- Adds support for abstract implementations
- Adds native windowing implementations for WIN32 and X11 as fallbacks
when present support is lacking (headless configs)
- For some reason the hardware forgets that primitive restart is enabled and tries to actually read vertex index 65535
- Works correctly if uint32 vertex indices are used instead of uint16 for cases where primitive restart is active
- Fix for texture barriers
- vulkan: Rework texture cache handling of depth surfaces
- Support for scaled depth blit using overlay pass
- Support proper readback of D24S8 in both D32F_S8 and D24U_S8 variants
- Optimize the depth conversion routines with SSE
- vulkan: Replace slow single element copy with std::memcpy
- Check heap status before attempting blit operations
- Bump guard size on upload buffer as well
- Implement flush-always behaviour to partially fix readback from a currently bound fbo
- Without this, only the first read is correct, as more draws are added the results become 'wrong'
- Fixes WCB and cpublit behviour
- Synchronize blit_dst surfaces to avoid data loss when gpu texture scaling is used
- Its still faster in such cases to disable gpu texture scaling but some types cannot be disabled without force cpu blit (e.g framebuffer transfers)
- Memory management tuning
- rsx: on-demand texture cache rescanning for unprotected sections
- rsx: Only framebuffer resources are upscaled
- Do not resize regular blit engine resources
- Lazy initialize readback buffer when using opengl
-- These measures should help minimize vram usage
Six instructions changed to use xmm registers instead of gpr.
ROTQBII, ROTQMBII, SHLQBII look better (shifts by imm)
ROTQBI, ROTQMBI, SHLQBI changed for consistency (shifts by variable)
Only AVX-512 path is changed (third version).
This instruction is extremely rare.
And the code is probably not optimal.
So this commit is pretty useless.
* Fix gs_frame spawning on a screen other than the one the RPCS3 window is on for multi-monitor setups
* Cleaned up code & refactored it into a utility function for reuse
* Qt: take gs_frame's framemargins into account by using showEvent
- For some surfaces, dimensions are passed via the log2 bits rather than surface pitch
-- This is similar to the setup for nv406e and probably means the surfaces are padded and swizzled
- Do not assume texture2D when creating new textures
- Flag invalid texture cache if readonly texture is trampled by fbo memory.
Avoids binding a stale handle to the pipeline and is rare enough that it should not hurt performance
- Do not add unused subroutines in shaders unless necessary
-- makes shaders easier to read and disassembled spir-v has less clutter
- glsl: Replace switch block with lookup table
- vulkan: Do not assume an aux frame context must exist in a well defined state as set in init_buffers() since the request might be external (via overlays path)
- gl: Do not bother waiting for idle before servicing external flip requests
- gl: Queue overlay cleanup requests to ensure only glthread attempts touching the context
- overlays: Do not compute size metrics for invalid/unsupported glyphs
- opengl driver optimization for nvidia. On nvidia glTextureBufferRange performance is horrendous
-- Initialize texture buffer to whole buffer at startup and use absolute offsets to read data instead
-- Over 2x performance in some cases (Resogun, TNT racers)
- gl/vk: Do not flip non-existent display buffers. Fixes spec violation at boot in TNT racers demo
- whitespace fixes for sys_rsx
- Implement low bit decode override flags for 2-component textures
- Properly implement alot of texture remaps according to the autotest results
rsx: Do not unnecessarily shuffle WZYX->RGBA unless we have proof
- From looking at format swizzles, this is incorrect
- Use edges of depth range to map clamped stuff
Disable range compression on regular draws vs extended range draws
- Some applications require full 0-1 usage without compromises.
-- TODO: This leaves the extended range z values to fight with regular draws in the .99 - 1.0 range
- The scale offset matrix is fine but on real hardware the z results seem to be independent of near/far clipping distances
-- If depth falls within near/far, clamp depth value to [0,1]
- Always flush the primary queue and wait if not involking readback from rsx thread
-- Should fix some instances of device_lost when using WCB
-- Marked remaining case as TODO
-- TODO: optimize amount of time rsx waits for external threads trying to read
fix input regression and fix input for FIFA games
fix input in NASCAR [BLUS30932]
fix port status query -> disconnected devices don't cripple following devices by decreased now_connect
- Track last working state and reset to it if RSX starts to desync
-- This is especially useful when running vulkan since the renderer will easily outpace the rest of the system when merely recording draw commands
- Ignore empty sets
-- Mark empty/invalid IB sets as having 0 element counts.
