Patched libpng to address CVE-2019-7317

Keep `_FILE_OFFSET_BITS` undefined in Android builds.

.editorconfig

@@ -0,0 +1,31 @@
+# https://editorconfig.org/
+
+root = true
+
+[*]
+end_of_line = lf
+charset = utf-8
+trim_trailing_whitespace = true
+insert_final_newline = true
+indent_style = space
+indent_size = 4
+
+[{CMakeLists.*,*.cmake}]
+indent_style = space
+indent_size = 2
+
+[Makefile]
+indent_style = tab
+
+[*.{bat,cmd,cmd.*}]
+end_of_line = crlf
+indent_style = space
+indent_size = 2
+
+[*.{ps1,ps1.*}]
+end_of_line = crlf
+indent_style = space
+indent_size = 4
+
+[*.{md,markdown}]
+indent_size = 2

.gitattributes

@@ -43,6 +43,8 @@
 
 # reST underlines/overlines can look like conflict markers
 *.rst          text conflict-marker-size=80
+# Markdown heading markers can look like conflict markers
+*.md           text conflict-marker-size=200
 
 *.cmake         text whitespace=tabwidth=2
 *.cmakein       text whitespace=tabwidth=2
@@ -79,6 +81,8 @@ org.eclipse.jdt.core.prefs -text whitespace=cr-at-eol merge=union
 *.cmd       text eol=crlf
 *.cmd.tmpl  text eol=crlf
 *.dsp       text eol=crlf -whitespace
+*.ps1       text eol=crlf
+*.ps1.in    text eol=crlf
 *.sln       text eol=crlf -whitespace
 *.vcproj    text eol=crlf -whitespace merge=union
 *.vcxproj   text eol=crlf -whitespace merge=union

.github/ISSUE_TEMPLATE.md

@@ -2,6 +2,14 @@
 If you have a question rather than reporting a bug please go to http://answers.opencv.org where you get much faster responses.
 If you need further assistance please read [How To Contribute](https://github.com/opencv/opencv/wiki/How_to_contribute).
 
+Please:
+
+* Read the documentation to test with the latest developer build.
+* Check if other person has already created the same issue to avoid duplicates. You can comment on it if there already is an issue.
+* Try to be as detailed as possible in your report.
+* Report only one problem per created issue.
+
+
 This is a template helping you to create an issue which can be processed as quickly as possible. This is the bug reporting section for the OpenCV library.
 -->
 
@@ -27,4 +35,4 @@ This is a template helping you to create an issue which can be processed as quic
     // C++ code example
     ```
  or attach as .txt or .zip file
--->
+-->

.gitignore

@@ -1,10 +1,11 @@
+# ignore dot files/directories
+.*
+!.gitignore
+
 *.autosave
 *.pyc
 *.user
 *~
-.*.swp
-.DS_Store
-.sw[a-z]
 Thumbs.db
 tags
 tegra/
@@ -21,4 +22,3 @@ bin/
 *.log
 *.tlog
 build
-.cache

.tgitconfig

@@ -1,2 +0,0 @@
-[tgit]
-    icon = doc/opencv.ico

3rdparty/carotene/CMakeLists.txt

@@ -37,6 +37,4 @@ if(WITH_NEON)
     target_compile_definitions(carotene_objs PRIVATE "-DWITH_NEON")
 endif()
 
-set_target_properties(carotene_objs PROPERTIES POSITION_INDEPENDENT_CODE TRUE)
-
 add_library(carotene STATIC EXCLUDE_FROM_ALL "$<TARGET_OBJECTS:carotene_objs>")

3rdparty/carotene/hal/CMakeLists.txt

@@ -3,8 +3,6 @@ cmake_minimum_required(VERSION 2.8.8 FATAL_ERROR)
 include(CheckCCompilerFlag)
 include(CheckCXXCompilerFlag)
 
-set(CMAKE_POSITION_INDEPENDENT_CODE ON)
-
 set(TEGRA_HAL_DIR "${CMAKE_CURRENT_SOURCE_DIR}")
 set(CAROTENE_DIR "${TEGRA_HAL_DIR}/../")
 
@@ -14,18 +12,20 @@ elseif (CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64.*|AARCH64.*")
   set(AARCH64 TRUE)
 endif()
 
+ocv_warnings_disable(CMAKE_CXX_FLAGS -Wunused-function)
+
 set(TEGRA_COMPILER_FLAGS "")
 
-if(CMAKE_COMPILER_IS_GNUCXX)
+if(CV_GCC OR CV_CLANG)
   # Generate unwind information even for functions that can't throw/propagate exceptions.
   # This lets debuggers and such get non-broken backtraces for such functions, even without debugging symbols.
   list(APPEND TEGRA_COMPILER_FLAGS -funwind-tables)
 endif()
 
-if(CMAKE_COMPILER_IS_GNUCXX)
+if(CV_GCC OR CV_CLANG)
   if(X86 OR ARMEABI_V6 OR (MIPS AND ANDROID_COMPILER_VERSION VERSION_LESS "4.6"))
     list(APPEND TEGRA_COMPILER_FLAGS -fweb -fwrapv -frename-registers -fsched-stalled-insns-dep=100 -fsched-stalled-insns=2)
-  elseif(CMAKE_COMPILER_IS_CLANGCXX)
+  elseif(CV_CLANG)
     list(APPEND TEGRA_COMPILER_FLAGS -fwrapv)
   else()
     list(APPEND TEGRA_COMPILER_FLAGS -fweb -fwrapv -frename-registers -fsched2-use-superblocks -fsched2-use-traces
@@ -42,7 +42,7 @@ set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${TEGRA_COMPILER_FLAGS}")
 set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${TEGRA_COMPILER_FLAGS}")
 
 if(ARMEABI_V7A)
-  if (CMAKE_COMPILER_IS_GNUCXX)
+  if(CV_GCC OR CV_CLANG)
     set( CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-tree-vectorize" )
     set( CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fno-tree-vectorize" )
   endif()
@@ -55,7 +55,7 @@ endif()
 set(CAROTENE_NS "carotene_o4t" CACHE STRING "" FORCE)
 
 function(compile_carotene)
-  if(ENABLE_NEON)
+  if(";${CPU_BASELINE_FINAL};" MATCHES ";NEON;")
     set(WITH_NEON ON)
   endif()
 
@@ -85,7 +85,7 @@ include_directories("${CAROTENE_DIR}/include")
 get_target_property(carotene_defs carotene_objs INTERFACE_COMPILE_DEFINITIONS)
 set_property(DIRECTORY APPEND PROPERTY COMPILE_DEFINITIONS ${carotene_defs})
 
-  if (CMAKE_COMPILER_IS_GNUCXX)
+  if(CV_GCC)
     # allow more inlines - these parameters improve performance for:
     #   matchTemplate about 5-10%
     #   goodFeaturesToTrack 10-20%
@@ -95,7 +95,6 @@ set_property(DIRECTORY APPEND PROPERTY COMPILE_DEFINITIONS ${carotene_defs})
   endif()
 
 add_library(tegra_hal STATIC $<TARGET_OBJECTS:carotene_objs>)
-set_target_properties(tegra_hal PROPERTIES POSITION_INDEPENDENT_CODE TRUE)
 set_target_properties(tegra_hal PROPERTIES ARCHIVE_OUTPUT_DIRECTORY ${3P_LIBRARY_OUTPUT_PATH})
 set(OPENCV_SRC_DIR "${CMAKE_SOURCE_DIR}")
 if(NOT BUILD_SHARED_LIBS)

3rdparty/carotene/hal/tegra_hal.hpp

@@ -67,7 +67,7 @@
                  size_t src2_step;
 
 #define DST_ARG1 DT * dst1_data_, size_t dst1_step_,
-#define DST_STORE1 dst1_data(dst1_data_), dst1_step(dst1_step_), 
+#define DST_STORE1 dst1_data(dst1_data_), dst1_step(dst1_step_),
 #define DST_VAR1 DT * dst1_data; \
                  size_t dst1_step;
 
@@ -1073,7 +1073,7 @@ struct FilterCtx
 inline int TEGRA_FILTERINIT(cvhalFilter2D **context, uchar *kernel_data, size_t kernel_step, int kernel_type, int kernel_width, int kernel_height,
                             int max_width, int max_height, int src_type, int dst_type, int borderType, double delta, int anchor_x, int anchor_y, bool allowSubmatrix, bool allowInplace)
 {
-    if(!context || !kernel_data || allowSubmatrix || allowInplace || 
+    if(!context || !kernel_data || allowSubmatrix || allowInplace ||
        src_type != CV_8UC1 || dst_type != CV_8UC1 ||
        delta != 0 || anchor_x != kernel_width / 2 || anchor_y != kernel_height / 2 )
         return CV_HAL_ERROR_NOT_IMPLEMENTED;
@@ -1105,7 +1105,7 @@ inline int TEGRA_FILTERINIT(cvhalFilter2D **context, uchar *kernel_data, size_t
         return CV_HAL_ERROR_NOT_IMPLEMENTED;
     }
 
-    if(!CAROTENE_NS::isConvolutionSupported(CAROTENE_NS::Size2D(max_width, max_height), ctx->ksize, ctx->border))    
+    if(!CAROTENE_NS::isConvolutionSupported(CAROTENE_NS::Size2D(max_width, max_height), ctx->ksize, ctx->border))
     {
         delete ctx;
         return CV_HAL_ERROR_NOT_IMPLEMENTED;
@@ -1212,7 +1212,7 @@ inline int TEGRA_SEPFILTERINIT(cvhalFilter2D **context, int src_type, int dst_ty
         return CV_HAL_ERROR_NOT_IMPLEMENTED;
     }
 
-    if(!CAROTENE_NS::isSeparableFilter3x3Supported(CAROTENE_NS::Size2D(16, 16), ctx->border, 3, 3))    
+    if(!CAROTENE_NS::isSeparableFilter3x3Supported(CAROTENE_NS::Size2D(16, 16), ctx->border, 3, 3))
     {
         delete ctx;
         return CV_HAL_ERROR_NOT_IMPLEMENTED;
@@ -1341,8 +1341,8 @@ inline int TEGRA_MORPHINIT(cvhalFilter2D **context, int operation, int src_type,
     ctx->anchor_y = anchor_y;
     switch(operation)
     {
-    case MORPH_ERODE:
-    case MORPH_DILATE:
+    case CV_HAL_MORPH_ERODE:
+    case CV_HAL_MORPH_DILATE:
         ctx->operation = operation;
         break;
     default:
@@ -1355,7 +1355,7 @@ inline int TEGRA_MORPHINIT(cvhalFilter2D **context, int operation, int src_type,
         ctx->border = CAROTENE_NS::BORDER_MODE_CONSTANT;
         if( borderValue[0] == DBL_MAX && borderValue[1] == DBL_MAX && borderValue[2] == DBL_MAX && borderValue[3] == DBL_MAX )
         {
-            if( operation == MORPH_ERODE )
+            if( operation == CV_HAL_MORPH_ERODE )
                 for(int i = 0; i < ctx->channels; ++i)
                     ctx->borderValues[i] = (CAROTENE_NS::u8)UCHAR_MAX;
             else
@@ -1404,14 +1404,14 @@ inline int TEGRA_MORPHFREE(cvhalFilter2D *context)
 ( \
     (void)dst_full_width, (void)dst_full_height, (void)dst_roi_x, (void)dst_roi_y, \
     context && CAROTENE_NS::isSupportedConfiguration() ? \
-        ((MorphCtx*)context)->operation == MORPH_ERODE ? \
+        ((MorphCtx*)context)->operation == CV_HAL_MORPH_ERODE ? \
         CAROTENE_NS::erode(CAROTENE_NS::Size2D(width, height), ((MorphCtx*)context)->channels, \
                            src_data, src_step, dst_data, dst_step, \
                            ((MorphCtx*)context)->ksize, ((MorphCtx*)context)->anchor_x, ((MorphCtx*)context)->anchor_y, \
                            ((MorphCtx*)context)->border, ((MorphCtx*)context)->border, ((MorphCtx*)context)->borderValues, \
                            CAROTENE_NS::Margin(src_roi_x, src_full_width - width - src_roi_x, src_roi_y, src_full_height - height - src_roi_y)), \
         CV_HAL_ERROR_OK : \
-        ((MorphCtx*)context)->operation == MORPH_DILATE ? \
+        ((MorphCtx*)context)->operation == CV_HAL_MORPH_DILATE ? \
         CAROTENE_NS::dilate(CAROTENE_NS::Size2D(width, height), ((MorphCtx*)context)->channels, \
                             src_data, src_step, dst_data, dst_step, \
                             ((MorphCtx*)context)->ksize, ((MorphCtx*)context)->anchor_x, ((MorphCtx*)context)->anchor_y, \
@@ -1531,7 +1531,7 @@ class TegraCvtColor_##name##_Invoker : public cv::ParallelLoopBody \
 public: \
     TegraCvtColor_##name##_Invoker(const uchar * src_data_, size_t src_step_, uchar * dst_data_, size_t dst_step_, int width_, int height_) : \
         cv::ParallelLoopBody(), src_data(src_data_), src_step(src_step_), dst_data(dst_data_), dst_step(dst_step_), width(width_), height(height_) {} \
-    virtual void operator()(const cv::Range& range) const \
+    virtual void operator()(const cv::Range& range) const CV_OVERRIDE \
     { \
         CAROTENE_NS::func(CAROTENE_NS::Size2D(width, range.end-range.start), __VA_ARGS__); \
     } \

3rdparty/carotene/src/channel_extract.cpp

@@ -49,7 +49,7 @@ void extract2(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#ifndef ANDROID
+#ifndef __ANDROID__
     size_t roiw32 = size.width >= 31 ? size.width - 31 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -60,7 +60,7 @@ void extract2(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
         for (; dj < roiw32; sj += 64, dj += 32)
         {
             internal::prefetch(src + sj);
@@ -101,7 +101,7 @@ void extract3(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#ifndef ANDROID
+#ifndef __ANDROID__
     size_t roiw32 = size.width >= 31 ? size.width - 31 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -112,7 +112,7 @@ void extract3(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
         for (; dj < roiw32; sj += 96, dj += 32)
         {
             internal::prefetch(src + sj);
@@ -153,7 +153,7 @@ void extract4(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#ifndef ANDROID
+#ifndef __ANDROID__
     size_t roiw32 = size.width >= 31 ? size.width - 31 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -164,7 +164,7 @@ void extract4(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
         for (; dj < roiw32; sj += 128, dj += 32)
         {
             internal::prefetch(src + sj);
@@ -231,7 +231,7 @@ void extract4(const Size2D &size,
                  srcStride == dst2Stride && \
                  srcStride == dst3Stride &&
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 
 #define SPLIT_ASM2(sgn, bits) __asm__ ( \
                                           "vld2." #bits " {d0, d2}, [%[in0]]            \n\t" \
@@ -351,7 +351,7 @@ void extract4(const Size2D &size,
     }                                                                                                   \
 }
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 
 #define ALPHA_QUAD(sgn, bits) { \
                                   internal::prefetch(src + sj); \

3rdparty/carotene/src/channels_combine.cpp

@@ -77,7 +77,7 @@ namespace CAROTENE_NS {
                  dstStride == src2Stride && \
                  dstStride == src3Stride &&
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 
 #define MERGE_ASM2(sgn, bits) __asm__ ( \
                                           "vld1." #bits " {d0-d1}, [%[in0]]             \n\t" \
@@ -240,7 +240,7 @@ void combineYUYV(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#ifndef ANDROID
+#ifndef __ANDROID__
     size_t roiw32 = size.width >= 31 ? size.width - 31 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -253,7 +253,7 @@ void combineYUYV(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t syj = 0u, sj = 0u, dj = 0u;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
         for (; sj < roiw32; sj += 32, syj += 64, dj += 128)
         {
             internal::prefetch(srcy + syj);
@@ -317,7 +317,7 @@ void combineUYVY(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#ifndef ANDROID
+#ifndef __ANDROID__
     size_t roiw32 = size.width >= 31 ? size.width - 31 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -330,7 +330,7 @@ void combineUYVY(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t syj = 0u, sj = 0u, dj = 0u;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
         for (; sj < roiw32; sj += 32, syj += 64, dj += 128)
         {
             internal::prefetch(srcy + syj);

3rdparty/carotene/src/colorconvert.cpp

@@ -49,12 +49,12 @@ namespace {
 
 enum
 {
-    SHIFT = 14,
+    SHIFT = 15,
     SHIFT_DELTA = 1 << (SHIFT - 1),
 
-    R2Y_BT601   = 4899,
-    G2Y_BT601   = 9617,
-    B2Y_BT601   = 1868,
+    R2Y_BT601   = 9798,
+    G2Y_BT601   = 19235,
+    B2Y_BT601   = 3735,
 
     R2Y_BT709   = 3483,
     G2Y_BT709   = 11718,
@@ -97,7 +97,7 @@ void rgb2gray(const Size2D &size, COLOR_SPACE color_space,
     const u32 G2Y = color_space == COLOR_SPACE_BT601 ? G2Y_BT601 : G2Y_BT709;
     const u32 B2Y = color_space == COLOR_SPACE_BT601 ? B2Y_BT601 : B2Y_BT709;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register int16x4_t v_r2y asm ("d31") = vmov_n_s16(R2Y);
     register int16x4_t v_g2y asm ("d30") = vmov_n_s16(G2Y);
     register int16x4_t v_b2y asm ("d29") = vmov_n_s16(B2Y);
@@ -116,7 +116,7 @@ void rgb2gray(const Size2D &size, COLOR_SPACE color_space,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
         for (; dj < roiw8; sj += 24, dj += 8)
         {
             internal::prefetch(src + sj);
@@ -198,7 +198,7 @@ void rgbx2gray(const Size2D &size, COLOR_SPACE color_space,
     const u32 G2Y = color_space == COLOR_SPACE_BT601 ? G2Y_BT601 : G2Y_BT709;
     const u32 B2Y = color_space == COLOR_SPACE_BT601 ? B2Y_BT601 : B2Y_BT709;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register int16x4_t v_r2y asm ("d31") = vmov_n_s16(R2Y);
     register int16x4_t v_g2y asm ("d30") = vmov_n_s16(G2Y);
     register int16x4_t v_b2y asm ("d29") = vmov_n_s16(B2Y);
@@ -217,7 +217,7 @@ void rgbx2gray(const Size2D &size, COLOR_SPACE color_space,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
         for (; dj < roiw8; sj += 32, dj += 8)
         {
             internal::prefetch(src + sj);
@@ -300,7 +300,7 @@ void bgr2gray(const Size2D &size, COLOR_SPACE color_space,
     const u32 G2Y = color_space == COLOR_SPACE_BT601 ? G2Y_BT601 : G2Y_BT709;
     const u32 B2Y = color_space == COLOR_SPACE_BT601 ? B2Y_BT601 : B2Y_BT709;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register int16x4_t v_r2y asm ("d31") = vmov_n_s16(R2Y);
     register int16x4_t v_g2y asm ("d30") = vmov_n_s16(G2Y);
     register int16x4_t v_b2y asm ("d29") = vmov_n_s16(B2Y);
@@ -319,7 +319,7 @@ void bgr2gray(const Size2D &size, COLOR_SPACE color_space,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
         for (; dj < roiw8; sj += 24, dj += 8)
         {
             internal::prefetch(src + sj);
@@ -402,7 +402,7 @@ void bgrx2gray(const Size2D &size, COLOR_SPACE color_space,
     const u32 G2Y = color_space == COLOR_SPACE_BT601 ? G2Y_BT601 : G2Y_BT709;
     const u32 B2Y = color_space == COLOR_SPACE_BT601 ? B2Y_BT601 : B2Y_BT709;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register int16x4_t v_r2y asm ("d31") = vmov_n_s16(R2Y);
     register int16x4_t v_g2y asm ("d30") = vmov_n_s16(G2Y);
     register int16x4_t v_b2y asm ("d29") = vmov_n_s16(B2Y);
@@ -421,7 +421,7 @@ void bgrx2gray(const Size2D &size, COLOR_SPACE color_space,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
         for (; dj < roiw8; sj += 32, dj += 8)
         {
             internal::prefetch(src + sj);
@@ -512,7 +512,7 @@ void gray2rgb(const Size2D &size,
         for (; sj < roiw16; sj += 16, dj += 48)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
             "vld1.8 {d0-d1}, [%[in0]]            \n\t"
             "vmov.8 q1, q0                       \n\t"
@@ -538,7 +538,7 @@ void gray2rgb(const Size2D &size,
 
         if (sj < roiw8)
         {
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
             "vld1.8 {d0}, [%[in]]                \n\t"
             "vmov.8 d1, d0                       \n\t"
@@ -584,7 +584,7 @@ void gray2rgbx(const Size2D &size,
     size_t roiw16 = size.width >= 15 ? size.width - 15 : 0;
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register uint8x16_t vc255   asm ("q4") = vmovq_n_u8(255);
 #else
     uint8x16x4_t vRgba;
@@ -602,7 +602,7 @@ void gray2rgbx(const Size2D &size,
         for (; sj < roiw16; sj += 16, dj += 64)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
             "vld1.8 {d0-d1}, [%[in0]]            \n\t"
             "vmov.8 q1, q0                       \n\t"
@@ -628,7 +628,7 @@ void gray2rgbx(const Size2D &size,
 
         if (sj < roiw8)
         {
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
             "vld1.8 {d5}, [%[in]]                \n\t"
             "vmov.8 d6, d5                       \n\t"
@@ -672,7 +672,7 @@ void rgb2rgbx(const Size2D &size,
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
     register uint8x8_t vc255_0  asm ("d3") = vmov_n_u8(255);
 #else
     size_t roiw16 = size.width >= 15 ? size.width - 15 : 0;
@@ -688,7 +688,7 @@ void rgb2rgbx(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u, j = 0u;
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
         for (; j < roiw8; sj += 24, dj += 32, j += 8)
         {
             internal::prefetch(src + sj);
@@ -742,7 +742,7 @@ void rgbx2rgb(const Size2D &size,
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
-#if !defined(__GNUC__) || !defined(__arm__)
+#if !(!defined(__aarch64__) && defined(__GNUC__) && defined(__arm__))
     size_t roiw16 = size.width >= 15 ? size.width - 15 : 0;
     union { uint8x16x4_t v4; uint8x16x3_t v3; } v_dst0;
     union { uint8x8x4_t v4; uint8x8x3_t v3; } v_dst;
@@ -754,7 +754,7 @@ void rgbx2rgb(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u, j = 0u;
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
         for (; j < roiw8; sj += 32, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
@@ -805,7 +805,7 @@ void rgb2bgr(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#if !defined(__GNUC__) || !defined(__arm__)
+#if !(!defined(__aarch64__) && defined(__GNUC__) && defined(__arm__))
     size_t roiw16 = size.width >= 15 ? size.width - 15 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -817,7 +817,7 @@ void rgb2bgr(const Size2D &size,
         size_t sj = 0u, dj = 0u, j = 0u;
 
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
         for (; j < roiw8; sj += 24, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
@@ -874,7 +874,7 @@ void rgbx2bgrx(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#if !defined(__GNUC__) || !defined(__arm__)
+#if !(!defined(__aarch64__) && defined(__GNUC__) && defined(__arm__))
     size_t roiw16 = size.width >= 15 ? size.width - 15 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -885,7 +885,7 @@ void rgbx2bgrx(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u, j = 0u;
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
         for (; j < roiw8; sj += 32, dj += 32, j += 8)
         {
             internal::prefetch(src + sj);
@@ -943,7 +943,7 @@ void rgbx2bgr(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#if !defined(__GNUC__) || !defined(__arm__)
+#if !(!defined(__aarch64__) && defined(__GNUC__) && defined(__arm__))
     size_t roiw16 = size.width >= 15 ? size.width - 15 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -954,7 +954,7 @@ void rgbx2bgr(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u, j = 0u;
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
         for (; j < roiw8; sj += 32, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
@@ -1010,7 +1010,7 @@ void rgb2bgrx(const Size2D &size,
 {
     internal::assertSupportedConfiguration();
 #ifdef CAROTENE_NEON
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
     register uint8x8_t vc255  asm ("d3") = vmov_n_u8(255);
 #else
     union { uint8x16x4_t v4; uint8x16x3_t v3; } vals0;
@@ -1019,7 +1019,7 @@ void rgb2bgrx(const Size2D &size,
     vals8.v4.val[3] = vmov_n_u8(255);
 #endif
 
-#if !defined(__GNUC__) || !defined(__arm__)
+#if !(!defined(__aarch64__) && defined(__GNUC__) && defined(__arm__))
     size_t roiw16 = size.width >= 15 ? size.width - 15 : 0;
 #endif
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
@@ -1030,7 +1030,7 @@ void rgb2bgrx(const Size2D &size,
         u8 * dst = internal::getRowPtr(dstBase, dstStride, i);
         size_t sj = 0u, dj = 0u, j = 0u;
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
         for (; j < roiw8; sj += 24, dj += 32, j += 8)
         {
             internal::prefetch(src + sj);
@@ -1087,12 +1087,12 @@ inline uint8x8x3_t convertToHSV(const uint8x8_t vR, const uint8x8_t vG, const ui
                                 const s32 hrange )
 {
     const s32 hsv_shift = 12;
-    register const f32 vsdiv_table = f32(255 << hsv_shift);
-    register f32 vhdiv_table = f32(hrange << hsv_shift);
-    register const s32 vhrange = hrange;
-    register const s32 v0 = s32(0);
-    register const s32 vshift = s32(1 << (hsv_shift-1));
-    register const s32 v6 = s32(6);
+    const f32 vsdiv_table = f32(255 << hsv_shift);
+    f32 vhdiv_table = f32(hrange << hsv_shift);
+    const s32 vhrange = hrange;
+    const s32 v0 = s32(0);
+    const s32 vshift = s32(1 << (hsv_shift-1));
+    const s32 v6 = s32(6);
 
     uint8x8_t vMin = vmin_u8(vR, vG);
     uint8x8_t vMax = vmax_u8(vR, vG);
@@ -1409,7 +1409,7 @@ inline void convertToHSV(const s32 r, const s32 g, const s32 b,
               "d24","d25","d26","d27","d28","d29","d30","d31"         \
             );
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 
 #define YCRCB_CONSTS                                                        \
     register int16x4_t vcYR  asm ("d31") = vmov_n_s16(4899);                \
@@ -1555,7 +1555,7 @@ inline uint8x8x3_t convertToYCrCb( const int16x8_t& vR, const int16x8_t& vG, con
 #define COEFF_G   (  8663)
 #define COEFF_B   (-17705)
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 #define YUV420ALPHA3_CONST
 #define YUV420ALPHA4_CONST register uint8x16_t c255  asm ("q13") = vmovq_n_u8(255);
 #define YUV420ALPHA3_CONVERT
@@ -1852,7 +1852,7 @@ void rgb2hsv(const Size2D &size,
 #ifdef CAROTENE_NEON
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
     const s32 hsv_shift = 12;
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register const f32 vsdiv_table = f32(255 << hsv_shift);
     register f32 vhdiv_table = f32(hrange << hsv_shift);
     register const s32 vhrange = hrange;
@@ -1871,7 +1871,7 @@ void rgb2hsv(const Size2D &size,
         for (; j < roiw8; sj += 24, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERT_TO_HSV_ASM(vld3.8 {d0-d2}, d0, d2)
 #else
             uint8x8x3_t vRgb = vld3_u8(src + sj);
@@ -1904,7 +1904,7 @@ void rgbx2hsv(const Size2D &size,
 #ifdef CAROTENE_NEON
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
     const s32 hsv_shift = 12;
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register const f32 vsdiv_table = f32(255 << hsv_shift);
     register f32 vhdiv_table = f32(hrange << hsv_shift);
     register const s32 vhrange = hrange;
@@ -1923,7 +1923,7 @@ void rgbx2hsv(const Size2D &size,
         for (; j < roiw8; sj += 32, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERT_TO_HSV_ASM(vld4.8 {d0-d3}, d0, d2)
 #else
             uint8x8x4_t vRgb = vld4_u8(src + sj);
@@ -1956,7 +1956,7 @@ void bgr2hsv(const Size2D &size,
 #ifdef CAROTENE_NEON
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
     const s32 hsv_shift = 12;
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register const f32 vsdiv_table = f32(255 << hsv_shift);
     register f32 vhdiv_table = f32(hrange << hsv_shift);
     register const s32 vhrange = hrange;
@@ -1975,7 +1975,7 @@ void bgr2hsv(const Size2D &size,
         for (; j < roiw8; sj += 24, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERT_TO_HSV_ASM(vld3.8 {d0-d2}, d2, d0)
 #else
             uint8x8x3_t vRgb = vld3_u8(src + sj);
@@ -2008,7 +2008,7 @@ void bgrx2hsv(const Size2D &size,
 #ifdef CAROTENE_NEON
     size_t roiw8 = size.width >= 7 ? size.width - 7 : 0;
     const s32 hsv_shift = 12;
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register const f32 vsdiv_table = f32(255 << hsv_shift);
     register f32 vhdiv_table = f32(hrange << hsv_shift);
     register const s32 vhrange = hrange;
@@ -2027,7 +2027,7 @@ void bgrx2hsv(const Size2D &size,
         for (; j < roiw8; sj += 32, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERT_TO_HSV_ASM(vld4.8 {d0-d3}, d2, d0)
 #else
             uint8x8x4_t vRgb = vld4_u8(src + sj);
@@ -2068,7 +2068,7 @@ void rgbx2bgr565(const Size2D &size,
         for (; j < roiw16; sj += 64, dj += 32, j += 16)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
                 "vld4.8 {d2, d4, d6, d8}, [%[in0]]        @  q0       q1       q2       q3       q4       \n\t"
                 "vld4.8 {d3, d5, d7, d9}, [%[in1]]        @  xxxxxxxx rrrrRRRR ggggGGGG bbbbBBBB xxxxxxxx \n\t"
@@ -2122,7 +2122,7 @@ void rgb2bgr565(const Size2D &size,
         for (; j < roiw16; sj += 48, dj += 32, j += 16)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
                 "vld3.8 {d2, d4, d6}, [%[in0]]       @  q0       q1       q2       q3       q4       \n\t"
                 "vld3.8 {d3, d5, d7}, [%[in1]]       @  xxxxxxxx rrrrRRRR ggggGGGG bbbbBBBB xxxxxxxx \n\t"
@@ -2176,7 +2176,7 @@ void rgbx2rgb565(const Size2D &size,
         for (; j < roiw16; sj += 64, dj += 32, j += 16)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
                 "vld4.8 {d0, d2, d4, d6}, [%[in0]]    @  q0       q1       q2       q3         \n\t"
                 "vld4.8 {d1, d3, d5, d7}, [%[in1]]    @  rrrrRRRR ggggGGGG bbbbBBBB aaaaAAAA   \n\t"
@@ -2230,7 +2230,7 @@ void rgb2rgb565(const Size2D &size,
         for (; j < roiw16; sj += 48, dj += 32, j += 16)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
                 "vld3.8 {d0, d2, d4}, [%[in0]]        @  q0       q1       q2       q3         \n\t"
                 "vld3.8 {d1, d3, d5}, [%[in1]]        @  rrrrRRRR ggggGGGG bbbbBBBB xxxxxxxx   \n\t"
@@ -2285,7 +2285,7 @@ void rgb2ycrcb(const Size2D &size,
         for (; j < roiw8; sj += 24, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTTOYCRCB(vld3.8 {d0-d2}, d0, d1, d2)
 #else
             uint8x8x3_t vRgb = vld3_u8(src + sj);
@@ -2329,7 +2329,7 @@ void rgbx2ycrcb(const Size2D &size,
         for (; j < roiw8; sj += 32, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTTOYCRCB(vld4.8 {d0-d3}, d0, d1, d2)
 #else
             uint8x8x4_t vRgba = vld4_u8(src + sj);
@@ -2373,7 +2373,7 @@ void bgr2ycrcb(const Size2D &size,
         for (; j < roiw8; sj += 24, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTTOYCRCB(vld3.8 {d0-d2}, d2, d1, d0)
 #else
             uint8x8x3_t vBgr = vld3_u8(src + sj);
@@ -2417,7 +2417,7 @@ void bgrx2ycrcb(const Size2D &size,
         for (; j < roiw8; sj += 32, dj += 24, j += 8)
         {
             internal::prefetch(src + sj);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTTOYCRCB(vld4.8 {d0-d3}, d2, d1, d0)
 #else
             uint8x8x4_t vBgra = vld4_u8(src + sj);
@@ -2499,7 +2499,7 @@ void yuv420sp2rgb(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(3, d1, d0, q5, q6)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);
@@ -2545,7 +2545,7 @@ void yuv420sp2rgbx(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(4, d1, d0, q5, q6)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);
@@ -2591,7 +2591,7 @@ void yuv420i2rgb(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(3, d0, d1, q5, q6)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);
@@ -2637,7 +2637,7 @@ void yuv420i2rgbx(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(4, d0, d1, q5, q6)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);
@@ -2683,7 +2683,7 @@ void yuv420sp2bgr(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(3, d1, d0, q6, q5)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);
@@ -2729,7 +2729,7 @@ void yuv420sp2bgrx(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(4, d1, d0, q6, q5)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);
@@ -2775,7 +2775,7 @@ void yuv420i2bgr(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(3, d0, d1, q6, q5)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);
@@ -2821,7 +2821,7 @@ void yuv420i2bgrx(const Size2D &size,
             internal::prefetch(uv + j);
             internal::prefetch(y1 + j);
             internal::prefetch(y2 + j);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             CONVERTYUV420TORGB(4, d0, d1, q6, q5)
 #else
             convertYUV420.ToRGB(y1 + j, y2 + j, uv + j, dst1 + dj, dst2 + dj);

3rdparty/carotene/src/convert.cpp

@@ -101,7 +101,7 @@ CVT_FUNC(u8, s8, 16,
      }
 })
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVT_FUNC(u8, u16, 16,
      register uint8x16_t zero0 asm ("q1") = vmovq_n_u8(0);,
 {
@@ -135,7 +135,7 @@ CVT_FUNC(u8, u16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVT_FUNC(u8, s32, 16,
      register uint8x16_t zero0 asm ("q1") = vmovq_n_u8(0);
      register uint8x16_t zero1 asm ("q2") = vmovq_n_u8(0);
@@ -173,7 +173,7 @@ CVT_FUNC(u8, s32, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(u8, f32, 16,
 ,
 {
@@ -248,7 +248,7 @@ CVT_FUNC(s8, u8, 16,
      }
 })
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVT_FUNC(s8, u16, 16,
      register uint8x16_t zero0 asm ("q1") = vmovq_n_u8(0);,
 {
@@ -284,7 +284,7 @@ CVT_FUNC(s8, u16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s8, s16, 16,
 ,
 {
@@ -323,7 +323,7 @@ CVT_FUNC(s8, s16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVT_FUNC(s8, s32, 16,
 ,
 {
@@ -377,7 +377,7 @@ CVT_FUNC(s8, s32, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s8, f32, 16,
 ,
 {
@@ -440,7 +440,7 @@ CVT_FUNC(s8, f32, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(u16, u8, 16,
 ,
 {
@@ -479,7 +479,7 @@ CVT_FUNC(u16, u8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(u16, s8, 16,
     register uint8x16_t v127 asm ("q4") = vmovq_n_u8(127);,
 {
@@ -522,7 +522,7 @@ CVT_FUNC(u16, s8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVT_FUNC(u16, s16, 8,
      register uint16x8_t v32767 asm ("q4") = vmovq_n_u16(0x7FFF);,
 {
@@ -555,7 +555,7 @@ CVT_FUNC(u16, s16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVT_FUNC(u16, s32, 8,
      register uint16x8_t zero0 asm ("q1") = vmovq_n_u16(0);,
 {
@@ -589,7 +589,7 @@ CVT_FUNC(u16, s32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(u16, f32, 8,
 ,
 {
@@ -633,7 +633,7 @@ CVT_FUNC(u16, f32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s16, u8, 16,
 ,
 {
@@ -672,7 +672,7 @@ CVT_FUNC(s16, u8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s16, s8, 16,
 ,
 {
@@ -711,7 +711,7 @@ CVT_FUNC(s16, s8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVT_FUNC(s16, u16, 8,
      register int16x8_t vZero asm ("q4") = vmovq_n_s16(0);,
 {
@@ -747,7 +747,7 @@ CVT_FUNC(s16, u16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s16, s32, 8,
 ,
 {
@@ -786,7 +786,7 @@ CVT_FUNC(s16, s32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s16, f32, 8,
 ,
 {
@@ -829,7 +829,7 @@ CVT_FUNC(s16, f32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s32, u8, 8,
 ,
 {
@@ -870,7 +870,7 @@ CVT_FUNC(s32, u8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s32, s8, 8,
 ,
 {
@@ -911,7 +911,7 @@ CVT_FUNC(s32, s8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s32, u16, 8,
 ,
 {
@@ -950,7 +950,7 @@ CVT_FUNC(s32, u16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s32, s16, 8,
 ,
 {
@@ -989,7 +989,7 @@ CVT_FUNC(s32, s16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(s32, f32, 8,
 ,
 {
@@ -1034,7 +1034,7 @@ CVT_FUNC(s32, f32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(f32, u8, 8,
     register float32x4_t vmult asm ("q0") = vdupq_n_f32((float)(1 << 16));
     register uint32x4_t  vmask asm ("q1") = vdupq_n_u32(1<<16);,
@@ -1101,7 +1101,7 @@ CVT_FUNC(f32, u8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(f32, s8, 8,
      register float32x4_t vhalf asm ("q0") = vdupq_n_f32(0.5f);,
 {
@@ -1153,7 +1153,7 @@ CVT_FUNC(f32, s8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(f32, u16, 8,
      register float32x4_t vhalf asm ("q0") = vdupq_n_f32(0.5f);,
 {
@@ -1212,7 +1212,7 @@ CVT_FUNC(f32, u16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(f32, s16, 8,
      register float32x4_t vhalf asm ("q0") = vdupq_n_f32(0.5f);,
 {
@@ -1271,7 +1271,7 @@ CVT_FUNC(f32, s16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
 CVT_FUNC(f32, s32, 8,
      register float32x4_t vhalf asm ("q0") = vdupq_n_f32(0.5f);,
 {

3rdparty/carotene/src/convert_scale.cpp

@@ -135,7 +135,7 @@ namespace CAROTENE_NS {
 
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC1(u8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -220,7 +220,7 @@ CVTS_FUNC1(u8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC(u8, s8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -305,7 +305,7 @@ CVTS_FUNC(u8, s8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC(u8, u16, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -389,7 +389,7 @@ CVTS_FUNC(u8, u16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC(u8, s16, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -473,7 +473,7 @@ CVTS_FUNC(u8, s16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(u8, s32, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -562,7 +562,7 @@ CVTS_FUNC(u8, s32, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(u8, f32, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta);,
@@ -643,7 +643,7 @@ CVTS_FUNC(u8, f32, 16,
 })
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC(s8, u8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -728,7 +728,7 @@ CVTS_FUNC(s8, u8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC1(s8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -813,7 +813,7 @@ CVTS_FUNC1(s8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC(s8, u16, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -899,7 +899,7 @@ CVTS_FUNC(s8, u16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
 CVTS_FUNC(s8, s16, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -985,7 +985,7 @@ CVTS_FUNC(s8, s16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s8, s32, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1074,7 +1074,7 @@ CVTS_FUNC(s8, s32, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s8, f32, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta);,
@@ -1155,7 +1155,7 @@ CVTS_FUNC(s8, f32, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(u16, u8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1214,7 +1214,7 @@ CVTS_FUNC(u16, u8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(u16, s8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1273,7 +1273,7 @@ CVTS_FUNC(u16, s8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC1(u16, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1330,7 +1330,7 @@ CVTS_FUNC1(u16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(u16, s16, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1387,7 +1387,7 @@ CVTS_FUNC(u16, s16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(u16, s32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1443,7 +1443,7 @@ CVTS_FUNC(u16, s32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(u16, f32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta);,
@@ -1495,7 +1495,7 @@ CVTS_FUNC(u16, f32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s16, u8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1554,7 +1554,7 @@ CVTS_FUNC(s16, u8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s16, s8, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1613,7 +1613,7 @@ CVTS_FUNC(s16, s8, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s16, u16, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1670,7 +1670,7 @@ CVTS_FUNC(s16, u16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC1(s16, 16,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1727,7 +1727,7 @@ CVTS_FUNC1(s16, 16,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s16, s32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1783,7 +1783,7 @@ CVTS_FUNC(s16, s32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s16, f32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta);,
@@ -1835,7 +1835,7 @@ CVTS_FUNC(s16, f32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s32, u8, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1893,7 +1893,7 @@ CVTS_FUNC(s32, u8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s32, s8, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -1951,7 +1951,7 @@ CVTS_FUNC(s32, s8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s32, u16, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -2007,7 +2007,7 @@ CVTS_FUNC(s32, u16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s32, s16, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -2063,7 +2063,7 @@ CVTS_FUNC(s32, s16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC1(s32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -2118,7 +2118,7 @@ CVTS_FUNC1(s32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(s32, f32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta);,
@@ -2169,7 +2169,7 @@ CVTS_FUNC(s32, f32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(f32, u8, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)((1 << 16)*alpha));
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)((1 << 16)*beta));
@@ -2239,7 +2239,7 @@ CVTS_FUNC(f32, u8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(f32, s8, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -2293,7 +2293,7 @@ CVTS_FUNC(f32, s8, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(f32, u16, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -2345,7 +2345,7 @@ CVTS_FUNC(f32, u16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(f32, s16, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -2397,7 +2397,7 @@ CVTS_FUNC(f32, s16, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC(f32, s32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta + 0.5f);,
@@ -2448,7 +2448,7 @@ CVTS_FUNC(f32, s32, 8,
 })
 #endif
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
 CVTS_FUNC1(f32, 8,
     register float32x4_t vscale asm ("q0") = vdupq_n_f32((f32)alpha);
     register float32x4_t vshift asm ("q1") = vdupq_n_f32((f32)beta);,

3rdparty/carotene/src/div.cpp

@@ -151,6 +151,10 @@ void div(const Size2D &size,
     typedef typename internal::VecTraits<T>::vec128 vec128;
     typedef typename internal::VecTraits<T>::vec64 vec64;
 
+#if defined(__GNUC__) && (defined(__GXX_EXPERIMENTAL_CXX0X__) || __cplusplus >= 201103L)
+    static_assert(std::numeric_limits<T>::is_integer, "template implementation is for integer types only");
+#endif
+
     if (scale == 0.0f ||
         (std::numeric_limits<T>::is_integer &&
          (scale * std::numeric_limits<T>::max()) <  1.0f &&
@@ -311,6 +315,10 @@ void recip(const Size2D &size,
     typedef typename internal::VecTraits<T>::vec128 vec128;
     typedef typename internal::VecTraits<T>::vec64 vec64;
 
+#if defined(__GNUC__) && (defined(__GXX_EXPERIMENTAL_CXX0X__) || __cplusplus >= 201103L)
+    static_assert(std::numeric_limits<T>::is_integer, "template implementation is for integer types only");
+#endif
+
     if (scale == 0.0f ||
         (std::numeric_limits<T>::is_integer &&
          scale <  1.0f &&
@@ -463,8 +471,6 @@ void div(const Size2D &size,
         return;
     }
 
-    float32x4_t v_zero = vdupq_n_f32(0.0f);
-
     size_t roiw128 = size.width >= 3 ? size.width - 3 : 0;
     size_t roiw64 = size.width >= 1 ? size.width - 1 : 0;
 
@@ -485,9 +491,7 @@ void div(const Size2D &size,
                 float32x4_t v_src0 = vld1q_f32(src0 + j);
                 float32x4_t v_src1 = vld1q_f32(src1 + j);
 
-                uint32x4_t v_mask = vceqq_f32(v_src1,v_zero);
-                vst1q_f32(dst + j, vreinterpretq_f32_u32(vbicq_u32(
-                                   vreinterpretq_u32_f32(vmulq_f32(v_src0, internal::vrecpq_f32(v_src1))), v_mask)));
+                vst1q_f32(dst + j, vmulq_f32(v_src0, internal::vrecpq_f32(v_src1)));
             }
 
             for (; j < roiw64; j += 2)
@@ -495,14 +499,12 @@ void div(const Size2D &size,
                 float32x2_t v_src0 = vld1_f32(src0 + j);
                 float32x2_t v_src1 = vld1_f32(src1 + j);
 
-                uint32x2_t v_mask = vceq_f32(v_src1,vget_low_f32(v_zero));
-                vst1_f32(dst + j, vreinterpret_f32_u32(vbic_u32(
-                                  vreinterpret_u32_f32(vmul_f32(v_src0, internal::vrecp_f32(v_src1))), v_mask)));
+                vst1_f32(dst + j, vmul_f32(v_src0, internal::vrecp_f32(v_src1)));
             }
 
             for (; j < size.width; j++)
             {
-                dst[j] = src1[j] ? src0[j] / src1[j] : 0.0f;
+                dst[j] = src0[j] / src1[j];
             }
         }
     }
@@ -523,10 +525,8 @@ void div(const Size2D &size,
                 float32x4_t v_src0 = vld1q_f32(src0 + j);
                 float32x4_t v_src1 = vld1q_f32(src1 + j);
 
-                uint32x4_t v_mask = vceqq_f32(v_src1,v_zero);
-                vst1q_f32(dst + j, vreinterpretq_f32_u32(vbicq_u32(
-                                   vreinterpretq_u32_f32(vmulq_f32(vmulq_n_f32(v_src0, scale),
-                                                         internal::vrecpq_f32(v_src1))), v_mask)));
+                vst1q_f32(dst + j, vmulq_f32(vmulq_n_f32(v_src0, scale),
+                                             internal::vrecpq_f32(v_src1)));
             }
 
             for (; j < roiw64; j += 2)
@@ -534,15 +534,13 @@ void div(const Size2D &size,
                 float32x2_t v_src0 = vld1_f32(src0 + j);
                 float32x2_t v_src1 = vld1_f32(src1 + j);
 
-                uint32x2_t v_mask = vceq_f32(v_src1,vget_low_f32(v_zero));
-                vst1_f32(dst + j, vreinterpret_f32_u32(vbic_u32(
-                                  vreinterpret_u32_f32(vmul_f32(vmul_n_f32(v_src0, scale),
-                                                                internal::vrecp_f32(v_src1))), v_mask)));
+                vst1_f32(dst + j, vmul_f32(vmul_n_f32(v_src0, scale),
+                                           internal::vrecp_f32(v_src1)));
             }
 
             for (; j < size.width; j++)
             {
-                dst[j] = src1[j] ? src0[j] * scale / src1[j] : 0.0f;
+                dst[j] = src0[j] * scale / src1[j];
             }
         }
     }
@@ -620,8 +618,6 @@ void reciprocal(const Size2D &size,
         return;
     }
 
-    float32x4_t v_zero = vdupq_n_f32(0.0f);
-
     size_t roiw128 = size.width >= 3 ? size.width - 3 : 0;
     size_t roiw64 = size.width >= 1 ? size.width - 1 : 0;
 
@@ -639,23 +635,19 @@ void reciprocal(const Size2D &size,
 
                 float32x4_t v_src1 = vld1q_f32(src1 + j);
 
-                uint32x4_t v_mask = vceqq_f32(v_src1,v_zero);
-                vst1q_f32(dst + j, vreinterpretq_f32_u32(vbicq_u32(
-                                   vreinterpretq_u32_f32(internal::vrecpq_f32(v_src1)), v_mask)));
+                vst1q_f32(dst + j, internal::vrecpq_f32(v_src1));
             }
 
             for (; j < roiw64; j += 2)
             {
                 float32x2_t v_src1 = vld1_f32(src1 + j);
 
-                uint32x2_t v_mask = vceq_f32(v_src1,vget_low_f32(v_zero));
-                vst1_f32(dst + j, vreinterpret_f32_u32(vbic_u32(
-                                  vreinterpret_u32_f32(internal::vrecp_f32(v_src1)), v_mask)));
+                vst1_f32(dst + j, internal::vrecp_f32(v_src1));
             }
 
             for (; j < size.width; j++)
             {
-                dst[j] = src1[j] ? 1.0f / src1[j] : 0;
+                dst[j] = 1.0f / src1[j];
             }
         }
     }
@@ -673,25 +665,19 @@ void reciprocal(const Size2D &size,
 
                 float32x4_t v_src1 = vld1q_f32(src1 + j);
 
-                uint32x4_t v_mask = vceqq_f32(v_src1,v_zero);
-                vst1q_f32(dst + j, vreinterpretq_f32_u32(vbicq_u32(
-                                   vreinterpretq_u32_f32(vmulq_n_f32(internal::vrecpq_f32(v_src1),
-                                                                     scale)),v_mask)));
+                vst1q_f32(dst + j, vmulq_n_f32(internal::vrecpq_f32(v_src1), scale));
             }
 
             for (; j < roiw64; j += 2)
             {
                 float32x2_t v_src1 = vld1_f32(src1 + j);
 
-                uint32x2_t v_mask = vceq_f32(v_src1,vget_low_f32(v_zero));
-                vst1_f32(dst + j, vreinterpret_f32_u32(vbic_u32(
-                                  vreinterpret_u32_f32(vmul_n_f32(internal::vrecp_f32(v_src1),
-                                                                  scale)), v_mask)));
+                vst1_f32(dst + j, vmul_n_f32(internal::vrecp_f32(v_src1), scale));
             }
 
             for (; j < size.width; j++)
             {
-                dst[j] = src1[j] ? scale / src1[j] : 0;
+                dst[j] = scale / src1[j];
             }
         }
     }

3rdparty/carotene/src/flip.cpp

@@ -105,12 +105,12 @@ void flip3(const Size2D & size,
 {
     using namespace internal;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
     typedef typename VecTraits<T, 3>::vec128 vec128;
 #endif
     typedef typename VecTraits<T, 3>::vec64 vec64;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
     u32 step_base = 16 / sizeof(T), step_base3 = step_base * 3;
     size_t roiw_base = size.width >= (step_base - 1) ? size.width - step_base + 1 : 0;
 #endif
@@ -123,7 +123,7 @@ void flip3(const Size2D & size,
         T * dst = getRowPtr((T *)dstBase, dstStride, (flipMode & FLIP_VERTICAL_MODE) != 0 ? size.height - i - 1 : i);
         size_t j = 0, js = 0, jd = size.width * 3;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
         for (; j < roiw_base; j += step_base, js += step_base3, jd -= step_base3)
         {
             prefetch(src + js);
@@ -139,7 +139,7 @@ void flip3(const Size2D & size,
 
             vst3q(dst + jd - step_base3, v_dst);
         }
-#endif // ANDROID
+#endif // __ANDROID__
 
         for (; j < roiw_tail; j += step_tail, js += step_tail3, jd -= step_tail3)
         {

3rdparty/carotene/src/gaussian_blur.cpp

@@ -327,7 +327,7 @@ void gaussianBlur5x5(const Size2D &size, s32 cn,
                 u16* lidx1 = lane + x - 1*2;
                 u16* lidx3 = lane + x + 1*2;
                 u16* lidx4 = lane + x + 2*2;
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
                 __asm__ __volatile__ (
                     "vld2.16 {d0, d2}, [%[in0]]!                              \n\t"
                     "vld2.16 {d1, d3}, [%[in0]]                               \n\t"
@@ -398,7 +398,7 @@ void gaussianBlur5x5(const Size2D &size, s32 cn,
                 u16* lidx1 = lane + x - 1*3;
                 u16* lidx3 = lane + x + 1*3;
                 u16* lidx4 = lane + x + 2*3;
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ __volatile__ (
                     "vld3.16 {d0, d2, d4}, [%[in0]]!                          \n\t"
                     "vld3.16 {d1, d3, d5}, [%[in0]]                           \n\t"
@@ -482,7 +482,7 @@ void gaussianBlur5x5(const Size2D &size, s32 cn,
                 u16* lidx1 = lane + x - 1*4;
                 u16* lidx3 = lane + x + 1*4;
                 u16* lidx4 = lane + x + 2*4;
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ __volatile__ (
                     "vld4.16 {d0, d2, d4, d6}, [%[in0]]!                      \n\t"
                     "vld4.16 {d1, d3, d5, d7}, [%[in0]]                       \n\t"

3rdparty/carotene/src/pyramid.cpp

@@ -331,7 +331,7 @@ void gaussianPyramidDown(const Size2D &srcSize,
             for (; x < roiw8; x += 8)
             {
                 internal::prefetch(lane + 2 * x);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
                 __asm__ (
                     "vld2.16 {d0-d3}, [%[in0]]                               \n\t"
                     "vld2.16 {d4-d7}, [%[in4]]                               \n\t"
@@ -538,7 +538,7 @@ void gaussianPyramidDown(const Size2D &srcSize,
             for (; x < roiw4; x += 4)
             {
                 internal::prefetch(lane + 2 * x);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
                 __asm__ (
                     "vld2.32 {d0-d3}, [%[in0]]                              \n\t"
                     "vld2.32 {d4-d7}, [%[in4]]                              \n\t"
@@ -672,7 +672,7 @@ void gaussianPyramidDown(const Size2D &srcSize,
     std::vector<f32> _buf(cn*(srcSize.width + 4) + 32/sizeof(f32));
     f32* lane = internal::alignPtr(&_buf[2*cn], 32);
 
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
     register float32x4_t vc6d4f32  asm ("q11") = vmovq_n_f32(1.5f);  // 6/4
     register float32x4_t vc1d4f32  asm ("q12") = vmovq_n_f32(0.25f); // 1/4
 
@@ -680,12 +680,12 @@ void gaussianPyramidDown(const Size2D &srcSize,
     register float32x4_t vc4d64f32 asm ("q14") = vmovq_n_f32(0.0625f);   //4/4/16
     register float32x4_t vc6d64f32 asm ("q15") = vmovq_n_f32(0.09375f);  //6/4/16
 #else
-    register float32x4_t vc6d4f32  = vmovq_n_f32(1.5f);  // 6/4
-    register float32x4_t vc1d4f32  = vmovq_n_f32(0.25f); // 1/4
+    float32x4_t vc6d4f32  = vmovq_n_f32(1.5f);  // 6/4
+    float32x4_t vc1d4f32  = vmovq_n_f32(0.25f); // 1/4
 
-    register float32x4_t vc1d64f32 = vmovq_n_f32(0.015625f); //1/4/16
-    register float32x4_t vc4d64f32 = vmovq_n_f32(0.0625f);   //4/4/16
-    register float32x4_t vc6d64f32 = vmovq_n_f32(0.09375f);  //6/4/16
+    float32x4_t vc1d64f32 = vmovq_n_f32(0.015625f); //1/4/16
+    float32x4_t vc4d64f32 = vmovq_n_f32(0.0625f);   //4/4/16
+    float32x4_t vc6d64f32 = vmovq_n_f32(0.09375f);  //6/4/16
 #endif
 
     for (size_t i = 0; i < dstSize.height; ++i)
@@ -739,7 +739,7 @@ void gaussianPyramidDown(const Size2D &srcSize,
             for (; x < roiw4; x += 4)
             {
                 internal::prefetch(lane + 2 * x);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
                 __asm__ __volatile__ (
                     "vld2.32 {d0-d3}, [%[in0]]                              \n\t"
                     "vld2.32 {d8-d11}, [%[in4]]                             \n\t"
@@ -932,7 +932,7 @@ pyrUp8uHorizontalConvolution:
             for (; x < lim; x += 8)
             {
                 internal::prefetch(lane + x);
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ (
                     "vld1.16 {d0-d1}, [%[in0]]       /*q0 = v0*/            \n\t"
                     "vld1.16 {d2-d3}, [%[in2]]       /*q1 = v2*/            \n\t"
@@ -973,7 +973,7 @@ pyrUp8uHorizontalConvolution:
             for (; x < lim; x += 24)
             {
                 internal::prefetch(lane + x);
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ (
                     "vmov.u16 q9, #6                                           \n\t"
                     "vld3.16 {d0, d2, d4}, [%[in0]]        /*v0*/              \n\t"
@@ -1064,7 +1064,7 @@ pyrUp8uHorizontalConvolution:
             for (; x < lim; x += 8)
             {
                 internal::prefetch(lane + x);
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ (
                     "vld1.16 {d0-d1}, [%[in0]]       /*q0 = v0*/            \n\t"
                     "vld1.16 {d2-d3}, [%[in2]]       /*q1 = v2*/            \n\t"
@@ -1210,7 +1210,7 @@ pyrUp16sHorizontalConvolution:
             for (; x < lim; x += 4)
             {
                 internal::prefetch(lane + x);
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ (
                     "vld1.32 {d0-d1}, [%[in0]]       /*q0 = v0*/            \n\t"
                     "vld1.32 {d2-d3}, [%[in2]]       /*q1 = v2*/            \n\t"
@@ -1251,7 +1251,7 @@ pyrUp16sHorizontalConvolution:
             for (; x < lim; x += 12)
             {
                 internal::prefetch(lane + x + 3);
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ (
                     "vmov.s32 q9, #6                                           \n\t"
                     "vld3.32 {d0, d2, d4}, [%[in0]]        /*v0*/              \n\t"
@@ -1343,7 +1343,7 @@ pyrUp16sHorizontalConvolution:
             for (; x < lim; x += 4)
             {
                 internal::prefetch(lane + x);
-#if defined(__GNUC__) && defined(__arm__)
+#if !defined(__aarch64__) && defined(__GNUC__) && defined(__arm__)
                 __asm__ (
                     "vld1.32 {d0-d1}, [%[in0]]       /*q0 = v0*/            \n\t"
                     "vld1.32 {d2-d3}, [%[in2]]       /*q1 = v2*/            \n\t"

3rdparty/carotene/src/resize.cpp

@@ -106,8 +106,10 @@ bool isResizeLinearOpenCVSupported(const Size2D &ssize, const Size2D &dsize, u32
             && !(ssize.width > 0xffffFFFF || ssize.height > 0xffffFFFF)// Restrict image size since internal index evaluation
                                                                        // is performed with u32
 #endif
-            && dsize.width >= 2 && dsize.height >= 8)
+            && dsize.width >= 2 && dsize.height >= 8
+            && (2*dsize.width != ssize.width || 2*dsize.height != ssize.height)) // 2x downscaling is performed as area in OpenCV which differs from this implementation
             return isSupportedConfiguration();
+        return false;
     default:
         return false;
     };
@@ -359,7 +361,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
         }
         else //if ((wr == 4.0f) && (hr == 4.0f)) //the only scale that lasts after isSupported check
         {
-#ifndef ANDROID
+#ifndef __ANDROID__
             size_t roiw16 = dsize.width >= 15 ? dsize.width - 15 : 0;
 #endif
             size_t roiw8 = dsize.width >= 7 ? dsize.width - 7 : 0;
@@ -373,7 +375,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
                 u8 * dst_row = internal::getRowPtr(dstBase, dstStride, i);
                 size_t sj = 0, dj = 0;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
                 for ( ; dj < roiw16; dj += 16, sj += 64)
                 {
                     internal::prefetch(src0_row + sj);
@@ -450,7 +452,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
     {
         if ((wr == 2.0f) && (hr == 2.0f))
         {
-#ifndef ANDROID
+#ifndef __ANDROID__
             size_t roiw4 = dsize.width >= 3 ? (dsize.width - 3) << 2 : 0;
 #endif
             size_t roiw2 = dsize.width >= 1 ? (dsize.width - 1) << 2 : 0;
@@ -462,7 +464,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
                 u8 * dst_row = internal::getRowPtr(dstBase, dstStride, i);
                 size_t sj = 0, dj = 0;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
                 for ( ; dj < roiw4; dj += 16, sj += 32)
                 {
                     internal::prefetch(src0_row + sj);
@@ -537,7 +539,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
         }
         else if ((wr == 0.5f) && (hr == 0.5f))
         {
-#ifndef ANDROID
+#ifndef __ANDROID__
             size_t roiw32 = dsize.width >= 31 ? (dsize.width - 31) << 2 : 0;
 #endif
             size_t roiw16 = dsize.width >= 15 ? (dsize.width - 15) << 2 : 0;
@@ -549,7 +551,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
                 u8 * dst1_row = internal::getRowPtr(dstBase, dstStride, std::min(i + 1, dsize.height - 1));
                 size_t sj = 0, dj = 0;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
                 for ( ; dj < roiw32; dj += 128, sj += 64)
                 {
                     internal::prefetch(src_row + sj);
@@ -758,7 +760,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
     {
         if ((wr == 2.0f) && (wr == 2.0f))
         {
-#ifndef ANDROID
+#ifndef __ANDROID__
             size_t roiw16 = dsize.width >= 15 ? (dsize.width - 15) * 3 : 0;
 #endif
             size_t roiw8 = dsize.width >= 7 ? (dsize.width - 7) * 3 : 0;
@@ -770,7 +772,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
                 u8 * dst_row = internal::getRowPtr(dstBase, dstStride, i);
                 size_t sj = 0, dj = 0;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
                 for ( ; dj < roiw16; dj += 48, sj += 96)
                 {
                     internal::prefetch(src0_row + sj);
@@ -856,7 +858,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
         }
         else if ((wr == 0.5f) && (hr == 0.5f))
         {
-#ifndef ANDROID
+#ifndef __ANDROID__
             size_t roiw32 = dsize.width >= 31 ? (dsize.width - 31) * 3 : 0;
 #endif
             size_t roiw16 = dsize.width >= 15 ? (dsize.width - 15) * 3 : 0;
@@ -868,7 +870,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
                 u8 * dst1_row = internal::getRowPtr(dstBase, dstStride, std::min(i + 1, dsize.height - 1));
                 size_t sj = 0, dj = 0;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
                 for ( ; dj < roiw32; dj += 96, sj += 48)
                 {
                     internal::prefetch(src_row + sj);
@@ -928,7 +930,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
         }
         else //if ((hr == 4.0f) && (wr == 4.0f)) //the only scale that lasts after isSupported check
         {
-#ifndef ANDROID
+#ifndef __ANDROID__
             size_t roiw8 = dsize.width >= 7 ? (dsize.width - 7) * 3 : 0;
 #endif
 
@@ -941,7 +943,7 @@ inline void resizeAreaRounding(const Size2D &ssize, const Size2D &dsize,
                 u8 * dst_row = internal::getRowPtr(dstBase, dstStride, i);
                 size_t sj = 0, dj = 0;
 
-#ifndef ANDROID
+#ifndef __ANDROID__
                 for ( ; dj < roiw8; dj += 24, sj += 96)
                 {
                     internal::prefetch(src0_row + sj);

3rdparty/carotene/src/saturate_cast.hpp

@@ -72,11 +72,11 @@ __declspec(naked) static void vcvtr_s32_f64_imp(f64 d)
 
 # if defined(__VFP_FP__) && !defined(__SOFTFP__) && !(defined _DEBUG || defined DEBUG) && !defined(__CUDACC__)
 #  define CAROTENE_ROUND_FLT(value) {                              \
-    register union { f32 f; s32 i; } result;                    \
+    union { f32 f; s32 i; } result; \
     asm ("ftosis  %0, %1 \n" : "=w" (result.f) : "w" (value) ); \
     return result.i; }
 #  define CAROTENE_ROUND_DBL(value) {                      \
-    register union {f32 f; s32 i;} __tegra_result;      \
+    union {f32 f; s32 i;} __tegra_result; \
     asm (                                               \
         "ftosid  %0, %P1\n"                             \
         : "=w" (__tegra_result.f)                       \

3rdparty/carotene/src/scharr.cpp

@@ -109,7 +109,7 @@ void ScharrDeriv(const Size2D &size, s32 cn,
             internal::prefetch(srow0 + x);
             internal::prefetch(srow1 + x);
             internal::prefetch(srow2 + x);
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 7
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 7 && !defined(__clang__)
             __asm__ (
                 "vld1.8 {d0}, [%[src0]]                                \n\t"
                 "vld1.8 {d2}, [%[src2]]                                \n\t"
@@ -161,7 +161,7 @@ void ScharrDeriv(const Size2D &size, s32 cn,
         x = 0;
         for( ; x < roiw8; x += 8 )
         {
-#if defined(__GNUC__) && __GNUC__ == 4 && __GNUC_MINOR__ < 6
+#if !defined(__aarch64__) && defined(__GNUC__) && __GNUC__ == 4 &&  __GNUC_MINOR__ < 6 && !defined(__clang__)
             __asm__ (
                 "vld1.16 {d4-d5}, [%[s2ptr]]                           \n\t"
                 "vld1.16 {d8-d9}, [%[s4ptr]]                           \n\t"

3rdparty/cpufeatures/CMakeLists.txt

@@ -2,19 +2,21 @@ if(NOT ANDROID)
   message("cpufeatures is ANDROID project")
 endif()
 
+ocv_update(OPENCV_CPUFEATURES_TARGET_NAME libcpufeatures)
+
 set(CPUFEATURES_ROOT "${CMAKE_CURRENT_SOURCE_DIR}" CACHE PATH "Android cpufeatures project sources (for example, <android-ndk>/sources/android/cpufeatures)")
 
 set(CPUFEATURES_INCLUDE_DIRS ${CPUFEATURES_ROOT} CACHE INTERNAL "")
-set(CPUFEATURES_LIBRARIES cpufeatures CACHE INTERNAL "")
+set(CPUFEATURES_LIBRARIES "${OPENCV_CPUFEATURES_TARGET_NAME}" CACHE INTERNAL "")
 
 if(NOT DEFINED CPUFEATURES_SOURCES)
   set(CPUFEATURES_SOURCES ${CPUFEATURES_ROOT}/cpu-features.c ${CPUFEATURES_ROOT}/cpu-features.h)
 endif()
 
 include_directories(${CPUFEATURES_INCLUDE_DIRS})
-add_library(cpufeatures STATIC ${CPUFEATURES_SOURCES})
+add_library(${OPENCV_CPUFEATURES_TARGET_NAME} STATIC ${CPUFEATURES_SOURCES})
 
-set_target_properties(cpufeatures
+set_target_properties(${OPENCV_CPUFEATURES_TARGET_NAME}
   PROPERTIES OUTPUT_NAME cpufeatures
   DEBUG_POSTFIX "${OPENCV_DEBUG_POSTFIX}"
   COMPILE_PDB_NAME cpufeatures
@@ -23,9 +25,11 @@ set_target_properties(cpufeatures
 )
 
 if(ENABLE_SOLUTION_FOLDERS)
-  set_target_properties(cpufeatures PROPERTIES FOLDER "3rdparty")
+  set_target_properties(${OPENCV_CPUFEATURES_TARGET_NAME} PROPERTIES FOLDER "3rdparty")
 endif()
 
 if(NOT BUILD_SHARED_LIBS)
-  ocv_install_target(cpufeatures EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev)
+  ocv_install_target(${OPENCV_CPUFEATURES_TARGET_NAME} EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev)
 endif()
+
+ocv_install_3rdparty_licenses(cpufeatures LICENSE README.md)

3rdparty/cpufeatures/cpu-features.h

@@ -30,6 +30,7 @@
 
 #include <sys/cdefs.h>
 #include <stdint.h>
+#include <string.h>
 
 __BEGIN_DECLS
 

3rdparty/ffmpeg/ffmpeg-download.ps1.in

@@ -0,0 +1,63 @@
+$url = "https://raw.githubusercontent.com/opencv/opencv_3rdparty/@FFMPEG_BINARIES_COMMIT@/ffmpeg/opencv_ffmpeg_64.dll"
+$expected_md5 = "@FFMPEG_FILE_HASH_BIN64@"
+$output = "$PSScriptRoot\@OPENCV_BIN_INSTALL_PATH@\opencv_ffmpeg@OPENCV_DLLVERSION@_64.dll"
+
+Write-Output ("=" * 120)
+try {
+    Get-content -Path "$PSScriptRoot\@OPENCV_LICENSES_INSTALL_PATH@\ffmpeg-readme.txt" -ErrorAction 'Stop'
+} catch {
+    Write-Output "Refer to OpenCV FFmpeg wrapper readme notes about library usage / licensing details."
+}
+Write-Output ("=" * 120)
+Write-Output ""
+
+if(![System.IO.File]::Exists($output)) {
+    try {
+        Write-Output ("Downloading: " + $output)
+        Import-Module BitsTransfer
+        $start_time = Get-Date
+        Start-BitsTransfer -Source $url -Destination $output -ErrorAction 'Stop'
+        Write-Output "Downloaded in $((Get-Date).Subtract($start_time).Seconds) seconds"
+    } catch {
+        $_ # Dump error
+        try {
+            Write-Output ("Downloading (second attempt): " + $output)
+            $start_time = Get-Date
+            Invoke-WebRequest -Uri $url -OutFile $output
+            Write-Output "Downloaded in $((Get-Date).Subtract($start_time).Seconds) seconds"
+        } catch {
+            Write-Output ("Can't download file: " + $output)
+            Write-Output ("URL: " + $url)
+            Write-Output "You need to download this file manually. Stop"
+            Pause
+            Exit
+        }
+    }
+} else {
+    Write-Output ("File exists: " + $output)
+    Write-Output ("Downloading is skipped. Remove this file and re-run this script to force downloading.")
+}
+
+if(![System.IO.File]::Exists($output)) {
+    Write-Output ("Destination file not found: " + $output)
+    Write-Output "Stop"
+    Pause
+    Exit
+}
+
+try {
+    $hash = Get-FileHash $output -Algorithm MD5 -ErrorAction 'Stop'
+
+    if($hash.Hash -eq $expected_md5) {
+        Write-Output "MD5 check passed"
+    } else {
+        Write-Output ("MD5     : " + $hash.Hash.toLower())
+        Write-Output ("Expected: " + $expected_md5)
+        Write-Output "MD5 hash mismatch"
+    }
+} catch {
+    $_ # Dump error
+    Write-Output "Can't check MD5 hash (requires PowerShell 4+)"
+}
+Pause
+Write-Output "Exit"

3rdparty/ffmpeg/ffmpeg.cmake

@@ -1,9 +1,9 @@
-# Binary branch name: ffmpeg/master_20170418
-# Binaries were created for OpenCV: b993b9b7c7f6f5f37d10acacb2962812228410ba
-set(FFMPEG_BINARIES_COMMIT "86c4a841055f2612774e85b4292bb20e5fe8a783")
-set(FFMPEG_FILE_HASH_BIN32 "3dea5f7f009b44601fe95728328e0f9e")
-set(FFMPEG_FILE_HASH_BIN64 "9debe701975ef074bd6661981f3f0716")
-set(FFMPEG_FILE_HASH_CMAKE "208c00f03d2f6f39fa6262649e0bfc8d")
+# Binaries branch name: ffmpeg/master_20181106
+# Binaries were created for OpenCV: 2c6f1ab57d4250ee46e32d1b51c056431965b470
+ocv_update(FFMPEG_BINARIES_COMMIT "759a23e24ab787a0979f8a93103dcc3105ec10c1")
+ocv_update(FFMPEG_FILE_HASH_BIN32 "849286ccc527c99e5a218b67f13c6e8c")
+ocv_update(FFMPEG_FILE_HASH_BIN64 "96444a4645753aaafa296479665c9185")
+ocv_update(FFMPEG_FILE_HASH_CMAKE "f710891525a04586d565d0e700e62a9c")
 
 function(download_win_ffmpeg script_var)
   set(${script_var} "" PARENT_SCOPE)
@@ -35,3 +35,10 @@ function(download_win_ffmpeg script_var)
     set(${script_var} "${FFMPEG_DOWNLOAD_DIR}/ffmpeg_version.cmake" PARENT_SCOPE)
   endif()
 endfunction()
+
+if(OPENCV_INSTALL_FFMPEG_DOWNLOAD_SCRIPT)
+  configure_file("${CMAKE_CURRENT_LIST_DIR}/ffmpeg-download.ps1.in" "${CMAKE_BINARY_DIR}/win-install/ffmpeg-download.ps1" @ONLY)
+  install(FILES "${CMAKE_BINARY_DIR}/win-install/ffmpeg-download.ps1" DESTINATION "." COMPONENT libs)
+endif()
+
+ocv_install_3rdparty_licenses(ffmpeg license.txt readme.txt)

3rdparty/ffmpeg/readme.txt

@@ -16,19 +16,18 @@
   The binaries are opencv_ffmpeg.dll (version for 32-bit Windows) and
   opencv_ffmpeg_64.dll (version for 64-bit Windows).
 
-  See build_win32.txt for the build instructions, if you want to rebuild opencv_ffmpeg*.dll from scratch.
-
   The pre-built opencv_ffmpeg*.dll is:
   * LGPL library, not BSD libraries.
   * Loaded at runtime by opencv_videoio module.
     If it succeeds, ffmpeg can be used to decode/encode videos;
     otherwise, other API is used.
 
-  FFMPEG build contains H264 encoder based on the OpenH264 library, that should be installed separatelly.
+  FFMPEG build includes support for H264 encoder based on the OpenH264 library.
   OpenH264 Video Codec provided by Cisco Systems, Inc.
   See https://github.com/cisco/openh264/releases for details and OpenH264 license.
-  Downloaded binary file can be placed into global system path (System32 or SysWOW64) or near application binaries.
-  You can also specify location of binary file via OPENH264_LIBRARY_PATH environment variable.
+  OpenH264 library should be installed separatelly. Downloaded binary file can be placed into global system path
+  (System32 or SysWOW64) or near application binaries (check documentation of "LoadLibrary" Win32 function from MSDN).
+  Or you can specify location of binary file via OPENH264_LIBRARY environment variable.
 
   If LGPL/GPL software can not be supplied with your OpenCV-based product, simply exclude
   opencv_ffmpeg*.dll from your distribution; OpenCV will stay fully functional except for the ability to

3rdparty/include/dshow/_mingw_dxhelper.h

@@ -1,110 +0,0 @@
-/**
- * This file has no copyright assigned and is placed in the Public Domain.
- * This file is part of the w64 mingw-runtime package.
- * No warranty is given; refer to the file DISCLAIMER within this package.
- */
-
-#if defined(_MSC_VER) && !defined(_MSC_EXTENSIONS)
-#define NONAMELESSUNION		1
-#endif
-#if defined(NONAMELESSSTRUCT) && \
-   !defined(NONAMELESSUNION)
-#define NONAMELESSUNION		1
-#endif
-#if defined(NONAMELESSUNION)  && \
-   !defined(NONAMELESSSTRUCT)
-#define NONAMELESSSTRUCT	1
-#endif
-
-#ifndef __ANONYMOUS_DEFINED
-#define __ANONYMOUS_DEFINED
-#if defined(__GNUC__) || defined(__GNUG__)
-#define _ANONYMOUS_UNION	__extension__
-#define _ANONYMOUS_STRUCT	__extension__
-#else
-#define _ANONYMOUS_UNION
-#define _ANONYMOUS_STRUCT
-#endif
-#ifndef NONAMELESSUNION
-#define _UNION_NAME(x)
-#define _STRUCT_NAME(x)
-#else /* NONAMELESSUNION */
-#define _UNION_NAME(x)  x
-#define _STRUCT_NAME(x) x
-#endif
-#endif	/* __ANONYMOUS_DEFINED */
-
-#ifndef DUMMYUNIONNAME
-# ifdef NONAMELESSUNION
-#  define DUMMYUNIONNAME  u
-#  define DUMMYUNIONNAME1 u1	/* Wine uses this variant */
-#  define DUMMYUNIONNAME2 u2
-#  define DUMMYUNIONNAME3 u3
-#  define DUMMYUNIONNAME4 u4
-#  define DUMMYUNIONNAME5 u5
-#  define DUMMYUNIONNAME6 u6
-#  define DUMMYUNIONNAME7 u7
-#  define DUMMYUNIONNAME8 u8
-#  define DUMMYUNIONNAME9 u9
-# else /* NONAMELESSUNION */
-#  define DUMMYUNIONNAME
-#  define DUMMYUNIONNAME1	/* Wine uses this variant */
-#  define DUMMYUNIONNAME2
-#  define DUMMYUNIONNAME3
-#  define DUMMYUNIONNAME4
-#  define DUMMYUNIONNAME5
-#  define DUMMYUNIONNAME6
-#  define DUMMYUNIONNAME7
-#  define DUMMYUNIONNAME8
-#  define DUMMYUNIONNAME9
-# endif
-#endif	/* DUMMYUNIONNAME */
-
-#if !defined(DUMMYUNIONNAME1)	/* MinGW does not define this one */
-# ifdef NONAMELESSUNION
-#  define DUMMYUNIONNAME1 u1	/* Wine uses this variant */
-# else
-#  define DUMMYUNIONNAME1	/* Wine uses this variant */
-# endif
-#endif	/* DUMMYUNIONNAME1 */
-
-#ifndef DUMMYSTRUCTNAME
-# ifdef NONAMELESSUNION
-#  define DUMMYSTRUCTNAME  s
-#  define DUMMYSTRUCTNAME1 s1	/* Wine uses this variant */
-#  define DUMMYSTRUCTNAME2 s2
-#  define DUMMYSTRUCTNAME3 s3
-#  define DUMMYSTRUCTNAME4 s4
-#  define DUMMYSTRUCTNAME5 s5
-# else
-#  define DUMMYSTRUCTNAME
-#  define DUMMYSTRUCTNAME1	/* Wine uses this variant */
-#  define DUMMYSTRUCTNAME2
-#  define DUMMYSTRUCTNAME3
-#  define DUMMYSTRUCTNAME4
-#  define DUMMYSTRUCTNAME5
-# endif
-#endif /* DUMMYSTRUCTNAME */
-
-/* These are for compatibility with the Wine source tree */
-
-#ifndef WINELIB_NAME_AW
-# ifdef __MINGW_NAME_AW
-#   define WINELIB_NAME_AW  __MINGW_NAME_AW
-# else
-#  ifdef UNICODE
-#   define WINELIB_NAME_AW(func) func##W
-#  else
-#   define WINELIB_NAME_AW(func) func##A
-#  endif
-# endif
-#endif	/* WINELIB_NAME_AW */
-
-#ifndef DECL_WINELIB_TYPE_AW
-# ifdef __MINGW_TYPEDEF_AW
-#  define DECL_WINELIB_TYPE_AW  __MINGW_TYPEDEF_AW
-# else
-#  define DECL_WINELIB_TYPE_AW(type)  typedef WINELIB_NAME_AW(type) type;
-# endif
-#endif	/* DECL_WINELIB_TYPE_AW */
-

3rdparty/include/dshow/_mingw_unicode.h

@@ -1,33 +0,0 @@
-/**
- * This file has no copyright assigned and is placed in the Public Domain.
- * This file is part of the w64 mingw-runtime package.
- * No warranty is given; refer to the file DISCLAIMER.PD within this package.
- */
-
-#if !defined(_INC_CRT_UNICODE_MACROS)
-/* _INC_CRT_UNICODE_MACROS defined based on UNICODE flag */
-
-#if defined(UNICODE)
-# define _INC_CRT_UNICODE_MACROS 1
-# define __MINGW_NAME_AW(func) func##W
-# define __MINGW_NAME_AW_EXT(func,ext) func##W##ext
-# define __MINGW_NAME_UAW(func) func##_W
-# define __MINGW_NAME_UAW_EXT(func,ext) func##_W_##ext
-# define __MINGW_STRING_AW(str) L##str	/* same as TEXT() from winnt.h */
-# define __MINGW_PROCNAMEEXT_AW "W"
-#else
-# define _INC_CRT_UNICODE_MACROS 2
-# define __MINGW_NAME_AW(func) func##A
-# define __MINGW_NAME_AW_EXT(func,ext) func##A##ext
-# define __MINGW_NAME_UAW(func) func##_A
-# define __MINGW_NAME_UAW_EXT(func,ext) func##_A_##ext
-# define __MINGW_STRING_AW(str) str	/* same as TEXT() from winnt.h */
-# define __MINGW_PROCNAMEEXT_AW "A"
-#endif
-
-#define __MINGW_TYPEDEF_AW(type)	\
-    typedef __MINGW_NAME_AW(type) type;
-#define __MINGW_TYPEDEF_UAW(type)	\
-    typedef __MINGW_NAME_UAW(type) type;
-
-#endif /* !defined(_INC_CRT_UNICODE_MACROS) */

3rdparty/include/dshow/audevcod.h

@@ -1,31 +0,0 @@
-#ifndef _AUDEVCOD_H
-#define _AUDEVCOD_H
-#if __GNUC__ >=3
-#pragma GCC system_header
-#endif
-
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-typedef enum _tagSND_DEVICE_ERROR {
-    SNDDEV_ERROR_Open = 1,
-    SNDDEV_ERROR_Close = 2,
-    SNDDEV_ERROR_GetCaps = 3,
-    SNDDEV_ERROR_PrepareHeader = 4,
-    SNDDEV_ERROR_UnprepareHeader = 5,
-    SNDDEV_ERROR_Reset = 6,
-    SNDDEV_ERROR_Restart = 7,
-    SNDDEV_ERROR_GetPosition = 8,
-    SNDDEV_ERROR_Write = 9,
-    SNDDEV_ERROR_Pause = 10,
-    SNDDEV_ERROR_Stop = 11,
-    SNDDEV_ERROR_Start = 12,
-    SNDDEV_ERROR_AddBuffer = 13,
-    SNDDEV_ERROR_Query = 14
-} SNDDEV_ERR;
-
-#ifdef __cplusplus
-}
-#endif
-#endif

3rdparty/include/dshow/bdatypes.h

@@ -1,32 +0,0 @@
-#ifndef _BDATYPES_H
-#define _BDATYPES_H
-#if __GNUC__ >= 3
-#pragma GCC system_header
-#endif
-
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/*--- DirectShow Reference - DirectShow Enumerated Types */
-typedef enum {
-    MEDIA_TRANSPORT_PACKET,
-    MEDIA_ELEMENTARY_STREAM,
-    MEDIA_MPEG2_PSI,
-    MEDIA_TRANSPORT_PAYLOAD
-} MEDIA_SAMPLE_CONTENT;
-/*--- DirectShow Reference - DirectShow Structures */
-typedef struct {
-    DWORD dwOffset;
-    DWORD dwPacketLength;
-    DWORD dwStride;
-} MPEG2_TRANSPORT_STRIDE;
-typedef struct {
-    ULONG ulPID;
-    MEDIA_SAMPLE_CONTENT MediaSampleContent ;
-} PID_MAP;
-
-#ifdef __cplusplus
-}
-#endif
-#endif

3rdparty/include/dshow/dshow.h

@@ -1,61 +0,0 @@
-/*
- * Copyright (C) 2002 Alexandre Julliard
- *
- * This library is free software; you can redistribute it and/or
- * modify it under the terms of the GNU Lesser General Public
- * License as published by the Free Software Foundation; either
- * version 2.1 of the License, or (at your option) any later version.
- *
- * This library 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
- * Lesser General Public License for more details.
- *
- * You should have received a copy of the GNU Lesser General Public
- * License along with this library; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
- */
-
-#ifndef __DSHOW_INCLUDED__
-#define __DSHOW_INCLUDED__
-
-#define AM_NOVTABLE
-
-#ifndef __WINESRC__
-# include <windows.h>
-# include <windowsx.h>
-#else
-# include <windef.h>
-# include <wingdi.h>
-# include <objbase.h>
-#endif
-#include <olectl.h>
-#include <dshow/ddraw.h>
-#include <mmsystem.h>
-/* FIXME: #include <strsafe.h>*/
-
-#ifndef NUMELMS
-#define NUMELMS(array) (sizeof(array)/sizeof((array)[0]))
-#endif
-
-#include <dshow/strmif.h>
-#include <dshow/amvideo.h>
-#ifdef DSHOW_USE_AMAUDIO
-/* FIXME: #include <amaudio.h>*/
-#endif
-#include <dshow/control.h>
-#include <dshow/evcode.h>
-#include <dshow/uuids.h>
-#include <dshow/errors.h>
-/* FIXME: #include <edevdefs.h> */
-#include <dshow/audevcod.h>
-/* FIXME: #include <dvdevcod.h> */
-
-#ifndef OATRUE
-#define OATRUE (-1)
-#endif
-#ifndef OAFALSE
-#define OAFALSE (0)
-#endif
-
-#endif /* __DSHOW_INCLUDED__ */

3rdparty/include/dshow/dvdmedia.h

@@ -1,75 +0,0 @@
-/*
- * Copyright (C) 2008 Maarten Lankhorst
- *
- * This library is free software; you can redistribute it and/or
- * modify it under the terms of the GNU Lesser General Public
- * License as published by the Free Software Foundation; either
- * version 2.1 of the License, or (at your option) any later version.
- *
- * This library 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
- * Lesser General Public License for more details.
- *
- * You should have received a copy of the GNU Lesser General Public
- * License along with this library; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
- */
-
-#ifndef __DVDMEDIA_H__
-#define __DVDMEDIA_H__
-
-#define AMCONTROL_USED 0x00000001
-#define AMCONTROL_PAD_TO_4x3 0x00000002
-#define AMCONTROL_PAD_TO_16x9 0x00000004
-
-enum AM_MPEG2Level {
-    AM_MPEG2Level_Low = 1,
-    AM_MPEG2Level_Main,
-    AM_MPEG2Level_High1440,
-    AM_MPEG2Level_High
-};
-enum AM_MPEG2Profile {
-    AM_MPEG2Profile_Simple = 1,
-    AM_MPEG2Profile_Main,
-    AM_MPEG2Profile_SNRScalable,
-    AM_MPEG2Profile_SpatiallyScalable,
-    AM_MPEG2Profile_High
-};
-typedef enum {
-    AM_RATE_ChangeRate = 1,
-    AM_RATE_FullDataRateMax = 2,
-    AM_RATE_ReverseDecode = 3,
-    AM_RATE_DecoderPosition = 4,
-    AM_RATE_DecoderVersion = 5
-} AM_PROPERTY_DVD_RATE_CHANGE;
-
-typedef struct tagVIDEOINFOHEADER2 {
-    RECT rcSource;
-    RECT rcTarget;
-    DWORD dwBitRate;
-    DWORD dwBitErrorRate;
-    REFERENCE_TIME AvgTimePerFrame;
-    DWORD dwInterlaceFlags;
-    DWORD dwCopyProtectFlags;
-    DWORD dwPictAspectRatioX;
-    DWORD dwPictAspectRatioY;
-    union {
-        DWORD dwControlFlags;
-        DWORD dwReserved1;
-    } DUMMYUNIONNAME;
-    DWORD dwReserved2;
-    BITMAPINFOHEADER bmiHeader;
-} VIDEOINFOHEADER2;
-
-typedef struct tagMPEG2VIDEOINFO {
-    VIDEOINFOHEADER2 hdr;
-    DWORD dwStartTimeCode;
-    DWORD cbSequenceHeader;
-    DWORD dwProfile;
-    DWORD dwLevel;
-    DWORD dwFlags;
-    DWORD dwSequenceHeader[1];
-} MPEG2VIDEOINFO;
-
-#endif /* __DVDMEDIA_H__ */

