summaryrefslogtreecommitdiffstats
path: root/llvm
Commit message (Collapse)AuthorAgeFilesLines
* [AVX512] Add 512-bit load/stores to fast isel.Craig Topper2016-06-021-0/+46
| | | | llvm-svn: 271486
* [X86] No need to use 256-bit VMOVNTPS for integer types when only AVX1 is ↵Craig Topper2016-06-021-15/+1
| | | | | | | | supported. VMOVNTDQ is available with AVX1. We were getting this right for v4i64 but not the other integer types. llvm-svn: 271482
* [X86] Add AVX 256-bit load and stores to fast isel.Craig Topper2016-06-022-18/+61
| | | | | | | | I'm not sure why this was missing for so long. This also exposed that we were picking floating point 256-bit VMOVNTPS for some integer types in normal isel for AVX1 even though VMOVNTDQ is available. In practice it doesn't matter due to the execution dependency fix pass, but it required extra isel patterns. Fixing that in a follow up commit. llvm-svn: 271481
* [X86] Use uint16_t for a couple arrays of instruction opcodes. NFCCraig Topper2016-06-021-2/+2
| | | | llvm-svn: 271480
* [AVX512] Remove LOADA/LOADU/STOREA/STOREU intrinsic types now that they are ↵Craig Topper2016-06-022-52/+3
| | | | | | unused. llvm-svn: 271479
* [AVX512] Remove masked load intrinsics. Clang now emits generic masked load ↵Craig Topper2016-06-0211-696/+626
| | | | | | | | intrinsics instead. The intrinsics will be autoupgraded to the same generic masked loads. llvm-svn: 271478
* Add "REQUIRES-ANY" feature testEric Fiselier2016-06-024-3/+19
| | | | | | | | | | | | | | | | Summary: This patch adds a "REQUIRES-ANY" feature test that is disjunctive. This marks a test as `UNSUPPORTED` if none of the specified features are available. Libc++ has the need to write feature test such as `// REQUIRES-ANY: c++98, c++03` when testing of behavior that is specific to older dialects but has since changed. Reviewers: rnk, ddunbar Subscribers: ddunbar, probinson, llvm-commits, cfe-commits Differential Revision: http://reviews.llvm.org/D20757 llvm-svn: 271468
* make icall pass name consistent /NFCXinliang David Li2016-06-023-7/+7
| | | | llvm-svn: 271467
* [libFuzzer] use __sanitizer_print_memory_profile to print the memory profile ↵Kostya Serebryany2016-06-022-1/+5
| | | | | | on OOM llvm-svn: 271465
* Inline isDereferenceableFromAttribute; NFCSanjoy Das2016-06-021-19/+8
| | | | | | | | Now that `Value::getPointerDereferenceableBytes` looks beyond just attributes, the name `isDereferenceableFromAttribute` is misleading. Just inline the function, since it is small and only used once. llvm-svn: 271456
* Remove Value::isPointerDereferenceable; NFCISanjoy Das2016-06-023-42/+21
| | | | | | | | ... and merge into `Value::getPointerDereferenceableBytes`. This was suggested by Artur Pilipenko in D20764 -- since we no longer allow loads of unsized types, there is no need anymore to have this special logic. llvm-svn: 271455
* Try to fix docs build after rL271440Tamas Berghammer2016-06-021-2/+2
| | | | llvm-svn: 271452
* [asan] Rename *UAR* into *UseAfterReturn*Vitaly Buka2016-06-021-7/+7
| | | | | | | | | | | | | | | Summary: To improve readability. PR27453 Reviewers: kcc, eugenis, aizatsky Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D20761 llvm-svn: 271447
* AMDGPU: Fix incorrectly setting kill flag when copying register tuplesMatt Arsenault2016-06-021-1/+1
| | | | | | | This fixes some verifier errors when trackLivenessAfterRegAlloc is enabled. llvm-svn: 271446
* Make MachineCopyPropagation preserve CFGMatt Arsenault2016-06-021-0/+5
| | | | | | This doesn't touch it as far as I can tell. llvm-svn: 271445
* AMDGPU: SIDebuggerInsertNops preserves CFGMatt Arsenault2016-06-022-0/+6
| | | | | | | This saves an additional run of the DominatorTree and MachineLoopInfo llvm-svn: 271444
* cmake: Fix color diags under ninja if the compiler is "AppleClang"Justin Bogner2016-06-011-1/+1
| | | | | | | Apparently cmake differentiates between Clang and AppleClang, and we fail to color our diagnostics if you have the latter. Fix that. llvm-svn: 271442
* Rework r271439. I forgot to save the buffer for editing.NAKAMURA Takumi2016-06-011-1/+1
| | | | llvm-svn: 271441
* Add new LLVM_EXTERNAL_PROJECTS option to cmakeTamas Berghammer2016-06-012-0/+13
| | | | | | | | | | | | | | | | | | | The new option makes it possible to build external projects as part of the llvm build without copying (or symlinking) then into llvm/tool with specifying a few additional cmake variables. Example usage (2 additional project called foo and bar): -DLLVM_EXTERNAL_PROJECTS="Foo;Bar" -DLLVM_EXTERNAL_FOO_SOURCE_DIR=/src/foo -DLLVM_EXTERNAL_BAR_SOURCE_DIR=/src/bar Note: This is the extension of the approach we already support for clang/lldb/poly with adding an option to specify additional supported projects. Differential revision: http://reviews.llvm.org/D20838 llvm-svn: 271440
* MappedBlockStreamTest.cpp: Simplify array initializers.NAKAMURA Takumi2016-06-011-2/+2
| | | | llvm-svn: 271439
* Avoid a load for local functions.Rafael Espindola2016-06-012-2/+17
| | | | llvm-svn: 271437
* [x86, AVX2] regenerate checksSanjay Patel2016-06-011-4/+10
| | | | llvm-svn: 271434
* [MemorySSA] Port to new pass managerGeoff Berry2016-06-0123-124/+132
| | | | | | | | | | | | | | | | | Add support for the new pass manager to MemorySSA pass. Change MemorySSA to be computed eagerly upon construction. Change MemorySSAWalker to be owned by the MemorySSA object that creates it. Reviewers: dberlin, george.burgess.iv Subscribers: mcrosier, llvm-commits Differential Revision: http://reviews.llvm.org/D19664 llvm-svn: 271432
* SDAG: Drop a redundant replace and move the dead node removal closer. NFCJustin Bogner2016-06-011-6/+4
| | | | llvm-svn: 271429
* [DAG] Improve legalization of INSERT_SUBVECTORMichael Kuperstein2016-06-012-115/+55
| | | | | | | | | When the index is known to be constant 0, insert directly into the the low half, instead of spilling, performing the insert in-memory, and reloading. Differential Revision: http://reviews.llvm.org/D20763 llvm-svn: 271428
* [PPC64] Fix SUBFC8 Defs listKeno Fischer2016-06-013-2/+147
| | | | | | | | | | | | | Fix PR27943 "Bad machine code: Using an undefined physical register". SUBFC8 implicitly defines the CR0 register, but this was omitted in the instruction definition. Patch by Jameson Nash <jameson@juliacomputing.com> Reviewers: hfinkel Differential Revision: http://reviews.llvm.org/D20802 llvm-svn: 271425
* [SCEV] Keep SCEVExpander insert points consistent.Geoff Berry2016-06-012-35/+102
| | | | | | | | | | | | | | | | | | | | | Summary: Make sure that the SCEVExpander Builder insert point and any saved/restored insert points are kept consistent (i.e. their Instruction and BasicBlock match) when moving instructions in SCEVExpander. This fixes an issue triggered by http://reviews.llvm.org/D18001 [LSR] Create fewer redundant instructions. Test case will be added in reapply commit of above change: http://reviews.llvm.org/D18480 Reapply [LSR] Create fewer redundant instructions. Reviewers: sanjoy Subscribers: mzolotukhin, sanjoy, qcolombet, mcrosier, llvm-commits Differential Revision: http://reviews.llvm.org/D20703 llvm-svn: 271424
* Revert "Claim NoAlias if two GEPs index different fields of the same struct"Daniel Berlin2016-06-013-149/+83
| | | | | | This reverts commit 2d5d6493f43eb68493a3852b8c226ac9fafdc7eb. llvm-svn: 271422
* [CFLAA] Recognize builtin allocation functions.George Burgess IV2016-06-013-34/+92
| | | | | | | | | | | This patch extends CFLAA to recognize allocation functions such as malloc, free, etc, so we can treat them more aggressively. Patch by Jia Chen. Differential Revision: http://reviews.llvm.org/D20776 llvm-svn: 271421
* [pdb] silence warnings about moving from a temporary.Zachary Turner2016-06-011-2/+2
| | | | llvm-svn: 271420
* [CodeView] Simplify StreamArray operator++David Majnemer2016-06-011-16/+24
| | | | llvm-svn: 271419
* [CodeView] Make sure StreamRef::readBytes doesn't read too muchDavid Majnemer2016-06-012-6/+21
| | | | llvm-svn: 271418
* [PDB] Make ModStream::symbols report errorsDavid Majnemer2016-06-011-1/+2
| | | | llvm-svn: 271417
* [PDB] Silence sign comparison warnings in MappedBlockStreamTestDavid Majnemer2016-06-011-8/+8
| | | | llvm-svn: 271416
* Claim NoAlias if two GEPs index different fields of the same structDaniel Berlin2016-06-013-83/+149
| | | | | | | | | | | | | | Patch by Taewook Oh Summary: Patch for Bug 27478. Make BasicAliasAnalysis claims NoAlias if two GEPs index different fields of the same structure. Reviewers: hfinkel, dberlin Subscribers: dberlin, mcrosier, llvm-commits Differential Revision: http://reviews.llvm.org/D20665 llvm-svn: 271415
* Better fix for PR27903.Than McIntosh2016-06-012-70/+80
| | | | | | | | | | | | | | | | | Summary: Re-enable lifetime-start-on-first-use for stack coloring, but explicitly disable it for slots with more than one start or end lifetime marker. Bug: 27903 Reviewers: wmi, tejohnson, qcolombet, gbiv Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D20739 llvm-svn: 271412
* Fix the NDEBUG buildReid Kleckner2016-06-011-0/+1
| | | | llvm-svn: 271411
* [LV] For some IVs, use vector phis instead of widening in the loop bodyMichael Kuperstein2016-06-019-40/+161
| | | | | | | | | | | | | Previously, whenever we needed a vector IV, we would create it on the fly, by splatting the scalar IV and adding a step vector. Instead, we can create a real vector IV. This tends to save a couple of instructions per iteration. This only changes the behavior for the most basic case - integer primary IVs with a constant step. Differential Revision: http://reviews.llvm.org/D20315 llvm-svn: 271410
* [codeview] Translate basic DITypes to CV type recordsReid Kleckner2016-06-014-34/+655
| | | | | | | | | | | | | | | Summary: This is meant to be the tiniest step towards DIType to CV type index translation that I could come up with. Whenever translation fails, we use type index zero, which is the unknown type. Reviewers: aaboud, zturner Subscribers: llvm-commits, amccarth Differential Revision: http://reviews.llvm.org/D20840 llvm-svn: 271408
* Reduce dependence on pointee types when deducing dereferenceabilitySanjoy Das2016-06-012-72/+117
| | | | | | | | | | | | | | | | | | | | | | | | | | Summary: Change some of the internal interfaces in Loads.cpp to keep track of the number of bytes we're trying to prove dereferenceable using an explicit `Size` parameter. Before this, the `Size` parameter was implicitly inferred from the pointee type of the pointer whose dereferenceability we were trying to prove, causing us to be conservative around bitcasts. This was unfortunate since bitcast instructions are no-ops and should never break optimizations. With an explicit `Size` parameter, we're more precise (as shown in the test cases), and the code is simpler. We should eventually move towards a `DerefQuery` struct that groups together a base pointer, an offset, a size and an alignment; but this patch is a first step. Reviewers: apilipenko, dblaikie, hfinkel, reames Subscribers: mcrosier, llvm-commits Differential Revision: http://reviews.llvm.org/D20764 llvm-svn: 271406
* [libFuzzer] when an invalid flag is given, warn, but don't crashKostya Serebryany2016-06-012-2/+7
| | | | llvm-svn: 271404
* [IR] Disallow loading and storing unsized typesSanjoy Das2016-06-013-13/+40
| | | | | | | | | | | | | | | | Summary: It isn't clear what is the operational meaning of loading or storing an unsized types, since it cannot be lowered into something meaningful. Since there does not seem to be any practical need for it either, make such loads and stores illegal IR. Reviewers: majnemer, chandlerc Subscribers: mcrosier, llvm-commits Differential Revision: http://reviews.llvm.org/D20846 llvm-svn: 271402
* MappedBlockStreamTest.cpp: Appease msc18 to avoid initializer for std::vector.NAKAMURA Takumi2016-06-011-2/+6
| | | | llvm-svn: 271397
* Update usage(), make sure srcdir is really empty.Joerg Sonnenberger2016-06-011-0/+2
| | | | llvm-svn: 271396
* [X86][SSE] Added non-temporal store tests for all 512-bit vector typesSimon Pilgrim2016-06-011-57/+224
| | | | llvm-svn: 271393
* [X86][SSE] Added non-temporal store tests for all 256-bit vector typesSimon Pilgrim2016-06-011-0/+185
| | | | | | Also added KNL AVX-512 checks llvm-svn: 271391
* [X86][SSE] Added non-temporal store tests for all 128-bit integer vector typesSimon Pilgrim2016-06-011-0/+45
| | | | llvm-svn: 271389
* Adding back-end support to two bit scanning intrinsicsMichael Zuckerman2016-06-013-0/+35
| | | | | | | | | | | | | | Adding LLVM back-end support to two intrinsics dealing with bit scan: _bit_scan_forward and _bit_scan_reverse. Their functionality is as described in Intel intrinsics guide: https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_bit_scan_forward&expand=371,370 https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_bit_scan_reverse&expand=371,370 Commit on behalf of Omer Paparo Bivas Differential Revision: http://reviews.llvm.org/D19915 llvm-svn: 271386
* [ARM] Add additional matching for UBFX instructionsOliver Stannard2016-06-012-0/+37
| | | | | | | | | | | This adds an additional matcher to select UBFX(..) from SRL(AND(..)) in ARMISelDAGToDAG to help with code size. Patch by David Green. Differential Revision: http://reviews.llvm.org/D20667 llvm-svn: 271384
* Fix off-by-one error in max integer functionsDylan McKay2016-06-011-4/+4
| | | | | | | | | I recently added these functions, but implemented them poorly. This fixes that. Sorry for the spam. llvm-svn: 271380
OpenPOWER on IntegriCloud