Akira Hatanaka [Fri, 27 Sep 2013 19:51:35 +0000 (19:51 +0000)]
[mips] Rewrite MipsTargetLowering::getAddr functions as template functions.
No intended functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191546
91177308-0d34-0410-b5e6-
96231b3b80d8
Yunzhong Gao [Fri, 27 Sep 2013 18:38:42 +0000 (18:38 +0000)]
Adding intrinsics to the llvm backend for TBM instruction set.
Phabricator code review is located here: http://llvm-reviews.chandlerc.com/D1750
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191539
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Fri, 27 Sep 2013 18:34:27 +0000 (18:34 +0000)]
TBAA: handle scalar TBAA format and struct-path aware TBAA format.
Remove the command line argument "struct-path-tbaa" since we should not depend
on command line argument to decide which format the IR file is using. Instead,
we check the first operand of the tbaa tag node, if it is a MDNode, we treat
it as struct-path aware TBAA format, otherwise, we treat it as scalar TBAA
format.
When clang starts to use struct-path aware TBAA format no matter whether
struct-path-tbaa is no, and we can auto-upgrade existing bc files, the support
for scalar TBAA format can be dropped.
Existing testing cases are updated to use the struct-path aware TBAA format.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191538
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Fri, 27 Sep 2013 15:30:25 +0000 (15:30 +0000)]
Transforms: Use getFirstNonPHI to set the insertion point for PHIs
We were previously using getFirstInsertionPt to insert PHI
instructions when vectorizing, but getFirstInsertionPt also skips past
landingpads, causing this to generate invalid IR.
We can avoid this issue by using getFirstNonPHI instead.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191526
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Fri, 27 Sep 2013 15:29:20 +0000 (15:29 +0000)]
[SystemZ] Rein back the use of block operations
The backend tries to use block operations like MVC, NC, OC and XC for
simple scalar operations. For correctness reasons, it rejects any case
in which the regions might partially overlap. However, for performance
reasons, it should also reject cases where the regions might be equal,
since the instruction might then not use the fast path.
This fixes a performance regression seen in bzip2. We may want to limit
the optimisation even more in future, or even remove it entirely, but I'll
try with this for now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191525
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Fri, 27 Sep 2013 15:14:04 +0000 (15:14 +0000)]
[SystemZ] Improve handling of PC-relative addresses
The backend previously folded offsets into PC-relative addresses
whereever possible. That's the right thing to do when the address
can be used directly in a PC-relative memory reference (using things
like LRL). But if we have a register-based memory reference and need
to load the PC-relative address separately, it's better to use an anchor
point that could be shared with other accesses to the same area of the
variable.
Fixes a FIXME.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191524
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 13:36:54 +0000 (13:36 +0000)]
[mips][msa] Implemented insert.d intrinsic.
This intrinsic is lowered into an equivalent INSERT_VECTOR_ELT which is
further lowered into a sequence of insert.w's on MIPS32.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191521
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Fri, 27 Sep 2013 13:28:17 +0000 (13:28 +0000)]
ARM: Teach assembler to enforce constraints for ARM LDRD destination register operands.
As specified in A8.8.72/A8.8.73/A8.8.74 in the ARM ARM, all variants of the ARM LDRD instruction have the following two constraints:
LDRD<c> <Rt>, <Rt2>, ...
(a) Rt must be even-numbered and not r14
(b) Rt2 must be R(t+1)
If those two constraints are not met the result of executing the instruction will be unpredictable.
Constraint (b) was already enforced, this commit adds support for constraint (a).
Fixes rdar://
14479793.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191520
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 13:20:41 +0000 (13:20 +0000)]
[mips][msa] Implemented fill.d intrinsic.
This intrinsic is lowered into an equivalent BUILD_VECTOR which is further
lowered into a sequence of insert.w's on MIPS32.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191519
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 13:04:21 +0000 (13:04 +0000)]
[mips][msa] Implemented copy_[us].d intrinsic.
