summaryrefslogtreecommitdiff
path: root/servers/rendering/renderer_rd/shaders
diff options
context:
space:
mode:
Diffstat (limited to 'servers/rendering/renderer_rd/shaders')
-rw-r--r--servers/rendering/renderer_rd/shaders/canvas.glsl77
-rw-r--r--servers/rendering/renderer_rd/shaders/cluster_render.glsl15
-rw-r--r--servers/rendering/renderer_rd/shaders/effects/blur_raster.glsl49
-rw-r--r--servers/rendering/renderer_rd/shaders/effects/copy.glsl12
-rw-r--r--servers/rendering/renderer_rd/shaders/effects/copy_to_fb.glsl59
-rw-r--r--servers/rendering/renderer_rd/shaders/effects/screen_space_reflection.glsl34
-rw-r--r--servers/rendering/renderer_rd/shaders/effects/taa_resolve.glsl2
-rw-r--r--servers/rendering/renderer_rd/shaders/effects/vrs.glsl14
-rw-r--r--servers/rendering/renderer_rd/shaders/environment/volumetric_fog_process.glsl2
-rw-r--r--servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered.glsl57
-rw-r--r--servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered_inc.glsl3
-rw-r--r--servers/rendering/renderer_rd/shaders/forward_mobile/scene_forward_mobile.glsl15
-rw-r--r--servers/rendering/renderer_rd/shaders/particles.glsl2
-rw-r--r--servers/rendering/renderer_rd/shaders/scene_forward_aa_inc.glsl7
-rw-r--r--servers/rendering/renderer_rd/shaders/skeleton.glsl13
15 files changed, 194 insertions, 167 deletions
diff --git a/servers/rendering/renderer_rd/shaders/canvas.glsl b/servers/rendering/renderer_rd/shaders/canvas.glsl
index 45dc63aa17..eb5f68849e 100644
--- a/servers/rendering/renderer_rd/shaders/canvas.glsl
+++ b/servers/rendering/renderer_rd/shaders/canvas.glsl
@@ -191,48 +191,6 @@ void main() {
uv += 1e-5;
}
-#ifdef USE_ATTRIBUTES
-#if 0
- if (bool(draw_data.flags & FLAGS_USE_SKELETON) && bone_weights != vec4(0.0)) { //must be a valid bone
- //skeleton transform
- ivec4 bone_indicesi = ivec4(bone_indices);
-
- uvec2 tex_ofs = bone_indicesi.x * 2;
-
- mat2x4 m;
- m = mat2x4(
- texelFetch(skeleton_buffer, tex_ofs + 0),
- texelFetch(skeleton_buffer, tex_ofs + 1)) *
- bone_weights.x;
-
- tex_ofs = bone_indicesi.y * 2;
-
- m += mat2x4(
- texelFetch(skeleton_buffer, tex_ofs + 0),
- texelFetch(skeleton_buffer, tex_ofs + 1)) *
- bone_weights.y;
-
- tex_ofs = bone_indicesi.z * 2;
-
- m += mat2x4(
- texelFetch(skeleton_buffer, tex_ofs + 0),
- texelFetch(skeleton_buffer, tex_ofs + 1)) *
- bone_weights.z;
-
- tex_ofs = bone_indicesi.w * 2;
-
- m += mat2x4(
- texelFetch(skeleton_buffer, tex_ofs + 0),
- texelFetch(skeleton_buffer, tex_ofs + 1)) *
- bone_weights.w;
-
- mat4 bone_matrix = skeleton_data.skeleton_transform * transpose(mat4(m[0], m[1], vec4(0.0, 0.0, 1.0, 0.0), vec4(0.0, 0.0, 0.0, 1.0))) * skeleton_data.skeleton_transform_inverse;
-
- //outvec = bone_matrix * outvec;
- }
-#endif
-#endif
-
vertex = (canvas_data.canvas_transform * vec4(vertex, 0.0, 1.0)).xy;
vertex_interp = vertex;
@@ -598,13 +556,11 @@ void main() {
normal = normalize((canvas_data.canvas_normal_transform * vec4(normal, 0.0)).xyz);
}
- vec3 base_color = color.rgb;
+ vec4 base_color = color;
if (bool(draw_data.flags & FLAGS_USING_LIGHT_MASK)) {
color = vec4(0.0); //invisible by default due to using light mask
}
- vec4 original_color = color;
-
#ifdef MODE_LIGHT_ONLY
color = vec4(0.0);
#elif !defined(MODE_UNSHADED)
@@ -624,12 +580,14 @@ void main() {
#ifdef LIGHT_CODE_USED
vec4 shadow_modulate = vec4(1.0);
- light_color = light_compute(light_vertex, vec3(direction, light_array.data[light_base].height), normal, light_color, light_color.a, specular_shininess, shadow_modulate, screen_uv, uv, color, true);
+ light_color = light_compute(light_vertex, vec3(direction, light_array.data[light_base].height), normal, light_color, light_color.a, specular_shininess, shadow_modulate, screen_uv, uv, base_color, true);
#else
if (normal_used) {
vec3 light_vec = normalize(mix(vec3(direction, 0.0), vec3(0, 0, 1), light_array.data[light_base].height));
- light_color.rgb = light_normal_compute(light_vec, normal, base_color, light_color.rgb, specular_shininess, specular_shininess_used);
