void ProgramSettings::readLocalStrainCalculatorSettings(const libconfig::Setting& stg)
{
m_engineSettings.localStrainCalculatorSettings->hstep = stg["hstep"];
m_engineSettings.localStrainCalculatorSettings->xrange = readRange(stg["xrange"]);
m_engineSettings.localStrainCalculatorSettings->yrange = readRange(stg["yrange"]);
m_engineSettings.localStrainCalculatorSettings->zrange = readRange(stg["zrange"]);
}
void ProgramSettings::readCoplanarCorrelationCalculatorSettings(const libconfig::Setting& stg)
{
m_engineSettings.coplanarCorrelationCalculatorSettings->Q = readMillerReciprocalHexIndices(stg["Q"]);
m_engineSettings.coplanarCorrelationCalculatorSettings->nbsteps = stg["nbsteps"];
m_engineSettings.coplanarCorrelationCalculatorSettings->xrange = readRange(stg["xrange"]);
m_engineSettings.coplanarCorrelationCalculatorSettings->z1range = readRange(stg["z1range"]);
m_engineSettings.coplanarCorrelationCalculatorSettings->z2range = readRange(stg["z2range"]);
}
void ProgramSettings::readCoplanarIntensityCalculatorSettings(const libconfig::Setting& stg)
{
m_engineSettings.coplanarIntensityCalculatorSettings->Q = readMillerReciprocalHexIndices(stg["Q"]);
m_engineSettings.coplanarIntensityCalculatorSettings->nbsteps = stg["nbsteps"];
m_engineSettings.coplanarIntensityCalculatorSettings->sigma_x = stg["sigma_x"];
m_engineSettings.coplanarIntensityCalculatorSettings->sigma_z = stg["sigma_z"];
m_engineSettings.coplanarIntensityCalculatorSettings->precision = stg["precision"];
m_engineSettings.coplanarIntensityCalculatorSettings->qxrange = readRange(stg["qxrange"]);
m_engineSettings.coplanarIntensityCalculatorSettings->qzrange = readRange(stg["qzrange"]);
}
void ProgramSettings::readLocalDisplacementCalculatorSettings(const libconfig::Setting& stg)
{
if (stg.exists("input"))
{
m_engineSettings.localDisplacementCalculatorSettings->infile = stg["input"].c_str();
}else
{
m_engineSettings.localDisplacementCalculatorSettings->infile = "";
m_engineSettings.localDisplacementCalculatorSettings->xrange =
readRange(stg["xrange"]);
m_engineSettings.localDisplacementCalculatorSettings->yrange =
readRange(stg["yrange"]);
m_engineSettings.localDisplacementCalculatorSettings->zrange =
readRange(stg["zrange"]);
}
}
FrameResource::FrameResource(ID3D12Device* pDevice, UINT cityRowCount, UINT cityColumnCount) :
m_fenceValue(0),
m_cityRowCount(cityRowCount),
m_cityColumnCount(cityColumnCount)
{
m_modelMatrices.resize(m_cityRowCount * m_cityColumnCount);
// The command allocator is used by the main sample class when
// resetting the command list in the main update loop. Each frame
// resource needs a command allocator because command allocators
// cannot be reused until the GPU is done executing the commands
// associated with it.
ThrowIfFailed(pDevice->CreateCommandAllocator(D3D12_COMMAND_LIST_TYPE_DIRECT, IID_PPV_ARGS(&m_commandAllocator)));
ThrowIfFailed(pDevice->CreateCommandAllocator(D3D12_COMMAND_LIST_TYPE_BUNDLE, IID_PPV_ARGS(&m_bundleAllocator)));
// Create an upload heap for the constant buffers.
ThrowIfFailed(pDevice->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(sizeof(SceneConstantBuffer) * m_cityRowCount * m_cityColumnCount),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_cbvUploadHeap)));
// Map the constant buffers. Note that unlike D3D11, the resource
// does not need to be unmapped for use by the GPU. In this sample,
// the resource stays 'permenantly' mapped to avoid overhead with
// mapping/unmapping each frame.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_cbvUploadHeap->Map(0, &readRange, reinterpret_cast<void**>(&m_pConstantBuffers)));
// Update all of the model matrices once; our cities don't move so
// we don't need to do this ever again.
SetCityPositions(8.0f, -8.0f);
}
void RezkonvImporter::loadIngredient( const QString &string, Recipe &recipe, bool &is_sub )
{
Ingredient new_ingredient;
new_ingredient.amount = 0; //amount not required, so give default of 0
QRegExp cont_test( "^-{1,2}" );
if ( string.trimmed().contains( cont_test ) ) {
QString name = string.trimmed();
name.remove( cont_test );
kDebug() << "Appending to last ingredient: " << name ;
if ( !recipe.ingList.isEmpty() ) //so it doesn't crash when the first ingredient appears to be a continuation of another
recipe.ingList.last().name += ' ' + name;
return ;
}
//amount
if ( !string.mid( 0, 7 ).trimmed().isEmpty() )
readRange( string.mid( 0, 7 ), new_ingredient.amount, new_ingredient.amount_offset );
//unit
QString unit_str = string.mid( 8, 9 ).trimmed();
new_ingredient.units = Unit( unit_str, new_ingredient.amount );
//name and preparation method
new_ingredient.name = string.mid( 18, string.length() - 18 ).trimmed();
//separate out the preparation method
QString name_and_prep = new_ingredient.name;
int separator_index = name_and_prep.indexOf( "," );
if ( separator_index != -1 ) {
new_ingredient.name = name_and_prep.mid( 0, separator_index ).trimmed();
new_ingredient.prepMethodList = ElementList::split(",",name_and_prep.mid( separator_index + 1, name_and_prep.length() ).trimmed() );
}
//header (if present)
new_ingredient.group = current_header;
bool last_is_sub = is_sub;
if ( !new_ingredient.prepMethodList.isEmpty() &&
new_ingredient.prepMethodList.last().name == "or" ) {
new_ingredient.prepMethodList.pop_back();
is_sub = true;
}
else
is_sub = false;
if ( last_is_sub )
recipe.ingList.last().substitutes.append(new_ingredient);
else
recipe.ingList.append( new_ingredient );
}
void StringReader :: readRangeList(dynaList< Range > &list, const char *source, const char *idString)
{
int li, hi;
const char *str1, *helpSource = source;
// Range* range;
// find first valid occurrence of idString
int len = strlen(idString);
do {
if ( ( str1 = strstr(helpSource, idString) ) == NULL ) {
return;
}
helpSource = str1 + 1;
} while ( !( isspace( * ( helpSource + len - 1 ) ) ) );
helpSource = str1 + len;
// find first non whitespace character
// skip whitespaces
while ( isspace(* helpSource) ) {
helpSource++;
}
// test if list left bracketed found
if ( * helpSource != '{' ) {
OOFEM_WARNING("StringReader::readRangeList: parse error - missing left '{'");
list.clear();
return;
}
helpSource++;
// read ranges
while ( readRange(& helpSource, li, hi) ) {
Range range(li, hi);
list.pushBack(range);
}
// skip whitespaces after last range
while ( isspace(* helpSource) ) {
helpSource++;
}
// test for enclosing bracket
if ( * helpSource != '}' ) {
OOFEM_WARNING("StringReader::readRangeList: parse error - missing end '}'");
list.clear();
return;
}
}
void DynamicConstantBuffer::Init(ID3D12Device* pDevice)
{
const UINT bufferSize = m_perFrameConstantBufferSize * m_frameCount;
ThrowIfFailed(pDevice->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(bufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_constantBuffer)
));
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_constantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pMappedConstantBuffer)));
}
//コンスタンスバッファ作成
void CBufferDraw::Impl::CreateConstantBuffer()
{
auto Dev = App::GetApp()->GetDeviceResources();
ThrowIfFailed(Dev->GetDevice()->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(sizeof(SpriteConstantBuffer)),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_ConstantBufferUploadHeap)),
L"コンスタントバッファ用のアップロードヒープ作成に失敗しました",
L"Dev->GetDevice()->CreateCommittedResource()",
L"TriangleSprite::CreateConstantBuffer()"
);
//コンスタントバッファのビューを作成
//TODO : デスクリプタとビューの違いについて
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.BufferLocation = m_ConstantBufferUploadHeap->GetGPUVirtualAddress();
//コンスタントバッファは256バイトにアラインメント
cbvDesc.SizeInBytes = (sizeof(SpriteConstantBuffer) + 255) & ~255;
//コンスタントバッファビューを作成すべきデスクプリタヒープ上のハンドルを取得
//シェーダリソースがある場合コンスタントバッファはシェーダリソースビューのあとに設置する
CD3DX12_CPU_DESCRIPTOR_HANDLE cbvSrvHandle(
m_CbvSrvUavDescriptorHeap->GetCPUDescriptorHandleForHeapStart(),
0,
0
);
Dev->GetDevice()->CreateConstantBufferView(&cbvDesc, cbvSrvHandle);
//コンスタントバッファのアップロードヒープのマップ
CD3DX12_RANGE readRange(0, 0);
ThrowIfFailed(m_ConstantBufferUploadHeap->Map(0, &readRange, reinterpret_cast<void**>(&m_pConstantBuffer)),
L"コンスタントバッファのマップに失敗しました",
L"pImpl->m_ConstantBufferUploadHeap->Map()",
L"TriangleSprite::CreateConstantBuffer()"
);
}
void RezkonvImporter::readRecipe( const QStringList &raw_recipe )
{
kapp->processEvents(); //don't want the user to think its frozen... especially for files with thousands of recipes
Recipe recipe;
QStringList::const_iterator text_it = raw_recipe.begin();
m_end_it = raw_recipe.end();
//title (Titel)
text_it++;
recipe.title = ( *text_it ).mid( ( *text_it ).indexOf( ":" ) + 1, ( *text_it ).length() ).trimmed();
kDebug() << "Found title: " << recipe.title ;
//categories (Kategorien):
text_it++;
QStringList categories;
if ( ( *text_it ).mid( ( *text_it ).indexOf( ":" ) + 1, ( *text_it ).length() ).isEmpty() )
categories = QStringList();
else
categories = ( *text_it ).mid( ( *text_it ).indexOf( ":" ) + 1, ( *text_it ).length() ).split( ',', QString::SkipEmptyParts );
for ( QStringList::const_iterator it = categories.constBegin(); it != categories.constEnd(); ++it ) {
Element new_cat;
new_cat.name = QString( *it ).trimmed();
kDebug() << "Found category: " << new_cat.name ;
recipe.categoryList.append( new_cat );
}
//yield (Menge)
text_it++;
//get the number between the ":" and the next space after it
QString yield_str = ( *text_it ).trimmed();
yield_str.remove( QRegExp( "^Menge:\\s*" ) );
int sep_index = yield_str.indexOf( ' ' );
if ( sep_index != -1 )
recipe.yield.setType(yield_str.mid( sep_index+1 ));
double amount = 0.0, amountOffset = 0.0;
readRange( yield_str.mid( 0, sep_index ), amount, amountOffset );
recipe.yield.setAmount(amount);
recipe.yield.setAmountOffset(amountOffset);
kDebug() << "Found yield: " << recipe.yield.amount();
bool is_sub = false;
bool last_line_empty = false;
text_it++;
while ( text_it != raw_recipe.end() ) {
if ( ( *text_it ).isEmpty() ) {
last_line_empty = true;
text_it++;
continue;
}
if ( ( *text_it ).contains( QRegExp( "^=====.*=$" ) ) ) //is a header
{
if ( ( *text_it ).contains( "quelle", Qt::CaseInsensitive ) )
{
loadReferences( text_it, recipe );
break; //reference lines are the last before the instructions
}
else
loadIngredientHeader( *text_it, recipe );
}
//if it has no more than two spaces followed by a non-digit
//then we'll assume it is a direction line
else if ( last_line_empty && ( *text_it ).contains( QRegExp( "^\\s{0,2}[^\\d\\s=]" ) ) )
break;
else
loadIngredient( *text_it, recipe, is_sub );
last_line_empty = false;
text_it++;
}
loadInstructions( text_it, recipe );
add
( recipe );
current_header.clear();
}
void D3D12Fullscreen::LoadSizeDependentResources()
{
m_viewport.Width = static_cast<float>(m_width);
m_viewport.Height = static_cast<float>(m_height);
m_viewport.MaxDepth = 1.0f;
m_scissorRect.right = static_cast<LONG>(m_width);
m_scissorRect.bottom = static_cast<LONG>(m_height);
// Create frame resources.
{
CD3DX12_CPU_DESCRIPTOR_HANDLE rtvHandle(m_rtvHeap->GetCPUDescriptorHandleForHeapStart());
// Create a RTV for each frame.
for (UINT n = 0; n < FrameCount; n++)
{
ThrowIfFailed(m_swapChain->GetBuffer(n, IID_PPV_ARGS(&m_renderTargets[n])));
m_device->CreateRenderTargetView(m_renderTargets[n].Get(), nullptr, rtvHandle);
rtvHandle.Offset(1, m_rtvDescriptorSize);
WCHAR name[25];
if (swprintf_s(name, L"m_renderTargets[%u]", n) > 0)
{
SetName(m_renderTargets[n].Get(), name);
}
}
}
// Create/update the vertex buffer. When updating the vertex buffer it is important
// to ensure that the GPU is finished using the resource before it is released.
// The OnSizeChanged method waits for the GPU to be idle before this method is
// called.
{
// Define the geometry for a triangle that stays the same size regardless
// of the window size. This is not the recommended way to transform vertices.
// The same effect could be achieved by using constant buffers and
// transforming a static set of vertices in the vertex shader, but this
// sample merely demonstrates modifying a resource that is tied to the render
// target size.
// Other apps might also resize intermediate render targets or depth stencils
// at this time.
float x = TriangleWidth / m_viewport.Width;
float y = TriangleWidth / m_viewport.Height;
Vertex triangleVertices[] =
{
{ { 0.0f, y, 0.0f }, { 1.0f, 0.0f, 0.0f, 1.0f } },
{ { x, -y, 0.0f }, { 0.0f, 1.0f, 0.0f, 1.0f } },
{ { -x, -y, 0.0f }, { 0.0f, 0.0f, 1.0f, 1.0f } }
};
const UINT vertexBufferSize = sizeof(triangleVertices);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_vertexBufferUpload)));
NAME_D3D12_OBJECT(m_vertexBuffer);
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_vertexBufferUpload->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin)));
memcpy(pVertexDataBegin, triangleVertices, sizeof(triangleVertices));
m_vertexBufferUpload->Unmap(0, nullptr);
m_commandList->CopyBufferRegion(m_vertexBuffer.Get(), 0, m_vertexBufferUpload.Get(), 0, vertexBufferSize);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer views.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = vertexBufferSize;
}
m_resizeResources = false;
}
void Plink::displayGeneReport()
{
// Simply read in any generic results file and list of SNPs by
// ranges (which may be subsetted).
// if ( false )
// readMapFile(par::mapfile,include,include_pos,nl_actual);
ofstream GREP;
GREP.open( (par::output_file_name + ".range.report").c_str() , ios::out);
map<string, set<Range> > ranges;
// Read list of ranges
ranges = readRange( par::greport_gene_list );
// Filter ranges
if ( par::greport_subset )
ranges = filterRanges( ranges, par::greport_subset_file );
// Open a single results file
ifstream RESIN;
RESIN.open( par::greport_results.c_str() , ios::in );
// Read first (header) row
char cline[par::MAX_LINE_LENGTH];
RESIN.getline(cline,par::MAX_LINE_LENGTH,'\n');
string sline = cline;
if (sline=="")
error("Problem reading [ " + par::greport_results + " ]\n");
string buf;
stringstream ss(sline);
vector<string> tokens;
while (ss >> buf)
tokens.push_back(buf);
int chr_column = -1;
int bp_column = -1;
int pval_column = -1;
int snp_column = -1;
for (int i=0; i<tokens.size(); i++)
{
if ( tokens[i] == "CHR" )
chr_column = i;
if ( tokens[i] == "BP" )
bp_column = i;
if ( tokens[i] == "SNP" )
snp_column = i;
if ( tokens[i] == "P" )
pval_column = i;
}
// Do we have a list of SNPs to specifically extract?
set<string> extractSNP;
if ( par::extract_set )
{
if ( snp_column == -1 )
error("Did not find a SNP field, so cannot use --extract");
checkFileExists( par::extract_file );
PP->printLOG("Only extracting SNPs listed in [ " + par::extract_file + " ]\n");
ifstream IN(par::extract_file.c_str(), ios::in);
while ( ! IN.eof() )
{
string snpname;
IN >> snpname;
if ( snpname=="" )
continue;
extractSNP.insert(snpname);
}
IN.close();
PP->printLOG("Read " + int2str( extractSNP.size() ) + " SNPs to extract\n");
}
if ( chr_column < 0 || bp_column < 0 )
error("Could not find CHR and BP fields in results file");
map<Range*,vector<string> > annotatedResults;
string headerline = sline;
int cnt = 0;
while ( ! RESIN.eof() )
{
// if ( ! par::silent )
// cout << "Processing results line " << ++cnt << " \r";
// vector<string> tokens = tokenizeLine( RESIN );
char cline[par::MAX_LINE_LENGTH];
RESIN.getline(cline,par::MAX_LINE_LENGTH,'\n');
string sline = cline;
if (sline=="")
continue;
string buf;
stringstream ss(sline);
vector<string> tokens;
while (ss >> buf)
tokens.push_back(buf);
if ( tokens.size() <= chr_column ||
tokens.size() <= bp_column )
continue;
// Using a p-value-filtering field?
double pvalue = 0;
if ( pval_column != -1 )
{
if ( tokens.size() <= pval_column )
continue;
if ( ! from_string<double>( pvalue, tokens[pval_column] , std::dec))
continue;
if ( par::pfilter && pvalue > par::pfvalue )
continue;
}
if ( par::extract_set )
{
if ( tokens.size() <= snp_column )
continue;
if ( extractSNP.find( tokens[snp_column] ) == extractSNP.end() )
continue;
}
int thisChr = -1;
int thisBP = -1;
if ( ! from_string<int>( thisChr, tokens[chr_column] , std::dec))
continue;
if ( ! from_string<int>( thisBP, tokens[bp_column] , std::dec))
continue;
// Do we need to store this? i.e. what ranges is it actually in?
