release: OpenCV 3.4.19

Merge pull request #23029 from savuor:backport3_fix_fisheye_aspect_ratio
backported changes
2026-07-21 19:33:03 +04:00 · 2022-12-27 03:50:12 +00:00 · 2022-12-27 03:47:01 +00:00 · 2022-12-27 00:54:39 +01:00 · 2022-12-25 18:39:54 +00:00 · 2022-12-25 16:59:47 +00:00
673 changed files with 41166 additions and 12696 deletions
@@ -0,0 +1,34 @@
+name: PR:3.4
+
+on:
+  pull_request:
+    branches:
+      - 3.4
+
+jobs:
+  Ubuntu2004-ARM64:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-ARM64.yaml@main
+
+  Ubuntu1404-x64:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-U14.yaml@main
+
+  Ubuntu2004-x64:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-U20.yaml@main
+
+  Windows10-x64:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-W10.yaml@main
+
+  macOS-ARM64:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-macOS-ARM64.yaml@main
+
+  macOS-x64:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-macOS-x86_64.yaml@main
+
+  iOS:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-iOS.yaml@main
+
+  Android:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-Android.yaml@main
+
+  docs:
+    uses: opencv/ci-gha-workflow/.github/workflows/OCV-PR-3.4-docs.yaml@main
@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 2.8.11 FATAL_ERROR)
+cmake_minimum_required(VERSION ${MIN_VER_CMAKE} FATAL_ERROR)

 project(Carotene)

@@ -27,6 +27,10 @@ if(CMAKE_COMPILER_IS_GNUCC)
    endif()
 endif()

+if(APPLE AND CV_CLANG AND WITH_NEON)
+    ocv_warnings_disable(CMAKE_CXX_FLAGS -Wno-unused-function)
+endif()
+
 add_library(carotene_objs OBJECT EXCLUDE_FROM_ALL
  ${carotene_headers}
  ${carotene_sources}
@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 2.8.8 FATAL_ERROR)
+cmake_minimum_required(VERSION ${MIN_VER_CMAKE} FATAL_ERROR)

 include(CheckCCompilerFlag)
 include(CheckCXXCompilerFlag)
@@ -1296,13 +1296,13 @@ struct MorphCtx
    CAROTENE_NS::BORDER_MODE border;
    uchar borderValues[4];
 };
-inline int TEGRA_MORPHINIT(cvhalFilter2D **context, int operation, int src_type, int dst_type, int, int,
+inline int TEGRA_MORPHINIT(cvhalFilter2D **context, int operation, int src_type, int dst_type, int width, int height,
                           int kernel_type, uchar *kernel_data, size_t kernel_step, int kernel_width, int kernel_height, int anchor_x, int anchor_y,
                           int borderType, const double borderValue[4], int iterations, bool allowSubmatrix, bool allowInplace)
 {
    if(!context || !kernel_data || src_type != dst_type ||
       CV_MAT_DEPTH(src_type) != CV_8U || src_type < 0 || (src_type >> CV_CN_SHIFT) > 3 ||
-
+       width < kernel_width || height < kernel_height ||
       allowSubmatrix || allowInplace || iterations != 1 ||
       !CAROTENE_NS::isSupportedConfiguration())
        return CV_HAL_ERROR_NOT_IMPLEMENTED;
@@ -1778,30 +1778,30 @@ TegraCvtColor_Invoker(bgrx2hsvf, bgrx2hsv, src_data + static_cast<size_t>(range.
    : CV_HAL_ERROR_NOT_IMPLEMENTED \
 )

-#define TEGRA_CVT2PYUVTOBGR(src_data, src_step, dst_data, dst_step, dst_width, dst_height, dcn, swapBlue, uIdx) \
+#define TEGRA_CVT2PYUVTOBGR_EX(y_data, y_step, uv_data, uv_step, dst_data, dst_step, dst_width, dst_height, dcn, swapBlue, uIdx) \
 ( \
    CAROTENE_NS::isSupportedConfiguration() ? \
        dcn == 3 ? \
            uIdx == 0 ? \
                (swapBlue ? \
                    CAROTENE_NS::yuv420i2rgb(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                             src_data, src_step, \
-                                             src_data + src_step * dst_height, src_step, \
+                                             y_data, y_step, \
+                                             uv_data, uv_step, \
                                             dst_data, dst_step) : \
                    CAROTENE_NS::yuv420i2bgr(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                             src_data, src_step, \
-                                             src_data + src_step * dst_height, src_step, \
+                                             y_data, y_step, \
+                                             uv_data, uv_step, \
                                             dst_data, dst_step)), \
                CV_HAL_ERROR_OK : \
            uIdx == 1 ? \
                (swapBlue ? \
                    CAROTENE_NS::yuv420sp2rgb(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                              src_data, src_step, \
-                                              src_data + src_step * dst_height, src_step, \
+                                              y_data, y_step, \
+                                              uv_data, uv_step, \
                                              dst_data, dst_step) : \
                    CAROTENE_NS::yuv420sp2bgr(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                              src_data, src_step, \
-                                              src_data + src_step * dst_height, src_step, \
+                                              y_data, y_step, \
+                                              uv_data, uv_step, \
                                              dst_data, dst_step)), \
                CV_HAL_ERROR_OK : \
            CV_HAL_ERROR_NOT_IMPLEMENTED : \
@@ -1809,29 +1809,32 @@ TegraCvtColor_Invoker(bgrx2hsvf, bgrx2hsv, src_data + static_cast<size_t>(range.
            uIdx == 0 ? \
                (swapBlue ? \
                    CAROTENE_NS::yuv420i2rgbx(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                              src_data, src_step, \
-                                              src_data + src_step * dst_height, src_step, \
+                                              y_data, y_step, \
+                                              uv_data, uv_step, \
                                              dst_data, dst_step) : \
                    CAROTENE_NS::yuv420i2bgrx(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                              src_data, src_step, \
-                                              src_data + src_step * dst_height, src_step, \
+                                              y_data, y_step, \
+                                              uv_data, uv_step, \
                                              dst_data, dst_step)), \
                CV_HAL_ERROR_OK : \
            uIdx == 1 ? \
                (swapBlue ? \
                    CAROTENE_NS::yuv420sp2rgbx(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                               src_data, src_step, \
-                                               src_data + src_step * dst_height, src_step, \
+                                               y_data, y_step, \
+                                               uv_data, uv_step, \
                                               dst_data, dst_step) : \
                    CAROTENE_NS::yuv420sp2bgrx(CAROTENE_NS::Size2D(dst_width, dst_height), \
-                                               src_data, src_step, \
-                                               src_data + src_step * dst_height, src_step, \
+                                               y_data, y_step, \
+                                               uv_data, uv_step, \
                                               dst_data, dst_step)), \
                CV_HAL_ERROR_OK : \
            CV_HAL_ERROR_NOT_IMPLEMENTED : \
        CV_HAL_ERROR_NOT_IMPLEMENTED \
    : CV_HAL_ERROR_NOT_IMPLEMENTED \
 )
+#define TEGRA_CVT2PYUVTOBGR(src_data, src_step, dst_data, dst_step, dst_width, dst_height, dcn, swapBlue, uIdx) \
+    TEGRA_CVT2PYUVTOBGR_EX(src_data, src_step, src_data + src_step * dst_height, src_step, dst_data, dst_step, \
+            dst_width, dst_height, dcn, swapBlue, uIdx);

 #undef cv_hal_cvtBGRtoBGR
 #define cv_hal_cvtBGRtoBGR TEGRA_CVTBGRTOBGR
@@ -1841,12 +1844,18 @@ TegraCvtColor_Invoker(bgrx2hsvf, bgrx2hsv, src_data + static_cast<size_t>(range.
 #define cv_hal_cvtBGRtoGray TEGRA_CVTBGRTOGRAY
 #undef cv_hal_cvtGraytoBGR
 #define cv_hal_cvtGraytoBGR TEGRA_CVTGRAYTOBGR
+#if 0  // bit-exact tests are failed
 #undef cv_hal_cvtBGRtoYUV
 #define cv_hal_cvtBGRtoYUV TEGRA_CVTBGRTOYUV
+#endif
 #undef cv_hal_cvtBGRtoHSV
 #define cv_hal_cvtBGRtoHSV TEGRA_CVTBGRTOHSV
+#if 0  // bit-exact tests are failed
 #undef cv_hal_cvtTwoPlaneYUVtoBGR
 #define cv_hal_cvtTwoPlaneYUVtoBGR TEGRA_CVT2PYUVTOBGR
+#undef cv_hal_cvtTwoPlaneYUVtoBGREx
+#define cv_hal_cvtTwoPlaneYUVtoBGREx TEGRA_CVT2PYUVTOBGR_EX
+#endif

 #endif // OPENCV_IMGPROC_HAL_INTERFACE_H

@@ -109,9 +109,9 @@ template <> struct wAdd<s32>
        vgamma = vdupq_n_f32(_gamma + 0.5);
    }

-    void operator() (const typename VecTraits<s32>::vec128 & v_src0,
-                     const typename VecTraits<s32>::vec128 & v_src1,
-                     typename VecTraits<s32>::vec128 & v_dst) const
+    void operator() (const VecTraits<s32>::vec128 & v_src0,
+                     const VecTraits<s32>::vec128 & v_src1,
+                     VecTraits<s32>::vec128 & v_dst) const
    {
        float32x4_t vs1 = vcvtq_f32_s32(v_src0);
        float32x4_t vs2 = vcvtq_f32_s32(v_src1);
@@ -121,9 +121,9 @@ template <> struct wAdd<s32>
        v_dst = vcvtq_s32_f32(vs1);
    }

-    void operator() (const typename VecTraits<s32>::vec64 & v_src0,
-                     const typename VecTraits<s32>::vec64 & v_src1,
-                     typename VecTraits<s32>::vec64 & v_dst) const
+    void operator() (const VecTraits<s32>::vec64 & v_src0,
+                     const VecTraits<s32>::vec64 & v_src1,
+                     VecTraits<s32>::vec64 & v_dst) const
    {
        float32x2_t vs1 = vcvt_f32_s32(v_src0);
        float32x2_t vs2 = vcvt_f32_s32(v_src1);
@@ -153,9 +153,9 @@ template <> struct wAdd<u32>
        vgamma = vdupq_n_f32(_gamma + 0.5);
    }

-    void operator() (const typename VecTraits<u32>::vec128 & v_src0,
-                     const typename VecTraits<u32>::vec128 & v_src1,
-                     typename VecTraits<u32>::vec128 & v_dst) const
+    void operator() (const VecTraits<u32>::vec128 & v_src0,
+                     const VecTraits<u32>::vec128 & v_src1,
+                     VecTraits<u32>::vec128 & v_dst) const
    {
        float32x4_t vs1 = vcvtq_f32_u32(v_src0);
        float32x4_t vs2 = vcvtq_f32_u32(v_src1);
@@ -165,9 +165,9 @@ template <> struct wAdd<u32>
        v_dst = vcvtq_u32_f32(vs1);
    }

-    void operator() (const typename VecTraits<u32>::vec64 & v_src0,
-                     const typename VecTraits<u32>::vec64 & v_src1,
-                     typename VecTraits<u32>::vec64 & v_dst) const
+    void operator() (const VecTraits<u32>::vec64 & v_src0,
+                     const VecTraits<u32>::vec64 & v_src1,
+                     VecTraits<u32>::vec64 & v_dst) const
    {
        float32x2_t vs1 = vcvt_f32_u32(v_src0);
        float32x2_t vs2 = vcvt_f32_u32(v_src1);
@@ -197,17 +197,17 @@ template <> struct wAdd<f32>
        vgamma = vdupq_n_f32(_gamma + 0.5);
    }

-    void operator() (const typename VecTraits<f32>::vec128 & v_src0,
-                     const typename VecTraits<f32>::vec128 & v_src1,
-                     typename VecTraits<f32>::vec128 & v_dst) const
+    void operator() (const VecTraits<f32>::vec128 & v_src0,
+                     const VecTraits<f32>::vec128 & v_src1,
+                     VecTraits<f32>::vec128 & v_dst) const
    {
        float32x4_t vs1 = vmlaq_f32(vgamma, v_src0, valpha);
        v_dst = vmlaq_f32(vs1, v_src1, vbeta);
    }

-    void operator() (const typename VecTraits<f32>::vec64 & v_src0,
-                     const typename VecTraits<f32>::vec64 & v_src1,
-                     typename VecTraits<f32>::vec64 & v_dst) const
+    void operator() (const VecTraits<f32>::vec64 & v_src0,
+                     const VecTraits<f32>::vec64 & v_src1,
+                     VecTraits<f32>::vec64 & v_dst) const
    {
        float32x2_t vs1 = vmla_f32(vget_low(vgamma), v_src0, vget_low(valpha));
        v_dst = vmla_f32(vs1, v_src1, vget_low(vbeta));
@@ -1,8 +1,8 @@
-# Binaries branch name: ffmpeg/3.4_20210302
-# Binaries were created for OpenCV: 2ab1f3f166fccc3a01497209cc01c5cea44ff201
-ocv_update(FFMPEG_BINARIES_COMMIT "e99214251d9f3cde7c48abd46b2259bddc9885b6")
-ocv_update(FFMPEG_FILE_HASH_BIN32 "fad5ada9be36120bba8966709e7953a8")
-ocv_update(FFMPEG_FILE_HASH_BIN64 "650e2272728491923e566f784f79cfef")
+# Binaries branch name: ffmpeg/3.4_20211220
+# Binaries were created for OpenCV: a22dd28e0272ec0f1cfee8811d3f5f0392827c65
+ocv_update(FFMPEG_BINARIES_COMMIT "5a7644ec3940c6eed41c6ebb5a0602a5615fdb3f")
+ocv_update(FFMPEG_FILE_HASH_BIN32 "8ad9de6f1f2ca77786748d1f3a4e83ea")
+ocv_update(FFMPEG_FILE_HASH_BIN64 "2c670f068252e7cd28d3883993dc1d6e")
 ocv_update(FFMPEG_FILE_HASH_CMAKE "3b90f67f4b429e77d3da36698cef700c")

 function(download_win_ffmpeg script_var)
@@ -3,10 +3,10 @@ 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 6)
+set(VERSION_MINOR 1)
+set(VERSION_REVISION 2)
 set(VERSION ${VERSION_MAJOR}.${VERSION_MINOR}.${VERSION_REVISION})
-set(LIBJPEG_TURBO_VERSION_NUMBER 2000006)
+set(LIBJPEG_TURBO_VERSION_NUMBER 2001002)

 string(TIMESTAMP BUILD "opencv-${OPENCV_VERSION}-libjpeg-turbo")
 if(CMAKE_BUILD_TYPE STREQUAL "Debug")
@@ -46,7 +46,6 @@ if(UNIX)
  ocv_update(HAVE_UNSIGNED_SHORT 1)
  # undef INCOMPLETE_TYPES_BROKEN
  ocv_update(RIGHT_SHIFT_IS_UNSIGNED 0)
-  ocv_update(__CHAR_UNSIGNED__ 0)
 endif()


@@ -91,7 +91,7 @@ best of our understanding.
 The Modified (3-clause) BSD License
 ===================================

-Copyright (C)2009-2020 D. R. Commander.  All Rights Reserved.
+Copyright (C)2009-2021 D. R. Commander.  All Rights Reserved.<br>
 Copyright (C)2015 Viktor Szathmáry.  All Rights Reserved.

 Redistribution and use in source and binary forms, with or without
@@ -128,7 +128,7 @@ 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.
+This software is copyright (C) 1991-2020, Thomas G. Lane, Guido Vollbeding.
 All Rights Reserved except as specified below.

 Permission is hereby granted to use, copy, modify, and distribute this
@@ -159,19 +159,6 @@ 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
 ==========

@@ -3,7 +3,7 @@ Background

 libjpeg-turbo is a JPEG image codec that uses SIMD instructions to accelerate
 baseline JPEG compression and decompression on x86, x86-64, Arm, PowerPC, and
-MIPS systems, as well as progressive JPEG compression on x86 and x86-64
+MIPS systems, as well as progressive JPEG compression on x86, x86-64, and Arm
 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
@@ -61,11 +61,6 @@
   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 */

@@ -18,7 +18,6 @@
 #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 */
@@ -40,3 +40,13 @@
 #define HAVE_BITSCANFORWARD
 #endif
 #endif
+
+#if defined(__has_attribute)
+#if __has_attribute(fallthrough)
+#define FALLTHROUGH  __attribute__((fallthrough));
+#else
+#define FALLTHROUGH
+#endif
+#else
+#define FALLTHROUGH
+#endif
@@ -48,9 +48,9 @@ rgb_ycc_convert_internal(j_compress_ptr cinfo, JSAMPARRAY input_buf,
    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]);
+      r = inptr[RGB_RED];
+      g = inptr[RGB_GREEN];
+      b = 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.
@@ -100,9 +100,9 @@ rgb_gray_convert_internal(j_compress_ptr cinfo, JSAMPARRAY 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]);
+      r = inptr[RGB_RED];
+      g = inptr[RGB_GREEN];
+      b = inptr[RGB_BLUE];
      inptr += RGB_PIXELSIZE;
      /* Y */
      outptr[col] = (JSAMPLE)((ctab[r + R_Y_OFF] + ctab[g + G_Y_OFF] +
@@ -135,9 +135,9 @@ rgb_rgb_convert_internal(j_compress_ptr cinfo, JSAMPARRAY input_buf,
    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]);
+      outptr0[col] = inptr[RGB_RED];
+      outptr1[col] = inptr[RGB_GREEN];
+      outptr2[col] = inptr[RGB_BLUE];
      inptr += RGB_PIXELSIZE;
    }
  }
@@ -392,11 +392,11 @@ cmyk_ycck_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf,
    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]);
+      r = MAXJSAMPLE - inptr[0];
+      g = MAXJSAMPLE - inptr[1];
+      b = MAXJSAMPLE - inptr[2];
      /* K passes through as-is */
-      outptr3[col] = inptr[3];  /* don't need GETJSAMPLE here */
+      outptr3[col] = inptr[3];
      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.
@@ -438,7 +438,7 @@ grayscale_convert(j_compress_ptr cinfo, JSAMPARRAY 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 */
+      outptr[col] = inptr[0];
      inptr += instride;
    }
  }
@@ -497,7 +497,7 @@ null_convert(j_compress_ptr cinfo, JSAMPARRAY input_buf, JSAMPIMAGE output_buf,
        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 */
+          outptr[col] = inptr[ci];
          inptr += nc;
        }
      }
@@ -381,19 +381,19 @@ convsamp(JSAMPARRAY sample_data, JDIMENSION start_col, DCTELEM *workspace)
    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;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
+    *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
 #else
    {
      register int elemc;
      for (elemc = DCTSIZE; elemc > 0; elemc--)
-        *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
+        *workspaceptr++ = (*elemptr++) - CENTERJSAMPLE;
    }
 #endif
  }
@@ -533,20 +533,19 @@ convsamp_float(JSAMPARRAY sample_data, JDIMENSION start_col,
  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);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
+    *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
 #else
    {
      register int elemc;
      for (elemc = DCTSIZE; elemc > 0; elemc--)
-        *workspaceptr++ = (FAST_FLOAT)
-                          (GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
+        *workspaceptr++ = (FAST_FLOAT)((*elemptr++) - CENTERJSAMPLE);
    }
 #endif
  }
@@ -4,8 +4,10 @@
 * 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, 2014-2016, 2018-2019, D. R. Commander.
+ * Copyright (C) 2009-2011, 2014-2016, 2018-2021, D. R. Commander.
 * Copyright (C) 2015, Matthieu Darbois.
+ * Copyright (C) 2018, Matthias Räncker.
+ * Copyright (C) 2020, Arm Limited.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -43,14 +45,19 @@
 */

 /* NOTE: Both GCC and Clang define __GNUC__ */
-#if defined(__GNUC__) && (defined(__arm__) || defined(__aarch64__))
+#if (defined(__GNUC__) && (defined(__arm__) || defined(__aarch64__))) || \
+    defined(_M_ARM) || defined(_M_ARM64)
 #if !defined(__thumb__) || defined(__thumb2__)
 #define USE_CLZ_INTRINSIC
 #endif
 #endif

 #ifdef USE_CLZ_INTRINSIC
+#if defined(_MSC_VER) && !defined(__clang__)
+#define JPEG_NBITS_NONZERO(x)  (32 - _CountLeadingZeros(x))
+#else
 #define JPEG_NBITS_NONZERO(x)  (32 - __builtin_clz(x))
+#endif
 #define JPEG_NBITS(x)          (x ? JPEG_NBITS_NONZERO(x) : 0)
 #else
 #include "jpeg_nbits_table.h"
@@ -65,32 +72,43 @@
 * but must not be updated permanently until we complete the MCU.
 */

+#if defined(__x86_64__) && defined(__ILP32__)
+typedef unsigned long long bit_buf_type;
+#else
+typedef size_t bit_buf_type;
+#endif
+
+/* NOTE: The more optimal Huffman encoding algorithm is only used by the
+ * intrinsics implementation of the Arm Neon SIMD extensions, which is why we
+ * retain the old Huffman encoder behavior when using the GAS implementation.
+ */
+#if defined(WITH_SIMD) && !(defined(__arm__) || defined(__aarch64__) || \
+                            defined(_M_ARM) || defined(_M_ARM64))
+typedef unsigned long long simd_bit_buf_type;
+#else
+typedef bit_buf_type simd_bit_buf_type;
+#endif
+
+#if (defined(SIZEOF_SIZE_T) && SIZEOF_SIZE_T == 8) || defined(_WIN64) || \
+    (defined(__x86_64__) && defined(__ILP32__))
+#define BIT_BUF_SIZE  64
+#elif (defined(SIZEOF_SIZE_T) && SIZEOF_SIZE_T == 4) || defined(_WIN32)
+#define BIT_BUF_SIZE  32
+#else
+#error Cannot determine word size
+#endif
+#define SIMD_BIT_BUF_SIZE  (sizeof(simd_bit_buf_type) * 8)
+
 typedef struct {
-  size_t put_buffer;                    /* current bit-accumulation buffer */
-  int put_bits;                         /* # of bits now in it */
+  union {
+    bit_buf_type c;
+    simd_bit_buf_type simd;
+  } put_buffer;                         /* current bit accumulation buffer */
+  int free_bits;                        /* # of bits available in it */
+                                        /* (Neon GAS: # of bits now in it) */
  int last_dc_val[MAX_COMPS_IN_SCAN];   /* last DC coef for each component */
 } savable_state;

-/* This macro is to work around compilers with missing or broken
- * structure assignment.  You'll need to fix this code if you have
- * such a compiler and you change MAX_COMPS_IN_SCAN.
- */
-
-#ifndef NO_STRUCT_ASSIGN
-#define ASSIGN_STATE(dest, src)  ((dest) = (src))
-#else
-#if MAX_COMPS_IN_SCAN == 4
-#define ASSIGN_STATE(dest, src) \
-  ((dest).put_buffer = (src).put_buffer, \
-   (dest).put_bits = (src).put_bits, \
-   (dest).last_dc_val[0] = (src).last_dc_val[0], \
-   (dest).last_dc_val[1] = (src).last_dc_val[1], \
-   (dest).last_dc_val[2] = (src).last_dc_val[2], \
-   (dest).last_dc_val[3] = (src).last_dc_val[3])
-#endif
-#endif
-
-
 typedef struct {
  struct jpeg_entropy_encoder pub; /* public fields */

@@ -123,6 +141,7 @@ typedef struct {
  size_t free_in_buffer;        /* # of byte spaces remaining in buffer */
  savable_state cur;            /* Current bit buffer & DC state */
  j_compress_ptr cinfo;         /* dump_buffer needs access to this */
+  int simd;
 } working_state;


@@ -201,8 +220,17 @@ start_pass_huff(j_compress_ptr cinfo, boolean gather_statistics)
  }

  /* Initialize bit buffer to empty */
-  entropy->saved.put_buffer = 0;
-  entropy->saved.put_bits = 0;
+  if (entropy->simd) {
+    entropy->saved.put_buffer.simd = 0;
+#if defined(__aarch64__) && !defined(NEON_INTRINSICS)
+    entropy->saved.free_bits = 0;
+#else
+    entropy->saved.free_bits = SIMD_BIT_BUF_SIZE;
+#endif
+  } else {
+    entropy->saved.put_buffer.c = 0;
+    entropy->saved.free_bits = BIT_BUF_SIZE;
+  }

  /* Initialize restart stuff */
  entropy->restarts_to_go = cinfo->restart_interval;
@@ -287,6 +315,7 @@ jpeg_make_c_derived_tbl(j_compress_ptr cinfo, boolean isDC, int tblno,
   * this lets us detect duplicate VAL entries here, and later
   * allows emit_bits to detect any attempt to emit such symbols.
   */
+  MEMZERO(dtbl->ehufco, sizeof(dtbl->ehufco));
  MEMZERO(dtbl->ehufsi, sizeof(dtbl->ehufsi));

  /* This is also a convenient place to check for out-of-range
@@ -334,94 +363,94 @@ dump_buffer(working_state *state)

 /* Outputting bits to the file */

-/* These macros perform the same task as the emit_bits() function in the
- * original libjpeg code.  In addition to reducing overhead by explicitly
- * inlining the code, additional performance is achieved by taking into
- * account the size of the bit buffer and waiting until it is almost full
- * before emptying it.  This mostly benefits 64-bit platforms, since 6
- * bytes can be stored in a 64-bit bit buffer before it has to be emptied.
+/* Output byte b and, speculatively, an additional 0 byte.  0xFF must be
+ * encoded as 0xFF 0x00, so the output buffer pointer is advanced by 2 if the
+ * byte is 0xFF.  Otherwise, the output buffer pointer is advanced by 1, and
+ * the speculative 0 byte will be overwritten by the next byte.
 */
-
-#define EMIT_BYTE() { \
-  JOCTET c; \
-  put_bits -= 8; \
-  c = (JOCTET)GETJOCTET(put_buffer >> put_bits); \
-  *buffer++ = c; \
-  if (c == 0xFF)  /* need to stuff a zero byte? */ \
-    *buffer++ = 0; \
+#define EMIT_BYTE(b) { \
+  buffer[0] = (JOCTET)(b); \
+  buffer[1] = 0; \
+  buffer -= -2 + ((JOCTET)(b) < 0xFF); \
 }

-#define PUT_BITS(code, size) { \
-  put_bits += size; \
-  put_buffer = (put_buffer << size) | code; \
-}
+/* Output the entire bit buffer.  If there are no 0xFF bytes in it, then write
+ * directly to the output buffer.  Otherwise, use the EMIT_BYTE() macro to
+ * encode 0xFF as 0xFF 0x00.
+ */
+#if BIT_BUF_SIZE == 64

-#if SIZEOF_SIZE_T != 8 && !defined(_WIN64)
-
-#define CHECKBUF15() { \
-  if (put_bits > 15) { \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
+#define FLUSH() { \
+  if (put_buffer & 0x8080808080808080 & ~(put_buffer + 0x0101010101010101)) { \
+    EMIT_BYTE(put_buffer >> 56) \
+    EMIT_BYTE(put_buffer >> 48) \
+    EMIT_BYTE(put_buffer >> 40) \
+    EMIT_BYTE(put_buffer >> 32) \
+    EMIT_BYTE(put_buffer >> 24) \
+    EMIT_BYTE(put_buffer >> 16) \
+    EMIT_BYTE(put_buffer >>  8) \
+    EMIT_BYTE(put_buffer      ) \
+  } else { \
+    buffer[0] = (JOCTET)(put_buffer >> 56); \
+    buffer[1] = (JOCTET)(put_buffer >> 48); \
+    buffer[2] = (JOCTET)(put_buffer >> 40); \
+    buffer[3] = (JOCTET)(put_buffer >> 32); \
+    buffer[4] = (JOCTET)(put_buffer >> 24); \
+    buffer[5] = (JOCTET)(put_buffer >> 16); \
+    buffer[6] = (JOCTET)(put_buffer >> 8); \
+    buffer[7] = (JOCTET)(put_buffer); \
+    buffer += 8; \
  } \
 }

-#endif
-
-#define CHECKBUF31() { \
-  if (put_bits > 31) { \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-  } \
-}
-
-#define CHECKBUF47() { \
-  if (put_bits > 47) { \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-    EMIT_BYTE() \
-  } \
-}
-
-#if !defined(_WIN32) && !defined(SIZEOF_SIZE_T)
-#error Cannot determine word size
-#endif
-
-#if SIZEOF_SIZE_T == 8 || defined(_WIN64)
-
-#define EMIT_BITS(code, size) { \
-  CHECKBUF47() \
-  PUT_BITS(code, size) \
-}
-
-#define EMIT_CODE(code, size) { \
-  temp2 &= (((JLONG)1) << nbits) - 1; \
-  CHECKBUF31() \
-  PUT_BITS(code, size) \
-  PUT_BITS(temp2, nbits) \
-}
-
 #else

-#define EMIT_BITS(code, size) { \
-  PUT_BITS(code, size) \
-  CHECKBUF15() \
-}
-
-#define EMIT_CODE(code, size) { \
-  temp2 &= (((JLONG)1) << nbits) - 1; \
-  PUT_BITS(code, size) \
-  CHECKBUF15() \
-  PUT_BITS(temp2, nbits) \
-  CHECKBUF15() \
+#define FLUSH() { \
+  if (put_buffer & 0x80808080 & ~(put_buffer + 0x01010101)) { \
+    EMIT_BYTE(put_buffer >> 24) \
+    EMIT_BYTE(put_buffer >> 16) \
+    EMIT_BYTE(put_buffer >>  8) \
+    EMIT_BYTE(put_buffer      ) \
+  } else { \
+    buffer[0] = (JOCTET)(put_buffer >> 24); \
+    buffer[1] = (JOCTET)(put_buffer >> 16); \
+    buffer[2] = (JOCTET)(put_buffer >> 8); \
+    buffer[3] = (JOCTET)(put_buffer); \
+    buffer += 4; \
+  } \
 }

 #endif

+/* Fill the bit buffer to capacity with the leading bits from code, then output
+ * the bit buffer and put the remaining bits from code into the bit buffer.
+ */
+#define PUT_AND_FLUSH(code, size) { \
+  put_buffer = (put_buffer << (size + free_bits)) | (code >> -free_bits); \
+  FLUSH() \
+  free_bits += BIT_BUF_SIZE; \
+  put_buffer = code; \
+}
+
+/* Insert code into the bit buffer and output the bit buffer if needed.
+ * NOTE: We can't flush with free_bits == 0, since the left shift in
+ * PUT_AND_FLUSH() would have undefined behavior.
+ */
+#define PUT_BITS(code, size) { \
+  free_bits -= size; \
+  if (free_bits < 0) \
+    PUT_AND_FLUSH(code, size) \
+  else \
+    put_buffer = (put_buffer << size) | code; \
+}
+
+#define PUT_CODE(code, size) { \
+  temp &= (((JLONG)1) << nbits) - 1; \
+  temp |= code << nbits; \
+  nbits += size; \
+  PUT_BITS(temp, nbits) \
+}
+

 /* Although it is exceedingly rare, it is possible for a Huffman-encoded
 * coefficient block to be larger than the 128-byte unencoded block.  For each
@@ -444,6 +473,7 @@ dump_buffer(working_state *state)

 #define STORE_BUFFER() { \
  if (localbuf) { \
+    size_t bytes, bytestocopy; \
    bytes = buffer - _buffer; \
    buffer = _buffer; \
    while (bytes > 0) { \
@@ -466,20 +496,46 @@ dump_buffer(working_state *state)
 LOCAL(boolean)
 flush_bits(working_state *state)
 {
-  JOCTET _buffer[BUFSIZE], *buffer;
-  size_t put_buffer;  int put_bits;
-  size_t bytes, bytestocopy;  int localbuf = 0;
+  JOCTET _buffer[BUFSIZE], *buffer, temp;
+  simd_bit_buf_type put_buffer;  int put_bits;
+  int localbuf = 0;
+
+  if (state->simd) {
+#if defined(__aarch64__) && !defined(NEON_INTRINSICS)
+    put_bits = state->cur.free_bits;
+#else
+    put_bits = SIMD_BIT_BUF_SIZE - state->cur.free_bits;
+#endif
+    put_buffer = state->cur.put_buffer.simd;
+  } else {
+    put_bits = BIT_BUF_SIZE - state->cur.free_bits;
+    put_buffer = state->cur.put_buffer.c;
+  }

-  put_buffer = state->cur.put_buffer;
-  put_bits = state->cur.put_bits;
  LOAD_BUFFER()

-  /* fill any partial byte with ones */
-  PUT_BITS(0x7F, 7)
-  while (put_bits >= 8) EMIT_BYTE()
+  while (put_bits >= 8) {
+    put_bits -= 8;
+    temp = (JOCTET)(put_buffer >> put_bits);
+    EMIT_BYTE(temp)
+  }
+  if (put_bits) {
+    /* fill partial byte with ones */
+    temp = (JOCTET)((put_buffer << (8 - put_bits)) | (0xFF >> put_bits));
+    EMIT_BYTE(temp)
+  }

-  state->cur.put_buffer = 0;    /* and reset bit-buffer to empty */
-  state->cur.put_bits = 0;
+  if (state->simd) {                    /* and reset bit buffer to empty */
+    state->cur.put_buffer.simd = 0;
+#if defined(__aarch64__) && !defined(NEON_INTRINSICS)
+    state->cur.free_bits = 0;
+#else
+    state->cur.free_bits = SIMD_BIT_BUF_SIZE;
+#endif
+  } else {
+    state->cur.put_buffer.c = 0;
+    state->cur.free_bits = BIT_BUF_SIZE;
+  }
  STORE_BUFFER()

  return TRUE;
@@ -493,7 +549,7 @@ encode_one_block_simd(working_state *state, JCOEFPTR block, int last_dc_val,
                      c_derived_tbl *dctbl, c_derived_tbl *actbl)
 {
  JOCTET _buffer[BUFSIZE], *buffer;
-  size_t bytes, bytestocopy;  int localbuf = 0;
+  int localbuf = 0;

  LOAD_BUFFER()

@@ -509,53 +565,41 @@ LOCAL(boolean)
 encode_one_block(working_state *state, JCOEFPTR block, int last_dc_val,
                 c_derived_tbl *dctbl, c_derived_tbl *actbl)
 {
-  int temp, temp2, temp3;
-  int nbits;
-  int r, code, size;
+  int temp, nbits, free_bits;
+  bit_buf_type put_buffer;
  JOCTET _buffer[BUFSIZE], *buffer;
-  size_t put_buffer;  int put_bits;
-  int code_0xf0 = actbl->ehufco[0xf0], size_0xf0 = actbl->ehufsi[0xf0];
-  size_t bytes, bytestocopy;  int localbuf = 0;
+  int localbuf = 0;

-  put_buffer = state->cur.put_buffer;
-  put_bits = state->cur.put_bits;
+  free_bits = state->cur.free_bits;
+  put_buffer = state->cur.put_buffer.c;
  LOAD_BUFFER()

  /* Encode the DC coefficient difference per section F.1.2.1 */

-  temp = temp2 = block[0] - last_dc_val;
+  temp = block[0] - last_dc_val;

  /* This is a well-known technique for obtaining the absolute value without a
   * branch.  It is derived from an assembly language technique presented in
   * "How to Optimize for the Pentium Processors", Copyright (c) 1996, 1997 by
-   * Agner Fog.
+   * Agner Fog.  This code assumes we are on a two's complement machine.
   */
-  temp3 = temp >> (CHAR_BIT * sizeof(int) - 1);
-  temp ^= temp3;
-  temp -= temp3;
-
-  /* For a negative input, want temp2 = bitwise complement of abs(input) */
-  /* This code assumes we are on a two's complement machine */
-  temp2 += temp3;
+  nbits = temp >> (CHAR_BIT * sizeof(int) - 1);
+  temp += nbits;
+  nbits ^= temp;

  /* Find the number of bits needed for the magnitude of the coefficient */
-  nbits = JPEG_NBITS(temp);
+  nbits = JPEG_NBITS(nbits);

-  /* Emit the Huffman-coded symbol for the number of bits */
-  code = dctbl->ehufco[nbits];
-  size = dctbl->ehufsi[nbits];
-  EMIT_BITS(code, size)
-
-  /* Mask off any extra bits in code */
-  temp2 &= (((JLONG)1) << nbits) - 1;
-
-  /* Emit that number of bits of the value, if positive, */
-  /* or the complement of its magnitude, if negative. */
-  EMIT_BITS(temp2, nbits)
+  /* Emit the Huffman-coded symbol for the number of bits.
+   * Emit that number of bits of the value, if positive,
+   * or the complement of its magnitude, if negative.
+   */
+  PUT_CODE(dctbl->ehufco[nbits], dctbl->ehufsi[nbits])

  /* Encode the AC coefficients per section F.1.2.2 */

-  r = 0;                        /* r = run length of zeros */
+  {
+    int r = 0;                  /* r = run length of zeros */

 /* Manually unroll the k loop to eliminate the counter variable.  This
 * improves performance greatly on systems with a limited number of
@@ -563,51 +607,46 @@ encode_one_block(working_state *state, JCOEFPTR block, int last_dc_val,
 */
 #define kloop(jpeg_natural_order_of_k) { \
  if ((temp = block[jpeg_natural_order_of_k]) == 0) { \
-    r++; \
+    r += 16; \
  } else { \
-    temp2 = temp; \
    /* Branch-less absolute value, bitwise complement, etc., same as above */ \
-    temp3 = temp >> (CHAR_BIT * sizeof(int) - 1); \
-    temp ^= temp3; \
-    temp -= temp3; \
-    temp2 += temp3; \
-    nbits = JPEG_NBITS_NONZERO(temp); \
+    nbits = temp >> (CHAR_BIT * sizeof(int) - 1); \
+    temp += nbits; \
+    nbits ^= temp; \
+    nbits = JPEG_NBITS_NONZERO(nbits); \
    /* if run length > 15, must emit special run-length-16 codes (0xF0) */ \
-    while (r > 15) { \
-      EMIT_BITS(code_0xf0, size_0xf0) \
-      r -= 16; \
+    while (r >= 16 * 16) { \
+      r -= 16 * 16; \
+      PUT_BITS(actbl->ehufco[0xf0], actbl->ehufsi[0xf0]) \
    } \
    /* Emit Huffman symbol for run length / number of bits */ \
-    temp3 = (r << 4) + nbits; \
-    code = actbl->ehufco[temp3]; \
-    size = actbl->ehufsi[temp3]; \
-    EMIT_CODE(code, size) \
+    r += nbits; \
+    PUT_CODE(actbl->ehufco[r], actbl->ehufsi[r]) \
    r = 0; \
  } \
 }

-  /* One iteration for each value in jpeg_natural_order[] */
-  kloop(1);   kloop(8);   kloop(16);  kloop(9);   kloop(2);   kloop(3);
-  kloop(10);  kloop(17);  kloop(24);  kloop(32);  kloop(25);  kloop(18);
-  kloop(11);  kloop(4);   kloop(5);   kloop(12);  kloop(19);  kloop(26);
-  kloop(33);  kloop(40);  kloop(48);  kloop(41);  kloop(34);  kloop(27);
-  kloop(20);  kloop(13);  kloop(6);   kloop(7);   kloop(14);  kloop(21);
-  kloop(28);  kloop(35);  kloop(42);  kloop(49);  kloop(56);  kloop(57);
-  kloop(50);  kloop(43);  kloop(36);  kloop(29);  kloop(22);  kloop(15);
-  kloop(23);  kloop(30);  kloop(37);  kloop(44);  kloop(51);  kloop(58);
-  kloop(59);  kloop(52);  kloop(45);  kloop(38);  kloop(31);  kloop(39);
-  kloop(46);  kloop(53);  kloop(60);  kloop(61);  kloop(54);  kloop(47);
-  kloop(55);  kloop(62);  kloop(63);
+    /* One iteration for each value in jpeg_natural_order[] */
+    kloop(1);   kloop(8);   kloop(16);  kloop(9);   kloop(2);   kloop(3);
+    kloop(10);  kloop(17);  kloop(24);  kloop(32);  kloop(25);  kloop(18);
+    kloop(11);  kloop(4);   kloop(5);   kloop(12);  kloop(19);  kloop(26);
+    kloop(33);  kloop(40);  kloop(48);  kloop(41);  kloop(34);  kloop(27);
+    kloop(20);  kloop(13);  kloop(6);   kloop(7);   kloop(14);  kloop(21);
+    kloop(28);  kloop(35);  kloop(42);  kloop(49);  kloop(56);  kloop(57);
+    kloop(50);  kloop(43);  kloop(36);  kloop(29);  kloop(22);  kloop(15);
+    kloop(23);  kloop(30);  kloop(37);  kloop(44);  kloop(51);  kloop(58);
+    kloop(59);  kloop(52);  kloop(45);  kloop(38);  kloop(31);  kloop(39);
+    kloop(46);  kloop(53);  kloop(60);  kloop(61);  kloop(54);  kloop(47);
+    kloop(55);  kloop(62);  kloop(63);

-  /* If the last coef(s) were zero, emit an end-of-block code */
-  if (r > 0) {
-    code = actbl->ehufco[0];
-    size = actbl->ehufsi[0];
-    EMIT_BITS(code, size)
+    /* If the last coef(s) were zero, emit an end-of-block code */
+    if (r > 0) {
+      PUT_BITS(actbl->ehufco[0], actbl->ehufsi[0])
+    }
  }

-  state->cur.put_buffer = put_buffer;
-  state->cur.put_bits = put_bits;
+  state->cur.put_buffer.c = put_buffer;
+  state->cur.free_bits = free_bits;
  STORE_BUFFER()

  return TRUE;
@@ -654,8 +693,9 @@ encode_mcu_huff(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
  /* Load up working state */
  state.next_output_byte = cinfo->dest->next_output_byte;
  state.free_in_buffer = cinfo->dest->free_in_buffer;
-  ASSIGN_STATE(state.cur, entropy->saved);
+  state.cur = entropy->saved;
  state.cinfo = cinfo;
+  state.simd = entropy->simd;

  /* Emit restart marker if needed */
  if (cinfo->restart_interval) {
@@ -694,7 +734,7 @@ encode_mcu_huff(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
  /* Completed MCU, so update state */
  cinfo->dest->next_output_byte = state.next_output_byte;
  cinfo->dest->free_in_buffer = state.free_in_buffer;
-  ASSIGN_STATE(entropy->saved, state.cur);
+  entropy->saved = state.cur;

  /* Update restart-interval state too */
  if (cinfo->restart_interval) {
@@ -723,8 +763,9 @@ finish_pass_huff(j_compress_ptr cinfo)
  /* Load up working state ... flush_bits needs it */
  state.next_output_byte = cinfo->dest->next_output_byte;
  state.free_in_buffer = cinfo->dest->free_in_buffer;
-  ASSIGN_STATE(state.cur, entropy->saved);
+  state.cur = entropy->saved;
  state.cinfo = cinfo;
+  state.simd = entropy->simd;

  /* Flush out the last data */
  if (!flush_bits(&state))
@@ -733,7 +774,7 @@ finish_pass_huff(j_compress_ptr cinfo)
  /* Update state */
  cinfo->dest->next_output_byte = state.next_output_byte;
  cinfo->dest->free_in_buffer = state.free_in_buffer;
-  ASSIGN_STATE(entropy->saved, state.cur);
+  entropy->saved = state.cur;
 }


@@ -493,7 +493,7 @@ prepare_for_pass(j_compress_ptr cinfo)
    master->pass_type = output_pass;
    master->pass_number++;
 #endif
-    /*FALLTHROUGH*/
+    FALLTHROUGH                 /*FALLTHROUGH*/
  case output_pass:
    /* Do a data-output pass. */
    /* We need not repeat per-scan setup if prior optimization pass did it. */
@@ -4,8 +4,10 @@
 * This file was part of the Independent JPEG Group's software:
 * Copyright (C) 1995-1997, Thomas G. Lane.
 * libjpeg-turbo Modifications:
- * Copyright (C) 2011, 2015, 2018, D. R. Commander.
+ * Copyright (C) 2011, 2015, 2018, 2021, D. R. Commander.
 * Copyright (C) 2016, 2018, Matthieu Darbois.
+ * Copyright (C) 2020, Arm Limited.
+ * Copyright (C) 2021, Alex Richardson.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -52,14 +54,19 @@
 */

 /* NOTE: Both GCC and Clang define __GNUC__ */
-#if defined(__GNUC__) && (defined(__arm__) || defined(__aarch64__))
+#if (defined(__GNUC__) && (defined(__arm__) || defined(__aarch64__))) || \
+    defined(_M_ARM) || defined(_M_ARM64)
 #if !defined(__thumb__) || defined(__thumb2__)
 #define USE_CLZ_INTRINSIC
 #endif
 #endif

 #ifdef USE_CLZ_INTRINSIC
+#if defined(_MSC_VER) && !defined(__clang__)
+#define JPEG_NBITS_NONZERO(x)  (32 - _CountLeadingZeros(x))
+#else
 #define JPEG_NBITS_NONZERO(x)  (32 - __builtin_clz(x))
+#endif
 #define JPEG_NBITS(x)          (x ? JPEG_NBITS_NONZERO(x) : 0)
 #else
 #include "jpeg_nbits_table.h"
@@ -169,24 +176,26 @@ INLINE
 METHODDEF(int)
 count_zeroes(size_t *x)
 {
-  int result;
 #if defined(HAVE_BUILTIN_CTZL)
+  int result;
  result = __builtin_ctzl(*x);
  *x >>= result;
 #elif defined(HAVE_BITSCANFORWARD64)
+  unsigned long result;
  _BitScanForward64(&result, *x);
  *x >>= result;
 #elif defined(HAVE_BITSCANFORWARD)
+  unsigned long result;
  _BitScanForward(&result, *x);
  *x >>= result;
 #else
-  result = 0;
+  int result = 0;
  while ((*x & 1) == 0) {
    ++result;
    *x >>= 1;
  }
 #endif
-  return result;
+  return (int)result;
 }


@@ -672,7 +681,7 @@ encode_mcu_AC_first(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
      emit_restart(entropy, entropy->next_restart_num);

 #ifdef WITH_SIMD
-  cvalue = values = (JCOEF *)PAD((size_t)values_unaligned, 16);
+  cvalue = values = (JCOEF *)PAD((JUINTPTR)values_unaligned, 16);
 #else
  /* Not using SIMD, so alignment is not needed */
  cvalue = values = values_unaligned;
@@ -860,7 +869,7 @@ encode_mcu_AC_refine_prepare(const JCOEF *block,

 #define ENCODE_COEFS_AC_REFINE(label) { \
  while (zerobits) { \
-    int idx = count_zeroes(&zerobits); \
+    idx = count_zeroes(&zerobits); \
    r += idx; \
    cabsvalue += idx; \
    signbits >>= idx; \
@@ -917,7 +926,7 @@ METHODDEF(boolean)
 encode_mcu_AC_refine(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
 {
  phuff_entropy_ptr entropy = (phuff_entropy_ptr)cinfo->entropy;
-  register int temp, r;
+  register int temp, r, idx;
  char *BR_buffer;
  unsigned int BR;
  int Sl = cinfo->Se - cinfo->Ss + 1;
@@ -937,7 +946,7 @@ encode_mcu_AC_refine(j_compress_ptr cinfo, JBLOCKROW *MCU_data)
      emit_restart(entropy, entropy->next_restart_num);

