2014-05-28 09:50:20

by Antonio Ospite

[permalink] [raw]
Subject: [PATCH BlueZ] plugins/sixaxis: Fix get_js_number() for devices connected via BT

Match hid devices and input devices using HID_UNIQ and UNIQ when these
are available, this is the correct way to get matching devices when the
controllers are connected via BT (UNIQ refers to the device bdaddr, PHYS
is the adapter bdaddr, so matching against PHYS will result in all
devices with the same LED number).

Fall back to HID_PHYS and PHYS when needed, hid devices do not define
HID_UNIQ when connected via USB.
---

Hi,

this is a possible fix for assigning the correct LED numbers via BT.

Frank, can you take a look?

I thought about using the information about the bus explicitly but it was
slightly more work; considering that I still intend to look into simplifying
get_js_number() and solving the race we have the first time the joydev module
is loaded the changes proposed in this patch seemed a viable temporary fix.

Successfully tested on kernel 3.13 and 3.15-rc4: the js number is retrieved
correctly for the second controller (even though hidraw over BT was broken in
3.13 so setting the actual LED failed there).

Thanks,
Antonio

plugins/sixaxis.c | 33 +++++++++++++++++++++++++--------
1 file changed, 25 insertions(+), 8 deletions(-)

diff --git a/plugins/sixaxis.c b/plugins/sixaxis.c
index de878d1..b802a9c 100644
--- a/plugins/sixaxis.c
+++ b/plugins/sixaxis.c
@@ -302,15 +302,24 @@ static int get_js_number(struct udev_device *udevice)
struct udev_enumerate *enumerate;
struct udev_device *hid_parent;
const char *hidraw_node;
- const char *hid_phys;
+ const char *hid_id;
int number = 0;

hid_parent = udev_device_get_parent_with_subsystem_devtype(udevice,
"hid", NULL);

- hid_phys = udev_device_get_property_value(hid_parent, "HID_PHYS");
+ /*
+ * Look for HID_UNIQ first for the correct behavior via BT, if
+ * HID_UNIQ is not available it means the USB bus is being used and we
+ * can rely on HID_PHYS.
+ */
+ hid_id = udev_device_get_property_value(hid_parent, "HID_UNIQ");
+ if (!hid_id)
+ hid_id = udev_device_get_property_value(hid_parent,
+ "HID_PHYS");
+
hidraw_node = udev_device_get_devnode(udevice);
- if (!hid_phys || !hidraw_node)
+ if (!hid_id || !hidraw_node)
return 0;

enumerate = udev_enumerate_new(udev_device_get_udev(udevice));
@@ -321,7 +330,7 @@ static int get_js_number(struct udev_device *udevice)
udev_list_entry_foreach(dev_list_entry, devices) {
struct udev_device *input_parent;
struct udev_device *js_dev;
- const char *input_phys;
+ const char *input_id;
const char *devname;

devname = udev_list_entry_get_name(dev_list_entry);
@@ -336,12 +345,20 @@ static int get_js_number(struct udev_device *udevice)

/* check if this is the joystick relative to the hidraw device
* above */
- input_phys = udev_device_get_sysattr_value(input_parent,
- "phys");
- if (!input_phys)
+ input_id = udev_device_get_sysattr_value(input_parent, "uniq");
+
+ /*
+ * A strlen() check is needed because input device over USB
+ * have the UNIQ attribute defined but with an empty value.
+ */
+ if (!input_id || strlen(input_id) == 0)
+ input_id = udev_device_get_sysattr_value(input_parent,
+ "phys");
+
+ if (!input_id)
goto next;

- if (!strcmp(input_phys, hid_phys)) {
+ if (!strcmp(input_id, hid_id)) {
number = atoi(udev_device_get_sysnum(js_dev));

/* joystick numbers start from 0, leds from 1 */
--
2.0.0.rc4



2014-05-30 16:31:14

by Frank Praznik

[permalink] [raw]
Subject: Re: [PATCH BlueZ] plugins/sixaxis: Fix get_js_number() for devices connected via BT

