asgardius/virtualx-engine
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
virtualx-engine/servers/server_wrap_mt_common.h

815 lines
64 KiB
C++
Raw Normal View History

One Copyright Update to rule them all As many open source projects have started doing it, we're removing the current year from the copyright notice, so that we don't need to bump it every year. It seems like only the first year of publication is technically relevant for copyright notices, and even that seems to be something that many companies stopped listing altogether (in a version controlled codebase, the commits are a much better source of date of publication than a hardcoded copyright statement). We also now list Godot Engine contributors first as we're collectively the current maintainers of the project, and we clarify that the "exclusive" copyright of the co-founders covers the timespan before opensourcing (their further contributions are included as part of Godot Engine contributors). Also fixed "cf." Frenchism - it's meant as "refer to / see".
2023-01-05 13:25:55 +01:00
/**************************************************************************/
/* server_wrap_mt_common.h */
/**************************************************************************/
/* This file is part of: */
/* GODOT ENGINE */
/* https://godotengine.org */
/**************************************************************************/
/* Copyright (c) 2014-present Godot Engine contributors (see AUTHORS.md). */
/* Copyright (c) 2007-2014 Juan Linietsky, Ariel Manzur. */
/* */
/* Permission is hereby granted, free of charge, to any person obtaining */
/* a copy of this software and associated documentation files (the */
/* "Software"), to deal in the Software without restriction, including */
/* without limitation the rights to use, copy, modify, merge, publish, */
/* distribute, sublicense, and/or sell copies of the Software, and to */
/* permit persons to whom the Software is furnished to do so, subject to */
/* the following conditions: */
/* */
/* The above copyright notice and this permission notice shall be */
/* included in all copies or substantial portions of the Software. */
/* */
/* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, */
/* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF */
/* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. */
/* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY */
/* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, */
/* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE */
/* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. */
/**************************************************************************/
ability to run 2D physics in a thread also, 2D physics is now thread safe too. see physics_2d/thread_model
2015-05-26 06:05:08 +02:00
Code quality: Fix header guards consistency Adds `header_guards.sh` bash script, used in CI to validate future changes. Can be run locally to fix invalid header guards.
2022-07-23 23:41:51 +02:00
#ifndef SERVER_WRAP_MT_COMMON_H
#define SERVER_WRAP_MT_COMMON_H
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
#ifdef DEBIG_ENABLED
#define MAIN_THREAD_SYNC_CHECK \
if (unlikely(Thread::is_main_thread() && Engine::get_singleton()->notify_frame_server_synced())) { \
MAIN_THREAD_SYNC_WARN \
}
#else
#define MAIN_THREAD_SYNC_CHECK
#endif
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
#define FUNC0R(m_r, m_type) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type() override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(); \
} \
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNCRIDSPLIT(m_type) \
virtual RID m_type##_create() override { \
RID ret = server_name->m_type##_allocate(); \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type##_initialize, ret); \
} else { \
server_name->m_type##_initialize(ret); \
} \
return ret; \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
//RID now returns directly, ensure thread safety yourself
#define FUNCRID(m_type) \
virtual RID m_type##_create() override { \
return server_name->m_type##_create(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
#define FUNC0RC(m_r, m_type) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type() const override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(); \
} \
}
#define FUNC0(m_type) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type() override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(); \
} \
}
#define FUNC0C(m_type) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type() const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(); \
} \
}
#define FUNC0S(m_type) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type() override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(); \
} \
}
#define FUNC0SC(m_type) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type() const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(); \
} \
ability to run 2D physics in a thread also, 2D physics is now thread safe too. see physics_2d/thread_model
2015-05-26 06:05:08 +02:00
}
///////////////////////////////////////////////
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
#define FUNC1R(m_r, m_type, m_arg1) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1); \
} \
}
#define FUNC1RC(m_r, m_type, m_arg1) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1); \
} \
}
#define FUNC1S(m_type, m_arg1) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1); \
} \
}
#define FUNC1SC(m_type, m_arg1) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1); \
} \
}
#define FUNC1(m_type, m_arg1) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type, p1); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1); \
} \
}
#define FUNC1C(m_type, m_arg1) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type, p1); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1); \
} \
}
#define FUNC2R(m_r, m_type, m_arg1, m_arg2) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1, m_arg2 p2) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2); \
} \
}
