RE: [PATCH] ahci: added a new driver for supporting Freescale AHCI sata

[Yuantian Tang]

Hi,

> -----Original Message-----
> From: Hans de Goede [mailto:hdegoede@xxxxxxxxxx]
> Sent: Wednesday, September 02, 2015 4:32 PM
> To: Tang Yuantian-B29983 <Yuantian.Tang@xxxxxxxxxxxxx>
> Cc: tj@xxxxxxxxxx; linux-ide@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
> Subject: Re: [PATCH] ahci: added a new driver for supporting Freescale AHCI
> sata
> 
> Hi,
> 
> On 02-09-15 04:25, Yuantian.Tang@xxxxxxxxxxxxx wrote:
> > From: Tang Yuantian <Yuantian.Tang@xxxxxxxxxxxxx>
> >
> > Currently Freescale QorIQ series SATA is supported by ahci_platform
> > driver. Some SoC specific settings have been put in uboot. So whether
> > SATA works or not heavily depends on uboot.
> > This patch will add a new driver to support QorIQ sata which removes
> > the dependency on any other boot loader.
> > Freescale QorIQ series sata, like ls1021a ls2085a ls1043a, is
> > compatible with serial ATA 3.0 and AHCI 1.3 specification.
> >
> > Signed-off-by: Yuantian Tang <Yuantian.Tang@xxxxxxxxxxxxx>
> 
> Thanks for the patch looks good overall.
> 
> You need to add a Documentation/devicetree/bindings/ata/ahci-fsl-qoriq.txt
> (or a similar named file) documenting the compatible strings and what the
> devicetree nodes should contain wrt reg, interrupts, etc.
> properties. See Documentation/devicetree/bindings/ata/ahci-platform.txt
> as an example.
> 
> Further comments inline.
> 
I was about to use ahci_platform driver, so I added the bindings stuff to
Documentation/devicetree/bindings/ata/ahci-platform.txt
So I need to revert the old bingings first and then add a new one.

> > ---
> >   drivers/ata/Kconfig         |   9 ++
> >   drivers/ata/Makefile        |   1 +
> >   drivers/ata/ahci_platform.c |   1 -
> >   drivers/ata/ahci_qoriq.c    | 308
> ++++++++++++++++++++++++++++++++++++++++++++
> >   4 files changed, 318 insertions(+), 1 deletion(-)
> >   create mode 100644 drivers/ata/ahci_qoriq.c
> >
> > diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig index
> > 15e40ee..6aaa3f8 100644
> > --- a/drivers/ata/Kconfig
> > +++ b/drivers/ata/Kconfig
> > @@ -175,6 +175,15 @@ config AHCI_XGENE
> >   	help
> >   	 This option enables support for APM X-Gene SoC SATA host
> controller.
> >
> > +config AHCI_QORIQ
> > +	tristate "Freescale QorIQ AHCI SATA support"
> > +	depends on OF
> > +	help
> > +	  This option enables support for the Freescale QorIQ AHCI SoC's
> > +	  onboard AHCI SATA.
> > +
> > +	  If unsure, say N.
