RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-21 20:12:56 +02:00
Code Issues Projects Releases Wiki Activity
9f798c877d
llvm-mirror/lib/Target/R600
History
Tom Stellard 9f798c877d R600: Set scheduling preference to Sched::Source 
R600 doesn't need to do any scheduling on the SelectionDAG now that it
has a very good MachineScheduler.  Also, using the VLIW SelectionDAG
scheduler was having a major impact on compile times. For example with
the phatk kernel here are the LLVM IR to machine code compile times:

With Sched::VLIW

Total Compile Time:                  1.4890 Seconds (User + System)
SelectionDAG Instruction Scheduling: 1.1670 Seconds (User + System)

With Sched::Source

Total Compile Time:                  0.3330 Seconds (User + System)
SelectionDAG Instruction Scheduling: 0.0070 Seconds (User + System)

The code ouput was identical with both schedulers.  This may not be true
for all programs, but it gives me confidence that there won't be much
reduction, if any, in code quality by using Sched::Source.

llvm-svn: 188215
2013-08-12 22:33:21 +00:00
..
InstPrinter R600: Bank Swizzle now display SCL equivalent 2013-06-29 19:32:29 +00:00
MCTargetDesc Remove address spaces from MC. 2013-07-02 15:49:13 +00:00
TargetInfo R600: Remove unnecessary include 2013-06-07 20:28:43 +00:00
AMDGPU.h R600/SI: Use VSrc_* register classes as the default classes for types 2013-08-06 23:08:28 +00:00
AMDGPU.td R600: Rework subtarget info and remove AMDILDevice classes 2013-06-07 20:37:48 +00:00
AMDGPUAsmPrinter.cpp R600/SI: Initial local memory support 2013-07-10 16:37:07 +00:00
AMDGPUAsmPrinter.h R600: Emit used GPRs count 2013-04-17 15:17:25 +00:00
AMDGPUCallingConv.td R600: Add 64-bit float load/store support 2013-08-01 15:23:42 +00:00
AMDGPUConvertToISA.cpp
AMDGPUFrameLowering.cpp R600: Fix calculation of stack offset in AMDGPUFrameLowering 2013-06-07 20:52:05 +00:00
AMDGPUFrameLowering.h
AMDGPUIndirectAddressing.cpp Don't cache the instruction and register info from the TargetMachine, because 2013-06-07 20:28:55 +00:00
AMDGPUInstrInfo.cpp R600: Use new getNamedOperandIdx function generated by TableGen 2013-06-25 21:22:18 +00:00
AMDGPUInstrInfo.h R600: Use new getNamedOperandIdx function generated by TableGen 2013-06-25 21:22:18 +00:00
AMDGPUInstrInfo.td Create an FPOW SDNode opcode def in the target independent .td file rather than in a specific backend. 2013-05-22 06:36:09 +00:00
AMDGPUInstructions.td R600: Add support for 24-bit MUL instructions 2013-07-23 01:48:42 +00:00
AMDGPUIntrinsics.td R600: Add support for GROUP_BARRIER instruction 2013-06-28 15:46:59 +00:00
AMDGPUISelDAGToDAG.cpp R600/SI: Use VSrc_* register classes as the default classes for types 2013-08-06 23:08:28 +00:00
AMDGPUISelLowering.cpp R600: Implement TargetLowering::getVectorIdxTy() 2013-08-05 22:22:07 +00:00
AMDGPUISelLowering.h R600: Implement TargetLowering::getVectorIdxTy() 2013-08-05 22:22:07 +00:00
AMDGPUMachineFunction.cpp Move string pointer from being a static class member to just a static global in the one file its needed in. 2013-07-17 00:31:35 +00:00
AMDGPUMachineFunction.h Move string pointer from being a static class member to just a static global in the one file its needed in. 2013-07-17 00:31:35 +00:00
AMDGPUMCInstLower.cpp
AMDGPUMCInstLower.h
AMDGPURegisterInfo.cpp Don't cache the instruction and register info from the TargetMachine, because 2013-06-07 20:28:55 +00:00
AMDGPURegisterInfo.h Don't cache the instruction and register info from the TargetMachine, because 2013-06-07 20:28:55 +00:00
AMDGPURegisterInfo.td Make SubRegIndex size mandatory, following r183020. 2013-05-31 23:45:26 +00:00
AMDGPUSubtarget.cpp R600: Rework subtarget info and remove AMDILDevice classes 2013-06-07 20:37:48 +00:00
AMDGPUSubtarget.h R600: Rework subtarget info and remove AMDILDevice classes 2013-06-07 20:37:48 +00:00
AMDGPUTargetMachine.cpp R600/SI: Use VSrc_* register classes as the default classes for types 2013-08-06 23:08:28 +00:00
AMDGPUTargetMachine.h SimplifyCFG: Use parallel-and and parallel-or mode to consolidate branch conditions 2013-07-27 00:01:07 +00:00
AMDGPUTargetTransformInfo.cpp SimplifyCFG: Use parallel-and and parallel-or mode to consolidate branch conditions 2013-07-27 00:01:07 +00:00
AMDILBase.td R600: Move Subtarget feature definitions into AMDGPU.td 2013-06-07 20:28:49 +00:00
AMDILCFGStructurizer.cpp R600: Remove predicated_break inst 2013-07-31 19:31:14 +00:00
AMDILInstrInfo.td R600: Rework subtarget info and remove AMDILDevice classes 2013-06-07 20:37:48 +00:00
AMDILIntrinsicInfo.cpp R600: Rework subtarget info and remove AMDILDevice classes 2013-06-07 20:37:48 +00:00
AMDILIntrinsicInfo.h
AMDILIntrinsics.td R600/SI: Use MULADD_IEEE/V_MAD_F32 instruction for mad pattern 2013-02-18 14:11:28 +00:00
AMDILISelLowering.cpp Make some arrays 'static const' 2013-07-15 06:39:13 +00:00
AMDILRegisterInfo.td
CMakeLists.txt R600: Add new file from r187831 to CMakeLists.txt 2013-08-06 23:12:34 +00:00
LLVMBuild.txt
Makefile
Processors.td Add a newline. 2013-07-01 21:31:10 +00:00
R600ControlFlowFinalizer.cpp R600: Add 64-bit float load/store support 2013-08-01 15:23:42 +00:00
R600Defines.h R600: Add local memory support via LDS 2013-06-28 15:47:08 +00:00
R600EmitClauseMarkers.cpp Use SmallVectorImpl& instead of SmallVector to avoid repeating small vector size. 2013-07-14 04:42:23 +00:00
R600ExpandSpecialInstrs.cpp R600: Remove predicated_break inst 2013-07-31 19:31:14 +00:00
R600InstrFormats.td Revert "R600: Use SchedModel enum for is{Trans,Vector}Only functions" 2013-07-31 20:43:03 +00:00
R600InstrInfo.cpp R600: Add 64-bit float load/store support 2013-08-01 15:23:42 +00:00
R600InstrInfo.h Revert "R600: Use SchedModel enum for is{Trans,Vector}Only functions" 2013-07-31 20:43:03 +00:00
R600Instructions.td R600: Add 64-bit float load/store support 2013-08-01 15:23:42 +00:00
R600Intrinsics.td R600: Improve texture handling 2013-05-17 16:50:20 +00:00
R600ISelLowering.cpp R600: Set scheduling preference to Sched::Source 2013-08-12 22:33:21 +00:00
R600ISelLowering.h R600: Use DAG lowering pass to handle fcos/fsin 2013-07-09 15:03:11 +00:00
R600MachineFunctionInfo.cpp R600/SI: Share code recording ShaderTypeAttribute between generations 2013-04-01 21:47:53 +00:00
R600MachineFunctionInfo.h Move passes from namespace llvm into anonymous namespaces. Sort includes while there. 2013-05-23 17:10:37 +00:00
R600MachineScheduler.cpp Revert "R600: Non vector only instruction can be scheduled on trans unit" 2013-07-31 20:43:27 +00:00
R600MachineScheduler.h Revert "R600: Non vector only instruction can be scheduled on trans unit" 2013-07-31 20:43:27 +00:00
R600OptimizeVectorRegisters.cpp R600: Do not mergevector after a vector reg is used 2013-07-31 19:32:12 +00:00
R600Packetizer.cpp Revert "R600: Non vector only instruction can be scheduled on trans unit" 2013-07-31 20:43:27 +00:00
R600RegisterInfo.cpp Don't cache the instruction and register info from the TargetMachine, because 2013-06-07 20:28:55 +00:00
R600RegisterInfo.h Don't cache the instruction and register info from the TargetMachine, because 2013-06-07 20:28:55 +00:00
R600RegisterInfo.td R600: Add 64-bit float load/store support 2013-08-01 15:23:42 +00:00
R600Schedule.td R600: Add local memory support via LDS 2013-06-28 15:47:08 +00:00
R600TextureIntrinsicsReplacer.cpp Move passes from namespace llvm into anonymous namespaces. Sort includes while there. 2013-05-23 17:10:37 +00:00
SIAnnotateControlFlow.cpp Add 'const' qualifiers to static const char* variables. 2013-07-16 01:17:10 +00:00
SIDefines.h R600/SI: Initial local memory support 2013-07-10 16:37:07 +00:00
SIFixSGPRCopies.cpp R600/SI: Use VSrc_* register classes as the default classes for types 2013-08-06 23:08:28 +00:00
SIInsertWaits.cpp Initialize SIInsertWaits::ExpInstrTypesSeen in the pass constructor. 2013-08-07 07:47:41 +00:00
SIInstrFormats.td R600/SI: Initial support for LDS/GDS instructions 2013-07-10 16:36:43 +00:00
SIInstrInfo.cpp Make some arrays 'static const' 2013-07-15 06:39:13 +00:00
SIInstrInfo.h R600/SI: adjust writemask to only the used components 2013-04-10 08:39:08 +00:00
SIInstrInfo.td R600/SI: Implement sint<->fp64 conversions 2013-08-08 16:06:08 +00:00
SIInstructions.td R600/SI: Add FMA pattern 2013-08-10 10:38:47 +00:00
SIIntrinsics.td R600/SI: Add intrinsic for retrieving the current thread ID 2013-07-10 16:36:52 +00:00
SIISelLowering.cpp R600/SI: FMA is faster than fmul and fadd for f64 2013-08-10 10:38:54 +00:00
SIISelLowering.h R600/SI: FMA is faster than fmul and fadd for f64 2013-08-10 10:38:54 +00:00
SILowerControlFlow.cpp R600/SI: Initial support for LDS/GDS instructions 2013-07-10 16:36:43 +00:00
SIMachineFunctionInfo.cpp R600/SI: Share code recording ShaderTypeAttribute between generations 2013-04-01 21:47:53 +00:00
SIMachineFunctionInfo.h R600/SI: Share code recording ShaderTypeAttribute between generations 2013-04-01 21:47:53 +00:00
SIRegisterInfo.cpp R600/SI: Add more special cases for opcodes to ensureSRegLimit() 2013-08-06 23:08:18 +00:00
SIRegisterInfo.h R600/SI: Add more special cases for opcodes to ensureSRegLimit() 2013-08-06 23:08:18 +00:00
SIRegisterInfo.td R600/SI: Add support for v2f32 loads 2013-07-18 21:43:48 +00:00
SISchedule.td