[libc++] <experimental/simd> Add implicit type conversion constructor for class simd/simd_mask #71132
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
llvmbot
added
the
libc++
joy2myself
force-pushed
the
split_simd
branch
from
5e80afc
to
101322c
Compare
|
Gentle ping. @philnik777 |
philnik777
approved these changes
Nov 14, 2023
| }; | ||
|
|
||
| template <class U, class T, class SimdAbi = ex::simd_abi::compatible<T>, class = void> | ||
| struct has_conversion_ctor : std::false_type {}; |
There was a problem hiding this comment.
I tried it but it seems is_convertible can not accurately test the type conversion constructor. When U is the same as T, simd[_mask]<U, Abi> is the same as simd[_mask]<T, Abi>. And the default copy constructor make them convertible no matter what the Abi is.
There was a problem hiding this comment.
Sorry, my previous understanding may not right. The new commit may be what you want.
joy2myself
force-pushed
the
split_simd
branch
from
101322c
to
b601f16
Compare
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
joy2myself
force-pushed
the
split_simd
branch
from
b601f16
to
f5bf3ab
Compare
… for class simd/simd_mask
joy2myself
force-pushed
the
split_simd
branch
from
f5bf3ab
to
acccf40
Compare
fadlyas07
pushed a commit
to greenforce-project/llvm-project
that referenced
this pull request
Nov 25, 2023
* llvm-project/main: [JITLink][AArch32] Add test fixture with helper functions in error tests [libc++] Updates mdspan synopsis, [libc++][doc] Fixes grammar issues. [JITLink][AArch32] Split invalid edge kind test out into readAddendDataErrors [llvm][NFC] Autoupdater x86 intrinsic selection (llvm#73046) [libc++] Removes basic_string::reserve(). (llvm#73354) [lld][LoongArch] Add a another corner testcase for elf::getLoongArchPageDelta [llvm][MC][ARM][Assembly] Emit relocations for LDRs (llvm#72873) [MLIR][Vector] Refactor tests for contract -> OP transforms (llvm#73348) [JITLink][AArch32] Fix GetEdgeKindName function in error tests [JITLink][AArch32] Reflow code structure after llvm::endianness refactor (NFC) [CI] Add check-mlir-python to MLIR pre-merge checks (llvm#72847) [GlobalISel] Add identity fold for fadd -0.0 (llvm#73296) [mlir][doc] Remove duplicate syntax formats (llvm#73343) [RISCV][GISel] Add more G_SEXTLOAD instruction selection tests. NFC [RISCV][GISel] Add G_FRAME_INDEX support to selectAddrRegImm. [MSP430] Add missed mayStore flag for MSP430 PUSH instructions (llvm#73377) [Flang] Fix a test case that depends on stderr output of nullptr. (llvm#73349) [llvm][X86WinEHState] Replace calls to 'Type::getPointerTo' (NFC) [llvm][Transforms][Utils] Remove no-op ptr-to-ptr bitcasts (NFC) [gn build] Port 1a3b14d [clang-linker-wrapper] Re-use type returned from 'PointerType::getUnqual(C)' (NFC) (llvm#73374) [RISCV] Add C intrinsics for scalar bitmanip and crypto [clang] Avoid memcopy for small structure with padding under -ftrivial-auto-var-init (llvm#71677) [libc++][NFC] Refactor _LIBCPP_AVAILABILITY_HAS_* macros to always be defined (llvm#71002) [NFC] fix failed ompt tests on M1 device (llvm#65696) [Clang][OpenMP] Emit unsupported directive error (llvm#70233) [JITLink][AArch32] Add dynamic lookup for relocation fixup infos (llvm#71649) [llvm-jitlink] Support plain AArch32 stubs in jitlink-check's stub_addr() expressions (llvm#73268) [Thumb] Add test case where the machine-outliner clobbers LR. [mlir][spirv] Split codegen for float min/max reductions and others v2. [NFC] (llvm#73363) Revert "[llvm-exegesis] Switch from MCJIT to LLJIT (llvm#72838)" [SimpleLoopUnswitch] Remove callbacks (llvm#73300) [llvm-exegesis] Switch from MCJIT to LLJIT (llvm#72838) [mlir][spirv] Handle all zero-element memref types (llvm#73351) [PowerPC] Do not string pool globals that are part of llvm used. (llvm#66848) [GlobalISel] Treat shift amounts as unsigned in `matchShiftImmedChain` [IR] Add disjoint flag for Or instructions. (llvm#72583) [GISel][RISCV] Fix several boundary cases in narrow G_SEXT_INREG. (llvm#72719) [libc++] Removes codecvt. (llvm#72496) [libc++] Implements Runtime format strings II. (llvm#72543) [MLIR][OpenMP] NFC: Remove unused variable [clang] Add missing LinkageSpec case to getCursorKindForDecl (llvm#72401) [SLP][NFC]Make collectValuesToDemote member of BoUpSLP to avoid using Expr container, NFC. [Flang] Add new Integration tests directory to Flang (llvm#73141) [libc] Provide compiler version properties (llvm#73344) [lldb] correct inconsistent order of messages on process launch (llvm#73173) [Clang] Fix `-Wdocumentation` warning (NFC) [libc][cmake] Add missing dependencies for type_traits (llvm#73339) [DebugInfo][RemoveDIs] Support cloning and remapping DPValues (llvm#72546) [DebugInfo][RemoveDIs] Allow speculative-DPMarker creation [libc][NFC] Move float macro into its own header / add target os detection (llvm#73311) [CVP] Regenerate test checks (NFC) [SCEV] Regenerate test checks (NFC) [CI] Skip libcxx in non-llvm repo (llvm#73282) [AArch64] Add