 #ifdef WITH_SIMD
-  cabsvalue = absvalues = (JCOEF *)PAD((size_t)absvalues_unaligned, 16);
+  cabsvalue = absvalues = (JCOEF *)PAD((JUINTPTR)absvalues_unaligned, 16);
 #else
  /* Not using SIMD, so alignment is not needed */
  cabsvalue = absvalues = absvalues_unaligned;
@@ -968,7 +977,7 @@ encode_mcu_AC_refine(j_compress_ptr cinfo, JBLOCKROW *MCU_data)

  if (zerobits) {
    int diff = ((absvalues + DCTSIZE2 / 2) - cabsvalue);
-    int idx = count_zeroes(&zerobits);
+    idx = count_zeroes(&zerobits);
    signbits >>= idx;
    idx += diff;
    r += idx;
@@ -6,7 +6,7 @@
 * 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.
+ * Copyright (C) 2015, 2019, D. R. Commander.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -103,7 +103,7 @@ expand_right_edge(JSAMPARRAY image_data, int num_rows, JDIMENSION 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 */
+      pixval = ptr[-1];
      for (count = numcols; count > 0; count--)
        *ptr++ = pixval;
    }
@@ -174,7 +174,7 @@ int_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
      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++);
+          outvalue += (JLONG)(*inptr++);
        }
      }
      *outptr++ = (JSAMPLE)((outvalue + numpix2) / numpix);
@@ -237,8 +237,7 @@ h2v1_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
    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);
+      *outptr++ = (JSAMPLE)((inptr[0] + inptr[1] + bias) >> 1);
      bias ^= 1;                /* 0=>1, 1=>0 */
      inptr += 2;
    }
@@ -277,8 +276,7 @@ h2v2_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
    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);
+        (JSAMPLE)((inptr0[0] + inptr0[1] + inptr1[0] + inptr1[1] + bias) >> 2);
      bias ^= 3;                /* 1=>2, 2=>1 */
      inptr0 += 2;  inptr1 += 2;
    }
@@ -337,33 +335,25 @@ h2v2_smooth_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
    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]);
+    membersum = inptr0[0] + inptr0[1] + inptr1[0] + inptr1[1];
+    neighsum = above_ptr[0] + above_ptr[1] + below_ptr[0] + below_ptr[1] +
+               inptr0[0] + inptr0[2] + inptr1[0] + inptr1[2];
    neighsum += neighsum;
-    neighsum += GETJSAMPLE(*above_ptr) + GETJSAMPLE(above_ptr[2]) +
-                GETJSAMPLE(*below_ptr) + GETJSAMPLE(below_ptr[2]);
+    neighsum += above_ptr[0] + above_ptr[2] + below_ptr[0] + 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]);
+      membersum = inptr0[0] + inptr0[1] + inptr1[0] + 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]);
+      neighsum = above_ptr[0] + above_ptr[1] + below_ptr[0] + below_ptr[1] +
+                 inptr0[-1] + inptr0[2] + inptr1[-1] + 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]);
+      neighsum += above_ptr[-1] + above_ptr[2] + below_ptr[-1] + below_ptr[2];
      /* form final output scaled up by 2^16 */
      membersum = membersum * memberscale + neighsum * neighscale;
      /* round, descale and output it */
@@ -372,15 +362,11 @@ h2v2_smooth_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
    }

    /* 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]);
+    membersum = inptr0[0] + inptr0[1] + inptr1[0] + inptr1[1];
+    neighsum = above_ptr[0] + above_ptr[1] + below_ptr[0] + below_ptr[1] +
+               inptr0[-1] + inptr0[1] + inptr1[-1] + inptr1[1];
    neighsum += neighsum;
-    neighsum += GETJSAMPLE(above_ptr[-1]) + GETJSAMPLE(above_ptr[1]) +
-                GETJSAMPLE(below_ptr[-1]) + GETJSAMPLE(below_ptr[1]);
+    neighsum += above_ptr[-1] + above_ptr[1] + below_ptr[-1] + below_ptr[1];
    membersum = membersum * memberscale + neighsum * neighscale;
    *outptr = (JSAMPLE)((membersum + 32768) >> 16);

@@ -429,21 +415,18 @@ fullsize_smooth_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
    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);
+    colsum = (*above_ptr++) + (*below_ptr++) + inptr[0];
+    membersum = *inptr++;
+    nextcolsum = above_ptr[0] + below_ptr[0] + inptr[0];
    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++);
+      membersum = *inptr++;
      above_ptr++;  below_ptr++;
-      nextcolsum = GETJSAMPLE(*above_ptr) + GETJSAMPLE(*below_ptr) +
-                   GETJSAMPLE(*inptr);
+      nextcolsum = above_ptr[0] + below_ptr[0] + inptr[0];
      neighsum = lastcolsum + (colsum - membersum) + nextcolsum;
      membersum = membersum * memberscale + neighsum * neighscale;
      *outptr++ = (JSAMPLE)((membersum + 32768) >> 16);
@@ -451,7 +434,7 @@ fullsize_smooth_downsample(j_compress_ptr cinfo, jpeg_component_info *compptr,
    }

    /* Special case for last column */
-    membersum = GETJSAMPLE(*inptr);
+    membersum = *inptr;
    neighsum = lastcolsum + (colsum - membersum) + colsum;
    membersum = membersum * memberscale + neighsum * neighscale;
    *outptr = (JSAMPLE)((membersum + 32768) >> 16);
@@ -23,6 +23,7 @@
 #include "jinclude.h"
 #include "jpeglib.h"
 #include "jdmaster.h"
+#include "jconfigint.h"


 /*
@@ -308,7 +309,7 @@ jpeg_consume_input(j_decompress_ptr cinfo)
    /* Initialize application's data source module */
    (*cinfo->src->init_source) (cinfo);
    cinfo->global_state = DSTATE_INHEADER;
-    /*FALLTHROUGH*/
+    FALLTHROUGH                 /*FALLTHROUGH*/
  case DSTATE_INHEADER:
    retcode = (*cinfo->inputctl->consume_input) (cinfo);
    if (retcode == JPEG_REACHED_SOS) { /* Found SOS, prepare to decompress */
@@ -4,7 +4,7 @@
 * 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, 2020, D. R. Commander.
+ * Copyright (C) 2010, 2015-2020, D. R. Commander.
 * Copyright (C) 2015, Google, Inc.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
@@ -319,6 +319,8 @@ read_and_discard_scanlines(j_decompress_ptr cinfo, JDIMENSION num_lines)
 {
  JDIMENSION n;
  my_master_ptr master = (my_master_ptr)cinfo->master;
+  JSAMPLE dummy_sample[1] = { 0 };
+  JSAMPROW dummy_row = dummy_sample;
  JSAMPARRAY scanlines = NULL;
  void (*color_convert) (j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
                         JDIMENSION input_row, JSAMPARRAY output_buf,
@@ -329,6 +331,10 @@ read_and_discard_scanlines(j_decompress_ptr cinfo, JDIMENSION num_lines)
  if (cinfo->cconvert && cinfo->cconvert->color_convert) {
    color_convert = cinfo->cconvert->color_convert;
    cinfo->cconvert->color_convert = noop_convert;
+    /* This just prevents UBSan from complaining about adding 0 to a NULL
+     * pointer.  The pointer isn't actually used.
+     */
+    scanlines = &dummy_row;
  }

  if (cinfo->cquantize && cinfo->cquantize->color_quantize) {
@@ -532,6 +538,8 @@ jpeg_skip_scanlines(j_decompress_ptr cinfo, JDIMENSION num_lines)
         * decoded coefficients.  This is ~5% faster for large subsets, but
         * it's tough to tell a difference for smaller images.
         */
+        if (!cinfo->entropy->insufficient_data)
+          cinfo->master->last_good_iMCU_row = cinfo->input_iMCU_row;
        (*cinfo->entropy->decode_mcu) (cinfo, NULL);
      }
    }
@@ -4,7 +4,7 @@
 * 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.
+ * Copyright (C) 2015-2020, D. R. Commander.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -80,7 +80,7 @@ get_byte(j_decompress_ptr cinfo)
    if (!(*src->fill_input_buffer) (cinfo))
      ERREXIT(cinfo, JERR_CANT_SUSPEND);
  src->bytes_in_buffer--;
-  return GETJOCTET(*src->next_input_byte++);
+  return *src->next_input_byte++;
 }


@@ -665,8 +665,16 @@ bad:
    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];
+      int *prev_coef_bit_ptr =
+        &cinfo->coef_bits[cindex + cinfo->num_components][0];
      if (cinfo->Ss && coef_bit_ptr[0] < 0) /* AC without prior DC scan */
        WARNMS2(cinfo, JWRN_BOGUS_PROGRESSION, cindex, 0);
+      for (coefi = MIN(cinfo->Ss, 1); coefi <= MAX(cinfo->Se, 9); coefi++) {
+        if (cinfo->input_scan_number > 1)
+          prev_coef_bit_ptr[coefi] = coef_bit_ptr[coefi];
+        else
+          prev_coef_bit_ptr[coefi] = 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)
@@ -727,6 +735,7 @@ bad:
  entropy->c = 0;
  entropy->a = 0;
  entropy->ct = -16;    /* force reading 2 initial bytes to fill C */
+  entropy->pub.insufficient_data = FALSE;

  /* Initialize restart counter */
  entropy->restarts_to_go = cinfo->restart_interval;
@@ -763,7 +772,7 @@ jinit_arith_decoder(j_decompress_ptr cinfo)
    int *coef_bit_ptr, ci;
    cinfo->coef_bits = (int (*)[DCTSIZE2])
      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
-                                  cinfo->num_components * DCTSIZE2 *
+                                  cinfo->num_components * 2 * DCTSIZE2 *
                                  sizeof(int));
    coef_bit_ptr = &cinfo->coef_bits[0][0];
    for (ci = 0; ci < cinfo->num_components; ci++)
@@ -5,7 +5,7 @@
 * 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) 2010, 2015-2016, 2019-2020, D. R. Commander.
 * Copyright (C) 2015, 2020, Google, Inc.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
@@ -102,6 +102,8 @@ decompress_onepass(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
      /* 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->insufficient_data)
+        cinfo->master->last_good_iMCU_row = cinfo->input_iMCU_row;
      if (!(*cinfo->entropy->decode_mcu) (cinfo, coef->MCU_buffer)) {
        /* Suspension forced; update state counters and exit */
        coef->MCU_vert_offset = yoffset;
@@ -227,6 +229,8 @@ consume_data(j_decompress_ptr cinfo)
          }
        }
      }
+      if (!cinfo->entropy->insufficient_data)
+        cinfo->master->last_good_iMCU_row = cinfo->input_iMCU_row;
      /* Try to fetch the MCU. */
      if (!(*cinfo->entropy->decode_mcu) (cinfo, coef->MCU_buffer)) {
        /* Suspension forced; update state counters and exit */
@@ -326,19 +330,22 @@ decompress_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
 #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.
+ * This code applies interblock smoothing; the first 9 AC coefficients are
+ * estimated from the DC values of a DCT block and its 24 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 */
+/* Natural-order array positions of the first 9 zigzag-order coefficients */
 #define Q01_POS  1
 #define Q10_POS  8
 #define Q20_POS  16
 #define Q11_POS  9
 #define Q02_POS  2
+#define Q03_POS  3
+#define Q12_POS  10
+#define Q21_POS  17
+#define Q30_POS  24

 /*
 * Determine whether block smoothing is applicable and safe.
@@ -356,8 +363,8 @@ smoothing_ok(j_decompress_ptr cinfo)
  int ci, coefi;
  jpeg_component_info *compptr;
  JQUANT_TBL *qtable;
-  int *coef_bits;
-  int *coef_bits_latch;
+  int *coef_bits, *prev_coef_bits;
+  int *coef_bits_latch, *prev_coef_bits_latch;

  if (!cinfo->progressive_mode || cinfo->coef_bits == NULL)
    return FALSE;
@@ -366,34 +373,47 @@ smoothing_ok(j_decompress_ptr cinfo)
  if (coef->coef_bits_latch == NULL)
    coef->coef_bits_latch = (int *)
      (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
-                                  cinfo->num_components *
+                                  cinfo->num_components * 2 *
                                  (SAVED_COEFS * sizeof(int)));
  coef_bits_latch = coef->coef_bits_latch;
+  prev_coef_bits_latch =
+    &coef->coef_bits_latch[cinfo->num_components * SAVED_COEFS];

  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. */
+    /* Verify DC & first 9 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)
+        qtable->quantval[Q02_POS] == 0 ||
+        qtable->quantval[Q03_POS] == 0 ||
+        qtable->quantval[Q12_POS] == 0 ||
+        qtable->quantval[Q21_POS] == 0 ||
+        qtable->quantval[Q30_POS] == 0)
      return FALSE;
    /* DC values must be at least partly known for all components. */
    coef_bits = cinfo->coef_bits[ci];
+    prev_coef_bits = cinfo->coef_bits[ci + cinfo->num_components];
    if (coef_bits[0] < 0)
      return FALSE;
+    coef_bits_latch[0] = coef_bits[0];
    /* Block smoothing is helpful if some AC coefficients remain inaccurate. */
-    for (coefi = 1; coefi <= 5; coefi++) {
+    for (coefi = 1; coefi < SAVED_COEFS; coefi++) {
+      if (cinfo->input_scan_number > 1)
+        prev_coef_bits_latch[coefi] = prev_coef_bits[coefi];
+      else
+        prev_coef_bits_latch[coefi] = -1;
      coef_bits_latch[coefi] = coef_bits[coefi];
      if (coef_bits[coefi] != 0)
        smoothing_useful = TRUE;
    }
    coef_bits_latch += SAVED_COEFS;
+    prev_coef_bits_latch += SAVED_COEFS;
  }

  return smoothing_useful;
@@ -412,17 +432,20 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
  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;
+  JBLOCKROW buffer_ptr, prev_prev_block_row, prev_block_row;
+  JBLOCKROW next_block_row, next_next_block_row;
  JSAMPARRAY output_ptr;
  JDIMENSION output_col;
  jpeg_component_info *compptr;
  inverse_DCT_method_ptr inverse_DCT;
-  boolean first_row, last_row;
+  boolean change_dc;
  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;
+  JLONG Q00, Q01, Q02, Q03 = 0, Q10, Q11, Q12 = 0, Q20, Q21 = 0, Q30 = 0, num;
+  int DC01, DC02, DC03, DC04, DC05, DC06, DC07, DC08, DC09, DC10, DC11, DC12,
+      DC13, DC14, DC15, DC16, DC17, DC18, DC19, DC20, DC21, DC22, DC23, DC24,
+      DC25;
  int Al, pred;

  /* Keep a local variable to avoid looking it up more than once */
@@ -434,10 +457,10 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
    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
+       * we want it to keep two rows ahead so that next two block rows' DC
       * values are up to date.
       */
-      JDIMENSION delta = (cinfo->Ss == 0) ? 1 : 0;
+      JDIMENSION delta = (cinfo->Ss == 0) ? 2 : 0;
      if (cinfo->input_iMCU_row > cinfo->output_iMCU_row + delta)
        break;
    }
@@ -452,34 +475,53 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
    if (!compptr->component_needed)
      continue;
    /* Count non-dummy DCT block rows in this iMCU row. */
-    if (cinfo->output_iMCU_row < last_iMCU_row) {
+    if (cinfo->output_iMCU_row < last_iMCU_row - 1) {
+      block_rows = compptr->v_samp_factor;
+      access_rows = block_rows * 3; /* this and next two iMCU rows */
+    } else 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 */
+    if (cinfo->output_iMCU_row > 1) {
+      access_rows += 2 * compptr->v_samp_factor; /* prior two iMCU rows too */
+      buffer = (*cinfo->mem->access_virt_barray)
+        ((j_common_ptr)cinfo, coef->whole_image[ci],
+         (cinfo->output_iMCU_row - 2) * compptr->v_samp_factor,
+         (JDIMENSION)access_rows, FALSE);
+      buffer += 2 * compptr->v_samp_factor; /* point to current iMCU row */
+    } else if (cinfo->output_iMCU_row > 0) {
      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);
+    /* Fetch component-dependent info.
+     * If the current scan is incomplete, then we use the component-dependent
+     * info from the previous scan.
+     */
+    if (cinfo->output_iMCU_row > cinfo->master->last_good_iMCU_row)
+      coef_bits =
+        coef->coef_bits_latch + ((ci + cinfo->num_components) * SAVED_COEFS);
+    else
+      coef_bits = coef->coef_bits_latch + (ci * SAVED_COEFS);
+
+    /* We only do DC interpolation if no AC coefficient data is available. */
+    change_dc =
+      coef_bits[1] == -1 && coef_bits[2] == -1 && coef_bits[3] == -1 &&
+      coef_bits[4] == -1 && coef_bits[5] == -1 && coef_bits[6] == -1 &&
+      coef_bits[7] == -1 && coef_bits[8] == -1 && coef_bits[9] == -1;
+
    quanttbl = compptr->quant_table;
    Q00 = quanttbl->quantval[0];
    Q01 = quanttbl->quantval[Q01_POS];
@@ -487,27 +529,51 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
    Q20 = quanttbl->quantval[Q20_POS];
    Q11 = quanttbl->quantval[Q11_POS];
    Q02 = quanttbl->quantval[Q02_POS];
+    if (change_dc) {
+      Q03 = quanttbl->quantval[Q03_POS];
+      Q12 = quanttbl->quantval[Q12_POS];
+      Q21 = quanttbl->quantval[Q21_POS];
+      Q30 = quanttbl->quantval[Q30_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)
+
+      if (block_row > 0 || cinfo->output_iMCU_row > 0)
+        prev_block_row =
+          buffer[block_row - 1] + cinfo->master->first_MCU_col[ci];
+      else
        prev_block_row = buffer_ptr;
+
+      if (block_row > 1 || cinfo->output_iMCU_row > 1)
+        prev_prev_block_row =
+          buffer[block_row - 2] + cinfo->master->first_MCU_col[ci];
+      else
+        prev_prev_block_row = prev_block_row;
+
+      if (block_row < block_rows - 1 || cinfo->output_iMCU_row < last_iMCU_row)
+        next_block_row =
+          buffer[block_row + 1] + cinfo->master->first_MCU_col[ci];
      else
-        prev_block_row = buffer[block_row - 1] +
-                         cinfo->master->first_MCU_col[ci];
-      if (last_row && block_row == block_rows - 1)
        next_block_row = buffer_ptr;
+
+      if (block_row < block_rows - 2 ||
+          cinfo->output_iMCU_row < last_iMCU_row - 1)
+        next_next_block_row =
+          buffer[block_row + 2] + cinfo->master->first_MCU_col[ci];
      else
-        next_block_row = buffer[block_row + 1] +
-                         cinfo->master->first_MCU_col[ci];
+        next_next_block_row = next_block_row;
+
      /* 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.
+       * Initialize all 25 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];
+      DC01 = DC02 = DC03 = DC04 = DC05 = (int)prev_prev_block_row[0][0];
+      DC06 = DC07 = DC08 = DC09 = DC10 = (int)prev_block_row[0][0];
+      DC11 = DC12 = DC13 = DC14 = DC15 = (int)buffer_ptr[0][0];
+      DC16 = DC17 = DC18 = DC19 = DC20 = (int)next_block_row[0][0];
+      DC21 = DC22 = DC23 = DC24 = DC25 = (int)next_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];
@@ -515,18 +581,39 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
        /* 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];
+        if (block_num == cinfo->master->first_MCU_col[ci] &&
+            block_num < last_block_column) {
+          DC04 = (int)prev_prev_block_row[1][0];
+          DC09 = (int)prev_block_row[1][0];
+          DC14 = (int)buffer_ptr[1][0];
+          DC19 = (int)next_block_row[1][0];
+          DC24 = (int)next_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.
+        if (block_num + 1 < last_block_column) {
+          DC05 = (int)prev_prev_block_row[2][0];
+          DC10 = (int)prev_block_row[2][0];
+          DC15 = (int)buffer_ptr[2][0];
+          DC20 = (int)next_block_row[2][0];
+          DC25 = (int)next_next_block_row[2][0];
+        }
+        /* If DC interpolation is enabled, compute coefficient estimates using
+         * a Gaussian-like kernel, keeping the averages of the DC values.
+         *
+         * If DC interpolation is disabled, compute coefficient estimates using
+         * an algorithm similar to the one described in Section K.8 of the JPEG
+         * standard, except applied to a 5x5 window rather than a 3x3 window.
+         *
+         * An estimate is applied only if the 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);
+          num = Q00 * (change_dc ?
+                (-DC01 - DC02 + DC04 + DC05 - 3 * DC06 + 13 * DC07 -
+                 13 * DC09 + 3 * DC10 - 3 * DC11 + 38 * DC12 - 38 * DC14 +
+                 3 * DC15 - 3 * DC16 + 13 * DC17 - 13 * DC19 + 3 * DC20 -
+                 DC21 - DC22 + DC24 + DC25) :
+                (-7 * DC11 + 50 * DC12 - 50 * DC14 + 7 * DC15));
          if (num >= 0) {
            pred = (int)(((Q01 << 7) + num) / (Q01 << 8));
            if (Al > 0 && pred >= (1 << Al))
@@ -541,7 +628,12 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
        }
        /* AC10 */
        if ((Al = coef_bits[2]) != 0 && workspace[8] == 0) {
-          num = 36 * Q00 * (DC2 - DC8);
+          num = Q00 * (change_dc ?
+                (-DC01 - 3 * DC02 - 3 * DC03 - 3 * DC04 - DC05 - DC06 +
+                 13 * DC07 + 38 * DC08 + 13 * DC09 - DC10 + DC16 -
+                 13 * DC17 - 38 * DC18 - 13 * DC19 + DC20 + DC21 +
+                 3 * DC22 + 3 * DC23 + 3 * DC24 + DC25) :
+                (-7 * DC03 + 50 * DC08 - 50 * DC18 + 7 * DC23));
          if (num >= 0) {
            pred = (int)(((Q10 << 7) + num) / (Q10 << 8));
            if (Al > 0 && pred >= (1 << Al))
@@ -556,7 +648,10 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
        }
        /* AC20 */
        if ((Al = coef_bits[3]) != 0 && workspace[16] == 0) {
-          num = 9 * Q00 * (DC2 + DC8 - 2 * DC5);
+          num = Q00 * (change_dc ?
+                (DC03 + 2 * DC07 + 7 * DC08 + 2 * DC09 - 5 * DC12 - 14 * DC13 -
+                 5 * DC14 + 2 * DC17 + 7 * DC18 + 2 * DC19 + DC23) :
+                (-DC03 + 13 * DC08 - 24 * DC13 + 13 * DC18 - DC23));
          if (num >= 0) {
            pred = (int)(((Q20 << 7) + num) / (Q20 << 8));
            if (Al > 0 && pred >= (1 << Al))
@@ -571,7 +666,11 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
        }
        /* AC11 */
        if ((Al = coef_bits[4]) != 0 && workspace[9] == 0) {
-          num = 5 * Q00 * (DC1 - DC3 - DC7 + DC9);
+          num = Q00 * (change_dc ?
+                (-DC01 + DC05 + 9 * DC07 - 9 * DC09 - 9 * DC17 +
+                 9 * DC19 + DC21 - DC25) :
+                (DC10 + DC16 - 10 * DC17 + 10 * DC19 - DC02 - DC20 + DC22 -
+                 DC24 + DC04 - DC06 + 10 * DC07 - 10 * DC09));
          if (num >= 0) {
            pred = (int)(((Q11 << 7) + num) / (Q11 << 8));
            if (Al > 0 && pred >= (1 << Al))
@@ -586,7 +685,10 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
        }
        /* AC02 */
        if ((Al = coef_bits[5]) != 0 && workspace[2] == 0) {
-          num = 9 * Q00 * (DC4 + DC6 - 2 * DC5);
+          num = Q00 * (change_dc ?
+                (2 * DC07 - 5 * DC08 + 2 * DC09 + DC11 + 7 * DC12 - 14 * DC13 +
+                 7 * DC14 + DC15 + 2 * DC17 - 5 * DC18 + 2 * DC19) :
+                (-DC11 + 13 * DC12 - 24 * DC13 + 13 * DC14 - DC15));
          if (num >= 0) {
            pred = (int)(((Q02 << 7) + num) / (Q02 << 8));
            if (Al > 0 && pred >= (1 << Al))
@@ -599,14 +701,96 @@ decompress_smooth_data(j_decompress_ptr cinfo, JSAMPIMAGE output_buf)
          }
          workspace[2] = (JCOEF)pred;
        }
+        if (change_dc) {
+          /* AC03 */
+          if ((Al = coef_bits[6]) != 0 && workspace[3] == 0) {
+            num = Q00 * (DC07 - DC09 + 2 * DC12 - 2 * DC14 + DC17 - DC19);
+            if (num >= 0) {
+              pred = (int)(((Q03 << 7) + num) / (Q03 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+            } else {
+              pred = (int)(((Q03 << 7) - num) / (Q03 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+              pred = -pred;
+            }
+            workspace[3] = (JCOEF)pred;
+          }
+          /* AC12 */
+          if ((Al = coef_bits[7]) != 0 && workspace[10] == 0) {
+            num = Q00 * (DC07 - 3 * DC08 + DC09 - DC17 + 3 * DC18 - DC19);
+            if (num >= 0) {
+              pred = (int)(((Q12 << 7) + num) / (Q12 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+            } else {
+              pred = (int)(((Q12 << 7) - num) / (Q12 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+              pred = -pred;
+            }
+            workspace[10] = (JCOEF)pred;
+          }
+          /* AC21 */
+          if ((Al = coef_bits[8]) != 0 && workspace[17] == 0) {
+            num = Q00 * (DC07 - DC09 - 3 * DC12 + 3 * DC14 + DC17 - DC19);
+            if (num >= 0) {
+              pred = (int)(((Q21 << 7) + num) / (Q21 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+            } else {
+              pred = (int)(((Q21 << 7) - num) / (Q21 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+              pred = -pred;
+            }
+            workspace[17] = (JCOEF)pred;
+          }
+          /* AC30 */
+          if ((Al = coef_bits[9]) != 0 && workspace[24] == 0) {
+            num = Q00 * (DC07 + 2 * DC08 + DC09 - DC17 - 2 * DC18 - DC19);
+            if (num >= 0) {
+              pred = (int)(((Q30 << 7) + num) / (Q30 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+            } else {
+              pred = (int)(((Q30 << 7) - num) / (Q30 << 8));
+              if (Al > 0 && pred >= (1 << Al))
+                pred = (1 << Al) - 1;
+              pred = -pred;
+            }
+            workspace[24] = (JCOEF)pred;
+          }
+          /* coef_bits[0] is non-negative.  Otherwise this function would not
+           * be called.
+           */
+          num = Q00 *
+                (-2 * DC01 - 6 * DC02 - 8 * DC03 - 6 * DC04 - 2 * DC05 -
+                 6 * DC06 + 6 * DC07 + 42 * DC08 + 6 * DC09 - 6 * DC10 -
+                 8 * DC11 + 42 * DC12 + 152 * DC13 + 42 * DC14 - 8 * DC15 -
+                 6 * DC16 + 6 * DC17 + 42 * DC18 + 6 * DC19 - 6 * DC20 -
+                 2 * DC21 - 6 * DC22 - 8 * DC23 - 6 * DC24 - 2 * DC25);
+          if (num >= 0) {
+            pred = (int)(((Q00 << 7) + num) / (Q00 << 8));
+          } else {
+            pred = (int)(((Q00 << 7) - num) / (Q00 << 8));
+            pred = -pred;
+          }
+          workspace[0] = (JCOEF)pred;
+        }  /* change_dc */
+
        /* 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++;
+        DC01 = DC02;  DC02 = DC03;  DC03 = DC04;  DC04 = DC05;
+        DC06 = DC07;  DC07 = DC08;  DC08 = DC09;  DC09 = DC10;
+        DC11 = DC12;  DC12 = DC13;  DC13 = DC14;  DC14 = DC15;
+        DC16 = DC17;  DC17 = DC18;  DC18 = DC19;  DC19 = DC20;
+        DC21 = DC22;  DC22 = DC23;  DC23 = DC24;  DC24 = DC25;
+        buffer_ptr++, prev_block_row++, next_block_row++,
+          prev_prev_block_row++, next_next_block_row++;
        output_col += compptr->_DCT_scaled_size;
      }
      output_ptr += compptr->_DCT_scaled_size;
@@ -655,7 +839,7 @@ jinit_d_coef_controller(j_decompress_ptr cinfo, boolean need_full_buffer)
 #ifdef BLOCK_SMOOTHING_SUPPORTED
      /* If block smoothing could be used, need a bigger window */
      if (cinfo->progressive_mode)
-        access_rows *= 3;
+        access_rows *= 5;
 #endif
      coef->whole_image[ci] = (*cinfo->mem->request_virt_barray)
        ((j_common_ptr)cinfo, JPOOL_IMAGE, TRUE,
@@ -5,6 +5,7 @@
 * Copyright (C) 1994-1997, Thomas G. Lane.
 * libjpeg-turbo Modifications:
 * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * Copyright (C) 2020, Google, Inc.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 */
@@ -51,7 +52,7 @@ typedef struct {
 #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] */
+#define SAVED_COEFS  10         /* we save coef_bits[0..9] */
 #endif
 } my_coef_controller;

@@ -45,9 +45,9 @@ ycc_rgb565_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    outptr = *output_buf++;

    if (PACK_NEED_ALIGNMENT(outptr)) {
-      y  = GETJSAMPLE(*inptr0++);
-      cb = GETJSAMPLE(*inptr1++);
-      cr = GETJSAMPLE(*inptr2++);
+      y  = *inptr0++;
+      cb = *inptr1++;
+      cr = *inptr2++;
      r = range_limit[y + Crrtab[cr]];
      g = range_limit[y + ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
                                            SCALEBITS))];
@@ -58,18 +58,18 @@ ycc_rgb565_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
      num_cols--;
    }
    for (col = 0; col < (num_cols >> 1); col++) {
-      y  = GETJSAMPLE(*inptr0++);
-      cb = GETJSAMPLE(*inptr1++);
-      cr = GETJSAMPLE(*inptr2++);
+      y  = *inptr0++;
+      cb = *inptr1++;
+      cr = *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++);
+      y  = *inptr0++;
+      cb = *inptr1++;
+      cr = *inptr2++;
      r = range_limit[y + Crrtab[cr]];
      g = range_limit[y + ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
                                            SCALEBITS))];
@@ -80,9 +80,9 @@ ycc_rgb565_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
      outptr += 4;
    }
    if (num_cols & 1) {
-      y  = GETJSAMPLE(*inptr0);
-      cb = GETJSAMPLE(*inptr1);
-      cr = GETJSAMPLE(*inptr2);
+      y  = *inptr0;
+      cb = *inptr1;
+      cr = *inptr2;
      r = range_limit[y + Crrtab[cr]];
      g = range_limit[y + ((int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr],
                                            SCALEBITS))];
@@ -125,9 +125,9 @@ ycc_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    input_row++;
    outptr = *output_buf++;
    if (PACK_NEED_ALIGNMENT(outptr)) {
-      y  = GETJSAMPLE(*inptr0++);
-      cb = GETJSAMPLE(*inptr1++);
-      cr = GETJSAMPLE(*inptr2++);
+      y  = *inptr0++;
+      cb = *inptr1++;
+      cr = *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],
@@ -139,9 +139,9 @@ ycc_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
      num_cols--;
    }
    for (col = 0; col < (num_cols >> 1); col++) {
-      y  = GETJSAMPLE(*inptr0++);
-      cb = GETJSAMPLE(*inptr1++);
-      cr = GETJSAMPLE(*inptr2++);
+      y  = *inptr0++;
+      cb = *inptr1++;
+      cr = *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],
@@ -150,9 +150,9 @@ ycc_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
      d0 = DITHER_ROTATE(d0);
      rgb = PACK_SHORT_565(r, g, b);

-      y  = GETJSAMPLE(*inptr0++);
-      cb = GETJSAMPLE(*inptr1++);
-      cr = GETJSAMPLE(*inptr2++);
+      y  = *inptr0++;
+      cb = *inptr1++;
+      cr = *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],
@@ -165,9 +165,9 @@ ycc_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
      outptr += 4;
    }
    if (num_cols & 1) {
-      y  = GETJSAMPLE(*inptr0);
-      cb = GETJSAMPLE(*inptr1);
-      cr = GETJSAMPLE(*inptr2);
+      y  = *inptr0;
+      cb = *inptr1;
+      cr = *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],
@@ -202,32 +202,32 @@ rgb_rgb565_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    input_row++;
    outptr = *output_buf++;
    if (PACK_NEED_ALIGNMENT(outptr)) {
-      r = GETJSAMPLE(*inptr0++);
-      g = GETJSAMPLE(*inptr1++);
-      b = GETJSAMPLE(*inptr2++);
+      r = *inptr0++;
+      g = *inptr1++;
+      b = *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++);
+      r = *inptr0++;
+      g = *inptr1++;
+      b = *inptr2++;
      rgb = PACK_SHORT_565(r, g, b);

-      r = GETJSAMPLE(*inptr0++);
-      g = GETJSAMPLE(*inptr1++);
-      b = GETJSAMPLE(*inptr2++);
+      r = *inptr0++;
+      g = *inptr1++;
+      b = *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);
+      r = *inptr0;
+      g = *inptr1;
+      b = *inptr2;
      rgb = PACK_SHORT_565(r, g, b);
      *(INT16 *)outptr = (INT16)rgb;
    }
@@ -259,24 +259,24 @@ rgb_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    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)];
+      r = range_limit[DITHER_565_R(*inptr0++, d0)];
+      g = range_limit[DITHER_565_G(*inptr1++, d0)];
+      b = range_limit[DITHER_565_B(*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)];
+      r = range_limit[DITHER_565_R(*inptr0++, d0)];
+      g = range_limit[DITHER_565_G(*inptr1++, d0)];
+      b = range_limit[DITHER_565_B(*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)];
+      r = range_limit[DITHER_565_R(*inptr0++, d0)];
+      g = range_limit[DITHER_565_G(*inptr1++, d0)];
+      b = range_limit[DITHER_565_B(*inptr2++, d0)];
      d0 = DITHER_ROTATE(d0);
      rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r, g, b));

@@ -284,9 +284,9 @@ rgb_rgb565D_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
      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)];
+      r = range_limit[DITHER_565_R(*inptr0, d0)];
+      g = range_limit[DITHER_565_G(*inptr1, d0)];
+      b = range_limit[DITHER_565_B(*inptr2, d0)];
      rgb = PACK_SHORT_565(r, g, b);
      *(INT16 *)outptr = (INT16)rgb;
    }
@@ -53,9 +53,9 @@ ycc_rgb_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    input_row++;
    outptr = *output_buf++;
    for (col = 0; col < num_cols; col++) {
-      y  = GETJSAMPLE(inptr0[col]);
-      cb = GETJSAMPLE(inptr1[col]);
-      cr = GETJSAMPLE(inptr2[col]);
+      y  = inptr0[col];
+      cb = inptr1[col];
+      cr = 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 +
@@ -93,7 +93,6 @@ gray_rgb_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    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 */
@@ -128,7 +127,6 @@ rgb_rgb_convert_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    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];
@@ -341,9 +341,9 @@ rgb_gray_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    input_row++;
    outptr = *output_buf++;
    for (col = 0; col < num_cols; col++) {
-      r = GETJSAMPLE(inptr0[col]);
-      g = GETJSAMPLE(inptr1[col]);
-      b = GETJSAMPLE(inptr2[col]);
+      r = inptr0[col];
+      g = inptr1[col];
+      b = inptr2[col];
      /* Y */
      outptr[col] = (JSAMPLE)((ctab[r + R_Y_OFF] + ctab[g + G_Y_OFF] +
                               ctab[b + B_Y_OFF]) >> SCALEBITS);
@@ -550,9 +550,9 @@ ycck_cmyk_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    input_row++;
    outptr = *output_buf++;
    for (col = 0; col < num_cols; col++) {
-      y  = GETJSAMPLE(inptr0[col]);
-      cb = GETJSAMPLE(inptr1[col]);
-      cr = GETJSAMPLE(inptr2[col]);
+      y  = inptr0[col];
+      cb = inptr1[col];
+      cr = 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 */
@@ -560,7 +560,7 @@ ycck_cmyk_convert(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
                                                 SCALEBITS)))];
      outptr[2] = range_limit[MAXJSAMPLE - (y + Cbbtab[cb])];   /* blue */
      /* K passes through unchanged */
-      outptr[3] = inptr3[col];  /* don't need GETJSAMPLE here */
+      outptr[3] = inptr3[col];
      outptr += 4;
    }
  }
@@ -5,6 +5,7 @@
 * Copyright (C) 1991-1997, Thomas G. Lane.
 * libjpeg-turbo Modifications:
 * Copyright (C) 2009-2011, 2016, 2018-2019, D. R. Commander.
+ * Copyright (C) 2018, Matthias Räncker.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -39,24 +40,6 @@ typedef struct {
  int last_dc_val[MAX_COMPS_IN_SCAN]; /* last DC coef for each component */
 } savable_state;

-/* This macro is to work around compilers with missing or broken
- * structure assignment.  You'll need to fix this code if you have
- * such a compiler and you change MAX_COMPS_IN_SCAN.
- */
-
-#ifndef NO_STRUCT_ASSIGN
-#define ASSIGN_STATE(dest, src)  ((dest) = (src))
-#else
-#if MAX_COMPS_IN_SCAN == 4
-#define ASSIGN_STATE(dest, src) \
-  ((dest).last_dc_val[0] = (src).last_dc_val[0], \
-   (dest).last_dc_val[1] = (src).last_dc_val[1], \
-   (dest).last_dc_val[2] = (src).last_dc_val[2], \
-   (dest).last_dc_val[3] = (src).last_dc_val[3])
-#endif
-#endif
-
-
 typedef struct {
  struct jpeg_entropy_decoder pub; /* public fields */

@@ -325,7 +308,7 @@ jpeg_fill_bit_buffer(bitread_working_state *state,
        bytes_in_buffer = cinfo->src->bytes_in_buffer;
      }
      bytes_in_buffer--;
-      c = GETJOCTET(*next_input_byte++);
+      c = *next_input_byte++;

      /* If it's 0xFF, check and discard stuffed zero byte */
      if (c == 0xFF) {
@@ -342,7 +325,7 @@ jpeg_fill_bit_buffer(bitread_working_state *state,
            bytes_in_buffer = cinfo->src->bytes_in_buffer;
          }
          bytes_in_buffer--;
-          c = GETJOCTET(*next_input_byte++);
+          c = *next_input_byte++;
        } while (c == 0xFF);

        if (c == 0) {
@@ -405,8 +388,8 @@ no_more_bytes:

 #define GET_BYTE { \
  register int c0, c1; \
-  c0 = GETJOCTET(*buffer++); \
-  c1 = GETJOCTET(*buffer); \
+  c0 = *buffer++; \
+  c1 = *buffer; \
  /* Pre-execute most common case */ \
  get_buffer = (get_buffer << 8) | c0; \
  bits_left += 8; \
@@ -423,7 +406,7 @@ no_more_bytes:
  } \
 }

-#if SIZEOF_SIZE_T == 8 || defined(_WIN64)
+#if SIZEOF_SIZE_T == 8 || defined(_WIN64) || (defined(__x86_64__) && defined(__ILP32__))

 /* Pre-fetch 48 bytes, because the holding register is 64-bit */
 #define FILL_BIT_BUFFER_FAST \
@@ -557,6 +540,12 @@ process_restart(j_decompress_ptr cinfo)
 }


+#if defined(__has_feature)
+#if __has_feature(undefined_behavior_sanitizer)
+__attribute__((no_sanitize("signed-integer-overflow"),
+               no_sanitize("unsigned-integer-overflow")))
+#endif
+#endif
 LOCAL(boolean)
 decode_mcu_slow(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
 {
@@ -568,7 +557,7 @@ decode_mcu_slow(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)

  /* Load up working state */
  BITREAD_LOAD_STATE(cinfo, entropy->bitstate);
-  ASSIGN_STATE(state, entropy->saved);
+  state = entropy->saved;

  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
    JBLOCKROW block = MCU_data ? MCU_data[blkn] : NULL;
@@ -589,11 +578,15 @@ decode_mcu_slow(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
    if (entropy->dc_needed[blkn]) {
      /* Convert DC difference to actual value, update last_dc_val */
      int ci = cinfo->MCU_membership[blkn];
-      /* This is really just
-       *   s += state.last_dc_val[ci];
-       * It is written this way in order to shut up UBSan.
+      /* Certain malformed JPEG images produce repeated DC coefficient
+       * differences of 2047 or -2047, which causes state.last_dc_val[ci] to
+       * grow until it overflows or underflows a 32-bit signed integer.  This
+       * behavior is, to the best of our understanding, innocuous, and it is
+       * unclear how to work around it without potentially affecting
+       * performance.  Thus, we (hopefully temporarily) suppress UBSan integer
+       * overflow errors for this function and decode_mcu_fast().
       */
-      s = (int)((unsigned int)s + (unsigned int)state.last_dc_val[ci]);
+      s += state.last_dc_val[ci];
      state.last_dc_val[ci] = s;
      if (block) {
        /* Output the DC coefficient (assumes jpeg_natural_order[0] = 0) */
@@ -653,11 +646,17 @@ decode_mcu_slow(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)

  /* Completed MCU, so update state */
  BITREAD_SAVE_STATE(cinfo, entropy->bitstate);
-  ASSIGN_STATE(entropy->saved, state);
+  entropy->saved = state;
  return TRUE;
 }


+#if defined(__has_feature)
+#if __has_feature(undefined_behavior_sanitizer)
+__attribute__((no_sanitize("signed-integer-overflow"),
+               no_sanitize("unsigned-integer-overflow")))
+#endif
+#endif
 LOCAL(boolean)
 decode_mcu_fast(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
 {
@@ -671,7 +670,7 @@ decode_mcu_fast(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
  /* Load up working state */
  BITREAD_LOAD_STATE(cinfo, entropy->bitstate);
  buffer = (JOCTET *)br_state.next_input_byte;
-  ASSIGN_STATE(state, entropy->saved);
+  state = entropy->saved;

  for (blkn = 0; blkn < cinfo->blocks_in_MCU; blkn++) {
    JBLOCKROW block = MCU_data ? MCU_data[blkn] : NULL;
@@ -688,7 +687,10 @@ decode_mcu_fast(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)

    if (entropy->dc_needed[blkn]) {
      int ci = cinfo->MCU_membership[blkn];
-      s = (int)((unsigned int)s + (unsigned int)state.last_dc_val[ci]);
+      /* Refer to the comment in decode_mcu_slow() regarding the supression of
+       * a UBSan integer overflow error in this line of code.
+       */
+      s += state.last_dc_val[ci];
      state.last_dc_val[ci] = s;
      if (block)
        (*block)[0] = (JCOEF)s;
@@ -740,7 +742,7 @@ decode_mcu_fast(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
  br_state.bytes_in_buffer -= (buffer - br_state.next_input_byte);
  br_state.next_input_byte = buffer;
  BITREAD_SAVE_STATE(cinfo, entropy->bitstate);
-  ASSIGN_STATE(entropy->saved, state);
+  entropy->saved = state;
  return TRUE;
 }

@@ -795,7 +797,8 @@ use_slow:
  }

  /* Account for restart interval (no-op if not using restarts) */
-  entropy->restarts_to_go--;
+  if (cinfo->restart_interval)
+    entropy->restarts_to_go--;

  return TRUE;
 }
@@ -4,7 +4,8 @@
 * This file was part of the Independent JPEG Group's software:
 * Copyright (C) 1991-1997, Thomas G. Lane.
 * libjpeg-turbo Modifications:
- * Copyright (C) 2010-2011, 2015-2016, D. R. Commander.
+ * Copyright (C) 2010-2011, 2015-2016, 2021, D. R. Commander.
+ * Copyright (C) 2018, Matthias Räncker.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -78,6 +79,11 @@ EXTERN(void) jpeg_make_d_derived_tbl(j_decompress_ptr cinfo, boolean isDC,
 typedef size_t bit_buf_type;            /* type of bit-extraction buffer */
 #define BIT_BUF_SIZE  64                /* size of buffer in bits */

+#elif defined(__x86_64__) && defined(__ILP32__)
+
+typedef unsigned long long bit_buf_type; /* type of bit-extraction buffer */
+#define BIT_BUF_SIZE  64                 /* size of buffer in bits */
+
 #else

 typedef unsigned long bit_buf_type;     /* type of bit-extraction buffer */
@@ -228,7 +234,10 @@ slowlabel: \
      s |= GET_BITS(1); \
      nb++; \
    } \
-    s = htbl->pub->huffval[(int)(s + htbl->valoffset[nb]) & 0xFF]; \
+    if (nb > 16) \
+      s = 0; \
+    else \
+      s = htbl->pub->huffval[(int)(s + htbl->valoffset[nb]) & 0xFF]; \
  }

 /* Out-of-line case for Huffman code fetching */
@@ -38,18 +38,18 @@ marker_is_icc(jpeg_saved_marker_ptr marker)
    marker->marker == ICC_MARKER &&
    marker->data_length >= ICC_OVERHEAD_LEN &&
    /* verify the identifying string */
-    GETJOCTET(marker->data[0]) == 0x49 &&
-    GETJOCTET(marker->data[1]) == 0x43 &&
-    GETJOCTET(marker->data[2]) == 0x43 &&
-    GETJOCTET(marker->data[3]) == 0x5F &&
-    GETJOCTET(marker->data[4]) == 0x50 &&
-    GETJOCTET(marker->data[5]) == 0x52 &&
-    GETJOCTET(marker->data[6]) == 0x4F &&
-    GETJOCTET(marker->data[7]) == 0x46 &&
-    GETJOCTET(marker->data[8]) == 0x49 &&
-    GETJOCTET(marker->data[9]) == 0x4C &&
-    GETJOCTET(marker->data[10]) == 0x45 &&
-    GETJOCTET(marker->data[11]) == 0x0;
+    marker->data[0] == 0x49 &&
+    marker->data[1] == 0x43 &&
+    marker->data[2] == 0x43 &&
+    marker->data[3] == 0x5F &&
+    marker->data[4] == 0x50 &&
+    marker->data[5] == 0x52 &&
+    marker->data[6] == 0x4F &&
+    marker->data[7] == 0x46 &&
+    marker->data[8] == 0x49 &&
+    marker->data[9] == 0x4C &&
+    marker->data[10] == 0x45 &&
+    marker->data[11] == 0x0;
 }


@@ -102,12 +102,12 @@ jpeg_read_icc_profile(j_decompress_ptr cinfo, JOCTET **icc_data_ptr,
  for (marker = cinfo->marker_list; marker != NULL; marker = marker->next) {
    if (marker_is_icc(marker)) {
      if (num_markers == 0)
-        num_markers = GETJOCTET(marker->data[13]);
-      else if (num_markers != GETJOCTET(marker->data[13])) {
+        num_markers = marker->data[13];
+      else if (num_markers != marker->data[13]) {
        WARNMS(cinfo, JWRN_BOGUS_ICC);  /* inconsistent num_markers fields */
        return FALSE;
      }
-      seq_no = GETJOCTET(marker->data[12]);
+      seq_no = marker->data[12];
      if (seq_no <= 0 || seq_no > num_markers) {
        WARNMS(cinfo, JWRN_BOGUS_ICC);  /* bogus sequence number */
        return FALSE;
@@ -154,7 +154,7 @@ jpeg_read_icc_profile(j_decompress_ptr cinfo, JOCTET **icc_data_ptr,
      JOCTET FAR *src_ptr;
      JOCTET *dst_ptr;
      unsigned int length;
-      seq_no = GETJOCTET(marker->data[12]);
+      seq_no = marker->data[12];
      dst_ptr = icc_data + data_offset[seq_no];
      src_ptr = marker->data + ICC_OVERHEAD_LEN;
      length = data_length[seq_no];
@@ -18,6 +18,7 @@

 #include "jinclude.h"
 #include "jdmainct.h"
+#include "jconfigint.h"


 /*
@@ -360,7 +361,7 @@ process_data_context_main(j_decompress_ptr cinfo, JSAMPARRAY output_buf,
    main_ptr->context_state = CTX_PREPARE_FOR_IMCU;
    if (*out_row_ctr >= out_rows_avail)
      return;                   /* Postprocessor exactly filled output buf */
-    /*FALLTHROUGH*/
+    FALLTHROUGH                 /*FALLTHROUGH*/
  case CTX_PREPARE_FOR_IMCU:
    /* Prepare to process first M-1 row groups of this iMCU row */
    main_ptr->rowgroup_ctr = 0;
@@ -371,7 +372,7 @@ process_data_context_main(j_decompress_ptr cinfo, JSAMPARRAY output_buf,
    if (main_ptr->iMCU_row_ctr == cinfo->total_iMCU_rows)
      set_bottom_pointers(cinfo);
    main_ptr->context_state = CTX_PROCESS_IMCU;
-    /*FALLTHROUGH*/
+    FALLTHROUGH                 /*FALLTHROUGH*/
  case CTX_PROCESS_IMCU:
    /* Call postprocessor using previously set pointers */
    (*cinfo->post->post_process_data) (cinfo,
@@ -151,7 +151,7 @@ typedef my_marker_reader *my_marker_ptr;
 #define INPUT_BYTE(cinfo, V, action) \
  MAKESTMT( MAKE_BYTE_AVAIL(cinfo, action); \
            bytes_in_buffer--; \
-            V = GETJOCTET(*next_input_byte++); )
+            V = *next_input_byte++; )

