AuroraMiddleware/OpenSubdiv
1
0
Fork 0
mirror of https://github.com/PixarAnimationStudios/OpenSubdiv synced 2024-11-09 22:00:06 +00:00
Code Issues Releases Wiki Activity

Removes FarKernelBatch.

Browse Source 
This is the first step to tease off Osd compute controller/contexts
from Far API.
Currently FarStencilTable only creates a kernelbatch for the entire range,
so we can use [0, numStencils) for all cases instead of KernelBatch.
This might not be true if we apply non-factorized level-wise stencils,
then we'll add another modular utility to serve those cases.
This commit is contained in:
Takahito Tejima 2015-04-09 11:16:54 -07:00
parent 50d30b101e
commit 8da827336d
39 changed files with 318 additions and 558 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
examples/glEvalLimit/glEvalLimit.cpp
View File

@ -197,8 +197,6 @@ Osd::CpuComputeContext * g_computeCtx = 0;
Osd::CpuComputeController g_computeCtrl;
Far::KernelBatchVector g_kernelBatches;
Osd::CpuEvalLimitContext * g_evalCtx = 0;
Osd::CpuEvalLimitController g_evalCtrl;
@ -242,7 +240,7 @@ updateGeom() {
g_vertexData->UpdateData( &g_positions[0], 0, nverts);
g_computeCtrl.Compute(g_computeCtx, g_kernelBatches, g_vertexData, g_varyingData);
g_computeCtrl.Compute(g_computeCtx, g_vertexData, g_varyingData);
s.Stop();
g_computeTime = float(s.GetElapsed() * 1000.0f);
@ -417,9 +415,6 @@ createOsdMesh(ShapeDesc const & shapeDesc, int level) {
delete g_evalCtx;
g_evalCtx = Osd::CpuEvalLimitContext::Create(*patchTables);
g_kernelBatches.clear();
g_kernelBatches.push_back(Far::StencilTablesFactory::Create(*concatStencils));
}
{ // Create vertex primvar buffer for the CVs

6
examples/glShareTopology/glShareTopology.cpp
View File

@ -295,8 +295,6 @@ public:
_computeContext = ComputeContext::Create(vertexStencils, varyingStencils);
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*vertexStencils));
_numVertices = vertexStencils->GetNumStencils() +
vertexStencils->GetNumControlVertices();
}
@ -328,7 +326,6 @@ public:
globalVaryingDesc.stride);
_computeController.Compute(_computeContext,
_kernelBatches,
typedInstance->GetVertexBuffer(),
typedInstance->GetVaryingBuffer(),
&vertexDesc,
@ -364,7 +361,6 @@ public:
private:
COMPUTE_CONTROLLER _computeController;
ComputeContext *_computeContext;
Far::KernelBatchVector _kernelBatches;
};
// ---------------------------------------------------------------------------
@ -394,8 +390,6 @@ Topology(Far::PatchTables const * patchTables,
_computeContext = ComputeContext::Create(g_clContext, vertexStencils, varyingStencils);
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*vertexStencils));
_numVertices = vertexStencils->GetNumStencils() +
vertexStencils->GetNumControlVertices();
}

2
opensubdiv/far/CMakeLists.txt
View File

@ -45,8 +45,6 @@ set(PUBLIC_HEADER_FILES
error.h
gregoryBasis.h
interpolate.h
kernelBatch.h
kernelBatchDispatcher.h
patchDescriptor.h
patchParam.h
patchMap.h

82
opensubdiv/far/kernelBatch.h
View File

@ -1,82 +0,0 @@
//
// Copyright 2013 Pixar
//
// Licensed under the Apache License, Version 2.0 (the "Apache License")
// with the following modification; you may not use this file except in
// compliance with the Apache License and the following modification to it:
// Section 6. Trademarks. is deleted and replaced with:
//
// 6. Trademarks. This License does not grant permission to use the trade
// names, trademarks, service marks, or product names of the Licensor
// and its affiliates, except as required to comply with Section 4(c) of
// the License and to reproduce the content of the NOTICE file.
//
// You may obtain a copy of the Apache License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the Apache License with the above modification is
// distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
// KIND, either express or implied. See the Apache License for the specific
// language governing permissions and limitations under the Apache License.
//
#ifndef FAR_KERNEL_BATCH_H
#define FAR_KERNEL_BATCH_H
#include "../version.h"
#include <vector>
namespace OpenSubdiv {
namespace OPENSUBDIV_VERSION {
namespace Far {
/// \brief A GP Compute Kernel descriptor.
///
/// Vertex refinement through subdivision schemes requires the successive
/// application of dedicated compute kernels. OpenSubdiv groups these vertices
/// in batches based on their topology in order to minimize the number of kernel
/// switches to process a given primitive.
///
struct KernelBatch {
public:
enum KernelType {
KERNEL_UNKNOWN=0,
KERNEL_STENCIL_TABLE,
KERNEL_USER_DEFINED
};
/// \brief Constructor.
///
/// @param _kernelType The type of compute kernel kernel
///
/// @param _level The level of subdivision of the vertices in the batch
///
/// @param _start Index of the first vertex in the batch
///
/// @param _end Index of the last vertex in the batch
///
KernelBatch( int _kernelType, int _level, int _start, int _end ) :
kernelType(_kernelType), level(_level), start(_start), end(_end) { }
int kernelType, // the type of compute kernel kernel
level, // the level of subdivision of the vertices in the batch
start, // index of the first vertex in the batch
end; // index of the last vertex in the batch
};
typedef std::vector<KernelBatch> KernelBatchVector;
} // end namespace Far
} // end namespace OPENSUBDIV_VERSION
using namespace OPENSUBDIV_VERSION;
} // end namespace OpenSubdiv
#endif /* FAR_KERNEL_BATCH_H */

