-
Notifications
You must be signed in to change notification settings - Fork 403
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
libusb: Allow building using non-POSIX runtime #582
libusb: Allow building using non-POSIX runtime #582
Conversation
I think CMake files need to be updated as well. |
I don't think so, it work as-is...? |
d22d740
to
1ffe07e
Compare
If you look at the CMake file, I think you need to add options to build both native Windows version and Windows libusb version, same for native macOS version and macOS libusb versoon. I think you can reuse Lines 41 to 44 in d3013f0
|
The macOS build can be made as a separate PR and isn't dependent on anything here. |
No problem, you can create a new PR for macOS. Thanks. |
BTW, just FYI, there is a bug with the libusb backend which is difficult to fix. It may not affect many devices though. |
I'd like to test this PR but unfortuantely I do not see how to build this PR without the changes to CMake or the legacy auto-tools script. Please help to explain how you build this PR. Thanks. As of now, Windows build will only use the native Windows HID backend and not libusb. |
windows/hid.c
Outdated
@@ -56,9 +56,11 @@ typedef LONG NTSTATUS; | |||
#include "hidapi_hidclass.h" | |||
#include "hidapi_hidsdi.h" | |||
|
|||
#ifndef HIDAPI_USING_SDL_RUNTIME |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This change looks unrelated
My guess if that SDL has something like: typedef struct
{
//...
} hid_device_thread_state;
// more thread_state defines
...
#define HIDAPI_USING_SDL_RUNTIME
#include "hidapi/libusb/hid.c" am I correct on the approach this is used? UPD: only now I checked #581, and my guess is right. |
While libsdl-org/SDL@57b33ae makes perfect sense in the context of SDL project, this change is not at its best in the context of HIDAPI project. First of all: the "SDL" term shouldn't be used among HIDAPI code. What I suggest:
It would also greatly simplify making a native HIDAPI port of libusb backend on Windows. UPD: I imaging it would require some (presumably small but) extra effort on SDL side to be ported to the upstream version of HIDAPI after the change with my suggestions accepted, but I have a strong feeling that it is worth doing anyway longterm. |
This change doesn't actually make it possible to build libusb backend of HIDAPI on Windows, but rather it is just a nesessary refactoring to allow replacing pthread with any other threading library. I'm generally agree with the aproach, just have a few conserns regarding implementation (my prev comment). |
1ffe07e
to
3510842
Compare
I've updated the pull request and scoped it just to create the threading abstraction. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Perfect, just a few small comments.
5e5c7bd
to
092d62b
Compare
…ms without pthreads
092d62b
to
081b696
Compare
Okay, I think this addresses everything. Please let me know if you have any other suggestions! |
clock_gettime(CLOCK_REALTIME, ts); | ||
} | ||
|
||
static void hidapi_thread_addtime(hidapi_timespec *ts, int milliseconds) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
oh, I haven't though of this one, but great suggestion, thanks
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Great work!
Now it should be relatively straight-forward to implement the native Windows port of libusb backend.
I'll keep this change in review for a few more days, so the comunity has a chance to catch up/review.
That will be great! |
static void hidapi_thread_gettime(hidapi_timespec *ts) | ||
{ | ||
clock_gettime(CLOCK_REALTIME, ts); | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I wonder if it would make sense to seperate the two time functions into an own entity like HIDAPI_TIME_REF_INCLUDE
.
This would allow to use a application specific time reference for future features like time stamping of incoming reports.
Other time references could be a the soundcard clock in an audio application or the network clock in a streaming application.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
As of this moment (and for many years before that) the only place where this time related thing is used - is for thread_cond_timedwait
.
As of this moment - this looks compact and we avoid additional abstraction/implementation.
If we need an abstraction in the future - we can easily introduce it later.
commit 4339113 Author: Sylvain <[email protected]> Date: Thu Jun 22 10:48:12 2023 +0200 SDL_DBus_AppendDictWithKeyValue: allows several key/value for other PRs commit 8aee690 Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 18:42:25 2023 -0700 Make sure we send update complete events for delayed guide buttons commit 4c9fb3e Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 13:59:53 2023 -0700 Added the events SDL_EVENT_JOYSTICK_UPDATE_COMPLETE and SDL_EVENT_GAMEPAD_UPDATE_COMPLETE This allows the application to tell when a joystick polling cycle is complete and can process state changes as a single atomic update. It is disabled by default, at least for now. commit 808d83d Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 13:49:09 2023 -0700 Allow specifying APP_PLATFORM and APP_ABI on the command line commit 214d5da Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 12:14:50 2023 -0700 Removed 100 ms hitch when querying third party Nintendo Switch controllers that don't respond to request for info Also take advantage of the fact that we know whether the device is connected over Bluetooth now. commit 3694dab Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 10:28:45 2023 -0700 Use default sensor calibration if we can't read it from the Nintendo Switch controller Fixes libsdl-org#7830 commit 6306ee9 Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 09:33:59 2023 -0700 List the available haptic devices in testhaptic commit 8cf5dc9 Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 09:18:19 2023 -0700 Build on Android targeting the arm64-v8a architecture by default This speeds up iteration time and covers most customer devices commit 5f3213e Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 08:40:30 2023 -0700 Added support for gamepad sensor fusion with the Razer Kishi commit 91198ba Author: Simon McVittie <[email protected]> Date: Tue Jun 20 13:22:55 2023 +0100 ibus: Handle error when getting the D-Bus machine ID It is possible for retrieving the machine ID to fail, either because dbus was installed incorrectly (machine ID absent or corrupt), or in 32-bit builds, because stat() on the machine ID fails with EOVERFLOW if it has an out-of-range timestamp or inode number. dbus has historically treated this as a faulty installation, raising a warning which by default causes the process to crash. Unfortunately, dbus_get_local_machine_id() never had a way to report errors, so it has no alternative for that (bad) error handling. In dbus >= 1.12.0, we can use dbus_try_get_local_machine_id() to get the same information, but with the ability to cope gracefully with errors. ibus won't work in this situation, but that's better than crashing. Mitigates: ValveSoftware/steam-for-linux#9605 Signed-off-by: Simon McVittie <[email protected]> commit 3ddbeab Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 01:17:44 2023 -0700 Moved Android sensor event handling to a separate thread This makes it so you can interact with sensors on multiple threads, as long as only one thread initializes and cleans up the sensor subsystem. This also has the benefit that sensor data is available as soon as possible. commit 329e1b8 Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 19:42:01 2023 -0700 Update the sensors before the joysticks so the gamepad code gets fresh sensor readings commit 1a9c04e Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 19:41:20 2023 -0700 Don't try to update the gamepad fusion sensors manually, instead rely on the normal update flow commit 20ea351 Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 16:17:34 2023 -0700 Use a separate sensor watching function for gamepad events to avoid reliance on system sensor events and prevent a potential deadlock commit 70e43c1 Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 08:36:30 2023 -0700 Added support for blitting SDL_PIXELFORMAT_INDEX1LSB Fixes libsdl-org#7844 (cherry picked from commit a3d4fd7) commit c6ee978 Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:54 2023 -0700 Fixed DualSense controllers not being picked up by the HIDAPI driver The hidraw device may take additional time to get the correct permissions for us to open it. In my tests on Steam Deck hardware, this ranges between 5-8ms. commit 4e81b4e Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:43 2023 -0700 Added SDL_HINT_VIDEO_X11_SCALING_FACTOR to allow overriding the content scale on X11 (thanks Andres!) commit 210c135 Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:38 2023 -0700 Implement SDL_HINT_WINDOW_ACTIVATE_WHEN_RAISED for X11 To match the focus stealing prevention logic from windows/osx. I didn't implement SDL_HINT_WINDOW_ACTIVATE_WHEN_SHOWN as there is no standard mechanism for us to do so. In most cases marking a window as OverrideRedirect will cause to not acquire focus when mapped. And in steam we are doing that when appropriate. I still left a note in X11_ShowWindow() regarding this behaviour. commit 9351bf6 Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:22 2023 -0700 Add handling for SDL_HINT_WINDOW_ACTIVATE_WHEN_RAISED and SDL_HINT_WINDOW_ACTIVATE_WHEN_SHOWN on macOS - Cocoa_ShowWindow will order a window that is not being activated below the key window (if one exists) then set visible - Cocoa_RaiseWindow calls -[NSWindow orderFront:] without changing the key window commit f168f9c Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 12:42:55 2023 -0700 Added support for the GameSir G4 Pro We can't read device info or IMU calibration from this controller, and it has no gyro or accelerometer, but is otherwise perfectly functional. commit b770644 Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 09:46:46 2023 -0700 Fixed building on older kernel headers commit b98494a Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 09:39:50 2023 -0700 Fixed building on older kernel headers commit 9fe384b Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 09:28:27 2023 -0700 Fixed display orientation function names for SDL 3.0 convention commit d91e96e Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 08:59:52 2023 -0700 Use SDL_HINT_GAMECONTROLLER_SENSOR_FUSION as a list of controllers to enable sensor fusion There are too many wraparound style controllers out there to enumerate them all, so instead make this a user option in applications that support it. commit 610c31c Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 08:28:37 2023 -0700 Generalized the sensor coordinate transform for wraparound gamepads commit 9eb5eab Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 08:04:34 2023 -0700 Use the correct orientation transformation based on whether the device is naturally landscape or portrait commit e6d1ba2 Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 00:52:40 2023 -0700 Added the concept of display natural orientation Also renamed SDL_GetDisplayOrientation() SDL_GetDisplayCurrentOrientation() The natural orientation of the primary display is the frame of reference for accelerometer and gyro sensor readings. commit 8de6ce7 Author: Sam Lantinga <[email protected]> Date: Fri Jun 