diff options
Diffstat (limited to 'llvm/utils')
25 files changed, 224 insertions, 72 deletions
diff --git a/llvm/utils/TableGen/Basic/CodeGenIntrinsics.cpp b/llvm/utils/TableGen/Basic/CodeGenIntrinsics.cpp index be7537c..cd86646 100644 --- a/llvm/utils/TableGen/Basic/CodeGenIntrinsics.cpp +++ b/llvm/utils/TableGen/Basic/CodeGenIntrinsics.cpp @@ -278,15 +278,21 @@ CodeGenIntrinsic::CodeGenIntrinsic(const Record *R, TargetPrefix = R->getValueAsString("TargetPrefix"); Name = R->getValueAsString("LLVMName").str(); + std::string DefaultName = "llvm." + EnumName.str(); + llvm::replace(DefaultName, '_', '.'); + if (Name == "") { // If an explicit name isn't specified, derive one from the DefName. - Name = "llvm." + EnumName.str(); - llvm::replace(Name, '_', '.'); + Name = std::move(DefaultName); } else { // Verify it starts with "llvm.". if (!StringRef(Name).starts_with("llvm.")) PrintFatalError(DefLoc, "Intrinsic '" + DefName + "'s name does not start with 'llvm.'!"); + + if (Name == DefaultName) + PrintNote(DefLoc, "Explicitly specified name matches default name, " + "consider dropping it"); } // If TargetPrefix is specified, make sure that Name starts with diff --git a/llvm/utils/TableGen/InstrInfoEmitter.cpp b/llvm/utils/TableGen/InstrInfoEmitter.cpp index d1b14fb..0b90f91 100644 --- a/llvm/utils/TableGen/InstrInfoEmitter.cpp +++ b/llvm/utils/TableGen/InstrInfoEmitter.cpp @@ -285,7 +285,7 @@ emitGetNamedOperandIdx(raw_ostream &OS, static void emitGetOperandIdxName(raw_ostream &OS, - MapVector<StringRef, unsigned> OperandNameToID, + const MapVector<StringRef, unsigned> &OperandNameToID, const MapVector<SmallVector<int>, unsigned> &OperandMap, unsigned MaxNumOperands, unsigned NumOperandNames) { OS << "LLVM_READONLY OpName getOperandIdxName(uint16_t Opcode, int16_t Idx) " diff --git a/llvm/utils/gn/secondary/clang/lib/AST/BUILD.gn b/llvm/utils/gn/secondary/clang/lib/AST/BUILD.gn index 9981d10..4da907c 100644 --- a/llvm/utils/gn/secondary/clang/lib/AST/BUILD.gn +++ b/llvm/utils/gn/secondary/clang/lib/AST/BUILD.gn @@ -121,6 +121,7 @@ static_library("AST") { "ExternalASTMerger.cpp", "ExternalASTSource.cpp", "FormatString.cpp", + "InferAlloc.cpp", "InheritViz.cpp", "ItaniumCXXABI.cpp", "ItaniumMangle.cpp", diff --git a/llvm/utils/gn/secondary/clang/lib/Headers/BUILD.gn b/llvm/utils/gn/secondary/clang/lib/Headers/BUILD.gn index 29c6178..3c523ae 100644 --- a/llvm/utils/gn/secondary/clang/lib/Headers/BUILD.gn +++ b/llvm/utils/gn/secondary/clang/lib/Headers/BUILD.gn @@ -104,6 +104,9 @@ copy("Headers") { "__clang_hip_runtime_wrapper.h", "__clang_hip_stdlib.h", "__clang_spirv_builtins.h", + "__float_float.h", + "__float_header_macro.h", + "__float_infinity_nan.h", "__stdarg___gnuc_va_list.h", "__stdarg___va_copy.h", "__stdarg_header_macro.h", diff --git a/llvm/utils/gn/secondary/libcxx/include/BUILD.gn b/llvm/utils/gn/secondary/libcxx/include/BUILD.gn index e747006..278c29c 100644 --- a/llvm/utils/gn/secondary/libcxx/include/BUILD.gn +++ b/llvm/utils/gn/secondary/libcxx/include/BUILD.gn @@ -911,7 +911,6 @@ if (current_toolchain == default_toolchain) { "__cxx03/cmath", "__cxx03/codecvt", "__cxx03/complex", - "__cxx03/complex.h", "__cxx03/condition_variable", "__cxx03/csetjmp", "__cxx03/csignal", @@ -924,25 +923,20 @@ if (current_toolchain == default_toolchain) { "__cxx03/cstring", "__cxx03/ctgmath", "__cxx03/ctime", - "__cxx03/ctype.h", "__cxx03/cuchar", "__cxx03/cwchar", "__cxx03/cwctype", "__cxx03/deque", - "__cxx03/errno.h", "__cxx03/exception", "__cxx03/experimental/__config", "__cxx03/experimental/utility", "__cxx03/ext/__hash", "__cxx03/ext/hash_map", "__cxx03/ext/hash_set", - "__cxx03/fenv.h", - "__cxx03/float.h", "__cxx03/forward_list", "__cxx03/fstream", "__cxx03/functional", "__cxx03/future", - "__cxx03/inttypes.h", "__cxx03/iomanip", "__cxx03/ios", "__cxx03/iosfwd", @@ -969,11 +963,8 @@ if (current_toolchain == default_toolchain) { "__cxx03/sstream", "__cxx03/stack", "__cxx03/stdatomic.h", - "__cxx03/stdbool.h", - "__cxx03/stddef.h", "__cxx03/stdexcept", "__cxx03/stdint.h", - "__cxx03/stdio.h", "__cxx03/stdlib.h", "__cxx03/streambuf", "__cxx03/string", @@ -981,7 +972,6 @@ if (current_toolchain == default_toolchain) { "__cxx03/string_view", "__cxx03/strstream", "__cxx03/system_error", - "__cxx03/tgmath.h", "__cxx03/thread", "__cxx03/type_traits", "__cxx03/typeindex", @@ -994,7 +984,6 @@ if (current_toolchain == default_toolchain) { "__cxx03/vector", "__cxx03/version", "__cxx03/wchar.h", - "__cxx03/wctype.h", "__debug_utils/randomize_range.h", "__debug_utils/sanitizers.h", "__debug_utils/strict_weak_ordering_check.h", diff --git a/llvm/utils/gn/secondary/lldb/source/Plugins/ExpressionParser/Clang/BUILD.gn b/llvm/utils/gn/secondary/lldb/source/Plugins/ExpressionParser/Clang/BUILD.gn index 5efc153..51911d7 100644 --- a/llvm/utils/gn/secondary/lldb/source/Plugins/ExpressionParser/Clang/BUILD.gn +++ b/llvm/utils/gn/secondary/lldb/source/Plugins/ExpressionParser/Clang/BUILD.gn @@ -47,7 +47,6 @@ static_library("Clang") { "ClangASTImporter.cpp", "ClangASTMetadata.cpp", "ClangASTSource.cpp", - "ClangDeclVendor.cpp", "ClangExpressionDeclMap.cpp", "ClangExpressionHelper.cpp", "ClangExpressionParser.cpp", diff --git a/llvm/utils/gn/secondary/llvm/include/llvm/Config/BUILD.gn b/llvm/utils/gn/secondary/llvm/include/llvm/Config/BUILD.gn index 022cd87..1ca9292 100644 --- a/llvm/utils/gn/secondary/llvm/include/llvm/Config/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/include/llvm/Config/BUILD.gn @@ -301,7 +301,6 @@ write_cmake_config("llvm-config") { "LLVM_BUILD_SHARED_LIBS=", "LLVM_ENABLE_LLVM_C_EXPORT_ANNOTATIONS=", "LLVM_ENABLE_TELEMETRY=", - "LLVM_ENABLE_ONDISK_CAS=", "LLVM_DEFAULT_TARGET_TRIPLE=$llvm_target_triple", "LLVM_ENABLE_DEBUGLOC_TRACKING_COVERAGE=", "LLVM_ENABLE_DEBUGLOC_TRACKING_ORIGIN=", @@ -367,6 +366,12 @@ write_cmake_config("llvm-config") { values += [ "LLVM_ENABLE_DIA_SDK=" ] } + if (llvm_enable_ondisk_cas) { + values += [ "LLVM_ENABLE_ONDISK_CAS=1" ] + } else { + values += [ "LLVM_ENABLE_ONDISK_CAS=" ] + } + if (llvm_enable_threads) { values += [ "LLVM_ENABLE_THREADS=1" ] } else { diff --git a/llvm/utils/gn/secondary/llvm/include/llvm/Config/config.gni b/llvm/utils/gn/secondary/llvm/include/llvm/Config/config.gni index 8c2ab8a..715b03e 100644 --- a/llvm/utils/gn/secondary/llvm/include/llvm/Config/config.gni +++ b/llvm/utils/gn/secondary/llvm/include/llvm/Config/config.gni @@ -1,4 +1,7 @@ declare_args() { # Iterate unordered llvm containers in reverse. llvm_enable_reverse_iteration = false + + # Iterate unordered llvm containers in reverse. + llvm_enable_ondisk_cas = false } diff --git a/llvm/utils/gn/secondary/llvm/lib/AsmParser/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/AsmParser/BUILD.gn index 9ff794f..2ad867d 100644 --- a/llvm/utils/gn/secondary/llvm/lib/AsmParser/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/lib/AsmParser/BUILD.gn @@ -6,6 +6,7 @@ static_library("AsmParser") { "//llvm/lib/Support", ] sources = [ + "AsmParserContext.cpp", "LLLexer.cpp", "LLParser.cpp", "Parser.cpp", diff --git a/llvm/utils/gn/secondary/llvm/lib/CAS/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/CAS/BUILD.gn index b4edd8d..5590b27 100644 --- a/llvm/utils/gn/secondary/llvm/lib/CAS/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/lib/CAS/BUILD.gn @@ -10,6 +10,8 @@ static_library("CAS") { "ObjectStore.cpp", "OnDiskCommon.cpp", "OnDiskDataAllocator.cpp", + "OnDiskGraphDB.cpp", + "OnDiskKeyValueDB.cpp", "OnDiskTrieRawHashMap.cpp", ] } diff --git a/llvm/utils/gn/secondary/llvm/lib/CodeGen/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/CodeGen/BUILD.gn index 39dacf7..4446702 100644 --- a/llvm/utils/gn/secondary/llvm/lib/CodeGen/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/lib/CodeGen/BUILD.gn @@ -122,6 +122,7 @@ static_library("CodeGen") { "MLRegAllocPriorityAdvisor.cpp", "MachineBasicBlock.cpp", "MachineBlockFrequencyInfo.cpp", + "MachineBlockHashInfo.cpp", "MachineBlockPlacement.cpp", "MachineBranchProbabilityInfo.cpp", "MachineCFGPrinter.cpp", diff --git a/llvm/utils/gn/secondary/llvm/lib/ExecutionEngine/Orc/TargetProcess/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/ExecutionEngine/Orc/TargetProcess/BUILD.gn index c4ce990..937e81b 100644 --- a/llvm/utils/gn/secondary/llvm/lib/ExecutionEngine/Orc/TargetProcess/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/lib/ExecutionEngine/Orc/TargetProcess/BUILD.gn @@ -12,6 +12,8 @@ static_library("TargetProcess") { "JITLoaderGDB.cpp", "JITLoaderPerf.cpp", "JITLoaderVTune.cpp", + "LibraryResolver.cpp", + "LibraryScanner.cpp", "OrcRTBootstrap.cpp", "RegisterEHFrames.cpp", "SimpleExecutorDylibManager.cpp", diff --git a/llvm/utils/gn/secondary/llvm/lib/Support/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/Support/BUILD.gn index 38ba466..df9ddf9 100644 --- a/llvm/utils/gn/secondary/llvm/lib/Support/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/lib/Support/BUILD.gn @@ -45,6 +45,7 @@ static_library("Support") { "ARMAttributeParser.cpp", "ARMBuildAttributes.cpp", "ARMWinEH.cpp", + "AllocToken.cpp", "Allocator.cpp", "AutoConvert.cpp", "BalancedPartitioning.cpp", diff --git a/llvm/utils/gn/secondary/llvm/lib/Target/AMDGPU/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/Target/AMDGPU/BUILD.gn index c89e335..e47ca1e 100644 --- a/llvm/utils/gn/secondary/llvm/lib/Target/AMDGPU/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/lib/Target/AMDGPU/BUILD.gn @@ -137,6 +137,7 @@ static_library("LLVMAMDGPUCodeGen") { "AMDGPUAsmPrinter.cpp", "AMDGPUAtomicOptimizer.cpp", "AMDGPUAttributor.cpp", + "AMDGPUBarrierLatency.cpp", "AMDGPUCallLowering.cpp", "AMDGPUCodeGenPrepare.cpp", "AMDGPUCombinerHelper.cpp", diff --git a/llvm/utils/gn/secondary/llvm/lib/Target/Hexagon/BUILD.gn b/llvm/utils/gn/secondary/llvm/lib/Target/Hexagon/BUILD.gn index c055001..065d33d 100644 --- a/llvm/utils/gn/secondary/llvm/lib/Target/Hexagon/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/lib/Target/Hexagon/BUILD.gn @@ -79,6 +79,7 @@ static_library("LLVMHexagonCodeGen") { "HexagonOptAddrMode.cpp", "HexagonOptimizeSZextends.cpp", "HexagonPeephole.cpp", + "HexagonQFPOptimizer.cpp", "HexagonRDFOpt.cpp", "HexagonRegisterInfo.cpp", "HexagonSelectionDAGInfo.cpp", diff --git a/llvm/utils/gn/secondary/llvm/tools/llvm-ir2vec/BUILD.gn b/llvm/utils/gn/secondary/llvm/tools/llvm-ir2vec/BUILD.gn index 07a7951..4d75201 100644 --- a/llvm/utils/gn/secondary/llvm/tools/llvm-ir2vec/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/tools/llvm-ir2vec/BUILD.gn @@ -1,9 +1,16 @@ executable("llvm-ir2vec") { deps = [ "//llvm/lib/Analysis", + "//llvm/lib/CodeGen", + "//llvm/lib/CodeGen/MIRParser", "//llvm/lib/IR", "//llvm/lib/IRReader", "//llvm/lib/Support", + "//llvm/lib/Target:AllTargetsAsmParsers", + "//llvm/lib/Target:AllTargetsCodeGens", + "//llvm/lib/Target:AllTargetsDescs", + "//llvm/lib/Target:AllTargetsInfos", + "//llvm/lib/TargetParser", ] sources = [ "llvm-ir2vec.cpp" ] } diff --git a/llvm/utils/gn/secondary/llvm/unittests/CAS/BUILD.gn b/llvm/utils/gn/secondary/llvm/unittests/CAS/BUILD.gn index 52a64be..2d9eb68 100644 --- a/llvm/utils/gn/secondary/llvm/unittests/CAS/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/unittests/CAS/BUILD.gn @@ -1,3 +1,4 @@ +import("//llvm/include/llvm/Config/config.gni") import("//third-party/unittest/unittest.gni") unittest("CASTests") { @@ -10,8 +11,15 @@ unittest("CASTests") { "ActionCacheTest.cpp", "CASTestConfig.cpp", "ObjectStoreTest.cpp", - "OnDiskDataAllocatorTest.cpp", - "OnDiskTrieRawHashMapTest.cpp", - "ProgramTest.cpp", ] + + if (llvm_enable_ondisk_cas) { + sources += [ + "OnDiskDataAllocatorTest.cpp", + "OnDiskGraphDBTest.cpp", + "OnDiskKeyValueDBTest.cpp", + "OnDiskTrieRawHashMapTest.cpp", + "ProgramTest.cpp", + ] + } } diff --git a/llvm/utils/gn/secondary/llvm/unittests/ExecutionEngine/Orc/BUILD.gn