linux-sg2042/drivers/spi/pxa2xx_spi.c

1673 lines
44 KiB
C
Raw Normal View History

/*
* Copyright (C) 2005 Stephen Street / StreetFire Sound Labs
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
#include <linux/init.h>
#include <linux/module.h>
#include <linux/device.h>
#include <linux/ioport.h>
#include <linux/errno.h>
#include <linux/interrupt.h>
#include <linux/platform_device.h>
#include <linux/dma-mapping.h>
#include <linux/spi/spi.h>
#include <linux/workqueue.h>
#include <linux/delay.h>
#include <linux/clk.h>
#include <asm/io.h>
#include <asm/irq.h>
#include <asm/delay.h>
#include <asm/dma.h>
#include <mach/hardware.h>
#include <mach/pxa-regs.h>
#include <mach/regs-ssp.h>
#include <mach/ssp.h>
#include <mach/pxa2xx_spi.h>
MODULE_AUTHOR("Stephen Street");
MODULE_DESCRIPTION("PXA2xx SSP SPI Controller");
MODULE_LICENSE("GPL");
MODULE_ALIAS("platform:pxa2xx-spi");
#define MAX_BUSES 3
#define DMA_INT_MASK (DCSR_ENDINTR | DCSR_STARTINTR | DCSR_BUSERR)
#define RESET_DMA_CHANNEL (DCSR_NODESC | DMA_INT_MASK)
#define IS_DMA_ALIGNED(x) ((((u32)(x)) & 0x07) == 0)
#define MAX_DMA_LEN 8191
/*
* for testing SSCR1 changes that require SSP restart, basically
* everything except the service and interrupt enables, the pxa270 developer
* manual says only SSCR1_SCFR, SSCR1_SPH, SSCR1_SPO need to be in this
* list, but the PXA255 dev man says all bits without really meaning the
* service and interrupt enables
*/
#define SSCR1_CHANGE_MASK (SSCR1_TTELP | SSCR1_TTE | SSCR1_SCFR \
| SSCR1_ECRA | SSCR1_ECRB | SSCR1_SCLKDIR \
| SSCR1_SFRMDIR | SSCR1_RWOT | SSCR1_TRAIL \
| SSCR1_IFS | SSCR1_STRF | SSCR1_EFWR \
| SSCR1_RFT | SSCR1_TFT | SSCR1_MWDS \
| SSCR1_SPH | SSCR1_SPO | SSCR1_LBM)
#define DEFINE_SSP_REG(reg, off) \
static inline u32 read_##reg(void const __iomem *p) \
{ return __raw_readl(p + (off)); } \
\
static inline void write_##reg(u32 v, void __iomem *p) \
{ __raw_writel(v, p + (off)); }
DEFINE_SSP_REG(SSCR0, 0x00)
DEFINE_SSP_REG(SSCR1, 0x04)
DEFINE_SSP_REG(SSSR, 0x08)
DEFINE_SSP_REG(SSITR, 0x0c)
DEFINE_SSP_REG(SSDR, 0x10)
DEFINE_SSP_REG(SSTO, 0x28)
DEFINE_SSP_REG(SSPSP, 0x2c)
#define START_STATE ((void*)0)
#define RUNNING_STATE ((void*)1)
#define DONE_STATE ((void*)2)
#define ERROR_STATE ((void*)-1)
#define QUEUE_RUNNING 0
#define QUEUE_STOPPED 1
struct driver_data {
/* Driver model hookup */
struct platform_device *pdev;
/* SSP Info */
struct ssp_device *ssp;
/* SPI framework hookup */
enum pxa_ssp_type ssp_type;
struct spi_master *master;
/* PXA hookup */
struct pxa2xx_spi_master *master_info;
/* DMA setup stuff */
int rx_channel;
int tx_channel;
u32 *null_dma_buf;
/* SSP register addresses */
void __iomem *ioaddr;
u32 ssdr_physical;
/* SSP masks*/
u32 dma_cr1;
u32 int_cr1;
u32 clear_sr;
u32 mask_sr;
/* Driver message queue */
struct workqueue_struct *workqueue;
struct work_struct pump_messages;
spinlock_t lock;
struct list_head queue;
int busy;
int run;
/* Message Transfer pump */
struct tasklet_struct pump_transfers;
/* Current message transfer state info */
struct spi_message* cur_msg;
struct spi_transfer* cur_transfer;
struct chip_data *cur_chip;
size_t len;
void *tx;
void *tx_end;
void *rx;
void *rx_end;
int dma_mapped;
dma_addr_t rx_dma;
dma_addr_t tx_dma;
size_t rx_map_len;
size_t tx_map_len;
u8 n_bytes;
u32 dma_width;
int (*write)(struct driver_data *drv_data);
int (*read)(struct driver_data *drv_data);
irqreturn_t (*transfer_handler)(struct driver_data *drv_data);
void (*cs_control)(u32 command);
};
struct chip_data {
u32 cr0;
u32 cr1;
u32 psp;
u32 timeout;
u8 n_bytes;
u32 dma_width;
u32 dma_burst_size;
u32 threshold;
u32 dma_threshold;
u8 enable_dma;
u8 bits_per_word;
u32 speed_hz;
int (*write)(struct driver_data *drv_data);
int (*read)(struct driver_data *drv_data);
void (*cs_control)(u32 command);
};
static void pump_messages(struct work_struct *work);
static int flush(struct driver_data *drv_data)
{
unsigned long limit = loops_per_jiffy << 1;
void __iomem *reg = drv_data->ioaddr;
do {
while (read_SSSR(reg) & SSSR_RNE) {
read_SSDR(reg);
}
} while ((read_SSSR(reg) & SSSR_BSY) && limit--);
write_SSSR(SSSR_ROR, reg);
return limit;
}
static void null_cs_control(u32 command)
{
}
static int null_writer(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
u8 n_bytes = drv_data->n_bytes;
if (((read_SSSR(reg) & 0x00000f00) == 0x00000f00)
|| (drv_data->tx == drv_data->tx_end))
return 0;
write_SSDR(0, reg);
drv_data->tx += n_bytes;
return 1;
}
static int null_reader(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
u8 n_bytes = drv_data->n_bytes;
while ((read_SSSR(reg) & SSSR_RNE)
&& (drv_data->rx < drv_data->rx_end)) {
read_SSDR(reg);
drv_data->rx += n_bytes;
}
return drv_data->rx == drv_data->rx_end;
}
static int u8_writer(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
if (((read_SSSR(reg) & 0x00000f00) == 0x00000f00)
|| (drv_data->tx == drv_data->tx_end))
return 0;
write_SSDR(*(u8 *)(drv_data->tx), reg);
++drv_data->tx;
return 1;
}
static int u8_reader(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
while ((read_SSSR(reg) & SSSR_RNE)
&& (drv_data->rx < drv_data->rx_end)) {
*(u8 *)(drv_data->rx) = read_SSDR(reg);
++drv_data->rx;
}
return drv_data->rx == drv_data->rx_end;
}
static int u16_writer(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
if (((read_SSSR(reg) & 0x00000f00) == 0x00000f00)
|| (drv_data->tx == drv_data->tx_end))
return 0;
write_SSDR(*(u16 *)(drv_data->tx), reg);
drv_data->tx += 2;
return 1;
}
static int u16_reader(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
while ((read_SSSR(reg) & SSSR_RNE)
&& (drv_data->rx < drv_data->rx_end)) {
*(u16 *)(drv_data->rx) = read_SSDR(reg);
drv_data->rx += 2;
}
return drv_data->rx == drv_data->rx_end;
}
static int u32_writer(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
if (((read_SSSR(reg) & 0x00000f00) == 0x00000f00)
|| (drv_data->tx == drv_data->tx_end))
return 0;
write_SSDR(*(u32 *)(drv_data->tx), reg);
drv_data->tx += 4;
return 1;
}
static int u32_reader(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
while ((read_SSSR(reg) & SSSR_RNE)
&& (drv_data->rx < drv_data->rx_end)) {
*(u32 *)(drv_data->rx) = read_SSDR(reg);
drv_data->rx += 4;
}
return drv_data->rx == drv_data->rx_end;
}
static void *next_transfer(struct driver_data *drv_data)
{
struct spi_message *msg = drv_data->cur_msg;
struct spi_transfer *trans = drv_data->cur_transfer;
/* Move to next transfer */
if (trans->transfer_list.next != &msg->transfers) {
drv_data->cur_transfer =
list_entry(trans->transfer_list.next,
struct spi_transfer,
transfer_list);
return RUNNING_STATE;
} else
return DONE_STATE;
}
static int map_dma_buffers(struct driver_data *drv_data)
{
struct spi_message *msg = drv_data->cur_msg;
struct device *dev = &msg->spi->dev;
if (!drv_data->cur_chip->enable_dma)
return 0;
if (msg->is_dma_mapped)
return drv_data->rx_dma && drv_data->tx_dma;
if (!IS_DMA_ALIGNED(drv_data->rx) || !IS_DMA_ALIGNED(drv_data->tx))
return 0;
/* Modify setup if rx buffer is null */
if (drv_data->rx == NULL) {
*drv_data->null_dma_buf = 0;
drv_data->rx = drv_data->null_dma_buf;
drv_data->rx_map_len = 4;
} else
drv_data->rx_map_len = drv_data->len;
/* Modify setup if tx buffer is null */