16 17:48:34 2023 -0700 Rotate the sensor axes to match gamepad orientation when using the device sensors for game controllers commit a9c86e5 Author: Sam Lantinga <[email protected]> Date: Fri Jun 16 17:48:18 2023 -0700 Added the Razer Edge controller to the list of wraparound controllers commit c207cd3 Author: Sam Lantinga <[email protected]> Date: Fri Jun 16 16:03:55 2023 -0700 Added the Razer Junglecat to the wraparound controller list commit e4f53e6 Author: Ozkan Sezer <[email protected]> Date: Sat Jun 17 01:01:10 2023 +0300 testevdev.c: comment out two unused data to fix build. commit 42e4639 Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 22:03:44 2023 -0700 For gamepads that don't have their own sensors, try to use the system sensors. This allows using the gyro and accelerometer in handheld devices in conjunction with built-in or wraparound controllers. commit d584592 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 15:16:44 2023 +0100 linux: If the kernel specifically tells us the device type, trust it If a device is positively identified as an accelerometer, pointing stick or clickpad, then we don't need to second-guess it. In practice this does not change the result for any device in our test data, so add some artificial records that exercise this. Signed-off-by: Simon McVittie <[email protected]> commit 0d5aa70 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 15:14:33 2023 +0100 linux: Pass evdev properties when guessing device type In newer kernels, devices that can be positively identified as a particular device type (for example accelerometers) get a property bit set. Plumb this information through into the function, but don't use it for anything just yet. Signed-off-by: Simon McVittie <[email protected]> commit a4ce721 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 15:15:31 2023 +0100 testevdev: Allow device properties to be fully populated The props array was too small for the highest property bits to be set, although in practice this didn't matter since only the lower-order bits have a meaning. Make it consistent with all the others. Signed-off-by: Simon McVittie <[email protected]> commit fa0ca3d Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:53:29 2023 +0100 linux: Distinguish between "device with keys" and a full keyboard udev distinguishes between ID_INPUT_KEY, a device with any keyboard keys at all (including for example USB numeric keypads, Bluetooth headsets with volume control buttons, and some game controllers; and ID_INPUT_KEYBOARD, a reasonably fully-featured keyboard that you could use for general-purpose text entry. If we do the same here, then it's useful input to our heuristics for identifying devices: for example, a device with ID_INPUT_KEY could reasonably be a gamepad, but a device with ID_INPUT_KEYBOARD certainly isn't. Resolves: libsdl-org#7827 Signed-off-by: Simon McVittie <[email protected]> commit 9b7a9ca Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:29:53 2023 +0100 testevdev: Add some more laptop built-in devices Signed-off-by: Simon McVittie <[email protected]> commit 00b6db6 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:29:41 2023 +0100 testevdev: Add some EVIOCGNAME and USB name strings to test data We don't currently use these for anything, but we might start using them as input to our heuristics as part of libsdl-org#7697. Signed-off-by: Simon McVittie <[email protected]> commit 3772d6c Author: Simon McVittie <[email protected]> Date: Thu Jun 15 19:46:35 2023 +0100 testevdev: Add raw HID report descriptors where available We don't currently use these in our device-classification heuristic, but it could be a useful input in future. Thanks to Sam Lantinga, Ben Fradella, kevenwyld and schlegp for providing some of these. Signed-off-by: Simon McVittie <[email protected]> commit 2b7556f Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:35:16 2023 +0100 testevdev: Correct typo in bus type for Xbox Series S|X via Bluetooth All Bluetooth devices are bus type 0x0005. Signed-off-by: Simon McVittie <[email protected]> commit c13e511 Author: Simon McVittie <[email protected]> Date: Tue Jun 13 17:48:50 2023 +0100 testevdev: Try to correct Wii devices guessed from kernel source The comments here disagreed with the actual bytes. According to https://elixir.bootlin.com/linux/v6.3.7/source/drivers/hid/hid-wiimote-modules.c, the Balance Board reports BTN_A and ABS_HAT0X, HAT0Y, HAT1X and HAT1Y. This means the comments here were correct, but the .abs bits shown were in the wrong byte. Matching the Wii U Pro Controller against the same kernel source, it appears to be correct: it's the same representation as a PS3 gamepad, except that it lacks the Z and RZ axes for analogue triggers. Signed-off-by: Simon McVittie <[email protected]> commit ffdafcd Author: Simon McVittie <[email protected]> Date: Tue Jun 13 17:26:51 2023 +0100 testevdev: Verify most Wii devices against real hardware Some of the test data previously seen here was guessed from kernel source code, and not all of it was correct. The following devices have now been verified against `evemu-describe` output with Linux 6.3 (thanks to Jeremy Whiting for collecting this): - basic Wiimote - buttons - 3-axis accelerometer - infra-red sensor for Sensor Bar location (precise aim) - Motion Plus accessory (3-axis gyroscope) - Nunchuck accessory (joystick, 2 buttons, second 3-axis accelerometer) - Classic Controller accessory (a complete traditional gamepad) Signed-off-by: Simon McVittie <[email protected]> commit 739f783 Author: SDL Wiki Bot <[email protected]> Date: Fri Jun 16 12:22:18 2023 +0000 Sync SDL3 wiki -> header commit 26df689 Author: Anonymous Maarten <[email protected]> Date: Fri Jun 16 14:21:33 2023 +0200 docs: expand CMake documentation + add minimal CMake project for reporting issues commit 378e33b Author: Sylvain <[email protected]> Date: Fri Jun 16 10:05:58 2023 +0200 Android: potential ANR during onKeyDown/Up SDLActivity may call onNativeKeyDown, while application is quitting commit e72935a Author: Sam Lantinga <[email protected]> Date: Thu Jun 15 16:43:35 2023 -0700 Check for modff in addition to modf commit dab4f29 Author: Sam Lantinga <[email protected]> Date: Thu Jun 15 16:43:22 2023 -0700 Cleanup spacing commit 16b57d2 Author: Simon McVittie <[email protected]> Date: Thu Jun 15 19:25:23 2023 +0100 testevdev: Add details of another driving simulator controller Thanks to Ben Fradella. Signed-off-by: Simon McVittie <[email protected]> commit 2b00751 Author: Simon McVittie <[email protected]> Date: Thu Jun 15 15:50:03 2023 +0100 testevdev: Expand test data for X-Box One Elite 2 This slightly newer device than the one from libsdl-org#7814 is functionally equivalent when connected via USB. When connected via Bluetooth, it has a different button mapping. Thanks to Sam Lantinga. Signed-off-by: Simon McVittie <[email protected]> commit 62ed6f4 Author: Mathieu Eyraud <[email protected]> Date: Thu Jun 15 09:59:07 2023 +0200 Use SDL_strdup instead of strdup commit 73927b0 Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 15:16:21 2023 -0700 Removed unused function commit 2e465ae Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 11:05:10 2023 -0700 Revert "Added SDL_nextafter() and SDL_nextafterf()" This reverts commit bc5d074. It's not clear that we need these yet, so I'm going to remove them for now. commit 0c16f4f Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 10:15:06 2023 -0700 Fixed third-party Nintendo Switch Pro controllers shutting down when we try to set the home LED. This fixes the PDP Afterglow Wireless Deluxe Controller. commit 23e007d Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 00:02:32 2023 -0700 Fixed third party Nintendo Switch Pro Controller resetting when being sent an unsupported command to set the Home LED Tested with the PowerA Fusion Pro Wireless Controller in Bluetooth mode commit 8c95bd8 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 22:40:51 2023 -0700 Allow switching licensed Nintendo Switch Pro controllers into gyro input mode commit cdfc0c5 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 22:20:58 2023 -0700 Added support for the PowerA Fusion Pro Wireless Controller in Bluetooth mode This controller shows up with a VID/PID of 0, but has full functionality over Bluetooth commit 0f4b15e Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:53:10 2023 -0700 Removed more Linux Xbox mappings in favor of the automatic mapping This fixes the Xbox Series X share button on Linux 5.x kernels. commit 883b0f4 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:49:33 2023 -0700 Cleanup for previous change, fixing typos, etc. commit 9567989 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:41:52 2023 -0700 Use the automatic mapping instead of a hardcoded one for Xbox controllers on Linux This is much more robust and able to dynamically create a mapping for Xbox One S, Xbox Series X, and Xbox Elite 2 controllers. commit db1d4d3 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:40:30 2023 -0700 Added automatic mapping support for Xbox controllers on the 6.x Linux kernels This automatically adds support for the share button and paddles when present. commit 0f24956 Author: Sylvain <[email protected]> Date: Tue Jun 13 23:10:29 2023 +0200 testautomation_hints.c: free hint memory commit 2c37178 Author: Sylvain <[email protected]> Date: Tue Jun 13 23:00:00 2023 +0200 testautomation_events.c: initialize "timestamp" to solve "conditional jump or move depends on uninitialised value" commit 56ba7f2 Author: Simon McVittie <[email protected]> Date: Tue Jun 13 16:53:56 2023 +0100 testevdev: Add details of X-Box One Elite 2 via USB Thanks to iacore for capturing these. Signed-off-by: Simon McVittie <[email protected]> commit 7448451 Author: Simon McVittie <[email protected]> Date: Tue Jun 13 16:47:56 2023 +0100 testevdev: Provide a pointer to more information about adding test-cases Signed-off-by: Simon McVittie <[email protected]> commit bc5d074 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 07:34:50 2023 -0700 Added SDL_nextafter() and SDL_nextafterf() commit b0677f4 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 10:06:36 2023 -0700 Added automatic mapping for Xbox Elite paddles using the xpadneo driver We can't actually tell yet whether a controller has paddles, so this code isn't effective, but I'll file an upstream issue and see if we can get that resolved. commit 071d1e2 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 10:05:19 2023 -0700 Fixed joystick vendor detection in Linux automatic gamepad mapping commit 5a62a45 Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 23:55:56 2023 -0700 Fixed building with the 16.1.4479499 Android toolchain (cherry picked from commit b9d1c48) commit cf1dc66 Author: Simon McVittie <[email protected]> Date: Fri Jun 9 13:24:21 2023 +0100 linux: Improve gamepad mapping heuristic to accept Android conventions This heuristic for gamepads without a more specific mapping already tried two incompatible conventions for handling triggers: the Linux Gamepad Specification uses hat switch 2 for the triggers (for whatever reason), but the de facto standard set by the drivers for older Xbox and Playstation controllers represents each trigger as the Z-axis of the nearest analog stick. Android documentation encourages Bluetooth gamepad manufacturers to use a third incompatible convention where the left and