AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

P4 to Git Change 1330796 by gandryey@gera-w8 on 2016/10/24 15:12:41

Browse Source 
SWDEV-86035 - Add PAL backend to OpenCL
	- Use loader for the code objects loading and avoid allocation of each individual kernel

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/Makefile#17 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palkernel.cpp#13 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palkernel.hpp#6 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palprogram.cpp#13 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palprogram.hpp#10 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#33 edit
This commit is contained in:
foreman
2016-10-24 15:25:54 -04:00
parent 5359f19211
commit 4e1f32bd71
5 changed files with 185 additions and 94 deletions
Download Patch File Download Diff File Expand all files Collapse all files
rocclr/runtime/device/pal/palkernel.cpp
+6 -27
View File
@@ -377,11 +377,11 @@ HSAILKernel::aqlCreateHWInfo(amd::hsa::loader::Symbol *sym)
if (!sym) {
return false;
}
uint64_t akc_addr = 0;
if (!sym->GetInfo(HSA_EXECUTABLE_SYMBOL_INFO_KERNEL_OBJECT, reinterpret_cast<void*>(&akc_addr))) {
if (!sym->GetInfo(HSA_EXECUTABLE_SYMBOL_INFO_KERNEL_OBJECT, reinterpret_cast<void*>(&code_))) {
return false;
}
amd_kernel_code_t *akc = reinterpret_cast<amd_kernel_code_t*>(akc_addr);
amd_kernel_code_t *akc = reinterpret_cast<amd_kernel_code_t*>(prog().findHostKernelAddress(code_));
cpuAqlCode_ = akc;
if (!sym->GetInfo(HSA_EXT_EXECUTABLE_SYMBOL_INFO_KERNEL_OBJECT_SIZE, reinterpret_cast<void*>(&codeSize_))) {
return false;
@@ -390,22 +390,6 @@ HSAILKernel::aqlCreateHWInfo(amd::hsa::loader::Symbol *sym)
if (!sym->GetInfo(HSA_EXT_EXECUTABLE_SYMBOL_INFO_KERNEL_OBJECT_ALIGN, reinterpret_cast<void*>(&akc_align))) {
return false;
}
// Allocate HW resources for the real program only
if (!prog().isNull()) {
code_ = new Memory(dev(), amd::alignUp(codeSize_, akc_align));
Resource::MemoryType type = Resource::Local;
// Initialize kernel ISA code
if (code_ && code_->create(type)) {
constexpr bool WaitForUpload = true;
code_->writeRawData(*code_->dev().xferQueue(), 0, codeSize_,
reinterpret_cast<void*>(akc), WaitForUpload);
}
else {
LogError("Failed to allocate ISA code!");
return false;
}
}
assert((akc->workitem_private_segment_byte_size & 3) == 0 &&
"Scratch must be DWORD aligned");
@@ -591,9 +575,8 @@ HSAILKernel::HSAILKernel(std::string name,
, dev_(prog->dev())
, prog_(*prog)
, index_(0)
, code_(nullptr)
, code_(0)
, codeSize_(0)
, hwMetaData_(nullptr)
, extraArgumentsNum_(extraArgsNum)
, waveLimiter_(this, (prog->isNull() ? 1 :
dev().properties().gfxipProperties.shaderCore.numCusPerShaderArray) * dev().hwInfo()->simdPerCU_)
@@ -608,10 +591,6 @@ HSAILKernel::~HSAILKernel()
delete arg;
arguments_.pop_back();
}
delete [] hwMetaData_;
delete code_;
}
bool
@@ -1217,7 +1196,7 @@ HSAILKernel::loadArguments(
// Initialize kernel ISA and execution buffer requirements
hsaDisp->private_segment_size = spillSegSize();
hsaDisp->group_segment_size = ldsAddress - ldsSize();
hsaDisp->kernel_object = gpuAqlCode()->vmAddress();
hsaDisp->kernel_object = gpuAqlCode();
ConstBuffer* cb = gpu.constBufs_[0];
cb->uploadDataToHw(argsBufferSize() + sizeof(hsa_kernel_dispatch_packet_t));
@@ -1228,7 +1207,7 @@ HSAILKernel::loadArguments(
hsaDisp->completion_signal.handle = 0;
memList.push_back(cb);
memList.push_back(gpuAqlCode());
memList.push_back(&prog().codeSegGpu());
for (pal::Memory * mem : prog().globalStores()) {
memList.push_back(mem);
}