Commit Graph

54399 Commits

Author SHA1 Message Date
Jakob Stoklund Olesen fffff88a3c Enable -split-phi-edges by default, except when -regalloc=local.
The local register allocator doesn't like it when LiveVariables is run.
We should also disable edge splitting under -O0, but that has to wait a bit.

llvm-svn: 89125
2009-11-17 19:15:50 +00:00
Jim Grosbach e4e018ae67 80-column violations
llvm-svn: 89123
2009-11-17 19:05:35 +00:00
Viktor Kutuzov dafdd883bc Added getArchNameForAssembler method to the Triple class for which returns OS and Vendor independent target assembler arch.
llvm-svn: 89122
2009-11-17 18:48:27 +00:00
Bob Wilson eeac7ce769 Remove a special case for tail merging that seems to be both broken and
unnecessary.  It is broken because the "isIdenticalTo" check should be
negated.  If that is fixed, this code causes the CodeGen/X86/tail-opts.ll
test to fail, in the dont_merge_oddly function.  And, I confirmed that the
regression is real -- the generated code is worse.  As far as I can tell,
that tail-opts.ll test is checking for what this code is supposed to handle
and we're doing the right thing anyway.

llvm-svn: 89121
2009-11-17 18:30:09 +00:00
Evan Cheng ba4e5da727 Generalize OptimizeLoopTermCond to optimize more loop terminating icmp to use postinc iv.
llvm-svn: 89116
2009-11-17 18:10:11 +00:00
Dan Gohman 2650723d81 Set MadeChange instead of MadeChangeThisIteration.
llvm-svn: 89114
2009-11-17 18:04:15 +00:00
Benjamin Kramer 713fd3564a Revert CPU detection code to return "generic" instead of an empty string in case
of failure. The x86 target didn't like empty cpu names and broke x86 tests on
non-x86 buildbots.

llvm-svn: 89111
2009-11-17 17:57:04 +00:00
Jim Grosbach 60f4854c76 Remove trailing whitespace
llvm-svn: 89110
2009-11-17 17:53:56 +00:00
Bob Wilson 724d8a45c1 Update a comment, now that tail duplication happens after other branch
folding optimizations.

llvm-svn: 89109
2009-11-17 17:40:31 +00:00
Johnny Chen 14b25eb87b Set Inst{15-12} (Rd/Rt) to 0b1111 (PC) for BR_JTadd, BR_JTr, and BR_JTm to
distinguish between them and the more generic instructions (add, mov, and ldr).

llvm-svn: 89108
2009-11-17 17:17:50 +00:00
Bob Wilson bd22f1965d Perform tail duplication only once, after tail merging is complete.
It was too difficult to keep the heuristics for merging and duplication
consistent.

llvm-svn: 89105
2009-11-17 17:06:18 +00:00
Nuno Lopes 6f6668cac2 add Case() with 5 args
llvm-svn: 89099
2009-11-17 15:35:39 +00:00
Jay Foad 658af5a9e2 Fix HTML formatting.
llvm-svn: 89093
2009-11-17 13:13:59 +00:00
Duncan Sands 7a04e7df6d 1. Allow SCCIterator to work with GraphT types that are constant.
2.  Allow SCCIterator to work with inverse graphs.
3.  Fix an incorrect comment in GraphTraits.h (the type in the comment 
was given as GraphType* when it is actually const GraphType &).
Patch by Patrick Alexander Simmons.

llvm-svn: 89091
2009-11-17 10:54:25 +00:00
Duncan Sands 3246fe0b6b Make bugpoint pass -load arguments to LLI. This lets one use bugpoint with
programs that depend on native shared libraries.  Patch by Timo Lindfors.

llvm-svn: 89087
2009-11-17 10:20:22 +00:00
Evan Cheng 84efacfaad Revert 89021. It's miscompiling llvm-gcc driver driver at -O0.
llvm-svn: 89082
2009-11-17 09:55:52 +00:00
Evan Cheng 5392cc9d14 Re-apply 89011. It's not to be blamed.
llvm-svn: 89081
2009-11-17 09:51:18 +00:00
Daniel Dunbar 1acdfbd60b "XFAIL" the Split2 StringReft test with Apple gcc, which miscompiles it.
- I plan on fixing/workarounding this, but until then I'd like the bots to stay
   green.

llvm-svn: 89077
2009-11-17 09:29:59 +00:00
Evan Cheng 05938e819b Revert 89011. Buildbot thinks it might be breaking stuff.
llvm-svn: 89076
2009-11-17 09:20:28 +00:00
Nick Lewycky a1c09d674e Remove VISIBILITY_HIDDEN from the classes in this directory. Fixes bug 5507.
llvm-svn: 89075
2009-11-17 09:17:08 +00:00
Duncan Sands 8b9336bd98 Following a suggestion of Daniel Dunbar, stop people passing the name
as the isSigned bool to CreateIntCast by having this resolve to a call
to a private method, rather than by using a gcc attribute.

llvm-svn: 89067
2009-11-17 08:34:52 +00:00
Nick Lewycky b7993d61f2 Revert r88939.
llvm-svn: 89066
2009-11-17 08:11:44 +00:00
Nick Lewycky 7efd07fd9d Fail less mysteriously; inform the user that their LLVM was not built with
libffi support and that the interpreter can't call external functions without
it. Patch by Timo Juhani Lindfors! Fixes PR5466.

