spi: reorganize drivers

Sort the SPI makefile and enforce the naming convention spi_*.c for
spi drivers.

This change also rolls the contents of atmel_spi.h into the .c file
since there is only one user of that particular include file.

v2: - Use 'spi-' prefix instead of 'spi_' to match what seems to be
      be the predominant pattern for subsystem prefixes.
    - Clean up filenames in Kconfig and header comment blocks

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Acked-by: Linus Walleij <linus.walleij@linaro.org>

1 files changed, 2342 insertions, 0 deletions

diff --git a/drivers/spi/spi-pl022.c b/drivers/spi/spi-pl022.c
new file mode 100644
index 000000000000..25417054a456
--- /dev/null
+++ b/drivers/spi/spi-pl022.c
@@ -0,0 +1,2342 @@
+/*
+ * A driver for the ARM PL022 PrimeCell SSP/SPI bus master.
+ *
+ * Copyright (C) 2008-2009 ST-Ericsson AB
+ * Copyright (C) 2006 STMicroelectronics Pvt. Ltd.
+ *
+ * Author: Linus Walleij <linus.walleij@stericsson.com>
+ *
+ * Initial version inspired by:
+ *	linux-2.6.17-rc3-mm1/drivers/spi/pxa2xx_spi.c
+ * Initial adoption to PL022 by:
+ *      Sachin Verma <sachin.verma@st.com>
+ *
+ * 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.
+ */
+
+#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/spi/spi.h>
+#include <linux/workqueue.h>
+#include <linux/delay.h>
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/amba/bus.h>
+#include <linux/amba/pl022.h>
+#include <linux/io.h>
+#include <linux/slab.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/scatterlist.h>
+
+/*
+ * This macro is used to define some register default values.
+ * reg is masked with mask, the OR:ed with an (again masked)
+ * val shifted sb steps to the left.
+ */
+#define SSP_WRITE_BITS(reg, val, mask, sb) \
+ ((reg) = (((reg) & ~(mask)) | (((val)<<(sb)) & (mask))))
+
+/*
+ * This macro is also used to define some default values.
+ * It will just shift val by sb steps to the left and mask
+ * the result with mask.
+ */
+#define GEN_MASK_BITS(val, mask, sb) \
+ (((val)<<(sb)) & (mask))
+
+#define DRIVE_TX		0
+#define DO_NOT_DRIVE_TX		1
+
+#define DO_NOT_QUEUE_DMA	0
+#define QUEUE_DMA		1
+
+#define RX_TRANSFER		1
+#define TX_TRANSFER		2
+
+/*
+ * Macros to access SSP Registers with their offsets
+ */
+#define SSP_CR0(r)	(r + 0x000)
+#define SSP_CR1(r)	(r + 0x004)
+#define SSP_DR(r)	(r + 0x008)
+#define SSP_SR(r)	(r + 0x00C)
+#define SSP_CPSR(r)	(r + 0x010)
+#define SSP_IMSC(r)	(r + 0x014)
+#define SSP_RIS(r)	(r + 0x018)
+#define SSP_MIS(r)	(r + 0x01C)
+#define SSP_ICR(r)	(r + 0x020)
+#define SSP_DMACR(r)	(r + 0x024)
+#define SSP_ITCR(r)	(r + 0x080)
+#define SSP_ITIP(r)	(r + 0x084)
+#define SSP_ITOP(r)	(r + 0x088)
+#define SSP_TDR(r)	(r + 0x08C)
+
+#define SSP_PID0(r)	(r + 0xFE0)
+#define SSP_PID1(r)	(r + 0xFE4)
+#define SSP_PID2(r)	(r + 0xFE8)
+#define SSP_PID3(r)	(r + 0xFEC)
+
+#define SSP_CID0(r)	(r + 0xFF0)
+#define SSP_CID1(r)	(r + 0xFF4)
+#define SSP_CID2(r)	(r + 0xFF8)
+#define SSP_CID3(r)	(r + 0xFFC)
+
+/*
+ * SSP Control Register 0  - SSP_CR0
+ */
+#define SSP_CR0_MASK_DSS	(0x0FUL << 0)
+#define SSP_CR0_MASK_FRF	(0x3UL << 4)
+#define SSP_CR0_MASK_SPO	(0x1UL << 6)
+#define SSP_CR0_MASK_SPH	(0x1UL << 7)
+#define SSP_CR0_MASK_SCR	(0xFFUL << 8)
+
+/*
+ * The ST version of this block moves som bits
+ * in SSP_CR0 and extends it to 32 bits
+ */
+#define SSP_CR0_MASK_DSS_ST	(0x1FUL << 0)
+#define SSP_CR0_MASK_HALFDUP_ST	(0x1UL << 5)
+#define SSP_CR0_MASK_CSS_ST	(0x1FUL << 16)
+#define SSP_CR0_MASK_FRF_ST	(0x3UL << 21)
+
+
+/*
+ * SSP Control Register 0  - SSP_CR1
+ */
+#define SSP_CR1_MASK_LBM	(0x1UL << 0)
+#define SSP_CR1_MASK_SSE	(0x1UL << 1)
+#define SSP_CR1_MASK_MS		(0x1UL << 2)
+#define SSP_CR1_MASK_SOD	(0x1UL << 3)
+
+/*
+ * The ST version of this block adds some bits
+ * in SSP_CR1