if (drv_data->tx == NULL) {
*drv_data->null_dma_buf = 0;
drv_data->tx = drv_data->null_dma_buf;
drv_data->tx_map_len = 4;
} else
drv_data->tx_map_len = drv_data->len;
/* Stream map the rx buffer */
drv_data->rx_dma = dma_map_single(dev, drv_data->rx,
drv_data->rx_map_len,
DMA_FROM_DEVICE);
dma-mapping: add the device argument to dma_mapping_error() Add per-device dma_mapping_ops support for CONFIG_X86_64 as POWER architecture does: This enables us to cleanly fix the Calgary IOMMU issue that some devices are not behind the IOMMU (http://lkml.org/lkml/2008/5/8/423). I think that per-device dma_mapping_ops support would be also helpful for KVM people to support PCI passthrough but Andi thinks that this makes it difficult to support the PCI passthrough (see the above thread). So I CC'ed this to KVM camp. Comments are appreciated. A pointer to dma_mapping_ops to struct dev_archdata is added. If the pointer is non NULL, DMA operations in asm/dma-mapping.h use it. If it's NULL, the system-wide dma_ops pointer is used as before. If it's useful for KVM people, I plan to implement a mechanism to register a hook called when a new pci (or dma capable) device is created (it works with hot plugging). It enables IOMMUs to set up an appropriate dma_mapping_ops per device. The major obstacle is that dma_mapping_error doesn't take a pointer to the device unlike other DMA operations. So x86 can't have dma_mapping_ops per device. Note all the POWER IOMMUs use the same dma_mapping_error function so this is not a problem for POWER but x86 IOMMUs use different dma_mapping_error functions. The first patch adds the device argument to dma_mapping_error. The patch is trivial but large since it touches lots of drivers and dma-mapping.h in all the architecture. This patch: dma_mapping_error() doesn't take a pointer to the device unlike other DMA operations. So we can't have dma_mapping_ops per device. Note that POWER already has dma_mapping_ops per device but all the POWER IOMMUs use the same dma_mapping_error function. x86 IOMMUs use device argument. [akpm@linux-foundation.org: fix sge] [akpm@linux-foundation.org: fix svc_rdma] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: fix bnx2x] [akpm@linux-foundation.org: fix s2io] [akpm@linux-foundation.org: fix pasemi_mac] [akpm@linux-foundation.org: fix sdhci] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: fix sparc] [akpm@linux-foundation.org: fix ibmvscsi] Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Muli Ben-Yehuda <muli@il.ibm.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Avi Kivity <avi@qumranet.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-26 10:44:49 +08:00
if (dma_mapping_error(dev, drv_data->rx_dma))
return 0;
/* Stream map the tx buffer */
drv_data->tx_dma = dma_map_single(dev, drv_data->tx,
drv_data->tx_map_len,
DMA_TO_DEVICE);
dma-mapping: add the device argument to dma_mapping_error() Add per-device dma_mapping_ops support for CONFIG_X86_64 as POWER architecture does: This enables us to cleanly fix the Calgary IOMMU issue that some devices are not behind the IOMMU (http://lkml.org/lkml/2008/5/8/423). I think that per-device dma_mapping_ops support would be also helpful for KVM people to support PCI passthrough but Andi thinks that this makes it difficult to support the PCI passthrough (see the above thread). So I CC'ed this to KVM camp. Comments are appreciated. A pointer to dma_mapping_ops to struct dev_archdata is added. If the pointer is non NULL, DMA operations in asm/dma-mapping.h use it. If it's NULL, the system-wide dma_ops pointer is used as before. If it's useful for KVM people, I plan to implement a mechanism to register a hook called when a new pci (or dma capable) device is created (it works with hot plugging). It enables IOMMUs to set up an appropriate dma_mapping_ops per device. The major obstacle is that dma_mapping_error doesn't take a pointer to the device unlike other DMA operations. So x86 can't have dma_mapping_ops per device. Note all the POWER IOMMUs use the same dma_mapping_error function so this is not a problem for POWER but x86 IOMMUs use different dma_mapping_error functions. The first patch adds the device argument to dma_mapping_error. The patch is trivial but large since it touches lots of drivers and dma-mapping.h in all the architecture. This patch: dma_mapping_error() doesn't take a pointer to the device unlike other DMA operations. So we can't have dma_mapping_ops per device. Note that POWER already has dma_mapping_ops per device but all the POWER IOMMUs use the same dma_mapping_error function. x86 IOMMUs use device argument. [akpm@linux-foundation.org: fix sge] [akpm@linux-foundation.org: fix svc_rdma] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: fix bnx2x] [akpm@linux-foundation.org: fix s2io] [akpm@linux-foundation.org: fix pasemi_mac] [akpm@linux-foundation.org: fix sdhci] [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: fix sparc] [akpm@linux-foundation.org: fix ibmvscsi] Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Muli Ben-Yehuda <muli@il.ibm.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Avi Kivity <avi@qumranet.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-26 10:44:49 +08:00
if (dma_mapping_error(dev, drv_data->tx_dma)) {
dma_unmap_single(dev, drv_data->rx_dma,
drv_data->rx_map_len, DMA_FROM_DEVICE);
return 0;
}
return 1;
}
static void unmap_dma_buffers(struct driver_data *drv_data)
{
struct device *dev;
if (!drv_data->dma_mapped)
return;
if (!drv_data->cur_msg->is_dma_mapped) {
dev = &drv_data->cur_msg->spi->dev;
dma_unmap_single(dev, drv_data->rx_dma,
drv_data->rx_map_len, DMA_FROM_DEVICE);
dma_unmap_single(dev, drv_data->tx_dma,
drv_data->tx_map_len, DMA_TO_DEVICE);
}
drv_data->dma_mapped = 0;
}
/* caller already set message->status; dma and pio irqs are blocked */
static void giveback(struct driver_data *drv_data)
{
struct spi_transfer* last_transfer;
unsigned long flags;
struct spi_message *msg;
spin_lock_irqsave(&drv_data->lock, flags);
msg = drv_data->cur_msg;
drv_data->cur_msg = NULL;
drv_data->cur_transfer = NULL;
drv_data->cur_chip = NULL;
queue_work(drv_data->workqueue, &drv_data->pump_messages);
spin_unlock_irqrestore(&drv_data->lock, flags);
last_transfer = list_entry(msg->transfers.prev,
struct spi_transfer,
transfer_list);
pxa2xx_spi: chipselect bugfixes Fixes several chipselect bugs in the pxa2xx_spi driver. These bugs are in all versions of this driver and prevent using it with chips like m25p16 flash. 1. The spi_transfer.cs_change flag is handled too early: before spi_transfer.delay_usecs applies, thus making the delay ineffective at holding chip select. 2. spi_transfer.delay_usecs is ignored on the last transfer of a message (likewise not holding chipselect long enough). 3. If spi_transfer.cs_change is set on the last transfer, the chip select is always disabled, instead of the intended meaning: optionally holding chip select enabled for the next message. Those first three bugs were fixed with a relocation of delays and chip select de-assertions. 4. If a message has the cs_change flag set on the last transfer, and had the chip select stayed enabled as requested (see 3, above), it would not have been disabled if the next message is for a different chip. Fixed by dropping chip select regardless of cs_change at end of a message, if there is no next message or if the next message is for a different chip. This patch should apply to all kernels back to and including 2.6.20; it was test patched against 2.6.20. An additional patch would be required for older kernels, but those versions are very buggy anyway. Signed-off-by: Ned Forrester <nforrester@whoi.edu> Cc: Vernon Sauder <vernoninhand@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-13 17:33:17 +08:00
/* Delay if requested before any change in chip select */
if (last_transfer->delay_usecs)
udelay(last_transfer->delay_usecs);
/* Drop chip select UNLESS cs_change is true or we are returning
* a message with an error, or next message is for another chip
*/
if (!last_transfer->cs_change)
drv_data->cs_control(PXA2XX_CS_DEASSERT);
pxa2xx_spi: chipselect bugfixes Fixes several chipselect bugs in the pxa2xx_spi driver. These bugs are in all versions of this driver and prevent using it with chips like m25p16 flash. 1. The spi_transfer.cs_change flag is handled too early: before spi_transfer.delay_usecs applies, thus making the delay ineffective at holding chip select. 2. spi_transfer.delay_usecs is ignored on the last transfer of a message (likewise not holding chipselect long enough). 3. If spi_transfer.cs_change is set on the last transfer, the chip select is always disabled, instead of the intended meaning: optionally holding chip select enabled for the next message. Those first three bugs were fixed with a relocation of delays and chip select de-assertions. 4. If a message has the cs_change flag set on the last transfer, and had the chip select stayed enabled as requested (see 3, above), it would not have been disabled if the next message is for a different chip. Fixed by dropping chip select regardless of cs_change at end of a message, if there is no next message or if the next message is for a different chip. This patch should apply to all kernels back to and including 2.6.20; it was test patched against 2.6.20. An additional patch would be required for older kernels, but those versions are very buggy anyway. Signed-off-by: Ned Forrester <nforrester@whoi.edu> Cc: Vernon Sauder <vernoninhand@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-13 17:33:17 +08:00
else {
struct spi_message *next_msg;
/* Holding of cs was hinted, but we need to make sure
* the next message is for the same chip. Don't waste
* time with the following tests unless this was hinted.
*
* We cannot postpone this until pump_messages, because
* after calling msg->complete (below) the driver that
* sent the current message could be unloaded, which
* could invalidate the cs_control() callback...
*/
/* get a pointer to the next message, if any */
spin_lock_irqsave(&drv_data->lock, flags);
if (list_empty(&drv_data->queue))
next_msg = NULL;
else
next_msg = list_entry(drv_data->queue.next,
struct spi_message, queue);
spin_unlock_irqrestore(&drv_data->lock, flags);
/* see if the next and current messages point
* to the same chip
*/
if (next_msg && next_msg->spi != msg->spi)
next_msg = NULL;
if (!next_msg || msg->state == ERROR_STATE)
drv_data->cs_control(PXA2XX_CS_DEASSERT);
}
msg->state = NULL;
if (msg->complete)
msg->complete(msg->context);
}
static int wait_ssp_rx_stall(void const __iomem *ioaddr)
{
unsigned long limit = loops_per_jiffy << 1;
while ((read_SSSR(ioaddr) & SSSR_BSY) && limit--)
cpu_relax();
return limit;
}
static int wait_dma_channel_stop(int channel)
{
unsigned long limit = loops_per_jiffy << 1;
while (!(DCSR(channel) & DCSR_STOPSTATE) && limit--)
cpu_relax();
return limit;
}
static void dma_error_stop(struct driver_data *drv_data, const char *msg)
{
void __iomem *reg = drv_data->ioaddr;
/* Stop and reset */
DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
write_SSSR(drv_data->clear_sr, reg);
write_SSCR1(read_SSCR1(reg) & ~drv_data->dma_cr1, reg);
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(0, reg);
flush(drv_data);
write_SSCR0(read_SSCR0(reg) & ~SSCR0_SSE, reg);
unmap_dma_buffers(drv_data);
dev_err(&drv_data->pdev->dev, "%s\n", msg);
drv_data->cur_msg->state = ERROR_STATE;
tasklet_schedule(&drv_data->pump_transfers);
}
static void dma_transfer_complete(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
struct spi_message *msg = drv_data->cur_msg;
/* Clear and disable interrupts on SSP and DMA channels*/
write_SSCR1(read_SSCR1(reg) & ~drv_data->dma_cr1, reg);
write_SSSR(drv_data->clear_sr, reg);
DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
if (wait_dma_channel_stop(drv_data->rx_channel) == 0)
dev_err(&drv_data->pdev->dev,
"dma_handler: dma rx channel stop failed\n");
if (wait_ssp_rx_stall(drv_data->ioaddr) == 0)
dev_err(&drv_data->pdev->dev,
"dma_transfer: ssp rx stall failed\n");
unmap_dma_buffers(drv_data);
/* update the buffer pointer for the amount completed in dma */
drv_data->rx += drv_data->len -
(DCMD(drv_data->rx_channel) & DCMD_LENGTH);
/* read trailing data from fifo, it does not matter how many
* bytes are in the fifo just read until buffer is full
* or fifo is empty, which ever occurs first */
drv_data->read(drv_data);
/* return count of what was actually read */
msg->actual_length += drv_data->len -
(drv_data->rx_end - drv_data->rx);
pxa2xx_spi: chipselect bugfixes Fixes several chipselect bugs in the pxa2xx_spi driver. These bugs are in all versions of this driver and prevent using it with chips like m25p16 flash. 1. The spi_transfer.cs_change flag is handled too early: before spi_transfer.delay_usecs applies, thus making the delay ineffective at holding chip select. 2. spi_transfer.delay_usecs is ignored on the last transfer of a message (likewise not holding chipselect long enough). 3. If spi_transfer.cs_change is set on the last transfer, the chip select is always disabled, instead of the intended meaning: optionally holding chip select enabled for the next message. Those first three bugs were fixed with a relocation of delays and chip select de-assertions. 4. If a message has the cs_change flag set on the last transfer, and had the chip select stayed enabled as requested (see 3, above), it would not have been disabled if the next message is for a different chip. Fixed by dropping chip select regardless of cs_change at end of a message, if there is no next message or if the next message is for a different chip. This patch should apply to all kernels back to and including 2.6.20; it was test patched against 2.6.20. An additional patch would be required for older kernels, but those versions are very buggy anyway. Signed-off-by: Ned Forrester <nforrester@whoi.edu> Cc: Vernon Sauder <vernoninhand@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-13 17:33:17 +08:00
/* Transfer delays and chip select release are
* handled in pump_transfers or giveback
*/
/* Move to next transfer */
msg->state = next_transfer(drv_data);
/* Schedule transfer tasklet */
tasklet_schedule(&drv_data->pump_transfers);
}
static void dma_handler(int channel, void *data)
{
struct driver_data *drv_data = data;
u32 irq_status = DCSR(channel) & DMA_INT_MASK;
