1
0
mirror of https://github.com/RPCS3/llvm.git synced 2024-12-23 12:40:17 +00:00
llvm/lib/CodeGen
Chris Lattner 892e182393 1. Make MCSection an abstract class.
2. Move section switch printing to MCSection virtual method which takes a
   TAI.  This eliminates textual formatting stuff from TLOF.
3. Eliminate SwitchToSectionDirective, getSectionFlagsAsString, and 
   TLOFELF::AtIsCommentChar.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@78510 91177308-0d34-0410-b5e6-96231b3b80d8
2009-08-08 22:41:53 +00:00
..
AsmPrinter 1. Make MCSection an abstract class. 2009-08-08 22:41:53 +00:00
PBQP Fix some -Asserts unused variable warnings. 2009-08-08 00:40:46 +00:00
SelectionDAG Don't build illegal ops in DAGCombiner::SimplifyBinOpWithSameOpcodeHands(). 2009-08-08 20:42:17 +00:00
BranchFolding.cpp Rewrite previous patch to follow Chris' stylistic 2009-08-07 17:41:29 +00:00
CMakeLists.txt Update CMake. 2009-08-07 03:59:06 +00:00
CodePlacementOpt.cpp
DeadMachineInstructionElim.cpp
Dump.cpp Re-apply LiveInterval index dumping patch, with fixes suggested by Bill 2009-08-03 21:55:09 +00:00
DwarfEHPrepare.cpp Move types back to the 2.5 API. 2009-07-29 22:17:13 +00:00
ELF.h ELF improvements: 2009-08-08 17:29:04 +00:00
ELFCodeEmitter.cpp - Remove custom handling of jumptables by the elf writter (this was 2009-08-05 06:57:03 +00:00
ELFCodeEmitter.h
ELFWriter.cpp ELF improvements: 2009-08-08 17:29:04 +00:00
ELFWriter.h ELF improvements: 2009-08-08 17:29:04 +00:00
GCMetadata.cpp Remove Value::getName{Start,End}, the last of the old Name APIs. 2009-07-26 09:48:23 +00:00
GCMetadataPrinter.cpp
GCStrategy.cpp
IfConversion.cpp
IntrinsicLowering.cpp Move a few more APIs back to 2.5 forms. The only remaining ones left to change back are 2009-07-31 20:28:14 +00:00
LatencyPriorityQueue.cpp
LazyLiveness.cpp
LiveInterval.cpp
LiveIntervalAnalysis.cpp Turn some insert_subreg, extract_subreg, subreg_to_reg into implicit_defs. 2009-08-05 03:53:14 +00:00
LiveStackAnalysis.cpp
LiveVariables.cpp Reapply r77654 with a fix: MachineFunctionPass's getAnalysisUsage 2009-07-31 18:16:33 +00:00
LLVMTargetMachine.cpp Disable stack coloring with register for now. It's not able to set kill markers. 2009-08-05 07:26:17 +00:00
LowerSubregs.cpp Remove RegisterScavenger::isSuperRegUsed(). This completely reverses the mistaken commit r77904. 2009-08-08 13:19:10 +00:00
MachineBasicBlock.cpp Re-apply LiveInterval index dumping patch, with fixes suggested by Bill 2009-08-03 21:55:09 +00:00
MachineDominators.cpp
MachineFunction.cpp Re-apply LiveInterval index dumping patch, with fixes suggested by Bill 2009-08-03 21:55:09 +00:00
MachineFunctionAnalysis.cpp Give MachineFunctionAnalysis a destructor so it can verify that 2009-08-01 04:19:43 +00:00
MachineFunctionPass.cpp Reapply r77654 with a fix: MachineFunctionPass's getAnalysisUsage 2009-07-31 18:16:33 +00:00
MachineInstr.cpp Don't tamper with <undef> operands in MachineInstr::addRegisterKilled. 2009-08-04 20:09:25 +00:00
MachineLICM.cpp
MachineLoopInfo.cpp Reapply r77654 with a fix: MachineFunctionPass's getAnalysisUsage 2009-07-31 18:16:33 +00:00
MachineModuleInfo.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
MachinePassRegistry.cpp
MachineRegisterInfo.cpp
MachineSink.cpp Various comment fixes. 2009-08-05 01:19:01 +00:00
MachineVerifier.cpp Clean out per-function data after the machine code verifier is done with it. 2009-08-08 15:34:50 +00:00
MachO.h
MachOCodeEmitter.cpp
MachOCodeEmitter.h
MachOWriter.cpp
MachOWriter.h
Makefile
ObjectCodeEmitter.cpp Remove accidental commited comment 2009-08-05 07:00:43 +00:00
OcamlGC.cpp
Passes.cpp
PHIElimination.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
PHIElimination.h
PostRASchedulerList.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
PreAllocSplitting.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
PrologEpilogInserter.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
PrologEpilogInserter.h
PseudoSourceValue.cpp
README.txt
RegAllocLinearScan.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
RegAllocLocal.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
RegAllocPBQP.cpp Fix some -Asserts unused variable warnings. 2009-08-08 00:40:46 +00:00
RegAllocSimple.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
RegisterCoalescer.cpp
RegisterScavenging.cpp Remove RegisterScavenger::isSuperRegUsed(). This completely reverses the mistaken commit r77904. 2009-08-08 13:19:10 +00:00
ScheduleDAG.cpp
ScheduleDAGEmit.cpp
ScheduleDAGInstrs.cpp Fix a typo in a comment. 2009-08-07 01:26:06 +00:00
ScheduleDAGInstrs.h
ScheduleDAGPrinter.cpp
ShadowStackGC.cpp Privatize the StructType table, which unfortunately involves routing contexts through a number of APIs. 2009-08-05 23:16:16 +00:00
ShrinkWrapping.cpp
SimpleRegisterCoalescing.cpp Another coalescer bug. When a dead copy is eliminated, transfer the kill to a def of the exact register rather than a super-register. 2009-08-07 07:14:14 +00:00
SimpleRegisterCoalescing.h
Spiller.cpp
Spiller.h
StackProtector.cpp
StackSlotColoring.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
StrongPHIElimination.cpp Use setPreservesAll and setPreservesCFG in CodeGen passes. 2009-07-31 23:37:33 +00:00
TargetInstrInfoImpl.cpp
TwoAddressInstructionPass.cpp Code clean up. 2009-08-07 00:28:58 +00:00
UnreachableBlockElim.cpp Make UnreachableMachineBlockElim preserve MachineDominatorTree and 2009-08-01 00:34:30 +00:00
VirtRegMap.cpp
VirtRegMap.h
VirtRegRewriter.cpp Fix a bunch of namespace pollution. 2009-08-07 01:32:21 +00:00
VirtRegRewriter.h

