From e376ffc5fe9c87d2779bf839278e08c2c2dffe70 Mon Sep 17 00:00:00 2001
From: Egor Pugin <egor.pugin@gmail.com>
Date: Thu, 23 Feb 2017 19:29:48 +0300
Subject: [PATCH] Fix avx on windows.

---
 CMakeLists.txt         | 12 +++++++++++-
 arch/dotproductavx.cpp | 11 ++++++++++-
 cppan.yml              | 16 +++++++++++-----
 3 files changed, 32 insertions(+), 7 deletions(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 9c74d6090f..d227b7dc36 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -84,7 +84,6 @@ if (STATIC)
 endif()
 
 if (WIN32)
-    add_definitions(-D__SSE4_1__)
     if (MSVC)
         add_definitions(-D_CRT_SECURE_NO_WARNINGS)
 
@@ -214,6 +213,17 @@ set(tesseract_src ${tesseract_src}
     api/pdfrenderer.cpp
 )
 
+if (WIN32)
+    set_source_files_properties(
+        ${CMAKE_CURRENT_SOURCE_DIR}/arch/dotproductsse.cpp
+        PROPERTIES COMPILE_DEFINITIONS __SSE4_1__)
+    if (MSVC)
+        set_source_files_properties(
+            ${CMAKE_CURRENT_SOURCE_DIR}/arch/dotproductavx.cpp
+            PROPERTIES COMPILE_FLAGS "/arch:AVX")
+    endif()
+endif()
+
 add_library                     (libtesseract ${LIBRARY_TYPE} ${tesseract_src} ${tesseract_hdr})
 if (NOT STATIC)
 target_compile_definitions      (libtesseract
diff --git a/arch/dotproductavx.cpp b/arch/dotproductavx.cpp
index 7ba784a7fc..9c84226439 100644
--- a/arch/dotproductavx.cpp
+++ b/arch/dotproductavx.cpp
@@ -90,8 +90,17 @@ double DotProductAVX(const double* u, const double* v, int n) {
   // _mm256_extract_f64 doesn't exist, but resist the temptation to use an sse
   // instruction, as that introduces a 70 cycle delay. All this casting is to
   // fool the instrinsics into thinking we are extracting the bottom int64.
+  auto cast_sum = _mm256_castpd_si256(sum);
   *(reinterpret_cast<inT64*>(&result)) =
-      _mm256_extract_epi64(_mm256_castpd_si256(sum), 0);
+#ifndef _WIN32
+      _mm256_extract_epi64(cast_sum, 0)
+#else
+      // this is a very simple workaround that probably could be activated
+      // for all other platforms that do not have _mm256_extract_epi64
+      // _mm256_extract_epi64(X, Y) == ((uint64_t*)&X)[Y]
+      ((uint64_t*)&cast_sum)[0]
+#endif
+      ;
   while (offset < n) {
     result += u[offset] * v[offset];
     ++offset;
diff --git a/cppan.yml b/cppan.yml
index b221ace1f4..dc0a1ffeda 100644
--- a/cppan.yml
+++ b/cppan.yml
@@ -108,13 +108,19 @@ projects:
           - _Bool
 
         pre_sources: |
-            # dummy config file
-            if (NOT EXISTS ${CMAKE_CURRENT_BINARY_DIR}/config_auto.h)
-                file(WRITE ${CMAKE_CURRENT_BINARY_DIR}/config_auto.h)
-            endif()
+            file_write_once(${BDIR}/config_auto.h "")
 
         post_sources: |
-            if (NOT WIN32)
+            if (WIN32)
+                set_source_files_properties(
+                    ${CMAKE_CURRENT_SOURCE_DIR}/arch/dotproductsse.cpp
+                    PROPERTIES COMPILE_DEFINITIONS __SSE4_1__)
+                if (MSVC)
+                    set_source_files_properties(
+                        ${CMAKE_CURRENT_SOURCE_DIR}/arch/dotproductavx.cpp
+                        PROPERTIES COMPILE_FLAGS "/arch:AVX")
+                endif()
+            else()
                 remove_src_dir(vs2010/port/*)
             endif()