/configs/omap1510inn.h b/include/configs/omap1510inn.h
> index 8408209..b0ebafd 100644
> --- a/include/configs/omap1510inn.h
> +++ b/include/configs/omap1510inn.h
> @@ -60,7 +60,8 @@
>  #define CONFIG_SMC9196_BASE 0x08000300
>  #define CONFIG_SMC9196_EXT_PHY
>  */
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE 0x08000300
>  #define CONFIG_LAN91C96_EXT_PHY
>  
> diff --git a/include/configs/omap1610h2.h b/include/configs/omap1610h2.h
> index 42e0198..0bbb5b3 100644
> --- a/include/configs/omap1610h2.h
> +++ b/include/configs/omap1610h2.h
> @@ -57,7 +57,8 @@
>  /*
>   * Hardware drivers
>   */
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE 0x04000300
>  #define CONFIG_LAN91C96_EXT_PHY
>  
> diff --git a/include/configs/omap1610inn.h b/include/configs/omap1610inn.h
> index 22c873e..832dd42 100644
> --- a/include/configs/omap1610inn.h
> +++ b/include/configs/omap1610inn.h
> @@ -58,7 +58,8 @@
>   */
>  /*
>  */
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE 0x04000300
>  #define CONFIG_LAN91C96_EXT_PHY
>  
> diff --git a/include/configs/omap2420h4.h b/include/configs/omap2420h4.h
> index 9c18842..6ab4438 100644
> --- a/include/configs/omap2420h4.h
> +++ b/include/configs/omap2420h4.h
> @@ -81,7 +81,8 @@
>  /*
>   * SMC91c96 Etherent
>   */
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE     (H4_CS1_BASE+0x300)
>  #define CONFIG_LAN91C96_EXT_PHY
>  
> diff --git a/include/configs/omap5912osk.h b/include/configs/omap5912osk.h
> index d0ce9dc..bc660e3 100644
> --- a/include/configs/omap5912osk.h
> +++ b/include/configs/omap5912osk.h
> @@ -61,7 +61,8 @@
>   */
>  /*
>  */
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE 0x04800300
>  #define CONFIG_LAN91C96_EXT_PHY
>  
> diff --git a/include/configs/omap730p2.h b/include/configs/omap730p2.h
> index 32a9b23..a6a8a02 100644
> --- a/include/configs/omap730p2.h
> +++ b/include/configs/omap730p2.h
> @@ -65,7 +65,8 @@
>   * Hardware drivers
>   */
>  
> -#define CONFIG_DRIVER_LAN91C96
> +#define CONFIG_NET_MULTI
> +#define CONFIG_LAN91C96
>  #define CONFIG_LAN91C96_BASE	   0x04000300
>  #define CONFIG_LAN91C96_EXT_PHY
>  
> diff --git a/include/configs/pleb2.h b/include/configs/pleb2.h
> index 635ef71..9e69411 100644
> --- a/include/configs/pleb2.h
> +++ b/include/configs/pleb2.h
> @@ -56,8 +56,10 @@
>   */
>  
>  /* None - PLEB 2 doesn't have any of this.
> -   #define CONFIG_DRIVER_LAN91C96
> -   #define CONFIG_LAN91C96_BASE 0x0C000000 */
> +	#define CONFIG_NET_MULTI
> +	#define CONFIG_LAN91C96
> +	#define CONFIG_LAN91C96_BASE 0x0C000000
> + */
>  
>  /*
>   * select serial console configuration

[U-Boot] [PATCH 3/3] LAN91C96: Enable NET_MULTI LAN driver

[Ben Warren]

Tom,

On Sun, Oct 18, 2009 at 12:57 PM, Tom <Tom.Rix@windriver.com> wrote:

> Nishanth Menon wrote:
>
>> This modification is NOT tested on any of the
>> platforms modified as I dont have them. please
>> help by testing+building+fixing
>>
>
> On the testing front..
> Regression testing on MAKEALL shows
>
> #warning Ethernet driver .. error killing for
> assabet, gcplus, lubbock, omap1510inn, omap1610h2,
> omap2420h4, omap3_sdp3430, omap5912osk and omap730p2
>
> There is a regression on B2.
>
> <snip>

