summaryrefslogtreecommitdiffstats
path: root/src
diff options
context:
space:
mode:
Diffstat (limited to 'src')
-rw-r--r--src/common/telemetry.cpp1
-rw-r--r--src/common/x64/cpu_detect.cpp5
-rw-r--r--src/common/x64/cpu_detect.h1
-rw-r--r--src/core/arm/dynarmic/arm_dynarmic_32.cpp2
-rw-r--r--src/core/arm/dynarmic/arm_dynarmic_64.cpp4
-rw-r--r--src/core/hle/service/acc/acc.cpp341
-rw-r--r--src/core/hle/service/acc/acc_aa.cpp4
-rw-r--r--src/core/hle/service/acc/acc_su.cpp34
-rw-r--r--src/core/hle/service/acc/acc_u0.cpp18
-rw-r--r--src/core/hle/service/acc/acc_u1.cpp29
-rw-r--r--src/core/hle/service/am/applets/software_keyboard.cpp4
-rw-r--r--src/core/hle/service/hid/controllers/npad.cpp8
-rw-r--r--src/core/hle/service/hid/controllers/npad.h10
-rw-r--r--src/core/hle/service/hid/hid.cpp42
-rw-r--r--src/core/hle/service/hid/hid.h2
-rw-r--r--src/core/hle/service/lm/manager.cpp3
-rw-r--r--src/video_core/CMakeLists.txt2
-rw-r--r--src/video_core/buffer_cache/buffer_cache.h44
-rw-r--r--src/video_core/engines/maxwell_3d.cpp2
-rw-r--r--src/video_core/engines/maxwell_3d.h8
-rw-r--r--src/video_core/macro/macro.cpp35
-rw-r--r--src/video_core/macro/macro.h19
-rw-r--r--src/video_core/macro/macro_hle.cpp113
-rw-r--r--src/video_core/macro/macro_hle.h44
-rw-r--r--src/video_core/macro/macro_interpreter.cpp3
-rw-r--r--src/video_core/macro/macro_jit_x64.cpp3
-rw-r--r--src/video_core/renderer_opengl/gl_buffer_cache.cpp62
-rw-r--r--src/video_core/renderer_opengl/gl_buffer_cache.h36
-rw-r--r--src/video_core/renderer_opengl/gl_device.cpp28
-rw-r--r--src/video_core/renderer_opengl/gl_device.h10
-rw-r--r--src/video_core/renderer_opengl/gl_rasterizer.cpp70
-rw-r--r--src/video_core/renderer_opengl/gl_shader_decompiler.cpp50
-rw-r--r--src/video_core/renderer_opengl/gl_stream_buffer.cpp56
-rw-r--r--src/video_core/renderer_opengl/gl_stream_buffer.h16
-rw-r--r--src/video_core/renderer_opengl/renderer_opengl.cpp17
-rw-r--r--src/video_core/renderer_opengl/renderer_opengl.h3
-rw-r--r--src/video_core/renderer_vulkan/vk_buffer_cache.cpp89
-rw-r--r--src/video_core/renderer_vulkan/vk_buffer_cache.h29
-rw-r--r--src/video_core/renderer_vulkan/vk_rasterizer.cpp31
-rw-r--r--src/video_core/renderer_vulkan/vk_stream_buffer.h6
-rw-r--r--src/video_core/shader/decode/image.cpp26
-rw-r--r--src/video_core/texture_cache/texture_cache.h2
-rw-r--r--src/yuzu/main.cpp15
-rw-r--r--src/yuzu_tester/service/yuzutest.cpp2
44 files changed, 1033 insertions, 296 deletions
diff --git a/src/common/telemetry.cpp b/src/common/telemetry.cpp
index 200c6489a..16d42facd 100644
--- a/src/common/telemetry.cpp
+++ b/src/common/telemetry.cpp
@@ -60,6 +60,7 @@ void AppendCPUInfo(FieldCollection& fc) {
fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AES", Common::GetCPUCaps().aes);
fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AVX", Common::GetCPUCaps().avx);
fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AVX2", Common::GetCPUCaps().avx2);
+ fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_AVX512", Common::GetCPUCaps().avx512);
fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_BMI1", Common::GetCPUCaps().bmi1);
fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_BMI2", Common::GetCPUCaps().bmi2);
fc.AddField(FieldType::UserSystem, "CPU_Extension_x64_FMA", Common::GetCPUCaps().fma);
diff --git a/src/common/x64/cpu_detect.cpp b/src/common/x64/cpu_detect.cpp
index c9349a6b4..f35dcb498 100644
--- a/src/common/x64/cpu_detect.cpp
+++ b/src/common/x64/cpu_detect.cpp
@@ -110,6 +110,11 @@ static CPUCaps Detect() {
caps.bmi1 = true;
if ((cpu_id[1] >> 8) & 1)
caps.bmi2 = true;
+ // Checks for AVX512F, AVX512CD, AVX512VL, AVX512DQ, AVX512BW (Intel Skylake-X/SP)
+ if ((cpu_id[1] >> 16) & 1 && (cpu_id[1] >> 28) & 1 && (cpu_id[1] >> 31) & 1 &&
+ (cpu_id[1] >> 17) & 1 && (cpu_id[1] >> 30) & 1) {
+ caps.avx512 = caps.avx2;
+ }
}
}
diff --git a/src/common/x64/cpu_detect.h b/src/common/x64/cpu_detect.h
index 20f2ba234..7606c3f7b 100644
--- a/src/common/x64/cpu_detect.h
+++ b/src/common/x64/cpu_detect.h
@@ -19,6 +19,7 @@ struct CPUCaps {
bool lzcnt;
bool avx;
bool avx2;
+ bool avx512;
bool bmi1;
bool bmi2;
bool fma;
diff --git a/src/core/arm/dynarmic/arm_dynarmic_32.cpp b/src/core/arm/dynarmic/arm_dynarmic_32.cpp
index 19d798dc7..4c8663d03 100644
--- a/src/core/arm/dynarmic/arm_dynarmic_32.cpp
+++ b/src/core/arm/dynarmic/arm_dynarmic_32.cpp
@@ -62,7 +62,7 @@ public:
case Dynarmic::A32::Exception::Breakpoint:
break;
}
- LOG_CRITICAL(HW_GPU, "ExceptionRaised(exception = {}, pc = {:08X}, code = {:08X})",
+ LOG_CRITICAL(Core_ARM, "ExceptionRaised(exception = {}, pc = {:08X}, code = {:08X})",
static_cast<std::size_t>(exception), pc, MemoryReadCode(pc));
UNIMPLEMENTED();
}
diff --git a/src/core/arm/dynarmic/arm_dynarmic_64.cpp b/src/core/arm/dynarmic/arm_dynarmic_64.cpp
index 337b97be9..5f5e36d94 100644
--- a/src/core/arm/dynarmic/arm_dynarmic_64.cpp
+++ b/src/core/arm/dynarmic/arm_dynarmic_64.cpp
@@ -98,8 +98,8 @@ public:
}
[[fallthrough]];
default:
- ASSERT_MSG(false, "ExceptionRaised(exception = {}, pc = {:X})",
- static_cast<std::size_t>(exception), pc);
+ ASSERT_MSG(false, "ExceptionRaised(exception = {}, pc = {:08X}, code = {:08X})",
+ static_cast<std::size_t>(exception), pc, MemoryReadCode(pc));
}
}
diff --git a/src/core/hle/service/acc/acc.cpp b/src/core/hle/service/acc/acc.cpp
index 630a8b048..94d8c1fc6 100644
--- a/src/core/hle/service/acc/acc.cpp
+++ b/src/core/hle/service/acc/acc.cpp
@@ -44,6 +44,218 @@ static constexpr u32 SanitizeJPEGSize(std::size_t size) {
return static_cast<u32>(std::min(size, max_jpeg_image_size));
}
+class IManagerForSystemService final : public ServiceFramework<IManagerForSystemService> {
+public:
+ explicit IManagerForSystemService(Common::UUID user_id)
+ : ServiceFramework("IManagerForSystemService") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "CheckAvailability"},
+ {1, nullptr, "GetAccountId"},
+ {2, nullptr, "EnsureIdTokenCacheAsync"},
+ {3, nullptr, "LoadIdTokenCache"},
+ {100, nullptr, "SetSystemProgramIdentification"},
+ {101, nullptr, "RefreshNotificationTokenAsync"}, // 7.0.0+
+ {110, nullptr, "GetServiceEntryRequirementCache"}, // 4.0.0+
+ {111, nullptr, "InvalidateServiceEntryRequirementCache"}, // 4.0.0+
+ {112, nullptr, "InvalidateTokenCache"}, // 4.0.0 - 6.2.0
+ {113, nullptr, "GetServiceEntryRequirementCacheForOnlinePlay"}, // 6.1.0+
+ {120, nullptr, "GetNintendoAccountId"},
+ {121, nullptr, "CalculateNintendoAccountAuthenticationFingerprint"}, // 9.0.0+
+ {130, nullptr, "GetNintendoAccountUserResourceCache"},
+ {131, nullptr, "RefreshNintendoAccountUserResourceCacheAsync"},
+ {132, nullptr, "RefreshNintendoAccountUserResourceCacheAsyncIfSecondsElapsed"},
+ {133, nullptr, "GetNintendoAccountVerificationUrlCache"}, // 9.0.0+
+ {134, nullptr, "RefreshNintendoAccountVerificationUrlCache"}, // 9.0.0+
+ {135, nullptr, "RefreshNintendoAccountVerificationUrlCacheAsyncIfSecondsElapsed"}, // 9.0.0+
+ {140, nullptr, "GetNetworkServiceLicenseCache"}, // 5.0.0+
+ {141, nullptr, "RefreshNetworkServiceLicenseCacheAsync"}, // 5.0.0+
+ {142, nullptr, "RefreshNetworkServiceLicenseCacheAsyncIfSecondsElapsed"}, // 5.0.0+
+ {150, nullptr, "CreateAuthorizationRequest"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+// 3.0.0+
+class IFloatingRegistrationRequest final : public ServiceFramework<IFloatingRegistrationRequest> {
+public:
+ explicit IFloatingRegistrationRequest(Common::UUID user_id)
+ : ServiceFramework("IFloatingRegistrationRequest") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "GetSessionId"},
+ {12, nullptr, "GetAccountId"},
+ {13, nullptr, "GetLinkedNintendoAccountId"},
+ {14, nullptr, "GetNickname"},
+ {15, nullptr, "GetProfileImage"},
+ {21, nullptr, "LoadIdTokenCache"},
+ {100, nullptr, "RegisterUser"}, // [1.0.0-3.0.2] RegisterAsync
+ {101, nullptr, "RegisterUserWithUid"}, // [1.0.0-3.0.2] RegisterWithUidAsync
+ {102, nullptr, "RegisterNetworkServiceAccountAsync"}, // 4.0.0+
+ {103, nullptr, "RegisterNetworkServiceAccountWithUidAsync"}, // 4.0.0+
+ {110, nullptr, "SetSystemProgramIdentification"},
+ {111, nullptr, "EnsureIdTokenCacheAsync"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class IAdministrator final : public ServiceFramework<IAdministrator> {
+public:
+ explicit IAdministrator(Common::UUID user_id) : ServiceFramework("IAdministrator") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "CheckAvailability"},
+ {1, nullptr, "GetAccountId"},
+ {2, nullptr, "EnsureIdTokenCacheAsync"},
+ {3, nullptr, "LoadIdTokenCache"},
+ {100, nullptr, "SetSystemProgramIdentification"},
+ {101, nullptr, "RefreshNotificationTokenAsync"}, // 7.0.0+
+ {110, nullptr, "GetServiceEntryRequirementCache"}, // 4.0.0+
+ {111, nullptr, "InvalidateServiceEntryRequirementCache"}, // 4.0.0+
+ {112, nullptr, "InvalidateTokenCache"}, // 4.0.0 - 6.2.0
+ {113, nullptr, "GetServiceEntryRequirementCacheForOnlinePlay"}, // 6.1.0+
+ {120, nullptr, "GetNintendoAccountId"},
+ {121, nullptr, "CalculateNintendoAccountAuthenticationFingerprint"}, // 9.0.0+
+ {130, nullptr, "GetNintendoAccountUserResourceCache"},
+ {131, nullptr, "RefreshNintendoAccountUserResourceCacheAsync"},
+ {132, nullptr, "RefreshNintendoAccountUserResourceCacheAsyncIfSecondsElapsed"},
+ {133, nullptr, "GetNintendoAccountVerificationUrlCache"}, // 9.0.0+
+ {134, nullptr, "RefreshNintendoAccountVerificationUrlCacheAsync"}, // 9.0.0+
+ {135, nullptr, "RefreshNintendoAccountVerificationUrlCacheAsyncIfSecondsElapsed"}, // 9.0.0+
+ {140, nullptr, "GetNetworkServiceLicenseCache"}, // 5.0.0+
+ {141, nullptr, "RefreshNetworkServiceLicenseCacheAsync"}, // 5.0.0+
+ {142, nullptr, "RefreshNetworkServiceLicenseCacheAsyncIfSecondsElapsed"}, // 5.0.0+
+ {150, nullptr, "CreateAuthorizationRequest"},
+ {200, nullptr, "IsRegistered"},
+ {201, nullptr, "RegisterAsync"},
+ {202, nullptr, "UnregisterAsync"},
+ {203, nullptr, "DeleteRegistrationInfoLocally"},
+ {220, nullptr, "SynchronizeProfileAsync"},
+ {221, nullptr, "UploadProfileAsync"},
+ {222, nullptr, "SynchronizaProfileAsyncIfSecondsElapsed"},
+ {250, nullptr, "IsLinkedWithNintendoAccount"},
+ {251, nullptr, "CreateProcedureToLinkWithNintendoAccount"},
+ {252, nullptr, "ResumeProcedureToLinkWithNintendoAccount"},
+ {255, nullptr, "CreateProcedureToUpdateLinkageStateOfNintendoAccount"},
+ {256, nullptr, "ResumeProcedureToUpdateLinkageStateOfNintendoAccount"},
+ {260, nullptr, "CreateProcedureToLinkNnidWithNintendoAccount"}, // 3.0.0+
+ {261, nullptr, "ResumeProcedureToLinkNnidWithNintendoAccount"}, // 3.0.0+
+ {280, nullptr, "ProxyProcedureToAcquireApplicationAuthorizationForNintendoAccount"},
+ {290, nullptr, "GetRequestForNintendoAccountUserResourceView"}, // 8.0.0+
+ {300, nullptr, "TryRecoverNintendoAccountUserStateAsync"}, // 6.0.0+
+ {400, nullptr, "IsServiceEntryRequirementCacheRefreshRequiredForOnlinePlay"}, // 6.1.0+
+ {401, nullptr, "RefreshServiceEntryRequirementCacheForOnlinePlayAsync"}, // 6.1.0+
+ {900, nullptr, "GetAuthenticationInfoForWin"}, // 9.0.0+
+ {901, nullptr, "ImportAsyncForWin"}, // 9.0.0+
+ {997, nullptr, "DebugUnlinkNintendoAccountAsync"},
+ {998, nullptr, "DebugSetAvailabilityErrorDetail"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class IAuthorizationRequest final : public ServiceFramework<IAuthorizationRequest> {
+public:
+ explicit IAuthorizationRequest(Common::UUID user_id)
+ : ServiceFramework("IAuthorizationRequest") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "GetSessionId"},
+ {10, nullptr, "InvokeWithoutInteractionAsync"},
+ {19, nullptr, "IsAuthorized"},
+ {20, nullptr, "GetAuthorizationCode"},
+ {21, nullptr, "GetIdToken"},
+ {22, nullptr, "GetState"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class IOAuthProcedure final : public ServiceFramework<IOAuthProcedure> {
+public:
+ explicit IOAuthProcedure(Common::UUID user_id) : ServiceFramework("IOAuthProcedure") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "PrepareAsync"},