164
opensubdiv/far/kernelBatchDispatcher.h
View File

@ -1,164 +0,0 @@
//
// Copyright 2013 Pixar
//
// Licensed under the Apache License, Version 2.0 (the "Apache License")
// with the following modification; you may not use this file except in
// compliance with the Apache License and the following modification to it:
// Section 6. Trademarks. is deleted and replaced with:
//
// 6. Trademarks. This License does not grant permission to use the trade
// names, trademarks, service marks, or product names of the Licensor
// and its affiliates, except as required to comply with Section 4(c) of
// the License and to reproduce the content of the NOTICE file.
//
// You may obtain a copy of the Apache License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the Apache License with the above modification is
// distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
// KIND, either express or implied. See the Apache License for the specific
// language governing permissions and limitations under the Apache License.
//
#ifndef FAR_KERNELBATCH_DISPATCHER_H
#define FAR_KERNELBATCH_DISPATCHER_H
#include "../version.h"
#include "../far/kernelBatch.h"
#include "../far/stencilTables.h"
#include <cassert>
namespace OpenSubdiv {
namespace OPENSUBDIV_VERSION {
namespace Far {
/// \brief Subdivision refinement encapsulation layer.
///
/// The kernel dispatcher allows client code to customize parts or the entire
/// computation process. This pattern aims at hiding the logic specific to
/// the subdivision algorithms and expose a simplified access to minimalistic
/// compute kernels. By default, meshes revert to a default dispatcher that
/// implements single-threaded CPU kernels.
///
/// - derive a dispatcher class from this one
/// - override the virtual functions
/// - pass the derived dispatcher to the factory (one instance can be shared by many meshes)
///
/// Note : the caller is responsible for deleting a custom dispatcher
///
class KernelBatchDispatcher {
public:
/// \brief Launches the processing of a vector of kernel batches
/// this is a convenient API for controllers which don't have any user defined kernels.
///
/// @param controller refinement controller implementation (vertex array)
///
/// @param context refinement context implementation (subdivision tables)
/// passed to the controller.
///
/// @param batches batches of kernels that need to be processed
///
/// @param maxlevel process vertex batches up to this level
///
template <class CONTROLLER, class CONTEXT> static void Apply(
CONTROLLER *controller, CONTEXT *context, KernelBatchVector const & batches, int maxlevel);
protected:
/// \brief Launches the processing of a kernel batch
/// returns true if the batch is handled, otherwise returns false (i.e. user defined kernel)
///
/// @param controller refinement controller implementation
///
/// @param context refinement context implementation
///
/// @param batch a batch of kernel that need to be processed
///
template <class CONTROLLER, class CONTEXT> static bool ApplyKernel(
CONTROLLER *controller, CONTEXT *context, KernelBatch const &batch);
};
///
/// \brief Far default controller implementation
///
/// This is Far's default implementation of a kernal batch controller.
///
class DefaultController {
public:
template <class CONTEXT> void ApplyStencilTableKernel(
KernelBatch const &batch, CONTEXT *context) const;
};
// Launches the processing of a kernel batch
template <class CONTROLLER, class CONTEXT> bool
KernelBatchDispatcher::ApplyKernel(CONTROLLER *controller, CONTEXT *context,
KernelBatch const &batch) {
if (batch.end==0) {
return true;
}
switch(batch.kernelType) {
case KernelBatch::KERNEL_UNKNOWN:
assert(0);
case KernelBatch::KERNEL_STENCIL_TABLE:
controller->ApplyStencilTableKernel(batch, context);
break;
default: // user defined kernel type
return false;
}
return true;
}
// Launches the processing of a vector of kernel batches
template <class CONTROLLER, class CONTEXT> void
KernelBatchDispatcher::Apply(CONTROLLER *controller, CONTEXT *context,
KernelBatchVector const & batches, int maxlevel) {
for (int i = 0; i < (int)batches.size(); ++i) {
const KernelBatch &batch = batches[i];
if (maxlevel>=0 and batch.level>=maxlevel) {
continue;
}
ApplyKernel(controller, context, batch);
}
}
template <class CONTEXT> void
DefaultController::ApplyStencilTableKernel(
KernelBatch const &batch, CONTEXT *context) const {
StencilTables const * stencilTables = context->GetStencilTables();
assert(stencilTables);
typename CONTEXT::VertexType *vsrc = &context->GetVertices().at(0),
*vdst = vsrc + batch.start + stencilTables->GetNumControlVertices();
stencilTables->UpdateValues(vsrc, vdst, batch.start, batch.end);
}
} // end namespace Far
} // end namespace OPENSUBDIV_VERSION
using namespace OPENSUBDIV_VERSION;
} // end namespace OpenSubdiv
#endif /* FAR_KERNELBATCH_DISPATCHER_H */

9
opensubdiv/far/stencilTablesFactory.cpp
View File

@ -395,15 +395,6 @@ LimitStencilTablesFactory::Create(TopologyRefiner const & refiner,
return result;
}
//------------------------------------------------------------------------------
KernelBatch
StencilTablesFactory::Create(StencilTables const &stencilTables) {
return KernelBatch( KernelBatch::KERNEL_STENCIL_TABLE,
-1, 0, stencilTables.GetNumStencils());
}
} // end namespace Far
} // end namespace OPENSUBDIV_VERSION

8
opensubdiv/far/stencilTablesFactory.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatch.h"
#include "../far/patchTables.h"
#include <vector>
@ -103,13 +102,6 @@ public:
///
static StencilTables const * Create(int numTables, StencilTables const ** tables);
/// \brief Returns a KernelBatch applying all the stencil in the tables
/// to primvar data.
///
/// @param stencilTables The stencil tables to batch
///
static KernelBatch Create(StencilTables const &stencilTables);
private:
// Generate stencils for the coarse control-vertices (single weight = 1.0f)

17
opensubdiv/osd/clComputeContext.cpp
View File

@ -60,6 +60,7 @@ public:
_offsets = createCLBuffer(stencilTables.GetOffsets(), clContext);
_indices = createCLBuffer(stencilTables.GetControlIndices(), clContext);
_weights = createCLBuffer(stencilTables.GetWeights(), clContext);
_numStencils = stencilTables.GetNumStencils();
}
~CLStencilTables() {
@ -89,12 +90,17 @@ public:
return _weights;
}
int GetNumStencils() const {
return _numStencils;
}
private:
cl_mem _sizes,
_offsets,
_indices,
_weights;
int _numStencils;
};
// -----------------------------------------------------------------------------
@ -104,7 +110,7 @@ CLComputeContext::CLComputeContext(
Far::StencilTables const * varyingStencilTables,
cl_context clContext) :
_vertexStencilTables(0), _varyingStencilTables(0),
_numControlVertices(0) {
_numControlVertices(0) {
if (vertexStencilTables) {
_vertexStencilTables = new CLStencilTables(*vertexStencilTables, clContext);
@ -139,6 +145,15 @@ CLComputeContext::HasVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->IsValid() : false;
}
int
CLComputeContext::GetNumStencilsInVertexStencilTables() const {
return _vertexStencilTables ? _vertexStencilTables->GetNumStencils() : 0;
}
int
CLComputeContext::GetNumStencilsInVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->GetNumStencils() : 0;
}
// ----------------------------------------------------------------------------
cl_mem

6
opensubdiv/osd/clComputeContext.h
View File

@ -80,6 +80,12 @@ public:
return _numControlVertices;
}
/// Returns the number of stencils in vertex stencil table
int GetNumStencilsInVertexStencilTables() const;
/// Returns the number of stencils in varying stencil table
int GetNumStencilsInVaryingStencilTables() const;
/// Returns the Cuda buffer containing vertex-stencil stencil sizes
cl_mem GetVertexStencilTablesSizes() const;

24
opensubdiv/osd/clComputeController.cpp
View File

@ -32,6 +32,7 @@
#include <algorithm>
#include <string.h>
#include <sstream>
#include <cassert>
namespace OpenSubdiv {
namespace OPENSUBDIV_VERSION {
@ -138,18 +139,20 @@ private:
// ----------------------------------------------------------------------------
void
CLComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch, ComputeContext const *context) {
CLComputeController::ApplyStencilTableKernel(ComputeContext const *context) {
assert(context);
cl_int errNum;
size_t globalWorkSize[1] = { (size_t)(batch.end - batch.start) };
size_t globalWorkSize = 0;
int ncvs = context->GetNumControlVertices();
if (context->HasVertexStencilTables()) {
int start = 0;
int end = context->GetNumStencilsInVertexStencilTables();
globalWorkSize = (size_t)(end - start);
KernelBundle const * bundle = getKernel(_currentBindState.vertexDesc);
@ -167,14 +170,14 @@ CLComputeController::ApplyStencilTableKernel(
clSetKernelArg(kernel, 3, sizeof(cl_mem), &indices);
clSetKernelArg(kernel, 4, sizeof(cl_mem), &weights);
clSetKernelArg(kernel, 5, sizeof(int), &batch.start);
clSetKernelArg(kernel, 6, sizeof(int), &batch.end);
clSetKernelArg(kernel, 5, sizeof(int), &start);
clSetKernelArg(kernel, 6, sizeof(int), &end);
clSetKernelArg(kernel, 7, sizeof(int), &_currentBindState.vertexDesc.offset);
clSetKernelArg(kernel, 8, sizeof(int), &ncvs);
errNum = clEnqueueNDRangeKernel(
_clQueue, kernel, 1, NULL, globalWorkSize, NULL, 0, NULL, NULL);
_clQueue, kernel, 1, NULL, &globalWorkSize, NULL, 0, NULL, NULL);
if (errNum!=CL_SUCCESS) {
Far::Error(Far::FAR_RUNTIME_ERROR,
"ApplyStencilTableKernel (%d) ", errNum);
@ -182,6 +185,9 @@ CLComputeController::ApplyStencilTableKernel(
}
if (context->HasVaryingStencilTables()) {
int start = 0;
int end = context->GetNumStencilsInVaryingStencilTables();
globalWorkSize = (size_t)(end - start);
KernelBundle const * bundle = getKernel(_currentBindState.varyingDesc);
@ -199,14 +205,14 @@ CLComputeController::ApplyStencilTableKernel(
clSetKernelArg(kernel, 3, sizeof(cl_mem), &indices);
clSetKernelArg(kernel, 4, sizeof(cl_mem), &weights);
clSetKernelArg(kernel, 5, sizeof(int), &batch.start);
clSetKernelArg(kernel, 6, sizeof(int), &batch.end);
clSetKernelArg(kernel, 5, sizeof(int), &start);
clSetKernelArg(kernel, 6, sizeof(int), &end);
clSetKernelArg(kernel, 7, sizeof(int), &_currentBindState.varyingDesc.offset);
clSetKernelArg(kernel, 8, sizeof(int), &ncvs);
errNum = clEnqueueNDRangeKernel(
_clQueue, kernel, 1, NULL, globalWorkSize, NULL, 0, NULL, NULL);
_clQueue, kernel, 1, NULL, &globalWorkSize, NULL, 0, NULL, NULL);
if (errNum!=CL_SUCCESS) {
Far::Error(Far::FAR_RUNTIME_ERROR,
"ApplyStencilTableKernel (%d)", errNum);

