static int __cpuinit smp_85xx_kick_cpu(int nr)
{
unsigned long flags;
const u64 *cpu_rel_addr;
__iomem struct epapr_spin_table *spin_table;
struct device_node *np;
int hw_cpu = get_hard_smp_processor_id(nr);
int ioremappable;
int ret = 0;
WARN_ON(nr < 0 || nr >= NR_CPUS);
WARN_ON(hw_cpu < 0 || hw_cpu >= NR_CPUS);
pr_debug("smp_85xx_kick_cpu: kick CPU #%d\n", nr);
np = of_get_cpu_node(nr, NULL);
cpu_rel_addr = of_get_property(np, "cpu-release-addr", NULL);
if (cpu_rel_addr == NULL) {
printk(KERN_ERR "No cpu-release-addr for cpu %d\n", nr);
return -ENOENT;
}
/*
* A secondary core could be in a spinloop in the bootpage
* (0xfffff000), somewhere in highmem, or somewhere in lowmem.
* The bootpage and highmem can be accessed via ioremap(), but
* we need to directly access the spinloop if its in lowmem.
*/
ioremappable = *cpu_rel_addr > virt_to_phys(high_memory);
/* Map the spin table */
if (ioremappable)
spin_table = ioremap(*cpu_rel_addr,
sizeof(struct epapr_spin_table));
else
spin_table = phys_to_virt(*cpu_rel_addr);
local_irq_save(flags);
#ifdef CONFIG_PPC32
#ifdef CONFIG_HOTPLUG_CPU
/* Corresponding to generic_set_cpu_dead() */
generic_set_cpu_up(nr);
if (system_state == SYSTEM_RUNNING) {
out_be32(&spin_table->addr_l, 0);
/*
* We don't set the BPTR register here since it already points
* to the boot page properly.
*/
mpic_reset_core(hw_cpu);
/* wait until core is ready... */
if (!spin_event_timeout(in_be32(&spin_table->addr_l) == 1,
10000, 100)) {
pr_err("%s: timeout waiting for core %d to reset\n",
__func__, hw_cpu);
ret = -ENOENT;
goto out;
}
/* clear the acknowledge status */
__secondary_hold_acknowledge = -1;
}
#endif
out_be32(&spin_table->pir, hw_cpu);
out_be32(&spin_table->addr_l, __pa(__early_start));
if (!ioremappable)
flush_dcache_range((ulong)spin_table,
(ulong)spin_table + sizeof(struct epapr_spin_table));
/* Wait a bit for the CPU to ack. */
if (!spin_event_timeout(__secondary_hold_acknowledge == hw_cpu,
10000, 100)) {
pr_err("%s: timeout waiting for core %d to ack\n",
__func__, hw_cpu);
ret = -ENOENT;
goto out;
}
out:
#else
smp_generic_kick_cpu(nr);
out_be32(&spin_table->pir, hw_cpu);
out_be64((u64 *)(&spin_table->addr_h),
__pa((u64)*((unsigned long long *)generic_secondary_smp_init)));
if (!ioremappable)
flush_dcache_range((ulong)spin_table,
(ulong)spin_table + sizeof(struct epapr_spin_table));
#endif
local_irq_restore(flags);
if (ioremappable)
iounmap(spin_table);
return ret;
}
int zynq_load(Xilinx_desc *desc, const void *buf, size_t bsize)
{
unsigned long ts; /* Timestamp */
u32 partialbit = 0;
u32 i, control, isr_status, status, swap, diff;
u32 *buf_start;
/* Detect if we are going working with partial or full bitstream */
if (bsize != desc->size) {
printf("%s: Working with partial bitstream\n", __func__);
partialbit = 1;
}
buf_start = check_data((u8 *)buf, bsize, &swap);
if (!buf_start)
return FPGA_FAIL;
/* Check if data is postpone from start */
diff = (u32)buf_start - (u32)buf;
if (diff) {
printf("%s: Bitstream is not validated yet (diff %x)\n",
__func__, diff);
return FPGA_FAIL;
}
if ((u32)buf < SZ_1M) {
printf("%s: Bitstream has to be placed up to 1MB (%x)\n",
__func__, (u32)buf);
return FPGA_FAIL;
}
if ((u32)buf != ALIGN((u32)buf, ARCH_DMA_MINALIGN)) {
u32 *new_buf = (u32 *)ALIGN((u32)buf, ARCH_DMA_MINALIGN);
printf("%s: Align buffer at %x to %x(swap %d)\n", __func__,
(u32)buf_start, (u32)new_buf, swap);
for (i = 0; i < (bsize/4); i++)
new_buf[i] = load_word(&buf_start[i], swap);
swap = SWAP_DONE;
buf = new_buf;
} else if (swap != SWAP_DONE) {
/* For bitstream which are aligned */
u32 *new_buf = (u32 *)buf;
printf("%s: Bitstream is not swapped(%d) - swap it\n", __func__,
swap);
for (i = 0; i < (bsize/4); i++)
new_buf[i] = load_word(&buf_start[i], swap);
swap = SWAP_DONE;
}
/* Clear loopback bit */
clrbits_le32(&devcfg_base->mctrl, DEVCFG_MCTRL_PCAP_LPBK);
if (!partialbit) {
zynq_slcr_devcfg_disable();
/* Setting PCFG_PROG_B signal to high */
control = readl(&devcfg_base->ctrl);
writel(control | DEVCFG_CTRL_PCFG_PROG_B, &devcfg_base->ctrl);
/* Setting PCFG_PROG_B signal to low */
writel(control & ~DEVCFG_CTRL_PCFG_PROG_B, &devcfg_base->ctrl);
/* Polling the PCAP_INIT status for Reset */
ts = get_timer(0);
while (readl(&devcfg_base->status) & DEVCFG_STATUS_PCFG_INIT) {
if (get_timer(ts) > CONFIG_SYS_FPGA_WAIT) {
printf("%s: Timeout wait for INIT to clear\n",
__func__);
return FPGA_FAIL;
}
}
/* Setting PCFG_PROG_B signal to high */
writel(control | DEVCFG_CTRL_PCFG_PROG_B, &devcfg_base->ctrl);
/* Polling the PCAP_INIT status for Set */
ts = get_timer(0);
while (!(readl(&devcfg_base->status) &
DEVCFG_STATUS_PCFG_INIT)) {
if (get_timer(ts) > CONFIG_SYS_FPGA_WAIT) {
printf("%s: Timeout wait for INIT to set\n",
__func__);
return FPGA_FAIL;
}
}
}
isr_status = readl(&devcfg_base->int_sts);
/* Clear it all, so if Boot ROM comes back, it can proceed */
writel(0xFFFFFFFF, &devcfg_base->int_sts);
if (isr_status & DEVCFG_ISR_FATAL_ERROR_MASK) {
debug("%s: Fatal errors in PCAP 0x%X\n", __func__, isr_status);
/* If RX FIFO overflow, need to flush RX FIFO first */
if (isr_status & DEVCFG_ISR_RX_FIFO_OV) {
writel(DEVCFG_MCTRL_RFIFO_FLUSH, &devcfg_base->mctrl);
writel(0xFFFFFFFF, &devcfg_base->int_sts);
}
return FPGA_FAIL;
}
status = readl(&devcfg_base->status);
debug("%s: Status = 0x%08X\n", __func__, status);
if (status & DEVCFG_STATUS_DMA_CMD_Q_F) {
debug("%s: Error: device busy\n", __func__);
return FPGA_FAIL;
}
debug("%s: Device ready\n", __func__);
if (!(status & DEVCFG_STATUS_DMA_CMD_Q_E)) {
if (!(readl(&devcfg_base->int_sts) & DEVCFG_ISR_DMA_DONE)) {
/* Error state, transfer cannot occur */
debug("%s: ISR indicates error\n", __func__);
return FPGA_FAIL;
} else {
/* Clear out the status */
writel(DEVCFG_ISR_DMA_DONE, &devcfg_base->int_sts);
}
}
if (status & DEVCFG_STATUS_DMA_DONE_CNT_MASK) {
/* Clear the count of completed DMA transfers */
writel(DEVCFG_STATUS_DMA_DONE_CNT_MASK, &devcfg_base->status);
}
debug("%s: Source = 0x%08X\n", __func__, (u32)buf);
debug("%s: Size = %zu\n", __func__, bsize);
/* flush(clean & invalidate) d-cache range buf */
flush_dcache_range((u32)buf, (u32)buf +
roundup(bsize, ARCH_DMA_MINALIGN));
/* Set up the transfer */
writel((u32)buf | 1, &devcfg_base->dma_src_addr);
writel(0xFFFFFFFF, &devcfg_base->dma_dst_addr);
writel(bsize >> 2, &devcfg_base->dma_src_len);
writel(0, &devcfg_base->dma_dst_len);
isr_status = readl(&devcfg_base->int_sts);
/* Polling the PCAP_INIT status for Set */
ts = get_timer(0);
while (!(isr_status & DEVCFG_ISR_DMA_DONE)) {
if (isr_status & DEVCFG_ISR_ERROR_FLAGS_MASK) {
debug("%s: Error: isr = 0x%08X\n", __func__,
isr_status);
debug("%s: Write count = 0x%08X\n", __func__,
readl(&devcfg_base->write_count));
debug("%s: Read count = 0x%08X\n", __func__,
readl(&devcfg_base->read_count));
return FPGA_FAIL;
}
if (get_timer(ts) > CONFIG_SYS_FPGA_PROG_TIME) {
printf("%s: Timeout wait for DMA to complete\n",
__func__);
return FPGA_FAIL;
}
isr_status = readl(&devcfg_base->int_sts);
}
debug("%s: DMA transfer is done\n", __func__);
/* Check FPGA configuration completion */
ts = get_timer(0);
while (!(isr_status & DEVCFG_ISR_PCFG_DONE)) {
if (get_timer(ts) > CONFIG_SYS_FPGA_WAIT) {
printf("%s: Timeout wait for FPGA to config\n",
__func__);
return FPGA_FAIL;
}
isr_status = readl(&devcfg_base->int_sts);
}
debug("%s: FPGA config done\n", __func__);
/* Clear out the DMA status */
writel(DEVCFG_ISR_DMA_DONE, &devcfg_base->int_sts);
if (!partialbit)
zynq_slcr_devcfg_enable();
return FPGA_SUCCESS;
}
/**
* Transmit one frame
* @param[in] dev Our ethernet device to handle
* @param[in] packet Pointer to the data to be transmitted
* @param[in] length Data count in bytes
* @return 0 on success
*/
static int fec_send(struct eth_device *dev, void *packet, int length)
{
unsigned int status;
uint32_t size, end;
uint32_t addr;
int timeout = FEC_XFER_TIMEOUT;
int ret = 0;
/*
* This routine transmits one frame. This routine only accepts
* 6-byte Ethernet addresses.
*/
struct fec_priv *fec = (struct fec_priv *)dev->priv;
/*
* Check for valid length of data.
*/
if ((length > 1500) || (length <= 0)) {
printf("Payload (%d) too large\n", length);
return -1;
}
/*
* Setup the transmit buffer. We are always using the first buffer for
* transmission, the second will be empty and only used to stop the DMA
* engine. We also flush the packet to RAM here to avoid cache trouble.
*/
#ifdef CONFIG_FEC_MXC_SWAP_PACKET
swap_packet((uint32_t *)packet, length);
#endif
addr = (uint32_t)packet;
end = roundup(addr + length, ARCH_DMA_MINALIGN);
addr &= ~(ARCH_DMA_MINALIGN - 1);
flush_dcache_range(addr, end);
writew(length, &fec->tbd_base[fec->tbd_index].data_length);
writel(addr, &fec->tbd_base[fec->tbd_index].data_pointer);
/*
* update BD's status now
* This block:
* - is always the last in a chain (means no chain)
* - should transmitt the CRC
* - might be the last BD in the list, so the address counter should
* wrap (-> keep the WRAP flag)
*/
status = readw(&fec->tbd_base[fec->tbd_index].status) & FEC_TBD_WRAP;
status |= FEC_TBD_LAST | FEC_TBD_TC | FEC_TBD_READY;
writew(status, &fec->tbd_base[fec->tbd_index].status);
/*
* Flush data cache. This code flushes both TX descriptors to RAM.
* After this code, the descriptors will be safely in RAM and we
* can start DMA.
*/
size = roundup(2 * sizeof(struct fec_bd), ARCH_DMA_MINALIGN);
addr = (uint32_t)fec->tbd_base;
flush_dcache_range(addr, addr + size);
/*
* Below we read the DMA descriptor's last four bytes back from the
* DRAM. This is important in order to make sure that all WRITE
* operations on the bus that were triggered by previous cache FLUSH
* have completed.
*
* Otherwise, on MX28, it is possible to observe a corruption of the
* DMA descriptors. Please refer to schematic "Figure 1-2" in MX28RM
* for the bus structure of MX28. The scenario is as follows:
*
* 1) ARM core triggers a series of WRITEs on the AHB_ARB2 bus going
* to DRAM due to flush_dcache_range()
* 2) ARM core writes the FEC registers via AHB_ARB2
* 3) FEC DMA starts reading/writing from/to DRAM via AHB_ARB3
*
* Note that 2) does sometimes finish before 1) due to reordering of
* WRITE accesses on the AHB bus, therefore triggering 3) before the
* DMA descriptor is fully written into DRAM. This results in occasional
* corruption of the DMA descriptor.
