wxgui branch: Maintenance merge against trunk, plus many cleanups and project-level changes.

* Moved the x86 emitter to /common, so that plugins can link against it if they wish.
 * Created a new "utility" class in /common which houses string utils, fast memcpy, common exception classes, and other handy dandies.
 * Removed old-style linux automake files from the pcsx2 dir since they were hopelessly out of date (and their multi-file-per-line format makes svn merging impossible >_<)

git-svn-id: http://pcsx2.googlecode.com/svn/branches/wxgui@1454 96395faa-99c1-11dd-bbfe-3dabce05a288

3rdparty/3rdpartyDLL.vsprops

@@ -0,0 +1,26 @@
+<?xml version="1.0" encoding="windows-1250"?>
+<VisualStudioPropertySheet
+	ProjectType="Visual C++"
+	Version="8.00"
+	Name="3rdpartyDLL"
+	OutputDirectory="$(SvnRootDir)\deps\$(PlatformName)\$(ConfigurationName)"
+	IntermediateDirectory="$(PlatformName)\$(ConfigurationName)"
+	>
+	<Tool
+		Name="VCCLCompilerTool"
+		AdditionalIncludeDirectories="&quot;$(ProjectDir)&quot;"
+		PreprocessorDefinitions="__WIN32__;WIN32;_WINDOWS;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE"
+		StructMemberAlignment="5"
+		RuntimeTypeInfo="false"
+		WarningLevel="3"
+		DebugInformationFormat="3"
+	/>
+	<Tool
+		Name="VCLinkerTool"
+		OutputFile="$(SolutionDir)\bin\$(ProjectName).dll"
+		LinkIncremental="1"
+		GenerateDebugInformation="true"
+		SubSystem="2"
+		ImportLibrary="$(OutDir)\$(ProjectName).lib"
+	/>
+</VisualStudioPropertySheet>

3rdparty/DefaultProjectRootDir.vsprops

@@ -0,0 +1,15 @@
+<?xml version="1.0" encoding="Windows-1252"?>
+<VisualStudioPropertySheet
+	ProjectType="Visual C++"
+	Version="8.00"
+	Name="DefaultProjectRootDir"
+	>
+	<UserMacro
+		Name="ProjectRootDir"
+		Value="$(ProjectDir)"
+	/>
+	<UserMacro
+		Name="SvnRootDir"
+		Value="$(ProjectRootDir)\..\.."
+	/>
+</VisualStudioPropertySheet>

3rdparty/SoundTouch/Makefile.am

@@ -24,13 +24,8 @@ AUTOMAKE_OPTIONS = foreign
 noinst_HEADERS=AAFilter.h cpu_detect.h FIRFilter.h RateTransposer.h TDStretch.h cpu_detect_x86_gcc.cpp
 noinst_LIBRARIES = libSoundTouch.a
 
-if X86_64
-libSoundTouch_a_CXXFLAGS = -fPIC
-libSoundTouch_a_CFLAGS = -fPIC
-else
 libSoundTouch_a_CXXFLAGS = -msse -mmmx
 libSoundTouch_a_CFLAGS = -msse -mmmx
-endif
 
 #lib_LTLIBRARIES=libSoundTouch.la
 # the mmx_gcc.cpp and cpu_detect_x86_gcc.cpp may need to be conditionally included here from things discovered in configure.ac

3rdparty/SoundTouch/SoundTouch.vcproj

@@ -18,7 +18,7 @@
 		<Configuration
 			Name="Debug|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Debug.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			>
 			<Tool
@@ -38,9 +38,6 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="0"
-				BasicRuntimeChecks="3"
-				RuntimeLibrary="1"
 				WarningLevel="3"
 			/>
 			<Tool
@@ -74,7 +71,7 @@
 		<Configuration
 			Name="Release|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Release.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="1"
 			>
@@ -95,12 +92,7 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				EnableIntrinsicFunctions="true"
-				RuntimeLibrary="0"
-				EnableFunctionLevelLinking="true"
 				WarningLevel="3"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -133,7 +125,7 @@
 		<Configuration
 			Name="Devel|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Devel.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="1"
 			>
@@ -154,12 +146,7 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				EnableIntrinsicFunctions="true"
-				RuntimeLibrary="0"
-				BufferSecurityCheck="false"
 				WarningLevel="3"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"

3rdparty/SoundTouch/configure.ac

@@ -11,33 +11,15 @@ AC_PROG_CXX
 AC_PROG_CC
 AC_PROG_RANLIB
 
-dnl Check for 64bit CPU
+CFLAGS=
-AC_MSG_CHECKING(for a x86-64 CPU)
+CPPFLAGS=
-dnl if test "$build_os" == "target_os"
+CXXFLAGS=
-dnl then
+CCASFLAGS=
-AC_TRY_RUN([
-int main()
-{
-int a = 0;
-int*pa = &a;
-asm(".intel_syntax\n"
-	"mov %%rax, %0\n"
-    "mov %%eax, [%%rax]\n"
-    ".att_syntax\n"
-    : : "r"(pa) : "%rax");
-return 0;
-}
-],cpu64=yes,cpu64=no,)
-dnl else
-dnl cpu64=no
-dnl fi
-if test "x$cpu64" == xyes
-then
-AC_DEFINE(__x86_64__,1,[__x86_64__])
-fi
-AC_MSG_RESULT($cpu64)
-AM_CONDITIONAL(X86_64, test x$cpu64 = xyes)
 
+CFLAGS+=" -m32 "
+CPPFLAGS+=" -m32 "
+CXXFLAGS+=" -m32 "
+CCASFLAGS+=" -m32 "
 
 # Checks for header files.
 AC_CHECK_HEADERS([limits.h memory.h stdlib.h string.h])

3rdparty/bzip2/bzip2.vcproj

@@ -18,7 +18,7 @@
 		<Configuration
 			Name="Debug|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Debug.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			>
 			<Tool
@@ -38,11 +38,6 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="0"
-				BasicRuntimeChecks="3"
-				RuntimeLibrary="1"
-				WarningLevel="3"
-				DebugInformationFormat="4"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -75,7 +70,7 @@
 		<Configuration
 			Name="Release|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Release.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="1"
 			>
@@ -96,17 +91,6 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				InlineFunctionExpansion="2"
-				EnableIntrinsicFunctions="true"
-				FavorSizeOrSpeed="1"
-				OmitFramePointers="true"
-				StringPooling="true"
-				RuntimeLibrary="0"
-				BufferSecurityCheck="false"
-				EnableFunctionLevelLinking="true"
-				WarningLevel="3"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -139,7 +123,7 @@
 		<Configuration
 			Name="Devel|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Devel.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="0"
 			>
@@ -160,17 +144,6 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				InlineFunctionExpansion="2"
-				EnableIntrinsicFunctions="true"
-				FavorSizeOrSpeed="1"
-				OmitFramePointers="true"
-				WholeProgramOptimization="false"
-				StringPooling="true"
-				RuntimeLibrary="0"
-				BufferSecurityCheck="false"
-				WarningLevel="3"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"

3rdparty/w32pthreads/PCSX2.txt

@@ -2,14 +2,19 @@
 -- Air's Notes on Using PThreads --
 -----------------------------------
 
-We use pthreads as a static link because it allows for inlined calls of several methods,
+The official name of the DLL is "w32pthreads.dll" -- I use the prefix to indicate with
-most notably the InterlockedExchange implementations.  These would be DLL invocations
+some vaugeness that the pthreads dll is a modified non-standard implementation (see below
-otherwise, and would be roughly twice to three times the total overhead.
+for details); just in case pthreads-win32 dlls in another universe were to be conflicting
+interests.
 
-I created a subfolder for pthreads' publicall exposed header files, since "config.h" was
+I created a subfolder for pthreads' publically exposed header files, since "config.h" was
 rather brutal include file namespace pollution. >_<
 
 Important: pthread_cond is an inherently flawed concept, and is quite incredibly slow
 on top of it.  It's primarily meant for inter-process thread regulation, of which we 
 do none.  Therefore it has been excluded from the library builds.  Don't use it!
 
+I've also disabled rwlocks, since they use pthread_cond internally.  Again, there should
+be little or no reason to have to use those within the context of a high performance app
+like Pcsx2.
+

3rdparty/w32pthreads/include/pthread.h

@@ -1077,6 +1077,7 @@ PTW32_DLLPORT int PTW32_CDECL pthread_condattr_setpshared (pthread_condattr_t *
 /*
  * Condition Variable Functions
  */
+#ifdef PCSX2_ALLOW_COND
 PTW32_DLLPORT int PTW32_CDECL pthread_cond_init (pthread_cond_t * cond,
                                const pthread_condattr_t * attr);
 
@@ -1092,6 +1093,7 @@ PTW32_DLLPORT int PTW32_CDECL pthread_cond_timedwait (pthread_cond_t * cond,
 PTW32_DLLPORT int PTW32_CDECL pthread_cond_signal (pthread_cond_t * cond);
 
 PTW32_DLLPORT int PTW32_CDECL pthread_cond_broadcast (pthread_cond_t * cond);
+#endif
 
 /*
  * Scheduling

3rdparty/w32pthreads/pthreads_2008.vcproj

@@ -2,7 +2,7 @@
 <VisualStudioProject
 	ProjectType="Visual C++"
 	Version="9.00"
-	Name="pthreads"
+	Name="w32pthreads"
 	ProjectGUID="{26511268-2902-4997-8421-ECD7055F9E28}"
 	RootNamespace="pthreads"
 	Keyword="Win32Proj"
@@ -18,8 +18,8 @@
 	<Configurations>
 		<Configuration
 			Name="Debug|Win32"
-			ConfigurationType="4"
+			ConfigurationType="2"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\DefaultProjectRootDir.vsprops;..\3rdpartyDLL.vsprops;..\..\common\vsprops\CodeGen_Debug.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			>
 			<Tool
@@ -39,15 +39,11 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="0"
 				AdditionalIncludeDirectories="&quot;$(ProjectDir)&quot;;&quot;$(ProjectDir)\include&quot;"
-				PreprocessorDefinitions="PTW32_STATIC_LIB;__CLEANUP_SEH;WIN32;_DEBUG;_LIB"
+				PreprocessorDefinitions="PTW32_BUILD;__CLEANUP_SEH;_DEBUG"
 				ExceptionHandling="2"
-				BasicRuntimeChecks="3"
-				RuntimeLibrary="1"
 				UsePrecompiledHeader="0"
 				WarningLevel="3"
-				DebugInformationFormat="4"
 				CompileAs="1"
 			/>
 			<Tool
@@ -60,11 +56,15 @@
 				Name="VCPreLinkEventTool"
 			/>
 			<Tool
-				Name="VCLibrarianTool"
+				Name="VCLinkerTool"
+				AdditionalDependencies="ws2_32.lib"
 			/>
 			<Tool
 				Name="VCALinkTool"
 			/>
+			<Tool
+				Name="VCManifestTool"
+			/>
 			<Tool
 				Name="VCXDCMakeTool"
 			/>
@@ -74,14 +74,18 @@
 			<Tool
 				Name="VCFxCopTool"
 			/>
+			<Tool
+				Name="VCAppVerifierTool"
+			/>
 			<Tool
 				Name="VCPostBuildEventTool"
 			/>
 		</Configuration>
 		<Configuration
 			Name="Release|Win32"
-			ConfigurationType="4"
+			IntermediateDirectory="$(ConfigurationName)"
-			InheritedPropertySheets="..\3rdparty.vsprops"
+			ConfigurationType="2"
+			InheritedPropertySheets="..\DefaultProjectRootDir.vsprops;..\3rdpartyDLL.vsprops;..\..\common\vsprops\CodeGen_Release.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="1"
 			>
@@ -102,23 +106,11 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				InlineFunctionExpansion="2"
-				EnableIntrinsicFunctions="true"
-				FavorSizeOrSpeed="1"
-				EnableFiberSafeOptimizations="true"
 				AdditionalIncludeDirectories="&quot;$(ProjectDir)&quot;;&quot;$(ProjectDir)\include&quot;"
-				PreprocessorDefinitions="PTW32_STATIC_LIB;PTW32_BUILD_INLINED;__CLEANUP_SEH;WIN32;NDEBUG;_LIB"
+				PreprocessorDefinitions="PTW32_BUILD;__CLEANUP_SEH;NDEBUG"
-				StringPooling="true"
 				ExceptionHandling="2"
-				RuntimeLibrary="0"
-				StructMemberAlignment="5"
-				BufferSecurityCheck="false"
-				EnableFunctionLevelLinking="false"
 				UsePrecompiledHeader="0"
-				BrowseInformation="1"
 				WarningLevel="3"
-				DebugInformationFormat="3"
 				CompileAs="1"
 			/>
 			<Tool
@@ -131,11 +123,15 @@
 				Name="VCPreLinkEventTool"
 			/>
 			<Tool
-				Name="VCLibrarianTool"
+				Name="VCLinkerTool"
+				AdditionalDependencies="ws2_32.lib"
 			/>
 			<Tool
 				Name="VCALinkTool"
 			/>
+			<Tool
+				Name="VCManifestTool"
+			/>
 			<Tool
 				Name="VCXDCMakeTool"
 			/>
@@ -145,14 +141,17 @@
 			<Tool
 				Name="VCFxCopTool"
 			/>
+			<Tool
+				Name="VCAppVerifierTool"
+			/>
 			<Tool
 				Name="VCPostBuildEventTool"
 			/>
 		</Configuration>
 		<Configuration
 			Name="Devel|Win32"
-			ConfigurationType="4"
+			ConfigurationType="2"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\DefaultProjectRootDir.vsprops;..\3rdpartyDLL.vsprops;..\..\common\vsprops\CodeGen_Devel.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="0"
 			>
@@ -173,22 +172,11 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				InlineFunctionExpansion="2"
-				EnableIntrinsicFunctions="true"
-				FavorSizeOrSpeed="1"
-				EnableFiberSafeOptimizations="true"
-				WholeProgramOptimization="false"
 				AdditionalIncludeDirectories="&quot;$(ProjectDir)&quot;;&quot;$(ProjectDir)\include&quot;"
-				PreprocessorDefinitions="PTW32_STATIC_LIB;PTW32_BUILD_INLINED;__CLEANUP_SEH;WIN32;NDEBUG;_LIB"
+				PreprocessorDefinitions="PTW32_BUILD;__CLEANUP_SEH;NDEBUG"
-				StringPooling="true"
 				ExceptionHandling="2"
-				RuntimeLibrary="0"
-				StructMemberAlignment="5"
-				BufferSecurityCheck="false"
 				UsePrecompiledHeader="0"
 				WarningLevel="3"
-				DebugInformationFormat="3"
 				CompileAs="1"
 			/>
 			<Tool
@@ -201,11 +189,15 @@
 				Name="VCPreLinkEventTool"
 			/>
 			<Tool
-				Name="VCLibrarianTool"
+				Name="VCLinkerTool"
+				AdditionalDependencies="ws2_32.lib"
 			/>
 			<Tool
 				Name="VCALinkTool"
 			/>
+			<Tool
+				Name="VCManifestTool"
+			/>
 			<Tool
 				Name="VCXDCMakeTool"
 			/>
@@ -215,6 +207,9 @@
 			<Tool
 				Name="VCFxCopTool"
 			/>
+			<Tool
+				Name="VCAppVerifierTool"
+			/>
 			<Tool
 				Name="VCPostBuildEventTool"
 			/>
@@ -439,6 +434,10 @@
 					/>
 				</FileConfiguration>
 			</File>
+			<File
+				RelativePath=".\dll.c"
+				>
+			</File>
 			<File
 				RelativePath=".\errno.c"
 				>
@@ -754,6 +753,7 @@
 				>
 				<FileConfiguration
 					Name="Debug|Win32"
+					ExcludedFromBuild="true"
 					>
 					<Tool
 						Name="VCCLCompilerTool"
@@ -764,6 +764,7 @@
 				</FileConfiguration>
 				<FileConfiguration
 					Name="Release|Win32"
+					ExcludedFromBuild="true"
 					>
 					<Tool
 						Name="VCCLCompilerTool"
@@ -774,6 +775,7 @@
 				</FileConfiguration>
 				<FileConfiguration
 					Name="Devel|Win32"
+					ExcludedFromBuild="true"
 					>
 					<Tool
 						Name="VCCLCompilerTool"

3rdparty/wxWidgets/build/msw/Common.vsprops

@@ -0,0 +1,12 @@
+<?xml version="1.0" encoding="Windows-1252"?>
+<VisualStudioPropertySheet
+	ProjectType="Visual C++"
+	Version="8.00"
+	Name="wxCommon"
+	IntermediateDirectory="$(PlatformName)\$(ConfigurationName)\$(ProjectName)"
+	>
+	<Tool
+		Name="VCCLCompilerTool"
+		PreprocessorDefinitions="WIN32;__WXMSW__;_LIB;wxUSE_UNICODE=1"
+	/>
+</VisualStudioPropertySheet>

3rdparty/wxWidgets/build/msw/ProjectRootDir.vsprops

@@ -0,0 +1,19 @@
+<?xml version="1.0" encoding="Windows-1252"?>
+<VisualStudioPropertySheet
+	ProjectType="Visual C++"
+	Version="8.00"
+	Name="ProjectRootDir"
+	>
+	<UserMacro
+		Name="ProjectRootDir"
+		Value="$(ProjectDir)\..\.."
+	/>
+	<UserMacro
+		Name="SvnRootDir"
+		Value="$(ProjectRootDir)\..\.."
+	/>
+	<UserMacro
+		Name="SvnCommonDir"
+		Value="$(SvnRootDir)\common"
+	/>
+</VisualStudioPropertySheet>

3rdparty/wxWidgets/build/msw/wx_base_2008.vcproj

@@ -20,10 +20,8 @@
 	<Configurations>
 		<Configuration
 			Name="Release|Win32"
-			OutputDirectory="..\..\..\..\deps\$(PlatformName)\$(ConfigurationName)"
-			IntermediateDirectory="$(ProjectDir)\$(PlatformName)\$(ConfigurationName)\$(TargetName)"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\..\..\3rdparty.vsprops"
+			InheritedPropertySheets=".\ProjectRootDir.vsprops;..\..\..\3rdparty.vsprops;.\Common.vsprops;..\..\..\3rdpartyDLL.vsprops;..\..\..\..\common\vsprops\CodeGen_Release.vsprops"
 			UseOfMFC="0"
 			ATLMinimizesCRunTimeLibraryUsage="false"
 			CharacterSet="1"
@@ -47,25 +45,12 @@
 			<Tool
 				Name="VCCLCompilerTool"
 				AdditionalOptions="/EHsc "
-				Optimization="3"
-				InlineFunctionExpansion="2"
-				FavorSizeOrSpeed="2"
-				OmitFramePointers="true"
-				WholeProgramOptimization="true"
 				AdditionalIncludeDirectories="..\..\Include;..\..\Include\msvc;..\..\..\zlib"
-				PreprocessorDefinitions="WIN32;_LIB;__WXMSW__;wxUSE_UNICODE=1;wxUSE_GUI=0;wxUSE_BASE=1;_CRT_SECURE_NO_WARNINGS"
+				PreprocessorDefinitions="wxUSE_GUI=0;wxUSE_BASE=1;NDEBUG"
-				StringPooling="true"
-				RuntimeLibrary="0"
-				BufferSecurityCheck="false"
-				EnableFunctionLevelLinking="false"
 				UsePrecompiledHeader="2"
 				PrecompiledHeaderThrough="wx/wxprec.h"
 				PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch"
-				AssemblerListingLocation="$(IntDir)\"
-				ObjectFile="$(IntDir)\"
-				ProgramDataBaseFileName="$(IntDir)\vc80.pdb"
 				WarningLevel="4"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -79,8 +64,6 @@
 			/>
 			<Tool
 				Name="VCLibrarianTool"
-				OutputFile="$(OutDir)/wxBase28.lib"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCALinkTool"
@@ -102,10 +85,8 @@
 		</Configuration>
 		<Configuration
 			Name="Debug|Win32"
-			OutputDirectory="..\..\..\..\deps\$(PlatformName)\$(ConfigurationName)"
-			IntermediateDirectory="$(ProjectDir)\$(PlatformName)\$(ConfigurationName)\$(TargetName)"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\..\..\3rdparty.vsprops"
+			InheritedPropertySheets=".\ProjectRootDir.vsprops;..\..\..\3rdparty.vsprops;.\Common.vsprops;..\..\..\..\common\vsprops\CodeGen_Debug.vsprops;..\..\..\..\common\vsprops\IncrementalLinking.vsprops"
 			UseOfMFC="0"
 			ATLMinimizesCRunTimeLibraryUsage="false"
 			CharacterSet="1"
@@ -128,18 +109,12 @@
 			<Tool
 				Name="VCCLCompilerTool"
 				AdditionalOptions="/EHsc "
-				Optimization="0"
 				AdditionalIncludeDirectories="..\..\Include;..\..\Include\msvc;..\..\..\zlib"
-				PreprocessorDefinitions="WIN32;_LIB;_DEBUG;__WXMSW__;__WXDEBUG__;wxUSE_GUI=0;wxUSE_BASE=1;wxUSE_UNICODE=1;_CRT_SECURE_NO_WARNINGS"
+				PreprocessorDefinitions="_DEBUG;__WXDEBUG__;wxUSE_BASE=1;wxUSE_GUI=0"
-				MinimalRebuild="true"
-				BasicRuntimeChecks="3"
-				RuntimeLibrary="1"
 				UsePrecompiledHeader="2"
 				PrecompiledHeaderThrough="wx/wxprec.h"
 				PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch"
 				WarningLevel="4"
-				SuppressStartupBanner="true"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -153,7 +128,6 @@
 			/>
 			<Tool
 				Name="VCLibrarianTool"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCALinkTool"
@@ -175,10 +149,8 @@
 		</Configuration>
 		<Configuration
 			Name="Devel|Win32"
-			OutputDirectory="..\..\..\..\deps\$(PlatformName)\$(ConfigurationName)"
-			IntermediateDirectory="$(ProjectDir)\$(PlatformName)\$(ConfigurationName)\$(TargetName)"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\..\..\3rdparty.vsprops"
+			InheritedPropertySheets=".\ProjectRootDir.vsprops;..\..\..\3rdparty.vsprops;.\Common.vsprops;..\..\..\..\common\vsprops\CodeGen_Devel.vsprops;..\..\..\..\common\vsprops\IncrementalLinking.vsprops"
 			UseOfMFC="0"
 			ATLMinimizesCRunTimeLibraryUsage="false"
 			CharacterSet="1"
@@ -207,16 +179,10 @@
 				FavorSizeOrSpeed="2"
 				WholeProgramOptimization="false"
 				AdditionalIncludeDirectories="..\..\Include;..\..\Include\msvc;..\..\..\zlib"
-				PreprocessorDefinitions="WIN32;_LIB;__WXMSW__;wxUSE_UNICODE=1;wxUSE_GUI=0;wxUSE_BASE=1;_CRT_SECURE_NO_WARNINGS"
+				PreprocessorDefinitions="wxUSE_GUI=0;wxUSE_BASE=1;NDEBUG"
-				StringPooling="true"
-				RuntimeLibrary="0"
-				EnableFunctionLevelLinking="true"
 				UsePrecompiledHeader="2"
 				PrecompiledHeaderThrough="wx/wxprec.h"
 				PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch"
-				AssemblerListingLocation="$(IntDir)\"
-				ObjectFile="$(IntDir)\"
-				ProgramDataBaseFileName="$(IntDir)\vc80.pdb"
 				WarningLevel="4"
 				SuppressStartupBanner="true"
 			/>
@@ -232,8 +198,6 @@
 			/>
 			<Tool
 				Name="VCLibrarianTool"
-				OutputFile="$(OutDir)/wxBase28.lib"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCALinkTool"

3rdparty/wxWidgets/build/msw/wx_core_2008.vcproj

@@ -20,10 +20,8 @@
 	<Configurations>
 		<Configuration
 			Name="Debug|Win32"
-			OutputDirectory="..\..\..\..\deps\$(PlatformName)\$(ConfigurationName)"
-			IntermediateDirectory="$(ProjectDir)\$(PlatformName)\$(ConfigurationName)\$(TargetName)"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\..\..\3rdparty.vsprops"
+			InheritedPropertySheets=".\ProjectRootDir.vsprops;..\..\..\3rdparty.vsprops;.\Common.vsprops;..\..\..\..\common\vsprops\CodeGen_Debug.vsprops;..\..\..\..\common\vsprops\IncrementalLinking.vsprops"
 			UseOfMFC="0"
 			ATLMinimizesCRunTimeLibraryUsage="false"
 			CharacterSet="1"
@@ -46,18 +44,12 @@
 			<Tool
 				Name="VCCLCompilerTool"
 				AdditionalOptions="/EHsc "
-				Optimization="0"
+				AdditionalIncludeDirectories="&quot;$(ProjectRootDir)\include&quot;;&quot;$(ProjectRootDir)\include\msvc&quot;;&quot;$(SvnRootDir)\3rdparty\zlib&quot;;&quot;$(ProjectRootDir)\src\png&quot;"
-				AdditionalIncludeDirectories="..\..\Include;..\..\Include\msvc;..\..\..\zlib;..\..\src\png"
+				PreprocessorDefinitions="_DEBUG;__WXDEBUG__;wxUSE_BASE=0;wxUSE_GUI=1"
-				PreprocessorDefinitions="WIN32;_LIB;_DEBUG;__WXMSW__;__WXDEBUG__;wxUSE_BASE=0;wxUSE_UNICODE=1;_CRT_SECURE_NO_WARNINGS"
-				MinimalRebuild="true"
-				BasicRuntimeChecks="3"
-				RuntimeLibrary="1"
 				UsePrecompiledHeader="2"
 				PrecompiledHeaderThrough="wx/wxprec.h"
 				PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch"
 				WarningLevel="4"
-				SuppressStartupBanner="true"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -71,7 +63,6 @@
 			/>
 			<Tool
 				Name="VCLibrarianTool"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCALinkTool"
@@ -93,10 +84,8 @@
 		</Configuration>
 		<Configuration
 			Name="Release|Win32"
-			OutputDirectory="..\..\..\..\deps\$(PlatformName)\$(ConfigurationName)"
-			IntermediateDirectory="$(ProjectDir)\$(PlatformName)\$(ConfigurationName)\$(TargetName)"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\..\..\3rdparty.vsprops"
+			InheritedPropertySheets=".\ProjectRootDir.vsprops;..\..\..\3rdparty.vsprops;.\Common.vsprops;..\..\..\..\common\vsprops\CodeGen_Release.vsprops"
 			UseOfMFC="0"
 			ATLMinimizesCRunTimeLibraryUsage="false"
 			CharacterSet="1"
@@ -120,25 +109,12 @@
 			<Tool
 				Name="VCCLCompilerTool"
 				AdditionalOptions="/EHsc "
-				Optimization="3"
-				InlineFunctionExpansion="2"
-				FavorSizeOrSpeed="2"
-				OmitFramePointers="true"
-				WholeProgramOptimization="true"
 				AdditionalIncludeDirectories="..\..\Include;..\..\Include\msvc;..\..\..\zlib;..\..\src\png"
-				PreprocessorDefinitions="WIN32;_LIB;__WXMSW__;wxUSE_BASE=0;wxUSE_UNICODE=1;_CRT_SECURE_NO_WARNINGS"
+				PreprocessorDefinitions="wxUSE_BASE=0;wxUSE_GUI=1"
-				StringPooling="true"
-				RuntimeLibrary="0"
-				BufferSecurityCheck="false"
-				EnableFunctionLevelLinking="false"
 				UsePrecompiledHeader="2"
 				PrecompiledHeaderThrough="wx/wxprec.h"
 				PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch"
-				AssemblerListingLocation="$(IntDir)\"
-				ObjectFile="$(IntDir)\"
-				ProgramDataBaseFileName="$(IntDir)\vc80.pdb"
 				WarningLevel="4"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -152,7 +128,6 @@
 			/>
 			<Tool
 				Name="VCLibrarianTool"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCALinkTool"
@@ -174,10 +149,8 @@
 		</Configuration>
 		<Configuration
 			Name="Devel|Win32"
-			OutputDirectory="..\..\..\..\deps\$(PlatformName)\$(ConfigurationName)"
-			IntermediateDirectory="$(ProjectDir)\$(PlatformName)\$(ConfigurationName)\$(TargetName)"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\..\..\3rdparty.vsprops"
+			InheritedPropertySheets=".\ProjectRootDir.vsprops;..\..\..\3rdparty.vsprops;.\Common.vsprops;..\..\..\..\common\vsprops\CodeGen_Devel.vsprops;..\..\..\..\common\vsprops\IncrementalLinking.vsprops"
 			UseOfMFC="0"
 			ATLMinimizesCRunTimeLibraryUsage="false"
 			CharacterSet="1"
@@ -201,23 +174,12 @@
 			<Tool
 				Name="VCCLCompilerTool"
 				AdditionalOptions="/EHsc "
-				Optimization="3"
-				InlineFunctionExpansion="1"
-				FavorSizeOrSpeed="2"
-				WholeProgramOptimization="false"
 				AdditionalIncludeDirectories="..\..\Include;..\..\Include\msvc;..\..\..\zlib;..\..\src\png"
-				PreprocessorDefinitions="WIN32;_LIB;__WXMSW__;wxUSE_BASE=0;wxUSE_UNICODE=1;_CRT_SECURE_NO_WARNINGS"
+				PreprocessorDefinitions="wxUSE_BASE=0;wxUSE_GUI=1"
-				StringPooling="true"
-				RuntimeLibrary="0"
-				EnableFunctionLevelLinking="true"
 				UsePrecompiledHeader="2"
 				PrecompiledHeaderThrough="wx/wxprec.h"
 				PrecompiledHeaderFile="$(IntDir)\$(TargetName).pch"
-				AssemblerListingLocation="$(IntDir)\"
-				ObjectFile="$(IntDir)\"
-				ProgramDataBaseFileName="$(IntDir)\vc80.pdb"
 				WarningLevel="4"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -231,7 +193,6 @@
 			/>
 			<Tool
 				Name="VCLibrarianTool"
-				SuppressStartupBanner="true"
 			/>
 			<Tool
 				Name="VCALinkTool"

