usb: typec: Add a sysfs node to manage port type

User space applications in some cases have the need to enforce a specific port type(DFP/UFP/DRP). This change allows userspace to attempt setting the desired port type. Low level drivers can however reject the request if the specific port type is not supported.

---
Changelog since v1:
- introduced a new variable port_type in struct typec_port to track the current port type instead of changing type member in typec_capability to address Heikki Krogerus comments.
- changed the output format and strings that would be displayed as suggested by Heikki Krogerus.

Changelog since v2:
- introduced a new mutex lock to protect port_type for addressing the race conditions identified by Guenter Roeck
- added typec_port_types_drp for printing port_type when cap->type is TYPE_PORT_DRP as suggested by Guenter Roeck
- Power role swap and data role swaps would be rejected unless port port_type == TYPE_PORT_DRP
- port_type_store would return immediately if the current port_type is same as the new port_type as suggested by Guenter Roeck

Changelog since v3:
- Moved as much as code outside port_type_lock as suggested by Guenter Roeck
- Removed Change-Id line from commit message identified by Greg Kroah-Hartman

Changelog since v4:
- Corrected return value and moved moved one more line of code outside port_type_lock in the port_type_store function as suggested by Guenter Roeck

Changelog since v6:
- Added back the ABI Documentation which was dropped accidentally during patch revisions.

 Documentation/ABI/testing/sysfs-class-typec |  15 ++++
 drivers/usb/typec/typec.c                   | 106 +++++++++++++++++++++++++---
 include/linux/usb/typec.h                   |   4 ++
 3 files changed, 115 insertions(+), 10 deletions(-)

index d4a3d23eb09c..5be552e255e9 100644
@@ -30,6 +30,21 @@ Description:
 
 		Valid values: source, sink
 
+What:           /sys/class/typec/<port>/port_type
+Date:           May 2017
+Contact:	Badhri Jagan Sridharan <[email protected]>
+Description:
+		Indicates the type of the port. This attribute can be used for
+		requesting a change in the port type. Port type change is
+		supported as a synchronous operation, so write(2) to the
+		attribute will not return until the operation has finished.
+
+		Valid values:
+		- source (The port will behave as source only DFP port)
+		- sink (The port will behave as sink only UFP port)
+		- dual (The port will behave as dual-role-data and
+			dual-role-power port)
+
 What:		/sys/class/typec/<port>/vconn_source
 Date:		April 2017
 Contact:	Heikki Krogerus <[email protected]>

index ec78204964ab..b174248af7c9 100644
@@ -190,6 +190,7 @@ struct typec_partner_desc {
  * @pr_set: Set Power Role
  * @vconn_set: Set VCONN Role
  * @activate_mode: Enter/exit given Alternate Mode
+ * @port_type_set: Set port type
  *
  * Static capabilities of a single USB Type-C port.
  */

@@ -214,6 +215,9 @@ struct typec_capability {
 
 	int		(*activate_mode)(const struct typec_capability *,
 					 int mode, int activate);
+	int		(*port_type_set)(const struct typec_capability *,
+					enum typec_port_type);
+
 };
 
 /* Specific to try_role(). Indicates the user want's to clear the preference. */

2.13.0.219.gdb65acc882-goog

Hi,

Sorry for the late reply. I was unable to follow my email last week.

It looks good to me. In case Greg has not yet picked this:

Thanks,

heikki

Hi Heikki,

Greg has already picked the change.
I got emails on Jun 3rd saying it has been picked on usb-next branch. I do see it there as well.
https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb.git/commit/?h=usb-next&id=bab3548078237706f53baafe43ae58257225549d
Happy to know that you are good with the CL as well :)

Thanks,
Badhri.