RE: [PATCH v2 1/1] mfd: intel_quark_i2c_gpio: Add Intel Quark X1000 I2C-GPIO MFD Driver

From: Tan, Raymond
Date: Mon Nov 24 2014 - 08:52:10 EST


Hi Lee Jones,

Thanks for the review and input. Here's the responses.

Warm Regards,

ÂRaymond Tan

> -----Original Message-----
> From: Lee Jones [mailto:lee.jones@xxxxxxxxxx]
> Sent: Wednesday, November 19, 2014 1:22 AM
> To: Tan, Raymond
> Cc: Samuel Ortiz; linux-kernel@xxxxxxxxxxxxxxx; Chen, Alvin; Shevchenko,
> Andriy
> Subject: Re: [PATCH v2 1/1] mfd: intel_quark_i2c_gpio: Add Intel Quark
> X1000 I2C-GPIO MFD Driver
>
> On Tue, 11 Nov 2014, Raymond Tan wrote:
>
> > In Quark X1000, there's a single PCI device that provides both an I2C
> > controller and a GPIO controller. This MFD driver will split the 2
> > devices for their respective drivers.
> >
> > This patch is based on Josef Ahmad's initial work for Quark enabling.
> >
> > Reviewed-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
> > Signed-off-by: Weike Chen <alvin.chen@xxxxxxxxx>
> > Signed-off-by: Raymond Tan <raymond.tan@xxxxxxxxx>
> > ---
> > drivers/mfd/Kconfig | 11 ++
> > drivers/mfd/Makefile | 1 +
> > drivers/mfd/intel_quark_i2c_gpio.c | 298
> > ++++++++++++++++++++++++++++++++++++
> > 3 files changed, 310 insertions(+)
> > create mode 100644 drivers/mfd/intel_quark_i2c_gpio.c
> >
> > diff --git a/drivers/mfd/Kconfig b/drivers/mfd/Kconfig index
> > b7c74a7..d01d042 100644
> > --- a/drivers/mfd/Kconfig
> > +++ b/drivers/mfd/Kconfig
> > @@ -219,6 +219,17 @@ config HTC_I2CPLD
> > This device provides input and output GPIOs through an I2C
> > interface to one or more sub-chips.
> >
> > +config MFD_INTEL_QUARK_I2C_GPIO
> > + tristate "Intel Quark MFD I2C GPIO"
> > + depends on PCI && X86
>
> Why have you bunched these two up?

Noted. I will separate them.
>
> > + depends on COMMON_CLK
>
> I don't think you should depend on this. Just use the API and fail if it doesn't
> find the clock you're after.

This was added to make sure the clk initialization and other clk related calls will work with the MFD.
>
> > + select MFD_CORE
> > + help
> > + This MFD provides support for I2C and GPIO that exist only
> > + in a single PCI device. It splits the 2 IO devices to
> > + their respective IO driver.
> > + The GPIO exports a total amount of 8 interrupt-capable GPIOs.
> > +
> > config LPC_ICH
> > tristate "Intel ICH LPC"
> > depends on PCI
> > diff --git a/drivers/mfd/Makefile b/drivers/mfd/Makefile index
> > 8a28dc9..d42652d 100644
> > --- a/drivers/mfd/Makefile
> > +++ b/drivers/mfd/Makefile
> > @@ -133,6 +133,7 @@ obj-$(CONFIG_AB8500_CORE) += ab8500-
> core.o ab8500-sysctrl.o
> > obj-$(CONFIG_MFD_TIMBERDALE) += timberdale.o
> > obj-$(CONFIG_PMIC_ADP5520) += adp5520.o
> > obj-$(CONFIG_MFD_KEMPLD) += kempld-core.o
> > +obj-$(CONFIG_MFD_INTEL_QUARK_I2C_GPIO) +=
> intel_quark_i2c_gpio.o
> > obj-$(CONFIG_LPC_SCH) += lpc_sch.o
> > obj-$(CONFIG_LPC_ICH) += lpc_ich.o
> > obj-$(CONFIG_MFD_RDC321X) += rdc321x-southbridge.o
> > diff --git a/drivers/mfd/intel_quark_i2c_gpio.c
> > b/drivers/mfd/intel_quark_i2c_gpio.c
> > new file mode 100644
> > index 0000000..eed95da
> > --- /dev/null
> > +++ b/drivers/mfd/intel_quark_i2c_gpio.c
> > @@ -0,0 +1,298 @@
> > +/*
> > + * Intel Quark MFD PCI driver for I2C & GPIO
> > + *
> > + * Copyright(c) 2014 Intel Corporation.
> > + *
> > + * This program is free software; you can redistribute it and/or
> > +modify it
> > + * under the terms and conditions of the GNU General Public License,
> > + * version 2, as published by the Free Software Foundation.
> > + *
> > + * This program is distributed in the hope 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.
> > + *
> > + * Intel Quark PCI device for I2C and GPIO controller sharing the
> > +same
> > + * PCI function. This PCI driver will split the 2 devices into their
> > + * respective drivers.
> > + */
> > +
> > +#include <linux/kernel.h>
> > +#include <linux/module.h>
> > +#include <linux/pci.h>
> > +#include <linux/mfd/core.h>
> > +#include <linux/clkdev.h>
> > +#include <linux/clk-provider.h>
> > +#include <linux/dmi.h>
> > +#include <linux/platform_data/gpio-dwapb.h>
> > +#include <linux/platform_data/i2c-designware.h>
> > +
> > +/* PCI BAR for register base address */
> > +#define MFD_I2C_BAR 0
> > +#define MFD_GPIO_BAR 1
> > +
> > +/* The base GPIO number under GPIOLIB framework */
> > +#define INTEL_QUARK_MFD_GPIO_BASE 8
> > +
> > +/* The default number of South-Cluster GPIO on Quark. */
> > +#define INTEL_QUARK_MFD_NGPIO 8
> > +
> > +/* The DesignWare GPIO ports on Quark. */
> > +#define INTEL_QUARK_GPIO_NPORTS 1
> > +
> > +#define INTEL_QUARK_IORES_MEM 0
> > +#define INTEL_QUARK_IORES_IRQ 1
> > +
> > +#define INTEL_QUARK_I2C_CONTROLLER_CLK "i2c_designware.0"
> > +
> > +/* The Quark I2C controller source clock */
> > +#define INTEL_QUARK_I2C_CLK_HZ 33000000
> > +
> > +#define INTEL_QUARK_I2C_NCLK 1
> > +
> > +struct clk *intel_quark_i2c_clk;
> > +struct clk_lookup *intel_quark_i2c_clk_lookups;
>
> Why do these need to be global?