+ {1, nullptr, "GetRequest"},
+ {2, nullptr, "ApplyResponse"},
+ {3, nullptr, "ApplyResponseAsync"},
+ {10, nullptr, "Suspend"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+// 3.0.0+
+class IOAuthProcedureForExternalNsa final : public ServiceFramework<IOAuthProcedureForExternalNsa> {
+public:
+ explicit IOAuthProcedureForExternalNsa(Common::UUID user_id)
+ : ServiceFramework("IOAuthProcedureForExternalNsa") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "PrepareAsync"},
+ {1, nullptr, "GetRequest"},
+ {2, nullptr, "ApplyResponse"},
+ {3, nullptr, "ApplyResponseAsync"},
+ {10, nullptr, "Suspend"},
+ {100, nullptr, "GetAccountId"},
+ {101, nullptr, "GetLinkedNintendoAccountId"},
+ {102, nullptr, "GetNickname"},
+ {103, nullptr, "GetProfileImage"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class IOAuthProcedureForNintendoAccountLinkage final
+ : public ServiceFramework<IOAuthProcedureForNintendoAccountLinkage> {
+public:
+ explicit IOAuthProcedureForNintendoAccountLinkage(Common::UUID user_id)
+ : ServiceFramework("IOAuthProcedureForNintendoAccountLinkage") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "PrepareAsync"},
+ {1, nullptr, "GetRequest"},
+ {2, nullptr, "ApplyResponse"},
+ {3, nullptr, "ApplyResponseAsync"},
+ {10, nullptr, "Suspend"},
+ {100, nullptr, "GetRequestWithTheme"},
+ {101, nullptr, "IsNetworkServiceAccountReplaced"},
+ {199, nullptr, "GetUrlForIntroductionOfExtraMembership"}, // 2.0.0 - 5.1.0
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class INotifier final : public ServiceFramework<INotifier> {
+public:
+ explicit INotifier(Common::UUID user_id) : ServiceFramework("INotifier") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "GetSystemEvent"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
class IProfileCommon : public ServiceFramework<IProfileCommon> {
public:
explicit IProfileCommon(const char* name, bool editor_commands, Common::UUID user_id,
@@ -226,6 +438,54 @@ public:
: IProfileCommon("IProfileEditor", true, user_id, profile_manager) {}
};
+class IAsyncContext final : public ServiceFramework<IAsyncContext> {
+public:
+ explicit IAsyncContext(Common::UUID user_id) : ServiceFramework("IAsyncContext") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "GetSystemEvent"},
+ {1, nullptr, "Cancel"},
+ {2, nullptr, "HasDone"},
+ {3, nullptr, "GetResult"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class ISessionObject final : public ServiceFramework<ISessionObject> {
+public:
+ explicit ISessionObject(Common::UUID user_id) : ServiceFramework("ISessionObject") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {999, nullptr, "Dummy"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class IGuestLoginRequest final : public ServiceFramework<IGuestLoginRequest> {
+public:
+ explicit IGuestLoginRequest(Common::UUID) : ServiceFramework("IGuestLoginRequest") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "GetSessionId"},
+ {11, nullptr, "Unknown"}, // 1.0.0 - 2.3.0 (the name is blank on Switchbrew)
+ {12, nullptr, "GetAccountId"},
+ {13, nullptr, "GetLinkedNintendoAccountId"},
+ {14, nullptr, "GetNickname"},
+ {15, nullptr, "GetProfileImage"},
+ {21, nullptr, "LoadIdTokenCache"}, // 3.0.0+
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
class IManagerForApplication final : public ServiceFramework<IManagerForApplication> {
public:
explicit IManagerForApplication(Common::UUID user_id)
@@ -265,6 +525,87 @@ private:
Common::UUID user_id;
};
+// 6.0.0+
+class IAsyncNetworkServiceLicenseKindContext final
+ : public ServiceFramework<IAsyncNetworkServiceLicenseKindContext> {
+public:
+ explicit IAsyncNetworkServiceLicenseKindContext(Common::UUID user_id)
+ : ServiceFramework("IAsyncNetworkServiceLicenseKindContext") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "GetSystemEvent"},
+ {1, nullptr, "Cancel"},
+ {2, nullptr, "HasDone"},
+ {3, nullptr, "GetResult"},
+ {4, nullptr, "GetNetworkServiceLicenseKind"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+// 8.0.0+
+class IOAuthProcedureForUserRegistration final
+ : public ServiceFramework<IOAuthProcedureForUserRegistration> {
+public:
+ explicit IOAuthProcedureForUserRegistration(Common::UUID user_id)
+ : ServiceFramework("IOAuthProcedureForUserRegistration") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "PrepareAsync"},
+ {1, nullptr, "GetRequest"},
+ {2, nullptr, "ApplyResponse"},
+ {3, nullptr, "ApplyResponseAsync"},
+ {10, nullptr, "Suspend"},
+ {100, nullptr, "GetAccountId"},
+ {101, nullptr, "GetLinkedNintendoAccountId"},
+ {102, nullptr, "GetNickname"},
+ {103, nullptr, "GetProfileImage"},
+ {110, nullptr, "RegisterUserAsync"},
+ {111, nullptr, "GetUid"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+class DAUTH_O final : public ServiceFramework<DAUTH_O> {
+public:
+ explicit DAUTH_O(Common::UUID) : ServiceFramework("dauth:o") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "EnsureAuthenticationTokenCacheAsync"}, // [5.0.0-5.1.0] GeneratePostData
+ {1, nullptr, "LoadAuthenticationTokenCache"}, // 6.0.0+
+ {2, nullptr, "InvalidateAuthenticationTokenCache"}, // 6.0.0+
+ {10, nullptr, "EnsureEdgeTokenCacheAsync"}, // 6.0.0+
+ {11, nullptr, "LoadEdgeTokenCache"}, // 6.0.0+
+ {12, nullptr, "InvalidateEdgeTokenCache"}, // 6.0.0+
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
+// 6.0.0+
+class IAsyncResult final : public ServiceFramework<IAsyncResult> {
+public:
+ explicit IAsyncResult(Common::UUID user_id) : ServiceFramework("IAsyncResult") {
+ // clang-format off
+ static const FunctionInfo functions[] = {
+ {0, nullptr, "GetResult"},
+ {1, nullptr, "Cancel"},
+ {2, nullptr, "IsAvailable"},
+ {3, nullptr, "GetSystemEvent"},
+ };
+ // clang-format on
+
+ RegisterHandlers(functions);
+ }
+};
+
void Module::Interface::GetUserCount(Kernel::HLERequestContext& ctx) {
LOG_DEBUG(Service_ACC, "called");
IPC::ResponseBuilder rb{ctx, 3};
diff --git a/src/core/hle/service/acc/acc_aa.cpp b/src/core/hle/service/acc/acc_aa.cpp
index 3bac6bcd1..51f119b12 100644
--- a/src/core/hle/service/acc/acc_aa.cpp
+++ b/src/core/hle/service/acc/acc_aa.cpp
@@ -13,8 +13,8 @@ ACC_AA::ACC_AA(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
{0, nullptr, "EnsureCacheAsync"},
{1, nullptr, "LoadCache"},
{2, nullptr, "GetDeviceAccountId"},
- {50, nullptr, "RegisterNotificationTokenAsync"},
- {51, nullptr, "UnregisterNotificationTokenAsync"},
+ {50, nullptr, "RegisterNotificationTokenAsync"}, // 1.0.0 - 6.2.0
+ {51, nullptr, "UnregisterNotificationTokenAsync"}, // 1.0.0 - 6.2.0
};
RegisterHandlers(functions);
}
diff --git a/src/core/hle/service/acc/acc_su.cpp b/src/core/hle/service/acc/acc_su.cpp
index 2eefc6df5..85620bde3 100644
--- a/src/core/hle/service/acc/acc_su.cpp
+++ b/src/core/hle/service/acc/acc_su.cpp
@@ -17,28 +17,28 @@ ACC_SU::ACC_SU(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
{3, &ACC_SU::ListOpenUsers, "ListOpenUsers"},
{4, &ACC_SU::GetLastOpenedUser, "GetLastOpenedUser"},
{5, &ACC_SU::GetProfile, "GetProfile"},
- {6, nullptr, "GetProfileDigest"},
+ {6, nullptr, "GetProfileDigest"}, // 3.0.0+
{50, &ACC_SU::IsUserRegistrationRequestPermitted, "IsUserRegistrationRequestPermitted"},
{51, &ACC_SU::TrySelectUserWithoutInteraction, "TrySelectUserWithoutInteraction"},
- {60, nullptr, "ListOpenContextStoredUsers"},
- {99, nullptr, "DebugActivateOpenContextRetention"},
+ {60, nullptr, "ListOpenContextStoredUsers"}, // 5.0.0 - 5.1.0
+ {99, nullptr, "DebugActivateOpenContextRetention"}, // 6.0.0+
{100, nullptr, "GetUserRegistrationNotifier"},
{101, nullptr, "GetUserStateChangeNotifier"},
{102, nullptr, "GetBaasAccountManagerForSystemService"},
{103, nullptr, "GetBaasUserAvailabilityChangeNotifier"},
{104, nullptr, "GetProfileUpdateNotifier"},
- {105, nullptr, "CheckNetworkServiceAvailabilityAsync"},
- {106, nullptr, "GetProfileSyncNotifier"},
+ {105, nullptr, "CheckNetworkServiceAvailabilityAsync"}, // 4.0.0+
+ {106, nullptr, "GetProfileSyncNotifier"}, // 9.0.0+
{110, nullptr, "StoreSaveDataThumbnail"},
{111, nullptr, "ClearSaveDataThumbnail"},
{112, nullptr, "LoadSaveDataThumbnail"},
- {113, nullptr, "GetSaveDataThumbnailExistence"},
- {120, nullptr, "ListOpenUsersInApplication"},
- {130, nullptr, "ActivateOpenContextRetention"},
- {140, &ACC_SU::ListQualifiedUsers, "ListQualifiedUsers"},
- {150, nullptr, "AuthenticateApplicationAsync"},
- {190, nullptr, "GetUserLastOpenedApplication"},
- {191, nullptr, "ActivateOpenContextHolder"},
+ {113, nullptr, "GetSaveDataThumbnailExistence"}, // 5.0.0+
+ {120, nullptr, "ListOpenUsersInApplication"}, // 10.0.0+
+ {130, nullptr, "ActivateOpenContextRetention"}, // 6.0.0+
+ {140, &ACC_SU::ListQualifiedUsers, "ListQualifiedUsers"}, // 6.0.0+
+ {150, nullptr, "AuthenticateApplicationAsync"}, // 10.0.0+
+ {190, nullptr, "GetUserLastOpenedApplication"}, // 1.0.0 - 9.2.0
+ {191, nullptr, "ActivateOpenContextHolder"}, // 7.0.0+
{200, nullptr, "BeginUserRegistration"},
{201, nullptr, "CompleteUserRegistration"},
{202, nullptr, "CancelUserRegistration"},
@@ -46,15 +46,15 @@ ACC_SU::ACC_SU(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
{204, nullptr, "SetUserPosition"},
{205, &ACC_SU::GetProfileEditor, "GetProfileEditor"},
{206, nullptr, "CompleteUserRegistrationForcibly"},
- {210, nullptr, "CreateFloatingRegistrationRequest"},
- {211, nullptr, "CreateProcedureToRegisterUserWithNintendoAccount"},
- {212, nullptr, "ResumeProcedureToRegisterUserWithNintendoAccount"},
+ {210, nullptr, "CreateFloatingRegistrationRequest"}, // 3.0.0+
+ {211, nullptr, "CreateProcedureToRegisterUserWithNintendoAccount"}, // 8.0.0+
+ {212, nullptr, "ResumeProcedureToRegisterUserWithNintendoAccount"}, // 8.0.0+
{230, nullptr, "AuthenticateServiceAsync"},
{250, nullptr, "GetBaasAccountAdministrator"},
{290, nullptr, "ProxyProcedureForGuestLoginWithNintendoAccount"},
- {291, nullptr, "ProxyProcedureForFloatingRegistrationWithNintendoAccount"},
+ {291, nullptr, "ProxyProcedureForFloatingRegistrationWithNintendoAccount"}, // 3.0.0+
{299, nullptr, "SuspendBackgroundDaemon"},
- {997, nullptr, "DebugInvalidateTokenCacheForUser"},
+ {997, nullptr, "DebugInvalidateTokenCacheForUser"}, // 3.0.0+
{998, nullptr, "DebugSetUserStateClose"},
{999, nullptr, "DebugSetUserStateOpen"},
};
diff --git a/src/core/hle/service/acc/acc_u0.cpp b/src/core/hle/service/acc/acc_u0.cpp
index fb4e7e772..49f6e20f1 100644
--- a/src/core/hle/service/acc/acc_u0.cpp
+++ b/src/core/hle/service/acc/acc_u0.cpp
@@ -17,23 +17,23 @@ ACC_U0::ACC_U0(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
{3, &ACC_U0::ListOpenUsers, "ListOpenUsers"},
{4, &ACC_U0::GetLastOpenedUser, "GetLastOpenedUser"},
{5, &ACC_U0::GetProfile, "GetProfile"},
- {6, nullptr, "GetProfileDigest"},
+ {6, nullptr, "GetProfileDigest"}, // 3.0.0+
{50, &ACC_U0::IsUserRegistrationRequestPermitted, "IsUserRegistrationRequestPermitted"},
{51, &ACC_U0::TrySelectUserWithoutInteraction, "TrySelectUserWithoutInteraction"},
- {60, nullptr, "ListOpenContextStoredUsers"},
- {99, nullptr, "DebugActivateOpenContextRetention"},
+ {60, nullptr, "ListOpenContextStoredUsers"}, // 5.0.0 - 5.1.0
+ {99, nullptr, "DebugActivateOpenContextRetention"}, // 6.0.0+
{100, &ACC_U0::InitializeApplicationInfo, "InitializeApplicationInfo"},
{101, &ACC_U0::GetBaasAccountManagerForApplication, "GetBaasAccountManagerForApplication"},
{102, nullptr, "AuthenticateApplicationAsync"},
- {103, nullptr, "CheckNetworkServiceAvailabilityAsync"},
+ {103, nullptr, "CheckNetworkServiceAvailabilityAsync"}, // 4.0.0+
{110, nullptr, "StoreSaveDataThumbnail"},
{111, nullptr, "ClearSaveDataThumbnail"},
{120, nullptr, "CreateGuestLoginRequest"},
- {130, nullptr, "LoadOpenContext"},
- {131, nullptr, "ListOpenContextStoredUsers"},
- {140, &ACC_U0::InitializeApplicationInfoRestricted, "InitializeApplicationInfoRestricted"},
- {141, &ACC_U0::ListQualifiedUsers, "ListQualifiedUsers"},
- {150, &ACC_U0::IsUserAccountSwitchLocked, "IsUserAccountSwitchLocked"},