llvm-svn: 89062
2009-11-17 07:52:09 +00:00
Lang Hames 48770eb61f Fixed call to wrong constructor.
llvm-svn: 89059
2009-11-17 07:19:50 +00:00
Owen Anderson 759d72dc8e Fix a race condition in the Timer class.
llvm-svn: 89056
2009-11-17 07:06:10 +00:00
Bill Wendling bc0020174b Refactor the code that creates the "dot-label" difference. This may be used in
more than one place. No intended functionality change.

llvm-svn: 89024
2009-11-17 01:23:53 +00:00
Jim Grosbach af1ad30a23 When moving a block for table jumps, make sure the prior block terminator
is analyzable so it can be updated. If it's not, be safe and don't move the
block.

llvm-svn: 89022
2009-11-17 01:21:04 +00:00
Jakob Stoklund Olesen 9f0d55d8d8 Enable -split-phi-edges by default
llvm-svn: 89021
2009-11-17 01:07:22 +00:00
Evan Cheng d33400e636 MOV64rm should be marked isReMaterializable.
llvm-svn: 89019
2009-11-17 00:55:55 +00:00
Dan Gohman 82e80019a5 Remove the optimizations that convert BRCOND and BR_CC into
unconditional branches or fallthroghes. Instcombine/SimplifyCFG
should be simplifying branches with known conditions.

This fixes some problems caused by these transformations not
updating the MachineBasicBlock CFG.

llvm-svn: 89017
2009-11-17 00:47:23 +00:00
Devang Patel 12144a2348 Remove debug info attached with an instruction.
llvm-svn: 89016
2009-11-17 00:47:06 +00:00
Jeffrey Yasskin 7c57c41d5f In GlobalVariable::setInitializer, assert that the initializer has the
right type.

llvm-svn: 89014
2009-11-17 00:43:13 +00:00
Evan Cheng ce28f6f478 A few more instructions that should be marked re-materializable.
llvm-svn: 89011
2009-11-17 00:23:22 +00:00
Jim Grosbach 0ad7efbace Convert to FileCheck
llvm-svn: 89007
2009-11-17 00:20:26 +00:00
Jim Grosbach 4781c3caf8 Convert to FileCheck
llvm-svn: 89002
2009-11-17 00:03:38 +00:00
Jim Grosbach 805d195649 Cleanup. Missed removing these when converting. Oops.
llvm-svn: 89001
2009-11-17 00:00:33 +00:00
Johnny Chen 18183b6d8e Set Rm bits of BX_RET to 0b1110 (R14); and set condition code bits of BRIND to
0b1110 (ALways).  This is so that the disassembler decoder can distinguish among
BX_RET, BRIND, and BXr9.

llvm-svn: 89000
2009-11-16 23:57:56 +00:00
Dan Gohman b43e1ff236 Fix this test - there don't appear to be any actual Reload Reuses
in this testcase.

llvm-svn: 88998
2009-11-16 23:49:55 +00:00
Dan Gohman 9dede3b383 Revert r87049, which was the workaround for the regression triggered
by the recent FixedStackPseudoSourceValue-related changes, now that
the specific bug that affected it is fixed, in r88954.

llvm-svn: 88997
2009-11-16 23:43:42 +00:00
Jeffrey Yasskin 0632b53bfe Revert the test from r88984. It relies on being able to mmap 16GB of
address space (though it only uses a small fraction of that), and the
buildbots disallow that.

Also add a comment to the Makefile's ulimit line warning future
developers that changing it won't work.

llvm-svn: 88994
2009-11-16 23:32:30 +00:00
Jim Grosbach 1deb0b9f53 Convert to FileCheck
llvm-svn: 88991
2009-11-16 23:19:29 +00:00
Dan Gohman 9b5eea30f7 Initialize the new AsmPrinterFlags field to 0, fixing uses of
uninitialized memory.

llvm-svn: 88985
2009-11-16 22:49:38 +00:00
Jeffrey Yasskin 10d3604a9e Make X86-64 in the Large model always emit 64-bit calls.
The large code model is documented at
http://www.x86-64.org/documentation/abi.pdf and says that calls should
assume their target doesn't live within the 32-bit pc-relative offset
that fits in the call instruction.

To do this, we turn off the global-address->target-global-address
conversion in X86TargetLowering::LowerCall(). The first attempt at
this broke the lazy JIT because it can separate the movabs(imm->reg)
from the actual call instruction. The lazy JIT receives the address of
the movabs as a relocation and needs to record the return address from
the call; and then when that call happens, it needs to patch the
movabs with the newly-compiled target. We could thread the call
instruction into the relocation and record the movabs<->call mapping
explicitly, but that seems to require at least as much new
complication in the code generator as this change.

To fix this, we make lazy functions _always_ go through a call
stub. You'd think we'd only have to force lazy calls through a stub on
difficult platforms, but that turns out to break indirect calls
through a function pointer. The right fix for that is to distinguish
between calls and address-of operations on uncompiled functions, but
that's complex enough to leave for someone else to do.

Another attempt at this defined a new CALL64i pseudo-instruction,
which expanded to a 2-instruction sequence in the assembly output and
was special-cased in the X86CodeEmitter's emitInstruction()
function. That broke indirect calls in the same way as above.

This patch also removes a hack forcing Darwin to the small code model.
Without far-call-stubs, the small code model requires things of the
JITMemoryManager that the DefaultJITMemoryManager can't provide.

