Silviu Baranga [Wed, 9 Dec 2015 15:03:52 +0000 (15:03 +0000)]
[LV][LAA] Add a layer over SCEV to apply run-time checked knowledge on SCEV expressions
Summary:
This change creates a layer over ScalarEvolution for LAA and LV, and centralizes the
usage of SCEV predicates. The SCEVPredicatedLayer takes the statically deduced knowledge
by ScalarEvolution and applies the knowledge from the SCEV predicates. The end goal is
that both LAA and LV should use this interface everywhere.
This also solves a problem involving the result of SCEV expression rewritting when
the predicate changes. Suppose we have the expression (sext {a,+,b}) and two predicates
P1: {a,+,b} has nsw
P2: b = 1.
Applying P1 and then P2 gives us {a,+,1}, while applying P2 and the P1 gives us
sext({a,+,1}) (the AddRec expression was changed by P2 so P1 no longer applies).
The SCEVPredicatedLayer maintains the order of transformations by feeding back
the results of previous transformations into new transformations, and therefore
avoiding this issue.
The SCEVPredicatedLayer maintains a cache to remember the results of previous
SCEV rewritting results. This also has the benefit of reducing the overall number
of expression rewrites.
Reviewers: mzolotukhin, anemet
Subscribers: jmolloy, sanjoy, llvm-commits
Differential Revision: http://reviews.llvm.org/D14296
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255115
91177308-0d34-0410-b5e6-
96231b3b80d8
Robert Lougher [Wed, 9 Dec 2015 14:34:10 +0000 (14:34 +0000)]
Fix cycle in selection DAG introduced by extractelement legalization
During selection DAG legalization, extractelement is replaced with a load
instruction. To do this, a temporary store to the stack is used unless an
existing store is found that can be re-used.
If re-using a store, the chain going out of the store must be replaced by
the one going out of the new load (this ensures that any stores that must
take place after the store happens after the load, else the value might
be overwritten before it is loaded).
The problem is, if the extractelement index is dependent on the store
replacing the chain will introduce a cycle in the selection DAG (the load
uses the index, and by replacing the chain we will make the index dependent
on the load).
To fix this, if the index is dependent on the store, the store is skipped.
This is conservative as we may end up creating an unnecessary extra store
to the stack. However, the situation is not expected to occur very often.
Differential Revision: http://reviews.llvm.org/D15330
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255114
91177308-0d34-0410-b5e6-
96231b3b80d8
Oliver Stannard [Wed, 9 Dec 2015 14:32:11 +0000 (14:32 +0000)]
[AArch64] Fix FP16 vector instructions that should only accept low registers
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255113
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Wed, 9 Dec 2015 13:48:05 +0000 (13:48 +0000)]
[mips][ias] Range check uimm10 operands
Summary:
Reviewers: vkalintiris
Subscribers: dsanders, llvm-commits
Differential Revision: http://reviews.llvm.org/D15229
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255112
91177308-0d34-0410-b5e6-
96231b3b80d8
JF Bastien [Wed, 9 Dec 2015 13:29:32 +0000 (13:29 +0000)]
WebAssembly: add known failures
The bots are now running the torture tests properly. Bin all failures from the GCC C torture tests so that we can tackle failures and make the tree go red on regressions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255111
91177308-0d34-0410-b5e6-
96231b3b80d8
Vasileios Kalintiris [Wed, 9 Dec 2015 13:24:22 +0000 (13:24 +0000)]
[mips] Use multiclass patterns for f32/f64 comparisons and i32 selects.
Summary:
Although the multiclass for i32 selects might seem redundant as it has
only one instantiation, we will use it to replace the correspondent
patterns in Mips64r6InstrInfo.td in follow-up commits.
Reviewers: dsanders
Subscribers: llvm-commits, dsanders
Differential Revision: http://reviews.llvm.org/D14612
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255110
91177308-0d34-0410-b5e6-
96231b3b80d8
Zlatko Buljan [Wed, 9 Dec 2015 13:07:45 +0000 (13:07 +0000)]
Revert r254897 "[mips][microMIPS] Implement LH, LHE, LHU and LHUE instructions"
Commited patch was intended to implement LH, LHE, LHU and LHUE instructions.
