Add quirks required to make a Shuttle PN-31 remote control work
PATCH 1/1 Add quirks required to make a Shuttle PN-31 remote control work
Add some quirks to make a Shuttle PN-31 remote control work.
Shuttle PN31 remote controls are supposed to act as a USB keyboard and mouse, but they (in particular the mouse part) don't follow the standards.
5 quirks are needed to get this device to work, 2 of which are unique to this device so far (it sets a superfluous extra bit on the GenericDesktop keycodes, and the mouse flags are wrong (_Constant_, Variable, Relative).
Tested with 2.6.16-rc6-mm1 and 2.6.16-rc6-mm2.
---
--- linux-2.6.15/drivers/usb/input/hid.h.pn31~ 2006-01-22 18:12:59.000000000
+0100
+++ linux-2.6.15/drivers/usb/input/hid.h 2006-01-22 19:47:13.000000000 +0100
@@ -249,6 +249,9 @@
#define HID_QUIRK_CYMOTION 0x00000800
#define HID_QUIRK_POWERBOOK_HAS_FN 0x00001000
#define HID_QUIRK_POWERBOOK_FN_ON 0x00002000
+#define HID_QUIRK_BAD_LOGICAL_RANGE 0x00004000
+#define HID_QUIRK_WRONG_CONSTANT 0x00008000
+#define HID_QUIRK_GENDESK_EXTRA_BIT 0x00010000
/*
* This is the global environment of the parser. This information is
--- linux-2.6.15/drivers/usb/input/hid-input.c.pn31~ 2006-01-22
18:12:59.000000000 +0100
+++ linux-2.6.15/drivers/usb/input/hid-input.c 2006-01-22 18:13:35.000000000
+0100
@@ -243,7 +243,7 @@
printk(" ---> ");
#endif
- if (field->flags & HID_MAIN_ITEM_CONSTANT)
+ if ((field->flags & HID_MAIN_ITEM_CONSTANT) && !(device->quirks &HID_QUIRK_WRONG_CONSTANT))
goto ignore;
switch (usage->hid & HID_USAGE_PAGE) {
@@ -326,16 +326,21 @@
break;
}
- switch (usage->hid) {
+ if (device->quirks & HID_QUIRK_GENDESK_EXTRA_BIT)
+ code = usage->hid & 0xfff9ffff;
+ else
+ code = usage->hid;
+
+ switch (code) {
/* These usage IDs map directly to the usage codes. */ case HID_GD_X: case HID_GD_Y: case HID_GD_Z:
case HID_GD_RX: case HID_GD_RY: case HID_GD_RZ:
case HID_GD_SLIDER: case HID_GD_DIAL: case HID_GD_WHEEL:
if (field->flags & HID_MAIN_ITEM_RELATIVE)
- map_rel(usage->hid & 0xf);
+ map_rel(code & 0xf);
else
- map_abs(usage->hid & 0xf);
+ map_abs(code & 0xf);
break;
case HID_GD_HATSWITCH:
@@ -569,6 +574,12 @@
set_bit(usage->type, input->evbit);
+ if ((usage->type == EV_REL) &&
+ (device->quirks & HID_QUIRK_BAD_LOGICAL_RANGE) && (field->logical_minimum
== field->logical_maximum)) {
+ field->logical_minimum = -127;
+ field->logical_maximum = 127;
+ }
+
while (usage->code <= max && test_and_set_bit(usage->code, bit))
usage->code = find_next_zero_bit(bit, max + 1, usage->code);
--- linux-2.6.15/drivers/usb/input/hid-core.c.pn31~ 2006-01-22
18:13:18.000000000 +0100
+++ linux-2.6.15/drivers/usb/input/hid-core.c 2006-01-22 18:13:35.000000000
+0100
@@ -1449,6 +1449,9 @@
* Alphabetically sorted blacklist by quirk type. */
+#define USB_VENDOR_ID_SHUTTLE_REMOTE 0x4572
+#define USB_DEVICE_ID_SHUTTLE_PN31 0x4572
+
static const struct hid_blacklist {
__u16 idVendor;
__u16 idProduct;
@@ -1585,6 +1588,8 @@
{ USB_VENDOR_ID_APPLE, 0x030A, HID_QUIRK_POWERBOOK_HAS_FN }, { USB_VENDOR_ID_APPLE, 0x030B, HID_QUIRK_POWERBOOK_HAS_FN },
+ { USB_VENDOR_ID_SHUTTLE_REMOTE, USB_DEVICE_ID_SHUTTLE_PN31,HID_QUIRK_BAD_LOGICAL_RANGE | HID_QUIRK_HIDDEV | HID_QUIRK_MULTI_INPUT | HID_QUIRK_WRONG_CONSTANT | HID_QUIRK_GENDESK_EXTRA_BIT },
+
{ 0, 0 }
};
Quoted 10 lines · Bernhard Rosenkraenzer
Add some quirks to make a Shuttle PN-31 remote control work.
Shuttle PN31 remote controls are supposed to act as a USB keyboard and mouse, but they (in particular the mouse part) don't follow the standards.
5 quirks are needed to get this device to work, 2 of which are unique to this device so far (it sets a superfluous extra bit on the GenericDesktop keycodes, and the mouse flags are wrong (_Constant_, Variable, Relative).