+ */
+#define SSP_CR1_MASK_RENDN_ST	(0x1UL << 4)
+#define SSP_CR1_MASK_TENDN_ST	(0x1UL << 5)
+#define SSP_CR1_MASK_MWAIT_ST	(0x1UL << 6)
+#define SSP_CR1_MASK_RXIFLSEL_ST (0x7UL << 7)
+#define SSP_CR1_MASK_TXIFLSEL_ST (0x7UL << 10)
+/* This one is only in the PL023 variant */
+#define SSP_CR1_MASK_FBCLKDEL_ST (0x7UL << 13)
+
+/*
+ * SSP Status Register - SSP_SR
+ */
+#define SSP_SR_MASK_TFE		(0x1UL << 0) /* Transmit FIFO empty */
+#define SSP_SR_MASK_TNF		(0x1UL << 1) /* Transmit FIFO not full */
+#define SSP_SR_MASK_RNE		(0x1UL << 2) /* Receive FIFO not empty */
+#define SSP_SR_MASK_RFF		(0x1UL << 3) /* Receive FIFO full */
+#define SSP_SR_MASK_BSY		(0x1UL << 4) /* Busy Flag */
+
+/*
+ * SSP Clock Prescale Register  - SSP_CPSR
+ */
+#define SSP_CPSR_MASK_CPSDVSR	(0xFFUL << 0)
+
+/*
+ * SSP Interrupt Mask Set/Clear Register - SSP_IMSC
+ */
+#define SSP_IMSC_MASK_RORIM (0x1UL << 0) /* Receive Overrun Interrupt mask */
+#define SSP_IMSC_MASK_RTIM  (0x1UL << 1) /* Receive timeout Interrupt mask */
+#define SSP_IMSC_MASK_RXIM  (0x1UL << 2) /* Receive FIFO Interrupt mask */
+#define SSP_IMSC_MASK_TXIM  (0x1UL << 3) /* Transmit FIFO Interrupt mask */
+
+/*
+ * SSP Raw Interrupt Status Register - SSP_RIS
+ */
+/* Receive Overrun Raw Interrupt status */
+#define SSP_RIS_MASK_RORRIS		(0x1UL << 0)
+/* Receive Timeout Raw Interrupt status */
+#define SSP_RIS_MASK_RTRIS		(0x1UL << 1)
+/* Receive FIFO Raw Interrupt status */
+#define SSP_RIS_MASK_RXRIS		(0x1UL << 2)
+/* Transmit FIFO Raw Interrupt status */
+#define SSP_RIS_MASK_TXRIS		(0x1UL << 3)
+
+/*
+ * SSP Masked Interrupt Status Register - SSP_MIS
+ */
+/* Receive Overrun Masked Interrupt status */
+#define SSP_MIS_MASK_RORMIS		(0x1UL << 0)
+/* Receive Timeout Masked Interrupt status */
+#define SSP_MIS_MASK_RTMIS		(0x1UL << 1)
+/* Receive FIFO Masked Interrupt status */
+#define SSP_MIS_MASK_RXMIS		(0x1UL << 2)
+/* Transmit FIFO Masked Interrupt status */
+#define SSP_MIS_MASK_TXMIS		(0x1UL << 3)
+
+/*
+ * SSP Interrupt Clear Register - SSP_ICR
+ */
+/* Receive Overrun Raw Clear Interrupt bit */
+#define SSP_ICR_MASK_RORIC		(0x1UL << 0)
+/* Receive Timeout Clear Interrupt bit */
+#define SSP_ICR_MASK_RTIC		(0x1UL << 1)
+
+/*
+ * SSP DMA Control Register - SSP_DMACR
+ */
+/* Receive DMA Enable bit */
+#define SSP_DMACR_MASK_RXDMAE		(0x1UL << 0)
+/* Transmit DMA Enable bit */
+#define SSP_DMACR_MASK_TXDMAE		(0x1UL << 1)
+
+/*
+ * SSP Integration Test control Register - SSP_ITCR
+ */
+#define SSP_ITCR_MASK_ITEN		(0x1UL << 0)
+#define SSP_ITCR_MASK_TESTFIFO		(0x1UL << 1)
+
+/*
+ * SSP Integration Test Input Register - SSP_ITIP
+ */
+#define ITIP_MASK_SSPRXD		 (0x1UL << 0)
+#define ITIP_MASK_SSPFSSIN		 (0x1UL << 1)
+#define ITIP_MASK_SSPCLKIN		 (0x1UL << 2)
+#define ITIP_MASK_RXDMAC		 (0x1UL << 3)
+#define ITIP_MASK_TXDMAC		 (0x1UL << 4)
+#define ITIP_MASK_SSPTXDIN		 (0x1UL << 5)
+
+/*
+ * SSP Integration Test output Register - SSP_ITOP
+ */
+#define ITOP_MASK_SSPTXD		 (0x1UL << 0)
+#define ITOP_MASK_SSPFSSOUT		 (0x1UL << 1)
+#define ITOP_MASK_SSPCLKOUT		 (0x1UL << 2)
+#define ITOP_MASK_SSPOEn		 (0x1UL << 3)
+#define ITOP_MASK_SSPCTLOEn		 (0x1UL << 4)
+#define ITOP_MASK_RORINTR		 (0x1UL << 5)
+#define ITOP_MASK_RTINTR		 (0x1UL << 6)
+#define ITOP_MASK_RXINTR		 (0x1UL << 7)
+#define ITOP_MASK_TXINTR		 (0x1UL << 8)
+#define ITOP_MASK_INTR			 (0x1UL << 9)
+#define ITOP_MASK_RXDMABREQ		 (0x1UL << 10)
+#define ITOP_MASK_RXDMASREQ		 (0x1UL << 11)
+#define ITOP_MASK_TXDMABREQ		 (0x1UL << 12)
+#define ITOP_MASK_TXDMASREQ		 (0x1UL << 13)
+
+/*
+ * SSP Test Data Register - SSP_TDR
+ */
+#define TDR_MASK_TESTDATA		(0xFFFFFFFF)
+
+/*
+ * Message State
+ * we use the spi_message.state (void *) pointer to
+ * hold a single state value, that's why all this
+ * (void *) casting is done here.