> > +
> >   config SATA_FSL
> >   	tristate "Freescale 3.0Gbps SATA support"
> >   	depends on FSL_SOC
> > diff --git a/drivers/ata/Makefile b/drivers/ata/Makefile index
> > af70919..af45eff 100644
> > --- a/drivers/ata/Makefile
> > +++ b/drivers/ata/Makefile
> > @@ -19,6 +19,7 @@ obj-$(CONFIG_AHCI_SUNXI)	+= ahci_sunxi.o
> libahci.o libahci_platform.o
> >   obj-$(CONFIG_AHCI_ST)		+= ahci_st.o libahci.o
> libahci_platform.o
> >   obj-$(CONFIG_AHCI_TEGRA)	+= ahci_tegra.o libahci.o
> libahci_platform.o
> >   obj-$(CONFIG_AHCI_XGENE)	+= ahci_xgene.o libahci.o
> libahci_platform.o
> > +obj-$(CONFIG_AHCI_QORIQ)	+= ahci_qoriq.o libahci.o
> libahci_platform.o
> >
> >   # SFF w/ custom DMA
> >   obj-$(CONFIG_PDC_ADMA)		+= pdc_adma.o
> > diff --git a/drivers/ata/ahci_platform.c b/drivers/ata/ahci_platform.c
> > index 1befb11..04975b8 100644
> > --- a/drivers/ata/ahci_platform.c
> > +++ b/drivers/ata/ahci_platform.c
> > @@ -76,7 +76,6 @@ static const struct of_device_id ahci_of_match[] = {
> >   	{ .compatible = "ibm,476gtr-ahci", },
> >   	{ .compatible = "snps,dwc-ahci", },
> >   	{ .compatible = "hisilicon,hisi-ahci", },
> > -	{ .compatible = "fsl,qoriq-ahci", },
> >   	{},
> >   };
> >   MODULE_DEVICE_TABLE(of, ahci_of_match);
> 
> This will break booting new kernels with old dtb files, something which in
> general is considered a big non-no, I suggest adding a comment that this has
> been superseded by the new ahci_qoriq.c code, and maybe add a date to
> retire the compatible in say a year from now.
> 
There is no old dtb because LS* platforms are not been upstreamed yet.
So I think we can safely replace it without breaking anything.

Regards,
Yuantian

> > diff --git a/drivers/ata/ahci_qoriq.c b/drivers/ata/ahci_qoriq.c new
> > file mode 100644 index 0000000..943b783
> > --- /dev/null
> > +++ b/drivers/ata/ahci_qoriq.c
> > @@ -0,0 +1,308 @@
> > +/*
> > + * Freescale QorIQ AHCI SATA platform driver
> > + *
> > + * Copyright 2015 Freescale, Inc.
> > + *   Tang Yuantian <Yuantian.Tang@xxxxxxxxxxxxx>
> > + *
> > + * 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.
> > + */
> > +
> > +#include <linux/kernel.h>
> > +#include <linux/module.h>
> > +#include <linux/pm.h>
> > +#include <linux/ahci_platform.h>
> > +#include <linux/device.h>
> > +#include <linux/of_address.h>
> > +#include <linux/of.h>
> > +#include <linux/of_device.h>
> > +#include <linux/platform_device.h>
> > +#include <linux/libata.h>
> > +#include "ahci.h"
> > +
> > +#define DRV_NAME "ahci-qoriq"
> > +
> > +#define AHCI_PORT_PHY_1_CFG	0xa003fffe
> > +#define AHCI_PORT_PHY_2_CFG	0x28183411
> > +#define AHCI_PORT_PHY_3_CFG	0x0e081004
> > +#define AHCI_PORT_PHY_4_CFG	0x00480811
> > +#define AHCI_PORT_PHY_5_CFG	0x192c96a4
> > +#define AHCI_PORT_TRANS_CFG	0x08000025
> > +
> > +#define SATA_ECC_REG_ADDR	0x20220520
> > +#define SATA_ECC_DISABLE	0x00020000
> > +
> > +enum ahci_qoriq_type {
> > +	AHCI_LS1021A,
> > +	AHCI_LS1043A,
> > +	AHCI_LS2085A,
> > +};
> > +
> > +struct ahci_qoriq_priv {
> > +	struct ccsr_ahci *reg_base;
> > +	enum ahci_qoriq_type type;
> > +	void __iomem *ecc_addr;
> > +};
> > +
> > +/* AHCI (sata) register map */
