* Netplay Stuff
## PROTOCOL FALLBACK
In order to support older clients a protocol fallback system was introduced.
The host will no longer send its header automatically after a TCP connection is established, instead, it awaits for the client to send his before determining which protocol this connection is going to operate on.
Netplay has now two protocols, a low protocol and a high protocol; the low protocol is the minimum protocol it supports, while the high protocol is the highest protocol it can operate on.
To fully support older clients, a hack was necessary: sending the high protocol in the unused client's header salt field, while keeping the protocol field to the low protocol. Without this hack we would only be able to support older clients if a newer client was the host.
Any future system can make use of this system by checking connection->netplay_protocol, which is available for both the client and host.
## NETPLAY CHAT
Starting with protocol 6, netplay chat is available through the new NETPLAY_CMD_PLAYER_CHAT command.
Limitations of the command code, which causes a disconnection on unknown commands, makes this system not possible on protocol 5.
Protocol 5 connections can neither send nor receive chat, but other netplay operations are unaffected.
Clients send chat as a string to the server, and it's the server's sole responsability to relay chat messages.
As of now, sending chat uses RetroArch's input menu, while the display of on-screen chat uses a widget overlay and RetroArch's notifications as a fallback.
If a new overlay and/or input system is desired, no backwards compatibility changes need to be made.
Only clients in playing mode (as opposed to spectating mode) can send and receive chat.
## SETTINGS SHARING
Some settings are better used when both host and clients share the same configuration.
As of protocol 6, the following settings will be shared from host to clients (without altering a client's configuration file): input latency frames and allow pausing.
## NETPLAY TUNNEL/MITM
With the current MITM system being defunct (at least as of 1.9.X), a new system was in order to solve most if not all of the problems with the current system.
This new system uses a tunneling approach, which is similar to most VPN and tunneling services around.
Tunnel commands:
RATS[unique id] (RetroArch Tunnel Session) - 16 bytes -> When this command is sent with a zeroed unique id, the tunnel server interprets this as a netplay host wanting to create a new session, in this case, the same command is returned to the host, but now with its unique session id. When a client needs to connect to a host, this command is sent with the unique session id of the host, causing the tunnel server to send a RATL command to the host.
RATL[unique id] (RetroArch Tunnel Link) - 16 bytes -> The tunnel server sends this command to the host when a client wants to connect to the host. Once the host receives this command, it establishes a new connection to the tunnel server, sending this command together with the client's unique id through this new connection, causing the tunnel server to link this connection to the connection of the client.
RATP (RetroArch Tunnel Ping) - 4 bytes -> The tunnel server sends this command to verify that the host, whom the session belongs to, is still around. The host replies with the same command. A session is closed if the tunnel server can not verify that the host is alive.
Operations:
Host -> Instead of listening and accepting connections, it connects to the tunnel server, requests a new session and then monitor this connection for new linking requests. Once a request is received, it establishes a new connection to the tunnel server for linking with a client. The tunnel server's address and port are obtained by querying the lobby server. The host will publish its session id together with the rest of its info to the lobby server.
Client -> It connects to the tunnel server and then sends the session id of the host it wants to connect to. A host's session id is obtained from the json data sent by the lobby server.
Improvements (from current MITM system):
No longer a risk of TCP port exhaustion; we only use one port now at the tunnel server.
Very little cpu usage. About 95% net I/O bound now.
Future backwards compatible with any and all changes to netplay as it no longer runs any netplay logic at MITM servers.
No longer operates the host in client mode, which was a source of many of the current problems.
Cleaner and more maintainable system and code.
Notable functions:
netplay_mitm_query -> Grabs the tunnel's address and port from the lobby server.
init_tcp_socket -> Handles the creation and operation mode of the TCP socket based on whether it's host, host+MITM or client.
handle_mitm_connection -> Creates and completes linking connections and replies to ping commands (only 1 of each per call to not affect performance).
## MISC
Ping Limiter: If a client's estimated latency to the server is higher than this value, connection will be dropped just before finishing the netplay handshake.
Ping Counter: A ping counter (similar to the FPS one) can be shown in the bottom right corner of the screen, if you are connected to a host.
LAN Discovery: Refactored and moved to its own "Refresh Netplay LAN List" button.
## FIXES
Many minor fixes to the current netplay implementation are also included.