+ {130, nullptr, "LoadOpenContext"}, // 5.0.0+
+ {131, nullptr, "ListOpenContextStoredUsers"}, // 6.0.0+
+ {140, &ACC_U0::InitializeApplicationInfoRestricted, "InitializeApplicationInfoRestricted"}, // 6.0.0+
+ {141, &ACC_U0::ListQualifiedUsers, "ListQualifiedUsers"}, // 6.0.0+
+ {150, &ACC_U0::IsUserAccountSwitchLocked, "IsUserAccountSwitchLocked"}, // 6.0.0+
};
// clang-format on
diff --git a/src/core/hle/service/acc/acc_u1.cpp b/src/core/hle/service/acc/acc_u1.cpp
index 9f29cdc82..f47004f84 100644
--- a/src/core/hle/service/acc/acc_u1.cpp
+++ b/src/core/hle/service/acc/acc_u1.cpp
@@ -17,28 +17,29 @@ ACC_U1::ACC_U1(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p
{3, &ACC_U1::ListOpenUsers, "ListOpenUsers"},
{4, &ACC_U1::GetLastOpenedUser, "GetLastOpenedUser"},
{5, &ACC_U1::GetProfile, "GetProfile"},
- {6, nullptr, "GetProfileDigest"},
+ {6, nullptr, "GetProfileDigest"}, // 3.0.0+
{50, &ACC_U1::IsUserRegistrationRequestPermitted, "IsUserRegistrationRequestPermitted"},
{51, &ACC_U1::TrySelectUserWithoutInteraction, "TrySelectUserWithoutInteraction"},
- {60, nullptr, "ListOpenContextStoredUsers"},
- {99, nullptr, "DebugActivateOpenContextRetention"},
+ {60, nullptr, "ListOpenContextStoredUsers"}, // 5.0.0 - 5.1.0
+ {99, nullptr, "DebugActivateOpenContextRetention"}, // 6.0.0+
{100, nullptr, "GetUserRegistrationNotifier"},
{101, nullptr, "GetUserStateChangeNotifier"},
{102, nullptr, "GetBaasAccountManagerForSystemService"},
- {103, nullptr, "GetProfileUpdateNotifier"},
- {104, nullptr, "CheckNetworkServiceAvailabilityAsync"},
- {105, nullptr, "GetBaasUserAvailabilityChangeNotifier"},
- {106, nullptr, "GetProfileSyncNotifier"},
+ {103, nullptr, "GetBaasUserAvailabilityChangeNotifier"},
+ {104, nullptr, "GetProfileUpdateNotifier"},
+ {105, nullptr, "CheckNetworkServiceAvailabilityAsync"}, // 4.0.0+
+ {106, nullptr, "GetProfileSyncNotifier"}, // 9.0.0+
{110, nullptr, "StoreSaveDataThumbnail"},
{111, nullptr, "ClearSaveDataThumbnail"},
{112, nullptr, "LoadSaveDataThumbnail"},
- {113, nullptr, "GetSaveDataThumbnailExistence"},
- {130, nullptr, "ActivateOpenContextRetention"},
- {140, &ACC_U1::ListQualifiedUsers, "ListQualifiedUsers"},
- {150, nullptr, "AuthenticateApplicationAsync"},
- {190, nullptr, "GetUserLastOpenedApplication"},
- {191, nullptr, "ActivateOpenContextHolder"},
- {997, nullptr, "DebugInvalidateTokenCacheForUser"},
+ {113, nullptr, "GetSaveDataThumbnailExistence"}, // 5.0.0+
+ {120, nullptr, "ListOpenUsersInApplication"}, // 10.0.0+
+ {130, nullptr, "ActivateOpenContextRetention"}, // 6.0.0+
+ {140, &ACC_U1::ListQualifiedUsers, "ListQualifiedUsers"}, // 6.0.0+
+ {150, nullptr, "AuthenticateApplicationAsync"}, // 10.0.0+
+ {190, nullptr, "GetUserLastOpenedApplication"}, // 1.0.0 - 9.2.0
+ {191, nullptr, "ActivateOpenContextHolder"}, // 7.0.0+
+ {997, nullptr, "DebugInvalidateTokenCacheForUser"}, // 3.0.0+
{998, nullptr, "DebugSetUserStateClose"},
{999, nullptr, "DebugSetUserStateOpen"},
};
diff --git a/src/core/hle/service/am/applets/software_keyboard.cpp b/src/core/hle/service/am/applets/software_keyboard.cpp
index 54e63c138..d14076b02 100644
--- a/src/core/hle/service/am/applets/software_keyboard.cpp
+++ b/src/core/hle/service/am/applets/software_keyboard.cpp
@@ -30,7 +30,7 @@ static Core::Frontend::SoftwareKeyboardParameters ConvertToFrontendParameters(
config.sub_text.size());
params.guide_text = Common::UTF16StringFromFixedZeroTerminatedBuffer(config.guide_text.data(),
config.guide_text.size());
- params.initial_text = initial_text;
+ params.initial_text = std::move(initial_text);
params.max_length = config.length_limit == 0 ? DEFAULT_MAX_LENGTH : config.length_limit;
params.password = static_cast<bool>(config.is_password);
params.cursor_at_beginning = static_cast<bool>(config.initial_cursor_position);
@@ -109,7 +109,7 @@ void SoftwareKeyboard::Execute() {
const auto parameters = ConvertToFrontendParameters(config, initial_text);
- frontend.RequestText([this](std::optional<std::u16string> text) { WriteText(text); },
+ frontend.RequestText([this](std::optional<std::u16string> text) { WriteText(std::move(text)); },
parameters);
}
diff --git a/src/core/hle/service/hid/controllers/npad.cpp b/src/core/hle/service/hid/controllers/npad.cpp
index c55d900e2..6fbee7efa 100644
--- a/src/core/hle/service/hid/controllers/npad.cpp
+++ b/src/core/hle/service/hid/controllers/npad.cpp
@@ -566,6 +566,14 @@ void Controller_NPad::DisconnectNPad(u32 npad_id) {
connected_controllers[NPadIdToIndex(npad_id)].is_connected = false;
}
+void Controller_NPad::SetGyroscopeZeroDriftMode(GyroscopeZeroDriftMode drift_mode) {
+ gyroscope_zero_drift_mode = drift_mode;
+}
+
+Controller_NPad::GyroscopeZeroDriftMode Controller_NPad::GetGyroscopeZeroDriftMode() const {
+ return gyroscope_zero_drift_mode;
+}
+
void Controller_NPad::StartLRAssignmentMode() {
// Nothing internally is used for lr assignment mode. Since we have the ability to set the
// controller types from boot, it doesn't really matter about showing a selection screen
diff --git a/src/core/hle/service/hid/controllers/npad.h b/src/core/hle/service/hid/controllers/npad.h
index 931f03430..5d4c58a43 100644
--- a/src/core/hle/service/hid/controllers/npad.h
+++ b/src/core/hle/service/hid/controllers/npad.h
@@ -58,6 +58,12 @@ public:
};
static_assert(sizeof(Vibration) == 0x10, "Vibration is an invalid size");
+ enum class GyroscopeZeroDriftMode : u32 {
+ Loose = 0,
+ Standard = 1,
+ Tight = 2,
+ };
+
enum class NpadHoldType : u64 {
Vertical = 0,
Horizontal = 1,
@@ -117,6 +123,8 @@ public:
void ConnectNPad(u32 npad_id);
void DisconnectNPad(u32 npad_id);
+ void SetGyroscopeZeroDriftMode(GyroscopeZeroDriftMode drift_mode);
+ GyroscopeZeroDriftMode GetGyroscopeZeroDriftMode() const;
LedPattern GetLedPattern(u32 npad_id);
void SetVibrationEnabled(bool can_vibrate);
bool IsVibrationEnabled() const;
@@ -324,8 +332,8 @@ private:
std::array<Kernel::EventPair, 10> styleset_changed_events;
Vibration last_processed_vibration{};
std::array<ControllerHolder, 10> connected_controllers{};
+ GyroscopeZeroDriftMode gyroscope_zero_drift_mode{GyroscopeZeroDriftMode::Standard};
bool can_controllers_vibrate{true};
-
std::array<ControllerPad, 10> npad_pad_states{};
bool is_in_lr_assignment_mode{false};
Core::System& system;
diff --git a/src/core/hle/service/hid/hid.cpp b/src/core/hle/service/hid/hid.cpp
index 1ee99bfeb..57d5edea7 100644
--- a/src/core/hle/service/hid/hid.cpp
+++ b/src/core/hle/service/hid/hid.cpp
@@ -185,8 +185,8 @@ Hid::Hid(Core::System& system) : ServiceFramework("hid"), system(system) {
{77, nullptr, "GetAccelerometerPlayMode"},
{78, nullptr, "ResetAccelerometerPlayMode"},
{79, &Hid::SetGyroscopeZeroDriftMode, "SetGyroscopeZeroDriftMode"},
- {80, nullptr, "GetGyroscopeZeroDriftMode"},
- {81, nullptr, "ResetGyroscopeZeroDriftMode"},
+ {80, &Hid::GetGyroscopeZeroDriftMode, "GetGyroscopeZeroDriftMode"},
+ {81, &Hid::ResetGyroscopeZeroDriftMode, "ResetGyroscopeZeroDriftMode"},
{82, &Hid::IsSixAxisSensorAtRest, "IsSixAxisSensorAtRest"},
{83, nullptr, "IsFirmwareUpdateAvailableForSixAxisSensor"},
{91, &Hid::ActivateGesture, "ActivateGesture"},
@@ -440,9 +440,41 @@ void Hid::SetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx) {
const auto drift_mode{rp.Pop<u32>()};
const auto applet_resource_user_id{rp.Pop<u64>()};
- LOG_WARNING(Service_HID,
- "(STUBBED) called, handle={}, drift_mode={}, applet_resource_user_id={}", handle,
- drift_mode, applet_resource_user_id);
+ applet_resource->GetController<Controller_NPad>(HidController::NPad)
+ .SetGyroscopeZeroDriftMode(Controller_NPad::GyroscopeZeroDriftMode{drift_mode});
+
+ LOG_DEBUG(Service_HID, "called, handle={}, drift_mode={}, applet_resource_user_id={}", handle,
+ drift_mode, applet_resource_user_id);
+
+ IPC::ResponseBuilder rb{ctx, 2};
+ rb.Push(RESULT_SUCCESS);
+}
+
+void Hid::GetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx) {
+ IPC::RequestParser rp{ctx};
+ const auto handle{rp.Pop<u32>()};
+ const auto applet_resource_user_id{rp.Pop<u64>()};
+
+ LOG_DEBUG(Service_HID, "called, handle={}, applet_resource_user_id={}", handle,
+ applet_resource_user_id);
+
+ IPC::ResponseBuilder rb{ctx, 3};
+ rb.Push(RESULT_SUCCESS);
+ rb.Push<u32>(
+ static_cast<u32>(applet_resource->GetController<Controller_NPad>(HidController::NPad)
+ .GetGyroscopeZeroDriftMode()));
+}
+
+void Hid::ResetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx) {
+ IPC::RequestParser rp{ctx};
+ const auto handle{rp.Pop<u32>()};
+ const auto applet_resource_user_id{rp.Pop<u64>()};
+
+ applet_resource->GetController<Controller_NPad>(HidController::NPad)
+ .SetGyroscopeZeroDriftMode(Controller_NPad::GyroscopeZeroDriftMode::Standard);
+
+ LOG_DEBUG(Service_HID, "called, handle={}, applet_resource_user_id={}", handle,
+ applet_resource_user_id);
IPC::ResponseBuilder rb{ctx, 2};
rb.Push(RESULT_SUCCESS);
diff --git a/src/core/hle/service/hid/hid.h b/src/core/hle/service/hid/hid.h
index ffce5ba45..6fb048360 100644
--- a/src/core/hle/service/hid/hid.h
+++ b/src/core/hle/service/hid/hid.h
@@ -97,6 +97,8 @@ private:
void StartSixAxisSensor(Kernel::HLERequestContext& ctx);
void StopSixAxisSensor(Kernel::HLERequestContext& ctx);
void SetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx);
+ void GetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx);
+ void ResetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx);
void IsSixAxisSensorAtRest(Kernel::HLERequestContext& ctx);
void SetSupportedNpadStyleSet(Kernel::HLERequestContext& ctx);
void GetSupportedNpadStyleSet(Kernel::HLERequestContext& ctx);
diff --git a/src/core/hle/service/lm/manager.cpp b/src/core/hle/service/lm/manager.cpp
index b67081b86..3ee2374e7 100644
--- a/src/core/hle/service/lm/manager.cpp
+++ b/src/core/hle/service/lm/manager.cpp
@@ -86,7 +86,8 @@ std::string FormatField(Field type, const std::vector<u8>& data) {
return Common::StringFromFixedZeroTerminatedBuffer(
reinterpret_cast<const char*>(data.data()), data.size());
default:
- UNIMPLEMENTED();
+ UNIMPLEMENTED_MSG("Unimplemented field type={}", type);
+ return "";
}
}
diff --git a/src/video_core/CMakeLists.txt b/src/video_core/CMakeLists.txt
index 099bb446e..2dc752aa9 100644
--- a/src/video_core/CMakeLists.txt
+++ b/src/video_core/CMakeLists.txt
@@ -27,6 +27,8 @@ add_library(video_core STATIC
engines/shader_type.h
macro/macro.cpp
macro/macro.h
+ macro/macro_hle.cpp
+ macro/macro_hle.h
macro/macro_interpreter.cpp
macro/macro_interpreter.h
macro/macro_jit_x64.cpp
diff --git a/src/video_core/buffer_cache/buffer_cache.h b/src/video_core/buffer_cache/buffer_cache.h
index bae1d527c..cf8bdd021 100644
--- a/src/video_core/buffer_cache/buffer_cache.h
+++ b/src/video_core/buffer_cache/buffer_cache.h
@@ -41,7 +41,11 @@ class BufferCache {
static constexpr u64 BLOCK_PAGE_SIZE = 1ULL << BLOCK_PAGE_BITS;
public:
- using BufferInfo = std::pair<BufferType, u64>;
+ struct BufferInfo {
+ BufferType handle;
+ u64 offset;
+ u64 address;
+ };
BufferInfo UploadMemory(GPUVAddr gpu_addr, std::size_t size, std::size_t alignment = 4,
bool is_written = false, bool use_fast_cbuf = false) {
@@ -50,7 +54,7 @@ public:
auto& memory_manager = system.GPU().MemoryManager();
const std::optional<VAddr> cpu_addr_opt = memory_manager.GpuToCpuAddress(gpu_addr);
if (!cpu_addr_opt) {
- return {GetEmptyBuffer(size), 0};
+ return GetEmptyBuffer(size);
}
const VAddr cpu_addr = *cpu_addr_opt;
@@ -88,7 +92,7 @@ public:
Buffer* const block = GetBlock(cpu_addr, size);
MapInterval* const map = MapAddress(block, gpu_addr, cpu_addr, size);
if (!map) {
- return {GetEmptyBuffer(size), 0};
+ return GetEmptyBuffer(size);
}
if (is_written) {
map->MarkAsModified(true, GetModifiedTicks());
@@ -101,7 +105,7 @@ public:
}
}
- return {block->Handle(), static_cast<u64>(block->Offset(cpu_addr))};
+ return BufferInfo{block->Handle(), block->Offset(cpu_addr), block->Address()};
}
/// Uploads from a host memory. Returns the OpenGL buffer where it's located and its offset.
