diff mbox

[v3,1/2] spi: add support for DLN-2 USB-SPI adapter

Message ID 1416312599-5176-2-git-send-email-laurentiu.palcu@intel.com (mailing list archive)
State New, archived
Headers show

Commit Message

Laurentiu Palcu Nov. 18, 2014, 12:09 p.m. UTC
This adds support for Diolan DLN2 USB-SPI adapter.

Information about the USB protocol interface can be found in the
Programmer's Reference Manual [1], see section 5.4.6 for the SPI
master module commands and responses.

[1] https://www.diolan.com/downloads/dln-api-manual.pdf

Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com>
---
 drivers/spi/Kconfig    |  10 +
 drivers/spi/Makefile   |   1 +
 drivers/spi/spi-dln2.c | 818 +++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 829 insertions(+)
 create mode 100644 drivers/spi/spi-dln2.c

Comments

Johan Hovold Nov. 18, 2014, 5:06 p.m. UTC | #1
On Tue, Nov 18, 2014 at 02:09:58PM +0200, Laurentiu Palcu wrote:

> +/*
> + * Copy the data to DLN2 buffer and change the byte order to LE, requested by
> + * DLN2 module. SPI core makes sure that the data length is a multiple of word
> + * size.
> + */
> +static int dln2_spi_copy_to_buf(u8 *dln2_buf, const u8 *src, u16 len, u8 bpw)
> +{
> +#ifdef __LITTLE_ENDIAN
> +	memcpy(dln2_buf, src, len);
> +#else
> +	if (bpw <= 8) {
> +		memcpy(dln2_buf, src, len);
> +	} else if (bpw <= 16) {
> +		__le16 *d = (__le16 *)dln2_buf;
> +		u16 *s = (u16 *)src;
> +
> +		len = len / 2;
> +		while (len--)
> +			put_unaligned_le16(get_unaligned(s++), d++);

You said the dln2 buffer was properly aligned, right? Then you don't
need to use put_unaligned_lexx here...

> +	} else {
> +		__le32 *d = (__le32 *)dln2_buf;
> +		u32 *s = (u32 *)src;
> +
> +		len = len / 4;
> +		while (len--)
> +			put_unaligned_le32(get_unaligned(s++), d++);
> +	}
> +#endif
> +
> +	return 0;
> +}
> +
> +/*
> + * Copy the data from DLN2 buffer and convert to CPU byte order since the DLN2
> + * buffer is LE ordered. SPI core makes sure that the data length is a multiple
> + * of word size.
> + */
> +static int dln2_spi_copy_from_buf(u8 *dest, const u8 *dln2_buf, u16 len, u8 bpw)
> +{
> +#ifdef __LITTLE_ENDIAN
> +	memcpy(dest, dln2_buf, len);
> +#else
> +	if (bpw <= 8) {
> +		memcpy(dest, dln2_buf, len);
> +	} else if (bpw <= 16) {
> +		u16 *d = (u16 *)dest;
> +		__le16 *s = (__le16 *)dln2_buf;
> +
> +		len = len / 2;
> +		while (len--)
> +			put_unaligned(get_unaligned_le16(s++), d++)

...or get_unaligned_lexx here.

> +	} else {
> +		u32 *d = (u32 *)dest;
> +		__le32 *s = (__le32 *)dln2_buf;
> +
> +		len = len / 4;
> +		while (len--)
> +			put_unaligned(get_unaligned_le32(s++), d++)
> +	}
> +#endif
> +
> +	return 0;
> +}

Did you check the alignment of the SPI buffers as well? I'd assume they
were DMA-able and thus properly aligned, and then you do not need to use
any unaligned helpers above at all.

Johan
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Laurentiu Palcu Nov. 19, 2014, 9:53 a.m. UTC | #2
On Tue, Nov 18, 2014 at 06:06:27PM +0100, Johan Hovold wrote:
> On Tue, Nov 18, 2014 at 02:09:58PM +0200, Laurentiu Palcu wrote:
> 
> > +/*
> > + * Copy the data to DLN2 buffer and change the byte order to LE, requested by
> > + * DLN2 module. SPI core makes sure that the data length is a multiple of word
> > + * size.
> > + */
> > +static int dln2_spi_copy_to_buf(u8 *dln2_buf, const u8 *src, u16 len, u8 bpw)
> > +{
> > +#ifdef __LITTLE_ENDIAN
> > +	memcpy(dln2_buf, src, len);
> > +#else
> > +	if (bpw <= 8) {
> > +		memcpy(dln2_buf, src, len);
> > +	} else if (bpw <= 16) {
> > +		__le16 *d = (__le16 *)dln2_buf;
> > +		u16 *s = (u16 *)src;
> > +
> > +		len = len / 2;
> > +		while (len--)
> > +			put_unaligned_le16(get_unaligned(s++), d++);
> 
> You said the dln2 buffer was properly aligned, right? Then you don't
> need to use put_unaligned_lexx here...
I know, and I was very close not to use it, but, as you said, others
might not be as careful and if Diolan FW changes anything in the
header's structure in the future, the alignment will be off. This is
just a precaution. I know it's a little bit costly but this will affect
only BE machines and only if bits-per-word is bigger than 8.

