Matt Arsenault fa711758df R600/SI: Implement enableAggressiveFMAFusion ...

Add tests for the various combines. This should
always be at least cycle neutral on all subtargets for f64,
and faster on some. For f32 we should prefer selecting
v_mad_f32 over v_fma_f32.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@227484 91177308-0d34-0410-b5e6-96231b3b80d8

2015-01-29 19:34:32 +00:00

..

AsmParser

R600/SI: Add a stub GCNTargetMachine

2015-01-06 18:00:21 +00:00

InstPrinter

[cleanup] Re-sort all the #include lines in LLVM using

2015-01-14 11:23:27 +00:00

MCTargetDesc

Add r224985 back with fixes.

2015-01-19 21:11:14 +00:00

TargetInfo

R600/SI: Add a stub GCNTargetMachine

2015-01-06 18:00:21 +00:00

AMDGPU.h

R600/SI: Use external symbols for scratch buffer

2015-01-20 17:49:47 +00:00

AMDGPU.td

R600/SI: Add subtarget feature for if f32 fma is fast

2015-01-29 19:34:25 +00:00

AMDGPUAlwaysInlinePass.cpp

Reapply: R600: Make sure to inline all internal functions

2014-11-03 19:49:05 +00:00

AMDGPUAsmPrinter.cpp

Compute the ELF SectionKind from the flags.

2015-01-29 17:33:21 +00:00

AMDGPUAsmPrinter.h

std::unique_ptrify the MCStreamer argument to createAsmPrinter

2015-01-18 20:29:04 +00:00

AMDGPUCallingConv.td

Remove the target machine from CCState. Previously it was only used

2014-08-06 18:45:26 +00:00

AMDGPUFrameLowering.cpp

Fix typo

2014-06-14 04:26:07 +00:00

AMDGPUFrameLowering.h

Canonicalize header guards into a common format.

2014-08-13 16:26:38 +00:00

AMDGPUInstrInfo.cpp

R600/SI: Don't shrink instructions whose e32 encoding doesn't exist

2015-01-15 18:42:51 +00:00

AMDGPUInstrInfo.h

R600/SI: Don't shrink instructions whose e32 encoding doesn't exist

2015-01-15 18:42:51 +00:00

AMDGPUInstrInfo.td

R600/SI: Add class intrinsic

2015-01-06 23:00:37 +00:00

AMDGPUInstructions.td

R600/SI: Add patterns for v_cvt_{flr|rpi}_i32_f32

2015-01-15 23:58:35 +00:00

AMDGPUIntrinsicInfo.cpp

Move R600 subtarget dependent variables onto the subtarget.

2014-07-25 22:22:39 +00:00

AMDGPUIntrinsicInfo.h

Canonicalize header guards into a common format.

2014-08-13 16:26:38 +00:00

AMDGPUIntrinsics.td

R600: Remove intrinsics that appear to be unused

2014-07-15 20:10:27 +00:00

AMDGPUISelDAGToDAG.cpp

R600/SI: Move i64 -> v2i32 load promotion into AMDGPUDAGToDAGISel::Select()

2015-01-23 22:05:45 +00:00

AMDGPUISelLowering.cpp

Move DataLayout back to the TargetMachine from TargetSubtargetInfo

2015-01-26 19:03:15 +00:00

AMDGPUISelLowering.h

R600: Try to use lower types for 64bit division if possible

2015-01-22 23:42:43 +00:00

AMDGPUMachineFunction.cpp

R600/SI: Add preliminary support for flat address space

2014-09-15 15:41:53 +00:00

AMDGPUMachineFunction.h

Reapply "R600: Add new intrinsic to read work dimensions"

2014-10-14 20:05:26 +00:00

AMDGPUMCInstLower.cpp

R600/SI: Use external symbols for scratch buffer

2015-01-20 17:49:47 +00:00

AMDGPUMCInstLower.h

R600/SI: Don't shrink instructions whose e32 encoding doesn't exist

2015-01-15 18:42:51 +00:00

AMDGPUPromoteAlloca.cpp

R600: Don't promote allocas when one of the users is a ptrtoint instruction

2014-10-31 20:52:04 +00:00

AMDGPURegisterInfo.cpp

R600/SI: Enable inline assembly

2014-12-03 04:08:00 +00:00

AMDGPURegisterInfo.h

Add override to overriden virtual methods, remove virtual keywords.