On 5/28/2014 05:50, Antonio Ospite wrote:
> Match hid devices and input devices using HID_UNIQ and UNIQ when these
> are available, this is the correct way to get matching devices when the
> controllers are connected via BT (UNIQ refers to the device bdaddr, PHYS
> is the adapter bdaddr, so matching against PHYS will result in all
> devices with the same LED number).
>
> Fall back to HID_PHYS and PHYS when needed, hid devices do not define
> HID_UNIQ when connected via USB.
> ---
>
> Hi,
>
> this is a possible fix for assigning the correct LED numbers via BT.
>
> Frank, can you take a look?
>
> I thought about using the information about the bus explicitly but it was
> slightly more work; considering that I still intend to look into simplifying
> get_js_number() and solving the race we have the first time the joydev module
> is loaded the changes proposed in this patch seemed a viable temporary fix.
>
> Successfully tested on kernel 3.13 and 3.15-rc4: the js number is retrieved
> correctly for the second controller (even though hidraw over BT was broken in
> 3.13 so setting the actual LED failed there).
>
> Thanks,
> Antonio
>
> plugins/sixaxis.c | 33 +++++++++++++++++++++++++--------
> 1 file changed, 25 insertions(+), 8 deletions(-)
>
> diff --git a/plugins/sixaxis.c b/plugins/sixaxis.c
> index de878d1..b802a9c 100644
> --- a/plugins/sixaxis.c
> +++ b/plugins/sixaxis.c
> @@ -302,15 +302,24 @@ static int get_js_number(struct udev_device *udevice)
> struct udev_enumerate *enumerate;
> struct udev_device *hid_parent;
> const char *hidraw_node;
> - const char *hid_phys;
> + const char *hid_id;
> int number = 0;
>
> hid_parent = udev_device_get_parent_with_subsystem_devtype(udevice,
> "hid", NULL);
>
> - hid_phys = udev_device_get_property_value(hid_parent, "HID_PHYS");
> + /*
> + * Look for HID_UNIQ first for the correct behavior via BT, if
> + * HID_UNIQ is not available it means the USB bus is being used and we
> + * can rely on HID_PHYS.
> + */
> + hid_id = udev_device_get_property_value(hid_parent, "HID_UNIQ");
> + if (!hid_id)
> + hid_id = udev_device_get_property_value(hid_parent,
> + "HID_PHYS");
> +
> hidraw_node = udev_device_get_devnode(udevice);
> - if (!hid_phys || !hidraw_node)
> + if (!hid_id || !hidraw_node)
> return 0;
>
> enumerate = udev_enumerate_new(udev_device_get_udev(udevice));
> @@ -321,7 +330,7 @@ static int get_js_number(struct udev_device *udevice)
> udev_list_entry_foreach(dev_list_entry, devices) {
> struct udev_device *input_parent;
> struct udev_device *js_dev;
> - const char *input_phys;
> + const char *input_id;
> const char *devname;
>
> devname = udev_list_entry_get_name(dev_list_entry);
> @@ -336,12 +345,20 @@ static int get_js_number(struct udev_device *udevice)
>
> /* check if this is the joystick relative to the hidraw device
> * above */
> - input_phys = udev_device_get_sysattr_value(input_parent,
> - "phys");
> - if (!input_phys)
> + input_id = udev_device_get_sysattr_value(input_parent, "uniq");
> +
> + /*
> + * A strlen() check is needed because input device over USB
> + * have the UNIQ attribute defined but with an empty value.
> + */
> + if (!input_id || strlen(input_id) == 0)
> + input_id = udev_device_get_sysattr_value(input_parent,
> + "phys");
> +
> + if (!input_id)
> goto next;
>
> - if (!strcmp(input_phys, hid_phys)) {
> + if (!strcmp(input_id, hid_id)) {
> number = atoi(udev_device_get_sysnum(js_dev));
>
> /* joystick numbers start from 0, leds from 1 */

I tested this against the latest hid-sony module code (queued for 3.16)
with a mix of USB and Bluetooth controllers and it seemed to work just fine.

