Skip to contentSkip to conversation

PXA DMA-capable PATA driver

PATCH Superseded Marek Vasut 3 participants 6 messages May 17, 2010 – May 19, 2010

Subsystems (4)

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 PXA DMA-capable PATA driver

This patch adds a driver for a harddrive attached to PXA address and data bus. Unlike pata_platform, this driver allows usage of PXA DMA controller, making the transmission speed 3x higher.

---
 arch/arm/mach-pxa/include/mach/pata_pxa.h |   33 +++
 drivers/ata/Kconfig                       |   11 +
 drivers/ata/Makefile                      |    1 +
 drivers/ata/pata_pxa.c                    |  404 +++++++++++++++++++++++++++++
 4 files changed, 449 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/mach-pxa/include/mach/pata_pxa.h
 create mode 100644 drivers/ata/pata_pxa.c
arch/arm/mach-pxa/include/mach/pata_pxa.h new file +33 -0
new file mode 100644
index 0000000..6cf7df1
@@ -0,0 +1,33 @@
+/*
+ * Generic PXA PATA driver
+ *
+ * Copyright (C) 2010 Marek Vasut <[email protected]>
+ *
+ *  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, 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; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef	__MACH_PATA_PXA_H__
+#define	__MACH_PATA_PXA_H__
+
+struct pata_pxa_pdata {
+	/* PXA DMA DREQ<0:2> pin */
+	uint32_t	dma_dreq;
+	/* Register shift */
+	uint32_t	reg_shift;
+	/* IRQ flags */
+	uint32_t	irq_flags;
+};
+
+#endif	/* __MACH_PATA_PXA_H__ */
drivers/ata/Kconfig +11 -0
index 01c52c4..5cd3e8c 100644
@@ -684,6 +684,17 @@ config PATA_VIA
 
 	  If unsure, say N.
 
+config PATA_PXA
+	tristate "PXA DMA-capable PATA support"
+	depends on ARCH_PXA
+	help
+	  This option enables support for harddrive attached to PXA CPU's bus.
+	  
+	  NOTE: This driver utilizes PXA DMA controller, in case your hardware
+	        is not capable of doing MWDMA, use pata_platform instead.
+
+	  If unsure, say N.
+
 config PATA_WINBOND
 	tristate "Winbond SL82C105 PATA support"
 	depends on PCI
drivers/ata/Makefile +1 -0
index fc936d4..5ecf45a 100644
@@ -79,6 +79,7 @@ obj-$(CONFIG_PATA_PLATFORM)	+= pata_platform.o
 obj-$(CONFIG_PATA_AT91)	+= pata_at91.o
 obj-$(CONFIG_PATA_OF_PLATFORM)	+= pata_of_platform.o
 obj-$(CONFIG_PATA_ICSIDE)	+= pata_icside.o
+obj-$(CONFIG_PATA_PXA)		+= pata_pxa.o
 # Should be last but two libata driver
 obj-$(CONFIG_PATA_ACPI)		+= pata_acpi.o
 # Should be last but one libata driver
drivers/ata/pata_pxa.c new file +404 -0

404 lines not on this page · raw message

Signature · 1 line
1.7.0
#1 lore raw

PATCH v3 1/1 PXA DMA-capable PATA driver

Dne ?t 18. kv?tna 2010 01:09:28 Marek Vasut napsal(a):

Quoted 3 lines

This patch adds a driver for a harddrive attached to PXA address and data bus. Unlike pata_platform, this driver allows usage of PXA DMA controller, making the transmission speed 3x higher.

Hey guys, all the issues pointed out should be addressed. See below. Cheers!

Quoted 158 lines
---
 arch/arm/mach-pxa/include/mach/pata_pxa.h |   33 +++
 drivers/ata/Kconfig                       |   11 +
 drivers/ata/Makefile                      |    1 +
 drivers/ata/pata_pxa.c                    |  404
+++++++++++++++++++++++++++++ 4 files changed, 449 insertions(+), 0
deletions(-)
 create mode 100644 arch/arm/mach-pxa/include/mach/pata_pxa.h
 create mode 100644 drivers/ata/pata_pxa.c
(quoted diff) +0 -0

b/arch/arm/mach-pxa/include/mach/pata_pxa.h new file mode 100644 index 0000000..6cf7df1

