When I was implementing the WebGPU extension depth_clamp on D3D12, I found D3D12_RASTERIZER_DESC.DepthClipEnable doesn't have the same functionality as VkPipelineRasterizationStateCreateInfo::depthClampEnable, so we cannot implement depth_clamping with D3D12_RASTERIZER_DESC.DepthClipEnable.

• According to Vulkan SPEC, "if VkPipelineRasterizationStateCreateInfo::depthClampEnable is enabled, before the sample’s zf is compared to za,zf is clamped to [min(n,f),max(n,f)], where n and f are the minDepth and maxDepth depth range values of the viewport used by this fragment, respectively", which implies in Vulkan "depthClamp" is done after fragment shader stage: clamp the depth value of the output fragment shader after the execution of fragment shader.

• According to D3D12 document, "the hardware always performs x and y clipping of rasterized coordinates. When DepthClipEnable is set to the default–TRUE, the hardware also clips the z value ", which means in D3D12 depthClip is executed before the pixel shader stage: discard the specific coordinates after rasterization.

I think the right way to implement depth_clamp is injecting the following at the end of fragment shaders, what do you think?

gl_FragDepth = clamp(0, 1, gl_FragDepth);