17
opensubdiv/osd/clComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/clComputeContext.h"
#include "../osd/vertexDescriptor.h"
#include "../osd/opencl.h"
@ -70,9 +69,6 @@ public:
///
/// @param context The CLContext to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
/// @param vertexDesc The descriptor of vertex elements to be refined.
@ -87,17 +83,14 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( CLComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
bind(vertexBuffer, varyingBuffer, vertexDesc, varyingDesc);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(context);
unbind();
}
@ -113,10 +106,9 @@ public:
///
template<class VERTEX_BUFFER>
void Compute(CLComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -130,10 +122,7 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context);
void ApplyStencilTableKernel(ComputeContext const *context);
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void bind( VERTEX_BUFFER * vertexBuffer,

2
opensubdiv/osd/cpuComputeContext.h
View File

@ -61,7 +61,7 @@ public:
/// interpolation
///
static CpuComputeContext * Create(Far::StencilTables const * vertexStencilTables,
Far::StencilTables const * varyingStencilTables=0);
Far::StencilTables const * varyingStencilTables=0);
/// Destructor
virtual ~CpuComputeContext();

36
opensubdiv/osd/cpuComputeController.cpp
View File

@ -46,7 +46,7 @@ CpuComputeController::Synchronize() {
void
CpuComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch, ComputeContext const *context) const {
ComputeContext const *context) const {
assert(context);
@ -61,14 +61,17 @@ CpuComputeController::ApplyStencilTableKernel(
float * destBuffer = _currentBindState.vertexBuffer + desc.offset +
vertexStencils->GetNumControlVertices() * desc.stride;
int start = 0;
int end = vertexStencils->GetNumStencils();
CpuComputeStencils(_currentBindState.vertexDesc,
srcBuffer, destBuffer,
&vertexStencils->GetSizes().at(0),
&vertexStencils->GetOffsets().at(0),
&vertexStencils->GetControlIndices().at(0),
&vertexStencils->GetWeights().at(0),
batch.start,
batch.end);
srcBuffer, destBuffer,
&vertexStencils->GetSizes().at(0),
&vertexStencils->GetOffsets().at(0),
&vertexStencils->GetControlIndices().at(0),
&vertexStencils->GetWeights().at(0),
start,
end);
}
Far::StencilTables const * varyingStencils = context->GetVaryingStencilTables();
@ -82,14 +85,17 @@ CpuComputeController::ApplyStencilTableKernel(
float * destBuffer = _currentBindState.varyingBuffer + desc.offset +
varyingStencils->GetNumControlVertices() * desc.stride;
int start = 0;
int end = varyingStencils->GetNumStencils();
CpuComputeStencils(_currentBindState.varyingDesc,
srcBuffer, destBuffer,
&varyingStencils->GetSizes().at(0),
&varyingStencils->GetOffsets().at(0),
&varyingStencils->GetControlIndices().at(0),
&varyingStencils->GetWeights().at(0),
batch.start,
batch.end);
srcBuffer, destBuffer,
&varyingStencils->GetSizes().at(0),
&varyingStencils->GetOffsets().at(0),
&varyingStencils->GetControlIndices().at(0),
&varyingStencils->GetWeights().at(0),
start,
end);
}
}

20
opensubdiv/osd/cpuComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/cpuComputeContext.h"
#include "../osd/vertexDescriptor.h"
@ -64,9 +63,6 @@ public:
///
/// @param context The CpuContext to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
/// @param vertexDesc The descriptor of vertex elements to be refined.
@ -81,17 +77,14 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( CpuComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
bind(vertexBuffer, varyingBuffer, vertexDesc, varyingDesc);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(context);
unbind();
}
@ -100,17 +93,13 @@ public:
///
/// @param context The CpuContext to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
template<class VERTEX_BUFFER>
void Compute(CpuComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -119,10 +108,7 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context) const;
void ApplyStencilTableKernel(ComputeContext const *context) const;
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void bind( VERTEX_BUFFER * vertexBuffer,

16
opensubdiv/osd/cudaComputeContext.cpp
View File

@ -67,6 +67,7 @@ public:
_offsets = createCudaBuffer(stencilTables.GetOffsets());
_indices = createCudaBuffer(stencilTables.GetControlIndices());
_weights = createCudaBuffer(stencilTables.GetWeights());
_numStencils = stencilTables.GetNumStencils();
}
~CudaStencilTables() {
@ -96,11 +97,16 @@ public:
return _weights;
}
int GetNumStencils() const {
return _numStencils;
}
private:
void * _sizes,
* _offsets,
* _indices,
* _weights;
int _numStencils;
};
// ----------------------------------------------------------------------------
@ -144,6 +150,16 @@ CudaComputeContext::HasVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->IsValid() : false;
}
int
CudaComputeContext::GetNumStencilsInVertexStencilTables() const {
return _vertexStencilTables ? _vertexStencilTables->GetNumStencils() : 0;
}
int
CudaComputeContext::GetNumStencilsInVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->GetNumStencils() : 0;
}
// ----------------------------------------------------------------------------
void *

6
opensubdiv/osd/cudaComputeContext.h
View File

@ -77,6 +77,12 @@ public:
return _numControlVertices;
}
/// Returns the number of stencils in vertex stencil tables
int GetNumStencilsInVertexStencilTables() const;
/// Returns the number of stencils in varying stencil tables
int GetNumStencilsInVaryingStencilTables() const;
/// Returns the Cuda buffer containing vertex-stencil stencil sizes
void * GetVertexStencilTablesSizes() const;

17
opensubdiv/osd/cudaComputeController.cpp
View File

@ -26,6 +26,7 @@
#include <cuda_runtime.h>
#include <string.h>
#include <cassert>
extern "C" {
@ -46,7 +47,7 @@ namespace Osd {
void
CudaComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch, ComputeContext const *context) const {
ComputeContext const *context) const {
assert(context);
@ -55,6 +56,9 @@ CudaComputeController::ApplyStencilTableKernel(
int length = _currentBindState.vertexDesc.length,
stride = _currentBindState.vertexDesc.stride;
int start = 0;
int end = context->GetNumStencilsInVertexStencilTables();
float const * src = _currentBindState.GetVertexBufferAtOffset();
float * dst = const_cast<float *>(src) +
@ -65,8 +69,8 @@ CudaComputeController::ApplyStencilTableKernel(
(int const *)context->GetVertexStencilTablesOffsets(),
(int const *)context->GetVertexStencilTablesIndices(),
(float const *)context->GetVertexStencilTablesWeights(),
batch.start,
batch.end);
start,
end);
}
if (context->HasVaryingStencilTables()) {
@ -74,6 +78,9 @@ CudaComputeController::ApplyStencilTableKernel(
int length = _currentBindState.varyingDesc.length,
stride = _currentBindState.varyingDesc.stride;
int start = 0;
int end = context->GetNumStencilsInVaryingStencilTables();
float const * src = _currentBindState.GetVaryingBufferAtOffset();
float * dst = const_cast<float *>(src) +
@ -84,8 +91,8 @@ CudaComputeController::ApplyStencilTableKernel(
(int const *)context->GetVaryingStencilTablesOffsets(),
(int const *)context->GetVaryingStencilTablesIndices(),
(float const *)context->GetVaryingStencilTablesWeights(),
batch.start,
batch.end);
start,
end);
}
}