Thanks to echristo for lots of testing!

llvm-svn: 88984
2009-11-16 22:41:33 +00:00
Daniel Dunbar 1aab2ded3a Don't build examples by default, use BUILD_EXAMPLES=1 to build them. The only utility of this is testing that we keep the examples up to date, I will just make the buildbots run with this flag.
llvm-svn: 88979
2009-11-16 22:38:00 +00:00
Daniel Dunbar 08d4ab9663 Add "Unoptimized" build (NO_DEBUG_SYMBOLS=1 ENABLE_OPTIMIZED=1), for reducing
disk space, and increasing battery lifetime. :)

llvm-svn: 88978
2009-11-16 22:37:52 +00:00
Eric Christopher dc679f88c8 Fix unused variables warnings.
llvm-svn: 88977
2009-11-16 22:34:32 +00:00
Evan Cheng f25ef4ffb0 - Check memoperand alignment instead of checking stack alignment. Most load / store folding instructions are not referencing spill stack slots.
- Mark MOVUPSrm re-materializable.

llvm-svn: 88974
2009-11-16 21:56:03 +00:00
Devang Patel f3d7c08e34 Revert r88939.
llvm-svn: 88973
2009-11-16 21:53:40 +00:00
David Greene a3ce7828b2 Fix an expensive-checks error.
The Mask and LHSMask may not be of the same size, so don't do the
transformation if they're different.

llvm-svn: 88972
2009-11-16 21:52:23 +00:00
Jim Grosbach 56ee15c162 Make the pass class name more explicit.
llvm-svn: 88964
2009-11-16 21:13:22 +00:00
Jim Grosbach 544b15cbc3 make pass name a bit more clear
llvm-svn: 88961
2009-11-16 21:03:58 +00:00
Dan Gohman f76945ff14 Revert 88957. This file uses CodeGenOpt, which is defined in TargetMachine.h.
llvm-svn: 88959
2009-11-16 20:45:50 +00:00
Dan Gohman 0bcda0d5bd Remove an unnecessary #include.
llvm-svn: 88957
2009-11-16 20:41:12 +00:00
Dan Gohman 93f529d514 Sink a #include <map> to where it's actually needed.
llvm-svn: 88956
2009-11-16 20:40:47 +00:00
Dan Gohman 4617f3e893 Make PseudoSourceValue's classof recognize
FixedStackPseudoSourceValueVal, to respect this isa relationship.

llvm-svn: 88954
2009-11-16 20:40:06 +00:00
Dan Gohman 6b3f32e6d7 Fix a typo in a comment.
llvm-svn: 88953
2009-11-16 20:35:59 +00:00
Jim Grosbach 9b32e22ad1 Convert to FileCheck
llvm-svn: 88947
2009-11-16 20:04:15 +00:00
Lang Hames 16f6b3e607 Added a testcase for PR5495.
llvm-svn: 88946
2009-11-16 20:03:13 +00:00
Rafael Espindola 65e9be6dc2 Add configure options for specifying where to look for libstdc++.
llvm-svn: 88943
2009-11-16 19:46:55 +00:00
Jim Grosbach 980d94164d Convert to FileCheck
llvm-svn: 88942
2009-11-16 19:46:46 +00:00
Bob Wilson 44724c4122 Fix a comment.
llvm-svn: 88940
2009-11-16 19:33:27 +00:00
Devang Patel 2ffd9935a1 Add VISIBILITY_HIDDEN marker.
llvm-svn: 88939
2009-11-16 19:20:48 +00:00
Jim Grosbach 9785e5914f Simplify thumb2 jump table adjustments. Remove unnecessary calculation and
usage of block sizes and offsets.

llvm-svn: 88935
2009-11-16 18:58:52 +00:00
Jim Grosbach 87b0f0d95c clarify comment
llvm-svn: 88933
2009-11-16 18:55:47 +00:00
Bob Wilson 1a8ea98b76 Fix some comments.
llvm-svn: 88932
2009-11-16 18:54:08 +00:00
Bob Wilson 3794ec2519 Whitespace: be consistent with pointer syntax.
llvm-svn: 88929
2009-11-16 18:08:46 +00:00
Bob Wilson 699f5b9a0a Clean up whitespace.
llvm-svn: 88927
2009-11-16 17:56:13 +00:00
Jim Grosbach c670bdc311 tbb opt off by default
llvm-svn: 88921
2009-11-16 17:24:45 +00:00
Jim Grosbach 9c148c2163 back off for a bit. tracking down weirdness
llvm-svn: 88919
2009-11-16 17:17:48 +00:00
Jim Grosbach 47d5e333ff Analyze has to be before checking the condition, obviously. Properly construct an iterator for prior.
llvm-svn: 88917
2009-11-16 17:10:56 +00:00
Douglas Gregor 16917aa19d Make ERROR_IF_USED macro work with GCC <= 4.2, Apple GCCs
llvm-svn: 88916
2009-11-16 16:56:48 +00:00
Duncan Sands 551861e7d9 Make sure that if anyone passes a name by accident for the isSigned
parameter of CreateIntCast then they get an error from the compiler
(or from the linker with a non-gcc compiler).  Another possibility
is to flip the order of the DestTy and isSigned parameters, since you
should then get a compiler warning if you try to use a char* for a
Type*.

llvm-svn: 88913
2009-11-16 15:28:17 +00:00
David Greene 25905c8336 Support spill comments.
Have the asm printer emit a comment if an instruction is a spill or
reload and have the spiller mark copies it introdues so the asm printer
can also annotate those.

llvm-svn: 88911
2009-11-16 15:12:23 +00:00
Duncan Sands 9649e6085d BuildIntCast takes an additional parameter, isSigned.
llvm-svn: 88910
2009-11-16 13:15:28 +00:00
Duncan Sands e5de4a9ad6 CreateIntCast takes an "isSigned" parameter. Pass "true" for it, rather than
a name.