> I am not in favor of including this patch in the 2009.11
> as we will likely apply the patch after waiting a reasonable
> amount of time on the runtime testing.  Which is not the
> same as knowing we aren't breaking the old boards.
>
> This is in no way a candidate for 2009.11.  As I mentioned in my initial
review, it will soon be queued up in net/next, which will be merged to
main/next soon.

regards,
Ben

[U-Boot] [PATCH 0/3] enable CONFIG_NET_MULTI for LAN91C96

[Menon, Nishanth]

> From: Ben Warren [mailto:biggerbadderben at gmail.com]
> Sent: Friday, October 16, 2009 11:30 AM
> 
> Nishanth Menon wrote:
> > Based on the discussion[1] here is the patchset
> >
> > Patchset is based on [2], and tested on SDP3430.
> > Other platforms would be prefered to be
> > build/tested/fixed as required
> >
> > Nishanth Menon (3):
> >   NET: LAN91C96 CONFIG_NET_MULTIify
> >   TI OMAP3: SDP3430 FIX NET_MULTI Warning
> >   LAN91C96: Enable NET_MULTI LAN driver
> >
> >
> On first pass review, this looks great!  There are a few things I'd like
> to do to improve the driver, but it has nothing to do with converting to
> NET_MULTI.
> 
> I'm going to fire up a net-next branch this weekend and will try to
> apply it there.
Did this get into the net-next branch? [1] did not seem to have it.

Regards,
Nishanth Menon
Ref:
[1] http://git.denx.de/?p=u-boot/u-boot-net.git;a=shortlog;h=refs/heads/next

[U-Boot] [PATCH 0/3] enable CONFIG_NET_MULTI for LAN91C96

[Ben Warren]

Menon, Nishanth wrote:
>> From: Ben Warren [mailto:biggerbadderben at gmail.com]
>> Sent: Friday, October 16, 2009 11:30 AM
>>
>> Nishanth Menon wrote:
>>     
>>> Based on the discussion[1] here is the patchset
>>>
>>> Patchset is based on [2], and tested on SDP3430.
>>> Other platforms would be prefered to be
>>> build/tested/fixed as required
>>>
>>> Nishanth Menon (3):
>>>   NET: LAN91C96 CONFIG_NET_MULTIify
>>>   TI OMAP3: SDP3430 FIX NET_MULTI Warning
>>>   LAN91C96: Enable NET_MULTI LAN driver
>>>
>>>
>>>       
>> On first pass review, this looks great!  There are a few things I'd like
>> to do to improve the driver, but it has nothing to do with converting to
>> NET_MULTI.
>>
>> I'm going to fire up a net-next branch this weekend and will try to
>> apply it there.
>>     
> Did this get into the net-next branch? [1] did not seem to have it.
>
>   
Unfortunately, for me 'this weekend' quite often leaves 'this' open to 
interpretation.  Bill Clinton would be proud.

Soon...

regards,
Ben

[U-Boot] [PATCH 1/3] NET: LAN91C96 CONFIG_NET_MULTIify

[Ben Warren]

Nishanth,

Nishanth Menon wrote:
> Make the lan91c96 driver capable of CONFIG_NET_MULTI
> to be clean for the new arch, add a a lil detect function
> Most of the formatting change was done to keep checkpatch
> silent, but a few functions and #if 0ed code which
> does not make sense for NET_MULTI have been removed
>
> Now, use the lan91c96_initialize() function to init the driver
>
> Signed-off-by: Nishanth Menon <nm@ti.com>
> Cc: Andrea Scian <andrea.scian@dave-tech.it>
> Cc: Ben Warren <biggerbadderben@gmail.com>
> Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
> Cc: George G. Davis <gdavis@mvista.com>
> Cc: Kyungmin Park <kyungmin.park@samsung.com>
> Cc: Nishant Kamat <nskamat@ti.com>
> Cc: Richard Woodruff <r-woodruff2@ti.com>
> Cc: Rishi Bhattacharya <rishi@ti.com>
> Cc: Sandeep Paulraj <s-paulraj@ti.com>
> ---
>  drivers/net/Makefile   |    2 +-
>  drivers/net/lan91c96.c |  452 +++++++++++++++++++----------------------------
>  drivers/net/lan91c96.h |  110 ++++++-------
>  include/netdev.h       |    1 +
>  4 files changed, 236 insertions(+), 329 deletions(-)
>   
applied to net/next

thanks,
Ben

[U-Boot] [PATCH 2/3] TI OMAP3: SDP3430 FIX NET_MULTI Warning

[Ben Warren]