artificial clobbers to swift async context test. [JITLink] Fix typos: symobls -> symbols (NFC) [llvm-jitlink] Avoid assertion failure in make_error parameter [CodeGen] Make some includes explicit (NFC) [clang] Classify vector types in __builtin_classify_type (llvm#73299) [clang][ASTImporter] Fix import of SubstTemplateTypeParmType in return type of function. (llvm#69724) [libc][NFC] Remove dead code (llvm#73315) [DebugInfo] Clone dbg.values in SimplifyCFG like normal instructions (llvm#72526) Remove extraneous ` in AttrDocs.td [TLI][AArch64] Add TLI Mappings of @llvm.exp10 for ArmPL and SLEEF. [clang] Fix sorting module headers (llvm#73146) [MLIR][NFC] Fix build on recent GCC with C++20 enabled (llvm#73308) [MLIR][OpenMP] Fix the assertion failure for VariableCaptureKind::ByCopy (llvm#72424) [Flang][OpenMP] Remove use of non reference values from MapInfoOp (llvm#72444) [TargetLowering] Don't include ComplexDeinterleavingPass.h (NFC) Fix build failure on certain bots [gn build] Port 28233b1 [mlir][vector][spirv] Lower vector.load and vector.store to SPIR-V (llvm#71674) [AMDGPU] New AMDGPUInsertSingleUseVDST pass (llvm#72388) [Github] Build clang docs in CI if autogenerated files change (llvm#72623) [llvm-exegesis] Refactor ExecutableFunction to use a named constructor (llvm#72837) [AArch64] Add extra vecreduce.fmul tests. NFC [MachineSink] Some more preserving of debug location when rematerialising an instruction to replace a COPY (llvm#73155) clang/APINotes: squelch a -Wparantheses warning (NFC) Update links in AttrDocs.td for coro_lifetimebound [clang][analyzer] Move `security.cert.env.InvalidPtr` out of `alpha` (llvm#71912) [clang][analyzer] Support `fprintf` in the SecuritySyntaxChecker (llvm#73247) [Support] Implement getMainExecutable for Haiku [RISCV] Support target attribute for function Try to fix good-first-issue bot (llvm#71816) [lld][ELF] Add getBitcodeMachineKind test for LoongArch (llvm#71931) [NFC][X86] Clang-format X86FrameLowering.cpp (llvm#73287) [NFC][X86] Move the comments to a right place (llvm#73284) Adding missing `useSoftFloat` check, NFCI [X86][AVX10] Allow AVX10 use VBMI2 instructions (llvm#73276) [X86][MC] Support encoding/decoding for PUSH2[P]/POP2[P] (llvm#73233) [SelectionDAG] Format BuiltinOpcodes [NFC][X86] Clang-format X86RecognizableInstr.h (llvm#73279) [NFC][X86] Clang-format X86RecognizableInstr.cpp (llvm#73278) [NFC][X86] Clang-format X86BaseInfo.h and refine some comments (llvm#73274) [NFC][X86] Clang-format X86DisassemblerDecoderCommon.h (llvm#73272) Revert "[SLP][NFC]Make collectValuesToDemote member of BoUpSLP to avoid using" [libc++] <experimental/simd> Add implicit type conversion constructor for class simd/simd_mask (llvm#71132) [libc++abi] Avoid raw calls to assert() in libc++abi (llvm#71121) [RISCV][GISel] Support G_CTTZ/CTLZ with Zbb. [LAA] Factor out logic to compute dependence distance. (NFCI) [mlir] fix LLVM type converter for structs (llvm#73231) [libc++][NFC] Move QoI attributes into a single place inside <__config> (llvm#70870) [RISCV][GISel] Rename some variables to be more consistent. NFC [RISCV][GISel] Support G_CTPOP with Zbb. [JITLink][AArch32] Add tests for BLX interworking calls [ConstraintElim] Add tests with GEPs with variable and negative offsets [gn build] Port aea7929 [BOLT] Extend calculateEmittedSize() for block size calculation (llvm#73076) Revert "mp" Revert "[clang] Avoid memcopy for small structure with padding under -ftrivial-auto-var-init (llvm#71677)" mp Revert "[ClangRepl] Type Directed Code Completion" (llvm#73259) [libc++] Unify __is_trivial_equality_predicate and __is_trivial_plus_operation into __desugars_to (llvm#68642) [APINotes] Upstream Driver and Frontend options that enable API Notes [GlobalISel] Fix a bit of formatting. NFC Extra assertions in RS4GC (llvm#71201) [Bazel] Add Bazel build files for Python bindings of the GPU dialect (llvm#73256) [ClangRepl] Type Directed Code Completion (llvm#67349) [SVE] Don't require lookup when demangling vector function mappings (llvm#72260) [StackColoring] Do not drop AA metadata when not doing remappings (llvm#71958) [DebugInfo][RemoveDIs] Instrument jump-threading to update DPValues (llvm#73127) [VPlan] Remove dead IsEpilogueVec argument from prepareToExecute (NFC). Revert rG67275263b3b781a "[X86] X86DAGToDAGISel - attempt to merge XMM/YMM loads with YMM/ZMM loads of the same ptr (llvm#73126)" [AMDGPU] Define new targets gfx1200 and gfx1201 (llvm#73133) [clang] Avoid memcopy for small structure with padding under -ftrivial-auto-var-init (llvm#71677) [clang][CodeGen] Remove ptr-to-ptr bitcasts (NFC) (llvm#73020) [libc++][NFC] Adjust synopsis for std::launder Recommit "[ConstraintElim] Treat ConstantPointerNull as constant offset 0." [NFC] Use TypeSize for comparison in EVT::isExtendedXBitVector functions (llvm#73131) [libc++] Refactor the creation of the global and classic locales (llvm#72581) [libc++][NFC] Run clang-format on uses_alloc_types.h [libc++][test][msan] Refine XFAIL after llvm#67799 (llvm#73213) [clang][CGExprConstant] Remove no-op ptr-to-ptr bitcast (NFC) [libc++] Simplify the conditions for generating a linker script (llvm#73151) [clang][analyzer] Support `fgetc` in StreamChecker (llvm#72627) [flang] Add runtimes using --dependent-lib on MSVC targets (llvm#72519) [X86] X86DAGToDAGISel - attempt to merge XMM/YMM loads with YMM/ZMM loads of the same ptr (llvm#73126) [Bazel] Fixes for c43c885 Reland [mlir] Workaround for export lib generation on Windows for mlir_arm_sme_abi_stubs llvm#73147 (llvm#73238) [mlir] fix CAPI/translation.c test [AMDGPU] Fix CPol operands of MUBUF atomics. (llvm#73118) [Flang] Add partial support for lowering procedure pointer assignment. (llvm#70461) [LV] Remove TODO addressed in 32d1197. [mlir] do not inject malloc/free in to-LLVM translation (llvm#73224) Revert "[mlir] Workaround for export lib generation on Windows for `mlir_arm_sme_abi_stubs` (llvm#73147)" [NFC][OpenMP][MLIR] Add MLIR test for lowering parallel if (llvm#71788) [SystemZ] Move new test into existing CodeGen test. (llvm#73230) [DebugInfo][RemoveDIs] Instrument loop-deletion for DPValues (llvm#73042) [mlir] Workaround for export lib generation on Windows for `mlir_arm_sme_abi_stubs` (llvm#73147) Fix MSVC "not all control paths return a value" warning. NFC. [LLVM] Make s_getpc_b64 rematerializable (llvm#71823) [mlir] Add mlirTranslateModuleToLLVMIR to MLIR-C (llvm#73117) [LV] Re-use existing compare if possible for diff checks. [LoongArch] Disable mulodi4 and muloti4 libcalls (llvm#73199) [gn build] Port b61ac4a [DWARFLinkerParallel] Add support for -odr mode. (llvm#68721) Revert "[clang] Avoid memcopy for small structure with padding under -ftrivial-auto-var-init (llvm#71677)" [libc][NFC] Sink "PlatformDefs.h" into "FloatProperties.h" (llvm#73226) [mlir][LLVM] Support `immargs` in LLVM_IntrOpBase intrinsics (llvm#73013) [AMDGPU] Add some clang-format off/on markers [AMDGPU] Reindent some tables [Gisel][AArch64] legalize G_IS_FPCLASS (llvm#72796) [libc][NFC] Remove unused define (llvm#73222) [libc][NFC] Split builtin_wrapper into bit and math_extras (llvm#73113) [mlir][linalg] Add an e2e test for linalg.matmul to ArmSME (llvm#72144) [RISCV][GISel] Suppport G_BSWAP with Zbb. [LoongArch][MC] Modify branch evaluation for MCInstrAnalysis (llvm#73205) [mlir][linalg] `BufferizeToAllocationOp`: fix side effects (llvm#72986) [AArch64] Allow LDR merge with same destination register by renaming (llvm#71908) [X86][MC] Support encoding/decoding for PUSHP/POPP (llvm#73092) [AMDGPU][NFC] Update GISel memory-legalizer-atomic-fence test (llvm#72829) clang/CodeGen/RISCV: test lowering of math builtins (llvm#71399) Revert "[Flang] Add partial support for lowering procedure pointer assignment. (llvm#70461)" [LoongArch] Precommit a test for smul with overflow (NFC) (llvm#73212) [flang][openacc][NFC] Merge acc-declare.f90 tests (llvm#73055) [RISCV] Add MinimumJumpTableEntries to TuneInfo (llvm#72963) [clang] Avoid memcopy for small structure with padding under -ftrivial-auto-var-init (llvm#71677) [nfc][InstrProfTest]Parameterize the edge cases of value profile merge by value kind (llvm#73165) [X86] Remove unused IES_IDENTIFIER state from IntelExprState. NFC [RISCV] Use Float type instead of Half type for Fixed RVV vector type mangling (llvm#73091) [X86] Reject fpsr in inline asm constraints other than clobber. [LinkerWrapper] Accept some needed lld-link linker arguments for COFF targets (llvm#72889) Revert "[CUDA][HIP] make trivial ctor/dtor host device (llvm#72394)" [compiler-rt] Fix the HWCAP2_EBF16 and HWCAP2_SVE_EBF16 macro value (llvm#70905) Revert "[CUDA][HIP] ignore implicit host/device attr for override (llvm#72815)" [Driver] Default Generic_GCC aarch64_be to -fasynchronous-unwind-tables (llvm#72971) [AArch64] Fix -Wunused-function of getLivePhysRegsUpTo in AArch64FrameLowering.cpp (NFC) [RISCV][GISel] Add support for G_IS_FPCLASS in F and D extensions (llvm#72000) [RISCV][NFC] Rename `RISCVISD::FPCLASS` to `RISCVISD::FCLASS` [flang] Make .exe extension of the linker optional (NFC) (llvm#73157) [CMake] Support building shared library for NetBSD [PowerPC] Silence -Woverloaded-virtual warning. NFC [libc++] Remove the deprecated _LIBCPP_AVAILABILITY_CUSTOM_VERBOSE_ABORT_PROVIDED macro (llvm#73164) Revert "[mlir] Fix `TileUsingForOp` attr-dict printing/parsing, cleanup assembly format" (llvm#73178) [AArch64] Use the same fast math preservation for MachineCombiner reassociation as X86/PowerPC/RISCV. (llvm#72820) [flang][OpenMP] Add semantic check for declare target (llvm#72770) [runtimes] Add missing test dependencies to check-all (llvm#72955) Revert "[ConstraintElim] Treat ConstantPointerNull as constant offset 0." [libc++] Add missing check for C++17 in test [mlir] Fix `TileUsingForOp` attr-dict printing/parsing (llvm#72745) [flang][openacc] Avoid crash when collapse loop nest has extra directive (llvm#73166) [SLP][NFC]Remove