diff --git a/intern/cycles/app/cycles_standalone.cpp b/intern/cycles/app/cycles_standalone.cpp
index 9871273381e..7ea1ca2d8fb 100644
--- a/intern/cycles/app/cycles_standalone.cpp
+++ b/intern/cycles/app/cycles_standalone.cpp
@@ -214,7 +214,7 @@ static void motion(int x, int y, int button)
 
 		/* Rotate */
 		else if(button == 2) {
-			float4 r1= make_float4((float)x * 0.1f, 0.0f, 1.0f, 0.0f);
+			float4 r1 = make_float4((float)x * 0.1f, 0.0f, 1.0f, 0.0f);
 			matrix = matrix * transform_rotate(DEG2RADF(r1.x), make_float3(r1.y, r1.z, r1.w));
 
 			float4 r2 = make_float4(y * 0.1f, 1.0f, 0.0f, 0.0f);
diff --git a/intern/cycles/app/cycles_xml.h b/intern/cycles/app/cycles_xml.h
index 2d225f3a013..96bc79c35d8 100644
--- a/intern/cycles/app/cycles_xml.h
+++ b/intern/cycles/app/cycles_xml.h
@@ -14,8 +14,8 @@
  * limitations under the License
  */
 
-#ifndef __CYCLES_XML__
-#define __CYCLES_XML__
+#ifndef __CYCLES_XML_H__
+#define __CYCLES_XML_H__
 
 CCL_NAMESPACE_BEGIN
 
@@ -29,5 +29,4 @@ void xml_read_file(Scene *scene, const char *filepath);
 
 CCL_NAMESPACE_END
 
-#endif /* __CYCLES_XML__ */
-
+#endif /* __CYCLES_XML_H__ */
diff --git a/intern/cycles/blender/CCL_api.h b/intern/cycles/blender/CCL_api.h
index 6532315cf39..2772b9ac8a7 100644
--- a/intern/cycles/blender/CCL_api.h
+++ b/intern/cycles/blender/CCL_api.h
@@ -14,8 +14,8 @@
  * limitations under the License
  */
 
-#ifndef CCL_API_H
-#define CCL_API_H
+#ifndef __CCL_API_H__
+#define __CCL_API_H__
 
 #ifdef __cplusplus
 extern "C" {
@@ -40,5 +40,4 @@ void *CCL_python_module_init(void);
 }
 #endif
 
-#endif /* CCL_API_H */
-
+#endif /* __CCL_API_H__ */
diff --git a/intern/cycles/blender/blender_mesh.cpp b/intern/cycles/blender/blender_mesh.cpp
index 29aecee09c2..fb343a74bd5 100644
--- a/intern/cycles/blender/blender_mesh.cpp
+++ b/intern/cycles/blender/blender_mesh.cpp
@@ -322,7 +322,8 @@ static void create_mesh(Scene *scene, Mesh *mesh, BL::Mesh b_mesh, const vector<
 		/* create triangles */
 		if(n == 4) {
 			if(is_zero(cross(mesh->verts[vi[1]] - mesh->verts[vi[0]], mesh->verts[vi[2]] - mesh->verts[vi[0]])) ||
-				is_zero(cross(mesh->verts[vi[2]] - mesh->verts[vi[0]], mesh->verts[vi[3]] - mesh->verts[vi[0]]))) {
+			   is_zero(cross(mesh->verts[vi[2]] - mesh->verts[vi[0]], mesh->verts[vi[3]] - mesh->verts[vi[0]])))
+			{
 				mesh->set_triangle(ti++, vi[0], vi[1], vi[3], shader, smooth);
 				mesh->set_triangle(ti++, vi[2], vi[3], vi[1], shader, smooth);
 			}
diff --git a/intern/cycles/blender/blender_python.cpp b/intern/cycles/blender/blender_python.cpp
index 658fccaf12e..872f891cc2a 100644
--- a/intern/cycles/blender/blender_python.cpp
+++ b/intern/cycles/blender/blender_python.cpp
@@ -317,7 +317,8 @@ static PyObject *osl_update_node_func(PyObject *self, PyObject *args)
 		}
 		else if(param->type.vecsemantics == TypeDesc::POINT ||
 		        param->type.vecsemantics == TypeDesc::VECTOR ||
-		        param->type.vecsemantics == TypeDesc::NORMAL) {
+		        param->type.vecsemantics == TypeDesc::NORMAL)
+		{
 			socket_type = "NodeSocketVector";
 			data_type = BL::NodeSocket::type_VECTOR;
 
diff --git a/intern/cycles/device/device_cuda.cpp b/intern/cycles/device/device_cuda.cpp
index 29b348d86ac..c667b8863d6 100644
--- a/intern/cycles/device/device_cuda.cpp
+++ b/intern/cycles/device/device_cuda.cpp
@@ -508,7 +508,7 @@ public:
 				if(interpolation == INTERPOLATION_CLOSEST) {
 					cuda_assert(cuTexRefSetFilterMode(texref, CU_TR_FILTER_MODE_POINT));
 				}
-				else if (interpolation == INTERPOLATION_LINEAR){
+				else if (interpolation == INTERPOLATION_LINEAR) {
 					cuda_assert(cuTexRefSetFilterMode(texref, CU_TR_FILTER_MODE_LINEAR));
 				}
 				else {/* CUBIC and SMART are unsupported for CUDA */
diff --git a/intern/cycles/device/device_network.cpp b/intern/cycles/device/device_network.cpp
index 8f00324c3e1..af051076009 100644
--- a/intern/cycles/device/device_network.cpp
+++ b/intern/cycles/device/device_network.cpp
@@ -326,7 +326,7 @@ class DeviceServer {
 public:
 	thread_mutex rpc_lock;
 
-	void network_error(const string &message){
+	void network_error(const string &message) {
 		error_func.network_error(message);
 	}
 
@@ -366,7 +366,7 @@ protected:
 	{
 		/* create a new DataVector and insert it into mem_data */
 		pair<DataMap::iterator,bool> data_ins = mem_data.insert(
-				DataMap::value_type(client_pointer, DataVector()));
+		        DataMap::value_type(client_pointer, DataVector()));
 
 		/* make sure it was a unique insertion */
 		assert(data_ins.second);
diff --git a/intern/cycles/device/device_network.h b/intern/cycles/device/device_network.h
index 8596b61b179..893841d1da7 100644
--- a/intern/cycles/device/device_network.h
+++ b/intern/cycles/device/device_network.h
@@ -209,7 +209,7 @@ public:
 		boost::system::error_code error;
 		size_t len = boost::asio::read(socket, boost::asio::buffer(header), error);
 