3rdparty/zlib/zlib.vcproj

@@ -18,7 +18,7 @@
 		<Configuration
 			Name="Debug|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Debug.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			>
 			<Tool
@@ -38,11 +38,7 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="0"
-				BasicRuntimeChecks="3"
-				RuntimeLibrary="1"
 				WarningLevel="3"
-				DebugInformationFormat="4"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -75,7 +71,7 @@
 		<Configuration
 			Name="Release|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Release.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="1"
 			>
@@ -96,15 +92,7 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				InlineFunctionExpansion="2"
-				EnableIntrinsicFunctions="true"
-				FavorSizeOrSpeed="1"
-				OmitFramePointers="true"
-				StringPooling="true"
-				RuntimeLibrary="0"
 				WarningLevel="3"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"
@@ -137,7 +125,7 @@
 		<Configuration
 			Name="Devel|Win32"
 			ConfigurationType="4"
-			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
+			InheritedPropertySheets="..\3rdparty.vsprops;..\..\common\vsprops\CodeGen_Devel.vsprops;..\..\common\vsprops\IncrementalLinking.vsprops"
 			CharacterSet="2"
 			WholeProgramOptimization="0"
 			>
@@ -158,17 +146,7 @@
 			/>
 			<Tool
 				Name="VCCLCompilerTool"
-				Optimization="2"
-				InlineFunctionExpansion="2"
-				EnableIntrinsicFunctions="true"
-				FavorSizeOrSpeed="1"
-				OmitFramePointers="true"
-				WholeProgramOptimization="false"
-				StringPooling="true"
-				RuntimeLibrary="0"
-				BufferSecurityCheck="false"
 				WarningLevel="3"
-				DebugInformationFormat="3"
 			/>
 			<Tool
 				Name="VCManagedResourceCompilerTool"

build.rb

@@ -0,0 +1,174 @@
+#!/usr/bin/env ruby
+
+# Usage: ruby build.rb [option] [pcsx2, plugins, <plugin name>] [dev,debug,release] [all, install, clean]
+# If you don't specify pcsx2, plugins, or a plugin name, it will assume you want to rebuild everything.
+# If you don't specify dev or debug, it assumes a release build.
+# If it isn't all, install, or clean, it assumes install.
+
+# If you want other options, add them to $pcsx2_build_types. This is still a work in progress...
+# --arcum42
+
+require "fileutils.rb"
+include FileUtils
+
+$main_dir = Dir.pwd
+$pcsx2_install_dir =  "#{$main_dir}/bin"
+$plugin_install_dir =  "#{$main_dir}/bin/plugins"
+
+$pcsx2_dir = "#{$main_dir}/pcsx2"
+$plugins_dir = "#{$main_dir}/plugins"
+
+$pcsx2_prefix = " --prefix #{$main_dir}"
+$plugins_prefix = " --prefix #{$plugin_install_dir}"
+
+$plugin_list=["CDVDnull",  "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP",  "CDVDlinuz"]
+$full_plugin_list=["CDVDnull",  "dev9null", "FWnull", "USBnull", "SPU2null", "zerogs", "zzogl", "zeropad", "zerospu2", "PeopsSPU2", "CDVDiso", "CDVDisoEFP",  "CDVDlinuz","GSnull","PadNull","wxpad"]
+
+$pcsx2_build_types = {
+	"dev" => " --enable-devbuild ",
+	"debug" => " --enable-debug ",
+	"release" => " "
+	}
+	
+$pcsx2_release_params=["dev","debug","release"]
+$make_params=["all", "clean","install"]
+
+$build_report =""
+$build_counter = 0
+
+def plugin_src_dir(plugin_name)
+	name = "#{$plugins_dir}/#{plugin_name}/"
+	case plugin_name
+		when "CDVDiso" then
+			name +=  "src"
+		when "CDVDisoEFP" then 
+			name +=  "src/Linux"
+		when "CDVDlinuz"
+			name +=  "Src/Linux"
+		when "zerogs", "zzogl"
+			name +=  "opengl"
+	end
+	
+	return name
+end
+
+def announce(my_program)
+	print "---------------\n"
+	print "Building #{my_program}\n"
+	print "---------------\n"
+end
+
+def make(options)
+	system("make #{options}")
+	($? == 0)
+end
+
+def rebuild(options)
+	system("aclocal")
+	system("automake")
+	system("autoconf")
+	system("chmod +x configure")
+	system("./configure #{options}")
+	make "clean"
+end
+
+def install(build_name)
+	ret = make "install"
+	
+	case build_name
+		# If the package isn't inclined to obey simple instructions...
+		when "CDVDisoEFP" then 
+			system("cp #{plugin_src_dir(build_name)}/cfgCDVDisoEFP #{$plugin_install_dir}")
+			system("cp #{plugin_src_dir(build_name)}/libCDVDisoEFP.so #{$plugin_install_dir}")
+		when "CDVDlinuz" then 
+			system("cp #{plugin_src_dir(build_name)}/cfgCDVDlinuz #{$plugin_install_dir}")
+			system("cp #{plugin_src_dir(build_name)}/libCDVDlinuz.so #{$plugin_install_dir}")
+		when "PeopsSPU2" then 
+			system("cp #{plugin_src_dir(build_name)}/libspu2Peops*.so* #{$plugin_install_dir}")
+			
+		# Copy the shaders over. Shouldn't the makefile do this?
+		when "zzogl","zerogs" then
+			system("cp #{plugin_src_dir(build_name)}/Win32/ps2hw.dat #{$plugin_install_dir}")
+			
+		#And while we have the opportunity...
+		when "pcsx2" then 
+		svn_revision = `svn info | grep Revision:`
+		svn_revision = /[0-9]+/.match(svn_revision)
+		system("cp #{$pcsx2_install_dir}/pcsx2 #{$pcsx2_install_dir}/pcsx2-#{svn_revision}")
+	end
+	
+	ret
+end
+
+def build(build_name, make_parameter)
+	completed = true
+	
+	announce "#{build_name.capitalize}"
+	
+	if build_name != "pcsx2" then
+		build_dir = plugin_src_dir(build_name)
+	else
+		build_dir = "#{$pcsx2_dir}"
+	end
+	
+	Dir.chdir build_dir
+	
+	case make_parameter
+		when "all" then
+			if build_name == "pcsx2"
+				rebuild($pcsx2_prefix)
+			else
+				rebuild($plugins_prefix)
+			end
+			completed = install(build_name)
+			
+		when "clean" then
+			make "clean"
+			
+		else
+			completed = install(build_name)
+		end
+		
+	Dir.chdir $main_dir
+	
+	if completed then
+		$build_report += "#{build_name} was built successfully.\n"
+		$build_counter += 1
+	else
+		$build_report += "#{build_name} was not built successfully.\n"
+	end
+			
+end
+
+build_parameter = "all"
+make_parameter = ""
+build_items = Array.new([])
+
+ARGV.each do |x|
+	make_parameter = x if $make_params.include?(x)
+	
+	build_items.push(x) if $full_plugin_list.include?(x) or (x == "pcsx2")
+	$pcsx2_prefix = $pcsx2_build_types[x] + $pcsx2_prefix if $pcsx2_release_params.include?(x)
+	
+	if (x == "plugins") then
+		x = $plugin_list 
+		build_items.push(x)
+	end
+end
+
+if build_items.empty? then
+	build_items.push($plugin_list)
+	build_items.push("pcsx2")
+end
+
+build_items.flatten!
+
+build_items.each do |x|
+	build(x,make_parameter)
+end
+
+print "\n--\n"
+print "Build Summary:\n"
+print $build_report 
+print "\n"
+print "#{$build_counter}/#{build_items.count} Successful.\n"

build.sh

@@ -11,8 +11,14 @@
 #Optimized, but a devbuild
 export PCSX2OPTIONS="--enable-devbuild --prefix `pwd`"
 
-#Debug / Devbuild version
+#Debug version (which now implies a devbuild)
-#export PCSX2OPTIONS="--enable-debug --enable-devbuild --prefix `pwd`"
+#export PCSX2OPTIONS="--enable-debug --prefix `pwd`"
+
+#Normal, but unoptimized
+#export PCSX2OPTIONS="--disable-optimization --prefix `pwd`"
+
+#Normal, with warnings
+#export PCSX2OPTIONS="--enable-warnings --prefix `pwd`"
 
 #ZeroGS Normal mode
 export ZEROGSOPTIONS="--enable-sse2"
@@ -23,12 +29,16 @@ export ZEROGSOPTIONS="--enable-sse2"
 #ZeroSPU2 Debug mode (Don't enable right now)
 #export ZEROSPU2OPTIONS="--enable-debug --enable-devbuild"
 
+#GSnull debug options.
+#export GSnullOPTIONS="--enable-debug"
+
 option=$@
 export PCSX2PLUGINS="`pwd`/bin/plugins"
 curdir=`pwd`
 
+echo
 echo "Building the Pcsx2 Suite."
-echo "Note: will not compile on Linux x64."
+echo "Note: binaries generated are 32 bit, and require 32 bit versions of all dependencies."
 cd ${curdir}/plugins
 sh build.sh $option
 
@@ -38,8 +48,9 @@ echo Error with building plugins
 exit 1
 fi
 
+echo
 echo "Building Pcsx2."
-echo "Note: will not compile on Linux x64."
+echo "Note: binaries generated are 32 bit, and require 32 bit versions of all dependencies."
 cd ${curdir}/pcsx2
 sh build.sh $option
 

common/build/Utilities/utilities.vcproj

@@ -0,0 +1,421 @@
+<?xml version="1.0" encoding="Windows-1252"?>
+<VisualStudioProject
+	ProjectType="Visual C++"
+	Version="9.00"
+	Name="Utilities"
+	ProjectGUID="{4639972E-424E-4E13-8B07-CA403C481346}"
+	RootNamespace="x86emitter"
+	Keyword="Win32Proj"
+	TargetFrameworkVersion="196613"
+	>
+	<Platforms>
+		<Platform
+			Name="Win32"
+		/>
+	</Platforms>
+	<ToolFiles>
+	</ToolFiles>
+	<Configurations>
+		<Configuration
+			Name="Debug|Win32"
+			ConfigurationType="4"
+			InheritedPropertySheets="..\..\vsprops\CommonLibrary.vsprops;..\..\vsprops\CodeGen_Debug.vsprops;..\..\vsprops\IncrementalLinking.vsprops;..\..\vsprops\pthreads.vsprops"
+			CharacterSet="1"
+			>
+			<Tool
+				Name="VCPreBuildEventTool"
+			/>
+			<Tool
+				Name="VCCustomBuildTool"
+			/>
+			<Tool
+				Name="VCXMLDataGeneratorTool"
+			/>
+			<Tool
+				Name="VCWebServiceProxyGeneratorTool"
+			/>
+			<Tool
+				Name="VCMIDLTool"
+			/>
+			<Tool
+				Name="VCCLCompilerTool"
+				Optimization="0"
+				PreprocessorDefinitions="WIN32;_DEBUG;_LIB"
+				ExceptionHandling="2"
+			/>
+			<Tool
+				Name="VCManagedResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCPreLinkEventTool"
+			/>
+			<Tool
+				Name="VCLibrarianTool"
+			/>
+			<Tool
+				Name="VCALinkTool"
+			/>
+			<Tool
+				Name="VCXDCMakeTool"
+			/>
+			<Tool
+				Name="VCBscMakeTool"
+			/>
+			<Tool
+				Name="VCFxCopTool"
+			/>
+			<Tool
+				Name="VCPostBuildEventTool"
+			/>
+		</Configuration>
+		<Configuration
+			Name="Release|Win32"
+			ConfigurationType="4"
+			InheritedPropertySheets="..\..\vsprops\CommonLibrary.vsprops;..\..\vsprops\CodeGen_Release.vsprops;..\..\vsprops\pthreads.vsprops"
+			CharacterSet="1"
+			WholeProgramOptimization="1"
+			>
+			<Tool
+				Name="VCPreBuildEventTool"
+			/>
+			<Tool
+				Name="VCCustomBuildTool"
+			/>
+			<Tool
+				Name="VCXMLDataGeneratorTool"
+			/>
+			<Tool
+				Name="VCWebServiceProxyGeneratorTool"
+			/>
+			<Tool
+				Name="VCMIDLTool"
+			/>
+			<Tool
+				Name="VCCLCompilerTool"
+				Optimization="2"
+				EnableIntrinsicFunctions="true"
+				PreprocessorDefinitions="WIN32;NDEBUG;_LIB"
+				ExceptionHandling="2"
+			/>
+			<Tool
+				Name="VCManagedResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCPreLinkEventTool"
+			/>
+			<Tool
+				Name="VCLibrarianTool"
+			/>
+			<Tool
+				Name="VCALinkTool"
+			/>
+			<Tool
+				Name="VCXDCMakeTool"
+			/>
+			<Tool
+				Name="VCBscMakeTool"
+			/>
+			<Tool
+				Name="VCFxCopTool"
+			/>
+			<Tool
+				Name="VCPostBuildEventTool"
+			/>
+		</Configuration>
+		<Configuration
+			Name="Devel|Win32"
+			ConfigurationType="4"
+			InheritedPropertySheets="..\..\vsprops\CommonLibrary.vsprops;..\..\vsprops\IncrementalLinking.vsprops;..\..\vsprops\CodeGen_Devel.vsprops;..\..\vsprops\pthreads.vsprops"
+			CharacterSet="1"
+			WholeProgramOptimization="1"
+			>
+			<Tool
+				Name="VCPreBuildEventTool"
+			/>
+			<Tool
+				Name="VCCustomBuildTool"
+			/>
+			<Tool
+				Name="VCXMLDataGeneratorTool"
+			/>
+			<Tool
+				Name="VCWebServiceProxyGeneratorTool"
+			/>
+			<Tool
+				Name="VCMIDLTool"
+			/>
+			<Tool
+				Name="VCCLCompilerTool"
+				Optimization="2"
+				EnableIntrinsicFunctions="true"
+				PreprocessorDefinitions="WIN32;NDEBUG;_LIB"
+				ExceptionHandling="2"
+			/>
+			<Tool
+				Name="VCManagedResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCPreLinkEventTool"
+			/>
+			<Tool
+				Name="VCLibrarianTool"
+			/>
+			<Tool
+				Name="VCALinkTool"
+			/>
+			<Tool
+				Name="VCXDCMakeTool"
+			/>
+			<Tool
+				Name="VCBscMakeTool"
+			/>
+			<Tool
+				Name="VCFxCopTool"
+			/>
+			<Tool
+				Name="VCPostBuildEventTool"
+			/>
+		</Configuration>
+	</Configurations>
+	<References>
+	</References>
+	<Files>
+		<Filter
+			Name="Source Files"
+			Filter="cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx"
+			UniqueIdentifier="{4FC737F1-C7A5-4376-A066-2A32D752A2FF}"
+			>
+			<File
+				RelativePath="..\..\src\Utilities\AlignedMalloc.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\Console.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\Exceptions.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\x86\MemcpyFast.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\StringHelpers.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\ThreadTools.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\vssprintf.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\Windows\WinHostSys.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\WinMisc.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\WinThreads.cpp"
+				>
+			</File>
+			<Filter
+				Name="Linux"
+				>
+				<File
+					RelativePath="..\..\src\Utilities\Linux\LnxHostSys.cpp"
+					>
+					<FileConfiguration
+						Name="Debug|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Release|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Devel|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+				</File>
+				<File
+					RelativePath="..\..\src\Utilities\LnxMisc.cpp"
+					>
+					<FileConfiguration
+						Name="Debug|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Release|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Devel|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+				</File>
+				<File
+					RelativePath="..\..\src\Utilities\LnxThreads.cpp"
+					>
+					<FileConfiguration
+						Name="Debug|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Release|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Devel|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCLCompilerTool"
+						/>
+					</FileConfiguration>
+				</File>
+				<File
+					RelativePath="..\..\src\Utilities\x86\MemcpyFast.S"
+					>
+					<FileConfiguration
+						Name="Debug|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCustomBuildTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Release|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCustomBuildTool"
+						/>
+					</FileConfiguration>
+					<FileConfiguration
+						Name="Devel|Win32"
+						ExcludedFromBuild="true"
+						>
+						<Tool
+							Name="VCCustomBuildTool"
+						/>
+					</FileConfiguration>
+				</File>
+			</Filter>
+		</Filter>
+		<Filter
+			Name="Header Files"
+			Filter="h;hpp;hxx;hm;inl;inc;xsd"
+			UniqueIdentifier="{93995380-89BD-4b04-88EB-625FBE52EBFB}"
+			>
+			<File
+				RelativePath="..\..\include\Utilities\Console.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\Dependencies.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\Exceptions.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\General.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\intrin_x86.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\lnx_memzero.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\MemcpyFast.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\Utilities\PrecompiledHeader.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\RedtapeWindows.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\SafeArray.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\StringHelpers.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\Threading.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\Utilities\win_memzero.h"
+				>
+			</File>
+		</Filter>
+		<Filter
+			Name="Resource Files"
+			Filter="rc;ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe;resx;tiff;tif;png;wav"
+			UniqueIdentifier="{67DA6AB6-F800-4c08-8B7A-83BB121AAD01}"
+			>
+		</Filter>
+	</Files>
+	<Globals>
+	</Globals>
+</VisualStudioProject>

common/build/x86emitter/x86emitter.vcproj

@@ -0,0 +1,366 @@
+<?xml version="1.0" encoding="Windows-1252"?>
+<VisualStudioProject
+	ProjectType="Visual C++"
+	Version="9.00"
+	Name="x86emitter"
+	ProjectGUID="{A51123F5-9505-4EAE-85E7-D320290A272C}"
+	RootNamespace="x86emitter"
+	Keyword="Win32Proj"
+	TargetFrameworkVersion="196613"
+	>
+	<Platforms>
+		<Platform
+			Name="Win32"
+		/>
+	</Platforms>
+	<ToolFiles>
+	</ToolFiles>
+	<Configurations>
+		<Configuration
+			Name="Debug|Win32"
+			ConfigurationType="4"
+			InheritedPropertySheets="..\..\vsprops\CommonLibrary.vsprops;..\..\vsprops\CodeGen_Debug.vsprops;..\..\vsprops\IncrementalLinking.vsprops;..\..\vsprops\pthreads.vsprops"
+			CharacterSet="1"
+			>
+			<Tool
+				Name="VCPreBuildEventTool"
+			/>
+			<Tool
+				Name="VCCustomBuildTool"
+			/>
+			<Tool
+				Name="VCXMLDataGeneratorTool"
+			/>
+			<Tool
+				Name="VCWebServiceProxyGeneratorTool"
+			/>
+			<Tool
+				Name="VCMIDLTool"
+			/>
+			<Tool
+				Name="VCCLCompilerTool"
+				Optimization="0"
+				PreprocessorDefinitions="WIN32;_DEBUG;_LIB"
+			/>
+			<Tool
+				Name="VCManagedResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCPreLinkEventTool"
+			/>
+			<Tool
+				Name="VCLibrarianTool"
+			/>
+			<Tool
+				Name="VCALinkTool"
+			/>
+			<Tool
+				Name="VCXDCMakeTool"
+			/>
+			<Tool
+				Name="VCBscMakeTool"
+			/>
+			<Tool
+				Name="VCFxCopTool"
+			/>
+			<Tool
+				Name="VCPostBuildEventTool"
+			/>
+		</Configuration>
+		<Configuration
+			Name="Release|Win32"
+			ConfigurationType="4"
+			InheritedPropertySheets="..\..\vsprops\CommonLibrary.vsprops;..\..\vsprops\CodeGen_Release.vsprops;..\..\vsprops\pthreads.vsprops"
+			CharacterSet="1"
+			WholeProgramOptimization="1"
+			>
+			<Tool
+				Name="VCPreBuildEventTool"
+			/>
+			<Tool
+				Name="VCCustomBuildTool"
+			/>
+			<Tool
+				Name="VCXMLDataGeneratorTool"
+			/>
+			<Tool
+				Name="VCWebServiceProxyGeneratorTool"
+			/>
+			<Tool
+				Name="VCMIDLTool"
+			/>
+			<Tool
+				Name="VCCLCompilerTool"
+				Optimization="2"
+				EnableIntrinsicFunctions="true"
+				PreprocessorDefinitions="WIN32;NDEBUG;_LIB"
+			/>
+			<Tool
+				Name="VCManagedResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCPreLinkEventTool"
+			/>
+			<Tool
+				Name="VCLibrarianTool"
+			/>
+			<Tool
+				Name="VCALinkTool"
+			/>
+			<Tool
+				Name="VCXDCMakeTool"
+			/>
+			<Tool
+				Name="VCBscMakeTool"
+			/>
+			<Tool
+				Name="VCFxCopTool"
+			/>
+			<Tool
+				Name="VCPostBuildEventTool"
+			/>
+		</Configuration>
+		<Configuration
+			Name="Devel|Win32"
+			ConfigurationType="4"
+			InheritedPropertySheets="..\..\vsprops\CommonLibrary.vsprops;..\..\vsprops\IncrementalLinking.vsprops;..\..\vsprops\CodeGen_Devel.vsprops;..\..\vsprops\pthreads.vsprops"
+			CharacterSet="1"
+			WholeProgramOptimization="1"
+			>
+			<Tool
+				Name="VCPreBuildEventTool"
+			/>
+			<Tool
+				Name="VCCustomBuildTool"
+			/>
+			<Tool
+				Name="VCXMLDataGeneratorTool"
+			/>
+			<Tool
+				Name="VCWebServiceProxyGeneratorTool"
+			/>
+			<Tool
+				Name="VCMIDLTool"
+			/>
+			<Tool
+				Name="VCCLCompilerTool"
+				Optimization="2"
+				EnableIntrinsicFunctions="true"
+				PreprocessorDefinitions="WIN32;NDEBUG;_LIB"
+			/>
+			<Tool
+				Name="VCManagedResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCResourceCompilerTool"
+			/>
+			<Tool
+				Name="VCPreLinkEventTool"
+			/>
+			<Tool
+				Name="VCLibrarianTool"
+			/>
+			<Tool
+				Name="VCALinkTool"
+			/>
+			<Tool
+				Name="VCXDCMakeTool"
+			/>
+			<Tool
+				Name="VCBscMakeTool"
+			/>
+			<Tool
+				Name="VCFxCopTool"
+			/>
+			<Tool
+				Name="VCPostBuildEventTool"
+			/>
+		</Configuration>
+	</Configurations>
+	<References>
+	</References>
+	<Files>
+		<Filter
+			Name="Source Files"
+			Filter="cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx"
+			UniqueIdentifier="{4FC737F1-C7A5-4376-A066-2A32D752A2FF}"
+			>
+			<File
+				RelativePath="..\..\src\x86emitter\3dnow.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\cpudetect.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\fpu.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\jmp.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\legacy.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\legacy_sse.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\PrecompiledHeader.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\simd.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\tools.cpp"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\x86emitter.cpp"
+				>
+			</File>
+		</Filter>
+		<Filter
+			Name="Header Files"
+			Filter="h;hpp;hxx;hm;inl;inc;xsd"
+			UniqueIdentifier="{93995380-89BD-4b04-88EB-625FBE52EBFB}"
+			>
+			<File
+				RelativePath="..\..\include\x86emitter\inlines.inl"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\instructions.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\internal.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\legacy_instructions.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\legacy_internal.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\legacy_types.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\macros.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\src\x86emitter\PrecompiledHeader.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\sse_helpers.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\tools.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\types.h"
+				>
+			</File>
+			<File
+				RelativePath="..\..\include\x86emitter\x86emitter.h"
+				>
+			</File>
+			<Filter
+				Name="Implement"
+				>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\dwshift.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\group1.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\group2.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\group3.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\helpers.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\incdec.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\jmpcall.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\movs.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\test.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\xchg.h"
+					>
+				</File>
+			</Filter>
+			<Filter
+				Name="Implement_Simd"
+				>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\xmm\arithmetic.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\xmm\basehelpers.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\xmm\comparisons.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\xmm\moremovs.h"
+					>
+				</File>
+				<File
+					RelativePath="..\..\include\x86emitter\implement\xmm\shufflepack.h"
+					>
+				</File>
+			</Filter>
+		</Filter>
+		<Filter
+			Name="Resource Files"
+			Filter="rc;ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe;resx;tiff;tif;png;wav"
+			UniqueIdentifier="{67DA6AB6-F800-4c08-8B7A-83BB121AAD01}"
+			>
+		</Filter>
+	</Files>
+	<Globals>
+	</Globals>
+</VisualStudioProject>

common/include/PS2Edefs.h

@@ -77,10 +77,18 @@
 #define PS2E_SIO_VERSION  0x0001
 #ifdef COMMONdefs
 
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 u32   CALLBACK PS2EgetLibType(void);
 u32   CALLBACK PS2EgetLibVersion2(u32 type);
 char* CALLBACK PS2EgetLibName(void);
 
+#ifdef __cplusplus
+}
+#endif
+
 #endif
 
 // key values:

common/include/Pcsx2Config.h

@@ -44,30 +44,34 @@ extern SessionOverrideFlags g_Session;
 //////////////////////////////////////////////////////////////////////////
 // Pcsx2 User Configuration Options!
 
-//#define PCSX2_MICROVU	// Use Micro VU recs instead of Zero VU Recs
+#define PCSX2_GSMULTITHREAD		0x0001 // Use Multi-Threaded GS
-//#define PCSX2_MICROVU_	// Fully enable Micro VU recs (temporary option for now)
+#define PCSX2_EEREC				0x0010
-#define PCSX2_GSMULTITHREAD 1 // uses multi-threaded gs
+#define PCSX2_VU0REC			0x0020
-#define PCSX2_EEREC 0x10
+#define PCSX2_VU1REC			0x0040
-#define PCSX2_VU0REC 0x20
+#define PCSX2_FRAMELIMIT_MASK	0x0c00
-#define PCSX2_VU1REC 0x40
+#define PCSX2_FRAMELIMIT_NORMAL	0x0000
-#define PCSX2_FRAMELIMIT_MASK 0xc00
+#define PCSX2_FRAMELIMIT_LIMIT	0x0400
-#define PCSX2_FRAMELIMIT_NORMAL 0x000
+#define PCSX2_FRAMELIMIT_SKIP	0x0800
-#define PCSX2_FRAMELIMIT_LIMIT 0x400
+#define PCSX2_MICROVU0			0x1000 // Use Micro VU0 recs instead of Zero VU0 Recs
-#define PCSX2_FRAMELIMIT_SKIP 0x800
+#define PCSX2_MICROVU1			0x2000 // Use Micro VU1 recs instead of Zero VU1 Recs
 
 #define CHECK_FRAMELIMIT (Config.Options&PCSX2_FRAMELIMIT_MASK)
 
 //------------ CPU Options!!! ---------------
 #define CHECK_MULTIGS	(Config.Options&PCSX2_GSMULTITHREAD)
+#define CHECK_MICROVU0	(Config.Options&PCSX2_MICROVU0)
+#define CHECK_MICROVU1	(Config.Options&PCSX2_MICROVU1)
 #define CHECK_EEREC (!g_Session.ForceDisableEErec && Config.Options&PCSX2_EEREC)
 #define CHECK_VU0REC (!g_Session.ForceDisableVU0rec && Config.Options&PCSX2_VU0REC)
 #define CHECK_VU1REC (!g_Session.ForceDisableVU1rec && (Config.Options&PCSX2_VU1REC))
 
 //------------ SPECIAL GAME FIXES!!! ---------------
-#define CHECK_VUADDSUBHACK	 (Config.GameFixes & 0x1) // Special Fix for Tri-ace games, they use an encryption algorithm that requires VU addi opcode to be bit-accurate.
+#define CHECK_VUADDSUBHACK	 (Config.GameFixes & 0x01) // Special Fix for Tri-ace games, they use an encryption algorithm that requires VU addi opcode to be bit-accurate.
-#define CHECK_FPUCOMPAREHACK (Config.GameFixes & 0x4) // Special Fix for Digimon Rumble Arena 2, fixes spinning/hanging on intro-menu.
+#define CHECK_FPUCOMPAREHACK (Config.GameFixes & 0x04) // Special Fix for Digimon Rumble Arena 2, fixes spinning/hanging on intro-menu.
-#define CHECK_VUCLIPFLAGHACK (Config.GameFixes & 0x2) // Special Fix for Persona games, maybe others. It's to do with the VU clip flag (again).
+#define CHECK_VUCLIPFLAGHACK (Config.GameFixes & 0x02) // Special Fix for Persona games, maybe others. It's to do with the VU clip flag (again).
-#define CHECK_FPUMULHACK	 (Config.GameFixes & 0x8) // Special Fix for Tales of Destiny hangs.
+#define CHECK_FPUMULHACK	 (Config.GameFixes & 0x08) // Special Fix for Tales of Destiny hangs.
+#define CHECK_DMAEXECHACK	 (Config.GameFixes & 0x10) // Special Fix for Fatal Frame; breaks Gust and Tri-Ace games.
+#define CHECK_XGKICKHACK	 (Config.GameFixes & 0x20) // Special Fix for Erementar Gerad, adds more delay to VU XGkick instructions. Corrects the color of some graphics.
 
 //------------ Advanced Options!!! ---------------
 #define CHECK_VU_OVERFLOW		 (Config.vuOptions & 0x1)
@@ -165,8 +169,10 @@ public:
 		bool IOPCycleDouble;
 		bool WaitCycleExt;
 		bool INTCSTATSlow;
-		int VUCycleSteal;
 		bool IdleLoopFF;
+		int  VUCycleSteal;
+		bool vuFlagHack;
+		bool vuMinMax;
 		bool ESCExits; // this is a hack!?
 	} Hacks;
 	int GameFixes;

common/include/Pcsx2Defs.h

@@ -19,21 +19,62 @@
 #ifndef __PCSX2DEFS_H__
 #define __PCSX2DEFS_H__
 
-#ifndef __LINUX__
+#if defined (__linux__) && !defined(__LINUX__)  // some distributions are lower case
-#	if defined (__linux__)   // some distributions are lower case
 #	define __LINUX__
-#	endif
+#endif
 
-#	ifdef __CYGWIN__
+#ifdef __CYGWIN__
 #	define __LINUX__
-#	endif
 #endif
 
 #include "Pcsx2Types.h"
 
+//////////////////////////////////////////////////////////////////////////////////////////
+// Microsoft specific STL extensions for bounds checking and stuff: Enabled in devbuilds,
+// disabled in release builds. :)
+//
+#ifdef _MSC_VER
+#	pragma warning(disable:4244)	// disable warning C4244: '=' : conversion from 'big' to 'small', possible loss of data
+#	ifdef PCSX2_DEVBUILD
+#		undef _SECURE_SCL
+#		define _SECURE_SCL 1
+#		undef _SECURE_SCL_THROWS
+#		define _SECURE_SCL_THROWS 1
+#	else
+#		undef _SECURE_SCL
+#		define _SECURE_SCL 0
+#		undef _SECURE_SCL_THROWS
+#		define _SECURE_SCL_THROWS 0
+#	endif
+#endif
+
+#ifdef _MSC_VER
+#	include <intrin.h>
+#else
+#	include <intrin_x86.h>
+#endif
+
 // Renamed ARRAYSIZE to ArraySize -- looks nice and gets rid of Windows.h conflicts (air)
+// Notes: I'd have used ARRAY_SIZE instead but ran into cross-platform lib conflicts with
+// that as well.  >_<
 #ifndef ArraySize
-#define ArraySize(x) (sizeof(x)/sizeof((x)[0]))
+#	define ArraySize(x) (sizeof(x)/sizeof((x)[0]))
+#endif
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// __releaseinline -- a forceinline macro that is enabled for RELEASE/PUBLIC builds ONLY.
+// This is useful because forceinline can make certain types of debugging problematic since
+// functions that look like they should be called won't breakpoint since their code is
+// inlined, and it can make stack traces confusing or near useless.
+//
+// Use __releaseinline for things which are generally large functions where trace debugging
+// from Devel builds is likely useful; but which should be inlined in an optimized Release
+// environment.
+//
+#ifdef PCSX2_DEVBUILD
+#	define __releaseinline
+#else
+#	define __releaseinline __forceinline
 #endif
 
 //////////////////////////////////////////////////////////////////////////////////////////
@@ -41,66 +82,182 @@
 //  This is primarily useful for the default case switch optimizer, which enables VC to
 //  generate more compact switches.
 
-#ifdef NDEBUG
+#ifndef jASSUME
+#	ifdef NDEBUG
 #		define jBREAKPOINT() ((void) 0)
 #		ifdef _MSC_VER
 #			define jASSUME(exp) (__assume(exp))
 #		else
 #			define jASSUME(exp) ((void) sizeof(exp))
 #		endif
-#else
-#	if defined(_MSC_VER)
-#		define jBREAKPOINT() do { __asm int 3 } while(0)
 #	else
-#		define jBREAKPOINT() ((void) *(volatile char *) 0)
+#		define jBREAKPOINT() __debugbreak();
+#		define jASSUME(exp) do { if(exp) ; else jBREAKPOINT(); } while(0);
 #	endif
-#	define jASSUME(exp) if(exp) ; else jBREAKPOINT()
 #endif
 
-// disable the default case in a switch
+//////////////////////////////////////////////////////////////////////////////////////////
-#define jNO_DEFAULT \
+// jNO_DEFAULT -- disables the default case in a switch, which improves switch optimization
-{ \
+// under MSVC.
-default: \
+//
+// How it Works: jASSUME turns into an __assume(0) under msvc compilers, which when specified
+// in the 'default:' case of a switch tells the compiler that the case is unreachable, so
+// that it will not generate any code, LUTs, or conditionals to handle it.
+//
+// * In debug builds the default case will cause an assertion.
+// * In devel builds the default case will cause a LogicError exception (C++ only)
+// (either meaning the jNO_DEFAULT has been used incorrectly, and that the default case is in
+//  fact used and needs to be handled).
+//
+// MSVC Note: To stacktrace LogicError exceptions, add Exception::LogicError to the C++ First-
+// Chance Exception list (under Debug->Exceptions menu).
+//
+#ifndef jNO_DEFAULT
+#if defined(__cplusplus) && defined(PCSX2_DEVBUILD)
+#	define jNO_DEFAULT \
+	{ \
+	default: \
+		assert(0); \
+		if( !IsDebugBuild ) throw Exception::LogicError( "Incorrect usage of jNO_DEFAULT detected (default case is not unreachable!)" ); \
+		break; \
+	}
+#else
+#	define jNO_DEFAULT \
+	default: \
+	{ \
 		jASSUME(0); \
 		break; \
-}
+	}
-
+#endif
-/* common defines */
-#ifndef C_ASSERT
-#define C_ASSERT(e) typedef char __C_ASSERT__[(e)?1:-1]
 #endif
 
-#ifdef _MSC_VER
+//////////////////////////////////////////////////////////////////////////////////////////
-// Note: building the 'extern' into PCSX2_ALIGNED16_DECL fixes Visual Assist X's intellisense.
+// compile-time assertion; usable at static variable define level.
+// (typically used to confirm the correct sizeof() for struct types where size
+//  restaints must be enforced).
+//
+#ifndef C_ASSERT
+#	define C_ASSERT(e) typedef char __C_ASSERT__[(e)?1:-1]
+#endif
 
-#define PCSX2_ALIGNED(alig,x) __declspec(align(alig)) x
+//////////////////////////////////////////////////////////////////////////////////////////
-#define PCSX2_ALIGNED_EXTERN(alig,x) extern __declspec(align(alig)) x
+// Dev / Debug conditionals - Consts for using if() statements instead of uglier #ifdef.
-#define PCSX2_ALIGNED16(x) __declspec(align(16)) x
+//
-#define PCSX2_ALIGNED16_EXTERN(x) extern __declspec(align(16)) x
+// Note: Using if() optimizes nicely in Devel and Release builds, but will generate extra
+// code overhead in debug builds (since debug neither inlines, nor optimizes out const-
+// level conditionals).  Normally not a concern, but if you stick if( IsDevbuild ) in
+// some tight loops it will likely make debug builds unusably slow.
+//
+#ifdef __cplusplus
+#	ifdef PCSX2_DEVBUILD
+		static const bool IsDevBuild = true;
+#	else
+		static const bool IsDevBuild = false;
+#	endif
 
-#define __naked __declspec(naked)
+#	ifdef PCSX2_DEBUG
-#define __unused /*unused*/
+		static const bool IsDebugBuild = true;
-#define __noinline __declspec(noinline) 
+#	else
-#define CALLBACK    __stdcall
+		static const bool IsDebugBuild = false;
+#	endif
 
