linux/drivers/net/irda/sh_sir.c

825 lines
18 KiB
C
Raw Normal View History

/*
* SuperH IrDA Driver
*
* Copyright (C) 2009 Renesas Solutions Corp.
* Kuninori Morimoto <morimoto.kuninori@renesas.com>
*
* Based on bfin_sir.c
* Copyright 2006-2009 Analog Devices Inc.
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*/
#include <linux/module.h>
#include <linux/platform_device.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
#include <net/irda/wrapper.h>
#include <net/irda/irda_device.h>
#include <asm/clock.h>
#define DRIVER_NAME "sh_sir"
#define RX_PHASE (1 << 0)
#define TX_PHASE (1 << 1)
#define TX_COMP_PHASE (1 << 2) /* tx complete */
#define NONE_PHASE (1 << 31)
#define IRIF_RINTCLR 0x0016 /* DMA rx interrupt source clear */
#define IRIF_TINTCLR 0x0018 /* DMA tx interrupt source clear */
#define IRIF_SIR0 0x0020 /* IrDA-SIR10 control */
#define IRIF_SIR1 0x0022 /* IrDA-SIR10 baudrate error correction */
#define IRIF_SIR2 0x0024 /* IrDA-SIR10 baudrate count */
#define IRIF_SIR3 0x0026 /* IrDA-SIR10 status */
#define IRIF_SIR_FRM 0x0028 /* Hardware frame processing set */
#define IRIF_SIR_EOF 0x002A /* EOF value */
#define IRIF_SIR_FLG 0x002C /* Flag clear */
#define IRIF_UART_STS2 0x002E /* UART status 2 */
#define IRIF_UART0 0x0030 /* UART control */
#define IRIF_UART1 0x0032 /* UART status */
#define IRIF_UART2 0x0034 /* UART mode */
#define IRIF_UART3 0x0036 /* UART transmit data */
#define IRIF_UART4 0x0038 /* UART receive data */
#define IRIF_UART5 0x003A /* UART interrupt mask */
#define IRIF_UART6 0x003C /* UART baud rate error correction */
#define IRIF_UART7 0x003E /* UART baud rate count set */
#define IRIF_CRC0 0x0040 /* CRC engine control */
#define IRIF_CRC1 0x0042 /* CRC engine input data */
#define IRIF_CRC2 0x0044 /* CRC engine calculation */
#define IRIF_CRC3 0x0046 /* CRC engine output data 1 */
#define IRIF_CRC4 0x0048 /* CRC engine output data 2 */
/* IRIF_SIR0 */
#define IRTPW (1 << 1) /* transmit pulse width select */
#define IRERRC (1 << 0) /* Clear receive pulse width error */
/* IRIF_SIR3 */
#define IRERR (1 << 0) /* received pulse width Error */
/* IRIF_SIR_FRM */
#define EOFD (1 << 9) /* EOF detection flag */
#define FRER (1 << 8) /* Frame Error bit */
#define FRP (1 << 0) /* Frame processing set */
/* IRIF_UART_STS2 */
#define IRSME (1 << 6) /* Receive Sum Error flag */
#define IROVE (1 << 5) /* Receive Overrun Error flag */
#define IRFRE (1 << 4) /* Receive Framing Error flag */
#define IRPRE (1 << 3) /* Receive Parity Error flag */
/* IRIF_UART0_*/
#define TBEC (1 << 2) /* Transmit Data Clear */
#define RIE (1 << 1) /* Receive Enable */
#define TIE (1 << 0) /* Transmit Enable */
/* IRIF_UART1 */
#define URSME (1 << 6) /* Receive Sum Error Flag */
#define UROVE (1 << 5) /* Receive Overrun Error Flag */
#define URFRE (1 << 4) /* Receive Framing Error Flag */
#define URPRE (1 << 3) /* Receive Parity Error Flag */
#define RBF (1 << 2) /* Receive Buffer Full Flag */
#define TSBE (1 << 1) /* Transmit Shift Buffer Empty Flag */
#define TBE (1 << 0) /* Transmit Buffer Empty flag */
#define TBCOMP (TSBE | TBE)
/* IRIF_UART5 */
#define RSEIM (1 << 6) /* Receive Sum Error Flag IRQ Mask */
#define RBFIM (1 << 2) /* Receive Buffer Full Flag IRQ Mask */
#define TSBEIM (1 << 1) /* Transmit Shift Buffer Empty Flag IRQ Mask */
#define TBEIM (1 << 0) /* Transmit Buffer Empty Flag IRQ Mask */
