1 files changed, 117 insertions, 0 deletions
diff --git a/src/mbgl/programs/gl/hillshade_prepare.cpp b/src/mbgl/programs/gl/hillshade_prepare.cpp
new file mode 100644
index 0000000000..b445713530
--- /dev/null
+++ b/src/mbgl/programs/gl/hillshade_prepare.cpp
@@ -0,0 +1,117 @@
+// NOTE: DO NOT CHANGE THIS FILE. IT IS AUTOMATICALLY GENERATED.
+
+#include <mbgl/programs/hillshade_prepare_program.hpp>
+#include <mbgl/programs/gl/shader_source.hpp>
+#include <mbgl/gl/program.hpp>
+
+namespace mbgl {
+namespace gfx {
+
+template <>
+std::unique_ptr<Program<HillshadePrepareProgram>>
+Context::createProgram<gl::Context>(const ProgramParameters& programParameters) {
+    return gl::Program<HillshadePrepareProgram>::createProgram(
+        reinterpret_cast<gl::Context&>(*this), programParameters, "hillshade_prepare",
+        programs::gl::shaderSource() + 27925, programs::gl::shaderSource() + 28218);
+}
+
+} // namespace gfx
+} // namespace mbgl
+
+// Uncompressed source of hillshade_prepare.vertex.glsl:
+/*
+uniform mat4 u_matrix;
+uniform vec2 u_dimension;
+
+attribute vec2 a_pos;
+attribute vec2 a_texture_pos;
+
+varying vec2 v_pos;
+
+void main() {
+    gl_Position = u_matrix * vec4(a_pos, 0, 1);
+
+    highp vec2 epsilon = 1.0 / u_dimension;
+    float scale = (u_dimension.x - 2.0) / u_dimension.x;
+    v_pos = (a_texture_pos / 8192.0) * scale + epsilon;
+}
+
+*/
+
+// Uncompressed source of hillshade_prepare.fragment.glsl:
+/*
+#ifdef GL_ES
+precision highp float;
+#endif
+
+uniform sampler2D u_image;
+varying vec2 v_pos;
+uniform vec2 u_dimension;
+uniform float u_zoom;
+uniform float u_maxzoom;
+
+float getElevation(vec2 coord, float bias) {
+    // Convert encoded elevation value to meters
+    vec4 data = texture2D(u_image, coord) * 255.0;
+    return (data.r + data.g * 256.0 + data.b * 256.0 * 256.0) / 4.0;
+}
+
+void main() {
+    vec2 epsilon = 1.0 / u_dimension;
+
+    // queried pixels:
+    // +-----------+
+    // |   |   |   |
+    // | a | b | c |
+    // |   |   |   |
+    // +-----------+
+    // |   |   |   |
+    // | d | e | f |
+    // |   |   |   |
+    // +-----------+
+    // |   |   |   |
+    // | g | h | i |
+    // |   |   |   |
+    // +-----------+
+
+    float a = getElevation(v_pos + vec2(-epsilon.x, -epsilon.y), 0.0);
+    float b = getElevation(v_pos + vec2(0, -epsilon.y), 0.0);
+    float c = getElevation(v_pos + vec2(epsilon.x, -epsilon.y), 0.0);
+    float d = getElevation(v_pos + vec2(-epsilon.x, 0), 0.0);
+    float e = getElevation(v_pos, 0.0);
+    float f = getElevation(v_pos + vec2(epsilon.x, 0), 0.0);
+    float g = getElevation(v_pos + vec2(-epsilon.x, epsilon.y), 0.0);
+    float h = getElevation(v_pos + vec2(0, epsilon.y), 0.0);
+    float i = getElevation(v_pos + vec2(epsilon.x, epsilon.y), 0.0);
+
+    // here we divide the x and y slopes by 8 * pixel size
+    // where pixel size (aka meters/pixel) is:
+    // circumference of the world / (pixels per tile * number of tiles)
+    // which is equivalent to: 8 * 40075016.6855785 / (512 * pow(2, u_zoom))
+    // which can be reduced to: pow(2, 19.25619978527 - u_zoom)
+    // we want to vertically exaggerate the hillshading though, because otherwise
+    // it is barely noticeable at low zooms. to do this, we multiply this by some
+    // scale factor pow(2, (u_zoom - u_maxzoom) * a) where a is an arbitrary value
+    // Here we use a=0.3 which works out to the expression below. see 
+    // nickidlugash's awesome breakdown for more info
+    // https://github.com/mapbox/mapbox-gl-js/pull/5286#discussion_r148419556
+    float exaggeration = u_zoom < 2.0 ? 0.4 : u_zoom < 4.5 ? 0.35 : 0.3;
+
+    vec2 deriv = vec2(
+        (c + f + f + i) - (a + d + d + g),
+        (g + h + h + i) - (a + b + b + c)
+    ) /  pow(2.0, (u_zoom - u_maxzoom) * exaggeration + 19.2562 - u_zoom);
+
+    gl_FragColor = clamp(vec4(
+        deriv.x / 2.0 + 0.5,
+        deriv.y / 2.0 + 0.5,
+        1.0,
+        1.0), 0.0, 1.0);
+
+#ifdef OVERDRAW_INSPECTOR
+    gl_FragColor = vec4(1.0);
+#endif
+}
+
+*/
+