 #else
 
+#	ifdef PCSX2_DEVBUILD
+		static const u8 IsDevBuild = 1;
+#	else
+		static const u8 IsDevBuild = 0;
+#	endif
+
+#	ifdef PCSX2_DEBUG
+		static const u8 IsDebugBuild = 1;
+#	else
+		static const u8 IsDebugBuild = 0;
+#	endif
+#endif
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// PCSX2_ALIGNED16  - helper macros for aligning variables in MSVC and GCC.
+//
+// GCC Warning!  The GCC linker (LD) typically fails to assure alignment of class members.
+// If you want alignment to be assured, the variable must either be a member of a struct
+// or a static global.
+//
+// General Performance Warning: Any function that specifies alignment on a local (stack)
+// variable will have to align the stack frame on enter, and restore it on exit (adds
+// overhead).  Furthermore, compilers cannot inline functions that have aligned local
+// vars.  So use local var alignment with much caution.
+//
+// Note: building the 'extern' into PCSX2_ALIGNED16 fixes Visual Assist X's intellisense.
+//
+#ifdef _MSC_VER
+
+#	define PCSX2_ALIGNED(alig,x) __declspec(align(alig)) x
+#	define PCSX2_ALIGNED_EXTERN(alig,x) extern __declspec(align(alig)) x
+#	define PCSX2_ALIGNED16(x) __declspec(align(16)) x
+#	define PCSX2_ALIGNED16_EXTERN(x) extern __declspec(align(16)) x
+
+#	define __naked __declspec(naked)
+#	define __unused /*unused*/
+#	define __noinline __declspec(noinline) 
+
+// Don't know if there are Visual C++ equivalents of these. 
+#	define __hot
+#	define __cold
+#	define likely(x) x
+#	define unlikely(x) x
+
+#	define CALLBACK    __stdcall
+
+#else
+
+// GCC 4.4.0 is a bit nutty, as compilers go. it gets a define to itself.
+#	define GCC_VERSION (__GNUC__ * 10000 \
+                               + __GNUC_MINOR__ * 100 \
+                               + __GNUC_PATCHLEVEL__)
+          
+/* Test for GCC > 4.4.0; Should be adjusted when new versions come out */
+#	if GCC_VERSION >= 40400
+#		define THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0
+#		define __nooptimization __attribute__((optimize("O0")))
+#	endif
+
+/*
+This theoretically unoptimizes. Not having much luck so far.
+#	ifdef THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0
+#		pragma GCC optimize ("O0")
+#	endif
+
+#	ifdef THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0
+#		pragma GCC reset_options
+#	endif
+
+*/
+
 // fixme - is this needed for recent versions of GCC?  Or can we just use the first two macros
 // instead for both definitions (implementations) and declarations (includes)? -- air
-#define PCSX2_ALIGNED(alig,x) x __attribute((aligned(alig)))
+#	define PCSX2_ALIGNED(alig,x) x __attribute((aligned(alig)))
-#define PCSX2_ALIGNED16(x) x __attribute((aligned(16)))
+#	define PCSX2_ALIGNED16(x) x __attribute((aligned(16)))
-#define PCSX2_ALIGNED_EXTERN(alig,x) extern x __attribute((aligned(alig)))
+#	define PCSX2_ALIGNED_EXTERN(alig,x) extern x __attribute((aligned(alig)))
-#define PCSX2_ALIGNED16_EXTERN(x) extern x __attribute((aligned(16)))
+#	define PCSX2_ALIGNED16_EXTERN(x) extern x __attribute((aligned(16)))
 
-#define __naked		// GCC lacks the naked specifier
+#	define __naked		// GCC lacks the naked specifier
-#define CALLBACK	// CALLBACK is a win32-specific mess
+#	define CALLBACK	// CALLBACK is a win32-specific mess
 
 // GCC uses attributes for a lot of things that Visual C+ doesn't.
-#define __fastcall __attribute__((fastcall))
+#	define __fastcall		__attribute__((fastcall))
-#define __unused __attribute__((unused))
+#	define __unused			__attribute__((unused))
-#define _inline __inline__ __attribute__((unused))
+#	define _inline			__inline__ __attribute__((unused))
-#define __forceinline __attribute__((always_inline,unused))
+#	define __forceinline	__attribute__((always_inline,unused))
-#define __noinline __attribute__((noinline))
+#	define __noinline		__attribute__((noinline))
+#	define __hot			__attribute__((hot))
+#	define __cold			__attribute__((cold))
+#	define likely(x)		__builtin_expect(!!(x), 1)
+#	define unlikely(x)		__builtin_expect(!!(x), 0)
+#endif
+
+#ifndef THE_UNBEARABLE_LIGHTNESS_OF_BEING_GCC_4_4_0
+#	define __nooptimization
 #endif
 
 typedef struct {

common/include/Pcsx2Types.h

@@ -107,6 +107,7 @@ typedef s32 sptr;
 
 //////////////////////////////////////////////////////////////////////////////////////////
 // A rough-and-ready cross platform 128-bit datatype, Non-SSE style.
+//
 #ifdef __cplusplus
 struct u128
 {

common/include/Utilities/Console.h

@@ -0,0 +1,142 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#pragma once
+
+#include "StringHelpers.h"
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// Console Namespace -- For printing messages to the console.
+//
+// SysPrintf is depreciated; We should phase these in over time.
+//
+namespace Console
+{
+	enum Colors
+	{
+		Color_Black = 0,
+		Color_Red,
+		Color_Green,
+		Color_Yellow,
+		Color_Blue,
+		Color_Magenta,
+		Color_Cyan,
+		Color_White
+	};
+
+	// va_args version of WriteLn, mostly for internal use only.
+	extern void __fastcall _WriteLn( Colors color, const char* fmt, va_list args );
+
+	extern void __fastcall SetTitle( const wxString& title );
+
+	// Changes the active console color.
+	// This color will be unset by calls to colored text methods
+	// such as ErrorMsg and Notice.
+	extern void __fastcall SetColor( Colors color );
+
+	// Restores the console color to default (usually low-intensity white on Win32)
+	extern void ClearColor();
+
+	// The following Write functions return bool so that we can use macros to exclude
+	// them from different build types.  The return values are always zero.
+
+	// Writes a newline to the console.
+	extern bool Newline();
+
+	// Writes an unformatted string of text to the console (fast!)
+	// No newline is appended.
+	extern bool __fastcall Write( const char* text );
+
+	// Writes an unformatted string of text to the console (fast!)
+	// A newline is automatically appended, and the console color reset to default
+	// after the log is written.
+	extern bool __fastcall Write( Colors color, const char* text );
+
+	// Writes an unformatted string of text to the console (fast!)
+	// A newline is automatically appended.
+	extern bool __fastcall WriteLn( const char* text );
+
+	// Writes an unformatted string of text to the console (fast!)
+	// A newline is automatically appended, and the console color reset to default
+	// after the log is written.
+	extern bool __fastcall WriteLn( Colors color, const char* text );
+
+	// Writes a line of colored text to the console, with automatic newline appendage.
+	// The console color is reset to default when the operation is complete.
+	extern bool WriteLn( Colors color, const char* fmt, VARG_PARAM dummy, ... );
+
+	// Writes a formatted message to the console, with appended newline.
+	extern bool WriteLn( const char* fmt, VARG_PARAM dummy, ... );
+
+	// Writes a line of colored text to the console (no newline).
+	// The console color is reset to default when the operation is complete.
+	extern bool Write( Colors color, const char* fmt, VARG_PARAM dummy, ... );
+
+	// Writes a formatted message to the console (no newline)
+	extern bool Write( const char* fmt, VARG_PARAM dummy, ... );
+
+	// Displays a message in the console with red emphasis.
+	// Newline is automatically appended.
+	extern bool Error( const char* fmt, VARG_PARAM dummy, ... );
+	extern bool __fastcall Error( const char* text );
+
+	// Displays a message in the console with yellow emphasis.
+	// Newline is automatically appended.
+	extern bool Notice( const char* fmt, VARG_PARAM dummy, ... );
+	extern bool __fastcall Notice( const char* text );
+
+	// Displays a message in the console with yellow emphasis.
+	// Newline is automatically appended.
+	extern bool Status( const char* fmt, VARG_PARAM dummy, ... );
+	extern bool __fastcall Status( const char* text );
+
+
+	extern bool __fastcall Write( const wxString& text );
+	extern bool __fastcall Write( Colors color, const wxString& text );
+	extern bool __fastcall WriteLn( const wxString& text );
+	extern bool __fastcall WriteLn( Colors color, const wxString& text );
+
+	extern bool __fastcall Error( const wxString& text );
+	extern bool __fastcall Notice( const wxString& text );
+	extern bool __fastcall Status( const wxString& text );
+}
+
+using Console::Color_Red;
+using Console::Color_Green;
+using Console::Color_Blue;
+using Console::Color_Magenta;
+using Console::Color_Cyan;
+using Console::Color_Yellow;
+using Console::Color_White;
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// DevCon / DbgCon
+
+#ifdef PCSX2_DEVBUILD
+#	define DevCon Console
+#	define DevMsg MsgBox
+#else
+#	define DevCon 0&&Console
+#	define DevMsg 
+#endif
+
+#ifdef PCSX2_DEBUG
+#	define DbgCon Console
+#else
+#	define DbgCon 0&&Console
+#endif

common/include/Utilities/Dependencies.h

@@ -1,5 +1,5 @@
 /*  Pcsx2 - Pc Ps2 Emulator
- *  Copyright (C) 2002-2008  Pcsx2 Team
+ *  Copyright (C) 2002-2009  Pcsx2 Team
  *
  *  This program is free software; you can redistribute it and/or modify
  *  it under the terms of the GNU General Public License as published by
@@ -15,29 +15,20 @@
  *  along with this program; if not, write to the Free Software
  *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
  */
-//============================================
-//=== Audio XA decoding
-//=== Kazzuya
-//============================================
 
-#ifndef DECODEXA_H
+#pragma once
-#define DECODEXA_H
 
-typedef struct {
+#include "Pcsx2Defs.h"
-	long	y0, y1;
-} ADPCM_Decode_t;
 
-typedef struct {
+#include <wx/string.h>
-	int				freq;
+#include <wx/tokenzr.h>
-	int				nbits;
+#include <wx/gdicmn.h>		// for wxPoint/wxRect stuff
-	int				stereo;
+#include <wx/intl.h>
-	int				nsamples;
+#include <wx/log.h>
-	ADPCM_Decode_t	left, right;
-	short			pcm[16384];
-} xa_decode_t;
 
-long xa_decode_sector( xa_decode_t *xdp,
+#include <stdexcept>
-					   unsigned char *sectorp,
+#include <algorithm>
-					   int is_first_sector );
+#include <string>
-
+#include <cstring>		// string.h under c++
-#endif
+#include <cstdio>		// stdio.h under c++
+#include <cstdlib>

common/include/Utilities/Exceptions.h

@@ -18,6 +18,10 @@
 
 #pragma once
 
+#include "Dependencies.h"
+
+extern void DevAssert( bool condition, const char* msg );
+
 //////////////////////////////////////////////////////////////////////////////////////////
 // This class provides an easy and clean method for ensuring objects are not copyable.
 class NoncopyableObject
@@ -56,11 +60,6 @@ protected:
 // for every class that derives from it. 
 #define Sealed private virtual __BaseSealed<__COUNTER__>
 
-extern wxLocale* g_EnglishLocale;
-
-extern wxString GetEnglish( const char* msg );
-extern wxString GetTranslation( const char* msg );
-
 namespace Exception
 {
 	//////////////////////////////////////////////////////////////////////////////////
@@ -104,7 +103,7 @@ namespace Exception
 	//////////////////////////////////////////////////////////////////////////////////////////
 	// This class is used as a base exception for things tossed by PS2 cpus (EE, IOP, etc).
 	// Translation Note: These exceptions are never translated, except to issue a general
-	// error message to the user (which is xspecified below).
+	// error message to the user (which is specified below).
 	//
 	class Ps2Generic : public BaseException
 	{
@@ -151,6 +150,18 @@ namespace Exception
 			BaseException( msg_eng, msg_xlt ) { }
 	};
 	
+	class AssertionFailure : public LogicError
+	{
+	public:
+		explicit AssertionFailure( const char* msg="Assertion Failure" ) :
+			LogicError( msg ) {}
+
+		explicit AssertionFailure( const wxString& msg_eng, const wxString& msg_xlt ) :
+			LogicError( msg_eng, msg_xlt ) { }
+
+		virtual ~AssertionFailure() throw() {}
+	};
+	
 	//////////////////////////////////////////////////////////////////////////////////
 	//
 	class OutOfMemory : public RuntimeError

common/include/Utilities/General.h

@@ -0,0 +1,57 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *  
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *  
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#pragma once
+
+enum PageProtectionMode
+{
+	Protect_NoAccess = 0,
+	Protect_ReadOnly,
+	Protect_ReadWrite
+};
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// HostSys - Namespace housing general system-level implementations relating to loading
+// plugins and allocating memory.  For now, these functions are all accessed via Sys*
+// versions defined in System.h/cpp.
+//
+namespace HostSys
+{
+	// Maps a block of memory for use as a recompiled code buffer.
+	// The allocated block has code execution privileges.
+	// Returns NULL on allocation failure.
+	extern void *Mmap(uptr base, u32 size);
+
+	// Unmaps a block allocated by SysMmap
+	extern void Munmap(uptr base, u32 size);
+
+	extern void MemProtect( void* baseaddr, size_t size, PageProtectionMode mode, bool allowExecution=false );
+
+	static __forceinline void Munmap( void* base, u32 size )
+	{
+		Munmap( (uptr)base, size );
+	}
+}
+
+
+//////////////////////////////////////////////////////////////////////////////////////////
+
+
+extern void InitCPUTicks();
+extern u64  GetTickFrequency();
+extern u64  GetCPUTicks();

common/include/Utilities/MemcpyFast.h

@@ -18,28 +18,26 @@
 
 #pragma once
 
-#if defined(_WIN32)
-	#include "windows/memzero.h"
-#else
-	#include "Linux/memzero.h"
-#endif // WIN32
-
-	// Only used in the Windows version of memzero.h. But it's in Misc.cpp for some reason.
-	void _memset16_unaligned( void* dest, u16 data, size_t size );
-
-	// The new simplified memcpy_amd_ is now faster than memcpy_raz_.
-	// memcpy_amd_ also does mmx register saving, negating the need for freezeregs (code cleanup!)
-	// Additionally, using one single memcpy implementation keeps the code cache cleaner.
-
 #ifdef __LINUX__
+
+#	include "lnx_memzero.h"
+
 	extern "C" void __fastcall memcpy_amd_(void *dest, const void *src, size_t bytes);
 	extern "C" u8 memcmp_mmx(const void* src1, const void* src2, int cmpsize);
 	extern "C" void memxor_mmx(void* dst, const void* src1, int cmpsize);
+
 #else
+
+#	include "win_memzero.h"
+
 	extern void __fastcall memcpy_amd_(void *dest, const void *src, size_t bytes);
 	extern u8 memcmp_mmx(const void* src1, const void* src2, int cmpsize);
 	extern void memxor_mmx(void* dst, const void* src1, int cmpsize);
+
 #endif
 
+// Only used in the Windows version of memzero.h. But it's in Misc.cpp for some reason.
+void _memset16_unaligned( void* dest, u16 data, size_t size );
+
 #define memcpy_fast memcpy_amd_
 #define memcpy_aligned memcpy_amd_

common/include/Utilities/RedtapeWindows.h

@@ -0,0 +1,45 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#pragma once
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// Windows Redtape!  No windows.h should be included without it!
+//
+// This header's purpose is to include windows.h with the correct OS version info, and
+// to undefine some of windows.h's more evil macros (min/max).  It also does a _WIN32
+// check, so that we don't have to do it explicitly in every instance where it might
+// be needed from non-Win32-specific files
+
+#define NOMINMAX		// Disables other libs inclusion of their own min/max macros (we use std instead)
+
+#ifdef _WIN32
+
+// Force availability of to WinNT APIs (change to 0x600 to enable XP-specific APIs)
+#ifndef WINVER 
+#define WINVER 0x0501
+#define _WIN32_WINNT 0x0501
+#endif
+
+#include <windows.h>
+
+// disable Windows versions of min/max -- we'll use the typesafe STL versions instead.
+#undef min
+#undef max
+
+#endif

common/include/Utilities/SafeArray.h

@@ -75,21 +75,19 @@ public:
 	static const int DefaultChunkSize = 0x1000 * sizeof(T);
 
 public: 
-	const wxString Name;		// user-assigned block name
+	const wxChar* Name;		// user-assigned block name
 	int ChunkSize;
 
 protected:
 	T* m_ptr;
 	int m_size;	// size of the allocation of memory
 
-	const static wxString m_str_Unnamed;
-
 protected:
 	// Internal constructor for use by derived classes.  This allows a derived class to
 	// use its own memory allocation (with an aligned memory, for example).
 	// Throws:
 	//   Exception::OutOfMemory if the allocated_mem pointer is NULL.
-	explicit SafeArray( const wxString& name, T* allocated_mem, int initSize ) :
+	explicit SafeArray( const wxChar* name, T* allocated_mem, int initSize ) : 
 	  Name( name )
 	, ChunkSize( DefaultChunkSize )
 	, m_ptr( allocated_mem )
@@ -101,7 +99,10 @@ protected:
 
 	virtual T* _virtual_realloc( int newsize )
 	{
-		return (T*)realloc( m_ptr, newsize * sizeof(T) );
+		return (T*)((m_ptr == NULL) ? 
+			malloc( newsize * sizeof(T) ) :
+			realloc( m_ptr, newsize * sizeof(T) )
+		);
 	}
 
 public:
@@ -110,7 +111,7 @@ public:
 		safe_free( m_ptr );
 	}
 
-	explicit SafeArray( const wxString& name = L"Unnamed" ) :
+	explicit SafeArray( const wxChar* name=L"Unnamed" ) : 
 	  Name( name )
 	, ChunkSize( DefaultChunkSize )
 	, m_ptr( NULL )
@@ -118,59 +119,64 @@ public:
 	{
 	}
 
-	explicit SafeArray( const char* name ) :
+	explicit SafeArray( int initialSize, const wxChar* name=L"Unnamed" ) : 
-		Name( wxString::FromAscii(name) )
-	,	ChunkSize( DefaultChunkSize )
-	,	m_ptr( NULL )
-	,	m_size( 0 )
-	{
-	}
-
-	explicit SafeArray( int initialSize, const wxString& name = L"Unnamed" ) :
 	  Name( name )
 	, ChunkSize( DefaultChunkSize )
-	,	m_ptr( (T*)malloc( initialSize * sizeof(T) ) )
+	, m_ptr( (initialSize==0) ? NULL : (T*)malloc( initialSize * sizeof(T) ) )
 	, m_size( initialSize )
 	{
-		if( m_ptr == NULL )
+		if( (initialSize != 0) && (m_ptr == NULL) )
 			throw Exception::OutOfMemory();
 	}
 	
-	explicit SafeArray( int initialSize, const char* name ) :
+	// Clears the contents of the array to zero, and frees all memory allocations.
-		Name( wxString::FromAscii(name) )
+	void Dispose()
-	,	ChunkSize( DefaultChunkSize )
-	,	m_ptr( (T*)malloc( initialSize * sizeof(T) ) )
-	,	m_size( initialSize )
 	{
-		if( m_ptr == NULL )
+		m_size = 0;
-			throw Exception::OutOfMemory();
+		safe_free( m_ptr );
 	}
 
+	bool IsDisposed() const { return (m_ptr==NULL); }
+
 	// Returns the size of the memory allocation, as according to the array type.
 	int GetLength() const { return m_size; }
 	// Returns the size of the memory allocation in bytes.
 	int GetSizeInBytes() const { return m_size * sizeof(T); }
 
-	// Ensures that the allocation is large enough to fit data of the
+	// reallocates the array to the explicit size.  Can be used to shrink or grow an
-	// amount requested.  The memory allocation is not resized smaller.
+	// array, and bypasses the internal threshold growth indicators.	
-	void MakeRoomFor( int blockSize )
+	void ExactAlloc( int newsize )
 	{
-		if( blockSize > m_size )
+		if( newsize == m_size ) return;
-		{
+
-			const uint newalloc = blockSize + ChunkSize;
+		m_ptr = _virtual_realloc( newsize );
-			m_ptr = _virtual_realloc( newalloc );
 		if( m_ptr == NULL )
 		{
+
 				throw Exception::OutOfMemory(
 					wxsFormat(	// english (for diagnostic)
 						L"Out-of-memory on SafeArray block re-allocation.\n"
 						L"Old size: %d bytes, New size: %d bytes.",
-						m_size, newalloc
+						m_size, newsize
 					)
 				);
 		}
-			m_size = newalloc;
+		m_size = newsize;
 	}
+
+	// Ensures that the allocation is large enough to fit data of the
+	// amount requested.  The memory allocation is not resized smaller.
+	void MakeRoomFor( int newsize )
+	{
+		if( newsize > m_size )
+			ExactAlloc( newsize );
+	}
+	
+	// Extends the containment area of the array.  Extensions are performed
+	// in chunks.
+	void GrowBy( int items )
+	{
+		MakeRoomFor( m_size + ChunkSize + items + 1 );
 	}
 
 	// Gets a pointer to the requested allocation index.
@@ -205,20 +211,26 @@ protected:
 #endif
 		return &m_ptr[i];
 	}
-
 };
 
 //////////////////////////////////////////////////////////////////////////////////////////
+// SafeList - Simple growable container without all the mess or hassle of std containers.
 //
+// This container is intended for reasonably simple class types only.  Things which this
+// container does not handle with desired robustness:
+//
+//  * Classes with non-trivial constructors (such that construction creates much overhead)
+//  * Classes with copy constructors (copying is done using performance memcpy)
+//  * Classes with destructors (they're not called, sorry!)
 //
 template< typename T >
-class SafeList
+class SafeList : public NoncopyableObject
 {
 public:
 	static const int DefaultChunkSize = 0x80 * sizeof(T);
 
 public: 
-	const wxString Name;		// user-assigned block name
+	const wxChar* Name;		// user-assigned block name
 	int ChunkSize;			// assigned DefaultChunkSize on init, reconfigurable at any time.
 
 protected:
@@ -226,8 +238,6 @@ protected:
 	int m_allocsize;	// size of the allocation of memory
 	uint m_length;		// length of the array (active items, not buffer allocation)
 
-	const static wxString m_str_Unnamed;
-
 protected:
 	virtual T* _virtual_realloc( int newsize )
 	{
@@ -243,9 +253,10 @@ protected:
 public:
 	virtual ~SafeList()
 	{
+		safe_free( m_ptr );
 	}
 
-	explicit SafeList( const wxString& name = L"Unnamed" ) :
+	explicit SafeList( const wxChar* name=L"Unnamed" ) : 
 		Name( name )
 	,	ChunkSize( DefaultChunkSize )
 	,	m_ptr( NULL )
@@ -254,16 +265,7 @@ public:
 	{
 	}
 
-	explicit SafeList( const char* name ) :
+	explicit SafeList( int initialSize, const wxChar* name=L"Unnamed" ) : 
-		Name( wxString::FromAscii(name) )
-	,	ChunkSize( DefaultChunkSize )
-	,	m_ptr( NULL )
-	,	m_allocsize( 0 )
-	,	m_length( 0 )
-	{
-	}
-
-	explicit SafeList( int initialSize, const wxString& name = L"Unnamed" ) :
 		Name( name )
 	,	ChunkSize( DefaultChunkSize )
 	,	m_ptr( (T*)malloc( initialSize * sizeof(T) ) )
@@ -272,17 +274,12 @@ public:
 	{
 		if( m_ptr == NULL )
 			throw Exception::OutOfMemory();
+
+		for( int i=0; i<m_allocsize; ++i )
+		{
+			new (&m_ptr[i]) T();
 		}
 
-	explicit SafeList( int initialSize, const char* name ) :
-		Name( wxString::FromAscii(name) )
-	,	ChunkSize( DefaultChunkSize )
-	,	m_ptr( (T*)malloc( initialSize * sizeof(T) ) )
-	,	m_allocsize( initialSize )
-	,	m_length( 0 )
-	{
-		if( m_ptr == NULL )
-			throw Exception::OutOfMemory();
 	}
 
 	// Returns the size of the list, as according to the array type.  This includes
@@ -299,7 +296,7 @@ public:
 	{
 		if( blockSize > m_allocsize )
 		{
-			const uint newalloc = blockSize + ChunkSize;
+			const int newalloc = blockSize + ChunkSize;
 			m_ptr = _virtual_realloc( newalloc );
 			if( m_ptr == NULL )
 			{
@@ -312,22 +309,58 @@ public:
 					)
 				);
 			}
-			m_allocsize = newalloc;
+
+			for( ; m_allocsize<newalloc; ++m_allocsize )
+			{
+				new (&m_ptr[m_allocsize]) T();
 			}
 		}
+	}
+
+	void GrowBy( int items )
+	{
+		MakeRoomFor( m_length + ChunkSize + items + 1 );
+	}
+
+	// Sets the item length to zero.  Does not free memory allocations.
+	void Clear()
+	{
+		m_length = 0;
+	}
+	
+	// Appends an item to the end of the list and returns a handle to it.
+	T& New()
+	{
+		_MakeRoomFor_threshold( m_length + 1 );
+		return m_ptr[m_length++];
+	}
 
 	// Gets an element of this memory allocation much as if it were an array.
 	// DevBuilds : Throws Exception::IndexBoundsFault() if the index is invalid.
 	T& operator[]( int idx )				{ return *_getPtr( (uint)idx ); }
 	const T& operator[]( int idx ) const	{ return *_getPtr( (uint)idx ); }
 
+	T* GetPtr()				{ return m_ptr; }
+	const T* GetPtr() const	{ return m_ptr; }
+
+	T& GetLast()			{ return m_ptr[m_length-1]; }
+	const T& GetLast() const{ return m_ptr[m_length-1]; }
+
 	int Add( const T& src )
 	{
-		MakeRoomFor( m_length + 1 );
+		_MakeRoomFor_threshold( m_length + 1 );
 		m_ptr[m_length] = src;
 		return m_length++;
 	}
 
+	// Same as Add, but returns the handle of the new object instead of it's array index.
+	T& AddNew( const T& src )
+	{
+		_MakeRoomFor_threshold( m_length + 1 );
+		m_ptr[m_length] = src;
+		return m_ptr[m_length];
+	}
+	
 	// Performs a standard array-copy removal of the given item.  All items past the
 	// given item are copied over.  Throws Exception::IndexBoundsFault() if the index
 	// is invalid (devbuilds only)
@@ -347,6 +380,12 @@ public:
 	}
 
 protected:
+
+	void _MakeRoomFor_threshold( int newsize )
+	{
+		MakeRoomFor( newsize + ChunkSize );
+	}
+
 	// A safe array index fetcher.  Throws an exception if the array index
 	// is outside the bounds of the array.
 	// Performance Considerations: This function adds quite a bit of overhead
@@ -371,18 +410,14 @@ class SafeAlignedArray : public SafeArray<T>
 protected:
 	T* _virtual_realloc( int newsize )
 	{
-		return (T*)_aligned_realloc( this->m_ptr, newsize * sizeof(T), Alignment );
+		return (T*)( ( this->m_ptr == NULL ) ?
+			_aligned_malloc( newsize * sizeof(T), Alignment ) :
+			_aligned_realloc( this->m_ptr, newsize * sizeof(T), Alignment )
+		);
 	}
 
 	// Appends "(align: xx)" to the name of the allocation in devel builds.
 	// Maybe useful,maybe not... no harm in attaching it. :D
-	wxString _getName( const wxString& src )
-	{
-		if( IsDevBuild )
-			return src + wxsFormat( L"(align: %d)", Alignment );
-		else
-			return src;
-	}
 
 public:
 	virtual ~SafeAlignedArray()
@@ -391,28 +426,14 @@ public:
 		// mptr is set to null, so the parent class's destructor won't re-free it.
 	}
 
-	explicit SafeAlignedArray( const wxString& name = L"Unnamed") :
+	explicit SafeAlignedArray( const wxChar* name=L"Unnamed" ) : 
 		SafeArray<T>::SafeArray( name )
 	{
 	}
 
-	explicit SafeAlignedArray( const char* name ) :
+	explicit SafeAlignedArray( int initialSize, const wxChar* name=L"Unnamed" ) : 
-		SafeArray<T>::SafeArray( name )
-	{
-	}
-
-	explicit SafeAlignedArray( int initialSize, const wxString& name = L"Unnamed") :
 		SafeArray<T>::SafeArray(
-			_getName(name),
+			name,
-			(T*)_aligned_malloc( initialSize * sizeof(T), Alignment ),
-			initialSize
-		)
-	{
-	}
-
-	explicit SafeAlignedArray( int initialSize, const char* name ) :
-		SafeArray<T>::SafeArray(
-			_getName(wxString::FromAscii(name)),
 			(T*)_aligned_malloc( initialSize * sizeof(T), Alignment ),
 			initialSize
 		)

common/include/Utilities/StringHelpers.h

@@ -0,0 +1,91 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#pragma once
+
+#include <wx/string.h>
+#include <wx/tokenzr.h>
+#include <wx/gdicmn.h>		// for wxPoint/wxRect stuff
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// Helpers for wxWidgets stuff!
+//
+
+// wxWidgets lacks one of its own...
+extern const wxRect wxDefaultRect;
+
+// This should prove useful....
+#define wxsFormat wxString::Format
+
+extern wxString ToString( const wxPoint& src, const wxString& separator=L"," );
+extern wxString ToString( const wxSize& src, const wxString& separator=L"," );
+extern wxString ToString( const wxRect& src, const wxString& separator=L"," );
+
+extern bool TryParse( wxPoint& dest, const wxStringTokenizer& parts );
+extern bool TryParse( wxSize& dest, const wxStringTokenizer& parts );
+
+extern bool TryParse( wxPoint& dest, const wxString& src, const wxPoint& defval=wxDefaultPosition, const wxString& separators=L",");
+extern bool TryParse( wxSize& dest, const wxString& src, const wxSize& defval=wxDefaultSize, const wxString& separators=L",");
+extern bool TryParse( wxRect& dest, const wxString& src, const wxRect& defval=wxDefaultRect, const wxString& separators=L",");
+
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// dummy structure used to type-guard the dummy parameter that's been inserted to
+// allow us to use the va_list feature on references.
+struct _VARG_PARAM
+{
+	// just some value to make the struct length 32bits instead of 8 bits, so that the
+	// compiler generates somewhat more efficient code.
+	uint someval;
+};
+
+#ifdef PCSX2_DEBUG
+
+#define params va_arg_dummy,
+#define varg_assert()  // jASSUME( dummy == &va_arg_dummy );
+// typedef the Va-Arg value to be a value type in debug builds.  The value
+// type requires a little more overhead in terms of code generation, but is always
+// type-safe.  The compiler will generate errors for any forgotten params value.
+typedef _VARG_PARAM VARG_PARAM;
+
+#else
+
+#define params NULL,	// using null is faster / more compact!
+#define varg_assert()	jASSUME( dummy == NULL );
+// typedef the Va-Arg value to be a pointer in release builds.  Pointers
+// generate more compact code by a small margin, but aren't entirely type safe since
+// the compiler won't generate errors if you pass NULL or other values.
+typedef _VARG_PARAM const * VARG_PARAM;
+
+#endif
+
+extern const _VARG_PARAM va_arg_dummy;
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// Custom internal sprintf functions, which are ASCII only (even in UNICODE builds)
+//
+// These functions are useful since they are ASCII always, even under Unicode.  Typically
+// even in a unicode app.
+
+extern void ssprintf(std::string& dest, const char* fmt, ...);
+extern void ssappendf(std::string& dest, const char* format, ...);
+extern void vssprintf(std::string& dest, const char* format, va_list args);
+extern void vssappendf(std::string& dest, const char* format, va_list args);
+
+extern std::string fmt_string( const char* fmt, ... );
+extern std::string vfmt_string( const char* fmt, va_list args );

common/include/Utilities/Threading.h

@@ -16,10 +16,10 @@
  *	Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
  */
  
-#ifndef _THREADING_H_
+ #pragma once
-#define _THREADING_H_
 
 #include <errno.h> // EBUSY
+#include <pthread.h>
 #include <semaphore.h>
 
 #include "Pcsx2Defs.h"
@@ -276,4 +276,3 @@ namespace Threading
 
 }
 
-#endif

common/include/Utilities/win_memzero.h

@@ -16,8 +16,7 @@
  *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
  */
 
-#ifndef _WIN_MEMZERO_H_
+#pragma once
-#define _WIN_MEMZERO_H_
 
 // These functions are meant for memset operations of constant length only.
 // For dynamic length clears, use the C-compiler provided memset instead.
@@ -36,6 +35,11 @@
 //  structures, which are constant in size, thus allowing us to generate optimal compile-
 //  time code for each use of the function.
 