+ */
+#define STATE_START			((void *) 0)
+#define STATE_RUNNING			((void *) 1)
+#define STATE_DONE			((void *) 2)
+#define STATE_ERROR			((void *) -1)
+
+/*
+ * SSP State - Whether Enabled or Disabled
+ */
+#define SSP_DISABLED			(0)
+#define SSP_ENABLED			(1)
+
+/*
+ * SSP DMA State - Whether DMA Enabled or Disabled
+ */
+#define SSP_DMA_DISABLED		(0)
+#define SSP_DMA_ENABLED			(1)
+
+/*
+ * SSP Clock Defaults
+ */
+#define SSP_DEFAULT_CLKRATE 0x2
+#define SSP_DEFAULT_PRESCALE 0x40
+
+/*
+ * SSP Clock Parameter ranges
+ */
+#define CPSDVR_MIN 0x02
+#define CPSDVR_MAX 0xFE
+#define SCR_MIN 0x00
+#define SCR_MAX 0xFF
+
+/*
+ * SSP Interrupt related Macros
+ */
+#define DEFAULT_SSP_REG_IMSC  0x0UL
+#define DISABLE_ALL_INTERRUPTS DEFAULT_SSP_REG_IMSC
+#define ENABLE_ALL_INTERRUPTS (~DEFAULT_SSP_REG_IMSC)
+
+#define CLEAR_ALL_INTERRUPTS  0x3
+
+#define SPI_POLLING_TIMEOUT 1000
+
+
+/*
+ * The type of reading going on on this chip
+ */
+enum ssp_reading {
+	READING_NULL,
+	READING_U8,
+	READING_U16,
+	READING_U32
+};
+
+/**
+ * The type of writing going on on this chip
+ */
+enum ssp_writing {
+	WRITING_NULL,
+	WRITING_U8,
+	WRITING_U16,
+	WRITING_U32
+};
+
+/**
+ * struct vendor_data - vendor-specific config parameters
+ * for PL022 derivates
+ * @fifodepth: depth of FIFOs (both)
+ * @max_bpw: maximum number of bits per word
+ * @unidir: supports unidirection transfers
+ * @extended_cr: 32 bit wide control register 0 with extra
+ * features and extra features in CR1 as found in the ST variants
+ * @pl023: supports a subset of the ST extensions called "PL023"
+ */
+struct vendor_data {
+	int fifodepth;
+	int max_bpw;
+	bool unidir;
+	bool extended_cr;
+	bool pl023;
+	bool loopback;
+};
+
+/**
+ * struct pl022 - This is the private SSP driver data structure
+ * @adev: AMBA device model hookup
+ * @vendor: vendor data for the IP block
+ * @phybase: the physical memory where the SSP device resides
+ * @virtbase: the virtual memory where the SSP is mapped
+ * @clk: outgoing clock "SPICLK" for the SPI bus
+ * @master: SPI framework hookup
+ * @master_info: controller-specific data from machine setup
+ * @workqueue: a workqueue on which any spi_message request is queued
+ * @pump_messages: work struct for scheduling work to the workqueue
+ * @queue_lock: spinlock to syncronise access to message queue
+ * @queue: message queue
+ * @busy: workqueue is busy
+ * @running: workqueue is running
+ * @pump_transfers: Tasklet used in Interrupt Transfer mode
+ * @cur_msg: Pointer to current spi_message being processed
+ * @cur_transfer: Pointer to current spi_transfer
+ * @cur_chip: pointer to current clients chip(assigned from controller_state)
+ * @tx: current position in TX buffer to be read
+ * @tx_end: end position in TX buffer to be read
+ * @rx: current position in RX buffer to be written
+ * @rx_end: end position in RX buffer to be written
+ * @read: the type of read currently going on
+ * @write: the type of write currently going on
+ * @exp_fifo_level: expected FIFO level
+ * @dma_rx_channel: optional channel for RX DMA
+ * @dma_tx_channel: optional channel for TX DMA
+ * @sgt_rx: scattertable for the RX transfer
+ * @sgt_tx: scattertable for the TX transfer
+ * @dummypage: a dummy page used for driving data on the bus with DMA
+ */
+struct pl022 {
+	struct amba_device		*adev;
+	struct vendor_data		*vendor;
+	resource_size_t			phybase;
+	void __iomem			*virtbase;
+	struct clk			*clk;
+	struct spi_master		*master;
+	struct pl022_ssp_controller	*master_info;
+	/* Driver message queue */
+	struct workqueue_struct		*workqueue;
+	struct work_struct		pump_messages;
+	spinlock_t			queue_lock;
+	struct list_head		queue;
+	bool				busy;
+	bool				running;
+	/* Message transfer pump */
+	struct tasklet_struct		pump_transfers;
+	struct spi_message		*cur_msg;
+	struct spi_transfer		*cur_transfer;
+	struct chip_data		*cur_chip;
+	void				*tx;
+	void				*tx_end;
+	void				*rx;
+	void				*rx_end;
+	enum ssp_reading		read;
+	enum ssp_writing		write;
+	u32				exp_fifo_level;
+	/* DMA settings */
+#ifdef CONFIG_DMA_ENGINE
+	struct dma_chan			*dma_rx_channel;
+	struct dma_chan			*dma_tx_channel;
+	struct sg_table			sgt_rx;
+	struct sg_table			sgt_tx;
