diff options
Diffstat (limited to 'arch/arm/cpu/ixp/npe/npe.c')
| -rw-r--r-- | arch/arm/cpu/ixp/npe/npe.c | 676 |
1 files changed, 676 insertions, 0 deletions
diff --git a/arch/arm/cpu/ixp/npe/npe.c b/arch/arm/cpu/ixp/npe/npe.c new file mode 100644 index 0000000..2e68689 --- /dev/null +++ b/arch/arm/cpu/ixp/npe/npe.c @@ -0,0 +1,676 @@ +/* + * (C) Copyright 2005-2006 + * Stefan Roese, DENX Software Engineering, sr@denx.de. + * + * See file CREDITS for list of people who contributed to this + * project. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation; either version 2 of + * the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, + * MA 02111-1307 USA + */ + +#if 0 +#define DEBUG /* define for debug output */ +#endif + +#include <config.h> +#include <common.h> +#include <net.h> +#include <miiphy.h> +#include <malloc.h> +#include <asm/processor.h> +#include <asm/arch-ixp/ixp425.h> + +#include <IxOsal.h> +#include <IxEthAcc.h> +#include <IxEthDB.h> +#include <IxNpeDl.h> +#include <IxQMgr.h> +#include <IxNpeMh.h> +#include <ix_ossl.h> +#include <IxFeatureCtrl.h> + +#include <npe.h> + +static IxQMgrDispatcherFuncPtr qDispatcherFunc = NULL; +static int npe_exists[NPE_NUM_PORTS]; +static int npe_used[NPE_NUM_PORTS]; + +/* A little extra so we can align to cacheline. */ +static u8 npe_alloc_pool[NPE_MEM_POOL_SIZE + CONFIG_SYS_CACHELINE_SIZE - 1]; +static u8 *npe_alloc_end; +static u8 *npe_alloc_free; + +static void *npe_alloc(int size) +{ + static int count = 0; + void *p = NULL; + + size = (size + (CONFIG_SYS_CACHELINE_SIZE-1)) & ~(CONFIG_SYS_CACHELINE_SIZE-1); + count++; + + if ((npe_alloc_free + size) < npe_alloc_end) { + p = npe_alloc_free; + npe_alloc_free += size; + } else { + printf("npe_alloc: failed (count=%d, size=%d)!\n", count, size); + } + return p; +} + +/* Not interrupt safe! */ +static void mbuf_enqueue(IX_OSAL_MBUF **q, IX_OSAL_MBUF *new) +{ + IX_OSAL_MBUF *m = *q; + + IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR(new) = NULL; + + if (m) { + while(IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR(m)) + m = IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR(m); + IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR(m) = new; + } else + *q = new; +} + +/* Not interrupt safe! */ +static IX_OSAL_MBUF *mbuf_dequeue(IX_OSAL_MBUF **q) +{ + IX_OSAL_MBUF *m = *q; + if (m) + *q = IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR(m); + return m; +} + +static void reset_tx_mbufs(struct npe* p_npe) +{ + IX_OSAL_MBUF *m; + int i; + + p_npe->txQHead = NULL; + + for (i = 0; i < CONFIG_DEVS_ETH_INTEL_NPE_MAX_TX_DESCRIPTORS; i++) { + m = &p_npe->tx_mbufs[i]; + + memset(m, 0, sizeof(*m)); + + IX_OSAL_MBUF_MDATA(m) = (void *)&p_npe->tx_pkts[i * NPE_PKT_SIZE]; + IX_OSAL_MBUF_MLEN(m) = IX_OSAL_MBUF_PKT_LEN(m) = NPE_PKT_SIZE; + mbuf_enqueue(&p_npe->txQHead, m); + } +} + +static void reset_rx_mbufs(struct npe* p_npe) +{ + IX_OSAL_MBUF *m; + int i; + + p_npe->rxQHead = NULL; + + HAL_DCACHE_INVALIDATE(p_npe->rx_pkts, NPE_PKT_SIZE * + CONFIG_DEVS_ETH_INTEL_NPE_MAX_RX_DESCRIPTORS); + + for (i = 0; i < CONFIG_DEVS_ETH_INTEL_NPE_MAX_RX_DESCRIPTORS; i++) { + m = &p_npe->rx_mbufs[i]; + + memset(m, 0, sizeof(*m)); + + IX_OSAL_MBUF_MDATA(m) = (void *)&p_npe->rx_pkts[i * NPE_PKT_SIZE]; + IX_OSAL_MBUF_MLEN(m) = IX_OSAL_MBUF_PKT_LEN(m) = NPE_PKT_SIZE; + + if(ixEthAccPortRxFreeReplenish(p_npe->eth_id, m) != IX_SUCCESS) { + printf("ixEthAccPortRxFreeReplenish failed for port %d\n", p_npe->eth_id); + break; + } + } +} + +static void init_rx_mbufs(struct npe* p_npe) +{ + p_npe->rxQHead = NULL; + + p_npe->rx_pkts = npe_alloc(NPE_PKT_SIZE * + CONFIG_DEVS_ETH_INTEL_NPE_MAX_RX_DESCRIPTORS); + if (p_npe->rx_pkts == NULL) { + printf("alloc of packets failed.\n"); + return; + } + + p_npe->rx_mbufs = (IX_OSAL_MBUF *) + npe_alloc(sizeof(IX_OSAL_MBUF) * + CONFIG_DEVS_ETH_INTEL_NPE_MAX_RX_DESCRIPTORS); + if (p_npe->rx_mbufs == NULL) { + printf("alloc of mbufs failed.\n"); + return; + } + + reset_rx_mbufs(p_npe); +} + +static void init_tx_mbufs(struct npe* p_npe) +{ + p_npe->tx_pkts = npe_alloc(NPE_PKT_SIZE * + CONFIG_DEVS_ETH_INTEL_NPE_MAX_TX_DESCRIPTORS); + if (p_npe->tx_pkts == NULL) { + printf("alloc of packets failed.\n"); + return; + } + + p_npe->tx_mbufs = (IX_OSAL_MBUF *) + npe_alloc(sizeof(IX_OSAL_MBUF) * + CONFIG_DEVS_ETH_INTEL_NPE_MAX_TX_DESCRIPTORS); + if (p_npe->tx_mbufs == NULL) { + printf("alloc of mbufs failed.\n"); + return; + } + + reset_tx_mbufs(p_npe); +} + +/* Convert IX_ETH_PORT_n to IX_NPEMH_NPEID_NPEx */ +static int __eth_to_npe(int eth_id) +{ + switch(eth_id) { + case IX_ETH_PORT_1: + return IX_NPEMH_NPEID_NPEB; + + case IX_ETH_PORT_2: + return IX_NPEMH_NPEID_NPEC; + + case IX_ETH_PORT_3: + return IX_NPEMH_NPEID_NPEA; + } + return 0; +} + +/* Poll the CSR machinery. */ +static void npe_poll(int eth_id) +{ + if (qDispatcherFunc != NULL) { + ixNpeMhMessagesReceive(__eth_to_npe(eth_id)); + (*qDispatcherFunc)(IX_QMGR_QUELOW_GROUP); + } +} + +/* ethAcc RX callback */ +static void npe_rx_callback(u32 cbTag, IX_OSAL_MBUF *m, IxEthAccPortId portid) +{ + struct npe* p_npe = (struct npe *)cbTag; + + if (IX_OSAL_MBUF_MLEN(m) > 0) { + mbuf_enqueue(&p_npe->rxQHead, m); + + if (p_npe->rx_write == ((p_npe->rx_read-1) & (PKTBUFSRX-1))) { + debug("Rx overflow: rx_write=%d rx_read=%d\n", + p_npe->rx_write, p_npe->rx_read); + } else { + debug("Received message #%d (len=%d)\n", p_npe->rx_write, + IX_OSAL_MBUF_MLEN(m)); + memcpy((void *)NetRxPackets[p_npe->rx_write], IX_OSAL_MBUF_MDATA(m), + IX_OSAL_MBUF_MLEN(m)); + p_npe->rx_len[p_npe->rx_write] = IX_OSAL_MBUF_MLEN(m); + p_npe->rx_write++; + if (p_npe->rx_write == PKTBUFSRX) + p_npe->rx_write = 0; + +#ifdef CONFIG_PRINT_RX_FRAMES + { + u8 *ptr = IX_OSAL_MBUF_MDATA(m); + int i; + + for (i=0; i<60; i++) { + debug("%02x ", *ptr++); + } + debug("\n"); + } +#endif + } + + m = mbuf_dequeue(&p_npe->rxQHead); + } else { + debug("Received frame with length 0!!!