> > +struct ccsr_ahci {
> > +	u32 res1[0xa4/4];	/* 0x0 - 0xa4 */
> > +	u32 pcfg;	/* port config */
> > +	u32 ppcfg;	/* port phy1 config */
> > +	u32 pp2c;	/* port phy2 config */
> > +	u32 pp3c;	/* port phy3 config */
> > +	u32 pp4c;	/* port phy4 config */
> > +	u32 pp5c;	/* port phy5 config */
> > +	u32 paxic;	/* port AXI config */
> > +	u32 axicc;	/* AXI cache control */
> > +	u32 axipc;	/* AXI PROT control */
> > +	u32 ptc;	/* port Trans Config */
> > +	u32 pts;	/* port Trans Status */
> > +	u32 plc;	/* port link config */
> > +	u32 plc1;	/* port link config1 */
> > +	u32 plc2;	/* port link config2 */
> > +	u32 pls;	/* port link status */
> > +	u32 pls1;	/* port link status1 */
> > +	u32 pcmdc;	/* port CMD config */
> > +	u32 ppcs;	/* port phy control status */
> > +	u32 pberr;	/* port 0/1 BIST error */
> > +	u32 cmds;	/* port 0/1 CMD status error */
> > +};
> > +
> > +static const struct of_device_id ahci_qoriq_of_match[] = {
> > +	{ .compatible = "fsl,ls1021a-ahci", .data = (void *)AHCI_LS1021A},
> > +	{ .compatible = "fsl,ls1043a-ahci", .data = (void *)AHCI_LS1043A},
> > +	{ .compatible = "fsl,ls2085a-ahci", .data = (void *)AHCI_LS2085A},
> > +	{},
> > +};
> > +MODULE_DEVICE_TABLE(of, ahci_qoriq_of_match);
> > +
> > +static int ahci_qoriq_hardreset(struct ata_link *link, unsigned int *class,
> > +			  unsigned long deadline)
> > +{
> > +	const unsigned long *timing = sata_ehc_deb_timing(&link-
> >eh_context);
> > +	void __iomem *port_mmio = ahci_port_base(link->ap);
> > +	u32 px_cmd, px_is, px_val;
> > +	struct ata_port *ap = link->ap;
> > +	struct ahci_port_priv *pp = ap->private_data;
> > +	struct ahci_host_priv *hpriv = ap->host->private_data;
> > +	struct ahci_qoriq_priv *qoriq_priv = hpriv->plat_data;
> > +	u8 *d2h_fis = pp->rx_fis + RX_FIS_D2H_REG;
> > +	struct ata_taskfile tf;
> > +	bool online;
> > +	int rc;
> > +
> > +	DPRINTK("ENTER\n");
> > +
> > +	ahci_stop_engine(ap);
> > +
> > +	/*
> > +	 * There is a errata on ls1021a Rev1.0 and Rev2.0 which is:
> > +	 * A-009042: The device detection initialization sequence
> > +	 * mistakenly resets some registers.
> > +	 *
> > +	 * Workaround for this is:
> > +	 * The software should read and store PxCMD and PxIS values
> > +	 * before issuing the device detection initialization sequence.
> > +	 * After the sequence is complete, software should restore the
> > +	 * PxCMD and PxIS with the stored values.
> > +	 */
> > +	if (qoriq_priv->type == AHCI_LS1021A) {
> > +		px_cmd = readl(port_mmio + PORT_CMD);
> > +		px_is = readl(port_mmio + PORT_IRQ_STAT);
> > +	}
> > +
> > +	/* clear D2H reception area to properly wait for D2H FIS */
> > +	ata_tf_init(link->device, &tf);
> > +	tf.command = ATA_BUSY;
> > +	ata_tf_to_fis(&tf, 0, 0, d2h_fis);
> > +
> > +	rc = sata_link_hardreset(link, timing, deadline, &online,
> > +				 ahci_check_ready);
> > +
> > +	/* restore the PxCMD and PxIS on ls1021 */
> > +	if (qoriq_priv->type == AHCI_LS1021A) {
> > +		px_val = readl(port_mmio + PORT_CMD);
> > +		if (px_val != px_cmd)
> > +			writel(px_cmd, port_mmio + PORT_CMD);
> > +
> > +		px_val = readl(port_mmio + PORT_IRQ_STAT);
> > +		if (px_val != px_is)