Nishanth,

Nishanth Menon wrote:
> Enable the NET MULTI option and remove build warning
>
> Tested: SDP3430
>
> Signed-off-by: Nishanth Menon <nm@ti.com>
> Cc: Ben Warren <biggerbadderben@gmail.com>
> Cc: Sandeep Paulraj <s-paulraj@ti.com>
> ---
>  board/ti/sdp3430/sdp.c          |   12 +++++++-----
>  include/configs/omap3_sdp3430.h |    3 ++-
>  2 files changed, 9 insertions(+), 6 deletions(-)
>   
Applied to net/next

thanks,
Ben

[U-Boot] [PATCH 3/3] LAN91C96: Enable NET_MULTI LAN driver

[Ben Warren]

Nishanth,

Nishanth Menon wrote:
> This modification is NOT tested on any of the
> platforms modified as I dont have them. please
> help by testing+building+fixing
>
> Signed-off-by: Nishanth Menon <nm@ti.com>
> Cc: Andrea Scian <andrea.scian@dave-tech.it>
> Cc: Ben Warren <biggerbadderben@gmail.com>
> Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
> Cc: George G. Davis <gdavis@mvista.com>
> Cc: Kyungmin Park <kyungmin.park@samsung.com>
> Cc: Nishant Kamat <nskamat@ti.com>
> Cc: Richard Woodruff <r-woodruff2@ti.com>
> Cc: Rishi Bhattacharya <rishi@ti.com>
> Cc: Sandeep Paulraj <s-paulraj@ti.com>
> ---
>  board/apollon/apollon.c       |   12 +++++++-----
>  include/configs/B2.h          |    2 +-
>  include/configs/apollon.h     |    3 ++-
>  include/configs/assabet.h     |    3 ++-
>  include/configs/gcplus.h      |    3 ++-
>  include/configs/lubbock.h     |    3 ++-
>  include/configs/omap1510inn.h |    3 ++-
>  include/configs/omap1610h2.h  |    3 ++-
>  include/configs/omap1610inn.h |    3 ++-
>  include/configs/omap2420h4.h  |    3 ++-
>  include/configs/omap5912osk.h |    3 ++-
>  include/configs/omap730p2.h   |    3 ++-
>  include/configs/pleb2.h       |    6 ++++--
>  13 files changed, 32 insertions(+), 18 deletions(-)
>   
I expect that all of these boards are now broken, since you haven't 
provided board_eth_init() functions for them.

Applying to net/next anyway to make it easy to get everything working.

thanks,
Ben

[U-Boot] [PATCH 1/3] NET: LAN91C96 CONFIG_NET_MULTIify

[Wolfgang Denk]

Dear Nishanth Menon,

In message <1255669597-10607-2-git-send-email-nm@ti.com> you wrote:
> Make the lan91c96 driver capable of CONFIG_NET_MULTI
> to be clean for the new arch, add a a lil detect function
--------------^^^^^^^^^^^^^^^^------^^^^^^^

"for the new arch" is not really helpful for the commit message - for
example, I don't know what you are referring to here. Also, your
keyboard seems to have got stuck at the "add a a lil" part.

Can you please update the commit message?


Eventually Ben accepts just the new commit message and edits it when
applying the patch?

