The results of Roughness=1.0 from a constant and from a parameter are different on 4.20.

The results of Roughness=1.0 from a constant and from a parameter are different on 4.20.

The static bool switch of the following material always returns 1.0. But the results are different.
[Link Removed]

[Link Removed]

From 4.20, Statically constant 1.0 roughness is compiled as a "Fully Rough" material. But if the roughness value can't be decided at material compiling time, it isn't compiled as a "Fully Rough". As a result, generated shader codes are different. "Fully Rough" sets FORCE_FULLY_ROUGH to 1.

(BasePassPixelShader.usf)

    #if FORCE_FULLY_ROUGH
		// Factors derived from EnvBRDFApprox( SpecularColor, 1, 1 ) == SpecularColor * 0.4524 - 0.0024
		GBuffer.DiffuseColor += GBuffer.SpecularColor * 0.45;
		GBuffer.SpecularColor = 0;
		GBuffer.Roughness = 1;
	#endif

This happens from 4.20. Because the following code was added to 4.20 from Fortnite(CL:3883284).

// Fully rough if we have a roughness code chunk and it's constant and evaluates to 1.
bIsFullyRough = Chunk[MP_Roughness] != INDEX_NONE && IsMaterialPropertyUsed(MP_Roughness, Chunk[MP_Roughness], FLinearColor(1, 0, 0, 0), 1) == false;

I think the results shouldn't be different so I report this as a bug and attached a repro project.

1. Open the attached project.
2. Build light
3. You can see two planes are lit differently even they have the same roughness (1.0).