// This information is in snp2range
Range r1(thisChr,thisBP,thisBP,"dummy");
set<Range*> implicated = rangeIntersect(r1,ranges);
set<Range*>::iterator ri = implicated.begin();
while ( ri != implicated.end() )
{
string distance = dbl2str(( thisBP - ((*ri)->start + par::make_set_border)) /1000.00 , 4 ) + "kb" ;
if ( annotatedResults.find( *ri ) == annotatedResults.end() )
{
vector<string> t(2);
t[0] = distance;
t[1] = sline;
annotatedResults.insert(make_pair( (Range *)(*ri) , t ) );
}
else
{
vector<string> & v = annotatedResults.find( *ri )->second;
v.push_back(distance);
v.push_back(sline);
}
++ri;
}
// Read next line of results
}
// Iterate through these -- they will be in genomic order, hopefully
map<string, set<Range> >::iterator ri = ranges.begin();
while ( ri != ranges.end() )
{
set<Range>::iterator si = ri->second.begin();
while ( si != ri->second.end() )
{
bool displayed = false;
map<Range*,vector<string> >::iterator ari;
ari = annotatedResults.find( (Range *)&(*si) );
if ( ari != annotatedResults.end() )
{
for (int l=0; l< ari->second.size(); l+=2)
{
if ( ! displayed )
{
GREP << ri->first << " -- chr"
<< chromosomeName( si->chr ) << ":"
<< si->start << ".."
<< si->stop << " ( "
<< (si->stop - si->start ) / 1000.00 << "kb ) ";
if ( par::make_set_border > 0 )
GREP << " including " << par::make_set_border/1000.00 << "kb border ";
GREP << "\n\n"
<< setw(12) << "DIST" << " "
<< headerline << "\n";
displayed = true;
}
GREP << setw(12) << ari->second[l] << " "
<< ari->second[l+1] << "\n";
}
}
if ( ! displayed )
{
if ( par::greport_display_empty )
{
GREP << ri->first << " -- chr"
<< chromosomeName( si->chr ) << ":"
<< si->start << ".."
<< si->stop << " ( "
<< (si->stop - si->start ) / 1000.00 << "kb ) ";
if ( par::make_set_border > 0 )
GREP << " including " << par::make_set_border/1000.00 << "kb border ";
GREP << " { nothing to report }\n\n";
}
}
else
GREP << "\n\n";
++si;
}
++ri;
}
RESIN.close();
GREP.close();
if ( ! par::silent )
cout << "\n";
printLOG("Writing per-range report to [ "
+ par::output_file_name
+ ".range.report ]\n");
shutdown();
}
HRESULT DX12Framework::AllocateVersionedBuffer(Buffer** ppBuffer)
{
//
// Grab a buffer from our lookaside list if possible.
//
if (!IsListEmpty(&m_DynamicBufferListHead))
{
LIST_ENTRY* pEntry = RemoveHeadList(&m_DynamicBufferListHead);
*ppBuffer = static_cast<Buffer*>(pEntry);
return S_OK;
}
//
// No available buffers, let's try to allocate a new one.
//
HRESULT hr;
ID3D12Resource* pD3DBuffer = nullptr;
CD3DX12_RANGE readRange(0, 0);
void* pBaseAddress = nullptr;
hr = m_pDevice->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(DYNAMIC_BUFFER_SIZE),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&pD3DBuffer));
if (FAILED(hr))
{
LOG_WARNING("Failed to create D3D12 buffer resource, hr=0x%.8x", hr);
goto cleanup;
}
//
// Obtain a permanent CPU visible address for this buffer. In D3D12, it is okay to
// keep a permanent virtual address mapped to an allocation while the resource is
// accessed by the GPU. In this case, the semantics of the mapping operation are
// always equivalent to the D3D11 MAP_WRITE_NO_OVERWRITE value.
//
hr = pD3DBuffer->Map(0, &readRange, &pBaseAddress);
if (FAILED(hr))
{
LOG_WARNING("Failed to map base address for buffer, hr=0x%.8x", hr);
goto cleanup;
}
Buffer* pBuffer = nullptr;
try
{
pBuffer = new Buffer();
}
catch (std::bad_alloc&)
{
LOG_WARNING("Failed to allocate dynamic buffer");
hr = E_OUTOFMEMORY;
goto cleanup;
}
pBuffer->pBaseAddress = (ULONG_PTR)pBaseAddress;
pBuffer->pBuffer = pD3DBuffer;
*ppBuffer = pBuffer;
return S_OK;
cleanup:
if (pBaseAddress)
{
pD3DBuffer->Unmap(0, nullptr);
}
SafeRelease(pD3DBuffer);
return hr;
}
void ProgramSettings::readMeanStrainCalculatorSettings(const libconfig::Setting& stg)
{
m_engineSettings.meanStrainCalculatorSettings->hstep = stg["hstep"];
m_engineSettings.meanStrainCalculatorSettings->nbsteps = stg["nbsteps"];
m_engineSettings.meanStrainCalculatorSettings->zrange = readRange(stg["zrange"]);
}
/**
***************************************************************************************************
* DX12ImageRenderer::CaptureImage
*
* @brief
* Convert a DX12 resource to a CPU-visible linear buffer of pixels. The data is filled
* in a user-provided CpuImage struct.
*
* IMPORTANT: Memory inside pImgOut is allocated on behalf of the caller, so it is their
* responsibility to free it.
*
* @return
* S_OK if successful.
***************************************************************************************************
*/
HRESULT DX12ImageRenderer::CaptureImage(
ID3D12Resource* pRes,
D3D12_RESOURCE_STATES prevState,
UINT newWidth,
UINT newHeight,
CpuImage* pImgOut,
bool bFlipX,
bool bFlipY)
{
HRESULT result = E_FAIL;
if ((pRes != nullptr) && (pImgOut != nullptr) && (newWidth > 0) && (newHeight > 0))
{
// Create temp assets
result = CreateCaptureAssets(pRes, newWidth, newHeight);
if (result == S_OK)
{
result = m_pCmdAllocator->Reset();
}
if (result == S_OK)
{
result = m_pCmdList->Reset(m_pCmdAllocator, m_pPipelineStateGraphics);
}
// Render work
if (result == S_OK)
{
// Set root sig
m_pCmdList->SetGraphicsRootSignature(m_pRootSignatureGraphics);
// Set descriptors and tables
ID3D12DescriptorHeap* ppHeaps[] = { m_pSrvUavCbHeap };
m_pCmdList->SetDescriptorHeaps(_countof(ppHeaps), ppHeaps);
CD3DX12_GPU_DESCRIPTOR_HANDLE srvHandle(m_pSrvUavCbHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterSRV, m_srvUavCbDescriptorSize);
CD3DX12_GPU_DESCRIPTOR_HANDLE uavHandle(m_pSrvUavCbHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterUAV, m_srvUavCbDescriptorSize);
CD3DX12_GPU_DESCRIPTOR_HANDLE cbHandle(m_pSrvUavCbHeap->GetGPUDescriptorHandleForHeapStart(), RootParameterCBV, m_srvUavCbDescriptorSize);
m_pCmdList->SetGraphicsRootDescriptorTable(RootParameterSRV, srvHandle);
m_pCmdList->SetGraphicsRootDescriptorTable(RootParameterUAV, uavHandle);
m_pCmdList->SetGraphicsRootDescriptorTable(RootParameterCBV, cbHandle);
// Set viewport
D3D12_VIEWPORT viewport = {};
viewport.Width = static_cast<float>(newWidth);
viewport.Height = static_cast<float>(newHeight);
viewport.MaxDepth = 1.0f;
m_pCmdList->RSSetViewports(1, &viewport);
// Set scissor
D3D12_RECT scissorRect = {};
scissorRect.right = static_cast<LONG>(newWidth);
scissorRect.bottom = static_cast<LONG>(newHeight);
m_pCmdList->RSSetScissorRects(1, &scissorRect);
// Update const buf
m_constantBufferData.rtWidth = newWidth;
m_constantBufferData.flipX = bFlipX ? 1 : 0;
m_constantBufferData.flipY = bFlipY ? 1 : 0;
CD3DX12_RANGE readRange(0, 0);
result = m_pConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin));
memcpy(m_pCbvDataBegin, &m_constantBufferData, sizeof(m_constantBufferData));
m_pConstantBuffer->Unmap(0, nullptr);
// Set RT
CD3DX12_CPU_DESCRIPTOR_HANDLE internalRtvHandle(m_pInternalRtvHeap->GetCPUDescriptorHandleForHeapStart());
m_config.pDevice->CreateRenderTargetView(m_pInternalRT, nullptr, internalRtvHandle);
m_pCmdList->OMSetRenderTargets(1, &internalRtvHandle, FALSE, nullptr);
// Record commands
m_pCmdList->ClearRenderTargetView(internalRtvHandle, ClearColor, 0, nullptr);
m_pCmdList->IASetPrimitiveTopology(D3D_PRIMITIVE_TOPOLOGY_TRIANGLELIST);
CD3DX12_RESOURCE_BARRIER barrier = {};
// Render full screen quad and write out UAV
barrier = CD3DX12_RESOURCE_BARRIER::Transition(pRes, prevState, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE);
m_pCmdList->ResourceBarrier(1, &barrier);
m_pCmdList->DrawInstanced(3, 1, 0, 0);
#if OVERWRITE_SRC_RES
barrier = CD3DX12_RESOURCE_BARRIER::Transition(pRes, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE, D3D12_RESOURCE_STATE_COPY_DEST);
m_pCmdList->ResourceBarrier(1, &barrier);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pInternalRT, D3D12_RESOURCE_STATE_RENDER_TARGET, D3D12_RESOURCE_STATE_COPY_SOURCE);
m_pCmdList->ResourceBarrier(1, &barrier);
m_pCmdList->CopyResource(pRes, m_pInternalRT);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pInternalRT, D3D12_RESOURCE_STATE_COPY_SOURCE, D3D12_RESOURCE_STATE_RENDER_TARGET);
m_pCmdList->ResourceBarrier(1, &barrier);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(pRes, D3D12_RESOURCE_STATE_COPY_DEST, prevState);
m_pCmdList->ResourceBarrier(1, &barrier);
#endif
// Copy UAV to CPU-visible buffer
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pPSWriteBuf, D3D12_RESOURCE_STATE_UNORDERED_ACCESS, D3D12_RESOURCE_STATE_COPY_SOURCE);
m_pCmdList->ResourceBarrier(1, &barrier);
m_pCmdList->CopyResource(m_pPSWriteBufReadBack, m_pPSWriteBuf);
barrier = CD3DX12_RESOURCE_BARRIER::Transition(m_pPSWriteBuf, D3D12_RESOURCE_STATE_COPY_SOURCE, D3D12_RESOURCE_STATE_UNORDERED_ACCESS);
m_pCmdList->ResourceBarrier(1, &barrier);
// Execute the command list
result = m_pCmdList->Close();
ID3D12CommandList* ppCommandLists[] = { m_pCmdList };
m_config.pCmdQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
WaitCmdListFinish();
// Read back UAV results
void* pUavData = nullptr;
result = m_pPSWriteBufReadBack->Map(0, &readRange, &pUavData);
if (result == S_OK)
{
const UINT totalBytes = newWidth * newHeight * BytesPerPixel;
pImgOut->pitch = newWidth * BytesPerPixel;
pImgOut->width = newWidth;
pImgOut->height = newHeight;
pImgOut->pData = new char[totalBytes];
memcpy(pImgOut->pData, pUavData, totalBytes);
m_pPSWriteBufReadBack->Unmap(0, nullptr);
}
}
// Free temp assets
FreeCaptureAssets();
}
return result;
}
bool D3DClass::Initialize(int screenHeight, int screenWidth, HWND hwnd, bool vsync, bool fullscreen)
{
D3D_FEATURE_LEVEL featureLevel;
HRESULT result;
D3D12_COMMAND_QUEUE_DESC commandQueueDesc;
IDXGIFactory4* factory;
IDXGIAdapter* adapter;
IDXGIOutput* adapterOutput;
unsigned int numModes, i, numerator, denominator, renderTargetViewDescriptorSize;
unsigned long long stringLength;
DXGI_MODE_DESC* displayModeList;
DXGI_ADAPTER_DESC adapterDesc;
DXGI_SWAP_CHAIN_DESC swapChainDesc;
IDXGISwapChain* swapChain;
D3D12_DESCRIPTOR_HEAP_DESC renderTargetViewHeapDesc;
D3D12_CPU_DESCRIPTOR_HANDLE renderTargetViewHandle;
// Store the vsync setting.
m_vsync_enabled = vsync;
// Set the feature level to DirectX 12.1 to enable using all the DirectX 12 features.
// Note: Not all cards support full DirectX 12, this feature level may need to be reduced on some cards to 12.0.
featureLevel = D3D_FEATURE_LEVEL_11_0;
// Create the Direct3D 12 device.
result = D3D12CreateDevice(NULL, featureLevel, __uuidof(ID3D12Device), (void**)&m_device);
if (FAILED(result))
{
MessageBox(hwnd, L"Could not create a DirectX 12.1 device. The default video card does not support DirectX 12.1.", L"DirectX Device Failure", MB_OK);
return false;
}
// Initialize the description of the command queue.
ZeroMemory(&commandQueueDesc, sizeof(commandQueueDesc));
// Set up the description of the command queue.
commandQueueDesc.Type = D3D12_COMMAND_LIST_TYPE_DIRECT;
commandQueueDesc.Priority = D3D12_COMMAND_QUEUE_PRIORITY_NORMAL;
commandQueueDesc.Flags = D3D12_COMMAND_QUEUE_FLAG_NONE;
commandQueueDesc.NodeMask = 0;
// Create the command queue.
m_device->CreateCommandQueue(&commandQueueDesc, __uuidof(ID3D12CommandQueue), (void**)&m_commandQueue);
// Create a DirectX graphics interface factory.
CreateDXGIFactory1(__uuidof(IDXGIFactory4), (void**)&factory);
// Use the factory to create an adapter for the primary graphics interface (video card).
factory->EnumAdapters(0, &adapter);
// Enumerate the primary adapter output (monitor).
adapter->EnumOutputs(0, &adapterOutput);
// Get the number of modes that fit the DXGI_FORMAT_R8G8B8A8_UNORM display format for the adapter output (monitor).
adapterOutput->GetDisplayModeList(DXGI_FORMAT_R8G8B8A8_UNORM, DXGI_ENUM_MODES_INTERLACED, &numModes, NULL);
// Create a list to hold all the possible display modes for this monitor/video card combination.
displayModeList = new DXGI_MODE_DESC[numModes];
// Now fill the display mode list structures.
adapterOutput->GetDisplayModeList(DXGI_FORMAT_R8G8B8A8_UNORM, DXGI_ENUM_MODES_INTERLACED, &numModes, displayModeList);
// Now go through all the display modes and find the one that matches the screen height and width.
// When a match is found store the numerator and denominator of the refresh rate for that monitor.
for (i = 0; i<numModes; i++)
{
if (displayModeList[i].Height == (unsigned int)screenHeight)
{
if (displayModeList[i].Width == (unsigned int)screenWidth)
{
numerator = displayModeList[i].RefreshRate.Numerator;
denominator = displayModeList[i].RefreshRate.Denominator;
}
}
}
// Get the adapter (video card) description.
adapter->GetDesc(&adapterDesc);
// Store the dedicated video card memory in megabytes.
m_videoCardMemory = (int)(adapterDesc.DedicatedVideoMemory / 1024 / 1024);
// Convert the name of the video card to a character array and store it.
wcstombs_s(&stringLength, m_videoCardDescription, 128, adapterDesc.Description, 128);
// Release the display mode list.
delete[] displayModeList;
displayModeList = nullptr;
// Release the adapter output.
adapterOutput->Release();
adapterOutput = nullptr;
// Release the adapter.
adapter->Release();
adapter = nullptr;
// Initialize the swap chain description.
ZeroMemory(&swapChainDesc, sizeof(swapChainDesc));
// Set the swap chain to use double buffering.
swapChainDesc.BufferCount = 2;
// Set the height and width of the back buffers in the swap chain.
swapChainDesc.BufferDesc.Height = screenHeight;
swapChainDesc.BufferDesc.Width = screenWidth;
// Set a regular 32-bit surface for the back buffers.
swapChainDesc.BufferDesc.Format = DXGI_FORMAT_R8G8B8A8_UNORM;
// Set the usage of the back buffers to be render target outputs.
swapChainDesc.BufferUsage = DXGI_USAGE_RENDER_TARGET_OUTPUT;
// Set the swap effect to discard the previous buffer contents after swapping.
swapChainDesc.SwapEffect = DXGI_SWAP_EFFECT_FLIP_DISCARD;
// Set the handle for the window to render to.
swapChainDesc.OutputWindow = hwnd;
// Set to full screen or windowed mode.
swapChainDesc.Windowed = !fullscreen;
// Set the refresh rate of the back buffer.
if (m_vsync_enabled)
{
swapChainDesc.BufferDesc.RefreshRate.Numerator = numerator;
swapChainDesc.BufferDesc.RefreshRate.Denominator = denominator;
}
else
{
swapChainDesc.BufferDesc.RefreshRate.Numerator = 0;
swapChainDesc.BufferDesc.RefreshRate.Denominator = 1;
}
// Turn multisampling off.
swapChainDesc.SampleDesc.Count = 1;
swapChainDesc.SampleDesc.Quality = 0;
// Set the scan line ordering and scaling to unspecified.
swapChainDesc.BufferDesc.ScanlineOrdering = DXGI_MODE_SCANLINE_ORDER_UNSPECIFIED;
swapChainDesc.BufferDesc.Scaling = DXGI_MODE_SCALING_UNSPECIFIED;
// Don't set the advanced flags.
swapChainDesc.Flags = 0;
// Finally create the swap chain using the swap chain description.
factory->CreateSwapChain(m_commandQueue, &swapChainDesc, &swapChain);
// Next upgrade the IDXGISwapChain to a IDXGISwapChain3 interface and store it in a private member variable named m_swapChain.