if (irq_status & DCSR_BUSERR) {
if (channel == drv_data->tx_channel)
dma_error_stop(drv_data,
"dma_handler: "
"bad bus address on tx channel");
else
dma_error_stop(drv_data,
"dma_handler: "
"bad bus address on rx channel");
return;
}
/* PXA255x_SSP has no timeout interrupt, wait for tailing bytes */
if ((channel == drv_data->tx_channel)
&& (irq_status & DCSR_ENDINTR)
&& (drv_data->ssp_type == PXA25x_SSP)) {
/* Wait for rx to stall */
if (wait_ssp_rx_stall(drv_data->ioaddr) == 0)
dev_err(&drv_data->pdev->dev,
"dma_handler: ssp rx stall failed\n");
/* finish this transfer, start the next */
dma_transfer_complete(drv_data);
}
}
static irqreturn_t dma_transfer(struct driver_data *drv_data)
{
u32 irq_status;
void __iomem *reg = drv_data->ioaddr;
irq_status = read_SSSR(reg) & drv_data->mask_sr;
if (irq_status & SSSR_ROR) {
dma_error_stop(drv_data, "dma_transfer: fifo overrun");
return IRQ_HANDLED;
}
/* Check for false positive timeout */
if ((irq_status & SSSR_TINT)
&& (DCSR(drv_data->tx_channel) & DCSR_RUN)) {
write_SSSR(SSSR_TINT, reg);
return IRQ_HANDLED;
}
if (irq_status & SSSR_TINT || drv_data->rx == drv_data->rx_end) {
/* Clear and disable timeout interrupt, do the rest in
* dma_transfer_complete */
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(0, reg);
/* finish this transfer, start the next */
dma_transfer_complete(drv_data);
return IRQ_HANDLED;
}
/* Opps problem detected */
return IRQ_NONE;
}
static void int_error_stop(struct driver_data *drv_data, const char* msg)
{
void __iomem *reg = drv_data->ioaddr;
/* Stop and reset SSP */
write_SSSR(drv_data->clear_sr, reg);
write_SSCR1(read_SSCR1(reg) & ~drv_data->int_cr1, reg);
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(0, reg);
flush(drv_data);
write_SSCR0(read_SSCR0(reg) & ~SSCR0_SSE, reg);
dev_err(&drv_data->pdev->dev, "%s\n", msg);
drv_data->cur_msg->state = ERROR_STATE;
tasklet_schedule(&drv_data->pump_transfers);
}
static void int_transfer_complete(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
/* Stop SSP */
write_SSSR(drv_data->clear_sr, reg);
write_SSCR1(read_SSCR1(reg) & ~drv_data->int_cr1, reg);
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(0, reg);
/* Update total byte transfered return count actual bytes read */
drv_data->cur_msg->actual_length += drv_data->len -
(drv_data->rx_end - drv_data->rx);
pxa2xx_spi: chipselect bugfixes Fixes several chipselect bugs in the pxa2xx_spi driver. These bugs are in all versions of this driver and prevent using it with chips like m25p16 flash. 1. The spi_transfer.cs_change flag is handled too early: before spi_transfer.delay_usecs applies, thus making the delay ineffective at holding chip select. 2. spi_transfer.delay_usecs is ignored on the last transfer of a message (likewise not holding chipselect long enough). 3. If spi_transfer.cs_change is set on the last transfer, the chip select is always disabled, instead of the intended meaning: optionally holding chip select enabled for the next message. Those first three bugs were fixed with a relocation of delays and chip select de-assertions. 4. If a message has the cs_change flag set on the last transfer, and had the chip select stayed enabled as requested (see 3, above), it would not have been disabled if the next message is for a different chip. Fixed by dropping chip select regardless of cs_change at end of a message, if there is no next message or if the next message is for a different chip. This patch should apply to all kernels back to and including 2.6.20; it was test patched against 2.6.20. An additional patch would be required for older kernels, but those versions are very buggy anyway. Signed-off-by: Ned Forrester <nforrester@whoi.edu> Cc: Vernon Sauder <vernoninhand@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-13 17:33:17 +08:00
/* Transfer delays and chip select release are
* handled in pump_transfers or giveback
*/
/* Move to next transfer */
drv_data->cur_msg->state = next_transfer(drv_data);
/* Schedule transfer tasklet */
tasklet_schedule(&drv_data->pump_transfers);
}
static irqreturn_t interrupt_transfer(struct driver_data *drv_data)
{
void __iomem *reg = drv_data->ioaddr;
u32 irq_mask = (read_SSCR1(reg) & SSCR1_TIE) ?
drv_data->mask_sr : drv_data->mask_sr & ~SSSR_TFS;
u32 irq_status = read_SSSR(reg) & irq_mask;
if (irq_status & SSSR_ROR) {
int_error_stop(drv_data, "interrupt_transfer: fifo overrun");
return IRQ_HANDLED;
}
if (irq_status & SSSR_TINT) {
write_SSSR(SSSR_TINT, reg);
if (drv_data->read(drv_data)) {
int_transfer_complete(drv_data);
return IRQ_HANDLED;
}
}
/* Drain rx fifo, Fill tx fifo and prevent overruns */
do {
if (drv_data->read(drv_data)) {
int_transfer_complete(drv_data);
return IRQ_HANDLED;
}
} while (drv_data->write(drv_data));
if (drv_data->read(drv_data)) {
int_transfer_complete(drv_data);
return IRQ_HANDLED;
}
if (drv_data->tx == drv_data->tx_end) {
write_SSCR1(read_SSCR1(reg) & ~SSCR1_TIE, reg);
/* PXA25x_SSP has no timeout, read trailing bytes */
if (drv_data->ssp_type == PXA25x_SSP) {
if (!wait_ssp_rx_stall(reg))
{
int_error_stop(drv_data, "interrupt_transfer: "
"rx stall failed");
return IRQ_HANDLED;
}
if (!drv_data->read(drv_data))
{
int_error_stop(drv_data,
"interrupt_transfer: "
"trailing byte read failed");
return IRQ_HANDLED;
}
int_transfer_complete(drv_data);
}
}
/* We did something */
return IRQ_HANDLED;
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t ssp_int(int irq, void *dev_id)
{
struct driver_data *drv_data = dev_id;
void __iomem *reg = drv_data->ioaddr;
if (!drv_data->cur_msg) {
write_SSCR0(read_SSCR0(reg) & ~SSCR0_SSE, reg);
write_SSCR1(read_SSCR1(reg) & ~drv_data->int_cr1, reg);
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(0, reg);
write_SSSR(drv_data->clear_sr, reg);
dev_err(&drv_data->pdev->dev, "bad message state "
"in interrupt handler\n");
/* Never fail */
return IRQ_HANDLED;
}
return drv_data->transfer_handler(drv_data);
}
static int set_dma_burst_and_threshold(struct chip_data *chip,
struct spi_device *spi,
u8 bits_per_word, u32 *burst_code,
u32 *threshold)
{
struct pxa2xx_spi_chip *chip_info =
(struct pxa2xx_spi_chip *)spi->controller_data;
int bytes_per_word;
int burst_bytes;
int thresh_words;
int req_burst_size;
int retval = 0;
/* Set the threshold (in registers) to equal the same amount of data
* as represented by burst size (in bytes). The computation below
* is (burst_size rounded up to nearest 8 byte, word or long word)
* divided by (bytes/register); the tx threshold is the inverse of
* the rx, so that there will always be enough data in the rx fifo
* to satisfy a burst, and there will always be enough space in the
* tx fifo to accept a burst (a tx burst will overwrite the fifo if
* there is not enough space), there must always remain enough empty
* space in the rx fifo for any data loaded to the tx fifo.
* Whenever burst_size (in bytes) equals bits/word, the fifo threshold
* will be 8, or half the fifo;
* The threshold can only be set to 2, 4 or 8, but not 16, because
* to burst 16 to the tx fifo, the fifo would have to be empty;
* however, the minimum fifo trigger level is 1, and the tx will
* request service when the fifo is at this level, with only 15 spaces.
