llvm-6502/lib/CodeGen
Bill Schmidt 5ff776bfde This patch addresses bug 15031.
The common code in the post-RA scheduler to break anti-dependencies on the
critical path contained a flaw.  In the reported case, an anti-dependency
between the overlapping registers %X4 and %R4 exists:

	%X29<def> = OR8 %X4, %X4
	%R4<def>, %X3<def,dead,tied3> = LBZU 1, %X3<kill,tied1>

The unpatched code breaks the dependency by replacing %R4 and its uses
with %R3, the first register on the available list.  However, %R3 and
%X3 overlap, so this creates two overlapping definitions on the same
instruction.

The fix is straightforward, preventing selection of a register that
overlaps any other defined register on the same instruction.

The test case is reduced from the bug report, and verifies that we no
longer produce "lbzu 3, 1(3)" when breaking this anti-dependency.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@173706 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-28 18:36:58 +00:00
..
AsmPrinter Fix comment. 2013-01-28 17:33:26 +00:00
SelectionDAG Legalizer: Reword comment again, per Duncan's suggestion. 2013-01-27 21:02:52 +00:00
AggressiveAntiDepBreaker.cpp Remove duplicate includes. 2012-12-21 17:06:44 +00:00
AggressiveAntiDepBreaker.h
AllocationOrder.cpp
AllocationOrder.h Limit the search space in RAGreedy::tryEvict(). 2013-01-12 00:57:44 +00:00
Analysis.cpp Remove unused parameter. Also use the AttributeSet query methods instead of the Attribute query methods. 2013-01-18 21:50:24 +00:00
AntiDepBreaker.h
BasicTargetTransformInfo.cpp Split TargetLowering into a CodeGen and a SelectionDAG part. 2013-01-11 20:05:37 +00:00
BranchFolding.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
BranchFolding.h
CalcSpillWeights.cpp
CallingConvLower.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
CMakeLists.txt Split TargetLowering into a CodeGen and a SelectionDAG part. 2013-01-11 20:05:37 +00:00
CodeGen.cpp Switch TargetTransformInfo from an immutable analysis pass that requires 2013-01-07 01:37:14 +00:00
CodePlacementOpt.cpp Remove the Function::getFnAttributes method in favor of using the AttributeSet 2012-12-30 10:32:01 +00:00
CriticalAntiDepBreaker.cpp This patch addresses bug 15031. 2013-01-28 18:36:58 +00:00
CriticalAntiDepBreaker.h This patch addresses bug 15031. 2013-01-28 18:36:58 +00:00
DeadMachineInstructionElim.cpp
DFAPacketizer.cpp
DwarfEHPrepare.cpp Split TargetLowering into a CodeGen and a SelectionDAG part. 2013-01-11 20:05:37 +00:00
EarlyIfConversion.cpp Move MachineTraceMetrics.h into include/llvm/CodeGen. 2013-01-17 01:06:04 +00:00
EdgeBundles.cpp
ExecutionDepsFix.cpp
ExpandISelPseudos.cpp
ExpandPostRAPseudos.cpp
GCMetadata.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
GCMetadataPrinter.cpp
GCStrategy.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
IfConversion.cpp Avoid creating duplicate CFG edges in the IfConversion pass. 2013-01-24 23:59:08 +00:00
InlineSpiller.cpp
InterferenceCache.cpp
InterferenceCache.h
IntrinsicLowering.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
JITCodeEmitter.cpp
LatencyPriorityQueue.cpp
LexicalScopes.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
LiveDebugVariables.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
LiveDebugVariables.h
LiveInterval.cpp
LiveIntervalAnalysis.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
LiveIntervalUnion.cpp
LiveRangeCalc.cpp
LiveRangeCalc.h
LiveRangeEdit.cpp
LiveRegMatrix.cpp
LiveStackAnalysis.cpp
LiveVariables.cpp
LLVMBuild.txt Extracted ObjCARC.cpp into its own library libLLVMObjCARCOpts in preparation for refactoring the ARC Optimizer. 2013-01-28 01:35:51 +00:00
LLVMTargetMachine.cpp Switch TargetTransformInfo from an immutable analysis pass that requires 2013-01-07 01:37:14 +00:00
LocalStackSlotAllocation.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
MachineBasicBlock.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
MachineBlockFrequencyInfo.cpp
MachineBlockPlacement.cpp Split TargetLowering into a CodeGen and a SelectionDAG part. 2013-01-11 20:05:37 +00:00
MachineBranchProbabilityInfo.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
MachineCodeEmitter.cpp
MachineCopyPropagation.cpp
MachineCSE.cpp
MachineDominators.cpp
MachineFunction.cpp Stack Alignment: throw error if we can't satisfy the minimal alignment 2013-01-10 01:10:10 +00:00
MachineFunctionAnalysis.cpp
MachineFunctionPass.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
MachineFunctionPrinterPass.cpp
MachineInstr.cpp ScheduleDAG: colorize the DOT graph and improve formatting. 2013-01-25 07:45:25 +00:00
MachineInstrBundle.cpp Move an assertion so it doesn't dereference end(). 2013-01-04 22:17:31 +00:00
MachineLICM.cpp Split TargetLowering into a CodeGen and a SelectionDAG part. 2013-01-11 20:05:37 +00:00
MachineLoopInfo.cpp
MachineModuleInfo.cpp small fixes to enable the reuse of the pass manager across multiple modules 2013-01-04 18:04:42 +00:00
MachineModuleInfoImpls.cpp
MachinePassRegistry.cpp
MachinePostDominators.cpp
MachineRegisterInfo.cpp Don't call destructors on MachineInstr and MachineOperand. 2013-01-05 05:05:51 +00:00
MachineScheduler.cpp Use const reference instead of vector copying. 2013-01-25 21:44:27 +00:00
MachineSink.cpp
MachineSSAUpdater.cpp Use MachineInstrBuilder in a few CodeGen passes. 2012-12-20 18:08:06 +00:00
MachineTraceMetrics.cpp Move MachineTraceMetrics.h into include/llvm/CodeGen. 2013-01-17 01:06:04 +00:00
MachineVerifier.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
Makefile
OcamlGC.cpp
OptimizePHIs.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
Passes.cpp Provide a place for targets to insert ILP optimization passes. 2013-01-17 00:58:38 +00:00
PeepholeOptimizer.cpp
PHIElimination.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
PHIEliminationUtils.cpp
PHIEliminationUtils.h
PostRASchedulerList.cpp Use MachineInstrBuilder in a few CodeGen passes. 2012-12-20 18:08:06 +00:00
ProcessImplicitDefs.cpp
PrologEpilogInserter.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
PrologEpilogInserter.h
PseudoSourceValue.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
README.txt
RegAllocBase.cpp
RegAllocBase.h
RegAllocBasic.cpp
RegAllocFast.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
RegAllocGreedy.cpp Limit the search space in RAGreedy::tryEvict(). 2013-01-12 00:57:44 +00:00
RegAllocPBQP.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
RegisterClassInfo.cpp Precompute some information about register costs. 2013-01-12 00:54:59 +00:00
RegisterCoalescer.cpp Fix PR14732 by handling all kinds of IMPLICIT_DEF live ranges. 2013-01-03 00:47:51 +00:00
RegisterCoalescer.h
RegisterPressure.cpp
RegisterScavenging.cpp
ScheduleDAG.cpp MIsched: Added biasCriticalPath. 2013-01-24 02:09:55 +00:00
ScheduleDAGInstrs.cpp ScheduleDAG: colorize the DOT graph and improve formatting. 2013-01-25 07:45:25 +00:00
ScheduleDAGPrinter.cpp ScheduleDAG: colorize the DOT graph and improve formatting. 2013-01-25 07:45:25 +00:00
ScoreboardHazardRecognizer.cpp
ShadowStackGC.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
ShrinkWrapping.cpp Remove duplicate includes. 2012-12-21 17:06:44 +00:00
SjLjEHPrepare.cpp Split TargetLowering into a CodeGen and a SelectionDAG part. 2013-01-11 20:05:37 +00:00
SlotIndexes.cpp
Spiller.cpp
Spiller.h
SpillPlacement.cpp
SpillPlacement.h
SplitKit.cpp
SplitKit.h
StackColoring.cpp Move MachineTraceMetrics.h into include/llvm/CodeGen. 2013-01-17 01:06:04 +00:00
StackProtector.cpp Add the heuristic to differentiate SSPStrong from SSPRequired. 2013-01-23 06:43:53 +00:00
StackSlotColoring.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
StrongPHIElimination.cpp
TailDuplication.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
TargetFrameLoweringImpl.cpp
TargetInstrInfo.cpp
TargetLoweringBase.cpp Split TargetLowering into a CodeGen and a SelectionDAG part. 2013-01-11 20:05:37 +00:00
TargetLoweringObjectFileImpl.cpp [MC/Mach-O] Implement integrated assembler support for linker options. 2013-01-18 19:37:00 +00:00
TargetOptionsImpl.cpp
TargetRegisterInfo.cpp
TargetSchedule.cpp
TwoAddressInstructionPass.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
UnreachableBlockElim.cpp Move all of the header files which are involved in modelling the LLVM IR 2013-01-02 11:36:10 +00:00
VirtRegMap.cpp