 /* As above, but read two bytes interpreted as an unsigned 16-bit integer.
 * V should be declared unsigned int or perhaps JLONG.
@@ -159,10 +159,10 @@ typedef my_marker_reader *my_marker_ptr;
 #define INPUT_2BYTES(cinfo, V, action) \
  MAKESTMT( MAKE_BYTE_AVAIL(cinfo, action); \
            bytes_in_buffer--; \
-            V = ((unsigned int)GETJOCTET(*next_input_byte++)) << 8; \
+            V = ((unsigned int)(*next_input_byte++)) << 8; \
            MAKE_BYTE_AVAIL(cinfo, action); \
            bytes_in_buffer--; \
-            V += GETJOCTET(*next_input_byte++); )
+            V += *next_input_byte++; )


 /*
@@ -608,18 +608,18 @@ examine_app0(j_decompress_ptr cinfo, JOCTET *data, unsigned int datalen,
  JLONG totallen = (JLONG)datalen + remaining;

  if (datalen >= APP0_DATA_LEN &&
-      GETJOCTET(data[0]) == 0x4A &&
-      GETJOCTET(data[1]) == 0x46 &&
-      GETJOCTET(data[2]) == 0x49 &&
-      GETJOCTET(data[3]) == 0x46 &&
-      GETJOCTET(data[4]) == 0) {
+      data[0] == 0x4A &&
+      data[1] == 0x46 &&
+      data[2] == 0x49 &&
+      data[3] == 0x46 &&
+      data[4] == 0) {
    /* Found JFIF APP0 marker: save info */
    cinfo->saw_JFIF_marker = TRUE;
-    cinfo->JFIF_major_version = GETJOCTET(data[5]);
-    cinfo->JFIF_minor_version = GETJOCTET(data[6]);
-    cinfo->density_unit = GETJOCTET(data[7]);
-    cinfo->X_density = (GETJOCTET(data[8]) << 8) + GETJOCTET(data[9]);
-    cinfo->Y_density = (GETJOCTET(data[10]) << 8) + GETJOCTET(data[11]);
+    cinfo->JFIF_major_version = data[5];
+    cinfo->JFIF_minor_version = data[6];
+    cinfo->density_unit = data[7];
+    cinfo->X_density = (data[8] << 8) + data[9];
+    cinfo->Y_density = (data[10] << 8) + data[11];
    /* Check version.
     * Major version must be 1, anything else signals an incompatible change.
     * (We used to treat this as an error, but now it's a nonfatal warning,
@@ -634,24 +634,22 @@ examine_app0(j_decompress_ptr cinfo, JOCTET *data, unsigned int datalen,
             cinfo->JFIF_major_version, cinfo->JFIF_minor_version,
             cinfo->X_density, cinfo->Y_density, cinfo->density_unit);
    /* Validate thumbnail dimensions and issue appropriate messages */
-    if (GETJOCTET(data[12]) | GETJOCTET(data[13]))
-      TRACEMS2(cinfo, 1, JTRC_JFIF_THUMBNAIL,
-               GETJOCTET(data[12]), GETJOCTET(data[13]));
+    if (data[12] | data[13])
+      TRACEMS2(cinfo, 1, JTRC_JFIF_THUMBNAIL, data[12], data[13]);
    totallen -= APP0_DATA_LEN;
-    if (totallen !=
-        ((JLONG)GETJOCTET(data[12]) * (JLONG)GETJOCTET(data[13]) * (JLONG)3))
+    if (totallen != ((JLONG)data[12] * (JLONG)data[13] * (JLONG)3))
      TRACEMS1(cinfo, 1, JTRC_JFIF_BADTHUMBNAILSIZE, (int)totallen);
  } else if (datalen >= 6 &&
-             GETJOCTET(data[0]) == 0x4A &&
-             GETJOCTET(data[1]) == 0x46 &&
-             GETJOCTET(data[2]) == 0x58 &&
-             GETJOCTET(data[3]) == 0x58 &&
-             GETJOCTET(data[4]) == 0) {
+             data[0] == 0x4A &&
+             data[1] == 0x46 &&
+             data[2] == 0x58 &&
+             data[3] == 0x58 &&
+             data[4] == 0) {
    /* Found JFIF "JFXX" extension APP0 marker */
    /* The library doesn't actually do anything with these,
     * but we try to produce a helpful trace message.
     */
-    switch (GETJOCTET(data[5])) {
+    switch (data[5]) {
    case 0x10:
      TRACEMS1(cinfo, 1, JTRC_THUMB_JPEG, (int)totallen);
      break;
@@ -662,8 +660,7 @@ examine_app0(j_decompress_ptr cinfo, JOCTET *data, unsigned int datalen,
      TRACEMS1(cinfo, 1, JTRC_THUMB_RGB, (int)totallen);
      break;
    default:
-      TRACEMS2(cinfo, 1, JTRC_JFIF_EXTENSION,
-               GETJOCTET(data[5]), (int)totallen);
+      TRACEMS2(cinfo, 1, JTRC_JFIF_EXTENSION, data[5], (int)totallen);
      break;
    }
  } else {
@@ -684,16 +681,16 @@ examine_app14(j_decompress_ptr cinfo, JOCTET *data, unsigned int datalen,
  unsigned int version, flags0, flags1, transform;

  if (datalen >= APP14_DATA_LEN &&
-      GETJOCTET(data[0]) == 0x41 &&
-      GETJOCTET(data[1]) == 0x64 &&
-      GETJOCTET(data[2]) == 0x6F &&
-      GETJOCTET(data[3]) == 0x62 &&
-      GETJOCTET(data[4]) == 0x65) {
+      data[0] == 0x41 &&
+      data[1] == 0x64 &&
+      data[2] == 0x6F &&
+      data[3] == 0x62 &&
+      data[4] == 0x65) {
    /* Found Adobe APP14 marker */
-    version = (GETJOCTET(data[5]) << 8) + GETJOCTET(data[6]);
-    flags0 = (GETJOCTET(data[7]) << 8) + GETJOCTET(data[8]);
-    flags1 = (GETJOCTET(data[9]) << 8) + GETJOCTET(data[10]);
-    transform = GETJOCTET(data[11]);
+    version = (data[5] << 8) + data[6];
+    flags0 = (data[7] << 8) + data[8];
+    flags1 = (data[9] << 8) + data[10];
+    transform = data[11];
    TRACEMS4(cinfo, 1, JTRC_ADOBE, version, flags0, flags1, transform);
    cinfo->saw_Adobe_marker = TRUE;
    cinfo->Adobe_transform = (UINT8)transform;
@@ -5,7 +5,7 @@
 * Copyright (C) 1991-1997, Thomas G. Lane.
 * Modified 2002-2009 by Guido Vollbeding.
 * libjpeg-turbo Modifications:
- * Copyright (C) 2009-2011, 2016, D. R. Commander.
+ * Copyright (C) 2009-2011, 2016, 2019, D. R. Commander.
 * Copyright (C) 2013, Linaro Limited.
 * Copyright (C) 2015, Google, Inc.
 * For conditions of distribution and use, see the accompanying README.ijg
@@ -22,7 +22,6 @@
 #include "jpeglib.h"
 #include "jpegcomp.h"
 #include "jdmaster.h"
-#include "jsimd.h"


 /*
@@ -70,17 +69,6 @@ use_merged_upsample(j_decompress_ptr cinfo)
      cinfo->comp_info[1]._DCT_scaled_size != cinfo->_min_DCT_scaled_size ||
      cinfo->comp_info[2]._DCT_scaled_size != cinfo->_min_DCT_scaled_size)
    return FALSE;
-#ifdef WITH_SIMD
-  /* If YCbCr-to-RGB color conversion is SIMD-accelerated but merged upsampling
-     isn't, then disabling merged upsampling is likely to be faster when
-     decompressing YCbCr JPEG images. */
-  if (!jsimd_can_h2v2_merged_upsample() && !jsimd_can_h2v1_merged_upsample() &&
-      jsimd_can_ycc_rgb() && cinfo->jpeg_color_space == JCS_YCbCr &&
-      (cinfo->out_color_space == JCS_RGB ||
-       (cinfo->out_color_space >= JCS_EXT_RGB &&
-        cinfo->out_color_space <= JCS_EXT_ARGB)))
-    return FALSE;
-#endif
  /* ??? also need to test for upsample-time rescaling, when & if supported */
  return TRUE;                  /* by golly, it'll work... */
 #else
@@ -580,6 +568,7 @@ master_selection(j_decompress_ptr cinfo)
   */
  cinfo->master->first_iMCU_col = 0;
  cinfo->master->last_iMCU_col = cinfo->MCUs_per_row - 1;
+  cinfo->master->last_good_iMCU_row = 0;

 #ifdef D_MULTISCAN_FILES_SUPPORTED
  /* If jpeg_start_decompress will read the whole file, initialize
@@ -43,20 +43,20 @@ h2v1_merged_upsample_565_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
  /* Loop for each pair of output pixels */
  for (col = cinfo->output_width >> 1; col > 0; col--) {
    /* Do the chroma part of the calculation */
-    cb = GETJSAMPLE(*inptr1++);
-    cr = GETJSAMPLE(*inptr2++);
+    cb = *inptr1++;
+    cr = *inptr2++;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];

    /* Fetch 2 Y values and emit 2 pixels */
-    y  = GETJSAMPLE(*inptr0++);
+    y  = *inptr0++;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
    rgb = PACK_SHORT_565(r, g, b);

-    y  = GETJSAMPLE(*inptr0++);
+    y  = *inptr0++;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
@@ -68,12 +68,12 @@ h2v1_merged_upsample_565_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,

  /* If image width is odd, do the last output column separately */
  if (cinfo->output_width & 1) {
-    cb = GETJSAMPLE(*inptr1);
-    cr = GETJSAMPLE(*inptr2);
+    cb = *inptr1;
+    cr = *inptr2;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];
-    y  = GETJSAMPLE(*inptr0);
+    y  = *inptr0;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
@@ -115,21 +115,21 @@ h2v1_merged_upsample_565D_internal(j_decompress_ptr cinfo,
  /* Loop for each pair of output pixels */
  for (col = cinfo->output_width >> 1; col > 0; col--) {
    /* Do the chroma part of the calculation */
-    cb = GETJSAMPLE(*inptr1++);
-    cr = GETJSAMPLE(*inptr2++);
+    cb = *inptr1++;
+    cr = *inptr2++;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];

    /* Fetch 2 Y values and emit 2 pixels */
-    y  = GETJSAMPLE(*inptr0++);
+    y  = *inptr0++;
    r = range_limit[DITHER_565_R(y + cred, d0)];
    g = range_limit[DITHER_565_G(y + cgreen, d0)];
    b = range_limit[DITHER_565_B(y + cblue, d0)];
    d0 = DITHER_ROTATE(d0);
    rgb = PACK_SHORT_565(r, g, b);

-    y  = GETJSAMPLE(*inptr0++);
+    y  = *inptr0++;
    r = range_limit[DITHER_565_R(y + cred, d0)];
    g = range_limit[DITHER_565_G(y + cgreen, d0)];
    b = range_limit[DITHER_565_B(y + cblue, d0)];
@@ -142,12 +142,12 @@ h2v1_merged_upsample_565D_internal(j_decompress_ptr cinfo,

  /* If image width is odd, do the last output column separately */
  if (cinfo->output_width & 1) {
-    cb = GETJSAMPLE(*inptr1);
-    cr = GETJSAMPLE(*inptr2);
+    cb = *inptr1;
+    cr = *inptr2;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];
-    y  = GETJSAMPLE(*inptr0);
+    y  = *inptr0;
    r = range_limit[DITHER_565_R(y + cred, d0)];
    g = range_limit[DITHER_565_G(y + cgreen, d0)];
    b = range_limit[DITHER_565_B(y + cblue, d0)];
@@ -189,20 +189,20 @@ h2v2_merged_upsample_565_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
  /* Loop for each group of output pixels */
  for (col = cinfo->output_width >> 1; col > 0; col--) {
    /* Do the chroma part of the calculation */
-    cb = GETJSAMPLE(*inptr1++);
-    cr = GETJSAMPLE(*inptr2++);
+    cb = *inptr1++;
+    cr = *inptr2++;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];

    /* Fetch 4 Y values and emit 4 pixels */
-    y  = GETJSAMPLE(*inptr00++);
+    y  = *inptr00++;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
    rgb = PACK_SHORT_565(r, g, b);

-    y  = GETJSAMPLE(*inptr00++);
+    y  = *inptr00++;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
@@ -211,13 +211,13 @@ h2v2_merged_upsample_565_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    WRITE_TWO_PIXELS(outptr0, rgb);
    outptr0 += 4;

-    y  = GETJSAMPLE(*inptr01++);
+    y  = *inptr01++;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
    rgb = PACK_SHORT_565(r, g, b);

-    y  = GETJSAMPLE(*inptr01++);
+    y  = *inptr01++;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
@@ -229,20 +229,20 @@ h2v2_merged_upsample_565_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,

  /* If image width is odd, do the last output column separately */
  if (cinfo->output_width & 1) {
-    cb = GETJSAMPLE(*inptr1);
-    cr = GETJSAMPLE(*inptr2);
+    cb = *inptr1;
+    cr = *inptr2;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];

-    y  = GETJSAMPLE(*inptr00);
+    y  = *inptr00;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
    rgb = PACK_SHORT_565(r, g, b);
    *(INT16 *)outptr0 = (INT16)rgb;

-    y  = GETJSAMPLE(*inptr01);
+    y  = *inptr01;
    r = range_limit[y + cred];
    g = range_limit[y + cgreen];
    b = range_limit[y + cblue];
@@ -287,21 +287,21 @@ h2v2_merged_upsample_565D_internal(j_decompress_ptr cinfo,
  /* Loop for each group of output pixels */
  for (col = cinfo->output_width >> 1; col > 0; col--) {
    /* Do the chroma part of the calculation */
-    cb = GETJSAMPLE(*inptr1++);
-    cr = GETJSAMPLE(*inptr2++);
+    cb = *inptr1++;
+    cr = *inptr2++;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];

    /* Fetch 4 Y values and emit 4 pixels */
-    y  = GETJSAMPLE(*inptr00++);
+    y  = *inptr00++;
    r = range_limit[DITHER_565_R(y + cred, d0)];
    g = range_limit[DITHER_565_G(y + cgreen, d0)];
    b = range_limit[DITHER_565_B(y + cblue, d0)];
    d0 = DITHER_ROTATE(d0);
    rgb = PACK_SHORT_565(r, g, b);

-    y  = GETJSAMPLE(*inptr00++);
+    y  = *inptr00++;
    r = range_limit[DITHER_565_R(y + cred, d0)];
    g = range_limit[DITHER_565_G(y + cgreen, d0)];
    b = range_limit[DITHER_565_B(y + cblue, d0)];
@@ -311,14 +311,14 @@ h2v2_merged_upsample_565D_internal(j_decompress_ptr cinfo,
    WRITE_TWO_PIXELS(outptr0, rgb);
    outptr0 += 4;

-    y  = GETJSAMPLE(*inptr01++);
+    y  = *inptr01++;
    r = range_limit[DITHER_565_R(y + cred, d1)];
    g = range_limit[DITHER_565_G(y + cgreen, d1)];
    b = range_limit[DITHER_565_B(y + cblue, d1)];
    d1 = DITHER_ROTATE(d1);
    rgb = PACK_SHORT_565(r, g, b);

-    y  = GETJSAMPLE(*inptr01++);
+    y  = *inptr01++;
    r = range_limit[DITHER_565_R(y + cred, d1)];
    g = range_limit[DITHER_565_G(y + cgreen, d1)];
    b = range_limit[DITHER_565_B(y + cblue, d1)];
@@ -331,20 +331,20 @@ h2v2_merged_upsample_565D_internal(j_decompress_ptr cinfo,

  /* If image width is odd, do the last output column separately */
  if (cinfo->output_width & 1) {
-    cb = GETJSAMPLE(*inptr1);
-    cr = GETJSAMPLE(*inptr2);
+    cb = *inptr1;
+    cr = *inptr2;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];

-    y  = GETJSAMPLE(*inptr00);
+    y  = *inptr00;
    r = range_limit[DITHER_565_R(y + cred, d0)];
    g = range_limit[DITHER_565_G(y + cgreen, d0)];
    b = range_limit[DITHER_565_B(y + cblue, d0)];
    rgb = PACK_SHORT_565(r, g, b);
    *(INT16 *)outptr0 = (INT16)rgb;

-    y  = GETJSAMPLE(*inptr01);
+    y  = *inptr01;
    r = range_limit[DITHER_565_R(y + cred, d1)];
    g = range_limit[DITHER_565_G(y + cgreen, d1)];
    b = range_limit[DITHER_565_B(y + cblue, d1)];
@@ -46,13 +46,13 @@ h2v1_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
  /* Loop for each pair of output pixels */
  for (col = cinfo->output_width >> 1; col > 0; col--) {
    /* Do the chroma part of the calculation */
-    cb = GETJSAMPLE(*inptr1++);
-    cr = GETJSAMPLE(*inptr2++);
+    cb = *inptr1++;
+    cr = *inptr2++;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];
    /* Fetch 2 Y values and emit 2 pixels */
-    y  = GETJSAMPLE(*inptr0++);
+    y  = *inptr0++;
    outptr[RGB_RED] =   range_limit[y + cred];
    outptr[RGB_GREEN] = range_limit[y + cgreen];
    outptr[RGB_BLUE] =  range_limit[y + cblue];
@@ -60,7 +60,7 @@ h2v1_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    outptr[RGB_ALPHA] = 0xFF;
 #endif
    outptr += RGB_PIXELSIZE;
-    y  = GETJSAMPLE(*inptr0++);
+    y  = *inptr0++;
    outptr[RGB_RED] =   range_limit[y + cred];
    outptr[RGB_GREEN] = range_limit[y + cgreen];
    outptr[RGB_BLUE] =  range_limit[y + cblue];
@@ -71,12 +71,12 @@ h2v1_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
  }
  /* If image width is odd, do the last output column separately */
  if (cinfo->output_width & 1) {
-    cb = GETJSAMPLE(*inptr1);
-    cr = GETJSAMPLE(*inptr2);
+    cb = *inptr1;
+    cr = *inptr2;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];
-    y  = GETJSAMPLE(*inptr0);
+    y  = *inptr0;
    outptr[RGB_RED] =   range_limit[y + cred];
    outptr[RGB_GREEN] = range_limit[y + cgreen];
    outptr[RGB_BLUE] =  range_limit[y + cblue];
@@ -120,13 +120,13 @@ h2v2_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
  /* Loop for each group of output pixels */
  for (col = cinfo->output_width >> 1; col > 0; col--) {
    /* Do the chroma part of the calculation */
-    cb = GETJSAMPLE(*inptr1++);
-    cr = GETJSAMPLE(*inptr2++);
+    cb = *inptr1++;
+    cr = *inptr2++;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];
    /* Fetch 4 Y values and emit 4 pixels */
-    y  = GETJSAMPLE(*inptr00++);
+    y  = *inptr00++;
    outptr0[RGB_RED] =   range_limit[y + cred];
    outptr0[RGB_GREEN] = range_limit[y + cgreen];
    outptr0[RGB_BLUE] =  range_limit[y + cblue];
@@ -134,7 +134,7 @@ h2v2_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    outptr0[RGB_ALPHA] = 0xFF;
 #endif
    outptr0 += RGB_PIXELSIZE;
-    y  = GETJSAMPLE(*inptr00++);
+    y  = *inptr00++;
    outptr0[RGB_RED] =   range_limit[y + cred];
    outptr0[RGB_GREEN] = range_limit[y + cgreen];
    outptr0[RGB_BLUE] =  range_limit[y + cblue];
@@ -142,7 +142,7 @@ h2v2_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    outptr0[RGB_ALPHA] = 0xFF;
 #endif
    outptr0 += RGB_PIXELSIZE;
-    y  = GETJSAMPLE(*inptr01++);
+    y  = *inptr01++;
    outptr1[RGB_RED] =   range_limit[y + cred];
    outptr1[RGB_GREEN] = range_limit[y + cgreen];
    outptr1[RGB_BLUE] =  range_limit[y + cblue];
@@ -150,7 +150,7 @@ h2v2_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
    outptr1[RGB_ALPHA] = 0xFF;
 #endif
    outptr1 += RGB_PIXELSIZE;
-    y  = GETJSAMPLE(*inptr01++);
+    y  = *inptr01++;
    outptr1[RGB_RED] =   range_limit[y + cred];
    outptr1[RGB_GREEN] = range_limit[y + cgreen];
    outptr1[RGB_BLUE] =  range_limit[y + cblue];
@@ -161,19 +161,19 @@ h2v2_merged_upsample_internal(j_decompress_ptr cinfo, JSAMPIMAGE input_buf,
  }
  /* If image width is odd, do the last output column separately */
  if (cinfo->output_width & 1) {
-    cb = GETJSAMPLE(*inptr1);
-    cr = GETJSAMPLE(*inptr2);
+    cb = *inptr1;
+    cr = *inptr2;
    cred = Crrtab[cr];
    cgreen = (int)RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
    cblue = Cbbtab[cb];
-    y  = GETJSAMPLE(*inptr00);
+    y  = *inptr00;
    outptr0[RGB_RED] =   range_limit[y + cred];
    outptr0[RGB_GREEN] = range_limit[y + cgreen];
    outptr0[RGB_BLUE] =  range_limit[y + cblue];
 #ifdef RGB_ALPHA
    outptr0[RGB_ALPHA] = 0xFF;
 #endif
-    y  = GETJSAMPLE(*inptr01);
+    y  = *inptr01;
    outptr1[RGB_RED] =   range_limit[y + cred];
    outptr1[RGB_GREEN] = range_limit[y + cgreen];
    outptr1[RGB_BLUE] =  range_limit[y + cblue];
@@ -4,7 +4,7 @@
 * This file was part of the Independent JPEG Group's software:
 * Copyright (C) 1995-1997, Thomas G. Lane.
 * libjpeg-turbo Modifications:
- * Copyright (C) 2015-2016, 2018, D. R. Commander.
+ * Copyright (C) 2015-2016, 2018-2021, D. R. Commander.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -41,25 +41,6 @@ typedef struct {
  int last_dc_val[MAX_COMPS_IN_SCAN];   /* last DC coef for each component */
 } savable_state;

-/* This macro is to work around compilers with missing or broken
- * structure assignment.  You'll need to fix this code if you have
- * such a compiler and you change MAX_COMPS_IN_SCAN.
- */
-
-#ifndef NO_STRUCT_ASSIGN
-#define ASSIGN_STATE(dest, src)  ((dest) = (src))
-#else
-#if MAX_COMPS_IN_SCAN == 4
-#define ASSIGN_STATE(dest, src) \
-  ((dest).EOBRUN = (src).EOBRUN, \
-   (dest).last_dc_val[0] = (src).last_dc_val[0], \
-   (dest).last_dc_val[1] = (src).last_dc_val[1], \
-   (dest).last_dc_val[2] = (src).last_dc_val[2], \
-   (dest).last_dc_val[3] = (src).last_dc_val[3])
-#endif
-#endif
-
-
 typedef struct {
  struct jpeg_entropy_decoder pub; /* public fields */

@@ -102,7 +83,7 @@ start_pass_phuff_decoder(j_decompress_ptr cinfo)
  boolean is_DC_band, bad;
  int ci, coefi, tbl;
  d_derived_tbl **pdtbl;
-  int *coef_bit_ptr;
+  int *coef_bit_ptr, *prev_coef_bit_ptr;
  jpeg_component_info *compptr;

  is_DC_band = (cinfo->Ss == 0);
@@ -143,8 +124,15 @@ start_pass_phuff_decoder(j_decompress_ptr cinfo)
  for (ci = 0; ci < cinfo->comps_in_scan; ci++) {
    int cindex = cinfo->cur_comp_info[ci]->component_index;
    coef_bit_ptr = &cinfo->coef_bits[cindex][0];
+    prev_coef_bit_ptr = &cinfo->coef_bits[cindex + cinfo->num_components][0];
    if (!is_DC_band && coef_bit_ptr[0] < 0) /* AC without prior DC scan */
      WARNMS2(cinfo, JWRN_BOGUS_PROGRESSION, cindex, 0);
+    for (coefi = MIN(cinfo->Ss, 1); coefi <= MAX(cinfo->Se, 9); coefi++) {
+      if (cinfo->input_scan_number > 1)
+        prev_coef_bit_ptr[coefi] = coef_bit_ptr[coefi];
+      else
+        prev_coef_bit_ptr[coefi] = 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)
@@ -323,7 +311,7 @@ decode_mcu_DC_first(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)

    /* Load up working state */
    BITREAD_LOAD_STATE(cinfo, entropy->bitstate);
-    ASSIGN_STATE(state, entropy->saved);
+    state = entropy->saved;

    /* Outer loop handles each block in the MCU */

@@ -356,11 +344,12 @@ decode_mcu_DC_first(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)

    /* Completed MCU, so update state */
    BITREAD_SAVE_STATE(cinfo, entropy->bitstate);
-    ASSIGN_STATE(entropy->saved, state);
+    entropy->saved = state;
  }

  /* Account for restart interval (no-op if not using restarts) */
-  entropy->restarts_to_go--;
+  if (cinfo->restart_interval)
+    entropy->restarts_to_go--;

  return TRUE;
 }
@@ -444,7 +433,8 @@ decode_mcu_AC_first(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
  }

  /* Account for restart interval (no-op if not using restarts) */
-  entropy->restarts_to_go--;
+  if (cinfo->restart_interval)
+    entropy->restarts_to_go--;

  return TRUE;
 }
@@ -495,7 +485,8 @@ decode_mcu_DC_refine(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
  BITREAD_SAVE_STATE(cinfo, entropy->bitstate);

  /* Account for restart interval (no-op if not using restarts) */
-  entropy->restarts_to_go--;
+  if (cinfo->restart_interval)
+    entropy->restarts_to_go--;

  return TRUE;
 }
@@ -638,7 +629,8 @@ decode_mcu_AC_refine(j_decompress_ptr cinfo, JBLOCKROW *MCU_data)
  }

  /* Account for restart interval (no-op if not using restarts) */
-  entropy->restarts_to_go--;
+  if (cinfo->restart_interval)
+    entropy->restarts_to_go--;

  return TRUE;

@@ -676,7 +668,7 @@ jinit_phuff_decoder(j_decompress_ptr cinfo)
  /* Create progression status table */
  cinfo->coef_bits = (int (*)[DCTSIZE2])
    (*cinfo->mem->alloc_small) ((j_common_ptr)cinfo, JPOOL_IMAGE,
-                                cinfo->num_components * DCTSIZE2 *
+                                cinfo->num_components * 2 * DCTSIZE2 *
                                sizeof(int));
  coef_bit_ptr = &cinfo->coef_bits[0][0];
  for (ci = 0; ci < cinfo->num_components; ci++)
@@ -8,7 +8,7 @@
 * Copyright (C) 2010, 2015-2016, D. R. Commander.
 * Copyright (C) 2014, MIPS Technologies, Inc., California.
 * Copyright (C) 2015, Google, Inc.
- * Copyright (C) 2019, Arm Limited.
+ * Copyright (C) 2019-2020, Arm Limited.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -177,7 +177,7 @@ int_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
    outptr = output_data[outrow];
    outend = outptr + cinfo->output_width;
    while (outptr < outend) {
-      invalue = *inptr++;       /* don't need GETJSAMPLE() here */
+      invalue = *inptr++;
      for (h = h_expand; h > 0; h--) {
        *outptr++ = invalue;
      }
@@ -213,7 +213,7 @@ h2v1_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
    outptr = output_data[inrow];
    outend = outptr + cinfo->output_width;
    while (outptr < outend) {
-      invalue = *inptr++;       /* don't need GETJSAMPLE() here */
+      invalue = *inptr++;
      *outptr++ = invalue;
      *outptr++ = invalue;
    }
@@ -242,7 +242,7 @@ h2v2_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
    outptr = output_data[outrow];
    outend = outptr + cinfo->output_width;
    while (outptr < outend) {
-      invalue = *inptr++;       /* don't need GETJSAMPLE() here */
+      invalue = *inptr++;
      *outptr++ = invalue;
      *outptr++ = invalue;
    }
@@ -283,20 +283,20 @@ h2v1_fancy_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
    inptr = input_data[inrow];
    outptr = output_data[inrow];
    /* Special case for first column */
-    invalue = GETJSAMPLE(*inptr++);
+    invalue = *inptr++;
    *outptr++ = (JSAMPLE)invalue;
-    *outptr++ = (JSAMPLE)((invalue * 3 + GETJSAMPLE(*inptr) + 2) >> 2);
+    *outptr++ = (JSAMPLE)((invalue * 3 + inptr[0] + 2) >> 2);

    for (colctr = compptr->downsampled_width - 2; colctr > 0; colctr--) {
      /* General case: 3/4 * nearer pixel + 1/4 * further pixel */
-      invalue = GETJSAMPLE(*inptr++) * 3;
-      *outptr++ = (JSAMPLE)((invalue + GETJSAMPLE(inptr[-2]) + 1) >> 2);
-      *outptr++ = (JSAMPLE)((invalue + GETJSAMPLE(*inptr) + 2) >> 2);
+      invalue = (*inptr++) * 3;
+      *outptr++ = (JSAMPLE)((invalue + inptr[-2] + 1) >> 2);
+      *outptr++ = (JSAMPLE)((invalue + inptr[0] + 2) >> 2);
    }

    /* Special case for last column */
-    invalue = GETJSAMPLE(*inptr);
-    *outptr++ = (JSAMPLE)((invalue * 3 + GETJSAMPLE(inptr[-1]) + 1) >> 2);
+    invalue = *inptr;
+    *outptr++ = (JSAMPLE)((invalue * 3 + inptr[-1] + 1) >> 2);
    *outptr++ = (JSAMPLE)invalue;
  }
 }
@@ -338,7 +338,7 @@ h1v2_fancy_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
      outptr = output_data[outrow++];

      for (colctr = 0; colctr < compptr->downsampled_width; colctr++) {
-        thiscolsum = GETJSAMPLE(*inptr0++) * 3 + GETJSAMPLE(*inptr1++);
+        thiscolsum = (*inptr0++) * 3 + (*inptr1++);
        *outptr++ = (JSAMPLE)((thiscolsum + bias) >> 2);
      }
    }
@@ -381,8 +381,8 @@ h2v2_fancy_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
      outptr = output_data[outrow++];

      /* Special case for first column */
-      thiscolsum = GETJSAMPLE(*inptr0++) * 3 + GETJSAMPLE(*inptr1++);
-      nextcolsum = GETJSAMPLE(*inptr0++) * 3 + GETJSAMPLE(*inptr1++);
+      thiscolsum = (*inptr0++) * 3 + (*inptr1++);
+      nextcolsum = (*inptr0++) * 3 + (*inptr1++);
      *outptr++ = (JSAMPLE)((thiscolsum * 4 + 8) >> 4);
      *outptr++ = (JSAMPLE)((thiscolsum * 3 + nextcolsum + 7) >> 4);
      lastcolsum = thiscolsum;  thiscolsum = nextcolsum;
@@ -390,7 +390,7 @@ h2v2_fancy_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
      for (colctr = compptr->downsampled_width - 2; colctr > 0; colctr--) {
        /* General case: 3/4 * nearer pixel + 1/4 * further pixel in each */
        /* dimension, thus 9/16, 3/16, 3/16, 1/16 overall */
-        nextcolsum = GETJSAMPLE(*inptr0++) * 3 + GETJSAMPLE(*inptr1++);
+        nextcolsum = (*inptr0++) * 3 + (*inptr1++);
        *outptr++ = (JSAMPLE)((thiscolsum * 3 + lastcolsum + 8) >> 4);
        *outptr++ = (JSAMPLE)((thiscolsum * 3 + nextcolsum + 7) >> 4);
        lastcolsum = thiscolsum;  thiscolsum = nextcolsum;
@@ -477,7 +477,13 @@ jinit_upsampler(j_decompress_ptr cinfo)
    } else if (h_in_group == h_out_group &&
               v_in_group * 2 == v_out_group && do_fancy) {
      /* Non-fancy upsampling is handled by the generic method */
-      upsample->methods[ci] = h1v2_fancy_upsample;
+#if defined(__arm__) || defined(__aarch64__) || \
+    defined(_M_ARM) || defined(_M_ARM64)
+      if (jsimd_can_h1v2_fancy_upsample())
+        upsample->methods[ci] = jsimd_h1v2_fancy_upsample;
+      else
+#endif
+        upsample->methods[ci] = h1v2_fancy_upsample;
      upsample->pub.need_context_rows = TRUE;
    } else if (h_in_group * 2 == h_out_group &&
               v_in_group * 2 == v_out_group) {
@@ -207,6 +207,10 @@ JMESSAGE(JWRN_ARITH_BAD_CODE, "Corrupt JPEG data: bad arithmetic code")
 #endif
 #endif
 JMESSAGE(JWRN_BOGUS_ICC, "Corrupt JPEG data: bad ICC marker")
+#if JPEG_LIB_VERSION < 70
+JMESSAGE(JERR_BAD_DROP_SAMPLING,
+         "Component index %d: mismatching sampling ratio %d:%d, %d:%d, %c")
+#endif

 #ifdef JMAKE_ENUM_LIST

@@ -252,6 +256,15 @@ JMESSAGE(JWRN_BOGUS_ICC, "Corrupt JPEG data: bad ICC marker")
   (cinfo)->err->msg_parm.i[2] = (p3), \
   (cinfo)->err->msg_parm.i[3] = (p4), \
   (*(cinfo)->err->error_exit) ((j_common_ptr)(cinfo)))
+#define ERREXIT6(cinfo, code, p1, p2, p3, p4, p5, p6) \
+  ((cinfo)->err->msg_code = (code), \
+   (cinfo)->err->msg_parm.i[0] = (p1), \
+   (cinfo)->err->msg_parm.i[1] = (p2), \
+   (cinfo)->err->msg_parm.i[2] = (p3), \
+   (cinfo)->err->msg_parm.i[3] = (p4), \
+   (cinfo)->err->msg_parm.i[4] = (p5), \
+   (cinfo)->err->msg_parm.i[5] = (p6), \
+   (*(cinfo)->err->error_exit) ((j_common_ptr)(cinfo)))
 #define ERREXITS(cinfo, code, str) \
  ((cinfo)->err->msg_code = (code), \
   strncpy((cinfo)->err->msg_parm.s, (str), JMSG_STR_PARM_MAX), \
@@ -3,7 +3,7 @@
 *
 * This file was part of the Independent JPEG Group's software:
 * Copyright (C) 1991-1998, Thomas G. Lane.
- * Modification developed 2002-2009 by Guido Vollbeding.
+ * Modification developed 2002-2018 by Guido Vollbeding.
 * libjpeg-turbo Modifications:
 * Copyright (C) 2015, 2020, D. R. Commander.
 * For conditions of distribution and use, see the accompanying README.ijg
@@ -417,7 +417,7 @@ jpeg_idct_islow(j_decompress_ptr cinfo, jpeg_component_info *compptr,

 /*
 * Perform dequantization and inverse DCT on one block of coefficients,
- * producing a 7x7 output block.
+ * producing a reduced-size 7x7 output block.
 *
 * Optimized algorithm with 12 multiplications in the 1-D kernel.
 * cK represents sqrt(2) * cos(K*pi/14).
@@ -1258,7 +1258,7 @@ jpeg_idct_10x10(j_decompress_ptr cinfo, jpeg_component_info *compptr,

 /*
 * Perform dequantization and inverse DCT on one block of coefficients,
- * producing a 11x11 output block.
+ * producing an 11x11 output block.
 *
 * Optimized algorithm with 24 multiplications in the 1-D kernel.
 * cK represents sqrt(2) * cos(K*pi/22).
@@ -2398,7 +2398,7 @@ jpeg_idct_16x16(j_decompress_ptr cinfo, jpeg_component_info *compptr,
    tmp0 = DEQUANTIZE(inptr[DCTSIZE * 0], quantptr[DCTSIZE * 0]);
    tmp0 = LEFT_SHIFT(tmp0, CONST_BITS);
    /* Add fudge factor here for final descale. */
-    tmp0 += 1 << (CONST_BITS - PASS1_BITS - 1);
+    tmp0 += ONE << (CONST_BITS - PASS1_BITS - 1);

    z1 = DEQUANTIZE(inptr[DCTSIZE * 4], quantptr[DCTSIZE * 4]);
    tmp1 = MULTIPLY(z1, FIX(1.306562965));      /* c4[16] = c2[8] */
@@ -4,7 +4,7 @@
 * This file was part of the Independent JPEG Group's software:
 * Copyright (C) 1991-1997, Thomas G. Lane.
 * libjpeg-turbo Modifications:
- * Copyright (C) 2016, D. R. Commander.
+ * Copyright (C) 2016, 2021, D. R. Commander.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -1032,7 +1032,7 @@ free_pool(j_common_ptr cinfo, int pool_id)
    large_pool_ptr next_lhdr_ptr = lhdr_ptr->next;
    space_freed = lhdr_ptr->bytes_used +
                  lhdr_ptr->bytes_left +
-                  sizeof(large_pool_hdr);
+                  sizeof(large_pool_hdr) + ALIGN_SIZE - 1;
    jpeg_free_large(cinfo, (void *)lhdr_ptr, space_freed);
    mem->total_space_allocated -= space_freed;
    lhdr_ptr = next_lhdr_ptr;
@@ -1045,7 +1045,7 @@ free_pool(j_common_ptr cinfo, int pool_id)
  while (shdr_ptr != NULL) {
    small_pool_ptr next_shdr_ptr = shdr_ptr->next;
    space_freed = shdr_ptr->bytes_used + shdr_ptr->bytes_left +
-                  sizeof(small_pool_hdr);
+                  sizeof(small_pool_hdr) + ALIGN_SIZE - 1;
    jpeg_free_small(cinfo, (void *)shdr_ptr, space_freed);
    mem->total_space_allocated -= space_freed;
    shdr_ptr = next_shdr_ptr;
@@ -43,25 +43,11 @@

 #if BITS_IN_JSAMPLE == 8
 /* JSAMPLE should be the smallest type that will hold the values 0..255.
- * You can use a signed char by having GETJSAMPLE mask it with 0xFF.
 */

-#ifdef HAVE_UNSIGNED_CHAR
-
 typedef unsigned char JSAMPLE;
 #define GETJSAMPLE(value)  ((int)(value))

-#else /* not HAVE_UNSIGNED_CHAR */
-
-typedef char JSAMPLE;
-#ifdef __CHAR_UNSIGNED__
-#define GETJSAMPLE(value)  ((int)(value))
-#else
-#define GETJSAMPLE(value)  ((int)(value) & 0xFF)
-#endif /* __CHAR_UNSIGNED__ */
-
-#endif /* HAVE_UNSIGNED_CHAR */
-
 #define MAXJSAMPLE      255
 #define CENTERJSAMPLE   128

@@ -97,22 +83,9 @@ typedef short JCOEF;
 * managers, this is also the data type passed to fread/fwrite.
 */

-#ifdef HAVE_UNSIGNED_CHAR
-
 typedef unsigned char JOCTET;
 #define GETJOCTET(value)  (value)

-#else /* not HAVE_UNSIGNED_CHAR */
-
-typedef char JOCTET;
-#ifdef __CHAR_UNSIGNED__
-#define GETJOCTET(value)  (value)
-#else
-#define GETJOCTET(value)  ((value) & 0xFF)
-#endif /* __CHAR_UNSIGNED__ */
-
-#endif /* HAVE_UNSIGNED_CHAR */
-

 /* These typedefs are used for various table entries and so forth.
 * They must be at least as wide as specified; but making them too big
@@ -123,15 +96,7 @@ typedef char JOCTET;

 /* UINT8 must hold at least the values 0..255. */

-#ifdef HAVE_UNSIGNED_CHAR
 typedef unsigned char UINT8;
-#else /* not HAVE_UNSIGNED_CHAR */
-#ifdef __CHAR_UNSIGNED__
-typedef char UINT8;
-#else /* not __CHAR_UNSIGNED__ */
-typedef short UINT8;
-#endif /* __CHAR_UNSIGNED__ */
-#endif /* HAVE_UNSIGNED_CHAR */

 /* UINT16 must hold at least the values 0..65535. */

@@ -5,8 +5,9 @@
 * Copyright (C) 1991-1997, Thomas G. Lane.
 * Modified 1997-2009 by Guido Vollbeding.
 * libjpeg-turbo Modifications:
- * Copyright (C) 2015-2016, D. R. Commander.
+ * Copyright (C) 2015-2016, 2019, 2021, D. R. Commander.
 * Copyright (C) 2015, Google, Inc.
+ * Copyright (C) 2021, Alex Richardson.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -47,6 +48,18 @@ typedef enum {            /* Operating modes for buffer controllers */
 /* JLONG must hold at least signed 32-bit values. */
 typedef long JLONG;

+/* JUINTPTR must hold pointer values. */
+#ifdef __UINTPTR_TYPE__
+/*
+ * __UINTPTR_TYPE__ is GNU-specific and available in GCC 4.6+ and Clang 3.0+.
+ * Fortunately, that is sufficient to support the few architectures for which
+ * sizeof(void *) != sizeof(size_t).  The only other options would require C99
+ * or Clang-specific builtins.
+ */
+typedef __UINTPTR_TYPE__ JUINTPTR;
+#else
+typedef size_t JUINTPTR;
+#endif

 /*
 * Left shift macro that handles a negative operand without causing any
@@ -158,6 +171,9 @@ struct jpeg_decomp_master {
  JDIMENSION first_MCU_col[MAX_COMPONENTS];
  JDIMENSION last_MCU_col[MAX_COMPONENTS];
  boolean jinit_upsampler_no_alloc;
+
+  /* Last iMCU row that was successfully decoded */
+  JDIMENSION last_good_iMCU_row;
 };

 /* Input control module */
@@ -479,7 +479,7 @@ color_quantize(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
    for (col = width; col > 0; col--) {
      pixcode = 0;
      for (ci = 0; ci < nc; ci++) {
-        pixcode += GETJSAMPLE(colorindex[ci][GETJSAMPLE(*ptrin++)]);
+        pixcode += colorindex[ci][*ptrin++];
      }
      *ptrout++ = (JSAMPLE)pixcode;
    }
@@ -506,9 +506,9 @@ color_quantize3(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
    ptrin = input_buf[row];
    ptrout = output_buf[row];
    for (col = width; col > 0; col--) {
-      pixcode  = GETJSAMPLE(colorindex0[GETJSAMPLE(*ptrin++)]);
-      pixcode += GETJSAMPLE(colorindex1[GETJSAMPLE(*ptrin++)]);
-      pixcode += GETJSAMPLE(colorindex2[GETJSAMPLE(*ptrin++)]);
+      pixcode  = colorindex0[*ptrin++];
+      pixcode += colorindex1[*ptrin++];
+      pixcode += colorindex2[*ptrin++];
      *ptrout++ = (JSAMPLE)pixcode;
    }
  }
@@ -552,7 +552,7 @@ quantize_ord_dither(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
         * required amount of padding.
         */
        *output_ptr +=
-          colorindex_ci[GETJSAMPLE(*input_ptr) + dither[col_index]];
+          colorindex_ci[*input_ptr + dither[col_index]];
        input_ptr += nc;
        output_ptr++;
        col_index = (col_index + 1) & ODITHER_MASK;
@@ -595,12 +595,9 @@ quantize3_ord_dither(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
    col_index = 0;

    for (col = width; col > 0; col--) {
-      pixcode  =
-        GETJSAMPLE(colorindex0[GETJSAMPLE(*input_ptr++) + dither0[col_index]]);
-      pixcode +=
-        GETJSAMPLE(colorindex1[GETJSAMPLE(*input_ptr++) + dither1[col_index]]);
-      pixcode +=
-        GETJSAMPLE(colorindex2[GETJSAMPLE(*input_ptr++) + dither2[col_index]]);
+      pixcode  = colorindex0[(*input_ptr++) + dither0[col_index]];
+      pixcode += colorindex1[(*input_ptr++) + dither1[col_index]];
+      pixcode += colorindex2[(*input_ptr++) + dither2[col_index]];
      *output_ptr++ = (JSAMPLE)pixcode;
      col_index = (col_index + 1) & ODITHER_MASK;
    }
@@ -677,15 +674,15 @@ quantize_fs_dither(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
         * The maximum error is +- MAXJSAMPLE; this sets the required size
         * of the range_limit array.
         */
-        cur += GETJSAMPLE(*input_ptr);
-        cur = GETJSAMPLE(range_limit[cur]);
+        cur += *input_ptr;
+        cur = range_limit[cur];
        /* Select output value, accumulate into output code for this pixel */
-        pixcode = GETJSAMPLE(colorindex_ci[cur]);
+        pixcode = colorindex_ci[cur];
        *output_ptr += (JSAMPLE)pixcode;
        /* Compute actual representation error at this pixel */
        /* Note: we can do this even though we don't have the final */
        /* pixel code, because the colormap is orthogonal. */
-        cur -= GETJSAMPLE(colormap_ci[pixcode]);
+        cur -= colormap_ci[pixcode];
        /* Compute error fractions to be propagated to adjacent pixels.
         * Add these into the running sums, and simultaneously shift the
         * next-line error sums left by 1 column.
@@ -215,9 +215,9 @@ prescan_quantize(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
    ptr = input_buf[row];
    for (col = width; col > 0; col--) {
      /* get pixel value and index into the histogram */
-      histp = &histogram[GETJSAMPLE(ptr[0]) >> C0_SHIFT]
-                        [GETJSAMPLE(ptr[1]) >> C1_SHIFT]
-                        [GETJSAMPLE(ptr[2]) >> C2_SHIFT];
+      histp = &histogram[ptr[0] >> C0_SHIFT]
+                        [ptr[1] >> C1_SHIFT]
+                        [ptr[2] >> C2_SHIFT];
      /* increment, check for overflow and undo increment if so. */
      if (++(*histp) <= 0)
        (*histp)--;
@@ -665,7 +665,7 @@ find_nearby_colors(j_decompress_ptr cinfo, int minc0, int minc1, int minc2,

  for (i = 0; i < numcolors; i++) {
    /* We compute the squared-c0-distance term, then add in the other two. */
-    x = GETJSAMPLE(cinfo->colormap[0][i]);
+    x = cinfo->colormap[0][i];
    if (x < minc0) {
      tdist = (x - minc0) * C0_SCALE;
      min_dist = tdist * tdist;
@@ -688,7 +688,7 @@ find_nearby_colors(j_decompress_ptr cinfo, int minc0, int minc1, int minc2,
      }
    }

-    x = GETJSAMPLE(cinfo->colormap[1][i]);
+    x = cinfo->colormap[1][i];
    if (x < minc1) {
      tdist = (x - minc1) * C1_SCALE;
      min_dist += tdist * tdist;
@@ -710,7 +710,7 @@ find_nearby_colors(j_decompress_ptr cinfo, int minc0, int minc1, int minc2,
      }
    }

