shader cleanup; minor bugfixes

data/pipelines/atmo.hlsl

@@ -14,10 +14,10 @@ cbuffer Data : register (b4) {
 	float u_fog_top;
 	float u_fog_enabled;
 	float u_godrays_enabled;
-	uint u_output;
-	uint u_optical_depth;
-	uint u_depth_buffer;
-	uint u_inscatter;
+	RWTextureHandle u_output;
+	TextureHandle u_optical_depth;
+	TextureHandle u_depth_buffer;
+	TextureHandle u_inscatter;
 };
 
 // mie - Schlick appoximation phase function of Henyey-Greenstein

data/pipelines/atmo_optical_depth.hlsl

@@ -14,7 +14,7 @@ cbuffer Data :register(b4) {
 	float u_fog_top;
 	float u_fog_enabled;
 	float u_godarys_enabled;
-	uint u_output;
+	RWTextureHandle u_output;
 };
 
 float3 getTopAtmo(float3 p, float3 dir) {

data/pipelines/atmo_scattering.hlsl

@@ -16,8 +16,8 @@ cbuffer Data : register(b4) {
 	float u_fog_top;
 	float u_fog_enabled;
 	float u_godarys_enabled;
-	uint u_inscatter;
-	uint u_optical_depth;
+	RWTextureHandle u_inscatter;
+	TextureHandle u_optical_depth;
 };
 
 float getOptDepthY(float3 position) {

data/pipelines/avg_luminance.hlsl

@@ -1,10 +1,9 @@
 #include "pipelines/common.hlsli"
 
 cbuffer Data : register (b4) {
-	float2 u_size;
 	float u_accomodation_speed;
-	uint u_image;
-	uint b_histogram;
+	TextureHandle u_image;
+	uint u_histogram;
 };
 
 groupshared uint histogram[256];
@@ -30,27 +29,27 @@ void main(uint3 local_thread_id : SV_GroupThreadID, uint3 thread_id : SV_Dispatc
 	const float range = 12.0;
 	const float min = -6;
 	#ifdef PASS0
-		bindless_rw_buffers[b_histogram].Store(local_thread_id.x * 4, 0);
+		bindless_rw_buffers[u_histogram].Store(local_thread_id.x * 4, 0);
 	#elif !defined PASS2
 		uint idx = local_thread_id.x + local_thread_id.y * 16;
 		histogram[idx] = 0;
 
 		AllMemoryBarrier();
 		GroupMemoryBarrierWithGroupSync();
 
-		if (all(thread_id.xy < uint2(u_size + 0.5))) {
-			float _luminance = luminance(sampleBindlessLod(LinearSamplerClamp, u_image, int2(thread_id.xy) / u_size, 0).rgb);
+		if (all(thread_id.xy < uint2(Global_framebuffer_size + 0.5))) {
+			float _luminance = luminance(sampleBindlessLod(LinearSamplerClamp, u_image, thread_id.xy * Global_rcp_framebuffer_size, 0).rgb);
 			uint bin = luminanceToBin(_luminance, min, 1 / range);
 			InterlockedAdd(histogram[bin], 1);
 		}
 		AllMemoryBarrier();
 		GroupMemoryBarrierWithGroupSync();
 
 		uint dummy;
-		bindless_rw_buffers[b_histogram].InterlockedAdd(idx * 4, histogram[idx], dummy);
+		bindless_rw_buffers[u_histogram].InterlockedAdd(idx * 4, histogram[idx], dummy);
 	#else
 		uint idx = local_thread_id.x;
-		histogram[idx] = bindless_rw_buffers[b_histogram].Load(idx * 4) * idx;
+		histogram[idx] = bindless_rw_buffers[u_histogram].Load(idx * 4) * idx;
 