*/
readl(addr + size - 4);
/*
* Enable SmartDMA transmit task
*/
fec_tx_task_enable(fec);
/*
* Wait until frame is sent. On each turn of the wait cycle, we must
* invalidate data cache to see what's really in RAM. Also, we need
* barrier here.
*/
while (--timeout) {
if (!(readl(&fec->eth->x_des_active) & FEC_X_DES_ACTIVE_TDAR))
break;
}
if (!timeout) {
ret = -EINVAL;
goto out;
}
/*
* The TDAR bit is cleared when the descriptors are all out from TX
* but on mx6solox we noticed that the READY bit is still not cleared
* right after TDAR.
* These are two distinct signals, and in IC simulation, we found that
* TDAR always gets cleared prior than the READY bit of last BD becomes
* cleared.
* In mx6solox, we use a later version of FEC IP. It looks like that
* this intrinsic behaviour of TDAR bit has changed in this newer FEC
* version.
*
* Fix this by polling the READY bit of BD after the TDAR polling,
* which covers the mx6solox case and does not harm the other SoCs.
*/
timeout = FEC_XFER_TIMEOUT;
while (--timeout) {
invalidate_dcache_range(addr, addr + size);
if (!(readw(&fec->tbd_base[fec->tbd_index].status) &
FEC_TBD_READY))
break;
}
if (!timeout)
ret = -EINVAL;
out:
debug("fec_send: status 0x%x index %d ret %i\n",
readw(&fec->tbd_base[fec->tbd_index].status),
fec->tbd_index, ret);
/* for next transmission use the other buffer */
if (fec->tbd_index)
fec->tbd_index = 0;
else
fec->tbd_index = 1;
return ret;
}
static int zynq_gem_init(struct udevice *dev)
{
u32 i, nwconfig;
int ret;
unsigned long clk_rate = 0;
struct zynq_gem_priv *priv = dev_get_priv(dev);
struct zynq_gem_regs *regs = priv->iobase;
struct emac_bd *dummy_tx_bd = &priv->tx_bd[TX_FREE_DESC];
struct emac_bd *dummy_rx_bd = &priv->tx_bd[TX_FREE_DESC + 2];
if (!priv->init) {
/* Disable all interrupts */
writel(0xFFFFFFFF, ®s->idr);
/* Disable the receiver & transmitter */
writel(0, ®s->nwctrl);
writel(0, ®s->txsr);
writel(0, ®s->rxsr);
writel(0, ®s->phymntnc);
/* Clear the Hash registers for the mac address
* pointed by AddressPtr
*/
writel(0x0, ®s->hashl);
/* Write bits [63:32] in TOP */
writel(0x0, ®s->hashh);
/* Clear all counters */
for (i = 0; i < STAT_SIZE; i++)
readl(®s->stat[i]);
/* Setup RxBD space */
memset(priv->rx_bd, 0, RX_BUF * sizeof(struct emac_bd));
for (i = 0; i < RX_BUF; i++) {
priv->rx_bd[i].status = 0xF0000000;
priv->rx_bd[i].addr =
((ulong)(priv->rxbuffers) +
(i * PKTSIZE_ALIGN));
}
/* WRAP bit to last BD */
priv->rx_bd[--i].addr |= ZYNQ_GEM_RXBUF_WRAP_MASK;
/* Write RxBDs to IP */
writel((ulong)priv->rx_bd, ®s->rxqbase);
/* Setup for DMA Configuration register */
writel(ZYNQ_GEM_DMACR_INIT, ®s->dmacr);
/* Setup for Network Control register, MDIO, Rx and Tx enable */
setbits_le32(®s->nwctrl, ZYNQ_GEM_NWCTRL_MDEN_MASK);
/* Disable the second priority queue */
dummy_tx_bd->addr = 0;
dummy_tx_bd->status = ZYNQ_GEM_TXBUF_WRAP_MASK |
ZYNQ_GEM_TXBUF_LAST_MASK|
ZYNQ_GEM_TXBUF_USED_MASK;
dummy_rx_bd->addr = ZYNQ_GEM_RXBUF_WRAP_MASK |
ZYNQ_GEM_RXBUF_NEW_MASK;
dummy_rx_bd->status = 0;
flush_dcache_range((ulong)&dummy_tx_bd, (ulong)&dummy_tx_bd +
sizeof(dummy_tx_bd));
flush_dcache_range((ulong)&dummy_rx_bd, (ulong)&dummy_rx_bd +
sizeof(dummy_rx_bd));
writel((ulong)dummy_tx_bd, ®s->transmit_q1_ptr);
writel((ulong)dummy_rx_bd, ®s->receive_q1_ptr);
priv->init++;
}
ret = phy_startup(priv->phydev);
if (ret)
return ret;
if (!priv->phydev->link) {
printf("%s: No link.\n", priv->phydev->dev->name);
return -1;
}
nwconfig = ZYNQ_GEM_NWCFG_INIT;
if (priv->interface == PHY_INTERFACE_MODE_SGMII) {
nwconfig |= ZYNQ_GEM_NWCFG_SGMII_ENBL |
ZYNQ_GEM_NWCFG_PCS_SEL;
#ifdef CONFIG_ARM64
writel(readl(®s->pcscntrl) | ZYNQ_GEM_PCS_CTL_ANEG_ENBL,
®s->pcscntrl);
#endif
}
switch (priv->phydev->speed) {
case SPEED_1000:
writel(nwconfig | ZYNQ_GEM_NWCFG_SPEED1000,
®s->nwcfg);
clk_rate = ZYNQ_GEM_FREQUENCY_1000;
break;
case SPEED_100:
writel(nwconfig | ZYNQ_GEM_NWCFG_SPEED100,
®s->nwcfg);
clk_rate = ZYNQ_GEM_FREQUENCY_100;
break;
case SPEED_10:
clk_rate = ZYNQ_GEM_FREQUENCY_10;
break;
}
/* Change the rclk and clk only not using EMIO interface */
if (!priv->emio)
#ifndef CONFIG_CLK_ZYNQMP
zynq_slcr_gem_clk_setup((ulong)priv->iobase !=
ZYNQ_GEM_BASEADDR0, clk_rate);
#else
ret = clk_set_rate(&priv->clk, clk_rate);
if (IS_ERR_VALUE(ret))
return -1;
#endif
setbits_le32(®s->nwctrl, ZYNQ_GEM_NWCTRL_RXEN_MASK |
ZYNQ_GEM_NWCTRL_TXEN_MASK);
return 0;
}
static inline void macb_flush_rx_buffer(struct macb_device *macb)
{
flush_dcache_range(macb->rx_buffer_dma, macb->rx_buffer_dma +
MACB_RX_BUFFER_SIZE);
}
static void mxs_nand_flush_cmd_buf(struct mxs_nand_info *info)
{
uint32_t addr = (uint32_t)info->cmd_buf;
flush_dcache_range(addr, addr + MXS_NAND_COMMAND_BUFFER_SIZE);
}
static int fec_init(struct eth_device *dev, bd_t* bd)
{
struct fec_priv *fec = (struct fec_priv *)dev->priv;
uint32_t mib_ptr = (uint32_t)&fec->eth->rmon_t_drop;
uint32_t size;
int i, ret;
/* Initialize MAC address */
fec_set_hwaddr(dev);
/*
* Allocate transmit descriptors, there are two in total. This
* allocation respects cache alignment.
*/
if (!fec->tbd_base) {
size = roundup(2 * sizeof(struct fec_bd),
ARCH_DMA_MINALIGN);
fec->tbd_base = memalign(ARCH_DMA_MINALIGN, size);
if (!fec->tbd_base) {
ret = -ENOMEM;
goto err1;
}
memset(fec->tbd_base, 0, size);
fec_tbd_init(fec);
}
/*
* Allocate receive descriptors. This allocation respects cache
* alignment.
*/
if (!fec->rbd_base) {
size = roundup(FEC_RBD_NUM * sizeof(struct fec_bd),
ARCH_DMA_MINALIGN);
fec->rbd_base = memalign(ARCH_DMA_MINALIGN, size);
if (!fec->rbd_base) {
ret = -ENOMEM;
goto err2;
}
memset(fec->rbd_base, 0, size);
/*
* Initialize RxBD ring
*/
if (fec_rbd_init(fec, FEC_RBD_NUM, FEC_MAX_PKT_SIZE) < 0) {
ret = -ENOMEM;
goto err3;
}
flush_dcache_range((unsigned)fec->rbd_base,
(unsigned)fec->rbd_base + size);
}
fec_reg_setup(fec);
if (fec->xcv_type != SEVENWIRE)
fec_mii_setspeed(fec->bus->priv);
/*
* Set Opcode/Pause Duration Register
*/
writel(0x00010020, &fec->eth->op_pause); /* FIXME 0xffff0020; */
writel(0x2, &fec->eth->x_wmrk);
/*
* Set multicast address filter
*/
writel(0x00000000, &fec->eth->gaddr1);
writel(0x00000000, &fec->eth->gaddr2);
/* clear MIB RAM */
for (i = mib_ptr; i <= mib_ptr + 0xfc; i += 4)
writel(0, i);
/* FIFO receive start register */
writel(0x520, &fec->eth->r_fstart);
/* size and address of each buffer */
writel(FEC_MAX_PKT_SIZE, &fec->eth->emrbr);
writel((uint32_t)fec->tbd_base, &fec->eth->etdsr);
writel((uint32_t)fec->rbd_base, &fec->eth->erdsr);
#ifndef CONFIG_PHYLIB
if (fec->xcv_type != SEVENWIRE)
miiphy_restart_aneg(dev);
#endif
fec_open(dev);
return 0;
err3:
free(fec->rbd_base);
err2:
free(fec->tbd_base);
err1:
return ret;
}
/*******************************************************************************
* This function initializes the power domain topology tree by querying the
* platform. The power domain nodes higher than the CPU are populated in the
* array psci_non_cpu_pd_nodes[] and the CPU power domains are populated in
* psci_cpu_pd_nodes[]. The platform exports its static topology map through the
* populate_power_domain_topology_tree() API. The algorithm populates the
* psci_non_cpu_pd_nodes and psci_cpu_pd_nodes iteratively by using this
* topology map. On a platform that implements two clusters of 2 cpus each, and
* supporting 3 domain levels, the populated psci_non_cpu_pd_nodes would look
* like this:
*
* ---------------------------------------------------
* | system node | cluster 0 node | cluster 1 node |
* ---------------------------------------------------
*
* And populated psci_cpu_pd_nodes would look like this :
* <- cpus cluster0 -><- cpus cluster1 ->
* ------------------------------------------------
* | CPU 0 | CPU 1 | CPU 2 | CPU 3 |
* ------------------------------------------------
******************************************************************************/
int psci_setup(void)
{
const unsigned char *topology_tree;
/* Query the topology map from the platform */
topology_tree = plat_get_power_domain_tree_desc();
/* Populate the power domain arrays using the platform topology map */
populate_power_domain_tree(topology_tree);
/* Update the CPU limits for each node in psci_non_cpu_pd_nodes */
psci_update_pwrlvl_limits();
/* Populate the mpidr field of cpu node for this CPU */
psci_cpu_pd_nodes[plat_my_core_pos()].mpidr =
read_mpidr() & MPIDR_AFFINITY_MASK;
#if !USE_COHERENT_MEM
/*
* The psci_non_cpu_pd_nodes only needs flushing when it's not allocated in
* coherent memory.
*/
flush_dcache_range((uintptr_t) &psci_non_cpu_pd_nodes,
sizeof(psci_non_cpu_pd_nodes));
#endif
flush_dcache_range((uintptr_t) &psci_cpu_pd_nodes,
sizeof(psci_cpu_pd_nodes));
psci_init_req_local_pwr_states();
/*
* Set the requested and target state of this CPU and all the higher
* power domain levels for this CPU to run.