-    x = GETJSAMPLE(cinfo->colormap[2][i]);
+    x = cinfo->colormap[2][i];
    if (x < minc2) {
      tdist = (x - minc2) * C2_SCALE;
      min_dist += tdist * tdist;
@@ -788,13 +788,13 @@ find_best_colors(j_decompress_ptr cinfo, int minc0, int minc1, int minc2,
 #define STEP_C2  ((1 << C2_SHIFT) * C2_SCALE)

  for (i = 0; i < numcolors; i++) {
-    icolor = GETJSAMPLE(colorlist[i]);
+    icolor = colorlist[i];
    /* Compute (square of) distance from minc0/c1/c2 to this color */
-    inc0 = (minc0 - GETJSAMPLE(cinfo->colormap[0][icolor])) * C0_SCALE;
+    inc0 = (minc0 - cinfo->colormap[0][icolor]) * C0_SCALE;
    dist0 = inc0 * inc0;
-    inc1 = (minc1 - GETJSAMPLE(cinfo->colormap[1][icolor])) * C1_SCALE;
+    inc1 = (minc1 - cinfo->colormap[1][icolor]) * C1_SCALE;
    dist0 += inc1 * inc1;
-    inc2 = (minc2 - GETJSAMPLE(cinfo->colormap[2][icolor])) * C2_SCALE;
+    inc2 = (minc2 - cinfo->colormap[2][icolor]) * C2_SCALE;
    dist0 += inc2 * inc2;
    /* Form the initial difference increments */
    inc0 = inc0 * (2 * STEP_C0) + STEP_C0 * STEP_C0;
@@ -879,7 +879,7 @@ fill_inverse_cmap(j_decompress_ptr cinfo, int c0, int c1, int c2)
    for (ic1 = 0; ic1 < BOX_C1_ELEMS; ic1++) {
      cachep = &histogram[c0 + ic0][c1 + ic1][c2];
      for (ic2 = 0; ic2 < BOX_C2_ELEMS; ic2++) {
-        *cachep++ = (histcell)(GETJSAMPLE(*cptr++) + 1);
+        *cachep++ = (histcell)((*cptr++) + 1);
      }
    }
  }
@@ -909,9 +909,9 @@ pass2_no_dither(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
    outptr = output_buf[row];
    for (col = width; col > 0; col--) {
      /* get pixel value and index into the cache */
-      c0 = GETJSAMPLE(*inptr++) >> C0_SHIFT;
-      c1 = GETJSAMPLE(*inptr++) >> C1_SHIFT;
-      c2 = GETJSAMPLE(*inptr++) >> C2_SHIFT;
+      c0 = (*inptr++) >> C0_SHIFT;
+      c1 = (*inptr++) >> C1_SHIFT;
+      c2 = (*inptr++) >> C2_SHIFT;
      cachep = &histogram[c0][c1][c2];
      /* If we have not seen this color before, find nearest colormap entry */
      /* and update the cache */
@@ -996,12 +996,12 @@ pass2_fs_dither(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
       * The maximum error is +- MAXJSAMPLE (or less with error limiting);
       * this sets the required size of the range_limit array.
       */
-      cur0 += GETJSAMPLE(inptr[0]);
-      cur1 += GETJSAMPLE(inptr[1]);
-      cur2 += GETJSAMPLE(inptr[2]);
-      cur0 = GETJSAMPLE(range_limit[cur0]);
-      cur1 = GETJSAMPLE(range_limit[cur1]);
-      cur2 = GETJSAMPLE(range_limit[cur2]);
+      cur0 += inptr[0];
+      cur1 += inptr[1];
+      cur2 += inptr[2];
+      cur0 = range_limit[cur0];
+      cur1 = range_limit[cur1];
+      cur2 = range_limit[cur2];
      /* Index into the cache with adjusted pixel value */
      cachep =
        &histogram[cur0 >> C0_SHIFT][cur1 >> C1_SHIFT][cur2 >> C2_SHIFT];
@@ -1015,9 +1015,9 @@ pass2_fs_dither(j_decompress_ptr cinfo, JSAMPARRAY input_buf,
        register int pixcode = *cachep - 1;
        *outptr = (JSAMPLE)pixcode;
        /* Compute representation error for this pixel */
-        cur0 -= GETJSAMPLE(colormap0[pixcode]);
-        cur1 -= GETJSAMPLE(colormap1[pixcode]);
-        cur2 -= GETJSAMPLE(colormap2[pixcode]);
+        cur0 -= colormap0[pixcode];
+        cur1 -= colormap1[pixcode];
+        cur2 -= colormap2[pixcode];
      }
      /* Compute error fractions to be propagated to adjacent pixels.
       * Add these into the running sums, and simultaneously shift the
@@ -4,6 +4,7 @@
 * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
 * Copyright (C) 2011, 2014, D. R. Commander.
 * Copyright (C) 2015-2016, 2018, Matthieu Darbois.
+ * Copyright (C) 2020, Arm Limited.
 *
 * Based on the x86 SIMD extension for IJG JPEG library,
 * Copyright (C) 1999-2006, MIYASAKA Masaru.
@@ -75,6 +76,7 @@ EXTERN(void) jsimd_int_upsample(j_decompress_ptr cinfo,

 EXTERN(int) jsimd_can_h2v2_fancy_upsample(void);
 EXTERN(int) jsimd_can_h2v1_fancy_upsample(void);
+EXTERN(int) jsimd_can_h1v2_fancy_upsample(void);

 EXTERN(void) jsimd_h2v2_fancy_upsample(j_decompress_ptr cinfo,
                                       jpeg_component_info *compptr,
@@ -84,6 +86,10 @@ EXTERN(void) jsimd_h2v1_fancy_upsample(j_decompress_ptr cinfo,
                                       jpeg_component_info *compptr,
                                       JSAMPARRAY input_data,
                                       JSAMPARRAY *output_data_ptr);
+EXTERN(void) jsimd_h1v2_fancy_upsample(j_decompress_ptr cinfo,
+                                       jpeg_component_info *compptr,
+                                       JSAMPARRAY input_data,
+                                       JSAMPARRAY *output_data_ptr);

 EXTERN(int) jsimd_can_h2v2_merged_upsample(void);
 EXTERN(int) jsimd_can_h2v1_merged_upsample(void);
@@ -4,6 +4,7 @@
 * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
 * Copyright (C) 2009-2011, 2014, D. R. Commander.
 * Copyright (C) 2015-2016, 2018, Matthieu Darbois.
+ * Copyright (C) 2020, Arm Limited.
 *
 * Based on the x86 SIMD extension for IJG JPEG library,
 * Copyright (C) 1999-2006, MIYASAKA Masaru.
@@ -169,6 +170,12 @@ jsimd_can_h2v1_fancy_upsample(void)
  return 0;
 }

+GLOBAL(int)
+jsimd_can_h1v2_fancy_upsample(void)
+{
+  return 0;
+}
+
 GLOBAL(void)
 jsimd_h2v2_fancy_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
                          JSAMPARRAY input_data, JSAMPARRAY *output_data_ptr)
@@ -181,6 +188,12 @@ jsimd_h2v1_fancy_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
 {
 }

+GLOBAL(void)
+jsimd_h1v2_fancy_upsample(j_decompress_ptr cinfo, jpeg_component_info *compptr,
+                          JSAMPARRAY input_data, JSAMPARRAY *output_data_ptr)
+{
+}
+
 GLOBAL(int)
 jsimd_can_h2v2_merged_upsample(void)
 {
@@ -2,9 +2,9 @@
 * jversion.h
 *
 * This file was part of the Independent JPEG Group's software:
- * Copyright (C) 1991-2012, Thomas G. Lane, Guido Vollbeding.
+ * Copyright (C) 1991-2020, Thomas G. Lane, Guido Vollbeding.
 * libjpeg-turbo Modifications:
- * Copyright (C) 2010, 2012-2020, D. R. Commander.
+ * Copyright (C) 2010, 2012-2021, D. R. Commander.
 * For conditions of distribution and use, see the accompanying README.ijg
 * file.
 *
@@ -37,9 +37,9 @@
 */

 #define JCOPYRIGHT \
-  "Copyright (C) 2009-2020 D. R. Commander\n" \
+  "Copyright (C) 2009-2021 D. R. Commander\n" \
  "Copyright (C) 2015, 2020 Google, Inc.\n" \
-  "Copyright (C) 2019 Arm Limited\n" \
+  "Copyright (C) 2019-2020 Arm Limited\n" \
  "Copyright (C) 2015-2016, 2018 Matthieu Darbois\n" \
  "Copyright (C) 2011-2016 Siarhei Siamashka\n" \
  "Copyright (C) 2015 Intel Corporation\n" \
@@ -48,7 +48,7 @@
  "Copyright (C) 2009, 2012 Pierre Ossman for Cendio AB\n" \
  "Copyright (C) 2009-2011 Nokia Corporation and/or its subsidiary(-ies)\n" \
  "Copyright (C) 1999-2006 MIYASAKA Masaru\n" \
-  "Copyright (C) 1991-2017 Thomas G. Lane, Guido Vollbeding"
+  "Copyright (C) 1991-2020 Thomas G. Lane, Guido Vollbeding"

 #define JCOPYRIGHT_SHORT \
-  "Copyright (C) 1991-2020 The libjpeg-turbo Project and many others"
+  "Copyright (C) 1991-2021 The libjpeg-turbo Project and many others"
@@ -77,7 +77,9 @@ endif(MSVC)
 add_library(${PNG_LIBRARY} STATIC ${OPENCV_3RDPARTY_EXCLUDE_FROM_ALL} ${lib_srcs} ${lib_hdrs})
 target_link_libraries(${PNG_LIBRARY} ${ZLIB_LIBRARIES})

-ocv_warnings_disable(CMAKE_C_FLAGS -Wundef -Wcast-align -Wimplicit-fallthrough -Wunused-parameter -Wsign-compare)
+ocv_warnings_disable(CMAKE_C_FLAGS -Wundef -Wcast-align -Wimplicit-fallthrough -Wunused-parameter -Wsign-compare
+    -Wmaybe-uninitialized
+)

 set_target_properties(${PNG_LIBRARY}
  PROPERTIES OUTPUT_NAME ${PNG_LIBRARY}
@@ -452,8 +452,9 @@ ocv_warnings_disable(CMAKE_C_FLAGS -Wno-unused-but-set-variable -Wmissing-protot
                                   -Wcast-align -Wshadow -Wno-maybe-uninitialized -Wno-pointer-to-int-cast -Wno-int-to-pointer-cast
                                   -Wmisleading-indentation
                                   -Wimplicit-fallthrough
+                                   -Wunused-parameter  # clang
+                                   -Warray-parameter
 )
-ocv_warnings_disable(CMAKE_C_FLAGS -Wunused-parameter) # clang
 ocv_warnings_disable(CMAKE_CXX_FLAGS -Wmissing-declarations -Wunused-parameter -Wmissing-prototypes
    -Wundef  # tiffiop.h: #if __clang_major__ >= 4
 )
@@ -119,6 +119,7 @@ ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4334) # vs2005 Win64
 ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4244) # vs2008
 ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4267) # vs2008 Win64
 ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4456) # vs2015
+ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4819) # vs2019 Win64

 if(MSVC AND CV_ICC)
  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /Qrestrict")
@@ -77,7 +77,7 @@ E.g. external ref-counting is implemented for 1.0 version and native OpenVX one

 Also there are some **C++ 11** features are used (e.g. rvalue ref-s) when their availability is detected at ***compile time***.

-C++ exceptions are used for errors indication instead of return codes. There are two types of exceptions are defined:  `RuntimeError` is thrown when OpenVX C call returned unsuccessful result and `WrapperError` is thrown when a problem is occured in the wrappers code. Both exception calsses are derived from `std::exception` (actually from its inheritants).
+C++ exceptions are used for errors indication instead of return codes. There are two types of exceptions are defined:  `RuntimeError` is thrown when OpenVX C call returned unsuccessful result and `WrapperError` is thrown when a problem is occurred in the wrappers code. Both exception calsses are derived from `std::exception` (actually from its inheritants).

 The so called **OpenVX objects** (e.g. `vx_image`) are represented as C++ classes in wrappers.
 All these classes use automatic ref-counting that allows development of exception-safe code.
@@ -923,6 +923,11 @@ int ovx_hal_cvtGraytoBGR(const uchar * a, size_t astep, uchar * b, size_t bstep,
 }

 int ovx_hal_cvtTwoPlaneYUVtoBGR(const uchar * a, size_t astep, uchar * b, size_t bstep, int w, int h, int bcn, bool swapBlue, int uIdx)
+{
+    return ovx_hal_cvtTwoPlaneYUVtoBGREx(a, astep, a + h * astep, astep, b, bstep, w, h, bcn, swapBlue, uIdx);
+}
+
+int ovx_hal_cvtTwoPlaneYUVtoBGREx(const uchar * a, size_t astep, const uchar * b, size_t bstep, uchar * c, size_t cstep, int w, int h, int bcn, bool swapBlue, int uIdx)
 {
    if (skipSmallImages<VX_KERNEL_COLOR_CONVERT>(w, h))
        return CV_HAL_ERROR_NOT_IMPLEMENTED;
@@ -933,8 +938,7 @@ int ovx_hal_cvtTwoPlaneYUVtoBGR(const uchar * a, size_t astep, uchar * b, size_t

    if (w & 1 || h & 1) // It's not described in spec but sample implementation unable to convert odd sized images
        return CV_HAL_ERROR_NOT_IMPLEMENTED;
-    refineStep(w, h, uIdx ? VX_DF_IMAGE_NV21 : VX_DF_IMAGE_NV12, astep);
-    refineStep(w, h, bcn == 3 ? VX_DF_IMAGE_RGB : VX_DF_IMAGE_RGBX, bstep);
+
    try
    {
        ivx::Context ctx = getOpenVXHALContext();
@@ -943,8 +947,8 @@ int ovx_hal_cvtTwoPlaneYUVtoBGR(const uchar * a, size_t astep, uchar * b, size_t
        std::vector<void *> ptrs;
            addr.push_back(ivx::Image::createAddressing(w, h, 1, (vx_int32)astep));
            ptrs.push_back((void*)a);
-            addr.push_back(ivx::Image::createAddressing(w / 2, h / 2, 2, (vx_int32)astep));
-            ptrs.push_back((void*)(a + h * astep));
+            addr.push_back(ivx::Image::createAddressing(w / 2, h / 2, 2, (vx_int32)bstep));
+            ptrs.push_back((void*)b);

        vxImage
            ia = ivx::Image::createFromHandle(ctx, uIdx ? VX_DF_IMAGE_NV21 : VX_DF_IMAGE_NV12, addr, ptrs);
@@ -952,7 +956,7 @@ int ovx_hal_cvtTwoPlaneYUVtoBGR(const uchar * a, size_t astep, uchar * b, size_t
            return CV_HAL_ERROR_NOT_IMPLEMENTED; // OpenCV store NV12/NV21 as RANGE_RESTRICTED while OpenVX expect RANGE_FULL
        vxImage
            ib = ivx::Image::createFromHandle(ctx, bcn == 3 ? VX_DF_IMAGE_RGB : VX_DF_IMAGE_RGBX,
-                ivx::Image::createAddressing(w, h, bcn, (vx_int32)bstep), b);
+                ivx::Image::createAddressing(w, h, bcn, (vx_int32)cstep), c);
        ivx::IVX_CHECK_STATUS(vxuColorConvert(ctx, ia, ib));
    }
    catch (ivx::RuntimeError & e)
@@ -49,6 +49,7 @@ int ovx_hal_morph(cvhalFilter2D *filter_context, uchar *a, size_t astep, uchar *
 int ovx_hal_cvtBGRtoBGR(const uchar * a, size_t astep, uchar * b, size_t bstep, int w, int h, int depth, int acn, int bcn, bool swapBlue);
 int ovx_hal_cvtGraytoBGR(const uchar * a, size_t astep, uchar * b, size_t bstep, int w, int h, int depth, int bcn);
 int ovx_hal_cvtTwoPlaneYUVtoBGR(const uchar * a, size_t astep, uchar * b, size_t bstep, int w, int h, int bcn, bool swapBlue, int uIdx);
+int ovx_hal_cvtTwoPlaneYUVtoBGREx(const uchar * a, size_t astep, const uchar * b, size_t bstep, uchar * c, size_t cstep, int w, int h, int bcn, bool swapBlue, int uIdx);
 int ovx_hal_cvtThreePlaneYUVtoBGR(const uchar * a, size_t astep, uchar * b, size_t bstep, int w, int h, int bcn, bool swapBlue, int uIdx);
 int ovx_hal_cvtBGRtoThreePlaneYUV(const uchar * a, size_t astep, uchar * b, size_t bstep, int w, int h, int acn, bool swapBlue, int uIdx);
 int ovx_hal_cvtOnePlaneYUVtoBGR(const uchar * a, size_t astep, uchar * b, size_t bstep, int w, int h, int bcn, bool swapBlue, int uIdx, int ycn);
@@ -130,6 +131,8 @@ int ovx_hal_integral(int depth, int sdepth, int, const uchar * a, size_t astep,
 #define cv_hal_cvtGraytoBGR ovx_hal_cvtGraytoBGR
 #undef cv_hal_cvtTwoPlaneYUVtoBGR
 #define cv_hal_cvtTwoPlaneYUVtoBGR ovx_hal_cvtTwoPlaneYUVtoBGR
+#undef cv_hal_cvtTwoPlaneYUVtoBGREx
+#define cv_hal_cvtTwoPlaneYUVtoBGREx ovx_hal_cvtTwoPlaneYUVtoBGREx
 #undef cv_hal_cvtThreePlaneYUVtoBGR
 #define cv_hal_cvtThreePlaneYUVtoBGR ovx_hal_cvtThreePlaneYUVtoBGR
 #undef cv_hal_cvtBGRtoThreePlaneYUV
@@ -31,13 +31,13 @@ libpng                Portable Network Graphics library.
 libtiff               Tag Image File Format (TIFF) Software
                      Copyright (c) 1988-1997 Sam Leffler
                      Copyright (c) 1991-1997 Silicon Graphics, Inc.
-                      See libtiff home page http://www.remotesensing.org/libtiff/
+                      See libtiff home page http://www.libtiff.org/
                      for details and links to the source code

                      WITH_TIFF CMake option must be ON to add libtiff & zlib support to imgcodecs.
 ------------------------------------------------------------------------------------
 zlib                  General purpose LZ77 compression library
-                      Copyright (C) 1995-2012 Jean-loup Gailly and Mark Adler.
+                      Copyright (C) 1995-2022 Jean-loup Gailly and Mark Adler.
                      See zlib home page http://www.zlib.net
                      for details and links to the source code
 ------------------------------------------------------------------------------------
@@ -51,7 +51,9 @@ jasper                JasPer is a collection of software
                      Copyright (c) 1999-2000 The University of British Columbia
                      Copyright (c) 2001-2003 Michael David Adams

-                      The JasPer license can be found in libjasper.
+                      See JasPer official GitHub repository
+                      https://github.com/jasper-software/jasper.git
+                      for details and links to source code
 ------------------------------------------------------------------------------------
 openexr               OpenEXR is a high dynamic-range (HDR) image file format developed
                      by Industrial Light & Magic for use in computer imaging applications.
@@ -83,6 +83,7 @@ ocv_warnings_disable(CMAKE_C_FLAGS -Wshorten-64-to-32 -Wattributes -Wstrict-prot
    -Wundef  # _LFS64_LARGEFILE is not defined
    /wd4267  # MSVS 2015 (x64) + zlib 1.2.11
    -Wimplicit-fallthrough
+    /wd4244  # MSVS + zlib 1.2.12: warning C4244: '=': conversion from 'ush' to 'uchf', possible loss of data
 )

 set_target_properties(${ZLIB_LIBRARY} PROPERTIES
@@ -101,4 +102,4 @@ if(NOT BUILD_SHARED_LIBS)
  ocv_install_target(${ZLIB_LIBRARY} EXPORT OpenCVModules ARCHIVE DESTINATION ${OPENCV_3P_LIB_INSTALL_PATH} COMPONENT dev)
 endif()

-ocv_install_3rdparty_licenses(zlib README)
+ocv_install_3rdparty_licenses(zlib LICENSE)
@@ -1,6 +1,81 @@

                ChangeLog file for zlib

+Changes in 1.2.13 (13 Oct 2022)
+- Fix configure issue that discarded provided CC definition
+- Correct incorrect inputs provided to the CRC functions
+- Repair prototypes and exporting of new CRC functions
+- Fix inflateBack to detect invalid input with distances too far
+- Have infback() deliver all of the available output up to any error
+- Fix a bug when getting a gzip header extra field with inflate()
+- Fix bug in block type selection when Z_FIXED used
+- Tighten deflateBound bounds
+- Remove deleted assembler code references
+- Various portability and appearance improvements
+
+Changes in 1.2.12 (27 Mar 2022)
+- Cygwin does not have _wopen(), so do not create gzopen_w() there
+- Permit a deflateParams() parameter change as soon as possible
+- Limit hash table inserts after switch from stored deflate
+- Fix bug when window full in deflate_stored()
+- Fix CLEAR_HASH macro to be usable as a single statement
+- Avoid a conversion error in gzseek when off_t type too small
+- Have Makefile return non-zero error code on test failure
+- Avoid some conversion warnings in gzread.c and gzwrite.c
+- Update use of errno for newer Windows CE versions
+- Small speedup to inflate [psumbera]
+- Return an error if the gzputs string length can't fit in an int
+- Add address checking in clang to -w option of configure
+- Don't compute check value for raw inflate if asked to validate
+- Handle case where inflateSync used when header never processed
+- Avoid the use of ptrdiff_t
+- Avoid an undefined behavior of memcpy() in gzappend()
+- Avoid undefined behaviors of memcpy() in gz*printf()
+- Avoid an undefined behavior of memcpy() in _tr_stored_block()
+- Make the names in functions declarations identical to definitions
+- Remove old assembler code in which bugs have manifested
+- Fix deflateEnd() to not report an error at start of raw deflate
+- Add legal disclaimer to README
+- Emphasize the need to continue decompressing gzip members
+- Correct the initialization requirements for deflateInit2()
+- Fix a bug that can crash deflate on some input when using Z_FIXED
+- Assure that the number of bits for deflatePrime() is valid
+- Use a structure to make globals in enough.c evident
+- Use a macro for the printf format of big_t in enough.c
+- Clean up code style in enough.c, update version
+- Use inline function instead of macro for index in enough.c
+- Clarify that prefix codes are counted in enough.c
+- Show all the codes for the maximum tables size in enough.c
+- Add gznorm.c example, which normalizes gzip files
+- Fix the zran.c example to work on a multiple-member gzip file
+- Add tables for crc32_combine(), to speed it up by a factor of 200
+- Add crc32_combine_gen() and crc32_combine_op() for fast combines
+- Speed up software CRC-32 computation by a factor of 1.5 to 3
+- Use atomic test and set, if available, for dynamic CRC tables
+- Don't bother computing check value after successful inflateSync()
+- Correct comment in crc32.c
+- Add use of the ARMv8 crc32 instructions when requested
+- Use ARM crc32 instructions if the ARM architecture has them
+- Explicitly note that the 32-bit check values are 32 bits
+- Avoid adding empty gzip member after gzflush with Z_FINISH
+- Fix memory leak on error in gzlog.c
+- Fix error in comment on the polynomial representation of a byte
+- Clarify gz* function interfaces, referring to parameter names
+- Change macro name in inflate.c to avoid collision in VxWorks
+- Correct typo in blast.c
+- Improve portability of contrib/minizip
+- Fix indentation in minizip's zip.c
+- Replace black/white with allow/block. (theresa-m)
+- minizip warning fix if MAXU32 already defined. (gvollant)
+- Fix unztell64() in minizip to work past 4GB. (Daniël Hörchner)
+- Clean up minizip to reduce warnings for testing
+- Add fallthrough comments for gcc
+- Eliminate use of ULL constants
+- Separate out address sanitizing from warnings in configure
+- Remove destructive aspects of make distclean
+- Check for cc masquerading as gcc or clang in configure
+- Fix crc32.c to compile local functions only if used
+
 Changes in 1.2.11 (15 Jan 2017)
 - Fix deflate stored bug when pulling last block from window
 - Permit immediate deflateParams changes before any deflate input
@@ -96,7 +171,7 @@ Changes in 1.2.7.1 (24 Mar 2013)
 - Fix types in contrib/minizip to match result of get_crc_table()
 - Simplify contrib/vstudio/vc10 with 'd' suffix
 - Add TOP support to win32/Makefile.msc
- Suport i686 and amd64 assembler builds in CMakeLists.txt
+- Support i686 and amd64 assembler builds in CMakeLists.txt
 - Fix typos in the use of _LARGEFILE64_SOURCE in zconf.h
 - Add vc11 and vc12 build files to contrib/vstudio
 - Add gzvprintf() as an undocumented function in zlib
@@ -296,14 +371,14 @@ Changes in 1.2.5.1 (10 Sep 2011)
 - Use u4 type for crc_table to avoid conversion warnings
 - Apply casts in zlib.h to avoid conversion warnings
 - Add OF to prototypes for adler32_combine_ and crc32_combine_ [Miller]
- Improve inflateSync() documentation to note indeterminancy
+- Improve inflateSync() documentation to note indeterminacy
 - Add deflatePending() function to return the amount of pending output
 - Correct the spelling of "specification" in FAQ [Randers-Pehrson]
 - Add a check in configure for stdarg.h, use for gzprintf()
 - Check that pointers fit in ints when gzprint() compiled old style
 - Add dummy name before $(SHAREDLIBV) in Makefile [Bar-Lev, Bowler]
 - Delete line in configure that adds -L. libz.a to LDFLAGS [Weigelt]
- Add debug records in assmebler code [Londer]
+- Add debug records in assembler code [Londer]
 - Update RFC references to use http://tools.ietf.org/html/... [Li]
 - Add --archs option, use of libtool to configure for Mac OS X [Borstel]

@@ -511,7 +586,7 @@ Changes in 1.2.3.5 (8 Jan 2010)
 - Don't use _vsnprintf on later versions of MSVC [Lowman]
 - Add CMake build script and input file [Lowman]
 - Update contrib/minizip to 1.1 [Svensson, Vollant]
- Moved nintendods directory from contrib to .
+- Moved nintendods directory from contrib to root
 - Replace gzio.c with a new set of routines with the same functionality
 - Add gzbuffer(), gzoffset(), gzclose_r(), gzclose_w() as part of above
 - Update contrib/minizip to 1.1b
@@ -685,7 +760,7 @@ Changes in 1.2.2.4 (11 July 2005)
 - Be more strict on incomplete code sets in inflate_table() and increase
  ENOUGH and MAXD -- this repairs a possible security vulnerability for
  invalid inflate input.  Thanks to Tavis Ormandy and Markus Oberhumer for
-  discovering the vulnerability and providing test cases.
+  discovering the vulnerability and providing test cases
 - Add ia64 support to configure for HP-UX [Smith]
 - Add error return to gzread() for format or i/o error [Levin]
 - Use malloc.h for OS/2 [Necasek]
@@ -721,7 +796,7 @@ Changes in 1.2.2.2 (30 December 2004)
 - Add Z_FIXED strategy option to deflateInit2() to force fixed trees
 - Add updated make_vms.com [Coghlan], update README
 - Create a new "examples" directory, move gzappend.c there, add zpipe.c,
-  fitblk.c, gzlog.[ch], gzjoin.c, and zlib_how.html.
+  fitblk.c, gzlog.[ch], gzjoin.c, and zlib_how.html
 - Add FAQ entry and comments in deflate.c on uninitialized memory access
 - Add Solaris 9 make options in configure [Gilbert]
 - Allow strerror() usage in gzio.c for STDC
@@ -792,7 +867,7 @@ Changes in 1.2.1.1 (9 January 2004)
 - Fix a big fat bug in inftrees.c that prevented decoding valid
  dynamic blocks with only literals and no distance codes --
  Thanks to "Hot Emu" for the bug report and sample file
- Add a note to puff.c on no distance codes case.
+- Add a note to puff.c on no distance codes case

 Changes in 1.2.1 (17 November 2003)
 - Remove a tab in contrib/gzappend/gzappend.c
@@ -970,7 +1045,7 @@ Changes in 1.2.0.1 (17 March 2003)
    - Include additional header file on VMS for off_t typedef
 - Try to use _vsnprintf where it supplants vsprintf [Vollant]
 - Add some casts in inffast.c
- Enchance comments in zlib.h on what happens if gzprintf() tries to
+- Enhance comments in zlib.h on what happens if gzprintf() tries to
  write more than 4095 bytes before compression
 - Remove unused state from inflateBackEnd()
 - Remove exit(0) from minigzip.c, example.c
@@ -1036,14 +1111,14 @@ Changes in 1.2.0 (9 March 2003)
 - Add contrib/puff/ simple inflate for deflate format description

 Changes in 1.1.4 (11 March 2002)
- ZFREE was repeated on same allocation on some error conditions.
+- ZFREE was repeated on same allocation on some error conditions
  This creates a security problem described in
  http://www.zlib.org/advisory-2002-03-11.txt
 - Returned incorrect error (Z_MEM_ERROR) on some invalid data
 - Avoid accesses before window for invalid distances with inflate window
-  less than 32K.
+  less than 32K
 - force windowBits > 8 to avoid a bug in the encoder for a window size
-  of 256 bytes. (A complete fix will be available in 1.1.5).
+  of 256 bytes. (A complete fix will be available in 1.1.5)

 Changes in 1.1.3 (9 July 1998)
 - fix "an inflate input buffer bug that shows up on rare but persistent
@@ -1117,7 +1192,7 @@ Changes in 1.1.1 (27 Feb 98)
 - remove block truncation heuristic which had very marginal effect for zlib
  (smaller lit_bufsize than in gzip 1.2.4) and degraded a little the
  compression ratio on some files. This also allows inlining _tr_tally for
-  matches in deflate_slow.
+  matches in deflate_slow
 - added msdos/Makefile.w32 for WIN32 Microsoft Visual C++ (Bob Frazier)

 Changes in 1.1.0 (24 Feb 98)
@@ -1148,7 +1223,7 @@ Changes in 1.0.9 (17 Feb 1998)
 - Avoid gcc 2.8.0 comparison bug a little differently than zlib 1.0.8
 - in inftrees.c, avoid cc -O bug on HP (Farshid Elahi)
 - in zconf.h move the ZLIB_DLL stuff earlier to avoid problems with
-  the declaration of FAR (Gilles VOllant)
+  the declaration of FAR (Gilles Vollant)
 - install libz.so* with mode 755 (executable) instead of 644 (Marc Lehmann)
 - read_buf buf parameter of type Bytef* instead of charf*
 - zmemcpy parameters are of type Bytef*, not charf* (Joseph Strout)
@@ -1162,7 +1237,7 @@ Changes in 1.0.8 (27 Jan 1998)
 - include sys/types.h to get off_t on some systems (Marc Lehmann & QingLong)
 - use constant arrays for the static trees in trees.c instead of computing
  them at run time (thanks to Ken Raeburn for this suggestion). To create
-  trees.h, compile with GEN_TREES_H and run "make test".
+  trees.h, compile with GEN_TREES_H and run "make test"
 - check return code of example in "make test" and display result
 - pass minigzip command line options to file_compress
 - simplifying code of inflateSync to avoid gcc 2.8 bug
@@ -1201,12 +1276,12 @@ Changes in 1.0.6 (19 Jan 1998)
 - add functions gzprintf, gzputc, gzgetc, gztell, gzeof, gzseek, gzrewind and
  gzsetparams (thanks to Roland Giersig and Kevin Ruland for some of this code)
 - Fix a deflate bug occurring only with compression level 0 (thanks to
-  Andy Buckler for finding this one).
- In minigzip, pass transparently also the first byte for .Z files.
+  Andy Buckler for finding this one)
+- In minigzip, pass transparently also the first byte for .Z files
 - return Z_BUF_ERROR instead of Z_OK if output buffer full in uncompress()
 - check Z_FINISH in inflate (thanks to Marc Schluper)
 - Implement deflateCopy (thanks to Adam Costello)
- make static libraries by default in configure, add --shared option.
+- make static libraries by default in configure, add --shared option
 - move MSDOS or Windows specific files to directory msdos
 - suppress the notion of partial flush to simplify the interface
  (but the symbol Z_PARTIAL_FLUSH is kept for compatibility with 1.0.4)
@@ -1218,7 +1293,7 @@ Changes in 1.0.6 (19 Jan 1998)
 - added Makefile.nt (thanks to Stephen Williams)
 - added the unsupported "contrib" directory:
   contrib/asm386/ by Gilles Vollant <info@winimage.com>
-        386 asm code replacing longest_match().
+        386 asm code replacing longest_match()
   contrib/iostream/ by Kevin Ruland <kevin@rodin.wustl.edu>
        A C++ I/O streams interface to the zlib gz* functions
   contrib/iostream2/  by Tyge Løvset <Tyge.Lovset@cmr.no>
@@ -1226,7 +1301,7 @@ Changes in 1.0.6 (19 Jan 1998)
   contrib/untgz/  by "Pedro A. Aranda Guti\irrez" <paag@tid.es>
        A very simple tar.gz file extractor using zlib
   contrib/visual-basic.txt by Carlos Rios <c_rios@sonda.cl>
-        How to use compress(), uncompress() and the gz* functions from VB.
+        How to use compress(), uncompress() and the gz* functions from VB
 - pass params -f (filtered data), -h (huffman only), -1 to -9 (compression
  level) in minigzip (thanks to Tom Lane)

@@ -1235,8 +1310,8 @@ Changes in 1.0.6 (19 Jan 1998)
 - add undocumented function inflateSyncPoint() (hack for Paul Mackerras)
 - add undocumented function zError to convert error code to string
  (for Tim Smithers)
- Allow compilation of gzio with -DNO_DEFLATE to avoid the compression code.
- Use default memcpy for Symantec MSDOS compiler.
+- Allow compilation of gzio with -DNO_DEFLATE to avoid the compression code
+- Use default memcpy for Symantec MSDOS compiler
 - Add EXPORT keyword for check_func (needed for Windows DLL)
 - add current directory to LD_LIBRARY_PATH for "make test"
 - create also a link for libz.so.1
@@ -1249,7 +1324,7 @@ Changes in 1.0.6 (19 Jan 1998)
 - allow compilation with ANSI keywords only enabled for TurboC in large model
 - avoid "versionString"[0] (Borland bug)
 - add NEED_DUMMY_RETURN for Borland
- use variable z_verbose for tracing in debug mode (L. Peter Deutsch).
+- use variable z_verbose for tracing in debug mode (L. Peter Deutsch)
 - allow compilation with CC
 - defined STDC for OS/2 (David Charlap)
 - limit external names to 8 chars for MVS (Thomas Lund)
@@ -1259,7 +1334,7 @@ Changes in 1.0.6 (19 Jan 1998)
 - use _fdopen instead of fdopen for MSC >= 6.0 (Thomas Fanslau)
 - added makelcc.bat for lcc-win32 (Tom St Denis)
 - in Makefile.dj2, use copy and del instead of install and rm (Frank Donahoe)
- Avoid expanded $Id$. Use "rcs -kb" or "cvs admin -kb" to avoid Id expansion.
+- Avoid expanded $Id$. Use "rcs -kb" or "cvs admin -kb" to avoid Id expansion
 - check for unistd.h in configure (for off_t)
 - remove useless check parameter in inflate_blocks_free
 - avoid useless assignment of s->check to itself in inflate_blocks_new
@@ -1280,7 +1355,7 @@ Changes in 1.0.5 (3 Jan 98)
 Changes in 1.0.4 (24 Jul 96)
 - In very rare conditions, deflate(s, Z_FINISH) could fail to produce an EOF
  bit, so the decompressor could decompress all the correct data but went
-  on to attempt decompressing extra garbage data. This affected minigzip too.
+  on to attempt decompressing extra garbage data. This affected minigzip too
 - zlibVersion and gzerror return const char* (needed for DLL)
 - port to RISCOS (no fdopen, no multiple dots, no unlink, no fileno)
 - use z_error only for DEBUG (avoid problem with DLLs)
@@ -1310,7 +1385,7 @@ Changes in 1.0.1 (20 May 96) [1.0 skipped to avoid confusion]
 - fix array overlay in deflate.c which sometimes caused bad compressed data
 - fix inflate bug with empty stored block
 - fix MSDOS medium model which was broken in 0.99
- fix deflateParams() which could generate bad compressed data.
+- fix deflateParams() which could generate bad compressed data
 - Bytef is define'd instead of typedef'ed (work around Borland bug)
 - added an INDEX file
 - new makefiles for DJGPP (Makefile.dj2), 32-bit Borland (Makefile.b32),
@@ -1331,7 +1406,7 @@ Changes in 0.99 (27 Jan 96)
 - allow preset dictionary shared between compressor and decompressor
 - allow compression level 0 (no compression)
 - add deflateParams in zlib.h: allow dynamic change of compression level
-  and compression strategy.
+  and compression strategy
 - test large buffers and deflateParams in example.c
 - add optional "configure" to build zlib as a shared library
 - suppress Makefile.qnx, use configure instead
@@ -1373,30 +1448,30 @@ Changes in 0.99 (27 Jan 96)
 - use STDC instead of __GO32__ to avoid redeclaring exit, calloc, etc...
 - use Z_BINARY instead of BINARY
 - document that gzclose after gzdopen will close the file
- allow "a" as mode in gzopen.
+- allow "a" as mode in gzopen
 - fix error checking in gzread
 - allow skipping .gz extra-field on pipes
 - added reference to Perl interface in README
 - put the crc table in FAR data (I dislike more and more the medium model :)
 - added get_crc_table
- added a dimension to all arrays (Borland C can't count).
+- added a dimension to all arrays (Borland C can't count)
 - workaround Borland C bug in declaration of inflate_codes_new & inflate_fast
 - guard against multiple inclusion of *.h (for precompiled header on Mac)
- Watcom C pretends to be Microsoft C small model even in 32 bit mode.
+- Watcom C pretends to be Microsoft C small model even in 32 bit mode
 - don't use unsized arrays to avoid silly warnings by Visual C++:
     warning C4746: 'inflate_mask' : unsized array treated as  '__far'
-     (what's wrong with far data in far model?).
+     (what's wrong with far data in far model?)
 - define enum out of inflate_blocks_state to allow compilation with C++

 Changes in 0.95 (16 Aug 95)
 - fix MSDOS small and medium model (now easier to adapt to any compiler)
 - inlined send_bits
 - fix the final (:-) bug for deflate with flush (output was correct but
-  not completely flushed in rare occasions).
+  not completely flushed in rare occasions)
 - default window size is same for compression and decompression
-  (it's now sufficient to set MAX_WBITS in zconf.h).
+  (it's now sufficient to set MAX_WBITS in zconf.h)
 - voidp -> voidpf and voidnp -> voidp (for consistency with other
-  typedefs and because voidnp was not near in large model).
+  typedefs and because voidnp was not near in large model)

 Changes in 0.94 (13 Aug 95)
 - support MSDOS medium model
@@ -1405,12 +1480,12 @@ Changes in 0.94 (13 Aug 95)
 - added support for VMS
 - allow a compression level in gzopen()
 - gzflush now calls fflush
- For deflate with flush, flush even if no more input is provided.
+- For deflate with flush, flush even if no more input is provided
 - rename libgz.a as libz.a
 - avoid complex expression in infcodes.c triggering Turbo C bug
 - work around a problem with gcc on Alpha (in INSERT_STRING)
 - don't use inline functions (problem with some gcc versions)
- allow renaming of Byte, uInt, etc... with #define.
+- allow renaming of Byte, uInt, etc... with #define
 - avoid warning about (unused) pointer before start of array in deflate.c
 - avoid various warnings in gzio.c, example.c, infblock.c, adler32.c, zutil.c
 - avoid reserved word 'new' in trees.c
@@ -1429,7 +1504,7 @@ Changes in 0.92 (3 May 95)
 - no memcpy on Pyramid
 - suppressed inftest.c
 - optimized fill_window, put longest_match inline for gcc
- optimized inflate on stored blocks.
+- optimized inflate on stored blocks
 - untabify all sources to simplify patches

 Changes in 0.91 (2 May 95)
@@ -1447,7 +1522,7 @@ Changes in 0.9 (1 May 95)
 - let again gzread copy uncompressed data unchanged (was working in 0.71)
 - deflate(Z_FULL_FLUSH), inflateReset and inflateSync are now fully implemented
 - added a test of inflateSync in example.c
- moved MAX_WBITS to zconf.h because users might want to change that.
+- moved MAX_WBITS to zconf.h because users might want to change that
 - document explicitly that zalloc(64K) on MSDOS must return a normalized
  pointer (zero offset)
 - added Makefiles for Microsoft C, Turbo C, Borland C++
@@ -1456,7 +1531,7 @@ Changes in 0.9 (1 May 95)
 Changes in 0.8 (29 April 95)
 - added fast inflate (inffast.c)
 - deflate(Z_FINISH) now returns Z_STREAM_END when done. Warning: this
-  is incompatible with previous versions of zlib which returned Z_OK.
+  is incompatible with previous versions of zlib which returned Z_OK
 - work around a TurboC compiler bug (bad code for b << 0, see infutil.h)
  (actually that was not a compiler bug, see 0.81 above)
 - gzread no longer reads one extra byte in certain cases
@@ -1466,50 +1541,50 @@ Changes in 0.8 (29 April 95)

 Changes in 0.71 (14 April 95)
 - Fixed more MSDOS compilation problems :( There is still a bug with
-  TurboC large model.
+  TurboC large model

 Changes in 0.7 (14 April 95)
- Added full inflate support.
+- Added full inflate support
 - Simplified the crc32() interface. The pre- and post-conditioning
  (one's complement) is now done inside crc32(). WARNING: this is
-  incompatible with previous versions; see zlib.h for the new usage.
+  incompatible with previous versions; see zlib.h for the new usage

 Changes in 0.61 (12 April 95)
- workaround for a bug in TurboC. example and minigzip now work on MSDOS.
+- workaround for a bug in TurboC. example and minigzip now work on MSDOS

 Changes in 0.6 (11 April 95)
 - added minigzip.c
 - added gzdopen to reopen a file descriptor as gzFile
- added transparent reading of non-gziped files in gzread.
+- added transparent reading of non-gziped files in gzread
 - fixed bug in gzread (don't read crc as data)
- fixed bug in destroy (gzio.c) (don't return Z_STREAM_END for gzclose).
+- fixed bug in destroy (gzio.c) (don't return Z_STREAM_END for gzclose)
 - don't allocate big arrays in the stack (for MSDOS)
 - fix some MSDOS compilation problems

 Changes in 0.5:
 - do real compression in deflate.c. Z_PARTIAL_FLUSH is supported but
-  not yet Z_FULL_FLUSH.
+  not yet Z_FULL_FLUSH
 - support decompression but only in a single step (forced Z_FINISH)
- added opaque object for zalloc and zfree.
+- added opaque object for zalloc and zfree
 - added deflateReset and inflateReset
- added a variable zlib_version for consistency checking.
- renamed the 'filter' parameter of deflateInit2 as 'strategy'.
-  Added Z_FILTERED and Z_HUFFMAN_ONLY constants.
+- added a variable zlib_version for consistency checking
+- renamed the 'filter' parameter of deflateInit2 as 'strategy'
+  Added Z_FILTERED and Z_HUFFMAN_ONLY constants

 Changes in 0.4:
- avoid "zip" everywhere, use zlib instead of ziplib.
+- avoid "zip" everywhere, use zlib instead of ziplib
 - suppress Z_BLOCK_FLUSH, interpret Z_PARTIAL_FLUSH as block flush
-  if compression method == 8.
+  if compression method == 8
 - added adler32 and crc32
 - renamed deflateOptions as deflateInit2, call one or the other but not both
- added the method parameter for deflateInit2.
+- added the method parameter for deflateInit2
 - added inflateInit2
- simplied considerably deflateInit and inflateInit by not supporting
+- simplified considerably deflateInit and inflateInit by not supporting
  user-provided history buffer. This is supported only in deflateInit2
-  and inflateInit2.
+  and inflateInit2

 Changes in 0.3:
 - prefix all macro names with Z_
- use Z_FINISH instead of deflateEnd to finish compression.
+- use Z_FINISH instead of deflateEnd to finish compression
 - added Z_HUFFMAN_ONLY
 - added gzerror()
@@ -0,0 +1,22 @@
+Copyright notice:
+
+ (C) 1995-2022 Jean-loup Gailly and Mark Adler
+
+  This software is provided 'as-is', without any express or implied
+  warranty.  In no event will the authors be held liable for any damages
+  arising from the use of this software.
+
+  Permission is granted to anyone to use this software for any purpose,
+  including commercial applications, and to alter it and redistribute it
+  freely, subject to the following restrictions:
+
+  1. The origin of this software must not be misrepresented; you must not
+     claim that you wrote the original software. If you use this software
+     in a product, an acknowledgment in the product documentation would be
+     appreciated but is not required.
+  2. Altered source versions must be plainly marked as such, and must not be
+     misrepresented as being the original software.
+  3. This notice may not be removed or altered from any source distribution.
+
+  Jean-loup Gailly        Mark Adler
+  jloup@gzip.org          madler@alumni.caltech.edu
@@ -1,6 +1,6 @@
 ZLIB DATA COMPRESSION LIBRARY

-zlib 1.2.11 is a general purpose data compression library.  All the code is
+zlib 1.2.13 is a general purpose data compression library.  All the code is
 thread safe.  The data format used by the zlib library is described by RFCs
 (Request for Comments) 1950 to 1952 in the files
 http://tools.ietf.org/html/rfc1950 (zlib format), rfc1951 (deflate format) and
@@ -31,7 +31,7 @@ Mark Nelson <markn@ieee.org> wrote an article about zlib for the Jan.  1997
 issue of Dr.  Dobb's Journal; a copy of the article is available at
 http://marknelson.us/1997/01/01/zlib-engine/ .

-The changes made in version 1.2.11 are documented in the file ChangeLog.
+The changes made in version 1.2.13 are documented in the file ChangeLog.

 Unsupported third party contributions are provided in directory contrib/ .

@@ -84,7 +84,7 @@ Acknowledgments:

 Copyright notice:

- (C) 1995-2017 Jean-loup Gailly and Mark Adler
+ (C) 1995-2022 Jean-loup Gailly and Mark Adler

  This software is provided 'as-is', without any express or implied
  warranty.  In no event will the authors be held liable for any damages
@@ -108,7 +108,10 @@ Copyright notice:
 If you use the zlib library in a product, we would appreciate *not* receiving
 lengthy legal documents to sign.  The sources are provided for free but without
 warranty of any kind.  The library has been entirely written by Jean-loup
-Gailly and Mark Adler; it does not include third-party code.
+Gailly and Mark Adler; it does not include third-party code.  We make all
+contributions to and distributions of this project solely in our personal
+capacity, and are not conveying any rights to any intellectual property of
+any third parties.