@@ -254,27 +258,17 @@ public:
committed_flushes.pop_front();
}
- virtual BufferType GetEmptyBuffer(std::size_t size) = 0;
+ virtual BufferInfo GetEmptyBuffer(std::size_t size) = 0;
protected:
explicit BufferCache(VideoCore::RasterizerInterface& rasterizer, Core::System& system,
- std::unique_ptr<StreamBuffer> stream_buffer_)
- : rasterizer{rasterizer}, system{system}, stream_buffer{std::move(stream_buffer_)},
- stream_buffer_handle{stream_buffer->Handle()} {}
+ std::unique_ptr<StreamBuffer> stream_buffer)
+ : rasterizer{rasterizer}, system{system}, stream_buffer{std::move(stream_buffer)} {}
~BufferCache() = default;
virtual std::shared_ptr<Buffer> CreateBlock(VAddr cpu_addr, std::size_t size) = 0;
- virtual void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- const u8* data) = 0;
-
- virtual void DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- u8* data) = 0;
-
- virtual void CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset,
- std::size_t dst_offset, std::size_t size) = 0;
-
virtual BufferInfo ConstBufferUpload(const void* raw_pointer, std::size_t size) {
return {};
}
@@ -336,11 +330,11 @@ private:
const VAddr cpu_addr_end = cpu_addr + size;
if (memory_manager.IsGranularRange(gpu_addr, size)) {
u8* host_ptr = memory_manager.GetPointer(gpu_addr);
- UploadBlockData(*block, block->Offset(cpu_addr), size, host_ptr);
+ block->Upload(block->Offset(cpu_addr), size, host_ptr);
} else {
staging_buffer.resize(size);
memory_manager.ReadBlockUnsafe(gpu_addr, staging_buffer.data(), size);
- UploadBlockData(*block, block->Offset(cpu_addr), size, staging_buffer.data());
+ block->Upload(block->Offset(cpu_addr), size, staging_buffer.data());
}
return Register(MapInterval(cpu_addr, cpu_addr_end, gpu_addr));
}
@@ -399,7 +393,7 @@ private:
}
staging_buffer.resize(size);
system.Memory().ReadBlockUnsafe(interval.lower(), staging_buffer.data(), size);
- UploadBlockData(*block, block->Offset(interval.lower()), size, staging_buffer.data());
+ block->Upload(block->Offset(interval.lower()), size, staging_buffer.data());
}
}
@@ -436,7 +430,7 @@ private:
const std::size_t size = map->end - map->start;
staging_buffer.resize(size);
- DownloadBlockData(*block, block->Offset(map->start), size, staging_buffer.data());
+ block->Download(block->Offset(map->start), size, staging_buffer.data());
system.Memory().WriteBlockUnsafe(map->start, staging_buffer.data(), size);
map->MarkAsModified(false, 0);
}
@@ -449,7 +443,7 @@ private:
buffer_ptr += size;
buffer_offset += size;
- return {stream_buffer_handle, uploaded_offset};
+ return BufferInfo{stream_buffer->Handle(), uploaded_offset, stream_buffer->Address()};
}
void AlignBuffer(std::size_t alignment) {
@@ -464,7 +458,7 @@ private:
const std::size_t new_size = old_size + BLOCK_PAGE_SIZE;
const VAddr cpu_addr = buffer->CpuAddr();
std::shared_ptr<Buffer> new_buffer = CreateBlock(cpu_addr, new_size);
- CopyBlock(*buffer, *new_buffer, 0, 0, old_size);
+ new_buffer->CopyFrom(*buffer, 0, 0, old_size);
QueueDestruction(std::move(buffer));
const VAddr cpu_addr_end = cpu_addr + new_size - 1;
@@ -486,8 +480,8 @@ private:
const std::size_t new_size = size_1 + size_2;
std::shared_ptr<Buffer> new_buffer = CreateBlock(new_addr, new_size);
- CopyBlock(*first, *new_buffer, 0, new_buffer->Offset(first_addr), size_1);
- CopyBlock(*second, *new_buffer, 0, new_buffer->Offset(second_addr), size_2);
+ new_buffer->CopyFrom(*first, 0, new_buffer->Offset(first_addr), size_1);
+ new_buffer->CopyFrom(*second, 0, new_buffer->Offset(second_addr), size_2);
QueueDestruction(std::move(first));
QueueDestruction(std::move(second));
diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp
index ea3c8a963..c01436295 100644
--- a/src/video_core/engines/maxwell_3d.cpp
+++ b/src/video_core/engines/maxwell_3d.cpp
@@ -128,7 +128,7 @@ void Maxwell3D::CallMacroMethod(u32 method, const std::vector<u32>& parameters)
((method - MacroRegistersStart) >> 1) % static_cast<u32>(macro_positions.size());
// Execute the current macro.
- macro_engine->Execute(macro_positions[entry], parameters);
+ macro_engine->Execute(*this, macro_positions[entry], parameters);
if (mme_draw.current_mode != MMEDrawMode::Undefined) {
FlushMMEInlineDraw();
}
diff --git a/src/video_core/engines/maxwell_3d.h b/src/video_core/engines/maxwell_3d.h
index d5fe25065..ef1618990 100644
--- a/src/video_core/engines/maxwell_3d.h
+++ b/src/video_core/engines/maxwell_3d.h
@@ -1418,6 +1418,14 @@ public:
return execute_on;
}
+ VideoCore::RasterizerInterface& GetRasterizer() {
+ return rasterizer;
+ }
+
+ const VideoCore::RasterizerInterface& GetRasterizer() const {
+ return rasterizer;
+ }
+
/// Notify a memory write has happened.
void OnMemoryWrite() {
dirty.flags |= dirty.on_write_stores;
diff --git a/src/video_core/macro/macro.cpp b/src/video_core/macro/macro.cpp
index 89077a2d8..ef7dad349 100644
--- a/src/video_core/macro/macro.cpp
+++ b/src/video_core/macro/macro.cpp
@@ -2,23 +2,37 @@
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
+#include <boost/container_hash/hash.hpp>
#include "common/assert.h"
#include "common/logging/log.h"
#include "core/settings.h"
+#include "video_core/engines/maxwell_3d.h"
#include "video_core/macro/macro.h"
+#include "video_core/macro/macro_hle.h"
#include "video_core/macro/macro_interpreter.h"
#include "video_core/macro/macro_jit_x64.h"
namespace Tegra {
+MacroEngine::MacroEngine(Engines::Maxwell3D& maxwell3d)
+ : hle_macros{std::make_unique<Tegra::HLEMacro>(maxwell3d)} {}
+
+MacroEngine::~MacroEngine() = default;
+
void MacroEngine::AddCode(u32 method, u32 data) {
uploaded_macro_code[method].push_back(data);
}
-void MacroEngine::Execute(u32 method, const std::vector<u32>& parameters) {
+void MacroEngine::Execute(Engines::Maxwell3D& maxwell3d, u32 method,
+ const std::vector<u32>& parameters) {
auto compiled_macro = macro_cache.find(method);
if (compiled_macro != macro_cache.end()) {
- compiled_macro->second->Execute(parameters, method);
+ const auto& cache_info = compiled_macro->second;
+ if (cache_info.has_hle_program) {
+ cache_info.hle_program->Execute(parameters, method);
+ } else {
+ cache_info.lle_program->Execute(parameters, method);
+ }
} else {
// Macro not compiled, check if it's uploaded and if so, compile it
auto macro_code = uploaded_macro_code.find(method);
@@ -26,8 +40,21 @@ void MacroEngine::Execute(u32 method, const std::vector<u32>& parameters) {
UNREACHABLE_MSG("Macro 0x{0:x} was not uploaded", method);
return;
}
- macro_cache[method] = Compile(macro_code->second);
- macro_cache[method]->Execute(parameters, method);
+ auto& cache_info = macro_cache[method];
+ cache_info.hash = boost::hash_value(macro_code->second);
+ cache_info.lle_program = Compile(macro_code->second);
+
+ auto hle_program = hle_macros->GetHLEProgram(cache_info.hash);
+ if (hle_program.has_value()) {
+ cache_info.has_hle_program = true;
+ cache_info.hle_program = std::move(hle_program.value());
+ }
+
+ if (cache_info.has_hle_program) {
+ cache_info.hle_program->Execute(parameters, method);
+ } else {
+ cache_info.lle_program->Execute(parameters, method);
+ }
}
}
diff --git a/src/video_core/macro/macro.h b/src/video_core/macro/macro.h
index b76ed891f..4d00b84b0 100644
--- a/src/video_core/macro/macro.h
+++ b/src/video_core/macro/macro.h
@@ -11,9 +11,11 @@
#include "common/common_types.h"
namespace Tegra {
+
namespace Engines {
class Maxwell3D;
}
+
namespace Macro {
constexpr std::size_t NUM_MACRO_REGISTERS = 8;
enum class Operation : u32 {
@@ -94,6 +96,8 @@ union MethodAddress {
} // namespace Macro
+class HLEMacro;
+
class CachedMacro {
public:
virtual ~CachedMacro() = default;
@@ -107,20 +111,29 @@ public:
class MacroEngine {
public:
- virtual ~MacroEngine() = default;
+ explicit MacroEngine(Engines::Maxwell3D& maxwell3d);
+ virtual ~MacroEngine();
// Store the uploaded macro code to compile them when they're called.
void AddCode(u32 method, u32 data);
// Compiles the macro if its not in the cache, and executes the compiled macro
- void Execute(u32 method, const std::vector<u32>& parameters);
+ void Execute(Engines::Maxwell3D& maxwell3d, u32 method, const std::vector<u32>& parameters);
protected:
virtual std::unique_ptr<CachedMacro> Compile(const std::vector<u32>& code) = 0;
private:
- std::unordered_map<u32, std::unique_ptr<CachedMacro>> macro_cache;
+ struct CacheInfo {
+ std::unique_ptr<CachedMacro> lle_program{};
+ std::unique_ptr<CachedMacro> hle_program{};
+ u64 hash{};
+ bool has_hle_program{};
+ };
+
+ std::unordered_map<u32, CacheInfo> macro_cache;
std::unordered_map<u32, std::vector<u32>> uploaded_macro_code;
+ std::unique_ptr<HLEMacro> hle_macros;
};
std::unique_ptr<MacroEngine> GetMacroEngine(Engines::Maxwell3D& maxwell3d);
diff --git a/src/video_core/macro/macro_hle.cpp b/src/video_core/macro/macro_hle.cpp
new file mode 100644
index 000000000..410f99018
--- /dev/null
+++ b/src/video_core/macro/macro_hle.cpp
@@ -0,0 +1,113 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <array>
+#include <vector>
+#include "video_core/engines/maxwell_3d.h"
+#include "video_core/macro/macro_hle.h"
+#include "video_core/rasterizer_interface.h"
+
+namespace Tegra {
+
+namespace {
+// HLE'd functions
+static void HLE_771BB18C62444DA0(Engines::Maxwell3D& maxwell3d,
+ const std::vector<u32>& parameters) {
+ const u32 instance_count = parameters[2] & maxwell3d.GetRegisterValue(0xD1B);
+
+ maxwell3d.regs.draw.topology.Assign(
+ static_cast<Tegra::Engines::Maxwell3D::Regs::PrimitiveTopology>(parameters[0] &
+ ~(0x3ffffff << 26)));
+ maxwell3d.regs.vb_base_instance = parameters[5];
+ maxwell3d.mme_draw.instance_count = instance_count;
+ maxwell3d.regs.vb_element_base = parameters[3];
+ maxwell3d.regs.index_array.count = parameters[1];
+ maxwell3d.regs.index_array.first = parameters[4];
+
+ if (maxwell3d.ShouldExecute()) {
+ maxwell3d.GetRasterizer().Draw(true, true);
+ }
+ maxwell3d.regs.index_array.count = 0;
+ maxwell3d.mme_draw.instance_count = 0;
+ maxwell3d.mme_draw.current_mode = Engines::Maxwell3D::MMEDrawMode::Undefined;
+}
+
+static void HLE_0D61FC9FAAC9FCAD(Engines::Maxwell3D& maxwell3d,
+ const std::vector<u32>& parameters) {
+ const u32 count = (maxwell3d.GetRegisterValue(0xD1B) & parameters[2]);
+
+ maxwell3d.regs.vertex_buffer.first = parameters[3];
+ maxwell3d.regs.vertex_buffer.count = parameters[1];
+ maxwell3d.regs.vb_base_instance = parameters[4];
+ maxwell3d.regs.draw.topology.Assign(
+ static_cast<Tegra::Engines::Maxwell3D::Regs::PrimitiveTopology>(parameters[0]));
+ maxwell3d.mme_draw.instance_count = count;
+
+ if (maxwell3d.ShouldExecute()) {
+ maxwell3d.GetRasterizer().Draw(false, true);
+ }
+ maxwell3d.regs.vertex_buffer.count = 0;
+ maxwell3d.mme_draw.instance_count = 0;
+ maxwell3d.mme_draw.current_mode = Engines::Maxwell3D::MMEDrawMode::Undefined;
+}
+
+static void HLE_0217920100488FF7(Engines::Maxwell3D& maxwell3d,
+ const std::vector<u32>& parameters) {
+ const u32 instance_count = (maxwell3d.GetRegisterValue(0xD1B) & parameters[2]);
+ const u32 element_base = parameters[4];
+ const u32 base_instance = parameters[5];
+ maxwell3d.regs.index_array.first = parameters[3];
+ maxwell3d.regs.reg_array[0x446] = element_base; // vertex id base?
+ maxwell3d.regs.index_array.count = parameters[1];
+ maxwell3d.regs.vb_element_base = element_base;
+ maxwell3d.regs.vb_base_instance = base_instance;
+ maxwell3d.mme_draw.instance_count = instance_count;
+ maxwell3d.CallMethodFromMME(0x8e3, 0x640);
+ maxwell3d.CallMethodFromMME(0x8e4, element_base);
+ maxwell3d.CallMethodFromMME(0x8e5, base_instance);
+ maxwell3d.regs.draw.topology.Assign(
+ static_cast<Tegra::Engines::Maxwell3D::Regs::PrimitiveTopology>(parameters[0]));
+ if (maxwell3d.ShouldExecute()) {
+ maxwell3d.GetRasterizer().Draw(true, true);
+ }
+ maxwell3d.regs.reg_array[0x446] = 0x0; // vertex id base?