llvm-svn: 88908
2009-11-16 12:32:28 +00:00
Evan Cheng 00e87d19f5 Special case FixedStackPseudoSourceValueVal as well. Do we really need to differentiate PseudoSourceValueVal from FixedStackPseudoSourceValueVal at this level?
llvm-svn: 88902
2009-11-16 07:10:36 +00:00
Evan Cheng 597f7b6ee3 Check if subreg index is zero.
llvm-svn: 88899
2009-11-16 06:31:49 +00:00
Evan Cheng 11bf4493d4 For some targets, a copy can use a register multiple times, e.g. ppc.
llvm-svn: 88895
2009-11-16 05:52:06 +00:00
Evan Cheng 8ca5d4b9ad xfail for now. It has been failing.
llvm-svn: 88892
2009-11-16 05:44:04 +00:00
Bruno Cardoso Lopes 5bf807688e Disable ldc1/sdc1 instructions for mips1 targets.
llvm-svn: 88887
2009-11-16 04:35:29 +00:00
Bruno Cardoso Lopes 537e409c58 - Fix a small bug while handling target constant pools (one param was missing).
- Add a smarter constant pool loading, instead of:

lui $2, %hi($CPI1_0)
addiu $2, $2, %lo($CPI1_0)
lwc1 $f0, 0($2)

Generate:

lui $2, %hi($CPI1_0)
lwc1 $f0, %lo($CPI1_0)($2)

llvm-svn: 88886
2009-11-16 04:33:42 +00:00
Chris Lattner e58c05780e typo spotted by duncan.
llvm-svn: 88884
2009-11-16 03:51:42 +00:00
Lang Hames 01f31c0a48 Fixes the bug exposed by Anton's test case in PR 5495:
Make sure when ProcessImplicitDefs removes a copy which kills its source reg that it
removes the copy from said reg's Kills list.

llvm-svn: 88881
2009-11-16 02:07:31 +00:00
Lang Hames 968bd90f88 Fix for the original bug in PR5495 - Look at uses as well as defs when determining the PHI-copy insert point.
- Patch by Andrew Canis!

llvm-svn: 88880
2009-11-16 02:00:09 +00:00
Jim Grosbach 01c1cae34d Detect need for autoalignment of the stack earlier to catch spills more
conservatively. eliminateFrameIndex() machinery adjust to handle addr mode
6 (vld1/vst1) used for spills. Fix tests to expect aligned Q-reg spilling

llvm-svn: 88874
2009-11-15 21:45:34 +00:00
Jim Grosbach 74ae3e5b0e set the def of the VLD1q64 properly
llvm-svn: 88873
2009-11-15 21:05:07 +00:00
Chris Lattner ee4cf81d9a disable copying, enforce some invariants.
llvm-svn: 88870
2009-11-15 20:03:53 +00:00
Chris Lattner 7735878591 teach LVI to infer edge information from switch instructions.
This allows JT to eliminate a ton of infeasible edges when
handling code like the templates in PatternMatch.h

llvm-svn: 88869
2009-11-15 20:02:12 +00:00
Chris Lattner b0c0a0df3e fix a logic error that would cause LVI-JT to miscompile
some conditionals

llvm-svn: 88868
2009-11-15 20:01:24 +00:00
Chris Lattner 2c708562f5 implement the first stab at caching queries. This isn't correct
(because the invalidation logic is missing) but LVI isn't enabled
by default anyway.

llvm-svn: 88867
2009-11-15 20:00:52 +00:00
Chris Lattner af025d3f6e refactor a bunch of code forming the new LazyValueInfoCache
and LVIQuery classes, no functionality change.

llvm-svn: 88866
2009-11-15 19:59:49 +00:00
Chris Lattner 9d9812a636 make PRE of loads preserve the alignment of the moved load instruction.
llvm-svn: 88865
2009-11-15 19:58:31 +00:00
Chris Lattner 5f037b6439 fix a bug handling 'not x' when x is undef.
llvm-svn: 88864
2009-11-15 19:57:43 +00:00
Chris Lattner fb7613a5d6 mark getIntrinsicID() 'readonly'. This allows various classof methods
(like DbgDeclareInst's) to shrink substantially.  It sucks that we have
to pull Compiler.h into such a public header, but at least Compiler.h
doesn't pull anything else in.

llvm-svn: 88863
2009-11-15 19:56:28 +00:00
Chris Lattner f456ae74c8 add attributes for readnone/readonly functions.
llvm-svn: 88862
2009-11-15 19:54:31 +00:00
Chris Lattner 8938c85eb5 add a version of array_pod_sort that takes a custom comparator function.
llvm-svn: 88861
2009-11-15 19:52:43 +00:00
Nick Lewycky ef4ea9a2a9 Add a complex missed optimization opportunity I came across while investigating
bug 5438.

llvm-svn: 88855
2009-11-15 17:51:23 +00:00
Edward O'Callaghan e3f3ca70e6 Add PSP OS Target to Triple, Credit to Bruno Cardoso Lopes.
llvm-svn: 88849
2009-11-15 10:18:17 +00:00
Daniel Dunbar 9a61bd5022 lit: Factor a new OneCommandPerFileTest out of SyntaxCheckTest.
- Used for running a single fixed command on a directory of files, with the
   option of deriving a temporary input file from the test source.

llvm-svn: 88844
2009-11-15 08:10:29 +00:00
Nick Lewycky 95148689c9 Revert r88830 and r88831 which appear to have caused a selfhost buildbot some
grief. I suspect this patch merely exposed a bug else.

llvm-svn: 88841
2009-11-15 07:47:32 +00:00
Daniel Dunbar d4764717de Remove duplicate implementation of excludes functionality, and support excluding
dirnames.

