Re: [PATCH] acpi:apd:add AMD ACPI2Platform device support for x86 system.

From: Mika Westerberg
Date: Thu Nov 27 2014 - 06:47:11 EST


On Wed, Nov 26, 2014 at 06:31:38PM +0800, Ken Xue wrote:
> On ä, 2014-11-24 at 02:47 +0100, Rafael J. Wysocki wrote:
> > On Monday, November 24, 2014 01:02:30 AM Xue, Ken wrote:
> > >
> > > On Tuesday, November 18, 2014 01:58:11 PM Ken Xue wrote:
> > > > This new feature is to interpret AMD specific ACPI device to platform
> > > > device such as I2C, UART found on AMD CZ and later chipsets. It is
> > > > based on example INTEL LPSS. Now, it can support AMD I2C & UART.
> > > >
> > > > Signed-off-by: Ken Xue <Ken.Xue@xxxxxxx>
> > > > Signed-off-by: Jeff Wu <Jeff.Wu@xxxxxxx>
> > >
> > > Generally speaking, this seems to duplicate much code from acpi_lpss which should be re-used instead. What about moving the code that will be common between acpi_lpss and the new driver into a new file (say acpi_soc.c)?
> > >
> > > Also, you need to avoid automatic creation of platform devices when !X86_AMD_PLATFORM_DEVICE in analogy with what acpi_lpss does, or bad things will happen.
> > >
> > > [ken] sounds fair enough. Let me take action to merge drivers to acpi_soc.c ? or you have other plan?
> >
> > I'd prefer the common code to reside in one file (or one .c file and one header
> > file), and the driver-specific code to stay in separate per-driver files.
> >
> [Ken] I wrote a proto type for acpi_soc.c. please help to confirm if it
> can match your ideal. if yes, i will submit a new patch after do more
> test and refine codes. I think it will impact lpss driver greatly, even
> i have taken it into account. below codes is for acpi_soc.c.

In general looks good. I have few comments though.

>
> >From fc323fb7b3b4cbb79bda05ce3b1d6d8dfe5e883b Mon Sep 17 00:00:00 2001
> From: Ken Xue <Ken.Xue@xxxxxxx>
> Date: Wed, 26 Nov 2014 17:15:30 +0800
> Subject: [PATCH] This is proto type for acpi_soc.
>
> Signed-off-by: Ken Xue <Ken.Xue@xxxxxxx>
> ---
> arch/x86/Kconfig | 11 +++
> drivers/acpi/Makefile | 2 +-
> drivers/acpi/acpi_apd.c | 123 ++++++++++++++++++++++++++++
> drivers/acpi/acpi_soc.c | 213
> ++++++++++++++++++++++++++++++++++++++++++++++++

This is line-wrapped, please make sure when you submit the actual patch
that it is formatted properly. Also you need proper changelog etc.

