|
LLVM 23.0.0git
|
Coexecution-focused scheduling strategy for AMDGPU. More...
Go to the source code of this file.
Classes | |
| class | llvm::HardwareUnitInfo |
| HardwareUnitInfo is a wrapper class which maps to some real hardware resource. More... | |
| class | llvm::CandidateHeuristics |
| CandidateHeuristics contains state and implementations to facilitate making per instruction scheduling decisions; it contains methods used in tryCandidate to decide which instruction to schedule next. More... | |
| class | llvm::AMDGPUCoExecSchedStrategy |
Namespaces | |
| namespace | llvm |
| This is an optimization pass for GlobalISel generic memory operations. | |
| namespace | llvm::AMDGPU |
| namespace | llvm::AMDGPU::FlavorGroups |
Typedefs | |
| using | llvm::AMDGPU::FlavorGroup = SmallVector<InstructionFlavor, 4> |
Enumerations | |
| enum class | llvm::AMDGPU::InstructionFlavor : uint8_t { llvm::AMDGPU::WMMA , llvm::AMDGPU::SingleCycleVALU , llvm::AMDGPU::TRANS , llvm::AMDGPU::MultiCycleVALU , llvm::AMDGPU::VMEM , llvm::AMDGPU::DS , llvm::AMDGPU::SALU , llvm::AMDGPU::DMA , llvm::AMDGPU::Fence , llvm::AMDGPU::Other , llvm::AMDGPU::NUM_FLAVORS } |
| enum class | llvm::AMDGPU::AMDGPUSchedReason : uint8_t { llvm::AMDGPU::None , llvm::AMDGPU::CritResourceBalance , llvm::AMDGPU::CritResourceDep , llvm::AMDGPU::NUM_REASONS } |
| AMDGPU-specific scheduling decision reasons. More... | |
Coexecution-focused scheduling strategy for AMDGPU.
Definition in file AMDGPUCoExecSchedStrategy.h.