Re: [PATCH v3 2/2] phy: cygnus: pcie: Add Cygnus PCIe PHY support

[Ray Jui]

On 9/17/2015 11:22 PM, Kishon Vijay Abraham I wrote:
> Hi,
> 
> On Tuesday 15 September 2015 11:26 PM, Ray Jui wrote:
>> This patch adds the PCIe PHY support for the Broadcom PCIe RC interface
>> on Cygnus
>>
>> Signed-off-by: Ray Jui <rjui@xxxxxxxxxxxx>
>> Reviewed-by: Arun Parameswaran <aparames@xxxxxxxxxxxx>
>> Reviewed-by: JD (Jiandong) Zheng <jdzheng@xxxxxxxxxxxx>
>> Reviewed-by: Scott Branden <sbranden@xxxxxxxxxxxx>
>> ---
>>  drivers/phy/Kconfig               |   9 ++
>>  drivers/phy/Makefile              |   1 +
>>  drivers/phy/phy-bcm-cygnus-pcie.c | 198 ++++++++++++++++++++++++++++++++++++++
>>  3 files changed, 208 insertions(+)
>>  create mode 100644 drivers/phy/phy-bcm-cygnus-pcie.c
>>
>> diff --git a/drivers/phy/Kconfig b/drivers/phy/Kconfig
>> index 47da573..947bae2 100644
>> --- a/drivers/phy/Kconfig
>> +++ b/drivers/phy/Kconfig
>> @@ -371,4 +371,13 @@ config PHY_BRCMSTB_SATA
>>  	  Enable this to support the SATA3 PHY on 28nm Broadcom STB SoCs.
>>  	  Likely useful only with CONFIG_SATA_BRCMSTB enabled.
>>  
>> +config PHY_CYGNUS_PCIE
>> +	tristate "Broadcom Cygnus PCIe PHY driver"
>> +	depends on OF && (ARCH_BCM_CYGNUS || COMPILE_TEST)
>> +	select GENERIC_PHY
>> +	default ARCH_BCM_CYGNUS
>> +	help
>> +	  Enable this to support the Broadcom Cygnus PCIe PHY.
>> +	  If unsure, say N.
>> +
>>  endmenu
>> diff --git a/drivers/phy/Makefile b/drivers/phy/Makefile
>> index a5b18c1..fdce78d 100644
>> --- a/drivers/phy/Makefile
>> +++ b/drivers/phy/Makefile
>> @@ -46,3 +46,4 @@ obj-$(CONFIG_PHY_QCOM_UFS) 	+= phy-qcom-ufs-qmp-14nm.o
>>  obj-$(CONFIG_PHY_TUSB1210)		+= phy-tusb1210.o
>>  obj-$(CONFIG_PHY_BRCMSTB_SATA)		+= phy-brcmstb-sata.o
>>  obj-$(CONFIG_PHY_PISTACHIO_USB)		+= phy-pistachio-usb.o
>> +obj-$(CONFIG_PHY_CYGNUS_PCIE)		+= phy-bcm-cygnus-pcie.o
>> diff --git a/drivers/phy/phy-bcm-cygnus-pcie.c b/drivers/phy/phy-bcm-cygnus-pcie.c
>> new file mode 100644
>> index 0000000..0fb6750
>> --- /dev/null
>> +++ b/drivers/phy/phy-bcm-cygnus-pcie.c
>> @@ -0,0 +1,198 @@
>> +/*
>> + * Copyright (C) 2015 Broadcom 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.
>> + *
>> + * This program is distributed "as is" WITHOUT ANY WARRANTY of any
>> + * kind, whether express or implied; without even the implied warranty
>> + * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> + * GNU General Public License for more details.
>> + */
>> +
>> +#include <linux/delay.h>
>> +#include <linux/io.h>
>> +#include <linux/module.h>
>> +#include <linux/of.h>
>> +#include <linux/phy/phy.h>
>> +#include <linux/platform_device.h>
>> +
>> +#define PCIE_CFG_OFFSET         0x00
>> +#define PCIE1_PHY_IDDQ_SHIFT    10
>> +#define PCIE0_PHY_IDDQ_SHIFT    2
>> +
>> +enum cygnus_pcie_phy_id {
>> +	CYGNUS_PHY_PCIE0 = 0,
>> +	CYGNUS_PHY_PCIE1,
>> +	MAX_NUM_PHYS,
>> +};
>> +
>> +struct cygnus_pcie_phy_core;
>> +
>> +/**
>> + * struct cygnus_pcie_phy - Cygnus PCIe PHY device
>> + * @core: pointer to the Cygnus PCIe PHY core control
>> + * @id: internal ID to identify the Cygnus PCIe PHY
>> + * @phy: pointer to the kernel PHY device
>> + */
>> +struct cygnus_pcie_phy {
>> +	struct cygnus_pcie_phy_core *core;
>> +	enum cygnus_pcie_phy_id id;
>> +	struct phy *phy;
>> +};
>> +
>> +/**
>> + * struct cygnus_pcie_phy_core - Cygnus PCIe PHY core control
>> + * @dev: pointer to device
>> + * @base: base register
>> + * @lock: mutex to protect access to individual PHYs
>> + * @phys: pointer to Cygnus PHY device
>> + */
>> +struct cygnus_pcie_phy_core {
>> +	struct device *dev;
>> +	void __iomem *base;
>> +	struct mutex lock;
>> +	struct cygnus_pcie_phy phys[MAX_NUM_PHYS];
>> +};
>> +
>> +static int cygnus_pcie_power_config(struct cygnus_pcie_phy *phy, bool enable)
>> +{
>> +	struct cygnus_pcie_phy_core *core = phy->core;
>> +	unsigned shift;
>> +	u32 val;
>> +
>> +	mutex_lock(&core->lock);
>> +
>> +	switch (phy->id) {
>> +	case CYGNUS_PHY_PCIE0:
>> +		shift = PCIE0_PHY_IDDQ_SHIFT;
>> +		break;
>> +
>> +	case CYGNUS_PHY_PCIE1:
>> +		shift = PCIE1_PHY_IDDQ_SHIFT;
>> +		break;
>> +
>> +	default:
>> +		mutex_unlock(&core->lock);
>> +		dev_err(core->dev, "PCIe PHY id%d invalid\n", phy->id);
>> +		return -EINVAL;
>> +	}
>> +
>> +	if (enable) {
>> +		val = readl(core->base + PCIE_CFG_OFFSET);
>> +		val &= ~BIT(shift);
>> +		writel(val, core->base + PCIE_CFG_OFFSET);
>> +		msleep(50);
>> +	} else {
>> +		val = readl(core->base + PCIE_CFG_OFFSET);
>> +		val |= BIT(shift);
>> +		writel(val, core->base + PCIE_CFG_OFFSET);
>> +	}
>> +
>> +	mutex_unlock(&core->lock);
>> +	dev_info(core->dev, "PCIe PHY %s\n", enable ? "enabled" : "disabled");
>> +	return 0;
>> +}
>> +
>> +static int cygnus_pcie_phy_power_on(struct phy *p)
>> +{
>> +	struct cygnus_pcie_phy *phy = phy_get_drvdata(p);
>> +
>> +	return cygnus_pcie_power_config(phy, true);
>> +}
>> +
>> +static int cygnus_pcie_phy_power_off(struct phy *p)
>> +{
>> +	struct cygnus_pcie_phy *phy = phy_get_drvdata(p);
>> +
>> +	return cygnus_pcie_power_config(phy, false);
>> +}
>> +
>> +static struct phy_ops cygnus_pcie_phy_ops = {
>> +	.power_on = cygnus_pcie_phy_power_on,
>> +	.power_off = cygnus_pcie_phy_power_off,
> 
> .owner is required here.

