| Message ID | 20170722185537.12696-1-hdegoede@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi, On 22-07-17 20:55, Hans de Goede wrote: > Some ACPI devices report multiple ids for a single i2c_client, while not > really implementing the hw-interface asociated with some of these ids. > > For some of these devices calling probe and having probe fail with > -ENODEV is a problem in itself as this causes the device to be > powered-up and down again (causes its PS0 and PS3 ACPI methods to be > executed) which puts some devices in an unusable state. > > This commit adds a match callback to i2c_driver, allowing drivers to > override the default i2c_bus match behavior and tell the core they > are not the right driver for the device, avoiding i2c_bus_type.probe > getting called, avoiding the undesirable power up / down cycle. > > Signed-off-by: Hans de Goede <hdegoede@redhat.com> What is the status of this series ? I thought there was agreement on merging this through the i2c (wsa/linux.git) tree ? Regards, Hans > --- > drivers/i2c/i2c-core-base.c | 12 +++++++++--- > include/linux/i2c.h | 6 ++++++ > 2 files changed, 15 insertions(+), 3 deletions(-) > > diff --git a/drivers/i2c/i2c-core-base.c b/drivers/i2c/i2c-core-base.c > index c89dac7fd2e7..2c6702f2347b 100644 > --- a/drivers/i2c/i2c-core-base.c > +++ b/drivers/i2c/i2c-core-base.c > @@ -98,8 +98,16 @@ EXPORT_SYMBOL_GPL(i2c_match_id); > static int i2c_device_match(struct device *dev, struct device_driver *drv) > { > struct i2c_client *client = i2c_verify_client(dev); > - struct i2c_driver *driver; > + struct i2c_driver *driver = to_i2c_driver(drv); > + int ret; > > + if (driver->match && client) { > + ret = driver->match(client); > + if (ret < 0) > + return 0; > + if (ret > 0) > + return 1; > + } > > /* Attempt an OF style match */ > if (i2c_of_match_device(drv->of_match_table, client)) > @@ -109,8 +117,6 @@ static int i2c_device_match(struct device *dev, struct device_driver *drv) > if (acpi_driver_match_device(dev, drv)) > return 1; > > - driver = to_i2c_driver(drv); > - > /* Finally an I2C match */ > if (i2c_match_id(driver->id_table, client)) > return 1; > diff --git a/include/linux/i2c.h b/include/linux/i2c.h > index 00ca5b86a753..670577c82bc3 100644 > --- a/include/linux/i2c.h > +++ b/include/linux/i2c.h > @@ -139,6 +139,9 @@ enum i2c_alert_protocol { > * struct i2c_driver - represent an I2C device driver > * @class: What kind of i2c device we instantiate (for detect) > * @attach_adapter: Callback for bus addition (deprecated) > + * @match: Allows the driver to override the default i2c_bus match behavior > + * return < 0 to fail the match, > 0 to force a match, 0 to fallback > + * to default id matching > * @probe: Callback for device binding - soon to be deprecated > * @probe_new: New callback for device binding > * @remove: Callback for device unbinding > @@ -180,6 +183,9 @@ struct i2c_driver { > */ > int (*attach_adapter)(struct i2c_adapter *) __deprecated; > > + /* Set this to override standard i2c_bus match behavior */ > + int (*match)(struct i2c_client *); > + > /* Standard driver model interfaces */ > int (*probe)(struct i2c_client *, const struct i2c_device_id *); > int (*remove)(struct i2c_client *); > -- To unsubscribe from this list: send the line "unsubscribe linux-input" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Aug 14, 2017 at 10:13:23PM +0200, Hans de Goede wrote: > Hi, > > On 22-07-17 20:55, Hans de Goede wrote: > >Some ACPI devices report multiple ids for a single i2c_client, while not > >really implementing the hw-interface asociated with some of these ids. > > > >For some of these devices calling probe and having probe fail with > >-ENODEV is a problem in itself as this causes the device to be > >powered-up and down again (causes its PS0 and PS3 ACPI methods to be > >executed) which puts some devices in an unusable state. > > > >This commit adds a match callback to i2c_driver, allowing drivers to > >override the default i2c_bus match behavior and tell the core they > >are not the right driver for the device, avoiding i2c_bus_type.probe > >getting called, avoiding the undesirable power up / down cycle. > > > >Signed-off-by: Hans de Goede <hdegoede@redhat.com> > > What is the status of this series ? I thought there was agreement > on merging this through the i2c (wsa/linux.git) tree ? Yes, but I need to review the I2C core changes first.
diff --git a/drivers/i2c/i2c-core-base.c b/drivers/i2c/i2c-core-base.c index c89dac7fd2e7..2c6702f2347b 100644 --- a/drivers/i2c/i2c-core-base.c +++ b/drivers/i2c/i2c-core-base.c @@ -98,8 +98,16 @@ EXPORT_SYMBOL_GPL(i2c_match_id); static int i2c_device_match(struct device *dev, struct device_driver *drv) { struct i2c_client *client = i2c_verify_client(dev); - struct i2c_driver *driver; + struct i2c_driver *driver = to_i2c_driver(drv); + int ret; + if (driver->match && client) { + ret = driver->match(client); + if (ret < 0) + return 0; + if (ret > 0) + return 1; + } /* Attempt an OF style match */ if (i2c_of_match_device(drv->of_match_table, client)) @@ -109,8 +117,6 @@ static int i2c_device_match(struct device *dev, struct device_driver *drv) if (acpi_driver_match_device(dev, drv)) return 1; - driver = to_i2c_driver(drv); - /* Finally an I2C match */ if (i2c_match_id(driver->id_table, client)) return 1; diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 00ca5b86a753..670577c82bc3 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -139,6 +139,9 @@ enum i2c_alert_protocol { * struct i2c_driver - represent an I2C device driver * @class: What kind of i2c device we instantiate (for detect) * @attach_adapter: Callback for bus addition (deprecated) + * @match: Allows the driver to override the default i2c_bus match behavior + * return < 0 to fail the match, > 0 to force a match, 0 to fallback + * to default id matching * @probe: Callback for device binding - soon to be deprecated * @probe_new: New callback for device binding * @remove: Callback for device unbinding @@ -180,6 +183,9 @@ struct i2c_driver { */ int (*attach_adapter)(struct i2c_adapter *) __deprecated; + /* Set this to override standard i2c_bus match behavior */ + int (*match)(struct i2c_client *); + /* Standard driver model interfaces */ int (*probe)(struct i2c_client *, const struct i2c_device_id *); int (*remove)(struct i2c_client *);
Some ACPI devices report multiple ids for a single i2c_client, while not really implementing the hw-interface asociated with some of these ids. For some of these devices calling probe and having probe fail with -ENODEV is a problem in itself as this causes the device to be powered-up and down again (causes its PS0 and PS3 ACPI methods to be executed) which puts some devices in an unusable state. This commit adds a match callback to i2c_driver, allowing drivers to override the default i2c_bus match behavior and tell the core they are not the right driver for the device, avoiding i2c_bus_type.probe getting called, avoiding the undesirable power up / down cycle. Signed-off-by: Hans de Goede <hdegoede@redhat.com> --- drivers/i2c/i2c-core-base.c | 12 +++++++++--- include/linux/i2c.h | 6 ++++++ 2 files changed, 15 insertions(+), 3 deletions(-)