+ maxwell3d.regs.index_array.count = 0;
+ maxwell3d.regs.vb_element_base = 0x0;
+ maxwell3d.regs.vb_base_instance = 0x0;
+ maxwell3d.mme_draw.instance_count = 0;
+ maxwell3d.CallMethodFromMME(0x8e3, 0x640);
+ maxwell3d.CallMethodFromMME(0x8e4, 0x0);
+ maxwell3d.CallMethodFromMME(0x8e5, 0x0);
+ maxwell3d.mme_draw.current_mode = Engines::Maxwell3D::MMEDrawMode::Undefined;
+}
+} // namespace
+
+constexpr std::array<std::pair<u64, HLEFunction>, 3> hle_funcs{{
+ std::make_pair<u64, HLEFunction>(0x771BB18C62444DA0, &HLE_771BB18C62444DA0),
+ std::make_pair<u64, HLEFunction>(0x0D61FC9FAAC9FCAD, &HLE_0D61FC9FAAC9FCAD),
+ std::make_pair<u64, HLEFunction>(0x0217920100488FF7, &HLE_0217920100488FF7),
+}};
+
+HLEMacro::HLEMacro(Engines::Maxwell3D& maxwell3d) : maxwell3d(maxwell3d) {}
+HLEMacro::~HLEMacro() = default;
+
+std::optional<std::unique_ptr<CachedMacro>> HLEMacro::GetHLEProgram(u64 hash) const {
+ const auto it = std::find_if(hle_funcs.cbegin(), hle_funcs.cend(),
+ [hash](const auto& pair) { return pair.first == hash; });
+ if (it == hle_funcs.end()) {
+ return std::nullopt;
+ }
+ return std::make_unique<HLEMacroImpl>(maxwell3d, it->second);
+}
+
+HLEMacroImpl::~HLEMacroImpl() = default;
+
+HLEMacroImpl::HLEMacroImpl(Engines::Maxwell3D& maxwell3d, HLEFunction func)
+ : maxwell3d(maxwell3d), func(func) {}
+
+void HLEMacroImpl::Execute(const std::vector<u32>& parameters, u32 method) {
+ func(maxwell3d, parameters);
+}
+
+} // namespace Tegra
diff --git a/src/video_core/macro/macro_hle.h b/src/video_core/macro/macro_hle.h
new file mode 100644
index 000000000..37af875a0
--- /dev/null
+++ b/src/video_core/macro/macro_hle.h
@@ -0,0 +1,44 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <memory>
+#include <optional>
+#include <vector>
+#include "common/common_types.h"
+#include "video_core/macro/macro.h"
+
+namespace Tegra {
+
+namespace Engines {
+class Maxwell3D;
+}
+
+using HLEFunction = void (*)(Engines::Maxwell3D& maxwell3d, const std::vector<u32>& parameters);
+
+class HLEMacro {
+public:
+ explicit HLEMacro(Engines::Maxwell3D& maxwell3d);
+ ~HLEMacro();
+
+ std::optional<std::unique_ptr<CachedMacro>> GetHLEProgram(u64 hash) const;
+
+private:
+ Engines::Maxwell3D& maxwell3d;
+};
+
+class HLEMacroImpl : public CachedMacro {
+public:
+ explicit HLEMacroImpl(Engines::Maxwell3D& maxwell3d, HLEFunction func);
+ ~HLEMacroImpl();
+
+ void Execute(const std::vector<u32>& parameters, u32 method) override;
+
+private:
+ Engines::Maxwell3D& maxwell3d;
+ HLEFunction func;
+};
+
+} // namespace Tegra
diff --git a/src/video_core/macro/macro_interpreter.cpp b/src/video_core/macro/macro_interpreter.cpp
index 5edff27aa..aa5256419 100644
--- a/src/video_core/macro/macro_interpreter.cpp
+++ b/src/video_core/macro/macro_interpreter.cpp
@@ -11,7 +11,8 @@
MICROPROFILE_DEFINE(MacroInterp, "GPU", "Execute macro interpreter", MP_RGB(128, 128, 192));
namespace Tegra {
-MacroInterpreter::MacroInterpreter(Engines::Maxwell3D& maxwell3d) : maxwell3d(maxwell3d) {}
+MacroInterpreter::MacroInterpreter(Engines::Maxwell3D& maxwell3d)
+ : MacroEngine::MacroEngine(maxwell3d), maxwell3d(maxwell3d) {}
std::unique_ptr<CachedMacro> MacroInterpreter::Compile(const std::vector<u32>& code) {
return std::make_unique<MacroInterpreterImpl>(maxwell3d, code);
diff --git a/src/video_core/macro/macro_jit_x64.cpp b/src/video_core/macro/macro_jit_x64.cpp
index 30abb66e5..07292702f 100644
--- a/src/video_core/macro/macro_jit_x64.cpp
+++ b/src/video_core/macro/macro_jit_x64.cpp
@@ -28,7 +28,8 @@ static const std::bitset<32> PERSISTENT_REGISTERS = Common::X64::BuildRegSet({
BRANCH_HOLDER,
});
-MacroJITx64::MacroJITx64(Engines::Maxwell3D& maxwell3d) : maxwell3d(maxwell3d) {}
+MacroJITx64::MacroJITx64(Engines::Maxwell3D& maxwell3d)
+ : MacroEngine::MacroEngine(maxwell3d), maxwell3d(maxwell3d) {}
std::unique_ptr<CachedMacro> MacroJITx64::Compile(const std::vector<u32>& code) {
return std::make_unique<MacroJITx64Impl>(maxwell3d, code);
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
index ad0577a4f..d9f7b4cc6 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
@@ -22,21 +22,46 @@ using Maxwell = Tegra::Engines::Maxwell3D::Regs;
MICROPROFILE_DEFINE(OpenGL_Buffer_Download, "OpenGL", "Buffer Download", MP_RGB(192, 192, 128));
-Buffer::Buffer(VAddr cpu_addr, const std::size_t size) : VideoCommon::BufferBlock{cpu_addr, size} {
+Buffer::Buffer(const Device& device, VAddr cpu_addr, std::size_t size)
+ : VideoCommon::BufferBlock{cpu_addr, size} {
gl_buffer.Create();
glNamedBufferData(gl_buffer.handle, static_cast<GLsizeiptr>(size), nullptr, GL_DYNAMIC_DRAW);
+ if (device.HasVertexBufferUnifiedMemory()) {
+ glMakeNamedBufferResidentNV(gl_buffer.handle, GL_READ_WRITE);
+ glGetNamedBufferParameterui64vNV(gl_buffer.handle, GL_BUFFER_GPU_ADDRESS_NV, &gpu_address);
+ }
}
Buffer::~Buffer() = default;
+void Buffer::Upload(std::size_t offset, std::size_t size, const u8* data) const {
+ glNamedBufferSubData(Handle(), static_cast<GLintptr>(offset), static_cast<GLsizeiptr>(size),
+ data);
+}
+
+void Buffer::Download(std::size_t offset, std::size_t size, u8* data) const {
+ MICROPROFILE_SCOPE(OpenGL_Buffer_Download);
+ glMemoryBarrier(GL_BUFFER_UPDATE_BARRIER_BIT);
+ glGetNamedBufferSubData(Handle(), static_cast<GLintptr>(offset), static_cast<GLsizeiptr>(size),
+ data);
+}
+
+void Buffer::CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset,
+ std::size_t size) const {
+ glCopyNamedBufferSubData(src.Handle(), Handle(), static_cast<GLintptr>(src_offset),
+ static_cast<GLintptr>(dst_offset), static_cast<GLsizeiptr>(size));
+}
+
OGLBufferCache::OGLBufferCache(RasterizerOpenGL& rasterizer, Core::System& system,
- const Device& device, std::size_t stream_size)
- : GenericBufferCache{rasterizer, system, std::make_unique<OGLStreamBuffer>(stream_size, true)} {
+ const Device& device_, std::size_t stream_size)
+ : GenericBufferCache{rasterizer, system,
+ std::make_unique<OGLStreamBuffer>(device_, stream_size, true)},
+ device{device_} {
if (!device.HasFastBufferSubData()) {
return;
}
- static constexpr auto size = static_cast<GLsizeiptr>(Maxwell::MaxConstBufferSize);
+ static constexpr GLsizeiptr size = static_cast<GLsizeiptr>(Maxwell::MaxConstBufferSize);
glCreateBuffers(static_cast<GLsizei>(std::size(cbufs)), std::data(cbufs));
for (const GLuint cbuf : cbufs) {
glNamedBufferData(cbuf, size, nullptr, GL_STREAM_DRAW);
@@ -48,39 +73,20 @@ OGLBufferCache::~OGLBufferCache() {
}
std::shared_ptr<Buffer> OGLBufferCache::CreateBlock(VAddr cpu_addr, std::size_t size) {
- return std::make_shared<Buffer>(cpu_addr, size);
+ return std::make_shared<Buffer>(device, cpu_addr, size);
}
-GLuint OGLBufferCache::GetEmptyBuffer(std::size_t) {
- return 0;
-}
-
-void OGLBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- const u8* data) {
- glNamedBufferSubData(buffer.Handle(), static_cast<GLintptr>(offset),
- static_cast<GLsizeiptr>(size), data);
-}
-
-void OGLBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- u8* data) {
- MICROPROFILE_SCOPE(OpenGL_Buffer_Download);
- glMemoryBarrier(GL_BUFFER_UPDATE_BARRIER_BIT);
- glGetNamedBufferSubData(buffer.Handle(), static_cast<GLintptr>(offset),
- static_cast<GLsizeiptr>(size), data);
-}
-
-void OGLBufferCache::CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset,
- std::size_t dst_offset, std::size_t size) {
- glCopyNamedBufferSubData(src.Handle(), dst.Handle(), static_cast<GLintptr>(src_offset),
- static_cast<GLintptr>(dst_offset), static_cast<GLsizeiptr>(size));
+OGLBufferCache::BufferInfo OGLBufferCache::GetEmptyBuffer(std::size_t) {
+ return {0, 0, 0};
}
OGLBufferCache::BufferInfo OGLBufferCache::ConstBufferUpload(const void* raw_pointer,
std::size_t size) {
DEBUG_ASSERT(cbuf_cursor < std::size(cbufs));
const GLuint cbuf = cbufs[cbuf_cursor++];
+
glNamedBufferSubData(cbuf, 0, static_cast<GLsizeiptr>(size), raw_pointer);
- return {cbuf, 0};
+ return {cbuf, 0, 0};
}
} // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.h b/src/video_core/renderer_opengl/gl_buffer_cache.h
index a49aaf9c4..59d95adbc 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.h
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.h
@@ -25,15 +25,27 @@ class RasterizerOpenGL;
class Buffer : public VideoCommon::BufferBlock {
public:
- explicit Buffer(VAddr cpu_addr, const std::size_t size);
+ explicit Buffer(const Device& device, VAddr cpu_addr, std::size_t size);
~Buffer();
- GLuint Handle() const {
+ void Upload(std::size_t offset, std::size_t size, const u8* data) const;
+
+ void Download(std::size_t offset, std::size_t size, u8* data) const;
+
+ void CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset,
+ std::size_t size) const;
+
+ GLuint Handle() const noexcept {
return gl_buffer.handle;
}
+ u64 Address() const noexcept {
+ return gpu_address;
+ }
+
private:
OGLBuffer gl_buffer;
+ u64 gpu_address = 0;
};
using GenericBufferCache = VideoCommon::BufferCache<Buffer, GLuint, OGLStreamBuffer>;
@@ -43,7 +55,7 @@ public:
const Device& device, std::size_t stream_size);
~OGLBufferCache();
- GLuint GetEmptyBuffer(std::size_t) override;
+ BufferInfo GetEmptyBuffer(std::size_t) override;
void Acquire() noexcept {
cbuf_cursor = 0;
@@ -52,22 +64,16 @@ public:
protected:
std::shared_ptr<Buffer> CreateBlock(VAddr cpu_addr, std::size_t size) override;
- void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- const u8* data) override;
-
- void DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- u8* data) override;
-
- void CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset,
- std::size_t dst_offset, std::size_t size) override;
-
BufferInfo ConstBufferUpload(const void* raw_pointer, std::size_t size) override;
private:
+ static constexpr std::size_t NUM_CBUFS = Tegra::Engines::Maxwell3D::Regs::MaxConstBuffers *
+ Tegra::Engines::Maxwell3D::Regs::MaxShaderProgram;
+
+ const Device& device;
+
std::size_t cbuf_cursor = 0;
- std::array<GLuint, Tegra::Engines::Maxwell3D::Regs::MaxConstBuffers *
- Tegra::Engines::Maxwell3D::Regs::MaxShaderProgram>
- cbufs;
+ std::array<GLuint, NUM_CBUFS> cbufs{};
};
} // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_device.cpp b/src/video_core/renderer_opengl/gl_device.cpp
index b31d604e4..447a19595 100644
--- a/src/video_core/renderer_opengl/gl_device.cpp
+++ b/src/video_core/renderer_opengl/gl_device.cpp
@@ -188,16 +188,32 @@ bool IsASTCSupported() {
return true;
}
+/// @brief Returns true when a GL_RENDERER is a Turing GPU
+/// @param renderer GL_RENDERER string
+bool IsTuring(std::string_view renderer) {
+ static constexpr std::array<std::string_view, 12> TURING_GPUS = {
+ "GTX 1650", "GTX 1660", "RTX 2060", "RTX 2070",
+ "RTX 2080", "TITAN RTX", "Quadro RTX 3000", "Quadro RTX 4000",
+ "Quadro RTX 5000", "Quadro RTX 6000", "Quadro RTX 8000", "Tesla T4",
+ };
+ return std::any_of(TURING_GPUS.begin(), TURING_GPUS.end(),
+ [renderer](std::string_view candidate) {
+ return renderer.find(candidate) != std::string_view::npos;
+ });
+}
+
} // Anonymous namespace
Device::Device()
: max_uniform_buffers{BuildMaxUniformBuffers()}, base_bindings{BuildBaseBindings()} {
const std::string_view vendor = reinterpret_cast<const char*>(glGetString(GL_VENDOR));
+ const std::string_view renderer = reinterpret_cast<const char*>(glGetString(GL_RENDERER));
const std::string_view version = reinterpret_cast<const char*>(glGetString(GL_VERSION));
const std::vector extensions = GetExtensions();
const bool is_nvidia = vendor == "NVIDIA Corporation";
const bool is_amd = vendor == "ATI Technologies Inc.";
+ const bool is_turing = is_nvidia && IsTuring(renderer);
bool disable_fast_buffer_sub_data = false;
if (is_nvidia && version == "4.6.0 NVIDIA 443.24") {
@@ -216,12 +232,21 @@ Device::Device()
has_shader_ballot = GLAD_GL_ARB_shader_ballot;
has_vertex_viewport_layer = GLAD_GL_ARB_shader_viewport_layer_array;
has_image_load_formatted = HasExtension(extensions, "GL_EXT_shader_image_load_formatted");
+ has_texture_shadow_lod = HasExtension(extensions, "GL_EXT_texture_shadow_lod");
has_astc = IsASTCSupported();
has_variable_aoffi = TestVariableAoffi();
has_component_indexing_bug = is_amd;
has_precise_bug = TestPreciseBug();
- has_fast_buffer_sub_data = is_nvidia && !disable_fast_buffer_sub_data;
has_nv_viewport_array2 = GLAD_GL_NV_viewport_array2;
+
+ // At the moment of writing this, only Nvidia's driver optimizes BufferSubData on exclusive
+ // uniform buffers as "push constants"
+ has_fast_buffer_sub_data = is_nvidia && !disable_fast_buffer_sub_data;
+
+ // Nvidia's driver on Turing GPUs randomly crashes when the buffer is made resident, or on
+ // DeleteBuffers. Disable unified memory on these devices.