+	char				*dummypage;
+#endif
+};
+
+/**
+ * struct chip_data - To maintain runtime state of SSP for each client chip
+ * @cr0: Value of control register CR0 of SSP - on later ST variants this
+ *       register is 32 bits wide rather than just 16
+ * @cr1: Value of control register CR1 of SSP
+ * @dmacr: Value of DMA control Register of SSP
+ * @cpsr: Value of Clock prescale register
+ * @n_bytes: how many bytes(power of 2) reqd for a given data width of client
+ * @enable_dma: Whether to enable DMA or not
+ * @read: function ptr to be used to read when doing xfer for this chip
+ * @write: function ptr to be used to write when doing xfer for this chip
+ * @cs_control: chip select callback provided by chip
+ * @xfer_type: polling/interrupt/DMA
+ *
+ * Runtime state of the SSP controller, maintained per chip,
+ * This would be set according to the current message that would be served
+ */
+struct chip_data {
+	u32 cr0;
+	u16 cr1;
+	u16 dmacr;
+	u16 cpsr;
+	u8 n_bytes;
+	bool enable_dma;
+	enum ssp_reading read;
+	enum ssp_writing write;
+	void (*cs_control) (u32 command);
+	int xfer_type;
+};
+
+/**
+ * null_cs_control - Dummy chip select function
+ * @command: select/delect the chip
+ *
+ * If no chip select function is provided by client this is used as dummy
+ * chip select
+ */
+static void null_cs_control(u32 command)
+{
+	pr_debug("pl022: dummy chip select control, CS=0x%x\n", command);
+}
+
+/**
+ * giveback - current spi_message is over, schedule next message and call
+ * callback of this message. Assumes that caller already
+ * set message->status; dma and pio irqs are blocked
+ * @pl022: SSP driver private data structure
+ */
+static void giveback(struct pl022 *pl022)
+{
+	struct spi_transfer *last_transfer;
+	unsigned long flags;
+	struct spi_message *msg;
+	void (*curr_cs_control) (u32 command);
+
+	/*
+	 * This local reference to the chip select function
+	 * is needed because we set curr_chip to NULL
+	 * as a step toward termininating the message.
+	 */
+	curr_cs_control = pl022->cur_chip->cs_control;
+	spin_lock_irqsave(&pl022->queue_lock, flags);
+	msg = pl022->cur_msg;
+	pl022->cur_msg = NULL;
+	pl022->cur_transfer = NULL;
+	pl022->cur_chip = NULL;
+	queue_work(pl022->workqueue, &pl022->pump_messages);
+	spin_unlock_irqrestore(&pl022->queue_lock, flags);
+
+	last_transfer = list_entry(msg->transfers.prev,
+					struct spi_transfer,
+					transfer_list);
+
+	/* Delay if requested before any change in chip select */
+	if (last_transfer->delay_usecs)
+		/*
+		 * FIXME: This runs in interrupt context.
+		 * Is this really smart?
+		 */
+		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)
+		curr_cs_control(SSP_CHIP_DESELECT);
+	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(&pl022->queue_lock, flags);
+		if (list_empty(&pl022->queue))
+			next_msg = NULL;
+		else
+			next_msg = list_entry(pl022->queue.next,
+					struct spi_message, queue);
+		spin_unlock_irqrestore(&pl022->queue_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 == STATE_ERROR)
+			curr_cs_control(SSP_CHIP_DESELECT);
+	}
+	msg->state = NULL;
+	if (msg->complete)
+		msg->complete(msg->context);
+	/* This message is completed, so let's turn off the clocks & power */
+	clk_disable(pl022->clk);
+	amba_pclk_disable(pl022->adev);
+	amba_vcore_disable(pl022->adev);
+}
+
+/**
+ * flush - flush the FIFO to reach a clean state
+ * @pl022: SSP driver private data structure
+ */
+static int flush(struct pl022 *pl022)
+{
+	unsigned long limit = loops_per_jiffy << 1;
+
+	dev_dbg(&pl022->adev->dev, "flush\n");
+	do {
+		while (readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RNE)
+			readw(SSP_DR(pl022->virtbase));
+	} while ((readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_BSY) && limit--);
+
+	pl022->exp_fifo_level = 0;
+
+	return limit;
+}
+
+/**
+ * restore_state - Load configuration of current chip
+ * @pl022: SSP driver private data structure
+ */
+static void restore_state(struct pl022 *pl022)
+{
+	struct chip_data *chip = pl022->cur_chip;
+
+	if (pl022->vendor->extended_cr)
+		writel(chip->cr0, SSP_CR0(pl022->virtbase));
+	else
+		writew(chip->cr0, SSP_CR0(pl022->virtbase));
+	writew(chip->cr1, SSP_CR1(pl022->virtbase));
+	writew(chip->dmacr, SSP_DMACR(pl022->virtbase));
+	writew(chip->cpsr, SSP_CPSR(pl022->virtbase));
+	writew(DISABLE_ALL_INTERRUPTS, SSP_IMSC(pl022->virtbase));
+	writew(CLEAR_ALL_INTERRUPTS, SSP_ICR(pl022->virtbase));
+}
+
+/*
+ * Default SSP Register Values
+ */
+#define DEFAULT_SSP_REG_CR0 ( \
+	GEN_MASK_BITS(SSP_DATA_BITS_12, SSP_CR0_MASK_DSS, 0)	| \
+	GEN_MASK_BITS(SSP_INTERFACE_MOTOROLA_SPI, SSP_CR0_MASK_FRF, 4) | \
+	GEN_MASK_BITS(SSP_CLK_POL_IDLE_LOW, SSP_CR0_MASK_SPO, 6) | \
+	GEN_MASK_BITS(SSP_CLK_SECOND_EDGE, SSP_CR0_MASK_SPH, 7) | \
+	GEN_MASK_BITS(SSP_DEFAULT_CLKRATE, SSP_CR0_MASK_SCR, 8) \
+)
+
+/* ST versions have slightly different bit layout */
+#define DEFAULT_SSP_REG_CR0_ST ( \
+	GEN_MASK_BITS(SSP_DATA_BITS_12, SSP_CR0_MASK_DSS_ST, 0)	| \
+	GEN_MASK_BITS(SSP_MICROWIRE_CHANNEL_FULL_DUPLEX, SSP_CR0_MASK_HALFDUP_ST, 5) | \
+	GEN_MASK_BITS(SSP_CLK_POL_IDLE_LOW, SSP_CR0_MASK_SPO, 6) | \
+	GEN_MASK_BITS(SSP_CLK_SECOND_EDGE, SSP_CR0_MASK_SPH, 7) | \
+	GEN_MASK_BITS(SSP_DEFAULT_CLKRATE, SSP_CR0_MASK_SCR, 8) | \
+	GEN_MASK_BITS(SSP_BITS_8, SSP_CR0_MASK_CSS_ST, 16)	| \
+	GEN_MASK_BITS(SSP_INTERFACE_MOTOROLA_SPI, SSP_CR0_MASK_FRF_ST, 21) \
+)
+
+/* The PL023 version is slightly different again */
+#define DEFAULT_SSP_REG_CR0_ST_PL023 ( \
+	GEN_MASK_BITS(SSP_DATA_BITS_12, SSP_CR0_MASK_DSS_ST, 0)	| \
+	GEN_MASK_BITS(SSP_CLK_POL_IDLE_LOW, SSP_CR0_MASK_SPO, 6) | \
+	GEN_MASK_BITS(SSP_CLK_SECOND_EDGE, SSP_CR0_MASK_SPH, 7) | \
+	GEN_MASK_BITS(SSP_DEFAULT_CLKRATE, SSP_CR0_MASK_SCR, 8) \
+)
+
+#define DEFAULT_SSP_REG_CR1 ( \
+	GEN_MASK_BITS(LOOPBACK_DISABLED, SSP_CR1_MASK_LBM, 0) | \
+	GEN_MASK_BITS(SSP_DISABLED, SSP_CR1_MASK_SSE, 1) | \
+	GEN_MASK_BITS(SSP_MASTER, SSP_CR1_MASK_MS, 2) | \
+	GEN_MASK_BITS(DO_NOT_DRIVE_TX, SSP_CR1_MASK_SOD, 3) \
+)
+
+/* ST versions extend this register to use all 16 bits */
+#define DEFAULT_SSP_REG_CR1_ST ( \
+	DEFAULT_SSP_REG_CR1 | \
+	GEN_MASK_BITS(SSP_RX_MSB, SSP_CR1_MASK_RENDN_ST, 4) | \
+	GEN_MASK_BITS(SSP_TX_MSB, SSP_CR1_MASK_TENDN_ST, 5) | \
+	GEN_MASK_BITS(SSP_MWIRE_WAIT_ZERO, SSP_CR1_MASK_MWAIT_ST, 6) |\
+	GEN_MASK_BITS(SSP_RX_1_OR_MORE_ELEM, SSP_CR1_MASK_RXIFLSEL_ST, 7) | \
+	GEN_MASK_BITS(SSP_TX_1_OR_MORE_EMPTY_LOC, SSP_CR1_MASK_TXIFLSEL_ST, 10) \
+)
+
+/*
+ * The PL023 variant has further differences: no loopback mode, no microwire
+ * support, and a new clock feedback delay setting.
+ */
+#define DEFAULT_SSP_REG_CR1_ST_PL023 ( \
+	GEN_MASK_BITS(SSP_DISABLED, SSP_CR1_MASK_SSE, 1) | \
+	GEN_MASK_BITS(SSP_MASTER, SSP_CR1_MASK_MS, 2) | \
+	GEN_MASK_BITS(DO_NOT_DRIVE_TX, SSP_CR1_MASK_SOD, 3) | \
+	GEN_MASK_BITS(SSP_RX_MSB, SSP_CR1_MASK_RENDN_ST, 4) | \
+	GEN_MASK_BITS(SSP_TX_MSB, SSP_CR1_MASK_TENDN_ST, 5) | \
+	GEN_MASK_BITS(SSP_RX_1_OR_MORE_ELEM, SSP_CR1_MASK_RXIFLSEL_ST, 7) | \
+	GEN_MASK_BITS(SSP_TX_1_OR_MORE_EMPTY_LOC, SSP_CR1_MASK_TXIFLSEL_ST, 10) | \
+	GEN_MASK_BITS(SSP_FEEDBACK_CLK_DELAY_NONE, SSP_CR1_MASK_FBCLKDEL_ST, 13) \
+)
+
+#define DEFAULT_SSP_REG_CPSR ( \
+	GEN_MASK_BITS(SSP_DEFAULT_PRESCALE, SSP_CPSR_MASK_CPSDVSR, 0) \
+)
+
+#define DEFAULT_SSP_REG_DMACR (\
+	GEN_MASK_BITS(SSP_DMA_DISABLED, SSP_DMACR_MASK_RXDMAE, 0) | \
+	GEN_MASK_BITS(SSP_DMA_DISABLED, SSP_DMACR_MASK_TXDMAE, 1) \
+)
+
+/**
+ * load_ssp_default_config - Load default configuration for SSP
+ * @pl022: SSP driver private data structure
+ */
+static void load_ssp_default_config(struct pl022 *pl022)
+{
+	if (pl022->vendor->pl023) {
+		writel(DEFAULT_SSP_REG_CR0_ST_PL023, SSP_CR0(pl022->virtbase));
+		writew(DEFAULT_SSP_REG_CR1_ST_PL023, SSP_CR1(pl022->virtbase));
+	} else if (pl022->vendor->extended_cr) {
+		writel(DEFAULT_SSP_REG_CR0_ST, SSP_CR0(pl022->virtbase));
+		writew(DEFAULT_SSP_REG_CR1_ST, SSP_CR1(pl022->virtbase));
+	} else {
+		writew(DEFAULT_SSP_REG_CR0, SSP_CR0(pl022->virtbase));
+		writew(DEFAULT_SSP_REG_CR1, SSP_CR1(pl022->virtbase));
+	}
+	writew(DEFAULT_SSP_REG_DMACR, SSP_DMACR(pl022->virtbase));
+	writew(DEFAULT_SSP_REG_CPSR, SSP_CPSR(pl022->virtbase));
+	writew(DISABLE_ALL_INTERRUPTS, SSP_IMSC(pl022->virtbase));
+	writew(CLEAR_ALL_INTERRUPTS, SSP_ICR(pl022->virtbase));
+}
+
+/**
+ * This will write to TX and read from RX according to the parameters
+ * set in pl022.