This intrinsic is lowered into equivalent copy_s.w instructions during
legalization.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191518
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 12:45:08 +0000 (12:45 +0000)]
[mips][msa] Rename arguments to MSA_INSERT_DESC_BASE to better match their expected values.
No functional change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191517
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 12:31:32 +0000 (12:31 +0000)]
[mips][msa] Implemented insert_vector_elt for v4f32 and v2f64.
For v4f32 and v2f64, INSERT_VECTOR_ELT is matched by a pseudo-insn which is
later expanded to appropriate insve.[wd] insns.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191515
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 12:17:32 +0000 (12:17 +0000)]
[mips][msa] Implemented extract_vector_elt for v4f32 or v2f64
For v4f32 and v2f64, EXTRACT_VECTOR_ELT is matched by a pseudo-insn which may
be expanded to subregister copies and/or instructions as appropriate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191514
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrea Di Biagio [Fri, 27 Sep 2013 12:13:58 +0000 (12:13 +0000)]
Remove superfluous comment accidentally checked-in.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191513
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 12:03:51 +0000 (12:03 +0000)]
[mips][msa] Added support for MSA registers to copyPhysReg
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191512
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 11:48:57 +0000 (11:48 +0000)]
[mips][msa] Added support for matching splati from normal IR (i.e. not intrinsics)
Updated some of the vshf since they (correctly) emit splati's now
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191511
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrea Di Biagio [Fri, 27 Sep 2013 11:37:05 +0000 (11:37 +0000)]
Re-apply the change from r191393 with fix for pr17380.
This change fixes the problem reported in pr17380 and re-add the dagcombine
transformation ensuring that the value types are always legal if the
transformation is triggered after Legalization took place.
Added the test case from pr17380.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191509
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 10:42:22 +0000 (10:42 +0000)]
[mips][msa] Added MSA.txt to describe instruction selection quirks.
This file contains notes about the instruction selection for MSA. For example,
it notes that ilvl.d is cannot be selected because ilvev.d covers the same
cases and is selected instead of ilvl.d.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191507
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Fri, 27 Sep 2013 10:38:11 +0000 (10:38 +0000)]
Fix comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191505
91177308-0d34-0410-b5e6-
96231b3b80d8
Tilmann Scheller [Fri, 27 Sep 2013 10:30:18 +0000 (10:30 +0000)]
ARM: Teach assembler to enforce constraint for Thumb2 LDRD (literal/immediate) destination register operands.
LDRD<c> <Rt>, <Rt2>, <label>
LDRD<c> <Rt>, <Rt2>, [<Rn>{, #+/-<imm>}]
LDRD<c> <Rt>, <Rt2>, [<Rn>], #+/-<imm>
LDRD<c> <Rt>, <Rt2>, [<Rn>, #+/-<imm>]!
As specified in A8.8.72/A8.8.73 in the ARM ARM, the T1 encoding has a constraint which enforces that Rt != Rt2.
If this constraint is not met the result of executing the instruction will be unpredictable.
Fixes rdar://
14479780.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191504
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 10:25:41 +0000 (10:25 +0000)]
[mips][msa] Tidy up
lowerMSABinaryIntr, lowerMSABinaryImmIntr, lowerMSABranchIntr,
and lowerMSAUnaryIntr were trivially small functions. Inlined them into
their callers.
lowerMSASplat now takes its callers SDLoc instead of making a new one.
No functional change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191503
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 10:08:31 +0000 (10:08 +0000)]
[mips][msa] MSA requires FR=1 mode (64-bit FPU register file). Report fatal error when using it in FR=0 mode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191498
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 09:44:59 +0000 (09:44 +0000)]
[mips][msa] Expand all truncstores and loadexts for MSA as well as DSP
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191496
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 27 Sep 2013 09:25:29 +0000 (09:25 +0000)]
[mips][msa] Added missing check in performSRACombine
Reviewers: jacksprat, dsanders
Reviewed By: dsanders
Differential Revision: http://llvm-reviews.chandlerc.com/D1755
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191495
91177308-0d34-0410-b5e6-
96231b3b80d8
Puyan Lotfi [Fri, 27 Sep 2013 07:36:10 +0000 (07:36 +0000)]
First check in. Modified a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191491
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Fri, 27 Sep 2013 07:20:47 +0000 (07:20 +0000)]
Put HasAVX512 predicate on some patterns to properly disable them when AVX512 isn't enabled. Currently it works simply because the SSE and AVX version of the same patterns are checked first in the DAG isel table.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191490
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Fri, 27 Sep 2013 07:16:24 +0000 (07:16 +0000)]
Switch HasAVX to UseAVX in one spot to ensure that AVX512 form of VINSERTPS is used in AVX512 mode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191489
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Fri, 27 Sep 2013 07:11:17 +0000 (07:11 +0000)]
Removal some duplicate patterns.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191488
91177308-0d34-0410-b5e6-
96231b3b80d8
Yunzhong Gao [Fri, 27 Sep 2013 01:44:23 +0000 (01:44 +0000)]
Fixing Intel format of the vshufpd instruction.
Phabricator code review is located at: http://llvm-reviews.chandlerc.com/D1759
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191481
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Fri, 27 Sep 2013 01:29:36 +0000 (01:29 +0000)]
Revert "llvm-objdump: Dump COFF import table if -private-headers option is given."
This reverts commit r191472 because it's failing on BE machine.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191480
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Fri, 27 Sep 2013 00:53:07 +0000 (00:53 +0000)]
Fix another -Wnon-pod-varargs error in r191472.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191474
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Fri, 27 Sep 2013 00:20:53 +0000 (00:20 +0000)]
Fix -Wnon-pod-varargs error in r191472.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191473
91177308-0d34-0410-b5e6-
96231b3b80d8
Rui Ueyama [Fri, 27 Sep 2013 00:07:01 +0000 (00:07 +0000)]
llvm-objdump: Dump COFF import table if -private-headers option is given.
This is a patch to add capability to llvm-objdump to dump COFF Import Table
entries, so that we can write tests for LLD checking Import Table contents.
llvm-objdump did not print anything but just file name if the format is COFF
and -private-headers option is given. This is a patch adds capability for
dumping DLL Import Table, which is specific to the COFF format.
In this patch I defined a new iterator to iterate over import table entries.
Also added a few functions to COFFObjectFile.cpp to access fields of the entry.
Differential Revision: http://llvm-reviews.chandlerc.com/D1719
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191472
91177308-0d34-0410-b5e6-
96231b3b80d8
Adrian Prantl [Thu, 26 Sep 2013 23:37:11 +0000 (23:37 +0000)]
MCParser/Debug info: Accept line number 0 as a legitimate value, since
CFE produces it to indicate artificial locations.
c.f.: DWARF standard, Table 6.2:
line -- An unsigned integer indicating a source line number. Lines are numbered beginning at 1. The compiler may emit the value 0 in cases where an instruction cannot be attributed to any source line.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191471
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Thu, 26 Sep 2013 21:31:43 +0000 (21:31 +0000)]
[mips][msa] Direct Object Emission for 3RF instructions.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191461
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Thu, 26 Sep 2013 21:18:57 +0000 (21:18 +0000)]
[mips][msa] Updates encoding of 3RF instructions to match the latest revision of the MSA spec (1.06).
This does not affect any of the existing output.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191460
91177308-0d34-0410-b5e6-
96231b3b80d8
Weiming Zhao [Thu, 26 Sep 2013 17:25:10 +0000 (17:25 +0000)]
Fix PR 17372: Emitting PLD for stack address for ARM Thumb2
t2PLDi12, t2PLDi8, t2PLDs was omitted in Thumb2InstrInfo.
This patch fixes it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191441
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Schmidt [Thu, 26 Sep 2013 17:09:28 +0000 (17:09 +0000)]
[PowerPC] Fix PR17354: Generate nop after local calls for PIC code.
When generating code for shared libraries, even local calls may be
intercepted, so we need a nop after the call for the linker to fix up the
TOC. Test case adapted from the one provided in PR17354.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191440
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrea Di Biagio [Thu, 26 Sep 2013 16:54:01 +0000 (16:54 +0000)]
Revert r191393 since it caused pr17380.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191438
91177308-0d34-0410-b5e6-
96231b3b80d8
Venkatraman Govindaraju [Thu, 26 Sep 2013 15:11:00 +0000 (15:11 +0000)]
[Sparc] Implements exception handling in SPARC with DwarfCFI.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191432
91177308-0d34-0410-b5e6-
96231b3b80d8
Venkatraman Govindaraju [Thu, 26 Sep 2013 14:49:40 +0000 (14:49 +0000)]
Implements parsing and emitting of .cfi_window_save in MC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191431
91177308-0d34-0410-b5e6-
96231b3b80d8
Amara Emerson [Thu, 26 Sep 2013 12:22:36 +0000 (12:22 +0000)]
[ARM] Use the load-acquire/store-release instructions optimally in AArch32.
Patch by Artyom Skrobov.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191428
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Thu, 26 Sep 2013 09:18:48 +0000 (09:18 +0000)]
PPC: Allow partial fills in writeNopData()
When asked to pad an irregular number of bytes, we should fill with
zeros. This is consistent with the behavior specified in the AIX
Assembler Language Reference as well as other LLVM and binutils
assemblers.
N.B. There is a small deviation from binutils' PPC assembler:
when handling pads which are greater than 4 bytes but not mod 4,
binutils will not emit any NOP sequences at all and only use zeros.
This may or may not be a bug but there is no excellent rationale as to
why that behavior is important to emulate. If that behavior is needed,
we can change writeNopData() to behave in the same way.
This fixes PR17352.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191426
91177308-0d34-0410-b5e6-
96231b3b80d8
Renato Golin [Thu, 26 Sep 2013 08:57:07 +0000 (08:57 +0000)]
Add links to cross-compilation docs from getting started
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191425
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Trick [Thu, 26 Sep 2013 05:53:35 +0000 (05:53 +0000)]
Added temp flag -misched-bench for staging in default changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191423
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Trick [Thu, 26 Sep 2013 05:53:31 +0000 (05:53 +0000)]
whitespace
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191422
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Thu, 26 Sep 2013 05:22:11 +0000 (05:22 +0000)]
PPC: Do not introduce ISD nodes for fctid and fctiw
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191421
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Thu, 26 Sep 2013 04:11:24 +0000 (04:11 +0000)]
PPC: Add support for fctid and fctiw
Encodings were checked against the Power ISA documents and double
checked against binutils.
This fixes PR17350.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191419
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Thu, 26 Sep 2013 00:09:46 +0000 (00:09 +0000)]
[mips][msa] Direct Object Emission for 3R instructions.
This is the first set of instructions with a ".b" modifier thus we need to add the required code to disassemble a MSA128B register class.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191415
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Thu, 26 Sep 2013 00:02:44 +0000 (00:02 +0000)]
[mips][msa] Updates encoding of 3R instructions to match the latest revision of the MSA spec (1.06).
Internal changes only.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191414
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Wed, 25 Sep 2013 23:56:25 +0000 (23:56 +0000)]
[mips][msa] Direct Object Emission for 2RF instructions.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191413
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Wed, 25 Sep 2013 23:50:44 +0000 (23:50 +0000)]
[mips][msa] Direct Object Emission support for the MSA instruction set.
In more detail, this patch adds the ability to parse, encode and decode MSA registers ($w0-$w31). The format of 2RF instructions (MipsMSAInstrFormat.td) was updated so that we could attach a test case to this patch i.e., the test case parses, encodes and decodes 2 MSA instructions. Following patches will add the remainder of the instructions.