* Remove NETPLAY_TEST_BUILD
This can fix a lot of performance issues, like audio crackling and frame
time spikes. This requires the GameMode package to be installed. See:
https://github.com/FeralInteractive/gamemode
This commit adds a "Game Mode" bool option to the "Power
Management" and "Latency" settings sections, and it can be toggled
on/off without restarting RA.
The actual toggling of game mode happens in a new frontend platform
interface function. Perhaps this will become useful for other platforms
that provide some equivalent of Linux GameMode.
Since the GameMode ABI is fixed, and the API comes as a single,
header-only file with no actual deps, we simply bundle the header
(deps/feralgamemode/gamemode_client.h.) That way, all Linux builds will
have support for GameMode regardless of whether the GameMode development
package is installed or not.
* Added setting to allow/disallow players other than the host from pausing the game.
* Added a sublabel for netplay max connections.
* Fixed port override macro from not being set immediately after the port setting.
* Fixed memory leaks.
Adds new entry (where the entry for dumping CD is) to eject the disc -
in case the CD drive has no physical button to eject the disc / the
button does not work. Useless for most, but quality of life improvement
for some.
Upstream of patch used by Lakka at build time
1943ad296e/packages/libretro/retroarch/patches/retroarch-99-eject_disc.patch
* Any pad can control the menu
== DETAILS
I am not sure I've quite got it so that any pad can *open* the
menu, but I do have it so any pad can control it.
- split out the input processing into a separate method
- track down and squish some hairy bugs that boiled down to
bad pointer math
- it looks like `menu_driver.c` has a mix of line endings, so I
ran it through `dos2unix` so it has consistent line endings
again.
- verified that this change did not impact actual cores
* optimize out cumulative_bits
* Incorporate PR feedback
Many thanks to @jdgleaver for providing these optimizations.
* apply one more optimization
* Fix for warning and fix for incorrect comment
* Fixed contrast to be more correct - now scales from 0-10 linearly and behaves more the way you'd expect it to - changed name to ditch legacy settings users may have
Added ability to skip inverse tonemapper to the shader via the constant buffer using 'inverse_tonemap' - set to 0.0f to skip
Fixed potential bug when swapping between hdr and sdr and the bit depth not being set correctly
Fixed dx11's blend, rasterizer and topology states not being set to the sames when using hdr and leaving the menu - caused issues with PCSX2's Shadow of the Colossus
Added numerous helper functions to help create the correct values to colour the UI - normally the white UI elements should be rendered at paper white not max brightness for various reasons
* Fix stylistic issues - * Don't use camelcase for variables and function names * Use '(void)' for function declarations instead of () in C code * Declare variables at the top of a function or code block * Make sure functions that return a value always have a default return path that is not encapsulated by an else block * Use more unique names for retro_math functions which are less likely to overlap with other libraries' function symbols
Co-authored-by: twinaphex <libretro@gmail.com>
* Add HDR support
* Attempt to fix Mingw build and Metal builds
* (D3D12) Fix relative header includes
* Add missing hdr_sm5.hlsl.h
* (d3d12_common.c) Some C89 build fixes
* Fix MSVC build
* - Attempt to fix build on mingw/msys unix with dirty hack
- Fix shader compilation of hdr_sm5.hlsl.h on MSVC/Visual Studio -
the define was seen as an error and was causing the first pipeline
to error out
- Make sure we manually set handle of backBuffer to NULL
* Moving the release of the texture above the freeing of desc.srv_heap
and desc.rtv_heap solves the hard crashes on teardown/setup in RA -
it was crashing hard in d3d12_release_texture before
* Add HAVE_D3D12_HDR ifdef - needs to be disabled for WinRT for now
because of several things that are Windows desktop-specific right now
(GetWindowRect)
* Add dirty GUID hack - should work for both mingw/msys on Windows/Linux
as well as MSVC/Visual Studio (hopefully)
* Change HAVE_D3D12_HDR to HAVE_DXGI_HDR
* Move away from camelcase named variables
* Fix RARCH_ERR logs - they need a newline at the end
* d3d12_check_display_hdr_support - make it return a bool on return
and set d3d12->hdr.support and d3d12->hdr.enable outside of the
function
* (DXGI) Remove D3D12 dependencies from dxgi_check_display_hdr_support and
move it to dxgi_common.c instead
* (DXGI) move d3d12_swapchain_color_space over to dxgi_common.c and
rename it dxgi_swapchain_color_space
* (DXGI) move d3d12_set_hdr_metadata to dxgi_common.c and
rename it dxgi_set_hdr_metadata
* (DXGI) dxgi_check_display_hdr_support - better error handling?