+ */
+static void readwriter(struct pl022 *pl022)
+{
+
+	/*
+	 * The FIFO depth is different between primecell variants.
+	 * I believe filling in too much in the FIFO might cause
+	 * errons in 8bit wide transfers on ARM variants (just 8 words
+	 * FIFO, means only 8x8 = 64 bits in FIFO) at least.
+	 *
+	 * To prevent this issue, the TX FIFO is only filled to the
+	 * unused RX FIFO fill length, regardless of what the TX
+	 * FIFO status flag indicates.
+	 */
+	dev_dbg(&pl022->adev->dev,
+		"%s, rx: %p, rxend: %p, tx: %p, txend: %p\n",
+		__func__, pl022->rx, pl022->rx_end, pl022->tx, pl022->tx_end);
+
+	/* Read as much as you can */
+	while ((readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RNE)
+	       && (pl022->rx < pl022->rx_end)) {
+		switch (pl022->read) {
+		case READING_NULL:
+			readw(SSP_DR(pl022->virtbase));
+			break;
+		case READING_U8:
+			*(u8 *) (pl022->rx) =
+				readw(SSP_DR(pl022->virtbase)) & 0xFFU;
+			break;
+		case READING_U16:
+			*(u16 *) (pl022->rx) =
+				(u16) readw(SSP_DR(pl022->virtbase));
+			break;
+		case READING_U32:
+			*(u32 *) (pl022->rx) =
+				readl(SSP_DR(pl022->virtbase));
+			break;
+		}
+		pl022->rx += (pl022->cur_chip->n_bytes);
+		pl022->exp_fifo_level--;
+	}
+	/*
+	 * Write as much as possible up to the RX FIFO size
+	 */
+	while ((pl022->exp_fifo_level < pl022->vendor->fifodepth)
+	       && (pl022->tx < pl022->tx_end)) {
+		switch (pl022->write) {
+		case WRITING_NULL:
+			writew(0x0, SSP_DR(pl022->virtbase));
+			break;
+		case WRITING_U8:
+			writew(*(u8 *) (pl022->tx), SSP_DR(pl022->virtbase));
+			break;
+		case WRITING_U16:
+			writew((*(u16 *) (pl022->tx)), SSP_DR(pl022->virtbase));
+			break;
+		case WRITING_U32:
+			writel(*(u32 *) (pl022->tx), SSP_DR(pl022->virtbase));
+			break;
+		}
+		pl022->tx += (pl022->cur_chip->n_bytes);
+		pl022->exp_fifo_level++;
+		/*
+		 * This inner reader takes care of things appearing in the RX
+		 * FIFO as we're transmitting. This will happen a lot since the
+		 * clock starts running when you put things into the TX FIFO,
+		 * and then things are continuously clocked into the RX FIFO.
+		 */
+		while ((readw(SSP_SR(pl022->virtbase)) & SSP_SR_MASK_RNE)
+		       && (pl022->rx < pl022->rx_end)) {
+			switch (pl022->read) {
+			case READING_NULL:
+				readw(SSP_DR(pl022->virtbase));
+				break;
+			case READING_U8:
+				*(u8 *) (pl022->rx) =
+					readw(SSP_DR(pl022->virtbase)) & 0xFFU;
+				break;
+			case READING_U16:
+				*(u16 *) (pl022->rx) =
+					(u16) readw(SSP_DR(pl022->virtbase));
+				break;
+			case READING_U32:
+				*(u32 *) (pl022->rx) =
+					readl(SSP_DR(pl022->virtbase));
+				break;
+			}
+			pl022->rx += (pl022->cur_chip->n_bytes);
+			pl022->exp_fifo_level--;
+		}
+	}
+	/*
+	 * When we exit here the TX FIFO should be full and the RX FIFO
+	 * should be empty
+	 */
+}
+
+
+/**
+ * next_transfer - Move to the Next transfer in the current spi message
+ * @pl022: SSP driver private data structure
+ *
+ * This function moves though the linked list of spi transfers in the
+ * current spi message and returns with the state of current spi
+ * message i.e whether its last transfer is done(STATE_DONE) or
+ * Next transfer is ready(STATE_RUNNING)
+ */
+static void *next_transfer(struct pl022 *pl022)
+{
+	struct spi_message *msg = pl022->cur_msg;
+	struct spi_transfer *trans = pl022->cur_transfer;
+
+	/* Move to next transfer */
+	if (trans->transfer_list.next != &msg->transfers) {
+		pl022->cur_transfer =
+		    list_entry(trans->transfer_list.next,
+			       struct spi_transfer, transfer_list);
+		return STATE_RUNNING;
+	}
+	return STATE_DONE;
+}
+
+/*
+ * This DMA functionality is only compiled in if we have
+ * access to the generic DMA devices/DMA engine.
