patch-2.2.18 linux/drivers/net/macmace.c
Next file: linux/drivers/net/macsonic.c
Previous file: linux/drivers/net/mac89x0.c
Back to the patch index
Back to the overall index
- Lines: 794
- Date:
Fri Oct 13 23:40:43 2000
- Orig file:
v2.2.17/drivers/net/macmace.c
- Orig date:
Thu Jan 1 01:00:00 1970
diff -u --new-file --recursive --exclude-from /usr/src/exclude v2.2.17/drivers/net/macmace.c linux/drivers/net/macmace.c
@@ -0,0 +1,793 @@
+/*
+ * Driver for the Macintosh 68K onboard MACE controller with PSC
+ * driven DMA. The MACE driver code is derived from mace.c. The
+ * Mac68k theory of operation is courtesy of the MacBSD wizards.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ *
+ * Copyright (C) 1996 Paul Mackerras.
+ * Copyright (C) 1998 Alan Cox <alan@redhat.com>
+ */
+
+
+#include <linux/kernel.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/delay.h>
+#include <linux/string.h>
+#include <linux/timer.h>
+#include <asm/io.h>
+#include <asm/pgtable.h>
+#include <asm/irq.h>
+#include <asm/macintosh.h>
+#include <asm/macints.h>
+#include <asm/mac_psc.h>
+#include "mace.h"
+
+#define N_RX_RING 8
+#define N_TX_RING 2
+#define MAX_TX_ACTIVE 1
+#define NCMDS_TX 1 /* dma commands per element in tx ring */
+#define RX_BUFLEN (ETH_FRAME_LEN + 8)
+#define TX_TIMEOUT HZ /* 1 second */
+
+/* Bits in transmit DMA status */
+#define TX_DMA_ERR 0x80
+
+/* The MACE is simply wired down on a Mac68K box */
+
+#define MACE_BASE (void *)(0x50F1C000)
+#define MACE_PROM (void *)(0x50F08001)
+
+struct mace68k_data
+{
+ volatile struct mace *mace;
+ volatile unsigned char *tx_ring;
+ volatile unsigned char *rx_ring;
+ int dma_intr;
+ unsigned char maccc;
+ struct net_device_stats stats;
+ struct timer_list tx_timeout;
+ int timeout_active;
+ int rx_slot, rx_done;
+ int tx_slot, tx_count;
+};
+
+struct mace_frame
+{
+ u16 len;
+ u16 status;
+ u16 rntpc;
+ u16 rcvcc;
+ u32 pad1;
+ u32 pad2;
+ u8 data[1];
+ /* And frame continues.. */
+};
+
+#define PRIV_BYTES sizeof(struct mace68k_data)
+
+static int mace68k_open(struct device *dev);
+static int mace68k_close(struct device *dev);
+static int mace68k_xmit_start(struct sk_buff *skb, struct device *dev);
+static struct net_device_stats *mace68k_stats(struct device *dev);
+static void mace68k_set_multicast(struct device *dev);
+static void mace68k_reset(struct device *dev);
+static int mace68k_set_address(struct device *dev, void *addr);
+static void mace68k_interrupt(int irq, void *dev_id, struct pt_regs *regs);
+static void mace68k_dma_intr(int irq, void *dev_id, struct pt_regs *regs);
+static void mace68k_set_timeout(struct device *dev);
+static void mace68k_tx_timeout(unsigned long data);
+
+/*
+ * PSC DMA engine control. As you'd expect on a macintosh its
+ * more like a lawnmower engine supplied without instructions
+ *
+ * The basic theory of operation appears to be as follows.
+ *
+ * There are two sets of receive DMA registers and two sets
+ * of transmit DMA registers. Instead of the more traditional
+ * "ring buffer" approach the Mac68K DMA engine expects you
+ * to be loading one chain while the other runs, and then
+ * to flip register set. Each entry in the chain is a fixed
+ * length.