-		if(error.value()){
+		if(error.value()) {
 			error_func->network_error(error.message());
 		}
 
@@ -276,7 +276,7 @@ public:
 		boost::system::error_code error;
 		size_t len = boost::asio::read(socket, boost::asio::buffer(buffer, size), error);
 
-		if(error.value()){
+		if(error.value()) {
 			error_func->network_error(error.message());
 		}
 
@@ -391,7 +391,7 @@ private:
 
 					/* add address if it's not already in the list */
 					bool found = std::find(servers.begin(), servers.end(),
-							address) != servers.end();
+					                       address) != servers.end();
 
 					if(!found)
 						servers.push_back(address);
diff --git a/intern/cycles/device/device_opencl.cpp b/intern/cycles/device/device_opencl.cpp
index 2e759d1a36f..694ec9db036 100644
--- a/intern/cycles/device/device_opencl.cpp
+++ b/intern/cycles/device/device_opencl.cpp
@@ -461,7 +461,7 @@ public:
 		vector<cl_platform_id> platforms(num_platforms, NULL);
 
 		ciErr = clGetPlatformIDs(num_platforms, &platforms[0], NULL);
-		if(opencl_error(ciErr)){
+		if(opencl_error(ciErr)) {
 			fprintf(stderr, "clGetPlatformIDs failed \n");
 			return;
 		}
@@ -489,7 +489,7 @@ public:
 			/* get devices */
 			vector<cl_device_id> device_ids(num_devices, NULL);
 
-			if(opencl_error(clGetDeviceIDs(cpPlatform, opencl_device_type(), num_devices, &device_ids[0], NULL))){
+			if(opencl_error(clGetDeviceIDs(cpPlatform, opencl_device_type(), num_devices, &device_ids[0], NULL))) {
 				fprintf(stderr, "clGetDeviceIDs failed \n");
 				return;
 			}
@@ -528,7 +528,7 @@ public:
 				cxContext = clCreateContext(context_props, 1, &cdDevice,
 					context_notify_callback, cdDevice, &ciErr);
 
-				if(opencl_error(ciErr)){
+				if(opencl_error(ciErr)) {
 					opencl_error("OpenCL: clCreateContext failed");
 					return;
 				}
diff --git a/intern/cycles/kernel/closure/bsdf_hair.h b/intern/cycles/kernel/closure/bsdf_hair.h
index 26e1198ffcf..19cdb773255 100644
--- a/intern/cycles/kernel/closure/bsdf_hair.h
+++ b/intern/cycles/kernel/closure/bsdf_hair.h
@@ -84,7 +84,7 @@ ccl_device float3 bsdf_hair_reflection_eval_reflect(const ShaderClosure *sc, con
 	float theta_i = M_PI_2_F - safe_acosf(omega_in_z);
 	float cosphi_i = dot(omega_in_y, locy);
 
-	if(M_PI_2_F - fabsf(theta_i) < 0.001f || cosphi_i < 0.0f){
+	if(M_PI_2_F - fabsf(theta_i) < 0.001f || cosphi_i < 0.0f) {
 		*pdf = 0.0f;
 		return make_float3(*pdf, *pdf, *pdf);
 	}
@@ -140,7 +140,7 @@ ccl_device float3 bsdf_hair_transmission_eval_transmit(const ShaderClosure *sc,
 	float theta_i = M_PI_2_F - safe_acosf(omega_in_z);
 	float phi_i = safe_acosf(dot(omega_in_y, locy));
 
-	if(M_PI_2_F - fabsf(theta_i) < 0.001f){
+	if(M_PI_2_F - fabsf(theta_i) < 0.001f) {
 		*pdf = 0.0f;
 		return make_float3(*pdf, *pdf, *pdf);
 	}
@@ -251,8 +251,8 @@ ccl_device int bsdf_hair_transmission_sample(const ShaderClosure *sc, float3 Ng,
 	float phi_pdf = roughness2 / (c_TT * (p * p + roughness2 * roughness2));
 
 	*omega_in =(cosf(phi) * costheta_i) * locy -
-			   (sinf(phi) * costheta_i) * locx +
-			   (            sintheta_i) * Tg;
+	           (sinf(phi) * costheta_i) * locx +
+	           (            sintheta_i) * Tg;
 
 	//differentials - TODO: find a better approximation for the transmission bounce
 #ifdef __RAY_DIFFERENTIALS__
@@ -261,7 +261,7 @@ ccl_device int bsdf_hair_transmission_sample(const ShaderClosure *sc, float3 Ng,
 #endif
 
 	*pdf = fabsf(phi_pdf * theta_pdf);
-	if(M_PI_2_F - fabsf(theta_i) < 0.001f){
+	if(M_PI_2_F - fabsf(theta_i) < 0.001f) {
 		*pdf = 0.0f;
 	}
 
diff --git a/intern/cycles/kernel/closure/bsdf_microfacet.h b/intern/cycles/kernel/closure/bsdf_microfacet.h
index dfa8886c113..1ec35e444fe 100644
--- a/intern/cycles/kernel/closure/bsdf_microfacet.h
+++ b/intern/cycles/kernel/closure/bsdf_microfacet.h
@@ -154,8 +154,8 @@ ccl_device int bsdf_microfacet_ggx_sample(const ShaderClosure *sc, float3 Ng, fl
 		float sinThetaM  = cosThetaM * safe_sqrtf(tanThetaM2);
 		float phiM = M_2PI_F * randv;
 		float3 m = (cosf(phiM) * sinThetaM) * X +
-				 (sinf(phiM) * sinThetaM) * Y +
-							   cosThetaM  * Z;
+		           (sinf(phiM) * sinThetaM) * Y +
+		           (             cosThetaM) * Z;
 		if(!m_refractive) {
 			float cosMO = dot(m, I);
 			if(cosMO > 0) {
@@ -383,8 +383,8 @@ ccl_device int bsdf_microfacet_beckmann_sample(const ShaderClosure *sc, float3 N
 		float sinThetaM = cosThetaM * tanThetaM;
 		float phiM = M_2PI_F * randv;
 		float3 m = (cosf(phiM) * sinThetaM) * X +
-				 (sinf(phiM) * sinThetaM) * Y +
-							   cosThetaM  * Z;
+		           (sinf(phiM) * sinThetaM) * Y +
+		           (             cosThetaM) * Z;
 