+ */
+#ifdef CONFIG_DMA_ENGINE
+static void unmap_free_dma_scatter(struct pl022 *pl022)
+{
+	/* Unmap and free the SG tables */
+	dma_unmap_sg(pl022->dma_tx_channel->device->dev, pl022->sgt_tx.sgl,
+		     pl022->sgt_tx.nents, DMA_TO_DEVICE);
+	dma_unmap_sg(pl022->dma_rx_channel->device->dev, pl022->sgt_rx.sgl,
+		     pl022->sgt_rx.nents, DMA_FROM_DEVICE);
+	sg_free_table(&pl022->sgt_rx);
+	sg_free_table(&pl022->sgt_tx);
+}
+
+static void dma_callback(void *data)
+{
+	struct pl022 *pl022 = data;
+	struct spi_message *msg = pl022->cur_msg;
+
+	BUG_ON(!pl022->sgt_rx.sgl);
+
+#ifdef VERBOSE_DEBUG
+	/*
+	 * Optionally dump out buffers to inspect contents, this is
+	 * good if you want to convince yourself that the loopback
+	 * read/write contents are the same, when adopting to a new
+	 * DMA engine.
+	 */
+	{
+		struct scatterlist *sg;
+		unsigned int i;
+
+		dma_sync_sg_for_cpu(&pl022->adev->dev,
+				    pl022->sgt_rx.sgl,
+				    pl022->sgt_rx.nents,
+				    DMA_FROM_DEVICE);
+
+		for_each_sg(pl022->sgt_rx.sgl, sg, pl022->sgt_rx.nents, i) {
+			dev_dbg(&pl022->adev->dev, "SPI RX SG ENTRY: %d", i);
+			print_hex_dump(KERN_ERR, "SPI RX: ",
+				       DUMP_PREFIX_OFFSET,
+				       16,
+				       1,
+				       sg_virt(sg),
+				       sg_dma_len(sg),
+				       1);
+		}
+		for_each_sg(pl022->sgt_tx.sgl, sg, pl022->sgt_tx.nents, i) {
+			dev_dbg(&pl022->adev->dev, "SPI TX SG ENTRY: %d", i);
+			print_hex_dump(KERN_ERR, "SPI TX: ",
+				       DUMP_PREFIX_OFFSET,
+				       16,
+				       1,
+				       sg_virt(sg),
+				       sg_dma_len(sg),
+				       1);
+		}
+	}
+#endif
+
+	unmap_free_dma_scatter(pl022);
+
+	/* Update total bytes transferred */
+	msg->actual_length += pl022->cur_transfer->len;
+	if (pl022->cur_transfer->cs_change)
+		pl022->cur_chip->
+			cs_control(SSP_CHIP_DESELECT);
+
+	/* Move to next transfer */
+	msg->state = next_transfer(pl022);
+	tasklet_schedule(&pl022->pump_transfers);
+}
+
+static void setup_dma_scatter(struct pl022 *pl022,
+			      void *buffer,
+			      unsigned int length,
+			      struct sg_table *sgtab)
+{
+	struct scatterlist *sg;
+	int bytesleft = length;
+	void *bufp = buffer;
+	int mapbytes;
+	int i;
+
+	if (buffer) {
+		for_each_sg(sgtab->sgl, sg, sgtab->nents, i) {
+			/*
+			 * If there are less bytes left than what fits
+			 * in the current page (plus page alignment offset)
+			 * we just feed in this, else we stuff in as much
+			 * as we can.
+			 */
+			if (bytesleft < (PAGE_SIZE - offset_in_page(bufp)))
+				mapbytes = bytesleft;
+			else
+				mapbytes = PAGE_SIZE - offset_in_page(bufp);
+			sg_set_page(sg, virt_to_page(bufp),
+				    mapbytes, offset_in_page(bufp));
+			bufp += mapbytes;
+			bytesleft -= mapbytes;
+			dev_dbg(&pl022->adev->dev,
+				"set RX/TX target page @ %p, %d bytes, %d left\n",
+				bufp, mapbytes, bytesleft);
+		}
+	} else {
+		/* Map the dummy buffer on every page */
+		for_each_sg(sgtab->sgl, sg, sgtab->nents, i) {
+			if (bytesleft < PAGE_SIZE)
+				mapbytes = bytesleft;
+			else
+				mapbytes = PAGE_SIZE;
+			sg_set_page(sg, virt_to_page(pl022->dummypage),
+				    mapbytes, 0);
+			bytesleft -= mapbytes;
+			dev_dbg(&pl022->adev->dev,
+				"set RX/TX to dummy page %d bytes, %d left\n",
+				mapbytes, bytesleft);
+
+		}
+	}
+	BUG_ON(bytesleft);
+}
+
+/**
+ * configure_dma - configures the channels for the next transfer
+ * @pl022: SSP driver's private data structure
+ */
+static int configure_dma(struct pl022 *pl022)
+{
+	struct dma_slave_config rx_conf = {
+		.src_addr = SSP_DR(pl022->phybase),
+		.direction = DMA_FROM_DEVICE,
+		.src_maxburst = pl022->vendor->fifodepth >> 1,
+	};
+	struct dma_slave_config tx_conf = {
+		.dst_addr = SSP_DR(pl022->phybase),
+		.direction = DMA_TO_DEVICE,
+		.dst_maxburst = pl022->vendor->fifodepth >> 1,
+	};
+	unsigned int pages;
+	int ret;
+	int rx_sglen, tx_sglen;
+	struct dma_chan *rxchan = pl022->dma_rx_channel;
+	struct dma_chan *txchan = pl022->dma_tx_channel;
+	struct dma_async_tx_descriptor *rxdesc;
+	struct dma_async_tx_descriptor *txdesc;
+
+	/* Check that the channels are available */
+	if (!rxchan || !txchan)
+		return -ENODEV;
+
+	switch (pl022->read) {
+	case READING_NULL:
+		/* Use the same as for writing */