3rdparty/include/dshow/errors.h

@@ -1,169 +0,0 @@
-#ifndef _ERRORS_H
-#define _ERRORS_H
-#if __GNUC__ >=3
-#pragma GCC system_header
-#endif
-
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/*--- DirectShow Reference - Constants and GUIDs - Error and Success Codes */
-#define VFW_S_NO_MORE_ITEMS 0x00040103
-#define VFW_S_DUPLICATE_NAME 0x0004022D
-#define VFW_S_STATE_INTERMEDIATE 0x00040237
-#define VFW_S_PARTIAL_RENDER 0x00040242
-#define VFW_S_SOME_DATA_IGNORED 0x00040245
-#define VFW_S_CONNECTIONS_DEFERRED 0x00040246
-#define VFW_S_RESOURCE_NOT_NEEDED 0x00040250
-#define VFW_S_MEDIA_TYPE_IGNORED 0x00040254
-#define VFW_S_VIDEO_NOT_RENDERED 0x00040257
-#define VFW_S_AUDIO_NOT_RENDERED 0x00040258
-#define VFW_S_RPZA 0x0004025A
-#define VFW_S_ESTIMATED 0x00040260
-#define VFW_S_RESERVED 0x00040263
-#define VFW_S_STREAM_OFF 0x00040267
-#define VFW_S_CANT_CUE 0x00040268
-#define VFW_S_NOPREVIEWPIN 0x0004027E
-#define VFW_S_DVD_NON_ONE_SEQUENTIAL 0x00040280
-#define VFW_S_DVD_CHANNEL_CONTENTS_NOT_AVAILABLE 0x0004028C
-#define VFW_S_DVD_NOT_ACCURATE 0x0004028D
-#define VFW_E_INVALIDMEDIATYPE 0x80040200
-#define VFW_E_INVALIDSUBTYPE 0x80040201
-#define VFW_E_NEED_OWNER 0x80040202
-#define VFW_E_ENUM_OUT_OF_SYNC 0x80040203
-#define VFW_E_ALREADY_CONNECTED 0x80040204
-#define VFW_E_FILTER_ACTIVE 0x80040205
-#define VFW_E_NO_TYPES 0x80040206
-#define VFW_E_NO_ACCEPTABLE_TYPES 0x80040207
-#define VFW_E_INVALID_DIRECTION 0x80040208
-#define VFW_E_NOT_CONNECTED 0x80040209
-#define VFW_E_NO_ALLOCATOR 0x8004020A
-#define VFW_E_RUNTIME_ERROR 0x8004020B
-#define VFW_E_BUFFER_NOTSET 0x8004020C
-#define VFW_E_BUFFER_OVERFLOW 0x8004020D
-#define VFW_E_BADALIGN 0x8004020E
-#define VFW_E_ALREADY_COMMITTED 0x8004020F
-#define VFW_E_BUFFERS_OUTSTANDING 0x80040210
-#define VFW_E_NOT_COMMITTED 0x80040211
-#define VFW_E_SIZENOTSET 0x80040212
-#define VFW_E_NO_CLOCK 0x80040213
-#define VFW_E_NO_SINK 0x80040214
-#define VFW_E_NO_INTERFACE 0x80040215
-#define VFW_E_NOT_FOUND 0x80040216
-#define VFW_E_CANNOT_CONNECT 0x80040217
-#define VFW_E_CANNOT_RENDER 0x80040218
-#define VFW_E_CHANGING_FORMAT 0x80040219
-#define VFW_E_NO_COLOR_KEY_SET 0x8004021A
-#define VFW_E_NOT_OVERLAY_CONNECTION 0x8004021B
-#define VFW_E_NOT_SAMPLE_CONNECTION 0x8004021C
-#define VFW_E_PALETTE_SET 0x8004021D
-#define VFW_E_COLOR_KEY_SET 0x8004021E
-#define VFW_E_NO_COLOR_KEY_FOUND 0x8004021F
-#define VFW_E_NO_PALETTE_AVAILABLE 0x80040220
-#define VFW_E_NO_DISPLAY_PALETTE 0x80040221
-#define VFW_E_TOO_MANY_COLORS 0x80040222
-#define VFW_E_STATE_CHANGED 0x80040223
-#define VFW_E_NOT_STOPPED 0x80040224
-#define VFW_E_NOT_PAUSED 0x80040225
-#define VFW_E_NOT_RUNNING 0x80040226
-#define VFW_E_WRONG_STATE 0x80040227
-#define VFW_E_START_TIME_AFTER_END 0x80040228
-#define VFW_E_INVALID_RECT 0x80040229
-#define VFW_E_TYPE_NOT_ACCEPTED 0x8004022A
-#define VFW_E_SAMPLE_REJECTED 0x8004022B
-#define VFW_E_SAMPLE_REJECTED_EOS 0x8004022C
-#define VFW_E_DUPLICATE_NAME 0x8004022D
-#define VFW_E_TIMEOUT 0x8004022E
-#define VFW_E_INVALID_FILE_FORMAT 0x8004022F
-#define VFW_E_ENUM_OUT_OF_RANGE 0x80040230
-#define VFW_E_CIRCULAR_GRAPH 0x80040231
-#define VFW_E_NOT_ALLOWED_TO_SAVE 0x80040232
-#define VFW_E_TIME_ALREADY_PASSED 0x80040233
-#define VFW_E_ALREADY_CANCELLED 0x80040234
-#define VFW_E_CORRUPT_GRAPH_FILE 0x80040235
-#define VFW_E_ADVISE_ALREADY_SET 0x80040236
-#define VFW_E_NO_MODEX_AVAILABLE 0x80040238
-#define VFW_E_NO_ADVISE_SET 0x80040239
-#define VFW_E_NO_FULLSCREEN 0x8004023A
-#define VFW_E_IN_FULLSCREEN_MODE 0x8004023B
-#define VFW_E_UNKNOWN_FILE_TYPE 0x80040240
-#define VFW_E_CANNOT_LOAD_SOURCE_FILTER 0x80040241
-#define VFW_E_FILE_TOO_SHORT 0x80040243
-#define VFW_E_INVALID_FILE_VERSION 0x80040244
-#define VFW_E_INVALID_CLSID 0x80040247
-#define VFW_E_INVALID_MEDIA_TYPE 0x80040248
-#define VFW_E_SAMPLE_TIME_NOT_SET 0x80040249
-#define VFW_E_MEDIA_TIME_NOT_SET 0x80040251
-#define VFW_E_NO_TIME_FORMAT_SET 0x80040252
-#define VFW_E_MONO_AUDIO_HW 0x80040253
-#define VFW_E_NO_DECOMPRESSOR 0x80040255
-#define VFW_E_NO_AUDIO_HARDWARE 0x80040256
-#define VFW_E_RPZA 0x80040259
-#define VFW_E_PROCESSOR_NOT_SUITABLE 0x8004025B
-#define VFW_E_UNSUPPORTED_AUDIO 0x8004025C
-#define VFW_E_UNSUPPORTED_VIDEO 0x8004025D
-#define VFW_E_MPEG_NOT_CONSTRAINED 0x8004025E
-#define VFW_E_NOT_IN_GRAPH 0x8004025F
-#define VFW_E_NO_TIME_FORMAT 0x80040261
-#define VFW_E_READ_ONLY 0x80040262
-#define VFW_E_BUFFER_UNDERFLOW 0x80040264
-#define VFW_E_UNSUPPORTED_STREAM 0x80040265
-#define VFW_E_NO_TRANSPORT 0x80040266
-#define VFW_E_BAD_VIDEOCD 0x80040269
-#define VFW_S_NO_STOP_TIME 0x80040270
-#define VFW_E_OUT_OF_VIDEO_MEMORY 0x80040271
-#define VFW_E_VP_NEGOTIATION_FAILED 0x80040272
-#define VFW_E_DDRAW_CAPS_NOT_SUITABLE 0x80040273
-#define VFW_E_NO_VP_HARDWARE 0x80040274
-#define VFW_E_NO_CAPTURE_HARDWARE 0x80040275
-#define VFW_E_DVD_OPERATION_INHIBITED 0x80040276
-#define VFW_E_DVD_INVALIDDOMAIN 0x80040277
-#define VFW_E_DVD_NO_BUTTON 0x80040278
-#define VFW_E_DVD_GRAPHNOTREADY 0x80040279
-#define VFW_E_DVD_RENDERFAIL 0x8004027A
-#define VFW_E_DVD_DECNOTENOUGH 0x8004027B
-#define VFW_E_DDRAW_VERSION_NOT_SUITABLE 0x8004027C
-#define VFW_E_COPYPROT_FAILED 0x8004027D
-#define VFW_E_TIME_EXPIRED 0x8004027F
-#define VFW_E_DVD_WRONG_SPEED 0x80040281
-#define VFW_E_DVD_MENU_DOES_NOT_EXIST 0x80040282
-#define VFW_E_DVD_CMD_CANCELLED 0x80040283
-#define VFW_E_DVD_STATE_WRONG_VERSION 0x80040284
-#define VFW_E_DVD_STATE_CORRUPT 0x80040285
-#define VFW_E_DVD_STATE_WRONG_DISC 0x80040286
-#define VFW_E_DVD_INCOMPATIBLE_REGION 0x80040287
-#define VFW_E_DVD_NO_ATTRIBUTES 0x80040288
-#define VFW_E_DVD_NO_GOUP_PGC 0x80040289
-#define VFW_E_DVD_LOW_PARENTAL_LEVEL 0x8004028A
-#define VFW_E_DVD_NOT_IN_KARAOKE_MODE 0x8004028B
-#define VFW_E_FRAME_STEP_UNSUPPORTED 0x8004028E
-#define VFW_E_DVD_STREAM_DISABLED 0x8004028F
-#define VFW_E_DVD_TITLE_UNKNOWN 0x80040290
-#define VFW_E_DVD_INVALID_DISC 0x80040291
-#define VFW_E_DVD_NO_RESUME_INFORMATION 0x80040292
-#define VFW_E_PIN_ALREADY_BLOCKED_ON_THIS_THREAD 0x80040293
-#define VFW_E_PIN_ALREADY_BLOCKED 0x80040294
-#define VFW_E_CERTIFICATION_FAILURE 0x80040295
-#define VFW_E_VMR_NOT_IN_MIXER_MODE 0x80040296
-#define VFW_E_VMR_NO_AP_SUPPLIED 0x80040297
-#define VFW_E_VMR_NO_DEINTERLACE_HW 0x80040298
-#define VFW_E_VMR_NO_PROCAMP_HW 0x80040299
-#define VFW_E_DVD_VMR9_INCOMPATIBLEDEC 0x8004029A
-#define VFW_E_NO_COPP_HW 0x8004029B
-#define VFW_E_BAD_KEY 0x800403F2
-/*--- DirectShow Reference - Functions */
-#define MAX_ERROR_TEXT_LEN 160
-/*--- DirectShow Reference - Functions */
-DWORD WINAPI AMGetErrorTextA(HRESULT,CHAR*,DWORD);
-DWORD WINAPI AMGetErrorTextW(HRESULT,WCHAR*,DWORD);
-#ifdef UNICODE
-#define AMGetErrorText AMGetErrorTextW
-#else
-#define AMGetErrorText AMGetErrorTextA
-#endif
-
-#ifdef __cplusplus
-}
-#endif
-#endif

3rdparty/include/dshow/evcode.h

@@ -1,68 +0,0 @@
-#ifndef _EVCODE_H
-#define _EVCODE_H
-#if __GNUC__ >=3
-#pragma GCC system_header
-#endif
-
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/*--- DirectShow Reference - Constants and GUIDs - Event Notification Codes */
-#define EC_ACTIVATE 0x0013
-#define EC_BUFFERING_DATA 0x0011
-#define EC_BUILT 0x0300
-#define EC_CLOCK_CHANGED 0x000D
-#define EC_CLOCK_UNSET 0x0051
-#define EC_CODECAPI_EVENT 0x0057
-#define EC_COMPLETE 0x0001
-#define EC_DEVICE_LOST 0x001F
-#define EC_DISPLAY_CHANGED 0x0016
-#define EC_END_OF_SEGMENT 0x001C
-#define EC_ERROR_STILLPLAYING 0x0008
-#define EC_ERRORABORT 0x0003
-#define EC_EXTDEVICE_MODE_CHANGE 0x0031
-#define EC_FULLSCREEN_LOST 0x0012
-#define EC_GRAPH_CHANGED 0x0050
-#define EC_LENGTH_CHANGED 0x001E
-#define EC_NEED_RESTART 0x0014
-#define EC_NOTIFY_WINDOW 0x0019
-#define EC_OLE_EVENT 0x0018
-#define EC_OPENING_FILE 0x0010
-#define EC_PALETTE_CHANGED 0x0009
-#define EC_PAUSED 0x000E
-#define EC_PREPROCESS_COMPLETE 0x0056
-#define EC_QUALITY_CHANGE 0x000B
-#define EC_REPAINT 0x0005
-#define EC_SEGMENT_STARTED 0x001D
-#define EC_SHUTTING_DOWN 0x000C
-#define EC_SNDDEV_IN_ERROR 0x0200
-#define EC_SNDDEV_OUT_ERROR 0x0201
-#define EC_STARVATION 0x0017
-#define EC_STATE_CHANGE 0x0032
-#define EC_STEP_COMPLETE 0x0024
-#define EC_STREAM_CONTROL_STARTED 0x001B
-#define EC_STREAM_CONTROL_STOPPED 0x001A
-#define EC_STREAM_ERROR_STILLPLAYING 0x0007
-#define EC_STREAM_ERROR_STOPPED 0x0006
-#define EC_TIMECODE_AVAILABLE 0x0030
-#define EC_UNBUILT 0x0301
-#define EC_USERABORT 0x0002
-#define EC_VIDEO_SIZE_CHANGED 0x000A
-#define EC_VMR_RENDERDEVICE_SET 0x0053
-#define EC_VMR_SURFACE_FLIPPED 0x0054
-#define EC_VMR_RECONNECTION_FAILED 0x0055
-#define EC_WINDOW_DESTROYED 0x0015
-#define EC_WMT_EVENT 0x0252
-#define EC_WMT_INDEX_EVENT 0x0251
-#define EC_USER 0x8000
-/*--- DirectShow Reference - DirectShow Structures */
-typedef struct {
-    HRESULT hrStatus;
-    void   *pData;
-} AM_WMT_EVENT_DATA;
-
-#ifdef __cplusplus
-}
-#endif
-#endif

3rdparty/include/dshow/ksuuids.h

@@ -1,191 +0,0 @@
-/**
- * This file has no copyright assigned and is placed in the Public Domain.
- * This file is part of the w64 mingw-runtime package.
- * No warranty is given; refer to the file DISCLAIMER.PD within this package.
- */
-
-OUR_GUID_ENTRY(MEDIATYPE_MPEG2_PACK,
-        0x36523B13,0x8EE5,0x11d1,0x8C,0xA3,0x00,0x60,0xB0,0x57,0x66,0x4A)
-
-OUR_GUID_ENTRY(MEDIATYPE_MPEG2_PES,
-        0xe06d8020,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x5f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_WMDRM_TRANSPORT,
-        0x18BEC4EA,0x4676,0x450e,0xB4,0x78,0x0C,0xD8,0x4C,0x54,0xB3,0x27)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_VIDEO,
-        0xe06d8026,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x5f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(FORMAT_MPEG2_VIDEO,
-        0xe06d80e3,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x5f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(FORMAT_VIDEOINFO2,
-        0xf72a76A0L,0xeb0a,0x11d0,0xac,0xe4,0x0,0x0,0xc0,0xcc,0x16,0xba)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_PROGRAM,
-        0xe06d8022,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_TRANSPORT,
-        0xe06d8023,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_AUDIO,
-        0xe06d802b,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DOLBY_AC3,
-        0xe06d802c,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVD_SUBPICTURE,
-        0xe06d802d,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVD_LPCM_AUDIO,
-        0xe06d8032,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIATYPE_DVD_ENCRYPTED_PACK,
-        0xed0b916a,0x044d,0x11d1,0xaa,0x78,0x00,0xc0,0x04f,0xc3,0x1d,0x60)
-
-OUR_GUID_ENTRY(MEDIATYPE_DVD_NAVIGATION,
-        0xe06d802e,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVD_NAVIGATION_PCI,
-        0xe06d802f,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVD_NAVIGATION_DSI,
-        0xe06d8030,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVD_NAVIGATION_PROVIDER,
-        0xe06d8031,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(FORMAT_MPEG2Video,
-        0xe06d80e3,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(FORMAT_DolbyAC3,
-        0xe06d80e4,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(FORMAT_MPEG2Audio,
-        0xe06d80e5,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(FORMAT_DVD_LPCMAudio,
-        0xe06d80e6,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_AC3,
-        0xBFABE720,0x6E1F,0x11D0,0xBC,0xF2,0x44,0x45,0x53,0x54,0x00,0x00)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_DvdSubPic,
-        0xac390460,0x43af,0x11d0,0xbd,0x6a,0x00,0x35,0x05,0xc1,0x03,0xa9)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_CopyProt,
-        0x0E8A0A40,0x6AEF,0x11D0,0x9E,0xD0,0x00,0xA0,0x24,0xCA,0x19,0xB3)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_TSRateChange,
-        0xa503c5c0,0x1d1d,0x11d1,0xad,0x80,0x44,0x45,0x53,0x54,0x0,0x0)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_MPEG4_MediaType_Attributes,
-        0xff6c4bfa,0x7a9,0x4c7b,0xa2,0x37,0x67,0x2f,0x9d,0x68,0x6,0x5f)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_CAPTURE,
-        0x65E8773DL,0x8F56,0x11D0,0xA3,0xB9,0x00,0xA0,0xC9,0x22,0x31,0x96)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_RENDER,
-        0x65E8773EL,0x8F56,0x11D0,0xA3,0xB9,0x00,0xA0,0xC9,0x22,0x31,0x96)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_DATACOMPRESSOR,
-        0x1E84C900L,0x7E70,0x11D0,0xA5,0xD6,0x28,0xDB,0x04,0xC1,0x00,0x00)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_AUDIO,
-        0x6994AD04L,0x93EF,0x11D0,0xA3,0xCC,0x00,0xA0,0xC9,0x22,0x31,0x96)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_VIDEO,
-        0x6994AD05L,0x93EF,0x11D0,0xA3,0xCC,0x00,0xA0,0xC9,0x22,0x31,0x96)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_TVTUNER,
-        0xa799a800L,0xa46d,0x11d0,0xa1,0x8c,0x00,0xa0,0x24,0x01,0xdc,0xd4)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_CROSSBAR,
-        0xa799a801L,0xa46d,0x11d0,0xa1,0x8c,0x00,0xa0,0x24,0x01,0xdc,0xd4)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_TVAUDIO,
-        0xa799a802L,0xa46d,0x11d0,0xa1,0x8c,0x00,0xa0,0x24,0x01,0xdc,0xd4)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_VBICODEC,
-        0x07dad660L,0x22f1,0x11d1,0xa9,0xf4,0x00,0xc0,0x4f,0xbb,0xde,0x8f)
-
-OUR_GUID_ENTRY(AM_KSCATEGORY_SPLITTER,
-        0x0A4252A0L,0x7E70,0x11D0,0xA5,0xD6,0x28,0xDB,0x04,0xC1,0x00,0x00)
-
-OUR_GUID_ENTRY(IID_IKsInterfaceHandler,
-        0xD3ABC7E0L,0x9A61,0x11D0,0xA4,0x0D,0x00,0xA0,0xC9,0x22,0x31,0x96)
-
-OUR_GUID_ENTRY(IID_IKsDataTypeHandler,
-        0x5FFBAA02L,0x49A3,0x11D0,0x9F,0x36,0x00,0xAA,0x00,0xA2,0x16,0xA1)
-
-OUR_GUID_ENTRY(IID_IKsPin,
-        0xb61178d1L,0xa2d9,0x11cf,0x9e,0x53,0x00,0xaa,0x00,0xa2,0x16,0xa1)
-
-OUR_GUID_ENTRY(IID_IKsControl,
-        0x28F54685L,0x06FD,0x11D2,0xB2,0x7A,0x00,0xA0,0xC9,0x22,0x31,0x96)
-
-OUR_GUID_ENTRY(IID_IKsPinFactory,
-        0xCD5EBE6BL,0x8B6E,0x11D1,0x8A,0xE0,0x00,0xA0,0xC9,0x22,0x31,0x96)
-
-OUR_GUID_ENTRY(AM_INTERFACESETID_Standard,
-        0x1A8766A0L,0x62CE,0x11CF,0xA5,0xD6,0x28,0xDB,0x04,0xC1,0x00,0x00)
-
-#if ( (NTDDI_VERSION >= NTDDI_WINXPSP2) && (NTDDI_VERSION < NTDDI_WS03) ) || (NTDDI_VERSION >= NTDDI_WS03SP1)
-OUR_GUID_ENTRY(MEDIATYPE_MPEG2_SECTIONS,
-        0x455f176c,0x4b06,0x47ce,0x9a,0xef,0x8c,0xae,0xf7,0x3d,0xf7,0xb5)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_VERSIONED_TABLES,
-        0x1ed988b0,0x3ffc,0x4523,0x87,0x25,0x34,0x7b,0xee,0xc1,0xa8,0xa0)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_ATSC_SI,
-        0xb3c7397c,0xd303,0x414d,0xb3,0x3c,0x4e,0xd2,0xc9,0xd2,0x97,0x33)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVB_SI,
-        0xe9dd31a3,0x221d,0x4adb,0x85,0x32,0x9a,0xf3,0x9,0xc1,0xa4,0x8)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_ISDB_SI,
-        0xe89ad298,0x3601,0x4b06,0xaa,0xec,0x9d,0xde,0xed,0xcc,0x5b,0xd0)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_TIF_SI,
-        0xec232eb2,0xcb96,0x4191,0xb2,0x26,0xe,0xa1,0x29,0xf3,0x82,0x50)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2DATA,
-        0xc892e55b,0x252d,0x42b5,0xa3,0x16,0xd9,0x97,0xe7,0xa5,0xd9,0x95)
-#endif
-/* ( (NTDDI_VERSION >= NTDDI_WINXPSP2) && (NTDDI_VERSION < NTDDI_WS03) ) ||
-                        (NTDDI_VERSION >= NTDDI_WS03SP1) */
-
-#if (NTDDI_VERSION >= NTDDI_WINXP)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_TRANSPORT_STRIDE,
-        0x138aa9a4,0x1ee2,0x4c5b,0x98,0x8e,0x19,0xab,0xfd,0xbc,0x8a,0x11)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_UDCR_TRANSPORT,
-        0x18BEC4EA,0x4676,0x450e,0xB4,0x78,0x0C,0xD8,0x4C,0x54,0xB3,0x27)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_PBDA_TRANSPORT_RAW,
-        0x0d7aed42,0xcb9a,0x11db,0x97,0x5,0x0,0x50,0x56,0xc0,0x0,0x8)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG2_PBDA_TRANSPORT_PROCESSED,
-        0xaf748dd4,0xd80,0x11db,0x97,0x5,0x0,0x50,0x56,0xc0,0x0,0x8)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_DTS,
-        0xe06d8033,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_SDDS,
-        0xe06d8034,0xdb46,0x11cf,0xb4,0xd1,0x00,0x80,0x05f,0x6c,0xbb,0xea)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_DVD_RateChange,
-        0x3577eb09,0x9582,0x477f,0xb2,0x9c,0xb0,0xc4,0x52,0xa4,0xff,0x9a)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_DvdKaraoke,
-        0xae4720ae,0xaa71,0x42d8,0xb8,0x2a,0xff,0xfd,0xf5,0x8b,0x76,0xfd)
-
-OUR_GUID_ENTRY(AM_KSPROPSETID_FrameStep,
-        0xc830acbd,0xab07,0x492f,0x88,0x52,0x45,0xb6,0x98,0x7c,0x29,0x79)
-#endif /* NTDDI_VERSION >= NTDDI_WINXP */
-
-#if (NTDDI_VERSION >= NTDDI_WS03SP1)
-OUR_GUID_ENTRY(AM_KSCATEGORY_VBICODEC_MI,
-        0x9c24a977,0x951,0x451a,0x80,0x6,0xe,0x49,0xbd,0x28,0xcd,0x5f)
-#endif /* NTDDI_VERSION >= NTDDI_WS03SP1 */
-

3rdparty/include/dshow/uuids.h

@@ -1,368 +0,0 @@
-/**
- * This file has no copyright assigned and is placed in the Public Domain.
- * This file is part of the w64 mingw-runtime package.
- * No warranty is given; refer to the file DISCLAIMER.PD within this package.
- */
-#ifndef OUR_GUID_ENTRY
-#define OUR_GUID_ENTRY(name,l,w1,w2,b1,b2,b3,b4,b5,b6,b7,b8) DEFINE_GUID(name,l,w1,w2,b1,b2,b3,b4,b5,b6,b7,b8);
-#endif
-
-#define MEDIATYPE_NULL GUID_NULL
-#define MEDIASUBTYPE_NULL GUID_NULL
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_None,0xe436eb8e,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIATYPE_Video,0x73646976,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_Audio,0x73647561,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_Text,0x73747874,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_Midi,0x7364696D,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_Stream,0xe436eb83,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIATYPE_Interleaved,0x73766169,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_File,0x656c6966,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_ScriptCommand,0x73636d64,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_AUXLine21Data,0x670aea80,0x3a82,0x11d0,0xb7,0x9b,0x0,0xaa,0x0,0x37,0x67,0xa7)
-OUR_GUID_ENTRY(MEDIATYPE_VBI,0xf72a76e1,0xeb0a,0x11d0,0xac,0xe4,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(MEDIATYPE_Timecode,0x482dee3,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIATYPE_LMRT,0x74726c6d,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_URL_STREAM,0x736c7275,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_CLPL,0x4C504C43,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_YUYV,0x56595559,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IYUV,0x56555949,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_YVU9,0x39555659,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Y411,0x31313459,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Y41P,0x50313459,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_YUY2,0x32595559,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_YVYU,0x55595659,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_UYVY,0x59565955,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Y211,0x31313259,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_CLJR,0x524a4c43,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IF09,0x39304649,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_CPLA,0x414c5043,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MJPG,0x47504A4D,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_TVMJ,0x4A4D5654,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_WAKE,0x454B4157,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_CFCC,0x43434643,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IJPG,0x47504A49,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Plum,0x6D756C50,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVCS,0x53435644,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_DVSD,0x44535644,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MDVF,0x4656444D,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB1,0xe436eb78,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB4,0xe436eb79,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB8,0xe436eb7a,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB565,0xe436eb7b,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB555,0xe436eb7c,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB24,0xe436eb7d,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB32,0xe436eb7e,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB1555,0x297c55af,0xe209,0x4cb3,0xb7,0x57,0xc7,0x6d,0x6b,0x9c,0x88,0xa8)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB4444,0x6e6415e6,0x5c24,0x425f,0x93,0xcd,0x80,0x10,0x2b,0x3d,0x1c,0xca)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB32,0x773c9ac0,0x3274,0x11d0,0xb7,0x24,0x0,0xaa,0x0,0x6c,0x1a,0x1)
-OUR_GUID_ENTRY(MEDIASUBTYPE_A2R10G10B10,0x2f8bb76d,0xb644,0x4550,0xac,0xf3,0xd3,0x0c,0xaa,0x65,0xd5,0xc5)
-OUR_GUID_ENTRY(MEDIASUBTYPE_A2B10G10R10,0x576f7893,0xbdf6,0x48c4,0x87,0x5f,0xae,0x7b,0x81,0x83,0x45,0x67)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AYUV,0x56555941,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AI44,0x34344941,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IA44,0x34344149,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB32_D3D_DX7_RT,0x32335237,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB16_D3D_DX7_RT,0x36315237,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB32_D3D_DX7_RT,0x38384137,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB4444_D3D_DX7_RT,0x34344137,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB1555_D3D_DX7_RT,0x35314137,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB32_D3D_DX9_RT,0x32335239,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RGB16_D3D_DX9_RT,0x36315239,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB32_D3D_DX9_RT,0x38384139,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB4444_D3D_DX9_RT,0x34344139,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_ARGB1555_D3D_DX9_RT,0x35314139,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-
-#define MEDIASUBTYPE_HASALPHA(mt) (((mt).subtype==MEDIASUBTYPE_ARGB4444) || ((mt).subtype==MEDIASUBTYPE_ARGB32) || ((mt).subtype==MEDIASUBTYPE_AYUV) || ((mt).subtype==MEDIASUBTYPE_AI44) || ((mt).subtype==MEDIASUBTYPE_IA44) || ((mt).subtype==MEDIASUBTYPE_ARGB1555) || ((mt).subtype==MEDIASUBTYPE_ARGB32_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB4444_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB1555_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB32_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB4444_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB1555_D3D_DX9_RT))
-#define MEDIASUBTYPE_HASALPHA7(mt) (((mt).subtype==MEDIASUBTYPE_ARGB32_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB4444_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB1555_D3D_DX7_RT))
-#define MEDIASUBTYPE_D3D_DX7_RT(mt) (((mt).subtype==MEDIASUBTYPE_ARGB32_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB4444_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB1555_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_RGB32_D3D_DX7_RT) || ((mt).subtype==MEDIASUBTYPE_RGB16_D3D_DX7_RT))
-#define MEDIASUBTYPE_HASALPHA9(mt) (((mt).subtype==MEDIASUBTYPE_ARGB32_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB4444_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB1555_D3D_DX9_RT))
-#define MEDIASUBTYPE_D3D_DX9_RT(mt) (((mt).subtype==MEDIASUBTYPE_ARGB32_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB4444_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_ARGB1555_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_RGB32_D3D_DX9_RT) || ((mt).subtype==MEDIASUBTYPE_RGB16_D3D_DX9_RT))
-
-OUR_GUID_ENTRY(MEDIASUBTYPE_YV12,0x32315659,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_NV12,0x3231564E,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IMC1,0x31434D49,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IMC2,0x32434D49,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IMC3,0x33434D49,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IMC4,0x34434D49,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_S340,0x30343353,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_S342,0x32343353,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Overlay,0xe436eb7f,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG1Packet,0xe436eb80,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG1Payload,0xe436eb81,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG1AudioPayload,0x00000050,0x0000,0x0010,0x80,0x00,0x00,0xAA,0x00,0x38,0x9B,0x71)
-OUR_GUID_ENTRY(MEDIATYPE_MPEG1SystemStream,0xe436eb82,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG1System,0xe436eb84,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG1VideoCD,0xe436eb85,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG1Video,0xe436eb86,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_MPEG1Audio,0xe436eb87,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Avi,0xe436eb88,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Asf,0x3db80f90,0x9412,0x11d1,0xad,0xed,0x0,0x0,0xf8,0x75,0x4b,0x99)
-OUR_GUID_ENTRY(MEDIASUBTYPE_QTMovie,0xe436eb89,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_QTRpza,0x617a7072,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_QTSmc,0x20636d73,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_QTRle,0x20656c72,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_QTJpeg,0x6765706a,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_PCMAudio_Obsolete,0xe436eb8a,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_PCM,0x00000001,0x0000,0x0010,0x80,0x00,0x00,0xAA,0x00,0x38,0x9B,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_WAVE,0xe436eb8b,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AU,0xe436eb8c,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AIFF,0xe436eb8d,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(MEDIASUBTYPE_dvsd,0x64737664,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_dvhd,0x64687664,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_dvsl,0x6c737664,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_dv25,0x35327664,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_dv50,0x30357664,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_dvh1,0x31687664,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Line21_BytePair,0x6e8d4a22,0x310c,0x11d0,0xb7,0x9a,0x0,0xaa,0x0,0x37,0x67,0xa7)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Line21_GOPPacket,0x6e8d4a23,0x310c,0x11d0,0xb7,0x9a,0x0,0xaa,0x0,0x37,0x67,0xa7)
-OUR_GUID_ENTRY(MEDIASUBTYPE_Line21_VBIRawData,0x6e8d4a24,0x310c,0x11d0,0xb7,0x9a,0x0,0xaa,0x0,0x37,0x67,0xa7)
-OUR_GUID_ENTRY(MEDIASUBTYPE_TELETEXT,0xf72a76e3,0xeb0a,0x11d0,0xac,0xe4,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(MEDIASUBTYPE_WSS,0x2791D576,0x8E7A,0x466F,0x9E,0x90,0x5D,0x3F,0x30,0x83,0x73,0x8B)
-OUR_GUID_ENTRY(MEDIASUBTYPE_VPS,0xa1b3f620,0x9792,0x4d8d,0x81,0xa4,0x86,0xaf,0x25,0x77,0x20,0x90)
-OUR_GUID_ENTRY(MEDIASUBTYPE_DRM_Audio,0x00000009,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_IEEE_FLOAT,0x00000003,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_DOLBY_AC3_SPDIF,0x00000092,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_RAW_SPORT,0x00000240,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_SPDIF_TAG_241h,0x00000241,0x0000,0x0010,0x80,0x00,0x00,0xaa,0x00,0x38,0x9b,0x71)
-OUR_GUID_ENTRY(MEDIASUBTYPE_DssVideo,0xa0af4f81,0xe163,0x11d0,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(MEDIASUBTYPE_DssAudio,0xa0af4f82,0xe163,0x11d0,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(MEDIASUBTYPE_VPVideo,0x5a9b6a40,0x1a22,0x11d1,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(MEDIASUBTYPE_VPVBI,0x5a9b6a41,0x1a22,0x11d1,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(CLSID_CaptureGraphBuilder,0xBF87B6E0,0x8C27,0x11d0,0xB3,0xF0,0x0,0xAA,0x00,0x37,0x61,0xC5)
-OUR_GUID_ENTRY(CLSID_CaptureGraphBuilder2,0xBF87B6E1,0x8C27,0x11d0,0xB3,0xF0,0x0,0xAA,0x00,0x37,0x61,0xC5)
-OUR_GUID_ENTRY(CLSID_ProtoFilterGraph,0xe436ebb0,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_SystemClock,0xe436ebb1,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_FilterMapper,0xe436ebb2,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_FilterGraph,0xe436ebb3,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_FilterGraphNoThread,0xe436ebb8,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_MPEG1Doc,0xe4bbd160,0x4269,0x11ce,0x83,0x8d,0x0,0xaa,0x0,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(CLSID_FileSource,0x701722e0,0x8ae3,0x11ce,0xa8,0x5c,0x00,0xaa,0x00,0x2f,0xea,0xb5)
-OUR_GUID_ENTRY(CLSID_MPEG1PacketPlayer,0x26c25940,0x4ca9,0x11ce,0xa8,0x28,0x0,0xaa,0x0,0x2f,0xea,0xb5)
-OUR_GUID_ENTRY(CLSID_MPEG1Splitter,0x336475d0,0x942a,0x11ce,0xa8,0x70,0x00,0xaa,0x00,0x2f,0xea,0xb5)
-OUR_GUID_ENTRY(CLSID_CMpegVideoCodec,0xfeb50740,0x7bef,0x11ce,0x9b,0xd9,0x0,0x0,0xe2,0x2,0x59,0x9c)
-OUR_GUID_ENTRY(CLSID_CMpegAudioCodec,0x4a2286e0,0x7bef,0x11ce,0x9b,0xd9,0x0,0x0,0xe2,0x2,0x59,0x9c)
-OUR_GUID_ENTRY(CLSID_TextRender,0xe30629d3,0x27e5,0x11ce,0x87,0x5d,0x0,0x60,0x8c,0xb7,0x80,0x66)
-OUR_GUID_ENTRY(CLSID_InfTee,0xf8388a40,0xd5bb,0x11d0,0xbe,0x5a,0x0,0x80,0xc7,0x6,0x56,0x8e)
-OUR_GUID_ENTRY(CLSID_AviSplitter,0x1b544c20,0xfd0b,0x11ce,0x8c,0x63,0x0,0xaa,0x00,0x44,0xb5,0x1e)
-OUR_GUID_ENTRY(CLSID_AviReader,0x1b544c21,0xfd0b,0x11ce,0x8c,0x63,0x0,0xaa,0x00,0x44,0xb5,0x1e)
-OUR_GUID_ENTRY(CLSID_VfwCapture,0x1b544c22,0xfd0b,0x11ce,0x8c,0x63,0x0,0xaa,0x00,0x44,0xb5,0x1e)
-OUR_GUID_ENTRY(CLSID_CaptureProperties,0x1B544c22,0xFD0B,0x11ce,0x8C,0x63,0x00,0xAA,0x00,0x44,0xB5,0x1F)
-OUR_GUID_ENTRY(CLSID_FGControl,0xe436ebb4,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_MOVReader,0x44584800,0xf8ee,0x11ce,0xb2,0xd4,0x00,0xdd,0x1,0x10,0x1b,0x85)
-OUR_GUID_ENTRY(CLSID_QuickTimeParser,0xd51bd5a0,0x7548,0x11cf,0xa5,0x20,0x0,0x80,0xc7,0x7e,0xf5,0x8a)
-OUR_GUID_ENTRY(CLSID_QTDec,0xfdfe9681,0x74a3,0x11d0,0xaf,0xa7,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(CLSID_AVIDoc,0xd3588ab0,0x0781,0x11ce,0xb0,0x3a,0x00,0x20,0xaf,0xb,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_VideoRenderer,0x70e102b0,0x5556,0x11ce,0x97,0xc0,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(CLSID_Colour,0x1643e180,0x90f5,0x11ce,0x97,0xd5,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(CLSID_Dither,0x1da08500,0x9edc,0x11cf,0xbc,0x10,0x00,0xaa,0x00,0xac,0x74,0xf6)
-OUR_GUID_ENTRY(CLSID_ModexRenderer,0x7167665,0x5011,0x11cf,0xbf,0x33,0x0,0xaa,0x0,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(CLSID_AudioRender,0xe30629d1,0x27e5,0x11ce,0x87,0x5d,0x0,0x60,0x8c,0xb7,0x80,0x66)
-OUR_GUID_ENTRY(CLSID_AudioProperties,0x05589faf,0xc356,0x11ce,0xbf,0x01,0x0,0xaa,0x0,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(CLSID_DSoundRender,0x79376820,0x07D0,0x11CF,0xA2,0x4D,0x0,0x20,0xAF,0xD7,0x97,0x67)
-OUR_GUID_ENTRY(CLSID_AudioRecord,0xe30629d2,0x27e5,0x11ce,0x87,0x5d,0x0,0x60,0x8c,0xb7,0x80,0x66)
-OUR_GUID_ENTRY(CLSID_AudioInputMixerProperties,0x2ca8ca52,0x3c3f,0x11d2,0xb7,0x3d,0x0,0xc0,0x4f,0xb6,0xbd,0x3d)
-OUR_GUID_ENTRY(CLSID_AVIDec,0xcf49d4e0,0x1115,0x11ce,0xb0,0x3a,0x0,0x20,0xaf,0xb,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_AVIDraw,0xa888df60,0x1e90,0x11cf,0xac,0x98,0x0,0xaa,0x0,0x4c,0xf,0xa9)
-OUR_GUID_ENTRY(CLSID_ACMWrapper,0x6a08cf80,0x0e18,0x11cf,0xa2,0x4d,0x0,0x20,0xaf,0xd7,0x97,0x67)
-OUR_GUID_ENTRY(CLSID_AsyncReader,0xe436ebb5,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_URLReader,0xe436ebb6,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_PersistMonikerPID,0xe436ebb7,0x524f,0x11ce,0x9f,0x53,0x00,0x20,0xaf,0x0b,0xa7,0x70)
-OUR_GUID_ENTRY(CLSID_AVICo,0xd76e2820,0x1563,0x11cf,0xac,0x98,0x0,0xaa,0x0,0x4c,0xf,0xa9)
-OUR_GUID_ENTRY(CLSID_FileWriter,0x8596e5f0,0xda5,0x11d0,0xbd,0x21,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_AviDest,0xe2510970,0xf137,0x11ce,0x8b,0x67,0x0,0xaa,0x0,0xa3,0xf1,0xa6)
-OUR_GUID_ENTRY(CLSID_AviMuxProptyPage,0xc647b5c0,0x157c,0x11d0,0xbd,0x23,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_AviMuxProptyPage1,0xa9ae910,0x85c0,0x11d0,0xbd,0x42,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_AVIMIDIRender,0x07b65360,0xc445,0x11ce,0xaf,0xde,0x00,0xaa,0x00,0x6c,0x14,0xf4)
-OUR_GUID_ENTRY(CLSID_WMAsfReader,0x187463a0,0x5bb7,0x11d3,0xac,0xbe,0x0,0x80,0xc7,0x5e,0x24,0x6e)
-OUR_GUID_ENTRY(CLSID_WMAsfWriter,0x7c23220e,0x55bb,0x11d3,0x8b,0x16,0x0,0xc0,0x4f,0xb6,0xbd,0x3d)
-OUR_GUID_ENTRY(CLSID_MPEG2Demultiplexer,0xafb6c280,0x2c41,0x11d3,0x8a,0x60,0x00,0x00,0xf8,0x1e,0x0e,0x4a)
-OUR_GUID_ENTRY(CLSID_MMSPLITTER,0x3ae86b20,0x7be8,0x11d1,0xab,0xe6,0x00,0xa0,0xc9,0x05,0xf3,0x75)
-OUR_GUID_ENTRY(CLSID_StreamBufferSink,0x2db47ae5,0xcf39,0x43c2,0xb4,0xd6,0xc,0xd8,0xd9,0x9,0x46,0xf4)
-OUR_GUID_ENTRY(CLSID_StreamBufferSource,0xc9f5fe02,0xf851,0x4eb5,0x99,0xee,0xad,0x60,0x2a,0xf1,0xe6,0x19)
-OUR_GUID_ENTRY(CLSID_StreamBufferConfig,0xfa8a68b2,0xc864,0x4ba2,0xad,0x53,0xd3,0x87,0x6a,0x87,0x49,0x4b)
-OUR_GUID_ENTRY(CLSID_Mpeg2VideoStreamAnalyzer,0x6cfad761,0x735d,0x4aa5,0x8a,0xfc,0xaf,0x91,0xa7,0xd6,0x1e,0xba)
-OUR_GUID_ENTRY(CLSID_StreamBufferRecordingAttributes,0xccaa63ac,0x1057,0x4778,0xae,0x92,0x12,0x6,0xab,0x9a,0xce,0xe6)
-OUR_GUID_ENTRY(CLSID_StreamBufferComposeRecording,0xd682c4ba,0xa90a,0x42fe,0xb9,0xe1,0x3,0x10,0x98,0x49,0xc4,0x23)
-OUR_GUID_ENTRY(CLSID_DVVideoCodec,0xb1b77c00,0xc3e4,0x11cf,0xaf,0x79,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(CLSID_DVVideoEnc,0x13aa3650,0xbb6f,0x11d0,0xaf,0xb9,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(CLSID_DVSplitter,0x4eb31670,0x9fc6,0x11cf,0xaf,0x6e,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(CLSID_DVMux,0x129d7e40,0xc10d,0x11d0,0xaf,0xb9,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(CLSID_SeekingPassThru,0x60af76c,0x68dd,0x11d0,0x8f,0xc1,0x0,0xc0,0x4f,0xd9,0x18,0x9d)
-OUR_GUID_ENTRY(CLSID_Line21Decoder,0x6e8d4a20,0x310c,0x11d0,0xb7,0x9a,0x0,0xaa,0x0,0x37,0x67,0xa7)
-OUR_GUID_ENTRY(CLSID_Line21Decoder2,0xe4206432,0x01a1,0x4bee,0xb3,0xe1,0x37,0x02,0xc8,0xed,0xc5,0x74)
-OUR_GUID_ENTRY(CLSID_OverlayMixer,0xcd8743a1,0x3736,0x11d0,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(CLSID_VBISurfaces,0x814b9800,0x1c88,0x11d1,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(CLSID_WSTDecoder,0x70bc06e0,0x5666,0x11d3,0xa1,0x84,0x0,0x10,0x5a,0xef,0x9f,0x33)
-OUR_GUID_ENTRY(CLSID_MjpegDec,0x301056d0,0x6dff,0x11d2,0x9e,0xeb,0x0,0x60,0x8,0x3,0x9e,0x37)
-OUR_GUID_ENTRY(CLSID_MJPGEnc,0xb80ab0a0,0x7416,0x11d2,0x9e,0xeb,0x0,0x60,0x8,0x3,0x9e,0x37)
-OUR_GUID_ENTRY(CLSID_SystemDeviceEnum,0x62BE5D10,0x60EB,0x11d0,0xBD,0x3B,0x00,0xA0,0xC9,0x11,0xCE,0x86)
-OUR_GUID_ENTRY(CLSID_CDeviceMoniker,0x4315D437,0x5B8C,0x11d0,0xBD,0x3B,0x00,0xA0,0xC9,0x11,0xCE,0x86)
-OUR_GUID_ENTRY(CLSID_VideoInputDeviceCategory,0x860BB310,0x5D01,0x11d0,0xBD,0x3B,0x00,0xA0,0xC9,0x11,0xCE,0x86)
-OUR_GUID_ENTRY(CLSID_CVidCapClassManager,0x860BB310,0x5D01,0x11d0,0xBD,0x3B,0x00,0xA0,0xC9,0x11,0xCE,0x86)
-OUR_GUID_ENTRY(CLSID_LegacyAmFilterCategory,0x083863F1,0x70DE,0x11d0,0xBD,0x40,0x00,0xA0,0xC9,0x11,0xCE,0x86)
-OUR_GUID_ENTRY(CLSID_CQzFilterClassManager,0x083863F1,0x70DE,0x11d0,0xBD,0x40,0x00,0xA0,0xC9,0x11,0xCE,0x86)
-OUR_GUID_ENTRY(CLSID_VideoCompressorCategory,0x33d9a760,0x90c8,0x11d0,0xbd,0x43,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_CIcmCoClassManager,0x33d9a760,0x90c8,0x11d0,0xbd,0x43,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_AudioCompressorCategory,0x33d9a761,0x90c8,0x11d0,0xbd,0x43,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_CAcmCoClassManager,0x33d9a761,0x90c8,0x11d0,0xbd,0x43,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_AudioInputDeviceCategory,0x33d9a762,0x90c8,0x11d0,0xbd,0x43,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_CWaveinClassManager,0x33d9a762,0x90c8,0x11d0,0xbd,0x43,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_AudioRendererCategory,0xe0f158e1,0xcb04,0x11d0,0xbd,0x4e,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_CWaveOutClassManager,0xe0f158e1,0xcb04,0x11d0,0xbd,0x4e,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_MidiRendererCategory,0x4EfE2452,0x168A,0x11d1,0xBC,0x76,0x0,0xc0,0x4F,0xB9,0x45,0x3B)
-OUR_GUID_ENTRY(CLSID_CMidiOutClassManager,0x4EfE2452,0x168A,0x11d1,0xBC,0x76,0x0,0xc0,0x4F,0xB9,0x45,0x3B)
-OUR_GUID_ENTRY(CLSID_TransmitCategory,0xcc7bfb41,0xf175,0x11d1,0xa3,0x92,0x0,0xe0,0x29,0x1f,0x39,0x59)
-OUR_GUID_ENTRY(CLSID_DeviceControlCategory,0xcc7bfb46,0xf175,0x11d1,0xa3,0x92,0x0,0xe0,0x29,0x1f,0x39,0x59)
-OUR_GUID_ENTRY(CLSID_ActiveMovieCategories,0xda4e3da0,0xd07d,0x11d0,0xbd,0x50,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_DVDHWDecodersCategory,0x2721AE20,0x7E70,0x11D0,0xA5,0xD6,0x28,0xDB,0x04,0xC1,0x00,0x00)
-OUR_GUID_ENTRY(CLSID_MediaEncoderCategory,0x7D22E920,0x5CA9,0x4787,0x8C,0x2B,0xA6,0x77,0x9B,0xD1,0x17,0x81)
-OUR_GUID_ENTRY(CLSID_MediaMultiplexerCategory,0x236C9559,0xADCE,0x4736,0xBF,0x72,0xBA,0xB3,0x4E,0x39,0x21,0x96)
-OUR_GUID_ENTRY(CLSID_FilterMapper2,0xcda42200,0xbd88,0x11d0,0xbd,0x4e,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_MemoryAllocator,0x1e651cc0,0xb199,0x11d0,0x82,0x12,0x00,0xc0,0x4f,0xc3,0x2c,0x45)
-OUR_GUID_ENTRY(CLSID_MediaPropertyBag,0xcdbd8d00,0xc193,0x11d0,0xbd,0x4e,0x0,0xa0,0xc9,0x11,0xce,0x86)
-OUR_GUID_ENTRY(CLSID_DvdGraphBuilder,0xFCC152B7,0xF372,0x11d0,0x8E,0x00,0x00,0xC0,0x4F,0xD7,0xC0,0x8B)
-OUR_GUID_ENTRY(CLSID_DVDNavigator,0x9b8c4620,0x2c1a,0x11d0,0x84,0x93,0x0,0xa0,0x24,0x38,0xad,0x48)
-OUR_GUID_ENTRY(CLSID_DVDState,0xf963c5cf,0xa659,0x4a93,0x96,0x38,0xca,0xf3,0xcd,0x27,0x7d,0x13)
-OUR_GUID_ENTRY(CLSID_SmartTee,0xcc58e280,0x8aa1,0x11d1,0xb3,0xf1,0x0,0xaa,0x0,0x37,0x61,0xc5)
-OUR_GUID_ENTRY(FORMAT_None,0x0F6417D6,0xc318,0x11d0,0xa4,0x3f,0x00,0xa0,0xc9,0x22,0x31,0x96)
-OUR_GUID_ENTRY(FORMAT_VideoInfo,0x05589f80,0xc356,0x11ce,0xbf,0x01,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(FORMAT_VideoInfo2,0xf72a76A0,0xeb0a,0x11d0,0xac,0xe4,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(FORMAT_WaveFormatEx,0x05589f81,0xc356,0x11ce,0xbf,0x01,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(FORMAT_MPEGVideo,0x05589f82,0xc356,0x11ce,0xbf,0x01,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(FORMAT_MPEGStreams,0x05589f83,0xc356,0x11ce,0xbf,0x01,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(FORMAT_DvInfo,0x05589f84,0xc356,0x11ce,0xbf,0x01,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(CLSID_DirectDrawProperties,0x944d4c00,0xdd52,0x11ce,0xbf,0x0e,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(CLSID_PerformanceProperties,0x59ce6880,0xacf8,0x11cf,0xb5,0x6e,0x00,0x80,0xc7,0xc4,0xb6,0x8a)
-OUR_GUID_ENTRY(CLSID_QualityProperties,0x418afb70,0xf8b8,0x11ce,0xaa,0xc6,0x00,0x20,0xaf,0x0b,0x99,0xa3)
-OUR_GUID_ENTRY(IID_IBaseVideoMixer,0x61ded640,0xe912,0x11ce,0xa0,0x99,0x00,0xaa,0x00,0x47,0x9a,0x58)
-OUR_GUID_ENTRY(IID_IDirectDrawVideo,0x36d39eb0,0xdd75,0x11ce,0xbf,0x0e,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(IID_IQualProp,0x1bd0ecb0,0xf8e2,0x11ce,0xaa,0xc6,0x00,0x20,0xaf,0x0b,0x99,0xa3)
-OUR_GUID_ENTRY(CLSID_VPObject,0xce292861,0xfc88,0x11d0,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(IID_IVPObject,0xce292862,0xfc88,0x11d0,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(IID_IVPControl,0x25df12c1,0x3de0,0x11d1,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(CLSID_VPVBIObject,0x814b9801,0x1c88,0x11d1,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(IID_IVPVBIObject,0x814b9802,0x1c88,0x11d1,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(IID_IVPConfig,0xbc29a660,0x30e3,0x11d0,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(IID_IVPNotify,0xc76794a1,0xd6c5,0x11d0,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(IID_IVPNotify2,0xebf47183,0x8764,0x11d1,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(IID_IVPVBIConfig,0xec529b00,0x1a1f,0x11d1,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(IID_IVPVBINotify,0xec529b01,0x1a1f,0x11d1,0xba,0xd9,0x0,0x60,0x97,0x44,0x11,0x1a)
-OUR_GUID_ENTRY(IID_IMixerPinConfig,0x593cdde1,0x759,0x11d1,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-OUR_GUID_ENTRY(IID_IMixerPinConfig2,0xebf47182,0x8764,0x11d1,0x9e,0x69,0x0,0xc0,0x4f,0xd7,0xc1,0x5b)
-#ifndef __DDRAW_INCLUDED__
-OUR_GUID_ENTRY(CLSID_DirectDraw,0xD7B70EE0,0x4340,0x11CF,0xB0,0x63,0x00,0x20,0xAF,0xC2,0xCD,0x35)
-OUR_GUID_ENTRY(CLSID_DirectDrawClipper,0x593817A0,0x7DB3,0x11CF,0xA2,0xDE,0x00,0xAA,0x00,0xb9,0x33,0x56)
-OUR_GUID_ENTRY(IID_IDirectDraw,0x6C14DB80,0xA733,0x11CE,0xA5,0x21,0x00,0x20,0xAF,0x0B,0xE5,0x60)
-OUR_GUID_ENTRY(IID_IDirectDraw2,0xB3A6F3E0,0x2B43,0x11CF,0xA2,0xDE,0x00,0xAA,0x00,0xB9,0x33,0x56)
-OUR_GUID_ENTRY(IID_IDirectDrawSurface,0x6C14DB81,0xA733,0x11CE,0xA5,0x21,0x00,0x20,0xAF,0x0B,0xE5,0x60)
-OUR_GUID_ENTRY(IID_IDirectDrawSurface2,0x57805885,0x6eec,0x11cf,0x94,0x41,0xa8,0x23,0x03,0xc1,0x0e,0x27)
-OUR_GUID_ENTRY(IID_IDirectDrawSurface3,0xDA044E00,0x69B2,0x11D0,0xA1,0xD5,0x00,0xAA,0x00,0xB8,0xDF,0xBB)
-OUR_GUID_ENTRY(IID_IDirectDrawSurface4,0x0B2B8630,0xAD35,0x11D0,0x8E,0xA6,0x00,0x60,0x97,0x97,0xEA,0x5B)
-OUR_GUID_ENTRY(IID_IDirectDrawSurface7,0x06675a80,0x3b9b,0x11d2,0xb9,0x2f,0x00,0x60,0x97,0x97,0xea,0x5b)
-OUR_GUID_ENTRY(IID_IDirectDrawPalette,0x6C14DB84,0xA733,0x11CE,0xA5,0x21,0x00,0x20,0xAF,0x0B,0xE5,0x60)
-OUR_GUID_ENTRY(IID_IDirectDrawClipper,0x6C14DB85,0xA733,0x11CE,0xA5,0x21,0x00,0x20,0xAF,0x0B,0xE5,0x60)
-OUR_GUID_ENTRY(IID_IDirectDrawColorControl,0x4B9F0EE0,0x0D7E,0x11D0,0x9B,0x06,0x00,0xA0,0xC9,0x03,0xA3,0xB8)
-#endif
-#ifndef __DVP_INCLUDED__
-OUR_GUID_ENTRY(IID_IDDVideoPortContainer,0x6C142760,0xA733,0x11CE,0xA5,0x21,0x00,0x20,0xAF,0x0B,0xE5,0x60)
-#endif
-#ifndef __DDKM_INCLUDED__
-OUR_GUID_ENTRY(IID_IDirectDrawKernel,0x8D56C120,0x6A08,0x11D0,0x9B,0x06,0x00,0xA0,0xC9,0x03,0xA3,0xB8)
-OUR_GUID_ENTRY(IID_IDirectDrawSurfaceKernel,0x60755DA0,0x6A40,0x11D0,0x9B,0x06,0x00,0xA0,0xC9,0x03,0xA3,0xB8)
-#endif
-OUR_GUID_ENTRY(CLSID_ModexProperties,0x0618aa30,0x6bc4,0x11cf,0xbf,0x36,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(IID_IFullScreenVideo,0xdd1d7110,0x7836,0x11cf,0xbf,0x47,0x00,0xaa,0x00,0x55,0x59,0x5a)
-OUR_GUID_ENTRY(IID_IFullScreenVideoEx,0x53479470,0xf1dd,0x11cf,0xbc,0x42,0x00,0xaa,0x00,0xac,0x74,0xf6)
-OUR_GUID_ENTRY(CLSID_DVDecPropertiesPage,0x101193c0,0xbfe,0x11d0,0xaf,0x91,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(CLSID_DVEncPropertiesPage,0x4150f050,0xbb6f,0x11d0,0xaf,0xb9,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(CLSID_DVMuxPropertyPage,0x4db880e0,0xc10d,0x11d0,0xaf,0xb9,0x0,0xaa,0x0,0xb6,0x7a,0x42)
-OUR_GUID_ENTRY(IID_IAMDirectSound,0x546f4260,0xd53e,0x11cf,0xb3,0xf0,0x0,0xaa,0x0,0x37,0x61,0xc5)
-OUR_GUID_ENTRY(IID_IMpegAudioDecoder,0xb45dd570,0x3c77,0x11d1,0xab,0xe1,0x00,0xa0,0xc9,0x05,0xf3,0x75)
-OUR_GUID_ENTRY(IID_IAMLine21Decoder,0x6e8d4a21,0x310c,0x11d0,0xb7,0x9a,0x0,0xaa,0x0,0x37,0x67,0xa7)
-OUR_GUID_ENTRY(IID_IAMWstDecoder,0xc056de21,0x75c2,0x11d3,0xa1,0x84,0x0,0x10,0x5a,0xef,0x9f,0x33)
-OUR_GUID_ENTRY(CLSID_WstDecoderPropertyPage,0x4e27f80,0x91e4,0x11d3,0xa1,0x84,0x0,0x10,0x5a,0xef,0x9f,0x33)
-OUR_GUID_ENTRY(FORMAT_AnalogVideo,0x482dde0,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIATYPE_AnalogVideo,0x482dde1,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_NTSC_M,0x482dde2,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_B,0x482dde5,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_D,0x482dde6,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_G,0x482dde7,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_H,0x482dde8,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_I,0x482dde9,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_M,0x482ddea,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_N,0x482ddeb,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_PAL_N_COMBO,0x482ddec,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_SECAM_B,0x482ddf0,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_SECAM_D,0x482ddf1,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_SECAM_G,0x482ddf2,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_SECAM_H,0x482ddf3,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_SECAM_K,0x482ddf4,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_SECAM_K1,0x482ddf5,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIASUBTYPE_AnalogVideo_SECAM_L,0x482ddf6,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(MEDIATYPE_AnalogAudio,0x482dee1,0x7817,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-
-#include "dshow/ksuuids.h"
-
-OUR_GUID_ENTRY(TIME_FORMAT_NONE,0L,0,0,0,0,0,0,0,0,0,0)
-OUR_GUID_ENTRY(TIME_FORMAT_FRAME,0x7b785570,0x8c82,0x11cf,0xbc,0xc,0x0,0xaa,0x0,0xac,0x74,0xf6)
-OUR_GUID_ENTRY(TIME_FORMAT_BYTE,0x7b785571,0x8c82,0x11cf,0xbc,0xc,0x0,0xaa,0x0,0xac,0x74,0xf6)
-OUR_GUID_ENTRY(TIME_FORMAT_SAMPLE,0x7b785572,0x8c82,0x11cf,0xbc,0xc,0x0,0xaa,0x0,0xac,0x74,0xf6)
-OUR_GUID_ENTRY(TIME_FORMAT_FIELD,0x7b785573,0x8c82,0x11cf,0xbc,0xc,0x0,0xaa,0x0,0xac,0x74,0xf6)
-OUR_GUID_ENTRY(TIME_FORMAT_MEDIA_TIME,0x7b785574,0x8c82,0x11cf,0xbc,0xc,0x0,0xaa,0x0,0xac,0x74,0xf6)
-OUR_GUID_ENTRY(AMPROPSETID_Pin,0x9b00f101,0x1567,0x11d1,0xb3,0xf1,0x0,0xaa,0x0,0x37,0x61,0xc5)
-OUR_GUID_ENTRY(PIN_CATEGORY_CAPTURE,0xfb6c4281,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_PREVIEW,0xfb6c4282,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_ANALOGVIDEOIN,0xfb6c4283,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_VBI,0xfb6c4284,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_VIDEOPORT,0xfb6c4285,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_NABTS,0xfb6c4286,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_EDS,0xfb6c4287,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_TELETEXT,0xfb6c4288,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_CC,0xfb6c4289,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_STILL,0xfb6c428a,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_TIMECODE,0xfb6c428b,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(PIN_CATEGORY_VIDEOPORT_VBI,0xfb6c428c,0x0353,0x11d1,0x90,0x5f,0x00,0x00,0xc0,0xcc,0x16,0xba)
-OUR_GUID_ENTRY(LOOK_UPSTREAM_ONLY,0xac798be0,0x98e3,0x11d1,0xb3,0xf1,0x0,0xaa,0x0,0x37,0x61,0xc5)
-OUR_GUID_ENTRY(LOOK_DOWNSTREAM_ONLY,0xac798be1,0x98e3,0x11d1,0xb3,0xf1,0x0,0xaa,0x0,0x37,0x61,0xc5)
-OUR_GUID_ENTRY(CLSID_TVTunerFilterPropertyPage,0x266eee41,0x6c63,0x11cf,0x8a,0x3,0x0,0xaa,0x0,0x6e,0xcb,0x65)
-OUR_GUID_ENTRY(CLSID_CrossbarFilterPropertyPage,0x71f96461,0x78f3,0x11d0,0xa1,0x8c,0x0,0xa0,0xc9,0x11,0x89,0x56)
-OUR_GUID_ENTRY(CLSID_TVAudioFilterPropertyPage,0x71f96463,0x78f3,0x11d0,0xa1,0x8c,0x0,0xa0,0xc9,0x11,0x89,0x56)
-OUR_GUID_ENTRY(CLSID_VideoProcAmpPropertyPage,0x71f96464,0x78f3,0x11d0,0xa1,0x8c,0x0,0xa0,0xc9,0x11,0x89,0x56)
-OUR_GUID_ENTRY(CLSID_CameraControlPropertyPage,0x71f96465,0x78f3,0x11d0,0xa1,0x8c,0x0,0xa0,0xc9,0x11,0x89,0x56)
-OUR_GUID_ENTRY(CLSID_AnalogVideoDecoderPropertyPage,0x71f96466,0x78f3,0x11d0,0xa1,0x8c,0x0,0xa0,0xc9,0x11,0x89,0x56)
-OUR_GUID_ENTRY(CLSID_VideoStreamConfigPropertyPage,0x71f96467,0x78f3,0x11d0,0xa1,0x8c,0x0,0xa0,0xc9,0x11,0x89,0x56)
-OUR_GUID_ENTRY(CLSID_AudioRendererAdvancedProperties,0x37e92a92,0xd9aa,0x11d2,0xbf,0x84,0x8e,0xf2,0xb1,0x55,0x5a,0xed)
-OUR_GUID_ENTRY(CLSID_VideoMixingRenderer,0xB87BEB7B,0x8D29,0x423f,0xAE,0x4D,0x65,0x82,0xC1,0x01,0x75,0xAC)
-OUR_GUID_ENTRY(CLSID_VideoRendererDefault,0x6BC1CFFA,0x8FC1,0x4261,0xAC,0x22,0xCF,0xB4,0xCC,0x38,0xDB,0x50)
-OUR_GUID_ENTRY(CLSID_AllocPresenter,0x99d54f63,0x1a69,0x41ae,0xaa,0x4d,0xc9,0x76,0xeb,0x3f,0x07,0x13)
-OUR_GUID_ENTRY(CLSID_AllocPresenterDDXclMode,0x4444ac9e,0x242e,0x471b,0xa3,0xc7,0x45,0xdc,0xd4,0x63,0x52,0xbc)
-OUR_GUID_ENTRY(CLSID_VideoPortManager,0x6f26a6cd,0x967b,0x47fd,0x87,0x4a,0x7a,0xed,0x2c,0x9d,0x25,0xa2)
-OUR_GUID_ENTRY(CLSID_VideoMixingRenderer9,0x51b4abf3,0x748f,0x4e3b,0xa2,0x76,0xc8,0x28,0x33,0x0e,0x92,0x6a)
-OUR_GUID_ENTRY(CLSID_ATSCNetworkProvider,0x0dad2fdd,0x5fd7,0x11d3,0x8f,0x50,0x00,0xc0,0x4f,0x79,0x71,0xe2)
-OUR_GUID_ENTRY(CLSID_ATSCNetworkPropertyPage,0xe3444d16,0x5ac4,0x4386,0x88,0xdf,0x13,0xfd,0x23,0x0e,0x1d,0xda)
-OUR_GUID_ENTRY(CLSID_DVBSNetworkProvider,0xfa4b375a,0x45b4,0x4d45,0x84,0x40,0x26,0x39,0x57,0xb1,0x16,0x23)
-OUR_GUID_ENTRY(CLSID_DVBTNetworkProvider,0x216c62df,0x6d7f,0x4e9a,0x85,0x71,0x5,0xf1,0x4e,0xdb,0x76,0x6a)
-OUR_GUID_ENTRY(CLSID_DVBCNetworkProvider,0xdc0c0fe7,0x485,0x4266,0xb9,0x3f,0x68,0xfb,0xf8,0xe,0xd8,0x34)
-OUR_GUID_ENTRY(CLSID_DShowTVEFilter,0x05500280,0xFAA5,0x4DF9,0x82,0x46,0xBF,0xC2,0x3A,0xC5,0xCE,0xA8)
-OUR_GUID_ENTRY(CLSID_TVEFilterTuneProperties,0x05500281,0xFAA5,0x4DF9,0x82,0x46,0xBF,0xC2,0x3A,0xC5,0xCE,0xA8)
-OUR_GUID_ENTRY(CLSID_TVEFilterCCProperties,0x05500282,0xFAA5,0x4DF9,0x82,0x46,0xBF,0xC2,0x3A,0xC5,0xCE,0xA8)
-OUR_GUID_ENTRY(CLSID_TVEFilterStatsProperties,0x05500283,0xFAA5,0x4DF9,0x82,0x46,0xBF,0xC2,0x3A,0xC5,0xCE,0xA8)
-OUR_GUID_ENTRY(CLSID_IVideoEncoderProxy,0xb43c4eec,0x8c32,0x4791,0x91,0x2,0x50,0x8a,0xda,0x5e,0xe8,0xe7)
-OUR_GUID_ENTRY(CLSID_ICodecAPIProxy,0x7ff0997a,0x1999,0x4286,0xa7,0x3c,0x62,0x2b,0x88,0x14,0xe7,0xeb)
-OUR_GUID_ENTRY(CLSID_IVideoEncoderCodecAPIProxy,0xb05dabd9,0x56e5,0x4fdc,0xaf,0xa4,0x8a,0x47,0xe9,0x1f,0x1c,0x9c)
-
-#ifndef __ENCODER_API_GUIDS__
-#define __ENCODER_API_GUIDS__
-OUR_GUID_ENTRY(ENCAPIPARAM_BITRATE,0x49cc4c43,0xca83,0x4ad4,0xa9,0xaf,0xf3,0x69,0x6a,0xf6,0x66,0xdf)
-OUR_GUID_ENTRY(ENCAPIPARAM_PEAK_BITRATE,0x703f16a9,0x3d48,0x44a1,0xb0,0x77,0x1,0x8d,0xff,0x91,0x5d,0x19)
-OUR_GUID_ENTRY(ENCAPIPARAM_BITRATE_MODE,0xee5fb25c,0xc713,0x40d1,0x9d,0x58,0xc0,0xd7,0x24,0x1e,0x25,0xf)
-OUR_GUID_ENTRY(CODECAPI_CHANGELISTS,0x62b12acf,0xf6b0,0x47d9,0x94,0x56,0x96,0xf2,0x2c,0x4e,0x0b,0x9d)
-OUR_GUID_ENTRY(CODECAPI_VIDEO_ENCODER,0x7112e8e1,0x3d03,0x47ef,0x8e,0x60,0x03,0xf1,0xcf,0x53,0x73,0x01)
-OUR_GUID_ENTRY(CODECAPI_AUDIO_ENCODER,0xb9d19a3e,0xf897,0x429c,0xbc,0x46,0x81,0x38,0xb7,0x27,0x2b,0x2d)
-OUR_GUID_ENTRY(CODECAPI_SETALLDEFAULTS,0x6c5e6a7c,0xacf8,0x4f55,0xa9,0x99,0x1a,0x62,0x81,0x09,0x05,0x1b)
-OUR_GUID_ENTRY(CODECAPI_ALLSETTINGS,0x6a577e92,0x83e1,0x4113,0xad,0xc2,0x4f,0xce,0xc3,0x2f,0x83,0xa1)
-OUR_GUID_ENTRY(CODECAPI_SUPPORTSEVENTS,0x0581af97,0x7693,0x4dbd,0x9d,0xca,0x3f,0x9e,0xbd,0x65,0x85,0xa1)
-OUR_GUID_ENTRY(CODECAPI_CURRENTCHANGELIST,0x1cb14e83,0x7d72,0x4657,0x83,0xfd,0x47,0xa2,0xc5,0xb9,0xd1,0x3d)
-#endif
-
-#undef OUR_GUID_ENTRY

3rdparty/include/msc_inttypes.h

@@ -1,301 +0,0 @@
-// ISO C9x  compliant inttypes.h for Microsoft Visual Studio
-// Based on ISO/IEC 9899:TC2 Committee draft (May 6, 2005) WG14/N1124
-//
-//  Copyright (c) 2006 Alexander Chemeris
-//
-// Redistribution and use in source and binary forms, with or without
-// modification, are permitted provided that the following conditions are met:
-//
-//   1. Redistributions of source code must retain the above copyright notice,
-//      this list of conditions and the following disclaimer.
-//
-//   2. Redistributions in binary form must reproduce the above copyright
-//      notice, this list of conditions and the following disclaimer in the
-//      documentation and/or other materials provided with the distribution.
-//
-//   3. The name of the author may be used to endorse or promote products
-//      derived from this software without specific prior written permission.
-//
-// THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
-// WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
-// MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
-// EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
-// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
-// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
-// OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
-// WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
-// OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
-// ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-//
-///////////////////////////////////////////////////////////////////////////////
-
-#ifndef _MSC_VER // [
-#error "Use this header only with Microsoft Visual C++ compilers!"
-#endif // _MSC_VER ]
-
-#ifndef _MSC_INTTYPES_H_ // [
-#define _MSC_INTTYPES_H_
-
-#if _MSC_VER > 1000
-#pragma once
-#endif
-
-#include <msc_stdint.h>
-
-// 7.8 Format conversion of integer types
-
-typedef struct {
-   intmax_t quot;
-   intmax_t rem;
-} imaxdiv_t;
-
-// 7.8.1 Macros for format specifiers
-
-// The fprintf macros for signed integers are:
-#define PRId8       "d"
-#define PRIi8       "i"
-#define PRIdLEAST8  "d"
-#define PRIiLEAST8  "i"
-#define PRIdFAST8   "d"
-#define PRIiFAST8   "i"
-
-#define PRId16       "hd"
-#define PRIi16       "hi"
-#define PRIdLEAST16  "hd"
-#define PRIiLEAST16  "hi"
-#define PRIdFAST16   "hd"
-#define PRIiFAST16   "hi"
-
-#define PRId32       "I32d"
-#define PRIi32       "I32i"
-#define PRIdLEAST32  "I32d"
-#define PRIiLEAST32  "I32i"
-#define PRIdFAST32   "I32d"
-#define PRIiFAST32   "I32i"
-
-#define PRId64       "I64d"
-#define PRIi64       "I64i"
-#define PRIdLEAST64  "I64d"
-#define PRIiLEAST64  "I64i"
-#define PRIdFAST64   "I64d"
-#define PRIiFAST64   "I64i"
-
-#define PRIdMAX     "I64d"
-#define PRIiMAX     "I64i"
-
-#define PRIdPTR     "Id"
-#define PRIiPTR     "Ii"
-
-// The fprintf macros for unsigned integers are:
-#define PRIo8       "o"
-#define PRIu8       "u"
-#define PRIx8       "x"
-#define PRIX8       "X"
-#define PRIoLEAST8  "o"
-#define PRIuLEAST8  "u"
-#define PRIxLEAST8  "x"
-#define PRIXLEAST8  "X"
-#define PRIoFAST8   "o"
-#define PRIuFAST8   "u"
-#define PRIxFAST8   "x"
-#define PRIXFAST8   "X"
-
-#define PRIo16       "ho"
-#define PRIu16       "hu"
-#define PRIx16       "hx"
-#define PRIX16       "hX"
-#define PRIoLEAST16  "ho"
-#define PRIuLEAST16  "hu"
-#define PRIxLEAST16  "hx"
-#define PRIXLEAST16  "hX"
-#define PRIoFAST16   "ho"
-#define PRIuFAST16   "hu"
-#define PRIxFAST16   "hx"
-#define PRIXFAST16   "hX"
-
-#define PRIo32       "I32o"
-#define PRIu32       "I32u"
-#define PRIx32       "I32x"
-#define PRIX32       "I32X"
-#define PRIoLEAST32  "I32o"
-#define PRIuLEAST32  "I32u"
-#define PRIxLEAST32  "I32x"
-#define PRIXLEAST32  "I32X"
-#define PRIoFAST32   "I32o"
-#define PRIuFAST32   "I32u"
-#define PRIxFAST32   "I32x"
-#define PRIXFAST32   "I32X"
-
-#define PRIo64       "I64o"
-#define PRIu64       "I64u"
-#define PRIx64       "I64x"
-#define PRIX64       "I64X"
-#define PRIoLEAST64  "I64o"
-#define PRIuLEAST64  "I64u"
-#define PRIxLEAST64  "I64x"
-#define PRIXLEAST64  "I64X"
-#define PRIoFAST64   "I64o"
-#define PRIuFAST64   "I64u"
-#define PRIxFAST64   "I64x"
-#define PRIXFAST64   "I64X"
-
-#define PRIoMAX     "I64o"
-#define PRIuMAX     "I64u"
-#define PRIxMAX     "I64x"
-#define PRIXMAX     "I64X"
-
-#define PRIoPTR     "Io"
-#define PRIuPTR     "Iu"
-#define PRIxPTR     "Ix"
-#define PRIXPTR     "IX"
-
-// The fscanf macros for signed integers are:
-#define SCNd8       "d"
-#define SCNi8       "i"
-#define SCNdLEAST8  "d"
-#define SCNiLEAST8  "i"
-#define SCNdFAST8   "d"
-#define SCNiFAST8   "i"
-
-#define SCNd16       "hd"
-#define SCNi16       "hi"
-#define SCNdLEAST16  "hd"
-#define SCNiLEAST16  "hi"
-#define SCNdFAST16   "hd"
-#define SCNiFAST16   "hi"
-
-#define SCNd32       "ld"
-#define SCNi32       "li"
-#define SCNdLEAST32  "ld"
-#define SCNiLEAST32  "li"
-#define SCNdFAST32   "ld"
-#define SCNiFAST32   "li"
-
-#define SCNd64       "I64d"
-#define SCNi64       "I64i"
-#define SCNdLEAST64  "I64d"
-#define SCNiLEAST64  "I64i"
-#define SCNdFAST64   "I64d"
-#define SCNiFAST64   "I64i"
-
-#define SCNdMAX     "I64d"
-#define SCNiMAX     "I64i"
-
-#ifdef _WIN64 // [
-#  define SCNdPTR     "I64d"
-#  define SCNiPTR     "I64i"
-#else  // _WIN64 ][
-#  define SCNdPTR     "ld"
-#  define SCNiPTR     "li"
-#endif  // _WIN64 ]
-
-// The fscanf macros for unsigned integers are:
-#define SCNo8       "o"
-#define SCNu8       "u"
-#define SCNx8       "x"
-#define SCNX8       "X"
-#define SCNoLEAST8  "o"
-#define SCNuLEAST8  "u"
-#define SCNxLEAST8  "x"
-#define SCNXLEAST8  "X"
-#define SCNoFAST8   "o"
-#define SCNuFAST8   "u"
-#define SCNxFAST8   "x"
-#define SCNXFAST8   "X"
-
-#define SCNo16       "ho"
-#define SCNu16       "hu"
-#define SCNx16       "hx"
-#define SCNX16       "hX"
-#define SCNoLEAST16  "ho"
-#define SCNuLEAST16  "hu"
-#define SCNxLEAST16  "hx"
-#define SCNXLEAST16  "hX"
-#define SCNoFAST16   "ho"
-#define SCNuFAST16   "hu"
-#define SCNxFAST16   "hx"
-#define SCNXFAST16   "hX"
-
-#define SCNo32       "lo"
-#define SCNu32       "lu"
-#define SCNx32       "lx"
-#define SCNX32       "lX"
-#define SCNoLEAST32  "lo"
-#define SCNuLEAST32  "lu"
-#define SCNxLEAST32  "lx"
-#define SCNXLEAST32  "lX"
-#define SCNoFAST32   "lo"
-#define SCNuFAST32   "lu"
-#define SCNxFAST32   "lx"
-#define SCNXFAST32   "lX"
-
-#define SCNo64       "I64o"
-#define SCNu64       "I64u"
-#define SCNx64       "I64x"
-#define SCNX64       "I64X"
-#define SCNoLEAST64  "I64o"
-#define SCNuLEAST64  "I64u"
-#define SCNxLEAST64  "I64x"
-#define SCNXLEAST64  "I64X"
-#define SCNoFAST64   "I64o"
-#define SCNuFAST64   "I64u"
-#define SCNxFAST64   "I64x"
-#define SCNXFAST64   "I64X"
-
-#define SCNoMAX     "I64o"
-#define SCNuMAX     "I64u"
-#define SCNxMAX     "I64x"
-#define SCNXMAX     "I64X"
-
-#ifdef _WIN64 // [
-#  define SCNoPTR     "I64o"
-#  define SCNuPTR     "I64u"
-#  define SCNxPTR     "I64x"
-#  define SCNXPTR     "I64X"
-#else  // _WIN64 ][
-#  define SCNoPTR     "lo"
-#  define SCNuPTR     "lu"
-#  define SCNxPTR     "lx"
-#  define SCNXPTR     "lX"
-#endif  // _WIN64 ]
-
-// 7.8.2 Functions for greatest-width integer types
-
-// 7.8.2.1 The imaxabs function
-#define imaxabs _abs64
-
-// 7.8.2.2 The imaxdiv function
-
-// This is modified version of div() function from Microsoft's div.c found
-// in %MSVC.NET%\crt\src\div.c
-#ifdef STATIC_IMAXDIV // [
-static
-#else // STATIC_IMAXDIV ][
-_inline
-#endif // STATIC_IMAXDIV ]
-imaxdiv_t __cdecl imaxdiv(intmax_t numer, intmax_t denom)
-{
-   imaxdiv_t result;
-
-   result.quot = numer / denom;
-   result.rem = numer % denom;
-
-   if (numer < 0 && result.rem > 0) {
-      // did division wrong; must fix up
-      ++result.quot;
-      result.rem -= denom;
-   }
-
-   return result;
-}
-
-// 7.8.2.3 The strtoimax and strtoumax functions
-#define strtoimax _strtoi64
-#define strtoumax _strtoui64
-
-// 7.8.2.4 The wcstoimax and wcstoumax functions
-#define wcstoimax _wcstoi64
-#define wcstoumax _wcstoui64
-
-
-#endif // _MSC_INTTYPES_H_ ]

3rdparty/include/msc_stdint.h

@@ -1,219 +0,0 @@
-// ISO C9x  compliant stdint.h for Microsoft Visual Studio
-// Based on ISO/IEC 9899:TC2 Committee draft (May 6, 2005) WG14/N1124
-//
-//  Copyright (c) 2006 Alexander Chemeris
-//
-// Redistribution and use in source and binary forms, with or without
-// modification, are permitted provided that the following conditions are met:
-//
-//   1. Redistributions of source code must retain the above copyright notice,
-//      this list of conditions and the following disclaimer.
-//
-//   2. Redistributions in binary form must reproduce the above copyright
-//      notice, this list of conditions and the following disclaimer in the
-//      documentation and/or other materials provided with the distribution.
-//
-//   3. The name of the author may be used to endorse or promote products
-//      derived from this software without specific prior written permission.
-//
-// THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
-// WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
-// MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
-// EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
-// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
-// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
-// OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
-// WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
-// OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
-// ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-//
-///////////////////////////////////////////////////////////////////////////////
-
-#ifndef _MSC_STDINT_H_ // [
-#define _MSC_STDINT_H_
-
-#include <limits.h>
-
-// For Visual Studio 6 in C++ mode wrap <wchar.h> include with 'extern "C++" {}'
-// or compiler give many errors like this:
-//   error C2733: second C linkage of overloaded function 'wmemchr' not allowed
-#if (_MSC_VER < 1300) && defined(__cplusplus)
-   extern "C++" {
-#endif
-#     include <wchar.h>
-#if (_MSC_VER < 1300) && defined(__cplusplus)
-   }
-#endif
-
-#if defined __GNUC__ || _MSC_VER >= 1600
-#include "stdint.h"
-#else
-
-// 7.18.1 Integer types
-
-// 7.18.1.1 Exact-width integer types
-typedef __int8            int8_t;
-typedef __int16           int16_t;
-typedef __int32           int32_t;
-typedef __int64           int64_t;
-typedef unsigned __int8   uint8_t;
-typedef unsigned __int16  uint16_t;
-typedef unsigned __int32  uint32_t;
-typedef unsigned __int64  uint64_t;
-
-// 7.18.1.2 Minimum-width integer types
-typedef int8_t    int_least8_t;
-typedef int16_t   int_least16_t;
-typedef int32_t   int_least32_t;
-typedef int64_t   int_least64_t;
-typedef uint8_t   uint_least8_t;
-typedef uint16_t  uint_least16_t;
-typedef uint32_t  uint_least32_t;
-typedef uint64_t  uint_least64_t;
-
-// 7.18.1.3 Fastest minimum-width integer types
-typedef int8_t    int_fast8_t;
-typedef int16_t   int_fast16_t;
-typedef int32_t   int_fast32_t;
-typedef int64_t   int_fast64_t;
-typedef uint8_t   uint_fast8_t;
-typedef uint16_t  uint_fast16_t;
-typedef uint32_t  uint_fast32_t;
-typedef uint64_t  uint_fast64_t;
-
-// 7.18.1.4 Integer types capable of holding object pointers
-#ifdef _WIN64 // [
-   typedef __int64           intptr_t;
-   typedef unsigned __int64  uintptr_t;
-#else // _WIN64 ][
-   typedef int               intptr_t;
-   typedef unsigned int      uintptr_t;
-#endif // _WIN64 ]
-
-// 7.18.1.5 Greatest-width integer types
-typedef int64_t   intmax_t;
-typedef uint64_t  uintmax_t;
-
-
-// 7.18.2 Limits of specified-width integer types
-
-#if !defined(__cplusplus) || defined(__STDC_LIMIT_MACROS) // [   See footnote 220 at page 257 and footnote 221 at page 259
-
-// 7.18.2.1 Limits of exact-width integer types
-#define INT8_MIN     _I8_MIN
-#define INT8_MAX     _I8_MAX
-#define INT16_MIN    _I16_MIN
-#define INT16_MAX    _I16_MAX
-#define INT32_MIN    _I32_MIN
-#define INT32_MAX    _I32_MAX
-#define INT64_MIN    _I64_MIN
-#define INT64_MAX    _I64_MAX
-#define UINT8_MAX    _UI8_MAX
-#define UINT16_MAX   _UI16_MAX
-#define UINT32_MAX   _UI32_MAX
-#define UINT64_MAX   _UI64_MAX
-
-// 7.18.2.2 Limits of minimum-width integer types
-#define INT_LEAST8_MIN    INT8_MIN
-#define INT_LEAST8_MAX    INT8_MAX
-#define INT_LEAST16_MIN   INT16_MIN
-#define INT_LEAST16_MAX   INT16_MAX
-#define INT_LEAST32_MIN   INT32_MIN
-#define INT_LEAST32_MAX   INT32_MAX
-#define INT_LEAST64_MIN   INT64_MIN
-#define INT_LEAST64_MAX   INT64_MAX
-#define UINT_LEAST8_MAX   UINT8_MAX
-#define UINT_LEAST16_MAX  UINT16_MAX
-#define UINT_LEAST32_MAX  UINT32_MAX
-#define UINT_LEAST64_MAX  UINT64_MAX
-
-// 7.18.2.3 Limits of fastest minimum-width integer types
-#define INT_FAST8_MIN    INT8_MIN
-#define INT_FAST8_MAX    INT8_MAX
-#define INT_FAST16_MIN   INT16_MIN
-#define INT_FAST16_MAX   INT16_MAX
-#define INT_FAST32_MIN   INT32_MIN
-#define INT_FAST32_MAX   INT32_MAX
-#define INT_FAST64_MIN   INT64_MIN
-#define INT_FAST64_MAX   INT64_MAX
-#define UINT_FAST8_MAX   UINT8_MAX
-#define UINT_FAST16_MAX  UINT16_MAX
-#define UINT_FAST32_MAX  UINT32_MAX
-#define UINT_FAST64_MAX  UINT64_MAX
-
-// 7.18.2.4 Limits of integer types capable of holding object pointers
-#ifdef _WIN64 // [
-#  define INTPTR_MIN   INT64_MIN
-#  define INTPTR_MAX   INT64_MAX
-#  define UINTPTR_MAX  UINT64_MAX
-#else // _WIN64 ][
-#  define INTPTR_MIN   INT32_MIN
-#  define INTPTR_MAX   INT32_MAX
-#  define UINTPTR_MAX  UINT32_MAX
-#endif // _WIN64 ]
-
-// 7.18.2.5 Limits of greatest-width integer types
-#define INTMAX_MIN   INT64_MIN
-#define INTMAX_MAX   INT64_MAX
-#define UINTMAX_MAX  UINT64_MAX
-
-// 7.18.3 Limits of other integer types
-
-#ifdef _WIN64 // [
-#  define PTRDIFF_MIN  _I64_MIN
-#  define PTRDIFF_MAX  _I64_MAX
-#else  // _WIN64 ][
-#  define PTRDIFF_MIN  _I32_MIN
-#  define PTRDIFF_MAX  _I32_MAX
-#endif  // _WIN64 ]
-
-#define SIG_ATOMIC_MIN  INT_MIN
-#define SIG_ATOMIC_MAX  INT_MAX
-
-#ifndef SIZE_MAX // [
-#  ifdef _WIN64 // [
-#     define SIZE_MAX  _UI64_MAX
-#  else // _WIN64 ][
-#     define SIZE_MAX  _UI32_MAX
-#  endif // _WIN64 ]
-#endif // SIZE_MAX ]
-
-// WCHAR_MIN and WCHAR_MAX are also defined in <wchar.h>
-#ifndef WCHAR_MIN // [
-#  define WCHAR_MIN  0
-#endif  // WCHAR_MIN ]
-#ifndef WCHAR_MAX // [
-#  define WCHAR_MAX  _UI16_MAX
-#endif  // WCHAR_MAX ]
-
-#define WINT_MIN  0
-#define WINT_MAX  _UI16_MAX
-
-#endif // __STDC_LIMIT_MACROS ]
-
-
-// 7.18.4 Limits of other integer types
-
-#if !defined(__cplusplus) || defined(__STDC_CONSTANT_MACROS) // [   See footnote 224 at page 260
-
-// 7.18.4.1 Macros for minimum-width integer constants
-
-#define INT8_C(val)  val
-#define INT16_C(val) val
-#define INT32_C(val) val##L
-#define INT64_C(val) val##i64
-
-#define UINT8_C(val)  val
-#define UINT16_C(val) val
-#define UINT32_C(val) val##UL
-#define UINT64_C(val) val##Ui64
-
-// 7.18.4.2 Macros for greatest-width integer constants
-#define INTMAX_C   INT64_C
-#define UINTMAX_C  UINT64_C
-
-#endif // __STDC_CONSTANT_MACROS ]
-
-#endif
-
-#endif // _MSC_STDINT_H_ ]

3rdparty/include/opencl/LICENSE.txt

@@ -0,0 +1,25 @@
+Copyright (c) 2008-2015 The Khronos Group Inc.
+
+Permission is hereby granted, free of charge, to any person obtaining a
+copy of this software and/or associated documentation files (the
+"Materials"), to deal in the Materials without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Materials, and to
+permit persons to whom the Materials are 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 Materials.
+
+MODIFICATIONS TO THIS FILE MAY MEAN IT NO LONGER ACCURATELY REFLECTS
+KHRONOS STANDARDS. THE UNMODIFIED, NORMATIVE VERSIONS OF KHRONOS
+SPECIFICATIONS AND HEADER INFORMATION ARE LOCATED AT
+   https://www.khronos.org/registry/
+
+THE MATERIALS ARE 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
+MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.