// This will allow us to use the newer functionality such as getting the current back buffer index.
swapChain->QueryInterface(__uuidof(IDXGISwapChain3), (void**)&m_swapChain);
// Clear pointer to original swap chain interface since we are using version 3 instead (m_swapChain).
swapChain = nullptr;
// Release the factory now that the swap chain has been created.
factory->Release();
factory = nullptr;
// Initialize the render target view heap description for the two back buffers.
ZeroMemory(&renderTargetViewHeapDesc, sizeof(renderTargetViewHeapDesc));
// Set the number of descriptors to two for our two back buffers. Also set the heap tyupe to render target views.
renderTargetViewHeapDesc.NumDescriptors = 2;
renderTargetViewHeapDesc.Type = D3D12_DESCRIPTOR_HEAP_TYPE_RTV;
renderTargetViewHeapDesc.Flags = D3D12_DESCRIPTOR_HEAP_FLAG_NONE;
// Create the render target view heap for the back buffers.
m_device->CreateDescriptorHeap(&renderTargetViewHeapDesc, __uuidof(ID3D12DescriptorHeap), (void**)&m_renderTargetViewHeap);
// Get a handle to the starting memory location in the render target view heap to identify where the render target views will be located for the two back buffers.
renderTargetViewHandle = m_renderTargetViewHeap->GetCPUDescriptorHandleForHeapStart();
// Get the size of the memory location for the render target view descriptors.
renderTargetViewDescriptorSize = m_device->GetDescriptorHandleIncrementSize(D3D12_DESCRIPTOR_HEAP_TYPE_RTV);
// Get a pointer to the first back buffer from the swap chain.
m_swapChain->GetBuffer(0, __uuidof(ID3D12Resource), (void**)&m_backBufferRenderTarget[0]);
// Create a render target view for the first back buffer.
m_device->CreateRenderTargetView(m_backBufferRenderTarget[0], NULL, renderTargetViewHandle);
// Increment the view handle to the next descriptor location in the render target view heap.
renderTargetViewHandle.ptr += renderTargetViewDescriptorSize;
// Get a pointer to the second back buffer from the swap chain.
m_swapChain->GetBuffer(1, __uuidof(ID3D12Resource), (void**)&m_backBufferRenderTarget[1]);
// Create a render target view for the second back buffer.
m_device->CreateRenderTargetView(m_backBufferRenderTarget[1], NULL, renderTargetViewHandle);
// Finally get the initial index to which buffer is the current back buffer.
m_bufferIndex = m_swapChain->GetCurrentBackBufferIndex();
// Create a command allocator.
m_device->CreateCommandAllocator(D3D12_COMMAND_LIST_TYPE_DIRECT, __uuidof(ID3D12CommandAllocator), (void**)&m_commandAllocator);
// Create a fence for GPU synchronization.
m_device->CreateFence(0, D3D12_FENCE_FLAG_NONE, __uuidof(ID3D12Fence), (void**)&m_fence);
// Create an event object for the fence.
m_fenceEvent = CreateEvent(nullptr, false, false, nullptr);
// Initialize the starting fence value.
m_fenceValue = 1;
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init(0, nullptr, 0, nullptr, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ID3DBlob* signature;
ID3DBlob* error;
D3D12SerializeRootSignature(
&rootSignatureDesc,
D3D_ROOT_SIGNATURE_VERSION_1,
&signature,
&error);
m_device->CreateRootSignature(
0,
signature->GetBufferPointer(),
signature->GetBufferSize(),
IID_PPV_ARGS(&m_rootSignature));
//Create pipeline state, load and compiler shaders.
//Note here to change D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION for debug
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
D3DReadFileToBlob(L"DefaultVS.cso", &m_vertexShader);
D3DReadFileToBlob(L"DefaultHS.cso", &m_hullShader);
D3DReadFileToBlob(L"DefaultDS.cso", &m_domainShader);
D3DReadFileToBlob(L"DefaultPS.cso", &m_pixelShader);
std::array<D3D12_INPUT_ELEMENT_DESC, 2> inputElementDescs =
{
D3D12_INPUT_ELEMENT_DESC{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
D3D12_INPUT_ELEMENT_DESC{ "COLOR", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 12, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
D3D12_GRAPHICS_PIPELINE_STATE_DESC pipelineDesc = {};
pipelineDesc.InputLayout = { inputElementDescs.data() , (UINT)inputElementDescs.size() };
pipelineDesc.pRootSignature = m_rootSignature;
pipelineDesc.VS = { m_vertexShader->GetBufferPointer(), m_vertexShader->GetBufferSize() };
pipelineDesc.PS = { m_pixelShader->GetBufferPointer(), m_pixelShader->GetBufferSize() };
pipelineDesc.HS = { m_hullShader->GetBufferPointer(), m_hullShader->GetBufferSize() };
pipelineDesc.DS = { m_domainShader->GetBufferPointer(), m_domainShader->GetBufferSize() };
pipelineDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT_WIREFRAME);
pipelineDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
pipelineDesc.DepthStencilState.DepthEnable = false;
pipelineDesc.DepthStencilState.StencilEnable = false;
pipelineDesc.SampleMask = UINT_MAX;
pipelineDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_PATCH;
pipelineDesc.NumRenderTargets = 1;
pipelineDesc.RTVFormats[0] = DXGI_FORMAT_B8G8R8A8_UNORM;
pipelineDesc.SampleDesc.Count = 1;
m_device->CreateGraphicsPipelineState(&pipelineDesc, IID_PPV_ARGS(&m_pipelineState));
// Create a basic command list.
m_device->CreateCommandList(
0,
D3D12_COMMAND_LIST_TYPE_DIRECT,
m_commandAllocator,
m_pipelineState,
IID_PPV_ARGS(&m_commandList));
// Initially we need to close the command list during initialization as it is created in a recording state.
m_commandList->Close();
Vertex triangleVertices[] =
{
{ { 0.0f, 0.5f, 0.f }, { 1.f, 0.f, 0.f, 1.f } },
{ { 0.5f, -0.5f, 0.f },{ 0.f, 1.f, 0.f, 1.f } },
{ { -0.5f, -0.5f, 0.f },{ 0.f, 0.f, 1.f, 1.f } },
};
const UINT vertexBufferSize = sizeof(triangleVertices);
m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer));
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0);
m_vertexBuffer->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin));
memcpy(pVertexDataBegin, triangleVertices, vertexBufferSize);
m_vertexBuffer->Unmap(0, nullptr);
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = vertexBufferSize;
m_viewport.Height = screenHeight;
m_viewport.Width = screenWidth;
m_viewport.MaxDepth = 1000.f;
m_viewport.MinDepth = 0.1f;
m_viewport.TopLeftX = 0.f;
m_viewport.TopLeftY = 0.f;
m_scissorRect.left = 0;
m_scissorRect.right = screenWidth;
m_scissorRect.top = 0;
m_scissorRect.bottom = screenHeight;
unsigned long long fenceToWaitFor = m_fenceValue;
m_commandQueue->Signal(m_fence, fenceToWaitFor);
m_fenceValue++;
// Wait until the GPU is done rendering.
if (m_fence->GetCompletedValue() < fenceToWaitFor)
{
m_fence->SetEventOnCompletion(fenceToWaitFor, m_fenceEvent);
WaitForSingleObject(m_fenceEvent, INFINITE);
}
return true;
}
// Load the sample assets.
void D3D12HeterogeneousMultiadapter::LoadAssets()
{
// Create the root signatures.
{
CD3DX12_ROOT_PARAMETER rootParameters[2];
rootParameters[0].InitAsConstantBufferView(0, 0, D3D12_SHADER_VISIBILITY_VERTEX);
rootParameters[1].InitAsConstantBufferView(1, 0, D3D12_SHADER_VISIBILITY_PIXEL);
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init(_countof(rootParameters), rootParameters, 0, nullptr, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3D12SerializeRootSignature(&rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_devices[Primary]->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
CD3DX12_DESCRIPTOR_RANGE ranges[1];
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 1, 0);
CD3DX12_ROOT_PARAMETER blurRootParameters[3];
blurRootParameters[0].InitAsConstantBufferView(0, 0, D3D12_SHADER_VISIBILITY_PIXEL);
blurRootParameters[1].InitAsDescriptorTable(_countof(ranges), ranges, D3D12_SHADER_VISIBILITY_PIXEL);
blurRootParameters[2].InitAsConstantBufferView(1, 0, D3D12_SHADER_VISIBILITY_PIXEL);
CD3DX12_STATIC_SAMPLER_DESC staticPointSampler(0);
staticPointSampler.Filter = D3D12_FILTER_MIN_MAG_MIP_POINT;
staticPointSampler.ShaderVisibility = D3D12_SHADER_VISIBILITY_PIXEL;
CD3DX12_STATIC_SAMPLER_DESC staticLinearSampler(1);
staticLinearSampler.Filter = D3D12_FILTER_MIN_MAG_MIP_LINEAR;
staticLinearSampler.ShaderVisibility = D3D12_SHADER_VISIBILITY_PIXEL;
D3D12_STATIC_SAMPLER_DESC staticSamplers[] = { staticPointSampler, staticLinearSampler };
rootSignatureDesc.Init(_countof(blurRootParameters), blurRootParameters, _countof(staticSamplers), staticSamplers, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ThrowIfFailed(D3D12SerializeRootSignature(&rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_devices[Secondary]->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_blurRootSignature)));
}
// Create the pipeline states, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
ComPtr<ID3DBlob> vertexShaderBlur;
ComPtr<ID3DBlob> pixelShaderBlurU;
ComPtr<ID3DBlob> pixelShaderBlurV;
ComPtr<ID3DBlob> error;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VShader", "vs_5_0", compileFlags, 0, &vertexShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PShader", "ps_5_0", compileFlags, 0, &pixelShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"blurShaders.hlsl").c_str(), nullptr, nullptr, "VSSimpleBlur", "vs_5_0", compileFlags, 0, &vertexShaderBlur, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"blurShaders.hlsl").c_str(), nullptr, nullptr, "PSSimpleBlurU", "ps_5_0", compileFlags, 0, &pixelShaderBlurU, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"blurShaders.hlsl").c_str(), nullptr, nullptr, "PSSimpleBlurV", "ps_5_0", compileFlags, 0, &pixelShaderBlurV, &error));
// Define the vertex input layouts.
const D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
};
const D3D12_INPUT_ELEMENT_DESC blurInputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, D3D12_APPEND_ALIGNED_ELEMENT, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
};
// Describe and create the graphics pipeline state objects (PSOs).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(vertexShader.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(pixelShader.Get());
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState = CD3DX12_DEPTH_STENCIL_DESC(D3D12_DEFAULT);
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.DSVFormat = DXGI_FORMAT_D32_FLOAT;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_devices[Primary]->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
psoDesc.InputLayout = { blurInputElementDescs, _countof(blurInputElementDescs) };
psoDesc.pRootSignature = m_blurRootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(vertexShaderBlur.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(pixelShaderBlurU.Get());
psoDesc.DepthStencilState.DepthEnable = false;
psoDesc.DSVFormat = DXGI_FORMAT_UNKNOWN;
ThrowIfFailed(m_devices[Secondary]->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_blurPipelineStates[0])));
psoDesc.PS = CD3DX12_SHADER_BYTECODE(pixelShaderBlurV.Get());
ThrowIfFailed(m_devices[Secondary]->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_blurPipelineStates[1])));
}
// Create the command lists.
ThrowIfFailed(m_devices[Primary]->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_directCommandAllocators[Primary][m_frameIndex].Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_directCommandLists[Primary])));
ThrowIfFailed(m_devices[Primary]->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_COPY, m_copyCommandAllocators[m_frameIndex].Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_copyCommandList)));
ThrowIfFailed(m_copyCommandList->Close());
ThrowIfFailed(m_devices[Secondary]->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_directCommandAllocators[Secondary][m_frameIndex].Get(), m_blurPipelineStates[0].Get(), IID_PPV_ARGS(&m_directCommandLists[Secondary])));
// Note: ComPtr's are CPU objects but these resources need to stay in scope until
// the command list that references them has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resources are not
// prematurely destroyed.
ComPtr<ID3D12Resource> vertexBufferUpload;
ComPtr<ID3D12Resource> fullscreenQuadVertexBufferUpload;
// Create the vertex buffer for the primary adapter.
{
// Define the geometry for a triangle.
Vertex triangleVertices[] =
{
{ { 0.0f, TriangleHalfWidth, TriangleDepth } },
{ { TriangleHalfWidth, -TriangleHalfWidth, TriangleDepth } },
{ { -TriangleHalfWidth, -TriangleHalfWidth, TriangleDepth } }
};
const UINT vertexBufferSize = sizeof(triangleVertices);
ThrowIfFailed(m_devices[Primary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
ThrowIfFailed(m_devices[Primary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&vertexBufferUpload)));
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
D3D12_SUBRESOURCE_DATA vertexData = {};
vertexData.pData = reinterpret_cast<UINT8*>(triangleVertices);
vertexData.RowPitch = vertexBufferSize;
vertexData.SlicePitch = vertexData.RowPitch;
UpdateSubresources<1>(m_directCommandLists[Primary].Get(), m_vertexBuffer.Get(), vertexBufferUpload.Get(), 0, 0, 1, &vertexData);
m_directCommandLists[Primary]->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = sizeof(triangleVertices);
}
// Create the vertex buffer for the secondary adapter.
{
// Define the geometry for a fullscreen triangle.
VertexPositionUV quadVertices[] =
{
{ { -1.0f, -1.0f, 0.0f, 1.0f }, { 0.0f, 0.0f } }, // Bottom left.
{ { -1.0f, 1.0f, 0.0f, 1.0f }, { 0.0f, 1.0f } }, // Top left.
{ { 1.0f, -1.0f, 0.0f, 1.0f }, { 1.0f, 0.0f } }, // Bottom right.
{ { 1.0f, 1.0f, 0.0f, 1.0f }, { 1.0f, 1.0f } }, // Top right.
};
const UINT vertexBufferSize = sizeof(quadVertices);
ThrowIfFailed(m_devices[Secondary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_fullscreenQuadVertexBuffer)));
ThrowIfFailed(m_devices[Secondary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&fullscreenQuadVertexBufferUpload)));
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
D3D12_SUBRESOURCE_DATA vertexData = {};
vertexData.pData = reinterpret_cast<UINT8*>(quadVertices);
vertexData.RowPitch = vertexBufferSize;
vertexData.SlicePitch = vertexData.RowPitch;
UpdateSubresources<1>(m_directCommandLists[Secondary].Get(), m_fullscreenQuadVertexBuffer.Get(), fullscreenQuadVertexBufferUpload.Get(), 0, 0, 1, &vertexData);
m_directCommandLists[Secondary]->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_fullscreenQuadVertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer view.
m_fullscreenQuadVertexBufferView.BufferLocation = m_fullscreenQuadVertexBuffer->GetGPUVirtualAddress();
m_fullscreenQuadVertexBufferView.StrideInBytes = sizeof(VertexPositionUV);
m_fullscreenQuadVertexBufferView.SizeInBytes = sizeof(quadVertices);
}
// Create the depth stencil view.
{
D3D12_DEPTH_STENCIL_VIEW_DESC depthStencilDesc = {};
depthStencilDesc.Format = DXGI_FORMAT_D32_FLOAT;
depthStencilDesc.ViewDimension = D3D12_DSV_DIMENSION_TEXTURE2D;
depthStencilDesc.Flags = D3D12_DSV_FLAG_NONE;
const CD3DX12_CLEAR_VALUE clearValue(DXGI_FORMAT_D32_FLOAT, 1.0f, 0);
ThrowIfFailed(m_devices[Primary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Tex2D(DXGI_FORMAT_D32_FLOAT, m_width, m_height, 1, 0, 1, 0, D3D12_RESOURCE_FLAG_ALLOW_DEPTH_STENCIL),
D3D12_RESOURCE_STATE_DEPTH_WRITE,
&clearValue,
IID_PPV_ARGS(&m_depthStencil)
));
m_devices[Primary]->CreateDepthStencilView(m_depthStencil.Get(), &depthStencilDesc, m_dsvHeap->GetCPUDescriptorHandleForHeapStart());
}
// Create the constant buffers.
{
{
const UINT64 constantBufferSize = sizeof(ConstantBufferData) * MaxTriangleCount * FrameCount;
ThrowIfFailed(m_devices[Primary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(constantBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_constantBuffer)));
// Setup constant buffer data.
for (UINT n = 0; n < MaxTriangleCount; n++)
{
m_constantBufferData[n].velocity = XMFLOAT4(GetRandomFloat(0.01f, 0.02f), 0.0f, 0.0f, 0.0f);
m_constantBufferData[n].offset = XMFLOAT4(GetRandomFloat(-5.0f, -1.5f), GetRandomFloat(-1.0f, 1.0f), GetRandomFloat(0.0f, 2.0f), 0.0f);
m_constantBufferData[n].color = XMFLOAT4(GetRandomFloat(0.5f, 1.0f), GetRandomFloat(0.5f, 1.0f), GetRandomFloat(0.5f, 1.0f), 1.0f);
XMStoreFloat4x4(&m_constantBufferData[n].projection, XMMatrixTranspose(XMMatrixPerspectiveFovLH(XM_PIDIV4, m_aspectRatio, 0.01f, 20.0f)));
}
// Map the constant buffer. We don't unmap this until the app closes.
// Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_constantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin)));
memcpy(m_pCbvDataBegin, &m_constantBufferData[0], constantBufferSize / FrameCount);
}
{
const UINT64 workloadConstantBufferSize = sizeof(WorkloadConstantBufferData) * FrameCount;
ThrowIfFailed(m_devices[Primary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(workloadConstantBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_workloadConstantBuffer)));
// Setup constant buffer data.
m_workloadConstantBufferData.loopCount = m_psLoopCount;
// Map the constant buffer. We don't unmap this until the app closes.
// Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_workloadConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pWorkloadCbvDataBegin)));
memcpy(m_pWorkloadCbvDataBegin, &m_workloadConstantBufferData, workloadConstantBufferSize / FrameCount);
}
{
const UINT64 blurWorkloadConstantBufferSize = sizeof(WorkloadConstantBufferData) * FrameCount;
ThrowIfFailed(m_devices[Secondary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(blurWorkloadConstantBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_blurWorkloadConstantBuffer)));
// Setup constant buffer data.
m_blurWorkloadConstantBufferData.loopCount = m_blurPSLoopCount;
// Map the constant buffer. We don't unmap this until the app closes.
// Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_blurWorkloadConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pBlurWorkloadCbvDataBegin)));
memcpy(m_pBlurWorkloadCbvDataBegin, &m_blurWorkloadConstantBufferData, blurWorkloadConstantBufferSize / FrameCount);
}
{
ThrowIfFailed(m_devices[Secondary]->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(sizeof(BlurConstantBufferData)),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_blurConstantBuffer)));
// Map the constant buffer.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_blurConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pBlurCbvDataBegin)));
// Setup constant buffer data.
m_pBlurCbvDataBegin[0].offset = 0.5f;
m_pBlurCbvDataBegin[0].textureDimensions.x = static_cast<float>(m_width);
m_pBlurCbvDataBegin[0].textureDimensions.y = static_cast<float>(m_height);
// Unmap the constant buffer because we don't update this again.
// If we ever do, it should be buffered by the number of frames like other constant buffers.
const CD3DX12_RANGE emptyRange(0, 0);
m_blurConstantBuffer->Unmap(0, &emptyRange);
m_pBlurCbvDataBegin = nullptr;
}
}
// Close the command lists and execute them to begin the vertex buffer copies into the default heaps.
for (UINT i = 0; i < GraphicsAdaptersCount; i++)
{
ThrowIfFailed(m_directCommandLists[i]->Close());
ID3D12CommandList* ppCommandLists[] = { m_directCommandLists[i].Get() };
m_directCommandQueues[i]->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
}
// Create synchronization objects and wait until assets have been uploaded to the GPU.
// We use a cross-adapter fence for handling Signals and Waits between adapters.
// We use regular fences for things that don't need to be cross adapter because they don't need the additional overhead associated with being cross-adapter.
{
// Fence used to control CPU pacing.
ThrowIfFailed(m_devices[Secondary]->CreateFence(0, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_frameFence)));
// Fence used by the primary adapter to signal its copy queue that it has completed rendering.
// When this is signaled, the primary adapter's copy queue can begin copying to the cross-adapter shared resource.
ThrowIfFailed(m_devices[Primary]->CreateFence(0, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_renderFence)));
// Cross-adapter shared fence used by both adapters.
// Used by the primary adapter to signal the secondary adapter that it has completed copying to the cross-adapter shared resource.
// When this is signaled, the secondary adapter can begin its work.
ThrowIfFailed(m_devices[Primary]->CreateFence(0, D3D12_FENCE_FLAG_SHARED | D3D12_FENCE_FLAG_SHARED_CROSS_ADAPTER, IID_PPV_ARGS(&m_crossAdapterFences[Primary])));
// For now, require GENERIC_ALL access.
HANDLE fenceHandle = nullptr;
ThrowIfFailed(m_devices[Primary]->CreateSharedHandle(
m_crossAdapterFences[Primary].Get(),
nullptr,
GENERIC_ALL,
nullptr,
&fenceHandle));
HRESULT openSharedHandleResult = m_devices[Secondary]->OpenSharedHandle(fenceHandle, IID_PPV_ARGS(&m_crossAdapterFences[Secondary]));
// We can close the handle after opening the cross-adapter shared fence.
CloseHandle(fenceHandle);
ThrowIfFailed(openSharedHandleResult);
for (UINT i = 0; i < GraphicsAdaptersCount; i++)
{
// Create an event handle to use for frame synchronization.
m_fenceEvents[i] = CreateEventEx(nullptr, FALSE, FALSE, EVENT_ALL_ACCESS);
if (m_fenceEvents == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForGpu(static_cast<GraphicsAdapter>(i));
}
}
}
// Load the sample assets.
void D3D12Fullscreen::LoadAssets()
{
// Create a root signature consisting of a descriptor table with a single CBV.
{
CD3DX12_DESCRIPTOR_RANGE ranges[1];
CD3DX12_ROOT_PARAMETER rootParameters[1];
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_CBV, 1, 0);
rootParameters[0].InitAsDescriptorTable(1, &ranges[0], D3D12_SHADER_VISIBILITY_VERTEX);
// Allow input layout and deny uneccessary access to certain pipeline stages.
D3D12_ROOT_SIGNATURE_FLAGS rootSignatureFlags =
D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT |
D3D12_ROOT_SIGNATURE_FLAG_DENY_HULL_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_DOMAIN_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_GEOMETRY_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_PIXEL_SHADER_ROOT_ACCESS;
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init(_countof(rootParameters), rootParameters, 0, nullptr, rootSignatureFlags);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3D12SerializeRootSignature(&rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
NAME_D3D12_OBJECT(m_rootSignature);
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
ComPtr<ID3DBlob> error;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &vertexShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &pixelShader, &error));
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "COLOR", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 12, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
// Describe and create the graphics pipeline state object (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(vertexShader.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(pixelShader.Get());
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState.DepthEnable = FALSE;
psoDesc.DepthStencilState.StencilEnable = FALSE;
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
NAME_D3D12_OBJECT(m_pipelineState);
}
// Create the command list.
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocators[m_frameIndex].Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_commandList)));
NAME_D3D12_OBJECT(m_commandList);
LoadSizeDependentResources();
// Create/update the vertex buffer.
{
// Define the geometry for a thin quad that will animate across the screen.
const float x = QuadWidth / 2.0f;
const float y = QuadHeight / 2.0f;
Vertex quadVertices[] =
{
{ { -x, -y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { -x, y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { x, -y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { x, y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } }
};
const UINT vertexBufferSize = sizeof(quadVertices);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_vertexBufferUpload)));
NAME_D3D12_OBJECT(m_vertexBuffer);
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_vertexBufferUpload->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin)));
memcpy(pVertexDataBegin, quadVertices, sizeof(quadVertices));
m_vertexBufferUpload->Unmap(0, nullptr);
m_commandList->CopyBufferRegion(m_vertexBuffer.Get(), 0, m_vertexBufferUpload.Get(), 0, vertexBufferSize);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer views.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = vertexBufferSize;
}
// Create the constant buffer.
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(sizeof(SceneConstantBuffer) * FrameCount),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_constantBuffer)));
NAME_D3D12_OBJECT(m_constantBuffer);
// Describe and create constant buffer views.
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.BufferLocation = m_constantBuffer->GetGPUVirtualAddress();
cbvDesc.SizeInBytes = sizeof(SceneConstantBuffer);
CD3DX12_CPU_DESCRIPTOR_HANDLE cpuHandle(m_cbvHeap->GetCPUDescriptorHandleForHeapStart());
for (UINT n = 0; n < FrameCount; n++)
{
m_device->CreateConstantBufferView(&cbvDesc, cpuHandle);
cbvDesc.BufferLocation += sizeof(SceneConstantBuffer);
cpuHandle.Offset(m_cbvDescriptorSize);
}
// Initialize and map the constant buffers. We don't unmap this until the
// app closes. Keeping things mapped for the lifetime of the resource is okay.
ZeroMemory(&m_constantBufferData, sizeof(m_constantBufferData));
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_constantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin)));
memcpy(m_pCbvDataBegin, &m_constantBufferData, sizeof(m_constantBufferData));
}
// Close the command list and execute it to begin the vertex buffer copy into
// the default heap.
ThrowIfFailed(m_commandList->Close());
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(m_fenceValues[m_frameIndex], D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValues[m_frameIndex]++;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForGpu();
}
}
FrameResource::FrameResource(ID3D12Device* pDevice, ID3D12PipelineState* pPso, ID3D12PipelineState* pShadowMapPso, ID3D12DescriptorHeap* pDsvHeap, ID3D12DescriptorHeap* pCbvSrvHeap, D3D12_VIEWPORT* pViewport, UINT frameResourceIndex) :
m_fenceValue(0),
m_pipelineState(pPso),
m_pipelineStateShadowMap(pShadowMapPso)
{
for (int i = 0; i < CommandListCount; i++)
{
ThrowIfFailed(pDevice->CreateCommandAllocator(D3D12_COMMAND_LIST_TYPE_DIRECT, IID_PPV_ARGS(&m_commandAllocators[i])));
ThrowIfFailed(pDevice->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocators[i].Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_commandLists[i])));
// Close these command lists; don't record into them for now.
ThrowIfFailed(m_commandLists[i]->Close());
}
for (int i = 0; i < NumContexts; i++)
{
// Create command list allocators for worker threads. One alloc is
// for the shadow pass command list, and one is for the scene pass.
ThrowIfFailed(pDevice->CreateCommandAllocator(D3D12_COMMAND_LIST_TYPE_DIRECT, IID_PPV_ARGS(&m_shadowCommandAllocators[i])));
ThrowIfFailed(pDevice->CreateCommandAllocator(D3D12_COMMAND_LIST_TYPE_DIRECT, IID_PPV_ARGS(&m_sceneCommandAllocators[i])));
ThrowIfFailed(pDevice->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_shadowCommandAllocators[i].Get(), m_pipelineStateShadowMap.Get(), IID_PPV_ARGS(&m_shadowCommandLists[i])));
ThrowIfFailed(pDevice->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_sceneCommandAllocators[i].Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_sceneCommandLists[i])));
// Close these command lists; don't record into them for now. We will
// reset them to a recording state when we start the render loop.
ThrowIfFailed(m_shadowCommandLists[i]->Close());
ThrowIfFailed(m_sceneCommandLists[i]->Close());
}
// Describe and create the shadow map texture.
CD3DX12_RESOURCE_DESC shadowTexDesc(
D3D12_RESOURCE_DIMENSION_TEXTURE2D,
0,
static_cast<UINT>(pViewport->Width),
static_cast<UINT>(pViewport->Height),
1,
1,
DXGI_FORMAT_R32_TYPELESS,
1,
0,
D3D12_TEXTURE_LAYOUT_UNKNOWN,
D3D12_RESOURCE_FLAG_ALLOW_DEPTH_STENCIL);
D3D12_CLEAR_VALUE clearValue; // Performance tip: Tell the runtime at resource creation the desired clear value.
clearValue.Format = DXGI_FORMAT_D32_FLOAT;
clearValue.DepthStencil.Depth = 1.0f;
clearValue.DepthStencil.Stencil = 0;
ThrowIfFailed(pDevice->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&shadowTexDesc,
D3D12_RESOURCE_STATE_DEPTH_WRITE,
&clearValue,
IID_PPV_ARGS(&m_shadowTexture)));
// Get a handle to the start of the descriptor heap then offset
// it based on the frame resource index.
const UINT dsvDescriptorSize = pDevice->GetDescriptorHandleIncrementSize(D3D12_DESCRIPTOR_HEAP_TYPE_DSV);
CD3DX12_CPU_DESCRIPTOR_HANDLE depthHandle(pDsvHeap->GetCPUDescriptorHandleForHeapStart(), 1 + frameResourceIndex, dsvDescriptorSize); // + 1 for the shadow map.
// Describe and create the shadow depth view and cache the CPU
// descriptor handle.
D3D12_DEPTH_STENCIL_VIEW_DESC depthStencilViewDesc = {};
depthStencilViewDesc.Format = DXGI_FORMAT_D32_FLOAT;
depthStencilViewDesc.ViewDimension = D3D12_DSV_DIMENSION_TEXTURE2D;
depthStencilViewDesc.Texture2D.MipSlice = 0;
pDevice->CreateDepthStencilView(m_shadowTexture.Get(), &depthStencilViewDesc, depthHandle);
m_shadowDepthView = depthHandle;
// Get a handle to the start of the descriptor heap then offset it
// based on the existing textures and the frame resource index. Each
// frame has 1 SRV (shadow tex) and 2 CBVs.
const UINT nullSrvCount = 2; // Null descriptors at the start of the heap.
const UINT textureCount = _countof(SampleAssets::Textures); // Diffuse + normal textures near the start of the heap. Ideally, track descriptor heap contents/offsets at a higher level.
const UINT cbvSrvDescriptorSize = pDevice->GetDescriptorHandleIncrementSize(D3D12_DESCRIPTOR_HEAP_TYPE_CBV_SRV_UAV);
CD3DX12_CPU_DESCRIPTOR_HANDLE cbvSrvCpuHandle(pCbvSrvHeap->GetCPUDescriptorHandleForHeapStart());
CD3DX12_GPU_DESCRIPTOR_HANDLE cbvSrvGpuHandle(pCbvSrvHeap->GetGPUDescriptorHandleForHeapStart());
m_nullSrvHandle = cbvSrvGpuHandle;
cbvSrvCpuHandle.Offset(nullSrvCount + textureCount + (frameResourceIndex * FrameCount), cbvSrvDescriptorSize);
cbvSrvGpuHandle.Offset(nullSrvCount + textureCount + (frameResourceIndex * FrameCount), cbvSrvDescriptorSize);
// Describe and create a shader resource view (SRV) for the shadow depth
// texture and cache the GPU descriptor handle. This SRV is for sampling
// the shadow map from our shader. It uses the same texture that we use
// as a depth-stencil during the shadow pass.
D3D12_SHADER_RESOURCE_VIEW_DESC shadowSrvDesc = {};
shadowSrvDesc.Format = DXGI_FORMAT_R32_FLOAT;
shadowSrvDesc.ViewDimension = D3D12_SRV_DIMENSION_TEXTURE2D;
shadowSrvDesc.Texture2D.MipLevels = 1;
shadowSrvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
pDevice->CreateShaderResourceView(m_shadowTexture.Get(), &shadowSrvDesc, cbvSrvCpuHandle);
m_shadowDepthHandle = cbvSrvGpuHandle;
// Increment the descriptor handles.
cbvSrvCpuHandle.Offset(cbvSrvDescriptorSize);
cbvSrvGpuHandle.Offset(cbvSrvDescriptorSize);
// Create the constant buffers.
const UINT constantBufferSize = (sizeof(ConstantBuffer) + (D3D12_CONSTANT_BUFFER_DATA_PLACEMENT_ALIGNMENT - 1)) & ~(D3D12_CONSTANT_BUFFER_DATA_PLACEMENT_ALIGNMENT - 1); // must be a multiple 256 bytes
ThrowIfFailed(pDevice->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(constantBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_shadowConstantBuffer)));
ThrowIfFailed(pDevice->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(constantBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_sceneConstantBuffer)));
// Map the constant buffers and cache their heap pointers.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_shadowConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&mp_shadowConstantBufferWO)));
ThrowIfFailed(m_sceneConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&mp_sceneConstantBufferWO)));
// Create the constant buffer views: one for the shadow pass and
// another for the scene pass.
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.SizeInBytes = constantBufferSize;
// Describe and create the shadow constant buffer view (CBV) and
// cache the GPU descriptor handle.
cbvDesc.BufferLocation = m_shadowConstantBuffer->GetGPUVirtualAddress();
pDevice->CreateConstantBufferView(&cbvDesc, cbvSrvCpuHandle);
m_shadowCbvHandle = cbvSrvGpuHandle;
// Increment the descriptor handles.
cbvSrvCpuHandle.Offset(cbvSrvDescriptorSize);
cbvSrvGpuHandle.Offset(cbvSrvDescriptorSize);
// Describe and create the scene constant buffer view (CBV) and
// cache the GPU descriptor handle.
cbvDesc.BufferLocation = m_sceneConstantBuffer->GetGPUVirtualAddress();
pDevice->CreateConstantBufferView(&cbvDesc, cbvSrvCpuHandle);
m_sceneCbvHandle = cbvSrvGpuHandle;
// Batch up command lists for execution later.
const UINT batchSize = _countof(m_sceneCommandLists) + _countof(m_shadowCommandLists) + 3;
m_batchSubmit[0] = m_commandLists[CommandListPre].Get();
memcpy(m_batchSubmit + 1, m_shadowCommandLists, _countof(m_shadowCommandLists) * sizeof(ID3D12CommandList*));
m_batchSubmit[_countof(m_shadowCommandLists) + 1] = m_commandLists[CommandListMid].Get();
memcpy(m_batchSubmit + _countof(m_shadowCommandLists) + 2, m_sceneCommandLists, _countof(m_sceneCommandLists) * sizeof(ID3D12CommandList*));
m_batchSubmit[batchSize - 1] = m_commandLists[CommandListPost].Get();
}
// Load the sample assets.
void D3D12ExecuteIndirect::LoadAssets()
{
// Create the root signatures.
{
CD3DX12_ROOT_PARAMETER rootParameters[GraphicsRootParametersCount];
rootParameters[Cbv].InitAsConstantBufferView(0, 0, D3D12_SHADER_VISIBILITY_VERTEX);
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init(_countof(rootParameters), rootParameters, 0, nullptr, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3D12SerializeRootSignature(&rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
// Create compute signature.
CD3DX12_DESCRIPTOR_RANGE ranges[2];
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 2, 0);
ranges[1].Init(D3D12_DESCRIPTOR_RANGE_TYPE_UAV, 1, 0);
CD3DX12_ROOT_PARAMETER computeRootParameters[ComputeRootParametersCount];
computeRootParameters[SrvUavTable].InitAsDescriptorTable(2, ranges);
computeRootParameters[RootConstants].InitAsConstants(4, 0);
CD3DX12_ROOT_SIGNATURE_DESC computeRootSignatureDesc;
computeRootSignatureDesc.Init(_countof(computeRootParameters), computeRootParameters);
ThrowIfFailed(D3D12SerializeRootSignature(&computeRootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_computeRootSignature)));
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
ComPtr<ID3DBlob> computeShader;
ComPtr<ID3DBlob> error;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &vertexShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &pixelShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"compute.hlsl").c_str(), nullptr, nullptr, "CSMain", "cs_5_0", compileFlags, 0, &computeShader, &error));
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
};
// Describe and create the graphics pipeline state objects (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = { reinterpret_cast<UINT8*>(vertexShader->GetBufferPointer()), vertexShader->GetBufferSize() };
psoDesc.PS = { reinterpret_cast<UINT8*>(pixelShader->GetBufferPointer()), pixelShader->GetBufferSize() };
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState = CD3DX12_DEPTH_STENCIL_DESC(D3D12_DEFAULT);
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.DSVFormat = DXGI_FORMAT_D32_FLOAT;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
// Describe and create the compute pipeline state object (PSO).