> 
> > +	} else {
> > +		__le32 *d = (__le32 *)dln2_buf;
> > +		u32 *s = (u32 *)src;
> > +
> > +		len = len / 4;
> > +		while (len--)
> > +			put_unaligned_le32(get_unaligned(s++), d++);
> > +	}
> > +#endif
> > +
> > +	return 0;
> > +}
> > +
> > +/*
> > + * Copy the data from DLN2 buffer and convert to CPU byte order since the DLN2
> > + * buffer is LE ordered. SPI core makes sure that the data length is a multiple
> > + * of word size.
> > + */
> > +static int dln2_spi_copy_from_buf(u8 *dest, const u8 *dln2_buf, u16 len, u8 bpw)
> > +{
> > +#ifdef __LITTLE_ENDIAN
> > +	memcpy(dest, dln2_buf, len);
> > +#else
> > +	if (bpw <= 8) {
> > +		memcpy(dest, dln2_buf, len);
> > +	} else if (bpw <= 16) {
> > +		u16 *d = (u16 *)dest;
> > +		__le16 *s = (__le16 *)dln2_buf;
> > +
> > +		len = len / 2;
> > +		while (len--)
> > +			put_unaligned(get_unaligned_le16(s++), d++)
> 
> ...or get_unaligned_lexx here.
> 
> > +	} else {
> > +		u32 *d = (u32 *)dest;
> > +		__le32 *s = (__le32 *)dln2_buf;
> > +
> > +		len = len / 4;
> > +		while (len--)
> > +			put_unaligned(get_unaligned_le32(s++), d++)
> > +	}
> > +#endif
> > +
> > +	return 0;
> > +}
> 
> Did you check the alignment of the SPI buffers as well? I'd assume they
> were DMA-able and thus properly aligned, and then you do not need to use
> any unaligned helpers above at all.
I did... The spi_transfer struct documentation says that those tx_buf
and rx_buf buffers are dma-safe but, apparently, I found drivers that do
not really use alligned buffers for transfers. Look, for example, at
max1111.c. Those buffers don't look aligned to me... Well, this driver
is probably not the best example since BPW is 8, but you got my point.
Other drivers, indeed, use the ____cacheline_aligned attribute and
should be dma-safe.

laurentiu
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Johan Hovold Nov. 19, 2014, 10:09 a.m. UTC | #3
On Wed, Nov 19, 2014 at 11:53:44AM +0200, Laurentiu Palcu wrote:
> On Tue, Nov 18, 2014 at 06:06:27PM +0100, Johan Hovold wrote:
> > On Tue, Nov 18, 2014 at 02:09:58PM +0200, Laurentiu Palcu wrote:
> > 
> > > +/*
> > > + * Copy the data to DLN2 buffer and change the byte order to LE, requested by
> > > + * DLN2 module. SPI core makes sure that the data length is a multiple of word
> > > + * size.
> > > + */
> > > +static int dln2_spi_copy_to_buf(u8 *dln2_buf, const u8 *src, u16 len, u8 bpw)
> > > +{
> > > +#ifdef __LITTLE_ENDIAN
> > > +	memcpy(dln2_buf, src, len);
> > > +#else
> > > +	if (bpw <= 8) {
> > > +		memcpy(dln2_buf, src, len);
> > > +	} else if (bpw <= 16) {
> > > +		__le16 *d = (__le16 *)dln2_buf;
> > > +		u16 *s = (u16 *)src;
> > > +
> > > +		len = len / 2;
> > > +		while (len--)
> > > +			put_unaligned_le16(get_unaligned(s++), d++);
> > 
> > You said the dln2 buffer was properly aligned, right? Then you don't
> > need to use put_unaligned_lexx here...
>
> I know, and I was very close not to use it, but, as you said, others
> might not be as careful

That was in a different context. :)

> and if Diolan FW changes anything in the
> header's structure in the future, the alignment will be off. This is
> just a precaution. I know it's a little bit costly but this will affect
> only BE machines and only if bits-per-word is bigger than 8.

I don't think you should worry about hypothetical changes to the
protocol.

You verified the that headers are a multiple of four, so no need for the
unaligned macros.

> > > +	} else {
> > > +		__le32 *d = (__le32 *)dln2_buf;
> > > +		u32 *s = (u32 *)src;
> > > +
> > > +		len = len / 4;
> > > +		while (len--)
> > > +			put_unaligned_le32(get_unaligned(s++), d++);
> > > +	}
> > > +#endif
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +/*
> > > + * Copy the data from DLN2 buffer and convert to CPU byte order since the DLN2
> > > + * buffer is LE ordered. SPI core makes sure that the data length is a multiple
> > > + * of word size.
> > > + */
> > > +static int dln2_spi_copy_from_buf(u8 *dest, const u8 *dln2_buf, u16 len, u8 bpw)
> > > +{
> > > +#ifdef __LITTLE_ENDIAN
> > > +	memcpy(dest, dln2_buf, len);
> > > +#else
> > > +	if (bpw <= 8) {
> > > +		memcpy(dest, dln2_buf, len);
> > > +	} else if (bpw <= 16) {
> > > +		u16 *d = (u16 *)dest;
> > > +		__le16 *s = (__le16 *)dln2_buf;
> > > +
> > > +		len = len / 2;
> > > +		while (len--)
> > > +			put_unaligned(get_unaligned_le16(s++), d++)
> > 
> > ...or get_unaligned_lexx here.
> > 
> > > +	} else {
> > > +		u32 *d = (u32 *)dest;
> > > +		__le32 *s = (__le32 *)dln2_buf;
> > > +
> > > +		len = len / 4;
> > > +		while (len--)
> > > +			put_unaligned(get_unaligned_le32(s++), d++)
> > > +	}
> > > +#endif
> > > +
> > > +	return 0;
> > > +}
> > 
> > Did you check the alignment of the SPI buffers as well? I'd assume they
> > were DMA-able and thus properly aligned, and then you do not need to use
> > any unaligned helpers above at all.
>
> I did... The spi_transfer struct documentation says that those tx_buf
> and rx_buf buffers are dma-safe but, apparently, I found drivers that do
> not really use alligned buffers for transfers. Look, for example, at
> max1111.c. Those buffers don't look aligned to me...