*/
/* find bytes/word */
if (bits_per_word <= 8)
bytes_per_word = 1;
else if (bits_per_word <= 16)
bytes_per_word = 2;
else
bytes_per_word = 4;
/* use struct pxa2xx_spi_chip->dma_burst_size if available */
if (chip_info)
req_burst_size = chip_info->dma_burst_size;
else {
switch (chip->dma_burst_size) {
default:
/* if the default burst size is not set,
* do it now */
chip->dma_burst_size = DCMD_BURST8;
case DCMD_BURST8:
req_burst_size = 8;
break;
case DCMD_BURST16:
req_burst_size = 16;
break;
case DCMD_BURST32:
req_burst_size = 32;
break;
}
}
if (req_burst_size <= 8) {
*burst_code = DCMD_BURST8;
burst_bytes = 8;
} else if (req_burst_size <= 16) {
if (bytes_per_word == 1) {
/* don't burst more than 1/2 the fifo */
*burst_code = DCMD_BURST8;
burst_bytes = 8;
retval = 1;
} else {
*burst_code = DCMD_BURST16;
burst_bytes = 16;
}
} else {
if (bytes_per_word == 1) {
/* don't burst more than 1/2 the fifo */
*burst_code = DCMD_BURST8;
burst_bytes = 8;
retval = 1;
} else if (bytes_per_word == 2) {
/* don't burst more than 1/2 the fifo */
*burst_code = DCMD_BURST16;
burst_bytes = 16;
retval = 1;
} else {
*burst_code = DCMD_BURST32;
burst_bytes = 32;
}
}
thresh_words = burst_bytes / bytes_per_word;
/* thresh_words will be between 2 and 8 */
*threshold = (SSCR1_RxTresh(thresh_words) & SSCR1_RFT)
| (SSCR1_TxTresh(16-thresh_words) & SSCR1_TFT);
return retval;
}
static unsigned int ssp_get_clk_div(struct ssp_device *ssp, int rate)
{
unsigned long ssp_clk = clk_get_rate(ssp->clk);
if (ssp->type == PXA25x_SSP)
return ((ssp_clk / (2 * rate) - 1) & 0xff) << 8;
else
return ((ssp_clk / rate - 1) & 0xfff) << 8;
}
static void pump_transfers(unsigned long data)
{
struct driver_data *drv_data = (struct driver_data *)data;
struct spi_message *message = NULL;
struct spi_transfer *transfer = NULL;
struct spi_transfer *previous = NULL;
struct chip_data *chip = NULL;
struct ssp_device *ssp = drv_data->ssp;
void __iomem *reg = drv_data->ioaddr;
u32 clk_div = 0;
u8 bits = 0;
u32 speed = 0;
u32 cr0;
u32 cr1;
u32 dma_thresh = drv_data->cur_chip->dma_threshold;
u32 dma_burst = drv_data->cur_chip->dma_burst_size;
/* Get current state information */
message = drv_data->cur_msg;
transfer = drv_data->cur_transfer;
chip = drv_data->cur_chip;
/* Handle for abort */
if (message->state == ERROR_STATE) {
message->status = -EIO;
giveback(drv_data);
return;
}
/* Handle end of message */
if (message->state == DONE_STATE) {
message->status = 0;
giveback(drv_data);
return;
}
pxa2xx_spi: chipselect bugfixes Fixes several chipselect bugs in the pxa2xx_spi driver. These bugs are in all versions of this driver and prevent using it with chips like m25p16 flash. 1. The spi_transfer.cs_change flag is handled too early: before spi_transfer.delay_usecs applies, thus making the delay ineffective at holding chip select. 2. spi_transfer.delay_usecs is ignored on the last transfer of a message (likewise not holding chipselect long enough). 3. If spi_transfer.cs_change is set on the last transfer, the chip select is always disabled, instead of the intended meaning: optionally holding chip select enabled for the next message. Those first three bugs were fixed with a relocation of delays and chip select de-assertions. 4. If a message has the cs_change flag set on the last transfer, and had the chip select stayed enabled as requested (see 3, above), it would not have been disabled if the next message is for a different chip. Fixed by dropping chip select regardless of cs_change at end of a message, if there is no next message or if the next message is for a different chip. This patch should apply to all kernels back to and including 2.6.20; it was test patched against 2.6.20. An additional patch would be required for older kernels, but those versions are very buggy anyway. Signed-off-by: Ned Forrester <nforrester@whoi.edu> Cc: Vernon Sauder <vernoninhand@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-13 17:33:17 +08:00
/* Delay if requested at end of transfer before CS change */
if (message->state == RUNNING_STATE) {
previous = list_entry(transfer->transfer_list.prev,
struct spi_transfer,
transfer_list);
if (previous->delay_usecs)
udelay(previous->delay_usecs);
pxa2xx_spi: chipselect bugfixes Fixes several chipselect bugs in the pxa2xx_spi driver. These bugs are in all versions of this driver and prevent using it with chips like m25p16 flash. 1. The spi_transfer.cs_change flag is handled too early: before spi_transfer.delay_usecs applies, thus making the delay ineffective at holding chip select. 2. spi_transfer.delay_usecs is ignored on the last transfer of a message (likewise not holding chipselect long enough). 3. If spi_transfer.cs_change is set on the last transfer, the chip select is always disabled, instead of the intended meaning: optionally holding chip select enabled for the next message. Those first three bugs were fixed with a relocation of delays and chip select de-assertions. 4. If a message has the cs_change flag set on the last transfer, and had the chip select stayed enabled as requested (see 3, above), it would not have been disabled if the next message is for a different chip. Fixed by dropping chip select regardless of cs_change at end of a message, if there is no next message or if the next message is for a different chip. This patch should apply to all kernels back to and including 2.6.20; it was test patched against 2.6.20. An additional patch would be required for older kernels, but those versions are very buggy anyway. Signed-off-by: Ned Forrester <nforrester@whoi.edu> Cc: Vernon Sauder <vernoninhand@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-09-13 17:33:17 +08:00
/* Drop chip select only if cs_change is requested */
if (previous->cs_change)
drv_data->cs_control(PXA2XX_CS_DEASSERT);
}
/* Check for transfers that need multiple DMA segments */
if (transfer->len > MAX_DMA_LEN && chip->enable_dma) {
/* reject already-mapped transfers; PIO won't always work */
if (message->is_dma_mapped
|| transfer->rx_dma || transfer->tx_dma) {
dev_err(&drv_data->pdev->dev,
"pump_transfers: mapped transfer length "
"of %u is greater than %d\n",
transfer->len, MAX_DMA_LEN);
message->status = -EINVAL;
giveback(drv_data);
return;
}
/* warn ... we force this to PIO mode */
if (printk_ratelimit())
dev_warn(&message->spi->dev, "pump_transfers: "
"DMA disabled for transfer length %ld "
"greater than %d\n",
(long)drv_data->len, MAX_DMA_LEN);
}
/* Setup the transfer state based on the type of transfer */
if (flush(drv_data) == 0) {
dev_err(&drv_data->pdev->dev, "pump_transfers: flush failed\n");
message->status = -EIO;
giveback(drv_data);
return;
}
drv_data->n_bytes = chip->n_bytes;
drv_data->dma_width = chip->dma_width;
drv_data->cs_control = chip->cs_control;
drv_data->tx = (void *)transfer->tx_buf;
drv_data->tx_end = drv_data->tx + transfer->len;
drv_data->rx = transfer->rx_buf;
drv_data->rx_end = drv_data->rx + transfer->len;
drv_data->rx_dma = transfer->rx_dma;
drv_data->tx_dma = transfer->tx_dma;
drv_data->len = transfer->len & DCMD_LENGTH;
drv_data->write = drv_data->tx ? chip->write : null_writer;
drv_data->read = drv_data->rx ? chip->read : null_reader;
/* Change speed and bit per word on a per transfer */
cr0 = chip->cr0;
if (transfer->speed_hz || transfer->bits_per_word) {
bits = chip->bits_per_word;
speed = chip->speed_hz;
if (transfer->speed_hz)
speed = transfer->speed_hz;
if (transfer->bits_per_word)
bits = transfer->bits_per_word;
clk_div = ssp_get_clk_div(ssp, speed);
if (bits <= 8) {
drv_data->n_bytes = 1;
drv_data->dma_width = DCMD_WIDTH1;
drv_data->read = drv_data->read != null_reader ?
u8_reader : null_reader;
drv_data->write = drv_data->write != null_writer ?
u8_writer : null_writer;
} else if (bits <= 16) {
drv_data->n_bytes = 2;
drv_data->dma_width = DCMD_WIDTH2;
drv_data->read = drv_data->read != null_reader ?
u16_reader : null_reader;
drv_data->write = drv_data->write != null_writer ?
u16_writer : null_writer;
} else if (bits <= 32) {
drv_data->n_bytes = 4;
drv_data->dma_width = DCMD_WIDTH4;
drv_data->read = drv_data->read != null_reader ?
u32_reader : null_reader;
drv_data->write = drv_data->write != null_writer ?
u32_writer : null_writer;
}
/* if bits/word is changed in dma mode, then must check the
* thresholds and burst also */
if (chip->enable_dma) {
if (set_dma_burst_and_threshold(chip, message->spi,
bits, &dma_burst,
&dma_thresh))
if (printk_ratelimit())
dev_warn(&message->spi->dev,
"pump_transfers: "
"DMA burst size reduced to "
"match bits_per_word\n");
}
cr0 = clk_div
| SSCR0_Motorola
| SSCR0_DataSize(bits > 16 ? bits - 16 : bits)
| SSCR0_SSE
| (bits > 16 ? SSCR0_EDSS : 0);
}
message->state = RUNNING_STATE;
/* Try to map dma buffer and do a dma transfer if successful, but
* only if the length is non-zero and less than MAX_DMA_LEN.