+// Use of CLD (Clear Direction Flag):
+//  On Windows platforms the ABI declares that the direction flag should be cleared upon
+//  entry of *any* function.  Therefore there is no need to have CLD prior to our use of
+//  rep strosd here.
+
 // Notes on XMM0's "storage" area (_xmm_backup):
 // Unfortunately there's no way to guarantee alignment for this variable.  If I use the
 // __declspec(aligned(16)) decorator, MSVC fails to inline the function since stack
@@ -197,7 +201,6 @@ static __forceinline void memzero_ptr( void *dest )
 		case 3:
 			__asm
 			{
-				cld;
 				mov edi, dest
 				xor eax, eax
 				stosd
@@ -209,7 +212,6 @@ static __forceinline void memzero_ptr( void *dest )
 		case 4:
 			__asm
 			{
-				cld;
 				mov edi, dest
 				xor eax, eax
 				stosd
@@ -222,7 +224,6 @@ static __forceinline void memzero_ptr( void *dest )
 		case 5:
 			__asm
 			{
-				cld;
 				mov edi, dest
 				xor eax, eax
 				stosd
@@ -236,7 +237,6 @@ static __forceinline void memzero_ptr( void *dest )
 		default:
 			__asm
 			{
-				cld;
 				mov ecx, remdat
 				mov edi, dest
 				xor eax, eax
@@ -336,7 +336,6 @@ static __forceinline void memset_8( void *dest )
 		case 3:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -348,7 +347,6 @@ static __forceinline void memset_8( void *dest )
 		case 4:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -361,7 +359,6 @@ static __forceinline void memset_8( void *dest )
 		case 5:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -375,7 +372,6 @@ static __forceinline void memset_8( void *dest )
 		default:
 			__asm
 			{
-				cld;
 				mov ecx, remdat;
 				mov edi, dest;
 				mov eax, data32;
@@ -428,7 +424,6 @@ static __forceinline void memset_16( void *dest )
 		case 3:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -440,7 +435,6 @@ static __forceinline void memset_16( void *dest )
 		case 4:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -453,7 +447,6 @@ static __forceinline void memset_16( void *dest )
 		case 5:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -467,7 +460,6 @@ static __forceinline void memset_16( void *dest )
 		default:
 			__asm
 			{
-				cld;
 				mov ecx, remdat;
 				mov edi, dest;
 				mov eax, data32;
@@ -515,7 +507,6 @@ static __forceinline void memset_32( void *dest )
 		case 3:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -527,7 +518,6 @@ static __forceinline void memset_32( void *dest )
 		case 4:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -540,7 +530,6 @@ static __forceinline void memset_32( void *dest )
 		case 5:
 			__asm
 			{
-				cld;
 				mov edi, dest;
 				mov eax, data32;
 				stosd;
@@ -554,7 +543,6 @@ static __forceinline void memset_32( void *dest )
 		default:
 			__asm
 			{
-				cld;
 				mov ecx, remdat;
 				mov edi, dest;
 				mov eax, data32;
@@ -594,5 +582,3 @@ static __forceinline void memset32_obj( T& object )
 	memset_32<data, sizeof(T)>( &object );
 }
 
-#endif
-

common/include/intrin_x86.h

@@ -0,0 +1,954 @@
+/*
+	Compatibility <intrin_x86.h> header for GCC -- GCC equivalents of intrinsic
+	Microsoft Visual C++ functions. Originally developed for the ReactOS
+	(<http://www.reactos.org/>) and TinyKrnl (<http://www.tinykrnl.org/>)
+	projects.
+
+	Copyright (c) 2006 KJK::Hyperion <hackbunny@reactos.com>
+
+	Permission is hereby granted, free of charge, to any person obtaining a
+	copy of this software and associated documentation files (the "Software"),
+	to deal in the Software without restriction, including without limitation
+	the rights to use, copy, modify, merge, publish, distribute, sublicense,
+	and/or sell copies of the Software, and to permit persons to whom the
+	Software is furnished to do so, subject to the following conditions:
+
+	The above copyright notice and this permission notice shall be included in
+	all copies or substantial portions of the Software.
+
+	THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+	IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+	FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+	AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+	LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+	FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+	DEALINGS IN THE SOFTWARE.
+*/
+
+#ifndef KJK_INTRIN_X86_H_
+#define KJK_INTRIN_X86_H_
+
+/*
+	FIXME: review all "memory" clobbers, add/remove to match Visual C++
+	behavior: some "obvious" memory barriers are not present in the Visual C++
+	implementation - e.g. __stosX; on the other hand, some memory barriers that
+	*are* present could have been missed
+*/
+
+/*
+	NOTE: this is a *compatibility* header. Some functions may look wrong at
+	first, but they're only "as wrong" as they would be on Visual C++. Our
+	priority is compatibility
+
+	NOTE: unlike most people who write inline asm for GCC, I didn't pull the
+	constraints and the uses of __volatile__ out of my... hat. Do not touch
+	them. I hate cargo cult programming
+
+	NOTE: be very careful with declaring "memory" clobbers. Some "obvious"
+	barriers aren't there in Visual C++ (e.g. __stosX)
+
+	NOTE: review all intrinsics with a return value, add/remove __volatile__
+	where necessary. If an intrinsic whose value is ignored generates a no-op
+	under Visual C++, __volatile__ must be omitted; if it always generates code
+	(for example, if it has side effects), __volatile__ must be specified. GCC
+	will only optimize out non-volatile asm blocks with outputs, so input-only
+	blocks are safe. Oddities such as the non-volatile 'rdmsr' are intentional
+	and follow Visual C++ behavior
+
+	NOTE: on GCC 4.1.0, please use the __sync_* built-ins for barriers and
+	atomic operations. Test the version like this:
+
+	#if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) > 40100
+		...
+
+	Pay attention to the type of barrier. Make it match with what Visual C++
+	would use in the same case
+*/
+
+
+#ifndef __SIZE_T
+#define __SIZE_T
+typedef unsigned int size_t;
+#endif
+
+/*** Stack frame juggling ***/
+#define _ReturnAddress() (__builtin_return_address(0))
+#define _AddressOfReturnAddress() (&(((void **)(__builtin_frame_address(0)))[1]))
+/* TODO: __getcallerseflags but how??? */
+
+
+/*** Atomic operations ***/
+
+#if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) > 40100
+#define _ReadWriteBarrier() __sync_synchronize()
+#else
+static void __inline__ __attribute__((always_inline)) _MemoryBarrier(void)
+{
+	__asm__ __volatile__("" : : : "memory");
+}
+#define _ReadWriteBarrier() _MemoryBarrier()
+#endif
+
+/* BUGBUG: GCC only supports full barriers */
+#define _ReadBarrier _ReadWriteBarrier
+#define _WriteBarrier _ReadWriteBarrier
+
+#if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) > 40100
+
+static __inline__ __attribute__((always_inline)) char _InterlockedCompareExchange8(volatile char * const Destination, const char Exchange, const char Comperand)
+{
+	return __sync_val_compare_and_swap(Destination, Comperand, Exchange);
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedCompareExchange16(volatile short * const Destination, const short Exchange, const short Comperand)
+{
+	return __sync_val_compare_and_swap(Destination, Comperand, Exchange);
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedCompareExchange(volatile long * const Destination, const long Exchange, const long Comperand)
+{
+	return __sync_val_compare_and_swap(Destination, Comperand, Exchange);
+}
+
+static __inline__ __attribute__((always_inline)) long long _InterlockedCompareExchange64(volatile long long * const Destination, const long long Exchange, const long long Comperand)
+{
+	return __sync_val_compare_and_swap(Destination, Comperand, Exchange);
+}
+
+static __inline__ __attribute__((always_inline)) void * _InterlockedCompareExchangePointer(void * volatile * const Destination, void * const Exchange, void * const Comperand)
+{
+	return __sync_val_compare_and_swap(Destination, Comperand, Exchange);
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedExchange(volatile long * const Target, const long Value)
+{
+	/* NOTE: __sync_lock_test_and_set would be an acquire barrier, so we force a full barrier */
+	__sync_synchronize();
+	return __sync_lock_test_and_set(Target, Value);
+}
+
+static __inline__ __attribute__((always_inline)) void * _InterlockedExchangePointer(void * volatile * const Target, void * const Value)
+{
+	/* NOTE: ditto */
+	__sync_synchronize();
+	return __sync_lock_test_and_set(Target, Value);
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedExchangeAdd(volatile long * const Addend, const long Value)
+{
+	return __sync_fetch_and_add(Addend, Value);
+}
+
+static __inline__ __attribute__((always_inline)) char _InterlockedAnd8(volatile char * const value, const char mask)
+{
+	return __sync_fetch_and_and(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedAnd16(volatile short * const value, const short mask)
+{
+	return __sync_fetch_and_and(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedAnd(volatile long * const value, const long mask)
+{
+	return __sync_fetch_and_and(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) char _InterlockedOr8(volatile char * const value, const char mask)
+{
+	return __sync_fetch_and_or(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedOr16(volatile short * const value, const short mask)
+{
+	return __sync_fetch_and_or(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedOr(volatile long * const value, const long mask)
+{
+	return __sync_fetch_and_or(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) char _InterlockedXor8(volatile char * const value, const char mask)
+{
+	return __sync_fetch_and_xor(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedXor16(volatile short * const value, const short mask)
+{
+	return __sync_fetch_and_xor(value, mask);
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedXor(volatile long * const value, const long mask)
+{
+	return __sync_fetch_and_xor(value, mask);
+}
+
+#else
+
+static __inline__ __attribute__((always_inline)) char _InterlockedCompareExchange8(volatile char * const Destination, const char Exchange, const char Comperand)
+{
+	char retval = Comperand;
+	__asm__("lock; cmpxchgb %b[Exchange], %[Destination]" : [retval] "+a" (retval) : [Destination] "m" (*Destination), [Exchange] "q" (Exchange) : "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedCompareExchange16(volatile short * const Destination, const short Exchange, const short Comperand)
+{
+	short retval = Comperand;
+	__asm__("lock; cmpxchgw %w[Exchange], %[Destination]" : [retval] "+a" (retval) : [Destination] "m" (*Destination), [Exchange] "q" (Exchange): "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedCompareExchange(volatile long * const Destination, const long Exchange, const long Comperand)
+{
+	long retval = Comperand;
+	__asm__("lock; cmpxchgl %k[Exchange], %[Destination]" : [retval] "+a" (retval) : [Destination] "m" (*Destination), [Exchange] "q" (Exchange): "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) long long _InterlockedCompareExchange64(volatile long long * const Destination, const long long Exchange, const long long Comperand)
+{
+	long long retval = Comperand;
+
+	__asm__
+	(
+		"cmpxchg8b %[Destination]" :
+		[retval] "+A" (retval) :
+			[Destination] "m" (*Destination),
+			"b" ((unsigned long)((Exchange >>  0) & 0xFFFFFFFF)),
+			"c" ((unsigned long)((Exchange >> 32) & 0xFFFFFFFF)) :
+		"memory"
+	);
+
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) void * _InterlockedCompareExchangePointer(void * volatile * const Destination, void * const Exchange, void * const Comperand)
+{
+	void * retval = (void *)Comperand;
+	__asm__("lock; cmpxchgl %k[Exchange], %[Destination]" : [retval] "=a" (retval) : "[retval]" (retval), [Destination] "m" (*Destination), [Exchange] "q" (Exchange) : "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedExchange(volatile long * const Target, const long Value)
+{
+	long retval = Value;
+	__asm__("xchgl %[retval], %[Target]" : [retval] "+r" (retval) : [Target] "m" (*Target) : "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) void * _InterlockedExchangePointer(void * volatile * const Target, void * const Value)
+{
+	void * retval = Value;
+	__asm__("xchgl %[retval], %[Target]" : [retval] "+r" (retval) : [Target] "m" (*Target) : "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedExchangeAdd(volatile long * const Addend, const long Value)
+{
+	long retval = Value;
+	__asm__("lock; xaddl %[retval], %[Addend]" : [retval] "+r" (retval) : [Addend] "m" (*Addend) : "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) char _InterlockedAnd8(volatile char * const value, const char mask)
+{
+	char x;
+	char y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange8(value, x & mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedAnd16(volatile short * const value, const short mask)
+{
+	short x;
+	short y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange16(value, x & mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedAnd(volatile long * const value, const long mask)
+{
+	long x;
+	long y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange(value, x & mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) char _InterlockedOr8(volatile char * const value, const char mask)
+{
+	char x;
+	char y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange8(value, x | mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedOr16(volatile short * const value, const short mask)
+{
+	short x;
+	short y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange16(value, x | mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedOr(volatile long * const value, const long mask)
+{
+	long x;
+	long y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange(value, x | mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) char _InterlockedXor8(volatile char * const value, const char mask)
+{
+	char x;
+	char y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange8(value, x ^ mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) short _InterlockedXor16(volatile short * const value, const short mask)
+{
+	short x;
+	short y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange16(value, x ^ mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedXor(volatile long * const value, const long mask)
+{
+	long x;
+	long y;
+
+	y = *value;
+
+	do
+	{
+		x = y;
+		y = _InterlockedCompareExchange(value, x ^ mask, x);
+	}
+	while(y != x);
+
+	return y;
+}
+
+#endif
+
+static __inline__ __attribute__((always_inline)) long _InterlockedAddLargeStatistic(volatile long long * const Addend, const long Value)
+{
+	__asm__
+	(
+		"lock; add %[Value], %[Lo32];"
+		"jae LABEL%=;"
+		"lock; adc $0, %[Hi32];"
+		"LABEL%=:;" :
+		[Lo32] "=m" (*((volatile long *)(Addend) + 0)), [Hi32] "=m" (*((volatile long *)(Addend) + 1)) :
+		[Value] "ir" (Value)
+	);
+
+	return Value;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedDecrement(volatile long * const lpAddend)
+{
+	return _InterlockedExchangeAdd(lpAddend, -1) - 1;
+}
+
+static __inline__ __attribute__((always_inline)) long _InterlockedIncrement(volatile long * const lpAddend)
+{
+	return _InterlockedExchangeAdd(lpAddend, 1) + 1;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char _interlockedbittestandreset(volatile long * a, const long b)
+{
+	unsigned char retval;
+	__asm__("lock; btrl %[b], %[a]; setb %b[retval]" : [retval] "=r" (retval), [a] "=m" (a) : [b] "Ir" (b) : "memory");
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char _interlockedbittestandset(volatile long * a, const long b)
+{
+	unsigned char retval;
+	__asm__("lock; btsl %[b], %[a]; setc %b[retval]" : [retval] "=r" (retval), [a] "=m" (a) : [b] "Ir" (b) : "memory");
+	return retval;
+}
+
+
+/*** String operations ***/
+/* NOTE: we don't set a memory clobber in the __stosX functions because Visual C++ doesn't */
+static __inline__ __attribute__((always_inline)) void __stosb(unsigned char * Dest, const unsigned char Data, size_t Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; stosb" :
+		[Dest] "=D" (Dest), [Count] "=c" (Count) :
+		"[Dest]" (Dest), "a" (Data), "[Count]" (Count)
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __stosw(unsigned short * Dest, const unsigned short Data, size_t Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; stosw" :
+		[Dest] "=D" (Dest), [Count] "=c" (Count) :
+		"[Dest]" (Dest), "a" (Data), "[Count]" (Count)
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __stosd(unsigned long * Dest, const unsigned long Data, size_t Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; stosl" :
+		[Dest] "=D" (Dest), [Count] "=c" (Count) :
+		"[Dest]" (Dest), "a" (Data), "[Count]" (Count)
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __movsb(unsigned char * Destination, const unsigned char * Source, size_t Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; movsb" :
+		[Destination] "=D" (Destination), [Source] "=S" (Source), [Count] "=c" (Count) :
+		"[Destination]" (Destination), "[Source]" (Source), "[Count]" (Count)
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __movsw(unsigned short * Destination, const unsigned short * Source, size_t Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; movsw" :
+		[Destination] "=D" (Destination), [Source] "=S" (Source), [Count] "=c" (Count) :
+		"[Destination]" (Destination), "[Source]" (Source), "[Count]" (Count)
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __movsd(unsigned long * Destination, const unsigned long * Source, size_t Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; movsd" :
+		[Destination] "=D" (Destination), [Source] "=S" (Source), [Count] "=c" (Count) :
+		"[Destination]" (Destination), "[Source]" (Source), "[Count]" (Count)
+	);
+}
+
+
+/*** FS segment addressing ***/
+static __inline__ __attribute__((always_inline)) void __writefsbyte(const unsigned long Offset, const unsigned char Data)
+{
+	__asm__("movb %b[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) void __writefsword(const unsigned long Offset, const unsigned short Data)
+{
+	__asm__("movw %w[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) void __writefsdword(const unsigned long Offset, const unsigned long Data)
+{
+	__asm__("movl %k[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char __readfsbyte(const unsigned long Offset)
+{
+	unsigned char value;
+	__asm__("movb %%fs:%a[Offset], %b[value]" : [value] "=q" (value) : [Offset] "irm" (Offset));
+	return value;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned short __readfsword(const unsigned long Offset)
+{
+	unsigned short value;
+	__asm__("movw %%fs:%a[Offset], %w[value]" : [value] "=q" (value) : [Offset] "irm" (Offset));
+	return value;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long __readfsdword(const unsigned long Offset)
+{
+	unsigned long value;
+	__asm__("movl %%fs:%a[Offset], %k[value]" : [value] "=q" (value) : [Offset] "irm" (Offset));
+	return value;
+}
+
+static __inline__ __attribute__((always_inline)) void __incfsbyte(const unsigned long Offset)
+{
+	__asm__("incb %%fs:%a[Offset]" : : [Offset] "ir" (Offset));
+}
+
+static __inline__ __attribute__((always_inline)) void __incfsword(const unsigned long Offset)
+{
+	__asm__("incw %%fs:%a[Offset]" : : [Offset] "ir" (Offset));
+}
+
+static __inline__ __attribute__((always_inline)) void __incfsdword(const unsigned long Offset)
+{
+	__asm__("incl %%fs:%a[Offset]" : : [Offset] "ir" (Offset));
+}
+
+/* NOTE: the bizarre implementation of __addfsxxx mimics the broken Visual C++ behavior */
+static __inline__ __attribute__((always_inline)) void __addfsbyte(const unsigned long Offset, const unsigned char Data)
+{
+	if(!__builtin_constant_p(Offset))
+		__asm__("addb %k[Offset], %%fs:%a[Offset]" : : [Offset] "r" (Offset));
+	else
+		__asm__("addb %b[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) void __addfsword(const unsigned long Offset, const unsigned short Data)
+{
+	if(!__builtin_constant_p(Offset))
+		__asm__("addw %k[Offset], %%fs:%a[Offset]" : : [Offset] "r" (Offset));
+	else
+		__asm__("addw %w[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) void __addfsdword(const unsigned long Offset, const unsigned int Data)
+{
+	if(!__builtin_constant_p(Offset))
+		__asm__("addl %k[Offset], %%fs:%a[Offset]" : : [Offset] "r" (Offset));
+	else
+		__asm__("addl %k[Data], %%fs:%a[Offset]" : : [Offset] "ir" (Offset), [Data] "iq" (Data));
+}
+
+
+/*** Bit manipulation ***/
+static  __inline__ __attribute__((always_inline)) unsigned char _BitScanForward(unsigned long * const Index, const unsigned long Mask)
+{
+	__asm__("bsfl %[Mask], %[Index]" : [Index] "=r" (*Index) : [Mask] "mr" (Mask));
+	return Mask ? 1 : 0;
+}
+
+static  __inline__ __attribute__((always_inline)) unsigned char _BitScanReverse(unsigned long * const Index, const unsigned long Mask)
+{
+	__asm__("bsrl %[Mask], %[Index]" : [Index] "=r" (*Index) : [Mask] "mr" (Mask));
+	return Mask ? 1 : 0;
+}
+
+/* NOTE: again, the bizarre implementation follows Visual C++ */
+static  __inline__ __attribute__((always_inline)) unsigned char _bittest(const long * const a, const long b)
+{
+	unsigned char retval;
+
+	if(__builtin_constant_p(b))
+		__asm__("bt %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32));
+	else
+		__asm__("bt %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b));
+
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char _bittestandcomplement(long * const a, const long b)
+{
+	unsigned char retval;
+
+	if(__builtin_constant_p(b))
+		__asm__("btc %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32));
+	else
+		__asm__("btc %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b));
+
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char _bittestandreset(long * const a, const long b)
+{
+	unsigned char retval;
+
+	if(__builtin_constant_p(b))
+		__asm__("btr %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32));
+	else
+		__asm__("btr %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b));
+
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char _bittestandset(long * const a, const long b)
+{
+	unsigned char retval;
+
+	if(__builtin_constant_p(b))
+		__asm__("bts %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*(a + (b / 32))), [b] "Ir" (b % 32));
+	else
+		__asm__("bts %[b], %[a]; setb %b[retval]" : [retval] "=q" (retval) : [a] "mr" (*a), [b] "r" (b));
+
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char _rotl8(const unsigned char value, const unsigned char shift)
+{
+	unsigned char retval;
+	__asm__("rolb %b[shift], %b[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift));
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned short _rotl16(const unsigned short value, const unsigned char shift)
+{
+	unsigned short retval;
+	__asm__("rolw %b[shift], %w[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift));
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned char _rotr8(const unsigned char value, const unsigned char shift)
+{
+	unsigned char retval;
+	__asm__("rorb %b[shift], %b[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift));
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned short _rotr16(const unsigned short value, const unsigned char shift)
+{
+	unsigned short retval;
+	__asm__("rorw %b[shift], %w[retval]" : [retval] "=rm" (retval) : "[retval]" (value), [shift] "Nc" (shift));
+	return retval;
+}
+
+/*
+	NOTE: in __ll_lshift, __ll_rshift and __ull_rshift we use the "A"
+	constraint (edx:eax) for the Mask argument, because it's the only way GCC
+	can pass 64-bit operands around - passing the two 32 bit parts separately
+	just confuses it. Also we declare Bit as an int and then truncate it to
+	match Visual C++ behavior
+*/
+static __inline__ __attribute__((always_inline)) unsigned long long __ll_lshift(const unsigned long long Mask, const int Bit)
+{
+	unsigned long long retval = Mask;
+
+	__asm__
+	(
+		"shldl %b[Bit], %%eax, %%edx; sall %b[Bit], %%eax" :
+		"+A" (retval) :
+		[Bit] "Nc" ((unsigned char)((unsigned long)Bit) & 0xFF)
+	);
+
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) long long __ll_rshift(const long long Mask, const int Bit)
+{
+	unsigned long long retval = Mask;
+
+	__asm__
+	(
+		"shldl %b[Bit], %%eax, %%edx; sarl %b[Bit], %%eax" :
+		"+A" (retval) :
+		[Bit] "Nc" ((unsigned char)((unsigned long)Bit) & 0xFF)
+	);
+
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long long __ull_rshift(const unsigned long long Mask, int Bit)
+{
+	unsigned long long retval = Mask;
+
+	__asm__
+	(
+		"shrdl %b[Bit], %%eax, %%edx; shrl %b[Bit], %%eax" :
+		"+A" (retval) :
+		[Bit] "Nc" ((unsigned char)((unsigned long)Bit) & 0xFF)
+	);
+
+	return retval;
+}
+
+
+/*** 64-bit math ***/
+static __inline__ __attribute__((always_inline)) long long __emul(const int a, const int b)
+{
+	long long retval;
+	__asm__("imull %[b]" : "=A" (retval) : [a] "a" (a), [b] "rm" (b));
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long long __emulu(const unsigned int a, const unsigned int b)
+{
+	unsigned long long retval;
+	__asm__("mull %[b]" : "=A" (retval) : [a] "a" (a), [b] "rm" (b));
+	return retval;
+}
+
+
+/*** Port I/O ***/
+static __inline__ __attribute__((always_inline)) unsigned char __inbyte(const unsigned short Port)
+{
+	unsigned char byte;
+	__asm__ __volatile__("inb %w[Port], %b[byte]" : [byte] "=a" (byte) : [Port] "Nd" (Port));
+	return byte;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned short __inword(const unsigned short Port)
+{
+	unsigned short word;
+	__asm__ __volatile__("inw %w[Port], %w[word]" : [word] "=a" (word) : [Port] "Nd" (Port));
+	return word;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long __indword(const unsigned short Port)
+{
+	unsigned long dword;
+	__asm__ __volatile__("inl %w[Port], %k[dword]" : [dword] "=a" (dword) : [Port] "Nd" (Port));
+	return dword;
+}
+
+static __inline__ __attribute__((always_inline)) void __inbytestring(unsigned short Port, unsigned char * Buffer, unsigned long Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; insb" :
+		[Buffer] "=D" (Buffer), [Count] "=c" (Count) :
+		"d" (Port), "[Buffer]" (Buffer), "[Count]" (Count) :
+		"memory"
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __inwordstring(unsigned short Port, unsigned short * Buffer, unsigned long Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; insw" :
+		[Buffer] "=D" (Buffer), [Count] "=c" (Count) :
+		"d" (Port), "[Buffer]" (Buffer), "[Count]" (Count) :
+		"memory"
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __indwordstring(unsigned short Port, unsigned long * Buffer, unsigned long Count)
+{
+	__asm__ __volatile__
+	(
+		"rep; insl" :
+		[Buffer] "=D" (Buffer), [Count] "=c" (Count) :
+		"d" (Port), "[Buffer]" (Buffer), "[Count]" (Count) :
+		"memory"
+	);
+}
+
+static __inline__ __attribute__((always_inline)) void __outbyte(unsigned short const Port, const unsigned char Data)
+{
+	__asm__ __volatile__("outb %b[Data], %w[Port]" : : [Port] "Nd" (Port), [Data] "a" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) void __outword(unsigned short const Port, const unsigned short Data)
+{
+	__asm__ __volatile__("outw %w[Data], %w[Port]" : : [Port] "Nd" (Port), [Data] "a" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) void __outdword(unsigned short const Port, const unsigned long Data)
+{
+	__asm__ __volatile__("outl %k[Data], %w[Port]" : : [Port] "Nd" (Port), [Data] "a" (Data));
+}
+
+static __inline__ __attribute__((always_inline)) void __outbytestring(unsigned short const Port, const unsigned char * const Buffer, const unsigned long Count)
+{
+	__asm__ __volatile__("rep; outsb" : : [Port] "d" (Port), [Buffer] "S" (Buffer), "c" (Count));
+}
+
+static __inline__ __attribute__((always_inline)) void __outwordstring(unsigned short const Port, const unsigned short * const Buffer, const unsigned long Count)
+{
+	__asm__ __volatile__("rep; outsw" : : [Port] "d" (Port), [Buffer] "S" (Buffer), "c" (Count));
+}
+
+static __inline__ __attribute__((always_inline)) void __outdwordstring(unsigned short const Port, const unsigned long * const Buffer, const unsigned long Count)
+{
+	__asm__ __volatile__("rep; outsl" : : [Port] "d" (Port), [Buffer] "S" (Buffer), "c" (Count));
+}
+
+
+/*** System information ***/
+static __inline__ __attribute__((always_inline)) void __cpuid(int CPUInfo[], const int InfoType)
+{
+	__asm__ __volatile__("cpuid" : "=a" (CPUInfo[0]), "=b" (CPUInfo[1]), "=c" (CPUInfo[2]), "=d" (CPUInfo[3]) : "a" (InfoType));
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long long __rdtsc(void)
+{
+	unsigned long long retval;
+	__asm__ __volatile__("rdtsc" : "=A"(retval));
+	return retval;
+}
+
+
+/*** Interrupts ***/
+static __inline__ __attribute__((always_inline)) void __debugbreak(void)
+{
+	__asm__("int $3");
+}
+
+static __inline__ __attribute__((always_inline)) void __int2c(void)
+{
+	__asm__("int $0x2c");
+}
+
+static __inline__ __attribute__((always_inline)) void _disable(void)
+{
+	__asm__("cli");
+}
+
+static __inline__ __attribute__((always_inline)) void _enable(void)
+{
+	__asm__("sti");
+}
+
+
+/*** Protected memory management ***/
+static __inline__ __attribute__((always_inline)) unsigned long __readcr0(void)
+{
+	unsigned long value;
+	__asm__ __volatile__("mov %%cr0, %[value]" : [value] "=q" (value));
+	return value;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long __readcr2(void)
+{
+	unsigned long value;
+	__asm__ __volatile__("mov %%cr2, %[value]" : [value] "=q" (value));
+	return value;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long __readcr3(void)
+{
+	unsigned long value;
+	__asm__ __volatile__("mov %%cr3, %[value]" : [value] "=q" (value));
+	return value;
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long __readcr4(void)
+{
+	unsigned long value;
+	__asm__ __volatile__("mov %%cr4, %[value]" : [value] "=q" (value));
+	return value;
+}
+
+static __inline__ __attribute__((always_inline)) void __writecr0(const unsigned long long Data)
+{
+	__asm__("mov %[Data], %%cr0" : : [Data] "q" ((const unsigned long)(Data & 0xFFFFFFFF)) : "memory");
+}
+
+static __inline__ __attribute__((always_inline)) void __writecr3(const unsigned long long Data)
+{
+	__asm__("mov %[Data], %%cr3" : : [Data] "q" ((const unsigned long)(Data & 0xFFFFFFFF)) : "memory");
+}
+
+static __inline__ __attribute__((always_inline)) void __writecr4(const unsigned long long Data)
+{
+	__asm__("mov %[Data], %%cr4" : : [Data] "q" ((const unsigned long)(Data & 0xFFFFFFFF)) : "memory");
+}
+
+static __inline__ __attribute__((always_inline)) void __invlpg(void * const Address)
+{
+	__asm__("invlpg %[Address]" : : [Address] "m" (*((unsigned char *)(Address))));
+}
+
+
+/*** System operations ***/
+static __inline__ __attribute__((always_inline)) unsigned long long __readmsr(const int reg)
+{
+	unsigned long long retval;
+	__asm__ __volatile__("rdmsr" : "=A" (retval) : "c" (reg));
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) void __writemsr(const unsigned long Register, const unsigned long long Value)
+{
+	__asm__ __volatile__("wrmsr" : : "A" (Value), "c" (Register));
+}
+
+static __inline__ __attribute__((always_inline)) unsigned long long __readpmc(const int counter)
+{
+	unsigned long long retval;
+	__asm__ __volatile__("rdpmc" : "=A" (retval) : "c" (counter));
+	return retval;
+}
+
+/* NOTE: an immediate value for 'a' will raise an ICE in Visual C++ */
+static __inline__ __attribute__((always_inline)) unsigned long __segmentlimit(const unsigned long a)
+{
+	unsigned long retval;
+	__asm__ __volatile__("lsl %[a], %[retval]" : [retval] "=r" (retval) : [a] "rm" (a));
+	return retval;
+}
+
+static __inline__ __attribute__((always_inline)) void __wbinvd(void)
+{
+	__asm__ __volatile__("wbinvd");
+}
+
+#endif /* KJK_INTRIN_X86_H_ */
+
+/* EOF */

common/include/x86emitter/implement/group1.h

@@ -47,22 +47,6 @@ public:
 		EmitSibMagic( from, to );
 	}
 
-	// ------------------------------------------------------------------------
-	template< typename T > __forceinline void operator()( const xRegister<T>& to, const void* src ) const
-	{
-		prefix16<T>();
-		xWrite8( (Is8BitOp<T>() ? 2 : 3) | (InstType<<3) );
-		EmitSibMagic( to, src );
-	}
-	
-	// ------------------------------------------------------------------------
-	template< typename T > __forceinline void operator()( void* dest, const xRegister<T>& from ) const
-	{
-		prefix16<T>();
-		xWrite8( (Is8BitOp<T>() ? 0 : 1) | (InstType<<3) ); 
-		EmitSibMagic( from, dest );
-	}
-
 	// ------------------------------------------------------------------------
 	template< typename T > __noinline void operator()( const ModSibBase& sibdest, const xRegister<T>& from ) const
 	{
@@ -127,6 +111,37 @@ public:
 		}
 	}
 	
+	// ------------------------------------------------------------------------
+	template< typename T > __noinline void operator()( const ModSibBase& to, const xImmReg<T>& immOrReg ) const
+	{
+		_DoI_helpermess( *this, to, immOrReg );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, const xImmReg<T>& immOrReg ) const
+	{
+		_DoI_helpermess( *this, to, immOrReg );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, int imm ) const
+	{
+		_DoI_helpermess( *this, to, imm );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, const xDirectOrIndirect<T>& from ) const
+	{
+		_DoI_helpermess( *this, to, from );
+	}
+
+	template< typename T > __noinline void operator()( const xRegister<T>& to, const xDirectOrIndirect<T>& from ) const
+	{
+		_DoI_helpermess( *this, xDirectOrIndirect<T>( to ), from );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, const xRegister<T>& from ) const
+	{
+		_DoI_helpermess( *this, to, xDirectOrIndirect<T>( from ) );
+	}
+
 	xImpl_Group1() {}		// Why does GCC need these?
 };
 
@@ -168,7 +183,6 @@ protected:
 	template< u8 Prefix > struct Woot
 	{
 		__forceinline void operator()( const xRegisterSSE& to, const xRegisterSSE& from, SSE2_ComparisonType cmptype ) const{ xOpWrite0F( Prefix, 0xc2, to, from, (u8)cmptype ); }
-		__forceinline void operator()( const xRegisterSSE& to, const void* from, SSE2_ComparisonType cmptype ) const		{ xOpWrite0F( Prefix, 0xc2, to, from, (u8)cmptype ); }
 		__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from, SSE2_ComparisonType cmptype ) const	{ xOpWrite0F( Prefix, 0xc2, to, from, (u8)cmptype ); }
 		Woot() {}
 	};

common/include/x86emitter/implement/group2.h

@@ -94,5 +94,17 @@ public:
 		}
 	}
 
+	// ------------------------------------------------------------------------
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, u8 imm ) const
+	{
+		_DoI_helpermess( *this, to, imm );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, const xRegisterCL& from ) const
+	{
+		_DoI_helpermess( *this, to, from );
+	}
+
+
 	Group2ImplAll() {}		// I am a class with no members, so I need an explicit constructor!  Sense abounds.
 };

common/include/x86emitter/implement/group3.h

@@ -53,6 +53,10 @@ public:
 		EmitSibMagic( InstType, from );
 	}
 
+	template< typename T > __emitinline void operator()( const xDirectOrIndirect<T>& from ) const
+	{
+		_DoI_helpermess( *this, from );
+	}
 	xImpl_Group3() {}
 };
 
@@ -97,13 +101,11 @@ public:
 	using ImplMulDivBase<G3Type_iMUL,0x59>::operator();
 	
 	__forceinline void operator()( const xRegister32& to,	const xRegister32& from ) const			{ xOpWrite0F( 0xaf, to, from ); }
-	__forceinline void operator()( const xRegister32& to,	const void* src ) const					{ xOpWrite0F( 0xaf, to, src ); }
 	__forceinline void operator()( const xRegister32& to,	const ModSibBase& src ) const			{ xOpWrite0F( 0xaf, to, src ); }
 	__forceinline void operator()( const xRegister32& to,	const xRegister32& from, s32 imm ) const{ ImmStyle( to, from, imm ); }
 	__forceinline void operator()( const xRegister32& to,	const ModSibBase& from, s32 imm ) const	{ ImmStyle( to, from, imm ); }
 