*/
psci_set_pwr_domains_to_run(PLAT_MAX_PWR_LVL);
plat_setup_psci_ops((uintptr_t)psci_entrypoint,
&psci_plat_pm_ops);
assert(psci_plat_pm_ops);
/* Initialize the psci capability */
psci_caps = PSCI_GENERIC_CAP;
if (psci_plat_pm_ops->pwr_domain_off)
psci_caps |= define_psci_cap(PSCI_CPU_OFF);
if (psci_plat_pm_ops->pwr_domain_on &&
psci_plat_pm_ops->pwr_domain_on_finish)
psci_caps |= define_psci_cap(PSCI_CPU_ON_AARCH64);
if (psci_plat_pm_ops->pwr_domain_suspend &&
psci_plat_pm_ops->pwr_domain_suspend_finish) {
psci_caps |= define_psci_cap(PSCI_CPU_SUSPEND_AARCH64);
if (psci_plat_pm_ops->get_sys_suspend_power_state)
psci_caps |= define_psci_cap(PSCI_SYSTEM_SUSPEND_AARCH64);
}
if (psci_plat_pm_ops->system_off)
psci_caps |= define_psci_cap(PSCI_SYSTEM_OFF);
if (psci_plat_pm_ops->system_reset)
psci_caps |= define_psci_cap(PSCI_SYSTEM_RESET);
return 0;
}
static int32_t tbase_init_entry()
{
DBG_PRINTF("tbase_init\n\r");
// Save el1 registers in case non-secure world has already been set up.
cm_el1_sysregs_context_save(NON_SECURE);
uint64_t mpidr = read_mpidr();
uint32_t linear_id = platform_get_core_pos(mpidr);
tbase_context *tbase_ctx = &secure_context[linear_id];
// Note: mapping is 1:1, so physical and virtual addresses are here the same.
cpu_context_t *ns_entry_context = (cpu_context_t *) cm_get_context(mpidr, NON_SECURE);
// ************************************************************************************
// Configure parameter passing to tbase
// Calculate page start addresses for register areas.
registerFileStart[REGISTER_FILE_NWD] = page_align((uint64_t)&ns_entry_context, DOWN);
registerFileStart[REGISTER_FILE_MONITOR] = page_align((uint64_t)&msm_area, DOWN);
// Calculate page end addresses for register areas.
registerFileEnd[REGISTER_FILE_NWD] = (uint64_t)(&ns_entry_context[TBASE_CORE_COUNT]);
registerFileEnd[REGISTER_FILE_MONITOR] = ((uint64_t)&msm_area) +sizeof(msm_area);
int32_t totalPages = 0;
for (int area=0; area<REGISTER_FILE_COUNT; area++) {
int32_t pages = page_align(registerFileEnd[area] - registerFileStart[area], UP) / PAGE_SIZE;
assert( pages +totalPages <= TBASE_INTERFACE_PAGES );
tbase_init_register_file(area, totalPages, pages);
totalPages += pages;
}
// ************************************************************************************
// Create boot structure
tbaseBootCfg.magic = TBASE_BOOTCFG_MAGIC;
tbaseBootCfg.length = sizeof(bootCfg_t);
tbaseBootCfg.version = TBASE_MONITOR_INTERFACE_VERSION;
tbaseBootCfg.dRamBase = TBASE_NWD_DRAM_BASE;
tbaseBootCfg.dRamSize = TBASE_NWD_DRAM_SIZE;
tbaseBootCfg.secDRamBase = TBASE_SWD_DRAM_BASE;
tbaseBootCfg.secDRamSize = TBASE_SWD_DRAM_SIZE;
tbaseBootCfg.secIRamBase = TBASE_SWD_IMEM_BASE;
tbaseBootCfg.secIRamSize = TBASE_SWD_IMEM_SIZE;
tbaseBootCfg.conf_mair_el3 = read_mair_el3();
tbaseBootCfg.MSMPteCount = totalPages;
tbaseBootCfg.MSMBase = (uint64_t)registerFileL2;
tbaseBootCfg.gic_distributor_base = TBASE_GIC_DIST_BASE;
tbaseBootCfg.gic_cpuinterface_base = TBASE_GIC_CPU_BASE;
tbaseBootCfg.gic_version = TBASE_GIC_VERSION;
tbaseBootCfg.total_number_spi = TBASE_SPI_COUNT;
tbaseBootCfg.ssiq_number = TBASE_SSIQ_NRO;
tbaseBootCfg.flags = TBASE_MONITOR_FLAGS;
DBG_PRINTF("*** tbase boot cfg ***\n\r");
DBG_PRINTF("* magic : 0x%.X\n\r",tbaseBootCfg.magic);
DBG_PRINTF("* length : 0x%.X\n\r",tbaseBootCfg.length);
DBG_PRINTF("* version : 0x%.X\n\r",tbaseBootCfg.version);
DBG_PRINTF("* dRamBase : 0x%.X\n\r",tbaseBootCfg.dRamBase);
DBG_PRINTF("* dRamSize : 0x%.X\n\r",tbaseBootCfg.dRamSize);
DBG_PRINTF("* secDRamBase : 0x%.X\n\r",tbaseBootCfg.secDRamBase);
DBG_PRINTF("* secDRamSize : 0x%.X\n\r",tbaseBootCfg.secDRamSize);
DBG_PRINTF("* secIRamBase : 0x%.X\n\r",tbaseBootCfg.secIRamBase);
DBG_PRINTF("* secIRamSize : 0x%.X\n\r",tbaseBootCfg.secIRamSize);
DBG_PRINTF("* conf_mair_el3 : 0x%.X\n\r",tbaseBootCfg.conf_mair_el3);
DBG_PRINTF("* MSMPteCount : 0x%.X\n\r",tbaseBootCfg.MSMPteCount);
DBG_PRINTF("* MSMBase : 0x%.X\n\r",tbaseBootCfg.MSMBase);
DBG_PRINTF("* gic_distributor_base : 0x%.X\n\r",tbaseBootCfg.gic_distributor_base);
DBG_PRINTF("* gic_cpuinterface_base : 0x%.X\n\r",tbaseBootCfg.gic_cpuinterface_base);
DBG_PRINTF("* gic_version : 0x%.X\n\r",tbaseBootCfg.gic_version);
DBG_PRINTF("* total_number_spi : 0x%.X\n\r",tbaseBootCfg.total_number_spi);
DBG_PRINTF("* ssiq_number : 0x%.X\n\r",tbaseBootCfg.ssiq_number);
DBG_PRINTF("* flags : 0x%.X\n\r",tbaseBootCfg.flags);
// ************************************************************************************
// tbaseBootCfg and l2 entries may be accesses uncached, so must flush those.
flush_dcache_range((unsigned long)&tbaseBootCfg, sizeof(bootCfg_t));
flush_dcache_range((unsigned long)®isterFileL2, sizeof(registerFileL2));
// ************************************************************************************
// Set registers for tbase initialization entry
cpu_context_t *s_entry_context = &tbase_ctx->cpu_ctx;
gp_regs_t *s_entry_gpregs = get_gpregs_ctx(s_entry_context);
write_ctx_reg(s_entry_gpregs, CTX_GPREG_X1, 0);
write_ctx_reg(s_entry_gpregs, CTX_GPREG_X1, (int64_t)&tbaseBootCfg);
// SPSR for SMC handling (FIQ mode)
tbaseEntrySpsr = TBASE_ENTRY_SPSR;
DBG_PRINTF("tbase init SPSR 0x%x\n\r", read_ctx_reg(get_el3state_ctx(&tbase_ctx->cpu_ctx),
CTX_SPSR_EL3) );
DBG_PRINTF("tbase SMC SPSR %x\nr\r", tbaseEntrySpsr );
// ************************************************************************************
// Start tbase
tbase_synchronous_sp_entry(tbase_ctx);
tbase_ctx->state = TBASE_STATE_ON;
#if TBASE_PM_ENABLE
// Register power managemnt hooks with PSCI
psci_register_spd_pm_hook(&tbase_pm);
#endif
cm_el1_sysregs_context_restore(NON_SECURE);
cm_set_next_eret_context(NON_SECURE);
return 1;
}
static int
cpm_enet_start_xmit(struct sk_buff *skb, struct device *dev)
{
struct cpm_enet_private *cep = (struct cpm_enet_private *)dev->priv;
volatile cbd_t *bdp;
unsigned long flags;
/* Transmitter timeout, serious problems. */
if (dev->tbusy) {
int tickssofar = jiffies - dev->trans_start;
if (tickssofar < 200)
return 1;
printk("%s: transmit timed out.\n", dev->name);
cep->stats.tx_errors++;
#ifndef final_version
{
int i;
cbd_t *bdp;
printk(" Ring data dump: cur_tx %p%s cur_rx %p.\n",
cep->cur_tx, cep->tx_full ? " (full)" : "",
cep->cur_rx);
bdp = cep->tx_bd_base;
for (i = 0 ; i < TX_RING_SIZE; i++, bdp++)
printk("%04x %04x %08x\n",
bdp->cbd_sc,
bdp->cbd_datlen,
bdp->cbd_bufaddr);
bdp = cep->rx_bd_base;
for (i = 0 ; i < RX_RING_SIZE; i++, bdp++)
printk("%04x %04x %08x\n",
bdp->cbd_sc,
bdp->cbd_datlen,
bdp->cbd_bufaddr);
}
#endif
dev->tbusy=0;
dev->trans_start = jiffies;
return 0;
}
/* Block a timer-based transmit from overlapping. This could better be
done with atomic_swap(1, dev->tbusy), but set_bit() works as well. */
if (test_and_set_bit(0, (void*)&dev->tbusy) != 0) {
printk("%s: Transmitter access conflict.\n", dev->name);
return 1;
}
if (test_and_set_bit(0, (void*)&cep->lock) != 0) {
printk("%s: tx queue lock!.\n", dev->name);
/* don't clear dev->tbusy flag. */
return 1;
}
/* Fill in a Tx ring entry */
bdp = cep->cur_tx;
#ifndef final_version
if (bdp->cbd_sc & BD_ENET_TX_READY) {
/* Ooops. All transmit buffers are full. Bail out.
* This should not happen, since dev->tbusy should be set.
*/
printk("%s: tx queue full!.\n", dev->name);
cep->lock = 0;
return 1;
}
#endif
/* Clear all of the status flags.
*/
bdp->cbd_sc &= ~BD_ENET_TX_STATS;
/* If the frame is short, tell CPM to pad it.
*/
if (skb->len <= ETH_ZLEN)
bdp->cbd_sc |= BD_ENET_TX_PAD;
else
bdp->cbd_sc &= ~BD_ENET_TX_PAD;
/* Set buffer length and buffer pointer.
*/
bdp->cbd_datlen = skb->len;
bdp->cbd_bufaddr = __pa(skb->data);
/* Save skb pointer.
*/
cep->tx_skbuff[cep->skb_cur] = skb;
cep->stats.tx_bytes += skb->len;
cep->skb_cur = (cep->skb_cur+1) & TX_RING_MOD_MASK;
/* Push the data cache so the CPM does not get stale memory
* data.
*/
flush_dcache_range(skb->data, skb->data + skb->len);
/* Send it on its way. Tell CPM its ready, interrupt when done,
* its the last BD of the frame, and to put the CRC on the end.
*/
bdp->cbd_sc |= (BD_ENET_TX_READY | BD_ENET_TX_INTR | BD_ENET_TX_LAST | BD_ENET_TX_TC);
dev->trans_start = jiffies;
/* If this was the last BD in the ring, start at the beginning again.
*/
if (bdp->cbd_sc & BD_ENET_TX_WRAP)
bdp = cep->tx_bd_base;
else
bdp++;
save_flags(flags);
cli();
cep->lock = 0;
if (bdp->cbd_sc & BD_ENET_TX_READY)
cep->tx_full = 1;
else
dev->tbusy=0;
restore_flags(flags);
cep->cur_tx = (cbd_t *)bdp;
return 0;
}
static int nand_read_page(const struct nfc_config *conf, u32 offs,
void *dest, int len)
{
dma_addr_t dst = (dma_addr_t)dest;
int nsectors = len / conf->ecc_size;
u16 rand_seed;
u32 val;
int page;
page = offs / conf->page_size;
if (offs % conf->page_size || len % conf->ecc_size ||
len > conf->page_size || len < 0)
return -EINVAL;
/* clear ecc status */
writel(0, SUNXI_NFC_BASE + NFC_ECC_ST);
/* Choose correct seed */
rand_seed = random_seed[page % conf->nseeds];
writel((rand_seed << 16) | (conf->ecc_strength << 12) |
(conf->randomize ? NFC_ECC_RANDOM_EN : 0) |
(conf->ecc_size == 512 ? NFC_ECC_BLOCK_SIZE : 0) |
NFC_ECC_EN | NFC_ECC_PIPELINE | NFC_ECC_EXCEPTION,
SUNXI_NFC_BASE + NFC_ECC_CTL);
flush_dcache_range(dst, ALIGN(dst + conf->ecc_size, ARCH_DMA_MINALIGN));
/* SUNXI_DMA */
writel(0x0, SUNXI_DMA_BASE + SUNXI_DMA_CFG_REG0); /* clr dma cmd */
/* read from REG_IO_DATA */
writel(SUNXI_NFC_BASE + NFC_IO_DATA,
SUNXI_DMA_BASE + SUNXI_DMA_SRC_START_ADDR_REG0);
/* read to RAM */
writel(dst, SUNXI_DMA_BASE + SUNXI_DMA_DEST_START_ADDRR_REG0);
writel(SUNXI_DMA_DDMA_PARA_REG_SRC_WAIT_CYC |
SUNXI_DMA_DDMA_PARA_REG_SRC_BLK_SIZE,
SUNXI_DMA_BASE + SUNXI_DMA_DDMA_PARA_REG0);
writel(len, SUNXI_DMA_BASE + SUNXI_DMA_DDMA_BC_REG0);
writel(SUNXI_DMA_DDMA_CFG_REG_LOADING |
SUNXI_DMA_DDMA_CFG_REG_DMA_DEST_DATA_WIDTH_32 |
SUNXI_DMA_DDMA_CFG_REG_DDMA_DST_DRQ_TYPE_DRAM |
SUNXI_DMA_DDMA_CFG_REG_DMA_SRC_DATA_WIDTH_32 |
SUNXI_DMA_DDMA_CFG_REG_DMA_SRC_ADDR_MODE_IO |
SUNXI_DMA_DDMA_CFG_REG_DDMA_SRC_DRQ_TYPE_NFC,
SUNXI_DMA_BASE + SUNXI_DMA_CFG_REG0);
writel(nsectors, SUNXI_NFC_BASE + NFC_SECTOR_NUM);
writel(NFC_ST_DMA_INT_FLAG, SUNXI_NFC_BASE + NFC_ST);
writel(NFC_DATA_TRANS | NFC_PAGE_CMD | NFC_DATA_SWAP_METHOD,
SUNXI_NFC_BASE + NFC_CMD);
if (!check_value(SUNXI_NFC_BASE + NFC_ST, NFC_ST_DMA_INT_FLAG,
DEFAULT_TIMEOUT_US)) {
printf("Error while initializing dma interrupt\n");
return -EIO;
}
writel(NFC_ST_DMA_INT_FLAG, SUNXI_NFC_BASE + NFC_ST);
if (!check_value_negated(SUNXI_DMA_BASE + SUNXI_DMA_CFG_REG0,
SUNXI_DMA_DDMA_CFG_REG_LOADING,
DEFAULT_TIMEOUT_US)) {
printf("Error while waiting for dma transfer to finish\n");
return -EIO;
}
invalidate_dcache_range(dst,
ALIGN(dst + conf->ecc_size, ARCH_DMA_MINALIGN));
val = readl(SUNXI_NFC_BASE + NFC_ECC_ST);
/* ECC error detected. */
if (val & 0xffff)
return -EIO;
/*
* Return 1 if the page is empty.