Note that DecodeMSA128BRegisterClass is missing from MipsDisassembler.td because it's not yet required at this stage and having it would cause a compiler warning (unused function).
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191412
91177308-0d34-0410-b5e6-
96231b3b80d8
Jack Carter [Wed, 25 Sep 2013 23:42:03 +0000 (23:42 +0000)]
[mips][msa] Updates encoding of 2RF instructions to match the latest revision of the MSA spec (1.06).
This only changes internal encodings and doesn't affect output.
Patch by Matheus Almeida
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191411
91177308-0d34-0410-b5e6-
96231b3b80d8
Weiming Zhao [Wed, 25 Sep 2013 23:12:06 +0000 (23:12 +0000)]
Fix PR 17368: disable vector mul distribution for square of add/sub for ARM
Generally, it is desirable to distribute (a + b) * c to a*c + b*c for
ARM with VMLx forwarding, where a, b and c are vectors.
However, for (a + b)*(a + b), distribution will result in one extra
instruction.
With distribution:
x = a + b (add)
y = a * x (mul)
z = y + b * y (mla)
Without distribution:
x = a + b (add)
z = x * x (mul)
This patch checks if a mul is a square of add/sub. If yes, skip
distribution.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191410
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Wed, 25 Sep 2013 23:02:44 +0000 (23:02 +0000)]
Add gnu pubsections as options to llvm-dwarfdump.
Argument spelling feedback welcome.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191409
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Wed, 25 Sep 2013 23:02:41 +0000 (23:02 +0000)]
Dump the normal dwarf pubtypes section as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191408
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Wed, 25 Sep 2013 23:02:36 +0000 (23:02 +0000)]
Unify pubsection/gnu pubsection printing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191407
91177308-0d34-0410-b5e6-
96231b3b80d8
Josh Magee [Wed, 25 Sep 2013 22:07:48 +0000 (22:07 +0000)]
Test commit. Removed trailing whitespace.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191402
91177308-0d34-0410-b5e6-
96231b3b80d8
Eric Christopher [Wed, 25 Sep 2013 21:17:37 +0000 (21:17 +0000)]
Slight formatting change for pubnames/pubtypes output.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191401
91177308-0d34-0410-b5e6-
96231b3b80d8
Reed Kotler [Wed, 25 Sep 2013 20:58:50 +0000 (20:58 +0000)]
Fix a bad typo in the inline assembly code for mips16 pic fp stubs
and make one cosmetic cleanup to make it look the same as gcc
in this area; adjusting test cases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191400
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrea Di Biagio [Wed, 25 Sep 2013 19:01:01 +0000 (19:01 +0000)]
Teach DAGCombiner how to canonicalize dags according to the rule
(shl (zext (shr A, X)), X) => (zext (shl (shr A, X), X)).
The rule only triggers when there are no other uses of the
zext to avoid materializing more instructions.
This helps the DAGCombiner understand that the shl/shr
sequence can then be converted into an and instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191393
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Trick [Wed, 25 Sep 2013 18:14:12 +0000 (18:14 +0000)]
Mark the x86 machine model as incomplete. PR17367.
Ideally, the machinel model is added at the time the instructions are
defined. But many instructions in X86InstrSSE.td still need a model.
Without this workaround the scheduler asserts because x86 already has
itinerary classes for these instructions, indicating they should be
modeled by the scheduler. Since we use the new machine model for other
instructions, it expects a new machine model for these too.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191391
91177308-0d34-0410-b5e6-
96231b3b80d8
Joerg Sonnenberger [Wed, 25 Sep 2013 17:49:57 +0000 (17:49 +0000)]
Undefine NetBSD, it may have been defined by an earlier include of
sys/param.h.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191384
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 25 Sep 2013 14:06:55 +0000 (14:06 +0000)]
Set the minimal stack size with msvc when using cmake >= 2.8.11.
This makes sure we get the same behavior with all supported cmake versions. Once
we support only versions >= 2.8.11 we can experiment with other values or just
setting it for some binaries.
Patch by Greg Bedwell.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191372
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnold Schwaighofer [Wed, 25 Sep 2013 14:02:32 +0000 (14:02 +0000)]
SLPVectorize: Put horizontal reductions feeding a store under separate flag
Put them under a separate flag for experimentation. They are more likely to
interfere with loop vectorization which happens later in the pass pipeline.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191371
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Wed, 25 Sep 2013 11:11:53 +0000 (11:11 +0000)]
[SystemZ] Define the GR64 low-word logic instructions as pseudo aliases.
Another patch to avoid duplication of encoding information. Things like
NILF, NILL and NILH are used as both 32-bit and 64-bit instructions.
Here the 64-bit versions are defined as aliases of the 32-bit ones.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191369
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Wed, 25 Sep 2013 10:47:21 +0000 (10:47 +0000)]
MC: Add support for treating $ as a reference to the PC
The binutils assembler supports a mode called DOLLAR_DOT which treats
the dollar sign token as a reference to the current program counter if
the dollar sign doesn't precede a constant or identifier.
This commit adds a new MCAsmInfo flag stating whether or not a given
target supports this interpretation of the dollar sign token; by
default, this flag is not enabled.
Further, enable this flag for PPC. The system assembler for AIX and
binutils both support using the dollar sign in this manner.
This fixes PR17353.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191368
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Wed, 25 Sep 2013 10:37:17 +0000 (10:37 +0000)]
[SystemZ] Define the call instructions as pseudo aliases.
Similar to r191364, but for calls. This patch also removes the shortening
of BRASL to BRAS within a TU. Doing that was a bit controversial internally,
since there's a strong expectation with the z assembler that WYWIWYG.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191366
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Wed, 25 Sep 2013 10:29:47 +0000 (10:29 +0000)]
[SystemZ] Use subregs for 64-bit truncating stores
Another patch to reduce the duplication of encoding information.
Rather than define separate patterns for truncating 64-bit stores,
use the 32-bit stores with a subreg. No behavioral changed intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191365
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Wed, 25 Sep 2013 10:20:08 +0000 (10:20 +0000)]
[SystemZ] Define the return instruction as a pseudo alias of BR
This is the first of a few patches to reduce the dupliation of encoding
information. The return instruction is a normal BR in which one of the
registers is fixed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191364
91177308-0d34-0410-b5e6-
96231b3b80d8
Richard Sandiford [Wed, 25 Sep 2013 10:11:07 +0000 (10:11 +0000)]
[SystemZ] Add instruction-shortening pass
When loading immediates into a GR32, the port prefered LHI, followed by
LLILH or LLILL, followed by IILF. LHI and IILF are natural 32-bit
operations, but LLILH and LLILL also clear the upper 32 bits of the register.
This was represented as taking a 32-bit subreg of a 64-bit assignment.
Using subregs for something as simple as a move immediate was probably
a bad idea. Also, I have patches to add support for the high-word facility,
and we don't want something like LLILH and LLILL to stop the high word of
the same GPR from being used.
This patch therefore uses LHI and IILF to begin with and adds a late
machine-specific pass to use LLILH and LLILL if the other half of the
register is not live. The high-word patches extend this behavior to
IIHF, LLIHL and LLIHH.
No behavioral change intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191363
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Wed, 25 Sep 2013 09:36:11 +0000 (09:36 +0000)]
MC: Remove vestigial PCSymbol field from AsmInfo
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191362
91177308-0d34-0410-b5e6-
96231b3b80d8
Evgeniy Stepanov [Wed, 25 Sep 2013 08:56:00 +0000 (08:56 +0000)]
[msan] Fix -Wreturn-type warnings in non-self-hosted build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191361
91177308-0d34-0410-b5e6-
96231b3b80d8
Peter Collingbourne [Wed, 25 Sep 2013 07:52:21 +0000 (07:52 +0000)]
Try again to fix the MSVC build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191359
91177308-0d34-0410-b5e6-
96231b3b80d8
Peter Collingbourne [Wed, 25 Sep 2013 07:11:58 +0000 (07:11 +0000)]
Wrap the #include of <stdbool.h> in an #ifndef __cplusplus.
This should fix the MSVC build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191357
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Wed, 25 Sep 2013 06:40:22 +0000 (06:40 +0000)]
Fix doxygen comments to use correct function name.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191356
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Wed, 25 Sep 2013 06:37:18 +0000 (06:37 +0000)]
Replace EVT with MVT in CodeGenDAGAPatterns.cpp.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191355
91177308-0d34-0410-b5e6-
96231b3b80d8
Akira Hatanaka [Wed, 25 Sep 2013 00:52:34 +0000 (00:52 +0000)]
Revert r191350.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191353
91177308-0d34-0410-b5e6-
96231b3b80d8
Akira Hatanaka [Wed, 25 Sep 2013 00:34:42 +0000 (00:34 +0000)]
[mips] Move public functions to the beginning of the class definition.
No intended functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191352
91177308-0d34-0410-b5e6-
96231b3b80d8
Akira Hatanaka [Wed, 25 Sep 2013 00:30:25 +0000 (00:30 +0000)]
[mips] Define getTargetNode as a template function.
No intended functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191350
91177308-0d34-0410-b5e6-
96231b3b80d8
Quentin Colombet [Wed, 25 Sep 2013 00:26:17 +0000 (00:26 +0000)]
[PR16882] Ignore noreturn definitions when setting isPhysRegUsed.
PEI inserts a save/restore sequence for the link register, according to the
information it gets from the MachineRegisterInfo.
MachineRegisterInfo is populated by the VirtRegMap pass.
This pass was not aware of noreturn calls and was registering the definitions of
these calls the same way as regular operations.
Modify VirtRegPass so that it does not set the isPhysRegUsed information for
registers only defined by noreturn calls.
The rational is that a noreturn call is the "last instruction" of the program
(if it returns the behavior is undefined), so everything that is defined by it
cannot be used and will not interfere with anything else. Therefore, it is
pointless to account for then.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191349
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Trick [Wed, 25 Sep 2013 00:26:16 +0000 (00:26 +0000)]
CriticalAntiDepBreaker is no longer needed for armv7 scheduling.
This is being disabled because it is no longer needed for
performance. It is only used by postRAscheduler which is also planned
for removal, and it is implemented with an out-dated view of register
liveness. It consideres aliases instead of register units, assumes
valid kill flags, and assumes implicit uses on partial register
defs. Kill flags and implicit operands are error prone and impossible
to verify. We should gradually eliminate dependence on them in the
postRA phases.
Targets that still benefit from this should move to the MI
scheduler. If that doesn't solve the problem, then we should add a
hook to regalloc to optimize reload placement.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191348
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Tue, 24 Sep 2013 23:56:31 +0000 (23:56 +0000)]
MachO: Improve backend diagnostic for overalignment.
Give the symbol's name and disengage the enchanced crash reporting.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191344
91177308-0d34-0410-b5e6-
96231b3b80d8
Peter Collingbourne [Tue, 24 Sep 2013 23:52:22 +0000 (23:52 +0000)]
Move LTO support library to a component, allowing it to be tested
more reliably across platforms. Patch by Tom Roeder!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191343
91177308-0d34-0410-b5e6-
96231b3b80d8
Eli Friedman [Tue, 24 Sep 2013 22:50:14 +0000 (22:50 +0000)]
Add missing check to SETCC optimization.
PR17338.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191337
91177308-0d34-0410-b5e6-
96231b3b80d8
David Blaikie [Tue, 24 Sep 2013 20:23:36 +0000 (20:23 +0000)]
llvm-dwarfdump: add missing opening quotation mark lost in r191330
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191333
91177308-0d34-0410-b5e6-
96231b3b80d8
Stepan Dyatkovskiy [Tue, 24 Sep 2013 20:06:31 +0000 (20:06 +0000)]
Patch that forces MergeFunctions pass for clang.
It is temporary patch. We need to keep it in trunk, since it makes easer to test it on buildbots on different platforms.
Once we see stable MergeFunctions behaviour with satisfied perfomance, this patch will be removed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191331
91177308-0d34-0410-b5e6-
96231b3b80d8
David Blaikie [Tue, 24 Sep 2013 19:56:27 +0000 (19:56 +0000)]
llvm-dwarfdump: re-add field formatting for the entry kind lost in r191329
CR feedback from Eric Christopher
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191330
91177308-0d34-0410-b5e6-
96231b3b80d8
David Blaikie [Tue, 24 Sep 2013 19:50:00 +0000 (19:50 +0000)]
llvm-dwarfdump support for gnu_pubtypes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191329
91177308-0d34-0410-b5e6-
96231b3b80d8
Yi Jiang [Tue, 24 Sep 2013 19:33:53 +0000 (19:33 +0000)]
Test case for r191314.
Some supplemental information for r191314: We would like to make sure SLP Vectorizer will not try to vectorize tiny trees even with a negative threshold so we set the cost to INT_MAX.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191327
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 24 Sep 2013 18:37:49 +0000 (18:37 +0000)]
Verify that we don't optimize null return checks to the nothrow_t version of operator new.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191325
91177308-0d34-0410-b5e6-
96231b3b80d8
Yunzhong Gao [Tue, 24 Sep 2013 18:21:52 +0000 (18:21 +0000)]
Adding a feature flag to the llvm backend for x86 TBM instruction set.
Adding TBM feature to bdver2 processor; piledriver supports this instruction set
according to the following document:
http://developer.amd.com/wordpress/media/2012/10/New-Bulldozer-and-Piledriver-Instructions.pdf
Phabricator code review is located here: http://llvm-reviews.chandlerc.com/D1692
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191324
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 24 Sep 2013 17:49:08 +0000 (17:49 +0000)]
MemoryBuiltins: Remove posix_memalign from the list and replace it with a TODO.
This code isn't ready to deal with allocation functions where the return is not
the allocated pointer. The checks below will reject posix_memalign anyways.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191319
91177308-0d34-0410-b5e6-
96231b3b80d8
Roman Divacky [Tue, 24 Sep 2013 17:44:41 +0000 (17:44 +0000)]
Make the size and expr arguments of .fill directive optional.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191318
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 24 Sep 2013 17:34:29 +0000 (17:34 +0000)]
MemoryBuiltins: Reinstate optimizing (uninitialized) loads from operator new.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191315
91177308-0d34-0410-b5e6-
96231b3b80d8
Yi Jiang [Tue, 24 Sep 2013 17:26:43 +0000 (17:26 +0000)]
set the cost of tiny trees to INT_MAX in SLP vectorizer to disable vectorization on them
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191314
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 24 Sep 2013 17:15:14 +0000 (17:15 +0000)]
MemoryBuiltins: Fix operator new bits.
We really don't want to optimize malloc return value checks away.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191313
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Trick [Tue, 24 Sep 2013 17:11:19 +0000 (17:11 +0000)]
Comment typo.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191312
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 24 Sep 2013 16:37:51 +0000 (16:37 +0000)]
Teach MemoryBuiltins and InstructionSimplify that operator new never returns NULL.
This is safe per C++11 18.6.1.1p3: [operator new returns] a non-null pointer to
suitably aligned storage (3.7.4), or else throw a bad_alloc exception. This
requirement is binding on a replacement version of this function.
Brings us a tiny bit closer to eliminating more vector push_backs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191310
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 24 Sep 2013 16:37:40 +0000 (16:37 +0000)]
Push analysis passes to InstSimplify when they're around anyways.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191309
91177308-0d34-0410-b5e6-
96231b3b80d8