I was trying to keep these as runtime allocated variables, and cleaned up with removal of the driver / upon error handling.
>
> > +struct i2c_mode_info {
> > + const char *name;
> > + unsigned int i2c_scl_freq;
> > +};
> > +
> > +static const struct i2c_mode_info platform_i2c_mode_info[] = {
> > + {
> > + .name = "Galileo",
> > + .i2c_scl_freq = 100000,
> > + },
> > + {
> > + .name = "GalileoGen2",
> > + .i2c_scl_freq = 400000,
> > + },
> > +};
> > +
> > +static struct resource intel_quark_i2c_res[] = {
> > + [INTEL_QUARK_IORES_MEM] = {
> > + .flags = IORESOURCE_MEM,
> > + },
> > + [INTEL_QUARK_IORES_IRQ] = {
> > + .flags = IORESOURCE_IRQ,
> > + },
> > +};
> > +
> > +static struct resource intel_quark_gpio_res[] = {
> > + [INTEL_QUARK_IORES_MEM] = {
> > + .flags = IORESOURCE_MEM,
> > + },
> > +};
> > +
> > +static struct mfd_cell intel_quark_mfd_cells[] = {
> > + {
> > + .id = MFD_I2C_BAR,
> > + .name = "i2c_designware",
> > + .num_resources = ARRAY_SIZE(intel_quark_i2c_res),
> > + .resources = intel_quark_i2c_res,
> > + .ignore_resource_conflicts = true,
> > + },
> > + {
> > + .id = MFD_GPIO_BAR,
> > + .name = "gpio-dwapb",
> > + .num_resources = ARRAY_SIZE(intel_quark_gpio_res),
> > + .resources = intel_quark_gpio_res,
> > + .ignore_resource_conflicts = true,
> > + },
> > +};
> > +
> > +static const struct pci_device_id intel_quark_mfd_ids[] = {
> > + { PCI_VDEVICE(INTEL, 0x0934), },
> > + { 0,}
> > +};
> > +MODULE_DEVICE_TABLE(pci, intel_quark_mfd_ids);
> > +
> > +static struct dwapb_platform_data *
> > +intel_quark_prepare_pdata(struct pci_dev *pdev) {
> > + struct dwapb_platform_data *pdata;
> > + struct device *dev = &pdev->dev;
> > +
> > + pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
> > + if (!pdata)
> > + return ERR_PTR(-ENOMEM);
> > +
> > + /* For intel quark x1000, it has only one port: portA */
> > + pdata->nports = INTEL_QUARK_GPIO_NPORTS;
> > + pdata->properties = devm_kcalloc(dev, pdata->nports,
> > + sizeof(*pdata->properties),
> > + GFP_KERNEL);
> > + if (!pdata->properties)
> > + return ERR_PTR(-ENOMEM);
> > +
> > + /* Set the properties for portA */
> > + pdata->properties->node = NULL;
> > + pdata->properties->name = "intel-quark-x1000-gpio-portA";
> > + pdata->properties->idx = 0;
> > + pdata->properties->ngpio = INTEL_QUARK_MFD_NGPIO;
> > + pdata->properties->gpio_base = INTEL_QUARK_MFD_GPIO_BASE;
> > + pdata->properties->irq = pdev->irq;
> > + pdata->properties->irq_shared = true;
> > +
> > + return pdata;
> > +}
> > +
> > +static struct dw_i2c_platform_data *
> > +intel_quark_get_i2c_mode(struct pci_dev *pdev) {
> > + const char *board_name =
> dmi_get_system_info(DMI_BOARD_NAME);
> > + struct dw_i2c_platform_data *pdata;
> > + struct device *dev = &pdev->dev;
> > + unsigned int i;
> > +
> > + pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
> > + if (!pdata)
> > + return ERR_PTR(-ENOMEM);
> > +
> > + /* Fast mode by default */
> > + pdata->i2c_scl_freq = 400000;
> > +
> > + for (i = 0; i < ARRAY_SIZE(platform_i2c_mode_info); i++)
> > + if (!strcmp(board_name, platform_i2c_mode_info[i].name))
> > + pdata->i2c_scl_freq
> > + = platform_i2c_mode_info[i].i2c_scl_freq;
> > +
> > + return pdata;
> > +}
>
> If this function going to be invoked from anywhere else? If not, I would
> suggest putting the content directly into
> intel_quark_i2c_setup() to simplify things a little.