That looks like a bug to me. DMA-buffers should never be allocated as
part of a larger structure.

> Well, this driver
> is probably not the best example since BPW is 8, but you got my point.
> Other drivers, indeed, use the ____cacheline_aligned attribute and
> should be dma-safe.

I think you can remove the unaligned macros altogether.

Johan
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index 62e2242..a52a910 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -176,6 +176,16 @@  config SPI_DAVINCI
 	help
 	  SPI master controller for DaVinci/DA8x/OMAP-L/AM1x SPI modules.
 
+config SPI_DLN2
+       tristate "Diolan DLN-2 USB SPI adapter"
+       depends on MFD_DLN2
+       help
+         If you say yes to this option, support will be included for Diolan
+         DLN2, a USB to SPI interface.
+
+         This driver can also be built as a module.  If so, the module
+         will be called spi-dln2.
+
 config SPI_EFM32
 	tristate "EFM32 SPI controller"
 	depends on OF && ARM && (ARCH_EFM32 || COMPILE_TEST)
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
index 762da07..b315da2 100644
--- a/drivers/spi/Makefile
+++ b/drivers/spi/Makefile
@@ -26,6 +26,7 @@  obj-$(CONFIG_SPI_CADENCE)		+= spi-cadence.o
 obj-$(CONFIG_SPI_CLPS711X)		+= spi-clps711x.o
 obj-$(CONFIG_SPI_COLDFIRE_QSPI)		+= spi-coldfire-qspi.o
 obj-$(CONFIG_SPI_DAVINCI)		+= spi-davinci.o
+obj-$(CONFIG_SPI_DLN2)			+= spi-dln2.o
 obj-$(CONFIG_SPI_DESIGNWARE)		+= spi-dw.o
 obj-$(CONFIG_SPI_DW_MMIO)		+= spi-dw-mmio.o
 obj-$(CONFIG_SPI_DW_PCI)		+= spi-dw-midpci.o
diff --git a/drivers/spi/spi-dln2.c b/drivers/spi/spi-dln2.c
new file mode 100644
index 0000000..1b26804
--- /dev/null
+++ b/drivers/spi/spi-dln2.c
@@ -0,0 +1,818 @@ 
+/*
+ * Driver for the Diolan DLN-2 USB-SPI adapter
+ *
+ * Copyright (c) 2014 Intel Corporation
+ *
+ * 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, version 2.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/dln2.h>
+#include <linux/spi/spi.h>
+
+#define DLN2_SPI_MODULE_ID		0x02
+#define DLN2_SPI_CMD(cmd)		DLN2_CMD(cmd, DLN2_SPI_MODULE_ID)
+
+/* SPI commands */
+#define DLN2_SPI_GET_PORT_COUNT			DLN2_SPI_CMD(0x00)
+#define DLN2_SPI_ENABLE				DLN2_SPI_CMD(0x11)
+#define DLN2_SPI_DISABLE			DLN2_SPI_CMD(0x12)
+#define DLN2_SPI_IS_ENABLED			DLN2_SPI_CMD(0x13)
+#define DLN2_SPI_SET_MODE			DLN2_SPI_CMD(0x14)
+#define DLN2_SPI_GET_MODE			DLN2_SPI_CMD(0x15)
+#define DLN2_SPI_SET_FRAME_SIZE			DLN2_SPI_CMD(0x16)
+#define DLN2_SPI_GET_FRAME_SIZE			DLN2_SPI_CMD(0x17)
+#define DLN2_SPI_SET_FREQUENCY			DLN2_SPI_CMD(0x18)
+#define DLN2_SPI_GET_FREQUENCY			DLN2_SPI_CMD(0x19)
+#define DLN2_SPI_READ_WRITE			DLN2_SPI_CMD(0x1A)
+#define DLN2_SPI_READ				DLN2_SPI_CMD(0x1B)
+#define DLN2_SPI_WRITE				DLN2_SPI_CMD(0x1C)
+#define DLN2_SPI_SET_DELAY_BETWEEN_SS		DLN2_SPI_CMD(0x20)
+#define DLN2_SPI_GET_DELAY_BETWEEN_SS		DLN2_SPI_CMD(0x21)
+#define DLN2_SPI_SET_DELAY_AFTER_SS		DLN2_SPI_CMD(0x22)
+#define DLN2_SPI_GET_DELAY_AFTER_SS		DLN2_SPI_CMD(0x23)
+#define DLN2_SPI_SET_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x24)
+#define DLN2_SPI_GET_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x25)
+#define DLN2_SPI_SET_SS				DLN2_SPI_CMD(0x26)
+#define DLN2_SPI_GET_SS				DLN2_SPI_CMD(0x27)
+#define DLN2_SPI_RELEASE_SS			DLN2_SPI_CMD(0x28)
+#define DLN2_SPI_SS_VARIABLE_ENABLE		DLN2_SPI_CMD(0x2B)
+#define DLN2_SPI_SS_VARIABLE_DISABLE		DLN2_SPI_CMD(0x2C)
+#define DLN2_SPI_SS_VARIABLE_IS_ENABLED		DLN2_SPI_CMD(0x2D)
+#define DLN2_SPI_SS_AAT_ENABLE			DLN2_SPI_CMD(0x2E)
+#define DLN2_SPI_SS_AAT_DISABLE			DLN2_SPI_CMD(0x2F)
+#define DLN2_SPI_SS_AAT_IS_ENABLED		DLN2_SPI_CMD(0x30)
