summaryrefslogtreecommitdiffstats
path: root/llvm
Commit message (Collapse)AuthorAgeFilesLines
* [Support/MemoryBuffer] Move the IsVolatile check inside shouldUseMmap() and ↵Argyrios Kyrtzidis2014-05-061-7/+13
| | | | | | | | make sure to zero-initialize the rest of the buffer if we unexpectedly reach end-of-file while reading. llvm-svn: 208021
* Detabify.Nick Lewycky2014-05-061-2/+2
| | | | llvm-svn: 208019
* Improve 'tail' call marking in TRE. A bootstrap of clang goes from 375k ↵Nick Lewycky2014-05-053-73/+269
| | | | | | | | | | calls marked tail in the IR to 470k, however this improvement does not carry into an improvement of the call/jmp ratio on x86. The most common pattern is a tail call + br to a block with nothing but a 'ret'. The number of tail call to loop conversions remains the same (1618 by my count). The new algorithm does a local scan over the use-def chains to identify local "alloca-derived" values, as well as points where the alloca could escape. Then, a visit over the CFG marks blocks as being before or after the allocas have escaped, and annotates the calls accordingly. llvm-svn: 208017
* Walk back commits for unused function parameters - they're still beingEric Christopher2014-05-052-6/+20
| | | | | | used via dragonegg for now. llvm-svn: 208016
* Reapply: Add slp vectorization to LTO passes. The bug it exposed has been ↵Yi Jiang2014-05-051-0/+3
| | | | | | fixed by r207983. <radar://16641956> llvm-svn: 208013
* Remove a now unnecessary function since all calls have one versionEric Christopher2014-05-051-17/+4
| | | | | | and inline it into its caller. llvm-svn: 208012
* Remove a call to std::exit in a library. Make "Help" returnEric Christopher2014-05-051-5/+7
| | | | | | a 0 as a default answer. llvm-svn: 208009
* [Support/MemoryBuffer] Introduce a boolean parameter (false by default) ↵Argyrios Kyrtzidis2014-05-052-24/+50
| | | | | | | | 'IsVolatile' for the open file functions. This provides a hint that the file may be changing often so mmap is avoided. llvm-svn: 208007
* Fix typo.Eric Christopher2014-05-051-1/+1
| | | | llvm-svn: 208006
* R600: Expand i64 ISD:SUBTom Stellard2014-05-052-18/+38
| | | | llvm-svn: 208005
* Remove unused argument from AddFeature.Eric Christopher2014-05-052-4/+3
| | | | llvm-svn: 208002
* Fix typo (also tab character).Eric Christopher2014-05-051-3/+3
| | | | llvm-svn: 208001
* Use a range loop.Rafael Espindola2014-05-051-3/+1
| | | | llvm-svn: 207996
* Revert "Optimize shufflevector that copies an i64/f64 and zeros the rest."Filipe Cabecinhas2014-05-053-37/+12
| | | | | | This reverts commit 207992. I misread the phab number on the LGTM. llvm-svn: 207993
* Optimize shufflevector that copies an i64/f64 and zeros the rest.Filipe Cabecinhas2014-05-053-12/+37
| | | | | | | | | | | | | | Summary: Also ran clang-format on the function. The code added is the last else if block. Reviewers: nadav, craig.topper Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D3518 llvm-svn: 207992
* R600/SI: allow 5 more input SGPRs to a shaderMarek Olsak2014-05-051-1/+1
| | | | | | | Our OpenGL driver needs 22 SGPRs (16 user SGPRs + 6 streamout non-user SGPRs). Signed-off-by: Marek Olšák <marek.olsak@amd.com> llvm-svn: 207990
* Move test from r207969 to another folder and rename it.Michael Zolotukhin2014-05-051-0/+0
| | | | llvm-svn: 207984
* Always set alignment of vectorized LD/ST in SLP-Vectorizer. ↵Yi Jiang2014-05-052-0/+31
| | | | | | <rdar://problem/16812145> llvm-svn: 207983
* Fix spelling.Joerg Sonnenberger2014-05-051-0/+0
| | | | llvm-svn: 207982
* Be a bit more specific in the release notes.Rafael Espindola2014-05-051-1/+2
| | | | llvm-svn: 207981
* LTO: -internalize sets visibility to defaultDuncan P. N. Exon Smith2014-05-052-0/+28
| | | | | | | | | Visibility is meaningless when the linkage is local. Change `-internalize` to reset the visibility to `default`. <rdar://problem/16141113> llvm-svn: 207979