> drivers/acpi/acpi_soc.h | 91 +++++++++++++++++++++
> drivers/acpi/internal.h | 6 ++
> drivers/acpi/scan.c | 1 +
> 7 files changed, 446 insertions(+), 1 deletion(-)
> create mode 100755 drivers/acpi/acpi_apd.c
> create mode 100755 drivers/acpi/acpi_soc.c
> create mode 100755 drivers/acpi/acpi_soc.h
>
> diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
> index ded8a67..6402c79f 100644
> --- a/arch/x86/Kconfig
> +++ b/arch/x86/Kconfig
> @@ -495,6 +495,17 @@ config X86_INTEL_LPSS
> things like clock tree (common clock framework) and pincontrol
> which are needed by the LPSS peripheral drivers.
>
> +config X86_AMD_PLATFORM_DEVICE
> + bool "AMD ACPI2Platform devices support"
> + depends on ACPI
> + select COMMON_CLK
> + select PINCTRL
> + ---help---
> + Select to interpret AMD specific ACPI device to platform device
> + such as I2C, UART found on AMD CARRIZO and later chipset. Selecting
> + this option enables things like clock tree (common clock framework)
> + and pinctrl.
> +
> config IOSF_MBI
> tristate "Intel SoC IOSF Sideband support for SoC platforms"
> depends on PCI
> diff --git a/drivers/acpi/Makefile b/drivers/acpi/Makefile
> index c3b2fcb..b07003a 100644
> --- a/drivers/acpi/Makefile
> +++ b/drivers/acpi/Makefile
> @@ -40,7 +40,7 @@ acpi-$(CONFIG_ARCH_MIGHT_HAVE_ACPI_PDC) +=
> processor_pdc.o
> acpi-y += ec.o
> acpi-$(CONFIG_ACPI_DOCK) += dock.o
> acpi-y += pci_root.o pci_link.o pci_irq.o
> -acpi-y += acpi_lpss.o
> +acpi-y += acpi_soc.o acpi_lpss.o acpi_apd.o
> acpi-y += acpi_platform.o
> acpi-y += acpi_pnp.o
> acpi-y += int340x_thermal.o
> diff --git a/drivers/acpi/acpi_apd.c b/drivers/acpi/acpi_apd.c
> new file mode 100755
> index 0000000..c7e916b
> --- /dev/null
> +++ b/drivers/acpi/acpi_apd.c
> @@ -0,0 +1,123 @@
> +/*
> + * AMD ACPI support for ACPI2platform device.
> + *
> + * Copyright (c) 2014, AMD Corporation.
> + * Authors: Ken Xue <Ken.Xue@xxxxxxx>
> + * Wu, Jeff <Jeff.Wu@xxxxxxx>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + */
> +
> +#include <linux/acpi.h>
> +#include <linux/clk.h>
> +#include <linux/clkdev.h>
> +#include <linux/clk-provider.h>
> +#include <linux/err.h>
> +#include <linux/io.h>
> +#include <linux/platform_device.h>
> +#include <linux/pm_runtime.h>
> +
> +#include "internal.h"
> +#include "acpi_soc.h"
> +
> +struct acpi_soc asoc;

static?

Also Is "asoc" good name? That might get confused with Alsa SoC (ASoC).

> +
> +#ifdef X86_AMD_PLATFORM_DEVICE
> +static int acpi_apd_setup(struct acpi_soc_dev_private_data *pdata)
> +{
> + struct acpi_soc_dev_desc *dev_desc = pdata->dev_desc;
> + struct clk *clk = ERR_PTR(-ENODEV);
> +
> + if (dev_desc->clk)
> + return 0;
> +
> + if (dev_desc->fixed_clk_rate) {
> + clk = clk_register_fixed_rate(&pdata->adev->dev,
> + dev_name(&pdata->adev->dev),
> + NULL, CLK_IS_ROOT, dev_desc->rate);
> + dev_desc->clk = clk;
> + clk_register_clkdev(clk, NULL, dev_name(&pdata->adev->dev));
> + }
> +
> + return 0;
> +}
> +
> +static struct acpi_soc_dev_desc cz_i2c_desc = {
> + .setup = acpi_apd_setup;
> + .fixed_clk_rate = 133000000,

Oh, good so now we can get rid the hack we did for
i2c-designware-platdrv.c with this commit:

a445900c906092f i2c: designware: Add support for AMD I2C controller

Since now you have means to pass clock to the driver.

Are you going to handle that driver as well?

> +};
> +
> +static struct acpi_soc_dev_desc cz_uart_desc = {
> + .setup = acpi_apd_setup;
> + .fixed_clk_rate = 48000000,
> +};
> +
> +#else
> +
> +#define APD_ADDR(desc) (0UL)
> +
> +#endif /* CONFIG_X86_INTEL_LPSS */
> +
> +static struct acpi_device_id acpi_apd_device_ids[] = {

const?