3rdparty/include/vulkan/vk_platform.h

@@ -0,0 +1,92 @@
+//
+// File: vk_platform.h
+//
+/*
+** Copyright (c) 2014-2017 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+
+#ifndef VK_PLATFORM_H_
+#define VK_PLATFORM_H_
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif // __cplusplus
+
+/*
+***************************************************************************************************
+*   Platform-specific directives and type declarations
+***************************************************************************************************
+*/
+
+/* Platform-specific calling convention macros.
+ *
+ * Platforms should define these so that Vulkan clients call Vulkan commands
+ * with the same calling conventions that the Vulkan implementation expects.
+ *
+ * VKAPI_ATTR - Placed before the return type in function declarations.
+ *              Useful for C++11 and GCC/Clang-style function attribute syntax.
+ * VKAPI_CALL - Placed after the return type in function declarations.
+ *              Useful for MSVC-style calling convention syntax.
+ * VKAPI_PTR  - Placed between the '(' and '*' in function pointer types.
+ *
+ * Function declaration:  VKAPI_ATTR void VKAPI_CALL vkCommand(void);
+ * Function pointer type: typedef void (VKAPI_PTR *PFN_vkCommand)(void);
+ */
+#if defined(_WIN32)
+    // On Windows, Vulkan commands use the stdcall convention
+    #define VKAPI_ATTR
+    #define VKAPI_CALL __stdcall
+    #define VKAPI_PTR  VKAPI_CALL
+#elif defined(__ANDROID__) && defined(__ARM_ARCH) && __ARM_ARCH < 7
+    #error "Vulkan isn't supported for the 'armeabi' NDK ABI"
+#elif defined(__ANDROID__) && defined(__ARM_ARCH) && __ARM_ARCH >= 7 && defined(__ARM_32BIT_STATE)
+    // On Android 32-bit ARM targets, Vulkan functions use the "hardfloat"
+    // calling convention, i.e. float parameters are passed in registers. This
+    // is true even if the rest of the application passes floats on the stack,
+    // as it does by default when compiling for the armeabi-v7a NDK ABI.
+    #define VKAPI_ATTR __attribute__((pcs("aapcs-vfp")))
+    #define VKAPI_CALL
+    #define VKAPI_PTR  VKAPI_ATTR
+#else
+    // On other platforms, use the default calling convention
+    #define VKAPI_ATTR
+    #define VKAPI_CALL
+    #define VKAPI_PTR
+#endif
+
+#include <stddef.h>
+
+#if !defined(VK_NO_STDINT_H)
+    #if defined(_MSC_VER) && (_MSC_VER < 1600)
+        typedef signed   __int8  int8_t;
+        typedef unsigned __int8  uint8_t;
+        typedef signed   __int16 int16_t;
+        typedef unsigned __int16 uint16_t;
+        typedef signed   __int32 int32_t;
+        typedef unsigned __int32 uint32_t;
+        typedef signed   __int64 int64_t;
+        typedef unsigned __int64 uint64_t;
+    #else
+        #include <stdint.h>
+    #endif
+#endif // !defined(VK_NO_STDINT_H)
+
+#ifdef __cplusplus
+} // extern "C"
+#endif // __cplusplus
+
+#endif

3rdparty/include/vulkan/vulkan.h

@@ -0,0 +1,83 @@
+#ifndef VULKAN_H_
+#define VULKAN_H_ 1
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+#include "vk_platform.h"
+#include "vulkan_core.h"
+
+#ifdef VK_USE_PLATFORM_ANDROID_KHR
+#include "vulkan_android.h"
+#endif
+
+#ifdef VK_USE_PLATFORM_FUCHSIA
+#include <zircon/types.h>
+#include "vulkan_fuchsia.h"
+#endif
+
+#ifdef VK_USE_PLATFORM_IOS_MVK
+#include "vulkan_ios.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_MACOS_MVK
+#include "vulkan_macos.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_MIR_KHR
+#include <mir_toolkit/client_types.h>
+#include "vulkan_mir.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_VI_NN
+#include "vulkan_vi.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_WAYLAND_KHR
+#include <wayland-client.h>
+#include "vulkan_wayland.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_WIN32_KHR
+#include <windows.h>
+#include "vulkan_win32.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_XCB_KHR
+#include <xcb/xcb.h>
+#include "vulkan_xcb.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_XLIB_KHR
+#include <X11/Xlib.h>
+#include "vulkan_xlib.h"
+#endif
+
+
+#ifdef VK_USE_PLATFORM_XLIB_XRANDR_EXT
+#include <X11/Xlib.h>
+#include <X11/extensions/Xrandr.h>
+#include "vulkan_xlib_xrandr.h"
+#endif
+
+#endif // VULKAN_H_

3rdparty/include/vulkan/vulkan_android.h

@@ -0,0 +1,126 @@
+#ifndef VULKAN_ANDROID_H_
+#define VULKAN_ANDROID_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_KHR_android_surface 1
+struct ANativeWindow;
+
+#define VK_KHR_ANDROID_SURFACE_SPEC_VERSION 6
+#define VK_KHR_ANDROID_SURFACE_EXTENSION_NAME "VK_KHR_android_surface"
+
+typedef VkFlags VkAndroidSurfaceCreateFlagsKHR;
+
+typedef struct VkAndroidSurfaceCreateInfoKHR {
+    VkStructureType                   sType;
+    const void*                       pNext;
+    VkAndroidSurfaceCreateFlagsKHR    flags;
+    struct ANativeWindow*             window;
+} VkAndroidSurfaceCreateInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateAndroidSurfaceKHR)(VkInstance instance, const VkAndroidSurfaceCreateInfoKHR* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateAndroidSurfaceKHR(
+    VkInstance                                  instance,
+    const VkAndroidSurfaceCreateInfoKHR*        pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+#endif
+
+#define VK_ANDROID_external_memory_android_hardware_buffer 1
+struct AHardwareBuffer;
+
+#define VK_ANDROID_EXTERNAL_MEMORY_ANDROID_HARDWARE_BUFFER_SPEC_VERSION 3
+#define VK_ANDROID_EXTERNAL_MEMORY_ANDROID_HARDWARE_BUFFER_EXTENSION_NAME "VK_ANDROID_external_memory_android_hardware_buffer"
+
+typedef struct VkAndroidHardwareBufferUsageANDROID {
+    VkStructureType    sType;
+    void*              pNext;
+    uint64_t           androidHardwareBufferUsage;
+} VkAndroidHardwareBufferUsageANDROID;
+
+typedef struct VkAndroidHardwareBufferPropertiesANDROID {
+    VkStructureType    sType;
+    void*              pNext;
+    VkDeviceSize       allocationSize;
+    uint32_t           memoryTypeBits;
+} VkAndroidHardwareBufferPropertiesANDROID;
+
+typedef struct VkAndroidHardwareBufferFormatPropertiesANDROID {
+    VkStructureType                  sType;
+    void*                            pNext;
+    VkFormat                         format;
+    uint64_t                         externalFormat;
+    VkFormatFeatureFlags             formatFeatures;
+    VkComponentMapping               samplerYcbcrConversionComponents;
+    VkSamplerYcbcrModelConversion    suggestedYcbcrModel;
+    VkSamplerYcbcrRange              suggestedYcbcrRange;
+    VkChromaLocation                 suggestedXChromaOffset;
+    VkChromaLocation                 suggestedYChromaOffset;
+} VkAndroidHardwareBufferFormatPropertiesANDROID;
+
+typedef struct VkImportAndroidHardwareBufferInfoANDROID {
+    VkStructureType            sType;
+    const void*                pNext;
+    struct AHardwareBuffer*    buffer;
+} VkImportAndroidHardwareBufferInfoANDROID;
+
+typedef struct VkMemoryGetAndroidHardwareBufferInfoANDROID {
+    VkStructureType    sType;
+    const void*        pNext;
+    VkDeviceMemory     memory;
+} VkMemoryGetAndroidHardwareBufferInfoANDROID;
+
+typedef struct VkExternalFormatANDROID {
+    VkStructureType    sType;
+    void*              pNext;
+    uint64_t           externalFormat;
+} VkExternalFormatANDROID;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkGetAndroidHardwareBufferPropertiesANDROID)(VkDevice device, const struct AHardwareBuffer* buffer, VkAndroidHardwareBufferPropertiesANDROID* pProperties);
+typedef VkResult (VKAPI_PTR *PFN_vkGetMemoryAndroidHardwareBufferANDROID)(VkDevice device, const VkMemoryGetAndroidHardwareBufferInfoANDROID* pInfo, struct AHardwareBuffer** pBuffer);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkGetAndroidHardwareBufferPropertiesANDROID(
+    VkDevice                                    device,
+    const struct AHardwareBuffer*               buffer,
+    VkAndroidHardwareBufferPropertiesANDROID*   pProperties);
+
+VKAPI_ATTR VkResult VKAPI_CALL vkGetMemoryAndroidHardwareBufferANDROID(
+    VkDevice                                    device,
+    const VkMemoryGetAndroidHardwareBufferInfoANDROID* pInfo,
+    struct AHardwareBuffer**                    pBuffer);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_fuchsia.h

@@ -0,0 +1,58 @@
+#ifndef VULKAN_FUCHSIA_H_
+#define VULKAN_FUCHSIA_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_FUCHSIA_imagepipe_surface 1
+#define VK_FUCHSIA_IMAGEPIPE_SURFACE_SPEC_VERSION 1
+#define VK_FUCHSIA_IMAGEPIPE_SURFACE_EXTENSION_NAME "VK_FUCHSIA_imagepipe_surface"
+
+typedef VkFlags VkImagePipeSurfaceCreateFlagsFUCHSIA;
+
+typedef struct VkImagePipeSurfaceCreateInfoFUCHSIA {
+    VkStructureType                         sType;
+    const void*                             pNext;
+    VkImagePipeSurfaceCreateFlagsFUCHSIA    flags;
+    zx_handle_t                             imagePipeHandle;
+} VkImagePipeSurfaceCreateInfoFUCHSIA;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateImagePipeSurfaceFUCHSIA)(VkInstance instance, const VkImagePipeSurfaceCreateInfoFUCHSIA* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateImagePipeSurfaceFUCHSIA(
+    VkInstance                                  instance,
+    const VkImagePipeSurfaceCreateInfoFUCHSIA*  pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_ios.h

@@ -0,0 +1,58 @@
+#ifndef VULKAN_IOS_H_
+#define VULKAN_IOS_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_MVK_ios_surface 1
+#define VK_MVK_IOS_SURFACE_SPEC_VERSION   2
+#define VK_MVK_IOS_SURFACE_EXTENSION_NAME "VK_MVK_ios_surface"
+
+typedef VkFlags VkIOSSurfaceCreateFlagsMVK;
+
+typedef struct VkIOSSurfaceCreateInfoMVK {
+    VkStructureType               sType;
+    const void*                   pNext;
+    VkIOSSurfaceCreateFlagsMVK    flags;
+    const void*                   pView;
+} VkIOSSurfaceCreateInfoMVK;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateIOSSurfaceMVK)(VkInstance instance, const VkIOSSurfaceCreateInfoMVK* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateIOSSurfaceMVK(
+    VkInstance                                  instance,
+    const VkIOSSurfaceCreateInfoMVK*            pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_macos.h

@@ -0,0 +1,58 @@
+#ifndef VULKAN_MACOS_H_
+#define VULKAN_MACOS_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_MVK_macos_surface 1
+#define VK_MVK_MACOS_SURFACE_SPEC_VERSION 2
+#define VK_MVK_MACOS_SURFACE_EXTENSION_NAME "VK_MVK_macos_surface"
+
+typedef VkFlags VkMacOSSurfaceCreateFlagsMVK;
+
+typedef struct VkMacOSSurfaceCreateInfoMVK {
+    VkStructureType                 sType;
+    const void*                     pNext;
+    VkMacOSSurfaceCreateFlagsMVK    flags;
+    const void*                     pView;
+} VkMacOSSurfaceCreateInfoMVK;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateMacOSSurfaceMVK)(VkInstance instance, const VkMacOSSurfaceCreateInfoMVK* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateMacOSSurfaceMVK(
+    VkInstance                                  instance,
+    const VkMacOSSurfaceCreateInfoMVK*          pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_mir.h

@@ -0,0 +1,65 @@
+#ifndef VULKAN_MIR_H_
+#define VULKAN_MIR_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_KHR_mir_surface 1
+#define VK_KHR_MIR_SURFACE_SPEC_VERSION   4
+#define VK_KHR_MIR_SURFACE_EXTENSION_NAME "VK_KHR_mir_surface"
+
+typedef VkFlags VkMirSurfaceCreateFlagsKHR;
+
+typedef struct VkMirSurfaceCreateInfoKHR {
+    VkStructureType               sType;
+    const void*                   pNext;
+    VkMirSurfaceCreateFlagsKHR    flags;
+    MirConnection*                connection;
+    MirSurface*                   mirSurface;
+} VkMirSurfaceCreateInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateMirSurfaceKHR)(VkInstance instance, const VkMirSurfaceCreateInfoKHR* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+typedef VkBool32 (VKAPI_PTR *PFN_vkGetPhysicalDeviceMirPresentationSupportKHR)(VkPhysicalDevice physicalDevice, uint32_t queueFamilyIndex, MirConnection* connection);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateMirSurfaceKHR(
+    VkInstance                                  instance,
+    const VkMirSurfaceCreateInfoKHR*            pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+
+VKAPI_ATTR VkBool32 VKAPI_CALL vkGetPhysicalDeviceMirPresentationSupportKHR(
+    VkPhysicalDevice                            physicalDevice,
+    uint32_t                                    queueFamilyIndex,
+    MirConnection*                              connection);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_vi.h

@@ -0,0 +1,58 @@
+#ifndef VULKAN_VI_H_
+#define VULKAN_VI_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_NN_vi_surface 1
+#define VK_NN_VI_SURFACE_SPEC_VERSION     1
+#define VK_NN_VI_SURFACE_EXTENSION_NAME   "VK_NN_vi_surface"
+
+typedef VkFlags VkViSurfaceCreateFlagsNN;
+
+typedef struct VkViSurfaceCreateInfoNN {
+    VkStructureType             sType;
+    const void*                 pNext;
+    VkViSurfaceCreateFlagsNN    flags;
+    void*                       window;
+} VkViSurfaceCreateInfoNN;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateViSurfaceNN)(VkInstance instance, const VkViSurfaceCreateInfoNN* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateViSurfaceNN(
+    VkInstance                                  instance,
+    const VkViSurfaceCreateInfoNN*              pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_wayland.h

@@ -0,0 +1,65 @@
+#ifndef VULKAN_WAYLAND_H_
+#define VULKAN_WAYLAND_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_KHR_wayland_surface 1
+#define VK_KHR_WAYLAND_SURFACE_SPEC_VERSION 6
+#define VK_KHR_WAYLAND_SURFACE_EXTENSION_NAME "VK_KHR_wayland_surface"
+
+typedef VkFlags VkWaylandSurfaceCreateFlagsKHR;
+
+typedef struct VkWaylandSurfaceCreateInfoKHR {
+    VkStructureType                   sType;
+    const void*                       pNext;
+    VkWaylandSurfaceCreateFlagsKHR    flags;
+    struct wl_display*                display;
+    struct wl_surface*                surface;
+} VkWaylandSurfaceCreateInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateWaylandSurfaceKHR)(VkInstance instance, const VkWaylandSurfaceCreateInfoKHR* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+typedef VkBool32 (VKAPI_PTR *PFN_vkGetPhysicalDeviceWaylandPresentationSupportKHR)(VkPhysicalDevice physicalDevice, uint32_t queueFamilyIndex, struct wl_display* display);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateWaylandSurfaceKHR(
+    VkInstance                                  instance,
+    const VkWaylandSurfaceCreateInfoKHR*        pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+
+VKAPI_ATTR VkBool32 VKAPI_CALL vkGetPhysicalDeviceWaylandPresentationSupportKHR(
+    VkPhysicalDevice                            physicalDevice,
+    uint32_t                                    queueFamilyIndex,
+    struct wl_display*                          display);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_win32.h

@@ -0,0 +1,276 @@
+#ifndef VULKAN_WIN32_H_
+#define VULKAN_WIN32_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_KHR_win32_surface 1
+#define VK_KHR_WIN32_SURFACE_SPEC_VERSION 6
+#define VK_KHR_WIN32_SURFACE_EXTENSION_NAME "VK_KHR_win32_surface"
+
+typedef VkFlags VkWin32SurfaceCreateFlagsKHR;
+
+typedef struct VkWin32SurfaceCreateInfoKHR {
+    VkStructureType                 sType;
+    const void*                     pNext;
+    VkWin32SurfaceCreateFlagsKHR    flags;
+    HINSTANCE                       hinstance;
+    HWND                            hwnd;
+} VkWin32SurfaceCreateInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateWin32SurfaceKHR)(VkInstance instance, const VkWin32SurfaceCreateInfoKHR* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+typedef VkBool32 (VKAPI_PTR *PFN_vkGetPhysicalDeviceWin32PresentationSupportKHR)(VkPhysicalDevice physicalDevice, uint32_t queueFamilyIndex);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateWin32SurfaceKHR(
+    VkInstance                                  instance,
+    const VkWin32SurfaceCreateInfoKHR*          pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+
+VKAPI_ATTR VkBool32 VKAPI_CALL vkGetPhysicalDeviceWin32PresentationSupportKHR(
+    VkPhysicalDevice                            physicalDevice,
+    uint32_t                                    queueFamilyIndex);
+#endif
+
+#define VK_KHR_external_memory_win32 1
+#define VK_KHR_EXTERNAL_MEMORY_WIN32_SPEC_VERSION 1
+#define VK_KHR_EXTERNAL_MEMORY_WIN32_EXTENSION_NAME "VK_KHR_external_memory_win32"
+
+typedef struct VkImportMemoryWin32HandleInfoKHR {
+    VkStructureType                       sType;
+    const void*                           pNext;
+    VkExternalMemoryHandleTypeFlagBits    handleType;
+    HANDLE                                handle;
+    LPCWSTR                               name;
+} VkImportMemoryWin32HandleInfoKHR;
+
+typedef struct VkExportMemoryWin32HandleInfoKHR {
+    VkStructureType               sType;
+    const void*                   pNext;
+    const SECURITY_ATTRIBUTES*    pAttributes;
+    DWORD                         dwAccess;
+    LPCWSTR                       name;
+} VkExportMemoryWin32HandleInfoKHR;
+
+typedef struct VkMemoryWin32HandlePropertiesKHR {
+    VkStructureType    sType;
+    void*              pNext;
+    uint32_t           memoryTypeBits;
+} VkMemoryWin32HandlePropertiesKHR;
+
+typedef struct VkMemoryGetWin32HandleInfoKHR {
+    VkStructureType                       sType;
+    const void*                           pNext;
+    VkDeviceMemory                        memory;
+    VkExternalMemoryHandleTypeFlagBits    handleType;
+} VkMemoryGetWin32HandleInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkGetMemoryWin32HandleKHR)(VkDevice device, const VkMemoryGetWin32HandleInfoKHR* pGetWin32HandleInfo, HANDLE* pHandle);
+typedef VkResult (VKAPI_PTR *PFN_vkGetMemoryWin32HandlePropertiesKHR)(VkDevice device, VkExternalMemoryHandleTypeFlagBits handleType, HANDLE handle, VkMemoryWin32HandlePropertiesKHR* pMemoryWin32HandleProperties);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkGetMemoryWin32HandleKHR(
+    VkDevice                                    device,
+    const VkMemoryGetWin32HandleInfoKHR*        pGetWin32HandleInfo,
+    HANDLE*                                     pHandle);
+
+VKAPI_ATTR VkResult VKAPI_CALL vkGetMemoryWin32HandlePropertiesKHR(
+    VkDevice                                    device,
+    VkExternalMemoryHandleTypeFlagBits          handleType,
+    HANDLE                                      handle,
+    VkMemoryWin32HandlePropertiesKHR*           pMemoryWin32HandleProperties);
+#endif
+
+#define VK_KHR_win32_keyed_mutex 1
+#define VK_KHR_WIN32_KEYED_MUTEX_SPEC_VERSION 1
+#define VK_KHR_WIN32_KEYED_MUTEX_EXTENSION_NAME "VK_KHR_win32_keyed_mutex"
+
+typedef struct VkWin32KeyedMutexAcquireReleaseInfoKHR {
+    VkStructureType          sType;
+    const void*              pNext;
+    uint32_t                 acquireCount;
+    const VkDeviceMemory*    pAcquireSyncs;
+    const uint64_t*          pAcquireKeys;
+    const uint32_t*          pAcquireTimeouts;
+    uint32_t                 releaseCount;
+    const VkDeviceMemory*    pReleaseSyncs;
+    const uint64_t*          pReleaseKeys;
+} VkWin32KeyedMutexAcquireReleaseInfoKHR;
+
+
+
+#define VK_KHR_external_semaphore_win32 1
+#define VK_KHR_EXTERNAL_SEMAPHORE_WIN32_SPEC_VERSION 1
+#define VK_KHR_EXTERNAL_SEMAPHORE_WIN32_EXTENSION_NAME "VK_KHR_external_semaphore_win32"
+
+typedef struct VkImportSemaphoreWin32HandleInfoKHR {
+    VkStructureType                          sType;
+    const void*                              pNext;
+    VkSemaphore                              semaphore;
+    VkSemaphoreImportFlags                   flags;
+    VkExternalSemaphoreHandleTypeFlagBits    handleType;
+    HANDLE                                   handle;
+    LPCWSTR                                  name;
+} VkImportSemaphoreWin32HandleInfoKHR;
+
+typedef struct VkExportSemaphoreWin32HandleInfoKHR {
+    VkStructureType               sType;
+    const void*                   pNext;
+    const SECURITY_ATTRIBUTES*    pAttributes;
+    DWORD                         dwAccess;
+    LPCWSTR                       name;
+} VkExportSemaphoreWin32HandleInfoKHR;
+
+typedef struct VkD3D12FenceSubmitInfoKHR {
+    VkStructureType    sType;
+    const void*        pNext;
+    uint32_t           waitSemaphoreValuesCount;
+    const uint64_t*    pWaitSemaphoreValues;
+    uint32_t           signalSemaphoreValuesCount;
+    const uint64_t*    pSignalSemaphoreValues;
+} VkD3D12FenceSubmitInfoKHR;
+
+typedef struct VkSemaphoreGetWin32HandleInfoKHR {
+    VkStructureType                          sType;
+    const void*                              pNext;
+    VkSemaphore                              semaphore;
+    VkExternalSemaphoreHandleTypeFlagBits    handleType;
+} VkSemaphoreGetWin32HandleInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkImportSemaphoreWin32HandleKHR)(VkDevice device, const VkImportSemaphoreWin32HandleInfoKHR* pImportSemaphoreWin32HandleInfo);
+typedef VkResult (VKAPI_PTR *PFN_vkGetSemaphoreWin32HandleKHR)(VkDevice device, const VkSemaphoreGetWin32HandleInfoKHR* pGetWin32HandleInfo, HANDLE* pHandle);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkImportSemaphoreWin32HandleKHR(
+    VkDevice                                    device,
+    const VkImportSemaphoreWin32HandleInfoKHR*  pImportSemaphoreWin32HandleInfo);
+
+VKAPI_ATTR VkResult VKAPI_CALL vkGetSemaphoreWin32HandleKHR(
+    VkDevice                                    device,
+    const VkSemaphoreGetWin32HandleInfoKHR*     pGetWin32HandleInfo,
+    HANDLE*                                     pHandle);
+#endif
+
+#define VK_KHR_external_fence_win32 1
+#define VK_KHR_EXTERNAL_FENCE_WIN32_SPEC_VERSION 1
+#define VK_KHR_EXTERNAL_FENCE_WIN32_EXTENSION_NAME "VK_KHR_external_fence_win32"
+
+typedef struct VkImportFenceWin32HandleInfoKHR {
+    VkStructureType                      sType;
+    const void*                          pNext;
+    VkFence                              fence;
+    VkFenceImportFlags                   flags;
+    VkExternalFenceHandleTypeFlagBits    handleType;
+    HANDLE                               handle;
+    LPCWSTR                              name;
+} VkImportFenceWin32HandleInfoKHR;
+
+typedef struct VkExportFenceWin32HandleInfoKHR {
+    VkStructureType               sType;
+    const void*                   pNext;
+    const SECURITY_ATTRIBUTES*    pAttributes;
+    DWORD                         dwAccess;
+    LPCWSTR                       name;
+} VkExportFenceWin32HandleInfoKHR;
+
+typedef struct VkFenceGetWin32HandleInfoKHR {
+    VkStructureType                      sType;
+    const void*                          pNext;
+    VkFence                              fence;
+    VkExternalFenceHandleTypeFlagBits    handleType;
+} VkFenceGetWin32HandleInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkImportFenceWin32HandleKHR)(VkDevice device, const VkImportFenceWin32HandleInfoKHR* pImportFenceWin32HandleInfo);
+typedef VkResult (VKAPI_PTR *PFN_vkGetFenceWin32HandleKHR)(VkDevice device, const VkFenceGetWin32HandleInfoKHR* pGetWin32HandleInfo, HANDLE* pHandle);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkImportFenceWin32HandleKHR(
+    VkDevice                                    device,
+    const VkImportFenceWin32HandleInfoKHR*      pImportFenceWin32HandleInfo);
+
+VKAPI_ATTR VkResult VKAPI_CALL vkGetFenceWin32HandleKHR(
+    VkDevice                                    device,
+    const VkFenceGetWin32HandleInfoKHR*         pGetWin32HandleInfo,
+    HANDLE*                                     pHandle);
+#endif
+
+#define VK_NV_external_memory_win32 1
+#define VK_NV_EXTERNAL_MEMORY_WIN32_SPEC_VERSION 1
+#define VK_NV_EXTERNAL_MEMORY_WIN32_EXTENSION_NAME "VK_NV_external_memory_win32"
+
+typedef struct VkImportMemoryWin32HandleInfoNV {
+    VkStructureType                      sType;
+    const void*                          pNext;
+    VkExternalMemoryHandleTypeFlagsNV    handleType;
+    HANDLE                               handle;
+} VkImportMemoryWin32HandleInfoNV;
+
+typedef struct VkExportMemoryWin32HandleInfoNV {
+    VkStructureType               sType;
+    const void*                   pNext;
+    const SECURITY_ATTRIBUTES*    pAttributes;
+    DWORD                         dwAccess;
+} VkExportMemoryWin32HandleInfoNV;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkGetMemoryWin32HandleNV)(VkDevice device, VkDeviceMemory memory, VkExternalMemoryHandleTypeFlagsNV handleType, HANDLE* pHandle);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkGetMemoryWin32HandleNV(
+    VkDevice                                    device,
+    VkDeviceMemory                              memory,
+    VkExternalMemoryHandleTypeFlagsNV           handleType,
+    HANDLE*                                     pHandle);
+#endif
+
+#define VK_NV_win32_keyed_mutex 1
+#define VK_NV_WIN32_KEYED_MUTEX_SPEC_VERSION 1
+#define VK_NV_WIN32_KEYED_MUTEX_EXTENSION_NAME "VK_NV_win32_keyed_mutex"
+
+typedef struct VkWin32KeyedMutexAcquireReleaseInfoNV {
+    VkStructureType          sType;
+    const void*              pNext;
+    uint32_t                 acquireCount;
+    const VkDeviceMemory*    pAcquireSyncs;
+    const uint64_t*          pAcquireKeys;
+    const uint32_t*          pAcquireTimeoutMilliseconds;
+    uint32_t                 releaseCount;
+    const VkDeviceMemory*    pReleaseSyncs;
+    const uint64_t*          pReleaseKeys;
+} VkWin32KeyedMutexAcquireReleaseInfoNV;
+
+
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_xcb.h

@@ -0,0 +1,66 @@
+#ifndef VULKAN_XCB_H_
+#define VULKAN_XCB_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_KHR_xcb_surface 1
+#define VK_KHR_XCB_SURFACE_SPEC_VERSION   6
+#define VK_KHR_XCB_SURFACE_EXTENSION_NAME "VK_KHR_xcb_surface"
+
+typedef VkFlags VkXcbSurfaceCreateFlagsKHR;
+
+typedef struct VkXcbSurfaceCreateInfoKHR {
+    VkStructureType               sType;
+    const void*                   pNext;
+    VkXcbSurfaceCreateFlagsKHR    flags;
+    xcb_connection_t*             connection;
+    xcb_window_t                  window;
+} VkXcbSurfaceCreateInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateXcbSurfaceKHR)(VkInstance instance, const VkXcbSurfaceCreateInfoKHR* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+typedef VkBool32 (VKAPI_PTR *PFN_vkGetPhysicalDeviceXcbPresentationSupportKHR)(VkPhysicalDevice physicalDevice, uint32_t queueFamilyIndex, xcb_connection_t* connection, xcb_visualid_t visual_id);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateXcbSurfaceKHR(
+    VkInstance                                  instance,
+    const VkXcbSurfaceCreateInfoKHR*            pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+
+VKAPI_ATTR VkBool32 VKAPI_CALL vkGetPhysicalDeviceXcbPresentationSupportKHR(
+    VkPhysicalDevice                            physicalDevice,
+    uint32_t                                    queueFamilyIndex,
+    xcb_connection_t*                           connection,
+    xcb_visualid_t                              visual_id);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_xlib.h

@@ -0,0 +1,66 @@
+#ifndef VULKAN_XLIB_H_
+#define VULKAN_XLIB_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_KHR_xlib_surface 1
+#define VK_KHR_XLIB_SURFACE_SPEC_VERSION  6
+#define VK_KHR_XLIB_SURFACE_EXTENSION_NAME "VK_KHR_xlib_surface"
+
+typedef VkFlags VkXlibSurfaceCreateFlagsKHR;
+
+typedef struct VkXlibSurfaceCreateInfoKHR {
+    VkStructureType                sType;
+    const void*                    pNext;
+    VkXlibSurfaceCreateFlagsKHR    flags;
+    Display*                       dpy;
+    Window                         window;
+} VkXlibSurfaceCreateInfoKHR;
+
+
+typedef VkResult (VKAPI_PTR *PFN_vkCreateXlibSurfaceKHR)(VkInstance instance, const VkXlibSurfaceCreateInfoKHR* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkSurfaceKHR* pSurface);
+typedef VkBool32 (VKAPI_PTR *PFN_vkGetPhysicalDeviceXlibPresentationSupportKHR)(VkPhysicalDevice physicalDevice, uint32_t queueFamilyIndex, Display* dpy, VisualID visualID);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkCreateXlibSurfaceKHR(
+    VkInstance                                  instance,
+    const VkXlibSurfaceCreateInfoKHR*           pCreateInfo,
+    const VkAllocationCallbacks*                pAllocator,
+    VkSurfaceKHR*                               pSurface);
+
+VKAPI_ATTR VkBool32 VKAPI_CALL vkGetPhysicalDeviceXlibPresentationSupportKHR(
+    VkPhysicalDevice                            physicalDevice,
+    uint32_t                                    queueFamilyIndex,
+    Display*                                    dpy,
+    VisualID                                    visualID);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/include/vulkan/vulkan_xlib_xrandr.h

@@ -0,0 +1,54 @@
+#ifndef VULKAN_XLIB_XRANDR_H_
+#define VULKAN_XLIB_XRANDR_H_ 1
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/*
+** Copyright (c) 2015-2018 The Khronos Group Inc.
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+**     http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+/*
+** This header is generated from the Khronos Vulkan XML API Registry.
+**
+*/
+
+
+#define VK_EXT_acquire_xlib_display 1
+#define VK_EXT_ACQUIRE_XLIB_DISPLAY_SPEC_VERSION 1
+#define VK_EXT_ACQUIRE_XLIB_DISPLAY_EXTENSION_NAME "VK_EXT_acquire_xlib_display"
+
+typedef VkResult (VKAPI_PTR *PFN_vkAcquireXlibDisplayEXT)(VkPhysicalDevice physicalDevice, Display* dpy, VkDisplayKHR display);
+typedef VkResult (VKAPI_PTR *PFN_vkGetRandROutputDisplayEXT)(VkPhysicalDevice physicalDevice, Display* dpy, RROutput rrOutput, VkDisplayKHR* pDisplay);
+
+#ifndef VK_NO_PROTOTYPES
+VKAPI_ATTR VkResult VKAPI_CALL vkAcquireXlibDisplayEXT(
+    VkPhysicalDevice                            physicalDevice,
+    Display*                                    dpy,
+    VkDisplayKHR                                display);
+
+VKAPI_ATTR VkResult VKAPI_CALL vkGetRandROutputDisplayEXT(
+    VkPhysicalDevice                            physicalDevice,
+    Display*                                    dpy,
+    RROutput                                    rrOutput,
+    VkDisplayKHR*                               pDisplay);
+#endif
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

3rdparty/ippicv/CMakeLists.txt

@@ -6,7 +6,7 @@ project(${IPP_IW_LIBRARY})
 
 ocv_include_directories(${IPP_INCLUDE_DIRS} ${IPP_IW_PATH}/include)
 add_definitions(-DIW_BUILD)
-if(HAVE_IPP_ICV_ONLY)
+if(HAVE_IPP_ICV)
   add_definitions(-DICV_BASE)
 endif()
 
@@ -20,8 +20,11 @@ file(GLOB lib_hdrs ${IPP_IW_PATH}/include/*.h ${IPP_IW_PATH}/include/iw/*.h ${IP
 add_library(${IPP_IW_LIBRARY} STATIC ${lib_srcs} ${lib_hdrs})
 
 if(UNIX)
-  if(CMAKE_COMPILER_IS_GNUCXX OR CV_ICC)
-    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC -Wno-unused-function")
+  if(CV_GCC OR CV_CLANG OR CV_ICC)
+    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Wno-unused-function -Wno-missing-braces -Wno-missing-field-initializers")
+  endif()
+  if(CV_CLANG)
+    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Wno-self-assign")
   endif()
 endif()
 

3rdparty/ippicv/ippicv.cmake

@@ -2,37 +2,37 @@ function(download_ippicv root_var)
   set(${root_var} "" PARENT_SCOPE)
 
   # Commit SHA in the opencv_3rdparty repo
-  set(IPPICV_COMMIT "a62e20676a60ee0ad6581e217fe7e4bada3b95db")
+  set(IPPICV_COMMIT "32e315a5b106a7b89dbed51c28f8120a48b368b4")
   # Define actual ICV versions
   if(APPLE)
     set(OPENCV_ICV_PLATFORM "macosx")
     set(OPENCV_ICV_PACKAGE_SUBDIR "ippicv_mac")
     if(X86_64)
-      set(OPENCV_ICV_NAME "ippicv_2017u2_mac_intel64_20170418.tgz")
-      set(OPENCV_ICV_HASH "0c25953c99dbb499ff502485a9356d8d")
+      set(OPENCV_ICV_NAME "ippicv_2019_mac_intel64_general_20180723.tgz")
+      set(OPENCV_ICV_HASH "fe6b2bb75ae0e3f19ad3ae1a31dfa4a2")
     else()
-      set(OPENCV_ICV_NAME "ippicv_2017u2_mac_ia32_20170418.tgz")
-      set(OPENCV_ICV_HASH "5f225948f3f64067c681293c098d50d8")
+      set(OPENCV_ICV_NAME "ippicv_2019_mac_ia32_general_20180723.tgz")
+      set(OPENCV_ICV_HASH "b5dfa78c87eb75c64470cbe5ec876f4f")
     endif()
   elseif((UNIX AND NOT ANDROID) OR (UNIX AND ANDROID_ABI MATCHES "x86"))
     set(OPENCV_ICV_PLATFORM "linux")
     set(OPENCV_ICV_PACKAGE_SUBDIR "ippicv_lnx")
     if(X86_64)
-      set(OPENCV_ICV_NAME "ippicv_2017u2_lnx_intel64_20170418.tgz")
-      set(OPENCV_ICV_HASH "87cbdeb627415d8e4bc811156289fa3a")
+      set(OPENCV_ICV_NAME "ippicv_2019_lnx_intel64_general_20180723.tgz")
+      set(OPENCV_ICV_HASH "c0bd78adb4156bbf552c1dfe90599607")
     else()
-      set(OPENCV_ICV_NAME "ippicv_2017u2_lnx_ia32_20170418.tgz")
-      set(OPENCV_ICV_HASH "f2cece00d802d4dea86df52ed095257e")
+      set(OPENCV_ICV_NAME "ippicv_2019_lnx_ia32_general_20180723.tgz")
+      set(OPENCV_ICV_HASH "4f38432c30bfd6423164b7a24bbc98a0")
     endif()
   elseif(WIN32 AND NOT ARM)
     set(OPENCV_ICV_PLATFORM "windows")
     set(OPENCV_ICV_PACKAGE_SUBDIR "ippicv_win")
     if(X86_64)
-      set(OPENCV_ICV_NAME "ippicv_2017u2_win_intel64_20170418.zip")
-      set(OPENCV_ICV_HASH "75060a0c662c0800f48995b7e9b085f6")
+      set(OPENCV_ICV_NAME "ippicv_2019_win_intel64_20180723_general.zip")
+      set(OPENCV_ICV_HASH "1d222685246896fe089f88b8858e4b2f")
     else()
-      set(OPENCV_ICV_NAME "ippicv_2017u2_win_ia32_20170418.zip")
-      set(OPENCV_ICV_HASH "60fcf3ccd9a2ebc9e432ffb5cb91638b")
+      set(OPENCV_ICV_NAME "ippicv_2019_win_ia32_20180723_general.zip")
+      set(OPENCV_ICV_HASH "0157251a2eb9cd63a3ebc7eed0f3e59e")
     endif()
   else()
     return()

3rdparty/ittnotify/CMakeLists.txt

@@ -15,6 +15,8 @@ if(NOT WIN32)
   endif()
 endif()
 
+ocv_warnings_disable(CMAKE_C_FLAGS -Wimplicit-fallthrough)
+
 ocv_include_directories("${CMAKE_CURRENT_SOURCE_DIR}/include")
 set(ITT_INCLUDE_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}/include")
 
@@ -43,12 +45,6 @@ if(NOT WIN32)
   endif()
 endif()
 
-if(UNIX)
-  if(CMAKE_COMPILER_IS_GNUCXX OR CV_ICC)
-     set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC")
-  endif()
-endif()
-
 set_target_properties(${ITT_LIBRARY} PROPERTIES
         OUTPUT_NAME ${ITT_LIBRARY}
         DEBUG_POSTFIX "${OPENCV_DEBUG_POSTFIX}"
@@ -66,3 +62,5 @@ endif()
 if(NOT BUILD_SHARED_LIBS)
   ocv_install_target(${ITT_LIBRARY} EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev)
 endif()
+
+ocv_install_3rdparty_licenses(ittnotify src/ittnotify/LICENSE.BSD src/ittnotify/LICENSE.GPL)

3rdparty/ittnotify/src/ittnotify/ittnotify_config.h

@@ -335,7 +335,7 @@ ITT_INLINE long __itt_interlocked_increment(volatile long* ptr)
 #ifdef SDL_STRNCPY_S
 #define __itt_fstrcpyn(s1, b, s2, l) SDL_STRNCPY_S(s1, b, s2, l)
 #else
-#define __itt_fstrcpyn(s1, b, s2, l) strncpy(s1, s2, l)
+#define __itt_fstrcpyn(s1, b, s2, l) strncpy(s1, s2, b)
 #endif /* SDL_STRNCPY_S */
 
 #define __itt_fstrdup(s)          strdup(s)

3rdparty/libjasper/CMakeLists.txt

@@ -25,16 +25,13 @@ endif(WIN32 AND NOT MINGW)
 
 ocv_warnings_disable(CMAKE_C_FLAGS -Wno-implicit-function-declaration -Wno-uninitialized -Wmissing-prototypes
                                    -Wno-unused-but-set-parameter -Wmissing-declarations -Wunused -Wshadow
-                                   -Wsign-compare -Wstrict-overflow)
-ocv_warnings_disable(CMAKE_C_FLAGS -Wunused-parameter) # clang
+                                   -Wsign-compare -Wstrict-overflow -Wpointer-compare
+                                   -Wabsolute-value  # clang on Linux
+                                   -Wimplicit-fallthrough
+)
+ocv_warnings_disable(CMAKE_C_FLAGS -Wunused-parameter -Wstrict-prototypes) # clang
 ocv_warnings_disable(CMAKE_C_FLAGS /wd4013 /wd4018 /wd4101 /wd4244 /wd4267 /wd4715) # vs2005
 
-if(UNIX)
-  if(CMAKE_COMPILER_IS_GNUCXX OR CV_ICC)
-    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fPIC")
-  endif()
-endif()
-
 set_target_properties(${JASPER_LIBRARY}
   PROPERTIES
   OUTPUT_NAME ${JASPER_LIBRARY}
@@ -51,3 +48,5 @@ endif()
 if(NOT BUILD_SHARED_LIBS)
   ocv_install_target(${JASPER_LIBRARY} EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev)
 endif()
+
+ocv_install_3rdparty_licenses(jasper LICENSE README copyright)

3rdparty/libjpeg-turbo/CMakeLists.txt

@@ -0,0 +1,124 @@
+project(${JPEG_LIBRARY} C)
+
+ocv_warnings_disable(CMAKE_C_FLAGS -Wunused-parameter -Wsign-compare -Wshorten-64-to-32 -Wimplicit-fallthrough)
+
+set(VERSION_MAJOR 2)
+set(VERSION_MINOR 0)
+set(VERSION_REVISION 2)
+set(VERSION ${VERSION_MAJOR}.${VERSION_MINOR}.${VERSION_REVISION})
+set(LIBJPEG_TURBO_VERSION_NUMBER 2000002)
+
+string(TIMESTAMP BUILD "opencv-${OPENCV_VERSION}-libjpeg-turbo")
+if(CMAKE_BUILD_TYPE STREQUAL "Debug")
+  set(BUILD "${BUILD}-debug")
+endif()
+
+message(STATUS "libjpeg-turbo: VERSION = ${VERSION}, BUILD = ${BUILD}")
+
+option(WITH_ARITH_ENC "Include arithmetic encoding support when emulating the libjpeg v6b API/ABI" TRUE)
+option(WITH_ARITH_DEC "Include arithmetic decoding support when emulating the libjpeg v6b API/ABI" TRUE)
+
+include(CheckCSourceCompiles)
+include(CheckIncludeFiles)
+include(CheckTypeSize)
+
+check_type_size("size_t" SIZEOF_SIZE_T)
+check_type_size("unsigned long" SIZEOF_UNSIGNED_LONG)
+
+if(SIZEOF_SIZE_T EQUAL SIZEOF_UNSIGNED_LONG)
+  check_c_source_compiles("int main(int argc, char **argv) { unsigned long a = argc;  return __builtin_ctzl(a); }"
+    HAVE_BUILTIN_CTZL)
+endif()
+if(MSVC)
+  check_include_files("intrin.h" HAVE_INTRIN_H)
+endif()
+
+if(UNIX)
+  # Check for headers
+  check_include_files(locale.h HAVE_LOCALE_H)
+  check_include_files(stddef.h HAVE_STDDEF_H)
+  check_include_files(stdlib.h HAVE_STDLIB_H)
+  check_include_files(sys/types.h NEED_SYS_TYPES_H)
+
+  # Other predefines
+  # undef NEED_BSD_STRINGS
+  ocv_update(HAVE_UNSIGNED_CHAR 1)
+  ocv_update(HAVE_UNSIGNED_SHORT 1)
+  # undef INCOMPLETE_TYPES_BROKEN
+  ocv_update(RIGHT_SHIFT_IS_UNSIGNED 0)
+  ocv_update(__CHAR_UNSIGNED__ 0)
+endif()
+
+
+set(BITS_IN_JSAMPLE 8)
+
+if(WITH_ARITH_ENC)
+  set(C_ARITH_CODING_SUPPORTED 1)
+endif()
+
+if(WITH_ARITH_DEC)
+  set(D_ARITH_CODING_SUPPORTED 1)
+endif()
+
+set(JPEG_LIB_VERSION 62)
+
+# OpenCV
+set(JPEG_LIB_VERSION "${VERSION}-${JPEG_LIB_VERSION}" PARENT_SCOPE)
+
+if(MSVC)
+  add_definitions(-W3 -wd4996 -wd4018)
+endif()
+
+if(WIN32)
+  configure_file(jconfig.h.win.in jconfig.h)
+else()
+  configure_file(jconfig.h.in jconfig.h)
+endif()
+configure_file(jconfigint.h.in jconfigint.h)
+
+include_directories(${CMAKE_CURRENT_BINARY_DIR} ${CMAKE_CURRENT_SOURCE_DIR}/src)
+
+set(JPEG_SOURCES jcapimin.c jcapistd.c jccoefct.c jccolor.c jcdctmgr.c jchuff.c
+  jcinit.c jcmainct.c jcmarker.c jcmaster.c jcomapi.c jcparam.c jcphuff.c
+  jcprepct.c jcsample.c jctrans.c jdapimin.c jdapistd.c jdatadst.c jdatasrc.c
+  jdcoefct.c jdcolor.c jddctmgr.c jdhuff.c jdinput.c jdmainct.c jdmarker.c
+  jdmaster.c jdmerge.c jdphuff.c jdpostct.c jdsample.c jdtrans.c jerror.c
+  jfdctflt.c jfdctfst.c jfdctint.c jidctflt.c jidctfst.c jidctint.c jidctred.c
+  jquant1.c jquant2.c jutils.c jmemmgr.c jmemnobs.c)
+
+if(WITH_ARITH_ENC OR WITH_ARITH_DEC)
+  set(JPEG_SOURCES ${JPEG_SOURCES} jaricom.c)
+endif()
+
+if(WITH_ARITH_ENC)
+  set(JPEG_SOURCES ${JPEG_SOURCES} jcarith.c)
+endif()
+
+if(WITH_ARITH_DEC)
+  set(JPEG_SOURCES ${JPEG_SOURCES} jdarith.c)
+endif()
+
+# No SIMD
+set(JPEG_SOURCES ${JPEG_SOURCES} jsimd_none.c)
+
+ocv_list_add_prefix(JPEG_SOURCES src/)
+
+add_library(${JPEG_LIBRARY} STATIC ${JPEG_SOURCES} ${SIMD_OBJS})
+
+set_target_properties(${JPEG_LIBRARY}
+  PROPERTIES OUTPUT_NAME ${JPEG_LIBRARY}
+  DEBUG_POSTFIX "${OPENCV_DEBUG_POSTFIX}"
+  COMPILE_PDB_NAME ${JPEG_LIBRARY}
+  COMPILE_PDB_NAME_DEBUG "${JPEG_LIBRARY}${OPENCV_DEBUG_POSTFIX}"
+  ARCHIVE_OUTPUT_DIRECTORY ${3P_LIBRARY_OUTPUT_PATH}
+  )
+
+if(ENABLE_SOLUTION_FOLDERS)
+  set_target_properties(${JPEG_LIBRARY} PROPERTIES FOLDER "3rdparty")
+endif()
+
+if(NOT BUILD_SHARED_LIBS)
+  ocv_install_target(${JPEG_LIBRARY} EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev)
+endif()
+
+ocv_install_3rdparty_licenses(libjpeg-turbo README.md LICENSE.md README.ijg)

3rdparty/libjpeg-turbo/LICENSE.md

@@ -0,0 +1,132 @@
+libjpeg-turbo Licenses
+======================
+
+libjpeg-turbo is covered by three compatible BSD-style open source licenses:
+
+- The IJG (Independent JPEG Group) License, which is listed in
+  [README.ijg](README.ijg)
+
+  This license applies to the libjpeg API library and associated programs
+  (any code inherited from libjpeg, and any modifications to that code.)
+
+- The Modified (3-clause) BSD License, which is listed below
+
+  This license covers the TurboJPEG API library and associated programs, as
+  well as the build system.
+
+- The [zlib License](https://opensource.org/licenses/Zlib)
+
+  This license is a subset of the other two, and it covers the libjpeg-turbo
+  SIMD extensions.
+
+
+Complying with the libjpeg-turbo Licenses
+=========================================
+
+This section provides a roll-up of the libjpeg-turbo licensing terms, to the
+best of our understanding.
+
+1.  If you are distributing a modified version of the libjpeg-turbo source,
+    then:
+
+    1.  You cannot alter or remove any existing copyright or license notices
+        from the source.
+
+        **Origin**
+        - Clause 1 of the IJG License
+        - Clause 1 of the Modified BSD License
+        - Clauses 1 and 3 of the zlib License
+
+    2.  You must add your own copyright notice to the header of each source
+        file you modified, so others can tell that you modified that file (if
+        there is not an existing copyright header in that file, then you can
+        simply add a notice stating that you modified the file.)
+
+        **Origin**
+        - Clause 1 of the IJG License
+        - Clause 2 of the zlib License
+
+    3.  You must include the IJG README file, and you must not alter any of the
+        copyright or license text in that file.
+
+        **Origin**
+        - Clause 1 of the IJG License
+
+2.  If you are distributing only libjpeg-turbo binaries without the source, or
+    if you are distributing an application that statically links with
+    libjpeg-turbo, then:
+
+    1.  Your product documentation must include a message stating:
+
+        This software is based in part on the work of the Independent JPEG
+        Group.
+
+        **Origin**
+        - Clause 2 of the IJG license
+
+    2.  If your binary distribution includes or uses the TurboJPEG API, then
+        your product documentation must include the text of the Modified BSD
+        License (see below.)
+
+        **Origin**
+        - Clause 2 of the Modified BSD License
+
+3.  You cannot use the name of the IJG or The libjpeg-turbo Project or the
+    contributors thereof in advertising, publicity, etc.
+
+    **Origin**
+    - IJG License
+    - Clause 3 of the Modified BSD License
+
+4.  The IJG and The libjpeg-turbo Project do not warrant libjpeg-turbo to be
+    free of defects, nor do we accept any liability for undesirable
+    consequences resulting from your use of the software.
+
+    **Origin**
+    - IJG License
+    - Modified BSD License
+    - zlib License
+
+
+The Modified (3-clause) BSD License
+===================================
+
+Copyright (C)2009-2019 D. R. Commander.  All Rights Reserved.
+Copyright (C)2015 Viktor Szathmáry.  All Rights Reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+- Redistributions of source code must retain the above copyright notice,
+  this list of conditions and the following disclaimer.
+- Redistributions in binary form must reproduce the above copyright notice,
+  this list of conditions and the following disclaimer in the documentation
+  and/or other materials provided with the distribution.
+- Neither the name of the libjpeg-turbo Project nor the names of its
+  contributors may be used to endorse or promote products derived from this
+  software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS",
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+POSSIBILITY OF SUCH DAMAGE.
+
+
+Why Three Licenses?
+===================
+
+The zlib License could have been used instead of the Modified (3-clause) BSD
+License, and since the IJG License effectively subsumes the distribution
+conditions of the zlib License, this would have effectively placed
+libjpeg-turbo binary distributions under the IJG License.  However, the IJG
+License specifically refers to the Independent JPEG Group and does not extend
+attribution and endorsement protections to other entities.  Thus, it was
+desirable to choose a license that granted us the same protections for new code
+that were granted to the IJG for code derived from their software.

3rdparty/libjpeg-turbo/README.ijg

@@ -0,0 +1,277 @@
+libjpeg-turbo note:  This file has been modified by The libjpeg-turbo Project
+to include only information relevant to libjpeg-turbo, to wordsmith certain
+sections, and to remove impolitic language that existed in the libjpeg v8
+README.  It is included only for reference.  Please see README.md for
+information specific to libjpeg-turbo.
+
+
+The Independent JPEG Group's JPEG software
+==========================================
+
+This distribution contains a release of the Independent JPEG Group's free JPEG
+software.  You are welcome to redistribute this software and to use it for any
+purpose, subject to the conditions under LEGAL ISSUES, below.
+
+This software is the work of Tom Lane, Guido Vollbeding, Philip Gladstone,
+Bill Allombert, Jim Boucher, Lee Crocker, Bob Friesenhahn, Ben Jackson,
+Julian Minguillon, Luis Ortiz, George Phillips, Davide Rossi, Ge' Weijers,
+and other members of the Independent JPEG Group.
+
+IJG is not affiliated with the ISO/IEC JTC1/SC29/WG1 standards committee
+(also known as JPEG, together with ITU-T SG16).
+
+
+DOCUMENTATION ROADMAP
+=====================
+
+This file contains the following sections:
+
+OVERVIEW            General description of JPEG and the IJG software.
+LEGAL ISSUES        Copyright, lack of warranty, terms of distribution.
+REFERENCES          Where to learn more about JPEG.
+ARCHIVE LOCATIONS   Where to find newer versions of this software.
+FILE FORMAT WARS    Software *not* to get.
+TO DO               Plans for future IJG releases.
+
+Other documentation files in the distribution are:
+
+User documentation:
+  usage.txt         Usage instructions for cjpeg, djpeg, jpegtran,
+                    rdjpgcom, and wrjpgcom.
+  *.1               Unix-style man pages for programs (same info as usage.txt).
+  wizard.txt        Advanced usage instructions for JPEG wizards only.
+  change.log        Version-to-version change highlights.
+Programmer and internal documentation:
+  libjpeg.txt       How to use the JPEG library in your own programs.
+  example.txt       Sample code for calling the JPEG library.
+  structure.txt     Overview of the JPEG library's internal structure.
+  coderules.txt     Coding style rules --- please read if you contribute code.
+
+Please read at least usage.txt.  Some information can also be found in the JPEG
+FAQ (Frequently Asked Questions) article.  See ARCHIVE LOCATIONS below to find
+out where to obtain the FAQ article.
+
+If you want to understand how the JPEG code works, we suggest reading one or
+more of the REFERENCES, then looking at the documentation files (in roughly
+the order listed) before diving into the code.
+
+
+OVERVIEW
+========
+
+This package contains C software to implement JPEG image encoding, decoding,
+and transcoding.  JPEG (pronounced "jay-peg") is a standardized compression
+method for full-color and grayscale images.  JPEG's strong suit is compressing
+photographic images or other types of images that have smooth color and
+brightness transitions between neighboring pixels.  Images with sharp lines or
+other abrupt features may not compress well with JPEG, and a higher JPEG
+quality may have to be used to avoid visible compression artifacts with such
+images.
+
+JPEG is lossy, meaning that the output pixels are not necessarily identical to
+the input pixels.  However, on photographic content and other "smooth" images,
+very good compression ratios can be obtained with no visible compression
+artifacts, and extremely high compression ratios are possible if you are
+willing to sacrifice image quality (by reducing the "quality" setting in the
+compressor.)
+
+This software implements JPEG baseline, extended-sequential, and progressive
+compression processes.  Provision is made for supporting all variants of these
+processes, although some uncommon parameter settings aren't implemented yet.
+We have made no provision for supporting the hierarchical or lossless
+processes defined in the standard.
+
+We provide a set of library routines for reading and writing JPEG image files,
+plus two sample applications "cjpeg" and "djpeg", which use the library to
+perform conversion between JPEG and some other popular image file formats.
+The library is intended to be reused in other applications.
+
+In order to support file conversion and viewing software, we have included
+considerable functionality beyond the bare JPEG coding/decoding capability;
+for example, the color quantization modules are not strictly part of JPEG
+decoding, but they are essential for output to colormapped file formats or
+colormapped displays.  These extra functions can be compiled out of the
+library if not required for a particular application.
+
+We have also included "jpegtran", a utility for lossless transcoding between
+different JPEG processes, and "rdjpgcom" and "wrjpgcom", two simple
+applications for inserting and extracting textual comments in JFIF files.
+
+The emphasis in designing this software has been on achieving portability and
+flexibility, while also making it fast enough to be useful.  In particular,
+the software is not intended to be read as a tutorial on JPEG.  (See the
+REFERENCES section for introductory material.)  Rather, it is intended to
+be reliable, portable, industrial-strength code.  We do not claim to have
+achieved that goal in every aspect of the software, but we strive for it.
+
+We welcome the use of this software as a component of commercial products.
+No royalty is required, but we do ask for an acknowledgement in product
+documentation, as described under LEGAL ISSUES.
+
+
+LEGAL ISSUES
+============
+
+In plain English:
+
+1. We don't promise that this software works.  (But if you find any bugs,
+   please let us know!)
+2. You can use this software for whatever you want.  You don't have to pay us.
+3. You may not pretend that you wrote this software.  If you use it in a
+   program, you must acknowledge somewhere in your documentation that
+   you've used the IJG code.
+
+In legalese:
+
+The authors make NO WARRANTY or representation, either express or implied,
+with respect to this software, its quality, accuracy, merchantability, or
+fitness for a particular purpose.  This software is provided "AS IS", and you,
+its user, assume the entire risk as to its quality and accuracy.
+
+This software is copyright (C) 1991-2016, Thomas G. Lane, Guido Vollbeding.
+All Rights Reserved except as specified below.
+
+Permission is hereby granted to use, copy, modify, and distribute this
+software (or portions thereof) for any purpose, without fee, subject to these
+conditions:
+(1) If any part of the source code for this software is distributed, then this
+README file must be included, with this copyright and no-warranty notice
+unaltered; and any additions, deletions, or changes to the original files
+must be clearly indicated in accompanying documentation.
+(2) If only executable code is distributed, then the accompanying
+documentation must state that "this software is based in part on the work of
+the Independent JPEG Group".
+(3) Permission for use of this software is granted only if the user accepts
+full responsibility for any undesirable consequences; the authors accept
+NO LIABILITY for damages of any kind.
+
+These conditions apply to any software derived from or based on the IJG code,
+not just to the unmodified library.  If you use our work, you ought to
+acknowledge us.
+
+Permission is NOT granted for the use of any IJG author's name or company name
+in advertising or publicity relating to this software or products derived from
+it.  This software may be referred to only as "the Independent JPEG Group's
+software".
+
+We specifically permit and encourage the use of this software as the basis of
+commercial products, provided that all warranty or liability claims are
+assumed by the product vendor.
+
+
+The IJG distribution formerly included code to read and write GIF files.
+To avoid entanglement with the Unisys LZW patent (now expired), GIF reading
+support has been removed altogether, and the GIF writer has been simplified
+to produce "uncompressed GIFs".  This technique does not use the LZW
+algorithm; the resulting GIF files are larger than usual, but are readable
+by all standard GIF decoders.
+
+We are required to state that
+    "The Graphics Interchange Format(c) is the Copyright property of
+    CompuServe Incorporated.  GIF(sm) is a Service Mark property of
+    CompuServe Incorporated."
+
+
+REFERENCES
+==========
+
+We recommend reading one or more of these references before trying to
+understand the innards of the JPEG software.
+
+The best short technical introduction to the JPEG compression algorithm is
+        Wallace, Gregory K.  "The JPEG Still Picture Compression Standard",
+        Communications of the ACM, April 1991 (vol. 34 no. 4), pp. 30-44.
+(Adjacent articles in that issue discuss MPEG motion picture compression,
+applications of JPEG, and related topics.)  If you don't have the CACM issue
+handy, a PDF file containing a revised version of Wallace's article is
+available at http://www.ijg.org/files/Wallace.JPEG.pdf.  The file (actually
+a preprint for an article that appeared in IEEE Trans. Consumer Electronics)
+omits the sample images that appeared in CACM, but it includes corrections
+and some added material.  Note: the Wallace article is copyright ACM and IEEE,
+and it may not be used for commercial purposes.
+
+A somewhat less technical, more leisurely introduction to JPEG can be found in
+"The Data Compression Book" by Mark Nelson and Jean-loup Gailly, published by
+M&T Books (New York), 2nd ed. 1996, ISBN 1-55851-434-1.  This book provides
+good explanations and example C code for a multitude of compression methods
+including JPEG.  It is an excellent source if you are comfortable reading C
+code but don't know much about data compression in general.  The book's JPEG
+sample code is far from industrial-strength, but when you are ready to look
+at a full implementation, you've got one here...
+
+The best currently available description of JPEG is the textbook "JPEG Still
+Image Data Compression Standard" by William B. Pennebaker and Joan L.
+Mitchell, published by Van Nostrand Reinhold, 1993, ISBN 0-442-01272-1.
+Price US$59.95, 638 pp.  The book includes the complete text of the ISO JPEG
+standards (DIS 10918-1 and draft DIS 10918-2).
+
+The original JPEG standard is divided into two parts, Part 1 being the actual
+specification, while Part 2 covers compliance testing methods.  Part 1 is
+titled "Digital Compression and Coding of Continuous-tone Still Images,
+Part 1: Requirements and guidelines" and has document numbers ISO/IEC IS
+10918-1, ITU-T T.81.  Part 2 is titled "Digital Compression and Coding of
+Continuous-tone Still Images, Part 2: Compliance testing" and has document
+numbers ISO/IEC IS 10918-2, ITU-T T.83.
+
+The JPEG standard does not specify all details of an interchangeable file
+format.  For the omitted details, we follow the "JFIF" conventions, revision
+1.02.  JFIF version 1 has been adopted as ISO/IEC 10918-5 (05/2013) and
+Recommendation ITU-T T.871 (05/2011): Information technology - Digital
+compression and coding of continuous-tone still images: JPEG File Interchange
+Format (JFIF).  It is available as a free download in PDF file format from
+https://www.iso.org/standard/54989.html and http://www.itu.int/rec/T-REC-T.871.
+A PDF file of the older JFIF 1.02 specification is available at
+http://www.w3.org/Graphics/JPEG/jfif3.pdf.
+
+The TIFF 6.0 file format specification can be obtained by FTP from
+ftp://ftp.sgi.com/graphics/tiff/TIFF6.ps.gz.  The JPEG incorporation scheme
+found in the TIFF 6.0 spec of 3-June-92 has a number of serious problems.
+IJG does not recommend use of the TIFF 6.0 design (TIFF Compression tag 6).
+Instead, we recommend the JPEG design proposed by TIFF Technical Note #2
+(Compression tag 7).  Copies of this Note can be obtained from
+http://www.ijg.org/files/.  It is expected that the next revision
+of the TIFF spec will replace the 6.0 JPEG design with the Note's design.
+Although IJG's own code does not support TIFF/JPEG, the free libtiff library
+uses our library to implement TIFF/JPEG per the Note.
+
+
+ARCHIVE LOCATIONS
+=================
+
+The "official" archive site for this software is www.ijg.org.
+The most recent released version can always be found there in
+directory "files".
+
+The JPEG FAQ (Frequently Asked Questions) article is a source of some
+general information about JPEG.
+It is available on the World Wide Web at http://www.faqs.org/faqs/jpeg-faq/
+and other news.answers archive sites, including the official news.answers
+archive at rtfm.mit.edu: ftp://rtfm.mit.edu/pub/usenet/news.answers/jpeg-faq/.
+If you don't have Web or FTP access, send e-mail to mail-server@rtfm.mit.edu
+with body
+        send usenet/news.answers/jpeg-faq/part1
+        send usenet/news.answers/jpeg-faq/part2
+
+
+FILE FORMAT COMPATIBILITY
+=========================
+
+This software implements ITU T.81 | ISO/IEC 10918 with some extensions from
+ITU T.871 | ISO/IEC 10918-5 (JPEG File Interchange Format-- see REFERENCES).
+Informally, the term "JPEG image" or "JPEG file" most often refers to JFIF or
+a subset thereof, but there are other formats containing the name "JPEG" that
+are incompatible with the DCT-based JPEG standard or with JFIF (for instance,
+JPEG 2000 and JPEG XR).  This software therefore does not support these
+formats.  Indeed, one of the original reasons for developing this free software
+was to help force convergence on a common, interoperable format standard for
+JPEG files.
+
+JFIF is a minimal or "low end" representation.  TIFF/JPEG (TIFF revision 6.0 as
+modified by TIFF Technical Note #2) can be used for "high end" applications
+that need to record a lot of additional data about an image.
+
+
+TO DO
+=====
+
+Please send bug reports, offers of help, etc. to jpeg-info@jpegclub.org.

3rdparty/libjpeg-turbo/README.md

@@ -0,0 +1,346 @@
+Background
+==========
+
+libjpeg-turbo is a JPEG image codec that uses SIMD instructions (MMX, SSE2,
+AVX2, NEON, AltiVec) to accelerate baseline JPEG compression and decompression
+on x86, x86-64, ARM, and PowerPC systems, as well as progressive JPEG
+compression on x86 and x86-64 systems.  On such systems, libjpeg-turbo is
+generally 2-6x as fast as libjpeg, all else being equal.  On other types of
+systems, libjpeg-turbo can still outperform libjpeg by a significant amount, by
+virtue of its highly-optimized Huffman coding routines.  In many cases, the
+performance of libjpeg-turbo rivals that of proprietary high-speed JPEG codecs.
+
+libjpeg-turbo implements both the traditional libjpeg API as well as the less
+powerful but more straightforward TurboJPEG API.  libjpeg-turbo also features
+colorspace extensions that allow it to compress from/decompress to 32-bit and
+big-endian pixel buffers (RGBX, XBGR, etc.), as well as a full-featured Java
+interface.
+
+libjpeg-turbo was originally based on libjpeg/SIMD, an MMX-accelerated
+derivative of libjpeg v6b developed by Miyasaka Masaru.  The TigerVNC and
+VirtualGL projects made numerous enhancements to the codec in 2009, and in
+early 2010, libjpeg-turbo spun off into an independent project, with the goal
+of making high-speed JPEG compression/decompression technology available to a
+broader range of users and developers.
+
+
+License
+=======
+
+libjpeg-turbo is covered by three compatible BSD-style open source licenses.
+Refer to [LICENSE.md](LICENSE.md) for a roll-up of license terms.
+
+
+Building libjpeg-turbo
+======================
+
+Refer to [BUILDING.md](BUILDING.md) for complete instructions.
+
+
+Using libjpeg-turbo
+===================
+
+libjpeg-turbo includes two APIs that can be used to compress and decompress
+JPEG images:
+
+- **TurboJPEG API**<br>
+  This API provides an easy-to-use interface for compressing and decompressing
+  JPEG images in memory.  It also provides some functionality that would not be
+  straightforward to achieve using the underlying libjpeg API, such as
+  generating planar YUV images and performing multiple simultaneous lossless
+  transforms on an image.  The Java interface for libjpeg-turbo is written on
+  top of the TurboJPEG API.  The TurboJPEG API is recommended for first-time
+  users of libjpeg-turbo.  Refer to [tjexample.c](tjexample.c) and
+  [TJExample.java](java/TJExample.java) for examples of its usage and to
+  <http://libjpeg-turbo.org/Documentation/Documentation> for API documentation.
+
+- **libjpeg API**<br>
+  This is the de facto industry-standard API for compressing and decompressing
+  JPEG images.  It is more difficult to use than the TurboJPEG API but also
+  more powerful.  The libjpeg API implementation in libjpeg-turbo is both
+  API/ABI-compatible and mathematically compatible with libjpeg v6b.  It can
+  also optionally be configured to be API/ABI-compatible with libjpeg v7 and v8
+  (see below.)  Refer to [cjpeg.c](cjpeg.c) and [djpeg.c](djpeg.c) for examples
+  of its usage and to [libjpeg.txt](libjpeg.txt) for API documentation.
+
+There is no significant performance advantage to either API when both are used
+to perform similar operations.
+
+Colorspace Extensions
+---------------------
+
+libjpeg-turbo includes extensions that allow JPEG images to be compressed
+directly from (and decompressed directly to) buffers that use BGR, BGRX,
+RGBX, XBGR, and XRGB pixel ordering.  This is implemented with ten new
+colorspace constants:
+
+    JCS_EXT_RGB   /* red/green/blue */
+    JCS_EXT_RGBX  /* red/green/blue/x */
+    JCS_EXT_BGR   /* blue/green/red */
+    JCS_EXT_BGRX  /* blue/green/red/x */
+    JCS_EXT_XBGR  /* x/blue/green/red */
+    JCS_EXT_XRGB  /* x/red/green/blue */
+    JCS_EXT_RGBA  /* red/green/blue/alpha */
+    JCS_EXT_BGRA  /* blue/green/red/alpha */
+    JCS_EXT_ABGR  /* alpha/blue/green/red */
+    JCS_EXT_ARGB  /* alpha/red/green/blue */
+
+Setting `cinfo.in_color_space` (compression) or `cinfo.out_color_space`
+(decompression) to one of these values will cause libjpeg-turbo to read the
+red, green, and blue values from (or write them to) the appropriate position in
+the pixel when compressing from/decompressing to an RGB buffer.
+
+Your application can check for the existence of these extensions at compile
+time with:
+
+    #ifdef JCS_EXTENSIONS
+
+At run time, attempting to use these extensions with a libjpeg implementation
+that does not support them will result in a "Bogus input colorspace" error.
+Applications can trap this error in order to test whether run-time support is
+available for the colorspace extensions.
+
+When using the RGBX, BGRX, XBGR, and XRGB colorspaces during decompression, the
+X byte is undefined, and in order to ensure the best performance, libjpeg-turbo
+can set that byte to whatever value it wishes.  If an application expects the X
+byte to be used as an alpha channel, then it should specify `JCS_EXT_RGBA`,
+`JCS_EXT_BGRA`, `JCS_EXT_ABGR`, or `JCS_EXT_ARGB`.  When these colorspace
+constants are used, the X byte is guaranteed to be 0xFF, which is interpreted
+as opaque.
+
+Your application can check for the existence of the alpha channel colorspace
+extensions at compile time with:
+
+    #ifdef JCS_ALPHA_EXTENSIONS
+
+[jcstest.c](jcstest.c), located in the libjpeg-turbo source tree, demonstrates
+how to check for the existence of the colorspace extensions at compile time and
+run time.
+
+libjpeg v7 and v8 API/ABI Emulation
+-----------------------------------
+
+With libjpeg v7 and v8, new features were added that necessitated extending the
+compression and decompression structures.  Unfortunately, due to the exposed
+nature of those structures, extending them also necessitated breaking backward
+ABI compatibility with previous libjpeg releases.  Thus, programs that were
+built to use libjpeg v7 or v8 did not work with libjpeg-turbo, since it is
+based on the libjpeg v6b code base.  Although libjpeg v7 and v8 are not
+as widely used as v6b, enough programs (including a few Linux distros) made
+the switch that there was a demand to emulate the libjpeg v7 and v8 ABIs
+in libjpeg-turbo.  It should be noted, however, that this feature was added
+primarily so that applications that had already been compiled to use libjpeg
+v7+ could take advantage of accelerated baseline JPEG encoding/decoding
+without recompiling.  libjpeg-turbo does not claim to support all of the
+libjpeg v7+ features, nor to produce identical output to libjpeg v7+ in all
+cases (see below.)
+
+By passing an argument of `--with-jpeg7` or `--with-jpeg8` to `configure`, or
+an argument of `-DWITH_JPEG7=1` or `-DWITH_JPEG8=1` to `cmake`, you can build a
+version of libjpeg-turbo that emulates the libjpeg v7 or v8 ABI, so that
+programs that are built against libjpeg v7 or v8 can be run with libjpeg-turbo.
+The following section describes which libjpeg v7+ features are supported and
+which aren't.
+
+### Support for libjpeg v7 and v8 Features
+
+#### Fully supported
+
+- **libjpeg: IDCT scaling extensions in decompressor**<br>
+  libjpeg-turbo supports IDCT scaling with scaling factors of 1/8, 1/4, 3/8,
+  1/2, 5/8, 3/4, 7/8, 9/8, 5/4, 11/8, 3/2, 13/8, 7/4, 15/8, and 2/1 (only 1/4
+  and 1/2 are SIMD-accelerated.)
+
+- **libjpeg: Arithmetic coding**
+
+- **libjpeg: In-memory source and destination managers**<br>
+  See notes below.
+
+- **cjpeg: Separate quality settings for luminance and chrominance**<br>
+  Note that the libpjeg v7+ API was extended to accommodate this feature only
+  for convenience purposes.  It has always been possible to implement this
+  feature with libjpeg v6b (see rdswitch.c for an example.)
+
+- **cjpeg: 32-bit BMP support**
+
+- **cjpeg: `-rgb` option**
+
+- **jpegtran: Lossless cropping**
+
+- **jpegtran: `-perfect` option**
+
+- **jpegtran: Forcing width/height when performing lossless crop**
+
+- **rdjpgcom: `-raw` option**
+
+- **rdjpgcom: Locale awareness**
+
+
+#### Not supported
+
+NOTE:  As of this writing, extensive research has been conducted into the
+usefulness of DCT scaling as a means of data reduction and SmartScale as a
+means of quality improvement.  The reader is invited to peruse the research at
+<http://www.libjpeg-turbo.org/About/SmartScale> and draw his/her own conclusions,
+but it is the general belief of our project that these features have not
+demonstrated sufficient usefulness to justify inclusion in libjpeg-turbo.
+
+- **libjpeg: DCT scaling in compressor**<br>
+  `cinfo.scale_num` and `cinfo.scale_denom` are silently ignored.
+  There is no technical reason why DCT scaling could not be supported when
+  emulating the libjpeg v7+ API/ABI, but without the SmartScale extension (see
+  below), only scaling factors of 1/2, 8/15, 4/7, 8/13, 2/3, 8/11, 4/5, and
+  8/9 would be available, which is of limited usefulness.
+
+- **libjpeg: SmartScale**<br>
+  `cinfo.block_size` is silently ignored.
+  SmartScale is an extension to the JPEG format that allows for DCT block
+  sizes other than 8x8.  Providing support for this new format would be
+  feasible (particularly without full acceleration.)  However, until/unless
+  the format becomes either an official industry standard or, at minimum, an
+  accepted solution in the community, we are hesitant to implement it, as
+  there is no sense of whether or how it might change in the future.  It is
+  our belief that SmartScale has not demonstrated sufficient usefulness as a
+  lossless format nor as a means of quality enhancement, and thus our primary
+  interest in providing this feature would be as a means of supporting
+  additional DCT scaling factors.
+
+- **libjpeg: Fancy downsampling in compressor**<br>
+  `cinfo.do_fancy_downsampling` is silently ignored.
+  This requires the DCT scaling feature, which is not supported.
+
+- **jpegtran: Scaling**<br>
+  This requires both the DCT scaling and SmartScale features, which are not
+  supported.
+
+- **Lossless RGB JPEG files**<br>
+  This requires the SmartScale feature, which is not supported.
+
+### What About libjpeg v9?
+
+libjpeg v9 introduced yet another field to the JPEG compression structure
+(`color_transform`), thus making the ABI backward incompatible with that of
+libjpeg v8.  This new field was introduced solely for the purpose of supporting
+lossless SmartScale encoding.  Furthermore, there was actually no reason to
+extend the API in this manner, as the color transform could have just as easily
+been activated by way of a new JPEG colorspace constant, thus preserving
+backward ABI compatibility.
+
+Our research (see link above) has shown that lossless SmartScale does not
+generally accomplish anything that can't already be accomplished better with
+existing, standard lossless formats.  Therefore, at this time it is our belief
+that there is not sufficient technical justification for software projects to
+upgrade from libjpeg v8 to libjpeg v9, and thus there is not sufficient
+technical justification for us to emulate the libjpeg v9 ABI.
+
+In-Memory Source/Destination Managers
+-------------------------------------
+
+By default, libjpeg-turbo 1.3 and later includes the `jpeg_mem_src()` and
+`jpeg_mem_dest()` functions, even when not emulating the libjpeg v8 API/ABI.
+Previously, it was necessary to build libjpeg-turbo from source with libjpeg v8
+API/ABI emulation in order to use the in-memory source/destination managers,
+but several projects requested that those functions be included when emulating
+the libjpeg v6b API/ABI as well.  This allows the use of those functions by
+programs that need them, without breaking ABI compatibility for programs that
+don't, and it allows those functions to be provided in the "official"
+libjpeg-turbo binaries.
+
+Those who are concerned about maintaining strict conformance with the libjpeg
+v6b or v7 API can pass an argument of `--without-mem-srcdst` to `configure` or
+an argument of `-DWITH_MEM_SRCDST=0` to `cmake` prior to building
+libjpeg-turbo.  This will restore the pre-1.3 behavior, in which
+`jpeg_mem_src()` and `jpeg_mem_dest()` are only included when emulating the
+libjpeg v8 API/ABI.
+
+On Un*x systems, including the in-memory source/destination managers changes
+the dynamic library version from 62.1.0 to 62.2.0 if using libjpeg v6b API/ABI
+emulation and from 7.1.0 to 7.2.0 if using libjpeg v7 API/ABI emulation.
+
+Note that, on most Un*x systems, the dynamic linker will not look for a
+function in a library until that function is actually used.  Thus, if a program
+is built against libjpeg-turbo 1.3+ and uses `jpeg_mem_src()` or
+`jpeg_mem_dest()`, that program will not fail if run against an older version
+of libjpeg-turbo or against libjpeg v7- until the program actually tries to
+call `jpeg_mem_src()` or `jpeg_mem_dest()`.  Such is not the case on Windows.
+If a program is built against the libjpeg-turbo 1.3+ DLL and uses
+`jpeg_mem_src()` or `jpeg_mem_dest()`, then it must use the libjpeg-turbo 1.3+
+DLL at run time.
+
+Both cjpeg and djpeg have been extended to allow testing the in-memory
+source/destination manager functions.  See their respective man pages for more
+details.
+
+
+Mathematical Compatibility
+==========================
+
+For the most part, libjpeg-turbo should produce identical output to libjpeg
+v6b.  The one exception to this is when using the floating point DCT/IDCT, in
+which case the outputs of libjpeg v6b and libjpeg-turbo can differ for the
+following reasons:
+
+- The SSE/SSE2 floating point DCT implementation in libjpeg-turbo is ever so
+  slightly more accurate than the implementation in libjpeg v6b, but not by
+  any amount perceptible to human vision (generally in the range of 0.01 to
+  0.08 dB gain in PNSR.)
+
+- When not using the SIMD extensions, libjpeg-turbo uses the more accurate
+  (and slightly faster) floating point IDCT algorithm introduced in libjpeg
+  v8a as opposed to the algorithm used in libjpeg v6b.  It should be noted,
+  however, that this algorithm basically brings the accuracy of the floating
+  point IDCT in line with the accuracy of the slow integer IDCT.  The floating
+  point DCT/IDCT algorithms are mainly a legacy feature, and they do not
+  produce significantly more accuracy than the slow integer algorithms (to put
+  numbers on this, the typical difference in PNSR between the two algorithms
+  is less than 0.10 dB, whereas changing the quality level by 1 in the upper
+  range of the quality scale is typically more like a 1.0 dB difference.)
+
+- If the floating point algorithms in libjpeg-turbo are not implemented using
+  SIMD instructions on a particular platform, then the accuracy of the
+  floating point DCT/IDCT can depend on the compiler settings.
+
+While libjpeg-turbo does emulate the libjpeg v8 API/ABI, under the hood it is
+still using the same algorithms as libjpeg v6b, so there are several specific
+cases in which libjpeg-turbo cannot be expected to produce the same output as
+libjpeg v8:
+
+- When decompressing using scaling factors of 1/2 and 1/4, because libjpeg v8
+  implements those scaling algorithms differently than libjpeg v6b does, and
+  libjpeg-turbo's SIMD extensions are based on the libjpeg v6b behavior.
+
+- When using chrominance subsampling, because libjpeg v8 implements this
+  with its DCT/IDCT scaling algorithms rather than with a separate
+  downsampling/upsampling algorithm.  In our testing, the subsampled/upsampled
+  output of libjpeg v8 is less accurate than that of libjpeg v6b for this
+  reason.
+
+- When decompressing using a scaling factor > 1 and merged (AKA "non-fancy" or
+  "non-smooth") chrominance upsampling, because libjpeg v8 does not support
+  merged upsampling with scaling factors > 1.
+
+
+Performance Pitfalls
+====================
+
+Restart Markers
+---------------
+
+The optimized Huffman decoder in libjpeg-turbo does not handle restart markers
+in a way that makes the rest of the libjpeg infrastructure happy, so it is
+necessary to use the slow Huffman decoder when decompressing a JPEG image that
+has restart markers.  This can cause the decompression performance to drop by
+as much as 20%, but the performance will still be much greater than that of
+libjpeg.  Many consumer packages, such as PhotoShop, use restart markers when
+generating JPEG images, so images generated by those programs will experience
+this issue.
+
+Fast Integer Forward DCT at High Quality Levels
+-----------------------------------------------
+
+The algorithm used by the SIMD-accelerated quantization function cannot produce
+correct results whenever the fast integer forward DCT is used along with a JPEG
+quality of 98-100.  Thus, libjpeg-turbo must use the non-SIMD quantization
+function in those cases.  This causes performance to drop by as much as 40%.
+It is therefore strongly advised that you use the slow integer forward DCT
+whenever encoding images with a JPEG quality of 98 or higher.