right triggers are represented as the brake and gas pedals of a driving simulator controller. The Android convention also changes the representation of the right stick: instead of using X and Y rotation as a second pair of axes, Android uses Z position as a second horizontal axis, and Z rotation as a second vertical axis. Try to cope gracefully with all of these. This will hopefully resolve the issue described in libsdl-org#5406 (when using unpatched kernels). Signed-off-by: Simon McVittie <[email protected]> commit c4d49fa Author: Simon McVittie <[email protected]> Date: Fri Jun 9 13:06:30 2023 +0100 linux: Reduce magic numbers when mapping gamepad axes The bitfield `mapped` has two different sets of meanings, depending whether we're setting up the triggers or the d-pad. Represent them as symbolic constants rather than opaque integers. Signed-off-by: Simon McVittie <[email protected]> commit dec0dbf Author: Sam Lantinga <[email protected]> Date: Mon Jun 12 16:57:41 2023 -0700 Fixed enumerating Steam Controllers on iOS commit d95dbe7 Author: Sam Lantinga <[email protected]> Date: Mon Jun 12 15:39:57 2023 -0700 Fixed n3ds build commit cdc40ee Author: Sam Lantinga <[email protected]> Date: Mon Jun 12 15:26:34 2023 -0700 Reduce the chance of destroying the joystick mutex while it's in use Fixes libsdl-org#7811 commit 6ab846b Author: Linus Probert <[email protected]> Date: Mon Jun 12 18:27:40 2023 +0200 clipboard: Fixes additional x11 clipboard bugs found in tests commit a2ba5e9 Author: Linus Probert <[email protected]> Date: Mon Jun 12 13:06:16 2023 +0200 clipboard: Fixes testautomation fails introduced by clipboard changes commit 4e0f94e Author: Sylvain <[email protected]> Date: Mon Jun 12 11:45:46 2023 +0200 Android: add timeout when waiting the SDL thread to finish C SDLmain() thread might have started (mSDLThread.start() called) while the SDL_Init() might not have been called yet, and so the previous QUIT event will be discarded by SDL_Init() and app is running, not exiting. This is reprocible by adding instrumentation code in the SDLActivity. And hopefully, this could fix an ANR, where SDLActivity is in WAITING state (in thread.join()): at java.lang.Thread.join (Thread.java:1519) at org.libsdl.app.SDLActivity.onDestroy (SDLActivity.java) while SDLThread seems to be running commit 125e742 Author: Ryan C. Gordon <[email protected]> Date: Sun Jun 11 12:43:47 2023 -0400 cocoa: Warp mouse to center of window before enabling relative mouse. This prevents the case where the mouse might be at the edge of the window when enabling relative mode, which confuses macOS, at it might believe the user is attempting to resize the window. Fixes libsdl-org#6994. (cherry picked from commit 2afb49b) commit 4cfacd5 Author: Ozkan Sezer <[email protected]> Date: Sat Jun 10 20:50:02 2023 +0300 SDL_dynapi.c: minor cosmetics. commit 0103ec1 Author: Sam Lantinga <[email protected]> Date: Sat Jun 10 08:43:43 2023 -0700 Fixed accidental commit commit 7f86415 Author: SDL Wiki Bot <[email protected]> Date: Sat Jun 10 15:42:15 2023 +0000 Sync SDL3 wiki -> header commit 281018f Author: Sam Lantinga <[email protected]> Date: Sat Jun 10 08:39:20 2023 -0700 Make it clear that you can't mix 2D rendering and the window surface API Also added functions to query and destroy the window surface so you can switch between modes if you want. See pygame-community/pygame-ce#2190 for more details. commit 5490873 Author: Sam Lantinga <[email protected]> Date: Fri Jun 9 18:41:21 2023 -0700 Fixed querying device info on the MOBAPAD M073 The query packet needs to contain valid rumble data in order to be accepted by the controller. Fixes libsdl-org#7788 commit 2042e9c Author: Sam Lantinga <[email protected]> Date: Fri Jun 9 17:34:12 2023 -0700 Only update the serial number if it hasn't already been set This fixes the serial number for Nintendo Switch Pro, which is queried from the hardware in device initialization, and was later clobbered by the USB string which isn't correct. commit c8051b1 Author: Sam Lantinga <[email protected]> Date: Fri Jun 9 14:34:59 2023 -0700 Fixed reading input from the Razer Atrox Arcade Stick using Windows Gaming Input commit aaccf3c Author: Sylvain <[email protected]> Date: Fri Jun 9 14:21:17 2023 +0200 X11: common function to wait for the WM before sending SDL Window events commit 8096f72 Author: Sylvain <[email protected]> Date: Fri Jun 9 09:41:54 2023 +0200 Android: add robustness. check that the native code is run for the first time. commit ae9d8ac Author: Sylvain <[email protected]> Date: Fri Jun 9 09:27:39 2023 +0200 Android: when an EventFilter is binded to SDL_EVENT_DID_ENTER_FOREGROUND event and triggered, GL context is expected to be already restored. commit 4c0758a Author: Sam Lantinga <[email protected]> Date: Thu Jun 8 12:39:32 2023 -0700 Fixed crash if display couldn't be found in SDL_UpdateFullscreenMode() commit d9c17e7 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 18:24:09 2023 +0100 testevdev: Add details of some more 8BitDo devices Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit a2ed083 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 17:30:30 2023 +0100 testevdev: Describe several equivalent devices ValveSoftware/steam-devices#34 lists several more devices that are functionally equivalent to this one from the point of view of their evdev metadata. Thanks to apgrc. Signed-off-by: Simon McVittie <[email protected]> commit f4a53e7 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:13:38 2023 +0100 testevdev: Add test data for a DualSense (PS5) gamepad Also make details of PS4 gamepads (which are very similar from an evdev point of view) more specific. Thanks to Sam Lantinga and Jeremy Whiting for recording these. Signed-off-by: Simon McVittie <[email protected]> commit 2fb1df1 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 16:00:27 2023 +0100 testevdev: Add test data for Nintendo Switch Joy-Cons via Bluetooth Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit 8a82e06 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:51:10 2023 +0100 testevdev: Add test data for Microsoft Xbox Series S|X Controller Like the Stadia controller, this is unusual because it represents the Share button as the Record key from a multimedia keyboard (as of Linux 6.2.11 with the xpad driver). Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit ff01b43 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:48:46 2023 +0100 testevdev: Add Google Stadia controller This is a bit unusual because it has a small number of what would ordinarily be keyboard keys. Thanks to Jeremy Whiting for recording this. Signed-off-by: Simon McVittie <[email protected]> commit 9ad0b8b Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:47:50 2023 +0100 testevdev: Add test data for another Switch Pro Controller A newer evemu-describe transcript has this same controller with its buttons mapped differently, presumably a result of driver changes in the Linux kernel. Either way, we should recognise it as a gamepad. Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit 7b526d0 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:46:52 2023 +0100 testevdev: Finish incomplete data for Switch Pro Controller via USB This didn't include any buttons, which I assume was because I transcribed them incorrectly rather than reflecting reality. Confirmed against another Switch Pro Controller on a more recent kernel (thanks to Jeremy Whiting). Signed-off-by: Simon McVittie <[email protected]> commit 4c33ef9 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:45:17 2023 +0100 testevdev: Note a functionally-equivalent device We don't need to re-test the heuristic with the same input data, but knowing that another device has equivalent evdev metadata is useful information to record. Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit 4c035dc Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:14:19 2023 +0100 testevdev: Add another laptop touchpad This was reported by Rémi Bernon as an example of older SDL's non-udev code path going wrong for touchpads when the invoking user happens to be in the input group, which I believe was fixed by fdd945f. Signed-off-by: Simon McVittie <[email protected]> commit 32d015a Author: Sam Lantinga <[email protected]> Date: Thu Jun 8 10:46:56 2023 -0700 Fixed PS3 controller initialization over Bluetooth Now that we have hidapi that knows whether the controller is connected via Bluetooth or USB, this is much easier to fix. commit e15da19 Author: Frank Praznik <[email protected]> Date: Wed Jun 7 15:04:49 2023 -0400 video: Apply cleared flag states when restoring hidden windows Hiding a window and then clearing the minimized, maximized, fullscreen, or grabbed flags wouldn't result in the new state being applied to the window when shown, as the pending flags would always be zero, resulting in the flag application function never being entered. Calling SDL_RestoreWindow() didn't result in the minimized or maximized state being cleared on the hidden window either. Save the current window flags to the pending flags when hiding a window, and universally apply the pending flags when the window is shown again to ensure that all state that was set or cleared when the window was hidden is applied. Any pending flags that match the existing window flags will be automatically deduplicated by the corresponding functions. commit 49b5cfa Author: Ryan C. Gordon <[email protected]> Date: Wed Jun 7 15:42:22 2023 -0400 x11: check if window size/position has changed during SDL_ShowWindow. Fixes libsdl-org#4216. commit dc06116 Author: Sam Lantinga <[email protected]> Date: Tue Jun 6 12:33:48 2023 -0700 Steam sets the device version of the Steam Virtual Gamepad to 0, for the best compatibility with old games commit d032492 Author: Sam Lantinga <[email protected]> Date: Mon Jun 5 21:28:12 2023 -0700 Fixed detecting Bluetooth Steam Controllers on Windows commit 17c397f Author: Sam Lantinga <[email protected]> Date: Mon Jun 5 21:18:49 2023 -0700 Save the error code before doing any other operation SDL_ClearError() resets the Win32 error code to 0 commit 6e7769c Author: Sam Lantinga <[email protected]> Date: Mon Jun 5 18:02:37 2023 -0700 Fixed detecting Bluetooth Steam Controllers commit 6150b5b Author: Sam Lantinga <[email protected]> Date: Sun Jun 4 05:37:06 2023 -0700 Don't bother re-encoding Latin1 characters in the ASCII range commit 491ae20 Author: Sam Lantinga <[email protected]> Date: Sun Jun 4 02:06:52 2023 -0700 Only convert the result of XLookupString() if it's not already UTF-8 Fixes libsdl-org#7766 commit c369b90 Author: Sam Lantinga <[email protected]> Date: Sun Jun 4 01:01:06 2023 -0700 Fixed SDL_iconv_string() truncation when handling SDL_ICONV_E2BIG commit d406951 Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 23:24:21 2023 -0700 Cleaned up Alt-Enter/Ctrl-Enter fullscreen toggle logic Alt-Enter will go from the current state to fullscreen desktop, or if already there, will leave fullscreen mode. Ctrl-Enter will go from the current state to exclusive fullscreen mode, or if already there, will leave fullscreen mode. commit 052a9d3 Author: Rusty Moyher <[email protected]> Date: Fri Jun 2 11:45:21 2023 -0500 macOS: Removed the fullscreen blanking window - Removed the blanking window used in SDL_WINDOW_FULLSCREEN. This allows CMD + Tab to work. - Changed the fullscreen NSWindow level so this works properly. Fixes issue libsdl-org#7776 