P4 to Git Change 1129348 by smekhano@stas-solstice-llvm-merge on 2015/03/10 20:49:28

EPR #414133 - HSA HLC: fixed device enqueue code for no-capture case
	A dead store to a context structure may be generated by clang, while our code was only
	expecting a bitcast.

Affected files ...

... //depot/stg/opencl/drivers/opencl/compiler/llvm36/lib/Transforms/Scalar/AMDLowerEnqueueKernel.cpp#6 edit
Tento commit je obsažen v:
foreman
2015-03-11 10:15:48 -04:00
rodič 607c6fd805
revize 51f603a4cc
2 změnil soubory, kde provedl 5 přidání a 3 odebrání
+4 -2
Zobrazit soubor
@@ -165,11 +165,13 @@ HSAILKernel::aqlCreateHWInfo(const void* shader, size_t shaderSize)
char* headerBaseAddress =
reinterpret_cast<char*>(siMetaData->common.hShaderMemHandle);
hsa_ext_code_descriptor_t* hcd =
reinterpret_cast<hsa_ext_code_descriptor_t*>(headerBaseAddress);
amd_kernel_code_t* akc = reinterpret_cast<amd_kernel_code_t*>(
headerBaseAddress);
headerBaseAddress + hcd->code.handle);
address codeStartAddress = reinterpret_cast<address>(akc);
address codeEndAddress = reinterpret_cast<address>(akc) + siMetaData->common.codeLenInByte;
address codeEndAddress = reinterpret_cast<address>(hcd) + siMetaData->common.codeLenInByte;
codeSize_ = codeEndAddress - codeStartAddress;
code_ = new gpu::Memory(dev(), amd::alignUp(codeSize_, gpu::ConstBuffer::VectorSize));
+1 -1
Zobrazit soubor
@@ -12,7 +12,7 @@
#endif // AMD_PLATFORM_NAME
#ifndef AMD_PLATFORM_BUILD_NUMBER
# define AMD_PLATFORM_BUILD_NUMBER 1753
# define AMD_PLATFORM_BUILD_NUMBER 1752
#endif // AMD_PLATFORM_BUILD_NUMBER
#ifndef AMD_PLATFORM_REVISION_NUMBER