 	__forceinline void operator()( const xRegister16& to,	const xRegister16& from ) const			{ xOpWrite0F( 0x66, 0xaf, to, from ); }
-	__forceinline void operator()( const xRegister16& to,	const void* src ) const					{ xOpWrite0F( 0x66, 0xaf, to, src ); }
 	__forceinline void operator()( const xRegister16& to,	const ModSibBase& src ) const			{ xOpWrite0F( 0x66, 0xaf, to, src ); }
 	__forceinline void operator()( const xRegister16& to,	const xRegister16& from, s16 imm ) const{ ImmStyle( to, from, imm ); }
 	__forceinline void operator()( const xRegister16& to,	const ModSibBase& from, s16 imm ) const	{ ImmStyle( to, from, imm ); }

common/include/x86emitter/implement/helpers.h

@@ -0,0 +1,157 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *  
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *  
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+// ------------------------------------------------------------------------
+// helpers.h -- Various universally helpful functions for emitter convenience!
+//
+// Note: Header file should be included from the x86Emitter::Internal namespace, such
+// that all members contained within are in said namespace.
+// ------------------------------------------------------------------------
+
+
+#pragma once
+
+extern void SimdPrefix( u8 prefix, u16 opcode );
+extern void EmitSibMagic( uint regfield, const void* address );
+extern void EmitSibMagic( uint regfield, const ModSibBase& info );
+extern void xJccKnownTarget( JccComparisonType comparison, const void* target, bool slideForward );
+
+template< typename T > bool Is8BitOp() { return sizeof(T) == 1; }
+template< typename T > void prefix16() { if( sizeof(T) == 2 ) xWrite8( 0x66 ); }
+
+
+// Writes a ModRM byte for "Direct" register access forms, which is used for all
+// instructions taking a form of [reg,reg].
+template< typename T > __emitinline
+void EmitSibMagic( uint reg1, const xRegisterBase<T>& reg2 )
+{
+	xWrite8( (Mod_Direct << 6) | (reg1 << 3) | reg2.Id );
+}
+
+template< typename T1, typename T2 > __emitinline
+void EmitSibMagic( const xRegisterBase<T1> reg1, const xRegisterBase<T2>& reg2 )
+{
+	xWrite8( (Mod_Direct << 6) | (reg1.Id << 3) | reg2.Id );
+}
+
+template< typename T1 > __emitinline
+void EmitSibMagic( const xRegisterBase<T1> reg1, const void* src )			{ EmitSibMagic( reg1.Id, src ); }
+
+template< typename T1 > __emitinline
+void EmitSibMagic( const xRegisterBase<T1> reg1, const ModSibBase& sib )	{ EmitSibMagic( reg1.Id, sib ); }
+
+// ------------------------------------------------------------------------
+template< typename T1, typename T2 > __emitinline
+void xOpWrite( u8 prefix, u8 opcode, const T1& param1, const T2& param2 )
+{
+	if( prefix != 0 )
+		xWrite16( (opcode<<8) | prefix );
+	else
+		xWrite8( opcode );
+
+	EmitSibMagic( param1, param2 );
+}
+
+// ------------------------------------------------------------------------
+template< typename T1, typename T2 > __emitinline
+void xOpWrite0F( u8 prefix, u16 opcode, const T1& param1, const T2& param2 )
+{
+	SimdPrefix( prefix, opcode );
+	EmitSibMagic( param1, param2 );
+}
+
+template< typename T1, typename T2 > __emitinline
+void xOpWrite0F( u8 prefix, u16 opcode, const T1& param1, const T2& param2, u8 imm8 )
+{
+	xOpWrite0F( prefix, opcode, param1, param2 );
+	xWrite8( imm8 );
+}
+
+template< typename T1, typename T2 > __emitinline
+void xOpWrite0F( u16 opcode, const T1& param1, const T2& param2 )			{ xOpWrite0F( 0, opcode, param1, param2 ); }
+
+template< typename T1, typename T2 > __emitinline
+void xOpWrite0F( u16 opcode, const T1& param1, const T2& param2, u8 imm8 )	{ xOpWrite0F( 0, opcode, param1, param2, imm8 ); }
+
+// ------------------------------------------------------------------------
+
+template< typename xImpl, typename T >
+void _DoI_helpermess( const xImpl& helpme, const xDirectOrIndirect<T>& to, const xImmReg<T>& immOrReg )
+{
+	if( to.IsDirect() )
+	{
+		if( immOrReg.IsReg() )
+			helpme( to.GetReg(), immOrReg.GetReg() );
+		else
+			helpme( to.GetReg(), immOrReg.GetImm() );
+	}
+	else
+	{
+		if( immOrReg.IsReg() )
+			helpme( to.GetMem(), immOrReg.GetReg() );
+		else
+			helpme( to.GetMem(), immOrReg.GetImm() );
+	}
+}
+
+template< typename xImpl, typename T >
+void _DoI_helpermess( const xImpl& helpme, const ModSibBase& to, const xImmReg<T>& immOrReg )
+{
+	if( immOrReg.IsReg() )
+		helpme( to, immOrReg.GetReg() );
+	else
+		helpme( ModSibStrict<T>(to), immOrReg.GetImm() );
+}
+
+template< typename xImpl, typename T >
+void _DoI_helpermess( const xImpl& helpme, const xDirectOrIndirect<T>& to, int imm )
+{
+	if( to.IsDirect() )
+		helpme( to.GetReg(), imm );
+	else
+		helpme( to.GetMem(), imm );
+}
+
+template< typename xImpl, typename T >
+void _DoI_helpermess( const xImpl& helpme, const xDirectOrIndirect<T>& parm )
+{
+	if( parm.IsDirect() )
+		helpme( parm.GetReg() );
+	else 
+		helpme( parm.GetMem() );
+}
+
+template< typename xImpl, typename T >
+void _DoI_helpermess( const xImpl& helpme, const xDirectOrIndirect<T>& to, const xDirectOrIndirect<T>& from )
+{
+	if( to.IsDirect() && from.IsDirect() )
+		helpme( to.GetReg(), from.GetReg() );
+
+	else if( to.IsDirect() )
+		helpme( to.GetReg(), from.GetMem() );
+
+	else if( from.IsDirect() )
+		helpme( to.GetMem(), from.GetReg() );
+
+	else
+
+		// One of the fields needs to be direct, or else we cannot complete the operation.
+		// (intel doesn't support indirects in both fields)
+
+		jASSUME( false );
+}

common/include/x86emitter/implement/movs.h

@@ -0,0 +1,239 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *  
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *  
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#pragma once
+
+// Header: ix86_impl_movs.h -- covers mov, cmov, movsx/movzx, and SETcc (which shares
+// with cmov many similarities).
+
+// Note: This header is meant to be included from within the x86Emitter::Internal namespace.
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// MOV instruction Implementation
+
+class MovImplAll
+{
+public:
+	// ------------------------------------------------------------------------
+	template< typename T > __forceinline void operator()( const xRegister<T>& to, const xRegister<T>& from ) const
+	{
+		if( to == from ) return;	// ignore redundant MOVs.
+
+		prefix16<T>();
+		xWrite8( Is8BitOp<T>() ? 0x88 : 0x89 );
+		EmitSibMagic( from, to );
+	}
+
+	// ------------------------------------------------------------------------
+	template< typename T > __noinline void operator()( const ModSibBase& dest, const xRegister<T>& from ) const
+	{
+		prefix16<T>();
+
+		// mov eax has a special from when writing directly to a DISP32 address
+		// (sans any register index/base registers).
+
+		if( from.IsAccumulator() && dest.Index.IsEmpty() && dest.Base.IsEmpty() )
+		{
+			xWrite8( Is8BitOp<T>() ? 0xa2 : 0xa3 );
+			xWrite32( dest.Displacement );
+		}
+		else
+		{
+			xWrite8( Is8BitOp<T>() ? 0x88 : 0x89 );
+			EmitSibMagic( from.Id, dest );
+		}
+	}
+
+	// ------------------------------------------------------------------------
+	template< typename T > __noinline void operator()( const xRegister<T>& to, const ModSibBase& src ) const
+	{
+		prefix16<T>();
+
+		// mov eax has a special from when reading directly from a DISP32 address
+		// (sans any register index/base registers).
+
+		if( to.IsAccumulator() && src.Index.IsEmpty() && src.Base.IsEmpty() )
+		{
+			xWrite8( Is8BitOp<T>() ? 0xa0 : 0xa1 );
+			xWrite32( src.Displacement );
+		}
+		else
+		{
+			xWrite8( Is8BitOp<T>() ? 0x8a : 0x8b );
+			EmitSibMagic( to, src );
+		}
+	}
+
+	// ------------------------------------------------------------------------
+	template< typename T > __noinline void operator()( const ModSibStrict<T>& dest, int imm ) const
+	{
+		prefix16<T>();
+		xWrite8( Is8BitOp<T>() ? 0xc6 : 0xc7 );
+		EmitSibMagic( 0, dest );
+		xWrite<T>( imm );
+	}
+
+	// ------------------------------------------------------------------------
+	// preserve_flags  - set to true to disable optimizations which could alter the state of
+	//   the flags (namely replacing mov reg,0 with xor).
+	template< typename T > __emitinline void operator()( const xRegister<T>& to, int imm, bool preserve_flags=false ) const
+	{
+		if( !preserve_flags && (imm == 0) )
+			xXOR( to, to );
+		else
+		{
+			// Note: MOV does not have (reg16/32,imm8) forms.
+
+			prefix16<T>();
+			xWrite8( (Is8BitOp<T>() ? 0xb0 : 0xb8) | to.Id ); 
+			xWrite<T>( imm );
+		}
+	}
+
+	// ------------------------------------------------------------------------
+	template< typename T > __noinline void operator()( const ModSibBase& to, const xImmReg<T>& immOrReg ) const
+	{
+		_DoI_helpermess( *this, to, immOrReg );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, const xImmReg<T>& immOrReg ) const
+	{
+		_DoI_helpermess( *this, to, immOrReg );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, int imm ) const
+	{
+		_DoI_helpermess( *this, to, imm );
+	}
+
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, const xDirectOrIndirect<T>& from ) const
+	{
+		if( to == from ) return;
+		_DoI_helpermess( *this, to, from );
+	}
+
+	template< typename T > __noinline void operator()( const xRegister<T>& to, const xDirectOrIndirect<T>& from ) const
+	{
+		_DoI_helpermess( *this, xDirectOrIndirect<T>( to ), from );
+	}
+	
+	template< typename T > __noinline void operator()( const xDirectOrIndirect<T>& to, const xRegister<T>& from ) const
+	{
+		_DoI_helpermess( *this, to, xDirectOrIndirect<T>( from ) );
+	}
+
+	MovImplAll() {} // Satisfy GCC's whims.
+};
+
+#define ccSane() jASSUME( ccType >= 0 && ccType <= 0x0f )
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// CMOV !!  [in all of it's disappointing lack-of glory]  .. and ..
+// SETcc!!  [more glory, less lack!]
+//
+// CMOV Disclaimer: Caution!  This instruction can look exciting and cool, until you
+// realize that it cannot load immediate values into registers. -_-
+//
+// I use explicit method declarations here instead of templates, in order to provide
+// *only* 32 and 16 bit register operand forms (8 bit registers are not valid in CMOV).
+//
+class CMovImplGeneric
+{
+public:
+	__forceinline void operator()( JccComparisonType ccType, const xRegister32& to, const xRegister32& from ) const		{ ccSane(); xOpWrite0F( 0x40 | ccType, to, from ); }
+	__noinline void operator()( JccComparisonType ccType, const xRegister32& to, const ModSibBase& sibsrc ) const		{ ccSane(); xOpWrite0F( 0x40 | ccType, to, sibsrc ); }
+	//__noinline void operator()( JccComparisonType ccType, const xDirectOrIndirect32& to, const xDirectOrIndirect32& from ) const { ccSane(); _DoI_helpermess( *this, to, from ); }	// too.. lazy.. to fix.
+
+	__forceinline void operator()( JccComparisonType ccType, const xRegister16& to, const xRegister16& from ) const		{ ccSane(); xOpWrite0F( 0x66, 0x40 | ccType, to, from ); }
+	__noinline void operator()( JccComparisonType ccType, const xRegister16& to, const ModSibBase& sibsrc ) const	{ ccSane(); xOpWrite0F( 0x66, 0x40 | ccType, to, sibsrc ); }
+	//__noinline void operator()( JccComparisonType ccType, const xDirectOrIndirect16& to, const xDirectOrIndirect16& from ) const { ccSane(); _DoI_helpermess( *this, to, from ); }
+
+	CMovImplGeneric() {}		// don't ask.
+};
+
+// ------------------------------------------------------------------------
+template< JccComparisonType ccType >
+class CMovImplAll
+{
+	static const u16 Opcode = 0x40 | ccType;
+
+public:
+	__forceinline void operator()( const xRegister32& to, const xRegister32& from ) const	{ ccSane(); xOpWrite0F( Opcode, to, from ); }
+	__noinline void operator()( const xRegister32& to, const ModSibBase& sibsrc ) const		{ ccSane(); xOpWrite0F( Opcode, to, sibsrc ); }
+	__noinline void operator()( const xDirectOrIndirect32& to, const xDirectOrIndirect32& from ) const { ccSane(); _DoI_helpermess( *this, to, from ); }
+
+	__forceinline void operator()( const xRegister16& to, const xRegister16& from ) const	{ ccSane(); xOpWrite0F( 0x66, Opcode, to, from ); }
+	__noinline void operator()( const xRegister16& to, const ModSibBase& sibsrc ) const		{ ccSane(); xOpWrite0F( 0x66, Opcode, to, sibsrc ); }
+	__noinline void operator()( const xDirectOrIndirect16& to, const xDirectOrIndirect16& from ) const { ccSane(); _DoI_helpermess( *this, to, from ); }
+
+	CMovImplAll() {}		// don't ask.
+};
+
+// ------------------------------------------------------------------------
+class SetImplGeneric
+{
+	// note: SETcc are 0x90, with 0 in the Reg field of ModRM.
+public:
+	__forceinline void operator()( JccComparisonType ccType, const xRegister8& to ) const		{ ccSane(); xOpWrite0F( 0x90 | ccType, 0, to ); }
+	__noinline void operator()( JccComparisonType ccType, const ModSibStrict<u8>& dest ) const	{ ccSane(); xOpWrite0F( 0x90 | ccType, 0, dest ); }
+
+	SetImplGeneric() {}		// if you do, ask GCC.
+};
+
+// ------------------------------------------------------------------------
+template< JccComparisonType ccType >
+class SetImplAll
+{
+	static const u16 Opcode = 0x90 | ccType;		// SETcc are 0x90 base opcode, with 0 in the Reg field of ModRM.
+
+public:
+	__forceinline void operator()( const xRegister8& to ) const			{ ccSane(); xOpWrite0F( Opcode, 0, to ); }
+	__noinline void operator()( const ModSibStrict<u8>& dest ) const	{ ccSane(); xOpWrite0F( Opcode, 0, dest ); }
+	__noinline void operator()( const xDirectOrIndirect8& dest ) const	{ ccSane(); _DoI_helpermess( *this, dest ); }
+	
+	SetImplAll() {}		// if you do, ask GCC.
+};
+
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// Mov with sign/zero extension implementations (movsx / movzx)
+//
+
+// ------------------------------------------------------------------------
+template< bool SignExtend >
+class MovExtendImplAll
+{
+protected:
+	static const u16 Opcode = 0xb6 | (SignExtend ? 8 : 0 );
+
+public:
+	__forceinline void operator()( const xRegister32& to, const xRegister16& from )	const		{ xOpWrite0F( Opcode+1, to, from ); }
+	__noinline void operator()( const xRegister32& to, const ModSibStrict<u16>& sibsrc ) const	{ xOpWrite0F( Opcode+1, to, sibsrc ); }
+	__noinline void operator()( const xRegister32& to, const xDirectOrIndirect16& src ) const	{ _DoI_helpermess( *this, to, src ); }
+	
+	__forceinline void operator()( const xRegister32& to, const xRegister8& from ) const		{ xOpWrite0F( Opcode, to, from ); }
+	__noinline void operator()( const xRegister32& to, const ModSibStrict<u8>& sibsrc ) const	{ xOpWrite0F( Opcode, to, sibsrc ); }
+	__noinline void operator()( const xRegister32& to, const xDirectOrIndirect8& src ) const	{ _DoI_helpermess( *this, to, src ); }
+
+	__forceinline void operator()( const xRegister16& to, const xRegister8& from ) const		{ xOpWrite0F( 0x66, Opcode, to, from ); }
+	__noinline void operator()( const xRegister16& to, const ModSibStrict<u8>& sibsrc ) const	{ xOpWrite0F( 0x66, Opcode, to, sibsrc ); }
+	__noinline void operator()( const xRegister16& to, const xDirectOrIndirect8& src ) const	{ _DoI_helpermess( *this, to, src ); }
+
+	MovExtendImplAll() {}		// don't ask.
+};
+

common/include/x86emitter/implement/test.h

@@ -86,8 +86,6 @@ public:
 
 	__forceinline void operator()( const xRegister32& to, const xRegister32& from ) const	{ xOpWrite0F( Opcode, to, from ); }
 	__forceinline void operator()( const xRegister16& to, const xRegister16& from ) const	{ xOpWrite0F( 0x66, Opcode, to, from ); }
-	__forceinline void operator()( const xRegister32& to, const void* src ) const			{ xOpWrite0F( Opcode, to, src ); }
-	__forceinline void operator()( const xRegister16& to, const void* src ) const			{ xOpWrite0F( 0x66, Opcode, to, src ); }
 	__forceinline void operator()( const xRegister32& to, const ModSibBase& sibsrc ) const	{ xOpWrite0F( Opcode, to, sibsrc ); }
 	__forceinline void operator()( const xRegister16& to, const ModSibBase& sibsrc ) const	{ xOpWrite0F( 0x66, Opcode, to, sibsrc ); }
 };
@@ -104,13 +102,9 @@ public:
 	__forceinline void operator()( const xRegister16& bitbase, const xRegister16& bitoffset ) const	{ xOpWrite0F( 0x66, RegFormOp, bitbase, bitoffset ); }
 	__forceinline void operator()( const ModSibBase& bitbase, const xRegister32& bitoffset ) const	{ xOpWrite0F( RegFormOp, bitoffset, bitbase ); }
 	__forceinline void operator()( const ModSibBase& bitbase, const xRegister16& bitoffset ) const	{ xOpWrite0F( 0x66, RegFormOp, bitoffset, bitbase ); }
-	__forceinline void operator()( const void* bitbase, const xRegister32& bitoffset ) const		{ xOpWrite0F( 0xab, bitoffset, bitbase ); }
-	__forceinline void operator()( const void* bitbase, const xRegister16& bitoffset ) const		{ xOpWrite0F( 0x66, RegFormOp, bitoffset, bitbase ); }
 
 	__forceinline void operator()( const ModSibStrict<u32>& bitbase, u8 bitoffset ) const			{ xOpWrite0F( 0xba, InstType, bitbase, bitoffset ); }
 	__forceinline void operator()( const ModSibStrict<u16>& bitbase, u8 bitoffset ) const			{ xOpWrite0F( 0x66, 0xba, InstType, bitbase, bitoffset ); }
-	__forceinline void operator()( const u32* bitbase, u8 bitoffset ) const							{ xOpWrite0F( 0xba, InstType, bitbase, bitoffset ); }
-	__forceinline void operator()( const u16* bitbase, u8 bitoffset ) const							{ xOpWrite0F( 0x66, 0xba, InstType, bitbase, bitoffset ); }
 	__forceinline void operator()( const xRegister<u32>& bitbase, u8 bitoffset ) const				{ xOpWrite0F( 0xba, InstType, bitbase, bitoffset ); }
 	__forceinline void operator()( const xRegister<u16>& bitbase, u8 bitoffset ) const				{ xOpWrite0F( 0x66, 0xba, InstType, bitbase, bitoffset ); }
 

common/include/x86emitter/implement/xmm/arithmetic.h

@@ -29,11 +29,9 @@ public:
 	_SimdShiftHelper() {}
 
 	__forceinline void operator()( const xRegisterSSE& to, const xRegisterSSE& from ) const { xOpWrite0F( 0x66, Opcode1, to, from ); }
-	__forceinline void operator()( const xRegisterSSE& to, const void* from ) const			{ xOpWrite0F( 0x66, Opcode1, to, from ); }
 	__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from ) const	{ xOpWrite0F( 0x66, Opcode1, to, from ); }
 
 	__forceinline void operator()( const xRegisterMMX& to, const xRegisterMMX& from ) const { xOpWrite0F( Opcode1, to, from ); }
-	__forceinline void operator()( const xRegisterMMX& to, const void* from ) const			{ xOpWrite0F( Opcode1, to, from ); }
 	__forceinline void operator()( const xRegisterMMX& to, const ModSibBase& from ) const	{ xOpWrite0F( Opcode1, to, from ); }
 
 

common/include/x86emitter/implement/xmm/basehelpers.h

@@ -30,7 +30,6 @@ class SimdImpl_DestRegSSE
 {
 public:
 	__forceinline void operator()( const xRegisterSSE& to, const xRegisterSSE& from ) const	{ xOpWrite0F( Prefix, Opcode, to, from ); }
-	__forceinline void operator()( const xRegisterSSE& to, const void* from ) const			{ xOpWrite0F( Prefix, Opcode, to, from ); }
 	__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from ) const	{ xOpWrite0F( Prefix, Opcode, to, from ); }
 
 	SimdImpl_DestRegSSE() {} //GCWho?
@@ -45,7 +44,6 @@ class SimdImpl_DestRegImmSSE
 {
 public:
 	__forceinline void operator()( const xRegisterSSE& to, const xRegisterSSE& from, u8 imm ) const	{ xOpWrite0F( Prefix, Opcode, to, from, imm ); }
-	__forceinline void operator()( const xRegisterSSE& to, const void* from, u8 imm ) const			{ xOpWrite0F( Prefix, Opcode, to, from, imm ); }
 	__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from, u8 imm ) const	{ xOpWrite0F( Prefix, Opcode, to, from, imm ); }
 
 	SimdImpl_DestRegImmSSE() {} //GCWho?
@@ -56,7 +54,6 @@ class SimdImpl_DestRegImmMMX
 {
 public:
 	__forceinline void operator()( const xRegisterMMX& to, const xRegisterMMX& from, u8 imm ) const	{ xOpWrite0F( Opcode, to, from, imm ); }
-	__forceinline void operator()( const xRegisterMMX& to, const void* from, u8 imm ) const			{ xOpWrite0F( Opcode, to, from, imm ); }
 	__forceinline void operator()( const xRegisterMMX& to, const ModSibBase& from, u8 imm ) const	{ xOpWrite0F( Opcode, to, from, imm ); }
 
 	SimdImpl_DestRegImmMMX() {} //GCWho?
@@ -71,11 +68,9 @@ class SimdImpl_DestRegEither
 {
 public:
 	__forceinline void operator()( const xRegisterSSE& to, const xRegisterSSE& from ) const	{ xOpWrite0F( Prefix, Opcode, to, from ); }
-	__forceinline void operator()( const xRegisterSSE& to, const void* from ) const			{ xOpWrite0F( Prefix, Opcode, to, from ); }
 	__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from ) const	{ xOpWrite0F( Prefix, Opcode, to, from ); }
 
 	__forceinline void operator()( const xRegisterMMX& to, const xRegisterMMX& from ) const	{ xOpWrite0F( Opcode, to, from ); }
-	__forceinline void operator()( const xRegisterMMX& to, const void* from ) const			{ xOpWrite0F( Opcode, to, from ); }
 	__forceinline void operator()( const xRegisterMMX& to, const ModSibBase& from ) const	{ xOpWrite0F( Opcode, to, from ); }
 
 	SimdImpl_DestRegEither() {} //GCWho?
@@ -95,7 +90,6 @@ class SimdImpl_DestRegStrict
 {
 public:
 	__forceinline void operator()( const DestRegType& to, const SrcRegType& from ) const					{ xOpWrite0F( Prefix, Opcode, to, from ); }
-	__forceinline void operator()( const DestRegType& to, const SrcOperandType* from ) const				{ xOpWrite0F( Prefix, Opcode, to, from ); }
 	__forceinline void operator()( const DestRegType& to, const ModSibStrict<SrcOperandType>& from ) const	{ xOpWrite0F( Prefix, Opcode, to, from ); }
 
 	SimdImpl_DestRegStrict() {} //GCWho?

common/include/x86emitter/implement/xmm/comparisons.h

@@ -42,7 +42,6 @@ protected:
 	template< u8 Prefix > struct Woot
 	{
 		__forceinline void operator()( const xRegisterSSE& to, const xRegisterSSE& from ) const	{ xOpWrite0F( Prefix, 0xc2, to, from ); xWrite8( CType ); }
-		__forceinline void operator()( const xRegisterSSE& to, const void* from ) const			{ xOpWrite0F( Prefix, 0xc2, to, from ); xWrite8( CType ); }
 		__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from ) const	{ xOpWrite0F( Prefix, 0xc2, to, from ); xWrite8( CType ); }
 		Woot() {}
 	};
@@ -55,6 +54,22 @@ public:
 	SimdImpl_Compare() {} //GCWhat?
 };
 
+//////////////////////////////////////////////////////////////////////////////////////////
+// Compare scalar floating point values and set EFLAGS (Ordered or Unordered)
+//
+template< bool Ordered >
+class SimdImpl_COMI
+{
+protected:
+	static const u16 OpcodeSSE = Ordered ? 0x2f : 0x2e;
+
+public:
+	const SimdImpl_DestRegSSE<0x00,OpcodeSSE> SS;
+	const SimdImpl_DestRegSSE<0x66,OpcodeSSE> SD;
+	
+	SimdImpl_COMI() {}
+};
+
 
 //////////////////////////////////////////////////////////////////////////////////////////
 //

common/include/x86emitter/implement/xmm/moremovs.h

@@ -30,8 +30,6 @@ protected:
 	struct Woot
 	{
 		Woot() {}
-		__forceinline void operator()( const xRegisterSSE& to, const void* from ) const			{ xOpWrite0F( Prefix, Opcode, to, from ); }
-		__forceinline void operator()( const void* to, const xRegisterSSE& from ) const			{ xOpWrite0F( Prefix, Opcode+1, from, to ); }
 		__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from ) const	{ xOpWrite0F( Prefix, Opcode, to, from ); }
 		__forceinline void operator()( const ModSibBase& to, const xRegisterSSE& from ) const	{ xOpWrite0F( Prefix, Opcode+1, from, to ); }
 	};
@@ -64,11 +62,12 @@ public:
 // This happens when using Mem,Reg or Reg,Mem forms where the address is simple displacement
 // which can be checked for alignment at runtime.
 // 
+	
 template< u8 Prefix, bool isAligned >
 class SimdImpl_MoveSSE
 {
-	static const uint OpcodeA = 0x28;		// Aligned [aps] form
+	static const u16 OpcodeA = 0x28;		// Aligned [aps] form
-	static const uint OpcodeU = 0x10;		// unaligned [ups] form
+	static const u16 OpcodeU = 0x10;		// unaligned [ups] form
 
 public:
 	SimdImpl_MoveSSE() {} //GCC.
@@ -78,21 +77,19 @@ public:
 		if( to != from ) xOpWrite0F( Prefix, OpcodeA, to, from );
 	}
 
-	__forceinline void operator()( const xRegisterSSE& to, const void* from ) const
-	{
-		xOpWrite0F( Prefix, (isAligned || ((uptr)from & 0x0f) == 0) ? OpcodeA : OpcodeU, to, from );
-	}
-
-	__forceinline void operator()( void* to, const xRegisterSSE& from ) const
-	{
-		xOpWrite0F( Prefix, (isAligned || ((uptr)to & 0x0f) == 0) ? OpcodeA+1 : OpcodeU+1, from, to );
-	}
-
 	__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from ) const
 	{
 		// ModSib form is aligned if it's displacement-only and the displacement is aligned:
 		bool isReallyAligned = isAligned || ( ((from.Displacement & 0x0f) == 0) && from.Index.IsEmpty() && from.Base.IsEmpty() );
-		xOpWrite0F( Prefix, isReallyAligned ? OpcodeA : OpcodeU, to, from );
+		u16 opcode;
+		
+		// See previous comment.
+		if (isReallyAligned) 
+			opcode = OpcodeA;
+		else 
+			opcode = OpcodeU;
+		
+		xOpWrite0F( Prefix, opcode, to, from );
 	}
 
 	__forceinline void operator()( const ModSibBase& to, const xRegisterSSE& from ) const
@@ -123,16 +120,6 @@ public:
 		if( to != from ) xOpWrite0F( PrefixA, Opcode, to, from );
 	}
 
-	__forceinline void operator()( const xRegisterSSE& to, const void* from ) const
-	{
-		xOpWrite0F( (isAligned || ((uptr)from & 0x0f) == 0) ? PrefixA : PrefixU, Opcode, to, from );
-	}
-
-	__forceinline void operator()( const void* to, const xRegisterSSE& from ) const
-	{
-		xOpWrite0F( (isAligned || ((uptr)from & 0x0f) == 0) ? PrefixA : PrefixU, Opcode_Alt, to, from );
-	}
-
 	__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from ) const
 	{
 		// ModSib form is aligned if it's displacement-only and the displacement is aligned:
@@ -148,18 +135,6 @@ public:
 	}
 };
 
-
-//////////////////////////////////////////////////////////////////////////////////////////
-//
-template< u8 AltPrefix, u16 OpcodeSSE >
-class SimdImpl_UcomI
-{
-public:
-	const SimdImpl_DestRegSSE<0x00,OpcodeSSE> SS;
-	const SimdImpl_DestRegSSE<AltPrefix,OpcodeSSE> SD;
-	SimdImpl_UcomI() {}
-};
-
 //////////////////////////////////////////////////////////////////////////////////////////
 // Blend - Conditional copying of values in src into dest.
 //

common/include/x86emitter/implement/xmm/shufflepack.h

@@ -27,7 +27,6 @@ protected:
 	template< u8 Prefix > struct Woot
 	{
 		__forceinline void operator()( const xRegisterSSE& to, const xRegisterSSE& from, u8 cmptype ) const	{ xOpWrite0F( Prefix, OpcodeSSE, to, from ); xWrite8( cmptype ); }
-		__forceinline void operator()( const xRegisterSSE& to, const void* from, u8 cmptype ) const			{ xOpWrite0F( Prefix, OpcodeSSE, to, from ); xWrite8( cmptype ); }
 		__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from, u8 cmptype ) const	{ xOpWrite0F( Prefix, OpcodeSSE, to, from ); xWrite8( cmptype ); }
 		Woot() {}
 	};
@@ -185,11 +184,6 @@ protected:
 			xOpWrite0F( 0x66, (Opcode<<8) | 0x3a, to, from, imm8 );
 		}
 
-		__forceinline void operator()( const xRegisterSSE& to, const void* from, u8 imm8 ) const
-		{
-			xOpWrite0F( 0x66, (Opcode<<8) | 0x3a, to, from, imm8 );
-		}
-
 		__forceinline void operator()( const xRegisterSSE& to, const ModSibBase& from, u8 imm8 ) const
 		{
 			xOpWrite0F( 0x66, (Opcode<<8) | 0x3a, to, from, imm8 );
@@ -201,11 +195,9 @@ public:
 
 	// Operation can be performed on either MMX or SSE src operands.
 	__forceinline void W( const xRegisterSSE& to, const xRegister32& from, u8 imm8 ) const	{ xOpWrite0F( 0x66, 0xc4, to, from, imm8 ); }
-	__forceinline void W( const xRegisterSSE& to, const void* from, u8 imm8 ) const			{ xOpWrite0F( 0x66, 0xc4, to, from, imm8 ); }
 	__forceinline void W( const xRegisterSSE& to, const ModSibBase& from, u8 imm8 ) const	{ xOpWrite0F( 0x66, 0xc4, to, from, imm8 ); }
 
 	__forceinline void W( const xRegisterMMX& to, const xRegister32& from, u8 imm8 ) const	{ xOpWrite0F( 0xc4, to, from, imm8 ); }
-	__forceinline void W( const xRegisterMMX& to, const void* from, u8 imm8 ) const			{ xOpWrite0F( 0xc4, to, from, imm8 ); }
 	__forceinline void W( const xRegisterMMX& to, const ModSibBase& from, u8 imm8 ) const	{ xOpWrite0F( 0xc4, to, from, imm8 ); }
 
 	// [SSE-4.1] 
@@ -235,11 +227,6 @@ protected:
 			xOpWrite0F( 0x66, (Opcode<<8) | 0x3a, to, from, imm8 );
 		}
 
-		__forceinline void operator()( void* dest, const xRegisterSSE& from, u8 imm8 ) const
-		{
-			xOpWrite0F( 0x66, (Opcode<<8) | 0x3a, from, dest, imm8 );
-		}
-
 		__forceinline void operator()( const ModSibBase& dest, const xRegisterSSE& from, u8 imm8 ) const
 		{
 			xOpWrite0F( 0x66, (Opcode<<8) | 0x3a, from, dest, imm8 );
@@ -258,7 +245,6 @@ public:
 	__forceinline void W( const xRegister32& to, const xRegisterSSE& from, u8 imm8 ) const		{ xOpWrite0F( 0x66, 0xc5, to, from, imm8 ); }
 	__forceinline void W( const xRegister32& to, const xRegisterMMX& from, u8 imm8 ) const		{ xOpWrite0F( 0xc5, to, from, imm8 ); }
 
-	__forceinline void W( void* dest, const xRegisterSSE& from, u8 imm8 ) const					{ xOpWrite0F( 0x66, 0x153a, from, dest, imm8 ); }
 	__forceinline void W( const ModSibBase& dest, const xRegisterSSE& from, u8 imm8 ) const		{ xOpWrite0F( 0x66, 0x153a, from, dest, imm8 ); }
 
 	// [SSE-4.1] Copies the byte element specified by imm8 from src to dest.  The upper bits

common/include/x86emitter/inlines.inl

@@ -46,35 +46,101 @@
 // global optimization fails to resolve the externals and junk.
 //   (score one for MSVC!)
 