 		if(!m_refractive) {
 			float cosMO = dot(m, I);
diff --git a/intern/cycles/kernel/closure/bsdf_phong_ramp.h b/intern/cycles/kernel/closure/bsdf_phong_ramp.h
index 219c5aea159..2b4e1c68640 100644
--- a/intern/cycles/kernel/closure/bsdf_phong_ramp.h
+++ b/intern/cycles/kernel/closure/bsdf_phong_ramp.h
@@ -109,8 +109,8 @@ ccl_device int bsdf_phong_ramp_sample(const ShaderClosure *sc, const float3 colo
 		float sinTheta2 = 1 - cosTheta * cosTheta;
 		float sinTheta = sinTheta2 > 0 ? sqrtf(sinTheta2) : 0;
 		*omega_in = (cosf(phi) * sinTheta) * T +
-				   (sinf(phi) * sinTheta) * B +
-				   (            cosTheta) * R;
+		            (sinf(phi) * sinTheta) * B +
+		            (            cosTheta) * R;
 		if (dot(Ng, *omega_in) > 0.0f)
 		{
 			// common terms for pdf and eval
diff --git a/intern/cycles/kernel/closure/bsdf_util.h b/intern/cycles/kernel/closure/bsdf_util.h
index f6dceb3ca82..b3dcb9dcc38 100644
--- a/intern/cycles/kernel/closure/bsdf_util.h
+++ b/intern/cycles/kernel/closure/bsdf_util.h
@@ -35,14 +35,15 @@
 
 CCL_NAMESPACE_BEGIN
 
-ccl_device float fresnel_dielectric(float eta, const float3 N,
-		const float3 I, float3 *R, float3 *T,
+ccl_device float fresnel_dielectric(
+        float eta, const float3 N,
+        const float3 I, float3 *R, float3 *T,
 #ifdef __RAY_DIFFERENTIALS__
-		const float3 dIdx, const float3 dIdy,
-		float3 *dRdx, float3 *dRdy,
-		float3 *dTdx, float3 *dTdy, 
+        const float3 dIdx, const float3 dIdy,
+        float3 *dRdx, float3 *dRdy,
+        float3 *dTdx, float3 *dTdy,
 #endif
-		bool *is_inside)
+        bool *is_inside)
 {
 	float cos = dot(N, I), neta;
 	float3 Nn;
diff --git a/intern/cycles/kernel/closure/bsdf_westin.h b/intern/cycles/kernel/closure/bsdf_westin.h
index ca4c05e91fe..9dc1c00bb3d 100644
--- a/intern/cycles/kernel/closure/bsdf_westin.h
+++ b/intern/cycles/kernel/closure/bsdf_westin.h
@@ -96,10 +96,9 @@ ccl_device int bsdf_westin_backscatter_sample(const ShaderClosure *sc, float3 Ng
 		float sinTheta2 = 1 - cosTheta * cosTheta;
 		float sinTheta = sinTheta2 > 0 ? sqrtf(sinTheta2) : 0;
 		*omega_in = (cosf(phi) * sinTheta) * T +
-				   (sinf(phi) * sinTheta) * B +
-				   (cosTheta) * I;
-		if(dot(Ng, *omega_in) > 0)
-		{
+		            (sinf(phi) * sinTheta) * B +
+		            (cosTheta) * I;
+		if(dot(Ng, *omega_in) > 0) {
 			// common terms for pdf and eval
 			float cosNI = dot(N, *omega_in);
 			// make sure the direction we chose is still in the right hemisphere
diff --git a/intern/cycles/kernel/geom/geom_bvh_shadow.h b/intern/cycles/kernel/geom/geom_bvh_shadow.h
index be9def02080..42ecd4bc54b 100644
--- a/intern/cycles/kernel/geom/geom_bvh_shadow.h
+++ b/intern/cycles/kernel/geom/geom_bvh_shadow.h
@@ -253,12 +253,13 @@ ccl_device bool BVH_FUNCTION_NAME
 							int shader = 0;
 
 #ifdef __HAIR__
-							if(kernel_tex_fetch(__prim_type, isect_array->prim) & PRIMITIVE_ALL_TRIANGLE) {
+							if(kernel_tex_fetch(__prim_type, isect_array->prim) & PRIMITIVE_ALL_TRIANGLE)
 #endif
+							{
 								float4 Ns = kernel_tex_fetch(__tri_normal, prim);
 								shader = __float_as_int(Ns.w);
-#ifdef __HAIR__
 							}
+#ifdef __HAIR__
 							else {
 								float4 str = kernel_tex_fetch(__curves, prim);
 								shader = __float_as_int(str.z);
diff --git a/intern/cycles/kernel/geom/geom_object.h b/intern/cycles/kernel/geom/geom_object.h
index 5b0127f8a3b..91edd5863ac 100644
--- a/intern/cycles/kernel/geom/geom_object.h
+++ b/intern/cycles/kernel/geom/geom_object.h
@@ -366,8 +366,8 @@ ccl_device_inline float3 bvh_clamp_direction(float3 dir)
 	/* clamp absolute values by exp2f(-80.0f) to avoid division by zero when calculating inverse direction */
 	float ooeps = 8.271806E-25f;
 	return make_float3((fabsf(dir.x) > ooeps)? dir.x: copysignf(ooeps, dir.x),
-					   (fabsf(dir.y) > ooeps)? dir.y: copysignf(ooeps, dir.y),
-					   (fabsf(dir.z) > ooeps)? dir.z: copysignf(ooeps, dir.z));
+	                   (fabsf(dir.y) > ooeps)? dir.y: copysignf(ooeps, dir.y),
+	                   (fabsf(dir.z) > ooeps)? dir.z: copysignf(ooeps, dir.z));
 }
 
 ccl_device_inline float3 bvh_inverse_direction(float3 dir)
diff --git a/intern/cycles/kernel/kernel_avx.cpp b/intern/cycles/kernel/kernel_avx.cpp
index d2a7142c551..354214c406e 100644
--- a/intern/cycles/kernel/kernel_avx.cpp
+++ b/intern/cycles/kernel/kernel_avx.cpp
@@ -77,6 +77,6 @@ CCL_NAMESPACE_END
 
 /* needed for some linkers in combination with scons making empty compilation unit in a library */
 void __dummy_function_cycles_avx(void);
-void __dummy_function_cycles_avx(void){}
+void __dummy_function_cycles_avx(void) {}
 
 #endif
diff --git a/intern/cycles/kernel/kernel_displace.h b/intern/cycles/kernel/kernel_displace.h
index a2709e711d2..6e14c8679a8 100644
--- a/intern/cycles/kernel/kernel_displace.h
+++ b/intern/cycles/kernel/kernel_displace.h
@@ -119,7 +119,7 @@ ccl_device void kernel_bake_evaluate(KernelGlobals *kg, ccl_global uint4 *input,
 	/* TODO, disable the closures we won't need */
 	shader_setup_from_sample(kg, &sd, P, Ng, I, shader, object, prim, u, v, t, time, bounce, transparent_bounce);
 
-	if(is_light_pass(type)){
+	if(is_light_pass(type)) {
 		RNG rng = cmj_hash(i, 0);
 		compute_light_pass(kg, &sd, &L, rng, (type == SHADER_EVAL_AO));
 	}
diff --git a/intern/cycles/kernel/kernel_emission.h b/intern/cycles/kernel/kernel_emission.h
index 44899e0bf97..deffa7f2ba2 100644
--- a/intern/cycles/kernel/kernel_emission.h
+++ b/intern/cycles/kernel/kernel_emission.h
@@ -166,10 +166,11 @@ ccl_device_noinline float3 indirect_primitive_emission(KernelGlobals *kg, Shader
 	float3 L = shader_emissive_eval(kg, sd);
 
 #ifdef __HAIR__
-	if(!(path_flag & PATH_RAY_MIS_SKIP) && (sd->flag & SD_USE_MIS) && (sd->type & PRIMITIVE_ALL_TRIANGLE)) {
+	if(!(path_flag & PATH_RAY_MIS_SKIP) && (sd->flag & SD_USE_MIS) && (sd->type & PRIMITIVE_ALL_TRIANGLE))
 #else
-	if(!(path_flag & PATH_RAY_MIS_SKIP) && (sd->flag & SD_USE_MIS)) {
+	if(!(path_flag & PATH_RAY_MIS_SKIP) && (sd->flag & SD_USE_MIS))
 #endif
+	{
 		/* multiple importance sampling, get triangle light pdf,
 		 * and compute weight with respect to BSDF pdf */
 		float pdf = triangle_light_pdf(kg, sd->Ng, sd->I, t);
diff --git a/intern/cycles/kernel/kernel_passes.h b/intern/cycles/kernel/kernel_passes.h
index 9cdcb8c5229..b3b6fc02894 100644
--- a/intern/cycles/kernel/kernel_passes.h
+++ b/intern/cycles/kernel/kernel_passes.h
@@ -51,7 +51,8 @@ ccl_device_inline void kernel_write_data_passes(KernelGlobals *kg, ccl_global fl
 	if(!(path_flag & PATH_RAY_SINGLE_PASS_DONE)) {
 		if(!(sd->flag & SD_TRANSPARENT) ||
 		   kernel_data.film.pass_alpha_threshold == 0.0f ||
-		   average(shader_bsdf_alpha(kg, sd)) >= kernel_data.film.pass_alpha_threshold) {
+		   average(shader_bsdf_alpha(kg, sd)) >= kernel_data.film.pass_alpha_threshold)
+		{
 
 			if(sample == 0) {
 				if(flag & PASS_DEPTH) {
diff --git a/intern/cycles/kernel/kernel_path.h b/intern/cycles/kernel/kernel_path.h
index e71b58aa709..bd7b53e3aec 100644
--- a/intern/cycles/kernel/kernel_path.h
+++ b/intern/cycles/kernel/kernel_path.h
@@ -1304,21 +1304,21 @@ ccl_device float4 kernel_branched_path_integrate(KernelGlobals *kg, RNG *rng, in
 				/* do subsurface scatter step with copy of shader data, this will
 				 * replace the BSSRDF with a diffuse BSDF closure */
 				for(int j = 0; j < num_samples; j++) {
-						ShaderData bssrdf_sd[BSSRDF_MAX_HITS];
-						float bssrdf_u, bssrdf_v;
-						path_branched_rng_2D(kg, &bssrdf_rng, &state, j, num_samples, PRNG_BSDF_U, &bssrdf_u, &bssrdf_v);
-						int num_hits = subsurface_scatter_multi_step(kg, &sd, bssrdf_sd, state.flag, sc, &lcg_state, bssrdf_u, bssrdf_v, true);
+					ShaderData bssrdf_sd[BSSRDF_MAX_HITS];
+					float bssrdf_u, bssrdf_v;
+					path_branched_rng_2D(kg, &bssrdf_rng, &state, j, num_samples, PRNG_BSDF_U, &bssrdf_u, &bssrdf_v);
+					int num_hits = subsurface_scatter_multi_step(kg, &sd, bssrdf_sd, state.flag, sc, &lcg_state, bssrdf_u, bssrdf_v, true);
 
-						/* compute lighting with the BSDF closure */
-						for(int hit = 0; hit < num_hits; hit++) {
-							PathState hit_state = state;
+					/* compute lighting with the BSDF closure */
+					for(int hit = 0; hit < num_hits; hit++) {
+						PathState hit_state = state;
 
-							path_state_branch(&hit_state, j, num_samples);
+						path_state_branch(&hit_state, j, num_samples);
 
-							kernel_branched_path_integrate_lighting(kg, rng,
-								&bssrdf_sd[hit], throughput, num_samples_inv,
-								&hit_state, &L, buffer);
-						}
+						kernel_branched_path_integrate_lighting(kg, rng,
+						                                        &bssrdf_sd[hit], throughput, num_samples_inv,
+						                                        &hit_state, &L, buffer);
+					}
 				}
 
 				state.flag &= ~PATH_RAY_BSSRDF_ANCESTOR;
diff --git a/intern/cycles/kernel/kernel_shader.h b/intern/cycles/kernel/kernel_shader.h
index 1351ab2e482..58cec090410 100644
--- a/intern/cycles/kernel/kernel_shader.h
+++ b/intern/cycles/kernel/kernel_shader.h
@@ -39,7 +39,7 @@ ccl_device void shader_setup_object_transforms(KernelGlobals *kg, ShaderData *sd
 {
 	if(sd->flag & SD_OBJECT_MOTION) {
 		sd->ob_tfm = object_fetch_transform_motion(kg, sd->object, time);
-		sd->ob_itfm= transform_quick_inverse(sd->ob_tfm);
+		sd->ob_itfm = transform_quick_inverse(sd->ob_tfm);
 	}
 	else {
 		sd->ob_tfm = object_fetch_transform(kg, sd->object, OBJECT_TRANSFORM);
diff --git a/intern/cycles/kernel/kernel_shadow.h b/intern/cycles/kernel/kernel_shadow.h
index b350e068177..ab7524c411a 100644
--- a/intern/cycles/kernel/kernel_shadow.h
+++ b/intern/cycles/kernel/kernel_shadow.h
@@ -190,7 +190,7 @@ ccl_device_inline bool shadow_blocked(KernelGlobals *kg, PathState *state, Ray *
 
 ccl_device_inline bool shadow_blocked(KernelGlobals *kg, PathState *state, Ray *ray, float3 *shadow)
 {
-*shadow = make_float3(1.0f, 1.0f, 1.0f);
+	*shadow = make_float3(1.0f, 1.0f, 1.0f);
 
 	if(ray->t == 0.0f)
 		return false;
@@ -217,10 +217,11 @@ ccl_device_inline bool shadow_blocked(KernelGlobals *kg, PathState *state, Ray *
 					return true;
 