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
A memorandum is written not to inform the reader, but to protect  the
writer.                                               -- Dean Acheson

[U-Boot] [PATCH 1/3] NET: LAN91C96 CONFIG_NET_MULTIify

[Menon, Nishanth]

Wolfgang Denk said the following on 11/22/2009 05:01 PM:
> Dear Nishanth Menon,
>
> In message <1255669597-10607-2-git-send-email-nm@ti.com> you wrote:
>   
>> Make the lan91c96 driver capable of CONFIG_NET_MULTI
>> to be clean for the new arch, add a a lil detect function
>>     
> --------------^^^^^^^^^^^^^^^^------^^^^^^^
>
> "for the new arch" is not really helpful for the commit message - for
> example, I don't know what you are referring to here. Also, your
> keyboard seems to have got stuck at the "add a a lil" part.
>
> Can you please update the commit message?
>   
ok.. here it goes:

Make the lan91c96 driver capable of CONFIG_NET_MULTI
to be clean for the new network driver architecture,
add a little detect function

>
> Eventually Ben accepts just the new commit message and edits it when
> applying the patch?
>   
let me know if you want me to post again.
> Best regards,
>
> Wolfgang Denk
>
>   

[U-Boot] [PATCH 3/3] LAN91C96: Enable NET_MULTI LAN driver

[Ben Warren]

Nishanth,

Nishanth Menon wrote:
> This modification is NOT tested on any of the
> platforms modified as I dont have them. please
> help by testing+building+fixing
>
> Signed-off-by: Nishanth Menon <nm@ti.com>
> Cc: Andrea Scian <andrea.scian@dave-tech.it>
> Cc: Ben Warren <biggerbadderben@gmail.com>
> Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
> Cc: George G. Davis <gdavis@mvista.com>
> Cc: Kyungmin Park <kyungmin.park@samsung.com>
> Cc: Nishant Kamat <nskamat@ti.com>
> Cc: Richard Woodruff <r-woodruff2@ti.com>
> Cc: Rishi Bhattacharya <rishi@ti.com>
> Cc: Sandeep Paulraj <s-paulraj@ti.com>
> ---
>  board/apollon/apollon.c       |   12 +++++++-----
>  include/configs/B2.h          |    2 +-
>  include/configs/apollon.h     |    3 ++-
>  include/configs/assabet.h     |    3 ++-
>  include/configs/gcplus.h      |    3 ++-
>  include/configs/lubbock.h     |    3 ++-
>  include/configs/omap1510inn.h |    3 ++-
>  include/configs/omap1610h2.h  |    3 ++-
>  include/configs/omap1610inn.h |    3 ++-
>  include/configs/omap2420h4.h  |    3 ++-
>  include/configs/omap5912osk.h |    3 ++-
>  include/configs/omap730p2.h   |    3 ++-
>  include/configs/pleb2.h       |    6 ++++--
>  13 files changed, 32 insertions(+), 18 deletions(-)
>   
You've update the config files for a bunch of boards, but have not added 
board_eth_init() functions for them.  Am I missing something?  If this 
is indeed the case, I can't push these patches upstream from the net 
tree because the associated boards will break.

regards,
Ben

[U-Boot] [PATCH 3/3] LAN91C96: Enable NET_MULTI LAN driver

[Nishanth Menon]

Ben Warren had written, on 12/07/2009 03:09 PM, the following:
> Nishanth,
> 
> Nishanth Menon wrote:
>> This modification is NOT tested on any of the
>> platforms modified as I dont have them. please
>> help by testing+building+fixing
>>
>> Signed-off-by: Nishanth Menon <nm@ti.com>
>> Cc: Andrea Scian <andrea.scian@dave-tech.it>
>> Cc: Ben Warren <biggerbadderben@gmail.com>
>> Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
>> Cc: George G. Davis <gdavis@mvista.com>
>> Cc: Kyungmin Park <kyungmin.park@samsung.com>
>> Cc: Nishant Kamat <nskamat@ti.com>
>> Cc: Richard Woodruff <r-woodruff2@ti.com>
>> Cc: Rishi Bhattacharya <rishi@ti.com>
>> Cc: Sandeep Paulraj <s-paulraj@ti.com>
>> ---
>>  board/apollon/apollon.c       |   12 +++++++-----
>>  include/configs/B2.h          |    2 +-
>>  include/configs/apollon.h     |    3 ++-
>>  include/configs/assabet.h     |    3 ++-
>>  include/configs/gcplus.h      |    3 ++-
>>  include/configs/lubbock.h     |    3 ++-
>>  include/configs/omap1510inn.h |    3 ++-
>>  include/configs/omap1610h2.h  |    3 ++-
>>  include/configs/omap1610inn.h |    3 ++-
>>  include/configs/omap2420h4.h  |    3 ++-
>>  include/configs/omap5912osk.h |    3 ++-
>>  include/configs/omap730p2.h   |    3 ++-
>>  include/configs/pleb2.h       |    6 ++++--
>>  13 files changed, 32 insertions(+), 18 deletions(-)
>>   
> You've update the config files for a bunch of boards, but have not added 
> board_eth_init() functions for them.  Am I missing something?  If this 
> is indeed the case, I can't push these patches upstream from the net 
> tree because the associated boards will break.
I have added the init where they have been previously used for the old 
lan91c96 driver. I think the only one which was actively using it was 
apollon, which means the rest of the platforms probably dont use the old 
lan driver properly in the first place. I have maintained status quo 
there instead of introducing a change which may break them.