Also, add support for the 'unsupported' config property.

llvm-svn: 88838
2009-11-15 07:22:58 +00:00
Nick Lewycky 6a6ac7e105 Correct typo.
llvm-svn: 88831
2009-11-15 06:16:57 +00:00
Nick Lewycky e29fa4c7a1 Teach instcombine to look for booleans in wider integers when it encounters a
zext(icmp). It may be able to optimize that away. This fixes one of the cases
in PR5438.

llvm-svn: 88830
2009-11-15 05:55:17 +00:00
Lang Hames 6da915ac86 Added an assert to the PBQP allocator to catch infinite cost solutions which might otherwise lead to miscompilations.
llvm-svn: 88829
2009-11-15 04:39:51 +00:00
Daniel Dunbar ad6c15465c lit: Add --repeat=N option, for running each test N times.
- Currently just useful for timing, although it could be extended as one (bad) way to deal with flaky tests.

llvm-svn: 88827
2009-11-15 01:02:09 +00:00
Daniel Dunbar 18f845fccd Remove bogus corei7 and atom entries, the family was incorrect.
llvm-svn: 88818
2009-11-14 22:04:42 +00:00
Jim Grosbach f16a3b7a9f remove xfail
llvm-svn: 88817
2009-11-14 21:57:35 +00:00
Daniel Dunbar 397235f11c Fill out X86 table, although we are missing lots of names for things. We now
properly detect my Xeon box though.

llvm-svn: 88814
2009-11-14 21:36:19 +00:00
Daniel Dunbar d90a9a04f3 Report the detected host CPU in --version.
llvm-svn: 88813
2009-11-14 21:36:07 +00:00
Jim Grosbach e07594df45 cleanup.
llvm-svn: 88812
2009-11-14 21:33:37 +00:00
Jim Grosbach ae84426ed8 Do not merge jump tables this early. Branch folding will do any necessary
merges, and until then, it's useful to keep the tables separate for ease
of manipulation.

llvm-svn: 88806
2009-11-14 20:15:03 +00:00
Jim Grosbach 43d2108546 Cleanup flow, and only update the jump table we're analyzing when replacing a destination MBB.
llvm-svn: 88805
2009-11-14 20:10:18 +00:00
Jim Grosbach 9c8609ea1c Add function to replace a destination MBB in a single jump table
llvm-svn: 88804
2009-11-14 20:09:13 +00:00
Benjamin Kramer 69865cc1db Remove dead variable found by clang++.
llvm-svn: 88803
2009-11-14 19:51:20 +00:00
Richard Osborne d5f2745965 Add XCore support for arbitrary-sized aggregate returns.
llvm-svn: 88802
2009-11-14 19:33:35 +00:00
Anton Korobeynikov fd0c7bae2a Temporary disable the error - it seems to be too conservative.
llvm-svn: 88800
2009-11-14 18:01:41 +00:00
Benjamin Kramer 8c19a8f17b Implement DISABLE_INLINE for MSVC. This required changing the position in all
forward declaration and patching tblgen to emit it right. Patch by Amine Khaldi!

llvm-svn: 88798
2009-11-14 16:37:18 +00:00
Benjamin Kramer 4197054ece This test doesn't work on arm either.
llvm-svn: 88794
2009-11-14 15:15:39 +00:00
Benjamin Kramer 56ffa48473 Make NORETURN working with MSVC. MSVC only accepts NORETURN in front of the
decl so move it there. GCC accepts it both in front and after decls.

llvm-svn: 88791
2009-11-14 14:14:58 +00:00
Daniel Dunbar 241d01b590 Add llvm::sys::getHostCPUName, for detecting the LLVM name for the host CPU.
- This is an initial step towards -march=native support in Clang, and towards
   eliminating host dependencies in the targets. See PR5389.

 - Patch by Roman Divacky!

llvm-svn: 88768
2009-11-14 10:09:12 +00:00
Nick Lewycky 7935bcb0fe Remove LLVMContext from reassociate. It was threaded through every function but
ultimately never used.

llvm-svn: 88763
2009-11-14 07:25:54 +00:00
Sanjiv Gupta d03742a122 revert 88761 as it fails builds.
llvm-svn: 88762
2009-11-14 07:22:25 +00:00
Sanjiv Gupta 8f9c22646b Fix debug info crashes for PIC16.
llvm-svn: 88761
2009-11-14 06:19:49 +00:00
Nick Lewycky c53e2ecf02 Teach BasicAA that a constant expression can't alias memory provably not
allocated until runtime (such as an alloca). Patch by Hans Wennborg!

llvm-svn: 88760
2009-11-14 06:15:14 +00:00
Evan Cheng 16797a1f55 Added getSubRegIndex(A,B) that returns subreg index of A to B. Use it to replace broken code in VirtRegRewriter.
llvm-svn: 88753
2009-11-14 03:42:17 +00:00
Evan Cheng 6ad7da96fe - Change TargetInstrInfo::reMaterialize to pass in TargetRegisterInfo.
- If destination is a physical register and it has a subreg index, use the
  sub-register instead.
This fixes PR5423.

llvm-svn: 88745
2009-11-14 02:55:43 +00:00
Dan Gohman 81132465d3 Add an option for running GVN with redundant load processing disabled.
llvm-svn: 88742
2009-11-14 02:27:51 +00:00
Evan Cheng e3b312fec9 Add radar number.
llvm-svn: 88739
2009-11-14 02:11:32 +00:00
Evan Cheng d2c10508cd Fix PR5412: Fix an inverted check and another missing sub-register check.
llvm-svn: 88738
2009-11-14 02:09:09 +00:00
Dan Gohman a627e26d39 Enable the tail call optimization when the caller returns undef.
llvm-svn: 88737
2009-11-14 02:06:30 +00:00
Evan Cheng 66401c90da When expanding t2STRDi8 r, r to two stores, add kill markers correctly.
llvm-svn: 88734
2009-11-14 01:50:00 +00:00
Jakob Stoklund Olesen 24b027401e Fix bug in -split-phi-edges.
When splitting an edge after a machine basic block with fall-through, we
forgot to insert a jump instruction. Fix this by calling updateTerminator() on
the fall-through block when relevant.