D3D12_COMPUTE_PIPELINE_STATE_DESC computePsoDesc = {};
computePsoDesc.pRootSignature = m_computeRootSignature.Get();
computePsoDesc.CS = { reinterpret_cast<UINT8*>(computeShader->GetBufferPointer()), computeShader->GetBufferSize() };
ThrowIfFailed(m_device->CreateComputePipelineState(&computePsoDesc, IID_PPV_ARGS(&m_computeState)));
}
// Create the command list.
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocators[m_frameIndex].Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_commandList)));
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_COMPUTE, m_computeCommandAllocators[m_frameIndex].Get(), m_computeState.Get(), IID_PPV_ARGS(&m_computeCommandList)));
ThrowIfFailed(m_computeCommandList->Close());
// Note: ComPtr's are CPU objects but these resources need to stay in scope until
// the command list that references them has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resources are not
// prematurely destroyed.
ComPtr<ID3D12Resource> vertexBufferUpload;
ComPtr<ID3D12Resource> commandBufferUpload;
// Create the vertex buffer.
{
// Define the geometry for a triangle.
Vertex triangleVertices[] =
{
{ { 0.0f, TriangleHalfWidth, TriangleDepth } },
{ { TriangleHalfWidth, -TriangleHalfWidth, TriangleDepth } },
{ { -TriangleHalfWidth, -TriangleHalfWidth, TriangleDepth } }
};
const UINT vertexBufferSize = sizeof(triangleVertices);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&vertexBufferUpload)));
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
D3D12_SUBRESOURCE_DATA vertexData = {};
vertexData.pData = reinterpret_cast<UINT8*>(triangleVertices);
vertexData.RowPitch = vertexBufferSize;
vertexData.SlicePitch = vertexData.RowPitch;
UpdateSubresources<1>(m_commandList.Get(), m_vertexBuffer.Get(), vertexBufferUpload.Get(), 0, 0, 1, &vertexData);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = sizeof(triangleVertices);
}
// Create the depth stencil view.
{
D3D12_DEPTH_STENCIL_VIEW_DESC depthStencilDesc = {};
depthStencilDesc.Format = DXGI_FORMAT_D32_FLOAT;
depthStencilDesc.ViewDimension = D3D12_DSV_DIMENSION_TEXTURE2D;
depthStencilDesc.Flags = D3D12_DSV_FLAG_NONE;
D3D12_CLEAR_VALUE depthOptimizedClearValue = {};
depthOptimizedClearValue.Format = DXGI_FORMAT_D32_FLOAT;
depthOptimizedClearValue.DepthStencil.Depth = 1.0f;
depthOptimizedClearValue.DepthStencil.Stencil = 0;
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Tex2D(DXGI_FORMAT_D32_FLOAT, m_width, m_height, 1, 0, 1, 0, D3D12_RESOURCE_FLAG_ALLOW_DEPTH_STENCIL),
D3D12_RESOURCE_STATE_DEPTH_WRITE,
&depthOptimizedClearValue,
IID_PPV_ARGS(&m_depthStencil)
));
m_device->CreateDepthStencilView(m_depthStencil.Get(), &depthStencilDesc, m_dsvHeap->GetCPUDescriptorHandleForHeapStart());
}
// Create the constant buffers.
{
const UINT constantBufferDataSize = TriangleResourceCount * sizeof(ConstantBufferData);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(constantBufferDataSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_constantBuffer)));
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.SizeInBytes = sizeof(ConstantBufferData);
// Create constant buffer views to access the upload buffer.
for (UINT n = 0; n < TriangleCount; n++)
{
m_constantBufferData[n].velocity = XMFLOAT4(GetRandomFloat(0.01f, 0.02f), 0.0f, 0.0f, 0.0f);
m_constantBufferData[n].offset = XMFLOAT4(GetRandomFloat(-5.0f, -1.5f), GetRandomFloat(-1.0f, 1.0f), GetRandomFloat(0.0f, 2.0f), 0.0f);
m_constantBufferData[n].color = XMFLOAT4(GetRandomFloat(0.5f, 1.0f), GetRandomFloat(0.5f, 1.0f), GetRandomFloat(0.5f, 1.0f), 1.0f);
XMStoreFloat4x4(&m_constantBufferData[n].projection, XMMatrixTranspose(XMMatrixPerspectiveFovLH(XM_PIDIV4, m_aspectRatio, 0.01f, 20.0f)));
}
// Map the constant buffers. We don't unmap this until the app closes.
// Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_constantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin)));
memcpy(m_pCbvDataBegin, &m_constantBufferData[0], TriangleCount * sizeof(ConstantBufferData));
// Create shader resource views (SRV) of the constant buffers for the
// compute shader to read from.
D3D12_SHADER_RESOURCE_VIEW_DESC srvDesc = {};
srvDesc.Format = DXGI_FORMAT_UNKNOWN;
srvDesc.ViewDimension = D3D12_SRV_DIMENSION_BUFFER;
srvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
srvDesc.Buffer.NumElements = TriangleCount;
srvDesc.Buffer.StructureByteStride = sizeof(ConstantBufferData);
srvDesc.Buffer.Flags = D3D12_BUFFER_SRV_FLAG_NONE;
CD3DX12_CPU_DESCRIPTOR_HANDLE cbvSrvHandle(m_cbvSrvUavHeap->GetCPUDescriptorHandleForHeapStart(), CbvSrvOffset, m_cbvSrvUavDescriptorSize);
for (UINT frame = 0; frame < FrameCount; frame++)
{
srvDesc.Buffer.FirstElement = frame * TriangleCount;
m_device->CreateShaderResourceView(m_constantBuffer.Get(), &srvDesc, cbvSrvHandle);
cbvSrvHandle.Offset(CbvSrvUavDescriptorCountPerFrame, m_cbvSrvUavDescriptorSize);
}
}
// Create the command signature used for indirect drawing.
{
// Each command consists of a CBV update and a DrawInstanced call.
D3D12_INDIRECT_ARGUMENT_DESC argumentDescs[2] = {};
argumentDescs[0].Type = D3D12_INDIRECT_ARGUMENT_TYPE_CONSTANT_BUFFER_VIEW;
argumentDescs[0].ConstantBufferView.RootParameterIndex = Cbv;
argumentDescs[1].Type = D3D12_INDIRECT_ARGUMENT_TYPE_DRAW;
D3D12_COMMAND_SIGNATURE_DESC commandSignatureDesc = {};
commandSignatureDesc.pArgumentDescs = argumentDescs;
commandSignatureDesc.NumArgumentDescs = _countof(argumentDescs);
commandSignatureDesc.ByteStride = sizeof(IndirectCommand);
ThrowIfFailed(m_device->CreateCommandSignature(&commandSignatureDesc, m_rootSignature.Get(), IID_PPV_ARGS(&m_commandSignature)));
}
// Create the command buffers and UAVs to store the results of the compute work.
{
std::vector<IndirectCommand> commands;
commands.resize(TriangleResourceCount);
const UINT commandBufferSize = CommandBufferSizePerFrame * FrameCount;
D3D12_RESOURCE_DESC commandBufferDesc = CD3DX12_RESOURCE_DESC::Buffer(commandBufferSize);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&commandBufferDesc,
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_commandBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(commandBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&commandBufferUpload)));
D3D12_GPU_VIRTUAL_ADDRESS gpuAddress = m_constantBuffer->GetGPUVirtualAddress();
UINT commandIndex = 0;
for (UINT frame = 0; frame < FrameCount; frame++)
{
for (UINT n = 0; n < TriangleCount; n++)
{
commands[commandIndex].cbv = gpuAddress;
commands[commandIndex].drawArguments.VertexCountPerInstance = 3;
commands[commandIndex].drawArguments.InstanceCount = 1;
commands[commandIndex].drawArguments.StartVertexLocation = 0;
commands[commandIndex].drawArguments.StartInstanceLocation = 0;
commandIndex++;
gpuAddress += sizeof(ConstantBufferData);
}
}
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the command buffer.
D3D12_SUBRESOURCE_DATA commandData = {};
commandData.pData = reinterpret_cast<UINT8*>(&commands[0]);
commandData.RowPitch = commandBufferSize;
commandData.SlicePitch = commandData.RowPitch;
UpdateSubresources<1>(m_commandList.Get(), m_commandBuffer.Get(), commandBufferUpload.Get(), 0, 0, 1, &commandData);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_commandBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_NON_PIXEL_SHADER_RESOURCE));
// Create SRVs for the command buffers.
D3D12_SHADER_RESOURCE_VIEW_DESC srvDesc = {};
srvDesc.Format = DXGI_FORMAT_UNKNOWN;
srvDesc.ViewDimension = D3D12_SRV_DIMENSION_BUFFER;
srvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
srvDesc.Buffer.NumElements = TriangleCount;
srvDesc.Buffer.StructureByteStride = sizeof(IndirectCommand);
srvDesc.Buffer.Flags = D3D12_BUFFER_SRV_FLAG_NONE;
CD3DX12_CPU_DESCRIPTOR_HANDLE commandsHandle(m_cbvSrvUavHeap->GetCPUDescriptorHandleForHeapStart(), CommandsOffset, m_cbvSrvUavDescriptorSize);
for (UINT frame = 0; frame < FrameCount; frame++)
{
srvDesc.Buffer.FirstElement = frame * TriangleCount;
m_device->CreateShaderResourceView(m_commandBuffer.Get(), &srvDesc, commandsHandle);
commandsHandle.Offset(CbvSrvUavDescriptorCountPerFrame, m_cbvSrvUavDescriptorSize);
}
// Create the unordered access views (UAVs) that store the results of the compute work.
CD3DX12_CPU_DESCRIPTOR_HANDLE processedCommandsHandle(m_cbvSrvUavHeap->GetCPUDescriptorHandleForHeapStart(), ProcessedCommandsOffset, m_cbvSrvUavDescriptorSize);
for (UINT frame = 0; frame < FrameCount; frame++)
{
// Allocate a buffer large enough to hold all of the indirect commands
// for a single frame as well as a UAV counter.
commandBufferDesc = CD3DX12_RESOURCE_DESC::Buffer(CommandBufferSizePerFrame + sizeof(UINT), D3D12_RESOURCE_FLAG_ALLOW_UNORDERED_ACCESS);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&commandBufferDesc,
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_processedCommandBuffers[frame])));
D3D12_UNORDERED_ACCESS_VIEW_DESC uavDesc = {};
uavDesc.Format = DXGI_FORMAT_UNKNOWN;
uavDesc.ViewDimension = D3D12_UAV_DIMENSION_BUFFER;
uavDesc.Buffer.FirstElement = 0;
uavDesc.Buffer.NumElements = TriangleCount;
uavDesc.Buffer.StructureByteStride = sizeof(IndirectCommand);
uavDesc.Buffer.CounterOffsetInBytes = CommandBufferSizePerFrame;
uavDesc.Buffer.Flags = D3D12_BUFFER_UAV_FLAG_NONE;
m_device->CreateUnorderedAccessView(
m_processedCommandBuffers[frame].Get(),
m_processedCommandBuffers[frame].Get(),
&uavDesc,
processedCommandsHandle);
processedCommandsHandle.Offset(CbvSrvUavDescriptorCountPerFrame, m_cbvSrvUavDescriptorSize);
}
// Allocate a buffer that can be used to reset the UAV counters and initialize
// it to 0.
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(sizeof(UINT)),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_processedCommandBufferCounterReset)));
UINT8* pMappedCounterReset = nullptr;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_processedCommandBufferCounterReset->Map(0, &readRange, reinterpret_cast<void**>(&pMappedCounterReset)));
ZeroMemory(pMappedCounterReset, sizeof(UINT));
m_processedCommandBufferCounterReset->Unmap(0, nullptr);
}
// Close the command list and execute it to begin the vertex buffer copy into
// the default heap.
ThrowIfFailed(m_commandList->Close());
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(m_fenceValues[m_frameIndex], D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
ThrowIfFailed(m_device->CreateFence(m_fenceValues[m_frameIndex], D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_computeFence)));
m_fenceValues[m_frameIndex]++;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForGpu();
}
}
// Load the sample assets.
void D3D12HelloTexture::LoadAssets()
{
// Create the root signature.
{
CD3DX12_DESCRIPTOR_RANGE ranges[1];
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 1, 0);
CD3DX12_ROOT_PARAMETER rootParameters[1];
rootParameters[0].InitAsDescriptorTable(1, &ranges[0], D3D12_SHADER_VISIBILITY_PIXEL);
D3D12_STATIC_SAMPLER_DESC sampler = {};
sampler.Filter = D3D12_FILTER_MIN_MAG_MIP_POINT;
sampler.AddressU = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressV = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressW = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.MipLODBias = 0;
sampler.MaxAnisotropy = 0;
sampler.ComparisonFunc = D3D12_COMPARISON_FUNC_NEVER;
sampler.BorderColor = D3D12_STATIC_BORDER_COLOR_TRANSPARENT_BLACK;
sampler.MinLOD = 0.0f;
sampler.MaxLOD = D3D12_FLOAT32_MAX;
sampler.ShaderRegister = 0;
sampler.RegisterSpace = 0;
sampler.ShaderVisibility = D3D12_SHADER_VISIBILITY_PIXEL;
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init(_countof(rootParameters), rootParameters, 1, &sampler, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3D12SerializeRootSignature(&rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &vertexShader, nullptr));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &pixelShader, nullptr));
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, 12, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
// Describe and create the graphics pipeline state object (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(vertexShader.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(pixelShader.Get());
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState.DepthEnable = FALSE;
psoDesc.DepthStencilState.StencilEnable = FALSE;
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
}
// Create the command list.
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocator.Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_commandList)));
// Create the vertex buffer.
{
// Define the geometry for a triangle.
Vertex triangleVertices[] =
{
{ { 0.0f, 0.25f * m_aspectRatio, 0.0f }, { 0.5f, 0.0f } },
{ { 0.25f, -0.25f * m_aspectRatio, 0.0f }, { 1.0f, 1.0f } },
{ { -0.25f, -0.25f * m_aspectRatio, 0.0f }, { 0.0f, 1.0f } }
};
const UINT vertexBufferSize = sizeof(triangleVertices);
// Note: using upload heaps to transfer static data like vert buffers is not
// recommended. Every time the GPU needs it, the upload heap will be marshalled
// over. Please read up on Default Heap usage. An upload heap is used here for
// code simplicity and because there are very few verts to actually transfer.
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
// Copy the triangle data to the vertex buffer.
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_vertexBuffer->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin)));
memcpy(pVertexDataBegin, triangleVertices, sizeof(triangleVertices));
m_vertexBuffer->Unmap(0, nullptr);
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = vertexBufferSize;
}
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> textureUploadHeap;
// Create the texture.
{
// Describe and create a Texture2D.
D3D12_RESOURCE_DESC textureDesc = {};
textureDesc.MipLevels = 1;
textureDesc.Format = DXGI_FORMAT_R8G8B8A8_UNORM;
textureDesc.Width = TextureWidth;
textureDesc.Height = TextureHeight;
textureDesc.Flags = D3D12_RESOURCE_FLAG_NONE;
textureDesc.DepthOrArraySize = 1;
textureDesc.SampleDesc.Count = 1;
textureDesc.SampleDesc.Quality = 0;
textureDesc.Dimension = D3D12_RESOURCE_DIMENSION_TEXTURE2D;
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&textureDesc,
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_texture)));
const UINT64 uploadBufferSize = GetRequiredIntermediateSize(m_texture.Get(), 0, 1);
// Create the GPU upload buffer.
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(uploadBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&textureUploadHeap)));
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the Texture2D.
std::vector<UINT8> texture = GenerateTextureData();
D3D12_SUBRESOURCE_DATA textureData = {};
textureData.pData = &texture[0];
textureData.RowPitch = TextureWidth * TexturePixelSize;
textureData.SlicePitch = textureData.RowPitch * TextureHeight;
UpdateSubresources(m_commandList.Get(), m_texture.Get(), textureUploadHeap.Get(), 0, 0, 1, &textureData);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_texture.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_PIXEL_SHADER_RESOURCE));
// Describe and create a SRV for the texture.
D3D12_SHADER_RESOURCE_VIEW_DESC srvDesc = {};
srvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
srvDesc.Format = textureDesc.Format;
srvDesc.ViewDimension = D3D12_SRV_DIMENSION_TEXTURE2D;
srvDesc.Texture2D.MipLevels = 1;
m_device->CreateShaderResourceView(m_texture.Get(), &srvDesc, m_srvHeap->GetCPUDescriptorHandleForHeapStart());
}
// Close the command list and execute it to begin the initial GPU setup.
ThrowIfFailed(m_commandList->Close());
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(0, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValue = 1;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForPreviousFrame();
}
}
void D3D12PipelineStateCache::LoadAssets()
{
// Create the root signature.
{
D3D12_FEATURE_DATA_ROOT_SIGNATURE featureData = {};
// This is the highest version the sample supports. If CheckFeatureSupport succeeds, the HighestVersion returned will not be greater than this.