* Remove the -disable-cfi option.Rafael Espindola2014-05-054-40/+2
| | | | | | | This also add a release note about it. If this stays I will cleanup MC next week. llvm-svn: 207977
* [Test] Remove substitution for clangAdam Nemet2014-05-051-1/+0
| | | | | | | | | clang should not be used in the llvm tests. The topic was discussed in this thread: http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20140428/214905.html llvm-svn: 207976
* Modify test to not use -disable-cfi.Rafael Espindola2014-05-052-34/+27
| | | | llvm-svn: 207974
* Select bdver2 instead of bdver1 if TBM support is present on models < 0x10.Kaelyn Takata2014-05-051-1/+2
| | | | | | | | | | Tested that the right -target-cpu is set in the clang -cc1 command line when running "clang -march=native -E -v - </dev/null" on both an FX-8150 and an FX-8350. Both are family 15h; the FX-8150 (Bulldozer processor) reports a model number of 1, and the FX-8350 (Piledriver processor) reports a model number of 2. llvm-svn: 207973
* Move test to the ARM64 directory.Rafael Espindola2014-05-051-0/+0
| | | | llvm-svn: 207972
* Convert a CodeGen test into a MC test.Rafael Espindola2014-05-052-161/+208
| | | | llvm-svn: 207971
* Fix test from r207966 and add a comment there.Michael Zolotukhin2014-05-051-2/+4
| | | | llvm-svn: 207969
* [ASan/Win] Fix issue 305 -- don't instrument .CRT initializer/terminator ↵Timur Iskhodzhanov2014-05-051-4/+14
| | | | | | | | | callbacks See https://code.google.com/p/address-sanitizer/issues/detail?id=305 Reviewed at http://reviews.llvm.org/D3607 llvm-svn: 207968
* Trivial simplification. No functionality change.Rafael Espindola2014-05-051-1/+1
| | | | llvm-svn: 207967
* Add regression test for r207692.Michael Zolotukhin2014-05-051-0/+23
| | | | llvm-svn: 207966
* Fix gcc -pedantic warning in lto.cpp.Patrik Hagglund2014-05-051-2/+2
| | | | llvm-svn: 207959
* Add range access to ELFFile's sections collection.Simon Atanasyan2014-05-051-16/+19
| | | | llvm-svn: 207952
* CodeGen: correct memset emittance for WoASaleem Abdulrasool2014-05-042-1/+20
| | | | | | | | Windows on ARM does not conform to AEABI. However, memset would be emitted using the AEABI signature, resulting in inverted parameters. Handle this special case appropriately. llvm-svn: 207943
* CodeGen: strengthen WoA AEABI avoidance testsSaleem Abdulrasool2014-05-041-0/+22
| | | | | | Add additional test cases for WoA AEABI avoidance checking. llvm-svn: 207942
* MC: support FK_SecRel_4 for Windows on ARMSaleem Abdulrasool2014-05-043-0/+48
| | | | | | | | | Add handling for FK_SecRel_4 (4-byte section relative relocations). These are used by the generation of DWARF debug information (the abbrevations use section relative relocations). This will also be used in generation of CodeView line tables. llvm-svn: 207941
* LoopUnroll: If we're doing partial unrolling, use the PartialThreshold to ↵Benjamin Kramer2014-05-042-3/+53
| | | | | | | | | | | limit unrolling. Otherwise we use the same threshold as for complete unrolling, which is way too high. This made us unroll any loop smaller than 150 instructions by 8 times, but only if someone specified -march=core2 or better, which happens to be the default on darwin. llvm-svn: 207940
* SLPVectorizer: Bring back the insertelement patch (r205965) with fixesArnold Schwaighofer2014-05-043-30/+211
| | | | | | | | | | | | | | | | When can't assume a vectorized tree is rooted in an instruction. The IRBuilder could have constant folded it. When we rebuild the build_vector (the series of InsertElement instructions) use the last original InsertElement instruction. The vectorized tree root is guaranteed to be before it. Also, we can't assume that the n-th InsertElement inserts the n-th element into a vector. This reverts r207746 which reverted the revert of the revert of r205018 or so. Fixes the test case in PR19621. llvm-svn: 207939
* AVX-512: minor change in rndscale intrinsicElena Demikhovsky2014-05-042-4/+13
