// SPDX-License-Identifier: GPL-2.0-or-later /* * Copyright (C) 2006-2009 Freescale Semicondutor, Inc. All rights reserved. * * Author: Shlomi Gridish * Li Yang * * Description: * QE UCC Gigabit Ethernet Driver */ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "ucc_geth.h" #undef DEBUG #define ugeth_printk(level, format, arg...) \ printk(level format "\n", ## arg) #define ugeth_dbg(format, arg...) \ ugeth_printk(KERN_DEBUG , format , ## arg) #ifdef UGETH_VERBOSE_DEBUG #define ugeth_vdbg ugeth_dbg #else #define ugeth_vdbg(fmt, args...) do { } while (0) #endif /* UGETH_VERBOSE_DEBUG */ #define UGETH_MSG_DEFAULT (NETIF_MSG_IFUP << 1 ) - 1 static DEFINE_SPINLOCK(ugeth_lock); static struct { u32 msg_enable; } debug = { -1 }; module_param_named(debug, debug.msg_enable, int, 0); MODULE_PARM_DESC(debug, "Debug verbosity level (0=none, ..., 0xffff=all)"); static int ucc_geth_thread_count(enum ucc_geth_num_of_threads idx) { static const u8 count[] = { [UCC_GETH_NUM_OF_THREADS_1] = 1, [UCC_GETH_NUM_OF_THREADS_2] = 2, [UCC_GETH_NUM_OF_THREADS_4] = 4, [UCC_GETH_NUM_OF_THREADS_6] = 6, [UCC_GETH_NUM_OF_THREADS_8] = 8, }; if (idx >= ARRAY_SIZE(count)) return 0; return count[idx]; } static inline int ucc_geth_tx_queues(const struct ucc_geth_info *info) { return 1; } static inline int ucc_geth_rx_queues(const struct ucc_geth_info *info) { return 1; } static const struct ucc_geth_info ugeth_primary_info = { .uf_info = { .rtsm = UCC_FAST_SEND_IDLES_BETWEEN_FRAMES, .max_rx_buf_length = 1536, /* adjusted at startup if max-speed 1000 */ .urfs = UCC_GETH_URFS_INIT, .urfet = UCC_GETH_URFET_INIT, .urfset = UCC_GETH_URFSET_INIT, .utfs = UCC_GETH_UTFS_INIT, .utfet = UCC_GETH_UTFET_INIT, .utftt = UCC_GETH_UTFTT_INIT, .ufpt = 256, .mode = UCC_FAST_PROTOCOL_MODE_ETHERNET, .ttx_trx = UCC_FAST_GUMR_TRANSPARENT_TTX_TRX_NORMAL, .tenc = UCC_FAST_TX_ENCODING_NRZ, .renc = UCC_FAST_RX_ENCODING_NRZ, .tcrc = UCC_FAST_16_BIT_CRC, .synl = UCC_FAST_SYNC_LEN_NOT_USED, }, .extendedFilteringChainPointer = ((uint32_t) NULL), .typeorlen = 3072 /*1536 */ , .nonBackToBackIfgPart1 = 0x40, .nonBackToBackIfgPart2 = 0x60, .miminumInterFrameGapEnforcement = 0x50, .backToBackInterFrameGap = 0x60, .mblinterval = 128, .nortsrbytetime = 5, .fracsiz = 1, .strictpriorityq = 0xff, .altBebTruncation = 0xa, .excessDefer = 1, .maxRetransmission = 0xf, .collisionWindow = 0x37, .receiveFlowControl = 1, .transmitFlowControl = 1, .maxGroupAddrInHash = 4, .maxIndAddrInHash = 4, .prel = 7, .maxFrameLength = 1518+16, /* Add extra bytes for VLANs etc. */ .minFrameLength = 64, .maxD1Length = 1520+16, /* Add extra bytes for VLANs etc. */ .maxD2Length = 1520+16, /* Add extra bytes for VLANs etc. */ .vlantype = 0x8100, .ecamptr = ((uint32_t) NULL), .eventRegMask = UCCE_OTHER, .pausePeriod = 0xf000, .interruptcoalescingmaxvalue = {1, 1, 1, 1, 1, 1, 1, 1}, .bdRingLenTx = { TX_BD_RING_LEN, TX_BD_RING_LEN, TX_BD_RING_LEN, TX_BD_RING_LEN, TX_BD_RING_LEN, TX_BD_RING_LEN, TX_BD_RING_LEN, TX_BD_RING_LEN}, .bdRingLenRx = { RX_BD_RING_LEN, RX_BD_RING_LEN, RX_BD_RING_LEN, RX_BD_RING_LEN, RX_BD_RING_LEN, RX_BD_RING_LEN, RX_BD_RING_LEN, RX_BD_RING_LEN}, .numStationAddresses = UCC_GETH_NUM_OF_STATION_ADDRESSES_1, .largestexternallookupkeysize = QE_FLTR_LARGEST_EXTERNAL_TABLE_LOOKUP_KEY_SIZE_NONE, .statisticsMode = UCC_GETH_STATISTICS_GATHERING_MODE_HARDWARE | UCC_GETH_STATISTICS_GATHERING_MODE_FIRMWARE_TX | UCC_GETH_STATISTICS_GATHERING_MODE_FIRMWARE_RX, .vlanOperationTagged = UCC_GETH_VLAN_OPERATION_TAGGED_NOP, .vlanOperationNonTagged = UCC_GETH_VLAN_OPERATION_NON_TAGGED_NOP, .rxQoSMode = UCC_GETH_QOS_MODE_DEFAULT, .aufc = UPSMR_AUTOMATIC_FLOW_CONTROL_MODE_NONE, .padAndCrc = MACCFG2_PAD_AND_CRC_MODE_PAD_AND_CRC, .numThreadsTx = UCC_GETH_NUM_OF_THREADS_1, .numThreadsRx = UCC_GETH_NUM_OF_THREADS_1, .riscTx = QE_RISC_ALLOCATION_RISC1_AND_RISC2, .riscRx = QE_RISC_ALLOCATION_RISC1_AND_RISC2, }; #ifdef DEBUG static void mem_disp(u8 *addr, int size) { u8 *i; int size16Aling = (size >> 4) << 4; int size4Aling = (size >> 2) << 2; int notAlign = 0; if (size % 16) notAlign = 1; for (i = addr; (u32) i < (u32) addr + size16Aling; i += 16) printk("0x%08x: %08x %08x %08x %08x\r\n", (u32) i, *((u32 *) (i)), *((u32 *) (i + 4)), *((u32 *) (i + 8)), *((u32 *) (i + 12))); if (notAlign == 1) printk("0x%08x: ", (u32) i); for (; (u32) i < (u32) addr + size4Aling; i += 4) printk("%08x ", *((u32 *) (i))); for (; (u32) i < (u32) addr + size; i++) printk("%02x", *((i))); if (notAlign == 1) printk("\r\n"); } #endif /* DEBUG */ static struct list_head *dequeue(struct list_head *lh) { unsigned long flags; spin_lock_irqsave(&ugeth_lock, flags); if (!list_empty(lh)) { struct list_head *node = lh->next; list_del(node); spin_unlock_irqrestore(&ugeth_lock, flags); return node; } else { spin_unlock_irqrestore(&ugeth_lock, flags); return NULL; } } static struct sk_buff *get_new_skb(struct ucc_geth_private *ugeth, u8 __iomem *bd) { struct sk_buff *skb; skb = netdev_alloc_skb(ugeth->ndev, ugeth->ug_info->uf_info.max_rx_buf_length + UCC_GETH_RX_DATA_BUF_ALIGNMENT); if (!skb) return NULL; /* We need the data buffer to be aligned properly. We will reserve * as many bytes as needed to align the data properly */ skb_reserve(skb, UCC_GETH_RX_DATA_BUF_ALIGNMENT - (((unsigned)skb->data) & (UCC_GETH_RX_DATA_BUF_ALIGNMENT - 1))); out_be32(&((struct qe_bd __iomem *)bd)->buf, dma_map_single(ugeth->dev, skb->data, ugeth->ug_info->uf_info.max_rx_buf_length + UCC_GETH_RX_DATA_BUF_ALIGNMENT, DMA_FROM_DEVICE)); out_be32((u32 __iomem *)bd, (R_E | R_I | (in_be32((u32 __iomem*)bd) & R_W))); return skb; } static int rx_bd_buffer_set(struct ucc_geth_private *ugeth, u8 rxQ) { u8 __iomem *bd; u32 bd_status; struct sk_buff *skb; int i; bd = ugeth->p_rx_bd_ring[rxQ]; i = 0; do { bd_status = in_be32((u32 __iomem *)bd); skb = get_new_skb(ugeth, bd); if (!skb) /* If can not allocate data buffer, abort. Cleanup will be elsewhere */ return -ENOMEM; ugeth->rx_skbuff[rxQ][i] = skb; /* advance the BD pointer */ bd += sizeof(struct qe_bd); i++; } while (!(bd_status & R_W)); return 0; } static int fill_init_enet_entries(struct ucc_geth_private *ugeth, u32 *p_start, u8 num_entries, u32 thread_size, u32 thread_alignment, unsigned int risc, int skip_page_for_first_entry) { u32 init_enet_offset; u8 i; int snum; for (i = 0; i < num_entries; i++) { if ((snum = qe_get_snum()) < 0) { if (netif_msg_ifup(ugeth)) pr_err("Can not get SNUM\n"); return snum; } if ((i == 0) && skip_page_for_first_entry) /* First entry of Rx does not have page */ init_enet_offset = 0; else { init_enet_offset = qe_muram_alloc(thread_size, thread_alignment); if (IS_ERR_VALUE(init_enet_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory\n"); qe_put_snum((u8) snum); return -ENOMEM; } } *(p_start++) = ((u8) snum << ENET_INIT_PARAM_SNUM_SHIFT) | init_enet_offset | risc; } return 0; } static int return_init_enet_entries(struct ucc_geth_private *ugeth, u32 *p_start, u8 num_entries, unsigned int risc, int skip_page_for_first_entry) { u32 init_enet_offset; u8 i; int snum; for (i = 0; i < num_entries; i++) { u32 val = *p_start; /* Check that this entry was actually valid -- needed in case failed in allocations */ if ((val & ENET_INIT_PARAM_RISC_MASK) == risc) { snum = (u32) (val & ENET_INIT_PARAM_SNUM_MASK) >> ENET_INIT_PARAM_SNUM_SHIFT; qe_put_snum((u8) snum); if (!((i == 0) && skip_page_for_first_entry)) { /* First entry of Rx does not have page */ init_enet_offset = (val & ENET_INIT_PARAM_PTR_MASK); qe_muram_free(init_enet_offset); } *p_start++ = 0; } } return 0; } #ifdef DEBUG static int dump_init_enet_entries(struct ucc_geth_private *ugeth, u32 __iomem *p_start, u8 num_entries, u32 thread_size, unsigned int risc, int skip_page_for_first_entry) { u32 init_enet_offset; u8 i; int snum; for (i = 0; i < num_entries; i++) { u32 val = in_be32(p_start); /* Check that this entry was actually valid -- needed in case failed in allocations */ if ((val & ENET_INIT_PARAM_RISC_MASK) == risc) { snum = (u32) (val & ENET_INIT_PARAM_SNUM_MASK) >> ENET_INIT_PARAM_SNUM_SHIFT; qe_put_snum((u8) snum); if (!((i == 0) && skip_page_for_first_entry)) { /* First entry of Rx does not have page */ init_enet_offset = (in_be32(p_start) & ENET_INIT_PARAM_PTR_MASK); pr_info("Init enet entry %d:\n", i); pr_info("Base address: 0x%08x\n", (u32)qe_muram_addr(init_enet_offset)); mem_disp(qe_muram_addr(init_enet_offset), thread_size); } p_start++; } } return 0; } #endif static void put_enet_addr_container(struct enet_addr_container *enet_addr_cont) { kfree(enet_addr_cont); } static void set_mac_addr(__be16 __iomem *reg, u8 *mac) { out_be16(®[0], ((u16)mac[5] << 8) | mac[4]); out_be16(®[1], ((u16)mac[3] << 8) | mac[2]); out_be16(®[2], ((u16)mac[1] << 8) | mac[0]); } static int hw_clear_addr_in_paddr(struct ucc_geth_private *ugeth, u8 paddr_num) { struct ucc_geth_82xx_address_filtering_pram __iomem *p_82xx_addr_filt; if (paddr_num >= NUM_OF_PADDRS) { pr_warn("%s: Invalid paddr_num: %u\n", __func__, paddr_num); return -EINVAL; } p_82xx_addr_filt = (struct ucc_geth_82xx_address_filtering_pram __iomem *) ugeth->p_rx_glbl_pram-> addressfiltering; /* Writing address ff.ff.ff.ff.ff.ff disables address recognition for this register */ out_be16(&p_82xx_addr_filt->paddr[paddr_num].h, 0xffff); out_be16(&p_82xx_addr_filt->paddr[paddr_num].m, 0xffff); out_be16(&p_82xx_addr_filt->paddr[paddr_num].l, 0xffff); return 0; } static void hw_add_addr_in_hash(struct ucc_geth_private *ugeth, u8 *p_enet_addr) { struct ucc_geth_82xx_address_filtering_pram __iomem *p_82xx_addr_filt; u32 cecr_subblock; p_82xx_addr_filt = (struct ucc_geth_82xx_address_filtering_pram __iomem *) ugeth->p_rx_glbl_pram-> addressfiltering; cecr_subblock = ucc_fast_get_qe_cr_subblock(ugeth->ug_info->uf_info.ucc_num); /* Ethernet frames are defined in Little Endian mode, therefore to insert */ /* the address to the hash (Big Endian mode), we reverse the bytes.*/ set_mac_addr(&p_82xx_addr_filt->taddr.h, p_enet_addr); qe_issue_cmd(QE_SET_GROUP_ADDRESS, cecr_subblock, QE_CR_PROTOCOL_ETHERNET, 0); } #ifdef DEBUG static void get_statistics(struct ucc_geth_private *ugeth, struct ucc_geth_tx_firmware_statistics * tx_firmware_statistics, struct ucc_geth_rx_firmware_statistics * rx_firmware_statistics, struct ucc_geth_hardware_statistics *hardware_statistics) { struct ucc_fast __iomem *uf_regs; struct ucc_geth __iomem *ug_regs; struct ucc_geth_tx_firmware_statistics_pram *p_tx_fw_statistics_pram; struct ucc_geth_rx_firmware_statistics_pram *p_rx_fw_statistics_pram; ug_regs = ugeth->ug_regs; uf_regs = (struct ucc_fast __iomem *) ug_regs; p_tx_fw_statistics_pram = ugeth->p_tx_fw_statistics_pram; p_rx_fw_statistics_pram = ugeth->p_rx_fw_statistics_pram; /* Tx firmware only if user handed pointer and driver actually gathers Tx firmware statistics */ if (tx_firmware_statistics && p_tx_fw_statistics_pram) { tx_firmware_statistics->sicoltx = in_be32(&p_tx_fw_statistics_pram->sicoltx); tx_firmware_statistics->mulcoltx = in_be32(&p_tx_fw_statistics_pram->mulcoltx); tx_firmware_statistics->latecoltxfr = in_be32(&p_tx_fw_statistics_pram->latecoltxfr); tx_firmware_statistics->frabortduecol = in_be32(&p_tx_fw_statistics_pram->frabortduecol); tx_firmware_statistics->frlostinmactxer = in_be32(&p_tx_fw_statistics_pram->frlostinmactxer); tx_firmware_statistics->carriersenseertx = in_be32(&p_tx_fw_statistics_pram->carriersenseertx); tx_firmware_statistics->frtxok = in_be32(&p_tx_fw_statistics_pram->frtxok); tx_firmware_statistics->txfrexcessivedefer = in_be32(&p_tx_fw_statistics_pram->txfrexcessivedefer); tx_firmware_statistics->txpkts256 = in_be32(&p_tx_fw_statistics_pram->txpkts256); tx_firmware_statistics->txpkts512 = in_be32(&p_tx_fw_statistics_pram->txpkts512); tx_firmware_statistics->txpkts1024 = in_be32(&p_tx_fw_statistics_pram->txpkts1024); tx_firmware_statistics->txpktsjumbo = in_be32(&p_tx_fw_statistics_pram->txpktsjumbo); } /* Rx firmware only if user handed pointer and driver actually * gathers Rx firmware statistics */ if (rx_firmware_statistics && p_rx_fw_statistics_pram) { int i; rx_firmware_statistics->frrxfcser = in_be32(&p_rx_fw_statistics_pram->frrxfcser); rx_firmware_statistics->fraligner = in_be32(&p_rx_fw_statistics_pram->fraligner); rx_firmware_statistics->inrangelenrxer = in_be32(&p_rx_fw_statistics_pram->inrangelenrxer); rx_firmware_statistics->outrangelenrxer = in_be32(&p_rx_fw_statistics_pram->outrangelenrxer); rx_firmware_statistics->frtoolong = in_be32(&p_rx_fw_statistics_pram->frtoolong); rx_firmware_statistics->runt = in_be32(&p_rx_fw_statistics_pram->runt); rx_firmware_statistics->verylongevent = in_be32(&p_rx_fw_statistics_pram->verylongevent); rx_firmware_statistics->symbolerror = in_be32(&p_rx_fw_statistics_pram->symbolerror); rx_firmware_statistics->dropbsy = in_be32(&p_rx_fw_statistics_pram->dropbsy); for (i = 0; i < 0x8; i++) rx_firmware_statistics->res0[i] = p_rx_fw_statistics_pram->res0[i]; rx_firmware_statistics->mismatchdrop = in_be32(&p_rx_fw_statistics_pram->mismatchdrop); rx_firmware_statistics->underpkts = in_be32(&p_rx_fw_statistics_pram->underpkts); rx_firmware_statistics->pkts256 = in_be32(&p_rx_fw_statistics_pram->pkts256); rx_firmware_statistics->pkts512 = in_be32(&p_rx_fw_statistics_pram->pkts512); rx_firmware_statistics->pkts1024 = in_be32(&p_rx_fw_statistics_pram->pkts1024); rx_firmware_statistics->pktsjumbo = in_be32(&p_rx_fw_statistics_pram->pktsjumbo); rx_firmware_statistics->frlossinmacer = in_be32(&p_rx_fw_statistics_pram->frlossinmacer); rx_firmware_statistics->pausefr = in_be32(&p_rx_fw_statistics_pram->pausefr); for (i = 0; i < 0x4; i++) rx_firmware_statistics->res1[i] = p_rx_fw_statistics_pram->res1[i]; rx_firmware_statistics->removevlan = in_be32(&p_rx_fw_statistics_pram->removevlan); rx_firmware_statistics->replacevlan = in_be32(&p_rx_fw_statistics_pram->replacevlan); rx_firmware_statistics->insertvlan = in_be32(&p_rx_fw_statistics_pram->insertvlan); } /* Hardware only if user handed pointer and driver actually gathers hardware statistics */ if (hardware_statistics && (in_be32(&uf_regs->upsmr) & UCC_GETH_UPSMR_HSE)) { hardware_statistics->tx64 = in_be32(&ug_regs->tx64); hardware_statistics->tx127 = in_be32(&ug_regs->tx127); hardware_statistics->tx255 = in_be32(&ug_regs->tx255); hardware_statistics->rx64 = in_be32(&ug_regs->rx64); hardware_statistics->rx127 = in_be32(&ug_regs->rx127); hardware_statistics->rx255 = in_be32(&ug_regs->rx255); hardware_statistics->txok = in_be32(&ug_regs->txok); hardware_statistics->txcf = in_be16(&ug_regs->txcf); hardware_statistics->tmca = in_be32(&ug_regs->tmca); hardware_statistics->tbca = in_be32(&ug_regs->tbca); hardware_statistics->rxfok = in_be32(&ug_regs->rxfok); hardware_statistics->rxbok = in_be32(&ug_regs->rxbok); hardware_statistics->rbyt = in_be32(&ug_regs->rbyt); hardware_statistics->rmca = in_be32(&ug_regs->rmca); hardware_statistics->rbca = in_be32(&ug_regs->rbca); } } static void dump_bds(struct ucc_geth_private *ugeth) { int i; int length; for (i = 0; i < ucc_geth_tx_queues(ugeth->ug_info); i++) { if (ugeth->p_tx_bd_ring[i]) { length = (ugeth->ug_info->bdRingLenTx[i] * sizeof(struct qe_bd)); pr_info("TX BDs[%d]\n", i); mem_disp(ugeth->p_tx_bd_ring[i], length); } } for (i = 0; i < ucc_geth_rx_queues(ugeth->ug_info); i++) { if (ugeth->p_rx_bd_ring[i]) { length = (ugeth->ug_info->bdRingLenRx[i] * sizeof(struct qe_bd)); pr_info("RX BDs[%d]\n", i); mem_disp(ugeth->p_rx_bd_ring[i], length); } } } static void dump_regs(struct ucc_geth_private *ugeth) { int i; pr_info("UCC%d Geth registers:\n", ugeth->ug_info->uf_info.ucc_num + 1); pr_info("Base address: 0x%08x\n", (u32)ugeth->ug_regs); pr_info("maccfg1 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->maccfg1, in_be32(&ugeth->ug_regs->maccfg1)); pr_info("maccfg2 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->maccfg2, in_be32(&ugeth->ug_regs->maccfg2)); pr_info("ipgifg : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->ipgifg, in_be32(&ugeth->ug_regs->ipgifg)); pr_info("hafdup : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->hafdup, in_be32(&ugeth->ug_regs->hafdup)); pr_info("ifctl : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->ifctl, in_be32(&ugeth->ug_regs->ifctl)); pr_info("ifstat : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->ifstat, in_be32(&ugeth->ug_regs->ifstat)); pr_info("macstnaddr1: addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->macstnaddr1, in_be32(&ugeth->ug_regs->macstnaddr1)); pr_info("macstnaddr2: addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->macstnaddr2, in_be32(&ugeth->ug_regs->macstnaddr2)); pr_info("uempr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->uempr, in_be32(&ugeth->ug_regs->uempr)); pr_info("utbipar : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->utbipar, in_be32(&ugeth->ug_regs->utbipar)); pr_info("uescr : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->ug_regs->uescr, in_be16(&ugeth->ug_regs->uescr)); pr_info("tx64 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->tx64, in_be32(&ugeth->ug_regs->tx64)); pr_info("tx127 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->tx127, in_be32(&ugeth->ug_regs->tx127)); pr_info("tx255 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->tx255, in_be32(&ugeth->ug_regs->tx255)); pr_info("rx64 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rx64, in_be32(&ugeth->ug_regs->rx64)); pr_info("rx127 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rx127, in_be32(&ugeth->ug_regs->rx127)); pr_info("rx255 : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rx255, in_be32(&ugeth->ug_regs->rx255)); pr_info("txok : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->txok, in_be32(&ugeth->ug_regs->txok)); pr_info("txcf : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->ug_regs->txcf, in_be16(&ugeth->ug_regs->txcf)); pr_info("tmca : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->tmca, in_be32(&ugeth->ug_regs->tmca)); pr_info("tbca : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->tbca, in_be32(&ugeth->ug_regs->tbca)); pr_info("rxfok : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rxfok, in_be32(&ugeth->ug_regs->rxfok)); pr_info("rxbok : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rxbok, in_be32(&ugeth->ug_regs->rxbok)); pr_info("rbyt : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rbyt, in_be32(&ugeth->ug_regs->rbyt)); pr_info("rmca : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rmca, in_be32(&ugeth->ug_regs->rmca)); pr_info("rbca : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->rbca, in_be32(&ugeth->ug_regs->rbca)); pr_info("scar : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->scar, in_be32(&ugeth->ug_regs->scar)); pr_info("scam : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->ug_regs->scam, in_be32(&ugeth->ug_regs->scam)); if (ugeth->p_thread_data_tx) { int count = ucc_geth_thread_count(ugeth->ug_info->numThreadsTx); pr_info("Thread data TXs:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_thread_data_tx); for (i = 0; i < count; i++) { pr_info("Thread data TX[%d]:\n", i); pr_info("Base address: 0x%08x\n", (u32)&ugeth->p_thread_data_tx[i]); mem_disp((u8 *) & ugeth->p_thread_data_tx[i], sizeof(struct ucc_geth_thread_data_tx)); } } if (ugeth->p_thread_data_rx) { int count = ucc_geth_thread_count(ugeth->ug_info->numThreadsRx); pr_info("Thread data RX:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_thread_data_rx); for (i = 0; i < count; i++) { pr_info("Thread data RX[%d]:\n", i); pr_info("Base address: 0x%08x\n", (u32)&ugeth->p_thread_data_rx[i]); mem_disp((u8 *) & ugeth->p_thread_data_rx[i], sizeof(struct ucc_geth_thread_data_rx)); } } if (ugeth->p_exf_glbl_param) { pr_info("EXF global param:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_exf_glbl_param); mem_disp((u8 *) ugeth->p_exf_glbl_param, sizeof(*ugeth->p_exf_glbl_param)); } if (ugeth->p_tx_glbl_pram) { pr_info("TX global param:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_tx_glbl_pram); pr_info("temoder : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_tx_glbl_pram->temoder, in_be16(&ugeth->p_tx_glbl_pram->temoder)); pr_info("sqptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->sqptr, in_be32(&ugeth->p_tx_glbl_pram->sqptr)); pr_info("schedulerbasepointer: addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->schedulerbasepointer, in_be32(&ugeth->p_tx_glbl_pram->schedulerbasepointer)); pr_info("txrmonbaseptr: addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->txrmonbaseptr, in_be32(&ugeth->p_tx_glbl_pram->txrmonbaseptr)); pr_info("tstate : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->tstate, in_be32(&ugeth->p_tx_glbl_pram->tstate)); pr_info("iphoffset[0] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[0], ugeth->p_tx_glbl_pram->iphoffset[0]); pr_info("iphoffset[1] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[1], ugeth->p_tx_glbl_pram->iphoffset[1]); pr_info("iphoffset[2] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[2], ugeth->p_tx_glbl_pram->iphoffset[2]); pr_info("iphoffset[3] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[3], ugeth->p_tx_glbl_pram->iphoffset[3]); pr_info("iphoffset[4] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[4], ugeth->p_tx_glbl_pram->iphoffset[4]); pr_info("iphoffset[5] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[5], ugeth->p_tx_glbl_pram->iphoffset[5]); pr_info("iphoffset[6] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[6], ugeth->p_tx_glbl_pram->iphoffset[6]); pr_info("iphoffset[7] : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_tx_glbl_pram->iphoffset[7], ugeth->p_tx_glbl_pram->iphoffset[7]); pr_info("vtagtable[0] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[0], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[0])); pr_info("vtagtable[1] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[1], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[1])); pr_info("vtagtable[2] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[2], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[2])); pr_info("vtagtable[3] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[3], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[3])); pr_info("vtagtable[4] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[4], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[4])); pr_info("vtagtable[5] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[5], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[5])); pr_info("vtagtable[6] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[6], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[6])); pr_info("vtagtable[7] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->vtagtable[7], in_be32(&ugeth->p_tx_glbl_pram->vtagtable[7])); pr_info("tqptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_tx_glbl_pram->tqptr, in_be32(&ugeth->p_tx_glbl_pram->tqptr)); } if (ugeth->p_rx_glbl_pram) { pr_info("RX global param:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_rx_glbl_pram); pr_info("remoder : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->remoder, in_be32(&ugeth->p_rx_glbl_pram->remoder)); pr_info("rqptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->rqptr, in_be32(&ugeth->p_rx_glbl_pram->rqptr)); pr_info("typeorlen : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->typeorlen, in_be16(&ugeth->p_rx_glbl_pram->typeorlen)); pr_info("rxgstpack : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_rx_glbl_pram->rxgstpack, ugeth->p_rx_glbl_pram->rxgstpack); pr_info("rxrmonbaseptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->rxrmonbaseptr, in_be32(&ugeth->p_rx_glbl_pram->rxrmonbaseptr)); pr_info("intcoalescingptr: addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->intcoalescingptr, in_be32(&ugeth->p_rx_glbl_pram->intcoalescingptr)); pr_info("rstate : addr - 0x%08x, val - 0x%02x\n", (u32)&ugeth->p_rx_glbl_pram->rstate, ugeth->p_rx_glbl_pram->rstate); pr_info("mrblr : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->mrblr, in_be16(&ugeth->p_rx_glbl_pram->mrblr)); pr_info("rbdqptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->rbdqptr, in_be32(&ugeth->p_rx_glbl_pram->rbdqptr)); pr_info("mflr : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->mflr, in_be16(&ugeth->p_rx_glbl_pram->mflr)); pr_info("minflr : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->minflr, in_be16(&ugeth->p_rx_glbl_pram->minflr)); pr_info("maxd1 : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->maxd1, in_be16(&ugeth->p_rx_glbl_pram->maxd1)); pr_info("maxd2 : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->maxd2, in_be16(&ugeth->p_rx_glbl_pram->maxd2)); pr_info("ecamptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->ecamptr, in_be32(&ugeth->p_rx_glbl_pram->ecamptr)); pr_info("l2qt : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l2qt, in_be32(&ugeth->p_rx_glbl_pram->l2qt)); pr_info("l3qt[0] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[0], in_be32(&ugeth->p_rx_glbl_pram->l3qt[0])); pr_info("l3qt[1] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[1], in_be32(&ugeth->p_rx_glbl_pram->l3qt[1])); pr_info("l3qt[2] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[2], in_be32(&ugeth->p_rx_glbl_pram->l3qt[2])); pr_info("l3qt[3] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[3], in_be32(&ugeth->p_rx_glbl_pram->l3qt[3])); pr_info("l3qt[4] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[4], in_be32(&ugeth->p_rx_glbl_pram->l3qt[4])); pr_info("l3qt[5] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[5], in_be32(&ugeth->p_rx_glbl_pram->l3qt[5])); pr_info("l3qt[6] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[6], in_be32(&ugeth->p_rx_glbl_pram->l3qt[6])); pr_info("l3qt[7] : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->l3qt[7], in_be32(&ugeth->p_rx_glbl_pram->l3qt[7])); pr_info("vlantype : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->vlantype, in_be16(&ugeth->p_rx_glbl_pram->vlantype)); pr_info("vlantci : addr - 0x%08x, val - 0x%04x\n", (u32)&ugeth->p_rx_glbl_pram->vlantci, in_be16(&ugeth->p_rx_glbl_pram->vlantci)); for (i = 0; i < 64; i++) pr_info("addressfiltering[%d]: addr - 0x%08x, val - 0x%02x\n", i, (u32)&ugeth->p_rx_glbl_pram->addressfiltering[i], ugeth->p_rx_glbl_pram->addressfiltering[i]); pr_info("exfGlobalParam : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_glbl_pram->exfGlobalParam, in_be32(&ugeth->p_rx_glbl_pram->exfGlobalParam)); } if (ugeth->p_send_q_mem_reg) { pr_info("Send Q memory registers:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_send_q_mem_reg); for (i = 0; i < ucc_geth_tx_queues(ugeth->ug_info); i++) { pr_info("SQQD[%d]:\n", i); pr_info("Base address: 0x%08x\n", (u32)&ugeth->p_send_q_mem_reg->sqqd[i]); mem_disp((u8 *) & ugeth->p_send_q_mem_reg->sqqd[i], sizeof(struct ucc_geth_send_queue_qd)); } } if (ugeth->p_scheduler) { pr_info("Scheduler:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_scheduler); mem_disp((u8 *) ugeth->p_scheduler, sizeof(*ugeth->p_scheduler)); } if (ugeth->p_tx_fw_statistics_pram) { pr_info("TX FW statistics pram:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_tx_fw_statistics_pram); mem_disp((u8 *) ugeth->p_tx_fw_statistics_pram, sizeof(*ugeth->p_tx_fw_statistics_pram)); } if (ugeth->p_rx_fw_statistics_pram) { pr_info("RX FW statistics pram:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_rx_fw_statistics_pram); mem_disp((u8 *) ugeth->p_rx_fw_statistics_pram, sizeof(*ugeth->p_rx_fw_statistics_pram)); } if (ugeth->p_rx_irq_coalescing_tbl) { pr_info("RX IRQ coalescing tables:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_rx_irq_coalescing_tbl); for (i = 0; i < ucc_geth_rx_queues(ugeth->ug_info); i++) { pr_info("RX IRQ coalescing table entry[%d]:\n", i); pr_info("Base address: 0x%08x\n", (u32)&ugeth->p_rx_irq_coalescing_tbl-> coalescingentry[i]); pr_info("interruptcoalescingmaxvalue: addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_irq_coalescing_tbl-> coalescingentry[i].interruptcoalescingmaxvalue, in_be32(&ugeth->p_rx_irq_coalescing_tbl-> coalescingentry[i]. interruptcoalescingmaxvalue)); pr_info("interruptcoalescingcounter : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_irq_coalescing_tbl-> coalescingentry[i].interruptcoalescingcounter, in_be32(&ugeth->p_rx_irq_coalescing_tbl-> coalescingentry[i]. interruptcoalescingcounter)); } } if (ugeth->p_rx_bd_qs_tbl) { pr_info("RX BD QS tables:\n"); pr_info("Base address: 0x%08x\n", (u32)ugeth->p_rx_bd_qs_tbl); for (i = 0; i < ucc_geth_rx_queues(ugeth->ug_info); i++) { pr_info("RX BD QS table[%d]:\n", i); pr_info("Base address: 0x%08x\n", (u32)&ugeth->p_rx_bd_qs_tbl[i]); pr_info("bdbaseptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_bd_qs_tbl[i].bdbaseptr, in_be32(&ugeth->p_rx_bd_qs_tbl[i].bdbaseptr)); pr_info("bdptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_bd_qs_tbl[i].bdptr, in_be32(&ugeth->p_rx_bd_qs_tbl[i].bdptr)); pr_info("externalbdbaseptr: addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_bd_qs_tbl[i].externalbdbaseptr, in_be32(&ugeth->p_rx_bd_qs_tbl[i]. externalbdbaseptr)); pr_info("externalbdptr : addr - 0x%08x, val - 0x%08x\n", (u32)&ugeth->p_rx_bd_qs_tbl[i].externalbdptr, in_be32(&ugeth->p_rx_bd_qs_tbl[i].externalbdptr)); pr_info("ucode RX Prefetched BDs:\n"); pr_info("Base address: 0x%08x\n", (u32)qe_muram_addr(in_be32 (&ugeth->p_rx_bd_qs_tbl[i]. bdbaseptr))); mem_disp((u8 *) qe_muram_addr(in_be32 (&ugeth->p_rx_bd_qs_tbl[i]. bdbaseptr)), sizeof(struct ucc_geth_rx_prefetched_bds)); } } if (ugeth->p_init_enet_param_shadow) { int size; pr_info("Init enet param shadow:\n"); pr_info("Base address: 0x%08x\n", (u32) ugeth->p_init_enet_param_shadow); mem_disp((u8 *) ugeth->p_init_enet_param_shadow, sizeof(*ugeth->p_init_enet_param_shadow)); size = sizeof(struct ucc_geth_thread_rx_pram); if (ugeth->ug_info->rxExtendedFiltering) { size += THREAD_RX_PRAM_ADDITIONAL_FOR_EXTENDED_FILTERING; if (ugeth->ug_info->largestexternallookupkeysize == QE_FLTR_TABLE_LOOKUP_KEY_SIZE_8_BYTES) size += THREAD_RX_PRAM_ADDITIONAL_FOR_EXTENDED_FILTERING_8; if (ugeth->ug_info->largestexternallookupkeysize == QE_FLTR_TABLE_LOOKUP_KEY_SIZE_16_BYTES) size += THREAD_RX_PRAM_ADDITIONAL_FOR_EXTENDED_FILTERING_16; } dump_init_enet_entries(ugeth, &(ugeth->p_init_enet_param_shadow-> txthread[0]), ENET_INIT_PARAM_MAX_ENTRIES_TX, sizeof(struct ucc_geth_thread_tx_pram), ugeth->ug_info->riscTx, 0); dump_init_enet_entries(ugeth, &(ugeth->p_init_enet_param_shadow-> rxthread[0]), ENET_INIT_PARAM_MAX_ENTRIES_RX, size, ugeth->ug_info->riscRx, 1); } } #endif /* DEBUG */ static void init_default_reg_vals(u32 __iomem *upsmr_register, u32 __iomem *maccfg1_register, u32 __iomem *maccfg2_register) { out_be32(upsmr_register, UCC_GETH_UPSMR_INIT); out_be32(maccfg1_register, UCC_GETH_MACCFG1_INIT); out_be32(maccfg2_register, UCC_GETH_MACCFG2_INIT); } static int init_half_duplex_params(int alt_beb, int back_pressure_no_backoff, int no_backoff, int excess_defer, u8 alt_beb_truncation, u8 max_retransmissions, u8 collision_window, u32 __iomem *hafdup_register) { u32 value = 0; if ((alt_beb_truncation > HALFDUP_ALT_BEB_TRUNCATION_MAX) || (max_retransmissions > HALFDUP_MAX_RETRANSMISSION_MAX) || (collision_window > HALFDUP_COLLISION_WINDOW_MAX)) return -EINVAL; value = (u32) (alt_beb_truncation << HALFDUP_ALT_BEB_TRUNCATION_SHIFT); if (alt_beb) value |= HALFDUP_ALT_BEB; if (back_pressure_no_backoff) value |= HALFDUP_BACK_PRESSURE_NO_BACKOFF; if (no_backoff) value |= HALFDUP_NO_BACKOFF; if (excess_defer) value |= HALFDUP_EXCESSIVE_DEFER; value |= (max_retransmissions << HALFDUP_MAX_RETRANSMISSION_SHIFT); value |= collision_window; out_be32(hafdup_register, value); return 0; } static int init_inter_frame_gap_params(u8 non_btb_cs_ipg, u8 non_btb_ipg, u8 min_ifg, u8 btb_ipg, u32 __iomem *ipgifg_register) { u32 value = 0; /* Non-Back-to-back IPG part 1 should be <= Non-Back-to-back IPG part 2 */ if (non_btb_cs_ipg > non_btb_ipg) return -EINVAL; if ((non_btb_cs_ipg > IPGIFG_NON_BACK_TO_BACK_IFG_PART1_MAX) || (non_btb_ipg > IPGIFG_NON_BACK_TO_BACK_IFG_PART2_MAX) || /*(min_ifg > IPGIFG_MINIMUM_IFG_ENFORCEMENT_MAX) || */ (btb_ipg > IPGIFG_BACK_TO_BACK_IFG_MAX)) return -EINVAL; value |= ((non_btb_cs_ipg << IPGIFG_NON_BACK_TO_BACK_IFG_PART1_SHIFT) & IPGIFG_NBTB_CS_IPG_MASK); value |= ((non_btb_ipg << IPGIFG_NON_BACK_TO_BACK_IFG_PART2_SHIFT) & IPGIFG_NBTB_IPG_MASK); value |= ((min_ifg << IPGIFG_MINIMUM_IFG_ENFORCEMENT_SHIFT) & IPGIFG_MIN_IFG_MASK); value |= (btb_ipg & IPGIFG_BTB_IPG_MASK); out_be32(ipgifg_register, value); return 0; } int init_flow_control_params(u32 automatic_flow_control_mode, int rx_flow_control_enable, int tx_flow_control_enable, u16 pause_period, u16 extension_field, u32 __iomem *upsmr_register, u32 __iomem *uempr_register, u32 __iomem *maccfg1_register) { u32 value = 0; /* Set UEMPR register */ value = (u32) pause_period << UEMPR_PAUSE_TIME_VALUE_SHIFT; value |= (u32) extension_field << UEMPR_EXTENDED_PAUSE_TIME_VALUE_SHIFT; out_be32(uempr_register, value); /* Set UPSMR register */ setbits32(upsmr_register, automatic_flow_control_mode); value = in_be32(maccfg1_register); if (rx_flow_control_enable) value |= MACCFG1_FLOW_RX; if (tx_flow_control_enable) value |= MACCFG1_FLOW_TX; out_be32(maccfg1_register, value); return 0; } static int init_hw_statistics_gathering_mode(int enable_hardware_statistics, int auto_zero_hardware_statistics, u32 __iomem *upsmr_register, u16 __iomem *uescr_register) { u16 uescr_value = 0; /* Enable hardware statistics gathering if requested */ if (enable_hardware_statistics) setbits32(upsmr_register, UCC_GETH_UPSMR_HSE); /* Clear hardware statistics counters */ uescr_value = in_be16(uescr_register); uescr_value |= UESCR_CLRCNT; /* Automatically zero hardware statistics counters on read, if requested */ if (auto_zero_hardware_statistics) uescr_value |= UESCR_AUTOZ; out_be16(uescr_register, uescr_value); return 0; } static int init_firmware_statistics_gathering_mode(int enable_tx_firmware_statistics, int enable_rx_firmware_statistics, u32 __iomem *tx_rmon_base_ptr, u32 tx_firmware_statistics_structure_address, u32 __iomem *rx_rmon_base_ptr, u32 rx_firmware_statistics_structure_address, u16 __iomem *temoder_register, u32 __iomem *remoder_register) { /* Note: this function does not check if */ /* the parameters it receives are NULL */ if (enable_tx_firmware_statistics) { out_be32(tx_rmon_base_ptr, tx_firmware_statistics_structure_address); setbits16(temoder_register, TEMODER_TX_RMON_STATISTICS_ENABLE); } if (enable_rx_firmware_statistics) { out_be32(rx_rmon_base_ptr, rx_firmware_statistics_structure_address); setbits32(remoder_register, REMODER_RX_RMON_STATISTICS_ENABLE); } return 0; } static int init_mac_station_addr_regs(u8 address_byte_0, u8 address_byte_1, u8 address_byte_2, u8 address_byte_3, u8 address_byte_4, u8 address_byte_5, u32 __iomem *macstnaddr1_register, u32 __iomem *macstnaddr2_register) { u32 value = 0; /* Example: for a station address of 0x12345678ABCD, */ /* 0x12 is byte 0, 0x34 is byte 1 and so on and 0xCD is byte 5 */ /* MACSTNADDR1 Register: */ /* 0 7 8 15 */ /* station address byte 5 station address byte 4 */ /* 16 23 24 31 */ /* station address byte 3 station address byte 2 */ value |= (u32) ((address_byte_2 << 0) & 0x000000FF); value |= (u32) ((address_byte_3 << 8) & 0x0000FF00); value |= (u32) ((address_byte_4 << 16) & 0x00FF0000); value |= (u32) ((address_byte_5 << 24) & 0xFF000000); out_be32(macstnaddr1_register, value); /* MACSTNADDR2 Register: */ /* 0 7 8 15 */ /* station address byte 1 station address byte 0 */ /* 16 23 24 31 */ /* reserved reserved */ value = 0; value |= (u32) ((address_byte_0 << 16) & 0x00FF0000); value |= (u32) ((address_byte_1 << 24) & 0xFF000000); out_be32(macstnaddr2_register, value); return 0; } static int init_check_frame_length_mode(int length_check, u32 __iomem *maccfg2_register) { u32 value = 0; value = in_be32(maccfg2_register); if (length_check) value |= MACCFG2_LC; else value &= ~MACCFG2_LC; out_be32(maccfg2_register, value); return 0; } static int init_preamble_length(u8 preamble_length, u32 __iomem *maccfg2_register) { if ((preamble_length < 3) || (preamble_length > 7)) return -EINVAL; clrsetbits_be32(maccfg2_register, MACCFG2_PREL_MASK, preamble_length << MACCFG2_PREL_SHIFT); return 0; } static int init_rx_parameters(int reject_broadcast, int receive_short_frames, int promiscuous, u32 __iomem *upsmr_register) { u32 value = 0; value = in_be32(upsmr_register); if (reject_broadcast) value |= UCC_GETH_UPSMR_BRO; else value &= ~UCC_GETH_UPSMR_BRO; if (receive_short_frames) value |= UCC_GETH_UPSMR_RSH; else value &= ~UCC_GETH_UPSMR_RSH; if (promiscuous) value |= UCC_GETH_UPSMR_PRO; else value &= ~UCC_GETH_UPSMR_PRO; out_be32(upsmr_register, value); return 0; } static int init_max_rx_buff_len(u16 max_rx_buf_len, u16 __iomem *mrblr_register) { /* max_rx_buf_len value must be a multiple of 128 */ if ((max_rx_buf_len == 0) || (max_rx_buf_len % UCC_GETH_MRBLR_ALIGNMENT)) return -EINVAL; out_be16(mrblr_register, max_rx_buf_len); return 0; } static int init_min_frame_len(u16 min_frame_length, u16 __iomem *minflr_register, u16 __iomem *mrblr_register) { u16 mrblr_value = 0; mrblr_value = in_be16(mrblr_register); if (min_frame_length >= (mrblr_value - 4)) return -EINVAL; out_be16(minflr_register, min_frame_length); return 0; } static int adjust_enet_interface(struct ucc_geth_private *ugeth) { struct ucc_geth_info *ug_info; struct ucc_geth __iomem *ug_regs; struct ucc_fast __iomem *uf_regs; int ret_val; u32 upsmr, maccfg2; u16 value; ugeth_vdbg("%s: IN", __func__); ug_info = ugeth->ug_info; ug_regs = ugeth->ug_regs; uf_regs = ugeth->uccf->uf_regs; /* Set MACCFG2 */ maccfg2 = in_be32(&ug_regs->maccfg2); maccfg2 &= ~MACCFG2_INTERFACE_MODE_MASK; if ((ugeth->max_speed == SPEED_10) || (ugeth->max_speed == SPEED_100)) maccfg2 |= MACCFG2_INTERFACE_MODE_NIBBLE; else if (ugeth->max_speed == SPEED_1000) maccfg2 |= MACCFG2_INTERFACE_MODE_BYTE; maccfg2 |= ug_info->padAndCrc; out_be32(&ug_regs->maccfg2, maccfg2); /* Set UPSMR */ upsmr = in_be32(&uf_regs->upsmr); upsmr &= ~(UCC_GETH_UPSMR_RPM | UCC_GETH_UPSMR_R10M | UCC_GETH_UPSMR_TBIM | UCC_GETH_UPSMR_RMM); if ((ugeth->phy_interface == PHY_INTERFACE_MODE_RMII) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII_ID) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII_RXID) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII_TXID) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RTBI)) { if (ugeth->phy_interface != PHY_INTERFACE_MODE_RMII) upsmr |= UCC_GETH_UPSMR_RPM; switch (ugeth->max_speed) { case SPEED_10: upsmr |= UCC_GETH_UPSMR_R10M; fallthrough; case SPEED_100: if (ugeth->phy_interface != PHY_INTERFACE_MODE_RTBI) upsmr |= UCC_GETH_UPSMR_RMM; } } if ((ugeth->phy_interface == PHY_INTERFACE_MODE_TBI) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RTBI)) { upsmr |= UCC_GETH_UPSMR_TBIM; } if (ugeth->phy_interface == PHY_INTERFACE_MODE_SGMII) upsmr |= UCC_GETH_UPSMR_SGMM; out_be32(&uf_regs->upsmr, upsmr); /* Disable autonegotiation in tbi mode, because by default it comes up in autonegotiation mode. */ /* Note that this depends on proper setting in utbipar register. */ if ((ugeth->phy_interface == PHY_INTERFACE_MODE_TBI) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RTBI)) { struct ucc_geth_info *ug_info = ugeth->ug_info; struct phy_device *tbiphy; if (!ug_info->tbi_node) pr_warn("TBI mode requires that the device tree specify a tbi-handle\n"); tbiphy = of_phy_find_device(ug_info->tbi_node); if (!tbiphy) pr_warn("Could not get TBI device\n"); value = phy_read(tbiphy, ENET_TBI_MII_CR); value &= ~0x1000; /* Turn off autonegotiation */ phy_write(tbiphy, ENET_TBI_MII_CR, value); put_device(&tbiphy->mdio.dev); } init_check_frame_length_mode(ug_info->lengthCheckRx, &ug_regs->maccfg2); ret_val = init_preamble_length(ug_info->prel, &ug_regs->maccfg2); if (ret_val != 0) { if (netif_msg_probe(ugeth)) pr_err("Preamble length must be between 3 and 7 inclusive\n"); return ret_val; } return 0; } static int ugeth_graceful_stop_tx(struct ucc_geth_private *ugeth) { struct ucc_fast_private *uccf; u32 cecr_subblock; u32 temp; int i = 10; uccf = ugeth->uccf; /* Mask GRACEFUL STOP TX interrupt bit and clear it */ clrbits32(uccf->p_uccm, UCC_GETH_UCCE_GRA); out_be32(uccf->p_ucce, UCC_GETH_UCCE_GRA); /* clear by writing 1 */ /* Issue host command */ cecr_subblock = ucc_fast_get_qe_cr_subblock(ugeth->ug_info->uf_info.ucc_num); qe_issue_cmd(QE_GRACEFUL_STOP_TX, cecr_subblock, QE_CR_PROTOCOL_ETHERNET, 0); /* Wait for command to complete */ do { msleep(10); temp = in_be32(uccf->p_ucce); } while (!