After commit test-suite failed with error message in the form of:
fatal error: error in backend: Cannot select: t124: i32,ch = load<LD2[%d](tbaa=<0x94acc48>), sext from i16> t0, t2, undef:i32
For that reason I decided to revert commit r254897 and make new patch which besides implementation and standard regression tests will also have dedicated tests (CodeGen) for the above error.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255109
91177308-0d34-0410-b5e6-
96231b3b80d8
JF Bastien [Wed, 9 Dec 2015 09:05:42 +0000 (09:05 +0000)]
EarlyCSE: fix typo from rL255054.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255102
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Wed, 9 Dec 2015 08:17:42 +0000 (08:17 +0000)]
Revert "Implement a new pass - LiveDebugValues - to compute the set of live DEBUG_VALUEs at each basic block and insert them. Reviewed and accepted at: reviews.llvm.org/D11933"
This reverts commit r255096.
Break the bots: http://lab.llvm.org:8080/green/job/clang-stage1-cmake-RA-incremental_check/16378/
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255101
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Wed, 9 Dec 2015 08:17:35 +0000 (08:17 +0000)]
The current importing scheme is processing one function at a time,
loading the source Module, linking the function in the destination
module, and destroying the source Module before repeating with the
next function to import (potentially from the same Module).
Ideally we would keep the source Module alive and import the next
Function needed from this Module. Unfortunately this is not possible
because the linker does not leave it in a usable state.
However we can do better by first computing the list of all candidates
per Module, and only then load the source Module and import all the
function we need for it.
The trick to process callees is to materialize function in the source
module when building the list of function to import, and inspect them
in their source module, collecting the list of callees for each
callee.
When we move the the actual import, we will import from each source
module exactly once. Each source module is loaded exactly once.
The only drawback it that it requires to have all the lazy-loaded
source Module in memory at the same time.
Currently this patch already improves considerably the link time,
a multithreaded link of llvm-dis on my laptop was:
real 1m12.175s user 6m32.430s sys 0m10.529s
and is now:
real 0m40.697s user 2m10.237s sys 0m4.375s
Note: this is the full link time (linker+Import+Optimizer+CodeGen)
Differential Revision: http://reviews.llvm.org/D15178
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255100
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram TV [Wed, 9 Dec 2015 05:49:14 +0000 (05:49 +0000)]
Implement a new pass - LiveDebugValues - to compute the set of live DEBUG_VALUEs at each basic block and insert them. Reviewed and accepted at: reviews.llvm.org/D11933
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255096
91177308-0d34-0410-b5e6-
96231b3b80d8
Vikram TV [Wed, 9 Dec 2015 05:16:24 +0000 (05:16 +0000)]
Test commit access - Fix few missing '.' in comments of LoopInterchange code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255095
91177308-0d34-0410-b5e6-
96231b3b80d8
Steven Wu [Wed, 9 Dec 2015 03:37:51 +0000 (03:37 +0000)]
Fix the order of destructors in LibLTOCodeGenerator
Summary:
The order of destructors in LTOCodeGenerator gets changed in r254696.
It is possible for LTOCodeGenerator to have a MergedModule created in
the OwnedContext, in which case the module must be destructed before
the context.
Reviewers: rafael, dexonsmith
Subscribers: llvm-commits, joker.eph
Differential Revision: http://reviews.llvm.org/D15346
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255092
91177308-0d34-0410-b5e6-
96231b3b80d8
Ahmed Bougacha [Wed, 9 Dec 2015 01:19:50 +0000 (01:19 +0000)]
[AArch64][ARM] Don't base interleaved op legality on type alloc size.
Otherwise, we think that most types that look like they'd fit in a
legal vector type are legal (so, basically, *any* vector type with a
size between 33 and 128 bits, I think, since we use pow2 alignment;
e.g., v2i25, v3f32, ...).
DataLayout::getTypeAllocSize rounds up based on alignment.
When checking for target intrinsic legality, that's not what we want:
if rounding makes a difference, the type isn't legal, and the
target intrinsics shouldn't be used, as they are always assumed legal.
One could make the argument that alloc size is ultimately the most
relevant here, since we're dealing with LD/ST intrinsics. That's only
true if we did legalize them though; that's a problem for another day.
Use DataLayout::getTypeSizeInBits instead of getTypeAllocSizeInBits.
Type::getSizeInBits can't be used because that'd gratuitously break
pointer vector support.