\n"); + m = mbuf_dequeue(&p_npe->rxQHead); + } + + /* Now return mbuf to NPE */ + IX_OSAL_MBUF_MLEN(m) = IX_OSAL_MBUF_PKT_LEN(m) = NPE_PKT_SIZE; + IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(m) = NULL; + IX_OSAL_MBUF_FLAGS(m) = 0; + + if(ixEthAccPortRxFreeReplenish(p_npe->eth_id, m) != IX_SUCCESS) { + debug("npe_rx_callback: Error returning mbuf.\n"); + } +} + +/* ethAcc TX callback */ +static void npe_tx_callback(u32 cbTag, IX_OSAL_MBUF *m) +{ + struct npe* p_npe = (struct npe *)cbTag; + + debug("%s\n", __FUNCTION__); + + IX_OSAL_MBUF_MLEN(m) = IX_OSAL_MBUF_PKT_LEN(m) = NPE_PKT_SIZE; + IX_OSAL_MBUF_NEXT_BUFFER_IN_PKT_PTR(m) = NULL; + IX_OSAL_MBUF_FLAGS(m) = 0; + + mbuf_enqueue(&p_npe->txQHead, m); +} + + +static int npe_set_mac_address(struct eth_device *dev) +{ + struct npe *p_npe = (struct npe *)dev->priv; + IxEthAccMacAddr npeMac; + + debug("%s\n", __FUNCTION__); + + /* Set MAC address */ + memcpy(npeMac.macAddress, dev->enetaddr, 6); + + if (ixEthAccPortUnicastMacAddressSet(p_npe->eth_id, &npeMac) != IX_ETH_ACC_SUCCESS) { + printf("Error setting unicast address! %02x:%02x:%02x:%02x:%02x:%02x\n", + npeMac.macAddress[0], npeMac.macAddress[1], + npeMac.macAddress[2], npeMac.macAddress[3], + npeMac.macAddress[4], npeMac.macAddress[5]); + return 0; + } + + return 1; +} + +/* Boot-time CSR library initialization. */ +static int npe_csr_load(void) +{ + int i; + + if (ixQMgrInit() != IX_SUCCESS) { + debug("Error initialising queue manager!\n"); + return 0; + } + + ixQMgrDispatcherLoopGet(&qDispatcherFunc); + + if(ixNpeMhInitialize(IX_NPEMH_NPEINTERRUPTS_YES) != IX_SUCCESS) { + printf("Error initialising NPE Message handler!\n"); + return 0; + } + + if (npe_used[IX_ETH_PORT_1] && npe_exists[IX_ETH_PORT_1] && + ixNpeDlNpeInitAndStart(IX_NPEDL_NPEIMAGE_NPEB_ETH_LEARN_FILTER_SPAN_FIREWALL_VLAN_QOS) + != IX_SUCCESS) { + printf("Error downloading firmware to NPE-B!\n"); + return 0; + } + + if (npe_used[IX_ETH_PORT_2] && npe_exists[IX_ETH_PORT_2] && + ixNpeDlNpeInitAndStart(IX_NPEDL_NPEIMAGE_NPEC_ETH_LEARN_FILTER_SPAN_FIREWALL_VLAN_QOS) + != IX_SUCCESS) { + printf("Error downloading firmware to NPE-C!\n"); + return 0; + } + + /* don't need this for U-Boot */ + ixFeatureCtrlSwConfigurationWrite(IX_FEATURECTRL_ETH_LEARNING, FALSE); + + if (ixEthAccInit() != IX_ETH_ACC_SUCCESS) { + printf("Error initialising Ethernet access driver!\n"); + return 0; + } + + for (i = 0; i < IX_ETH_ACC_NUMBER_OF_PORTS; i++) { + if (!npe_used[i] || !npe_exists[i]) + continue; + if (ixEthAccPortInit(i) != IX_ETH_ACC_SUCCESS) { + printf("Error initialising Ethernet port%d!\n", i); + } + if (ixEthAccTxSchedulingDisciplineSet(i, FIFO_NO_PRIORITY) != IX_ETH_ACC_SUCCESS) { + printf("Error setting scheduling discipline for port %d.