#define FUNC2RC(m_r, m_type, m_arg1, m_arg2) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1, m_arg2 p2) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2); \
} \
}
#define FUNC2S(m_type, m_arg1, m_arg2) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2); \
} \
}
#define FUNC2SC(m_type, m_arg1, m_arg2) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2); \
} \
}
#define FUNC2(m_type, m_arg1, m_arg2) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type, p1, p2); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2); \
} \
}
#define FUNC2C(m_type, m_arg1, m_arg2) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type, p1, p2); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2); \
} \
}
#define FUNC3R(m_r, m_type, m_arg1, m_arg2, m_arg3) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2, p3); \
} \
}
#define FUNC3RC(m_r, m_type, m_arg1, m_arg2, m_arg3) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, &ret); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2, p3); \
} \
}
#define FUNC3S(m_type, m_arg1, m_arg2, m_arg3) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3); \
} \
}
#define FUNC3SC(m_type, m_arg1, m_arg2, m_arg3) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3); \
} \
}
#define FUNC3(m_type, m_arg1, m_arg2, m_arg3) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3); \
} \
}
#define FUNC3C(m_type, m_arg1, m_arg2, m_arg3) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3); \
} \
}
#define FUNC4R(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2, p3, p4); \
} \
}
#define FUNC4RC(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2, p3, p4); \
} \
}
#define FUNC4S(m_type, m_arg1, m_arg2, m_arg3, m_arg4) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3, p4); \
} \
}
#define FUNC4SC(m_type, m_arg1, m_arg2, m_arg3, m_arg4) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3, p4); \
} \
}
#define FUNC4(m_type, m_arg1, m_arg2, m_arg3, m_arg4) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4); \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3, p4); \
} \
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC4C(m_type, m_arg1, m_arg2, m_arg3, m_arg4) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4); \
} \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
#define FUNC5R(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5) \
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5) { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2, p3, p4, p5); \
} \
}
#define FUNC5RC(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5) const override { \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2, p3, p4, p5); \
} \
}
#define FUNC5S(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5) \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5) override { \
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5); \
CommandQueueMT doesn't lock during command execution
2017-10-13 19:13:52 +02:00
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
server_name->m_type(p1, p2, p3, p4, p5); \
} \
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC5SC(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5); \
} \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC5(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5); \
} \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC5C(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5); \
} \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
#define FUNC6R(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6) \
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6) { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
WRITE_ACTION \
Makes all Godot API's methods Lower Case
2017-08-07 12:17:31 +02:00
if (Thread::get_caller_id() != server_thread) { \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return ret; \
} else { \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
command_queue.flush_if_pending(); \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
return server_name->m_type(p1, p2, p3, p4, p5, p6); \
} \
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC6RC(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6) \
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6) const override { \
if (Thread::get_caller_id() != server_thread) { \
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
return ret; \
} else { \
command_queue.flush_if_pending(); \
return server_name->m_type(p1, p2, p3, p4, p5, p6); \
} \
}
#define FUNC6S(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6); \
} \
}
#define FUNC6SC(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6); \
} \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC6(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6); \
} \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC6C(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6); \
} \
A Whole New World (clang-format edition) I can show you the code Pretty, with proper whitespace Tell me, coder, now when did You last write readable code? I can open your eyes Make you see your bad indent Force you to respect the style The core devs agreed upon A whole new world A new fantastic code format A de facto standard With some sugar Enforced with clang-format A whole new world A dazzling style we all dreamed of And when we read it through It's crystal clear That now we're in a whole new world of code
2017-03-05 16:44:50 +01:00
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC7R(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7) \
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
return ret; \
} else { \
command_queue.flush_if_pending(); \
return server_name->m_type(p1, p2, p3, p4, p5, p6, p7); \
} \
ability to run 2D physics in a thread also, 2D physics is now thread safe too. see physics_2d/thread_model
2015-05-26 06:05:08 +02:00
}
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
#define FUNC7RC(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7) \
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7) const override { \
if (Thread::get_caller_id() != server_thread) { \
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
