system: add console language conf

Extracted from Stefanos Costa's pthread fix in #147.

.github/workflows/build.yml

@@ -73,7 +73,7 @@ jobs:
       - name: Configure
         shell: cmd
         run: |
-          cmake -S src -B _Build/windows ^
+          cmake -S . -B _Build/windows ^
             -G Ninja ^
             -DCMAKE_BUILD_TYPE=Release ^
             -DCMAKE_C_COMPILER=clang-cl ^
@@ -147,7 +147,7 @@ jobs:
       - name: Configure
         shell: bash
         run: |
-          cmake -S src -B _Build/macos \
+          cmake -S . -B _Build/macos \
             -G Ninja \
             -DCMAKE_BUILD_TYPE=Release \
             -DCMAKE_OSX_ARCHITECTURES=x86_64 \
@@ -270,7 +270,7 @@ jobs:
         shell: bash
         run: |
           mkdir -p _Build
-          cmake -S src -B _Build/linux \
+          cmake -S . -B _Build/linux \
             -G Ninja \
             -DCMAKE_BUILD_TYPE=Release \
             -DCMAKE_C_COMPILER=clang \

.gitignore

@@ -3,4 +3,5 @@
 .idea/
 build/
 _Build/vscode-clang/
-_Build/
+_Build/
+cmake-build-debug/

.vscode/settings.json

@@ -1,5 +1,5 @@
 {
-    "cmake.sourceDirectory": "${workspaceFolder}/src",
+    "cmake.sourceDirectory": "${workspaceFolder}",
     "cmake.buildDirectory": "${workspaceFolder}/_Build/vscode-clang",
     "cmake.generator": "Ninja",
     "cmake.environment": {

CMakeLists.txt

@@ -6,6 +6,10 @@ endif()
 
 project(Kyty)
 
+set(KYTY_SOURCE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/src")
+set(KYTY_TESTS_DIR "${CMAKE_CURRENT_SOURCE_DIR}/tests")
+set(KYTY_THIRD_PARTY_DIR "${CMAKE_CURRENT_SOURCE_DIR}/3rdparty")
+
 if(CMAKE_SYSTEM_NAME MATCHES ".*Linux")
 	set(LINUX TRUE)
 endif()
@@ -19,13 +23,11 @@ set(CMAKE_CXX_STANDARD_REQUIRED ON)
 set(CMAKE_CXX_EXTENSIONS OFF)
 set(CMAKE_CXX_SCAN_FOR_MODULES OFF)
 
-include(utils.cmake)
+include("${KYTY_SOURCE_DIR}/utils.cmake")
 include(CTest)
 
 option(KYTY_ENABLE_CLANG_TIDY "Run clang-tidy checks during builds" OFF)
 
-set(KYTY_THIRD_PARTY_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../3rdparty")
-
 set(CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS ON)
 
 include(TestBigEndian)
@@ -97,36 +99,36 @@ endif()
 
 project(Kyty${KYTY_PROJECT_NAME}${CMAKE_BUILD_TYPE}${KYTY_COMPILER} VERSION 0.2.2)
 
-include_directories(	
+include_directories(
 	${KYTY_THIRD_PARTY_DIR}/gtest/include
 	${KYTY_THIRD_PARTY_DIR}/gtest
 	${KYTY_THIRD_PARTY_DIR}/fmt/include
 	${KYTY_THIRD_PARTY_DIR}/magic_enum/include/magic_enum
 	${PROJECT_BINARY_DIR}
-	${CMAKE_CURRENT_SOURCE_DIR}
+	${KYTY_SOURCE_DIR}
 )
 
 set(KYTY_VERSION "${PROJECT_VERSION}")
 
 configure_file(
-  ${PROJECT_SOURCE_DIR}/cmake_config.h.in
+  ${KYTY_SOURCE_DIR}/cmake_config.h.in
   ${PROJECT_BINARY_DIR}/cmake_config.h
   )
 
 find_package(Git)
-  
+
 add_custom_target( KytyGitVersion
-    COMMAND ${CMAKE_COMMAND} 
-	-D INPUT_FILE=${CMAKE_CURRENT_SOURCE_DIR}/kytyGitVersion.h.in
-    -D OUTPUT_FILE=${CMAKE_CURRENT_BINARY_DIR}/kytyGitVersion.h
+    COMMAND ${CMAKE_COMMAND}
+	-D INPUT_FILE=${KYTY_SOURCE_DIR}/kytyGitVersion.h.in
+	-D OUTPUT_FILE=${PROJECT_BINARY_DIR}/kytyGitVersion.h
 	-D GIT_EXECUTABLE=${GIT_EXECUTABLE}
 	-D GIT_WORKING_DIRECTORY=${CMAKE_CURRENT_SOURCE_DIR}
-	-P ${CMAKE_CURRENT_SOURCE_DIR}/generate_version.cmake
+	-P ${KYTY_SOURCE_DIR}/generate_version.cmake
 	COMMENT "Generate kytyGitVersion.h"
 )
 
 if (CMAKE_CXX_COMPILER_VERSION VERSION_GREATER_EQUAL 12.0.0)
-	list(APPEND KYTY_IWYU 
+	list(APPEND KYTY_IWYU
 		kyty_emulator
 		common
 		#launcher
@@ -139,55 +141,55 @@ if (CMAKE_CXX_COMPILER_VERSION VERSION_GREATER_EQUAL 12.0.0)
 endif()
 
 option(KYTY_BUILD_LAUNCHER "Build Qt launcher" ON)
-  
-config_compiler_and_linker()  
-  
+
+config_compiler_and_linker()
+
 add_subdirectory("${KYTY_THIRD_PARTY_DIR}" "${CMAKE_CURRENT_BINARY_DIR}/3rdparty")
-add_subdirectory(common)
+add_subdirectory("${KYTY_SOURCE_DIR}/common" "${CMAKE_CURRENT_BINARY_DIR}/common")
 
 file(GLOB kyty_emulator_src CONFIGURE_DEPENDS
-	libs/*.cpp
-	libs/*.h
-	graphics/*.cpp
-	graphics/*.h
-	graphics/guest_gpu/*.cpp
-	graphics/guest_gpu/*.h
-	graphics/guest_gpu/command_processor/*.cpp
-	graphics/guest_gpu/command_processor/*.h
-	graphics/host_gpu/*.cpp
-	graphics/host_gpu/*.h
-	graphics/host_gpu/renderer/*.cpp
-	graphics/host_gpu/renderer/*.h
-	graphics/host_gpu/renderer/cache/*.cpp
-	graphics/host_gpu/renderer/cache/*.h
-	graphics/host_gpu/renderer/image/*.cpp
-	graphics/host_gpu/renderer/image/*.h
-	graphics/host_gpu/renderer/pipeline/*.cpp
-	graphics/host_gpu/renderer/pipeline/*.h
-	graphics/shader/*.cpp
-	graphics/shader/*.h
-	graphics/shader/recompiler/*.cpp
-	graphics/shader/recompiler/*.h
-	graphics/shader/recompiler/cfg/*.cpp
-	graphics/shader/recompiler/cfg/*.h
-	graphics/shader/recompiler/decompiler/*.cpp
-	graphics/shader/recompiler/decompiler/*.h
-	graphics/shader/recompiler/emitter/*.cpp
-	graphics/shader/recompiler/emitter/*.h
-	graphics/shader/recompiler/ir/*.cpp
-	graphics/shader/recompiler/ir/*.h
-	graphics/presentation/*.cpp
-	graphics/presentation/*.h
-	graphics/presentation/window/*.cpp
-	graphics/presentation/window/*.h
-	kernel/*.cpp
-	kernel/*.h
-	loader/*.cpp
-	loader/*.h
+	"${KYTY_SOURCE_DIR}/libs/*.cpp"
+	"${KYTY_SOURCE_DIR}/libs/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/guest_gpu/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/guest_gpu/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/guest_gpu/command_processor/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/guest_gpu/command_processor/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/cache/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/cache/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/image/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/image/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/pipeline/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/pipeline/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/shader/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/cfg/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/cfg/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/decompiler/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/decompiler/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/emitter/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/emitter/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/presentation/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/presentation/*.h"
+	"${KYTY_SOURCE_DIR}/graphics/presentation/window/*.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/presentation/window/*.h"
+	"${KYTY_SOURCE_DIR}/kernel/*.cpp"
+	"${KYTY_SOURCE_DIR}/kernel/*.h"
+	"${KYTY_SOURCE_DIR}/loader/*.cpp"
+	"${KYTY_SOURCE_DIR}/loader/*.h"
 )
 
 find_program(KYTY_GLSLANG_VALIDATOR glslangValidator REQUIRED)
-set(gpu_tiler_shader_dir "${CMAKE_CURRENT_SOURCE_DIR}/graphics/host_gpu/shaders")
+set(gpu_tiler_shader_dir "${KYTY_SOURCE_DIR}/graphics/host_gpu/shaders")
 set(gpu_tiler_generated_dir "${PROJECT_BINARY_DIR}/gpu_tiler_shaders")
 set(gpu_tiler_shader_names
 	standard256
@@ -215,7 +217,7 @@ foreach(shader_name IN LISTS gpu_tiler_shader_names)
 		COMMAND "${KYTY_GLSLANG_VALIDATOR}" -V --target-env vulkan1.0 -Os
 		        "-I${gpu_tiler_shader_dir}" -o "${shader_spv}" "${shader_source}"
 		COMMAND ${CMAKE_COMMAND} -DINPUT=${shader_spv} -DOUTPUT=${shader_header}
-		        -DSYMBOL=${shader_symbol} -P "${CMAKE_CURRENT_SOURCE_DIR}/embed_spirv.cmake"
+		        -DSYMBOL=${shader_symbol} -P "${KYTY_SOURCE_DIR}/embed_spirv.cmake"
 		DEPENDS "${shader_source}" ${gpu_tiler_shader_includes}
 		VERBATIM
 	)
@@ -239,7 +241,7 @@ foreach(shader_source IN LISTS gpu_blit_shader_sources)
 		COMMAND "${KYTY_GLSLANG_VALIDATOR}" -V --target-env vulkan1.0 -Os
 		        "-I${gpu_tiler_shader_dir}" -o "${shader_spv}" "${shader_source}"
 		COMMAND ${CMAKE_COMMAND} -DINPUT=${shader_spv} -DOUTPUT=${shader_header}
-		        -DSYMBOL=${shader_symbol} -P "${CMAKE_CURRENT_SOURCE_DIR}/embed_spirv.cmake"
+		        -DSYMBOL=${shader_symbol} -P "${KYTY_SOURCE_DIR}/embed_spirv.cmake"
 		DEPENDS "${shader_source}"
 		VERBATIM
 	)
@@ -248,8 +250,8 @@ endforeach()
 list(APPEND kyty_emulator_src ${gpu_blit_shader_headers})
 
 list(APPEND kyty_emulator_src
-	emulator.h
-	emulator.cpp
+	"${KYTY_SOURCE_DIR}/emulator.h"
+	"${KYTY_SOURCE_DIR}/emulator.cpp"
 )
 list(REMOVE_DUPLICATES kyty_emulator_src)
 
@@ -280,7 +282,7 @@ if(LINUX)
 endif()
 
 set(inc_headers
-	${CMAKE_CURRENT_SOURCE_DIR}
+	${KYTY_SOURCE_DIR}
 	${KYTY_THIRD_PARTY_DIR}/SDL2/include
 	${KYTY_THIRD_PARTY_DIR}/VulkanMemoryAllocator/include
 	${KYTY_THIRD_PARTY_DIR}/SPIRV-Tools/include
@@ -292,13 +294,13 @@ set(inc_headers
 
 if (KYTY_CLANG_CL)
 	list(APPEND kyty_emulator_link_libraries winpthread)
-	list(APPEND inc_headers 
+	list(APPEND inc_headers
 		${KYTY_THIRD_PARTY_DIR}/winpthread/include
 	)
 endif()
 
 list(APPEND check_headers
-	${CMAKE_CURRENT_SOURCE_DIR}
+	${KYTY_SOURCE_DIR}
 )
 
 function(add_kyty_full_emulator_test target source)
@@ -320,108 +322,108 @@ endfunction()
 function(configure_macos_guest_address_space target)
 	if(APPLE AND (CMAKE_OSX_ARCHITECTURES STREQUAL "x86_64" OR
 			(NOT CMAKE_OSX_ARCHITECTURES AND CMAKE_SYSTEM_PROCESSOR MATCHES "^(x86_64|AMD64)$")))
-		target_sources(${target} PRIVATE kernel/macosGuestAddressSpace.cpp)
+		target_sources(${target} PRIVATE "${KYTY_SOURCE_DIR}/kernel/macosGuestAddressSpace.cpp")
 		target_compile_definitions(${target} PRIVATE KYTY_LINKED_GUEST_ADDRESS_SPACE=1)
 		target_link_options(${target} PRIVATE
 			-Wl,-ld_classic,-no_pie,-no_fixup_chains,-no_huge,-pagezero_size,0x40000,-segaddr,SYSTEM_MANAGED,0x40000,-segaddr,SYSTEM_RESERVED,0x7ffffc000,-segaddr,USER_AREA,0x7000000000,-image_base,0x700000000000)
 	endif()
 endfunction()
 
-add_kyty_full_emulator_test(shader_cfg_tests ../tests/shaderCfgTests.cpp)
+add_kyty_full_emulator_test(shader_cfg_tests "${KYTY_TESTS_DIR}/shaderCfgTests.cpp")
 
 add_executable(scalar_provenance_tests EXCLUDE_FROM_ALL
-	../tests/ScalarProvenanceTests.cpp
-	graphics/host_gpu/hostMemory.cpp
-	graphics/shader/recompiler/ir/ReadLaneElimination.cpp
-	graphics/shader/recompiler/ir/ScalarProvenance.cpp
-	graphics/shader/recompiler/ir/SrtWalker.cpp
+	"${KYTY_TESTS_DIR}/ScalarProvenanceTests.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/hostMemory.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ReadLaneElimination.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ScalarProvenance.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/SrtWalker.cpp"
 )
 target_link_libraries(scalar_provenance_tests fmt::fmt)
 target_include_directories(scalar_provenance_tests PRIVATE ${inc_headers})
 
 add_executable(page_manager_tests EXCLUDE_FROM_ALL
-	../tests/PageManagerTests.cpp
-	graphics/host_gpu/pageManager.cpp
+	"${KYTY_TESTS_DIR}/PageManagerTests.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/pageManager.cpp"
 )
 target_include_directories(page_manager_tests PRIVATE ${inc_headers})
 
 add_executable(bit_array_tests EXCLUDE_FROM_ALL
-	../tests/BitArrayTests.cpp
+	"${KYTY_TESTS_DIR}/BitArrayTests.cpp"
 )
 target_include_directories(bit_array_tests PRIVATE ${inc_headers})
 
 add_executable(memory_tracker_tests EXCLUDE_FROM_ALL
-	../tests/MemoryTrackerTests.cpp
-	graphics/host_gpu/pageManager.cpp
-	graphics/host_gpu/memoryTracker.cpp
+	"${KYTY_TESTS_DIR}/MemoryTrackerTests.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/pageManager.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/memoryTracker.cpp"
 )
 target_link_libraries(memory_tracker_tests fmt::fmt common)
 target_include_directories(memory_tracker_tests PRIVATE ${inc_headers})
 
 add_executable(shader_vertex_metadata_tests EXCLUDE_FROM_ALL
-	../tests/ShaderVertexMetadataTests.cpp
-	graphics/host_gpu/hostMemory.cpp
-	graphics/shader/shaderVertexMetadata.cpp
+	"${KYTY_TESTS_DIR}/ShaderVertexMetadataTests.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/hostMemory.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/shaderVertexMetadata.cpp"
 )
 target_include_directories(shader_vertex_metadata_tests PRIVATE ${inc_headers})
 
 add_executable(shader_stage_runtime_tests EXCLUDE_FROM_ALL
-	../tests/ShaderStageRuntimeTests.cpp
-	graphics/guest_gpu/gpu_format.cpp
-	graphics/host_gpu/hostMemory.cpp
-	graphics/shader/shaderStageRuntime.cpp
-	graphics/shader/recompiler/ir/ResourceMaterialization.cpp
-	graphics/shader/recompiler/ir/ScalarProvenance.cpp
-	graphics/shader/recompiler/ir/SrtWalker.cpp
+	"${KYTY_TESTS_DIR}/ShaderStageRuntimeTests.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/guest_gpu/gpu_format.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/hostMemory.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/shaderStageRuntime.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ResourceMaterialization.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ScalarProvenance.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/SrtWalker.cpp"
 )
 target_link_libraries(shader_stage_runtime_tests fmt::fmt)
 target_include_directories(shader_stage_runtime_tests PRIVATE ${inc_headers})
 
 add_executable(resource_tracking_tests EXCLUDE_FROM_ALL
-	../tests/ResourceTrackingTests.cpp
-	graphics/guest_gpu/gpu_format.cpp
-	graphics/host_gpu/hostMemory.cpp
-	graphics/shader/recompiler/ir/ScalarProvenance.cpp
-	graphics/shader/recompiler/ir/SrtWalker.cpp
-	graphics/shader/recompiler/ir/SrtPatcher.cpp
-	graphics/shader/recompiler/ir/ResourceTracking.cpp
-	graphics/shader/recompiler/ir/ResourceMaterialization.cpp
-	graphics/shader/recompiler/ir/ShaderInfoCollection.cpp
-	graphics/shader/recompiler/ir/BindingLayout.cpp
+	"${KYTY_TESTS_DIR}/ResourceTrackingTests.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/guest_gpu/gpu_format.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/hostMemory.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ScalarProvenance.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/SrtWalker.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/SrtPatcher.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ResourceTracking.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ResourceMaterialization.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/ShaderInfoCollection.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/shader/recompiler/ir/BindingLayout.cpp"
 )
 target_link_libraries(resource_tracking_tests fmt::fmt)
 target_include_directories(resource_tracking_tests PRIVATE ${inc_headers})
 
 add_executable(resource_mutex_tests EXCLUDE_FROM_ALL
-	../tests/ResourceMutexTests.cpp
-	graphics/host_gpu/renderer/cache/resourceMutex.cpp
+	"${KYTY_TESTS_DIR}/ResourceMutexTests.cpp"
+	"${KYTY_SOURCE_DIR}/graphics/host_gpu/renderer/cache/resourceMutex.cpp"
 )
 target_link_libraries(resource_mutex_tests common)
 target_include_directories(resource_mutex_tests PRIVATE ${inc_headers})
 
 add_executable(audio_out2_port_tests EXCLUDE_FROM_ALL
-	../tests/AudioOut2PortTests.cpp
-	libs/libAudio2.cpp
-	loader/timer.cpp
+	"${KYTY_TESTS_DIR}/AudioOut2PortTests.cpp"
+	"${KYTY_SOURCE_DIR}/libs/libAudio2.cpp"
+	"${KYTY_SOURCE_DIR}/loader/timer.cpp"
 )
 target_link_libraries(audio_out2_port_tests common fmt::fmt)
 target_include_directories(audio_out2_port_tests PRIVATE ${inc_headers})
 
 add_executable(event_queue_lifetime_tests EXCLUDE_FROM_ALL
-	../tests/EventQueueLifetimeTests.cpp
-	kernel/eventQueue.cpp
-	loader/timer.cpp
+	"${KYTY_TESTS_DIR}/EventQueueLifetimeTests.cpp"
+	"${KYTY_SOURCE_DIR}/kernel/eventQueue.cpp"
+	"${KYTY_SOURCE_DIR}/loader/timer.cpp"
 )
 target_link_libraries(event_queue_lifetime_tests common fmt::fmt)
 target_include_directories(event_queue_lifetime_tests PRIVATE ${inc_headers})
 
 add_executable(image_page_table_tests EXCLUDE_FROM_ALL
-	../tests/ImagePageTableTests.cpp
+	"${KYTY_TESTS_DIR}/ImagePageTableTests.cpp"
 )
 target_link_libraries(image_page_table_tests common fmt::fmt)
 target_include_directories(image_page_table_tests PRIVATE ${inc_headers})
 