Some of these uses are currently fine, because we only hit them when
the type is already known legal (e.g., r114454). Update them for
consistency. It's faster to avoid the rounding anyway!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255089
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Wed, 9 Dec 2015 01:01:28 +0000 (01:01 +0000)]
Don't drop attributes when inlining through "deopt" operand bundles
Test case attached (test case also checks that we don't drop the calling
convention, but that functionality was correct before this patch).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255088
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 9 Dec 2015 00:55:05 +0000 (00:55 +0000)]
Simplify testMergedProgram.
It now receives and returns std::unique_ptr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255087
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 9 Dec 2015 00:51:06 +0000 (00:51 +0000)]
Simplify memory management. NFC.
This passes std::unique_ptr to predicates that are expected to delete
their argument.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255086
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 9 Dec 2015 00:34:10 +0000 (00:34 +0000)]
Return std::unique_ptr from SplitFunctionsOutOfModule. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255084
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 9 Dec 2015 00:18:41 +0000 (00:18 +0000)]
Simplify memory management. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255082
91177308-0d34-0410-b5e6-
96231b3b80d8
Vyacheslav Klochkov [Wed, 9 Dec 2015 00:12:13 +0000 (00:12 +0000)]
X86-FMA3: Defined the ExeDomain property for Scalar FMA3 opcodes.
Reviewer: Simon Pilgrim.
Differential Revision: http://reviews.llvm.org/D15317
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255080
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Wed, 9 Dec 2015 00:08:22 +0000 (00:08 +0000)]
Simplify memory management a bit. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255079
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Tue, 8 Dec 2015 23:57:17 +0000 (23:57 +0000)]
Return a std::unique_ptr from CloneModule. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255078
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 23:52:58 +0000 (23:52 +0000)]
[IndVars] Use any_of and foreach instead of explicit for loops; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255077
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 23:16:52 +0000 (23:16 +0000)]
[OperandBundles] Have PruneEH work correct with operand bundles.
For an invoke with operand bundles, the [op_begin(), op_end()-3] range
can contain things other than invoke arguments. This change teaches
PruneEH to use arg_begin() and arg_end() explicitly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255073
91177308-0d34-0410-b5e6-
96231b3b80d8
Pirama Arumuga Nainar [Tue, 8 Dec 2015 23:07:06 +0000 (23:07 +0000)]
Define selection for v4f16, v8f16 scalar_to_vector
Summary:
This fixes failure when trying to select
insertelement <4 x half> undef, half %a, i64 0
which gets transformed to a scalar_to_vector node.
The accompanying v4 and v8 tests fail instruction selection without this
patch.
Reviewers: ab, jmolloy
Subscribers: srhines, llvm-commits
Differential Revision: http://reviews.llvm.org/D15322
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255072
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Tue, 8 Dec 2015 23:04:19 +0000 (23:04 +0000)]
Fix/Improve Debug print in FunctionImport pass
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255071
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Kleckner [Tue, 8 Dec 2015 23:00:03 +0000 (23:00 +0000)]
[CGP] Reimplement r255055 a different way
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255070
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 22:53:36 +0000 (22:53 +0000)]
[SCEV] Use for-each; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255069
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Tue, 8 Dec 2015 22:39:40 +0000 (22:39 +0000)]
Remove caching in FunctionImport: a Module can't be reused after being linked from
The Linker destroys the source module (API change coming to make it explicit)
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255064
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Kleckner [Tue, 8 Dec 2015 22:33:23 +0000 (22:33 +0000)]
Revert "[CGP] Check that we have an insert point before moving llvm.dbg.value around"
This reverts commit r255055.
Breakage has been reported.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255063
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 22:26:08 +0000 (22:26 +0000)]
[OperandBundles] Fix a transform in simplifycfg
Reviewers: pcc, majnemer, reames
Subscribers: reames, llvm-commits
Differential Revision: http://reviews.llvm.org/D15345
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255062
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Tue, 8 Dec 2015 22:17:11 +0000 (22:17 +0000)]
[X86][AVX] Fold loads + splats into broadcast instructions
On AVX and AVX2, BROADCAST instructions can load a scalar into all elements of a target vector.
This patch improves the lowering of 'splat' shuffles of a loaded vector into a broadcast - currently the lowering only works for cases where we are splatting the zero'th element, which is now generalised to any element.
Fix for PR23022
Differential Revision: http://reviews.llvm.org/D15310
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255061
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Bieneman [Tue, 8 Dec 2015 21:51:48 +0000 (21:51 +0000)]
[CMake] Ignore externalizing debuginfo for unit tests
If you externalize debug info for unit tests the test runner finds the mach-o inside the dsym bundle and tries to execute it as a test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255056
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Kleckner [Tue, 8 Dec 2015 21:50:52 +0000 (21:50 +0000)]
[CGP] Check that we have an insert point before moving llvm.dbg.value around
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255055
91177308-0d34-0410-b5e6-
96231b3b80d8
Philip Reames [Tue, 8 Dec 2015 21:45:41 +0000 (21:45 +0000)]
[EarlyCSE] Value forwarding for unordered atomics
This patch teaches the fully redundant load part of EarlyCSE how to forward from atomic and volatile loads and stores, and how to eliminate unordered atomics (only). This patch does not include dead store elimination support for unordered atomics, that will follow in the near future.
The basic idea is that we allow all loads and stores to be tracked by the AvailableLoad table. We store a bit in the table which tracks whether load/store was atomic, and then only replace atomic loads with ones which were also atomic.
No attempt is made to refine our handling of ordered loads or stores. Those are still treated as full fences. We could pretty easily extend the release fence handling to release stores, but that should be a separate patch.
Differential Revision: http://reviews.llvm.org/D15337
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255054
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Tue, 8 Dec 2015 21:43:41 +0000 (21:43 +0000)]
[X86][SSE4A] Added fast-isel intrinsics tests
As discussed on PR24580, this patch adds fast-isel codegen tests to match the IR generated in clang/test/CodeGen/sse4a-builtins.c
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255053
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Tue, 8 Dec 2015 21:32:08 +0000 (21:32 +0000)]
[X86][SSSE3] Added fast-isel intrinsics tests
As discussed on PR24580, this patch adds fast-isel codegen tests to match the IR generated in clang/test/CodeGen/ssse3-builtins.c
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255052
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Tue, 8 Dec 2015 21:27:19 +0000 (21:27 +0000)]
[X86][SSE3] Added fast-isel intrinsics tests
As discussed on PR24580, this patch adds fast-isel codegen tests to match the IR generated in clang/test/CodeGen/sse3-builtins.c
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255051
91177308-0d34-0410-b5e6-
96231b3b80d8
Artyom Skrobov [Tue, 8 Dec 2015 19:59:01 +0000 (19:59 +0000)]
Fix ARMv4T (Thumb1) epilogue generation
Summary:
Before ARMv5T, Thumb1 code could not pop PC, as described at D14357 and D14986;
so we need the special fixup in the epilogue.
Reviewers: jroelofs, qcolombet
Subscribers: aemerson, llvm-commits, rengolin
Differential Revision: http://reviews.llvm.org/D15126
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255047
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Tue, 8 Dec 2015 19:13:31 +0000 (19:13 +0000)]
Revert "Add Available Externally linkage type to isWeakForLinker()"
This reverts r255043, as per post-review concern were raised on the correctness.
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255045
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Tue, 8 Dec 2015 19:02:55 +0000 (19:02 +0000)]
Cleanup test: remove useless alignment
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255044
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Tue, 8 Dec 2015 19:01:29 +0000 (19:01 +0000)]
Add Available Externally linkage type to isWeakForLinker()
Per LangRef: "Globals with available_externally linkage are
allowed to be discarded at will, and are otherwise the same
as linkonce_odr", since linkonce_odr is in this list it makes
sense to have available_externally there as well.
Reviewers: rafael
Differential Revision: http://reviews.llvm.org/D15323
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255043
91177308-0d34-0410-b5e6-
96231b3b80d8
Tim Northover [Tue, 8 Dec 2015 18:31:35 +0000 (18:31 +0000)]
X86: produce more friendly errors during MachO relocation handling
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255036
91177308-0d34-0410-b5e6-
96231b3b80d8
Renato Golin [Tue, 8 Dec 2015 18:10:58 +0000 (18:10 +0000)]
[ARM] Allowing SP/PC for AND/BIC mod_imm_not
AND/BIC instructions do accept SP/PC, so the register class should be
more generic (rGPR -> GPR) to cope with that case. Adding more tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255034
91177308-0d34-0410-b5e6-
96231b3b80d8
Mike Aizatsky [Tue, 8 Dec 2015 17:44:51 +0000 (17:44 +0000)]
adding readability-identifier-naming to llvm clang-tidy configuration.
Differential Revision: http://reviews.llvm.org/D15196
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255028
91177308-0d34-0410-b5e6-
96231b3b80d8
Ron Lieberman [Tue, 8 Dec 2015 16:28:32 +0000 (16:28 +0000)]
[Hexagon] Add NewValueJump support for C4_cmpneq, C4_cmplte, C4_cmplteu
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255027
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Tue, 8 Dec 2015 14:54:49 +0000 (14:54 +0000)]
Move all private members together. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255021
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Tue, 8 Dec 2015 14:42:10 +0000 (14:42 +0000)]
[mips][ias] Range check uimm8 operands
Summary:
Reviewers: vkalintiris
Subscribers: llvm-commits, dsanders
Differential Revision: http://reviews.llvm.org/D15226
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255018
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Tue, 8 Dec 2015 13:49:19 +0000 (13:49 +0000)]
[mips][ias] Range check uimm6 operands and fix a bug this revealed.
Summary:
We don't check the size operand on ext/dext*/ins/dins* yet because the
permitted range depends on the pos argument and we can't check that using
this mechanism.
The bug was that dextu/dinsu accepted 0..31 in the pos operand instead of 32..63.
Reviewers: vkalintiris
Subscribers: llvm-commits, dsanders
Differential Revision: http://reviews.llvm.org/D15190
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255015
91177308-0d34-0410-b5e6-
96231b3b80d8
Asaf Badouh [Tue, 8 Dec 2015 12:34:34 +0000 (12:34 +0000)]
[x86][avx512] more changes in intrinsics to be align with gcc format
Differential Revision: http://reviews.llvm.org/D15329
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255011
91177308-0d34-0410-b5e6-
96231b3b80d8
Oliver Stannard [Tue, 8 Dec 2015 12:16:10 +0000 (12:16 +0000)]
[AArch64] Add ARMv8.2-A FP16 vector instructions
ARMv8.2-A adds 16-bit floating point versions of all existing SIMD
floating-point instructions. This is an optional extension, so all of
these instructions require the FeatureFullFP16 subtarget feature.
Note that VFP without SIMD is not a valid combination for any version of
ARMv8-A, but I have ensured that these instructions all depend on both
FeatureNEON and FeatureFullFP16 for consistency.
The ".2h" vector type specifier is now legal (for the scalar pairwise
reduction instructions), so some unrelated tests have been modified as
different error messages are emitted. This is not a problem as the
invalid operands are still caught.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255010
91177308-0d34-0410-b5e6-
96231b3b80d8
Michael Zuckerman [Tue, 8 Dec 2015 12:00:24 +0000 (12:00 +0000)]
dding test for fnstsw
continue of Wrong FNSTSW size operator
url: http://reviews.llvm.org/D14953
Differential Revision: http://reviews.llvm.org/D15155
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255007
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 04:32:54 +0000 (04:32 +0000)]
[SCEV] Move some struct declarations inside functions; NFC
Reduces the scope over which the struct is visible, making its usages
obvious. I did not move structs in cases where this wasn't a clear
win (the struct is too large, or is grouped in some other interesting
way).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255003
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 04:32:51 +0000 (04:32 +0000)]
[SCEV] Fix indentation; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255002
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 03:50:32 +0000 (03:50 +0000)]
[OperandBundles] Remove unncessary constructor
The StringRef constructor is unnecessary (since we're converting to
std::string anyway), and having it requires an explicit call to
StringRef's or std::string's constructor.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@255000
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 8 Dec 2015 03:43:03 +0000 (03:43 +0000)]
[WebAssembly] Fix a typo in a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254999
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 8 Dec 2015 03:42:50 +0000 (03:42 +0000)]
[WebAssembly] Remove an unneeded static_cast.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254998
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 8 Dec 2015 03:36:00 +0000 (03:36 +0000)]
[WebAssembly] Fix an emacs syntax highlighting comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254997
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 8 Dec 2015 03:33:51 +0000 (03:33 +0000)]
[WebAssembly] Convert a file-level comment to doxygen style.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254996
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 8 Dec 2015 03:30:42 +0000 (03:30 +0000)]
[WebAssembly] Assert MRI.isSSA() in passes that depend on SSA form.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254995
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 8 Dec 2015 03:25:35 +0000 (03:25 +0000)]
[WebAssembly] Trim some unneeded #includes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254994
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Tue, 8 Dec 2015 03:22:33 +0000 (03:22 +0000)]
[WebAssembly] Remove the override of haveFastSqrt.
The default implementation in BasicTTI already checks TLI and does
the right thing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254993
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Tue, 8 Dec 2015 03:01:16 +0000 (03:01 +0000)]
IR: Allow vectors of halfs to be ConstantDataVectors
Currently, vectors of halfs end up as ConstantVectors, but there isn't
a good reason they can't be ConstantDataVectors. This should save some
memory.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254991
91177308-0d34-0410-b5e6-
96231b3b80d8
Davide Italiano [Tue, 8 Dec 2015 02:45:59 +0000 (02:45 +0000)]
[llvm-objdump/MachO] Don't cut'n'paste the same code over and over.
Use the appropriate helper instead.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254990
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Tue, 8 Dec 2015 02:38:14 +0000 (02:38 +0000)]
Add a test showing that we internalize lazily linked GVs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254989
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Tue, 8 Dec 2015 02:37:48 +0000 (02:37 +0000)]
AsmPrinter: Use emitGlobalConstantFP to emit elements of constant data
It's strange to duplicate the logic for emitting FP values into
emitGlobalConstantDataSequential, and it's even stranger that we end
up printing the verbose assembly comments differently between the two
paths. Just call into emitGlobalConstantFP rather than crudely
duplicating its logic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254988
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Tue, 8 Dec 2015 02:29:45 +0000 (02:29 +0000)]
Simplify test. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254987
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Tue, 8 Dec 2015 00:14:38 +0000 (00:14 +0000)]
[CXX TLS calling convention] Add support for AArch64.
rdar://
9001553
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254978
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 00:13:21 +0000 (00:13 +0000)]
[IndVars] Have getInsertPointForUses preserve LCSSA
Summary:
Also add a stricter post-condition for IndVarSimplify.
Fixes PR25578. Test case by Michael Zolotukhin.
Reviewers: hfinkel, atrick, mzolotukhin
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D15059
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254977
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 00:13:17 +0000 (00:13 +0000)]
[SCEVExpander] Have hoistIVInc preserve LCSSA
Summary:
(Note: the problematic invocation of hoistIVInc that caused PR24804 came
from IndVarSimplify, not from SCEVExpander itself)
Fixes PR24804. Test case by David Majnemer.
Reviewers: hfinkel, majnemer, atrick, mzolotukhin
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D15058
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254976
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Tue, 8 Dec 2015 00:13:12 +0000 (00:13 +0000)]
Add Instruction::getFunction; NFC
Will be used in a upcoming patch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254975
91177308-0d34-0410-b5e6-
96231b3b80d8
Philip Reames [Tue, 8 Dec 2015 00:10:56 +0000 (00:10 +0000)]
[PassManager] Tuning Memory Usage of AnalysisUsage
We were using unneccessarily large initial sizes for these SmallVectors. This was wasting around 50kb of memory for the O3 pipeline, even after the uniquing changes. We're still using around 20kb which is a bit much, but it's definitely better. This is about a 6% improvement in total O3 memory usage.
Note: The raw data on structure size which were used to pick these thresholds can be found in the review thread.
Differential Revision: http://reviews.llvm.org/D15244
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254974
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Mon, 7 Dec 2015 23:34:30 +0000 (23:34 +0000)]
fix return values to match bool return type; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254968
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 7 Dec 2015 23:32:39 +0000 (23:32 +0000)]
Factor two calls to a common location.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254967
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Mon, 7 Dec 2015 23:15:57 +0000 (23:15 +0000)]
Stabilize llvm/test/Object/archive-update.test a bit.
A manipulation (in this case, mkdir) can make slack between creating and touching %t.older/evenlen.
I would make this rewrote with python if this were still unstable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254965
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Mon, 7 Dec 2015 23:12:26 +0000 (23:12 +0000)]
Support: Teach Asan about BumpPtrAllocator
Based on patch by Pete Cooper.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254964
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Mon, 7 Dec 2015 23:07:16 +0000 (23:07 +0000)]
Let llvm-lto installed. A few tests in clang/test are using it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254963
91177308-0d34-0410-b5e6-
96231b3b80d8
Philip Reames [Mon, 7 Dec 2015 22:43:56 +0000 (22:43 +0000)]
Remove debug output that snuck into 254957
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254960
91177308-0d34-0410-b5e6-
96231b3b80d8
Philip Reames [Mon, 7 Dec 2015 22:41:23 +0000 (22:41 +0000)]
Reapply 254950 w/fix
254950 ended up being not NFC. The previous code was overriding the flags for whether an instruction read or wrote memory using the target specific flags returned via TTI. I'd missed this in my refactoring. Since I mistakenly built only x86 and didn't notice the number of unsupported tests, I didn't catch that before the original checkin.
This raises an interesting issue though. Given we have function attributes (i.e. readonly, readnone, argmemonly) which describe the aliasing of intrinsics, why does TTI have this information overriding the instruction definition at all? I see no reason for this, but decided to preserve existing behavior for the moment. The root issue might be that we don't have a "writeonly" attribute.
Original commit message:
[EarlyCSE] Simplify and invert ParseMemoryInst [NFCI]
Restructure ParseMemoryInst - which was introduced to abstract over target specific load and stores instructions - to just query the underlying instructions. In theory, this could be slightly slower than caching the results, but in practice, it's very unlikely to be measurable.
The simple query scheme makes it far easier to understand, and much easier to extend with new queries. Given I'm about to need to add new query types, doing the cleanup first seemed worthwhile.
Do we still believe the target specific intrinsic handling is worthwhile in EarlyCSE? It adds quite a bit of complexity and makes the code harder to read. Being able to delete the abstraction entirely would be wonderful.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254957
91177308-0d34-0410-b5e6-
96231b3b80d8
Mehdi Amini [Mon, 7 Dec 2015 22:27:19 +0000 (22:27 +0000)]
Remove useless hack that avoids calling LLVMLinkInInterpreter()
This is supposed to force-link the Interpreter, by inserting a dead
call to LLVMLinkInInterpreter().
Since it is actually an empty function, there is no reason for the
call to be dead.
From: Mehdi Amini <mehdi.amini@apple.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254956
91177308-0d34-0410-b5e6-
96231b3b80d8
Philip Reames [Mon, 7 Dec 2015 21:41:29 +0000 (21:41 +0000)]
Revert 254950
It's causing test failures on AArch64. Due to a bad build config on my part, I apparently wasn't running the tests I thought I was.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254954
91177308-0d34-0410-b5e6-
96231b3b80d8
Manman Ren [Mon, 7 Dec 2015 21:40:09 +0000 (21:40 +0000)]
Update doc for C++ TLS calling convention.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254953
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 7 Dec 2015 21:28:22 +0000 (21:28 +0000)]
Simplify the error handling a bit. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254952
91177308-0d34-0410-b5e6-
96231b3b80d8
Philip Reames [Mon, 7 Dec 2015 21:27:15 +0000 (21:27 +0000)]
[EarlyCSE] Simplify and invert ParseMemoryInst [NFCI]
Restructure ParseMemoryInst - which was introduced to abstract over target specific load and stores instructions - to just query the underlying instructions. In theory, this could be slightly slower than caching the results, but in practice, it's very unlikely to be measurable.
The simple query scheme makes it far easier to understand, and much easier to extend with new queries. Given I'm about to need to add new query types, doing the cleanup first seemed worthwhile.
Do we still believe the target specific intrinsic handling is worthwhile in EarlyCSE? It adds quite a bit of complexity and makes the code harder to read. Being able to delete the abstraction entirely would be wonderful.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254950
91177308-0d34-0410-b5e6-
96231b3b80d8
Easwaran Raman [Mon, 7 Dec 2015 21:21:20 +0000 (21:21 +0000)]
Use updated threshold for indirect call bonus
When considering foo->bar inlining, if there is an indirect call in foo which gets resolved to a direct call (say baz), then we try to inline baz into bar with a threshold T and subtract max(T - Cost(bar->baz), 0) from Cost(foo->bar). This patch uses max(Threshold(bar->baz) - Cost(bar->baz)) instead, where Thresheld(bar->baz) could be different from T due to bonuses or subtractions. Threshold(bar->baz) - Cost(bar->baz) better represents the desirability of inlining baz into bar.
Differential Revision: http://reviews.llvm.org/D14309
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254945
91177308-0d34-0410-b5e6-
96231b3b80d8
Kit Barton [Mon, 7 Dec 2015 20:50:29 +0000 (20:50 +0000)]
[PPC64] Convert bool literals to i32
Convert i1 values to i32 values if they should be allocated in GPRs instead of CRs.
Phabricator: http://reviews.llvm.org/D14064
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254942
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Mon, 7 Dec 2015 20:36:00 +0000 (20:36 +0000)]
Fix line endings
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254939
91177308-0d34-0410-b5e6-
96231b3b80d8
Teresa Johnson [Mon, 7 Dec 2015 19:53:38 +0000 (19:53 +0000)]
Fix function return type in declaration (bot errors from r254926).
Try to fix bot build errors from r254926 by correcting the function
return type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254934
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Mon, 7 Dec 2015 19:31:34 +0000 (19:31 +0000)]
don't repeat function names in comments; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254930
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Mon, 7 Dec 2015 19:21:39 +0000 (19:21 +0000)]
fix 'the the '; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254928
91177308-0d34-0410-b5e6-
96231b3b80d8
Teresa Johnson [Mon, 7 Dec 2015 19:21:11 +0000 (19:21 +0000)]
[ThinLTO] Support for specifying function index from pass manager
Summary:
Add a field on the PassManagerBuilder that clang or gold can use to pass
down a pointer to the function index in memory to use for importing when
the ThinLTO backend is triggered. Add support to supply this to the
function import pass.
Reviewers: joker.eph, dexonsmith
Subscribers: davidxl, llvm-commits, joker.eph
Differential Revision: http://reviews.llvm.org/D15024
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254926
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Mon, 7 Dec 2015 19:13:40 +0000 (19:13 +0000)]
remove redundant check: optForSize() includes a check for the minsize attribute; NFCI
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254925
91177308-0d34-0410-b5e6-
96231b3b80d8
Ron Lieberman [Mon, 7 Dec 2015 18:52:39 +0000 (18:52 +0000)]
[Hexagon] Adding v60 test, vasr in particular.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254923
91177308-0d34-0410-b5e6-
96231b3b80d8
David Blaikie [Mon, 7 Dec 2015 18:46:41 +0000 (18:46 +0000)]
[llvm-dwp] Restructure inputs for test case so they're all grouped together
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254922
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Mon, 7 Dec 2015 17:39:48 +0000 (17:39 +0000)]
Tighten checks so we can see existing codegen
The 2-element vector case shows a surprising bug: we failed to
eliminate ops on undefs, so there are 4 fmax calls even though
there can only be 2 valid elements in the inputs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254920
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Mon, 7 Dec 2015 17:35:56 +0000 (17:35 +0000)]
[Orc] Removing traces of takeOwnershipOfBuffers left after r251560.
Patch by Joshua Gerrard. Thanks Joshua!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254919
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 7 Dec 2015 16:31:41 +0000 (16:31 +0000)]
Link declaration lazily.
We already linked available_externally and linkonce lazily, this just
adds declarations to the list.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254917
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 7 Dec 2015 16:01:40 +0000 (16:01 +0000)]
Simplify test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254916
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Ballman [Mon, 7 Dec 2015 15:44:34 +0000 (15:44 +0000)]
Silence all C4592 warnings with MSVC 2015 Update 1. This warning produces false positives that Microsoft says will be fixed in Update 2. Until this produces reliable diagnostics, it is safe to disable the diagnostic -- the compiler is not doing anything different than it previously did aside from issuing the diagnostic.
(Note, this silences at least one false positive in LLVM with FeatureBitset uses.)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254915
91177308-0d34-0410-b5e6-
96231b3b80d8
Teresa Johnson [Mon, 7 Dec 2015 15:05:44 +0000 (15:05 +0000)]
[ThinLTO] Support cloning of temporary DILocation metadata
This is needed to support linking of module-level metadata as a
postpass after function importing, where we will be leaving temporary
metadata on imported instructions until the postpass metadata import.
Also added unittest. Split from D14838.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254914
91177308-0d34-0410-b5e6-
96231b3b80d8