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

P4 to Git Change 1598525 by gandryey@gera-ocl-lc on 2018/08/27 14:28:05

Browse Source 
SWDEV-79445 - OCL generic changes and code clean-up
	- Setup KernelParameterDescriptor directly from the binary's metadata without any intermediate arguments presentation.

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_program.cpp#50 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#312 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palkernel.cpp#60 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palkernel.hpp#19 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#122 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocvirtual.cpp#66 edit
This commit is contained in:
foreman
2018-08-27 14:44:08 -04:00
parent e2bbfbe758
commit 4d46463948
5 changed files with 570 additions and 736 deletions
Download Patch File Download Diff File Expand all files Collapse all files
rocclr/runtime/device/pal/palvirtual.cpp
+32 -26
View File
@@ -1918,46 +1918,50 @@ void VirtualGPU::PrintChildren(const HSAILKernel& hsaKernel, VirtualGPU* gpuDefQ
uint offsArg = kernarg_address - gpuDefQueue->virtualQueue_->vmAddress();
address argum = gpuDefQueue->virtualQueue_->data() + offsArg;
print << "Kernel: " << child->name() << "\n";
for (auto arg : child->arguments()) {
const amd::KernelSignature& signature = child->signature();
// Check if runtime has to setup hidden arguments
for (const auto it : signature.parameters()) {
const char* extraArgName = nullptr;
switch (arg->type_) {
case HSAIL_ARGTYPE_HIDDEN_GLOBAL_OFFSET_X:
switch (it.info_.oclObject_) {
case amd::KernelParameterDescriptor::HiddenNone:
// void* zero = 0;
// WriteAqlArgAt(const_cast<address>(parameters), &zero, it.size_, it.offset_);
break;
case amd::KernelParameterDescriptor::HiddenGlobalOffsetX:
extraArgName = "Offset0: ";
break;
case HSAIL_ARGTYPE_HIDDEN_GLOBAL_OFFSET_Y:
case amd::KernelParameterDescriptor::HiddenGlobalOffsetY:
extraArgName = "Offset1: ";
break;
case HSAIL_ARGTYPE_HIDDEN_GLOBAL_OFFSET_Z:
case amd::KernelParameterDescriptor::HiddenGlobalOffsetZ:
extraArgName = "Offset2: ";
break;
case HSAIL_ARGTYPE_HIDDEN_PRINTF_BUFFER:
case amd::KernelParameterDescriptor::HiddenPrintfBuffer:
extraArgName = "PrintfBuf: ";
break;
case HSAIL_ARGTYPE_HIDDEN_DEFAULT_QUEUE:
case amd::KernelParameterDescriptor::HiddenDefaultQueue:
extraArgName = "VqueuePtr: ";
break;
case HSAIL_ARGTYPE_HIDDEN_COMPLETION_ACTION:
case amd::KernelParameterDescriptor::HiddenCompletionAction:
extraArgName = "AqlWrap: ";
break;
case HSAIL_ARGTYPE_HIDDEN_NONE:
extraArgName = "Unknown: ";
break;
default:
default:
break;
}
if (extraArgName) {
print << "\t" << extraArgName << *(size_t*)argum;
print << "\n";
argum += sizeof(size_t);
continue;
print << "\t" << extraArgName << *reinterpret_cast<size_t*>(argum);
print << "\n";
argum += sizeof(size_t);
continue;
}
print << "\t" << arg->name_ << ": ";
for (int s = arg->size_ - 1; s >= 0; --s) {
print.width(2);
print.fill('0');
print << (uint32_t)(argum[s]);
print << "\t" << it.name_ << ": ";
for (int s = it.size_- 1; s >= 0; --s) {
print.width(2);
print.fill('0');
print << static_cast<uint32_t>(argum[s]);
}
argum += arg->size_;
argum += it.offset_;
print << "\n";
}
printf("%s", print.str().c_str());
@@ -3141,10 +3145,12 @@ bool VirtualGPU::processMemObjectsHSA(const amd::Kernel& kernel, const_address p
gpuMem->wait(*this, WaitOnBusyEngine);
addVmMemory(gpuMem);
void* globalAddress = *(void**)(const_cast<address>(params) + desc.offset_);
LogPrintfInfo("!\targ%d: %s %s = ptr:%p obj:[%p-%p] threadId : %zx\n", index, desc.typeName_, desc.name_,
globalAddress, (void*)gpuMem->vmAddress(),
(void*)((intptr_t)gpuMem->vmAddress() + gpuMem->size()), std::this_thread::get_id());
const void* globalAddress = *reinterpret_cast<const void* const*>(params + desc.offset_);
LogPrintfInfo("!\targ%d: %s %s = ptr:%p obj:[%p-%p] threadId : %zx\n", index,
desc.typeName_.c_str(), desc.name_.c_str(),
globalAddress, reinterpret_cast<void*>(gpuMem->vmAddress()),
reinterpret_cast<void*>(gpuMem->vmAddress() + gpuMem->size()),
std::this_thread::get_id());
//! Check if compiler expects read/write.
//! Note: SVM with subbuffers has an issue with tracking.