+ light_color.rgb = light_normal_compute(light_vec, normal, base_color.rgb, light_color.rgb, specular_shininess, specular_shininess_used);
+ } else {
+ light_color.rgb *= base_color.rgb;
}
#endif
@@ -646,8 +604,6 @@ void main() {
);
}
- light_color.rgb *= original_color.rgb;
-
light_blend_compute(light_base, light_color, color.rgb);
}
@@ -657,20 +613,7 @@ void main() {
if (i >= light_count) {
break;
}
- uint light_base;
- if (i < 8) {
- if (i < 4) {
- light_base = draw_data.lights[0];
- } else {
- light_base = draw_data.lights[1];
- }
- } else {
- if (i < 12) {
- light_base = draw_data.lights[2];
- } else {
- light_base = draw_data.lights[3];
- }
- }
+ uint light_base = draw_data.lights[i >> 2];
light_base >>= (i & 3) * 8;
light_base &= 0xFF;
@@ -685,7 +628,7 @@ void main() {
vec3 light_position = vec3(light_array.data[light_base].position, light_array.data[light_base].height);
light_color.rgb *= light_base_color.rgb;
- light_color = light_compute(light_vertex, light_position, normal, light_color, light_base_color.a, specular_shininess, shadow_modulate, screen_uv, uv, color, false);
+ light_color = light_compute(light_vertex, light_position, normal, light_color, light_base_color.a, specular_shininess, shadow_modulate, screen_uv, uv, base_color, false);
#else
light_color.rgb *= light_base_color.rgb * light_base_color.a;
@@ -695,7 +638,9 @@ void main() {
vec3 pos = light_vertex;
vec3 light_vec = normalize(light_pos - pos);
- light_color.rgb = light_normal_compute(light_vec, normal, base_color, light_color.rgb, specular_shininess, specular_shininess_used);
+ light_color.rgb = light_normal_compute(light_vec, normal, base_color.rgb, light_color.rgb, specular_shininess, specular_shininess_used);
+ } else {
+ light_color.rgb *= base_color.rgb;
}
#endif
if (any(lessThan(tex_uv, vec2(0.0, 0.0))) || any(greaterThanEqual(tex_uv, vec2(1.0, 1.0)))) {
@@ -743,8 +688,6 @@ void main() {
);
}
- light_color.rgb *= original_color.rgb;
-
light_blend_compute(light_base, light_color, color.rgb);
}
#endif
diff --git a/servers/rendering/renderer_rd/shaders/cluster_render.glsl b/servers/rendering/renderer_rd/shaders/cluster_render.glsl
index 2fe230f0bf..8c26a67926 100644
--- a/servers/rendering/renderer_rd/shaders/cluster_render.glsl
+++ b/servers/rendering/renderer_rd/shaders/cluster_render.glsl
@@ -64,7 +64,7 @@ void main() {
#version 450
#VERSION_DEFINES
-
+#ifndef MOLTENVK_USED // Metal will corrupt GPU state otherwise
#if defined(has_GL_KHR_shader_subgroup_ballot) && defined(has_GL_KHR_shader_subgroup_arithmetic) && defined(has_GL_KHR_shader_subgroup_vote)
#extension GL_KHR_shader_subgroup_ballot : enable
@@ -73,6 +73,7 @@ void main() {
#define USE_SUBGROUPS
#endif
+#endif
layout(location = 0) in float depth_interp;
layout(location = 1) in flat uint element_index;
@@ -141,7 +142,11 @@ void main() {
}
}
#else
- if (!gl_HelperInvocation) {
+// MoltenVK/Metal fails to compile shaders using gl_HelperInvocation for some GPUs
+#ifndef MOLTENVK_USED
+ if (!gl_HelperInvocation)
+#endif
+ {
atomicOr(cluster_render.data[usage_write_offset], usage_write_bit);
}
#endif
@@ -161,7 +166,11 @@ void main() {
}
}
#else
- if (!gl_HelperInvocation) {
+// MoltenVK/Metal fails to compile shaders using gl_HelperInvocation for some GPUs
+#ifndef MOLTENVK_USED
+ if (!gl_HelperInvocation)
+#endif
+ {
atomicOr(cluster_render.data[z_write_offset], z_write_bit);
}
#endif
diff --git a/servers/rendering/renderer_rd/shaders/effects/blur_raster.glsl b/servers/rendering/renderer_rd/shaders/effects/blur_raster.glsl
index cb06250cf2..31aabbe9d2 100644
--- a/servers/rendering/renderer_rd/shaders/effects/blur_raster.glsl
+++ b/servers/rendering/renderer_rd/shaders/effects/blur_raster.glsl
@@ -53,30 +53,31 @@ void main() {
#ifdef MODE_GAUSSIAN_BLUR
- // Simpler blur uses SIGMA2 for the gaussian kernel for a stronger effect
-
- // note, for blur blur.luminance_multiplier is irrelavant, we would be multiplying and then dividing by this amount.