- Doesn't account for dynamic libraries loaded after the fact,
but usually good enough since
1) Those aren't even present in some games
2) They usually only have about 1 or 2 fragments (dialogs) each.
- Make all texture access on non-existent textures return 0
- If border color is closer to 0, then set alpha to 0 as well (might break some corner cases with alpha test)
- Zero initialize null sampler
- Force mipmap count to 1 if sampling from an RTV/DSV
- TODO: Better wcb flush detection, it should be better to re-upload the texture after it has been dwnloaded if expected mipmaps are > 1
- Sometimes square renders are done to surfaces with pitch=64 and re-uploaded with swizzle scanning
-- This setup avoids discarding targets if they are square and pitch == 64
- Abort nv406e semaphore acquire if the rsx thread stalls/crashes
- Fix texture size approximation to take mipmaps into account. Fixes some games hanging with WCB
- Avoid unprotecting memory until just before we have to write the data
- Avoids race conditions where the caller thread takes too long to enter the second phase and another thread accesses the "bad" memory
- Reorganize storage hash vs ucode hash
- Scan for actual fragment program start in case leading NOPed code precedes the actual instructions
-- e.g FEAR2 Demo has over 32k of padding before actual program code that messes up hashes
- Fix VERTEX_DATA_3F_M element alignment (its 16 bytes per attribute)
- Fix DATA_2S_X interpretation type. Its signed 16-bit unnormalized (s32k) and not signed normalized (s1)
- Zeta pitch is ignored by real HW for some reason
- Monitor ptch value changes as well since they may affect disabled surfaces
- TODO: Verify if MRT pitch is really taken into consideration
- Discard intentionally invalidated framebuffer resources. These are created after a flush has happened, forcing reupload since contents cannot be guaranteed (strict mode only)
- Fix for blits using vulkan; dont use the copy method if formats do not match, use generic blit instead
- Handle aliased depth + color target by disabling depth writes. This looks to be the correct way
- Add support for generic passes that cannot be done using general imaging operations. Lays the framework for tons of features and effects
- Implement RGBA->D24D8 casting. Sometimes games will split depth texture into RGBA8 then use the new RGBA8 as a depth texture directly
-- This happens alot in ps3 games and I'm not sure why. Its likely the ps3 did not sample fp values with linear filtering so this is a workaround
-- Only implemented for openGL at the moment
-- Requires a workaround for an AMD driver bug
* Input: further work on remapping Xinput and begin work on remapping DS4
* Input: Improve pad_settings_dialog a bit and begin Remapping for XInput
* Input: begin evdev remapping and change all handlers to use cfg::string
* Input: finish work on remapping evdev
and some more crap
* Input: finish work on remapping Xinput and DS4
* Input: add DS4 Colors to DS4 config
* Input: Improve DS4 deadzone scaling
Jarves made some mistakes, so I'll fix them in the follow up commit
* Input: fix Jarves fixes on DS4 deadzone
and remove unnecessary usage of toUtf8
* Input: add primitive batterychecks to XInput and DS4
* Input: add mmjoystick remapping
* Input: Fix evdev and some Vibration issues
* Input: adjust capabilities to fix stick input for games like LoS 2
also fix threshold slider minimum
also add ps button to all the handlers
* Input: Further evdev work
based on danilaml code review and own debugging:
Fixed path issue, <= 0 issue, some captures, const, axis with same codes.
Adds a map to each device that differentiates negative and positive axis mappings.
adjusted rest of the file to tabs (ListDevices and beginning of threadProc)
* Input: use 20ms vibration update time for xbox one elite controllers.