-add_kyty_full_emulator_test(shader_recompiler_compute_tests ../tests/ShaderRecompilerComputeTests.cpp)
+add_kyty_full_emulator_test(shader_recompiler_compute_tests "${KYTY_TESTS_DIR}/ShaderRecompilerComputeTests.cpp")
 
 set(gpu_test_generated_dir "${PROJECT_BINARY_DIR}/gpu_test_shaders")
 set(gpu_test_ms_depth_source
@@ -437,14 +439,14 @@ add_custom_command(
 	        -o "${gpu_test_ms_depth_spv}" "${gpu_test_ms_depth_source}"
 	COMMAND ${CMAKE_COMMAND} -DINPUT=${gpu_test_ms_depth_spv}
 	        -DOUTPUT=${gpu_test_ms_depth_header} -DSYMBOL=GPU_TEST_MS_DEPTH_SPV
-	        -P "${CMAKE_CURRENT_SOURCE_DIR}/embed_spirv.cmake"
+	        -P "${KYTY_SOURCE_DIR}/embed_spirv.cmake"
 	DEPENDS "${gpu_test_ms_depth_source}"
 	VERBATIM
 )
 target_sources(shader_recompiler_compute_tests PRIVATE
 	"${gpu_test_ms_depth_header}")
 
-add_kyty_full_emulator_test(virtual_memory_allocation_tests ../tests/VirtualMemoryAllocationTests.cpp)
+add_kyty_full_emulator_test(virtual_memory_allocation_tests "${KYTY_TESTS_DIR}/VirtualMemoryAllocationTests.cpp")
 target_compile_definitions(virtual_memory_allocation_tests PRIVATE
 	KYTY_VIRTUAL_MEMORY_ALLOCATION_TESTS=1)
 
@@ -520,7 +522,7 @@ if(BUILD_TESTING)
 endif()
 
 
-add_executable(kyty_emulator main.cpp ${kyty_emulator_src})
+add_executable(kyty_emulator "${KYTY_SOURCE_DIR}/main.cpp" ${kyty_emulator_src})
 configure_macos_guest_address_space(kyty_emulator)
 
 target_link_libraries(kyty_emulator ${kyty_emulator_link_libraries})
@@ -570,14 +572,14 @@ if(APPLE)
 	# reverts to the hardened defaults and aborts when it first executes written code).
 	add_custom_command(TARGET kyty_emulator POST_BUILD
 		COMMAND codesign -s - --force --options runtime
-			--entitlements "${CMAKE_CURRENT_SOURCE_DIR}/macos_jit.entitlements"
+			--entitlements "${KYTY_SOURCE_DIR}/macos_jit.entitlements"
 			$<TARGET_FILE:kyty_emulator>
 		COMMENT "Codesign kyty_emulator with JIT entitlements (macOS)")
 endif()
 
 install(TARGETS kyty_emulator DESTINATION .)
 if(KYTY_BUILD_LAUNCHER)
-	add_subdirectory(launcher)
+	add_subdirectory("${KYTY_SOURCE_DIR}/launcher" "${CMAKE_CURRENT_BINARY_DIR}/launcher")
 endif()
 if(KYTY_CLANG_CL)
 	install(FILES "${KYTY_THIRD_PARTY_DIR}/winpthread/bin/libwinpthread-1.dll" DESTINATION .)

README.md

@@ -141,7 +141,7 @@ git submodule update --init --recursive
 Configure the project. Replace the Qt path with the version installed on your system:
 
 ```powershell
-cmake -S src -B _Build/windows -G Ninja -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=clang-cl -DCMAKE_CXX_COMPILER=clang-cl -DCMAKE_PREFIX_PATH="C:/Qt/6.x.x/msvc2022_64"
+cmake -S . -B _Build/windows -G Ninja -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=clang-cl -DCMAKE_CXX_COMPILER=clang-cl -DCMAKE_PREFIX_PATH="C:/Qt/6.x.x/msvc2022_64"
 ```
 
 Build the launcher and stage a runnable installation:
@@ -174,7 +174,7 @@ Qt 6 (Concurrent, Network, Widgets) is also required — either the distribution
 ```bash
 git submodule update --init --recursive
 
-cmake -S src -B _Build/linux -G Ninja -DCMAKE_BUILD_TYPE=Release \
+cmake -S . -B _Build/linux -G Ninja -DCMAKE_BUILD_TYPE=Release \
   -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ \
   -DCMAKE_PREFIX_PATH="$Qt6_DIR"
 
@@ -188,7 +188,7 @@ The install step copies the Qt libraries and plugins next to the binaries, so
 As on Windows, the MSVC compiler is not used; Clang is required. `cl.exe` is rejected at configure
 time.
 
-Note that the CMake source root is `src`, not the repository root.
+The CMake source root is the repository root.
 
 ### Building on macOS
 
@@ -207,7 +207,7 @@ Requirements:
 ```bash
 git submodule update --init --recursive
 
-cmake -S src -B _Build/macos -G Ninja -DCMAKE_BUILD_TYPE=Release \
+cmake -S . -B _Build/macos -G Ninja -DCMAKE_BUILD_TYPE=Release \
   -DCMAKE_OSX_ARCHITECTURES=x86_64 \
   -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ \
   -DCMAKE_PREFIX_PATH="$Qt6_DIR"

src/common/CMakeLists.txt

@@ -13,7 +13,7 @@ endif()
 add_library(common STATIC ${common_src} ${common_headers})
 
 target_include_directories(common
-	PUBLIC ${PROJECT_SOURCE_DIR}
+	PUBLIC ${KYTY_SOURCE_DIR}
 	PRIVATE
 		${KYTY_THIRD_PARTY_DIR}/cpuinfo/include
 )

src/common/emulatorConfig.cpp

@@ -37,6 +37,10 @@ uint32_t GetVblankFrequency() {
 	return std::clamp(g_config->vblank_frequency, 30u, 360u);
 }
 
+uint32_t GetConsoleLanguage() {
+	return g_config->console_language;
+}
+
 bool VulkanValidationEnabled() {
 	return g_config->vulkan_validation_enabled;
 }
@@ -89,10 +93,6 @@ bool RenderDocEnabled() {
 	return g_config->renderdoc_enabled;
 }
 
-bool NggRectlistDrawEnabled() {
-	return g_config->ngg_rectlist_draw_enabled;
-}
-
 bool ReadbackLinearImagesEnabled() {
 	return g_config->readback_linear_images;
 }

src/common/emulatorConfig.h

@@ -18,10 +18,14 @@ enum class ProfilerDirection { None, Network };
 
 enum class OutputDirection { Silent, Console, File };
 
+constexpr uint32_t DEFAULT_CONSOLE_LANGUAGE = 1;
+constexpr uint32_t MAX_CONSOLE_LANGUAGE     = 29;
+
 struct ConfigOptions {
 	uint32_t               screen_width                = 1280;
 	uint32_t               screen_height               = 720;
 	uint32_t               vblank_frequency            = 60;
+	uint32_t               console_language            = DEFAULT_CONSOLE_LANGUAGE;
 	bool                   vulkan_validation_enabled   = false;
 	bool                   shader_validation_enabled   = false;
 	ShaderOptimizationType shader_optimization_type    = ShaderOptimizationType::None;
@@ -35,7 +39,6 @@ struct ConfigOptions {
 	ProfilerDirection      profiler_direction          = ProfilerDirection::None;
 	bool                   spirv_debug_printf_enabled  = false;
 	bool                   renderdoc_enabled           = false;
-	bool                   ngg_rectlist_draw_enabled   = true;
 	bool                   readback_linear_images      = false;
 };
 
@@ -44,6 +47,7 @@ void Load(const ConfigOptions& cfg);
 uint32_t GetScreenWidth();
 uint32_t GetScreenHeight();
 uint32_t GetVblankFrequency();
+uint32_t GetConsoleLanguage();
 bool     VulkanValidationEnabled();
 
 bool                   ShaderValidationEnabled();
@@ -64,7 +68,6 @@ ProfilerDirection GetProfilerDirection();
 bool SpirvDebugPrintfEnabled();
 
 bool RenderDocEnabled();
-bool NggRectlistDrawEnabled();
 bool ReadbackLinearImagesEnabled();
 
 } // namespace Config

src/graphics/guest_gpu/command_processor/commandProcessor.h

@@ -83,8 +83,7 @@ public:
 	               uint32_t first_instance = 0);
 	void DrawIndexOffset(uint32_t index_offset, uint32_t index_count, uint32_t flags);
 	void DrawIndexAuto(uint32_t index_count, uint32_t flags,
-	                   uint32_t render_target_slice_offset = 0, uint32_t instance_count = 1,
-	                   uint32_t first_vertex = 0, uint32_t first_instance = 0);
+	                   uint32_t render_target_slice_offset = 0);
 	void DrawIndirect(uint32_t data_offset, uint32_t draw_initiator, bool indexed);
 	void DrawIndirectMulti(uint32_t data_offset, uint32_t max_count_or_count,
 	                       const volatile uint32_t* count_addr, uint32_t stride_in_bytes,
@@ -152,6 +151,9 @@ private:
 	                      uint32_t interrupt_context_id);
 	void ProcessPm4(Pm4Execution& execution, size_t stop_depth);
 	void SuspendPm4();
+	void SubmitNonIndexedDraw(uint32_t vertex_count, uint32_t flags,
+	                          uint32_t render_target_slice_offset, uint32_t first_vertex,
+	                          uint32_t first_instance);
 
 	CommandScheduler&    GetScheduler() const { return m_renderer.GetCommandScheduler(); }
 	RenderCommandBuffer& CurrentBuffer() { return GetScheduler().Current(); }
@@ -168,7 +170,8 @@ private:
 	uint64_t         m_index_base_addr                  = 0;
 	uint64_t         m_draw_indirect_args_base_addr     = 0;
 	uint64_t         m_dispatch_indirect_args_base_addr = 0;
-	uint32_t         m_num_instances                    = 1;
+	// Persistent draw state: indirect draws update it for subsequent draws.
+	uint32_t m_num_instances = 1;
 
 	uint32_t m_de_count        = 0;
 	uint32_t m_ce_count        = 0;

src/graphics/guest_gpu/graphicsRun.cpp

@@ -32,11 +32,10 @@
 
 namespace Libs::Graphics {
 
-static thread_local CommandProcessor* g_current_processor      = nullptr;
-static thread_local Pm4Execution*     g_current_execution      = nullptr;
-static thread_local uint32_t          g_submission_pause_depth = 0;
-static thread_local bool              g_gpu_mutex_owned        = false;
-static thread_local bool              g_gpu_thread             = false;
+static thread_local CommandProcessor* g_current_processor = nullptr;
+static thread_local Pm4Execution*     g_current_execution = nullptr;
+static thread_local bool              g_gpu_mutex_owned   = false;
+static thread_local bool              g_gpu_thread        = false;
 
 class GpuMutexLock final {
 public:
@@ -98,8 +97,6 @@ public:
 	                   bool trigger_agc_interrupt_on_done);
 	void SubmitFlipPreparation(uint64_t request_id);
 	void Done();
-	void PauseSubmissions();
-	void ResumeSubmissions();
 	void Shutdown();
 	[[nodiscard]] bool IsStopping();
 	void               SendCommand(Common::UniqueFunction<void>&& command);
@@ -407,7 +404,7 @@ void CommandProcessor::WriteData(uint32_t* dst, const uint32_t* src, uint32_t dw
                                  uint32_t write_control) {
 	const uint32_t dst_sel      = ((write_control >> 30u) & 0x1u) | ((write_control >> 7u) & 0x1eu);
 	const uint32_t cache_policy = (write_control >> 25u) & 0x3u;
-	const uint32_t increment     = (write_control >> 16u) & 0x1u;
+	const uint32_t increment    = (write_control >> 16u) & 0x1u;
 	const uint32_t write_confirm = (write_control >> 20u) & 0x1u;
 
 	switch (dst_sel) {
@@ -696,26 +693,6 @@ bool GpuState::Process(Submission& submission) {
 	return complete;
 }
 
-void GpuState::PauseSubmissions() {
-	if (g_gpu_mutex_owned) {
-		EXIT("GPU submissions are already paused by this thread\n");
-	}
-	g_gpu_mutex_owned = true;
-	m_submission_mutex.Lock();
-	if (!IsGpuThread()) {
-		WaitLocked();
-	}
-	m_renderer.GetCommandScheduler().DrainPriorityOperations();
-}
-
-void GpuState::ResumeSubmissions() {
-	if (!g_gpu_mutex_owned) {
-		EXIT("GPU submissions resumed without an active pause\n");
-	}
-	m_submission_mutex.Unlock();
-	g_gpu_mutex_owned = false;
-}
-
 Pm4ProcessResult CommandProcessor::Process(Pm4Execution& execution, uint32_t* buffer,
                                            uint32_t size_dw) {
 	KYTY_PROFILER_BLOCK("CommandProcessor::Process");
@@ -1006,8 +983,9 @@ void CommandProcessor::DrawIndirect(uint32_t data_offset, uint32_t draw_initiato
 				     args.start_vertex_location, args.start_instance_location);
 			}
 		}
-		DrawIndexAuto(args.vertex_count_per_instance, 0, 0, args.instance_count,
-		              args.start_vertex_location, args.start_instance_location);
+		m_num_instances = args.instance_count;
+		SubmitNonIndexedDraw(args.vertex_count_per_instance, 0, 0, args.start_vertex_location,
+		                     args.start_instance_location);
 		return;
 	}
 
@@ -1047,6 +1025,7 @@ void CommandProcessor::DrawIndirect(uint32_t data_offset, uint32_t draw_initiato
 		}
 	}
 
+	m_num_instances = args.instance_count;
 	DrawIndex(index_count, index_addr, 0, 1, args.instance_count, nullptr, 0,
 	          static_cast<int32_t>(args.base_vertex_location), args.start_instance_location);
 }
@@ -1104,8 +1083,9 @@ void CommandProcessor::DrawIndirectMulti(uint32_t data_offset, uint32_t max_coun
 					     args->start_vertex_location, args->start_instance_location);
 				}
 			}
-			DrawIndexAuto(args->vertex_count_per_instance, 0, 0, args->instance_count,
-			              args->start_vertex_location, args->start_instance_location);
+			m_num_instances = args->instance_count;
+			SubmitNonIndexedDraw(args->vertex_count_per_instance, 0, 0, args->start_vertex_location,
+			                     args->start_instance_location);
 			continue;
 		}
 
@@ -1146,6 +1126,7 @@ void CommandProcessor::DrawIndirectMulti(uint32_t data_offset, uint32_t max_coun
 			}
 		}
 
+		m_num_instances = args->instance_count;
 		DrawIndex(index_count, index_addr, 0, 1, args->instance_count, nullptr, 0,
 		          static_cast<int32_t>(args->base_vertex_location), args->start_instance_location);
 	}
@@ -1237,12 +1218,17 @@ void CommandProcessor::DispatchIndirect(uint32_t data_offset, uint32_t mode) {
 }
 
 void CommandProcessor::DrawIndexAuto(uint32_t index_count, uint32_t flags,
-                                     uint32_t render_target_slice_offset, uint32_t instance_count,
-                                     uint32_t first_vertex, uint32_t first_instance) {
+                                     uint32_t render_target_slice_offset) {
+	SubmitNonIndexedDraw(index_count, flags, render_target_slice_offset, 0, 0);
+}
+
+void CommandProcessor::SubmitNonIndexedDraw(uint32_t vertex_count, uint32_t flags,
+                                            uint32_t render_target_slice_offset,
+                                            uint32_t first_vertex, uint32_t first_instance) {
 	CheckBuffer();
 
-	m_renderer.GetRenderExecutor().DrawAuto(m_submit_id, CurrentBuffer(), index_count, flags,
-	                                        render_target_slice_offset, instance_count,
+	m_renderer.GetRenderExecutor().DrawAuto(m_submit_id, CurrentBuffer(), vertex_count, flags,
+	                                        render_target_slice_offset, m_num_instances,
 	                                        first_vertex, first_instance);
 }
 
@@ -1693,32 +1679,6 @@ int Gpu::GetFrameNum() const {
 	return m_state->GetFrameNum();
 }
 
-void Gpu::PauseSubmissions() {
-	m_state->PauseSubmissions();
-}
-
-void Gpu::ResumeSubmissions() {
-	m_state->ResumeSubmissions();
-}
-
-Gpu::SubmissionLock::SubmissionLock(Gpu& gpu): m_gpu(gpu) {
-	if (g_current_processor != nullptr || g_submission_pause_depth == UINT32_MAX) {
-		EXIT("cannot acquire GPU submission lock in the current state\n");
-	}
-	if (g_submission_pause_depth++ == 0) {
-		m_gpu.PauseSubmissions();
-	}
-}
-
-Gpu::SubmissionLock::~SubmissionLock() {
-	if (g_submission_pause_depth == 0) {
-		EXIT("GPU submission lock released without ownership\n");
-	}
-	if (--g_submission_pause_depth == 0) {
-		m_gpu.ResumeSubmissions();
-	}
-}
-
 bool Gpu::IsCommandProcessorThread() noexcept {
 	return g_current_processor != nullptr;
 }
@@ -1727,12 +1687,4 @@ CommandProcessor* Gpu::CurrentCommandProcessor() noexcept {
 	return g_current_processor;
 }
 
-bool Gpu::SubmissionLockHeld() noexcept {
-	return g_submission_pause_depth != 0;
-}
-
-bool Gpu::MutexHeld() noexcept {
-	return g_gpu_mutex_owned;
-}
-
 } // namespace Libs::Graphics

src/graphics/guest_gpu/graphicsRun.h

@@ -35,25 +35,8 @@ public:
 
 	[[nodiscard]] static bool              IsCommandProcessorThread() noexcept;
 	[[nodiscard]] static CommandProcessor* CurrentCommandProcessor() noexcept;
-	[[nodiscard]] static bool              SubmissionLockHeld() noexcept;
-	[[nodiscard]] static bool              MutexHeld() noexcept;
-
-	class SubmissionLock final {
-	public:
-		explicit SubmissionLock(Gpu& gpu);
-		~SubmissionLock();
-		KYTY_CLASS_NO_COPY(SubmissionLock);
-
-	private:
-		Gpu& m_gpu;
-	};
 
 private:
-	friend class SubmissionLock;
-
-	void PauseSubmissions();
-	void ResumeSubmissions();
-
 	std::unique_ptr<GpuState> m_state;
 };
 } // namespace Libs::Graphics

src/graphics/host_gpu/renderer/cache/gpuResourceManager.cpp

@@ -7,7 +7,8 @@
 namespace Libs::Graphics {
 
 GpuResourceManager::GpuResourceManager(GraphicContext& graphics, CommandScheduler& scheduler)
-    : m_buffer_cache(graphics, scheduler, m_page_manager, m_texture_cache, m_resource_mutex),
+    : m_scheduler(scheduler),
+      m_buffer_cache(graphics, scheduler, m_page_manager, m_texture_cache, m_resource_mutex),
       m_texture_cache(graphics, scheduler, m_page_manager, m_buffer_cache, m_resource_mutex) {}
 