+ */
+
+/*
+ * Load a receive DMA channel with a base address and ring length
+ */
+
+static void psc_load_rxdma_base(int set, void *base)
+{
+ psc_write_word(PSC_ENETRD_CMD + set, 0x0100);
+ psc_write_long(PSC_ENETRD_ADDR + set, (u32)base);
+ psc_write_long(PSC_ENETRD_LEN + set, N_RX_RING);
+ psc_write_word(PSC_ENETRD_CMD + set, 0x9800);
+}
+
+/*
+ * Reset the receive DMA subsystem
+ */
+
+static void mace68k_rxdma_reset(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mace = mp->mace;
+ u8 mcc = mace->maccc;
+
+ /*
+ * Turn off receive
+ */
+
+ mcc&=~ENRCV;
+ mace->maccc=mcc;
+
+ /*
+ * Program the DMA
+ */
+
+ psc_write_word(PSC_ENETRD_CTL, 0x8800);
+ psc_load_rxdma_base(0x0, (void *)virt_to_bus(mp->rx_ring));
+ psc_write_word(PSC_ENETRD_CTL, 0x0400);
+
+ psc_write_word(PSC_ENETRD_CTL, 0x8800);
+ psc_load_rxdma_base(0x10, (void *)virt_to_bus(mp->rx_ring));
+ psc_write_word(PSC_ENETRD_CTL, 0x0400);
+
+ mace->maccc=mcc|ENRCV;
+
+ psc_write_word(PSC_ENETRD_CTL, 0x9800);
+ psc_write_word(PSC_ENETRD_CTL+0x10, 0x9800);
+}
+
+/*
+ * Reset the transmit DMA subsystem
+ */
+
+static void mace68k_txdma_reset(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mace = mp->mace;
+ u8 mcc = mace->maccc;
+
+ psc_write_word(PSC_ENETWR_CTL,0x8800);
+
+ mace->maccc = mcc&~ENXMT;
+ psc_write_word(PSC_ENETWR_CTL,0x400);
+ mace->maccc = mcc;
+}
+
+/*
+ * Disable DMA
+ */
+
+static void mace68k_dma_off(struct device *dev)
+{
+ psc_write_word(PSC_ENETRD_CTL, 0x8800);
+ psc_write_word(PSC_ENETRD_CTL, 0x1000);
+ psc_write_word(PSC_ENETRD_CMD, 0x1100);
+ psc_write_word(PSC_ENETRD_CMD+0x10, 0x1100);
+
+ psc_write_word(PSC_ENETWR_CTL, 0x8800);
+ psc_write_word(PSC_ENETWR_CTL, 0x1000);
+ psc_write_word(PSC_ENETWR_CMD, 0x1100);
+ psc_write_word(PSC_ENETWR_CMD+0x10, 0x1100);
+}
+
+/* Bit-reverse one byte of an ethernet hardware address. */
+
+static int
+bitrev(int b)
+{
+ int d = 0, i;
+
+ for (i = 0; i < 8; ++i, b >>= 1)
+ d = (d << 1) | (b & 1);
+ return d;
+}
+
+/*
+ * Not really much of a probe. The hardware table tells us if this
+ * model of Macintrash has a MACE (AV macintoshes)
+ */
+
+int mace68k_probe(struct device *unused)
+{
+ int j;
+ static int once=0;
+ struct mace68k_data *mp;
+ unsigned char *addr;
+ struct device *dev;
+ unsigned char checksum = 0;
+
+ /*
+ * There can be only one...
+ */
+
+ if (once) return -ENODEV;
+
+ once = 1;
+
+ if (macintosh_config->ether_type != MAC_ETHER_MACE) return -ENODEV;
+
+ printk("MACE ethernet should be present ");
+
+ dev = init_etherdev(0, PRIV_BYTES);
+ if(dev==NULL)
+ {
+ printk("no free memory.\n");
+ return -ENOMEM;
+ }
+ mp = (struct mace68k_data *) dev->priv;
+ dev->base_addr = (u32)MACE_BASE;
+ mp->mace = (volatile struct mace *) MACE_BASE;
+
+ printk("at 0x%p", mp->mace);
+
+ /*
+ * 16K RX ring and 4K TX ring should do nicely
+ */
+
+ mp->rx_ring=(void *)__get_free_pages(GFP_KERNEL, 2);
+ mp->tx_ring=(void *)__get_free_page(GFP_KERNEL);
+
+ printk(".");
+
+ if(mp->tx_ring==NULL || mp->rx_ring==NULL)
+ {
+ if(mp->tx_ring)
+ free_page((u32)mp->tx_ring);
+// if(mp->rx_ring)
+// __free_pages(mp->rx_ring,2);
+ printk("\nNo memory for ring buffers.\n");
+ return -ENOMEM;
+ }
+
+ /* We want the receive data to be uncached. We dont care about the
+ byte reading order */
+
+ printk(".");
+ kernel_set_cachemode((void *)mp->rx_ring, 16384, IOMAP_NOCACHE_NONSER);
+
+ printk(".");
+ /* The transmit buffer needs to be write through */
+ kernel_set_cachemode((void *)mp->tx_ring, 4096, IOMAP_WRITETHROUGH);
+
+ printk(" Ok\n");
+ dev->irq = IRQ_MAC_MACE;
+ printk(KERN_INFO "%s: MACE at", dev->name);
+
+ /*
+ * The PROM contains 8 bytes which total 0xFF when XOR'd
+ * together. Due to the usual peculiar apple brain damage
+ * the bytes are spaced out in a strange boundary and the
+ * bits are reversed.
