Craig Topper [Wed, 22 Apr 2015 04:18:27 +0000 (04:18 +0000)]
Don't use 'nullptr' in comment. Just use 'null'.
llvm-svn: 235476
David Blaikie [Wed, 22 Apr 2015 04:14:46 +0000 (04:14 +0000)]
[opaque pointer types] Serialize the value type for atomic store instructions
Without pointee types the space optimization of storing only the pointer
type and not the value type won't be viable - so add the extra type
information that would be missing.
llvm-svn: 235475
David Blaikie [Wed, 22 Apr 2015 04:14:42 +0000 (04:14 +0000)]
[opaque pointer types] Serialize the value type for store instructions
Without pointee types the space optimization of storing only the pointer
type and not the value type won't be viable - so add the extra type
information that would be missing.
Storeatomic coming soon.
llvm-svn: 235474
Duncan P. N. Exon Smith [Wed, 22 Apr 2015 04:11:00 +0000 (04:11 +0000)]
Linker: Add flag to override linkage rules
Add a flag to lib/Linker (and `llvm-link`) to override linkage rules.
When set, the functions in the source module *always* replace those in
the destination module.
The `llvm-link` option is `-override=abc.ll`. All the "regular" modules
are loaded and linked first, followed by the `-override` modules. This
is useful for debugging workflows where some subset of the module (e.g.,
a single function) is extracted into a separate file where it's
optimized differently, before being merged back in.
Patch by Luqman Aden!
llvm-svn: 235473
Duncan P. N. Exon Smith [Wed, 22 Apr 2015 04:08:22 +0000 (04:08 +0000)]
llvm-link: Factor out loop over input files, NFC
Factor the loop for linking input files together into a combined module
into a separate function. This is in preparation for an upcoming patch
that runs the logic twice.
Patch by Luqman Aden!
llvm-svn: 235472
Hans Wennborg [Wed, 22 Apr 2015 04:05:17 +0000 (04:05 +0000)]
Don't dllimport/export class members with internal linkage (PR23308)
For example, a function taking a parameter with internal linkage will
itself have internal linkage since it cannot be called outside the
translation unit.
Differential Revision: http://reviews.llvm.org/D9182
llvm-svn: 235471
Nico Weber [Wed, 22 Apr 2015 03:44:51 +0000 (03:44 +0000)]
Wrap to 80 columns, fix typo in comment. No behavior change.
llvm-svn: 235470
Craig Topper [Wed, 22 Apr 2015 02:59:06 +0000 (02:59 +0000)]
Revert "[TableGen] Use cast instead of dyn_cast where result isn't checked before being dereferenced."
Turns out I misread the parentheses. Though I'm pretty sure its always a RecordRecTy and non of the callers really seem to expect null. But until I'm completely sure I'm going to revert this.
llvm-svn: 235469
Craig Topper [Wed, 22 Apr 2015 02:59:03 +0000 (02:59 +0000)]
Fix stale comment that mentioned 0 instead of nullptr. NFC.
llvm-svn: 235468
Craig Topper [Wed, 22 Apr 2015 02:20:44 +0000 (02:20 +0000)]
[TableGen] Remove Pool helper class and just use unique_ptr in the maps.
llvm-svn: 235467
Craig Topper [Wed, 22 Apr 2015 02:09:47 +0000 (02:09 +0000)]
[TableGen] Use StringRecTy::get() instead of allocating (and leaking) a StringRecTy object.
llvm-svn: 235466
Craig Topper [Wed, 22 Apr 2015 02:09:45 +0000 (02:09 +0000)]
[TableGen] Use 'isa' to identify UnsetInits rather than comparing with the singleton object created by UnsetInit::get(). Makes it more consistent with the other types.
llvm-svn: 235465
Richard Smith [Wed, 22 Apr 2015 02:09:43 +0000 (02:09 +0000)]
[modules] Cope with partial module macro information, fix memory leak found by buildbot.
llvm-svn: 235464
Craig Topper [Wed, 22 Apr 2015 02:09:42 +0000 (02:09 +0000)]
[TableGen] Use cast instead of dyn_cast where result isn't checked before being dereferenced.
llvm-svn: 235463
Craig Topper [Wed, 22 Apr 2015 02:09:39 +0000 (02:09 +0000)]
Cleanup formatting for consistency with surrounding code.
llvm-svn: 235462
Richard Smith [Wed, 22 Apr 2015 00:26:11 +0000 (00:26 +0000)]
[modules] Build a DAG of module macros for each identifier.
This graph will be used to determine the current set of active macros. This is
foundation work for getting macro visibility correct across submodules of the
current module. No functionality change for now.
llvm-svn: 235461
Sanjay Patel [Wed, 22 Apr 2015 00:24:30 +0000 (00:24 +0000)]
[x86] allow 64-bit extracted vector element integer stores on a 32-bit system
With SSE2, we can generate a 'movq' or other 64-bit store op on a 32-bit system
even though 64-bit integers are not legal types.