Also be more precise in PHIElimination::isLiveIn.

llvm-svn: 88728
2009-11-14 00:38:13 +00:00
Jakob Stoklund Olesen 15ca009ee7 Update MachineDominator information
llvm-svn: 88727
2009-11-14 00:38:06 +00:00
Lang Hames 6b7233afc6 Added an API to the SlotIndexes pass to allow new instructions to be inserted into the numbering.
PreAllocSplitting is now using this API to insert code.

llvm-svn: 88725
2009-11-14 00:02:51 +00:00
Evan Cheng 78fa302e7d Fix PR5411. Bug in UpdateKills. A reg def partially define its super-registers.
llvm-svn: 88719
2009-11-13 23:16:41 +00:00
Eric Christopher 96b26acd0f Remove extraneous commit.
llvm-svn: 88716
2009-11-13 23:08:47 +00:00
Eric Christopher 0cabd30b22 Print out something, even if it's non-parseable later when we've
got ghost linkage. It's better than aborting.

llvm-svn: 88715
2009-11-13 23:00:14 +00:00
Dan Gohman 6e99014947 Move the FixedStackPseudoSourceValueVal enum value before InstructionVal
so that isa<Instructon> doesn't return true for FixedStackPseudoSourceValue
values. This fixes a variety of problems, including crashes with -debug
and -print-machineinstrs. Also, add a comment to warn about this.

llvm-svn: 88711
2009-11-13 22:24:13 +00:00
Bill Wendling 6787318557 Disable the JITTest.NoStubs test for Darwin PPC. It apparently doesn't implement
emitFunctionStubAtAddr.

llvm-svn: 88708
2009-11-13 21:58:54 +00:00
Jakob Stoklund Olesen ad205d6102 Fix PHIElimination optimization that uses MBB->getBasicBlock.
The BasicBlock associated with a MachineBasicBlock does not necessarily
correspond to the code in the MBB.

Don't insert a new IR BasicBlock when splitting critical edges. We are not
supposed to modify the IR during codegen, and we should be able to do just
fine with a NULL BB.

llvm-svn: 88707
2009-11-13 21:56:15 +00:00
Jakob Stoklund Olesen 27440e71fc Add MachineFunction::verify() to call the machine code verifier directly.
llvm-svn: 88706
2009-11-13 21:56:09 +00:00
Jakob Stoklund Olesen ff0302489b The instruction pointer %RIP is a reserved register on x86_64.
llvm-svn: 88705
2009-11-13 21:56:01 +00:00
Jakob Stoklund Olesen 1ecc8b2de1 Fix polarity of a CFG check in machine verifier.
llvm-svn: 88704
2009-11-13 21:55:54 +00:00
Dan Gohman 7474546e4d Use .data() instead of .c_str() when nul-termination is not needed.
llvm-svn: 88703
2009-11-13 21:55:31 +00:00
Devang Patel 452c1646c6 Do not use value handle to wrap MDNode in DIDescriptor.
llvm-svn: 88700
2009-11-13 21:45:04 +00:00
David Greene 659c1a9d78 Move DebugInfo checks into EmitComments and remove them from
target-specific AsmPrinters.  Not all comments need DebugInfo.

Re-enable the line numbers comment test.

llvm-svn: 88697
2009-11-13 21:34:57 +00:00
Dan Gohman 225fa59cac When optimizing for size, don't tail-merge unless it's likely to be a
code-size win, and not when it's only likely to be code-size neutral,
such as when only a single instruction would be eliminated and a new
branch would be required.

This fixes rdar://7392894.

llvm-svn: 88692
2009-11-13 21:02:15 +00:00
Evan Cheng d190b8216f Fix PR5410: LiveVariables lost subreg def:
D0<def,dead> = ...
...
             = S0<use, kill>
S0<def>      = ...
...
D0<def>      = 

The first D0 def is correctly marked dead, however, livevariables should have
added an implicit def of S0 or we end up with a use without a def.

llvm-svn: 88690
2009-11-13 20:36:40 +00:00
David Goodwin b9fe5d5d02 Allow target to specify regclass for which antideps will only be broken along the critical path.
llvm-svn: 88682
2009-11-13 19:52:48 +00:00
Bruno Cardoso Lopes a03b5b44fc Support fp64 immediate zero, this fixes only part of PR5445
because the testcase is triggering one more bug.

llvm-svn: 88674
2009-11-13 18:49:59 +00:00
Dan Gohman f80dc08059 Don't let a noalias difference disrupt the tailcall optimization.
llvm-svn: 88672
2009-11-13 18:49:38 +00:00
David Greene a32c214b23 Remove duplicate APIs and state WRT spill objects.
llvm-svn: 87106
2009-11-13 14:42:06 +00:00
Rafael Espindola ff2c72b858 Distinguish "a," from "a". The first one splits into "a" + "" and the second one into
"a" + 0.