 		AllMemoryBarrier();
 		GroupMemoryBarrierWithGroupSync();
@@ -65,12 +64,12 @@ void main(uint3 local_thread_id : SV_GroupThreadID, uint3 thread_id : SV_Dispatc
 
 		if (idx == 0) {
 			uint sum = histogram[0];
-			float avg_bin = sum / float(u_size.x * u_size.y);
+			float avg_bin = sum / float(Global_framebuffer_size.x * Global_framebuffer_size.y);
 			float avg_lum = binToLuminance(avg_bin, min, range);
-			float accumulator = asfloat(bindless_rw_buffers[b_histogram].Load(256 * 4));
+			float accumulator = asfloat(bindless_rw_buffers[u_histogram].Load(256 * 4));
 			accumulator += (avg_lum - accumulator) * (1 - exp(-Global_frame_time_delta * u_accomodation_speed));
 
-			bindless_rw_buffers[b_histogram].Store(256 * 4, asuint(accumulator));
+			bindless_rw_buffers[u_histogram].Store(256 * 4, asuint(accumulator));
 		}
 	#endif
 }

data/pipelines/blit.hlsl

@@ -8,8 +8,8 @@ cbuffer Drawcall : register(b4) {
 	float4 u_offsets;
 	uint2 u_position;
 	int2 u_scale;
-	uint u_input;
-	uint u_output;
+	TextureHandle u_input;
+	RWTextureHandle u_output;
 };
 
 [numthreads(16, 16, 1)]

data/pipelines/bloom.hlsl

@@ -1,51 +1,14 @@
 #include "pipelines/common.hlsli"
 
-#ifdef EXTRACT
-	cbuffer Data : register(b4) {
-		float u_avg_lum_multiplier;
-		uint b_histogram;
-		uint u_input;
-		uint u_output;
-	};
-
-	[numthreads(16, 16, 1)]
-	void main(uint3 thread_id : SV_DispatchThreadID) {
-		float avg_lum = asfloat(bindless_buffers[b_histogram].Load(256 * 4));
-		float3 c = bindless_textures[u_input][thread_id.xy * 2].rgb
-			+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 0)].rgb
-			+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 1)].rgb
-			+ bindless_textures[u_input][thread_id.xy * 2 + uint2(0, 1)].rgb;
-		float _luminance = luminance(c * 0.25);
-
-		float multiplier = saturate(1 + _luminance - avg_lum * u_avg_lum_multiplier);
-		bindless_rw_textures[u_output][thread_id.xy] = float4(c * multiplier, 1);
-	}
-#elif defined DOWNSCALE
-	cbuffer Data : register(b4) {
-		uint u_input;
-		uint u_output;
-	};
-
-	[numthreads(16, 16, 1)]
-	void main(uint3 thread_id : SV_DispatchThreadID) {
-		float4 o_color = bindless_textures[u_input][thread_id.xy * 2]
-			+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 0)]
-			+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 1)]
-			+ bindless_textures[u_input][thread_id.xy * 2 + uint2(0, 1)];
-		o_color *= 0.25;
-		bindless_rw_textures[u_output][thread_id.xy] = o_color;
-	}
-#else
-	cbuffer Data : register(b4) {
-		uint u_input;
-		uint u_output;
-	};
-
-	[numthreads(16, 16, 1)]
-	void main(uint3 thread_id : SV_DispatchThreadID) {
-		float2 uv = thread_id.xy / float2(Global_framebuffer_size.xy);
-		float4 a = sampleBindlessLod(LinearSamplerClamp, u_input, uv, 0);
-		float4 b = bindless_rw_textures[u_output][thread_id.xy];
-		bindless_rw_textures[u_output][thread_id.xy] = a + b;
-	}
-#endif
+cbuffer Data : register(b4) {
+	TextureHandle u_input;
+	RWTextureHandle u_output;
+};
+
+[numthreads(16, 16, 1)]
+void main(uint3 thread_id : SV_DispatchThreadID) {
+	float2 uv = thread_id.xy * Global_rcp_framebuffer_size;
+	float4 a = sampleBindlessLod(LinearSamplerClamp, u_input, uv, 0);
+	float4 b = bindless_rw_textures[u_output][thread_id.xy];
+	bindless_rw_textures[u_output][thread_id.xy] = a + b;
+}

data/pipelines/bloom_blur.hlsl

@@ -2,9 +2,9 @@
 #include "pipelines/common.hlsli"
 