(temp & UCC_GETH_UCCE_GRA) && --i); uccf->stopped_tx = 1; return 0; } static int ugeth_graceful_stop_rx(struct ucc_geth_private *ugeth) { struct ucc_fast_private *uccf; u32 cecr_subblock; u8 temp; int i = 10; uccf = ugeth->uccf; /* Clear acknowledge bit */ temp = in_8(&ugeth->p_rx_glbl_pram->rxgstpack); temp &= ~GRACEFUL_STOP_ACKNOWLEDGE_RX; out_8(&ugeth->p_rx_glbl_pram->rxgstpack, temp); /* Keep issuing command and checking acknowledge bit until it is asserted, according to spec */ do { /* Issue host command */ cecr_subblock = ucc_fast_get_qe_cr_subblock(ugeth->ug_info->uf_info. ucc_num); qe_issue_cmd(QE_GRACEFUL_STOP_RX, cecr_subblock, QE_CR_PROTOCOL_ETHERNET, 0); msleep(10); temp = in_8(&ugeth->p_rx_glbl_pram->rxgstpack); } while (!(temp & GRACEFUL_STOP_ACKNOWLEDGE_RX) && --i); uccf->stopped_rx = 1; return 0; } static int ugeth_restart_tx(struct ucc_geth_private *ugeth) { struct ucc_fast_private *uccf; u32 cecr_subblock; uccf = ugeth->uccf; cecr_subblock = ucc_fast_get_qe_cr_subblock(ugeth->ug_info->uf_info.ucc_num); qe_issue_cmd(QE_RESTART_TX, cecr_subblock, QE_CR_PROTOCOL_ETHERNET, 0); uccf->stopped_tx = 0; return 0; } static int ugeth_restart_rx(struct ucc_geth_private *ugeth) { struct ucc_fast_private *uccf; u32 cecr_subblock; uccf = ugeth->uccf; cecr_subblock = ucc_fast_get_qe_cr_subblock(ugeth->ug_info->uf_info.ucc_num); qe_issue_cmd(QE_RESTART_RX, cecr_subblock, QE_CR_PROTOCOL_ETHERNET, 0); uccf->stopped_rx = 0; return 0; } static int ugeth_enable(struct ucc_geth_private *ugeth, enum comm_dir mode) { struct ucc_fast_private *uccf; int enabled_tx, enabled_rx; uccf = ugeth->uccf; /* check if the UCC number is in range. */ if (ugeth->ug_info->uf_info.ucc_num >= UCC_MAX_NUM) { if (netif_msg_probe(ugeth)) pr_err("ucc_num out of range\n"); return -EINVAL; } enabled_tx = uccf->enabled_tx; enabled_rx = uccf->enabled_rx; /* Get Tx and Rx going again, in case this channel was actively disabled. */ if ((mode & COMM_DIR_TX) && (!enabled_tx) && uccf->stopped_tx) ugeth_restart_tx(ugeth); if ((mode & COMM_DIR_RX) && (!enabled_rx) && uccf->stopped_rx) ugeth_restart_rx(ugeth); ucc_fast_enable(uccf, mode); /* OK to do even if not disabled */ return 0; } static int ugeth_disable(struct ucc_geth_private *ugeth, enum comm_dir mode) { struct ucc_fast_private *uccf; uccf = ugeth->uccf; /* check if the UCC number is in range. */ if (ugeth->ug_info->uf_info.ucc_num >= UCC_MAX_NUM) { if (netif_msg_probe(ugeth)) pr_err("ucc_num out of range\n"); return -EINVAL; } /* Stop any transmissions */ if ((mode & COMM_DIR_TX) && uccf->enabled_tx && !uccf->stopped_tx) ugeth_graceful_stop_tx(ugeth); /* Stop any receptions */ if ((mode & COMM_DIR_RX) && uccf->enabled_rx && !uccf->stopped_rx) ugeth_graceful_stop_rx(ugeth); ucc_fast_disable(ugeth->uccf, mode); /* OK to do even if not enabled */ return 0; } static void ugeth_quiesce(struct ucc_geth_private *ugeth) { /* Prevent any further xmits */ netif_tx_stop_all_queues(ugeth->ndev); /* Disable the interrupt to avoid NAPI rescheduling. */ disable_irq(ugeth->ug_info->uf_info.irq); /* Stop NAPI, and possibly wait for its completion. */ napi_disable(&ugeth->napi); } static void ugeth_activate(struct ucc_geth_private *ugeth) { napi_enable(&ugeth->napi); enable_irq(ugeth->ug_info->uf_info.irq); /* allow to xmit again */ netif_tx_wake_all_queues(ugeth->ndev); __netdev_watchdog_up(ugeth->ndev); } /* Called every time the controller might need to be made * aware of new link state. The PHY code conveys this * information through variables in the ugeth structure, and this * function converts those variables into the appropriate * register values, and can bring down the device if needed. */ static void adjust_link(struct net_device *dev) { struct ucc_geth_private *ugeth = netdev_priv(dev); struct ucc_geth __iomem *ug_regs; struct ucc_fast __iomem *uf_regs; struct phy_device *phydev = ugeth->phydev; int new_state = 0; ug_regs = ugeth->ug_regs; uf_regs = ugeth->uccf->uf_regs; if (phydev->link) { u32 tempval = in_be32(&ug_regs->maccfg2); u32 upsmr = in_be32(&uf_regs->upsmr); /* Now we make sure that we can be in full duplex mode. * If not, we operate in half-duplex mode. */ if (phydev->duplex != ugeth->oldduplex) { new_state = 1; if (!(phydev->duplex)) tempval &= ~(MACCFG2_FDX); else tempval |= MACCFG2_FDX; ugeth->oldduplex = phydev->duplex; } if (phydev->speed != ugeth->oldspeed) { new_state = 1; switch (phydev->speed) { case SPEED_1000: tempval = ((tempval & ~(MACCFG2_INTERFACE_MODE_MASK)) | MACCFG2_INTERFACE_MODE_BYTE); break; case SPEED_100: case SPEED_10: tempval = ((tempval & ~(MACCFG2_INTERFACE_MODE_MASK)) | MACCFG2_INTERFACE_MODE_NIBBLE); /* if reduced mode, re-set UPSMR.R10M */ if ((ugeth->phy_interface == PHY_INTERFACE_MODE_RMII) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII_ID) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII_RXID) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RGMII_TXID) || (ugeth->phy_interface == PHY_INTERFACE_MODE_RTBI)) { if (phydev->speed == SPEED_10) upsmr |= UCC_GETH_UPSMR_R10M; else upsmr &= ~UCC_GETH_UPSMR_R10M; } break; default: if (netif_msg_link(ugeth)) pr_warn( "%s: Ack! Speed (%d) is not 10/100/1000!", dev->name, phydev->speed); break; } ugeth->oldspeed = phydev->speed; } if (!ugeth->oldlink) { new_state = 1; ugeth->oldlink = 1; } if (new_state) { /* * To change the MAC configuration we need to disable * the controller. To do so, we have to either grab * ugeth->lock, which is a bad idea since 'graceful * stop' commands might take quite a while, or we can * quiesce driver's activity. */ ugeth_quiesce(ugeth); ugeth_disable(ugeth, COMM_DIR_RX_AND_TX); out_be32(&ug_regs->maccfg2, tempval); out_be32(&uf_regs->upsmr, upsmr); ugeth_enable(ugeth, COMM_DIR_RX_AND_TX); ugeth_activate(ugeth); } } else if (ugeth->oldlink) { new_state = 1; ugeth->oldlink = 0; ugeth->oldspeed = 0; ugeth->oldduplex = -1; } if (new_state && netif_msg_link(ugeth)) phy_print_status(phydev); } /* Initialize TBI PHY interface for communicating with the * SERDES lynx PHY on the chip. We communicate with this PHY * through the MDIO bus on each controller, treating it as a * "normal" PHY at the address found in the UTBIPA register. We assume * that the UTBIPA register is valid. Either the MDIO bus code will set * it to a value that doesn't conflict with other PHYs on the bus, or the * value doesn't matter, as there are no other PHYs on the bus. */ static void uec_configure_serdes(struct net_device *dev) { struct ucc_geth_private *ugeth = netdev_priv(dev); struct ucc_geth_info *ug_info = ugeth->ug_info; struct phy_device *tbiphy; if (!ug_info->tbi_node) { dev_warn(&dev->dev, "SGMII mode requires that the device " "tree specify a tbi-handle\n"); return; } tbiphy = of_phy_find_device(ug_info->tbi_node); if (!tbiphy) { dev_err(&dev->dev, "error: Could not get TBI device\n"); return; } /* * If the link is already up, we must already be ok, and don't need to * configure and reset the TBI<->SerDes link. Maybe U-Boot configured * everything for us? Resetting it takes the link down and requires * several seconds for it to come back. */ if (phy_read(tbiphy, ENET_TBI_MII_SR) & TBISR_LSTATUS) { put_device(&tbiphy->mdio.dev); return; } /* Single clk mode, mii mode off(for serdes communication) */ phy_write(tbiphy, ENET_TBI_MII_ANA, TBIANA_SETTINGS); phy_write(tbiphy, ENET_TBI_MII_TBICON, TBICON_CLK_SELECT); phy_write(tbiphy, ENET_TBI_MII_CR, TBICR_SETTINGS); put_device(&tbiphy->mdio.dev); } /* Configure the PHY for dev. * returns 0 if success. -1 if failure */ static int init_phy(struct net_device *dev) { struct ucc_geth_private *priv = netdev_priv(dev); struct ucc_geth_info *ug_info = priv->ug_info; struct phy_device *phydev; priv->oldlink = 0; priv->oldspeed = 0; priv->oldduplex = -1; phydev = of_phy_connect(dev, ug_info->phy_node, &adjust_link, 0, priv->phy_interface); if (!phydev) { dev_err(&dev->dev, "Could not attach to PHY\n"); return -ENODEV; } if (priv->phy_interface == PHY_INTERFACE_MODE_SGMII) uec_configure_serdes(dev); phy_set_max_speed(phydev, priv->max_speed); priv->phydev = phydev; return 0; } static void ugeth_dump_regs(struct ucc_geth_private *ugeth) { #ifdef DEBUG ucc_fast_dump_regs(ugeth->uccf); dump_regs(ugeth); dump_bds(ugeth); #endif } static int ugeth_82xx_filtering_clear_all_addr_in_hash(struct ucc_geth_private * ugeth, enum enet_addr_type enet_addr_type) { struct ucc_geth_82xx_address_filtering_pram __iomem *p_82xx_addr_filt; struct ucc_fast_private *uccf; enum comm_dir comm_dir; struct list_head *p_lh; u16 i, num; u32 __iomem *addr_h; u32 __iomem *addr_l; u8 *p_counter; uccf = ugeth->uccf; p_82xx_addr_filt = (struct ucc_geth_82xx_address_filtering_pram __iomem *) ugeth->p_rx_glbl_pram->addressfiltering; if (enet_addr_type == ENET_ADDR_TYPE_GROUP) { addr_h = &(p_82xx_addr_filt->gaddr_h); addr_l = &(p_82xx_addr_filt->gaddr_l); p_lh = &ugeth->group_hash_q; p_counter = &(ugeth->numGroupAddrInHash); } else if (enet_addr_type == ENET_ADDR_TYPE_INDIVIDUAL) { addr_h = &(p_82xx_addr_filt->iaddr_h); addr_l = &(p_82xx_addr_filt->iaddr_l); p_lh = &ugeth->ind_hash_q; p_counter = &(ugeth->numIndAddrInHash); } else return -EINVAL; comm_dir = 0; if (uccf->enabled_tx) comm_dir |= COMM_DIR_TX; if (uccf->enabled_rx) comm_dir |= COMM_DIR_RX; if (comm_dir) ugeth_disable(ugeth, comm_dir); /* Clear the hash table. */ out_be32(addr_h, 0x00000000); out_be32(addr_l, 0x00000000); if (!p_lh) return 0; num = *p_counter; /* Delete all remaining CQ elements */ for (i = 0; i < num; i++) put_enet_addr_container(ENET_ADDR_CONT_ENTRY(dequeue(p_lh))); *p_counter = 0; if (comm_dir) ugeth_enable(ugeth, comm_dir); return 0; } static int ugeth_82xx_filtering_clear_addr_in_paddr(struct ucc_geth_private *ugeth, u8 paddr_num) { ugeth->indAddrRegUsed[paddr_num] = 0; /* mark this paddr as not used */ return hw_clear_addr_in_paddr(ugeth, paddr_num);/* clear in hardware */ } static void ucc_geth_free_rx(struct ucc_geth_private *ugeth) { struct ucc_geth_info *ug_info; struct ucc_fast_info *uf_info; u16 i, j; u8 __iomem *bd; ug_info = ugeth->ug_info; uf_info = &ug_info->uf_info; for (i = 0; i < ucc_geth_rx_queues(ugeth->ug_info); i++) { if (ugeth->p_rx_bd_ring[i]) { /* Return existing data buffers in ring */ bd = ugeth->p_rx_bd_ring[i]; for (j = 0; j < ugeth->ug_info->bdRingLenRx[i]; j++) { if (ugeth->rx_skbuff[i][j]) { dma_unmap_single(ugeth->dev, in_be32(&((struct qe_bd __iomem *)bd)->buf), ugeth->ug_info-> uf_info.max_rx_buf_length + UCC_GETH_RX_DATA_BUF_ALIGNMENT, DMA_FROM_DEVICE); dev_kfree_skb_any( ugeth->rx_skbuff[i][j]); ugeth->rx_skbuff[i][j] = NULL; } bd += sizeof(struct qe_bd); } kfree(ugeth->rx_skbuff[i]); kfree(ugeth->p_rx_bd_ring[i]); ugeth->p_rx_bd_ring[i] = NULL; } } } static void ucc_geth_free_tx(struct ucc_geth_private *ugeth) { struct ucc_geth_info *ug_info; struct ucc_fast_info *uf_info; u16 i, j; u8 __iomem *bd; netdev_reset_queue(ugeth->ndev); ug_info = ugeth->ug_info; uf_info = &ug_info->uf_info; for (i = 0; i < ucc_geth_tx_queues(ugeth->ug_info); i++) { bd = ugeth->p_tx_bd_ring[i]; if (!bd) continue; for (j = 0; j < ugeth->ug_info->bdRingLenTx[i]; j++) { if (ugeth->tx_skbuff[i][j]) { dma_unmap_single(ugeth->dev, in_be32(&((struct qe_bd __iomem *)bd)->buf), (in_be32((u32 __iomem *)bd) & BD_LENGTH_MASK), DMA_TO_DEVICE); dev_kfree_skb_any(ugeth->tx_skbuff[i][j]); ugeth->tx_skbuff[i][j] = NULL; } } kfree(ugeth->tx_skbuff[i]); kfree(ugeth->p_tx_bd_ring[i]); ugeth->p_tx_bd_ring[i] = NULL; } } static void ucc_geth_memclean(struct