arch/arm/mach-pxa/include/mach/pata_pxa.h new file +33 -0
@@ -0,0 +1,33 @@
+/*
+ * Generic PXA PATA driver
+ *
+ * Copyright (C) 2010 Marek Vasut <[email protected]>
+ *
+ *  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, 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; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef	__MACH_PATA_PXA_H__
+#define	__MACH_PATA_PXA_H__
+
+struct pata_pxa_pdata {
+	/* PXA DMA DREQ<0:2> pin */
+	uint32_t	dma_dreq;
+	/* Register shift */
+	uint32_t	reg_shift;
+	/* IRQ flags */
+	uint32_t	irq_flags;
+};
+
+#endif	/* __MACH_PATA_PXA_H__ */
drivers/ata/Kconfig +11 -0
index 01c52c4..5cd3e8c 100644
@@ -684,6 +684,17 @@ config PATA_VIA
 
 	  If unsure, say N.
 
+config PATA_PXA
+	tristate "PXA DMA-capable PATA support"
+	depends on ARCH_PXA
+	help
+	  This option enables support for harddrive attached to PXA CPU's bus.
+
+	  NOTE: This driver utilizes PXA DMA controller, in case your hardware
+	        is not capable of doing MWDMA, use pata_platform instead.
+
+	  If unsure, say N.
+
 config PATA_WINBOND
 	tristate "Winbond SL82C105 PATA support"
 	depends on PCI
drivers/ata/Makefile +1 -0
index fc936d4..5ecf45a 100644
@@ -79,6 +79,7 @@ obj-$(CONFIG_PATA_PLATFORM)	+= pata_platform.o
 obj-$(CONFIG_PATA_AT91)	+= pata_at91.o
 obj-$(CONFIG_PATA_OF_PLATFORM)	+= pata_of_platform.o
 obj-$(CONFIG_PATA_ICSIDE)	+= pata_icside.o
+obj-$(CONFIG_PATA_PXA)		+= pata_pxa.o
 # Should be last but two libata driver
 obj-$(CONFIG_PATA_ACPI)		+= pata_acpi.o
 # Should be last but one libata driver
drivers/ata/pata_pxa.c new file +58 -0
new file mode 100644
index 0000000..3f3071e
@@ -0,0 +1,404 @@
+/*
+ * Generic PXA PATA driver
+ *
+ * Copyright (C) 2010 Marek Vasut <[email protected]>
+ *
+ *  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, 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; see the file COPYING.  If not, write to
+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/ata.h>
+#include <linux/libata.h>
+#include <linux/platform_device.h>
+#include <linux/gpio.h>
+#include <linux/slab.h>
+#include <linux/completion.h>
+
+#include <scsi/scsi_host.h>
+
+#include <mach/pxa2xx-regs.h>
+#include <mach/pata_pxa.h>
+#include <mach/dma.h>
+
+#define DRV_NAME	"pata_pxa"
+#define DRV_VERSION	"0.1"
+
+struct pata_pxa_data {
+	uint32_t		dma_channel;
+	struct pxa_dma_desc	*dma_desc;
+	dma_addr_t		dma_desc_addr;
+	uint32_t		dma_desc_id;
+
+	/* DMA IO physical address */
+	uint32_t		dma_io_addr;
+	/* PXA DREQ<0:2> pin selector */
+	uint32_t		dma_dreq;
+	/* DMA DCSR register value */
+	uint32_t		dma_dcsr;
+
+	struct completion	dma_done;
+};
+
+/*
+ * Setup the DMA descriptors. The size is transfer capped at 4k per

descriptor, + * if the transfer is longer, it is split into multiple chained descriptors. + */
+static void pxa_load_dmac(struct scatterlist *sg, struct ata_queued_cmd *qc) +{

drivers/ata/pata_pxa.c new file +5 -0
+	struct pata_pxa_data *pd = qc->ap->private_data;
+
+	uint32_t cpu_len, seg_len;
+	dma_addr_t cpu_addr;
+