3rdparty/libjpeg-turbo/jconfig.h.in

@@ -0,0 +1,73 @@
+/* Version ID for the JPEG library.
+ * Might be useful for tests like "#if JPEG_LIB_VERSION >= 60".
+ */
+#define JPEG_LIB_VERSION  @JPEG_LIB_VERSION@
+
+/* libjpeg-turbo version */
+#define LIBJPEG_TURBO_VERSION  @VERSION@
+
+/* libjpeg-turbo version in integer form */
+#define LIBJPEG_TURBO_VERSION_NUMBER  @LIBJPEG_TURBO_VERSION_NUMBER@
+
+/* Support arithmetic encoding */
+#cmakedefine C_ARITH_CODING_SUPPORTED 1
+
+/* Support arithmetic decoding */
+#cmakedefine D_ARITH_CODING_SUPPORTED 1
+
+/* Support in-memory source/destination managers */
+#cmakedefine MEM_SRCDST_SUPPORTED 1
+
+/* Use accelerated SIMD routines. */
+#cmakedefine WITH_SIMD 1
+
+/*
+ * Define BITS_IN_JSAMPLE as either
+ *   8   for 8-bit sample values (the usual setting)
+ *   12  for 12-bit sample values
+ * Only 8 and 12 are legal data precisions for lossy JPEG according to the
+ * JPEG standard, and the IJG code does not support anything else!
+ * We do not support run-time selection of data precision, sorry.
+ */
+
+#define BITS_IN_JSAMPLE  @BITS_IN_JSAMPLE@      /* use 8 or 12 */
+
+/* Define to 1 if you have the <locale.h> header file. */
+#cmakedefine HAVE_LOCALE_H 1
+
+/* Define to 1 if you have the <stddef.h> header file. */
+#cmakedefine HAVE_STDDEF_H 1
+
+/* Define to 1 if you have the <stdlib.h> header file. */
+#cmakedefine HAVE_STDLIB_H 1
+
+/* Define if you need to include <sys/types.h> to get size_t. */
+#cmakedefine NEED_SYS_TYPES_H 1
+
+/* Define if you have BSD-like bzero and bcopy in <strings.h> rather than
+   memset/memcpy in <string.h>. */
+#cmakedefine NEED_BSD_STRINGS 1
+
+/* Define to 1 if the system has the type `unsigned char'. */
+#cmakedefine HAVE_UNSIGNED_CHAR 1
+
+/* Define to 1 if the system has the type `unsigned short'. */
+#cmakedefine HAVE_UNSIGNED_SHORT 1
+
+/* Compiler does not support pointers to undefined structures. */
+#cmakedefine INCOMPLETE_TYPES_BROKEN 1
+
+/* Define if your (broken) compiler shifts signed values as if they were
+   unsigned. */
+#cmakedefine RIGHT_SHIFT_IS_UNSIGNED 1
+
+/* Define to 1 if type `char' is unsigned and you are not using gcc.  */
+#ifndef __CHAR_UNSIGNED__
+  #cmakedefine __CHAR_UNSIGNED__ 1
+#endif
+
+/* Define to empty if `const' does not conform to ANSI C. */
+/* #undef const */
+
+/* Define to `unsigned int' if <sys/types.h> does not define. */
+/* #undef size_t */

3rdparty/libjpeg-turbo/jconfig.h.win.in

@@ -0,0 +1,34 @@
+#define JPEG_LIB_VERSION  @JPEG_LIB_VERSION@
+#define LIBJPEG_TURBO_VERSION  @VERSION@
+#define LIBJPEG_TURBO_VERSION_NUMBER  @LIBJPEG_TURBO_VERSION_NUMBER@
+
+#cmakedefine C_ARITH_CODING_SUPPORTED
+#cmakedefine D_ARITH_CODING_SUPPORTED
+#cmakedefine MEM_SRCDST_SUPPORTED
+#cmakedefine WITH_SIMD
+
+#define BITS_IN_JSAMPLE  @BITS_IN_JSAMPLE@      /* use 8 or 12 */
+
+#define HAVE_STDDEF_H
+#define HAVE_STDLIB_H
+#undef NEED_SYS_TYPES_H
+#undef NEED_BSD_STRINGS
+
+#define HAVE_UNSIGNED_CHAR
+#define HAVE_UNSIGNED_SHORT
+#undef INCOMPLETE_TYPES_BROKEN
+#undef RIGHT_SHIFT_IS_UNSIGNED
+#undef __CHAR_UNSIGNED__
+
+/* Define "boolean" as unsigned char, not int, per Windows custom */
+#ifndef __RPCNDR_H__            /* don't conflict if rpcndr.h already read */
+typedef unsigned char boolean;
+#endif
+#define HAVE_BOOLEAN            /* prevent jmorecfg.h from redefining it */
+
+/* Define "INT32" as int, not long, per Windows custom */
+#if !(defined(_BASETSD_H_) || defined(_BASETSD_H))   /* don't conflict if basetsd.h already read */
+typedef short INT16;
+typedef signed int INT32;
+#endif
+#define XMD_H                   /* prevent jmorecfg.h from redefining it */

3rdparty/libjpeg-turbo/jconfigint.h.in

@@ -0,0 +1,39 @@
+/* libjpeg-turbo build number */
+#define BUILD  "@BUILD@"
+
+/* Compiler's inline keyword */
+#undef inline
+
+/* How to obtain function inlining. */
+#ifndef INLINE
+#if defined(__GNUC__)
+#define INLINE inline __attribute__((always_inline))
+#elif defined(_MSC_VER)
+#define INLINE __forceinline
+#else
+#define INLINE
+#endif
+#endif
+
+/* Define to the full name of this package. */
+#define PACKAGE_NAME  "@CMAKE_PROJECT_NAME@"
+
+/* Version number of package */
+#define VERSION  "@VERSION@"
+
+/* The size of `size_t', as computed by sizeof. */
+#define SIZEOF_SIZE_T  @SIZEOF_SIZE_T@
+
+/* Define if your compiler has __builtin_ctzl() and sizeof(unsigned long) == sizeof(size_t). */
+#cmakedefine HAVE_BUILTIN_CTZL
+
+/* Define to 1 if you have the <intrin.h> header file. */
+#cmakedefine HAVE_INTRIN_H
+
+#if defined(_MSC_VER) && defined(HAVE_INTRIN_H)
+#if (SIZEOF_SIZE_T == 8)
+#define HAVE_BITSCANFORWARD64
+#elif (SIZEOF_SIZE_T == 4)
+#define HAVE_BITSCANFORWARD
+#endif
+#endif

3rdparty/libjpeg-turbo/src/jaricom.c

@@ -0,0 +1,157 @@
+/*
+ * jaricom.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Developed 1997-2009 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2015, 2018, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains probability estimation tables for common use in
+ * arithmetic entropy encoding and decoding routines.
+ *
+ * This data represents Table D.2 in
+ * Recommendation ITU-T T.81 (1992) | ISO/IEC 10918-1:1994 and Table 24 in
+ * Recommendation ITU-T T.82 (1993) | ISO/IEC 11544:1993.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+/* The following #define specifies the packing of the four components
+ * into the compact JLONG representation.
+ * Note that this formula must match the actual arithmetic encoder
+ * and decoder implementation.  The implementation has to be changed
+ * if this formula is changed.
+ * The current organization is leaned on Markus Kuhn's JBIG
+ * implementation (jbig_tab.c).
+ */
+
+#define V(i, a, b, c, d) \
+  (((JLONG)a << 16) | ((JLONG)c << 8) | ((JLONG)d << 7) | b)
+
+const JLONG jpeg_aritab[113 + 1] = {
+/*
+ * Index, Qe_Value, Next_Index_LPS, Next_Index_MPS, Switch_MPS
+ */
+  V(   0, 0x5a1d,   1,   1, 1 ),
+  V(   1, 0x2586,  14,   2, 0 ),
+  V(   2, 0x1114,  16,   3, 0 ),
+  V(   3, 0x080b,  18,   4, 0 ),
+  V(   4, 0x03d8,  20,   5, 0 ),
+  V(   5, 0x01da,  23,   6, 0 ),
+  V(   6, 0x00e5,  25,   7, 0 ),
+  V(   7, 0x006f,  28,   8, 0 ),
+  V(   8, 0x0036,  30,   9, 0 ),
+  V(   9, 0x001a,  33,  10, 0 ),
+  V(  10, 0x000d,  35,  11, 0 ),
+  V(  11, 0x0006,   9,  12, 0 ),
+  V(  12, 0x0003,  10,  13, 0 ),
+  V(  13, 0x0001,  12,  13, 0 ),
+  V(  14, 0x5a7f,  15,  15, 1 ),
+  V(  15, 0x3f25,  36,  16, 0 ),
+  V(  16, 0x2cf2,  38,  17, 0 ),
+  V(  17, 0x207c,  39,  18, 0 ),
+  V(  18, 0x17b9,  40,  19, 0 ),
+  V(  19, 0x1182,  42,  20, 0 ),
+  V(  20, 0x0cef,  43,  21, 0 ),
+  V(  21, 0x09a1,  45,  22, 0 ),
+  V(  22, 0x072f,  46,  23, 0 ),
+  V(  23, 0x055c,  48,  24, 0 ),
+  V(  24, 0x0406,  49,  25, 0 ),
+  V(  25, 0x0303,  51,  26, 0 ),
+  V(  26, 0x0240,  52,  27, 0 ),
+  V(  27, 0x01b1,  54,  28, 0 ),
+  V(  28, 0x0144,  56,  29, 0 ),
+  V(  29, 0x00f5,  57,  30, 0 ),
+  V(  30, 0x00b7,  59,  31, 0 ),
+  V(  31, 0x008a,  60,  32, 0 ),
+  V(  32, 0x0068,  62,  33, 0 ),
+  V(  33, 0x004e,  63,  34, 0 ),
+  V(  34, 0x003b,  32,  35, 0 ),
+  V(  35, 0x002c,  33,   9, 0 ),
+  V(  36, 0x5ae1,  37,  37, 1 ),
+  V(  37, 0x484c,  64,  38, 0 ),
+  V(  38, 0x3a0d,  65,  39, 0 ),
+  V(  39, 0x2ef1,  67,  40, 0 ),
+  V(  40, 0x261f,  68,  41, 0 ),
+  V(  41, 0x1f33,  69,  42, 0 ),
+  V(  42, 0x19a8,  70,  43, 0 ),
+  V(  43, 0x1518,  72,  44, 0 ),
+  V(  44, 0x1177,  73,  45, 0 ),
+  V(  45, 0x0e74,  74,  46, 0 ),
+  V(  46, 0x0bfb,  75,  47, 0 ),
+  V(  47, 0x09f8,  77,  48, 0 ),
+  V(  48, 0x0861,  78,  49, 0 ),
+  V(  49, 0x0706,  79,  50, 0 ),
+  V(  50, 0x05cd,  48,  51, 0 ),
+  V(  51, 0x04de,  50,  52, 0 ),
+  V(  52, 0x040f,  50,  53, 0 ),
+  V(  53, 0x0363,  51,  54, 0 ),
+  V(  54, 0x02d4,  52,  55, 0 ),
+  V(  55, 0x025c,  53,  56, 0 ),
+  V(  56, 0x01f8,  54,  57, 0 ),
+  V(  57, 0x01a4,  55,  58, 0 ),
+  V(  58, 0x0160,  56,  59, 0 ),
+  V(  59, 0x0125,  57,  60, 0 ),
+  V(  60, 0x00f6,  58,  61, 0 ),
+  V(  61, 0x00cb,  59,  62, 0 ),
+  V(  62, 0x00ab,  61,  63, 0 ),
+  V(  63, 0x008f,  61,  32, 0 ),
+  V(  64, 0x5b12,  65,  65, 1 ),
+  V(  65, 0x4d04,  80,  66, 0 ),
+  V(  66, 0x412c,  81,  67, 0 ),
+  V(  67, 0x37d8,  82,  68, 0 ),
+  V(  68, 0x2fe8,  83,  69, 0 ),
+  V(  69, 0x293c,  84,  70, 0 ),
+  V(  70, 0x2379,  86,  71, 0 ),
+  V(  71, 0x1edf,  87,  72, 0 ),
+  V(  72, 0x1aa9,  87,  73, 0 ),
+  V(  73, 0x174e,  72,  74, 0 ),
+  V(  74, 0x1424,  72,  75, 0 ),
+  V(  75, 0x119c,  74,  76, 0 ),
+  V(  76, 0x0f6b,  74,  77, 0 ),
+  V(  77, 0x0d51,  75,  78, 0 ),
+  V(  78, 0x0bb6,  77,  79, 0 ),
+  V(  79, 0x0a40,  77,  48, 0 ),
+  V(  80, 0x5832,  80,  81, 1 ),
+  V(  81, 0x4d1c,  88,  82, 0 ),
+  V(  82, 0x438e,  89,  83, 0 ),
+  V(  83, 0x3bdd,  90,  84, 0 ),
+  V(  84, 0x34ee,  91,  85, 0 ),
+  V(  85, 0x2eae,  92,  86, 0 ),
+  V(  86, 0x299a,  93,  87, 0 ),
+  V(  87, 0x2516,  86,  71, 0 ),
+  V(  88, 0x5570,  88,  89, 1 ),
+  V(  89, 0x4ca9,  95,  90, 0 ),
+  V(  90, 0x44d9,  96,  91, 0 ),
+  V(  91, 0x3e22,  97,  92, 0 ),
+  V(  92, 0x3824,  99,  93, 0 ),
+  V(  93, 0x32b4,  99,  94, 0 ),
+  V(  94, 0x2e17,  93,  86, 0 ),
+  V(  95, 0x56a8,  95,  96, 1 ),
+  V(  96, 0x4f46, 101,  97, 0 ),
+  V(  97, 0x47e5, 102,  98, 0 ),
+  V(  98, 0x41cf, 103,  99, 0 ),
+  V(  99, 0x3c3d, 104, 100, 0 ),
+  V( 100, 0x375e,  99,  93, 0 ),
+  V( 101, 0x5231, 105, 102, 0 ),
+  V( 102, 0x4c0f, 106, 103, 0 ),
+  V( 103, 0x4639, 107, 104, 0 ),
+  V( 104, 0x415e, 103,  99, 0 ),
+  V( 105, 0x5627, 105, 106, 1 ),
+  V( 106, 0x50e7, 108, 107, 0 ),
+  V( 107, 0x4b85, 109, 103, 0 ),
+  V( 108, 0x5597, 110, 109, 0 ),
+  V( 109, 0x504f, 111, 107, 0 ),
+  V( 110, 0x5a10, 110, 111, 1 ),
+  V( 111, 0x5522, 112, 109, 0 ),
+  V( 112, 0x59eb, 112, 111, 1 ),
+/*
+ * This last entry is used for fixed probability estimate of 0.5
+ * as recommended in Section 10.3 Table 5 of ITU-T Rec. T.851.
+ */
+  V( 113, 0x5a1d, 113, 113, 0 )
+};

3rdparty/libjpeg-turbo/src/jcapimin.c

@@ -0,0 +1,295 @@
+/*
+ * jcapimin.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1998, Thomas G. Lane.
+ * Modified 2003-2010 by Guido Vollbeding.
+ * It was modified by The libjpeg-turbo Project to include only code relevant
+ * to libjpeg-turbo.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains application interface code for the compression half
+ * of the JPEG library.  These are the "minimum" API routines that may be
+ * needed in either the normal full-compression case or the transcoding-only
+ * case.
+ *
+ * Most of the routines intended to be called directly by an application
+ * are in this file or in jcapistd.c.  But also see jcparam.c for
+ * parameter-setup helper routines, jcomapi.c for routines shared by
+ * compression and decompression, and jctrans.c for the transcoding case.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/*
+ * Initialization of a JPEG compression object.
+ * The error manager must already be set up (in case memory manager fails).
+ */
+
+GLOBAL(void)
+jpeg_CreateCompress(j_compress_ptr cinfo, int version, size_t structsize)
+{
+  int i;
+
+  /* Guard against version mismatches between library and caller. */
+  cinfo->mem = NULL;            /* so jpeg_destroy knows mem mgr not called */
+  if (version != JPEG_LIB_VERSION)
+    ERREXIT2(cinfo, JERR_BAD_LIB_VERSION, JPEG_LIB_VERSION, version);
+  if (structsize != sizeof(struct jpeg_compress_struct))
+    ERREXIT2(cinfo, JERR_BAD_STRUCT_SIZE,
+             (int)sizeof(struct jpeg_compress_struct), (int)structsize);
+
+  /* For debugging purposes, we zero the whole master structure.
+   * But the application has already set the err pointer, and may have set
+   * client_data, so we have to save and restore those fields.
+   * Note: if application hasn't set client_data, tools like Purify may
+   * complain here.
+   */
+  {
+    struct jpeg_error_mgr *err = cinfo->err;
+    void *client_data = cinfo->client_data; /* ignore Purify complaint here */
+    MEMZERO(cinfo, sizeof(struct jpeg_compress_struct));
+    cinfo->err = err;
+    cinfo->client_data = client_data;
+  }
+  cinfo->is_decompressor = FALSE;
+
+  /* Initialize a memory manager instance for this object */
+  jinit_memory_mgr((j_common_ptr)cinfo);
+
+  /* Zero out pointers to permanent structures. */
+  cinfo->progress = NULL;
+  cinfo->dest = NULL;
+
+  cinfo->comp_info = NULL;
+
+  for (i = 0; i < NUM_QUANT_TBLS; i++) {
+    cinfo->quant_tbl_ptrs[i] = NULL;
+#if JPEG_LIB_VERSION >= 70
+    cinfo->q_scale_factor[i] = 100;
+#endif
+  }
+
+  for (i = 0; i < NUM_HUFF_TBLS; i++) {
+    cinfo->dc_huff_tbl_ptrs[i] = NULL;
+    cinfo->ac_huff_tbl_ptrs[i] = NULL;
+  }
+
+#if JPEG_LIB_VERSION >= 80
+  /* Must do it here for emit_dqt in case jpeg_write_tables is used */
+  cinfo->block_size = DCTSIZE;
+  cinfo->natural_order = jpeg_natural_order;
+  cinfo->lim_Se = DCTSIZE2 - 1;
+#endif
+
+  cinfo->script_space = NULL;
+
+  cinfo->input_gamma = 1.0;     /* in case application forgets */
+
+  /* OK, I'm ready */
+  cinfo->global_state = CSTATE_START;
+}
+
+
+/*
+ * Destruction of a JPEG compression object
+ */
+
+GLOBAL(void)
+jpeg_destroy_compress(j_compress_ptr cinfo)
+{
+  jpeg_destroy((j_common_ptr)cinfo); /* use common routine */
+}
+
+
+/*
+ * Abort processing of a JPEG compression operation,
+ * but don't destroy the object itself.
+ */
+
+GLOBAL(void)
+jpeg_abort_compress(j_compress_ptr cinfo)
+{
+  jpeg_abort((j_common_ptr)cinfo); /* use common routine */
+}
+
+
+/*
+ * Forcibly suppress or un-suppress all quantization and Huffman tables.
+ * Marks all currently defined tables as already written (if suppress)
+ * or not written (if !suppress).  This will control whether they get emitted
+ * by a subsequent jpeg_start_compress call.
+ *
+ * This routine is exported for use by applications that want to produce
+ * abbreviated JPEG datastreams.  It logically belongs in jcparam.c, but
+ * since it is called by jpeg_start_compress, we put it here --- otherwise
+ * jcparam.o would be linked whether the application used it or not.
+ */
+
+GLOBAL(void)
+jpeg_suppress_tables(j_compress_ptr cinfo, boolean suppress)
+{
+  int i;
+  JQUANT_TBL *qtbl;
+  JHUFF_TBL *htbl;
+
+  for (i = 0; i < NUM_QUANT_TBLS; i++) {
+    if ((qtbl = cinfo->quant_tbl_ptrs[i]) != NULL)
+      qtbl->sent_table = suppress;
+  }
+
+  for (i = 0; i < NUM_HUFF_TBLS; i++) {
+    if ((htbl = cinfo->dc_huff_tbl_ptrs[i]) != NULL)
+      htbl->sent_table = suppress;
+    if ((htbl = cinfo->ac_huff_tbl_ptrs[i]) != NULL)
+      htbl->sent_table = suppress;
+  }
+}
+
+
+/*
+ * Finish JPEG compression.
+ *
+ * If a multipass operating mode was selected, this may do a great deal of
+ * work including most of the actual output.
+ */
+
+GLOBAL(void)
+jpeg_finish_compress(j_compress_ptr cinfo)
+{
+  JDIMENSION iMCU_row;
+
+  if (cinfo->global_state == CSTATE_SCANNING ||
+      cinfo->global_state == CSTATE_RAW_OK) {
+    /* Terminate first pass */
+    if (cinfo->next_scanline < cinfo->image_height)
+      ERREXIT(cinfo, JERR_TOO_LITTLE_DATA);
+    (*cinfo->master->finish_pass) (cinfo);
+  } else if (cinfo->global_state != CSTATE_WRCOEFS)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  /* Perform any remaining passes */
+  while (!cinfo->master->is_last_pass) {
+    (*cinfo->master->prepare_for_pass) (cinfo);
+    for (iMCU_row = 0; iMCU_row < cinfo->total_iMCU_rows; iMCU_row++) {
+      if (cinfo->progress != NULL) {
+        cinfo->progress->pass_counter = (long)iMCU_row;
+        cinfo->progress->pass_limit = (long)cinfo->total_iMCU_rows;
+        (*cinfo->progress->progress_monitor) ((j_common_ptr)cinfo);
+      }
+      /* We bypass the main controller and invoke coef controller directly;
+       * all work is being done from the coefficient buffer.
+       */
+      if (!(*cinfo->coef->compress_data) (cinfo, (JSAMPIMAGE)NULL))
+        ERREXIT(cinfo, JERR_CANT_SUSPEND);
+    }
+    (*cinfo->master->finish_pass) (cinfo);
+  }
+  /* Write EOI, do final cleanup */
+  (*cinfo->marker->write_file_trailer) (cinfo);
+  (*cinfo->dest->term_destination) (cinfo);
+  /* We can use jpeg_abort to release memory and reset global_state */
+  jpeg_abort((j_common_ptr)cinfo);
+}
+
+
+/*
+ * Write a special marker.
+ * This is only recommended for writing COM or APPn markers.
+ * Must be called after jpeg_start_compress() and before
+ * first call to jpeg_write_scanlines() or jpeg_write_raw_data().
+ */
+
+GLOBAL(void)
+jpeg_write_marker(j_compress_ptr cinfo, int marker, const JOCTET *dataptr,
+                  unsigned int datalen)
+{
+  void (*write_marker_byte) (j_compress_ptr info, int val);
+
+  if (cinfo->next_scanline != 0 ||
+      (cinfo->global_state != CSTATE_SCANNING &&
+       cinfo->global_state != CSTATE_RAW_OK &&
+       cinfo->global_state != CSTATE_WRCOEFS))
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  (*cinfo->marker->write_marker_header) (cinfo, marker, datalen);
+  write_marker_byte = cinfo->marker->write_marker_byte; /* copy for speed */
+  while (datalen--) {
+    (*write_marker_byte) (cinfo, *dataptr);
+    dataptr++;
+  }
+}
+
+/* Same, but piecemeal. */
+
+GLOBAL(void)
+jpeg_write_m_header(j_compress_ptr cinfo, int marker, unsigned int datalen)
+{
+  if (cinfo->next_scanline != 0 ||
+      (cinfo->global_state != CSTATE_SCANNING &&
+       cinfo->global_state != CSTATE_RAW_OK &&
+       cinfo->global_state != CSTATE_WRCOEFS))
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  (*cinfo->marker->write_marker_header) (cinfo, marker, datalen);
+}
+
+GLOBAL(void)
+jpeg_write_m_byte(j_compress_ptr cinfo, int val)
+{
+  (*cinfo->marker->write_marker_byte) (cinfo, val);
+}
+
+
+/*
+ * Alternate compression function: just write an abbreviated table file.
+ * Before calling this, all parameters and a data destination must be set up.
+ *
+ * To produce a pair of files containing abbreviated tables and abbreviated
+ * image data, one would proceed as follows:
+ *
+ *              initialize JPEG object
+ *              set JPEG parameters
+ *              set destination to table file
+ *              jpeg_write_tables(cinfo);
+ *              set destination to image file
+ *              jpeg_start_compress(cinfo, FALSE);
+ *              write data...
+ *              jpeg_finish_compress(cinfo);
+ *
+ * jpeg_write_tables has the side effect of marking all tables written
+ * (same as jpeg_suppress_tables(..., TRUE)).  Thus a subsequent start_compress
+ * will not re-emit the tables unless it is passed write_all_tables=TRUE.
+ */
+
+GLOBAL(void)
+jpeg_write_tables(j_compress_ptr cinfo)
+{
+  if (cinfo->global_state != CSTATE_START)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  /* (Re)initialize error mgr and destination modules */
+  (*cinfo->err->reset_error_mgr) ((j_common_ptr)cinfo);
+  (*cinfo->dest->init_destination) (cinfo);
+  /* Initialize the marker writer ... bit of a crock to do it here. */
+  jinit_marker_writer(cinfo);
+  /* Write them tables! */
+  (*cinfo->marker->write_tables_only) (cinfo);
+  /* And clean up. */
+  (*cinfo->dest->term_destination) (cinfo);
+  /*
+   * In library releases up through v6a, we called jpeg_abort() here to free
+   * any working memory allocated by the destination manager and marker
+   * writer.  Some applications had a problem with that: they allocated space
+   * of their own from the library memory manager, and didn't want it to go
+   * away during write_tables.  So now we do nothing.  This will cause a
+   * memory leak if an app calls write_tables repeatedly without doing a full
+   * compression cycle or otherwise resetting the JPEG object.  However, that
+   * seems less bad than unexpectedly freeing memory in the normal case.
+   * An app that prefers the old behavior can call jpeg_abort for itself after
+   * each call to jpeg_write_tables().
+   */
+}

3rdparty/libjpeg-turbo/src/jcapistd.c

@@ -0,0 +1,162 @@
+/*
+ * jcapistd.c
+ *
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * This file is part of the Independent JPEG Group's software.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains application interface code for the compression half
+ * of the JPEG library.  These are the "standard" API routines that are
+ * used in the normal full-compression case.  They are not used by a
+ * transcoding-only application.  Note that if an application links in
+ * jpeg_start_compress, it will end up linking in the entire compressor.
+ * We thus must separate this file from jcapimin.c to avoid linking the
+ * whole compression library into a transcoder.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/*
+ * Compression initialization.
+ * Before calling this, all parameters and a data destination must be set up.
+ *
+ * We require a write_all_tables parameter as a failsafe check when writing
+ * multiple datastreams from the same compression object.  Since prior runs
+ * will have left all the tables marked sent_table=TRUE, a subsequent run
+ * would emit an abbreviated stream (no tables) by default.  This may be what
+ * is wanted, but for safety's sake it should not be the default behavior:
+ * programmers should have to make a deliberate choice to emit abbreviated
+ * images.  Therefore the documentation and examples should encourage people
+ * to pass write_all_tables=TRUE; then it will take active thought to do the
+ * wrong thing.
+ */
+
+GLOBAL(void)
+jpeg_start_compress(j_compress_ptr cinfo, boolean write_all_tables)
+{
+  if (cinfo->global_state != CSTATE_START)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  if (write_all_tables)
+    jpeg_suppress_tables(cinfo, FALSE); /* mark all tables to be written */
+
+  /* (Re)initialize error mgr and destination modules */
+  (*cinfo->err->reset_error_mgr) ((j_common_ptr)cinfo);
+  (*cinfo->dest->init_destination) (cinfo);
+  /* Perform master selection of active modules */
+  jinit_compress_master(cinfo);
+  /* Set up for the first pass */
+  (*cinfo->master->prepare_for_pass) (cinfo);
+  /* Ready for application to drive first pass through jpeg_write_scanlines
+   * or jpeg_write_raw_data.
+   */
+  cinfo->next_scanline = 0;
+  cinfo->global_state = (cinfo->raw_data_in ? CSTATE_RAW_OK : CSTATE_SCANNING);
+}
+
+
+/*
+ * Write some scanlines of data to the JPEG compressor.
+ *
+ * The return value will be the number of lines actually written.
+ * This should be less than the supplied num_lines only in case that
+ * the data destination module has requested suspension of the compressor,
+ * or if more than image_height scanlines are passed in.
+ *
+ * Note: we warn about excess calls to jpeg_write_scanlines() since
+ * this likely signals an application programmer error.  However,
+ * excess scanlines passed in the last valid call are *silently* ignored,
+ * so that the application need not adjust num_lines for end-of-image
+ * when using a multiple-scanline buffer.
+ */
+
+GLOBAL(JDIMENSION)
+jpeg_write_scanlines(j_compress_ptr cinfo, JSAMPARRAY scanlines,
+                     JDIMENSION num_lines)
+{
+  JDIMENSION row_ctr, rows_left;
+
+  if (cinfo->global_state != CSTATE_SCANNING)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  if (cinfo->next_scanline >= cinfo->image_height)
+    WARNMS(cinfo, JWRN_TOO_MUCH_DATA);
+
+  /* Call progress monitor hook if present */
+  if (cinfo->progress != NULL) {
+    cinfo->progress->pass_counter = (long)cinfo->next_scanline;
+    cinfo->progress->pass_limit = (long)cinfo->image_height;
+    (*cinfo->progress->progress_monitor) ((j_common_ptr)cinfo);
+  }
+
+  /* Give master control module another chance if this is first call to
+   * jpeg_write_scanlines.  This lets output of the frame/scan headers be
+   * delayed so that application can write COM, etc, markers between
+   * jpeg_start_compress and jpeg_write_scanlines.
+   */
+  if (cinfo->master->call_pass_startup)
+    (*cinfo->master->pass_startup) (cinfo);
+
+  /* Ignore any extra scanlines at bottom of image. */
+  rows_left = cinfo->image_height - cinfo->next_scanline;
+  if (num_lines > rows_left)
+    num_lines = rows_left;
+
+  row_ctr = 0;
+  (*cinfo->main->process_data) (cinfo, scanlines, &row_ctr, num_lines);
+  cinfo->next_scanline += row_ctr;
+  return row_ctr;
+}
+
+
+/*
+ * Alternate entry point to write raw data.
+ * Processes exactly one iMCU row per call, unless suspended.
+ */
+
+GLOBAL(JDIMENSION)
+jpeg_write_raw_data(j_compress_ptr cinfo, JSAMPIMAGE data,
+                    JDIMENSION num_lines)
+{
+  JDIMENSION lines_per_iMCU_row;
+
+  if (cinfo->global_state != CSTATE_RAW_OK)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  if (cinfo->next_scanline >= cinfo->image_height) {
+    WARNMS(cinfo, JWRN_TOO_MUCH_DATA);
+    return 0;
+  }
+
+  /* Call progress monitor hook if present */
+  if (cinfo->progress != NULL) {
+    cinfo->progress->pass_counter = (long)cinfo->next_scanline;
+    cinfo->progress->pass_limit = (long)cinfo->image_height;
+    (*cinfo->progress->progress_monitor) ((j_common_ptr)cinfo);
+  }
+
+  /* Give master control module another chance if this is first call to
+   * jpeg_write_raw_data.  This lets output of the frame/scan headers be
+   * delayed so that application can write COM, etc, markers between
+   * jpeg_start_compress and jpeg_write_raw_data.
+   */
+  if (cinfo->master->call_pass_startup)
+    (*cinfo->master->pass_startup) (cinfo);
+
+  /* Verify that at least one iMCU row has been passed. */
+  lines_per_iMCU_row = cinfo->max_v_samp_factor * DCTSIZE;
+  if (num_lines < lines_per_iMCU_row)
+    ERREXIT(cinfo, JERR_BUFFER_SIZE);
+
+  /* Directly compress the row. */
+  if (!(*cinfo->coef->compress_data) (cinfo, data)) {
+    /* If compressor did not consume the whole row, suspend processing. */
+    return 0;
+  }
+
+  /* OK, we processed one iMCU row. */
+  cinfo->next_scanline += lines_per_iMCU_row;
+  return lines_per_iMCU_row;
+}

3rdparty/libjpeg-turbo/src/jcarith.c

@@ -0,0 +1,932 @@
+/*
+ * jcarith.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Developed 1997-2009 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2015, 2018, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains portable arithmetic entropy encoding routines for JPEG
+ * (implementing Recommendation ITU-T T.81 | ISO/IEC 10918-1).
+ *
+ * Both sequential and progressive modes are supported in this single module.
+ *
+ * Suspension is not currently supported in this module.
+ *
+ * NOTE: All referenced figures are from
+ * Recommendation ITU-T T.81 (1992) | ISO/IEC 10918-1:1994.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/* Expanded entropy encoder object for arithmetic encoding. */
+
+typedef struct {
+  struct jpeg_entropy_encoder pub; /* public fields */
+
+  JLONG c; /* C register, base of coding interval, layout as in sec. D.1.3 */
+  JLONG a;               /* A register, normalized size of coding interval */
+  JLONG sc;        /* counter for stacked 0xFF values which might overflow */
+  JLONG zc;          /* counter for pending 0x00 output values which might *
+                          * be discarded at the end ("Pacman" termination) */
+  int ct;  /* bit shift counter, determines when next byte will be written */
+  int buffer;                /* buffer for most recent output byte != 0xFF */
+
+  int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */
+  int dc_context[MAX_COMPS_IN_SCAN]; /* context index for DC conditioning */
+
+  unsigned int restarts_to_go;  /* MCUs left in this restart interval */
+  int next_restart_num;         /* next restart number to write (0-7) */
+
+  /* Pointers to statistics areas (these workspaces have image lifespan) */
+  unsigned char *dc_stats[NUM_ARITH_TBLS];
+  unsigned char *ac_stats[NUM_ARITH_TBLS];
+
+  /* Statistics bin for coding with fixed probability 0.5 */
+  unsigned char fixed_bin[4];
+} arith_entropy_encoder;
+
+typedef arith_entropy_encoder *arith_entropy_ptr;
+
+/* The following two definitions specify the allocation chunk size
+ * for the statistics area.
+ * According to sections F.1.4.4.1.3 and F.1.4.4.2, we need at least
+ * 49 statistics bins for DC, and 245 statistics bins for AC coding.
+ *
+ * We use a compact representation with 1 byte per statistics bin,
+ * thus the numbers directly represent byte sizes.
+ * This 1 byte per statistics bin contains the meaning of the MPS
+ * (more probable symbol) in the highest bit (mask 0x80), and the
+ * index into the probability estimation state machine table
+ * in the lower bits (mask 0x7F).
+ */
+
+#define DC_STAT_BINS  64
+#define AC_STAT_BINS  256
+
+/* NOTE: Uncomment the following #define if you want to use the
+ * given formula for calculating the AC conditioning parameter Kx
+ * for spectral selection progressive coding in section G.1.3.2
+ * of the spec (Kx = Kmin + SRL (8 + Se - Kmin) 4).
+ * Although the spec and P&M authors claim that this "has proven
+ * to give good results for 8 bit precision samples", I'm not
+ * convinced yet that this is really beneficial.
+ * Early tests gave only very marginal compression enhancements
+ * (a few - around 5 or so - bytes even for very large files),
+ * which would turn out rather negative if we'd suppress the
+ * DAC (Define Arithmetic Conditioning) marker segments for
+ * the default parameters in the future.
+ * Note that currently the marker writing module emits 12-byte
+ * DAC segments for a full-component scan in a color image.
+ * This is not worth worrying about IMHO. However, since the
+ * spec defines the default values to be used if the tables
+ * are omitted (unlike Huffman tables, which are required
+ * anyway), one might optimize this behaviour in the future,
+ * and then it would be disadvantageous to use custom tables if
+ * they don't provide sufficient gain to exceed the DAC size.
+ *
+ * On the other hand, I'd consider it as a reasonable result
+ * that the conditioning has no significant influence on the
+ * compression performance. This means that the basic
+ * statistical model is already rather stable.
+ *
+ * Thus, at the moment, we use the default conditioning values
+ * anyway, and do not use the custom formula.
+ *
+#define CALCULATE_SPECTRAL_CONDITIONING
+ */
+
+/* IRIGHT_SHIFT is like RIGHT_SHIFT, but works on int rather than JLONG.
+ * We assume that int right shift is unsigned if JLONG right shift is,
+ * which should be safe.
+ */
+
+#ifdef RIGHT_SHIFT_IS_UNSIGNED
+#define ISHIFT_TEMPS    int ishift_temp;
+#define IRIGHT_SHIFT(x, shft) \
+  ((ishift_temp = (x)) < 0 ? \
+   (ishift_temp >> (shft)) | ((~0) << (16 - (shft))) : \
+   (ishift_temp >> (shft)))
+#else
+#define ISHIFT_TEMPS
+#define IRIGHT_SHIFT(x, shft)   ((x) >> (shft))
+#endif
+
+
+LOCAL(void)
+emit_byte(int val, j_compress_ptr cinfo)
+/* Write next output byte; we do not support suspension in this module. */
+{
+  struct jpeg_destination_mgr *dest = cinfo->dest;
+
+  *dest->next_output_byte++ = (JOCTET)val;
+  if (--dest->free_in_buffer == 0)
+    if (!(*dest->empty_output_buffer) (cinfo))
+      ERREXIT(cinfo, JERR_CANT_SUSPEND);
+}
+
+
+/*
+ * Finish up at the end of an arithmetic-compressed scan.
+ */
+
+METHODDEF(void)
+finish_pass(j_compress_ptr cinfo)
+{
+  arith_entropy_ptr e = (arith_entropy_ptr)cinfo->entropy;
+  JLONG temp;
+
+  /* Section D.1.8: Termination of encoding */
+
+  /* Find the e->c in the coding interval with the largest
+   * number of trailing zero bits */
+  if ((temp = (e->a - 1 + e->c) & 0xFFFF0000UL) < e->c)
+    e->c = temp + 0x8000L;
+  else
+    e->c = temp;
+  /* Send remaining bytes to output */
+  e->c <<= e->ct;
+  if (e->c & 0xF8000000UL) {
+    /* One final overflow has to be handled */
+    if (e->buffer >= 0) {
+      if (e->zc)
+        do emit_byte(0x00, cinfo);
+        while (--e->zc);
+      emit_byte(e->buffer + 1, cinfo);
+      if (e->buffer + 1 == 0xFF)
+        emit_byte(0x00, cinfo);
+    }
+    e->zc += e->sc;  /* carry-over converts stacked 0xFF bytes to 0x00 */
+    e->sc = 0;
+  } else {
+    if (e->buffer == 0)
+      ++e->zc;
+    else if (e->buffer >= 0) {
+      if (e->zc)
+        do emit_byte(0x00, cinfo);
+        while (--e->zc);
+      emit_byte(e->buffer, cinfo);
+    }
+    if (e->sc) {
+      if (e->zc)
+        do emit_byte(0x00, cinfo);
+        while (--e->zc);
+      do {
+        emit_byte(0xFF, cinfo);
+        emit_byte(0x00, cinfo);
+      } while (--e->sc);
+    }
+  }
+  /* Output final bytes only if they are not 0x00 */
+  if (e->c & 0x7FFF800L) {
+    if (e->zc)  /* output final pending zero bytes */
+      do emit_byte(0x00, cinfo);
+      while (--e->zc);
+    emit_byte((e->c >> 19) & 0xFF, cinfo);
+    if (((e->c >> 19) & 0xFF) == 0xFF)
+      emit_byte(0x00, cinfo);
+    if (e->c & 0x7F800L) {
+      emit_byte((e->c >> 11) & 0xFF, cinfo);
+      if (((e->c >> 11) & 0xFF) == 0xFF)
+        emit_byte(0x00, cinfo);
+    }
+  }
+}
+
+
+/*
+ * The core arithmetic encoding routine (common in JPEG and JBIG).
+ * This needs to go as fast as possible.
+ * Machine-dependent optimization facilities
+ * are not utilized in this portable implementation.
+ * However, this code should be fairly efficient and
+ * may be a good base for further optimizations anyway.
+ *
+ * Parameter 'val' to be encoded may be 0 or 1 (binary decision).
+ *
+ * Note: I've added full "Pacman" termination support to the
+ * byte output routines, which is equivalent to the optional
+ * Discard_final_zeros procedure (Figure D.15) in the spec.
+ * Thus, we always produce the shortest possible output
+ * stream compliant to the spec (no trailing zero bytes,
+ * except for FF stuffing).
+ *
+ * I've also introduced a new scheme for accessing
+ * the probability estimation state machine table,
+ * derived from Markus Kuhn's JBIG implementation.
+ */
+
+LOCAL(void)
+arith_encode(j_compress_ptr cinfo, unsigned char *st, int val)
+{
+  register arith_entropy_ptr e = (arith_entropy_ptr)cinfo->entropy;
+  register unsigned char nl, nm;
+  register JLONG qe, temp;
+  register int sv;
+
+  /* Fetch values from our compact representation of Table D.2:
+   * Qe values and probability estimation state machine
+   */
+  sv = *st;
+  qe = jpeg_aritab[sv & 0x7F];  /* => Qe_Value */
+  nl = qe & 0xFF;  qe >>= 8;    /* Next_Index_LPS + Switch_MPS */
+  nm = qe & 0xFF;  qe >>= 8;    /* Next_Index_MPS */
+
+  /* Encode & estimation procedures per sections D.1.4 & D.1.5 */
+  e->a -= qe;
+  if (val != (sv >> 7)) {
+    /* Encode the less probable symbol */
+    if (e->a >= qe) {
+      /* If the interval size (qe) for the less probable symbol (LPS)
+       * is larger than the interval size for the MPS, then exchange
+       * the two symbols for coding efficiency, otherwise code the LPS
+       * as usual: */
+      e->c += e->a;
+      e->a = qe;
+    }
+    *st = (sv & 0x80) ^ nl;     /* Estimate_after_LPS */
+  } else {
+    /* Encode the more probable symbol */
+    if (e->a >= 0x8000L)
+      return;  /* A >= 0x8000 -> ready, no renormalization required */
+    if (e->a < qe) {
+      /* If the interval size (qe) for the less probable symbol (LPS)
+       * is larger than the interval size for the MPS, then exchange
+       * the two symbols for coding efficiency: */
+      e->c += e->a;
+      e->a = qe;
+    }
+    *st = (sv & 0x80) ^ nm;     /* Estimate_after_MPS */
+  }
+
+  /* Renormalization & data output per section D.1.6 */
+  do {
+    e->a <<= 1;
+    e->c <<= 1;
+    if (--e->ct == 0) {
+      /* Another byte is ready for output */
+      temp = e->c >> 19;
+      if (temp > 0xFF) {
+        /* Handle overflow over all stacked 0xFF bytes */
+        if (e->buffer >= 0) {
+          if (e->zc)
+            do emit_byte(0x00, cinfo);
+            while (--e->zc);
+          emit_byte(e->buffer + 1, cinfo);
+          if (e->buffer + 1 == 0xFF)
+            emit_byte(0x00, cinfo);
+        }
+        e->zc += e->sc;  /* carry-over converts stacked 0xFF bytes to 0x00 */
+        e->sc = 0;
+        /* Note: The 3 spacer bits in the C register guarantee
+         * that the new buffer byte can't be 0xFF here
+         * (see page 160 in the P&M JPEG book). */
+        e->buffer = temp & 0xFF;  /* new output byte, might overflow later */
+      } else if (temp == 0xFF) {
+        ++e->sc;  /* stack 0xFF byte (which might overflow later) */
+      } else {
+        /* Output all stacked 0xFF bytes, they will not overflow any more */
+        if (e->buffer == 0)
+          ++e->zc;
+        else if (e->buffer >= 0) {
+          if (e->zc)
+            do emit_byte(0x00, cinfo);
+            while (--e->zc);
+          emit_byte(e->buffer, cinfo);
+        }
+        if (e->sc) {
+          if (e->zc)
+            do emit_byte(0x00, cinfo);
+            while (--e->zc);
+          do {
+            emit_byte(0xFF, cinfo);
+            emit_byte(0x00, cinfo);
+          } while (--e->sc);
+        }
+        e->buffer = temp & 0xFF;  /* new output byte (can still overflow) */
+      }
+      e->c &= 0x7FFFFL;
+      e->ct += 8;
+    }
+  } while (e->a < 0x8000L);
+}
+
+
+/*
+ * Emit a restart marker & resynchronize predictions.
+ */
+
+LOCAL(void)
+emit_restart(j_compress_ptr cinfo, int restart_num)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  int ci;
+  jpeg_component_info *compptr;
+
+  finish_pass(cinfo);
+
+  emit_byte(0xFF, cinfo);
+  emit_byte(JPEG_RST0 + restart_num, cinfo);
+
+  /* Re-initialize statistics areas */
+  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+    compptr = cinfo->cur_comp_info[ci];
+    /* DC needs no table for refinement scan */
+    if (cinfo->progressive_mode == 0 || (cinfo->Ss == 0 && cinfo->Ah == 0)) {
+      MEMZERO(entropy->dc_stats[compptr->dc_tbl_no], DC_STAT_BINS);
+      /* Reset DC predictions to 0 */
+      entropy->last_dc_val[ci] = 0;
+      entropy->dc_context[ci] = 0;
+    }
+    /* AC needs no table when not present */
+    if (cinfo->progressive_mode == 0 || cinfo->Se) {
+      MEMZERO(entropy->ac_stats[compptr->ac_tbl_no], AC_STAT_BINS);
+    }
+  }
+
+  /* Reset arithmetic encoding variables */
+  entropy->c = 0;
+  entropy->a = 0x10000L;
+  entropy->sc = 0;
+  entropy->zc = 0;
+  entropy->ct = 11;
+  entropy->buffer = -1;  /* empty */
+}
+
+
+/*
+ * MCU encoding for DC initial scan (either spectral selection,
+ * or first pass of successive approximation).
+ */
+
+METHODDEF(boolean)
+encode_mcu_DC_first(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  JBLOCKROW block;
+  unsigned char *st;
+  int blkn, ci, tbl;
+  int v, v2, m;
+  ISHIFT_TEMPS
+
+  /* Emit restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0) {
+      emit_restart(cinfo, entropy->next_restart_num);
+      entropy->restarts_to_go = cinfo->restart_interval;
+      entropy->next_restart_num++;
+      entropy->next_restart_num &= 7;
+    }
+    entropy->restarts_to_go--;
+  }
+
+  /* Encode the MCU data blocks */
+  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
+    block = MCU_data[blkn];
+    ci = cinfo->MCU_membership[blkn];
+    tbl = cinfo->cur_comp_info[ci]->dc_tbl_no;
+
+    /* Compute the DC value after the required point transform by Al.
+     * This is simply an arithmetic right shift.
+     */
+    m = IRIGHT_SHIFT((int)((*block)[0]), cinfo->Al);
+
+    /* Sections F.1.4.1 & F.1.4.4.1: Encoding of DC coefficients */
+
+    /* Table F.4: Point to statistics bin S0 for DC coefficient coding */
+    st = entropy->dc_stats[tbl] + entropy->dc_context[ci];
+
+    /* Figure F.4: Encode_DC_DIFF */
+    if ((v = m - entropy->last_dc_val[ci]) == 0) {
+      arith_encode(cinfo, st, 0);
+      entropy->dc_context[ci] = 0;      /* zero diff category */
+    } else {
+      entropy->last_dc_val[ci] = m;
+      arith_encode(cinfo, st, 1);
+      /* Figure F.6: Encoding nonzero value v */
+      /* Figure F.7: Encoding the sign of v */
+      if (v > 0) {
+        arith_encode(cinfo, st + 1, 0); /* Table F.4: SS = S0 + 1 */
+        st += 2;                        /* Table F.4: SP = S0 + 2 */
+        entropy->dc_context[ci] = 4;    /* small positive diff category */
+      } else {
+        v = -v;
+        arith_encode(cinfo, st + 1, 1); /* Table F.4: SS = S0 + 1 */
+        st += 3;                        /* Table F.4: SN = S0 + 3 */
+        entropy->dc_context[ci] = 8;    /* small negative diff category */
+      }
+      /* Figure F.8: Encoding the magnitude category of v */
+      m = 0;
+      if (v -= 1) {
+        arith_encode(cinfo, st, 1);
+        m = 1;
+        v2 = v;
+        st = entropy->dc_stats[tbl] + 20; /* Table F.4: X1 = 20 */
+        while (v2 >>= 1) {
+          arith_encode(cinfo, st, 1);
+          m <<= 1;
+          st += 1;
+        }
+      }
+      arith_encode(cinfo, st, 0);
+      /* Section F.1.4.4.1.2: Establish dc_context conditioning category */
+      if (m < (int)((1L << cinfo->arith_dc_L[tbl]) >> 1))
+        entropy->dc_context[ci] = 0;    /* zero diff category */
+      else if (m > (int)((1L << cinfo->arith_dc_U[tbl]) >> 1))
+        entropy->dc_context[ci] += 8;   /* large diff category */
+      /* Figure F.9: Encoding the magnitude bit pattern of v */
+      st += 14;
+      while (m >>= 1)
+        arith_encode(cinfo, st, (m & v) ? 1 : 0);
+    }
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * MCU encoding for AC initial scan (either spectral selection,
+ * or first pass of successive approximation).
+ */
+
+METHODDEF(boolean)
+encode_mcu_AC_first(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  JBLOCKROW block;
+  unsigned char *st;
+  int tbl, k, ke;
+  int v, v2, m;
+
+  /* Emit restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0) {
+      emit_restart(cinfo, entropy->next_restart_num);
+      entropy->restarts_to_go = cinfo->restart_interval;
+      entropy->next_restart_num++;
+      entropy->next_restart_num &= 7;
+    }
+    entropy->restarts_to_go--;
+  }
+
+  /* Encode the MCU data block */
+  block = MCU_data[0];
+  tbl = cinfo->cur_comp_info[0]->ac_tbl_no;
+
+  /* Sections F.1.4.2 & F.1.4.4.2: Encoding of AC coefficients */
+
+  /* Establish EOB (end-of-block) index */
+  for (ke = cinfo->Se; ke > 0; ke--)
+    /* We must apply the point transform by Al.  For AC coefficients this
+     * is an integer division with rounding towards 0.  To do this portably
+     * in C, we shift after obtaining the absolute value.
+     */
+    if ((v = (*block)[jpeg_natural_order[ke]]) >= 0) {
+      if (v >>= cinfo->Al) break;
+    } else {
+      v = -v;
+      if (v >>= cinfo->Al) break;
+    }
+
+  /* Figure F.5: Encode_AC_Coefficients */
+  for (k = cinfo->Ss; k <= ke; k++) {
+    st = entropy->ac_stats[tbl] + 3 * (k - 1);
+    arith_encode(cinfo, st, 0);         /* EOB decision */
+    for (;;) {
+      if ((v = (*block)[jpeg_natural_order[k]]) >= 0) {
+        if (v >>= cinfo->Al) {
+          arith_encode(cinfo, st + 1, 1);
+          arith_encode(cinfo, entropy->fixed_bin, 0);
+          break;
+        }
+      } else {
+        v = -v;
+        if (v >>= cinfo->Al) {
+          arith_encode(cinfo, st + 1, 1);
+          arith_encode(cinfo, entropy->fixed_bin, 1);
+          break;
+        }
+      }
+      arith_encode(cinfo, st + 1, 0);  st += 3;  k++;
+    }
+    st += 2;
+    /* Figure F.8: Encoding the magnitude category of v */
+    m = 0;
+    if (v -= 1) {
+      arith_encode(cinfo, st, 1);
+      m = 1;
+      v2 = v;
+      if (v2 >>= 1) {
+        arith_encode(cinfo, st, 1);
+        m <<= 1;
+        st = entropy->ac_stats[tbl] +
+             (k <= cinfo->arith_ac_K[tbl] ? 189 : 217);
+        while (v2 >>= 1) {
+          arith_encode(cinfo, st, 1);
+          m <<= 1;
+          st += 1;
+        }
+      }
+    }
+    arith_encode(cinfo, st, 0);
+    /* Figure F.9: Encoding the magnitude bit pattern of v */
+    st += 14;
+    while (m >>= 1)
+      arith_encode(cinfo, st, (m & v) ? 1 : 0);
+  }
+  /* Encode EOB decision only if k <= cinfo->Se */
+  if (k <= cinfo->Se) {
+    st = entropy->ac_stats[tbl] + 3 * (k - 1);
+    arith_encode(cinfo, st, 1);
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * MCU encoding for DC successive approximation refinement scan.
+ */
+
+METHODDEF(boolean)
+encode_mcu_DC_refine(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  unsigned char *st;
+  int Al, blkn;
+
+  /* Emit restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0) {
+      emit_restart(cinfo, entropy->next_restart_num);
+      entropy->restarts_to_go = cinfo->restart_interval;
+      entropy->next_restart_num++;
+      entropy->next_restart_num &= 7;
+    }
+    entropy->restarts_to_go--;
+  }
+
+  st = entropy->fixed_bin;      /* use fixed probability estimation */
+  Al = cinfo->Al;
+
+  /* Encode the MCU data blocks */
+  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
+    /* We simply emit the Al'th bit of the DC coefficient value. */
+    arith_encode(cinfo, st, (MCU_data[blkn][0][0] >> Al) & 1);
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * MCU encoding for AC successive approximation refinement scan.
+ */
+
+METHODDEF(boolean)
+encode_mcu_AC_refine(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  JBLOCKROW block;
+  unsigned char *st;
+  int tbl, k, ke, kex;
+  int v;
+
+  /* Emit restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0) {
+      emit_restart(cinfo, entropy->next_restart_num);
+      entropy->restarts_to_go = cinfo->restart_interval;
+      entropy->next_restart_num++;
+      entropy->next_restart_num &= 7;
+    }
+    entropy->restarts_to_go--;
+  }
+
+  /* Encode the MCU data block */
+  block = MCU_data[0];
+  tbl = cinfo->cur_comp_info[0]->ac_tbl_no;
+
+  /* Section G.1.3.3: Encoding of AC coefficients */
+
+  /* Establish EOB (end-of-block) index */
+  for (ke = cinfo->Se; ke > 0; ke--)
+    /* We must apply the point transform by Al.  For AC coefficients this
+     * is an integer division with rounding towards 0.  To do this portably
+     * in C, we shift after obtaining the absolute value.
+     */
+    if ((v = (*block)[jpeg_natural_order[ke]]) >= 0) {
+      if (v >>= cinfo->Al) break;
+    } else {
+      v = -v;
+      if (v >>= cinfo->Al) break;
+    }
+
+  /* Establish EOBx (previous stage end-of-block) index */
+  for (kex = ke; kex > 0; kex--)
+    if ((v = (*block)[jpeg_natural_order[kex]]) >= 0) {
+      if (v >>= cinfo->Ah) break;
+    } else {
+      v = -v;
+      if (v >>= cinfo->Ah) break;
+    }
+
+  /* Figure G.10: Encode_AC_Coefficients_SA */
+  for (k = cinfo->Ss; k <= ke; k++) {
+    st = entropy->ac_stats[tbl] + 3 * (k - 1);
+    if (k > kex)
+      arith_encode(cinfo, st, 0);       /* EOB decision */
+    for (;;) {
+      if ((v = (*block)[jpeg_natural_order[k]]) >= 0) {
+        if (v >>= cinfo->Al) {
+          if (v >> 1)                   /* previously nonzero coef */
+            arith_encode(cinfo, st + 2, (v & 1));
+          else {                        /* newly nonzero coef */
+            arith_encode(cinfo, st + 1, 1);
+            arith_encode(cinfo, entropy->fixed_bin, 0);
+          }
+          break;
+        }
+      } else {
+        v = -v;
+        if (v >>= cinfo->Al) {
+          if (v >> 1)                   /* previously nonzero coef */
+            arith_encode(cinfo, st + 2, (v & 1));
+          else {                        /* newly nonzero coef */
+            arith_encode(cinfo, st + 1, 1);
+            arith_encode(cinfo, entropy->fixed_bin, 1);
+          }
+          break;
+        }
+      }
+      arith_encode(cinfo, st + 1, 0);  st += 3;  k++;
+    }
+  }
+  /* Encode EOB decision only if k <= cinfo->Se */
+  if (k <= cinfo->Se) {
+    st = entropy->ac_stats[tbl] + 3 * (k - 1);
+    arith_encode(cinfo, st, 1);
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * Encode and output one MCU's worth of arithmetic-compressed coefficients.
+ */
+
+METHODDEF(boolean)
+encode_mcu(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  jpeg_component_info *compptr;
+  JBLOCKROW block;
+  unsigned char *st;
+  int blkn, ci, tbl, k, ke;
+  int v, v2, m;
+
+  /* Emit restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0) {
+      emit_restart(cinfo, entropy->next_restart_num);
+      entropy->restarts_to_go = cinfo->restart_interval;
+      entropy->next_restart_num++;
+      entropy->next_restart_num &= 7;
+    }
+    entropy->restarts_to_go--;
+  }
+
+  /* Encode the MCU data blocks */
+  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
+    block = MCU_data[blkn];
+    ci = cinfo->MCU_membership[blkn];
+    compptr = cinfo->cur_comp_info[ci];
+
+    /* Sections F.1.4.1 & F.1.4.4.1: Encoding of DC coefficients */
+
+    tbl = compptr->dc_tbl_no;
+
+    /* Table F.4: Point to statistics bin S0 for DC coefficient coding */
+    st = entropy->dc_stats[tbl] + entropy->dc_context[ci];
+
+    /* Figure F.4: Encode_DC_DIFF */
+    if ((v = (*block)[0] - entropy->last_dc_val[ci]) == 0) {
+      arith_encode(cinfo, st, 0);
+      entropy->dc_context[ci] = 0;      /* zero diff category */
+    } else {
+      entropy->last_dc_val[ci] = (*block)[0];
+      arith_encode(cinfo, st, 1);
+      /* Figure F.6: Encoding nonzero value v */
+      /* Figure F.7: Encoding the sign of v */
+      if (v > 0) {
+        arith_encode(cinfo, st + 1, 0); /* Table F.4: SS = S0 + 1 */
+        st += 2;                        /* Table F.4: SP = S0 + 2 */
+        entropy->dc_context[ci] = 4;    /* small positive diff category */
+      } else {
+        v = -v;
+        arith_encode(cinfo, st + 1, 1); /* Table F.4: SS = S0 + 1 */
+        st += 3;                        /* Table F.4: SN = S0 + 3 */
+        entropy->dc_context[ci] = 8;    /* small negative diff category */
+      }
+      /* Figure F.8: Encoding the magnitude category of v */
+      m = 0;
+      if (v -= 1) {
+        arith_encode(cinfo, st, 1);
+        m = 1;
+        v2 = v;
+        st = entropy->dc_stats[tbl] + 20; /* Table F.4: X1 = 20 */
+        while (v2 >>= 1) {
+          arith_encode(cinfo, st, 1);
+          m <<= 1;
+          st += 1;
+        }
+      }
+      arith_encode(cinfo, st, 0);
+      /* Section F.1.4.4.1.2: Establish dc_context conditioning category */
+      if (m < (int)((1L << cinfo->arith_dc_L[tbl]) >> 1))
+        entropy->dc_context[ci] = 0;    /* zero diff category */
+      else if (m > (int)((1L << cinfo->arith_dc_U[tbl]) >> 1))
+        entropy->dc_context[ci] += 8;   /* large diff category */
+      /* Figure F.9: Encoding the magnitude bit pattern of v */
+      st += 14;
+      while (m >>= 1)
+        arith_encode(cinfo, st, (m & v) ? 1 : 0);
+    }
+
+    /* Sections F.1.4.2 & F.1.4.4.2: Encoding of AC coefficients */
+
+    tbl = compptr->ac_tbl_no;
+
+    /* Establish EOB (end-of-block) index */
+    for (ke = DCTSIZE2 - 1; ke > 0; ke--)
+      if ((*block)[jpeg_natural_order[ke]]) break;
+
+    /* Figure F.5: Encode_AC_Coefficients */
+    for (k = 1; k <= ke; k++) {
+      st = entropy->ac_stats[tbl] + 3 * (k - 1);
+      arith_encode(cinfo, st, 0);       /* EOB decision */
+      while ((v = (*block)[jpeg_natural_order[k]]) == 0) {
+        arith_encode(cinfo, st + 1, 0);  st += 3;  k++;
+      }
+      arith_encode(cinfo, st + 1, 1);
+      /* Figure F.6: Encoding nonzero value v */
+      /* Figure F.7: Encoding the sign of v */
+      if (v > 0) {
+        arith_encode(cinfo, entropy->fixed_bin, 0);
+      } else {
+        v = -v;
+        arith_encode(cinfo, entropy->fixed_bin, 1);
+      }
+      st += 2;
+      /* Figure F.8: Encoding the magnitude category of v */
+      m = 0;
+      if (v -= 1) {
+        arith_encode(cinfo, st, 1);
+        m = 1;
+        v2 = v;
+        if (v2 >>= 1) {
+          arith_encode(cinfo, st, 1);
+          m <<= 1;
+          st = entropy->ac_stats[tbl] +
+               (k <= cinfo->arith_ac_K[tbl] ? 189 : 217);
+          while (v2 >>= 1) {
+            arith_encode(cinfo, st, 1);
+            m <<= 1;
+            st += 1;
+          }
+        }
+      }
+      arith_encode(cinfo, st, 0);
+      /* Figure F.9: Encoding the magnitude bit pattern of v */
+      st += 14;
+      while (m >>= 1)
+        arith_encode(cinfo, st, (m & v) ? 1 : 0);
+    }
+    /* Encode EOB decision only if k <= DCTSIZE2 - 1 */
+    if (k <= DCTSIZE2 - 1) {
+      st = entropy->ac_stats[tbl] + 3 * (k - 1);
+      arith_encode(cinfo, st, 1);
+    }
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * Initialize for an arithmetic-compressed scan.
+ */
+
+METHODDEF(void)
+start_pass(j_compress_ptr cinfo, boolean gather_statistics)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  int ci, tbl;
+  jpeg_component_info *compptr;
+
+  if (gather_statistics)
+    /* Make sure to avoid that in the master control logic!
+     * We are fully adaptive here and need no extra
+     * statistics gathering pass!
+     */
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+
+  /* We assume jcmaster.c already validated the progressive scan parameters. */
+
+  /* Select execution routines */
+  if (cinfo->progressive_mode) {
+    if (cinfo->Ah == 0) {
+      if (cinfo->Ss == 0)
+        entropy->pub.encode_mcu = encode_mcu_DC_first;
+      else
+        entropy->pub.encode_mcu = encode_mcu_AC_first;
+    } else {
+      if (cinfo->Ss == 0)
+        entropy->pub.encode_mcu = encode_mcu_DC_refine;
+      else
+        entropy->pub.encode_mcu = encode_mcu_AC_refine;
+    }
+  } else
+    entropy->pub.encode_mcu = encode_mcu;
+
+  /* Allocate & initialize requested statistics areas */
+  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+    compptr = cinfo->cur_comp_info[ci];
+    /* DC needs no table for refinement scan */
+    if (cinfo->progressive_mode == 0 || (cinfo->Ss == 0 && cinfo->Ah == 0)) {
+      tbl = compptr->dc_tbl_no;
+      if (tbl < 0 || tbl >= NUM_ARITH_TBLS)
+        ERREXIT1(cinfo, JERR_NO_ARITH_TABLE, tbl);
+      if (entropy->dc_stats[tbl] == NULL)
+        entropy->dc_stats[tbl] = (unsigned char *)(*cinfo->mem->alloc_small)
+          ((j_common_ptr)cinfo, JPOOL_IMAGE, DC_STAT_BINS);
+      MEMZERO(entropy->dc_stats[tbl], DC_STAT_BINS);
+      /* Initialize DC predictions to 0 */
+      entropy->last_dc_val[ci] = 0;
+      entropy->dc_context[ci] = 0;
+    }
+    /* AC needs no table when not present */
+    if (cinfo->progressive_mode == 0 || cinfo->Se) {
+      tbl = compptr->ac_tbl_no;
+      if (tbl < 0 || tbl >= NUM_ARITH_TBLS)
+        ERREXIT1(cinfo, JERR_NO_ARITH_TABLE, tbl);
+      if (entropy->ac_stats[tbl] == NULL)
+        entropy->ac_stats[tbl] = (unsigned char *)(*cinfo->mem->alloc_small)
+          ((j_common_ptr)cinfo, JPOOL_IMAGE, AC_STAT_BINS);
+      MEMZERO(entropy->ac_stats[tbl], AC_STAT_BINS);
+#ifdef CALCULATE_SPECTRAL_CONDITIONING
+      if (cinfo->progressive_mode)
+        /* Section G.1.3.2: Set appropriate arithmetic conditioning value Kx */
+        cinfo->arith_ac_K[tbl] = cinfo->Ss +
+                                 ((8 + cinfo->Se - cinfo->Ss) >> 4);
+#endif
+    }
+  }
+
+  /* Initialize arithmetic encoding variables */
+  entropy->c = 0;
+  entropy->a = 0x10000L;
+  entropy->sc = 0;
+  entropy->zc = 0;
+  entropy->ct = 11;
+  entropy->buffer = -1;  /* empty */
+
+  /* Initialize restart stuff */
+  entropy->restarts_to_go = cinfo->restart_interval;
+  entropy->next_restart_num = 0;
+}
+
+
+/*
+ * Module initialization routine for arithmetic entropy encoding.
+ */
+
+GLOBAL(void)
+jinit_arith_encoder(j_compress_ptr cinfo)
+{
+  arith_entropy_ptr entropy;
+  int i;
+
+  entropy = (arith_entropy_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(arith_entropy_encoder));
+  cinfo->entropy = (struct jpeg_entropy_encoder *)entropy;
+  entropy->pub.start_pass = start_pass;
+  entropy->pub.finish_pass = finish_pass;
+
+  /* Mark tables unallocated */
+  for (i = 0; i < NUM_ARITH_TBLS; i++) {
+    entropy->dc_stats[i] = NULL;
+    entropy->ac_stats[i] = NULL;
+  }
+
+  /* Initialize index for fixed probability estimation */
+  entropy->fixed_bin[0] = 113;
+}

3rdparty/libjpeg-turbo/src/jccoefct.c

@@ -0,0 +1,449 @@
+/*
+ * jccoefct.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1997, Thomas G. Lane.
+ * It was modified by The libjpeg-turbo Project to include only code and
+ * information relevant to libjpeg-turbo.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains the coefficient buffer controller for compression.
+ * This controller is the top level of the JPEG compressor proper.
+ * The coefficient buffer lies between forward-DCT and entropy encoding steps.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/* We use a full-image coefficient buffer when doing Huffman optimization,
+ * and also for writing multiple-scan JPEG files.  In all cases, the DCT
+ * step is run during the first pass, and subsequent passes need only read
+ * the buffered coefficients.
+ */
+#ifdef ENTROPY_OPT_SUPPORTED
+#define FULL_COEF_BUFFER_SUPPORTED
+#else
+#ifdef C_MULTISCAN_FILES_SUPPORTED
+#define FULL_COEF_BUFFER_SUPPORTED
+#endif
+#endif
+
+
+/* Private buffer controller object */
+
+typedef struct {
+  struct jpeg_c_coef_controller pub; /* public fields */
+
+  JDIMENSION iMCU_row_num;      /* iMCU row # within image */
+  JDIMENSION mcu_ctr;           /* counts MCUs processed in current row */
+  int MCU_vert_offset;          /* counts MCU rows within iMCU row */
+  int MCU_rows_per_iMCU_row;    /* number of such rows needed */
+
+  /* For single-pass compression, it's sufficient to buffer just one MCU
+   * (although this may prove a bit slow in practice).  We allocate a
+   * workspace of C_MAX_BLOCKS_IN_MCU coefficient blocks, and reuse it for each
+   * MCU constructed and sent.  In multi-pass modes, this array points to the
+   * current MCU's blocks within the virtual arrays.
+   */
+  JBLOCKROW MCU_buffer[C_MAX_BLOCKS_IN_MCU];
+
+  /* In multi-pass modes, we need a virtual block array for each component. */
+  jvirt_barray_ptr whole_image[MAX_COMPONENTS];
+} my_coef_controller;
+
+typedef my_coef_controller *my_coef_ptr;
+
+
+/* Forward declarations */
+METHODDEF(boolean) compress_data(j_compress_ptr cinfo, JSAMPIMAGE input_buf);
+#ifdef FULL_COEF_BUFFER_SUPPORTED
+METHODDEF(boolean) compress_first_pass(j_compress_ptr cinfo,
+                                       JSAMPIMAGE input_buf);
+METHODDEF(boolean) compress_output(j_compress_ptr cinfo, JSAMPIMAGE input_buf);
+#endif
+
+
+LOCAL(void)
+start_iMCU_row(j_compress_ptr cinfo)
+/* Reset within-iMCU-row counters for a new row */
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+
+  /* In an interleaved scan, an MCU row is the same as an iMCU row.
+   * In a noninterleaved scan, an iMCU row has v_samp_factor MCU rows.
+   * But at the bottom of the image, process only what's left.
+   */
+  if (cinfo->comps_in_scan > 1) {
+    coef->MCU_rows_per_iMCU_row = 1;
+  } else {
+    if (coef->iMCU_row_num < (cinfo->total_iMCU_rows - 1))
+      coef->MCU_rows_per_iMCU_row = cinfo->cur_comp_info[0]->v_samp_factor;
+    else
+      coef->MCU_rows_per_iMCU_row = cinfo->cur_comp_info[0]->last_row_height;
+  }
+
+  coef->mcu_ctr = 0;
+  coef->MCU_vert_offset = 0;
+}
+
+
+/*
+ * Initialize for a processing pass.
+ */
+
+METHODDEF(void)
+start_pass_coef(j_compress_ptr cinfo, J_BUF_MODE pass_mode)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+
+  coef->iMCU_row_num = 0;
+  start_iMCU_row(cinfo);
+
+  switch (pass_mode) {
+  case JBUF_PASS_THRU:
+    if (coef->whole_image[0] != NULL)
+      ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+    coef->pub.compress_data = compress_data;
+    break;
+#ifdef FULL_COEF_BUFFER_SUPPORTED
+  case JBUF_SAVE_AND_PASS:
+    if (coef->whole_image[0] == NULL)
+      ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+    coef->pub.compress_data = compress_first_pass;
+    break;
+  case JBUF_CRANK_DEST:
+    if (coef->whole_image[0] == NULL)
+      ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+    coef->pub.compress_data = compress_output;
+    break;
+#endif
+  default:
+    ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+    break;
+  }
+}
+
+
+/*
+ * Process some data in the single-pass case.
+ * We process the equivalent of one fully interleaved MCU row ("iMCU" row)
+ * per call, ie, v_samp_factor block rows for each component in the image.
+ * Returns TRUE if the iMCU row is completed, FALSE if suspended.
+ *
+ * NB: input_buf contains a plane for each component in image,
+ * which we index according to the component's SOF position.
+ */
+
+METHODDEF(boolean)
+compress_data(j_compress_ptr cinfo, JSAMPIMAGE input_buf)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION MCU_col_num;       /* index of current MCU within row */
+  JDIMENSION last_MCU_col = cinfo->MCUs_per_row - 1;
+  JDIMENSION last_iMCU_row = cinfo->total_iMCU_rows - 1;
+  int blkn, bi, ci, yindex, yoffset, blockcnt;
+  JDIMENSION ypos, xpos;
+  jpeg_component_info *compptr;
+
+  /* Loop to write as much as one whole iMCU row */
+  for (yoffset = coef->MCU_vert_offset; yoffset < coef->MCU_rows_per_iMCU_row;
+       yoffset++) {
+    for (MCU_col_num = coef->mcu_ctr; MCU_col_num <= last_MCU_col;
+         MCU_col_num++) {
+      /* Determine where data comes from in input_buf and do the DCT thing.
+       * Each call on forward_DCT processes a horizontal row of DCT blocks
+       * as wide as an MCU; we rely on having allocated the MCU_buffer[] blocks
+       * sequentially.  Dummy blocks at the right or bottom edge are filled in
+       * specially.  The data in them does not matter for image reconstruction,
+       * so we fill them with values that will encode to the smallest amount of
+       * data, viz: all zeroes in the AC entries, DC entries equal to previous
+       * block's DC value.  (Thanks to Thomas Kinsman for this idea.)
+       */
+      blkn = 0;
+      for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+        compptr = cinfo->cur_comp_info[ci];
+        blockcnt = (MCU_col_num < last_MCU_col) ? compptr->MCU_width :
+                                                  compptr->last_col_width;
+        xpos = MCU_col_num * compptr->MCU_sample_width;
+        ypos = yoffset * DCTSIZE; /* ypos == (yoffset+yindex) * DCTSIZE */
+        for (yindex = 0; yindex < compptr->MCU_height; yindex++) {
+          if (coef->iMCU_row_num < last_iMCU_row ||
+              yoffset + yindex < compptr->last_row_height) {
+            (*cinfo->fdct->forward_DCT) (cinfo, compptr,
+                                         input_buf[compptr->component_index],
+                                         coef->MCU_buffer[blkn],
+                                         ypos, xpos, (JDIMENSION)blockcnt);
+            if (blockcnt < compptr->MCU_width) {
+              /* Create some dummy blocks at the right edge of the image. */
+              jzero_far((void *)coef->MCU_buffer[blkn + blockcnt],
+                        (compptr->MCU_width - blockcnt) * sizeof(JBLOCK));
+              for (bi = blockcnt; bi < compptr->MCU_width; bi++) {
+                coef->MCU_buffer[blkn + bi][0][0] =
+                  coef->MCU_buffer[blkn + bi - 1][0][0];
+              }
+            }
+          } else {
+            /* Create a row of dummy blocks at the bottom of the image. */
+            jzero_far((void *)coef->MCU_buffer[blkn],
+                      compptr->MCU_width * sizeof(JBLOCK));
+            for (bi = 0; bi < compptr->MCU_width; bi++) {
+              coef->MCU_buffer[blkn + bi][0][0] =
+                coef->MCU_buffer[blkn - 1][0][0];
+            }
+          }
+          blkn += compptr->MCU_width;
+          ypos += DCTSIZE;
+        }
+      }
+      /* Try to write the MCU.  In event of a suspension failure, we will
+       * re-DCT the MCU on restart (a bit inefficient, could be fixed...)
+       */
+      if (!(*cinfo->entropy->encode_mcu) (cinfo, coef->MCU_buffer)) {
+        /* Suspension forced; update state counters and exit */
+        coef->MCU_vert_offset = yoffset;
+        coef->mcu_ctr = MCU_col_num;
+        return FALSE;
+      }
+    }
+    /* Completed an MCU row, but perhaps not an iMCU row */
+    coef->mcu_ctr = 0;
+  }
+  /* Completed the iMCU row, advance counters for next one */
+  coef->iMCU_row_num++;
+  start_iMCU_row(cinfo);
+  return TRUE;
+}
+
+
+#ifdef FULL_COEF_BUFFER_SUPPORTED
+
+/*
+ * Process some data in the first pass of a multi-pass case.
+ * We process the equivalent of one fully interleaved MCU row ("iMCU" row)
+ * per call, ie, v_samp_factor block rows for each component in the image.
+ * This amount of data is read from the source buffer, DCT'd and quantized,
+ * and saved into the virtual arrays.  We also generate suitable dummy blocks
+ * as needed at the right and lower edges.  (The dummy blocks are constructed
+ * in the virtual arrays, which have been padded appropriately.)  This makes
+ * it possible for subsequent passes not to worry about real vs. dummy blocks.
+ *
+ * We must also emit the data to the entropy encoder.  This is conveniently
+ * done by calling compress_output() after we've loaded the current strip
+ * of the virtual arrays.
+ *
+ * NB: input_buf contains a plane for each component in image.  All
+ * components are DCT'd and loaded into the virtual arrays in this pass.
+ * However, it may be that only a subset of the components are emitted to
+ * the entropy encoder during this first pass; be careful about looking
+ * at the scan-dependent variables (MCU dimensions, etc).
+ */
+
+METHODDEF(boolean)
+compress_first_pass(j_compress_ptr cinfo, JSAMPIMAGE input_buf)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION last_iMCU_row = cinfo->total_iMCU_rows - 1;
+  JDIMENSION blocks_across, MCUs_across, MCUindex;
+  int bi, ci, h_samp_factor, block_row, block_rows, ndummy;
+  JCOEF lastDC;
+  jpeg_component_info *compptr;
+  JBLOCKARRAY buffer;
+  JBLOCKROW thisblockrow, lastblockrow;
+
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    /* Align the virtual buffer for this component. */
+    buffer = (*cinfo->mem->access_virt_barray)
+      ((j_common_ptr)cinfo, coef->whole_image[ci],
+       coef->iMCU_row_num * compptr->v_samp_factor,
+       (JDIMENSION)compptr->v_samp_factor, TRUE);
+    /* Count non-dummy DCT block rows in this iMCU row. */
+    if (coef->iMCU_row_num < last_iMCU_row)
+      block_rows = compptr->v_samp_factor;
+    else {
+      /* NB: can't use last_row_height here, since may not be set! */
+      block_rows = (int)(compptr->height_in_blocks % compptr->v_samp_factor);
+      if (block_rows == 0) block_rows = compptr->v_samp_factor;
+    }
+    blocks_across = compptr->width_in_blocks;
+    h_samp_factor = compptr->h_samp_factor;
+    /* Count number of dummy blocks to be added at the right margin. */
+    ndummy = (int)(blocks_across % h_samp_factor);
+    if (ndummy > 0)
+      ndummy = h_samp_factor - ndummy;
+    /* Perform DCT for all non-dummy blocks in this iMCU row.  Each call
+     * on forward_DCT processes a complete horizontal row of DCT blocks.
+     */
+    for (block_row = 0; block_row < block_rows; block_row++) {
+      thisblockrow = buffer[block_row];
+      (*cinfo->fdct->forward_DCT) (cinfo, compptr,
+                                   input_buf[ci], thisblockrow,
+                                   (JDIMENSION)(block_row * DCTSIZE),
+                                   (JDIMENSION)0, blocks_across);
+      if (ndummy > 0) {
+        /* Create dummy blocks at the right edge of the image. */
+        thisblockrow += blocks_across; /* => first dummy block */
+        jzero_far((void *)thisblockrow, ndummy * sizeof(JBLOCK));
+        lastDC = thisblockrow[-1][0];
+        for (bi = 0; bi < ndummy; bi++) {
+          thisblockrow[bi][0] = lastDC;
+        }
+      }
+    }
+    /* If at end of image, create dummy block rows as needed.
+     * The tricky part here is that within each MCU, we want the DC values
+     * of the dummy blocks to match the last real block's DC value.
+     * This squeezes a few more bytes out of the resulting file...
+     */
+    if (coef->iMCU_row_num == last_iMCU_row) {
+      blocks_across += ndummy;  /* include lower right corner */
+      MCUs_across = blocks_across / h_samp_factor;
+      for (block_row = block_rows; block_row < compptr->v_samp_factor;
+           block_row++) {
+        thisblockrow = buffer[block_row];
+        lastblockrow = buffer[block_row - 1];
+        jzero_far((void *)thisblockrow,
+                  (size_t)(blocks_across * sizeof(JBLOCK)));
+        for (MCUindex = 0; MCUindex < MCUs_across; MCUindex++) {
+          lastDC = lastblockrow[h_samp_factor - 1][0];
+          for (bi = 0; bi < h_samp_factor; bi++) {
+            thisblockrow[bi][0] = lastDC;
+          }
+          thisblockrow += h_samp_factor; /* advance to next MCU in row */
+          lastblockrow += h_samp_factor;
+        }
+      }
+    }
+  }
+  /* NB: compress_output will increment iMCU_row_num if successful.
+   * A suspension return will result in redoing all the work above next time.
+   */
+
+  /* Emit data to the entropy encoder, sharing code with subsequent passes */
+  return compress_output(cinfo, input_buf);
+}
+
+
+/*
+ * Process some data in subsequent passes of a multi-pass case.
+ * We process the equivalent of one fully interleaved MCU row ("iMCU" row)
+ * per call, ie, v_samp_factor block rows for each component in the scan.
+ * The data is obtained from the virtual arrays and fed to the entropy coder.
+ * Returns TRUE if the iMCU row is completed, FALSE if suspended.
+ *
+ * NB: input_buf is ignored; it is likely to be a NULL pointer.
+ */
+
+METHODDEF(boolean)
+compress_output(j_compress_ptr cinfo, JSAMPIMAGE input_buf)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION MCU_col_num;       /* index of current MCU within row */
+  int blkn, ci, xindex, yindex, yoffset;
+  JDIMENSION start_col;
+  JBLOCKARRAY buffer[MAX_COMPS_IN_SCAN];
+  JBLOCKROW buffer_ptr;
+  jpeg_component_info *compptr;
+
+  /* Align the virtual buffers for the components used in this scan.
+   * NB: during first pass, this is safe only because the buffers will
+   * already be aligned properly, so jmemmgr.c won't need to do any I/O.
+   */
+  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+    compptr = cinfo->cur_comp_info[ci];
+    buffer[ci] = (*cinfo->mem->access_virt_barray)
+      ((j_common_ptr)cinfo, coef->whole_image[compptr->component_index],
+       coef->iMCU_row_num * compptr->v_samp_factor,
+       (JDIMENSION)compptr->v_samp_factor, FALSE);
+  }
+
+  /* Loop to process one whole iMCU row */
+  for (yoffset = coef->MCU_vert_offset; yoffset < coef->MCU_rows_per_iMCU_row;
+       yoffset++) {
+    for (MCU_col_num = coef->mcu_ctr; MCU_col_num < cinfo->MCUs_per_row;
+         MCU_col_num++) {
+      /* Construct list of pointers to DCT blocks belonging to this MCU */
+      blkn = 0;                 /* index of current DCT block within MCU */
+      for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+        compptr = cinfo->cur_comp_info[ci];
+        start_col = MCU_col_num * compptr->MCU_width;
+        for (yindex = 0; yindex < compptr->MCU_height; yindex++) {
+          buffer_ptr = buffer[ci][yindex + yoffset] + start_col;
+          for (xindex = 0; xindex < compptr->MCU_width; xindex++) {
+            coef->MCU_buffer[blkn++] = buffer_ptr++;
+          }
+        }
+      }
+      /* Try to write the MCU. */
+      if (!(*cinfo->entropy->encode_mcu) (cinfo, coef->MCU_buffer)) {
+        /* Suspension forced; update state counters and exit */
+        coef->MCU_vert_offset = yoffset;
+        coef->mcu_ctr = MCU_col_num;
+        return FALSE;
+      }
+    }
+    /* Completed an MCU row, but perhaps not an iMCU row */
+    coef->mcu_ctr = 0;
+  }
+  /* Completed the iMCU row, advance counters for next one */
+  coef->iMCU_row_num++;
+  start_iMCU_row(cinfo);
+  return TRUE;
+}
+
+#endif /* FULL_COEF_BUFFER_SUPPORTED */
+
+
+/*
+ * Initialize coefficient buffer controller.
+ */
+
+GLOBAL(void)
+jinit_c_coef_controller(j_compress_ptr cinfo, boolean need_full_buffer)
+{
+  my_coef_ptr coef;
+
+  coef = (my_coef_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_coef_controller));
+  cinfo->coef = (struct jpeg_c_coef_controller *)coef;
+  coef->pub.start_pass = start_pass_coef;
+
+  /* Create the coefficient buffer. */
+  if (need_full_buffer) {
+#ifdef FULL_COEF_BUFFER_SUPPORTED
+    /* Allocate a full-image virtual array for each component, */
+    /* padded to a multiple of samp_factor DCT blocks in each direction. */
+    int ci;
+    jpeg_component_info *compptr;
+
+    for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+         ci++, compptr++) {
+      coef->whole_image[ci] = (*cinfo->mem->request_virt_barray)
+        ((j_common_ptr)cinfo, JPOOL_IMAGE, FALSE,
+         (JDIMENSION)jround_up((long)compptr->width_in_blocks,
+                               (long)compptr->h_samp_factor),
+         (JDIMENSION)jround_up((long)compptr->height_in_blocks,
+                               (long)compptr->v_samp_factor),
+         (JDIMENSION)compptr->v_samp_factor);
+    }
+#else
+    ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+#endif
+  } else {
+    /* We only need a single-MCU buffer. */
+    JBLOCKROW buffer;
+    int i;
+
+    buffer = (JBLOCKROW)
+      (*cinfo->mem->alloc_large) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                  C_MAX_BLOCKS_IN_MCU * sizeof(JBLOCK));
+    for (i = 0; i < C_MAX_BLOCKS_IN_MCU; i++) {
+      coef->MCU_buffer[i] = buffer + i;
+    }
+    coef->whole_image[0] = NULL; /* flag for no virtual arrays */
+  }
+}

3rdparty/libjpeg-turbo/src/jccolext.c

@@ -0,0 +1,144 @@
+/*
+ * jccolext.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1996, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2009-2012, 2015, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains input colorspace conversion routines.
+ */
+
+
+/* This file is included by jccolor.c */
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ *
+ * Note that we change from the application's interleaved-pixel format
+ * to our internal noninterleaved, one-plane-per-component format.
+ * The input buffer is therefore three times as wide as the output buffer.
+ *
+ * A starting row offset is provided only for the output buffer.  The caller
+ * can easily adjust the passed input_buf value to accommodate any row
+ * offset required on that side.
+ */
+
+INLINE
+LOCAL(void)
+rgb_ycc_convert_internal(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                         JSAMPIMAGE output_buf, JDIMENSION output_row,
+                         int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int r, g, b;
+  register JLONG *ctab = cconvert->rgb_ycc_tab;
+  register JSAMPROW inptr;
+  register JSAMPROW outptr0, outptr1, outptr2;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->image_width;
+
+  while (--num_rows >= 0) {
+    inptr = *input_buf++;
+    outptr0 = output_buf[0][output_row];
+    outptr1 = output_buf[1][output_row];
+    outptr2 = output_buf[2][output_row];
+    output_row++;
+    for (col = 0; col < num_cols; col++) {
+      r = GETJSAMPLE(inptr[RGB_RED]);
+      g = GETJSAMPLE(inptr[RGB_GREEN]);
+      b = GETJSAMPLE(inptr[RGB_BLUE]);
+      inptr += RGB_PIXELSIZE;
+      /* If the inputs are 0..MAXJSAMPLE, the outputs of these equations
+       * must be too; we do not need an explicit range-limiting operation.
+       * Hence the value being shifted is never negative, and we don't
+       * need the general RIGHT_SHIFT macro.
+       */
+      /* Y */
+      outptr0[col] = (JSAMPLE)((ctab[r + R_Y_OFF] + ctab[g + G_Y_OFF] +
+                                ctab[b + B_Y_OFF]) >> SCALEBITS);
+      /* Cb */
+      outptr1[col] = (JSAMPLE)((ctab[r + R_CB_OFF] + ctab[g + G_CB_OFF] +
+                                ctab[b + B_CB_OFF]) >> SCALEBITS);
+      /* Cr */
+      outptr2[col] = (JSAMPLE)((ctab[r + R_CR_OFF] + ctab[g + G_CR_OFF] +
+                                ctab[b + B_CR_OFF]) >> SCALEBITS);
+    }
+  }
+}
+
+
+/**************** Cases other than RGB -> YCbCr **************/
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ * This version handles RGB->grayscale conversion, which is the same
+ * as the RGB->Y portion of RGB->YCbCr.
+ * We assume rgb_ycc_start has been called (we only use the Y tables).
+ */
+
+INLINE
+LOCAL(void)
+rgb_gray_convert_internal(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                          JSAMPIMAGE output_buf, JDIMENSION output_row,
+                          int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int r, g, b;
+  register JLONG *ctab = cconvert->rgb_ycc_tab;
+  register JSAMPROW inptr;
+  register JSAMPROW outptr;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->image_width;
+
+  while (--num_rows >= 0) {
+    inptr = *input_buf++;
+    outptr = output_buf[0][output_row];
+    output_row++;
+    for (col = 0; col < num_cols; col++) {
+      r = GETJSAMPLE(inptr[RGB_RED]);
+      g = GETJSAMPLE(inptr[RGB_GREEN]);
+      b = GETJSAMPLE(inptr[RGB_BLUE]);
+      inptr += RGB_PIXELSIZE;
+      /* Y */
+      outptr[col] = (JSAMPLE)((ctab[r + R_Y_OFF] + ctab[g + G_Y_OFF] +
+                               ctab[b + B_Y_OFF]) >> SCALEBITS);
+    }
+  }
+}
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ * This version handles extended RGB->plain RGB conversion
+ */
+
+INLINE
+LOCAL(void)
+rgb_rgb_convert_internal(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                         JSAMPIMAGE output_buf, JDIMENSION output_row,
+                         int num_rows)
+{
+  register JSAMPROW inptr;
+  register JSAMPROW outptr0, outptr1, outptr2;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->image_width;
+
+  while (--num_rows >= 0) {
+    inptr = *input_buf++;
+    outptr0 = output_buf[0][output_row];
+    outptr1 = output_buf[1][output_row];
+    outptr2 = output_buf[2][output_row];
+    output_row++;
+    for (col = 0; col < num_cols; col++) {
+      outptr0[col] = GETJSAMPLE(inptr[RGB_RED]);
+      outptr1[col] = GETJSAMPLE(inptr[RGB_GREEN]);
+      outptr2[col] = GETJSAMPLE(inptr[RGB_BLUE]);
+      inptr += RGB_PIXELSIZE;
+    }
+  }
+}