* We consider the page as empty if the first ECC block is marked
* empty.
*/
return (val & 0x10000) ? 1 : 0;
}
static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
sigset_t *set, struct pt_regs *regs)
{
struct rt_sigframe __user *frame;
int err = 0;
int signal;
unsigned long address = 0;
#ifdef CONFIG_MMU
pmd_t *pmdp;
pte_t *ptep;
#endif
frame = get_sigframe(ka, regs, sizeof(*frame));
if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
goto give_sigsegv;
signal = current_thread_info()->exec_domain
&& current_thread_info()->exec_domain->signal_invmap
&& sig < 32
? current_thread_info()->exec_domain->signal_invmap[sig]
: sig;
if (info)
err |= copy_siginfo_to_user(&frame->info, info);
/* Create the ucontext. */
err |= __put_user(0, &frame->uc.uc_flags);
err |= __put_user(NULL, &frame->uc.uc_link);
err |= __save_altstack(&frame->uc.uc_stack, regs->r1);
err |= setup_sigcontext(&frame->uc.uc_mcontext,
regs, set->sig[0]);
err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
/* Set up to return from userspace. If provided, use a stub
already in userspace. */
/* minus 8 is offset to cater for "rtsd r15,8" */
/* addi r12, r0, __NR_sigreturn */
err |= __put_user(0x31800000 | __NR_rt_sigreturn ,
frame->tramp + 0);
/* brki r14, 0x8 */
err |= __put_user(0xb9cc0008, frame->tramp + 1);
/* Return from sighandler will jump to the tramp.
Negative 8 offset because return is rtsd r15, 8 */
regs->r15 = ((unsigned long)frame->tramp)-8;
address = ((unsigned long)frame->tramp);
#ifdef CONFIG_MMU
pmdp = pmd_offset(pud_offset(
pgd_offset(current->mm, address),
address), address);
preempt_disable();
ptep = pte_offset_map(pmdp, address);
if (pte_present(*ptep)) {
address = (unsigned long) page_address(pte_page(*ptep));
/* MS: I need add offset in page */
address += ((unsigned long)frame->tramp) & ~PAGE_MASK;
/* MS address is virtual */
address = __virt_to_phys(address);
invalidate_icache_range(address, address + 8);
flush_dcache_range(address, address + 8);
}
pte_unmap(ptep);
preempt_enable();
#else
flush_icache_range(address, address + 8);
flush_dcache_range(address, address + 8);
#endif
if (err)
goto give_sigsegv;
/* Set up registers for signal handler */
regs->r1 = (unsigned long) frame;
/* Signal handler args: */
regs->r5 = signal; /* arg 0: signum */
regs->r6 = (unsigned long) &frame->info; /* arg 1: siginfo */
regs->r7 = (unsigned long) &frame->uc; /* arg2: ucontext */
/* Offset to handle microblaze rtid r14, 0 */
regs->pc = (unsigned long)ka->sa.sa_handler;
set_fs(USER_DS);
#ifdef DEBUG_SIG
pr_info("SIG deliver (%s:%d): sp=%p pc=%08lx\n",
current->comm, current->pid, frame, regs->pc);
#endif
return 0;
give_sigsegv:
force_sigsegv(sig, current);
return -EFAULT;
}
static int mv_ata_exec_ata_cmd(int port, struct sata_fis_h2d *cfis,
u8 *buffer, u32 len, u32 iswrite)
{
struct mv_priv *priv = (struct mv_priv *)sata_dev_desc[port].priv;
struct crqb *req;
int slot;
u32 start;
if (len >= 64 * 1024) {
printf("We only support <64K transfers for now\n");
return -1;
}
/* Initialize request */
slot = get_reqip(port);
memset(&priv->request[slot], 0, sizeof(struct crqb));
req = &priv->request[slot];
req->dtb_low = (u32)buffer;
/* Dont use PRDs */
req->control_flags = CRQB_CNTRLFLAGS_PRDMODE;
req->control_flags |= iswrite ? 0 : CRQB_CNTRLFLAGS_DIR;
req->control_flags |=
((cfis->pm_port_c << CRQB_CNTRLFLAGS_PMPORTSHIFT)
& CRQB_CNTRLFLAGS_PMPORTMASK);
req->drb_count = len;
req->ata_cmd_feat = (cfis->command << CRQB_CMDFEAT_CMDSHIFT) &
CRQB_CMDFEAT_CMDMASK;
req->ata_cmd_feat |= (cfis->features << CRQB_CMDFEAT_FEATSHIFT) &
CRQB_CMDFEAT_FEATMASK;
req->ata_addr = (cfis->lba_low << CRQB_ADDR_LBA_LOWSHIFT) &
CRQB_ADDR_LBA_LOWMASK;
req->ata_addr |= (cfis->lba_mid << CRQB_ADDR_LBA_MIDSHIFT) &
CRQB_ADDR_LBA_MIDMASK;
req->ata_addr |= (cfis->lba_high << CRQB_ADDR_LBA_HIGHSHIFT) &
CRQB_ADDR_LBA_HIGHMASK;
req->ata_addr |= (cfis->device << CRQB_ADDR_DEVICE_SHIFT) &
CRQB_ADDR_DEVICE_MASK;
req->ata_addr_exp = (cfis->lba_low_exp << CRQB_ADDR_LBA_LOW_EXP_SHIFT) &
CRQB_ADDR_LBA_LOW_EXP_MASK;
req->ata_addr_exp |=
(cfis->lba_mid_exp << CRQB_ADDR_LBA_MID_EXP_SHIFT) &
CRQB_ADDR_LBA_MID_EXP_MASK;
req->ata_addr_exp |=
(cfis->lba_high_exp << CRQB_ADDR_LBA_HIGH_EXP_SHIFT) &
CRQB_ADDR_LBA_HIGH_EXP_MASK;
req->ata_addr_exp |=
(cfis->features_exp << CRQB_ADDR_FEATURE_EXP_SHIFT) &
CRQB_ADDR_FEATURE_EXP_MASK;
req->ata_sect_count =
(cfis->sector_count << CRQB_SECTCOUNT_COUNT_SHIFT) &
CRQB_SECTCOUNT_COUNT_MASK;
req->ata_sect_count |=
(cfis->sector_count_exp << CRQB_SECTCOUNT_COUNT_EXP_SHIFT) &
CRQB_SECTCOUNT_COUNT_EXP_MASK;
/* Flush data */
start = (u32)req & ~(ARCH_DMA_MINALIGN - 1);
flush_dcache_range(start,
start + ALIGN(sizeof(*req), ARCH_DMA_MINALIGN));
/* Trigger operation */
slot = get_next_reqip(port);
set_reqip(port, slot);
/* Wait for completion */
if (wait_dma_completion(port, slot, 10000)) {
printf("ATA operation timed out\n");
return -1;
}
process_responses(port);
/* Invalidate data on read */
if (buffer && len) {
start = (u32)buffer & ~(ARCH_DMA_MINALIGN - 1);
invalidate_dcache_range(start,
start + ALIGN(len, ARCH_DMA_MINALIGN));
}
return len;
}
/*
* This mmu table looks as below
* Level 0 table contains two entries to 512GB sizes. One is Level1 Table 0
* and other Level1 Table1.
* Level1 Table0 contains entries for each 1GB from 0 to 511GB.
* Level1 Table1 contains entries for each 1GB from 512GB to 1TB.
* Level2 Table0, Level2 Table1, Level2 Table2 and Level2 Table3 contains
* entries for each 2MB starting from 0GB, 1GB, 2GB and 3GB respectively.
*/
static void zynqmp_mmu_setup(void)
{
int el;
u32 index_attr;
u64 i, section_l1t0, section_l1t1;
u64 section_l2t0, section_l2t1, section_l2t2, section_l2t3;
u64 *level0_table = (u64 *)gd->arch.tlb_addr;
u64 *level1_table_0 = (u64 *)(gd->arch.tlb_addr + TLB_TABLE_SIZE);
u64 *level1_table_1 = (u64 *)(gd->arch.tlb_addr + (2 * TLB_TABLE_SIZE));
u64 *level2_table_0 = (u64 *)(gd->arch.tlb_addr + (3 * TLB_TABLE_SIZE));
u64 *level2_table_1 = (u64 *)(gd->arch.tlb_addr + (4 * TLB_TABLE_SIZE));
u64 *level2_table_2 = (u64 *)(gd->arch.tlb_addr + (5 * TLB_TABLE_SIZE));
u64 *level2_table_3 = (u64 *)(gd->arch.tlb_addr + (6 * TLB_TABLE_SIZE));
level0_table[0] =
(u64)level1_table_0 | PMD_TYPE_TABLE;
level0_table[1] =
(u64)level1_table_1 | PMD_TYPE_TABLE;
/*
* set level 1 table 0, covering 0 to 512GB
* set level 1 table 1, covering 512GB to 1TB
*/
section_l1t0 = 0;
section_l1t1 = BLOCK_SIZE_L0;
index_attr = 0;
for (i = 0; i < 512; i++) {
level1_table_0[i] = section_l1t0;
level1_table_0[i] |= attr_tbll1t0[index_attr].attr;
attr_tbll1t0[index_attr].num--;
if (attr_tbll1t0[index_attr].num == 0)
index_attr++;
level1_table_1[i] = section_l1t1;
level1_table_1[i] |= DEVICE_ATTR;
section_l1t0 += BLOCK_SIZE_L1;
section_l1t1 += BLOCK_SIZE_L1;
}
level1_table_0[0] =
(u64)level2_table_0 | PMD_TYPE_TABLE;
level1_table_0[1] =
(u64)level2_table_1 | PMD_TYPE_TABLE;
level1_table_0[2] =
(u64)level2_table_2 | PMD_TYPE_TABLE;
level1_table_0[3] =
(u64)level2_table_3 | PMD_TYPE_TABLE;
section_l2t0 = 0;
section_l2t1 = section_l2t0 + BLOCK_SIZE_L1; /* 1GB */
section_l2t2 = section_l2t1 + BLOCK_SIZE_L1; /* 2GB */
section_l2t3 = section_l2t2 + BLOCK_SIZE_L1; /* 3GB */
index_attr = 0;
for (i = 0; i < 512; i++) {
level2_table_0[i] = section_l2t0 | MEMORY_ATTR;
level2_table_1[i] = section_l2t1 | MEMORY_ATTR;
level2_table_2[i] = section_l2t2 | DEVICE_ATTR;
level2_table_3[i] = section_l2t3 |
attr_tbll2t3[index_attr].attr;
attr_tbll2t3[index_attr].num--;
if (attr_tbll2t3[index_attr].num == 0)
index_attr++;
section_l2t0 += BLOCK_SIZE_L2;
section_l2t1 += BLOCK_SIZE_L2;
section_l2t2 += BLOCK_SIZE_L2;
section_l2t3 += BLOCK_SIZE_L2;
}
/* flush new MMU table */
flush_dcache_range(gd->arch.tlb_addr,
gd->arch.tlb_addr + gd->arch.tlb_size);
/* point TTBR to the new table */
el = current_el();
set_ttbr_tcr_mair(el, gd->arch.tlb_addr,
ZYNQMP_TCR, MEMORY_ATTRIBUTES);
set_sctlr(get_sctlr() | CR_M);
}
/*
* Flush range from all levels of d-cache/unified-cache.
* Affects the range [start, start + size - 1].