-#include "System.h"
-
 namespace x86Emitter
 {
+	extern const char *const x86_regnames_gpr8[8];
+	extern const char *const x86_regnames_gpr16[8];
+	extern const char *const x86_regnames_gpr32[8];
+
+	extern const char *const x86_regnames_sse[8];
+	extern const char *const x86_regnames_mmx[8];
+
+	//////////////////////////////////////////////////////////////////////////////////////////
+	// Diagnostic -- returns a string representation of this register.
+	//
+	template< typename T >
+	const char* xGetRegName( const xRegister<T>& src )
+	{
+		if( src.IsEmpty() ) return "empty";
+		
+		switch( sizeof(T) )
+		{
+			case 1: return x86_regnames_gpr8[ src.Id ];
+			case 2: return x86_regnames_gpr16[ src.Id ];
+			case 4: return x86_regnames_gpr32[ src.Id ];
+			
+			jNO_DEFAULT
+		}
+
+		return "oops?";
+	}
+
+	template< typename T >
+	const char* xGetRegName( const xRegisterSIMD<T>& src )
+	{
+		if( src.IsEmpty() ) return "empty";
+		
+		switch( sizeof(T) )
+		{
+			case 8: return x86_regnames_mmx[ src.Id ];
+			case 16: return x86_regnames_sse[ src.Id ];
+			
+			jNO_DEFAULT
+		}
+
+		return "oops?";
+	}
+
 	//////////////////////////////////////////////////////////////////////////////////////////
 	// x86Register Method Implementations
 	//
 	__forceinline xAddressInfo xAddressReg::operator+( const xAddressReg& right ) const
 	{
+		jASSUME( Id != -1 );
 		return xAddressInfo( *this, right );
 	}
 
 	__forceinline xAddressInfo xAddressReg::operator+( const xAddressInfo& right ) const
 	{
+		jASSUME( Id != -1 );
 		return right + *this;
 	}
 
 	__forceinline xAddressInfo xAddressReg::operator+( s32 right ) const
 	{
+		jASSUME( Id != -1 );
 		return xAddressInfo( *this, right );
 	}
 
+	__forceinline xAddressInfo xAddressReg::operator+( const void* right ) const
+	{
+		jASSUME( Id != -1 );
+		return xAddressInfo( *this, (s32)right );
+	}
+
+	// ------------------------------------------------------------------------
+	__forceinline xAddressInfo xAddressReg::operator-( s32 right ) const
+	{
+		jASSUME( Id != -1 );
+		return xAddressInfo( *this, -right );
+	}
+
+	__forceinline xAddressInfo xAddressReg::operator-( const void* right ) const
+	{
+		jASSUME( Id != -1 );
+		return xAddressInfo( *this, -(s32)right );
+	}
+
+	// ------------------------------------------------------------------------
 	__forceinline xAddressInfo xAddressReg::operator*( u32 right ) const
 	{
+		jASSUME( Id != -1 );
 		return xAddressInfo( Empty, *this, right );
 	}
 
 	__forceinline xAddressInfo xAddressReg::operator<<( u32 shift ) const
 	{
+		jASSUME( Id != -1 );
 		return xAddressInfo( Empty, *this, 1<<shift );
 	}
 
@@ -112,6 +178,16 @@ namespace x86Emitter
 		// no reduction necessary :D
 	}
 
+	// ------------------------------------------------------------------------
+	__forceinline ModSibBase::ModSibBase( const void* target ) :
+		Base(),
+		Index(),
+		Scale(0),
+		Displacement( (s32)target )
+	{
+		// no reduction necessary :D
+	}
+
 	//////////////////////////////////////////////////////////////////////////////////////////
 	// xAddressInfo Method Implementations
 	//
@@ -211,7 +287,11 @@ namespace x86Emitter
 			if( !is_s8( displacement ) )
 			{
 				assert( false );
+// Don't ask. --arcum42
+#if !defined(__LINUX__) || !defined(DEBUG)
+
 				Console::Error( "Emitter Error: Invalid short jump displacement = 0x%x", params (int)displacement );
+#endif
 			}
 			BasePtr[-1] = (s8)displacement;
 		}
@@ -222,4 +302,15 @@ namespace x86Emitter
 		}
 	}
 
+	// ------------------------------------------------------------------------
+	// returns the inverted conditional type for this Jcc condition.  Ie, JNS will become JS.
+	//
+	static __forceinline JccComparisonType xInvertCond( JccComparisonType src )
+	{
+		jASSUME( src != Jcc_Unknown );
+		if( Jcc_Unconditional == src ) return Jcc_Unconditional;
+
+		// x86 conditionals are clever!  To invert conditional types, just invert the lower bit:
+		return (JccComparisonType)((int)src ^ 1);
+	}
 }

common/include/x86emitter/instructions.h

@@ -179,11 +179,9 @@ namespace x86Emitter
 	extern void xPUSH( const ModSibBase& from );
 
 	extern void xPOP( xRegister32 from );
-	extern void xPOP( void* from );
 
 	extern void xPUSH( u32 imm );
 	extern void xPUSH( xRegister32 from );
-	extern void xPUSH( void* from );
 
 	// pushes the EFLAGS register onto the stack
 	extern void xPUSHFD();
@@ -212,6 +210,8 @@ namespace x86Emitter
 	// JMP / Jcc Instructions!
 
 	extern void xJcc( JccComparisonType comparison, const void* target );
+	extern s8* xJcc8( JccComparisonType comparison=Jcc_Unconditional, s8 displacement=0 );
+	extern s32* xJcc32( JccComparisonType comparison=Jcc_Unconditional, s32 displacement=0 );
 
 	// ------------------------------------------------------------------------
 	// Conditional jumps to fixed targets.
@@ -344,58 +344,41 @@ namespace x86Emitter
 	extern void xLDMXCSR( const u32* src );
 
 	extern void xMOVDZX( const xRegisterSSE& to, const xRegister32& from );
-	extern void xMOVDZX( const xRegisterSSE& to, const void* src );
 	extern void xMOVDZX( const xRegisterSSE& to, const ModSibBase& src );
 
 	extern void xMOVDZX( const xRegisterMMX& to, const xRegister32& from );
-	extern void xMOVDZX( const xRegisterMMX& to, const void* src );
 	extern void xMOVDZX( const xRegisterMMX& to, const ModSibBase& src );
 
 	extern void xMOVD( const xRegister32& to, const xRegisterSSE& from );
-	extern void xMOVD( void* dest, const xRegisterSSE& from );
 	extern void xMOVD( const ModSibBase& dest, const xRegisterSSE& from );
 
 	extern void xMOVD( const xRegister32& to, const xRegisterMMX& from );
-	extern void xMOVD( void* dest, const xRegisterMMX& from );
 	extern void xMOVD( const ModSibBase& dest, const xRegisterMMX& from );
 
 	extern void xMOVQ( const xRegisterMMX& to, const xRegisterMMX& from );
 	extern void xMOVQ( const xRegisterMMX& to, const xRegisterSSE& from );
 	extern void xMOVQ( const xRegisterSSE& to, const xRegisterMMX& from );
 
-	extern void xMOVQ( void* dest, const xRegisterSSE& from );
 	extern void xMOVQ( const ModSibBase& dest, const xRegisterSSE& from );
-	extern void xMOVQ( void* dest, const xRegisterMMX& from );
 	extern void xMOVQ( const ModSibBase& dest, const xRegisterMMX& from );
-	extern void xMOVQ( const xRegisterMMX& to, const void* src );
 	extern void xMOVQ( const xRegisterMMX& to, const ModSibBase& src );
 
-	extern void xMOVQZX( const xRegisterSSE& to, const void* src );
 	extern void xMOVQZX( const xRegisterSSE& to, const ModSibBase& src );
 	extern void xMOVQZX( const xRegisterSSE& to, const xRegisterSSE& from );
 	
 	extern void xMOVSS( const xRegisterSSE& to, const xRegisterSSE& from );
-	extern void xMOVSS( const void* to, const xRegisterSSE& from );
 	extern void xMOVSS( const ModSibBase& to, const xRegisterSSE& from );
 	extern void xMOVSD( const xRegisterSSE& to, const xRegisterSSE& from );
-	extern void xMOVSD( const void* to, const xRegisterSSE& from );
 	extern void xMOVSD( const ModSibBase& to, const xRegisterSSE& from );
 
-	extern void xMOVSSZX( const xRegisterSSE& to, const void* from );
 	extern void xMOVSSZX( const xRegisterSSE& to, const ModSibBase& from );
-	extern void xMOVSDZX( const xRegisterSSE& to, const void* from );
 	extern void xMOVSDZX( const xRegisterSSE& to, const ModSibBase& from );
 
-	extern void xMOVNTDQA( const xRegisterSSE& to, const void* from );
 	extern void xMOVNTDQA( const xRegisterSSE& to, const ModSibBase& from );
-	extern void xMOVNTDQ( void* to, const xRegisterSSE& from );
 	extern void xMOVNTDQA( const ModSibBase& to, const xRegisterSSE& from );
 
-	extern void xMOVNTPD( void* to, const xRegisterSSE& from );
 	extern void xMOVNTPD( const ModSibBase& to, const xRegisterSSE& from );
-	extern void xMOVNTPS( void* to, const xRegisterSSE& from );
 	extern void xMOVNTPS( const ModSibBase& to, const xRegisterSSE& from );
-	extern void xMOVNTQ( void* to, const xRegisterMMX& from );
 	extern void xMOVNTQ( const ModSibBase& to, const xRegisterMMX& from );
 
 	extern void xMOVMSKPS( const xRegister32& to, const xRegisterSSE& from );
@@ -435,11 +418,9 @@ namespace x86Emitter
 	extern const Internal::SimdImpl_DestRegSSE<0xf3,0x16> xMOVSHDUP;
 
 	extern void xINSERTPS( const xRegisterSSE& to, const xRegisterSSE& from, u8 imm8 );
-	extern void xINSERTPS( const xRegisterSSE& to, const u32* from, u8 imm8 );
 	extern void xINSERTPS( const xRegisterSSE& to, const ModSibStrict<u32>& from, u8 imm8 );
 
 	extern void xEXTRACTPS( const xRegister32& to, const xRegisterSSE& from, u8 imm8 );
-	extern void xEXTRACTPS( u32* dest, const xRegisterSSE& from, u8 imm8 );
 	extern void xEXTRACTPS( const ModSibStrict<u32>& dest, const xRegisterSSE& from, u8 imm8 );
 
 	// ------------------------------------------------------------------------
@@ -451,7 +432,8 @@ namespace x86Emitter
 	
 	extern const Internal::SimdImpl_AndNot			xANDN;
 
-	extern const Internal::SimdImpl_UcomI<0x66,0x2e> xUCOMI;
+	extern const Internal::SimdImpl_COMI<true>		xCOMI;
+	extern const Internal::SimdImpl_COMI<false>		xUCOMI;
 	extern const Internal::SimdImpl_rSqrt<0x53>		xRCP;
 	extern const Internal::SimdImpl_rSqrt<0x52>		xRSQRT;
 	extern const Internal::SimdImpl_Sqrt<0x51>		xSQRT;

common/include/x86emitter/internal.h

@@ -18,5 +18,5 @@
 
 #pragma once
 
-#include "ix86_types.h"
+#include "x86types.h"
-#include "ix86_instructions.h"
+#include "instructions.h"

common/include/x86emitter/legacy_internal.h

@@ -18,7 +18,7 @@
 
 #pragma once
 
-#include "ix86_internal.h"
+#include "internal.h"
 
 //------------------------------------------------------------------
 // Legacy Helper Macros and Functions (depreciated)
@@ -31,8 +31,8 @@ using x86Emitter::xWrite16;
 using x86Emitter::xWrite32;
 using x86Emitter::xWrite64;
 
-#include "ix86_legacy_types.h"
+#include "legacy_types.h"
-#include "ix86_legacy_instructions.h"
+#include "legacy_instructions.h"
 
 #define MEMADDR(addr, oplen)	(addr)
 
@@ -49,5 +49,4 @@ extern void SibSB( uint ss, uint index, uint base );
 extern void SET8R( int cc, int to );
 extern u8*  J8Rel( int cc, int to );
 extern u32* J32Rel( int cc, u32 to );
-extern u64  GetCPUTick( void );
 

common/include/x86emitter/tools.h

@@ -0,0 +1,106 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *  
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *  
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#pragma once
+
+// this is all that needs to be called and will fill up the below structs
+extern void cpudetectInit();
+
+//////////////////////////////////////////////////////////////////////////////////////////
+// cpu capabilities structure
+//
+struct CAPABILITIES
+{
+   u32 hasFloatingPointUnit;
+   u32 hasVirtual8086ModeEnhancements;
+   u32 hasDebuggingExtensions;
+   u32 hasPageSizeExtensions;
+   u32 hasTimeStampCounter;
+   u32 hasModelSpecificRegisters;
+   u32 hasPhysicalAddressExtension;
+   u32 hasCOMPXCHG8BInstruction;
+   u32 hasAdvancedProgrammableInterruptController;
+   u32 hasSEPFastSystemCall;
+   u32 hasMemoryTypeRangeRegisters;
+   u32 hasPTEGlobalFlag;
+   u32 hasMachineCheckArchitecture;
+   u32 hasConditionalMoveAndCompareInstructions;
+   u32 hasFGPageAttributeTable;
+   u32 has36bitPageSizeExtension;
+   u32 hasProcessorSerialNumber;
+   u32 hasCFLUSHInstruction;
+   u32 hasDebugStore;
+   u32 hasACPIThermalMonitorAndClockControl;
+   u32 hasMultimediaExtensions;
+   u32 hasFastStreamingSIMDExtensionsSaveRestore;
+   u32 hasStreamingSIMDExtensions;
+   u32 hasStreamingSIMD2Extensions;
+   u32 hasSelfSnoop;
+   u32 hasMultiThreading;			// is TRUE for both mutli-core and Hyperthreaded CPUs.
+   u32 hasThermalMonitor;
+   u32 hasIntel64BitArchitecture;
+   u32 hasStreamingSIMD3Extensions;
+   u32 hasSupplementalStreamingSIMD3Extensions;
+   u32 hasStreamingSIMD4Extensions;
+   u32 hasStreamingSIMD4Extensions2;
+
+   // AMD-specific CPU Features
+   u32 hasMultimediaExtensionsExt;
+   u32 hasAMD64BitArchitecture;
+   u32 has3DNOWInstructionExtensionsExt;
+   u32 has3DNOWInstructionExtensions;
+   u32 hasStreamingSIMD4ExtensionsA;
+};
+
+//////////////////////////////////////////////////////////////////////////////////////////
+//
+struct CPUINFO
+{
+   u32 x86Family;	   // Processor Family
+   u32 x86Model;	   // Processor Model
+   u32 x86PType;	   // Processor Type
+   u32 x86StepID;	   // Stepping ID
+   u32 x86Flags;	   // Feature Flags
+   u32 x86Flags2;	   // More Feature Flags
+   u32 x86EFlags;	   // Extended Feature Flags
+   u32 x86EFlags2;	   // Extended Feature Flags pg2
+
+   u32 PhysicalCores;
+   u32 LogicalCores;
+
+   char x86ID[16];	   // Vendor ID  //the vendor creator (in %s)
+   char x86Type[20];   //cpu type in char format //the cpu type (in %s)
+   char x86Fam[50];    // family in char format //the original cpu name string (in %s)
+   u32 cpuspeed;      // speed of cpu //this will give cpu speed (in %d)
+};
+
+
+//////////////////////////////////////////////////////////////////////////////////////////
+
+
+extern CAPABILITIES cpucaps;
+extern CPUINFO cpuinfo;
+
+extern u8 g_globalMMXSaved, g_globalXMMSaved;
+extern bool g_EEFreezeRegs;
+
+// when using mmx/xmm regs, use; 0 is load
+// freezes no matter the state
+extern void FreezeXMMRegs(int save);
+extern void FreezeMMXRegs(int save);
+extern void FreezeRegs(int save);

common/include/x86emitter/x86emitter.h

@@ -45,10 +45,11 @@
 
 #pragma once
 
-#include "ix86_types.h"
+#include "x86types.h"
-#include "ix86_instructions.h"
+#include "tools.h"
+#include "instructions.h"
 
 // Including legacy items for now, but these should be removed eventually,
 // once most code is no longer dependent on them.
-#include "ix86_legacy_types.h"
+#include "legacy_types.h"
-#include "ix86_legacy_instructions.h"
+#include "legacy_instructions.h"

common/include/x86emitter/x86types.h

@@ -18,76 +18,6 @@
 
 #pragma once
 
-extern void cpudetectInit();//this is all that needs to be called and will fill up the below structs
-
-//cpu capabilities structure
-struct CAPABILITIES
-{
-   u32 hasFloatingPointUnit;
-   u32 hasVirtual8086ModeEnhancements;
-   u32 hasDebuggingExtensions;
-   u32 hasPageSizeExtensions;
-   u32 hasTimeStampCounter;
-   u32 hasModelSpecificRegisters;
-   u32 hasPhysicalAddressExtension;
-   u32 hasCOMPXCHG8BInstruction;
-   u32 hasAdvancedProgrammableInterruptController;
-   u32 hasSEPFastSystemCall;
-   u32 hasMemoryTypeRangeRegisters;
-   u32 hasPTEGlobalFlag;
-   u32 hasMachineCheckArchitecture;
-   u32 hasConditionalMoveAndCompareInstructions;
-   u32 hasFGPageAttributeTable;
-   u32 has36bitPageSizeExtension;
-   u32 hasProcessorSerialNumber;
-   u32 hasCFLUSHInstruction;
-   u32 hasDebugStore;
-   u32 hasACPIThermalMonitorAndClockControl;
-   u32 hasMultimediaExtensions;
-   u32 hasFastStreamingSIMDExtensionsSaveRestore;
-   u32 hasStreamingSIMDExtensions;
-   u32 hasStreamingSIMD2Extensions;
-   u32 hasSelfSnoop;
-   u32 hasMultiThreading;			// is TRUE for both mutli-core and Hyperthreaded CPUs.
-   u32 hasThermalMonitor;
-   u32 hasIntel64BitArchitecture;
-   u32 hasStreamingSIMD3Extensions;
-   u32 hasSupplementalStreamingSIMD3Extensions;
-   u32 hasStreamingSIMD4Extensions;
-   u32 hasStreamingSIMD4Extensions2;
-
-   // AMD-specific CPU Features
-   u32 hasMultimediaExtensionsExt;
-   u32 hasAMD64BitArchitecture;
-   u32 has3DNOWInstructionExtensionsExt;
-   u32 has3DNOWInstructionExtensions;
-   u32 hasStreamingSIMD4ExtensionsA;
-};
-
-extern CAPABILITIES cpucaps;
-
-struct CPUINFO
-{
-   u32 x86Family;	   // Processor Family
-   u32 x86Model;	   // Processor Model
-   u32 x86PType;	   // Processor Type
-   u32 x86StepID;	   // Stepping ID
-   u32 x86Flags;	   // Feature Flags
-   u32 x86Flags2;	   // More Feature Flags
-   u32 x86EFlags;	   // Extended Feature Flags
-   u32 x86EFlags2;	   // Extended Feature Flags pg2
-
-   u32 PhysicalCores;
-   u32 LogicalCores;
-
-   char x86ID[16];	   // Vendor ID  //the vendor creator (in %s)
-   char x86Type[20];   //cpu type in char format //the cpu type (in %s)
-   char x86Fam[50];    // family in char format //the original cpu name string (in %s)
-   u32 cpuspeed;      // speed of cpu //this will give cpu speed (in %d)
-};
-
-extern CPUINFO cpuinfo;
-
 //------------------------------------------------------------------
 #ifdef _MSC_VER
 #define __threadlocal __declspec(thread)
@@ -206,9 +136,8 @@ __forceinline void xWrite( T val )
 
 		int Id;
 
-		xRegisterBase( const xRegisterBase<OperandType>& src ) : Id( src.Id ) {}
 		xRegisterBase(): Id( -1 ) {}
-		explicit xRegisterBase( int regId ) : Id( regId ) { jASSUME( Id >= -1 && Id < 8 ); }
+		explicit xRegisterBase( int regId ) : Id( regId ) { jASSUME( Id >= -2 && Id < 8 ); }	// allow -2 for user-custom identifiers.
 
 		bool IsEmpty() const { return Id < 0; }
 
@@ -220,12 +149,6 @@ __forceinline void xWrite( T val )
 
 		bool operator==( const xRegisterBase<OperandType>& src ) const	{ return (Id == src.Id); }
 		bool operator!=( const xRegisterBase<OperandType>& src ) const	{ return (Id != src.Id); }
-
-		xRegisterBase<OperandType>& operator=( const xRegisterBase<OperandType>& src )
-		{
-			Id = src.Id;
-			return *this;
-		}
 	};
 
 	//////////////////////////////////////////////////////////////////////////////////////////
@@ -233,19 +156,15 @@ __forceinline void xWrite( T val )
 	template< typename OperandType >
 	class xRegister : public xRegisterBase<OperandType>
 	{
-	public:
-		static const xRegister Empty;		// defined as an empty/unused value (-1)
-
 	public:
 		xRegister(): xRegisterBase<OperandType>() {}
-		xRegister( const xRegister& src ) : xRegisterBase<OperandType>( src.Id ) {}
+		xRegister( const xRegisterBase<OperandType>& src ) : xRegisterBase<OperandType>( src ) {}
-		explicit xRegister( const xRegisterBase<OperandType>& src ) : xRegisterBase<OperandType>( src ) {}
 		explicit xRegister( int regId ) : xRegisterBase<OperandType>( regId ) {}
 		
 		bool operator==( const xRegister<OperandType>& src ) const	{ return this->Id == src.Id; }
 		bool operator!=( const xRegister<OperandType>& src ) const	{ return this->Id != src.Id; }
 
-		xRegister<OperandType>& operator=( const xRegister<OperandType>& src )
+		xRegister<OperandType>& operator=( const xRegisterBase<OperandType>& src )
 		{
 			this->Id = src.Id;
 			return *this;
@@ -262,14 +181,13 @@ __forceinline void xWrite( T val )
 
 	public:
 		xRegisterSIMD(): xRegisterBase<OperandType>() {}
-		xRegisterSIMD( const xRegisterSIMD& src ) : xRegisterBase<OperandType>( src.Id ) {}
 		explicit xRegisterSIMD( const xRegisterBase<OperandType>& src ) : xRegisterBase<OperandType>( src ) {}
 		explicit xRegisterSIMD( int regId ) : xRegisterBase<OperandType>( regId ) {}
 		
 		bool operator==( const xRegisterSIMD<OperandType>& src ) const	{ return this->Id == src.Id; }
 		bool operator!=( const xRegisterSIMD<OperandType>& src ) const	{ return this->Id != src.Id; }
 
-		xRegisterSIMD<OperandType>& operator=( const xRegisterSIMD<OperandType>& src )
+		xRegisterSIMD<OperandType>& operator=( const xRegisterBase<OperandType>& src )
 		{
 			this->Id = src.Id;
 			return *this;
@@ -317,6 +235,10 @@ __forceinline void xWrite( T val )
 		xAddressInfo operator+( const xAddressReg& right ) const;
 		xAddressInfo operator+( const xAddressInfo& right ) const;
 		xAddressInfo operator+( s32 right ) const;
+		xAddressInfo operator+( const void* right ) const;
+
+		xAddressInfo operator-( s32 right ) const;
+		xAddressInfo operator-( const void* right ) const;
 
 		xAddressInfo operator*( u32 factor ) const;
 		xAddressInfo operator<<( u32 shift ) const;
@@ -381,6 +303,7 @@ __forceinline void xWrite( T val )
 		__forceinline xAddressInfo operator+( const xAddressInfo& right ) const { return xAddressInfo( *this ).Add( right ); }
 		__forceinline xAddressInfo operator+( s32 imm ) const { return xAddressInfo( *this ).Add( imm ); }
 		__forceinline xAddressInfo operator-( s32 imm ) const { return xAddressInfo( *this ).Add( -imm ); }
+		__forceinline xAddressInfo operator+( const void* addr ) const { return xAddressInfo( *this ).Add( (uptr)addr ); }
 	};
 
 	extern const xRegisterSSE
@@ -405,6 +328,33 @@ __forceinline void xWrite( T val )
 
 	extern const xRegisterCL cl;		// I'm special!
 
+	//////////////////////////////////////////////////////////////////////////////////////////
+	// xImmReg - used to represent an immediate value which can also be optimized to a register.
+	// Note that the immediate value represented by this structure is *always* legal.  The
+	// register assignment is an optional optimization which can be implemented in cases where
+	// an immediate is used enough times to merit allocating it to a register.
+	//
+	// Note: not all instructions support this operand type (yet).  You can always implement it
+	// manually by checking the status of IsReg() and generating the xOP conditionally.
+	//
+	template< typename OperandType >
+	class xImmReg
+	{
+		xRegister<OperandType> m_reg;
+		int m_imm;
+		
+	public:
+		xImmReg() :
+			m_reg(), m_imm( 0 ) { }
+
+		xImmReg( int imm, const xRegister<OperandType>& reg=xRegister<OperandType>() ) :
+			m_reg( reg ), m_imm( imm ) { }
+			
+		const xRegister<OperandType>& GetReg() const { return m_reg; }
+		const int GetImm() const { return m_imm; }
+		bool IsReg() const { return !m_reg.IsEmpty(); }
+	};
+	
 	//////////////////////////////////////////////////////////////////////////////////////////
 	// ModSib - Internal low-level representation of the ModRM/SIB information.
 	//
@@ -429,6 +379,7 @@ __forceinline void xWrite( T val )
 		explicit ModSibBase( const xAddressInfo& src );
 		explicit ModSibBase( s32 disp );
 		ModSibBase( xAddressReg base, xAddressReg index, int scale=0, s32 displacement=0 );
+		ModSibBase( const void* target );
 		
 		bool IsByteSizeDisp() const { return is_s8( Displacement ); }
 
@@ -455,8 +406,10 @@ __forceinline void xWrite( T val )
 	public:
 		static const uint OperandSize = sizeof( OperandType );
 
+		__forceinline explicit ModSibStrict( const ModSibBase& src ) : ModSibBase( src ) {}
 		__forceinline explicit ModSibStrict( const xAddressInfo& src ) : ModSibBase( src ) {}
 		__forceinline explicit ModSibStrict( s32 disp ) : ModSibBase( disp ) {}
+		__forceinline ModSibStrict( const OperandType* target ) : ModSibBase( target ) {}
 		__forceinline ModSibStrict( xAddressReg base, xAddressReg index, int scale=0, s32 displacement=0 ) :
 			ModSibBase( base, index, scale, displacement ) {}
 		
@@ -468,6 +421,18 @@ __forceinline void xWrite( T val )
 
 		__forceinline ModSibStrict<OperandType> operator+( const s32 imm ) const { return ModSibStrict<OperandType>( *this ).Add( imm ); }
 		__forceinline ModSibStrict<OperandType> operator-( const s32 imm ) const { return ModSibStrict<OperandType>( *this ).Add( -imm ); }
+
+		bool operator==( const ModSibStrict<OperandType>& src ) const
+		{
+			return
+				( Base == src.Base ) && ( Index == src.Index ) &&
+				( Scale == src.Scale ) && ( Displacement == src.Displacement );
+		}
+
+		bool operator!=( const ModSibStrict<OperandType>& src ) const
+		{
+			return !operator==( src );
+		}
 	};
 
 	//////////////////////////////////////////////////////////////////////////////////////////
@@ -547,6 +512,58 @@ __forceinline void xWrite( T val )
 	extern const xAddressIndexer<u16>	ptr16;	// explicitly typed addressing, usually needed for '[dest],imm' instruction forms
 	extern const xAddressIndexer<u8>	ptr8;	// explicitly typed addressing, usually needed for '[dest],imm' instruction forms
 
+	//////////////////////////////////////////////////////////////////////////////////////////
+	//
+	// [TODO] - make SSE version of thise, perhaps?
+	//
+	template< typename OperandType >
+	class xDirectOrIndirect
+	{
+		xRegister<OperandType> m_RegDirect;
+		ModSibStrict<OperandType> m_MemIndirect;
+
+	public:
+		xDirectOrIndirect() :
+			m_RegDirect(), m_MemIndirect( 0 ) {}
+
+		xDirectOrIndirect( const xRegister<OperandType>& srcreg ) :
+			m_RegDirect( srcreg ), m_MemIndirect( 0 ) {}
+
+		xDirectOrIndirect( const ModSibBase& srcmem ) :
+			m_RegDirect(), m_MemIndirect( srcmem ) {}
+
+		xDirectOrIndirect( const ModSibStrict<OperandType>& srcmem ) :
+			m_RegDirect(), m_MemIndirect( srcmem ) {}
+		
+		const xRegister<OperandType>& GetReg() const { return m_RegDirect; }
+		const ModSibStrict<OperandType>& GetMem() const { return m_MemIndirect; }
+		bool IsDirect() const { return !m_RegDirect.IsEmpty(); }
+		bool IsIndirect() const { return m_RegDirect.IsEmpty(); }
+
+		bool operator==( const xDirectOrIndirect<OperandType>& src ) const
+		{
+			return IsDirect() ? 
+				(m_RegDirect == src.m_RegDirect) :
+				(m_MemIndirect == src.m_MemIndirect);
+		}
+		
+		bool operator!=( const xDirectOrIndirect<OperandType>& src ) const 
+		{
+			return !operator==( src );
+		}
+
+		bool operator==( const xRegister<OperandType>& src ) const	{ return (m_RegDirect == src); }
+		bool operator!=( const xRegister<OperandType>& src ) const	{ return (m_RegDirect == src); }
+	};
+
+	typedef xImmReg<u8>		xImmOrReg8;
+	typedef xImmReg<u16>	xImmOrReg16;
+	typedef xImmReg<u32>	xImmOrReg32;
+
+	typedef xDirectOrIndirect<u8>	xDirectOrIndirect8;
+	typedef xDirectOrIndirect<u16>	xDirectOrIndirect16;
+	typedef xDirectOrIndirect<u32>	xDirectOrIndirect32;
+
 	//////////////////////////////////////////////////////////////////////////////////////////
 	// JccComparisonType - enumerated possibilities for inspired code branching!
 	//
@@ -633,7 +650,7 @@ __forceinline void xWrite( T val )
 		JccComparisonType m_cc;		// comparison type of the instruction
 
 	public:
-		const int GetMaxInstructionSize() const
+		int GetMaxInstructionSize() const
 		{
 			jASSUME( m_cc != Jcc_Unknown );
 			return ( m_cc == Jcc_Unconditional ) ? 5 : 6;
@@ -686,70 +703,7 @@ __forceinline void xWrite( T val )
 	//	
 	namespace Internal
 	{
-		extern void SimdPrefix( u8 prefix, u16 opcode );
+		#include "implement/helpers.h"
-		extern void EmitSibMagic( uint regfield, const void* address );
-		extern void EmitSibMagic( uint regfield, const ModSibBase& info );
-		extern void xJccKnownTarget( JccComparisonType comparison, const void* target, bool slideForward );
-
-
-		// Writes a ModRM byte for "Direct" register access forms, which is used for all
-		// instructions taking a form of [reg,reg].
-		template< typename T > __emitinline
-		void EmitSibMagic( uint reg1, const xRegisterBase<T>& reg2 )
-		{
-			xWrite8( (Mod_Direct << 6) | (reg1 << 3) | reg2.Id );
-		}
-
-		template< typename T1, typename T2 > __emitinline
-		void EmitSibMagic( const xRegisterBase<T1> reg1, const xRegisterBase<T2>& reg2 )
-		{
-			xWrite8( (Mod_Direct << 6) | (reg1.Id << 3) | reg2.Id );
-		}
-
-		template< typename T1 > __emitinline
-		void EmitSibMagic( const xRegisterBase<T1> reg1, const void* src )			{ EmitSibMagic( reg1.Id, src ); }
-
-		template< typename T1 > __emitinline
-		void EmitSibMagic( const xRegisterBase<T1> reg1, const ModSibBase& sib )	{ EmitSibMagic( reg1.Id, sib ); }
-
-		// ------------------------------------------------------------------------
-		template< typename T1, typename T2 > __emitinline
-		void xOpWrite( u8 prefix, u8 opcode, const T1& param1, const T2& param2 )
-		{
-			if( prefix != 0 )
-				xWrite16( (opcode<<8) | prefix );
-			else
-				xWrite8( opcode );
-
-			EmitSibMagic( param1, param2 );
-		}
-
-		// ------------------------------------------------------------------------
-		template< typename T1, typename T2 > __emitinline
-		void xOpWrite0F( u8 prefix, u16 opcode, const T1& param1, const T2& param2 )
-		{
-			SimdPrefix( prefix, opcode );
-			EmitSibMagic( param1, param2 );
-		}
-
-		template< typename T1, typename T2 > __emitinline
-		void xOpWrite0F( u8 prefix, u16 opcode, const T1& param1, const T2& param2, u8 imm8 )
-		{
-			xOpWrite0F( prefix, opcode, param1, param2 );
-			xWrite8( imm8 );
-		}
-
-		template< typename T1, typename T2 > __emitinline
-		void xOpWrite0F( u16 opcode, const T1& param1, const T2& param2 )			{ xOpWrite0F( 0, opcode, param1, param2 ); }
-
-		template< typename T1, typename T2 > __emitinline
-		void xOpWrite0F( u16 opcode, const T1& param1, const T2& param2, u8 imm8 )	{ xOpWrite0F( 0, opcode, param1, param2, imm8 ); }
-
-		// ------------------------------------------------------------------------
-
-		template< typename T > bool Is8BitOp() { return sizeof(T) == 1; }
-		template< typename T > void prefix16() { if( sizeof(T) == 2 ) xWrite8( 0x66 ); }
-
 		#include "implement/xmm/basehelpers.h"
 		#include "implement/xmm/moremovs.h"
 		#include "implement/xmm/arithmetic.h"
@@ -764,6 +718,26 @@ __forceinline void xWrite( T val )
 		#include "implement/test.h"
 		#include "implement/jmpcall.h"
 	}
+
+	static __forceinline xAddressInfo operator+( const void* addr, const xAddressReg& reg )
+	{
+		return xAddressInfo( reg, (sptr)addr );
+	}
+
+	static __forceinline xAddressInfo operator+( const void* addr, const xAddressInfo& reg )
+	{
+		return xAddressInfo( (sptr)addr ).Add( reg );
+	}
+
+	static __forceinline xAddressInfo operator+( s32 addr, const xAddressReg& reg )
+	{
+		return xAddressInfo( reg, (sptr)addr );
+	}
+
+	static __forceinline xAddressInfo operator+( s32 addr, const xAddressInfo& reg )
+	{
+		return xAddressInfo( (sptr)addr ).Add( reg );
+	}
 }
 
-#include "ix86_inlines.inl"
+#include "inlines.inl"

common/src/Utilities/AlignedMalloc.cpp

@@ -20,7 +20,6 @@
 // it built into their CRT/libc.
 