+#define DLN2_SPI_SS_BETWEEN_FRAMES_ENABLE	DLN2_SPI_CMD(0x31)
+#define DLN2_SPI_SS_BETWEEN_FRAMES_DISABLE	DLN2_SPI_CMD(0x32)
+#define DLN2_SPI_SS_BETWEEN_FRAMES_IS_ENABLED	DLN2_SPI_CMD(0x33)
+#define DLN2_SPI_SET_CPHA			DLN2_SPI_CMD(0x34)
+#define DLN2_SPI_GET_CPHA			DLN2_SPI_CMD(0x35)
+#define DLN2_SPI_SET_CPOL			DLN2_SPI_CMD(0x36)
+#define DLN2_SPI_GET_CPOL			DLN2_SPI_CMD(0x37)
+#define DLN2_SPI_SS_MULTI_ENABLE		DLN2_SPI_CMD(0x38)
+#define DLN2_SPI_SS_MULTI_DISABLE		DLN2_SPI_CMD(0x39)
+#define DLN2_SPI_SS_MULTI_IS_ENABLED		DLN2_SPI_CMD(0x3A)
+#define DLN2_SPI_GET_SUPPORTED_MODES		DLN2_SPI_CMD(0x40)
+#define DLN2_SPI_GET_SUPPORTED_CPHA_VALUES	DLN2_SPI_CMD(0x41)
+#define DLN2_SPI_GET_SUPPORTED_CPOL_VALUES	DLN2_SPI_CMD(0x42)
+#define DLN2_SPI_GET_SUPPORTED_FRAME_SIZES	DLN2_SPI_CMD(0x43)
+#define DLN2_SPI_GET_SS_COUNT			DLN2_SPI_CMD(0x44)
+#define DLN2_SPI_GET_MIN_FREQUENCY		DLN2_SPI_CMD(0x45)
+#define DLN2_SPI_GET_MAX_FREQUENCY		DLN2_SPI_CMD(0x46)
+#define DLN2_SPI_GET_MIN_DELAY_BETWEEN_SS	DLN2_SPI_CMD(0x47)
+#define DLN2_SPI_GET_MAX_DELAY_BETWEEN_SS	DLN2_SPI_CMD(0x48)
+#define DLN2_SPI_GET_MIN_DELAY_AFTER_SS		DLN2_SPI_CMD(0x49)
+#define DLN2_SPI_GET_MAX_DELAY_AFTER_SS		DLN2_SPI_CMD(0x4A)
+#define DLN2_SPI_GET_MIN_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x4B)
+#define DLN2_SPI_GET_MAX_DELAY_BETWEEN_FRAMES	DLN2_SPI_CMD(0x4C)
+
+#define DLN2_SPI_MAX_XFER_SIZE			256
+#define DLN2_SPI_BUF_SIZE			(DLN2_SPI_MAX_XFER_SIZE + 16)
+#define DLN2_SPI_ATTR_LEAVE_SS_LOW		BIT(0)
+#define DLN2_TRANSFERS_WAIT_COMPLETE		1
+#define DLN2_TRANSFERS_CANCEL			0
+
+struct dln2_spi {
+	struct platform_device *pdev;
+	struct spi_master *master;
+	u8 port;
+
+	/* no need to protect this buffer because all SPI communication is
+	 * serialized by the SPI core
+	 */
+	void *buf;
+
+	u8 bpw;
+	u32 speed;
+	u16 mode;
+	u8 cs;
+};
+
+/*
+ * Enable/Disable SPI module. The disable command will wait for transfers to
+ * complete first.
+ */
+static int dln2_spi_enable(struct dln2_spi *dln2, bool enable)
+{
+	int ret;
+	u16 cmd;
+	struct {
+		u8 port;
+		u8 wait_for_completion;
+	} tx;
+	unsigned len = sizeof(tx);
+
+	tx.port = dln2->port;
+
+	if (enable) {
+		cmd = DLN2_SPI_ENABLE;
+		len -= sizeof(tx.wait_for_completion);
+	} else {
+		tx.wait_for_completion = DLN2_TRANSFERS_WAIT_COMPLETE;
+		cmd = DLN2_SPI_DISABLE;
+	}
+
+	ret = dln2_transfer_tx(dln2->pdev, cmd, &tx, len);
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+/*
+ * Select/unselect multiple CS lines. The selected lines will be automatically
+ * toggled LOW/HIGH by the board firmware during transfers, provided they're
+ * enabled first.
+ *
+ * Ex: cs_mask = 0x03 -> CS0 & CS1 will be selected and the next WR/RD operation
+ *                       will toggle the lines LOW/HIGH automatically.
+ */
+static int dln2_spi_cs_set(struct dln2_spi *dln2, u8 cs_mask)
+{
+	struct {
+		u8 port;
+		u8 cs;
+	} tx;
+
+	tx.port = dln2->port;
+
+	/* According to Diolan docs, "a slave device can be selected by changing
+	 * the corresponding bit value to 0". The rest must be set to 1. Hence
+	 * the bitwise NOT in front.
+	 */
+	tx.cs = ~cs_mask;
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_SET_SS, &tx, sizeof(tx));
+}
+
+/*
+ * Select one CS line. The other lines will be un-selected.
+ */
+static int dln2_spi_cs_set_one(struct dln2_spi *dln2, u8 cs)
+{
+	return dln2_spi_cs_set(dln2, BIT(cs));
+}
+
+/*
+ * Enable/disable CS lines for usage. The module has to be disabled first.
+ */
+static int dln2_spi_cs_enable(struct dln2_spi *dln2, u8 cs_mask, bool enable)
+{
+	struct {
+		u8 port;
+		u8 cs;
+	} tx;
+	u16 cmd;
+
+	tx.port = dln2->port;