 GpuResourceManager::~GpuResourceManager() = default;
@@ -101,7 +102,22 @@ void GpuResourceManager::MapMemory(uint64_t vaddr, uint64_t size) {
 }
 
 void GpuResourceManager::UnmapMemory(uint64_t vaddr, uint64_t size) {
+	if (CommandScheduler::InDeferredOperation()) {
+		EXIT("unsupported memory unmap from an asynchronous GPU completion, "
+		     "addr=0x%016" PRIx64 " size=0x%016" PRIx64 "\n",
+		     vaddr, size);
+	}
+	if (m_resource_mutex.IsOwnedByCurrentThread()) {
+		EXIT("unsupported memory unmap from a pre-owned resource transaction, "
+		     "addr=0x%016" PRIx64 " size=0x%016" PRIx64 "\n",
+		     vaddr, size);
+	}
 	const auto unmap = [this, vaddr, size] {
+		if (m_scheduler.Active()) {
+			const auto tick = m_scheduler.CurrentTick();
+			m_scheduler.FinishCurrent();
+			m_scheduler.WaitPriorityOperations(tick);
+		}
 		m_buffer_cache.UnmapMemory(vaddr, size);
 		m_texture_cache.UnmapMemory(vaddr, size);
 		m_page_manager.OnGpuUnmap(vaddr, size);
@@ -109,13 +125,9 @@ void GpuResourceManager::UnmapMemory(uint64_t vaddr, uint64_t size) {
 		m_mapped_ranges.Subtract(vaddr, size);
 	};
 	if (m_gpu == nullptr) {
-		if (m_resource_mutex.IsOwnedByCurrentThread()) {
-			EXIT("cannot synchronously unmap from a resource transaction\n");
-		}
 		unmap();
 		return;
 	}
-	Gpu::SubmissionLock submissions(*m_gpu);
 	m_gpu->SendCommandSync(unmap);
 }
 

src/graphics/host_gpu/renderer/cache/gpuResourceManager.h

@@ -36,6 +36,7 @@ public:
 private:
 	PageManager               m_page_manager;
 	ResourceMutex             m_resource_mutex;
+	CommandScheduler&         m_scheduler;
 	BufferCache               m_buffer_cache;
 	TextureCache              m_texture_cache;
 	mutable std::shared_mutex m_mapped_ranges_mutex;

src/graphics/host_gpu/renderer/cache/textureCache.cpp

@@ -838,11 +838,20 @@ struct TextureCache::ColorTransferPlan {
 	TextureUploadLayout              layout;
 	std::vector<vk::BufferImageCopy> regions;
 	std::vector<GpuTileInfo>         tiles;
+	uint64_t                         linear_size = 0;
 	bool                             tiled       = false;
 	bool                             swap_bgra16 = false;
 	bool                             valid       = false;
 };
 
+static uint64_t GetLinearSize(std::span<const GpuTileInfo> tiles) {
+	uint64_t size = 0;
+	for (const auto& tile: tiles) {
+		size = std::max(size, tile.linear_offset + tile.linear_size);
+	}
+	return size;
+}
+
 struct TextureCache::DownloadPlan {
 	ColorTransferPlan color;
 	bool              depth = false;
@@ -919,6 +928,7 @@ TextureCache::BuildColorTransfer(const Image& image, BindingType binding,
 		                              info.resources.levels, plan.tiles)) {
 			return plan;
 		}
+		plan.linear_size = GetLinearSize(plan.tiles);
 	}
 	plan.valid = true;
 	return plan;
@@ -956,11 +966,19 @@ void TextureCache::UploadImage(Image& image, const ImageDesc& desc, Buffer& sour
 
 	if (desc.type != BindingType::DepthTarget) {
 		auto plan = BuildColorTransfer(image, desc.type, TransferDirection::Upload);
-		EXIT_NOT_IMPLEMENTED(!plan.valid);
+		if (!plan.valid) {
+			EXIT("TextureCache: invalid color upload: binding=%u addr=0x%016" PRIx64
+			     " size=0x%016" PRIx64 " format=%u tile=%u family=%u extent=%ux%ux%u "
+			     "pitch=%u levels=%u layers=%u samples=%u\n",
+			     static_cast<uint32_t>(desc.type), info.data.address, info.data.size,
+			     info.guest_format, info.tile_mode, static_cast<uint32_t>(plan.layout.tile_family),
+			     info.extent.width, info.extent.height, info.extent.depth, info.pitch,
+			     info.resources.levels, info.resources.layers, info.samples);
+		}
 		TileManager::Result linear {source.Handle(), source_offset, info.data.size};
 		if (plan.tiled) {
-			linear = m_tiler->Detile(source.Handle(), source_offset, info.data.size, info.data.size,
-			                         plan.tiles);
+			linear = m_tiler->Detile(source.Handle(), source_offset, info.data.size,
+			                         plan.linear_size, plan.tiles);
 		}
 		if (plan.swap_bgra16) {
 			linear = m_tiler->SwapBgra16(linear);
@@ -1566,7 +1584,7 @@ void TextureCache::DownloadImageData(Image& image, Buffer& destination, uint64_t
 	}
 
 	m_tiler->TileImage(image, color.regions, destination.Handle(), destination_offset,
-	                   destination_size, destination_size, color.tiles, transform);
+	                   destination_size, color.linear_size, color.tiles, transform);
 }
 
 bool BufferCache::SynchronizeBufferFromImage(Buffer& buffer, uint64_t vaddr, uint64_t size) {
@@ -1659,6 +1677,7 @@ bool BufferCache::SynchronizeBufferFromImage(Buffer& buffer, uint64_t vaddr, uin
 			                              image.info.TransferLayers(), levels, color.tiles)) {
 				return false;
 			}
+			color.linear_size = GetLinearSize(color.tiles);
 		}
 	}
 	m_texture_cache.DownloadImageData(image, buffer, buf_offset, copy_size, std::move(plan));

src/graphics/host_gpu/renderer/depthRenderTarget.cpp

@@ -96,9 +96,11 @@ void RenderExecutor::ResolveRenderDepthTarget(uint64_t submit_id, RenderCommandB
 	const auto& dc = hw.GetDepthControl();
 	const auto& sc = hw.GetStencilControl();
 	const auto& sm = hw.GetStencilMask();
-	const bool  depth_active =
+	const bool  has_stencil =
+	    z.stencil_info.format != Prospero::GpuEnumValue(Prospero::StencilFormat::kInvalid);
+	const bool depth_active =
 	    dc.z_enable || dc.z_write_enable || dc.depth_bounds_enable || rc.depth_clear_enable;
-	const bool stencil_active = dc.stencil_enable || rc.stencil_clear_enable;
+	const bool stencil_active = has_stencil && (dc.stencil_enable || rc.stencil_clear_enable);
 	if (!depth_active && !stencil_active) {
 		return;
 	}
@@ -139,8 +141,6 @@ void RenderExecutor::ResolveRenderDepthTarget(uint64_t submit_id, RenderCommandB
 		}
 		return;
 	}
-	const bool has_stencil =
-	    z.stencil_info.format != Prospero::GpuEnumValue(Prospero::StencilFormat::kInvalid);
 	const bool has_htile = z.z_info.htile_acceleration;
 	const auto samples   = render_sample_count(z.z_info.num_samples);
 	if (samples == 0) {
@@ -156,8 +156,8 @@ void RenderExecutor::ResolveRenderDepthTarget(uint64_t submit_id, RenderCommandB
 			DepthFatal("invalid depth view: base=%u last=%u", z.depth_view.slice_start,
 			           z.depth_view.slice_max);
 	}
-	if ((stencil_active && !has_stencil) || rc.resummarize_enable || rc.copy_centroid ||
-	    rc.copy_sample != 0 || z.z_info.expclear_enabled || z.stencil_info.expclear_enabled ||
+	if (rc.resummarize_enable || rc.copy_centroid || rc.copy_sample != 0 ||
+	    z.z_info.expclear_enabled || z.stencil_info.expclear_enabled ||
 	    z.z_info.partially_resident || z.stencil_info.partially_resident ||
 	    z.z_info.max_mip_level != 0 || z.depth_view.current_mip_level != 0 ||
 	    z.depth_info.addr5_swizzle_mask != 0 || z.depth_info.array_mode != 0 ||
@@ -279,10 +279,10 @@ void RenderExecutor::ResolveRenderDepthTarget(uint64_t submit_id, RenderCommandB
 	r.depth_min_bounds         = hw.GetDepthBoundsMin();
 	r.depth_max_bounds         = hw.GetDepthBoundsMax();
 
-	r.stencil_clear_enable = rc.stencil_clear_enable;
+	r.stencil_clear_enable = has_stencil && rc.stencil_clear_enable;
 	r.stencil_clear_value  = hw.GetStencilClearValue();
-	r.stencil_test_enable  = dc.stencil_enable;
-	if (dc.stencil_enable) {
+	r.stencil_test_enable  = has_stencil && dc.stencil_enable;
+	if (r.stencil_test_enable) {
 		if (dc.stencilfunc > static_cast<uint8_t>(vk::CompareOp::eAlways) ||
 		    (dc.backface_enable &&
 		     dc.stencilfunc_bf > static_cast<uint8_t>(vk::CompareOp::eAlways)) ||

src/graphics/host_gpu/renderer/image/textureCommon.cpp

@@ -482,10 +482,10 @@ static bool SetGpuTileSize(uint64_t offset, uint64_t length, uint64_t capacity,
 	return true;
 }
 
-bool TextureBuildGpuTileInfos(uint64_t size, const std::vector<vk::BufferImageCopy>& regions,
+bool TextureBuildGpuTileInfos(uint64_t tiled_size, const std::vector<vk::BufferImageCopy>& regions,
                               const TextureUploadLayout& layout, uint32_t fmt, uint32_t depth,
                               uint64_t levels, std::vector<GpuTileInfo>& out_infos) {
-	if (size == 0 || levels == 0 || levels > 16 || depth == 0 ||
+	if (tiled_size == 0 || levels == 0 || levels > 16 || depth == 0 ||
 	    regions.size() != GetTextureRegionCount(depth, levels, layout.volume_texture) ||
 	    Prospero::IsFmaskTextureFormat(fmt)) {
 		return false;
@@ -538,8 +538,9 @@ bool TextureBuildGpuTileInfos(uint64_t size, const std::vector<vk::BufferImageCo
 				const uint64_t linear_span =
 				    static_cast<uint64_t>(copy_depth - 1u) * linear_stride +
 				    layout.level_sizes[level].size;
-				if (!SetGpuTileSize(info.linear_offset, linear_span, size, info.linear_size) ||
-				    !SetGpuTileSize(info.tiled_offset, volume.level_sizes[level], size,
+				if (!SetGpuTileSize(info.linear_offset, linear_span, UINT64_MAX,
+				                    info.linear_size) ||
+				    !SetGpuTileSize(info.tiled_offset, volume.level_sizes[level], tiled_size,
 				                    info.tiled_size)) {
 					return false;
 				}
@@ -588,8 +589,9 @@ bool TextureBuildGpuTileInfos(uint64_t size, const std::vector<vk::BufferImageCo
 				info.bytes_per_element = block.bytes_per_element;
 				info.linear_offset     = region.bufferOffset;
 				info.tiled_offset      = TextureUploadSliceSourceOffset(layout, level, z);
-				if (!SetGpuTileSize(info.linear_offset, level_size.size, size, info.linear_size) ||
-				    !SetGpuTileSize(info.tiled_offset, GetLevelSrcSize(level_size), size,
+				if (!SetGpuTileSize(info.linear_offset, level_size.size, UINT64_MAX,
+				                    info.linear_size) ||
+				    !SetGpuTileSize(info.tiled_offset, GetLevelSrcSize(level_size), tiled_size,
 				                    info.tiled_size)) {
 					return false;
 				}

src/graphics/host_gpu/renderer/image/textureCommon.h

@@ -44,7 +44,7 @@ std::vector<vk::BufferImageCopy> TextureBuildImageCopies(const TextureUploadLayo
                                                          uint32_t width, uint32_t height,
                                                          uint32_t depth, uint64_t levels,
                                                          bool array_texture, bool volume_texture);
-bool TextureBuildGpuTileInfos(uint64_t size, const std::vector<vk::BufferImageCopy>& regions,
+bool TextureBuildGpuTileInfos(uint64_t tiled_size, const std::vector<vk::BufferImageCopy>& regions,
                               const TextureUploadLayout& layout, uint32_t fmt, uint32_t depth,
                               uint64_t levels, std::vector<GpuTileInfo>& infos);
 

src/graphics/host_gpu/renderer/pipeline/descriptors.cpp

@@ -391,9 +391,8 @@ static bool IsSupportedStorageTextureDescriptor(const ShaderRecompiler::IR::Imag
 	const bool supported_swizzle =
 	    IsValidImageSwizzle(swizzle) &&
 	    (swizzle == DstSel(4, 5, 6, 7) || !resource.read || resource.atomic);
-	const bool supported_mip_view = descriptor.BaseLevel() == 0 || is_1d || is_2d;
 	return (is_1d || is_1d_array || is_2d || is_2d_array || is_3d) && supported_tile &&
-	       supported_mip_view && descriptor.BaseLevel() == descriptor.LastLevel() &&
+	       descriptor.BaseLevel() == descriptor.LastLevel() &&
 	       descriptor.LastLevel() <= descriptor.MaxMip() && descriptor.MinLod() == 0 &&
 	       supported_swizzle && descriptor.BCSwizzle() == 0 && !descriptor.MsaaDepth();
 }
@@ -618,14 +617,15 @@ RenderExecutor::ResolveTexture(const ShaderRecompiler::IR::ImageResource&   reso
 	const bool multisampled = IsMultisampledTexture(type);
 	const auto levels       = multisampled ? 1u : static_cast<uint32_t>(descriptor.MaxMip()) + 1u;
 	const auto tile         = descriptor.TileMode();
-	const bool depth_tile = tile == Prospero::GpuEnumValue(Prospero::TileMode::kDepth);
+	const bool depth_tile   = tile == Prospero::GpuEnumValue(Prospero::TileMode::kDepth);
 	const bool msaa_tile =
 	    depth_tile || tile == Prospero::GpuEnumValue(Prospero::TileMode::kRenderTarget);
 	const bool msaa_array = type == Prospero::ImageType::kColor2DMsaaArray;
 	if ((!multisampled && (base_level > last_level || last_level >= levels)) ||
 	    (multisampled &&
 	     (base_level != 0 || last_level == 0 || last_level > 3 ||
-	      descriptor.MaxMip() != last_level || !msaa_tile || (descriptor.MsaaDepth() && !depth_tile) ||
+	      descriptor.MaxMip() != last_level || !msaa_tile ||
+	      (descriptor.MsaaDepth() && !depth_tile) ||
 	      (!msaa_array && (descriptor.Depth() != 0 || descriptor.BaseArray5() != 0))))) {
 		EXIT("unsupported texture mip view: base=%u last=%u levels=%u max=%u type=%u tile=%u "
 		     "kind=%u dimension=%u mip_mode=%u read=%d written=%d "
@@ -634,7 +634,8 @@ RenderExecutor::ResolveTexture(const ShaderRecompiler::IR::ImageResource&   reso
 		     static_cast<uint32_t>(resource.kind), static_cast<uint32_t>(resource.dimension),
 		     static_cast<uint32_t>(resource.mip_mode), resource.read, resource.written,
 		     descriptor.fields[0], descriptor.fields[1], descriptor.fields[2], descriptor.fields[3],
-		     descriptor.fields[4], descriptor.fields[5], descriptor.fields[6], descriptor.fields[7]);
+		     descriptor.fields[4], descriptor.fields[5], descriptor.fields[6],
+		     descriptor.fields[7]);
 	}
 	const auto samples = multisampled ? 1u << last_level : 1u;
 	const auto view_levels =
@@ -661,8 +662,8 @@ RenderExecutor::ResolveTexture(const ShaderRecompiler::IR::ImageResource&   reso
 	TileSizeAlign size {};
 	if (multisampled) {
 		const auto bytes = Prospero::NumBytesPerElement(format);
-		pitch = depth_tile ? TileGetDepthPitch(width, bytes, last_level)
-		                   : TileGetRenderTargetPitch(width, bytes, last_level);
+		pitch            = depth_tile ? TileGetDepthPitch(width, bytes, last_level)
+		                              : TileGetRenderTargetPitch(width, bytes, last_level);
 		if (pitch == 0 || !TileGetRenderTargetSize(width, height, pitch, bytes, size, last_level) ||
 		    size.size > UINT32_MAX / image_layers) {
 			EXIT("unsupported multisample texture layout\n");
@@ -679,8 +680,8 @@ RenderExecutor::ResolveTexture(const ShaderRecompiler::IR::ImageResource&   reso
 		ValidateStorageTexture(resource, descriptor, size.size);
 	}
 
-	const auto              pixel_format = TextureGetFormat(format);
-	const auto              storage_view_format =
+	const auto pixel_format = TextureGetFormat(format);
+	const auto storage_view_format =
 	    storage && format == Prospero::GpuEnumValue(Prospero::BufferFormat::k32SInt)
 	        ? vk::Format::eR32Uint
 	        : SrgbStorageViewFormat(pixel_format);

src/graphics/host_gpu/renderer/pipeline/pipelineCache.cpp

@@ -154,8 +154,9 @@ PipelineCache::GraphicsPipeline& PipelineCache::CreateGraphicsPipeline(
 	for (uint32_t i = 0; i < RENDER_COLOR_ATTACHMENTS_MAX; i++) {
 		static_params.color_mask[i] = color_mask[i];
 	}
-	static_params.cull_back  = mc.cull_back;
-	static_params.cull_front = mc.cull_front;
+	const bool rect_list     = topology == vk::PrimitiveTopology::ePatchList;
+	static_params.cull_back  = !rect_list && mc.cull_back;
+	static_params.cull_front = !rect_list && mc.cull_front;
 	static_params.face       = mc.face;
 
 	for (uint32_t i = 0; i < color_count; i++) {

src/graphics/host_gpu/renderer/pipeline/shaders.cpp

@@ -14,6 +14,7 @@
 #include "graphics/host_gpu/renderer/renderTarget.h"
 #include "graphics/host_gpu/vulkanCommon.h"
 #include "graphics/shader/recompiler/ir/ShaderIR.h"
+#include "graphics/shader/rectListShader.h"
 #include "graphics/shader/shader.h"
 
 #include <algorithm>
@@ -463,8 +464,12 @@ void CreatePipelineInternal(
     uint32_t ps_hash0, uint32_t ps_crc32, bool ps_active) {
 	EXIT_IF(ps_active && ps_input_info == nullptr);
 
-	vk::ShaderModule vert_shader_module = nullptr;
-	vk::ShaderModule frag_shader_module = nullptr;
+	const bool rect_list = static_params.topology == vk::PrimitiveTopology::ePatchList;
+
+	vk::ShaderModule vert_shader_module         = nullptr;
+	vk::ShaderModule tess_control_shader_module = nullptr;
+	vk::ShaderModule tess_eval_shader_module    = nullptr;
+	vk::ShaderModule frag_shader_module         = nullptr;
 
 	vk::ShaderModuleCreateInfo create_info {};
 
@@ -491,8 +496,33 @@ void CreatePipelineInternal(
 		}
 		EXIT_NOT_IMPLEMENTED(result != vk::Result::eSuccess);
 	}
+	if (rect_list) {
+		const auto shaders =
+		    BuildRectListShaders(vs_input_info, ps_active ? ps_input_info : nullptr);
+		create_info.codeSize = shaders.control.size() * 4;
+		create_info.pCode    = shaders.control.data();
+		result =
+		    graphics.device.createShaderModule(&create_info, nullptr, &tess_control_shader_module);
+		if (graphics_debug_dump_enabled()) {
+			LOGF("PipelineTrace: vkCreateShaderModule RectList TCS done result=%s module=%p\n",
+			     VulkanToString(result).c_str(), static_cast<void*>(tess_control_shader_module));
+		}
+		EXIT_NOT_IMPLEMENTED(result != vk::Result::eSuccess);
+
+		create_info.codeSize = shaders.evaluation.size() * 4;
+		create_info.pCode    = shaders.evaluation.data();
+		result =
+		    graphics.device.createShaderModule(&create_info, nullptr, &tess_eval_shader_module);
+		if (graphics_debug_dump_enabled()) {
+			LOGF("PipelineTrace: vkCreateShaderModule RectList TES done result=%s module=%p\n",
+			     VulkanToString(result).c_str(), static_cast<void*>(tess_eval_shader_module));
+		}
+		EXIT_NOT_IMPLEMENTED(result != vk::Result::eSuccess);
+	}
 
 	EXIT_NOT_IMPLEMENTED(vert_shader_module == nullptr);
+	EXIT_NOT_IMPLEMENTED(
+	    rect_list && (tess_control_shader_module == nullptr || tess_eval_shader_module == nullptr));
 	EXIT_NOT_IMPLEMENTED(ps_active && frag_shader_module == nullptr);
 
 	vk::PipelineShaderStageCreateInfo                     vert_shader_stage_info {};
@@ -525,9 +555,28 @@ void CreatePipelineInternal(
 		                      frag_shader_stage_info);
 	}
 
