Commit Graph

94 Commits

Author SHA1 Message Date
Dan Gohman
3cc6243ddf Change SUnit's dump method to take a ScheduleDAG* instead of
a SelectionDAG*.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59488 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18 02:06:40 +00:00
Dan Gohman
e45b6f10e2 There is no InstrStage class.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59480 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-18 01:05:04 +00:00
Dan Gohman
7e6a1bc0fa Add support for building a ScheduleDAG from MachineInstrs. This is currently
fairly conservative; it doesn't do alias-analysis queries and it doesn't
attempt to break anti-dependencies.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59324 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-14 21:47:58 +00:00
Dan Gohman
59d5f1d13b Add a version of NewSUnit for creating units with MachineInstrs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59282 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-14 00:41:36 +00:00
Dan Gohman
f449bf36ef Initial support for carrying MachineInstrs in SUnits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59278 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-14 00:06:09 +00:00
Dan Gohman
ae78a88e12 Clarify the comment about NumPreds and NumSuccs to reflect that they only
count non-control preds and succs.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59276 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-13 23:41:41 +00:00
Dan Gohman
d23e0f81bc Remove the FlaggedNodes member from SUnit. Instead of requiring each SUnit
to carry a SmallVector of flagged nodes, just calculate the flagged nodes
dynamically when they are needed.

The local-liveness change is due to a trivial scheduling change where
the scheduler arbitrary decision differently.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59273 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-13 23:24:17 +00:00
Dan Gohman
550f5afb68 Make the Node member of SUnit private, and add accessors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59264 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-13 21:36:12 +00:00
Dan Gohman
a23b3b803e Change ScheduleDAG's DAG member from a reference to a pointer, to prepare
for the possibility of scheduling without a SelectionDAG being present.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59263 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-13 21:21:28 +00:00
Dan Gohman
9b75b37375 Change the scheduler accessor methods to accept an explicit TargetMachine
argument instead of taking the SelectionDAG's TargetMachine. This is
needed for some upcoming scheduler changes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@59055 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-11 17:50:47 +00:00
Dale Johannesen
86b49f8e2d Next round of earlyclobber handling. Approach the
RA problem by expanding the live interval of an
earlyclobber def back one slot.  Remove
overlap-earlyclobber throughout.  Remove 
earlyclobber bits and their handling from
live internals.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56539 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-24 01:07:17 +00:00
Dan Gohman
ee2e403545 Add a new "fast" scheduler. This is currently basically just a
copy of the BURRList scheduler, but with several parts ripped
out, such as backtracking, online topological sort maintenance
(needed by backtracking), the priority queue, and Sethi-Ullman
number computation and maintenance (needed by the priority
queue). As a result of all this, it generates somewhat lower
quality code, but that's its tradeoff for running about 30%
faster than list-burr in -fast mode in many cases.

This is somewhat experimental. Moving forward, major pieces of
this can be refactored with pieces in common with
ScheduleDAGRRList.cpp.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56307 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-18 16:26:26 +00:00
Dale Johannesen
91aac1015e Add a bit to mark operands of asm's that conflict
with an earlyclobber operand elsewhere.  Propagate
this bit and the earlyclobber bit through SDISel.
Change linear-scan RA not to allocate regs in a way 
that conflicts with an earlyclobber.  See also comments.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56290 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-17 21:13:11 +00:00
Bill Wendling
056292fd73 Reverting r56249. On further investigation, this functionality isn't needed.
Apologies for the thrashing.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56251 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-16 21:48:12 +00:00
Bill Wendling
9468a9b6be - Change "ExternalSymbolSDNode" to "SymbolSDNode".
- Add linkage to SymbolSDNode (default to external).
- Change ISD::ExternalSymbol to ISD::Symbol.
- Change ISD::TargetExternalSymbol to ISD::TargetSymbol