b/llvm/utils/gn/secondary/llvm/unittests/ExecutionEngine/Orc/BUILD.gn index 376f689..111e4c9 100644 --- a/llvm/utils/gn/secondary/llvm/unittests/ExecutionEngine/Orc/BUILD.gn +++ b/llvm/utils/gn/secondary/llvm/unittests/ExecutionEngine/Orc/BUILD.gn @@ -24,6 +24,7 @@ unittest("OrcJITTests") { "JITLinkRedirectionManagerTest.cpp", "JITTargetMachineBuilderTest.cpp", "LazyCallThroughAndReexportsTest.cpp", + "LibraryResolverTest.cpp", "LookupAndRecordAddrsTest.cpp", "MachOPlatformTest.cpp", "MapperJITLinkMemoryManagerTest.cpp", @@ -42,6 +43,7 @@ unittest("OrcJITTests") { "SymbolStringPoolTest.cpp", "TaskDispatchTest.cpp", "ThreadSafeModuleTest.cpp", + "WaitingOnGraphTest.cpp", "WrapperFunctionUtilsTest.cpp", ] diff --git a/llvm/utils/lit/lit/TestRunner.py b/llvm/utils/lit/lit/TestRunner.py index a7e2705..f883145 100644 --- a/llvm/utils/lit/lit/TestRunner.py +++ b/llvm/utils/lit/lit/TestRunner.py @@ -92,12 +92,12 @@ class ShellEnvironment(object): we maintain a dir stack for pushd/popd. """ - def __init__(self, cwd, env, umask=-1, ulimit={}): + def __init__(self, cwd, env, umask=-1, ulimit=None): self.cwd = cwd self.env = dict(env) self.umask = umask self.dirStack = [] - self.ulimit = ulimit + self.ulimit = ulimit if ulimit else {} def change_dir(self, newdir): if os.path.isabs(newdir): diff --git a/llvm/utils/lit/lit/run.py b/llvm/utils/lit/lit/run.py index 62070e8..55de914 100644 --- a/llvm/utils/lit/lit/run.py +++ b/llvm/utils/lit/lit/run.py @@ -137,6 +137,10 @@ class Run(object): "Raised process limit from %d to %d" % (soft_limit, desired_limit) ) except Exception as ex: - # Warn, unless this is Windows or z/OS, in which case this is expected. - if os.name != "nt" and platform.system() != "OS/390": + # Warn, unless this is Windows, z/OS, or Cygwin in which case this is expected. + if ( + os.name != "nt" + and platform.system() != "OS/390" + and platform.sys.platform != "cygwin" + ): self.lit_config.warning("Failed to raise process limit: %s" % ex) diff --git a/llvm/utils/lit/tests/Inputs/shtest-ulimit/ulimit_reset.txt b/llvm/utils/lit/tests/Inputs/shtest-ulimit/ulimit_reset.txt new file mode 100644 index 0000000..011d6db --- /dev/null +++ b/llvm/utils/lit/tests/Inputs/shtest-ulimit/ulimit_reset.txt @@ -0,0 +1,3 @@ +# RUN: %{python} %S/print_limits.py +# Fail the test so that we can assert on the output. +# RUN: not echo return diff --git a/llvm/utils/lit/tests/shtest-ulimit.py b/llvm/utils/lit/tests/shtest-ulimit.py index e843277..09cd475 100644 --- a/llvm/utils/lit/tests/shtest-ulimit.py +++ b/llvm/utils/lit/tests/shtest-ulimit.py @@ -3,11 +3,15 @@ # ulimit does not work on non-POSIX platforms. # Solaris for some reason does not respect ulimit -n, so mark it unsupported # as well. -# UNSUPPORTED: system-windows, system-solaris +# UNSUPPORTED: system-windows, system-cygwin, system-solaris -# RUN: not %{lit} -a -v %{inputs}/shtest-ulimit | FileCheck %s +# RUN: %{python} %S/Inputs/shtest-ulimit/print_limits.py | grep RLIMIT_NOFILE \ +# RUN: | sed -n -e 's/.