compacted skinning bone buffer

WickedEngine/skinningCS.hlsl

@@ -3,7 +3,9 @@
 
 struct Bone
 {
-	float4x4 pose;
+	float4 pose0;
+	float4 pose1;
+	float4 pose2;
 };
 STRUCTUREDBUFFER(boneBuffer, Bone, SKINNINGSLOT_IN_BONEBUFFER);
 
@@ -30,7 +32,15 @@ inline void Skinning(inout float4 pos, inout float4 nor, in float4 inBon, in flo
 	[loop]
 	for (uint i = 0; ((i < 4) && (weisum<1.0f)); ++i)
 	{
-		m = boneBuffer[(uint)inBon[i]].pose;
+		float4 pose0 = boneBuffer[(uint)inBon[i]].pose0;
+		float4 pose1 = boneBuffer[(uint)inBon[i]].pose1;
+		float4 pose2 = boneBuffer[(uint)inBon[i]].pose2;
+		m = float4x4(
+			float4(pose0.x, pose1.x, pose2.x, 0),
+			float4(pose0.y, pose1.y, pose2.y, 0),
+			float4(pose0.z, pose1.z, pose2.z, 0),
+			float4(pose0.w, pose1.w, pose2.w, 1)
+			);
 		m3 = (float3x3)m;
 
 		p += mul(float4(pos.xyz, 1), m)*inWei[i];

WickedEngine/wiLoader.h

@@ -718,7 +718,16 @@ public:
 
 	GFX_STRUCT ShaderBoneType
 	{
-		XMMATRIX pose;
+		XMFLOAT4A pose0;
+		XMFLOAT4A pose1;
+		XMFLOAT4A pose2; 
+		
+		void Create(const XMFLOAT4X4& matIn)
+		{
+			pose0 = XMFLOAT4A(matIn._11, matIn._21, matIn._31, matIn._41);
+			pose1 = XMFLOAT4A(matIn._12, matIn._22, matIn._32, matIn._42);
+			pose2 = XMFLOAT4A(matIn._13, matIn._23, matIn._33, matIn._43);
+		}
 
 		STRUCTUREDBUFFER_SETBINDSLOT(SKINNINGSLOT_IN_BONEBUFFER)
 

WickedEngine/wiRenderer.cpp

@@ -1827,7 +1827,7 @@ void wiRenderer::UpdateRenderData(GRAPHICSTHREAD threadID)
 					for (unsigned int k = 0; k < armature->boneCollection.size(); k++) 
 					{
 						// Note the transpose: we NEED to transpose so that loading from the structured buffer is easier in the shader (avoid the many mov operations)
-						armature->boneData[k].pose = XMMatrixTranspose(XMLoadFloat4x4(&armature->boneCollection[k]->boneRelativity));
+						armature->boneData[k].Create(armature->boneCollection[k]->boneRelativity);
 					}
 					GetDevice()->UpdateBuffer(&armature->boneBuffer, armature->boneData.data(), threadID, (int)(sizeof(Armature::ShaderBoneType) * armature->boneCollection.size()));
 					GetDevice()->BindResourceCS(&armature->boneBuffer, STRUCTUREDBUFFER_GETBINDSLOT(Armature::ShaderBoneType), threadID);

WickedEngine/wiVersion.cpp

@@ -9,7 +9,7 @@ namespace wiVersion
 	// minor features, major updates
 	const int minor = 13;
 	// minor bug fixes, alterations, refactors, updates
-	const int revision = 13;
+	const int revision = 14;
 
 
 	long GetVersion()