> > +			writel(px_is, port_mmio + PORT_IRQ_STAT);
> > +	}
> > +
> > +	hpriv->start_engine(ap);
> > +
> > +	if (online)
> > +		*class = ahci_dev_classify(ap);
> > +
> > +	DPRINTK("EXIT, rc=%d, class=%u\n", rc, *class);
> > +	return rc;
> > +}
> > +
> > +static struct ata_port_operations ahci_qoriq_ops = {
> > +	.inherits	= &ahci_ops,
> > +	.hardreset	= ahci_qoriq_hardreset,
> > +};
> > +
> > +static const struct ata_port_info ahci_qoriq_port_info = {
> > +	.flags		= AHCI_FLAG_COMMON | ATA_FLAG_NCQ,
> > +	.pio_mask	= ATA_PIO4,
> > +	.udma_mask	= ATA_UDMA6,
> > +	.port_ops	= &ahci_qoriq_ops,
> > +};
> > +
> > +static struct scsi_host_template ahci_qoriq_sht = {
> > +	AHCI_SHT(DRV_NAME),
> > +};
> > +
> > +static int ahci_qoriq_phy_init(struct ahci_qoriq_priv *qpriv) {
> > +	struct ccsr_ahci *reg_base = qpriv->reg_base;
> > +
> > +	switch (qpriv->type) {
> > +	case AHCI_LS1021A:
> > +		writel(SATA_ECC_DISABLE, qpriv->ecc_addr);
> > +		writel(AHCI_PORT_PHY_1_CFG, &reg_base->ppcfg);
> > +		writel(AHCI_PORT_PHY_2_CFG, &reg_base->pp2c);
> > +		writel(AHCI_PORT_PHY_3_CFG, &reg_base->pp3c);
> > +		writel(AHCI_PORT_PHY_4_CFG, &reg_base->pp4c);
> > +		writel(AHCI_PORT_PHY_5_CFG, &reg_base->pp5c);
> > +		writel(AHCI_PORT_TRANS_CFG, &reg_base->ptc);
> > +		break;
> > +
> > +	case AHCI_LS1043A:
> > +	case AHCI_LS2085A:
> > +		writel(AHCI_PORT_PHY_1_CFG, &reg_base->ppcfg);
> > +		break;
> > +	}
> > +
> > +	return 0;
> > +}
> > +
> > +static int ahci_qoriq_probe(struct platform_device *pdev) {
> > +	struct device_node *np = pdev->dev.of_node;
> > +	struct device *dev = &pdev->dev;
> > +	struct ahci_host_priv *hpriv;
> > +	struct ahci_qoriq_priv *qoriq_priv;
> > +	const struct of_device_id *of_id;
> > +	int rc;
> > +
> > +	hpriv = ahci_platform_get_resources(pdev);
> > +	if (IS_ERR(hpriv))
> > +		return PTR_ERR(hpriv);
> > +
> > +	rc = ahci_platform_enable_resources(hpriv);
> > +	if (rc)
> > +		return rc;
> 
> You may want to move this down in the function to remove unnecessary
> goto disable_foo error handling, you can do this directly before the phy_init.
> 
> > +
> > +	of_id = of_match_node(ahci_qoriq_of_match, np);
> > +	if (!of_id) {
> > +		rc = -ENODEV;
> > +		goto disable_resources;
> > +	}
> > +
> > +	qoriq_priv = devm_kzalloc(dev, sizeof(*qoriq_priv), GFP_KERNEL);
> > +	if (!qoriq_priv) {
> > +		rc = -ENOMEM;
> > +		goto disable_resources;
> > +	}
> > +
> > +	qoriq_priv->reg_base = of_iomap(np, 0);
> > +	if (!qoriq_priv->reg_base) {
> > +		rc = -ENOMEM;
> > +		goto free_qpriv;
> > +	}
> 
> You should use devm_ioremap_resource() to map registers so that
> resources get released automatically.
> 
> In this case hower you already have access to there registers through
> hpriv->mmio, so there is no need to map them a second time.
> 
> > +
> > +	qoriq_priv->type = (enum ahci_qoriq_type)of_id->data;
> > +
> > +	if (qoriq_priv->type == AHCI_LS1021A) {
> > +		qoriq_priv->ecc_addr =
> > +			ioremap((phys_addr_t)SATA_ECC_REG_ADDR, 4);
> 
> In a devicetree enabled driver you should never hardcode register addresses
> like this, instead they should be specified in the reg property of the driver.
