Databases
StackOverflow2013
Postsdbo
POHLSL buffer stride and threading - what is happening here?
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POHLSL buffer stride and threading - what is happening here?
    primarykey
    data
    text
    <p>I'm really new to DirectCompute technologies, and have been attempting to learn from the documentation on the msdn website, which is.. dense, to say the least.</p> <p>I'd like to make a basic hlsl file that takes in a 4x4 matrix and a 4xN matrix and returns the multiplied result. But after spending some time playing with the code, I've found some weird stuff I don't understand - mainly with how the threads I pass in process the buffers and output data.</p> <p>With all of these examples, I pass in two 16 float buffers and get out a 16 float buffer and then Dispatch with a 4x1x1 grouping - I can show you code, but I honestly dont yet know what would help you help me. Let me know if there's a section of my C++ code you want to see.</p> <p>with the following code:</p> <pre><code>StructuredBuffer&lt;float4x4&gt; base_matrix : register(t0); // byteWidth = 64 StructuredBuffer&lt;float4&gt; extended_matrix : register(t1); // byteWidth = 64 RWStructuredBuffer&lt;float4&gt; BufferOut : register(u0); // byteWidth = 64, zeroed out before reading from the GPU [numthreads(1, 1, 1)] void CSMain( uint3 DTid : SV_DispatchThreadID ) { BufferOut[DTid.x].x = 1; } </code></pre> <p>I get the following values out:</p> <pre><code>1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 </code></pre> <p>This makes sense to me - the buffer is parsed as 4 threads, each executing 1 float4 grouping.</p> <p>with the following code:</p> <pre><code>StructuredBuffer&lt;float4x4&gt; base_matrix : register(t0); // byteWidth = 64 StructuredBuffer&lt;float4&gt; extended_matrix : register(t1); // byteWidth = 64 RWStructuredBuffer&lt;float4&gt; BufferOut : register(u0); // byteWidth = 64, zeroed out before reading from the GPU [numthreads(1, 1, 1)] void CSMain( uint3 DTid : SV_DispatchThreadID ) { BufferOut[DTid.x].x = 1; BufferOut[DTid.x].y = 2; BufferOut[DTid.x].z = 3; BufferOut[DTid.x].w = 4; } </code></pre> <p>I get the following values out:</p> <pre><code>1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 </code></pre> <p>and with the actual code I want to run:</p> <pre><code>StructuredBuffer&lt;float4x4&gt; base_matrix : register(t0); StructuredBuffer&lt;float4&gt; extended_matrix : register(t1); RWStructuredBuffer&lt;float4&gt; BufferOut : register(u0); [numthreads(1, 1, 1)] void CSMain( uint3 DTid : SV_DispatchThreadID ) { BufferOut[DTid.x] = mul(base_matrix[0],extended_matrix[DTid.x]) } </code></pre> <p>I get the following values out:</p> <pre><code>0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 </code></pre> <p>I can tell I'm missing a critical thing here, but for the life of me I cant find the appropriate documentation telling me how these work. Could someone help me understand whats going on in this code?</p> <p>Thanks for your time,</p> <p>Zach</p> <p>As another note, this code was cribbed together using the Microsoft DirectX SDK (June 2010)\Samples\C++\Direct3D11\BasicCompute11 Sample available. If I'm doing something terribly wrong, feel free to let me know. I'm REALLY new at HLSL.</p> <p>Edit: My buffer creation code.</p> <pre><code>CreateStructuredBuffer( g_pDevice, sizeof(float)*16, 1, g_matrix, &amp;g_pBuf0 ); CreateStructuredBuffer( g_pDevice, sizeof(float)*4, NUM_ELEMENTS, g_extended_matrix, &amp;g_pBuf1 ); CreateStructuredBuffer( g_pDevice, sizeof(float)*4, NUM_ELEMENTS, NULL, &amp;g_pBufResult ); //-------------------------------------------------------------------------------------- // Create Structured Buffer //-------------------------------------------------------------------------------------- HRESULT CreateStructuredBuffer( ID3D11Device* pDevice, UINT uElementSize, UINT uCount, VOID* pInitData, ID3D11Buffer** ppBufOut ) { *ppBufOut = NULL; D3D11_BUFFER_DESC desc; ZeroMemory( &amp;desc, sizeof(desc) ); desc.BindFlags = D3D11_BIND_UNORDERED_ACCESS | D3D11_BIND_SHADER_RESOURCE; desc.ByteWidth = uElementSize * uCount; desc.MiscFlags = D3D11_RESOURCE_MISC_BUFFER_STRUCTURED; desc.StructureByteStride = uElementSize; if ( pInitData ) { D3D11_SUBRESOURCE_DATA InitData; InitData.pSysMem = pInitData; return pDevice-&gt;CreateBuffer( &amp;desc, &amp;InitData, ppBufOut ); } else return pDevice-&gt;CreateBuffer( &amp;desc, NULL, ppBufOut ); } </code></pre> <p>Trying .1,.2,.3,.4 ...</p> <pre><code>StructuredBuffer&lt;float4x4&gt; base_matrix : register(t0); StructuredBuffer&lt;float4&gt; extended_matrix : register(t1); StructuredBuffer&lt;uint&gt; loop_multiplier : register(t2); RWStructuredBuffer&lt;float4&gt; BufferOut : register(u0); [numthreads(1, 1, 1)] void CSMain( uint3 DTid : SV_DispatchThreadID ) { BufferOut[DTid.x].x = .1; BufferOut[DTid.x].y = .2; BufferOut[DTid.x].z = .3; BufferOut[DTid.x].w = .4; } </code></pre> <p>got this out:</p> <pre><code>0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 </code></pre>
    singulars
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PTQuestion
    1. USZach H
    1. USZach H
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. VO
      singulars
      1. POHLSL buffer stride and threading - what is happening here?
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POHLSL buffer stride and threading - what is happening here?
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POHLSL buffer stride and threading - what is happening here?
      1. This table or related slice is empty.
      1. VTUpMod
    1. COCould you post the code where you create your ID3D11Buffers and the corresponding ID3D11ShaderResourceView and ID3D11UnorderedAccessView objects, including the contents of the various *_DESC objects used to create them? Also, random hunch: can you try writing out 0.1, 0.2, 0.3, 0.4 instead of 1,2,3,4 in the second example?
      singulars
      1. POHLSL buffer stride and threading - what is happening here?
      1. USpostgoodism
    2. COI'll post this in the morning (eta 8 hours). Didnt catch this until later. But thanks in advance!
      singulars
      1. POHLSL buffer stride and threading - what is happening here?
      1. USZach H
    3. CO@postgoodism : Added my buffer creation code. Let me know if you need anything else.
      singulars
      1. POHLSL buffer stride and threading - what is happening here?
      1. USZach H
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload