Fixed ARM-NEON GCC conformance issue with half-precision optimization…

…s on ARM64
microsoft · Jan 11, 2021 · 5a5d2bd · 5a5d2bd · walbourn · Jan 11, 2021
1 parent 5d15267
commit 5a5d2bd
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 9 deletions.
diff --git a/Inc/DirectXPackedVector.inl b/Inc/DirectXPackedVector.inl
@@ -644,13 +644,13 @@ inline HALF* XMConvertFloatToHalfStream
 
                     uint16x4_t vHalf = vreinterpret_u16_f16(vcvt_f16_f32(vFloat));
 
-                    vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 0);
+                    vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 0);
                     pHalf += OutputStride;
-                    vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 1);
+                    vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 1);
                     pHalf += OutputStride;
-                    vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 2);
+                    vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 2);
                     pHalf += OutputStride;
-                    vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 3);
+                    vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 3);
                     pHalf += OutputStride;
                     i += 4;
                 }
@@ -701,13 +701,13 @@ inline HALF* XMConvertFloatToHalfStream
 
                 uint16x4_t vHalf = vreinterpret_u16_f16(vcvt_f16_f32(vFloat));
 
-                vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 0);
+                vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 0);
                 pHalf += OutputStride;
-                vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 1);
+                vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 1);
                 pHalf += OutputStride;
-                vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 2);
+                vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 2);
                 pHalf += OutputStride;
-                vst1_lane_u16(reinterpret_cast<float*>(pHalf), vHalf, 3);
+                vst1_lane_u16(reinterpret_cast<uint16_t*>(pHalf), vHalf, 3);
                 pHalf += OutputStride;
                 i += 4;
             }

diff --git a/LICENSE b/LICENSE
@@ -1,6 +1,6 @@
                                The MIT License (MIT)
 
-Copyright (c) 2011-2020 Microsoft Corp
+Copyright (c) 2011-2021 Microsoft Corp
 
 Permission is hereby granted, free of charge, to any person obtaining a copy of this 
 software and associated documentation files (the "Software"), to deal in the Software