> 
> You should put something like this in the dts:
> 
> 	reg = <0x01c13400 0x10>, <0x01c14800 0x4>;
> 	reg-names = "ahci", "sata_ecc";
> 
> And then in the code do:
> 
> 	struct resource *res = platform_get_resource_byname(pdev,
> IORESOURCE_MEM, "sata_ecc");
> 	qoriq_priv->ecc_addr = devm_ioremap_resource(dev, res);
> 	if (IS_ERR(qoriq_priv->ecc_addr))
> 		return PTR_ERR(qoriq_priv->ecc_addr);
> 
> Note the direct return. This is ok to do if you move enable_resources down
> as discussed below.
> 
> > +		if (!qoriq_priv->ecc_addr) {
> > +			rc = -ENOMEM;
> > +			goto err_iomap;
> > +		}
> > +	}
> > +	hpriv->plat_data = qoriq_priv;
> > +	rc = ahci_qoriq_phy_init(qoriq_priv);
> > +	if (rc) {
> > +		rc = -ENOMEM;
> > +		goto err_iomap2;
> > +	}
> > +
> > +	rc = ahci_platform_init_host(pdev, hpriv, &ahci_qoriq_port_info,
> > +				     &ahci_qoriq_sht);
> > +	if (rc)
> > +		goto err_iomap2;
> > +
> > +	return 0;
> > +
> > +err_iomap2:
> > +	iounmap(qoriq_priv->ecc_addr);
> > +
> > +err_iomap:
> > +	iounmap(qoriq_priv->reg_base);
> > +
> > +free_qpriv:
> > +	devm_kfree(dev, qoriq_priv);
> 
> All these 3 labels + code can go away since devm will do this automatically
> when you fail the probe method (assuming you use
> devm_ioremap_resource as discussed above).
> 
> > +
> > +disable_resources:
> > +	ahci_platform_disable_resources(hpriv);
> > +
> > +	return rc;
> > +}
> > +
> > +#ifdef CONFIG_PM_SLEEP
> > +static int ahci_qoriq_resume(struct device *dev) {
> > +	struct ata_host *host = dev_get_drvdata(dev);
> > +	struct ahci_host_priv *hpriv = host->private_data;
> > +	int rc;
> > +
> > +	rc = ahci_platform_enable_resources(hpriv);
> > +	if (rc)
> > +		return rc;
> > +
> > +	rc = ahci_qoriq_phy_init(hpriv->plat_data);
> > +	if (rc)
> > +		goto disable_resources;
> > +
> > +	rc = ahci_platform_resume_host(dev);
> > +	if (rc)
> > +		goto disable_resources;
> > +
> > +	/* We resumed so update PM runtime state */
> > +	pm_runtime_disable(dev);
> > +	pm_runtime_set_active(dev);
> > +	pm_runtime_enable(dev);
> > +
> > +	return 0;
> > +
> > +disable_resources:
> > +	ahci_platform_disable_resources(hpriv);
> > +
> > +	return rc;
> > +}
> > +#endif
> > +
> > +static SIMPLE_DEV_PM_OPS(ahci_qoriq_pm_ops,
> ahci_platform_suspend,
> > +			 ahci_qoriq_resume);
> > +
> > +static struct platform_driver ahci_qoriq_driver = {
> > +	.probe = ahci_qoriq_probe,
> > +	.remove = ata_platform_remove_one,
> > +	.driver = {
> > +		.name = DRV_NAME,
> > +		.of_match_table = ahci_qoriq_of_match,
> > +		.pm = &ahci_qoriq_pm_ops,
> > +	},
> > +};
> > +module_platform_driver(ahci_qoriq_driver);
> > +
> > +MODULE_DESCRIPTION("Freescale QorIQ AHCI SATA platform driver");
> > +MODULE_AUTHOR("Tang Yuantian <Yuantian.Tang@xxxxxxxxxxxxx>");
> > +MODULE_LICENSE("GPL");
> >
> 
> 
> Regards,
> 
> Hans
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/