14
opensubdiv/osd/cudaComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/cudaComputeContext.h"
#include "../osd/vertexDescriptor.h"
@ -77,17 +76,14 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( CudaComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
bind(vertexBuffer, varyingBuffer, vertexDesc, varyingDesc);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(context);
unbind();
}
@ -103,10 +99,9 @@ public:
///
template<class VERTEX_BUFFER>
void Compute(CudaComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -114,10 +109,7 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context) const;
void ApplyStencilTableKernel(ComputeContext const *context) const;
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void bind( VERTEX_BUFFER * vertexBuffer,

18
opensubdiv/osd/d3d11ComputeContext.cpp
View File

@ -125,6 +125,8 @@ public:
_indices.initialize(stencilTables.GetControlIndices(), DXGI_FORMAT_R32_SINT, deviceContext);
_weights.initialize(stencilTables.GetWeights(), DXGI_FORMAT_R32_FLOAT, deviceContext);
_numStencils = stencilTables.GetNumStencils();
}
bool IsValid() const {
@ -148,6 +150,10 @@ public:
return _weights;
}
int GetNumStencils() const {
return _numStencils;
}
void Bind(ID3D11DeviceContext * deviceContext) const {
ID3D11ShaderResourceView *SRViews[] = {
_sizes.srv,
@ -170,6 +176,8 @@ private:
_offsets,
_indices,
_weights;
int _numStencils;
};
// ----------------------------------------------------------------------------
@ -217,6 +225,16 @@ D3D11ComputeContext::HasVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->IsValid() : false;
}
int
D3D11ComputeContext::GetNumStencilsInVertexStencilTables() const {
return _vertexStencilTables ? _vertexStencilTables->GetNumStencils() : 0;
}
int
D3D11ComputeContext::GetNumStencilsInVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->GetNumStencils() : 0;
}
// ----------------------------------------------------------------------------
void

6
opensubdiv/osd/d3d11ComputeContext.h
View File

@ -78,6 +78,12 @@ public:
return _numControlVertices;
}
/// Returns the number of stencils in vertex stencil table
int GetNumStencilsInVertexStencilTables() const;
/// Returns the number of stencils in varying stencil table
int GetNumStencilsInVaryingStencilTables() const;
/// Binds D3D11 buffers containing stencils for 'vertex' interpolation
///
/// @param deviceContext The D3D device

16
opensubdiv/osd/d3d11ComputeController.cpp
View File

@ -155,11 +155,11 @@ public:
}
void ApplyStencilTableKernel(ID3D11DeviceContext *deviceContext,
Far::KernelBatch const &batch, int offset, int numCVs) {
int offset, int numCVs, int start, int end) {
KernelUniformArgs args;
args.uniformStart = batch.start;
args.uniformEnd = batch.end;
args.uniformStart = start;
args.uniformEnd = end;
args.uniformOffset = offset;
args.uniformNumCVs = numCVs;
@ -299,7 +299,7 @@ D3D11ComputeController::unbindBuffer() {
void
D3D11ComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch, D3D11ComputeContext const *context) const {
D3D11ComputeContext const *context, int numStencils) const {
assert(context);
@ -307,8 +307,12 @@ D3D11ComputeController::ApplyStencilTableKernel(
D3D11ComputeController::KernelBundle * bundle =
const_cast<D3D11ComputeController::KernelBundle *>(_currentBindState.kernelBundle);
bundle->ApplyStencilTableKernel(_deviceContext,
batch, _currentBindState.desc.offset, context->GetNumControlVertices());
bundle->ApplyStencilTableKernel(
_deviceContext,
_currentBindState.desc.offset,
context->GetNumControlVertices(),
0,
numStencils);
}

25
opensubdiv/osd/d3d11ComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/d3d11ComputeContext.h"
#include "../osd/vertexDescriptor.h"
@ -69,9 +68,6 @@ public:
///
/// @param context The D3D11Context to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
/// @param vertexDesc The descriptor of vertex elements to be refined.
@ -86,20 +82,18 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( D3D11ComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
if (vertexBuffer) {
bind(vertexBuffer, vertexDesc);
context->BindVertexStencilTables(_deviceContext);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(
context, context->GetNumStencilsInVertexStencilTables());
}
if (varyingBuffer) {
@ -107,7 +101,8 @@ public:
context->BindVaryingStencilTables(_deviceContext);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(
context, context->GetNumStencilsInVaryingStencilTables());
}
context->UnbindStencilTables(_deviceContext);
@ -119,17 +114,13 @@ public:
///
/// @param context The D3D11Context to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
template<class VERTEX_BUFFER>
void Compute(D3D11ComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -137,10 +128,8 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context) const;
void ApplyStencilTableKernel(ComputeContext const *context,
int numStencils) const;
template<class BUFFER>
void bind( BUFFER * buffer,

21
opensubdiv/osd/d3d11Mesh.h
View File

@ -78,7 +78,6 @@ public:
Mesh(ComputeController * computeController,
Far::TopologyRefiner * refiner,
Far::PatchTables * patchTables,
Far::KernelBatchVector const & kernelBatches,
VertexBuffer * vertexBuffer,
VertexBuffer * varyingBuffer,
ComputeContext * computeContext,
@ -87,7 +86,6 @@ public:
_refiner(refiner),
_patchTables(patchTables),
_kernelBatches(kernelBatches),
_vertexBuffer(vertexBuffer),
_varyingBuffer(varyingBuffer),
_computeContext(computeContext),
@ -119,13 +117,14 @@ public:
_varyingBuffer->UpdateData(varyingData, startVertex, numVerts, _d3d11DeviceContext);
}
virtual void Refine() {
_computeController->Compute(_computeContext, _kernelBatches, _vertexBuffer, _varyingBuffer);
_computeController->Compute(_computeContext, _vertexBuffer, _varyingBuffer);
}
virtual void Refine(VertexBufferDescriptor const *vertexDesc,
VertexBufferDescriptor const *varyingDesc,
bool interleaved) {
_computeController->Compute(_computeContext, _kernelBatches,
_vertexBuffer, (interleaved ? _vertexBuffer : _varyingBuffer),
_computeController->Compute(_computeContext,
_vertexBuffer,
(interleaved ? _vertexBuffer : _varyingBuffer),
vertexDesc, varyingDesc);
}
virtual void Synchronize() {
@ -176,8 +175,6 @@ private:
if (numVertexElements>0) {
vertexStencils = Far::StencilTablesFactory::Create(*_refiner, options);
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*vertexStencils));
}
if (numVaryingElements>0) {
@ -235,7 +232,6 @@ private:
Far::TopologyRefiner * _refiner;
Far::PatchTables * _patchTables;
Far::KernelBatchVector _kernelBatches;
VertexBuffer *_vertexBuffer;
VertexBuffer *_varyingBuffer;
@ -286,7 +282,6 @@ public:
Mesh(ComputeController * computeController,
Far::TopologyRefiner * refiner,
Far::PatchTables * patchTables,
Far::KernelBatchVector const & kernelBatches,
VertexBuffer * vertexBuffer,
VertexBuffer * varyingBuffer,
ComputeContext * computeContext,
@ -295,7 +290,6 @@ public:
_refiner(refiner),
_patchTables(patchTables),
_kernelBatches(kernelBatches),
_vertexBuffer(vertexBuffer),
_varyingBuffer(varyingBuffer),
_computeContext(computeContext),
@ -324,12 +318,12 @@ public:
_varyingBuffer->UpdateData(varyingData, startVertex, numVerts, _d3d11DeviceContext);
}
virtual void Refine() {
_computeController->Compute(_computeContext, _kernelBatches, _vertexBuffer, _varyingBuffer);
_computeController->Compute(_computeContext, _vertexBuffer, _varyingBuffer);
}
virtual void Refine(VertexBufferDescriptor const *vertexDesc,
VertexBufferDescriptor const *varyingDesc,
bool interleaved) {
_computeController->Compute(_computeContext, _kernelBatches,
_computeController->Compute(_computeContext,
_vertexBuffer, (interleaved ? _vertexBuffer : _varyingBuffer),
vertexDesc, varyingDesc);
}
@ -382,8 +376,6 @@ private:
if (numVertexElements>0) {
vertexStencils = Far::StencilTablesFactory::Create(*_refiner, options);
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*vertexStencils));
}
if (numVaryingElements>0) {
@ -441,7 +433,6 @@ private:
Far::TopologyRefiner * _refiner;
Far::PatchTables * _patchTables;
Far::KernelBatchVector _kernelBatches;
VertexBuffer *_vertexBuffer;
VertexBuffer *_varyingBuffer;