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_1;
if (FAILED(m_device->CheckFeatureSupport(D3D12_FEATURE_ROOT_SIGNATURE, &featureData, sizeof(featureData))))
{
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_0;
}
CD3DX12_DESCRIPTOR_RANGE1 ranges[RootParametersCount];
ranges[RootParameterSRV].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 1, 0, 0, D3D12_DESCRIPTOR_RANGE_FLAG_DATA_STATIC);
CD3DX12_ROOT_PARAMETER1 rootParameters[RootParametersCount];
rootParameters[RootParameterUberShaderCB].InitAsConstantBufferView(0, 0, D3D12_ROOT_DESCRIPTOR_FLAG_DATA_STATIC, D3D12_SHADER_VISIBILITY_ALL);
rootParameters[RootParameterCB].InitAsConstantBufferView(1, 0, D3D12_ROOT_DESCRIPTOR_FLAG_DATA_STATIC, D3D12_SHADER_VISIBILITY_ALL);
rootParameters[RootParameterSRV].InitAsDescriptorTable(1, &ranges[RootParameterSRV]);
D3D12_STATIC_SAMPLER_DESC sampler = {};
sampler.Filter = D3D12_FILTER_MIN_MAG_MIP_POINT;
sampler.AddressU = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressV = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressW = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.MipLODBias = 0;
sampler.MaxAnisotropy = 0;
sampler.ComparisonFunc = D3D12_COMPARISON_FUNC_NEVER;
sampler.BorderColor = D3D12_STATIC_BORDER_COLOR_TRANSPARENT_BLACK;
sampler.MinLOD = 0.0f;
sampler.MaxLOD = 9999.0f;
sampler.ShaderRegister = 0;
sampler.RegisterSpace = 0;
sampler.ShaderVisibility = D3D12_SHADER_VISIBILITY_ALL;
CD3DX12_VERSIONED_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init_1_1(_countof(rootParameters), rootParameters, 1, &sampler, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3DX12SerializeVersionedRootSignature(&rootSignatureDesc, featureData.HighestVersion, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
NAME_D3D12_OBJECT(m_rootSignature);
}
// Create the command list.
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocators[m_frameIndex].Get(), nullptr, IID_PPV_ARGS(&m_commandList)));
NAME_D3D12_OBJECT(m_commandList);
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> vertexIndexBufferUpload;
// Vertex and Index Buffer.
{
const VertexPositionColor cubeVertices[] = {
{ { -1.0f, 1.0f, -1.0f, 1.0f }, { GetRandomColor(),GetRandomColor(), GetRandomColor() } }, // Back Top Left
{ { 1.0f, 1.0f, -1.0f, 1.0f }, { GetRandomColor(), GetRandomColor(), GetRandomColor() } }, // Back Top Right
{ { 1.0f, 1.0f, 1.0f, 1.0f }, { GetRandomColor(), GetRandomColor(), GetRandomColor() } }, // Front Top Right
{ { -1.0f, 1.0f, 1.0f, 1.0f }, { GetRandomColor(), GetRandomColor(), GetRandomColor() } }, // Front Top Left
{ { -1.0f, -1.0f, -1.0f, 1.0f }, { GetRandomColor(),GetRandomColor(), GetRandomColor() } }, // Back Bottom Left
{ { 1.0f, -1.0f, -1.0f, 1.0f }, { GetRandomColor(),GetRandomColor(), GetRandomColor() } }, // Back Bottom Right
{ { 1.0f, -1.0f, 1.0f, 1.0f }, { GetRandomColor(),GetRandomColor(), GetRandomColor() } }, // Front Bottom Right
{ { -1.0f, -1.0f, 1.0f, 1.0f }, { GetRandomColor(),GetRandomColor(), GetRandomColor() } }, // Front Bottom Left
};
const UINT cubeIndices[] =
{
0, 1, 3,
1, 2, 3,
3, 2, 7,
6, 7, 2,
2, 1, 6,
5, 6, 1,
1, 0, 5,
4, 5, 0,
0, 3, 4,
7, 4, 3,
7, 6, 4,
5, 4, 6,
};
static const VertexPositionUV quadVertices[] =
{
{ { -1.0f, -1.0f, 0.0f, 1.0f }, { 0.0f, 1.0f } }, // Bottom Left
{ { -1.0f, 1.0f, 0.0f, 1.0f }, { 0.0f, 0.0f } }, // Top Left
{ { 1.0f, -1.0f, 0.0f, 1.0f }, { 1.0f, 1.0f } }, // Bottom Right
{ { 1.0f, 1.0f, 0.0f, 1.0f }, { 1.0f, 0.0f } }, // Top Right
};
const UINT vertexIndexBufferSize = sizeof(cubeIndices) + sizeof(cubeVertices) + sizeof(quadVertices);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexIndexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_vertexIndexBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexIndexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&vertexIndexBufferUpload)));
NAME_D3D12_OBJECT(m_vertexIndexBuffer);
UINT8* mappedUploadHeap = nullptr;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(vertexIndexBufferUpload->Map(0, &readRange, reinterpret_cast<void**>(&mappedUploadHeap)));
// Fill in part of the upload heap with our index and vertex data.
UINT8* heapLocation = static_cast<UINT8*>(mappedUploadHeap);
memcpy(heapLocation, cubeVertices, sizeof(cubeVertices));
heapLocation += sizeof(cubeVertices);
memcpy(heapLocation, cubeIndices, sizeof(cubeIndices));
heapLocation += sizeof(cubeIndices);
memcpy(heapLocation, quadVertices, sizeof(quadVertices));
// Pack the vertices and indices into their destination by copying from the upload heap.
m_commandList->CopyBufferRegion(m_vertexIndexBuffer.Get(), 0, vertexIndexBufferUpload.Get(), 0, vertexIndexBufferSize);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_vertexIndexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER | D3D12_RESOURCE_STATE_INDEX_BUFFER));
// Create the index and vertex buffer views.
m_cubeVbv.BufferLocation = m_vertexIndexBuffer.Get()->GetGPUVirtualAddress();
m_cubeVbv.SizeInBytes = sizeof(cubeVertices);
m_cubeVbv.StrideInBytes = sizeof(VertexPositionColor);
m_cubeIbv.BufferLocation = m_cubeVbv.BufferLocation + sizeof(cubeVertices);
m_cubeIbv.SizeInBytes = sizeof(cubeIndices);
m_cubeIbv.Format = DXGI_FORMAT_R32_UINT;
m_quadVbv.BufferLocation = m_cubeIbv.BufferLocation + sizeof(cubeIndices);
m_quadVbv.SizeInBytes = sizeof(quadVertices);
m_quadVbv.StrideInBytes = sizeof(VertexPositionUV);
}
// Create the constant buffer.
m_dynamicCB.Init(m_device.Get());
// Close the command list and execute it to begin the vertex/index buffer copy
// into the default heap.
ThrowIfFailed(m_commandList->Close());
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(m_fenceValues[m_frameIndex], D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValues[m_frameIndex]++;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForGpu();
}
m_psoLibrary.Build(m_device.Get(), m_rootSignature.Get());
UpdateWindowTextPso();
}
// Load the assets.
HRESULT VolumetricAnimation::LoadAssets()
{
HRESULT hr;
// Create a root signature consisting of a descriptor table with a CBV SRV and a sampler.
{
CD3DX12_DESCRIPTOR_RANGE ranges[3];
CD3DX12_ROOT_PARAMETER rootParameters[3];
ranges[0].Init( D3D12_DESCRIPTOR_RANGE_TYPE_CBV, 1, 0 );
ranges[1].Init( D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 1, 0 );
ranges[2].Init( D3D12_DESCRIPTOR_RANGE_TYPE_UAV, 1, 0 );
rootParameters[RootParameterCBV].InitAsDescriptorTable( 1, &ranges[0], D3D12_SHADER_VISIBILITY_ALL );
rootParameters[RootParameterSRV].InitAsDescriptorTable( 1, &ranges[1], D3D12_SHADER_VISIBILITY_PIXEL );
rootParameters[RootParameterUAV].InitAsDescriptorTable( 1, &ranges[2], D3D12_SHADER_VISIBILITY_ALL );
D3D12_STATIC_SAMPLER_DESC sampler = {};
sampler.Filter = D3D12_FILTER_MIN_MAG_MIP_POINT;
sampler.AddressU = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressV = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressW = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.MipLODBias = 0;
sampler.MaxAnisotropy = 0;
sampler.ComparisonFunc = D3D12_COMPARISON_FUNC_NEVER;
sampler.BorderColor = D3D12_STATIC_BORDER_COLOR_TRANSPARENT_BLACK;
sampler.MinLOD = 0.0f;
sampler.MaxLOD = D3D12_FLOAT32_MAX;
sampler.ShaderRegister = 0;
sampler.RegisterSpace = 0;
sampler.ShaderVisibility = D3D12_SHADER_VISIBILITY_PIXEL;
// Allow input layout and deny unnecessary access to certain pipeline stages.
D3D12_ROOT_SIGNATURE_FLAGS rootSignatureFlags =
D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT |
D3D12_ROOT_SIGNATURE_FLAG_DENY_HULL_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_DOMAIN_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_GEOMETRY_SHADER_ROOT_ACCESS;
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init( _countof(rootParameters), rootParameters, 1, &sampler, rootSignatureFlags );
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
V( D3D12SerializeRootSignature( &rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error ) );
if ( error ) PRINTERROR( reinterpret_cast< const char* >( error->GetBufferPointer() ) );
VRET( m_device->CreateRootSignature( 0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS( &m_graphicsRootSignature ) ) );
DXDebugName( m_graphicsRootSignature );
// Create compute signature. Must change visibility for the SRV.
rootParameters[RootParameterSRV].ShaderVisibility = D3D12_SHADER_VISIBILITY_ALL;
CD3DX12_ROOT_SIGNATURE_DESC computeRootSignatureDesc( _countof( rootParameters ), rootParameters, 0, nullptr );
VRET( D3D12SerializeRootSignature( &computeRootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error ) );
VRET( m_device->CreateRootSignature( 0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS( &m_computeRootSignature ) ) );
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
ComPtr<ID3DBlob> computeShader;
UINT compileFlags = 0;
VRET( CompileShaderFromFile( GetAssetFullPath( _T( "VolumetricAnimation_shader.hlsl" ) ).c_str(), nullptr, D3D_COMPILE_STANDARD_FILE_INCLUDE, "vsmain", "vs_5_0", compileFlags, 0, &vertexShader ) );
VRET( CompileShaderFromFile( GetAssetFullPath( _T( "VolumetricAnimation_shader.hlsl" ) ).c_str(), nullptr, D3D_COMPILE_STANDARD_FILE_INCLUDE, "psmain", "ps_5_0", compileFlags, 0, &pixelShader ) );
VRET( CompileShaderFromFile( GetAssetFullPath( _T( "VolumetricAnimation_shader.hlsl" ) ).c_str(), nullptr, D3D_COMPILE_STANDARD_FILE_INCLUDE, "csmain", "cs_5_0", compileFlags, 0, &computeShader ) );
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, D3D12_APPEND_ALIGNED_ELEMENT, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
CD3DX12_DEPTH_STENCIL_DESC depthStencilDesc( D3D12_DEFAULT );
depthStencilDesc.DepthEnable = true;
depthStencilDesc.DepthWriteMask = D3D12_DEPTH_WRITE_MASK_ALL;
depthStencilDesc.DepthFunc = D3D12_COMPARISON_FUNC_LESS_EQUAL;
depthStencilDesc.StencilEnable = FALSE;
// Describe and create the graphics pipeline state object (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof( inputElementDescs ) };
psoDesc.pRootSignature = m_graphicsRootSignature.Get();
psoDesc.VS = { reinterpret_cast< UINT8* >( vertexShader->GetBufferPointer() ), vertexShader->GetBufferSize() };
psoDesc.PS = { reinterpret_cast< UINT8* >( pixelShader->GetBufferPointer() ), pixelShader->GetBufferSize() };
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC( D3D12_DEFAULT );
psoDesc.BlendState = CD3DX12_BLEND_DESC( D3D12_DEFAULT );
psoDesc.DepthStencilState = depthStencilDesc;
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.DSVFormat = DXGI_FORMAT_D32_FLOAT;
psoDesc.SampleDesc.Count = 1;
VRET( m_device->CreateGraphicsPipelineState( &psoDesc, IID_PPV_ARGS( &m_pipelineState ) ) );
DXDebugName( m_pipelineState );
// Describe and create the compute pipeline state object (PSO).
D3D12_COMPUTE_PIPELINE_STATE_DESC computePsoDesc = {};
computePsoDesc.pRootSignature = m_computeRootSignature.Get();
computePsoDesc.CS = { reinterpret_cast< UINT8* >( computeShader->GetBufferPointer() ), computeShader->GetBufferSize() };
VRET( m_device->CreateComputePipelineState( &computePsoDesc, IID_PPV_ARGS( &m_computeState ) ) );
DXDebugName( m_computeState );
}
// Create the compute command list.
VRET( m_device->CreateCommandList( 0, D3D12_COMMAND_LIST_TYPE_COMPUTE, m_computeCmdAllocator.Get(),m_computeState.Get(), IID_PPV_ARGS( &m_computeCmdList ) ) );
DXDebugName( m_computeCmdList );
VRET( m_computeCmdList->Close() );
// Create the graphics command list.
VRET( m_device->CreateCommandList( 0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_graphicCmdAllocator.Get(), m_pipelineState.Get(), IID_PPV_ARGS( &m_graphicCmdList ) ) );
DXDebugName( m_graphicCmdList );
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> volumeBufferUploadHeap;
// Create the volumeBuffer.
{
UINT volumeBufferSize = m_volumeDepth*m_volumeHeight*m_volumeWidth * 4 * sizeof( UINT8 );
D3D12_RESOURCE_DESC bufferDesc = CD3DX12_RESOURCE_DESC::Buffer( volumeBufferSize, D3D12_RESOURCE_FLAG_ALLOW_UNORDERED_ACCESS );
D3D12_RESOURCE_DESC uploadBufferDesc = CD3DX12_RESOURCE_DESC::Buffer( volumeBufferSize );
VRET( m_device->CreateCommittedResource(&CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_DEFAULT ),D3D12_HEAP_FLAG_NONE,
&bufferDesc,D3D12_RESOURCE_STATE_COPY_DEST,nullptr,IID_PPV_ARGS( &m_volumeBuffer ) ) );
const UINT64 uploadBufferSize = GetRequiredIntermediateSize( m_volumeBuffer.Get(), 0, 1 );
// Create the GPU upload buffer.
VRET( m_device->CreateCommittedResource(&CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ),D3D12_HEAP_FLAG_NONE,
&uploadBufferDesc,D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,IID_PPV_ARGS( &volumeBufferUploadHeap ) ) );
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the Texture2D.
UINT8* volumeBuffer = ( UINT8* ) malloc( volumeBufferSize );
memset( volumeBuffer, 64, volumeBufferSize );
//float radius = m_volumeHeight / 2.f;
float a = m_volumeWidth / 2.f;
float b = m_volumeHeight / 2.f;
float c = m_volumeDepth / 2.f;
float radius = sqrt( a*a + b*b + c*c );
for ( UINT z = 0; z < m_volumeDepth; z++ )
for ( UINT y = 0; y < m_volumeHeight; y++ )
for ( UINT x = 0; x < m_volumeWidth; x++ )
{
float _x = x - m_volumeWidth / 2.f;
float _y = y - m_volumeHeight / 2.f;
float _z = z - m_volumeDepth / 2.f;
//float currentRaidus =abs(_x)+abs(_y)+abs(_z);
float currentRaidus = sqrt( _x*_x + _y*_y + _z*_z );
float scale = currentRaidus *3.f / radius;
UINT idx = 4 - (UINT)floor( scale );
UINT interm = ( UINT ) ( 192 * scale +0.5f );
UINT8 col = interm % 192+1;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 0] += col * m_constantBufferData.colVal[idx].x;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 1] += col * m_constantBufferData.colVal[idx].y;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 2] += col * m_constantBufferData.colVal[idx].z;
volumeBuffer[( x + y*m_volumeWidth + z*m_volumeHeight*m_volumeWidth ) * 4 + 3] = m_constantBufferData.colVal[idx].w;
}
D3D12_SUBRESOURCE_DATA volumeBufferData = {};
volumeBufferData.pData = &volumeBuffer[0];
volumeBufferData.RowPitch = volumeBufferSize;
volumeBufferData.SlicePitch = volumeBufferData.RowPitch;
UpdateSubresources( m_graphicCmdList.Get(), m_volumeBuffer.Get(), volumeBufferUploadHeap.Get(), 0, 0, 1, &volumeBufferData );
m_graphicCmdList->ResourceBarrier( 1, &CD3DX12_RESOURCE_BARRIER::Transition( m_volumeBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_UNORDERED_ACCESS ) );
// Describe and create a SRV for the volumeBuffer.
D3D12_SHADER_RESOURCE_VIEW_DESC srvDesc = {};
srvDesc.Shader4ComponentMapping = D3D12_DEFAULT_SHADER_4_COMPONENT_MAPPING;
srvDesc.Format = DXGI_FORMAT_UNKNOWN;
srvDesc.ViewDimension = D3D12_SRV_DIMENSION_BUFFER;
srvDesc.Buffer.FirstElement = 0;
srvDesc.Buffer.NumElements = m_volumeDepth*m_volumeHeight*m_volumeWidth;
srvDesc.Buffer.StructureByteStride = 4 * sizeof( UINT8 );
srvDesc.Buffer.Flags = D3D12_BUFFER_SRV_FLAG_NONE;
CD3DX12_CPU_DESCRIPTOR_HANDLE srvHandle( m_cbvsrvuavHeap->GetCPUDescriptorHandleForHeapStart(), RootParameterSRV, m_cbvsrvuavDescriptorSize );
m_device->CreateShaderResourceView( m_volumeBuffer.Get(), &srvDesc, srvHandle );
// Describe and create a UAV for the volumeBuffer.
D3D12_UNORDERED_ACCESS_VIEW_DESC uavDesc = {};
uavDesc.Format = DXGI_FORMAT_UNKNOWN;
uavDesc.ViewDimension = D3D12_UAV_DIMENSION_BUFFER;
uavDesc.Buffer.FirstElement = 0;
uavDesc.Buffer.NumElements = m_volumeWidth*m_volumeHeight*m_volumeDepth;
uavDesc.Buffer.StructureByteStride = 4 * sizeof( UINT8 );
uavDesc.Buffer.CounterOffsetInBytes = 0;
uavDesc.Buffer.Flags = D3D12_BUFFER_UAV_FLAG_NONE;
CD3DX12_CPU_DESCRIPTOR_HANDLE uavHandle( m_cbvsrvuavHeap->GetCPUDescriptorHandleForHeapStart(), RootParameterUAV, m_cbvsrvuavDescriptorSize );
m_device->CreateUnorderedAccessView( m_volumeBuffer.Get(), nullptr, &uavDesc, uavHandle );
free( volumeBuffer );
}
// Create the vertex buffer.
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> vertexBufferUpload;
{
// Define the geometry for a triangle.