So instead of producing this:
pshufd $229, %xmm0, %xmm1 ## xmm1 = xmm0[1,1,2,3]
movd %xmm0, (%eax)
movd %xmm1, 4(%eax)
We can do:
movq %xmm0, (%eax)
This is a fix for the problem noted in D7296.
Differential Revision: http://reviews.llvm.org/D9134
llvm-svn: 235460
Reid Kleckner [Wed, 22 Apr 2015 00:07:52 +0000 (00:07 +0000)]
[WinEH] Correctly handle inlined __finally blocks with captures
We should also teach the inliner to collapse framerecover of
frameaddress of the current frame down to an alloca, but that can happen
later.
llvm-svn: 235459
David Blaikie [Tue, 21 Apr 2015 23:26:57 +0000 (23:26 +0000)]
[opaque pointer type] Avoid using PointerType::getElementType for a few cases of CallInst
Calls to llvm::Value::mutateType are becoming extra-sensitive now that
instructions have extra type information that will not be derived from
operands or result type (alloca, gep, load, call/invoke, etc... ). The
special-handling for mutateType will get more complicated as this work
continues - it might be worth making mutateType virtual & pushing the
complexity down into the classes that need special handling. But with
only two significant uses of mutateType (vectorization and linking) this
seems OK for now.
Totally open to ideas/suggestions/improvements, of course.
With this, and a bunch of exceptions, we can roundtrip an indirect call
site through bitcode and IR. (a direct call site is actually trickier...
I haven't figured out how to deal with the IR deserializer's lazy
construction of Function/GlobalVariable decl's based on the type of the
entity which means looking through the "pointer to T" type referring to
the global)
llvm-svn: 235458
NAKAMURA Takumi [Tue, 21 Apr 2015 23:14:33 +0000 (23:14 +0000)]
Remove a zero-length file of llvm/test/Transforms/InstCombine/descale-zero.ll.
llvm-svn: 235457
Hans Wennborg [Tue, 21 Apr 2015 23:12:03 +0000 (23:12 +0000)]
Expose -fdiagnostics-parseable-fixits to clang-cl
Patch by Daniel Cheng!
Differential Revision: http://reviews.llvm.org/D9175
llvm-svn: 235456
Wei Mi [Tue, 21 Apr 2015 23:02:15 +0000 (23:02 +0000)]
Limiting gep merging to fix the performance problem described in
https://llvm.org/bugs/show_bug.cgi?id=23163.
Gep merging sometimes behaves like a reverse CSE/LICM optimization,
which has negative impact on performance. In this patch we restrict
gep merging to happen only when the indexes to be merged are both consts,
which ensures such merge is always beneficial.
The patch makes gep merging only happen in very restrictive cases.
It is possible that some analysis/optimization passes rely on the merged
geps to get better result, and we havn't notice them yet. We will be ready
to further improve it once we see the cases.
Differential Revision: http://reviews.llvm.org/D8911
llvm-svn: 235455
Chaoren Lin [Tue, 21 Apr 2015 23:00:58 +0000 (23:00 +0000)]
Same issue as in D8685 but for i386.
llvm-svn: 235454
Wei Mi [Tue, 21 Apr 2015 22:56:09 +0000 (22:56 +0000)]
Revert r235451 since it is attached to a wrong Differential Revision. Sorry.
llvm-svn: 235453
Artem Belevich [Tue, 21 Apr 2015 22:55:54 +0000 (22:55 +0000)]
[cuda] Allow using integral non-type template parameters as launch_bounds attribute arguments.
- Changed CUDALaunchBounds arguments from integers to Expr* so they can
be saved in AST for instantiation.
- Added support for template instantiation of launch_bounds attrubute.
- Moved evaluation of launch_bounds arguments to NVPTXTargetCodeGenInfo::
SetTargetAttributes() where it can be done after template instantiation.
- Added a warning on negative launch_bounds arguments.
- Amended test cases.
Differential Revision: http://reviews.llvm.org/D8985
llvm-svn: 235452
Wei Mi [Tue, 21 Apr 2015 22:37:09 +0000 (22:37 +0000)]
Limiting gep merging to fix the performance problem described in
https://llvm.org/bugs/show_bug.cgi?id=23163.
Gep merging sometimes behaves like a reverse CSE/LICM optimizations,
which has negative impact on performance. In this patch we restrict
gep merging to happen only when the indexes to be merged are both consts,
which ensures such merge is always beneficial.
The patch makes gep merging only happen in very restrictive cases.
It is possible that some analysis/optimization passes rely on the merged
geps to get better result, and we havn't notice them yet. We will be ready
to further improve it once we see the cases.
Differential Revision: http://reviews.llvm.org/D9007
llvm-svn: 235451
Andrew Kaylor [Tue, 21 Apr 2015 22:29:38 +0000 (22:29 +0000)]
Silence MSVC build alignment warnings
llvm-svn: 235450
Artem Belevich [Tue, 21 Apr 2015 22:14:13 +0000 (22:14 +0000)]
[cuda] Added support for CUDA built-in variables.
