| Age | Commit message (Collapse) | Author |
|---|
|
Drop the display lock when we exit from XChangeDeviceControl(), instead of
leaking it.
|
|
I wish that there was a better way to do this, but c'est la vie. It's our
exposed API, so we've just got to deal for now.
|
|
|
|
Don't throw a printf every time we get a DPN, and fill in all the fields
when copying the structure.
|
|
|
|
(We now compile again.)
Track the DEVICE_TOUCHSCREEN -> DEVICE_ABS_CALIB and new DEVICE_ABS_AREA
changes to the input protocol.
|
|
Add DEVICE_ENABLE for enabling/disabling devices, and an 'iscore'
flag to DEVICE_CORE that specifies whether or not the device is a
virtual core device.
|
|
held.
All callers of _XiCheckExtInit have already called XInput_find_display
first outside the lock, so just pass their copy of the XExtDisplayInfo
structure down. Besides being more correct, this should be slightly
faster. :-)
Thanks to Magnus Kessler <Magnus.Kessler@gmx.net> for identifying
the bug and proposing a workaround.
|
|
lock is held.
_XiGetExtensionVersion has two callers. One had already called
XInput_find_display, and the other could easily do so outside the
Display lock. So I change the _XiGetExtensionVersion interface to
accept a previously-computed XExtDisplayInfo from the caller.
Besides being more correct, this should be slightly faster. :-)
Thanks to Magnus Kessler <Magnus.Kessler@gmx.net> for identifying
the bug and proposing a workaround.
|
|
|
|
Add a DevicePresenceNotify event, which notes that something about the
device list changed.
|
|
Add a more sensible DEVICE_TOUCHSCREEN control, which allows you to
clip the x and y values.
Add a DEVICE_CORE control, which toggles the sending of core events
by extended devices.
|
|
Another brief run with indent.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