3rdparty/libjpeg-turbo/src/jccolor.c

@@ -0,0 +1,710 @@
+/*
+ * jccolor.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1996, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * Copyright (C) 2009-2012, 2015, D. R. Commander.
+ * Copyright (C) 2014, MIPS Technologies, Inc., California.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains input colorspace conversion routines.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jsimd.h"
+#include "jconfigint.h"
+
+
+/* Private subobject */
+
+typedef struct {
+  struct jpeg_color_converter pub; /* public fields */
+
+  /* Private state for RGB->YCC conversion */
+  JLONG *rgb_ycc_tab;           /* => table for RGB to YCbCr conversion */
+} my_color_converter;
+
+typedef my_color_converter *my_cconvert_ptr;
+
+
+/**************** RGB -> YCbCr conversion: most common case **************/
+
+/*
+ * YCbCr is defined per CCIR 601-1, except that Cb and Cr are
+ * normalized to the range 0..MAXJSAMPLE rather than -0.5 .. 0.5.
+ * The conversion equations to be implemented are therefore
+ *      Y  =  0.29900 * R + 0.58700 * G + 0.11400 * B
+ *      Cb = -0.16874 * R - 0.33126 * G + 0.50000 * B  + CENTERJSAMPLE
+ *      Cr =  0.50000 * R - 0.41869 * G - 0.08131 * B  + CENTERJSAMPLE
+ * (These numbers are derived from TIFF 6.0 section 21, dated 3-June-92.)
+ * Note: older versions of the IJG code used a zero offset of MAXJSAMPLE/2,
+ * rather than CENTERJSAMPLE, for Cb and Cr.  This gave equal positive and
+ * negative swings for Cb/Cr, but meant that grayscale values (Cb=Cr=0)
+ * were not represented exactly.  Now we sacrifice exact representation of
+ * maximum red and maximum blue in order to get exact grayscales.
+ *
+ * To avoid floating-point arithmetic, we represent the fractional constants
+ * as integers scaled up by 2^16 (about 4 digits precision); we have to divide
+ * the products by 2^16, with appropriate rounding, to get the correct answer.
+ *
+ * For even more speed, we avoid doing any multiplications in the inner loop
+ * by precalculating the constants times R,G,B for all possible values.
+ * For 8-bit JSAMPLEs this is very reasonable (only 256 entries per table);
+ * for 12-bit samples it is still acceptable.  It's not very reasonable for
+ * 16-bit samples, but if you want lossless storage you shouldn't be changing
+ * colorspace anyway.
+ * The CENTERJSAMPLE offsets and the rounding fudge-factor of 0.5 are included
+ * in the tables to save adding them separately in the inner loop.
+ */
+
+#define SCALEBITS       16      /* speediest right-shift on some machines */
+#define CBCR_OFFSET     ((JLONG)CENTERJSAMPLE << SCALEBITS)
+#define ONE_HALF        ((JLONG)1 << (SCALEBITS - 1))
+#define FIX(x)          ((JLONG)((x) * (1L << SCALEBITS) + 0.5))
+
+/* We allocate one big table and divide it up into eight parts, instead of
+ * doing eight alloc_small requests.  This lets us use a single table base
+ * address, which can be held in a register in the inner loops on many
+ * machines (more than can hold all eight addresses, anyway).
+ */
+
+#define R_Y_OFF         0                       /* offset to R => Y section */
+#define G_Y_OFF         (1 * (MAXJSAMPLE + 1))  /* offset to G => Y section */
+#define B_Y_OFF         (2 * (MAXJSAMPLE + 1))  /* etc. */
+#define R_CB_OFF        (3 * (MAXJSAMPLE + 1))
+#define G_CB_OFF        (4 * (MAXJSAMPLE + 1))
+#define B_CB_OFF        (5 * (MAXJSAMPLE + 1))
+#define R_CR_OFF        B_CB_OFF                /* B=>Cb, R=>Cr are the same */
+#define G_CR_OFF        (6 * (MAXJSAMPLE + 1))
+#define B_CR_OFF        (7 * (MAXJSAMPLE + 1))
+#define TABLE_SIZE      (8 * (MAXJSAMPLE + 1))
+
+
+/* Include inline routines for colorspace extensions */
+
+#include "jccolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+
+#define RGB_RED  EXT_RGB_RED
+#define RGB_GREEN  EXT_RGB_GREEN
+#define RGB_BLUE  EXT_RGB_BLUE
+#define RGB_PIXELSIZE  EXT_RGB_PIXELSIZE
+#define rgb_ycc_convert_internal  extrgb_ycc_convert_internal
+#define rgb_gray_convert_internal  extrgb_gray_convert_internal
+#define rgb_rgb_convert_internal  extrgb_rgb_convert_internal
+#include "jccolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef rgb_ycc_convert_internal
+#undef rgb_gray_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_RGBX_RED
+#define RGB_GREEN  EXT_RGBX_GREEN
+#define RGB_BLUE  EXT_RGBX_BLUE
+#define RGB_PIXELSIZE  EXT_RGBX_PIXELSIZE
+#define rgb_ycc_convert_internal  extrgbx_ycc_convert_internal
+#define rgb_gray_convert_internal  extrgbx_gray_convert_internal
+#define rgb_rgb_convert_internal  extrgbx_rgb_convert_internal
+#include "jccolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef rgb_ycc_convert_internal
+#undef rgb_gray_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_BGR_RED
+#define RGB_GREEN  EXT_BGR_GREEN
+#define RGB_BLUE  EXT_BGR_BLUE
+#define RGB_PIXELSIZE  EXT_BGR_PIXELSIZE
+#define rgb_ycc_convert_internal  extbgr_ycc_convert_internal
+#define rgb_gray_convert_internal  extbgr_gray_convert_internal
+#define rgb_rgb_convert_internal  extbgr_rgb_convert_internal
+#include "jccolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef rgb_ycc_convert_internal
+#undef rgb_gray_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_BGRX_RED
+#define RGB_GREEN  EXT_BGRX_GREEN
+#define RGB_BLUE  EXT_BGRX_BLUE
+#define RGB_PIXELSIZE  EXT_BGRX_PIXELSIZE
+#define rgb_ycc_convert_internal  extbgrx_ycc_convert_internal
+#define rgb_gray_convert_internal  extbgrx_gray_convert_internal
+#define rgb_rgb_convert_internal  extbgrx_rgb_convert_internal
+#include "jccolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef rgb_ycc_convert_internal
+#undef rgb_gray_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_XBGR_RED
+#define RGB_GREEN  EXT_XBGR_GREEN
+#define RGB_BLUE  EXT_XBGR_BLUE
+#define RGB_PIXELSIZE  EXT_XBGR_PIXELSIZE
+#define rgb_ycc_convert_internal  extxbgr_ycc_convert_internal
+#define rgb_gray_convert_internal  extxbgr_gray_convert_internal
+#define rgb_rgb_convert_internal  extxbgr_rgb_convert_internal
+#include "jccolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef rgb_ycc_convert_internal
+#undef rgb_gray_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_XRGB_RED
+#define RGB_GREEN  EXT_XRGB_GREEN
+#define RGB_BLUE  EXT_XRGB_BLUE
+#define RGB_PIXELSIZE  EXT_XRGB_PIXELSIZE
+#define rgb_ycc_convert_internal  extxrgb_ycc_convert_internal
+#define rgb_gray_convert_internal  extxrgb_gray_convert_internal
+#define rgb_rgb_convert_internal  extxrgb_rgb_convert_internal
+#include "jccolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef rgb_ycc_convert_internal
+#undef rgb_gray_convert_internal
+#undef rgb_rgb_convert_internal
+
+
+/*
+ * Initialize for RGB->YCC colorspace conversion.
+ */
+
+METHODDEF(void)
+rgb_ycc_start(j_compress_ptr cinfo)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  JLONG *rgb_ycc_tab;
+  JLONG i;
+
+  /* Allocate and fill in the conversion tables. */
+  cconvert->rgb_ycc_tab = rgb_ycc_tab = (JLONG *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                (TABLE_SIZE * sizeof(JLONG)));
+
+  for (i = 0; i <= MAXJSAMPLE; i++) {
+    rgb_ycc_tab[i + R_Y_OFF] = FIX(0.29900) * i;
+    rgb_ycc_tab[i + G_Y_OFF] = FIX(0.58700) * i;
+    rgb_ycc_tab[i + B_Y_OFF] = FIX(0.11400) * i   + ONE_HALF;
+    rgb_ycc_tab[i + R_CB_OFF] = (-FIX(0.16874)) * i;
+    rgb_ycc_tab[i + G_CB_OFF] = (-FIX(0.33126)) * i;
+    /* We use a rounding fudge-factor of 0.5-epsilon for Cb and Cr.
+     * This ensures that the maximum output will round to MAXJSAMPLE
+     * not MAXJSAMPLE+1, and thus that we don't have to range-limit.
+     */
+    rgb_ycc_tab[i + B_CB_OFF] = FIX(0.50000) * i  + CBCR_OFFSET + ONE_HALF - 1;
+/*  B=>Cb and R=>Cr tables are the same
+    rgb_ycc_tab[i + R_CR_OFF] = FIX(0.50000) * i  + CBCR_OFFSET + ONE_HALF - 1;
+*/
+    rgb_ycc_tab[i + G_CR_OFF] = (-FIX(0.41869)) * i;
+    rgb_ycc_tab[i + B_CR_OFF] = (-FIX(0.08131)) * i;
+  }
+}
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ */
+
+METHODDEF(void)
+rgb_ycc_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                JSAMPIMAGE output_buf, JDIMENSION output_row, int num_rows)
+{
+  switch (cinfo->in_color_space) {
+  case JCS_EXT_RGB:
+    extrgb_ycc_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                num_rows);
+    break;
+  case JCS_EXT_RGBX:
+  case JCS_EXT_RGBA:
+    extrgbx_ycc_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_BGR:
+    extbgr_ycc_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                num_rows);
+    break;
+  case JCS_EXT_BGRX:
+  case JCS_EXT_BGRA:
+    extbgrx_ycc_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_XBGR:
+  case JCS_EXT_ABGR:
+    extxbgr_ycc_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_XRGB:
+  case JCS_EXT_ARGB:
+    extxrgb_ycc_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  default:
+    rgb_ycc_convert_internal(cinfo, input_buf, output_buf, output_row,
+                             num_rows);
+    break;
+  }
+}
+
+
+/**************** Cases other than RGB -> YCbCr **************/
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ */
+
+METHODDEF(void)
+rgb_gray_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                 JSAMPIMAGE output_buf, JDIMENSION output_row, int num_rows)
+{
+  switch (cinfo->in_color_space) {
+  case JCS_EXT_RGB:
+    extrgb_gray_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_RGBX:
+  case JCS_EXT_RGBA:
+    extrgbx_gray_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                  num_rows);
+    break;
+  case JCS_EXT_BGR:
+    extbgr_gray_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_BGRX:
+  case JCS_EXT_BGRA:
+    extbgrx_gray_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                  num_rows);
+    break;
+  case JCS_EXT_XBGR:
+  case JCS_EXT_ABGR:
+    extxbgr_gray_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                  num_rows);
+    break;
+  case JCS_EXT_XRGB:
+  case JCS_EXT_ARGB:
+    extxrgb_gray_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                  num_rows);
+    break;
+  default:
+    rgb_gray_convert_internal(cinfo, input_buf, output_buf, output_row,
+                              num_rows);
+    break;
+  }
+}
+
+
+/*
+ * Extended RGB to plain RGB conversion
+ */
+
+METHODDEF(void)
+rgb_rgb_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                JSAMPIMAGE output_buf, JDIMENSION output_row, int num_rows)
+{
+  switch (cinfo->in_color_space) {
+  case JCS_EXT_RGB:
+    extrgb_rgb_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                num_rows);
+    break;
+  case JCS_EXT_RGBX:
+  case JCS_EXT_RGBA:
+    extrgbx_rgb_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_BGR:
+    extbgr_rgb_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                num_rows);
+    break;
+  case JCS_EXT_BGRX:
+  case JCS_EXT_BGRA:
+    extbgrx_rgb_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_XBGR:
+  case JCS_EXT_ABGR:
+    extxbgr_rgb_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  case JCS_EXT_XRGB:
+  case JCS_EXT_ARGB:
+    extxrgb_rgb_convert_internal(cinfo, input_buf, output_buf, output_row,
+                                 num_rows);
+    break;
+  default:
+    rgb_rgb_convert_internal(cinfo, input_buf, output_buf, output_row,
+                             num_rows);
+    break;
+  }
+}
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ * This version handles Adobe-style CMYK->YCCK conversion,
+ * where we convert R=1-C, G=1-M, and B=1-Y to YCbCr using the same
+ * conversion as above, while passing K (black) unchanged.
+ * We assume rgb_ycc_start has been called.
+ */
+
+METHODDEF(void)
+cmyk_ycck_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                  JSAMPIMAGE output_buf, JDIMENSION output_row, int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int r, g, b;
+  register JLONG *ctab = cconvert->rgb_ycc_tab;
+  register JSAMPROW inptr;
+  register JSAMPROW outptr0, outptr1, outptr2, outptr3;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->image_width;
+
+  while (--num_rows >= 0) {
+    inptr = *input_buf++;
+    outptr0 = output_buf[0][output_row];
+    outptr1 = output_buf[1][output_row];
+    outptr2 = output_buf[2][output_row];
+    outptr3 = output_buf[3][output_row];
+    output_row++;
+    for (col = 0; col < num_cols; col++) {
+      r = MAXJSAMPLE - GETJSAMPLE(inptr[0]);
+      g = MAXJSAMPLE - GETJSAMPLE(inptr[1]);
+      b = MAXJSAMPLE - GETJSAMPLE(inptr[2]);
+      /* K passes through as-is */
+      outptr3[col] = inptr[3];  /* don't need GETJSAMPLE here */
+      inptr += 4;
+      /* If the inputs are 0..MAXJSAMPLE, the outputs of these equations
+       * must be too; we do not need an explicit range-limiting operation.
+       * Hence the value being shifted is never negative, and we don't
+       * need the general RIGHT_SHIFT macro.
+       */
+      /* Y */
+      outptr0[col] = (JSAMPLE)((ctab[r + R_Y_OFF] + ctab[g + G_Y_OFF] +
+                                ctab[b + B_Y_OFF]) >> SCALEBITS);
+      /* Cb */
+      outptr1[col] = (JSAMPLE)((ctab[r + R_CB_OFF] + ctab[g + G_CB_OFF] +
+                                ctab[b + B_CB_OFF]) >> SCALEBITS);
+      /* Cr */
+      outptr2[col] = (JSAMPLE)((ctab[r + R_CR_OFF] + ctab[g + G_CR_OFF] +
+                                ctab[b + B_CR_OFF]) >> SCALEBITS);
+    }
+  }
+}
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ * This version handles grayscale output with no conversion.
+ * The source can be either plain grayscale or YCbCr (since Y == gray).
+ */
+
+METHODDEF(void)
+grayscale_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                  JSAMPIMAGE output_buf, JDIMENSION output_row, int num_rows)
+{
+  register JSAMPROW inptr;
+  register JSAMPROW outptr;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->image_width;
+  int instride = cinfo->input_components;
+
+  while (--num_rows >= 0) {
+    inptr = *input_buf++;
+    outptr = output_buf[0][output_row];
+    output_row++;
+    for (col = 0; col < num_cols; col++) {
+      outptr[col] = inptr[0];   /* don't need GETJSAMPLE() here */
+      inptr += instride;
+    }
+  }
+}
+
+
+/*
+ * Convert some rows of samples to the JPEG colorspace.
+ * This version handles multi-component colorspaces without conversion.
+ * We assume input_components == num_components.
+ */
+
+METHODDEF(void)
+null_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf, JSAMPIMAGE output_buf,
+             JDIMENSION output_row, int num_rows)
+{
+  register JSAMPROW inptr;
+  register JSAMPROW outptr, outptr0, outptr1, outptr2, outptr3;
+  register JDIMENSION col;
+  register int ci;
+  int nc = cinfo->num_components;
+  JDIMENSION num_cols = cinfo->image_width;
+
+  if (nc == 3) {
+    while (--num_rows >= 0) {
+      inptr = *input_buf++;
+      outptr0 = output_buf[0][output_row];
+      outptr1 = output_buf[1][output_row];
+      outptr2 = output_buf[2][output_row];
+      output_row++;
+      for (col = 0; col < num_cols; col++) {
+        outptr0[col] = *inptr++;
+        outptr1[col] = *inptr++;
+        outptr2[col] = *inptr++;
+      }
+    }
+  } else if (nc == 4) {
+    while (--num_rows >= 0) {
+      inptr = *input_buf++;
+      outptr0 = output_buf[0][output_row];
+      outptr1 = output_buf[1][output_row];
+      outptr2 = output_buf[2][output_row];
+      outptr3 = output_buf[3][output_row];
+      output_row++;
+      for (col = 0; col < num_cols; col++) {
+        outptr0[col] = *inptr++;
+        outptr1[col] = *inptr++;
+        outptr2[col] = *inptr++;
+        outptr3[col] = *inptr++;
+      }
+    }
+  } else {
+    while (--num_rows >= 0) {
+      /* It seems fastest to make a separate pass for each component. */
+      for (ci = 0; ci < nc; ci++) {
+        inptr = *input_buf;
+        outptr = output_buf[ci][output_row];
+        for (col = 0; col < num_cols; col++) {
+          outptr[col] = inptr[ci]; /* don't need GETJSAMPLE() here */
+          inptr += nc;
+        }
+      }
+      input_buf++;
+      output_row++;
+    }
+  }
+}
+
+
+/*
+ * Empty method for start_pass.
+ */
+
+METHODDEF(void)
+null_method(j_compress_ptr cinfo)
+{
+  /* no work needed */
+}
+
+
+/*
+ * Module initialization routine for input colorspace conversion.
+ */
+
+GLOBAL(void)
+jinit_color_converter(j_compress_ptr cinfo)
+{
+  my_cconvert_ptr cconvert;
+
+  cconvert = (my_cconvert_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_color_converter));
+  cinfo->cconvert = (struct jpeg_color_converter *)cconvert;
+  /* set start_pass to null method until we find out differently */
+  cconvert->pub.start_pass = null_method;
+
+  /* Make sure input_components agrees with in_color_space */
+  switch (cinfo->in_color_space) {
+  case JCS_GRAYSCALE:
+    if (cinfo->input_components != 1)
+      ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE);
+    break;
+
+  case JCS_RGB:
+  case JCS_EXT_RGB:
+  case JCS_EXT_RGBX:
+  case JCS_EXT_BGR:
+  case JCS_EXT_BGRX:
+  case JCS_EXT_XBGR:
+  case JCS_EXT_XRGB:
+  case JCS_EXT_RGBA:
+  case JCS_EXT_BGRA:
+  case JCS_EXT_ABGR:
+  case JCS_EXT_ARGB:
+    if (cinfo->input_components != rgb_pixelsize[cinfo->in_color_space])
+      ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE);
+    break;
+
+  case JCS_YCbCr:
+    if (cinfo->input_components != 3)
+      ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE);
+    break;
+
+  case JCS_CMYK:
+  case JCS_YCCK:
+    if (cinfo->input_components != 4)
+      ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE);
+    break;
+
+  default:                      /* JCS_UNKNOWN can be anything */
+    if (cinfo->input_components < 1)
+      ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE);
+    break;
+  }
+
+  /* Check num_components, set conversion method based on requested space */
+  switch (cinfo->jpeg_color_space) {
+  case JCS_GRAYSCALE:
+    if (cinfo->num_components != 1)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    if (cinfo->in_color_space == JCS_GRAYSCALE)
+      cconvert->pub.color_convert = grayscale_convert;
+    else if (cinfo->in_color_space == JCS_RGB ||
+             cinfo->in_color_space == JCS_EXT_RGB ||
+             cinfo->in_color_space == JCS_EXT_RGBX ||
+             cinfo->in_color_space == JCS_EXT_BGR ||
+             cinfo->in_color_space == JCS_EXT_BGRX ||
+             cinfo->in_color_space == JCS_EXT_XBGR ||
+             cinfo->in_color_space == JCS_EXT_XRGB ||
+             cinfo->in_color_space == JCS_EXT_RGBA ||
+             cinfo->in_color_space == JCS_EXT_BGRA ||
+             cinfo->in_color_space == JCS_EXT_ABGR ||
+             cinfo->in_color_space == JCS_EXT_ARGB) {
+      if (jsimd_can_rgb_gray())
+        cconvert->pub.color_convert = jsimd_rgb_gray_convert;
+      else {
+        cconvert->pub.start_pass = rgb_ycc_start;
+        cconvert->pub.color_convert = rgb_gray_convert;
+      }
+    } else if (cinfo->in_color_space == JCS_YCbCr)
+      cconvert->pub.color_convert = grayscale_convert;
+    else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  case JCS_RGB:
+    if (cinfo->num_components != 3)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    if (rgb_red[cinfo->in_color_space] == 0 &&
+        rgb_green[cinfo->in_color_space] == 1 &&
+        rgb_blue[cinfo->in_color_space] == 2 &&
+        rgb_pixelsize[cinfo->in_color_space] == 3) {
+#if defined(__mips__)
+      if (jsimd_c_can_null_convert())
+        cconvert->pub.color_convert = jsimd_c_null_convert;
+      else
+#endif
+        cconvert->pub.color_convert = null_convert;
+    } else if (cinfo->in_color_space == JCS_RGB ||
+               cinfo->in_color_space == JCS_EXT_RGB ||
+               cinfo->in_color_space == JCS_EXT_RGBX ||
+               cinfo->in_color_space == JCS_EXT_BGR ||
+               cinfo->in_color_space == JCS_EXT_BGRX ||
+               cinfo->in_color_space == JCS_EXT_XBGR ||
+               cinfo->in_color_space == JCS_EXT_XRGB ||
+               cinfo->in_color_space == JCS_EXT_RGBA ||
+               cinfo->in_color_space == JCS_EXT_BGRA ||
+               cinfo->in_color_space == JCS_EXT_ABGR ||
+               cinfo->in_color_space == JCS_EXT_ARGB)
+      cconvert->pub.color_convert = rgb_rgb_convert;
+    else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  case JCS_YCbCr:
+    if (cinfo->num_components != 3)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    if (cinfo->in_color_space == JCS_RGB ||
+        cinfo->in_color_space == JCS_EXT_RGB ||
+        cinfo->in_color_space == JCS_EXT_RGBX ||
+        cinfo->in_color_space == JCS_EXT_BGR ||
+        cinfo->in_color_space == JCS_EXT_BGRX ||
+        cinfo->in_color_space == JCS_EXT_XBGR ||
+        cinfo->in_color_space == JCS_EXT_XRGB ||
+        cinfo->in_color_space == JCS_EXT_RGBA ||
+        cinfo->in_color_space == JCS_EXT_BGRA ||
+        cinfo->in_color_space == JCS_EXT_ABGR ||
+        cinfo->in_color_space == JCS_EXT_ARGB) {
+      if (jsimd_can_rgb_ycc())
+        cconvert->pub.color_convert = jsimd_rgb_ycc_convert;
+      else {
+        cconvert->pub.start_pass = rgb_ycc_start;
+        cconvert->pub.color_convert = rgb_ycc_convert;
+      }
+    } else if (cinfo->in_color_space == JCS_YCbCr) {
+#if defined(__mips__)
+      if (jsimd_c_can_null_convert())
+        cconvert->pub.color_convert = jsimd_c_null_convert;
+      else
+#endif
+        cconvert->pub.color_convert = null_convert;
+    } else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  case JCS_CMYK:
+    if (cinfo->num_components != 4)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    if (cinfo->in_color_space == JCS_CMYK) {
+#if defined(__mips__)
+      if (jsimd_c_can_null_convert())
+        cconvert->pub.color_convert = jsimd_c_null_convert;
+      else
+#endif
+        cconvert->pub.color_convert = null_convert;
+    } else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  case JCS_YCCK:
+    if (cinfo->num_components != 4)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    if (cinfo->in_color_space == JCS_CMYK) {
+      cconvert->pub.start_pass = rgb_ycc_start;
+      cconvert->pub.color_convert = cmyk_ycck_convert;
+    } else if (cinfo->in_color_space == JCS_YCCK) {
+#if defined(__mips__)
+      if (jsimd_c_can_null_convert())
+        cconvert->pub.color_convert = jsimd_c_null_convert;
+      else
+#endif
+        cconvert->pub.color_convert = null_convert;
+    } else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  default:                      /* allow null conversion of JCS_UNKNOWN */
+    if (cinfo->jpeg_color_space != cinfo->in_color_space ||
+        cinfo->num_components != cinfo->input_components)
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+#if defined(__mips__)
+    if (jsimd_c_can_null_convert())
+      cconvert->pub.color_convert = jsimd_c_null_convert;
+    else
+#endif
+      cconvert->pub.color_convert = null_convert;
+    break;
+  }
+}

3rdparty/libjpeg-turbo/src/jcdctmgr.c

@@ -0,0 +1,721 @@
+/*
+ * jcdctmgr.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 1999-2006, MIYASAKA Masaru.
+ * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * Copyright (C) 2011, 2014-2015, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains the forward-DCT management logic.
+ * This code selects a particular DCT implementation to be used,
+ * and it performs related housekeeping chores including coefficient
+ * quantization.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jdct.h"               /* Private declarations for DCT subsystem */
+#include "jsimddct.h"
+
+
+/* Private subobject for this module */
+
+typedef void (*forward_DCT_method_ptr) (DCTELEM *data);
+typedef void (*float_DCT_method_ptr) (FAST_FLOAT *data);
+
+typedef void (*convsamp_method_ptr) (JSAMPARRAY sample_data,
+                                     JDIMENSION start_col,
+                                     DCTELEM *workspace);
+typedef void (*float_convsamp_method_ptr) (JSAMPARRAY sample_data,
+                                           JDIMENSION start_col,
+                                           FAST_FLOAT *workspace);
+
+typedef void (*quantize_method_ptr) (JCOEFPTR coef_block, DCTELEM *divisors,
+                                     DCTELEM *workspace);
+typedef void (*float_quantize_method_ptr) (JCOEFPTR coef_block,
+                                           FAST_FLOAT *divisors,
+                                           FAST_FLOAT *workspace);
+
+METHODDEF(void) quantize(JCOEFPTR, DCTELEM *, DCTELEM *);
+
+typedef struct {
+  struct jpeg_forward_dct pub;  /* public fields */
+
+  /* Pointer to the DCT routine actually in use */
+  forward_DCT_method_ptr dct;
+  convsamp_method_ptr convsamp;
+  quantize_method_ptr quantize;
+
+  /* The actual post-DCT divisors --- not identical to the quant table
+   * entries, because of scaling (especially for an unnormalized DCT).
+   * Each table is given in normal array order.
+   */
+  DCTELEM *divisors[NUM_QUANT_TBLS];
+
+  /* work area for FDCT subroutine */
+  DCTELEM *workspace;
+
+#ifdef DCT_FLOAT_SUPPORTED
+  /* Same as above for the floating-point case. */
+  float_DCT_method_ptr float_dct;
+  float_convsamp_method_ptr float_convsamp;
+  float_quantize_method_ptr float_quantize;
+  FAST_FLOAT *float_divisors[NUM_QUANT_TBLS];
+  FAST_FLOAT *float_workspace;
+#endif
+} my_fdct_controller;
+
+typedef my_fdct_controller *my_fdct_ptr;
+
+
+#if BITS_IN_JSAMPLE == 8
+
+/*
+ * Find the highest bit in an integer through binary search.
+ */
+
+LOCAL(int)
+flss(UINT16 val)
+{
+  int bit;
+
+  bit = 16;
+
+  if (!val)
+    return 0;
+
+  if (!(val & 0xff00)) {
+    bit -= 8;
+    val <<= 8;
+  }
+  if (!(val & 0xf000)) {
+    bit -= 4;
+    val <<= 4;
+  }
+  if (!(val & 0xc000)) {
+    bit -= 2;
+    val <<= 2;
+  }
+  if (!(val & 0x8000)) {
+    bit -= 1;
+    val <<= 1;
+  }
+
+  return bit;
+}
+
+
+/*
+ * Compute values to do a division using reciprocal.
+ *
+ * This implementation is based on an algorithm described in
+ *   "How to optimize for the Pentium family of microprocessors"
+ *   (http://www.agner.org/assem/).
+ * More information about the basic algorithm can be found in
+ * the paper "Integer Division Using Reciprocals" by Robert Alverson.
+ *
+ * The basic idea is to replace x/d by x * d^-1. In order to store
+ * d^-1 with enough precision we shift it left a few places. It turns
+ * out that this algoright gives just enough precision, and also fits
+ * into DCTELEM:
+ *
+ *   b = (the number of significant bits in divisor) - 1
+ *   r = (word size) + b
+ *   f = 2^r / divisor
+ *
+ * f will not be an integer for most cases, so we need to compensate
+ * for the rounding error introduced:
+ *
+ *   no fractional part:
+ *
+ *       result = input >> r
+ *
+ *   fractional part of f < 0.5:
+ *
+ *       round f down to nearest integer
+ *       result = ((input + 1) * f) >> r
+ *
+ *   fractional part of f > 0.5:
+ *
+ *       round f up to nearest integer
+ *       result = (input * f) >> r
+ *
+ * This is the original algorithm that gives truncated results. But we
+ * want properly rounded results, so we replace "input" with
+ * "input + divisor/2".
+ *
+ * In order to allow SIMD implementations we also tweak the values to
+ * allow the same calculation to be made at all times:
+ *
+ *   dctbl[0] = f rounded to nearest integer
+ *   dctbl[1] = divisor / 2 (+ 1 if fractional part of f < 0.5)
+ *   dctbl[2] = 1 << ((word size) * 2 - r)
+ *   dctbl[3] = r - (word size)
+ *
+ * dctbl[2] is for stupid instruction sets where the shift operation
+ * isn't member wise (e.g. MMX).
+ *
+ * The reason dctbl[2] and dctbl[3] reduce the shift with (word size)
+ * is that most SIMD implementations have a "multiply and store top
+ * half" operation.
+ *
+ * Lastly, we store each of the values in their own table instead
+ * of in a consecutive manner, yet again in order to allow SIMD
+ * routines.
+ */
+
+LOCAL(int)
+compute_reciprocal(UINT16 divisor, DCTELEM *dtbl)
+{
+  UDCTELEM2 fq, fr;
+  UDCTELEM c;
+  int b, r;
+
+  if (divisor == 1) {
+    /* divisor == 1 means unquantized, so these reciprocal/correction/shift
+     * values will cause the C quantization algorithm to act like the
+     * identity function.  Since only the C quantization algorithm is used in
+     * these cases, the scale value is irrelevant.
+     */
+    dtbl[DCTSIZE2 * 0] = (DCTELEM)1;                        /* reciprocal */
+    dtbl[DCTSIZE2 * 1] = (DCTELEM)0;                        /* correction */
+    dtbl[DCTSIZE2 * 2] = (DCTELEM)1;                        /* scale */
+    dtbl[DCTSIZE2 * 3] = -(DCTELEM)(sizeof(DCTELEM) * 8);   /* shift */
+    return 0;
+  }
+
+  b = flss(divisor) - 1;
+  r  = sizeof(DCTELEM) * 8 + b;
+
+  fq = ((UDCTELEM2)1 << r) / divisor;
+  fr = ((UDCTELEM2)1 << r) % divisor;
+
+  c = divisor / 2;                      /* for rounding */
+
+  if (fr == 0) {                        /* divisor is power of two */
+    /* fq will be one bit too large to fit in DCTELEM, so adjust */
+    fq >>= 1;
+    r--;
+  } else if (fr <= (divisor / 2U)) {    /* fractional part is < 0.5 */
+    c++;
+  } else {                              /* fractional part is > 0.5 */
+    fq++;
+  }
+
+  dtbl[DCTSIZE2 * 0] = (DCTELEM)fq;     /* reciprocal */
+  dtbl[DCTSIZE2 * 1] = (DCTELEM)c;      /* correction + roundfactor */
+#ifdef WITH_SIMD
+  dtbl[DCTSIZE2 * 2] = (DCTELEM)(1 << (sizeof(DCTELEM) * 8 * 2 - r)); /* scale */
+#else
+  dtbl[DCTSIZE2 * 2] = 1;
+#endif
+  dtbl[DCTSIZE2 * 3] = (DCTELEM)r - sizeof(DCTELEM) * 8; /* shift */
+
+  if (r <= 16) return 0;
+  else return 1;
+}
+
+#endif
+
+
+/*
+ * Initialize for a processing pass.
+ * Verify that all referenced Q-tables are present, and set up
+ * the divisor table for each one.
+ * In the current implementation, DCT of all components is done during
+ * the first pass, even if only some components will be output in the
+ * first scan.  Hence all components should be examined here.
+ */
+
+METHODDEF(void)
+start_pass_fdctmgr(j_compress_ptr cinfo)
+{
+  my_fdct_ptr fdct = (my_fdct_ptr)cinfo->fdct;
+  int ci, qtblno, i;
+  jpeg_component_info *compptr;
+  JQUANT_TBL *qtbl;
+  DCTELEM *dtbl;
+
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    qtblno = compptr->quant_tbl_no;
+    /* Make sure specified quantization table is present */
+    if (qtblno < 0 || qtblno >= NUM_QUANT_TBLS ||
+        cinfo->quant_tbl_ptrs[qtblno] == NULL)
+      ERREXIT1(cinfo, JERR_NO_QUANT_TABLE, qtblno);
+    qtbl = cinfo->quant_tbl_ptrs[qtblno];
+    /* Compute divisors for this quant table */
+    /* We may do this more than once for same table, but it's not a big deal */
+    switch (cinfo->dct_method) {
+#ifdef DCT_ISLOW_SUPPORTED
+    case JDCT_ISLOW:
+      /* For LL&M IDCT method, divisors are equal to raw quantization
+       * coefficients multiplied by 8 (to counteract scaling).
+       */
+      if (fdct->divisors[qtblno] == NULL) {
+        fdct->divisors[qtblno] = (DCTELEM *)
+          (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                      (DCTSIZE2 * 4) * sizeof(DCTELEM));
+      }
+      dtbl = fdct->divisors[qtblno];
+      for (i = 0; i < DCTSIZE2; i++) {
+#if BITS_IN_JSAMPLE == 8
+        if (!compute_reciprocal(qtbl->quantval[i] << 3, &dtbl[i]) &&
+            fdct->quantize == jsimd_quantize)
+          fdct->quantize = quantize;
+#else
+        dtbl[i] = ((DCTELEM)qtbl->quantval[i]) << 3;
+#endif
+      }
+      break;
+#endif
+#ifdef DCT_IFAST_SUPPORTED
+    case JDCT_IFAST:
+      {
+        /* For AA&N IDCT method, divisors are equal to quantization
+         * coefficients scaled by scalefactor[row]*scalefactor[col], where
+         *   scalefactor[0] = 1
+         *   scalefactor[k] = cos(k*PI/16) * sqrt(2)    for k=1..7
+         * We apply a further scale factor of 8.
+         */
+#define CONST_BITS  14
+        static const INT16 aanscales[DCTSIZE2] = {
+          /* precomputed values scaled up by 14 bits */
+          16384, 22725, 21407, 19266, 16384, 12873,  8867,  4520,
+          22725, 31521, 29692, 26722, 22725, 17855, 12299,  6270,
+          21407, 29692, 27969, 25172, 21407, 16819, 11585,  5906,
+          19266, 26722, 25172, 22654, 19266, 15137, 10426,  5315,
+          16384, 22725, 21407, 19266, 16384, 12873,  8867,  4520,
+          12873, 17855, 16819, 15137, 12873, 10114,  6967,  3552,
+           8867, 12299, 11585, 10426,  8867,  6967,  4799,  2446,
+           4520,  6270,  5906,  5315,  4520,  3552,  2446,  1247
+        };
+        SHIFT_TEMPS
+
+        if (fdct->divisors[qtblno] == NULL) {
+          fdct->divisors[qtblno] = (DCTELEM *)
+            (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                        (DCTSIZE2 * 4) * sizeof(DCTELEM));
+        }
+        dtbl = fdct->divisors[qtblno];
+        for (i = 0; i < DCTSIZE2; i++) {
+#if BITS_IN_JSAMPLE == 8
+          if (!compute_reciprocal(
+                DESCALE(MULTIPLY16V16((JLONG)qtbl->quantval[i],
+                                      (JLONG)aanscales[i]),
+                        CONST_BITS - 3), &dtbl[i]) &&
+              fdct->quantize == jsimd_quantize)
+            fdct->quantize = quantize;
+#else
+          dtbl[i] = (DCTELEM)
+            DESCALE(MULTIPLY16V16((JLONG)qtbl->quantval[i],
+                                  (JLONG)aanscales[i]),
+                    CONST_BITS - 3);
+#endif
+        }
+      }
+      break;
+#endif
+#ifdef DCT_FLOAT_SUPPORTED
+    case JDCT_FLOAT:
+      {
+        /* For float AA&N IDCT method, divisors are equal to quantization
+         * coefficients scaled by scalefactor[row]*scalefactor[col], where
+         *   scalefactor[0] = 1
+         *   scalefactor[k] = cos(k*PI/16) * sqrt(2)    for k=1..7
+         * We apply a further scale factor of 8.
+         * What's actually stored is 1/divisor so that the inner loop can
+         * use a multiplication rather than a division.
+         */
+        FAST_FLOAT *fdtbl;
+        int row, col;
+        static const double aanscalefactor[DCTSIZE] = {
+          1.0, 1.387039845, 1.306562965, 1.175875602,
+          1.0, 0.785694958, 0.541196100, 0.275899379
+        };
+
+        if (fdct->float_divisors[qtblno] == NULL) {
+          fdct->float_divisors[qtblno] = (FAST_FLOAT *)
+            (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                        DCTSIZE2 * sizeof(FAST_FLOAT));
+        }
+        fdtbl = fdct->float_divisors[qtblno];
+        i = 0;
+        for (row = 0; row < DCTSIZE; row++) {
+          for (col = 0; col < DCTSIZE; col++) {
+            fdtbl[i] = (FAST_FLOAT)
+              (1.0 / (((double)qtbl->quantval[i] *
+                       aanscalefactor[row] * aanscalefactor[col] * 8.0)));
+            i++;
+          }
+        }
+      }
+      break;
+#endif
+    default:
+      ERREXIT(cinfo, JERR_NOT_COMPILED);
+      break;
+    }
+  }
+}
+
+
+/*
+ * Load data into workspace, applying unsigned->signed conversion.
+ */
+
+METHODDEF(void)
+convsamp(JSAMPARRAY sample_data, JDIMENSION start_col, DCTELEM *workspace)
+{
+  register DCTELEM *workspaceptr;
+  register JSAMPROW elemptr;
+  register int elemr;
+
+  workspaceptr = workspace;
+  for (elemr = 0; elemr < DCTSIZE; elemr++) {
+    elemptr = sample_data[elemr] + start_col;
+
+#if DCTSIZE == 8                /* unroll the inner loop */
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+#else
+    {
+      register int elemc;
+      for (elemc = DCTSIZE; elemc > 0; elemc--)
+        *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+    }
+#endif
+  }
+}
+
+
+/*
+ * Quantize/descale the coefficients, and store into coef_blocks[].
+ */
+
+METHODDEF(void)
+quantize(JCOEFPTR coef_block, DCTELEM *divisors, DCTELEM *workspace)
+{
+  int i;
+  DCTELEM temp;
+  JCOEFPTR output_ptr = coef_block;
+
+#if BITS_IN_JSAMPLE == 8
+
+  UDCTELEM recip, corr;
+  int shift;
+  UDCTELEM2 product;
+
+  for (i = 0; i < DCTSIZE2; i++) {
+    temp = workspace[i];
+    recip = divisors[i + DCTSIZE2 * 0];
+    corr =  divisors[i + DCTSIZE2 * 1];
+    shift = divisors[i + DCTSIZE2 * 3];
+
+    if (temp < 0) {
+      temp = -temp;
+      product = (UDCTELEM2)(temp + corr) * recip;
+      product >>= shift + sizeof(DCTELEM) * 8;
+      temp = (DCTELEM)product;
+      temp = -temp;
+    } else {
+      product = (UDCTELEM2)(temp + corr) * recip;
+      product >>= shift + sizeof(DCTELEM) * 8;
+      temp = (DCTELEM)product;
+    }
+    output_ptr[i] = (JCOEF)temp;
+  }
+
+#else
+
+  register DCTELEM qval;
+
+  for (i = 0; i < DCTSIZE2; i++) {
+    qval = divisors[i];
+    temp = workspace[i];
+    /* Divide the coefficient value by qval, ensuring proper rounding.
+     * Since C does not specify the direction of rounding for negative
+     * quotients, we have to force the dividend positive for portability.
+     *
+     * In most files, at least half of the output values will be zero
+     * (at default quantization settings, more like three-quarters...)
+     * so we should ensure that this case is fast.  On many machines,
+     * a comparison is enough cheaper than a divide to make a special test
+     * a win.  Since both inputs will be nonnegative, we need only test
+     * for a < b to discover whether a/b is 0.
+     * If your machine's division is fast enough, define FAST_DIVIDE.
+     */
+#ifdef FAST_DIVIDE
+#define DIVIDE_BY(a, b)  a /= b
+#else
+#define DIVIDE_BY(a, b)  if (a >= b) a /= b;  else a = 0
+#endif
+    if (temp < 0) {
+      temp = -temp;
+      temp += qval >> 1;        /* for rounding */
+      DIVIDE_BY(temp, qval);
+      temp = -temp;
+    } else {
+      temp += qval >> 1;        /* for rounding */
+      DIVIDE_BY(temp, qval);
+    }
+    output_ptr[i] = (JCOEF)temp;
+  }
+
+#endif
+
+}
+
+
+/*
+ * Perform forward DCT on one or more blocks of a component.
+ *
+ * The input samples are taken from the sample_data[] array starting at
+ * position start_row/start_col, and moving to the right for any additional
+ * blocks. The quantized coefficients are returned in coef_blocks[].
+ */
+
+METHODDEF(void)
+forward_DCT(j_compress_ptr cinfo, jpeg_component_info *compptr,
+            JSAMPARRAY sample_data, JBLOCKROW coef_blocks,
+            JDIMENSION start_row, JDIMENSION start_col, JDIMENSION num_blocks)
+/* This version is used for integer DCT implementations. */
+{
+  /* This routine is heavily used, so it's worth coding it tightly. */
+  my_fdct_ptr fdct = (my_fdct_ptr)cinfo->fdct;
+  DCTELEM *divisors = fdct->divisors[compptr->quant_tbl_no];
+  DCTELEM *workspace;
+  JDIMENSION bi;
+
+  /* Make sure the compiler doesn't look up these every pass */
+  forward_DCT_method_ptr do_dct = fdct->dct;
+  convsamp_method_ptr do_convsamp = fdct->convsamp;
+  quantize_method_ptr do_quantize = fdct->quantize;
+  workspace = fdct->workspace;
+
+  sample_data += start_row;     /* fold in the vertical offset once */
+
+  for (bi = 0; bi < num_blocks; bi++, start_col += DCTSIZE) {
+    /* Load data into workspace, applying unsigned->signed conversion */
+    (*do_convsamp) (sample_data, start_col, workspace);
+
+    /* Perform the DCT */
+    (*do_dct) (workspace);
+
+    /* Quantize/descale the coefficients, and store into coef_blocks[] */
+    (*do_quantize) (coef_blocks[bi], divisors, workspace);
+  }
+}
+
+
+#ifdef DCT_FLOAT_SUPPORTED
+
+METHODDEF(void)
+convsamp_float(JSAMPARRAY sample_data, JDIMENSION start_col,
+               FAST_FLOAT *workspace)
+{
+  register FAST_FLOAT *workspaceptr;
+  register JSAMPROW elemptr;
+  register int elemr;
+
+  workspaceptr = workspace;
+  for (elemr = 0; elemr < DCTSIZE; elemr++) {
+    elemptr = sample_data[elemr] + start_col;
+#if DCTSIZE == 8                /* unroll the inner loop */
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+#else
+    {
+      register int elemc;
+      for (elemc = DCTSIZE; elemc > 0; elemc--)
+        *workspaceptr++ = (FAST_FLOAT)
+                          (GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+    }
+#endif
+  }
+}
+
+
+METHODDEF(void)
+quantize_float(JCOEFPTR coef_block, FAST_FLOAT *divisors,
+               FAST_FLOAT *workspace)
+{
+  register FAST_FLOAT temp;
+  register int i;
+  register JCOEFPTR output_ptr = coef_block;
+
+  for (i = 0; i < DCTSIZE2; i++) {
+    /* Apply the quantization and scaling factor */
+    temp = workspace[i] * divisors[i];
+
+    /* Round to nearest integer.
+     * Since C does not specify the direction of rounding for negative
+     * quotients, we have to force the dividend positive for portability.
+     * The maximum coefficient size is +-16K (for 12-bit data), so this
+     * code should work for either 16-bit or 32-bit ints.
+     */
+    output_ptr[i] = (JCOEF)((int)(temp + (FAST_FLOAT)16384.5) - 16384);
+  }
+}
+
+
+METHODDEF(void)
+forward_DCT_float(j_compress_ptr cinfo, jpeg_component_info *compptr,
+                  JSAMPARRAY sample_data, JBLOCKROW coef_blocks,
+                  JDIMENSION start_row, JDIMENSION start_col,
+                  JDIMENSION num_blocks)
+/* This version is used for floating-point DCT implementations. */
+{
+  /* This routine is heavily used, so it's worth coding it tightly. */
+  my_fdct_ptr fdct = (my_fdct_ptr)cinfo->fdct;
+  FAST_FLOAT *divisors = fdct->float_divisors[compptr->quant_tbl_no];
+  FAST_FLOAT *workspace;
+  JDIMENSION bi;
+
+
+  /* Make sure the compiler doesn't look up these every pass */
+  float_DCT_method_ptr do_dct = fdct->float_dct;
+  float_convsamp_method_ptr do_convsamp = fdct->float_convsamp;
+  float_quantize_method_ptr do_quantize = fdct->float_quantize;
+  workspace = fdct->float_workspace;
+
+  sample_data += start_row;     /* fold in the vertical offset once */
+
+  for (bi = 0; bi < num_blocks; bi++, start_col += DCTSIZE) {
+    /* Load data into workspace, applying unsigned->signed conversion */
+    (*do_convsamp) (sample_data, start_col, workspace);
+
+    /* Perform the DCT */
+    (*do_dct) (workspace);
+
+    /* Quantize/descale the coefficients, and store into coef_blocks[] */
+    (*do_quantize) (coef_blocks[bi], divisors, workspace);
+  }
+}
+
+#endif /* DCT_FLOAT_SUPPORTED */
+
+
+/*
+ * Initialize FDCT manager.
+ */
+
+GLOBAL(void)
+jinit_forward_dct(j_compress_ptr cinfo)
+{
+  my_fdct_ptr fdct;
+  int i;
+
+  fdct = (my_fdct_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_fdct_controller));
+  cinfo->fdct = (struct jpeg_forward_dct *)fdct;
+  fdct->pub.start_pass = start_pass_fdctmgr;
+
+  /* First determine the DCT... */
+  switch (cinfo->dct_method) {
+#ifdef DCT_ISLOW_SUPPORTED
+  case JDCT_ISLOW:
+    fdct->pub.forward_DCT = forward_DCT;
+    if (jsimd_can_fdct_islow())
+      fdct->dct = jsimd_fdct_islow;
+    else
+      fdct->dct = jpeg_fdct_islow;
+    break;
+#endif
+#ifdef DCT_IFAST_SUPPORTED
+  case JDCT_IFAST:
+    fdct->pub.forward_DCT = forward_DCT;
+    if (jsimd_can_fdct_ifast())
+      fdct->dct = jsimd_fdct_ifast;
+    else
+      fdct->dct = jpeg_fdct_ifast;
+    break;
+#endif
+#ifdef DCT_FLOAT_SUPPORTED
+  case JDCT_FLOAT:
+    fdct->pub.forward_DCT = forward_DCT_float;
+    if (jsimd_can_fdct_float())
+      fdct->float_dct = jsimd_fdct_float;
+    else
+      fdct->float_dct = jpeg_fdct_float;
+    break;
+#endif
+  default:
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+    break;
+  }
+
+  /* ...then the supporting stages. */
+  switch (cinfo->dct_method) {
+#ifdef DCT_ISLOW_SUPPORTED
+  case JDCT_ISLOW:
+#endif
+#ifdef DCT_IFAST_SUPPORTED
+  case JDCT_IFAST:
+#endif
+#if defined(DCT_ISLOW_SUPPORTED) || defined(DCT_IFAST_SUPPORTED)
+    if (jsimd_can_convsamp())
+      fdct->convsamp = jsimd_convsamp;
+    else
+      fdct->convsamp = convsamp;
+    if (jsimd_can_quantize())
+      fdct->quantize = jsimd_quantize;
+    else
+      fdct->quantize = quantize;
+    break;
+#endif
+#ifdef DCT_FLOAT_SUPPORTED
+  case JDCT_FLOAT:
+    if (jsimd_can_convsamp_float())
+      fdct->float_convsamp = jsimd_convsamp_float;
+    else
+      fdct->float_convsamp = convsamp_float;
+    if (jsimd_can_quantize_float())
+      fdct->float_quantize = jsimd_quantize_float;
+    else
+      fdct->float_quantize = quantize_float;
+    break;
+#endif
+  default:
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+    break;
+  }
+
+  /* Allocate workspace memory */
+#ifdef DCT_FLOAT_SUPPORTED
+  if (cinfo->dct_method == JDCT_FLOAT)
+    fdct->float_workspace = (FAST_FLOAT *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                  sizeof(FAST_FLOAT) * DCTSIZE2);
+  else
+#endif
+    fdct->workspace = (DCTELEM *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                  sizeof(DCTELEM) * DCTSIZE2);
+
+  /* Mark divisor tables unallocated */
+  for (i = 0; i < NUM_QUANT_TBLS; i++) {
+    fdct->divisors[i] = NULL;
+#ifdef DCT_FLOAT_SUPPORTED
+    fdct->float_divisors[i] = NULL;
+#endif
+  }
+}

3rdparty/libjpeg-turbo/src/jchuff.h

@@ -0,0 +1,42 @@
+/*
+ * jchuff.h
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1997, Thomas G. Lane.
+ * It was modified by The libjpeg-turbo Project to include only code relevant
+ * to libjpeg-turbo.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains declarations for Huffman entropy encoding routines
+ * that are shared between the sequential encoder (jchuff.c) and the
+ * progressive encoder (jcphuff.c).  No other modules need to see these.
+ */
+
+/* The legal range of a DCT coefficient is
+ *  -1024 .. +1023  for 8-bit data;
+ * -16384 .. +16383 for 12-bit data.
+ * Hence the magnitude should always fit in 10 or 14 bits respectively.
+ */
+
+#if BITS_IN_JSAMPLE == 8
+#define MAX_COEF_BITS  10
+#else
+#define MAX_COEF_BITS  14
+#endif
+
+/* Derived data constructed for each Huffman table */
+
+typedef struct {
+  unsigned int ehufco[256];     /* code for each symbol */
+  char ehufsi[256];             /* length of code for each symbol */
+  /* If no code has been allocated for a symbol S, ehufsi[S] contains 0 */
+} c_derived_tbl;
+
+/* Expand a Huffman table definition into the derived format */
+EXTERN(void) jpeg_make_c_derived_tbl(j_compress_ptr cinfo, boolean isDC,
+                                     int tblno, c_derived_tbl **pdtbl);
+
+/* Generate an optimal table definition given the specified counts */
+EXTERN(void) jpeg_gen_optimal_table(j_compress_ptr cinfo, JHUFF_TBL *htbl,
+                                    long freq[]);

3rdparty/libjpeg-turbo/src/jcinit.c

@@ -0,0 +1,77 @@
+/*
+ * jcinit.c
+ *
+ * Copyright (C) 1991-1997, Thomas G. Lane.
+ * This file is part of the Independent JPEG Group's software.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains initialization logic for the JPEG compressor.
+ * This routine is in charge of selecting the modules to be executed and
+ * making an initialization call to each one.
+ *
+ * Logically, this code belongs in jcmaster.c.  It's split out because
+ * linking this routine implies linking the entire compression library.
+ * For a transcoding-only application, we want to be able to use jcmaster.c
+ * without linking in the whole library.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/*
+ * Master selection of compression modules.
+ * This is done once at the start of processing an image.  We determine
+ * which modules will be used and give them appropriate initialization calls.
+ */
+
+GLOBAL(void)
+jinit_compress_master(j_compress_ptr cinfo)
+{
+  /* Initialize master control (includes parameter checking/processing) */
+  jinit_c_master_control(cinfo, FALSE /* full compression */);
+
+  /* Preprocessing */
+  if (!cinfo->raw_data_in) {
+    jinit_color_converter(cinfo);
+    jinit_downsampler(cinfo);
+    jinit_c_prep_controller(cinfo, FALSE /* never need full buffer here */);
+  }
+  /* Forward DCT */
+  jinit_forward_dct(cinfo);
+  /* Entropy encoding: either Huffman or arithmetic coding. */
+  if (cinfo->arith_code) {
+#ifdef C_ARITH_CODING_SUPPORTED
+    jinit_arith_encoder(cinfo);
+#else
+    ERREXIT(cinfo, JERR_ARITH_NOTIMPL);
+#endif
+  } else {
+    if (cinfo->progressive_mode) {
+#ifdef C_PROGRESSIVE_SUPPORTED
+      jinit_phuff_encoder(cinfo);
+#else
+      ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif
+    } else
+      jinit_huff_encoder(cinfo);
+  }
+
+  /* Need a full-image coefficient buffer in any multi-pass mode. */
+  jinit_c_coef_controller(cinfo, (boolean)(cinfo->num_scans > 1 ||
+                                           cinfo->optimize_coding));
+  jinit_c_main_controller(cinfo, FALSE /* never need full buffer here */);
+
+  jinit_marker_writer(cinfo);
+
+  /* We can now tell the memory manager to allocate virtual arrays. */
+  (*cinfo->mem->realize_virt_arrays) ((j_common_ptr)cinfo);
+
+  /* Write the datastream header (SOI) immediately.
+   * Frame and scan headers are postponed till later.
+   * This lets application insert special markers after the SOI.
+   */
+  (*cinfo->marker->write_file_header) (cinfo);
+}

3rdparty/libjpeg-turbo/src/jcmainct.c

@@ -0,0 +1,162 @@
+/*
+ * jcmainct.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * It was modified by The libjpeg-turbo Project to include only code relevant
+ * to libjpeg-turbo.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains the main buffer controller for compression.
+ * The main buffer lies between the pre-processor and the JPEG
+ * compressor proper; it holds downsampled data in the JPEG colorspace.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/* Private buffer controller object */
+
+typedef struct {
+  struct jpeg_c_main_controller pub; /* public fields */
+
+  JDIMENSION cur_iMCU_row;      /* number of current iMCU row */
+  JDIMENSION rowgroup_ctr;      /* counts row groups received in iMCU row */
+  boolean suspended;            /* remember if we suspended output */
+  J_BUF_MODE pass_mode;         /* current operating mode */
+
+  /* If using just a strip buffer, this points to the entire set of buffers
+   * (we allocate one for each component).  In the full-image case, this
+   * points to the currently accessible strips of the virtual arrays.
+   */
+  JSAMPARRAY buffer[MAX_COMPONENTS];
+} my_main_controller;
+
+typedef my_main_controller *my_main_ptr;
+
+
+/* Forward declarations */
+METHODDEF(void) process_data_simple_main(j_compress_ptr cinfo,
+                                         JSAMPARRAY input_buf,
+                                         JDIMENSION *in_row_ctr,
+                                         JDIMENSION in_rows_avail);
+
+
+/*
+ * Initialize for a processing pass.
+ */
+
+METHODDEF(void)
+start_pass_main(j_compress_ptr cinfo, J_BUF_MODE pass_mode)
+{
+  my_main_ptr main_ptr = (my_main_ptr)cinfo->main;
+
+  /* Do nothing in raw-data mode. */
+  if (cinfo->raw_data_in)
+    return;
+
+  if (pass_mode != JBUF_PASS_THRU)
+    ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+
+  main_ptr->cur_iMCU_row = 0;   /* initialize counters */
+  main_ptr->rowgroup_ctr = 0;
+  main_ptr->suspended = FALSE;
+  main_ptr->pass_mode = pass_mode;      /* save mode for use by process_data */
+  main_ptr->pub.process_data = process_data_simple_main;
+}
+
+
+/*
+ * Process some data.
+ * This routine handles the simple pass-through mode,
+ * where we have only a strip buffer.
+ */
+
+METHODDEF(void)
+process_data_simple_main(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                         JDIMENSION *in_row_ctr, JDIMENSION in_rows_avail)
+{
+  my_main_ptr main_ptr = (my_main_ptr)cinfo->main;
+
+  while (main_ptr->cur_iMCU_row < cinfo->total_iMCU_rows) {
+    /* Read input data if we haven't filled the main buffer yet */
+    if (main_ptr->rowgroup_ctr < DCTSIZE)
+      (*cinfo->prep->pre_process_data) (cinfo, input_buf, in_row_ctr,
+                                        in_rows_avail, main_ptr->buffer,
+                                        &main_ptr->rowgroup_ctr,
+                                        (JDIMENSION)DCTSIZE);
+
+    /* If we don't have a full iMCU row buffered, return to application for
+     * more data.  Note that preprocessor will always pad to fill the iMCU row
+     * at the bottom of the image.
+     */
+    if (main_ptr->rowgroup_ctr != DCTSIZE)
+      return;
+
+    /* Send the completed row to the compressor */
+    if (!(*cinfo->coef->compress_data) (cinfo, main_ptr->buffer)) {
+      /* If compressor did not consume the whole row, then we must need to
+       * suspend processing and return to the application.  In this situation
+       * we pretend we didn't yet consume the last input row; otherwise, if
+       * it happened to be the last row of the image, the application would
+       * think we were done.
+       */
+      if (!main_ptr->suspended) {
+        (*in_row_ctr)--;
+        main_ptr->suspended = TRUE;
+      }
+      return;
+    }
+    /* We did finish the row.  Undo our little suspension hack if a previous
+     * call suspended; then mark the main buffer empty.
+     */
+    if (main_ptr->suspended) {
+      (*in_row_ctr)++;
+      main_ptr->suspended = FALSE;
+    }
+    main_ptr->rowgroup_ctr = 0;
+    main_ptr->cur_iMCU_row++;
+  }
+}
+
+
+/*
+ * Initialize main buffer controller.
+ */
+
+GLOBAL(void)
+jinit_c_main_controller(j_compress_ptr cinfo, boolean need_full_buffer)
+{
+  my_main_ptr main_ptr;
+  int ci;
+  jpeg_component_info *compptr;
+
+  main_ptr = (my_main_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_main_controller));
+  cinfo->main = (struct jpeg_c_main_controller *)main_ptr;
+  main_ptr->pub.start_pass = start_pass_main;
+
+  /* We don't need to create a buffer in raw-data mode. */
+  if (cinfo->raw_data_in)
+    return;
+
+  /* Create the buffer.  It holds downsampled data, so each component
+   * may be of a different size.
+   */
+  if (need_full_buffer) {
+    ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+  } else {
+    /* Allocate a strip buffer for each component */
+    for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+         ci++, compptr++) {
+      main_ptr->buffer[ci] = (*cinfo->mem->alloc_sarray)
+        ((j_common_ptr)cinfo, JPOOL_IMAGE,
+         compptr->width_in_blocks * DCTSIZE,
+         (JDIMENSION)(compptr->v_samp_factor * DCTSIZE));
+    }
+  }
+}

3rdparty/libjpeg-turbo/src/jcmarker.c

@@ -0,0 +1,664 @@
+/*
+ * jcmarker.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1998, Thomas G. Lane.
+ * Modified 2003-2010 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2010, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains routines to write JPEG datastream markers.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jpegcomp.h"
+
+
+typedef enum {                  /* JPEG marker codes */
+  M_SOF0  = 0xc0,
+  M_SOF1  = 0xc1,
+  M_SOF2  = 0xc2,
+  M_SOF3  = 0xc3,
+
+  M_SOF5  = 0xc5,
+  M_SOF6  = 0xc6,
+  M_SOF7  = 0xc7,
+
+  M_JPG   = 0xc8,
+  M_SOF9  = 0xc9,
+  M_SOF10 = 0xca,
+  M_SOF11 = 0xcb,
+
+  M_SOF13 = 0xcd,
+  M_SOF14 = 0xce,
+  M_SOF15 = 0xcf,
+
+  M_DHT   = 0xc4,
+
+  M_DAC   = 0xcc,
+
+  M_RST0  = 0xd0,
+  M_RST1  = 0xd1,
+  M_RST2  = 0xd2,
+  M_RST3  = 0xd3,
+  M_RST4  = 0xd4,
+  M_RST5  = 0xd5,
+  M_RST6  = 0xd6,
+  M_RST7  = 0xd7,
+
+  M_SOI   = 0xd8,
+  M_EOI   = 0xd9,
+  M_SOS   = 0xda,
+  M_DQT   = 0xdb,
+  M_DNL   = 0xdc,
+  M_DRI   = 0xdd,
+  M_DHP   = 0xde,
+  M_EXP   = 0xdf,
+
+  M_APP0  = 0xe0,
+  M_APP1  = 0xe1,
+  M_APP2  = 0xe2,
+  M_APP3  = 0xe3,
+  M_APP4  = 0xe4,
+  M_APP5  = 0xe5,
+  M_APP6  = 0xe6,
+  M_APP7  = 0xe7,
+  M_APP8  = 0xe8,
+  M_APP9  = 0xe9,
+  M_APP10 = 0xea,
+  M_APP11 = 0xeb,
+  M_APP12 = 0xec,
+  M_APP13 = 0xed,
+  M_APP14 = 0xee,
+  M_APP15 = 0xef,
+
+  M_JPG0  = 0xf0,
+  M_JPG13 = 0xfd,
+  M_COM   = 0xfe,
+
+  M_TEM   = 0x01,
+
+  M_ERROR = 0x100
+} JPEG_MARKER;
+
+
+/* Private state */
+
+typedef struct {
+  struct jpeg_marker_writer pub; /* public fields */
+
+  unsigned int last_restart_interval; /* last DRI value emitted; 0 after SOI */
+} my_marker_writer;
+
+typedef my_marker_writer *my_marker_ptr;
+
+
+/*
+ * Basic output routines.
+ *
+ * Note that we do not support suspension while writing a marker.
+ * Therefore, an application using suspension must ensure that there is
+ * enough buffer space for the initial markers (typ. 600-700 bytes) before
+ * calling jpeg_start_compress, and enough space to write the trailing EOI
+ * (a few bytes) before calling jpeg_finish_compress.  Multipass compression
+ * modes are not supported at all with suspension, so those two are the only
+ * points where markers will be written.
+ */
+
+LOCAL(void)
+emit_byte(j_compress_ptr cinfo, int val)
+/* Emit a byte */
+{
+  struct jpeg_destination_mgr *dest = cinfo->dest;
+
+  *(dest->next_output_byte)++ = (JOCTET)val;
+  if (--dest->free_in_buffer == 0) {
+    if (!(*dest->empty_output_buffer) (cinfo))
+      ERREXIT(cinfo, JERR_CANT_SUSPEND);
+  }
+}
+
+
+LOCAL(void)
+emit_marker(j_compress_ptr cinfo, JPEG_MARKER mark)
+/* Emit a marker code */
+{
+  emit_byte(cinfo, 0xFF);
+  emit_byte(cinfo, (int)mark);
+}
+
+
+LOCAL(void)
+emit_2bytes(j_compress_ptr cinfo, int value)
+/* Emit a 2-byte integer; these are always MSB first in JPEG files */
+{
+  emit_byte(cinfo, (value >> 8) & 0xFF);
+  emit_byte(cinfo, value & 0xFF);
+}
+
+
+/*
+ * Routines to write specific marker types.
+ */
+
+LOCAL(int)
+emit_dqt(j_compress_ptr cinfo, int index)
+/* Emit a DQT marker */
+/* Returns the precision used (0 = 8bits, 1 = 16bits) for baseline checking */
+{
+  JQUANT_TBL *qtbl = cinfo->quant_tbl_ptrs[index];
+  int prec;
+  int i;
+
+  if (qtbl == NULL)
+    ERREXIT1(cinfo, JERR_NO_QUANT_TABLE, index);
+
+  prec = 0;
+  for (i = 0; i < DCTSIZE2; i++) {
+    if (qtbl->quantval[i] > 255)
+      prec = 1;
+  }
+
+  if (!qtbl->sent_table) {
+    emit_marker(cinfo, M_DQT);
+
+    emit_2bytes(cinfo, prec ? DCTSIZE2 * 2 + 1 + 2 : DCTSIZE2 + 1 + 2);
+
+    emit_byte(cinfo, index + (prec << 4));
+
+    for (i = 0; i < DCTSIZE2; i++) {
+      /* The table entries must be emitted in zigzag order. */
+      unsigned int qval = qtbl->quantval[jpeg_natural_order[i]];
+      if (prec)
+        emit_byte(cinfo, (int)(qval >> 8));
+      emit_byte(cinfo, (int)(qval & 0xFF));
+    }
+
+    qtbl->sent_table = TRUE;
+  }
+
+  return prec;
+}
+
+
+LOCAL(void)
+emit_dht(j_compress_ptr cinfo, int index, boolean is_ac)
+/* Emit a DHT marker */
+{
+  JHUFF_TBL *htbl;
+  int length, i;
+
+  if (is_ac) {
+    htbl = cinfo->ac_huff_tbl_ptrs[index];
+    index += 0x10;              /* output index has AC bit set */
+  } else {
+    htbl = cinfo->dc_huff_tbl_ptrs[index];
+  }
+
+  if (htbl == NULL)
+    ERREXIT1(cinfo, JERR_NO_HUFF_TABLE, index);
+
+  if (!htbl->sent_table) {
+    emit_marker(cinfo, M_DHT);
+
+    length = 0;
+    for (i = 1; i <= 16; i++)
+      length += htbl->bits[i];
+
+    emit_2bytes(cinfo, length + 2 + 1 + 16);
+    emit_byte(cinfo, index);
+
+    for (i = 1; i <= 16; i++)
+      emit_byte(cinfo, htbl->bits[i]);
+
+    for (i = 0; i < length; i++)
+      emit_byte(cinfo, htbl->huffval[i]);
+
+    htbl->sent_table = TRUE;
+  }
+}
+
+
+LOCAL(void)
+emit_dac(j_compress_ptr cinfo)
+/* Emit a DAC marker */
+/* Since the useful info is so small, we want to emit all the tables in */
+/* one DAC marker.  Therefore this routine does its own scan of the table. */
+{
+#ifdef C_ARITH_CODING_SUPPORTED
+  char dc_in_use[NUM_ARITH_TBLS];
+  char ac_in_use[NUM_ARITH_TBLS];
+  int length, i;
+  jpeg_component_info *compptr;
+
+  for (i = 0; i < NUM_ARITH_TBLS; i++)
+    dc_in_use[i] = ac_in_use[i] = 0;
+
+  for (i = 0; i < cinfo->comps_in_scan; i++) {
+    compptr = cinfo->cur_comp_info[i];
+    /* DC needs no table for refinement scan */
+    if (cinfo->Ss == 0 && cinfo->Ah == 0)
+      dc_in_use[compptr->dc_tbl_no] = 1;
+    /* AC needs no table when not present */
+    if (cinfo->Se)
+      ac_in_use[compptr->ac_tbl_no] = 1;
+  }
+
+  length = 0;
+  for (i = 0; i < NUM_ARITH_TBLS; i++)
+    length += dc_in_use[i] + ac_in_use[i];
+
+  if (length) {
+    emit_marker(cinfo, M_DAC);
+
+    emit_2bytes(cinfo, length * 2 + 2);
+
+    for (i = 0; i < NUM_ARITH_TBLS; i++) {
+      if (dc_in_use[i]) {
+        emit_byte(cinfo, i);
+        emit_byte(cinfo, cinfo->arith_dc_L[i] + (cinfo->arith_dc_U[i] << 4));
+      }
+      if (ac_in_use[i]) {
+        emit_byte(cinfo, i + 0x10);
+        emit_byte(cinfo, cinfo->arith_ac_K[i]);
+      }
+    }
+  }
+#endif /* C_ARITH_CODING_SUPPORTED */
+}
+
+
+LOCAL(void)
+emit_dri(j_compress_ptr cinfo)
+/* Emit a DRI marker */
+{
+  emit_marker(cinfo, M_DRI);
+
+  emit_2bytes(cinfo, 4);        /* fixed length */
+
+  emit_2bytes(cinfo, (int)cinfo->restart_interval);
+}
+
+
+LOCAL(void)
+emit_sof(j_compress_ptr cinfo, JPEG_MARKER code)
+/* Emit a SOF marker */
+{
+  int ci;
+  jpeg_component_info *compptr;
+
+  emit_marker(cinfo, code);
+
+  emit_2bytes(cinfo, 3 * cinfo->num_components + 2 + 5 + 1); /* length */
+
+  /* Make sure image isn't bigger than SOF field can handle */
+  if ((long)cinfo->_jpeg_height > 65535L || (long)cinfo->_jpeg_width > 65535L)
+    ERREXIT1(cinfo, JERR_IMAGE_TOO_BIG, (unsigned int)65535);
+
+  emit_byte(cinfo, cinfo->data_precision);
+  emit_2bytes(cinfo, (int)cinfo->_jpeg_height);
+  emit_2bytes(cinfo, (int)cinfo->_jpeg_width);
+
+  emit_byte(cinfo, cinfo->num_components);
+
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    emit_byte(cinfo, compptr->component_id);
+    emit_byte(cinfo, (compptr->h_samp_factor << 4) + compptr->v_samp_factor);
+    emit_byte(cinfo, compptr->quant_tbl_no);
+  }
+}
+
+
+LOCAL(void)
+emit_sos(j_compress_ptr cinfo)
+/* Emit a SOS marker */
+{
+  int i, td, ta;
+  jpeg_component_info *compptr;
+
+  emit_marker(cinfo, M_SOS);
+
+  emit_2bytes(cinfo, 2 * cinfo->comps_in_scan + 2 + 1 + 3); /* length */
+
+  emit_byte(cinfo, cinfo->comps_in_scan);
+
+  for (i = 0; i < cinfo->comps_in_scan; i++) {
+    compptr = cinfo->cur_comp_info[i];
+    emit_byte(cinfo, compptr->component_id);
+
+    /* We emit 0 for unused field(s); this is recommended by the P&M text
+     * but does not seem to be specified in the standard.
+     */
+
+    /* DC needs no table for refinement scan */
+    td = cinfo->Ss == 0 && cinfo->Ah == 0 ? compptr->dc_tbl_no : 0;
+    /* AC needs no table when not present */
+    ta = cinfo->Se ? compptr->ac_tbl_no : 0;
+
+    emit_byte(cinfo, (td << 4) + ta);
+  }
+
+  emit_byte(cinfo, cinfo->Ss);
+  emit_byte(cinfo, cinfo->Se);
+  emit_byte(cinfo, (cinfo->Ah << 4) + cinfo->Al);
+}
+
+
+LOCAL(void)
+emit_jfif_app0(j_compress_ptr cinfo)
+/* Emit a JFIF-compliant APP0 marker */
+{
+  /*
+   * Length of APP0 block       (2 bytes)
+   * Block ID                   (4 bytes - ASCII "JFIF")
+   * Zero byte                  (1 byte to terminate the ID string)
+   * Version Major, Minor       (2 bytes - major first)
+   * Units                      (1 byte - 0x00 = none, 0x01 = inch, 0x02 = cm)
+   * Xdpu                       (2 bytes - dots per unit horizontal)
+   * Ydpu                       (2 bytes - dots per unit vertical)
+   * Thumbnail X size           (1 byte)
+   * Thumbnail Y size           (1 byte)
+   */
+
+  emit_marker(cinfo, M_APP0);
+
+  emit_2bytes(cinfo, 2 + 4 + 1 + 2 + 1 + 2 + 2 + 1 + 1); /* length */
+
+  emit_byte(cinfo, 0x4A);       /* Identifier: ASCII "JFIF" */
+  emit_byte(cinfo, 0x46);
+  emit_byte(cinfo, 0x49);
+  emit_byte(cinfo, 0x46);
+  emit_byte(cinfo, 0);
+  emit_byte(cinfo, cinfo->JFIF_major_version); /* Version fields */
+  emit_byte(cinfo, cinfo->JFIF_minor_version);
+  emit_byte(cinfo, cinfo->density_unit); /* Pixel size information */
+  emit_2bytes(cinfo, (int)cinfo->X_density);
+  emit_2bytes(cinfo, (int)cinfo->Y_density);
+  emit_byte(cinfo, 0);          /* No thumbnail image */
+  emit_byte(cinfo, 0);
+}
+
+
+LOCAL(void)
+emit_adobe_app14(j_compress_ptr cinfo)
+/* Emit an Adobe APP14 marker */
+{
+  /*
+   * Length of APP14 block      (2 bytes)
+   * Block ID                   (5 bytes - ASCII "Adobe")
+   * Version Number             (2 bytes - currently 100)
+   * Flags0                     (2 bytes - currently 0)
+   * Flags1                     (2 bytes - currently 0)
+   * Color transform            (1 byte)
+   *
+   * Although Adobe TN 5116 mentions Version = 101, all the Adobe files
+   * now in circulation seem to use Version = 100, so that's what we write.
+   *
+   * We write the color transform byte as 1 if the JPEG color space is
+   * YCbCr, 2 if it's YCCK, 0 otherwise.  Adobe's definition has to do with
+   * whether the encoder performed a transformation, which is pretty useless.
+   */
+
+  emit_marker(cinfo, M_APP14);
+
+  emit_2bytes(cinfo, 2 + 5 + 2 + 2 + 2 + 1); /* length */
+
+  emit_byte(cinfo, 0x41);       /* Identifier: ASCII "Adobe" */
+  emit_byte(cinfo, 0x64);
+  emit_byte(cinfo, 0x6F);
+  emit_byte(cinfo, 0x62);
+  emit_byte(cinfo, 0x65);
+  emit_2bytes(cinfo, 100);      /* Version */
+  emit_2bytes(cinfo, 0);        /* Flags0 */
+  emit_2bytes(cinfo, 0);        /* Flags1 */
+  switch (cinfo->jpeg_color_space) {
+  case JCS_YCbCr:
+    emit_byte(cinfo, 1);        /* Color transform = 1 */
+    break;
+  case JCS_YCCK:
+    emit_byte(cinfo, 2);        /* Color transform = 2 */
+    break;
+  default:
+    emit_byte(cinfo, 0);        /* Color transform = 0 */
+    break;
+  }
+}
+
+
+/*
+ * These routines allow writing an arbitrary marker with parameters.
+ * The only intended use is to emit COM or APPn markers after calling
+ * write_file_header and before calling write_frame_header.
+ * Other uses are not guaranteed to produce desirable results.
+ * Counting the parameter bytes properly is the caller's responsibility.
+ */
+
+METHODDEF(void)
+write_marker_header(j_compress_ptr cinfo, int marker, unsigned int datalen)
+/* Emit an arbitrary marker header */
+{
+  if (datalen > (unsigned int)65533)            /* safety check */
+    ERREXIT(cinfo, JERR_BAD_LENGTH);
+
+  emit_marker(cinfo, (JPEG_MARKER)marker);
+
+  emit_2bytes(cinfo, (int)(datalen + 2));       /* total length */
+}
+
+METHODDEF(void)
+write_marker_byte(j_compress_ptr cinfo, int val)
+/* Emit one byte of marker parameters following write_marker_header */
+{
+  emit_byte(cinfo, val);
+}
+
+
+/*
+ * Write datastream header.
+ * This consists of an SOI and optional APPn markers.
+ * We recommend use of the JFIF marker, but not the Adobe marker,
+ * when using YCbCr or grayscale data.  The JFIF marker should NOT
+ * be used for any other JPEG colorspace.  The Adobe marker is helpful
+ * to distinguish RGB, CMYK, and YCCK colorspaces.
+ * Note that an application can write additional header markers after
+ * jpeg_start_compress returns.
+ */
+
+METHODDEF(void)
+write_file_header(j_compress_ptr cinfo)
+{
+  my_marker_ptr marker = (my_marker_ptr)cinfo->marker;
+
+  emit_marker(cinfo, M_SOI);    /* first the SOI */
+
+  /* SOI is defined to reset restart interval to 0 */
+  marker->last_restart_interval = 0;
+
+  if (cinfo->write_JFIF_header) /* next an optional JFIF APP0 */
+    emit_jfif_app0(cinfo);
+  if (cinfo->write_Adobe_marker) /* next an optional Adobe APP14 */
+    emit_adobe_app14(cinfo);
+}
+
+
+/*
+ * Write frame header.
+ * This consists of DQT and SOFn markers.
+ * Note that we do not emit the SOF until we have emitted the DQT(s).
+ * This avoids compatibility problems with incorrect implementations that
+ * try to error-check the quant table numbers as soon as they see the SOF.
+ */
+
+METHODDEF(void)
+write_frame_header(j_compress_ptr cinfo)
+{
+  int ci, prec;
+  boolean is_baseline;
+  jpeg_component_info *compptr;
+
+  /* Emit DQT for each quantization table.
+   * Note that emit_dqt() suppresses any duplicate tables.
+   */
+  prec = 0;
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    prec += emit_dqt(cinfo, compptr->quant_tbl_no);
+  }
+  /* now prec is nonzero iff there are any 16-bit quant tables. */
+
+  /* Check for a non-baseline specification.
+   * Note we assume that Huffman table numbers won't be changed later.
+   */
+  if (cinfo->arith_code || cinfo->progressive_mode ||
+      cinfo->data_precision != 8) {
+    is_baseline = FALSE;
+  } else {
+    is_baseline = TRUE;
+    for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+         ci++, compptr++) {
+      if (compptr->dc_tbl_no > 1 || compptr->ac_tbl_no > 1)
+        is_baseline = FALSE;
+    }
+    if (prec && is_baseline) {
+      is_baseline = FALSE;
+      /* If it's baseline except for quantizer size, warn the user */
+      TRACEMS(cinfo, 0, JTRC_16BIT_TABLES);
+    }
+  }
+
+  /* Emit the proper SOF marker */
+  if (cinfo->arith_code) {
+    if (cinfo->progressive_mode)
+      emit_sof(cinfo, M_SOF10); /* SOF code for progressive arithmetic */
+    else
+      emit_sof(cinfo, M_SOF9);  /* SOF code for sequential arithmetic */
+  } else {
+    if (cinfo->progressive_mode)
+      emit_sof(cinfo, M_SOF2);  /* SOF code for progressive Huffman */
+    else if (is_baseline)
+      emit_sof(cinfo, M_SOF0);  /* SOF code for baseline implementation */
+    else
+      emit_sof(cinfo, M_SOF1);  /* SOF code for non-baseline Huffman file */
+  }
+}
+
+
+/*
+ * Write scan header.
+ * This consists of DHT or DAC markers, optional DRI, and SOS.
+ * Compressed data will be written following the SOS.
+ */
+
+METHODDEF(void)
+write_scan_header(j_compress_ptr cinfo)
+{
+  my_marker_ptr marker = (my_marker_ptr)cinfo->marker;
+  int i;
+  jpeg_component_info *compptr;
+
+  if (cinfo->arith_code) {
+    /* Emit arith conditioning info.  We may have some duplication
+     * if the file has multiple scans, but it's so small it's hardly
+     * worth worrying about.
+     */
+    emit_dac(cinfo);
+  } else {
+    /* Emit Huffman tables.
+     * Note that emit_dht() suppresses any duplicate tables.
+     */
+    for (i = 0; i < cinfo->comps_in_scan; i++) {
+      compptr = cinfo->cur_comp_info[i];
+      /* DC needs no table for refinement scan */
+      if (cinfo->Ss == 0 && cinfo->Ah == 0)
+        emit_dht(cinfo, compptr->dc_tbl_no, FALSE);
+      /* AC needs no table when not present */
+      if (cinfo->Se)
+        emit_dht(cinfo, compptr->ac_tbl_no, TRUE);
+    }
+  }
+
+  /* Emit DRI if required --- note that DRI value could change for each scan.
+   * We avoid wasting space with unnecessary DRIs, however.
+   */
+  if (cinfo->restart_interval != marker->last_restart_interval) {
+    emit_dri(cinfo);
+    marker->last_restart_interval = cinfo->restart_interval;
+  }
+
+  emit_sos(cinfo);
+}
+
+
+/*
+ * Write datastream trailer.
+ */
+
+METHODDEF(void)
+write_file_trailer(j_compress_ptr cinfo)
+{
+  emit_marker(cinfo, M_EOI);
+}
+
+
+/*
+ * Write an abbreviated table-specification datastream.
+ * This consists of SOI, DQT and DHT tables, and EOI.
+ * Any table that is defined and not marked sent_table = TRUE will be
+ * emitted.  Note that all tables will be marked sent_table = TRUE at exit.
+ */
+
+METHODDEF(void)
+write_tables_only(j_compress_ptr cinfo)
+{
+  int i;
+
+  emit_marker(cinfo, M_SOI);
+
+  for (i = 0; i < NUM_QUANT_TBLS; i++) {
+    if (cinfo->quant_tbl_ptrs[i] != NULL)
+      (void)emit_dqt(cinfo, i);
+  }
+
+  if (!cinfo->arith_code) {
+    for (i = 0; i < NUM_HUFF_TBLS; i++) {
+      if (cinfo->dc_huff_tbl_ptrs[i] != NULL)
+        emit_dht(cinfo, i, FALSE);
+      if (cinfo->ac_huff_tbl_ptrs[i] != NULL)
+        emit_dht(cinfo, i, TRUE);
+    }
+  }
+
+  emit_marker(cinfo, M_EOI);
+}
+
+
+/*
+ * Initialize the marker writer module.
+ */
+
+GLOBAL(void)
+jinit_marker_writer(j_compress_ptr cinfo)
+{
+  my_marker_ptr marker;
+
+  /* Create the subobject */
+  marker = (my_marker_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_marker_writer));
+  cinfo->marker = (struct jpeg_marker_writer *)marker;
+  /* Initialize method pointers */
+  marker->pub.write_file_header = write_file_header;
+  marker->pub.write_frame_header = write_frame_header;
+  marker->pub.write_scan_header = write_scan_header;
+  marker->pub.write_file_trailer = write_file_trailer;
+  marker->pub.write_tables_only = write_tables_only;
+  marker->pub.write_marker_header = write_marker_header;
+  marker->pub.write_marker_byte = write_marker_byte;
+  /* Initialize private state */
+  marker->last_restart_interval = 0;
+}

3rdparty/libjpeg-turbo/src/jcmaster.c

@@ -0,0 +1,640 @@
+/*
+ * jcmaster.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1997, Thomas G. Lane.
+ * Modified 2003-2010 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2010, 2016, 2018, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains master control logic for the JPEG compressor.
+ * These routines are concerned with parameter validation, initial setup,
+ * and inter-pass control (determining the number of passes and the work
+ * to be done in each pass).
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jpegcomp.h"
+#include "jconfigint.h"
+
+
+/* Private state */
+
+typedef enum {
+  main_pass,                    /* input data, also do first output step */
+  huff_opt_pass,                /* Huffman code optimization pass */
+  output_pass                   /* data output pass */
+} c_pass_type;
+
+typedef struct {
+  struct jpeg_comp_master pub;  /* public fields */
+
+  c_pass_type pass_type;        /* the type of the current pass */
+
+  int pass_number;              /* # of passes completed */
+  int total_passes;             /* total # of passes needed */
+
+  int scan_number;              /* current index in scan_info[] */
+
+  /*
+   * This is here so we can add libjpeg-turbo version/build information to the
+   * global string table without introducing a new global symbol.  Adding this
+   * information to the global string table allows one to examine a binary
+   * object and determine which version of libjpeg-turbo it was built from or
+   * linked against.
+   */
+  const char *jpeg_version;
+
+} my_comp_master;
+
+typedef my_comp_master *my_master_ptr;
+
+
+/*
+ * Support routines that do various essential calculations.
+ */
+
+#if JPEG_LIB_VERSION >= 70
+/*
+ * Compute JPEG image dimensions and related values.
+ * NOTE: this is exported for possible use by application.
+ * Hence it mustn't do anything that can't be done twice.
+ */
+
+GLOBAL(void)
+jpeg_calc_jpeg_dimensions(j_compress_ptr cinfo)
+/* Do computations that are needed before master selection phase */
+{
+  /* Hardwire it to "no scaling" */
+  cinfo->jpeg_width = cinfo->image_width;
+  cinfo->jpeg_height = cinfo->image_height;
+  cinfo->min_DCT_h_scaled_size = DCTSIZE;
+  cinfo->min_DCT_v_scaled_size = DCTSIZE;
+}
+#endif
+
+
+LOCAL(void)
+initial_setup(j_compress_ptr cinfo, boolean transcode_only)
+/* Do computations that are needed before master selection phase */
+{
+  int ci;
+  jpeg_component_info *compptr;
+  long samplesperrow;
+  JDIMENSION jd_samplesperrow;
+
+#if JPEG_LIB_VERSION >= 70
+#if JPEG_LIB_VERSION >= 80
+  if (!transcode_only)
+#endif
+    jpeg_calc_jpeg_dimensions(cinfo);
+#endif
+
+  /* Sanity check on image dimensions */
+  if (cinfo->_jpeg_height <= 0 || cinfo->_jpeg_width <= 0 ||
+      cinfo->num_components <= 0 || cinfo->input_components <= 0)
+    ERREXIT(cinfo, JERR_EMPTY_IMAGE);
+
+  /* Make sure image isn't bigger than I can handle */
+  if ((long)cinfo->_jpeg_height > (long)JPEG_MAX_DIMENSION ||
+      (long)cinfo->_jpeg_width > (long)JPEG_MAX_DIMENSION)
+    ERREXIT1(cinfo, JERR_IMAGE_TOO_BIG, (unsigned int)JPEG_MAX_DIMENSION);
+
+  /* Width of an input scanline must be representable as JDIMENSION. */
+  samplesperrow = (long)cinfo->image_width * (long)cinfo->input_components;
+  jd_samplesperrow = (JDIMENSION)samplesperrow;
+  if ((long)jd_samplesperrow != samplesperrow)
+    ERREXIT(cinfo, JERR_WIDTH_OVERFLOW);
+
+  /* For now, precision must match compiled-in value... */
+  if (cinfo->data_precision != BITS_IN_JSAMPLE)
+    ERREXIT1(cinfo, JERR_BAD_PRECISION, cinfo->data_precision);
+
+  /* Check that number of components won't exceed internal array sizes */
+  if (cinfo->num_components > MAX_COMPONENTS)
+    ERREXIT2(cinfo, JERR_COMPONENT_COUNT, cinfo->num_components,
+             MAX_COMPONENTS);
+
+  /* Compute maximum sampling factors; check factor validity */
+  cinfo->max_h_samp_factor = 1;
+  cinfo->max_v_samp_factor = 1;
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    if (compptr->h_samp_factor <= 0 ||
+        compptr->h_samp_factor > MAX_SAMP_FACTOR ||
+        compptr->v_samp_factor <= 0 ||
+        compptr->v_samp_factor > MAX_SAMP_FACTOR)
+      ERREXIT(cinfo, JERR_BAD_SAMPLING);
+    cinfo->max_h_samp_factor = MAX(cinfo->max_h_samp_factor,
+                                   compptr->h_samp_factor);
+    cinfo->max_v_samp_factor = MAX(cinfo->max_v_samp_factor,
+                                   compptr->v_samp_factor);
+  }
+
+  /* Compute dimensions of components */
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    /* Fill in the correct component_index value; don't rely on application */
+    compptr->component_index = ci;
+    /* For compression, we never do DCT scaling. */
+#if JPEG_LIB_VERSION >= 70
+    compptr->DCT_h_scaled_size = compptr->DCT_v_scaled_size = DCTSIZE;
+#else
+    compptr->DCT_scaled_size = DCTSIZE;
+#endif
+    /* Size in DCT blocks */
+    compptr->width_in_blocks = (JDIMENSION)
+      jdiv_round_up((long)cinfo->_jpeg_width * (long)compptr->h_samp_factor,
+                    (long)(cinfo->max_h_samp_factor * DCTSIZE));
+    compptr->height_in_blocks = (JDIMENSION)
+      jdiv_round_up((long)cinfo->_jpeg_height * (long)compptr->v_samp_factor,
+                    (long)(cinfo->max_v_samp_factor * DCTSIZE));
+    /* Size in samples */
+    compptr->downsampled_width = (JDIMENSION)
+      jdiv_round_up((long)cinfo->_jpeg_width * (long)compptr->h_samp_factor,
+                    (long)cinfo->max_h_samp_factor);
+    compptr->downsampled_height = (JDIMENSION)
+      jdiv_round_up((long)cinfo->_jpeg_height * (long)compptr->v_samp_factor,
+                    (long)cinfo->max_v_samp_factor);
+    /* Mark component needed (this flag isn't actually used for compression) */
+    compptr->component_needed = TRUE;
+  }
+
+  /* Compute number of fully interleaved MCU rows (number of times that
+   * main controller will call coefficient controller).
+   */
+  cinfo->total_iMCU_rows = (JDIMENSION)
+    jdiv_round_up((long)cinfo->_jpeg_height,
+                  (long)(cinfo->max_v_samp_factor * DCTSIZE));
+}
+
+
+#ifdef C_MULTISCAN_FILES_SUPPORTED
+
+LOCAL(void)
+validate_script(j_compress_ptr cinfo)
+/* Verify that the scan script in cinfo->scan_info[] is valid; also
+ * determine whether it uses progressive JPEG, and set cinfo->progressive_mode.
+ */
+{
+  const jpeg_scan_info *scanptr;
+  int scanno, ncomps, ci, coefi, thisi;
+  int Ss, Se, Ah, Al;
+  boolean component_sent[MAX_COMPONENTS];
+#ifdef C_PROGRESSIVE_SUPPORTED
+  int *last_bitpos_ptr;
+  int last_bitpos[MAX_COMPONENTS][DCTSIZE2];
+  /* -1 until that coefficient has been seen; then last Al for it */
+#endif
+
+  if (cinfo->num_scans <= 0)
+    ERREXIT1(cinfo, JERR_BAD_SCAN_SCRIPT, 0);
+
+  /* For sequential JPEG, all scans must have Ss=0, Se=DCTSIZE2-1;
+   * for progressive JPEG, no scan can have this.
+   */
+  scanptr = cinfo->scan_info;
+  if (scanptr->Ss != 0 || scanptr->Se != DCTSIZE2 - 1) {
+#ifdef C_PROGRESSIVE_SUPPORTED
+    cinfo->progressive_mode = TRUE;
+    last_bitpos_ptr = &last_bitpos[0][0];
+    for (ci = 0; ci < cinfo->num_components; ci++)
+      for (coefi = 0; coefi < DCTSIZE2; coefi++)
+        *last_bitpos_ptr++ = -1;
+#else
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif
+  } else {
+    cinfo->progressive_mode = FALSE;
+    for (ci = 0; ci < cinfo->num_components; ci++)
+      component_sent[ci] = FALSE;
+  }
+
+  for (scanno = 1; scanno <= cinfo->num_scans; scanptr++, scanno++) {
+    /* Validate component indexes */
+    ncomps = scanptr->comps_in_scan;
+    if (ncomps <= 0 || ncomps > MAX_COMPS_IN_SCAN)
+      ERREXIT2(cinfo, JERR_COMPONENT_COUNT, ncomps, MAX_COMPS_IN_SCAN);
+    for (ci = 0; ci < ncomps; ci++) {
+      thisi = scanptr->component_index[ci];
+      if (thisi < 0 || thisi >= cinfo->num_components)
+        ERREXIT1(cinfo, JERR_BAD_SCAN_SCRIPT, scanno);
+      /* Components must appear in SOF order within each scan */
+      if (ci > 0 && thisi <= scanptr->component_index[ci - 1])
+        ERREXIT1(cinfo, JERR_BAD_SCAN_SCRIPT, scanno);
+    }
+    /* Validate progression parameters */
+    Ss = scanptr->Ss;
+    Se = scanptr->Se;
+    Ah = scanptr->Ah;
+    Al = scanptr->Al;
+    if (cinfo->progressive_mode) {
+#ifdef C_PROGRESSIVE_SUPPORTED
+      /* Rec. ITU-T T.81 | ISO/IEC 10918-1 simply gives the ranges 0..13 for Ah
+       * and Al, but that seems wrong: the upper bound ought to depend on data
+       * precision.  Perhaps they really meant 0..N+1 for N-bit precision.
+       * Here we allow 0..10 for 8-bit data; Al larger than 10 results in
+       * out-of-range reconstructed DC values during the first DC scan,
+       * which might cause problems for some decoders.
+       */
+#if BITS_IN_JSAMPLE == 8
+#define MAX_AH_AL  10
+#else
+#define MAX_AH_AL  13
+#endif
+      if (Ss < 0 || Ss >= DCTSIZE2 || Se < Ss || Se >= DCTSIZE2 ||
+          Ah < 0 || Ah > MAX_AH_AL || Al < 0 || Al > MAX_AH_AL)
+        ERREXIT1(cinfo, JERR_BAD_PROG_SCRIPT, scanno);
+      if (Ss == 0) {
+        if (Se != 0)            /* DC and AC together not OK */
+          ERREXIT1(cinfo, JERR_BAD_PROG_SCRIPT, scanno);
+      } else {
+        if (ncomps != 1)        /* AC scans must be for only one component */
+          ERREXIT1(cinfo, JERR_BAD_PROG_SCRIPT, scanno);
+      }
+      for (ci = 0; ci < ncomps; ci++) {
+        last_bitpos_ptr = &last_bitpos[scanptr->component_index[ci]][0];
+        if (Ss != 0 && last_bitpos_ptr[0] < 0) /* AC without prior DC scan */
+          ERREXIT1(cinfo, JERR_BAD_PROG_SCRIPT, scanno);
+        for (coefi = Ss; coefi <= Se; coefi++) {
+          if (last_bitpos_ptr[coefi] < 0) {
+            /* first scan of this coefficient */
+            if (Ah != 0)
+              ERREXIT1(cinfo, JERR_BAD_PROG_SCRIPT, scanno);
+          } else {
+            /* not first scan */
+            if (Ah != last_bitpos_ptr[coefi] || Al != Ah - 1)
+              ERREXIT1(cinfo, JERR_BAD_PROG_SCRIPT, scanno);
+          }
+          last_bitpos_ptr[coefi] = Al;
+        }
+      }
+#endif
+    } else {
+      /* For sequential JPEG, all progression parameters must be these: */
+      if (Ss != 0 || Se != DCTSIZE2 - 1 || Ah != 0 || Al != 0)
+        ERREXIT1(cinfo, JERR_BAD_PROG_SCRIPT, scanno);
+      /* Make sure components are not sent twice */
+      for (ci = 0; ci < ncomps; ci++) {
+        thisi = scanptr->component_index[ci];
+        if (component_sent[thisi])
+          ERREXIT1(cinfo, JERR_BAD_SCAN_SCRIPT, scanno);
+        component_sent[thisi] = TRUE;
+      }
+    }
+  }
+
+  /* Now verify that everything got sent. */
+  if (cinfo->progressive_mode) {
+#ifdef C_PROGRESSIVE_SUPPORTED
+    /* For progressive mode, we only check that at least some DC data
+     * got sent for each component; the spec does not require that all bits
+     * of all coefficients be transmitted.  Would it be wiser to enforce
+     * transmission of all coefficient bits??
+     */
+    for (ci = 0; ci < cinfo->num_components; ci++) {
+      if (last_bitpos[ci][0] < 0)
+        ERREXIT(cinfo, JERR_MISSING_DATA);
+    }
+#endif
+  } else {
+    for (ci = 0; ci < cinfo->num_components; ci++) {
+      if (!component_sent[ci])
+        ERREXIT(cinfo, JERR_MISSING_DATA);
+    }
+  }
+}
+
+#endif /* C_MULTISCAN_FILES_SUPPORTED */
+
+
+LOCAL(void)
+select_scan_parameters(j_compress_ptr cinfo)
+/* Set up the scan parameters for the current scan */
+{
+  int ci;
+
+#ifdef C_MULTISCAN_FILES_SUPPORTED
+  if (cinfo->scan_info != NULL) {
+    /* Prepare for current scan --- the script is already validated */
+    my_master_ptr master = (my_master_ptr)cinfo->master;
+    const jpeg_scan_info *scanptr = cinfo->scan_info + master->scan_number;
+
+    cinfo->comps_in_scan = scanptr->comps_in_scan;
+    for (ci = 0; ci < scanptr->comps_in_scan; ci++) {
+      cinfo->cur_comp_info[ci] =
+        &cinfo->comp_info[scanptr->component_index[ci]];
+    }
+    cinfo->Ss = scanptr->Ss;
+    cinfo->Se = scanptr->Se;
+    cinfo->Ah = scanptr->Ah;
+    cinfo->Al = scanptr->Al;
+  } else
+#endif
+  {
+    /* Prepare for single sequential-JPEG scan containing all components */
+    if (cinfo->num_components > MAX_COMPS_IN_SCAN)
+      ERREXIT2(cinfo, JERR_COMPONENT_COUNT, cinfo->num_components,
+               MAX_COMPS_IN_SCAN);
+    cinfo->comps_in_scan = cinfo->num_components;
+    for (ci = 0; ci < cinfo->num_components; ci++) {
+      cinfo->cur_comp_info[ci] = &cinfo->comp_info[ci];
+    }
+    cinfo->Ss = 0;
+    cinfo->Se = DCTSIZE2 - 1;
+    cinfo->Ah = 0;
+    cinfo->Al = 0;
+  }
+}
+
+
+LOCAL(void)
+per_scan_setup(j_compress_ptr cinfo)
+/* Do computations that are needed before processing a JPEG scan */
+/* cinfo->comps_in_scan and cinfo->cur_comp_info[] are already set */
+{
+  int ci, mcublks, tmp;
+  jpeg_component_info *compptr;
+
+  if (cinfo->comps_in_scan == 1) {
+
+    /* Noninterleaved (single-component) scan */
+    compptr = cinfo->cur_comp_info[0];
+
+    /* Overall image size in MCUs */
+    cinfo->MCUs_per_row = compptr->width_in_blocks;
+    cinfo->MCU_rows_in_scan = compptr->height_in_blocks;
+
+    /* For noninterleaved scan, always one block per MCU */
+    compptr->MCU_width = 1;
+    compptr->MCU_height = 1;
+    compptr->MCU_blocks = 1;
+    compptr->MCU_sample_width = DCTSIZE;
+    compptr->last_col_width = 1;
+    /* For noninterleaved scans, it is convenient to define last_row_height
+     * as the number of block rows present in the last iMCU row.
+     */
+    tmp = (int)(compptr->height_in_blocks % compptr->v_samp_factor);
+    if (tmp == 0) tmp = compptr->v_samp_factor;
+    compptr->last_row_height = tmp;
+
+    /* Prepare array describing MCU composition */
+    cinfo->blocks_in_MCU = 1;
+    cinfo->MCU_membership[0] = 0;
+
+  } else {
+
+    /* Interleaved (multi-component) scan */
+    if (cinfo->comps_in_scan <= 0 || cinfo->comps_in_scan > MAX_COMPS_IN_SCAN)
+      ERREXIT2(cinfo, JERR_COMPONENT_COUNT, cinfo->comps_in_scan,
+               MAX_COMPS_IN_SCAN);
+
+    /* Overall image size in MCUs */
+    cinfo->MCUs_per_row = (JDIMENSION)
+      jdiv_round_up((long)cinfo->_jpeg_width,
+                    (long)(cinfo->max_h_samp_factor * DCTSIZE));
+    cinfo->MCU_rows_in_scan = (JDIMENSION)
+      jdiv_round_up((long)cinfo->_jpeg_height,
+                    (long)(cinfo->max_v_samp_factor * DCTSIZE));
+
+    cinfo->blocks_in_MCU = 0;
+
+    for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+      compptr = cinfo->cur_comp_info[ci];
+      /* Sampling factors give # of blocks of component in each MCU */
+      compptr->MCU_width = compptr->h_samp_factor;
+      compptr->MCU_height = compptr->v_samp_factor;
+      compptr->MCU_blocks = compptr->MCU_width * compptr->MCU_height;
+      compptr->MCU_sample_width = compptr->MCU_width * DCTSIZE;
+      /* Figure number of non-dummy blocks in last MCU column & row */
+      tmp = (int)(compptr->width_in_blocks % compptr->MCU_width);
+      if (tmp == 0) tmp = compptr->MCU_width;
+      compptr->last_col_width = tmp;
+      tmp = (int)(compptr->height_in_blocks % compptr->MCU_height);
+      if (tmp == 0) tmp = compptr->MCU_height;
+      compptr->last_row_height = tmp;
+      /* Prepare array describing MCU composition */
+      mcublks = compptr->MCU_blocks;
+      if (cinfo->blocks_in_MCU + mcublks > C_MAX_BLOCKS_IN_MCU)
+        ERREXIT(cinfo, JERR_BAD_MCU_SIZE);
+      while (mcublks-- > 0) {
+        cinfo->MCU_membership[cinfo->blocks_in_MCU++] = ci;
+      }
+    }
+
+  }
+
+  /* Convert restart specified in rows to actual MCU count. */
+  /* Note that count must fit in 16 bits, so we provide limiting. */
+  if (cinfo->restart_in_rows > 0) {
+    long nominal = (long)cinfo->restart_in_rows * (long)cinfo->MCUs_per_row;
+    cinfo->restart_interval = (unsigned int)MIN(nominal, 65535L);
+  }
+}
+
+
+/*
+ * Per-pass setup.
+ * This is called at the beginning of each pass.  We determine which modules
+ * will be active during this pass and give them appropriate start_pass calls.
+ * We also set is_last_pass to indicate whether any more passes will be
+ * required.
+ */
+
+METHODDEF(void)
+prepare_for_pass(j_compress_ptr cinfo)
+{
+  my_master_ptr master = (my_master_ptr)cinfo->master;
+
+  switch (master->pass_type) {
+  case main_pass:
+    /* Initial pass: will collect input data, and do either Huffman
+     * optimization or data output for the first scan.
+     */
+    select_scan_parameters(cinfo);
+    per_scan_setup(cinfo);
+    if (!cinfo->raw_data_in) {
+      (*cinfo->cconvert->start_pass) (cinfo);
+      (*cinfo->downsample->start_pass) (cinfo);
+      (*cinfo->prep->start_pass) (cinfo, JBUF_PASS_THRU);
+    }
+    (*cinfo->fdct->start_pass) (cinfo);
+    (*cinfo->entropy->start_pass) (cinfo, cinfo->optimize_coding);
+    (*cinfo->coef->start_pass) (cinfo,
+                                (master->total_passes > 1 ?
+                                 JBUF_SAVE_AND_PASS : JBUF_PASS_THRU));
+    (*cinfo->main->start_pass) (cinfo, JBUF_PASS_THRU);
+    if (cinfo->optimize_coding) {
+      /* No immediate data output; postpone writing frame/scan headers */
+      master->pub.call_pass_startup = FALSE;
+    } else {
+      /* Will write frame/scan headers at first jpeg_write_scanlines call */
+      master->pub.call_pass_startup = TRUE;
+    }
+    break;
+#ifdef ENTROPY_OPT_SUPPORTED
+  case huff_opt_pass:
+    /* Do Huffman optimization for a scan after the first one. */
+    select_scan_parameters(cinfo);
+    per_scan_setup(cinfo);
+    if (cinfo->Ss != 0 || cinfo->Ah == 0 || cinfo->arith_code) {
+      (*cinfo->entropy->start_pass) (cinfo, TRUE);
+      (*cinfo->coef->start_pass) (cinfo, JBUF_CRANK_DEST);
+      master->pub.call_pass_startup = FALSE;
+      break;
+    }
+    /* Special case: Huffman DC refinement scans need no Huffman table
+     * and therefore we can skip the optimization pass for them.
+     */
+    master->pass_type = output_pass;
+    master->pass_number++;
+    /*FALLTHROUGH*/
+#endif
+  case output_pass:
+    /* Do a data-output pass. */
+    /* We need not repeat per-scan setup if prior optimization pass did it. */
+    if (!cinfo->optimize_coding) {
+      select_scan_parameters(cinfo);
+      per_scan_setup(cinfo);
+    }
+    (*cinfo->entropy->start_pass) (cinfo, FALSE);
+    (*cinfo->coef->start_pass) (cinfo, JBUF_CRANK_DEST);
+    /* We emit frame/scan headers now */
+    if (master->scan_number == 0)
+      (*cinfo->marker->write_frame_header) (cinfo);
+    (*cinfo->marker->write_scan_header) (cinfo);
+    master->pub.call_pass_startup = FALSE;
+    break;
+  default:
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+  }
+
+  master->pub.is_last_pass = (master->pass_number == master->total_passes - 1);
+
+  /* Set up progress monitor's pass info if present */
+  if (cinfo->progress != NULL) {
+    cinfo->progress->completed_passes = master->pass_number;
+    cinfo->progress->total_passes = master->total_passes;
+  }
+}
+
+
+/*
+ * Special start-of-pass hook.
+ * This is called by jpeg_write_scanlines if call_pass_startup is TRUE.
+ * In single-pass processing, we need this hook because we don't want to
+ * write frame/scan headers during jpeg_start_compress; we want to let the
+ * application write COM markers etc. between jpeg_start_compress and the
+ * jpeg_write_scanlines loop.
+ * In multi-pass processing, this routine is not used.
+ */
+
+METHODDEF(void)
+pass_startup(j_compress_ptr cinfo)
+{
+  cinfo->master->call_pass_startup = FALSE; /* reset flag so call only once */
+
+  (*cinfo->marker->write_frame_header) (cinfo);
+  (*cinfo->marker->write_scan_header) (cinfo);
+}
+
+
+/*
+ * Finish up at end of pass.
+ */
+
+METHODDEF(void)
+finish_pass_master(j_compress_ptr cinfo)
+{
+  my_master_ptr master = (my_master_ptr)cinfo->master;
+
+  /* The entropy coder always needs an end-of-pass call,
+   * either to analyze statistics or to flush its output buffer.
+   */
+  (*cinfo->entropy->finish_pass) (cinfo);
+
+  /* Update state for next pass */
+  switch (master->pass_type) {
+  case main_pass:
+    /* next pass is either output of scan 0 (after optimization)
+     * or output of scan 1 (if no optimization).
+     */
+    master->pass_type = output_pass;
+    if (!cinfo->optimize_coding)
+      master->scan_number++;
+    break;
+  case huff_opt_pass:
+    /* next pass is always output of current scan */
+    master->pass_type = output_pass;
+    break;
+  case output_pass:
+    /* next pass is either optimization or output of next scan */
+    if (cinfo->optimize_coding)
+      master->pass_type = huff_opt_pass;
+    master->scan_number++;
+    break;
+  }
+
+  master->pass_number++;
+}
+
+
+/*
+ * Initialize master compression control.
+ */
+
+GLOBAL(void)
+jinit_c_master_control(j_compress_ptr cinfo, boolean transcode_only)
+{
+  my_master_ptr master;
+
+  master = (my_master_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_comp_master));
+  cinfo->master = (struct jpeg_comp_master *)master;
+  master->pub.prepare_for_pass = prepare_for_pass;
+  master->pub.pass_startup = pass_startup;
+  master->pub.finish_pass = finish_pass_master;
+  master->pub.is_last_pass = FALSE;
+
+  /* Validate parameters, determine derived values */
+  initial_setup(cinfo, transcode_only);
+
+  if (cinfo->scan_info != NULL) {
+#ifdef C_MULTISCAN_FILES_SUPPORTED
+    validate_script(cinfo);
+#else
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif
+  } else {
+    cinfo->progressive_mode = FALSE;
+    cinfo->num_scans = 1;
+  }
+
+  if (cinfo->progressive_mode && !cinfo->arith_code)  /*  TEMPORARY HACK ??? */
+    cinfo->optimize_coding = TRUE; /* assume default tables no good for progressive mode */
+
+  /* Initialize my private state */
+  if (transcode_only) {
+    /* no main pass in transcoding */
+    if (cinfo->optimize_coding)
+      master->pass_type = huff_opt_pass;
+    else
+      master->pass_type = output_pass;
+  } else {
+    /* for normal compression, first pass is always this type: */
+    master->pass_type = main_pass;
+  }
+  master->scan_number = 0;
+  master->pass_number = 0;
+  if (cinfo->optimize_coding)
+    master->total_passes = cinfo->num_scans * 2;
+  else
+    master->total_passes = cinfo->num_scans;
+
+  master->jpeg_version = PACKAGE_NAME " version " VERSION " (build " BUILD ")";
+}