ucc_geth_private *ugeth) { if (!ugeth) return; if (ugeth->uccf) { ucc_fast_free(ugeth->uccf); ugeth->uccf = NULL; } qe_muram_free_addr(ugeth->p_thread_data_tx); ugeth->p_thread_data_tx = NULL; qe_muram_free_addr(ugeth->p_thread_data_rx); ugeth->p_thread_data_rx = NULL; qe_muram_free_addr(ugeth->p_exf_glbl_param); ugeth->p_exf_glbl_param = NULL; qe_muram_free_addr(ugeth->p_rx_glbl_pram); ugeth->p_rx_glbl_pram = NULL; qe_muram_free_addr(ugeth->p_tx_glbl_pram); ugeth->p_tx_glbl_pram = NULL; qe_muram_free_addr(ugeth->p_send_q_mem_reg); ugeth->p_send_q_mem_reg = NULL; qe_muram_free_addr(ugeth->p_scheduler); ugeth->p_scheduler = NULL; qe_muram_free_addr(ugeth->p_tx_fw_statistics_pram); ugeth->p_tx_fw_statistics_pram = NULL; qe_muram_free_addr(ugeth->p_rx_fw_statistics_pram); ugeth->p_rx_fw_statistics_pram = NULL; qe_muram_free_addr(ugeth->p_rx_irq_coalescing_tbl); ugeth->p_rx_irq_coalescing_tbl = NULL; qe_muram_free_addr(ugeth->p_rx_bd_qs_tbl); ugeth->p_rx_bd_qs_tbl = NULL; if (ugeth->p_init_enet_param_shadow) { return_init_enet_entries(ugeth, &(ugeth->p_init_enet_param_shadow-> rxthread[0]), ENET_INIT_PARAM_MAX_ENTRIES_RX, ugeth->ug_info->riscRx, 1); return_init_enet_entries(ugeth, &(ugeth->p_init_enet_param_shadow-> txthread[0]), ENET_INIT_PARAM_MAX_ENTRIES_TX, ugeth->ug_info->riscTx, 0); kfree(ugeth->p_init_enet_param_shadow); ugeth->p_init_enet_param_shadow = NULL; } ucc_geth_free_tx(ugeth); ucc_geth_free_rx(ugeth); while (!list_empty(&ugeth->group_hash_q)) put_enet_addr_container(ENET_ADDR_CONT_ENTRY (dequeue(&ugeth->group_hash_q))); while (!list_empty(&ugeth->ind_hash_q)) put_enet_addr_container(ENET_ADDR_CONT_ENTRY (dequeue(&ugeth->ind_hash_q))); if (ugeth->ug_regs) { iounmap(ugeth->ug_regs); ugeth->ug_regs = NULL; } } static void ucc_geth_set_multi(struct net_device *dev) { struct ucc_geth_private *ugeth; struct netdev_hw_addr *ha; struct ucc_fast __iomem *uf_regs; struct ucc_geth_82xx_address_filtering_pram __iomem *p_82xx_addr_filt; ugeth = netdev_priv(dev); uf_regs = ugeth->uccf->uf_regs; if (dev->flags & IFF_PROMISC) { setbits32(&uf_regs->upsmr, UCC_GETH_UPSMR_PRO); } else { clrbits32(&uf_regs->upsmr, UCC_GETH_UPSMR_PRO); p_82xx_addr_filt = (struct ucc_geth_82xx_address_filtering_pram __iomem *) ugeth-> p_rx_glbl_pram->addressfiltering; if (dev->flags & IFF_ALLMULTI) { /* Catch all multicast addresses, so set the * filter to all 1's. */ out_be32(&p_82xx_addr_filt->gaddr_h, 0xffffffff); out_be32(&p_82xx_addr_filt->gaddr_l, 0xffffffff); } else { /* Clear filter and add the addresses in the list. */ out_be32(&p_82xx_addr_filt->gaddr_h, 0x0); out_be32(&p_82xx_addr_filt->gaddr_l, 0x0); netdev_for_each_mc_addr(ha, dev) { /* Ask CPM to run CRC and set bit in * filter mask. */ hw_add_addr_in_hash(ugeth, ha->addr); } } } } static void ucc_geth_stop(struct ucc_geth_private *ugeth) { struct ucc_geth __iomem *ug_regs = ugeth->ug_regs; struct phy_device *phydev = ugeth->phydev; ugeth_vdbg("%s: IN", __func__); /* * Tell the kernel the link is down. * Must be done before disabling the controller * or deadlock may happen. */ phy_stop(phydev); /* Disable the controller */ ugeth_disable(ugeth, COMM_DIR_RX_AND_TX); /* Mask all interrupts */ out_be32(ugeth->uccf->p_uccm, 0x00000000); /* Clear all interrupts */ out_be32(ugeth->uccf->p_ucce, 0xffffffff); /* Disable Rx and Tx */ clrbits32(&ug_regs->maccfg1, MACCFG1_ENABLE_RX | MACCFG1_ENABLE_TX); ucc_geth_memclean(ugeth); } static int ucc_struct_init(struct ucc_geth_private *ugeth) { struct ucc_geth_info *ug_info; struct ucc_fast_info *uf_info; int i; ug_info = ugeth->ug_info; uf_info = &ug_info->uf_info; /* Rx BD lengths */ for (i = 0; i < ucc_geth_rx_queues(ug_info); i++) { if ((ug_info->bdRingLenRx[i] < UCC_GETH_RX_BD_RING_SIZE_MIN) || (ug_info->bdRingLenRx[i] % UCC_GETH_RX_BD_RING_SIZE_ALIGNMENT)) { if (netif_msg_probe(ugeth)) pr_err("Rx BD ring length must be multiple of 4, no smaller than 8\n"); return -EINVAL; } } /* Tx BD lengths */ for (i = 0; i < ucc_geth_tx_queues(ug_info); i++) { if (ug_info->bdRingLenTx[i] < UCC_GETH_TX_BD_RING_SIZE_MIN) { if (netif_msg_probe(ugeth)) pr_err("Tx BD ring length must be no smaller than 2\n"); return -EINVAL; } } /* mrblr */ if ((uf_info->max_rx_buf_length == 0) || (uf_info->max_rx_buf_length % UCC_GETH_MRBLR_ALIGNMENT)) { if (netif_msg_probe(ugeth)) pr_err("max_rx_buf_length must be non-zero multiple of 128\n"); return -EINVAL; } /* num Tx queues */ if (ucc_geth_tx_queues(ug_info) > NUM_TX_QUEUES) { if (netif_msg_probe(ugeth)) pr_err("number of tx queues too large\n"); return -EINVAL; } /* num Rx queues */ if (ucc_geth_rx_queues(ug_info) > NUM_RX_QUEUES) { if (netif_msg_probe(ugeth)) pr_err("number of rx queues too large\n"); return -EINVAL; } /* l2qt */ for (i = 0; i < UCC_GETH_VLAN_PRIORITY_MAX; i++) { if (ug_info->l2qt[i] >= ucc_geth_rx_queues(ug_info)) { if (netif_msg_probe(ugeth)) pr_err("VLAN priority table entry must not be larger than number of Rx queues\n"); return -EINVAL; } } /* l3qt */ for (i = 0; i < UCC_GETH_IP_PRIORITY_MAX; i++) { if (ug_info->l3qt[i] >= ucc_geth_rx_queues(ug_info)) { if (netif_msg_probe(ugeth)) pr_err("IP priority table entry must not be larger than number of Rx queues\n"); return -EINVAL; } } if (ug_info->cam && !ug_info->ecamptr) { if (netif_msg_probe(ugeth)) pr_err("If cam mode is chosen, must supply cam ptr\n"); return -EINVAL; } if ((ug_info->numStationAddresses != UCC_GETH_NUM_OF_STATION_ADDRESSES_1) && ug_info->rxExtendedFiltering) { if (netif_msg_probe(ugeth)) pr_err("Number of station addresses greater than 1 not allowed in extended parsing mode\n"); return -EINVAL; } /* Generate uccm_mask for receive */ uf_info->uccm_mask = ug_info->eventRegMask & UCCE_OTHER;/* Errors */ for (i = 0; i < ucc_geth_rx_queues(ug_info); i++) uf_info->uccm_mask |= (UCC_GETH_UCCE_RXF0 << i); for (i = 0; i < ucc_geth_tx_queues(ug_info); i++) uf_info->uccm_mask |= (UCC_GETH_UCCE_TXB0 << i); /* Initialize the general fast UCC block. */ if (ucc_fast_init(uf_info, &ugeth->uccf)) { if (netif_msg_probe(ugeth)) pr_err("Failed to init uccf\n"); return -ENOMEM; } /* read the number of risc engines, update the riscTx and riscRx * if there are 4 riscs in QE */ if (qe_get_num_of_risc() == 4) { ug_info->riscTx = QE_RISC_ALLOCATION_FOUR_RISCS; ug_info->riscRx = QE_RISC_ALLOCATION_FOUR_RISCS; } ugeth->ug_regs = ioremap(uf_info->regs, sizeof(*ugeth->ug_regs)); if (!ugeth->ug_regs) { if (netif_msg_probe(ugeth)) pr_err("Failed to ioremap regs\n"); return -ENOMEM; } return 0; } static int ucc_geth_alloc_tx(struct ucc_geth_private *ugeth) { struct ucc_geth_info *ug_info; struct ucc_fast_info *uf_info; int length; u16 i, j; u8 __iomem *bd; ug_info = ugeth->ug_info; uf_info = &ug_info->uf_info; /* Allocate Tx bds */ for (j = 0; j < ucc_geth_tx_queues(ug_info); j++) { u32 align = max(UCC_GETH_TX_BD_RING_ALIGNMENT, UCC_GETH_TX_BD_RING_SIZE_MEMORY_ALIGNMENT); u32 alloc; length = ug_info->bdRingLenTx[j] * sizeof(struct qe_bd); alloc = round_up(length, align); alloc = roundup_pow_of_two(alloc); ugeth->p_tx_bd_ring[j] = kmalloc(alloc, GFP_KERNEL); if (!ugeth->p_tx_bd_ring[j]) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate memory for Tx bd rings\n"); return -ENOMEM; } /* Zero unused end of bd ring, according to spec */ memset(ugeth->p_tx_bd_ring[j] + length, 0, alloc - length); } /* Init Tx bds */ for (j = 0; j < ucc_geth_tx_queues(ug_info); j++) { /* Setup the skbuff rings */ ugeth->tx_skbuff[j] = kcalloc(ugeth->ug_info->bdRingLenTx[j], sizeof(struct sk_buff *), GFP_KERNEL); if (ugeth->tx_skbuff[j] == NULL) { if (netif_msg_ifup(ugeth)) pr_err("Could not allocate tx_skbuff\n"); return -ENOMEM; } ugeth->skb_curtx[j] = ugeth->skb_dirtytx[j] = 0; bd = ugeth->confBd[j] = ugeth->txBd[j] = ugeth->p_tx_bd_ring[j]; for (i = 0; i < ug_info->bdRingLenTx[j]; i++) { /* clear bd buffer */ out_be32(&((struct qe_bd __iomem *)bd)->buf, 0); /* set bd status and length */ out_be32((u32 __iomem *)bd, 0); bd += sizeof(struct qe_bd); } bd -= sizeof(struct qe_bd); /* set bd status and length */ out_be32((u32 __iomem *)bd, T_W); /* for last BD set Wrap bit */ } return 0; } static int ucc_geth_alloc_rx(struct ucc_geth_private *ugeth) { struct ucc_geth_info *ug_info; struct ucc_fast_info *uf_info; int length; u16 i, j; u8 __iomem *bd; ug_info = ugeth->ug_info; uf_info = &ug_info->uf_info; /* Allocate Rx bds */ for (j = 0; j < ucc_geth_rx_queues(ug_info); j++) { u32 align = UCC_GETH_RX_BD_RING_ALIGNMENT; u32 alloc; length = ug_info->bdRingLenRx[j] * sizeof(struct qe_bd); alloc = round_up(length, align); alloc = roundup_pow_of_two(alloc); ugeth->p_rx_bd_ring[j] = kmalloc(alloc, GFP_KERNEL); if (!ugeth->p_rx_bd_ring[j]) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate memory for Rx bd rings\n"); return -ENOMEM; } } /* Init Rx bds */ for (j = 0; j < ucc_geth_rx_queues(ug_info); j++) { /* Setup the skbuff rings */ ugeth->rx_skbuff[j] = kcalloc(ugeth->ug_info->bdRingLenRx[j], sizeof(struct sk_buff *), GFP_KERNEL); if (ugeth->rx_skbuff[j] == NULL) { if (netif_msg_ifup(ugeth)) pr_err("Could not allocate rx_skbuff\n"); return -ENOMEM; } ugeth->skb_currx[j] = 0; bd = ugeth->rxBd[j] = ugeth->p_rx_bd_ring[j]; for (i = 0; i < ug_info->bdRingLenRx[j]; i++) { /* set bd status and length */ out_be32((u32 __iomem *)bd, R_I); /* clear bd buffer */ out_be32(&((struct qe_bd __iomem *)bd)->buf, 0); bd += sizeof(struct qe_bd); } bd -= sizeof(struct qe_bd); /* set bd status and length */ out_be32((u32 __iomem *)bd, R_W); /* for last BD set Wrap bit */ } return 0; } static int ucc_geth_startup(struct ucc_geth_private *ugeth) { struct ucc_geth_82xx_address_filtering_pram __iomem *p_82xx_addr_filt; struct ucc_geth_init_pram __iomem *p_init_enet_pram; struct ucc_fast_private *uccf; struct ucc_geth_info *ug_info; struct ucc_fast_info *uf_info; struct ucc_fast __iomem *uf_regs; struct ucc_geth __iomem *ug_regs; int ret_val = -EINVAL; u32 remoder = UCC_GETH_REMODER_INIT; u32 init_enet_pram_offset, cecr_subblock, command; u32 ifstat, i, j, size, l2qt, l3qt; u16 temoder = UCC_GETH_TEMODER_INIT; u8 function_code = 0; u8 __iomem *endOfRing; u8 numThreadsRxNumerical, numThreadsTxNumerical; s32 rx_glbl_pram_offset, tx_glbl_pram_offset; ugeth_vdbg("%s: IN", __func__); uccf = ugeth->uccf; ug_info = ugeth->ug_info; uf_info = &ug_info->uf_info; uf_regs = uccf->uf_regs; ug_regs = ugeth->ug_regs; numThreadsRxNumerical = ucc_geth_thread_count(ug_info->numThreadsRx); if (!numThreadsRxNumerical) { if (netif_msg_ifup(ugeth)) pr_err("Bad number of Rx threads value\n"); return -EINVAL; } numThreadsTxNumerical = ucc_geth_thread_count(ug_info->numThreadsTx); if (!numThreadsTxNumerical) { if (netif_msg_ifup(ugeth)) pr_err("Bad number of Tx threads value\n"); return -EINVAL; } /* Calculate rx_extended_features */ ugeth->rx_non_dynamic_extended_features = ug_info->ipCheckSumCheck || ug_info->ipAddressAlignment || (ug_info->numStationAddresses != UCC_GETH_NUM_OF_STATION_ADDRESSES_1); ugeth->rx_extended_features = ugeth->rx_non_dynamic_extended_features || (ug_info->vlanOperationTagged != UCC_GETH_VLAN_OPERATION_TAGGED_NOP) || (ug_info->vlanOperationNonTagged != UCC_GETH_VLAN_OPERATION_NON_TAGGED_NOP); init_default_reg_vals(&uf_regs->upsmr, &ug_regs->maccfg1, &ug_regs->maccfg2); /* Set UPSMR */ /* For more details see the hardware spec. */ init_rx_parameters(ug_info->bro, ug_info->rsh, ug_info->pro, &uf_regs->upsmr); /* We're going to ignore other registers for now, */ /* except as needed to get up and running */ /* Set MACCFG1 */ /* For more details see the hardware spec. */ init_flow_control_params(ug_info->aufc, ug_info->receiveFlowControl, ug_info->transmitFlowControl, ug_info->pausePeriod, ug_info->extensionField, &uf_regs->upsmr, &ug_regs->uempr, &ug_regs->maccfg1); setbits32(&ug_regs->maccfg1, MACCFG1_ENABLE_RX | MACCFG1_ENABLE_TX); /* Set IPGIFG */ /* For more details see the hardware spec. */ ret_val = init_inter_frame_gap_params(ug_info->nonBackToBackIfgPart1, ug_info->nonBackToBackIfgPart2, ug_info-> miminumInterFrameGapEnforcement, ug_info->backToBackInterFrameGap, &ug_regs->ipgifg); if (ret_val != 0) { if (netif_msg_ifup(ugeth)) pr_err("IPGIFG initialization parameter too large\n"); return ret_val; } /* Set HAFDUP */ /* For more details see the hardware spec. */ ret_val = init_half_duplex_params(ug_info->altBeb, ug_info->backPressureNoBackoff, ug_info->noBackoff, ug_info->excessDefer, ug_info->altBebTruncation, ug_info->maxRetransmission, ug_info->collisionWindow, &ug_regs->hafdup); if (ret_val != 0) { if (netif_msg_ifup(ugeth)) pr_err("Half Duplex initialization parameter too large\n"); return ret_val; } /* Set IFSTAT */ /* For more details see the hardware spec. */ /* Read only - resets upon read */ ifstat = in_be32(&ug_regs->ifstat); /* Clear UEMPR */ /* For more details see the hardware spec. */ out_be32(&ug_regs->uempr, 0); /* Set UESCR */ /* For more details see the hardware spec. */ init_hw_statistics_gathering_mode((ug_info->statisticsMode & UCC_GETH_STATISTICS_GATHERING_MODE_HARDWARE), 0, &uf_regs->upsmr, &ug_regs->uescr); ret_val = ucc_geth_alloc_tx(ugeth); if (ret_val != 0) return ret_val; ret_val = ucc_geth_alloc_rx(ugeth); if (ret_val != 0) return ret_val; /* * Global PRAM */ /* Tx global PRAM */ /* Allocate global tx parameter RAM page */ tx_glbl_pram_offset = qe_muram_alloc(sizeof(struct ucc_geth_tx_global_pram), UCC_GETH_TX_GLOBAL_PRAM_ALIGNMENT); if (tx_glbl_pram_offset < 0) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_tx_glbl_pram\n"); return -ENOMEM; } ugeth->p_tx_glbl_pram = qe_muram_addr(tx_glbl_pram_offset); /* Fill global PRAM */ /* TQPTR */ /* Size varies with number of Tx threads */ ugeth->thread_dat_tx_offset = qe_muram_alloc(numThreadsTxNumerical * sizeof(struct ucc_geth_thread_data_tx) + 32 * (numThreadsTxNumerical == 1), UCC_GETH_THREAD_DATA_ALIGNMENT); if (IS_ERR_VALUE(ugeth->thread_dat_tx_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_thread_data_tx\n"); return -ENOMEM; } ugeth->p_thread_data_tx = (struct ucc_geth_thread_data_tx __iomem *) qe_muram_addr(ugeth-> thread_dat_tx_offset); out_be32(&ugeth->p_tx_glbl_pram->tqptr, ugeth->thread_dat_tx_offset); /* vtagtable */ for (i = 0; i < UCC_GETH_TX_VTAG_TABLE_ENTRY_MAX; i++) out_be32(&ugeth->p_tx_glbl_pram->vtagtable[i], ug_info->vtagtable[i]); /* iphoffset */ for (i = 0; i < TX_IP_OFFSET_ENTRY_MAX; i++) out_8(&ugeth->p_tx_glbl_pram->iphoffset[i], ug_info->iphoffset[i]); /* SQPTR */ /* Size varies with number of Tx queues */ ugeth->send_q_mem_reg_offset = qe_muram_alloc(ucc_geth_tx_queues(ug_info) * sizeof(struct ucc_geth_send_queue_qd), UCC_GETH_SEND_QUEUE_QUEUE_DESCRIPTOR_ALIGNMENT); if (IS_ERR_VALUE(ugeth->send_q_mem_reg_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_send_q_mem_reg\n"); return -ENOMEM; } ugeth->p_send_q_mem_reg = (struct ucc_geth_send_queue_mem_region __iomem *) qe_muram_addr(ugeth-> send_q_mem_reg_offset); out_be32(&ugeth->p_tx_glbl_pram->sqptr, ugeth->send_q_mem_reg_offset); /* Setup the table */ /* Assume BD rings are already established */ for (i = 0; i < ucc_geth_tx_queues(ug_info); i++) { endOfRing = ugeth->p_tx_bd_ring[i] + (ug_info->bdRingLenTx[i] - 1) * sizeof(struct qe_bd); out_be32(&ugeth->p_send_q_mem_reg->sqqd[i].bd_ring_base, (u32) virt_to_phys(ugeth->p_tx_bd_ring[i])); out_be32(&ugeth->p_send_q_mem_reg->sqqd[i]. last_bd_completed_address, (u32) virt_to_phys(endOfRing)); } /* schedulerbasepointer */ if (ucc_geth_tx_queues(ug_info) > 1) { /* scheduler exists only if more than 1 tx queue */ ugeth->scheduler_offset = qe_muram_alloc(sizeof(struct ucc_geth_scheduler), UCC_GETH_SCHEDULER_ALIGNMENT); if (IS_ERR_VALUE(ugeth->scheduler_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_scheduler\n"); return -ENOMEM; } ugeth->p_scheduler = (struct ucc_geth_scheduler __iomem *) qe_muram_addr(ugeth-> scheduler_offset); out_be32(&ugeth->p_tx_glbl_pram->schedulerbasepointer, ugeth->scheduler_offset); /* Set values in scheduler */ out_be32(&ugeth->p_scheduler->mblinterval, ug_info->mblinterval); out_be16(&ugeth->p_scheduler->nortsrbytetime, ug_info->nortsrbytetime); out_8(&ugeth->p_scheduler->fracsiz, ug_info->fracsiz); out_8(&ugeth->p_scheduler->strictpriorityq, ug_info->strictpriorityq); out_8(&ugeth->p_scheduler->txasap, ug_info->txasap); out_8(&ugeth->p_scheduler->extrabw, ug_info->extrabw); for (i = 0; i < NUM_TX_QUEUES; i++) out_8(&ugeth->p_scheduler->weightfactor[i], ug_info->weightfactor[i]); /* Set pointers to cpucount registers in scheduler */ ugeth->p_cpucount[0] = &(ugeth->p_scheduler->cpucount0); ugeth->p_cpucount[1] = &(ugeth->p_scheduler->cpucount1); ugeth->p_cpucount[2] = &(ugeth->p_scheduler->cpucount2); ugeth->p_cpucount[3] = &(ugeth->p_scheduler->cpucount3); ugeth->p_cpucount[4] = &(ugeth->p_scheduler->cpucount4); ugeth->p_cpucount[5] = &(ugeth->p_scheduler->cpucount5); ugeth->p_cpucount[6] = &(ugeth->p_scheduler->cpucount6); ugeth->p_cpucount[7] = &(ugeth->p_scheduler->cpucount7); } /* schedulerbasepointer */ /* TxRMON_PTR (statistics) */ if (ug_info-> statisticsMode & UCC_GETH_STATISTICS_GATHERING_MODE_FIRMWARE_TX) { ugeth->tx_fw_statistics_pram_offset = qe_muram_alloc(sizeof (struct ucc_geth_tx_firmware_statistics_pram), UCC_GETH_TX_STATISTICS_ALIGNMENT); if (IS_ERR_VALUE(ugeth->tx_fw_statistics_pram_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_tx_fw_statistics_pram\n"); return -ENOMEM; } ugeth->p_tx_fw_statistics_pram = (struct ucc_geth_tx_firmware_statistics_pram __iomem *) qe_muram_addr(ugeth->tx_fw_statistics_pram_offset); } /* temoder */ /* Already has speed set */ if (ucc_geth_tx_queues(ug_info) > 1) temoder |= TEMODER_SCHEDULER_ENABLE; if (ug_info->ipCheckSumGenerate) temoder |= TEMODER_IP_CHECKSUM_GENERATE; temoder |= ((ucc_geth_tx_queues(ug_info) - 1) << TEMODER_NUM_OF_QUEUES_SHIFT); out_be16(&ugeth->p_tx_glbl_pram->temoder, temoder); /* Function code register value to be used later */ function_code = UCC_BMR_BO_BE | UCC_BMR_GBL; /* Required for QE */ /* function code register */ out_be32(&ugeth->p_tx_glbl_pram->tstate, ((u32) function_code) << 24); /* Rx global PRAM */ /* Allocate global rx parameter RAM page */ rx_glbl_pram_offset = qe_muram_alloc(sizeof(struct ucc_geth_rx_global_pram), UCC_GETH_RX_GLOBAL_PRAM_ALIGNMENT); if (rx_glbl_pram_offset < 0) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_rx_glbl_pram\n"); return -ENOMEM; } ugeth->p_rx_glbl_pram = qe_muram_addr(rx_glbl_pram_offset); /* Fill global PRAM */ /* RQPTR */ /* Size varies with number of Rx threads */ ugeth->thread_dat_rx_offset = qe_muram_alloc(numThreadsRxNumerical * sizeof(struct ucc_geth_thread_data_rx), UCC_GETH_THREAD_DATA_ALIGNMENT); if (IS_ERR_VALUE(ugeth->thread_dat_rx_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_thread_data_rx\n"); return -ENOMEM; } ugeth->p_thread_data_rx = (struct ucc_geth_thread_data_rx __iomem *) qe_muram_addr(ugeth-> thread_dat_rx_offset); out_be32(&ugeth->p_rx_glbl_pram->rqptr, ugeth->thread_dat_rx_offset); /* typeorlen */ out_be16(&ugeth->p_rx_glbl_pram->typeorlen, ug_info->typeorlen); /* rxrmonbaseptr (statistics) */ if (ug_info-> statisticsMode & UCC_GETH_STATISTICS_GATHERING_MODE_FIRMWARE_RX) { ugeth->rx_fw_statistics_pram_offset = qe_muram_alloc(sizeof (struct ucc_geth_rx_firmware_statistics_pram), UCC_GETH_RX_STATISTICS_ALIGNMENT); if (IS_ERR_VALUE(ugeth->rx_fw_statistics_pram_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_rx_fw_statistics_pram\n"); return -ENOMEM; } ugeth->p_rx_fw_statistics_pram = (struct ucc_geth_rx_firmware_statistics_pram __iomem *) qe_muram_addr(ugeth->rx_fw_statistics_pram_offset); } /* intCoalescingPtr */ /* Size varies with number of Rx queues */ ugeth->rx_irq_coalescing_tbl_offset = qe_muram_alloc(ucc_geth_rx_queues(ug_info) * sizeof(struct ucc_geth_rx_interrupt_coalescing_entry) + 4, UCC_GETH_RX_INTERRUPT_COALESCING_ALIGNMENT); if (IS_ERR_VALUE(ugeth->rx_irq_coalescing_tbl_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_rx_irq_coalescing_tbl\n"); return -ENOMEM; } ugeth->p_rx_irq_coalescing_tbl = (struct ucc_geth_rx_interrupt_coalescing_table __iomem *) qe_muram_addr(ugeth->rx_irq_coalescing_tbl_offset); out_be32(&ugeth->p_rx_glbl_pram->intcoalescingptr, ugeth->rx_irq_coalescing_tbl_offset); /* Fill interrupt coalescing table */ for (i = 0; i < ucc_geth_rx_queues(ug_info); i++) { out_be32(&ugeth->p_rx_irq_coalescing_tbl->coalescingentry[i]. interruptcoalescingmaxvalue, ug_info->interruptcoalescingmaxvalue[i]); out_be32(&ugeth->p_rx_irq_coalescing_tbl->coalescingentry[i]. interruptcoalescingcounter, ug_info->interruptcoalescingmaxvalue[i]); } /* MRBLR */ init_max_rx_buff_len(uf_info->max_rx_buf_length, &ugeth->p_rx_glbl_pram->mrblr); /* MFLR */ out_be16(&ugeth->p_rx_glbl_pram->mflr, ug_info->maxFrameLength); /* MINFLR */ init_min_frame_len(ug_info->minFrameLength, &ugeth->p_rx_glbl_pram->minflr, &ugeth->p_rx_glbl_pram->mrblr); /* MAXD1 */ out_be16(&ugeth->p_rx_glbl_pram->maxd1, ug_info->maxD1Length); /* MAXD2 */ out_be16(&ugeth->p_rx_glbl_pram->maxd2, ug_info->maxD2Length); /* l2qt */ l2qt = 0; for (i = 0; i < UCC_GETH_VLAN_PRIORITY_MAX; i++) l2qt |= (ug_info->l2qt[i] << (28 - 4 * i)); out_be32(&ugeth->p_rx_glbl_pram->l2qt, l2qt); /* l3qt */ for (j = 0; j < UCC_GETH_IP_PRIORITY_MAX; j += 8) { l3qt = 0; for (i = 0; i < 8; i++) l3qt |= (ug_info->l3qt[j + i] << (28 - 4 * i)); out_be32(&ugeth->p_rx_glbl_pram->l3qt[j/8], l3qt); } /* vlantype */ out_be16(&ugeth->p_rx_glbl_pram->vlantype, ug_info->vlantype); /* vlantci */ out_be16(&ugeth->p_rx_glbl_pram->vlantci, ug_info->vlantci); /* ecamptr */ out_be32(&ugeth->p_rx_glbl_pram->ecamptr, ug_info->ecamptr); /* RBDQPTR */ /* Size varies with number of Rx queues */ ugeth->rx_bd_qs_tbl_offset = qe_muram_alloc(ucc_geth_rx_queues(ug_info) * (sizeof(struct ucc_geth_rx_bd_queues_entry) + sizeof(struct ucc_geth_rx_prefetched_bds)), UCC_GETH_RX_BD_QUEUES_ALIGNMENT); if (IS_ERR_VALUE(ugeth->rx_bd_qs_tbl_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_rx_bd_qs_tbl\n"); return -ENOMEM; } ugeth->p_rx_bd_qs_tbl = (struct ucc_geth_rx_bd_queues_entry __iomem *) qe_muram_addr(ugeth-> rx_bd_qs_tbl_offset); out_be32(&ugeth->p_rx_glbl_pram->rbdqptr, ugeth->rx_bd_qs_tbl_offset); /* Setup the table */ /* Assume BD rings are already established */ for (i = 0; i < ucc_geth_rx_queues(ug_info); i++) { out_be32(&ugeth->p_rx_bd_qs_tbl[i].externalbdbaseptr, (u32) virt_to_phys(ugeth->p_rx_bd_ring[i])); /* rest of fields handled by QE */ } /* remoder */ /* Already has speed set */ if (ugeth->rx_extended_features) remoder |= REMODER_RX_EXTENDED_FEATURES; if (ug_info->rxExtendedFiltering) remoder |= REMODER_RX_EXTENDED_FILTERING; if (ug_info->dynamicMaxFrameLength) remoder |= REMODER_DYNAMIC_MAX_FRAME_LENGTH; if (ug_info->dynamicMinFrameLength) remoder |= REMODER_DYNAMIC_MIN_FRAME_LENGTH; remoder |= ug_info->vlanOperationTagged << REMODER_VLAN_OPERATION_TAGGED_SHIFT; remoder |= ug_info-> vlanOperationNonTagged << REMODER_VLAN_OPERATION_NON_TAGGED_SHIFT; remoder |= ug_info->rxQoSMode << REMODER_RX_QOS_MODE_SHIFT; remoder |= ((ucc_geth_rx_queues(ug_info) - 1) << REMODER_NUM_OF_QUEUES_SHIFT); if (ug_info->ipCheckSumCheck) remoder |= REMODER_IP_CHECKSUM_CHECK; if (ug_info->ipAddressAlignment) remoder |= REMODER_IP_ADDRESS_ALIGNMENT; out_be32(&ugeth->p_rx_glbl_pram->remoder, remoder); /* Note that this function must be called */ /* ONLY AFTER p_tx_fw_statistics_pram */ /* andp_UccGethRxFirmwareStatisticsPram are allocated ! */ init_firmware_statistics_gathering_mode((ug_info-> statisticsMode & UCC_GETH_STATISTICS_GATHERING_MODE_FIRMWARE_TX), (ug_info->statisticsMode & UCC_GETH_STATISTICS_GATHERING_MODE_FIRMWARE_RX), &ugeth->p_tx_glbl_pram->txrmonbaseptr, ugeth->tx_fw_statistics_pram_offset, &ugeth->p_rx_glbl_pram->rxrmonbaseptr, ugeth->rx_fw_statistics_pram_offset, &ugeth->p_tx_glbl_pram->temoder, &ugeth->p_rx_glbl_pram->remoder); /* function code register */ out_8(&ugeth->p_rx_glbl_pram->rstate, function_code); /* initialize extended filtering */ if (ug_info->rxExtendedFiltering) { if (!ug_info->extendedFilteringChainPointer) { if (netif_msg_ifup(ugeth)) pr_err("Null Extended Filtering Chain Pointer\n"); return -EINVAL; } /* Allocate memory for extended filtering Mode Global Parameters */ ugeth->exf_glbl_param_offset = qe_muram_alloc(sizeof(struct ucc_geth_exf_global_pram), UCC_GETH_RX_EXTENDED_FILTERING_GLOBAL_PARAMETERS_ALIGNMENT); if (IS_ERR_VALUE(ugeth->exf_glbl_param_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_exf_glbl_param\n"); return -ENOMEM; } ugeth->p_exf_glbl_param = (struct ucc_geth_exf_global_pram __iomem *) qe_muram_addr(ugeth-> exf_glbl_param_offset); out_be32(&ugeth->p_rx_glbl_pram->exfGlobalParam, ugeth->exf_glbl_param_offset); out_be32(&ugeth->p_exf_glbl_param->l2pcdptr, (u32) ug_info->extendedFilteringChainPointer); } else { /* initialize 82xx style address filtering */ /* Init individual address recognition registers to disabled */ for (j = 0; j < NUM_OF_PADDRS; j++) ugeth_82xx_filtering_clear_addr_in_paddr(ugeth, (u8) j); p_82xx_addr_filt = (struct ucc_geth_82xx_address_filtering_pram __iomem *) ugeth-> p_rx_glbl_pram->addressfiltering; ugeth_82xx_filtering_clear_all_addr_in_hash(ugeth, ENET_ADDR_TYPE_GROUP); ugeth_82xx_filtering_clear_all_addr_in_hash(ugeth, ENET_ADDR_TYPE_INDIVIDUAL); } /* * Initialize UCC at QE level */ command = QE_INIT_TX_RX; /* Allocate shadow InitEnet command parameter structure. * This is needed because after the InitEnet command is executed, * the structure in DPRAM is released, because DPRAM is a premium * resource. * This shadow structure keeps a copy of what was done so that the * allocated resources can be released when the channel is freed. */ if (!