*/
__weak void flush_cache(unsigned long start, unsigned long size)
{
flush_dcache_range(start, start + size);
}
static int ldpaa_eth_tx(struct eth_device *net_dev, void *buf, int len)
{
struct ldpaa_eth_priv *priv = (struct ldpaa_eth_priv *)net_dev->priv;
struct dpaa_fd fd;
u64 buffer_start;
int data_offset, err;
u32 timeo = (CONFIG_SYS_HZ * 10) / 1000;
u32 time_start;
struct qbman_swp *swp = dflt_dpio->sw_portal;
struct qbman_eq_desc ed;
struct qbman_release_desc releasedesc;
/* Setup the FD fields */
memset(&fd, 0, sizeof(fd));
data_offset = priv->tx_data_offset;
do {
err = qbman_swp_acquire(dflt_dpio->sw_portal,
dflt_dpbp->dpbp_attr.bpid,
&buffer_start, 1);
} while (err == -EBUSY);
if (err < 0) {
printf("qbman_swp_acquire() failed\n");
return -ENOMEM;
}
debug("TX data: malloc buffer start=0x%p\n", (u64 *)buffer_start);
memcpy(((uint8_t *)(buffer_start) + data_offset), buf, len);
flush_dcache_range(buffer_start, buffer_start +
LDPAA_ETH_RX_BUFFER_SIZE);
ldpaa_fd_set_addr(&fd, (u64)buffer_start);
ldpaa_fd_set_offset(&fd, (uint16_t)(data_offset));
ldpaa_fd_set_bpid(&fd, dflt_dpbp->dpbp_attr.bpid);
ldpaa_fd_set_len(&fd, len);
fd.simple.ctrl = LDPAA_FD_CTRL_ASAL | LDPAA_FD_CTRL_PTA |
LDPAA_FD_CTRL_PTV1;
qbman_eq_desc_clear(&ed);
qbman_eq_desc_set_no_orp(&ed, 0);
qbman_eq_desc_set_qd(&ed, priv->tx_qdid, priv->tx_flow_id, 0);
time_start = get_timer(0);
while (get_timer(time_start) < timeo) {
err = qbman_swp_enqueue(swp, &ed,
(const struct qbman_fd *)(&fd));
if (err != -EBUSY)
break;
}
if (err < 0) {
printf("error enqueueing Tx frame\n");
goto error;
}
return err;
error:
qbman_release_desc_clear(&releasedesc);
qbman_release_desc_set_bpid(&releasedesc, dflt_dpbp->dpbp_attr.bpid);
time_start = get_timer(0);
do {
/* Release buffer into the QBMAN */
err = qbman_swp_release(swp, &releasedesc, &buffer_start, 1);
} while (get_timer(time_start) < timeo && err == -EBUSY);
if (err == -EBUSY)
printf("TX data: QBMAN buffer release fails\n");
return err;
}
static void mxs_nand_flush_data_buf(struct mxs_nand_info *info)
{
uint32_t addr = (uint32_t)info->data_buf;
flush_dcache_range(addr, addr + info->data_buf_size);
}
/*******************************************************************************
* Generic function to load an image into the trusted RAM,
* given a name, extents of free memory & whether the image should be loaded at
* the bottom or top of the free memory. It updates the memory layout if the
* load is successful. It also updates the image information and the entry point
* information in the params passed
******************************************************************************/
int load_image(meminfo_t *mem_layout,
const char *image_name,
unsigned int load_type,
unsigned long fixed_addr,
image_info_t *image_data,
entry_point_info_t *entry_point_info)
{
uintptr_t dev_handle;
uintptr_t image_handle;
uintptr_t image_spec;
unsigned long temp_image_base = 0;
unsigned long image_base = 0;
long offset = 0;
size_t image_size = 0;
size_t bytes_read = 0;
int io_result = IO_FAIL;
assert(mem_layout != NULL);
assert(image_name != NULL);
assert(image_data->h.version >= VERSION_1);
/* Obtain a reference to the image by querying the platform layer */
io_result = plat_get_image_source(image_name, &dev_handle, &image_spec);
if (io_result != IO_SUCCESS) {
WARN("Failed to obtain reference to image '%s' (%i)\n",
image_name, io_result);
return io_result;
}
/* Attempt to access the image */
io_result = io_open(dev_handle, image_spec, &image_handle);
if (io_result != IO_SUCCESS) {
WARN("Failed to access image '%s' (%i)\n",
image_name, io_result);
return io_result;
}
/* Find the size of the image */
io_result = io_size(image_handle, &image_size);
if ((io_result != IO_SUCCESS) || (image_size == 0)) {
WARN("Failed to determine the size of the image '%s' file (%i)\n",
image_name, io_result);
goto exit;
}
/* See if we have enough space */
if (image_size > mem_layout->free_size) {
WARN("Cannot load '%s' file: Not enough space.\n",
image_name);
dump_load_info(0, image_size, mem_layout);
goto exit;
}
switch (load_type) {
case TOP_LOAD:
/* Load the image in the top of free memory */
temp_image_base = mem_layout->free_base + mem_layout->free_size;
temp_image_base -= image_size;
/* Page align base address and check whether the image still fits */
image_base = page_align(temp_image_base, DOWN);
assert(image_base <= temp_image_base);
if (image_base < mem_layout->free_base) {
WARN("Cannot load '%s' file: Not enough space.\n",
image_name);
dump_load_info(image_base, image_size, mem_layout);
io_result = -ENOMEM;
goto exit;
}
/* Calculate the amount of extra memory used due to alignment */
offset = temp_image_base - image_base;
break;
case BOT_LOAD:
/* Load the BL2 image in the bottom of free memory */
temp_image_base = mem_layout->free_base;
image_base = page_align(temp_image_base, UP);
assert(image_base >= temp_image_base);
/* Page align base address and check whether the image still fits */
if (image_base + image_size >
mem_layout->free_base + mem_layout->free_size) {
WARN("Cannot load '%s' file: Not enough space.\n",
image_name);
dump_load_info(image_base, image_size, mem_layout);
io_result = -ENOMEM;
goto exit;
}
/* Calculate the amount of extra memory used due to alignment */
offset = image_base - temp_image_base;
break;
default:
assert(0);
}
/*
* Some images must be loaded at a fixed address, not a dynamic one.
*
* This has been implemented as a hack on top of the existing dynamic
* loading mechanism, for the time being. If the 'fixed_addr' function
* argument is different from zero, then it will force the load address.
* So we still have this principle of top/bottom loading but the code
* determining the load address is bypassed and the load address is
* forced to the fixed one.
*
* This can result in quite a lot of wasted space because we still use
* 1 sole meminfo structure to represent the extents of free memory,
* where we should use some sort of linked list.
*
* E.g. we want to load BL2 at address 0x04020000, the resulting memory
* layout should look as follows:
* ------------ 0x04040000
* | | <- Free space (1)
* |----------|
* | BL2 |
* |----------| 0x04020000
* | | <- Free space (2)
* |----------|
* | BL1 |
* ------------ 0x04000000
*
* But in the current hacky implementation, we'll need to specify
* whether BL2 is loaded at the top or bottom of the free memory.
* E.g. if BL2 is considered as top-loaded, the meminfo structure
* will give the following view of the memory, hiding the chunk of
* free memory above BL2:
* ------------ 0x04040000
* | |
* | |
* | BL2 |
* |----------| 0x04020000
* | | <- Free space (2)
* |----------|
* | BL1 |
* ------------ 0x04000000
*/
if (fixed_addr != 0) {
/* Load the image at the given address. */
image_base = fixed_addr;
/* Check whether the image fits. */
if ((image_base < mem_layout->free_base) ||
(image_base + image_size >
mem_layout->free_base + mem_layout->free_size)) {
WARN("Cannot load '%s' file: Not enough space.\n",
image_name);
dump_load_info(image_base, image_size, mem_layout);
io_result = -ENOMEM;
goto exit;
}
/* Check whether the fixed load address is page-aligned. */
if (!is_page_aligned(image_base)) {
WARN("Cannot load '%s' file at unaligned address 0x%lx\n",
image_name, fixed_addr);
io_result = -ENOMEM;
goto exit;
}
/*
* Calculate the amount of extra memory used due to fixed
* loading.
*/
if (load_type == TOP_LOAD) {
unsigned long max_addr, space_used;
/*
* ------------ max_addr
* | /wasted/ | | offset
* |..........|..............................
* | image | | image_flen
* |----------| fixed_addr
* | |
* | |
* ------------ total_base
*/
max_addr = mem_layout->total_base + mem_layout->total_size;
/*
* Compute the amount of memory used by the image.
* Corresponds to all space above the image load
* address.
*/
space_used = max_addr - fixed_addr;
/*
* Calculate the amount of wasted memory within the
* amount of memory used by the image.
*/
offset = space_used - image_size;
} else /* BOT_LOAD */
/*
* ------------
* | |
* | |
* |----------|
* | image |
* |..........| fixed_addr
* | /wasted/ | | offset
* ------------ total_base
*/
offset = fixed_addr - mem_layout->total_base;
}
/* We have enough space so load the image now */
/* TODO: Consider whether to try to recover/retry a partially successful read */
io_result = io_read(image_handle, image_base, image_size, &bytes_read);
if ((io_result != IO_SUCCESS) || (bytes_read < image_size)) {
WARN("Failed to load '%s' file (%i)\n", image_name, io_result);
goto exit;
}
image_data->image_base = image_base;
image_data->image_size = image_size;
entry_point_info->pc = image_base;
/*
* File has been successfully loaded. Update the free memory
* data structure & flush the contents of the TZRAM so that
* the next EL can see it.
*/
/* Update the memory contents */
flush_dcache_range(image_base, image_size);
mem_layout->free_size -= image_size + offset;
/* Update the base of free memory since its moved up */
if (load_type == BOT_LOAD)
mem_layout->free_base += offset + image_size;
exit:
io_close(image_handle);
/* Ignore improbable/unrecoverable error in 'close' */
/* TODO: Consider maintaining open device connection from this bootloader stage */
io_dev_close(dev_handle);
/* Ignore improbable/unrecoverable error in 'dev_close' */
return io_result;
}
/* -1 --- error, can't enqueue -- no space available */
static int jr_enqueue(uint32_t *desc_addr,
void (*callback)(uint32_t status, void *arg),
void *arg)
{
struct jr_regs *regs = (struct jr_regs *)CONFIG_SYS_FSL_JR0_ADDR;
int head = jr.head;
uint32_t desc_word;
int length = desc_len(desc_addr);
int i;
#ifdef CONFIG_PHYS_64BIT
uint32_t *addr_hi, *addr_lo;
#endif
/* The descriptor must be submitted to SEC block as per endianness
* of the SEC Block.
* So, if the endianness of Core and SEC block is different, each word
* of the descriptor will be byte-swapped.
*/
for (i = 0; i < length; i++) {
desc_word = desc_addr[i];
sec_out32((uint32_t *)&desc_addr[i], desc_word);
}
phys_addr_t desc_phys_addr = virt_to_phys(desc_addr);
if (sec_in32(®s->irsa) == 0 ||
CIRC_SPACE(jr.head, jr.tail, jr.size) <= 0)
return -1;
jr.info[head].desc_phys_addr = desc_phys_addr;
jr.info[head].callback = (void *)callback;
jr.info[head].arg = arg;
jr.info[head].op_done = 0;
unsigned long start = (unsigned long)&jr.info[head] &
~(ARCH_DMA_MINALIGN - 1);
unsigned long end = ALIGN((unsigned long)&jr.info[head] +
sizeof(struct jr_info), ARCH_DMA_MINALIGN);
flush_dcache_range(start, end);
#ifdef CONFIG_PHYS_64BIT
/* Write the 64 bit Descriptor address on Input Ring.
* The 32 bit hign and low part of the address will
* depend on endianness of SEC block.
*/
#ifdef CONFIG_SYS_FSL_SEC_LE
addr_lo = (uint32_t *)(&jr.input_ring[head]);
addr_hi = (uint32_t *)(&jr.input_ring[head]) + 1;
#elif defined(CONFIG_SYS_FSL_SEC_BE)
addr_hi = (uint32_t *)(&jr.input_ring[head]);
addr_lo = (uint32_t *)(&jr.input_ring[head]) + 1;
#endif /* ifdef CONFIG_SYS_FSL_SEC_LE */
sec_out32(addr_hi, (uint32_t)(desc_phys_addr >> 32));
sec_out32(addr_lo, (uint32_t)(desc_phys_addr));
#else
/* Write the 32 bit Descriptor address on Input Ring. */
sec_out32(&jr.input_ring[head], desc_phys_addr);
#endif /* ifdef CONFIG_PHYS_64BIT */
start = (unsigned long)&jr.input_ring[head] & ~(ARCH_DMA_MINALIGN - 1);
end = ALIGN((unsigned long)&jr.input_ring[head] +
sizeof(dma_addr_t), ARCH_DMA_MINALIGN);
flush_dcache_range(start, end);
jr.head = (head + 1) & (jr.size - 1);
/* Invalidate output ring */
start = (unsigned long)jr.output_ring &
~(ARCH_DMA_MINALIGN - 1);
end = ALIGN((unsigned long)jr.output_ring + jr.op_size,
ARCH_DMA_MINALIGN);
invalidate_dcache_range(start, end);
sec_out32(®s->irja, 1);
return 0;
}
static void mxc_udc_queue_update(u8 epnum,
u8 *data, u32 len, u32 tx)
{
struct mxc_ep_t *ep;
struct ep_queue_item *tqi, *head, *last;
int send = 0;
int in;
head = last = NULL;
in = ep_is_in(epnum, tx);
ep = mxc_udc.mxc_ep + (epnum * 2 + in);
DBG("epnum = %d, in = %d\n", epnum, in);
do {
tqi = ep->ep_dtd[ep->index];
DBG("%s, index = %d, tqi = %p\n", __func__, ep->index, tqi);
while (mxc_tqi_is_busy(tqi))
;
mxc_tqi_init_page(tqi);
DBG("%s, line = %d, len = %d\n", __func__, __LINE__, len);
inc_index(ep->index);
send = min(len, ep->max_pkt_size);
if (data) {
memcpy((void *)tqi->page_vir, (void *)data, send);
_dump_buf((u8 *)(tqi->page_vir), send);
}
flush_dcache_range((unsigned long)(tqi->page_vir),
CACHE_ALIGNED_END(tqi->page_vir, ep->max_pkt_size));
if (!head)
last = head = tqi;
else {
last->next_item_ptr = virt_to_phys(tqi);
last->next_item_vir = tqi;
last = tqi;
}
if (!tx)
tqi->reserved[0] = send;
/* we set IOC for every dtd */
tqi->info = ((send << 16) | (1 << 15) | (1 << 7));
data += send;
len -= send;
flush_dcache_range((unsigned long)tqi,
CACHE_ALIGNED_END(tqi, sizeof(struct ep_queue_item)));
} while (len);
last->next_item_ptr = 0x1; /* end */
flush_dcache_range((unsigned long)last,
CACHE_ALIGNED_END(last, sizeof(struct ep_queue_item)));
if (ep->tail) {
ep->tail->next_item_ptr = virt_to_phys(head);
ep->tail->next_item_vir = head;
flush_dcache_range((unsigned long)(ep->tail),
CACHE_ALIGNED_END(ep->tail, sizeof(struct ep_queue_item)));
if (mxc_ep_xfer_is_working(ep, in)) {
DBG("ep is working\n");
goto out;
}
}
mxc_update_qh(ep, head, in);
out:
ep->tail = last;
}
static int zynqmp_qspi_genfifo_fill_rx(struct zynqmp_qspi_priv *priv)
{
u32 gen_fifo_cmd;
u32 *buf;
u32 addr;
u32 size, len;
u32 timeout = 10000000;
u32 actuallen = priv->len;
struct zynqmp_qspi_regs *regs = priv->regs;
struct zynqmp_qspi_dma_regs *dma_regs = priv->dma_regs;
gen_fifo_cmd = zynqmp_qspi_bus_select(priv);
gen_fifo_cmd |= ZYNQMP_QSPI_GFIFO_RX |
ZYNQMP_QSPI_GFIFO_DATA_XFR_MASK;
if (last_cmd == QUAD_OUT_READ_CMD)
gen_fifo_cmd |= ZYNQMP_QSPI_SPI_MODE_QSPI;
else
gen_fifo_cmd |= ZYNQMP_QSPI_SPI_MODE_SPI;
if (priv->stripe)
gen_fifo_cmd |= ZYNQMP_QSPI_GFIFO_STRIPE_MASK;
if (!((u32)priv->rx_buf & 0x3) && !(actuallen % 4)) {
buf = (u32 *)priv->rx_buf;
} else {
ALLOC_CACHE_ALIGN_BUFFER(u8, tmp, roundup(priv->len, 4));
buf = (u32 *)tmp;
}
writel((u32)buf, &dma_regs->dmadst);
writel(roundup(priv->len, 4), &dma_regs->dmasize);
writel(ZYNQMP_QSPI_DMA_DST_I_STS_MASK, &dma_regs->dmaier);
addr = (u32)buf;
size = roundup(priv->len, ARCH_DMA_MINALIGN);
flush_dcache_range(addr, addr+size);
while (priv->len) {
len = zynqmp_qspi_calc_exp(priv, &gen_fifo_cmd);
if (!(gen_fifo_cmd & ZYNQMP_QSPI_GFIFO_EXP_MASK) &&
(len % 4)) {
gen_fifo_cmd &= ~(0xFF);
gen_fifo_cmd |= (len/4 + 1) * 4;
}
writel(gen_fifo_cmd, ®s->genfifo);
debug("GFIFO_CMD_RX:0x%x\n", gen_fifo_cmd);
}
while (timeout) {
if (readl(&dma_regs->dmaisr) &
ZYNQMP_QSPI_DMA_DST_I_STS_DONE) {
writel(ZYNQMP_QSPI_DMA_DST_I_STS_DONE,
&dma_regs->dmaisr);
break;
}
timeout--;
}
debug("buf:0x%lx, rxbuf:0x%lx, *buf:0x%x len: 0x%x\n",
(unsigned long)buf, (unsigned long)priv->rx_buf, *buf,
actuallen);
if (!timeout) {
debug("DMA Timeout:0x%x\n", readl(&dma_regs->dmaisr));
return -1;
}
if (buf != priv->rx_buf)
memcpy(priv->rx_buf, buf, actuallen);
return 0;
}
static int
ehci_submit_async(struct usb_device *dev, unsigned long pipe, void *buffer,
int length, struct devrequest *req)
{
static struct QH qh __attribute__((aligned(32)));
static struct qTD qtd[3] __attribute__((aligned (32)));
int qtd_counter = 0;
volatile struct qTD *vtd;
unsigned long ts;
uint32_t *tdp;
uint32_t endpt, token, usbsts;
uint32_t c, toggle;
uint32_t cmd;
int timeout;
int ret = 0;
debug("dev=%p, pipe=%lx, buffer=%p, length=%d, req=%p\n", dev, pipe,
buffer, length, req);
if (req != NULL)
debug("req=%u (%#x), type=%u (%#x), value=%u (%#x), index=%u\n",
req->request, req->request,
req->requesttype, req->requesttype,
le16_to_cpu(req->value), le16_to_cpu(req->value),
le16_to_cpu(req->index));
memset(&qh, 0, sizeof(struct QH));
memset(qtd, 0, sizeof(qtd));
toggle = usb_gettoggle(dev, usb_pipeendpoint(pipe), usb_pipeout(pipe));
/*
* Setup QH (3.6 in ehci-r10.pdf)
*
* qh_link ................. 03-00 H
* qh_endpt1 ............... 07-04 H
* qh_endpt2 ............... 0B-08 H
* - qh_curtd
* qh_overlay.qt_next ...... 13-10 H
* - qh_overlay.qt_altnext
*/
qh.qh_link = cpu_to_hc32((uint32_t)PHYSICAL_ADDR(&qh_list) | QH_LINK_TYPE_QH);
c = (usb_pipespeed(pipe) != USB_SPEED_HIGH &&
usb_pipeendpoint(pipe) == 0) ? 1 : 0;
endpt = (8 << 28) |
(c << 27) |
(usb_maxpacket(dev, pipe) << 16) |
(0 << 15) |
(1 << 14) |
(usb_pipespeed(pipe) << 12) |
(usb_pipeendpoint(pipe) << 8) |
(0 << 7) | (usb_pipedevice(pipe) << 0);
qh.qh_endpt1 = cpu_to_hc32(endpt);
endpt = (1 << 30) |
(dev->portnr << 23) |
(dev->parent->devnum << 16) | (0 << 8) | (0 << 0);
qh.qh_endpt2 = cpu_to_hc32(endpt);
qh.qh_overlay.qt_next = cpu_to_hc32(QT_NEXT_TERMINATE);
tdp = &qh.qh_overlay.qt_next;
if (req != NULL) {
/*
* Setup request qTD (3.5 in ehci-r10.pdf)
*
* qt_next ................ 03-00 H
* qt_altnext ............. 07-04 H
* qt_token ............... 0B-08 H
*
* [ buffer, buffer_hi ] loaded with "req".
*/
qtd[qtd_counter].qt_next = cpu_to_hc32(QT_NEXT_TERMINATE);
qtd[qtd_counter].qt_altnext = cpu_to_hc32(QT_NEXT_TERMINATE);
token = (0 << 31) |
(sizeof(*req) << 16) |
(0 << 15) | (0 << 12) | (3 << 10) | (2 << 8) | (0x80 << 0);
qtd[qtd_counter].qt_token = cpu_to_hc32(token);
if (ehci_td_buffer(&qtd[qtd_counter], req, sizeof(*req)) != 0) {
debug("unable construct SETUP td\n");
goto fail;
}
/* Update previous qTD! */
*tdp = cpu_to_hc32((uint32_t)PHYSICAL_ADDR(&qtd[qtd_counter]));
tdp = &qtd[qtd_counter++].qt_next;
toggle = 1;
}
if (length > 0 || req == NULL) {
/*
* Setup request qTD (3.5 in ehci-r10.pdf)
*
* qt_next ................ 03-00 H
* qt_altnext ............. 07-04 H
* qt_token ............... 0B-08 H
*
* [ buffer, buffer_hi ] loaded with "buffer".
*/
qtd[qtd_counter].qt_next = cpu_to_hc32(QT_NEXT_TERMINATE);
qtd[qtd_counter].qt_altnext = cpu_to_hc32(QT_NEXT_TERMINATE);
token = (toggle << 31) |
(length << 16) |
((req == NULL ? 1 : 0) << 15) |
(0 << 12) |
(3 << 10) |
((usb_pipein(pipe) ? 1 : 0) << 8) | (0x80 << 0);
qtd[qtd_counter].qt_token = cpu_to_hc32(token);
if (ehci_td_buffer(&qtd[qtd_counter], buffer, length) != 0) {
debug("unable construct DATA td\n");
goto fail;
}
/* Update previous qTD! */
*tdp = cpu_to_hc32((uint32_t)PHYSICAL_ADDR(&qtd[qtd_counter]));
tdp = &qtd[qtd_counter++].qt_next;
}
if (req != NULL) {
/*
* Setup request qTD (3.5 in ehci-r10.pdf)
*
* qt_next ................ 03-00 H
* qt_altnext ............. 07-04 H
* qt_token ............... 0B-08 H
*/
qtd[qtd_counter].qt_next = cpu_to_hc32(QT_NEXT_TERMINATE);
qtd[qtd_counter].qt_altnext = cpu_to_hc32(QT_NEXT_TERMINATE);
token = (toggle << 31) |
(0 << 16) |
(1 << 15) |
(0 << 12) |
(3 << 10) |
((usb_pipein(pipe) ? 0 : 1) << 8) | (0x80 << 0);
qtd[qtd_counter].qt_token = cpu_to_hc32(token);
/* Update previous qTD! */
*tdp = cpu_to_hc32((uint32_t)PHYSICAL_ADDR(&qtd[qtd_counter]));
tdp = &qtd[qtd_counter++].qt_next;
}
qh_list.qh_link = cpu_to_hc32((uint32_t) PHYSICAL_ADDR(&qh) | QH_LINK_TYPE_QH);
/* Flush dcache */
flush_dcache_range((uint32_t)&qh_list,
(uint32_t)&qh_list + sizeof(struct QH));
flush_dcache_range((uint32_t)&qh, (uint32_t)&qh + sizeof(struct QH));
flush_dcache_range((uint32_t)qtd, (uint32_t)qtd + sizeof(qtd));
usbsts = ehci_readl(&hcor->or_usbsts);
ehci_writel(&hcor->or_usbsts, (usbsts & 0x3f));
/* Enable async. schedule. */
cmd = ehci_readl(&hcor->or_usbcmd);
cmd |= CMD_ASE;
ehci_writel(&hcor->or_usbcmd, cmd);
ret = handshake((uint32_t *)&hcor->or_usbsts, STD_ASS, STD_ASS,
100 * 1000);
if (ret < 0) {
printf("EHCI fail timeout STD_ASS set\n");
goto fail;
}
/* Wait for TDs to be processed. */
ts = get_timer(0);
vtd = &qtd[qtd_counter - 1];
timeout = USB_TIMEOUT_MS(pipe);
do {
/* Invalidate dcache */
invalidate_dcache_range((uint32_t)&qh_list,
(uint32_t)&qh_list + sizeof(struct QH));
invalidate_dcache_range((uint32_t)&qh,
(uint32_t)&qh + sizeof(struct QH));
invalidate_dcache_range((uint32_t)qtd,
(uint32_t)qtd + sizeof(qtd));
token = hc32_to_cpu(vtd->qt_token);
if (!(token & 0x80))
break;
WATCHDOG_RESET();
} while (get_timer(ts) < timeout);
/* Invalidate the memory area occupied by buffer */
if (buffer!=NULL) {
invalidate_dcache_range(((uint32_t)buffer & ~31),
((uint32_t)buffer & ~31) + roundup(length, 32));
}
/* Check that the TD processing happened */
if (token & 0x80) {
printf("EHCI timed out on TD - token=%#x\n", token);
}
/* Disable async schedule. */
cmd = ehci_readl(&hcor->or_usbcmd);
cmd &= ~CMD_ASE;
ehci_writel(&hcor->or_usbcmd, cmd);
ret = handshake((uint32_t *)&hcor->or_usbsts, STD_ASS, 0,
100 * 1000);
if (ret < 0) {
printf("EHCI fail timeout STD_ASS reset\n");
goto fail;
}
qh_list.qh_link = cpu_to_hc32((uint32_t)PHYSICAL_ADDR(&qh_list) | QH_LINK_TYPE_QH);
token = hc32_to_cpu(qh.qh_overlay.qt_token);
if (!(token & 0x80)) {
debug("TOKEN=%#x\n", token);
switch (token & 0xfc) {
case 0:
toggle = token >> 31;
usb_settoggle(dev, usb_pipeendpoint(pipe),
usb_pipeout(pipe), toggle);
dev->status = 0;
break;
case 0x40:
dev->status = USB_ST_STALLED;
break;
case 0xa0:
case 0x20:
dev->status = USB_ST_BUF_ERR;
break;
case 0x50:
case 0x10:
dev->status = USB_ST_BABBLE_DET;
break;
default:
dev->status = USB_ST_CRC_ERR;
if ((token & 0x40) == 0x40)
dev->status |= USB_ST_STALLED;
break;
}
dev->act_len = length - ((token >> 16) & 0x7fff);
} else {
/* Flush the TF params and the TF plat params */
void bl2_plat_flush_bl31_params(void)
{
flush_dcache_range((unsigned long)&bl31_params_mem,
sizeof(bl2_to_bl31_params_mem_t));
}
int gmac_check_rx_done(struct eth_dma *dma, uint8_t *buf)
{
void *bufp, *datap;
size_t rcvlen = 0, buflen = 0;
uint32_t stat0 = 0, stat1 = 0;
uint32_t control, offset;
uint8_t statbuf[HWRXOFF*2];
int index, curr, active;
dma64dd_t *descp = NULL;
/* udelay(50); */
/*
* this api will check if a packet has been received.