* Input: Fix return type of Clamp()
* Input: Evdev Fix
* Input: Evdev Optional GetNextButtonPress
presumably better than the other
* Input: review changes
* Input: evdev: fix wrong index in axis handling
move bindpadtodevice down to keep consistency between handlers and not get crazy
* Input: evdev: fix expensive add_device in GetNextButtonPress
* cleanup
* Input: mmjoy: fix type
* Input: evdev: final fixes
* Input: evdev: exclude unnecessary buttons while mapping Xbox 360 or DS4
* Input: add deadzone preview by passing necessary values in callback
use 0.5 of max value for threshold in pad dialog
* Input: get rid of all-uppercase variables
* cmake: simplify GLEW integration
* cmake: simplify ZLIB integration
* cmake: simplify ALSA detection
* cmake: check C++ compiler and flags instead of C ones
* cmake: bring consistency in Qt packages detection and remove unnecessary related "include_directories"
* cmake: harmonize coding style
* cmake: set CMAKE_INCLUDE_CURRENT_DIR to ON only when necessary
* Trophy: Save dialog state and add show type settings
* SaveDataManager: Save Dialog State
* SaveDataList: Minor Optimization
* Qt: Save icon size on mouseevent resizes
it's a bit slower than using the slider because it saves every single resize. But better than not saving at all for now
* SaveData: Optimize saving to settings a bit
No Saving needed there
* Qt: get rid of all-uppercase enums and namespaces
* Qt/Linux: adjust remaining DX12 tooltip
* Qt: prevent dockwidget contextmenu
Added force_boot to force boot on cmdline boot.
Load() caused a Stop() that exited the application with "Exit RPCS3 when process finishes" enabled. Now Stop is only called if the emu is not stopped
* Thread: unbreak on BSDs after dbc9bdfe02
Utilities/Thread.cpp:1920:2: error: unknown type name 'cpu_set_t'; did you mean 'cpusetid_t'?
cpu_set_t cs;
^~~~~~~~~
cpusetid_t
/usr/include/sys/types.h:84:22: note: 'cpusetid_t' declared here
typedef __cpusetid_t cpusetid_t;
^
Utilities/Thread.cpp:1921:2: error: use of undeclared identifier 'CPU_ZERO'
CPU_ZERO(&cs);
^
Utilities/Thread.cpp:1922:2: error: use of undeclared identifier 'CPU_SET'
CPU_SET(core, &cs);
^
Utilities/Thread.cpp:1923:48: error: unknown type name 'cpu_set_t'; did you mean 'cpusetid_t'?
pthread_setaffinity_np(pthread_self(), sizeof(cpu_set_t), &cs);
^~~~~~~~~
cpusetid_t
* JIT: use MAP_32BIT on Linux and FreeBSD
Unless RLIMIT_DATA is low enough FreeBSD by default reserves lower 2Gb
for brk(2) style heap, ignoring mmap(2) address hint requested by RPCS3.
Passing MAP_32BIT fixes the following crash
Assertion failed: ((Type == ELF::R_X86_64_32 && (Value <= UINT32_MAX)) || (Type == ELF::R_X86_64_32S && ((int64_t)Value <= INT32_MAX && (int64_t)Value >= INT32_MIN))), function resolveX86_64Relocation, file /usr/ports/devel/llvm40/work/llvm-4.0.1.src/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp, line 287.
* build: unbreak -DVULKAN_PREBUILT with system glslang on Unix
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:4:10: fatal error: '../../../../Vulkan/glslang/SPIRV/GlslangToSpv.h' file not found
#include "../../../../Vulkan/glslang/SPIRV/GlslangToSpv.h"
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
rpcs3/CMakeFiles/rpcs3.dir/Emu/RSX/VK/VKCommonDecompiler.cpp.o: In function `vk::compile_glsl_to_spv(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >&, glsl::program_domain, std::__1::vector<unsigned int, std::__1::allocator<unsigned int> >&)':
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x50e): undefined reference to `glslang::TProgram::TProgram()'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x51d): undefined reference to `glslang::TShader::TShader(EShLanguage)'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x542): undefined reference to `glslang::TShader::setStrings(char const* const*, int)'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x581): undefined reference to `glslang::TShader::parse(TBuiltInResource const*, int, EProfile, bool, bool, EShMessages, glslang::TShader::Includer&)'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x5d6): undefined reference to `glslang::TProgram::link(EShMessages)'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x5f1): undefined reference to `glslang::GlslangToSpv(glslang::TIntermediate const&, std::__1::vector<unsigned int, std::__1::allocator<unsigned int> >&, glslang::SpvOptions*)'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x5ff): undefined reference to `glslang::TShader::getInfoLog()'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x61a): undefined reference to `glslang::TShader::getInfoDebugLog()'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x630): undefined reference to `glslang::TShader::~TShader()'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x63c): undefined reference to `glslang::TProgram::~TProgram()'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x6d2): undefined reference to `glslang::TShader::~TShader()'