> + /* Generic apd devices */
> + { "AMD0010", APD_ADDR(cz_i2c_desc) },
> + { "AMD0020", APD_ADDR(cz_uart_desc) },
> + { }
> +};
> +
> +#ifdef X86_AMD_PLATFORM_DEVICE
> +
> +static ssize_t apd_device_desc_show(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + int ret;
> + struct acpi_device *adev;
> + struct acpi_soc_dev_private_data *pdata;
> + struct acpi_soc_dev_desc *dev_desc;
> +
> + ret = acpi_bus_get_device(ACPI_HANDLE(dev), &adev);
> + if (WARN_ON(ret))
> + return ret;
> +
> + pdata = acpi_driver_data(adev);
> + if (WARN_ON(!pdata || !pdata->dev_desc))
> + return -ENODEV;
> +
> + dev_desc = pdata->dev_desc;
> + if (dev_desc->fixed_clk_rate)
> + return sprintf(buf, "Required fix rate clk %s: %ld\n",
> + dev_desc->clk->name,
> + dev_desc->fixed_clk_rate);
> + else
> + return sprintf(buf, "No need clk\n");
> +}
> +
> +static DEVICE_ATTR(device_desc, S_IRUSR, apd_device_desc_show, NULL);
> +
> +static struct attribute *apd_attrs[] = {
> + &dev_attr_device_desc.attr,
> + NULL,
> +};
> +
> +static struct attribute_group apd_attr_group = {
> + .attrs = apd_attrs,
> + .name = "apd",
> +};

This requires updating sysfs ABI but then again, do you really need the
above? And does it belong to sysfs in the first place?

> +
> +void __init acpi_apd_init(void)
> +{
> + asoc.ids = acpi_apd_device_ids;
> + asoc.attr_group = &apd_attr_group;
> + register_acpi_soc(&asoc, true);
> +}
> +
> +#else
> +
> +void __init acpi_apd_init(void)
> +{
> + asoc.ids = acpi_apd_device_ids;
> + register_acpi_soc(&asoc, false);
> +}
> +#endif
> diff --git a/drivers/acpi/acpi_soc.c b/drivers/acpi/acpi_soc.c
> new file mode 100755
> index 0000000..99df1ab
> --- /dev/null
> +++ b/drivers/acpi/acpi_soc.c
> @@ -0,0 +1,213 @@
> +/*
> + * ACPI SOC support for Intel Lynxpoint LPSS and AMD APD.
> + *
> + * Copyright (C) 2015, Intel Corporation & AMD Corporation
> + * Authors: Ken Xue <Ken.Xue@xxxxxxx>
> + * Mika Westerberg <mika.westerberg@xxxxxxxxxxxxxxx>
> + * Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + */
> +
> +#include <linux/acpi.h>
> +#include <linux/err.h>
> +#include <linux/io.h>
> +#include <linux/platform_device.h>
> +#include <linux/pm_runtime.h>
> +#include <linux/delay.h>
> +#include <linux/pm_domain.h>
> +
> +#include "internal.h"
> +#include "acpi_soc.h"
> +
> +