drivers/ata/pata_pxa.c

Quoted 10 lines
drivers/ata/pata_pxa.c new file +10 -0
+	cpu_addr = sg_dma_address(sg);
+	cpu_len = sg_dma_len(sg);
+
+	do {
+		seg_len = (cpu_len > 0x1000) ? 0x1000 : cpu_len;
+
+		pd->dma_desc[pd->dma_desc_id].ddadr = pd->dma_desc_addr +
+			((pd->dma_desc_id + 1) * sizeof(struct pxa_dma_desc));
+
+		pd->dma_desc[pd->dma_desc_id].dcmd = DCMD_BURST32 | DCMD_WIDTH2 

|

Quoted 44 lines
drivers/ata/pata_pxa.c new file +44 -0
+					(DCMD_LENGTH & seg_len);
+
+		if (qc->tf.flags & ATA_TFLAG_WRITE) {
+			pd->dma_desc[pd->dma_desc_id].dsadr = cpu_addr;
+			pd->dma_desc[pd->dma_desc_id].dtadr = pd->dma_io_addr;
+			pd->dma_desc[pd->dma_desc_id].dcmd |= DCMD_INCSRCADDR |
+						DCMD_FLOWTRG;
+		} else {
+			pd->dma_desc[pd->dma_desc_id].dsadr = pd->dma_io_addr;
+			pd->dma_desc[pd->dma_desc_id].dtadr = cpu_addr;
+			pd->dma_desc[pd->dma_desc_id].dcmd |= DCMD_INCTRGADDR |
+						DCMD_FLOWSRC;
+		}
+
+		cpu_len -= seg_len;
+		cpu_addr += seg_len;
+		pd->dma_desc_id++;
+
+	} while(cpu_len);
+
+	/* Should not happen */
+	if (seg_len & 0x1f)
+		DALGN |= (1 << pd->dma_dreq);
+}
+
+/*
+ * Prepare taskfile for submission.
+ */
+static void pxa_qc_prep(struct ata_queued_cmd *qc)
+{
+	struct pata_pxa_data *pd = qc->ap->private_data;
+	int si = 0;
+	struct scatterlist *sg;
+
+	if (!(qc->flags & ATA_QCFLAG_DMAMAP))
+		return;
+
+	pd->dma_desc_id = 0;
+
+	DCSR(pd->dma_channel) = 0;
+	DALGN &= ~(1 << pd->dma_dreq);
+
+	for_each_sg(qc->sg, sg, qc->n_elem, si)
+		pxa_load_dmac(sg, qc);

drivers/ata/pata_pxa.c line 131

Quoted 10 lines
drivers/ata/pata_pxa.c new file +10 -0
+
+	pd->dma_desc[pd->dma_desc_id - 1].ddadr = DDADR_STOP;
+
+	/* Fire IRQ only at the end of last block */
+	pd->dma_desc[pd->dma_desc_id - 1].dcmd |= DCMD_ENDIRQEN;
+
+	DDADR(pd->dma_channel) = pd->dma_desc_addr;
+	DRCMR(pd->dma_dreq) = DRCMR_MAPVLD | pd->dma_channel;
+
+}

Jeff, this is now moved.

Quoted 38 lines
drivers/ata/pata_pxa.c new file +3 -0
+
+/*
+ * Configure the DMA controller, load the DMA descriptors, but don't start

the + * DMA controller yet. Only issue the ATA command.
+ */

drivers/ata/pata_pxa.c new file +31 -0
+static void pxa_bmdma_setup(struct ata_queued_cmd *qc)
+{
+	qc->ap->ops->sff_exec_command(qc->ap, &qc->tf);
+}
+
+/*
+ * Execute the DMA transfer.
+ */
+static void pxa_bmdma_start(struct ata_queued_cmd *qc)
+{
+	struct pata_pxa_data *pd = qc->ap->private_data;
+	init_completion(&pd->dma_done);
+	DCSR(pd->dma_channel) = DCSR_RUN;
+}
+
+/*
+ * Wait until the DMA transfer completes, then stop the DMA controller.
+ */
+static void pxa_bmdma_stop(struct ata_queued_cmd *qc)
+{
+	struct pata_pxa_data *pd = qc->ap->private_data;
+
+	if ((DCSR(pd->dma_channel) & DCSR_RUN) &&
+		wait_for_completion_timeout(&pd->dma_done, HZ))
+		dev_err(qc->ap->dev, "Timeout waiting for DMA completion!");
+
+	DCSR(pd->dma_channel) = 0;
+}
+
+/*
+ * Read DMA status. The bmdma_stop() will take care of properly finishing

the + * DMA transfer so we always have DMA-complete interrupt here.
+ */

drivers/ata/pata_pxa.c line 179

Quoted 10 lines
drivers/ata/pata_pxa.c new file +10 -0
+static unsigned char pxa_bmdma_status(struct ata_port *ap)
+{
+	struct pata_pxa_data *pd = ap->private_data;
+	unsigned char ret = ATA_DMA_INTR;
+
+	if (pd->dma_dcsr & DCSR_BUSERR)
+		ret |= ATA_DMA_ERR;
+
+	return ret;
+}