-
- if (bool(blur.flags & FLAG_HORIZONTAL)) {
- vec2 pix_size = blur.pixel_size;
- pix_size *= 0.5; //reading from larger buffer, so use more samples
- vec4 color = texture(source_color, uv_interp + vec2(0.0, 0.0) * pix_size) * 0.214607;
- color += texture(source_color, uv_interp + vec2(1.0, 0.0) * pix_size) * 0.189879;
- color += texture(source_color, uv_interp + vec2(2.0, 0.0) * pix_size) * 0.131514;
- color += texture(source_color, uv_interp + vec2(3.0, 0.0) * pix_size) * 0.071303;
- color += texture(source_color, uv_interp + vec2(-1.0, 0.0) * pix_size) * 0.189879;
- color += texture(source_color, uv_interp + vec2(-2.0, 0.0) * pix_size) * 0.131514;
- color += texture(source_color, uv_interp + vec2(-3.0, 0.0) * pix_size) * 0.071303;
- frag_color = color;
- } else {
- vec2 pix_size = blur.pixel_size;
- vec4 color = texture(source_color, uv_interp + vec2(0.0, 0.0) * pix_size) * 0.38774;
- color += texture(source_color, uv_interp + vec2(0.0, 1.0) * pix_size) * 0.24477;
- color += texture(source_color, uv_interp + vec2(0.0, 2.0) * pix_size) * 0.06136;
- color += texture(source_color, uv_interp + vec2(0.0, -1.0) * pix_size) * 0.24477;
- color += texture(source_color, uv_interp + vec2(0.0, -2.0) * pix_size) * 0.06136;
- frag_color = color;
- }
+ // For Gaussian Blur we use 13 taps in a single pass instead of 12 taps over 2 passes.
+ // This minimizes the number of times we change framebuffers which is very important for mobile.
+ // Source: http://www.iryoku.com/next-generation-post-processing-in-call-of-duty-advanced-warfare
+ vec4 A = texture(source_color, uv_interp + blur.pixel_size * vec2(-1.0, -1.0));
+ vec4 B = texture(source_color, uv_interp + blur.pixel_size * vec2(0.0, -1.0));
+ vec4 C = texture(source_color, uv_interp + blur.pixel_size * vec2(1.0, -1.0));
+ vec4 D = texture(source_color, uv_interp + blur.pixel_size * vec2(-0.5, -0.5));
+ vec4 E = texture(source_color, uv_interp + blur.pixel_size * vec2(0.5, -0.5));
+ vec4 F = texture(source_color, uv_interp + blur.pixel_size * vec2(-1.0, 0.0));
+ vec4 G = texture(source_color, uv_interp);
+ vec4 H = texture(source_color, uv_interp + blur.pixel_size * vec2(1.0, 0.0));
+ vec4 I = texture(source_color, uv_interp + blur.pixel_size * vec2(-0.5, 0.5));
+ vec4 J = texture(source_color, uv_interp + blur.pixel_size * vec2(0.5, 0.5));
+ vec4 K = texture(source_color, uv_interp + blur.pixel_size * vec2(-1.0, 1.0));
+ vec4 L = texture(source_color, uv_interp + blur.pixel_size * vec2(0.0, 1.0));
+ vec4 M = texture(source_color, uv_interp + blur.pixel_size * vec2(1.0, 1.0));
+
+ float base_weight = 0.5 / 4.0;
+ float lesser_weight = 0.125 / 4.0;
+
+ frag_color = (D + E + I + J) * base_weight;
+ frag_color += (A + B + G + F) * lesser_weight;
+ frag_color += (B + C + H + G) * lesser_weight;
+ frag_color += (F + G + L + K) * lesser_weight;
+ frag_color += (G + H + M + L) * lesser_weight;
#endif
#ifdef MODE_GAUSSIAN_GLOW
diff --git a/servers/rendering/renderer_rd/shaders/effects/copy.glsl b/servers/rendering/renderer_rd/shaders/effects/copy.glsl
index bfe329b8ec..5cc2ed7622 100644
--- a/servers/rendering/renderer_rd/shaders/effects/copy.glsl
+++ b/servers/rendering/renderer_rd/shaders/effects/copy.glsl
@@ -194,10 +194,10 @@ void main() {
color = min(color * feedback, vec4(params.glow_luminance_cap));
}
-#endif
+#endif // MODE_GLOW
imageStore(dest_buffer, pos + params.target, color);
-#endif
+#endif // MODE_GAUSSIAN_BLUR
#ifdef MODE_SIMPLE_COPY
@@ -227,7 +227,7 @@ void main() {
imageStore(dest_buffer, pos + params.target, color);
-#endif
+#endif // MODE_SIMPLE_COPY
#ifdef MODE_SIMPLE_COPY_DEPTH
@@ -239,7 +239,7 @@ void main() {
imageStore(dest_buffer, pos + params.target, vec4(color.r));
-#endif
+#endif // MODE_SIMPLE_COPY_DEPTH
#ifdef MODE_LINEARIZE_DEPTH_COPY
@@ -253,7 +253,7 @@ void main() {
}
imageStore(dest_buffer, pos + params.target, color);
-#endif
+#endif // MODE_LINEARIZE_DEPTH_COPY
#if defined(MODE_CUBEMAP_TO_PANORAMA) || defined(MODE_CUBEMAP_ARRAY_TO_PANORAMA)
@@ -276,7 +276,7 @@ void main() {
vec4 color = textureLod(source_color, vec4(normal, params.camera_z_far), 0.0); //the biggest the lod the least the acne
#endif
imageStore(dest_buffer, pos + params.target, color);
-#endif
