aboutsummaryrefslogtreecommitdiff
path: root/llvm
AgeCommit message (Collapse)AuthorFilesLines
2014-09-12Add Tom Stellard's role as 3.5 release manager.Joerg Sonnenberger1-1/+1
llvm-svn: 217659
2014-09-12llvm-cov: Move FunctionCoverageMapping into CoverageMapping.h (NFC)Justin Bogner8-47/+22
llvm-svn: 217657
2014-09-12Remove a temporary variable and just construct a unique_ptr directly using ↵Craig Topper1-9/+6
make_unique. llvm-svn: 217655
2014-09-11Object: Small cleanup in COFFObjectFile::getSymbolDavid Majnemer1-3/+3
NFC. llvm-svn: 217648
2014-09-11Revert "llvm-cov: Remove an overly system specific test"Justin Bogner3-1/+32
This fixes a call to sys::fs::equivalent that should've been to CodeCoverageTool::equivalentFiles, which lets us restore the test of r217476 that was removed in r217478. This reverts r217478, but the test works this time. llvm-svn: 217646
2014-09-11[MCJIT] Improve the "stub not found" diagnostic in RuntimeDyldChecker.Lang Hames1-1/+4
A "stub found found" diagnostic is emitted when RuntimeDyldChecker's stub lookup logic fails to find the requested stub. The obvious reason for the failure is that no such stub has been created, but it can also fail for internal symbols if the symbol offset is not computed correctly (E.g. due to a mangled relocation addend). This patch adds a comment about the latter case so that it's not overlooked. Inspired by confusion experienced during test case construction for r217635. llvm-svn: 217643
2014-09-11[Support][Endian] Overload += and -=Rui Ueyama1-0/+10
This patch is to overload operator+= and operator-= for {u}{little}{big}{16,32,64}_t. llvm-svn: 217637
2014-09-11R600/SI: Fix off by 1 error in used register countMatt Arsenault2-3/+12
The register numbers start at 0, so if only 1 register was used, this was reported as 0. llvm-svn: 217636
2014-09-11[MCJIT] Make sure we test ARM BR24 relocations with both internal and externalLang Hames1-2/+7
symbols. Previously we have only been testing these relocations with external symbols. <rdar://problem/18308413> llvm-svn: 217635
2014-09-11Support: Use llvm::COFF::BigObjMagicRui Ueyama2-8/+16
Use llvm::COFF::BigObjMagic insetad of the string literal. Also checks the version number. llvm-svn: 217633
2014-09-11Support: Delete {aligned_,}{u,}{little,big}8_tRui Ueyama7-81/+60
The byte has no endianness, so these types don't make sense. uint8_t should be used instead. llvm-svn: 217631
2014-09-11[C API] Make the 'lower switch' pass available via the C API.Juergen Ributzka2-0/+7
llvm-svn: 217630
2014-09-11[CodeGenPrepare] Teach the addressing mode matcher how to promote zext.Quentin Colombet2-13/+71
I.e., teach it about 'sext (zext a to ty) to ty2' => zext a to ty2. llvm-svn: 217629
2014-09-11Remove the unused string section symbol parameter from DwarfFile::emitStringsDavid Blaikie9-61/+43
And since it /looked/ like the DwarfStrSectionSym was unused, I tried removing it - but then it turned out that DwarfStringPool was reconstructing the same label (and expecting it to have already been emitted) and uses that. So I kept it around, but wanted to pass it in to users - since it seemed a bit silly for DwarfStringPool to have it passed in and returned but itself have no use for it. The only two users don't handle strings in both .dwo and .o files so they only ever need the one symbol - no need to keep it (and have an unused symbol) in the DwarfStringPool used for fission/.dwo. Refactor a bunch of accelerator table usage to remove duplication so I didn't have to touch 4-5 callers. llvm-svn: 217628
2014-09-11Support: improve identify_magic to recognize COFF bigobjRui Ueyama2-3/+12
identify_magic recognized a COFF bigobj as an import library file. This patch fixes that. llvm-svn: 217627
2014-09-11Misc cleanups to the FileSytem api.Rafael Espindola7-86/+66
The main difference is the removal of std::error_code exists(const Twine &path, bool &result); It was an horribly redundant interface since a file not existing is also a valid error_code. Now we have an access function that returns just an error_code. This is the only function that has to be implemented for Unix and Windows. The functions can_write, exists and can_execute an now just wrappers. One still has to be very careful using these function to avoid introducing race conditions (Time of check to time of use). llvm-svn: 217625
2014-09-11Add missing colon to RUN line...Bill Schmidt1-1/+1
llvm-svn: 217623
2014-09-11[PATCH, PowerPC] Accept 'U' and 'X' constraints in inline asmBill Schmidt2-0/+42
Inline asm may specify 'U' and 'X' constraints to print a 'u' for an update-form memory reference, or an 'x' for an indexed-form memory reference. However, these are really only useful in GCC internal code generation. In inline asm the operand of the memory constraint is typically just a register containing the address, so 'U' and 'X' make no sense. This patch quietly accepts 'U' and 'X' in inline asm patterns, but otherwise does nothing. If we ever unexpectedly see a non-register, we'll assert and sort it out afterwards. I've added a new test for these constraints; the test case should be used for other asm-constraints changes down the road. llvm-svn: 217622
2014-09-11[MCJIT] Add support for ARM HALF_DIFF relocations to MCJIT.Lang Hames5-73/+202
Fixes <rdar://problem/18297804>. llvm-svn: 217620
2014-09-11Use simpler version of sys::fs::exists. NFC.Rafael Espindola1-3/+2
llvm-svn: 217618
2014-09-11Use the simpler sys::fs:;exists. NFC.Rafael Espindola1-2/+1
llvm-svn: 217617
2014-09-11Add triple to test to fix botsMatt Arsenault1-1/+1
llvm-svn: 217612
2014-09-11Provide an implementation of getNoopForMachoTarget for SPARC.Brad Smith3-0/+15
llvm-svn: 217611
2014-09-11Add DAG combine for shl + add of constants.Matt Arsenault6-41/+157
Do (shl (add x, c1), c2) -> (add (shl x, c2), c1 << c2) This is already done for multiplies, but since multiplies by powers of two are turned into shifts, we also need to handle it here. This might want checks for isLegalAddImmediate to avoid transforming an add of a legal immediate with one that isn't. llvm-svn: 217610
2014-09-11[MCJIT] Take the relocation addend into account when applying ARM MachO VANILLALang Hames2-1/+5
and BR24 relocations. <rdar://problem/18296496> llvm-svn: 217605
2014-09-11Added missing LLVM_NOEXCEPT to the definition of _obj2yaml_error_category::nameIsmail Pazarbasi1-1/+3
LLVM_NOEXCEPT was added in r210591. llvm-svn: 217603
2014-09-11[AVX512] Fix miscompile for unpackAdam Nemet2-60/+53
r189189 implemented AVX512 unpack by essentially performing a 256-bit unpack between the low and the high 256 bits of src1 into the low part of the destination and another unpack of the low and high 256 bits of src2 into the high part of the destination. I don't think that's how unpack works. AVX512 unpack simply has more 128-bit lanes but other than it works the same way as AVX. So in each 128-bit lane, we're always interleaving certain parts of both operands rather different parts of one of the operands. E.g. for this: __v16sf a = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 }; __v16sf b = { 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 }; __v16sf c = __builtin_shufflevector(a, b, 0, 8, 1, 9, 4, 12, 5, 13, 16, 24, 17, 25, 20, 28, 21, 29); we generated punpcklps (notice how the elements of a and b are not interleaved in the shuffle). In turn, c was set to this: 0 16 1 17 4 20 5 21 8 24 9 25 12 28 13 29 Obviously this should have just returned the mask vector of the shuffle vector. I mostly reverted this change and made sure the original AVX code worked for 512-bit vectors as well. Also updated the tests because they matched the logic from the code. llvm-svn: 217602
2014-09-11Add triple and remove hashes to account for buildbot differences in comment ↵Sanjay Patel1-13/+13
strings. llvm-svn: 217601
2014-09-11Move constant-sized bitvector to the stack.Benjamin Kramer1-2/+2
llvm-svn: 217600
2014-09-11Combine fmul vector FP constants when unsafe math is allowed.Sanjay Patel2-6/+70
This is an extension of the change made with r215820: http://llvm.org/viewvc/llvm-project?view=revision&revision=215820 That patch allowed combining of splatted vector FP constants that are multiplied. This patch allows combining non-uniform vector FP constants too by relaxing the check on the type of vector. Also, canonicalize a vector fmul in the same way that we already do for scalars - if only one operand of the fmul is a constant, make it operand 1. Otherwise, we miss potential folds. This fold is also done by -instcombine, but it's possible that extra fmuls may have been generated during lowering. Differential Revision: http://reviews.llvm.org/D5254 llvm-svn: 217599
