summaryrefslogtreecommitdiffstats
path: root/llvm/lib
Commit message (Collapse)AuthorAgeFilesLines
...
* Revert "Remove a use of PathV1.h."Rafael Espindola2013-06-171-5/+3
| | | | | | | | This reverts commit r183985. We were missing the checks for not deleting things like /dev/null. llvm-svn: 184111
* Switch spill weights from a basic loop depth estimation to BlockFrequencyInfo.Benjamin Kramer2013-06-1711-50/+60
| | | | | | | | | | | | | | | | | | The main advantages here are way better heuristics, taking into account not just loop depth but also __builtin_expect and other static heuristics and will eventually learn how to use profile info. Most of the work in this patch is pushing the MachineBlockFrequencyInfo analysis into the right places. This is good for a 5% speedup on zlib's deflate (x86_64), there were some very unfortunate spilling decisions in its hottest loop in longest_match(). Other benchmarks I tried were mostly neutral. This changes register allocation in subtle ways, update the tests for it. 2012-02-20-MachineCPBug.ll was deleted as it's very fragile and the instruction it looked for was gone already (but the FileCheck pattern picked up unrelated stuff). llvm-svn: 184105
* Fix for a regression caused by the LoopVectorizer whenPekka Jaaskelainen2013-06-171-3/+3
| | | | | | | vectorizing loops with memory accesses to non-zero address spaces. It simply dropped the AS info. Fixes PR16306. llvm-svn: 184103
* Don't use PathV1.h in LTOCodeGenerator.cppRafael Espindola2013-06-171-0/+4
| | | | | | | This patch also adds a simpler version of sys::fs::remove and a tool_output_file constructor for when we already have an open file. llvm-svn: 184095
* Disable vectorization for -Oz.Nadav Rotem2013-06-171-1/+1
| | | | llvm-svn: 184089
* Enable the loop vectorizer by default for -Os and -O2.Nadav Rotem2013-06-171-7/+1
| | | | llvm-svn: 184084
* Move lib/Archive to tools/llvm-ar.Rafael Espindola2013-06-1710-1374/+2
| | | | | | | llvm-ar is the only tool that needs to write archive files. Every other tool should be able to use the lib/Object interface. llvm-svn: 184083
* Fix incorrectly finding 'executable' directories instead of files.Manuel Klimek2013-06-171-1/+8
| | | | | | | | | | | | | | | This broke for example the 'not' utility, if a directory called 'FileCheck' is executable and in the path before the actual 'FileCheck'. This patch steals the implementation of the "old" PathV1 canExecute implementation: - checks for R_OK (file readable): this is necessary for executing scripts; we should not regress here unless we have good reasons - checks for S_ISREG; if we want to get rid of this, we'd need to change all callers who already made the assumption when depending on Path V1. llvm-svn: 184074
* DebugInfo: remove target-specific Frame Index handling for DBG_VALUE ↵David Blaikie2013-06-1630-377/+9
| | | | | | | | | | MachineInstrs Frame index handling is now target-agnostic, so delete the target hooks for creation & asm printing of target-specific addressing in DBG_VALUEs and any related functions. llvm-svn: 184067
* Debug Info: Simplify Frame Index handling in DBG_VALUE Machine InstructionsDavid Blaikie2013-06-1612-137/+113
| | | | | | | | | | | | | | | | | | | | Rather than using the full power of target-specific addressing modes in DBG_VALUEs with Frame Indicies, simply use Frame Index + Offset. This reduces the complexity of debug info handling down to two representations of values (reg+offset and frame index+offset) rather than three or four. Ideally we could ensure that frame indicies had been eliminated by the time we reached an assembly or dwarf generation, but I haven't spent the time to figure out where the FIs are leaking through into that & whether there's a good place to convert them. Some FI+offset=>reg+offset conversion is done (see PrologEpilogInserter, for example) which is necessary for some SelectionDAG assumptions about registers, I believe, but it might be possible to make this a more thorough conversion & ensure there are no remaining FIs no matter how instruction selection is performed. llvm-svn: 184066
* Silence warning in Release builds.Benjamin Kramer2013-06-161-0/+1
| | | | llvm-svn: 184059
* DebugInfo: print DBG_VALUE MachineInstrs with [] for deref and drop the ↵David Blaikie2013-06-151-3/+13
| | | | | | offset when it's zero llvm-svn: 184045
* Use 0 instead of NULL.Jakub Staszak2013-06-151-5/+5
| | | | llvm-svn: 184044
* PruneEH: Only merge attribute sets when used. No functionality change.Benjamin Kramer2013-06-151-7/+5
| | | | llvm-svn: 184041
* Missing NDEBUGs.Andrew Trick2013-06-151-0/+6
| | | | llvm-svn: 184039
