Commit Graph

24428 Commits

Author SHA1 Message Date
Tim Northover ce6538c38d AArch64 & ARM: remove undefined behaviour from some tests.
llvm-svn: 209880
2014-05-30 08:59:55 +00:00
Hao Liu 5c7314b68d Test cases named with dates is a legacy rule not used now. Rename several test cases.
llvm-svn: 209877
2014-05-30 05:58:19 +00:00
Karthik Bhat 5ab7795649 Allow vectorization of intrinsics such as powi,cttz and ctlz in Loop and SLP Vectorizer.
This patch adds support to vectorize intrinsics such as powi, cttz and ctlz in Vectorizer. These intrinsics are different from other
intrinsics as second argument to these function must be same in order to vectorize them and it should be represented as a scalar.
Review: http://reviews.llvm.org/D3851#inline-32769 and http://reviews.llvm.org/D3937#inline-32857

llvm-svn: 209873
2014-05-30 04:31:24 +00:00
Nick Lewycky 59633cb478 When analyzing params/args for readnone/readonly, don't forget to consider that a pointer argument may be passed through a callsite to the return, and that we may need to analyze it. Fixes a bug reported on llvm-dev: http://lists.cs.uiuc.edu/pipermail/llvmdev/2014-May/073098.html
llvm-svn: 209870
2014-05-30 02:31:27 +00:00
Adam Nemet b3587e98b7 [X86] Move test from r209863 to CodeGen/X86
We should only run this if X86 is in the targets.

llvm-svn: 209866
2014-05-29 23:52:53 +00:00
Adam Nemet 35b80eaef1 [X86] Remove AVX1 vbroadcast intrinsics
The corresponding CFE patch replaces these intrinsics with vector initializers
in avxintrin.h.  This patch removes the LLVM intrinsics from the backend.

We now stop lowering at X86ISD::VBROADCAST custom node rather than lowering
that further to the intrinsics.

The patch only changes VBROADCASTS* and leaves VBROADCAST[FI]128 to continue
to use intrinsics.  As explained in the CFE patch, the reason is that we
currently don't generate as good code for them without the intrinsics.

CodeGen/X86/avx-vbroadcast.ll already provides coverage for this change.  It
checks that for a series of insertelements we generate the appropriate
vbroadcast instruction.

Also verified that there was no assembly change in the test-suite before and
after this patch.

llvm-svn: 209864
2014-05-29 23:35:36 +00:00
Adam Nemet 39066800e9 [X86] Auto-upgrade AVX1 vbroadcast intrinsics
They are replaced with the same IR that is generated for the
vector-initializers in avxintrin.h.

The test verifies that we get back the original instruction.  I haven't seen
this approach to be used in other auto-upgrade tests (i.e. llc + FileCheck)
but I think it's the most direct way to test this case.  I believe this should
work because llc upgrades calls during parsing.  (Other tests mostly check
that assembling and disassembling yields the upgraded IR.)

llvm-svn: 209863
2014-05-29 23:35:33 +00:00
Arnold Schwaighofer e2067680a6 LoopVectorizer: Add a check that the backedge taken count + 1 does not overflow
The loop vectorizer instantiates be-taken-count + 1 as the loop iteration count.
If this expression overflows the generated code was invalid.

In case of overflow the code now jumps to the scalar loop.

Fixes PR17288.

llvm-svn: 209854
2014-05-29 22:10:01 +00:00
Filipe Cabecinhas 229dc17610 Added tests for shufflevector lowering to blend instrs.
These tests ensure that a change I will propose in clang works as
expected.

Summary:
Added tests for the generation of blend+immediate instructions from a
shufflevector.
These tests were proposed along with a patch that was dropped. I'm
committing the tests anyway to protect against possible regressions in
codegen.

