/* * Ethernet on Serial Communications Controller (SCC) driver for Motorola MPC8xx and MPC82xx. * * Copyright (c) 2003 Intracom S.A. * by Pantelis Antoniou <panto@intracom.gr> * * 2005 (c) MontaVista Software, Inc. * Vitaly Bordug <vbordug@ru.mvista.com> * * This file is licensed under the terms of the GNU General Public License * version 2. This program is licensed "as is" without any warranty of any * kind, whether express or implied. */ #include <linux/module.h> #include <linux/kernel.h> #include <linux/types.h> #include <linux/string.h> #include <linux/ptrace.h> #include <linux/errno.h> #include <linux/ioport.h> #include <linux/interrupt.h> #include <linux/delay.h> #include <linux/netdevice.h> #include <linux/etherdevice.h> #include <linux/skbuff.h> #include <linux/spinlock.h> #include <linux/mii.h> #include <linux/ethtool.h> #include <linux/bitops.h> #include <linux/fs.h> #include <linux/platform_device.h> #include <linux/of_address.h> #include <linux/of_irq.h> #include <asm/irq.h> #include <linux/uaccess.h> #include "fs_enet.h" /*************************************************/ #if defined(CONFIG_CPM1) /* for a 8xx __raw_xxx's are sufficient */ #define __fs_out32(addr, x) __raw_writel(x, addr) #define __fs_out16(addr, x) __raw_writew(x, addr) #define __fs_out8(addr, x) __raw_writeb(x, addr) #define __fs_in32(addr) __raw_readl(addr) #define __fs_in16(addr) __raw_readw(addr) #define __fs_in8(addr) __raw_readb(addr) #else /* for others play it safe */ #define __fs_out32(addr, x) out_be32(addr, x) #define __fs_out16(addr, x) out_be16(addr, x) #define __fs_in32(addr) in_be32(addr) #define __fs_in16(addr) in_be16(addr) #define __fs_out8(addr, x) out_8(addr, x) #define __fs_in8(addr) in_8(addr) #endif /* write, read, set bits, clear bits */ #define W32(_p, _m, _v) __fs_out32(&(_p)->_m, (_v)) #define R32(_p, _m) __fs_in32(&(_p)->_m) #define S32(_p, _m, _v) W32(_p, _m, R32(_p, _m) | (_v)) #define C32(_p, _m, _v) W32(_p, _m, R32(_p, _m) & ~(_v)) #define W16(_p, _m, _v) __fs_out16(&(_p)->_m, (_v)) #define R16(_p, _m) __fs_in16(&(_p)->_m) #define S16(_p, _m, _v) W16(_p, _m, R16(_p, _m) | (_v)) #define C16(_p, _m, _v) W16(_p, _m, R16(_p, _m) & ~(_v)) #define W8(_p, _m, _v) __fs_out8(&(_p)->_m, (_v)) #define R8(_p, _m) __fs_in8(&(_p)->_m) #define S8(_p, _m, _v) W8(_p, _m, R8(_p, _m) | (_v)) #define C8(_p, _m, _v) W8(_p, _m, R8(_p, _m) & ~(_v)) #define SCC_MAX_MULTICAST_ADDRS 64 /* * Delay to wait for SCC reset command to complete (in us) */ #define SCC_RESET_DELAY 50 static inline int scc_cr_cmd(struct fs_enet_private *fep, u32 op) { const struct fs_platform_info *fpi = fep->fpi; return cpm_command(fpi->cp_command, op); } static int do_pd_setup(struct fs_enet_private *fep) { struct platform_device *ofdev = to_platform_device(fep->dev); fep->interrupt = irq_of_parse_and_map(ofdev->dev.of_node, 0); if (!fep->interrupt) return -EINVAL; fep->scc.sccp = of_iomap(ofdev->dev.of_node, 0); if (!fep->scc.sccp) return -EINVAL; fep->scc.ep = of_iomap(ofdev->dev.of_node, 1); if (!fep->scc.ep) { iounmap(fep->scc.sccp); return -EINVAL; } return 0; } #define SCC_NAPI_EVENT_MSK (SCCE_ENET_RXF | SCCE_ENET_RXB | SCCE_ENET_TXB) #define SCC_EVENT (SCCE_ENET_RXF | SCCE_ENET_TXB) #define SCC_ERR_EVENT_MSK (SCCE_ENET_TXE | SCCE_ENET_BSY) static int setup_data(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); do_pd_setup(fep); fep->scc.hthi = 0; fep->scc.htlo = 0; fep->ev_napi = SCC_NAPI_EVENT_MSK; fep->ev = SCC_EVENT | SCCE_ENET_TXE; fep->ev_err = SCC_ERR_EVENT_MSK; return 0; } static int allocate_bd(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); const struct fs_platform_info *fpi = fep->fpi; fep->ring_mem_addr = cpm_muram_alloc((fpi->tx_ring + fpi->rx_ring) * sizeof(cbd_t), 8); if (IS_ERR_VALUE(fep->ring_mem_addr)) return -ENOMEM; fep->ring_base = (void __iomem __force*) cpm_muram_addr(fep->ring_mem_addr); return 0; } static void free_bd(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); if (fep->ring_base) cpm_muram_free(fep->ring_mem_addr); } static void cleanup_data(struct net_device *dev) { /* nothing */ } static void set_promiscuous_mode(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; S16(sccp, scc_psmr, SCC_PSMR_PRO); } static void set_multicast_start(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_enet_t __iomem *ep = fep->scc.ep; W16(ep, sen_gaddr1, 0); W16(ep, sen_gaddr2, 0); W16(ep, sen_gaddr3, 0); W16(ep, sen_gaddr4, 0); } static void set_multicast_one(struct net_device *dev, const u8 * mac) { struct fs_enet_private *fep = netdev_priv(dev); scc_enet_t __iomem *ep = fep->scc.ep; u16 taddrh, taddrm, taddrl; taddrh = ((u16) mac[5] << 8) | mac[4]; taddrm = ((u16) mac[3] << 8) | mac[2]; taddrl = ((u16) mac[1] << 8) | mac[0]; W16(ep, sen_taddrh, taddrh); W16(ep, sen_taddrm, taddrm); W16(ep, sen_taddrl, taddrl); scc_cr_cmd(fep, CPM_CR_SET_GADDR); } static void set_multicast_finish(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; scc_enet_t __iomem *ep = fep->scc.ep; /* clear promiscuous always */ C16(sccp, scc_psmr, SCC_PSMR_PRO); /* if all multi or too many multicasts; just enable all */ if ((dev->flags & IFF_ALLMULTI) != 0 || netdev_mc_count(dev) > SCC_MAX_MULTICAST_ADDRS) { W16(ep, sen_gaddr1, 0xffff); W16(ep, sen_gaddr2, 0xffff); W16(ep, sen_gaddr3, 0xffff); W16(ep, sen_gaddr4, 0xffff); } } static void set_multicast_list(struct net_device *dev) { struct netdev_hw_addr *ha; if ((dev->flags & IFF_PROMISC) == 0) { set_multicast_start(dev); netdev_for_each_mc_addr(ha, dev) set_multicast_one(dev, ha->addr); set_multicast_finish(dev); } else set_promiscuous_mode(dev); } /* * This function is called to start or restart the FEC during a link * change. This only happens when switching between half and full * duplex. */ static void restart(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; scc_enet_t __iomem *ep = fep->scc.ep; const struct fs_platform_info *fpi = fep->fpi; u16 paddrh, paddrm, paddrl; const unsigned char *mac; int i; C32(sccp, scc_gsmrl, SCC_GSMRL_ENR | SCC_GSMRL_ENT); /* clear everything (slow & steady does it) */ for (i = 0; i < sizeof(*ep); i++) __fs_out8((u8 __iomem *)ep + i, 0); /* point to bds */ W16(ep, sen_genscc.scc_rbase, fep->ring_mem_addr); W16(ep, sen_genscc.scc_tbase, fep->ring_mem_addr + sizeof(cbd_t) * fpi->rx_ring); /* Initialize function code registers for big-endian. */ #ifndef CONFIG_NOT_COHERENT_CACHE W8(ep, sen_genscc.scc_rfcr, SCC_EB | SCC_GBL); W8(ep, sen_genscc.scc_tfcr, SCC_EB | SCC_GBL); #else W8(ep, sen_genscc.scc_rfcr, SCC_EB); W8(ep, sen_genscc.scc_tfcr, SCC_EB); #endif /* Set maximum bytes per receive buffer. * This appears to be an Ethernet frame size, not the buffer * fragment size. It must be a multiple of four. */ W16(ep, sen_genscc.scc_mrblr, 0x5f0); /* Set CRC preset and mask. */ W32(ep, sen_cpres, 0xffffffff); W32(ep, sen_cmask, 0xdebb20e3); W32(ep, sen_crcec, 0); /* CRC Error counter */ W32(ep, sen_alec, 0); /* alignment error counter */ W32(ep, sen_disfc, 0); /* discard frame counter */ W16(ep, sen_pads, 0x8888); /* Tx short frame pad character */ W16(ep, sen_retlim, 15); /* Retry limit threshold */ W16(ep, sen_maxflr, 0x5ee); /* maximum frame length register */ W16(ep, sen_minflr, PKT_MINBUF_SIZE); /* minimum frame length register */ W16(ep, sen_maxd1, 0x000005f0); /* maximum DMA1 length */ W16(ep, sen_maxd2, 0x000005f0); /* maximum DMA2 length */ /* Clear hash tables. */ W16(ep, sen_gaddr1, 0); W16(ep, sen_gaddr2, 0); W16(ep, sen_gaddr3, 0); W16(ep, sen_gaddr4, 0); W16(ep, sen_iaddr1, 0); W16(ep, sen_iaddr2, 0); W16(ep, sen_iaddr3, 0); W16(ep, sen_iaddr4, 