* MI-Sched: heuristics using the new latency and machine model.Andrew Trick2013-06-151-234/+330
| | | | llvm-svn: 184038
* Support BufferSize on ProcResGroup for unified MOp schedulers.Andrew Trick2013-06-152-0/+11
| | | | | | And add Sandybridge/Haswell resource buffers. llvm-svn: 184034
* Update machine models. Specify buffer sizes for OOO processors.Andrew Trick2013-06-156-10/+10
| | | | llvm-svn: 184033
* Machine Model: Add MicroOpBufferSize and resource BufferSize.Andrew Trick2013-06-1512-118/+51
| | | | | | | | | | | | | Replace the ill-defined MinLatency and ILPWindow properties with with straightforward buffer sizes: MCSchedMode::MicroOpBufferSize MCProcResourceDesc::BufferSize These can be used to more precisely model instruction execution if desired. Disabled some misched tests temporarily. They'll be reenabled in a few commits. llvm-svn: 184032
* MI-Sched: Rename IssueCount to CurrMOps.Andrew Trick2013-06-151-8/+8
| | | | | | | "Counts" refer to scaled resource counts within a region. CurrMOps is simply the number of micro-ops to be issue in the current cycle. llvm-svn: 184031
* MI-Sched: Remove the temporary EnableCopyConstrain flag.Andrew Trick2013-06-151-7/+1
| | | | llvm-svn: 184030
* MI-Sched: added tracking of dependent latency for better heuristics.Andrew Trick2013-06-151-13/+26
| | | | | | | | | | | | | | Heuristics compare the critical path in the scheduled code, called ExpectedLatency, with the latency of instructions remaining to be scheduled. There are two ways to look at remaining latency: (1) Dependent latency includes the latency between unscheduled and scheduled instructions. (2) Independent latency is simply the height (bottom-up) or depth (top-down) of instructions currently in the ready Q. llvm-svn: 184029
* MI-Sched: DEBUG: print critical resource.Andrew Trick2013-06-151-0/+4
| | | | llvm-svn: 184028
* SelectionDAG: slightly refactor DAGCombiner::visitSELECT_CC to avoid ↵Stephen Lin2013-06-151-12/+14
| | | | | | | | redudant checks... This doesn't really effect performance due to all the relevant calls being transparent but is clearer. llvm-svn: 184027
* Debug Info: Don't print the display name and colon prefix for DEBUG_VALUE ↵David Blaikie2013-06-151-2/+5
| | | | | | comments if the display name is empty llvm-svn: 184026
* [yaml2obj] Add support for sh_link via `Link` key.Sean Silva2013-06-151-0/+1
| | | | llvm-svn: 184022
* R600: Add SI load support for v[24]i32 and store for v2i32Tom Stellard2013-06-151-0/+5
| | | | | | | | | | | Also add a seperate vector lit test file, since r600 doesn't seem to handle v2i32 load/store yet, but we can test both for SI. Patch by: Aaron Watry Reviewed-by: Tom Stellard <thomas.stellard@amd.com> Signed-off-by: Aaron Watry <awatry@gmail.com> llvm-svn: 184021
* Remove the LLVM specific archive index.Rafael Espindola2013-06-145-150/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Archive files (.a) can have a symbol table indicating which object files in them define which symbols. The purpose of this symbol table is to speed up linking by allowing the linker the read only the .o files it is actually going to use instead of having to parse every object's symbol table. LLVM's archive library currently supports a LLVM specific format for such table. It is hard to see any value in that now that llvm-ld is gone: * System linkers don't use it: GNU ar uses the same plugin as the linker to create archive files with a regular index. The OS X ar creates no symbol table for IL files, I assume the linker just parses all IL files. * It doesn't interact well with archives having both IL and native objects. * We probably don't want to be responsible for yet another archive format variant. This patch then: * Removes support for creating and reading such index from lib/Archive. * Remove llvm-ranlib, since there is nothing left for it to do. We should in the future add support for regular indexes to llvm-ar for both native and IL objects. When we do that, llvm-ranlib should be reimplemented as a symlink to llvm-ar, as it is equivalent to "ar s". llvm-svn: 184019
* R600: Use correct encoding for Vertex Fetch instructions on CaymanTom Stellard2013-06-143-156/+294
| | | | | Reviewed-by: Vincent Lejeune<vljn at ovi.com> llvm-svn: 184016
* R600: Use EXPORT_RAT_INST_STORE_DWORD for stores on CaymanTom Stellard2013-06-142-37/+57
| | | | | | | | | We were using RAT_INST_STORE_RAW, which seemed to work, but the docs say this instruction doesn't exist for Cayman, so it's probably safer to use a documented instruction instead. Reviewed-by: Vincent Lejeune<vljn at ovi.com> llvm-svn: 184015
* R600: Factor the instruction encoding out the RAT_WRITE_CACHELESS_eg classTom Stellard2013-06-142-50/+68