+#endif // defined(MODE_CUBEMAP_TO_PANORAMA) || defined(MODE_CUBEMAP_ARRAY_TO_PANORAMA)
#ifdef MODE_SET_COLOR
imageStore(dest_buffer, pos + params.target, params.set_color);
diff --git a/servers/rendering/renderer_rd/shaders/effects/copy_to_fb.glsl b/servers/rendering/renderer_rd/shaders/effects/copy_to_fb.glsl
index 1c17eabb56..46bb99794d 100644
--- a/servers/rendering/renderer_rd/shaders/effects/copy_to_fb.glsl
+++ b/servers/rendering/renderer_rd/shaders/effects/copy_to_fb.glsl
@@ -13,6 +13,14 @@
#endif // has_VK_KHR_multiview
#endif //MULTIVIEW
+#define FLAG_FLIP_Y (1 << 0)
+#define FLAG_USE_SECTION (1 << 1)
+#define FLAG_FORCE_LUMINANCE (1 << 2)
+#define FLAG_ALPHA_TO_ZERO (1 << 3)
+#define FLAG_SRGB (1 << 4)
+#define FLAG_ALPHA_TO_ONE (1 << 5)
+#define FLAG_LINEAR (1 << 6)
+
#ifdef MULTIVIEW
layout(location = 0) out vec3 uv_interp;
#else
@@ -22,11 +30,10 @@ layout(location = 0) out vec2 uv_interp;
layout(push_constant, std430) uniform Params {
vec4 section;
vec2 pixel_size;
- bool flip_y;
- bool use_section;
+ float luminance_multiplier;
+ uint flags;
- bool force_luminance;
- uint pad[3];
+ vec4 color;
}
params;
@@ -37,13 +44,13 @@ void main() {
uv_interp.z = ViewIndex;
#endif
vec2 vpos = uv_interp.xy;
- if (params.use_section) {
+ if (bool(params.flags & FLAG_USE_SECTION)) {
vpos = params.section.xy + vpos * params.section.zw;
}
gl_Position = vec4(vpos * 2.0 - 1.0, 0.0, 1.0);
- if (params.flip_y) {
+ if (bool(params.flags & FLAG_FLIP_Y)) {
uv_interp.y = 1.0 - uv_interp.y;
}
}
@@ -63,16 +70,21 @@ void main() {
#endif // has_VK_KHR_multiview
#endif //MULTIVIEW
+#define FLAG_FLIP_Y (1 << 0)
+#define FLAG_USE_SECTION (1 << 1)
+#define FLAG_FORCE_LUMINANCE (1 << 2)
+#define FLAG_ALPHA_TO_ZERO (1 << 3)
+#define FLAG_SRGB (1 << 4)
+#define FLAG_ALPHA_TO_ONE (1 << 5)
+#define FLAG_LINEAR (1 << 6)
+
layout(push_constant, std430) uniform Params {
vec4 section;
vec2 pixel_size;
- bool flip_y;
- bool use_section;
+ float luminance_multiplier;
+ uint flags;
- bool force_luminance;
- bool alpha_to_zero;
- bool srgb;
- uint pad;
+ vec4 color;
}
params;
@@ -104,7 +116,15 @@ vec3 linear_to_srgb(vec3 color) {
return mix((vec3(1.0f) + a) * pow(color.rgb, vec3(1.0f / 2.4f)) - a, 12.92f * color.rgb, lessThan(color.rgb, vec3(0.0031308f)));
}
+vec3 srgb_to_linear(vec3 color) {
+ return mix(pow((color.rgb + vec3(0.055)) * (1.0 / (1.0 + 0.055)), vec3(2.4)), color.rgb * (1.0 / 12.92), lessThan(color.rgb, vec3(0.04045)));
+}
+
void main() {
+#ifdef MODE_SET_COLOR
+ frag_color = params.color;
+#else
+
#ifdef MULTIVIEW
vec3 uv = uv_interp;
#else
@@ -155,15 +175,22 @@ void main() {
#endif /* MODE_TWO_SOURCES */
#endif /* MULTIVIEW */
- if (params.force_luminance) {
+ if (bool(params.flags & FLAG_FORCE_LUMINANCE)) {
color.rgb = vec3(max(max(color.r, color.g), color.b));
}
- if (params.alpha_to_zero) {
+ if (bool(params.flags & FLAG_ALPHA_TO_ZERO)) {
color.rgb *= color.a;
}
- if (params.srgb) {
+ if (bool(params.flags & FLAG_SRGB)) {
color.rgb = linear_to_srgb(color.rgb);
}
+ if (bool(params.flags & FLAG_ALPHA_TO_ONE)) {
+ color.a = 1.0;
+ }
+ if (bool(params.flags & FLAG_LINEAR)) {
+ color.rgb = srgb_to_linear(color.rgb);
+ }
- frag_color = color;
+ frag_color = color / params.luminance_multiplier;
+#endif // MODE_SET_COLOR
}
diff --git a/servers/rendering/renderer_rd/shaders/effects/screen_space_reflection.glsl b/servers/rendering/renderer_rd/shaders/effects/screen_space_reflection.glsl
index 9f86643e52..9d8294a7da 100644
--- a/servers/rendering/renderer_rd/shaders/effects/screen_space_reflection.glsl
+++ b/servers/rendering/renderer_rd/shaders/effects/screen_space_reflection.glsl
@@ -121,7 +121,7 @@ void main() {
// clip z and w advance to line advance
vec2 line_advance = normalize(line_dir); // down to pixel
- float step_size = length(line_advance) / length(line_dir);
+ float step_size = 1.0 / length(line_dir);
float z_advance = z_dir * step_size; // adapt z advance to line advance
float w_advance = w_dir * step_size; // adapt w advance to line advance
@@ -139,6 +139,14 @@ void main() {
float depth;
vec2 prev_pos = pos;
+ if (ivec2(pos + line_advance - 0.5) == ssC) {
+ // It is possible for rounding to cause our first pixel to check to be the pixel we're reflecting.
+ // Make sure we skip it
+ pos += line_advance;
+ z += z_advance;
+ w += w_advance;
+ }
+
bool found = false;
float steps_taken = 0.0;
@@ -149,8 +157,8 @@ void main() {
w += w_advance;
// convert to linear depth
-
- depth = imageLoad(source_depth, ivec2(pos - 0.5)).r;
+ ivec2 test_pos = ivec2(pos - 0.5);
+ depth = imageLoad(source_depth, test_pos).r;
if (sc_multiview) {
depth = depth * 2.0 - 1.0;
depth = 2.0 * params.camera_z_near * params.camera_z_far / (params.camera_z_far + params.camera_z_near - depth * (params.camera_z_far - params.camera_z_near));
@@ -161,13 +169,21 @@ void main() {
z_to = z / w;
if (depth > z_to) {
- // if depth was surpassed
- if (depth <= max(z_to, z_from) + params.depth_tolerance && -depth < params.camera_z_far * 0.95) {
- // check the depth tolerance and far clip
- // check that normal is valid
- found = true;
+ // Test if our ray is hitting the "right" side of the surface, if not we're likely self reflecting and should skip.
+ vec4 test_normal_roughness = imageLoad(source_normal_roughness, test_pos);
+ vec3 test_normal = test_normal_roughness.xyz * 2.0 - 1.0;
+ test_normal = normalize(test_normal);
+ test_normal.y = -test_normal.y; //because this code reads flipped
+
+ if (dot(ray_dir, test_normal) < 0.001) {
+ // if depth was surpassed
+ if (depth <= max(z_to, z_from) + params.depth_tolerance && -depth < params.camera_z_far * 0.95) {
+ // check the depth tolerance and far clip
+ // check that normal is valid
+ found = true;
+ }
+ break;
}
- break;
}
steps_taken += 1.0;
diff --git a/servers/rendering/renderer_rd/shaders/effects/taa_resolve.glsl b/servers/rendering/renderer_rd/shaders/effects/taa_resolve.glsl
index b0a0839836..02566d8e35 100644
--- a/servers/rendering/renderer_rd/shaders/effects/taa_resolve.glsl
+++ b/servers/rendering/renderer_rd/shaders/effects/taa_resolve.glsl
@@ -32,7 +32,9 @@
// Based on Spartan Engine's TAA implementation (without TAA upscale).
// <https://github.com/PanosK92/SpartanEngine/blob/a8338d0609b85dc32f3732a5c27fb4463816a3b9/Data/shaders/temporal_antialiasing.hlsl>
+#ifndef MOLTENVK_USED
#define USE_SUBGROUPS
+#endif // MOLTENVK_USED
#define GROUP_SIZE 8
#define FLT_MIN 0.00000001
diff --git a/servers/rendering/renderer_rd/shaders/effects/vrs.glsl b/servers/rendering/renderer_rd/shaders/effects/vrs.glsl
index 5ef83c0b44..b450bb9fe9 100644
--- a/servers/rendering/renderer_rd/shaders/effects/vrs.glsl
+++ b/servers/rendering/renderer_rd/shaders/effects/vrs.glsl
@@ -63,10 +63,18 @@ void main() {
#ifdef MULTIVIEW
vec4 color = textureLod(source_color, uv, 0.0);
+ frag_color = uint(color.r * 255.0);
#else /* MULTIVIEW */
vec4 color = textureLod(source_color, uv, 0.0);
-#endif /* MULTIVIEW */
- // See if we can change the sampler to one that returns int...
