On Mon, Nov 01, 2021 at 07:55:17PM +0100, Hans de Goede wrote:
The Xiaomi Mi Pad 2 tablet uses an ACPI enumerated LP8556 backlight controller for its LCD-panel, with a Xiaomi specific ACPI HID of "XMCC0001", add support for this.
Note the new "if (id)" check also fixes a NULL pointer deref when a user tries to manually bind the driver from sysfs.
When CONFIG_ACPI is disabled acpi_match_device() will always return NULL, so the lp855x_parse_acpi() call will get optimized away.
Signed-off-by: Hans de Goede hdegoede@redhat.com
drivers/video/backlight/lp855x_bl.c | 70 ++++++++++++++++++++++++----- 1 file changed, 60 insertions(+), 10 deletions(-)
diff --git a/drivers/video/backlight/lp855x_bl.c b/drivers/video/backlight/lp855x_bl.c index d1d27d5eb0f2..f075ec84acfb 100644 --- a/drivers/video/backlight/lp855x_bl.c +++ b/drivers/video/backlight/lp855x_bl.c @@ -338,10 +339,6 @@ static int lp855x_parse_dt(struct lp855x *lp) return -EINVAL; }
- pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
- if (!pdata)
return -ENOMEM;- of_property_read_string(node, "bl-name", &pdata->name); of_property_read_u8(node, "dev-ctrl", &pdata->device_control); of_property_read_u8(node, "init-brt", &pdata->initial_brightness);
Shouldn't there be a removal of an `lp->pdata = pdata` from somewhere in this function?
@@ -379,8 +376,31 @@ static int lp855x_parse_dt(struct lp855x *lp) } #endif
+static int lp855x_parse_acpi(struct lp855x *lp) +{
- int ret;
- /*
* On ACPI the device has already been initialized by the firmware
Perhaps nitpicking but ideally I'd like "and is in register mode" here since I presume it can also be assumed that everything with this HID has adopted that).
Other than these, LGTM.
Daniel.