//===---------------------------------------------------------------------===//

Common register allocation / spilling problem:

        mul lr, r4, lr
        str lr, [sp, #+52]
        ldr lr, [r1, #+32]
        sxth r3, r3
        ldr r4, [sp, #+52]
        mla r4, r3, lr, r4

can be:

        mul lr, r4, lr
        mov r4, lr
        str lr, [sp, #+52]
        ldr lr, [r1, #+32]
        sxth r3, r3
        mla r4, r3, lr, r4

and then "merge" mul and mov:

        mul r4, r4, lr
        str lr, [sp, #+52]
        ldr lr, [r1, #+32]
        sxth r3, r3
        mla r4, r3, lr, r4

It also increase the likelihood the store may become dead.

//===---------------------------------------------------------------------===//

bb27 ...
        ...
        %reg1037 = ADDri %reg1039, 1
        %reg1038 = ADDrs %reg1032, %reg1039, %NOREG, 10
    Successors according to CFG: 0x8b03bf0 (#5)

bb76 (0x8b03bf0, LLVM BB @0x8b032d0, ID#5):
    Predecessors according to CFG: 0x8b0c5f0 (#3) 0x8b0a7c0 (#4)
        %reg1039 = PHI %reg1070, mbb<bb76.outer,0x8b0c5f0>, %reg1037, mbb<bb27,0x8b0a7c0>

Note ADDri is not a two-address instruction. However, its result %reg1037 is an
operand of the PHI node in bb76 and its operand %reg1039 is the result of the
PHI node. We should treat it as a two-address code and make sure the ADDri is
scheduled after any node that reads %reg1039.

//===---------------------------------------------------------------------===//

Use local info (i.e. register scavenger) to assign it a free register to allow
reuse:
        ldr r3, [sp, #+4]
        add r3, r3, #3
        ldr r2, [sp, #+8]
        add r2, r2, #2
        ldr r1, [sp, #+4]  <==
        add r1, r1, #1
        ldr r0, [sp, #+4]
        add r0, r0, #2

//===---------------------------------------------------------------------===//

LLVM aggressively lift CSE out of loop. Sometimes this can be negative side-
effects:

R1 = X + 4
R2 = X + 7
R3 = X + 15

loop:
load [i + R1]
...
load [i + R2]
...
load [i + R3]