Delete the extra blank line

> +ACPI_MODULE_NAME("acpi_soc");
> +
> +/*A list for all acpi soc device*/
> +static LIST_HEAD(asoc_list);
> +
> +static int is_memory(struct acpi_resource *res, void *not_used)
> +{
> + struct resource r;
> + return !acpi_dev_resource_memory(res, &r);
> +}
> +
> +static int acpi_soc_create_device(struct acpi_device *adev,
> + const struct acpi_device_id *id)
> +{
> + struct acpi_soc_dev_desc *dev_desc;
> + struct acpi_soc_dev_private_data *pdata;
> + struct resource_list_entry *rentry;
> + struct list_head resource_list;
> + struct platform_device *pdev;
> + int ret;
> +
> + dev_desc = (struct acpi_soc_dev_desc *)id->driver_data;
> + if (!dev_desc) {
> + pdev = acpi_create_platform_device(adev);
> + return IS_ERR_OR_NULL(pdev) ? PTR_ERR(pdev) : 1;
> + }
> + pdata = kzalloc(sizeof(*pdata), GFP_KERNEL);
> + if (!pdata)
> + return -ENOMEM;
> +
> + INIT_LIST_HEAD(&resource_list);
> + ret = acpi_dev_get_resources(adev, &resource_list, is_memory, NULL);
> + if (ret < 0)
> + goto err_out;
> +
> + list_for_each_entry(rentry, &resource_list, node)
> + if (resource_type(&rentry->res) == IORESOURCE_MEM) {
> + if (dev_desc->mem_size_override)
> + pdata->mmio_size = dev_desc->mem_size_override;
> + else
> + pdata->mmio_size = resource_size(&rentry->res);
> + pdata->mmio_base = ioremap(rentry->res.start,
> + pdata->mmio_size);
> + break;
> + }
> +
> + acpi_dev_free_resource_list(&resource_list);
> +
> + pdata->dev_desc = dev_desc;
> +
> + /*setup device by a hook routine*/

The comment should look like this

/* Setup device by hook routine */

but I think it does not provide any new information so you can just drop
it.