-	vk::PipelineShaderStageCreateInfo shader_stages[]    = {vert_shader_stage_info,
-	                                                        frag_shader_stage_info};
-	const uint32_t                    shader_stage_count = ps_active ? 2u : 1u;
+	vk::PipelineShaderStageCreateInfo tess_control_shader_stage_info {};
+	tess_control_shader_stage_info.sType  = vk::StructureType::ePipelineShaderStageCreateInfo;
+	tess_control_shader_stage_info.stage  = vk::ShaderStageFlagBits::eTessellationControl;
+	tess_control_shader_stage_info.module = tess_control_shader_module;
+	tess_control_shader_stage_info.pName  = "main";
+
+	vk::PipelineShaderStageCreateInfo tess_eval_shader_stage_info {};
+	tess_eval_shader_stage_info.sType  = vk::StructureType::ePipelineShaderStageCreateInfo;
+	tess_eval_shader_stage_info.stage  = vk::ShaderStageFlagBits::eTessellationEvaluation;
+	tess_eval_shader_stage_info.module = tess_eval_shader_module;
+	tess_eval_shader_stage_info.pName  = "main";
+
+	vk::PipelineShaderStageCreateInfo shader_stages[4]   = {};
+	uint32_t                          shader_stage_count = 0;
+	shader_stages[shader_stage_count++]                  = vert_shader_stage_info;
+	if (rect_list) {
+		shader_stages[shader_stage_count++] = tess_control_shader_stage_info;
+		shader_stages[shader_stage_count++] = tess_eval_shader_stage_info;
+	}
+	if (ps_active) {
+		shader_stages[shader_stage_count++] = frag_shader_stage_info;
+	}
 
 	vk::VertexInputAttributeDescription input_attr[ShaderVertexInputInfo::RES_MAX];
 	vk::VertexInputBindingDescription   input_desc[ShaderVertexInputInfo::RES_MAX];
@@ -929,10 +978,13 @@ void CreatePipelineInternal(
 	pipeline_info.pStages                  = shader_stages;
 	pipeline_info.pVertexInputState        = &vertex_input_info;
 	pipeline_info.pInputAssemblyState      = &input_assembly;
-	pipeline_info.pTessellationState       = nullptr;
-	pipeline_info.pViewportState           = &viewport_state;
-	pipeline_info.pRasterizationState      = &rasterizer;
-	pipeline_info.pMultisampleState        = &multisampling;
+	vk::PipelineTessellationStateCreateInfo tessellation_state {};
+	tessellation_state.sType              = vk::StructureType::ePipelineTessellationStateCreateInfo;
+	tessellation_state.patchControlPoints = 3;
+	pipeline_info.pTessellationState      = (rect_list ? &tessellation_state : nullptr);
+	pipeline_info.pViewportState          = &viewport_state;
+	pipeline_info.pRasterizationState     = &rasterizer;
+	pipeline_info.pMultisampleState       = &multisampling;
 	pipeline_info.pDepthStencilState = (static_params.with_depth ? &depth_stencil_info : nullptr);
 	pipeline_info.pColorBlendState   = &color_blending;
 	pipeline_info.pDynamicState      = &dynamic_state;
@@ -968,6 +1020,12 @@ void CreatePipelineInternal(
 	if (frag_shader_module != nullptr) {
 		graphics.device.destroyShaderModule(frag_shader_module, nullptr);
 	}
+	if (tess_control_shader_module != nullptr) {
+		graphics.device.destroyShaderModule(tess_control_shader_module, nullptr);
+	}
+	if (tess_eval_shader_module != nullptr) {
+		graphics.device.destroyShaderModule(tess_eval_shader_module, nullptr);
+	}
 	graphics.device.destroyShaderModule(vert_shader_module, nullptr);
 }
 

src/graphics/host_gpu/renderer/renderDraw.cpp

@@ -2,7 +2,6 @@
 
 #include "common/assert.h"
 #include "common/common.h"
-#include "common/emulatorConfig.h"
 #include "common/file.h"
 #include "common/logging/log.h"
 #include "common/profiler.h"
@@ -650,11 +649,9 @@ static bool ConsumeMetadataColorOperation(const RenderCommandBuffer& buffer) {
 }
 
 struct DrawEmitInfo {
-	bool     indexed           = false;
-	bool     draw_prim7_as_ngg = false;
-	uint32_t draw_vertex_count = 0;
-	int32_t  vertex_offset     = 0;
-	uint32_t first_vertex      = 0;
+	bool     indexed       = false;
+	int32_t  vertex_offset = 0;
+	uint32_t first_vertex  = 0;
 };
 
 struct DrawIndexBufferSource {
@@ -767,7 +764,7 @@ static void SetDrawDebugPhase(RenderCommandBuffer& buffer, uint64_t submit_id,
 	                    draw.flags, draw.instance_count, draw.first_instance);
 }
 
-static bool GetDrawTopology(const HW::UserConfig& ucfg, bool auto_draw, bool use_ngg_rectlist_draw,
+static bool GetDrawTopology(const HW::UserConfig& ucfg, bool auto_draw,
                             vk::PrimitiveTopology& topology) {
 
 	topology = vk::PrimitiveTopology::ePointList;
@@ -791,8 +788,7 @@ static bool GetDrawTopology(const HW::UserConfig& ucfg, bool auto_draw, bool use
 			topology = vk::PrimitiveTopology::eTriangleStrip;
 			break;
 		case Prospero::PrimitiveType::kRectList:
-			topology = (auto_draw && use_ngg_rectlist_draw ? vk::PrimitiveTopology::eTriangleStrip
-			                                               : vk::PrimitiveTopology::eTriangleList);
+			topology = vk::PrimitiveTopology::ePatchList;
 			break;
 		case Prospero::PrimitiveType::kRectListLegacy:
 			if (!auto_draw) {
@@ -991,20 +987,6 @@ static void LogDrawStateIfNeeded(const RenderCommandBuffer& buffer, const DrawCa
 	// LogDrawTextureState(draw.name, state.color_info[0], state.ps_input_info);
 }
 
-static bool IsHostExpandedRectListDrawSupported(const ShaderVertexInputInfo& vs_input_info,
-                                                const DrawCallInfo&          draw,
-                                                const DrawEmitInfo&          emit) {
-	if (!emit.draw_prim7_as_ngg) {
-		return true;
-	}
-
-	if (vs_input_info.buffers_num != 0) {
-		return false;
-	}
-
-	return draw.index_count == 3 || draw.index_count == emit.draw_vertex_count;
-}
-
 static void EmitDrawPrimitives(const HW::UserConfig& ucfg, vk::CommandBuffer vk_buffer,
                                const ShaderVertexInputInfo& vs_input_info, const DrawCallInfo& draw,
                                const DrawEmitInfo& emit) {
@@ -1017,22 +999,12 @@ static void EmitDrawPrimitives(const HW::UserConfig& ucfg, vk::CommandBuffer vk_
 		case Prospero::PrimitiveType::kTriList:
 		case Prospero::PrimitiveType::kTriFan:
 		case Prospero::PrimitiveType::kTriStrip:
-			if (emit.indexed) {
-				vk_buffer.drawIndexed(draw.index_count, draw.instance_count, 0, emit.vertex_offset,
-				                      draw.first_instance);
-			} else {
-				vk_buffer.draw(draw.index_count, draw.instance_count, emit.first_vertex,
-				               draw.first_instance);
-			}
-			break;
 		case Prospero::PrimitiveType::kRectList:
 			if (emit.indexed) {
 				vk_buffer.drawIndexed(draw.index_count, draw.instance_count, 0, emit.vertex_offset,
 				                      draw.first_instance);
 			} else {
-				EXIT_NOT_IMPLEMENTED(
-				    !IsHostExpandedRectListDrawSupported(vs_input_info, draw, emit));
-				vk_buffer.draw(emit.draw_vertex_count, draw.instance_count, emit.first_vertex,
+				vk_buffer.draw(draw.index_count, draw.instance_count, emit.first_vertex,
 				               draw.first_instance);
 			}
 			break;
@@ -1159,7 +1131,7 @@ void RenderExecutor::DrawIndex(uint64_t submit_id, RenderCommandBuffer& buffer,
 	                    reinterpret_cast<uint64_t>(index_addr));
 
 	Common::LockGuard lock(m_context.GetMutex());
-	if (index_count == 0) {
+	if (index_count == 0 || instance_count == 0) {
 		return;
 	}
 
@@ -1201,7 +1173,7 @@ void RenderExecutor::DrawIndex(uint64_t submit_id, RenderCommandBuffer& buffer,
 	hw_check(buffer);
 
 	vk::PrimitiveTopology topology = vk::PrimitiveTopology::ePointList;
-	if (!GetDrawTopology(ucfg, false, false, topology)) {
+	if (!GetDrawTopology(ucfg, false, topology)) {
 		return;
 	}
 
@@ -1229,10 +1201,6 @@ void RenderExecutor::DrawIndex(uint64_t submit_id, RenderCommandBuffer& buffer,
 
 	EXIT_NOT_IMPLEMENTED(flags != 0);
 	EXIT_NOT_IMPLEMENTED(type != 1);
-	if (instance_count == 0) {
-		instance_count = 1;
-	}
-
 	const DrawCallInfo    draw {"DrawIndex",    CommandBufferDebugOp::DrawIndex,
 	                            index_count,    flags,
 	                            instance_count, first_instance};
@@ -1292,7 +1260,7 @@ void RenderExecutor::DrawAuto(uint64_t submit_id, RenderCommandBuffer& buffer, u
 	                    index_count, flags, first_vertex, instance_count, first_instance);
 
 	Common::LockGuard lock(m_context.GetMutex());
-	if (index_count == 0) {
+	if (index_count == 0 || instance_count == 0) {
 		return;
 	}
 
@@ -1330,10 +1298,6 @@ void RenderExecutor::DrawAuto(uint64_t submit_id, RenderCommandBuffer& buffer, u
 	hw_check(buffer);
 
 	EXIT_NOT_IMPLEMENTED(flags != 0);
-	if (instance_count == 0) {
-		instance_count = 1;
-	}
-
 	const DrawCallInfo draw {"DrawIndexAuto", CommandBufferDebugOp::DrawIndexAuto,
 	                         index_count,     flags,
 	                         instance_count,  first_instance};
@@ -1345,20 +1309,15 @@ void RenderExecutor::DrawAuto(uint64_t submit_id, RenderCommandBuffer& buffer, u
 		return;
 	}
 
-	vk::PrimitiveTopology topology              = vk::PrimitiveTopology::ePointList;
-	const bool            use_ngg_rectlist_draw = Config::NggRectlistDrawEnabled();
-
-	if (!GetDrawTopology(ucfg, true, use_ngg_rectlist_draw, topology)) {
+	vk::PrimitiveTopology topology = vk::PrimitiveTopology::ePointList;
+	if (!GetDrawTopology(ucfg, true, topology)) {
 		ResetBindings();
 		return;
 	}
-	const bool draw_prim7_as_ngg =
-	    (use_ngg_rectlist_draw &&
-	     ucfg.GetPrimType() == Prospero::GpuEnumValue(Prospero::PrimitiveType::kRectList));
-
 	RefreshShaders(buffer, draw, false, state);
 
-	if (draw_prim7_as_ngg && state.vs_input_info.buffers_num == 0 &&
+	const bool rect_list = topology == vk::PrimitiveTopology::ePatchList;
+	if (rect_list && state.vs_input_info.buffers_num == 0 &&
 	    state.vs_input_info.param_export_mask == 0 && state.ps_input_info.input_num != 0) {
 		if (graphics_debug_dump_enabled()) {
 			LOGF("DrawIndexAuto: skipping rect-list draw with no VS param exports and PS inputs: "
@@ -1375,13 +1334,10 @@ void RenderExecutor::DrawAuto(uint64_t submit_id, RenderCommandBuffer& buffer, u
 	                         Prospero::GpuEnumValue(Prospero::PrimitiveType::kRectListLegacy),
 	                     0, nullptr);
 
-	const uint32_t draw_vertex_count = (draw_prim7_as_ngg ? 4u : index_count);
-	const auto     vertex_offset = ResolveVertexOffset(ucfg.GetIndexOffset(), state.vs_input_info) +
-	                               static_cast<int32_t>(first_vertex);
-	DrawEmitInfo   emit {};
-	emit.draw_prim7_as_ngg = draw_prim7_as_ngg;
-	emit.draw_vertex_count = draw_vertex_count;
-	emit.first_vertex      = static_cast<uint32_t>(vertex_offset);
+	const auto   vertex_offset = ResolveVertexOffset(ucfg.GetIndexOffset(), state.vs_input_info) +
+	                             static_cast<int32_t>(first_vertex);
+	DrawEmitInfo emit {};
+	emit.first_vertex = static_cast<uint32_t>(vertex_offset);
 
 	DrawIndexBufferSource index_source {};
 	ExecutePreparedDraw(submit_id, buffer, draw, state, topology, emit, index_source, false, false,

src/graphics/shader/recompiler/emitter/SpirvBuilder.cpp

@@ -19,7 +19,7 @@ static void AppendInstructionWords(std::vector<uint32_t>& section, const uint32_
 	section.insert(section.end(), words + 1, words + words_num);
 }
 
-Builder::Builder() {
+Builder::Builder(uint32_t version): m_version(version) {
 	m_debug.reserve(InitialSpirvSectionReserve);
 	m_annotations.reserve(InitialSpirvSectionReserve);
 	m_types.reserve(InitialSpirvSectionReserve);
@@ -138,7 +138,7 @@ std::vector<uint32_t> Builder::Build() const {
 	               m_debug.size() + m_annotations.size() + m_types.size() + m_functions.size());
 
 	module.push_back(0x07230203u);
-	module.push_back(0x00010300u);
+	module.push_back(m_version);
 	module.push_back(0u);
 	module.push_back(m_next_id);
 	module.push_back(0u);

src/graphics/shader/recompiler/emitter/SpirvBuilder.h

@@ -10,7 +10,7 @@ namespace Libs::Graphics::ShaderRecompiler::Spirv {
 
 class Builder {
 public:
-	Builder();
+	explicit Builder(uint32_t version = 0x00010300u);
 	~Builder() = default;
 	KYTY_CLASS_DEFAULT_COPY(Builder);
 
@@ -39,6 +39,7 @@ private:
 	static void AppendString(std::vector<uint32_t>& words, const char* text);
 
 	uint32_t              m_next_id = 1;
+	uint32_t              m_version = 0;
 	std::vector<uint32_t> m_capabilities;
 	std::vector<uint32_t> m_extensions;
 	std::vector<uint32_t> m_ext_inst_imports;

src/graphics/shader/recompiler/emitter/spirvEmitterAnalysis.cpp

@@ -7,51 +7,30 @@ namespace Libs::Graphics::ShaderRecompiler::Spirv::Emitter {
 
 uint32_t PixelParameterMappedLocation(const EmitterState& state, uint32_t attr) {
 	const auto* ps = state.pixel_input_info;
-	if (state.stage != ShaderType::Pixel || ps == nullptr || attr >= ps->input_num) {
+	if (state.stage != ShaderType::Pixel || ps == nullptr) {
 		return attr;
 	}
-	// VINTRP ATTR selects the PS input slot. SPI_PS_INPUT_CNTL maps that slot to a
-	// VS parameter export, which is the SPIR-V location we must link against.
-	return ps->interpolator_settings[attr] & PsInputOffsetMask;
+	return ShaderPixelParameterMappedLocation(*ps, attr);
 }
 
 uint32_t PixelParameterLocation(const EmitterState& state, uint32_t attr) {
-	bool used_locations[32] = {};
-
+	std::array<uint32_t, 32> active_inputs {};
+	uint32_t                 active_count = 0;
 	for (const auto& input: state.inputs) {
-		if (input.kind != IR::StageInputKind::Parameter) {
-			continue;
-		}
-
-		auto location = PixelParameterMappedLocation(state, input.location);
-		if (location < std::size(used_locations) && used_locations[location]) {
-			auto fallback_location = input.location;
-			while (fallback_location < std::size(used_locations) &&
-			       used_locations[fallback_location]) {
-				fallback_location++;
-			}
-			EXIT_NOT_IMPLEMENTED(fallback_location >= std::size(used_locations));
-			location = fallback_location;
-		}
-
-		if (input.location == attr) {
-			return location;
-		}
-
-		if (location < std::size(used_locations)) {
-			used_locations[location] = true;
+		if (input.kind == IR::StageInputKind::Parameter) {
+			active_inputs[active_count++] = input.location;
 		}
 	}
-
-	return PixelParameterMappedLocation(state, attr);
+	return state.stage == ShaderType::Pixel && state.pixel_input_info != nullptr
+	           ? ShaderPixelParameterLocation(*state.pixel_input_info,
+	                                          {active_inputs.data(), active_count}, attr)
+	           : attr;
 }
 
 bool PixelParameterIsFlat(const EmitterState& state, uint32_t attr) {
 	const auto* ps = state.pixel_input_info;
-	if (state.stage != ShaderType::Pixel || ps == nullptr || attr >= ps->input_num) {
-		return false;
-	}
-	return (ps->interpolator_settings[attr] & PsInputFlatShade) != 0;
+	return state.stage == ShaderType::Pixel && ps != nullptr &&
+	       ShaderPixelParameterIsFlat(*ps, attr);
 }
 
 void SetError(std::string* error, const char* message) {

src/graphics/shader/recompiler/emitter/spirvEmitterInternal.h

@@ -425,9 +425,6 @@ struct EmitterState {
 	std::map<uint32_t, uint32_t>                     float_constants;
 };
 
