Richard Sandiford [Tue, 1 Oct 2013 14:31:11 +0000 (14:31 +0000)]
[SystemZ] Optimize 32-bit FPR<->GPR moves for z196 and above
Floats are stored in the high 32 bits of an FPR, and the only GPR<->FPR
transfers are full-register transfers. This patch optimizes GPR<->FPR
float transfers when the high word of a GPR is directly accessible.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191764
91177308-0d34-0410-b5e6-
96231b3b80d8
Tareq A. Siraj [Tue, 1 Oct 2013 14:28:18 +0000 (14:28 +0000)]
Add non-blocking Wait() for launched processes
- New ProcessInfo class to encapsulate information about child processes.
- Generalized the Wait() to support non-blocking wait on child processes.
- ExecuteNoWait() now returns a ProcessInfo object with information about
the launched child. Users will be able to use this object to
perform non-blocking wait.
- ExecuteNoWait() now accepts an ExecutionFailed param that tells if execution
failed or not.
These changes will allow users to implement basic process parallel
tools.
Differential Revision: http://llvm-reviews.chandlerc.com/D1728
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191763
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 14:20:41 +0000 (14:20 +0000)]
[SystemZ] Allow integer AND involving high words
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191762
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 14:08:44 +0000 (14:08 +0000)]
[SystemZ] Allow integer XOR involving high words
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191759
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Tue, 1 Oct 2013 13:32:03 +0000 (13:32 +0000)]
Remove several unused variables.
Patch by Alp Toker.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191757
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 13:22:41 +0000 (13:22 +0000)]
[SystemZ] Allow integer OR involving high words
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191755
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 13:18:56 +0000 (13:18 +0000)]
[SystemZ] Allow integer insertions with a high-word destination
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191753
91177308-0d34-0410-b5e6-
96231b3b80d8
Sylvestre Ledru [Tue, 1 Oct 2013 13:17:09 +0000 (13:17 +0000)]
Fix a typo in the documentation. Thanks to Diana Vasile for the patch
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191752
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 13:10:16 +0000 (13:10 +0000)]
[SystemZ] Allow selects with a high-word destination
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191751
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 13:02:28 +0000 (13:02 +0000)]
[SystemZ] Add patterns to load a constant into a high word (IIHF)
Similar to low words, we can use the shorter LLIHL and LLIHH if it turns
out that the other half of the GR64 isn't live.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191750
91177308-0d34-0410-b5e6-
96231b3b80d8
Joey Gouly [Tue, 1 Oct 2013 13:01:10 +0000 (13:01 +0000)]
[ARM] Remove an unused function from the disassembler.
Pointed out by Joerg.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191749
91177308-0d34-0410-b5e6-
96231b3b80d8
Matheus Almeida [Tue, 1 Oct 2013 12:53:00 +0000 (12:53 +0000)]
Test commit. Updated comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191748
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 12:49:07 +0000 (12:49 +0000)]
[SystemZ] Add register zero extensions involving at least one high word
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191746
91177308-0d34-0410-b5e6-
96231b3b80d8
Joerg Sonnenberger [Tue, 1 Oct 2013 12:42:48 +0000 (12:42 +0000)]
Remove empty directory.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191745
91177308-0d34-0410-b5e6-
96231b3b80d8
Joey Gouly [Tue, 1 Oct 2013 12:39:11 +0000 (12:39 +0000)]
[ARM] Introduce the 'sevl' instruction in ARMv8.
This also removes the restriction on the immediate field of the 'hint'
instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191744
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 12:22:49 +0000 (12:22 +0000)]
[SystemZ] Add truncating high-word stores (STCH and STHH)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191743
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 12:19:08 +0000 (12:19 +0000)]
[SystemZ] Add zero-extending high-word loads (LLCH and LLHH)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191742
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 1 Oct 2013 12:17:11 +0000 (12:17 +0000)]
SCEVExpander: Fix a regression I introduced by to eagerly adding RAII objects.
PR17425.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191741
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 12:11:47 +0000 (12:11 +0000)]
[SystemZ] Add sign-extending high-word loads (LBH and LHH)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191740
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 11:26:28 +0000 (11:26 +0000)]
[SystemZ] Use upper words of GR64s for codegen
This just adds the basics necessary for allocating the upper words to
virtual registers (move, load and store). The move support is parameterised
in a way that makes it easy to handle zero extensions, but the associated
zero-extend patterns are added by a later patch.