2014-09-11typoSanjay Patel1-1/+1
llvm-svn: 217597
2014-09-11R600: Add cmpxchg instruction for evergreenAaron Watry2-5/+29
Refactored the R600_LDS_1A2D class a bit to get it to actually work. It seemed to be previously unused and broken. We also have to disable the conversion to the noret variant for now in R600ISelLowering because the getLDSNoRetOp method only handles 1A1D LDS ops. Someone can feel free to modify the AMDGPU::getLDSNoRetOp method to work for more than 1A1D variants of LDS operations. It's being left as a future TODO for now. Signed-off-by: Aaron Watry <awatry at gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> llvm-svn: 217596
2014-09-11R600: Test local atomics for evergreenAaron Watry1-0/+27
Now that the operations are all implemented, we can test this sub-arch here. Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> llvm-svn: 217595
2014-09-11R600: Add LDS_WRXCHG[_RET] instructions for Evergreen.Aaron Watry1-0/+4
Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> llvm-svn: 217594
2014-09-11R600: Add LDS_MIN_[U]INT[_RET] instructions for EvergreenAaron Watry1-0/+8
Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> llvm-svn: 217593
2014-09-11R600: Add LDS_XOR[_RET] instructions for EvergreenAaron Watry1-0/+4
Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> llvm-svn: 217592
2014-09-11R600: Add LDS_OR[_RET] instructions for EvergreenAaron Watry1-0/+4
Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> llvm-svn: 217591
2014-09-11R600: Add LDS_AND[_RET] instructions for EvergreenAaron Watry1-0/+4
Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> llvm-svn: 217590
2014-09-11R600: Add LDS_MAX_[U]INT[_RET] instructions for EvergreenAaron Watry1-0/+8
This was only present for SI before. Cayman may still be missing, but I am unable to test that currently. v2: Don't create atomicrmw max tests in separate file Signed-off-by: Aaron Watry <awatry@gmail.com> Reviewed-by: Matt Arsenault <matthew.arsenault@amd.com> CC: Tom Stellard <thomas.stellard@amd.com> llvm-svn: 217589
2014-09-11[docs] Mention character array constants in docs/LangRef.rstDaniel Sanders1-1/+3
Summary: They were used in the 'Module Structure' example but weren't otherwise documented. Credit to Reed Kotler for noticing. Reviewers: hans Reviewed By: hans Subscribers: hans, llvm-commits Differential Revision: http://reviews.llvm.org/D5191 llvm-svn: 217583
2014-09-11[ARM] Add Thumb-2 code size optimization regression test for LSR (register).Tilmann Scheller1-0/+9
llvm-svn: 217582
2014-09-11[ARM] Add Thumb-2 code size optimization regression test for LSR (immediate).Tilmann Scheller1-0/+9
llvm-svn: 217581
2014-09-11[AArch64] Reenable the PBQP test now that the leak issue has been fixed.Arnaud A. de Grandmaison1-0/+14
David Blaikie's commits r217563 & r217564, which added shared_ptr to the CostPool have fixed some memory leak issues exposed by the PBQP with coalescing constraints. The sanitizer bot was failing because of those leaks. Now that the leaks are gone, we can reenable the aarch64/pbqp test. llvm-svn: 217580
2014-09-11[ARM] Add Thumb-2 code size optimization regression test for LSL (register).Tilmann Scheller1-0/+9
llvm-svn: 217579
2014-09-11[ARM] Add Thumb2 code size optimization regression test for LSL (immediate).Tilmann Scheller1-0/+9
llvm-svn: 217576
2014-09-11[x86] Fixup r217565 which baked in an assumption about the functionChandler Carruth1-1/+1
name that breaks on some platforms. This part of the test just doesn't matter... llvm-svn: 217575
2014-09-11[AlignmentFromAssumptions] Don't crash just because the target is 32-bitHal Finkel2-0/+219
We used to crash processing any relevant @llvm.assume on a 32-bit target (because we'd ask SE to subtract expressions of differing types). I've copied our 'simple.ll' test, but with the data layout from arm-linux-gnueabihf to get some meaningful test coverage here. llvm-svn: 217574
2014-09-11Build correct vector filled with undef nodesDavid Xu2-4/+62
llvm-svn: 217570
2014-09-11utils: Teach lldbDataFormatters how to format ArrayRefsJustin Bogner1-0/+33
llvm-svn: 217567
2014-09-11[x86] FileCheck-ize this test.Chandler Carruth1-5/+24
llvm-svn: 217565
generated by cgit v1.1 at 2026-01-05 23:59:09 +0000