commit fa41ece Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 11:53:36 2023 -0700 Allow other applications to open controllers on macOS commit 6815e75 Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 11:39:03 2023 -0700 Steam uses a different VID/PID for the Steam Virtual Gamepad on Windows commit 5f00147 Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 11:35:27 2023 -0700 Revert "Fixed detection of the Steam Virtual Gamepad on macOS" This reverts commit 5fcd705. Steam has been updated to send a version of 1 to avoid conflicts with controllers that report a version of 0. commit 0c862d9 Author: Ozkan Sezer <[email protected]> Date: Sat Jun 3 17:23:40 2023 +0300 added compiler support comment about #pragma push_macro/pop_macro. commit 5fcd705 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 18:37:45 2023 -0700 Fixed detection of the Steam Virtual Gamepad on macOS commit 9837653 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 15:52:56 2023 -0700 Allow the application to send commands to Nintendo Switch controllers commit 7c55845 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 11:20:31 2023 -0700 Fixed build if SDL_JOYSTICK_RAWINPUT_MATCHING isn't enabled commit 767507f Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 11:16:07 2023 -0700 Don't use raw input while Remote Desktop is active Raw input will not send game controller events while Remote Desktop is active, so dynamically switch between XInput and raw input when Remote Desktop state changes. Fixes libsdl-org#7759 commit a5a1844 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 10:20:29 2023 -0700 Fixed build error commit e8b5b48 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 09:38:18 2023 -0700 Fixed WGI immediately being correlated with raw input devices with no input The at rest match state is 0x0000008800000000, not 0 commit 2e27812 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 08:51:47 2023 -0700 Fixed event sequence when Remote Desktop connects and disconnects 1. Send display disconnected events for all displays no longer connected 2. Send display connected events for all displays that have been connected 3. Send window display events for any windows that have changed displays commit 6aaf032 Author: Alibek Omarov <[email protected]> Date: Fri Jun 2 05:55:42 2023 +0300 wayland: reset orientation bitmask before reading values from hint on QtWayland Signed-off-by: Alibek Omarov <[email protected]> (cherry picked from commit 68e3e99) commit 5652037 Author: Alibek Omarov <[email protected]> Date: Fri Jun 2 05:59:06 2023 +0300 hints: clarify support for comma-separated values for QtWayland orientation that's available since 2.0.22 Signed-off-by: Alibek Omarov <[email protected]> (cherry picked from commit c395240) commit 8c476ca Author: Ozkan Sezer <[email protected]> Date: Fri Jun 2 14:56:40 2023 +0300 hidapi/libusb: need libusb >= 1.0.16 because of libusb_get_port_numbers commit e27c10a Author: Ozkan Sezer <[email protected]> Date: Fri Jun 2 14:55:24 2023 +0300 hidapi/libusb: don't try to dlsym libusb_wrap_sys_device not used in SDL. commit 3ecdbf2 Author: Ozkan Sezer <[email protected]> Date: Fri Jun 2 14:36:14 2023 +0300 hidapi: silence redefinition warnings commit f082c68 Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 10:48:15 2023 -0700 Updated with upstream suggestions in libusb/hidapi#582 commit 0ad089d Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 08:30:10 2023 -0700 Backed out local changes to build with libusb on Windows commit b36679b Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 08:27:56 2023 -0700 Updated with upstream suggestions in libusb/hidapi#582 commit 20182ee Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 07:47:28 2023 -0700 Updated with upstream suggestions in libusb/hidapi#572 commit 0cff44d Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 07:27:51 2023 -0700 Updated with upstream suggestions in libusb/hidapi#577 commit 4f58445 Author: Sam Lantinga <[email protected]> Date: Wed May 31 14:28:50 2023 -0700 Ignore both the mouse and keyboard endpoints of Steam Controllers when enumerating game controllers commit c886e80 Author: Nikita Krapivin <[email protected]> Date: Thu Jun 1 01:09:34 2023 +0500 gdk: Virtual keyboard and text input backend commit b6a88b0 Author: Sam Lantinga <[email protected]> Date: Wed May 31 10:49:46 2023 -0700 Don't try to interpret Xbox controllers as raw HID devices on Windows These generate fake reports which don't include separate trigger axes, and the HIDAPI joystick driver doesn't understand this report format. commit adad651 Author: Frank Praznik <[email protected]> Date: Mon May 15 11:37:43 2023 -0400 wayland: Add aspect-correct output for scaled modes Add aspect-correct output of scaled video modes and a hint to control this behavior (aspect, stretch, or none). The Wayland spec states that fullscreen surfaces that do not cover the entire output shall be centered with the borders masked by the compositor, so no additional work is required aside from calculating the proper window dimensions. The default is still 'stretch' mode, as some window managers as of this time (KDE and older versions of GNOME still found in LTS distros) don't behave according to the spec and present an unmasked window that is not centered, so it's not yet safe to change the default. commit 4644ac9 Author: Sam Lantinga <[email protected]> Date: Wed May 31 08:27:27 2023 -0700 Remove unneeded property type check commit cddcc4f Author: Wohlstand <[email protected]> Date: Wed May 31 03:54:38 2023 +0300 Don't ignore "build-scripts" commit 8e0b473 Author: Wohlstand <[email protected]> Date: Wed May 31 03:38:25 2023 +0300 Also ignore "build-*/" sub-directories Locally, I do make multiple build directories for various purposes, like "build-debug", "build-release", "build-novideo", etc. Also, Qt Creator (if configure it) may automatically create build directories that always starts with a "build-*" prefix. commit e9e4a91 Author: Ozkan Sezer <[email protected]> Date: Wed May 31 04:50:02 2023 +0300 testautomation_guid.c: fix integer warning in 32 bit builds. commit 0b28cbe Author: Sam Lantinga <[email protected]> Date: Tue May 30 16:36:13 2023 -0700 Allow building on really old Linux kernels (thanks @sezero!) commit 9b147eb Author: Ozkan Sezer <[email protected]> Date: Wed May 31 01:55:50 2023 +0300 SDL_x11window.c: apply missing SDL2-to-SDL3 portage to fix build. commit c521e12 Author: Ozkan Sezer <[email protected]> Date: Wed May 31 01:55:40 2023 +0300 SDL_dbus.h: define DBUS_TIMEOUT_USE_DEFAULT if not already defined. commit 20dfb7d Author: Ozkan Sezer <[email protected]> Date: Wed May 31 01:55:40 2023 +0300 hidapi/linux/hid.c: define BUS_SPI if not already defined. commit 19c10b4 Author: Ryan C. Gordon <[email protected]> Date: Tue May 30 17:04:31 2023 -0400 x11: Attempt to wait for SDL_MaximizeWindow to complete before returning. Fixes libsdl-org#7070. (cherry picked from commit 379a6f4) commit 1fdcb61 Author: Sam Lantinga <[email protected]> Date: Tue May 30 05:52:07 2023 -0700 Fixed build commit 4c36726 Author: Sam Lantinga <[email protected]> Date: Mon May 29 16:44:17 2023 -0700 macOS child window fixes - Fix places working with window coordinates that need to call SDL_RelativeToGlobalForWindow or SDL_GlobalToRelativeForWindow - Remove NSScreen param from ConvertNSRect(). Reflecting the Y coordinate is done relative to the main screen height (which ConvertNSRect was already doing) so the explicit screen isn't needed. - Refactor NSScreen lookups for point/rect and fix getting the screen for Cocoa_SetWindowPosition() to get the screen for the new position and not the window's current screen (which may not exist if the window is off-screen). - Fix re-associating the popup and parent window when the child window is shown. Hiding a child window removes it from the window hierarchy and so must be added when the window is shown again. - Allow popup windows that are not tooltips to gain key focus. commit ecccf6f Author: Ryan C. Gordon <[email protected]> Date: Mon May 29 14:46:58 2023 -0400 windows: Don't allow non-resizable windows to be maximized. Fixes libsdl-org#6346. (cherry picked from commit d275851) commit 52b73d4 Author: David Gow <[email protected]> Date: Mon May 29 15:36:59 2023 +0800 pipewire: Set 'application.id' if SDL_HINT_APP_ID set If SDL_HINT_APP_ID is set, pass it as the application.id to pipewire. This gives any pipewire-based tools a hint to find an associated .desktop file for icons, etc. commit f74549e Author: Sam Lantinga <[email protected]> Date: Sun May 28 19:23:56 2023 -0700 Fixed build warning (cherry picked from commit 5007b96) commit 1da5b2e Author: Sam Lantinga <[email protected]> Date: Sun May 28 18:41:21 2023 -0700 Disable SDL_JOYSTICK_RAWINPUT_XINPUT Apparently when using the Xbox One Wireless Adapter, using XInput at the same time as raw input will cause the controller to turn off immediately after connecting. This appears to be a bug in the Windows 11 driver stack, but since WGI provides all the extended functionality we need, this can be turned off for now. Fixes libsdl-org#3468 (cherry picked from commit b2e88ec) commit 95f2445 Author: Ryan C. Gordon <[email protected]> Date: Thu May 18 13:26:55 2023 -0400 power: On Linux, compare status strings as case-insensitive. In case something reports "Device" when we expected "device", etc. Reference Issue libsdl-org#6835. (cherry picked from commit df9d0fb) commit 2f75596 Author: Frank Praznik <[email protected]> Date: Thu May 25 20:37:49 2023 -0400 Consolidate the X11 WM_CLASS and Wayland app ID envvars Consolidate the X11_WMCLASS and WAYLAND_WMCLASS envvars into one SDL_HINT_APP_ID hint. This hint serves the same purpose on both windowing systems to allow desktop compositors to identify and group windows together, as well as associate applications with their desktop settings and icons. The common code for retrieving the value is now consolidated under core/unix/SDL_appid.c as it's common to *nix platforms, and the value is now retrieved at window creation time instead of being cached by the video driver at startup so that changes to the hint after video initialization and before window creation will be seen, as well as to accommodate cases where applications want to use different values for different windows.