+ has_vertex_buffer_unified_memory = GLAD_GL_NV_vertex_buffer_unified_memory && !is_turing;
+
use_assembly_shaders = Settings::values.use_assembly_shaders && GLAD_GL_NV_gpu_program5 &&
GLAD_GL_NV_compute_program5 && GLAD_GL_NV_transform_feedback &&
GLAD_GL_NV_transform_feedback2;
@@ -245,6 +270,7 @@ Device::Device(std::nullptr_t) {
has_shader_ballot = true;
has_vertex_viewport_layer = true;
has_image_load_formatted = true;
+ has_texture_shadow_lod = true;
has_variable_aoffi = true;
}
diff --git a/src/video_core/renderer_opengl/gl_device.h b/src/video_core/renderer_opengl/gl_device.h
index 145347943..e1d811966 100644
--- a/src/video_core/renderer_opengl/gl_device.h
+++ b/src/video_core/renderer_opengl/gl_device.h
@@ -68,6 +68,14 @@ public:
return has_image_load_formatted;
}
+ bool HasTextureShadowLod() const {
+ return has_texture_shadow_lod;
+ }
+
+ bool HasVertexBufferUnifiedMemory() const {
+ return has_vertex_buffer_unified_memory;
+ }
+
bool HasASTC() const {
return has_astc;
}
@@ -110,6 +118,8 @@ private:
bool has_shader_ballot{};
bool has_vertex_viewport_layer{};
bool has_image_load_formatted{};
+ bool has_texture_shadow_lod{};
+ bool has_vertex_buffer_unified_memory{};
bool has_astc{};
bool has_variable_aoffi{};
bool has_component_indexing_bug{};
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp
index 2d6c11320..362457ffe 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp
@@ -61,7 +61,8 @@ constexpr std::size_t NUM_CONST_BUFFERS_BYTES_PER_STAGE =
constexpr std::size_t TOTAL_CONST_BUFFER_BYTES =
NUM_CONST_BUFFERS_BYTES_PER_STAGE * Maxwell::MaxShaderStage;
-constexpr std::size_t NumSupportedVertexAttributes = 16;
+constexpr std::size_t NUM_SUPPORTED_VERTEX_ATTRIBUTES = 16;
+constexpr std::size_t NUM_SUPPORTED_VERTEX_BINDINGS = 16;
template <typename Engine, typename Entry>
Tegra::Texture::FullTextureInfo GetTextureInfo(const Engine& engine, const Entry& entry,
@@ -193,7 +194,7 @@ void RasterizerOpenGL::SetupVertexFormat() {
// avoid OpenGL errors.
// TODO(Subv): Analyze the shader to identify which attributes are actually used and don't
// assume every shader uses them all.
- for (std::size_t index = 0; index < NumSupportedVertexAttributes; ++index) {
+ for (std::size_t index = 0; index < NUM_SUPPORTED_VERTEX_ATTRIBUTES; ++index) {
if (!flags[Dirty::VertexFormat0 + index]) {
continue;
}
@@ -231,9 +232,11 @@ void RasterizerOpenGL::SetupVertexBuffer() {
MICROPROFILE_SCOPE(OpenGL_VB);
+ const bool use_unified_memory = device.HasVertexBufferUnifiedMemory();
+
// Upload all guest vertex arrays sequentially to our buffer
const auto& regs = gpu.regs;
- for (std::size_t index = 0; index < Maxwell::NumVertexArrays; ++index) {
+ for (std::size_t index = 0; index < NUM_SUPPORTED_VERTEX_BINDINGS; ++index) {
if (!flags[Dirty::VertexBuffer0 + index]) {
continue;
}
@@ -246,16 +249,25 @@ void RasterizerOpenGL::SetupVertexBuffer() {
const GPUVAddr start = vertex_array.StartAddress();
const GPUVAddr end = regs.vertex_array_limit[index].LimitAddress();
-
ASSERT(end >= start);
+
+ const GLuint gl_index = static_cast<GLuint>(index);
const u64 size = end - start;
if (size == 0) {
- glBindVertexBuffer(static_cast<GLuint>(index), 0, 0, vertex_array.stride);
+ glBindVertexBuffer(gl_index, 0, 0, vertex_array.stride);
+ if (use_unified_memory) {
+ glBufferAddressRangeNV(GL_VERTEX_ATTRIB_ARRAY_ADDRESS_NV, gl_index, 0, 0);
+ }
continue;
}
- const auto [vertex_buffer, vertex_buffer_offset] = buffer_cache.UploadMemory(start, size);
- glBindVertexBuffer(static_cast<GLuint>(index), vertex_buffer, vertex_buffer_offset,
- vertex_array.stride);
+ const auto info = buffer_cache.UploadMemory(start, size);
+ if (use_unified_memory) {
+ glBindVertexBuffer(gl_index, 0, 0, vertex_array.stride);
+ glBufferAddressRangeNV(GL_VERTEX_ATTRIB_ARRAY_ADDRESS_NV, gl_index,
+ info.address + info.offset, size);
+ } else {
+ glBindVertexBuffer(gl_index, info.handle, info.offset, vertex_array.stride);
+ }
}
}
@@ -268,7 +280,7 @@ void RasterizerOpenGL::SetupVertexInstances() {
flags[Dirty::VertexInstances] = false;
const auto& regs = gpu.regs;
- for (std::size_t index = 0; index < NumSupportedVertexAttributes; ++index) {
+ for (std::size_t index = 0; index < NUM_SUPPORTED_VERTEX_ATTRIBUTES; ++index) {
if (!flags[Dirty::VertexInstance0 + index]) {
continue;
}
@@ -285,9 +297,9 @@ GLintptr RasterizerOpenGL::SetupIndexBuffer() {
MICROPROFILE_SCOPE(OpenGL_Index);
const auto& regs = system.GPU().Maxwell3D().regs;
const std::size_t size = CalculateIndexBufferSize();
- const auto [buffer, offset] = buffer_cache.UploadMemory(regs.index_array.IndexStart(), size);
- glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, buffer);
- return offset;
+ const auto info = buffer_cache.UploadMemory(regs.index_array.IndexStart(), size);
+ glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, info.handle);
+ return info.offset;
}
void RasterizerOpenGL::SetupShaders(GLenum primitive_mode) {
@@ -643,9 +655,9 @@ void RasterizerOpenGL::Draw(bool is_indexed, bool is_instanced) {
if (!device.UseAssemblyShaders()) {
MaxwellUniformData ubo;
ubo.SetFromRegs(gpu);
- const auto [buffer, offset] =
+ const auto info =
buffer_cache.UploadHostMemory(&ubo, sizeof(ubo), device.GetUniformBufferAlignment());
- glBindBufferRange(GL_UNIFORM_BUFFER, EmulationUniformBlockBinding, buffer, offset,
+ glBindBufferRange(GL_UNIFORM_BUFFER, EmulationUniformBlockBinding, info.handle, info.offset,
static_cast<GLsizeiptr>(sizeof(ubo)));
}
@@ -956,8 +968,7 @@ void RasterizerOpenGL::SetupConstBuffer(GLenum stage, u32 binding,
if (device.UseAssemblyShaders()) {
glBindBufferRangeNV(stage, entry.GetIndex(), 0, 0, 0);
} else {
- glBindBufferRange(GL_UNIFORM_BUFFER, binding,
- buffer_cache.GetEmptyBuffer(sizeof(float)), 0, sizeof(float));
+ glBindBufferRange(GL_UNIFORM_BUFFER, binding, 0, 0, sizeof(float));
}
return;
}
@@ -970,24 +981,25 @@ void RasterizerOpenGL::SetupConstBuffer(GLenum stage, u32 binding,
const std::size_t alignment = use_unified ? 4 : device.GetUniformBufferAlignment();
const GPUVAddr gpu_addr = buffer.address;
- auto [cbuf, offset] = buffer_cache.UploadMemory(gpu_addr, size, alignment, false, fast_upload);
+ auto info = buffer_cache.UploadMemory(gpu_addr, size, alignment, false, fast_upload);
if (device.UseAssemblyShaders()) {
UNIMPLEMENTED_IF(use_unified);
- if (offset != 0) {
+ if (info.offset != 0) {
const GLuint staging_cbuf = staging_cbufs[current_cbuf++];
- glCopyNamedBufferSubData(cbuf, staging_cbuf, offset, 0, size);
- cbuf = staging_cbuf;
- offset = 0;
+ glCopyNamedBufferSubData(info.handle, staging_cbuf, info.offset, 0, size);
+ info.handle = staging_cbuf;
+ info.offset = 0;
}
- glBindBufferRangeNV(stage, binding, cbuf, offset, size);
+ glBindBufferRangeNV(stage, binding, info.handle, info.offset, size);
return;
}
if (use_unified) {
- glCopyNamedBufferSubData(cbuf, unified_uniform_buffer.handle, offset, unified_offset, size);
+ glCopyNamedBufferSubData(info.handle, unified_uniform_buffer.handle, info.offset,
+ unified_offset, size);
} else {
- glBindBufferRange(GL_UNIFORM_BUFFER, binding, cbuf, offset, size);
+ glBindBufferRange(GL_UNIFORM_BUFFER, binding, info.handle, info.offset, size);
}
}
@@ -1023,9 +1035,8 @@ void RasterizerOpenGL::SetupComputeGlobalMemory(Shader* kernel) {
void RasterizerOpenGL::SetupGlobalMemory(u32 binding, const GlobalMemoryEntry& entry,
GPUVAddr gpu_addr, std::size_t size) {
const auto alignment{device.GetShaderStorageBufferAlignment()};
- const auto [ssbo, buffer_offset] =
- buffer_cache.UploadMemory(gpu_addr, size, alignment, entry.is_written);
- glBindBufferRange(GL_SHADER_STORAGE_BUFFER, binding, ssbo, buffer_offset,
+ const auto info = buffer_cache.UploadMemory(gpu_addr, size, alignment, entry.is_written);
+ glBindBufferRange(GL_SHADER_STORAGE_BUFFER, binding, info.handle, info.offset,
static_cast<GLsizeiptr>(size));
}
@@ -1712,8 +1723,9 @@ void RasterizerOpenGL::EndTransformFeedback() {
const GLuint handle = transform_feedback_buffers[index].handle;
const GPUVAddr gpu_addr = binding.Address();
const std::size_t size = binding.buffer_size;
- const auto [dest_buffer, offset] = buffer_cache.UploadMemory(gpu_addr, size, 4, true);
- glCopyNamedBufferSubData(handle, dest_buffer, 0, offset, static_cast<GLsizeiptr>(size));
+ const auto info = buffer_cache.UploadMemory(gpu_addr, size, 4, true);
+ glCopyNamedBufferSubData(handle, info.handle, 0, info.offset,
+ static_cast<GLsizeiptr>(size));
}
}
diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
index d6e30b321..2c49aeaac 100644
--- a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
@@ -37,6 +37,7 @@ using Tegra::Shader::IpaMode;
using Tegra::Shader::IpaSampleMode;
using Tegra::Shader::PixelImap;
using Tegra::Shader::Register;
+using Tegra::Shader::TextureType;
using VideoCommon::Shader::BuildTransformFeedback;
using VideoCommon::Shader::Registry;
@@ -526,6 +527,9 @@ private:
if (device.HasImageLoadFormatted()) {
code.AddLine("#extension GL_EXT_shader_image_load_formatted : require");
}
+ if (device.HasTextureShadowLod()) {
+ code.AddLine("#extension GL_EXT_texture_shadow_lod : require");
+ }
if (device.HasWarpIntrinsics()) {
code.AddLine("#extension GL_NV_gpu_shader5 : require");
code.AddLine("#extension GL_NV_shader_thread_group : require");
@@ -909,13 +913,13 @@ private:
return "samplerBuffer";
}
switch (sampler.type) {
- case Tegra::Shader::TextureType::Texture1D:
+ case TextureType::Texture1D:
return "sampler1D";
- case Tegra::Shader::TextureType::Texture2D:
+ case TextureType::Texture2D:
return "sampler2D";
- case Tegra::Shader::TextureType::Texture3D:
+ case TextureType::Texture3D:
return "sampler3D";
- case Tegra::Shader::TextureType::TextureCube:
+ case TextureType::TextureCube:
return "samplerCube";
default:
UNREACHABLE();
@@ -1380,8 +1384,19 @@ private:
const std::size_t count = operation.GetOperandsCount();
const bool has_array = meta->sampler.is_array;
const bool has_shadow = meta->sampler.is_shadow;
+ const bool workaround_lod_array_shadow_as_grad =
+ !device.HasTextureShadowLod() && function_suffix == "Lod" && meta->sampler.is_shadow &&
+ ((meta->sampler.type == TextureType::Texture2D && meta->sampler.is_array) ||
+ meta->sampler.type == TextureType::TextureCube);
+
+ std::string expr = "texture";
+
+ if (workaround_lod_array_shadow_as_grad) {
+ expr += "Grad";
+ } else {
+ expr += function_suffix;
+ }
- std::string expr = "texture" + function_suffix;
if (!meta->aoffi.empty()) {
expr += "Offset";
} else if (!meta->ptp.empty()) {
@@ -1415,6 +1430,16 @@ private:
expr += ')';
}
+ if (workaround_lod_array_shadow_as_grad) {
+ switch (meta->sampler.type) {
+ case TextureType::Texture2D:
+ return expr + ", vec2(0.0), vec2(0.0))";
+ case TextureType::TextureCube:
+ return expr + ", vec3(0.0), vec3(0.0))";
+ }
+ UNREACHABLE();
+ }
+
for (const auto& variant : extras) {
if (const auto argument = std::get_if<TextureArgument>(&variant)) {
expr += GenerateTextureArgument(*argument);
@@ -2041,8 +2066,19 @@ private:
const auto meta = std::get_if<MetaTexture>(&operation.GetMeta());
ASSERT(meta);
- std::string expr = GenerateTexture(
- operation, "Lod", {TextureArgument{Type::Float, meta->lod}, TextureOffset{}});
+ std::string expr{};
+
+ if (!device.HasTextureShadowLod() && meta->sampler.is_shadow &&
+ ((meta->sampler.type == TextureType::Texture2D && meta->sampler.is_array) ||
+ meta->sampler.type == TextureType::TextureCube)) {
+ LOG_ERROR(Render_OpenGL,
+ "Device lacks GL_EXT_texture_shadow_lod, using textureGrad as a workaround");
+ expr = GenerateTexture(operation, "Lod", {});
+ } else {
+ expr = GenerateTexture(operation, "Lod",
+ {TextureArgument{Type::Float, meta->lod}, TextureOffset{}});
+ }
+
if (meta->sampler.is_shadow) {
expr = "vec4(" + expr + ')';
}
diff --git a/src/video_core/renderer_opengl/gl_stream_buffer.cpp b/src/video_core/renderer_opengl/gl_stream_buffer.cpp
index 932a2f69e..3655ff629 100644
--- a/src/video_core/renderer_opengl/gl_stream_buffer.cpp
+++ b/src/video_core/renderer_opengl/gl_stream_buffer.cpp
@@ -2,11 +2,13 @@
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
-#include <deque>
+#include <tuple>
#include <vector>
+