Yes Robert, you were right. Fixed. Thanks

Quoted 12 lines
drivers/ata/pata_pxa.c new file +10 -0
+
+/*
+ * No IRQ register present so we do nothing.
+ */
+static void pxa_irq_clear(struct ata_port *ap)
+{
+}
+
+/*
+ * Check for ATAPI DMA. ATAPI DMA is unsupported by this driver. It's

still + * unclear why ATAPI has DMA issues.
+ */

drivers/ata/pata_pxa.c line 196

Quoted 5 lines
drivers/ata/pata_pxa.c new file +5 -0
+static int pxa_check_atapi_dma(struct ata_queued_cmd *qc)
+{
+	return -EOPNOTSUPP;
+}
+

See libata-core.c:

4766 /**

4767  *      atapi_check_dma - Check whether ATAPI DMA can be supported
4768  *      @qc: Metadata associated with taskfile to check
4769  *
4770  *      Allow low-level driver to filter ATA PACKET commands, returning
4771  *      a status indicating whether or not it is OK to use DMA for the
4772  *      supplied PACKET command.
4773  *
4774  *      LOCKING:
4775  *      spin_lock_irqsave(host lock)
4776  *
4777  *      RETURNS: 0 when ATAPI DMA can be used
4778  *               nonzero otherwise
4779  */

And since ATAPI DMA can NOT be used and -EOPNOTSUPP is nonzero, this is fine.