extra unused vars, add TODO, NFC. [IR] Replace uses of IRBuilder::getInt8PtrTy with getPtrTy. NFC (llvm#73154) [libc++] Remove the ignore_format.txt file (llvm#73135) [libc++] Document upcoming removal of _LIBCPP_ENABLE_NARROWING_CONVERSIONS_IN_VARIANT in LLVM 19 [CMake] Fix the condition to include asan_rtl_x86_64.S in libclang_rt.asan_static.a [ConstraintElim] Treat ConstantPointerNull as constant offset 0. [libc++] Refactor atomic_{unsigned,signed}_lock_free (llvm#73041) [libc++] Re-introduce special support for narrowing conversions to bool in variant (llvm#73121) [SLP]Improve detection of gathered loads, if no other deps are detected. [IROutliner] Skip dbg values during the candidate search. (llvm#72945) [libc++][test][msan] Fix bots after llvm#67799 (llvm#73152) Sink variable into #ifndef NDEBUG where it is used Remove `__cdecl` from _ReturnAddress (llvm#72919) [code-format] Also include libc++ extensionless headers and .inc and .cppm (llvm#73142) [MCP] fix PowerPC redundant copy instructions removal fail test cases, NFC [mlir][sparse] change dim level type -> level type (llvm#73058) [Flang] Add partial support for lowering procedure pointer assignment. (llvm#70461) [AArch64] Add check that prologue insertion doesn't clobber live regs. (llvm#71826) [SLP][NFC]Use SmallVector instead of std::vector and remove unused includes, NFC. [MCP] Enhance MCP copy Instruction removal for special case (llvm#70778) Revert "[Flang][OpenMP] NFC: Minor refactoring of Reduction lowering code" (llvm#73139) [libc][bazel] Enable __support tests (llvm#73125) [clang][dataflow] Clear `ExprToLoc` and `ExprToVal` at the start of a block. (llvm#72985) [mlir][vector] Add patterns to simplify chained reductions (llvm#73048) [Flang][OpenMP] NFC: Minor refactoring of Reduction lowering code (llvm#70790) [mlgo] Fix test post 42d4840 [DebugInfo][RemoveDIs] Handle DPValues in LCSSA (llvm#72996) [Driver] Simply some gcc search logic (llvm#72558) [mlir] `im2col` & `l2cache` on cp.async.bulk.tensor.shared.cluster.global` (llvm#72967) [AArch64] Add quadword gather load/scatter store intrinsics with unscaled vector offset (llvm#71290) [Bazel] Fixes for dff97c1 [X86][AVX10] Add no-evex512 for MMX intrinsics, NFCI [libc] Disable nexttoward tests on the GPU [mlir][ArmSME] Move ArmSME -> intrinsics lowerings to `convert-arm-sme-to-llvm` pass (llvm#72890) Revert "Revert "[SVE2.1][Clang][LLVM]Add BFloat16 builtin in Clang an… (llvm#73110) [Tosa] Add local_bound attribute (llvm#73001) Fix for TOSA-to-linalg lowering of tosa.transpose op (llvm#72698) [X86] combineFMulcFCMulc - use KnownBits to detect conjugate patterns. [LV] Use SCEV for subtraction of src/sink for diff runtime checks. [mlir][tosa] Fix lowering of tosa.matmul with dynamic outputs (llvm#72724) [LV] Add test with a number of redundant runtime check instructions. [coroutines] Introduce [[clang::coro_lifetimebound]] (llvm#72851) [libc++] Floating Point Atomic (llvm#67799) [flang][OpenMP] Fix min reduction initialization (llvm#73102) [X86] combineFaddCFmul - use KnownBits to detect FNEG patterns. [LV] Reduce memory-check-threshold for test to preserve original test. [DomTree] Remove unnecessary check (NFC) [FuncSpec] Update function specialization to handle phi-chains (llvm#72903) [clang] Ensure minimal alignment of global vars of incomplete type. (llvm#72886) [IR][TLI] Cache getLibFunc() result on Function (NFC) (llvm#72867) [llvm][ARM] Emit MVE .arch_extension after .fpu directive if it does not include MVE features (llvm#71545) [mlir] Fix a few more TypeSize::Fixed->TypeSize::getFixed following llvm#72979. [llvm][TypeSize] Fix addition/subtraction in TypeSize. (llvm#72979) [IR] Don't include GenericDomTreeConstruction.h in header (NFC) [RISCV] Fix the order of arguments of setTruncStoreAction and setLoadExtAction (llvm#73090) [RISCV] Add more Zbs patterns for -riscv-experimental-rv64-legal-i32. [Clang][Coroutines] Properly emit EH code for initial suspend `await_resume` (llvm#73073) [test][hwasan] Deflake release-shadow.c [BOLT] Fix type mismatch error (llvm#73016) [X86][MC] Support encoding/decoding for JMPABS (llvm#72835) Support BranchProbabilityInfo in update_analyze_test_checks.py (llvm#72943) [mlir][sparse] fixed naming consistency (llvm#73053) [TargetLowering][RISCV] Introduce shouldFoldSelectWithSingleBitTest and RISC-V implement. (llvm#72978) [flang][openacc][NFC] Check only HLFIR lowering for remaining tests (llvm#73054) [flang][openacc][NFC] Check only HLFIR lowering for compute construct tests (llvm#73051) [flang][openacc][NFC] Check only HLFIR lowering for enter/exit data tests (llvm#73035) Supports viewing class member variables in lambda when using the vs debugger (llvm#71564) [flang][openacc][NFC] Remove run line for FIR only checks (llvm#73050) [mlir][sparse] code cleanup (llvm#73047) [nfc][InstrProfTest]Add a test fixture to parameterize the read-write test of value profiles (llvm#73038) [sanitizer_symbolizer] Add MarkupStackTracePrinter (llvm#73032) [NFC sanitizer_symbolizer] Make some functions non virtual in StackTracePrinter. (llvm#73029) [libc++] Make common_iterator's data member private (llvm#72564) test-release.sh: Only build the clang target in stage 1 and 2 (llvm#72703) Allow multiple sanitizers on baremetal targets. (llvm#72933) [clang][NFC] Reorder Atomic builtins to be consistent. (llvm#72718) [SelectionDAG] Fix copy/paste mistake in SDNodeFlags::intersectWith [AMDGPU] NFC: Add flag to disable clustered low occupancy phase (llvm#73025) [nfc][InstrProfTest]Un-parameterize test cases that doesn't use profile reader and writer (llvm#73026) workflows/release-binaries: Do a preliminary build to fill ccache (llvm#72576) [flang] Remove dead code and update test (NFC) (llvm#73004) [libc++][hardening] Categorize all `ryu` assertions as internal (llvm#71853) [Bazel][clang] Fix build for e6ef315 [compiler-rt] Fix interceptors with Solaris as (llvm#72973) [libc++] Promote android to supported. (llvm#72949) [clang codegen][regression] Add dso_local/hidden/etc. markings to VTT definitions and declarations (llvm#72452) [RISCV] Replace XLenVT in RV64 only pattern with i64. NFC Revert "[SVE2.1][Clang][LLVM]Add BFloat16 builtin in Clang and LLVM intrinisc (llvm#70362)" [OpenACC] Implement Atomic construct variants (llvm#73015) [mlir][vector][nfc] Refactor vector.contract matvec tests (llvm#72832) [RISCV] Use short forward branch for ISD::ABS. [RISCV] Add rv32 command line to short-forward-branch-opt.ll. NFC [AMDGPU] NFC. Run auto-update on a few tests Enable customer lowering for fabs_v16f16 with AVX2 (llvm#72914) Reland "[MC][AsmParser] Diagnose improperly nested .cfi frames" [X86] X86FixupVectorConstantsPass - attempt to match VEX logic ops back to EVEX if we can create a broadcast fold [NFC] Fix typos in llvm-mc doc (llvm#72457) [llvm][NFC] Autoupdater x86 detection (llvm#72808) [OpenMP] Optimized trivial multiple edges from task dependency graph [mlir] fix overflow in LLVM dialect inlining (llvm#72878) [mlir][sparse] test on read/convert permuted 3d sparse tensors (llvm#72925) [gn build] Add a generic `compiler_wrapper` gn arg (llvm#72757) [flang][NFC] Move lowering test to correct folder [mlir][sparse] add a csr x bsr matmul test case (llvm#73012) [flang][openacc][NFC] Check only HLFIR lowering for declare tests (llvm#73011) [DebugInfo][RemoveDIs] Interpret DPValue objects in SelectionDAG (llvm#72253) reapply "[TargetInstrInfo] enable foldMemoryOperand for InlineAsm (llvm#70743)" (llvm#72910) [AArch64] Add SVE2.1 intrinsics for indexed quadword gather loads and scatter stores (llvm#70476) [llvm][IRMover] Remove no-op ptr-to-ptr bitcast (NFC) [OpenACC] Implement enter data/exit data construct parsing (llvm#72916) [ELF] Support R_RISCV_SET_ULEB128/R_RISCV_SUB_ULEB128 in non-SHF_ALLOC sections (llvm#72610) [mlir][spirv] Add more CL math ops (llvm#72995) [AMDGPU] - Add constant folding to s_wqm intrinsic (llvm#72382) [AArch64][SVE2.1] Add intrinsics for quadword loads/stores with unscaled offset (llvm#70474) [BasicAA] Optimize index size adjustment (NFC) [RISCV] Convert all floating point vector type operands to integer vector type (llvm#69559) [mlir] use TypeSize and uint64_t in DataLayout (llvm#72874) [clang][dataflow] Remove deprecated synonyms from `Environment`. (llvm#72987) [mlir][mesh] Add collective communication operations (llvm#71960) [InstCombine] Fix incorrect nneg inference on shift amount [InstCombine] Add tests for incorrect shift nneg inference (NFC) [LV] Add test case for diff checks with nested AddRecs. [clang] Fix lit test failure caused by llvm#70762 (llvm#72928) [MLIR][OpenMP] remove now unnecessary getUsedValuesDefinedAbove call from convertTargetOp (llvm#72904) [gn] port e6ef315 [BasicAA] Don't use MinAbsVarIndex = 1. (llvm#72993) Fold `linalg.fill` -> `linalg.copy` (llvm#72920) Revert "[mlir][vector] Move transpose with unit-dim to shape_cast pattern (llvm#72493)" (llvm#72918) [libc][math] Implement nexttoward functions (llvm#72763) [SVE2.1][Clang][LLVM]Add BFloat16 builtin in Clang and LLVM intrinisc (llvm#70362) [BasicAA] Add wrapping test for llvm#72831. [APINotes] Introduce APINotes infrastructure in Clang Sema and Frontend [ValueTracking] Handle operand bundle assumes in same loop (NFCI) [gn build] Port 527fcb8 [libc] Update the AMDGPU implementation to use code object 5 (llvm#72580) [LTO] [LLD] Don't alias the __imp_func and func symbol resolutions (llvm#71376) [analyzer] Add std::variant checker (llvm#66481) [X86] Regenerate ispow2.ll. NFC. [CodeGen] getPointerMemTy - move FIXME to start of comment line so editors are more likely to detect it. NFC. [RegBankInfo] Add brackets around || in assert. NFC [SystemZ][z/OS] Replace unconventional characters that are not within the ASCII range (llvm#72906) [Offload] Initial support for registering offloading entries on COFF targets (llvm#72697) [clang][Analyzer] Move checker 'alpha.unix.Errno' to 'unix.Errno'. (llvm#69469) [mlir][spirv] Add `CL.mix` op (llvm#72800) [mlir][spirv] Add some op decorations (llvm#72809) [BOLT][NFC] Extract a function for dump MCInst (llvm#67225) [BasicAA] Don't assume DT is nonnull [LoopInfo] Avoid redundant DomTree lookup (NFC) [LoongArch][MC] Support to get the FixupKind for BL (llvm#72938) [LoopUtils] Freeze compare results for diff checks instead of pointers. [LLD] [COFF] Add tests to observe details about LTO and __imp_ prefixes. NFC. (llvm#72764) [DebugInfo][RemoveDIs] Implement redundant elimination for DPValues (llvm#72284) [MachineSink][AArch64] Preserve debug location when rematerialising an instruction to replace a COPY (llvm#72685) [ConstraintElim] Use isKnownNonNegative for condition transfer. (llvm#72879) [SVE2.1][Clang][LLVM]Add 128bits builtin in Clang and LLVM intrinisc (llvm#71930) [llvm] Specialize 'NoCFIValue::getType()' (llvm#72923) [NewPM] Remove Delinearization legacy pass (llvm#72942) [NewPM] Remove MakeGuardsExplicitLegacyPass (llvm#72946) [NewPM] Remove LoopPredicationLegacyPass (llvm#72944) [NewPM] Remove CostModelAnalysis Legacy Pass (llvm#72941) [NewPM] Remove LoopGuardWideningLegacyPass (llvm#72937) [NewPM] Remove InstCountLegacyPass (llvm#72936) [NewPM] Remove SimpleLoopUnswitchLegacyPass (llvm#72934) [NewPM] Remove LoopSimplifyCFGLegacyPass (llvm#72930) [BasicAA] Make isNotCapturedBeforeOrAt() check for calls more precise (llvm#69931) [mlir][tensor] Fix crash when canonicalizing invalid IR (llvm#72888) [AMDGPU] - Add bitreplicate const folding tests (llvm#72649) Fix clang/test/Sema/code_align.c for Arm v7 llvm#70762 [mlir][async] Avoid crash when not using `func.func` (llvm#72801) [AMDGPU] Move ballot64 wave32 mode tests to the appropriate place. NFC. (llvm#72959) [Driver] Make ELF -nopie specific to OpenBSD (llvm#72578) Fix build error from llvm#72178 (llvm#72905) Revert "Fix tab spaces in coroutine-hostile-raii.cpp" [Driver,test] Remove -v from mkdir & ln -s from a test since it fails on AIX (llvm#72924) Fix tab spaces in coroutine-hostile-raii.cpp [MachineBlockPlacement][X86] Use max of MDAlign and TLIAlign to align Loops. (llvm#71026) [RISCV] Fix spelling error in test name. NFC [RISCV] DAG combine (mul (add x, 1), y) -> vmadd (llvm#71495) [flang][openacc][NFC] Check only HLFIR lowering for data tests (llvm#72926) [lld][ELF] Add armeb support when incoming bc is arm big endian (llvm#72604) [BOLT] Fix C++ exceptions when LPStart is specified (llvm#72737) Make MLIR Value more consistent in terms of `const` "correctness" (NFC) (llvm#72765) Reland [clang-format][NFC] Remove a redundant isLiteral() call [flang][doc] Added remark about array element references in data clauses. (llvm#72332) [mlir][python] remove eager loading of dialect module (for type and value casting) (llvm#72338) [flang][OpenMP] Add semantic check for device clause (llvm#72789) [OpenMP] Be more forgiving during record and replay [OpenMP][FIX] Ensure recording works properly w/ late allocations [OpenMP][NFC] Remove std::move to silence warnings [mlir] Non-void lambda does not return a value in all control paths in yieldReplacementForFusedProducer (NFC) [LoongArch][MC] Pre-commit tests for instr bl fixupkind testing (llvm#72826) [Clang] Fix finding instantiated decls for class template specializations during instantiation (llvm#72346) [DebugInfo][RemoveDIs] Make dropping variable locations explicit (llvm#72399) [InstCombine] Add transform for `~X + ~Y)` -> `-2 - Y - X` [InstCombine] Replace `isFreeToInvert` + `CreateNot` with `getFreelyInverted` [InstCombine] add `getFreeInverted` to perform folds for free inversion of op [InstCombine] Add additional tests for free inversion; NFC [InstCombine] Make `isFreeToInvert` check recursively. [InstCombine] Add folds for `(X + Y) - (W + Z)` [InstCombine] Add tests for folding `(X + Y) - (W + Z)`; NFC Fix command escape bug in lldb-dap (llvm#72902) [flang][openacc][NFC] Check only HLFIR lowering for atomic tests (llvm#72922) [libc] Remove the optional arguments for NVPTX constructors (llvm#69536) Revert "[nfc][InstrProfTest]Factor out common code for value profile test" (llvm#72921) [nfc][InstrProfTest]Factor out common code for value profile test (llvm#72611) [MemorySSA] Update test to use NewPM (llvm#72915) [ConstraintElim] Add extra info transfer tests for llvm#72879. [CodeGenModule] Remove no-op ptr-to-ptr bitcasts (NFC) Revert "[clang-format][NFC] Remove a redundant isLiteral() call" [clang] Add support for new loop attribute [[clang::code_align()]] (llvm#70762) [SelectionDAG] Add support to filter SelectionDAG dumps during ISel by function names (llvm#72696) [CUDA][HIP] ignore implicit host/device attr for override (llvm#72815) [DebugInfo][RemoveDIs] Don't convert debug-intrinsics to Unreachable (llvm#72380) [mlir][sparse] test four row/col major versions of BSR (llvm#72898) [lldb][split-dwarf] implement GetSeparateDebugInfo for SymbolFileOnDemand (llvm#71230) Move all libc++ builders to one machine type. [OpenACC] Implement