- 11 Oct, 2021 12 commits
-
-
Alexandre Julliard authored
Signed-off-by:Alexandre Julliard <julliard@winehq.org>
-
Alexandre Julliard authored
Signed-off-by: -
Alexandre Julliard authored
Signed-off-by: -
Alexandre Julliard authored
Signed-off-by: -
Alexandre Julliard authored
Signed-off-by: -
Rémi Bernon authored
Signed-off-by:
Rémi Bernon <rbernon@codeweavers.com> Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Huw Davies authored
Signed-off-by:
Huw Davies <huw@codeweavers.com> Signed-off-by:
-
- 08 Oct, 2021 28 commits
-
-
Alexandre Julliard authored
Signed-off-by: -
Zebediah Figura authored
Fixes Collins English Dictionary. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=51802Signed-off-by:
Zebediah Figura <zfigura@codeweavers.com> Signed-off-by:
-
Robert Wilhelm authored
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=50758Signed-off-by:
Robert Wilhelm <robert.wilhelm@gmx.net> Signed-off-by:
Jacek Caban <jacek@codeweavers.com> Signed-off-by:
-
Jacek Caban authored
Signed-off-by:
-
Alexandre Julliard authored
Signed-off-by: -
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
And remove the device from the list of acquired devices if the callback indicates a read error. Signed-off-by:
-
Rémi Bernon authored
Otherwise we may later write the cancelled status to invalid memory. Also use a manual-reset event, as it should be for overlapped I/O, so all waiters are woken up on cancel. Signed-off-by:
-
Rémi Bernon authored
The read buffer is only used by the reading thread but the device state is not, we should only update it while holding the CS. Signed-off-by:
-
Rémi Bernon authored
This can cause some deadlocks as the internal thread will acquire the device CS when reading the state. Signed-off-by:
-
Rémi Bernon authored
Based on a patch from Andrew Eikum <aeikum@codeweavers.com>. Dark Souls Remasters checks for xinput devices when it receives a WM_DEVICECHANGE message. We would only poll for new devices if it had been at least 2 seconds since the last check. So often, DS would receive the message, but we would refuse to poll for devices, so the game would think no controller was present. This commit fixes that by also subscribing to event notifications and triggering a poll. Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
Signed-off-by:
-
Rémi Bernon authored
Otherwise we may later write the cancelled status to invalid memory. Also use a manual-reset event, as it should be for overlapped I/O, so all waiters are waken up on cancel. Signed-off-by:
-
Ivo Ivanov authored
Signed-off-by:
Ivo Ivanov <logos128@gmail.com> Signed-off-by:
-
Ivo Ivanov authored
Signed-off-by:
-
Huw Davies authored
Signed-off-by:
-
Jacek Caban authored
Signed-off-by:
-
Jacek Caban authored
Signed-off-by:
-
Jacek Caban authored
Link directly to win32u instead. Signed-off-by:
-
Jacek Caban authored
Signed-off-by:
-
Jacek Caban authored
Signed-off-by:
-
Daniel Lehman authored
Signed-off-by:
Daniel Lehman <dlehman@esri.com> Signed-off-by:
-
Rémi Bernon authored
If the hidclass.sys read requests aren't done fast enough. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=51824Signed-off-by:
-
Rémi Bernon authored
And use them for IOCTL_HID_GET_INPUT_REPORT ioctls. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=51824Signed-off-by:
-
Rémi Bernon authored
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=51824Signed-off-by:
-
Rémi Bernon authored
The design of hidclass.sys prevents any concurrent irps, there's no need to queue more. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=51824Signed-off-by:
-