+	tx.cs = cs_mask;
+	cmd = enable ? DLN2_SPI_SS_MULTI_ENABLE : DLN2_SPI_SS_MULTI_DISABLE;
+
+	return dln2_transfer_tx(dln2->pdev, cmd, &tx, sizeof(tx));
+}
+
+static int dln2_spi_cs_enable_all(struct dln2_spi *dln2, bool enable)
+{
+	u8 cs_mask = GENMASK(dln2->master->num_chipselect - 1, 0);
+
+	return dln2_spi_cs_enable(dln2, cs_mask, enable);
+}
+
+static int dln2_spi_get_cs_num(struct dln2_spi *dln2, u16 *cs_num)
+{
+	int ret;
+	struct {
+		u8 port;
+	} tx;
+	struct {
+		__le16 cs_count;
+	} rx;
+	unsigned rx_len = sizeof(rx);
+
+	if (!cs_num)
+		return -EINVAL;
+
+	tx.port = dln2->port;
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_GET_SS_COUNT, &tx, sizeof(tx),
+			    &rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx))
+		return -EPROTO;
+
+	*cs_num = le16_to_cpu(rx.cs_count);
+
+	dev_dbg(&dln2->pdev->dev, "cs_num = %d\n", *cs_num);
+
+	return 0;
+}
+
+static int dln2_spi_get_speed(struct dln2_spi *dln2, u16 cmd, u32 *freq)
+{
+	int ret;
+	struct {
+		u8 port;
+	} tx;
+	struct {
+		__le32 speed;
+	} rx;
+	unsigned rx_len = sizeof(rx);
+
+	if (!freq)
+		return -EINVAL;
+
+	tx.port = dln2->port;
+
+	ret = dln2_transfer(dln2->pdev, cmd, &tx, sizeof(tx), &rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx))
+		return -EPROTO;
+
+	*freq = le32_to_cpu(rx.speed);
+
+	return 0;
+}
+
+/*
+ * Get bus min/max frequencies.
+ */
+static int dln2_spi_get_speed_range(struct dln2_spi *dln2, u32 *fmin, u32 *fmax)
+{
+	int ret;
+
+	if (!fmin || !fmax)
+		return -EINVAL;
+
+	ret = dln2_spi_get_speed(dln2, DLN2_SPI_GET_MIN_FREQUENCY,
+				 fmin);
+	if (ret < 0)
+		return ret;
+
+	ret = dln2_spi_get_speed(dln2, DLN2_SPI_GET_MAX_FREQUENCY,
+				 fmax);
+	if (ret < 0)
+		return ret;
+
+	dev_dbg(&dln2->pdev->dev, "freq_min = %d, freq_max = %d\n",
+		*fmin, *fmax);
+
+	return 0;
+}
+
+/*
+ * Set the bus speed. The module will automatically round down to the closest
+ * available frequency and returns it. The module has to be disabled first.
+ */
+static int dln2_spi_set_speed(struct dln2_spi *dln2, u32 speed)
+{
+	int ret;
+	struct {
+		u8 port;
+		__le32 speed;
+	} __packed tx;
+	struct {
+		__le32 speed;
+	} rx;
+	int rx_len = sizeof(rx);
+
+	tx.port = dln2->port;
+	tx.speed = cpu_to_le32(speed);
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_SET_FREQUENCY, &tx, sizeof(tx),
+			    &rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx))
+		return -EPROTO;
+
+	return 0;
+}
+
+/*
+ * Change CPOL & CPHA. The module has to be disabled first.
+ */
+static int dln2_spi_set_mode(struct dln2_spi *dln2, u8 mode)
+{
+	struct {
+		u8 port;
+		u8 mode;
+	} tx;
+
+	tx.port = dln2->port;
+	tx.mode = mode;
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_SET_MODE, &tx, sizeof(tx));
+}
+
+/*
+ * Change frame size. The module has to be disabled first.
+ */
+static int dln2_spi_set_bpw(struct dln2_spi *dln2, u8 bpw)
+{
+	struct {
+		u8 port;
+		u8 bpw;
+	} tx;
+
+	tx.port = dln2->port;
+	tx.bpw = bpw;
+
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_SET_FRAME_SIZE,
+				&tx, sizeof(tx));
+}
+
+static int dln2_spi_get_supported_frame_sizes(struct dln2_spi *dln2,
+					      u32 *bpw_mask)
+{
+	int ret;
+	struct {
+		u8 port;
+	} tx;
+	struct {
+		u8 count;
+		u8 frame_sizes[36];
+	} *rx = dln2->buf;
+	unsigned rx_len = sizeof(*rx);
+	int i;
+
+	if (!bpw_mask)
+		return -EINVAL;
+
+	tx.port = dln2->port;
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_GET_SUPPORTED_FRAME_SIZES,
+			    &tx, sizeof(tx), rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len != sizeof(*rx))
+		return -EPROTO;
+	if (rx->count > ARRAY_SIZE(rx->frame_sizes))
+		return -EPROTO;
+
+	*bpw_mask = 0;
+	for (i = 0; i < rx->count; i++)
+		*bpw_mask |= BIT(rx->frame_sizes[i] - 1);
+
+	dev_dbg(&dln2->pdev->dev, "bpw_mask = 0x%X\n", *bpw_mask);
+
+	return 0;
+}
+
+/*
+ * Copy the data to DLN2 buffer and change the byte order to LE, requested by
+ * DLN2 module. SPI core makes sure that the data length is a multiple of word
+ * size.
+ */
+static int dln2_spi_copy_to_buf(u8 *dln2_buf, const u8 *src, u16 len, u8 bpw)
+{
+#ifdef __LITTLE_ENDIAN
+	memcpy(dln2_buf, src, len);
+#else
+	if (bpw <= 8) {
+		memcpy(dln2_buf, src, len);
+	} else if (bpw <= 16) {
+		__le16 *d = (__le16 *)dln2_buf;
+		u16 *s = (u16 *)src;
+
+		len = len / 2;
+		while (len--)
+			put_unaligned_le16(get_unaligned(s++), d++);
+	} else {
+		__le32 *d = (__le32 *)dln2_buf;
+		u32 *s = (u32 *)src;
+
+		len = len / 4;
+		while (len--)
+			put_unaligned_le32(get_unaligned(s++), d++);
+	}
+#endif
+
+	return 0;
+}
+
+/*
+ * Copy the data from DLN2 buffer and convert to CPU byte order since the DLN2
+ * buffer is LE ordered. SPI core makes sure that the data length is a multiple
+ * of word size.
+ */
+static int dln2_spi_copy_from_buf(u8 *dest, const u8 *dln2_buf, u16 len, u8 bpw)
+{
+#ifdef __LITTLE_ENDIAN
+	memcpy(dest, dln2_buf, len);
+#else
+	if (bpw <= 8) {
+		memcpy(dest, dln2_buf, len);
+	} else if (bpw <= 16) {
+		u16 *d = (u16 *)dest;
+		__le16 *s = (__le16 *)dln2_buf;
+
+		len = len / 2;
+		while (len--)
+			put_unaligned(get_unaligned_le16(s++), d++)
+	} else {
+		u32 *d = (u32 *)dest;
+		__le32 *s = (__le32 *)dln2_buf;
+
+		len = len / 4;
+		while (len--)
+			put_unaligned(get_unaligned_le32(s++), d++)
+	}
+#endif
+
+	return 0;
+}
+
+/*
+ * Perform one write operation.
+ */
+static int dln2_spi_write_one(struct dln2_spi *dln2, const u8 *data,
+			      u16 data_len, u8 attr)
+{
+	struct {
+		u8 port;
+		__le16 size;
+		u8 attr;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *tx = dln2->buf;
+	unsigned tx_len;
+
+	BUILD_BUG_ON(sizeof(*tx) > DLN2_SPI_BUF_SIZE);
+
+	if (data_len > DLN2_SPI_MAX_XFER_SIZE)
+		return -EINVAL;
+
+	tx->port = dln2->port;
+	tx->size = cpu_to_le16(data_len);
+	tx->attr = attr;
+
+	dln2_spi_copy_to_buf(tx->buf, data, data_len, dln2->bpw);
+
+	tx_len = sizeof(*tx) + data_len - DLN2_SPI_MAX_XFER_SIZE;
+	return dln2_transfer_tx(dln2->pdev, DLN2_SPI_WRITE, tx, tx_len);
+}
+
+/*
+ * Perform one read operation.
+ */
+static int dln2_spi_read_one(struct dln2_spi *dln2, u8 *data,
+			     u16 data_len, u8 attr)
+{
+	int ret;
+	struct {
+		u8 port;
+		__le16 size;
+		u8 attr;
+	} __packed tx;
+	struct {
+		__le16 size;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *rx = dln2->buf;
+	unsigned rx_len = sizeof(*rx);
+
+	BUILD_BUG_ON(sizeof(*rx) > DLN2_SPI_BUF_SIZE);
+
+	if (data_len > DLN2_SPI_MAX_XFER_SIZE)
+		return -EINVAL;
+
+	tx.port = dln2->port;
+	tx.size = cpu_to_le16(data_len);
+	tx.attr = attr;
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_READ, &tx, sizeof(tx),
+			    rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx->size) + data_len)
+		return -EPROTO;
+	if (le16_to_cpu(rx->size) != data_len)
+		return -EPROTO;
+
+	dln2_spi_copy_from_buf(data, rx->buf, data_len, dln2->bpw);
+
+	return 0;
+}
+
+/*
+ * Perform one write & read operation.
+ */
+static int dln2_spi_read_write_one(struct dln2_spi *dln2, const u8 *tx_data,
+				   u8 *rx_data, u16 data_len, u8 attr)
+{
+	int ret;
+	struct {
+		u8 port;
+		__le16 size;
+		u8 attr;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *tx;
+	struct {
+		__le16 size;
+		u8 buf[DLN2_SPI_MAX_XFER_SIZE];
+	} __packed *rx;
+	unsigned tx_len, rx_len;
+
+	BUILD_BUG_ON(sizeof(*tx) > DLN2_SPI_BUF_SIZE ||