Noted. Will keep this simple. I was trying to keep the call structure of the overall and various functions reusable in future.
>
> > +static int intel_quark_register_i2c_clk(struct pci_dev *pdev) {
> > + intel_quark_i2c_clk_lookups = devm_kcalloc(
>
> I suggest putting this into a device container (struct) and simplifying the
> name, allot.

Thanks for the tips. Will do this.
>
> > + &pdev->dev, INTEL_QUARK_I2C_NCLK,
> > + sizeof(*intel_quark_i2c_clk_lookups), GFP_KERNEL);
> > +
> > + if (!intel_quark_i2c_clk_lookups)
> > + return -ENOMEM;
> > +
> > + intel_quark_i2c_clk_lookups[0].dev_id =
> > +INTEL_QUARK_I2C_CONTROLLER_CLK;
> > +
> > + intel_quark_i2c_clk = clk_register_fixed_rate(
> > + &pdev->dev, INTEL_QUARK_I2C_CONTROLLER_CLK, NULL,
> > + CLK_IS_ROOT, INTEL_QUARK_I2C_CLK_HZ);
> > +
> > + return clk_register_clkdevs(intel_quark_i2c_clk,
> > + intel_quark_i2c_clk_lookups, INTEL_QUARK_I2C_NCLK);
>
> We don't normally register clks from MFD. Normally they are registered in
> drivers/clk and fetched when a device requires them.
> Why is this different?

This is a static clk and on this platform, there's no clk hardware to have the clk driver to initialize this and add into the global struct for use
later by i2c controller. The hardware on the platform itself is same, and the current driver (i2c_designware_platform) requires the clk
for its operation. Due to this, I decided to have the clk initialize in the MFD before the device being added, which later will trigger the probe on the i2c controller driver.
>
> > +}
> > +
> > +static void intel_quark_unregister_i2c_clk(void)
> > +{
> > + if (!intel_quark_i2c_clk || !intel_quark_i2c_clk_lookups)
> > + return;
> > +
> > + clkdev_drop(intel_quark_i2c_clk_lookups);
> > + clk_unregister(intel_quark_i2c_clk);
> > +}
> > +
> > +static int intel_quark_i2c_setup(struct pci_dev *pdev, struct
> > +mfd_cell *cell) {
> > + struct dw_i2c_platform_data *pdata;
> > + struct resource *res = (struct resource *)cell->resources;
> > + int retval;
> > +
> > + res[INTEL_QUARK_IORES_MEM].start =
> > + pci_resource_start(pdev, MFD_I2C_BAR);
> > + res[INTEL_QUARK_IORES_MEM].end =
> > + pci_resource_end(pdev, MFD_I2C_BAR);
> > +
> > + res[INTEL_QUARK_IORES_IRQ].start = pdev->irq;
> > + res[INTEL_QUARK_IORES_IRQ].end = pdev->irq;
> > +
> > + retval = intel_quark_register_i2c_clk(pdev);
> > + if (retval) {
> > + dev_err(&pdev->dev, "Fixed clk register failed: %d\n",
> retval);
> > + return retval;
> > + }
> > +
> > + pdata = intel_quark_get_i2c_mode(pdev);
> > + if (IS_ERR(pdata))
> > + return PTR_ERR(pdata);
>
> Bring this functionality into here?