3rdparty/libjpeg-turbo/src/jcomapi.c

@@ -0,0 +1,109 @@
+/*
+ * jcomapi.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1997, Thomas G. Lane.
+ * It was modified by The libjpeg-turbo Project to include only code relevant
+ * to libjpeg-turbo.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains application interface routines that are used for both
+ * compression and decompression.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/*
+ * Abort processing of a JPEG compression or decompression operation,
+ * but don't destroy the object itself.
+ *
+ * For this, we merely clean up all the nonpermanent memory pools.
+ * Note that temp files (virtual arrays) are not allowed to belong to
+ * the permanent pool, so we will be able to close all temp files here.
+ * Closing a data source or destination, if necessary, is the application's
+ * responsibility.
+ */
+
+GLOBAL(void)
+jpeg_abort(j_common_ptr cinfo)
+{
+  int pool;
+
+  /* Do nothing if called on a not-initialized or destroyed JPEG object. */
+  if (cinfo->mem == NULL)
+    return;
+
+  /* Releasing pools in reverse order might help avoid fragmentation
+   * with some (brain-damaged) malloc libraries.
+   */
+  for (pool = JPOOL_NUMPOOLS - 1; pool > JPOOL_PERMANENT; pool--) {
+    (*cinfo->mem->free_pool) (cinfo, pool);
+  }
+
+  /* Reset overall state for possible reuse of object */
+  if (cinfo->is_decompressor) {
+    cinfo->global_state = DSTATE_START;
+    /* Try to keep application from accessing now-deleted marker list.
+     * A bit kludgy to do it here, but this is the most central place.
+     */
+    ((j_decompress_ptr)cinfo)->marker_list = NULL;
+  } else {
+    cinfo->global_state = CSTATE_START;
+  }
+}
+
+
+/*
+ * Destruction of a JPEG object.
+ *
+ * Everything gets deallocated except the master jpeg_compress_struct itself
+ * and the error manager struct.  Both of these are supplied by the application
+ * and must be freed, if necessary, by the application.  (Often they are on
+ * the stack and so don't need to be freed anyway.)
+ * Closing a data source or destination, if necessary, is the application's
+ * responsibility.
+ */
+
+GLOBAL(void)
+jpeg_destroy(j_common_ptr cinfo)
+{
+  /* We need only tell the memory manager to release everything. */
+  /* NB: mem pointer is NULL if memory mgr failed to initialize. */
+  if (cinfo->mem != NULL)
+    (*cinfo->mem->self_destruct) (cinfo);
+  cinfo->mem = NULL;            /* be safe if jpeg_destroy is called twice */
+  cinfo->global_state = 0;      /* mark it destroyed */
+}
+
+
+/*
+ * Convenience routines for allocating quantization and Huffman tables.
+ * (Would jutils.c be a more reasonable place to put these?)
+ */
+
+GLOBAL(JQUANT_TBL *)
+jpeg_alloc_quant_table(j_common_ptr cinfo)
+{
+  JQUANT_TBL *tbl;
+
+  tbl = (JQUANT_TBL *)
+    (*cinfo->mem->alloc_small) (cinfo, JPOOL_PERMANENT, sizeof(JQUANT_TBL));
+  tbl->sent_table = FALSE;      /* make sure this is false in any new table */
+  return tbl;
+}
+
+
+GLOBAL(JHUFF_TBL *)
+jpeg_alloc_huff_table(j_common_ptr cinfo)
+{
+  JHUFF_TBL *tbl;
+
+  tbl = (JHUFF_TBL *)
+    (*cinfo->mem->alloc_small) (cinfo, JPOOL_PERMANENT, sizeof(JHUFF_TBL));
+  tbl->sent_table = FALSE;      /* make sure this is false in any new table */
+  return tbl;
+}

3rdparty/libjpeg-turbo/src/jcparam.c

@@ -0,0 +1,541 @@
+/*
+ * jcparam.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1998, Thomas G. Lane.
+ * Modified 2003-2008 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2009-2011, 2018, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains optional default-setting code for the JPEG compressor.
+ * Applications do not have to use this file, but those that don't use it
+ * must know a lot more about the innards of the JPEG code.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jstdhuff.c"
+
+
+/*
+ * Quantization table setup routines
+ */
+
+GLOBAL(void)
+jpeg_add_quant_table(j_compress_ptr cinfo, int which_tbl,
+                     const unsigned int *basic_table, int scale_factor,
+                     boolean force_baseline)
+/* Define a quantization table equal to the basic_table times
+ * a scale factor (given as a percentage).
+ * If force_baseline is TRUE, the computed quantization table entries
+ * are limited to 1..255 for JPEG baseline compatibility.
+ */
+{
+  JQUANT_TBL **qtblptr;
+  int i;
+  long temp;
+
+  /* Safety check to ensure start_compress not called yet. */
+  if (cinfo->global_state != CSTATE_START)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  if (which_tbl < 0 || which_tbl >= NUM_QUANT_TBLS)
+    ERREXIT1(cinfo, JERR_DQT_INDEX, which_tbl);
+
+  qtblptr = &cinfo->quant_tbl_ptrs[which_tbl];
+
+  if (*qtblptr == NULL)
+    *qtblptr = jpeg_alloc_quant_table((j_common_ptr)cinfo);
+
+  for (i = 0; i < DCTSIZE2; i++) {
+    temp = ((long)basic_table[i] * scale_factor + 50L) / 100L;
+    /* limit the values to the valid range */
+    if (temp <= 0L) temp = 1L;
+    if (temp > 32767L) temp = 32767L; /* max quantizer needed for 12 bits */
+    if (force_baseline && temp > 255L)
+      temp = 255L;              /* limit to baseline range if requested */
+    (*qtblptr)->quantval[i] = (UINT16)temp;
+  }
+
+  /* Initialize sent_table FALSE so table will be written to JPEG file. */
+  (*qtblptr)->sent_table = FALSE;
+}
+
+
+/* These are the sample quantization tables given in Annex K (Clause K.1) of
+ * Recommendation ITU-T T.81 (1992) | ISO/IEC 10918-1:1994.
+ * The spec says that the values given produce "good" quality, and
+ * when divided by 2, "very good" quality.
+ */
+static const unsigned int std_luminance_quant_tbl[DCTSIZE2] = {
+  16,  11,  10,  16,  24,  40,  51,  61,
+  12,  12,  14,  19,  26,  58,  60,  55,
+  14,  13,  16,  24,  40,  57,  69,  56,
+  14,  17,  22,  29,  51,  87,  80,  62,
+  18,  22,  37,  56,  68, 109, 103,  77,
+  24,  35,  55,  64,  81, 104, 113,  92,
+  49,  64,  78,  87, 103, 121, 120, 101,
+  72,  92,  95,  98, 112, 100, 103,  99
+};
+static const unsigned int std_chrominance_quant_tbl[DCTSIZE2] = {
+  17,  18,  24,  47,  99,  99,  99,  99,
+  18,  21,  26,  66,  99,  99,  99,  99,
+  24,  26,  56,  99,  99,  99,  99,  99,
+  47,  66,  99,  99,  99,  99,  99,  99,
+  99,  99,  99,  99,  99,  99,  99,  99,
+  99,  99,  99,  99,  99,  99,  99,  99,
+  99,  99,  99,  99,  99,  99,  99,  99,
+  99,  99,  99,  99,  99,  99,  99,  99
+};
+
+
+#if JPEG_LIB_VERSION >= 70
+GLOBAL(void)
+jpeg_default_qtables(j_compress_ptr cinfo, boolean force_baseline)
+/* Set or change the 'quality' (quantization) setting, using default tables
+ * and straight percentage-scaling quality scales.
+ * This entry point allows different scalings for luminance and chrominance.
+ */
+{
+  /* Set up two quantization tables using the specified scaling */
+  jpeg_add_quant_table(cinfo, 0, std_luminance_quant_tbl,
+                       cinfo->q_scale_factor[0], force_baseline);
+  jpeg_add_quant_table(cinfo, 1, std_chrominance_quant_tbl,
+                       cinfo->q_scale_factor[1], force_baseline);
+}
+#endif
+
+
+GLOBAL(void)
+jpeg_set_linear_quality(j_compress_ptr cinfo, int scale_factor,
+                        boolean force_baseline)
+/* Set or change the 'quality' (quantization) setting, using default tables
+ * and a straight percentage-scaling quality scale.  In most cases it's better
+ * to use jpeg_set_quality (below); this entry point is provided for
+ * applications that insist on a linear percentage scaling.
+ */
+{
+  /* Set up two quantization tables using the specified scaling */
+  jpeg_add_quant_table(cinfo, 0, std_luminance_quant_tbl,
+                       scale_factor, force_baseline);
+  jpeg_add_quant_table(cinfo, 1, std_chrominance_quant_tbl,
+                       scale_factor, force_baseline);
+}
+
+
+GLOBAL(int)
+jpeg_quality_scaling(int quality)
+/* Convert a user-specified quality rating to a percentage scaling factor
+ * for an underlying quantization table, using our recommended scaling curve.
+ * The input 'quality' factor should be 0 (terrible) to 100 (very good).
+ */
+{
+  /* Safety limit on quality factor.  Convert 0 to 1 to avoid zero divide. */
+  if (quality <= 0) quality = 1;
+  if (quality > 100) quality = 100;
+
+  /* The basic table is used as-is (scaling 100) for a quality of 50.
+   * Qualities 50..100 are converted to scaling percentage 200 - 2*Q;
+   * note that at Q=100 the scaling is 0, which will cause jpeg_add_quant_table
+   * to make all the table entries 1 (hence, minimum quantization loss).
+   * Qualities 1..50 are converted to scaling percentage 5000/Q.
+   */
+  if (quality < 50)
+    quality = 5000 / quality;
+  else
+    quality = 200 - quality * 2;
+
+  return quality;
+}
+
+
+GLOBAL(void)
+jpeg_set_quality(j_compress_ptr cinfo, int quality, boolean force_baseline)
+/* Set or change the 'quality' (quantization) setting, using default tables.
+ * This is the standard quality-adjusting entry point for typical user
+ * interfaces; only those who want detailed control over quantization tables
+ * would use the preceding three routines directly.
+ */
+{
+  /* Convert user 0-100 rating to percentage scaling */
+  quality = jpeg_quality_scaling(quality);
+
+  /* Set up standard quality tables */
+  jpeg_set_linear_quality(cinfo, quality, force_baseline);
+}
+
+
+/*
+ * Default parameter setup for compression.
+ *
+ * Applications that don't choose to use this routine must do their
+ * own setup of all these parameters.  Alternately, you can call this
+ * to establish defaults and then alter parameters selectively.  This
+ * is the recommended approach since, if we add any new parameters,
+ * your code will still work (they'll be set to reasonable defaults).
+ */
+
+GLOBAL(void)
+jpeg_set_defaults(j_compress_ptr cinfo)
+{
+  int i;
+
+  /* Safety check to ensure start_compress not called yet. */
+  if (cinfo->global_state != CSTATE_START)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  /* Allocate comp_info array large enough for maximum component count.
+   * Array is made permanent in case application wants to compress
+   * multiple images at same param settings.
+   */
+  if (cinfo->comp_info == NULL)
+    cinfo->comp_info = (jpeg_component_info *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                                  MAX_COMPONENTS * sizeof(jpeg_component_info));
+
+  /* Initialize everything not dependent on the color space */
+
+#if JPEG_LIB_VERSION >= 70
+  cinfo->scale_num = 1;         /* 1:1 scaling */
+  cinfo->scale_denom = 1;
+#endif
+  cinfo->data_precision = BITS_IN_JSAMPLE;
+  /* Set up two quantization tables using default quality of 75 */
+  jpeg_set_quality(cinfo, 75, TRUE);
+  /* Set up two Huffman tables */
+  std_huff_tables((j_common_ptr)cinfo);
+
+  /* Initialize default arithmetic coding conditioning */
+  for (i = 0; i < NUM_ARITH_TBLS; i++) {
+    cinfo->arith_dc_L[i] = 0;
+    cinfo->arith_dc_U[i] = 1;
+    cinfo->arith_ac_K[i] = 5;
+  }
+
+  /* Default is no multiple-scan output */
+  cinfo->scan_info = NULL;
+  cinfo->num_scans = 0;
+
+  /* Expect normal source image, not raw downsampled data */
+  cinfo->raw_data_in = FALSE;
+
+  /* Use Huffman coding, not arithmetic coding, by default */
+  cinfo->arith_code = FALSE;
+
+  /* By default, don't do extra passes to optimize entropy coding */
+  cinfo->optimize_coding = FALSE;
+  /* The standard Huffman tables are only valid for 8-bit data precision.
+   * If the precision is higher, force optimization on so that usable
+   * tables will be computed.  This test can be removed if default tables
+   * are supplied that are valid for the desired precision.
+   */
+  if (cinfo->data_precision > 8)
+    cinfo->optimize_coding = TRUE;
+
+  /* By default, use the simpler non-cosited sampling alignment */
+  cinfo->CCIR601_sampling = FALSE;
+
+#if JPEG_LIB_VERSION >= 70
+  /* By default, apply fancy downsampling */
+  cinfo->do_fancy_downsampling = TRUE;
+#endif
+
+  /* No input smoothing */
+  cinfo->smoothing_factor = 0;
+
+  /* DCT algorithm preference */
+  cinfo->dct_method = JDCT_DEFAULT;
+
+  /* No restart markers */
+  cinfo->restart_interval = 0;
+  cinfo->restart_in_rows = 0;
+
+  /* Fill in default JFIF marker parameters.  Note that whether the marker
+   * will actually be written is determined by jpeg_set_colorspace.
+   *
+   * By default, the library emits JFIF version code 1.01.
+   * An application that wants to emit JFIF 1.02 extension markers should set
+   * JFIF_minor_version to 2.  We could probably get away with just defaulting
+   * to 1.02, but there may still be some decoders in use that will complain
+   * about that; saying 1.01 should minimize compatibility problems.
+   */
+  cinfo->JFIF_major_version = 1; /* Default JFIF version = 1.01 */
+  cinfo->JFIF_minor_version = 1;
+  cinfo->density_unit = 0;      /* Pixel size is unknown by default */
+  cinfo->X_density = 1;         /* Pixel aspect ratio is square by default */
+  cinfo->Y_density = 1;
+
+  /* Choose JPEG colorspace based on input space, set defaults accordingly */
+
+  jpeg_default_colorspace(cinfo);
+}
+
+
+/*
+ * Select an appropriate JPEG colorspace for in_color_space.
+ */
+
+GLOBAL(void)
+jpeg_default_colorspace(j_compress_ptr cinfo)
+{
+  switch (cinfo->in_color_space) {
+  case JCS_GRAYSCALE:
+    jpeg_set_colorspace(cinfo, JCS_GRAYSCALE);
+    break;
+  case JCS_RGB:
+  case JCS_EXT_RGB:
+  case JCS_EXT_RGBX:
+  case JCS_EXT_BGR:
+  case JCS_EXT_BGRX:
+  case JCS_EXT_XBGR:
+  case JCS_EXT_XRGB:
+  case JCS_EXT_RGBA:
+  case JCS_EXT_BGRA:
+  case JCS_EXT_ABGR:
+  case JCS_EXT_ARGB:
+    jpeg_set_colorspace(cinfo, JCS_YCbCr);
+    break;
+  case JCS_YCbCr:
+    jpeg_set_colorspace(cinfo, JCS_YCbCr);
+    break;
+  case JCS_CMYK:
+    jpeg_set_colorspace(cinfo, JCS_CMYK); /* By default, no translation */
+    break;
+  case JCS_YCCK:
+    jpeg_set_colorspace(cinfo, JCS_YCCK);
+    break;
+  case JCS_UNKNOWN:
+    jpeg_set_colorspace(cinfo, JCS_UNKNOWN);
+    break;
+  default:
+    ERREXIT(cinfo, JERR_BAD_IN_COLORSPACE);
+  }
+}
+
+
+/*
+ * Set the JPEG colorspace, and choose colorspace-dependent default values.
+ */
+
+GLOBAL(void)
+jpeg_set_colorspace(j_compress_ptr cinfo, J_COLOR_SPACE colorspace)
+{
+  jpeg_component_info *compptr;
+  int ci;
+
+#define SET_COMP(index, id, hsamp, vsamp, quant, dctbl, actbl) \
+  (compptr = &cinfo->comp_info[index], \
+   compptr->component_id = (id), \
+   compptr->h_samp_factor = (hsamp), \
+   compptr->v_samp_factor = (vsamp), \
+   compptr->quant_tbl_no = (quant), \
+   compptr->dc_tbl_no = (dctbl), \
+   compptr->ac_tbl_no = (actbl) )
+
+  /* Safety check to ensure start_compress not called yet. */
+  if (cinfo->global_state != CSTATE_START)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  /* For all colorspaces, we use Q and Huff tables 0 for luminance components,
+   * tables 1 for chrominance components.
+   */
+
+  cinfo->jpeg_color_space = colorspace;
+
+  cinfo->write_JFIF_header = FALSE; /* No marker for non-JFIF colorspaces */
+  cinfo->write_Adobe_marker = FALSE; /* write no Adobe marker by default */
+
+  switch (colorspace) {
+  case JCS_GRAYSCALE:
+    cinfo->write_JFIF_header = TRUE; /* Write a JFIF marker */
+    cinfo->num_components = 1;
+    /* JFIF specifies component ID 1 */
+    SET_COMP(0, 1, 1, 1, 0, 0, 0);
+    break;
+  case JCS_RGB:
+    cinfo->write_Adobe_marker = TRUE; /* write Adobe marker to flag RGB */
+    cinfo->num_components = 3;
+    SET_COMP(0, 0x52 /* 'R' */, 1, 1, 0, 0, 0);
+    SET_COMP(1, 0x47 /* 'G' */, 1, 1, 0, 0, 0);
+    SET_COMP(2, 0x42 /* 'B' */, 1, 1, 0, 0, 0);
+    break;
+  case JCS_YCbCr:
+    cinfo->write_JFIF_header = TRUE; /* Write a JFIF marker */
+    cinfo->num_components = 3;
+    /* JFIF specifies component IDs 1,2,3 */
+    /* We default to 2x2 subsamples of chrominance */
+    SET_COMP(0, 1, 2, 2, 0, 0, 0);
+    SET_COMP(1, 2, 1, 1, 1, 1, 1);
+    SET_COMP(2, 3, 1, 1, 1, 1, 1);
+    break;
+  case JCS_CMYK:
+    cinfo->write_Adobe_marker = TRUE; /* write Adobe marker to flag CMYK */
+    cinfo->num_components = 4;
+    SET_COMP(0, 0x43 /* 'C' */, 1, 1, 0, 0, 0);
+    SET_COMP(1, 0x4D /* 'M' */, 1, 1, 0, 0, 0);
+    SET_COMP(2, 0x59 /* 'Y' */, 1, 1, 0, 0, 0);
+    SET_COMP(3, 0x4B /* 'K' */, 1, 1, 0, 0, 0);
+    break;
+  case JCS_YCCK:
+    cinfo->write_Adobe_marker = TRUE; /* write Adobe marker to flag YCCK */
+    cinfo->num_components = 4;
+    SET_COMP(0, 1, 2, 2, 0, 0, 0);
+    SET_COMP(1, 2, 1, 1, 1, 1, 1);
+    SET_COMP(2, 3, 1, 1, 1, 1, 1);
+    SET_COMP(3, 4, 2, 2, 0, 0, 0);
+    break;
+  case JCS_UNKNOWN:
+    cinfo->num_components = cinfo->input_components;
+    if (cinfo->num_components < 1 || cinfo->num_components > MAX_COMPONENTS)
+      ERREXIT2(cinfo, JERR_COMPONENT_COUNT, cinfo->num_components,
+               MAX_COMPONENTS);
+    for (ci = 0; ci < cinfo->num_components; ci++) {
+      SET_COMP(ci, ci, 1, 1, 0, 0, 0);
+    }
+    break;
+  default:
+    ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+  }
+}
+
+
+#ifdef C_PROGRESSIVE_SUPPORTED
+
+LOCAL(jpeg_scan_info *)
+fill_a_scan(jpeg_scan_info *scanptr, int ci, int Ss, int Se, int Ah, int Al)
+/* Support routine: generate one scan for specified component */
+{
+  scanptr->comps_in_scan = 1;
+  scanptr->component_index[0] = ci;
+  scanptr->Ss = Ss;
+  scanptr->Se = Se;
+  scanptr->Ah = Ah;
+  scanptr->Al = Al;
+  scanptr++;
+  return scanptr;
+}
+
+LOCAL(jpeg_scan_info *)
+fill_scans(jpeg_scan_info *scanptr, int ncomps, int Ss, int Se, int Ah, int Al)
+/* Support routine: generate one scan for each component */
+{
+  int ci;
+
+  for (ci = 0; ci < ncomps; ci++) {
+    scanptr->comps_in_scan = 1;
+    scanptr->component_index[0] = ci;
+    scanptr->Ss = Ss;
+    scanptr->Se = Se;
+    scanptr->Ah = Ah;
+    scanptr->Al = Al;
+    scanptr++;
+  }
+  return scanptr;
+}
+
+LOCAL(jpeg_scan_info *)
+fill_dc_scans(jpeg_scan_info *scanptr, int ncomps, int Ah, int Al)
+/* Support routine: generate interleaved DC scan if possible, else N scans */
+{
+  int ci;
+
+  if (ncomps <= MAX_COMPS_IN_SCAN) {
+    /* Single interleaved DC scan */
+    scanptr->comps_in_scan = ncomps;
+    for (ci = 0; ci < ncomps; ci++)
+      scanptr->component_index[ci] = ci;
+    scanptr->Ss = scanptr->Se = 0;
+    scanptr->Ah = Ah;
+    scanptr->Al = Al;
+    scanptr++;
+  } else {
+    /* Noninterleaved DC scan for each component */
+    scanptr = fill_scans(scanptr, ncomps, 0, 0, Ah, Al);
+  }
+  return scanptr;
+}
+
+
+/*
+ * Create a recommended progressive-JPEG script.
+ * cinfo->num_components and cinfo->jpeg_color_space must be correct.
+ */
+
+GLOBAL(void)
+jpeg_simple_progression(j_compress_ptr cinfo)
+{
+  int ncomps = cinfo->num_components;
+  int nscans;
+  jpeg_scan_info *scanptr;
+
+  /* Safety check to ensure start_compress not called yet. */
+  if (cinfo->global_state != CSTATE_START)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  /* Figure space needed for script.  Calculation must match code below! */
+  if (ncomps == 3 && cinfo->jpeg_color_space == JCS_YCbCr) {
+    /* Custom script for YCbCr color images. */
+    nscans = 10;
+  } else {
+    /* All-purpose script for other color spaces. */
+    if (ncomps > MAX_COMPS_IN_SCAN)
+      nscans = 6 * ncomps;      /* 2 DC + 4 AC scans per component */
+    else
+      nscans = 2 + 4 * ncomps;  /* 2 DC scans; 4 AC scans per component */
+  }
+
+  /* Allocate space for script.
+   * We need to put it in the permanent pool in case the application performs
+   * multiple compressions without changing the settings.  To avoid a memory
+   * leak if jpeg_simple_progression is called repeatedly for the same JPEG
+   * object, we try to re-use previously allocated space, and we allocate
+   * enough space to handle YCbCr even if initially asked for grayscale.
+   */
+  if (cinfo->script_space == NULL || cinfo->script_space_size < nscans) {
+    cinfo->script_space_size = MAX(nscans, 10);
+    cinfo->script_space = (jpeg_scan_info *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                        cinfo->script_space_size * sizeof(jpeg_scan_info));
+  }
+  scanptr = cinfo->script_space;
+  cinfo->scan_info = scanptr;
+  cinfo->num_scans = nscans;
+
+  if (ncomps == 3 && cinfo->jpeg_color_space == JCS_YCbCr) {
+    /* Custom script for YCbCr color images. */
+    /* Initial DC scan */
+    scanptr = fill_dc_scans(scanptr, ncomps, 0, 1);
+    /* Initial AC scan: get some luma data out in a hurry */
+    scanptr = fill_a_scan(scanptr, 0, 1, 5, 0, 2);
+    /* Chroma data is too small to be worth expending many scans on */
+    scanptr = fill_a_scan(scanptr, 2, 1, 63, 0, 1);
+    scanptr = fill_a_scan(scanptr, 1, 1, 63, 0, 1);
+    /* Complete spectral selection for luma AC */
+    scanptr = fill_a_scan(scanptr, 0, 6, 63, 0, 2);
+    /* Refine next bit of luma AC */
+    scanptr = fill_a_scan(scanptr, 0, 1, 63, 2, 1);
+    /* Finish DC successive approximation */
+    scanptr = fill_dc_scans(scanptr, ncomps, 1, 0);
+    /* Finish AC successive approximation */
+    scanptr = fill_a_scan(scanptr, 2, 1, 63, 1, 0);
+    scanptr = fill_a_scan(scanptr, 1, 1, 63, 1, 0);
+    /* Luma bottom bit comes last since it's usually largest scan */
+    scanptr = fill_a_scan(scanptr, 0, 1, 63, 1, 0);
+  } else {
+    /* All-purpose script for other color spaces. */
+    /* Successive approximation first pass */
+    scanptr = fill_dc_scans(scanptr, ncomps, 0, 1);
+    scanptr = fill_scans(scanptr, ncomps, 1, 5, 0, 2);
+    scanptr = fill_scans(scanptr, ncomps, 6, 63, 0, 2);
+    /* Successive approximation second pass */
+    scanptr = fill_scans(scanptr, ncomps, 1, 63, 2, 1);
+    /* Successive approximation final pass */
+    scanptr = fill_dc_scans(scanptr, ncomps, 1, 0);
+    scanptr = fill_scans(scanptr, ncomps, 1, 63, 1, 0);
+  }
+}
+
+#endif /* C_PROGRESSIVE_SUPPORTED */

3rdparty/libjpeg-turbo/src/jcprepct.c

@@ -0,0 +1,351 @@
+/*
+ * jcprepct.c
+ *
+ * This file is part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * It was modified by The libjpeg-turbo Project to include only code relevant
+ * to libjpeg-turbo.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains the compression preprocessing controller.
+ * This controller manages the color conversion, downsampling,
+ * and edge expansion steps.
+ *
+ * Most of the complexity here is associated with buffering input rows
+ * as required by the downsampler.  See the comments at the head of
+ * jcsample.c for the downsampler's needs.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/* At present, jcsample.c can request context rows only for smoothing.
+ * In the future, we might also need context rows for CCIR601 sampling
+ * or other more-complex downsampling procedures.  The code to support
+ * context rows should be compiled only if needed.
+ */
+#ifdef INPUT_SMOOTHING_SUPPORTED
+#define CONTEXT_ROWS_SUPPORTED
+#endif
+
+
+/*
+ * For the simple (no-context-row) case, we just need to buffer one
+ * row group's worth of pixels for the downsampling step.  At the bottom of
+ * the image, we pad to a full row group by replicating the last pixel row.
+ * The downsampler's last output row is then replicated if needed to pad
+ * out to a full iMCU row.
+ *
+ * When providing context rows, we must buffer three row groups' worth of
+ * pixels.  Three row groups are physically allocated, but the row pointer
+ * arrays are made five row groups high, with the extra pointers above and
+ * below "wrapping around" to point to the last and first real row groups.
+ * This allows the downsampler to access the proper context rows.
+ * At the top and bottom of the image, we create dummy context rows by
+ * copying the first or last real pixel row.  This copying could be avoided
+ * by pointer hacking as is done in jdmainct.c, but it doesn't seem worth the
+ * trouble on the compression side.
+ */
+
+
+/* Private buffer controller object */
+
+typedef struct {
+  struct jpeg_c_prep_controller pub; /* public fields */
+
+  /* Downsampling input buffer.  This buffer holds color-converted data
+   * until we have enough to do a downsample step.
+   */
+  JSAMPARRAY color_buf[MAX_COMPONENTS];
+
+  JDIMENSION rows_to_go;        /* counts rows remaining in source image */
+  int next_buf_row;             /* index of next row to store in color_buf */
+
+#ifdef CONTEXT_ROWS_SUPPORTED   /* only needed for context case */
+  int this_row_group;           /* starting row index of group to process */
+  int next_buf_stop;            /* downsample when we reach this index */
+#endif
+} my_prep_controller;
+
+typedef my_prep_controller *my_prep_ptr;
+
+
+/*
+ * Initialize for a processing pass.
+ */
+
+METHODDEF(void)
+start_pass_prep(j_compress_ptr cinfo, J_BUF_MODE pass_mode)
+{
+  my_prep_ptr prep = (my_prep_ptr)cinfo->prep;
+
+  if (pass_mode != JBUF_PASS_THRU)
+    ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+
+  /* Initialize total-height counter for detecting bottom of image */
+  prep->rows_to_go = cinfo->image_height;
+  /* Mark the conversion buffer empty */
+  prep->next_buf_row = 0;
+#ifdef CONTEXT_ROWS_SUPPORTED
+  /* Preset additional state variables for context mode.
+   * These aren't used in non-context mode, so we needn't test which mode.
+   */
+  prep->this_row_group = 0;
+  /* Set next_buf_stop to stop after two row groups have been read in. */
+  prep->next_buf_stop = 2 * cinfo->max_v_samp_factor;
+#endif
+}
+
+
+/*
+ * Expand an image vertically from height input_rows to height output_rows,
+ * by duplicating the bottom row.
+ */
+
+LOCAL(void)
+expand_bottom_edge(JSAMPARRAY image_data, JDIMENSION num_cols, int input_rows,
+                   int output_rows)
+{
+  register int row;
+
+  for (row = input_rows; row < output_rows; row++) {
+    jcopy_sample_rows(image_data, input_rows - 1, image_data, row, 1,
+                      num_cols);
+  }
+}
+
+
+/*
+ * Process some data in the simple no-context case.
+ *
+ * Preprocessor output data is counted in "row groups".  A row group
+ * is defined to be v_samp_factor sample rows of each component.
+ * Downsampling will produce this much data from each max_v_samp_factor
+ * input rows.
+ */
+
+METHODDEF(void)
+pre_process_data(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                 JDIMENSION *in_row_ctr, JDIMENSION in_rows_avail,
+                 JSAMPIMAGE output_buf, JDIMENSION *out_row_group_ctr,
+                 JDIMENSION out_row_groups_avail)
+{
+  my_prep_ptr prep = (my_prep_ptr)cinfo->prep;
+  int numrows, ci;
+  JDIMENSION inrows;
+  jpeg_component_info *compptr;
+
+  while (*in_row_ctr < in_rows_avail &&
+         *out_row_group_ctr < out_row_groups_avail) {
+    /* Do color conversion to fill the conversion buffer. */
+    inrows = in_rows_avail - *in_row_ctr;
+    numrows = cinfo->max_v_samp_factor - prep->next_buf_row;
+    numrows = (int)MIN((JDIMENSION)numrows, inrows);
+    (*cinfo->cconvert->color_convert) (cinfo, input_buf + *in_row_ctr,
+                                       prep->color_buf,
+                                       (JDIMENSION)prep->next_buf_row,
+                                       numrows);
+    *in_row_ctr += numrows;
+    prep->next_buf_row += numrows;
+    prep->rows_to_go -= numrows;
+    /* If at bottom of image, pad to fill the conversion buffer. */
+    if (prep->rows_to_go == 0 &&
+        prep->next_buf_row < cinfo->max_v_samp_factor) {
+      for (ci = 0; ci < cinfo->num_components; ci++) {
+        expand_bottom_edge(prep->color_buf[ci], cinfo->image_width,
+                           prep->next_buf_row, cinfo->max_v_samp_factor);
+      }
+      prep->next_buf_row = cinfo->max_v_samp_factor;
+    }
+    /* If we've filled the conversion buffer, empty it. */
+    if (prep->next_buf_row == cinfo->max_v_samp_factor) {
+      (*cinfo->downsample->downsample) (cinfo,
+                                        prep->color_buf, (JDIMENSION)0,
+                                        output_buf, *out_row_group_ctr);
+      prep->next_buf_row = 0;
+      (*out_row_group_ctr)++;
+    }
+    /* If at bottom of image, pad the output to a full iMCU height.
+     * Note we assume the caller is providing a one-iMCU-height output buffer!
+     */
+    if (prep->rows_to_go == 0 && *out_row_group_ctr < out_row_groups_avail) {
+      for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+           ci++, compptr++) {
+        expand_bottom_edge(output_buf[ci], compptr->width_in_blocks * DCTSIZE,
+                           (int)(*out_row_group_ctr * compptr->v_samp_factor),
+                           (int)(out_row_groups_avail * compptr->v_samp_factor));
+      }
+      *out_row_group_ctr = out_row_groups_avail;
+      break;                    /* can exit outer loop without test */
+    }
+  }
+}
+
+
+#ifdef CONTEXT_ROWS_SUPPORTED
+
+/*
+ * Process some data in the context case.
+ */
+
+METHODDEF(void)
+pre_process_context(j_compress_ptr cinfo, JSAMPARRAY input_buf,
+                    JDIMENSION *in_row_ctr, JDIMENSION in_rows_avail,
+                    JSAMPIMAGE output_buf, JDIMENSION *out_row_group_ctr,
+                    JDIMENSION out_row_groups_avail)
+{
+  my_prep_ptr prep = (my_prep_ptr)cinfo->prep;
+  int numrows, ci;
+  int buf_height = cinfo->max_v_samp_factor * 3;
+  JDIMENSION inrows;
+
+  while (*out_row_group_ctr < out_row_groups_avail) {
+    if (*in_row_ctr < in_rows_avail) {
+      /* Do color conversion to fill the conversion buffer. */
+      inrows = in_rows_avail - *in_row_ctr;
+      numrows = prep->next_buf_stop - prep->next_buf_row;
+      numrows = (int)MIN((JDIMENSION)numrows, inrows);
+      (*cinfo->cconvert->color_convert) (cinfo, input_buf + *in_row_ctr,
+                                         prep->color_buf,
+                                         (JDIMENSION)prep->next_buf_row,
+                                         numrows);
+      /* Pad at top of image, if first time through */
+      if (prep->rows_to_go == cinfo->image_height) {
+        for (ci = 0; ci < cinfo->num_components; ci++) {
+          int row;
+          for (row = 1; row <= cinfo->max_v_samp_factor; row++) {
+            jcopy_sample_rows(prep->color_buf[ci], 0, prep->color_buf[ci],
+                              -row, 1, cinfo->image_width);
+          }
+        }
+      }
+      *in_row_ctr += numrows;
+      prep->next_buf_row += numrows;
+      prep->rows_to_go -= numrows;
+    } else {
+      /* Return for more data, unless we are at the bottom of the image. */
+      if (prep->rows_to_go != 0)
+        break;
+      /* When at bottom of image, pad to fill the conversion buffer. */
+      if (prep->next_buf_row < prep->next_buf_stop) {
+        for (ci = 0; ci < cinfo->num_components; ci++) {
+          expand_bottom_edge(prep->color_buf[ci], cinfo->image_width,
+                             prep->next_buf_row, prep->next_buf_stop);
+        }
+        prep->next_buf_row = prep->next_buf_stop;
+      }
+    }
+    /* If we've gotten enough data, downsample a row group. */
+    if (prep->next_buf_row == prep->next_buf_stop) {
+      (*cinfo->downsample->downsample) (cinfo, prep->color_buf,
+                                        (JDIMENSION)prep->this_row_group,
+                                        output_buf, *out_row_group_ctr);
+      (*out_row_group_ctr)++;
+      /* Advance pointers with wraparound as necessary. */
+      prep->this_row_group += cinfo->max_v_samp_factor;
+      if (prep->this_row_group >= buf_height)
+        prep->this_row_group = 0;
+      if (prep->next_buf_row >= buf_height)
+        prep->next_buf_row = 0;
+      prep->next_buf_stop = prep->next_buf_row + cinfo->max_v_samp_factor;
+    }
+  }
+}
+
+
+/*
+ * Create the wrapped-around downsampling input buffer needed for context mode.
+ */
+
+LOCAL(void)
+create_context_buffer(j_compress_ptr cinfo)
+{
+  my_prep_ptr prep = (my_prep_ptr)cinfo->prep;
+  int rgroup_height = cinfo->max_v_samp_factor;
+  int ci, i;
+  jpeg_component_info *compptr;
+  JSAMPARRAY true_buffer, fake_buffer;
+
+  /* Grab enough space for fake row pointers for all the components;
+   * we need five row groups' worth of pointers for each component.
+   */
+  fake_buffer = (JSAMPARRAY)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                (cinfo->num_components * 5 * rgroup_height) *
+                                sizeof(JSAMPROW));
+
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    /* Allocate the actual buffer space (3 row groups) for this component.
+     * We make the buffer wide enough to allow the downsampler to edge-expand
+     * horizontally within the buffer, if it so chooses.
+     */
+    true_buffer = (*cinfo->mem->alloc_sarray)
+      ((j_common_ptr)cinfo, JPOOL_IMAGE,
+       (JDIMENSION)(((long)compptr->width_in_blocks * DCTSIZE *
+                     cinfo->max_h_samp_factor) / compptr->h_samp_factor),
+       (JDIMENSION)(3 * rgroup_height));
+    /* Copy true buffer row pointers into the middle of the fake row array */
+    MEMCOPY(fake_buffer + rgroup_height, true_buffer,
+            3 * rgroup_height * sizeof(JSAMPROW));
+    /* Fill in the above and below wraparound pointers */
+    for (i = 0; i < rgroup_height; i++) {
+      fake_buffer[i] = true_buffer[2 * rgroup_height + i];
+      fake_buffer[4 * rgroup_height + i] = true_buffer[i];
+    }
+    prep->color_buf[ci] = fake_buffer + rgroup_height;
+    fake_buffer += 5 * rgroup_height; /* point to space for next component */
+  }
+}
+
+#endif /* CONTEXT_ROWS_SUPPORTED */
+
+
+/*
+ * Initialize preprocessing controller.
+ */
+
+GLOBAL(void)
+jinit_c_prep_controller(j_compress_ptr cinfo, boolean need_full_buffer)
+{
+  my_prep_ptr prep;
+  int ci;
+  jpeg_component_info *compptr;
+
+  if (need_full_buffer)         /* safety check */
+    ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+
+  prep = (my_prep_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_prep_controller));
+  cinfo->prep = (struct jpeg_c_prep_controller *)prep;
+  prep->pub.start_pass = start_pass_prep;
+
+  /* Allocate the color conversion buffer.
+   * We make the buffer wide enough to allow the downsampler to edge-expand
+   * horizontally within the buffer, if it so chooses.
+   */
+  if (cinfo->downsample->need_context_rows) {
+    /* Set up to provide context rows */
+#ifdef CONTEXT_ROWS_SUPPORTED
+    prep->pub.pre_process_data = pre_process_context;
+    create_context_buffer(cinfo);
+#else
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif
+  } else {
+    /* No context, just make it tall enough for one row group */
+    prep->pub.pre_process_data = pre_process_data;
+    for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+         ci++, compptr++) {
+      prep->color_buf[ci] = (*cinfo->mem->alloc_sarray)
+        ((j_common_ptr)cinfo, JPOOL_IMAGE,
+         (JDIMENSION)(((long)compptr->width_in_blocks * DCTSIZE *
+                       cinfo->max_h_samp_factor) / compptr->h_samp_factor),
+         (JDIMENSION)cinfo->max_v_samp_factor);
+    }
+  }
+}

3rdparty/libjpeg-turbo/src/jcsample.c

@@ -0,0 +1,539 @@
+/*
+ * jcsample.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1996, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * Copyright (C) 2014, MIPS Technologies, Inc., California.
+ * Copyright (C) 2015, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains downsampling routines.
+ *
+ * Downsampling input data is counted in "row groups".  A row group
+ * is defined to be max_v_samp_factor pixel rows of each component,
+ * from which the downsampler produces v_samp_factor sample rows.
+ * A single row group is processed in each call to the downsampler module.
+ *
+ * The downsampler is responsible for edge-expansion of its output data
+ * to fill an integral number of DCT blocks horizontally.  The source buffer
+ * may be modified if it is helpful for this purpose (the source buffer is
+ * allocated wide enough to correspond to the desired output width).
+ * The caller (the prep controller) is responsible for vertical padding.
+ *
+ * The downsampler may request "context rows" by setting need_context_rows
+ * during startup.  In this case, the input arrays will contain at least
+ * one row group's worth of pixels above and below the passed-in data;
+ * the caller will create dummy rows at image top and bottom by replicating
+ * the first or last real pixel row.
+ *
+ * An excellent reference for image resampling is
+ *   Digital Image Warping, George Wolberg, 1990.
+ *   Pub. by IEEE Computer Society Press, Los Alamitos, CA. ISBN 0-8186-8944-7.
+ *
+ * The downsampling algorithm used here is a simple average of the source
+ * pixels covered by the output pixel.  The hi-falutin sampling literature
+ * refers to this as a "box filter".  In general the characteristics of a box
+ * filter are not very good, but for the specific cases we normally use (1:1
+ * and 2:1 ratios) the box is equivalent to a "triangle filter" which is not
+ * nearly so bad.  If you intend to use other sampling ratios, you'd be well
+ * advised to improve this code.
+ *
+ * A simple input-smoothing capability is provided.  This is mainly intended
+ * for cleaning up color-dithered GIF input files (if you find it inadequate,
+ * we suggest using an external filtering program such as pnmconvol).  When
+ * enabled, each input pixel P is replaced by a weighted sum of itself and its
+ * eight neighbors.  P's weight is 1-8*SF and each neighbor's weight is SF,
+ * where SF = (smoothing_factor / 1024).
+ * Currently, smoothing is only supported for 2h2v sampling factors.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jsimd.h"
+
+
+/* Pointer to routine to downsample a single component */
+typedef void (*downsample1_ptr) (j_compress_ptr cinfo,
+                                 jpeg_component_info *compptr,
+                                 JSAMPARRAY input_data,
+                                 JSAMPARRAY output_data);
+
+/* Private subobject */
+
+typedef struct {
+  struct jpeg_downsampler pub;  /* public fields */
+
+  /* Downsampling method pointers, one per component */
+  downsample1_ptr methods[MAX_COMPONENTS];
+} my_downsampler;
+
+typedef my_downsampler *my_downsample_ptr;
+
+
+/*
+ * Initialize for a downsampling pass.
+ */
+
+METHODDEF(void)
+start_pass_downsample(j_compress_ptr cinfo)
+{
+  /* no work for now */
+}
+
+
+/*
+ * Expand a component horizontally from width input_cols to width output_cols,
+ * by duplicating the rightmost samples.
+ */
+
+LOCAL(void)
+expand_right_edge(JSAMPARRAY image_data, int num_rows, JDIMENSION input_cols,
+                  JDIMENSION output_cols)
+{
+  register JSAMPROW ptr;
+  register JSAMPLE pixval;
+  register int count;
+  int row;
+  int numcols = (int)(output_cols - input_cols);
+
+  if (numcols > 0) {
+    for (row = 0; row < num_rows; row++) {
+      ptr = image_data[row] + input_cols;
+      pixval = ptr[-1];         /* don't need GETJSAMPLE() here */
+      for (count = numcols; count > 0; count--)
+        *ptr++ = pixval;
+    }
+  }
+}
+
+
+/*
+ * Do downsampling for a whole row group (all components).
+ *
+ * In this version we simply downsample each component independently.
+ */
+
+METHODDEF(void)
+sep_downsample(j_compress_ptr cinfo, JSAMPIMAGE input_buf,
+               JDIMENSION in_row_index, JSAMPIMAGE output_buf,
+               JDIMENSION out_row_group_index)
+{
+  my_downsample_ptr downsample = (my_downsample_ptr)cinfo->downsample;
+  int ci;
+  jpeg_component_info *compptr;
+  JSAMPARRAY in_ptr, out_ptr;
+
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    in_ptr = input_buf[ci] + in_row_index;
+    out_ptr = output_buf[ci] + (out_row_group_index * compptr->v_samp_factor);
+    (*downsample->methods[ci]) (cinfo, compptr, in_ptr, out_ptr);
+  }
+}
+
+
+/*
+ * Downsample pixel values of a single component.
+ * One row group is processed per call.
+ * This version handles arbitrary integral sampling ratios, without smoothing.
+ * Note that this version is not actually used for customary sampling ratios.
+ */
+
+METHODDEF(void)
+int_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
+               JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+  int inrow, outrow, h_expand, v_expand, numpix, numpix2, h, v;
+  JDIMENSION outcol, outcol_h;  /* outcol_h == outcol*h_expand */
+  JDIMENSION output_cols = compptr->width_in_blocks * DCTSIZE;
+  JSAMPROW inptr, outptr;
+  JLONG outvalue;
+
+  h_expand = cinfo->max_h_samp_factor / compptr->h_samp_factor;
+  v_expand = cinfo->max_v_samp_factor / compptr->v_samp_factor;
+  numpix = h_expand * v_expand;
+  numpix2 = numpix / 2;
+
+  /* Expand input data enough to let all the output samples be generated
+   * by the standard loop.  Special-casing padded output would be more
+   * efficient.
+   */
+  expand_right_edge(input_data, cinfo->max_v_samp_factor, cinfo->image_width,
+                    output_cols * h_expand);
+
+  inrow = 0;
+  for (outrow = 0; outrow < compptr->v_samp_factor; outrow++) {
+    outptr = output_data[outrow];
+    for (outcol = 0, outcol_h = 0; outcol < output_cols;
+         outcol++, outcol_h += h_expand) {
+      outvalue = 0;
+      for (v = 0; v < v_expand; v++) {
+        inptr = input_data[inrow + v] + outcol_h;
+        for (h = 0; h < h_expand; h++) {
+          outvalue += (JLONG)GETJSAMPLE(*inptr++);
+        }
+      }
+      *outptr++ = (JSAMPLE)((outvalue + numpix2) / numpix);
+    }
+    inrow += v_expand;
+  }
+}
+
+
+/*
+ * Downsample pixel values of a single component.
+ * This version handles the special case of a full-size component,
+ * without smoothing.
+ */
+
+METHODDEF(void)
+fullsize_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
+                    JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+  /* Copy the data */
+  jcopy_sample_rows(input_data, 0, output_data, 0, cinfo->max_v_samp_factor,
+                    cinfo->image_width);
+  /* Edge-expand */
+  expand_right_edge(output_data, cinfo->max_v_samp_factor, cinfo->image_width,
+                    compptr->width_in_blocks * DCTSIZE);
+}
+
+
+/*
+ * Downsample pixel values of a single component.
+ * This version handles the common case of 2:1 horizontal and 1:1 vertical,
+ * without smoothing.
+ *
+ * A note about the "bias" calculations: when rounding fractional values to
+ * integer, we do not want to always round 0.5 up to the next integer.
+ * If we did that, we'd introduce a noticeable bias towards larger values.
+ * Instead, this code is arranged so that 0.5 will be rounded up or down at
+ * alternate pixel locations (a simple ordered dither pattern).
+ */
+
+METHODDEF(void)
+h2v1_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
+                JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+  int outrow;
+  JDIMENSION outcol;
+  JDIMENSION output_cols = compptr->width_in_blocks * DCTSIZE;
+  register JSAMPROW inptr, outptr;
+  register int bias;
+
+  /* Expand input data enough to let all the output samples be generated
+   * by the standard loop.  Special-casing padded output would be more
+   * efficient.
+   */
+  expand_right_edge(input_data, cinfo->max_v_samp_factor, cinfo->image_width,
+                    output_cols * 2);
+
+  for (outrow = 0; outrow < compptr->v_samp_factor; outrow++) {
+    outptr = output_data[outrow];
+    inptr = input_data[outrow];
+    bias = 0;                   /* bias = 0,1,0,1,... for successive samples */
+    for (outcol = 0; outcol < output_cols; outcol++) {
+      *outptr++ =
+        (JSAMPLE)((GETJSAMPLE(*inptr) + GETJSAMPLE(inptr[1]) + bias) >> 1);
+      bias ^= 1;                /* 0=>1, 1=>0 */
+      inptr += 2;
+    }
+  }
+}
+
+
+/*
+ * Downsample pixel values of a single component.
+ * This version handles the standard case of 2:1 horizontal and 2:1 vertical,
+ * without smoothing.
+ */
+
+METHODDEF(void)
+h2v2_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
+                JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+  int inrow, outrow;
+  JDIMENSION outcol;
+  JDIMENSION output_cols = compptr->width_in_blocks * DCTSIZE;
+  register JSAMPROW inptr0, inptr1, outptr;
+  register int bias;
+
+  /* Expand input data enough to let all the output samples be generated
+   * by the standard loop.  Special-casing padded output would be more
+   * efficient.
+   */
+  expand_right_edge(input_data, cinfo->max_v_samp_factor, cinfo->image_width,
+                    output_cols * 2);
+
+  inrow = 0;
+  for (outrow = 0; outrow < compptr->v_samp_factor; outrow++) {
+    outptr = output_data[outrow];
+    inptr0 = input_data[inrow];
+    inptr1 = input_data[inrow + 1];
+    bias = 1;                   /* bias = 1,2,1,2,... for successive samples */
+    for (outcol = 0; outcol < output_cols; outcol++) {
+      *outptr++ =
+        (JSAMPLE)((GETJSAMPLE(*inptr0) + GETJSAMPLE(inptr0[1]) +
+                   GETJSAMPLE(*inptr1) + GETJSAMPLE(inptr1[1]) + bias) >> 2);
+      bias ^= 3;                /* 1=>2, 2=>1 */
+      inptr0 += 2;  inptr1 += 2;
+    }
+    inrow += 2;
+  }
+}
+
+
+#ifdef INPUT_SMOOTHING_SUPPORTED
+
+/*
+ * Downsample pixel values of a single component.
+ * This version handles the standard case of 2:1 horizontal and 2:1 vertical,
+ * with smoothing.  One row of context is required.
+ */
+
+METHODDEF(void)
+h2v2_smooth_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
+                       JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+  int inrow, outrow;
+  JDIMENSION colctr;
+  JDIMENSION output_cols = compptr->width_in_blocks * DCTSIZE;
+  register JSAMPROW inptr0, inptr1, above_ptr, below_ptr, outptr;
+  JLONG membersum, neighsum, memberscale, neighscale;
+
+  /* Expand input data enough to let all the output samples be generated
+   * by the standard loop.  Special-casing padded output would be more
+   * efficient.
+   */
+  expand_right_edge(input_data - 1, cinfo->max_v_samp_factor + 2,
+                    cinfo->image_width, output_cols * 2);
+
+  /* We don't bother to form the individual "smoothed" input pixel values;
+   * we can directly compute the output which is the average of the four
+   * smoothed values.  Each of the four member pixels contributes a fraction
+   * (1-8*SF) to its own smoothed image and a fraction SF to each of the three
+   * other smoothed pixels, therefore a total fraction (1-5*SF)/4 to the final
+   * output.  The four corner-adjacent neighbor pixels contribute a fraction
+   * SF to just one smoothed pixel, or SF/4 to the final output; while the
+   * eight edge-adjacent neighbors contribute SF to each of two smoothed
+   * pixels, or SF/2 overall.  In order to use integer arithmetic, these
+   * factors are scaled by 2^16 = 65536.
+   * Also recall that SF = smoothing_factor / 1024.
+   */
+
+  memberscale = 16384 - cinfo->smoothing_factor * 80; /* scaled (1-5*SF)/4 */
+  neighscale = cinfo->smoothing_factor * 16; /* scaled SF/4 */
+
+  inrow = 0;
+  for (outrow = 0; outrow < compptr->v_samp_factor; outrow++) {
+    outptr = output_data[outrow];
+    inptr0 = input_data[inrow];
+    inptr1 = input_data[inrow + 1];
+    above_ptr = input_data[inrow - 1];
+    below_ptr = input_data[inrow + 2];
+
+    /* Special case for first column: pretend column -1 is same as column 0 */
+    membersum = GETJSAMPLE(*inptr0) + GETJSAMPLE(inptr0[1]) +
+                GETJSAMPLE(*inptr1) + GETJSAMPLE(inptr1[1]);
+    neighsum = GETJSAMPLE(*above_ptr) + GETJSAMPLE(above_ptr[1]) +
+               GETJSAMPLE(*below_ptr) + GETJSAMPLE(below_ptr[1]) +
+               GETJSAMPLE(*inptr0) + GETJSAMPLE(inptr0[2]) +
+               GETJSAMPLE(*inptr1) + GETJSAMPLE(inptr1[2]);
+    neighsum += neighsum;
+    neighsum += GETJSAMPLE(*above_ptr) + GETJSAMPLE(above_ptr[2]) +
+                GETJSAMPLE(*below_ptr) + GETJSAMPLE(below_ptr[2]);
+    membersum = membersum * memberscale + neighsum * neighscale;
+    *outptr++ = (JSAMPLE)((membersum + 32768) >> 16);
+    inptr0 += 2;  inptr1 += 2;  above_ptr += 2;  below_ptr += 2;
+
+    for (colctr = output_cols - 2; colctr > 0; colctr--) {
+      /* sum of pixels directly mapped to this output element */
+      membersum = GETJSAMPLE(*inptr0) + GETJSAMPLE(inptr0[1]) +
+                  GETJSAMPLE(*inptr1) + GETJSAMPLE(inptr1[1]);
+      /* sum of edge-neighbor pixels */
+      neighsum = GETJSAMPLE(*above_ptr) + GETJSAMPLE(above_ptr[1]) +
+                 GETJSAMPLE(*below_ptr) + GETJSAMPLE(below_ptr[1]) +
+                 GETJSAMPLE(inptr0[-1]) + GETJSAMPLE(inptr0[2]) +
+                 GETJSAMPLE(inptr1[-1]) + GETJSAMPLE(inptr1[2]);
+      /* The edge-neighbors count twice as much as corner-neighbors */
+      neighsum += neighsum;
+      /* Add in the corner-neighbors */
+      neighsum += GETJSAMPLE(above_ptr[-1]) + GETJSAMPLE(above_ptr[2]) +
+                  GETJSAMPLE(below_ptr[-1]) + GETJSAMPLE(below_ptr[2]);
+      /* form final output scaled up by 2^16 */
+      membersum = membersum * memberscale + neighsum * neighscale;
+      /* round, descale and output it */
+      *outptr++ = (JSAMPLE)((membersum + 32768) >> 16);
+      inptr0 += 2;  inptr1 += 2;  above_ptr += 2;  below_ptr += 2;
+    }
+
+    /* Special case for last column */
+    membersum = GETJSAMPLE(*inptr0) + GETJSAMPLE(inptr0[1]) +
+                GETJSAMPLE(*inptr1) + GETJSAMPLE(inptr1[1]);
+    neighsum = GETJSAMPLE(*above_ptr) + GETJSAMPLE(above_ptr[1]) +
+               GETJSAMPLE(*below_ptr) + GETJSAMPLE(below_ptr[1]) +
+               GETJSAMPLE(inptr0[-1]) + GETJSAMPLE(inptr0[1]) +
+               GETJSAMPLE(inptr1[-1]) + GETJSAMPLE(inptr1[1]);
+    neighsum += neighsum;
+    neighsum += GETJSAMPLE(above_ptr[-1]) + GETJSAMPLE(above_ptr[1]) +
+                GETJSAMPLE(below_ptr[-1]) + GETJSAMPLE(below_ptr[1]);
+    membersum = membersum * memberscale + neighsum * neighscale;
+    *outptr = (JSAMPLE)((membersum + 32768) >> 16);
+
+    inrow += 2;
+  }
+}
+
+
+/*
+ * Downsample pixel values of a single component.
+ * This version handles the special case of a full-size component,
+ * with smoothing.  One row of context is required.
+ */
+
+METHODDEF(void)
+fullsize_smooth_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
+                           JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+  int outrow;
+  JDIMENSION colctr;
+  JDIMENSION output_cols = compptr->width_in_blocks * DCTSIZE;
+  register JSAMPROW inptr, above_ptr, below_ptr, outptr;
+  JLONG membersum, neighsum, memberscale, neighscale;
+  int colsum, lastcolsum, nextcolsum;
+
+  /* Expand input data enough to let all the output samples be generated
+   * by the standard loop.  Special-casing padded output would be more
+   * efficient.
+   */
+  expand_right_edge(input_data - 1, cinfo->max_v_samp_factor + 2,
+                    cinfo->image_width, output_cols);
+
+  /* Each of the eight neighbor pixels contributes a fraction SF to the
+   * smoothed pixel, while the main pixel contributes (1-8*SF).  In order
+   * to use integer arithmetic, these factors are multiplied by 2^16 = 65536.
+   * Also recall that SF = smoothing_factor / 1024.
+   */
+
+  memberscale = 65536L - cinfo->smoothing_factor * 512L; /* scaled 1-8*SF */
+  neighscale = cinfo->smoothing_factor * 64; /* scaled SF */
+
+  for (outrow = 0; outrow < compptr->v_samp_factor; outrow++) {
+    outptr = output_data[outrow];
+    inptr = input_data[outrow];
+    above_ptr = input_data[outrow - 1];
+    below_ptr = input_data[outrow + 1];
+
+    /* Special case for first column */
+    colsum = GETJSAMPLE(*above_ptr++) + GETJSAMPLE(*below_ptr++) +
+             GETJSAMPLE(*inptr);
+    membersum = GETJSAMPLE(*inptr++);
+    nextcolsum = GETJSAMPLE(*above_ptr) + GETJSAMPLE(*below_ptr) +
+                 GETJSAMPLE(*inptr);
+    neighsum = colsum + (colsum - membersum) + nextcolsum;
+    membersum = membersum * memberscale + neighsum * neighscale;
+    *outptr++ = (JSAMPLE)((membersum + 32768) >> 16);
+    lastcolsum = colsum;  colsum = nextcolsum;
+
+    for (colctr = output_cols - 2; colctr > 0; colctr--) {
+      membersum = GETJSAMPLE(*inptr++);
+      above_ptr++;  below_ptr++;
+      nextcolsum = GETJSAMPLE(*above_ptr) + GETJSAMPLE(*below_ptr) +
+                   GETJSAMPLE(*inptr);
+      neighsum = lastcolsum + (colsum - membersum) + nextcolsum;
+      membersum = membersum * memberscale + neighsum * neighscale;
+      *outptr++ = (JSAMPLE)((membersum + 32768) >> 16);
+      lastcolsum = colsum;  colsum = nextcolsum;
+    }
+
+    /* Special case for last column */
+    membersum = GETJSAMPLE(*inptr);
+    neighsum = lastcolsum + (colsum - membersum) + colsum;
+    membersum = membersum * memberscale + neighsum * neighscale;
+    *outptr = (JSAMPLE)((membersum + 32768) >> 16);
+
+  }
+}
+
+#endif /* INPUT_SMOOTHING_SUPPORTED */
+
+
+/*
+ * Module initialization routine for downsampling.
+ * Note that we must select a routine for each component.
+ */
+
+GLOBAL(void)
+jinit_downsampler(j_compress_ptr cinfo)
+{
+  my_downsample_ptr downsample;
+  int ci;
+  jpeg_component_info *compptr;
+  boolean smoothok = TRUE;
+
+  downsample = (my_downsample_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_downsampler));
+  cinfo->downsample = (struct jpeg_downsampler *)downsample;
+  downsample->pub.start_pass = start_pass_downsample;
+  downsample->pub.downsample = sep_downsample;
+  downsample->pub.need_context_rows = FALSE;
+
+  if (cinfo->CCIR601_sampling)
+    ERREXIT(cinfo, JERR_CCIR601_NOTIMPL);
+
+  /* Verify we can handle the sampling factors, and set up method pointers */
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    if (compptr->h_samp_factor == cinfo->max_h_samp_factor &&
+        compptr->v_samp_factor == cinfo->max_v_samp_factor) {
+#ifdef INPUT_SMOOTHING_SUPPORTED
+      if (cinfo->smoothing_factor) {
+        downsample->methods[ci] = fullsize_smooth_downsample;
+        downsample->pub.need_context_rows = TRUE;
+      } else
+#endif
+        downsample->methods[ci] = fullsize_downsample;
+    } else if (compptr->h_samp_factor * 2 == cinfo->max_h_samp_factor &&
+               compptr->v_samp_factor == cinfo->max_v_samp_factor) {
+      smoothok = FALSE;
+      if (jsimd_can_h2v1_downsample())
+        downsample->methods[ci] = jsimd_h2v1_downsample;
+      else
+        downsample->methods[ci] = h2v1_downsample;
+    } else if (compptr->h_samp_factor * 2 == cinfo->max_h_samp_factor &&
+               compptr->v_samp_factor * 2 == cinfo->max_v_samp_factor) {
+#ifdef INPUT_SMOOTHING_SUPPORTED
+      if (cinfo->smoothing_factor) {
+#if defined(__mips__)
+        if (jsimd_can_h2v2_smooth_downsample())
+          downsample->methods[ci] = jsimd_h2v2_smooth_downsample;
+        else
+#endif
+          downsample->methods[ci] = h2v2_smooth_downsample;
+        downsample->pub.need_context_rows = TRUE;
+      } else
+#endif
+      {
+        if (jsimd_can_h2v2_downsample())
+          downsample->methods[ci] = jsimd_h2v2_downsample;
+        else
+          downsample->methods[ci] = h2v2_downsample;
+      }
+    } else if ((cinfo->max_h_samp_factor % compptr->h_samp_factor) == 0 &&
+               (cinfo->max_v_samp_factor % compptr->v_samp_factor) == 0) {
+      smoothok = FALSE;
+      downsample->methods[ci] = int_downsample;
+    } else
+      ERREXIT(cinfo, JERR_FRACT_SAMPLE_NOTIMPL);
+  }
+
+#ifdef INPUT_SMOOTHING_SUPPORTED
+  if (cinfo->smoothing_factor && !smoothok)
+    TRACEMS(cinfo, 0, JTRC_SMOOTH_NOTIMPL);
+#endif
+}

3rdparty/libjpeg-turbo/src/jctrans.c

@@ -0,0 +1,400 @@
+/*
+ * jctrans.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1995-1998, Thomas G. Lane.
+ * Modified 2000-2009 by Guido Vollbeding.
+ * It was modified by The libjpeg-turbo Project to include only code relevant
+ * to libjpeg-turbo.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains library routines for transcoding compression,
+ * that is, writing raw DCT coefficient arrays to an output JPEG file.
+ * The routines in jcapimin.c will also be needed by a transcoder.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+/* Forward declarations */
+LOCAL(void) transencode_master_selection(j_compress_ptr cinfo,
+                                         jvirt_barray_ptr *coef_arrays);
+LOCAL(void) transencode_coef_controller(j_compress_ptr cinfo,
+                                        jvirt_barray_ptr *coef_arrays);
+
+
+/*
+ * Compression initialization for writing raw-coefficient data.
+ * Before calling this, all parameters and a data destination must be set up.
+ * Call jpeg_finish_compress() to actually write the data.
+ *
+ * The number of passed virtual arrays must match cinfo->num_components.
+ * Note that the virtual arrays need not be filled or even realized at
+ * the time write_coefficients is called; indeed, if the virtual arrays
+ * were requested from this compression object's memory manager, they
+ * typically will be realized during this routine and filled afterwards.
+ */
+
+GLOBAL(void)
+jpeg_write_coefficients(j_compress_ptr cinfo, jvirt_barray_ptr *coef_arrays)
+{
+  if (cinfo->global_state != CSTATE_START)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  /* Mark all tables to be written */
+  jpeg_suppress_tables(cinfo, FALSE);
+  /* (Re)initialize error mgr and destination modules */
+  (*cinfo->err->reset_error_mgr) ((j_common_ptr)cinfo);
+  (*cinfo->dest->init_destination) (cinfo);
+  /* Perform master selection of active modules */
+  transencode_master_selection(cinfo, coef_arrays);
+  /* Wait for jpeg_finish_compress() call */
+  cinfo->next_scanline = 0;     /* so jpeg_write_marker works */
+  cinfo->global_state = CSTATE_WRCOEFS;
+}
+
+
+/*
+ * Initialize the compression object with default parameters,
+ * then copy from the source object all parameters needed for lossless
+ * transcoding.  Parameters that can be varied without loss (such as
+ * scan script and Huffman optimization) are left in their default states.
+ */
+
+GLOBAL(void)
+jpeg_copy_critical_parameters(j_decompress_ptr srcinfo, j_compress_ptr dstinfo)
+{
+  JQUANT_TBL **qtblptr;
+  jpeg_component_info *incomp, *outcomp;
+  JQUANT_TBL *c_quant, *slot_quant;
+  int tblno, ci, coefi;
+
+  /* Safety check to ensure start_compress not called yet. */
+  if (dstinfo->global_state != CSTATE_START)
+    ERREXIT1(dstinfo, JERR_BAD_STATE, dstinfo->global_state);
+  /* Copy fundamental image dimensions */
+  dstinfo->image_width = srcinfo->image_width;
+  dstinfo->image_height = srcinfo->image_height;
+  dstinfo->input_components = srcinfo->num_components;
+  dstinfo->in_color_space = srcinfo->jpeg_color_space;
+#if JPEG_LIB_VERSION >= 70
+  dstinfo->jpeg_width = srcinfo->output_width;
+  dstinfo->jpeg_height = srcinfo->output_height;
+  dstinfo->min_DCT_h_scaled_size = srcinfo->min_DCT_h_scaled_size;
+  dstinfo->min_DCT_v_scaled_size = srcinfo->min_DCT_v_scaled_size;
+#endif
+  /* Initialize all parameters to default values */
+  jpeg_set_defaults(dstinfo);
+  /* jpeg_set_defaults may choose wrong colorspace, eg YCbCr if input is RGB.
+   * Fix it to get the right header markers for the image colorspace.
+   */
+  jpeg_set_colorspace(dstinfo, srcinfo->jpeg_color_space);
+  dstinfo->data_precision = srcinfo->data_precision;
+  dstinfo->CCIR601_sampling = srcinfo->CCIR601_sampling;
+  /* Copy the source's quantization tables. */
+  for (tblno = 0; tblno < NUM_QUANT_TBLS; tblno++) {
+    if (srcinfo->quant_tbl_ptrs[tblno] != NULL) {
+      qtblptr = &dstinfo->quant_tbl_ptrs[tblno];
+      if (*qtblptr == NULL)
+        *qtblptr = jpeg_alloc_quant_table((j_common_ptr)dstinfo);
+      MEMCOPY((*qtblptr)->quantval, srcinfo->quant_tbl_ptrs[tblno]->quantval,
+              sizeof((*qtblptr)->quantval));
+      (*qtblptr)->sent_table = FALSE;
+    }
+  }
+  /* Copy the source's per-component info.
+   * Note we assume jpeg_set_defaults has allocated the dest comp_info array.
+   */
+  dstinfo->num_components = srcinfo->num_components;
+  if (dstinfo->num_components < 1 || dstinfo->num_components > MAX_COMPONENTS)
+    ERREXIT2(dstinfo, JERR_COMPONENT_COUNT, dstinfo->num_components,
+             MAX_COMPONENTS);
+  for (ci = 0, incomp = srcinfo->comp_info, outcomp = dstinfo->comp_info;
+       ci < dstinfo->num_components; ci++, incomp++, outcomp++) {
+    outcomp->component_id = incomp->component_id;
+    outcomp->h_samp_factor = incomp->h_samp_factor;
+    outcomp->v_samp_factor = incomp->v_samp_factor;
+    outcomp->quant_tbl_no = incomp->quant_tbl_no;
+    /* Make sure saved quantization table for component matches the qtable
+     * slot.  If not, the input file re-used this qtable slot.
+     * IJG encoder currently cannot duplicate this.
+     */
+    tblno = outcomp->quant_tbl_no;
+    if (tblno < 0 || tblno >= NUM_QUANT_TBLS ||
+        srcinfo->quant_tbl_ptrs[tblno] == NULL)
+      ERREXIT1(dstinfo, JERR_NO_QUANT_TABLE, tblno);
+    slot_quant = srcinfo->quant_tbl_ptrs[tblno];
+    c_quant = incomp->quant_table;
+    if (c_quant != NULL) {
+      for (coefi = 0; coefi < DCTSIZE2; coefi++) {
+        if (c_quant->quantval[coefi] != slot_quant->quantval[coefi])
+          ERREXIT1(dstinfo, JERR_MISMATCHED_QUANT_TABLE, tblno);
+      }
+    }
+    /* Note: we do not copy the source's Huffman table assignments;
+     * instead we rely on jpeg_set_colorspace to have made a suitable choice.
+     */
+  }
+  /* Also copy JFIF version and resolution information, if available.
+   * Strictly speaking this isn't "critical" info, but it's nearly
+   * always appropriate to copy it if available.  In particular,
+   * if the application chooses to copy JFIF 1.02 extension markers from
+   * the source file, we need to copy the version to make sure we don't
+   * emit a file that has 1.02 extensions but a claimed version of 1.01.
+   * We will *not*, however, copy version info from mislabeled "2.01" files.
+   */
+  if (srcinfo->saw_JFIF_marker) {
+    if (srcinfo->JFIF_major_version == 1) {
+      dstinfo->JFIF_major_version = srcinfo->JFIF_major_version;
+      dstinfo->JFIF_minor_version = srcinfo->JFIF_minor_version;
+    }
+    dstinfo->density_unit = srcinfo->density_unit;
+    dstinfo->X_density = srcinfo->X_density;
+    dstinfo->Y_density = srcinfo->Y_density;
+  }
+}
+
+
+/*
+ * Master selection of compression modules for transcoding.
+ * This substitutes for jcinit.c's initialization of the full compressor.
+ */
+
+LOCAL(void)
+transencode_master_selection(j_compress_ptr cinfo,
+                             jvirt_barray_ptr *coef_arrays)
+{
+  /* Although we don't actually use input_components for transcoding,
+   * jcmaster.c's initial_setup will complain if input_components is 0.
+   */
+  cinfo->input_components = 1;
+  /* Initialize master control (includes parameter checking/processing) */
+  jinit_c_master_control(cinfo, TRUE /* transcode only */);
+
+  /* Entropy encoding: either Huffman or arithmetic coding. */
+  if (cinfo->arith_code) {
+#ifdef C_ARITH_CODING_SUPPORTED
+    jinit_arith_encoder(cinfo);
+#else
+    ERREXIT(cinfo, JERR_ARITH_NOTIMPL);
+#endif
+  } else {
+    if (cinfo->progressive_mode) {
+#ifdef C_PROGRESSIVE_SUPPORTED
+      jinit_phuff_encoder(cinfo);
+#else
+      ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif
+    } else
+      jinit_huff_encoder(cinfo);
+  }
+
+  /* We need a special coefficient buffer controller. */
+  transencode_coef_controller(cinfo, coef_arrays);
+
+  jinit_marker_writer(cinfo);
+
+  /* We can now tell the memory manager to allocate virtual arrays. */
+  (*cinfo->mem->realize_virt_arrays) ((j_common_ptr)cinfo);
+
+  /* Write the datastream header (SOI, JFIF) immediately.
+   * Frame and scan headers are postponed till later.
+   * This lets application insert special markers after the SOI.
+   */
+  (*cinfo->marker->write_file_header) (cinfo);
+}
+
+
+/*
+ * The rest of this file is a special implementation of the coefficient
+ * buffer controller.  This is similar to jccoefct.c, but it handles only
+ * output from presupplied virtual arrays.  Furthermore, we generate any
+ * dummy padding blocks on-the-fly rather than expecting them to be present
+ * in the arrays.
+ */
+
+/* Private buffer controller object */
+
+typedef struct {
+  struct jpeg_c_coef_controller pub; /* public fields */
+
+  JDIMENSION iMCU_row_num;      /* iMCU row # within image */
+  JDIMENSION mcu_ctr;           /* counts MCUs processed in current row */
+  int MCU_vert_offset;          /* counts MCU rows within iMCU row */
+  int MCU_rows_per_iMCU_row;    /* number of such rows needed */
+
+  /* Virtual block array for each component. */
+  jvirt_barray_ptr *whole_image;
+
+  /* Workspace for constructing dummy blocks at right/bottom edges. */
+  JBLOCKROW dummy_buffer[C_MAX_BLOCKS_IN_MCU];
+} my_coef_controller;
+
+typedef my_coef_controller *my_coef_ptr;
+
+
+LOCAL(void)
+start_iMCU_row(j_compress_ptr cinfo)
+/* Reset within-iMCU-row counters for a new row */
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+
+  /* In an interleaved scan, an MCU row is the same as an iMCU row.
+   * In a noninterleaved scan, an iMCU row has v_samp_factor MCU rows.
+   * But at the bottom of the image, process only what's left.
+   */
+  if (cinfo->comps_in_scan > 1) {
+    coef->MCU_rows_per_iMCU_row = 1;
+  } else {
+    if (coef->iMCU_row_num < (cinfo->total_iMCU_rows - 1))
+      coef->MCU_rows_per_iMCU_row = cinfo->cur_comp_info[0]->v_samp_factor;
+    else
+      coef->MCU_rows_per_iMCU_row = cinfo->cur_comp_info[0]->last_row_height;
+  }
+
+  coef->mcu_ctr = 0;
+  coef->MCU_vert_offset = 0;
+}
+
+
+/*
+ * Initialize for a processing pass.
+ */
+
+METHODDEF(void)
+start_pass_coef(j_compress_ptr cinfo, J_BUF_MODE pass_mode)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+
+  if (pass_mode != JBUF_CRANK_DEST)
+    ERREXIT(cinfo, JERR_BAD_BUFFER_MODE);
+
+  coef->iMCU_row_num = 0;
+  start_iMCU_row(cinfo);
+}
+
+
+/*
+ * Process some data.
+ * We process the equivalent of one fully interleaved MCU row ("iMCU" row)
+ * per call, ie, v_samp_factor block rows for each component in the scan.
+ * The data is obtained from the virtual arrays and fed to the entropy coder.
+ * Returns TRUE if the iMCU row is completed, FALSE if suspended.
+ *
+ * NB: input_buf is ignored; it is likely to be a NULL pointer.
+ */
+
+METHODDEF(boolean)
+compress_output(j_compress_ptr cinfo, JSAMPIMAGE input_buf)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION MCU_col_num;       /* index of current MCU within row */
+  JDIMENSION last_MCU_col = cinfo->MCUs_per_row - 1;
+  JDIMENSION last_iMCU_row = cinfo->total_iMCU_rows - 1;
+  int blkn, ci, xindex, yindex, yoffset, blockcnt;
+  JDIMENSION start_col;
+  JBLOCKARRAY buffer[MAX_COMPS_IN_SCAN];
+  JBLOCKROW MCU_buffer[C_MAX_BLOCKS_IN_MCU];
+  JBLOCKROW buffer_ptr;
+  jpeg_component_info *compptr;
+
+  /* Align the virtual buffers for the components used in this scan. */
+  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+    compptr = cinfo->cur_comp_info[ci];
+    buffer[ci] = (*cinfo->mem->access_virt_barray)
+      ((j_common_ptr)cinfo, coef->whole_image[compptr->component_index],
+       coef->iMCU_row_num * compptr->v_samp_factor,
+       (JDIMENSION)compptr->v_samp_factor, FALSE);
+  }
+
+  /* Loop to process one whole iMCU row */
+  for (yoffset = coef->MCU_vert_offset; yoffset < coef->MCU_rows_per_iMCU_row;
+       yoffset++) {
+    for (MCU_col_num = coef->mcu_ctr; MCU_col_num < cinfo->MCUs_per_row;
+         MCU_col_num++) {
+      /* Construct list of pointers to DCT blocks belonging to this MCU */
+      blkn = 0;                 /* index of current DCT block within MCU */
+      for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+        compptr = cinfo->cur_comp_info[ci];
+        start_col = MCU_col_num * compptr->MCU_width;
+        blockcnt = (MCU_col_num < last_MCU_col) ? compptr->MCU_width :
+                                                  compptr->last_col_width;
+        for (yindex = 0; yindex < compptr->MCU_height; yindex++) {
+          if (coef->iMCU_row_num < last_iMCU_row ||
+              yindex + yoffset < compptr->last_row_height) {
+            /* Fill in pointers to real blocks in this row */
+            buffer_ptr = buffer[ci][yindex + yoffset] + start_col;
+            for (xindex = 0; xindex < blockcnt; xindex++)
+              MCU_buffer[blkn++] = buffer_ptr++;
+          } else {
+            /* At bottom of image, need a whole row of dummy blocks */
+            xindex = 0;
+          }
+          /* Fill in any dummy blocks needed in this row.
+           * Dummy blocks are filled in the same way as in jccoefct.c:
+           * all zeroes in the AC entries, DC entries equal to previous
+           * block's DC value.  The init routine has already zeroed the
+           * AC entries, so we need only set the DC entries correctly.
+           */
+          for (; xindex < compptr->MCU_width; xindex++) {
+            MCU_buffer[blkn] = coef->dummy_buffer[blkn];
+            MCU_buffer[blkn][0][0] = MCU_buffer[blkn - 1][0][0];
+            blkn++;
+          }
+        }
+      }
+      /* Try to write the MCU. */
+      if (!(*cinfo->entropy->encode_mcu) (cinfo, MCU_buffer)) {
+        /* Suspension forced; update state counters and exit */
+        coef->MCU_vert_offset = yoffset;
+        coef->mcu_ctr = MCU_col_num;
+        return FALSE;
+      }
+    }
+    /* Completed an MCU row, but perhaps not an iMCU row */
+    coef->mcu_ctr = 0;
+  }
+  /* Completed the iMCU row, advance counters for next one */
+  coef->iMCU_row_num++;
+  start_iMCU_row(cinfo);
+  return TRUE;
+}
+
+
+/*
+ * Initialize coefficient buffer controller.
+ *
+ * Each passed coefficient array must be the right size for that
+ * coefficient: width_in_blocks wide and height_in_blocks high,
+ * with unitheight at least v_samp_factor.
+ */
+
+LOCAL(void)
+transencode_coef_controller(j_compress_ptr cinfo,
+                            jvirt_barray_ptr *coef_arrays)
+{
+  my_coef_ptr coef;
+  JBLOCKROW buffer;
+  int i;
+
+  coef = (my_coef_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_coef_controller));
+  cinfo->coef = (struct jpeg_c_coef_controller *)coef;
+  coef->pub.start_pass = start_pass_coef;
+  coef->pub.compress_data = compress_output;
+
+  /* Save pointer to virtual arrays */
+  coef->whole_image = coef_arrays;
+
+  /* Allocate and pre-zero space for dummy DCT blocks. */
+  buffer = (JBLOCKROW)
+    (*cinfo->mem->alloc_large) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                C_MAX_BLOCKS_IN_MCU * sizeof(JBLOCK));
+  jzero_far((void *)buffer, C_MAX_BLOCKS_IN_MCU * sizeof(JBLOCK));
+  for (i = 0; i < C_MAX_BLOCKS_IN_MCU; i++) {
+    coef->dummy_buffer[i] = buffer + i;
+  }
+}