* Fix typo
* Remove video_force_resolution
* (D3D12) Address TODO/FIXME
* (D3D12) Backport
c1b6c0bff2
- Fixed resource transition for present when HDR is off
Fixed cel shader displaying all black as blending was enabled when the hdr shader was being applied - turned off blending during this shader
* Move d3d12_hdr_uniform_t to dxgi_common.h and
rename it dxgi_hdr_uniform_t
* (D3D11) Add HDR support
* Add TODO/FIXME notes
* Cache hdr_enable in video_frame_info_t
* Update comment
with latest Xbox Series dashboard; DXGIResizeBuffers passing 0, 0 as
width/height is apparently problematic as it changes 0, 0 to 8,8
instead,
breaking the program
Fixed SR close match refresh bug.
Added menu high resolution option.
Fixed desktop restore bug cuusing endless resolution change requests.
Fixed file conflicts
Added destop restore resolution back in for manu only.
Pulled Switchres fixes.
Added better PI rsolution support.
Ver 0.7 SR2 (Switchres API) Implimantation
Removed HH experimetal check. This is better done via teh switchres.ini at present.
Fixed refresh rate bug. Now new resolution and refesh is added correctly.
Removed SR deinit from menu restore. Meanu now stays at last content resolution.
Ver 0.6.2 SR2 (Switchres API) Implimantation
Fixed super resolution bug casuing abnormal video size and aspect ratio
Fixed logging issue casuing seg falts on RA exit
Ver 0.6 SR2 (Switchres API) Implimantation
Ver 0.2 SR2 (Switchres API) Implimantation
Added forced super resolutions.
Added Multi-monitor/monitor selection support.
Added desktop resolution restore when switching back to menu only.
Added new menu items for 31KHz standard and 120hz monitor profiles.
Added new menu item INI. load monitor profile from switchrss.ini.
Fixed winraw driver. Coordinates new refreshed after a resolution change.
Fixed Menu aspect ratio in super resolutions.
Removed static glabals. These have been added to videocrt_switch struct.
Ver 0.1 SR2 (Switchres API) Implimantation
Removed old CRTSwitchRes method. Added new SR2 API implimantaion.
Resolution swithcing is now done by switchres libs. Both Linux and Windows
working with native and super resolutions. Working multi-monitor support
with monitor index selection. Working 31KHz support with standard and 120Hxz
modes. The monitor index selection is still done via the RA UI. Only choose
native and 15KHz form the CRT options in the RA UI as all options are now set
in the switchres.ini. All other CRT optoins in the RA UI currently do nothing.
Added SR wrapper to fix compile issues. Added back RPi functionality
Fixed windows resize/scaling issues on resolution change
Thanks @Calamity no more need for crt_switch_driver_refresh()
Fix broken case after prevous commit
Monitor preset options 15/31KHz now active. Added new meu option.
Moitor persets can now be choosen fom the RA UI. 15KHz and 31KHz will set
arcade_15 and aracde_31 respectivly. New option INI, if this is chosen your
monitor preset will be selected from your switchres.ini file.
Added 3KHhz, 120Hz. for old RA users. Renamed 31KHz to 31 KHz, Standard
Fixed winraw input coordinates after switching resolution.
Code cleanup
Fixed menu aspect ratio issue
Added menu resolution restore after closing content
Fixed aspect ratio after menu resolution restore.
code clean up
Fxed menu Resulition Restore Aspect Ratio. When SR uses non integer scalled resolution.
super width bug with restoring menu resolution fix
added super resolution check after setting desktop resolutoion variables
when menu active only sr_deinit() used to restore desktop mode.
Fixed menu sr_deinit bug. now setting sr_active false
Removed static globals, added them to video_switch struct
Fixex compile bug due to comment //
Fixed compile issues doe to c++ comments in teh switchres_wrapper.h
Temporarily removed SR2 logging to fix compile isses for c90
added logging back in. Removed support for winnt and osx
Added define for C89. Disabled SR if defined C89
Removed all RA compile fixes fro C89 C90 etc. Swithing now working again.