-constexpr uint32_t PsInputOffsetMask = 0x0000001fu;
-constexpr uint32_t PsInputFlatShade  = 0x00000400u;
-
 enum class VertexInputScalarKind { Float, Sint, Uint };
 
 constexpr uint32_t NoImageComponent = 0xffffffffu;

src/graphics/shader/rectListShader.cpp

@@ -0,0 +1,397 @@
+#include "graphics/shader/rectListShader.h"
+
+#include "common/assert.h"
+#include "graphics/shader/recompiler/emitter/SpirvBuilder.h"
+#include "graphics/shader/recompiler/ir/ShaderIR.h"
+#include "graphics/shader/shader.h"
+#include "spirv/unified1/spirv.hpp11"
+
+#include <array>
+#include <bit>
+#include <cstdint>
+#include <utility>
+
+namespace Libs::Graphics {
+
+namespace {
+
+using ShaderRecompiler::Spirv::Builder;
+
+constexpr uint32_t SpirvVersion15 = 0x00010500u;
+
+template <typename T>
+constexpr uint32_t Word(T value) {
+	return static_cast<uint32_t>(value);
+}
+
+struct Parameter {
+	uint32_t input_location  = 0;
+	uint32_t output_location = 0;
+	bool     flat            = false;
+};
+
+std::vector<Parameter> GetParameters(const ShaderVertexInputInfo& vertex_info,
+                                     const ShaderPixelInputInfo*  pixel_info) {
+	if (pixel_info == nullptr) {
+		return {};
+	}
+	EXIT_IF(pixel_info->input_num > ShaderVertexInputInfo::RES_MAX);
+	EXIT_IF(pixel_info->stage.program == nullptr);
+
+	std::vector<uint32_t> active_inputs;
+	for (const auto& input: pixel_info->stage.program->info.inputs) {
+		if (input.kind == ShaderRecompiler::IR::StageInputKind::Parameter) {
+			active_inputs.push_back(input.location);
+		}
+	}
+
+	std::vector<Parameter> parameters;
+	for (const auto input: active_inputs) {
+		const auto input_location = ShaderPixelParameterMappedLocation(*pixel_info, input);
+		if ((vertex_info.param_export_mask & (1u << input_location)) != 0) {
+			parameters.push_back({input_location,
+			                      ShaderPixelParameterLocation(*pixel_info, active_inputs, input),
+			                      ShaderPixelParameterIsFlat(*pixel_info, input)});
+		}
+	}
+	return parameters;
+}
+
+class RectListEmitter {
+public:
+	RectListEmitter(const std::vector<Parameter>& parameters_, spv::ExecutionModel model)
+	    : parameters(parameters_) {
+		builder.AddMemoryModel(
+		    {Word(spv::AddressingModel::Logical), Word(spv::MemoryModel::GLSL450)});
+
+		void_type       = Type(spv::Op::OpTypeVoid);
+		uint_type       = Type(spv::Op::OpTypeInt, 32u, 0u);
+		int_type        = Type(spv::Op::OpTypeInt, 32u, 1u);
+		float_type      = Type(spv::Op::OpTypeFloat, 32u);
+		vec4_float_type = Type(spv::Op::OpTypeVector, float_type, 4u);
+		function_type   = Type(spv::Op::OpTypeFunction, void_type);
+
+		per_vertex_type = Type(spv::Op::OpTypeStruct, vec4_float_type);
+		builder.AddAnnotation({Word(spv::Op::OpMemberDecorate), per_vertex_type, 0u,
+		                       Word(spv::Decoration::BuiltIn), Word(spv::BuiltIn::Position)});
+		builder.AddAnnotation(
+		    {Word(spv::Op::OpDecorate), per_vertex_type, Word(spv::Decoration::Block)});
+
+		ptr_input_vec4_float  = Pointer(spv::StorageClass::Input, vec4_float_type);
+		ptr_output_vec4_float = Pointer(spv::StorageClass::Output, vec4_float_type);
+		if (model == spv::ExecutionModel::TessellationControl) {
+			bool_type        = Type(spv::Op::OpTypeBool);
+			vec2_bool_type   = Type(spv::Op::OpTypeVector, bool_type, 2u);
+			vec2_float_type  = Type(spv::Op::OpTypeVector, float_type, 2u);
+			ptr_output_float = Pointer(spv::StorageClass::Output, float_type);
+		} else {
+			vec3_float_type = Type(spv::Op::OpTypeVector, float_type, 3u);
+			ptr_input_float = Pointer(spv::StorageClass::Input, float_type);
+		}
+	}
+
+	std::vector<uint32_t> EmitControl() {
+		DefineEntry(spv::ExecutionModel::TessellationControl);
+		const auto float_one = Constant(float_type, std::bit_cast<uint32_t>(1.0f));
+
+		for (uint32_t i = 0; i < 4; i++) {
+			Store(Access(ptr_output_float, tess_outer, Int(i)), float_one);
+		}
+		for (uint32_t i = 0; i < 2; i++) {
+			Store(Access(ptr_output_float, tess_inner, Int(i)), float_one);
+		}
+
+		std::array<uint32_t, 3> positions {};
+		for (uint32_t i = 0; i < positions.size(); i++) {
+			positions[i] =
+			    Load(vec4_float_type, Access(ptr_input_vec4_float, gl_in, Int(i), Int(0)));
+		}
+
+		std::array<uint32_t, 3> coordinate_equal {};
+		for (uint32_t i = 0; i < coordinate_equal.size(); i++) {
+			const auto left  = Result(spv::Op::OpVectorShuffle, vec2_float_type, positions[i],
+			                          positions[i], 0u, 1u);
+			const auto right = Result(spv::Op::OpVectorShuffle, vec2_float_type,
+			                          positions[(i + 1u) % 3u], positions[(i + 1u) % 3u], 0u, 1u);
+			coordinate_equal[i] = Result(spv::Op::OpFOrdEqual, vec2_bool_type, left, right);
+		}
+
+		std::array<uint32_t, 3> barycentric {};
+		std::array<uint32_t, 3> edge_vertex {};
+		const auto float_minus_one = Constant(float_type, std::bit_cast<uint32_t>(-1.0f));
+		for (uint32_t i = 0; i < edge_vertex.size(); i++) {
+			const auto previous = (i + 2u) % 3u;
+			const auto xy       = Result(
+			    spv::Op::OpLogicalAnd, bool_type,
+			    Result(spv::Op::OpCompositeExtract, bool_type, coordinate_equal[i], 0u),
+			    Result(spv::Op::OpCompositeExtract, bool_type, coordinate_equal[previous], 1u));
+			const auto yx = Result(
+			    spv::Op::OpLogicalAnd, bool_type,
+			    Result(spv::Op::OpCompositeExtract, bool_type, coordinate_equal[i], 1u),
+			    Result(spv::Op::OpCompositeExtract, bool_type, coordinate_equal[previous], 0u));
+			edge_vertex[i] = Result(spv::Op::OpLogicalOr, bool_type, xy, yx);
+			barycentric[i] =
+			    Result(spv::Op::OpSelect, float_type, edge_vertex[i], float_minus_one, float_one);
+		}
+
+		auto vertex_index = Result(spv::Op::OpSelect, int_type, edge_vertex[2], Int(2), Int(0));
+		vertex_index = Result(spv::Op::OpSelect, int_type, edge_vertex[1], Int(1), vertex_index);
+		const auto invocation = Load(int_type, invocation_id);
+		const auto is_fourth  = Result(spv::Op::OpIEqual, bool_type, invocation, Int(3));
+		const auto index =
+		    Result(spv::Op::OpSMod, int_type,
+		           Result(spv::Op::OpIAdd, int_type, vertex_index, invocation), Int(3));
+
+		const auto position3 = Interpolate(positions[0], positions[1], positions[2], barycentric);
+		const auto position =
+		    Result(spv::Op::OpSelect, vec4_float_type, is_fourth, position3,
+		           Load(vec4_float_type, Access(ptr_input_vec4_float, gl_in, index, Int(0))));
+		Store(Access(ptr_output_vec4_float, gl_out, invocation, Int(0)), position);
+
+		for (uint32_t i = 0; i < parameters.size(); i++) {
+			const auto input0 =
+			    Load(vec4_float_type, Access(ptr_input_vec4_float, inputs[i], Int(0)));
+			if (parameters[i].flat) {
+				Store(Access(ptr_output_vec4_float, outputs[i], invocation), input0);
+				continue;
+			}
+			const auto input1 =
+			    Load(vec4_float_type, Access(ptr_input_vec4_float, inputs[i], Int(1)));
+			const auto input2 =
+			    Load(vec4_float_type, Access(ptr_input_vec4_float, inputs[i], Int(2)));
+			const auto input3 = Interpolate(input0, input1, input2, barycentric);
+			const auto value =
+			    Result(spv::Op::OpSelect, vec4_float_type, is_fourth, input3,
+			           Load(vec4_float_type, Access(ptr_input_vec4_float, inputs[i], index)));
+			Store(Access(ptr_output_vec4_float, outputs[i], invocation), value);
+		}
+
+		Emit(spv::Op::OpReturn);
+		Emit(spv::Op::OpFunctionEnd);
+		return builder.Build();
+	}
+
+	std::vector<uint32_t> EmitEvaluation() {
+		DefineEntry(spv::ExecutionModel::TessellationEvaluation);
+
+		const auto x     = Load(float_type, Access(ptr_input_float, tess_coord, Int(0)));
+		const auto y     = Load(float_type, Access(ptr_input_float, tess_coord, Int(1)));
+		const auto index = Result(
+		    spv::Op::OpIAdd, int_type,
+		    Result(spv::Op::OpIMul, int_type, Result(spv::Op::OpConvertFToS, int_type, y), Int(2)),
+		    Result(spv::Op::OpConvertFToS, int_type, x));
+
+		const auto position =
+		    Load(vec4_float_type, Access(ptr_input_vec4_float, gl_in, index, Int(0)));
+		Store(Access(ptr_output_vec4_float, gl_out, Int(0)), position);
+		for (uint32_t i = 0; i < parameters.size(); i++) {
+			Store(outputs[i],
+			      Load(vec4_float_type, Access(ptr_input_vec4_float, inputs[i], index)));
+		}
+
+		Emit(spv::Op::OpReturn);
+		Emit(spv::Op::OpFunctionEnd);
+		return builder.Build();
+	}
+
+private:
+	template <typename... Args>
+	uint32_t Type(spv::Op opcode, Args... operands) {
+		const auto id = builder.AllocateId();
+		builder.AddType({Word(opcode), id, Word(operands)...});
+		return id;
+	}
+
+	uint32_t Constant(uint32_t type, uint32_t value) {
+		const auto id = builder.AllocateId();
+		builder.AddType({Word(spv::Op::OpConstant), type, id, value});
+		return id;
+	}
+
+	uint32_t Pointer(spv::StorageClass storage, uint32_t type) {
+		return Type(spv::Op::OpTypePointer, storage, type);
+	}
+
+	uint32_t Array(uint32_t type, uint32_t size) {
+		return Type(spv::Op::OpTypeArray, type, Uint(size));
+	}
+
+	template <typename... Args>
+	uint32_t Result(spv::Op opcode, uint32_t type, Args... operands) {
+		const auto id = builder.AllocateId();
+		builder.AddFunction({Word(opcode), type, id, Word(operands)...});
+		return id;
+	}
+
+	template <typename... Args>
+	uint32_t ResultWithoutType(spv::Op opcode, Args... operands) {
+		const auto id = builder.AllocateId();
+		builder.AddFunction({Word(opcode), id, Word(operands)...});
+		return id;
+	}
+
+	template <typename... Args>
+	void Emit(spv::Op opcode, Args... operands) {
+		builder.AddFunction({Word(opcode), Word(operands)...});
+	}
+
+	template <typename... Args>
+	uint32_t Access(uint32_t pointer_type, uint32_t base, Args... indices) {
+		return Result(spv::Op::OpAccessChain, pointer_type, base, Word(indices)...);
+	}
+
+	uint32_t Load(uint32_t type, uint32_t pointer) {
+		return Result(spv::Op::OpLoad, type, pointer);
+	}
+
+	void Store(uint32_t pointer, uint32_t value) { Emit(spv::Op::OpStore, pointer, value); }
+
+	uint32_t Int(uint32_t value) {
+		auto& id = int_constants[value];
+		if (id == 0) {
+			id = Constant(int_type, value);
+		}
+		return id;
+	}
+
+	uint32_t Uint(uint32_t value) {
+		auto& id = uint_constants[value];
+		if (id == 0) {
+			id = Constant(uint_type, value);
+		}
+		return id;
+	}
+
+	uint32_t AddInterface(spv::StorageClass storage, uint32_t type) {
+		const auto variable = builder.AllocateId();
+		builder.AddType(
+		    {Word(spv::Op::OpVariable), Pointer(storage, type), variable, Word(storage)});
+		interfaces.push_back(variable);
+		return variable;
+	}
+
+	void Decorate(uint32_t target, spv::Decoration decoration, uint32_t value) {
+		builder.AddAnnotation({Word(spv::Op::OpDecorate), target, Word(decoration), value});
+	}
+
+	void DefineEntry(spv::ExecutionModel model) {
+		builder.AddCapability({Word(spv::Capability::Shader)});
+		builder.AddCapability({Word(spv::Capability::Tessellation)});
+		main = Result(spv::Op::OpFunction, void_type, spv::FunctionControlMask::MaskNone,
+		              function_type);
+		if (model == spv::ExecutionModel::TessellationControl) {
+			builder.AddExecutionMode({main, Word(spv::ExecutionMode::OutputVertices), 4u});
+		} else {
+			builder.AddExecutionMode({main, Word(spv::ExecutionMode::Quads)});
+			builder.AddExecutionMode({main, Word(spv::ExecutionMode::SpacingEqual)});
+			builder.AddExecutionMode({main, Word(spv::ExecutionMode::VertexOrderCw)});
+		}
+		DefineInputs(model);
+		DefineOutputs(model);
+		builder.AddEntryPoint(Word(model), main, "main", interfaces);
+		ResultWithoutType(spv::Op::OpLabel);
+	}
+
+	void DefineInputs(spv::ExecutionModel model) {
+		const auto tess_control = model == spv::ExecutionModel::TessellationControl;
+		if (tess_control) {
+			invocation_id = AddInterface(spv::StorageClass::Input, int_type);
+			Decorate(invocation_id, spv::Decoration::BuiltIn, Word(spv::BuiltIn::InvocationId));
+		} else {
+			tess_coord = AddInterface(spv::StorageClass::Input, vec3_float_type);
+			Decorate(tess_coord, spv::Decoration::BuiltIn, Word(spv::BuiltIn::TessCoord));
+		}
+		gl_in =
+		    AddInterface(spv::StorageClass::Input, Array(per_vertex_type, tess_control ? 3u : 4u));
+
+		inputs.resize(parameters.size());
+		std::array<uint32_t, ShaderVertexInputInfo::RES_MAX> locations {};
+		for (uint32_t i = 0; i < parameters.size(); i++) {
+			const auto location =
+			    tess_control ? parameters[i].input_location : parameters[i].output_location;
+			if (tess_control && locations[location] != 0) {
+				inputs[i] = locations[location];
+				continue;
+			}
+			inputs[i] = AddInterface(spv::StorageClass::Input,
+			                         Array(vec4_float_type, tess_control ? 3u : 4u));
+			Decorate(inputs[i], spv::Decoration::Location, location);
+			locations[location] = inputs[i];
+		}
+	}
+
+	void DefineOutputs(spv::ExecutionModel model) {
+		const auto tess_control = model == spv::ExecutionModel::TessellationControl;
+		if (tess_control) {
+			gl_out     = AddInterface(spv::StorageClass::Output, Array(per_vertex_type, 4u));
+			tess_inner = AddInterface(spv::StorageClass::Output, Array(float_type, 2u));
+			Decorate(tess_inner, spv::Decoration::BuiltIn, Word(spv::BuiltIn::TessLevelInner));
+			builder.AddAnnotation(
+			    {Word(spv::Op::OpDecorate), tess_inner, Word(spv::Decoration::Patch)});
+			tess_outer = AddInterface(spv::StorageClass::Output, Array(float_type, 4u));
+			Decorate(tess_outer, spv::Decoration::BuiltIn, Word(spv::BuiltIn::TessLevelOuter));
+			builder.AddAnnotation(
+			    {Word(spv::Op::OpDecorate), tess_outer, Word(spv::Decoration::Patch)});
+		} else {
+			gl_out = AddInterface(spv::StorageClass::Output, per_vertex_type);
+		}
+
+		outputs.resize(parameters.size());
+		for (uint32_t i = 0; i < parameters.size(); i++) {
+			outputs[i] = AddInterface(spv::StorageClass::Output,
+			                          tess_control ? Array(vec4_float_type, 4u) : vec4_float_type);
+			Decorate(outputs[i], spv::Decoration::Location, parameters[i].output_location);
+		}
+	}
+
+	uint32_t Interpolate(uint32_t v0, uint32_t v1, uint32_t v2,
+	                     const std::array<uint32_t, 3>& barycentric) {
+		const auto p0 = Result(spv::Op::OpVectorTimesScalar, vec4_float_type, v0, barycentric[0]);
+		const auto p1 = Result(spv::Op::OpVectorTimesScalar, vec4_float_type, v1, barycentric[1]);
+		const auto p2 = Result(spv::Op::OpVectorTimesScalar, vec4_float_type, v2, barycentric[2]);
+		return Result(spv::Op::OpFAdd, vec4_float_type, p0,
+		              Result(spv::Op::OpFAdd, vec4_float_type, p1, p2));
+	}
+
+	Builder                       builder {SpirvVersion15};
+	const std::vector<Parameter>& parameters;
+	std::vector<uint32_t>         interfaces;
+	std::vector<uint32_t>         inputs;
+	std::vector<uint32_t>         outputs;
+	std::array<uint32_t, 5>       int_constants {};
+	std::array<uint32_t, 5>       uint_constants {};
+	uint32_t                      main                  = 0;
+	uint32_t                      void_type             = 0;
+	uint32_t                      bool_type             = 0;
+	uint32_t                      uint_type             = 0;
+	uint32_t                      int_type              = 0;
+	uint32_t                      float_type            = 0;
+	uint32_t                      vec2_bool_type        = 0;
+	uint32_t                      vec2_float_type       = 0;
+	uint32_t                      vec3_float_type       = 0;
+	uint32_t                      vec4_float_type       = 0;
+	uint32_t                      function_type         = 0;
+	uint32_t                      per_vertex_type       = 0;
+	uint32_t                      ptr_input_float       = 0;
+	uint32_t                      ptr_input_vec4_float  = 0;
+	uint32_t                      ptr_output_float      = 0;
+	uint32_t                      ptr_output_vec4_float = 0;
+	uint32_t                      gl_in                 = 0;
+	uint32_t                      gl_out                = 0;
+	uint32_t                      tess_inner            = 0;
+	uint32_t                      tess_outer            = 0;
+	uint32_t                      tess_coord            = 0;
+	uint32_t                      invocation_id         = 0;
+};
+
+} // namespace
+
+RectListShaders BuildRectListShaders(const ShaderVertexInputInfo& vertex_info,
+                                     const ShaderPixelInputInfo*  pixel_info) {
+	const auto      parameters = GetParameters(vertex_info, pixel_info);
+	RectListEmitter control(parameters, spv::ExecutionModel::TessellationControl);
+	RectListEmitter evaluation(parameters, spv::ExecutionModel::TessellationEvaluation);
+	return {control.EmitControl(), evaluation.EmitEvaluation()};
+}
+
+} // namespace Libs::Graphics

src/graphics/shader/rectListShader.h

@@ -0,0 +1,22 @@
+#ifndef EMULATOR_SRC_GRAPHICS_SHADER_RECTLISTSHADER_H_
+#define EMULATOR_SRC_GRAPHICS_SHADER_RECTLISTSHADER_H_
+
+#include <cstdint>
+#include <vector>
+
+namespace Libs::Graphics {
+
+struct ShaderPixelInputInfo;
+struct ShaderVertexInputInfo;
+
+struct RectListShaders {
+	std::vector<uint32_t> control;
+	std::vector<uint32_t> evaluation;
+};
+
+RectListShaders BuildRectListShaders(const ShaderVertexInputInfo& vertex_info,
+                                     const ShaderPixelInputInfo*  pixel_info);
+
+} // namespace Libs::Graphics
+
+#endif // EMULATOR_SRC_GRAPHICS_SHADER_RECTLISTSHADER_H_

src/graphics/shader/shader.cpp

@@ -45,6 +45,42 @@
 
 namespace Libs::Graphics {
 
+namespace {
+
+constexpr uint32_t PsInputOffsetMask = 0x0000001fu;
+constexpr uint32_t PsInputFlatShade  = 0x00000400u;
+
+} // namespace
+
+uint32_t ShaderPixelParameterMappedLocation(const ShaderPixelInputInfo& info, uint32_t input) {
+	return input < info.input_num ? info.interpolator_settings[input] & PsInputOffsetMask : input;
+}
+
+uint32_t ShaderPixelParameterLocation(const ShaderPixelInputInfo& info,
+                                      std::span<const uint32_t> active_inputs, uint32_t input) {
+	std::array<bool, 32> used_locations {};
+	for (const auto active_input: active_inputs) {
+		auto location = ShaderPixelParameterMappedLocation(info, active_input);
+		if (location < used_locations.size() && used_locations[location]) {
+			location = active_input;
+			while (location < used_locations.size() && used_locations[location]) {
+				location++;
+			}
+			EXIT_NOT_IMPLEMENTED(location >= used_locations.size());
+		}
+
+		if (active_input == input) {
+			return location;
+		}
+		used_locations[location] = true;
+	}
+	return ShaderPixelParameterMappedLocation(info, input);
+}
+
+bool ShaderPixelParameterIsFlat(const ShaderPixelInputInfo& info, uint32_t input) {
+	return input < info.input_num && (info.interpolator_settings[input] & PsInputFlatShade) != 0;
+}
+
 struct ShaderBinaryInfo {
 	uint8_t  signature[7];
 	uint8_t  version;
@@ -1606,6 +1642,7 @@ ShaderId ShaderGetIdPS(const HW::PixelShaderInfo& regs, const ShaderPixelInputIn
 	ret.ids.push_back(static_cast<uint32_t>(input_info.ps_pos_z));
 	ret.ids.push_back(static_cast<uint32_t>(input_info.ps_pos_w));
 	ret.ids.push_back(static_cast<uint32_t>(input_info.ps_front_face));
+	ret.ids.push_back(static_cast<uint32_t>(input_info.ps_no_perspective));
 	ret.ids.push_back(static_cast<uint32_t>(input_info.ps_pixel_kill_enable));
 	ret.ids.push_back(static_cast<uint32_t>(input_info.ps_sample_mask_export_enable));
 	ret.ids.push_back(static_cast<uint32_t>(input_info.ps_early_z));

src/graphics/shader/shader.h

@@ -122,6 +122,11 @@ struct ShaderPixelInputInfo {
 	bool HasPositionInput() const { return ps_pos_x || ps_pos_y || ps_pos_z || ps_pos_w; }
 };
 