+ */
+
+ addr = (void *)MACE_PROM;
+
+ for (j = 0; j < 6; ++j)
+ {
+ u8 v=bitrev(addr[j<<4]);
+ checksum^=v;
+ dev->dev_addr[j] = v;
+ printk("%c%.2x", (j ? ':' : ' '), dev->dev_addr[j]);
+ }
+ for (; j < 8; ++j)
+ {
+ checksum^=bitrev(addr[j<<4]);
+ }
+
+ if(checksum!=0xFF)
+ {
+ printk(" (invalid checksum)\n");
+ return -ENODEV;
+ }
+ printk("\n");
+
+ memset(&mp->stats, 0, sizeof(mp->stats));
+ init_timer(&mp->tx_timeout);
+ mp->timeout_active = 0;
+
+ dev->open = mace68k_open;
+ dev->stop = mace68k_close;
+ dev->hard_start_xmit = mace68k_xmit_start;
+ dev->get_stats = mace68k_stats;
+ dev->set_multicast_list = mace68k_set_multicast;
+ dev->set_mac_address = mace68k_set_address;
+
+ ether_setup(dev);
+
+ mp = (struct mace68k_data *) dev->priv;
+ mp->maccc = ENXMT | ENRCV;
+ mp->dma_intr = IRQ_MAC_MACE_DMA;
+
+ psc_write_word(PSC_ENETWR_CTL, 0x9000);
+ psc_write_word(PSC_ENETRD_CTL, 0x9000);
+ psc_write_word(PSC_ENETWR_CTL, 0x0400);
+ psc_write_word(PSC_ENETRD_CTL, 0x0400);
+
+ mace68k_dma_off(dev);
+
+ return 0;
+}
+
+/*
+ * Reset a MACE controller
+ */
+
+static void mace68k_reset(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mb = mp->mace;
+ int i;
+
+ /* soft-reset the chip */
+ mb->biucc = SWRST;
+ udelay(100);
+
+ mb->biucc = XMTSP_64;
+ mb->imr = 0xff; /* disable all intrs for now */
+ i = mb->ir;
+ mb->maccc = 0; /* turn off tx, rx */
+ mb->utr = RTRD;
+ mb->fifocc = RCVFW_64;
+ mb->xmtfc = AUTO_PAD_XMIT; /* auto-pad short frames */
+
+ /* load up the hardware address */
+
+ mb->iac = ADDRCHG | PHYADDR;
+
+ while ((mb->iac & ADDRCHG) != 0);
+
+ for (i = 0; i < 6; ++i)
+ mb->padr = dev->dev_addr[i];
+
+ /* clear the multicast filter */
+ mb->iac = ADDRCHG | LOGADDR;
+
+ while ((mb->iac & ADDRCHG) != 0);
+
+ for (i = 0; i < 8; ++i)
+ mb->ladrf = 0;
+
+ mb->plscc = PORTSEL_GPSI + ENPLSIO;
+}
+
+/*
+ * Load the address on a mace controller.