#define RX_MASK (RSEIM | RBFIM)
/* IRIF_CRC0 */
#define CRC_RST (1 << 15) /* CRC Engine Reset */
#define CRC_CT_MASK 0x0FFF
/************************************************************************
structure
************************************************************************/
struct sh_sir_self {
void __iomem *membase;
unsigned int irq;
struct clk *clk;
struct net_device *ndev;
struct irlap_cb *irlap;
struct qos_info qos;
iobuff_t tx_buff;
iobuff_t rx_buff;
};
/************************************************************************
common function
************************************************************************/
static void sh_sir_write(struct sh_sir_self *self, u32 offset, u16 data)
{
iowrite16(data, self->membase + offset);
}
static u16 sh_sir_read(struct sh_sir_self *self, u32 offset)
{
return ioread16(self->membase + offset);
}
static void sh_sir_update_bits(struct sh_sir_self *self, u32 offset,
u16 mask, u16 data)
{
u16 old, new;
old = sh_sir_read(self, offset);
new = (old & ~mask) | data;
if (old != new)
sh_sir_write(self, offset, new);
}
/************************************************************************
CRC function
************************************************************************/
static void sh_sir_crc_reset(struct sh_sir_self *self)
{
sh_sir_write(self, IRIF_CRC0, CRC_RST);
}
static void sh_sir_crc_add(struct sh_sir_self *self, u8 data)
{
sh_sir_write(self, IRIF_CRC1, (u16)data);
}
static u16 sh_sir_crc_cnt(struct sh_sir_self *self)
{
return CRC_CT_MASK & sh_sir_read(self, IRIF_CRC0);
}
static u16 sh_sir_crc_out(struct sh_sir_self *self)
{
return sh_sir_read(self, IRIF_CRC4);
}
static int sh_sir_crc_init(struct sh_sir_self *self)
{
struct device *dev = &self->ndev->dev;
int ret = -EIO;
u16 val;
sh_sir_crc_reset(self);
sh_sir_crc_add(self, 0xCC);
sh_sir_crc_add(self, 0xF5);
sh_sir_crc_add(self, 0xF1);
sh_sir_crc_add(self, 0xA7);
val = sh_sir_crc_cnt(self);
if (4 != val) {
dev_err(dev, "CRC count error %x\n", val);
goto crc_init_out;
}
val = sh_sir_crc_out(self);
if (0x51DF != val) {
dev_err(dev, "CRC result error%x\n", val);
goto crc_init_out;
}
ret = 0;
crc_init_out:
sh_sir_crc_reset(self);
return ret;
}
/************************************************************************
baud rate functions
************************************************************************/
#define SCLK_BASE 1843200 /* 1.8432MHz */
static u32 sh_sir_find_sclk(struct clk *irda_clk)
{
struct cpufreq_frequency_table *freq_table = irda_clk->freq_table;
struct clk *pclk = clk_get(NULL, "peripheral_clk");
u32 limit, min = 0xffffffff, tmp;
int i, index = 0;
limit = clk_get_rate(pclk);
clk_put(pclk);
/* IrDA can not set over peripheral_clk */
for (i = 0;
freq_table[i].frequency != CPUFREQ_TABLE_END;
i++) {
u32 freq = freq_table[i].frequency;
if (freq == CPUFREQ_ENTRY_INVALID)
continue;
/* IrDA should not over peripheral_clk */
if (freq > limit)
continue;
tmp = freq % SCLK_BASE;
if (tmp < min) {
min = tmp;
index = i;
}
}
return freq_table[index].frequency;
}
#define ERR_ROUNDING(a) ((a + 5000) / 10000)
static int sh_sir_set_baudrate(struct sh_sir_self *self, u32 baudrate)
{
struct clk *clk;
struct device *dev = &self->ndev->dev;
u32 rate;
u16 uabca, uabc;
u16 irbca, irbc;
u32 min, rerr, tmp;
int i;
/* Baud Rate Error Correction x 10000 */
u32 rate_err_array[] = {