 If you redistribute modified sources, we would appreciate that you include in
 the file ChangeLog history information documenting your changes.  Please read
@@ -19,7 +19,7 @@
   memory, Z_BUF_ERROR if there was not enough room in the output buffer,
   Z_STREAM_ERROR if the level parameter is invalid.
 */
-int ZEXPORT compress2 (dest, destLen, source, sourceLen, level)
+int ZEXPORT compress2(dest, destLen, source, sourceLen, level)
    Bytef *dest;
    uLongf *destLen;
    const Bytef *source;
@@ -65,7 +65,7 @@ int ZEXPORT compress2 (dest, destLen, source, sourceLen, level)

 /* ===========================================================================
 */
-int ZEXPORT compress (dest, destLen, source, sourceLen)
+int ZEXPORT compress(dest, destLen, source, sourceLen)
    Bytef *dest;
    uLongf *destLen;
    const Bytef *source;
@@ -78,7 +78,7 @@ int ZEXPORT compress (dest, destLen, source, sourceLen)
     If the default memLevel or windowBits for deflateInit() is changed, then
   this function needs to be updated.
 */
-uLong ZEXPORT compressBound (sourceLen)
+uLong ZEXPORT compressBound(sourceLen)
    uLong sourceLen;
 {
    return sourceLen + (sourceLen >> 12) + (sourceLen >> 14) +
@@ -1,5 +1,5 @@
 /* deflate.c -- compress data using the deflation algorithm
- * Copyright (C) 1995-2017 Jean-loup Gailly and Mark Adler
+ * Copyright (C) 1995-2022 Jean-loup Gailly and Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -52,7 +52,7 @@
 #include "deflate.h"

 const char deflate_copyright[] =
-   " deflate 1.2.11 Copyright 1995-2017 Jean-loup Gailly and Mark Adler ";
+   " deflate 1.2.13 Copyright 1995-2022 Jean-loup Gailly and Mark Adler ";
 /*
  If you use the zlib library in a product, an acknowledgment is welcome
  in the documentation of your product. If for some reason you cannot
@@ -87,13 +87,7 @@ local void lm_init        OF((deflate_state *s));
 local void putShortMSB    OF((deflate_state *s, uInt b));
 local void flush_pending  OF((z_streamp strm));
 local unsigned read_buf   OF((z_streamp strm, Bytef *buf, unsigned size));
-#ifdef ASMV
-#  pragma message("Assembler code may have bugs -- use at your own risk")
-      void match_init OF((void)); /* asm code initialization */
-      uInt longest_match  OF((deflate_state *s, IPos cur_match));
-#else
 local uInt longest_match  OF((deflate_state *s, IPos cur_match));
-#endif

 #ifdef ZLIB_DEBUG
 local  void check_match OF((deflate_state *s, IPos start, IPos match,
@@ -160,7 +154,7 @@ local const config configuration_table[10] = {
 *    characters, so that a running hash key can be computed from the previous
 *    key instead of complete recalculation each time.
 */
-#define UPDATE_HASH(s,h,c) (h = (((h)<<s->hash_shift) ^ (c)) & s->hash_mask)
+#define UPDATE_HASH(s,h,c) (h = (((h) << s->hash_shift) ^ (c)) & s->hash_mask)


 /* ===========================================================================
@@ -190,8 +184,11 @@ local const config configuration_table[10] = {
 * prev[] will be initialized on the fly.
 */
 #define CLEAR_HASH(s) \
-    s->head[s->hash_size-1] = NIL; \
-    zmemzero((Bytef *)s->head, (unsigned)(s->hash_size-1)*sizeof(*s->head));
+    do { \
+        s->head[s->hash_size - 1] = NIL; \
+        zmemzero((Bytef *)s->head, \
+                 (unsigned)(s->hash_size - 1)*sizeof(*s->head)); \
+    } while (0)

 /* ===========================================================================
 * Slide the hash table when sliding the window down (could be avoided with 32
@@ -252,11 +249,6 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
    int wrap = 1;
    static const char my_version[] = ZLIB_VERSION;

-    ushf *overlay;
-    /* We overlay pending_buf and d_buf+l_buf. This works since the average
-     * output size for (length,distance) codes is <= 24 bits.
-     */
-
    if (version == Z_NULL || version[0] != my_version[0] ||
        stream_size != sizeof(z_stream)) {
        return Z_VERSION_ERROR;
@@ -287,6 +279,8 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,

    if (windowBits < 0) { /* suppress zlib wrapper */
        wrap = 0;
+        if (windowBits < -15)
+            return Z_STREAM_ERROR;
        windowBits = -windowBits;
    }
 #ifdef GZIP
@@ -316,7 +310,7 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
    s->hash_bits = (uInt)memLevel + 7;
    s->hash_size = 1 << s->hash_bits;
    s->hash_mask = s->hash_size - 1;
-    s->hash_shift =  ((s->hash_bits+MIN_MATCH-1)/MIN_MATCH);
+    s->hash_shift =  ((s->hash_bits + MIN_MATCH-1) / MIN_MATCH);

    s->window = (Bytef *) ZALLOC(strm, s->w_size, 2*sizeof(Byte));
    s->prev   = (Posf *)  ZALLOC(strm, s->w_size, sizeof(Pos));
@@ -326,9 +320,47 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,

    s->lit_bufsize = 1 << (memLevel + 6); /* 16K elements by default */

-    overlay = (ushf *) ZALLOC(strm, s->lit_bufsize, sizeof(ush)+2);
-    s->pending_buf = (uchf *) overlay;
-    s->pending_buf_size = (ulg)s->lit_bufsize * (sizeof(ush)+2L);
+    /* We overlay pending_buf and sym_buf. This works since the average size
+     * for length/distance pairs over any compressed block is assured to be 31
+     * bits or less.
+     *
+     * Analysis: The longest fixed codes are a length code of 8 bits plus 5
+     * extra bits, for lengths 131 to 257. The longest fixed distance codes are
+     * 5 bits plus 13 extra bits, for distances 16385 to 32768. The longest
+     * possible fixed-codes length/distance pair is then 31 bits total.
+     *
+     * sym_buf starts one-fourth of the way into pending_buf. So there are
+     * three bytes in sym_buf for every four bytes in pending_buf. Each symbol
+     * in sym_buf is three bytes -- two for the distance and one for the
+     * literal/length. As each symbol is consumed, the pointer to the next
+     * sym_buf value to read moves forward three bytes. From that symbol, up to
+     * 31 bits are written to pending_buf. The closest the written pending_buf
+     * bits gets to the next sym_buf symbol to read is just before the last
+     * code is written. At that time, 31*(n - 2) bits have been written, just
+     * after 24*(n - 2) bits have been consumed from sym_buf. sym_buf starts at
+     * 8*n bits into pending_buf. (Note that the symbol buffer fills when n - 1
+     * symbols are written.) The closest the writing gets to what is unread is
+     * then n + 14 bits. Here n is lit_bufsize, which is 16384 by default, and
+     * can range from 128 to 32768.
+     *
+     * Therefore, at a minimum, there are 142 bits of space between what is
+     * written and what is read in the overlain buffers, so the symbols cannot
+     * be overwritten by the compressed data. That space is actually 139 bits,
+     * due to the three-bit fixed-code block header.
+     *
+     * That covers the case where either Z_FIXED is specified, forcing fixed
+     * codes, or when the use of fixed codes is chosen, because that choice
+     * results in a smaller compressed block than dynamic codes. That latter
+     * condition then assures that the above analysis also covers all dynamic
+     * blocks. A dynamic-code block will only be chosen to be emitted if it has
+     * fewer bits than a fixed-code block would for the same set of symbols.
+     * Therefore its average symbol length is assured to be less than 31. So
+     * the compressed data for a dynamic block also cannot overwrite the
+     * symbols from which it is being constructed.
+     */
+
+    s->pending_buf = (uchf *) ZALLOC(strm, s->lit_bufsize, 4);
+    s->pending_buf_size = (ulg)s->lit_bufsize * 4;

    if (s->window == Z_NULL || s->prev == Z_NULL || s->head == Z_NULL ||
        s->pending_buf == Z_NULL) {
@@ -337,8 +369,12 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
        deflateEnd (strm);
        return Z_MEM_ERROR;
    }
-    s->d_buf = overlay + s->lit_bufsize/sizeof(ush);
-    s->l_buf = s->pending_buf + (1+sizeof(ush))*s->lit_bufsize;
+    s->sym_buf = s->pending_buf + s->lit_bufsize;
+    s->sym_end = (s->lit_bufsize - 1) * 3;
+    /* We avoid equality with lit_bufsize*3 because of wraparound at 64K
+     * on 16 bit machines and because stored blocks are restricted to
+     * 64K-1 bytes.
+     */

    s->level = level;
    s->strategy = strategy;
@@ -350,7 +386,7 @@ int ZEXPORT deflateInit2_(strm, level, method, windowBits, memLevel, strategy,
 /* =========================================================================
 * Check for a valid deflate stream state. Return 0 if ok, 1 if not.
 */
-local int deflateStateCheck (strm)
+local int deflateStateCheck(strm)
    z_streamp strm;
 {
    deflate_state *s;
@@ -373,7 +409,7 @@ local int deflateStateCheck (strm)
 }

 /* ========================================================================= */
-int ZEXPORT deflateSetDictionary (strm, dictionary, dictLength)
+int ZEXPORT deflateSetDictionary(strm, dictionary, dictLength)
    z_streamp strm;
    const Bytef *dictionary;
    uInt  dictLength;
@@ -442,7 +478,7 @@ int ZEXPORT deflateSetDictionary (strm, dictionary, dictLength)
 }

 /* ========================================================================= */
-int ZEXPORT deflateGetDictionary (strm, dictionary, dictLength)
+int ZEXPORT deflateGetDictionary(strm, dictionary, dictLength)
    z_streamp strm;
    Bytef *dictionary;
    uInt  *dictLength;
@@ -464,7 +500,7 @@ int ZEXPORT deflateGetDictionary (strm, dictionary, dictLength)
 }

 /* ========================================================================= */
-int ZEXPORT deflateResetKeep (strm)
+int ZEXPORT deflateResetKeep(strm)
    z_streamp strm;
 {
    deflate_state *s;
@@ -488,13 +524,13 @@ int ZEXPORT deflateResetKeep (strm)
 #ifdef GZIP
        s->wrap == 2 ? GZIP_STATE :
 #endif
-        s->wrap ? INIT_STATE : BUSY_STATE;
+        INIT_STATE;
    strm->adler =
 #ifdef GZIP
        s->wrap == 2 ? crc32(0L, Z_NULL, 0) :
 #endif
        adler32(0L, Z_NULL, 0);
-    s->last_flush = Z_NO_FLUSH;
+    s->last_flush = -2;

    _tr_init(s);

@@ -502,7 +538,7 @@ int ZEXPORT deflateResetKeep (strm)
 }

 /* ========================================================================= */
-int ZEXPORT deflateReset (strm)
+int ZEXPORT deflateReset(strm)
    z_streamp strm;
 {
    int ret;
@@ -514,7 +550,7 @@ int ZEXPORT deflateReset (strm)
 }

 /* ========================================================================= */
-int ZEXPORT deflateSetHeader (strm, head)
+int ZEXPORT deflateSetHeader(strm, head)
    z_streamp strm;
    gz_headerp head;
 {
@@ -525,7 +561,7 @@ int ZEXPORT deflateSetHeader (strm, head)
 }

 /* ========================================================================= */
-int ZEXPORT deflatePending (strm, pending, bits)
+int ZEXPORT deflatePending(strm, pending, bits)
    unsigned *pending;
    int *bits;
    z_streamp strm;
@@ -539,7 +575,7 @@ int ZEXPORT deflatePending (strm, pending, bits)
 }

 /* ========================================================================= */
-int ZEXPORT deflatePrime (strm, bits, value)
+int ZEXPORT deflatePrime(strm, bits, value)
    z_streamp strm;
    int bits;
    int value;
@@ -549,7 +585,8 @@ int ZEXPORT deflatePrime (strm, bits, value)

    if (deflateStateCheck(strm)) return Z_STREAM_ERROR;
    s = strm->state;
-    if ((Bytef *)(s->d_buf) < s->pending_out + ((Buf_size + 7) >> 3))
+    if (bits < 0 || bits > 16 ||
+        s->sym_buf < s->pending_out + ((Buf_size + 7) >> 3))
        return Z_BUF_ERROR;
    do {
        put = Buf_size - s->bi_valid;
@@ -587,12 +624,12 @@ int ZEXPORT deflateParams(strm, level, strategy)
    func = configuration_table[s->level].func;

    if ((strategy != s->strategy || func != configuration_table[level].func) &&
-        s->high_water) {
+        s->last_flush != -2) {
        /* Flush the last buffer: */
        int err = deflate(strm, Z_BLOCK);
        if (err == Z_STREAM_ERROR)
            return err;
-        if (strm->avail_out == 0)
+        if (strm->avail_in || (s->strstart - s->block_start) + s->lookahead)
            return Z_BUF_ERROR;
    }
    if (s->level != level) {
@@ -633,36 +670,50 @@ int ZEXPORT deflateTune(strm, good_length, max_lazy, nice_length, max_chain)
 }

 /* =========================================================================
- * For the default windowBits of 15 and memLevel of 8, this function returns
- * a close to exact, as well as small, upper bound on the compressed size.
- * They are coded as constants here for a reason--if the #define's are
- * changed, then this function needs to be changed as well.  The return
- * value for 15 and 8 only works for those exact settings.
+ * For the default windowBits of 15 and memLevel of 8, this function returns a
+ * close to exact, as well as small, upper bound on the compressed size. This
+ * is an expansion of ~0.03%, plus a small constant.
 *
- * For any setting other than those defaults for windowBits and memLevel,
- * the value returned is a conservative worst case for the maximum expansion
- * resulting from using fixed blocks instead of stored blocks, which deflate
- * can emit on compressed data for some combinations of the parameters.
+ * For any setting other than those defaults for windowBits and memLevel, one
+ * of two worst case bounds is returned. This is at most an expansion of ~4% or
+ * ~13%, plus a small constant.
 *
- * This function could be more sophisticated to provide closer upper bounds for
- * every combination of windowBits and memLevel.  But even the conservative
- * upper bound of about 14% expansion does not seem onerous for output buffer
- * allocation.
+ * Both the 0.03% and 4% derive from the overhead of stored blocks. The first
+ * one is for stored blocks of 16383 bytes (memLevel == 8), whereas the second
+ * is for stored blocks of 127 bytes (the worst case memLevel == 1). The
+ * expansion results from five bytes of header for each stored block.
+ *
+ * The larger expansion of 13% results from a window size less than or equal to
+ * the symbols buffer size (windowBits <= memLevel + 7). In that case some of
+ * the data being compressed may have slid out of the sliding window, impeding
+ * a stored block from being emitted. Then the only choice is a fixed or
+ * dynamic block, where a fixed block limits the maximum expansion to 9 bits
+ * per 8-bit byte, plus 10 bits for every block. The smallest block size for
+ * which this can occur is 255 (memLevel == 2).
+ *
+ * Shifts are used to approximate divisions, for speed.
 */
 uLong ZEXPORT deflateBound(strm, sourceLen)
    z_streamp strm;
    uLong sourceLen;
 {
    deflate_state *s;
-    uLong complen, wraplen;
+    uLong fixedlen, storelen, wraplen;

-    /* conservative upper bound for compressed data */
-    complen = sourceLen +
-              ((sourceLen + 7) >> 3) + ((sourceLen + 63) >> 6) + 5;
+    /* upper bound for fixed blocks with 9-bit literals and length 255
+       (memLevel == 2, which is the lowest that may not use stored blocks) --
+       ~13% overhead plus a small constant */
+    fixedlen = sourceLen + (sourceLen >> 3) + (sourceLen >> 8) +
+               (sourceLen >> 9) + 4;

-    /* if can't get parameters, return conservative bound plus zlib wrapper */
+    /* upper bound for stored blocks with length 127 (memLevel == 1) --
+       ~4% overhead plus a small constant */
+    storelen = sourceLen + (sourceLen >> 5) + (sourceLen >> 7) +
+               (sourceLen >> 11) + 7;
+
+    /* if can't get parameters, return larger bound plus a zlib wrapper */
    if (deflateStateCheck(strm))
-        return complen + 6;
+        return (fixedlen > storelen ? fixedlen : storelen) + 6;

    /* compute wrapper length */
    s = strm->state;
@@ -699,11 +750,12 @@ uLong ZEXPORT deflateBound(strm, sourceLen)
        wraplen = 6;
    }

-    /* if not default parameters, return conservative bound */
+    /* if not default parameters, return one of the conservative bounds */
    if (s->w_bits != 15 || s->hash_bits != 8 + 7)
-        return complen + wraplen;
+        return (s->w_bits <= s->hash_bits ? fixedlen : storelen) + wraplen;

-    /* default settings: return tight bound for that case */
+    /* default settings: return tight bound for that case -- ~0.03% overhead
+       plus a small constant */
    return sourceLen + (sourceLen >> 12) + (sourceLen >> 14) +
           (sourceLen >> 25) + 13 - 6 + wraplen;
 }
@@ -713,7 +765,7 @@ uLong ZEXPORT deflateBound(strm, sourceLen)
 * IN assertion: the stream state is correct and there is enough room in
 * pending_buf.
 */
-local void putShortMSB (s, b)
+local void putShortMSB(s, b)
    deflate_state *s;
    uInt b;
 {
@@ -760,7 +812,7 @@ local void flush_pending(strm)
    } while (0)

 /* ========================================================================= */
-int ZEXPORT deflate (strm, flush)
+int ZEXPORT deflate(strm, flush)
    z_streamp strm;
    int flush;
 {
@@ -811,9 +863,11 @@ int ZEXPORT deflate (strm, flush)
    }

    /* Write the header */
+    if (s->status == INIT_STATE && s->wrap == 0)
+        s->status = BUSY_STATE;
    if (s->status == INIT_STATE) {
        /* zlib header */
-        uInt header = (Z_DEFLATED + ((s->w_bits-8)<<4)) << 8;
+        uInt header = (Z_DEFLATED + ((s->w_bits - 8) << 4)) << 8;
        uInt level_flags;

        if (s->strategy >= Z_HUFFMAN_ONLY || s->level < 2)
@@ -1073,7 +1127,7 @@ int ZEXPORT deflate (strm, flush)
 }

 /* ========================================================================= */
-int ZEXPORT deflateEnd (strm)
+int ZEXPORT deflateEnd(strm)
    z_streamp strm;
 {
    int status;
@@ -1099,7 +1153,7 @@ int ZEXPORT deflateEnd (strm)
 * To simplify the source, this is not supported for 16-bit MSDOS (which
 * doesn't have enough memory anyway to duplicate compression states).
 */
-int ZEXPORT deflateCopy (dest, source)
+int ZEXPORT deflateCopy(dest, source)
    z_streamp dest;
    z_streamp source;
 {
@@ -1108,7 +1162,6 @@ int ZEXPORT deflateCopy (dest, source)
 #else
    deflate_state *ds;
    deflate_state *ss;
-    ushf *overlay;


    if (deflateStateCheck(source) || dest == Z_NULL) {
@@ -1128,8 +1181,7 @@ int ZEXPORT deflateCopy (dest, source)
    ds->window = (Bytef *) ZALLOC(dest, ds->w_size, 2*sizeof(Byte));
    ds->prev   = (Posf *)  ZALLOC(dest, ds->w_size, sizeof(Pos));
    ds->head   = (Posf *)  ZALLOC(dest, ds->hash_size, sizeof(Pos));
-    overlay = (ushf *) ZALLOC(dest, ds->lit_bufsize, sizeof(ush)+2);
-    ds->pending_buf = (uchf *) overlay;
+    ds->pending_buf = (uchf *) ZALLOC(dest, ds->lit_bufsize, 4);

    if (ds->window == Z_NULL || ds->prev == Z_NULL || ds->head == Z_NULL ||
        ds->pending_buf == Z_NULL) {
@@ -1143,8 +1195,7 @@ int ZEXPORT deflateCopy (dest, source)
    zmemcpy(ds->pending_buf, ss->pending_buf, (uInt)ds->pending_buf_size);

    ds->pending_out = ds->pending_buf + (ss->pending_out - ss->pending_buf);
-    ds->d_buf = overlay + ds->lit_bufsize/sizeof(ush);
-    ds->l_buf = ds->pending_buf + (1+sizeof(ush))*ds->lit_bufsize;
+    ds->sym_buf = ds->pending_buf + ds->lit_bufsize;

    ds->l_desc.dyn_tree = ds->dyn_ltree;
    ds->d_desc.dyn_tree = ds->dyn_dtree;
@@ -1191,7 +1242,7 @@ local unsigned read_buf(strm, buf, size)
 /* ===========================================================================
 * Initialize the "longest match" routines for a new zlib stream
 */
-local void lm_init (s)
+local void lm_init(s)
    deflate_state *s;
 {
    s->window_size = (ulg)2L*s->w_size;
@@ -1212,11 +1263,6 @@ local void lm_init (s)
    s->match_length = s->prev_length = MIN_MATCH-1;
    s->match_available = 0;
    s->ins_h = 0;
-#ifndef FASTEST
-#ifdef ASMV
-    match_init(); /* initialize the asm code */
-#endif
-#endif
 }

 #ifndef FASTEST
@@ -1229,10 +1275,6 @@ local void lm_init (s)
 *   string (strstart) and its distance is <= MAX_DIST, and prev_length >= 1
 * OUT assertion: the match length is not greater than s->lookahead.
 */
-#ifndef ASMV
-/* For 80x86 and 680x0, an optimized version will be provided in match.asm or
- * match.S. The code will be functionally equivalent.
- */
 local uInt longest_match(s, cur_match)
    deflate_state *s;
    IPos cur_match;                             /* current match */
@@ -1257,10 +1299,10 @@ local uInt longest_match(s, cur_match)
     */
    register Bytef *strend = s->window + s->strstart + MAX_MATCH - 1;
    register ush scan_start = *(ushf*)scan;
-    register ush scan_end   = *(ushf*)(scan+best_len-1);
+    register ush scan_end   = *(ushf*)(scan + best_len - 1);
 #else
    register Bytef *strend = s->window + s->strstart + MAX_MATCH;
-    register Byte scan_end1  = scan[best_len-1];
+    register Byte scan_end1  = scan[best_len - 1];
    register Byte scan_end   = scan[best_len];
 #endif

@@ -1278,7 +1320,8 @@ local uInt longest_match(s, cur_match)
     */
    if ((uInt)nice_match > s->lookahead) nice_match = (int)s->lookahead;

-    Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
+    Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD,
+           "need lookahead");

    do {
        Assert(cur_match < s->strstart, "no future");
@@ -1296,43 +1339,44 @@ local uInt longest_match(s, cur_match)
        /* This code assumes sizeof(unsigned short) == 2. Do not use
         * UNALIGNED_OK if your compiler uses a different size.
         */
-        if (*(ushf*)(match+best_len-1) != scan_end ||
+        if (*(ushf*)(match + best_len - 1) != scan_end ||
            *(ushf*)match != scan_start) continue;

        /* It is not necessary to compare scan[2] and match[2] since they are
         * always equal when the other bytes match, given that the hash keys
         * are equal and that HASH_BITS >= 8. Compare 2 bytes at a time at
-         * strstart+3, +5, ... up to strstart+257. We check for insufficient
+         * strstart + 3, + 5, up to strstart + 257. We check for insufficient
         * lookahead only every 4th comparison; the 128th check will be made
-         * at strstart+257. If MAX_MATCH-2 is not a multiple of 8, it is
+         * at strstart + 257. If MAX_MATCH-2 is not a multiple of 8, it is
         * necessary to put more guard bytes at the end of the window, or
         * to check more often for insufficient lookahead.
         */
        Assert(scan[2] == match[2], "scan[2]?");
        scan++, match++;
        do {
-        } while (*(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
-                 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
-                 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
-                 *(ushf*)(scan+=2) == *(ushf*)(match+=2) &&
+        } while (*(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
+                 *(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
+                 *(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
+                 *(ushf*)(scan += 2) == *(ushf*)(match += 2) &&
                 scan < strend);
        /* The funny "do {}" generates better code on most compilers */

-        /* Here, scan <= window+strstart+257 */
-        Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
+        /* Here, scan <= window + strstart + 257 */
+        Assert(scan <= s->window + (unsigned)(s->window_size - 1),
+               "wild scan");
        if (*scan == *match) scan++;

-        len = (MAX_MATCH - 1) - (int)(strend-scan);
+        len = (MAX_MATCH - 1) - (int)(strend - scan);
        scan = strend - (MAX_MATCH-1);

 #else /* UNALIGNED_OK */

-        if (match[best_len]   != scan_end  ||
-            match[best_len-1] != scan_end1 ||
-            *match            != *scan     ||
-            *++match          != scan[1])      continue;
+        if (match[best_len]     != scan_end  ||
+            match[best_len - 1] != scan_end1 ||
+            *match              != *scan     ||
+            *++match            != scan[1])      continue;

-        /* The check at best_len-1 can be removed because it will be made
+        /* The check at best_len - 1 can be removed because it will be made
         * again later. (This heuristic is not always a win.)
         * It is not necessary to compare scan[2] and match[2] since they
         * are always equal when the other bytes match, given that
@@ -1342,7 +1386,7 @@ local uInt longest_match(s, cur_match)
        Assert(*scan == *match, "match[2]?");

        /* We check for insufficient lookahead only every 8th comparison;
-         * the 256th check will be made at strstart+258.
+         * the 256th check will be made at strstart + 258.
         */
        do {
        } while (*++scan == *++match && *++scan == *++match &&
@@ -1351,7 +1395,8 @@ local uInt longest_match(s, cur_match)
                 *++scan == *++match && *++scan == *++match &&
                 scan < strend);

-        Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
+        Assert(scan <= s->window + (unsigned)(s->window_size - 1),
+               "wild scan");

        len = MAX_MATCH - (int)(strend - scan);
        scan = strend - MAX_MATCH;
@@ -1363,9 +1408,9 @@ local uInt longest_match(s, cur_match)
            best_len = len;
            if (len >= nice_match) break;
 #ifdef UNALIGNED_OK
-            scan_end = *(ushf*)(scan+best_len-1);
+            scan_end = *(ushf*)(scan + best_len - 1);
 #else
-            scan_end1  = scan[best_len-1];
+            scan_end1  = scan[best_len - 1];
            scan_end   = scan[best_len];
 #endif
        }
@@ -1375,7 +1420,6 @@ local uInt longest_match(s, cur_match)
    if ((uInt)best_len <= s->lookahead) return (uInt)best_len;
    return s->lookahead;
 }
-#endif /* ASMV */

 #else /* FASTEST */

@@ -1396,7 +1440,8 @@ local uInt longest_match(s, cur_match)
     */
    Assert(s->hash_bits >= 8 && MAX_MATCH == 258, "Code too clever");

-    Assert((ulg)s->strstart <= s->window_size-MIN_LOOKAHEAD, "need lookahead");
+    Assert((ulg)s->strstart <= s->window_size - MIN_LOOKAHEAD,
+           "need lookahead");

    Assert(cur_match < s->strstart, "no future");

@@ -1406,7 +1451,7 @@ local uInt longest_match(s, cur_match)
     */
    if (match[0] != scan[0] || match[1] != scan[1]) return MIN_MATCH-1;

-    /* The check at best_len-1 can be removed because it will be made
+    /* The check at best_len - 1 can be removed because it will be made
     * again later. (This heuristic is not always a win.)
     * It is not necessary to compare scan[2] and match[2] since they
     * are always equal when the other bytes match, given that
@@ -1416,7 +1461,7 @@ local uInt longest_match(s, cur_match)
    Assert(*scan == *match, "match[2]?");

    /* We check for insufficient lookahead only every 8th comparison;
-     * the 256th check will be made at strstart+258.
+     * the 256th check will be made at strstart + 258.
     */
    do {
    } while (*++scan == *++match && *++scan == *++match &&
@@ -1425,7 +1470,7 @@ local uInt longest_match(s, cur_match)
             *++scan == *++match && *++scan == *++match &&
             scan < strend);

-    Assert(scan <= s->window+(unsigned)(s->window_size-1), "wild scan");
+    Assert(scan <= s->window + (unsigned)(s->window_size - 1), "wild scan");

    len = MAX_MATCH - (int)(strend - scan);

@@ -1461,7 +1506,7 @@ local void check_match(s, start, match, length)
        z_error("invalid match");
    }
    if (z_verbose > 1) {
-        fprintf(stderr,"\\[%d,%d]", start-match, length);
+        fprintf(stderr,"\\[%d,%d]", start - match, length);
        do { putc(s->window[start++], stderr); } while (--length != 0);
    }
 }
@@ -1507,12 +1552,14 @@ local void fill_window(s)
        /* If the window is almost full and there is insufficient lookahead,
         * move the upper half to the lower one to make room in the upper half.
         */
-        if (s->strstart >= wsize+MAX_DIST(s)) {
+        if (s->strstart >= wsize + MAX_DIST(s)) {

-            zmemcpy(s->window, s->window+wsize, (unsigned)wsize - more);
+            zmemcpy(s->window, s->window + wsize, (unsigned)wsize - more);
            s->match_start -= wsize;
            s->strstart    -= wsize; /* we now have strstart >= MAX_DIST */
            s->block_start -= (long) wsize;
+            if (s->insert > s->strstart)
+                s->insert = s->strstart;
            slide_hash(s);
            more += wsize;
        }
@@ -1638,7 +1685,7 @@ local void fill_window(s)
 *
 * deflate_stored() is written to minimize the number of times an input byte is
 * copied. It is most efficient with large input and output buffers, which
- * maximizes the opportunites to have a single copy from next_in to next_out.
+ * maximizes the opportunities to have a single copy from next_in to next_out.
 */
 local block_state deflate_stored(s, flush)
    deflate_state *s;
@@ -1742,6 +1789,7 @@ local block_state deflate_stored(s, flush)
            s->matches = 2;         /* clear hash */
            zmemcpy(s->window, s->strm->next_in - s->w_size, s->w_size);
            s->strstart = s->w_size;
+            s->insert = s->strstart;
        }
        else {
            if (s->window_size - s->strstart <= used) {
@@ -1750,12 +1798,14 @@ local block_state deflate_stored(s, flush)
                zmemcpy(s->window, s->window + s->w_size, s->strstart);
                if (s->matches < 2)
                    s->matches++;   /* add a pending slide_hash() */
+                if (s->insert > s->strstart)
+                    s->insert = s->strstart;
            }
            zmemcpy(s->window + s->strstart, s->strm->next_in - used, used);
            s->strstart += used;
+            s->insert += MIN(used, s->w_size - s->insert);
        }
        s->block_start = s->strstart;
-        s->insert += MIN(used, s->w_size - s->insert);
    }
    if (s->high_water < s->strstart)
        s->high_water = s->strstart;
@@ -1770,7 +1820,7 @@ local block_state deflate_stored(s, flush)
        return block_done;

    /* Fill the window with any remaining input. */
-    have = s->window_size - s->strstart - 1;
+    have = s->window_size - s->strstart;
    if (s->strm->avail_in > have && s->block_start >= (long)s->w_size) {
        /* Slide the window down. */
        s->block_start -= s->w_size;
@@ -1779,12 +1829,15 @@ local block_state deflate_stored(s, flush)
        if (s->matches < 2)
            s->matches++;           /* add a pending slide_hash() */
        have += s->w_size;          /* more space now */
+        if (s->insert > s->strstart)
+            s->insert = s->strstart;
    }
    if (have > s->strm->avail_in)
        have = s->strm->avail_in;
    if (have) {
        read_buf(s->strm, s->window + s->strstart, have);
        s->strstart += have;
+        s->insert += MIN(have, s->w_size - s->insert);
    }
    if (s->high_water < s->strstart)
        s->high_water = s->strstart;
@@ -1842,7 +1895,7 @@ local block_state deflate_fast(s, flush)
            if (s->lookahead == 0) break; /* flush the current block */
        }

-        /* Insert the string window[strstart .. strstart+2] in the
+        /* Insert the string window[strstart .. strstart + 2] in the
         * dictionary, and set hash_head to the head of the hash chain:
         */
        hash_head = NIL;
@@ -1890,7 +1943,7 @@ local block_state deflate_fast(s, flush)
                s->strstart += s->match_length;
                s->match_length = 0;
                s->ins_h = s->window[s->strstart];
-                UPDATE_HASH(s, s->ins_h, s->window[s->strstart+1]);
+                UPDATE_HASH(s, s->ins_h, s->window[s->strstart + 1]);
 #if MIN_MATCH != 3
                Call UPDATE_HASH() MIN_MATCH-3 more times
 #endif
@@ -1901,7 +1954,7 @@ local block_state deflate_fast(s, flush)
        } else {
            /* No match, output a literal byte */
            Tracevv((stderr,"%c", s->window[s->strstart]));
-            _tr_tally_lit (s, s->window[s->strstart], bflush);
+            _tr_tally_lit(s, s->window[s->strstart], bflush);
            s->lookahead--;
            s->strstart++;
        }
@@ -1912,7 +1965,7 @@ local block_state deflate_fast(s, flush)
        FLUSH_BLOCK(s, 1);
        return finish_done;
    }
-    if (s->last_lit)
+    if (s->sym_next)
        FLUSH_BLOCK(s, 0);
    return block_done;
 }
@@ -1945,7 +1998,7 @@ local block_state deflate_slow(s, flush)
            if (s->lookahead == 0) break; /* flush the current block */
        }

-        /* Insert the string window[strstart .. strstart+2] in the
+        /* Insert the string window[strstart .. strstart + 2] in the
         * dictionary, and set hash_head to the head of the hash chain:
         */
        hash_head = NIL;
@@ -1987,17 +2040,17 @@ local block_state deflate_slow(s, flush)
            uInt max_insert = s->strstart + s->lookahead - MIN_MATCH;
            /* Do not insert strings in hash table beyond this. */

-            check_match(s, s->strstart-1, s->prev_match, s->prev_length);
+            check_match(s, s->strstart - 1, s->prev_match, s->prev_length);

-            _tr_tally_dist(s, s->strstart -1 - s->prev_match,
+            _tr_tally_dist(s, s->strstart - 1 - s->prev_match,
                           s->prev_length - MIN_MATCH, bflush);

            /* Insert in hash table all strings up to the end of the match.
-             * strstart-1 and strstart are already inserted. If there is not
+             * strstart - 1 and strstart are already inserted. If there is not
             * enough lookahead, the last two strings are not inserted in
             * the hash table.
             */
-            s->lookahead -= s->prev_length-1;
+            s->lookahead -= s->prev_length - 1;
            s->prev_length -= 2;
            do {
                if (++s->strstart <= max_insert) {
@@ -2015,8 +2068,8 @@ local block_state deflate_slow(s, flush)
             * single literal. If there was a match but the current match
             * is longer, truncate the previous match to a single literal.
             */
-            Tracevv((stderr,"%c", s->window[s->strstart-1]));
-            _tr_tally_lit(s, s->window[s->strstart-1], bflush);
+            Tracevv((stderr,"%c", s->window[s->strstart - 1]));
+            _tr_tally_lit(s, s->window[s->strstart - 1], bflush);
            if (bflush) {
                FLUSH_BLOCK_ONLY(s, 0);
            }
@@ -2034,8 +2087,8 @@ local block_state deflate_slow(s, flush)
    }
    Assert (flush != Z_NO_FLUSH, "no flush?");
    if (s->match_available) {
-        Tracevv((stderr,"%c", s->window[s->strstart-1]));
-        _tr_tally_lit(s, s->window[s->strstart-1], bflush);
+        Tracevv((stderr,"%c", s->window[s->strstart - 1]));
+        _tr_tally_lit(s, s->window[s->strstart - 1], bflush);
        s->match_available = 0;
    }
    s->insert = s->strstart < MIN_MATCH-1 ? s->strstart : MIN_MATCH-1;
@@ -2043,7 +2096,7 @@ local block_state deflate_slow(s, flush)
        FLUSH_BLOCK(s, 1);
        return finish_done;
    }
-    if (s->last_lit)
+    if (s->sym_next)
        FLUSH_BLOCK(s, 0);
    return block_done;
 }
@@ -2092,7 +2145,8 @@ local block_state deflate_rle(s, flush)
                if (s->match_length > s->lookahead)
                    s->match_length = s->lookahead;
            }
-            Assert(scan <= s->window+(uInt)(s->window_size-1), "wild scan");
+            Assert(scan <= s->window + (uInt)(s->window_size - 1),
+                   "wild scan");
        }

        /* Emit match if have run of MIN_MATCH or longer, else emit literal */
@@ -2107,7 +2161,7 @@ local block_state deflate_rle(s, flush)
        } else {
            /* No match, output a literal byte */
            Tracevv((stderr,"%c", s->window[s->strstart]));
-            _tr_tally_lit (s, s->window[s->strstart], bflush);
+            _tr_tally_lit(s, s->window[s->strstart], bflush);
            s->lookahead--;
            s->strstart++;
        }
@@ -2118,7 +2172,7 @@ local block_state deflate_rle(s, flush)
        FLUSH_BLOCK(s, 1);
        return finish_done;
    }
-    if (s->last_lit)
+    if (s->sym_next)
        FLUSH_BLOCK(s, 0);
    return block_done;
 }
@@ -2147,7 +2201,7 @@ local block_state deflate_huff(s, flush)
        /* Output a literal byte */
        s->match_length = 0;
        Tracevv((stderr,"%c", s->window[s->strstart]));
-        _tr_tally_lit (s, s->window[s->strstart], bflush);
+        _tr_tally_lit(s, s->window[s->strstart], bflush);
        s->lookahead--;
        s->strstart++;
        if (bflush) FLUSH_BLOCK(s, 0);
@@ -2157,7 +2211,7 @@ local block_state deflate_huff(s, flush)
        FLUSH_BLOCK(s, 1);
        return finish_done;
    }
-    if (s->last_lit)
+    if (s->sym_next)
        FLUSH_BLOCK(s, 0);
    return block_done;
 }
@@ -1,5 +1,5 @@
 /* deflate.h -- internal compression state
- * Copyright (C) 1995-2016 Jean-loup Gailly
+ * Copyright (C) 1995-2018 Jean-loup Gailly
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -217,7 +217,7 @@ typedef struct internal_state {
    /* Depth of each subtree used as tie breaker for trees of equal frequency
     */

-    uchf *l_buf;          /* buffer for literals or lengths */
+    uchf *sym_buf;        /* buffer for distances and literals/lengths */

    uInt  lit_bufsize;
    /* Size of match buffer for literals/lengths.  There are 4 reasons for
@@ -239,13 +239,8 @@ typedef struct internal_state {
     *   - I can't count above 4
     */

-    uInt last_lit;      /* running index in l_buf */
-
-    ushf *d_buf;
-    /* Buffer for distances. To simplify the code, d_buf and l_buf have
-     * the same number of elements. To use different lengths, an extra flag
-     * array would be necessary.
-     */
+    uInt sym_next;      /* running index in sym_buf */
+    uInt sym_end;       /* symbol table full when sym_next reaches this */

    ulg opt_len;        /* bit length of current block with optimal trees */
    ulg static_len;     /* bit length of current block with static trees */
@@ -325,20 +320,22 @@ void ZLIB_INTERNAL _tr_stored_block OF((deflate_state *s, charf *buf,

 # define _tr_tally_lit(s, c, flush) \
  { uch cc = (c); \
-    s->d_buf[s->last_lit] = 0; \
-    s->l_buf[s->last_lit++] = cc; \
+    s->sym_buf[s->sym_next++] = 0; \
+    s->sym_buf[s->sym_next++] = 0; \
+    s->sym_buf[s->sym_next++] = cc; \
    s->dyn_ltree[cc].Freq++; \
-    flush = (s->last_lit == s->lit_bufsize-1); \
+    flush = (s->sym_next == s->sym_end); \
   }
 # define _tr_tally_dist(s, distance, length, flush) \
  { uch len = (uch)(length); \
    ush dist = (ush)(distance); \
-    s->d_buf[s->last_lit] = dist; \
-    s->l_buf[s->last_lit++] = len; \
+    s->sym_buf[s->sym_next++] = (uch)dist; \
+    s->sym_buf[s->sym_next++] = (uch)(dist >> 8); \
+    s->sym_buf[s->sym_next++] = len; \
    dist--; \
    s->dyn_ltree[_length_code[len]+LITERALS+1].Freq++; \
    s->dyn_dtree[d_code(dist)].Freq++; \
-    flush = (s->last_lit == s->lit_bufsize-1); \
+    flush = (s->sym_next == s->sym_end); \
  }
 #else
 # define _tr_tally_lit(s, c, flush) flush = _tr_tally(s, 0, c)
@@ -1,5 +1,5 @@
 /* gzguts.h -- zlib internal header definitions for gz* operations
- * Copyright (C) 2004, 2005, 2010, 2011, 2012, 2013, 2016 Mark Adler
+ * Copyright (C) 2004-2019 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -39,7 +39,7 @@
 #  include <io.h>
 #endif

-#if defined(_WIN32) || defined(__CYGWIN__)
+#if defined(_WIN32)
 #  define WIDECHAR
 #endif

@@ -190,6 +190,7 @@ typedef struct {
        /* just for writing */
    int level;              /* compression level */
    int strategy;           /* compression strategy */
+    int reset;              /* true if a reset is pending after a Z_FINISH */
        /* seek request */
    z_off64_t skip;         /* amount to skip (already rewound if backwards) */
    int seek;               /* true if seek request pending */
@@ -1,11 +1,11 @@
 /* gzlib.c -- zlib functions common to reading and writing gzip files
- * Copyright (C) 2004-2017 Mark Adler
+ * Copyright (C) 2004-2019 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

 #include "gzguts.h"

-#if defined(_WIN32) && !defined(__BORLANDC__) && !defined(__MINGW32__)
+#if defined(_WIN32) && !defined(__BORLANDC__)
 #  define LSEEK _lseeki64
 #else
 #if defined(_LARGEFILE64_SOURCE) && _LFS64_LARGEFILE-0
@@ -30,7 +30,7 @@ local gzFile gz_open OF((const void *, int, const char *));

   The gz_strwinerror function does not change the current setting of
   GetLastError. */
-char ZLIB_INTERNAL *gz_strwinerror (error)
+char ZLIB_INTERNAL *gz_strwinerror(error)
     DWORD error;
 {
    static char buf[1024];
@@ -81,6 +81,8 @@ local void gz_reset(state)
        state->past = 0;            /* have not read past end yet */
        state->how = LOOK;          /* look for gzip header */
    }
+    else                            /* for writing ... */
+        state->reset = 0;           /* no deflateReset pending */
    state->seek = 0;                /* no seek request pending */
    gz_error(state, Z_OK, NULL);    /* clear error */
    state->x.pos = 0;               /* no uncompressed data yet */
@@ -397,7 +399,7 @@ z_off64_t ZEXPORT gzseek64(file, offset, whence)
    /* if within raw area while reading, just go there */
    if (state->mode == GZ_READ && state->how == COPY &&
            state->x.pos + offset >= 0) {
-        ret = LSEEK(state->fd, offset - state->x.have, SEEK_CUR);
+        ret = LSEEK(state->fd, offset - (z_off64_t)state->x.have, SEEK_CUR);
        if (ret == -1)
            return -1;
        state->x.have = 0;
@@ -1,5 +1,5 @@
 /* gzread.c -- zlib functions for reading gzip files
- * Copyright (C) 2004, 2005, 2010, 2011, 2012, 2013, 2016 Mark Adler
+ * Copyright (C) 2004-2017 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -157,11 +157,9 @@ local int gz_look(state)
       the output buffer is larger than the input buffer, which also assures
       space for gzungetc() */
    state->x.next = state->out;
-    if (strm->avail_in) {
-        memcpy(state->x.next, strm->next_in, strm->avail_in);
-        state->x.have = strm->avail_in;
-        strm->avail_in = 0;
-    }
+    memcpy(state->x.next, strm->next_in, strm->avail_in);
+    state->x.have = strm->avail_in;
+    strm->avail_in = 0;
    state->how = COPY;
    state->direct = 1;
    return 0;
@@ -314,9 +312,9 @@ local z_size_t gz_read(state, buf, len)
    got = 0;
    do {
        /* set n to the maximum amount of len that fits in an unsigned int */
-        n = -1;
+        n = (unsigned)-1;
        if (n > len)
-            n = (int)len;
+            n = (unsigned)len;

        /* first just try copying data from the output buffer */
        if (state->x.have) {
@@ -397,7 +395,7 @@ int ZEXPORT gzread(file, buf, len)
    }

    /* read len or fewer bytes to buf */
-    len = (int)gz_read(state, buf, len);
+    len = (unsigned)gz_read(state, buf, len);

    /* check for an error */
    if (len == 0 && state->err != Z_OK && state->err != Z_BUF_ERROR)
@@ -447,7 +445,6 @@ z_size_t ZEXPORT gzfread(buf, size, nitems, file)
 int ZEXPORT gzgetc(file)
    gzFile file;
 {
-    int ret;
    unsigned char buf[1];
    gz_statep state;

@@ -469,8 +466,7 @@ int ZEXPORT gzgetc(file)
    }

    /* nothing there -- try gz_read() */
-    ret = (int)gz_read(state, buf, 1);
-    return ret < 1 ? -1 : buf[0];
+    return gz_read(state, buf, 1) < 1 ? -1 : buf[0];
 }

 int ZEXPORT gzgetc_(file)
@@ -1,5 +1,5 @@
 /* gzwrite.c -- zlib functions for writing gzip files
- * Copyright (C) 2004-2017 Mark Adler
+ * Copyright (C) 2004-2019 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -97,6 +97,15 @@ local int gz_comp(state, flush)
        return 0;
    }

+    /* check for a pending reset */
+    if (state->reset) {
+        /* don't start a new gzip member unless there is data to write */
+        if (strm->avail_in == 0)
+            return 0;
+        deflateReset(strm);
+        state->reset = 0;
+    }
+
    /* run deflate() on provided input until it produces no more output */
    ret = Z_OK;
    do {
@@ -134,7 +143,7 @@ local int gz_comp(state, flush)

    /* if that completed a deflate stream, allow another to start */
    if (flush == Z_FINISH)
-        deflateReset(strm);
+        state->reset = 1;

    /* all done, no errors */
    return 0;
@@ -209,7 +218,7 @@ local z_size_t gz_write(state, buf, len)
                              state->in);
            copy = state->size - have;
            if (copy > len)
-                copy = (int)len;
+                copy = (unsigned)len;
            memcpy(state->in + have, buf, copy);
            state->strm.avail_in += copy;
            state->x.pos += copy;
@@ -229,7 +238,7 @@ local z_size_t gz_write(state, buf, len)
        do {
            unsigned n = (unsigned)-1;
            if (n > len)
-                n = (int)len;
+                n = (unsigned)len;
            state->strm.avail_in = n;
            state->x.pos += n;
            if (gz_comp(state, Z_NO_FLUSH) == -1)
@@ -349,12 +358,11 @@ int ZEXPORT gzputc(file, c)
 }

 /* -- see zlib.h -- */
-int ZEXPORT gzputs(file, str)
+int ZEXPORT gzputs(file, s)
    gzFile file;
-    const char *str;
+    const char *s;
 {
-    int ret;
-    z_size_t len;
+    z_size_t len, put;
    gz_statep state;

    /* get internal structure */
@@ -367,9 +375,13 @@ int ZEXPORT gzputs(file, str)
        return -1;

    /* write string */
-    len = strlen(str);
-    ret = (int)gz_write(state, str, len);
-    return ret == 0 && len != 0 ? -1 : ret;
+    len = strlen(s);
+    if ((int)len < 0 || (unsigned)len != len) {
+        gz_error(state, Z_STREAM_ERROR, "string length does not fit in int");
+        return -1;
+    }
+    put = gz_write(state, s, len);
+    return put < len ? -1 : (int)len;
 }

 #if defined(STDC) || defined(Z_HAVE_STDARG_H)
@@ -441,7 +453,7 @@ int ZEXPORTVA gzvprintf(gzFile file, const char *format, va_list va)
        strm->avail_in = state->size;
        if (gz_comp(state, Z_NO_FLUSH) == -1)
            return state->err;
-        memcpy(state->in, state->in + state->size, left);
+        memmove(state->in, state->in + state->size, left);
        strm->next_in = state->in;
        strm->avail_in = left;
    }
@@ -462,7 +474,7 @@ int ZEXPORTVA gzprintf(gzFile file, const char *format, ...)
 #else /* !STDC && !Z_HAVE_STDARG_H */

 /* -- see zlib.h -- */
-int ZEXPORTVA gzprintf (file, format, a1, a2, a3, a4, a5, a6, a7, a8, a9, a10,
+int ZEXPORTVA gzprintf(file, format, a1, a2, a3, a4, a5, a6, a7, a8, a9, a10,
                       a11, a12, a13, a14, a15, a16, a17, a18, a19, a20)
    gzFile file;
    const char *format;
@@ -540,7 +552,7 @@ int ZEXPORTVA gzprintf (file, format, a1, a2, a3, a4, a5, a6, a7, a8, a9, a10,
        strm->avail_in = state->size;
        if (gz_comp(state, Z_NO_FLUSH) == -1)
            return state->err;
-        memcpy(state->in, state->in + state->size, left);
+        memmove(state->in, state->in + state->size, left);
        strm->next_in = state->in;
        strm->avail_in = left;
    }
@@ -1,5 +1,5 @@
 /* infback.c -- inflate using a call-back interface
- * Copyright (C) 1995-2016 Mark Adler
+ * Copyright (C) 1995-2022 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -66,6 +66,7 @@ int stream_size;
    state->window = window;
    state->wnext = 0;
    state->whave = 0;
+    state->sane = 1;
    return Z_OK;
 }