25
opensubdiv/osd/glMesh.h
View File

@ -98,7 +98,6 @@ public:
Mesh(ComputeController * computeController,
Far::TopologyRefiner * refiner,
Far::PatchTables * patchTables,
Far::KernelBatchVector const & kernelBatches,
VertexBuffer * vertexBuffer,
VertexBuffer * varyingBuffer,
ComputeContext * computeContext,
@ -106,7 +105,6 @@ public:
_refiner(refiner),
_patchTables(patchTables),
_kernelBatches(kernelBatches),
_vertexBuffer(vertexBuffer),
_varyingBuffer(varyingBuffer),
_computeContext(computeContext),
@ -140,14 +138,15 @@ public:
}
virtual void Refine() {
_computeController->Compute(_computeContext, _kernelBatches, _vertexBuffer, _varyingBuffer);
_computeController->Compute(_computeContext, _vertexBuffer, _varyingBuffer);
}
virtual void Refine(VertexBufferDescriptor const * vertexDesc,
VertexBufferDescriptor const * varyingDesc,
bool interleaved) {
_computeController->Compute(_computeContext, _kernelBatches,
_vertexBuffer, (interleaved ? _vertexBuffer : _varyingBuffer),
_computeController->Compute(_computeContext,
_vertexBuffer,
(interleaved ? _vertexBuffer : _varyingBuffer),
vertexDesc, varyingDesc);
}
@ -255,8 +254,6 @@ private:
varyingStencils = concatVaryingStencils;
}
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*vertexStencils));
_computeContext = ComputeContext::Create(vertexStencils,
varyingStencils);
@ -279,7 +276,6 @@ private:
Far::TopologyRefiner * _refiner;
Far::PatchTables * _patchTables;
Far::KernelBatchVector _kernelBatches;
VertexBuffer *_vertexBuffer;
VertexBuffer *_varyingBuffer;
@ -353,7 +349,6 @@ public:
Mesh(ComputeController * computeController,
Far::TopologyRefiner * refiner,
Far::PatchTables * patchTables,
Far::KernelBatchVector const & kernelBatches,
VertexBuffer * vertexBuffer,
VertexBuffer * varyingBuffer,
ComputeContext * computeContext,
@ -363,7 +358,6 @@ public:
_refiner(refiner),
_patchTables(patchTables),
_kernelBatches(kernelBatches),
_vertexBuffer(vertexBuffer),
_varyingBuffer(varyingBuffer),
_computeContext(computeContext),
@ -395,14 +389,15 @@ public:
}
virtual void Refine() {
_computeController->Compute(_computeContext, _kernelBatches, _vertexBuffer, _varyingBuffer);
_computeController->Compute(_computeContext, _vertexBuffer, _varyingBuffer);
}
virtual void Refine(VertexBufferDescriptor const *vertexDesc,
VertexBufferDescriptor const *varyingDesc,
bool interleaved) {
_computeController->Compute(_computeContext, _kernelBatches,
_vertexBuffer, (interleaved ? _vertexBuffer : _varyingBuffer),
_computeController->Compute(_computeContext,
_vertexBuffer,
(interleaved ? _vertexBuffer : _varyingBuffer),
vertexDesc, varyingDesc);
}
@ -499,8 +494,6 @@ private:
Far::StencilTables const *concatStencils =
Far::StencilTablesFactory::Create(2, inStencils);
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*concatStencils));
Far::StencilTables const *inVaryingStencils[] = {
varyingStencils, endCapVaryingStencils
};
@ -513,7 +506,6 @@ private:
delete varyingStencils;
varyingStencils = concatVaryingStencils;
}
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*vertexStencils));
_computeContext = ComputeContext::Create(_clContext,
vertexStencils,
@ -538,7 +530,6 @@ private:
Far::TopologyRefiner * _refiner;
Far::PatchTables * _patchTables;
Far::KernelBatchVector _kernelBatches;
VertexBuffer *_vertexBuffer;
VertexBuffer *_varyingBuffer;

20
opensubdiv/osd/glslComputeContext.cpp
View File

@ -71,6 +71,7 @@ public:
_offsets = createGLSLBuffer(stencilTables.GetOffsets());
_indices = createGLSLBuffer(stencilTables.GetControlIndices());
_weights = createGLSLBuffer(stencilTables.GetWeights());
_numStencils = stencilTables.GetNumStencils();
}
~GLSLStencilTables() {
@ -100,6 +101,10 @@ public:
return _weights;
}
int GetNumStencils() const {
return _numStencils;
}
void Bind() const {
glBindBufferBase(GL_SHADER_STORAGE_BUFFER, 1, _sizes);
glBindBufferBase(GL_SHADER_STORAGE_BUFFER, 2, _offsets);
@ -122,6 +127,7 @@ private:
_offsets,
_indices,
_weights;
int _numStencils;
};
// -----------------------------------------------------------------------------
@ -130,7 +136,8 @@ GLSLComputeContext::GLSLComputeContext(
Far::StencilTables const * vertexStencilTables,
Far::StencilTables const * varyingStencilTables) :
_vertexStencilTables(0), _varyingStencilTables(0),
_numControlVertices(0) {
_numControlVertices(0),
_numStencils(0) {
if (vertexStencilTables) {
_vertexStencilTables = new GLSLStencilTables(*vertexStencilTables);
@ -165,8 +172,17 @@ GLSLComputeContext::HasVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->IsValid() : false;
}
// ----------------------------------------------------------------------------
int
GLSLComputeContext::GetNumStencilsInVertexStencilTables() const {
return _vertexStencilTables ? _vertexStencilTables->GetNumStencils() : false;
}
int
GLSLComputeContext::GetNumStencilsInVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->GetNumStencils() : false;
}
// ----------------------------------------------------------------------------
void
GLSLComputeContext::BindVertexStencilTables() const {