 #ifdef __HAIR__
-				if(!scene_intersect(kg, ray, PATH_RAY_SHADOW_TRANSPARENT, &isect, NULL, 0.0f, 0.0f)) {
+				if(!scene_intersect(kg, ray, PATH_RAY_SHADOW_TRANSPARENT, &isect, NULL, 0.0f, 0.0f))
 #else
-				if(!scene_intersect(kg, ray, PATH_RAY_SHADOW_TRANSPARENT, &isect)) {
+				if(!scene_intersect(kg, ray, PATH_RAY_SHADOW_TRANSPARENT, &isect))
 #endif
+				{
 
 #ifdef __VOLUME__
 					/* attenuation for last line segment towards light */
diff --git a/intern/cycles/kernel/kernel_sse2.cpp b/intern/cycles/kernel/kernel_sse2.cpp
index 6a2a7804146..2d5f6091908 100644
--- a/intern/cycles/kernel/kernel_sse2.cpp
+++ b/intern/cycles/kernel/kernel_sse2.cpp
@@ -75,6 +75,6 @@ CCL_NAMESPACE_END
 
 /* needed for some linkers in combination with scons making empty compilation unit in a library */
 void __dummy_function_cycles_sse2(void);
-void __dummy_function_cycles_sse2(void){}
+void __dummy_function_cycles_sse2(void) {}
 
 #endif
diff --git a/intern/cycles/kernel/kernel_sse3.cpp b/intern/cycles/kernel/kernel_sse3.cpp
index 9d0abb93cc6..1062fd0c990 100644
--- a/intern/cycles/kernel/kernel_sse3.cpp
+++ b/intern/cycles/kernel/kernel_sse3.cpp
@@ -76,6 +76,6 @@ CCL_NAMESPACE_END
 
 /* needed for some linkers in combination with scons making empty compilation unit in a library */
 void __dummy_function_cycles_sse3(void);
-void __dummy_function_cycles_sse3(void){}
+void __dummy_function_cycles_sse3(void) {}
 
 #endif
diff --git a/intern/cycles/kernel/kernel_sse41.cpp b/intern/cycles/kernel/kernel_sse41.cpp
index bc20de0ec20..ba3b4887650 100644
--- a/intern/cycles/kernel/kernel_sse41.cpp
+++ b/intern/cycles/kernel/kernel_sse41.cpp
@@ -77,6 +77,6 @@ CCL_NAMESPACE_END
 
 /* needed for some linkers in combination with scons making empty compilation unit in a library */
 void __dummy_function_cycles_sse41(void);
-void __dummy_function_cycles_sse41(void){}
+void __dummy_function_cycles_sse41(void) {}
 
 #endif
diff --git a/intern/cycles/kernel/osl/osl_services.cpp b/intern/cycles/kernel/osl/osl_services.cpp
index 9861e527afd..2984cc97437 100644
--- a/intern/cycles/kernel/osl/osl_services.cpp
+++ b/intern/cycles/kernel/osl/osl_services.cpp
@@ -524,7 +524,8 @@ static bool get_mesh_element_attribute(KernelGlobals *kg, const ShaderData *sd,
                                const TypeDesc& type, bool derivatives, void *val)
 {
 	if (attr.type == TypeDesc::TypePoint || attr.type == TypeDesc::TypeVector ||
-	    attr.type == TypeDesc::TypeNormal || attr.type == TypeDesc::TypeColor) {
+	    attr.type == TypeDesc::TypeNormal || attr.type == TypeDesc::TypeColor)
+	{
 		float3 fval[3];
 		fval[0] = primitive_attribute_float3(kg, sd, attr.elem, attr.offset,
 		                                     (derivatives) ? &fval[1] : NULL, (derivatives) ? &fval[2] : NULL);
@@ -606,7 +607,7 @@ bool OSLRenderServices::get_object_standard_attribute(KernelGlobals *kg, ShaderD
 	}
 	else if (name == u_particle_lifetime) {
 		int particle_id = object_particle_id(kg, sd->object);
-		float f= particle_lifetime(kg, particle_id);
+		float f = particle_lifetime(kg, particle_id);
 		return set_attribute_float(f, type, derivatives, val);
 	}
 	else if (name == u_particle_location) {
@@ -643,10 +644,11 @@ bool OSLRenderServices::get_object_standard_attribute(KernelGlobals *kg, ShaderD
 	}
 	else if ((name == u_geom_trianglevertices || name == u_geom_polyvertices)
 #ifdef __HAIR__
-		     && sd->type & PRIMITIVE_ALL_TRIANGLE) {
+		     && sd->type & PRIMITIVE_ALL_TRIANGLE)
 #else
-		) {
+		)
 #endif
+	{
 		float3 P[3];
 
 		if(sd->type & PRIMITIVE_TRIANGLE)
diff --git a/intern/cycles/kernel/osl/osl_services.h b/intern/cycles/kernel/osl/osl_services.h
index c3ea12253e9..54ddd7a98b4 100644
--- a/intern/cycles/kernel/osl/osl_services.h
+++ b/intern/cycles/kernel/osl/osl_services.h
@@ -110,9 +110,9 @@ public:
 	                      ustring dataname, TypeDesc datatype, void *data);
 
 	static bool get_background_attribute(KernelGlobals *kg, ShaderData *sd, ustring name,
-			TypeDesc type, bool derivatives, void *val);
+	                                     TypeDesc type, bool derivatives, void *val);
 	static bool get_object_standard_attribute(KernelGlobals *kg, ShaderData *sd, ustring name,
-			TypeDesc type, bool derivatives, void *val);
+	                                          TypeDesc type, bool derivatives, void *val);
 
 	static ustring u_distance;
 	static ustring u_index;
diff --git a/intern/cycles/kernel/osl/osl_shader.cpp b/intern/cycles/kernel/osl/osl_shader.cpp
index 64c734dec64..843dcdd0985 100644
--- a/intern/cycles/kernel/osl/osl_shader.cpp
+++ b/intern/cycles/kernel/osl/osl_shader.cpp
@@ -409,8 +409,9 @@ static void flatten_volume_closure_tree(ShaderData *sd,
 					sc.data1 = volume->sc.data1;
 
