Re: [PATCH] HID: hid-logitech-hidpp: driver for m545/m546 mouse
flat view
From: VRan Liu <hidden>
Date: 2015-10-09 02:18:37
This is duplicated mail in plain text for content policy of mail list. Hi all, I use `solaar show` in https://github.com/pwr/Solaar to detect features. My m546 support 0x1B00(REPROG CONTROLS) and 0x1B03(REPROG CONTROLS V3), but doesn't have 0x1B04(REPROG CONTROLS V4). Rgd, VRan On Fri, Oct 9, 2015 at 4:57 AM, Benjamin Tissoires [off-list ref] wrote:
Hi Goffredo, On Thu, Oct 8, 2015 at 1:07 PM, Goffredo Baroncelli [off-list ref] wrote:quoted
Ping, Benjamin, Nestor, do you have any comments ?Sorry for not going back to you earlier. I do not see any obvious problems in the patch (few comments though), but I think we can do better (and for the M560 too). I have now a much better understanding of the MX Master, and this mouse also sends some Windows 8 gestures. We can disable them in a similar way the M560 does, but we have the per button release information. Also, I think with my latest hidpp patches, we should be able to drop the manual parsing of the mouse report. Could you please download libratbag from https://github.com/libratbag/libratbag, add your device id in src/driver-hidpp20.c, make and run: $ sudo ./tools/ratbag-command --verbose=raw list This should dump the list of supported features your mouse has, and if the feature 0x1b04 ("programmable keys/buttons") shows up, then we will be able to clean up the code and have it working for these 3 mice. Few remarks inlined:quoted
BR G.Baroncelli On 2015-09-29 20:20, Goffredo Baroncelli wrote:quoted
From: Goffredo Baroncelli <redacted> Add support for the Logitech m545/m546 mouse. This mouse is designed for windows 8. So when the side buttons are pressed, it sends some keyboard keys instead of mouse buttons. Original-author: VRan Liu [off-list ref] Signed-off-by: Goffredo Baroncelli <redacted> --- In the comments of the code there is a brief description of the protocol. I developed a similar solution for the Logitech m560 mouse, which was included in the kernel v4.2.1. Then I was contacted by VRan Liu; he adapted my patch for this mouse and he asked me to send its patch to the mailing list. drivers/hid/hid-logitech-hidpp.c | 143 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 142 insertions(+), 1 deletion(-)diff --git a/drivers/hid/hid-logitech-hidpp.c b/drivers/hid/hid-logitech-hidpp.c index 4841964..6de57c0 100644 --- a/drivers/hid/hid-logitech-hidpp.c +++ b/drivers/hid/hid-logitech-hidpp.c@@ -41,8 +41,9 @@ MODULE_PARM_DESC(disable_raw_mode, #define HIDPP_QUIRK_CLASS_WTP BIT(0) #define HIDPP_QUIRK_CLASS_M560 BIT(1) +#define HIDPP_QUIRK_CLASS_M545 BIT(2) -/* bits 2..20 are reserved for classes */ +/* bits 3..20 are reserved for classes */ #define HIDPP_QUIRK_DELAYED_INIT BIT(21) #define HIDPP_QUIRK_WTP_PHYSICAL_BUTTONS BIT(22)@@ -1132,6 +1133,131 @@ static int m560_input_mapping(struct hid_device *hdev, struct hid_input *hi, return -1; } +/* ------------------------------------------------------------------------- */ +/* Logitech M545/M546 devices */ +/* ------------------------------------------------------------------------- */ + +/* + * Logitech M545 protocol overview + * + * The Logitech M545 mouse, is designed for windows 8. When the sides buttons + * are pressed, it sends some keyboard keys events instead of buttons ones. + * + * forward button -> Super_R + * backward button -> Super_L+'d' (press only) + * NB: press-only means that when the button is pressed, the + * KeyPress/ButtonPress and KeyRelease/ButtonRelease events are generated + * together sequentially; instead when the button is released, no event is + * generated ! + * + * for the sides button it sends: + * side 1 button (forward) press 11<xx>15 0000a900... + * (then keyboard events) + * side 2 button (backward) press 11<xx>15 0000ae00... + * (then keyboard events) + */ + + +struct m545_private_data { + struct input_dev *input; +}; + +#define M545_SUB_ID 0x15 + +static int m545_allocate(struct hid_device *hdev) +{ + struct hidpp_device *hidpp = hid_get_drvdata(hdev); + struct m545_private_data *d; + + d = devm_kzalloc(&hdev->dev, sizeof(struct m545_private_data), + GFP_KERNEL); + if (!d) + return -ENOMEM; + + hidpp->private_data = d; + + return 0; +}; + +static int m545_raw_event(struct hid_device *hdev, u8 *data, int size) +{ + struct hidpp_device *hidpp = hid_get_drvdata(hdev); + struct m545_private_data *mydata = hidpp->private_data; + + /* sanity check */ + if (!mydata || !mydata->input) { + hid_err(hdev, "error in parameter\n"); + return -EINVAL; + } + + if (size < 7) { + hid_err(hdev, "error in report\n"); + return 0; + } + + if (data[0] == REPORT_ID_HIDPP_LONG && + data[2] == M545_SUB_ID && data[6] == 0x00) { + /* + * m545 mouse report for middle, forward and backward button + * + * data[0] = 0x11 + * data[1] = device-id + * data[2] = 0x15 + * data[5] = 0xa9 -> forward + * 0xae -> backward + * 0x00 -> release all + * data[6] = 0x00 + */ + + switch (data[5]) { + case 0xa9: + input_report_key(mydata->input, BTN_FORWARD, 1); + break; + case 0xae: + input_report_key(mydata->input, BTN_BACK, 1); + break; + case 0x00: + input_report_key(mydata->input, BTN_BACK, 0); + input_report_key(mydata->input, BTN_FORWARD, 0); + break; + default: + hid_err(hdev, "error in report\n"); + return 0; + } + input_sync(mydata->input); + } + + return 1; +} + +static void m545_populate_input(struct hidpp_device *hidpp, + struct input_dev *input_dev, bool origin_is_hid_core) +{ + struct m545_private_data *mydata = hidpp->private_data; + + mydata->input = input_dev; + + __set_bit(EV_KEY, mydata->input->evbit); + __set_bit(BTN_MIDDLE, mydata->input->keybit); + __set_bit(BTN_RIGHT, mydata->input->keybit); + __set_bit(BTN_LEFT, mydata->input->keybit); + __set_bit(BTN_BACK, mydata->input->keybit); + __set_bit(BTN_FORWARD, mydata->input->keybit); + + __set_bit(EV_REL, mydata->input->evbit); + __set_bit(REL_X, mydata->input->relbit); + __set_bit(REL_Y, mydata->input->relbit); + __set_bit(REL_WHEEL, mydata->input->relbit); + __set_bit(REL_HWHEEL, mydata->input->relbit);Not a big fan of this, given that the default one should already have this.quoted
quoted
+} + +static int m545_input_mapping(struct hid_device *hdev, struct hid_input *hi, + struct hid_field *field, struct hid_usage *usage, + unsigned long **bit, int *max) +{ + return -1;If we just ignore the keyboard mapping, the mouse should be populated properly (hopefully). Again, I will need to push the MX Master code I have somewhere to check if this will also work for your mice. Cheers, Benjaminquoted
quoted
+} + /* -------------------------------------------------------------------------- */ /* Generic HID++ devices */ /* -------------------------------------------------------------------------- */@@ -1147,6 +1273,9 @@ static int hidpp_input_mapping(struct hid_device *hdev, struct hid_input *hi, else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M560 && field->application != HID_GD_MOUSE) return m560_input_mapping(hdev, hi, field, usage, bit, max); + else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M545 && + field->application != HID_GD_MOUSE) + return m545_input_mapping(hdev, hi, field, usage, bit, max); return 0; }@@ -1158,6 +1287,8 @@ static void hidpp_populate_input(struct hidpp_device *hidpp, wtp_populate_input(hidpp, input, origin_is_hid_core); else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M560) m560_populate_input(hidpp, input, origin_is_hid_core); + else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M545) + m545_populate_input(hidpp, input, origin_is_hid_core); } static void hidpp_input_configured(struct hid_device *hdev,@@ -1247,6 +1378,8 @@ static int hidpp_raw_event(struct hid_device *hdev, struct hid_report *report, return wtp_raw_event(hdev, data, size); else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M560) return m560_raw_event(hdev, data, size); + else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M545) + return m545_raw_event(hdev, data, size); return 0; }@@ -1408,6 +1541,10 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id) ret = m560_allocate(hdev); if (ret) goto allocate_fail; + } else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M545) { + ret = m545_allocate(hdev); + if (ret) + goto allocate_fail; } INIT_WORK(&hidpp->work, delayed_work_cb);@@ -1502,6 +1639,10 @@ static const struct hid_device_id hidpp_devices[] = { HID_DEVICE(BUS_USB, HID_GROUP_LOGITECH_DJ_DEVICE, USB_VENDOR_ID_LOGITECH, 0x402d), .driver_data = HIDPP_QUIRK_DELAYED_INIT | HIDPP_QUIRK_CLASS_M560 }, + { /* Mouse logitech M545/M546 */ + HID_DEVICE(BUS_USB, HID_GROUP_LOGITECH_DJ_DEVICE, + USB_VENDOR_ID_LOGITECH, 0x4028), + .driver_data = HIDPP_QUIRK_CLASS_M545 }, { HID_DEVICE(BUS_USB, HID_GROUP_LOGITECH_DJ_DEVICE, USB_VENDOR_ID_LOGITECH, HID_ANY_ID)},-- gpg @keyserver.linux.it: Goffredo Baroncelli <kreijackATinwind.it> Key fingerprint BBF5 1610 0B64 DAC6 5F7D 17B2 0EDA 9B37 8B82 E0B5