The new implementation uses a message-only window to register raw input events.
XInput is still used when possible, but raw input is being used to handle device detection. (XInput is only enabled when an XInput device is plugged in, which also prevents us from having to load and poll the XInput library unnecessarily.)
For raw devices, the Windows HID parser library is used to parse the raw data. Unfortunately, the Windows 7.1 SDK does not ship hid.lib (which is in the WDK 7.1.0), so I am dynamically loading hid.dll to prevent pulling in a dependency on the WDK. (But perhaps we can ship hid.lib in the thirdparty libraries?)
Input devices other than XInput gamepads and the 3D mouse have not been tested very well yet. Adding keyboard and mouse support is still a TODO.
This also splits out the Windows implementation of InputDeviceManager into a separate subclass; it is intended that the other implementations will follow suit.
This fixes error messages being generated of the form:
AL lib: (WW) alSetError: Error generated on context 0x801cf8800, code 0xa003
This is caused by alSourceUnqueueBuffers being called without first checking whether processed buffers are available using alGetSourcei.
Fixes#180
This is required to make the test suite run with pipelining enabled.
It's necessary to let the draw thread do these tasks because that's where the OpenGL context is bound to. However, we need to let it access the data from the calling thread, so we have to temporarily change the pipeline stage of the draw thread.
I'm not really happy about this solution; it would be better to temporarily make the context current to the app thread, but we need a window for that, which we don't currently require to be passed into that method.
This deadlock happens when another thread holds a cycler lock and then attempts to call Pipeline::remove_cycler() while Pipeline::cycle() is running on the main thread.
The fix for this problem is threefold:
* Allow remove_cycler (and add_cycler) to be called while a cycle is in progress, by introducing a second lock
* Let cycle() not block if a dirty cycler can't be locked, instead trying other cyclers first
* Adding a way to let remove_cycler() check whether a cycler is currently in use by the main thread, and yielding if so
More information is on https://github.com/panda3d/panda3d/issues/217Fixes#217 (also see LP 1186880)
This adds a list of known devices so that we can assign custom behavior to them. I tested this with the T.Flight Hotas X flight stick and the NVIDIA Shield Controller.
I tested gamepads with a wireless 360 controller and a wired Xbox One controller using the 360Controller driver. I tested the joystick support using a T.Flight Hotas X.
This changes behaviour for sRGB textures, which weren't previously converting to the correct color space.
Also add unit tests for storing to PNMImage.
Closes: #212
rdb
Younguk Kim