The easiest way of testing this seemed to be add a new "h" register
constraint for high words. I don't expect the constraint to be useful
in real inline asms, but it should work, so I didn't try to hide it
behind an option.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191739
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 10:31:04 +0000 (10:31 +0000)]
[SystemZ] Reapply: Add definitions of LFH and STFH
Originally committed as r191661, but reverted because it changed the matching
order of comparisons on some hosts. That should have been fixed by r191735.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191738
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Tue, 1 Oct 2013 10:22:35 +0000 (10:22 +0000)]
[mips][msa] Added support for matching mod_[us] from normal IR (i.e. not intrinsics)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191737
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Tue, 1 Oct 2013 09:49:01 +0000 (09:49 +0000)]
Fix pattern sort in DAGISelEmitter.cpp
The old code skipped one of the sorting criteria if either pattern had
no types. This could lead to cycles of the form X < Y, Y < Z, Z < X.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191735
91177308-0d34-0410-b5e6-
96231b3b80d8
Vladimir Medic [Tue, 1 Oct 2013 09:48:56 +0000 (09:48 +0000)]
This patch adds aliases for Mips sub instruction with immediate operands. Corresponding test cases are added.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191734
91177308-0d34-0410-b5e6-
96231b3b80d8
Elena Demikhovsky [Tue, 1 Oct 2013 08:38:02 +0000 (08:38 +0000)]
AVX-512: Added X86vzmovl patterns
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191733
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Tue, 1 Oct 2013 07:10:28 +0000 (07:10 +0000)]
Remove 0 as a valid encoding for the m-mmmm field.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191732
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Tue, 1 Oct 2013 06:56:57 +0000 (06:56 +0000)]
Remove unneeded fields from disassembler internal instruction format.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191731
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Tue, 1 Oct 2013 03:48:26 +0000 (03:48 +0000)]
BEXTR should be defined to take same type for bother operands.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191728
91177308-0d34-0410-b5e6-
96231b3b80d8
Tom Stellard [Tue, 1 Oct 2013 02:09:00 +0000 (02:09 +0000)]
SelectionDAG: Clarify comments from r191600
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191724
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Kaylor [Tue, 1 Oct 2013 01:48:36 +0000 (01:48 +0000)]
Tests for MCJIT multiple module support
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191723
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Kaylor [Tue, 1 Oct 2013 01:47:35 +0000 (01:47 +0000)]
Adding multiple module support for MCJIT.
Tests to follow.
PIC with small code model and EH frame handling will not work with multiple modules. There are also some rough edges to be smoothed out for remote target support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191722
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Tue, 1 Oct 2013 00:43:36 +0000 (00:43 +0000)]
Add the DW_AT_GNU_ranges_base attribute if we've emitted any ranges
into the debug_ranges section.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191721
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Tue, 1 Oct 2013 00:43:31 +0000 (00:43 +0000)]
Update comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191720
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Tue, 1 Oct 2013 00:01:14 +0000 (00:01 +0000)]
Fix code duplication
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191716
91177308-0d34-0410-b5e6-
96231b3b80d8
Preston Gurd [Mon, 30 Sep 2013 23:51:22 +0000 (23:51 +0000)]
Forgot to add a break statement.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191715
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Mon, 30 Sep 2013 23:31:55 +0000 (23:31 +0000)]
Use CHECK-LABEL
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191713
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Mon, 30 Sep 2013 23:31:50 +0000 (23:31 +0000)]
Reuse variable
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191712
91177308-0d34-0410-b5e6-
96231b3b80d8
Preston Gurd [Mon, 30 Sep 2013 23:18:42 +0000 (23:18 +0000)]
The X86FixupLEAs pass for Intel Atom must not call convertToThreeAddress
on ADD16rr opcodes, if src1 != src, since that would cause
convertToThreeAddress to try to create a virtual register. This is not
permitted after register allocation, which is when the X86FixupLEAs pass
runs.
This patch fixes PR16785.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191711
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Mon, 30 Sep 2013 23:14:16 +0000 (23:14 +0000)]
The DW_AT_GNU_pubnames/pubtypes attributes are actually form
SEC_OFFSET from the beginning of the section so go ahead and emit
a label at the beginning of each one.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191710
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Mon, 30 Sep 2013 21:55:01 +0000 (21:55 +0000)]
Add llvm-readobj to the list of programs to find in the freshly built
toolchain.
Patch by Richard Pennington.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191706
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Mon, 30 Sep 2013 21:23:03 +0000 (21:23 +0000)]
Fix getOrInsertGlobal dropping the address space.
Currently it will insert an illegal bitcast.
Arguably, the address space argument should be
added for the creation case.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191702
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Mon, 30 Sep 2013 21:11:01 +0000 (21:11 +0000)]
Use right address space size in InstCombineCompares
The test's output doesn't change, but this ensures
this is actually hit with a different address space.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191701
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Mon, 30 Sep 2013 21:06:18 +0000 (21:06 +0000)]
Constant fold ptrtoint + compare with address spaces
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191699
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Mon, 30 Sep 2013 19:42:10 +0000 (19:42 +0000)]
Debug Info: constify and rename from generateRef to getRef.
No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191696
91177308-0d34-0410-b5e6-
96231b3b80d8
Anders Waldenborg [Mon, 30 Sep 2013 19:11:32 +0000 (19:11 +0000)]
llvm-c: use typedef for function pointers
This makes it consistent with other function pointers used in llvm-c
Differential Revision: http://llvm-reviews.chandlerc.com/D1712
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191693
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Mon, 30 Sep 2013 18:50:51 +0000 (18:50 +0000)]
[ARM] Fix Thumb(-2) diagnostic tests.
Changing the diagnostic message for out of range branch targets in 191686 broke the tests.
The diagnostic message for out of range branch targets was changed to be more consistent with the other diagnostics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191691
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Mon, 30 Sep 2013 18:17:55 +0000 (18:17 +0000)]
TBAA: update tbaa format from scalar format to struct-path aware format.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191690
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Mon, 30 Sep 2013 18:17:35 +0000 (18:17 +0000)]
TBAA: remove !tbaa from testing cases when they are not needed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191689
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Mon, 30 Sep 2013 18:05:18 +0000 (18:05 +0000)]
[mips][msa] Direct Object Emission for I8 instructions.
This patch adds Direct Object Emission support for I8 instructions: andi.b, bmnzi.b, bmzi.b, bseli.b, nori.b, ori.b, shf.{b,h,w} and xori.b.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191688
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Mon, 30 Sep 2013 17:58:07 +0000 (17:58 +0000)]
[mips][msa] Direct Object Emission for I5 instructions.
This patch adds Direct Object Emission support for I5 instructions: addvi.{b,h,w,d}, ceqi.{b,h,w,d}, clei_s.{b,h,w,d}, clei_u.{b,h,w,d}, clti_s.{b,h,w,d}, clti_u.{b,h,w,d}, maxi_s.{b,h,w,d}, maxi_u.{b,h,w,d}, mini_s.{b,h,w,d}, mini_u.{b,h,w,d}, subvi.{b,h,w,d}.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191687
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Mon, 30 Sep 2013 17:57:30 +0000 (17:57 +0000)]
[ARM] Clean up ARMAsmParser::validateInstruction().
Fix some LLVM Coding Standards violations.
No changes in functionality.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191686
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Mon, 30 Sep 2013 17:52:33 +0000 (17:52 +0000)]
[mips][msa] Direct Object Emission for 2R instructions.
This patch adds Direct Object Emission support for 2R instructions: nloc.{b,h,w}, nlzc.{b,h,w}, pcnt.{b,w,d}.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191685
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Mon, 30 Sep 2013 17:43:04 +0000 (17:43 +0000)]
[PATCH 1/4] [mips][msa] Source register of FILL instructions is GPR
and not an MSA register
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191684
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Mon, 30 Sep 2013 17:31:26 +0000 (17:31 +0000)]
[ARM] Use FileCheck instead of grep for ARM LDRD negative tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191683
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 30 Sep 2013 16:39:19 +0000 (16:39 +0000)]
Move command line options to the users of libLTO. Fixes --enable-shared build.
Patch by Richard Sandiford.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191680
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 30 Sep 2013 16:32:51 +0000 (16:32 +0000)]
Revert "Enable building LTO on WIN32."
This reverts commit r191670.
It was causing build failures on the msvc bots:
http://bb.pgr.jp/builders/ninja-clang-i686-msc17-R/builds/5166/steps/compile/logs/stdio
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191679
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Mon, 30 Sep 2013 16:11:48 +0000 (16:11 +0000)]
[ARM] Assembler: ARM LDRD with writeback requires the base register to be different from the destination registers.
See ARM ARM A8.8.72.
Violating this constraint results in unpredictable behavior.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191678
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnold Schwaighofer [Mon, 30 Sep 2013 15:56:34 +0000 (15:56 +0000)]
Swift model: Fix uop description on some writes
Those writes really need two/three uops.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191677
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 30 Sep 2013 15:52:50 +0000 (15:52 +0000)]
BoundsChecking: Fix refacto.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191676
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 30 Sep 2013 15:40:17 +0000 (15:40 +0000)]
Convert manual insert point restores to the new RAII object.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191675
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 30 Sep 2013 15:39:59 +0000 (15:39 +0000)]
InstCombine: Replace manual fast math flag copying with the new IRBuilder RAII helper.
Defines away the issue where cast<Instruction> would fail because constant
folding happened. Also slightly cleaner.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191674
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 30 Sep 2013 15:39:48 +0000 (15:39 +0000)]
IRBuilder: Add RAII objects to reset insertion points or fast math flags.
Inspired by the object from the SLPVectorizer. This found a minor bug in the
debug loc restoration in the vectorizer where the location of a following
instruction was attached instead of the location from the original instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191673
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 30 Sep 2013 15:39:27 +0000 (15:39 +0000)]
IRBuilder: Move fast math flags to IRBuilderBase.
They don't depend on the templated stuff.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191672
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnold Schwaighofer [Mon, 30 Sep 2013 15:28:56 +0000 (15:28 +0000)]
IfConverter: Use TargetSchedule for instruction latencies
For targets that have instruction itineraries this means no change. Targets
that move over to the new schedule model will use be able the new schedule
module for instruction latencies in the if-converter (the logic is such that if
there is no itineary we will use the new sched model for the latencies).
Before, we queried "TTI->getInstructionLatency()" for the instruction latency
and the extra prediction cost. Now, we query the TargetSchedule abstraction for
the instruction latency and TargetInstrInfo for the extra predictation cost. The
TargetSchedule abstraction will internally call "TTI->getInstructionLatency" if
an itinerary exists, otherwise it will use the new schedule model.
ATTENTION: Out of tree targets!
(I will also send out an email later to LLVMDev)
This means, if your target implements
unsigned getInstrLatency(const InstrItineraryData *ItinData,
const MachineInstr *MI,
unsigned *PredCost);
and returns a value for "PredCost", you now also need to implement
unsigned getPredictationCost(const MachineInstr *MI);
(if your target uses the IfConversion.cpp pass)
radar://
15077010
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191671
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 30 Sep 2013 15:28:14 +0000 (15:28 +0000)]
Enable building LTO on WIN32.
Enable building the LTO library (.lib and.dll) and llvm-lto.exe on Windows with
MSVC and Mingw as well as re-enabling the associated test.
Patch by Greg Bedwell!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191670
91177308-0d34-0410-b5e6-
96231b3b80d8
Joey Gouly [Mon, 30 Sep 2013 14:18:35 +0000 (14:18 +0000)]
Fix a bug in InstCombine where it attempted to cast a Value* to an Instruction*
when it was actually a Constant*.
There are quite a few other casts to Instruction that might have the same problem,
but this is the only one I have a test case for.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191668
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Mon, 30 Sep 2013 13:04:22 +0000 (13:04 +0000)]
[ARM] Assembler: Add more negative tests for ARM LDRD.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191664
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Mon, 30 Sep 2013 12:01:35 +0000 (12:01 +0000)]
[SystemZ] Revert r191661: Add definitions of LFH and STFH
For some reason, adding definitions for these load and store
instructions changed whether some of the build bots matched
comparisons as signed or unsigned.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191663
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Mon, 30 Sep 2013 10:50:33 +0000 (10:50 +0000)]
[SystemZ] Add definitions of LFH and STFH
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191661
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Mon, 30 Sep 2013 10:45:16 +0000 (10:45 +0000)]
[SystemZ] Add GRH32 for the high word of a GR64
The only thing this does on its own is make the definitions of RISB[HL]G
a bit more precise. Those instructions are only used by the MC layer at
the moment, so no behavioral change is intended. The class is needed by
later patches though.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191660
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Mon, 30 Sep 2013 10:28:35 +0000 (10:28 +0000)]
[SystemZ] Rename subregs and add subreg_h32
Use subreg_hNN and subreg_lNN for the high and low NN bits of a register.