- frag_color = uint(color.r * 256.0);
+ // for user supplied VRS map we do a color mapping
+ color.r *= 3.0;
+ frag_color = int(color.r) << 2;
+
+ color.g *= 3.0;
+ frag_color += int(color.g);
+
+ // note 1x4, 4x1, 1x8, 8x1, 2x8 and 8x2 are not supported
+ // 4x8, 8x4 and 8x8 are only available on some GPUs
+#endif /* MULTIVIEW */
}
diff --git a/servers/rendering/renderer_rd/shaders/environment/volumetric_fog_process.glsl b/servers/rendering/renderer_rd/shaders/environment/volumetric_fog_process.glsl
index eed9038502..28507e6c12 100644
--- a/servers/rendering/renderer_rd/shaders/environment/volumetric_fog_process.glsl
+++ b/servers/rendering/renderer_rd/shaders/environment/volumetric_fog_process.glsl
@@ -381,7 +381,7 @@ void main() {
float cell_depth_size = abs(view_pos.z - get_depth_at_pos(fog_cell_size.z, pos.z + 1));
//compute directional lights
- if (total_density > 0.001) {
+ if (total_density > 0.00005) {
for (uint i = 0; i < params.directional_light_count; i++) {
if (directional_lights.data[i].volumetric_fog_energy > 0.001) {
vec3 shadow_attenuation = vec3(1.0);
diff --git a/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered.glsl b/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered.glsl
index 1a8a1f3aa3..896f51ca01 100644
--- a/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered.glsl
+++ b/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered.glsl
@@ -62,7 +62,7 @@ vec3 oct_to_vec3(vec2 e) {
vec3 v = vec3(e.xy, 1.0 - abs(e.x) - abs(e.y));
float t = max(-v.z, 0.0);
v.xy += t * -sign(v.xy);
- return v;
+ return normalize(v);
}
/* Varyings */
@@ -97,9 +97,7 @@ layout(location = 8) out vec4 prev_screen_position;
#ifdef MATERIAL_UNIFORMS_USED
layout(set = MATERIAL_UNIFORM_SET, binding = 0, std140) uniform MaterialUniforms{
-
#MATERIAL_UNIFORMS
-
} material;
#endif
@@ -691,7 +689,7 @@ vec4 fog_process(vec3 vertex) {
void cluster_get_item_range(uint p_offset, out uint item_min, out uint item_max, out uint item_from, out uint item_to) {
uint item_min_max = cluster_buffer.data[p_offset];
- item_min = item_min_max & 0xFFFF;
+ item_min = item_min_max & 0xFFFFu;
item_max = item_min_max >> 16;
item_from = item_min >> 5;
@@ -828,7 +826,8 @@ void fragment_shader(in SceneData scene_data) {
// alpha hash can be used in unison with alpha antialiasing
#ifdef ALPHA_HASH_USED
- if (alpha < compute_alpha_hash_threshold(vertex, alpha_hash_scale)) {
+ vec3 object_pos = (inverse(read_model_matrix) * inv_view_matrix * vec4(vertex, 1.0)).xyz;
+ if (alpha < compute_alpha_hash_threshold(object_pos, alpha_hash_scale)) {
discard;
}
#endif // ALPHA_HASH_USED
@@ -958,9 +957,9 @@ void fragment_shader(in SceneData scene_data) {
while (merged_mask != 0) {
uint bit = findMSB(merged_mask);
- merged_mask &= ~(1 << bit);
+ merged_mask &= ~(1u << bit);
#ifdef USE_SUBGROUPS
- if (((1 << bit) & mask) == 0) { //do not process if not originally here
+ if (((1u << bit) & mask) == 0) { //do not process if not originally here
continue;
}
#endif
@@ -1419,9 +1418,9 @@ void fragment_shader(in SceneData scene_data) {
while (merged_mask != 0) {
uint bit = findMSB(merged_mask);
- merged_mask &= ~(1 << bit);
+ merged_mask &= ~(1u << bit);
#ifdef USE_SUBGROUPS
- if (((1 << bit) & mask) == 0) { //do not process if not originally here
+ if (((1u << bit) & mask) == 0) { //do not process if not originally here
continue;
}
#endif
@@ -1447,18 +1446,24 @@ void fragment_shader(in SceneData scene_data) {
}
//finalize ambient light here
- ambient_light *= albedo.rgb;
- ambient_light *= ao;
+ {
+#if defined(AMBIENT_LIGHT_DISABLED)
+ ambient_light = vec3(0.0, 0.0, 0.0);
+#else
+ ambient_light *= albedo.rgb;
+ ambient_light *= ao;
+
+ if (bool(implementation_data.ss_effects_flags & SCREEN_SPACE_EFFECTS_FLAGS_USE_SSIL)) {
+ vec4 ssil = textureLod(sampler2D(ssil_buffer, material_samplers[SAMPLER_LINEAR_CLAMP]), screen_uv, 0.0);
+ ambient_light *= 1.0 - ssil.a;
+ ambient_light += ssil.rgb * albedo.rgb;
+ }
+#endif // AMBIENT_LIGHT_DISABLED
+ }
// convert ao to direct light ao
ao = mix(1.0, ao, ao_light_affect);