These changes pave the way to allowing SymbolSDNodes with non-external linkage.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56249 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-16 21:12:30 +00:00
Dan Gohman
475871a144 Rename SDOperand to SDValue.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54128 91177308-0d34-0410-b5e6-96231b3b80d8
2008-07-27 21:46:04 +00:00
Dan Gohman
5e84368b26 Reapply 53476 and 53480, with a fix so that it properly updates
the BB member to the current basic block after emitting
instructions.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@53567 91177308-0d34-0410-b5e6-96231b3b80d8
2008-07-14 18:19:29 +00:00
Evan Cheng
4576f6d7a9 Do not use computationally expensive scheduling heuristics with -fast.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52971 91177308-0d34-0410-b5e6-96231b3b80d8
2008-07-01 18:05:03 +00:00
Dan Gohman
a0b50d7f0d Comment fixes, and make Schedule() pure virtual.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52658 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-24 00:02:44 +00:00
Dan Gohman
94d7a5f815 Remove ScheduleDAG's SUnitMap altogether. Instead, use SDNode's NodeId
field, which is otherwise unused after instruction selection, as an index
into the SUnit array.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52583 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-21 19:18:17 +00:00
Dan Gohman
3627e34486 Add a priority queue class, which is a wrapper around std::priority_queue
and provides fairly efficient removal of arbitrary elements. Switch
ScheduleDAGRRList from std::set to this new priority queue.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52582 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-21 18:35:25 +00:00
Dan Gohman
4c8c83022b Change ScheduleDAG's SUnitMap from DenseMap<SDNode*, vector<SUnit*> >
to DenseMap<SDNode*, SUnit*>, and adjust the way cloned SUnit nodes are
handled so that only the original node needs to be in the map.
This speeds up llc on 447.dealII.llvm.bc by about 2%.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52576 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-21 15:52:51 +00:00
Bill Wendling
13d57320bd Remove warnings about unused parameters and shadowed variables.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@51266 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-19 20:15:12 +00:00
Evan Cheng
34cd4a484e Fix more -Wshorten-64-to-32 warnings.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50659 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-05 18:30:58 +00:00
Roman Levenstein
a0201d5204 Use std::set instead of std::priority_queue for the RegReductionPriorityQueue.
This removes the existing bottleneck related to the removal of elements from 
the middle of the queue.

Also fixes a subtle bug in ScheduleDAGRRList::CapturePred:
It was updating the state of the SUnit before removing it. As a result, the
comparison operators were working incorrectly and this SUnit could not be removed 
from the queue properly.

Reviewed by Evan and Dan. Approved by Dan.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50412 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-29 09:07:59 +00:00
Roman Levenstein
9cac5259fe Ongoing work on improving the instruction selection infrastructure:
Rename SDOperandImpl back to SDOperand.
Introduce the SDUse class that represents a use of the SDNode referred by
an SDOperand. Now it is more similar to Use/Value classes.

Patch is approved by Dan Gohman.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49795 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-16 16:15:27 +00:00
Dan Gohman
80792f3dde Treat EntryToken nodes as "passive" so that they aren't added to the
ScheduleDAG; they don't correspond to any actual instructions so they
don't need to be scheduled.

This fixes a bug where the EntryToken was being scheduled multiple
times in some cases, though it ended up not causing any trouble because 
EntryToken doesn't expand into anything. With this fixed the schedulers
reliably schedule the expected number of units, so we can check this
with an assertion.

