LLVM: lib/Target/AMDGPU/AMDGPUCallLowering.cpp File Reference (original) (raw)
This file implements the lowering of LLVM calls to machine code calls for GlobalISel. More...
Go to the source code of this file.
Functions | |
---|---|
static ISD::NodeType | extOpcodeToISDExtOpcode (unsigned MIOpc) |
static void | allocateHSAUserSGPRs (CCState &CCInfo, MachineIRBuilder &B, MachineFunction &MF, const SIRegisterInfo &TRI, SIMachineFunctionInfo &Info) |
static std::pair< CCAssignFn *, CCAssignFn * > | getAssignFnsForCC (CallingConv::ID CC, const SITargetLowering &TLI) |
Returns a pair containing the fixed CCAssignFn and the vararg CCAssignFn for CC. | |
static unsigned | getCallOpcode (const MachineFunction &CallerF, bool IsIndirect, bool IsTailCall, bool isWave32, CallingConv::ID CC) |
static bool | addCallTargetOperands (MachineInstrBuilder &CallInst, MachineIRBuilder &MIRBuilder, AMDGPUCallLowering::CallLoweringInfo &Info) |
static bool | canGuaranteeTCO (CallingConv::ID CC) |
Return true if the calling convention is one that we can guarantee TCO for. | |
static bool | mayTailCallThisCC (CallingConv::ID CC) |
Return true if we might ever do TCO for calls with this calling convention. | |
This file implements the lowering of LLVM calls to machine code calls for GlobalISel.
Definition in file AMDGPUCallLowering.cpp.
◆ DEBUG_TYPE
#define DEBUG_TYPE "amdgpu-call-lowering"
◆ addCallTargetOperands()
◆ allocateHSAUserSGPRs()
Definition at line 448 of file AMDGPUCallLowering.cpp.
References llvm::MachineFunction::addLiveIn(), llvm::CCState::AllocateReg(), B, llvm::AMDGPUAS::CONSTANT_ADDRESS, llvm::MachineFunction::getRegInfo(), llvm::GCNUserSGPRUsageInfo::hasDispatchID(), llvm::GCNUserSGPRUsageInfo::hasDispatchPtr(), llvm::GCNUserSGPRUsageInfo::hasFlatScratchInit(), llvm::GCNUserSGPRUsageInfo::hasKernargSegmentPtr(), llvm::GCNUserSGPRUsageInfo::hasPrivateSegmentBuffer(), llvm::GCNUserSGPRUsageInfo::hasQueuePtr(), Info, MRI, llvm::LLT::pointer(), and TRI.
Referenced by llvm::AMDGPUCallLowering::lowerFormalArgumentsKernel().