#include "common/alignment.h"
#include "common/assert.h"
#include "common/microprofile.h"
+#include "video_core/renderer_opengl/gl_device.h"
#include "video_core/renderer_opengl/gl_stream_buffer.h"
MICROPROFILE_DEFINE(OpenGL_StreamBuffer, "OpenGL", "Stream Buffer Orphaning",
@@ -14,8 +16,7 @@ MICROPROFILE_DEFINE(OpenGL_StreamBuffer, "OpenGL", "Stream Buffer Orphaning",
namespace OpenGL {
-OGLStreamBuffer::OGLStreamBuffer(GLsizeiptr size, bool vertex_data_usage, bool prefer_coherent,
- bool use_persistent)
+OGLStreamBuffer::OGLStreamBuffer(const Device& device, GLsizeiptr size, bool vertex_data_usage)
: buffer_size(size) {
gl_buffer.Create();
@@ -29,23 +30,19 @@ OGLStreamBuffer::OGLStreamBuffer(GLsizeiptr size, bool vertex_data_usage, bool p
allocate_size *= 2;
}
- if (use_persistent) {
- persistent = true;
- coherent = prefer_coherent;
- const GLbitfield flags =
- GL_MAP_WRITE_BIT | GL_MAP_PERSISTENT_BIT | (coherent ? GL_MAP_COHERENT_BIT : 0);
- glNamedBufferStorage(gl_buffer.handle, allocate_size, nullptr, flags);
- mapped_ptr = static_cast<u8*>(glMapNamedBufferRange(
- gl_buffer.handle, 0, buffer_size, flags | (coherent ? 0 : GL_MAP_FLUSH_EXPLICIT_BIT)));
- } else {
- glNamedBufferData(gl_buffer.handle, allocate_size, nullptr, GL_STREAM_DRAW);
+ static constexpr GLbitfield flags = GL_MAP_WRITE_BIT | GL_MAP_PERSISTENT_BIT;
+ glNamedBufferStorage(gl_buffer.handle, allocate_size, nullptr, flags);
+ mapped_ptr = static_cast<u8*>(
+ glMapNamedBufferRange(gl_buffer.handle, 0, buffer_size, flags | GL_MAP_FLUSH_EXPLICIT_BIT));
+
+ if (device.HasVertexBufferUnifiedMemory()) {
+ glMakeNamedBufferResidentNV(gl_buffer.handle, GL_READ_ONLY);
+ glGetNamedBufferParameterui64vNV(gl_buffer.handle, GL_BUFFER_GPU_ADDRESS_NV, &gpu_address);
}
}
OGLStreamBuffer::~OGLStreamBuffer() {
- if (persistent) {
- glUnmapNamedBuffer(gl_buffer.handle);
- }
+ glUnmapNamedBuffer(gl_buffer.handle);
gl_buffer.Release();
}
@@ -60,36 +57,21 @@ std::tuple<u8*, GLintptr, bool> OGLStreamBuffer::Map(GLsizeiptr size, GLintptr a
bool invalidate = false;
if (buffer_pos + size > buffer_size) {
+ MICROPROFILE_SCOPE(OpenGL_StreamBuffer);
+ glInvalidateBufferData(gl_buffer.handle);
+
buffer_pos = 0;
invalidate = true;
-
- if (persistent) {
- glUnmapNamedBuffer(gl_buffer.handle);
- }
}
- if (invalidate || !persistent) {
- MICROPROFILE_SCOPE(OpenGL_StreamBuffer);
- GLbitfield flags = GL_MAP_WRITE_BIT | (persistent ? GL_MAP_PERSISTENT_BIT : 0) |
- (coherent ? GL_MAP_COHERENT_BIT : GL_MAP_FLUSH_EXPLICIT_BIT) |
- (invalidate ? GL_MAP_INVALIDATE_BUFFER_BIT : GL_MAP_UNSYNCHRONIZED_BIT);
- mapped_ptr = static_cast<u8*>(
- glMapNamedBufferRange(gl_buffer.handle, buffer_pos, buffer_size - buffer_pos, flags));
- mapped_offset = buffer_pos;
- }
-
- return std::make_tuple(mapped_ptr + buffer_pos - mapped_offset, buffer_pos, invalidate);
+ return std::make_tuple(mapped_ptr + buffer_pos, buffer_pos, invalidate);
}
void OGLStreamBuffer::Unmap(GLsizeiptr size) {
ASSERT(size <= mapped_size);
- if (!coherent && size > 0) {
- glFlushMappedNamedBufferRange(gl_buffer.handle, buffer_pos - mapped_offset, size);
- }
-
- if (!persistent) {
- glUnmapNamedBuffer(gl_buffer.handle);
+ if (size > 0) {
+ glFlushMappedNamedBufferRange(gl_buffer.handle, buffer_pos, size);
}
buffer_pos += size;
diff --git a/src/video_core/renderer_opengl/gl_stream_buffer.h b/src/video_core/renderer_opengl/gl_stream_buffer.h
index 866da3594..307a67113 100644
--- a/src/video_core/renderer_opengl/gl_stream_buffer.h
+++ b/src/video_core/renderer_opengl/gl_stream_buffer.h
@@ -11,10 +11,11 @@
namespace OpenGL {
+class Device;
+
class OGLStreamBuffer : private NonCopyable {
public:
- explicit OGLStreamBuffer(GLsizeiptr size, bool vertex_data_usage, bool prefer_coherent = false,
- bool use_persistent = true);
+ explicit OGLStreamBuffer(const Device& device, GLsizeiptr size, bool vertex_data_usage);
~OGLStreamBuffer();
/*
@@ -33,19 +34,20 @@ public:
return gl_buffer.handle;
}
- GLsizeiptr Size() const {
+ u64 Address() const {
+ return gpu_address;
+ }
+
+ GLsizeiptr Size() const noexcept {
return buffer_size;
}
private:
OGLBuffer gl_buffer;
- bool coherent = false;
- bool persistent = false;
-
+ GLuint64EXT gpu_address = 0;
GLintptr buffer_pos = 0;
GLsizeiptr buffer_size = 0;
- GLintptr mapped_offset = 0;
GLsizeiptr mapped_size = 0;
u8* mapped_ptr = nullptr;
};
diff --git a/src/video_core/renderer_opengl/renderer_opengl.cpp b/src/video_core/renderer_opengl/renderer_opengl.cpp
index 6214fcbc3..c40adb6e7 100644
--- a/src/video_core/renderer_opengl/renderer_opengl.cpp
+++ b/src/video_core/renderer_opengl/renderer_opengl.cpp
@@ -488,6 +488,15 @@ void RendererOpenGL::InitOpenGLObjects() {
// Clear screen to black
LoadColorToActiveGLTexture(0, 0, 0, 0, screen_info.texture);
+
+ // Enable unified vertex attributes and query vertex buffer address when the driver supports it
+ if (device.HasVertexBufferUnifiedMemory()) {
+ glEnableClientState(GL_VERTEX_ATTRIB_ARRAY_UNIFIED_NV);
+
+ glMakeNamedBufferResidentNV(vertex_buffer.handle, GL_READ_ONLY);
+ glGetNamedBufferParameterui64vNV(vertex_buffer.handle, GL_BUFFER_GPU_ADDRESS_NV,
+ &vertex_buffer_address);
+ }
}
void RendererOpenGL::AddTelemetryFields() {
@@ -656,7 +665,13 @@ void RendererOpenGL::DrawScreen(const Layout::FramebufferLayout& layout) {
offsetof(ScreenRectVertex, tex_coord));
glVertexAttribBinding(PositionLocation, 0);
glVertexAttribBinding(TexCoordLocation, 0);
- glBindVertexBuffer(0, vertex_buffer.handle, 0, sizeof(ScreenRectVertex));
+ if (device.HasVertexBufferUnifiedMemory()) {
+ glBindVertexBuffer(0, 0, 0, sizeof(ScreenRectVertex));
+ glBufferAddressRangeNV(GL_VERTEX_ATTRIB_ARRAY_ADDRESS_NV, 0, vertex_buffer_address,
+ sizeof(vertices));
+ } else {
+ glBindVertexBuffer(0, vertex_buffer.handle, 0, sizeof(ScreenRectVertex));
+ }
glBindTextureUnit(0, screen_info.display_texture);
glBindSampler(0, 0);
diff --git a/src/video_core/renderer_opengl/renderer_opengl.h b/src/video_core/renderer_opengl/renderer_opengl.h
index 61bf507f4..8b18d32e6 100644
--- a/src/video_core/renderer_opengl/renderer_opengl.h
+++ b/src/video_core/renderer_opengl/renderer_opengl.h
@@ -107,6 +107,9 @@ private:
OGLPipeline pipeline;
OGLFramebuffer screenshot_framebuffer;
+ // GPU address of the vertex buffer
+ GLuint64EXT vertex_buffer_address = 0;
+
/// Display information for Switch screen
ScreenInfo screen_info;
diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
index 1fde38328..f10f96cd8 100644
--- a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
+++ b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
@@ -37,9 +37,9 @@ std::unique_ptr<VKStreamBuffer> CreateStreamBuffer(const VKDevice& device, VKSch
} // Anonymous namespace
-Buffer::Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VAddr cpu_addr,
- std::size_t size)
- : VideoCommon::BufferBlock{cpu_addr, size} {
+Buffer::Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VKScheduler& scheduler_,
+ VKStagingBufferPool& staging_pool_, VAddr cpu_addr, std::size_t size)
+ : VideoCommon::BufferBlock{cpu_addr, size}, scheduler{scheduler_}, staging_pool{staging_pool_} {
VkBufferCreateInfo ci;
ci.sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO;
ci.pNext = nullptr;
@@ -56,40 +56,15 @@ Buffer::Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VAddr cp
Buffer::~Buffer() = default;
-VKBufferCache::VKBufferCache(VideoCore::RasterizerInterface& rasterizer, Core::System& system,
- const VKDevice& device, VKMemoryManager& memory_manager,
- VKScheduler& scheduler, VKStagingBufferPool& staging_pool)
- : VideoCommon::BufferCache<Buffer, VkBuffer, VKStreamBuffer>{rasterizer, system,
- CreateStreamBuffer(device,
- scheduler)},
- device{device}, memory_manager{memory_manager}, scheduler{scheduler}, staging_pool{
- staging_pool} {}
-
-VKBufferCache::~VKBufferCache() = default;
-
-std::shared_ptr<Buffer> VKBufferCache::CreateBlock(VAddr cpu_addr, std::size_t size) {
- return std::make_shared<Buffer>(device, memory_manager, cpu_addr, size);
-}
-
-VkBuffer VKBufferCache::GetEmptyBuffer(std::size_t size) {
- size = std::max(size, std::size_t(4));
- const auto& empty = staging_pool.GetUnusedBuffer(size, false);
- scheduler.RequestOutsideRenderPassOperationContext();
- scheduler.Record([size, buffer = *empty.handle](vk::CommandBuffer cmdbuf) {
- cmdbuf.FillBuffer(buffer, 0, size, 0);
- });
- return *empty.handle;
-}
-
-void VKBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- const u8* data) {
+void Buffer::Upload(std::size_t offset, std::size_t size, const u8* data) const {
const auto& staging = staging_pool.GetUnusedBuffer(size, true);
std::memcpy(staging.commit->Map(size), data, size);
scheduler.RequestOutsideRenderPassOperationContext();
- scheduler.Record([staging = *staging.handle, buffer = buffer.Handle(), offset,
- size](vk::CommandBuffer cmdbuf) {
- cmdbuf.CopyBuffer(staging, buffer, VkBufferCopy{0, offset, size});
+
+ const VkBuffer handle = Handle();
+ scheduler.Record([staging = *staging.handle, handle, offset, size](vk::CommandBuffer cmdbuf) {
+ cmdbuf.CopyBuffer(staging, handle, VkBufferCopy{0, offset, size});
VkBufferMemoryBarrier barrier;
barrier.sType = VK_STRUCTURE_TYPE_BUFFER_MEMORY_BARRIER;
@@ -98,7 +73,7 @@ void VKBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, st
barrier.dstAccessMask = UPLOAD_ACCESS_BARRIERS;
barrier.srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
barrier.dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
- barrier.buffer = buffer;
+ barrier.buffer = handle;
barrier.offset = offset;
barrier.size = size;
cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_TRANSFER_BIT, UPLOAD_PIPELINE_STAGE, 0, {},
@@ -106,12 +81,12 @@ void VKBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, st
});
}
-void VKBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- u8* data) {
+void Buffer::Download(std::size_t offset, std::size_t size, u8* data) const {
const auto& staging = staging_pool.GetUnusedBuffer(size, true);
scheduler.RequestOutsideRenderPassOperationContext();
- scheduler.Record([staging = *staging.handle, buffer = buffer.Handle(), offset,
- size](vk::CommandBuffer cmdbuf) {
+
+ const VkBuffer handle = Handle();
+ scheduler.Record([staging = *staging.handle, handle, offset, size](vk::CommandBuffer cmdbuf) {
VkBufferMemoryBarrier barrier;
barrier.sType = VK_STRUCTURE_TYPE_BUFFER_MEMORY_BARRIER;
barrier.pNext = nullptr;
@@ -119,7 +94,7 @@ void VKBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset,
barrier.dstAccessMask = VK_ACCESS_TRANSFER_READ_BIT;
barrier.srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
barrier.dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
- barrier.buffer = buffer;
+ barrier.buffer = handle;
barrier.offset = offset;
barrier.size = size;
@@ -127,17 +102,19 @@ void VKBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset,
VK_PIPELINE_STAGE_FRAGMENT_SHADER_BIT |
VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT,
VK_PIPELINE_STAGE_TRANSFER_BIT, 0, {}, barrier, {});
- cmdbuf.CopyBuffer(buffer, staging, VkBufferCopy{offset, 0, size});
+ cmdbuf.CopyBuffer(handle, staging, VkBufferCopy{offset, 0, size});
});
scheduler.Finish();
std::memcpy(data, staging.commit->Map(size), size);
}
-void VKBufferCache::CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset,
- std::size_t dst_offset, std::size_t size) {
+void Buffer::CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset,
+ std::size_t size) const {
scheduler.RequestOutsideRenderPassOperationContext();
- scheduler.Record([src_buffer = src.Handle(), dst_buffer = dst.Handle(), src_offset, dst_offset,
+
+ const VkBuffer dst_buffer = Handle();
+ scheduler.Record([src_buffer = src.Handle(), dst_buffer, src_offset, dst_offset,
size](vk::CommandBuffer cmdbuf) {
cmdbuf.CopyBuffer(src_buffer, dst_buffer, VkBufferCopy{src_offset, dst_offset, size});
@@ -165,4 +142,30 @@ void VKBufferCache::CopyBlock(const Buffer& src, const Buffer& dst, std::size_t
});
}
+VKBufferCache::VKBufferCache(VideoCore::RasterizerInterface& rasterizer, Core::System& system,
+ const VKDevice& device, VKMemoryManager& memory_manager,
+ VKScheduler& scheduler, VKStagingBufferPool& staging_pool)