+uint32_t ShaderPixelParameterMappedLocation(const ShaderPixelInputInfo& info, uint32_t input);
+uint32_t ShaderPixelParameterLocation(const ShaderPixelInputInfo& info,
+                                      std::span<const uint32_t> active_inputs, uint32_t input);
+bool     ShaderPixelParameterIsFlat(const ShaderPixelInputInfo& info, uint32_t input);
+
 struct ShaderSharp {
 	uint16_t offset_dw : 15;
 	uint16_t size      : 1;

src/kernel/pthread.cpp

@@ -365,6 +365,7 @@ struct PthreadAttrPrivate {
 	uint64_t       stack_map_addr;
 	size_t         stack_map_size;
 	int            policy;
+	int            guest_priority;
 	int            inherit_sched;
 	int            solosched;
 	bool           detached;
@@ -2128,13 +2129,8 @@ int KYTY_SYSV_ABI PthreadAttrGetschedparam(const PthreadAttr* attr, KernelSchedP
 
 	int result = pthread_attr_getschedparam(&(*attr)->p, param);
 
-	if (param->sched_priority <= -2) {
-		param->sched_priority = 767;
-	} else if (param->sched_priority >= +2) {
-		param->sched_priority = 256;
-	} else {
-		param->sched_priority = 700;
-	}
+	// Host priority mapping is lossy; return the exact guest value.
+	param->sched_priority = (*attr)->guest_priority;
 
 	if (result == 0) {
 		return OK;
@@ -2298,6 +2294,7 @@ int KYTY_SYSV_ABI PthreadAttrSetschedparam(PthreadAttr* attr, const KernelSchedP
 		return KERNEL_ERROR_EINVAL;
 	}
 
+#if KYTY_PLATFORM == KYTY_PLATFORM_WINDOWS
 	KernelSchedParam pparam {};
 	if (param->sched_priority <= 478) {
 		pparam.sched_priority = +2;
@@ -2307,12 +2304,13 @@ int KYTY_SYSV_ABI PthreadAttrSetschedparam(PthreadAttr* attr, const KernelSchedP
 		pparam.sched_priority = 0;
 	}
 
-	int result = pthread_attr_setschedparam(&attr_value->p, &pparam);
-
-	if (result == 0) {
-		return OK;
+	if (pthread_attr_setschedparam(&attr_value->p, &pparam) != 0) {
+		return KERNEL_ERROR_EINVAL;
 	}
-	return KERNEL_ERROR_EINVAL;
+#endif
+
+	attr_value->guest_priority = param->sched_priority;
+	return OK;
 }
 
 int KYTY_SYSV_ABI PthreadAttrSetschedpolicy(PthreadAttr* attr, int policy) {
@@ -3639,26 +3637,15 @@ int KYTY_SYSV_ABI PthreadGetprio(Pthread thread, int* prio) {
 
 	EXIT_NOT_IMPLEMENTED(prio == nullptr);
 
-	sched_param param {};
-	int         pol = 0;
-
-	int result = pthread_getschedparam(thread->p, &pol, &param);
-
-	if (result == 0) {
-		if (param.sched_priority <= -2) {
-			*prio = 767;
-		} else if (param.sched_priority >= +2) {
-			*prio = 256;
-		} else {
-			*prio = 700;
-		}
-
-		LOGF("\t PthreadGetprio: %d, %d\n", thread->unique_id, *prio);
-
-		return OK;
+	sched_param native_param {};
+	int         native_policy = 0;
+	if (pthread_getschedparam(thread->p, &native_policy, &native_param) != 0) {
+		return KERNEL_ERROR_EINVAL;
 	}
 
-	return KERNEL_ERROR_EINVAL;
+	*prio = thread->attr->guest_priority;
+	LOGF("\t PthreadGetprio: %d, %d\n", thread->unique_id, *prio);
+	return OK;
 }
 
 int KYTY_SYSV_ABI PthreadSetprio(Pthread thread, int prio) {
@@ -3673,25 +3660,27 @@ int KYTY_SYSV_ABI PthreadSetprio(Pthread thread, int prio) {
 
 	int result = pthread_getschedparam(thread->p, &pol, &param);
 
-	if (result == 0) {
-		if (prio <= 478) {
-			param.sched_priority = +2;
-		} else if (prio >= 733) {
-			param.sched_priority = -2;
-		} else {
-			param.sched_priority = 0;
-		}
-
-		result = pthread_setschedparam(thread->p, pol, &param);
-
-		if (result == 0) {
-			LOGF("\t PthreadSetprio: %d, %d\n", thread->unique_id, prio);
-
-			return OK;
-		}
+	if (result != 0) {
+		return KERNEL_ERROR_EINVAL;
 	}
 
-	return KERNEL_ERROR_EINVAL;
+#if KYTY_PLATFORM == KYTY_PLATFORM_WINDOWS
+	if (prio <= 478) {
+		param.sched_priority = +2;
+	} else if (prio >= 733) {
+		param.sched_priority = -2;
+	} else {
+		param.sched_priority = 0;
+	}
+
+	if (pthread_setschedparam(thread->p, pol, &param) != 0) {
+		return KERNEL_ERROR_EINVAL;
+	}
+#endif
+
+	thread->attr->guest_priority = prio;
+	LOGF("\t PthreadSetprio: %d, %d\n", thread->unique_id, prio);
+	return OK;
 }
 
 void KYTY_SYSV_ABI PthreadTestcancel() {

src/launcher/CMakeLists.txt

@@ -141,14 +141,14 @@ elseif(LINUX)
 endif()
 
 
-set(tidy_dirs "${CMAKE_SOURCE_DIR}/launcher/include") 
-set(iwyu_maps "${CMAKE_SOURCE_DIR}/launcher/utils/qt6_16.imp")
+set(tidy_dirs "${CMAKE_CURRENT_SOURCE_DIR}/include")
+set(iwyu_maps "${CMAKE_CURRENT_SOURCE_DIR}/utils/qt6_16.imp")
 
 get_property(inc_headers TARGET launcher PROPERTY INCLUDE_DIRECTORIES)
 
 list(APPEND inc_headers
-	${CMAKE_SOURCE_DIR}/launcher
-	${CMAKE_BINARY_DIR}/launcher/launcher_autogen/include
+	${CMAKE_CURRENT_SOURCE_DIR}
+	${CMAKE_CURRENT_BINARY_DIR}/launcher_autogen/include
 	${Qt6Widgets_INCLUDE_DIRS}
 	${Qt6Core_INCLUDE_DIRS}
 	${Qt6Gui_INCLUDE_DIRS}

src/launcher/forms/configuration_edit_dialog.ui

@@ -7,7 +7,7 @@
     <x>0</x>
     <y>0</y>
     <width>560</width>
-    <height>420</height>
+    <height>450</height>
    </rect>
   </property>
   <property name="windowTitle">
@@ -78,19 +78,6 @@
            </property>
           </widget>
          </item>
-         <item row="2" column="0" colspan="2">
-          <widget class="QCheckBox" name="checkBox_ngg_rectlist_draw">
-           <property name="toolTip">
-            <string>Use the NGG 4-vertex path for rect-list DrawIndexAuto primitive 7</string>
-           </property>
-           <property name="text">
-            <string>Use NGG rect-list draw</string>
-           </property>
-           <property name="checked">
-            <bool>true</bool>
-           </property>
-          </widget>
-         </item>
         </layout>
        </widget>
       </item>
@@ -132,41 +119,55 @@
        </widget>
       </item>
       <item row="3" column="0">
+       <widget class="QLabel" name="label_console_language">
+        <property name="text">
+         <string>Console language:</string>
+        </property>
+       </widget>
+      </item>
+      <item row="3" column="1">
+       <widget class="QComboBox" name="comboBox_console_language">
+        <property name="toolTip">
+         <string>Language</string>
+        </property>
+       </widget>
+      </item>
+      <item row="4" column="0">
        <widget class="QLabel" name="label_20">
         <property name="text">
          <string>Shader optimization type:</string>
         </property>
        </widget>
       </item>
-      <item row="3" column="1">
+      <item row="4" column="1">
        <widget class="QComboBox" name="comboBox_shader_optimization_type">
         <property name="toolTip">
          <string>Optimize shaders for code size or performance</string>
         </property>
        </widget>
       </item>
-      <item row="4" column="0">
+      <item row="5" column="0">
        <widget class="QLabel" name="label_21">
         <property name="text">
          <string>Shader log direction:</string>
         </property>
        </widget>
       </item>
-      <item row="4" column="1">
+      <item row="5" column="1">
        <widget class="QComboBox" name="comboBox_shader_log_direction">
         <property name="toolTip">
          <string>Dump shaders to file or console window. If enabled may decrease emulator performance</string>
         </property>
        </widget>
       </item>
-      <item row="5" column="0">
+      <item row="6" column="0">
        <widget class="QLabel" name="label_22">
         <property name="text">
          <string>Shader log folder:</string>
         </property>
        </widget>
       </item>
-      <item row="5" column="1">
+      <item row="6" column="1">
        <widget class="MandatoryLineEdit" name="lineEdit_shader_log_folder">
         <property name="toolTip">
          <string>Specify directory to dump shaders</string>
@@ -176,14 +177,14 @@
         </property>
        </widget>
       </item>
-      <item row="6" column="0">
+      <item row="7" column="0">
        <widget class="QLabel" name="label_24">
         <property name="text">
          <string>Command buffer dump folder:</string>
         </property>
        </widget>
       </item>
-      <item row="6" column="1">
+      <item row="7" column="1">
        <widget class="MandatoryLineEdit" name="lineEdit_cmd_dump_folder">
         <property name="toolTip">
          <string>Specify directory to dump command buffers</string>
@@ -193,28 +194,28 @@
         </property>
        </widget>
       </item>
-      <item row="7" column="0">
+      <item row="8" column="0">
        <widget class="QLabel" name="label_25">
         <property name="text">
          <string>Printf direction:</string>
         </property>
        </widget>
       </item>
-      <item row="7" column="1">
+      <item row="8" column="1">
        <widget class="QComboBox" name="comboBox_printf_direction">
         <property name="toolTip">
          <string>Print logs to file or console window. If enabled may decrease emulator performance</string>
         </property>
        </widget>
       </item>
-      <item row="8" column="0">
+      <item row="9" column="0">
        <widget class="QLabel" name="label_26">
         <property name="text">
          <string>Printf output file:</string>
         </property>
        </widget>
       </item>
-      <item row="8" column="1">
+      <item row="9" column="1">
        <widget class="MandatoryLineEdit" name="lineEdit_printf_file">
         <property name="toolTip">
          <string>Specify file to dump logs</string>
@@ -224,14 +225,14 @@
         </property>
        </widget>
       </item>
-      <item row="9" column="0">
+      <item row="10" column="0">
        <widget class="QLabel" name="label_27">
         <property name="text">
          <string>Profiler direction:</string>
         </property>
        </widget>
       </item>
-      <item row="9" column="1">
+      <item row="10" column="1">
        <widget class="QComboBox" name="comboBox_profiler_direction">
         <property name="toolTip">
          <string>Enable or disable profiler. If enabled may decrease emulator performance</string>

src/launcher/include/configuration.h

@@ -48,6 +48,9 @@ class Configuration: public QObject {
 	Q_OBJECT
 
 public:
+	static constexpr int DEFAULT_CONSOLE_LANGUAGE = 1;
+	static constexpr int MAX_CONSOLE_LANGUAGE     = 29;
+
 	enum class Resolution {
 		R1280X720,
 		R1920X1080,
@@ -83,6 +86,7 @@ public:
 
 	Resolution             screen_resolution           = Resolution::R1280X720;
 	int                    vblank_frequency            = 60;
+	int                    console_language            = DEFAULT_CONSOLE_LANGUAGE;
 	bool                   vulkan_validation_enabled   = true;
 	bool                   shader_validation_enabled   = true;
 	ShaderOptimizationType shader_optimization_type    = ShaderOptimizationType::Performance;
@@ -94,13 +98,13 @@ public:
 	QString                printf_output_file          = "_kyty.txt";
 	ProfilerDirection      profiler_direction          = ProfilerDirection::None;
 	bool                   renderdoc_enabled           = false;
-	bool                   ngg_rectlist_draw_enabled   = true;
 
 	QString elf = QStringLiteral("eboot.bin");
 
 	void CopyEmulatorSettingsFrom(const Configuration& other) {
 		screen_resolution           = other.screen_resolution;
 		vblank_frequency            = other.vblank_frequency;
+		console_language            = other.console_language;
 		vulkan_validation_enabled   = other.vulkan_validation_enabled;
 		shader_validation_enabled   = other.shader_validation_enabled;
 		shader_optimization_type    = other.shader_optimization_type;
@@ -112,7 +116,6 @@ public:
 		printf_output_file          = other.printf_output_file;
 		profiler_direction          = other.profiler_direction;
 		renderdoc_enabled           = other.renderdoc_enabled;
-		ngg_rectlist_draw_enabled   = other.ngg_rectlist_draw_enabled;
 	}
 
 	void CopyFrom(const Configuration& other) {
@@ -136,6 +139,7 @@ public:
 		KYTY_CFG_SET(custom_settings);
 		KYTY_CFG_SET(screen_resolution);
 		KYTY_CFG_SET(vblank_frequency);
+		KYTY_CFG_SET(console_language);
 		KYTY_CFG_SET(vulkan_validation_enabled);
 		KYTY_CFG_SET(shader_validation_enabled);
 		KYTY_CFG_SET(shader_optimization_type);
@@ -147,7 +151,6 @@ public:
 		KYTY_CFG_SET(printf_output_file);
 		KYTY_CFG_SET(profiler_direction);
 		KYTY_CFG_SET(renderdoc_enabled);
-		KYTY_CFG_SET(ngg_rectlist_draw_enabled);
 		KYTY_CFG_SET(elf);
 	}
 
@@ -158,6 +161,10 @@ public:
 		KYTY_CFG_GET(custom_settings);
 		KYTY_CFG_GET(screen_resolution);
 		vblank_frequency = s->value("vblank_frequency", vblank_frequency).toInt();
+		console_language = s->value("console_language", console_language).toInt();
+		if (console_language < 0 || console_language > MAX_CONSOLE_LANGUAGE) {
+			console_language = DEFAULT_CONSOLE_LANGUAGE;
+		}
 		KYTY_CFG_GET(vulkan_validation_enabled);
 		KYTY_CFG_GET(shader_validation_enabled);
 		KYTY_CFG_GET(shader_optimization_type);
@@ -169,8 +176,6 @@ public:
 		KYTY_CFG_GET(printf_output_file);
 		KYTY_CFG_GET(profiler_direction);
 		KYTY_CFG_GET(renderdoc_enabled);
-		ngg_rectlist_draw_enabled =
-		    s->value("ngg_rectlist_draw_enabled", ngg_rectlist_draw_enabled).toBool();
 		elf = s->value("elf", elf).toString();
 	}
 };

src/launcher/src/configurationEditDialog.cpp

@@ -30,6 +30,39 @@ constexpr char SETTINGS_CFG_DIALOG[]               = "ConfigurationEditDialog";
 constexpr char SETTINGS_CFG_LAST_GEOMETRY[]        = "geometry";
 constexpr int  GLOBAL_SETTINGS_GAME_DIRS_MIN_WIDTH = 560;
 
+const QStringList CONSOLE_LANGUAGE_NAMES = {
+    "Japanese",
+    "English (United States)",
+    "French (France)",
+    "Spanish (Spain)",
+    "German",
+    "Italian",
+    "Dutch",
+    "Portuguese (Portugal)",
+    "Russian",
+    "Korean",
+    "Chinese (Traditional)",
+    "Chinese (Simplified)",
+    "Finnish",
+    "Swedish",
+    "Danish",
+    "Norwegian",
+    "Polish",
+    "Portuguese (Brazil)",
+    "English (United Kingdom)",
+    "Turkish",
+    "Spanish (Latin America)",
+    "Arabic",
+    "French (Canada)",
+    "Czech",
+    "Hungarian",
+    "Greek",
+    "Romanian",
+    "Thai",
+    "Vietnamese",
+    "Indonesian",
+};
+
 static QString NormalizeGameDirectory(const QString& dir) {
 	const auto trimmed = dir.trimmed();
 	if (trimmed.isEmpty()) {
@@ -121,10 +154,15 @@ static void ListInit(QComboBox* combo, T value) {
 void ConfigurationEditDialog::Init(const Configuration& info) {
 	ListInit(m_ui->comboBox_screen_resolution, info.screen_resolution);
 	m_ui->spinBox_vblank_frequency->setValue(info.vblank_frequency);
+	m_ui->comboBox_console_language->clear();
+	m_ui->comboBox_console_language->addItems(CONSOLE_LANGUAGE_NAMES);
+	m_ui->comboBox_console_language->setCurrentIndex(
+	    info.console_language >= 0 && info.console_language < CONSOLE_LANGUAGE_NAMES.size()
+	        ? info.console_language
+	        : Configuration::DEFAULT_CONSOLE_LANGUAGE);
 	m_ui->checkBox_shader_validation->setChecked(info.shader_validation_enabled);
 	m_ui->checkBox_vulkan_validation->setChecked(info.vulkan_validation_enabled);
 	m_ui->checkBox_renderdoc_capture->setChecked(info.renderdoc_enabled);
-	m_ui->checkBox_ngg_rectlist_draw->setChecked(info.ngg_rectlist_draw_enabled);
 	ListInit(m_ui->comboBox_shader_optimization_type, info.shader_optimization_type);
 	ListInit(m_ui->comboBox_shader_log_direction, info.shader_log_direction);
 	m_ui->lineEdit_shader_log_folder->setText(info.shader_log_folder);
@@ -242,10 +280,10 @@ static void UpdateInfo(Configuration& info, Ui::ConfigurationEditDialog& ui) {
 	info.screen_resolution =
 	    TextToEnum<Configuration::Resolution>(ui.comboBox_screen_resolution->currentText());
 	info.vblank_frequency          = ui.spinBox_vblank_frequency->value();
+	info.console_language          = ui.comboBox_console_language->currentIndex();
 	info.vulkan_validation_enabled = ui.checkBox_vulkan_validation->isChecked();
 	info.shader_validation_enabled = ui.checkBox_shader_validation->isChecked();
 	info.renderdoc_enabled         = ui.checkBox_renderdoc_capture->isChecked();
-	info.ngg_rectlist_draw_enabled = ui.checkBox_ngg_rectlist_draw->isChecked();
 	info.shader_optimization_type  = TextToEnum<Configuration::ShaderOptimizationType>(
 	    ui.comboBox_shader_optimization_type->currentText());
 	info.shader_log_direction = TextToEnum<Configuration::ShaderLogDirection>(

src/launcher/src/mainDialog.cpp

@@ -205,6 +205,7 @@ static QStringList CreateEmulatorArgs(const Configuration& info) {
 	args << "--screen-width" << r.at(0);
 	args << "--screen-height" << r.at(1);
 	args << "--vblank-frequency" << QString::number(info.vblank_frequency);
+	args << "--console-language" << QString::number(info.console_language);
 	args << "--vulkan-validation" << BoolArg(info.vulkan_validation_enabled);
 	args << "--shader-validation" << BoolArg(info.shader_validation_enabled);
 	args << "--shader-optimization-type" << EnumToText(info.shader_optimization_type);
@@ -216,7 +217,6 @@ static QStringList CreateEmulatorArgs(const Configuration& info) {
 	args << "--printf-output-file" << info.printf_output_file;
 	args << "--profiler-direction" << EnumToText(info.profiler_direction);
 	args << "--spirv-debug-printf" << "false";
-	args << "--ngg-rectlist-draw" << BoolArg(info.ngg_rectlist_draw_enabled);
 	if (info.renderdoc_enabled) {
 		args << "--rd";
 	}

src/libs/agc.cpp

@@ -221,57 +221,6 @@ static RegisterDefaults* get_internal_register_defaults(uint32_t ver) {
 	return get_register_defaults(g_agc_internal_reg_defaults_by_version[index], &storage[index]);
 }
 