llvm-svn: 87084
2009-11-13 04:55:09 +00:00
Devang Patel 76a06074eb Revert r87059 for now. It is failing clang tests.
llvm-svn: 87070
2009-11-13 02:27:33 +00:00
Devang Patel 97f99fa554 Ignore nameless variables.
llvm-svn: 87069
2009-11-13 02:25:26 +00:00
Rafael Espindola 7c6854995d Switch to smallvector. Also fix issue with using unsigend for MaxSplit.
llvm-svn: 87068
2009-11-13 02:18:25 +00:00
Dale Johannesen 5f4eecf961 Adjust isConstantSplat to allow for big-endian targets.
PPC is such a target; make it work.

llvm-svn: 87060
2009-11-13 01:45:18 +00:00
Victor Hernandez 41e7648e09 Remove unnecessary llvm.dbg.declare bitcast
llvm-svn: 87059
2009-11-13 01:44:55 +00:00
Rafael Espindola d554e44092 Add a new split method to StringRef that puts the substrings in a vector.
llvm-svn: 87058
2009-11-13 01:24:40 +00:00
Jim Grosbach 46a524c3e8 Block renumbering
llvm-svn: 87056
2009-11-13 01:19:24 +00:00
Jim Grosbach 969910b3e8 use lower case for readability
llvm-svn: 87054
2009-11-13 01:17:22 +00:00
Daniel Dunbar 3f75f5ddcb Update test.
llvm-svn: 87049
2009-11-13 01:01:58 +00:00
David Greene 2f4c37425b Fix a bootstrap failure.
Provide special isLoadFromStackSlotPostFE and isStoreToStackSlotPostFE
interfaces to explicitly request checking for post-frame ptr elimination
operands.  This uses a heuristic so it isn't reliable for correctness.

llvm-svn: 87047
2009-11-13 00:29:53 +00:00
Owen Anderson e96b2111b1 Re-enable this code, since redundant PHIs are now being better nuked.
llvm-svn: 87042
2009-11-12 23:22:41 +00:00
Bill Wendling c781d7a072 Simplify code a bit
llvm-svn: 87040
2009-11-12 23:13:08 +00:00
Bill Wendling e412064c4c Refactor code that checks if it's a call to a "nounwind" function.
llvm-svn: 87036
2009-11-12 21:59:20 +00:00
Chris Lattner 5c89f4b4ef use isInstructionTriviallyDead, as pointed out by Duncan
llvm-svn: 87035
2009-11-12 21:58:18 +00:00
David Greene 033d65581d Do some cleanups suggested by Chris.
llvm-svn: 87034
2009-11-12 21:49:55 +00:00
Daniel Dunbar 5fb4e7553c StringRef(const char*) should not be used to turn null pointers into empty
strings.

llvm-svn: 87031
2009-11-12 21:26:11 +00:00
David Greene 7cf326aed4 Set the ReloadReuse AsmPrinter flag where appropriate.
llvm-svn: 87030
2009-11-12 21:07:54 +00:00
Daniel Dunbar c71bf467d5 Remove my Value.h build fix.
llvm-svn: 87029
2009-11-12 21:07:02 +00:00
David Greene f3ecd97cd8 Fix a build error by providing a missing enum value.
llvm-svn: 87028
2009-11-12 21:04:19 +00:00
David Greene be851acfb0 Make the MachineFunction argument of getFrameRegister const.
This also fixes a build error.

llvm-svn: 87027
2009-11-12 21:00:03 +00:00
David Greene 70fdd57dc1 Add hasLoadFromStackSlot and hasStoreToStackSlot to return whether a
machine instruction loads or stores from/to a stack slot.  Unlike
isLoadFromStackSlot and isStoreFromStackSlot, the instruction may be
something other than a pure load/store (e.g. it may be an arithmetic
operation with a memory operand).  This helps AsmPrinter determine when
to print a spill/reload comment.

This is only a hint since we may not be able to figure this out in all
cases.  As such, it should not be relied upon for correctness.

Implement for X86.  Return false by default for other architectures.

llvm-svn: 87026
2009-11-12 20:55:29 +00:00
Daniel Dunbar 22b75873b9 Attempt to unbreak LLVM build, David G. please check.
llvm-svn: 87025
2009-11-12 20:53:56 +00:00
Daniel Dunbar 5026bc3a86 Fix -Asserts warning.
llvm-svn: 87024
2009-11-12 20:53:43 +00:00
Bill Wendling e3ae25b3d8 If there's more than one function operand to a call instruction, be conservative
and don't assume that the call doesn't throw. It would be nice if there were a
way to determine which is the callee and which is a parameter. In practice, the
architecture we care about normally only have one operand for a call instruction
(x86 and arm).

llvm-svn: 87023
2009-11-12 20:51:53 +00:00
David Greene 1fbe054450 Add a bool flag to StackObjects telling whether they reference spill
slots.  The AsmPrinter will use this information to determine whether to
print a spill/reload comment.

Remove default argument values.  It's too easy to pass a wrong argument
value when multiple arguments have default values.  Make everything
explicit to trap bugs early.

Update all targets to adhere to the new interfaces..

llvm-svn: 87022
2009-11-12 20:49:22 +00:00
Benjamin Kramer 68e4945c03 Add compare_lower and equals_lower methods to StringRef. Switch all users of
StringsEqualNoCase (from StringExtras.h) to it.

llvm-svn: 87020
2009-11-12 20:36:59 +00:00
David Greene 81eb42d8ca Make FixedStackPseudoSourceValue a first-class PseudoSourceValue by
making it visible to clients and adding LLVM-style cast capability.
This will be used by AsmPrinter to determine when to emit spill comments
for an instruction.

llvm-svn: 87019
2009-11-12 20:25:07 +00:00
David Greene 815863f157 Add AsmPrinter comment flags to machine instructions so that AsmPrinter
can emit extra information in comments.

llvm-svn: 87018
2009-11-12 20:21:09 +00:00
David Greene 33935767b9 Add comment flags so AsmPrinter can output additional information when
emitting comments.  These flags carry semantic information not otherwise
easily derivable from the IR text.

llvm-svn: 87016
2009-11-12 20:13:34 +00:00
David Goodwin da83f7d58b Rename registers to break output dependencies in addition to anti-dependencies.
llvm-svn: 87015
2009-11-12 19:08:21 +00:00
Devang Patel 2904aa9f6e "Attach debug info with llvm instructions" mode was enabled a month ago. Now make it permanent and remove old way of inserting intrinsics to encode debug info for line number and scopes.
llvm-svn: 87014
2009-11-12 19:02:56 +00:00
Dan Gohman 29271e7585 Mark DBG_LABEL, EH_LABEL, and GC_LABEL as not-duplicable, since
they really are not duplicable.

llvm-svn: 87009
2009-11-12 18:36:19 +00:00
Benjamin Kramer 3628bdab14 Silence a warning on targets with unsigned chars.
llvm-svn: 87002
2009-11-12 17:59:45 +00:00
Jim Grosbach 5d57714e6c Update TB[BH] layout optimization. Add support for moving the target block
to directly follow the jump table. Move the layout changes to prior to any
constant island handling.

llvm-svn: 86999
2009-11-12 17:25:07 +00:00
Jim Grosbach 1025a4998b Clean up testcase a bit. Simplify case blocks and adjust switch instruction to not take an undefined value as input.
llvm-svn: 86997
2009-11-12 17:19:09 +00:00
Nuno Lopes f8fcac7470 fix crash in my previous patch
llvm-svn: 86987
2009-11-12 15:10:33 +00:00
Nuno Lopes 60d5b1cfdb implement shl, ashr, and lshr methods. shl is not fully implemented as it is quite tricky.
llvm-svn: 86986
2009-11-12 14:53:53 +00:00
Benjamin Kramer 5218176bc6 Fix typo in run line.
llvm-svn: 86984
2009-11-12 12:35:27 +00:00
Gabor Greif 13431c6cdf typo
llvm-svn: 86980
2009-11-12 09:44:17 +00:00
Chris Lattner eb9acbfb05 implement a nice little efficiency hack in the inliner. Since we're now
running IPSCCP early, and we run functionattrs interlaced with the inliner,
we often (particularly for small or noop functions) completely propagate
all of the information about a call to its call site in IPSSCP (making a call
dead) and functionattrs is smart enough to realize that the function is
readonly (because it is interlaced with inliner).

To improve compile time and make the inliner threshold more accurate, realize
that we don't have to inline dead readonly function calls.  Instead, just 
delete the call.  This happens all the time for C++ codes, here are some
counters from opt/llvm-ld counting the number of times calls were deleted vs
inlined on various apps:

Tramp3d opt:
  5033 inline                - Number of call sites deleted, not inlined
 24596 inline                - Number of functions inlined
llvm-ld:
  667 inline           - Number of functions deleted because all callers found
  699 inline           - Number of functions inlined

483.xalancbmk opt:
  8096 inline                - Number of call sites deleted, not inlined
 62528 inline                - Number of functions inlined
llvm-ld:
   217 inline           - Number of allocas merged together
  2158 inline           - Number of functions inlined

471.omnetpp:
  331 inline                - Number of call sites deleted, not inlined
 8981 inline                - Number of functions inlined
llvm-ld:
  171 inline           - Number of functions deleted because all callers found
  629 inline           - Number of functions inlined


Deleting a call is much faster than inlining it, and is insensitive to the
size of the callee. :)

llvm-svn: 86975
2009-11-12 07:56:08 +00:00
Evan Cheng 5d85a46f76 RegScavenger::enterBasicBlock should always reset register state.
llvm-svn: 86972
2009-11-12 07:49:10 +00:00
Evan Cheng 85a9f430e9 - Teach LSR to avoid changing cmp iv stride if it will create an immediate that
cannot be folded into target cmp instruction.
- Avoid a phase ordering issue where early cmp optimization would prevent the
  later count-to-zero optimization.
- Add missing checks which could cause LSR to reuse stride that does not have
  users.
- Fix a bug in count-to-zero optimization code which failed to find the pre-inc
  iv's phi node.
- Remove, tighten, loosen some incorrect checks disable valid transformations.
- Quite a bit of code clean up.

llvm-svn: 86969
2009-11-12 07:35:05 +00:00
Evan Cheng e129dd311e Use table to separate opcode from operands.
llvm-svn: 86965
2009-11-12 07:16:34 +00:00
Evan Cheng 15b80e4a9f isLegalICmpImmediate should take a signed integer; code clean up.
llvm-svn: 86964
2009-11-12 07:13:11 +00:00
Oscar Fuentes 700205c49f CMake: Hopefully unbreak the build by mimicking the changes on the
other build system about the new C_INCLUDE_DIRS configure option.

llvm-svn: 86960
2009-11-12 06:48:09 +00:00
Rafael Espindola d95960be25 Add the --with-c-include-dirs to llvm's configure.
The clang patch is next.

llvm-svn: 86955
2009-11-12 05:46:09 +00:00
Oscar Fuentes d533e1d46b CMake: Pass -lm to check_symbol_exists for detecting several math
functions like floorf, ceilf, ... Add test for detecting nearbyintf.

This change was prompted by test/Transforms/SimplifyLibCalls/floor.ll

llvm-svn: 86954
2009-11-12 05:36:09 +00:00