Vertex cubeVertices[] =
{
{ XMFLOAT3( -128.f, -128.f, -128.f ) },
{ XMFLOAT3( -128.f, -128.f, 128.f ) },
{ XMFLOAT3( -128.f, 128.f, -128.f ) },
{ XMFLOAT3( -128.f, 128.f, 128.f ) },
{ XMFLOAT3( 128.f, -128.f, -128.f )},
{ XMFLOAT3( 128.f, -128.f, 128.f )},
{ XMFLOAT3( 128.f, 128.f, -128.f )},
{ XMFLOAT3( 128.f, 128.f, 128.f )},
};
const UINT vertexBufferSize = sizeof( cubeVertices );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( vertexBufferSize ), D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr, IID_PPV_ARGS( &vertexBufferUpload ) ) );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_DEFAULT ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( vertexBufferSize ), D3D12_RESOURCE_STATE_COPY_DEST,
nullptr, IID_PPV_ARGS( &m_vertexBuffer ) ) );
DXDebugName( m_vertexBuffer );
D3D12_SUBRESOURCE_DATA vertexData = {};
vertexData.pData = reinterpret_cast< UINT8* >( cubeVertices );
vertexData.RowPitch = vertexBufferSize;
vertexData.SlicePitch = vertexBufferSize;
UpdateSubresources<1>( m_graphicCmdList.Get(), m_vertexBuffer.Get(), vertexBufferUpload.Get(), 0, 0, 1, &vertexData );
m_graphicCmdList->ResourceBarrier( 1, &CD3DX12_RESOURCE_BARRIER::Transition( m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST,
D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER ));
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof( Vertex );
m_vertexBufferView.SizeInBytes = vertexBufferSize;
}
// Create the index buffer
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> indexBufferUpload;
{
uint16_t cubeIndices[] =
{
0,2,1, 1,2,3, 4,5,6, 5,7,6, 0,1,5, 0,5,4, 2,6,7, 2,7,3, 0,4,6, 0,6,2, 1,3,7, 1,7,5,
};
const UINT indexBufferSize = sizeof( cubeIndices );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( indexBufferSize ), D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr, IID_PPV_ARGS( &indexBufferUpload ) ) );
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_DEFAULT ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( indexBufferSize ), D3D12_RESOURCE_STATE_COPY_DEST,
nullptr, IID_PPV_ARGS( &m_indexBuffer ) ) );
DXDebugName( m_indexBuffer );
D3D12_SUBRESOURCE_DATA indexData = {};
indexData.pData = reinterpret_cast< UINT8* >( cubeIndices );
indexData.RowPitch = indexBufferSize;
indexData.SlicePitch = indexBufferSize;
UpdateSubresources<1>( m_graphicCmdList.Get(), m_indexBuffer.Get(), indexBufferUpload.Get(), 0, 0, 1, &indexData );
m_graphicCmdList->ResourceBarrier( 1, &CD3DX12_RESOURCE_BARRIER::Transition( m_indexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST,
D3D12_RESOURCE_STATE_INDEX_BUFFER ) );
m_indexBufferView.BufferLocation = m_indexBuffer->GetGPUVirtualAddress();
m_indexBufferView.SizeInBytes = sizeof( cubeIndices );
m_indexBufferView.Format = DXGI_FORMAT_R16_UINT;
}
// Create the constant buffer
{
VRET( m_device->CreateCommittedResource( &CD3DX12_HEAP_PROPERTIES( D3D12_HEAP_TYPE_UPLOAD ), D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer( 1024 * 64 ), D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr, IID_PPV_ARGS( &m_constantBuffer ) ) );
DXDebugName( m_constantBuffer );
// Describe and create a constant buffer view.
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.BufferLocation = m_constantBuffer->GetGPUVirtualAddress();
cbvDesc.SizeInBytes = ( sizeof( ConstantBuffer ) + 255 ) & ~255; // CB size is required to be 256-byte aligned.
m_device->CreateConstantBufferView( &cbvDesc, m_cbvsrvuavHeap->GetCPUDescriptorHandleForHeapStart() );
// Initialize and map the constant buffers. We don't unmap this until the
// app closes. Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange( 0, 0 ); // We do not intend to read from this resource on the CPU.
VRET( m_constantBuffer->Map( 0, &readRange, reinterpret_cast< void** >( &m_pCbvDataBegin ) ) );
memcpy( m_pCbvDataBegin, &m_constantBufferData, sizeof( m_constantBufferData ) );
}
// Close the command list and execute it to begin the initial GPU setup.
VRET( m_graphicCmdList->Close() );
ID3D12CommandList* ppCommandLists[] = { m_graphicCmdList.Get() };
m_graphicCmdQueue->ExecuteCommandLists( _countof( ppCommandLists ), ppCommandLists );
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
VRET( m_device->CreateFence( 0, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS( &m_fence ) ) );
DXDebugName( m_fence );
m_fenceValue = 1;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent( nullptr, FALSE, FALSE, nullptr );
if ( m_fenceEvent == nullptr )
{
VRET( HRESULT_FROM_WIN32( GetLastError() ) );
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForGraphicsCmd();
}
XMVECTORF32 vecEye = { 500.0f, 500.0f, -500.0f };
XMVECTORF32 vecAt = { 0.0f, 0.0f, 0.0f };
m_camera.SetViewParams( vecEye, vecAt );
m_camera.SetEnablePositionMovement( true );
m_camera.SetButtonMasks( MOUSE_RIGHT_BUTTON, MOUSE_WHEEL, MOUSE_LEFT_BUTTON );
return S_OK;
}
// Load the sample assets.
void D3D12PredicationQueries::LoadAssets()
{
// Create a root signature consisting of a single CBV parameter.
{
D3D12_FEATURE_DATA_ROOT_SIGNATURE featureData = {};
// This is the highest version the sample supports. If CheckFeatureSupport succeeds, the HighestVersion returned will not be greater than this.
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_1;
if (FAILED(m_device->CheckFeatureSupport(D3D12_FEATURE_ROOT_SIGNATURE, &featureData, sizeof(featureData))))
{
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_0;
}
CD3DX12_DESCRIPTOR_RANGE1 ranges[1];
CD3DX12_ROOT_PARAMETER1 rootParameters[1];
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_CBV, 1, 0, 0, D3D12_DESCRIPTOR_RANGE_FLAG_DATA_STATIC);
rootParameters[0].InitAsDescriptorTable(1, &ranges[0], D3D12_SHADER_VISIBILITY_VERTEX);
// Allow input layout and deny uneccessary access to certain pipeline stages.
D3D12_ROOT_SIGNATURE_FLAGS rootSignatureFlags =
D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT |
D3D12_ROOT_SIGNATURE_FLAG_DENY_HULL_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_DOMAIN_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_GEOMETRY_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_PIXEL_SHADER_ROOT_ACCESS;
CD3DX12_VERSIONED_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init_1_1(_countof(rootParameters), rootParameters, 0, nullptr, rootSignatureFlags);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3DX12SerializeVersionedRootSignature(&rootSignatureDesc, featureData.HighestVersion, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
NAME_D3D12_OBJECT(m_rootSignature);
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &vertexShader, nullptr));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &pixelShader, nullptr));
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "COLOR", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 12, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
// Enable alpha blending so we can visualize the occlusion query results.
CD3DX12_BLEND_DESC blendDesc(D3D12_DEFAULT);
blendDesc.RenderTarget[0] =
{
TRUE, FALSE,
D3D12_BLEND_SRC_ALPHA, D3D12_BLEND_INV_SRC_ALPHA, D3D12_BLEND_OP_ADD,
D3D12_BLEND_ONE, D3D12_BLEND_ZERO, D3D12_BLEND_OP_ADD,
D3D12_LOGIC_OP_NOOP,
D3D12_COLOR_WRITE_ENABLE_ALL,
};
// Describe and create the graphics pipeline state objects (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(vertexShader.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(pixelShader.Get());
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = blendDesc;
psoDesc.DepthStencilState = CD3DX12_DEPTH_STENCIL_DESC(D3D12_DEFAULT);
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.DSVFormat = DXGI_FORMAT_D32_FLOAT;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
NAME_D3D12_OBJECT(m_pipelineState);
// Disable color writes and depth writes for the occlusion query's state.
psoDesc.BlendState.RenderTarget[0].RenderTargetWriteMask = 0;
psoDesc.DepthStencilState.DepthWriteMask = D3D12_DEPTH_WRITE_MASK_ZERO;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_queryState)));
NAME_D3D12_OBJECT(m_queryState);
}
// Create the command list.
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocators[m_frameIndex].Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_commandList)));
NAME_D3D12_OBJECT(m_commandList);
// Note: ComPtr's are CPU objects but this resource needs to stay in scope until
// the command list that references it has finished executing on the GPU.
// We will flush the GPU at the end of this method to ensure the resource is not
// prematurely destroyed.
ComPtr<ID3D12Resource> vertexBufferUpload;
// Create the vertex buffer.
{
// Create geometry for two quads and a bounding box for the occlusion query.
// Geometry will be rendered back-to-front to support transparency in the scene.
Vertex quadVertices[] =
{
// Far quad - in practice this would be a complex geometry.
{ { -0.25f, -0.25f * m_aspectRatio, 0.5f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { -0.25f, 0.25f * m_aspectRatio, 0.5f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { 0.25f, -0.25f * m_aspectRatio, 0.5f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { 0.25f, 0.25f * m_aspectRatio, 0.5f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
// Near quad.
{ { -0.5f, -0.35f * m_aspectRatio, 0.0f }, { 1.0f, 0.0f, 0.0f, 0.65f } },
{ { -0.5f, 0.35f * m_aspectRatio, 0.0f }, { 1.0f, 0.0f, 0.0f, 0.65f } },
{ { 0.5f, -0.35f * m_aspectRatio, 0.0f }, { 1.0f, 1.0f, 0.0f, 0.65f } },
{ { 0.5f, 0.35f * m_aspectRatio, 0.0f }, { 1.0f, 1.0f, 0.0f, 0.65f } },
// Far quad bounding box used for occlusion query (offset slightly to avoid z-fighting).
{ { -0.25f, -0.25f * m_aspectRatio, 0.4999f }, { 0.0f, 0.0f, 0.0f, 1.0f } },
{ { -0.25f, 0.25f * m_aspectRatio, 0.4999f }, { 0.0f, 0.0f, 0.0f, 1.0f } },
{ { 0.25f, -0.25f * m_aspectRatio, 0.4999f }, { 0.0f, 0.0f, 0.0f, 1.0f } },
{ { 0.25f, 0.25f * m_aspectRatio, 0.4999f }, { 0.0f, 0.0f, 0.0f, 1.0f } },
};
const UINT vertexBufferSize = sizeof(quadVertices);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&vertexBufferUpload)));
NAME_D3D12_OBJECT(m_vertexBuffer);
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
D3D12_SUBRESOURCE_DATA vertexData = {};
vertexData.pData = reinterpret_cast<UINT8*>(quadVertices);
vertexData.RowPitch = vertexBufferSize;
vertexData.SlicePitch = vertexData.RowPitch;
UpdateSubresources<1>(m_commandList.Get(), m_vertexBuffer.Get(), vertexBufferUpload.Get(), 0, 0, 1, &vertexData);
m_commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_vertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = sizeof(quadVertices);
}
// Create the constant buffers.
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(FrameCount * sizeof(m_constantBufferData)),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_constantBuffer)));
NAME_D3D12_OBJECT(m_constantBuffer);
// Map and initialize the constant buffer. We don't unmap this until the
// app closes. Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_constantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin)));
ZeroMemory(m_pCbvDataBegin, FrameCount * sizeof(m_constantBufferData));
// Create constant buffer views to access the upload buffer.
CD3DX12_CPU_DESCRIPTOR_HANDLE cpuHandle(m_cbvHeap->GetCPUDescriptorHandleForHeapStart());
D3D12_GPU_VIRTUAL_ADDRESS gpuAddress = m_constantBuffer->GetGPUVirtualAddress();
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.SizeInBytes = sizeof(SceneConstantBuffer);
for (UINT n = 0; n < FrameCount; n++)
{
cbvDesc.BufferLocation = gpuAddress;
m_device->CreateConstantBufferView(&cbvDesc, cpuHandle);
cpuHandle.Offset(m_cbvSrvDescriptorSize);
gpuAddress += cbvDesc.SizeInBytes;
cbvDesc.BufferLocation = gpuAddress;
m_device->CreateConstantBufferView(&cbvDesc, cpuHandle);
cpuHandle.Offset(m_cbvSrvDescriptorSize);
gpuAddress += cbvDesc.SizeInBytes;
}
}
// Create the depth stencil view.
{
D3D12_DEPTH_STENCIL_VIEW_DESC depthStencilDesc = {};
depthStencilDesc.Format = DXGI_FORMAT_D32_FLOAT;
depthStencilDesc.ViewDimension = D3D12_DSV_DIMENSION_TEXTURE2D;
depthStencilDesc.Flags = D3D12_DSV_FLAG_NONE;
D3D12_CLEAR_VALUE depthOptimizedClearValue = {};
depthOptimizedClearValue.Format = DXGI_FORMAT_D32_FLOAT;
depthOptimizedClearValue.DepthStencil.Depth = 1.0f;
depthOptimizedClearValue.DepthStencil.Stencil = 0;
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Tex2D(DXGI_FORMAT_D32_FLOAT, m_width, m_height, 1, 0, 1, 0, D3D12_RESOURCE_FLAG_ALLOW_DEPTH_STENCIL),
D3D12_RESOURCE_STATE_DEPTH_WRITE,
&depthOptimizedClearValue,
IID_PPV_ARGS(&m_depthStencil)
));
NAME_D3D12_OBJECT(m_depthStencil);
m_device->CreateDepthStencilView(m_depthStencil.Get(), &depthStencilDesc, m_dsvHeap->GetCPUDescriptorHandleForHeapStart());
}
// Create the query result buffer.
{
D3D12_RESOURCE_DESC queryResultDesc = CD3DX12_RESOURCE_DESC::Buffer(8);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&queryResultDesc,
D3D12_RESOURCE_STATE_PREDICATION,
nullptr,
IID_PPV_ARGS(&m_queryResult)
));
NAME_D3D12_OBJECT(m_queryResult);
}
// Close the command list and execute it to begin the vertex buffer copy into
// the default heap.
ThrowIfFailed(m_commandList->Close());
ID3D12CommandList* ppCommandLists[] = { m_commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(m_fenceValues[m_frameIndex], D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValues[m_frameIndex]++;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForGpu();
}
}
// Load the sample assets.
void D3D12Fullscreen::LoadAssets()
{
D3D12_FEATURE_DATA_ROOT_SIGNATURE featureData = {};
// This is the highest version the sample supports. If CheckFeatureSupport succeeds, the HighestVersion returned will not be greater than this.
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_1;
if (FAILED(m_device->CheckFeatureSupport(D3D12_FEATURE_ROOT_SIGNATURE, &featureData, sizeof(featureData))))
{
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_0;
}
// Create a root signature consisting of a descriptor table with a single CBV.
{
CD3DX12_DESCRIPTOR_RANGE1 ranges[1];
CD3DX12_ROOT_PARAMETER1 rootParameters[1];
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_CBV, 1, 0, 0, D3D12_DESCRIPTOR_RANGE_FLAG_DATA_STATIC);
rootParameters[0].InitAsDescriptorTable(1, &ranges[0], D3D12_SHADER_VISIBILITY_VERTEX);
// Allow input layout and deny uneccessary access to certain pipeline stages.
D3D12_ROOT_SIGNATURE_FLAGS rootSignatureFlags =
D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT |
D3D12_ROOT_SIGNATURE_FLAG_DENY_HULL_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_DOMAIN_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_GEOMETRY_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_PIXEL_SHADER_ROOT_ACCESS;
CD3DX12_VERSIONED_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init_1_1(_countof(rootParameters), rootParameters, 0, nullptr, rootSignatureFlags);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3DX12SerializeVersionedRootSignature(&rootSignatureDesc, featureData.HighestVersion, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_sceneRootSignature)));
NAME_D3D12_OBJECT(m_sceneRootSignature);
}
// Create a root signature consisting of a descriptor table with a SRV and a sampler.
{
CD3DX12_DESCRIPTOR_RANGE1 ranges[1];
CD3DX12_ROOT_PARAMETER1 rootParameters[1];
// We don't modify the SRV in the post-processing command list after
// SetGraphicsRootDescriptorTable is executed on the GPU so we can use the default
// range behavior: D3D12_DESCRIPTOR_RANGE_FLAG_DATA_STATIC_WHILE_SET_AT_EXECUTE
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_SRV, 1, 0);
rootParameters[0].InitAsDescriptorTable(1, &ranges[0], D3D12_SHADER_VISIBILITY_PIXEL);
// Allow input layout and pixel shader access and deny uneccessary access to certain pipeline stages.
D3D12_ROOT_SIGNATURE_FLAGS rootSignatureFlags =
D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT |
D3D12_ROOT_SIGNATURE_FLAG_DENY_HULL_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_DOMAIN_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_GEOMETRY_SHADER_ROOT_ACCESS;
// Create a sampler.