Suppose there is high register pressure, R1, R2, R3, can be spilled. We need
to implement proper re-materialization to handle this:

R1 = X + 4
R2 = X + 7
R3 = X + 15

loop:
R1 = X + 4  @ re-materialized
load [i + R1]
...
R2 = X + 7 @ re-materialized
load [i + R2]
...
R3 = X + 15 @ re-materialized
load [i + R3]

Furthermore, with re-association, we can enable sharing:

R1 = X + 4
R2 = X + 7
R3 = X + 15

loop:
T = i + X
load [T + 4]
...
load [T + 7]
...
load [T + 15]
//===---------------------------------------------------------------------===//

It's not always a good idea to choose rematerialization over spilling. If all
the load / store instructions would be folded then spilling is cheaper because
it won't require new live intervals / registers. See 2003-05-31-LongShifts for
an example.

//===---------------------------------------------------------------------===//

With a copying garbage collector, derived pointers must not be retained across
collector safe points; the collector could move the objects and invalidate the
derived pointer. This is bad enough in the first place, but safe points can
crop up unpredictably. Consider:

        %array = load { i32, [0 x %obj] }** %array_addr
        %nth_el = getelementptr { i32, [0 x %obj] }* %array, i32 0, i32 %n
        %old = load %obj** %nth_el
        %z = div i64 %x, %y
        store %obj* %new, %obj** %nth_el

If the i64 division is lowered to a libcall, then a safe point will (must)
appear for the call site. If a collection occurs, %array and %nth_el no longer
point into the correct object.

The fix for this is to copy address calculations so that dependent pointers
are never live across safe point boundaries. But the loads cannot be copied
like this if there was an intervening store, so may be hard to get right.

Only a concurrent mutator can trigger a collection at the libcall safe point.
So single-threaded programs do not have this requirement, even with a copying
collector. Still, LLVM optimizations would probably undo a front-end's careful
work.

//===---------------------------------------------------------------------===//

The ocaml frametable structure supports liveness information. It would be good
to support it.

//===---------------------------------------------------------------------===//

The FIXME in ComputeCommonTailLength in BranchFolding.cpp needs to be
revisited. The check is there to work around a misuse of directives in inline
assembly.

//===---------------------------------------------------------------------===//

It would be good to detect collector/target compatibility instead of silently
doing the wrong thing.

//===---------------------------------------------------------------------===//

It would be really nice to be able to write patterns in .td files for copies,
which would eliminate a bunch of explicit predicates on them (e.g. no side 
effects).  Once this is in place, it would be even better to have tblgen 
synthesize the various copy insertion/inspection methods in TargetInstrInfo.

//===---------------------------------------------------------------------===//

Stack coloring improvements:

1. Do proper LiveStackAnalysis on all stack objects including those which are
   not spill slots.
2. Reorder objects to fill in gaps between objects.
   e.g. 4, 1, <gap>, 4, 1, 1, 1, <gap>, 4 => 4, 1, 1, 1, 1, 4, 4

//===---------------------------------------------------------------------===//

The scheduler should be able to sort nearby instructions by their address. For
example, in an expanded memset sequence it's not uncommon to see code like this:

  movl $0, 4(%rdi)
  movl $0, 8(%rdi)
  movl $0, 12(%rdi)
  movl $0, 0(%rdi)

Each of the stores is independent, and the scheduler is currently making an
arbitrary decision about the order.

//===---------------------------------------------------------------------===//

Another opportunitiy in this code is that the $0 could be moved to a register:

  movl $0, 4(%rdi)
  movl $0, 8(%rdi)
  movl $0, 12(%rdi)
  movl $0, 0(%rdi)

This would save substantial code size, especially for longer sequences like
this. It would be easy to have a rule telling isel to avoid matching MOV32mi
if the immediate has more than some fixed number of uses. It's more involved
to teach the register allocator how to do late folding to recover from
excessive register pressure.