summaryrefslogtreecommitdiffstats
path: root/llvm
Commit message (Collapse)AuthorAgeFilesLines
...
* Add AVX512 masked leadz instrinsic support.Cameron McInally2014-06-113-1/+65
| | | | llvm-svn: 210652
* Improve the test for inlining of __no_debug__ functions.Evgeniy Stepanov2014-06-111-1/+1
| | | | llvm-svn: 210645
* [X86] Refactor the logic to select horizontal adds/subs to a helper function.Andrea Di Biagio2014-06-112-108/+201
| | | | | | | | | | | | | | | This patch moves part of the logic implemented by the target specific combine rules added at r210477 to a separate helper function. This should make easier to add more rules for matching AVX/AVX2 horizontal adds/subs. This patch also fixes a problem caused by a wrong check performed on indices of extract_vector_elt dag nodes in input to the scalar adds/subs. New tests have been added to verify that we correctly check indices of extract_vector_elt dag nodes when selecting a horizontal operation. llvm-svn: 210644
* Create macro INITIALIZE_TM_PASS.Jiangning Liu2014-06-114-42/+19
| | | | | | | | Pass initialization requires to initialize TargetMachine for back-end specific passes. This commit creates a new macro INITIALIZE_TM_PASS to simplify this kind of initialization. llvm-svn: 210641
* Global merge for global symbols.Jiangning Liu2014-06-119-75/+284
| | | | | | | | This commit is to improve global merge pass and support global symbol merge. The global symbol merge is not enabled by default. For aarch64, we need some more back-end fix to make it really benifit ADRP CSE. llvm-svn: 210640
* Rename global-merge to enable-global-merge.Jiangning Liu2014-06-113-9/+9
| | | | llvm-svn: 210639
* Convert StringMapEntry::Create to use StringRef instead of start/end ↵Craig Topper2014-06-114-18/+15
| | | | | | pointers. Simpliies all in tree call sites. No functional change. llvm-svn: 210638
* Try to fix the msvc build.Rafael Espindola2014-06-111-1/+2
| | | | llvm-svn: 210636
* Uses generic_category instead of system_category.Rafael Espindola2014-06-112-25/+25
| | | | | | | Some c++ libraries (libstdc++ at least) don't seem to map to the generic category in in the system_category's default_error_condition. llvm-svn: 210635
* MC: add enumeration of WinEH data encodingSaleem Abdulrasool2014-06-112-2/+21
| | | | | | | | | | | | | | | Most Windows platforms use auxiliary data for unwinding. This information is stored in the .pdata section. The encoding format for the data differs between architectures and Windows variants. Windows MIPS and Alpha use identical formats; Alpha64 is the same with different widths. Windows x86_64 and Itanium share the representation. All Windows CE entries are identical irrespective of the architecture. ARMv7 (Windows [NT] on ARM) has its own format. This enumeration will become the differentiator once the windows EH emission infrastructure is generalised, allowing us to emit the necessary unwinding information for Windows on ARM. llvm-svn: 210634
* MC: clang-format MCAsmInfoSaleem Abdulrasool2014-06-111-528/+459
| | | | | | | Apply clang-format over the header. Reformat the docs to current LLVM style. NFC. llvm-svn: 210633
* Remove windows_error.Rafael Espindola2014-06-115-101/+50
| | | | | | | | | | MSVC doesn't seem to provide any is_error_code_enum enumeration for the windows errors. Fortunately very few places in llvm have to handle raw windows errors, so we can just construct the corresponding error_code directly. llvm-svn: 210631
* There is no posix_category in std, use generic_category.Rafael Espindola2014-06-113-17/+3
| | | | llvm-svn: 210630
* Use cast instead of assert + dyn_castMatt Arsenault2014-06-111-3/+2
| | | | llvm-svn: 210628
* R600: Add helper functions.Matt Arsenault2014-06-112-0/+21
| | | | | | | Extract these from some of my other patches, since this is the only thing really making them dependent on each other. llvm-svn: 210627
* Use an enum class.Rafael Espindola2014-06-112-22/+13
| | | | llvm-svn: 210623
* CodeGen: refactor DwarfExceptionSaleem Abdulrasool2014-06-119-206/+201
| | | | | | | | | | | | DwarfException served as a base class for exception handling directive emission. However, this is also used by other exception models (e.g. Win64EH). Rename this class to EHStreamer and split it out of DwarfException.h. NFC. Use the opportunity to fix up some of the documentation comments to match current LLVM style. Also rename some functions to conform better with current LLVM coding style. llvm-svn: 210622
* Use an enum class.Rafael Espindola2014-06-112-18/+9
| | | | llvm-svn: 210620
* Remove duplicate copy of InstrItineraryData from the TargetMachine,Eric Christopher2014-06-112-13/+8
| | | | | | it's already on the subtarget. llvm-svn: 210619
* Move to a private function to initialize the subtarget dependenciesEric Christopher2014-06-112-11/+20
| | | | | | so that we can use initializer lists for the AArch64 Subtarget. llvm-svn: 210616
* Move to a private function to initialize the subtarget dependenciesEric Christopher2014-06-112-22/+25
| | | | | | so that we can use initializer lists for the X86Subtarget. llvm-svn: 210614
* Sort includes.Eric Christopher2014-06-111-1/+1
| | | | llvm-svn: 210613
* [FastISel][X86] Extend support for {s|u}{add|sub|mul}.with.overflow intrinsics.Juergen Ributzka2014-06-103-33/+429
| | | | llvm-svn: 210610