return ret; \
} else { \
command_queue.flush_if_pending(); \
return server_name->m_type(p1, p2, p3, p4, p5, p6, p7); \
} \
}
#define FUNC7S(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7); \
} \
}
#define FUNC7SC(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7); \
} \
}
#define FUNC7(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7); \
} \
}
#define FUNC7C(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7); \
} \
}
#define FUNC8R(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8) \
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
return ret; \
} else { \
command_queue.flush_if_pending(); \
return server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8); \
} \
}
#define FUNC8RC(m_r, m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8) \
virtual m_r m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8) const override { \
if (Thread::get_caller_id() != server_thread) { \
m_r ret; \
command_queue.push_and_ret(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, &ret); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
return ret; \
} else { \
command_queue.flush_if_pending(); \
return server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8); \
} \
}
#define FUNC8S(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8); \
} \
}
#define FUNC8SC(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push_and_sync(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8); \
SYNC_DEBUG \
Use WorkerThreadPool for Server threads * Servers now use WorkerThreadPool for background computation. * This helps keep the number of threads used fixed at all times. * It also ensures everything works on HTML5 with threads. * And makes it easier to support disabling threads for also HTML5. CommandQueueMT now syncs with the servers via the WorkerThreadPool yielding mechanism, which makes its classic main sync semaphore superfluous. Also, some warnings about calls that kill performance when using threaded rendering are removed because there's a mechanism that warns about that in a more general fashion. Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-04 10:18:08 +02:00
MAIN_THREAD_SYNC_CHECK \
Make Servers truly Thread Safe -Rendering server now uses a split RID allocate/initialize internally, this allows generating RIDs immediately but initialization to happen later on the proper thread (as rendering APIs generally requiere to call on the right thread). -RenderingServerWrapMT is no more, multithreading is done in RenderingServerDefault. -Some functions like texture or mesh creation, when renderer supports it, can register and return immediately (so no waiting for server API to flush, and saving staging and command buffer memory). -3D physics server changed to be made multithread friendly. -Added PhysicsServer3DWrapMT to use 3D physics server from multiple threads. -Disablet Bullet (too much effort to make multithread friendly, this needs to be fixed eventually).
2021-02-09 17:19:03 +01:00
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8); \
} \
}
#define FUNC8(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8); \
} \
}
#define FUNC8C(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8) const override { \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8); \
} \
}
#define FUNC9(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8, m_arg9) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8, m_arg9 p9) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, p9); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8, p9); \
} \
}
#define FUNC10(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8, m_arg9, m_arg10) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8, m_arg9 p9, m_arg10 p10) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8, p9, p10); \
} \
}
#define FUNC11(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8, m_arg9, m_arg10, m_arg11) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8, m_arg9 p9, m_arg10 p10, m_arg11 p11) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11); \
} \
}
#define FUNC12(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8, m_arg9, m_arg10, m_arg11, m_arg12) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8, m_arg9 p9, m_arg10 p10, m_arg11 p11, m_arg12 p12) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12); \
} \
}
#define FUNC13(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8, m_arg9, m_arg10, m_arg11, m_arg12, m_arg13) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8, m_arg9 p9, m_arg10 p10, m_arg11 p11, m_arg12 p12, m_arg13 p13) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13); \
} \
}
#define FUNC14(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8, m_arg9, m_arg10, m_arg11, m_arg12, m_arg13, m_arg14) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8, m_arg9 p9, m_arg10 p10, m_arg11 p11, m_arg12 p12, m_arg13 p13, m_arg14 p14) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13, p14); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13, p14); \
} \
}
#define FUNC15(m_type, m_arg1, m_arg2, m_arg3, m_arg4, m_arg5, m_arg6, m_arg7, m_arg8, m_arg9, m_arg10, m_arg11, m_arg12, m_arg13, m_arg14, m_arg15) \
virtual void m_type(m_arg1 p1, m_arg2 p2, m_arg3 p3, m_arg4 p4, m_arg5 p5, m_arg6 p6, m_arg7 p7, m_arg8 p8, m_arg9 p9, m_arg10 p10, m_arg11 p11, m_arg12 p12, m_arg13 p13, m_arg14 p14, m_arg15 p15) override { \
WRITE_ACTION \
if (Thread::get_caller_id() != server_thread) { \
command_queue.push(server_name, &ServerName::m_type, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13, p14, p15); \
} else { \
command_queue.flush_if_pending(); \
server_name->m_type(p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13, p14, p15); \
} \
A lot of progress with canvas rendering, still far from working.
2019-06-16 04:45:24 +02:00
}
Code quality: Fix header guards consistency Adds `header_guards.sh` bash script, used in CI to validate future changes. Can be run locally to fix invalid header guards.
2022-07-23 23:41:51 +02:00
#endif // SERVER_WRAP_MT_COMMON_H
Reference in a new issue Copy permalink