commit c2350b23720e4e33aab926b37fd3229ec8e3d846 Author: Sylvain <[email protected]> Date: Thu Jun 22 18:22:54 2023 +0200 Revert "Squashed commit of the following:" This reverts commit 1e8eef0ec49d65c0edf89e444fc16d9a0f6dbbf3. commit 1e8eef0ec49d65c0edf89e444fc16d9a0f6dbbf3 Author: Sylvain <[email protected]> Date: Thu Jun 22 15:19:30 2023 +0200 Squashed commit of the following: commit 433911307237afaec03e2dcb976f4dd1efb62782 Author: Sylvain <[email protected]> Date: Thu Jun 22 10:48:12 2023 +0200 SDL_DBus_AppendDictWithKeyValue: allows several key/value for other PRs commit 8aee6908bbae5d57bc88d35b81c2d57a727dee95 Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 18:42:25 2023 -0700 Make sure we send update complete events for delayed guide buttons commit 4c9fb3e16902607d978a2c2f9ade777ad232b628 Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 13:59:53 2023 -0700 Added the events SDL_EVENT_JOYSTICK_UPDATE_COMPLETE and SDL_EVENT_GAMEPAD_UPDATE_COMPLETE This allows the application to tell when a joystick polling cycle is complete and can process state changes as a single atomic update. It is disabled by default, at least for now. commit 808d83dd67ca3cbedb76c846172f76735203db87 Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 13:49:09 2023 -0700 Allow specifying APP_PLATFORM and APP_ABI on the command line commit 214d5daa3c741d7672e53dd741a78b3dc61c4848 Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 12:14:50 2023 -0700 Removed 100 ms hitch when querying third party Nintendo Switch controllers that don't respond to request for info Also take advantage of the fact that we know whether the device is connected over Bluetooth now. commit 3694dabe7c16bb8e76f6f4820a1987ce16c4a3c9 Author: Sam Lantinga <[email protected]> Date: Wed Jun 21 10:28:45 2023 -0700 Use default sensor calibration if we can't read it from the Nintendo Switch controller Fixes https://github.com/libsdl-org/SDL/issues/7830 commit 6306ee9f421fac43891b22a417d7c5d5eb4e8b31 Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 09:33:59 2023 -0700 List the available haptic devices in testhaptic commit 8cf5dc9963211c077b3e274886753e12ebfd5107 Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 09:18:19 2023 -0700 Build on Android targeting the arm64-v8a architecture by default This speeds up iteration time and covers most customer devices commit 5f3213eb0e51c9c46e6c3b76d003e59046095c67 Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 08:40:30 2023 -0700 Added support for gamepad sensor fusion with the Razer Kishi commit 91198baed40d5709020c3001e9234f4580df696a Author: Simon McVittie <[email protected]> Date: Tue Jun 20 13:22:55 2023 +0100 ibus: Handle error when getting the D-Bus machine ID It is possible for retrieving the machine ID to fail, either because dbus was installed incorrectly (machine ID absent or corrupt), or in 32-bit builds, because stat() on the machine ID fails with EOVERFLOW if it has an out-of-range timestamp or inode number. dbus has historically treated this as a faulty installation, raising a warning which by default causes the process to crash. Unfortunately, dbus_get_local_machine_id() never had a way to report errors, so it has no alternative for that (bad) error handling. In dbus >= 1.12.0, we can use dbus_try_get_local_machine_id() to get the same information, but with the ability to cope gracefully with errors. ibus won't work in this situation, but that's better than crashing. Mitigates: https://github.com/ValveSoftware/steam-for-linux/issues/9605 Signed-off-by: Simon McVittie <[email protected]> commit 3ddbeab88fcf5bc1b09fb664403b1240d342a6ae Author: Sam Lantinga <[email protected]> Date: Tue Jun 20 01:17:44 2023 -0700 Moved Android sensor event handling to a separate thread This makes it so you can interact with sensors on multiple threads, as long as only one thread initializes and cleans up the sensor subsystem. This also has the benefit that sensor data is available as soon as possible. commit 329e1b8b6aba88fe7271a112fc90cf2bdc10f4ca Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 19:42:01 2023 -0700 Update the sensors before the joysticks so the gamepad code gets fresh sensor readings commit 1a9c04e9f1a276f85e08dcb0c0dc0908dc0cfb1a Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 19:41:20 2023 -0700 Don't try to update the gamepad fusion sensors manually, instead rely on the normal update flow commit 20ea35138f2f11dc8a0f35e949e372ae8b916534 Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 16:17:34 2023 -0700 Use a separate sensor watching function for gamepad events to avoid reliance on system sensor events and prevent a potential deadlock commit 70e43c150ec377d5aad2ca238752b1140abc03f0 Author: Sam Lantinga <[email protected]> Date: Mon Jun 19 08:36:30 2023 -0700 Added support for blitting SDL_PIXELFORMAT_INDEX1LSB Fixes https://github.com/libsdl-org/SDL/issues/7844 (cherry picked from commit a3d4fd71c3163b2b827f9e87930dec72c93b2fb2) commit c6ee9780df4286f66c38f3fa9732daa9afe0a8a3 Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:54 2023 -0700 Fixed DualSense controllers not being picked up by the HIDAPI driver The hidraw device may take additional time to get the correct permissions for us to open it. In my tests on Steam Deck hardware, this ranges between 5-8ms. commit 4e81b4e8dee50bbe7a510ca3f11069e2c6fc4c90 Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:43 2023 -0700 Added SDL_HINT_VIDEO_X11_SCALING_FACTOR to allow overriding the content scale on X11 (thanks Andres!) commit 210c135f744b757872eb0811988b53dca8e46462 Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:38 2023 -0700 Implement SDL_HINT_WINDOW_ACTIVATE_WHEN_RAISED for X11 To match the focus stealing prevention logic from windows/osx. I didn't implement SDL_HINT_WINDOW_ACTIVATE_WHEN_SHOWN as there is no standard mechanism for us to do so. In most cases marking a window as OverrideRedirect will cause to not acquire focus when mapped. And in steam we are doing that when appropriate. I still left a note in X11_ShowWindow() regarding this behaviour. commit 9351bf6dd10734c7dc43442537341bdb19801999 Author: Sam Lantinga <[email protected]> Date: Sun Jun 18 12:19:22 2023 -0700 Add handling for SDL_HINT_WINDOW_ACTIVATE_WHEN_RAISED and SDL_HINT_WINDOW_ACTIVATE_WHEN_SHOWN on macOS - Cocoa_ShowWindow will order a window that is not being activated below the key window (if one exists) then set visible - Cocoa_RaiseWindow calls -[NSWindow orderFront:] without changing the key window commit f168f9c81326ad374aade49d1dc46f245b20d07a Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 12:42:55 2023 -0700 Added support for the GameSir G4 Pro We can't read device info or IMU calibration from this controller, and it has no gyro or accelerometer, but is otherwise perfectly functional. commit b77064441149413543f190ee5067760e6acc1358 Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 09:46:46 2023 -0700 Fixed building on older kernel headers commit b98494a10154d4e6944279125a2194cac20d7c85 Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 09:39:50 2023 -0700 Fixed building on older kernel headers commit 9fe384b69636cf672cee09b185cae4762b5aecfe Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 09:28:27 2023 -0700 Fixed display orientation function names for SDL 3.0 convention commit d91e96e7f50b9c3cb0e1dacda39274b4af037a96 Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 08:59:52 2023 -0700 Use SDL_HINT_GAMECONTROLLER_SENSOR_FUSION as a list of controllers to enable sensor fusion There are too many wraparound style controllers out there to enumerate them all, so instead make this a user option in applications that support it. commit 610c31c7b7957e6f670cc94698f04861ff839d3f Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 08:28:37 2023 -0700 Generalized the sensor coordinate transform for wraparound gamepads commit 9eb5eab0adb032336ebaf5d980d5a246977e8272 Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 08:04:34 2023 -0700 Use the correct orientation transformation based on whether the device is naturally landscape or portrait commit e6d1ba2a17a7e157d5aa36957f16842ac17cba89 Author: Sam Lantinga <[email protected]> Date: Sat Jun 17 00:52:40 2023 -0700 Added the concept of display natural orientation Also renamed SDL_GetDisplayOrientation() SDL_GetDisplayCurrentOrientation() The natural orientation of the primary display is the frame of reference for accelerometer and gyro sensor readings. commit 8de6ce7e92907c5f509c6b33c5ee2a5f7b3c1c0c Author: Sam Lantinga <[email protected]> Date: Fri Jun 16 17:48:34 2023 -0700 Rotate the sensor axes to match gamepad orientation when using the device sensors for game controllers commit a9c86e518aa1b90fd5e8f1a6f3b1d23ce76d845b Author: Sam Lantinga <[email protected]> Date: Fri Jun 16 17:48:18 2023 -0700 Added the Razer Edge controller to the list of wraparound controllers commit c207cd3f568d17679a6a728a8a157811326c7fdd Author: Sam Lantinga <[email protected]> Date: Fri Jun 16 16:03:55 2023 -0700 Added the Razer Junglecat to the wraparound controller list commit e4f53e6b214cf476c9ad4b035ead97fb62da81c0 Author: Ozkan Sezer <[email protected]> Date: Sat Jun 17 01:01:10 2023 +0300 testevdev.c: comment out two unused data to fix build. commit 42e4639a5eab214370b1d6939ba09f65d553c9f7 Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 22:03:44 2023 -0700 For gamepads that don't have their own sensors, try to use the system sensors. This allows using the gyro and accelerometer in handheld devices in conjunction with built-in or wraparound controllers. commit d5845928222f24be91fc27f68e846c1b8bcd9a93 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 15:16:44 2023 +0100 linux: If the kernel specifically tells us the device type, trust it If a device is positively identified as an accelerometer, pointing stick or clickpad, then we don't need to second-guess it. In practice this does not change the result for any device in our test data, so add some artificial records that exercise this. Signed-off-by: Simon McVittie <[email protected]> commit 0d5aa70e62e477e99d024287070c70e3ee7b9da1 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 15:14:33 2023 +0100 linux: Pass evdev properties when guessing device type In newer kernels, devices that can be positively identified as a particular device type (for example accelerometers) get a property bit set. Plumb this information through into the function, but don't use it for anything just yet. Signed-off-by: Simon McVittie <[email protected]> commit a4ce721d7a3c9b6062d172192edb82b71836f0a2 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 15:15:31 2023 +0100 testevdev: Allow device properties to be fully populated The props array was too small for the highest property bits to be set, although in practice this didn't matter since only the lower-order bits have a meaning. Make it consistent with all the others. Signed-off-by: Simon McVittie <[email protected]> commit fa0ca3d41e5638949bf4d6fa44ee713fd1f62d38 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:53:29 2023 +0100 linux: Distinguish between "device with keys" and a full keyboard udev distinguishes between ID_INPUT_KEY, a device with any keyboard keys at all (including for example USB numeric keypads, Bluetooth headsets with volume control buttons, and some game controllers; and ID_INPUT_KEYBOARD, a reasonably fully-featured keyboard that you could use for general-purpose text entry. If we do the same here, then it's useful input to our heuristics for identifying devices: for example, a device with ID_INPUT_KEY could reasonably be a gamepad, but a device with ID_INPUT_KEYBOARD certainly isn't. Resolves: https://github.com/libsdl-org/SDL/issues/7827 Signed-off-by: Simon McVittie <[email protected]> commit 9b7a9ca6669326e61be3e98e0fbceb93d03ca48e Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:29:53 2023 +0100 testevdev: Add some more laptop built-in devices Signed-off-by: Simon McVittie <[email protected]> commit 00b6db68defb31fc3f2a0f1de6edad9e418fcad2 Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:29:41 2023 +0100 testevdev: Add some EVIOCGNAME and USB name strings to test data We don't currently use these for anything, but we might start using them as input to our heuristics as part of #7697. Signed-off-by: Simon McVittie <[email protected]> commit 3772d6cc9926f294f04d6a25be9a2c0f7a43b66a Author: Simon McVittie <[email protected]> Date: Thu Jun 15 19:46:35 2023 +0100 testevdev: Add raw HID report descriptors where available We don't currently use these