compound construct parsing (llvm#72692) [llvm-profdata] Fix binary ids with multiple raw profiles in a single… (llvm#72740) Fix to attribute plugins reaching an unreachable (llvm#70877) LoopVectorize: Add better heuristic for vectorized epilogue skip test (llvm#72589) [Libomptarget] Handle dynamic stack sizes for AMD COV5 (llvm#72606) [libc++] Don't open-close namespace std in __config (llvm#72695) [libc++] Reduce the compilation time required by SIMD tests (llvm#72602) [mlir][sparse] code cleanup using the assumption that dim2lvl maps ar… (llvm#72894) [BOLT][TEST] Remove LTO flag from a test (llvm#72896) [build_symbolizer] Fix a missing mkdir cmd [Clang] Correct handling of negative and out-of-bounds indices (llvm#71877) [mlir][spirv][nfc] Sort CL ops (llvm#72868) [libc] Adjust headers for some implementations of 'stdio.h' [NFC] Simplify the tiling implementation using cloning. (llvm#72178) [DebugInfo][RemoveDIs] Add local-utility plumbing for DPValues (llvm#72276) [clang][Interp][NFC] Add const InterpBlock::deref() overload [flang] Remove extra space added with --dependent-lib option Revert "[OpenMP] atomic compare fail : Parser & AST support" [Flang][OpenMP] NFC: Formatting change [SLP]Fix PR72833: do not crash if only operand is casted but the use instruction. [libc++] Stop checking for trailing whitespace in check-generated-output (llvm#72711) [clang][driver] Add <executable>/../include/c++/v1 to include path on Darwin (llvm#70817) NFCI: update debug-names-types test to use an output file unique to the test [C23] Complete support for WG14 N2508 (llvm#71398) Add @ftynse as GitHub codeowner for MLIR Transform dialect (llvm#72882) [X86] combineLoad - try to reuse existing constant pool entries for smaller vector constant data (REAPPLIED) [InstCombine] Propagate NUW flags for `shl (lshr X, C1), C2 -> shl X, C2-C1` (llvm#72525) [InstSimplify] Remove redundant gep zero fold (NFC) [InstrProf] Add pgo use block coverage test (llvm#72443) [X86] constant-pool-sharing.ll - add test showing failure to reuse subvectors when storing larger vector types [X86] Regenerate constant-pool-sharing.ll with AVX test coverage [LV] Stability fix for outerloop vectorization (llvm#68118) [SLP]Do not emit int bitcast after minbitwidth analysis. [mlir][ArmSME] Move vector.extract/insert lowerings to vector-to-arm-sme (NFC) (llvm#72852) [AMDGPU] Fix PromoteAlloca size check of alloca for store (llvm#72528) [LV] Precommit tests for uniform arguments for vector function variants [LV] Refactor vector function variant selection to prepare for uniform args (llvm#68879) [mlir][Docs] Code review is now done on GitHub [llvm-c] Fix outdated comment (NFC) [NFC][SROA] Remove implementation details from SROA header (llvm#72846) [AMDGPU] Add live-through register set printing to GCNRegPressurePrinter pass. (llvm#71096) Add llvm-mca to the list of toolchain tools (llvm#72840) [clang][Interp][NFC] Use isArrayRoot() when comparing pointers [DAG] clang-format createBranchMacroFusionDAGMutation calls. NFC. [X86] vector-half-conversions.ll - regenerate with AVX512 slow/fast lane shuffles [clang][Interp][NFC] Factor array element init into its own function [AArch64] Add missing bf16 store pattern (llvm#72844) [LAA] Add extra test for llvm#70819 showing incorrect Forward dep. Fix typo in DiagnosticSemaKinds.td [DAG] narrowExtractedVectorBinOp - ensure we limit late node creation to LegalOperations only (llvm#72130) [InstSimplify] Fold converted urem to 0 if there's no overlapping bits (llvm#71528) [mlir][vector] Modernize `vector.transpose` op (llvm#72594) [AArch64] Fix big endian shuffle vector miscompile (llvm#68673) [lit] Use raw strings for backslash escapes to fix SyntaxWarnings (llvm#70907) [MachineLICM][AArch64] Hoist COPY instructions with other uses in the loop (llvm#71403) [AArch64][SME] Remove immediate argument restriction for svldr and svstr (llvm#68565) [PowerPC] Use MIR test so that it's not affected by instruction selection. NFC. Apply clang-tidy fixes for llvm-else-after-return in GPUToLLVMConversion.cpp (NFC) Apply clang-tidy fixes for llvm-qualified-auto in ToLLVMInterface.cpp (NFC) Apply clang-tidy fixes for llvm-qualified-auto in ConvertToLLVMPass.cpp (NFC) Apply clang-tidy fixes for llvm-include-order in ComplexToLLVM.cpp (NFC) Apply clang-tidy fixes for llvm-else-after-return in AMDGPUToROCDL.cpp (NFC) [llvm-exegesis] Preserve rcx and r11 around system call (llvm#72807) [llvm-exegesis] Fix race condition in subprocess mode (llvm#72778) [NewPM] Remove ScalarizerLegacyPass (llvm#72814) [OpenMP] atomic compare fail : Parser & AST support [NewPM] Remove LowerWidenableConditionLegacyPass (llvm#72818) [NewPM] Remove AssumeBundleBuilderPassLegacyPass (llvm#72817) [NewPM] Remove UnifyFunctionExitNodesLegacyPass (llvm#72816) [AMDGPU] Emit backend_stack_size PAL metadata (llvm#72509) Revert "Apply clang-tidy fixes for misc-include-cleaner in IRCore.cpp (NFC)" [NewPM] Remove MergedLoadStoreMotionLegacyPass (llvm#72813) [NewPM] Remove LoopInstSimplifyLegacyPass (llvm#72812) [NewPM] Remove LoopSinkLegacy Pass (llvm#72811) [NewPM] Remove GuardWideningLegacyPass (llvm#72810) [mlir][vector] Extend TransferReadDropUnitDimsPattern to support partially-static memrefs (llvm#72142) [AArch64][SME] Add support for sme-fa64 (llvm#70809) [ARM][FPEnv] Lowering of fpenv intrinsics [PowerPC] Precommit test to show codegen while `isel` is unavailable. NFC. [PowerPC] Disable float128 on AIX in Clang (llvm#67298) Revert "Add new API in SBTarget for loading core from SBFile (llvm#71769)" Add @MaheshRavishankar to CODEOWNERS on relevant source files. (llvm#72449) [RISCV] Remove checks that MI's info is valid. NFC [RISCV] Use DemandedFields instead of checking for vmv.s.x/vmv.x.s. NFC [InstCombine] Convert or concat to fshl if opposite or concat exists (llvm#68502) [RISCV] postpone removal in initundef pass (llvm#71661) [NFC][Clang] Refactor code to calculate flexible array member size (llvm#72790) [PowerPC][EarlyIfConversion] Do not insert `isel` if subtarget doesn't support `isel` (llvm#72211) [mailmap] Add my entry [clang-format] Correctly annotate braces of empty functions (llvm#72733) [clang-format] Fix a bug in isStartOfName() on macro definitions (llvm#72768) [clang-format] Fix a bug in aligning comments above PPDirective (llvm#72791) [clang-format][NFC] Skip alignArrayInitializers() for 1-row matrices (llvm#72166) Revert "[MC][AsmParser] Diagnose improperly nested .cfi frames" [RISCV] Don't set nsw/nuw/exact flag after MachineCombiner reassociation. Revert "[mlir][affine] implement `promoteIfSingleIteration` for `AffineForOp` (llvm#72547)" [test] Fix misused Joined -e options [mlir][affine] implement `promoteIfSingleIteration` for `AffineForOp` (llvm#72547) [InstCombine] Add transforms for `(icmp uPred (trunc x),(truncOrZext(y)))`->`(icmp uPred x,y)` [InstCombine] Add tests for transforming `(icmp eq/ne trunc(x), truncOrZext(y))`; NFC Recommit "[DAGCombiner] Transform `(icmp eq/ne (and X,C0),(shift X,C1))` to use rotate or to getter constants." (2nd Try) [X86] Add more tests for transform `(icmp eq/ne (and X,C0),(shift X,C1))`; PR71598 [InstCombine] Don't transform `sub X, ~Y` -> `add X, -Y` unless `Y` is actually negatable [InstCombine] Add tests for improving `sub X, ~Y` -> `add X, -Y`; NFC [InstCombine] Propagate NSW/NUW flags for `(X - Y) - Z -> X - (Y + Z)` (llvm#72693) [libc++] Use __is_pointer_in_range for char_traits checks (llvm#72643) Revert "[TargetInstrInfo] enable foldMemoryOperand for InlineAsm (llvm#70743)" [Driver] Enable __float128 support on X86 on FreeBSD / NetBSD (llvm#72788) [Github] Fix typo [Github] Prevent scorecard action from running on forks (llvm#72780) [NFC] Fix typos in comments [clang] Remove unused selStructPtrTy in CGObjCGNU.cpp (NFC) [clang] Remove ConstantAggregateBuilderBase::addBitCast (NFC) [DebugInfo][RemoveDIs] Support finding DPValues like dbg.values (llvm#71952) Apply clang-tidy fixes for performance-unnecessary-value-param in IRInterfaces.cpp (NFC) Apply clang-tidy fixes for misc-include-cleaner in IRInterfaces.cpp (NFC) Apply clang-tidy fixes for readability-identifier-naming in IRCore.cpp (NFC) Apply clang-tidy fixes for misc-include-cleaner in IRCore.cpp (NFC) Apply clang-tidy fixes for llvm-else-after-return in IRCore.cpp (NFC) Apply clang-tidy fixes for performance-unnecessary-value-param in IRAttributes.cpp (NFC) [CGObjCMac] Replace calls to ConstantAggregateBuilderBase::addBitCast (NFC) Revert rGbfbfd1caa4da "[X86] combineLoad - try to reuse existing constant pool entries for smaller vector constant data" [DAG] Fix ShrinkDemandedOp doxygen description to match behaviour. NFC. [LV] Don't crash on vector masks during scalar VPReductionRecipe::exec. [LV] Retain mask-reversal comment as suggested after e5e71af. [RISCV][GISel] Instruction selection for G_JUMP_TABLE and G_BRJT. (llvm#71987) [RISCV][GISel] Add s32 G_SELECT instruction select test for RV64. NFC [Github] Add build Flang docs in CI if autogenerated files change (llvm#72721) [AArch64] Add some testing for BE shuffles. NFC [CGObjCGNU] Remove unneeded method 'CGObjCGNUstep2::EnforceType' (NFC) [bazel] Port dc4786b [mlir][affine] Remove unused captures [RISCV][GISel] Remove the rv32/rv64 subdirectories for legalizer tests. NFC [MC] Remove duplicate Contents field from MCLEBFragment. [mlir][affine] remove divide zero check when simplifer affineMap (llvm#64622) (llvm#68519) Change-Id: I30498962c4a19676f8d3a534a80fab36dbfdf0a2 Signed-off-by: greenforce-auto-merge <greenforce-auto-merge@users.noreply.github.com>
Guzhu-AMD
pushed a commit
to GPUOpen-Drivers/llvm-project
that referenced
this pull request
Nov 30, 2023
Local branch amd-gfx 461597c Merged main:0a9c6bea6b08 into amd-gfx:fad78e377956 Remote branch main ba89749 [libc++] <experimental/simd> Add implicit type conversion constructor for class simd/simd_mask (llvm#71132)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.