+ : VideoCommon::BufferCache<Buffer, VkBuffer, VKStreamBuffer>{rasterizer, system,
+ CreateStreamBuffer(device,
+ scheduler)},
+ device{device}, memory_manager{memory_manager}, scheduler{scheduler}, staging_pool{
+ staging_pool} {}
+
+VKBufferCache::~VKBufferCache() = default;
+
+std::shared_ptr<Buffer> VKBufferCache::CreateBlock(VAddr cpu_addr, std::size_t size) {
+ return std::make_shared<Buffer>(device, memory_manager, scheduler, staging_pool, cpu_addr,
+ size);
+}
+
+VKBufferCache::BufferInfo VKBufferCache::GetEmptyBuffer(std::size_t size) {
+ size = std::max(size, std::size_t(4));
+ const auto& empty = staging_pool.GetUnusedBuffer(size, false);
+ scheduler.RequestOutsideRenderPassOperationContext();
+ scheduler.Record([size, buffer = *empty.handle](vk::CommandBuffer cmdbuf) {
+ cmdbuf.FillBuffer(buffer, 0, size, 0);
+ });
+ return {*empty.handle, 0, 0};
+}
+
} // namespace Vulkan
diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.h b/src/video_core/renderer_vulkan/vk_buffer_cache.h
index 9ebbef835..3630aca77 100644
--- a/src/video_core/renderer_vulkan/vk_buffer_cache.h
+++ b/src/video_core/renderer_vulkan/vk_buffer_cache.h
@@ -25,15 +25,29 @@ class VKScheduler;
class Buffer final : public VideoCommon::BufferBlock {
public:
- explicit Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VAddr cpu_addr,
- std::size_t size);
+ explicit Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VKScheduler& scheduler,
+ VKStagingBufferPool& staging_pool, VAddr cpu_addr, std::size_t size);
~Buffer();
+ void Upload(std::size_t offset, std::size_t size, const u8* data) const;
+
+ void Download(std::size_t offset, std::size_t size, u8* data) const;
+
+ void CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset,
+ std::size_t size) const;
+
VkBuffer Handle() const {
return *buffer.handle;
}
+ u64 Address() const {
+ return 0;
+ }
+
private:
+ VKScheduler& scheduler;
+ VKStagingBufferPool& staging_pool;
+
VKBuffer buffer;
};
@@ -44,20 +58,11 @@ public:
VKScheduler& scheduler, VKStagingBufferPool& staging_pool);
~VKBufferCache();
- VkBuffer GetEmptyBuffer(std::size_t size) override;
+ BufferInfo GetEmptyBuffer(std::size_t size) override;
protected:
std::shared_ptr<Buffer> CreateBlock(VAddr cpu_addr, std::size_t size) override;
- void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- const u8* data) override;
-
- void DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size,
- u8* data) override;
-
- void CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset,
- std::size_t dst_offset, std::size_t size) override;
-
private:
const VKDevice& device;
VKMemoryManager& memory_manager;
diff --git a/src/video_core/renderer_vulkan/vk_rasterizer.cpp b/src/video_core/renderer_vulkan/vk_rasterizer.cpp
index 29001953c..e3714ee6d 100644
--- a/src/video_core/renderer_vulkan/vk_rasterizer.cpp
+++ b/src/video_core/renderer_vulkan/vk_rasterizer.cpp
@@ -870,10 +870,10 @@ void RasterizerVulkan::BeginTransformFeedback() {
UNIMPLEMENTED_IF(binding.buffer_offset != 0);
const GPUVAddr gpu_addr = binding.Address();
- const auto size = static_cast<VkDeviceSize>(binding.buffer_size);
- const auto [buffer, offset] = buffer_cache.UploadMemory(gpu_addr, size, 4, true);
+ const VkDeviceSize size = static_cast<VkDeviceSize>(binding.buffer_size);
+ const auto info = buffer_cache.UploadMemory(gpu_addr, size, 4, true);
- scheduler.Record([buffer = buffer, offset = offset, size](vk::CommandBuffer cmdbuf) {
+ scheduler.Record([buffer = info.handle, offset = info.offset, size](vk::CommandBuffer cmdbuf) {
cmdbuf.BindTransformFeedbackBuffersEXT(0, 1, &buffer, &offset, &size);
cmdbuf.BeginTransformFeedbackEXT(0, 0, nullptr, nullptr);
});
@@ -925,8 +925,8 @@ void RasterizerVulkan::SetupVertexArrays(FixedPipelineState::VertexInput& vertex
buffer_bindings.AddVertexBinding(DefaultBuffer(), 0);
continue;
}
- const auto [buffer, offset] = buffer_cache.UploadMemory(start, size);
- buffer_bindings.AddVertexBinding(buffer, offset);
+ const auto info = buffer_cache.UploadMemory(start, size);
+ buffer_bindings.AddVertexBinding(info.handle, info.offset);
}
}
@@ -948,7 +948,9 @@ void RasterizerVulkan::SetupIndexBuffer(BufferBindings& buffer_bindings, DrawPar
break;
}
const GPUVAddr gpu_addr = regs.index_array.IndexStart();
- auto [buffer, offset] = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize());
+ const auto info = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize());
+ VkBuffer buffer = info.handle;
+ u64 offset = info.offset;
std::tie(buffer, offset) = quad_indexed_pass.Assemble(
regs.index_array.format, params.num_vertices, params.base_vertex, buffer, offset);
@@ -962,7 +964,9 @@ void RasterizerVulkan::SetupIndexBuffer(BufferBindings& buffer_bindings, DrawPar
break;
}
const GPUVAddr gpu_addr = regs.index_array.IndexStart();
- auto [buffer, offset] = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize());
+ const auto info = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize());
+ VkBuffer buffer = info.handle;
+ u64 offset = info.offset;
auto format = regs.index_array.format;
const bool is_uint8 = format == Maxwell::IndexFormat::UnsignedByte;
@@ -1109,10 +1113,9 @@ void RasterizerVulkan::SetupConstBuffer(const ConstBufferEntry& entry,
Common::AlignUp(CalculateConstBufferSize(entry, buffer), 4 * sizeof(float));
ASSERT(size <= MaxConstbufferSize);
- const auto [buffer_handle, offset] =
+ const auto info =
buffer_cache.UploadMemory(buffer.address, size, device.GetUniformBufferAlignment());
-
- update_descriptor_queue.AddBuffer(buffer_handle, offset, size);
+ update_descriptor_queue.AddBuffer(info.handle, info.offset, size);
}
void RasterizerVulkan::SetupGlobalBuffer(const GlobalBufferEntry& entry, GPUVAddr address) {
@@ -1126,14 +1129,14 @@ void RasterizerVulkan::SetupGlobalBuffer(const GlobalBufferEntry& entry, GPUVAdd
// Note: Do *not* use DefaultBuffer() here, storage buffers can be written breaking the
// default buffer.
static constexpr std::size_t dummy_size = 4;
- const auto buffer = buffer_cache.GetEmptyBuffer(dummy_size);
- update_descriptor_queue.AddBuffer(buffer, 0, dummy_size);
+ const auto info = buffer_cache.GetEmptyBuffer(dummy_size);
+ update_descriptor_queue.AddBuffer(info.handle, info.offset, dummy_size);
return;
}
- const auto [buffer, offset] = buffer_cache.UploadMemory(
+ const auto info = buffer_cache.UploadMemory(
actual_addr, size, device.GetStorageBufferAlignment(), entry.IsWritten());
- update_descriptor_queue.AddBuffer(buffer, offset, size);
+ update_descriptor_queue.AddBuffer(info.handle, info.offset, size);
}
void RasterizerVulkan::SetupUniformTexels(const Tegra::Texture::TICEntry& tic,
diff --git a/src/video_core/renderer_vulkan/vk_stream_buffer.h b/src/video_core/renderer_vulkan/vk_stream_buffer.h
index c765c60a0..689f0d276 100644
--- a/src/video_core/renderer_vulkan/vk_stream_buffer.h
+++ b/src/video_core/renderer_vulkan/vk_stream_buffer.h
@@ -35,10 +35,14 @@ public:
/// Ensures that "size" bytes of memory are available to the GPU, potentially recording a copy.
void Unmap(u64 size);
- VkBuffer Handle() const {
+ VkBuffer Handle() const noexcept {
return *buffer;
}
+ u64 Address() const noexcept {
+ return 0;
+ }
+
private:
struct Watch final {
VKFenceWatch fence;
diff --git a/src/video_core/shader/decode/image.cpp b/src/video_core/shader/decode/image.cpp
index 60b6ad72a..07778dc3e 100644
--- a/src/video_core/shader/decode/image.cpp
+++ b/src/video_core/shader/decode/image.cpp
@@ -97,6 +97,7 @@ ComponentType GetComponentType(Tegra::Engines::SamplerDescriptor descriptor,
break;
case TextureFormat::B5G6R5:
case TextureFormat::B6G5R5:
+ case TextureFormat::BF10GF11RF11:
if (component == 0) {
return descriptor.b_type;
}
@@ -119,7 +120,7 @@ ComponentType GetComponentType(Tegra::Engines::SamplerDescriptor descriptor,
}
break;
}
- UNIMPLEMENTED_MSG("texture format not implement={}", format);
+ UNIMPLEMENTED_MSG("Texture format not implemented={}", format);
return ComponentType::FLOAT;
}
@@ -191,6 +192,14 @@ u32 GetComponentSize(TextureFormat format, std::size_t component) {
return 6;
}
return 0;
+ case TextureFormat::BF10GF11RF11:
+ if (component == 1 || component == 2) {
+ return 11;
+ }
+ if (component == 0) {
+ return 10;
+ }
+ return 0;
case TextureFormat::G8R24:
if (component == 0) {
return 8;
@@ -211,10 +220,9 @@ u32 GetComponentSize(TextureFormat format, std::size_t component) {
return (component == 0 || component == 1) ? 8 : 0;
case TextureFormat::G4R4:
return (component == 0 || component == 1) ? 4 : 0;
- default:
- UNIMPLEMENTED_MSG("texture format not implement={}", format);
- return 0;
}
+ UNIMPLEMENTED_MSG("Texture format not implemented={}", format);
+ return 0;
}
std::size_t GetImageComponentMask(TextureFormat format) {
@@ -235,6 +243,7 @@ std::size_t GetImageComponentMask(TextureFormat format) {
case TextureFormat::R32_B24G8:
case TextureFormat::B5G6R5:
case TextureFormat::B6G5R5:
+ case TextureFormat::BF10GF11RF11:
return std::size_t{R | G | B};
case TextureFormat::R32_G32:
case TextureFormat::R16_G16:
@@ -248,10 +257,9 @@ std::size_t GetImageComponentMask(TextureFormat format) {
case TextureFormat::R8:
case TextureFormat::R1:
return std::size_t{R};
- default:
- UNIMPLEMENTED_MSG("texture format not implement={}", format);
- return std::size_t{R | G | B | A};
}
+ UNIMPLEMENTED_MSG("Texture format not implemented={}", format);
+ return std::size_t{R | G | B | A};
}
std::size_t GetImageTypeNumCoordinates(Tegra::Shader::ImageType image_type) {
@@ -299,7 +307,7 @@ std::pair<Node, bool> ShaderIR::GetComponentValue(ComponentType component_type,
return {std::move(original_value), true};
}
default:
- UNIMPLEMENTED_MSG("Unimplement component type={}", component_type);
+ UNIMPLEMENTED_MSG("Unimplemented component type={}", component_type);
return {std::move(original_value), true};
}
}
@@ -459,7 +467,7 @@ u32 ShaderIR::DecodeImage(NodeBlock& bb, u32 pc) {
default:
break;
}
- UNIMPLEMENTED_MSG("Unimplemented operation={} type={}",
+ UNIMPLEMENTED_MSG("Unimplemented operation={}, type={}",
static_cast<u64>(instr.suatom_d.operation.Value()),
static_cast<u64>(instr.suatom_d.operation_type.Value()));
return OperationCode::AtomicImageAdd;
diff --git a/src/video_core/texture_cache/texture_cache.h b/src/video_core/texture_cache/texture_cache.h
index b543fc8c0..85075e868 100644
--- a/src/video_core/texture_cache/texture_cache.h
+++ b/src/video_core/texture_cache/texture_cache.h
@@ -1053,7 +1053,7 @@ private:
void DeduceBestBlit(SurfaceParams& src_params, SurfaceParams& dst_params,
const GPUVAddr src_gpu_addr, const GPUVAddr dst_gpu_addr) {
auto deduced_src = DeduceSurface(src_gpu_addr, src_params);
- auto deduced_dst = DeduceSurface(src_gpu_addr, src_params);
+ auto deduced_dst = DeduceSurface(dst_gpu_addr, dst_params);
if (deduced_src.Failed() || deduced_dst.Failed()) {
return;
}
diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp
index 4119d7907..059b96e70 100644
--- a/src/yuzu/main.cpp
+++ b/src/yuzu/main.cpp
@@ -217,7 +217,20 @@ GMainWindow::GMainWindow()
LOG_INFO(Frontend, "yuzu Version: {} | {}-{}", yuzu_build_version, Common::g_scm_branch,
Common::g_scm_desc);
#ifdef ARCHITECTURE_x86_64
- LOG_INFO(Frontend, "Host CPU: {}", Common::GetCPUCaps().cpu_string);
+ const auto& caps = Common::GetCPUCaps();
+ std::string cpu_string = caps.cpu_string;
+ if (caps.avx || caps.avx2 || caps.avx512) {
+ cpu_string += " | AVX";
+ if (caps.avx512) {
+ cpu_string += "512";
+ } else if (caps.avx2) {
+ cpu_string += '2';
+ }
+ if (caps.fma || caps.fma4) {
+ cpu_string += " | FMA";
+ }
+ }
+ LOG_INFO(Frontend, "Host CPU: {}", cpu_string);
#endif
LOG_INFO(Frontend, "Host OS: {}", QSysInfo::prettyProductName().toStdString());
LOG_INFO(Frontend, "Host RAM: {:.2f} GB",
diff --git a/src/yuzu_tester/service/yuzutest.cpp b/src/yuzu_tester/service/yuzutest.cpp
index 85d3f436b..2d3f6e3a7 100644
--- a/src/yuzu_tester/service/yuzutest.cpp
+++ b/src/yuzu_tester/service/yuzutest.cpp
@@ -53,7 +53,7 @@ private:
IPC::ResponseBuilder rb{ctx, 3};
rb.Push(RESULT_SUCCESS);
- rb.Push<u32>(write_size);
+ rb.Push<u32>(static_cast<u32>(write_size));
}
void StartIndividual(Kernel::HLERequestContext& ctx) {