\n", i); + } + if (ixEthAccPortRxFrameAppendFCSDisable(i) != IX_ETH_ACC_SUCCESS) { + printf("Error disabling RX FCS for port %d.\n", i); + } + if (ixEthAccPortTxFrameAppendFCSEnable(i) != IX_ETH_ACC_SUCCESS) { + printf("Error enabling TX FCS for port %d.\n", i); + } + } + + return 1; +} + +static int npe_init(struct eth_device *dev, bd_t * bis) +{ + struct npe *p_npe = (struct npe *)dev->priv; + int i; + u16 reg_short; + int speed; + int duplex; + + debug("%s: 1\n", __FUNCTION__); + + miiphy_read (dev->name, p_npe->phy_no, PHY_BMSR, ®_short); + + /* + * Wait if PHY is capable of autonegotiation and autonegotiation is not complete + */ + if ((reg_short & PHY_BMSR_AUTN_ABLE) && !(reg_short & PHY_BMSR_AUTN_COMP)) { + puts ("Waiting for PHY auto negotiation to complete"); + i = 0; + while (!(reg_short & PHY_BMSR_AUTN_COMP)) { + /* + * Timeout reached ? + */ + if (i > PHY_AUTONEGOTIATE_TIMEOUT) { + puts (" TIMEOUT !\n"); + break; + } + + if ((i++ % 1000) == 0) { + putc ('.'); + miiphy_read (dev->name, p_npe->phy_no, PHY_BMSR, ®_short); + } + udelay (1000); /* 1 ms */ + } + puts (" done\n"); + udelay (500000); /* another 500 ms (results in faster booting) */ + } + + speed = miiphy_speed (dev->name, p_npe->phy_no); + duplex = miiphy_duplex (dev->name, p_npe->phy_no); + + if (p_npe->print_speed) { + p_npe->print_speed = 0; + printf ("ENET Speed is %d Mbps - %s duplex connection\n", + (int) speed, (duplex == HALF) ? "HALF" : "FULL"); + } + + npe_alloc_end = npe_alloc_pool + sizeof(npe_alloc_pool); + npe_alloc_free = (u8 *)(((unsigned)npe_alloc_pool + + CONFIG_SYS_CACHELINE_SIZE - 1) & ~(CONFIG_SYS_CACHELINE_SIZE - 1)); + + /* initialize mbuf pool */ + init_rx_mbufs(p_npe); + init_tx_mbufs(p_npe); + + if (ixEthAccPortRxCallbackRegister(p_npe->eth_id, npe_rx_callback, + (u32)p_npe) != IX_ETH_ACC_SUCCESS) { + printf("can't register RX callback!\n"); + return -1; + } + + if (ixEthAccPortTxDoneCallbackRegister(p_npe->eth_id, npe_tx_callback, + (u32)p_npe) != IX_ETH_ACC_SUCCESS) { + printf("can't register TX callback!\n"); + return -1; + } + + npe_set_mac_address(dev); + + if (ixEthAccPortEnable(p_npe->eth_id) != IX_ETH_ACC_SUCCESS) { + printf("can't enable port!\n"); + return -1; + } + + p_npe->active = 1; + + return 0; +} + +#if 0 /* test-only: probably have to deal with it when booting linux (for a clean state) */ +/* Uninitialize CSR library. */ +static void npe_csr_unload(void) +{ + ixEthAccUnload(); + ixEthDBUnload(); + ixNpeMhUnload(); + ixQMgrUnload(); +} + +/* callback which is used by ethAcc to recover RX buffers when stopping */ +static void npe_rx_stop_callback(u32 cbTag, IX_OSAL_MBUF *m, IxEthAccPortId portid) +{ + debug("%s\n", __FUNCTION__); +} + +/* callback which is used by ethAcc to recover TX buffers when stopping */ +static void npe_tx_stop_callback(u32 cbTag, IX_OSAL_MBUF *m) +{ + debug("%s\n", __FUNCTION__); +} +#endif + +static void npe_halt(struct eth_device *dev) +{ + struct npe *p_npe = (struct npe *)dev->priv; + int i; + + debug("%s\n", __FUNCTION__); + + /* Delay to give time for recovery of mbufs */ + for (i = 0; i < 100; i++) { + npe_poll(p_npe->eth_id); + udelay(100); + } + +#if 0 /* test-only: probably have to deal with it when booting linux (for a clean state) */ + if (ixEthAccPortRxCallbackRegister(p_npe->eth_id, npe_rx_stop_callback, + (u32)p_npe) != IX_ETH_ACC_SUCCESS) { + debug("Error registering rx callback!\n"); + } + + if (ixEthAccPortTxDoneCallbackRegister(p_npe->eth_id, npe_tx_stop_callback, + (u32)p_npe) != IX_ETH_ACC_SUCCESS) { + debug("Error registering tx callback!\n"); + } + + if (ixEthAccPortDisable(p_npe->eth_id) != IX_ETH_ACC_SUCCESS) { + debug("npe_stop: Error disabling NPEB!\n"); + } + + /* Delay to give time for recovery of mbufs */ + for (i = 0; i < 100; i++) { + npe_poll(p_npe->eth_id); + udelay(10000); + } + + /* + * For U-Boot only, we are probably launching Linux or other OS that + * needs a clean slate for its NPE library. + */ +#if 0 /* test-only */ + for (i = 0; i < IX_ETH_ACC_NUMBER_OF_PORTS; i++) { + if (npe_used[i] && npe_exists[i]) + if (ixNpeDlNpeStopAndReset(__eth_to_npe(i)) != IX_SUCCESS) + printf("Failed to stop and reset NPE B.\n"); + } +#endif + +#endif + p_npe->active = 0; +} + + +static int npe_send(struct eth_device *dev, volatile void *packet, int len) +{ + struct npe *p_npe = (struct npe *)dev->priv; + u8 *dest; + int err; + IX_OSAL_MBUF *m; + + debug("%s\n", __FUNCTION__); + m = mbuf_dequeue(&p_npe->txQHead); + dest = IX_OSAL_MBUF_MDATA(m); + IX_OSAL_MBUF_PKT_LEN(m) = IX_OSAL_MBUF_MLEN(m) = len; + IX_OSAL_MBUF_NEXT_PKT_IN_CHAIN_PTR(m) = NULL; + + memcpy(dest, (char *)packet, len); + + if ((err = ixEthAccPortTxFrameSubmit(p_npe->eth_id, m, IX_ETH_ACC_TX_DEFAULT_PRIORITY)) + != IX_ETH_ACC_SUCCESS) { + printf("npe_send: Can't submit frame. err[%d]\n", err); + mbuf_enqueue(&p_npe->txQHead, m); + return 0; + } + +#ifdef DEBUG_PRINT_TX_FRAMES + { + u8 *ptr = IX_OSAL_MBUF_MDATA(m); + int i; + + for (i=0; i<IX_OSAL_MBUF_MLEN(m); i++) { + printf("%02x ", *ptr++); + } + printf(" (tx-len=%d)\n", IX_OSAL_MBUF_MLEN(m)); + } +#endif + + npe_poll(p_npe->eth_id); + + return len; +} + +static int npe_rx(struct eth_device *dev) +{ + struct npe *p_npe = (struct npe *)dev->priv; + + debug("%s\n", __FUNCTION__); + npe_poll(p_npe->eth_id); + + debug("%s: rx_write=%d rx_read=%d\n", __FUNCTION__, p_npe->rx_write, p_npe->rx_read); + while (p_npe->rx_write != p_npe->rx_read) { + debug("Reading message #%d\n", p_npe->rx_read); + NetReceive(NetRxPackets[p_npe->rx_read], p_npe->rx_len[p_npe->rx_read]); + p_npe->rx_read++; + if (p_npe->rx_read == PKTBUFSRX) + p_npe->rx_read = 0; + } + + return 0; +} + +int npe_initialize(bd_t * bis) +{ + static int virgin = 0; + struct eth_device *dev; + int eth_num = 0; + struct npe *p_npe = NULL; + uchar enetaddr[6]; + + for (eth_num = 0; eth_num < CONFIG_SYS_NPE_NUMS; eth_num++) { + + /* See if we can actually bring up the interface, otherwise, skip it */ +#ifdef CONFIG_HAS_ETH1 + if (eth_num == 1) { + if (!eth_getenv_enetaddr("eth1addr", enetaddr)) + continue; + } else +#endif + if (!eth_getenv_enetaddr("ethaddr", enetaddr)) + continue; + + /* Allocate device structure */ + dev = (struct eth_device *)malloc(sizeof(*dev)); + if (dev == NULL) { + printf ("%s: Cannot allocate eth_device %d\n", __FUNCTION__, eth_num); + return -1; + } + memset(dev, 0, sizeof(*dev)); + + /* Allocate our private use data */ + p_npe = (struct npe *)malloc(sizeof(struct npe)); + if (p_npe == NULL) { + printf("%s: Cannot allocate private hw data for eth_device %d", + __FUNCTION__, eth_num); + free(dev); + return -1; + } + memset(p_npe, 0, sizeof(struct npe)); + + p_npe->eth_id = eth_num; + memcpy(dev->enetaddr, enetaddr, 6); +#ifdef CONFIG_HAS_ETH1 + if (eth_num == 1) + p_npe->phy_no = CONFIG_PHY1_ADDR; + else +#endif + p_npe->phy_no = CONFIG_PHY_ADDR; + + sprintf(dev->name, "NPE%d", eth_num); + dev->priv = (void *)p_npe; + dev->init = npe_init; + dev->halt = npe_halt; + dev->send = npe_send; + dev->recv = npe_rx; + + p_npe->print_speed = 1; + + if (0 == virgin) { + virgin = 1; + + if (ixFeatureCtrlDeviceRead() == IX_FEATURE_CTRL_DEVICE_TYPE_IXP42X) { + switch (ixFeatureCtrlProductIdRead() & IX_FEATURE_CTRL_SILICON_STEPPING_MASK) { + case IX_FEATURE_CTRL_SILICON_TYPE_B0: + /* + * If it is B0 Silicon, we only enable port when its corresponding + * Eth Coprocessor is available. + */ + if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH0) == + IX_FEATURE_CTRL_COMPONENT_ENABLED) + npe_exists[IX_ETH_PORT_1] = TRUE; + + if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH1) == + IX_FEATURE_CTRL_COMPONENT_ENABLED) + npe_exists[IX_ETH_PORT_2] = TRUE; + break; + case IX_FEATURE_CTRL_SILICON_TYPE_A0: + /* + * If it is A0 Silicon, we enable both as both Eth Coprocessors + * are available. + */ + npe_exists[IX_ETH_PORT_1] = TRUE; + npe_exists[IX_ETH_PORT_2] = TRUE; + break; + } + } else if (ixFeatureCtrlDeviceRead() == IX_FEATURE_CTRL_DEVICE_TYPE_IXP46X) { + if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH0) == + IX_FEATURE_CTRL_COMPONENT_ENABLED) + npe_exists[IX_ETH_PORT_1] = TRUE; + + if (ixFeatureCtrlComponentCheck(IX_FEATURECTRL_ETH1) == + IX_FEATURE_CTRL_COMPONENT_ENABLED) + npe_exists[IX_ETH_PORT_2] = TRUE; + } + + npe_used[IX_ETH_PORT_1] = 1; + npe_used[IX_ETH_PORT_2] = 1; + + npe_alloc_end = npe_alloc_pool + sizeof(npe_alloc_pool); + npe_alloc_free = (u8 *)(((unsigned)npe_alloc_pool + + CONFIG_SYS_CACHELINE_SIZE - 1) + & ~(CONFIG_SYS_CACHELINE_SIZE - 1)); + + if (!npe_csr_load()) + return 0; + } + + eth_register(dev); + +#if defined(CONFIG_MII) || defined(CONFIG_CMD_MII) + miiphy_register(dev->name, npe_miiphy_read, npe_miiphy_write); +#endif + + } /* end for each supported device */ + + return 1; +} |