(ugeth->p_init_enet_param_shadow = kzalloc(sizeof(struct ucc_geth_init_pram), GFP_KERNEL))) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate memory for p_UccInitEnetParamShadows\n"); return -ENOMEM; } /* Fill shadow InitEnet command parameter structure */ ugeth->p_init_enet_param_shadow->resinit1 = ENET_INIT_PARAM_MAGIC_RES_INIT1; ugeth->p_init_enet_param_shadow->resinit2 = ENET_INIT_PARAM_MAGIC_RES_INIT2; ugeth->p_init_enet_param_shadow->resinit3 = ENET_INIT_PARAM_MAGIC_RES_INIT3; ugeth->p_init_enet_param_shadow->resinit4 = ENET_INIT_PARAM_MAGIC_RES_INIT4; ugeth->p_init_enet_param_shadow->resinit5 = ENET_INIT_PARAM_MAGIC_RES_INIT5; ugeth->p_init_enet_param_shadow->rgftgfrxglobal |= ((u32) ug_info->numThreadsRx) << ENET_INIT_PARAM_RGF_SHIFT; ugeth->p_init_enet_param_shadow->rgftgfrxglobal |= ((u32) ug_info->numThreadsTx) << ENET_INIT_PARAM_TGF_SHIFT; ugeth->p_init_enet_param_shadow->rgftgfrxglobal |= rx_glbl_pram_offset | ug_info->riscRx; if ((ug_info->largestexternallookupkeysize != QE_FLTR_LARGEST_EXTERNAL_TABLE_LOOKUP_KEY_SIZE_NONE) && (ug_info->largestexternallookupkeysize != QE_FLTR_LARGEST_EXTERNAL_TABLE_LOOKUP_KEY_SIZE_8_BYTES) && (ug_info->largestexternallookupkeysize != QE_FLTR_LARGEST_EXTERNAL_TABLE_LOOKUP_KEY_SIZE_16_BYTES)) { if (netif_msg_ifup(ugeth)) pr_err("Invalid largest External Lookup Key Size\n"); return -EINVAL; } ugeth->p_init_enet_param_shadow->largestexternallookupkeysize = ug_info->largestexternallookupkeysize; size = sizeof(struct ucc_geth_thread_rx_pram); if (ug_info->rxExtendedFiltering) { size += THREAD_RX_PRAM_ADDITIONAL_FOR_EXTENDED_FILTERING; if (ug_info->largestexternallookupkeysize == QE_FLTR_LARGEST_EXTERNAL_TABLE_LOOKUP_KEY_SIZE_8_BYTES) size += THREAD_RX_PRAM_ADDITIONAL_FOR_EXTENDED_FILTERING_8; if (ug_info->largestexternallookupkeysize == QE_FLTR_LARGEST_EXTERNAL_TABLE_LOOKUP_KEY_SIZE_16_BYTES) size += THREAD_RX_PRAM_ADDITIONAL_FOR_EXTENDED_FILTERING_16; } if ((ret_val = fill_init_enet_entries(ugeth, &(ugeth-> p_init_enet_param_shadow->rxthread[0]), (u8) (numThreadsRxNumerical + 1) /* Rx needs one extra for terminator */ , size, UCC_GETH_THREAD_RX_PRAM_ALIGNMENT, ug_info->riscRx, 1)) != 0) { if (netif_msg_ifup(ugeth)) pr_err("Can not fill p_init_enet_param_shadow\n"); return ret_val; } ugeth->p_init_enet_param_shadow->txglobal = tx_glbl_pram_offset | ug_info->riscTx; if ((ret_val = fill_init_enet_entries(ugeth, &(ugeth->p_init_enet_param_shadow-> txthread[0]), numThreadsTxNumerical, sizeof(struct ucc_geth_thread_tx_pram), UCC_GETH_THREAD_TX_PRAM_ALIGNMENT, ug_info->riscTx, 0)) != 0) { if (netif_msg_ifup(ugeth)) pr_err("Can not fill p_init_enet_param_shadow\n"); return ret_val; } /* Load Rx bds with buffers */ for (i = 0; i < ucc_geth_rx_queues(ug_info); i++) { if ((ret_val = rx_bd_buffer_set(ugeth, (u8) i)) != 0) { if (netif_msg_ifup(ugeth)) pr_err("Can not fill Rx bds with buffers\n"); return ret_val; } } /* Allocate InitEnet command parameter structure */ init_enet_pram_offset = qe_muram_alloc(sizeof(struct ucc_geth_init_pram), 4); if (IS_ERR_VALUE(init_enet_pram_offset)) { if (netif_msg_ifup(ugeth)) pr_err("Can not allocate DPRAM memory for p_init_enet_pram\n"); return -ENOMEM; } p_init_enet_pram = (struct ucc_geth_init_pram __iomem *) qe_muram_addr(init_enet_pram_offset); /* Copy shadow InitEnet command parameter structure into PRAM */ out_8(&p_init_enet_pram->resinit1, ugeth->p_init_enet_param_shadow->resinit1); out_8(&p_init_enet_pram->resinit2, ugeth->p_init_enet_param_shadow->resinit2); out_8(&p_init_enet_pram->resinit3, ugeth->p_init_enet_param_shadow->resinit3); out_8(&p_init_enet_pram->resinit4, ugeth->p_init_enet_param_shadow->resinit4); out_be16(&p_init_enet_pram->resinit5, ugeth->p_init_enet_param_shadow->resinit5); out_8(&p_init_enet_pram->largestexternallookupkeysize, ugeth->p_init_enet_param_shadow->largestexternallookupkeysize); out_be32(&p_init_enet_pram->rgftgfrxglobal, ugeth->p_init_enet_param_shadow->rgftgfrxglobal); for (i = 0; i < ENET_INIT_PARAM_MAX_ENTRIES_RX; i++) out_be32(&p_init_enet_pram->rxthread[i], ugeth->p_init_enet_param_shadow->rxthread[i]); out_be32(&p_init_enet_pram->txglobal, ugeth->p_init_enet_param_shadow->txglobal); for (i = 0; i < ENET_INIT_PARAM_MAX_ENTRIES_TX; i++) out_be32(&p_init_enet_pram->txthread[i], ugeth->p_init_enet_param_shadow->txthread[i]); /* Issue QE command */ cecr_subblock = ucc_fast_get_qe_cr_subblock(ugeth->ug_info->uf_info.ucc_num); qe_issue_cmd(command, cecr_subblock, QE_CR_PROTOCOL_ETHERNET, init_enet_pram_offset); /* Free InitEnet command parameter */ qe_muram_free(init_enet_pram_offset); return 0; } /* This is called by the kernel when a frame is ready for transmission. */ /* It is pointed to by the dev->hard_start_xmit function pointer */ static netdev_tx_t ucc_geth_start_xmit(struct sk_buff *skb, struct net_device *dev) { struct ucc_geth_private *ugeth = netdev_priv(dev); #ifdef CONFIG_UGETH_TX_ON_DEMAND struct ucc_fast_private *uccf; #endif u8 __iomem *bd; /* BD pointer */ u32 bd_status; u8 txQ = 0; unsigned long flags; ugeth_vdbg("%s: IN", __func__); netdev_sent_queue(dev, skb->len); spin_lock_irqsave(&ugeth->lock, flags); dev->stats.tx_bytes += skb->len; /* Start from the next BD that should be filled */ bd = ugeth->txBd[txQ]; bd_status = in_be32((u32 __iomem *)bd); /* Save the skb pointer so we can free it later */ ugeth->tx_skbuff[txQ][ugeth->skb_curtx[txQ]] = skb; /* Update the current skb pointer (wrapping if this was the last) */ ugeth->skb_curtx[txQ] = (ugeth->skb_curtx[txQ] + 1) & TX_RING_MOD_MASK(ugeth->ug_info->bdRingLenTx[txQ]); /* set up the buffer descriptor */ out_be32(&((struct qe_bd __iomem *)bd)->buf, dma_map_single(ugeth->dev, skb->data, skb->len, DMA_TO_DEVICE)); /* printk(KERN_DEBUG"skb->data is 0x%x\n",skb->data); */ bd_status = (bd_status & T_W) | T_R | T_I | T_L | skb->len; /* set bd status and length */ out_be32((u32 __iomem *)bd, bd_status); /* Move to next BD in the ring */ if (!(bd_status & T_W)) bd += sizeof(struct qe_bd); else bd = ugeth->p_tx_bd_ring[txQ]; /* If the next BD still needs to be cleaned up, then the bds are full. We need to tell the kernel to stop sending us stuff. */ if (bd == ugeth->confBd[txQ]) { if (!netif_queue_stopped(dev)) netif_stop_queue(dev); } ugeth->txBd[txQ] = bd; skb_tx_timestamp(skb); if (ugeth->p_scheduler) { ugeth->cpucount[txQ]++; /* Indicate to QE that there are more Tx bds ready for transmission */ /* This is done by writing a running counter of the bd count to the scheduler PRAM. */ out_be16(ugeth->p_cpucount[txQ], ugeth->cpucount[txQ]); } #ifdef CONFIG_UGETH_TX_ON_DEMAND uccf = ugeth->uccf; out_be16(uccf->p_utodr, UCC_FAST_TOD); #endif spin_unlock_irqrestore(&ugeth->lock, flags); return NETDEV_TX_OK; } static int ucc_geth_rx(struct ucc_geth_private *ugeth, u8 rxQ, int rx_work_limit) { struct sk_buff *skb; u8 __iomem *bd; u16 length, howmany = 0; u32 bd_status; u8 *bdBuffer; struct net_device *dev; ugeth_vdbg("%s: IN", __func__); dev = ugeth->ndev; /* collect received buffers */ bd = ugeth->rxBd[rxQ]; bd_status = in_be32((u32 __iomem *)bd); /* while there are received buffers and BD is full (~R_E) */ while (!((bd_status & (R_E)) || (--rx_work_limit < 0))) { bdBuffer = (u8 *) in_be32(&((struct qe_bd __iomem *)bd)->buf); length = (u16) ((bd_status & BD_LENGTH_MASK) - 4); skb = ugeth->rx_skbuff[rxQ][ugeth->skb_currx[rxQ]]; /* determine whether buffer is first, last, first and last (single buffer frame) or middle (not first and not last) */ if (!skb || (!(bd_status & (R_F | R_L))) || (bd_status & R_ERRORS_FATAL)) { if (netif_msg_rx_err(ugeth)) pr_err("%d: ERROR!!! skb - 0x%08x\n", __LINE__, (u32)skb); dev_kfree_skb(skb); ugeth->rx_skbuff[rxQ][ugeth->skb_currx[rxQ]] = NULL; dev->stats.rx_dropped++; } else { dev->stats.rx_packets++; howmany++; /* Prep the skb for the packet */ skb_put(skb, length); /* Tell the skb what kind of packet this is */ skb->protocol = eth_type_trans(skb, ugeth->ndev); dev->stats.rx_bytes += length; /* Send the packet up the stack */ netif_receive_skb(skb); } skb = get_new_skb(ugeth, bd); if (!skb) { if (netif_msg_rx_err(ugeth)) pr_warn("No Rx Data Buffer\n"); dev->stats.rx_dropped++; break; } ugeth->rx_skbuff[rxQ][ugeth->skb_currx[rxQ]] = skb; /* update to point at the next skb */ ugeth->skb_currx[rxQ] = (ugeth->skb_currx[rxQ] + 1) & RX_RING_MOD_MASK(ugeth->ug_info->bdRingLenRx[rxQ]); if (bd_status & R_W) bd = ugeth->p_rx_bd_ring[rxQ]; else bd += sizeof(struct qe_bd); bd_status = in_be32((u32 __iomem *)bd); } ugeth->rxBd[rxQ] = bd; return howmany; } static int ucc_geth_tx(struct net_device *dev, u8 txQ) { /* Start from the next BD that should be filled */ struct ucc_geth_private *ugeth = netdev_priv(dev); unsigned int bytes_sent = 0; int howmany = 0; u8 __iomem *bd; /* BD pointer */ u32 bd_status; bd = ugeth->confBd[txQ]; bd_status = in_be32((u32 __iomem *)bd); /* Normal processing. */ while ((bd_status & T_R) == 0) { struct sk_buff *skb; /* BD contains already transmitted buffer. */ /* Handle the transmitted buffer and release */ /* the BD to be used with the current frame */ skb = ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]]; if (!skb) break; howmany++; bytes_sent += skb->len; dev->stats.tx_packets++; dev_consume_skb_any(skb); ugeth->tx_skbuff[txQ][ugeth->skb_dirtytx[txQ]] = NULL; ugeth->skb_dirtytx[txQ] = (ugeth->skb_dirtytx[txQ] + 1) & TX_RING_MOD_MASK(ugeth->ug_info->bdRingLenTx[txQ]); /* We freed a buffer, so now we can restart transmission */ if (netif_queue_stopped(dev)) netif_wake_queue(dev); /* Advance the confirmation BD pointer */ if (!(bd_status & T_W)) bd += sizeof(struct qe_bd); else bd = ugeth->p_tx_bd_ring[txQ]; bd_status = in_be32((u32 __iomem *)bd); } ugeth->confBd[txQ] = bd; netdev_completed_queue(dev, howmany, bytes_sent); return 0; } static int ucc_geth_poll(struct napi_struct *napi, int budget) { struct ucc_geth_private *ugeth = container_of(napi, struct ucc_geth_private, napi); struct ucc_geth_info *ug_info; int howmany, i; ug_info = ugeth->ug_info; /* Tx event processing */ spin_lock(&ugeth->lock); for (i = 0; i < ucc_geth_tx_queues(ug_info); i++) ucc_geth_tx(ugeth->ndev, i); spin_unlock(&ugeth->lock); howmany = 0; for (i = 0; i < ucc_geth_rx_queues(ug_info); i++) howmany += ucc_geth_rx(ugeth, i, budget - howmany); if (howmany < budget) { napi_complete_done(napi, howmany); setbits32(ugeth->uccf->p_uccm, UCCE_RX_EVENTS | UCCE_TX_EVENTS); } return howmany; } static irqreturn_t ucc_geth_irq_handler(int irq, void *info) { struct net_device *dev = info; struct ucc_geth_private *ugeth = netdev_priv(dev); struct ucc_fast_private *uccf; struct ucc_geth_info *ug_info; register u32 ucce; register u32 uccm; ugeth_vdbg("%s: IN", __func__); uccf = ugeth->uccf; ug_info = ugeth->ug_info; /* read and clear events */ ucce = (u32) in_be32(uccf->p_ucce); uccm = (u32) in_be32(uccf->p_uccm); ucce &= uccm; out_be32(uccf->p_ucce, ucce); /* check for receive events that require processing */ if (ucce & (UCCE_RX_EVENTS | UCCE_TX_EVENTS)) { if (napi_schedule_prep(&ugeth->napi)) { uccm &= ~(UCCE_RX_EVENTS | UCCE_TX_EVENTS); out_be32(uccf->p_uccm, uccm); __napi_schedule(&ugeth->napi); } } /* Errors and other events */ if (ucce & UCCE_OTHER) { if (ucce & UCC_GETH_UCCE_BSY) dev->stats.rx_errors++; if (ucce & UCC_GETH_UCCE_TXE) dev->stats.tx_errors++; } return IRQ_HANDLED; } #ifdef CONFIG_NET_POLL_CONTROLLER /* * Polling 'interrupt' - used by things like netconsole to send skbs * without having to re-enable interrupts. It's not called while * the interrupt routine is executing. */ static void ucc_netpoll(struct net_device *dev) { struct ucc_geth_private *ugeth = netdev_priv(dev); int irq = ugeth->ug_info->uf_info.irq; disable_irq(irq); ucc_geth_irq_handler(irq, dev); enable_irq(irq); } #endif /* CONFIG_NET_POLL_CONTROLLER */ static int ucc_geth_set_mac_addr(struct net_device *dev, void *p) { struct ucc_geth_private *ugeth = netdev_priv(dev); struct sockaddr *addr = p; if (!is_valid_ether_addr(addr->sa_data)) return -EADDRNOTAVAIL; memcpy(dev->dev_addr, addr->sa_data, dev->addr_len); /* * If device is not running, we will set mac addr register * when opening the device. */ if (!netif_running(dev)) return 0; spin_lock_irq(&ugeth->lock); init_mac_station_addr_regs(dev->dev_addr[0], dev->dev_addr[1], dev->dev_addr[2], dev->dev_addr[3], dev->dev_addr[4], dev->dev_addr[5], &ugeth->ug_regs->macstnaddr1, &ugeth->ug_regs->macstnaddr2); spin_unlock_irq(&ugeth->lock); return 0; } static int ucc_geth_init_mac(struct ucc_geth_private *ugeth) { struct net_device *dev = ugeth->ndev; int err; err = ucc_struct_init(ugeth); if (err) { netif_err(ugeth, ifup, dev, "Cannot configure internal struct, aborting\n"); goto err; } err = ucc_geth_startup(ugeth); if (err) { netif_err(ugeth, ifup, dev, "Cannot configure net device, aborting\n"); goto err; } err = adjust_enet_interface(ugeth); if (err) { netif_err(ugeth, ifup, dev, "Cannot configure net device, aborting\n"); goto err; } /* Set MACSTNADDR1, MACSTNADDR2 */ /* For more details see the hardware spec. */ init_mac_station_addr_regs(dev->dev_addr[0], dev->dev_addr[1], dev->dev_addr[2], dev->dev_addr[3], dev->dev_addr[4], dev->dev_addr[5], &ugeth->ug_regs->macstnaddr1, &ugeth->ug_regs->macstnaddr2); err = ugeth_enable(ugeth, COMM_DIR_RX_AND_TX); if (err) { netif_err(ugeth, ifup, dev, "Cannot enable net device, aborting\n"); goto err; } return 0; err: ucc_geth_stop(ugeth); return err; } /* Called when something needs to use the ethernet device */ /* Returns 0 for success. */ static int ucc_geth_open(struct net_device *dev) { struct ucc_geth_private *ugeth = netdev_priv(dev); int err; ugeth_vdbg("%s: IN", __func__); /* Test station address */ if (dev->dev_addr[0] & ENET_GROUP_ADDR) { netif_err(ugeth, ifup, dev, "Multicast address used for station address - is this what you wanted?