This requires a tweak to test/CodeGen/X86/loop-hoist.ll because it
ends up getting scheduled differently in a trivial way, though it was
enough to fool the prcontext+grep that the test does.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49701 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-15 01:22:18 +00:00
Dan Gohman
89bf0a6b05 In -view-sunit-dags, display "special" chain dependencies as cyan
instead of blue to distinguish them from regular dependencies.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49696 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-14 23:15:07 +00:00
Dan Gohman
b1e07ec6a4 Fix a typo in a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49502 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-10 22:27:06 +00:00
Dan Gohman
36b5c1338a Rename MemOperand to MachineMemOperand. This was suggested by
review feedback from Chris quite a while ago. No functionality
change.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49348 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-07 19:35:22 +00:00
Roman Levenstein
dc1adac582 Re-commit of the r48822, where the infinite looping problem discovered
by Dan Gohman is fixed.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49330 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-07 10:06:32 +00:00
Evan Cheng
8a50f1fcf0 Start of a series of patches related to implicit_def.
There is no point in creating a long live range defined by an implicit_def. Scheduler now duplicates implicit_def instruction for each of its uses. Therefore, if an implicit_def node has multiple uses, it will become a number of very short live ranges, rather than a long one. This will make coalescer's job easier.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49164 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-03 16:36:07 +00:00
Evan Cheng
6397c64441 Backing out 48222 temporarily.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49124 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-03 03:13:16 +00:00
Roman Levenstein
e326332acd Use a linked data structure for the uses lists of an SDNode, just like
LLVM Value/Use does and MachineRegisterInfo/MachineOperand does.
This allows constant time for all uses list maintenance operations.

The idea was suggested by Chris. Reviewed by Evan and Dan.
Patch is tested and approved by Dan.

On normal use-cases compilation speed is not affected. On very big basic
blocks there are compilation speedups in the range of 15-20% or even better. 



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@48822 91177308-0d34-0410-b5e6-96231b3b80d8
2008-03-26 12:39:26 +00:00
Evan Cheng
1090fc99cd Forgot this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@48349 91177308-0d34-0410-b5e6-96231b3b80d8
2008-03-14 00:17:29 +00:00
Evan Cheng
9e23336d0c Experimental scheduler change to schedule / coalesce the copies added for function livein's. Take 2008-03-10-RegAllocInfLoop.ll, the schedule looks like this after these copies are inserted:
entry: 0x12049d0, LLVM BB @0x1201fd0, ID#0:
Live Ins: %EAX %EDX %ECX
        %reg1031<def> = MOVPC32r 0
        %reg1032<def> = ADD32ri %reg1031, <es:_GLOBAL_OFFSET_TABLE_>, %EFLAGS<imp-def>
        %reg1028<def> = MOV32rr %EAX
        %reg1029<def> = MOV32rr %EDX
        %reg1030<def> = MOV32rr %ECX
        %reg1027<def> = MOV8rm %reg0, 1, %reg0, 0, Mem:LD(1,1) [0x1201910 + 0]
        %reg1025<def> = MOV32rr %reg1029
        %reg1026<def> = MOV32rr %reg1030
        %reg1024<def> = MOV32rr %reg1028

The copies unnecessarily increase register pressure and it will end up requiring a physical register to be spilled.

With -schedule-livein-copies:
entry: 0x12049d0, LLVM BB @0x1201fa0, ID#0:
Live Ins: %EAX %EDX %ECX
        %reg1031<def> = MOVPC32r 0
        %reg1032<def> = ADD32ri %reg1031, <es:_GLOBAL_OFFSET_TABLE_>, %EFLAGS<imp-def>
        %reg1024<def> = MOV32rr %EAX
        %reg1025<def> = MOV32rr %EDX
        %reg1026<def> = MOV32rr %ECX
        %reg1027<def> = MOV8rm %reg0, 1, %reg0, 0, Mem:LD(1,1) [0x12018e0 + 0]

Much better!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@48307 91177308-0d34-0410-b5e6-96231b3b80d8
2008-03-12 22:19:41 +00:00
Dan Gohman
42a77880a8 Rename CountMemOperands to ComputeMemOperandsEnd to reflect what
it actually does. Simplify CountOperands a little by reusing
ComputeMemOperandsEnd. And reword some comments for both.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47198 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-16 00:36:48 +00:00
Nate Begeman
e179584f9b Change how FP immediates are handled.
1) ConstantFP is now expand by default
2) ConstantFP is not turned into TargetConstantFP during Legalize
   if it is legal.

This allows ConstantFP to be handled like Constant, allowing for 
targets that can encode FP immediates as MachineOperands.

As a bonus, fix up Itanium FP constants, which now correctly match,
and match more constants!  Hooray.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47121 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-14 08:57:00 +00:00
Dan Gohman
6f0d024a53 Rename MRegisterInfo to TargetRegisterInfo.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46930 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-10 18:45:23 +00:00
Dan Gohman
69de1932b3 Re-apply the memory operand changes, with a fix for the static
initializer problem, a minor tweak to the way the
DAGISelEmitter finds load/store nodes, and a renaming of the
new PseudoSourceValue objects.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46827 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-06 22:27:42 +00:00
Evan Cheng
334dc1f58d Revert 46556 and 46585. Dan please fix the PseudoSourceValue problem and re-commit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46623 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-31 21:00:00 +00:00
Dan Gohman
c6c391dadd Create a new class, MemOperand, for describing memory references
in the backend. Introduce a new SDNode type, MemOperandSDNode, for
holding a MemOperand in the SelectionDAG IR, and add a MemOperand
list to MachineInstr, and code to manage them. Remove the offset
field from SrcValueSDNode; uses of SrcValueSDNode that were using
it are all all using MemOperandSDNode now.

Also, begin updating some getLoad and getStore calls to use the
PseudoSourceValue objects.

Most of this was written by Florian Brander, some
reorganization and updating to TOT by me.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46585 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-31 00:25:39 +00:00
Evan Cheng
6b2cf285bd A semi-gross fix for a debug info issue. When inserting the "function start" label (i.e. first label in the entry block) take care to insert it at the beginning of the block.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@46568 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-30 19:35:32 +00:00
Chris Lattner
749c6f6b5e rename TargetInstrDescriptor -> TargetInstrDesc.
Make MachineInstr::getDesc return a reference instead
of a pointer, since it can never be null.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@45695 91177308-0d34-0410-b5e6-96231b3b80d8
2008-01-07 07:27:27 +00:00
Chris Lattner
84bc5427d6 Rename SSARegMap -> MachineRegisterInfo in keeping with the idea
that "machine" classes are used to represent the current state of
the code being compiled.  Given this expanded name, we can start 
moving other stuff into it.  For now, move the UsedPhysRegs and
LiveIn/LoveOuts vectors from MachineFunction into it.

Update all the clients to match.

This also reduces some needless #includes, such as MachineModuleInfo
from MachineFunction.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@45467 91177308-0d34-0410-b5e6-96231b3b80d8
2007-12-31 04:13:23 +00:00
Chris Lattner
7ed47a1335 Don't attribute in file headers anymore. See llvmdev for the
discussion of this change.  Boy are my fingers tired. ;-)



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@45411 91177308-0d34-0410-b5e6-96231b3b80d8
2007-12-29 19:59:42 +00:00
Evan Cheng
f10c973797 If a node that defines a physical register that is expensive to copy. The
scheduler will try a number of tricks in order to avoid generating the
copies. This may not be possible in case the node produces a chain value
that prevent movement. Try unfolding the load from the node before to allow
it to be moved / cloned.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42625 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-05 01:39:18 +00:00
Evan Cheng
450ff01faf Remove simple scheduler.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42499 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-01 20:44:07 +00:00
Evan Cheng
22a529990b If two instructions are both two-address code, favors (schedule closer to
terminator) the one that has a CopyToReg use. This fixes
2006-05-11-InstrSched.ll with -new-cc-modeling-scheme.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42453 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-28 22:32:30 +00:00
Evan Cheng
74d2fd8dd8 Trim some unneeded fields.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42442 91177308-0d34-0410-b5e6-96231b3b80d8
2007-09-28 19:24:24 +00:00