*=//p' | tr -d '\n' > %t.nofile_limit -# CHECK: -- Testing: 2 tests{{.*}} +# RUN: not %{lit} -a -v %{inputs}/shtest-ulimit --order=lexical \ +# RUN: | FileCheck -DBASE_NOFILE_LIMIT=%{readfile:%t.nofile_limit} %s + +# CHECK: -- Testing: 3 tests{{.*}} # CHECK-LABEL: FAIL: shtest-ulimit :: ulimit-bad-arg.txt ({{[^)]*}}) # CHECK: ulimit -n @@ -16,3 +20,6 @@ # CHECK-LABEL: FAIL: shtest-ulimit :: ulimit_okay.txt ({{[^)]*}}) # CHECK: ulimit -n 50 # CHECK: RLIMIT_NOFILE=50 + +# CHECK-LABEL: FAIL: shtest-ulimit :: ulimit_reset.txt ({{[^)]*}}) +# CHECK: RLIMIT_NOFILE=[[BASE_NOFILE_LIMIT]] diff --git a/llvm/utils/lldbDataFormatters.py b/llvm/utils/lldbDataFormatters.py index 5e553ca..a3e4ae1 100644 --- a/llvm/utils/lldbDataFormatters.py +++ b/llvm/utils/lldbDataFormatters.py @@ -197,6 +197,11 @@ def StringRefSummaryProvider(valobj, internal_dict): return '""' data = data_pointer.deref + # StringRef may be uninitialized with length exceeding available memory, + # potentially causing bad_alloc exceptions. Limit the length to max string summary setting. + limit_obj = valobj.target.debugger.GetSetting("target.max-string-summary-length") + if limit_obj: + length = min(length, limit_obj.GetUnsignedIntegerValue()) # Get a char[N] type, from the underlying char type. array_type = data.type.GetArrayType(length) # Cast the char* string data to a char[N] array. diff --git a/llvm/utils/mlgo-utils/IR2Vec/generateTriplets.py b/llvm/utils/mlgo-utils/IR2Vec/generateTriplets.py index 80ac4c6..dba9e2c 100644 --- a/llvm/utils/mlgo-utils/IR2Vec/generateTriplets.py +++ b/llvm/utils/mlgo-utils/IR2Vec/generateTriplets.py @@ -1,14 +1,19 @@ # Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. # See https://llvm.org/LICENSE.txt for license information. # SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception -"""IR2Vec Triplet Generator +"""IR2Vec/MIR2Vec Triplet Generator -Generates IR2Vec triplets by applying random optimization levels to LLVM IR files -and extracting triplets using llvm-ir2vec. Automatically generates preprocessed -files: entity2id.txt, relation2id.txt, and train2id.txt. +Generates IR2Vec or MIR2Vec triplets by applying random optimization levels to +LLVM IR files (or processing MIR files) and extracting triplets using llvm-ir2vec. +Automatically generates preprocessed files (entity2id.txt, relation2id.txt, and +train2id.txt) necessary for training IR2Vec or MIR2Vec vocabularies. Usage: - python generateTriplets.py <llvm_build_dir> <num_optimizations> <ll_file_list> <output_dir> + For LLVM IR: + python generateTriplets.py <llvm_build_dir> <num_optimizations> <ll_file_list> <output_dir> + + For Machine IR: + python generateTriplets.py --mode=mir <llvm_build_dir> <mir_file_list> <output_dir> """ import argparse @@ -41,7 +46,7 @@ class TripletResult: class IR2VecTripletGenerator: - """Main class for generating IR2Vec triplets""" + """Main class for generating IR2Vec or