Direct3D 12 / PIX - Pixel Shader not getting invoked

Question

I'm very new to D3D12-programming, and is just beginning to set up a little dummy-project to get the hang of the general process.

However, I have gotten stuck.

Currently, when debugging my little program in PIX, I can confirm that the Vertex Shader is running as it should, but the Pixel Shader does not get invoked.

I have tried checking my Scissor, Viewports and Depthbuffer, to confirm that the vertices does not get clipped, but I have not managed to solve the issue.

Does anyone have any idea what to check for here?

Thank you in advance!

Pipeline state in PIX Vertex shader output in PIX

A few code fragments which i think is relevant:

(Rasterizer state, DepthStencil state, Swapchain creation, RTV creation)

    D3D12_RASTERIZER_DESC rasterizer{};

    rasterizer.FillMode = D3D12_FILL_MODE_SOLID;
    rasterizer.CullMode = D3D12_CULL_MODE_NONE;
    rasterizer.FrontCounterClockwise = FALSE;
    rasterizer.DepthClipEnable = TRUE; 
    rasterizer.MultisampleEnable = FALSE;
    rasterizer.ConservativeRaster = D3D12_CONSERVATIVE_RASTERIZATION_MODE_OFF;

    pipelineDesc.RasterizerState = rasterizer;

    D3D12_DEPTH_STENCIL_DESC depthStencil{};

    depthStencil.DepthEnable = TRUE; 
    depthStencil.DepthWriteMask = D3D12_DEPTH_WRITE_MASK_ALL;
    depthStencil.DepthFunc = D3D12_COMPARISON_FUNC_LESS;
    depthStencil.StencilEnable = FALSE;
    depthStencil.FrontFace.StencilFailOp = D3D12_STENCIL_OP_KEEP;
    depthStencil.BackFace.StencilFailOp = D3D12_STENCIL_OP_KEEP;
    depthStencil.FrontFace.StencilPassOp = D3D12_STENCIL_OP_KEEP;
    depthStencil.BackFace.StencilPassOp = D3D12_STENCIL_OP_KEEP;
    depthStencil.FrontFace.StencilFunc = D3D12_COMPARISON_FUNC_ALWAYS;
    depthStencil.BackFace.StencilFunc = D3D12_COMPARISON_FUNC_ALWAYS;


    DXGI_SWAP_CHAIN_DESC1 swapchainDescription{};
    DXGI_SWAP_CHAIN_FULLSCREEN_DESC fullscreenDescription{};

    swapchainDescription.Width = 0;
    swapchainDescription.Height = 0;
    swapchainDescription.Format = DXGI_FORMAT_R16G16B16A16_FLOAT;
    swapchainDescription.Stereo = FALSE;
    swapchainDescription.SampleDesc = { 1, 0 };
    swapchainDescription.BufferUsage = DXGI_USAGE_RENDER_TARGET_OUTPUT;
    swapchainDescription.BufferCount = 2;
    swapchainDescription.Scaling = DXGI_SCALING_NONE;
    swapchainDescription.SwapEffect = DXGI_SWAP_EFFECT_FLIP_DISCARD;
    swapchainDescription.AlphaMode = DXGI_ALPHA_MODE_IGNORE;
    swapchainDescription.Flags = 0;

    fullscreenDescription.RefreshRate = { 1, 60 };
    fullscreenDescription.ScanlineOrdering = DXGI_MODE_SCANLINE_ORDER_PROGRESSIVE;
    fullscreenDescription.Scaling = DXGI_MODE_SCALING_CENTERED;
    fullscreenDescription.Windowed = TRUE;


    CheckHresult(factory->CreateSwapChainForHwnd(commandQueue, hwnd, &swapchainDescription, &fullscreenDescription, NULL, &t_swapchain));


    HRESULT sres = t_swapchain->QueryInterface(IID_PPV_ARGS(&swapchain));


    D3D12_DESCRIPTOR_HEAP_DESC descriptorHeapInfoRtv{};

    descriptorHeapInfoRtv.Type = D3D12_DESCRIPTOR_HEAP_TYPE_RTV;
    descriptorHeapInfoRtv.Flags = D3D12_DESCRIPTOR_HEAP_FLAG_NONE;
    descriptorHeapInfoRtv.NumDescriptors = 2;
    descriptorHeapInfoRtv.NodeMask = 0;


    device->CreateDescriptorHeap(&descriptorHeapInfoRtv, IID_PPV_ARGS(&rtv_descriptorHeap.descriptorHeap));
    rtv_descriptorHeap.incrementSize = device->GetDescriptorHandleIncrementSize(descriptorHeapInfoRtv.Type);
    rtv_descriptorHeap.startHandle = rtv_descriptorHeap.descriptorHeap->GetCPUDescriptorHandleForHeapStart();

    device->CreateRenderTargetView(swapchainBuffers[0], NULL, rtv_descriptorHeap.startHandle);
    device->CreateRenderTargetView(swapchainBuffers[1], NULL, rtv_descriptorHeap.AddOffset(1));

Shader code:

// Vertex Shader

    struct VS_IN
    {
        float4 pos : POSITION;
    };


    struct VS_OUT
    {

        float4 pos : SV_Position;
    
    };
    

    VS_OUT main(VS_IN inp) 
    {
        VS_OUT outp;
    
        outp.pos = float4(inp.pos.x, inp.pos.y, inp.pos.z, 1.0f);
    
        return outp;
    }




// Pixel Shader

    struct PS_OUT
    {

        float4 col_in : COLOR;
    
    };

    struct VS_OUT
    {

        float4 pos : SV_Position;
    
    };

    float4 main(VS_OUT inp) : SV_Target
    {
    

    
        PS_OUT outp;
    
        outp.col_in = float4(1.0f, 1.0f, 1.0f, 1.0f);
   
    
        return outp.col_in;
    }

I have tried using both default settings (passing NULL in device->CreateRenderTargetView) and also passing in a custom description. I have tried disabling depth testing, and changed the pipeline states samplemask.

Answer 1

Found the problem: I called commandList->ClearDepthStencilView(dsv_descriptorHeap.startHandle, D3D12_CLEAR_FLAG_DEPTH | D3D12_CLEAR_FLAG_STENCIL, 1.0f, 0, 0, &rect); With the 1.0f, 0, 0 parameters in the wrong order. I had them 0, 1, 0, thus giving the effect of clearing the depth buffer to the wrong value, presumably causing my pixels to fail the depth test.