3rdparty/libjpeg-turbo/src/jdapimin.c

@@ -0,0 +1,407 @@
+/*
+ * jdapimin.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1998, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2016, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains application interface code for the decompression half
+ * of the JPEG library.  These are the "minimum" API routines that may be
+ * needed in either the normal full-decompression case or the
+ * transcoding-only case.
+ *
+ * Most of the routines intended to be called directly by an application
+ * are in this file or in jdapistd.c.  But also see jcomapi.c for routines
+ * shared by compression and decompression, and jdtrans.c for the transcoding
+ * case.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jdmaster.h"
+
+
+/*
+ * Initialization of a JPEG decompression object.
+ * The error manager must already be set up (in case memory manager fails).
+ */
+
+GLOBAL(void)
+jpeg_CreateDecompress(j_decompress_ptr cinfo, int version, size_t structsize)
+{
+  int i;
+
+  /* Guard against version mismatches between library and caller. */
+  cinfo->mem = NULL;            /* so jpeg_destroy knows mem mgr not called */
+  if (version != JPEG_LIB_VERSION)
+    ERREXIT2(cinfo, JERR_BAD_LIB_VERSION, JPEG_LIB_VERSION, version);
+  if (structsize != sizeof(struct jpeg_decompress_struct))
+    ERREXIT2(cinfo, JERR_BAD_STRUCT_SIZE,
+             (int)sizeof(struct jpeg_decompress_struct), (int)structsize);
+
+  /* For debugging purposes, we zero the whole master structure.
+   * But the application has already set the err pointer, and may have set
+   * client_data, so we have to save and restore those fields.
+   * Note: if application hasn't set client_data, tools like Purify may
+   * complain here.
+   */
+  {
+    struct jpeg_error_mgr *err = cinfo->err;
+    void *client_data = cinfo->client_data; /* ignore Purify complaint here */
+    MEMZERO(cinfo, sizeof(struct jpeg_decompress_struct));
+    cinfo->err = err;
+    cinfo->client_data = client_data;
+  }
+  cinfo->is_decompressor = TRUE;
+
+  /* Initialize a memory manager instance for this object */
+  jinit_memory_mgr((j_common_ptr)cinfo);
+
+  /* Zero out pointers to permanent structures. */
+  cinfo->progress = NULL;
+  cinfo->src = NULL;
+
+  for (i = 0; i < NUM_QUANT_TBLS; i++)
+    cinfo->quant_tbl_ptrs[i] = NULL;
+
+  for (i = 0; i < NUM_HUFF_TBLS; i++) {
+    cinfo->dc_huff_tbl_ptrs[i] = NULL;
+    cinfo->ac_huff_tbl_ptrs[i] = NULL;
+  }
+
+  /* Initialize marker processor so application can override methods
+   * for COM, APPn markers before calling jpeg_read_header.
+   */
+  cinfo->marker_list = NULL;
+  jinit_marker_reader(cinfo);
+
+  /* And initialize the overall input controller. */
+  jinit_input_controller(cinfo);
+
+  /* OK, I'm ready */
+  cinfo->global_state = DSTATE_START;
+
+  /* The master struct is used to store extension parameters, so we allocate it
+   * here.
+   */
+  cinfo->master = (struct jpeg_decomp_master *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                                sizeof(my_decomp_master));
+  MEMZERO(cinfo->master, sizeof(my_decomp_master));
+}
+
+
+/*
+ * Destruction of a JPEG decompression object
+ */
+
+GLOBAL(void)
+jpeg_destroy_decompress(j_decompress_ptr cinfo)
+{
+  jpeg_destroy((j_common_ptr)cinfo); /* use common routine */
+}
+
+
+/*
+ * Abort processing of a JPEG decompression operation,
+ * but don't destroy the object itself.
+ */
+
+GLOBAL(void)
+jpeg_abort_decompress(j_decompress_ptr cinfo)
+{
+  jpeg_abort((j_common_ptr)cinfo); /* use common routine */
+}
+
+
+/*
+ * Set default decompression parameters.
+ */
+
+LOCAL(void)
+default_decompress_parms(j_decompress_ptr cinfo)
+{
+  /* Guess the input colorspace, and set output colorspace accordingly. */
+  /* (Wish JPEG committee had provided a real way to specify this...) */
+  /* Note application may override our guesses. */
+  switch (cinfo->num_components) {
+  case 1:
+    cinfo->jpeg_color_space = JCS_GRAYSCALE;
+    cinfo->out_color_space = JCS_GRAYSCALE;
+    break;
+
+  case 3:
+    if (cinfo->saw_JFIF_marker) {
+      cinfo->jpeg_color_space = JCS_YCbCr; /* JFIF implies YCbCr */
+    } else if (cinfo->saw_Adobe_marker) {
+      switch (cinfo->Adobe_transform) {
+      case 0:
+        cinfo->jpeg_color_space = JCS_RGB;
+        break;
+      case 1:
+        cinfo->jpeg_color_space = JCS_YCbCr;
+        break;
+      default:
+        WARNMS1(cinfo, JWRN_ADOBE_XFORM, cinfo->Adobe_transform);
+        cinfo->jpeg_color_space = JCS_YCbCr; /* assume it's YCbCr */
+        break;
+      }
+    } else {
+      /* Saw no special markers, try to guess from the component IDs */
+      int cid0 = cinfo->comp_info[0].component_id;
+      int cid1 = cinfo->comp_info[1].component_id;
+      int cid2 = cinfo->comp_info[2].component_id;
+
+      if (cid0 == 1 && cid1 == 2 && cid2 == 3)
+        cinfo->jpeg_color_space = JCS_YCbCr; /* assume JFIF w/out marker */
+      else if (cid0 == 82 && cid1 == 71 && cid2 == 66)
+        cinfo->jpeg_color_space = JCS_RGB; /* ASCII 'R', 'G', 'B' */
+      else {
+        TRACEMS3(cinfo, 1, JTRC_UNKNOWN_IDS, cid0, cid1, cid2);
+        cinfo->jpeg_color_space = JCS_YCbCr; /* assume it's YCbCr */
+      }
+    }
+    /* Always guess RGB is proper output colorspace. */
+    cinfo->out_color_space = JCS_RGB;
+    break;
+
+  case 4:
+    if (cinfo->saw_Adobe_marker) {
+      switch (cinfo->Adobe_transform) {
+      case 0:
+        cinfo->jpeg_color_space = JCS_CMYK;
+        break;
+      case 2:
+        cinfo->jpeg_color_space = JCS_YCCK;
+        break;
+      default:
+        WARNMS1(cinfo, JWRN_ADOBE_XFORM, cinfo->Adobe_transform);
+        cinfo->jpeg_color_space = JCS_YCCK; /* assume it's YCCK */
+        break;
+      }
+    } else {
+      /* No special markers, assume straight CMYK. */
+      cinfo->jpeg_color_space = JCS_CMYK;
+    }
+    cinfo->out_color_space = JCS_CMYK;
+    break;
+
+  default:
+    cinfo->jpeg_color_space = JCS_UNKNOWN;
+    cinfo->out_color_space = JCS_UNKNOWN;
+    break;
+  }
+
+  /* Set defaults for other decompression parameters. */
+  cinfo->scale_num = 1;         /* 1:1 scaling */
+  cinfo->scale_denom = 1;
+  cinfo->output_gamma = 1.0;
+  cinfo->buffered_image = FALSE;
+  cinfo->raw_data_out = FALSE;
+  cinfo->dct_method = JDCT_DEFAULT;
+  cinfo->do_fancy_upsampling = TRUE;
+  cinfo->do_block_smoothing = TRUE;
+  cinfo->quantize_colors = FALSE;
+  /* We set these in case application only sets quantize_colors. */
+  cinfo->dither_mode = JDITHER_FS;
+#ifdef QUANT_2PASS_SUPPORTED
+  cinfo->two_pass_quantize = TRUE;
+#else
+  cinfo->two_pass_quantize = FALSE;
+#endif
+  cinfo->desired_number_of_colors = 256;
+  cinfo->colormap = NULL;
+  /* Initialize for no mode change in buffered-image mode. */
+  cinfo->enable_1pass_quant = FALSE;
+  cinfo->enable_external_quant = FALSE;
+  cinfo->enable_2pass_quant = FALSE;
+}
+
+
+/*
+ * Decompression startup: read start of JPEG datastream to see what's there.
+ * Need only initialize JPEG object and supply a data source before calling.
+ *
+ * This routine will read as far as the first SOS marker (ie, actual start of
+ * compressed data), and will save all tables and parameters in the JPEG
+ * object.  It will also initialize the decompression parameters to default
+ * values, and finally return JPEG_HEADER_OK.  On return, the application may
+ * adjust the decompression parameters and then call jpeg_start_decompress.
+ * (Or, if the application only wanted to determine the image parameters,
+ * the data need not be decompressed.  In that case, call jpeg_abort or
+ * jpeg_destroy to release any temporary space.)
+ * If an abbreviated (tables only) datastream is presented, the routine will
+ * return JPEG_HEADER_TABLES_ONLY upon reaching EOI.  The application may then
+ * re-use the JPEG object to read the abbreviated image datastream(s).
+ * It is unnecessary (but OK) to call jpeg_abort in this case.
+ * The JPEG_SUSPENDED return code only occurs if the data source module
+ * requests suspension of the decompressor.  In this case the application
+ * should load more source data and then re-call jpeg_read_header to resume
+ * processing.
+ * If a non-suspending data source is used and require_image is TRUE, then the
+ * return code need not be inspected since only JPEG_HEADER_OK is possible.
+ *
+ * This routine is now just a front end to jpeg_consume_input, with some
+ * extra error checking.
+ */
+
+GLOBAL(int)
+jpeg_read_header(j_decompress_ptr cinfo, boolean require_image)
+{
+  int retcode;
+
+  if (cinfo->global_state != DSTATE_START &&
+      cinfo->global_state != DSTATE_INHEADER)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  retcode = jpeg_consume_input(cinfo);
+
+  switch (retcode) {
+  case JPEG_REACHED_SOS:
+    retcode = JPEG_HEADER_OK;
+    break;
+  case JPEG_REACHED_EOI:
+    if (require_image)          /* Complain if application wanted an image */
+      ERREXIT(cinfo, JERR_NO_IMAGE);
+    /* Reset to start state; it would be safer to require the application to
+     * call jpeg_abort, but we can't change it now for compatibility reasons.
+     * A side effect is to free any temporary memory (there shouldn't be any).
+     */
+    jpeg_abort((j_common_ptr)cinfo); /* sets state = DSTATE_START */
+    retcode = JPEG_HEADER_TABLES_ONLY;
+    break;
+  case JPEG_SUSPENDED:
+    /* no work */
+    break;
+  }
+
+  return retcode;
+}
+
+
+/*
+ * Consume data in advance of what the decompressor requires.
+ * This can be called at any time once the decompressor object has
+ * been created and a data source has been set up.
+ *
+ * This routine is essentially a state machine that handles a couple
+ * of critical state-transition actions, namely initial setup and
+ * transition from header scanning to ready-for-start_decompress.
+ * All the actual input is done via the input controller's consume_input
+ * method.
+ */
+
+GLOBAL(int)
+jpeg_consume_input(j_decompress_ptr cinfo)
+{
+  int retcode = JPEG_SUSPENDED;
+
+  /* NB: every possible DSTATE value should be listed in this switch */
+  switch (cinfo->global_state) {
+  case DSTATE_START:
+    /* Start-of-datastream actions: reset appropriate modules */
+    (*cinfo->inputctl->reset_input_controller) (cinfo);
+    /* Initialize application's data source module */
+    (*cinfo->src->init_source) (cinfo);
+    cinfo->global_state = DSTATE_INHEADER;
+    /*FALLTHROUGH*/
+  case DSTATE_INHEADER:
+    retcode = (*cinfo->inputctl->consume_input) (cinfo);
+    if (retcode == JPEG_REACHED_SOS) { /* Found SOS, prepare to decompress */
+      /* Set up default parameters based on header data */
+      default_decompress_parms(cinfo);
+      /* Set global state: ready for start_decompress */
+      cinfo->global_state = DSTATE_READY;
+    }
+    break;
+  case DSTATE_READY:
+    /* Can't advance past first SOS until start_decompress is called */
+    retcode = JPEG_REACHED_SOS;
+    break;
+  case DSTATE_PRELOAD:
+  case DSTATE_PRESCAN:
+  case DSTATE_SCANNING:
+  case DSTATE_RAW_OK:
+  case DSTATE_BUFIMAGE:
+  case DSTATE_BUFPOST:
+  case DSTATE_STOPPING:
+    retcode = (*cinfo->inputctl->consume_input) (cinfo);
+    break;
+  default:
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  }
+  return retcode;
+}
+
+
+/*
+ * Have we finished reading the input file?
+ */
+
+GLOBAL(boolean)
+jpeg_input_complete(j_decompress_ptr cinfo)
+{
+  /* Check for valid jpeg object */
+  if (cinfo->global_state < DSTATE_START ||
+      cinfo->global_state > DSTATE_STOPPING)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  return cinfo->inputctl->eoi_reached;
+}
+
+
+/*
+ * Is there more than one scan?
+ */
+
+GLOBAL(boolean)
+jpeg_has_multiple_scans(j_decompress_ptr cinfo)
+{
+  /* Only valid after jpeg_read_header completes */
+  if (cinfo->global_state < DSTATE_READY ||
+      cinfo->global_state > DSTATE_STOPPING)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  return cinfo->inputctl->has_multiple_scans;
+}
+
+
+/*
+ * Finish JPEG decompression.
+ *
+ * This will normally just verify the file trailer and release temp storage.
+ *
+ * Returns FALSE if suspended.  The return value need be inspected only if
+ * a suspending data source is used.
+ */
+
+GLOBAL(boolean)
+jpeg_finish_decompress(j_decompress_ptr cinfo)
+{
+  if ((cinfo->global_state == DSTATE_SCANNING ||
+       cinfo->global_state == DSTATE_RAW_OK) && !cinfo->buffered_image) {
+    /* Terminate final pass of non-buffered mode */
+    if (cinfo->output_scanline < cinfo->output_height)
+      ERREXIT(cinfo, JERR_TOO_LITTLE_DATA);
+    (*cinfo->master->finish_output_pass) (cinfo);
+    cinfo->global_state = DSTATE_STOPPING;
+  } else if (cinfo->global_state == DSTATE_BUFIMAGE) {
+    /* Finishing after a buffered-image operation */
+    cinfo->global_state = DSTATE_STOPPING;
+  } else if (cinfo->global_state != DSTATE_STOPPING) {
+    /* STOPPING = repeat call after a suspension, anything else is error */
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  }
+  /* Read until EOI */
+  while (!cinfo->inputctl->eoi_reached) {
+    if ((*cinfo->inputctl->consume_input) (cinfo) == JPEG_SUSPENDED)
+      return FALSE;             /* Suspend, come back later */
+  }
+  /* Do final cleanup */
+  (*cinfo->src->term_source) (cinfo);
+  /* We can use jpeg_abort to release memory and reset global_state */
+  jpeg_abort((j_common_ptr)cinfo);
+  return TRUE;
+}

3rdparty/libjpeg-turbo/src/jdapistd.c

@@ -0,0 +1,639 @@
+/*
+ * jdapistd.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2010, 2015-2018, D. R. Commander.
+ * Copyright (C) 2015, Google, Inc.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains application interface code for the decompression half
+ * of the JPEG library.  These are the "standard" API routines that are
+ * used in the normal full-decompression case.  They are not used by a
+ * transcoding-only application.  Note that if an application links in
+ * jpeg_start_decompress, it will end up linking in the entire decompressor.
+ * We thus must separate this file from jdapimin.c to avoid linking the
+ * whole decompression library into a transcoder.
+ */
+
+#include "jinclude.h"
+#include "jdmainct.h"
+#include "jdcoefct.h"
+#include "jdsample.h"
+#include "jmemsys.h"
+
+/* Forward declarations */
+LOCAL(boolean) output_pass_setup(j_decompress_ptr cinfo);
+
+
+/*
+ * Decompression initialization.
+ * jpeg_read_header must be completed before calling this.
+ *
+ * If a multipass operating mode was selected, this will do all but the
+ * last pass, and thus may take a great deal of time.
+ *
+ * Returns FALSE if suspended.  The return value need be inspected only if
+ * a suspending data source is used.
+ */
+
+GLOBAL(boolean)
+jpeg_start_decompress(j_decompress_ptr cinfo)
+{
+  if (cinfo->global_state == DSTATE_READY) {
+    /* First call: initialize master control, select active modules */
+    jinit_master_decompress(cinfo);
+    if (cinfo->buffered_image) {
+      /* No more work here; expecting jpeg_start_output next */
+      cinfo->global_state = DSTATE_BUFIMAGE;
+      return TRUE;
+    }
+    cinfo->global_state = DSTATE_PRELOAD;
+  }
+  if (cinfo->global_state == DSTATE_PRELOAD) {
+    /* If file has multiple scans, absorb them all into the coef buffer */
+    if (cinfo->inputctl->has_multiple_scans) {
+#ifdef D_MULTISCAN_FILES_SUPPORTED
+      for (;;) {
+        int retcode;
+        /* Call progress monitor hook if present */
+        if (cinfo->progress != NULL)
+          (*cinfo->progress->progress_monitor) ((j_common_ptr)cinfo);
+        /* Absorb some more input */
+        retcode = (*cinfo->inputctl->consume_input) (cinfo);
+        if (retcode == JPEG_SUSPENDED)
+          return FALSE;
+        if (retcode == JPEG_REACHED_EOI)
+          break;
+        /* Advance progress counter if appropriate */
+        if (cinfo->progress != NULL &&
+            (retcode == JPEG_ROW_COMPLETED || retcode == JPEG_REACHED_SOS)) {
+          if (++cinfo->progress->pass_counter >= cinfo->progress->pass_limit) {
+            /* jdmaster underestimated number of scans; ratchet up one scan */
+            cinfo->progress->pass_limit += (long)cinfo->total_iMCU_rows;
+          }
+        }
+      }
+#else
+      ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif /* D_MULTISCAN_FILES_SUPPORTED */
+    }
+    cinfo->output_scan_number = cinfo->input_scan_number;
+  } else if (cinfo->global_state != DSTATE_PRESCAN)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  /* Perform any dummy output passes, and set up for the final pass */
+  return output_pass_setup(cinfo);
+}
+
+
+/*
+ * Set up for an output pass, and perform any dummy pass(es) needed.
+ * Common subroutine for jpeg_start_decompress and jpeg_start_output.
+ * Entry: global_state = DSTATE_PRESCAN only if previously suspended.
+ * Exit: If done, returns TRUE and sets global_state for proper output mode.
+ *       If suspended, returns FALSE and sets global_state = DSTATE_PRESCAN.
+ */
+
+LOCAL(boolean)
+output_pass_setup(j_decompress_ptr cinfo)
+{
+  if (cinfo->global_state != DSTATE_PRESCAN) {
+    /* First call: do pass setup */
+    (*cinfo->master->prepare_for_output_pass) (cinfo);
+    cinfo->output_scanline = 0;
+    cinfo->global_state = DSTATE_PRESCAN;
+  }
+  /* Loop over any required dummy passes */
+  while (cinfo->master->is_dummy_pass) {
+#ifdef QUANT_2PASS_SUPPORTED
+    /* Crank through the dummy pass */
+    while (cinfo->output_scanline < cinfo->output_height) {
+      JDIMENSION last_scanline;
+      /* Call progress monitor hook if present */
+      if (cinfo->progress != NULL) {
+        cinfo->progress->pass_counter = (long)cinfo->output_scanline;
+        cinfo->progress->pass_limit = (long)cinfo->output_height;
+        (*cinfo->progress->progress_monitor) ((j_common_ptr)cinfo);
+      }
+      /* Process some data */
+      last_scanline = cinfo->output_scanline;
+      (*cinfo->main->process_data) (cinfo, (JSAMPARRAY)NULL,
+                                    &cinfo->output_scanline, (JDIMENSION)0);
+      if (cinfo->output_scanline == last_scanline)
+        return FALSE;           /* No progress made, must suspend */
+    }
+    /* Finish up dummy pass, and set up for another one */
+    (*cinfo->master->finish_output_pass) (cinfo);
+    (*cinfo->master->prepare_for_output_pass) (cinfo);
+    cinfo->output_scanline = 0;
+#else
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif /* QUANT_2PASS_SUPPORTED */
+  }
+  /* Ready for application to drive output pass through
+   * jpeg_read_scanlines or jpeg_read_raw_data.
+   */
+  cinfo->global_state = cinfo->raw_data_out ? DSTATE_RAW_OK : DSTATE_SCANNING;
+  return TRUE;
+}
+
+
+/*
+ * Enable partial scanline decompression
+ *
+ * Must be called after jpeg_start_decompress() and before any calls to
+ * jpeg_read_scanlines() or jpeg_skip_scanlines().
+ *
+ * Refer to libjpeg.txt for more information.
+ */
+
+GLOBAL(void)
+jpeg_crop_scanline(j_decompress_ptr cinfo, JDIMENSION *xoffset,
+                   JDIMENSION *width)
+{
+  int ci, align, orig_downsampled_width;
+  JDIMENSION input_xoffset;
+  boolean reinit_upsampler = FALSE;
+  jpeg_component_info *compptr;
+
+  if (cinfo->global_state != DSTATE_SCANNING || cinfo->output_scanline != 0)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  if (!xoffset || !width)
+    ERREXIT(cinfo, JERR_BAD_CROP_SPEC);
+
+  /* xoffset and width must fall within the output image dimensions. */
+  if (*width == 0 || *xoffset + *width > cinfo->output_width)
+    ERREXIT(cinfo, JERR_WIDTH_OVERFLOW);
+
+  /* No need to do anything if the caller wants the entire width. */
+  if (*width == cinfo->output_width)
+    return;
+
+  /* Ensuring the proper alignment of xoffset is tricky.  At minimum, it
+   * must align with an MCU boundary, because:
+   *
+   *   (1) The IDCT is performed in blocks, and it is not feasible to modify
+   *       the algorithm so that it can transform partial blocks.
+   *   (2) Because of the SIMD extensions, any input buffer passed to the
+   *       upsampling and color conversion routines must be aligned to the
+   *       SIMD word size (for instance, 128-bit in the case of SSE2.)  The
+   *       easiest way to accomplish this without copying data is to ensure
+   *       that upsampling and color conversion begin at the start of the
+   *       first MCU column that will be inverse transformed.
+   *
+   * In practice, we actually impose a stricter alignment requirement.  We
+   * require that xoffset be a multiple of the maximum MCU column width of all
+   * of the components (the "iMCU column width.")  This is to simplify the
+   * single-pass decompression case, allowing us to use the same MCU column
+   * width for all of the components.
+   */
+  if (cinfo->comps_in_scan == 1 && cinfo->num_components == 1)
+    align = cinfo->_min_DCT_scaled_size;
+  else
+    align = cinfo->_min_DCT_scaled_size * cinfo->max_h_samp_factor;
+
+  /* Adjust xoffset to the nearest iMCU boundary <= the requested value */
+  input_xoffset = *xoffset;
+  *xoffset = (input_xoffset / align) * align;
+
+  /* Adjust the width so that the right edge of the output image is as
+   * requested (only the left edge is altered.)  It is important that calling
+   * programs check this value after this function returns, so that they can
+   * allocate an output buffer with the appropriate size.
+   */
+  *width = *width + input_xoffset - *xoffset;
+  cinfo->output_width = *width;
+
+  /* Set the first and last iMCU columns that we must decompress.  These values
+   * will be used in single-scan decompressions.
+   */
+  cinfo->master->first_iMCU_col = (JDIMENSION)(long)(*xoffset) / (long)align;
+  cinfo->master->last_iMCU_col =
+    (JDIMENSION)jdiv_round_up((long)(*xoffset + cinfo->output_width),
+                              (long)align) - 1;
+
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    int hsf = (cinfo->comps_in_scan == 1 && cinfo->num_components == 1) ?
+              1 : compptr->h_samp_factor;
+
+    /* Set downsampled_width to the new output width. */
+    orig_downsampled_width = compptr->downsampled_width;
+    compptr->downsampled_width =
+      (JDIMENSION)jdiv_round_up((long)(cinfo->output_width *
+                                       compptr->h_samp_factor),
+                                (long)cinfo->max_h_samp_factor);
+    if (compptr->downsampled_width < 2 && orig_downsampled_width >= 2)
+      reinit_upsampler = TRUE;
+
+    /* Set the first and last iMCU columns that we must decompress.  These
+     * values will be used in multi-scan decompressions.
+     */
+    cinfo->master->first_MCU_col[ci] =
+      (JDIMENSION)(long)(*xoffset * hsf) / (long)align;
+    cinfo->master->last_MCU_col[ci] =
+      (JDIMENSION)jdiv_round_up((long)((*xoffset + cinfo->output_width) * hsf),
+                                (long)align) - 1;
+  }
+
+  if (reinit_upsampler) {
+    cinfo->master->jinit_upsampler_no_alloc = TRUE;
+    jinit_upsampler(cinfo);
+    cinfo->master->jinit_upsampler_no_alloc = FALSE;
+  }
+}
+
+
+/*
+ * Read some scanlines of data from the JPEG decompressor.
+ *
+ * The return value will be the number of lines actually read.
+ * This may be less than the number requested in several cases,
+ * including bottom of image, data source suspension, and operating
+ * modes that emit multiple scanlines at a time.
+ *
+ * Note: we warn about excess calls to jpeg_read_scanlines() since
+ * this likely signals an application programmer error.  However,
+ * an oversize buffer (max_lines > scanlines remaining) is not an error.
+ */
+
+GLOBAL(JDIMENSION)
+jpeg_read_scanlines(j_decompress_ptr cinfo, JSAMPARRAY scanlines,
+                    JDIMENSION max_lines)
+{
+  JDIMENSION row_ctr;
+
+  if (cinfo->global_state != DSTATE_SCANNING)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  if (cinfo->output_scanline >= cinfo->output_height) {
+    WARNMS(cinfo, JWRN_TOO_MUCH_DATA);
+    return 0;
+  }
+
+  /* Call progress monitor hook if present */
+  if (cinfo->progress != NULL) {
+    cinfo->progress->pass_counter = (long)cinfo->output_scanline;
+    cinfo->progress->pass_limit = (long)cinfo->output_height;
+    (*cinfo->progress->progress_monitor) ((j_common_ptr)cinfo);
+  }
+
+  /* Process some data */
+  row_ctr = 0;
+  (*cinfo->main->process_data) (cinfo, scanlines, &row_ctr, max_lines);
+  cinfo->output_scanline += row_ctr;
+  return row_ctr;
+}
+
+
+/* Dummy color convert function used by jpeg_skip_scanlines() */
+LOCAL(void)
+noop_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+             JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+}
+
+
+/* Dummy quantize function used by jpeg_skip_scanlines() */
+LOCAL(void)
+noop_quantize(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
+              JSAMPARRAY output_buf, int num_rows)
+{
+}
+
+
+/*
+ * In some cases, it is best to call jpeg_read_scanlines() and discard the
+ * output, rather than skipping the scanlines, because this allows us to
+ * maintain the internal state of the context-based upsampler.  In these cases,
+ * we set up and tear down a dummy color converter in order to avoid valgrind
+ * errors and to achieve the best possible performance.
+ */
+
+LOCAL(void)
+read_and_discard_scanlines(j_decompress_ptr cinfo, JDIMENSION num_lines)
+{
+  JDIMENSION n;
+  void (*color_convert) (j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                         JDIMENSION input_row, JSAMPARRAY output_buf,
+                         int num_rows) = NULL;
+  void (*color_quantize) (j_decompress_ptr cinfo, JSAMPARRAY input_buf,
+                          JSAMPARRAY output_buf, int num_rows) = NULL;
+
+  if (cinfo->cconvert && cinfo->cconvert->color_convert) {
+    color_convert = cinfo->cconvert->color_convert;
+    cinfo->cconvert->color_convert = noop_convert;
+  }
+
+  if (cinfo->cquantize && cinfo->cquantize->color_quantize) {
+    color_quantize = cinfo->cquantize->color_quantize;
+    cinfo->cquantize->color_quantize = noop_quantize;
+  }
+
+  for (n = 0; n < num_lines; n++)
+    jpeg_read_scanlines(cinfo, NULL, 1);
+
+  if (color_convert)
+    cinfo->cconvert->color_convert = color_convert;
+
+  if (color_quantize)
+    cinfo->cquantize->color_quantize = color_quantize;
+}
+
+
+/*
+ * Called by jpeg_skip_scanlines().  This partially skips a decompress block by
+ * incrementing the rowgroup counter.
+ */
+
+LOCAL(void)
+increment_simple_rowgroup_ctr(j_decompress_ptr cinfo, JDIMENSION rows)
+{
+  JDIMENSION rows_left;
+  my_main_ptr main_ptr = (my_main_ptr)cinfo->main;
+
+  /* Increment the counter to the next row group after the skipped rows. */
+  main_ptr->rowgroup_ctr += rows / cinfo->max_v_samp_factor;
+
+  /* Partially skipping a row group would involve modifying the internal state
+   * of the upsampler, so read the remaining rows into a dummy buffer instead.
+   */
+  rows_left = rows % cinfo->max_v_samp_factor;
+  cinfo->output_scanline += rows - rows_left;
+
+  read_and_discard_scanlines(cinfo, rows_left);
+}
+
+/*
+ * Skips some scanlines of data from the JPEG decompressor.
+ *
+ * The return value will be the number of lines actually skipped.  If skipping
+ * num_lines would move beyond the end of the image, then the actual number of
+ * lines remaining in the image is returned.  Otherwise, the return value will
+ * be equal to num_lines.
+ *
+ * Refer to libjpeg.txt for more information.
+ */
+
+GLOBAL(JDIMENSION)
+jpeg_skip_scanlines(j_decompress_ptr cinfo, JDIMENSION num_lines)
+{
+  my_main_ptr main_ptr = (my_main_ptr)cinfo->main;
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  my_upsample_ptr upsample = (my_upsample_ptr)cinfo->upsample;
+  JDIMENSION i, x;
+  int y;
+  JDIMENSION lines_per_iMCU_row, lines_left_in_iMCU_row, lines_after_iMCU_row;
+  JDIMENSION lines_to_skip, lines_to_read;
+
+  if (cinfo->global_state != DSTATE_SCANNING)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+
+  /* Do not skip past the bottom of the image. */
+  if (cinfo->output_scanline + num_lines >= cinfo->output_height) {
+    cinfo->output_scanline = cinfo->output_height;
+    (*cinfo->inputctl->finish_input_pass) (cinfo);
+    cinfo->inputctl->eoi_reached = TRUE;
+    return cinfo->output_height - cinfo->output_scanline;
+  }
+
+  if (num_lines == 0)
+    return 0;
+
+  lines_per_iMCU_row = cinfo->_min_DCT_scaled_size * cinfo->max_v_samp_factor;
+  lines_left_in_iMCU_row =
+    (lines_per_iMCU_row - (cinfo->output_scanline % lines_per_iMCU_row)) %
+    lines_per_iMCU_row;
+  lines_after_iMCU_row = num_lines - lines_left_in_iMCU_row;
+
+  /* Skip the lines remaining in the current iMCU row.  When upsampling
+   * requires context rows, we need the previous and next rows in order to read
+   * the current row.  This adds some complexity.
+   */
+  if (cinfo->upsample->need_context_rows) {
+    /* If the skipped lines would not move us past the current iMCU row, we
+     * read the lines and ignore them.  There might be a faster way of doing
+     * this, but we are facing increasing complexity for diminishing returns.
+     * The increasing complexity would be a by-product of meddling with the
+     * state machine used to skip context rows.  Near the end of an iMCU row,
+     * the next iMCU row may have already been entropy-decoded.  In this unique
+     * case, we will read the next iMCU row if we cannot skip past it as well.
+     */
+    if ((num_lines < lines_left_in_iMCU_row + 1) ||
+        (lines_left_in_iMCU_row <= 1 && main_ptr->buffer_full &&
+         lines_after_iMCU_row < lines_per_iMCU_row + 1)) {
+      read_and_discard_scanlines(cinfo, num_lines);
+      return num_lines;
+    }
+
+    /* If the next iMCU row has already been entropy-decoded, make sure that
+     * we do not skip too far.
+     */
+    if (lines_left_in_iMCU_row <= 1 && main_ptr->buffer_full) {
+      cinfo->output_scanline += lines_left_in_iMCU_row + lines_per_iMCU_row;
+      lines_after_iMCU_row -= lines_per_iMCU_row;
+    } else {
+      cinfo->output_scanline += lines_left_in_iMCU_row;
+    }
+
+    /* If we have just completed the first block, adjust the buffer pointers */
+    if (main_ptr->iMCU_row_ctr == 0 ||
+        (main_ptr->iMCU_row_ctr == 1 && lines_left_in_iMCU_row > 2))
+      set_wraparound_pointers(cinfo);
+    main_ptr->buffer_full = FALSE;
+    main_ptr->rowgroup_ctr = 0;
+    main_ptr->context_state = CTX_PREPARE_FOR_IMCU;
+    upsample->next_row_out = cinfo->max_v_samp_factor;
+    upsample->rows_to_go = cinfo->output_height - cinfo->output_scanline;
+  }
+
+  /* Skipping is much simpler when context rows are not required. */
+  else {
+    if (num_lines < lines_left_in_iMCU_row) {
+      increment_simple_rowgroup_ctr(cinfo, num_lines);
+      return num_lines;
+    } else {
+      cinfo->output_scanline += lines_left_in_iMCU_row;
+      main_ptr->buffer_full = FALSE;
+      main_ptr->rowgroup_ctr = 0;
+      upsample->next_row_out = cinfo->max_v_samp_factor;
+      upsample->rows_to_go = cinfo->output_height - cinfo->output_scanline;
+    }
+  }
+
+  /* Calculate how many full iMCU rows we can skip. */
+  if (cinfo->upsample->need_context_rows)
+    lines_to_skip = ((lines_after_iMCU_row - 1) / lines_per_iMCU_row) *
+                    lines_per_iMCU_row;
+  else
+    lines_to_skip = (lines_after_iMCU_row / lines_per_iMCU_row) *
+                    lines_per_iMCU_row;
+  /* Calculate the number of lines that remain to be skipped after skipping all
+   * of the full iMCU rows that we can.  We will not read these lines unless we
+   * have to.
+   */
+  lines_to_read = lines_after_iMCU_row - lines_to_skip;
+
+  /* For images requiring multiple scans (progressive, non-interleaved, etc.),
+   * all of the entropy decoding occurs in jpeg_start_decompress(), assuming
+   * that the input data source is non-suspending.  This makes skipping easy.
+   */
+  if (cinfo->inputctl->has_multiple_scans) {
+    if (cinfo->upsample->need_context_rows) {
+      cinfo->output_scanline += lines_to_skip;
+      cinfo->output_iMCU_row += lines_to_skip / lines_per_iMCU_row;
+      main_ptr->iMCU_row_ctr += lines_to_skip / lines_per_iMCU_row;
+      /* It is complex to properly move to the middle of a context block, so
+       * read the remaining lines instead of skipping them.
+       */
+      read_and_discard_scanlines(cinfo, lines_to_read);
+    } else {
+      cinfo->output_scanline += lines_to_skip;
+      cinfo->output_iMCU_row += lines_to_skip / lines_per_iMCU_row;
+      increment_simple_rowgroup_ctr(cinfo, lines_to_read);
+    }
+    upsample->rows_to_go = cinfo->output_height - cinfo->output_scanline;
+    return num_lines;
+  }
+
+  /* Skip the iMCU rows that we can safely skip. */
+  for (i = 0; i < lines_to_skip; i += lines_per_iMCU_row) {
+    for (y = 0; y < coef->MCU_rows_per_iMCU_row; y++) {
+      for (x = 0; x < cinfo->MCUs_per_row; x++) {
+        /* Calling decode_mcu() with a NULL pointer causes it to discard the
+         * decoded coefficients.  This is ~5% faster for large subsets, but
+         * it's tough to tell a difference for smaller images.
+         */
+        (*cinfo->entropy->decode_mcu) (cinfo, NULL);
+      }
+    }
+    cinfo->input_iMCU_row++;
+    cinfo->output_iMCU_row++;
+    if (cinfo->input_iMCU_row < cinfo->total_iMCU_rows)
+      start_iMCU_row(cinfo);
+    else
+      (*cinfo->inputctl->finish_input_pass) (cinfo);
+  }
+  cinfo->output_scanline += lines_to_skip;
+
+  if (cinfo->upsample->need_context_rows) {
+    /* Context-based upsampling keeps track of iMCU rows. */
+    main_ptr->iMCU_row_ctr += lines_to_skip / lines_per_iMCU_row;
+
+    /* It is complex to properly move to the middle of a context block, so
+     * read the remaining lines instead of skipping them.
+     */
+    read_and_discard_scanlines(cinfo, lines_to_read);
+  } else {
+    increment_simple_rowgroup_ctr(cinfo, lines_to_read);
+  }
+
+  /* Since skipping lines involves skipping the upsampling step, the value of
+   * "rows_to_go" will become invalid unless we set it here.  NOTE: This is a
+   * bit odd, since "rows_to_go" seems to be redundantly keeping track of
+   * output_scanline.
+   */
+  upsample->rows_to_go = cinfo->output_height - cinfo->output_scanline;
+
+  /* Always skip the requested number of lines. */
+  return num_lines;
+}
+
+/*
+ * Alternate entry point to read raw data.
+ * Processes exactly one iMCU row per call, unless suspended.
+ */
+
+GLOBAL(JDIMENSION)
+jpeg_read_raw_data(j_decompress_ptr cinfo, JSAMPIMAGE data,
+                   JDIMENSION max_lines)
+{
+  JDIMENSION lines_per_iMCU_row;
+
+  if (cinfo->global_state != DSTATE_RAW_OK)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  if (cinfo->output_scanline >= cinfo->output_height) {
+    WARNMS(cinfo, JWRN_TOO_MUCH_DATA);
+    return 0;
+  }
+
+  /* Call progress monitor hook if present */
+  if (cinfo->progress != NULL) {
+    cinfo->progress->pass_counter = (long)cinfo->output_scanline;
+    cinfo->progress->pass_limit = (long)cinfo->output_height;
+    (*cinfo->progress->progress_monitor) ((j_common_ptr)cinfo);
+  }
+
+  /* Verify that at least one iMCU row can be returned. */
+  lines_per_iMCU_row = cinfo->max_v_samp_factor * cinfo->_min_DCT_scaled_size;
+  if (max_lines < lines_per_iMCU_row)
+    ERREXIT(cinfo, JERR_BUFFER_SIZE);
+
+  /* Decompress directly into user's buffer. */
+  if (!(*cinfo->coef->decompress_data) (cinfo, data))
+    return 0;                   /* suspension forced, can do nothing more */
+
+  /* OK, we processed one iMCU row. */
+  cinfo->output_scanline += lines_per_iMCU_row;
+  return lines_per_iMCU_row;
+}
+
+
+/* Additional entry points for buffered-image mode. */
+
+#ifdef D_MULTISCAN_FILES_SUPPORTED
+
+/*
+ * Initialize for an output pass in buffered-image mode.
+ */
+
+GLOBAL(boolean)
+jpeg_start_output(j_decompress_ptr cinfo, int scan_number)
+{
+  if (cinfo->global_state != DSTATE_BUFIMAGE &&
+      cinfo->global_state != DSTATE_PRESCAN)
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  /* Limit scan number to valid range */
+  if (scan_number <= 0)
+    scan_number = 1;
+  if (cinfo->inputctl->eoi_reached && scan_number > cinfo->input_scan_number)
+    scan_number = cinfo->input_scan_number;
+  cinfo->output_scan_number = scan_number;
+  /* Perform any dummy output passes, and set up for the real pass */
+  return output_pass_setup(cinfo);
+}
+
+
+/*
+ * Finish up after an output pass in buffered-image mode.
+ *
+ * Returns FALSE if suspended.  The return value need be inspected only if
+ * a suspending data source is used.
+ */
+
+GLOBAL(boolean)
+jpeg_finish_output(j_decompress_ptr cinfo)
+{
+  if ((cinfo->global_state == DSTATE_SCANNING ||
+       cinfo->global_state == DSTATE_RAW_OK) && cinfo->buffered_image) {
+    /* Terminate this pass. */
+    /* We do not require the whole pass to have been completed. */
+    (*cinfo->master->finish_output_pass) (cinfo);
+    cinfo->global_state = DSTATE_BUFPOST;
+  } else if (cinfo->global_state != DSTATE_BUFPOST) {
+    /* BUFPOST = repeat call after a suspension, anything else is error */
+    ERREXIT1(cinfo, JERR_BAD_STATE, cinfo->global_state);
+  }
+  /* Read markers looking for SOS or EOI */
+  while (cinfo->input_scan_number <= cinfo->output_scan_number &&
+         !cinfo->inputctl->eoi_reached) {
+    if ((*cinfo->inputctl->consume_input) (cinfo) == JPEG_SUSPENDED)
+      return FALSE;             /* Suspend, come back later */
+  }
+  cinfo->global_state = DSTATE_BUFIMAGE;
+  return TRUE;
+}
+
+#endif /* D_MULTISCAN_FILES_SUPPORTED */

3rdparty/libjpeg-turbo/src/jdarith.c

@@ -0,0 +1,773 @@
+/*
+ * jdarith.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Developed 1997-2015 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2015-2018, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains portable arithmetic entropy encoding routines for JPEG
+ * (implementing Recommendation ITU-T T.81 | ISO/IEC 10918-1).
+ *
+ * Both sequential and progressive modes are supported in this single module.
+ *
+ * Suspension is not currently supported in this module.
+ *
+ * NOTE: All referenced figures are from
+ * Recommendation ITU-T T.81 (1992) | ISO/IEC 10918-1:1994.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+
+#define NEG_1  ((unsigned int)-1)
+
+
+/* Expanded entropy decoder object for arithmetic decoding. */
+
+typedef struct {
+  struct jpeg_entropy_decoder pub; /* public fields */
+
+  JLONG c;       /* C register, base of coding interval + input bit buffer */
+  JLONG a;               /* A register, normalized size of coding interval */
+  int ct;     /* bit shift counter, # of bits left in bit buffer part of C */
+                                                         /* init: ct = -16 */
+                                                         /* run: ct = 0..7 */
+                                                         /* error: ct = -1 */
+  int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */
+  int dc_context[MAX_COMPS_IN_SCAN]; /* context index for DC conditioning */
+
+  unsigned int restarts_to_go;  /* MCUs left in this restart interval */
+
+  /* Pointers to statistics areas (these workspaces have image lifespan) */
+  unsigned char *dc_stats[NUM_ARITH_TBLS];
+  unsigned char *ac_stats[NUM_ARITH_TBLS];
+
+  /* Statistics bin for coding with fixed probability 0.5 */
+  unsigned char fixed_bin[4];
+} arith_entropy_decoder;
+
+typedef arith_entropy_decoder *arith_entropy_ptr;
+
+/* The following two definitions specify the allocation chunk size
+ * for the statistics area.
+ * According to sections F.1.4.4.1.3 and F.1.4.4.2, we need at least
+ * 49 statistics bins for DC, and 245 statistics bins for AC coding.
+ *
+ * We use a compact representation with 1 byte per statistics bin,
+ * thus the numbers directly represent byte sizes.
+ * This 1 byte per statistics bin contains the meaning of the MPS
+ * (more probable symbol) in the highest bit (mask 0x80), and the
+ * index into the probability estimation state machine table
+ * in the lower bits (mask 0x7F).
+ */
+
+#define DC_STAT_BINS  64
+#define AC_STAT_BINS  256
+
+
+LOCAL(int)
+get_byte(j_decompress_ptr cinfo)
+/* Read next input byte; we do not support suspension in this module. */
+{
+  struct jpeg_source_mgr *src = cinfo->src;
+
+  if (src->bytes_in_buffer == 0)
+    if (!(*src->fill_input_buffer) (cinfo))
+      ERREXIT(cinfo, JERR_CANT_SUSPEND);
+  src->bytes_in_buffer--;
+  return GETJOCTET(*src->next_input_byte++);
+}
+
+
+/*
+ * The core arithmetic decoding routine (common in JPEG and JBIG).
+ * This needs to go as fast as possible.
+ * Machine-dependent optimization facilities
+ * are not utilized in this portable implementation.
+ * However, this code should be fairly efficient and
+ * may be a good base for further optimizations anyway.
+ *
+ * Return value is 0 or 1 (binary decision).
+ *
+ * Note: I've changed the handling of the code base & bit
+ * buffer register C compared to other implementations
+ * based on the standards layout & procedures.
+ * While it also contains both the actual base of the
+ * coding interval (16 bits) and the next-bits buffer,
+ * the cut-point between these two parts is floating
+ * (instead of fixed) with the bit shift counter CT.
+ * Thus, we also need only one (variable instead of
+ * fixed size) shift for the LPS/MPS decision, and
+ * we can do away with any renormalization update
+ * of C (except for new data insertion, of course).
+ *
+ * I've also introduced a new scheme for accessing
+ * the probability estimation state machine table,
+ * derived from Markus Kuhn's JBIG implementation.
+ */
+
+LOCAL(int)
+arith_decode(j_decompress_ptr cinfo, unsigned char *st)
+{
+  register arith_entropy_ptr e = (arith_entropy_ptr)cinfo->entropy;
+  register unsigned char nl, nm;
+  register JLONG qe, temp;
+  register int sv, data;
+
+  /* Renormalization & data input per section D.2.6 */
+  while (e->a < 0x8000L) {
+    if (--e->ct < 0) {
+      /* Need to fetch next data byte */
+      if (cinfo->unread_marker)
+        data = 0;               /* stuff zero data */
+      else {
+        data = get_byte(cinfo); /* read next input byte */
+        if (data == 0xFF) {     /* zero stuff or marker code */
+          do data = get_byte(cinfo);
+          while (data == 0xFF); /* swallow extra 0xFF bytes */
+          if (data == 0)
+            data = 0xFF;        /* discard stuffed zero byte */
+          else {
+            /* Note: Different from the Huffman decoder, hitting
+             * a marker while processing the compressed data
+             * segment is legal in arithmetic coding.
+             * The convention is to supply zero data
+             * then until decoding is complete.
+             */
+            cinfo->unread_marker = data;
+            data = 0;
+          }
+        }
+      }
+      e->c = (e->c << 8) | data; /* insert data into C register */
+      if ((e->ct += 8) < 0)      /* update bit shift counter */
+        /* Need more initial bytes */
+        if (++e->ct == 0)
+          /* Got 2 initial bytes -> re-init A and exit loop */
+          e->a = 0x8000L; /* => e->a = 0x10000L after loop exit */
+    }
+    e->a <<= 1;
+  }
+
+  /* Fetch values from our compact representation of Table D.2:
+   * Qe values and probability estimation state machine
+   */
+  sv = *st;
+  qe = jpeg_aritab[sv & 0x7F];  /* => Qe_Value */
+  nl = qe & 0xFF;  qe >>= 8;    /* Next_Index_LPS + Switch_MPS */
+  nm = qe & 0xFF;  qe >>= 8;    /* Next_Index_MPS */
+
+  /* Decode & estimation procedures per sections D.2.4 & D.2.5 */
+  temp = e->a - qe;
+  e->a = temp;
+  temp <<= e->ct;
+  if (e->c >= temp) {
+    e->c -= temp;
+    /* Conditional LPS (less probable symbol) exchange */
+    if (e->a < qe) {
+      e->a = qe;
+      *st = (sv & 0x80) ^ nm;   /* Estimate_after_MPS */
+    } else {
+      e->a = qe;
+      *st = (sv & 0x80) ^ nl;   /* Estimate_after_LPS */
+      sv ^= 0x80;               /* Exchange LPS/MPS */
+    }
+  } else if (e->a < 0x8000L) {
+    /* Conditional MPS (more probable symbol) exchange */
+    if (e->a < qe) {
+      *st = (sv & 0x80) ^ nl;   /* Estimate_after_LPS */
+      sv ^= 0x80;               /* Exchange LPS/MPS */
+    } else {
+      *st = (sv & 0x80) ^ nm;   /* Estimate_after_MPS */
+    }
+  }
+
+  return sv >> 7;
+}
+
+
+/*
+ * Check for a restart marker & resynchronize decoder.
+ */
+
+LOCAL(void)
+process_restart(j_decompress_ptr cinfo)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  int ci;
+  jpeg_component_info *compptr;
+
+  /* Advance past the RSTn marker */
+  if (!(*cinfo->marker->read_restart_marker) (cinfo))
+    ERREXIT(cinfo, JERR_CANT_SUSPEND);
+
+  /* Re-initialize statistics areas */
+  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+    compptr = cinfo->cur_comp_info[ci];
+    if (!cinfo->progressive_mode || (cinfo->Ss == 0 && cinfo->Ah == 0)) {
+      MEMZERO(entropy->dc_stats[compptr->dc_tbl_no], DC_STAT_BINS);
+      /* Reset DC predictions to 0 */
+      entropy->last_dc_val[ci] = 0;
+      entropy->dc_context[ci] = 0;
+    }
+    if (!cinfo->progressive_mode || cinfo->Ss) {
+      MEMZERO(entropy->ac_stats[compptr->ac_tbl_no], AC_STAT_BINS);
+    }
+  }
+
+  /* Reset arithmetic decoding variables */
+  entropy->c = 0;
+  entropy->a = 0;
+  entropy->ct = -16;    /* force reading 2 initial bytes to fill C */
+
+  /* Reset restart counter */
+  entropy->restarts_to_go = cinfo->restart_interval;
+}
+
+
+/*
+ * Arithmetic MCU decoding.
+ * Each of these routines decodes and returns one MCU's worth of
+ * arithmetic-compressed coefficients.
+ * The coefficients are reordered from zigzag order into natural array order,
+ * but are not dequantized.
+ *
+ * The i'th block of the MCU is stored into the block pointed to by
+ * MCU_data[i].  WE ASSUME THIS AREA IS INITIALLY ZEROED BY THE CALLER.
+ */
+
+/*
+ * MCU decoding for DC initial scan (either spectral selection,
+ * or first pass of successive approximation).
+ */
+
+METHODDEF(boolean)
+decode_mcu_DC_first(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  JBLOCKROW block;
+  unsigned char *st;
+  int blkn, ci, tbl, sign;
+  int v, m;
+
+  /* Process restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0)
+      process_restart(cinfo);
+    entropy->restarts_to_go--;
+  }
+
+  if (entropy->ct == -1) return TRUE;   /* if error do nothing */
+
+  /* Outer loop handles each block in the MCU */
+
+  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
+    block = MCU_data[blkn];
+    ci = cinfo->MCU_membership[blkn];
+    tbl = cinfo->cur_comp_info[ci]->dc_tbl_no;
+
+    /* Sections F.2.4.1 & F.1.4.4.1: Decoding of DC coefficients */
+
+    /* Table F.4: Point to statistics bin S0 for DC coefficient coding */
+    st = entropy->dc_stats[tbl] + entropy->dc_context[ci];
+
+    /* Figure F.19: Decode_DC_DIFF */
+    if (arith_decode(cinfo, st) == 0)
+      entropy->dc_context[ci] = 0;
+    else {
+      /* Figure F.21: Decoding nonzero value v */
+      /* Figure F.22: Decoding the sign of v */
+      sign = arith_decode(cinfo, st + 1);
+      st += 2;  st += sign;
+      /* Figure F.23: Decoding the magnitude category of v */
+      if ((m = arith_decode(cinfo, st)) != 0) {
+        st = entropy->dc_stats[tbl] + 20;       /* Table F.4: X1 = 20 */
+        while (arith_decode(cinfo, st)) {
+          if ((m <<= 1) == 0x8000) {
+            WARNMS(cinfo, JWRN_ARITH_BAD_CODE);
+            entropy->ct = -1;                   /* magnitude overflow */
+            return TRUE;
+          }
+          st += 1;
+        }
+      }
+      /* Section F.1.4.4.1.2: Establish dc_context conditioning category */
+      if (m < (int)((1L << cinfo->arith_dc_L[tbl]) >> 1))
+        entropy->dc_context[ci] = 0;               /* zero diff category */
+      else if (m > (int)((1L << cinfo->arith_dc_U[tbl]) >> 1))
+        entropy->dc_context[ci] = 12 + (sign * 4); /* large diff category */
+      else
+        entropy->dc_context[ci] = 4 + (sign * 4);  /* small diff category */
+      v = m;
+      /* Figure F.24: Decoding the magnitude bit pattern of v */
+      st += 14;
+      while (m >>= 1)
+        if (arith_decode(cinfo, st)) v |= m;
+      v += 1;  if (sign) v = -v;
+      entropy->last_dc_val[ci] = (entropy->last_dc_val[ci] + v) & 0xffff;
+    }
+
+    /* Scale and output the DC coefficient (assumes jpeg_natural_order[0]=0) */
+    (*block)[0] = (JCOEF)LEFT_SHIFT(entropy->last_dc_val[ci], cinfo->Al);
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * MCU decoding for AC initial scan (either spectral selection,
+ * or first pass of successive approximation).
+ */
+
+METHODDEF(boolean)
+decode_mcu_AC_first(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  JBLOCKROW block;
+  unsigned char *st;
+  int tbl, sign, k;
+  int v, m;
+
+  /* Process restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0)
+      process_restart(cinfo);
+    entropy->restarts_to_go--;
+  }
+
+  if (entropy->ct == -1) return TRUE;   /* if error do nothing */
+
+  /* There is always only one block per MCU */
+  block = MCU_data[0];
+  tbl = cinfo->cur_comp_info[0]->ac_tbl_no;
+
+  /* Sections F.2.4.2 & F.1.4.4.2: Decoding of AC coefficients */
+
+  /* Figure F.20: Decode_AC_coefficients */
+  for (k = cinfo->Ss; k <= cinfo->Se; k++) {
+    st = entropy->ac_stats[tbl] + 3 * (k - 1);
+    if (arith_decode(cinfo, st)) break;         /* EOB flag */
+    while (arith_decode(cinfo, st + 1) == 0) {
+      st += 3;  k++;
+      if (k > cinfo->Se) {
+        WARNMS(cinfo, JWRN_ARITH_BAD_CODE);
+        entropy->ct = -1;                       /* spectral overflow */
+        return TRUE;
+      }
+    }
+    /* Figure F.21: Decoding nonzero value v */
+    /* Figure F.22: Decoding the sign of v */
+    sign = arith_decode(cinfo, entropy->fixed_bin);
+    st += 2;
+    /* Figure F.23: Decoding the magnitude category of v */
+    if ((m = arith_decode(cinfo, st)) != 0) {
+      if (arith_decode(cinfo, st)) {
+        m <<= 1;
+        st = entropy->ac_stats[tbl] +
+             (k <= cinfo->arith_ac_K[tbl] ? 189 : 217);
+        while (arith_decode(cinfo, st)) {
+          if ((m <<= 1) == 0x8000) {
+            WARNMS(cinfo, JWRN_ARITH_BAD_CODE);
+            entropy->ct = -1;                   /* magnitude overflow */
+            return TRUE;
+          }
+          st += 1;
+        }
+      }
+    }
+    v = m;
+    /* Figure F.24: Decoding the magnitude bit pattern of v */
+    st += 14;
+    while (m >>= 1)
+      if (arith_decode(cinfo, st)) v |= m;
+    v += 1;  if (sign) v = -v;
+    /* Scale and output coefficient in natural (dezigzagged) order */
+    (*block)[jpeg_natural_order[k]] = (JCOEF)((unsigned)v << cinfo->Al);
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * MCU decoding for DC successive approximation refinement scan.
+ */
+
+METHODDEF(boolean)
+decode_mcu_DC_refine(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  unsigned char *st;
+  int p1, blkn;
+
+  /* Process restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0)
+      process_restart(cinfo);
+    entropy->restarts_to_go--;
+  }
+
+  st = entropy->fixed_bin;      /* use fixed probability estimation */
+  p1 = 1 << cinfo->Al;          /* 1 in the bit position being coded */
+
+  /* Outer loop handles each block in the MCU */
+
+  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
+    /* Encoded data is simply the next bit of the two's-complement DC value */
+    if (arith_decode(cinfo, st))
+      MCU_data[blkn][0][0] |= p1;
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * MCU decoding for AC successive approximation refinement scan.
+ */
+
+METHODDEF(boolean)
+decode_mcu_AC_refine(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  JBLOCKROW block;
+  JCOEFPTR thiscoef;
+  unsigned char *st;
+  int tbl, k, kex;
+  int p1, m1;
+
+  /* Process restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0)
+      process_restart(cinfo);
+    entropy->restarts_to_go--;
+  }
+
+  if (entropy->ct == -1) return TRUE;   /* if error do nothing */
+
+  /* There is always only one block per MCU */
+  block = MCU_data[0];
+  tbl = cinfo->cur_comp_info[0]->ac_tbl_no;
+
+  p1 = 1 << cinfo->Al;          /* 1 in the bit position being coded */
+  m1 = (NEG_1) << cinfo->Al;    /* -1 in the bit position being coded */
+
+  /* Establish EOBx (previous stage end-of-block) index */
+  for (kex = cinfo->Se; kex > 0; kex--)
+    if ((*block)[jpeg_natural_order[kex]]) break;
+
+  for (k = cinfo->Ss; k <= cinfo->Se; k++) {
+    st = entropy->ac_stats[tbl] + 3 * (k - 1);
+    if (k > kex)
+      if (arith_decode(cinfo, st)) break;       /* EOB flag */
+    for (;;) {
+      thiscoef = *block + jpeg_natural_order[k];
+      if (*thiscoef) {                          /* previously nonzero coef */
+        if (arith_decode(cinfo, st + 2)) {
+          if (*thiscoef < 0)
+            *thiscoef += m1;
+          else
+            *thiscoef += p1;
+        }
+        break;
+      }
+      if (arith_decode(cinfo, st + 1)) {        /* newly nonzero coef */
+        if (arith_decode(cinfo, entropy->fixed_bin))
+          *thiscoef = m1;
+        else
+          *thiscoef = p1;
+        break;
+      }
+      st += 3;  k++;
+      if (k > cinfo->Se) {
+        WARNMS(cinfo, JWRN_ARITH_BAD_CODE);
+        entropy->ct = -1;                       /* spectral overflow */
+        return TRUE;
+      }
+    }
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * Decode one MCU's worth of arithmetic-compressed coefficients.
+ */
+
+METHODDEF(boolean)
+decode_mcu(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  jpeg_component_info *compptr;
+  JBLOCKROW block;
+  unsigned char *st;
+  int blkn, ci, tbl, sign, k;
+  int v, m;
+
+  /* Process restart marker if needed */
+  if (cinfo->restart_interval) {
+    if (entropy->restarts_to_go == 0)
+      process_restart(cinfo);
+    entropy->restarts_to_go--;
+  }
+
+  if (entropy->ct == -1) return TRUE;   /* if error do nothing */
+
+  /* Outer loop handles each block in the MCU */
+
+  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
+    block = MCU_data ? MCU_data[blkn] : NULL;
+    ci = cinfo->MCU_membership[blkn];
+    compptr = cinfo->cur_comp_info[ci];
+
+    /* Sections F.2.4.1 & F.1.4.4.1: Decoding of DC coefficients */
+
+    tbl = compptr->dc_tbl_no;
+
+    /* Table F.4: Point to statistics bin S0 for DC coefficient coding */
+    st = entropy->dc_stats[tbl] + entropy->dc_context[ci];
+
+    /* Figure F.19: Decode_DC_DIFF */
+    if (arith_decode(cinfo, st) == 0)
+      entropy->dc_context[ci] = 0;
+    else {
+      /* Figure F.21: Decoding nonzero value v */
+      /* Figure F.22: Decoding the sign of v */
+      sign = arith_decode(cinfo, st + 1);
+      st += 2;  st += sign;
+      /* Figure F.23: Decoding the magnitude category of v */
+      if ((m = arith_decode(cinfo, st)) != 0) {
+        st = entropy->dc_stats[tbl] + 20;       /* Table F.4: X1 = 20 */
+        while (arith_decode(cinfo, st)) {
+          if ((m <<= 1) == 0x8000) {
+            WARNMS(cinfo, JWRN_ARITH_BAD_CODE);
+            entropy->ct = -1;                   /* magnitude overflow */
+            return TRUE;
+          }
+          st += 1;
+        }
+      }
+      /* Section F.1.4.4.1.2: Establish dc_context conditioning category */
+      if (m < (int)((1L << cinfo->arith_dc_L[tbl]) >> 1))
+        entropy->dc_context[ci] = 0;               /* zero diff category */
+      else if (m > (int)((1L << cinfo->arith_dc_U[tbl]) >> 1))
+        entropy->dc_context[ci] = 12 + (sign * 4); /* large diff category */
+      else
+        entropy->dc_context[ci] = 4 + (sign * 4);  /* small diff category */
+      v = m;
+      /* Figure F.24: Decoding the magnitude bit pattern of v */
+      st += 14;
+      while (m >>= 1)
+        if (arith_decode(cinfo, st)) v |= m;
+      v += 1;  if (sign) v = -v;
+      entropy->last_dc_val[ci] = (entropy->last_dc_val[ci] + v) & 0xffff;
+    }
+
+    if (block)
+      (*block)[0] = (JCOEF)entropy->last_dc_val[ci];
+
+    /* Sections F.2.4.2 & F.1.4.4.2: Decoding of AC coefficients */
+
+    tbl = compptr->ac_tbl_no;
+
+    /* Figure F.20: Decode_AC_coefficients */
+    for (k = 1; k <= DCTSIZE2 - 1; k++) {
+      st = entropy->ac_stats[tbl] + 3 * (k - 1);
+      if (arith_decode(cinfo, st)) break;       /* EOB flag */
+      while (arith_decode(cinfo, st + 1) == 0) {
+        st += 3;  k++;
+        if (k > DCTSIZE2 - 1) {
+          WARNMS(cinfo, JWRN_ARITH_BAD_CODE);
+          entropy->ct = -1;                     /* spectral overflow */
+          return TRUE;
+        }
+      }
+      /* Figure F.21: Decoding nonzero value v */
+      /* Figure F.22: Decoding the sign of v */
+      sign = arith_decode(cinfo, entropy->fixed_bin);
+      st += 2;
+      /* Figure F.23: Decoding the magnitude category of v */
+      if ((m = arith_decode(cinfo, st)) != 0) {
+        if (arith_decode(cinfo, st)) {
+          m <<= 1;
+          st = entropy->ac_stats[tbl] +
+               (k <= cinfo->arith_ac_K[tbl] ? 189 : 217);
+          while (arith_decode(cinfo, st)) {
+            if ((m <<= 1) == 0x8000) {
+              WARNMS(cinfo, JWRN_ARITH_BAD_CODE);
+              entropy->ct = -1;                 /* magnitude overflow */
+              return TRUE;
+            }
+            st += 1;
+          }
+        }
+      }
+      v = m;
+      /* Figure F.24: Decoding the magnitude bit pattern of v */
+      st += 14;
+      while (m >>= 1)
+        if (arith_decode(cinfo, st)) v |= m;
+      v += 1;  if (sign) v = -v;
+      if (block)
+        (*block)[jpeg_natural_order[k]] = (JCOEF)v;
+    }
+  }
+
+  return TRUE;
+}
+
+
+/*
+ * Initialize for an arithmetic-compressed scan.
+ */
+
+METHODDEF(void)
+start_pass(j_decompress_ptr cinfo)
+{
+  arith_entropy_ptr entropy = (arith_entropy_ptr)cinfo->entropy;
+  int ci, tbl;
+  jpeg_component_info *compptr;
+
+  if (cinfo->progressive_mode) {
+    /* Validate progressive scan parameters */
+    if (cinfo->Ss == 0) {
+      if (cinfo->Se != 0)
+        goto bad;
+    } else {
+      /* need not check Ss/Se < 0 since they came from unsigned bytes */
+      if (cinfo->Se < cinfo->Ss || cinfo->Se > DCTSIZE2 - 1)
+        goto bad;
+      /* AC scans may have only one component */
+      if (cinfo->comps_in_scan != 1)
+        goto bad;
+    }
+    if (cinfo->Ah != 0) {
+      /* Successive approximation refinement scan: must have Al = Ah-1. */
+      if (cinfo->Ah - 1 != cinfo->Al)
+        goto bad;
+    }
+    if (cinfo->Al > 13) {       /* need not check for < 0 */
+bad:
+      ERREXIT4(cinfo, JERR_BAD_PROGRESSION,
+               cinfo->Ss, cinfo->Se, cinfo->Ah, cinfo->Al);
+    }
+    /* Update progression status, and verify that scan order is legal.
+     * Note that inter-scan inconsistencies are treated as warnings
+     * not fatal errors ... not clear if this is right way to behave.
+     */
+    for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+      int coefi, cindex = cinfo->cur_comp_info[ci]->component_index;
+      int *coef_bit_ptr = &cinfo->coef_bits[cindex][0];
+      if (cinfo->Ss && coef_bit_ptr[0] < 0) /* AC without prior DC scan */
+        WARNMS2(cinfo, JWRN_BOGUS_PROGRESSION, cindex, 0);
+      for (coefi = cinfo->Ss; coefi <= cinfo->Se; coefi++) {
+        int expected = (coef_bit_ptr[coefi] < 0) ? 0 : coef_bit_ptr[coefi];
+        if (cinfo->Ah != expected)
+          WARNMS2(cinfo, JWRN_BOGUS_PROGRESSION, cindex, coefi);
+        coef_bit_ptr[coefi] = cinfo->Al;
+      }
+    }
+    /* Select MCU decoding routine */
+    if (cinfo->Ah == 0) {
+      if (cinfo->Ss == 0)
+        entropy->pub.decode_mcu = decode_mcu_DC_first;
+      else
+        entropy->pub.decode_mcu = decode_mcu_AC_first;
+    } else {
+      if (cinfo->Ss == 0)
+        entropy->pub.decode_mcu = decode_mcu_DC_refine;
+      else
+        entropy->pub.decode_mcu = decode_mcu_AC_refine;
+    }
+  } else {
+    /* Check that the scan parameters Ss, Se, Ah/Al are OK for sequential JPEG.
+     * This ought to be an error condition, but we make it a warning.
+     */
+    if (cinfo->Ss != 0 || cinfo->Ah != 0 || cinfo->Al != 0 ||
+        (cinfo->Se < DCTSIZE2 && cinfo->Se != DCTSIZE2 - 1))
+      WARNMS(cinfo, JWRN_NOT_SEQUENTIAL);
+    /* Select MCU decoding routine */
+    entropy->pub.decode_mcu = decode_mcu;
+  }
+
+  /* Allocate & initialize requested statistics areas */
+  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+    compptr = cinfo->cur_comp_info[ci];
+    if (!cinfo->progressive_mode || (cinfo->Ss == 0 && cinfo->Ah == 0)) {
+      tbl = compptr->dc_tbl_no;
+      if (tbl < 0 || tbl >= NUM_ARITH_TBLS)
+        ERREXIT1(cinfo, JERR_NO_ARITH_TABLE, tbl);
+      if (entropy->dc_stats[tbl] == NULL)
+        entropy->dc_stats[tbl] = (unsigned char *)(*cinfo->mem->alloc_small)
+          ((j_common_ptr)cinfo, JPOOL_IMAGE, DC_STAT_BINS);
+      MEMZERO(entropy->dc_stats[tbl], DC_STAT_BINS);
+      /* Initialize DC predictions to 0 */
+      entropy->last_dc_val[ci] = 0;
+      entropy->dc_context[ci] = 0;
+    }
+    if (!cinfo->progressive_mode || cinfo->Ss) {
+      tbl = compptr->ac_tbl_no;
+      if (tbl < 0 || tbl >= NUM_ARITH_TBLS)
+        ERREXIT1(cinfo, JERR_NO_ARITH_TABLE, tbl);
+      if (entropy->ac_stats[tbl] == NULL)
+        entropy->ac_stats[tbl] = (unsigned char *)(*cinfo->mem->alloc_small)
+          ((j_common_ptr)cinfo, JPOOL_IMAGE, AC_STAT_BINS);
+      MEMZERO(entropy->ac_stats[tbl], AC_STAT_BINS);
+    }
+  }
+
+  /* Initialize arithmetic decoding variables */
+  entropy->c = 0;
+  entropy->a = 0;
+  entropy->ct = -16;    /* force reading 2 initial bytes to fill C */
+
+  /* Initialize restart counter */
+  entropy->restarts_to_go = cinfo->restart_interval;
+}
+
+
+/*
+ * Module initialization routine for arithmetic entropy decoding.
+ */
+
+GLOBAL(void)
+jinit_arith_decoder(j_decompress_ptr cinfo)
+{
+  arith_entropy_ptr entropy;
+  int i;
+
+  entropy = (arith_entropy_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(arith_entropy_decoder));
+  cinfo->entropy = (struct jpeg_entropy_decoder *)entropy;
+  entropy->pub.start_pass = start_pass;
+
+  /* Mark tables unallocated */
+  for (i = 0; i < NUM_ARITH_TBLS; i++) {
+    entropy->dc_stats[i] = NULL;
+    entropy->ac_stats[i] = NULL;
+  }
+
+  /* Initialize index for fixed probability estimation */
+  entropy->fixed_bin[0] = 113;
+
+  if (cinfo->progressive_mode) {
+    /* Create progression status table */
+    int *coef_bit_ptr, ci;
+    cinfo->coef_bits = (int (*)[DCTSIZE2])
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                  cinfo->num_components * DCTSIZE2 *
+                                  sizeof(int));
+    coef_bit_ptr = &cinfo->coef_bits[0][0];
+    for (ci = 0; ci < cinfo->num_components; ci++)
+      for (i = 0; i < DCTSIZE2; i++)
+        *coef_bit_ptr++ = -1;
+  }
+}

3rdparty/libjpeg-turbo/src/jdatadst.c

@@ -0,0 +1,293 @@
+/*
+ * jdatadst.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * Modified 2009-2012 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2013, 2016, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains compression data destination routines for the case of
+ * emitting JPEG data to memory or to a file (or any stdio stream).
+ * While these routines are sufficient for most applications,
+ * some will want to use a different destination manager.
+ * IMPORTANT: we assume that fwrite() will correctly transcribe an array of
+ * JOCTETs into 8-bit-wide elements on external storage.  If char is wider
+ * than 8 bits on your machine, you may need to do some tweaking.
+ */
+
+/* this is not a core library module, so it doesn't define JPEG_INTERNALS */
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jerror.h"
+
+#ifndef HAVE_STDLIB_H           /* <stdlib.h> should declare malloc(),free() */
+extern void *malloc(size_t size);
+extern void free(void *ptr);
+#endif
+
+
+/* Expanded data destination object for stdio output */
+
+typedef struct {
+  struct jpeg_destination_mgr pub; /* public fields */
+
+  FILE *outfile;                /* target stream */
+  JOCTET *buffer;               /* start of buffer */
+} my_destination_mgr;
+
+typedef my_destination_mgr *my_dest_ptr;
+
+#define OUTPUT_BUF_SIZE  4096   /* choose an efficiently fwrite'able size */
+
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+/* Expanded data destination object for memory output */
+
+typedef struct {
+  struct jpeg_destination_mgr pub; /* public fields */
+
+  unsigned char **outbuffer;    /* target buffer */
+  unsigned long *outsize;
+  unsigned char *newbuffer;     /* newly allocated buffer */
+  JOCTET *buffer;               /* start of buffer */
+  size_t bufsize;
+} my_mem_destination_mgr;
+
+typedef my_mem_destination_mgr *my_mem_dest_ptr;
+#endif
+
+
+/*
+ * Initialize destination --- called by jpeg_start_compress
+ * before any data is actually written.
+ */
+
+METHODDEF(void)
+init_destination(j_compress_ptr cinfo)
+{
+  my_dest_ptr dest = (my_dest_ptr)cinfo->dest;
+
+  /* Allocate the output buffer --- it will be released when done with image */
+  dest->buffer = (JOCTET *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                OUTPUT_BUF_SIZE * sizeof(JOCTET));
+
+  dest->pub.next_output_byte = dest->buffer;
+  dest->pub.free_in_buffer = OUTPUT_BUF_SIZE;
+}
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+METHODDEF(void)
+init_mem_destination(j_compress_ptr cinfo)
+{
+  /* no work necessary here */
+}
+#endif
+
+
+/*
+ * Empty the output buffer --- called whenever buffer fills up.
+ *
+ * In typical applications, this should write the entire output buffer
+ * (ignoring the current state of next_output_byte & free_in_buffer),
+ * reset the pointer & count to the start of the buffer, and return TRUE
+ * indicating that the buffer has been dumped.
+ *
+ * In applications that need to be able to suspend compression due to output
+ * overrun, a FALSE return indicates that the buffer cannot be emptied now.
+ * In this situation, the compressor will return to its caller (possibly with
+ * an indication that it has not accepted all the supplied scanlines).  The
+ * application should resume compression after it has made more room in the
+ * output buffer.  Note that there are substantial restrictions on the use of
+ * suspension --- see the documentation.
+ *
+ * When suspending, the compressor will back up to a convenient restart point
+ * (typically the start of the current MCU). next_output_byte & free_in_buffer
+ * indicate where the restart point will be if the current call returns FALSE.
+ * Data beyond this point will be regenerated after resumption, so do not
+ * write it out when emptying the buffer externally.
+ */
+
+METHODDEF(boolean)
+empty_output_buffer(j_compress_ptr cinfo)
+{
+  my_dest_ptr dest = (my_dest_ptr)cinfo->dest;
+
+  if (JFWRITE(dest->outfile, dest->buffer, OUTPUT_BUF_SIZE) !=
+      (size_t)OUTPUT_BUF_SIZE)
+    ERREXIT(cinfo, JERR_FILE_WRITE);
+
+  dest->pub.next_output_byte = dest->buffer;
+  dest->pub.free_in_buffer = OUTPUT_BUF_SIZE;
+
+  return TRUE;
+}
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+METHODDEF(boolean)
+empty_mem_output_buffer(j_compress_ptr cinfo)
+{
+  size_t nextsize;
+  JOCTET *nextbuffer;
+  my_mem_dest_ptr dest = (my_mem_dest_ptr)cinfo->dest;
+
+  /* Try to allocate new buffer with double size */
+  nextsize = dest->bufsize * 2;
+  nextbuffer = (JOCTET *)malloc(nextsize);
+
+  if (nextbuffer == NULL)
+    ERREXIT1(cinfo, JERR_OUT_OF_MEMORY, 10);
+
+  MEMCOPY(nextbuffer, dest->buffer, dest->bufsize);
+
+  if (dest->newbuffer != NULL)
+    free(dest->newbuffer);
+
+  dest->newbuffer = nextbuffer;
+
+  dest->pub.next_output_byte = nextbuffer + dest->bufsize;
+  dest->pub.free_in_buffer = dest->bufsize;
+
+  dest->buffer = nextbuffer;
+  dest->bufsize = nextsize;
+
+  return TRUE;
+}
+#endif
+
+
+/*
+ * Terminate destination --- called by jpeg_finish_compress
+ * after all data has been written.  Usually needs to flush buffer.
+ *
+ * NB: *not* called by jpeg_abort or jpeg_destroy; surrounding
+ * application must deal with any cleanup that should happen even
+ * for error exit.
+ */
+
+METHODDEF(void)
+term_destination(j_compress_ptr cinfo)
+{
+  my_dest_ptr dest = (my_dest_ptr)cinfo->dest;
+  size_t datacount = OUTPUT_BUF_SIZE - dest->pub.free_in_buffer;
+
+  /* Write any data remaining in the buffer */
+  if (datacount > 0) {
+    if (JFWRITE(dest->outfile, dest->buffer, datacount) != datacount)
+      ERREXIT(cinfo, JERR_FILE_WRITE);
+  }
+  fflush(dest->outfile);
+  /* Make sure we wrote the output file OK */
+  if (ferror(dest->outfile))
+    ERREXIT(cinfo, JERR_FILE_WRITE);
+}
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+METHODDEF(void)
+term_mem_destination(j_compress_ptr cinfo)
+{
+  my_mem_dest_ptr dest = (my_mem_dest_ptr)cinfo->dest;
+
+  *dest->outbuffer = dest->buffer;
+  *dest->outsize = (unsigned long)(dest->bufsize - dest->pub.free_in_buffer);
+}
+#endif
+
+
+/*
+ * Prepare for output to a stdio stream.
+ * The caller must have already opened the stream, and is responsible
+ * for closing it after finishing compression.
+ */
+
+GLOBAL(void)
+jpeg_stdio_dest(j_compress_ptr cinfo, FILE *outfile)
+{
+  my_dest_ptr dest;
+
+  /* The destination object is made permanent so that multiple JPEG images
+   * can be written to the same file without re-executing jpeg_stdio_dest.
+   */
+  if (cinfo->dest == NULL) {    /* first time for this JPEG object? */
+    cinfo->dest = (struct jpeg_destination_mgr *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                                  sizeof(my_destination_mgr));
+  } else if (cinfo->dest->init_destination != init_destination) {
+    /* It is unsafe to reuse the existing destination manager unless it was
+     * created by this function.  Otherwise, there is no guarantee that the
+     * opaque structure is the right size.  Note that we could just create a
+     * new structure, but the old structure would not be freed until
+     * jpeg_destroy_compress() was called.
+     */
+    ERREXIT(cinfo, JERR_BUFFER_SIZE);
+  }
+
+  dest = (my_dest_ptr)cinfo->dest;
+  dest->pub.init_destination = init_destination;
+  dest->pub.empty_output_buffer = empty_output_buffer;
+  dest->pub.term_destination = term_destination;
+  dest->outfile = outfile;
+}
+
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+/*
+ * Prepare for output to a memory buffer.
+ * The caller may supply an own initial buffer with appropriate size.
+ * Otherwise, or when the actual data output exceeds the given size,
+ * the library adapts the buffer size as necessary.
+ * The standard library functions malloc/free are used for allocating
+ * larger memory, so the buffer is available to the application after
+ * finishing compression, and then the application is responsible for
+ * freeing the requested memory.
+ * Note:  An initial buffer supplied by the caller is expected to be
+ * managed by the application.  The library does not free such buffer
+ * when allocating a larger buffer.
+ */
+
+GLOBAL(void)
+jpeg_mem_dest(j_compress_ptr cinfo, unsigned char **outbuffer,
+              unsigned long *outsize)
+{
+  my_mem_dest_ptr dest;
+
+  if (outbuffer == NULL || outsize == NULL)     /* sanity check */
+    ERREXIT(cinfo, JERR_BUFFER_SIZE);
+
+  /* The destination object is made permanent so that multiple JPEG images
+   * can be written to the same buffer without re-executing jpeg_mem_dest.
+   */
+  if (cinfo->dest == NULL) {    /* first time for this JPEG object? */
+    cinfo->dest = (struct jpeg_destination_mgr *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                                  sizeof(my_mem_destination_mgr));
+  } else if (cinfo->dest->init_destination != init_mem_destination) {
+    /* It is unsafe to reuse the existing destination manager unless it was
+     * created by this function.
+     */
+    ERREXIT(cinfo, JERR_BUFFER_SIZE);
+  }
+
+  dest = (my_mem_dest_ptr)cinfo->dest;
+  dest->pub.init_destination = init_mem_destination;
+  dest->pub.empty_output_buffer = empty_mem_output_buffer;
+  dest->pub.term_destination = term_mem_destination;
+  dest->outbuffer = outbuffer;
+  dest->outsize = outsize;
+  dest->newbuffer = NULL;
+
+  if (*outbuffer == NULL || *outsize == 0) {
+    /* Allocate initial buffer */
+    dest->newbuffer = *outbuffer = (unsigned char *)malloc(OUTPUT_BUF_SIZE);
+    if (dest->newbuffer == NULL)
+      ERREXIT1(cinfo, JERR_OUT_OF_MEMORY, 10);
+    *outsize = OUTPUT_BUF_SIZE;
+  }
+
+  dest->pub.next_output_byte = dest->buffer = *outbuffer;
+  dest->pub.free_in_buffer = dest->bufsize = *outsize;
+}
+#endif

3rdparty/libjpeg-turbo/src/jdatasrc.c

@@ -0,0 +1,295 @@
+/*
+ * jdatasrc.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * Modified 2009-2011 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2013, 2016, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains decompression data source routines for the case of
+ * reading JPEG data from memory or from a file (or any stdio stream).
+ * While these routines are sufficient for most applications,
+ * some will want to use a different source manager.
+ * IMPORTANT: we assume that fread() will correctly transcribe an array of
+ * JOCTETs from 8-bit-wide elements on external storage.  If char is wider
+ * than 8 bits on your machine, you may need to do some tweaking.
+ */
+
+/* this is not a core library module, so it doesn't define JPEG_INTERNALS */
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jerror.h"
+
+
+/* Expanded data source object for stdio input */
+
+typedef struct {
+  struct jpeg_source_mgr pub;   /* public fields */
+
+  FILE *infile;                 /* source stream */
+  JOCTET *buffer;               /* start of buffer */
+  boolean start_of_file;        /* have we gotten any data yet? */
+} my_source_mgr;
+
+typedef my_source_mgr *my_src_ptr;
+
+#define INPUT_BUF_SIZE  4096    /* choose an efficiently fread'able size */
+
+
+/*
+ * Initialize source --- called by jpeg_read_header
+ * before any data is actually read.
+ */
+
+METHODDEF(void)
+init_source(j_decompress_ptr cinfo)
+{
+  my_src_ptr src = (my_src_ptr)cinfo->src;
+
+  /* We reset the empty-input-file flag for each image,
+   * but we don't clear the input buffer.
+   * This is correct behavior for reading a series of images from one source.
+   */
+  src->start_of_file = TRUE;
+}
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+METHODDEF(void)
+init_mem_source(j_decompress_ptr cinfo)
+{
+  /* no work necessary here */
+}
+#endif
+
+
+/*
+ * Fill the input buffer --- called whenever buffer is emptied.
+ *
+ * In typical applications, this should read fresh data into the buffer
+ * (ignoring the current state of next_input_byte & bytes_in_buffer),
+ * reset the pointer & count to the start of the buffer, and return TRUE
+ * indicating that the buffer has been reloaded.  It is not necessary to
+ * fill the buffer entirely, only to obtain at least one more byte.
+ *
+ * There is no such thing as an EOF return.  If the end of the file has been
+ * reached, the routine has a choice of ERREXIT() or inserting fake data into
+ * the buffer.  In most cases, generating a warning message and inserting a
+ * fake EOI marker is the best course of action --- this will allow the
+ * decompressor to output however much of the image is there.  However,
+ * the resulting error message is misleading if the real problem is an empty
+ * input file, so we handle that case specially.
+ *
+ * In applications that need to be able to suspend compression due to input
+ * not being available yet, a FALSE return indicates that no more data can be
+ * obtained right now, but more may be forthcoming later.  In this situation,
+ * the decompressor will return to its caller (with an indication of the
+ * number of scanlines it has read, if any).  The application should resume
+ * decompression after it has loaded more data into the input buffer.  Note
+ * that there are substantial restrictions on the use of suspension --- see
+ * the documentation.
+ *
+ * When suspending, the decompressor will back up to a convenient restart point
+ * (typically the start of the current MCU). next_input_byte & bytes_in_buffer
+ * indicate where the restart point will be if the current call returns FALSE.
+ * Data beyond this point must be rescanned after resumption, so move it to
+ * the front of the buffer rather than discarding it.
+ */
+
+METHODDEF(boolean)
+fill_input_buffer(j_decompress_ptr cinfo)
+{
+  my_src_ptr src = (my_src_ptr)cinfo->src;
+  size_t nbytes;
+
+  nbytes = JFREAD(src->infile, src->buffer, INPUT_BUF_SIZE);
+
+  if (nbytes <= 0) {
+    if (src->start_of_file)     /* Treat empty input file as fatal error */
+      ERREXIT(cinfo, JERR_INPUT_EMPTY);
+    WARNMS(cinfo, JWRN_JPEG_EOF);
+    /* Insert a fake EOI marker */
+    src->buffer[0] = (JOCTET)0xFF;
+    src->buffer[1] = (JOCTET)JPEG_EOI;
+    nbytes = 2;
+  }
+
+  src->pub.next_input_byte = src->buffer;
+  src->pub.bytes_in_buffer = nbytes;
+  src->start_of_file = FALSE;
+
+  return TRUE;
+}
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+METHODDEF(boolean)
+fill_mem_input_buffer(j_decompress_ptr cinfo)
+{
+  static const JOCTET mybuffer[4] = {
+    (JOCTET)0xFF, (JOCTET)JPEG_EOI, 0, 0
+  };
+
+  /* The whole JPEG data is expected to reside in the supplied memory
+   * buffer, so any request for more data beyond the given buffer size
+   * is treated as an error.
+   */
+  WARNMS(cinfo, JWRN_JPEG_EOF);
+
+  /* Insert a fake EOI marker */
+
+  cinfo->src->next_input_byte = mybuffer;
+  cinfo->src->bytes_in_buffer = 2;
+
+  return TRUE;
+}
+#endif
+
+
+/*
+ * Skip data --- used to skip over a potentially large amount of
+ * uninteresting data (such as an APPn marker).
+ *
+ * Writers of suspendable-input applications must note that skip_input_data
+ * is not granted the right to give a suspension return.  If the skip extends
+ * beyond the data currently in the buffer, the buffer can be marked empty so
+ * that the next read will cause a fill_input_buffer call that can suspend.
+ * Arranging for additional bytes to be discarded before reloading the input
+ * buffer is the application writer's problem.
+ */
+
+METHODDEF(void)
+skip_input_data(j_decompress_ptr cinfo, long num_bytes)
+{
+  struct jpeg_source_mgr *src = cinfo->src;
+
+  /* Just a dumb implementation for now.  Could use fseek() except
+   * it doesn't work on pipes.  Not clear that being smart is worth
+   * any trouble anyway --- large skips are infrequent.
+   */
+  if (num_bytes > 0) {
+    while (num_bytes > (long)src->bytes_in_buffer) {
+      num_bytes -= (long)src->bytes_in_buffer;
+      (void)(*src->fill_input_buffer) (cinfo);
+      /* note we assume that fill_input_buffer will never return FALSE,
+       * so suspension need not be handled.
+       */
+    }
+    src->next_input_byte += (size_t)num_bytes;
+    src->bytes_in_buffer -= (size_t)num_bytes;
+  }
+}
+
+
+/*
+ * An additional method that can be provided by data source modules is the
+ * resync_to_restart method for error recovery in the presence of RST markers.
+ * For the moment, this source module just uses the default resync method
+ * provided by the JPEG library.  That method assumes that no backtracking
+ * is possible.
+ */
+
+
+/*
+ * Terminate source --- called by jpeg_finish_decompress
+ * after all data has been read.  Often a no-op.
+ *
+ * NB: *not* called by jpeg_abort or jpeg_destroy; surrounding
+ * application must deal with any cleanup that should happen even
+ * for error exit.
+ */
+
+METHODDEF(void)
+term_source(j_decompress_ptr cinfo)
+{
+  /* no work necessary here */
+}
+
+
+/*
+ * Prepare for input from a stdio stream.
+ * The caller must have already opened the stream, and is responsible
+ * for closing it after finishing decompression.
+ */
+
+GLOBAL(void)
+jpeg_stdio_src(j_decompress_ptr cinfo, FILE *infile)
+{
+  my_src_ptr src;
+
+  /* The source object and input buffer are made permanent so that a series
+   * of JPEG images can be read from the same file by calling jpeg_stdio_src
+   * only before the first one.  (If we discarded the buffer at the end of
+   * one image, we'd likely lose the start of the next one.)
+   */
+  if (cinfo->src == NULL) {     /* first time for this JPEG object? */
+    cinfo->src = (struct jpeg_source_mgr *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                                  sizeof(my_source_mgr));
+    src = (my_src_ptr)cinfo->src;
+    src->buffer = (JOCTET *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                                  INPUT_BUF_SIZE * sizeof(JOCTET));
+  } else if (cinfo->src->init_source != init_source) {
+    /* It is unsafe to reuse the existing source manager unless it was created
+     * by this function.  Otherwise, there is no guarantee that the opaque
+     * structure is the right size.  Note that we could just create a new
+     * structure, but the old structure would not be freed until
+     * jpeg_destroy_decompress() was called.
+     */
+    ERREXIT(cinfo, JERR_BUFFER_SIZE);
+  }
+
+  src = (my_src_ptr)cinfo->src;
+  src->pub.init_source = init_source;
+  src->pub.fill_input_buffer = fill_input_buffer;
+  src->pub.skip_input_data = skip_input_data;
+  src->pub.resync_to_restart = jpeg_resync_to_restart; /* use default method */
+  src->pub.term_source = term_source;
+  src->infile = infile;
+  src->pub.bytes_in_buffer = 0; /* forces fill_input_buffer on first read */
+  src->pub.next_input_byte = NULL; /* until buffer loaded */
+}
+
+
+#if JPEG_LIB_VERSION >= 80 || defined(MEM_SRCDST_SUPPORTED)
+/*
+ * Prepare for input from a supplied memory buffer.
+ * The buffer must contain the whole JPEG data.
+ */
+
+GLOBAL(void)
+jpeg_mem_src(j_decompress_ptr cinfo, const unsigned char *inbuffer,
+             unsigned long insize)
+{
+  struct jpeg_source_mgr *src;
+
+  if (inbuffer == NULL || insize == 0)  /* Treat empty input as fatal error */
+    ERREXIT(cinfo, JERR_INPUT_EMPTY);
+
+  /* The source object is made permanent so that a series of JPEG images
+   * can be read from the same buffer by calling jpeg_mem_src only before
+   * the first one.
+   */
+  if (cinfo->src == NULL) {     /* first time for this JPEG object? */
+    cinfo->src = (struct jpeg_source_mgr *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_PERMANENT,
+                                  sizeof(struct jpeg_source_mgr));
+  } else if (cinfo->src->init_source != init_mem_source) {
+    /* It is unsafe to reuse the existing source manager unless it was created
+     * by this function.
+     */
+    ERREXIT(cinfo, JERR_BUFFER_SIZE);
+  }
+
+  src = cinfo->src;
+  src->init_source = init_mem_source;
+  src->fill_input_buffer = fill_mem_input_buffer;
+  src->skip_input_data = skip_input_data;
+  src->resync_to_restart = jpeg_resync_to_restart; /* use default method */
+  src->term_source = term_source;
+  src->bytes_in_buffer = (size_t)insize;
+  src->next_input_byte = (const JOCTET *)inbuffer;
+}
+#endif