@@ -477,6 +478,7 @@ void FAR *out_desc;
            }
            Tracev((stderr, "inflate:       codes ok\n"));
            state->mode = LEN;
+                /* fallthrough */

        case LEN:
            /* use inflate_fast() if we have enough input and output */
@@ -604,25 +606,27 @@ void FAR *out_desc;
            break;

        case DONE:
-            /* inflate stream terminated properly -- write leftover output */
+            /* inflate stream terminated properly */
            ret = Z_STREAM_END;
-            if (left < state->wsize) {
-                if (out(out_desc, state->window, state->wsize - left))
-                    ret = Z_BUF_ERROR;
-            }
            goto inf_leave;

        case BAD:
            ret = Z_DATA_ERROR;
            goto inf_leave;

-        default:                /* can't happen, but makes compilers happy */
+        default:
+            /* can't happen, but makes compilers happy */
            ret = Z_STREAM_ERROR;
            goto inf_leave;
        }

-    /* Return unused input */
+    /* Write leftover output and return unused input */
  inf_leave:
+    if (left < state->wsize) {
+        if (out(out_desc, state->window, state->wsize - left) &&
+            ret == Z_STREAM_END)
+            ret = Z_BUF_ERROR;
+    }
    strm->next_in = next;
    strm->avail_in = have;
    return ret;
@@ -70,7 +70,7 @@ unsigned start;         /* inflate()'s starting value for strm->avail_out */
    code const FAR *dcode;      /* local strm->distcode */
    unsigned lmask;             /* mask for first level of length codes */
    unsigned dmask;             /* mask for first level of distance codes */
-    code here;                  /* retrieved table entry */
+    code const *here;           /* retrieved table entry */
    unsigned op;                /* code bits, operation, extra bits, or */
                                /*  window position, window bytes to copy */
    unsigned len;               /* match length, unused bytes */
@@ -107,20 +107,20 @@ unsigned start;         /* inflate()'s starting value for strm->avail_out */
            hold += (unsigned long)(*in++) << bits;
            bits += 8;
        }
-        here = lcode[hold & lmask];
+        here = lcode + (hold & lmask);
      dolen:
-        op = (unsigned)(here.bits);
+        op = (unsigned)(here->bits);
        hold >>= op;
        bits -= op;
-        op = (unsigned)(here.op);
+        op = (unsigned)(here->op);
        if (op == 0) {                          /* literal */
-            Tracevv((stderr, here.val >= 0x20 && here.val < 0x7f ?
+            Tracevv((stderr, here->val >= 0x20 && here->val < 0x7f ?
                    "inflate:         literal '%c'\n" :
-                    "inflate:         literal 0x%02x\n", here.val));
-            *out++ = (unsigned char)(here.val);
+                    "inflate:         literal 0x%02x\n", here->val));
+            *out++ = (unsigned char)(here->val);
        }
        else if (op & 16) {                     /* length base */
-            len = (unsigned)(here.val);
+            len = (unsigned)(here->val);
            op &= 15;                           /* number of extra bits */
            if (op) {
                if (bits < op) {
@@ -138,14 +138,14 @@ unsigned start;         /* inflate()'s starting value for strm->avail_out */
                hold += (unsigned long)(*in++) << bits;
                bits += 8;
            }
-            here = dcode[hold & dmask];
+            here = dcode + (hold & dmask);
          dodist:
-            op = (unsigned)(here.bits);
+            op = (unsigned)(here->bits);
            hold >>= op;
            bits -= op;
-            op = (unsigned)(here.op);
+            op = (unsigned)(here->op);
            if (op & 16) {                      /* distance base */
-                dist = (unsigned)(here.val);
+                dist = (unsigned)(here->val);
                op &= 15;                       /* number of extra bits */
                if (bits < op) {
                    hold += (unsigned long)(*in++) << bits;
@@ -264,7 +264,7 @@ unsigned start;         /* inflate()'s starting value for strm->avail_out */
                }
            }
            else if ((op & 64) == 0) {          /* 2nd level distance code */
-                here = dcode[here.val + (hold & ((1U << op) - 1))];
+                here = dcode + here->val + (hold & ((1U << op) - 1));
                goto dodist;
            }
            else {
@@ -274,7 +274,7 @@ unsigned start;         /* inflate()'s starting value for strm->avail_out */
            }
        }
        else if ((op & 64) == 0) {              /* 2nd level length code */
-            here = lcode[here.val + (hold & ((1U << op) - 1))];
+            here = lcode + here->val + (hold & ((1U << op) - 1));
            goto dolen;
        }
        else if (op & 32) {                     /* end-of-block */
@@ -1,5 +1,5 @@
 /* inflate.c -- zlib decompression
- * Copyright (C) 1995-2016 Mark Adler
+ * Copyright (C) 1995-2022 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -130,6 +130,7 @@ z_streamp strm;
    state->mode = HEAD;
    state->last = 0;
    state->havedict = 0;
+    state->flags = -1;
    state->dmax = 32768U;
    state->head = Z_NULL;
    state->hold = 0;
@@ -167,6 +168,8 @@ int windowBits;

    /* extract wrap request from windowBits parameter */
    if (windowBits < 0) {
+        if (windowBits < -15)
+            return Z_STREAM_ERROR;
        wrap = 0;
        windowBits = -windowBits;
    }
@@ -448,10 +451,10 @@ unsigned copy;

 /* check function to use adler32() for zlib or crc32() for gzip */
 #ifdef GUNZIP
-#  define UPDATE(check, buf, len) \
+#  define UPDATE_CHECK(check, buf, len) \
    (state->flags ? crc32(check, buf, len) : adler32(check, buf, len))
 #else
-#  define UPDATE(check, buf, len) adler32(check, buf, len)
+#  define UPDATE_CHECK(check, buf, len) adler32(check, buf, len)
 #endif

 /* check macros for header crc */
@@ -671,7 +674,6 @@ int flush;
                state->mode = FLAGS;
                break;
            }
-            state->flags = 0;           /* expect zlib header */
            if (state->head != Z_NULL)
                state->head->done = -1;
            if (!(state->wrap & 1) ||   /* check if zlib header allowed */
@@ -698,6 +700,7 @@ int flush;
                break;
            }
            state->dmax = 1U << len;
+            state->flags = 0;               /* indicate zlib header */
            Tracev((stderr, "inflate:   zlib header ok\n"));
            strm->adler = state->check = adler32(0L, Z_NULL, 0);
            state->mode = hold & 0x200 ? DICTID : TYPE;
@@ -723,6 +726,7 @@ int flush;
                CRC2(state->check, hold);
            INITBITS();
            state->mode = TIME;
+                /* fallthrough */
        case TIME:
            NEEDBITS(32);
            if (state->head != Z_NULL)
@@ -731,6 +735,7 @@ int flush;
                CRC4(state->check, hold);
            INITBITS();
            state->mode = OS;
+                /* fallthrough */
        case OS:
            NEEDBITS(16);
            if (state->head != Z_NULL) {
@@ -741,6 +746,7 @@ int flush;
                CRC2(state->check, hold);
            INITBITS();
            state->mode = EXLEN;
+                /* fallthrough */
        case EXLEN:
            if (state->flags & 0x0400) {
                NEEDBITS(16);
@@ -754,14 +760,16 @@ int flush;
            else if (state->head != Z_NULL)
                state->head->extra = Z_NULL;
            state->mode = EXTRA;
+                /* fallthrough */
        case EXTRA:
            if (state->flags & 0x0400) {
                copy = state->length;
                if (copy > have) copy = have;
                if (copy) {
                    if (state->head != Z_NULL &&
-                        state->head->extra != Z_NULL) {
-                        len = state->head->extra_len - state->length;
+                        state->head->extra != Z_NULL &&
+                        (len = state->head->extra_len - state->length) <
+                            state->head->extra_max) {
                        zmemcpy(state->head->extra + len, next,
                                len + copy > state->head->extra_max ?
                                state->head->extra_max - len : copy);
@@ -776,6 +784,7 @@ int flush;
            }
            state->length = 0;
            state->mode = NAME;
+                /* fallthrough */
        case NAME:
            if (state->flags & 0x0800) {
                if (have == 0) goto inf_leave;
@@ -797,6 +806,7 @@ int flush;
                state->head->name = Z_NULL;
            state->length = 0;
            state->mode = COMMENT;
+                /* fallthrough */
        case COMMENT:
            if (state->flags & 0x1000) {
                if (have == 0) goto inf_leave;
@@ -817,6 +827,7 @@ int flush;
            else if (state->head != Z_NULL)
                state->head->comment = Z_NULL;
            state->mode = HCRC;
+                /* fallthrough */
        case HCRC:
            if (state->flags & 0x0200) {
                NEEDBITS(16);
@@ -840,6 +851,7 @@ int flush;
            strm->adler = state->check = ZSWAP32(hold);
            INITBITS();
            state->mode = DICT;
+                /* fallthrough */
        case DICT:
            if (state->havedict == 0) {
                RESTORE();
@@ -847,8 +859,10 @@ int flush;
            }
            strm->adler = state->check = adler32(0L, Z_NULL, 0);
            state->mode = TYPE;
+                /* fallthrough */
        case TYPE:
            if (flush == Z_BLOCK || flush == Z_TREES) goto inf_leave;
+                /* fallthrough */
        case TYPEDO:
            if (state->last) {
                BYTEBITS();
@@ -899,8 +913,10 @@ int flush;
            INITBITS();
            state->mode = COPY_;
            if (flush == Z_TREES) goto inf_leave;
+                /* fallthrough */
        case COPY_:
            state->mode = COPY;
+                /* fallthrough */
        case COPY:
            copy = state->length;
            if (copy) {
@@ -936,6 +952,7 @@ int flush;
            Tracev((stderr, "inflate:       table sizes ok\n"));
            state->have = 0;
            state->mode = LENLENS;
+                /* fallthrough */
        case LENLENS:
            while (state->have < state->ncode) {
                NEEDBITS(3);
@@ -957,6 +974,7 @@ int flush;
            Tracev((stderr, "inflate:       code lengths ok\n"));
            state->have = 0;
            state->mode = CODELENS;
+                /* fallthrough */
        case CODELENS:
            while (state->have < state->nlen + state->ndist) {
                for (;;) {
@@ -1040,8 +1058,10 @@ int flush;
            Tracev((stderr, "inflate:       codes ok\n"));
            state->mode = LEN_;
            if (flush == Z_TREES) goto inf_leave;
+                /* fallthrough */
        case LEN_:
            state->mode = LEN;
+                /* fallthrough */
        case LEN:
            if (have >= 6 && left >= 258) {
                RESTORE();
@@ -1091,6 +1111,7 @@ int flush;
            }
            state->extra = (unsigned)(here.op) & 15;
            state->mode = LENEXT;
+                /* fallthrough */
        case LENEXT:
            if (state->extra) {
                NEEDBITS(state->extra);
@@ -1101,6 +1122,7 @@ int flush;
            Tracevv((stderr, "inflate:         length %u\n", state->length));
            state->was = state->length;
            state->mode = DIST;
+                /* fallthrough */
        case DIST:
            for (;;) {
                here = state->distcode[BITS(state->distbits)];
@@ -1128,6 +1150,7 @@ int flush;
            state->offset = (unsigned)here.val;
            state->extra = (unsigned)(here.op) & 15;
            state->mode = DISTEXT;
+                /* fallthrough */
        case DISTEXT:
            if (state->extra) {
                NEEDBITS(state->extra);
@@ -1144,6 +1167,7 @@ int flush;
 #endif
            Tracevv((stderr, "inflate:         distance %u\n", state->offset));
            state->mode = MATCH;
+                /* fallthrough */
        case MATCH:
            if (left == 0) goto inf_leave;
            copy = out - left;
@@ -1203,7 +1227,7 @@ int flush;
                state->total += out;
                if ((state->wrap & 4) && out)
                    strm->adler = state->check =
-                        UPDATE(state->check, put - out, out);
+                        UPDATE_CHECK(state->check, put - out, out);
                out = left;
                if ((state->wrap & 4) && (
 #ifdef GUNZIP
@@ -1219,10 +1243,11 @@ int flush;
            }
 #ifdef GUNZIP
            state->mode = LENGTH;
+                /* fallthrough */
        case LENGTH:
            if (state->wrap && state->flags) {
                NEEDBITS(32);
-                if (hold != (state->total & 0xffffffffUL)) {
+                if ((state->wrap & 4) && hold != (state->total & 0xffffffff)) {
                    strm->msg = (char *)"incorrect length check";
                    state->mode = BAD;
                    break;
@@ -1232,6 +1257,7 @@ int flush;
            }
 #endif
            state->mode = DONE;
+                /* fallthrough */
        case DONE:
            ret = Z_STREAM_END;
            goto inf_leave;
@@ -1241,6 +1267,7 @@ int flush;
        case MEM:
            return Z_MEM_ERROR;
        case SYNC:
+                /* fallthrough */
        default:
            return Z_STREAM_ERROR;
        }
@@ -1266,7 +1293,7 @@ int flush;
    state->total += out;
    if ((state->wrap & 4) && out)
        strm->adler = state->check =
-            UPDATE(state->check, strm->next_out - out, out);
+            UPDATE_CHECK(state->check, strm->next_out - out, out);
    strm->data_type = (int)state->bits + (state->last ? 64 : 0) +
                      (state->mode == TYPE ? 128 : 0) +
                      (state->mode == LEN_ || state->mode == COPY_ ? 256 : 0);
@@ -1402,6 +1429,7 @@ int ZEXPORT inflateSync(strm)
 z_streamp strm;
 {
    unsigned len;               /* number of bytes to look at or looked at */
+    int flags;                  /* temporary to save header status */
    unsigned long in, out;      /* temporary to save total_in and total_out */
    unsigned char buf[4];       /* to restore bit buffer to byte string */
    struct inflate_state FAR *state;
@@ -1434,9 +1462,15 @@ z_streamp strm;

    /* return no joy or set up to restart inflate() on a new block */
    if (state->have != 4) return Z_DATA_ERROR;
+    if (state->flags == -1)
+        state->wrap = 0;    /* if no header yet, treat as raw */
+    else
+        state->wrap &= ~4;  /* no point in computing a check value now */
+    flags = state->flags;
    in = strm->total_in;  out = strm->total_out;
    inflateReset(strm);
    strm->total_in = in;  strm->total_out = out;
+    state->flags = flags;
    state->mode = TYPE;
    return Z_OK;
 }
@@ -1532,7 +1566,7 @@ int check;

    if (inflateStateCheck(strm)) return Z_STREAM_ERROR;
    state = (struct inflate_state FAR *)strm->state;
-    if (check)
+    if (check && state->wrap)
        state->wrap |= 4;
    else
        state->wrap &= ~4;
@@ -1,5 +1,5 @@
 /* inflate.h -- internal inflate state definition
- * Copyright (C) 1995-2016 Mark Adler
+ * Copyright (C) 1995-2019 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -86,7 +86,8 @@ struct inflate_state {
    int wrap;                   /* bit 0 true for zlib, bit 1 true for gzip,
                                   bit 2 true to validate check value */
    int havedict;               /* true if dictionary provided */
-    int flags;                  /* gzip header method and flags (0 if zlib) */
+    int flags;                  /* gzip header method and flags, 0 if zlib, or
+                                   -1 if raw or no header yet */
    unsigned dmax;              /* zlib header max distance (INFLATE_STRICT) */
    unsigned long check;        /* protected copy of check value */
    unsigned long total;        /* protected copy of output count */
@@ -1,5 +1,5 @@
 /* inftrees.c -- generate Huffman trees for efficient decoding
- * Copyright (C) 1995-2017 Mark Adler
+ * Copyright (C) 1995-2022 Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -9,7 +9,7 @@
 #define MAXBITS 15

 const char inflate_copyright[] =
-   " inflate 1.2.11 Copyright 1995-2017 Mark Adler ";
+   " inflate 1.2.13 Copyright 1995-2022 Mark Adler ";
 /*
  If you use the zlib library in a product, an acknowledgment is welcome
  in the documentation of your product. If for some reason you cannot
@@ -62,7 +62,7 @@ unsigned short FAR *work;
        35, 43, 51, 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0};
    static const unsigned short lext[31] = { /* Length codes 257..285 extra */
        16, 16, 16, 16, 16, 16, 16, 16, 17, 17, 17, 17, 18, 18, 18, 18,
-        19, 19, 19, 19, 20, 20, 20, 20, 21, 21, 21, 21, 16, 77, 202};
+        19, 19, 19, 19, 20, 20, 20, 20, 21, 21, 21, 21, 16, 194, 65};
    static const unsigned short dbase[32] = { /* Distance codes 0..29 base */
        1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193,
        257, 385, 513, 769, 1025, 1537, 2049, 3073, 4097, 6145,
@@ -38,7 +38,7 @@ typedef struct {
 /* Maximum size of the dynamic table.  The maximum number of code structures is
   1444, which is the sum of 852 for literal/length codes and 592 for distance
   codes.  These values were found by exhaustive searches using the program
-   examples/enough.c found in the zlib distribtution.  The arguments to that
+   examples/enough.c found in the zlib distribution.  The arguments to that
   program are the number of symbols, the initial root table size, and the
   maximum bit length of a code.  "enough 286 9 15" for literal/length codes
   returns returns 852, and "enough 30 6 15" for distance codes returns 592.
@@ -1,5 +1,5 @@
 /* trees.c -- output deflated data using Huffman coding
- * Copyright (C) 1995-2017 Jean-loup Gailly
+ * Copyright (C) 1995-2021 Jean-loup Gailly
 * detect_data_type() function provided freely by Cosmin Truta, 2006
 * For conditions of distribution and use, see copyright notice in zlib.h
 */
@@ -149,7 +149,7 @@ local void send_all_trees OF((deflate_state *s, int lcodes, int dcodes,
 local void compress_block OF((deflate_state *s, const ct_data *ltree,
                              const ct_data *dtree));
 local int  detect_data_type OF((deflate_state *s));
-local unsigned bi_reverse OF((unsigned value, int length));
+local unsigned bi_reverse OF((unsigned code, int len));
 local void bi_windup      OF((deflate_state *s));
 local void bi_flush       OF((deflate_state *s));

@@ -193,7 +193,7 @@ local void send_bits(s, value, length)
    s->bits_sent += (ulg)length;

    /* If not enough room in bi_buf, use (valid) bits from bi_buf and
-     * (16 - bi_valid) bits from value, leaving (width - (16-bi_valid))
+     * (16 - bi_valid) bits from value, leaving (width - (16 - bi_valid))
     * unused bits in value.
     */
    if (s->bi_valid > (int)Buf_size - length) {
@@ -256,7 +256,7 @@ local void tr_static_init()
    length = 0;
    for (code = 0; code < LENGTH_CODES-1; code++) {
        base_length[code] = length;
-        for (n = 0; n < (1<<extra_lbits[code]); n++) {
+        for (n = 0; n < (1 << extra_lbits[code]); n++) {
            _length_code[length++] = (uch)code;
        }
    }
@@ -265,13 +265,13 @@ local void tr_static_init()
     * in two different ways: code 284 + 5 bits or code 285, so we
     * overwrite length_code[255] to use the best encoding:
     */
-    _length_code[length-1] = (uch)code;
+    _length_code[length - 1] = (uch)code;

    /* Initialize the mapping dist (0..32K) -> dist code (0..29) */
    dist = 0;
    for (code = 0 ; code < 16; code++) {
        base_dist[code] = dist;
-        for (n = 0; n < (1<<extra_dbits[code]); n++) {
+        for (n = 0; n < (1 << extra_dbits[code]); n++) {
            _dist_code[dist++] = (uch)code;
        }
    }
@@ -279,11 +279,11 @@ local void tr_static_init()
    dist >>= 7; /* from now on, all distances are divided by 128 */
    for ( ; code < D_CODES; code++) {
        base_dist[code] = dist << 7;
-        for (n = 0; n < (1<<(extra_dbits[code]-7)); n++) {
+        for (n = 0; n < (1 << (extra_dbits[code] - 7)); n++) {
            _dist_code[256 + dist++] = (uch)code;
        }
    }
-    Assert (dist == 256, "tr_static_init: 256+dist != 512");
+    Assert (dist == 256, "tr_static_init: 256 + dist != 512");

    /* Construct the codes of the static literal tree */
    for (bits = 0; bits <= MAX_BITS; bits++) bl_count[bits] = 0;
@@ -312,7 +312,7 @@ local void tr_static_init()
 }

 /* ===========================================================================
- * Genererate the file trees.h describing the static trees.
+ * Generate the file trees.h describing the static trees.
 */
 #ifdef GEN_TREES_H
 #  ifndef ZLIB_DEBUG
@@ -321,7 +321,7 @@ local void tr_static_init()

 #  define SEPARATOR(i, last, width) \
      ((i) == (last)? "\n};\n\n" :    \
-       ((i) % (width) == (width)-1 ? ",\n" : ", "))
+       ((i) % (width) == (width) - 1 ? ",\n" : ", "))

 void gen_trees_header()
 {
@@ -416,7 +416,7 @@ local void init_block(s)

    s->dyn_ltree[END_BLOCK].Freq = 1;
    s->opt_len = s->static_len = 0L;
-    s->last_lit = s->matches = 0;
+    s->sym_next = s->matches = 0;
 }

 #define SMALLEST 1
@@ -458,7 +458,7 @@ local void pqdownheap(s, tree, k)
    while (j <= s->heap_len) {
        /* Set j to the smallest of the two sons: */
        if (j < s->heap_len &&
-            smaller(tree, s->heap[j+1], s->heap[j], s->depth)) {
+            smaller(tree, s->heap[j + 1], s->heap[j], s->depth)) {
            j++;
        }
        /* Exit if v is smaller than both sons */
@@ -507,7 +507,7 @@ local void gen_bitlen(s, desc)
     */
    tree[s->heap[s->heap_max]].Len = 0; /* root of the heap */

-    for (h = s->heap_max+1; h < HEAP_SIZE; h++) {
+    for (h = s->heap_max + 1; h < HEAP_SIZE; h++) {
        n = s->heap[h];
        bits = tree[tree[n].Dad].Len + 1;
        if (bits > max_length) bits = max_length, overflow++;
@@ -518,7 +518,7 @@ local void gen_bitlen(s, desc)

        s->bl_count[bits]++;
        xbits = 0;
-        if (n >= base) xbits = extra[n-base];
+        if (n >= base) xbits = extra[n - base];
        f = tree[n].Freq;
        s->opt_len += (ulg)f * (unsigned)(bits + xbits);
        if (stree) s->static_len += (ulg)f * (unsigned)(stree[n].Len + xbits);
@@ -530,10 +530,10 @@ local void gen_bitlen(s, desc)

    /* Find the first bit length which could increase: */
    do {
-        bits = max_length-1;
+        bits = max_length - 1;
        while (s->bl_count[bits] == 0) bits--;
-        s->bl_count[bits]--;      /* move one leaf down the tree */
-        s->bl_count[bits+1] += 2; /* move one overflow item as its brother */
+        s->bl_count[bits]--;        /* move one leaf down the tree */
+        s->bl_count[bits + 1] += 2; /* move one overflow item as its brother */
        s->bl_count[max_length]--;
        /* The brother of the overflow item also moves one step up,
         * but this does not affect bl_count[max_length]
@@ -569,7 +569,7 @@ local void gen_bitlen(s, desc)
 * OUT assertion: the field code is set for all tree elements of non
 *     zero code length.
 */
-local void gen_codes (tree, max_code, bl_count)
+local void gen_codes(tree, max_code, bl_count)
    ct_data *tree;             /* the tree to decorate */
    int max_code;              /* largest code with non zero frequency */
    ushf *bl_count;            /* number of codes at each bit length */
@@ -583,13 +583,13 @@ local void gen_codes (tree, max_code, bl_count)
     * without bit reversal.
     */
    for (bits = 1; bits <= MAX_BITS; bits++) {
-        code = (code + bl_count[bits-1]) << 1;
+        code = (code + bl_count[bits - 1]) << 1;
        next_code[bits] = (ush)code;
    }
    /* Check that the bit counts in bl_count are consistent. The last code
     * must be all ones.
     */
-    Assert (code + bl_count[MAX_BITS]-1 == (1<<MAX_BITS)-1,
+    Assert (code + bl_count[MAX_BITS] - 1 == (1 << MAX_BITS) - 1,
            "inconsistent bit counts");
    Tracev((stderr,"\ngen_codes: max_code %d ", max_code));

@@ -600,7 +600,7 @@ local void gen_codes (tree, max_code, bl_count)
        tree[n].Code = (ush)bi_reverse(next_code[len]++, len);

        Tracecv(tree != static_ltree, (stderr,"\nn %3d %c l %2d c %4x (%x) ",
-             n, (isgraph(n) ? n : ' '), len, tree[n].Code, next_code[len]-1));
+            n, (isgraph(n) ? n : ' '), len, tree[n].Code, next_code[len] - 1));
    }
 }

@@ -624,7 +624,7 @@ local void build_tree(s, desc)
    int node;          /* new node being created */

    /* Construct the initial heap, with least frequent element in
-     * heap[SMALLEST]. The sons of heap[n] are heap[2*n] and heap[2*n+1].
+     * heap[SMALLEST]. The sons of heap[n] are heap[2*n] and heap[2*n + 1].
     * heap[0] is not used.
     */
    s->heap_len = 0, s->heap_max = HEAP_SIZE;
@@ -652,7 +652,7 @@ local void build_tree(s, desc)
    }
    desc->max_code = max_code;

-    /* The elements heap[heap_len/2+1 .. heap_len] are leaves of the tree,
+    /* The elements heap[heap_len/2 + 1 .. heap_len] are leaves of the tree,
     * establish sub-heaps of increasing lengths:
     */
    for (n = s->heap_len/2; n >= 1; n--) pqdownheap(s, tree, n);
@@ -700,7 +700,7 @@ local void build_tree(s, desc)
 * Scan a literal or distance tree to determine the frequencies of the codes
 * in the bit length tree.
 */
-local void scan_tree (s, tree, max_code)
+local void scan_tree(s, tree, max_code)
    deflate_state *s;
    ct_data *tree;   /* the tree to be scanned */
    int max_code;    /* and its largest code of non zero frequency */
@@ -714,10 +714,10 @@ local void scan_tree (s, tree, max_code)
    int min_count = 4;         /* min repeat count */

    if (nextlen == 0) max_count = 138, min_count = 3;
-    tree[max_code+1].Len = (ush)0xffff; /* guard */
+    tree[max_code + 1].Len = (ush)0xffff; /* guard */

    for (n = 0; n <= max_code; n++) {
-        curlen = nextlen; nextlen = tree[n+1].Len;
+        curlen = nextlen; nextlen = tree[n + 1].Len;
        if (++count < max_count && curlen == nextlen) {
            continue;
        } else if (count < min_count) {
@@ -745,7 +745,7 @@ local void scan_tree (s, tree, max_code)
 * Send a literal or distance tree in compressed form, using the codes in
 * bl_tree.
 */
-local void send_tree (s, tree, max_code)
+local void send_tree(s, tree, max_code)
    deflate_state *s;
    ct_data *tree; /* the tree to be scanned */
    int max_code;       /* and its largest code of non zero frequency */
@@ -758,11 +758,11 @@ local void send_tree (s, tree, max_code)
    int max_count = 7;         /* max repeat count */
    int min_count = 4;         /* min repeat count */

-    /* tree[max_code+1].Len = -1; */  /* guard already set */
+    /* tree[max_code + 1].Len = -1; */  /* guard already set */
    if (nextlen == 0) max_count = 138, min_count = 3;

    for (n = 0; n <= max_code; n++) {
-        curlen = nextlen; nextlen = tree[n+1].Len;
+        curlen = nextlen; nextlen = tree[n + 1].Len;
        if (++count < max_count && curlen == nextlen) {
            continue;
        } else if (count < min_count) {
@@ -773,13 +773,13 @@ local void send_tree (s, tree, max_code)
                send_code(s, curlen, s->bl_tree); count--;
            }
            Assert(count >= 3 && count <= 6, " 3_6?");
-            send_code(s, REP_3_6, s->bl_tree); send_bits(s, count-3, 2);
+            send_code(s, REP_3_6, s->bl_tree); send_bits(s, count - 3, 2);

        } else if (count <= 10) {
-            send_code(s, REPZ_3_10, s->bl_tree); send_bits(s, count-3, 3);
+            send_code(s, REPZ_3_10, s->bl_tree); send_bits(s, count - 3, 3);

        } else {
-            send_code(s, REPZ_11_138, s->bl_tree); send_bits(s, count-11, 7);
+            send_code(s, REPZ_11_138, s->bl_tree); send_bits(s, count - 11, 7);
        }
        count = 0; prevlen = curlen;
        if (nextlen == 0) {
@@ -807,8 +807,8 @@ local int build_bl_tree(s)

    /* Build the bit length tree: */
    build_tree(s, (tree_desc *)(&(s->bl_desc)));
-    /* opt_len now includes the length of the tree representations, except
-     * the lengths of the bit lengths codes and the 5+5+4 bits for the counts.
+    /* opt_len now includes the length of the tree representations, except the
+     * lengths of the bit lengths codes and the 5 + 5 + 4 bits for the counts.
     */

    /* Determine the number of bit length codes to send. The pkzip format
@@ -819,7 +819,7 @@ local int build_bl_tree(s)
        if (s->bl_tree[bl_order[max_blindex]].Len != 0) break;
    }
    /* Update opt_len to include the bit length tree and counts */
-    s->opt_len += 3*((ulg)max_blindex+1) + 5+5+4;
+    s->opt_len += 3*((ulg)max_blindex + 1) + 5 + 5 + 4;
    Tracev((stderr, "\ndyn trees: dyn %ld, stat %ld",
            s->opt_len, s->static_len));

@@ -841,19 +841,19 @@ local void send_all_trees(s, lcodes, dcodes, blcodes)
    Assert (lcodes <= L_CODES && dcodes <= D_CODES && blcodes <= BL_CODES,
            "too many codes");
    Tracev((stderr, "\nbl counts: "));
-    send_bits(s, lcodes-257, 5); /* not +255 as stated in appnote.txt */
-    send_bits(s, dcodes-1,   5);
-    send_bits(s, blcodes-4,  4); /* not -3 as stated in appnote.txt */
+    send_bits(s, lcodes - 257, 5);  /* not +255 as stated in appnote.txt */
+    send_bits(s, dcodes - 1,   5);
+    send_bits(s, blcodes - 4,  4);  /* not -3 as stated in appnote.txt */
    for (rank = 0; rank < blcodes; rank++) {
        Tracev((stderr, "\nbl code %2d ", bl_order[rank]));
        send_bits(s, s->bl_tree[bl_order[rank]].Len, 3);
    }
    Tracev((stderr, "\nbl tree: sent %ld", s->bits_sent));

-    send_tree(s, (ct_data *)s->dyn_ltree, lcodes-1); /* literal tree */
+    send_tree(s, (ct_data *)s->dyn_ltree, lcodes - 1);  /* literal tree */
    Tracev((stderr, "\nlit tree: sent %ld", s->bits_sent));

-    send_tree(s, (ct_data *)s->dyn_dtree, dcodes-1); /* distance tree */
+    send_tree(s, (ct_data *)s->dyn_dtree, dcodes - 1);  /* distance tree */
    Tracev((stderr, "\ndist tree: sent %ld", s->bits_sent));
 }

@@ -866,17 +866,18 @@ void ZLIB_INTERNAL _tr_stored_block(s, buf, stored_len, last)
    ulg stored_len;   /* length of input block */
    int last;         /* one if this is the last block for a file */
 {
-    send_bits(s, (STORED_BLOCK<<1)+last, 3);    /* send block type */
+    send_bits(s, (STORED_BLOCK<<1) + last, 3);  /* send block type */
    bi_windup(s);        /* align on byte boundary */
    put_short(s, (ush)stored_len);
    put_short(s, (ush)~stored_len);
-    zmemcpy(s->pending_buf + s->pending, (Bytef *)buf, stored_len);
+    if (stored_len)
+        zmemcpy(s->pending_buf + s->pending, (Bytef *)buf, stored_len);
    s->pending += stored_len;
 #ifdef ZLIB_DEBUG
    s->compressed_len = (s->compressed_len + 3 + 7) & (ulg)~7L;
    s->compressed_len += (stored_len + 4) << 3;
    s->bits_sent += 2*16;
-    s->bits_sent += stored_len<<3;
+    s->bits_sent += stored_len << 3;
 #endif
 }

@@ -942,14 +943,17 @@ void ZLIB_INTERNAL _tr_flush_block(s, buf, stored_len, last)
        max_blindex = build_bl_tree(s);

        /* Determine the best encoding. Compute the block lengths in bytes. */
-        opt_lenb = (s->opt_len+3+7)>>3;
-        static_lenb = (s->static_len+3+7)>>3;
+        opt_lenb = (s->opt_len + 3 + 7) >> 3;
+        static_lenb = (s->static_len + 3 + 7) >> 3;

        Tracev((stderr, "\nopt %lu(%lu) stat %lu(%lu) stored %lu lit %u ",
                opt_lenb, s->opt_len, static_lenb, s->static_len, stored_len,
-                s->last_lit));
+                s->sym_next / 3));

-        if (static_lenb <= opt_lenb) opt_lenb = static_lenb;
+#ifndef FORCE_STATIC
+        if (static_lenb <= opt_lenb || s->strategy == Z_FIXED)
+#endif
+            opt_lenb = static_lenb;

    } else {
        Assert(buf != (char*)0, "lost buf");
@@ -959,7 +963,7 @@ void ZLIB_INTERNAL _tr_flush_block(s, buf, stored_len, last)
 #ifdef FORCE_STORED
    if (buf != (char*)0) { /* force stored block */
 #else
-    if (stored_len+4 <= opt_lenb && buf != (char*)0) {
+    if (stored_len + 4 <= opt_lenb && buf != (char*)0) {
                       /* 4: two words for the lengths */
 #endif
        /* The test buf != NULL is only necessary if LIT_BUFSIZE > WSIZE.
@@ -970,21 +974,17 @@ void ZLIB_INTERNAL _tr_flush_block(s, buf, stored_len, last)
         */
        _tr_stored_block(s, buf, stored_len, last);

-#ifdef FORCE_STATIC
-    } else if (static_lenb >= 0) { /* force static trees */
-#else
-    } else if (s->strategy == Z_FIXED || static_lenb == opt_lenb) {
-#endif
-        send_bits(s, (STATIC_TREES<<1)+last, 3);
+    } else if (static_lenb == opt_lenb) {
+        send_bits(s, (STATIC_TREES<<1) + last, 3);
        compress_block(s, (const ct_data *)static_ltree,
                       (const ct_data *)static_dtree);
 #ifdef ZLIB_DEBUG
        s->compressed_len += 3 + s->static_len;
 #endif
    } else {
-        send_bits(s, (DYN_TREES<<1)+last, 3);
-        send_all_trees(s, s->l_desc.max_code+1, s->d_desc.max_code+1,
-                       max_blindex+1);
+        send_bits(s, (DYN_TREES<<1) + last, 3);
+        send_all_trees(s, s->l_desc.max_code + 1, s->d_desc.max_code + 1,
+                       max_blindex + 1);
        compress_block(s, (const ct_data *)s->dyn_ltree,
                       (const ct_data *)s->dyn_dtree);
 #ifdef ZLIB_DEBUG
@@ -1003,21 +1003,22 @@ void ZLIB_INTERNAL _tr_flush_block(s, buf, stored_len, last)
        s->compressed_len += 7;  /* align on byte boundary */
 #endif
    }
-    Tracev((stderr,"\ncomprlen %lu(%lu) ", s->compressed_len>>3,
-           s->compressed_len-7*last));
+    Tracev((stderr,"\ncomprlen %lu(%lu) ", s->compressed_len >> 3,
+           s->compressed_len - 7*last));
 }

 /* ===========================================================================
 * Save the match info and tally the frequency counts. Return true if
 * the current block must be flushed.
 */
-int ZLIB_INTERNAL _tr_tally (s, dist, lc)
+int ZLIB_INTERNAL _tr_tally(s, dist, lc)
    deflate_state *s;
    unsigned dist;  /* distance of matched string */
-    unsigned lc;    /* match length-MIN_MATCH or unmatched char (if dist==0) */
+    unsigned lc;    /* match length - MIN_MATCH or unmatched char (dist==0) */
 {
-    s->d_buf[s->last_lit] = (ush)dist;
-    s->l_buf[s->last_lit++] = (uch)lc;
+    s->sym_buf[s->sym_next++] = (uch)dist;
+    s->sym_buf[s->sym_next++] = (uch)(dist >> 8);
+    s->sym_buf[s->sym_next++] = (uch)lc;
    if (dist == 0) {
        /* lc is the unmatched char */
        s->dyn_ltree[lc].Freq++;
@@ -1029,33 +1030,10 @@ int ZLIB_INTERNAL _tr_tally (s, dist, lc)
               (ush)lc <= (ush)(MAX_MATCH-MIN_MATCH) &&
               (ush)d_code(dist) < (ush)D_CODES,  "_tr_tally: bad match");

-        s->dyn_ltree[_length_code[lc]+LITERALS+1].Freq++;
+        s->dyn_ltree[_length_code[lc] + LITERALS + 1].Freq++;
        s->dyn_dtree[d_code(dist)].Freq++;
    }
-
-#ifdef TRUNCATE_BLOCK
-    /* Try to guess if it is profitable to stop the current block here */
-    if ((s->last_lit & 0x1fff) == 0 && s->level > 2) {
-        /* Compute an upper bound for the compressed length */
-        ulg out_length = (ulg)s->last_lit*8L;
-        ulg in_length = (ulg)((long)s->strstart - s->block_start);
-        int dcode;
-        for (dcode = 0; dcode < D_CODES; dcode++) {
-            out_length += (ulg)s->dyn_dtree[dcode].Freq *
-                (5L+extra_dbits[dcode]);
-        }
-        out_length >>= 3;
-        Tracev((stderr,"\nlast_lit %u, in %ld, out ~%ld(%ld%%) ",
-               s->last_lit, in_length, out_length,
-               100L - out_length*100L/in_length));
-        if (s->matches < s->last_lit/2 && out_length < in_length/2) return 1;
-    }
-#endif
-    return (s->last_lit == s->lit_bufsize-1);
-    /* We avoid equality with lit_bufsize because of wraparound at 64K
-     * on 16 bit machines and because stored blocks are restricted to
-     * 64K-1 bytes.
-     */
+    return (s->sym_next == s->sym_end);
 }

 /* ===========================================================================
@@ -1068,20 +1046,21 @@ local void compress_block(s, ltree, dtree)
 {
    unsigned dist;      /* distance of matched string */
    int lc;             /* match length or unmatched char (if dist == 0) */
-    unsigned lx = 0;    /* running index in l_buf */
+    unsigned sx = 0;    /* running index in sym_buf */
    unsigned code;      /* the code to send */
    int extra;          /* number of extra bits to send */

-    if (s->last_lit != 0) do {
-        dist = s->d_buf[lx];
-        lc = s->l_buf[lx++];
+    if (s->sym_next != 0) do {
+        dist = s->sym_buf[sx++] & 0xff;
+        dist += (unsigned)(s->sym_buf[sx++] & 0xff) << 8;
+        lc = s->sym_buf[sx++];
        if (dist == 0) {
            send_code(s, lc, ltree); /* send a literal byte */
            Tracecv(isgraph(lc), (stderr," '%c' ", lc));
        } else {
            /* Here, lc is the match length - MIN_MATCH */
            code = _length_code[lc];
-            send_code(s, code+LITERALS+1, ltree); /* send the length code */
+            send_code(s, code + LITERALS + 1, ltree);   /* send length code */
            extra = extra_lbits[code];
            if (extra != 0) {
                lc -= base_length[code];
@@ -1099,11 +1078,10 @@ local void compress_block(s, ltree, dtree)
            }
        } /* literal or match pair ? */

-        /* Check that the overlay between pending_buf and d_buf+l_buf is ok: */
-        Assert((uInt)(s->pending) < s->lit_bufsize + 2*lx,
-               "pendingBuf overflow");
+        /* Check that the overlay between pending_buf and sym_buf is ok: */
+        Assert(s->pending < s->lit_bufsize + sx, "pendingBuf overflow");

-    } while (lx < s->last_lit);
+    } while (sx < s->sym_next);

    send_code(s, END_BLOCK, ltree);
 }
@@ -1112,9 +1090,9 @@ local void compress_block(s, ltree, dtree)
 * Check if the data type is TEXT or BINARY, using the following algorithm:
 * - TEXT if the two conditions below are satisfied:
 *    a) There are no non-portable control characters belonging to the
- *       "black list" (0..6, 14..25, 28..31).
+ *       "block list" (0..6, 14..25, 28..31).
 *    b) There is at least one printable character belonging to the
- *       "white list" (9 {TAB}, 10 {LF}, 13 {CR}, 32..255).
+ *       "allow list" (9 {TAB}, 10 {LF}, 13 {CR}, 32..255).
 * - BINARY otherwise.
 * - The following partially-portable control characters form a
 *   "gray list" that is ignored in this detection algorithm:
@@ -1124,19 +1102,19 @@ local void compress_block(s, ltree, dtree)
 local int detect_data_type(s)
    deflate_state *s;
 {
-    /* black_mask is the bit mask of black-listed bytes
+    /* block_mask is the bit mask of block-listed bytes
     * set bits 0..6, 14..25, and 28..31
     * 0xf3ffc07f = binary 11110011111111111100000001111111
     */
-    unsigned long black_mask = 0xf3ffc07fUL;
+    unsigned long block_mask = 0xf3ffc07fUL;
    int n;

-    /* Check for non-textual ("black-listed") bytes. */
-    for (n = 0; n <= 31; n++, black_mask >>= 1)
-        if ((black_mask & 1) && (s->dyn_ltree[n].Freq != 0))
+    /* Check for non-textual ("block-listed") bytes. */
+    for (n = 0; n <= 31; n++, block_mask >>= 1)
+        if ((block_mask & 1) && (s->dyn_ltree[n].Freq != 0))
            return Z_BINARY;

-    /* Check for textual ("white-listed") bytes. */
+    /* Check for textual ("allow-listed") bytes. */
    if (s->dyn_ltree[9].Freq != 0 || s->dyn_ltree[10].Freq != 0
            || s->dyn_ltree[13].Freq != 0)
        return Z_TEXT;
@@ -1144,7 +1122,7 @@ local int detect_data_type(s)
        if (s->dyn_ltree[n].Freq != 0)
            return Z_TEXT;

-    /* There are no "black-listed" or "white-listed" bytes:
+    /* There are no "block-listed" or "allow-listed" bytes:
     * this stream either is empty or has tolerated ("gray-listed") bytes only.
     */
    return Z_BINARY;
@@ -1198,6 +1176,6 @@ local void bi_windup(s)
    s->bi_buf = 0;
    s->bi_valid = 0;
 #ifdef ZLIB_DEBUG
-    s->bits_sent = (s->bits_sent+7) & ~7;
+    s->bits_sent = (s->bits_sent + 7) & ~7;
 #endif
 }
@@ -24,7 +24,7 @@
   Z_DATA_ERROR if the input data was corrupted, including if the input data is
   an incomplete zlib stream.
 */
-int ZEXPORT uncompress2 (dest, destLen, source, sourceLen)
+int ZEXPORT uncompress2(dest, destLen, source, sourceLen)
    Bytef *dest;
    uLongf *destLen;
    const Bytef *source;
@@ -83,7 +83,7 @@ int ZEXPORT uncompress2 (dest, destLen, source, sourceLen)
           err;
 }

-int ZEXPORT uncompress (dest, destLen, source, sourceLen)
+int ZEXPORT uncompress(dest, destLen, source, sourceLen)
    Bytef *dest;
    uLongf *destLen;
    const Bytef *source;
@@ -38,6 +38,9 @@
 #  define crc32                 z_crc32
 #  define crc32_combine         z_crc32_combine
 #  define crc32_combine64       z_crc32_combine64
+#  define crc32_combine_gen     z_crc32_combine_gen
+#  define crc32_combine_gen64   z_crc32_combine_gen64
+#  define crc32_combine_op      z_crc32_combine_op
 #  define crc32_z               z_crc32_z
 #  define deflate               z_deflate
 #  define deflateBound          z_deflateBound
@@ -349,6 +352,9 @@
 #    ifdef FAR
 #      undef FAR
 #    endif
+#    ifndef WIN32_LEAN_AND_MEAN
+#      define WIN32_LEAN_AND_MEAN
+#    endif
 #    include <windows.h>
     /* No need for _export, use ZLIB.DEF instead. */
     /* For complete Windows compatibility, use WINAPI, not __stdcall. */
@@ -467,11 +473,18 @@ typedef uLong FAR uLongf;
 #  undef _LARGEFILE64_SOURCE
 #endif

-#if defined(__WATCOMC__) && !defined(Z_HAVE_UNISTD_H)
-#  define Z_HAVE_UNISTD_H
+#ifndef Z_HAVE_UNISTD_H
+#  ifdef __WATCOMC__
+#    define Z_HAVE_UNISTD_H
+#  endif
+#endif
+#ifndef Z_HAVE_UNISTD_H
+#  if defined(_LARGEFILE64_SOURCE) && !defined(_WIN32)
+#    define Z_HAVE_UNISTD_H
+#  endif
 #endif
 #ifndef Z_SOLO
-#  if defined(Z_HAVE_UNISTD_H) || defined(_LARGEFILE64_SOURCE)
+#  if defined(Z_HAVE_UNISTD_H)
 #    include <unistd.h>         /* for SEEK_*, off_t, and _LFS64_LARGEFILE */
 #    ifdef VMS
 #      include <unixio.h>       /* for off_t */
@@ -1,7 +1,7 @@
 /* zlib.h -- interface of the 'zlib' general purpose compression library
-  version 1.2.11, January 15th, 2017
+  version 1.2.13, October 13th, 2022

-  Copyright (C) 1995-2017 Jean-loup Gailly and Mark Adler
+  Copyright (C) 1995-2022 Jean-loup Gailly and Mark Adler

  This software is provided 'as-is', without any express or implied
  warranty.  In no event will the authors be held liable for any damages
@@ -37,11 +37,11 @@
 extern "C" {
 #endif

-#define ZLIB_VERSION "1.2.11"
-#define ZLIB_VERNUM 0x12b0
+#define ZLIB_VERSION "1.2.13"
+#define ZLIB_VERNUM 0x12d0
 #define ZLIB_VER_MAJOR 1
 #define ZLIB_VER_MINOR 2
-#define ZLIB_VER_REVISION 11
+#define ZLIB_VER_REVISION 13
 #define ZLIB_VER_SUBREVISION 0

 /*
@@ -276,7 +276,7 @@ ZEXTERN int ZEXPORT deflate OF((z_streamp strm, int flush));
  == 0), or after each call of deflate().  If deflate returns Z_OK and with
  zero avail_out, it must be called again after making room in the output
  buffer because there might be more output pending. See deflatePending(),
-  which can be used if desired to determine whether or not there is more ouput
+  which can be used if desired to determine whether or not there is more output
  in that case.

    Normally the parameter flush is set to Z_NO_FLUSH, which allows deflate to
@@ -543,8 +543,7 @@ ZEXTERN int ZEXPORT deflateInit2 OF((z_streamp strm,
                                     int  strategy));

     This is another version of deflateInit with more compression options.  The
-   fields next_in, zalloc, zfree and opaque must be initialized before by the
-   caller.
+   fields zalloc, zfree and opaque must be initialized before by the caller.