0, 625, 1250, 1875,
2500, 3125, 3750, 4375,
5000, 5625, 6250, 6875,
7500, 8125, 8750, 9375,
};
/*
* FIXME
*
* it support 9600 only now
*/
switch (baudrate) {
case 9600:
break;
default:
dev_err(dev, "un-supported baudrate %d\n", baudrate);
return -EIO;
}
clk = clk_get(NULL, "irda_clk");
if (!clk) {
dev_err(dev, "can not get irda_clk\n");
return -EIO;
}
clk_set_rate(clk, sh_sir_find_sclk(clk));
rate = clk_get_rate(clk);
clk_put(clk);
dev_dbg(dev, "selected sclk = %d\n", rate);
/*
* CALCULATION
*
* 1843200 = system rate / (irbca + (irbc + 1))
*/
irbc = rate / SCLK_BASE;
tmp = rate - (SCLK_BASE * irbc);
tmp *= 10000;
rerr = tmp / SCLK_BASE;
min = 0xffffffff;
irbca = 0;
for (i = 0; i < ARRAY_SIZE(rate_err_array); i++) {
tmp = abs(rate_err_array[i] - rerr);
if (min > tmp) {
min = tmp;
irbca = i;
}
}
tmp = rate / (irbc + ERR_ROUNDING(rate_err_array[irbca]));
if ((SCLK_BASE / 100) < abs(tmp - SCLK_BASE))
dev_warn(dev, "IrDA freq error margin over %d\n", tmp);
dev_dbg(dev, "target = %d, result = %d, infrared = %d.%d\n",
SCLK_BASE, tmp, irbc, rate_err_array[irbca]);
irbca = (irbca & 0xF) << 4;
irbc = (irbc - 1) & 0xF;
if (!irbc) {
dev_err(dev, "sh_sir can not set 0 in IRIF_SIR2\n");
return -EIO;
}
sh_sir_write(self, IRIF_SIR0, IRTPW | IRERRC);
sh_sir_write(self, IRIF_SIR1, irbca);
sh_sir_write(self, IRIF_SIR2, irbc);
/*
* CALCULATION
*
* BaudRate[bps] = system rate / (uabca + (uabc + 1) x 16)
*/
uabc = rate / baudrate;
uabc = (uabc / 16) - 1;
uabc = (uabc + 1) * 16;
tmp = rate - (uabc * baudrate);
tmp *= 10000;
rerr = tmp / baudrate;
min = 0xffffffff;
uabca = 0;
for (i = 0; i < ARRAY_SIZE(rate_err_array); i++) {
tmp = abs(rate_err_array[i] - rerr);
if (min > tmp) {
min = tmp;
uabca = i;
}
}
tmp = rate / (uabc + ERR_ROUNDING(rate_err_array[uabca]));
if ((baudrate / 100) < abs(tmp - baudrate))
dev_warn(dev, "UART freq error margin over %d\n", tmp);
dev_dbg(dev, "target = %d, result = %d, uart = %d.%d\n",
baudrate, tmp,
uabc, rate_err_array[uabca]);
uabca = (uabca & 0xF) << 4;
uabc = (uabc / 16) - 1;
sh_sir_write(self, IRIF_UART6, uabca);
sh_sir_write(self, IRIF_UART7, uabc);
return 0;
}
/************************************************************************
iobuf function
************************************************************************/
static int __sh_sir_init_iobuf(iobuff_t *io, int size)
{
io->head = kmalloc(size, GFP_KERNEL);
if (!io->head)
return -ENOMEM;
io->truesize = size;
io->in_frame = FALSE;
io->state = OUTSIDE_FRAME;
io->data = io->head;
return 0;
}
static void sh_sir_remove_iobuf(struct sh_sir_self *self)
{
kfree(self->rx_buff.head);
kfree(self->tx_buff.head);
self->rx_buff.head = NULL;
self->tx_buff.head = NULL;
}
static int sh_sir_init_iobuf(struct sh_sir_self *self, int rxsize, int txsize)
{
int err = -ENOMEM;
if (self->rx_buff.head ||
self->tx_buff.head) {
dev_err(&self->ndev->dev, "iobuff has already existed.");
return err;
}
err = __sh_sir_init_iobuf(&self->rx_buff, rxsize);
if (err)
goto iobuf_err;
err = __sh_sir_init_iobuf(&self->tx_buff, txsize);
iobuf_err:
if (err)
sh_sir_remove_iobuf(self);
return err;
}
/************************************************************************
status function
************************************************************************/
static void sh_sir_clear_all_err(struct sh_sir_self *self)
{
/* Clear error flag for receive pulse width */