-struct PendingGraphicsSegment {
-	uint32_t* start     = nullptr;
-	uint32_t* end       = nullptr;
-	uint32_t* range_end = nullptr;
-};
-
-static std::mutex             g_pending_graphics_segment_mutex;
-static PendingGraphicsSegment g_pending_graphics_segment;
-
-static void track_pending_graphics_segment_after_submit(uint32_t* dcb, uint32_t size_in_dwords) {
-	if (dcb == nullptr || size_in_dwords == 0) {
-		return;
-	}
-
-	auto* segment_start = dcb + size_in_dwords;
-	auto* range_end     = segment_start + 0xfffffu;
-
-	std::lock_guard lock(g_pending_graphics_segment_mutex);
-	g_pending_graphics_segment.start     = segment_start;
-	g_pending_graphics_segment.end       = segment_start;
-	g_pending_graphics_segment.range_end = range_end;
-}
-
-static void track_pending_graphics_allocation(uint32_t* cmd, uint32_t size_dw) {
-	if (cmd == nullptr || size_dw == 0) {
-		return;
-	}
-
-	std::lock_guard lock(g_pending_graphics_segment_mutex);
-	auto*           range_start = g_pending_graphics_segment.start;
-	auto*           range_end   = g_pending_graphics_segment.range_end;
-	if (range_start == nullptr || range_end == nullptr || cmd < range_start || cmd >= range_end) {
-		return;
-	}
-
-	auto* cmd_end = cmd + size_dw;
-	if (cmd > g_pending_graphics_segment.end) {
-		static std::atomic<uint32_t> log_count {0};
-		if (log_count.fetch_add(1) < 64) {
-			LOGF("\t pending graphics segment: ignoring non-contiguous allocation cmd = "
-			     "0x%016" PRIx64 ", tracked_end = 0x%016" PRIx64 "\n",
-			     reinterpret_cast<uint64_t>(cmd),
-			     reinterpret_cast<uint64_t>(g_pending_graphics_segment.end));
-		}
-		return;
-	}
-	if (cmd_end > g_pending_graphics_segment.end && cmd_end <= range_end) {
-		g_pending_graphics_segment.end = cmd_end;
-	}
-}
-
 struct CommandBuffer {
 	using Callback = KYTY_SYSV_ABI bool (*)(CommandBuffer*, uint32_t, void*);
 
@@ -370,7 +319,6 @@ struct CommandBuffer {
 		}
 		auto* ret_ptr = cursor_up;
 		cursor_up += size_dw;
-		track_pending_graphics_allocation(ret_ptr, size_dw);
 		return ret_ptr;
 	}
 };
@@ -1918,7 +1866,6 @@ uint32_t* KYTY_SYSV_ABI GraphicsCbReleaseMem(CommandBuffer* buf, uint8_t action,
 	cmd[5] = static_cast<uint32_t>(packet_data & 0xffffffffu);
 	cmd[6] = static_cast<uint32_t>((packet_data >> 32u) & 0xffffffffu);
 	cmd[7] = interrupt_ctx_id & 0x07ffffffu;
-
 	return cmd;
 }
 
@@ -3815,150 +3762,6 @@ static void submit_dcb(uint32_t* dcb, uint32_t size_in_dwords) {
 	EXIT_IF(g_renderer == nullptr);
 	g_renderer->GetGpu().Submit(dcb, size_in_dwords, nullptr, 0,
 	                            !dcb_has_queued_interrupt(dcb, size_in_dwords));
-	Gen5::track_pending_graphics_segment_after_submit(dcb, size_in_dwords);
-}
-
-static std::vector<uint64_t> collect_acb_wait_addresses(const uint32_t* acb,
-                                                        uint32_t        size_in_dwords) {
-	std::vector<uint64_t> addresses;
-
-	for (uint32_t offset = 0; offset < size_in_dwords;) {
-		auto cmd_id = acb[offset];
-		auto len    = KYTY_PM4_LEN(cmd_id);
-		if (len == 0 || len > size_in_dwords - offset) {
-			return addresses;
-		}
-
-		auto op = (cmd_id >> 8u) & 0xffu;
-		if (op == Pm4::IT_NOP && KYTY_PM4_R(cmd_id) == Pm4::R_WAIT_MEM_32 && len >= 7) {
-			auto address = static_cast<uint64_t>(acb[offset + 1]) |
-			               (static_cast<uint64_t>(acb[offset + 2]) << 32u);
-			if (address != 0) {
-				addresses.push_back(address);
-			}
-		} else if (op == Pm4::IT_NOP && KYTY_PM4_R(cmd_id) == Pm4::R_WAIT_MEM_64 && len >= 9) {
-			auto address = static_cast<uint64_t>(acb[offset + 1]) |
-			               (static_cast<uint64_t>(acb[offset + 2]) << 32u);
-			if (address != 0) {
-				addresses.push_back(address);
-			}
-		}
-
-		offset += len;
-	}
-
-	return addresses;
-}
-
-static bool acb_waits_for_address(const std::vector<uint64_t>& wait_addresses,
-                                  uint64_t                     release_address) {
-	for (auto address: wait_addresses) {
-		if (address == release_address) {
-			return true;
-		}
-	}
-
-	return false;
-}
-
-static void flush_pending_graphics_segment_before_acb(const uint32_t* acb,
-                                                      uint32_t        acb_size_in_dwords) {
-	uint32_t* dcb            = nullptr;
-	uint32_t  size_in_dwords = 0;
-	auto      wait_addresses = collect_acb_wait_addresses(acb, acb_size_in_dwords);
-
-	{
-		std::lock_guard lock(Gen5::g_pending_graphics_segment_mutex);
-		if (!wait_addresses.empty() && Gen5::g_pending_graphics_segment.start != nullptr) {
-			auto* scan        = Gen5::g_pending_graphics_segment.start;
-			auto* matched_end = Gen5::g_pending_graphics_segment.start;
-			while (scan < Gen5::g_pending_graphics_segment.end) {
-				auto cmd_id = *scan;
-				if (cmd_id == 0x80000000u) {
-					scan++;
-					continue;
-				}
-				if ((cmd_id & 0xC0000000u) != 0xC0000000u) {
-					break;
-				}
-
-				auto len = KYTY_PM4_LEN(cmd_id);
-				if (len == 0 ||
-				    len > static_cast<uint32_t>(Gen5::g_pending_graphics_segment.end - scan)) {
-					break;
-				}
-
-				if (((cmd_id >> 8u) & 0xffu) == Pm4::IT_NOP &&
-				    KYTY_PM4_R(cmd_id) == Pm4::R_RELEASE_MEM && len >= 7) {
-					auto release_addr =
-					    static_cast<uint64_t>(scan[3]) | (static_cast<uint64_t>(scan[4]) << 32u);
-					if (acb_waits_for_address(wait_addresses, release_addr)) {
-						matched_end = scan + len;
-					}
-				}
-
-				scan += len;
-			}
-
-			if (matched_end > Gen5::g_pending_graphics_segment.start) {
-				Gen5::g_pending_graphics_segment.end = matched_end;
-			}
-		}
-
-		if (Gen5::g_pending_graphics_segment.start != nullptr &&
-		    Gen5::g_pending_graphics_segment.end > Gen5::g_pending_graphics_segment.start) {
-			auto* scan      = Gen5::g_pending_graphics_segment.start;
-			auto* valid_end = Gen5::g_pending_graphics_segment.start;
-			while (scan < Gen5::g_pending_graphics_segment.end) {
-				auto cmd_id = *scan;
-				if (cmd_id == 0x80000000u) {
-					scan++;
-					valid_end = scan;
-					continue;
-				}
-				if ((cmd_id & 0xC0000000u) != 0xC0000000u) {
-					break;
-				}
-
-				auto len = KYTY_PM4_LEN(cmd_id);
-				if (len == 0 ||
-				    len > static_cast<uint32_t>(Gen5::g_pending_graphics_segment.end - scan)) {
-					break;
-				}
-
-				scan += len;
-				valid_end = scan;
-			}
-
-			if (valid_end < Gen5::g_pending_graphics_segment.end) {
-				static std::atomic<uint32_t> log_count {0};
-				if (log_count.fetch_add(1) < 64) {
-					LOGF("\t trimming pending graphics segment: addr = 0x%016" PRIx64
-					     ", old_dw = 0x%08" PRIx32 ", new_dw = 0x%08" PRIx32 "\n",
-					     reinterpret_cast<uint64_t>(Gen5::g_pending_graphics_segment.start),
-					     static_cast<uint32_t>(Gen5::g_pending_graphics_segment.end -
-					                           Gen5::g_pending_graphics_segment.start),
-					     static_cast<uint32_t>(valid_end - Gen5::g_pending_graphics_segment.start));
-				}
-				Gen5::g_pending_graphics_segment.end = valid_end;
-			}
-		}
-
-		if (Gen5::g_pending_graphics_segment.start == nullptr ||
-		    Gen5::g_pending_graphics_segment.end <= Gen5::g_pending_graphics_segment.start) {
-			return;
-		}
-
-		dcb            = Gen5::g_pending_graphics_segment.start;
-		size_in_dwords = static_cast<uint32_t>(Gen5::g_pending_graphics_segment.end -
-		                                       Gen5::g_pending_graphics_segment.start);
-	}
-
-	LOGF("\t flushing pending graphics segment before ACB: addr = 0x%016" PRIx64
-	     ", dw_num = 0x%08" PRIx32 "\n",
-	     reinterpret_cast<uint64_t>(dcb), size_in_dwords);
-
-	submit_dcb(dcb, size_in_dwords);
 }
 
 int KYTY_SYSV_ABI GraphicsDriverSubmitDcb(const Packet* packet) {
@@ -4074,8 +3877,6 @@ static void submit_acb(uint32_t queue, uint32_t* acb, uint32_t size_in_dwords) {
 		LOGF("\t acb[%u] = 0x%08" PRIx32 "\n", i, acb[i]);
 	}
 
-	flush_pending_graphics_segment_before_acb(acb, size_in_dwords);
-
 	GraphicsDbgDumpDcb("a", size_in_dwords, acb);
 
 	const bool trigger_interrupt_on_done = !dcb_has_queued_interrupt(acb, size_in_dwords);

src/libs/libNet.cpp

@@ -97,6 +97,10 @@ int KYTY_SYSV_ABI NetShutdown(int s, int how) {
 	return FinishSocketCall(Net::Shutdown(s, how));
 }
 
+int KYTY_SYSV_ABI NetGetsockname(int s, void* addr, uint32_t* addrlen) {
+	return FinishSocketCall(Net::Getsockname(s, addr, addrlen));
+}
+
 int KYTY_SYSV_ABI NetPoolCreate(const char* name, int size, int flags) {
 	return NET_CALL(Net::NetPoolCreate(name, size, flags));
 }
@@ -196,6 +200,7 @@ LIB_DEFINE(InitNet_1_Net) {
 	LIB_FUNC("v6M4txecCuo", LibNet::NetEtherNtostr);
 	LIB_FUNC("6Oc0bLsIYe0", LibNet::NetGetMacAddress);
 	LIB_FUNC("hLuXdjHnhiI", LibNet::NetGetSockInfo);
+	LIB_FUNC("hoOAofhhRvE", LibNet::NetGetsockname);
 	LIB_FUNC("SF47kB2MNTo", LibNet::NetEpollCreate);
 	LIB_FUNC("ZVw46bsasAk", LibNet::NetEpollControl);
 	LIB_FUNC("drjIbDbA7UQ", LibNet::NetEpollWait);

src/libs/libSystemService.cpp

@@ -1,6 +1,7 @@
 #include "common/abi.h"
 #include "common/assert.h"
 #include "common/common.h"
+#include "common/emulatorConfig.h"
 #include "common/logging/log.h"
 #include "common/stringUtils.h"
 #include "libs/errno.h"
@@ -25,37 +26,6 @@ namespace SystemService {
 [[maybe_unused]] constexpr int PARAM_ID_SCREEN_READER       = 208;
 [[maybe_unused]] constexpr int PARAM_ID_ENTER_BUTTON_ASSIGN = 1000;
 
-[[maybe_unused]] constexpr int PARAM_LANG_JAPANESE      = 0;
-[[maybe_unused]] constexpr int PARAM_LANG_ENGLISH_US    = 1;
-[[maybe_unused]] constexpr int PARAM_LANG_FRENCH        = 2;
-[[maybe_unused]] constexpr int PARAM_LANG_SPANISH       = 3;
-[[maybe_unused]] constexpr int PARAM_LANG_GERMAN        = 4;
-[[maybe_unused]] constexpr int PARAM_LANG_ITALIAN       = 5;
-[[maybe_unused]] constexpr int PARAM_LANG_DUTCH         = 6;
-[[maybe_unused]] constexpr int PARAM_LANG_PORTUGUESE_PT = 7;
-[[maybe_unused]] constexpr int PARAM_LANG_RUSSIAN       = 8;
-[[maybe_unused]] constexpr int PARAM_LANG_KOREAN        = 9;
-[[maybe_unused]] constexpr int PARAM_LANG_CHINESE_T     = 10;
-[[maybe_unused]] constexpr int PARAM_LANG_CHINESE_S     = 11;
-[[maybe_unused]] constexpr int PARAM_LANG_FINNISH       = 12;
-[[maybe_unused]] constexpr int PARAM_LANG_SWEDISH       = 13;
-[[maybe_unused]] constexpr int PARAM_LANG_DANISH        = 14;
-[[maybe_unused]] constexpr int PARAM_LANG_NORWEGIAN     = 15;
-[[maybe_unused]] constexpr int PARAM_LANG_POLISH        = 16;
-[[maybe_unused]] constexpr int PARAM_LANG_PORTUGUESE_BR = 17;
-[[maybe_unused]] constexpr int PARAM_LANG_ENGLISH_GB    = 18;
-[[maybe_unused]] constexpr int PARAM_LANG_TURKISH       = 19;
-[[maybe_unused]] constexpr int PARAM_LANG_SPANISH_LA    = 20;
-[[maybe_unused]] constexpr int PARAM_LANG_ARABIC        = 21;
-[[maybe_unused]] constexpr int PARAM_LANG_FRENCH_CA     = 22;
-[[maybe_unused]] constexpr int PARAM_LANG_CZECH         = 23;
-[[maybe_unused]] constexpr int PARAM_LANG_HUNGARIAN     = 24;
-[[maybe_unused]] constexpr int PARAM_LANG_GREEK         = 25;
-[[maybe_unused]] constexpr int PARAM_LANG_ROMANIAN      = 26;
-[[maybe_unused]] constexpr int PARAM_LANG_THAI          = 27;
-[[maybe_unused]] constexpr int PARAM_LANG_VIETNAMESE    = 28;
-[[maybe_unused]] constexpr int PARAM_LANG_INDONESIAN    = 29;
-
 [[maybe_unused]] constexpr int PARAM_DATE_FORMAT_YYYYMMDD = 0;
 [[maybe_unused]] constexpr int PARAM_DATE_FORMAT_DDMMYYYY = 1;
 [[maybe_unused]] constexpr int PARAM_DATE_FORMAT_MMDDYYYY = 2;
@@ -121,7 +91,7 @@ static int KYTY_SYSV_ABI SystemServiceParamGetInt(int param_id, int* value) {
 	int v = 0;
 
 	switch (param_id) {
-		case PARAM_ID_LANG: v = PARAM_LANG_ENGLISH_US; break;
+		case PARAM_ID_LANG: v = static_cast<int>(Config::GetConsoleLanguage()); break;
 		case PARAM_ID_DATE_FORMAT: v = PARAM_DATE_FORMAT_DDMMYYYY; break;
 		case PARAM_ID_TIME_FORMAT: v = PARAM_TIME_FORMAT_24HOUR; break;
 		case PARAM_ID_TIME_ZONE: v = +180; break;

src/main.cpp

@@ -10,6 +10,7 @@
 #include "emulator.h"
 #include "kytyGitVersion.h"
 
+#include <charconv>
 #include <cstdio>
 #include <fmt/format.h>
 
@@ -47,6 +48,7 @@ static void PrintUsage() {
 	::printf("  --screen-width <num>                 Window width. Default: 1280.\n");
 	::printf("  --screen-height <num>                Window height. Default: 720.\n");
 	::printf("  --vblank-frequency <num>             Virtual vblank frequency. Default: 60.\n");
+	::printf("  --console-language <0-29>            Console language. Default: 1 (English US).\n");
 	::printf("  --vulkan-validation <true|false>     Enable Vulkan validation.\n");
 	::printf("  --shader-validation <true|false>     Enable shader validation.\n");
 	::printf("  --shader-optimization-type <value>   None, Size, or Performance.\n");
@@ -59,8 +61,6 @@ static void PrintUsage() {
 	::printf("  --printf-output-file <path>          Guest printf output file.\n");
 	::printf("  --profiler-direction <value>         None or Network.\n");
 	::printf("  --spirv-debug-printf <true|false>    Enable SPIR-V debug printf.\n");
-	::printf("  --ngg-rectlist-draw <true|false>     Draw rect-list auto draws using the NGG "
-	         "4-vertex path.\n");
 	::printf(
 	    "  --readback-linear-images <true|false> Read back writable linear images on submit.\n");
 	::printf("  --rd                                 Enable RenderDoc capture.\n");
@@ -103,6 +103,17 @@ static bool ParseEnum(const std::string& value, E& out) {
 	return true;
 }
 
+static bool ParseConsoleLanguage(const std::string& value, uint32_t& out) {
+	uint32_t language = 0;
+	auto [end, error] = std::from_chars(value.data(), value.data() + value.size(), language);
+	if (error != std::errc {} || end != value.data() + value.size() ||
+	    language > Config::MAX_CONSOLE_LANGUAGE) {
+		return false;
+	}
+	out = language;
+	return true;
+}
+
 static bool ParseArgs(int argc, char* argv[], RunOptions& options, bool& show_help) {
 	show_help = false;
 
@@ -169,6 +180,11 @@ static bool ParseArgs(int argc, char* argv[], RunOptions& options, bool& show_he
 			const int32_t vblank_frequency = Common::ToInt32(value);
 			options.config.vblank_frequency =
 			    static_cast<uint32_t>(vblank_frequency < 0 ? 0 : vblank_frequency);
+		} else if (arg == "--console-language") {
+			if (!ParseConsoleLanguage(value, options.config.console_language)) {
+				::printf("invalid console language: %s\n", value.c_str());
+				return false;
+			}
 		} else if (arg == "--vulkan-validation") {
 			if (!ParseBool(value, options.config.vulkan_validation_enabled)) {
 				::printf("invalid boolean for %s: %s\n", arg.c_str(), value.c_str());
@@ -220,11 +236,6 @@ static bool ParseArgs(int argc, char* argv[], RunOptions& options, bool& show_he
 				::printf("invalid boolean for %s: %s\n", arg.c_str(), value.c_str());
 				return false;
 			}
-		} else if (arg == "--ngg-rectlist-draw") {
-			if (!ParseBool(value, options.config.ngg_rectlist_draw_enabled)) {
-				::printf("invalid boolean for %s: %s\n", arg.c_str(), value.c_str());
-				return false;
-			}
 		} else if (arg == "--readback-linear-images") {
 			if (!ParseBool(value, options.config.readback_linear_images)) {
 				::printf("invalid boolean for %s: %s\n", arg.c_str(), value.c_str());

tests/ShaderRecompilerComputeTests.cpp

@@ -39,6 +39,7 @@
 #include "graphics/shader/recompiler/emitter/SpirvBuilder.h"
 #include "graphics/shader/recompiler/emitter/SpirvEmitter.h"
 #include "graphics/shader/recompiler/ir/BindingLayout.h"
+#include "graphics/shader/rectListShader.h"
 #include "graphics/shader/shader.h"
 #include "kernel/memory.h"
 #include "spirv-tools/libspirv.hpp"
@@ -58,6 +59,7 @@
 #include <algorithm>
 #include <array>
 #include <bit>
+#include <chrono>
 #include <cinttypes>
 #include <cmath>
 #include <cstdint>
@@ -770,6 +772,91 @@ void ValidateSpirv(const char* shader_name, const std::vector<u32>& spirv) {
 	}
 }
 