> + if (dev_desc->setup)
> + dev_desc->setup(pdata);
> +
> + /*
> + * This works around a known issue in ACPI tables where acpi soc
> devices
> + * have _PS0 and _PS3 without _PSC (and no power resources), so
> + * acpi_bus_init_power() will assume that the BIOS has put them into
> D0.
> + */
> + ret = acpi_device_fix_up_power(adev);
> + if (ret) {
> + /* Skip the device, but continue the namespace scan. */
> + ret = 0;
> + goto err_out;
> + }
> +
> + adev->driver_data = pdata;
> + pdev = acpi_create_platform_device(adev);
> + if (!IS_ERR_OR_NULL(pdev)) {
> + return 1;
> + }
> +
> + ret = PTR_ERR(pdev);
> + adev->driver_data = NULL;
> +
> + err_out:
> + kfree(pdata);
> + return ret;
> +}
> +
> +static int acpi_soc_platform_notify(struct notifier_block *nb,
> + unsigned long action, void *data)
> +{
> + struct platform_device *pdev = to_platform_device(data);
> + struct acpi_soc_dev_private_data *pdata;
> + struct acpi_device *adev;
> + struct acpi_soc *asoc_entry;
> + const struct acpi_device_id *id = NULL;
> +
> + list_for_each_entry(asoc_entry, &asoc_list, list) {
> + id = acpi_match_device(asoc_entry->ids, &pdev->dev);
> + if (!id)
> + break;
> + }
> +
> + if (!id || !id->driver_data)
> + return 0;
> +
> + if (acpi_bus_get_device(ACPI_HANDLE(&pdev->dev), &adev))
> + return 0;
> +
> + pdata = acpi_driver_data(adev);
> + if (!pdata || !pdata->mmio_base)
> + return 0;
> +
> + switch (action) {
> + case BUS_NOTIFY_BOUND_DRIVER:
> + if ((pdata->dev_desc->flags & ACPI_SOC_PM)){
> + if (asoc_entry->pm_domain)
> + pdev->dev.pm_domain = asoc_entry->pm_domain;
> + else if (pdata->dev_desc->flags & ACPI_SOC_PM_ON)
> + dev_pm_domain_attach(&pdev->dev, true);
> + else
> + dev_pm_domain_attach(&pdev->dev, false);
> + }
> + break;
> + case BUS_NOTIFY_UNBOUND_DRIVER:
> + if ((pdata->dev_desc->flags & ACPI_SOC_PM)){
> + if (asoc_entry->pm_domain)
> + pdev->dev.pm_domain = asoc_entry->pm_domain;
> + else if (pdata->dev_desc->flags & ACPI_SOC_PM_ON)
> + dev_pm_domain_detach(&pdev->dev, true);
> + else
> + dev_pm_domain_detach(&pdev->dev, false);
> + }
> + break;
> + case BUS_NOTIFY_ADD_DEVICE:
> + if ((pdata->dev_desc->flags & ACPI_SOC_SYSFS) &&
> asoc_entry->pm_domain)
> + return sysfs_create_group(&pdev->dev.kobj,
> + asoc_entry->attr_group);
> + case BUS_NOTIFY_DEL_DEVICE:
> + if ((pdata->dev_desc->flags & ACPI_SOC_SYSFS) &&
> asoc_entry->pm_domain)
> + sysfs_remove_group(&pdev->dev.kobj, asoc_entry->attr_group);
> + default:
> + break;
> + }
> +
> + return 0;
> +}
> +
> +static struct notifier_block acpi_soc_nb = {
> + .notifier_call = acpi_soc_platform_notify,
> +};
> +
> +static void acpi_soc_bind(struct device *dev)
> +{
> + struct acpi_soc_dev_private_data *pdata;
> +
> + pdata = acpi_driver_data(ACPI_COMPANION(dev));
> +
> + if (!pdata || !pdata->dev_desc || !pdata->dev_desc->bind)
> + return;
> +
> + pdata->dev_desc->bind(pdata);
> + return;
> +}
> +
> +static void acpi_soc_unbind(struct device *dev)
> +{
> + struct acpi_soc_dev_private_data *pdata;
> +
> + pdata = acpi_driver_data(ACPI_COMPANION(dev));
> +
> + if (!pdata || !pdata->dev_desc || !pdata->dev_desc->unbind)
> + return;
> +
> + pdata->dev_desc->unbind(pdata);
> + return;
> +}
> +
> +void register_acpi_soc(struct acpi_soc *asoc, bool
> disable_scan_handler)
> +{
> + struct acpi_scan_handler *acpi_soc_handler;
> +
> + INIT_LIST_HEAD(&asoc->list);
> + list_add(&asoc->list, &asoc_list);
> +
> + acpi_soc_handler = kzalloc(sizeof(struct acpi_scan_handler),
> GFP_KERNEL);
> + acpi_soc_handler->ids = asoc->ids;
> + if (!disable_scan_handler) {
> + acpi_soc_handler->attach = acpi_soc_create_device;
> + acpi_soc_handler->bind = acpi_soc_bind;
> + acpi_soc_handler->unbind = acpi_soc_unbind;
> + }
> + acpi_scan_add_handler(acpi_soc_handler);
> + bus_register_notifier(&platform_bus_type, &acpi_soc_nb);
> +}
> +EXPORT_SYMBOL_GPL(register_acpi_soc);

I don't think we want to export these to modules.

> diff --git a/drivers/acpi/acpi_soc.h b/drivers/acpi/acpi_soc.h
> new file mode 100755
> index 0000000..96db30e
> --- /dev/null
> +++ b/drivers/acpi/acpi_soc.h

And all this belongs to drivers/acpi/internal.h.

