P4 to Git Change 1501660 by gandryey@gera-w8 on 2018/01/09 14:04:56
SWDEV-79445 - OCL generic changes and code clean-up - Code style clean-up. No functional changes. Affected files ... ... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#127 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.hpp#164 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#321 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuprogram.cpp#236 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#412 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.hpp#142 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palblit.cpp#15 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palcompiler.cpp#20 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#70 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.hpp#22 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevicegl.cpp#9 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palkernel.cpp#42 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palmemory.cpp#18 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palprintf.cpp#7 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palprogram.cpp#53 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.cpp#45 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#69 edit
This commit is contained in:
@@ -1031,7 +1031,6 @@ void Kernel::findLocalWorkSize(size_t workDim, const amd::NDRange& gblWorkSize,
|
||||
if (workGroupInfo()->compileSize_[0] == 0) {
|
||||
// Find the default local workgroup size, if it wasn't specified
|
||||
if (lclWorkSize[0] == 0) {
|
||||
size_t thrPerGrp;
|
||||
bool b1DOverrideSet = !flagIsDefault(GPU_MAX_WORKGROUP_SIZE);
|
||||
bool b2DOverrideSet = !flagIsDefault(GPU_MAX_WORKGROUP_SIZE_2D_X) ||
|
||||
!flagIsDefault(GPU_MAX_WORKGROUP_SIZE_2D_Y);
|
||||
@@ -1043,7 +1042,7 @@ void Kernel::findLocalWorkSize(size_t workDim, const amd::NDRange& gblWorkSize,
|
||||
((workDim == 3) && b3DOverrideSet);
|
||||
if (!overrideSet) {
|
||||
// Find threads per group
|
||||
thrPerGrp = workGroupInfo()->size_;
|
||||
size_t thrPerGrp = workGroupInfo()->size_;
|
||||
|
||||
// Check if kernel uses images
|
||||
if ((flags() & ImageEnable) &&
|
||||
@@ -1376,8 +1375,6 @@ bool Kernel::bindConstantBuffers(VirtualGPU& gpu) const {
|
||||
|
||||
void Kernel::processMemObjects(VirtualGPU& gpu, const amd::Kernel& kernel, const_address params,
|
||||
bool nativeMem) const {
|
||||
VirtualGPU::MemoryDependency& dependecy = gpu.memoryDependency();
|
||||
|
||||
// Mark the tracker with a new kernel,
|
||||
// so we can avoid checks of the aliased objects
|
||||
gpu.memoryDependency().newKernel();
|
||||
@@ -3323,7 +3320,6 @@ void HSAILKernel::findLocalWorkSize(size_t workDim, const amd::NDRange& gblWorkS
|
||||
if (workGroupInfo()->compileSize_[0] == 0) {
|
||||
// Find the default local workgroup size, if it wasn't specified
|
||||
if (lclWorkSize[0] == 0) {
|
||||
size_t thrPerGrp;
|
||||
bool b1DOverrideSet = !flagIsDefault(GPU_MAX_WORKGROUP_SIZE);
|
||||
bool b2DOverrideSet = !flagIsDefault(GPU_MAX_WORKGROUP_SIZE_2D_X) ||
|
||||
!flagIsDefault(GPU_MAX_WORKGROUP_SIZE_2D_Y);
|
||||
@@ -3335,7 +3331,7 @@ void HSAILKernel::findLocalWorkSize(size_t workDim, const amd::NDRange& gblWorkS
|
||||
((workDim == 3) && b3DOverrideSet);
|
||||
if (!overrideSet) {
|
||||
// Find threads per group
|
||||
thrPerGrp = workGroupInfo()->size_;
|
||||
size_t thrPerGrp = workGroupInfo()->size_;
|
||||
|
||||
// Check if kernel uses images
|
||||
if (flags_.imageEna_ &&
|
||||
|
||||
Reference in New Issue
Block a user