replica
/
hanchenye-llvm-project
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[AMDGPU] Distinguish between S/VGPR allocation and encoding granularities 

Differential Revision: https://reviews.llvm.org/D29633

llvm-svn: 294441
This commit is contained in:
Konstantin Zhuravlyov 2017-02-08 13:18:40 +00:00
parent e03b1d7b6a
commit e22fbcb264
2 changed files with 16 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
View File

@ -546,13 +546,13 @@ void AMDGPUAsmPrinter::getSIProgramInfo(SIProgramInfo &ProgInfo,
// SGPRBlocks is actual number of SGPR blocks minus 1.
ProgInfo.SGPRBlocks = alignTo(ProgInfo.NumSGPRsForWavesPerEU,
STM.getSGPRAllocGranule());
ProgInfo.SGPRBlocks = ProgInfo.SGPRBlocks / STM.getSGPRAllocGranule() - 1;
STM.getSGPREncodingGranule());
ProgInfo.SGPRBlocks = ProgInfo.SGPRBlocks / STM.getSGPREncodingGranule() - 1;
// VGPRBlocks is actual number of VGPR blocks minus 1.
ProgInfo.VGPRBlocks = alignTo(ProgInfo.NumVGPRsForWavesPerEU,
STM.getVGPRAllocGranule());
ProgInfo.VGPRBlocks = ProgInfo.VGPRBlocks / STM.getVGPRAllocGranule() - 1;
STM.getVGPREncodingGranule());
ProgInfo.VGPRBlocks = ProgInfo.VGPRBlocks / STM.getVGPREncodingGranule() - 1;
// Record first reserved register and reserved register count fields, and
// update max register counts if "amdgpu-debugger-reserve-regs" attribute was

12
llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
View File

@ -644,6 +644,13 @@ public:
/// \returns SGPR allocation granularity supported by the subtarget.
unsigned getSGPRAllocGranule() const {
if (getGeneration() >= AMDGPUSubtarget::VOLCANIC_ISLANDS)
return 16;
return 8;
}
/// \returns SGPR encoding granularity supported by the subtarget.
unsigned getSGPREncodingGranule() const {
return 8;
}
@ -689,6 +696,11 @@ public:
return 4;
}
/// \returns VGPR encoding granularity supported by the subtarget.
unsigned getVGPREncodingGranule() const {
return getVGPRAllocGranule();
}
/// \returns Total number of VGPRs supported by the subtarget.
unsigned getTotalNumVGPRs() const {
return 256;