D3D12_STATIC_SAMPLER_DESC sampler = {};
sampler.Filter = D3D12_FILTER_MIN_MAG_MIP_LINEAR;
sampler.AddressU = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressV = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.AddressW = D3D12_TEXTURE_ADDRESS_MODE_BORDER;
sampler.MipLODBias = 0;
sampler.MaxAnisotropy = 0;
sampler.ComparisonFunc = D3D12_COMPARISON_FUNC_NEVER;
sampler.BorderColor = D3D12_STATIC_BORDER_COLOR_TRANSPARENT_BLACK;
sampler.MinLOD = 0.0f;
sampler.MaxLOD = D3D12_FLOAT32_MAX;
sampler.ShaderRegister = 0;
sampler.RegisterSpace = 0;
sampler.ShaderVisibility = D3D12_SHADER_VISIBILITY_PIXEL;
CD3DX12_VERSIONED_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init_1_1(_countof(rootParameters), rootParameters, 1, &sampler, rootSignatureFlags);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3DX12SerializeVersionedRootSignature(&rootSignatureDesc, featureData.HighestVersion, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_postRootSignature)));
NAME_D3D12_OBJECT(m_postRootSignature);
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> sceneVertexShader;
ComPtr<ID3DBlob> scenePixelShader;
ComPtr<ID3DBlob> postVertexShader;
ComPtr<ID3DBlob> postPixelShader;
ComPtr<ID3DBlob> error;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"sceneShaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &sceneVertexShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"sceneShaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &scenePixelShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"postShaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &postVertexShader, &error));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"postShaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &postPixelShader, &error));
// Define the vertex input layouts.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "COLOR", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 12, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
D3D12_INPUT_ELEMENT_DESC scaleInputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, D3D12_APPEND_ALIGNED_ELEMENT, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
// Describe and create the graphics pipeline state objects (PSOs).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_sceneRootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(sceneVertexShader.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(scenePixelShader.Get());
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState.DepthEnable = FALSE;
psoDesc.DepthStencilState.StencilEnable = FALSE;
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_scenePipelineState)));
NAME_D3D12_OBJECT(m_scenePipelineState);
psoDesc.InputLayout = { scaleInputElementDescs, _countof(scaleInputElementDescs) };
psoDesc.pRootSignature = m_postRootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(postVertexShader.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(postPixelShader.Get());
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_postPipelineState)));
NAME_D3D12_OBJECT(m_postPipelineState);
}
// Single-use command allocator and command list for creating resources.
ComPtr<ID3D12CommandAllocator> commandAllocator;
ComPtr<ID3D12GraphicsCommandList> commandList;
ThrowIfFailed(m_device->CreateCommandAllocator(D3D12_COMMAND_LIST_TYPE_DIRECT, IID_PPV_ARGS(&commandAllocator)));
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, commandAllocator.Get(), nullptr, IID_PPV_ARGS(&commandList)));
// Create the command lists.
{
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_sceneCommandAllocators[m_frameIndex].Get(), m_scenePipelineState.Get(), IID_PPV_ARGS(&m_sceneCommandList)));
NAME_D3D12_OBJECT(m_sceneCommandList);
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_postCommandAllocators[m_frameIndex].Get(), m_postPipelineState.Get(), IID_PPV_ARGS(&m_postCommandList)));
NAME_D3D12_OBJECT(m_postCommandList);
// Close the command lists.
ThrowIfFailed(m_sceneCommandList->Close());
ThrowIfFailed(m_postCommandList->Close());
}
LoadSizeDependentResources();
LoadSceneResolutionDependentResources();
// Create/update the vertex buffer.
ComPtr<ID3D12Resource> sceneVertexBufferUpload;
{
// Define the geometry for a thin quad that will animate across the screen.
const float x = QuadWidth / 2.0f;
const float y = QuadHeight / 2.0f;
SceneVertex quadVertices[] =
{
{ { -x, -y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { -x, y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { x, -y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } },
{ { x, y, 1.0f }, { 1.0f, 1.0f, 1.0f, 1.0f } }
};
const UINT vertexBufferSize = sizeof(quadVertices);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_sceneVertexBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&sceneVertexBufferUpload)));
NAME_D3D12_OBJECT(m_sceneVertexBuffer);
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(sceneVertexBufferUpload->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin)));
memcpy(pVertexDataBegin, quadVertices, sizeof(quadVertices));
sceneVertexBufferUpload->Unmap(0, nullptr);
commandList->CopyBufferRegion(m_sceneVertexBuffer.Get(), 0, sceneVertexBufferUpload.Get(), 0, vertexBufferSize);
commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_sceneVertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer views.
m_sceneVertexBufferView.BufferLocation = m_sceneVertexBuffer->GetGPUVirtualAddress();
m_sceneVertexBufferView.StrideInBytes = sizeof(SceneVertex);
m_sceneVertexBufferView.SizeInBytes = vertexBufferSize;
}
// Create/update the fullscreen quad vertex buffer.
ComPtr<ID3D12Resource> postVertexBufferUpload;
{
// Define the geometry for a fullscreen quad.
PostVertex quadVertices[] =
{
{ { -1.0f, -1.0f, 0.0f, 1.0f }, { 0.0f, 0.0f } }, // Bottom left.
{ { -1.0f, 1.0f, 0.0f, 1.0f }, { 0.0f, 1.0f } }, // Top left.
{ { 1.0f, -1.0f, 0.0f, 1.0f }, { 1.0f, 0.0f } }, // Bottom right.
{ { 1.0f, 1.0f, 0.0f, 1.0f }, { 1.0f, 1.0f } } // Top right.
};
const UINT vertexBufferSize = sizeof(quadVertices);
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_DEFAULT),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_COPY_DEST,
nullptr,
IID_PPV_ARGS(&m_postVertexBuffer)));
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&postVertexBufferUpload)));
NAME_D3D12_OBJECT(m_postVertexBuffer);
// Copy data to the intermediate upload heap and then schedule a copy
// from the upload heap to the vertex buffer.
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(postVertexBufferUpload->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin)));
memcpy(pVertexDataBegin, quadVertices, sizeof(quadVertices));
postVertexBufferUpload->Unmap(0, nullptr);
commandList->CopyBufferRegion(m_postVertexBuffer.Get(), 0, postVertexBufferUpload.Get(), 0, vertexBufferSize);
commandList->ResourceBarrier(1, &CD3DX12_RESOURCE_BARRIER::Transition(m_postVertexBuffer.Get(), D3D12_RESOURCE_STATE_COPY_DEST, D3D12_RESOURCE_STATE_VERTEX_AND_CONSTANT_BUFFER));
// Initialize the vertex buffer views.
m_postVertexBufferView.BufferLocation = m_postVertexBuffer->GetGPUVirtualAddress();
m_postVertexBufferView.StrideInBytes = sizeof(PostVertex);
m_postVertexBufferView.SizeInBytes = vertexBufferSize;
}
// Create the constant buffer.
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(sizeof(SceneConstantBuffer) * FrameCount),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_sceneConstantBuffer)));
NAME_D3D12_OBJECT(m_sceneConstantBuffer);
// Describe and create constant buffer views.
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.BufferLocation = m_sceneConstantBuffer->GetGPUVirtualAddress();
cbvDesc.SizeInBytes = sizeof(SceneConstantBuffer);
CD3DX12_CPU_DESCRIPTOR_HANDLE cpuHandle(m_cbvSrvHeap->GetCPUDescriptorHandleForHeapStart(), 1, m_cbvSrvDescriptorSize);
for (UINT n = 0; n < FrameCount; n++)
{
m_device->CreateConstantBufferView(&cbvDesc, cpuHandle);
cbvDesc.BufferLocation += sizeof(SceneConstantBuffer);
cpuHandle.Offset(m_cbvSrvDescriptorSize);
}
// Map and initialize the constant buffer. We don't unmap this until the
// app closes. Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_sceneConstantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin)));
memcpy(m_pCbvDataBegin, &m_sceneConstantBufferData, sizeof(m_sceneConstantBufferData));
}
// Close the resource creation command list and execute it to begin the vertex buffer copy into
// the default heap.
ThrowIfFailed(commandList->Close());
ID3D12CommandList* ppCommandLists[] = { commandList.Get() };
m_commandQueue->ExecuteCommandLists(_countof(ppCommandLists), ppCommandLists);
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(m_fenceValues[m_frameIndex], D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValues[m_frameIndex]++;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute before continuing.
WaitForGpu();
}
}
// Load the sample assets.
void D3D12HelloTriangle::LoadAssets()
{
// Create an empty root signature.
{
CD3DX12_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init(0, nullptr, 0, nullptr, D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3D12SerializeRootSignature(&rootSignatureDesc, D3D_ROOT_SIGNATURE_VERSION_1, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &vertexShader, nullptr));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &pixelShader, nullptr));
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "COLOR", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 12, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
// Describe and create the graphics pipeline state object (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = { reinterpret_cast<UINT8*>(vertexShader->GetBufferPointer()), vertexShader->GetBufferSize() };
psoDesc.PS = { reinterpret_cast<UINT8*>(pixelShader->GetBufferPointer()), pixelShader->GetBufferSize() };
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState.DepthEnable = FALSE;
psoDesc.DepthStencilState.StencilEnable = FALSE;
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
}
// Create the command list.
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocator.Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_commandList)));
// Command lists are created in the recording state, but there is nothing
// to record yet. The main loop expects it to be closed, so close it now.
ThrowIfFailed(m_commandList->Close());
// Create the vertex buffer.
{
// Define the geometry for a triangle.
Vertex triangleVertices[] =
{
{ { 0.0f, 0.25f * m_aspectRatio, 0.0f }, { 1.0f, 0.0f, 0.0f, 1.0f } },
{ { 0.25f, -0.25f * m_aspectRatio, 0.0f }, { 0.0f, 1.0f, 0.0f, 1.0f } },
{ { -0.25f, -0.25f * m_aspectRatio, 0.0f }, { 0.0f, 0.0f, 1.0f, 1.0f } }
};
const UINT vertexBufferSize = sizeof(triangleVertices);
// Note: using upload heaps to transfer static data like vert buffers is not
// recommended. Every time the GPU needs it, the upload heap will be marshalled
// over. Please read up on Default Heap usage. An upload heap is used here for
// code simplicity and because there are very few verts to actually transfer.
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
// Copy the triangle data to the vertex buffer.
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_vertexBuffer->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin)));
memcpy(pVertexDataBegin, triangleVertices, sizeof(triangleVertices));
m_vertexBuffer->Unmap(0, nullptr);
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = vertexBufferSize;
}
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(0, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValue = 1;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForPreviousFrame();
}
}
// Load the sample assets.
void D3D12HelloConstBuffers::LoadAssets()
{
// Create a root signature consisting of a descriptor table with a single CBV.
{
D3D12_FEATURE_DATA_ROOT_SIGNATURE featureData = {};
// This is the highest version the sample supports. If CheckFeatureSupport succeeds, the HighestVersion returned will not be greater than this.
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_1;
if (FAILED(m_device->CheckFeatureSupport(D3D12_FEATURE_ROOT_SIGNATURE, &featureData, sizeof(featureData))))
{
featureData.HighestVersion = D3D_ROOT_SIGNATURE_VERSION_1_0;
}
CD3DX12_DESCRIPTOR_RANGE1 ranges[1];
CD3DX12_ROOT_PARAMETER1 rootParameters[1];
ranges[0].Init(D3D12_DESCRIPTOR_RANGE_TYPE_CBV, 1, 0, 0, D3D12_DESCRIPTOR_RANGE_FLAG_DATA_STATIC);
rootParameters[0].InitAsDescriptorTable(1, &ranges[0], D3D12_SHADER_VISIBILITY_VERTEX);
// Allow input layout and deny uneccessary access to certain pipeline stages.
D3D12_ROOT_SIGNATURE_FLAGS rootSignatureFlags =
D3D12_ROOT_SIGNATURE_FLAG_ALLOW_INPUT_ASSEMBLER_INPUT_LAYOUT |
D3D12_ROOT_SIGNATURE_FLAG_DENY_HULL_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_DOMAIN_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_GEOMETRY_SHADER_ROOT_ACCESS |
D3D12_ROOT_SIGNATURE_FLAG_DENY_PIXEL_SHADER_ROOT_ACCESS;
CD3DX12_VERSIONED_ROOT_SIGNATURE_DESC rootSignatureDesc;
rootSignatureDesc.Init_1_1(_countof(rootParameters), rootParameters, 0, nullptr, rootSignatureFlags);
ComPtr<ID3DBlob> signature;
ComPtr<ID3DBlob> error;
ThrowIfFailed(D3DX12SerializeVersionedRootSignature(&rootSignatureDesc, featureData.HighestVersion, &signature, &error));
ThrowIfFailed(m_device->CreateRootSignature(0, signature->GetBufferPointer(), signature->GetBufferSize(), IID_PPV_ARGS(&m_rootSignature)));
}
// Create the pipeline state, which includes compiling and loading shaders.
{
ComPtr<ID3DBlob> vertexShader;
ComPtr<ID3DBlob> pixelShader;
#if defined(_DEBUG)
// Enable better shader debugging with the graphics debugging tools.
UINT compileFlags = D3DCOMPILE_DEBUG | D3DCOMPILE_SKIP_OPTIMIZATION;
#else
UINT compileFlags = 0;
#endif
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "VSMain", "vs_5_0", compileFlags, 0, &vertexShader, nullptr));
ThrowIfFailed(D3DCompileFromFile(GetAssetFullPath(L"shaders.hlsl").c_str(), nullptr, nullptr, "PSMain", "ps_5_0", compileFlags, 0, &pixelShader, nullptr));
// Define the vertex input layout.
D3D12_INPUT_ELEMENT_DESC inputElementDescs[] =
{
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, 0, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 },
{ "COLOR", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, 12, D3D12_INPUT_CLASSIFICATION_PER_VERTEX_DATA, 0 }
};
// Describe and create the graphics pipeline state object (PSO).
D3D12_GRAPHICS_PIPELINE_STATE_DESC psoDesc = {};
psoDesc.InputLayout = { inputElementDescs, _countof(inputElementDescs) };
psoDesc.pRootSignature = m_rootSignature.Get();
psoDesc.VS = CD3DX12_SHADER_BYTECODE(vertexShader.Get());
psoDesc.PS = CD3DX12_SHADER_BYTECODE(pixelShader.Get());
psoDesc.RasterizerState = CD3DX12_RASTERIZER_DESC(D3D12_DEFAULT);
psoDesc.BlendState = CD3DX12_BLEND_DESC(D3D12_DEFAULT);
psoDesc.DepthStencilState.DepthEnable = FALSE;
psoDesc.DepthStencilState.StencilEnable = FALSE;
psoDesc.SampleMask = UINT_MAX;
psoDesc.PrimitiveTopologyType = D3D12_PRIMITIVE_TOPOLOGY_TYPE_TRIANGLE;
psoDesc.NumRenderTargets = 1;
psoDesc.RTVFormats[0] = DXGI_FORMAT_R8G8B8A8_UNORM;
psoDesc.SampleDesc.Count = 1;
ThrowIfFailed(m_device->CreateGraphicsPipelineState(&psoDesc, IID_PPV_ARGS(&m_pipelineState)));
}
// Create the command list.
ThrowIfFailed(m_device->CreateCommandList(0, D3D12_COMMAND_LIST_TYPE_DIRECT, m_commandAllocator.Get(), m_pipelineState.Get(), IID_PPV_ARGS(&m_commandList)));
// Command lists are created in the recording state, but there is nothing
// to record yet. The main loop expects it to be closed, so close it now.
ThrowIfFailed(m_commandList->Close());
// Create the vertex buffer.
{
// Define the geometry for a triangle.
Vertex triangleVertices[] =
{
{ { 0.0f, 0.25f * m_aspectRatio, 0.0f }, { 1.0f, 0.0f, 0.0f, 1.0f } },
{ { 0.25f, -0.25f * m_aspectRatio, 0.0f }, { 0.0f, 1.0f, 0.0f, 1.0f } },
{ { -0.25f, -0.25f * m_aspectRatio, 0.0f }, { 0.0f, 0.0f, 1.0f, 1.0f } }
};
const UINT vertexBufferSize = sizeof(triangleVertices);
// Note: using upload heaps to transfer static data like vert buffers is not
// recommended. Every time the GPU needs it, the upload heap will be marshalled
// over. Please read up on Default Heap usage. An upload heap is used here for
// code simplicity and because there are very few verts to actually transfer.
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(vertexBufferSize),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_vertexBuffer)));
// Copy the triangle data to the vertex buffer.
UINT8* pVertexDataBegin;
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_vertexBuffer->Map(0, &readRange, reinterpret_cast<void**>(&pVertexDataBegin)));
memcpy(pVertexDataBegin, triangleVertices, sizeof(triangleVertices));
m_vertexBuffer->Unmap(0, nullptr);
// Initialize the vertex buffer view.
m_vertexBufferView.BufferLocation = m_vertexBuffer->GetGPUVirtualAddress();
m_vertexBufferView.StrideInBytes = sizeof(Vertex);
m_vertexBufferView.SizeInBytes = vertexBufferSize;
}
// Create the constant buffer.
{
ThrowIfFailed(m_device->CreateCommittedResource(
&CD3DX12_HEAP_PROPERTIES(D3D12_HEAP_TYPE_UPLOAD),
D3D12_HEAP_FLAG_NONE,
&CD3DX12_RESOURCE_DESC::Buffer(1024 * 64),
D3D12_RESOURCE_STATE_GENERIC_READ,
nullptr,
IID_PPV_ARGS(&m_constantBuffer)));
// Describe and create a constant buffer view.
D3D12_CONSTANT_BUFFER_VIEW_DESC cbvDesc = {};
cbvDesc.BufferLocation = m_constantBuffer->GetGPUVirtualAddress();
cbvDesc.SizeInBytes = (sizeof(SceneConstantBuffer) + 255) & ~255; // CB size is required to be 256-byte aligned.
m_device->CreateConstantBufferView(&cbvDesc, m_cbvHeap->GetCPUDescriptorHandleForHeapStart());
// Map and initialize the constant buffer. We don't unmap this until the
// app closes. Keeping things mapped for the lifetime of the resource is okay.
CD3DX12_RANGE readRange(0, 0); // We do not intend to read from this resource on the CPU.
ThrowIfFailed(m_constantBuffer->Map(0, &readRange, reinterpret_cast<void**>(&m_pCbvDataBegin)));
memcpy(m_pCbvDataBegin, &m_constantBufferData, sizeof(m_constantBufferData));
}
// Create synchronization objects and wait until assets have been uploaded to the GPU.
{
ThrowIfFailed(m_device->CreateFence(0, D3D12_FENCE_FLAG_NONE, IID_PPV_ARGS(&m_fence)));
m_fenceValue = 1;
// Create an event handle to use for frame synchronization.
m_fenceEvent = CreateEvent(nullptr, FALSE, FALSE, nullptr);
if (m_fenceEvent == nullptr)
{
ThrowIfFailed(HRESULT_FROM_WIN32(GetLastError()));
}
// Wait for the command list to execute; we are reusing the same command
// list in our main loop but for now, we just want to wait for setup to
// complete before continuing.
WaitForPreviousFrame();
}
}