in our device-classification heuristic, but it could be a useful input in future. Thanks to Sam Lantinga, Ben Fradella, kevenwyld and schlegp for providing some of these. Signed-off-by: Simon McVittie <[email protected]> commit 2b7556fea27e517338135e6cfa934222409d88ca Author: Simon McVittie <[email protected]> Date: Fri Jun 16 14:35:16 2023 +0100 testevdev: Correct typo in bus type for Xbox Series S|X via Bluetooth All Bluetooth devices are bus type 0x0005. Signed-off-by: Simon McVittie <[email protected]> commit c13e511844c870d8bae8816ca6c07c7cd8c12a0e Author: Simon McVittie <[email protected]> Date: Tue Jun 13 17:48:50 2023 +0100 testevdev: Try to correct Wii devices guessed from kernel source The comments here disagreed with the actual bytes. According to https://elixir.bootlin.com/linux/v6.3.7/source/drivers/hid/hid-wiimote-modules.c, the Balance Board reports BTN_A and ABS_HAT0X, HAT0Y, HAT1X and HAT1Y. This means the comments here were correct, but the .abs bits shown were in the wrong byte. Matching the Wii U Pro Controller against the same kernel source, it appears to be correct: it's the same representation as a PS3 gamepad, except that it lacks the Z and RZ axes for analogue triggers. Signed-off-by: Simon McVittie <[email protected]> commit ffdafcd880fd2c7680b2b2b87d5c7a6e4654ff17 Author: Simon McVittie <[email protected]> Date: Tue Jun 13 17:26:51 2023 +0100 testevdev: Verify most Wii devices against real hardware Some of the test data previously seen here was guessed from kernel source code, and not all of it was correct. The following devices have now been verified against `evemu-describe` output with Linux 6.3 (thanks to Jeremy Whiting for collecting this): - basic Wiimote - buttons - 3-axis accelerometer - infra-red sensor for Sensor Bar location (precise aim) - Motion Plus accessory (3-axis gyroscope) - Nunchuck accessory (joystick, 2 buttons, second 3-axis accelerometer) - Classic Controller accessory (a complete traditional gamepad) Signed-off-by: Simon McVittie <[email protected]> commit 739f78302ba96afe88b84592a909c6b6a37644c6 Author: SDL Wiki Bot <[email protected]> Date: Fri Jun 16 12:22:18 2023 +0000 Sync SDL3 wiki -> header commit 26df6899356afcbea628e85be79d3063ccdc4c71 Author: Anonymous Maarten <[email protected]> Date: Fri Jun 16 14:21:33 2023 +0200 docs: expand CMake documentation + add minimal CMake project for reporting issues commit 378e33bb2cae7e475f1cda8c75f71f6232455833 Author: Sylvain <[email protected]> Date: Fri Jun 16 10:05:58 2023 +0200 Android: potential ANR during onKeyDown/Up SDLActivity may call onNativeKeyDown, while application is quitting commit e72935a445448104d02d4ecca2d24fe06141b15b Author: Sam Lantinga <[email protected]> Date: Thu Jun 15 16:43:35 2023 -0700 Check for modff in addition to modf commit dab4f296b8adde71f982dde703c79a3d9be5be65 Author: Sam Lantinga <[email protected]> Date: Thu Jun 15 16:43:22 2023 -0700 Cleanup spacing commit 16b57d2ff109e095e0c4024070075ddd1d5f410b Author: Simon McVittie <[email protected]> Date: Thu Jun 15 19:25:23 2023 +0100 testevdev: Add details of another driving simulator controller Thanks to Ben Fradella. Signed-off-by: Simon McVittie <[email protected]> commit 2b007512065af0f64b1bfed291e26e771f70728d Author: Simon McVittie <[email protected]> Date: Thu Jun 15 15:50:03 2023 +0100 testevdev: Expand test data for X-Box One Elite 2 This slightly newer device than the one from #7814 is functionally equivalent when connected via USB. When connected via Bluetooth, it has a different button mapping. Thanks to Sam Lantinga. Signed-off-by: Simon McVittie <[email protected]> commit 62ed6f453f6226784c1a47baf6bc544aed15fa7f Author: Mathieu Eyraud <[email protected]> Date: Thu Jun 15 09:59:07 2023 +0200 Use SDL_strdup instead of strdup commit 73927b09480fd5d664f7077a70a8b6df318cdd2d Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 15:16:21 2023 -0700 Removed unused function commit 2e465ae31bd3a8ad2a4b239c4d6ff1e8c57efe65 Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 11:05:10 2023 -0700 Revert "Added SDL_nextafter() and SDL_nextafterf()" This reverts commit bc5d074818bfb08b7429d5e6d8ef69215acb408e. It's not clear that we need these yet, so I'm going to remove them for now. commit 0c16f4faf00eb573d051cac7d60ed8dd22e5596a Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 10:15:06 2023 -0700 Fixed third-party Nintendo Switch Pro controllers shutting down when we try to set the home LED. This fixes the PDP Afterglow Wireless Deluxe Controller. commit 23e007d3b7c8f8f5ccd99e58cd52bb5636fc8d04 Author: Sam Lantinga <[email protected]> Date: Wed Jun 14 00:02:32 2023 -0700 Fixed third party Nintendo Switch Pro Controller resetting when being sent an unsupported command to set the Home LED Tested with the PowerA Fusion Pro Wireless Controller in Bluetooth mode commit 8c95bd814bf1cf0ea1f12aa724938176a7dfd780 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 22:40:51 2023 -0700 Allow switching licensed Nintendo Switch Pro controllers into gyro input mode commit cdfc0c5a3314e4e0cd5152feddd8950c7eb797f1 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 22:20:58 2023 -0700 Added support for the PowerA Fusion Pro Wireless Controller in Bluetooth mode This controller shows up with a VID/PID of 0, but has full functionality over Bluetooth commit 0f4b15e16b7f07a46db6dc8e651f8c1849d658c5 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:53:10 2023 -0700 Removed more Linux Xbox mappings in favor of the automatic mapping This fixes the Xbox Series X share button on Linux 5.x kernels. commit 883b0f4071687f2e5c8329963f97504e9679394b Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:49:33 2023 -0700 Cleanup for previous change, fixing typos, etc. commit 9567989eb3ce9c858f0fe76806c5ccad69da89ba Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:41:52 2023 -0700 Use the automatic mapping instead of a hardcoded one for Xbox controllers on Linux This is much more robust and able to dynamically create a mapping for Xbox One S, Xbox Series X, and Xbox Elite 2 controllers. commit db1d4d3d76f5e21b2547463710b513fe0ebd7fad Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 16:40:30 2023 -0700 Added automatic mapping support for Xbox controllers on the 6.x Linux kernels This automatically adds support for the share button and paddles when present. commit 0f24956b0af20c8dc3bde9bc1c921f2d127297be Author: Sylvain <[email protected]> Date: Tue Jun 13 23:10:29 2023 +0200 testautomation_hints.c: free hint memory commit 2c3717881f3500ae73ece85bffb70dd868535dbf Author: Sylvain <[email protected]> Date: Tue Jun 13 23:00:00 2023 +0200 testautomation_events.c: initialize "timestamp" to solve "conditional jump or move depends on uninitialised value" commit 56ba7f2ff073b50f82d28f98667c2d14845750fd Author: Simon McVittie <[email protected]> Date: Tue Jun 13 16:53:56 2023 +0100 testevdev: Add details of X-Box One Elite 2 via USB Thanks to iacore for capturing these. Signed-off-by: Simon McVittie <[email protected]> commit 74484511eb55c88e91fdd256828d11598a699550 Author: Simon McVittie <[email protected]> Date: Tue Jun 13 16:47:56 2023 +0100 testevdev: Provide a pointer to more information about adding test-cases Signed-off-by: Simon McVittie <[email protected]> commit bc5d074818bfb08b7429d5e6d8ef69215acb408e Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 07:34:50 2023 -0700 Added SDL_nextafter() and SDL_nextafterf() commit b0677f476fa43f4a113b04a959228fd38f95d740 Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 10:06:36 2023 -0700 Added automatic mapping for Xbox Elite paddles using the xpadneo driver We can't actually tell yet whether a controller has paddles, so this code isn't effective, but I'll file an upstream issue and see if we can get that resolved. commit 071d1e29dd4f80605932f320aa2e05fda383944c Author: Sam Lantinga <[email protected]> Date: Tue Jun 13 10:05:19 2023 -0700 Fixed joystick vendor detection in Linux automatic gamepad mapping commit 5a62a4596e33d4a4488543c7ca5aaadd4df36760 Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 23:55:56 2023 -0700 Fixed building with the 16.1.4479499 Android toolchain (cherry picked from commit b9d1c483b90e2a1c23053692e06d15f9dbd5fe99) commit cf1dc66e2cfc7a65374c5fea681dd31c50363a2c Author: Simon McVittie <[email protected]> Date: Fri Jun 9 13:24:21 2023 +0100 linux: Improve gamepad mapping heuristic to accept Android conventions This heuristic for gamepads without a more specific mapping already tried two incompatible conventions for handling triggers: the Linux Gamepad Specification uses hat switch 2 for the triggers (for whatever reason), but the de facto standard set by the drivers for older Xbox and Playstation controllers represents each trigger as the Z-axis of the nearest analog stick. Android documentation encourages Bluetooth gamepad manufacturers to use a third incompatible convention where the left and right triggers are represented as the brake and gas pedals of a driving simulator controller. The Android convention also changes the representation of the right stick: instead of using X and Y rotation as a second pair of axes, Android uses Z position as a second horizontal axis, and Z rotation as a second vertical axis. Try to cope gracefully with all of these. This will hopefully resolve the issue described in #5406 (when using unpatched kernels). Signed-off-by: Simon McVittie <[email protected]> commit c4d49fadd4500cb522b2a98e5a42d024c566d8bf Author: Simon McVittie <[email protected]> Date: Fri Jun 9 13:06:30 2023 +0100 linux: Reduce magic numbers when mapping gamepad axes The bitfield `mapped` has two different sets of meanings, depending whether we're setting up the triggers or the d-pad. Represent them as symbolic constants rather than opaque integers. Signed-off-by: Simon McVittie <[email protected]> commit dec0dbff13d4091035209016eb2d0dd82c9aba58 Author: Sam Lantinga <[email protected]> Date: Mon Jun 12 16:57:41 2023 -0700 Fixed enumerating Steam Controllers on iOS commit d95dbe78bb7a5087ecfb8d301417b1108d868738 Author: Sam Lantinga <[email protected]> Date: Mon Jun 12 15:39:57 2023 -0700 Fixed n3ds build commit cdc40ee0546f6a761c369076a9e7a11fa0ea35b4 Author: Sam Lantinga <[email protected]> Date: Mon Jun 12 15:26:34 2023 -0700 Reduce the chance of destroying the joystick mutex while it's in use Fixes https://github.com/libsdl-org/SDL/issues/7811 commit 6ab846b6885a147013527d187c21eb8f686ca615 Author: Linus Probert <[email protected]> Date: Mon Jun 12 18:27:40 2023 +0200 clipboard: Fixes additional x11 clipboard bugs found in tests commit a2ba5e90524a7e4f5bcbfc2bf53c737fecc09b26 Author: Linus Probert <[email protected]> Date: Mon Jun 12 13:06:16 2023 +0200 clipboard: Fixes testautomation fails introduced by clipboard changes commit 4e0f94ea7d58d4a1aac624421d73110232df3efd Author: Sylvain <[email protected]> Date: Mon Jun 12 11:45:46 2023 +0200 Android: add timeout when waiting the SDL thread to finish C SDLmain() thread might have started (mSDLThread.start() called) while the SDL_Init() might not have been called yet, and so the previous QUIT event will be discarded by SDL_Init() and app is running, not exiting. This is reprocible by adding instrumentation code in the SDLActivity. And hopefully, this could fix an ANR, where SDLActivity is in WAITING state (in thread.join()): at java.lang.Thread.join (Thread.java:1519) at org.libsdl.app.SDLActivity.onDestroy (SDLActivity.java) while SDLThread seems to be running commit 125e7420ecd2b0d1847aef804f53e614fbc68253 Author: Ryan C. Gordon <[email protected]> Date: Sun Jun 11 12:43:47 2023 -0400 cocoa: Warp mouse to center of window before enabling relative mouse. This prevents the case where the mouse might be at the edge of the window when enabling relative mode, which confuses macOS, at it might believe the user is attempting to resize the window. Fixes #6994. (cherry picked from commit 2afb49ba9a2edc361a3b35d8e30be977a5bcc0c4) commit 4cfacd5cb7ab79cc09d718871810973a80f0c86c Author: Ozkan Sezer <[email protected]> Date: Sat Jun 10 20:50:02 2023 +0300 SDL_dynapi.c: minor cosmetics. commit 0103ec112610b67d2f8f603140e4b2c4e98e4d99 Author: Sam Lantinga <[email protected]> Date: Sat Jun 10 08:43:43 2023 -0700 Fixed accidental commit commit 7f864151464441fde7be300aaecab80887a99fa8 Author: SDL Wiki Bot <[email protected]> Date: Sat Jun 10 15:42:15 2023 +0000 Sync SDL3 wiki -> header commit 281018f169b8b7edbcaa48dab25627b359b9294a Author: Sam Lantinga <[email protected]> Date: Sat Jun 10 08:39:20 2023 -0700 Make it clear that you can't mix 2D rendering and the window surface API Also added functions to query and destroy the window surface so you can switch between modes if you want. See https://github.com/pygame-community/pygame-ce/issues/2190 for more details. commit 5490873daa12a57ff4b36090f25f04530323c6e4 Author: Sam Lantinga <[email protected]> Date: Fri Jun 9 18:41:21 2023 -0700 Fixed querying device info on the MOBAPAD M073 The query packet needs to contain valid rumble data in order to be accepted by the controller. Fixes https://github.com/libsdl-org/SDL/issues/7788 commit 2042e9c4e3cba6ffa9c34abac14828e31365f98b Author: Sam Lantinga <[email protected]> Date: Fri Jun 9 17:34:12 2023 -0700 Only update the serial number if it hasn't already been set This fixes the serial number for Nintendo Switch Pro, which is queried from the hardware in device initialization, and was later clobbered by the USB string which isn't correct. commit c8051b11e84abdda49d5755c911457b5105d8d93 Author: Sam Lantinga <[email protected]> Date: Fri Jun 9 14:34:59 2023 -0700 Fixed reading input from the Razer Atrox Arcade Stick using Windows Gaming Input commit aaccf3cd52882d3ac227541976cd90b75a2f15b5 Author: Sylvain <[email protected]> Date: Fri Jun 9 14:21:17 2023 +0200 X11: common function to wait for the WM before sending SDL Window events commit 8096f7269b0b0692bdb08cf0f41e411ae9203619 Author: Sylvain <[email protected]> Date: Fri Jun 9 09:41:54 2023 +0200 Android: add robustness. check that the native code is run for the first time. commit ae9d8acc3b405be05f1d790489ed261ba7cfc518 Author: Sylvain <[email protected]> Date: Fri Jun 9 09:27:39 2023 +0200 Android: when an EventFilter is binded to SDL_EVENT_DID_ENTER_FOREGROUND event and triggered, GL context is expected to be already restored. commit 4c0758a234658a0f48a56d2263c744aaa9cee898 Author: Sam Lantinga <[email protected]> Date: Thu Jun 8 12:39:32 2023 -0700 Fixed crash if display couldn't be found in SDL_UpdateFullscreenMode() commit d9c17e70552fb9a085de5bac0bc7675b4ea593e0 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 18:24:09 2023 +0100 testevdev: Add details of some more 8BitDo devices Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit a2ed083285e2997ed14b4607b31954d99d7245b2 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 17:30:30 2023 +0100 testevdev: Describe several equivalent devices https://github.com/ValveSoftware/steam-devices/pull/34 lists several more devices that are functionally equivalent to this one from the point of view of their evdev metadata. Thanks to apgrc. Signed-off-by: Simon McVittie <[email protected]> commit f4a53e78c8f027efe6db2d5edd84630f21c8c22e Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:13:38 2023 +0100 testevdev: Add test data for a DualSense (PS5) gamepad Also make details of PS4 gamepads (which are very similar from an evdev point of view) more specific. Thanks to Sam Lantinga and Jeremy Whiting for recording these. Signed-off-by: Simon McVittie <[email protected]> commit 2fb1df155138accdf77700ee7f5a76d4bd0895ac Author: Simon McVittie <[email protected]> Date: Thu Jun 8 16:00:27 2023 +0100 testevdev: Add test data for Nintendo Switch Joy-Cons via Bluetooth Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit 8a82e06075628525e6dbffbcb6a277e864d183e1 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:51:10 2023 +0100 testevdev: Add test data for Microsoft Xbox Series S|X Controller Like the Stadia controller, this is unusual because it represents the Share button as the Record key from a multimedia keyboard (as of Linux 6.2.11 with the xpad driver). Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit ff01b43459b8cd9cd24ce6f2a91d86e03f0f7bc9 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:48:46 2023 +0100 testevdev: Add Google Stadia controller This is a bit unusual because it has a small number of what would ordinarily be keyboard keys. Thanks to Jeremy Whiting for recording this. Signed-off-by: Simon McVittie <[email protected]> commit 9ad0b8b47f351fb30c6cc35bd47b1f09872836d4 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:47:50 2023 +0100 testevdev: Add test data for another Switch Pro Controller A newer evemu-describe transcript has this same controller with its buttons mapped differently, presumably a result of driver changes in the Linux kernel. Either way, we should recognise it as a gamepad. Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit 7b526d0cad2b1238427f44c40f3279e7196a602f Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:46:52 2023 +0100 testevdev: Finish incomplete data for Switch Pro Controller via USB This didn't include any buttons, which I assume was because I transcribed them incorrectly rather than reflecting reality. Confirmed against another Switch Pro Controller on a more recent kernel (thanks to Jeremy Whiting). Signed-off-by: Simon McVittie <[email protected]> commit 4c33ef94c6d64836321fa24269c45a2dff4106d4 Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:45:17 2023 +0100 testevdev: Note a functionally-equivalent device We don't need to re-test the heuristic with the same input data, but knowing that another device has equivalent evdev metadata is useful information to record. Thanks to Jeremy Whiting. Signed-off-by: Simon McVittie <[email protected]> commit 4c035dc1e38d1a47718bfea8e6921289da68296a Author: Simon McVittie <[email protected]> Date: Thu Jun 8 15:14:19 2023 +0100 testevdev: Add another laptop touchpad This was reported by Rémi Bernon as an example of older SDL's non-udev code path going wrong for touchpads when the invoking user happens to be in the input group, which I believe was fixed by fdd945f2. Signed-off-by: Simon McVittie <[email protected]> commit 32d015a6a0c2a4fcfc8aad74fa60a3b8c5f6f53b Author: Sam Lantinga <[email protected]> Date: Thu Jun 8 10:46:56 2023 -0700 Fixed PS3 controller initialization over Bluetooth Now that we have hidapi that knows whether the controller is connected via Bluetooth or USB, this is much easier to fix. commit e15da1985a23bdc6e93701e728c327ba664dc1b8 Author: Frank Praznik <[email protected]> Date: Wed Jun 7 15:04:49 2023 -0400 video: Apply cleared flag states when restoring hidden windows Hiding a window and then clearing the minimized, maximized, fullscreen, or grabbed flags wouldn't result in the new state being applied to the window when shown, as the pending flags would always be zero, resulting in the flag application function never being entered. Calling SDL_RestoreWindow() didn't result in the minimized or maximized state being cleared on the hidden window either. Save the current window flags to the pending flags when hiding a window, and universally apply the pending flags when the window is shown again to ensure that all state that was set or cleared when the window was hidden is applied. Any pending flags that match the existing window flags will be automatically deduplicated by the corresponding functions. commit 49b5cfa6c5cb9bb8cecce508e12fa7f8e0cc302f Author: Ryan C. Gordon <[email protected]> Date: Wed Jun 7 15:42:22 2023 -0400 x11: check if window size/position has changed during SDL_ShowWindow. Fixes #4216. commit dc06116c718692d94145656926445da76acd24fe Author: Sam Lantinga <[email protected]> Date: Tue Jun 6 12:33:48 2023 -0700 Steam sets the device version of the Steam Virtual Gamepad to 0, for the best compatibility with old games commit d032492aaba19c108dfd5e52866573bd02bf6148 Author: Sam Lantinga <[email protected]> Date: Mon Jun 5 21:28:12 2023 -0700 Fixed detecting Bluetooth Steam Controllers on Windows commit 17c397f7ece85e1948a2eca739dd99122f761ea2 Author: Sam Lantinga <[email protected]> Date: Mon Jun 5 21:18:49 2023 -0700 Save the error code before doing any other operation SDL_ClearError() resets the Win32 error code to 0 commit 6e7769cde1e29b65f2e89688fe0d990e755a7017 Author: Sam Lantinga <[email protected]> Date: Mon Jun 5 18:02:37 2023 -0700 Fixed detecting Bluetooth Steam Controllers commit 6150b5b3cbde0e592c4ffe822f66aa5f9c90c3d9 Author: Sam Lantinga <[email protected]> Date: Sun Jun 4 05:37:06 2023 -0700 Don't bother re-encoding Latin1 characters in the ASCII range commit 491ae20d963cff397b5980b31d142d576e74becb Author: Sam Lantinga <[email protected]> Date: Sun Jun 4 02:06:52 2023 -0700 Only convert the result of XLookupString() if it's not already UTF-8 Fixes https://github.com/libsdl-org/SDL/issues/7766 commit c369b900199f905c51edb372bd7780f67a544a8c Author: Sam Lantinga <[email protected]> Date: Sun Jun 4 01:01:06 2023 -0700 Fixed SDL_iconv_string() truncation when handling SDL_ICONV_E2BIG commit d40695115f8fa5d7c81ac5bc5e8b0188d040b3b5 Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 23:24:21 2023 -0700 Cleaned up Alt-Enter/Ctrl-Enter fullscreen toggle logic Alt-Enter will go from the current state to fullscreen desktop, or if already there, will leave fullscreen mode. Ctrl-Enter will go from the current state to exclusive fullscreen mode, or if already there, will leave fullscreen mode. commit 052a9d3284396e1db8fb26642067c344ef6e1467 Author: Rusty Moyher <[email protected]> Date: Fri Jun 2 11:45:21 2023 -0500 macOS: Removed the fullscreen blanking window - Removed the blanking window used in SDL_WINDOW_FULLSCREEN. This allows CMD + Tab to work. - Changed the fullscreen NSWindow level so this works properly. Fixes issue #7776 commit fa41ece65f50cbfe5043a5d49911e37e60ebad5e Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 11:53:36 2023 -0700 Allow other applications to open controllers on macOS commit 6815e75cafee8e7caaacf1f428b6539c8bbc7f7d Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 11:39:03 2023 -0700 Steam uses a different VID/PID for the Steam Virtual Gamepad on Windows commit 5f00147e61017b0cdae73d9c5752e0cf136b06bc Author: Sam Lantinga <[email protected]> Date: Sat Jun 3 11:35:27 2023 -0700 Revert "Fixed detection of the Steam Virtual Gamepad on macOS" This reverts commit 5fcd70578bdde3eaaa51aa4ab6c24ed08bd49db3. Steam has been updated to send a version of 1 to avoid conflicts with controllers that report a version of 0. commit 0c862d9a554bb7e3fb6092dd07f6e928bf814736 Author: Ozkan Sezer <[email protected]> Date: Sat Jun 3 17:23:40 2023 +0300 added compiler support comment about #pragma push_macro/pop_macro. commit 5fcd70578bdde3eaaa51aa4ab6c24ed08bd49db3 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 