+		     sizeof(*rx) > DLN2_SPI_BUF_SIZE);
+
+	if (data_len > DLN2_SPI_MAX_XFER_SIZE)
+		return -EINVAL;
+
+	/* Since this is a pseudo full-duplex communication, we're perfectly
+	 * safe to use the same buffer for both tx and rx. When DLN2 sends the
+	 * response back, with the rx data, we don't need the tx buffer anymore.
+	 */
+	tx = dln2->buf;
+	rx = dln2->buf;
+
+	tx->port = dln2->port;
+	tx->size = cpu_to_le16(data_len);
+	tx->attr = attr;
+
+	dln2_spi_copy_to_buf(tx->buf, tx_data, data_len, dln2->bpw);
+
+	tx_len = sizeof(*tx) + data_len - DLN2_SPI_MAX_XFER_SIZE;
+	rx_len = sizeof(*rx);
+
+	ret = dln2_transfer(dln2->pdev, DLN2_SPI_READ_WRITE, tx, tx_len,
+			    rx, &rx_len);
+	if (ret < 0)
+		return ret;
+	if (rx_len < sizeof(rx->size) + data_len)
+		return -EPROTO;
+	if (le16_to_cpu(rx->size) != data_len)
+		return -EPROTO;
+
+	dln2_spi_copy_from_buf(rx_data, rx->buf, data_len, dln2->bpw);
+
+	return 0;
+}
+
+/*
+ * Read/Write wrapper. It will automatically split an operation into multiple
+ * single ones due to device buffer constraints.
+ */
+static int dln2_spi_rdwr(struct dln2_spi *dln2, const u8 *tx_data,
+			 u8 *rx_data, u16 data_len, u8 attr) {
+	int ret;
+	u16 len;
+	u8 temp_attr;
+	u16 remaining = data_len;
+	u16 offset;
+
+	do {
+		if (remaining > DLN2_SPI_MAX_XFER_SIZE) {
+			len = DLN2_SPI_MAX_XFER_SIZE;
+			temp_attr = DLN2_SPI_ATTR_LEAVE_SS_LOW;
+		} else {
+			len = remaining;
+			temp_attr = attr;
+		}
+
+		offset = data_len - remaining;
+
+		if (tx_data && rx_data) {
+			ret = dln2_spi_read_write_one(dln2,
+						      tx_data + offset,
+						      rx_data + offset,
+						      len, temp_attr);
+		} else if (tx_data) {
+			ret = dln2_spi_write_one(dln2,
+						 tx_data + offset,
+						 len, temp_attr);
+		} else if (rx_data) {
+			ret = dln2_spi_read_one(dln2,
+						rx_data + offset,
+						len, temp_attr);
+		 } else {
+			return -EINVAL;
+		 }
+
+		if (ret < 0)
+			return ret;
+
+		remaining -= len;
+	} while (remaining);
+
+	return 0;
+}
+
+static int dln2_spi_prepare_message(struct spi_master *master,
+				    struct spi_message *message)
+{
+	int ret;
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+	struct spi_device *spi = message->spi;
+
+	if (dln2->cs != spi->chip_select) {
+		ret = dln2_spi_cs_set_one(dln2, spi->chip_select);
+		if (ret < 0)
+			return ret;
+
+		dln2->cs = spi->chip_select;
+	}
+
+	return 0;
+}
+
+static int dln2_spi_transfer_setup(struct dln2_spi *dln2, u32 speed,
+				   u8 bpw, u8 mode)
+{
+	int ret;
+	bool bus_setup_change;
+
+	bus_setup_change = dln2->speed != speed || dln2->mode != mode ||
+			   dln2->bpw != bpw;
+
+	if (!bus_setup_change)
+		return 0;
+
+	ret = dln2_spi_enable(dln2, false);
+	if (ret < 0)
+		return ret;
+
+	if (dln2->speed != speed) {
+		ret = dln2_spi_set_speed(dln2, speed);
+		if (ret < 0)
+			return ret;
+
+		dln2->speed = speed;
+	}
+
+	if (dln2->mode != mode) {
+		ret = dln2_spi_set_mode(dln2, mode & 0x3);
+		if (ret < 0)
+			return ret;
+
+		dln2->mode = mode;
+	}
+
+	if (dln2->bpw != bpw) {
+		ret = dln2_spi_set_bpw(dln2, bpw);
+		if (ret < 0)
+			return ret;
+
+		dln2->bpw = bpw;
+	}
+
+	ret = dln2_spi_enable(dln2, true);
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+static int dln2_spi_transfer_one_message(struct spi_master *master,
+					 struct spi_message *msg)
+{
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+	struct spi_device *spi = msg->spi;
+	struct spi_transfer *xfer;
+	int status;
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		u8 attr = 0;
+
+		status = dln2_spi_transfer_setup(dln2, xfer->speed_hz,
+						 xfer->bits_per_word,
+						 spi->mode);