\n"); return -EINVAL; } err = init_phy(dev); if (err) { netif_err(ugeth, ifup, dev, "Cannot initialize PHY, aborting\n"); return err; } err = ucc_geth_init_mac(ugeth); if (err) { netif_err(ugeth, ifup, dev, "Cannot initialize MAC, aborting\n"); goto err; } err = request_irq(ugeth->ug_info->uf_info.irq, ucc_geth_irq_handler, 0, "UCC Geth", dev); if (err) { netif_err(ugeth, ifup, dev, "Cannot get IRQ for net device, aborting\n"); goto err; } phy_start(ugeth->phydev); napi_enable(&ugeth->napi); netdev_reset_queue(dev); netif_start_queue(dev); device_set_wakeup_capable(&dev->dev, qe_alive_during_sleep() || ugeth->phydev->irq); device_set_wakeup_enable(&dev->dev, ugeth->wol_en); return err; err: ucc_geth_stop(ugeth); return err; } /* Stops the kernel queue, and halts the controller */ static int ucc_geth_close(struct net_device *dev) { struct ucc_geth_private *ugeth = netdev_priv(dev); ugeth_vdbg("%s: IN", __func__); napi_disable(&ugeth->napi); cancel_work_sync(&ugeth->timeout_work); ucc_geth_stop(ugeth); phy_disconnect(ugeth->phydev); ugeth->phydev = NULL; free_irq(ugeth->ug_info->uf_info.irq, ugeth->ndev); netif_stop_queue(dev); netdev_reset_queue(dev); return 0; } /* Reopen device. This will reset the MAC and PHY. */ static void ucc_geth_timeout_work(struct work_struct *work) { struct ucc_geth_private *ugeth; struct net_device *dev; ugeth = container_of(work, struct ucc_geth_private, timeout_work); dev = ugeth->ndev; ugeth_vdbg("%s: IN", __func__); dev->stats.tx_errors++; ugeth_dump_regs(ugeth); if (dev->flags & IFF_UP) { /* * Must reset MAC *and* PHY. This is done by reopening * the device. */ netif_tx_stop_all_queues(dev); ucc_geth_stop(ugeth); ucc_geth_init_mac(ugeth); /* Must start PHY here */ phy_start(ugeth->phydev); netif_tx_start_all_queues(dev); } netif_tx_schedule_all(dev); } /* * ucc_geth_timeout gets called when a packet has not been * transmitted after a set amount of time. */ static void ucc_geth_timeout(struct net_device *dev, unsigned int txqueue) { struct ucc_geth_private *ugeth = netdev_priv(dev); schedule_work(&ugeth->timeout_work); } #ifdef CONFIG_PM static int ucc_geth_suspend(struct platform_device *ofdev, pm_message_t state) { struct net_device *ndev = platform_get_drvdata(ofdev); struct ucc_geth_private *ugeth = netdev_priv(ndev); if (!netif_running(ndev)) return 0; netif_device_detach(ndev); napi_disable(&ugeth->napi); /* * Disable the controller, otherwise we'll wakeup on any network * activity. */ ugeth_disable(ugeth, COMM_DIR_RX_AND_TX); if (ugeth->wol_en & WAKE_MAGIC) { setbits32(ugeth->uccf->p_uccm, UCC_GETH_UCCE_MPD); setbits32(&ugeth->ug_regs->maccfg2, MACCFG2_MPE); ucc_fast_enable(ugeth->uccf, COMM_DIR_RX_AND_TX); } else if (!(ugeth->wol_en & WAKE_PHY)) { phy_stop(ugeth->phydev); } return 0; } static int ucc_geth_resume(struct platform_device *ofdev) { struct net_device *ndev = platform_get_drvdata(ofdev); struct ucc_geth_private *ugeth = netdev_priv(ndev); int err; if (!netif_running(ndev)) return 0; if (qe_alive_during_sleep()) { if (ugeth->wol_en & WAKE_MAGIC) { ucc_fast_disable(ugeth->uccf, COMM_DIR_RX_AND_TX); clrbits32(&ugeth->ug_regs->maccfg2, MACCFG2_MPE); clrbits32(ugeth->uccf->p_uccm, UCC_GETH_UCCE_MPD); } ugeth_enable(ugeth, COMM_DIR_RX_AND_TX); } else { /* * Full reinitialization is required if QE shuts down * during sleep. */ ucc_geth_memclean(ugeth); err = ucc_geth_init_mac(ugeth); if (err) { netdev_err(ndev, "Cannot initialize MAC, aborting\n"); return err; } } ugeth->oldlink = 0; ugeth->oldspeed = 0; ugeth->oldduplex = -1; phy_stop(ugeth->phydev); phy_start(ugeth->phydev); napi_enable(&ugeth->napi); netif_device_attach(ndev); return 0; } #else #define ucc_geth_suspend NULL #define ucc_geth_resume NULL #endif static phy_interface_t to_phy_interface(const char *phy_connection_type) { if (strcasecmp(phy_connection_type, "mii") == 0) return PHY_INTERFACE_MODE_MII; if (strcasecmp(phy_connection_type, "gmii") == 0) return PHY_INTERFACE_MODE_GMII; if (strcasecmp(phy_connection_type, "tbi") == 0) return PHY_INTERFACE_MODE_TBI; if (strcasecmp(phy_connection_type, "rmii") == 0) return PHY_INTERFACE_MODE_RMII; if (strcasecmp(phy_connection_type, "rgmii") == 0) return PHY_INTERFACE_MODE_RGMII; if (strcasecmp(phy_connection_type, "rgmii-id") == 0) return PHY_INTERFACE_MODE_RGMII_ID; if (strcasecmp(phy_connection_type, "rgmii-txid") == 0) return PHY_INTERFACE_MODE_RGMII_TXID; if (strcasecmp(phy_connection_type, "rgmii-rxid") == 0) return PHY_INTERFACE_MODE_RGMII_RXID; if (strcasecmp(phy_connection_type, "rtbi") == 0) return PHY_INTERFACE_MODE_RTBI; if (strcasecmp(phy_connection_type, "sgmii") == 0) return PHY_INTERFACE_MODE_SGMII; return PHY_INTERFACE_MODE_MII; } static int ucc_geth_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) { struct ucc_geth_private *ugeth = netdev_priv(dev); if (!netif_running(dev)) return -EINVAL; if (!ugeth->phydev) return -ENODEV; return phy_mii_ioctl(ugeth->phydev, rq, cmd); } static const struct net_device_ops ucc_geth_netdev_ops = { .ndo_open = ucc_geth_open, .ndo_stop = ucc_geth_close, .ndo_start_xmit = ucc_geth_start_xmit, .ndo_validate_addr = eth_validate_addr, .ndo_change_carrier = fixed_phy_change_carrier, .ndo_set_mac_address = ucc_geth_set_mac_addr, .ndo_set_rx_mode = ucc_geth_set_multi, .ndo_tx_timeout = ucc_geth_timeout, .ndo_do_ioctl = ucc_geth_ioctl, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = ucc_netpoll, #endif }; static int ucc_geth_parse_clock(struct device_node *np, const char *which, enum qe_clock *out) { const char *sprop; char buf[24]; snprintf(buf, sizeof(buf), "%s-clock-name", which); sprop = of_get_property(np, buf, NULL); if (sprop) { *out = qe_clock_source(sprop); } else { u32 val; snprintf(buf, sizeof(buf), "%s-clock", which); if (of_property_read_u32(np, buf, &val)) { /* If both *-clock-name and *-clock are missing, * we want to tell people to use *-clock-name. */ pr_err("missing %s-clock-name property\n", buf); return -EINVAL; } *out = val; } if (*out < QE_CLK_NONE || *out > QE_CLK24) { pr_err("invalid %s property\n", buf); return -EINVAL; } return 0; } static int ucc_geth_probe(struct platform_device* ofdev) { struct device *device = &ofdev->dev; struct device_node *np = ofdev->dev.of_node; struct net_device *dev = NULL; struct ucc_geth_private *ugeth = NULL; struct ucc_geth_info *ug_info; struct resource res; int err, ucc_num, max_speed = 0; const unsigned int *prop; phy_interface_t phy_interface; static const int enet_to_speed[] = { SPEED_10, SPEED_10, SPEED_10, SPEED_100, SPEED_100, SPEED_100, SPEED_1000, SPEED_1000, SPEED_1000, SPEED_1000, }; static const phy_interface_t enet_to_phy_interface[] = { PHY_INTERFACE_MODE_MII, PHY_INTERFACE_MODE_RMII, PHY_INTERFACE_MODE_RGMII, PHY_INTERFACE_MODE_MII, PHY_INTERFACE_MODE_RMII, PHY_INTERFACE_MODE_RGMII, PHY_INTERFACE_MODE_GMII, PHY_INTERFACE_MODE_RGMII, PHY_INTERFACE_MODE_TBI, PHY_INTERFACE_MODE_RTBI, PHY_INTERFACE_MODE_SGMII, }; ugeth_vdbg("%s: IN", __func__); prop = of_get_property(np, "cell-index", NULL); if (!prop) { prop = of_get_property(np, "device-id", NULL); if (!prop) return -ENODEV; } ucc_num = *prop - 1; if ((ucc_num < 0) || (ucc_num > 7)) return -ENODEV; ug_info = kmalloc(sizeof(*ug_info), GFP_KERNEL); if (ug_info == NULL) return -ENOMEM; memcpy(ug_info, &ugeth_primary_info, sizeof(*ug_info)); ug_info->uf_info.ucc_num = ucc_num; err = ucc_geth_parse_clock(np, "rx", &ug_info->uf_info.rx_clock); if (err) goto err_free_info; err = ucc_geth_parse_clock(np, "tx", &ug_info->uf_info.tx_clock); if (err) goto err_free_info; err = of_address_to_resource(np, 0, &res); if (err) goto err_free_info; ug_info->uf_info.regs = res.start; ug_info->uf_info.irq = irq_of_parse_and_map(np, 0); ug_info->phy_node = of_parse_phandle(np, "phy-handle", 0); if (!ug_info->phy_node && of_phy_is_fixed_link(np)) { /* * In the case of a fixed PHY, the DT node associated * to the PHY is the Ethernet MAC DT node. */ err = of_phy_register_fixed_link(np); if (err) goto err_free_info; ug_info->phy_node = of_node_get(np); } /* Find the TBI PHY node. If it's not there, we don't support SGMII */ ug_info->tbi_node = of_parse_phandle(np, "tbi-handle", 0); /* get the phy interface type, or default to MII */ prop = of_get_property(np, "phy-connection-type", NULL); if (!prop) { /* handle interface property present in old trees */ prop = of_get_property(ug_info->phy_node, "interface", NULL); if (prop != NULL) { phy_interface = enet_to_phy_interface[*prop]; max_speed = enet_to_speed[*prop]; } else phy_interface = PHY_INTERFACE_MODE_MII; } else { phy_interface = to_phy_interface((const char *)prop); } /* get speed, or derive from PHY interface */ if (max_speed == 0) switch (phy_interface) { case PHY_INTERFACE_MODE_GMII: case PHY_INTERFACE_MODE_RGMII: case PHY_INTERFACE_MODE_RGMII_ID: case PHY_INTERFACE_MODE_RGMII_RXID: case PHY_INTERFACE_MODE_RGMII_TXID: case PHY_INTERFACE_MODE_TBI: case PHY_INTERFACE_MODE_RTBI: case PHY_INTERFACE_MODE_SGMII: max_speed = SPEED_1000; break; default: max_speed = SPEED_100; break; } if (max_speed == SPEED_1000) { unsigned int snums = qe_get_num_of_snums(); /* configure muram FIFOs for gigabit operation */ ug_info->uf_info.urfs = UCC_GETH_URFS_GIGA_INIT; ug_info->uf_info.urfet = UCC_GETH_URFET_GIGA_INIT; ug_info->uf_info.urfset = UCC_GETH_URFSET_GIGA_INIT; ug_info->uf_info.utfs = UCC_GETH_UTFS_GIGA_INIT; ug_info->uf_info.utfet = UCC_GETH_UTFET_GIGA_INIT; ug_info->uf_info.utftt = UCC_GETH_UTFTT_GIGA_INIT; ug_info->numThreadsTx = UCC_GETH_NUM_OF_THREADS_4; /* If QE's snum number is 46/76 which means we need to support * 4 UECs at 1000Base-T simultaneously, we need to allocate * more Threads to Rx. */ if ((snums == 76) || (snums == 46)) ug_info->numThreadsRx = UCC_GETH_NUM_OF_THREADS_6; else ug_info->numThreadsRx = UCC_GETH_NUM_OF_THREADS_4; } if (netif_msg_probe(&debug)) pr_info("UCC%1d at 0x%8llx (irq = %d)\n", ug_info->uf_info.ucc_num + 1, (u64)ug_info->uf_info.regs, ug_info->uf_info.irq); /* Create an ethernet device instance */ dev = alloc_etherdev(sizeof(*ugeth)); if (dev == NULL) { err = -ENOMEM; goto err_deregister_fixed_link; } ugeth = netdev_priv(dev); spin_lock_init(&ugeth->lock); /* Create CQs for hash tables */ INIT_LIST_HEAD(&ugeth->group_hash_q); INIT_LIST_HEAD(&ugeth->ind_hash_q); dev_set_drvdata(device, dev); /* Set the dev->base_addr to the gfar reg region */ dev->base_addr = (unsigned long)(ug_info->uf_info.regs); SET_NETDEV_DEV(dev, device); /* Fill in the dev structure */ uec_set_ethtool_ops(dev); dev->netdev_ops = &ucc_geth_netdev_ops; dev->watchdog_timeo = TX_TIMEOUT; INIT_WORK(&ugeth->timeout_work, ucc_geth_timeout_work); netif_napi_add(dev, &ugeth->napi, ucc_geth_poll, 64); dev->mtu = 1500; dev->max_mtu = 1518; ugeth->msg_enable = netif_msg_init(debug.msg_enable, UGETH_MSG_DEFAULT); ugeth->phy_interface = phy_interface; ugeth->max_speed = max_speed; /* Carrier starts down, phylib will bring it up */ netif_carrier_off(dev); err = register_netdev(dev); if (err) { if (netif_msg_probe(ugeth)) pr_err("%s: Cannot register net device, aborting\n", dev->name); goto err_free_netdev; } of_get_mac_address(np, dev->dev_addr); ugeth->ug_info = ug_info; ugeth->dev = device; ugeth->ndev = dev; ugeth->node = np; return 0; err_free_netdev: free_netdev(dev); err_deregister_fixed_link: if (of_phy_is_fixed_link(np)) of_phy_deregister_fixed_link(np); of_node_put(ug_info->tbi_node); of_node_put(ug_info->phy_node); err_free_info: kfree(ug_info); return err; } static int ucc_geth_remove(struct platform_device* ofdev) { struct net_device *dev = platform_get_drvdata(ofdev); struct ucc_geth_private *ugeth = netdev_priv(dev); struct device_node *np = ofdev->dev.of_node; unregister_netdev(dev); ucc_geth_memclean(ugeth); if (of_phy_is_fixed_link(np)) of_phy_deregister_fixed_link(np); of_node_put(ugeth->ug_info->tbi_node); of_node_put(ugeth->ug_info->phy_node); kfree(ugeth->ug_info); free_netdev(dev); return 0; } static const struct of_device_id ucc_geth_match[] = { { .type = "network", .compatible = "ucc_geth", }, {}, }; MODULE_DEVICE_TABLE(of, ucc_geth_match); static struct platform_driver ucc_geth_driver = { .driver = { .name = DRV_NAME, .of_match_table = ucc_geth_match, }, .probe = ucc_geth_probe, .remove = ucc_geth_remove, .suspend = ucc_geth_suspend, .resume = ucc_geth_resume, }; static int __init ucc_geth_init(void) { if (netif_msg_drv(&debug)) pr_info(DRV_DESC "\n"); return platform_driver_register(&ucc_geth_driver); } static void __exit ucc_geth_exit(void) { platform_driver_unregister(&ucc_geth_driver); } module_init(ucc_geth_init); module_exit(ucc_geth_exit); MODULE_AUTHOR("Freescale Semiconductor, Inc"); MODULE_DESCRIPTION(DRV_DESC); MODULE_LICENSE("GPL");