Were you referring the contents of intel_quark_get_i2c_mode(), and avoid the additional function call instead?
>
> > + cell->platform_data = pdata;
> > + cell->pdata_size = sizeof(*pdata);
> > +
> > + return 0;
> > +}
> > +
> > +static int intel_quark_gpio_setup(struct pci_dev *pdev, struct
> > +mfd_cell *cell) {
> > + struct dwapb_platform_data *pdata;
> > + struct resource *res = (struct resource *)cell->resources;
> > +
> > + res[INTEL_QUARK_IORES_MEM].start =
> > + pci_resource_start(pdev, MFD_GPIO_BAR);
> > + res[INTEL_QUARK_IORES_MEM].end =
> > + pci_resource_end(pdev, MFD_GPIO_BAR);
> > +
> > + pdata = intel_quark_prepare_pdata(pdev);
> > + if (IS_ERR(pdata))
> > + return PTR_ERR(pdata);
> > +
> > + cell->platform_data = pdata;
> > + cell->pdata_size = sizeof(*pdata);
> > +
> > + return 0;
> > +}
>
> The different ways in which you deal with platform_data in _gpio_setup and
> _i2c_setup is confusing. _prepare_pdata insinuates that it's a generic
> method for creating platform data, but it's actually only used for the GPIO
> device. I'm sure you can find a way to standardise these call paths.
>
> Can you just allocate and populate pdata in the setup functions.
> Making all sorts of unnecessary (i.e. only called once) extra calls into oddly
> named functions is not helping the readability of this driver.

Noted. I was trying to keep the overall call structure reusable. Will change this to avoid further confusion as noted.
>
> > +struct intel_quark_mfd_dev {
> > + int (*setup)(struct pci_dev *pdev, struct mfd_cell *cell);
> > + struct mfd_cell *cell;
> > +};
> > +
> > +static struct intel_quark_mfd_dev intel_quark_mfd_devs[] = {
> > + {
> > + .cell = &intel_quark_mfd_cells[MFD_I2C_BAR],
> > + .setup = intel_quark_i2c_setup,
> > + },
> > + {
> > + .cell = &intel_quark_mfd_cells[MFD_GPIO_BAR],
> > + .setup = intel_quark_gpio_setup,
> > + },
> > + {
> > + /* terminator */
> > + },
>
> {} will do just fine.

Noted.
>
> > +};
> > +
> > +static int intel_quark_mfd_probe(struct pci_dev *pdev,
> > + const struct pci_device_id *id)
> > +{
> > + struct intel_quark_mfd_dev *mfd_dev;
> > + int retval;
> > +
> > + retval = pcim_enable_device(pdev);
> > + if (retval)
> > + return retval;
> > +
> > + for (mfd_dev = intel_quark_mfd_devs; mfd_dev->cell; mfd_dev++)
> > + if (mfd_dev->setup) {
> > + retval = mfd_dev->setup(pdev, mfd_dev->cell);
> > + if (retval)
> > + return retval;
> > + }
>
> I don't see the benifit in doing this at all. Just:
>
> intel_quark_i2c_setup();
> intel_quark_gpio_setup();

Noted. Will modify this and keep this direct.
>
>
> > + return mfd_add_devices(&pdev->dev, 0, intel_quark_mfd_cells,
> > + ARRAY_SIZE(intel_quark_mfd_cells), NULL, 0, NULL); }
> > +
> > +static void intel_quark_mfd_remove(struct pci_dev *pdev) {
> > + intel_quark_unregister_i2c_clk();
> > + mfd_remove_devices(&pdev->dev);
> > +}
> > +
> > +static struct pci_driver intel_quark_mfd_driver = {
> > + .name = "intel_quark_mfd_i2c_gpio",
> > + .id_table = intel_quark_mfd_ids,
> > + .probe = intel_quark_mfd_probe,
> > + .remove = intel_quark_mfd_remove,
> > +};
> > +
> > +module_pci_driver(intel_quark_mfd_driver);
> > +
> > +MODULE_AUTHOR("Raymond Tan <raymond.tan@xxxxxxxxx>");
> > +MODULE_DESCRIPTION("Intel Quark MFD PCI driver for I2C & GPIO");
> > +MODULE_LICENSE("GPL v2");
>
> --
> Lee Jones
> Linaro STMicroelectronics Landing Team Lead Linaro.org â Open source
> software for ARM SoCs Follow Linaro: Facebook | Twitter | Blog
N‹§²æ¸›yú²X¬¶ÇvØ–)Þ{.nlj·¥Š{±‘êX§¶›¡Ü}©ž²ÆzÚj:+v‰¨¾«‘êZ+€Êzf£¢·hšˆ§~†­†Ûÿû®w¥¢¸?™¨è&¢)ßf”ùy§m…á«a¶Úÿ 0¶ìå