| | | | | Reviewed-by: Vincent Lejeune<vljn at ovi.com> llvm-svn: 184014
* R600: Move instruction encoding definitions into a separate .td fileTom Stellard2013-06-142-362/+393
| | | | | Reviewed-by: Vincent Lejeune<vljn at ovi.com> llvm-svn: 184013
* Introduce getSelect usage and use more getSelectCCMatt Arsenault2013-06-146-87/+90
| | | | llvm-svn: 184012
* Add GetCurrentDirectory back.Rafael Espindola2013-06-142-0/+21
| | | | | | | | It looks like clang-tools-extra/unittests/cpp11-migrate/TransformTest.cpp depends on the behaviour of the old one on Windows. Maybe a difference between GetCurrentDirectoryA and GetCurrentDirectoryW? llvm-svn: 184009
* SelectionDAG: minor fix to order of operands in comments to match the code Stephen Lin2013-06-141-3/+3
| | | | llvm-svn: 184008
* Replace use of PathV1.h in MCContext.cpp.Rafael Espindola2013-06-143-23/+4
| | | | | | GetCurrentDirectory is now unused. Remove it. llvm-svn: 184003
* Mark rematerialized super/sub registers as dead.Tim Northover2013-06-141-0/+1
| | | | | | | | | | | | | When we're rematerializing into a not-quite-right register we already add the real definition as an imp-def, but we should also be marking the "official" register as dead, since nothing else is going to use it as a result of this remat. Not doing this can affect pressure tracking. rdar://problem/14158833 llvm-svn: 184002
* Replace use of PathV1.h in Program.cpp.Rafael Espindola2013-06-145-75/+63
| | | | llvm-svn: 183996
* SelectionDAG: Fix incorrect condition checks in some cases of folding ↵Stephen Lin2013-06-141-16/+16
| | | | | | FADD/FMUL combinations; also improve accuracy of comments llvm-svn: 183993
* Remove unused argument.Rafael Espindola2013-06-141-3/+3
| | | | llvm-svn: 183992
* Don't use PathV1.h in GraphWriter.cpp.Rafael Espindola2013-06-141-5/+2
| | | | llvm-svn: 183988
* Convert a use of sys::Path::GetTemporaryDirectory.Rafael Espindola2013-06-141-10/+9
| | | | llvm-svn: 183987
* Remove a use of PathV1.h.Rafael Espindola2013-06-141-3/+5
| | | | llvm-svn: 183985
* Make PrologEpilogInserter save/restore all callee saved registersDerek Schuff2013-06-141-1/+3
| | | | | | | | | | | in functions which call __builtin_unwind_init() __builtin_unwind_init() is an undocumented gcc intrinsic which has this effect, and is used in libgcc_eh. Goes part of the way toward fixing PR8541. llvm-svn: 183984
* Remove a use of PathV1.h.Rafael Espindola2013-06-141-15/+9
| | | | llvm-svn: 183982
* Move PrependMainExecutablePath next to its only user.Rafael Espindola2013-06-141-23/+0
| | | | llvm-svn: 183980
* Remove a use of sys::Path.Rafael Espindola2013-06-141-2/+3
| | | | llvm-svn: 183979
* ARM: fix thumb coprocessor instruction with pre-writeback disassemblyAmaury de la Vieuville2013-06-141-1/+1
| | | | | | | was stc2 p0, c0, [r0]! instead of stc2 p0, c0, [r0,#0]! llvm-svn: 183975
* X86: cvtpi2ps is just an SSE instruction with MMX operands. It has no AVX ↵Benjamin Kramer2013-06-141-7/+8
| | | | | | | | equivalent. Give it the right register format so we can also emit it when AVX is enabled. llvm-svn: 183971
* Enable FastISel on ARM for Linux and NaCl, not MCJITJF Bastien2013-06-142-3/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a resubmit of r182877, which was reverted because it broken MCJIT tests on ARM. The patch leaves MCJIT on ARM as it was before: only enabled for iOS. I've CC'ed people from the original review and revert. FastISel was only enabled for iOS ARM and Thumb2, this patch enables it for ARM (not Thumb2) on Linux and NaCl, but not MCJIT. Thumb2 support needs a bit more work, mainly around register class restrictions. The patch punts to SelectionDAG when doing TLS relocation on non-Darwin targets. I will fix this and other FastISel-to-SelectionDAG failures in a separate patch. The patch also forces FastISel to retain frame pointers: iOS always keeps them for backtracking (so emitted code won't change because of this), but Linux was getting much worse code that was incorrect when using big frames (such as test-suite's lencod). I'll also fix this in a later patch, it will probably require a peephole so that FastISel doesn't rematerialize frame pointers back-to-back. The test changes are straightforward, similar to: http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20130513/174279.html They also add a vararg test that got dropped in that change. I ran all of lnt test-suite on A15 hardware with --optimize-option=-O0 and all the tests pass. All the tests also pass on x86 make check-all. I also re-ran the check-all tests that failed on ARM, and they all seem to pass. llvm-svn: 183966
OpenPOWER on IntegriCloud