Tested-by: Frank Praznik <[email protected]>


2014-06-08 12:58:39

by Szymon Janc

[permalink] [raw]
Subject: Re: [PATCH BlueZ] plugins/sixaxis: Fix get_js_number() for devices connected via BT

Hi Antonio,

On Wednesday 28 May 2014 11:50:20 Antonio Ospite wrote:
> Match hid devices and input devices using HID_UNIQ and UNIQ when these
> are available, this is the correct way to get matching devices when the
> controllers are connected via BT (UNIQ refers to the device bdaddr, PHYS
> is the adapter bdaddr, so matching against PHYS will result in all
> devices with the same LED number).
>
> Fall back to HID_PHYS and PHYS when needed, hid devices do not define
> HID_UNIQ when connected via USB.
> ---
>
> Hi,
>
> this is a possible fix for assigning the correct LED numbers via BT.
>
> Frank, can you take a look?
>
> I thought about using the information about the bus explicitly but it was
> slightly more work; considering that I still intend to look into simplifying
> get_js_number() and solving the race we have the first time the joydev
> module is loaded the changes proposed in this patch seemed a viable
> temporary fix.
>
> Successfully tested on kernel 3.13 and 3.15-rc4: the js number is retrieved
> correctly for the second controller (even though hidraw over BT was broken
> in 3.13 so setting the actual LED failed there).
>
> Thanks,
> Antonio
>
> plugins/sixaxis.c | 33 +++++++++++++++++++++++++--------
> 1 file changed, 25 insertions(+), 8 deletions(-)
>
> diff --git a/plugins/sixaxis.c b/plugins/sixaxis.c
> index de878d1..b802a9c 100644
> --- a/plugins/sixaxis.c
> +++ b/plugins/sixaxis.c
> @@ -302,15 +302,24 @@ static int get_js_number(struct udev_device *udevice)
> struct udev_enumerate *enumerate;
> struct udev_device *hid_parent;
> const char *hidraw_node;
> - const char *hid_phys;
> + const char *hid_id;
> int number = 0;
>
> hid_parent = udev_device_get_parent_with_subsystem_devtype(udevice,
> "hid", NULL);
>
> - hid_phys = udev_device_get_property_value(hid_parent, "HID_PHYS");
> + /*
> + * Look for HID_UNIQ first for the correct behavior via BT, if
> + * HID_UNIQ is not available it means the USB bus is being used and we
> + * can rely on HID_PHYS.
> + */
> + hid_id = udev_device_get_property_value(hid_parent, "HID_UNIQ");
> + if (!hid_id)
> + hid_id = udev_device_get_property_value(hid_parent,
> + "HID_PHYS");
> +
> hidraw_node = udev_device_get_devnode(udevice);
> - if (!hid_phys || !hidraw_node)
> + if (!hid_id || !hidraw_node)
> return 0;
>
> enumerate = udev_enumerate_new(udev_device_get_udev(udevice));
> @@ -321,7 +330,7 @@ static int get_js_number(struct udev_device *udevice)
> udev_list_entry_foreach(dev_list_entry, devices) {
> struct udev_device *input_parent;
> struct udev_device *js_dev;
> - const char *input_phys;
> + const char *input_id;
> const char *devname;
>
> devname = udev_list_entry_get_name(dev_list_entry);
> @@ -336,12 +345,20 @@ static int get_js_number(struct udev_device *udevice)
>
> /* check if this is the joystick relative to the hidraw device
> * above */
> - input_phys = udev_device_get_sysattr_value(input_parent,
> - "phys");
> - if (!input_phys)
> + input_id = udev_device_get_sysattr_value(input_parent, "uniq");
> +
> + /*
> + * A strlen() check is needed because input device over USB
> + * have the UNIQ attribute defined but with an empty value.
> + */
> + if (!input_id || strlen(input_id) == 0)
> + input_id = udev_device_get_sysattr_value(input_parent,
> + "phys");
> +
> + if (!input_id)
> goto next;
>
> - if (!strcmp(input_phys, hid_phys)) {
> + if (!strcmp(input_id, hid_id)) {
> number = atoi(udev_device_get_sysnum(js_dev));
>
> /* joystick numbers start from 0, leds from 1 */

Patch applied, thanks.

--
Szymon K. Janc
[email protected]