+ */
+
+static int mace68k_set_address(struct device *dev, void *addr)
+{
+ unsigned char *p = addr;
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mb = mp->mace;
+ int i;
+ unsigned long flags;
+
+ save_flags(flags);
+ cli();
+
+ /* load up the hardware address */
+ mb->iac = ADDRCHG | PHYADDR;
+ while ((mb->iac & ADDRCHG) != 0);
+
+ for (i = 0; i < 6; ++i)
+ mb->padr = dev->dev_addr[i] = p[i];
+ /* note: setting ADDRCHG clears ENRCV */
+ mb->maccc = mp->maccc;
+ restore_flags(flags);
+ return 0;
+}
+
+/*
+ * Open the Macintosh MACE. Most of this is playing with the DMA
+ * engine. The ethernet chip is quite friendly.
+ */
+
+static int mace68k_open(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mb = mp->mace;
+
+ /* reset the chip */
+ mace68k_reset(dev);
+
+ mp->rx_done = 0;
+ mace68k_rxdma_reset(dev);
+
+ /*
+ * The interrupt is fixed and comes off the PSC.
+ */
+
+ if (request_irq(dev->irq, mace68k_interrupt, 0, "68K MACE", dev))
+ {
+ printk(KERN_ERR "MACE: can't get irq %d\n", dev->irq);
+ return -EAGAIN;
+ }
+
+ /*
+ * Ditto the DMA interrupt.
+ */
+
+ if (request_irq(IRQ_MAC_MACE_DMA, mace68k_dma_intr, 0, "68K MACE DMA",
+ dev))
+ {
+ printk(KERN_ERR "MACE: can't get irq %d\n", IRQ_MAC_MACE_DMA);
+ return -EAGAIN;
+ }
+
+ /* Activate the Mac DMA engine */
+
+ mp->tx_slot = 0; /* Using register set 0 */
+ mp->tx_count = 1; /* 1 Buffer ready for use */
+ mace68k_txdma_reset(dev);
+
+ /* turn it on! */
+ mb->maccc = mp->maccc;
+ /* enable all interrupts except receive interrupts */
+ mb->imr = RCVINT;
+ return 0;
+}
+
+/*
+ * Shut down the mace and its interrupt channel
+ */
+
+static int mace68k_close(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mb = mp->mace;
+
+ /* disable rx and tx */
+ mb->maccc = 0;
+ mb->imr = 0xff; /* disable all intrs */
+
+ /* disable rx and tx dma */
+
+ mace68k_dma_off(dev);
+
+ free_irq(dev->irq, dev);
+ free_irq(IRQ_MAC_MACE_DMA, dev);
+ return 0;
+}
+
+static inline void mace68k_set_timeout(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ unsigned long flags;
+
+ save_flags(flags);
+ cli();
+ if (mp->timeout_active)
+ del_timer(&mp->tx_timeout);
+ mp->tx_timeout.expires = jiffies + TX_TIMEOUT;
+ mp->tx_timeout.function = mace68k_tx_timeout;
+ mp->tx_timeout.data = (unsigned long) dev;
+ add_timer(&mp->tx_timeout);
+ mp->timeout_active = 1;
+ restore_flags(flags);
+}
+
+/*
+ * Transmit a frame
+ */
+
+static int mace68k_xmit_start(struct sk_buff *skb, struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ /*
+ * This may need atomic types ???
+ */
+ if(mp->tx_count == 0)
+ {
+ dev->tbusy=1;
+ return 1;
+ }
+ mp->tx_count--;
+
+ /*
+ * FIXME:
+ * This is hackish. The memcpy probably isnt needed but
+ * the rules for alignment are not known. Ideally we'd like
+ * to just blast the skb directly to ethernet. We also don't
+ * use the ring properly - just a one frame buffer. That
+ * also requires cache pushes ;).
+ */
+ memcpy((void *)mp->tx_ring, skb, skb->len);
+ psc_write_long(PSC_ENETWR_ADDR + mp->tx_slot, virt_to_bus(mp->tx_ring));
+ psc_write_long(PSC_ENETWR_LEN + mp->tx_slot, skb->len);
+ psc_write_word(PSC_ENETWR_CMD + mp->tx_slot, 0x9800);
+ mp->stats.tx_packets++;
+ mp->stats.tx_bytes+=skb->len;
+ dev_kfree_skb(skb);
+ return 0;
+}
+
+static struct net_device_stats *mace68k_stats(struct device *dev)
+{
+ struct mace68k_data *p = (struct mace68k_data *) dev->priv;
+ return &p->stats;
+}
+
+/*
+ * CRC polynomial - used in working out multicast filter bits.