11
opensubdiv/osd/glslComputeContext.h
View File

@ -76,10 +76,16 @@ public:
return _numControlVertices;
}
/// Returns the Cuda buffer containing vertex-stencil stencil sizes
/// Returns the number of stencils in vertex stencil table
int GetNumStencilsInVertexStencilTables() const;
/// Returns the number of stencils in varying stencil table
int GetNumStencilsInVaryingStencilTables() const;
/// Returns the GL buffer containing vertex-stencil stencil sizes
GLuint GetVertexStencilTablesSizes() const;
/// Returns the Cuda buffer containing vertex-stencil stencil offsets
/// Returns the GL buffer containing vertex-stencil stencil offsets
GLuint GetVertexStencilTablesOffsets() const;
/// Binds GL buffers containing stencils for 'vertex' interpolation
@ -104,6 +110,7 @@ private:
* _varyingStencilTables;
int _numControlVertices;
int _numStencils;
};
} // end namespace Osd

18
opensubdiv/osd/glslComputeController.cpp
View File

@ -24,7 +24,6 @@
#include "../osd/glslComputeController.h"
#include "../osd/vertexDescriptor.h"
//#include "../osd/debug.h"
#include "../osd/opengl.h"
#include "../far/error.h"
@ -138,12 +137,13 @@ public:
return true;
}
void ApplyStencilTableKernel(Far::KernelBatch const &batch, int offset, int numCVs) const {
void ApplyStencilTableKernel(int offset, int numCVs,
int start, int end) const {
// select stencil GLSL subroutine
glUniformSubroutinesuiv(GL_COMPUTE_SHADER, 1, &_subStencilKernel);
dispatchCompute(offset, numCVs, batch.start, batch.end);
dispatchCompute(offset, numCVs, start, end);
}
struct Match {
@ -214,12 +214,20 @@ private:
void
GLSLComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch, ComputeContext const *context) const {
ComputeContext const *context, int numStencils) const {
assert(context);
// Note: GLSLComputeContext has a state, knowing whether vertex or
// varying stencil tables are being bound. GetNumStencils() reflects it.
// This structure will likely be revisited.
int start = 0;
int end = numStencils;
_currentBindState.kernelBundle->ApplyStencilTableKernel(
batch, _currentBindState.desc.offset, context->GetNumControlVertices());
_currentBindState.desc.offset, context->GetNumControlVertices(),
start, end);
}
// ----------------------------------------------------------------------------

21
opensubdiv/osd/glslComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/glslComputeContext.h"
#include "../osd/vertexDescriptor.h"
@ -79,28 +78,27 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( GLSLComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
if (vertexBuffer) {
bind(vertexBuffer, vertexDesc);
context->BindVertexStencilTables();
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(
context, context->GetNumStencilsInVertexStencilTables());
}
if (varyingBuffer) {
bind(varyingBuffer, varyingDesc);
context->BindVaryingStencilTables();
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(
context, context->GetNumStencilsInVaryingStencilTables());
}
context->UnbindStencilTables();
@ -119,10 +117,9 @@ public:
///
template<class VERTEX_BUFFER>
void Compute(GLSLComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -130,10 +127,8 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context) const;
void ApplyStencilTableKernel(ComputeContext const *context,
int numStencils) const;
template<class BUFFER>
void bind( BUFFER * buffer,

18
opensubdiv/osd/glslTransformFeedbackComputeContext.cpp
View File

@ -86,6 +86,7 @@ public:
_offsets = createGLTextureBuffer(stencilTables.GetOffsets(), GL_R32I);
_indices = createGLTextureBuffer(stencilTables.GetControlIndices(), GL_R32I);
_weights = createGLTextureBuffer(stencilTables.GetWeights(), GL_R32F);
_numStencils = stencilTables.GetNumStencils();
}
~GLStencilTables() {
@ -115,12 +116,18 @@ public:
return _weights;
}
int GetNumStencils() const {
return _numStencils;
}
private:
GLuint _sizes,
_offsets,
_indices,
_weights;
int _numStencils;
};
// -----------------------------------------------------------------------------
@ -164,8 +171,17 @@ GLSLTransformFeedbackComputeContext::HasVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->IsValid() : false;
}
// ----------------------------------------------------------------------------
int
GLSLTransformFeedbackComputeContext::GetNumStencilsInVertexStencilTables() const {
return _vertexStencilTables ? _vertexStencilTables->GetNumStencils() : 0;
}
int
GLSLTransformFeedbackComputeContext::GetNumStencilsInVaryingStencilTables() const {
return _varyingStencilTables ? _varyingStencilTables->GetNumStencils() : 0;
}
// ----------------------------------------------------------------------------
GLuint
GLSLTransformFeedbackComputeContext::GetVertexStencilTablesSizes() const {
return _vertexStencilTables ? _vertexStencilTables->GetSizes() : 0;

8
opensubdiv/osd/glslTransformFeedbackComputeContext.h
View File

@ -59,7 +59,7 @@ public:
/// interpolation
///
static GLSLTransformFeedbackComputeContext * Create(Far::StencilTables const * vertexStencilTables,
Far::StencilTables const * varyingStencilTables=0);
Far::StencilTables const * varyingStencilTables=0);
/// Destructor
virtual ~GLSLTransformFeedbackComputeContext();
@ -75,6 +75,12 @@ public:
return _numControlVertices;
}
/// Returns the number of stencils in vertex stencil table
int GetNumStencilsInVertexStencilTables() const;
/// Returns the number of stencils in varying stencil table
int GetNumStencilsInVaryingStencilTables() const;
/// Returns the GL texture buffer containing vertex-stencil stencil sizes
GLuint GetVertexStencilTablesSizes() const;

44
opensubdiv/osd/glslTransformFeedbackComputeController.cpp
View File

@ -87,7 +87,7 @@ public:
_uniformWeights(0),
_uniformStart(0),
_uniformEnd(0),
_uniformOffset(0) { }
_uniformPrimvarOffset(0) { }
~KernelBundle() {
if (_program) {
@ -95,9 +95,8 @@ public:
}
}
void UseProgram(int primvarOffset) const {
void UseProgram() const {
glUseProgram(_program);
glUniform1i(_uniformOffset, primvarOffset);
}
bool Compile(VertexBufferDescriptor const & desc) {
@ -138,15 +137,19 @@ public:
// outVertexData[2]
// (gl_SkipComponents1)
//
// note that "primvarOffset" in shader is still needed to read
// interleaved components even if gl_SkipComponents is used.
//
char attrName[32];
for (int i = 0; i < desc.offset; ++i) {
int primvarOffset = (desc.offset % desc.stride);
for (int i = 0; i < primvarOffset; ++i) {
outputs.push_back("gl_SkipComponents1");
}
for (int i = 0; i < desc.length; ++i) {
snprintf(attrName, 32, "outVertexBuffer[%d]", i);
outputs.push_back(attrName);
}
for (int i = desc.offset + desc.length; i < desc.stride; ++i) {
for (int i = primvarOffset + desc.length; i < desc.stride; ++i) {
outputs.push_back("gl_SkipComponents1");
}
@ -192,7 +195,7 @@ public:
_uniformStart = glGetUniformLocation(_program, "batchStart");
_uniformEnd = glGetUniformLocation(_program, "batchEnd");
_uniformOffset = glGetUniformLocation(_program, "primvarOffset");
_uniformPrimvarOffset = glGetUniformLocation(_program, "primvarOffset");
OSD_DEBUG_CHECK_GL_ERROR("KernelBundle::Compile");
@ -225,14 +228,16 @@ public:
// set batch range
glUniform1i(_uniformStart, start);
glUniform1i(_uniformEnd, end);
glUniform1i(_uniformOffset, offset);
glUniform1i(_uniformPrimvarOffset, offset);
int count = end - start,
stride = _desc.stride*sizeof(float);
// note: offset includes both "batching offset" and "primvar offset".
//
glBindBufferRange(GL_TRANSFORM_FEEDBACK_BUFFER,
0, primvarBuffer,
(start + numCVs)*stride + offset*sizeof(float),
(start + numCVs)*stride + (offset - offset%stride)*sizeof(float),
count*stride);
glBeginTransformFeedback(GL_POINTS);
@ -246,13 +251,14 @@ public:
//OSD_DEBUG_CHECK_GL_ERROR("TransformPrimvarBuffer\n");
}
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
GLuint primvarBuffer, int offset, int numCVs) const {
void ApplyStencilTableKernel(GLuint primvarBuffer,
int offset, int numCVs,
int start, int end) const {
glUniformSubroutinesuiv(GL_VERTEX_SHADER, 1, &_subStencilKernel);
TransformPrimvarBuffer(primvarBuffer,
offset, numCVs, batch.start, batch.end);
offset, numCVs, start, end);
}
struct Match {
@ -283,7 +289,7 @@ private:
_uniformStart, // batch
_uniformEnd,
_uniformOffset; // GL primvar buffer descriptor
_uniformPrimvarOffset;
VertexBufferDescriptor _desc; // primvar buffer descriptor
};
@ -294,7 +300,7 @@ GLSLTransformFeedbackComputeController::bindBufferAndProgram(
GLuint & feedbackTexture) {
glEnable(GL_RASTERIZER_DISCARD);
_currentBindState.kernelBundle->UseProgram(/*primvarOffset*/0);
_currentBindState.kernelBundle->UseProgram();
if (not feedbackTexture) {
glGenTextures(1, &feedbackTexture);
@ -385,14 +391,18 @@ GLSLTransformFeedbackComputeController::getKernel(
void
GLSLTransformFeedbackComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch,
GLSLTransformFeedbackComputeContext const *context) const {
GLSLTransformFeedbackComputeContext const *context, int numStencils) const {
assert(context);
_currentBindState.kernelBundle->ApplyStencilTableKernel(batch,
int start = 0;
int end = numStencils;
_currentBindState.kernelBundle->ApplyStencilTableKernel(
_currentBindState.buffer, _currentBindState.desc.offset,
context->GetNumControlVertices());
context->GetNumControlVertices(),
start,
end);
}

