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

P4 to Git Change 1568521 by gandryey@gera-w8 on 2018/06/14 17:43:52

Browse Source 
SWDEV-79445 - OCL generic changes and code clean-up
	- Change LDS setup to account the size, since LC forces 4 bytes for LDS offsets always

	http://ocltc.amd.com/reviews/r/15197/

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_execute.cpp#28 edit
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_program.cpp#49 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#130 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#327 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palblit.cpp#25 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palkernel.cpp#56 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#109 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocblit.hpp#11 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rockernel.cpp#38 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/command.cpp#88 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/kernel.cpp#34 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/kernel.hpp#25 edit


[ROCm/clr commit: f6d8feabeb]
This commit is contained in:
foreman
2018-06-14 17:51:35 -04:00
parent cf0a8d99de
commit 7e6f2cf63b
10 changed files with 127 additions and 131 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/clr/rocclr/runtime/device/pal/palblit.cpp
+15 -15
View File
@@ -938,14 +938,15 @@ static void setArgument(amd::Kernel* kernel, size_t index, size_t size, const vo
const amd::KernelParameterDescriptor& desc = kernel->signature().at(index);
void* param = kernel->parameters().values() + desc.offset_;
assert((desc.type_ == T_POINTER || value != NULL || desc.size_ == 0) &&
"not a valid local mem arg");
assert((desc.type_ == T_POINTER || value != NULL ||
(desc.addressQualifier_ == CL_KERNEL_ARG_ADDRESS_LOCAL)) &&
"not a valid local mem arg");
uint32_t uint32_value = 0;
uint64_t uint64_value = 0;
size_t argSize = desc.size_;
if (desc.type_ == T_POINTER && desc.size_ != 0) {
if (desc.type_ == T_POINTER && (desc.addressQualifier_ != CL_KERNEL_ARG_ADDRESS_LOCAL)) {
if ((value == NULL) || (static_cast<const cl_mem*>(value) == NULL)) {
reinterpret_cast<Memory**>(kernel->parameters().values() +
kernel->parameters().memoryObjOffset())[desc.info_.arrayIndex_] = nullptr;
@@ -966,27 +967,26 @@ static void setArgument(amd::Kernel* kernel, size_t index, size_t size, const vo
} else if (desc.type_ == T_SAMPLER) {
assert(false && "No sampler support in blit manager! Use internal samplers!");
} else
switch (argSize) {
case 1:
uint32_value = *static_cast<const uint8_t*>(value);
break;
case 2:
uint32_value = *static_cast<const uint16_t*>(value);
break;
switch (desc.size_) {
case 4:
uint32_value = *static_cast<const uint32_t*>(value);
if (desc.addressQualifier_ == CL_KERNEL_ARG_ADDRESS_LOCAL) {
uint32_value = size;
} else {
uint32_value = *static_cast<const uint32_t*>(value);
}
break;
case 8:
uint64_value = *static_cast<const uint64_t*>(value);
if (desc.addressQualifier_ == CL_KERNEL_ARG_ADDRESS_LOCAL) {
uint64_value = size;
} else {
uint64_value = *static_cast<const uint64_t*>(value);
}
break;
default:
break;
}
switch (argSize) {
case 0 /*local mem*/:
*static_cast<size_t*>(param) = size;
break;
case sizeof(uint32_t):
*static_cast<uint32_t*>(param) = uint32_value;
break;