+size_t CountText(const std::string& text, const std::string& needle) {
+	size_t count = 0;
+	for (size_t offset = 0; (offset = text.find(needle, offset)) != std::string::npos;
+	     offset += needle.size()) {
+		count++;
+	}
+	return count;
+}
+
+void CheckRectListShaders() {
+	constexpr const char* name = "RectListShaders";
+
+	auto program = std::make_shared<ShaderRecompiler::IR::Program>();
+	program->info.inputs.push_back(
+	    {ShaderRecompiler::IR::StageInputKind::Parameter, 0, 4, "in_param_0"});
+	program->info.inputs.push_back(
+	    {ShaderRecompiler::IR::StageInputKind::Parameter, 1, 4, "in_param_1"});
+
+	ShaderVertexInputInfo vertex {};
+	vertex.param_export_mask = 1u;
+	ShaderPixelInputInfo pixel {};
+	pixel.input_num                = 2;
+	pixel.interpolator_settings[0] = 0x400u;
+	pixel.interpolator_settings[1] = 0;
+	pixel.stage.program            = program;
+	HW::PixelShaderInfo ps_regs {};
+	const auto          perspective_id = ShaderGetIdPS(ps_regs, pixel, false);
+	pixel.ps_no_perspective            = true;
+	const auto no_perspective_id       = ShaderGetIdPS(ps_regs, pixel, false);
+	pixel.ps_no_perspective            = false;
+	Require(name, "pipeline identity", perspective_id != no_perspective_id,
+	        "pixel interpolation mode must participate in the shader and pipeline key");
+
+	const std::array<uint32_t, 2> active_inputs = {0, 1};
+	Require(name, "duplicate mapping",
+	        ShaderPixelParameterLocation(pixel, active_inputs, 0) == 0 &&
+	            ShaderPixelParameterLocation(pixel, active_inputs, 1) == 1,
+	        "duplicate pixel mappings must receive distinct effective output locations");
+
+	const auto shaders = BuildRectListShaders(vertex, &pixel);
+	Require(name, "SPIR-V version",
+	        shaders.control.size() > 1 && shaders.control[1] == 0x00010500u &&
+	            shaders.evaluation.size() > 1 && shaders.evaluation[1] == 0x00010500u,
+	        "shadPS4-compatible vector selection requires SPIR-V 1.5");
+	ValidateSpirv(name, shaders.control);
+	ValidateSpirv(name, shaders.evaluation);
+
+	spvtools::SpirvTools tools(SPV_ENV_VULKAN_1_2);
+	std::string          control_text;
+	std::string          evaluation_text;
+	Require(name, "control disassembly", tools.Disassemble(shaders.control, &control_text),
+	        "failed to disassemble rectangle-list tessellation control shader");
+	Require(name, "evaluation disassembly",
+	        tools.Disassemble(shaders.evaluation, &evaluation_text),
+	        "failed to disassemble rectangle-list tessellation evaluation shader");
+	Require(name, "control execution mode",
+	        control_text.find("TessellationControl") != std::string::npos &&
+	            control_text.find("OutputVertices 4") != std::string::npos,
+	        "rectangle-list control shader must produce four control points");
+	Require(name, "evaluation execution modes",
+	        evaluation_text.find("TessellationEvaluation") != std::string::npos &&
+	            evaluation_text.find("Quads") != std::string::npos &&
+	            evaluation_text.find("SpacingEqual") != std::string::npos &&
+	            evaluation_text.find("VertexOrderCw") != std::string::npos,
+	        "rectangle-list evaluation shader has the wrong patch modes");
+	Require(name, "no geometry stage",
+	        control_text.find("Geometry") == std::string::npos &&
+	            evaluation_text.find("Geometry") == std::string::npos,
+	        "rectangle-list expansion must not use geometry shaders");
+	Require(name, "flat broadcast",
+	        CountText(control_text, "OpVectorTimesScalar") == 6 &&
+	            CountText(control_text, "OpSelect %v4float") == 2,
+	        "flat parameters must use guest vertex zero instead of reconstructed values");
+	Require(name, "remapped interface",
+	        CountText(control_text, " Location 0") == 2 &&
+	            CountText(control_text, " Location 1") == 1 &&
+	            CountText(evaluation_text, " Location 0") == 2 &&
+	            CountText(evaluation_text, " Location 1") == 2,
+	        "duplicate pixel mappings must share one vertex input and keep distinct patch outputs");
+
+	const auto position_only = BuildRectListShaders(vertex, nullptr);
+	ValidateSpirv(name, position_only.control);
+	ValidateSpirv(name, position_only.evaluation);
+}
+
 void CheckSpirvText(const TestCase& test, const std::vector<u32>& spirv) {
 	if (test.required_spirv.empty() && test.forbidden_spirv.empty()) {
 		return;
@@ -1450,22 +1537,91 @@ public:
 		uint32_t              ordered_suffix = 0;
 		std::jthread          ordered([&] {
 			ordered_started.release();
-			Gpu::SubmissionLock submissions(gpu);
-			gpu.SendCommandSync([&] {
-				ordered_suffix   = suffix;
-				ordered_finished = true;
-			});
+			gpu.Done();
+			ordered_suffix   = suffix;
+			ordered_finished = true;
 		});
 		ordered_started.acquire();
 		gpu.SendCommandSync([&] {
-			Require("GpuCommandLane", "ordered barrier", !ordered_finished.load(),
-			        "ordered host command overtook a queued submission");
+			Require("GpuCommandLane", "submit done barrier", !ordered_finished.load(),
+			        "submit done returned before a queued submission");
 			label = 1;
 		});
 		ordered.join();
 		Require("GpuCommandLane", "ordered completion",
 		        prefix == 11 && suffix == 22 && ordered_suffix == 22 && ordered_finished.load(),
-		        "submission barrier did not drain prior PM4 work");
+		        "submit done did not drain prior PM4 work");
+
+		auto&              resources         = context.GetGpuResources();
+		constexpr uint64_t empty_unmap_base = 0x0000000200400000ull;
+		constexpr uint64_t empty_unmap_size = 0x4000;
+		resources.MapMemory(empty_unmap_base, empty_unmap_size);
+		label  = 0;
+		prefix = 0;
+		suffix = 0;
+		gpu.Submit(commands.data(), static_cast<uint32_t>(commands.size()), nullptr, 0);
+
+		std::binary_semaphore unmap_complete {0};
+		std::jthread unmap_thread([&] {
+			resources.UnmapMemory(empty_unmap_base, empty_unmap_size);
+			unmap_complete.release();
+		});
+		const bool unmap_returned = unmap_complete.try_acquire_for(std::chrono::seconds(2));
+		gpu.SendCommandSync([&] { label = 1; });
+		if (!unmap_returned) {
+			unmap_complete.acquire();
+		}
+		unmap_thread.join();
+		gpu.Done();
+		Require("GpuCommandLane", "unmap queue progress",
+		        unmap_returned && !resources.IsMapped(empty_unmap_base, empty_unmap_size) &&
+		            prefix == 11 && suffix == 22,
+		        "an unrelated unmap waited for a blocked PM4 submission");
+
+		auto&                 scheduler = context.GetCommandScheduler();
+		std::atomic<bool>      normal_completed {false};
+		gpu.SendCommandSync(
+		    [&] { scheduler.DeferOperation([&] { normal_completed = true; }); });
+		resources.MapMemory(empty_unmap_base, empty_unmap_size);
+		resources.UnmapMemory(empty_unmap_base, empty_unmap_size);
+		Require("GpuCommandLane", "unmap native completion",
+		        normal_completed.load() &&
+		            !resources.IsMapped(empty_unmap_base, empty_unmap_size),
+		        "unmap returned before an earlier native guest-memory callback");
+
+		std::binary_semaphore priority_entered {0};
+		std::binary_semaphore release_priority {0};
+		gpu.SendCommandSync([&] {
+			scheduler.DeferPriorityOperation([&] {
+				priority_entered.release();
+				release_priority.acquire();
+			});
+			scheduler.Flush();
+		});
+		priority_entered.acquire();
+		resources.MapMemory(empty_unmap_base, empty_unmap_size);
+
+		std::binary_semaphore priority_unmap_entered {0};
+		std::binary_semaphore priority_unmap_complete {0};
+		std::jthread priority_unmap_thread([&] {
+			gpu.SendCommandSync([&] {
+				priority_unmap_entered.release();
+				resources.UnmapMemory(empty_unmap_base, empty_unmap_size);
+			});
+			priority_unmap_complete.release();
+		});
+		priority_unmap_entered.acquire();
+		const bool unmap_overtook_priority =
+		    priority_unmap_complete.try_acquire_for(std::chrono::seconds(1));
+		release_priority.release();
+		if (!unmap_overtook_priority) {
+			priority_unmap_complete.acquire();
+		}
+		priority_unmap_thread.join();
+		Require("GpuCommandLane", "unmap priority ordering",
+		        !unmap_overtook_priority &&
+		            !resources.IsMapped(empty_unmap_base, empty_unmap_size),
+		        "unmap returned before an earlier guest-memory callback");
 
 		constexpr uintptr_t fault_base          = 0x0000000200500000ull;
 		constexpr uint64_t  fault_size          = 0x10000;
@@ -1483,7 +1639,6 @@ public:
 		Require("GpuCommandLane", "processor fault allocation",
 		        fault_memory == reinterpret_cast<void*>(fault_base),
 		        "fixed processor-fault allocation failed");
-		auto& resources = context.GetGpuResources();
 		resources.MapMemory(fault_base, fault_size);
 
 		constexpr uint64_t                immediate_dst         = fault_base + 0x1000;
@@ -1530,9 +1685,7 @@ public:
 		Require("GpuCommandLane", "DMA_DATA packet assembly", dma_cursor == dma_commands.size(),
 		        "DMA_DATA GDS packet stream has the wrong size");
 		gpu.Submit(dma_commands.data(), static_cast<uint32_t>(dma_commands.size()), nullptr, 0);
-		{
-			Gpu::SubmissionLock submissions(gpu);
-		}
+		gpu.Done();
 		constexpr uint32_t clean_fill_value = 0xdecafbad;
 		gpu.SendCommandSyncWithProcessor([&](CommandProcessor&) {
 			auto& buffer_cache = resources.GetBufferCache();
@@ -5453,6 +5606,43 @@ public:
 			    "state at the final acquisition boundary");
 			RenderExecutorTestAccess::ResetBindings(executor);
 
+			constexpr uint64_t depth_only_address = base + 0x140000;
+			HW::DepthRenderTarget depth_only_target {};
+			depth_only_target.z_info.format =
+			    Prospero::GpuEnumValue(Prospero::DepthFormat::kZ32F);
+			depth_only_target.z_info.z_compare_base = Prospero::ZCompareBase::kZMax;
+			depth_only_target.stencil_info.htile_stencil_disabled = true;
+			depth_only_target.z_read_base_addr                    = depth_only_address;
+			depth_only_target.z_write_base_addr                   = depth_only_address;
+			depth_only_target.size                                = {63, 63, true};
+			registers.SetDepthRenderTarget(depth_only_target);
+			HW::DepthControl depth_only_control {};
+			depth_only_control.stencil_enable = true;
+			depth_only_control.z_enable       = true;
+			depth_only_control.z_write_enable = true;
+			depth_only_control.zfunc          = static_cast<uint8_t>(vk::CompareOp::eAlways);
+			registers.SetDepthControl(depth_only_control);
+			HW::RenderControl depth_only_render_control {};
+			depth_only_render_control.depth_clear_enable   = true;
+			depth_only_render_control.stencil_clear_enable = true;
+			registers.SetRenderControl(depth_only_render_control);
+			RenderDepthInfo depth_only {};
+			RenderExecutorTestAccess::ResolveRenderDepthTarget(executor, 1, scheduler.Current(),
+			                                                   depth_only);
+			Require(
+			    name, "depth-only target with stale stencil state",
+			    depth_only.image_id && depth_only.format == vk::Format::eD32Sfloat &&
+			        depth_only.depth_test_enable && depth_only.depth_write_enable &&
+			        depth_only.depth_clear_enable && !depth_only.stencil_test_enable &&
+			        !depth_only.stencil_clear_enable && depth_only.stencil_buffer_vaddr == 0 &&
+			        depth_only.stencil_buffer_size == 0 && depth_only.desc.info.stencil.Empty() &&
+			        depth_only.depth_buffer_size != 0 &&
+			        depth_only_address + depth_only.depth_buffer_size <= base + allocation_size &&
+			        depth_only.vaddr_num == 1 &&
+			        depth_only.AttachmentWriteAspects() == vk::ImageAspectFlagBits::eDepth,
+			    "raw stencil test or clear state leaked into a depth-only attachment");
+			RenderExecutorTestAccess::ResetBindings(executor);
+
 			auto video_subresource = make_target_desc(base + 0x20000, target_mip_size, {1, 1, 1});
 			video_subresource.type = BindingType::VideoOut;
 			video_subresource.info.pixel_format = vk::Format::eR8G8B8A8Srgb;
@@ -7175,26 +7365,30 @@ public:
 		}
 
 		u32  case_index       = 0;
-		auto check_round_trip = [&](const char* stage, uint64_t size,
+		auto check_round_trip = [&](const char* stage, uint64_t tiled_size,
 		                            std::span<const GpuTileInfo> infos) {
-			std::vector<uint8_t> tiled(size);
-			std::vector<uint8_t> cpu(size, 0);
-			std::vector<uint8_t> gpu(size, 0xab);
+			uint64_t linear_size = 0;
+			for (const auto& info: infos) {
+				linear_size = std::max(linear_size, info.linear_offset + info.linear_size);
+			}
+			std::vector<uint8_t> tiled(tiled_size);
+			std::vector<uint8_t> cpu(linear_size, 0);
+			std::vector<uint8_t> gpu(linear_size, 0xab);
 			fill(&tiled, ++case_index);
 			for (const auto& info: infos) {
 				convert_reference(false, &cpu, tiled, info);
 			}
-			gpu_detile(tiled, &gpu, size, size, infos);
+			gpu_detile(tiled, &gpu, tiled_size, linear_size, infos);
 			compare((std::string(stage) + " detile bytes").c_str(), cpu, gpu);
 
-			std::vector<uint8_t> linear(size);
-			std::vector<uint8_t> cpu_tiled(size, 0xab);
-			std::vector<uint8_t> gpu_tiled(size, 0xab);
+			std::vector<uint8_t> linear(linear_size);
+			std::vector<uint8_t> cpu_tiled(tiled_size, 0xab);
+			std::vector<uint8_t> gpu_tiled(tiled_size, 0xab);
 			fill(&linear, 0x280u + case_index);
 			for (const auto& info: infos) {
 				convert_reference(true, &cpu_tiled, linear, info);
 			}
-			gpu_tile(linear, &gpu_tiled, size, size, infos);
+			gpu_tile(linear, &gpu_tiled, tiled_size, linear_size, infos);
 			compare((std::string(stage) + " tile bytes").c_str(), cpu_tiled, gpu_tiled);
 		};
 		for (const auto family: families) {
@@ -7332,6 +7526,32 @@ public:
 		Require(name, "format coverage", format_cases != 0,
 		        "no CPU-supported standard formats were tested");
 
+		{
+			constexpr u32 format = Prospero::GpuEnumValue(Prospero::BufferFormat::kBc1UNorm);
+			constexpr u32 tile = Prospero::GpuEnumValue(Prospero::TileMode::kStandard64KB);
+			constexpr u32 width = 256, height = 256, levels = 9;
+			const u32     pitch = TileGetTexturePitch(format, width, levels, tile);
+			TileSizeAlign total {};
+			TileGetTextureSize(format, width, height, pitch, levels, tile, &total, nullptr,
+			                   nullptr);
+			const auto layout = TextureCalcUploadLayout(format, width, height, levels, 1, pitch,
+			                                            tile, total.size, false, false, name);
+			const auto regions =
+			    TextureBuildImageCopies(layout, width, height, 1, levels, false, false);
+			std::vector<GpuTileInfo> infos;
+			const bool built =
+			    TextureBuildGpuTileInfos(total.size, regions, layout, format, 1, levels, infos);
+			uint64_t linear_size = 0;
+			for (const auto& info: infos) {
+				linear_size = std::max(linear_size, info.linear_offset + info.linear_size);
+			}
+			Require(name, "BC1 mip-tail capacities",
+			        built && total.size == 0x10000 && layout.first_tail_level == 0 &&
+			            linear_size == 0x15560 && linear_size > total.size,
+			        "BC1 mip tail conflated tiled and linear capacities");
+			check_round_trip("BC1 mip tail", total.size, infos);
+		}
+
 		{
 			constexpr u32 format = Prospero::GpuEnumValue(Prospero::BufferFormat::k32Float);
 			constexpr u32 tile   = Prospero::GpuEnumValue(Prospero::TileMode::kRenderTarget);
@@ -15855,11 +16075,6 @@ ShaderTextureResource AtomicStorageTextureDescriptor() {
 		resource   = BasicArrayStorageTextureResource();
 		descriptor = BasicArrayStorageTextureDescriptor();
 		descriptor.fields[4] |= 1u << 16u;
-	} else if (std::strcmp(kind, "array-mip-view") == 0) {
-		resource   = BasicArrayStorageTextureResource();
-		descriptor = BasicArrayStorageTextureDescriptor();
-		descriptor.fields[3] |= (1u << 12u) | (1u << 16u);
-		descriptor.fields[5] |= 1u << 4u;
 	} else if (std::strcmp(kind, "reserved") == 0) {
 		descriptor.fields[1] |= 1u << 29u;
 	} else if (std::strcmp(kind, "uint-format") == 0) {
@@ -16035,6 +16250,16 @@ void CheckBasicStorageTextureDescriptor() {
 	            array.DstSelXYZW() == DstSel(6, 5, 4, 7),
 	        "PPSA21268 2D-array storage descriptor fixture is malformed");
 	ValidateStorageTexture(BasicArrayStorageTextureResource(), array, 0x10000);
+	const ShaderTextureResource mip_array {{0x20268d00u, 0xc4700000u, 0x001fc01fu,
+	                                        0xd1b11facu, 0x00000000u, 0x00700070u,
+	                                        0x00000000u, 0x00000000u}};
+	Require("BasicStorageTexture", "PPSA14457 mip-one 2D-array descriptor",
+	        mip_array.BaseLevel() == 1 && mip_array.LastLevel() == 1 &&
+	            mip_array.MaxMip() == 7 &&
+	            mip_array.Type() == Prospero::GpuEnumValue(Prospero::ImageType::kColor2DArray) &&
+	            mip_array.Depth() == 0 && mip_array.BaseArray5() == 0,
+	        "PPSA14457 mip-one 2D-array storage descriptor fixture is malformed");
+	ValidateStorageTexture(BasicArrayStorageTextureResource(), mip_array, 0x30000);
 
 	const auto uint_array = BasicUintArrayStorageTextureDescriptor();
 	Require("BasicStorageTexture", "uint 2D-array descriptor",
@@ -16195,7 +16420,6 @@ void CheckBasicStorageTextureDescriptor() {
 	                        "yzwx-read",
 	                        "reserved-swizzle",
 	                        "array-base-out-of-range",
-	                        "array-mip-view",
 	                        "reserved",
 	                        "uint-format",
 	                        "uint-resource-float-format",
@@ -17364,6 +17588,7 @@ int main(int argc, char** argv) {
 	CheckPm4CeCompletion(vulkan.RuntimeRenderer());
 	CheckEmbeddedFetchVertexOffset();
 	CheckEmbeddedFetchLaneSpill();
+	CheckRectListShaders();
 	CheckPs5GameExampleImageClearRuntimeShape();
 	vulkan.CheckSchedulerTimeline();
 	vulkan.CheckGpuMappedRangeLifecycle();