25
opensubdiv/osd/glslTransformFeedbackComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/glslTransformFeedbackComputeContext.h"
#include "../osd/vertexDescriptor.h"
@ -67,9 +66,6 @@ public:
/// @param context The GLSLTransformFeedbackComputeContext to apply
/// refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
/// @param vertexDesc The descriptor of vertex elements to be refined.
@ -84,21 +80,19 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( GLSLTransformFeedbackComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
if (vertexBuffer) {
bind(vertexBuffer, vertexDesc, _vertexTexture);
bindContextStencilTables(context, false);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(
context, context->GetNumStencilsInVertexStencilTables());
}
if (varyingBuffer) {
@ -107,7 +101,8 @@ public:
bindContextStencilTables(context, true);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(
context, context->GetNumStencilsInVaryingStencilTables());
}
unbind();
}
@ -117,17 +112,13 @@ public:
/// @param context The GLSLTransformFeedbackComputeContext to apply
/// refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
template<class VERTEX_BUFFER>
void Compute(GLSLTransformFeedbackComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -135,10 +126,8 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context) const;
void ApplyStencilTableKernel(ComputeContext const *context,
int numStencils) const;
template<class BUFFER>
void bind( BUFFER * buffer, VertexBufferDescriptor const * desc,

10
opensubdiv/osd/mesh.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatch.h"
#include "../far/topologyRefiner.h"
#include "../far/patchTablesFactory.h"
#include "../far/stencilTables.h"
@ -154,7 +153,6 @@ public:
Mesh(ComputeController * computeController,
Far::TopologyRefiner * refiner,
Far::PatchTables * patchTables,
Far::KernelBatchVector const & kernelBatches,
VertexBuffer * vertexBuffer,
VertexBuffer * varyingBuffer,
ComputeContext * computeContext,
@ -162,7 +160,6 @@ public:
_refiner(refiner),
_patchTables(patchTables),
_kernelBatches(kernelBatches),
_vertexBuffer(vertexBuffer),
_varyingBuffer(varyingBuffer),
_computeContext(computeContext),
@ -192,11 +189,11 @@ public:
}
virtual void Refine() {
_computeController->Compute(_computeContext, _kernelBatches, _vertexBuffer, _varyingBuffer);
_computeController->Compute(_computeContext, _vertexBuffer, _varyingBuffer);
}
virtual void Refine(VertexBufferDescriptor const *vertexDesc, VertexBufferDescriptor const *varyingDesc) {
_computeController->Refine(_computeContext, _kernelBatches, _vertexBuffer, _varyingBuffer, vertexDesc, varyingDesc);
_computeController->Refine(_computeContext, _vertexBuffer, _varyingBuffer, vertexDesc, varyingDesc);
}
virtual void Synchronize() {
@ -237,8 +234,6 @@ private:
if (numVertexElements>0) {
vertexStencils = Far::StencilTablesFactory::Create(*_refiner, options);
_kernelBatches.push_back(Far::StencilTablesFactory::Create(*vertexStencils));
}
if (numVaryingElements>0) {
@ -291,7 +286,6 @@ private:
Far::TopologyRefiner * _refiner;
Far::PatchTables * _patchTables;
Far::KernelBatchVector _kernelBatches;
VertexBuffer * _vertexBuffer,
* _varyingBuffer;

36
opensubdiv/osd/ompComputeController.cpp
View File

@ -40,7 +40,7 @@ OmpComputeController::OmpComputeController(int numThreads) {
void
OmpComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch, ComputeContext const *context) const {
ComputeContext const *context) const {
assert(context);
@ -48,6 +48,9 @@ OmpComputeController::ApplyStencilTableKernel(
if (vertexStencils and _currentBindState.vertexBuffer) {
int start = 0;
int end = vertexStencils->GetNumStencils();
VertexBufferDescriptor const & desc = _currentBindState.vertexDesc;
float const * srcBuffer = _currentBindState.vertexBuffer + desc.offset;
@ -56,19 +59,22 @@ OmpComputeController::ApplyStencilTableKernel(
vertexStencils->GetNumControlVertices() * desc.stride;
OmpComputeStencils(_currentBindState.vertexDesc,
srcBuffer, destBuffer,
&vertexStencils->GetSizes().at(0),
&vertexStencils->GetOffsets().at(0),
&vertexStencils->GetControlIndices().at(0),
&vertexStencils->GetWeights().at(0),
batch.start,
batch.end);
srcBuffer, destBuffer,
&vertexStencils->GetSizes().at(0),
&vertexStencils->GetOffsets().at(0),
&vertexStencils->GetControlIndices().at(0),
&vertexStencils->GetWeights().at(0),
start,
end);
}
Far::StencilTables const * varyingStencils = context->GetVaryingStencilTables();
if (varyingStencils and _currentBindState.varyingBuffer) {
int start = 0;
int end = varyingStencils->GetNumStencils();
VertexBufferDescriptor const & desc = _currentBindState.varyingDesc;
float const * srcBuffer = _currentBindState.varyingBuffer + desc.offset;
@ -77,13 +83,13 @@ OmpComputeController::ApplyStencilTableKernel(
varyingStencils->GetNumControlVertices() * desc.stride;
OmpComputeStencils(_currentBindState.varyingDesc,
srcBuffer, destBuffer,
&varyingStencils->GetSizes().at(0),
&varyingStencils->GetOffsets().at(0),
&varyingStencils->GetControlIndices().at(0),
&varyingStencils->GetWeights().at(0),
batch.start,
batch.end);
srcBuffer, destBuffer,
&varyingStencils->GetSizes().at(0),
&varyingStencils->GetOffsets().at(0),
&varyingStencils->GetControlIndices().at(0),
&varyingStencils->GetWeights().at(0),
start,
end);
}
}