 cbuffer Drawcall : register(b4) {
-	float4 u_inv_sm_size;
-	uint u_big;
-	uint u_small;
+	float4 u_rcp_size;
+	TextureHandle u_big;
+	TextureHandle u_small;
 };
 
 struct VSOutput {
@@ -21,21 +21,21 @@ VSOutput mainVS(uint vertex_id : SV_VertexID) {
 	output.tc0 = uv;
 	#ifdef BLUR_H
 		output.tc1 = float4(
-			uv.x + 1.3846153846 * u_inv_sm_size.x, uv.y, 
-			uv.x + 3.2307692308 * u_inv_sm_size.x, uv.y
+			uv.x + 1.3846153846 * u_rcp_size.x, uv.y, 
+			uv.x + 3.2307692308 * u_rcp_size.x, uv.y
 		);
 		output.tc2 = float4(
-			uv.x - 1.3846153846 * u_inv_sm_size.x, uv.y, 
-			uv.x - 3.2307692308 * u_inv_sm_size.x, uv.y
+			uv.x - 1.3846153846 * u_rcp_size.x, uv.y, 
+			uv.x - 3.2307692308 * u_rcp_size.x, uv.y
 		);
 	#else
 		output.tc1 = float4(
-			uv.x, uv.y + 1.3846153846 * u_inv_sm_size.y,
-			uv.x, uv.y + 3.2307692308 * u_inv_sm_size.y
+			uv.x, uv.y + 1.3846153846 * u_rcp_size.y,
+			uv.x, uv.y + 3.2307692308 * u_rcp_size.y
 		);
 		output.tc2 = float4(
-			uv.x, uv.y - 1.3846153846 * u_inv_sm_size.y, 
-			uv.x, uv.y - 3.2307692308 * u_inv_sm_size.y
+			uv.x, uv.y - 1.3846153846 * u_rcp_size.y, 
+			uv.x, uv.y - 3.2307692308 * u_rcp_size.y
 		);
 	#endif
 	return output;

data/pipelines/bloom_downscale.hlsl

@@ -0,0 +1,16 @@
+#include "pipelines/common.hlsli"
+
+cbuffer Data : register(b4) {
+	TextureHandle u_input;
+	RWTextureHandle u_output;
+};
+
+[numthreads(16, 16, 1)]
+void main(uint3 thread_id : SV_DispatchThreadID) {
+	float4 o_color = bindless_textures[u_input][thread_id.xy * 2]
+		+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 0)]
+		+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 1)]
+		+ bindless_textures[u_input][thread_id.xy * 2 + uint2(0, 1)];
+	o_color *= 0.25;
+	bindless_rw_textures[u_output][thread_id.xy] = o_color;
+}

data/pipelines/bloom_extract.hlsl

@@ -0,0 +1,21 @@
+#include "pipelines/common.hlsli"
+
+cbuffer Data : register(b4) {
+	float u_avg_lum_multiplier;
+	uint u_histogram;
+	TextureHandle u_input;
+	RWTextureHandle u_output;
+};
+
+[numthreads(16, 16, 1)]
+void main(uint3 thread_id : SV_DispatchThreadID) {
+	float avg_lum = asfloat(bindless_buffers[u_histogram].Load(256 * 4));
+	float3 c = bindless_textures[u_input][thread_id.xy * 2].rgb
+		+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 0)].rgb
+		+ bindless_textures[u_input][thread_id.xy * 2 + uint2(1, 1)].rgb
+		+ bindless_textures[u_input][thread_id.xy * 2 + uint2(0, 1)].rgb;
+	float _luminance = luminance(c * 0.25);
+
+	float multiplier = saturate(1 + _luminance - avg_lum * u_avg_lum_multiplier);
+	bindless_rw_textures[u_output][thread_id.xy] = float4(c * multiplier, 1);
+}

data/pipelines/bloom_tonemap.hlsl

@@ -2,9 +2,9 @@
 
 cbuffer Drawcall : register (b4) {
 	float u_exposure;
-	uint u_input;
+	TextureHandle u_input;
 	uint u_accum;
-	uint u_output;
+	RWTextureHandle u_output;
 };
 