Quoted 208 lines
drivers/ata/pata_pxa.c new file +208 -0
+static struct scsi_host_template pxa_ata_sht = {
+	ATA_BMDMA_SHT(DRV_NAME),
+};
+
+static struct ata_port_operations pxa_ata_port_ops = {
+	.inherits		= &ata_bmdma_port_ops,
+	.cable_detect		= ata_cable_40wire,
+
+	.bmdma_setup		= pxa_bmdma_setup,
+	.bmdma_start		= pxa_bmdma_start,
+	.bmdma_stop		= pxa_bmdma_stop,
+	.bmdma_status		= pxa_bmdma_status,
+
+	.check_atapi_dma	= pxa_check_atapi_dma,
+
+	.sff_irq_clear		= pxa_irq_clear,
+
+	.qc_prep		= pxa_qc_prep,
+};
+
+/*
+ * DMA interrupt handler.
+ */
+static void pxa_ata_dma_irq(int dma, void *port)
+{
+	struct ata_port *ap = port;
+	struct pata_pxa_data *pd = ap->private_data;
+
+	pd->dma_dcsr = DCSR(dma);
+	DCSR(dma) = pd->dma_dcsr;
+
+	if (pd->dma_dcsr & DCSR_STOPSTATE)
+		complete(&pd->dma_done);
+}
+
+static int __devinit pxa_ata_probe(struct platform_device *pdev)
+{
+	struct ata_host *host;
+	struct ata_port *ap;
+	struct pata_pxa_data *data;
+	struct resource *cmd_res;
+	struct resource *ctl_res;
+	struct resource *dma_res;
+	struct resource *irq_res;
+	struct pata_pxa_pdata *pdata = pdev->dev.platform_data;
+
+	/*
+	 * Resource validation, three resources are needed:
+	 *  - CMD port base address
+	 *  - CTL port base address
+	 *  - DMA port base address
+	 *  - IRQ pin
+	 */
+	if (pdev->num_resources != 4) {
+		dev_err(&pdev->dev, "invalid number of resources\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * CMD port base address
+	 */
+	cmd_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (unlikely(cmd_res == NULL))
+		return -EINVAL;
+
+	/*
+	 * CTL port base address
+	 */
+	ctl_res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (unlikely(ctl_res == NULL))
+		return -EINVAL;
+
+	/*
+	 * DMA port base address
+	 */
+	dma_res = platform_get_resource(pdev, IORESOURCE_DMA, 0);
+	if (unlikely(dma_res == NULL))
+		return -EINVAL;
+
+	/*
+	 * IRQ pin
+	 */
+	irq_res = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
+	if (unlikely(irq_res == NULL))
+		return -EINVAL;
+
+	/*
+	 * Allocate the host
+	 */
+	host = ata_host_alloc(&pdev->dev, 1);
+	if (!host)
+		return -ENOMEM;
+
+	ap		= host->ports[0];
+	ap->ops		= &pxa_ata_port_ops;
+	ap->pio_mask	= ATA_PIO4;
+	ap->mwdma_mask	= ATA_MWDMA2;
+	ap->flags	= ATA_FLAG_MMIO;
+
+	ap->ioaddr.cmd_addr	= devm_ioremap(&pdev->dev, cmd_res->start,
+						resource_size(cmd_res));
+	ap->ioaddr.ctl_addr	= devm_ioremap(&pdev->dev, ctl_res->start,
+						resource_size(ctl_res));
+	ap->ioaddr.bmdma_addr	= devm_ioremap(&pdev->dev, dma_res->start,
+						resource_size(dma_res));
+
+	/*
+	 * Adjust register offsets
+	 */
+	ap->ioaddr.altstatus_addr = ap->ioaddr.ctl_addr;
+	ap->ioaddr.data_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_DATA << pdata->reg_shift);
+	ap->ioaddr.error_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_ERR << pdata->reg_shift);
+	ap->ioaddr.feature_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_FEATURE << pdata->reg_shift);
+	ap->ioaddr.nsect_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_NSECT << pdata->reg_shift);
+	ap->ioaddr.lbal_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_LBAL << pdata->reg_shift);
+	ap->ioaddr.lbam_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_LBAM << pdata->reg_shift);
+	ap->ioaddr.lbah_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_LBAH << pdata->reg_shift);
+	ap->ioaddr.device_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_DEVICE << pdata->reg_shift);
+	ap->ioaddr.status_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_STATUS << pdata->reg_shift);
+	ap->ioaddr.command_addr	= ap->ioaddr.cmd_addr +
+					(ATA_REG_CMD << pdata->reg_shift);
+
+	/*
+	 * Allocate and load driver's internal data structure
+	 */
+	data = devm_kzalloc(&pdev->dev, sizeof(struct pata_pxa_data),
+								GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	ap->private_data = data;
+	data->dma_dreq = pdata->dma_dreq;
+	data->dma_io_addr = dma_res->start;
+
+	/*
+	 * Request the DMA channel
+	 */
+	data->dma_channel = pxa_request_dma(DRV_NAME, DMA_PRIO_LOW,
+						pxa_ata_dma_irq, ap);
+	if (data->dma_channel < 0)
+		return -EBUSY;
+
+	/*
+	 * Stop and clear the DMA channel
+	 */
+	DCSR(data->dma_channel) = 0;
+
+	/*
+	 * Allocate space for the DMA descriptors
+	 */
+	data->dma_desc = dmam_alloc_coherent(&pdev->dev, PAGE_SIZE,
+					&data->dma_desc_addr, GFP_KERNEL);
+
+	/*
+	 * Activate the ATA host
+	 */
+	return ata_host_activate(host, irq_res->start, ata_sff_interrupt,
+				pdata->irq_flags, &pxa_ata_sht);
+}
+
+static int __devexit pxa_ata_remove(struct platform_device *pdev)
+{
+	struct ata_host *host = dev_get_drvdata(&pdev->dev);
+	struct pata_pxa_data *data = host->ports[0]->private_data;
+
+	pxa_free_dma(data->dma_channel);
+
+	ata_host_detach(host);
+
+	return 0;
+}
+
+static struct platform_driver pxa_ata_driver = {
+	.probe		= pxa_ata_probe,
+	.remove		= __devexit_p(pxa_ata_remove),
+	.driver		= {
+		.name		= DRV_NAME,
+		.owner		= THIS_MODULE,
+	},
+};
+
+static int __init pxa_ata_init(void)
+{
+	return platform_driver_register(&pxa_ata_driver);
+}
+
+static void __exit pxa_ata_exit(void)
+{
+	platform_driver_unregister(&pxa_ata_driver);
+}
+
+module_init(pxa_ata_init);
+module_exit(pxa_ata_exit);
+
+MODULE_AUTHOR("Marek Vasut <[email protected]>");
+MODULE_DESCRIPTION("DMA-capable driver for PATA on PXA CPU");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+MODULE_ALIAS("platform:" DRV_NAME);
maintainer
#2 lore raw