2014-09-03 11:41:21 +00:00

AMDGPURegisterInfo.td

…

AMDGPUSubtarget.cpp

R600/SI: Add subtarget feature for if f32 fma is fast

2015-01-29 19:34:25 +00:00

AMDGPUSubtarget.h

R600/SI: Add subtarget feature for if f32 fma is fast

2015-01-29 19:34:25 +00:00

AMDGPUTargetMachine.cpp

R600: Move DataLayout to AMDGPUTargetMachine

2015-01-28 16:04:26 +00:00

AMDGPUTargetMachine.h

R600: Move DataLayout to AMDGPUTargetMachine

2015-01-28 16:04:26 +00:00

AMDGPUTargetTransformInfo.cpp

Fix broken doxygen annotations, NFC

2014-11-12 18:25:06 +00:00

AMDILCFGStructurizer.cpp

Fix typos:

2014-08-11 18:04:46 +00:00

AMDKernelCodeT.h

R600/SI: Emit amd_kernel_code_t header for AMDGPU environment

2014-12-02 22:00:07 +00:00

CaymanInstructions.td

R600: Match rcp node on pre-SI

2014-07-24 06:59:24 +00:00

CIInstructions.td

R600/SI: Add VI instructions

2014-12-07 12:18:57 +00:00

CMakeLists.txt

R600/SI: Spill VGPRs to scratch space for compute shaders

2015-01-14 15:42:31 +00:00

EvergreenInstructions.td

R600/SI: Custom lower fround

2015-01-21 18:18:25 +00:00

LLVMBuild.txt

R600/SI: Start implementing an assembler

2014-11-14 14:08:00 +00:00

Makefile

R600/SI: Start implementing an assembler

2014-11-14 14:08:00 +00:00

Processors.td

R600/SI: Add subtarget feature for if f32 fma is fast

2015-01-29 19:34:25 +00:00

R600ClauseMergePass.cpp

Have MachineFunction cache a pointer to the subtarget to make lookups

2014-08-05 02:39:49 +00:00

R600ControlFlowFinalizer.cpp

Remove unnecessary copying or replace it with moves in a bunch of places.

2014-10-04 16:55:56 +00:00

R600Defines.h

Canonicalize header guards into a common format.

2014-08-13 16:26:38 +00:00

R600EmitClauseMarkers.cpp

Have MachineFunction cache a pointer to the subtarget to make lookups

2014-08-05 02:39:49 +00:00

R600ExpandSpecialInstrs.cpp

Have MachineFunction cache a pointer to the subtarget to make lookups

2014-08-05 02:39:49 +00:00

R600InstrFormats.td

R600/SI: Start implementing an assembler

2014-11-14 14:08:00 +00:00

R600InstrInfo.cpp

Remove unused argument to CreateTargetScheduleState and change

2014-10-09 01:59:35 +00:00

R600InstrInfo.h

Remove unused argument to CreateTargetScheduleState and change

2014-10-09 01:59:35 +00:00

R600Instructions.td

R600/SI: Custom lower fround

2015-01-21 18:18:25 +00:00

R600Intrinsics.td

…

R600ISelLowering.cpp

[SelectionDAG] Allow targets to specify legality of extloads' result

2015-01-08 00:51:32 +00:00

R600ISelLowering.h

Canonicalize header guards into a common format.

2014-08-13 16:26:38 +00:00

R600MachineFunctionInfo.cpp

…

R600MachineFunctionInfo.h

Canonicalize header guards into a common format.

2014-08-13 16:26:38 +00:00

R600MachineScheduler.cpp

Fix float division-by-zero in R600 scheduler.

2014-09-17 17:47:21 +00:00

R600MachineScheduler.h

Canonicalize header guards into a common format.

2014-08-13 16:26:38 +00:00

R600OptimizeVectorRegisters.cpp

Eliminate some deep std::vector copies. NFC.

2014-10-03 18:33:16 +00:00

R600Packetizer.cpp

Cleanup: Delete seemingly unused reference to MachineDominatorTree from ScheduleDAGInstrs.

2014-08-20 20:57:26 +00:00

R600RegisterInfo.cpp

R600: Remove AMDIL instruction and register definitions

2014-06-13 16:38:59 +00:00

R600RegisterInfo.h

Canonicalize header guards into a common format.

2014-08-13 16:26:38 +00:00

R600RegisterInfo.td

R600: Use LDS and vectors for private memory

2014-06-17 16:53:14 +00:00

R600Schedule.td

…

R600TextureIntrinsicsReplacer.cpp

[C++11] Add 'override' keywords and remove 'virtual'. Additionally add 'final' and leave 'virtual' on some methods that are marked virtual without overriding anything and have no obvious overrides themselves. R600 edition

2014-04-29 07:57:24 +00:00

R700Instructions.td

…

SIAnnotateControlFlow.cpp

R600/SI: SI Control Flow Annotation bug fixed

2014-06-20 17:06:02 +00:00

SIDefines.h

R600/SI: Add subtarget feature to enable VGPR spilling for all shader types

2015-01-20 19:33:04 +00:00

SIFixSGPRCopies.cpp

R600/SI: Remove VReg_32 register class

2015-01-07 20:59:25 +00:00

SIFixSGPRLiveRanges.cpp

R600/SI: Fix the FixSGPRLiveRanges pass

2014-09-24 01:33:24 +00:00

SIFoldOperands.cpp

R600/SI: Add pattern for bitcasting fp immediates to integers

2015-01-13 22:59:41 +00:00

SIInsertWaits.cpp

R600/SI: Insert s_waitcnt before s_barrier instructions.

2015-01-06 19:52:07 +00:00

SIInstrFormats.td

R600/SI: Add VI versions of LDS atomics

2015-01-27 17:25:07 +00:00

SIInstrInfo.cpp

R600/SI: Add subtarget feature to enable VGPR spilling for all shader types

2015-01-20 19:33:04 +00:00

SIInstrInfo.h

R600/SI: Don't shrink instructions whose e32 encoding doesn't exist

2015-01-15 18:42:51 +00:00

SIInstrInfo.td

R600/SI: Don't set patterns for chip-specific instructions while having pseudos

2015-01-27 17:25:11 +00:00

SIInstructions.td

R600/SI: Fix MIN3/MAX3 on VI, define MED3

2015-01-27 17:25:15 +00:00

SIIntrinsics.td

R600/SI: Use i32 vectors for resources and samplers

2014-07-11 17:11:52 +00:00

SIISelLowering.cpp

R600/SI: Implement enableAggressiveFMAFusion

2015-01-29 19:34:32 +00:00

SIISelLowering.h

R600/SI: Implement enableAggressiveFMAFusion

2015-01-29 19:34:32 +00:00

SILoadStoreOptimizer.cpp

R600/SI: Fix live range error hidden by SIFoldOperands

2014-12-03 05:22:29 +00:00

SILowerControlFlow.cpp

R600/SI: Add pattern for bitcasting fp immediates to integers

2015-01-13 22:59:41 +00:00

SILowerI1Copies.cpp

R600/SI: Remove VReg_32 register class

2015-01-07 20:59:25 +00:00

SIMachineFunctionInfo.cpp

R600/SI: Spill VGPRs to scratch space for compute shaders

2015-01-14 15:42:31 +00:00

SIMachineFunctionInfo.h

R600/SI: Add subtarget feature to enable VGPR spilling for all shader types

2015-01-20 19:33:04 +00:00

SIPrepareScratchRegs.cpp

R600/SI: Fix simple-loop.ll test

2015-01-20 19:33:02 +00:00

SIRegisterInfo.cpp

R600/SI: Remove stray debug statements

2015-01-29 16:55:28 +00:00

SIRegisterInfo.h

R600/SI: Define a schedule model and enable the generic machine scheduler

2015-01-29 16:55:25 +00:00

SIRegisterInfo.td

R600/SI: Use RegisterOperands to specify which operands can accept immediates

2015-01-12 19:33:18 +00:00

SISchedule.td

R600/SI: Define a schedule model

2015-01-14 01:13:19 +00:00

SIShrinkInstructions.cpp

R600/SI: Don't shrink instructions whose e32 encoding doesn't exist

2015-01-15 18:42:51 +00:00

SITypeRewriter.cpp

Revert "IR: MDNode => Value"

2014-11-11 21:30:22 +00:00

VIInstrFormats.td

R600/SI: Add VI instructions

2014-12-07 12:18:57 +00:00

VIInstructions.td

R600/SI: Add VI versions of MUBUF loads and stores

2015-01-27 17:24:58 +00:00