*
* Zero-length non-descriptor DMA is illegal on PXA2xx; force use
* of PIO instead. Care is needed above because the transfer may
* have have been passed with buffers that are already dma mapped.
* A zero-length transfer in PIO mode will not try to write/read
* to/from the buffers
*
* REVISIT large transfers are exactly where we most want to be
* using DMA. If this happens much, split those transfers into
* multiple DMA segments rather than forcing PIO.
*/
drv_data->dma_mapped = 0;
if (drv_data->len > 0 && drv_data->len <= MAX_DMA_LEN)
drv_data->dma_mapped = map_dma_buffers(drv_data);
if (drv_data->dma_mapped) {
/* Ensure we have the correct interrupt handler */
drv_data->transfer_handler = dma_transfer;
/* Setup rx DMA Channel */
DCSR(drv_data->rx_channel) = RESET_DMA_CHANNEL;
DSADR(drv_data->rx_channel) = drv_data->ssdr_physical;
DTADR(drv_data->rx_channel) = drv_data->rx_dma;
if (drv_data->rx == drv_data->null_dma_buf)
/* No target address increment */
DCMD(drv_data->rx_channel) = DCMD_FLOWSRC
| drv_data->dma_width
| dma_burst
| drv_data->len;
else
DCMD(drv_data->rx_channel) = DCMD_INCTRGADDR
| DCMD_FLOWSRC
| drv_data->dma_width
| dma_burst
| drv_data->len;
/* Setup tx DMA Channel */
DCSR(drv_data->tx_channel) = RESET_DMA_CHANNEL;
DSADR(drv_data->tx_channel) = drv_data->tx_dma;
DTADR(drv_data->tx_channel) = drv_data->ssdr_physical;
if (drv_data->tx == drv_data->null_dma_buf)
/* No source address increment */
DCMD(drv_data->tx_channel) = DCMD_FLOWTRG
| drv_data->dma_width
| dma_burst
| drv_data->len;
else
DCMD(drv_data->tx_channel) = DCMD_INCSRCADDR
| DCMD_FLOWTRG
| drv_data->dma_width
| dma_burst
| drv_data->len;
/* Enable dma end irqs on SSP to detect end of transfer */
if (drv_data->ssp_type == PXA25x_SSP)
DCMD(drv_data->tx_channel) |= DCMD_ENDIRQEN;
/* Clear status and start DMA engine */
cr1 = chip->cr1 | dma_thresh | drv_data->dma_cr1;
write_SSSR(drv_data->clear_sr, reg);
DCSR(drv_data->rx_channel) |= DCSR_RUN;
DCSR(drv_data->tx_channel) |= DCSR_RUN;
} else {
/* Ensure we have the correct interrupt handler */
drv_data->transfer_handler = interrupt_transfer;
/* Clear status */
cr1 = chip->cr1 | chip->threshold | drv_data->int_cr1;
write_SSSR(drv_data->clear_sr, reg);
}
/* see if we need to reload the config registers */
if ((read_SSCR0(reg) != cr0)
|| (read_SSCR1(reg) & SSCR1_CHANGE_MASK) !=
(cr1 & SSCR1_CHANGE_MASK)) {
/* stop the SSP, and update the other bits */
write_SSCR0(cr0 & ~SSCR0_SSE, reg);
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(chip->timeout, reg);
/* first set CR1 without interrupt and service enables */
write_SSCR1(cr1 & SSCR1_CHANGE_MASK, reg);
/* restart the SSP */
write_SSCR0(cr0, reg);
} else {
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(chip->timeout, reg);
}
/* FIXME, need to handle cs polarity,
* this driver uses struct pxa2xx_spi_chip.cs_control to
* specify a CS handling function, and it ignores most
* struct spi_device.mode[s], including SPI_CS_HIGH */
drv_data->cs_control(PXA2XX_CS_ASSERT);
/* after chip select, release the data by enabling service
* requests and interrupts, without changing any mode bits */
write_SSCR1(cr1, reg);
}
static void pump_messages(struct work_struct *work)
{
struct driver_data *drv_data =
container_of(work, struct driver_data, pump_messages);
unsigned long flags;
/* Lock queue and check for queue work */
spin_lock_irqsave(&drv_data->lock, flags);
if (list_empty(&drv_data->queue) || drv_data->run == QUEUE_STOPPED) {
drv_data->busy = 0;
spin_unlock_irqrestore(&drv_data->lock, flags);
return;
}
/* Make sure we are not already running a message */
if (drv_data->cur_msg) {
spin_unlock_irqrestore(&drv_data->lock, flags);
return;
}
/* Extract head of queue */
drv_data->cur_msg = list_entry(drv_data->queue.next,
struct spi_message, queue);
list_del_init(&drv_data->cur_msg->queue);
/* Initial message state*/
drv_data->cur_msg->state = START_STATE;
drv_data->cur_transfer = list_entry(drv_data->cur_msg->transfers.next,
struct spi_transfer,
transfer_list);
/* prepare to setup the SSP, in pump_transfers, using the per
* chip configuration */
drv_data->cur_chip = spi_get_ctldata(drv_data->cur_msg->spi);
/* Mark as busy and launch transfers */
tasklet_schedule(&drv_data->pump_transfers);
drv_data->busy = 1;
spin_unlock_irqrestore(&drv_data->lock, flags);
}
static int transfer(struct spi_device *spi, struct spi_message *msg)
{
struct driver_data *drv_data = spi_master_get_devdata(spi->master);
unsigned long flags;
spin_lock_irqsave(&drv_data->lock, flags);
if (drv_data->run == QUEUE_STOPPED) {
spin_unlock_irqrestore(&drv_data->lock, flags);
return -ESHUTDOWN;
}
msg->actual_length = 0;
msg->status = -EINPROGRESS;
msg->state = START_STATE;
list_add_tail(&msg->queue, &drv_data->queue);
if (drv_data->run == QUEUE_RUNNING && !drv_data->busy)
queue_work(drv_data->workqueue, &drv_data->pump_messages);
spin_unlock_irqrestore(&drv_data->lock, flags);
return 0;
}
/* the spi->mode bits understood by this driver: */
#define MODEBITS (SPI_CPOL | SPI_CPHA)
static int setup(struct spi_device *spi)
{
struct pxa2xx_spi_chip *chip_info = NULL;
struct chip_data *chip;
struct driver_data *drv_data = spi_master_get_devdata(spi->master);
struct ssp_device *ssp = drv_data->ssp;
unsigned int clk_div;
if (!spi->bits_per_word)
spi->bits_per_word = 8;
if (drv_data->ssp_type != PXA25x_SSP
&& (spi->bits_per_word < 4 || spi->bits_per_word > 32)) {
dev_err(&spi->dev, "failed setup: ssp_type=%d, bits/wrd=%d "
"b/w not 4-32 for type non-PXA25x_SSP\n",
drv_data->ssp_type, spi->bits_per_word);
return -EINVAL;
}
else if (drv_data->ssp_type == PXA25x_SSP
&& (spi->bits_per_word < 4
|| spi->bits_per_word > 16)) {
dev_err(&spi->dev, "failed setup: ssp_type=%d, bits/wrd=%d "
"b/w not 4-16 for type PXA25x_SSP\n",
drv_data->ssp_type, spi->bits_per_word);
return -EINVAL;
}
if (spi->mode & ~MODEBITS) {
dev_dbg(&spi->dev, "setup: unsupported mode bits %x\n",
spi->mode & ~MODEBITS);
return -EINVAL;
}
/* Only alloc on first setup */
chip = spi_get_ctldata(spi);
if (!chip) {
chip = kzalloc(sizeof(struct chip_data), GFP_KERNEL);
if (!chip) {
dev_err(&spi->dev,
"failed setup: can't allocate chip data\n");
return -ENOMEM;
}
chip->cs_control = null_cs_control;
chip->enable_dma = 0;
chip->timeout = 1000;
chip->threshold = SSCR1_RxTresh(1) | SSCR1_TxTresh(1);
chip->dma_burst_size = drv_data->master_info->enable_dma ?