* If so it will return the address of the buffer and current
* descriptor index will be incremented to the
* next descriptor. Once done with the frame the buffer should be
* added back onto the descriptor and the lastdscr should be updated
* to this descriptor.
*/
index = dma->cur_rx_index;
offset = (uint32_t)(dma->rx_desc_aligned);
stat0 = readl(GMAC0_DMA_RX_STATUS0_ADDR) & D64_RS0_CD_MASK;
stat1 = readl(GMAC0_DMA_RX_STATUS1_ADDR) & D64_RS0_CD_MASK;
curr = ((stat0 - offset) & D64_RS0_CD_MASK) / sizeof(dma64dd_t);
active = ((stat1 - offset) & D64_RS0_CD_MASK) / sizeof(dma64dd_t);
/* check if any frame */
if (index == curr)
return -1;
debug("received packet\n");
debug("expect(0x%x) curr(0x%x) active(0x%x)\n", index, curr, active);
/* remove warning */
if (index == active)
;
/* get the packet pointer that corresponds to the rx descriptor */
bufp = dma->rx_buf + index * RX_BUF_SIZE_ALIGNED;
descp = (dma64dd_t *)(dma->rx_desc_aligned) + index;
/* flush descriptor and buffer */
flush_dcache_range((unsigned long)dma->rx_desc_aligned,
(unsigned long)dma->rx_desc_aligned +
DESCP_SIZE_ALIGNED * RX_BUF_NUM);
flush_dcache_range((unsigned long)bufp,
(unsigned long)bufp + RX_BUF_SIZE_ALIGNED);
buflen = (descp->ctrl2 & D64_CTRL2_BC_MASK);
stat0 = readl(GMAC0_DMA_RX_STATUS0_ADDR);
stat1 = readl(GMAC0_DMA_RX_STATUS1_ADDR);
debug("bufp(0x%x) index(0x%x) buflen(0x%x) stat0(0x%x) stat1(0x%x)\n",
(uint32_t)bufp, index, buflen, stat0, stat1);
dma->cur_rx_index = (index + 1) & (RX_BUF_NUM - 1);
/* get buffer offset */
control = readl(GMAC0_DMA_RX_CTRL_ADDR);
offset = (control & D64_RC_RO_MASK) >> D64_RC_RO_SHIFT;
rcvlen = *(uint16_t *)bufp;
debug("Received %d bytes\n", rcvlen);
/* copy status into temp buf then copy data from rx buffer */
memcpy(statbuf, bufp, offset);
datap = (void *)((uint32_t)bufp + offset);
memcpy(buf, datap, rcvlen);
/* update descriptor that is being added back on ring */
descp->ctrl2 = RX_BUF_SIZE_ALIGNED;
descp->addrlow = (uint32_t)bufp;
descp->addrhigh = 0;
/* flush descriptor */
flush_dcache_range((unsigned long)dma->rx_desc_aligned,
(unsigned long)dma->rx_desc_aligned +
DESCP_SIZE_ALIGNED * RX_BUF_NUM);
/* set the lastdscr for the rx ring */
writel(((uint32_t)descp) & D64_XP_LD_MASK, GMAC0_DMA_RX_PTR_ADDR);
return (int)rcvlen;
}
static int
scc_enet_start_xmit(struct rtskb *skb, struct rtnet_device *rtdev)
{
struct scc_enet_private *cep = (struct scc_enet_private *)rtdev->priv;
volatile cbd_t *bdp;
RT_DEBUG(__FUNCTION__": ...\n");
/* Fill in a Tx ring entry */
bdp = cep->cur_tx;
#ifndef final_version
if (bdp->cbd_sc & BD_ENET_TX_READY) {
/* Ooops. All transmit buffers are full. Bail out.
* This should not happen, since cep->tx_busy should be set.
*/
printk("%s: tx queue full!.\n", rtdev->name);
return 1;
}
#endif
/* Clear all of the status flags.
*/
bdp->cbd_sc &= ~BD_ENET_TX_STATS;
/* If the frame is short, tell CPM to pad it.
*/
if (skb->len <= ETH_ZLEN)
bdp->cbd_sc |= BD_ENET_TX_PAD;
else
bdp->cbd_sc &= ~BD_ENET_TX_PAD;
/* Set buffer length and buffer pointer.
*/
bdp->cbd_datlen = skb->len;
bdp->cbd_bufaddr = __pa(skb->data);
/* Save skb pointer.
*/
cep->tx_skbuff[cep->skb_cur] = skb;
cep->stats.tx_bytes += skb->len;
cep->skb_cur = (cep->skb_cur+1) & TX_RING_MOD_MASK;
/* Push the data cache so the CPM does not get stale memory
* data.
*/
flush_dcache_range((unsigned long)(skb->data),
(unsigned long)(skb->data + skb->len));
/* Prevent interrupts from changing the Tx ring from underneath us. */
// *** RTnet ***
rt_sem_wait(&rtdev->xmit_sem);
rt_disable_irq(rtdev->irq);
rt_spin_lock(&cep->lock);
/* Send it on its way. Tell CPM its ready, interrupt when done,
* its the last BD of the frame, and to put the CRC on the end.
*/
bdp->cbd_sc |= (BD_ENET_TX_READY | BD_ENET_TX_INTR | BD_ENET_TX_LAST | BD_ENET_TX_TC);
#if 0
dev->trans_start = jiffies;
#endif
/* If this was the last BD in the ring, start at the beginning again.
*/
if (bdp->cbd_sc & BD_ENET_TX_WRAP)
bdp = cep->tx_bd_base;
else
bdp++;
if (bdp->cbd_sc & BD_ENET_TX_READY) {
rtnetif_stop_queue(rtdev);
cep->tx_full = 1;
}
cep->cur_tx = (cbd_t *)bdp;
rt_spin_unlock(&cep->lock);
rt_enable_irq(rtdev->irq);
rt_sem_signal(&rtdev->xmit_sem);
return 0;
}
static int
fec_enet_start_xmit(struct sk_buff *skb, struct net_device *dev)
{
struct fec_enet_private *fep;
volatile fec_t *fecp;
volatile cbd_t *bdp;
fep = dev->priv;
fecp = (volatile fec_t*)dev->base_addr;
if (!fep->link) {
/* Link is down or autonegotiation is in progress. */
return 1;
}
/* Fill in a Tx ring entry */
bdp = fep->cur_tx;
#ifndef final_version
if (bdp->cbd_sc & BD_ENET_TX_READY) {
/* Ooops. All transmit buffers are full. Bail out.
* This should not happen, since dev->tbusy should be set.
*/
printk("%s: tx queue full!.\n", dev->name);
return 1;
}
#endif
/* Clear all of the status flags.
*/
bdp->cbd_sc &= ~BD_ENET_TX_STATS;
/* Set buffer length and buffer pointer.
*/
bdp->cbd_bufaddr = __pa(skb->data);
bdp->cbd_datlen = skb->len;
/* Save skb pointer.
*/
fep->tx_skbuff[fep->skb_cur] = skb;
fep->stats.tx_bytes += skb->len;
fep->skb_cur = (fep->skb_cur+1) & TX_RING_MOD_MASK;
/* Push the data cache so the CPM does not get stale memory
* data.
*/
flush_dcache_range((unsigned long)skb->data,
(unsigned long)skb->data + skb->len);
/* disable interrupts while triggering transmit */
spin_lock_irq(&fep->lock);
/* Send it on its way. Tell FEC its ready, interrupt when done,
* its the last BD of the frame, and to put the CRC on the end.
*/
bdp->cbd_sc |= (BD_ENET_TX_READY | BD_ENET_TX_INTR
| BD_ENET_TX_LAST | BD_ENET_TX_TC);
dev->trans_start = jiffies;
/* Trigger transmission start */
fecp->fec_x_des_active = 0x01000000;
/* If this was the last BD in the ring, start at the beginning again.
*/
if (bdp->cbd_sc & BD_ENET_TX_WRAP) {
bdp = fep->tx_bd_base;
} else {
bdp++;
}
if (bdp->cbd_sc & BD_ENET_TX_READY) {
netif_stop_queue(dev);
fep->tx_full = 1;
}
fep->cur_tx = (cbd_t *)bdp;
spin_unlock_irq(&fep->lock);
return 0;
}
/**
* Start the FEC engine
* @param[in] dev Our device to handle
*/
static int fec_open(struct eth_device *edev)
{
struct fec_priv *fec = (struct fec_priv *)edev->priv;
int speed;
uint32_t addr, size;
int i;
debug("fec_open: fec_open(dev)\n");
/* full-duplex, heartbeat disabled */
writel(1 << 2, &fec->eth->x_cntrl);
fec->rbd_index = 0;
/* Invalidate all descriptors */
for (i = 0; i < FEC_RBD_NUM - 1; i++)
fec_rbd_clean(0, &fec->rbd_base[i]);
fec_rbd_clean(1, &fec->rbd_base[i]);
/* Flush the descriptors into RAM */
size = roundup(FEC_RBD_NUM * sizeof(struct fec_bd),
ARCH_DMA_MINALIGN);
addr = (uint32_t)fec->rbd_base;
flush_dcache_range(addr, addr + size);
#ifdef FEC_QUIRK_ENET_MAC
/* Enable ENET HW endian SWAP */
writel(readl(&fec->eth->ecntrl) | FEC_ECNTRL_DBSWAP,
&fec->eth->ecntrl);
/* Enable ENET store and forward mode */
writel(readl(&fec->eth->x_wmrk) | FEC_X_WMRK_STRFWD,
&fec->eth->x_wmrk);
#endif
/*
* Enable FEC-Lite controller
*/
writel(readl(&fec->eth->ecntrl) | FEC_ECNTRL_ETHER_EN,
&fec->eth->ecntrl);
#if defined(CONFIG_MX25) || defined(CONFIG_MX53) || defined(CONFIG_MX6SL)
udelay(100);
/*
* setup the MII gasket for RMII mode
*/
/* disable the gasket */
writew(0, &fec->eth->miigsk_enr);
/* wait for the gasket to be disabled */
while (readw(&fec->eth->miigsk_enr) & MIIGSK_ENR_READY)
udelay(2);
/* configure gasket for RMII, 50 MHz, no loopback, and no echo */
writew(MIIGSK_CFGR_IF_MODE_RMII, &fec->eth->miigsk_cfgr);
/* re-enable the gasket */
writew(MIIGSK_ENR_EN, &fec->eth->miigsk_enr);
/* wait until MII gasket is ready */
int max_loops = 10;
while ((readw(&fec->eth->miigsk_enr) & MIIGSK_ENR_READY) == 0) {
if (--max_loops <= 0) {
printf("WAIT for MII Gasket ready timed out\n");
break;
}
}
#endif
#ifdef CONFIG_PHYLIB
{
/* Start up the PHY */
int ret = phy_startup(fec->phydev);
if (ret) {
printf("Could not initialize PHY %s\n",
fec->phydev->dev->name);
return ret;
}
speed = fec->phydev->speed;
}
#elif CONFIG_FEC_FIXED_SPEED
speed = CONFIG_FEC_FIXED_SPEED;
#else
miiphy_wait_aneg(edev);
speed = miiphy_speed(edev->name, fec->phy_id);
miiphy_duplex(edev->name, fec->phy_id);
#endif
#ifdef FEC_QUIRK_ENET_MAC
{
u32 ecr = readl(&fec->eth->ecntrl) & ~FEC_ECNTRL_SPEED;
u32 rcr = readl(&fec->eth->r_cntrl) & ~FEC_RCNTRL_RMII_10T;
if (speed == _1000BASET)
ecr |= FEC_ECNTRL_SPEED;
else if (speed != _100BASET)
rcr |= FEC_RCNTRL_RMII_10T;
writel(ecr, &fec->eth->ecntrl);
writel(rcr, &fec->eth->r_cntrl);
}
#endif
debug("%s:Speed=%i\n", __func__, speed);
/*
* Enable SmartDMA receive task
*/
fec_rx_task_enable(fec);
udelay(100000);
return 0;
}
/*******************************************************************************
* Generic function to load an image into the trusted RAM using semihosting
* given a name, extents of free memory & whether the image should be loaded at
* the bottom or top of the free memory. It updates the memory layout if the
* load is successful.