+		if (status < 0) {
+			dev_err(&dln2->pdev->dev, "Cannot setup transfer\n");
+			break;
+		}
+
+		if (!xfer->cs_change &&
+		    !list_is_last(&xfer->transfer_list, &msg->transfers))
+			attr = xfer->cs_change ? 0 : DLN2_SPI_ATTR_LEAVE_SS_LOW;
+
+		status = dln2_spi_rdwr(dln2, xfer->tx_buf, xfer->rx_buf,
+				       xfer->len, attr);
+		if (status < 0) {
+			dev_err(&dln2->pdev->dev, "write/read failed!\n");
+			break;
+		}
+
+		status = 0;
+	}
+
+	msg->status = status;
+	spi_finalize_current_message(master);
+	return status;
+}
+
+static int dln2_spi_probe(struct platform_device *pdev)
+{
+	struct spi_master *master;
+	struct dln2_spi *dln2;
+	struct dln2_platform_data *pdata = dev_get_platdata(&pdev->dev);
+	int ret;
+
+	master = spi_alloc_master(&pdev->dev, sizeof(*dln2));
+	if (!master)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, master);
+
+	dln2 = spi_master_get_devdata(master);
+
+	dln2->buf = devm_kmalloc(&pdev->dev, DLN2_SPI_BUF_SIZE, GFP_KERNEL);
+	if (!dln2->buf) {
+		ret = -ENOMEM;
+		goto exit_free_master;
+	}
+
+	dln2->master = master;
+	dln2->pdev = pdev;
+	dln2->port = pdata->port;
+	/* cs number can never be 0xff, so the first transfer will set the cs */
+	dln2->cs = 0xff;
+
+	/* disable SPI module before continuing with the setup */
+	ret = dln2_spi_enable(dln2, false);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to disable SPI module\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_get_cs_num(dln2, &master->num_chipselect);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to get number of CS pins\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_get_speed_range(dln2,
+				       &master->min_speed_hz,
+				       &master->max_speed_hz);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to read bus min/max freqs\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_get_supported_frame_sizes(dln2,
+						 &master->bits_per_word_mask);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to read supported frame sizes\n");
+		goto exit_free_master;
+	}
+
+	ret = dln2_spi_cs_enable_all(dln2, true);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to enable CS pins\n");
+		goto exit_free_master;
+	}
+
+	master->bus_num = -1;
+	master->mode_bits = SPI_CPOL | SPI_CPHA;
+	master->prepare_message = dln2_spi_prepare_message;
+	master->transfer_one_message = dln2_spi_transfer_one_message;
+
+	/* enable SPI module, we're good to go */
+	ret = dln2_spi_enable(dln2, true);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to enable SPI module\n");
+		goto exit_free_master;
+	}
+
+	ret = devm_spi_register_master(&pdev->dev, master);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to register master\n");
+		goto exit_register;
+	}
+
+	return ret;
+
+exit_register:
+	if (dln2_spi_enable(dln2, false) < 0)
+		dev_err(&pdev->dev, "Failed to disable SPI module\n");
+exit_free_master:
+	spi_master_put(master);
+
+	return ret;
+}
+
+static int dln2_spi_remove(struct platform_device *pdev)
+{
+	struct spi_master *master = spi_master_get(platform_get_drvdata(pdev));
+	struct dln2_spi *dln2 = spi_master_get_devdata(master);
+
+	if (dln2_spi_enable(dln2, false) < 0)
+		dev_err(&pdev->dev, "Failed to disable SPI module\n");
+
+	return 0;
+}
+
+static struct platform_driver spi_dln2_driver = {
+	.driver.name	= "dln2-spi",
+	.probe		= dln2_spi_probe,
+	.remove		= dln2_spi_remove,
+};
+module_platform_driver(spi_dln2_driver);
+
+MODULE_DESCRIPTION("Driver for the Diolan DLN2 SPI master interface");
+MODULE_AUTHOR("Laurentiu Palcu <laurentiu.palcu@intel.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_ALIAS("platform:dln2-spi");