 					/* add */
-					if(sc.sample_weight > CLOSURE_WEIGHT_CUTOFF &&
-					   sd->num_closure < MAX_CLOSURE) {
+					if((sc.sample_weight > CLOSURE_WEIGHT_CUTOFF) &&
+					   (sd->num_closure < MAX_CLOSURE))
+					{
 						sd->closure[sd->num_closure++] = sc;
 						sd->flag |= volume->shaderdata_flag();
 					}
diff --git a/intern/cycles/kernel/svm/svm_geometry.h b/intern/cycles/kernel/svm/svm_geometry.h
index 0ec61fdebf0..fe681ec92af 100644
--- a/intern/cycles/kernel/svm/svm_geometry.h
+++ b/intern/cycles/kernel/svm/svm_geometry.h
@@ -117,13 +117,13 @@ ccl_device void svm_node_particle_info(KernelGlobals *kg, ShaderData *sd, float
 			stack_store_float3(stack, out_offset, particle_location(kg, particle_id));
 			break;
 		}
-		#if 0	/* XXX float4 currently not supported in SVM stack */
+#if 0	/* XXX float4 currently not supported in SVM stack */
 		case NODE_INFO_PAR_ROTATION: {
 			int particle_id = object_particle_id(kg, sd->object);
 			stack_store_float4(stack, out_offset, particle_rotation(kg, particle_id));
 			break;
 		}
-		#endif
+#endif
 		case NODE_INFO_PAR_SIZE: {
 			int particle_id = object_particle_id(kg, sd->object);
 			stack_store_float(stack, out_offset, particle_size(kg, particle_id));
diff --git a/intern/cycles/kernel/svm/svm_image.h b/intern/cycles/kernel/svm/svm_image.h
index ce2893e1893..6627786725f 100644
--- a/intern/cycles/kernel/svm/svm_image.h
+++ b/intern/cycles/kernel/svm/svm_image.h
@@ -65,7 +65,7 @@ ccl_device float4 svm_image_texture(KernelGlobals *kg, int id, float x, float y,
 
 	float4 r;
 	int ix, iy, nix, niy;
-	if (interpolation == INTERPOLATION_CLOSEST){
+	if (interpolation == INTERPOLATION_CLOSEST) {
 		svm_image_texture_frac(x*width, &ix);
 		svm_image_texture_frac(y*height, &iy);
 
diff --git a/intern/cycles/kernel/svm/svm_sepcomb_hsv.h b/intern/cycles/kernel/svm/svm_sepcomb_hsv.h
index 0f68ecbea03..111d5d47988 100644
--- a/intern/cycles/kernel/svm/svm_sepcomb_hsv.h
+++ b/intern/cycles/kernel/svm/svm_sepcomb_hsv.h
@@ -42,12 +42,12 @@ ccl_device void svm_node_separate_hsv(KernelGlobals *kg, ShaderData *sd, float *
 	/* Convert to HSV */
 	color = rgb_to_hsv(color);
 
-	if (stack_valid(hue_out)) 
-			stack_store_float(stack, hue_out, color.x);
-	if (stack_valid(saturation_out)) 
-			stack_store_float(stack, saturation_out, color.y);
-	if (stack_valid(value_out)) 
-			stack_store_float(stack, value_out, color.z);
+	if (stack_valid(hue_out))
+		stack_store_float(stack, hue_out, color.x);
+	if (stack_valid(saturation_out))
+		stack_store_float(stack, saturation_out, color.y);
+	if (stack_valid(value_out))
+		stack_store_float(stack, value_out, color.z);
 }
 
 CCL_NAMESPACE_END
diff --git a/intern/cycles/kernel/svm/svm_vector_transform.h b/intern/cycles/kernel/svm/svm_vector_transform.h
index 890a9ddda69..61d33aeb8cf 100644
--- a/intern/cycles/kernel/svm/svm_vector_transform.h
+++ b/intern/cycles/kernel/svm/svm_vector_transform.h
@@ -91,9 +91,9 @@ ccl_device void svm_node_vector_transform(KernelGlobals *kg, ShaderData *sd, flo
 	if(type == NODE_VECTOR_TRANSFORM_TYPE_NORMAL)
 		in = normalize(in);
 	
-	/* Output */	
+	/* Output */
 	if(stack_valid(vector_out)) {
-			stack_store_float3(stack, vector_out, in);
+		stack_store_float3(stack, vector_out, in);
 	}
 }
 
diff --git a/intern/cycles/kernel/svm/svm_wireframe.h b/intern/cycles/kernel/svm/svm_wireframe.h
index 7262a7904bb..660e6e2ca47 100644
--- a/intern/cycles/kernel/svm/svm_wireframe.h
+++ b/intern/cycles/kernel/svm/svm_wireframe.h
@@ -45,10 +45,11 @@ ccl_device void svm_node_wireframe(KernelGlobals *kg, ShaderData *sd, float *sta
 
 	/* Calculate wireframe */
 #ifdef __HAIR__
-	if (sd->prim != PRIM_NONE && sd->type & PRIMITIVE_ALL_TRIANGLE) {
+	if (sd->prim != PRIM_NONE && sd->type & PRIMITIVE_ALL_TRIANGLE)
 #else
-	if (sd->prim != PRIM_NONE) {
+	if (sd->prim != PRIM_NONE)
 #endif
+	{
 		float3 Co[3];
 		float pixelwidth = 1.0f;
 
diff --git a/intern/cycles/render/curves.cpp b/intern/cycles/render/curves.cpp
index 6e6b11ca92f..2c96ffa655e 100644
--- a/intern/cycles/render/curves.cpp
+++ b/intern/cycles/render/curves.cpp
@@ -110,7 +110,7 @@ void CurveSystemManager::device_update(Device *device, DeviceScene *dscene, Scen
 
 	progress.set_status("Updating Hair settings", "Copying Hair settings to device");
 
