Merge branch 'hotfix/fbx'

.github/workflows/linux.yml

@@ -80,7 +80,7 @@ jobs:
       run: |
         mkdir fbx
         cd fbx
-        sudo wget --user-agent='Mozilla/5.0' 'https://www.autodesk.com/content/dam/autodesk/www/adn/fbx/2020-2/fbx20202_fbxsdk_linux.tar.gz' -O 'fbx.tar.gz'
+        sudo wget --user-agent='Mozilla/5.0' 'https://damassets.autodesk.net/content/dam/autodesk/www/adn/fbx/2020-2/fbx20202_fbxsdk_linux.tar.gz' -O 'fbx.tar.gz'
         sudo tar -xf 'fbx.tar.gz'
         (yes yes | sudo ./*_fbxsdk_linux /usr/local) || true
         sudo chmod -R 755 /usr/local/lib

.github/workflows/macos.yml

@@ -19,7 +19,7 @@ jobs:
       run: |
         mkdir fbx
         cd fbx
-        sudo wget --user-agent='Mozilla/5.0' 'https://www.autodesk.com/content/dam/autodesk/www/adn/fbx/2020-0-1/fbx202001_fbxsdk_clang_mac.pkg.tgz' -O 'fbx.tgz'
+        sudo wget --user-agent='Mozilla/5.0' 'https://damassets.autodesk.net/content/dam/autodesk/www/adn/fbx/2020-0-1/fbx202001_fbxsdk_clang_mac.pkg.tgz' -O 'fbx.tgz'
         sudo tar -xf 'fbx.tgz'
         sudo installer -pkg *_fbxsdk_clang_macos.pkg -target /
         cd ..

.github/workflows/windows.yml

@@ -19,7 +19,7 @@ jobs:
       run: |
         md fbx
         cd fbx
-        Start-BitsTransfer -Source 'https://www.autodesk.com/content/dam/autodesk/www/adn/fbx/2020-2/fbx20202_fbxsdk_vs2019_win.exe' -Destination 'fbxsdk.exe'
+        Start-BitsTransfer -Source 'https://damassets.autodesk.net/content/dam/autodesk/www/adn/fbx/2020-3-4/fbx202034_fbxsdk_vs2022_win.exe' -Destination 'fbxsdk.exe'
         Start-Process -FilePath 'fbxsdk.exe' /S -Wait
         cd ..
     

CHANGES.md

@@ -1,9 +1,15 @@
+Release version 0.14.3
+----------------------
+
+* Build pipeline
+  - Adds vs2022 compiler support for fbxsdk (#170)
+
 Release version 0.14.2
 ----------------------
 
 * Library
   - Transitions away from sprintf to the more secure snprintf.
-  - #147 Works around gcc 11 error stringop-overflow which emits false positives for ozz math serialisation.
+  - #147 Works around gcc 11 error stringop-overflow which emits false positives for ozz math serialization.
 
 * Build pipeline
   - Updates CI compiler versions.

CMakeLists.txt

@@ -9,7 +9,7 @@ get_directory_property(is_sub_project PARENT_DIRECTORY)
 # Current version
 set(OZZ_VERSION_MAJOR 0)
 set(OZZ_VERSION_MINOR 14)
-set(OZZ_VERSION_PATCH 2)
+set(OZZ_VERSION_PATCH 3)
 set(OZZ_VERSION ${OZZ_VERSION_MAJOR}.${OZZ_VERSION_MINOR}.${OZZ_VERSION_PATCH})
 
 # Add project build options

build-utils/cmake/modules/FindFbx.cmake

@@ -67,7 +67,9 @@ function(FindFbxLibrariesGeneric _FBX_ROOT_DIR _OUT_FBX_LIBRARIES _OUT_FBX_LIBRA
   # Figures out matching compiler/os directory.
 