Added cuda_builtin_vars.h which implements built-in CUDA variables
using __declattr(property).
Fields of built-in variables (except for warpSize) are implemented
using __declattr(property) which replaces read/write of a member field
with a call to a getter/setter member function, in this case with
appropriate NVPTX builtin.
Added a test case to check diagnostics on attempt to construct or
improperly access a built-in variable.
Differential Revision: http://reviews.llvm.org/D9064
llvm-svn: 235448
Enrico Granata [Tue, 21 Apr 2015 22:09:12 +0000 (22:09 +0000)]
Add properties to SBExecutionContext to access the several entities it stores in a more Pythonic fashion
llvm-svn: 235447
Richard Smith [Tue, 21 Apr 2015 21:46:32 +0000 (21:46 +0000)]
[modules] Move list of exported module macros from IdentifierInfo lookup table to separate storage, adjacent to the macro directive history.
This is substantially simpler, provides better space usage accounting in bcanalyzer,
and gives a more compact representation. No functionality change intended.
llvm-svn: 235420
Ahmed Bougacha [Tue, 21 Apr 2015 21:28:33 +0000 (21:28 +0000)]
[MemCpyOpt] Use the raw i8* dest when optimizing memset+memcpy.
MemIntrinsic::getDest() looks through pointer casts, and using it
directly when building the new GEP+memset results in stuff like:
%0 = getelementptr i64* %p, i32 16
%1 = bitcast i64* %0 to i8*
call ..memset(i8* %1, ...)
instead of the correct:
%0 = bitcast i64* %p to i8*
%1 = getelementptr i8* %0, i32 16
call ..memset(i8* %1, ...)
Instead, use getRawDest, which just gives you the i8* value.
While there, use the memcpy's dest, as it's live anyway.
In most cases, when the optimization triggers, the memset and memcpy
sizes are the same, so the built memset is 0-sized and eliminated.
The problem occurs when they're different.
Fixes a regression caused by r235232: PR23300.
llvm-svn: 235419
Krzysztof Parzyszek [Tue, 21 Apr 2015 21:28:03 +0000 (21:28 +0000)]
[Hexagon] Patterns for frame index with offset for isel
llvm-svn: 235418
Lang Hames [Tue, 21 Apr 2015 21:15:35 +0000 (21:15 +0000)]
Update comment. NFC.
llvm-svn: 235417
Daniel Berlin [Tue, 21 Apr 2015 21:11:50 +0000 (21:11 +0000)]
Revamp PredIteratorCache interface to be cleaner.
Summary:
This lets us use range based for loops.
Reviewers: chandlerc
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9169
llvm-svn: 235416
Jingyue Wu [Tue, 21 Apr 2015 20:47:15 +0000 (20:47 +0000)]
[NVPTX] do not run DCE after SLSR and SeparateConstOffsetFromGEP
Summary:
With D9096 and D9101, there's no need to run DCE after SLSR and
SeparateConstOffsetFromGEP.
Test Plan: no regression
Reviewers: jholewinski, meheff
Subscribers: jholewinski, llvm-commits
Differential Revision: http://reviews.llvm.org/D9172
llvm-svn: 235415
Sanjoy Das [Tue, 21 Apr 2015 20:42:50 +0000 (20:42 +0000)]
[LSR][NFC] Remove a stale comment.
The comment was made stale in r171735.
llvm-svn: 235414
Duncan P. N. Exon Smith [Tue, 21 Apr 2015 20:07:38 +0000 (20:07 +0000)]
DebugInfo: Remove DIArray and DITypeArray typedefs
Remove the `DIArray` and `DITypeArray` typedefs, preferring the
underlying types (`DebugNodeArray` and `MDTypeRefArray`, respectively).
llvm-svn: 235413
Duncan P. N. Exon Smith [Tue, 21 Apr 2015 20:07:29 +0000 (20:07 +0000)]
DebugInfo: Prepare for removal of DIArray and DITypeArray typedefs
An upcoming LLVM commit will remove the `DIArray` and `DITypeArray`
typedefs that shadow `DebugNodeArray` and `MDTypeRefArray`,
respectively. Use those types directly.
llvm-svn: 235412
Sergey Matveev [Tue, 21 Apr 2015 20:05:10 +0000 (20:05 +0000)]
[sanitizer] Plug a memory leak in symbolization code.
llvm-svn: 235411
Jingyue Wu [Tue, 21 Apr 2015 19:56:18 +0000 (19:56 +0000)]
[SLSR] garbage-collect unused instructions
Summary:
After we rewrite a candidate, the instructions used by the old form may
become unused. This patch cleans up these unused instructions so that we
needn't run DCE after SLSR.
Test Plan: removed -dce in all the SLSR tests
Reviewers: broune, meheff
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9101
llvm-svn: 235410
Jingyue Wu [Tue, 21 Apr 2015 19:53:18 +0000 (19:53 +0000)]
[SeparateConstOffsetFromGEP] garbage-collect intermediate instructions
Summary: so that we needn't run DCE after this pass.
Test Plan: removed -dce from the commandline in split-gep.ll and split-gep-and-gvn.ll
Reviewers: meheff
Subscribers: llvm-commits, HaoLiu, hfinkel, jholewinski
Differential Revision: http://reviews.llvm.org/D9096
llvm-svn: 235409
Yaron Keren [Tue, 21 Apr 2015 19:25:11 +0000 (19:25 +0000)]
Remove FilesToRemove->push_back(Filename) from sys::DontRemoveFileOnSignal.
llvm-svn: 235408
Duncan P. N. Exon Smith [Tue, 21 Apr 2015 19:17:20 +0000 (19:17 +0000)]
DebugInfo: Use MDType in DITypeIdentifierMap
Use `MDType*` instead of `MDNode*` in `DITypeIdentifierMap`, since all
the members should be types.
llvm-svn: 235407
Daniel Berlin [Tue, 21 Apr 2015 19:13:02 +0000 (19:13 +0000)]
Move IDF Calculation to a separate file, expose an interface to it.
Summary:
MemorySSA uses this algorithm as well, and this enables us to reuse the code in both places.
There are no actual algorithm or datastructure changes in here, just code movement.
Reviewers: qcolombet, chandlerc
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9118
llvm-svn: 235406
Duncan P. N. Exon Smith [Tue, 21 Apr 2015 19:00:26 +0000 (19:00 +0000)]
DebugInfo: Prune unnecessary forward declarations
Probably these forward declarations were once useful, but they certainly
don't belong here now.
llvm-svn: 235405
Duncan P. N. Exon Smith [Tue, 21 Apr 2015 18:44:06 +0000 (18:44 +0000)]
DebugInfo: Drop rest of DIDescriptor subclasses
Delete the remaining subclasses of (the already deleted) `DIDescriptor`.
Part of PR23080.
llvm-svn: 235404
Duncan P. N. Exon Smith [Tue, 21 Apr 2015 18:43:54 +0000 (18:43 +0000)]
DebugInfo: Prepare for deletion of DIDescriptor subclasses
An upcoming LLVM commit will delete all the remaining subclasses of (the
already deleted) `DIDescriptor`. Stop using them.
llvm-svn: 235403
Alexey Samsonov [Tue, 21 Apr 2015 18:38:31 +0000 (18:38 +0000)]
Recommend to use CMake on the Clang Getting Started page.
llvm-svn: 235402
Artem Belevich [Tue, 21 Apr 2015 18:36:42 +0000 (18:36 +0000)]
Revert r235398 "[cuda] Added support for CUDA built-in variables."
r235398 was causing buildbot break due to missing Makefile changes.
llvm-svn: 235401
Duncan P. N. Exon Smith [Tue, 21 Apr 2015 18:24:23 +0000 (18:24 +0000)]
DebugInfo: Assert dbg.declare/value insts are valid
Remove early returns for when `getVariable()` is null, and just assert
that it never happens. The Verifier already confirms that there's a
valid variable on these intrinsics, so we should assume the debug info
isn't broken. I also updated a check for a `!dbg` attachment, which the
Verifier similarly guarantees.
llvm-svn: 235400
Reid Kleckner [Tue, 21 Apr 2015 18:23:57 +0000 (18:23 +0000)]
Re-land r235154-r235156 under the existing -sehprepare flag
Keep the old SEH fan-in lowering on by default for now, since projects
rely on it. This will make it easy to test this change with a simple
flag flip.
llvm-svn: 235399
Artem Belevich [Tue, 21 Apr 2015 17:39:06 +0000 (17:39 +0000)]
[cuda] Added support for CUDA built-in variables.
Added cuda_builtin_vars.h which implements built-in CUDA variables
using __declattr(property).
Fields of built-in variables (except for warpSize) are implemented
using __declattr(property) which replaces read/write of a member field
with a call to a getter/setter member function, in this case with
appropriate NVPTX builtin.
Added a test case to check diagnostics on attempt to construct or
improperly access a built-in variable.
Differential Revision: http://reviews.llvm.org/D9064
llvm-svn: 235398
Ulrich Weigand [Tue, 21 Apr 2015 17:29:35 +0000 (17:29 +0000)]
Implement target-specific __attribute__((aligned)) value
The GCC construct __attribute__((aligned)) is defined to set alignment
to "the default alignment for the target architecture" according to
the GCC documentation:
The default alignment is sufficient for all scalar types, but may not be
enough for all vector types on a target that supports vector operations.
The default alignment is fixed for a particular target ABI.
clang currently hard-coded an alignment of 16 bytes for that construct,
which is correct on some platforms (including X86), but wrong on others
(including SystemZ). Since this value is ABI-relevant, it is important
to get correct for compatibility purposes.
This patch adds a new TargetInfo member "DefaultAlignForAttributeAligned"
that targets can set to the appropriate default __attribute__((aligned))
value.
Note that I'm deliberately *not* using the existing "SuitableAlign"
value, which is used to set the pre-defined macro __BIGGEST_ALIGNMENT__,
since those two values may not be the same on all platforms. In fact,
on X86, __attribute__((aligned)) always uses 16-byte alignment, while
__BIGGEST_ALIGNMENT__ may be larger if AVX-2 or AVX-512 are supported.
(This is actually not yet correctly implemented in clang either.)
The patch provides a value for DefaultAlignForAttributeAligned only for
SystemZ, and leaves the default for all other targets at 16, which means
no visible change in behavior on all other targets. (The value is still
wrong for some other targets, but I'd prefer to leave it to the target
maintainers for those platforms to fix.)
llvm-svn: 235397
Ulrich Weigand [Tue, 21 Apr 2015 17:27:59 +0000 (17:27 +0000)]
Provide alignment info on LLVM external symbols
Code in CodeGenModule::GetOrCreateLLVMGlobal that sets up GlobalValue
object for LLVM external symbols has this comment:
// FIXME: This code is overly simple and should be merged with other global
// handling.
One part does seems to be "overly simple" currently is that this code
never sets any alignment info on the GlobalValue, so that the emitted
IR does not have any align attribute on external globals. This can
lead to unnecessarily inefficient code generation.
This patch adds a GV->setAlignment call to set alignment info.
llvm-svn: 235396
Ulrich Weigand [Tue, 21 Apr 2015 17:26:18 +0000 (17:26 +0000)]
Fix __alignof__ of global variables on SystemZ
SystemZ prefers to align all global variables to two bytes, which is
implemented by setting the TargetInfo member MinGlobalAlign.
However, for compatibility with existing compilers this should *not*
change the ABI alignment value as retrieved via __alignof__, which
it currently does.
This patch fixes the issue by having ASTContext::getDeclAlign ignore
the MinGlobalAlign setting in the ForAlignof case.
Since SystemZ is the only platform setting MinGlobalAlign, this should
cause no change for any other target.
llvm-svn: 235395
Matthias Braun [Tue, 21 Apr 2015 17:21:36 +0000 (17:21 +0000)]
X86: Match for X86ISD nodes in LowerBUILD_VECTOR instead of BUILD_VECTORCombine
There doesn't seem to be a reason to perform this target ISD node matching
in an DAGCombine, moving it to lowering fixes PR23296.
Differential Revision: http://reviews.llvm.org/D9137
llvm-svn: 235394
Simon Atanasyan [Tue, 21 Apr 2015 17:02:49 +0000 (17:02 +0000)]
[ELF] Change return type of the OutputSection::flags() method
This method returns _flags field which has `uint64_t` type. Now the
field's type and the method's return type are consistent.
llvm-svn: 235393
Simon Atanasyan [Tue, 21 Apr 2015 17:02:42 +0000 (17:02 +0000)]
[ELF] Narrow down the type of OutputSection::_sections container
The `OutputSection::appendSection()` method always gets a pointer
to the `Section` class descendants. So it is not necessary to keep them
in the vector of `Chunk` pointers.
No functional changes.
llvm-svn: 235392
Simon Atanasyan [Tue, 21 Apr 2015 17:02:35 +0000 (17:02 +0000)]
[Mips] Write .reginfo section into the separate PT_MIPS_REGINFO segment
llvm-svn: 235391
Simon Atanasyan [Tue, 21 Apr 2015 17:02:27 +0000 (17:02 +0000)]
[ELF] Simplify class memeber function declaration
No functional changes.
llvm-svn: 235390
Bob Wilson [Tue, 21 Apr 2015 16:32:02 +0000 (16:32 +0000)]
Minor edits to the llvm-cov documentation.
This just changes a few places to use a slightly more formal style.
llvm-svn: 235389
Davide Italiano [Tue, 21 Apr 2015 16:06:17 +0000 (16:06 +0000)]
Placate clang. lldb can build on FreeBSD with -Werror again.
llvm-svn: 235387
Reid Kleckner [Tue, 21 Apr 2015 15:56:21 +0000 (15:56 +0000)]
s/NULL/nullptr/ in OrcTestCommon.h to silence -Wsentinel in clang-cl
llvm-svn: 235386
Elena Demikhovsky [Tue, 21 Apr 2015 14:38:31 +0000 (14:38 +0000)]
AVX-512: Added VPMOVx2M instructions for SKX,
fixed encoding of VPMOVM2x.
llvm-svn: 235385
Tobias Grosser [Tue, 21 Apr 2015 14:28:02 +0000 (14:28 +0000)]
Remove target triples from test cases
I just learned that target triples prevent test cases to be run on other
architectures. Polly test cases are until now sufficiently target independent
to not require any target triples. Hence, we drop them.
llvm-svn: 235384
Elena Demikhovsky [Tue, 21 Apr 2015 13:13:46 +0000 (13:13 +0000)]
AVX-512: Added VPTESTM and VPTESTNM instructions for SKX
llvm-svn: 235383
Toma Tabacu [Tue, 21 Apr 2015 11:50:52 +0000 (11:50 +0000)]
[mips] [IAS] Implement the .asciiz directive.
Summary:
This directive is exactly the same as .asciz, except it's only used by MIPS.
It is used to store null terminated strings in object files.
Reviewers: rafael, dsanders, echristo
Reviewed By: dsanders, echristo
Subscribers: echristo, llvm-commits
Differential Revision: http://reviews.llvm.org/D7530
llvm-svn: 235382
Tobias Grosser [Tue, 21 Apr 2015 11:42:01 +0000 (11:42 +0000)]
Fix some formatting issues
llvm-svn: 235381
Tobias Grosser [Tue, 21 Apr 2015 11:37:25 +0000 (11:37 +0000)]
Rename 'scattering' to 'schedule'
In Polly we used both the term 'scattering' and the term 'schedule' to describe
the execution order of a statement without actually distinguishing between them.
We now uniformly use the term 'schedule' for the execution order. This
corresponds to the terminology of isl.
History: CLooG introduced the term scattering as the generated code can be used
as a sequential execution order (schedule) or as a parallel dimension
enumerating different threads of execution (placement). In Polly and/or isl the
term placement was never used, but we uniformly refer to an execution order as a
schedule and only later introduce parallelism. When doing so we do not talk
about about specific placement dimensions.
llvm-svn: 235380
Jozef Kolek [Tue, 21 Apr 2015 11:17:25 +0000 (11:17 +0000)]
[mips][microMIPSr6] Implement CACHE and PREF instructions
Implement CACHE and PREF instructions using mapping.
Differential Revision: http://reviews.llvm.org/D8893
llvm-svn: 235379
Tobias Grosser [Tue, 21 Apr 2015 11:01:34 +0000 (11:01 +0000)]
Pass an actual schedule to the isl_union_access [NFC]
This change is a step towards using a single isl_schedule object throughout
Polly. At the moment the schedule is just constructed from the flat
isl_union_map that defines the schedule. Later we will obtain it directly
from the scop and potentially obtain a schedule with a non-trivial internal
structure that will allow faster dependence analysis.
llvm-svn: 235378
Vasileios Kalintiris [Tue, 21 Apr 2015 10:53:57 +0000 (10:53 +0000)]
[mips] Cleanup old floating-point flag conditions definitions. NFC.
Reviewers: dsanders
Differential Revision: http://reviews.llvm.org/D7947
llvm-svn: 235377
Vasileios Kalintiris [Tue, 21 Apr 2015 10:49:03 +0000 (10:49 +0000)]
[mips] Optimize code generation for 64-bit variable shift instructions.
Summary:
The 64-bit version of the variable shift instructions uses the
shift_rotate_reg class which uses a GPR32Opnd to specify the variable
shift amount. With this patch we avoid the generation of a redundant
SLL instruction for the variable shift instructions in 64-bit targets.
Reviewers: dsanders
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D7413
llvm-svn: 235376
Elena Demikhovsky [Tue, 21 Apr 2015 10:27:40 +0000 (10:27 +0000)]
AVX-512: Added logical and arithmetic instructions for SKX
by Asaf Badouh (asaf.badouh@intel.com)
llvm-svn: 235375
Tobias Grosser [Tue, 21 Apr 2015 09:17:52 +0000 (09:17 +0000)]
Drop unneccessary -basicaa passes in DependenceInfo test cases
llvm-svn: 235374
Tobias Grosser [Tue, 21 Apr 2015 08:47:29 +0000 (08:47 +0000)]
Move to isl's new dependence analysis interface [NFC]
isl_union_map_compute_flow() has been replaced by
isl_union_access_info_compute_flow(). This change does not intend to
change funcitonality, yet. However, it will allow us to pass in subsequent
changes schedule trees to the dependence analysis instead of flat schedules.
This should speed up dependence analysis for important cases significantly.
llvm-svn: 235373
Simon Pilgrim [Tue, 21 Apr 2015 08:40:22 +0000 (08:40 +0000)]
[X86][SSE] Provide execution domains for scalar floating point operations
This is an updated version of Chandler's patch D7402 that got accepted but never committed, and has bit-rotted a bit since.
I've updated the execution domain declarations to match the approach of the packed templates and also added some extra scalar unary tests.
Differential Revision: http://reviews.llvm.org/D9095
llvm-svn: 235372
Simon Pilgrim [Tue, 21 Apr 2015 08:05:43 +0000 (08:05 +0000)]
CONCAT_VECTOR of BUILD_VECTOR - minor fix
Fixed issue with the combine of CONCAT_VECTOR of 2 BUILD_VECTOR nodes - the optimisation wasn't ensuring that the scalar operands of both nodes were the same type/size for implicit truncation.
Test case spotted by Patrik Hagglund
llvm-svn: 235371
Pawel Bylica [Tue, 21 Apr 2015 06:28:36 +0000 (06:28 +0000)]
Fix generic shift expansion when shift amount is 0
Summary:
This fixes http://llvm.org/bugs/show_bug.cgi?id=16439.
This is one possible way to approach this. The other would be to split InL>>(nbits-Amt) into (InL>>(nbits-1-Amt))>>1, which is also valid since since we only need to care about Amt up nbits-1. It's hard to tell which one is better since the shift might be expensive if this stage of expansion is not yet a legal machine integer, whereas comparisons with zero are relatively cheap at all sizes, but more expensive than a shift if the shift is on a legal machine type.
Patch by Keno Fischer!
Test Plan: regression test from http://reviews.llvm.org/D7752
Reviewers: chfast, resistor
Reviewed By: chfast, resistor
Subscribers: sanjoy, resistor, chfast, llvm-commits
Differential Revision: http://reviews.llvm.org/D4978
llvm-svn: 235370
Matthias Braun [Tue, 21 Apr 2015 01:35:42 +0000 (01:35 +0000)]
Enhanced vim support.
This brings the utils/vim folder into a more vim-like format by moving
the syntax hightlighting files into a syntax subdirectory. It adds
some minimal settings that everyone should agree on to ftdetect/ftplugin and
features a new indentation plugin for .ll files.
llvm-svn: 235369
Ying Chen [Tue, 21 Apr 2015 01:15:47 +0000 (01:15 +0000)]
XFAIL tests that are failed on linux with gcc-4.9.2
Summary:
- add decorator functions to xfail and skip test on specific os, architecture and version of comipler
- xfail failing test with gcc-4.9.2 on linux
- add one usage of skipIf function
Test Plan:
Run tests with different archs, and version of compilers to verify decorator function working as expected
Run tests with gcc-4.9.2 and no failure reported
Reviewers: sivachandra, ovyalov, vharron, chaoren
Subscribers: lldb-commits
Differential Revision: http://reviews.llvm.org/D8851
llvm-svn: 235368
Matthias Braun [Tue, 21 Apr 2015 01:13:41 +0000 (01:13 +0000)]
X86: Do not select X86 custom vector nodes if operand types don't match
X86ISD::ADDSUB, X86ISD::(F)HADD, X86ISD::(F)HSUB should not be selected
if the operand types do not match the result type because vector type
legalization cannot deal with this for custom nodes.
Testcase X86ISD::ADDSUB is attached. I could not create a testcase for
the FHADD/FHSUB cases because of: https://llvm.org/bugs/show_bug.cgi?id=23296
Differential Revision: http://reviews.llvm.org/D9120
llvm-svn: 235367
Derek Schuff [Tue, 21 Apr 2015 00:17:59 +0000 (00:17 +0000)]
Tighten bundling section alignment test.
Leftover comment from http://reviews.llvm.org/D9131
llvm-svn: 235366
Derek Schuff [Tue, 21 Apr 2015 00:14:25 +0000 (00:14 +0000)]
[MC] When using bundle aligment, align sections to bundle size
Summary:
Bundle aligment requires that the functions always start at an aligned address.
Usually this is ensured by the compiler, but assembly code does not always
begin with a .align directive.
This change ensures that sections get the correct alignment if they contain
any instructions and bundling is enabled. (It also makes LLVM match the
behavior of GNU as).
Differential Revision: http://reviews.llvm.org/D9131
llvm-svn: 235365
Fiona Glaser [Tue, 21 Apr 2015 00:05:41 +0000 (00:05 +0000)]
InstCombine: fold (sitofp (zext x)) to (uitofp x)
This is okay because the zext guarantees the high bit is zero,
and so the value is unsigned.
llvm-svn: 235364
Pirama Arumuga Nainar [Mon, 20 Apr 2015 23:54:41 +0000 (23:54 +0000)]
Fix flakiness in fp16-promote.ll
Summary:
In the f16-promote test, make the checks for native conversion instructions
similar to the libcall checks:
- Remove hard coded register names
- Do not check exact instruction sequences.
This fixes test flakiness due to non-determinism in instruction
scheduling and register allocation. I also fixed a few minor things in
the CHECK-LIBCALL checks.
I'll try to find a way to check that unnecessary loads, stores, or
conversions don't happen.
Reviewers: mzolotukhin, srhines, ab
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9112
llvm-svn: 235363
JF Bastien [Mon, 20 Apr 2015 23:42:22 +0000 (23:42 +0000)]
bugpoint Enhancement.
Summary:
This patch adds two flags to `bugpoint`: "-replace-funcs-with-null" and "-disable-pass-list-reduction".
When "-replace-funcs-with-null" is specified, bugpoint will, instead of simply deleting function bodies, replace all uses of functions and then will delete functions completely from the test module, correctly handling aliasing and @llvm.used && @llvm.compiler.used. This part was conceived while trying to debug the PNaCl IR simplification passes, which don't allow undefined functions (ie no declarations).