0); /* set address */ mac = dev->dev_addr; paddrh = ((u16) mac[5] << 8) | mac[4]; paddrm = ((u16) mac[3] << 8) | mac[2]; paddrl = ((u16) mac[1] << 8) | mac[0]; W16(ep, sen_paddrh, paddrh); W16(ep, sen_paddrm, paddrm); W16(ep, sen_paddrl, paddrl); W16(ep, sen_pper, 0); W16(ep, sen_taddrl, 0); W16(ep, sen_taddrm, 0); W16(ep, sen_taddrh, 0); fs_init_bds(dev); scc_cr_cmd(fep, CPM_CR_INIT_TRX); W16(sccp, scc_scce, 0xffff); /* Enable interrupts we wish to service. */ W16(sccp, scc_sccm, SCCE_ENET_TXE | SCCE_ENET_RXF | SCCE_ENET_TXB); /* Set GSMR_H to enable all normal operating modes. * Set GSMR_L to enable Ethernet to MC68160. */ W32(sccp, scc_gsmrh, 0); W32(sccp, scc_gsmrl, SCC_GSMRL_TCI | SCC_GSMRL_TPL_48 | SCC_GSMRL_TPP_10 | SCC_GSMRL_MODE_ENET); /* Set sync/delimiters. */ W16(sccp, scc_dsr, 0xd555); /* Set processing mode. Use Ethernet CRC, catch broadcast, and * start frame search 22 bit times after RENA. */ W16(sccp, scc_psmr, SCC_PSMR_ENCRC | SCC_PSMR_NIB22); /* Set full duplex mode if needed */ if (dev->phydev->duplex) S16(sccp, scc_psmr, SCC_PSMR_LPB | SCC_PSMR_FDE); /* Restore multicast and promiscuous settings */ set_multicast_list(dev); S32(sccp, scc_gsmrl, SCC_GSMRL_ENR | SCC_GSMRL_ENT); } static void stop(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; int i; for (i = 0; (R16(sccp, scc_sccm) == 0) && i < SCC_RESET_DELAY; i++) udelay(1); if (i == SCC_RESET_DELAY) dev_warn(fep->dev, "SCC timeout on graceful transmit stop\n"); W16(sccp, scc_sccm, 0); C32(sccp, scc_gsmrl, SCC_GSMRL_ENR | SCC_GSMRL_ENT); fs_cleanup_bds(dev); } static void napi_clear_event_fs(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; W16(sccp, scc_scce, SCC_NAPI_EVENT_MSK); } static void napi_enable_fs(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; S16(sccp, scc_sccm, SCC_NAPI_EVENT_MSK); } static void napi_disable_fs(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; C16(sccp, scc_sccm, SCC_NAPI_EVENT_MSK); } static void rx_bd_done(struct net_device *dev) { /* nothing */ } static void tx_kickstart(struct net_device *dev) { /* nothing */ } static u32 get_int_events(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; return (u32) R16(sccp, scc_scce); } static void clear_int_events(struct net_device *dev, u32 int_events) { struct fs_enet_private *fep = netdev_priv(dev); scc_t __iomem *sccp = fep->scc.sccp; W16(sccp, scc_scce, int_events & 0xffff); } static void ev_error(struct net_device *dev, u32 int_events) { struct fs_enet_private *fep = netdev_priv(dev); dev_warn(fep->dev, "SCC ERROR(s) 0x%x\n", int_events); } static int get_regs(struct net_device *dev, void *p, int *sizep) { struct fs_enet_private *fep = netdev_priv(dev); if (*sizep < sizeof(scc_t) + sizeof(scc_enet_t __iomem *)) return -EINVAL; memcpy_fromio(p, fep->scc.sccp, sizeof(scc_t)); p = (char *)p + sizeof(scc_t); memcpy_fromio(p, fep->scc.ep, sizeof(scc_enet_t __iomem *)); return 0; } static int get_regs_len(struct net_device *dev) { return sizeof(scc_t) + sizeof(scc_enet_t __iomem *); } static void tx_restart(struct net_device *dev) { struct fs_enet_private *fep = netdev_priv(dev); scc_cr_cmd(fep, CPM_CR_RESTART_TX); } /*************************************************************************/ const struct fs_ops fs_scc_ops = { .setup_data = setup_data, .cleanup_data = cleanup_data, .set_multicast_list = set_multicast_list, .restart = restart, .stop = stop, .napi_clear_event = napi_clear_event_fs, .napi_enable = napi_enable_fs, .napi_disable = napi_disable_fs, .rx_bd_done = rx_bd_done, .tx_kickstart = tx_kickstart, .get_int_events = get_int_events, .clear_int_events = clear_int_events, .ev_error = ev_error, .get_regs = get_regs, .get_regs_len = get_regs_len, .tx_restart = tx_restart, .allocate_bd = allocate_bd, .free_bd = free_bd, };