Reviewers: nadav, bkramer

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D3600

llvm-svn: 209853
2014-05-29 22:04:42 +00:00
Louis Gerbarg c6b506a0ae Add support for combining GEPs across PHI nodes
Currently LLVM will generally merge GEPs. This allows backends to use more
complex addressing modes. In some cases this is not happening because there
is PHI inbetween the two GEPs:

  GEP1--\
        |-->PHI1-->GEP3
  GEP2--/

This patch checks to see if GEP1 and GEP2 are similiar enough that they can be
cloned (GEP12) in GEP3's BB, allowing GEP->GEP merging (GEP123):

  GEP1--\                     --\                           --\
        |-->PHI1-->GEP3  ==>    |-->PHI2->GEP12->GEP3 == >    |-->PHI2->GEP123
  GEP2--/                     --/                           --/

This also breaks certain use chains that are preventing GEP->GEP merges that the
the existing instcombine would merge otherwise.

Tests included.

llvm-svn: 209843
2014-05-29 20:29:47 +00:00
Rafael Espindola 4d16e11922 Rename alias variables to make it easier to add new tests to the file.
llvm-svn: 209822
2014-05-29 16:16:12 +00:00
Rafael Espindola 04902862a8 [PPC] Use alias symbols in address computation.
This seems to match what gcc does for ppc and what every other llvm
backend does.

This is a fixed version of r209638. The difference is to avoid any change
in behavior for functions. The logic for using constant pools for function
addresseses is spread over a few places and we have to keep them in sync.

llvm-svn: 209821
2014-05-29 15:41:38 +00:00
Rafael Espindola 1613650e90 Add a test showing the ppc code sequence for getting a function pointer.
This would have found the miscompile in r209638.

llvm-svn: 209820
2014-05-29 15:13:23 +00:00
Rafael Espindola a248f536b3 Revert "Revert "Revert "InstCombine: Improvement to check if signed addition overflows."""
This reverts commit r209776.

It was miscompiling llvm::SelectionDAGISel::MorphNode.

llvm-svn: 209817
2014-05-29 14:39:16 +00:00
Artyom Skrobov 0d22e2afe7 Restore getInvertedCondCode() from the phased-out backend, fixing disassembly for NV
llvm-svn: 209803
2014-05-29 11:34:50 +00:00
Artyom Skrobov 7e9e31ebaf Add missing check when MatchInstructionImpl() reports failure
llvm-svn: 209802
2014-05-29 11:26:15 +00:00
Simon Atanasyan 8745993ec4 [elf2yaml][ELF] Move Info field to the RelocationSection structure. This
field represents ELF section header sh_info field and does not have any
sense for regular sections. Its interpretation depends on section type.

llvm-svn: 209801
2014-05-29 11:05:31 +00:00
Hao Liu d670c7eee0 Rename a test case to contain correct date info.
llvm-svn: 209799
2014-05-29 09:21:23 +00:00
Hao Liu 4091450181 Fix an assertion failure caused by v1i64 in DAGCombiner Shrink.
llvm-svn: 209798
2014-05-29 09:19:07 +00:00
Dinesh Dwivedi d266cb1a0b LCSSA should be performed on the outermost affected loop while unrolling loop.
During loop-unroll, loop exits from the current loop may end up in in different
outer loop. This requires to re-form LCSSA recursively for one level down from
the outer most loop where loop exits are landed during unroll. This fixes PR18861.

Differential Revision: http://reviews.llvm.org/D2976

llvm-svn: 209796
2014-05-29 06:47:23 +00:00
Michael J. Spencer 289067cc3d Add LoadCombine pass.
This pass is disabled by default. Use -combine-loads to enable in -O[1-3]

Differential revision: http://reviews.llvm.org/D3580

llvm-svn: 209791
2014-05-29 01:55:07 +00:00
Michael J. Spencer f375d80635 [x86] Fold extract_vector_elt of a load into the Load's address computation.
An address only use of an extract element of a load can be simplified to a
load. Without this the result of the extract element is spilled to the
stack so that an address is available.

llvm-svn: 209788
2014-05-29 01:42:45 +00:00
Alexey Samsonov 96e239f564 [ASan] Use llvm.global_ctors to insert init-order checking calls into ASan runtime.
Don't assume that dynamically initialized globals are all initialized from
_GLOBAL__<module_name>I_ function. Instead, scan the llvm.global_ctors and
insert poison/unpoison calls to each function there.

Patch by Nico Weber!

llvm-svn: 209780
2014-05-29 00:51:15 +00:00
Rafael Espindola 6196b7430e Revert "Revert "InstCombine: Improvement to check if signed addition overflows.""
This reverts commit r209762, bringing back r209746. It was not responsible for the libc++ build failure

llvm-svn: 209776
2014-05-28 21:43:52 +00:00
Rafael Espindola 910528a3eb Revert "Add support for combining GEPs across PHI nodes"
This reverts commit r209755.

it was the real cause of the libc++ build failure.

llvm-svn: 209775
2014-05-28 21:41:21 +00:00
Rafael Espindola fb59b05ca4 Revert "InstCombine: Improvement to check if signed addition overflows."
This reverts commit r209746.

It looks it is causing a crash while building libcxx. I am trying to get a
reduced testcase.

llvm-svn: 209762
2014-05-28 18:48:10 +00:00
Rafael Espindola 59f7eba2b5 [pr19844] Add thread local mode to aliases.
This matches gcc's behavior. It also seems natural given that aliases
contain other properties that govern how it is accessed (linkage,
visibility, dll storage).

Clang still has to be updated to expose this feature to C.

llvm-svn: 209759
2014-05-28 18:15:43 +00:00
Louis Gerbarg 727f1cbb17 Add support for combining GEPs across PHI nodes
Currently LLVM will generally merge GEPs. This allows backends to use more
complex addressing modes. In some cases this is not happening because there
is PHI inbetween the two GEPs:

  GEP1--\
        |-->PHI1-->GEP3
  GEP2--/

This patch checks to see if GEP1 and GEP2 are similiar enough that they can be
cloned (GEP12) in GEP3's BB, allowing GEP->GEP merging (GEP123):

  GEP1--\                     --\                           --\
        |-->PHI1-->GEP3  ==>    |-->PHI2->GEP12->GEP3 == >    |-->PHI2->GEP123
  GEP2--/                     --/                           --/

This also breaks certain use chains that are preventing GEP->GEP merges that the
the existing instcombine would merge otherwise.

Tests included.

llvm-svn: 209755
2014-05-28 17:38:31 +00:00
Hal Finkel 2c77fe59d9 Revert "[DAGCombiner] Split up an indexed load if only the base pointer value is live"
This reverts r208640 (I've just XFAILed the test) because it broke ppc64/Linux
self-hosting. Because nearly every regression test triggers a segfault, I hope
this will be easy to fix.

llvm-svn: 209747
2014-05-28 15:33:19 +00:00
Rafael Espindola 085b57941f InstCombine: Improvement to check if signed addition overflows.
This patch implements two things:

1. If we know one number is positive and another is negative, we return true as
   signed addition of two opposite signed numbers will never overflow.

2. Implemented TODO : If one of the operands only has one non-zero bit, and if
   the other operand has a known-zero bit in a more significant place than it
   (not including the sign bit) the ripple may go up to and fill the zero, but
   won't change the sign. e.x -  (x & ~4) + 1

We make sure that we are ignoring 0 at MSB.

Patch by Suyog Sarda.

llvm-svn: 209746
2014-05-28 15:30:40 +00:00
Hal Finkel f5c07ada1d Revert "[PPC] Use alias symbols in address computation."
This reverts commit r209638 because it broke self-hosting on ppc64/Linux. (the
Clang-compiled TableGen would segfault because it jumped to an invalid address
from within _ZNK4llvm17ManagedStaticBase21RegisterManagedStaticEPFPvvEPFvS1_E
(which is within the command-line parameter registration process)).

llvm-svn: 209745
2014-05-28 15:25:06 +00:00
Evgeniy Stepanov 386b58d056 [asancov] Don't emit extra runtime calls when compiling without coverage.
llvm-svn: 209721
2014-05-28 09:26:46 +00:00
Tilmann Scheller 7c747fc7c6 [AArch64] Add store post-index update folding regression tests for the load/store optimizer.
Add regression tests for the following transformation:

  str X, [x20]
   ...
  add x20, x20, #32
   ->
  str X, [x20], #32

with X being either w0, x0, s0, d0 or q0.

llvm-svn: 209715
2014-05-28 06:43:00 +00:00
Tilmann Scheller 35e451461f [AArch64] Add load post-index update folding regression tests for the load/store optimizer.
Add regression tests for the following transformation:

 ldr X, [x20]
  ...
 add x20, x20, #32
  ->
 ldr X, [x20], #32

 with X being either w0, x0, s0, d0 or q0.

llvm-svn: 209711
2014-05-28 05:44:14 +00:00
Sebastian Pop a6e5860513 remove constant terms
The delinearization is needed only to remove the non linearity induced by
expressions involving multiplications of parameters and induction variables.
There is no problem in dealing with constant times parameters, or constant times
an induction variable.

For this reason, the current patch discards all constant terms and multipliers
before running the delinearization algorithm on the terms. The only thing
remaining in the term expressions are parameters and multiply expressions of
parameters: these simplified term expressions are passed to the array shape
recognizer that will not recognize constant dimensions anymore: these will be
recognized as different strides in parametric subscripts.

The only important special case of a constant dimension is the size of elements.
Instead of relying on the delinearization to infer the size of an element,
compute the element size from the base address type. This is a much more precise
way of computing the element size than before, as we would have mixed together
the size of an element with the strides of the innermost dimension.

llvm-svn: 209691
2014-05-27 22:41:45 +00:00
Arnaud A. de Grandmaison de5ff26865 No need for those tests to go thru llvm-as and/or llvm-dis.
opt can handle them by itself.

llvm-svn: 209689
2014-05-27 22:03:28 +00:00
Sasa Stankovic e41db2fe31 [mips] Optimize long branch for MIPS64 by removing %higher and %highest.
%higher and %highest can have non-zero values only for offsets greater
than 2GB, which is highly unlikely, if not impossible when compiling a
single function. This makes long branch for MIPS64 3 instructions smaller.

Differential Revision: http://llvm-reviews.chandlerc.com/D3281.diff

llvm-svn: 209678
2014-05-27 18:53:06 +00:00
David Blaikie 482097d098 DebugInfo: Create abstract function definitions even when concrete definitions preceed inline definitions.
After much puppetry, here's the major piece of the work to ensure that
even when a concrete definition preceeds all inline definitions, an
abstract definition is still created and referenced from both concrete
and inline definitions.

Variables are still broken in this case (see comment in
dbg-value-inlined-parameter.ll test case) and will be addressed in
follow up work.

llvm-svn: 209677
2014-05-27 18:37:55 +00:00
David Blaikie 3c2fff3fe6 DebugInfo: Lazily construct subprogram definition DIEs.
A further step to correctly emitting concrete out of line definitions
preceeding inlined instances of the same program.

To do this, emission of subprograms must be delayed until required since
we don't know which (abstract only (if there's no out of line
definition), concrete only (if there are no inlined instances), or both)
DIEs are required at the start of the module.

To reduce the test churn in the following commit that actually fixes the
bug, this commit introduces the lazy DIE construction and cleans up test
cases that are impacted by the changes in the resulting DIE ordering.

llvm-svn: 209675
2014-05-27 18:37:48 +00:00
David Blaikie f7221adb8e DebugInfo: Lazily attach definition attributes to definitions.
This is a precursor to fixing inlined debug info where the concrete,
out-of-line definition may preceed any inlined usage. To cope with this,
the attributes that may appear on the concrete definition or the
abstract definition are delayed until the end of the module. Then, if an
abstract definition was created, it is referenced (and no other
attributes are added to the out-of-line definition), otherwise the
attributes are added directly to the out-of-line definition.

In a couple of cases this causes not just reordering of attributes, but
reordering of types. When the creation of the attribute is delayed, if
that creation would create a type (such as for a DW_AT_type attribute)
then other top level DIEs may've been constructed during the delay,
causing the referenced type to be created and added after those
intervening DIEs. In the extreme case, in cross-cu-inlining.ll, this
actually causes the DW_TAG_basic_type for "int" to move from one CU to
another.

llvm-svn: 209674
2014-05-27 18:37:43 +00:00
Jingyue Wu 76cbea6b6d Fixed a test in r209670
The test was outdated with r209537.

llvm-svn: 209671
2014-05-27 18:12:55 +00:00
Jingyue Wu 80a738dc62 Distribute sext/zext to the operands of and/or/xor
This is an enhancement to SeparateConstOffsetFromGEP. With this patch, we can
extract a constant offset from "s/zext and/or/xor A, B".

Added a new test @ext_or to verify this enhancement.

Refactoring the code, I also extracted some common logic to function
Distributable. 

llvm-svn: 209670
2014-05-27 18:00:00 +00:00
David Blaikie 66c3077470 DebugInfo: Fix argument ordering in test by adding argument numbering.
This old test didn't have the argument numbering that's now squirelled
away in the high bits of the line number in the DW_TAG_arg_variable
metadata.

Add the numbering and update the test to ensure arguments are in-order.

llvm-svn: 209669
2014-05-27 17:57:14 +00:00
Filipe Cabecinhas e8d6a1e82f Post-commit fixes for r209643
Detected by Daniel Jasper, Ilia Filippov, and Andrea Di Biagio
Fixed the argument order to select (the mask semantics to blendv* are the
inverse of select) and fixed the tests
Added parenthesis to the assert condition
Ran clang-format

llvm-svn: 209667
2014-05-27 16:54:33 +00:00
Tim Northover 9a6217b650 AArch64: add test for NZCV cross-copy save.
llvm-svn: 209665
2014-05-27 16:50:09 +00:00
Tim Northover de9402d345 AArch64: add AArch64-specific test for 'c' and 'n'.
llvm-svn: 209664
2014-05-27 16:50:03 +00:00
Bill Schmidt 71dddd51d9 [PATCH] Correct type used for VADD_SPLAT optimization on PowerPC
In PPCISelLowering.cpp: PPCTargetLowering::LowerBUILD_VECTOR(), there
is an optimization for certain patterns to generate one or two vector
splats followed by a vector add or subtract.  This operation is
represented by a VADD_SPLAT in the selection DAG.  Prior to this
patch, it was possible for the VADD_SPLAT to be assigned the wrong
data type, causing incorrect code generation.  This patch corrects the
problem.

Specifically, the code previously assigned the value type of the
BUILD_VECTOR node to the newly generated VADD_SPLAT node.  This is
correct much of the time, but not always.  The problem is that the
call to isConstantSplat() may return a SplatBitSize that is not the
same as the number of bits in the original element vector type.  The
correct type to assign is a vector type with the same element bit size
as SplatBitSize.

The included test case shows an example of this, where the
BUILD_VECTOR node has a type of v16i8.  The vector to be built is {0,
16, 0, 16, 0, 16, 0, 16, 0, 16, 0, 16, 0, 16, 0, 16}.  isConstantSplat
detects that we can generate a splat of 16 for type v8i16, which is
the type we must assign to the VADD_SPLAT node.  If we do not, we
generate a vspltisb of 8 and a vaddubm, which generates the incorrect
result {16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16,
16}.  The correct code generation is a vspltish of 8 and a vadduhm.

This patch also corrected code generation for
CodeGen/PowerPC/2008-07-10-SplatMiscompile.ll, which had been marked
as an XFAIL, so we can remove the XFAIL from the test case.

llvm-svn: 209662
2014-05-27 15:57:51 +00:00
Zoran Jovanovic b355e8f604 [mips][mips64r6] Add Relocations R_MIPS_PCHI16, R_MIPS_PCLO16
Differential Revision: http://reviews.llvm.org/D3860

llvm-svn: 209659
2014-05-27 14:58:51 +00:00
Amara Emerson ceeb1c4830 [ARM] Emit correct build attributes for the relocation models.
Patch by Asiri Rathnayake.

llvm-svn: 209656
2014-05-27 13:30:21 +00:00
Zoran Jovanovic 10e06da031 [mips][mips64r6] Add relocations R_MIPS_PC21_S2, R_MIPS_PC26_S2
Differential Revision: http://reviews.llvm.org/D3824

llvm-svn: 209655
2014-05-27 12:55:40 +00:00