* Use unique_ptr for X86Subtarget pointer members.Eric Christopher2014-06-102-22/+16
| | | | llvm-svn: 210606
* Move AArch64TargetLowering to AArch64Subtarget.Eric Christopher2014-06-106-9/+13
| | | | | | | This currently necessitates a TargetMachine for the TargetLowering constructor and TLOF. llvm-svn: 210605
* Revert "Remove support for runtime multi-threading."Zachary Turner2014-06-108-43/+80
| | | | | | This reverts revision r210600. llvm-svn: 210603
* Remove support for runtime multi-threading.Zachary Turner2014-06-108-80/+43
| | | | | | | | | | | | | | | | | | | | | This patch removes the functions llvm_start_multithreaded() and llvm_stop_multithreaded(), and changes llvm_is_multithreaded() to return a constant value based on the value of the compile-time definition LLVM_ENABLE_THREADS. Previously, it was possible to have compile-time support for threads on, and runtime support for threads off, in which case certain mutexes were not allocated or ever acquired. Now, if the build is created with threads enabled, mutexes are always acquired. A test before/after patch of compiling a very large TU showed no noticeable performance impact of this change. Reviewers: rnk Differential Revision: http://reviews.llvm.org/D4076 llvm-svn: 210600
* Move AArch64InstrInfo to AArch64Subtarget.Eric Christopher2014-06-104-6/+9
| | | | llvm-svn: 210599
* Remove a method that was just replacing direct access to a member.Eric Christopher2014-06-102-20/+18
| | | | llvm-svn: 210598
* lit: warn when passed invalid pathnameHans Wennborg2014-06-101-3/+5
| | | | | | | | | | | | | | | It would previously say things like warning: input 'test/Frontend/foo.c' contained no tests and have the user pull their hair trying to figure out what's wrong with that file. This patch changes the message to the much clearer: warning: no such file or directory: 'test/Frontend/foo.c' Differential Revision: http://reviews.llvm.org/D4097 llvm-svn: 210597
* Remove the use of TargetMachine from X86InstrInfo.Eric Christopher2014-06-103-61/+60
| | | | llvm-svn: 210596
* Move X86RegisterInfo away from using the TargetMachine and onlyEric Christopher2014-06-105-33/+30
| | | | | | using the subtarget. llvm-svn: 210595
* Mark a few functions noexcept.Rafael Espindola2014-06-107-15/+17
| | | | | | This reduces the difference between std::error_code and llvm::error_code. llvm-svn: 210591
* Use the TargetMachine on the DAG or the MachineFunction insteadEric Christopher2014-06-101-82/+80
| | | | | | of using the cached TargetMachine. llvm-svn: 210589
* R600/SI: Emit an error when attempting to spill VGPRs v4Tom Stellard2014-06-101-10/+24
| | | | | | | | | | | | | | | | | | I can't get VGPR spilling to work reliable, so for now just emit an error when the register allocator tries to spill VGPRs. v2: - Fix build v3: - Added crash fix when spilling SPGRs v4: - Use V_MOV_B32 as a dummy instruction instead of S_NOP Patch by: Darren Powell https://bugs.freedesktop.org/show_bug.cgi?id=75276 llvm-svn: 210588
* R600/SI: Fix a crash when spilling SGPRsTom Stellard2014-06-102-11/+8
| | | | | | | | | | | We need to make sure only one new instruction is added when spilling otherwise the register allocator may crash. This fixes a crash in the game Antichamber. https://bugs.freedesktop.org/show_bug.cgi?id=75276 llvm-svn: 210587
* Add a LLVM_NOEXCEPT to Compiler.h.Rafael Espindola2014-06-101-0/+6
| | | | | | This will be needed for the switch to std::error_code. llvm-svn: 210581
* We already have a reference to the TargetMachine, use that.Eric Christopher2014-06-101-2/+1
| | | | llvm-svn: 210580
* Have isInTailCallPosition take the DAG so that we can use theEric Christopher2014-06-103-6/+7
| | | | | | | version of TargetLowering/Machine from there on the way to avoiding TargetMachine in TargetLowering. llvm-svn: 210579
* Reorder includes to be sorted.Eric Christopher2014-06-101-1/+1
| | | | llvm-svn: 210578
* Rearrange the CHECK lines in this test to make failure more obvious.Reid Kleckner2014-06-101-17/+16
| | | | llvm-svn: 210575
* Revert "Patch by Ray Donnelly to print register names instead of numbers."Reid Kleckner2014-06-102-15/+8
| | | | | | | | | | | | | This reverts commit r206683. The code was confusing SEH register numbers with DWARF register numbers. The test case it was committed with was obviously incorrect. The disassembler was roundtripping '.seh_pushreg %rsi' as '.seh_pushreg %rbp', and other exciting things. Noticed by Vadim Chugunov. llvm-svn: 210574
* Fix error in tablegen when either operand of !if is an empty list.Matt Arsenault2014-06-103-7/+16
| | | | | | !if([Something], []) would error with "No type for list". llvm-svn: 210572
* Fix typos.Eric Christopher2014-06-102-2/+2
| | | | llvm-svn: 210571
* R600: Use BCNT_INT for evergreenMatt Arsenault2014-06-105-5/+67
| | | | llvm-svn: 210569
* R600/SI: Implement i64 ctpopMatt Arsenault2014-06-106-1/+155
| | | | llvm-svn: 210568
* R600/SI: Use bcnt instruction for ctpopMatt Arsenault2014-06-107-11/+230
| | | | llvm-svn: 210567
* Use an enum class now that they are available.Rafael Espindola2014-06-101-20/+12
| | | | llvm-svn: 210566
* R600: Handle fcopysignMatt Arsenault2014-06-107-3/+114
| | | | llvm-svn: 210564
* R600/SI: Handle sign_extend and zero_extend to i64 with patterns.Matt Arsenault2014-06-104-49/+92
| | | | llvm-svn: 210563
OpenPOWER on IntegriCloud