DCMD_BURST8 : 0;
}
/* protocol drivers may change the chip settings, so...
* if chip_info exists, use it */
chip_info = spi->controller_data;
/* chip_info isn't always needed */
chip->cr1 = 0;
if (chip_info) {
if (chip_info->cs_control)
chip->cs_control = chip_info->cs_control;
chip->timeout = chip_info->timeout;
chip->threshold = (SSCR1_RxTresh(chip_info->rx_threshold) &
SSCR1_RFT) |
(SSCR1_TxTresh(chip_info->tx_threshold) &
SSCR1_TFT);
chip->enable_dma = chip_info->dma_burst_size != 0
&& drv_data->master_info->enable_dma;
chip->dma_threshold = 0;
if (chip_info->enable_loopback)
chip->cr1 = SSCR1_LBM;
}
/* set dma burst and threshold outside of chip_info path so that if
* chip_info goes away after setting chip->enable_dma, the
* burst and threshold can still respond to changes in bits_per_word */
if (chip->enable_dma) {
/* set up legal burst and threshold for dma */
if (set_dma_burst_and_threshold(chip, spi, spi->bits_per_word,
&chip->dma_burst_size,
&chip->dma_threshold)) {
dev_warn(&spi->dev, "in setup: DMA burst size reduced "
"to match bits_per_word\n");
}
}
clk_div = ssp_get_clk_div(ssp, spi->max_speed_hz);
chip->speed_hz = spi->max_speed_hz;
chip->cr0 = clk_div
| SSCR0_Motorola
| SSCR0_DataSize(spi->bits_per_word > 16 ?
spi->bits_per_word - 16 : spi->bits_per_word)
| SSCR0_SSE
| (spi->bits_per_word > 16 ? SSCR0_EDSS : 0);
chip->cr1 &= ~(SSCR1_SPO | SSCR1_SPH);
chip->cr1 |= (((spi->mode & SPI_CPHA) != 0) ? SSCR1_SPH : 0)
| (((spi->mode & SPI_CPOL) != 0) ? SSCR1_SPO : 0);
/* NOTE: PXA25x_SSP _could_ use external clocking ... */
if (drv_data->ssp_type != PXA25x_SSP)
dev_dbg(&spi->dev, "%d bits/word, %ld Hz, mode %d\n",
spi->bits_per_word,
clk_get_rate(ssp->clk)
/ (1 + ((chip->cr0 & SSCR0_SCR) >> 8)),
spi->mode & 0x3);
else
dev_dbg(&spi->dev, "%d bits/word, %ld Hz, mode %d\n",
spi->bits_per_word,
clk_get_rate(ssp->clk)
/ (1 + ((chip->cr0 & SSCR0_SCR) >> 8)),
spi->mode & 0x3);
if (spi->bits_per_word <= 8) {
chip->n_bytes = 1;
chip->dma_width = DCMD_WIDTH1;
chip->read = u8_reader;
chip->write = u8_writer;
} else if (spi->bits_per_word <= 16) {
chip->n_bytes = 2;
chip->dma_width = DCMD_WIDTH2;
chip->read = u16_reader;
chip->write = u16_writer;
} else if (spi->bits_per_word <= 32) {
chip->cr0 |= SSCR0_EDSS;
chip->n_bytes = 4;
chip->dma_width = DCMD_WIDTH4;
chip->read = u32_reader;
chip->write = u32_writer;
} else {
dev_err(&spi->dev, "invalid wordsize\n");
return -ENODEV;
}
chip->bits_per_word = spi->bits_per_word;
spi_set_ctldata(spi, chip);
return 0;
}
static void cleanup(struct spi_device *spi)
{
struct chip_data *chip = spi_get_ctldata(spi);
kfree(chip);
}
static int __init init_queue(struct driver_data *drv_data)
{
INIT_LIST_HEAD(&drv_data->queue);
spin_lock_init(&drv_data->lock);
drv_data->run = QUEUE_STOPPED;
drv_data->busy = 0;
tasklet_init(&drv_data->pump_transfers,
pump_transfers, (unsigned long)drv_data);
INIT_WORK(&drv_data->pump_messages, pump_messages);
drv_data->workqueue = create_singlethread_workqueue(
drv_data->master->dev.parent->bus_id);
if (drv_data->workqueue == NULL)
return -EBUSY;
return 0;
}
static int start_queue(struct driver_data *drv_data)
{
unsigned long flags;
spin_lock_irqsave(&drv_data->lock, flags);
if (drv_data->run == QUEUE_RUNNING || drv_data->busy) {
spin_unlock_irqrestore(&drv_data->lock, flags);
return -EBUSY;
}
drv_data->run = QUEUE_RUNNING;
drv_data->cur_msg = NULL;
drv_data->cur_transfer = NULL;
drv_data->cur_chip = NULL;
spin_unlock_irqrestore(&drv_data->lock, flags);
queue_work(drv_data->workqueue, &drv_data->pump_messages);
return 0;
}
static int stop_queue(struct driver_data *drv_data)
{
unsigned long flags;
unsigned limit = 500;
int status = 0;
spin_lock_irqsave(&drv_data->lock, flags);
/* This is a bit lame, but is optimized for the common execution path.
* A wait_queue on the drv_data->busy could be used, but then the common
* execution path (pump_messages) would be required to call wake_up or
* friends on every SPI message. Do this instead */
drv_data->run = QUEUE_STOPPED;
while (!list_empty(&drv_data->queue) && drv_data->busy && limit--) {
spin_unlock_irqrestore(&drv_data->lock, flags);
msleep(10);
spin_lock_irqsave(&drv_data->lock, flags);
}
if (!list_empty(&drv_data->queue) || drv_data->busy)
status = -EBUSY;
spin_unlock_irqrestore(&drv_data->lock, flags);
return status;
}
static int destroy_queue(struct driver_data *drv_data)
{
int status;
status = stop_queue(drv_data);
/* we are unloading the module or failing to load (only two calls
* to this routine), and neither call can handle a return value.