> @@ -0,0 +1,91 @@
> +/*
> + * ACPI SOC support for Intel Lynxpoint LPSS and AMD APD.
> + *
> + * Copyright (C) 2015, Intel Corporation & AMD Corporation
> + * Authors: Ken Xue <Ken.Xue@xxxxxxx>
> + * Mika Westerberg <mika.westerberg@xxxxxxxxxxxxxxx>
> + * Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + */
> +#ifndef _ACPI_SOC_H
> +#define _ACPI_SOC_H
> +
> +#include <linux/acpi.h>
> +#include <linux/clk.h>
> +#include <linux/pm.h>
> +
> +/* Flags */
> +#define ACPI_SOC_SYSFS BIT(0)
> +#define ACPI_SOC_PM BIT(1)
> +#define ACPI_SOC_PM_ON BIT(2)
> +
> +struct acpi_soc_dev_private_data;
> +
> +/**
> + * struct acpi_soc - acpi soc
> + * @list: list head
> + * @ids: all acpi device ids for acpi soc
> + * @pm_domain: power domain for all acpi device;can be NULL
> + * @attr_group: attribute group for sysfs support of acpi soc;can be
> NULL
> + */
> +struct acpi_soc {
> + struct list_head list;
> + struct acpi_device_id *ids;
> + struct dev_pm_domain *pm_domain;
> + struct attribute_group *attr_group;
> +};
> +
> +/**
> + * struct acpi_soc_dev_desc - a descriptor for acpi device
> + * @flags: some device feature flags
> + * @clk: clock device
> + * @fixed_clk_rate: fixed rate input clock source for acpi device;
> + * 0 means no fixed rate input clock source
> + * @mem_size_override: a workaround for override device memsize;
> + * 0 means no needs for this WA
> + * @setup: a hook routine to set device resource during create platform
> device
> + * @bind: a hook of acpi_scan_handler.bind
> + * @unbind: a hook of acpi_scan_handler.unbind
> + *
> + *device description defined as acpi_device_id.driver_data
> + */
> +struct acpi_soc_dev_desc {
> + unsigned int flags;
> + struct clk *clk;
> + unsigned int fixed_clk_rate;
> + size_t mem_size_override;
> + int (*setup)(struct acpi_soc_dev_private_data *pdata);
> + void (*bind)(struct acpi_soc_dev_private_data *pdata);
> + void (*unbind)(struct acpi_soc_dev_private_data *pdata);
> +};
> +
> +/**
> + * struct acpi_soc_dev_private_data - acpi device private data
> + * @mmio_base: virtual memory base addr of the device
> + * @mmio_size: device memory size
> + * @dev_desc: device description
> + * @adev: apci device
> + *

Delete the above line.

> + */
> +struct acpi_soc_dev_private_data {
> + void __iomem *mmio_base;
> + resource_size_t mmio_size;
> +
> + const struct acpi_soc_dev_desc *dev_desc;
> + struct acpi_device *adev;
> +};
> +
> +/**
> + * register_acpi_soc - register a new acpi soc
> + * @asoc: acpi soc
> + * @disable_scan_handler: true means remove deafult scan handle
> + * false means use deafult scan handle
> + *
> + * register a new acpi soc into asoc_list and install deafult scan
> handle.
> + */
> +void register_acpi_soc(struct acpi_soc *asoc, bool
> disable_scan_handler);
> +
> +#endif
> diff --git a/drivers/acpi/internal.h b/drivers/acpi/internal.h
> index 447f6d6..c8a0e8e 100644
> --- a/drivers/acpi/internal.h
> +++ b/drivers/acpi/internal.h
> @@ -68,6 +68,12 @@ static inline void acpi_debugfs_init(void)
> { return; }
> #endif
> void acpi_lpss_init(void);
>
> +#ifdef CONFIG_X86_AMD_PLATFORM_DEVICE
> +void acpi_apd_init(void);
> +#else
> +static inline void acpi_apd_init(void) {}
> +#endif
> +
> acpi_status acpi_hotplug_schedule(struct acpi_device *adev, u32 src);
> bool acpi_queue_hotplug_work(struct work_struct *work);
> void acpi_device_hotplug(struct acpi_device *adev, u32 src);
> diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
> index 0476e90..24fef2b 100644
> --- a/drivers/acpi/scan.c
> +++ b/drivers/acpi/scan.c
> @@ -2349,6 +2349,7 @@ int __init acpi_scan_init(void)
> acpi_pci_link_init();
> acpi_processor_init();
> acpi_lpss_init();
> + acpi_apd_init();
> acpi_cmos_rtc_init();
> acpi_container_init();
> acpi_memory_hotplug_init();
> --
> 1.9.1
>
--
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/