Clamp chroma texcoords to avoid sampling alignment padding

This resolves #885 without the massive perf hit on lower end Intel GPUs.

app/shaders/d3d11_video_pixel_end.hlsli

@@ -1,7 +1,8 @@
 min16float4 main(ShaderInput input) : SV_TARGET
 {
+    // Clamp the chrominance texcoords to avoid sampling the row of texels adjacent to the alignment padding
     min16float3 yuv = min16float3(luminancePlane.Sample(theSampler, input.tex),
-                                  chrominancePlane.Sample(theSampler, input.tex));
+                                  chrominancePlane.Sample(theSampler, min(input.tex, chromaTexMax.rg)));
 
     // Subtract the YUV offset for limited vs full range
     yuv -= offsets;