sh_sir_update_bits(self, IRIF_SIR0, IRERRC, IRERRC);
/* Clear frame / EOF error flag */
sh_sir_write(self, IRIF_SIR_FLG, 0xffff);
/* Clear all status error */
sh_sir_write(self, IRIF_UART_STS2, 0);
}
static void sh_sir_set_phase(struct sh_sir_self *self, int phase)
{
u16 uart5 = 0;
u16 uart0 = 0;
switch (phase) {
case TX_PHASE:
uart5 = TBEIM;
uart0 = TBEC | TIE;
break;
case TX_COMP_PHASE:
uart5 = TSBEIM;
uart0 = TIE;
break;
case RX_PHASE:
uart5 = RX_MASK;
uart0 = RIE;
break;
default:
break;
}
sh_sir_write(self, IRIF_UART5, uart5);
sh_sir_write(self, IRIF_UART0, uart0);
}
static int sh_sir_is_which_phase(struct sh_sir_self *self)
{
u16 val = sh_sir_read(self, IRIF_UART5);
if (val & TBEIM)
return TX_PHASE;
if (val & TSBEIM)
return TX_COMP_PHASE;
if (val & RX_MASK)
return RX_PHASE;
return NONE_PHASE;
}
static void sh_sir_tx(struct sh_sir_self *self, int phase)
{
switch (phase) {
case TX_PHASE:
if (0 >= self->tx_buff.len) {
sh_sir_set_phase(self, TX_COMP_PHASE);
} else {
sh_sir_write(self, IRIF_UART3, self->tx_buff.data[0]);
self->tx_buff.len--;
self->tx_buff.data++;
}
break;
case TX_COMP_PHASE:
sh_sir_set_phase(self, RX_PHASE);
netif_wake_queue(self->ndev);
break;
default:
dev_err(&self->ndev->dev, "should not happen\n");
break;
}
}
static int sh_sir_read_data(struct sh_sir_self *self)
{
u16 val;
int timeout = 1024;
while (timeout--) {
val = sh_sir_read(self, IRIF_UART1);
/* data get */
if (val & RBF) {
if (val & (URSME | UROVE | URFRE | URPRE))
break;
return (int)sh_sir_read(self, IRIF_UART4);
}
udelay(1);
}
dev_err(&self->ndev->dev, "UART1 %04x : STATUS %04x\n",
val, sh_sir_read(self, IRIF_UART_STS2));
/* read data register for clear error */
sh_sir_read(self, IRIF_UART4);
return -1;
}
static void sh_sir_rx(struct sh_sir_self *self)
{
int timeout = 1024;
int data;
while (timeout--) {
data = sh_sir_read_data(self);
if (data < 0)
break;
async_unwrap_char(self->ndev, &self->ndev->stats,
&self->rx_buff, (u8)data);
self->ndev->last_rx = jiffies;
if (EOFD & sh_sir_read(self, IRIF_SIR_FRM))
continue;
break;
}
}
static irqreturn_t sh_sir_irq(int irq, void *dev_id)
{
struct sh_sir_self *self = dev_id;
struct device *dev = &self->ndev->dev;
int phase = sh_sir_is_which_phase(self);
switch (phase) {
case TX_COMP_PHASE:
case TX_PHASE:
sh_sir_tx(self, phase);
break;
case RX_PHASE:
if (sh_sir_read(self, IRIF_SIR3))
dev_err(dev, "rcv pulse width error occurred\n");
sh_sir_rx(self);
sh_sir_clear_all_err(self);
break;
default:
dev_err(dev, "unknown interrupt\n");
}
return IRQ_HANDLED;
}
/************************************************************************
net_device_ops function
************************************************************************/
static int sh_sir_hard_xmit(struct sk_buff *skb, struct net_device *ndev)
{
struct sh_sir_self *self = netdev_priv(ndev);
int speed = irda_get_next_speed(skb);
if ((0 < speed) &&
(9600 != speed)) {
dev_err(&ndev->dev, "support 9600 only (%d)\n", speed);
return -EIO;
}
netif_stop_queue(ndev);
self->tx_buff.data = self->tx_buff.head;
self->tx_buff.len = 0;
if (skb->len)
self->tx_buff.len = async_wrap_skb(skb, self->tx_buff.data,
self->tx_buff.truesize);
sh_sir_set_phase(self, TX_PHASE);
dev_kfree_skb(skb);
return 0;
}
static int sh_sir_ioctl(struct net_device *ndev, struct ifreq *ifreq, int cmd)
{
/*
* FIXME
*
* This function is needed for irda framework.