Tested with 2.6.16-rc6-mm1 and 2.6.16-rc6-mm2.
Can you send me the #define DEBUG output from hid-core and hid-input? I'm quite curious how much broken the device is and if it weren't easier to implement a way how to replace the HID Report descriptor inside hid-core instead of adding more quirks.
Quoted 98 lines
Signed-off-by Bernhard Rosenkraenzer <[email protected]>---
--- linux-2.6.15/drivers/usb/input/hid.h.pn31~ 2006-01-22 18:12:59.000000000
+0100
+++linux-2.6.15/drivers/usb/input/hid.h2006-01-22 19:47:13.000000000 +0100(quoted diff) +3
-1@@ -249,6 +249,9 @@ #define HID_QUIRK_CYMOTION 0x00000800 #define HID_QUIRK_POWERBOOK_HAS_FN 0x00001000 #define HID_QUIRK_POWERBOOK_FN_ON 0x00002000 +#define HID_QUIRK_BAD_LOGICAL_RANGE 0x00004000 +#define HID_QUIRK_WRONG_CONSTANT 0x00008000 +#define HID_QUIRK_GENDESK_EXTRA_BIT 0x00010000 /* * This is the global environment of the parser. This information is --- linux-2.6.15/drivers/usb/input/hid-input.c.pn31~ 2006-01-2218:12:59.000000000 +0100
+++linux-2.6.15/drivers/usb/input/hid-input.c2006-01-22 18:13:35.000000000
+0100(quoted diff) +1
-1@@ -243,7 +243,7 @@ printk(" ---> "); #endif - if (field->flags & HID_MAIN_ITEM_CONSTANT) + if ((field->flags & HID_MAIN_ITEM_CONSTANT) && !(device->quirks & HID_QUIRK_WRONG_CONSTANT)) goto ignore;(quoted diff) +14
-4switch (usage->hid & HID_USAGE_PAGE) {@@ -326,16 +326,21 @@ break; } - switch (usage->hid) { + if (device->quirks & HID_QUIRK_GENDESK_EXTRA_BIT) + code = usage->hid & 0xfff9ffff; + else + code = usage->hid; + + switch (code) { /* These usage IDs map directly to the usage codes. */ case HID_GD_X: case HID_GD_Y: case HID_GD_Z: case HID_GD_RX: case HID_GD_RY: case HID_GD_RZ: case HID_GD_SLIDER: case HID_GD_DIAL: case HID_GD_WHEEL: if (field->flags & HID_MAIN_ITEM_RELATIVE) - map_rel(usage->hid & 0xf); + map_rel(code & 0xf); else - map_abs(usage->hid & 0xf); + map_abs(code & 0xf); break; case HID_GD_HATSWITCH:@@ -569,6 +574,12 @@ set_bit(usage->type, input->evbit); + if ((usage->type == EV_REL) && + (device->quirks & HID_QUIRK_BAD_LOGICAL_RANGE) && (field->logical_minimum == field->logical_maximum)) { + field->logical_minimum = -127; + field->logical_maximum = 127; + } + while (usage->code <= max && test_and_set_bit(usage->code, bit)) usage->code = find_next_zero_bit(bit, max + 1, usage->code); --- linux-2.6.15/drivers/usb/input/hid-core.c.pn31~ 2006-01-2218:13:18.000000000 +0100
+++linux-2.6.15/drivers/usb/input/hid-core.c2006-01-22 18:13:35.000000000
+0100(quoted diff) +4
-0@@ -1449,6 +1449,9 @@ * Alphabetically sorted blacklist by quirk type. */ +#define USB_VENDOR_ID_SHUTTLE_REMOTE 0x4572 +#define USB_DEVICE_ID_SHUTTLE_PN31 0x4572 + static const struct hid_blacklist { __u16 idVendor; __u16 idProduct;@@ -1585,6 +1588,8 @@ { USB_VENDOR_ID_APPLE, 0x030A, HID_QUIRK_POWERBOOK_HAS_FN }, { USB_VENDOR_ID_APPLE, 0x030B, HID_QUIRK_POWERBOOK_HAS_FN }, + { USB_VENDOR_ID_SHUTTLE_REMOTE, USB_DEVICE_ID_SHUTTLE_PN31,
HID_QUIRK_BAD_LOGICAL_RANGE|HID_QUIRK_HIDDEV|HID_QUIRK_MULTI_INPUT|HID_QUIRK_WRONG_CONSTANT|HID_QUIRK_GENDESK_EXTRA_BIT},(quoted diff) +1
-0+ { 0, 0 } };
Signature · 2 lines
Vojtech Pavlik Director SuSE Labs
Quoted 7 lines · Vojtech Pavlik
Quoted 1 line · Bernhard Rosenkraenzer
Add some quirks to make a Shuttle PN-31 remote control work.
Can you send me the #define DEBUG output from hid-core and hid-input? I'm quite curious how much broken the device is and if it weren't easier to implement a way how to replace the HID Report descriptor inside hid-core instead of adding more quirks.
Unfortunately not ATM, I don't have the hardware anymore (it was part of the "If you get any new hardware, give it to me for a day before selling it, to make sure it supports Linux" deal I have with a local computer shop).
I'll gather the debug output if/when he sells another one.