3rdparty/libjpeg-turbo/src/jdcoefct.c

@@ -0,0 +1,692 @@
+/*
+ * jdcoefct.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1997, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * Copyright (C) 2010, 2015-2016, D. R. Commander.
+ * Copyright (C) 2015, Google, Inc.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains the coefficient buffer controller for decompression.
+ * This controller is the top level of the JPEG decompressor proper.
+ * The coefficient buffer lies between entropy decoding and inverse-DCT steps.
+ *
+ * In buffered-image mode, this controller is the interface between
+ * input-oriented processing and output-oriented processing.
+ * Also, the input side (only) is used when reading a file for transcoding.
+ */
+
+#include "jinclude.h"
+#include "jdcoefct.h"
+#include "jpegcomp.h"
+
+
+/* Forward declarations */
+METHODDEF(int) decompress_onepass(j_decompress_ptr cinfo,
+                                  JSAMPIMAGE output_buf);
+#ifdef D_MULTISCAN_FILES_SUPPORTED
+METHODDEF(int) decompress_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf);
+#endif
+#ifdef BLOCK_SMOOTHING_SUPPORTED
+LOCAL(boolean) smoothing_ok(j_decompress_ptr cinfo);
+METHODDEF(int) decompress_smooth_data(j_decompress_ptr cinfo,
+                                      JSAMPIMAGE output_buf);
+#endif
+
+
+/*
+ * Initialize for an input processing pass.
+ */
+
+METHODDEF(void)
+start_input_pass(j_decompress_ptr cinfo)
+{
+  cinfo->input_iMCU_row = 0;
+  start_iMCU_row(cinfo);
+}
+
+
+/*
+ * Initialize for an output processing pass.
+ */
+
+METHODDEF(void)
+start_output_pass(j_decompress_ptr cinfo)
+{
+#ifdef BLOCK_SMOOTHING_SUPPORTED
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+
+  /* If multipass, check to see whether to use block smoothing on this pass */
+  if (coef->pub.coef_arrays != NULL) {
+    if (cinfo->do_block_smoothing && smoothing_ok(cinfo))
+      coef->pub.decompress_data = decompress_smooth_data;
+    else
+      coef->pub.decompress_data = decompress_data;
+  }
+#endif
+  cinfo->output_iMCU_row = 0;
+}
+
+
+/*
+ * Decompress and return some data in the single-pass case.
+ * Always attempts to emit one fully interleaved MCU row ("iMCU" row).
+ * Input and output must run in lockstep since we have only a one-MCU buffer.
+ * Return value is JPEG_ROW_COMPLETED, JPEG_SCAN_COMPLETED, or JPEG_SUSPENDED.
+ *
+ * NB: output_buf contains a plane for each component in image,
+ * which we index according to the component's SOF position.
+ */
+
+METHODDEF(int)
+decompress_onepass(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION MCU_col_num;       /* index of current MCU within row */
+  JDIMENSION last_MCU_col = cinfo->MCUs_per_row - 1;
+  JDIMENSION last_iMCU_row = cinfo->total_iMCU_rows - 1;
+  int blkn, ci, xindex, yindex, yoffset, useful_width;
+  JSAMPARRAY output_ptr;
+  JDIMENSION start_col, output_col;
+  jpeg_component_info *compptr;
+  inverse_DCT_method_ptr inverse_DCT;
+
+  /* Loop to process as much as one whole iMCU row */
+  for (yoffset = coef->MCU_vert_offset; yoffset < coef->MCU_rows_per_iMCU_row;
+       yoffset++) {
+    for (MCU_col_num = coef->MCU_ctr; MCU_col_num <= last_MCU_col;
+         MCU_col_num++) {
+      /* Try to fetch an MCU.  Entropy decoder expects buffer to be zeroed. */
+      jzero_far((void *)coef->MCU_buffer[0],
+                (size_t)(cinfo->blocks_in_MCU * sizeof(JBLOCK)));
+      if (!(*cinfo->entropy->decode_mcu) (cinfo, coef->MCU_buffer)) {
+        /* Suspension forced; update state counters and exit */
+        coef->MCU_vert_offset = yoffset;
+        coef->MCU_ctr = MCU_col_num;
+        return JPEG_SUSPENDED;
+      }
+
+      /* Only perform the IDCT on blocks that are contained within the desired
+       * cropping region.
+       */
+      if (MCU_col_num >= cinfo->master->first_iMCU_col &&
+          MCU_col_num <= cinfo->master->last_iMCU_col) {
+        /* Determine where data should go in output_buf and do the IDCT thing.
+         * We skip dummy blocks at the right and bottom edges (but blkn gets
+         * incremented past them!).  Note the inner loop relies on having
+         * allocated the MCU_buffer[] blocks sequentially.
+         */
+        blkn = 0;               /* index of current DCT block within MCU */
+        for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+          compptr = cinfo->cur_comp_info[ci];
+          /* Don't bother to IDCT an uninteresting component. */
+          if (!compptr->component_needed) {
+            blkn += compptr->MCU_blocks;
+            continue;
+          }
+          inverse_DCT = cinfo->idct->inverse_DCT[compptr->component_index];
+          useful_width = (MCU_col_num < last_MCU_col) ?
+                         compptr->MCU_width : compptr->last_col_width;
+          output_ptr = output_buf[compptr->component_index] +
+                       yoffset * compptr->_DCT_scaled_size;
+          start_col = (MCU_col_num - cinfo->master->first_iMCU_col) *
+                      compptr->MCU_sample_width;
+          for (yindex = 0; yindex < compptr->MCU_height; yindex++) {
+            if (cinfo->input_iMCU_row < last_iMCU_row ||
+                yoffset + yindex < compptr->last_row_height) {
+              output_col = start_col;
+              for (xindex = 0; xindex < useful_width; xindex++) {
+                (*inverse_DCT) (cinfo, compptr,
+                                (JCOEFPTR)coef->MCU_buffer[blkn + xindex],
+                                output_ptr, output_col);
+                output_col += compptr->_DCT_scaled_size;
+              }
+            }
+            blkn += compptr->MCU_width;
+            output_ptr += compptr->_DCT_scaled_size;
+          }
+        }
+      }
+    }
+    /* Completed an MCU row, but perhaps not an iMCU row */
+    coef->MCU_ctr = 0;
+  }
+  /* Completed the iMCU row, advance counters for next one */
+  cinfo->output_iMCU_row++;
+  if (++(cinfo->input_iMCU_row) < cinfo->total_iMCU_rows) {
+    start_iMCU_row(cinfo);
+    return JPEG_ROW_COMPLETED;
+  }
+  /* Completed the scan */
+  (*cinfo->inputctl->finish_input_pass) (cinfo);
+  return JPEG_SCAN_COMPLETED;
+}
+
+
+/*
+ * Dummy consume-input routine for single-pass operation.
+ */
+
+METHODDEF(int)
+dummy_consume_data(j_decompress_ptr cinfo)
+{
+  return JPEG_SUSPENDED;        /* Always indicate nothing was done */
+}
+
+
+#ifdef D_MULTISCAN_FILES_SUPPORTED
+
+/*
+ * Consume input data and store it in the full-image coefficient buffer.
+ * We read as much as one fully interleaved MCU row ("iMCU" row) per call,
+ * ie, v_samp_factor block rows for each component in the scan.
+ * Return value is JPEG_ROW_COMPLETED, JPEG_SCAN_COMPLETED, or JPEG_SUSPENDED.
+ */
+
+METHODDEF(int)
+consume_data(j_decompress_ptr cinfo)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION MCU_col_num;       /* index of current MCU within row */
+  int blkn, ci, xindex, yindex, yoffset;
+  JDIMENSION start_col;
+  JBLOCKARRAY buffer[MAX_COMPS_IN_SCAN];
+  JBLOCKROW buffer_ptr;
+  jpeg_component_info *compptr;
+
+  /* Align the virtual buffers for the components used in this scan. */
+  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+    compptr = cinfo->cur_comp_info[ci];
+    buffer[ci] = (*cinfo->mem->access_virt_barray)
+      ((j_common_ptr)cinfo, coef->whole_image[compptr->component_index],
+       cinfo->input_iMCU_row * compptr->v_samp_factor,
+       (JDIMENSION)compptr->v_samp_factor, TRUE);
+    /* Note: entropy decoder expects buffer to be zeroed,
+     * but this is handled automatically by the memory manager
+     * because we requested a pre-zeroed array.
+     */
+  }
+
+  /* Loop to process one whole iMCU row */
+  for (yoffset = coef->MCU_vert_offset; yoffset < coef->MCU_rows_per_iMCU_row;
+       yoffset++) {
+    for (MCU_col_num = coef->MCU_ctr; MCU_col_num < cinfo->MCUs_per_row;
+         MCU_col_num++) {
+      /* Construct list of pointers to DCT blocks belonging to this MCU */
+      blkn = 0;                 /* index of current DCT block within MCU */
+      for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
+        compptr = cinfo->cur_comp_info[ci];
+        start_col = MCU_col_num * compptr->MCU_width;
+        for (yindex = 0; yindex < compptr->MCU_height; yindex++) {
+          buffer_ptr = buffer[ci][yindex + yoffset] + start_col;
+          for (xindex = 0; xindex < compptr->MCU_width; xindex++) {
+            coef->MCU_buffer[blkn++] = buffer_ptr++;
+          }
+        }
+      }
+      /* Try to fetch the MCU. */
+      if (!(*cinfo->entropy->decode_mcu) (cinfo, coef->MCU_buffer)) {
+        /* Suspension forced; update state counters and exit */
+        coef->MCU_vert_offset = yoffset;
+        coef->MCU_ctr = MCU_col_num;
+        return JPEG_SUSPENDED;
+      }
+    }
+    /* Completed an MCU row, but perhaps not an iMCU row */
+    coef->MCU_ctr = 0;
+  }
+  /* Completed the iMCU row, advance counters for next one */
+  if (++(cinfo->input_iMCU_row) < cinfo->total_iMCU_rows) {
+    start_iMCU_row(cinfo);
+    return JPEG_ROW_COMPLETED;
+  }
+  /* Completed the scan */
+  (*cinfo->inputctl->finish_input_pass) (cinfo);
+  return JPEG_SCAN_COMPLETED;
+}
+
+
+/*
+ * Decompress and return some data in the multi-pass case.
+ * Always attempts to emit one fully interleaved MCU row ("iMCU" row).
+ * Return value is JPEG_ROW_COMPLETED, JPEG_SCAN_COMPLETED, or JPEG_SUSPENDED.
+ *
+ * NB: output_buf contains a plane for each component in image.
+ */
+
+METHODDEF(int)
+decompress_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION last_iMCU_row = cinfo->total_iMCU_rows - 1;
+  JDIMENSION block_num;
+  int ci, block_row, block_rows;
+  JBLOCKARRAY buffer;
+  JBLOCKROW buffer_ptr;
+  JSAMPARRAY output_ptr;
+  JDIMENSION output_col;
+  jpeg_component_info *compptr;
+  inverse_DCT_method_ptr inverse_DCT;
+
+  /* Force some input to be done if we are getting ahead of the input. */
+  while (cinfo->input_scan_number < cinfo->output_scan_number ||
+         (cinfo->input_scan_number == cinfo->output_scan_number &&
+          cinfo->input_iMCU_row <= cinfo->output_iMCU_row)) {
+    if ((*cinfo->inputctl->consume_input) (cinfo) == JPEG_SUSPENDED)
+      return JPEG_SUSPENDED;
+  }
+
+  /* OK, output from the virtual arrays. */
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    /* Don't bother to IDCT an uninteresting component. */
+    if (!compptr->component_needed)
+      continue;
+    /* Align the virtual buffer for this component. */
+    buffer = (*cinfo->mem->access_virt_barray)
+      ((j_common_ptr)cinfo, coef->whole_image[ci],
+       cinfo->output_iMCU_row * compptr->v_samp_factor,
+       (JDIMENSION)compptr->v_samp_factor, FALSE);
+    /* Count non-dummy DCT block rows in this iMCU row. */
+    if (cinfo->output_iMCU_row < last_iMCU_row)
+      block_rows = compptr->v_samp_factor;
+    else {
+      /* NB: can't use last_row_height here; it is input-side-dependent! */
+      block_rows = (int)(compptr->height_in_blocks % compptr->v_samp_factor);
+      if (block_rows == 0) block_rows = compptr->v_samp_factor;
+    }
+    inverse_DCT = cinfo->idct->inverse_DCT[ci];
+    output_ptr = output_buf[ci];
+    /* Loop over all DCT blocks to be processed. */
+    for (block_row = 0; block_row < block_rows; block_row++) {
+      buffer_ptr = buffer[block_row] + cinfo->master->first_MCU_col[ci];
+      output_col = 0;
+      for (block_num = cinfo->master->first_MCU_col[ci];
+           block_num <= cinfo->master->last_MCU_col[ci]; block_num++) {
+        (*inverse_DCT) (cinfo, compptr, (JCOEFPTR)buffer_ptr, output_ptr,
+                        output_col);
+        buffer_ptr++;
+        output_col += compptr->_DCT_scaled_size;
+      }
+      output_ptr += compptr->_DCT_scaled_size;
+    }
+  }
+
+  if (++(cinfo->output_iMCU_row) < cinfo->total_iMCU_rows)
+    return JPEG_ROW_COMPLETED;
+  return JPEG_SCAN_COMPLETED;
+}
+
+#endif /* D_MULTISCAN_FILES_SUPPORTED */
+
+
+#ifdef BLOCK_SMOOTHING_SUPPORTED
+
+/*
+ * This code applies interblock smoothing as described by section K.8
+ * of the JPEG standard: the first 5 AC coefficients are estimated from
+ * the DC values of a DCT block and its 8 neighboring blocks.
+ * We apply smoothing only for progressive JPEG decoding, and only if
+ * the coefficients it can estimate are not yet known to full precision.
+ */
+
+/* Natural-order array positions of the first 5 zigzag-order coefficients */
+#define Q01_POS  1
+#define Q10_POS  8
+#define Q20_POS  16
+#define Q11_POS  9
+#define Q02_POS  2
+
+/*
+ * Determine whether block smoothing is applicable and safe.
+ * We also latch the current states of the coef_bits[] entries for the
+ * AC coefficients; otherwise, if the input side of the decompressor
+ * advances into a new scan, we might think the coefficients are known
+ * more accurately than they really are.
+ */
+
+LOCAL(boolean)
+smoothing_ok(j_decompress_ptr cinfo)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  boolean smoothing_useful = FALSE;
+  int ci, coefi;
+  jpeg_component_info *compptr;
+  JQUANT_TBL *qtable;
+  int *coef_bits;
+  int *coef_bits_latch;
+
+  if (!cinfo->progressive_mode || cinfo->coef_bits == NULL)
+    return FALSE;
+
+  /* Allocate latch area if not already done */
+  if (coef->coef_bits_latch == NULL)
+    coef->coef_bits_latch = (int *)
+      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                  cinfo->num_components *
+                                  (SAVED_COEFS * sizeof(int)));
+  coef_bits_latch = coef->coef_bits_latch;
+
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    /* All components' quantization values must already be latched. */
+    if ((qtable = compptr->quant_table) == NULL)
+      return FALSE;
+    /* Verify DC & first 5 AC quantizers are nonzero to avoid zero-divide. */
+    if (qtable->quantval[0] == 0 ||
+        qtable->quantval[Q01_POS] == 0 ||
+        qtable->quantval[Q10_POS] == 0 ||
+        qtable->quantval[Q20_POS] == 0 ||
+        qtable->quantval[Q11_POS] == 0 ||
+        qtable->quantval[Q02_POS] == 0)
+      return FALSE;
+    /* DC values must be at least partly known for all components. */
+    coef_bits = cinfo->coef_bits[ci];
+    if (coef_bits[0] < 0)
+      return FALSE;
+    /* Block smoothing is helpful if some AC coefficients remain inaccurate. */
+    for (coefi = 1; coefi <= 5; coefi++) {
+      coef_bits_latch[coefi] = coef_bits[coefi];
+      if (coef_bits[coefi] != 0)
+        smoothing_useful = TRUE;
+    }
+    coef_bits_latch += SAVED_COEFS;
+  }
+
+  return smoothing_useful;
+}
+
+
+/*
+ * Variant of decompress_data for use when doing block smoothing.
+ */
+
+METHODDEF(int)
+decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+  JDIMENSION last_iMCU_row = cinfo->total_iMCU_rows - 1;
+  JDIMENSION block_num, last_block_column;
+  int ci, block_row, block_rows, access_rows;
+  JBLOCKARRAY buffer;
+  JBLOCKROW buffer_ptr, prev_block_row, next_block_row;
+  JSAMPARRAY output_ptr;
+  JDIMENSION output_col;
+  jpeg_component_info *compptr;
+  inverse_DCT_method_ptr inverse_DCT;
+  boolean first_row, last_row;
+  JCOEF *workspace;
+  int *coef_bits;
+  JQUANT_TBL *quanttbl;
+  JLONG Q00, Q01, Q02, Q10, Q11, Q20, num;
+  int DC1, DC2, DC3, DC4, DC5, DC6, DC7, DC8, DC9;
+  int Al, pred;
+
+  /* Keep a local variable to avoid looking it up more than once */
+  workspace = coef->workspace;
+
+  /* Force some input to be done if we are getting ahead of the input. */
+  while (cinfo->input_scan_number <= cinfo->output_scan_number &&
+         !cinfo->inputctl->eoi_reached) {
+    if (cinfo->input_scan_number == cinfo->output_scan_number) {
+      /* If input is working on current scan, we ordinarily want it to
+       * have completed the current row.  But if input scan is DC,
+       * we want it to keep one row ahead so that next block row's DC
+       * values are up to date.
+       */
+      JDIMENSION delta = (cinfo->Ss == 0) ? 1 : 0;
+      if (cinfo->input_iMCU_row > cinfo->output_iMCU_row + delta)
+        break;
+    }
+    if ((*cinfo->inputctl->consume_input) (cinfo) == JPEG_SUSPENDED)
+      return JPEG_SUSPENDED;
+  }
+
+  /* OK, output from the virtual arrays. */
+  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+       ci++, compptr++) {
+    /* Don't bother to IDCT an uninteresting component. */
+    if (!compptr->component_needed)
+      continue;
+    /* Count non-dummy DCT block rows in this iMCU row. */
+    if (cinfo->output_iMCU_row < last_iMCU_row) {
+      block_rows = compptr->v_samp_factor;
+      access_rows = block_rows * 2; /* this and next iMCU row */
+      last_row = FALSE;
+    } else {
+      /* NB: can't use last_row_height here; it is input-side-dependent! */
+      block_rows = (int)(compptr->height_in_blocks % compptr->v_samp_factor);
+      if (block_rows == 0) block_rows = compptr->v_samp_factor;
+      access_rows = block_rows; /* this iMCU row only */
+      last_row = TRUE;
+    }
+    /* Align the virtual buffer for this component. */
+    if (cinfo->output_iMCU_row > 0) {
+      access_rows += compptr->v_samp_factor; /* prior iMCU row too */
+      buffer = (*cinfo->mem->access_virt_barray)
+        ((j_common_ptr)cinfo, coef->whole_image[ci],
+         (cinfo->output_iMCU_row - 1) * compptr->v_samp_factor,
+         (JDIMENSION)access_rows, FALSE);
+      buffer += compptr->v_samp_factor; /* point to current iMCU row */
+      first_row = FALSE;
+    } else {
+      buffer = (*cinfo->mem->access_virt_barray)
+        ((j_common_ptr)cinfo, coef->whole_image[ci],
+         (JDIMENSION)0, (JDIMENSION)access_rows, FALSE);
+      first_row = TRUE;
+    }
+    /* Fetch component-dependent info */
+    coef_bits = coef->coef_bits_latch + (ci * SAVED_COEFS);
+    quanttbl = compptr->quant_table;
+    Q00 = quanttbl->quantval[0];
+    Q01 = quanttbl->quantval[Q01_POS];
+    Q10 = quanttbl->quantval[Q10_POS];
+    Q20 = quanttbl->quantval[Q20_POS];
+    Q11 = quanttbl->quantval[Q11_POS];
+    Q02 = quanttbl->quantval[Q02_POS];
+    inverse_DCT = cinfo->idct->inverse_DCT[ci];
+    output_ptr = output_buf[ci];
+    /* Loop over all DCT blocks to be processed. */
+    for (block_row = 0; block_row < block_rows; block_row++) {
+      buffer_ptr = buffer[block_row] + cinfo->master->first_MCU_col[ci];
+      if (first_row && block_row == 0)
+        prev_block_row = buffer_ptr;
+      else
+        prev_block_row = buffer[block_row - 1];
+      if (last_row && block_row == block_rows - 1)
+        next_block_row = buffer_ptr;
+      else
+        next_block_row = buffer[block_row + 1];
+      /* We fetch the surrounding DC values using a sliding-register approach.
+       * Initialize all nine here so as to do the right thing on narrow pics.
+       */
+      DC1 = DC2 = DC3 = (int)prev_block_row[0][0];
+      DC4 = DC5 = DC6 = (int)buffer_ptr[0][0];
+      DC7 = DC8 = DC9 = (int)next_block_row[0][0];
+      output_col = 0;
+      last_block_column = compptr->width_in_blocks - 1;
+      for (block_num = cinfo->master->first_MCU_col[ci];
+           block_num <= cinfo->master->last_MCU_col[ci]; block_num++) {
+        /* Fetch current DCT block into workspace so we can modify it. */
+        jcopy_block_row(buffer_ptr, (JBLOCKROW)workspace, (JDIMENSION)1);
+        /* Update DC values */
+        if (block_num < last_block_column) {
+          DC3 = (int)prev_block_row[1][0];
+          DC6 = (int)buffer_ptr[1][0];
+          DC9 = (int)next_block_row[1][0];
+        }
+        /* Compute coefficient estimates per K.8.
+         * An estimate is applied only if coefficient is still zero,
+         * and is not known to be fully accurate.
+         */
+        /* AC01 */
+        if ((Al = coef_bits[1]) != 0 && workspace[1] == 0) {
+          num = 36 * Q00 * (DC4 - DC6);
+          if (num >= 0) {
+            pred = (int)(((Q01 << 7) + num) / (Q01 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+          } else {
+            pred = (int)(((Q01 << 7) - num) / (Q01 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+            pred = -pred;
+          }
+          workspace[1] = (JCOEF)pred;
+        }
+        /* AC10 */
+        if ((Al = coef_bits[2]) != 0 && workspace[8] == 0) {
+          num = 36 * Q00 * (DC2 - DC8);
+          if (num >= 0) {
+            pred = (int)(((Q10 << 7) + num) / (Q10 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+          } else {
+            pred = (int)(((Q10 << 7) - num) / (Q10 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+            pred = -pred;
+          }
+          workspace[8] = (JCOEF)pred;
+        }
+        /* AC20 */
+        if ((Al = coef_bits[3]) != 0 && workspace[16] == 0) {
+          num = 9 * Q00 * (DC2 + DC8 - 2 * DC5);
+          if (num >= 0) {
+            pred = (int)(((Q20 << 7) + num) / (Q20 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+          } else {
+            pred = (int)(((Q20 << 7) - num) / (Q20 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+            pred = -pred;
+          }
+          workspace[16] = (JCOEF)pred;
+        }
+        /* AC11 */
+        if ((Al = coef_bits[4]) != 0 && workspace[9] == 0) {
+          num = 5 * Q00 * (DC1 - DC3 - DC7 + DC9);
+          if (num >= 0) {
+            pred = (int)(((Q11 << 7) + num) / (Q11 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+          } else {
+            pred = (int)(((Q11 << 7) - num) / (Q11 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+            pred = -pred;
+          }
+          workspace[9] = (JCOEF)pred;
+        }
+        /* AC02 */
+        if ((Al = coef_bits[5]) != 0 && workspace[2] == 0) {
+          num = 9 * Q00 * (DC4 + DC6 - 2 * DC5);
+          if (num >= 0) {
+            pred = (int)(((Q02 << 7) + num) / (Q02 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+          } else {
+            pred = (int)(((Q02 << 7) - num) / (Q02 << 8));
+            if (Al > 0 && pred >= (1 << Al))
+              pred = (1 << Al) - 1;
+            pred = -pred;
+          }
+          workspace[2] = (JCOEF)pred;
+        }
+        /* OK, do the IDCT */
+        (*inverse_DCT) (cinfo, compptr, (JCOEFPTR)workspace, output_ptr,
+                        output_col);
+        /* Advance for next column */
+        DC1 = DC2;  DC2 = DC3;
+        DC4 = DC5;  DC5 = DC6;
+        DC7 = DC8;  DC8 = DC9;
+        buffer_ptr++, prev_block_row++, next_block_row++;
+        output_col += compptr->_DCT_scaled_size;
+      }
+      output_ptr += compptr->_DCT_scaled_size;
+    }
+  }
+
+  if (++(cinfo->output_iMCU_row) < cinfo->total_iMCU_rows)
+    return JPEG_ROW_COMPLETED;
+  return JPEG_SCAN_COMPLETED;
+}
+
+#endif /* BLOCK_SMOOTHING_SUPPORTED */
+
+
+/*
+ * Initialize coefficient buffer controller.
+ */
+
+GLOBAL(void)
+jinit_d_coef_controller(j_decompress_ptr cinfo, boolean need_full_buffer)
+{
+  my_coef_ptr coef;
+
+  coef = (my_coef_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_coef_controller));
+  cinfo->coef = (struct jpeg_d_coef_controller *)coef;
+  coef->pub.start_input_pass = start_input_pass;
+  coef->pub.start_output_pass = start_output_pass;
+#ifdef BLOCK_SMOOTHING_SUPPORTED
+  coef->coef_bits_latch = NULL;
+#endif
+
+  /* Create the coefficient buffer. */
+  if (need_full_buffer) {
+#ifdef D_MULTISCAN_FILES_SUPPORTED
+    /* Allocate a full-image virtual array for each component, */
+    /* padded to a multiple of samp_factor DCT blocks in each direction. */
+    /* Note we ask for a pre-zeroed array. */
+    int ci, access_rows;
+    jpeg_component_info *compptr;
+
+    for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
+         ci++, compptr++) {
+      access_rows = compptr->v_samp_factor;
+#ifdef BLOCK_SMOOTHING_SUPPORTED
+      /* If block smoothing could be used, need a bigger window */
+      if (cinfo->progressive_mode)
+        access_rows *= 3;
+#endif
+      coef->whole_image[ci] = (*cinfo->mem->request_virt_barray)
+        ((j_common_ptr)cinfo, JPOOL_IMAGE, TRUE,
+         (JDIMENSION)jround_up((long)compptr->width_in_blocks,
+                               (long)compptr->h_samp_factor),
+         (JDIMENSION)jround_up((long)compptr->height_in_blocks,
+                               (long)compptr->v_samp_factor),
+         (JDIMENSION)access_rows);
+    }
+    coef->pub.consume_data = consume_data;
+    coef->pub.decompress_data = decompress_data;
+    coef->pub.coef_arrays = coef->whole_image; /* link to virtual arrays */
+#else
+    ERREXIT(cinfo, JERR_NOT_COMPILED);
+#endif
+  } else {
+    /* We only need a single-MCU buffer. */
+    JBLOCKROW buffer;
+    int i;
+
+    buffer = (JBLOCKROW)
+      (*cinfo->mem->alloc_large) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                  D_MAX_BLOCKS_IN_MCU * sizeof(JBLOCK));
+    for (i = 0; i < D_MAX_BLOCKS_IN_MCU; i++) {
+      coef->MCU_buffer[i] = buffer + i;
+    }
+    coef->pub.consume_data = dummy_consume_data;
+    coef->pub.decompress_data = decompress_onepass;
+    coef->pub.coef_arrays = NULL; /* flag for no virtual arrays */
+  }
+
+  /* Allocate the workspace buffer */
+  coef->workspace = (JCOEF *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(JCOEF) * DCTSIZE2);
+}

3rdparty/libjpeg-turbo/src/jdcoefct.h

@@ -0,0 +1,82 @@
+/*
+ * jdcoefct.h
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1997, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ */
+
+#define JPEG_INTERNALS
+#include "jpeglib.h"
+
+
+/* Block smoothing is only applicable for progressive JPEG, so: */
+#ifndef D_PROGRESSIVE_SUPPORTED
+#undef BLOCK_SMOOTHING_SUPPORTED
+#endif
+
+
+/* Private buffer controller object */
+
+typedef struct {
+  struct jpeg_d_coef_controller pub; /* public fields */
+
+  /* These variables keep track of the current location of the input side. */
+  /* cinfo->input_iMCU_row is also used for this. */
+  JDIMENSION MCU_ctr;           /* counts MCUs processed in current row */
+  int MCU_vert_offset;          /* counts MCU rows within iMCU row */
+  int MCU_rows_per_iMCU_row;    /* number of such rows needed */
+
+  /* The output side's location is represented by cinfo->output_iMCU_row. */
+
+  /* In single-pass modes, it's sufficient to buffer just one MCU.
+   * We allocate a workspace of D_MAX_BLOCKS_IN_MCU coefficient blocks,
+   * and let the entropy decoder write into that workspace each time.
+   * In multi-pass modes, this array points to the current MCU's blocks
+   * within the virtual arrays; it is used only by the input side.
+   */
+  JBLOCKROW MCU_buffer[D_MAX_BLOCKS_IN_MCU];
+
+  /* Temporary workspace for one MCU */
+  JCOEF *workspace;
+
+#ifdef D_MULTISCAN_FILES_SUPPORTED
+  /* In multi-pass modes, we need a virtual block array for each component. */
+  jvirt_barray_ptr whole_image[MAX_COMPONENTS];
+#endif
+
+#ifdef BLOCK_SMOOTHING_SUPPORTED
+  /* When doing block smoothing, we latch coefficient Al values here */
+  int *coef_bits_latch;
+#define SAVED_COEFS  6          /* we save coef_bits[0..5] */
+#endif
+} my_coef_controller;
+
+typedef my_coef_controller *my_coef_ptr;
+
+
+LOCAL(void)
+start_iMCU_row(j_decompress_ptr cinfo)
+/* Reset within-iMCU-row counters for a new row (input side) */
+{
+  my_coef_ptr coef = (my_coef_ptr)cinfo->coef;
+
+  /* In an interleaved scan, an MCU row is the same as an iMCU row.
+   * In a noninterleaved scan, an iMCU row has v_samp_factor MCU rows.
+   * But at the bottom of the image, process only what's left.
+   */
+  if (cinfo->comps_in_scan > 1) {
+    coef->MCU_rows_per_iMCU_row = 1;
+  } else {
+    if (cinfo->input_iMCU_row < (cinfo->total_iMCU_rows - 1))
+      coef->MCU_rows_per_iMCU_row = cinfo->cur_comp_info[0]->v_samp_factor;
+    else
+      coef->MCU_rows_per_iMCU_row = cinfo->cur_comp_info[0]->last_row_height;
+  }
+
+  coef->MCU_ctr = 0;
+  coef->MCU_vert_offset = 0;
+}

3rdparty/libjpeg-turbo/src/jdcol565.c

@@ -0,0 +1,384 @@
+/*
+ * jdcol565.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1997, Thomas G. Lane.
+ * Modifications:
+ * Copyright (C) 2013, Linaro Limited.
+ * Copyright (C) 2014-2015, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains output colorspace conversion routines.
+ */
+
+/* This file is included by jdcolor.c */
+
+
+INLINE
+LOCAL(void)
+ycc_rgb565_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                            JDIMENSION input_row, JSAMPARRAY output_buf,
+                            int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int y, cb, cr;
+  register JSAMPROW outptr;
+  register JSAMPROW inptr0, inptr1, inptr2;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE *range_limit = cinfo->sample_range_limit;
+  register int *Crrtab = cconvert->Cr_r_tab;
+  register int *Cbbtab = cconvert->Cb_b_tab;
+  register JLONG *Crgtab = cconvert->Cr_g_tab;
+  register JLONG *Cbgtab = cconvert->Cb_g_tab;
+  SHIFT_TEMPS
+
+  while (--num_rows >= 0) {
+    JLONG rgb;
+    unsigned int r, g, b;
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    input_row++;
+    outptr = *output_buf++;
+
+    if (PACK_NEED_ALIGNMENT(outptr)) {
+      y  = GETJSAMPLE(*inptr0++);
+      cb = GETJSAMPLE(*inptr1++);
+      cr = GETJSAMPLE(*inptr2++);
+      r = range_limit[y + Crrtab[cr]];
+      g = range_limit[y + ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                            SCALEBITS))];
+      b = range_limit[y + Cbbtab[cb]];
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+      outptr += 2;
+      num_cols--;
+    }
+    for (col = 0; col < (num_cols >> 1); col++) {
+      y  = GETJSAMPLE(*inptr0++);
+      cb = GETJSAMPLE(*inptr1++);
+      cr = GETJSAMPLE(*inptr2++);
+      r = range_limit[y + Crrtab[cr]];
+      g = range_limit[y + ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                            SCALEBITS))];
+      b = range_limit[y + Cbbtab[cb]];
+      rgb = PACK_SHORT_565(r, g, b);
+
+      y  = GETJSAMPLE(*inptr0++);
+      cb = GETJSAMPLE(*inptr1++);
+      cr = GETJSAMPLE(*inptr2++);
+      r = range_limit[y + Crrtab[cr]];
+      g = range_limit[y + ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                            SCALEBITS))];
+      b = range_limit[y + Cbbtab[cb]];
+      rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r, g, b));
+
+      WRITE_TWO_ALIGNED_PIXELS(outptr, rgb);
+      outptr += 4;
+    }
+    if (num_cols & 1) {
+      y  = GETJSAMPLE(*inptr0);
+      cb = GETJSAMPLE(*inptr1);
+      cr = GETJSAMPLE(*inptr2);
+      r = range_limit[y + Crrtab[cr]];
+      g = range_limit[y + ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                            SCALEBITS))];
+      b = range_limit[y + Cbbtab[cb]];
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+    }
+  }
+}
+
+
+INLINE
+LOCAL(void)
+ycc_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                             JDIMENSION input_row, JSAMPARRAY output_buf,
+                             int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int y, cb, cr;
+  register JSAMPROW outptr;
+  register JSAMPROW inptr0, inptr1, inptr2;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE *range_limit = cinfo->sample_range_limit;
+  register int *Crrtab = cconvert->Cr_r_tab;
+  register int *Cbbtab = cconvert->Cb_b_tab;
+  register JLONG *Crgtab = cconvert->Cr_g_tab;
+  register JLONG *Cbgtab = cconvert->Cb_g_tab;
+  JLONG d0 = dither_matrix[cinfo->output_scanline & DITHER_MASK];
+  SHIFT_TEMPS
+
+  while (--num_rows >= 0) {
+    JLONG rgb;
+    unsigned int r, g, b;
+
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    input_row++;
+    outptr = *output_buf++;
+    if (PACK_NEED_ALIGNMENT(outptr)) {
+      y  = GETJSAMPLE(*inptr0++);
+      cb = GETJSAMPLE(*inptr1++);
+      cr = GETJSAMPLE(*inptr2++);
+      r = range_limit[DITHER_565_R(y + Crrtab[cr], d0)];
+      g = range_limit[DITHER_565_G(y +
+                                   ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                                     SCALEBITS)), d0)];
+      b = range_limit[DITHER_565_B(y + Cbbtab[cb], d0)];
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+      outptr += 2;
+      num_cols--;
+    }
+    for (col = 0; col < (num_cols >> 1); col++) {
+      y  = GETJSAMPLE(*inptr0++);
+      cb = GETJSAMPLE(*inptr1++);
+      cr = GETJSAMPLE(*inptr2++);
+      r = range_limit[DITHER_565_R(y + Crrtab[cr], d0)];
+      g = range_limit[DITHER_565_G(y +
+                                   ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                                     SCALEBITS)), d0)];
+      b = range_limit[DITHER_565_B(y + Cbbtab[cb], d0)];
+      d0 = DITHER_ROTATE(d0);
+      rgb = PACK_SHORT_565(r, g, b);
+
+      y  = GETJSAMPLE(*inptr0++);
+      cb = GETJSAMPLE(*inptr1++);
+      cr = GETJSAMPLE(*inptr2++);
+      r = range_limit[DITHER_565_R(y + Crrtab[cr], d0)];
+      g = range_limit[DITHER_565_G(y +
+                                   ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                                     SCALEBITS)), d0)];
+      b = range_limit[DITHER_565_B(y + Cbbtab[cb], d0)];
+      d0 = DITHER_ROTATE(d0);
+      rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r, g, b));
+
+      WRITE_TWO_ALIGNED_PIXELS(outptr, rgb);
+      outptr += 4;
+    }
+    if (num_cols & 1) {
+      y  = GETJSAMPLE(*inptr0);
+      cb = GETJSAMPLE(*inptr1);
+      cr = GETJSAMPLE(*inptr2);
+      r = range_limit[DITHER_565_R(y + Crrtab[cr], d0)];
+      g = range_limit[DITHER_565_G(y +
+                                   ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                                     SCALEBITS)), d0)];
+      b = range_limit[DITHER_565_B(y + Cbbtab[cb], d0)];
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+    }
+  }
+}
+
+
+INLINE
+LOCAL(void)
+rgb_rgb565_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                            JDIMENSION input_row, JSAMPARRAY output_buf,
+                            int num_rows)
+{
+  register JSAMPROW outptr;
+  register JSAMPROW inptr0, inptr1, inptr2;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+  SHIFT_TEMPS
+
+  while (--num_rows >= 0) {
+    JLONG rgb;
+    unsigned int r, g, b;
+
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    input_row++;
+    outptr = *output_buf++;
+    if (PACK_NEED_ALIGNMENT(outptr)) {
+      r = GETJSAMPLE(*inptr0++);
+      g = GETJSAMPLE(*inptr1++);
+      b = GETJSAMPLE(*inptr2++);
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+      outptr += 2;
+      num_cols--;
+    }
+    for (col = 0; col < (num_cols >> 1); col++) {
+      r = GETJSAMPLE(*inptr0++);
+      g = GETJSAMPLE(*inptr1++);
+      b = GETJSAMPLE(*inptr2++);
+      rgb = PACK_SHORT_565(r, g, b);
+
+      r = GETJSAMPLE(*inptr0++);
+      g = GETJSAMPLE(*inptr1++);
+      b = GETJSAMPLE(*inptr2++);
+      rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r, g, b));
+
+      WRITE_TWO_ALIGNED_PIXELS(outptr, rgb);
+      outptr += 4;
+    }
+    if (num_cols & 1) {
+      r = GETJSAMPLE(*inptr0);
+      g = GETJSAMPLE(*inptr1);
+      b = GETJSAMPLE(*inptr2);
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+    }
+  }
+}
+
+
+INLINE
+LOCAL(void)
+rgb_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                             JDIMENSION input_row, JSAMPARRAY output_buf,
+                             int num_rows)
+{
+  register JSAMPROW outptr;
+  register JSAMPROW inptr0, inptr1, inptr2;
+  register JDIMENSION col;
+  register JSAMPLE *range_limit = cinfo->sample_range_limit;
+  JDIMENSION num_cols = cinfo->output_width;
+  JLONG d0 = dither_matrix[cinfo->output_scanline & DITHER_MASK];
+  SHIFT_TEMPS
+
+  while (--num_rows >= 0) {
+    JLONG rgb;
+    unsigned int r, g, b;
+
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    input_row++;
+    outptr = *output_buf++;
+    if (PACK_NEED_ALIGNMENT(outptr)) {
+      r = range_limit[DITHER_565_R(GETJSAMPLE(*inptr0++), d0)];
+      g = range_limit[DITHER_565_G(GETJSAMPLE(*inptr1++), d0)];
+      b = range_limit[DITHER_565_B(GETJSAMPLE(*inptr2++), d0)];
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+      outptr += 2;
+      num_cols--;
+    }
+    for (col = 0; col < (num_cols >> 1); col++) {
+      r = range_limit[DITHER_565_R(GETJSAMPLE(*inptr0++), d0)];
+      g = range_limit[DITHER_565_G(GETJSAMPLE(*inptr1++), d0)];
+      b = range_limit[DITHER_565_B(GETJSAMPLE(*inptr2++), d0)];
+      d0 = DITHER_ROTATE(d0);
+      rgb = PACK_SHORT_565(r, g, b);
+
+      r = range_limit[DITHER_565_R(GETJSAMPLE(*inptr0++), d0)];
+      g = range_limit[DITHER_565_G(GETJSAMPLE(*inptr1++), d0)];
+      b = range_limit[DITHER_565_B(GETJSAMPLE(*inptr2++), d0)];
+      d0 = DITHER_ROTATE(d0);
+      rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r, g, b));
+
+      WRITE_TWO_ALIGNED_PIXELS(outptr, rgb);
+      outptr += 4;
+    }
+    if (num_cols & 1) {
+      r = range_limit[DITHER_565_R(GETJSAMPLE(*inptr0), d0)];
+      g = range_limit[DITHER_565_G(GETJSAMPLE(*inptr1), d0)];
+      b = range_limit[DITHER_565_B(GETJSAMPLE(*inptr2), d0)];
+      rgb = PACK_SHORT_565(r, g, b);
+      *(INT16 *)outptr = (INT16)rgb;
+    }
+  }
+}
+
+
+INLINE
+LOCAL(void)
+gray_rgb565_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                             JDIMENSION input_row, JSAMPARRAY output_buf,
+                             int num_rows)
+{
+  register JSAMPROW inptr, outptr;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+
+  while (--num_rows >= 0) {
+    JLONG rgb;
+    unsigned int g;
+
+    inptr = input_buf[0][input_row++];
+    outptr = *output_buf++;
+    if (PACK_NEED_ALIGNMENT(outptr)) {
+      g = *inptr++;
+      rgb = PACK_SHORT_565(g, g, g);
+      *(INT16 *)outptr = (INT16)rgb;
+      outptr += 2;
+      num_cols--;
+    }
+    for (col = 0; col < (num_cols >> 1); col++) {
+      g = *inptr++;
+      rgb = PACK_SHORT_565(g, g, g);
+      g = *inptr++;
+      rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(g, g, g));
+      WRITE_TWO_ALIGNED_PIXELS(outptr, rgb);
+      outptr += 4;
+    }
+    if (num_cols & 1) {
+      g = *inptr;
+      rgb = PACK_SHORT_565(g, g, g);
+      *(INT16 *)outptr = (INT16)rgb;
+    }
+  }
+}
+
+
+INLINE
+LOCAL(void)
+gray_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                              JDIMENSION input_row, JSAMPARRAY output_buf,
+                              int num_rows)
+{
+  register JSAMPROW inptr, outptr;
+  register JDIMENSION col;
+  register JSAMPLE *range_limit = cinfo->sample_range_limit;
+  JDIMENSION num_cols = cinfo->output_width;
+  JLONG d0 = dither_matrix[cinfo->output_scanline & DITHER_MASK];
+
+  while (--num_rows >= 0) {
+    JLONG rgb;
+    unsigned int g;
+
+    inptr = input_buf[0][input_row++];
+    outptr = *output_buf++;
+    if (PACK_NEED_ALIGNMENT(outptr)) {
+      g = *inptr++;
+      g = range_limit[DITHER_565_R(g, d0)];
+      rgb = PACK_SHORT_565(g, g, g);
+      *(INT16 *)outptr = (INT16)rgb;
+      outptr += 2;
+      num_cols--;
+    }
+    for (col = 0; col < (num_cols >> 1); col++) {
+      g = *inptr++;
+      g = range_limit[DITHER_565_R(g, d0)];
+      rgb = PACK_SHORT_565(g, g, g);
+      d0 = DITHER_ROTATE(d0);
+
+      g = *inptr++;
+      g = range_limit[DITHER_565_R(g, d0)];
+      rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(g, g, g));
+      d0 = DITHER_ROTATE(d0);
+
+      WRITE_TWO_ALIGNED_PIXELS(outptr, rgb);
+      outptr += 4;
+    }
+    if (num_cols & 1) {
+      g = *inptr;
+      g = range_limit[DITHER_565_R(g, d0)];
+      rgb = PACK_SHORT_565(g, g, g);
+      *(INT16 *)outptr = (INT16)rgb;
+    }
+  }
+}

3rdparty/libjpeg-turbo/src/jdcolext.c

@@ -0,0 +1,143 @@
+/*
+ * jdcolext.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1997, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2009, 2011, 2015, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains output colorspace conversion routines.
+ */
+
+
+/* This file is included by jdcolor.c */
+
+
+/*
+ * Convert some rows of samples to the output colorspace.
+ *
+ * Note that we change from noninterleaved, one-plane-per-component format
+ * to interleaved-pixel format.  The output buffer is therefore three times
+ * as wide as the input buffer.
+ * A starting row offset is provided only for the input buffer.  The caller
+ * can easily adjust the passed output_buf value to accommodate any row
+ * offset required on that side.
+ */
+
+INLINE
+LOCAL(void)
+ycc_rgb_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                         JDIMENSION input_row, JSAMPARRAY output_buf,
+                         int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int y, cb, cr;
+  register JSAMPROW outptr;
+  register JSAMPROW inptr0, inptr1, inptr2;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE *range_limit = cinfo->sample_range_limit;
+  register int *Crrtab = cconvert->Cr_r_tab;
+  register int *Cbbtab = cconvert->Cb_b_tab;
+  register JLONG *Crgtab = cconvert->Cr_g_tab;
+  register JLONG *Cbgtab = cconvert->Cb_g_tab;
+  SHIFT_TEMPS
+
+  while (--num_rows >= 0) {
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    input_row++;
+    outptr = *output_buf++;
+    for (col = 0; col < num_cols; col++) {
+      y  = GETJSAMPLE(inptr0[col]);
+      cb = GETJSAMPLE(inptr1[col]);
+      cr = GETJSAMPLE(inptr2[col]);
+      /* Range-limiting is essential due to noise introduced by DCT losses. */
+      outptr[RGB_RED] =   range_limit[y + Crrtab[cr]];
+      outptr[RGB_GREEN] = range_limit[y +
+                              ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                                SCALEBITS))];
+      outptr[RGB_BLUE] =  range_limit[y + Cbbtab[cb]];
+      /* Set unused byte to 0xFF so it can be interpreted as an opaque */
+      /* alpha channel value */
+#ifdef RGB_ALPHA
+      outptr[RGB_ALPHA] = 0xFF;
+#endif
+      outptr += RGB_PIXELSIZE;
+    }
+  }
+}
+
+
+/*
+ * Convert grayscale to RGB: just duplicate the graylevel three times.
+ * This is provided to support applications that don't want to cope
+ * with grayscale as a separate case.
+ */
+
+INLINE
+LOCAL(void)
+gray_rgb_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                          JDIMENSION input_row, JSAMPARRAY output_buf,
+                          int num_rows)
+{
+  register JSAMPROW inptr, outptr;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+
+  while (--num_rows >= 0) {
+    inptr = input_buf[0][input_row++];
+    outptr = *output_buf++;
+    for (col = 0; col < num_cols; col++) {
+      /* We can dispense with GETJSAMPLE() here */
+      outptr[RGB_RED] = outptr[RGB_GREEN] = outptr[RGB_BLUE] = inptr[col];
+      /* Set unused byte to 0xFF so it can be interpreted as an opaque */
+      /* alpha channel value */
+#ifdef RGB_ALPHA
+      outptr[RGB_ALPHA] = 0xFF;
+#endif
+      outptr += RGB_PIXELSIZE;
+    }
+  }
+}
+
+
+/*
+ * Convert RGB to extended RGB: just swap the order of source pixels
+ */
+
+INLINE
+LOCAL(void)
+rgb_rgb_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                         JDIMENSION input_row, JSAMPARRAY output_buf,
+                         int num_rows)
+{
+  register JSAMPROW inptr0, inptr1, inptr2;
+  register JSAMPROW outptr;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+
+  while (--num_rows >= 0) {
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    input_row++;
+    outptr = *output_buf++;
+    for (col = 0; col < num_cols; col++) {
+      /* We can dispense with GETJSAMPLE() here */
+      outptr[RGB_RED] = inptr0[col];
+      outptr[RGB_GREEN] = inptr1[col];
+      outptr[RGB_BLUE] = inptr2[col];
+      /* Set unused byte to 0xFF so it can be interpreted as an opaque */
+      /* alpha channel value */
+#ifdef RGB_ALPHA
+      outptr[RGB_ALPHA] = 0xFF;
+#endif
+      outptr += RGB_PIXELSIZE;
+    }
+  }
+}

3rdparty/libjpeg-turbo/src/jdcolor.c

@@ -0,0 +1,883 @@
+/*
+ * jdcolor.c
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1991-1997, Thomas G. Lane.
+ * Modified 2011 by Guido Vollbeding.
+ * libjpeg-turbo Modifications:
+ * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * Copyright (C) 2009, 2011-2012, 2014-2015, D. R. Commander.
+ * Copyright (C) 2013, Linaro Limited.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This file contains output colorspace conversion routines.
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+#include "jsimd.h"
+#include "jconfigint.h"
+
+
+/* Private subobject */
+
+typedef struct {
+  struct jpeg_color_deconverter pub; /* public fields */
+
+  /* Private state for YCC->RGB conversion */
+  int *Cr_r_tab;                /* => table for Cr to R conversion */
+  int *Cb_b_tab;                /* => table for Cb to B conversion */
+  JLONG *Cr_g_tab;              /* => table for Cr to G conversion */
+  JLONG *Cb_g_tab;              /* => table for Cb to G conversion */
+
+  /* Private state for RGB->Y conversion */
+  JLONG *rgb_y_tab;             /* => table for RGB to Y conversion */
+} my_color_deconverter;
+
+typedef my_color_deconverter *my_cconvert_ptr;
+
+
+/**************** YCbCr -> RGB conversion: most common case **************/
+/****************   RGB -> Y   conversion: less common case **************/
+
+/*
+ * YCbCr is defined per CCIR 601-1, except that Cb and Cr are
+ * normalized to the range 0..MAXJSAMPLE rather than -0.5 .. 0.5.
+ * The conversion equations to be implemented are therefore
+ *
+ *      R = Y                + 1.40200 * Cr
+ *      G = Y - 0.34414 * Cb - 0.71414 * Cr
+ *      B = Y + 1.77200 * Cb
+ *
+ *      Y = 0.29900 * R + 0.58700 * G + 0.11400 * B
+ *
+ * where Cb and Cr represent the incoming values less CENTERJSAMPLE.
+ * (These numbers are derived from TIFF 6.0 section 21, dated 3-June-92.)
+ *
+ * To avoid floating-point arithmetic, we represent the fractional constants
+ * as integers scaled up by 2^16 (about 4 digits precision); we have to divide
+ * the products by 2^16, with appropriate rounding, to get the correct answer.
+ * Notice that Y, being an integral input, does not contribute any fraction
+ * so it need not participate in the rounding.
+ *
+ * For even more speed, we avoid doing any multiplications in the inner loop
+ * by precalculating the constants times Cb and Cr for all possible values.
+ * For 8-bit JSAMPLEs this is very reasonable (only 256 entries per table);
+ * for 12-bit samples it is still acceptable.  It's not very reasonable for
+ * 16-bit samples, but if you want lossless storage you shouldn't be changing
+ * colorspace anyway.
+ * The Cr=>R and Cb=>B values can be rounded to integers in advance; the
+ * values for the G calculation are left scaled up, since we must add them
+ * together before rounding.
+ */
+
+#define SCALEBITS       16      /* speediest right-shift on some machines */
+#define ONE_HALF        ((JLONG)1 << (SCALEBITS - 1))
+#define FIX(x)          ((JLONG)((x) * (1L << SCALEBITS) + 0.5))
+
+/* We allocate one big table for RGB->Y conversion and divide it up into
+ * three parts, instead of doing three alloc_small requests.  This lets us
+ * use a single table base address, which can be held in a register in the
+ * inner loops on many machines (more than can hold all three addresses,
+ * anyway).
+ */
+
+#define R_Y_OFF         0                       /* offset to R => Y section */
+#define G_Y_OFF         (1 * (MAXJSAMPLE + 1))  /* offset to G => Y section */
+#define B_Y_OFF         (2 * (MAXJSAMPLE + 1))  /* etc. */
+#define TABLE_SIZE      (3 * (MAXJSAMPLE + 1))
+
+
+/* Include inline routines for colorspace extensions */
+
+#include "jdcolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+
+#define RGB_RED  EXT_RGB_RED
+#define RGB_GREEN  EXT_RGB_GREEN
+#define RGB_BLUE  EXT_RGB_BLUE
+#define RGB_PIXELSIZE  EXT_RGB_PIXELSIZE
+#define ycc_rgb_convert_internal  ycc_extrgb_convert_internal
+#define gray_rgb_convert_internal  gray_extrgb_convert_internal
+#define rgb_rgb_convert_internal  rgb_extrgb_convert_internal
+#include "jdcolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef ycc_rgb_convert_internal
+#undef gray_rgb_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_RGBX_RED
+#define RGB_GREEN  EXT_RGBX_GREEN
+#define RGB_BLUE  EXT_RGBX_BLUE
+#define RGB_ALPHA  3
+#define RGB_PIXELSIZE  EXT_RGBX_PIXELSIZE
+#define ycc_rgb_convert_internal  ycc_extrgbx_convert_internal
+#define gray_rgb_convert_internal  gray_extrgbx_convert_internal
+#define rgb_rgb_convert_internal  rgb_extrgbx_convert_internal
+#include "jdcolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_ALPHA
+#undef RGB_PIXELSIZE
+#undef ycc_rgb_convert_internal
+#undef gray_rgb_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_BGR_RED
+#define RGB_GREEN  EXT_BGR_GREEN
+#define RGB_BLUE  EXT_BGR_BLUE
+#define RGB_PIXELSIZE  EXT_BGR_PIXELSIZE
+#define ycc_rgb_convert_internal  ycc_extbgr_convert_internal
+#define gray_rgb_convert_internal  gray_extbgr_convert_internal
+#define rgb_rgb_convert_internal  rgb_extbgr_convert_internal
+#include "jdcolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_PIXELSIZE
+#undef ycc_rgb_convert_internal
+#undef gray_rgb_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_BGRX_RED
+#define RGB_GREEN  EXT_BGRX_GREEN
+#define RGB_BLUE  EXT_BGRX_BLUE
+#define RGB_ALPHA  3
+#define RGB_PIXELSIZE  EXT_BGRX_PIXELSIZE
+#define ycc_rgb_convert_internal  ycc_extbgrx_convert_internal
+#define gray_rgb_convert_internal  gray_extbgrx_convert_internal
+#define rgb_rgb_convert_internal  rgb_extbgrx_convert_internal
+#include "jdcolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_ALPHA
+#undef RGB_PIXELSIZE
+#undef ycc_rgb_convert_internal
+#undef gray_rgb_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_XBGR_RED
+#define RGB_GREEN  EXT_XBGR_GREEN
+#define RGB_BLUE  EXT_XBGR_BLUE
+#define RGB_ALPHA  0
+#define RGB_PIXELSIZE  EXT_XBGR_PIXELSIZE
+#define ycc_rgb_convert_internal  ycc_extxbgr_convert_internal
+#define gray_rgb_convert_internal  gray_extxbgr_convert_internal
+#define rgb_rgb_convert_internal  rgb_extxbgr_convert_internal
+#include "jdcolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_ALPHA
+#undef RGB_PIXELSIZE
+#undef ycc_rgb_convert_internal
+#undef gray_rgb_convert_internal
+#undef rgb_rgb_convert_internal
+
+#define RGB_RED  EXT_XRGB_RED
+#define RGB_GREEN  EXT_XRGB_GREEN
+#define RGB_BLUE  EXT_XRGB_BLUE
+#define RGB_ALPHA  0
+#define RGB_PIXELSIZE  EXT_XRGB_PIXELSIZE
+#define ycc_rgb_convert_internal  ycc_extxrgb_convert_internal
+#define gray_rgb_convert_internal  gray_extxrgb_convert_internal
+#define rgb_rgb_convert_internal  rgb_extxrgb_convert_internal
+#include "jdcolext.c"
+#undef RGB_RED
+#undef RGB_GREEN
+#undef RGB_BLUE
+#undef RGB_ALPHA
+#undef RGB_PIXELSIZE
+#undef ycc_rgb_convert_internal
+#undef gray_rgb_convert_internal
+#undef rgb_rgb_convert_internal
+
+
+/*
+ * Initialize tables for YCC->RGB colorspace conversion.
+ */
+
+LOCAL(void)
+build_ycc_rgb_table(j_decompress_ptr cinfo)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  int i;
+  JLONG x;
+  SHIFT_TEMPS
+
+  cconvert->Cr_r_tab = (int *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                (MAXJSAMPLE + 1) * sizeof(int));
+  cconvert->Cb_b_tab = (int *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                (MAXJSAMPLE + 1) * sizeof(int));
+  cconvert->Cr_g_tab = (JLONG *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                (MAXJSAMPLE + 1) * sizeof(JLONG));
+  cconvert->Cb_g_tab = (JLONG *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                (MAXJSAMPLE + 1) * sizeof(JLONG));
+
+  for (i = 0, x = -CENTERJSAMPLE; i <= MAXJSAMPLE; i++, x++) {
+    /* i is the actual input pixel value, in the range 0..MAXJSAMPLE */
+    /* The Cb or Cr value we are thinking of is x = i - CENTERJSAMPLE */
+    /* Cr=>R value is nearest int to 1.40200 * x */
+    cconvert->Cr_r_tab[i] = (int)
+                    RIGHT_SHIFT(FIX(1.40200) * x + ONE_HALF, SCALEBITS);
+    /* Cb=>B value is nearest int to 1.77200 * x */
+    cconvert->Cb_b_tab[i] = (int)
+                    RIGHT_SHIFT(FIX(1.77200) * x + ONE_HALF, SCALEBITS);
+    /* Cr=>G value is scaled-up -0.71414 * x */
+    cconvert->Cr_g_tab[i] = (-FIX(0.71414)) * x;
+    /* Cb=>G value is scaled-up -0.34414 * x */
+    /* We also add in ONE_HALF so that need not do it in inner loop */
+    cconvert->Cb_g_tab[i] = (-FIX(0.34414)) * x + ONE_HALF;
+  }
+}
+
+
+/*
+ * Convert some rows of samples to the output colorspace.
+ */
+
+METHODDEF(void)
+ycc_rgb_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  switch (cinfo->out_color_space) {
+  case JCS_EXT_RGB:
+    ycc_extrgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                num_rows);
+    break;
+  case JCS_EXT_RGBX:
+  case JCS_EXT_RGBA:
+    ycc_extrgbx_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_BGR:
+    ycc_extbgr_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                num_rows);
+    break;
+  case JCS_EXT_BGRX:
+  case JCS_EXT_BGRA:
+    ycc_extbgrx_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_XBGR:
+  case JCS_EXT_ABGR:
+    ycc_extxbgr_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_XRGB:
+  case JCS_EXT_ARGB:
+    ycc_extxrgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  default:
+    ycc_rgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                             num_rows);
+    break;
+  }
+}
+
+
+/**************** Cases other than YCbCr -> RGB **************/
+
+
+/*
+ * Initialize for RGB->grayscale colorspace conversion.
+ */
+
+LOCAL(void)
+build_rgb_y_table(j_decompress_ptr cinfo)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  JLONG *rgb_y_tab;
+  JLONG i;
+
+  /* Allocate and fill in the conversion tables. */
+  cconvert->rgb_y_tab = rgb_y_tab = (JLONG *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                (TABLE_SIZE * sizeof(JLONG)));
+
+  for (i = 0; i <= MAXJSAMPLE; i++) {
+    rgb_y_tab[i + R_Y_OFF] = FIX(0.29900) * i;
+    rgb_y_tab[i + G_Y_OFF] = FIX(0.58700) * i;
+    rgb_y_tab[i + B_Y_OFF] = FIX(0.11400) * i + ONE_HALF;
+  }
+}
+
+
+/*
+ * Convert RGB to grayscale.
+ */
+
+METHODDEF(void)
+rgb_gray_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                 JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int r, g, b;
+  register JLONG *ctab = cconvert->rgb_y_tab;
+  register JSAMPROW outptr;
+  register JSAMPROW inptr0, inptr1, inptr2;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+
+  while (--num_rows >= 0) {
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    input_row++;
+    outptr = *output_buf++;
+    for (col = 0; col < num_cols; col++) {
+      r = GETJSAMPLE(inptr0[col]);
+      g = GETJSAMPLE(inptr1[col]);
+      b = GETJSAMPLE(inptr2[col]);
+      /* Y */
+      outptr[col] = (JSAMPLE)((ctab[r + R_Y_OFF] + ctab[g + G_Y_OFF] +
+                               ctab[b + B_Y_OFF]) >> SCALEBITS);
+    }
+  }
+}
+
+
+/*
+ * Color conversion for no colorspace change: just copy the data,
+ * converting from separate-planes to interleaved representation.
+ */
+
+METHODDEF(void)
+null_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+             JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  register JSAMPROW inptr, inptr0, inptr1, inptr2, inptr3, outptr;
+  register JDIMENSION col;
+  register int num_components = cinfo->num_components;
+  JDIMENSION num_cols = cinfo->output_width;
+  int ci;
+
+  if (num_components == 3) {
+    while (--num_rows >= 0) {
+      inptr0 = input_buf[0][input_row];
+      inptr1 = input_buf[1][input_row];
+      inptr2 = input_buf[2][input_row];
+      input_row++;
+      outptr = *output_buf++;
+      for (col = 0; col < num_cols; col++) {
+        *outptr++ = inptr0[col];
+        *outptr++ = inptr1[col];
+        *outptr++ = inptr2[col];
+      }
+    }
+  } else if (num_components == 4) {
+    while (--num_rows >= 0) {
+      inptr0 = input_buf[0][input_row];
+      inptr1 = input_buf[1][input_row];
+      inptr2 = input_buf[2][input_row];
+      inptr3 = input_buf[3][input_row];
+      input_row++;
+      outptr = *output_buf++;
+      for (col = 0; col < num_cols; col++) {
+        *outptr++ = inptr0[col];
+        *outptr++ = inptr1[col];
+        *outptr++ = inptr2[col];
+        *outptr++ = inptr3[col];
+      }
+    }
+  } else {
+    while (--num_rows >= 0) {
+      for (ci = 0; ci < num_components; ci++) {
+        inptr = input_buf[ci][input_row];
+        outptr = *output_buf;
+        for (col = 0; col < num_cols; col++) {
+          outptr[ci] = inptr[col];
+          outptr += num_components;
+        }
+      }
+      output_buf++;
+      input_row++;
+    }
+  }
+}
+
+
+/*
+ * Color conversion for grayscale: just copy the data.
+ * This also works for YCbCr -> grayscale conversion, in which
+ * we just copy the Y (luminance) component and ignore chrominance.
+ */
+
+METHODDEF(void)
+grayscale_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                  JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  jcopy_sample_rows(input_buf[0], (int)input_row, output_buf, 0, num_rows,
+                    cinfo->output_width);
+}
+
+
+/*
+ * Convert grayscale to RGB
+ */
+
+METHODDEF(void)
+gray_rgb_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                 JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  switch (cinfo->out_color_space) {
+  case JCS_EXT_RGB:
+    gray_extrgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_RGBX:
+  case JCS_EXT_RGBA:
+    gray_extrgbx_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                  num_rows);
+    break;
+  case JCS_EXT_BGR:
+    gray_extbgr_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_BGRX:
+  case JCS_EXT_BGRA:
+    gray_extbgrx_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                  num_rows);
+    break;
+  case JCS_EXT_XBGR:
+  case JCS_EXT_ABGR:
+    gray_extxbgr_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                  num_rows);
+    break;
+  case JCS_EXT_XRGB:
+  case JCS_EXT_ARGB:
+    gray_extxrgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                  num_rows);
+    break;
+  default:
+    gray_rgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                              num_rows);
+    break;
+  }
+}
+
+
+/*
+ * Convert plain RGB to extended RGB
+ */
+
+METHODDEF(void)
+rgb_rgb_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  switch (cinfo->out_color_space) {
+  case JCS_EXT_RGB:
+    rgb_extrgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                num_rows);
+    break;
+  case JCS_EXT_RGBX:
+  case JCS_EXT_RGBA:
+    rgb_extrgbx_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_BGR:
+    rgb_extbgr_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                num_rows);
+    break;
+  case JCS_EXT_BGRX:
+  case JCS_EXT_BGRA:
+    rgb_extbgrx_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_XBGR:
+  case JCS_EXT_ABGR:
+    rgb_extxbgr_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  case JCS_EXT_XRGB:
+  case JCS_EXT_ARGB:
+    rgb_extxrgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                                 num_rows);
+    break;
+  default:
+    rgb_rgb_convert_internal(cinfo, input_buf, input_row, output_buf,
+                             num_rows);
+    break;
+  }
+}
+
+
+/*
+ * Adobe-style YCCK->CMYK conversion.
+ * We convert YCbCr to R=1-C, G=1-M, and B=1-Y using the same
+ * conversion as above, while passing K (black) unchanged.
+ * We assume build_ycc_rgb_table has been called.
+ */
+
+METHODDEF(void)
+ycck_cmyk_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                  JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  my_cconvert_ptr cconvert = (my_cconvert_ptr)cinfo->cconvert;
+  register int y, cb, cr;
+  register JSAMPROW outptr;
+  register JSAMPROW inptr0, inptr1, inptr2, inptr3;
+  register JDIMENSION col;
+  JDIMENSION num_cols = cinfo->output_width;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE *range_limit = cinfo->sample_range_limit;
+  register int *Crrtab = cconvert->Cr_r_tab;
+  register int *Cbbtab = cconvert->Cb_b_tab;
+  register JLONG *Crgtab = cconvert->Cr_g_tab;
+  register JLONG *Cbgtab = cconvert->Cb_g_tab;
+  SHIFT_TEMPS
+
+  while (--num_rows >= 0) {
+    inptr0 = input_buf[0][input_row];
+    inptr1 = input_buf[1][input_row];
+    inptr2 = input_buf[2][input_row];
+    inptr3 = input_buf[3][input_row];
+    input_row++;
+    outptr = *output_buf++;
+    for (col = 0; col < num_cols; col++) {
+      y  = GETJSAMPLE(inptr0[col]);
+      cb = GETJSAMPLE(inptr1[col]);
+      cr = GETJSAMPLE(inptr2[col]);
+      /* Range-limiting is essential due to noise introduced by DCT losses. */
+      outptr[0] = range_limit[MAXJSAMPLE - (y + Crrtab[cr])];   /* red */
+      outptr[1] = range_limit[MAXJSAMPLE - (y +                 /* green */
+                              ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
+                                                 SCALEBITS)))];
+      outptr[2] = range_limit[MAXJSAMPLE - (y + Cbbtab[cb])];   /* blue */
+      /* K passes through unchanged */
+      outptr[3] = inptr3[col];  /* don't need GETJSAMPLE here */
+      outptr += 4;
+    }
+  }
+}
+
+
+/*
+ * RGB565 conversion
+ */
+
+#define PACK_SHORT_565_LE(r, g, b)  ((((r) << 8) & 0xF800) | \
+                                     (((g) << 3) & 0x7E0) | ((b) >> 3))
+#define PACK_SHORT_565_BE(r, g, b)  (((r) & 0xF8) | ((g) >> 5) | \
+                                     (((g) << 11) & 0xE000) | \
+                                     (((b) << 5) & 0x1F00))
+
+#define PACK_TWO_PIXELS_LE(l, r)    ((r << 16) | l)
+#define PACK_TWO_PIXELS_BE(l, r)    ((l << 16) | r)
+
+#define PACK_NEED_ALIGNMENT(ptr)    (((size_t)(ptr)) & 3)
+
+#define WRITE_TWO_ALIGNED_PIXELS(addr, pixels)  ((*(int *)(addr)) = pixels)
+
+#define DITHER_565_R(r, dither)  ((r) + ((dither) & 0xFF))
+#define DITHER_565_G(g, dither)  ((g) + (((dither) & 0xFF) >> 1))
+#define DITHER_565_B(b, dither)  ((b) + ((dither) & 0xFF))
+
+
+/* Declarations for ordered dithering
+ *
+ * We use a 4x4 ordered dither array packed into 32 bits.  This array is
+ * sufficient for dithering RGB888 to RGB565.
+ */
+
+#define DITHER_MASK       0x3
+#define DITHER_ROTATE(x)  ((((x) & 0xFF) << 24) | (((x) >> 8) & 0x00FFFFFF))
+static const JLONG dither_matrix[4] = {
+  0x0008020A,
+  0x0C040E06,
+  0x030B0109,
+  0x0F070D05
+};
+
+
+static INLINE boolean is_big_endian(void)
+{
+  int test_value = 1;
+  if (*(char *)&test_value != 1)
+    return TRUE;
+  return FALSE;
+}
+
+
+/* Include inline routines for RGB565 conversion */
+
+#define PACK_SHORT_565  PACK_SHORT_565_LE
+#define PACK_TWO_PIXELS  PACK_TWO_PIXELS_LE
+#define ycc_rgb565_convert_internal  ycc_rgb565_convert_le
+#define ycc_rgb565D_convert_internal  ycc_rgb565D_convert_le
+#define rgb_rgb565_convert_internal  rgb_rgb565_convert_le
+#define rgb_rgb565D_convert_internal  rgb_rgb565D_convert_le
+#define gray_rgb565_convert_internal  gray_rgb565_convert_le
+#define gray_rgb565D_convert_internal  gray_rgb565D_convert_le
+#include "jdcol565.c"
+#undef PACK_SHORT_565
+#undef PACK_TWO_PIXELS
+#undef ycc_rgb565_convert_internal
+#undef ycc_rgb565D_convert_internal
+#undef rgb_rgb565_convert_internal
+#undef rgb_rgb565D_convert_internal
+#undef gray_rgb565_convert_internal
+#undef gray_rgb565D_convert_internal
+
+#define PACK_SHORT_565  PACK_SHORT_565_BE
+#define PACK_TWO_PIXELS  PACK_TWO_PIXELS_BE
+#define ycc_rgb565_convert_internal  ycc_rgb565_convert_be
+#define ycc_rgb565D_convert_internal  ycc_rgb565D_convert_be
+#define rgb_rgb565_convert_internal  rgb_rgb565_convert_be
+#define rgb_rgb565D_convert_internal  rgb_rgb565D_convert_be
+#define gray_rgb565_convert_internal  gray_rgb565_convert_be
+#define gray_rgb565D_convert_internal  gray_rgb565D_convert_be
+#include "jdcol565.c"
+#undef PACK_SHORT_565
+#undef PACK_TWO_PIXELS
+#undef ycc_rgb565_convert_internal
+#undef ycc_rgb565D_convert_internal
+#undef rgb_rgb565_convert_internal
+#undef rgb_rgb565D_convert_internal
+#undef gray_rgb565_convert_internal
+#undef gray_rgb565D_convert_internal
+
+
+METHODDEF(void)
+ycc_rgb565_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                   JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  if (is_big_endian())
+    ycc_rgb565_convert_be(cinfo, input_buf, input_row, output_buf, num_rows);
+  else
+    ycc_rgb565_convert_le(cinfo, input_buf, input_row, output_buf, num_rows);
+}
+
+
+METHODDEF(void)
+ycc_rgb565D_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                    JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  if (is_big_endian())
+    ycc_rgb565D_convert_be(cinfo, input_buf, input_row, output_buf, num_rows);
+  else
+    ycc_rgb565D_convert_le(cinfo, input_buf, input_row, output_buf, num_rows);
+}
+
+
+METHODDEF(void)
+rgb_rgb565_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                   JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  if (is_big_endian())
+    rgb_rgb565_convert_be(cinfo, input_buf, input_row, output_buf, num_rows);
+  else
+    rgb_rgb565_convert_le(cinfo, input_buf, input_row, output_buf, num_rows);
+}
+
+
+METHODDEF(void)
+rgb_rgb565D_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                    JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  if (is_big_endian())
+    rgb_rgb565D_convert_be(cinfo, input_buf, input_row, output_buf, num_rows);
+  else
+    rgb_rgb565D_convert_le(cinfo, input_buf, input_row, output_buf, num_rows);
+}
+
+
+METHODDEF(void)
+gray_rgb565_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                    JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  if (is_big_endian())
+    gray_rgb565_convert_be(cinfo, input_buf, input_row, output_buf, num_rows);
+  else
+    gray_rgb565_convert_le(cinfo, input_buf, input_row, output_buf, num_rows);
+}
+
+
+METHODDEF(void)
+gray_rgb565D_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
+                     JDIMENSION input_row, JSAMPARRAY output_buf, int num_rows)
+{
+  if (is_big_endian())
+    gray_rgb565D_convert_be(cinfo, input_buf, input_row, output_buf, num_rows);
+  else
+    gray_rgb565D_convert_le(cinfo, input_buf, input_row, output_buf, num_rows);
+}
+
+
+/*
+ * Empty method for start_pass.
+ */
+
+METHODDEF(void)
+start_pass_dcolor(j_decompress_ptr cinfo)
+{
+  /* no work needed */
+}
+
+
+/*
+ * Module initialization routine for output colorspace conversion.
+ */
+
+GLOBAL(void)
+jinit_color_deconverter(j_decompress_ptr cinfo)
+{
+  my_cconvert_ptr cconvert;
+  int ci;
+
+  cconvert = (my_cconvert_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
+                                sizeof(my_color_deconverter));
+  cinfo->cconvert = (struct jpeg_color_deconverter *)cconvert;
+  cconvert->pub.start_pass = start_pass_dcolor;
+
+  /* Make sure num_components agrees with jpeg_color_space */
+  switch (cinfo->jpeg_color_space) {
+  case JCS_GRAYSCALE:
+    if (cinfo->num_components != 1)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    break;
+
+  case JCS_RGB:
+  case JCS_YCbCr:
+    if (cinfo->num_components != 3)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    break;
+
+  case JCS_CMYK:
+  case JCS_YCCK:
+    if (cinfo->num_components != 4)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    break;
+
+  default:                      /* JCS_UNKNOWN can be anything */
+    if (cinfo->num_components < 1)
+      ERREXIT(cinfo, JERR_BAD_J_COLORSPACE);
+    break;
+  }
+
+  /* Set out_color_components and conversion method based on requested space.
+   * Also clear the component_needed flags for any unused components,
+   * so that earlier pipeline stages can avoid useless computation.
+   */
+
+  switch (cinfo->out_color_space) {
+  case JCS_GRAYSCALE:
+    cinfo->out_color_components = 1;
+    if (cinfo->jpeg_color_space == JCS_GRAYSCALE ||
+        cinfo->jpeg_color_space == JCS_YCbCr) {
+      cconvert->pub.color_convert = grayscale_convert;
+      /* For color->grayscale conversion, only the Y (0) component is needed */
+      for (ci = 1; ci < cinfo->num_components; ci++)
+        cinfo->comp_info[ci].component_needed = FALSE;
+    } else if (cinfo->jpeg_color_space == JCS_RGB) {
+      cconvert->pub.color_convert = rgb_gray_convert;
+      build_rgb_y_table(cinfo);
+    } else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  case JCS_RGB:
+  case JCS_EXT_RGB:
+  case JCS_EXT_RGBX:
+  case JCS_EXT_BGR:
+  case JCS_EXT_BGRX:
+  case JCS_EXT_XBGR:
+  case JCS_EXT_XRGB:
+  case JCS_EXT_RGBA:
+  case JCS_EXT_BGRA:
+  case JCS_EXT_ABGR:
+  case JCS_EXT_ARGB:
+    cinfo->out_color_components = rgb_pixelsize[cinfo->out_color_space];
+    if (cinfo->jpeg_color_space == JCS_YCbCr) {
+      if (jsimd_can_ycc_rgb())
+        cconvert->pub.color_convert = jsimd_ycc_rgb_convert;
+      else {
+        cconvert->pub.color_convert = ycc_rgb_convert;
+        build_ycc_rgb_table(cinfo);
+      }
+    } else if (cinfo->jpeg_color_space == JCS_GRAYSCALE) {
+      cconvert->pub.color_convert = gray_rgb_convert;
+    } else if (cinfo->jpeg_color_space == JCS_RGB) {
+      if (rgb_red[cinfo->out_color_space] == 0 &&
+          rgb_green[cinfo->out_color_space] == 1 &&
+          rgb_blue[cinfo->out_color_space] == 2 &&
+          rgb_pixelsize[cinfo->out_color_space] == 3)
+        cconvert->pub.color_convert = null_convert;
+      else
+        cconvert->pub.color_convert = rgb_rgb_convert;
+    } else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  case JCS_RGB565:
+    cinfo->out_color_components = 3;
+    if (cinfo->dither_mode == JDITHER_NONE) {
+      if (cinfo->jpeg_color_space == JCS_YCbCr) {
+        if (jsimd_can_ycc_rgb565())
+          cconvert->pub.color_convert = jsimd_ycc_rgb565_convert;
+        else {
+          cconvert->pub.color_convert = ycc_rgb565_convert;
+          build_ycc_rgb_table(cinfo);
+        }
+      } else if (cinfo->jpeg_color_space == JCS_GRAYSCALE) {
+        cconvert->pub.color_convert = gray_rgb565_convert;
+      } else if (cinfo->jpeg_color_space == JCS_RGB) {
+        cconvert->pub.color_convert = rgb_rgb565_convert;
+      } else
+        ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    } else {
+      /* only ordered dithering is supported */
+      if (cinfo->jpeg_color_space == JCS_YCbCr) {
+        cconvert->pub.color_convert = ycc_rgb565D_convert;
+        build_ycc_rgb_table(cinfo);
+      } else if (cinfo->jpeg_color_space == JCS_GRAYSCALE) {
+        cconvert->pub.color_convert = gray_rgb565D_convert;
+      } else if (cinfo->jpeg_color_space == JCS_RGB) {
+        cconvert->pub.color_convert = rgb_rgb565D_convert;
+      } else
+        ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    }
+    break;
+
+  case JCS_CMYK:
+    cinfo->out_color_components = 4;
+    if (cinfo->jpeg_color_space == JCS_YCCK) {
+      cconvert->pub.color_convert = ycck_cmyk_convert;
+      build_ycc_rgb_table(cinfo);
+    } else if (cinfo->jpeg_color_space == JCS_CMYK) {
+      cconvert->pub.color_convert = null_convert;
+    } else
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+
+  default:
+    /* Permit null conversion to same output space */
+    if (cinfo->out_color_space == cinfo->jpeg_color_space) {
+      cinfo->out_color_components = cinfo->num_components;
+      cconvert->pub.color_convert = null_convert;
+    } else                      /* unsupported non-null conversion */
+      ERREXIT(cinfo, JERR_CONVERSION_NOTIMPL);
+    break;
+  }
+
+  if (cinfo->quantize_colors)
+    cinfo->output_components = 1; /* single colormapped output component */
+  else
+    cinfo->output_components = cinfo->out_color_components;
+}

3rdparty/libjpeg-turbo/src/jdct.h

@@ -0,0 +1,208 @@
+/*
+ * jdct.h
+ *
+ * This file was part of the Independent JPEG Group's software:
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * libjpeg-turbo Modifications:
+ * Copyright (C) 2015, D. R. Commander.
+ * For conditions of distribution and use, see the accompanying README.ijg
+ * file.
+ *
+ * This include file contains common declarations for the forward and
+ * inverse DCT modules.  These declarations are private to the DCT managers
+ * (jcdctmgr.c, jddctmgr.c) and the individual DCT algorithms.
+ * The individual DCT algorithms are kept in separate files to ease
+ * machine-dependent tuning (e.g., assembly coding).
+ */
+
+
+/*
+ * A forward DCT routine is given a pointer to a work area of type DCTELEM[];
+ * the DCT is to be performed in-place in that buffer.  Type DCTELEM is int
+ * for 8-bit samples, JLONG for 12-bit samples.  (NOTE: Floating-point DCT
+ * implementations use an array of type FAST_FLOAT, instead.)
+ * The DCT inputs are expected to be signed (range +-CENTERJSAMPLE).
+ * The DCT outputs are returned scaled up by a factor of 8; they therefore
+ * have a range of +-8K for 8-bit data, +-128K for 12-bit data.  This
+ * convention improves accuracy in integer implementations and saves some
+ * work in floating-point ones.
+ * Quantization of the output coefficients is done by jcdctmgr.c. This
+ * step requires an unsigned type and also one with twice the bits.
+ */
+
+#if BITS_IN_JSAMPLE == 8
+#ifndef WITH_SIMD
+typedef int DCTELEM;            /* 16 or 32 bits is fine */
+typedef unsigned int UDCTELEM;
+typedef unsigned long long UDCTELEM2;
+#else
+typedef short DCTELEM;          /* prefer 16 bit with SIMD for parellelism */
+typedef unsigned short UDCTELEM;
+typedef unsigned int UDCTELEM2;
+#endif
+#else
+typedef JLONG DCTELEM;          /* must have 32 bits */
+typedef unsigned long long UDCTELEM2;
+#endif
+
+
+/*
+ * An inverse DCT routine is given a pointer to the input JBLOCK and a pointer
+ * to an output sample array.  The routine must dequantize the input data as
+ * well as perform the IDCT; for dequantization, it uses the multiplier table
+ * pointed to by compptr->dct_table.  The output data is to be placed into the
+ * sample array starting at a specified column.  (Any row offset needed will
+ * be applied to the array pointer before it is passed to the IDCT code.)
+ * Note that the number of samples emitted by the IDCT routine is
+ * DCT_scaled_size * DCT_scaled_size.
+ */
+
+/* typedef inverse_DCT_method_ptr is declared in jpegint.h */
+
+/*
+ * Each IDCT routine has its own ideas about the best dct_table element type.
+ */
+
+typedef MULTIPLIER ISLOW_MULT_TYPE;  /* short or int, whichever is faster */
+#if BITS_IN_JSAMPLE == 8
+typedef MULTIPLIER IFAST_MULT_TYPE;  /* 16 bits is OK, use short if faster */
+#define IFAST_SCALE_BITS  2          /* fractional bits in scale factors */
+#else
+typedef JLONG IFAST_MULT_TYPE;       /* need 32 bits for scaled quantizers */
+#define IFAST_SCALE_BITS  13         /* fractional bits in scale factors */
+#endif
+typedef FAST_FLOAT FLOAT_MULT_TYPE;  /* preferred floating type */
+
+
+/*
+ * Each IDCT routine is responsible for range-limiting its results and
+ * converting them to unsigned form (0..MAXJSAMPLE).  The raw outputs could
+ * be quite far out of range if the input data is corrupt, so a bulletproof
+ * range-limiting step is required.  We use a mask-and-table-lookup method
+ * to do the combined operations quickly.  See the comments with
+ * prepare_range_limit_table (in jdmaster.c) for more info.
+ */
+
+#define IDCT_range_limit(cinfo)  ((cinfo)->sample_range_limit + CENTERJSAMPLE)
+
+#define RANGE_MASK  (MAXJSAMPLE * 4 + 3) /* 2 bits wider than legal samples */
+
+
+/* Extern declarations for the forward and inverse DCT routines. */
+
+EXTERN(void) jpeg_fdct_islow(DCTELEM *data);
+EXTERN(void) jpeg_fdct_ifast(DCTELEM *data);
+EXTERN(void) jpeg_fdct_float(FAST_FLOAT *data);
+
+EXTERN(void) jpeg_idct_islow(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_ifast(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_float(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_7x7(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_6x6(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_5x5(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_4x4(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_3x3(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_2x2(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_1x1(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_9x9(j_decompress_ptr cinfo,
+                           jpeg_component_info *compptr, JCOEFPTR coef_block,
+                           JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_10x10(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_11x11(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_12x12(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_13x13(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_14x14(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_15x15(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+EXTERN(void) jpeg_idct_16x16(j_decompress_ptr cinfo,
+                             jpeg_component_info *compptr, JCOEFPTR coef_block,
+                             JSAMPARRAY output_buf, JDIMENSION output_col);
+
+
+/*
+ * Macros for handling fixed-point arithmetic; these are used by many
+ * but not all of the DCT/IDCT modules.
+ *
+ * All values are expected to be of type JLONG.
+ * Fractional constants are scaled left by CONST_BITS bits.
+ * CONST_BITS is defined within each module using these macros,
+ * and may differ from one module to the next.
+ */
+
+#define ONE          ((JLONG)1)
+#define CONST_SCALE  (ONE << CONST_BITS)
+
+/* Convert a positive real constant to an integer scaled by CONST_SCALE.
+ * Caution: some C compilers fail to reduce "FIX(constant)" at compile time,
+ * thus causing a lot of useless floating-point operations at run time.
+ */
+
+#define FIX(x)  ((JLONG)((x) * CONST_SCALE + 0.5))
+
+/* Descale and correctly round a JLONG value that's scaled by N bits.
+ * We assume RIGHT_SHIFT rounds towards minus infinity, so adding
+ * the fudge factor is correct for either sign of X.
+ */
+
+#define DESCALE(x, n)  RIGHT_SHIFT((x) + (ONE << ((n) - 1)), n)
+
+/* Multiply a JLONG variable by a JLONG constant to yield a JLONG result.
+ * This macro is used only when the two inputs will actually be no more than
+ * 16 bits wide, so that a 16x16->32 bit multiply can be used instead of a
+ * full 32x32 multiply.  This provides a useful speedup on many machines.
+ * Unfortunately there is no way to specify a 16x16->32 multiply portably
+ * in C, but some C compilers will do the right thing if you provide the
+ * correct combination of casts.
+ */
+
+#ifdef SHORTxSHORT_32           /* may work if 'int' is 32 bits */
+#define MULTIPLY16C16(var, const)  (((INT16)(var)) * ((INT16)(const)))
+#endif
+#ifdef SHORTxLCONST_32          /* known to work with Microsoft C 6.0 */
+#define MULTIPLY16C16(var, const)  (((INT16)(var)) * ((JLONG)(const)))
+#endif
+
+#ifndef MULTIPLY16C16           /* default definition */
+#define MULTIPLY16C16(var, const)  ((var) * (const))
+#endif
+
+/* Same except both inputs are variables. */
+
+#ifdef SHORTxSHORT_32           /* may work if 'int' is 32 bits */
+#define MULTIPLY16V16(var1, var2)  (((INT16)(var1)) * ((INT16)(var2)))
+#endif
+
+#ifndef MULTIPLY16V16           /* default definition */
+#define MULTIPLY16V16(var1, var2)  ((var1) * (var2))
+#endif