-- 
Regards,
Nishanth Menon

[U-Boot] [PATCH 3/3] LAN91C96: Enable NET_MULTI LAN driver

[Ben Warren]

<recipient list trimmed>

Nishanth,

Nishanth Menon wrote:
> Ben Warren had written, on 12/07/2009 03:09 PM, the following:
>> Nishanth,
>>
>> Nishanth Menon wrote:
>>> This modification is NOT tested on any of the
>>> platforms modified as I dont have them. please
>>> help by testing+building+fixing
>>>
>>> Signed-off-by: Nishanth Menon <nm@ti.com>
>>> Cc: Andrea Scian <andrea.scian@dave-tech.it>
>>> Cc: Ben Warren <biggerbadderben@gmail.com>
>>> Cc: Dave Peverley <dpeverley@mpc-data.co.uk>
>>> Cc: George G. Davis <gdavis@mvista.com>
>>> Cc: Kyungmin Park <kyungmin.park@samsung.com>
>>> Cc: Nishant Kamat <nskamat@ti.com>
>>> Cc: Richard Woodruff <r-woodruff2@ti.com>
>>> Cc: Rishi Bhattacharya <rishi@ti.com>
>>> Cc: Sandeep Paulraj <s-paulraj@ti.com>
>>> ---
>>>  board/apollon/apollon.c       |   12 +++++++-----
>>>  include/configs/B2.h          |    2 +-
>>>  include/configs/apollon.h     |    3 ++-
>>>  include/configs/assabet.h     |    3 ++-
>>>  include/configs/gcplus.h      |    3 ++-
>>>  include/configs/lubbock.h     |    3 ++-
>>>  include/configs/omap1510inn.h |    3 ++-
>>>  include/configs/omap1610h2.h  |    3 ++-
>>>  include/configs/omap1610inn.h |    3 ++-
>>>  include/configs/omap2420h4.h  |    3 ++-
>>>  include/configs/omap5912osk.h |    3 ++-
>>>  include/configs/omap730p2.h   |    3 ++-
>>>  include/configs/pleb2.h       |    6 ++++--
>>>  13 files changed, 32 insertions(+), 18 deletions(-)
>>>   
>> You've update the config files for a bunch of boards, but have not 
>> added board_eth_init() functions for them.  Am I missing something?  
>> If this is indeed the case, I can't push these patches upstream from 
>> the net tree because the associated boards will break.
> I have added the init where they have been previously used for the old 
> lan91c96 driver. I think the only one which was actively using it was 
> apollon, which means the rest of the platforms probably dont use the 
> old lan driver properly in the first place. I have maintained status 
> quo there instead of introducing a change which may break them.
>
I understand where you're coming from, but having the CONFIG option in 
the board's header file indicates that this board probably has a LAN 
chip.  Without board_eth_init() support, the CONFIG option is 
pointless.  I'd prefer to add full support, then if the user doesn't 
want it they can #undef in the header.

If you don't have time to do this, I can take care of it.

regards,
Ben