     The method parameter is the compression method.  It must be Z_DEFLATED in
   this version of the library.
@@ -661,7 +660,7 @@ ZEXTERN int ZEXPORT deflateGetDictionary OF((z_streamp strm,
   to dictionary.  dictionary must have enough space, where 32768 bytes is
   always enough.  If deflateGetDictionary() is called with dictionary equal to
   Z_NULL, then only the dictionary length is returned, and nothing is copied.
-   Similary, if dictLength is Z_NULL, then it is not set.
+   Similarly, if dictLength is Z_NULL, then it is not set.

     deflateGetDictionary() may return a length less than the window size, even
   when more than the window size in input has been provided. It may return up
@@ -712,11 +711,12 @@ ZEXTERN int ZEXPORT deflateParams OF((z_streamp strm,
   used to switch between compression and straight copy of the input data, or
   to switch to a different kind of input data requiring a different strategy.
   If the compression approach (which is a function of the level) or the
-   strategy is changed, and if any input has been consumed in a previous
-   deflate() call, then the input available so far is compressed with the old
-   level and strategy using deflate(strm, Z_BLOCK).  There are three approaches
-   for the compression levels 0, 1..3, and 4..9 respectively.  The new level
-   and strategy will take effect at the next call of deflate().
+   strategy is changed, and if there have been any deflate() calls since the
+   state was initialized or reset, then the input available so far is
+   compressed with the old level and strategy using deflate(strm, Z_BLOCK).
+   There are three approaches for the compression levels 0, 1..3, and 4..9
+   respectively.  The new level and strategy will take effect at the next call
+   of deflate().

     If a deflate(strm, Z_BLOCK) is performed by deflateParams(), and it does
   not have enough output space to complete, then the parameter change will not
@@ -865,9 +865,11 @@ ZEXTERN int ZEXPORT inflateInit2 OF((z_streamp strm,
   detection, or add 16 to decode only the gzip format (the zlib format will
   return a Z_DATA_ERROR).  If a gzip stream is being decoded, strm->adler is a
   CRC-32 instead of an Adler-32.  Unlike the gunzip utility and gzread() (see
-   below), inflate() will not automatically decode concatenated gzip streams.
-   inflate() will return Z_STREAM_END at the end of the gzip stream.  The state
-   would need to be reset to continue decoding a subsequent gzip stream.
+   below), inflate() will *not* automatically decode concatenated gzip members.
+   inflate() will return Z_STREAM_END at the end of the gzip member.  The state
+   would need to be reset to continue decoding a subsequent gzip member.  This
+   *must* be done if there is more data after a gzip member, in order for the
+   decompression to be compliant with the gzip standard (RFC 1952).

     inflateInit2 returns Z_OK if success, Z_MEM_ERROR if there was not enough
   memory, Z_VERSION_ERROR if the zlib library version is incompatible with the
@@ -913,7 +915,7 @@ ZEXTERN int ZEXPORT inflateGetDictionary OF((z_streamp strm,
   to dictionary.  dictionary must have enough space, where 32768 bytes is
   always enough.  If inflateGetDictionary() is called with dictionary equal to
   Z_NULL, then only the dictionary length is returned, and nothing is copied.
-   Similary, if dictLength is Z_NULL, then it is not set.
+   Similarly, if dictLength is Z_NULL, then it is not set.

     inflateGetDictionary returns Z_OK on success, or Z_STREAM_ERROR if the
   stream state is inconsistent.
@@ -1302,14 +1304,14 @@ typedef struct gzFile_s *gzFile;    /* semi-opaque gzip file descriptor */
 /*
 ZEXTERN gzFile ZEXPORT gzopen OF((const char *path, const char *mode));

-     Opens a gzip (.gz) file for reading or writing.  The mode parameter is as
-   in fopen ("rb" or "wb") but can also include a compression level ("wb9") or
-   a strategy: 'f' for filtered data as in "wb6f", 'h' for Huffman-only
-   compression as in "wb1h", 'R' for run-length encoding as in "wb1R", or 'F'
-   for fixed code compression as in "wb9F".  (See the description of
-   deflateInit2 for more information about the strategy parameter.)  'T' will
-   request transparent writing or appending with no compression and not using
-   the gzip format.
+     Open the gzip (.gz) file at path for reading and decompressing, or
+   compressing and writing.  The mode parameter is as in fopen ("rb" or "wb")
+   but can also include a compression level ("wb9") or a strategy: 'f' for
+   filtered data as in "wb6f", 'h' for Huffman-only compression as in "wb1h",
+   'R' for run-length encoding as in "wb1R", or 'F' for fixed code compression
+   as in "wb9F".  (See the description of deflateInit2 for more information
+   about the strategy parameter.)  'T' will request transparent writing or
+   appending with no compression and not using the gzip format.

     "a" can be used instead of "w" to request that the gzip stream that will
   be written be appended to the file.  "+" will result in an error, since
@@ -1339,9 +1341,9 @@ ZEXTERN gzFile ZEXPORT gzopen OF((const char *path, const char *mode));

 ZEXTERN gzFile ZEXPORT gzdopen OF((int fd, const char *mode));
 /*
-     gzdopen associates a gzFile with the file descriptor fd.  File descriptors
-   are obtained from calls like open, dup, creat, pipe or fileno (if the file
-   has been previously opened with fopen).  The mode parameter is as in gzopen.
+     Associate a gzFile with the file descriptor fd.  File descriptors are
+   obtained from calls like open, dup, creat, pipe or fileno (if the file has
+   been previously opened with fopen).  The mode parameter is as in gzopen.

     The next call of gzclose on the returned gzFile will also close the file
   descriptor fd, just like fclose(fdopen(fd, mode)) closes the file descriptor
@@ -1362,13 +1364,13 @@ ZEXTERN gzFile ZEXPORT gzdopen OF((int fd, const char *mode));

 ZEXTERN int ZEXPORT gzbuffer OF((gzFile file, unsigned size));
 /*
-     Set the internal buffer size used by this library's functions.  The
-   default buffer size is 8192 bytes.  This function must be called after
-   gzopen() or gzdopen(), and before any other calls that read or write the
-   file.  The buffer memory allocation is always deferred to the first read or
-   write.  Three times that size in buffer space is allocated.  A larger buffer
-   size of, for example, 64K or 128K bytes will noticeably increase the speed
-   of decompression (reading).
+     Set the internal buffer size used by this library's functions for file to
+   size.  The default buffer size is 8192 bytes.  This function must be called
+   after gzopen() or gzdopen(), and before any other calls that read or write
+   the file.  The buffer memory allocation is always deferred to the first read
+   or write.  Three times that size in buffer space is allocated.  A larger
+   buffer size of, for example, 64K or 128K bytes will noticeably increase the
+   speed of decompression (reading).

     The new buffer size also affects the maximum length for gzprintf().

@@ -1378,9 +1380,9 @@ ZEXTERN int ZEXPORT gzbuffer OF((gzFile file, unsigned size));

 ZEXTERN int ZEXPORT gzsetparams OF((gzFile file, int level, int strategy));
 /*
-     Dynamically update the compression level or strategy.  See the description
-   of deflateInit2 for the meaning of these parameters.  Previously provided
-   data is flushed before the parameter change.
+     Dynamically update the compression level and strategy for file.  See the
+   description of deflateInit2 for the meaning of these parameters. Previously
+   provided data is flushed before applying the parameter changes.

     gzsetparams returns Z_OK if success, Z_STREAM_ERROR if the file was not
   opened for writing, Z_ERRNO if there is an error writing the flushed data,
@@ -1389,7 +1391,7 @@ ZEXTERN int ZEXPORT gzsetparams OF((gzFile file, int level, int strategy));

 ZEXTERN int ZEXPORT gzread OF((gzFile file, voidp buf, unsigned len));
 /*
-     Reads the given number of uncompressed bytes from the compressed file.  If
+     Read and decompress up to len uncompressed bytes from file into buf.  If
   the input file is not in gzip format, gzread copies the given number of
   bytes into the buffer directly from the file.

@@ -1420,11 +1422,11 @@ ZEXTERN int ZEXPORT gzread OF((gzFile file, voidp buf, unsigned len));
 ZEXTERN z_size_t ZEXPORT gzfread OF((voidp buf, z_size_t size, z_size_t nitems,
                                     gzFile file));
 /*
-     Read up to nitems items of size size from file to buf, otherwise operating
-   as gzread() does.  This duplicates the interface of stdio's fread(), with
-   size_t request and return types.  If the library defines size_t, then
-   z_size_t is identical to size_t.  If not, then z_size_t is an unsigned
-   integer type that can contain a pointer.
+     Read and decompress up to nitems items of size size from file into buf,
+   otherwise operating as gzread() does.  This duplicates the interface of
+   stdio's fread(), with size_t request and return types.  If the library
+   defines size_t, then z_size_t is identical to size_t.  If not, then z_size_t
+   is an unsigned integer type that can contain a pointer.

     gzfread() returns the number of full items read of size size, or zero if
   the end of the file was reached and a full item could not be read, or if
@@ -1435,26 +1437,24 @@ ZEXTERN z_size_t ZEXPORT gzfread OF((voidp buf, z_size_t size, z_size_t nitems,

     In the event that the end of file is reached and only a partial item is
   available at the end, i.e. the remaining uncompressed data length is not a
-   multiple of size, then the final partial item is nevetheless read into buf
+   multiple of size, then the final partial item is nevertheless read into buf
   and the end-of-file flag is set.  The length of the partial item read is not
   provided, but could be inferred from the result of gztell().  This behavior
   is the same as the behavior of fread() implementations in common libraries,
   but it prevents the direct use of gzfread() to read a concurrently written
-   file, reseting and retrying on end-of-file, when size is not 1.
+   file, resetting and retrying on end-of-file, when size is not 1.
 */

-ZEXTERN int ZEXPORT gzwrite OF((gzFile file,
-                                voidpc buf, unsigned len));
+ZEXTERN int ZEXPORT gzwrite OF((gzFile file, voidpc buf, unsigned len));
 /*
-     Writes the given number of uncompressed bytes into the compressed file.
-   gzwrite returns the number of uncompressed bytes written or 0 in case of
-   error.
+     Compress and write the len uncompressed bytes at buf to file. gzwrite
+   returns the number of uncompressed bytes written or 0 in case of error.
 */

 ZEXTERN z_size_t ZEXPORT gzfwrite OF((voidpc buf, z_size_t size,
                                      z_size_t nitems, gzFile file));
 /*
-     gzfwrite() writes nitems items of size size from buf to file, duplicating
+     Compress and write nitems items of size size from buf to file, duplicating
   the interface of stdio's fwrite(), with size_t request and return types.  If
   the library defines size_t, then z_size_t is identical to size_t.  If not,
   then z_size_t is an unsigned integer type that can contain a pointer.
@@ -1467,22 +1467,22 @@ ZEXTERN z_size_t ZEXPORT gzfwrite OF((voidpc buf, z_size_t size,

 ZEXTERN int ZEXPORTVA gzprintf Z_ARG((gzFile file, const char *format, ...));
 /*
-     Converts, formats, and writes the arguments to the compressed file under
-   control of the format string, as in fprintf.  gzprintf returns the number of
+     Convert, format, compress, and write the arguments (...) to file under
+   control of the string format, as in fprintf.  gzprintf returns the number of
   uncompressed bytes actually written, or a negative zlib error code in case
   of error.  The number of uncompressed bytes written is limited to 8191, or
   one less than the buffer size given to gzbuffer().  The caller should assure
   that this limit is not exceeded.  If it is exceeded, then gzprintf() will
   return an error (0) with nothing written.  In this case, there may also be a
   buffer overflow with unpredictable consequences, which is possible only if
-   zlib was compiled with the insecure functions sprintf() or vsprintf()
+   zlib was compiled with the insecure functions sprintf() or vsprintf(),
   because the secure snprintf() or vsnprintf() functions were not available.
   This can be determined using zlibCompileFlags().
 */

 ZEXTERN int ZEXPORT gzputs OF((gzFile file, const char *s));
 /*
-     Writes the given null-terminated string to the compressed file, excluding
+     Compress and write the given null-terminated string s to file, excluding
   the terminating null character.

     gzputs returns the number of characters written, or -1 in case of error.
@@ -1490,11 +1490,12 @@ ZEXTERN int ZEXPORT gzputs OF((gzFile file, const char *s));

 ZEXTERN char * ZEXPORT gzgets OF((gzFile file, char *buf, int len));
 /*
-     Reads bytes from the compressed file until len-1 characters are read, or a
-   newline character is read and transferred to buf, or an end-of-file
-   condition is encountered.  If any characters are read or if len == 1, the
-   string is terminated with a null character.  If no characters are read due
-   to an end-of-file or len < 1, then the buffer is left untouched.
+     Read and decompress bytes from file into buf, until len-1 characters are
+   read, or until a newline character is read and transferred to buf, or an
+   end-of-file condition is encountered.  If any characters are read or if len
+   is one, the string is terminated with a null character.  If no characters
+   are read due to an end-of-file or len is less than one, then the buffer is
+   left untouched.

     gzgets returns buf which is a null-terminated string, or it returns NULL
   for end-of-file or in case of error.  If there was an error, the contents at
@@ -1503,13 +1504,13 @@ ZEXTERN char * ZEXPORT gzgets OF((gzFile file, char *buf, int len));

 ZEXTERN int ZEXPORT gzputc OF((gzFile file, int c));
 /*
-     Writes c, converted to an unsigned char, into the compressed file.  gzputc
+     Compress and write c, converted to an unsigned char, into file.  gzputc
   returns the value that was written, or -1 in case of error.
 */

 ZEXTERN int ZEXPORT gzgetc OF((gzFile file));
 /*
-     Reads one byte from the compressed file.  gzgetc returns this byte or -1
+     Read and decompress one byte from file.  gzgetc returns this byte or -1
   in case of end of file or error.  This is implemented as a macro for speed.
   As such, it does not do all of the checking the other functions do.  I.e.
   it does not check to see if file is NULL, nor whether the structure file
@@ -1518,8 +1519,8 @@ ZEXTERN int ZEXPORT gzgetc OF((gzFile file));

 ZEXTERN int ZEXPORT gzungetc OF((int c, gzFile file));
 /*
-     Push one character back onto the stream to be read as the first character
-   on the next read.  At least one character of push-back is allowed.
+     Push c back onto the stream for file to be read as the first character on
+   the next read.  At least one character of push-back is always allowed.
   gzungetc() returns the character pushed, or -1 on failure.  gzungetc() will
   fail if c is -1, and may fail if a character has been pushed but not read
   yet.  If gzungetc is used immediately after gzopen or gzdopen, at least the
@@ -1530,9 +1531,9 @@ ZEXTERN int ZEXPORT gzungetc OF((int c, gzFile file));

 ZEXTERN int ZEXPORT gzflush OF((gzFile file, int flush));
 /*
-     Flushes all pending output into the compressed file.  The parameter flush
-   is as in the deflate() function.  The return value is the zlib error number
-   (see function gzerror below).  gzflush is only permitted when writing.
+     Flush all pending output to file.  The parameter flush is as in the
+   deflate() function.  The return value is the zlib error number (see function
+   gzerror below).  gzflush is only permitted when writing.

     If the flush parameter is Z_FINISH, the remaining data is written and the
   gzip stream is completed in the output.  If gzwrite() is called again, a new
@@ -1547,8 +1548,8 @@ ZEXTERN int ZEXPORT gzflush OF((gzFile file, int flush));
 ZEXTERN z_off_t ZEXPORT gzseek OF((gzFile file,
                                   z_off_t offset, int whence));

-     Sets the starting position for the next gzread or gzwrite on the given
-   compressed file.  The offset represents a number of bytes in the
+     Set the starting position to offset relative to whence for the next gzread
+   or gzwrite on file.  The offset represents a number of bytes in the
   uncompressed data stream.  The whence parameter is defined as in lseek(2);
   the value SEEK_END is not supported.

@@ -1565,18 +1566,18 @@ ZEXTERN z_off_t ZEXPORT gzseek OF((gzFile file,

 ZEXTERN int ZEXPORT    gzrewind OF((gzFile file));
 /*
-     Rewinds the given file. This function is supported only for reading.
+     Rewind file. This function is supported only for reading.

-     gzrewind(file) is equivalent to (int)gzseek(file, 0L, SEEK_SET)
+     gzrewind(file) is equivalent to (int)gzseek(file, 0L, SEEK_SET).
 */

 /*
 ZEXTERN z_off_t ZEXPORT    gztell OF((gzFile file));

-     Returns the starting position for the next gzread or gzwrite on the given
-   compressed file.  This position represents a number of bytes in the
-   uncompressed data stream, and is zero when starting, even if appending or
-   reading a gzip stream from the middle of a file using gzdopen().
+     Return the starting position for the next gzread or gzwrite on file.
+   This position represents a number of bytes in the uncompressed data stream,
+   and is zero when starting, even if appending or reading a gzip stream from
+   the middle of a file using gzdopen().

     gztell(file) is equivalent to gzseek(file, 0L, SEEK_CUR)
 */
@@ -1584,22 +1585,22 @@ ZEXTERN z_off_t ZEXPORT    gztell OF((gzFile file));
 /*
 ZEXTERN z_off_t ZEXPORT gzoffset OF((gzFile file));

-     Returns the current offset in the file being read or written.  This offset
-   includes the count of bytes that precede the gzip stream, for example when
-   appending or when using gzdopen() for reading.  When reading, the offset
-   does not include as yet unused buffered input.  This information can be used
-   for a progress indicator.  On error, gzoffset() returns -1.
+     Return the current compressed (actual) read or write offset of file.  This
+   offset includes the count of bytes that precede the gzip stream, for example
+   when appending or when using gzdopen() for reading.  When reading, the
+   offset does not include as yet unused buffered input.  This information can
+   be used for a progress indicator.  On error, gzoffset() returns -1.
 */

 ZEXTERN int ZEXPORT gzeof OF((gzFile file));
 /*
-     Returns true (1) if the end-of-file indicator has been set while reading,
-   false (0) otherwise.  Note that the end-of-file indicator is set only if the
-   read tried to go past the end of the input, but came up short.  Therefore,
-   just like feof(), gzeof() may return false even if there is no more data to
-   read, in the event that the last read request was for the exact number of
-   bytes remaining in the input file.  This will happen if the input file size
-   is an exact multiple of the buffer size.
+     Return true (1) if the end-of-file indicator for file has been set while
+   reading, false (0) otherwise.  Note that the end-of-file indicator is set
+   only if the read tried to go past the end of the input, but came up short.
+   Therefore, just like feof(), gzeof() may return false even if there is no
+   more data to read, in the event that the last read request was for the exact
+   number of bytes remaining in the input file.  This will happen if the input
+   file size is an exact multiple of the buffer size.

     If gzeof() returns true, then the read functions will return no more data,
   unless the end-of-file indicator is reset by gzclearerr() and the input file
@@ -1608,7 +1609,7 @@ ZEXTERN int ZEXPORT gzeof OF((gzFile file));

 ZEXTERN int ZEXPORT gzdirect OF((gzFile file));
 /*
-     Returns true (1) if file is being copied directly while reading, or false
+     Return true (1) if file is being copied directly while reading, or false
   (0) if file is a gzip stream being decompressed.

     If the input file is empty, gzdirect() will return true, since the input
@@ -1629,8 +1630,8 @@ ZEXTERN int ZEXPORT gzdirect OF((gzFile file));

 ZEXTERN int ZEXPORT    gzclose OF((gzFile file));
 /*
-     Flushes all pending output if necessary, closes the compressed file and
-   deallocates the (de)compression state.  Note that once file is closed, you
+     Flush all pending output for file, if necessary, close file and
+   deallocate the (de)compression state.  Note that once file is closed, you
   cannot call gzerror with file, since its structures have been deallocated.
   gzclose must not be called more than once on the same file, just as free
   must not be called more than once on the same allocation.
@@ -1654,10 +1655,10 @@ ZEXTERN int ZEXPORT gzclose_w OF((gzFile file));

 ZEXTERN const char * ZEXPORT gzerror OF((gzFile file, int *errnum));
 /*
-     Returns the error message for the last error which occurred on the given
-   compressed file.  errnum is set to zlib error number.  If an error occurred
-   in the file system and not in the compression library, errnum is set to
-   Z_ERRNO and the application may consult errno to get the exact error code.
+     Return the error message for the last error which occurred on file.
+   errnum is set to zlib error number.  If an error occurred in the file system
+   and not in the compression library, errnum is set to Z_ERRNO and the
+   application may consult errno to get the exact error code.

     The application must not modify the returned string.  Future calls to
   this function may invalidate the previously returned string.  If file is
@@ -1670,7 +1671,7 @@ ZEXTERN const char * ZEXPORT gzerror OF((gzFile file, int *errnum));

 ZEXTERN void ZEXPORT gzclearerr OF((gzFile file));
 /*
-     Clears the error and end-of-file flags for file.  This is analogous to the
+     Clear the error and end-of-file flags for file.  This is analogous to the
   clearerr() function in stdio.  This is useful for continuing to read a gzip
   file that is being written concurrently.
 */
@@ -1688,8 +1689,9 @@ ZEXTERN void ZEXPORT gzclearerr OF((gzFile file));
 ZEXTERN uLong ZEXPORT adler32 OF((uLong adler, const Bytef *buf, uInt len));
 /*
     Update a running Adler-32 checksum with the bytes buf[0..len-1] and
-   return the updated checksum.  If buf is Z_NULL, this function returns the
-   required initial value for the checksum.
+   return the updated checksum. An Adler-32 value is in the range of a 32-bit
+   unsigned integer. If buf is Z_NULL, this function returns the required
+   initial value for the checksum.

     An Adler-32 checksum is almost as reliable as a CRC-32 but can be computed
   much faster.
@@ -1722,12 +1724,13 @@ ZEXTERN uLong ZEXPORT adler32_combine OF((uLong adler1, uLong adler2,
   negative, the result has no meaning or utility.
 */

-ZEXTERN uLong ZEXPORT crc32   OF((uLong crc, const Bytef *buf, uInt len));
+ZEXTERN uLong ZEXPORT crc32 OF((uLong crc, const Bytef *buf, uInt len));
 /*
     Update a running CRC-32 with the bytes buf[0..len-1] and return the
-   updated CRC-32.  If buf is Z_NULL, this function returns the required
-   initial value for the crc.  Pre- and post-conditioning (one's complement) is
-   performed within this function so it shouldn't be done by the application.
+   updated CRC-32. A CRC-32 value is in the range of a 32-bit unsigned integer.
+   If buf is Z_NULL, this function returns the required initial value for the
+   crc. Pre- and post-conditioning (one's complement) is performed within this
+   function so it shouldn't be done by the application.

   Usage example:

@@ -1739,7 +1742,7 @@ ZEXTERN uLong ZEXPORT crc32   OF((uLong crc, const Bytef *buf, uInt len));
     if (crc != original_crc) error();
 */

-ZEXTERN uLong ZEXPORT crc32_z OF((uLong adler, const Bytef *buf,
+ZEXTERN uLong ZEXPORT crc32_z OF((uLong crc, const Bytef *buf,
                                  z_size_t len));
 /*
     Same as crc32(), but with a size_t length.
@@ -1755,6 +1758,20 @@ ZEXTERN uLong ZEXPORT crc32_combine OF((uLong crc1, uLong crc2, z_off_t len2));
   len2.
 */

+/*
+ZEXTERN uLong ZEXPORT crc32_combine_gen OF((z_off_t len2));
+
+     Return the operator corresponding to length len2, to be used with
+   crc32_combine_op().
+*/
+
+ZEXTERN uLong ZEXPORT crc32_combine_op OF((uLong crc1, uLong crc2, uLong op));
+/*
+     Give the same result as crc32_combine(), using op in place of len2. op is
+   is generated from len2 by crc32_combine_gen(). This will be faster than
+   crc32_combine() if the generated op is used more than once.
+*/
+

                        /* various hacks, don't look :) */

@@ -1842,6 +1859,7 @@ ZEXTERN int ZEXPORT gzgetc_ OF((gzFile file));  /* backward compatibility */
   ZEXTERN z_off64_t ZEXPORT gzoffset64 OF((gzFile));
   ZEXTERN uLong ZEXPORT adler32_combine64 OF((uLong, uLong, z_off64_t));
   ZEXTERN uLong ZEXPORT crc32_combine64 OF((uLong, uLong, z_off64_t));
+   ZEXTERN uLong ZEXPORT crc32_combine_gen64 OF((z_off64_t));
 #endif

 #if !defined(ZLIB_INTERNAL) && defined(Z_WANT64)
@@ -1852,6 +1870,7 @@ ZEXTERN int ZEXPORT gzgetc_ OF((gzFile file));  /* backward compatibility */
 #    define z_gzoffset z_gzoffset64
 #    define z_adler32_combine z_adler32_combine64
 #    define z_crc32_combine z_crc32_combine64
+#    define z_crc32_combine_gen z_crc32_combine_gen64
 #  else
 #    define gzopen gzopen64
 #    define gzseek gzseek64
@@ -1859,6 +1878,7 @@ ZEXTERN int ZEXPORT gzgetc_ OF((gzFile file));  /* backward compatibility */
 #    define gzoffset gzoffset64
 #    define adler32_combine adler32_combine64
 #    define crc32_combine crc32_combine64
+#    define crc32_combine_gen crc32_combine_gen64
 #  endif
 #  ifndef Z_LARGE64
     ZEXTERN gzFile ZEXPORT gzopen64 OF((const char *, const char *));
@@ -1867,6 +1887,7 @@ ZEXTERN int ZEXPORT gzgetc_ OF((gzFile file));  /* backward compatibility */
     ZEXTERN z_off_t ZEXPORT gzoffset64 OF((gzFile));
     ZEXTERN uLong ZEXPORT adler32_combine64 OF((uLong, uLong, z_off_t));
     ZEXTERN uLong ZEXPORT crc32_combine64 OF((uLong, uLong, z_off_t));
+     ZEXTERN uLong ZEXPORT crc32_combine_gen64 OF((z_off_t));
 #  endif
 #else
   ZEXTERN gzFile ZEXPORT gzopen OF((const char *, const char *));
@@ -1875,12 +1896,14 @@ ZEXTERN int ZEXPORT gzgetc_ OF((gzFile file));  /* backward compatibility */
   ZEXTERN z_off_t ZEXPORT gzoffset OF((gzFile));
   ZEXTERN uLong ZEXPORT adler32_combine OF((uLong, uLong, z_off_t));
   ZEXTERN uLong ZEXPORT crc32_combine OF((uLong, uLong, z_off_t));
+   ZEXTERN uLong ZEXPORT crc32_combine_gen OF((z_off_t));
 #endif

 #else /* Z_SOLO */

   ZEXTERN uLong ZEXPORT adler32_combine OF((uLong, uLong, z_off_t));
   ZEXTERN uLong ZEXPORT crc32_combine OF((uLong, uLong, z_off_t));
+   ZEXTERN uLong ZEXPORT crc32_combine_gen OF((z_off_t));

 #endif /* !Z_SOLO */

@@ -1890,10 +1913,10 @@ ZEXTERN int            ZEXPORT inflateSyncPoint OF((z_streamp));
 ZEXTERN const z_crc_t FAR * ZEXPORT get_crc_table    OF((void));
 ZEXTERN int            ZEXPORT inflateUndermine OF((z_streamp, int));
 ZEXTERN int            ZEXPORT inflateValidate OF((z_streamp, int));
-ZEXTERN unsigned long  ZEXPORT inflateCodesUsed OF ((z_streamp));
+ZEXTERN unsigned long  ZEXPORT inflateCodesUsed OF((z_streamp));
 ZEXTERN int            ZEXPORT inflateResetKeep OF((z_streamp));
 ZEXTERN int            ZEXPORT deflateResetKeep OF((z_streamp));
-#if (defined(_WIN32) || defined(__CYGWIN__)) && !defined(Z_SOLO)
+#if defined(_WIN32) && !defined(Z_SOLO)
 ZEXTERN gzFile         ZEXPORT gzopen_w OF((const wchar_t *path,
                                            const char *mode));
 #endif
@@ -61,9 +61,11 @@ uLong ZEXPORT zlibCompileFlags()
 #ifdef ZLIB_DEBUG
    flags += 1 << 8;
 #endif
+    /*
 #if defined(ASMV) || defined(ASMINF)
    flags += 1 << 9;
 #endif
+     */
 #ifdef ZLIB_WINAPI
    flags += 1 << 10;
 #endif
@@ -119,7 +121,7 @@ uLong ZEXPORT zlibCompileFlags()
 #  endif
 int ZLIB_INTERNAL z_verbose = verbose;

-void ZLIB_INTERNAL z_error (m)
+void ZLIB_INTERNAL z_error(m)
    char *m;
 {
    fprintf(stderr, "%s\n", m);
@@ -137,7 +139,7 @@ const char * ZEXPORT zError(err)
 }

 #if defined(_WIN32_WCE) && _WIN32_WCE < 0x800
-    /* The Microsoft C Run-Time Library for Windows CE doesn't have
+    /* The older Microsoft C Run-Time Library for Windows CE doesn't have
     * errno.  We define it as a global variable to simplify porting.
     * Its value is always 0 and should not be used.
     */
@@ -214,7 +216,7 @@ local ptr_table table[MAX_PTR];
 * a protected system like OS/2. Use Microsoft C instead.
 */

-voidpf ZLIB_INTERNAL zcalloc (voidpf opaque, unsigned items, unsigned size)
+voidpf ZLIB_INTERNAL zcalloc(voidpf opaque, unsigned items, unsigned size)
 {
    voidpf buf;
    ulg bsize = (ulg)items*size;
@@ -240,7 +242,7 @@ voidpf ZLIB_INTERNAL zcalloc (voidpf opaque, unsigned items, unsigned size)
    return buf;
 }

-void ZLIB_INTERNAL zcfree (voidpf opaque, voidpf ptr)
+void ZLIB_INTERNAL zcfree(voidpf opaque, voidpf ptr)
 {
    int n;

@@ -277,13 +279,13 @@ void ZLIB_INTERNAL zcfree (voidpf opaque, voidpf ptr)
 #  define _hfree   hfree
 #endif

-voidpf ZLIB_INTERNAL zcalloc (voidpf opaque, uInt items, uInt size)
+voidpf ZLIB_INTERNAL zcalloc(voidpf opaque, uInt items, uInt size)
 {
    (void)opaque;
    return _halloc((long)items, size);
 }

-void ZLIB_INTERNAL zcfree (voidpf opaque, voidpf ptr)
+void ZLIB_INTERNAL zcfree(voidpf opaque, voidpf ptr)
 {
    (void)opaque;
    _hfree(ptr);
@@ -302,7 +304,7 @@ extern voidp  calloc OF((uInt items, uInt size));
 extern void   free   OF((voidpf ptr));
 #endif

-voidpf ZLIB_INTERNAL zcalloc (opaque, items, size)
+voidpf ZLIB_INTERNAL zcalloc(opaque, items, size)
    voidpf opaque;
    unsigned items;
    unsigned size;
@@ -312,7 +314,7 @@ voidpf ZLIB_INTERNAL zcalloc (opaque, items, size)
                              (voidpf)calloc(items, size);
 }

-void ZLIB_INTERNAL zcfree (opaque, ptr)
+void ZLIB_INTERNAL zcfree(opaque, ptr)
    voidpf opaque;
    voidpf ptr;
 {
@@ -1,5 +1,5 @@
 /* zutil.h -- internal interface and configuration of the compression library
- * Copyright (C) 1995-2016 Jean-loup Gailly, Mark Adler
+ * Copyright (C) 1995-2022 Jean-loup Gailly, Mark Adler
 * For conditions of distribution and use, see copyright notice in zlib.h
 */

@@ -29,10 +29,6 @@
 #  include <stdlib.h>
 #endif

-#ifdef Z_SOLO
-   typedef long ptrdiff_t;  /* guess -- will be caught if guess is wrong */
-#endif
-
 #ifndef local
 #  define local static
 #endif
@@ -46,6 +42,17 @@ typedef unsigned short ush;
 typedef ush FAR ushf;
 typedef unsigned long  ulg;

+#if !defined(Z_U8) && !defined(Z_SOLO) && defined(STDC)
+#  include <limits.h>
+#  if (ULONG_MAX == 0xffffffffffffffff)
+#    define Z_U8 unsigned long
+#  elif (ULLONG_MAX == 0xffffffffffffffff)
+#    define Z_U8 unsigned long long
+#  elif (UINT_MAX == 0xffffffffffffffff)
+#    define Z_U8 unsigned
+#  endif
+#endif
+
 extern z_const char * const z_errmsg[10]; /* indexed by 2-zlib_error */
 /* (size given to avoid silly warnings with Visual C++) */

@@ -169,13 +176,7 @@ extern z_const char * const z_errmsg[10]; /* indexed by 2-zlib_error */

 #if (defined(_MSC_VER) && (_MSC_VER > 600)) && !defined __INTERIX
 #  if defined(_WIN32_WCE)
-#    if _WIN32_WCE < 0x800
-#      define fdopen(fd,mode) NULL /* No fdopen() */
-#      ifndef _PTRDIFF_T_DEFINED
-         typedef int ptrdiff_t;
-#        define _PTRDIFF_T_DEFINED
-#      endif
-#  endif
+#    define fdopen(fd,mode) NULL /* No fdopen() */
 #  else
 #    define fdopen(fd,type)  _fdopen(fd,type)
 #  endif
@@ -192,6 +193,7 @@ extern z_const char * const z_errmsg[10]; /* indexed by 2-zlib_error */
    (!defined(_LARGEFILE64_SOURCE) || _LFS64_LARGEFILE-0 == 0)
    ZEXTERN uLong ZEXPORT adler32_combine64 OF((uLong, uLong, z_off_t));
    ZEXTERN uLong ZEXPORT crc32_combine64 OF((uLong, uLong, z_off_t));
+    ZEXTERN uLong ZEXPORT crc32_combine_gen64 OF((z_off_t));
 #endif

        /* common defaults */
@@ -98,6 +98,12 @@ ocv_cmake_hook(CMAKE_INIT)

 # must go before the project()/enable_language() commands
 ocv_update(CMAKE_CONFIGURATION_TYPES "Debug;Release" CACHE STRING "Configs" FORCE)
+if(NOT DEFINED CMAKE_BUILD_TYPE
+    AND NOT OPENCV_SKIP_DEFAULT_BUILD_TYPE
+)
+  message(STATUS "'Release' build type is used by default. Use CMAKE_BUILD_TYPE to specify build type (Release or Debug)")
+  set(CMAKE_BUILD_TYPE "Release" CACHE STRING "Choose the type of build")
+endif()
 if(DEFINED CMAKE_BUILD_TYPE)
  set_property(CACHE CMAKE_BUILD_TYPE PROPERTY STRINGS "${CMAKE_CONFIGURATION_TYPES}")
 endif()
@@ -107,7 +113,7 @@ set(CMAKE_POSITION_INDEPENDENT_CODE ${ENABLE_PIC})

 ocv_cmake_hook(PRE_CMAKE_BOOTSTRAP)

-# Bootstap CMake system: setup CMAKE_SYSTEM_NAME and other vars
+# Bootstrap CMake system: setup CMAKE_SYSTEM_NAME and other vars
 enable_language(CXX C)

 ocv_cmake_hook(POST_CMAKE_BOOTSTRAP)
@@ -218,7 +224,7 @@ OCV_OPTION(BUILD_TIFF               "Build libtiff from source"          (WIN32
 OCV_OPTION(BUILD_JASPER             "Build libjasper from source"        (WIN32 OR ANDROID OR APPLE OR OPENCV_FORCE_3RDPARTY_BUILD) )
 OCV_OPTION(BUILD_JPEG               "Build libjpeg from source"          (WIN32 OR ANDROID OR APPLE OR OPENCV_FORCE_3RDPARTY_BUILD) )
 OCV_OPTION(BUILD_PNG                "Build libpng from source"           (WIN32 OR ANDROID OR APPLE OR OPENCV_FORCE_3RDPARTY_BUILD) )
-OCV_OPTION(BUILD_OPENEXR            "Build openexr from source"          (((WIN32 OR ANDROID OR APPLE) AND NOT WINRT) OR OPENCV_FORCE_3RDPARTY_BUILD) )
+OCV_OPTION(BUILD_OPENEXR            "Build openexr from source"          (OPENCV_FORCE_3RDPARTY_BUILD) )
 OCV_OPTION(BUILD_WEBP               "Build WebP from source"             (((WIN32 OR ANDROID OR APPLE) AND NOT WINRT) OR OPENCV_FORCE_3RDPARTY_BUILD) )
 OCV_OPTION(BUILD_TBB                "Download and build TBB from source" (ANDROID OR OPENCV_FORCE_3RDPARTY_BUILD) )
 OCV_OPTION(BUILD_IPP_IW             "Build IPP IW from source"           (NOT MINGW OR OPENCV_FORCE_3RDPARTY_BUILD) IF (X86_64 OR X86) AND NOT WINRT )
@@ -238,7 +244,7 @@ OCV_OPTION(WITH_AVFOUNDATION "Use AVFoundation for Video I/O (iOS/Mac)" ON
 OCV_OPTION(WITH_CARBON "Use Carbon for UI instead of Cocoa (OBSOLETE)" OFF
  VISIBLE_IF (APPLE OR HAVE_COCOA)
  VERIFY HAVE_CARBON)
-OCV_OPTION(WITH_CAROTENE "Use NVidia carotene acceleration library for ARM platform" ON
+OCV_OPTION(WITH_CAROTENE "Use NVidia carotene acceleration library for ARM platform" (NOT CV_DISABLE_OPTIMIZATION)
  VISIBLE_IF (ARM OR AARCH64) AND NOT IOS AND NOT (CMAKE_VERSION VERSION_LESS "2.8.11"))
 OCV_OPTION(WITH_CPUFEATURES "Use cpufeatures Android library" ON
  VISIBLE_IF ANDROID
@@ -285,12 +291,10 @@ OCV_OPTION(WITH_IPP "Include Intel IPP support" (NOT MINGW AND NOT CV_DISABLE_OP
 OCV_OPTION(WITH_HALIDE "Include Halide support" OFF
  VISIBLE_IF TRUE
  VERIFY HAVE_HALIDE)
-OCV_OPTION(WITH_INF_ENGINE "Include Intel Inference Engine support" OFF
+# replacement for deprecated options: WITH_INF_ENGINE, WITH_NGRAPH
+OCV_OPTION(WITH_OPENVINO "Include Intel OpenVINO toolkit support" (WITH_INF_ENGINE)
  VISIBLE_IF TRUE
-  VERIFY INF_ENGINE_TARGET)
-OCV_OPTION(WITH_NGRAPH "Include nGraph support" WITH_INF_ENGINE
-  VISIBLE_IF TRUE
-  VERIFY TARGET ngraph::ngraph)
+  VERIFY TARGET ocv.3rdparty.openvino)
 OCV_OPTION(WITH_JASPER "Include JPEG2K support" ON
  VISIBLE_IF NOT IOS
  VERIFY HAVE_JASPER)
@@ -300,7 +304,7 @@ OCV_OPTION(WITH_JPEG "Include JPEG support" ON
 OCV_OPTION(WITH_WEBP "Include WebP support" ON
  VISIBLE_IF NOT WINRT
  VERIFY HAVE_WEBP)
-OCV_OPTION(WITH_OPENEXR "Include ILM support via OpenEXR" BUILD_OPENEXR OR NOT CMAKE_CROSSCOMPILING
+OCV_OPTION(WITH_OPENEXR "Include ILM support via OpenEXR" ((WIN32 OR ANDROID OR APPLE) OR BUILD_OPENEXR) OR NOT CMAKE_CROSSCOMPILING
  VISIBLE_IF NOT APPLE_FRAMEWORK AND NOT WINRT
  VERIFY HAVE_OPENEXR)
 OCV_OPTION(WITH_OPENGL "Include OpenGL support" OFF
@@ -599,10 +603,6 @@ endif()
 # ----------------------------------------------------------------------------
 # OpenCV compiler and linker options
 # ----------------------------------------------------------------------------
-# In case of Makefiles if the user does not setup CMAKE_BUILD_TYPE, assume it's Release:
-if(CMAKE_GENERATOR MATCHES "Makefiles|Ninja" AND "${CMAKE_BUILD_TYPE}" STREQUAL "")
-  set(CMAKE_BUILD_TYPE Release)
-endif()

 ocv_cmake_hook(POST_CMAKE_BUILD_OPTIONS)

@@ -648,6 +648,8 @@ if(UNIX)
      set(OPENCV_LINKER_LIBS ${OPENCV_LINKER_LIBS} m pthread)
    elseif(EMSCRIPTEN)
      # no need to link to system libs with emscripten
+    elseif(QNXNTO)
+      set(OPENCV_LINKER_LIBS ${OPENCV_LINKER_LIBS} m)
    else()
      set(OPENCV_LINKER_LIBS ${OPENCV_LINKER_LIBS} dl m pthread rt)
    endif()
@@ -760,7 +762,7 @@ if(WITH_HALIDE)
 endif()

 # --- Inference Engine ---
-if(WITH_INF_ENGINE)
+if(WITH_INF_ENGINE OR WITH_OPENVINO)
  include(cmake/OpenCVDetectInferenceEngine.cmake)
 endif()

@@ -1230,12 +1232,17 @@ status("")
 status("  GUI: ")

 if(WITH_QT OR HAVE_QT)
-  if(HAVE_QT5)
-    status("    QT:" "YES (ver ${Qt5Core_VERSION_STRING})")
-    status("      QT OpenGL support:" HAVE_QT_OPENGL THEN "YES (${Qt5OpenGL_LIBRARIES} ${Qt5OpenGL_VERSION_STRING})" ELSE NO)
-  elseif(HAVE_QT)
+  if(HAVE_QT)
    status("    QT:" "YES (ver ${QT_VERSION_MAJOR}.${QT_VERSION_MINOR}.${QT_VERSION_PATCH} ${QT_EDITION})")
-    status("      QT OpenGL support:" HAVE_QT_OPENGL THEN "YES (${QT_QTOPENGL_LIBRARY})" ELSE NO)
+    if(HAVE_QT_OPENGL)
+      if(Qt${QT_VERSION_MAJOR}OpenGL_LIBRARIES)
+        status("      QT OpenGL support:" HAVE_QT_OPENGL THEN "YES (${Qt${QT_VERSION_MAJOR}OpenGL_LIBRARIES} ${Qt${QT_VERSION_MAJOR}OpenGL_VERSION_STRING})" ELSE NO)
+      else()
+        status("      QT OpenGL support:" HAVE_QT_OPENGL THEN "YES (${QT_QTOPENGL_LIBRARY})" ELSE NO)
+      endif()
+    else()
+      status("      QT OpenGL support:" "NO")
+    endif()
  else()
    status("    QT:" "NO")
  endif()
@@ -1515,54 +1522,61 @@ if(WITH_HALIDE OR HAVE_HALIDE)
  status("    Halide:"     HAVE_HALIDE      THEN "YES (${HALIDE_LIBRARIES} ${HALIDE_INCLUDE_DIRS})" ELSE NO)
 endif()

-if(WITH_INF_ENGINE OR INF_ENGINE_TARGET)
-  if(INF_ENGINE_TARGET)
-    list(GET INF_ENGINE_TARGET 0 ie_target)
-    set(__msg "YES (${INF_ENGINE_RELEASE} / ${INF_ENGINE_VERSION})")
-    get_target_property(_lib ${ie_target} IMPORTED_LOCATION)
-    get_target_property(_lib_imp_rel ${ie_target} IMPORTED_IMPLIB_RELEASE)
-    get_target_property(_lib_imp_dbg ${ie_target} IMPORTED_IMPLIB_DEBUG)
-    get_target_property(_lib_rel ${ie_target} IMPORTED_LOCATION_RELEASE)
-    get_target_property(_lib_dbg ${ie_target} IMPORTED_LOCATION_DEBUG)
-    ocv_build_features_string(_lib
-      IF _lib THEN "${_lib}"
-      IF _lib_imp_rel AND _lib_imp_dbg THEN "${_lib_imp_rel} / ${_lib_imp_dbg}"
-      IF _lib_rel AND _lib_dbg THEN "${_lib_rel} / ${_lib_dbg}"
-      IF _lib_rel  THEN "${_lib_rel}"
-      IF _lib_dbg  THEN "${_lib_dbg}"
-      ELSE "unknown"
-    )
-    get_target_property(_inc ${ie_target} INTERFACE_INCLUDE_DIRECTORIES)
-    status("    Inference Engine:" "${__msg}")
-    status("        * libs:" "${_lib}")
-    status("        * includes:" "${_inc}")
-  else()
-    status("    Inference Engine:"     "NO")
+if(HAVE_OPENVINO
+    OR (WITH_OPENVINO AND NOT WITH_INF_ENGINE AND NOT INF_ENGINE_TARGET)
+)
+  status("    OpenVINO:" TARGET openvino::runtime THEN "YES (${OpenVINO_VERSION})" ELSE "NO")
+else()
+  if(WITH_INF_ENGINE OR INF_ENGINE_TARGET)
+    if(INF_ENGINE_TARGET)
+      list(GET INF_ENGINE_TARGET 0 ie_target)
+      set(__msg "YES (${INF_ENGINE_RELEASE} / ${INF_ENGINE_VERSION})")
+      ocv_get_imported_target(ie_target "${ie_target}")
+      get_target_property(_lib ${ie_target} IMPORTED_LOCATION)
+      get_target_property(_lib_imp_rel ${ie_target} IMPORTED_IMPLIB_RELEASE)
+      get_target_property(_lib_imp_dbg ${ie_target} IMPORTED_IMPLIB_DEBUG)
+      get_target_property(_lib_rel ${ie_target} IMPORTED_LOCATION_RELEASE)
+      get_target_property(_lib_dbg ${ie_target} IMPORTED_LOCATION_DEBUG)
+      ocv_build_features_string(_lib
+        IF _lib THEN "${_lib}"
+        IF _lib_imp_rel AND _lib_imp_dbg THEN "${_lib_imp_rel} / ${_lib_imp_dbg}"
+        IF _lib_rel AND _lib_dbg THEN "${_lib_rel} / ${_lib_dbg}"
+        IF _lib_rel  THEN "${_lib_rel}"
+        IF _lib_dbg  THEN "${_lib_dbg}"
+        ELSE "unknown"
+      )
+      get_target_property(_inc ${ie_target} INTERFACE_INCLUDE_DIRECTORIES)
+      status("    Inference Engine:" "${__msg}")
+      status("        * libs:" "${_lib}")
+      status("        * includes:" "${_inc}")
+    else()
+      status("    Inference Engine:"     "NO")
+    endif()
  endif()
-endif()
-if(WITH_NGRAPH OR HAVE_NGRAPH)
-  if(HAVE_NGRAPH)
-    set(__target ngraph::ngraph)
-    set(__msg "YES (${ngraph_VERSION})")
-    get_target_property(_lib ${__target} IMPORTED_LOCATION)
-    get_target_property(_lib_imp_rel ${__target} IMPORTED_IMPLIB_RELEASE)
-    get_target_property(_lib_imp_dbg ${__target} IMPORTED_IMPLIB_DEBUG)
-    get_target_property(_lib_rel ${__target} IMPORTED_LOCATION_RELEASE)
-    get_target_property(_lib_dbg ${__target} IMPORTED_LOCATION_DEBUG)
-    ocv_build_features_string(_lib
-      IF _lib THEN "${_lib}"
-      IF _lib_imp_rel AND _lib_imp_dbg THEN "${_lib_imp_rel} / ${_lib_imp_dbg}"
-      IF _lib_rel AND _lib_dbg THEN "${_lib_rel} / ${_lib_dbg}"
-      IF _lib_rel  THEN "${_lib_rel}"
-      IF _lib_dbg  THEN "${_lib_dbg}"
-      ELSE "unknown"
-    )
-    get_target_property(_inc ${__target} INTERFACE_INCLUDE_DIRECTORIES)
-    status("    nGraph:" "${__msg}")
-    status("        * libs:" "${_lib}")
-    status("        * includes:" "${_inc}")
-  else()
-    status("    nGraph:"     "NO")
+  if(WITH_NGRAPH OR HAVE_NGRAPH)
+    if(HAVE_NGRAPH)
+      ocv_get_imported_target(__target ngraph::ngraph)
+      set(__msg "YES (${ngraph_VERSION})")
+      get_target_property(_lib ${__target} IMPORTED_LOCATION)
+      get_target_property(_lib_imp_rel ${__target} IMPORTED_IMPLIB_RELEASE)
+      get_target_property(_lib_imp_dbg ${__target} IMPORTED_IMPLIB_DEBUG)
+      get_target_property(_lib_rel ${__target} IMPORTED_LOCATION_RELEASE)
+      get_target_property(_lib_dbg ${__target} IMPORTED_LOCATION_DEBUG)
+      ocv_build_features_string(_lib
+        IF _lib THEN "${_lib}"
+        IF _lib_imp_rel AND _lib_imp_dbg THEN "${_lib_imp_rel} / ${_lib_imp_dbg}"
+        IF _lib_rel AND _lib_dbg THEN "${_lib_rel} / ${_lib_dbg}"
+        IF _lib_rel  THEN "${_lib_rel}"
+        IF _lib_dbg  THEN "${_lib_dbg}"
+        ELSE "unknown"
+      )
+      get_target_property(_inc ${__target} INTERFACE_INCLUDE_DIRECTORIES)
+      status("    nGraph:" "${__msg}")
+      status("        * libs:" "${_lib}")
+      status("        * includes:" "${_inc}")
+    else()
+      status("    nGraph:"     "NO")
+    endif()
  endif()
 endif()

@@ -7,13 +7,13 @@ copy or use the software.
               For Open Source Computer Vision Library
                       (3-clause BSD License)

-Copyright (C) 2000-2021, Intel Corporation, all rights reserved.
+Copyright (C) 2000-2022, Intel Corporation, all rights reserved.
 Copyright (C) 2009-2011, Willow Garage Inc., all rights reserved.
 Copyright (C) 2009-2016, NVIDIA Corporation, all rights reserved.
 Copyright (C) 2010-2013, Advanced Micro Devices, Inc., all rights reserved.
-Copyright (C) 2015-2021, OpenCV Foundation, all rights reserved.
+Copyright (C) 2015-2022, OpenCV Foundation, all rights reserved.
 Copyright (C) 2015-2016, Itseez Inc., all rights reserved.
-Copyright (C) 2019-2021, Xperience AI, all rights reserved.
+Copyright (C) 2019-2022, Xperience AI, all rights reserved.
 Third party copyrights are property of their respective owners.