+ */
+#define CRC_POLY 0xedb88320
+
+static void mace68k_set_multicast(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mb = mp->mace;
+ int i, j, k, b;
+ unsigned long crc;
+
+ mp->maccc &= ~PROM;
+ if (dev->flags & IFF_PROMISC)
+ {
+ mp->maccc |= PROM;
+ } else
+ {
+ unsigned char multicast_filter[8];
+ struct dev_mc_list *dmi = dev->mc_list;
+
+ if (dev->flags & IFF_ALLMULTI)
+ {
+ for (i = 0; i < 8; i++)
+ multicast_filter[i] = 0xff;
+ } else
+ {
+ for (i = 0; i < 8; i++)
+ multicast_filter[i] = 0;
+ for (i = 0; i < dev->mc_count; i++)
+ {
+ crc = ~0;
+ for (j = 0; j < 6; ++j)
+ {
+ b = dmi->dmi_addr[j];
+ for (k = 0; k < 8; ++k)
+ {
+ if ((crc ^ b) & 1)
+ crc = (crc >> 1) ^ CRC_POLY;
+ else
+ crc >>= 1;
+ b >>= 1;
+ }
+ }
+ j = crc >> 26; /* bit number in multicast_filter */
+ multicast_filter[j >> 3] |= 1 << (j & 7);
+ dmi = dmi->next;
+ }
+ }
+#if 0
+ printk("Multicast filter :");
+ for (i = 0; i < 8; i++)
+ printk("%02x ", multicast_filter[i]);
+ printk("\n");
+#endif
+
+ mb->iac = ADDRCHG | LOGADDR;
+ while ((mb->iac & ADDRCHG) != 0);
+
+ for (i = 0; i < 8; ++i)
+ mb->ladrf = multicast_filter[i];
+ }
+ /* reset maccc */
+ mb->maccc = mp->maccc;
+}
+
+/*
+ * Miscellaneous interrupts are handled here. We may end up
+ * having to bash the chip on the head for bad errors
+ */
+
+static void mace68k_handle_misc_intrs(struct mace68k_data *mp, int intr)
+{
+ volatile struct mace *mb = mp->mace;
+ static int mace68k_babbles, mace68k_jabbers;
+
+ if (intr & MPCO)
+ mp->stats.rx_missed_errors += 256;
+ mp->stats.rx_missed_errors += mb->mpc; /* reading clears it */
+ if (intr & RNTPCO)
+ mp->stats.rx_length_errors += 256;
+ mp->stats.rx_length_errors += mb->rntpc; /* reading clears it */
+ if (intr & CERR)
+ ++mp->stats.tx_heartbeat_errors;
+ if (intr & BABBLE)
+ if (mace68k_babbles++ < 4)
+ printk(KERN_DEBUG "mace: babbling transmitter\n");
+ if (intr & JABBER)
+ if (mace68k_jabbers++ < 4)
+ printk(KERN_DEBUG "mace: jabbering transceiver\n");
+}
+
+/*
+ * A transmit error has occured. (We kick the transmit side from
+ * the DMA completion)
+ */
+
+static void mace68k_xmit_error(struct device *dev)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mb = mp->mace;
+ u8 xmtfs, xmtrc;
+
+ xmtfs = mb->xmtfs;
+ xmtrc = mb->xmtrc;
+
+ if(xmtfs & XMTSV)
+ {
+ if(xmtfs & UFLO)
+ {
+ printk("%s: DMA underrun.\n", dev->name);
+ mp->stats.tx_errors++;
+ mp->stats.tx_fifo_errors++;
+ mace68k_reset(dev);
+ }
+ if(xmtfs & RTRY)
+ mp->stats.collisions++;
+ }
+ mark_bh(NET_BH);
+}
+
+/*
+ * A receive interrupt occured.