Okay will add that.

>> +};
>> +
>> +static struct phy *cygnus_pcie_phy_xlate(struct device *dev,
>> +					 struct of_phandle_args *args)
>> +{
>> +	struct cygnus_pcie_phy_core *core;
>> +	int id;
>> +
>> +	core = dev_get_drvdata(dev);
>> +	if (!core)
>> +		return ERR_PTR(-EINVAL);
>> +
>> +	id = args->args[0];
>> +
>> +	if (WARN_ON(id >= MAX_NUM_PHYS))
>> +		return ERR_PTR(-ENODEV);
>> +
>> +	return core->phys[id].phy;
>> +}
>> +
>> +static int cygnus_pcie_phy_probe(struct platform_device *pdev)
>> +{
>> +	struct device *dev = &pdev->dev;
>> +	struct cygnus_pcie_phy_core *core;
>> +	struct phy_provider *provider;
>> +	struct resource *res;
>> +	int i = 0;
>> +
>> +	core = devm_kzalloc(dev, sizeof(*core), GFP_KERNEL);
>> +	if (!core)
>> +		return -ENOMEM;
>> +
>> +	core->dev = dev;
>> +
>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>> +	core->base = devm_ioremap_resource(dev, res);
>> +	if (IS_ERR(core->base))
>> +		return PTR_ERR(core->base);
>> +
>> +	mutex_init(&core->lock);
>> +
>> +	for (i = 0; i < MAX_NUM_PHYS; i++) {
> 
> Don't have to hard code MAX_NUM_PHYS. This can be modelled in dt itself.
> Create a single PHY provider node and model each individual PHYs as the
> child node of the PHY provider.

Ok. I'll make the change. Note MAX_NUM_PHYS will stay, just to check
against the max number of PCIe PHYs that can be supported. But each
individual PHY will be modeled by a child node, and #phy-cells become zero.

>> +		struct cygnus_pcie_phy *p = &core->phys[i];
>> +
>> +		p->phy = devm_phy_create(dev, NULL, &cygnus_pcie_phy_ops);
>> +		if (IS_ERR(p->phy)) {
>> +			dev_err(dev, "failed to create PHY\n");
>> +			return PTR_ERR(p->phy);
>> +		}
>> +
>> +		p->core = core;
>> +		p->id = i;
>> +		phy_set_drvdata(p->phy, p);
>> +	}
>> +
>> +	dev_set_drvdata(dev, core);
>> +
>> +	provider = devm_of_phy_provider_register(dev, cygnus_pcie_phy_xlate);
> 
> this custom xlate function is not required at all.

Yes, now #phy-cells is zero, I can use of_phy_simple_xlate.

> 
> Thanks
> Kishon
> 

Thanks,

Ray
--
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/