List the low registers first, so that subreg_l32 also means the low 32
bits of a 128-bit register.
Floats are stored in the upper 32 bits of a 64-bit register, so they
should use subreg_h32 rather than subreg_l32.
No behavioral change intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191659
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Mon, 30 Sep 2013 09:35:37 +0000 (09:35 +0000)]
[mips] Fix a broken link to mips.com in the documentation.
It now points to the equivalent page on imgtec.com
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191658
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Mon, 30 Sep 2013 08:54:17 +0000 (08:54 +0000)]
[SystemZ] Add change missing from previous commit
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191656
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Mon, 30 Sep 2013 08:48:38 +0000 (08:48 +0000)]
[SystemZ] Rename 32-bit GPR registers
I'm about to add support for high-word operations, so it seemed better
for the low-word registers to have names like R0L rather than R0W.
No behavioral change intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191655
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Mon, 30 Sep 2013 06:23:19 +0000 (06:23 +0000)]
Filter out repeated sections from the X86 disassembler modRMTable. Saves about ~43K from a released build. Unfortunately the disassembler tables are still upwards of 800K.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191652
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Mon, 30 Sep 2013 02:50:51 +0000 (02:50 +0000)]
Add a few more FMA4 disassembler test cases to match the scalar set with regards to combinations of L and W-bits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191650
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Mon, 30 Sep 2013 02:46:36 +0000 (02:46 +0000)]
Various x86 disassembler fixes.
Add VEX_LIG to scalar FMA4 instructions.
Use VEX_LIG in some of the inheriting checks in disassembler table generator.
Make use of VEX_L_W, VEX_L_W_XS, VEX_L_W_XD contexts.
Don't let VEX_L_W, VEX_L_W_XS, VEX_L_W_XD, VEX_L_W_OPSIZE inherit from their non-L forms unless VEX_LIG is set.
Let VEX_L_W, VEX_L_W_XS, VEX_L_W_XD, VEX_L_W_OPSIZE inherit from all of their non-L or non-W cases.
Increase ranking on VEX_L_W, VEX_L_W_XS, VEX_L_W_XD, VEX_L_W_OPSIZE so they get chosen over non-L/non-W forms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191649
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sun, 29 Sep 2013 19:39:13 +0000 (19:39 +0000)]
ObjectSizeOffsetEvaluator: Don't run into infinite recursion if we have a cyclic GEP.
Those can occur in dead code. PR17402.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191644
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sun, 29 Sep 2013 19:39:02 +0000 (19:39 +0000)]
Remove an old workaround for a compiler that EOL'd years ago.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191643
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sun, 29 Sep 2013 11:29:20 +0000 (11:29 +0000)]
Plug a memory leak in a unit test. Stack allocation is sufficient here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191638
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sun, 29 Sep 2013 11:24:02 +0000 (11:24 +0000)]
Deallocate type units when destroying a DWARFContext.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191637
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sun, 29 Sep 2013 11:18:56 +0000 (11:18 +0000)]
Allocate AtomicSDNode operands in SelectionDAG's allocator to stop leakage.
SDNode destructors are never called. As an optimization use AtomicSDNode's
internal storage if we have a small number of operands.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191636
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Sun, 29 Sep 2013 08:35:51 +0000 (08:35 +0000)]
Revert accidental commit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191633
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Sun, 29 Sep 2013 08:33:34 +0000 (08:33 +0000)]
Change type of XOP flag in code emitters to a bool. Remove a some unneeded cases from switch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191632
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Sun, 29 Sep 2013 06:31:18 +0000 (06:31 +0000)]
Add comments for XOPA map introduced with TBM instructions.a
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191630
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sat, 28 Sep 2013 21:27:49 +0000 (21:27 +0000)]
Add a test that large offsets on GEPs on 32 bits targets are handled correctly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191628
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Sat, 28 Sep 2013 18:17:10 +0000 (18:17 +0000)]
Enable libcxx as part of the top level CMake build when it is checked
out in projects. This appears to be working on my system, and I will be
watching build bots to see if there are any issues on other platforms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191624
91177308-0d34-0410-b5e6-
96231b3b80d8
Robert Wilhelm [Sat, 28 Sep 2013 13:42:22 +0000 (13:42 +0000)]
Even more spelling fixes for "instruction".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191611
91177308-0d34-0410-b5e6-
96231b3b80d8
Robert Wilhelm [Sat, 28 Sep 2013 11:46:15 +0000 (11:46 +0000)]
Fix spelling intruction -> instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191610
91177308-0d34-0410-b5e6-
96231b3b80d8
Tom Stellard [Sat, 28 Sep 2013 03:10:17 +0000 (03:10 +0000)]
SelectionDAG: Silence unused variable warning on release builds
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191604
91177308-0d34-0410-b5e6-
96231b3b80d8
Tom Stellard [Sat, 28 Sep 2013 02:50:50 +0000 (02:50 +0000)]
R600: Fix handling of NAN in comparison instructions
We were completely ignoring the unorder/ordered attributes of condition
codes and also incorrectly lowering seto and setuo.
Reviewed-by: Vincent Lejeune<vljn at ovi.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191603
91177308-0d34-0410-b5e6-
96231b3b80d8
Tom Stellard [Sat, 28 Sep 2013 02:50:43 +0000 (02:50 +0000)]
SelectionDAG: Improve legalization of SELECT_CC with illegal condition codes
SelectionDAG will now attempt to inverse an illegal conditon in order to
find a legal one and if that doesn't work, it will attempt to swap the
operands using the inverted condition.
There are no new test cases for this, but a nubmer of the existing R600
tests hit this path.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191602
91177308-0d34-0410-b5e6-
96231b3b80d8
Tom Stellard [Sat, 28 Sep 2013 02:50:38 +0000 (02:50 +0000)]
SelectionDAG: Try to expand all condition codes using getCCSwappedOperands()
This is useful for targets like R600, which only support GT, GE, NE, and EQ
condition codes as it removes the need to handle unsupported condition
codes in target specific code.
There are no tests with this commit, but R600 has been updated to take
advantage of this new feature, so its existing selectcc tests are now
testing the swapped operands path.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191601
91177308-0d34-0410-b5e6-
96231b3b80d8
Tom Stellard [Sat, 28 Sep 2013 02:50:32 +0000 (02:50 +0000)]
SelectionDAG: Clean up LegalizeSetCCCondCode() function
Interpreting the results of this function is not very intuitive, so I
cleaned it up to make it more clear whether or not a SETCC op was
legalized and how it was legalized (either by swapping LHS and RHS or
replacing with AND/OR).
This patch does change functionality in the LHS and RHS swapping case,
but unfortunately there are no in-tree tests for this. However, this
patch is a prerequisite for R600 to take advantage of the LHS and RHS
swapping, so tests will be added in subsequent commits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191600
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Sat, 28 Sep 2013 01:35:07 +0000 (01:35 +0000)]
MipsMachineFunction.cpp: Add missing #include <raw_ostream.h>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191597
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 28 Sep 2013 01:08:00 +0000 (01:08 +0000)]
Fix typo
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191595
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Sat, 28 Sep 2013 00:22:27 +0000 (00:22 +0000)]
AutoUpgrade: upgrade from scalar TBAA format to struct-path aware TBAA format.
We treat TBAA tags as struct-path aware TBAA format when the first operand
is a MDNode and the tag has 3 or more operands.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191593
91177308-0d34-0410-b5e6-
96231b3b80d8
Akira Hatanaka [Sat, 28 Sep 2013 00:12:32 +0000 (00:12 +0000)]
[mips] Make sure loads from lazy-binding entries do not get CSE'd or hoisted out
of loops.
Previously, two consecutive calls to function "func" would result in the
following sequence of instructions:
1. load $16, %got(func)($gp) // load address of lazy-binding stub.
2. move $25, $16
3. jalr $25 // jump to lazy-binding stub.
4. nop
5. move $25, $16
6. jalr $25 // jump to lazy-binding stub again.
With this patch, the second call directly jumps to func's address, bypassing
the lazy-binding resolution routine:
1. load $25, %got(func)($gp) // load address of lazy-binding stub.
2. jalr $25 // jump to lazy-binding stub.
3. nop
4. load $25, %got(func)($gp) // load resolved address of func.
5. jalr $25 // directly jump to func.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191591
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Fri, 27 Sep 2013 22:59:21 +0000 (22:59 +0000)]
TBAA: try to fix the dragonegg bots.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191585
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Fri, 27 Sep 2013 22:50:48 +0000 (22:50 +0000)]
Unify conditionals and reformat.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191582
91177308-0d34-0410-b5e6-
96231b3b80d8