******************************************************************************/
unsigned long load_image(meminfo *mem_layout,
const char *image_name,
unsigned int load_type,
unsigned long fixed_addr)
{
unsigned long temp_image_base, image_base;
long offset;
int image_flen;
/* Find the size of the image */
image_flen = semihosting_get_flen(image_name);
if (image_flen < 0) {
printf("ERROR: Cannot access '%s' file (%i).\r\n",
image_name, image_flen);
return 0;
}
/* See if we have enough space */
if (image_flen > mem_layout->free_size) {
printf("ERROR: Cannot load '%s' file: Not enough space.\r\n",
image_name);
dump_load_info(0, image_flen, mem_layout);
return 0;
}
switch (load_type) {
case TOP_LOAD:
/* Load the image in the top of free memory */
temp_image_base = mem_layout->free_base + mem_layout->free_size;
temp_image_base -= image_flen;
/* Page align base address and check whether the image still fits */
image_base = page_align(temp_image_base, DOWN);
assert(image_base <= temp_image_base);
if (image_base < mem_layout->free_base) {
printf("ERROR: Cannot load '%s' file: Not enough space.\r\n",
image_name);
dump_load_info(image_base, image_flen, mem_layout);
return 0;
}
/* Calculate the amount of extra memory used due to alignment */
offset = temp_image_base - image_base;
break;
case BOT_LOAD:
/* Load the BL2 image in the bottom of free memory */
temp_image_base = mem_layout->free_base;
image_base = page_align(temp_image_base, UP);
assert(image_base >= temp_image_base);
/* Page align base address and check whether the image still fits */
if (image_base + image_flen >
mem_layout->free_base + mem_layout->free_size) {
printf("ERROR: Cannot load '%s' file: Not enough space.\r\n",
image_name);
dump_load_info(image_base, image_flen, mem_layout);
return 0;
}
/* Calculate the amount of extra memory used due to alignment */
offset = image_base - temp_image_base;
break;
default:
assert(0);
}
/*
* Some images must be loaded at a fixed address, not a dynamic one.
*
* This has been implemented as a hack on top of the existing dynamic
* loading mechanism, for the time being. If the 'fixed_addr' function
* argument is different from zero, then it will force the load address.
* So we still have this principle of top/bottom loading but the code
* determining the load address is bypassed and the load address is
* forced to the fixed one.
*
* This can result in quite a lot of wasted space because we still use
* 1 sole meminfo structure to represent the extents of free memory,
* where we should use some sort of linked list.
*
* E.g. we want to load BL2 at address 0x04020000, the resulting memory
* layout should look as follows:
* ------------ 0x04040000
* | | <- Free space (1)
* |----------|
* | BL2 |
* |----------| 0x04020000
* | | <- Free space (2)
* |----------|
* | BL1 |
* ------------ 0x04000000
*
* But in the current hacky implementation, we'll need to specify
* whether BL2 is loaded at the top or bottom of the free memory.
* E.g. if BL2 is considered as top-loaded, the meminfo structure
* will give the following view of the memory, hiding the chunk of
* free memory above BL2:
* ------------ 0x04040000
* | |
* | |
* | BL2 |
* |----------| 0x04020000
* | | <- Free space (2)
* |----------|
* | BL1 |
* ------------ 0x04000000
*/
if (fixed_addr != 0) {
/* Load the image at the given address. */
image_base = fixed_addr;
/* Check whether the image fits. */
if ((image_base < mem_layout->free_base) ||
(image_base + image_flen >
mem_layout->free_base + mem_layout->free_size)) {
printf("ERROR: Cannot load '%s' file: Not enough space.\r\n",
image_name);
dump_load_info(image_base, image_flen, mem_layout);
return 0;
}
/* Check whether the fixed load address is page-aligned. */
if (!is_page_aligned(image_base)) {
printf("ERROR: Cannot load '%s' file at unaligned address 0x%lx.\r\n",
image_name, fixed_addr);
return 0;
}
/*
* Calculate the amount of extra memory used due to fixed
* loading.
*/
if (load_type == TOP_LOAD) {
unsigned long max_addr, space_used;
/*
* ------------ max_addr
* | /wasted/ | | offset
* |..........|..............................
* | image | | image_flen
* |----------| fixed_addr
* | |
* | |
* ------------ total_base
*/
max_addr = mem_layout->total_base + mem_layout->total_size;
/*
* Compute the amount of memory used by the image.
* Corresponds to all space above the image load
* address.
*/
space_used = max_addr - fixed_addr;
/*
* Calculate the amount of wasted memory within the
* amount of memory used by the image.
*/
offset = space_used - image_flen;
} else /* BOT_LOAD */
/*
* ------------
* | |
* | |
* |----------|
* | image |
* |..........| fixed_addr
* | /wasted/ | | offset
* ------------ total_base
*/
offset = fixed_addr - mem_layout->total_base;
}
/* We have enough space so load the image now */
image_flen = semihosting_download_file(image_name,
image_flen,
(void *) image_base);
if (image_flen <= 0) {
printf("ERROR: Failed to load '%s' file from semihosting (%i).\r\n",
image_name, image_flen);
return 0;
}
/*
* File has been successfully loaded. Update the free memory
* data structure & flush the contents of the TZRAM so that
* the next EL can see it.
*/
/* Update the memory contents */
flush_dcache_range(image_base, image_flen);
mem_layout->free_size -= image_flen + offset;
/* Update the base of free memory since its moved up */
if (load_type == BOT_LOAD)
mem_layout->free_base += offset + image_flen;
return image_base;
}
/**
* Pull one frame from the card
* @param[in] dev Our ethernet device to handle
* @return Length of packet read
*/
static int fec_recv(struct eth_device *dev)
{
struct fec_priv *fec = (struct fec_priv *)dev->priv;
struct fec_bd *rbd = &fec->rbd_base[fec->rbd_index];
unsigned long ievent;
int frame_length, len = 0;
uint16_t bd_status;
uint32_t addr, size, end;
int i;
ALLOC_CACHE_ALIGN_BUFFER(uchar, buff, FEC_MAX_PKT_SIZE);
/*
* Check if any critical events have happened
*/
ievent = readl(&fec->eth->ievent);
writel(ievent, &fec->eth->ievent);
debug("fec_recv: ievent 0x%lx\n", ievent);
if (ievent & FEC_IEVENT_BABR) {
fec_halt(dev);
fec_init(dev, fec->bd);
printf("some error: 0x%08lx\n", ievent);
return 0;
}
if (ievent & FEC_IEVENT_HBERR) {
/* Heartbeat error */
writel(0x00000001 | readl(&fec->eth->x_cntrl),
&fec->eth->x_cntrl);
}
if (ievent & FEC_IEVENT_GRA) {
/* Graceful stop complete */
if (readl(&fec->eth->x_cntrl) & 0x00000001) {
fec_halt(dev);
writel(~0x00000001 & readl(&fec->eth->x_cntrl),
&fec->eth->x_cntrl);
fec_init(dev, fec->bd);
}
}
/*
* Read the buffer status. Before the status can be read, the data cache
* must be invalidated, because the data in RAM might have been changed
* by DMA. The descriptors are properly aligned to cachelines so there's
* no need to worry they'd overlap.
*
* WARNING: By invalidating the descriptor here, we also invalidate
* the descriptors surrounding this one. Therefore we can NOT change the
* contents of this descriptor nor the surrounding ones. The problem is
* that in order to mark the descriptor as processed, we need to change
* the descriptor. The solution is to mark the whole cache line when all
* descriptors in the cache line are processed.
*/
addr = (uint32_t)rbd;
addr &= ~(ARCH_DMA_MINALIGN - 1);
size = roundup(sizeof(struct fec_bd), ARCH_DMA_MINALIGN);
invalidate_dcache_range(addr, addr + size);
bd_status = readw(&rbd->status);
debug("fec_recv: status 0x%x\n", bd_status);
if (!(bd_status & FEC_RBD_EMPTY)) {
if ((bd_status & FEC_RBD_LAST) && !(bd_status & FEC_RBD_ERR) &&
((readw(&rbd->data_length) - 4) > 14)) {
/*
* Get buffer address and size
*/
addr = readl(&rbd->data_pointer);
frame_length = readw(&rbd->data_length) - 4;
/*
* Invalidate data cache over the buffer
*/
end = roundup(addr + frame_length, ARCH_DMA_MINALIGN);
addr &= ~(ARCH_DMA_MINALIGN - 1);
invalidate_dcache_range(addr, end);
/*
* Fill the buffer and pass it to upper layers
*/
#ifdef CONFIG_FEC_MXC_SWAP_PACKET
swap_packet((uint32_t *)addr, frame_length);
#endif
memcpy(buff, (char *)addr, frame_length);
net_process_received_packet(buff, frame_length);
len = frame_length;
} else {
if (bd_status & FEC_RBD_ERR)
printf("error frame: 0x%08x 0x%08x\n",
addr, bd_status);
}
/*
* Free the current buffer, restart the engine and move forward
* to the next buffer. Here we check if the whole cacheline of
* descriptors was already processed and if so, we mark it free
* as whole.
*/
size = RXDESC_PER_CACHELINE - 1;
if ((fec->rbd_index & size) == size) {
i = fec->rbd_index - size;
addr = (uint32_t)&fec->rbd_base[i];
for (; i <= fec->rbd_index ; i++) {
fec_rbd_clean(i == (FEC_RBD_NUM - 1),
&fec->rbd_base[i]);
}
flush_dcache_range(addr,
addr + ARCH_DMA_MINALIGN);
}
fec_rx_task_enable(fec);
fec->rbd_index = (fec->rbd_index + 1) % FEC_RBD_NUM;
}
debug("fec_recv: stop\n");
return len;
}
/*
* Consistent memory allocators. Used for DMA devices that want to
* share uncached memory with the processor core.
* My crufty no-MMU approach is simple. In the HW platform we can optionally
* mirror the DDR up above the processor cacheable region. So, memory accessed
* in this mirror region will not be cached. It's alloced from the same
* pool as normal memory, but the handle we return is shifted up into the
* uncached region. This will no doubt cause big problems if memory allocated
* here is not also freed properly. -- JW
*/
void *consistent_alloc(gfp_t gfp, size_t size, dma_addr_t *dma_handle)
{
unsigned long order, vaddr;
void *ret;
unsigned int i, err = 0;
struct page *page, *end;
#ifdef CONFIG_MMU
phys_addr_t pa;
struct vm_struct *area;
unsigned long va;
#endif
if (in_interrupt())
BUG();
/* Only allocate page size areas. */
size = PAGE_ALIGN(size);
order = get_order(size);
vaddr = __get_free_pages(gfp, order);
if (!vaddr)
return NULL;
/*
* we need to ensure that there are no cachelines in use,
* or worse dirty in this area.
*/
flush_dcache_range(virt_to_phys((void *)vaddr),
virt_to_phys((void *)vaddr) + size);
#ifndef CONFIG_MMU
ret = (void *)vaddr;
/*
* Here's the magic! Note if the uncached shadow is not implemented,
* it's up to the calling code to also test that condition and make
* other arranegments, such as manually flushing the cache and so on.
*/
# ifdef CONFIG_XILINX_UNCACHED_SHADOW
ret = (void *)((unsigned) ret | UNCACHED_SHADOW_MASK);
# endif
if ((unsigned int)ret > cpuinfo.dcache_base &&
(unsigned int)ret < cpuinfo.dcache_high)
pr_warn("ERROR: Your cache coherent area is CACHED!!!\n");
/* dma_handle is same as physical (shadowed) address */
*dma_handle = (dma_addr_t)ret;
#else
/* Allocate some common virtual space to map the new pages. */
area = get_vm_area(size, VM_ALLOC);
if (!area) {
free_pages(vaddr, order);
return NULL;
}
va = (unsigned long) area->addr;
ret = (void *)va;
/* This gives us the real physical address of the first page. */
*dma_handle = pa = __virt_to_phys(vaddr);
#endif
/*
* free wasted pages. We skip the first page since we know
* that it will have count = 1 and won't require freeing.
* We also mark the pages in use as reserved so that
* remap_page_range works.
*/
page = virt_to_page(vaddr);
end = page + (1 << order);
split_page(page, order);
for (i = 0; i < size && err == 0; i += PAGE_SIZE) {
#ifdef CONFIG_MMU
/* MS: This is the whole magic - use cache inhibit pages */
err = map_page(va + i, pa + i, _PAGE_KERNEL | _PAGE_NO_CACHE);
#endif
SetPageReserved(page);
page++;
}
/* Free the otherwise unused pages. */
while (page < end) {
__free_page(page);
page++;
}
if (err) {
free_pages(vaddr, order);
return NULL;
}
return ret;
}