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

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 likelyhood the store may become dead.

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

I think we should have a "hasSideEffects" flag (which is automatically set for
stuff that "isLoad" "isCall" etc), and the remat pass should eventually be able
to remat any instruction that has no side effects, if it can handle it and if
profitable.

For now, I'd suggest having the remat stuff work like this:

1. I need to spill/reload this thing.
2. Check to see if it has side effects.
3. Check to see if it is simple enough: e.g. it only has one register
destination and no register input.
4. If so, clone the instruction, do the xform, etc.

Advantages of this are:

1. the .td file describes the behavior of the instructions, not the way the
   algorithm should work.
2. as remat gets smarter in the future, we shouldn't have to be changing the .td
   files.
3. it is easier to explain what the flag means in the .td file, because you
   don't have to pull in the explanation of how the current remat algo works.

Some potential added complexities:

1. Some instructions have to be glued to it's predecessor or successor. All of
   the PC relative instructions and condition code setting instruction. We could
   mark them as hasSideEffects, but that's not quite right. PC relative loads
   from constantpools can be remat'ed, for example. But it requires more than
   just cloning the instruction. Some instructions can be remat'ed but it
   expands to more than one instruction. But allocator will have to make a
   decision.

4. As stated in 3, not as simple as cloning in some cases. The target will have
   to decide how to remat it. For example, an ARM 2-piece constant generation
   instruction is remat'ed as a load from constantpool.

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

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 improvments:

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