+ */
+
+static void mace68k_recv_interrupt(struct device *dev)
+{
+// struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+// volatile struct mace *mb = mp->mace;
+}
+
+/*
+ * Process the chip interrupt
+ */
+
+static void mace68k_interrupt(int irq, void *dev_id, struct pt_regs *regs)
+{
+ struct device *dev = (struct device *) dev_id;
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ volatile struct mace *mb = mp->mace;
+ u8 ir;
+
+ ir = mb->ir;
+ mace68k_handle_misc_intrs(mp, ir);
+
+ if(ir&XMTINT)
+ mace68k_xmit_error(dev);
+ if(ir&RCVINT)
+ mace68k_recv_interrupt(dev);
+}
+
+static void mace68k_tx_timeout(unsigned long data)
+{
+// struct device *dev = (struct device *) data;
+// struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+// volatile struct mace *mb = mp->mace;
+}
+
+/*
+ * Handle a newly arrived frame
+ */
+
+static void mace_dma_rx_frame(struct device *dev, struct mace_frame *mf)
+{
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+ struct sk_buff *skb;
+
+ if(mf->status&RS_OFLO)
+ {
+ printk("%s: fifo overflow.\n", dev->name);
+ mp->stats.rx_errors++;
+ mp->stats.rx_fifo_errors++;
+ }
+ if(mf->status&(RS_CLSN|RS_FRAMERR|RS_FCSERR))
+ mp->stats.rx_errors++;
+
+ if(mf->status&RS_CLSN)
+ mp->stats.collisions++;
+ if(mf->status&RS_FRAMERR)
+ mp->stats.rx_frame_errors++;
+ if(mf->status&RS_FCSERR)
+ mp->stats.rx_crc_errors++;
+
+ skb = dev_alloc_skb(mf->len+2);
+ if(skb==NULL)
+ {
+ mp->stats.rx_dropped++;
+ return;
+ }
+ skb_reserve(skb,2);
+ memcpy(skb_put(skb, mf->len), mf->data, mf->len);
+
+ skb->protocol = eth_type_trans(skb, dev);
+ netif_rx(skb);
+ mp->stats.rx_packets++;
+ mp->stats.rx_bytes+=mf->len;
+}
+
+/*
+ * The PSC has passed us a DMA interrupt event.
+ */
+
+static void mace68k_dma_intr(int irq, void *dev_id, struct pt_regs *regs)
+{
+ struct device *dev = (struct device *) dev_id;
+ struct mace68k_data *mp = (struct mace68k_data *) dev->priv;
+
+ u32 psc_status;
+
+ /* It seems this must be allowed to stabilise ?? */
+
+ while((psc_status=psc_read_long(0x0804))!=psc_read_long(0x0804));
+
+ /*
+ * Was this an ethernet event ?
+ */
+
+ if(psc_status&0x60000000)
+ {
+ /*
+ * Process the read queue
+ */
+
+ u16 psc_status = psc_read_word(PSC_ENETRD_CTL);
+
+ if(psc_status&0x2000)
+ {
+ mace68k_rxdma_reset(dev);
+ mp->rx_done = 0;
+ }
+
+ else if(psc_status&0x100)
+ {
+ int left;
+
+ psc_write_word(PSC_ENETRD_CMD+mp->rx_slot, 0x1100);
+ left=psc_read_long(PSC_ENETRD_LEN+mp->rx_slot);
+ /* read packets */
+
+ while(mp->rx_done < left)
+ {
+ struct mace_frame *mf=((struct mace_frame *)
+ mp->rx_ring)+mp->rx_done++;
+ mace_dma_rx_frame(dev, mf);
+ }
+
+ if(left == 0) /* Out of DMA room */
+ {
+ psc_load_rxdma_base(mp->rx_slot,
+ (void *)virt_to_phys(mp->rx_ring));
+ mp->rx_slot^=16;
+ mp->rx_done = 0;
+ }
+ else
+ {
+ psc_write_word(PSC_ENETRD_CMD+mp->rx_slot,
+ 0x9800);
+ }
+
+ }
+
+ /*
+ * Process the write queue
+ */
+
+ psc_status = psc_read_word(PSC_ENETWR_CTL);
+ if(psc_status&0x2000) {
+ mace68k_txdma_reset(dev);
+ } else if(psc_status&0x0100) {
+ psc_write_word(PSC_ENETWR_CMD+mp->tx_slot, 0x100);
+ mp->tx_slot^=16;
+ mp->tx_count++;
+ dev->tbusy = 0;
+ mark_bh(NET_BH);
+ }
+ }
+}
FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen (who was at: slshen@lbl.gov)