 Redistribution and use in source and binary forms, with or without modification,
@@ -252,7 +252,7 @@ bool CvCascadeClassifier::train( const string _cascadeDirName,
            fs << "}";
        }
        // save current stage
-        char buf[10];
+        char buf[32];
        sprintf(buf, "%s%d", "stage", i );
        string stageFilename = dirName + buf + ".xml";
        FileStorage fs( stageFilename, FileStorage::WRITE );
@@ -86,6 +86,9 @@ int main( int argc, const char** argv )
        "{ image i        |      | (required) path to reference image }"
        "{ model m        |      | (required) path to cascade xml file }"
        "{ data d         |      | (optional) path to video output folder }"
+        "{ ext            | avi  | (optional) output video file extension e.g. avi (default) or mp4 }"
+        "{ fourcc         | XVID | (optional) output video file's 4-character codec e.g. XVID (default) or H264 }"
+        "{ fps            |   15 | (optional) output video file's frames-per-second rate }"
    );
    // Read in the input arguments
    if (parser.has("help")){
@@ -96,7 +99,9 @@ int main( int argc, const char** argv )
    string model(parser.get<string>("model"));
    string output_folder(parser.get<string>("data"));
    string image_ref = (parser.get<string>("image"));
-    if (model.empty() || image_ref.empty()){
+    string fourcc = (parser.get<string>("fourcc"));
+    int fps = parser.get<int>("fps");
+    if (model.empty() || image_ref.empty() || fourcc.size()!=4 || fps<1){
        parser.printMessage();
        printLimits();
        return -1;
@@ -166,11 +171,19 @@ int main( int argc, const char** argv )
    // each stage, containing all weak classifiers for that stage.
    bool draw_planes = false;
    stringstream output_video;
-    output_video << output_folder << "model_visualization.avi";
+    output_video << output_folder << "model_visualization." << parser.get<string>("ext");
    VideoWriter result_video;
    if( output_folder.compare("") != 0 ){
        draw_planes = true;
-        result_video.open(output_video.str(), VideoWriter::fourcc('X','V','I','D'), 15, Size(reference_image.cols * resize_factor, reference_image.rows * resize_factor), false);
+        result_video.open(output_video.str(), VideoWriter::fourcc(fourcc[0],fourcc[1],fourcc[2],fourcc[3]), fps, visualization.size(), false);
+        if (!result_video.isOpened()){
+            cerr << "the output video '" << output_video.str() << "' could not be opened."
+                 << " fourcc=" << fourcc
+                 << " fps=" << fps
+                 << " frameSize=" << visualization.size()
+                 << endl;
+            return -1;
+        }
    }

    if(haar){
@@ -87,11 +87,3 @@ endif()
 set( CMAKE_SHARED_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS} ${OPENCV_LINKER_DEFENSES_FLAGS_COMMON}" )
 set( CMAKE_MODULE_LINKER_FLAGS "${CMAKE_MODULE_LINKER_FLAGS} ${OPENCV_LINKER_DEFENSES_FLAGS_COMMON}" )
 set( CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} ${OPENCV_LINKER_DEFENSES_FLAGS_COMMON}" )
-
-if(CV_GCC OR CV_CLANG)
-  foreach(flags
-          CMAKE_CXX_FLAGS CMAKE_CXX_FLAGS_RELEASE CMAKE_CXX_FLAGS_DEBUG
-          CMAKE_C_FLAGS CMAKE_C_FLAGS_RELEASE CMAKE_C_FLAGS_DEBUG)
-    string(REPLACE "-O3" "-O2" ${flags} "${${flags}}")
-  endforeach()
-endif()
@@ -46,7 +46,7 @@

 set(CPU_ALL_OPTIMIZATIONS "SSE;SSE2;SSE3;SSSE3;SSE4_1;SSE4_2;POPCNT;AVX;FP16;AVX2;FMA3;AVX_512F")
 list(APPEND CPU_ALL_OPTIMIZATIONS "AVX512_COMMON;AVX512_KNL;AVX512_KNM;AVX512_SKX;AVX512_CNL;AVX512_CLX;AVX512_ICL")
-list(APPEND CPU_ALL_OPTIMIZATIONS NEON VFPV3 FP16)
+list(APPEND CPU_ALL_OPTIMIZATIONS NEON VFPV3 FP16 NEON_DOTPROD)
 list(APPEND CPU_ALL_OPTIMIZATIONS MSA)
 list(APPEND CPU_ALL_OPTIMIZATIONS VSX VSX3)
 list(REMOVE_DUPLICATES CPU_ALL_OPTIMIZATIONS)
@@ -238,7 +238,7 @@ if(X86 OR X86_64)
    ocv_intel_compiler_optimization_option(FP16 "-mavx" "/arch:AVX")
    ocv_intel_compiler_optimization_option(AVX "-mavx" "/arch:AVX")
    ocv_intel_compiler_optimization_option(FMA3 "" "")
-    ocv_intel_compiler_optimization_option(POPCNT "" "")
+    ocv_intel_compiler_optimization_option(POPCNT "-mpopcnt" "")  # -mpopcnt is available since ICC 19.0.0
    ocv_intel_compiler_optimization_option(SSE4_2 "-msse4.2" "/arch:SSE4.2")
    ocv_intel_compiler_optimization_option(SSE4_1 "-msse4.1" "/arch:SSE4.1")
    ocv_intel_compiler_optimization_option(SSE3 "-msse3" "/arch:SSE3")
@@ -326,6 +326,7 @@ if(X86 OR X86_64)
 elseif(ARM OR AARCH64)
  ocv_update(CPU_NEON_TEST_FILE "${OpenCV_SOURCE_DIR}/cmake/checks/cpu_neon.cpp")
  ocv_update(CPU_FP16_TEST_FILE "${OpenCV_SOURCE_DIR}/cmake/checks/cpu_fp16.cpp")
+  ocv_update(CPU_NEON_DOTPROD_TEST_FILE "${OpenCV_SOURCE_DIR}/cmake/checks/cpu_dotprod.cpp")
  if(NOT AARCH64)
    ocv_update(CPU_KNOWN_OPTIMIZATIONS "VFPV3;NEON;FP16")
    if(NOT MSVC)
@@ -337,9 +338,11 @@ elseif(ARM OR AARCH64)
    endif()
    ocv_update(CPU_FP16_IMPLIES "NEON")
  else()
-    ocv_update(CPU_KNOWN_OPTIMIZATIONS "NEON;FP16")
+    ocv_update(CPU_KNOWN_OPTIMIZATIONS "NEON;FP16;NEON_DOTPROD")
    ocv_update(CPU_NEON_FLAGS_ON "")
    ocv_update(CPU_FP16_IMPLIES "NEON")
+    ocv_update(CPU_NEON_DOTPROD_FLAGS_ON "-march=armv8.2-a+dotprod")
+    ocv_update(CPU_NEON_DOTPROD_IMPLIES "NEON")
    set(CPU_BASELINE "NEON;FP16" CACHE STRING "${HELP_CPU_BASELINE}")
  endif()
 elseif(MIPS)
@@ -677,7 +680,7 @@ macro(ocv_compiler_optimization_process_sources SOURCES_VAR_NAME LIBS_VAR_NAME T
          if(fname_LOWER MATCHES "\\.${OPT_LOWER}\\.cpp$")
 #message("${fname} BASELINE-${OPT}")
            set(__opt_found 1)
-            list(APPEND __result "${fname}")
+            list(APPEND __result_${OPT} "${fname}")
            break()
          endif()
        endforeach()
@@ -711,7 +714,7 @@ macro(ocv_compiler_optimization_process_sources SOURCES_VAR_NAME LIBS_VAR_NAME T
    endif()
  endforeach()

-  foreach(OPT ${CPU_DISPATCH_FINAL})
+  foreach(OPT ${CPU_BASELINE_FINAL} ${CPU_DISPATCH_FINAL})
    if(__result_${OPT})
 #message("${OPT}: ${__result_${OPT}}")
      if(CMAKE_GENERATOR MATCHES "^Visual"
@@ -119,12 +119,12 @@ if(CV_GCC OR CV_CLANG)
    # we want.
    add_extra_compiler_option(-Wall)
  endif()
-  add_extra_compiler_option(-Werror=return-type)
-  add_extra_compiler_option(-Werror=non-virtual-dtor)
-  add_extra_compiler_option(-Werror=address)
-  add_extra_compiler_option(-Werror=sequence-point)
+  add_extra_compiler_option(-Wreturn-type)
+  add_extra_compiler_option(-Wnon-virtual-dtor)
+  add_extra_compiler_option(-Waddress)
+  add_extra_compiler_option(-Wsequence-point)
  add_extra_compiler_option(-Wformat)
-  add_extra_compiler_option(-Werror=format-security -Wformat)
+  add_extra_compiler_option(-Wformat-security -Wformat)
  add_extra_compiler_option(-Wmissing-declarations)
  add_extra_compiler_option(-Wmissing-prototypes)
  add_extra_compiler_option(-Wstrict-prototypes)
@@ -134,7 +134,7 @@ if(CV_GCC OR CV_CLANG)
  add_extra_compiler_option(-Wshadow)
  add_extra_compiler_option(-Wsign-promo)
  add_extra_compiler_option(-Wuninitialized)
-  if(CV_GCC AND (CMAKE_CXX_COMPILER_VERSION VERSION_GREATER 6.0) AND (CMAKE_CXX_COMPILER_VERSION VERSION_LESS 7.0))
+  if(CV_GCC AND (CMAKE_CXX_COMPILER_VERSION VERSION_GREATER 6.0) AND (CMAKE_CXX_COMPILER_VERSION VERSION_LESS 7.0 OR ARM))
    add_extra_compiler_option(-Wno-psabi)
  endif()
  if(HAVE_CXX11)
@@ -177,7 +177,13 @@ if(CV_GCC OR CV_CLANG)
  endif()

  # We need pthread's
-  if(UNIX AND NOT ANDROID AND NOT (APPLE AND CV_CLANG)) # TODO
+  if((UNIX
+          AND NOT ANDROID
+          AND NOT (APPLE AND CV_CLANG)
+          AND NOT EMSCRIPTEN
+      )
+      OR (EMSCRIPTEN AND WITH_PTHREADS_PF)  # https://github.com/opencv/opencv/issues/20285
+  )
    add_extra_compiler_option(-pthread)
  endif()

@@ -225,9 +231,11 @@ if(CV_GCC OR CV_CLANG)
        if(APPLE)
          set(OPENCV_EXTRA_EXE_LINKER_FLAGS "${OPENCV_EXTRA_EXE_LINKER_FLAGS} -Wl,-dead_strip")
          set(OPENCV_EXTRA_SHARED_LINKER_FLAGS "${OPENCV_EXTRA_SHARED_LINKER_FLAGS} -Wl,-dead_strip")
+          set(OPENCV_EXTRA_MODULE_LINKER_FLAGS "${OPENCV_EXTRA_MODULE_LINKER_FLAGS} -Wl,-dead_strip")
        else()
          set(OPENCV_EXTRA_EXE_LINKER_FLAGS "${OPENCV_EXTRA_EXE_LINKER_FLAGS} -Wl,--gc-sections")
          set(OPENCV_EXTRA_SHARED_LINKER_FLAGS "${OPENCV_EXTRA_SHARED_LINKER_FLAGS} -Wl,--gc-sections")
+          set(OPENCV_EXTRA_MODULE_LINKER_FLAGS "${OPENCV_EXTRA_MODULE_LINKER_FLAGS} -Wl,--gc-sections")
        endif()
      endif()
    endif()
@@ -279,6 +287,7 @@ if(MSVC)
    set(OPENCV_EXTRA_FLAGS_RELEASE "${OPENCV_EXTRA_FLAGS_RELEASE} /Zi")
    set(OPENCV_EXTRA_EXE_LINKER_FLAGS_RELEASE "${OPENCV_EXTRA_EXE_LINKER_FLAGS_RELEASE} /debug")
    set(OPENCV_EXTRA_SHARED_LINKER_FLAGS_RELEASE "${OPENCV_EXTRA_SHARED_LINKER_FLAGS_RELEASE} /debug")
+    set(OPENCV_EXTRA_MODULE_LINKER_FLAGS_RELEASE "${OPENCV_EXTRA_MODULE_LINKER_FLAGS_RELEASE} /debug")
  endif()

  # Remove unreferenced functions: function level linking
@@ -346,6 +355,23 @@ if(NOT OPENCV_SKIP_LINK_AS_NEEDED)
    if(HAVE_LINK_AS_NEEDED)
      set(OPENCV_EXTRA_EXE_LINKER_FLAGS "${OPENCV_EXTRA_EXE_LINKER_FLAGS} ${_option}")
      set(OPENCV_EXTRA_SHARED_LINKER_FLAGS "${OPENCV_EXTRA_SHARED_LINKER_FLAGS} ${_option}")
+      set(OPENCV_EXTRA_MODULE_LINKER_FLAGS "${OPENCV_EXTRA_MODULE_LINKER_FLAGS} ${_option}")
+    endif()
+  endif()
+endif()
+
+# Apply "-Wl,--no-undefined" linker flags: https://github.com/opencv/opencv/pull/21347
+if(NOT OPENCV_SKIP_LINK_NO_UNDEFINED)
+  if(UNIX AND (NOT APPLE OR NOT CMAKE_VERSION VERSION_LESS "3.2"))
+    set(_option "-Wl,--no-undefined")
+    set(_saved_CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS}")
+    set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} ${_option}")  # requires CMake 3.2+ and CMP0056
+    ocv_check_compiler_flag(CXX "" HAVE_LINK_NO_UNDEFINED)
+    set(CMAKE_EXE_LINKER_FLAGS "${_saved_CMAKE_EXE_LINKER_FLAGS}")
+    if(HAVE_LINK_NO_UNDEFINED)
+      set(OPENCV_EXTRA_EXE_LINKER_FLAGS "${OPENCV_EXTRA_EXE_LINKER_FLAGS} ${_option}")
+      set(OPENCV_EXTRA_SHARED_LINKER_FLAGS "${OPENCV_EXTRA_SHARED_LINKER_FLAGS} ${_option}")
+      set(OPENCV_EXTRA_MODULE_LINKER_FLAGS "${OPENCV_EXTRA_MODULE_LINKER_FLAGS} ${_option}")
    endif()
  endif()
 endif()
@@ -364,6 +390,9 @@ if(NOT OPENCV_SKIP_EXTRA_COMPILER_FLAGS)
  set(CMAKE_SHARED_LINKER_FLAGS         "${CMAKE_SHARED_LINKER_FLAGS} ${OPENCV_EXTRA_SHARED_LINKER_FLAGS}")
  set(CMAKE_SHARED_LINKER_FLAGS_RELEASE "${CMAKE_SHARED_LINKER_FLAGS_RELEASE} ${OPENCV_EXTRA_SHARED_LINKER_FLAGS_RELEASE}")
  set(CMAKE_SHARED_LINKER_FLAGS_DEBUG   "${CMAKE_SHARED_LINKER_FLAGS_DEBUG} ${OPENCV_EXTRA_SHARED_LINKER_FLAGS_DEBUG}")
+  set(CMAKE_MODULE_LINKER_FLAGS         "${CMAKE_MODULE_LINKER_FLAGS} ${OPENCV_EXTRA_MODULE_LINKER_FLAGS}")
+  set(CMAKE_MODULE_LINKER_FLAGS_RELEASE "${CMAKE_MODULE_LINKER_FLAGS_RELEASE} ${OPENCV_EXTRA_MODULE_LINKER_FLAGS_RELEASE}")
+  set(CMAKE_MODULE_LINKER_FLAGS_DEBUG   "${CMAKE_MODULE_LINKER_FLAGS_DEBUG} ${OPENCV_EXTRA_MODULE_LINKER_FLAGS_DEBUG}")
 endif()

 if(MSVC)
@@ -387,6 +416,9 @@ if(MSVC)
    endif()
  endif()

+  # Enable [[attribute]] syntax checking to prevent silent failure: "attribute is ignored in this syntactic position"
+  add_extra_compiler_option("/w15240")
+
  if(NOT ENABLE_NOISY_WARNINGS)
    ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4127) # conditional expression is constant
    ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4251) # class 'std::XXX' needs to have dll-interface to be used by clients of YYY
@@ -394,6 +426,7 @@ if(MSVC)
    ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4275) # non dll-interface class 'std::exception' used as base for dll-interface class 'cv::Exception'
    ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4512) # Assignment operator could not be generated
    ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4589) # Constructor of abstract class 'cv::ORB' ignores initializer for virtual base class 'cv::Algorithm'
+    ocv_warnings_disable(CMAKE_CXX_FLAGS /wd4819) # Symbols like delta or epsilon cannot be represented
  endif()

  if(CV_ICC AND NOT ENABLE_NOISY_WARNINGS)
@@ -102,7 +102,7 @@ if(CUDA_FOUND)
  if(CUDA_GENERATION)
    if(NOT ";${_generations};" MATCHES ";${CUDA_GENERATION};")
      string(REPLACE ";" ", " _generations "${_generations}")
-      message(FATAL_ERROR "ERROR: ${_generations} Generations are suppered.")
+      message(FATAL_ERROR "ERROR: ${_generations} Generations are supported.")
    endif()
    unset(CUDA_ARCH_BIN CACHE)
    unset(CUDA_ARCH_PTX CACHE)
@@ -243,12 +243,13 @@ if(CUDA_FOUND)
      endif()
      if(NOT _nvcc_res EQUAL 0)
        message(STATUS "Automatic detection of CUDA generation failed. Going to build for all known architectures.")
-        # TX1 (5.3) TX2 (6.2) Xavier (7.2) V100 (7.0)
+        # TX1 (5.3) TX2 (6.2) Xavier (7.2) V100 (7.0) Orin (8.7)
        ocv_filter_available_architecture(__cuda_arch_bin
            5.3
            6.2
            7.2
            7.0
+            8.7
        )
      else()
        set(__cuda_arch_bin "${_nvcc_out}")
@@ -169,6 +169,8 @@ elseif(MSVC)
    set(OpenCV_RUNTIME vc15)
  elseif(MSVC_VERSION MATCHES "^192[0-9]$")
    set(OpenCV_RUNTIME vc16)
+  elseif(MSVC_VERSION MATCHES "^193[0-9]$")
+    set(OpenCV_RUNTIME vc17)
  else()
    message(WARNING "OpenCV does not recognize MSVC_VERSION \"${MSVC_VERSION}\". Cannot set OpenCV_RUNTIME")
  endif()
@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 3.1)
+cmake_minimum_required(VERSION ${MIN_VER_CMAKE})

 if(" ${HALIDE_ROOT_DIR}" STREQUAL " ")
  unset(HALIDE_ROOT_DIR CACHE)
@@ -9,9 +9,14 @@ set(HALIDE_ROOT_DIR "${HALIDE_ROOT_DIR}" CACHE PATH "Halide root directory")
 if(NOT HAVE_HALIDE)
  find_package(Halide QUIET) # Try CMake-based config files
  if(Halide_FOUND)
-    set(HALIDE_INCLUDE_DIRS "${Halide_INCLUDE_DIRS}" CACHE PATH "Halide include directories" FORCE)
-    set(HALIDE_LIBRARIES "${Halide_LIBRARIES}" CACHE PATH "Halide libraries" FORCE)
-    set(HAVE_HALIDE TRUE)
+    if(TARGET Halide::Halide)  # modern Halide scripts defines imported target
+      set(HALIDE_INCLUDE_DIRS "")
+      set(HALIDE_LIBRARIES "Halide::Halide")
+      set(HAVE_HALIDE TRUE)
+    else()
+      # using HALIDE_INCLUDE_DIRS / Halide_LIBRARIES
+      set(HAVE_HALIDE TRUE)
+    endif()
  endif()
 endif()

@@ -28,18 +33,15 @@ if(NOT HAVE_HALIDE AND HALIDE_ROOT_DIR)
  )
  if(HALIDE_LIBRARY AND HALIDE_INCLUDE_DIR)
    # TODO try_compile
-    set(HALIDE_INCLUDE_DIRS "${HALIDE_INCLUDE_DIR}" CACHE PATH "Halide include directories" FORCE)
-    set(HALIDE_LIBRARIES "${HALIDE_LIBRARY}" CACHE PATH "Halide libraries" FORCE)
+    set(HALIDE_INCLUDE_DIRS "${HALIDE_INCLUDE_DIR}")
+    set(HALIDE_LIBRARIES "${HALIDE_LIBRARY}")
    set(HAVE_HALIDE TRUE)
  endif()
-  if(NOT HAVE_HALIDE)
-    ocv_clear_vars(HALIDE_LIBRARIES HALIDE_INCLUDE_DIRS CACHE)
-  endif()
 endif()

 if(HAVE_HALIDE)
-  include_directories(${HALIDE_INCLUDE_DIRS})
+  if(HALIDE_INCLUDE_DIRS)
+    include_directories(${HALIDE_INCLUDE_DIRS})
+  endif()
  list(APPEND OPENCV_LINKER_LIBS ${HALIDE_LIBRARIES})
-else()
-  ocv_clear_vars(HALIDE_INCLUDE_DIRS HALIDE_LIBRARIES)
 endif()
@@ -1,21 +1,7 @@
-# The script detects Intel(R) Inference Engine installation
-#
-# Cache variables:
-# INF_ENGINE_RELEASE - a number reflecting IE source interface (linked with OpenVINO release)
-#
-# Detect parameters:
-# 1. Native cmake IE package:
-#    - environment variable InferenceEngine_DIR is set to location of cmake module
-# 2. Custom location:
-#    - INF_ENGINE_INCLUDE_DIRS - headers search location
-#    - INF_ENGINE_LIB_DIRS     - library search location
-# 3. OpenVINO location:
-#    - environment variable INTEL_OPENVINO_DIR is set to location of OpenVINO installation dir
-#    - INF_ENGINE_PLATFORM - part of name of library directory representing its platform
+# The script detects Intel(R) OpenVINO(TM) runtime installation
 #
 # Result:
-# INF_ENGINE_TARGET - set to name of imported library target representing InferenceEngine
-#
+# - target ocv.3rdparty.openvino

 if(NOT HAVE_CXX11)
    message(WARNING "DL Inference engine requires C++11. You can turn it on via ENABLE_CXX11=ON CMake flag.")
@@ -24,132 +10,53 @@ endif()

 # =======================

-macro(ocv_ie_find_extra_libraries find_prefix find_suffix)
-  file(GLOB libraries "${INF_ENGINE_LIB_DIRS}/${find_prefix}inference_engine*${find_suffix}")
-  foreach(full_path IN LISTS libraries)
-    get_filename_component(library "${full_path}" NAME_WE)
-    string(REPLACE "${find_prefix}" "" library "${library}")
-    if(library STREQUAL "inference_engine" OR library STREQUAL "inference_engined")
-      # skip
-    else()
-      add_library(${library} UNKNOWN IMPORTED)
-      set_target_properties(${library} PROPERTIES
-          IMPORTED_LOCATION "${full_path}")
-      list(APPEND custom_libraries ${library})
-    endif()
-  endforeach()
-endmacro()
-
-function(add_custom_ie_build _inc _lib _lib_rel _lib_dbg _msg)
-  if(NOT _inc OR NOT (_lib OR _lib_rel OR _lib_dbg))
+if(WITH_OPENVINO)
+  find_package(OpenVINO QUIET)
+  if(OpenVINO_FOUND)
+    message(STATUS "OpenVINO FOUND: ${OpenVINO_VERSION}")
+    math(EXPR ver "${OpenVINO_VERSION_MAJOR} * 1000000 + ${OpenVINO_VERSION_MINOR} * 10000 + ${OpenVINO_VERSION_PATCH} * 100")
+    ocv_add_external_target(openvino "" "openvino::runtime" "INF_ENGINE_RELEASE=${ver};HAVE_NGRAPH;HAVE_DNN_NGRAPH;HAVE_INF_ENGINE")
+    set(HAVE_OPENVINO 1)
    return()
  endif()
-  if(NOT _lib)
-    if(_lib_rel)
-      set(_lib "${_lib_rel}")
-    else()
-      set(_lib "${_lib_dbg}")
-    endif()
-  endif()
-  add_library(inference_engine UNKNOWN IMPORTED)
-  set_target_properties(inference_engine PROPERTIES
-    IMPORTED_LOCATION "${_lib}"
-    IMPORTED_IMPLIB_RELEASE "${_lib_rel}"
-    IMPORTED_IMPLIB_DEBUG "${_lib_dbg}"
-    INTERFACE_INCLUDE_DIRECTORIES "${_inc}"
-  )
-
-  set(custom_libraries "")
-  set(__prefixes "${CMAKE_FIND_LIBRARY_PREFIXES}")
-  if(NOT __prefixes)
-    set(__prefixes "_empty_")
-  endif()
-  foreach(find_prefix ${__prefixes})
-    if(find_prefix STREQUAL "_empty_")  # foreach doesn't iterate over empty elements
-      set(find_prefix "")
-    endif()
-    if(NOT DEFINED INFERENCE_ENGINE_FIND_LIBRARY_SUFFIXES)  # allow custom override
-      set(INFERENCE_ENGINE_FIND_LIBRARY_SUFFIXES ${CMAKE_FIND_LIBRARY_SUFFIXES})
-      if(APPLE)
-        ocv_list_filterout(INFERENCE_ENGINE_FIND_LIBRARY_SUFFIXES "^.so$")  # skip plugins (can't be linked)
-      endif()
-    endif()
-    foreach(find_suffix ${INFERENCE_ENGINE_FIND_LIBRARY_SUFFIXES})
-      ocv_ie_find_extra_libraries("${find_prefix}" "${find_suffix}")
-    endforeach()
-    if(NOT CMAKE_FIND_LIBRARY_SUFFIXES)
-      ocv_ie_find_extra_libraries("${find_prefix}" "")
-    endif()
-  endforeach()
-
-  if(NOT INF_ENGINE_RELEASE VERSION_GREATER "2018050000")
-    find_library(INF_ENGINE_OMP_LIBRARY iomp5 PATHS "${INF_ENGINE_OMP_DIR}" NO_DEFAULT_PATH)
-    if(NOT INF_ENGINE_OMP_LIBRARY)
-      message(WARNING "OpenMP for IE have not been found. Set INF_ENGINE_OMP_DIR variable if you experience build errors.")
-    endif()
-  endif()
-  if(EXISTS "${INF_ENGINE_OMP_LIBRARY}")
-    set_target_properties(inference_engine PROPERTIES IMPORTED_LINK_INTERFACE_LIBRARIES "${INF_ENGINE_OMP_LIBRARY}")
-  endif()
-  set(INF_ENGINE_VERSION "Unknown" CACHE STRING "")
-  set(INF_ENGINE_TARGET "inference_engine;${custom_libraries}" PARENT_SCOPE)
-  message(STATUS "Detected InferenceEngine: ${_msg}")
-endfunction()
+endif()

 # ======================

 find_package(InferenceEngine QUIET)
 if(InferenceEngine_FOUND)
  set(INF_ENGINE_TARGET ${InferenceEngine_LIBRARIES})
-  set(INF_ENGINE_VERSION "${InferenceEngine_VERSION}" CACHE STRING "")
+  set(INF_ENGINE_VERSION "${InferenceEngine_VERSION}")
  message(STATUS "Detected InferenceEngine: cmake package (${InferenceEngine_VERSION})")
 endif()

-if(NOT INF_ENGINE_TARGET AND INF_ENGINE_LIB_DIRS AND INF_ENGINE_INCLUDE_DIRS)
-  find_path(ie_custom_inc "inference_engine.hpp" PATHS "${INF_ENGINE_INCLUDE_DIRS}" NO_DEFAULT_PATH)
-  if(CMAKE_BUILD_TYPE STREQUAL "Debug")
-    find_library(ie_custom_lib_dbg "inference_engined" PATHS "${INF_ENGINE_LIB_DIRS}" NO_DEFAULT_PATH)  # Win32 and MacOSX
+if(DEFINED InferenceEngine_VERSION)
+  message(STATUS "InferenceEngine: ${InferenceEngine_VERSION}")
+  if(NOT INF_ENGINE_RELEASE AND NOT (InferenceEngine_VERSION VERSION_LESS "2021.4"))
+    math(EXPR INF_ENGINE_RELEASE_INIT "${InferenceEngine_VERSION_MAJOR} * 1000000 + ${InferenceEngine_VERSION_MINOR} * 10000 + ${InferenceEngine_VERSION_PATCH} * 100")
  endif()
-  find_library(ie_custom_lib "inference_engine" PATHS "${INF_ENGINE_LIB_DIRS}" NO_DEFAULT_PATH)
-  find_library(ie_custom_lib_rel "inference_engine" PATHS "${INF_ENGINE_LIB_DIRS}/Release" NO_DEFAULT_PATH)
-  find_library(ie_custom_lib_dbg "inference_engine" PATHS "${INF_ENGINE_LIB_DIRS}/Debug" NO_DEFAULT_PATH)
-  add_custom_ie_build("${ie_custom_inc}" "${ie_custom_lib}" "${ie_custom_lib_rel}" "${ie_custom_lib_dbg}" "INF_ENGINE_{INCLUDE,LIB}_DIRS")
 endif()
+if(NOT INF_ENGINE_RELEASE AND NOT INF_ENGINE_RELEASE_INIT)
+  message(STATUS "WARNING: InferenceEngine version has not been set, 2021.4.2 will be used by default. Set INF_ENGINE_RELEASE variable if you experience build errors.")
+  set(INF_ENGINE_RELEASE_INIT "2021040200")
+elseif(DEFINED INF_ENGINE_RELEASE)
+  set(INF_ENGINE_RELEASE_INIT "${INF_ENGINE_RELEASE}")
+endif()
+set(INF_ENGINE_RELEASE "${INF_ENGINE_RELEASE_INIT}" CACHE STRING "Force IE version, should be in form YYYYAABBCC (e.g. 2020.1.0.2 -> 2020010002)")

-set(_loc "$ENV{INTEL_OPENVINO_DIR}")
-if(NOT _loc AND DEFINED ENV{INTEL_CVSDK_DIR})
-  set(_loc "$ENV{INTEL_CVSDK_DIR}")  # OpenVINO 2018.x
-endif()
-if(NOT INF_ENGINE_TARGET AND _loc)
-  if(NOT INF_ENGINE_RELEASE VERSION_GREATER "2018050000")
-    set(INF_ENGINE_PLATFORM_DEFAULT "ubuntu_16.04")
-  else()
-    set(INF_ENGINE_PLATFORM_DEFAULT "")
-  endif()
-  set(INF_ENGINE_PLATFORM "${INF_ENGINE_PLATFORM_DEFAULT}" CACHE STRING "InferenceEngine platform (library dir)")
-  find_path(ie_custom_env_inc "inference_engine.hpp" PATHS "${_loc}/deployment_tools/inference_engine/include" NO_DEFAULT_PATH)
-  if(CMAKE_BUILD_TYPE STREQUAL "Debug")
-    find_library(ie_custom_env_lib_dbg "inference_engined" PATHS "${_loc}/deployment_tools/inference_engine/lib/${INF_ENGINE_PLATFORM}/intel64" NO_DEFAULT_PATH)
-  endif()
-  find_library(ie_custom_env_lib "inference_engine" PATHS "${_loc}/deployment_tools/inference_engine/lib/${INF_ENGINE_PLATFORM}/intel64" NO_DEFAULT_PATH)
-  find_library(ie_custom_env_lib_rel "inference_engine" PATHS "${_loc}/deployment_tools/inference_engine/lib/intel64/Release" NO_DEFAULT_PATH)
-  find_library(ie_custom_env_lib_dbg "inference_engine" PATHS "${_loc}/deployment_tools/inference_engine/lib/intel64/Debug" NO_DEFAULT_PATH)
-  add_custom_ie_build("${ie_custom_env_inc}" "${ie_custom_env_lib}" "${ie_custom_env_lib_rel}" "${ie_custom_env_lib_dbg}" "OpenVINO (${_loc})")
-endif()
+set(tgts)
+set(defs)

 # Add more features to the target
-
 if(INF_ENGINE_TARGET)
-  if(NOT INF_ENGINE_RELEASE)
-    message(WARNING "InferenceEngine version has not been set, 2021.3 will be used by default. Set INF_ENGINE_RELEASE variable if you experience build errors.")
-  endif()
-  set(INF_ENGINE_RELEASE "2021030000" CACHE STRING "Force IE version, should be in form YYYYAABBCC (e.g. 2020.1.0.2 -> 2020010002)")
  set_target_properties(${INF_ENGINE_TARGET} PROPERTIES
-    INTERFACE_COMPILE_DEFINITIONS "HAVE_INF_ENGINE=1;INF_ENGINE_RELEASE=${INF_ENGINE_RELEASE}"
+      INTERFACE_COMPILE_DEFINITIONS "HAVE_INF_ENGINE=1;INF_ENGINE_RELEASE=${INF_ENGINE_RELEASE}"
  )
+  list(APPEND tgts ${INF_ENGINE_TARGET})
+  list(APPEND defs "INF_ENGINE_RELEASE=${INF_ENGINE_RELEASE}" "HAVE_INF_ENGINE")
 endif()

-if(WITH_NGRAPH)
+if(WITH_NGRAPH OR NOT DEFINED WITH_NGRAPH)
  find_package(ngraph QUIET)
  if(ngraph_FOUND)
    ocv_assert(TARGET ngraph::ngraph)
@@ -158,5 +65,9 @@ if(WITH_NGRAPH)
    endif()
    message(STATUS "Detected ngraph: cmake package (${ngraph_VERSION})")
    set(HAVE_NGRAPH ON)
+    list(APPEND tgts ngraph::ngraph)
+    list(APPEND defs "HAVE_NGRAPH" "HAVE_DNN_NGRAPH")
  endif()
 endif()
+
+ocv_add_external_target(openvino "" "${tgts}" "${defs}")
@@ -177,7 +177,7 @@ if(NOT ${found})

    if(NOT ANDROID AND NOT IOS)
      if(CMAKE_HOST_UNIX)
-        execute_process(COMMAND ${_executable} -c "from distutils.sysconfig import *; print(get_python_lib())"
+        execute_process(COMMAND ${_executable} -c "from sysconfig import *; print(get_path('purelib'))"
                        RESULT_VARIABLE _cvpy_process
                        OUTPUT_VARIABLE _std_packages_path
                        OUTPUT_STRIP_TRAILING_WHITESPACE)
@@ -26,31 +26,14 @@ if(VTK_VERSION VERSION_LESS "5.8.0")
 endif()

 # Different Qt versions can't be linked together
-if(HAVE_QT5 AND VTK_VERSION VERSION_LESS "6.0.0")
-  if(VTK_USE_QT)
-    message(STATUS "VTK support is disabled. Incompatible combination: OpenCV + Qt5 and VTK ver.${VTK_VERSION} + Qt4")
-  endif()
-endif()
-
-# Different Qt versions can't be linked together. VTK 6.0.0 doesn't provide a way to get Qt version it was linked with
-if(HAVE_QT5 AND VTK_VERSION VERSION_EQUAL "6.0.0" AND NOT DEFINED FORCE_VTK)
-  message(STATUS "VTK support is disabled. Possible incompatible combination: OpenCV+Qt5, and VTK ver.${VTK_VERSION} with Qt4")
-  message(STATUS "If it is known that VTK was compiled without Qt4, please define '-DFORCE_VTK=TRUE' flag in CMake")
+if((HAVE_QT AND VTK_USE_QT)
+    AND NOT DEFINED FORCE_VTK  # deprecated
+    AND NOT DEFINED OPENCV_FORCE_VTK
+)
+  message(STATUS "VTK support is disabled. Possible incompatible combination: OpenCV+Qt, and VTK ver.${VTK_VERSION} with Qt")
+  message(STATUS "If it is known that VTK was compiled without Qt4, please define '-DOPENCV_FORCE_VTK=TRUE' flag in CMake")
  return()
 endif()

-# Different Qt versions can't be linked together
-if(HAVE_QT AND VTK_VERSION VERSION_GREATER "6.0.0" AND NOT ${VTK_QT_VERSION} STREQUAL "")
-  if(HAVE_QT5 AND ${VTK_QT_VERSION} EQUAL "4")
-    message(STATUS "VTK support is disabled. Incompatible combination: OpenCV + Qt5 and VTK ver.${VTK_VERSION} + Qt4")
-    return()
-  endif()
-
-  if(NOT HAVE_QT5 AND ${VTK_QT_VERSION} EQUAL "5")
-    message(STATUS "VTK support is disabled. Incompatible combination: OpenCV + Qt4 and VTK ver.${VTK_VERSION} + Qt5")
-    return()
-  endif()
-endif()
-
 set(HAVE_VTK ON)
 message(STATUS "Found VTK ${VTK_VERSION} (${VTK_USE_FILE})")
@@ -23,7 +23,7 @@ set(OPENCV_DOWNLOAD_LOG "${OpenCV_BINARY_DIR}/CMakeDownloadLog.txt")
 set(OPENCV_DOWNLOAD_WITH_CURL "${OpenCV_BINARY_DIR}/download_with_curl.sh")
 set(OPENCV_DOWNLOAD_WITH_WGET "${OpenCV_BINARY_DIR}/download_with_wget.sh")
 set(OPENCV_DOWNLOAD_TRIES_LIST 1 CACHE STRING "List of download tries") # a list
-set(OPENCV_DOWNLOAD_PARAMS INACTIVITY_TIMEOUT 60 TIMEOUT 600 CACHE STRING "Download parameters to be passed to file(DOWNLAOD ...)")
+set(OPENCV_DOWNLOAD_PARAMS INACTIVITY_TIMEOUT 60 TIMEOUT 600 CACHE STRING "Download parameters to be passed to file(DOWNLOAD ...)")
 mark_as_advanced(OPENCV_DOWNLOAD_TRIES_LIST OPENCV_DOWNLOAD_PARAMS)

 # Init download cache directory and log file and helper scripts
@@ -151,7 +151,7 @@ macro(ipp_detect_version)
        if("${name}" STREQUAL "core")  # https://github.com/opencv/opencv/pull/19681
          if(OPENCV_FORCE_IPP_EXCLUDE_LIBS OR OPENCV_FORCE_IPP_EXCLUDE_LIBS_CORE
              OR (UNIX AND NOT ANDROID AND NOT APPLE
-                  AND (CMAKE_CXX_COMPILER_ID MATCHES "GNU" OR CMAKE_CXX_COMPILER_ID MATCHES "Clang")
+                  AND CMAKE_CXX_COMPILER_ID MATCHES "GNU|Clang|Intel"
              )
              AND NOT OPENCV_SKIP_IPP_EXCLUDE_LIBS_CORE
          )
@@ -51,6 +51,23 @@ macro(ocv_lapack_check)
    if(NOT "${OPENCV_CBLAS_H_PATH_${_lapack_impl}}" STREQUAL "${OPENCV_LAPACKE_H_PATH_${_lapack_impl}}")
      list(APPEND _lapack_content "#include \"${OPENCV_LAPACKE_H_PATH_${_lapack_impl}}\"")
    endif()
+    list(APPEND _lapack_content "
+#if defined(LAPACK_GLOBAL) || defined(LAPACK_NAME)
+/*
+ * Using netlib's reference LAPACK implementation version >= 3.4.0 (first with C interface).
+ * Use LAPACK_xxxx to transparently (via predefined lapack macros) deal with pre and post 3.9.1 versions.
+ * LAPACK 3.9.1 introduces LAPACK_FORTRAN_STRLEN_END and modifies (through preprocessing) the declarations of the following functions used in opencv
+ *        sposv_, dposv_, spotrf_, dpotrf_, sgesdd_, dgesdd_, sgels_, dgels_
+ * which end up with an extra parameter.
+ * So we also need to preprocess the function calls in opencv coding by prefixing them with LAPACK_.
+ * The good news is the preprocessing works fine whatever netlib's LAPACK version.
+ */
+#define OCV_LAPACK_FUNC(f) LAPACK_##f
+#else
+/* Using other LAPACK implementations so fall back to opencv's assumption until now */
+#define OCV_LAPACK_FUNC(f) f##_
+#endif
+")
    if(${_lapack_add_extern_c})
      list(APPEND _lapack_content "}")
    endif()
@@ -11,25 +11,50 @@ if(WITH_WIN32UI)
    CMAKE_FLAGS "-DLINK_LIBRARIES:STRING=user32;gdi32")
 endif()

-# --- QT4 ---
-ocv_clear_vars(HAVE_QT HAVE_QT5)
-if(WITH_QT)
-  if(NOT WITH_QT EQUAL 4)
-    find_package(Qt5 COMPONENTS Core Gui Widgets Test Concurrent REQUIRED NO_MODULE)
-    if(Qt5_FOUND)
-      set(HAVE_QT5 ON)
-      set(HAVE_QT  ON)
-      find_package(Qt5 COMPONENTS OpenGL QUIET)
-      if(Qt5OpenGL_FOUND)
-        set(QT_QTOPENGL_FOUND ON)
-      endif()
-    endif()
+macro(ocv_find_package_Qt4)
+  find_package(Qt4 COMPONENTS QtCore QtGui QtTest ${ARGN})
+  if(QT4_FOUND)
+    set(QT_FOUND 1)
+    ocv_assert(QT_VERSION_MAJOR EQUAL 4)
  endif()
+endmacro()

-  if(NOT HAVE_QT)
-    find_package(Qt4 REQUIRED QtCore QtGui QtTest)
-    if(QT4_FOUND)
-      set(HAVE_QT TRUE)
+macro(ocv_find_package_Qt OCV_QT_VER)
+  find_package(Qt${OCV_QT_VER} COMPONENTS Core Gui Widgets Test Concurrent ${ARGN} NO_MODULE)
+  if(Qt${OCV_QT_VER}_FOUND)
+    set(QT_FOUND 1)
+    set(QT_VERSION "${Qt${OCV_QT_VER}_VERSION}")
+    set(QT_VERSION_MAJOR "${Qt${OCV_QT_VER}_VERSION_MAJOR}")
+    set(QT_VERSION_MINOR "${Qt${OCV_QT_VER}_VERSION_MINOR}")
+    set(QT_VERSION_PATCH "${Qt${OCV_QT_VER}_VERSION_PATCH}")
+    set(QT_VERSION_TWEAK "${Qt${OCV_QT_VER}_VERSION_TWEAK}")
+    set(QT_VERSION_COUNT "${Qt${OCV_QT_VER}_VERSION_COUNT}")
+  endif()
+endmacro()
+
+# --- QT4 ---
+if(WITH_QT)
+  if(NOT WITH_QT GREATER 0)
+    # BUG: Qt5Config.cmake script can't handle components properly: find_package(QT NAMES Qt6 Qt5 REQUIRED NO_MODULE COMPONENTS Core Gui Widgets Test Concurrent)
+    ocv_find_package_Qt(6 QUIET)
+    if(NOT QT_FOUND)
+      ocv_find_package_Qt(5 QUIET)
+    endif()
+    if(NOT QT_FOUND)
+      ocv_find_package_Qt4(QUIET)
+    endif()
+  elseif(WITH_QT EQUAL 4)
+    ocv_find_package_Qt4(REQUIRED)
+  else()  # WITH_QT=<major version>
+    ocv_find_package_Qt("${WITH_QT}" REQUIRED)
+  endif()
+  if(QT_FOUND)
+    set(HAVE_QT ON)
+    if(QT_VERSION_MAJOR GREATER 4)
+      find_package(Qt${QT_VERSION_MAJOR} COMPONENTS OpenGL QUIET)
+      if(Qt${QT_VERSION_MAJOR}OpenGL_FOUND)
+        set(QT_QTOPENGL_FOUND ON)  # HAVE_QT_OPENGL is defined below
+      endif()
    endif()
  endif()
 endif()
@@ -76,7 +101,6 @@ if(WITH_OPENGL)
    find_package (OpenGL QUIET)
    if(OPENGL_FOUND)
      set(HAVE_OPENGL TRUE)
-      list(APPEND OPENCV_LINKER_LIBS ${OPENGL_LIBRARIES})
      if(QT_QTOPENGL_FOUND)
        set(HAVE_QT_OPENGL TRUE)
      else()
@@ -240,7 +240,9 @@ if(WITH_OPENEXR)
    set(OPENEXR_LIBRARIES IlmImf)
    add_subdirectory("${OpenCV_SOURCE_DIR}/3rdparty/openexr")
    if(OPENEXR_VERSION)  # check via TARGET doesn't work
+      set(BUILD_OPENEXR ON)
      set(HAVE_OPENEXR YES)
+      set(BUILD_OPENEXR ON)
    endif()
  endif()
 endif()
@@ -29,7 +29,7 @@ if(WITH_IPP)
    if(OPENCV_FORCE_IPP_EXCLUDE_LIBS
        OR (HAVE_IPP_ICV
            AND UNIX AND NOT ANDROID AND NOT APPLE
-            AND (CMAKE_CXX_COMPILER_ID MATCHES "GNU" OR CMAKE_CXX_COMPILER_ID MATCHES "Clang")
+            AND CMAKE_CXX_COMPILER_ID MATCHES "GNU|Clang|Intel"
        )
        AND NOT OPENCV_SKIP_IPP_EXCLUDE_LIBS
    )
@@ -9,6 +9,27 @@
 # OPENEXR_LIBRARIES = libraries that are needed to use OpenEXR.
 #

+if(NOT HAVE_CXX11)
+  message(STATUS "OpenEXR: enable C++11 to use external OpenEXR")
+  return()
+endif()
+
+if(NOT OPENCV_SKIP_OPENEXR_FIND_PACKAGE)
+  find_package(OpenEXR 3 QUIET)
+  #ocv_cmake_dump_vars(EXR)
+  if(OpenEXR_FOUND)
+    if(TARGET OpenEXR::OpenEXR)  # OpenEXR 3+
+      set(OPENEXR_LIBRARIES OpenEXR::OpenEXR)
+      set(OPENEXR_INCLUDE_PATHS "")
+      set(OPENEXR_VERSION "${OpenEXR_VERSION}")
+      set(OPENEXR_FOUND 1)
+      return()
+    else()
+      message(STATUS "Unsupported find_package(OpenEXR) - missing OpenEXR::OpenEXR target (version ${OpenEXR_VERSION})")
+    endif()
+  endif()
+endif()
+
 SET(OPENEXR_LIBRARIES "")
 SET(OPENEXR_LIBSEARCH_SUFFIXES "")
 file(TO_CMAKE_PATH "$ENV{ProgramFiles}" ProgramFiles_ENV_PATH)
--- a/Show More
+++ b/Show More