 [numthreads(16, 16, 1)]

data/pipelines/blur.hlsl

@@ -3,7 +3,7 @@
 
 cbuffer Drawcall : register(b4) {
 	float4 u_inv_sm_size;
-	uint u_input;
+	TextureHandle u_input;
 };
 
 struct VSOutput {
@@ -41,11 +41,11 @@ VSOutput mainVS(uint vertex_id : SV_VertexID) {
 }
 
 float4 mainPS(VSOutput input) : SV_Target {
-	float2 uv0 = saturate(input.tc0.xy);
-	float2 uv1 = saturate(input.tc1.xy);
-	float2 uv2 = saturate(input.tc1.zw);
-	float2 uv3 = saturate(input.tc2.xy);
-	float2 uv4 = saturate(input.tc2.zw);
+	float2 uv0 = input.tc0.xy;
+	float2 uv1 = input.tc1.xy;
+	float2 uv2 = input.tc1.zw;
+	float2 uv3 = input.tc2.xy;
+	float2 uv4 = input.tc2.zw;
 	float4 c0 = sampleBindless(LinearSamplerClamp, u_input, uv0);
 	float4 c1 = sampleBindless(LinearSamplerClamp, u_input, uv1);
 	float4 c2 = sampleBindless(LinearSamplerClamp, u_input, uv2);

data/pipelines/ibl_filter.hlsl

@@ -1,20 +1,21 @@
 //@surface
+
 #include "pipelines/common.hlsli"
 
 cbuffer Drawcall : register(b4) {
 	float u_filter_roughness;
 	int u_face;
 	int u_mip;
-	uint u_texture;
+	TextureHandle u_texture;
 };
 
-struct Output {
+struct VSOutput {
 	float2 uv : TEXCOORD0;
 	float4 position : SV_POSITION;
 };
 
-Output mainVS(uint vertex_id : SV_VertexID) {
-	Output output;
+VSOutput mainVS(uint vertex_id : SV_VertexID) {
+	VSOutput output;
 	float4 pos = fullscreenQuad(vertex_id, output.uv);
 	pos.xy = pos.xy;
 	pos.y = -pos.y;

data/pipelines/impostor.hlsl

@@ -1,5 +1,4 @@
 //@surface
-#include "pipelines/common.hlsli"
 
 //@texture_slot "Albedo", "textures/common/white.tga"
 //@texture_slot "Normal", "", "HAS_NORMAL"
@@ -16,6 +15,8 @@
 //@uniform "Center", "float3", {0, 0, 0}
 //@uniform "Radius", "float", 1
 
+#include "pipelines/common.hlsli"
+
 struct VSOutput {
 	float2 uv : TEXCOORD0;
 	float3 normal : TEXCOORD1;
@@ -27,7 +28,7 @@ struct VSOutput {
 	float lod : TEXCOORD5;
 	float4 position : SV_POSITION;
 };
-struct Input {
+struct VSInput {
 	float3 position : TEXCOORD0;
 	float2 uv  : TEXCOORD1;
 	#define ATTR(X) TEXCOORD##X
@@ -59,7 +60,7 @@ float2 dirToGrid(float3 vec) {
 
 
 #ifdef USE_MATRIX
-	VSOutput mainVS(Input input) {
+	VSOutput mainVS(VSInput input) {
 		VSOutput output;
 		float3x3 to_model_space = (float3x3)model_mtx;
 		#ifdef DEPTH
@@ -101,7 +102,7 @@ float2 dirToGrid(float3 vec) {
 		return output;
 	}
 #else
-	VSOutput mainVS(Input input) {
+	VSOutput mainVS(VSInput input) {
 		VSOutput output;
 		float3x3 tangent_space;
 

data/pipelines/impostor_shadow.hlsl

@@ -9,9 +9,9 @@ cbuffer Data : register(b4){
 	int2 u_tile_size;
 	int u_size;
 	float u_radius;
-	uint u_depth;
-	uint u_normalmap;
-	uint u_output;
+	TextureHandle u_depth;
+	TextureHandle u_normalmap;
+	RWTextureHandle u_output;
 };
 
 float3 impostorToWorld(float2 uv) {