Fix D_GGX code which can cause divide-by-zero val

When given roughness is lower than 0.01, d value in original code will
be zero. This can make last return value as NAN because of
divide-by-zero. This is well addressed in issue #56373.

Modified code is referenced on D_GGX function of google/filament
(https://github.com/google/filament/blob/main/shaders/src/brdf.fs#L54-L79)

Signed-off-by: snowapril <[email protected]>

servers/rendering/renderer_rd/shaders/scene_forward_lights_inc.glsl

@@ -1,9 +1,9 @@
 // Functions related to lighting
 
 float D_GGX(float cos_theta_m, float alpha) {
-	float alpha2 = alpha * alpha;
-	float d = 1.0 + (alpha2 - 1.0) * cos_theta_m * cos_theta_m;
-	return alpha2 / (M_PI * d * d);
+	float a = cos_theta_m * alpha;
+	float k = alpha / (1.0 - cos_theta_m * cos_theta_m + a * a);
+	return k * k * (1.0 / M_PI);
 }
 
 // From Earl Hammon, Jr. "PBR Diffuse Lighting for GGX+Smith Microsurfaces" https://www.gdcvault.com/play/1024478/PBR-Diffuse-Lighting-for-GGX