| | | | llvm-svn: 207937
* [LCG] Add the last (and most complex) of the edge insertion mutationChandler Carruth2014-05-043-0/+290
| | | | | | | | | | | | | operations on the call graph. This one forms a cycle, and while not as complex as removing an internal edge from an SCC, it involves a reasonable amount of work to find all of the nodes newly connected in a cycle. Also somewhat alarming is the worst case complexity here: it might have to walk roughly the entire SCC inverse DAG to insert a single edge. This is carefully documented in the API (I hope). llvm-svn: 207935
* [LCG] Reorder the tests to be a bit more logical: inter-SCC mutationChandler Carruth2014-05-041-53/+53
| | | | | | | | before intra-SCC mutation, insertion before removal. No functionality changed. llvm-svn: 207934
* X86: further range-loopify AsmPrinterSaleem Abdulrasool2014-05-041-18/+16
| | | | | | Use more range loops in the X86AsmPrinter. NFC. llvm-svn: 207928
* X86: remove X86COFFMachineModuleInfoSaleem Abdulrasool2014-05-044-82/+0
| | | | | | Remove dead code. This is vestigial after r98384. llvm-svn: 207927
* X86: repair export compatibility with MinGW/cygwinSaleem Abdulrasool2014-05-042-37/+47
| | | | | | | | | | | | | | Both MinGW and cygwin (i686) construct export directives without the global leader prefix. This is mostly due to the fact that they use GNU ld which does not correctly handle the export directive. This apparently has been been broken for a while. However, this was recently reported as being broken by mingwandroid and diorcety of the msys2 project. Remove the global leader prefix if targeting MinGW or cygwin, otherwise, retain the global leader prefix. Add an explicit test for cygwin's behaviour of export directives. llvm-svn: 207926
* X86: refactor export directive generationSaleem Abdulrasool2014-05-042-22/+27
| | | | | | | | | Create a helper function to generate the export directive. This was previously duplicated inline to handle export directives for variables and functions. This also enables the use of range-based iterators for the generation of the directive rather than the traditional loops. NFC. llvm-svn: 207925
* IR: Cleanup AttributeSet::get for AttrBuilderDavid Majnemer2014-05-032-5/+13
| | | | | | | We don't modify the AttrBuilder in AttributeSet::get, make the reference argument const. llvm-svn: 207924
* [TBAA] Fix handling of mixed TBAA (path-aware and non-path-aware TBAA).Juergen Ributzka2014-05-033-2/+85
| | | | | | | | | | | | | | This fix simply ensures that both metadata nodes are path-aware before performing path-aware alias analysis. This issue isn't normally triggered in LLVM, because we perform an autoupgrade of the TBAA metadata to the new format when reading in LL or BC files. This issue only appears when a client creates the IR manually and mixes old and new TBAA metadata format. This fixes <rdar://problem/16760860>. llvm-svn: 207923
* Fix pr19645.Rafael Espindola2014-05-0313-58/+55
| | | | | | | | | | | | | | | | The fix itself is fairly simple: move getAccessVariant to MCValue so that we replace the old weak expression evaluation with the far more general EvaluateAsRelocatable. This then requires that EvaluateAsRelocatable stop when it finds a non trivial reference kind. And that in turn requires the ELF writer to look harder for weak references. Last but not least, this found a case where we were being bug by bug compatible with gas and accepting an invalid input. I reported pr19647 to track it. llvm-svn: 207920
* [ARM64] Correctly select ANDWri in FastISel.Joey Gouly2014-05-032-7/+14
| | | | | | http://reviews.llvm.org/D3598 llvm-svn: 207917
* SLPVectorizer: Lazily allocate the map for block numbering.Benjamin Kramer2014-05-032-27/+26
| | | | | | | | There is no point in creating it if we're not going to vectorize anything. Creating the map is expensive as it creates large values. No functionality change. llvm-svn: 207916
OpenPOWER on IntegriCloud