- if (bool(implementation_data.ss_effects_flags & SCREEN_SPACE_EFFECTS_FLAGS_USE_SSIL)) {
- vec4 ssil = textureLod(sampler2D(ssil_buffer, material_samplers[SAMPLER_LINEAR_CLAMP]), screen_uv, 0.0);
- ambient_light *= 1.0 - ssil.a;
- ambient_light += ssil.rgb * albedo.rgb;
- }
-
//this saves some VGPRs
vec3 f0 = F0(metallic, specular, albedo);
@@ -1775,9 +1780,9 @@ void fragment_shader(in SceneData scene_data) {
float shadow = 1.0;
#ifndef SHADOWS_DISABLED
if (i < 4) {
- shadow = float(shadow0 >> (i * 8) & 0xFF) / 255.0;
+ shadow = float(shadow0 >> (i * 8u) & 0xFFu) / 255.0;
} else {
- shadow = float(shadow1 >> ((i - 4) * 8) & 0xFF) / 255.0;
+ shadow = float(shadow1 >> ((i - 4u) * 8u) & 0xFFu) / 255.0;
}
shadow = shadow * directional_lights.data[i].shadow_opacity + 1.0 - directional_lights.data[i].shadow_opacity;
@@ -1839,9 +1844,9 @@ void fragment_shader(in SceneData scene_data) {
while (merged_mask != 0) {
uint bit = findMSB(merged_mask);
- merged_mask &= ~(1 << bit);
+ merged_mask &= ~(1u << bit);
#ifdef USE_SUBGROUPS
- if (((1 << bit) & mask) == 0) { //do not process if not originally here
+ if (((1u << bit) & mask) == 0) { //do not process if not originally here
continue;
}
#endif
@@ -1910,9 +1915,9 @@ void fragment_shader(in SceneData scene_data) {
while (merged_mask != 0) {
uint bit = findMSB(merged_mask);
- merged_mask &= ~(1 << bit);
+ merged_mask &= ~(1u << bit);
#ifdef USE_SUBGROUPS
- if (((1 << bit) & mask) == 0) { //do not process if not originally here
+ if (((1u << bit) & mask) == 0) { //do not process if not originally here
continue;
}
#endif
@@ -2065,7 +2070,7 @@ void fragment_shader(in SceneData scene_data) {
float sGreen = floor((cGreen / pow(2.0f, exps - B - N)) + 0.5f);
float sBlue = floor((cBlue / pow(2.0f, exps - B - N)) + 0.5f);
//store as 8985 to have 2 extra neighbour bits
- uint light_rgbe = ((uint(sRed) & 0x1FF) >> 1) | ((uint(sGreen) & 0x1FF) << 8) | (((uint(sBlue) & 0x1FF) >> 1) << 17) | ((uint(exps) & 0x1F) << 25);
+ uint light_rgbe = ((uint(sRed) & 0x1FFu) >> 1) | ((uint(sGreen) & 0x1FFu) << 8) | (((uint(sBlue) & 0x1FFu) >> 1) << 17) | ((uint(exps) & 0x1Fu) << 25);
imageStore(emission_grid, grid_pos, uvec4(light_rgbe));
imageStore(emission_aniso_grid, grid_pos, uvec4(light_aniso));
@@ -2099,8 +2104,8 @@ void fragment_shader(in SceneData scene_data) {
if (bool(instances.data[instance_index].flags & INSTANCE_FLAGS_USE_VOXEL_GI)) { // process voxel_gi_instances
uint index1 = instances.data[instance_index].gi_offset & 0xFFFF;
uint index2 = instances.data[instance_index].gi_offset >> 16;
- voxel_gi_buffer.x = index1 & 0xFF;
- voxel_gi_buffer.y = index2 & 0xFF;
+ voxel_gi_buffer.x = index1 & 0xFFu;
+ voxel_gi_buffer.y = index2 & 0xFFu;
} else {
voxel_gi_buffer.x = 0xFF;
voxel_gi_buffer.y = 0xFF;
diff --git a/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered_inc.glsl b/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered_inc.glsl
index 0d36b98645..3a45ab0059 100644
--- a/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered_inc.glsl
+++ b/servers/rendering/renderer_rd/shaders/forward_clustered/scene_forward_clustered_inc.glsl
@@ -4,14 +4,15 @@
#define MAX_VOXEL_GI_INSTANCES 8
#define MAX_VIEWS 2
+#ifndef MOLTENVK_USED
#if defined(has_GL_KHR_shader_subgroup_ballot) && defined(has_GL_KHR_shader_subgroup_arithmetic)
#extension GL_KHR_shader_subgroup_ballot : enable
#extension GL_KHR_shader_subgroup_arithmetic : enable
#define USE_SUBGROUPS
-
#endif
+#endif // MOLTENVK_USED
#if defined(USE_MULTIVIEW) && defined(has_VK_KHR_multiview)
#extension GL_EXT_multiview : enable
diff --git a/servers/rendering/renderer_rd/shaders/forward_mobile/scene_forward_mobile.glsl b/servers/rendering/renderer_rd/shaders/forward_mobile/scene_forward_mobile.glsl
index 33fd4c35b1..d50749306e 100644
--- a/servers/rendering/renderer_rd/shaders/forward_mobile/scene_forward_mobile.glsl
+++ b/servers/rendering/renderer_rd/shaders/forward_mobile/scene_forward_mobile.glsl
@@ -63,7 +63,7 @@ vec3 oct_to_vec3(vec2 e) {
vec3 v = vec3(e.xy, 1.0 - abs(e.x) - abs(e.y));
float t = max(-v.z, 0.0);
v.xy += t * -sign(v.xy);
- return v;
+ return normalize(v);
