platform/x86: intel_cht_int33fe: Update fusb302 type string, add properties
authorHans de Goede <hdegoede@redhat.com>
Wed, 11 Oct 2017 09:41:21 +0000 (11:41 +0200)
committerAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Fri, 3 Nov 2017 11:52:28 +0000 (13:52 +0200)
The fusb302 driver as merged in staging uses "typec_fusb302" as i2c-id
rather then just "fusb302" and needs us to set a number of device-
properties, adjust the intel_cht_int33fe driver accordingly.

One of the properties set is max-snk-mv which makes the fusb302 driver
negotiate up to 12V charging voltage, which is a bad idea on boards
which are not setup to handle this, so this commit also adds 2 extra
sanity checks to make sure that the expected Whiskey Cove PMIC +
TI bq24292i charger combo, which can handle 12V, is present.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
drivers/platform/x86/Kconfig
drivers/platform/x86/intel_cht_int33fe.c

index 0fdf688..87c8c36 100644 (file)
@@ -806,7 +806,7 @@ config ACPI_CMPC
 
 config INTEL_CHT_INT33FE
        tristate "Intel Cherry Trail ACPI INT33FE Driver"
-       depends on X86 && ACPI && I2C
+       depends on X86 && ACPI && I2C && REGULATOR
        ---help---
          This driver add support for the INT33FE ACPI device found on
          some Intel Cherry Trail devices.
@@ -817,6 +817,10 @@ config INTEL_CHT_INT33FE
          This driver instantiates i2c-clients for these, so that standard
          i2c drivers for these chips can bind to the them.
 
+         If you enable this driver it is advised to also select
+         CONFIG_TYPEC_FUSB302=m, CONFIG_CHARGER_BQ24190=m and
+         CONFIG_BATTERY_MAX17042=m.
+
 config INTEL_INT0002_VGPIO
        tristate "Intel ACPI INT0002 Virtual GPIO driver"
        depends on GPIOLIB && ACPI
index 286c620..380ef7e 100644 (file)
@@ -24,6 +24,7 @@
 #include <linux/i2c.h>
 #include <linux/interrupt.h>
 #include <linux/module.h>
+#include <linux/regulator/consumer.h>
 #include <linux/slab.h>
 
 #define EXPECTED_PTYPE         4
@@ -77,12 +78,21 @@ static const struct property_entry max17047_props[] = {
        { }
 };
 
+static const struct property_entry fusb302_props[] = {
+       PROPERTY_ENTRY_STRING("fcs,extcon-name", "cht_wcove_pwrsrc"),
+       PROPERTY_ENTRY_U32("fcs,max-sink-microvolt", 12000000),
+       PROPERTY_ENTRY_U32("fcs,max-sink-microamp",   3000000),
+       PROPERTY_ENTRY_U32("fcs,max-sink-microwatt", 36000000),
+       { }
+};
+
 static int cht_int33fe_probe(struct i2c_client *client)
 {
        struct device *dev = &client->dev;
        struct i2c_board_info board_info;
        struct cht_int33fe_data *data;
        struct i2c_client *max17047;
+       struct regulator *regulator;
        unsigned long long ptyp;
        acpi_status status;
        int fusb302_irq;
@@ -101,6 +111,34 @@ static int cht_int33fe_probe(struct i2c_client *client)
        if (ptyp != EXPECTED_PTYPE)
                return -ENODEV;
 
+       /* Check presence of INT34D3 (hardware-rev 3) expected for ptype == 4 */
+       if (!acpi_dev_present("INT34D3", "1", 3)) {
+               dev_err(dev, "Error PTYPE == %d, but no INT34D3 device\n",
+                       EXPECTED_PTYPE);
+               return -ENODEV;
+       }
+
+       /*
+        * We expect the WC PMIC to be paired with a TI bq24292i charger-IC.
+        * We check for the bq24292i vbus regulator here, this has 2 purposes:
+        * 1) The bq24292i allows charging with up to 12V, setting the fusb302's
+        *    max-snk voltage to 12V with another charger-IC is not good.
+        * 2) For the fusb302 driver to get the bq24292i vbus regulator, the
+        *    regulator-map, which is part of the bq24292i regulator_init_data,
+        *    must be registered before the fusb302 is instantiated, otherwise
+        *    it will end up with a dummy-regulator.
+        * Note "cht_wc_usb_typec_vbus" comes from the regulator_init_data
+        * which is defined in i2c-cht-wc.c from where the bq24292i i2c-client
+        * gets instantiated. We use regulator_get_optional here so that we
+        * don't end up getting a dummy-regulator ourselves.
+        */
+       regulator = regulator_get_optional(dev, "cht_wc_usb_typec_vbus");
+       if (IS_ERR(regulator)) {
+               ret = PTR_ERR(regulator);
+               return (ret == -ENODEV) ? -EPROBE_DEFER : ret;
+       }
+       regulator_put(regulator);
+
        /* The FUSB302 uses the irq at index 1 and is the only irq user */
        fusb302_irq = acpi_dev_gpio_irq_get(ACPI_COMPANION(dev), 1);
        if (fusb302_irq < 0) {
@@ -127,6 +165,7 @@ static int cht_int33fe_probe(struct i2c_client *client)
        } else {
                memset(&board_info, 0, sizeof(board_info));
                strlcpy(board_info.type, "max17047", I2C_NAME_SIZE);
+               board_info.dev_name = "max17047";
                board_info.properties = max17047_props;
                data->max17047 = i2c_acpi_new_device(dev, 1, &board_info);
                if (!data->max17047)
@@ -134,7 +173,9 @@ static int cht_int33fe_probe(struct i2c_client *client)
        }
 
        memset(&board_info, 0, sizeof(board_info));
-       strlcpy(board_info.type, "fusb302", I2C_NAME_SIZE);
+       strlcpy(board_info.type, "typec_fusb302", I2C_NAME_SIZE);
+       board_info.dev_name = "fusb302";
+       board_info.properties = fusb302_props;
        board_info.irq = fusb302_irq;
 
        data->fusb302 = i2c_acpi_new_device(dev, 2, &board_info);
@@ -142,6 +183,7 @@ static int cht_int33fe_probe(struct i2c_client *client)
                goto out_unregister_max17047;
 
        memset(&board_info, 0, sizeof(board_info));
+       board_info.dev_name = "pi3usb30532";
        strlcpy(board_info.type, "pi3usb30532", I2C_NAME_SIZE);
 
        data->pi3usb30532 = i2c_acpi_new_device(dev, 3, &board_info);