The main exits are initiated from code that cannot filter this back to D_DoomMain easily so the exception is the only way to get there.
The 3 main points of exit are:
* quit/exit CCMD
* quitting the menu through ST_Endoom
* receiving a quit message on the main window.
# Conflicts:
# src/posix/cocoa/st_start.mm
These ones were particularly bad examples of misusing the exit handlers by temporarily installing one themselves and then calling exit to clean stuff up.
Now they just return an error code to D_DoomMain to perform a regular exit.
The main catch is now in D_DoomMain, only calling platform specific functions to handle the output for the error.
As a nice side effect, -norun can now be done without an exception, just by exiting D_DoomMain with a special exit code.
# Conflicts:
# src/win32/i_main.cpp
Now there is only one single entry point for both, instead of previously 2 entry and 4 exit points.
This also eliminates the explicit shutdown of ZMusic. Timidity++'s two buffers have been put in containers that self-destruct on shutdown and calling dumb_exit is not necessary because the only feature requiring it is not used by any code in the music library.
Instead of trying a homegrown way to avoid recursive exceptions, let's do it with the defined procedure C++ has for this case: call std::terminate.
This allowed removing some old hackery inherited from Boom and will now hopefully allow sanitizing the exit procedure to the point that it can be done without depending on exit handlers.
# Conflicts:
# src/files_decompress.cpp
# Conflicts:
# src/d_main.cpp
The only difference here were the size values on Windows but for this initial call they have been useless for a long time. When this code was written the console buffer still had a fixed width that needed to be set before adding any text.
* - add cvar 'cl_disableinvertedcolormap' - changes the invulnerability sphere to instead be a regular desaturated colormap that transitions from deep blue to pale yellow
* - add menu option for cl_disableinvertedcolormap
* - added customization for invulnerability colormap
* - fixed custom colormap being calculated incorrectly
* - disable custom invulnerability map before the main game loop
# Conflicts:
# src/d_main.cpp
When exception is thrown from JITed code, VM stask isn't cleared during unwinding
It needs to be clear explicitly to avoid memory leaks and references to destructed objects on shutdown
https://forum.zdoom.org/viewtopic.php?t=66172
# Conflicts:
# src/dthinker.cpp
It's not advisable to rely on atexit() to invoke SDL_Quit(), as it
causes a segmentation fault on exit for the SDL2 KMSDRM driver due to
a conflict with the Mesa drivers* which upstream Mesa may not elect to fix.
The issue can be resolved by replacing the atexit() call with a signal
handler that will interrupt the D_DoomMain() and D_DoomLoop()
functions, then invoke call_terms() within the main thread before exiting.
We can leave the atexit() hook intact to handle edge cases such as abnormal
process exit, or exit methods which do not produce signals such as Alt+F4 or
window close via GUI (neither of which are possible in a KMS context, so
should not affect KMSDRM sessions).
Fixes a segmentation fault/uninterruptible application hang on exit
for all KMS targets, including Raspberry Pi 4B, 3B and Intel i965.
* See: https://bugzilla.libsdl.org/show_bug.cgi?id=4530 and
https://lists.freedesktop.org/archives/mesa-users/2019-March/001519.html
This was done for two reasons:
1. It forces loading of zd_extra.pk3 for the stock IWADs, eliminating a certain kind of user error.
2. It removes the stock IWAD definitions for custom games that distribute the engine along with their data, so that it doesn't pick up on Steam and GOG installations and shows an inappropriate IWAD picker.
With localization for non-Latin languages on the support list the multibyte API doesn't cut it anymore. It neither can handle system text output outside the local code page nor can an ANSI window receive text input outside its own code page.
Similar problems exist for file names. With the multibyte API it is impossible to handle any file containing characters outside the active local code page.
So as of now, everything that may pass along some Unicode text will use the Unicode API with some text conversion functions. The only places where calls to the multibyte API were left are those where known string literals are passed or where the information is not used for anything but comparing it to other return values from the same API.
# Conflicts:
# src/rendering/hwrenderer/postprocessing/hw_postprocess.h
# src/win32/base_sysfb.cpp
# src/win32/i_main.cpp
# src/win32/win32basevideo.cpp
# src/win32/win32glvideo.cpp
# Conflicts:
# src/version.h
# src/win32/i_main.cpp
# src/win32/i_system.cpp
# src/win32/optwin32.h
# src/win32/win32gliface.cpp
# wadsrc/static/language.enu
There are two options here - one only disables the vertical thrust and the other goes back fully to the original non-z-aware code.
Both options are settable through MAPINFO.
For the compatibility presets, the normal ones only disable the vertical thrust, the strict ones force use of the old code entirely.
# Conflicts:
# wadsrc/static/language.csv
It now reads everything into a two-dimensional TMap and creates a list of mappings that apply to the current setting.
The constant need for reloading was the main blocker in redesigning how Dehacked strings get inserted. Currently they override everything, but IWAD-based Dehacked text shouldn't block PWAD overrides from PWADs' LANGUAGE lumps and instead be treated as coming from an [en default] block.
This also renames the main block from [enu default] to [en default], because it should be treated as the English default for all English locales and not just make it fall through to the base default as it did before.
# Conflicts:
# src/doomstat.cpp
# Conflicts:
# src/stringtable.cpp
This was not applicable: Warning C4150 deletion of pointer to incomplete type 'FLightDefaults'; no destructor called (compiling source file C:\DEV\lzdoom\src\d_main.cpp)
Doing this intermingled with the thinkers is highly unsafe because there are absolutely no guarantees about order of execution.
Effectively it ran these commands right in the middle of the playsim which could cause all sorts of synchronization issues, because CCMDs are part of the UI, not the playsim.
- pass a const string to AddCommandString.
This function manipulated the input buffer, leading to all sorts of code contortions to make sure that the passed parameter is clean for that.
This function will now create a copy of the passed parameter which it can manipulate without complicating its calling code.
# Conflicts:
# src/c_dispatch.cpp
The mod which prompted me to add this is "The Chosen" which is a Dehacked-based TC and repurposes many original actors for something entirely different.
The stock lights are not usable for this and would make it impossible to add a GAMEINFO lump to it because then there is no way to disable loading of lights in the startup screen.
- moved the ALTHUDCF parser PClass::StaticInit, so that it gets done right after creating the actor definitions.
All left to do is not to reallocate the AltHud object for each frame but store it in a better suited place.
# Conflicts:
# src/d_main.cpp
# Conflicts:
# src/info.cpp
- split gl_postprocessshader.h in two so that the hardware independent part can be used by GLDEFS without pulling in all of OpenGL.
# Conflicts:
# src/CMakeLists.txt
# src/gl/dynlights/gl_glow.cpp
# src/gl/renderer/gl_postprocess.cpp
# src/gl/textures/gl_texture.cpp
It needed more than using the animated accessor. The code here nearly went out of its way to circumvent the texture manager's built in logic.
(cherry picked from commit 79b3c41677)
Remaining object(s) led to a potential crash on the next garbage collection cycle
Assertion failure was triggered during restarting in Debug configuration
(cherry picked from commit e085a8d58a)
This was done mainly to reduce the amount of occurences of the word FTexture but it immediately helped detect two small and mostly harmless bugs that were found due to the stricter type checks.