}
/* Varyings */
@@ -779,7 +779,8 @@ void main() {
// alpha hash can be used in unison with alpha antialiasing
#ifdef ALPHA_HASH_USED
- if (alpha < compute_alpha_hash_threshold(vertex, alpha_hash_scale)) {
+ vec3 object_pos = (inverse(read_model_matrix) * inv_view_matrix * vec4(vertex, 1.0)).xyz;
+ if (alpha < compute_alpha_hash_threshold(object_pos, alpha_hash_scale)) {
discard;
}
#endif // ALPHA_HASH_USED
@@ -1172,8 +1173,14 @@ void main() {
} //Reflection probes
// finalize ambient light here
- ambient_light *= albedo.rgb;
- ambient_light *= ao;
+ {
+#if defined(AMBIENT_LIGHT_DISABLED)
+ ambient_light = vec3(0.0, 0.0, 0.0);
+#else
+ ambient_light *= albedo.rgb;
+ ambient_light *= ao;
+#endif // AMBIENT_LIGHT_DISABLED
+ }
// convert ao to direct light ao
ao = mix(1.0, ao, ao_light_affect);
diff --git a/servers/rendering/renderer_rd/shaders/particles.glsl b/servers/rendering/renderer_rd/shaders/particles.glsl
index 3a6dd579b9..9f6aa7adc0 100644
--- a/servers/rendering/renderer_rd/shaders/particles.glsl
+++ b/servers/rendering/renderer_rd/shaders/particles.glsl
@@ -462,7 +462,7 @@ void main() {
if (any(lessThan(uvw_pos, vec3(0.0))) || any(greaterThan(uvw_pos, vec3(1.0)))) {
continue;
}
- vec3 s = texture(sampler3D(sdf_vec_textures[FRAME.attractors[i].texture_index], material_samplers[SAMPLER_LINEAR_CLAMP]), uvw_pos).xyz * 2.0 - 1.0;
+ vec3 s = texture(sampler3D(sdf_vec_textures[FRAME.attractors[i].texture_index], material_samplers[SAMPLER_LINEAR_CLAMP]), uvw_pos).xyz * -2.0 + 1.0;
dir = mat3(FRAME.attractors[i].transform) * safe_normalize(s); //revert direction
amount = length(s);
diff --git a/servers/rendering/renderer_rd/shaders/scene_forward_aa_inc.glsl b/servers/rendering/renderer_rd/shaders/scene_forward_aa_inc.glsl
index 97c913d489..71510ee06a 100644
--- a/servers/rendering/renderer_rd/shaders/scene_forward_aa_inc.glsl
+++ b/servers/rendering/renderer_rd/shaders/scene_forward_aa_inc.glsl
@@ -11,7 +11,8 @@ float hash_3d(vec3 p) {
float compute_alpha_hash_threshold(vec3 pos, float hash_scale) {
vec3 dx = dFdx(pos);
- vec3 dy = dFdx(pos);
+ vec3 dy = dFdy(pos);
+
float delta_max_sqr = max(length(dx), length(dy));
float pix_scale = 1.0 / (hash_scale * delta_max_sqr);
@@ -32,9 +33,9 @@ float compute_alpha_hash_threshold(vec3 pos, float hash_scale) {
1.0 - ((1.0 - a_interp) * (1.0 - a_interp) / (2.0 * min_lerp * (1.0 - min_lerp))));
float alpha_hash_threshold =
- (lerp_factor < (1.0 - min_lerp)) ? ((lerp_factor < min_lerp) ? cases.x : cases.y) : cases.z;
+ (a_interp < (1.0 - min_lerp)) ? ((a_interp < min_lerp) ? cases.x : cases.y) : cases.z;
- return clamp(alpha_hash_threshold, 0.0, 1.0);
+ return clamp(alpha_hash_threshold, 0.00001, 1.0);
}
#endif // ALPHA_HASH_USED
diff --git a/servers/rendering/renderer_rd/shaders/skeleton.glsl b/servers/rendering/renderer_rd/shaders/skeleton.glsl
index 75bea9300b..f5b233cca0 100644
--- a/servers/rendering/renderer_rd/shaders/skeleton.glsl
+++ b/servers/rendering/renderer_rd/shaders/skeleton.glsl
@@ -63,7 +63,7 @@ vec3 oct_to_vec3(vec2 oct) {
vec3 v = vec3(oct.xy, 1.0 - abs(oct.x) - abs(oct.y));
float t = max(-v.z, 0.0);
v.xy += t * -sign(v.xy);
- return v;
+ return normalize(v);
}
vec3 decode_uint_oct_to_norm(uint base) {
@@ -143,8 +143,8 @@ void main() {
uint skin_offset = params.skin_stride * index;
uvec2 bones = uvec2(src_bone_weights.data[skin_offset + 0], src_bone_weights.data[skin_offset + 1]);
- uvec2 bones_01 = uvec2(bones.x & 0xFFFF, bones.x >> 16) * 3; //pre-add xform offset
- uvec2 bones_23 = uvec2(bones.y & 0xFFFF, bones.y >> 16) * 3;
+ uvec2 bones_01 = uvec2(bones.x & 0xFFFF, bones.x >> 16) * 2; //pre-add xform offset
+ uvec2 bones_23 = uvec2(bones.y & 0xFFFF, bones.y >> 16) * 2;
skin_offset += params.skin_weight_offset;
@@ -161,6 +161,13 @@ void main() {
//reverse order because its transposed
vertex = (vec4(vertex, 0.0, 1.0) * m).xy;
}
+
+ uint dst_offset = index * params.vertex_stride;
+
+ uvec2 uvertex = floatBitsToUint(vertex);
+ dst_vertices.data[dst_offset + 0] = uvertex.x;
+ dst_vertices.data[dst_offset + 1] = uvertex.y;
+
#else
vec3 vertex;
vec3 normal;