-	KernelCurves *kcurve= &dscene->data.curve;
+	KernelCurves *kcurve = &dscene->data.curve;
 
 	kcurve->curveflags = 0;
 
diff --git a/intern/cycles/render/mesh.cpp b/intern/cycles/render/mesh.cpp
index ddcc42a9ae9..9607892ceb8 100644
--- a/intern/cycles/render/mesh.cpp
+++ b/intern/cycles/render/mesh.cpp
@@ -554,7 +554,7 @@ bool Mesh::has_motion_blur() const
 {
 	return (use_motion_blur &&
 	        (attributes.find(ATTR_STD_MOTION_VERTEX_POSITION) ||
-			 curve_attributes.find(ATTR_STD_MOTION_VERTEX_POSITION)));
+	         curve_attributes.find(ATTR_STD_MOTION_VERTEX_POSITION)));
 }
 
 /* Mesh Manager */
diff --git a/intern/cycles/render/nodes.cpp b/intern/cycles/render/nodes.cpp
index 8748294acb0..28d2a41a113 100644
--- a/intern/cycles/render/nodes.cpp
+++ b/intern/cycles/render/nodes.cpp
@@ -317,7 +317,7 @@ void ImageTextureNode::compile(OSLCompiler& compiler)
 	compiler.parameter("is_float", is_float);
 	compiler.parameter("use_alpha", !alpha_out->links.empty());
 
-	switch (interpolation){
+	switch (interpolation) {
 		case INTERPOLATION_CLOSEST:
 			compiler.parameter("interpolation", "closest");
 			break;
diff --git a/intern/cycles/render/svm.cpp b/intern/cycles/render/svm.cpp
index db80aad74d4..576c176759c 100644
--- a/intern/cycles/render/svm.cpp
+++ b/intern/cycles/render/svm.cpp
@@ -226,7 +226,8 @@ void SVMCompiler::stack_assign(ShaderInput *input)
 			else if(input->type == SHADER_SOCKET_VECTOR ||
 			        input->type == SHADER_SOCKET_NORMAL ||
 			        input->type == SHADER_SOCKET_POINT ||
-			        input->type == SHADER_SOCKET_COLOR) {
+			        input->type == SHADER_SOCKET_COLOR)
+			{
 
 				add_node(NODE_VALUE_V, input->stack_offset);
 				add_node(NODE_VALUE_V, input->value);
@@ -506,8 +507,8 @@ void SVMCompiler::generate_multi_closure(ShaderNode *node, set<ShaderNode*>& don
 			find_dependencies(cl2deps, done, cl2in);
 
 			set_intersection(cl1deps.begin(), cl1deps.end(),
-							 cl2deps.begin(), cl2deps.end(),
-							 std::inserter(shareddeps, shareddeps.begin()));
+			                 cl2deps.begin(), cl2deps.end(),
+			                 std::inserter(shareddeps, shareddeps.begin()));
 			
 			if(!shareddeps.empty()) {
 				if(cl1in->link)
diff --git a/intern/cycles/util/util_color.h b/intern/cycles/util/util_color.h
index 0cfa4049d3e..50d9f59ea27 100644
--- a/intern/cycles/util/util_color.h
+++ b/intern/cycles/util/util_color.h
@@ -132,8 +132,8 @@ ccl_device float3 xyY_to_xyz(float x, float y, float Y)
 ccl_device float3 xyz_to_rgb(float x, float y, float z)
 {
 	return make_float3(3.240479f * x + -1.537150f * y + -0.498535f * z,
-					  -0.969256f * x +  1.875991f * y +  0.041556f * z,
-					   0.055648f * x + -0.204043f * y +  1.057311f * z);
+	                  -0.969256f * x +  1.875991f * y +  0.041556f * z,
+	                   0.055648f * x + -0.204043f * y +  1.057311f * z);
 }
 
 #ifndef __KERNEL_OPENCL__
diff --git a/intern/cycles/util/util_hash.h b/intern/cycles/util/util_hash.h
index ded25c92b90..edd2448efa4 100644
--- a/intern/cycles/util/util_hash.h
+++ b/intern/cycles/util/util_hash.h
@@ -23,7 +23,7 @@ CCL_NAMESPACE_BEGIN
 
 static inline uint hash_int_2d(uint kx, uint ky)
 {
-	#define rot(x,k) (((x)<<(k)) | ((x)>>(32-(k))))
+#define rot(x,k) (((x)<<(k)) | ((x)>>(32-(k))))
 
 	uint a, b, c;
 
@@ -41,7 +41,7 @@ static inline uint hash_int_2d(uint kx, uint ky)
 
 	return c;
 
-	#undef rot
+#undef rot
 }
 
 static inline uint hash_int(uint k)
diff --git a/intern/cycles/util/util_opencl.h b/intern/cycles/util/util_opencl.h
index 5f3f1667bcc..141c5e38273 100644
--- a/intern/cycles/util/util_opencl.h
+++ b/intern/cycles/util/util_opencl.h
@@ -304,7 +304,9 @@ typedef struct _cl_kernel *         cl_kernel;
 typedef struct _cl_event *          cl_event;
 typedef struct _cl_sampler *        cl_sampler;
 
-typedef cl_uint             cl_bool;                     /* WARNING!  Unlike cl_ types in cl_platform.h, cl_bool is not guaranteed to be the same size as the bool in kernels. */ 
+/* WARNING!  Unlike cl_ types in cl_platform.h,
+ * cl_bool is not guaranteed to be the same size as the bool in kernels. */
+typedef cl_uint             cl_bool;
 typedef cl_ulong            cl_bitfield;
 typedef cl_bitfield         cl_device_type;
 typedef cl_uint             cl_platform_info;
diff --git a/intern/cycles/util/util_simd.h b/intern/cycles/util/util_simd.h
index 679556ee59b..f0f37fa57aa 100644
--- a/intern/cycles/util/util_simd.h
+++ b/intern/cycles/util/util_simd.h
@@ -71,7 +71,7 @@ ccl_device_inline const __m128 shuffle_swap(const __m128& a, shuffle_swap_t shuf
 
 #ifdef __KERNEL_SSE41__
 ccl_device_inline void gen_idirsplat_swap(const __m128 &pn, const shuffle_swap_t &shuf_identity, const shuffle_swap_t &shuf_swap,
-										  const float3& idir, __m128 idirsplat[3], shuffle_swap_t shufflexyz[3])
+                                          const float3& idir, __m128 idirsplat[3], shuffle_swap_t shufflexyz[3])
 {
 	const __m128 idirsplat_raw[] = { _mm_set_ps1(idir.x), _mm_set_ps1(idir.y), _mm_set_ps1(idir.z) };
 	idirsplat[0] = _mm_xor_ps(idirsplat_raw[0], pn);
@@ -87,7 +87,7 @@ ccl_device_inline void gen_idirsplat_swap(const __m128 &pn, const shuffle_swap_t
 }
 #else
 ccl_device_inline void gen_idirsplat_swap(const __m128 &pn, const shuffle_swap_t &shuf_identity, const shuffle_swap_t &shuf_swap,
-										  const float3& idir, __m128 idirsplat[3], shuffle_swap_t shufflexyz[3])
+                                          const float3& idir, __m128 idirsplat[3], shuffle_swap_t shufflexyz[3])
 {
 	idirsplat[0] = _mm_xor_ps(_mm_set_ps1(idir.x), pn);
 	idirsplat[1] = _mm_xor_ps(_mm_set_ps1(idir.y), pn);
diff --git a/intern/cycles/util/util_system.cpp b/intern/cycles/util/util_system.cpp
index 9a7f1427cbe..0764f7d9345 100644
--- a/intern/cycles/util/util_system.cpp
+++ b/intern/cycles/util/util_system.cpp
@@ -169,14 +169,15 @@ static CPUCapabilities& system_cpu_capabilities()
 			if( os_uses_xsave_xrestore && cpu_avx_support) {
 				// Check if the OS will save the YMM registers
 				uint32_t xcr_feature_mask;
-				#if defined(__GNUC__)
-					int edx; // not used
-					__asm__ (".byte 0x0f, 0x01, 0xd0" : "=a" (xcr_feature_mask) , "=d" (edx) : "c" (0) ); /* actual opcode for xgetbv */
-				#elif defined(_MSC_VER) && defined(_XCR_XFEATURE_ENABLED_MASK)
-					xcr_feature_mask = (uint32_t)_xgetbv(_XCR_XFEATURE_ENABLED_MASK);  /* min VS2010 SP1 compiler is required */
-				#else
-					xcr_feature_mask = 0;
-				#endif
+#if defined(__GNUC__)
+				int edx; /* not used */
+				/* actual opcode for xgetbv */
+				__asm__ (".byte 0x0f, 0x01, 0xd0" : "=a" (xcr_feature_mask) , "=d" (edx) : "c" (0) );
+#elif defined(_MSC_VER) && defined(_XCR_XFEATURE_ENABLED_MASK)
+				xcr_feature_mask = (uint32_t)_xgetbv(_XCR_XFEATURE_ENABLED_MASK);  /* min VS2010 SP1 compiler is required */
+#else
+				xcr_feature_mask = 0;
+#endif
 				caps.avx = (xcr_feature_mask & 0x6) == 0x6;
 			}
 		}
diff --git a/intern/cycles/util/util_transform.h b/intern/cycles/util/util_transform.h
index 4c7ce12d1de..5b3dbe42f69 100644
--- a/intern/cycles/util/util_transform.h
+++ b/intern/cycles/util/util_transform.h
@@ -108,9 +108,9 @@ ccl_device_inline Transform transform_transpose(const Transform a)
 }
 
 ccl_device_inline Transform make_transform(float a, float b, float c, float d,
-									float e, float f, float g, float h,
-									float i, float j, float k, float l,
-									float m, float n, float o, float p)
+                                           float e, float f, float g, float h,
+                                           float i, float j, float k, float l,
+                                           float m, float n, float o, float p)
 {
 	Transform t;
 
diff --git a/intern/cycles/util/util_types.h b/intern/cycles/util/util_types.h
index 5d2037960fc..8ac0b11678e 100644
--- a/intern/cycles/util/util_types.h
+++ b/intern/cycles/util/util_types.h
@@ -159,8 +159,8 @@ struct int2 {
 	__forceinline int& operator[](int i) { return *(&x + i); }
 };
 
-#ifdef __KERNEL_SSE__
 struct ccl_try_align(16) int3 {
+#ifdef __KERNEL_SSE__
 	union {
 		__m128i m128;
 		struct { int x, y, z, w; };
@@ -171,7 +171,6 @@ struct ccl_try_align(16) int3 {
 	__forceinline operator const __m128i&(void) const { return m128; }
 	__forceinline operator __m128i&(void) { return m128; }
 #else
-struct ccl_try_align(16) int3 {
 	int x, y, z, w;
 #endif
 
@@ -179,8 +178,8 @@ struct ccl_try_align(16) int3 {
 	__forceinline int& operator[](int i) { return *(&x + i); }
 };
 
-#ifdef __KERNEL_SSE__
 struct ccl_try_align(16) int4 {
+#ifdef __KERNEL_SSE__
 	union {
 		__m128i m128;
 		struct { int x, y, z, w; };
@@ -191,7 +190,6 @@ struct ccl_try_align(16) int4 {
 	__forceinline operator const __m128i&(void) const { return m128; }
 	__forceinline operator __m128i&(void) { return m128; }
 #else
-struct ccl_try_align(16) int4 {
 	int x, y, z, w;
 #endif
 
@@ -227,8 +225,8 @@ struct float2 {
 	__forceinline float& operator[](int i) { return *(&x + i); }
 };
 
-#ifdef __KERNEL_SSE__
 struct ccl_try_align(16) float3 {
+#ifdef __KERNEL_SSE__
 	union {
 		__m128 m128;
 		struct { float x, y, z, w; };
@@ -239,7 +237,6 @@ struct ccl_try_align(16) float3 {
 	__forceinline operator const __m128&(void) const { return m128; }
 	__forceinline operator __m128&(void) { return m128; }
 #else
-struct ccl_try_align(16) float3 {
 	float x, y, z, w;
 #endif
 
@@ -247,8 +244,8 @@ struct ccl_try_align(16) float3 {
 	__forceinline float& operator[](int i) { return *(&x + i); }
 };
 
-#ifdef __KERNEL_SSE__
 struct ccl_try_align(16) float4 {
+#ifdef __KERNEL_SSE__
 	union {
 		__m128 m128;
 		struct { float x, y, z, w; };
@@ -259,7 +256,6 @@ struct ccl_try_align(16) float4 {
 	__forceinline operator const __m128&(void) const { return m128; }
 	__forceinline operator __m128&(void) { return m128; }
 #else
-struct ccl_try_align(16) float4 {
 	float x, y, z, w;
 #endif