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
+		break;
+	case READING_U8:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case READING_U16:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case READING_U32:
+		rx_conf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	}
+
+	switch (pl022->write) {
+	case WRITING_NULL:
+		/* Use the same as for reading */
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_UNDEFINED;
+		break;
+	case WRITING_U8:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
+		break;
+	case WRITING_U16:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
+		break;
+	case WRITING_U32:
+		tx_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		break;
+	}
+
+	/* SPI pecularity: we need to read and write the same width */
+	if (rx_conf.src_addr_width == DMA_SLAVE_BUSWIDTH_UNDEFINED)
+		rx_conf.src_addr_width = tx_conf.dst_addr_width;
+	if (tx_conf.dst_addr_width == DMA_SLAVE_BUSWIDTH_UNDEFINED)
+		tx_conf.dst_addr_width = rx_conf.src_addr_width;
+	BUG_ON(rx_conf.src_addr_width != tx_conf.dst_addr_width);
+
+	dmaengine_slave_config(rxchan, &rx_conf);
+	dmaengine_slave_config(txchan, &tx_conf);
+
+	/* Create sglists for the transfers */
+	pages = (pl022->cur_transfer->len >> PAGE_SHIFT) + 1;
+	dev_dbg(&pl022->adev->dev, "using %d pages for transfer\n", pages);
+
+	ret = sg_alloc_table(&pl022->sgt_rx, pages, GFP_KERNEL);
+	if (ret)
+		goto err_alloc_rx_sg;
+
+	ret = sg_alloc_table(&pl022->sgt_tx, pages, GFP_KERNEL);
+	if (ret)
+		goto err_alloc_tx_sg;
+
+	/* Fill in the scatterlists for the RX+TX buffers */
+	setup_dma_scatter(pl022, pl022->rx,
+			  pl022->cur_transfer->len, &pl022->sgt_rx);
+	setup_dma_scatter(pl022, pl022->tx,
+			  pl022->cur_transfer->len, &pl022->sgt_tx);
+
+	/* Map DMA buffers */
+	rx_sglen = dma_map_sg(rxchan->device->dev, pl022->sgt_rx.sgl,
+			   pl022->sgt_rx.nents, DMA_FROM_DEVICE);
+	if (!rx_sglen)
+		goto err_rx_sgmap;
+
+	tx_sglen = dma_map_sg(txchan->device->dev, pl022->sgt_tx.sgl,
+			   pl022->sgt_tx.nents, DMA_TO_DEVICE);
+	if (!tx_sglen)
+		goto err_tx_sgmap;
+
+	/* Send both scatterlists */
+	rxdesc = rxchan->device->device_prep_slave_sg(rxchan,
+				      pl022->sgt_rx.sgl,
+				      rx_sglen,
+				      DMA_FROM_DEVICE,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!rxdesc)
+		goto err_rxdesc;
+
+	txdesc = txchan->device->device_prep_slave_sg(txchan,
+				      pl022->sgt_tx.sgl,
+				      tx_sglen,
+				      DMA_TO_DEVICE,
+				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
+	if (!txdesc)
+		goto err_txdesc;
+
+	/* Put the callback on the RX transfer only, that should finish last */
+	rxdesc->callback = dma_callback;
+	rxdesc->callback_param = pl022;
+
+	/* Submit and fire RX and TX with TX last so we're ready to read! */
+	dmaengine_submit(rxdesc);
+	dmaengine_submit(txdesc);
+	dma_async_issue_pending(rxchan);
+	dma_async_issue_pending(txchan);
+
+	return 0;
+
+err_txdesc:
+	dmaengine_terminate_all(txchan);
+err_rxdesc:
+	dmaengine_terminate_all(rxchan);
+	dma_unmap_sg(txchan->device->dev, pl022->sgt_tx.sgl,
+		     pl022->sgt_tx.nents, DMA_TO_DEVICE);
+err_tx_sgmap:
+	dma_unmap_sg(rxchan->device->dev, pl022->sgt_rx.sgl,
+		     pl022->sgt_tx.nents, DMA_FROM_DEVICE);
+err_rx_sgmap:
+	sg_free_table(&pl022->sgt_tx);
+err_alloc_tx_sg:
+	sg_free_table(&pl022->sgt_rx);
+err_alloc_rx_sg:
+	return -ENOMEM;
+}
+
+static int __init pl022_dma_probe(struct pl022 *pl022)
+{
+	dma_cap_mask_t mask;
+
+	/* Try to acquire a generic DMA engine slave channel */
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_SLAVE, mask);
+	/*
+	 * We need both RX and TX channels to do DMA, else do none
+	 * of them.
+	 */
+	pl022->dma_rx_channel = dma_request_channel(mask,
+					    pl022->master_info->dma_filter,
+					    pl022->master_info->dma_rx_param);
+	if (!pl022->dma_rx_channel) {
+		dev_dbg(&pl022->adev->dev, "no RX DMA channel!\n");
+		goto err_no_rxchan;
+	}
+
+	pl022->dma_tx_channel = dma_request_channel(mask,
+					    pl022->master_info->dma_filter,
+					    pl022->master_info->dma_tx_param);
+	if (!pl022->dma_tx_channel) {
+		dev_dbg(&pl022->adev->dev, "no TX DMA channel!\n");
+		goto err_no_txchan;
+	}
+
+	pl022->dummypage = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!pl022->dummypage) {
+		dev_dbg(