Put Switchres behind HAVE_SR2. HAVE_SR2=no by default. --enable-sr2
Ver 0.5 SR2 Implimentation.
Ver 0.4 SR2 Implimantation.
Bake SR inside RA
Removed temporary log files
Disable switchres when C89/C99 builds.
Removed C89 and C90 checks for SR
Fixed switchres_wrapper.h location
Ver 0.3 SR2 Implimenation
Dissable logging for C89 __STDC__
Fix For RPi
fixed missing EOL
fixed RPi function definition
added vidrocrt_switch stuct to RPI funcion
fixed xoffset for RPi
Removed old RPi function call
SR disabled for videocore until VC4 switching ported
Reverted back to state 5c8a56c Bake SR inside RA
Use native win32 api for threads.
Fix static lib linking
LIBERROR would be defined twice otherwise + improper function names prefixed by __imp_
Added lidstc++ to makefile.common for switchres
Fixed RPi switching. Disabled Switchres for videocore unill it is ported.
removed RAA.log. Should not exist
Added check for when SR fails to set mode with an aspect ratio fix.
added video driver re init for RPi
GB, GBA and GBC core check, adjusted reseolutions and scale. Please turn on integer scalling in the RA UI
Added logas back in. Checking STDC verstion >= C11
Fixed c89 for loop declaration.
Code clean up. Added new functions
Fix resolution switching bug introduces with HH code clean up.
Fixed menu restore bug on closw content after code clean up
Moved SR logging to relevant RA logs
Update makefile. Checks for X11 and xrandr
fixed makefile
Use native win32 api for threads.
Fix static lib linking
LIBERROR would be defined twice otherwise + improper function names prefixed by __imp_
Update switchres_wrapper.* header comments
Update year copyright
DRMKMS: build only if libdrm has the required version
XRANDR: build only if xrandr is available
Simplified maklefile
Fixed RPI compile error with unsued functions.
As before
Disable Griffin. No switching support available. Never has been
Removed log file 1
Added Win32 static define
Added SR source
Removed Videocore check on destroy SR
Moved SR deinit to trigger earlier on RA exit.
Fixed compile error after upstream rebase
Fixed aspect ration bug cused by super resolutions. Temporarily disbabled SR logging
Re inabled runtim eSR loggind. Disableed all RARCH logging on retro_deinit_drivers
Removed srdeinit from menu restore. Menu stays in current reolution until a fix can be found
Fixed refresh rate changes when no reolution change is detected.
Forgot to add teh resolution cahge in with the refresh change oops
Fixed endless no detection log.
Removed HH check. This can been better adjusted using the switchres.ini
fixed compile issue
Added better PI crt switching and fixed typo
Pulled Swicthres fixes. Updated desktop restore resolution.
removed unused makefile
Lockec menu refresh to 60hz
fixed missing new line
Fixed file conflicts
Forced 640x480@60 for menu
Added high resolution menu option
Removed item logg checker
Fixed typos
Removed unused functions
Fixed SR close match refesh bug.
Fixed typo
- Core info cache can now be enabled/disabled on all platforms via a new `Settings > Core > Cache Core Info Files` option
- Core info cache file has been renamed from `.cache` to `core_info.cache` (i.e. it is no longer a 'hidden' file on Unix platforms, so can be deleted easily)
- The core info cache file is now compressed (rzip) to further reduce disk IO
- The presence of a `core_info.refresh` file in the core info directory will force a one-time refresh of the info cache. This file is generated automatically when toggling on the `Cache Core Info Files` option, and we will also add it to core info file packaging such that updating info files (either manually or via the online updater) will force a refresh
- The core info cache no longer contains 'core is locked' and 'firmware missing' data fields; these are 'dynamic' properties that must be determined at runtime
- The 'core is locked' status is now determined on core info intialisation by parsing the core directory listing, rather than by performing indivdual 'lock file exists' checks. This minimises file IO, and greatly improves performance on devices with slow storage
- While parsing the core info cache file, we now avoid unnecessary `strdup()`s when adding entries to the resultant cache list
- Memory leaks (potential and real) have been fixed, and safety checks added
- Build errors have been fixed
So far, if display is scaled, overlays gets correctly drawn but touch
input is not correctly scaled, resulting in an unusable overlay.
This happens for touches in menu too.
This commit aims to introduce this scaling factor adding it to the
config file, eg:
input_touch_scale = "2"