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x6de): undefined reference to `glslang::TProgram::~TProgram()'
rpcs3/CMakeFiles/rpcs3.dir/Emu/RSX/VK/VKCommonDecompiler.cpp.o: In function `vk::initialize_compiler_context()':
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x6f5): undefined reference to `glslang::InitializeProcess()'
rpcs3/CMakeFiles/rpcs3.dir/Emu/RSX/VK/VKCommonDecompiler.cpp.o: In function `vk::finalize_compiler_context()':
rpcs3/Emu/RSX/VK/VKCommonDecompiler.cpp:(.text+0x856): undefined reference to `glslang::FinalizeProcess()'
* build/msvc: add missing glslang include directory after 6bb3f1b4d75c
"c:\projects\rpcs3\rpcs3\VKGSRender.vcxproj" (default target) (15) ->
(ClCompile target) ->
Emu\RSX\VK\VKCommonDecompiler.cpp(4): fatal error C1083: Cannot open include file: 'SPIRV/GlslangToSpv.h': No such file or directory [c:\projects\rpcs3\rpcs3\VKGSRender.vcxproj]
- Fix texture cache blit behaviour when src has AA enabled and dst is a blit dst texture with or without AA
-- This requires handling AA resolve by removing a half downscale on multisampled axes
- Return all ones when a vertex attribute is disabled.
-- Some games forget to enable vertex attributes actually needed by the fs
- texture_cache: Fix internal size calculation for subresources
- vk: Delay dynamic state updates until just about to draw to ensure no flush has discarded the cb state
- In rare cases the driver derefs a nullptr and dies, taking the emulator with it
- From testing, it seems the vram is indeed freed when this happens so its "safe" to continue
- Support for raster offsets in surface descriptors (looks to be unused)
- Do not tag disabled render targets when using MRT (pitch = 64)
- Add missing notify_surface_changed() call for openGL
- Use AA mode to predict surface compression. Compression mode is useless without AA activated
- Rewrites most image subresource fetch routines to use the new heuristic
- Fix rsx:🧵:find_tile. FEED000(X) can be substituted for (X) in the code
-- Fixes alot of failures when looking for tiled regions
rsx: Fix antialiased unnormalized coords
- scaling factors are inverse to allow proper coordinates to be computed in fs
- Remove generic throws from the rsx pipeline. Stops the rsx thread from silently dying leaving the emulator in a hung state
- Hackplement add_signed and reverse_subtract_signed blend modes
- Reimplement fragment program fetch and rewrite texture upload mechanism
-- All of these steps should only be done at most once per draw call
-- Eliminates continously checking the surface store for overlapping addresses as well
addenda - critical fixes
- gl: Bind TIU before starting texture operations as they will affect the currently bound texture
- vk: Reuse sampler objects if possible
- rsx: Support for depth resampling for depth textures obtained via blit engine
vk/rsx: Minor fixes
- Fix accidental imageview dereference when using WCB if texture memory occupies FB memory
- Invalidate dirty framebuffers (strict mode only)
- Normalize line endings because VS is dumb
Added a list widget that displays all the currently active breakpoints. You can rename each one for convenience, delete them in bulk, or double click to jump to that address.
As soon as I tried the debugger, i wanted to move it to the middle almost instantly. Checked the code and found it sitting right here but commented out. Why? This is awesome.
to flush_all is up to date. Should prevent recursive exceptions
Partially revert Jarves' fix to invalidate cache on tile unbind. This will
need alot more work. Fixes hangs
- Identify minimize/restore events as separate from regular resize and do not react to them
- Enable message queue consumption after loading the shaders cache. Also hides the frame in this step
-- This fixes the 'start fullscreen' bug when running vulkan
- This allows for better handling of deferred flushes.
-- There's still no guarantee that cache contents will have changed between the set acquisition and following flush operation
-- Hopefully this is rare enough that it doesnt cause serious issues for now
- vk: Always reopen primary command buffers. They should only be closed in flush_command_queue
- If uploading a texture and there are collisions with protected buffers, do not rebuild the cache
- Perform writes via flush before reprotecting pages that were not trampled
- Only flush no pages once