With "-disable-pass-list-reduction", bugpoint won't try to reduce the set of passes causing the "crash". This is needed in cases where one is trying to debug an issue inside the PNaCl IR simplification passes which is causing an PNaCl ABI verification error, for example.
Reviewers: jfb
Reviewed By: jfb
Subscribers: jfb, llvm-commits
Differential Revision: http://reviews.llvm.org/D8555
llvm-svn: 235362
Sanjay Patel [Mon, 20 Apr 2015 23:31:53 +0000 (23:31 +0000)]
use update_llc_test_checks.py to tighten checking
Also, replace win and linux runs with a generic run because that
makes no difference in what this test is checking.
llvm-svn: 235361
David Majnemer [Mon, 20 Apr 2015 23:19:10 +0000 (23:19 +0000)]
Add myself as the Constant Folder owner.
llvm-svn: 235360
Davide Italiano [Mon, 20 Apr 2015 22:54:50 +0000 (22:54 +0000)]
--discard-locals should be --discard-all here.
Sorry for the mistake/confusion.
llvm-svn: 235359
Andrew Kaylor [Mon, 20 Apr 2015 22:53:42 +0000 (22:53 +0000)]
[WinEH] Fix problem with landing pad return values used in PHI nodes during outlining.
llvm-svn: 235358
Davide Italiano [Mon, 20 Apr 2015 22:52:56 +0000 (22:52 +0000)]
[GNU/ELF] Add support for -X/--discard-locals.
There's (almost) never need to keep .L symbols around for production
builds. In fact, the FreeBSD kernel explicitly specify -X beacuse the
size impact (and the subsequent performance impact) might be significant,
because we keep symbols in memory.
I was tempted to make this the default, but I haven't (yet).
PR: 23232
llvm-svn: 235357
Duncan P. N. Exon Smith [Mon, 20 Apr 2015 22:10:08 +0000 (22:10 +0000)]
DebugInfo: Delete subclasses of DIScope
Delete subclasses of (the already defunct) `DIScope`, updating users to
use the raw pointers from the `Metadata` hierarchy directly.
llvm-svn: 235356
Duncan P. N. Exon Smith [Mon, 20 Apr 2015 22:09:57 +0000 (22:09 +0000)]
DebugInfo: Prepare for deletion of subclasses of DIScope
Prepare for the deletion in LLVM of the subclasses of (the already
deleted) `DIScope` by using the raw pointers they were wrapping
directly.
llvm-svn: 235355
Andrew Kaylor [Mon, 20 Apr 2015 22:04:09 +0000 (22:04 +0000)]
[WinEH] Fix problem with mapping shared empty handler blocks.
Differential Revision: http://reviews.llvm.org/D9125
llvm-svn: 235354
Duncan P. N. Exon Smith [Mon, 20 Apr 2015 21:29:44 +0000 (21:29 +0000)]
DebugInfo: Fix Kaleidoscope Ch. 8 after r235327
Pretty sure the build was broken by r235327 (I updated it there, but
apparently didn't check if it compiled).
llvm-svn: 235353
Duncan P. N. Exon Smith [Mon, 20 Apr 2015 21:17:32 +0000 (21:17 +0000)]
DebugInfo: Delete old subclasses of DIType
Delete subclasses of (the already deleted) `DIType` in favour of
directly using pointers from the `Metadata` hierarchy.
While `DICompositeType` wraps `MDCompositeTypeBase` and `DIDerivedType`
wraps `MDDerivedTypeBase`, most uses of each really meant the more
specific `MDCompositeType` and `MDDerivedType`.
llvm-svn: 235351
Duncan P. N. Exon Smith [Mon, 20 Apr 2015 21:17:26 +0000 (21:17 +0000)]
DebugInfo: Prepare for deletion of subclasses of DIType
Subclasses of (the already deleted) `DIType` will be deleted by an
upcoming LLVM commit. Remove references.
While `DICompositeType` wraps `MDCompositeTypeBase` and `DIDerivedType`
wraps `MDDerivedTypeBase`, most uses of each really meant the more
specific `MDCompositeType` and `MDDerivedType`. I updated accordingly.
llvm-svn: 235350
Duncan P. N. Exon Smith [Mon, 20 Apr 2015 21:04:33 +0000 (21:04 +0000)]
DwarfUnit: Split MDSubroutineType version of constructTypeDIE()
The version of `constructTypeDIE()` for `MDSubroutineType` is unrelated
to (and has different callers than) the `MDCompositeType`. Split the
two in half.
This simplifies an upcoming patch to delete `DICompositeType`. There
shouldn't be any real functionality change here. `createTypeDIE()` is
`cast<>`'ing where it didn't need to before, but that function in turn
is only called for true `MDCompositeType`s.
llvm-svn: 235349
Samuel Benzaquen [Mon, 20 Apr 2015 20:58:50 +0000 (20:58 +0000)]
Add conversionDecl matcher for node CXXConversionDecl.
llvm-svn: 235348