   if("x${CMAKE_CXX_COMPILER_ID}" STREQUAL "xMSVC")
-    if(NOT CMAKE_CXX_COMPILER_VERSION VERSION_LESS 19.20)
+    if(NOT CMAKE_CXX_COMPILER_VERSION VERSION_LESS 19.30)
+      set(FBX_CP_PATH "vs2022")
+    elseif(NOT CMAKE_CXX_COMPILER_VERSION VERSION_LESS 19.20)
       set(FBX_CP_PATH "vs2019")
     elseif(NOT CMAKE_CXX_COMPILER_VERSION VERSION_LESS 19.10)
       set(FBX_CP_PATH "vs2017")

test/animation/runtime/ik_aim_job_tests.cc

@@ -25,15 +25,13 @@
 //                                                                            //
 //----------------------------------------------------------------------------//
 
+#include "gtest/gtest.h"
 #include "ozz/animation/runtime/ik_aim_job.h"
-
+#include "ozz/base/maths/gtest_math_helper.h"
 #include "ozz/base/maths/quaternion.h"
 #include "ozz/base/maths/simd_math.h"
 #include "ozz/base/maths/simd_quaternion.h"
 
-#include "gtest/gtest.h"
-#include "ozz/base/maths/gtest_math_helper.h"
-
 TEST(JobValidity, IKAimJob) {
   const ozz::math::Float4x4 joint = ozz::math::Float4x4::identity();
   ozz::math::SimdQuaternion quat;
@@ -428,43 +426,24 @@ TEST(Twist, IKAimJob) {
     EXPECT_SIMDQUATERNION_EQ_TOL(quat, 0.f, 0.f, 0.f, 1.f, 2e-3f);
   }
 
-  {  // Pole y, twist pi
+  {  // Pole y, twist pi / 2
     job.pole_vector = ozz::math::simd_float4::y_axis();
-    job.twist_angle = ozz::math::kPi;
+    job.twist_angle = ozz::math::kPi_2;
     EXPECT_TRUE(job.Run());
     const ozz::math::Quaternion x_Pi = ozz::math::Quaternion::FromAxisAngle(
-        ozz::math::Float3::x_axis(), -ozz::math::kPi);
+        ozz::math::Float3::x_axis(), ozz::math::kPi_2);
     EXPECT_SIMDQUATERNION_EQ_TOL(quat, x_Pi.x, x_Pi.y, x_Pi.z, x_Pi.w, 2e-3f);
   }
 
-  {  // Pole y, twist -pi
+  {  // Pole y, twist -pi / 2
     job.pole_vector = ozz::math::simd_float4::y_axis();
-    job.twist_angle = -ozz::math::kPi;
+    job.twist_angle = -ozz::math::kPi_2;
     EXPECT_TRUE(job.Run());
     const ozz::math::Quaternion x_mPi = ozz::math::Quaternion::FromAxisAngle(
-        ozz::math::Float3::x_axis(), -ozz::math::kPi);
+        ozz::math::Float3::x_axis(), -ozz::math::kPi_2);
     EXPECT_SIMDQUATERNION_EQ_TOL(quat, x_mPi.x, x_mPi.y, x_mPi.z, x_mPi.w,
                                  2e-3f);
   }
-
-  {  // Pole y, twist pi/2
-    job.pole_vector = ozz::math::simd_float4::y_axis();
-    job.twist_angle = ozz::math::kPi_2;
-    EXPECT_TRUE(job.Run());
-    const ozz::math::Quaternion x_Pi_2 = ozz::math::Quaternion::FromAxisAngle(
-        ozz::math::Float3::x_axis(), ozz::math::kPi_2);
-    EXPECT_SIMDQUATERNION_EQ_TOL(quat, x_Pi_2.x, x_Pi_2.y, x_Pi_2.z, x_Pi_2.w,
-                                 2e-3f);
-  }
-
-  {  // Pole z, twist pi/2
-    job.pole_vector = ozz::math::simd_float4::z_axis();
-    job.twist_angle = ozz::math::kPi_2;
-    EXPECT_TRUE(job.Run());
-    const ozz::math::Quaternion x_Pi = ozz::math::Quaternion::FromAxisAngle(
-        ozz::math::Float3::x_axis(), ozz::math::kPi);
-    EXPECT_SIMDQUATERNION_EQ_TOL(quat, x_Pi.x, x_Pi.y, x_Pi.z, x_Pi.w, 2e-3f);
-  }
 }
 
 TEST(AlignedTargetUp, IKAimJob) {