* However, destroy_workqueue calls flush_workqueue, and that will
* block until all work is done. If the reason that stop_queue
* timed out is that the work will never finish, then it does no
* good to call destroy_workqueue, so return anyway. */
if (status != 0)
return status;
destroy_workqueue(drv_data->workqueue);
return 0;
}
static int __init pxa2xx_spi_probe(struct platform_device *pdev)
{
struct device *dev = &pdev->dev;
struct pxa2xx_spi_master *platform_info;
struct spi_master *master;
struct driver_data *drv_data = NULL;
struct ssp_device *ssp;
int status = 0;
platform_info = dev->platform_data;
ssp = ssp_request(pdev->id, pdev->name);
if (ssp == NULL) {
dev_err(&pdev->dev, "failed to request SSP%d\n", pdev->id);
return -ENODEV;
}
/* Allocate master with space for drv_data and null dma buffer */
master = spi_alloc_master(dev, sizeof(struct driver_data) + 16);
if (!master) {
dev_err(&pdev->dev, "can not alloc spi_master\n");
ssp_free(ssp);
return -ENOMEM;
}
drv_data = spi_master_get_devdata(master);
drv_data->master = master;
drv_data->master_info = platform_info;
drv_data->pdev = pdev;
drv_data->ssp = ssp;
master->bus_num = pdev->id;
master->num_chipselect = platform_info->num_chipselect;
master->cleanup = cleanup;
master->setup = setup;
master->transfer = transfer;
drv_data->ssp_type = ssp->type;
drv_data->null_dma_buf = (u32 *)ALIGN((u32)(drv_data +
sizeof(struct driver_data)), 8);
drv_data->ioaddr = ssp->mmio_base;
drv_data->ssdr_physical = ssp->phys_base + SSDR;
if (ssp->type == PXA25x_SSP) {
drv_data->int_cr1 = SSCR1_TIE | SSCR1_RIE;
drv_data->dma_cr1 = 0;
drv_data->clear_sr = SSSR_ROR;
drv_data->mask_sr = SSSR_RFS | SSSR_TFS | SSSR_ROR;
} else {
drv_data->int_cr1 = SSCR1_TIE | SSCR1_RIE | SSCR1_TINTE;
drv_data->dma_cr1 = SSCR1_TSRE | SSCR1_RSRE | SSCR1_TINTE;
drv_data->clear_sr = SSSR_ROR | SSSR_TINT;
drv_data->mask_sr = SSSR_TINT | SSSR_RFS | SSSR_TFS | SSSR_ROR;
}
status = request_irq(ssp->irq, ssp_int, 0, dev->bus_id, drv_data);
if (status < 0) {
dev_err(&pdev->dev, "can not get IRQ\n");
goto out_error_master_alloc;
}
/* Setup DMA if requested */
drv_data->tx_channel = -1;
drv_data->rx_channel = -1;
if (platform_info->enable_dma) {
/* Get two DMA channels (rx and tx) */
drv_data->rx_channel = pxa_request_dma("pxa2xx_spi_ssp_rx",
DMA_PRIO_HIGH,
dma_handler,
drv_data);
if (drv_data->rx_channel < 0) {
dev_err(dev, "problem (%d) requesting rx channel\n",
drv_data->rx_channel);
status = -ENODEV;
goto out_error_irq_alloc;
}
drv_data->tx_channel = pxa_request_dma("pxa2xx_spi_ssp_tx",
DMA_PRIO_MEDIUM,
dma_handler,
drv_data);
if (drv_data->tx_channel < 0) {
dev_err(dev, "problem (%d) requesting tx channel\n",
drv_data->tx_channel);
status = -ENODEV;
goto out_error_dma_alloc;
}
DRCMR(ssp->drcmr_rx) = DRCMR_MAPVLD | drv_data->rx_channel;
DRCMR(ssp->drcmr_tx) = DRCMR_MAPVLD | drv_data->tx_channel;
}
/* Enable SOC clock */
clk_enable(ssp->clk);
/* Load default SSP configuration */
write_SSCR0(0, drv_data->ioaddr);
write_SSCR1(SSCR1_RxTresh(4) | SSCR1_TxTresh(12), drv_data->ioaddr);
write_SSCR0(SSCR0_SerClkDiv(2)
| SSCR0_Motorola
| SSCR0_DataSize(8),
drv_data->ioaddr);
if (drv_data->ssp_type != PXA25x_SSP)
write_SSTO(0, drv_data->ioaddr);
write_SSPSP(0, drv_data->ioaddr);
/* Initial and start queue */
status = init_queue(drv_data);
if (status != 0) {
dev_err(&pdev->dev, "problem initializing queue\n");
goto out_error_clock_enabled;
}
status = start_queue(drv_data);
if (status != 0) {
dev_err(&pdev->dev, "problem starting queue\n");
goto out_error_clock_enabled;
}
/* Register with the SPI framework */
platform_set_drvdata(pdev, drv_data);
status = spi_register_master(master);
if (status != 0) {
dev_err(&pdev->dev, "problem registering spi master\n");
goto out_error_queue_alloc;
}
return status;
out_error_queue_alloc:
destroy_queue(drv_data);
out_error_clock_enabled:
clk_disable(ssp->clk);
out_error_dma_alloc:
if (drv_data->tx_channel != -1)
pxa_free_dma(drv_data->tx_channel);
if (drv_data->rx_channel != -1)
pxa_free_dma(drv_data->rx_channel);
out_error_irq_alloc:
free_irq(ssp->irq, drv_data);
out_error_master_alloc:
spi_master_put(master);
ssp_free(ssp);
return status;
}
static int pxa2xx_spi_remove(struct platform_device *pdev)
{
struct driver_data *drv_data = platform_get_drvdata(pdev);
struct ssp_device *ssp = drv_data->ssp;
int status = 0;
if (!drv_data)
return 0;
/* Remove the queue */
status = destroy_queue(drv_data);
if (status != 0)
/* the kernel does not check the return status of this
* this routine (mod->exit, within the kernel). Therefore
* nothing is gained by returning from here, the module is
* going away regardless, and we should not leave any more
* resources allocated than necessary. We cannot free the
* message memory in drv_data->queue, but we can release the
* resources below. I think the kernel should honor -EBUSY
* returns but... */
dev_err(&pdev->dev, "pxa2xx_spi_remove: workqueue will not "
"complete, message memory not freed\n");
/* Disable the SSP at the peripheral and SOC level */
write_SSCR0(0, drv_data->ioaddr);
clk_disable(ssp->clk);
/* Release DMA */
if (drv_data->master_info->enable_dma) {
DRCMR(ssp->drcmr_rx) = 0;
DRCMR(ssp->drcmr_tx) = 0;
pxa_free_dma(drv_data->tx_channel);
pxa_free_dma(drv_data->rx_channel);
}
/* Release IRQ */
free_irq(ssp->irq, drv_data);
/* Release SSP */
ssp_free(ssp);
/* Disconnect from the SPI framework */
spi_unregister_master(drv_data->master);
/* Prevent double remove */
platform_set_drvdata(pdev, NULL);
return 0;
}
static void pxa2xx_spi_shutdown(struct platform_device *pdev)
{
int status = 0;
if ((status = pxa2xx_spi_remove(pdev)) != 0)
dev_err(&pdev->dev, "shutdown failed with %d\n", status);
}
#ifdef CONFIG_PM
static int pxa2xx_spi_suspend(struct platform_device *pdev, pm_message_t state)
{
struct driver_data *drv_data = platform_get_drvdata(pdev);
struct ssp_device *ssp = drv_data->ssp;
int status = 0;
status = stop_queue(drv_data);
if (status != 0)
return status;
write_SSCR0(0, drv_data->ioaddr);
clk_disable(ssp->clk);
return 0;
}
static int pxa2xx_spi_resume(struct platform_device *pdev)
{
struct driver_data *drv_data = platform_get_drvdata(pdev);
struct ssp_device *ssp = drv_data->ssp;
int status = 0;
/* Enable the SSP clock */
clk_enable(ssp->clk);
/* Start the queue running */
status = start_queue(drv_data);
if (status != 0) {
dev_err(&pdev->dev, "problem starting queue (%d)\n", status);
return status;
}
return 0;
}
#else
#define pxa2xx_spi_suspend NULL
#define pxa2xx_spi_resume NULL
#endif /* CONFIG_PM */
static struct platform_driver driver = {
.driver = {
.name = "pxa2xx-spi",
.owner = THIS_MODULE,
},
.remove = pxa2xx_spi_remove,
.shutdown = pxa2xx_spi_shutdown,
.suspend = pxa2xx_spi_suspend,
.resume = pxa2xx_spi_resume,
};
static int __init pxa2xx_spi_init(void)
{
return platform_driver_probe(&driver, pxa2xx_spi_probe);
}
module_init(pxa2xx_spi_init);
static void __exit pxa2xx_spi_exit(void)
{
platform_driver_unregister(&driver);
}
module_exit(pxa2xx_spi_exit);