18:37:45 2023 -0700 Fixed detection of the Steam Virtual Gamepad on macOS commit 9837653b9d20d48b30caa66f93951b8604e7e9ea Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 15:52:56 2023 -0700 Allow the application to send commands to Nintendo Switch controllers commit 7c55845c804e2272df8bc111dea5d3e9d11670a0 Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 11:20:31 2023 -0700 Fixed build if SDL_JOYSTICK_RAWINPUT_MATCHING isn't enabled commit 767507fcf6f4607edb413acb53a6c029863701ed Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 11:16:07 2023 -0700 Don't use raw input while Remote Desktop is active Raw input will not send game controller events while Remote Desktop is active, so dynamically switch between XInput and raw input when Remote Desktop state changes. Fixes https://github.com/libsdl-org/SDL/issues/7759 commit a5a1844681d87950db50a39bc2b359d45bc7b08d Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 10:20:29 2023 -0700 Fixed build error commit e8b5b4881c8eb31bb7de5adef3479305a73fb0bf Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 09:38:18 2023 -0700 Fixed WGI immediately being correlated with raw input devices with no input The at rest match state is 0x0000008800000000, not 0 commit 2e27812d39d17831da0a0d8afa1add6c8795af5a Author: Sam Lantinga <[email protected]> Date: Fri Jun 2 08:51:47 2023 -0700 Fixed event sequence when Remote Desktop connects and disconnects 1. Send display disconnected events for all displays no longer connected 2. Send display connected events for all displays that have been connected 3. Send window display events for any windows that have changed displays commit 6aaf03232060a9a89faed1004c9afb0b31899d6a Author: Alibek Omarov <[email protected]> Date: Fri Jun 2 05:55:42 2023 +0300 wayland: reset orientation bitmask before reading values from hint on QtWayland Signed-off-by: Alibek Omarov <[email protected]> (cherry picked from commit 68e3e99087544c64abb534613643080312aced96) commit 56520372cad6915bc954dc33a4be67b07556307d Author: Alibek Omarov <[email protected]> Date: Fri Jun 2 05:59:06 2023 +0300 hints: clarify support for comma-separated values for QtWayland orientation that's available since 2.0.22 Signed-off-by: Alibek Omarov <[email protected]> (cherry picked from commit c39524042520d6962434228ab8af9ba7e0542308) commit 8c476ca1b0aafd090028d7ec32f417cb4bf05c62 Author: Ozkan Sezer <[email protected]> Date: Fri Jun 2 14:56:40 2023 +0300 hidapi/libusb: need libusb >= 1.0.16 because of libusb_get_port_numbers commit e27c10a5b1e79d1eac3a679d749a29d35ce7b109 Author: Ozkan Sezer <[email protected]> Date: Fri Jun 2 14:55:24 2023 +0300 hidapi/libusb: don't try to dlsym libusb_wrap_sys_device not used in SDL. commit 3ecdbf27cc2bef2501af8a85287590d3be1200a0 Author: Ozkan Sezer <[email protected]> Date: Fri Jun 2 14:36:14 2023 +0300 hidapi: silence redefinition warnings commit f082c68b2ff564a7c63edba3551fadfe3b02970a Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 10:48:15 2023 -0700 Updated with upstream suggestions in https://github.com/libusb/hidapi/pull/582 commit 0ad089de2150977e1d34cf2798e4cf2e97b981be Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 08:30:10 2023 -0700 Backed out local changes to build with libusb on Windows commit b36679b90e5df70a2b63b3eacbbd013f2afd0395 Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 08:27:56 2023 -0700 Updated with upstream suggestions in https://github.com/libusb/hidapi/pull/582 commit 20182eed7c14ca172c746d28ac08855894f43249 Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 07:47:28 2023 -0700 Updated with upstream suggestions in https://github.com/libusb/hidapi/pull/572 commit 0cff44dbd46e9ad19cf03856c4819f49d0f5e296 Author: Sam Lantinga <[email protected]> Date: Thu Jun 1 07:27:51 2023 -0700 Updated with upstream suggestions in https://github.com/libusb/hidapi/pull/577 commit 4f58445f0389ca1a2a7d4bb03248007484038df1 Author: Sam Lantinga <[email protected]> Date: Wed May 31 14:28:50 2023 -0700 Ignore both the mouse and keyboard endpoints of Steam Controllers when enumerating game controllers commit c886e8067507b3de47b3a7ad6a37aa8cac12e6f1 Author: Nikita Krapivin <[email protected]> Date: Thu Jun 1 01:09:34 2023 +0500 gdk: Virtual keyboard and text input backend commit b6a88b0339537d842fc59b2095d4eda80934b214 Author: Sam Lantinga <[email protected]> Date: Wed May 31 10:49:46 2023 -0700 Don't try to interpret Xbox controllers as raw HID devices on Windows These generate fake reports which don't include separate trigger axes, and the HIDAPI joystick driver doesn't understand this report format. commit adad6514d4f42998b8ae16aa0ca1ea8031568743 Author: Frank Praznik <[email protected]> Date: Mon May 15 11:37:43 2023 -0400 wayland: Add aspect-correct output for scaled modes Add aspect-correct output of scaled video modes and a hint to control this behavior (aspect, stretch, or none). The Wayland spec states that fullscreen surfaces that do not cover the entire output shall be centered with the borders masked by the compositor, so no additional work is required aside from calculating the proper window dimensions. The default is still 'stretch' mode, as some window managers as of this time (KDE and older versions of GNOME still found in LTS distros) don't behave according to the spec and present an unmasked window that is not centered, so it's not yet safe to change the default. commit 4644ac99fa32bc10b27784c5f9530f2cf0aadab0 Author: Sam Lantinga <[email protected]> Date: Wed May 31 08:27:27 2023 -0700 Remove unneeded property type check commit cddcc4f406cd77acfa95a95870886bebc29682ad Author: Wohlstand <[email protected]> Date: Wed May 31 03:54:38 2023 +0300 Don't ignore "build-scripts" commit 8e0b47331f33ec378a3375d50e230d1ec34b27b2 Author: Wohlstand <[email protected]> Date: Wed May 31 03:38:25 2023 +0300 Also ignore "build-*/" sub-directories Locally, I do make multiple build directories for various purposes, like "build-debug", "build-release", "build-novideo", etc. Also, Qt Creator (if configure it) may automatically create build directories that always starts with a "build-*" prefix. commit e9e4a9175f9acbff039d714feb1929384ba87145 Author: Ozkan Sezer <[email protected]> Date: Wed May 31 04:50:02 2023 +0300 testautomation_guid.c: fix integer warning in 32 bit builds. commit 0b28cbe385c808b3cbe89b8eb7c2b10d7849599b Author: Sam Lantinga <[email protected]> Date: Tue May 30 16:36:13 2023 -0700 Allow building on really old Linux kernels (thanks @sezero!) commit 9b147eb80b29ee42d68c87041900f988ba068d65 Author: Ozkan Sezer <[email protected]> Date: Wed May 31 01:55:50 2023 +0300 SDL_x11window.c: apply missing SDL2-to-SDL3 portage to fix build. commit c521e1286be9284d3652569fab02718d9faa4d06 Author: Ozkan Sezer <[email protected]> Date: Wed May 31 01:55:40 2023 +0300 SDL_dbus.h: define DBUS_TIMEOUT_USE_DEFAULT if not already defined. commit 20dfb7df2c7b658ddacf6a835f9a9c6bd6f9c07a Author: Ozkan Sezer <[email protected]> Date: Wed May 31 01:55:40 2023 +0300 hidapi/linux/hid.c: define BUS_SPI if not already defined. commit 19c10b41fd0435e259c1da675733cc446541f9fa Author: Ryan C. Gordon <[email protected]> Date: Tue May 30 17:04:31 2023 -0400 x11: Attempt to wait for SDL_MaximizeWindow to complete before returning. Fixes #7070. (cherry picked from commit 379a6f4dabc448d37a5823724d507967e2e069eb) commit 1fdcb610ff4d26bc95a7425a628a1c0ec14f7e32 Author: Sam Lantinga <[email protected]> Date: Tue May 30 05:52:07 2023 -0700 Fixed build commit 4c36726a319f86aaa1c16c865705f4ec54bcfcd8 Author: Sam Lantinga <[email protected]> Date: Mon May 29 16:44:17 2023 -0700 macOS child window fixes - Fix places working with window coordinates that need to call SDL_RelativeToGlobalForWindow or SDL_GlobalToRelativeForWindow - Remove NSScreen param from ConvertNSRect(). Reflecting the Y coordinate is done relative to the main screen height (which ConvertNSRect was already doing) so the explicit screen isn't needed. - Refactor NSScreen lookups for point/rect and fix getting the screen for Cocoa_SetWindowPosition() to get the screen for the new position and not the window's current screen (which may not exist if the window is off-screen). - Fix re-associating the popup and parent window when the child window is shown. Hiding a child window removes it from the window hierarchy and so must be added when the window is shown again. - Allow popup windows that are not tooltips to gain key focus. commit ecccf6f9caedb9c8e8d50ec2d6c26b188f8466c1 Author: Ryan C. Gordon <[email protected]> Date: Mon May 29 14:46:58 2023 -0400 windows: Don't allow non-resizable windows to be maximized. Fixes #6346. (cherry picked from commit d275851dfbd586910ddf1fbed8ef54714afbe6c4) commit 52b73d411524567ba0a8b522fbc0a5c41a45720c Author: David Gow <[email protected]> Date: Mon May 29 15:36:59 2023 +0800 pipewire: Set 'application.id' if SDL_HINT_APP_ID set If SDL_HINT_APP_ID is set, pass it as the application.id to pipewire. This gives any pipewire-based tools a hint to find an associated .desktop file for icons, etc. commit f74549ef2ea6e8067261f0e3dceb1f2bdfdce6c7 Author: Sam Lantinga <[email protected]> Date: Sun May 28 19:23:56 2023 -0700 Fixed build warning (cherry picked from commit 5007b963639bff2fa3359abbfe20d53ea9111fd9) commit 1da5b2e488cc53534de939ae3c56e5cca7230f73 Author: Sam Lantinga <[email protected]> Date: Sun May 28 18:41:21 2023 -0700 Disable SDL_JOYSTICK_RAWINPUT_XINPUT Apparently when using the Xbox One Wireless Adapter, using XInput at the same time as raw input will cause the controller to turn off immediately after connecting. This appears to be a bug in the Windows 11 driver stack, but since WGI provides all the extended functionality we need, this can be turned off for now. Fixes https://github.com/libsdl-org/SDL/issues/3468 (cherry picked from commit b2e88ecfeb5e4d7db021e43c1b9bc4c9d14f615c) commit 95f244598bb17455dc7651b0f588965e5c351086 Author: Ryan C. Gordon <[email protected]> Date: Thu May 18 13:26:55 2023 -0400 power: On Linux, compare status strings as case-insensitive. In case something reports "Device" when we expected "device", etc. Reference Issue #6835. (cherry picked from commit df9d0fb332ea65c3fc47f72574851c91da2c912b) commit 2f75596d5a531f731f2128a8b08ba8c036582382 Author: Frank Praznik <[email protected]> Date: Thu May 25 20:37:49 2023 -0400 Consolidate the X11 WM_CLASS and Wayland app ID envvars Consolidate the X11_WMCLASS and WAYLAND_WMCLASS envvars into one SDL_HINT_APP_ID hint. This hint serves the same purpose on both windowing systems to allow desktop compositors to identify and group windows together, as well as associate applications with their desktop settings and icons. The common code for retrieving the value is now consolidated under core/unix/SDL_appid.c as it's common to *nix platforms, and the value is now retrieved at window creation time instead of being cached by the video driver at startup so that changes to the hint after video initialization and before window creation will be seen, as well as to accommodate cases where applications want to use different values for different windows. commit 2f4076aba3521c2969bb0d0106b03fa62cc8e61f Author: Sylvain <[email protected]> Date: Sun May 28 22:24:33 2023 +0200 Add SDL Video Capture, with back-end for linux/macos/ios/android
This isn't a full Windows or macOS port, but abstracts the libusb threading code so that it can be easily ported to other platforms.
Fixes #581