 #include "PrecompiledHeader.h"
-#include "System.h"
 
 
 struct AlignedMallocHeader

common/src/Utilities/Console.cpp

@@ -0,0 +1,236 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#include "PrecompiledHeader.h"
+#include "Threading.h"
+
+using namespace Threading;
+using namespace std;
+
+const _VARG_PARAM va_arg_dummy = { 0 };
+
+namespace Console
+{
+	MutexLock m_writelock;
+	std::string m_format_buffer;
+
+	// ------------------------------------------------------------------------
+	bool __fastcall Write( Colors color, const char* fmt )
+	{
+		SetColor( color );
+		Write( fmt );
+		ClearColor();
+
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool __fastcall Write( Colors color, const wxString& fmt )
+	{
+		SetColor( color );
+		Write( fmt );
+		ClearColor();
+
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool __fastcall WriteLn( const char* fmt )
+	{
+		Write( fmt );
+		Newline();
+		return false;
+	}
+
+	bool __fastcall WriteLn( Colors color, const char* fmt )
+	{
+		SetColor( color );
+		Write( fmt );
+		Newline();
+		ClearColor();
+
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool __fastcall WriteLn( const wxString& fmt )
+	{
+		Write( fmt );
+		Newline();
+		return false;
+	}
+
+	bool __fastcall WriteLn( Colors color, const wxString& fmt )
+	{
+		SetColor( color );
+		Write( fmt );
+		Newline();
+		ClearColor();
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	__forceinline void __fastcall _Write( const char* fmt, va_list args )
+	{
+		ScopedLock locker( m_writelock );
+		vssprintf( m_format_buffer, fmt, args );
+		const char* cstr = m_format_buffer.c_str();
+
+		Write( fmt );
+	}
+
+	__forceinline void __fastcall _WriteLn( const char* fmt, va_list args )
+	{
+		_Write( fmt, args );
+		Newline();
+	}
+
+	__forceinline void __fastcall _WriteLn( Colors color, const char* fmt, va_list args )
+	{
+		SetColor( color );
+		_WriteLn( fmt, args );
+		ClearColor();
+	}
+		
+	// ------------------------------------------------------------------------
+	bool Write( const char* fmt, VARG_PARAM dummy, ... )
+	{
+		varg_assert();
+
+		va_list args;
+		va_start(args,dummy);
+		_Write( fmt, args );
+		va_end(args);
+
+		return false;
+	}
+
+	bool Write( Colors color, const char* fmt, VARG_PARAM dummy, ... )
+	{
+		varg_assert();
+
+		va_list args;
+		va_start(args,dummy);
+		SetColor( color );
+		_Write( fmt, args );
+		ClearColor();
+		va_end(args);
+
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool WriteLn( const char* fmt, VARG_PARAM dummy, ... )
+	{
+		varg_assert();
+		va_list args;
+		va_start(args,dummy);
+		_WriteLn( fmt, args );
+		va_end(args);
+
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool WriteLn( Colors color, const char* fmt, VARG_PARAM dummy, ... )
+	{
+		varg_assert();
+
+		va_list args;
+		va_start(args,dummy);
+		_WriteLn( color, fmt, args );
+		va_end(args);
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool Error( const char* fmt, VARG_PARAM dummy, ... )
+	{
+		varg_assert();
+
+		va_list args;
+		va_start(args,dummy);
+		_WriteLn( Color_Red, fmt, args );
+		va_end(args);
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool Notice( const char* fmt, VARG_PARAM dummy, ... )
+	{
+		varg_assert();
+
+		va_list list;
+		va_start(list,dummy);
+		_WriteLn( Color_Yellow, fmt, list );
+		va_end(list);
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool Status( const char* fmt, VARG_PARAM dummy, ... )
+	{
+		varg_assert();
+
+		va_list list;
+		va_start(list,dummy);
+		_WriteLn( Color_Green, fmt, list );
+		va_end(list);
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool __fastcall Error( const char* fmt )
+	{
+		WriteLn( Color_Red, fmt );
+		return false;
+	}
+
+	bool __fastcall Notice( const char* fmt )
+	{
+		WriteLn( Color_Yellow, fmt );
+		return false;
+	}
+
+	bool __fastcall Status( const char* fmt )
+	{
+		WriteLn( Color_Green, fmt );
+		return false;
+	}
+
+	// ------------------------------------------------------------------------
+	bool __fastcall Error( const wxString& src )
+	{
+		WriteLn( Color_Red, src );
+		return false;
+	}
+
+	bool __fastcall Notice( const wxString& src )
+	{
+		WriteLn( Color_Yellow, src );
+		return false;
+	}
+
+	bool __fastcall Status( const wxString& src )
+	{
+		WriteLn( Color_Green, src );
+		return false;
+	}
+
+}
+

common/src/Utilities/Exceptions.cpp

@@ -163,4 +163,3 @@ namespace Exception
 		return m_message;
 	}
 }
-

common/src/Utilities/Linux/LnxHostSys.cpp

@@ -0,0 +1,57 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#include "PrecompiledHeader.h"
+
+#include <sys/mman.h>
+#include <signal.h>
+
+namespace HostSys
+{
+	void *Mmap(uptr base, u32 size)
+	{
+		u8 *Mem;
+		Mem = (u8*)mmap((uptr*)base, size, PROT_EXEC | PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, 0, 0);
+		if (Mem == MAP_FAILED) Console::Notice("Mmap Failed!");
+
+		return Mem;
+	}
+
+	void Munmap(uptr base, u32 size)
+	{
+		munmap((uptr*)base, size);
+	}
+
+	void MemProtect( void* baseaddr, size_t size, PageProtectionMode mode, bool allowExecution )
+	{
+		int lnxmode = 0;
+
+		// make sure size is aligned to the system page size:
+		size = (size + m_pagemask) & ~m_pagemask;
+
+		switch( mode )
+		{
+			case Protect_NoAccess: break;
+			case Protect_ReadOnly: lnxmode = PROT_READ; break;
+			case Protect_ReadWrite: lnxmode = PROT_READ | PROT_WRITE; break;
+		}
+
+		if( allowExecution ) lnxmode |= PROT_EXEC;
+		mprotect( baseaddr, size, lnxmode );
+	}
+}

common/src/Utilities/LnxMisc.cpp

@@ -22,8 +22,6 @@
 #include <time.h>
 #include <sys/time.h>
 
-#include "CDVD.h"
-
 void InitCPUTicks()
 {
 }

common/src/Utilities/LnxThreads.cpp

@@ -17,7 +17,7 @@
  */
 
 #include "PrecompiledHeader.h"
-#include "ix86/ix86.h"
+#include "Threading.h"
 
 // Note: assuming multicore is safer because it forces the interlocked routines to use
 // the LOCK prefix.  The prefix works on single core CPUs fine (but is slow), but not
@@ -169,28 +169,4 @@ namespace Threading
 
 		return result;
 	}
-
-	#ifdef __x86_64__
-	__forceinline void pcsx2_InterlockedExchange64(volatile s64* Target, s64 Value)
-	{
-		__asm__ __volatile__(
-			".intel_syntax noprefix\n"
-			 "lock xchg [%0], rax\n"
-			 ".att_syntax\n" : : "r"(Target), "a"(Value) : "memory"
-		);
-		return 0;
-	}
-
-	__forceinline s64 pcsx2_InterlockedCompareExchange64(volatile s64* dest, s64 exch, s64 comp)
-	{
-		s64 old;
-		__asm__ __volatile__(
-			"lock; cmpxchgq %q2, %q1"
-			: "=a" (old)
-			: "r" (exch), "m" (*dest), "a" (comp)
-		);
-		return old;
-	}
-#endif
-
 }

common/src/Utilities/PrecompiledHeader.cpp

@@ -0,0 +1,2 @@
+
+#include "PrecompiledHeader.h"

common/src/Utilities/PrecompiledHeader.h

@@ -0,0 +1,17 @@
+#ifndef UTILITIES_PRECOMPILED_HEADER
+#define UTILITIES_PRECOMPILED_HEADER
+
+#include "Dependencies.h"
+
+using std::string;
+using std::min;
+using std::max;
+
+#include "MemcpyFast.h"
+#include "Console.h"
+#include "Exceptions.h"
+#include "SafeArray.h"
+#include "General.h"
+
+#endif
+

common/src/Utilities/StringHelpers.cpp

@@ -17,7 +17,8 @@
  */
  
 #include "PrecompiledHeader.h"
-#include "SafeArray.h"
+
+const wxRect wxDefaultRect( wxDefaultCoord, wxDefaultCoord, wxDefaultCoord, wxDefaultCoord );
 
 //////////////////////////////////////////////////////////////////////////////////////////
 // Splits a string into parts and adds the parts into the given SafeList.

common/src/Utilities/WinMisc.cpp

@@ -18,7 +18,6 @@
  
 #include "PrecompiledHeader.h"
 #include "RedtapeWindows.h"
-#include "cdvd.h"
 
 static LARGE_INTEGER lfreq;
 

common/src/Utilities/WinThreads.cpp

@@ -19,9 +19,9 @@
 
 #include "PrecompiledHeader.h"
 #include "RedtapeWindows.h"
+#include "x86emitter/tools.h"
+#include "Threading.h"
 
-#include "System.h"
-#include "ix86/ix86_types.h"
 
 #ifdef _WIN32
 #include "implement.h"		// win32 pthreads implementations.
@@ -53,7 +53,7 @@ namespace Threading
 			LogicalCoresPerPhysicalCPU = CPUs;
 
 		cpuinfo.PhysicalCores = ( CPUs / LogicalCoresPerPhysicalCPU ) * PhysicalCoresPerPhysicalCPU;
-		ptw32_smp_system = ( cpuinfo.LogicalCores > 1 ) ? TRUE : FALSE;
+		//ptw32_smp_system = ( cpuinfo.LogicalCores > 1 ) ? TRUE : FALSE;
 	}
 
 	__forceinline void Timeslice()
@@ -77,13 +77,13 @@ namespace Threading
 	{
 		jASSUME( itsme != NULL );
 
-		pthread_win32_thread_attach_np();
+		//pthread_win32_thread_attach_np();
 
 		Thread& owner = *((Thread*)itsme);
 		owner.m_returncode = owner.Callback();
 		owner.m_terminated = true;
 
-		pthread_win32_thread_detach_np();
+		//pthread_win32_thread_detach_np();
 
 		return NULL;
 	}
@@ -98,14 +98,13 @@ namespace Threading
 
 	__forceinline long pcsx2_InterlockedExchange( volatile long* target, long srcval )
 	{
-		// Use the pthreads-win32 implementation...
+		return _InterlockedExchange( target, srcval );
-		return ptw32_InterlockedExchange( target, srcval );
 	}
 
 	__forceinline long pcsx2_InterlockedCompareExchange( volatile long* target, long srcval, long comp )
 	{
 		// Use the pthreads-win32 implementation...
-		return ptw32_InterlockedCompareExchange( target, srcval, comp );
+		return _InterlockedCompareExchange( target, srcval, comp );
 	}
 
 	__forceinline long pcsx2_InterlockedExchangeAdd( volatile long* target, long srcval )

common/src/Utilities/Windows/WinHostSys.cpp

@@ -16,8 +16,8 @@
  *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
  */
 
-#include "PrecompiledHeader.h"
+#include "../PrecompiledHeader.h"
-#include "Win32.h"
+#include "Utilities/RedtapeWindows.h"
 #include <winnt.h>
 
 namespace HostSys

common/src/Utilities/x86/MemcpyFast.S

@@ -27,8 +27,6 @@
 // zerofrog(@gmail.com)
 // and added to by arcum42@gmail.com
 .intel_syntax noprefix
-.extern g_EEFreezeRegs
-.extern FreezeMMXRegs_
 .extern _mmx_backup
 
 // mmx memcmp implementation, size has to be a multiple of 8
@@ -227,15 +225,7 @@ memcmp_End:
 
 .global memxor_mmx
 memxor_mmx:
-		// make sure mmx regs are stored
-		// FreezeMMXRegs(1);
-		cmp dword ptr [g_EEFreezeRegs], 0
-		je memxor_mmx_begin
-		push 1
-		call FreezeMMXRegs_
-		add esp, 4
 
-memxor_mmx_begin:
 	push esi
 	mov MEMXOR_SRC1, dword ptr [esp+8]
 	mov MEMXOR_SRC2, dword ptr [esp+12]

common/src/Utilities/x86/MemcpyFast.cpp

@@ -31,7 +31,7 @@
  3dsdk.support@amd.com
 ******************************************************************************/
 
-#include "PrecompiledHeader.h"
+#include "..\PrecompiledHeader.h"
 
 #ifdef _MSC_VER
 #pragma warning(disable:4414)
@@ -72,19 +72,13 @@ MEMCPY_AMD.CPP
 // This is faster than using software prefetch.  The technique is great for
 // getting maximum read bandwidth, especially in DDR memory systems.
 
-#include "Misc.h"
-
 // Inline assembly syntax for use with Visual C++
 
 #if defined(_MSC_VER)
 
-#ifdef _DEBUG
+#ifdef PCSX2_DEBUG
 extern u8 g_globalMMXSaved;
 
-void checkregs()
-{
-	if( g_EEFreezeRegs ) assert( g_globalMMXSaved );
-}
 #endif
 
 
@@ -772,7 +766,6 @@ End:
 // returns the xor of all elements, cmpsize has to be mult of 8
 void memxor_mmx(void* dst, const void* src1, int cmpsize)
 {
-	FreezeMMXRegs(1);
 	assert( (cmpsize&7) == 0 );
 
 	__asm {
@@ -875,7 +868,6 @@ Setup1:
 End:
 		emms
 	}
-	FreezeMMXRegs(0);
 }
 
 #endif

common/src/x86emitter/3dnow.cpp

@@ -17,7 +17,7 @@
  */
 
 #include "PrecompiledHeader.h"
-#include "ix86_legacy_internal.h"
+#include "legacy_internal.h"
 
 //------------------------------------------------------------------
 // 3DNOW instructions [Anyone caught dead using these will be re-killed]

common/src/x86emitter/PrecompiledHeader.cpp

@@ -0,0 +1,2 @@
+
+#include "PrecompiledHeader.h"

common/src/x86emitter/PrecompiledHeader.h

@@ -0,0 +1,29 @@
+#ifndef EMITTER_PRECOMPILED_HEADER
+#define EMITTER_PRECOMPILED_HEADER
+
+#include "Pcsx2Defs.h"
+
+#include <wx/string.h>
+#include <wx/tokenzr.h>
+#include <wx/gdicmn.h>		// for wxPoint/wxRect stuff
+#include <wx/intl.h>
+#include <wx/log.h>
+
+#include <stdexcept>
+#include <algorithm>
+#include <string>
+#include <cstring>		// string.h under c++
+#include <cstdio>		// stdio.h under c++
+#include <cstdlib>
+
+using std::string;
+using std::min;
+using std::max;
+
+#include "Utilities/Console.h"
+#include "Utilities/Exceptions.h"
+#include "Utilities/General.h"
+#include "Utilities/MemcpyFast.h"
+
+#endif
+

common/src/x86emitter/cpudetect.cpp

@@ -17,143 +17,41 @@
  */
 
 #include "PrecompiledHeader.h"
-
+#include "internal.h"
-#include "System.h"
+#include "tools.h"
-#include "ix86_legacy_internal.h"
+#include "Utilities/RedtapeWindows.h"
-
+#include "Utilities/Threading.h"
-#include "RedtapeWindows.h"
 
 using namespace x86Emitter;
 
-#if defined (_MSC_VER) && _MSC_VER >= 1400
-
-	extern "C"
-	{
-		void __cpuid(int* CPUInfo, int InfoType);
-		unsigned __int64 __rdtsc();
-#ifndef __INTEL_COMPILER
-#		pragma intrinsic(__cpuid)
-#		pragma intrinsic(__rdtsc)
-#endif
-	}
-
-#endif
-
 CAPABILITIES cpucaps;
 CPUINFO cpuinfo;
 
-#define cpuid(cmd,a,b,c,d) \
+static s32 iCpuId( u32 cmd, u32 *regs )
-  __asm__ __volatile__("xchgl %%ebx, %1; cpuid; xchgl %%ebx, %1" \
-		: "=a" (a), "=r" (b), "=c" (c), "=d" (d)  : "0" (cmd), "c" (0))
-
-extern s32 iCpuId( u32 cmd, u32 *regs )
 {
-   int flag=1;
+#ifdef _MSC_VER
-
-#if defined (_MSC_VER) && _MSC_VER >= 1400
-
 	__asm
 	{
-      xor ecx, ecx; /* ecx should be zero for CPUID(4) */
+		xor ecx, ecx;		// ecx should be zero for CPUID(4)
 	}
-   __cpuid( (int*)regs, cmd );
-
-   return 0;
-
-#elif defined (_MSC_VER)
-   __asm
-   {
-      push ebx;
-      push edi;
-
-      pushfd;
-      pop eax;
-      mov edx, eax;
-      xor eax, 1 << 21;
-      push eax;
-      popfd;
-      pushfd;
-      pop eax;
-      xor eax, edx;
-      mov flag, eax;
-   }
-   if ( ! flag )
-   {
-      return -1;
-   }
-
-   __asm
-   {
-      mov eax, cmd;
-      xor ecx, ecx; /* ecx should be zero for CPUID(4) */
-      cpuid;
-      mov edi, [regs]
-      mov [edi], eax;
-      mov [edi+4], ebx;
-      mov [edi+8], ecx;
-      mov [edi+12], edx;
-
-      pop edi;
-      pop ebx;
-   }
-   return 0;
-
 #else
+	__asm__ ( "xor ecx, ecx" );
+#endif
 
-   // GCC Assembly Code -->
+   __cpuid( (int*)regs, cmd );
-
-   // see if we can use cpuid
-   __asm__ __volatile__ (
-      "sub $0x18, %%esp\n"
-      "pushf\n"
-      "pop %%eax\n"
-      "mov %%eax, %%edx\n"
-      "xor $0x200000, %%eax\n"
-      "push %%eax\n"
-      "popf\n"
-      "pushf\n"
-      "pop %%eax\n"
-      "xor %%edx, %%eax\n"
-      "mov %%eax, %[flag]\n"
-	  "add $0x18, %%esp\n"
-      "cmpl $0x0,%%eax\n"
-      "jne 1f\n"
-      "mov $0xffffffff, %%eax\n"
-      "leave\n"
-      "ret\n"
-      "1:\n"
-      : [flag]"=r"(flag) :
-   );
-
-   cpuid(cmd, regs[0], regs[1], regs[2], regs[3]);
    return 0;
-
-#endif // _MSC_VER
 }
 
-u64 GetCPUTick( void )
+static u64 GetRdtsc( void )
 {
-#if defined (_MSC_VER) && _MSC_VER >= 1400
-
    return __rdtsc();
-
-#elif defined(_WIN32)
-
-   __asm rdtsc;
-
-#else
-
-   u32 _a, _d;
-	__asm__ __volatile__ ("rdtsc" : "=a"(_a), "=d"(_d));
-	return (u64)_a | ((u64)_d << 32);
-
-#endif
 }
 
 //////////////////////////////////////////////////////////////////////////////////////////
 // Note: This function doesn't support GCC/Linux.  Looking online it seems the only 
-// way to simulate the Micrsoft SEH model is to use unix signals, and the 'sigaction'
+// way to simulate the Microsoft SEH model is to use unix signals, and the 'sigaction'
 // function specifically.  Maybe a project for a linux developer at a later date. :)
+//
 #ifdef _MSC_VER
 static bool _test_instruction( void* pfnCall )
 {
@@ -173,47 +71,83 @@ static char* bool_to_char( bool testcond )
 
 #endif
 
-#if defined __LINUX__
+#ifdef __LINUX__
-
+#	include <sys/time.h>
-#include <sys/time.h>
+#	include <errno.h>
-#include <errno.h>
-
 #endif
 
+#ifdef _WINDOWS_
+static HANDLE s_threadId = NULL;
+static DWORD s_oldmask = ERROR_INVALID_PARAMETER;
+#endif
+
+static void SetSingleAffinity()
+{
+#ifdef _WINDOWS_
+	// Assign a single CPU thread affinity to ensure rdtsc() accuracy.
+	// (rdtsc for each CPU/core can differ, causing skewed results)
+
+	DWORD_PTR availProcCpus, availSysCpus;
+	if( !GetProcessAffinityMask( GetCurrentProcess(), &availProcCpus, &availSysCpus ) ) return;
+	
+	int i;
+	for( i=0; i<32; ++i )
+	{
+		if( availProcCpus & (1<<i) ) break;
+	}
+
+	HANDLE s_threadId = GetCurrentThread();
+	DWORD s_oldmask = SetThreadAffinityMask( s_threadId, (1UL<<i) );
+
+	if( s_oldmask == ERROR_INVALID_PARAMETER )
+	{
+		Console::Notice(
+			"CpuDetect: SetThreadAffinityMask failed...\n"
+			"\tSystem Affinity : 0x%08x"
+			"\tProcess Affinity: 0x%08x"
+			"\tAttempted Thread Affinity CPU: i",
+			params availProcCpus, availSysCpus, i
+		);
+	}
+#endif
+}
+
 //////////////////////////////////////////////////////////////////////////////////////////
 //
-s64 CPUSpeedHz( int time )
+static s64 CPUSpeedHz( u64 time )
 {
-   int timeStart, 
+	u64 timeStart, timeStop;
-            timeStop;
+	s64 startTick, endTick;
-   s64 startTick, 
-            endTick;
 
 	if( ! cpucaps.hasTimeStampCounter )
-   {
 		return 0; //check if function is supported
-   }
 
-	// Align the cpu execution to a timeGetTime boundary.
+	SetSingleAffinity();
-	// Without this the result could be skewed by up to several milliseconds.
 
-	do { timeStart = timeGetTime( );
+	// Align the cpu execution to a cpuTick boundary.
-	} while( timeGetTime( ) == timeStart );
+
+	do { timeStart = GetCPUTicks();
+	} while( GetCPUTicks() == timeStart );
 
 	do
 	{
-		timeStop = timeGetTime( );
+		timeStop = GetCPUTicks( );
-		startTick = GetCPUTick( );
+		startTick = GetRdtsc( );
-	} while( ( timeStop - timeStart ) < 1 );
+	} while( ( timeStop - timeStart ) == 0 );
 
 	timeStart = timeStop;
 	do
 	{
-		timeStop = timeGetTime();
+		timeStop = GetCPUTicks();
-		endTick = GetCPUTick();
+		endTick = GetRdtsc();
 	}
 	while( ( timeStop - timeStart ) < time );
 
+#ifdef _WINDOWS_
+	if( s_oldmask != ERROR_INVALID_PARAMETER )
+		SetThreadAffinityMask( s_threadId, s_oldmask );
+#endif
+
 	return (s64)( endTick - startTick );
 }
 
@@ -265,7 +199,7 @@ void cpudetectInit()
          cpuinfo.x86Family = (regs[ 0 ] >>  8) & 0xf;
          cpuinfo.x86PType  = (regs[ 0 ] >> 12) & 0x3;
 		 LogicalCoresPerPhysicalCPU = ( regs[1] >> 16 ) & 0xff;
-         x86_64_8BITBRANDID = regs[1] & 0xff;
+         x86_64_8BITBRANDID = regs[ 1 ] & 0xff;
          cpuinfo.x86Flags  =  regs[ 3 ];
          cpuinfo.x86Flags2 =  regs[ 2 ];
       }
@@ -366,8 +300,13 @@ void cpudetectInit()
 	cpucaps.has3DNOWInstructionExtensions                = ( cpuinfo.x86EFlags >> 31 ) & 1; //3dnow   
 	cpucaps.hasStreamingSIMD4ExtensionsA                 = ( cpuinfo.x86EFlags2 >> 6 ) & 1; //INSERTQ / EXTRQ / MOVNT
 
+	InitCPUTicks();
+	u64 span = GetTickFrequency();
 
-	cpuinfo.cpuspeed = (u32)(CPUSpeedHz( 400 ) / 400000 );
+	if( (span % 1000) < 400 )	// helps minimize rounding errors
+		cpuinfo.cpuspeed = (u32)( CPUSpeedHz( span / 1000 ) / 1000 );
+	else
+		cpuinfo.cpuspeed = (u32)( CPUSpeedHz( span / 500 ) / 2000 );
 
 	// --> SSE3 / SSSE3 / SSE4.1 / SSE 4.2 detection <--
 
@@ -376,9 +315,18 @@ void cpudetectInit()
 	cpucaps.hasStreamingSIMD4Extensions  = ( cpuinfo.x86Flags2 >> 19 ) & 1; //sse4.1   
 	cpucaps.hasStreamingSIMD4Extensions2 = ( cpuinfo.x86Flags2 >> 20 ) & 1; //sse4.2
 
+	//////////////////////////////////////////////////////////////////////////////////////////
+	// SIMD Instruction Support Detection
+	//
 	// Can the SSE3 / SSE4.1 bits be trusted?  Using an instruction test is a very "complete"
-	// approach to ensuring the bit is accurate, and at least one reported case of a Q9550 not
+	// approach to ensuring the instruction set is supported, and at least one reported case
-	// having SSE 4.1 set but still supporting it properly is fixed by this --air
+	// of a Q9550 not having it's SSE 4.1 bit set but still supporting it properly is fixed
+	// by this --air
+	//  (note: the user who reported the case later fixed the problem by doing a CMOS reset)
+	//
+	// Linux support note: Linux/GCC doesn't have SEH-style exceptions which allow handling of
+	// CPU-level exceptions (__try/__except in msvc) so this code is disabled on GCC, and
+	// detection relies on the CPUID bits alone.
 
 	#ifdef _MSC_VER
 	u8* recSSE = (u8*)HostSys::Mmap( NULL, 0x1000 );
@@ -386,15 +334,15 @@ void cpudetectInit()
 	{
 		xSetPtr( recSSE );
 		xMOVSLDUP( xmm1, xmm0 );
-		RET();
+		xRET();
 
 		u8* funcSSSE3 = xGetPtr();
 		xPABS.W( xmm0, xmm1 );
-		RET();
+		xRET();
 
 		u8* funcSSE41 = xGetPtr();
 		xBLEND.VPD( xmm1, xmm0 );
-		RET();
+		xRET();
 		
 		bool sse3_result = _test_instruction( recSSE );  // sse3
 		bool ssse3_result = _test_instruction( funcSSSE3 );
@@ -404,7 +352,8 @@ void cpudetectInit()
 
 		// Test for and log any irregularities here.
 		// We take the instruction test result over cpuid since (in theory) it should be a
-		// more reliable gauge of the cpu's actual ability.
+		// more reliable gauge of the cpu's actual ability.  But since a difference in bit
+		// and actual ability may represent a cmos/bios problem, we report it to the user.
 
 		if( sse3_result != !!cpucaps.hasStreamingSIMD3Extensions )
 		{
@@ -432,13 +381,15 @@ void cpudetectInit()
 
 	}
 	else
+	{
 		Console::Notice(
 			"Notice: Could not allocate memory for SSE3/4 detection.\n"
 			"\tRelying on CPUID results. [this is not an error]"
 		);
+	}
 	#endif
 
-	//////////////////////////////////////
+	////////////////////////////////////////////////////////////////////////////////////////////
 	//  Core Counting!
 
 	if( !cpucaps.hasMultiThreading || LogicalCoresPerPhysicalCPU == 0 )

common/src/x86emitter/fpu.cpp

@@ -17,7 +17,7 @@
  */
 
 #include "PrecompiledHeader.h"
-#include "ix86_legacy_internal.h"
+#include "legacy_internal.h"
 
 //------------------------------------------------------------------
 // FPU instructions

common/src/x86emitter/jmp.cpp

@@ -32,9 +32,7 @@
  */
 
 #include "PrecompiledHeader.h"
-
+#include "internal.h"
-#include "System.h"
-#include "ix86_internal.h"
 
 namespace x86Emitter {
 
@@ -73,6 +71,34 @@ xSmartJump::~xSmartJump()
 	m_baseptr = NULL;	// just in case (sometimes helps in debugging too)
 }
 
+// ------------------------------------------------------------------------
+// Emits a 32 bit jump, and returns a pointer to the 32 bit displacement.
+// (displacements should be assigned relative to the end of the jump instruction,
+// or in other words *(retval+1) )
+__emitinline s32* xJcc32( JccComparisonType comparison, s32 displacement )
+{
+	if( comparison == Jcc_Unconditional )
+		xWrite8( 0xe9 );
+	else
+	{
+		xWrite8( 0x0f );
+		xWrite8( 0x80 | comparison );
+	}
+	xWrite<s32>( displacement );
+
+	return ((s32*)xGetPtr()) - 1;
+}
+
+// ------------------------------------------------------------------------
+// Emits a 32 bit jump, and returns a pointer to the 8 bit displacement.
+// (displacements should be assigned relative to the end of the jump instruction,
+// or in other words *(retval+1) )
+__emitinline s8* xJcc8( JccComparisonType comparison, s8 displacement )
+{
+	xWrite8( (comparison == Jcc_Unconditional) ? 0xeb : (0x70 | comparison) );
+	xWrite<s8>( displacement );
+	return (s8*)xGetPtr() - 1;
+}
 
 // ------------------------------------------------------------------------
 // Writes a jump at the current x86Ptr, which targets a pre-established target address.
@@ -84,32 +110,26 @@ xSmartJump::~xSmartJump()
 __emitinline void Internal::xJccKnownTarget( JccComparisonType comparison, const void* target, bool slideForward )
 {
 	// Calculate the potential j8 displacement first, assuming an instruction length of 2:
-	sptr displacement8 = (sptr)target - ((sptr)xGetPtr() + 2);
+	sptr displacement8 = (sptr)target - (sptr)(xGetPtr() + 2);
 
 	const int slideVal = slideForward ? ((comparison == Jcc_Unconditional) ? 3 : 4) : 0;
 	displacement8 -= slideVal;
 
 	// if the following assert fails it means we accidentally used slideForard on a backward
 	// jump (which is an invalid operation since there's nothing to slide forward).
-	if( slideForward ) jASSUME( displacement8 >= 0 );
+	if( slideForward ) 
+	{
+		// jASSUME has an  else statement in it that would be abiguous without the brackets.
+		jASSUME( displacement8 >= 0 );
+	}
 	
 	if( is_s8( displacement8 ) )
-	{
+		xJcc8( comparison, displacement8 );
-		xWrite8( (comparison == Jcc_Unconditional) ? 0xeb : (0x70 | comparison) );
-		xWrite<s8>( displacement8 );
-	}
 	else
 	{
 		// Perform a 32 bit jump instead. :(
-
+		s32* bah = xJcc32( comparison );
-		if( comparison == Jcc_Unconditional )
+		*bah = (s32)target - (s32)xGetPtr();
-			xWrite8( 0xe9 );
-		else
-		{
-			xWrite8( 0x0f );
-			xWrite8( 0x80 | comparison );
-		}
-		xWrite<s32>( (sptr)target - ((sptr)xGetPtr() + 4) );
 	}
 }
 

common/src/x86emitter/legacy.cpp

@@ -29,8 +29,7 @@
 //------------------------------------------------------------------
 
 #include "PrecompiledHeader.h"
-#include "System.h"
+#include "legacy_internal.h"
-#include "ix86_legacy_internal.h"
 
 //////////////////////////////////////////////////////////////////////////////////////////
 //

common/src/x86emitter/legacy_sse.cpp

@@ -17,8 +17,8 @@
  */
 
 #include "PrecompiledHeader.h"
-#include "ix86_legacy_internal.h"
+#include "legacy_internal.h"
-#include "ix86_sse_helpers.h"
+#include "sse_helpers.h"
 
 using namespace x86Emitter;
 
@@ -209,6 +209,7 @@ DEFINE_LEGACY_CMP_OPCODE( NLT )
 DEFINE_LEGACY_CMP_OPCODE( NLE )
 DEFINE_LEGACY_CMP_OPCODE( ORD )
 
+DEFINE_LEGACY_SSSD_OPCODE( COMI )
 DEFINE_LEGACY_SSSD_OPCODE( UCOMI )
 DEFINE_LEGACY_RSQRT_OPCODE( RCP )
 DEFINE_LEGACY_RSQRT_OPCODE( RSQRT )

common/src/x86emitter/simd.cpp

@@ -17,9 +17,7 @@
  */
 
 #include "PrecompiledHeader.h"
-
+#include "internal.h"
-#include "System.h"
-#include "ix86_internal.h"
 
 namespace x86Emitter {
 
@@ -95,14 +93,9 @@ const MovhlImplAll<0x12> xMOVL;
 const MovhlImpl_RtoR<0x16>		xMOVLH;
 const MovhlImpl_RtoR<0x12>		xMOVHL;
 