20
opensubdiv/osd/ompComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/cpuComputeContext.h"
#include "../osd/vertexDescriptor.h"
@ -66,9 +65,6 @@ public:
///
/// @param context The CpuContext to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
/// @param vertexDesc The descriptor of vertex elements to be refined.
@ -83,19 +79,16 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( CpuComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
omp_set_num_threads(_numThreads);
bind(vertexBuffer, varyingBuffer, vertexDesc, varyingDesc);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(context);
unbind();
}
@ -104,17 +97,13 @@ public:
///
/// @param context The CpuContext to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
template<class VERTEX_BUFFER>
void Compute(CpuComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -122,10 +111,7 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context) const;
void ApplyStencilTableKernel(ComputeContext const *context) const;
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void bind( VERTEX_BUFFER * vertexBuffer,

36
opensubdiv/osd/tbbComputeController.cpp
View File

@ -49,7 +49,7 @@ TbbComputeController::TbbComputeController(int numThreads)
void
TbbComputeController::ApplyStencilTableKernel(
Far::KernelBatch const &batch, ComputeContext const *context) const {
ComputeContext const *context) const {
assert(context);
@ -57,6 +57,9 @@ TbbComputeController::ApplyStencilTableKernel(
if (vertexStencils and _currentBindState.vertexBuffer) {
int start = 0;
int end = vertexStencils->GetNumStencils();
VertexBufferDescriptor const & desc = _currentBindState.vertexDesc;
float const * srcBuffer = _currentBindState.vertexBuffer + desc.offset;
@ -65,19 +68,22 @@ TbbComputeController::ApplyStencilTableKernel(
vertexStencils->GetNumControlVertices() * desc.stride;
TbbComputeStencils(_currentBindState.vertexDesc,
srcBuffer, destBuffer,
&vertexStencils->GetSizes().at(0),
&vertexStencils->GetOffsets().at(0),
&vertexStencils->GetControlIndices().at(0),
&vertexStencils->GetWeights().at(0),
batch.start,
batch.end);
srcBuffer, destBuffer,
&vertexStencils->GetSizes().at(0),
&vertexStencils->GetOffsets().at(0),
&vertexStencils->GetControlIndices().at(0),
&vertexStencils->GetWeights().at(0),
start,
end);
}
Far::StencilTables const * varyingStencils = context->GetVaryingStencilTables();
if (varyingStencils and _currentBindState.varyingBuffer) {
int start = 0;
int end = varyingStencils->GetNumStencils();
VertexBufferDescriptor const & desc = _currentBindState.varyingDesc;
float const * srcBuffer = _currentBindState.varyingBuffer + desc.offset;
@ -86,13 +92,13 @@ TbbComputeController::ApplyStencilTableKernel(
varyingStencils->GetNumControlVertices() * desc.stride;
TbbComputeStencils(_currentBindState.varyingDesc,
srcBuffer, destBuffer,
&varyingStencils->GetSizes().at(0),
&varyingStencils->GetOffsets().at(0),
&varyingStencils->GetControlIndices().at(0),
&varyingStencils->GetWeights().at(0),
batch.start,
batch.end);
srcBuffer, destBuffer,
&varyingStencils->GetSizes().at(0),
&varyingStencils->GetOffsets().at(0),
&varyingStencils->GetControlIndices().at(0),
&varyingStencils->GetWeights().at(0),
start,
end);
}
}

20
opensubdiv/osd/tbbComputeController.h
View File

@ -27,7 +27,6 @@
#include "../version.h"
#include "../far/kernelBatchDispatcher.h"
#include "../osd/cpuComputeContext.h"
#include "../osd/vertexDescriptor.h"
@ -62,9 +61,6 @@ public:
///
/// @param context The CpuContext to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
/// @param vertexDesc The descriptor of vertex elements to be refined.
@ -79,17 +75,14 @@ public:
///
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void Compute( CpuComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER * vertexBuffer,
VARYING_BUFFER * varyingBuffer,
VertexBufferDescriptor const * vertexDesc=NULL,
VertexBufferDescriptor const * varyingDesc=NULL ){
if (batches.empty()) return;
bind(vertexBuffer, varyingBuffer, vertexDesc, varyingDesc);
Far::KernelBatchDispatcher::Apply(this, context, batches, /*maxlevel*/ -1);
ApplyStencilTableKernel(context);
unbind();
}
@ -98,17 +91,13 @@ public:
///
/// @param context The CpuContext to apply refinement operations to
///
/// @param batches Vector of batches of vertices organized by operative
/// kernel
///
/// @param vertexBuffer Vertex-interpolated data buffer
///
template<class VERTEX_BUFFER>
void Compute(CpuComputeContext const * context,
Far::KernelBatchVector const & batches,
VERTEX_BUFFER *vertexBuffer) {
Compute<VERTEX_BUFFER>(context, batches, vertexBuffer, (VERTEX_BUFFER*)0);
Compute<VERTEX_BUFFER>(context, vertexBuffer, (VERTEX_BUFFER*)0);
}
/// Waits until all running subdivision kernels finish.
@ -116,10 +105,7 @@ public:
protected:
friend class Far::KernelBatchDispatcher;
void ApplyStencilTableKernel(Far::KernelBatch const &batch,
ComputeContext const *context) const;
void ApplyStencilTableKernel(ComputeContext const *context) const;
template<class VERTEX_BUFFER, class VARYING_BUFFER>
void bind( VERTEX_BUFFER * vertexBuffer,

18
regression/osd_regression/main.cpp
View File

@ -303,11 +303,7 @@ checkMeshCPU( FarTopologyRefiner *refiner,
vb->UpdateData( coarseverts[0].GetPos(), 0, (int)coarseverts.size() );
Far::KernelBatchVector kernelBatches;
kernelBatches.push_back(
Far::StencilTablesFactory::Create(*vertexStencils));
controller->Compute( context, kernelBatches, vb );
controller->Compute( context, vb );
int result = checkVertexBuffer(*refiner, refmesh, vb->BindCpuBuffer(),
vb->GetNumElements());
@ -341,11 +337,7 @@ checkMeshCPUGL(FarTopologyRefiner *refiner,
vb->UpdateData( coarseverts[0].GetPos(), 0, (int)coarseverts.size() );
Far::KernelBatchVector kernelBatches;
kernelBatches.push_back(
Far::StencilTablesFactory::Create(*vertexStencils));
controller->Compute( context, kernelBatches, vb );
controller->Compute( context, vb );
int result = checkVertexBuffer(*refiner, refmesh,
vb->BindCpuBuffer(), vb->GetNumElements());
@ -383,11 +375,7 @@ checkMeshCL( FarTopologyRefiner *refiner,
vb->UpdateData( coarseverts[0].GetPos(), 0, (int)coarseverts.size(),
g_clQueue );
Far::KernelBatchVector kernelBatches;
kernelBatches.push_back(
Far::StencilTablesFactory::Create(*vertexStencils));
controller->Compute( context, kernelBatches, vb );
controller->Compute( context, vb );
// read data back from CL buffer
size_t dataSize = vb->GetNumVertices() * vb->GetNumElements();

7
tutorials/osd/tutorial_0/osd_tutorial_0.cpp
View File

@ -75,8 +75,6 @@ int main(int, char **) {
Osd::CpuComputeContext * context=0;
Far::KernelBatchVector batches;
//
// Setup phase
//
@ -92,9 +90,6 @@ int main(int, char **) {
Far::StencilTables const * stencilTables =
Far::StencilTablesFactory::Create(*refiner, options);
// We need a kernel batch to dispatch Compute launches
batches.push_back(Far::StencilTablesFactory::Create(*stencilTables));
// Create an Osd Compute Context from the stencil tables
context = Osd::CpuComputeContext::Create(stencilTables);
@ -122,7 +117,7 @@ int main(int, char **) {
vbuffer->UpdateData(g_verts, 0, nCoarseVerts);
// Launch the computation
controller.Compute(context, batches, vbuffer);
controller.Compute(context, vbuffer);
}
{ // Visualization with Maya : print a MEL script that generates particles