* But nothing to do now
*/
return 0;
}
static struct net_device_stats *sh_sir_stats(struct net_device *ndev)
{
struct sh_sir_self *self = netdev_priv(ndev);
return &self->ndev->stats;
}
static int sh_sir_open(struct net_device *ndev)
{
struct sh_sir_self *self = netdev_priv(ndev);
int err;
clk_enable(self->clk);
err = sh_sir_crc_init(self);
if (err)
goto open_err;
sh_sir_set_baudrate(self, 9600);
self->irlap = irlap_open(ndev, &self->qos, DRIVER_NAME);
if (!self->irlap) {
err = -ENODEV;
goto open_err;
}
/*
* Now enable the interrupt then start the queue
*/
sh_sir_update_bits(self, IRIF_SIR_FRM, FRP, FRP);
sh_sir_read(self, IRIF_UART1); /* flag clear */
sh_sir_read(self, IRIF_UART4); /* flag clear */
sh_sir_set_phase(self, RX_PHASE);
netif_start_queue(ndev);
dev_info(&self->ndev->dev, "opened\n");
return 0;
open_err:
clk_disable(self->clk);
return err;
}
static int sh_sir_stop(struct net_device *ndev)
{
struct sh_sir_self *self = netdev_priv(ndev);
/* Stop IrLAP */
if (self->irlap) {
irlap_close(self->irlap);
self->irlap = NULL;
}
netif_stop_queue(ndev);
dev_info(&ndev->dev, "stoped\n");
return 0;
}
static const struct net_device_ops sh_sir_ndo = {
.ndo_open = sh_sir_open,
.ndo_stop = sh_sir_stop,
.ndo_start_xmit = sh_sir_hard_xmit,
.ndo_do_ioctl = sh_sir_ioctl,
.ndo_get_stats = sh_sir_stats,
};
/************************************************************************
platform_driver function
************************************************************************/
static int __devinit sh_sir_probe(struct platform_device *pdev)
{
struct net_device *ndev;
struct sh_sir_self *self;
struct resource *res;
char clk_name[8];
int irq;
int err = -ENOMEM;
res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
irq = platform_get_irq(pdev, 0);
if (!res || irq < 0) {
dev_err(&pdev->dev, "Not enough platform resources.\n");
goto exit;
}
ndev = alloc_irdadev(sizeof(*self));
if (!ndev)
goto exit;
self = netdev_priv(ndev);
self->membase = ioremap_nocache(res->start, resource_size(res));
if (!self->membase) {
err = -ENXIO;
dev_err(&pdev->dev, "Unable to ioremap.\n");
goto err_mem_1;
}
err = sh_sir_init_iobuf(self, IRDA_SKB_MAX_MTU, IRDA_SIR_MAX_FRAME);
if (err)
goto err_mem_2;
snprintf(clk_name, sizeof(clk_name), "irda%d", pdev->id);
self->clk = clk_get(&pdev->dev, clk_name);
if (IS_ERR(self->clk)) {
dev_err(&pdev->dev, "cannot get clock \"%s\"\n", clk_name);
goto err_mem_3;
}
irda_init_max_qos_capabilies(&self->qos);
ndev->netdev_ops = &sh_sir_ndo;
ndev->irq = irq;
self->ndev = ndev;
self->qos.baud_rate.bits &= IR_9600; /* FIXME */
self->qos.min_turn_time.bits = 1; /* 10 ms or more */
irda_qos_bits_to_value(&self->qos);
err = register_netdev(ndev);
if (err)
goto err_mem_4;
platform_set_drvdata(pdev, ndev);
if (request_irq(irq, sh_sir_irq, IRQF_DISABLED, "sh_sir", self)) {
dev_warn(&pdev->dev, "Unable to attach sh_sir interrupt\n");
goto err_mem_4;
}
dev_info(&pdev->dev, "SuperH IrDA probed\n");
goto exit;
err_mem_4:
clk_put(self->clk);
err_mem_3:
sh_sir_remove_iobuf(self);
err_mem_2:
iounmap(self->membase);
err_mem_1:
free_netdev(ndev);
exit:
return err;
}
static int __devexit sh_sir_remove(struct platform_device *pdev)
{
struct net_device *ndev = platform_get_drvdata(pdev);
struct sh_sir_self *self = netdev_priv(ndev);
if (!self)
return 0;
unregister_netdev(ndev);
clk_put(self->clk);
sh_sir_remove_iobuf(self);
iounmap(self->membase);
free_netdev(ndev);
platform_set_drvdata(pdev, NULL);
return 0;
}
static struct platform_driver sh_sir_driver = {
.probe = sh_sir_probe,
.remove = __devexit_p(sh_sir_remove),
.driver = {
.name = DRIVER_NAME,
},
};
static int __init sh_sir_init(void)
{
return platform_driver_register(&sh_sir_driver);
}
static void __exit sh_sir_exit(void)
{
platform_driver_unregister(&sh_sir_driver);
}
module_init(sh_sir_init);
module_exit(sh_sir_exit);
MODULE_AUTHOR("Kuninori Morimoto <morimoto.kuninori@renesas.com>");
MODULE_DESCRIPTION("SuperH IrDA driver");
MODULE_LICENSE("GPL");