-const SimdImpl_DestRegEither<0x66,0xdb> xPAND;
-const SimdImpl_DestRegEither<0x66,0xdf> xPANDN;
-const SimdImpl_DestRegEither<0x66,0xeb> xPOR;
-const SimdImpl_DestRegEither<0x66,0xef> xPXOR;
-
 const SimdImpl_AndNot			xANDN;
-
+const SimdImpl_COMI<true>		xCOMI;
-const SimdImpl_UcomI<0x66,0x2e> xUCOMI;
+const SimdImpl_COMI<false>		xUCOMI;
 const SimdImpl_rSqrt<0x53>		xRCP;
 const SimdImpl_rSqrt<0x52>		xRSQRT;
 const SimdImpl_Sqrt<0x51>		xSQRT;
@@ -111,6 +104,11 @@ const SimdImpl_MinMax<0x5f> xMAX;
 const SimdImpl_MinMax<0x5d>		xMIN;
 const SimdImpl_Shuffle<0xc6>	xSHUF;
 
+const SimdImpl_DestRegEither<0x66,0xdb> xPAND;
+const SimdImpl_DestRegEither<0x66,0xdf> xPANDN;
+const SimdImpl_DestRegEither<0x66,0xeb> xPOR;
+const SimdImpl_DestRegEither<0x66,0xef> xPXOR;
+
 // ------------------------------------------------------------------------
 
 // [SSE-4.1] Performs a bitwise AND of dest against src, and sets the ZF flag
@@ -238,19 +236,15 @@ __emitinline void xLDMXCSR( const u32* src )
 //
 
 __forceinline void xMOVDZX( const xRegisterSSE& to, const xRegister32& from )	{ xOpWrite0F( 0x66, 0x6e, to, from ); }
-__forceinline void xMOVDZX( const xRegisterSSE& to, const void* src )			{ xOpWrite0F( 0x66, 0x6e, to, src ); }
 __forceinline void xMOVDZX( const xRegisterSSE& to, const ModSibBase& src )		{ xOpWrite0F( 0x66, 0x6e, to, src ); }
 
 __forceinline void xMOVDZX( const xRegisterMMX& to, const xRegister32& from )	{ xOpWrite0F( 0x6e, to, from ); }
-__forceinline void xMOVDZX( const xRegisterMMX& to, const void* src )			{ xOpWrite0F( 0x6e, to, src ); }
 __forceinline void xMOVDZX( const xRegisterMMX& to, const ModSibBase& src )		{ xOpWrite0F( 0x6e, to, src ); }
 
 __forceinline void xMOVD( const xRegister32& to, const xRegisterSSE& from )		{ xOpWrite0F( 0x66, 0x7e, from, to ); }
-__forceinline void xMOVD( void* dest, const xRegisterSSE& from )				{ xOpWrite0F( 0x66, 0x7e, from, dest ); }
 __forceinline void xMOVD( const ModSibBase& dest, const xRegisterSSE& from )	{ xOpWrite0F( 0x66, 0x7e, from, dest ); }
 
 __forceinline void xMOVD( const xRegister32& to, const xRegisterMMX& from )		{ xOpWrite0F( 0x7e, from, to ); }
-__forceinline void xMOVD( void* dest, const xRegisterMMX& from )				{ xOpWrite0F( 0x7e, from, dest ); }
 __forceinline void xMOVD( const ModSibBase& dest, const xRegisterMMX& from )	{ xOpWrite0F( 0x7e, from, dest ); }
 
 
@@ -268,14 +262,10 @@ __forceinline void xMOVQZX( const xRegisterSSE& to, const void* src )			{ xOpWri
 
 // Moves lower quad of XMM to ptr64 (no bits are cleared)
 __forceinline void xMOVQ( const ModSibBase& dest, const xRegisterSSE& from )	{ xOpWrite0F( 0x66, 0xd6, from, dest ); }
-// Moves lower quad of XMM to ptr64 (no bits are cleared)
-__forceinline void xMOVQ( void* dest, const xRegisterSSE& from )				{ xOpWrite0F( 0x66, 0xd6, from, dest ); }
 
 __forceinline void xMOVQ( const xRegisterMMX& to, const xRegisterMMX& from )	{ if( to != from ) xOpWrite0F( 0x6f, to, from ); }
 __forceinline void xMOVQ( const xRegisterMMX& to, const ModSibBase& src )		{ xOpWrite0F( 0x6f, to, src ); }
-__forceinline void xMOVQ( const xRegisterMMX& to, const void* src )				{ xOpWrite0F( 0x6f, to, src ); }
 __forceinline void xMOVQ( const ModSibBase& dest, const xRegisterMMX& from )	{ xOpWrite0F( 0x7f, from, dest ); }
-__forceinline void xMOVQ( void* dest, const xRegisterMMX& from )				{ xOpWrite0F( 0x7f, from, dest ); }
 
 // This form of xMOVQ is Intel's adeptly named 'MOVQ2DQ'
 __forceinline void xMOVQ( const xRegisterSSE& to, const xRegisterMMX& from )	{ xOpWrite0F( 0xf3, 0xd6, to, from ); }
@@ -295,9 +285,7 @@ __forceinline void xMOVQ( const xRegisterMMX& to, const xRegisterSSE& from )
 
 #define IMPLEMENT_xMOVS( ssd, prefix ) \
 	__forceinline void xMOV##ssd( const xRegisterSSE& to, const xRegisterSSE& from )	{ if( to != from ) xOpWrite0F( prefix, 0x10, to, from ); } \
-	__forceinline void xMOV##ssd##ZX( const xRegisterSSE& to, const void* from )		{ xOpWrite0F( prefix, 0x10, to, from ); } \
 	__forceinline void xMOV##ssd##ZX( const xRegisterSSE& to, const ModSibBase& from )	{ xOpWrite0F( prefix, 0x10, to, from ); } \
-	__forceinline void xMOV##ssd( const void* to, const xRegisterSSE& from )			{ xOpWrite0F( prefix, 0x11, from, to ); } \
 	__forceinline void xMOV##ssd( const ModSibBase& to, const xRegisterSSE& from )		{ xOpWrite0F( prefix, 0x11, from, to ); }
 
 IMPLEMENT_xMOVS( SS, 0xf3 )
@@ -307,27 +295,17 @@ IMPLEMENT_xMOVS( SD, 0xf2 )
 // Non-temporal movs only support a register as a target (ie, load form only, no stores)
 //
 
-__forceinline void xMOVNTDQA( const xRegisterSSE& to, const void* from )
-{
-	xWrite32( 0x2A380f66 );
-	EmitSibMagic( to.Id, from );
-}
-
 __forceinline void xMOVNTDQA( const xRegisterSSE& to, const ModSibBase& from )
 {
 	xWrite32( 0x2A380f66 );
 	EmitSibMagic( to.Id, from );
 }
 
-__forceinline void xMOVNTDQ( void* to, const xRegisterSSE& from )				{ xOpWrite0F( 0x66, 0xe7, from, to ); }
 __forceinline void xMOVNTDQA( const ModSibBase& to, const xRegisterSSE& from )	{ xOpWrite0F( 0x66, 0xe7, from, to ); }
 
-__forceinline void xMOVNTPD( void* to, const xRegisterSSE& from )				{ xOpWrite0F( 0x66, 0x2b, from, to ); }
 __forceinline void xMOVNTPD( const ModSibBase& to, const xRegisterSSE& from )	{ xOpWrite0F( 0x66, 0x2b, from, to ); }
-__forceinline void xMOVNTPS( void* to, const xRegisterSSE& from )				{ xOpWrite0F( 0x2b, from, to ); }
 __forceinline void xMOVNTPS( const ModSibBase& to, const xRegisterSSE& from )	{ xOpWrite0F( 0x2b, from, to ); }
 
-__forceinline void xMOVNTQ( void* to, const xRegisterMMX& from )				{ xOpWrite0F( 0xe7, from, to ); }
 __forceinline void xMOVNTQ( const ModSibBase& to, const xRegisterMMX& from )	{ xOpWrite0F( 0xe7, from, to ); }
 
 // ------------------------------------------------------------------------
@@ -382,7 +360,6 @@ __forceinline void xPALIGNR( const xRegisterMMX& to, const xRegisterMMX& from, u
 //    with 0.0 if set to 1.
 //
 __emitinline void xINSERTPS( const xRegisterSSE& to, const xRegisterSSE& from, u8 imm8 )		{ xOpWrite0F( 0x66, 0x213a, to, from, imm8 ); }
-__emitinline void xINSERTPS( const xRegisterSSE& to, const u32* from, u8 imm8 )					{ xOpWrite0F( 0x66, 0x213a, to, from, imm8 ); }
 __emitinline void xINSERTPS( const xRegisterSSE& to, const ModSibStrict<u32>& from, u8 imm8 )	{ xOpWrite0F( 0x66, 0x213a, to, from, imm8 ); }
 
 // [SSE-4.1] Extract a single-precision floating-point value from src at an offset
@@ -390,7 +367,6 @@ __emitinline void xINSERTPS( const xRegisterSSE& to, const ModSibStrict<u32>& fr
 // is stored into the low 32-bits of dest (or at a 32-bit memory pointer).
 //
 __emitinline void xEXTRACTPS( const xRegister32& to, const xRegisterSSE& from, u8 imm8 )		{ xOpWrite0F( 0x66, 0x173a, to, from, imm8 ); }
-__emitinline void xEXTRACTPS( u32* dest, const xRegisterSSE& from, u8 imm8 )					{ xOpWrite0F( 0x66, 0x173a, from, dest, imm8 ); }
 __emitinline void xEXTRACTPS( const ModSibStrict<u32>& dest, const xRegisterSSE& from, u8 imm8 ){ xOpWrite0F( 0x66, 0x173a, from, dest, imm8 ); }
 
 }

common/src/x86emitter/tools.cpp

@@ -17,9 +17,8 @@
  */
 
 #include "PrecompiledHeader.h"
-
+#include "internal.h"
-#include "System.h"
+#include "tools.h"
-#include "ix86.h"
 
 // used to make sure regs don't get changed while in recompiler
 // use FreezeMMXRegs, FreezeXMMRegs
@@ -27,64 +26,12 @@
 u8 g_globalMMXSaved = 0;
 u8 g_globalXMMSaved = 0;
 
-PCSX2_ALIGNED16( static u64 g_globalMMXData[8] );
+PCSX2_ALIGNED16( u64 g_globalMMXData[8] );
-PCSX2_ALIGNED16( static u64 g_globalXMMData[2*iREGCNT_XMM] );
+PCSX2_ALIGNED16( u64 g_globalXMMData[2*iREGCNT_XMM] );
-
-namespace x86Emitter
-{
-	void xStoreReg( const xRegisterSSE& src )
-	{
-		xMOVDQA( &g_globalXMMData[src.Id*2], src );
-	}
-
-	void xRestoreReg( const xRegisterSSE& dest )
-	{
-		xMOVDQA( dest, &g_globalXMMData[dest.Id*2] );
-	}
-}
-
-
-/////////////////////////////////////////////////////////////////////
-// SetCPUState -- for assignment of SSE roundmodes and clampmodes.
-
-u32 g_sseMXCSR = DEFAULT_sseMXCSR; 
-u32 g_sseVUMXCSR = DEFAULT_sseVUMXCSR;
-
-void SetCPUState(u32 sseMXCSR, u32 sseVUMXCSR)
-{
-	//Msgbox::Alert("SetCPUState: Config.sseMXCSR = %x; Config.sseVUMXCSR = %x \n", Config.sseMXCSR, Config.sseVUMXCSR);
-	// SSE STATE //
-	// WARNING: do not touch unless you know what you are doing
-
-	sseMXCSR &= 0xffff; // clear the upper 16 bits since they shouldn't be set
-	sseVUMXCSR &= 0xffff;
-
-	if( !cpucaps.hasStreamingSIMD2Extensions )
-	{
-		// SSE1 cpus do not support Denormals Are Zero flag (throws an exception
-		// if we don't mask them off)
-
-		sseMXCSR &= ~0x0040;
-		sseVUMXCSR &= ~0x0040;
-	}
-
-	g_sseMXCSR = sseMXCSR;
-	g_sseVUMXCSR = sseVUMXCSR;
-
-#ifdef _MSC_VER
-	__asm ldmxcsr g_sseMXCSR; // set the new sse control
-#else
-    __asm__("ldmxcsr %[g_sseMXCSR]" : : [g_sseMXCSR]"m"(g_sseMXCSR) );
-#endif
-	//g_sseVUMXCSR = g_sseMXCSR|0x6000;
-}
 
 /////////////////////////////////////////////////////////////////////
 // MMX Register Freezing
-#ifndef __INTEL_COMPILER
+//
-extern "C"
-{
-#endif
 
 __forceinline void FreezeRegs(int save)
 {
@@ -92,14 +39,16 @@ __forceinline void FreezeRegs(int save)
 	FreezeMMXRegs(save);
 }
 
-__forceinline void FreezeMMXRegs_(int save)
+__forceinline void FreezeMMXRegs(int save)
 {
-	//DevCon::Notice("FreezeMMXRegs_(%d); [%d]\n", save, g_globalMMXSaved);
+	if( !g_EEFreezeRegs ) return;
-	assert( g_EEFreezeRegs );
 
-	if( save ) {
+	//DevCon::Notice("FreezeMMXRegs_(%d); [%d]\n", save, g_globalMMXSaved);
+
+	if( save )
+	{
 		g_globalMMXSaved++;
-		if( g_globalMMXSaved>1 )
+		if( g_globalMMXSaved > 1 )
 		{
 			//DevCon::Notice("MMX Already Saved!\n");
 			return;
@@ -178,10 +127,12 @@ __forceinline void FreezeMMXRegs_(int save)
 
 //////////////////////////////////////////////////////////////////////
 // XMM Register Freezing
-__forceinline void FreezeXMMRegs_(int save)
+//
+__forceinline void FreezeXMMRegs(int save)
 {
+	if( !g_EEFreezeRegs ) return;
+
 	//DevCon::Notice("FreezeXMMRegs_(%d); [%d]\n", save, g_globalXMMSaved);
-	assert( g_EEFreezeRegs );
 
 	if( save )
 	{
@@ -264,7 +215,4 @@ __forceinline void FreezeXMMRegs_(int save)
 #endif // _MSC_VER
 	}
 }
-#ifndef __INTEL_COMPILER
-}
-#endif
 

common/src/x86emitter/x86emitter.cpp

@@ -32,9 +32,11 @@
  */
 
 #include "PrecompiledHeader.h"
+#include "internal.h"
+
+// defined in tools.cpp
+PCSX2_ALIGNED16_EXTERN( u64 g_globalXMMData[2*iREGCNT_XMM] );
 
-#include "System.h"
-#include "ix86_internal.h"
 
 // ------------------------------------------------------------------------
 // Notes on Thread Local Storage:
@@ -96,6 +98,7 @@ const xAddressIndexer<u8> ptr8;
 // ------------------------------------------------------------------------
 
 template< typename OperandType > const xRegisterBase<OperandType> xRegisterBase<OperandType>::Empty;
+
 const xAddressReg xAddressReg::Empty;
 
 const xRegisterSSE
@@ -113,14 +116,14 @@ const xRegisterMMX
 const xAddressReg
 	eax( 0 ), ebx( 3 ),
 	ecx( 1 ), edx( 2 ),
-	esi( 6 ), edi( 7 ),
+	esp( 4 ), ebp( 5 ),
-	ebp( 5 ), esp( 4 );
+	esi( 6 ), edi( 7 );
 
 const xRegister16
 	ax( 0 ), bx( 3 ),
 	cx( 1 ), dx( 2 ),
-	si( 6 ), di( 7 ),
+	sp( 4 ), bp( 5 ),
-	bp( 5 ), sp( 4 );
+	si( 6 ), di( 7 );
 
 const xRegister8
 	al( 0 ),
@@ -130,10 +133,53 @@ const xRegister8
 	
 const xRegisterCL cl;
 
+const char *const x86_regnames_gpr8[8] =
+{
+	"al", "cl", "dl", "bl",
+	"ah", "ch", "dh", "bh"
+};
+
+const char *const x86_regnames_gpr16[8] =
+{
+	"ax", "cx", "dx", "bx",
+	"sp", "bp", "si", "di"
+};
+
+const char *const x86_regnames_gpr32[8] =
+{
+	"eax", "ecx", "edx", "ebx",
+	"esp", "ebp", "esi", "edi"
+};
+
+const char *const x86_regnames_sse[8] =
+{
+	"xmm0", "xmm1", "xmm2", "xmm3",
+	"xmm4", "xmm5", "xmm6", "xmm7"
+};
+
+const char *const x86_regnames_mmx[8] =
+{
+	"mm0", "mm1", "mm2", "mm3",
+	"mm4", "mm5", "mm6", "mm7"
+};
+
 //////////////////////////////////////////////////////////////////////////////////////////
 
 namespace Internal
 {
+
+	template< typename T >
+	const char* xGetRegName( const xRegister<T>& src )
+	{
+		if( src.IsEmpty() ) return "empty";
+		switch( sizeof(T) )
+		{
+		case 1: return tbl_regnames_gpr8[ src.Id ];
+		case 2: return tbl_regnames_gpr16[ src.Id ];
+		case 4: return tbl_regnames_gpr32[ src.Id ];
+		}
+	}
+
 	//////////////////////////////////////////////////////////////////////////////////////////
 	// Performance note: VC++ wants to use byte/word register form for the following
 	// ModRM/SibSB constructors when we use xWrite<u8>, and furthermore unrolls the
@@ -201,11 +247,6 @@ namespace Internal
 		xOpWrite0F( 0, opcode, instId, sib );
 	}
 
-	__emitinline void xOpWrite0F( u16 opcode, int instId, const void* data )
-	{
-		xOpWrite0F( 0, opcode, instId, data );
-	}
-
 
 	//////////////////////////////////////////////////////////////////////////////////////////
 	// returns TRUE if this instruction requires SIB to be encoded, or FALSE if the
@@ -674,11 +715,9 @@ __emitinline void xPUSH( const ModSibBase& from )
 }
 
 __forceinline void xPOP( xRegister32 from )		{ xWrite8( 0x58 | from.Id ); }
-__forceinline void xPOP( void* from )			{ xPOP( ptr[from] ); }
 
 __forceinline void xPUSH( u32 imm )				{ xWrite8( 0x68 ); xWrite32( imm ); }
 __forceinline void xPUSH( xRegister32 from )	{ xWrite8( 0x50 | from.Id ); }
-__forceinline void xPUSH( void* from )			{ xPUSH( ptr[from] ); }
 
 // pushes the EFLAGS register onto the stack
 __forceinline void xPUSHFD()					{ xWrite8( 0x9C ); }
@@ -710,4 +749,14 @@ __emitinline void xBSWAP( const xRegister32& to )
 	xWrite8( 0xC8 | to.Id );
 }
 
+__emitinline void xStoreReg( const xRegisterSSE& src )
+{
+	xMOVDQA( &g_globalXMMData[src.Id*2], src );
+}
+
+__emitinline void xRestoreReg( const xRegisterSSE& dest )
+{
+	xMOVDQA( dest, &g_globalXMMData[dest.Id*2] );
+}
+
 }

common/vsprops/CodeGen_Debug.vsprops

@@ -2,17 +2,13 @@
 <VisualStudioPropertySheet
 	ProjectType="Visual C++"
 	Version="8.00"
-	Name="debug"
+	Name="CodeGen_Debug"
 	>
 	<Tool
 		Name="VCCLCompilerTool"
 		Optimization="0"
-		PreprocessorDefinitions="_DEBUG"
+		PreprocessorDefinitions="PCSX2_DEBUG"
 		BasicRuntimeChecks="3"
-		RuntimeLibrary="1"
+		RuntimeLibrary="3"
-	/>
-	<Tool
-		Name="VCResourceCompilerTool"
-		PreprocessorDefinitions="_DEBUG"
 	/>
 </VisualStudioPropertySheet>

common/vsprops/CodeGen_Devel.vsprops

@@ -0,0 +1,18 @@
+<?xml version="1.0" encoding="windows-1250"?>
+<VisualStudioPropertySheet
+	ProjectType="Visual C++"
+	Version="8.00"
+	Name="CodeGen_Devel"
+	>
+	<Tool
+		Name="VCCLCompilerTool"
+		Optimization="2"
+		InlineFunctionExpansion="2"
+		EnableIntrinsicFunctions="true"
+		FavorSizeOrSpeed="1"
+		EnableFiberSafeOptimizations="true"
+		PreprocessorDefinitions="PCSX2_DEVEL;PCSX2_DEVBUILD;NDEBUG"
+		RuntimeLibrary="2"
+		BufferSecurityCheck="false"
+	/>
+</VisualStudioPropertySheet>

common/vsprops/CodeGen_Release.vsprops

@@ -2,7 +2,7 @@
 <VisualStudioPropertySheet
 	ProjectType="Visual C++"
 	Version="8.00"
-	Name="GlobalOptimizations"
+	Name="CodeGen_Release"
 	>
 	<Tool
 		Name="VCCLCompilerTool"
@@ -11,9 +11,12 @@
 		EnableIntrinsicFunctions="true"
 		FavorSizeOrSpeed="1"
 		OmitFramePointers="true"
+		EnableFiberSafeOptimizations="true"
 		WholeProgramOptimization="true"
+		PreprocessorDefinitions="NDEBUG"
 		StringPooling="true"
 		MinimalRebuild="true"
+		RuntimeLibrary="2"
 		BufferSecurityCheck="false"
 	/>
 	<Tool

common/vsprops/CommonLibrary.vsprops

@@ -0,0 +1,33 @@
+<?xml version="1.0" encoding="windows-1250"?>
+<VisualStudioPropertySheet
+	ProjectType="Visual C++"
+	Version="8.00"
+	Name="CommonLibrary"
+	OutputDirectory="..\..\deps\$(PlatformName)\$(ConfigurationName)"
+	IntermediateDirectory="$(PlatformName)\$(ConfigurationName)"
+	>
+	<Tool
+		Name="VCCLCompilerTool"
+		AdditionalIncludeDirectories="&quot;$(SvnRootDir)\common\include\$(ProjectName)&quot;;&quot;$(SvnRootDir)\common\include&quot;;&quot;$(SvnRootDir)\3rdparty\wxWidgets\include&quot;"
+		PreprocessorDefinitions="__WIN32__;WIN32;_WINDOWS;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE"
+		StructMemberAlignment="5"
+		RuntimeTypeInfo="false"
+		WarningLevel="3"
+		DebugInformationFormat="3"
+		CompileAs="0"
+	/>
+	<Tool
+		Name="VCLinkerTool"
+		LinkIncremental="1"
+		GenerateDebugInformation="true"
+		SubSystem="2"
+	/>
+	<UserMacro
+		Name="ProjectRootDir"
+		Value="$(ProjectDir)\..\.."
+	/>
+	<UserMacro
+		Name="SvnRootDir"
+		Value="$(ProjectRootDir)\.."
+	/>
+</VisualStudioPropertySheet>

common/vsprops/pthreads.vsprops

@@ -4,17 +4,13 @@
 	Version="8.00"
 	Name="w32pthreads"
 	>
-	<Tool
-		Name="VCCLCompilerTool"
-		AdditionalIncludeDirectories="&quot;$(SvnRootDir)\3rdparty&quot;"
-		PreprocessorDefinitions="PTW32_STATIC_LIB;WIN32_PTHREADS;__CLEANUP_SEH"
-	/>
 	<Tool
 		Name="VCCLCompilerTool"
 		AdditionalIncludeDirectories="&quot;$(SvnRootDir)\3rdparty\w32pthreads\include&quot;"
+		PreprocessorDefinitions="WIN32_PTHREADS;__CLEANUP_SEH"
 	/>
 	<Tool
 		Name="VCLinkerTool"
-		AdditionalDependencies="pthreads.lib Ws2_32.lib"
+		AdditionalDependencies="w32pthreads.lib"
 	/>
 </VisualStudioPropertySheet>

pcsx2/3rdparty

@@ -0,0 +1 @@
+../3rdparty/

pcsx2/CDVD/CDVD.cpp

@@ -23,6 +23,7 @@
 
 #include "IopCommon.h"
 #include "CDVDiso.h"
+#include "CDVD_internal.h"
 
 static cdvdStruct cdvd;
 
@@ -77,7 +78,7 @@ FILE *_cdvdOpenMechaVer() {
 	// get the name of the bios file
 	
 	// use the bios filename to get the name of the mecha ver file
-	// [TODO] : Upgrade this to use wxString!
+	// [TODO] : Upgrade this to use std::string!
 
 	strcpy(file, g_Conf.Files.Bios().ToAscii().data() );
 	
@@ -91,7 +92,7 @@ FILE *_cdvdOpenMechaVer() {
 	// if file doesnt exist, create empty one
 	fd = fopen(file, "r+b");
 	if (fd == NULL) {
-		Console::Notice( "MEC File Not Found, Creating Blank File..." );
+		Console::Notice("MEC File Not Found , Creating Blank File");
 		fd = fopen(file, "wb");
 		if (fd == NULL)
 		{
@@ -399,7 +400,6 @@ void cdvdReadKey(u8 arg0, u16 arg1, u32 arg2, u8* key) {
 	{
 		ElfCRC = loadElfCRC( str );
 		ElfApplyPatches();
-		LoadGameSpecificSettings();
 		GSsetGameCRC( ElfCRC, 0 );
 	}
 }
@@ -536,12 +536,6 @@ void cdvdReset()
 	cdvd.Action = cdvdAction_None;
 	cdvd.ReadTime = cdvdBlockReadTime( MODE_DVDROM );
 
-    // any random valid date will do
-    //cdvd.RTC.hour = 1;
-    //cdvd.RTC.day = 25;
-    //cdvd.RTC.month = 5;
-    //cdvd.RTC.year = 7; //2007
-
 	wxDateTime curtime( wxDateTime::GetTimeNow() );
 	cdvd.RTC.second = (u8)curtime.GetSecond();
 	cdvd.RTC.minute = (u8)curtime.GetMinute();
@@ -580,13 +574,24 @@ void cdvdNewDiskCB()
 {
 	cdvd.Type = CDVDgetDiskType();
 	
+	wxString str;
+	int result = GetPS2ElfName(str);
+
 	if (cdvd.Type == CDVD_TYPE_PS2CD) 
 	{
-		wxString str;
 		
 		 // Does the SYSTEM.CNF file only say "BOOT="? PS1 CD then.
-		if(GetPS2ElfName(str) == 1) cdvd.Type = CDVD_TYPE_PSCD;
+		if(result == 1) cdvd.Type = CDVD_TYPE_PSCD;
 	}
+	
+	// Now's a good time to reload the ELF info...
+	if( ElfCRC == 0 )
+	{
+		ElfCRC = loadElfCRC( str.ToAscii().data() );
+		ElfApplyPatches();
+		GSsetGameCRC( ElfCRC, 0 );
+	}
+
 }
 
 void mechaDecryptBytes( u32 madr, int size )
@@ -823,7 +828,7 @@ static uint cdvdStartSeek( uint newsector, CDVD_MODE_TYPE mode )
 {
 	cdvd.SeekToSector = newsector;
 
-	uint delta = abs((int)cdvd.SeekToSector - (int)cdvd.Sector);
+	uint delta = abs( (s32)(cdvd.SeekToSector - cdvd.Sector) );
 	uint seektime;
 
 	cdvd.Ready = CDVD_NOTREADY;

pcsx2/CDVD/CDVD.h

@@ -0,0 +1,103 @@
+/*  Pcsx2 - Pc Ps2 Emulator
+ *  Copyright (C) 2002-2009  Pcsx2 Team
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ */
+
+#pragma once
+
+#include "IopCommon.h"
+
+#define btoi(b)		((b)/16*10 + (b)%16)		/* BCD to u_char */
+#define itob(i)		((i)/10*16 + (i)%10)		/* u_char to BCD */
+
+struct cdvdRTC {
+	u8 status;
+	u8 second;
+	u8 minute;
+	u8 hour;
+	u8 pad;
+	u8 day;
+	u8 month;
+	u8 year;
+};
+
+struct cdvdStruct {
+	u8 nCommand;
+	u8 Ready;
+	u8 Error;
+	u8 PwOff;
+	u8 Status;
+	u8 Type;
+	u8 sCommand;
+	u8 sDataIn;
+	u8 sDataOut;
+	u8 HowTo;
+
+	u8 Param[32];
+	u8 Result[32];
+
+	u8 ParamC;
+	u8 ParamP;
+	u8 ResultC;
+	u8 ResultP;
+
+	u8 CBlockIndex;
+	u8 COffset;
+	u8 CReadWrite;
+	u8 CNumBlocks;
+
+	int RTCcount;
+	cdvdRTC RTC;
+
+	u32 Sector;
+	int nSectors;
+	int Readed; // change to bool. --arcum42
+	int Reading; // same here.
+	int ReadMode;
+	int BlockSize; // Total bytes transfered at 1x speed
+	int Speed;
+	int RetryCnt;
+	int RetryCntP;
+	int RErr;
+	int SpindlCtrl;
+
+	u8 Key[16];
+	u8 KeyXor;
+	u8 decSet;
+
+	u8  mg_buffer[65536];
+	int mg_size;
+	int mg_maxsize;
+	int mg_datatype;//0-data(encrypted); 1-header
+	u8	mg_kbit[16];//last BIT key 'seen'
+	u8	mg_kcon[16];//last content key 'seen'
+
+	u8  Action;			// the currently scheduled emulated action
+	u32 SeekToSector;	// Holds the destination sector during seek operations.
+	u32 ReadTime;		// Avg. time to read one block of data (in Iop cycles)
+	bool Spinning;		// indicates if the Cdvd is spinning or needs a spinup delay
+};
+
+extern void cdvdReset();
+extern void cdvdVsync();
+extern void cdvdActionInterrupt();
+extern void cdvdReadInterrupt();
+
+// We really should not have a function with the exact same name as a callback except for case!
+extern void cdvdNewDiskCB();
+extern u8 cdvdRead(u8 key);
+extern void cdvdWrite(u8 key, u8 rt);
+

pcsx2/CDVD/CDVD_internal.h

@@ -16,78 +16,8 @@
  *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
  */
  
-#ifndef __CDVD_H__
+#ifndef __CDVD_INTERNAL_H__
-#define __CDVD_H__
+#define __CDVD_INTERNAL_H__
-
-#include "IopCommon.h"
-
-struct cdvdRTC {
-	u8 status;
-	u8 second;
-	u8 minute;
-	u8 hour;
-	u8 pad;
-	u8 day;
-	u8 month;
-	u8 year;
-};
-
-struct cdvdStruct {
-	u8 nCommand;
-	u8 Ready;
-	u8 Error;
-	u8 PwOff;
-	u8 Status;
-	u8 Type;
-	u8 sCommand;
-	u8 sDataIn;
-	u8 sDataOut;
-	u8 HowTo;
-
-	u8 Param[32];
-	u8 Result[32];
-
-	u8 ParamC;
-	u8 ParamP;
-	u8 ResultC;
-	u8 ResultP;
-
-	u8 CBlockIndex;
-	u8 COffset;
-	u8 CReadWrite;
-	u8 CNumBlocks;
-
-	int RTCcount;
-	cdvdRTC RTC;
-
-	u32 Sector;
-	int nSectors;
-	int Readed; // change to bool. --arcum42
-	int Reading; // same here.
-	int ReadMode;
-	int BlockSize; // Total bytes transfered at 1x speed
-	int Speed;
-	int RetryCnt;
-	int RetryCntP;
-	int RErr;
-	int SpindlCtrl;
-
-	u8 Key[16];
-	u8 KeyXor;
-	u8 decSet;
-
-	u8  mg_buffer[65536];
-	int mg_size;
-	int mg_maxsize;
-	int mg_datatype;//0-data(encrypted); 1-header
-	u8	mg_kbit[16];//last BIT key 'seen'
-	u8	mg_kcon[16];//last content key 'seen'
-
-	u8  Action;			// the currently scheduled emulated action
-	u32 SeekToSector;	// Holds the destination sector during seek operations.
-	u32 ReadTime;		// Avg. time to read one block of data (in Iop cycles)
-	bool Spinning;		// indicates if the Cdvd is spinning or needs a spinup delay
-};
 
 /*
 Interrupts - values are flag bits.
@@ -191,7 +121,6 @@ static const uint tbl_ContigiousSeekDelta[3] =
 static const uint PSX_CD_READSPEED = 153600;   // 1 Byte Time @ x1 (150KB = cd x 1)
 static const uint PSX_DVD_READSPEED = 1382400 + 256000; // normal is 1 Byte Time @ x1 (1350KB = dvd x 1).
 
-
 // Legacy Note: FullSeek timing causes many games to load very slow, but it likely not the real problem.
 // Games breaking with it set to PSXCLK*40 : "wrath unleashed" and "Shijou Saikyou no Deshi Kenichi".
 
@@ -301,15 +230,4 @@ static NVMLayout nvmlayouts[NVM_FORMAT_MAX] =
 	{0x146,  0x270, 0x2B0, 0x200, 0x1C8, 0x1E0, 0x1B0, 0x180, 0x198},	// eeproms from bios v1.70 and up
 };
 
-#define btoi(b)		((b)/16*10 + (b)%16)		/* BCD to u_char */
+#endif
-#define itob(i)		((i)/10*16 + (i)%10)		/* u_char to BCD */
-
-void cdvdReset();
-void cdvdVsync();
-extern void  cdvdActionInterrupt();
-extern void  cdvdReadInterrupt();
-void cdvdNewDiskCB();
-u8   cdvdRead(u8 key);
-void cdvdWrite(u8 key, u8 rt);
-
-#endif /* __CDVD_H__ */