HID: logitech-hidpp: Don't restart IO, instead defer hid_connect() only
authorHans de Goede <hdegoede@redhat.com>
Tue, 10 Oct 2023 10:20:18 +0000 (12:20 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 20 Nov 2023 10:59:22 +0000 (11:59 +0100)
[ Upstream commit 11ca0322a41920df2b462d2e45b0731e47ff475b ]

Restarting IO causes 2 problems:

1. Some devices do not like IO being restarted this was addressed in
   commit 498ba2069035 ("HID: logitech-hidpp: Don't restart communication
   if not necessary"), but that change has issues of its own and needs to
   be reverted.

2. Restarting IO and specifically calling hid_device_io_stop() causes
   received packets to be missed, which may cause connect-events to
   get missed.

Restarting IO was introduced in commit 91cf9a98ae41 ("HID: logitech-hidpp:
make .probe usbhid capable") to allow to retrieve the device's name and
serial number and store these in hdev->name and hdev->uniq before
connecting any hid subdrivers (hid-input, hidraw) exporting this info
to userspace.

But this does not require restarting IO, this merely requires deferring
calling hid_connect(). Calling hid_hw_start() with a connect-mask of
0 makes it skip calling hid_connect(), so hidpp_probe() can simply call
hid_connect() later without needing to restart IO.

Remove the stop + restart of IO and instead just call hid_connect() later
to avoid the issues caused by restarting IO.

Now that IO is no longer stopped, hid_hw_close() must be called at the end
of probe() to balance the hid_hw_open() done at the beginning probe().

This series has been tested on the following devices:
Logitech Bluetooth Laser Travel Mouse (bluetooth, HID++ 1.0)
Logitech M720 Triathlon (bluetooth, HID++ 4.5)
Logitech M720 Triathlon (unifying, HID++ 4.5)
Logitech K400 Pro (unifying, HID++ 4.1)
Logitech K270 (eQUAD nano Lite, HID++ 2.0)
Logitech M185 (eQUAD nano Lite, HID++ 4.5)
Logitech LX501 keyboard (27 Mhz, HID++ builtin scroll-wheel, HID++ 1.0)
Logitech M-RAZ105 mouse (27 Mhz, HID++ extra mouse buttons, HID++ 1.0)

And by bentiss:
Logitech Touchpad T650 (unifying)
Logitech Touchpad T651 (bluetooth)
Logitech MX Master 3B (BLE)
Logitech G403 (plain USB / Gaming receiver)

Fixes: 498ba2069035 ("HID: logitech-hidpp: Don't restart communication if not necessary")
Suggested-by: Benjamin Tissoires <bentiss@kernel.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20231010102029.111003-2-hdegoede@redhat.com
Signed-off-by: Benjamin Tissoires <bentiss@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/hid/hid-logitech-hidpp.c

index a209d51..aa4f232 100644 (file)
@@ -4460,8 +4460,10 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id)
                         hdev->name);
 
        /*
-        * Plain USB connections need to actually call start and open
-        * on the transport driver to allow incoming data.
+        * First call hid_hw_start(hdev, 0) to allow IO without connecting any
+        * hid subdrivers (hid-input, hidraw). This allows retrieving the dev's
+        * name and serial number and store these in hdev->name and hdev->uniq,
+        * before the hid-input and hidraw drivers expose these to userspace.
         */
        ret = hid_hw_start(hdev, will_restart ? 0 : connect_mask);
        if (ret) {
@@ -4519,19 +4521,14 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id)
        flush_work(&hidpp->work);
 
        if (will_restart) {
-               /* Reset the HID node state */
-               hid_device_io_stop(hdev);
-               hid_hw_close(hdev);
-               hid_hw_stop(hdev);
-
                if (hidpp->quirks & HIDPP_QUIRK_DELAYED_INIT)
                        connect_mask &= ~HID_CONNECT_HIDINPUT;
 
                /* Now export the actual inputs and hidraw nodes to the world */
-               ret = hid_hw_start(hdev, connect_mask);
+               ret = hid_connect(hdev, connect_mask);
                if (ret) {
-                       hid_err(hdev, "%s:hid_hw_start returned error\n", __func__);
-                       goto hid_hw_start_fail;
+                       hid_err(hdev, "%s:hid_connect returned error %d\n", __func__, ret);
+                       goto hid_hw_init_fail;
                }
        }
 
@@ -4543,6 +4540,11 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id)
                                 ret);
        }
 
+       /*
+        * This relies on logi_dj_ll_close() being a no-op so that DJ connection
+        * events will still be received.
+        */
+       hid_hw_close(hdev);
        return ret;
 
 hid_hw_init_fail: