330 Commits

Author SHA1 Message Date
Owen Anderson
76a561f35e Don't assign a new stack slot if the pre-alloc splitter already assigned one.
llvm-svn: 67764
2009-03-26 18:53:38 +00:00
Evan Cheng
f858466018 Fix PR3391 and PR3864. Reg allocator infinite looping.
llvm-svn: 67544
2009-03-23 18:24:37 +00:00
Evan Cheng
968c3b0d6e Model inline asm constraint which ties an input to an output register as machine operand TIED_TO constraint. This eliminated the need to pre-allocate registers for these. This also allows register allocator can eliminate the unneeded copies.
llvm-svn: 67512
2009-03-23 08:01:15 +00:00
Sebastian Redl
8d5baa09f8 Fix the Win32 VS2008 build:
- Make type declarations match the struct/class keyword of the definition.
 - Move AddSignalHandler into the namespace where it belongs.
 - Correctly call functions from template base.
 - Some other small changes.
With this patch, LLVM and Clang should build properly and with far less noise under VS2008.

llvm-svn: 67347
2009-03-19 23:26:52 +00:00
Evan Cheng
1361cbbb0b Added MachineInstr::isRegTiedToDefOperand to check for two-addressness.
llvm-svn: 67335
2009-03-19 20:30:06 +00:00
Evan Cheng
f0bfc6a675 Fix how livein live intervals are handled. Previously it could end at MBB start. Sorry, no small test case possible.
llvm-svn: 66129
2009-03-05 03:34:26 +00:00
Evan Cheng
b3783639cb Fix PR3486. Fix a bug in code that manually patch physical register live interval after its sub-register is coalesced with a virtual register.
llvm-svn: 64082
2009-02-08 11:04:35 +00:00
Evan Cheng
76a2736c74 Exit with nice warnings when register allocator run out of registers.
llvm-svn: 63267
2009-01-29 02:20:59 +00:00
Evan Cheng
c544cb0eca Change TargetInstrInfo::isMoveInstr to return source and destination sub-register indices as well.
llvm-svn: 62600
2009-01-20 19:12:24 +00:00
Evan Cheng
f6768bd9cb The coalescer does not coalesce a virtual register to a physical register if any of the physical register's sub-register live intervals overlaps with the virtual register. This is overly conservative. It prevents a extract_subreg from being coalesced away:
v1024 = EDI  // not killed
      =
      = EDI

One possible solution is for the coalescer to examine the sub-register live intervals in the same manner as the physical register. Another possibility is to examine defs and uses (when needed) of sub-registers. Both solutions are too expensive. For now, look for "short virtual intervals" and scan instructions to look for conflict instead.

This is a small win on x86-64. e.g. It shaves 403.gcc by ~80 instructions.

llvm-svn: 61847
2009-01-07 02:08:57 +00:00
Evan Cheng
0869f78555 Fix PR3149. If an early clobber def is a physical register and it is tied to an input operand, it effectively extends the live range of the physical register. Currently we do not have a good way to represent this.
172     %ECX<def> = MOV32rr %reg1039<kill>
180     INLINEASM <es:subl $5,$1
        sbbl $3,$0>, 10, %EAX<def>, 14, %ECX<earlyclobber,def>, 9, %EAX<kill>,
36, <fi#0>, 1, %reg0, 0, 9, %ECX<kill>, 36, <fi#1>, 1, %reg0, 0
188     %EAX<def> = MOV32rr %EAX<kill>
196     %ECX<def> = MOV32rr %ECX<kill>
204     %ECX<def> = MOV32rr %ECX<kill>
212     %EAX<def> = MOV32rr %EAX<kill>
220     %EAX<def> = MOV32rr %EAX
228     %reg1039<def> = MOV32rr %ECX<kill>

The early clobber operand ties ECX input to the ECX def.

The live interval of ECX is represented as this:
%reg20,inf = [46,47:1)[174,230:0)  0@174-(230) 1@46-(47)

The right way to represent this is something like
%reg20,inf = [46,47:2)[174,182:1)[181:230:0)  0@174-(182) 1@181-230 @2@46-(47)

Of course that won't work since that means overlapping live ranges defined by two val#.

The workaround for now is to add a bit to val# which says the val# is redefined by a early clobber def somewhere. This prevents the move at 228 from being optimized away by SimpleRegisterCoalescing::AdjustCopiesBackFrom.

llvm-svn: 61259
2008-12-19 20:58:01 +00:00
Dan Gohman
9de2ac8e8b Clarify some comments.
llvm-svn: 60683
2008-12-08 04:53:23 +00:00
Evan Cheng
43c0891838 Reason #3 from 60595 doesn't hold true. If we can fold a PIC load from constpool into a use, the rewrite happens at time of spill (not in VirtRegMap). Later on, if the GlobalBaseReg is spilled, the spiller can see the use uses GlobalBaseReg and do the right thing.
llvm-svn: 60596
2008-12-05 17:41:31 +00:00
Evan Cheng
994faaf218 Fix comment.
llvm-svn: 60592
2008-12-05 17:00:16 +00:00
Dan Gohman
d24be45d99 Drop the reg argument to isRegReDefinedByTwoAddr, which was redundant.
llvm-svn: 60586
2008-12-05 05:45:42 +00:00
Dan Gohman
69cc2cbbff Rename isSimpleLoad to canFoldAsLoad, to better reflect its meaning.
llvm-svn: 60487
2008-12-03 18:15:48 +00:00
Dan Gohman
3336b1f06b LiveRanges are represented as half-open ranges. Fix the findLiveInMBBs code
and the LiveInterval.h top-level comment and accordingly. This fixes blocks
having spurious live-in registers in boundary cases.

llvm-svn: 60092
2008-11-26 05:50:31 +00:00
Devang Patel
cb181bb203 Silence unused variable warnings.
llvm-svn: 59841
2008-11-21 20:00:59 +00:00
Dan Gohman
8fed4ce0b8 Use find_first/find_next to iterate through all the set bits in a
BitVector, instead of manually testing each bit.

llvm-svn: 59246
2008-11-13 16:31:27 +00:00
Dan Gohman
5efcb61337 Remove some debugging code made redundant by the change to do
coalescing as a separate pass rather than inside of
LiveIntervalAnalysis.

llvm-svn: 59146
2008-11-12 17:09:23 +00:00
Evan Cheng
938e201528 - Rewrite code that update register live interval that's split.
- Create and update spill slot live intervals.
- Lots of bug fixes.

llvm-svn: 58367
2008-10-29 05:06:14 +00:00
David Greene
b00b267bcc Fix PR2634. Create new virtual registers from spills early so that we
can give it the same stack slot as the spilled interval if it is folded.
This prevents the fold/unfold code from pointing to the wrong register.

llvm-svn: 58255
2008-10-27 17:38:59 +00:00
Evan Cheng
4bac4d0a16 Avoid splitting an interval multiple times; avoid splitting re-materializable val# (for now).
llvm-svn: 58068
2008-10-24 02:05:00 +00:00
Evan Cheng
2dadd3bbc6 By min, I mean max.
llvm-svn: 57766
2008-10-18 05:21:37 +00:00
Evan Cheng
ac4e70d946 When creating intervals, leave min(1, numdefs) holes after each instruction.
llvm-svn: 57765
2008-10-18 05:18:55 +00:00
Owen Anderson
1d338fc6a4 Add an option to enable StrongPHIElimination, for ease of testing.
llvm-svn: 57259
2008-10-07 20:22:28 +00:00
Dan Gohman
0d1e9a8e04 Switch the MachineOperand accessors back to the short names like
isReg, etc., from isRegister, etc.

llvm-svn: 57006
2008-10-03 15:45:36 +00:00
Owen Anderson
f4a36407c7 Fix a simple error in renumbering kill markaers, that took an inordinant amount of time to track down.
llvm-svn: 56889
2008-09-30 22:51:54 +00:00
Evan Cheng
9156bd2f48 Re-apply 56835 along with header file changes.
llvm-svn: 56848
2008-09-30 15:44:16 +00:00
Duncan Sands
2b9adce1d0 Revert commit 56835 since it breaks the build.
"If a re-materializable instruction has a register
operand, the spiller will change the register operand's
spill weight to HUGE_VAL to avoid it being spilled.
However, if the operand is already in the queue ready
to be spilled, avoid re-materializing it".

llvm-svn: 56837
2008-09-30 10:00:30 +00:00
Evan Cheng
9469049f7d If a re-materializable instruction has a register operand, the spiller will change the register operand's spill weight to HUGE_VAL to avoid it being spilled. However, if the operand is already in the queue ready to be spilled, avoid re-materializing it.
llvm-svn: 56835
2008-09-30 06:36:58 +00:00
Dale Johannesen
c36660d756 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.

llvm-svn: 56539
2008-09-24 01:07:17 +00:00
Owen Anderson
df8f1cb995 Fetch the starting index of the block when assigning intervals. This gets live-in indices
correct in the presence of things like EH labels.

llvm-svn: 56410
2008-09-21 20:43:24 +00:00
Dale Johannesen
e519bd4183 Remove AsmThatEarlyClobber etc. from LiveIntervalAnalysis
and redo as linked list walk.  Logic moved into RA.
Per review feedback.

llvm-svn: 56326
2008-09-19 01:02:35 +00:00
Dale Johannesen
f8610ebebc 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.

llvm-svn: 56290
2008-09-17 21:13:11 +00:00
Owen Anderson
82ab1e7280 Live intervals for live-in registers should begin at the beginning of a basic block, not at the first
instruction.  Also, their valno's should have an unknown def.  This has no effect currently, but was
causing issues when StrongPHIElimination was enabled.

llvm-svn: 56231
2008-09-15 22:00:38 +00:00
Dan Gohman
38453eebdc Remove isImm(), isReg(), and friends, in favor of
isImmediate(), isRegister(), and friends, to avoid confusion
about having two different names with the same meaning. I'm
not attached to the longer names, and would be ok with
changing to the shorter names if others prefer it.

llvm-svn: 56189
2008-09-13 17:58:21 +00:00
Owen Anderson
2a458775db Allow the fast-path spilling code to attempt folding, but still leaving out remat and splitting.
llvm-svn: 55012
2008-08-19 22:12:11 +00:00
Owen Anderson
c75ae13566 The fast-path still needs to set kill markers and spill/restore points as appropriate.
With this patch, all of MultiSource/Applications and all of SPEC2000/2006 pass with
the SimpleSpiller and this fast-path enabled.

llvm-svn: 55000
2008-08-19 20:09:52 +00:00
Owen Anderson
442dc6fed0 Add a flag to enable the fast spilling path.
llvm-svn: 54958
2008-08-19 00:17:30 +00:00
Owen Anderson
a57c57067d Fix a few more bugs:
1) Assign stack slots to new temporaries.
  2) Don't insert an interval into the return vector more than once.

llvm-svn: 54956
2008-08-18 23:41:04 +00:00
Owen Anderson
60f6b2cac0 Fix several bugs in the new fast-path:
1) Remove an incorrect assertion.
  2) Set the stack slot weight properly.
  3) Resize the VirtRegMap when needed.

llvm-svn: 54949
2008-08-18 21:20:32 +00:00
Owen Anderson
e996a12f27 Clients of addIntervalForSpills expect the added intervals to be returned sorted by starting index.
llvm-svn: 54939
2008-08-18 19:52:22 +00:00
Owen Anderson
75e27d2402 Simplify the fast-patch interval spilling by using MachineRegisterInfo::reg_iterator.
llvm-svn: 54930
2008-08-18 18:38:12 +00:00
Owen Anderson
cc9314ba32 Resurrect some ancient code to add spill ranges without attempting folding, remat, or splitting. This code has been updated to current APIs
in so far as it compiles and, in theory, works, but does not take advantage of recent advancements.  For instance, it could be improved by using
MachineRegisterInfo::use_iterator.

llvm-svn: 54924
2008-08-18 18:05:32 +00:00
Owen Anderson
706f6b7899 Expunge the last uses of std::map from LiveIntervals.
llvm-svn: 54766
2008-08-13 22:28:50 +00:00
Owen Anderson
767b5cc7fd Move r2iMap_ over to DenseMap from std::map.
llvm-svn: 54765
2008-08-13 22:08:30 +00:00
Owen Anderson
51f689a652 Make the allocation of LiveIntervals explicit, rather than holding them in the r2iMap_ by value. This will prevent references to them from being invalidated
if the map is changed.

llvm-svn: 54763
2008-08-13 21:49:13 +00:00
Owen Anderson
d184929176 Oops, didn't mean to commit this.
llvm-svn: 54425
2008-08-06 20:58:38 +00:00
Owen Anderson
03dddbbed5 Only remap each VNInfo once when doing renumbering.
llvm-svn: 54420
2008-08-06 18:35:45 +00:00