PATCH v3 1/1 PXA DMA-capable PATA driver

Quoted 7 lines · Marek Vasut

Dne ?t 18. kv?tna 2010 01:09:28 Marek Vasut napsal(a):

Quoted 3 lines

This patch adds a driver for a harddrive attached to PXA address and data bus. Unlike pata_platform, this driver allows usage of PXA DMA controller, making the transmission speed 3x higher.

Hey guys, all the issues pointed out should be addressed. See below. Cheers!

For the first iteration, you can have my :

For the next iteration, it would be good to :
- fix the error path in pxa_ata_probe()
=> if dmam_alloc_coherent() fails
=> if pxa_request_dma() fails
- in pxa_ata_remove()
=> maybe call devm_iounmap()

--
Robert

#3 lore raw

PATCH v3 1/1 PXA DMA-capable PATA driver

Dne St 19. kv?tna 2010 13:19:14 Robert Jarzmik napsal(a):

Quoted 10 lines
Quoted 7 lines · Marek Vasut

Dne ?t 18. kv?tna 2010 01:09:28 Marek Vasut napsal(a):

Quoted 3 lines

This patch adds a driver for a harddrive attached to PXA address and data bus. Unlike pata_platform, this driver allows usage of PXA DMA controller, making the transmission speed 3x higher.

Hey guys, all the issues pointed out should be addressed. See below. Cheers!

For the first iteration, you can have my :

Quoted 9 lines

For the next iteration, it would be good to :
- fix the error path in pxa_ata_probe()
=> if dmam_alloc_coherent() fails
=> if pxa_request_dma() fails
- in pxa_ata_remove()
=> maybe call devm_iounmap()

It'd be awesome if you could just point out all the issues at once ... about devm_x(), you probably don't need to, it's 'managed'.

Quoted 2 lines

--
Robert

#4 lore raw

PATCH v3 1/1 PXA DMA-capable PATA driver

In reply to Marek Vasut
Quoted 5 lines · Marek Vasut

This patch adds a driver for a harddrive attached to PXA address and data bus. Unlike pata_platform, this driver allows usage of PXA DMA controller, making the transmission speed 3x higher.

Quoted 8 lines
---
  arch/arm/mach-pxa/include/mach/pata_pxa.h |   33 +++
  drivers/ata/Kconfig                       |   11 +
  drivers/ata/Makefile                      |    1 +
  drivers/ata/pata_pxa.c                    |  404 +++++++++++++++++++++++++++++
  4 files changed, 449 insertions(+), 0 deletions(-)

create mode 100644 arch/arm/mach-pxa/include/mach/pata_pxa.h create mode 100644 drivers/ata/pata_pxa.c

There are a few minor issues remaining, but we can take care of them post-merge. I presume you'll send this via arm tree.

#5 lore raw

PATCH v3 1/1 PXA DMA-capable PATA driver

Dne ?t 18. kv?tna 2010 05:56:38 Jeff Garzik napsal(a):

Quoted 17 lines · Marek Vasut
Quoted 7 lines

This patch adds a driver for a harddrive attached to PXA address and data bus. Unlike pata_platform, this driver allows usage of PXA DMA controller, making the transmission speed 3x higher.

---

Quoted 8 lines
  arch/arm/mach-pxa/include/mach/pata_pxa.h |   33 +++
  drivers/ata/Kconfig                       |   11 +
  drivers/ata/Makefile                      |    1 +
  drivers/ata/pata_pxa.c                    |  404

+++++++++++++++++++++++++++++ 4 files changed, 449 insertions(+), 0 deletions(-)
create mode 100644 arch/arm/mach-pxa/include/mach/pata_pxa.h create mode 100644 drivers/ata/pata_pxa.c

Quoted 4 lines

There are a few minor issues remaining, but we can take care of them post-merge. I presume you'll send this via arm tree.

If you're fine with it, I guess that'd be the best way.

Thanks!