MIR2Vec triplets""" def __init__( self, @@ -49,11 +54,13 @@ class IR2VecTripletGenerator: num_optimizations: int, output_dir: Path, max_workers: int = DEFAULT_MAX_WORKERS, + mode: str = "llvm", ): self.llvm_build_dir = llvm_build_dir self.num_optimizations = num_optimizations self.output_dir = output_dir self.max_workers = max_workers + self.mode = mode # "llvm" or "mir" # Tool paths self.opt_binary = os.path.join(llvm_build_dir, "bin", "opt") @@ -85,7 +92,11 @@ class IR2VecTripletGenerator: f"llvm-ir2vec binary not found or not executable: {self.ir2vec_binary}" ) - if not (1 <= self.num_optimizations <= len(OPT_LEVELS)): + if self.mode not in ["llvm", "mir"]: + raise ValueError(f"Mode must be 'llvm' or 'mir', got: {self.mode}") + + # For LLVM IR mode, validate optimization count + if self.mode == "llvm" and not (1 <= self.num_optimizations <= len(OPT_LEVELS)): raise ValueError( f"Number of optimizations must be between 1-{len(OPT_LEVELS)}" ) @@ -95,19 +106,28 @@ class IR2VecTripletGenerator: return random.sample(OPT_LEVELS, self.num_optimizations) def _process_single_file(self, input_file: Path) -> TripletResult: - """Process a single LLVM IR file with multiple optimization levels""" + """Process a single LLVM IR or MIR file""" all_triplets = set() max_relation = 1 - opt_levels = self._select_optimization_levels() - for opt_level in opt_levels: - triplets, file_max_relation = self._run_pipeline(input_file, opt_level) + if self.mode == "mir": + # For MIR files, process directly without optimization + triplets, file_max_relation = self._run_mir_pipeline(input_file) if triplets: all_triplets.update(triplets) max_relation = max(max_relation, file_max_relation) - logger.debug( - f"Generated {len(triplets)} triplets for {input_file} with {opt_level}" - ) + logger.debug(f"Generated {len(triplets)} triplets for {input_file}") + else: + # For LLVM IR files, apply multiple optimization levels + opt_levels = self._select_optimization_levels() + for opt_level in opt_levels: + triplets, file_max_relation = self._run_pipeline(input_file, opt_level) + if triplets: + all_triplets.update(triplets) + max_relation = max(max_relation, file_max_relation) + logger.debug( + f"Generated {len(triplets)} triplets for {input_file} with {opt_level}" + ) return TripletResult(all_triplets, max_relation) @@ -124,7 +144,7 @@ class IR2VecTripletGenerator: # Run llvm-ir2vec with opt's output as input ir2vec_proc = subprocess.Popen( - [self.ir2vec_binary, "triplets", "-", "-o", "-"], + [self.ir2vec_binary, "triplets", "--mode=llvm", "-", "-o", "-"], stdin=opt_proc.stdout, stdout=subprocess.PIPE, stderr=subprocess.PIPE, @@ -143,6 +163,32 @@ class IR2VecTripletGenerator: except (subprocess.SubprocessError, OSError): return set(), 1 + def _run_mir_pipeline(self, input_file: Path) -> Tuple[Set[str], int]: + """Run llvm-ir2vec pipeline for MIR files.""" + try: + # Run llvm-ir2vec directly on MIR file + result = subprocess.run( + [ + self.ir2vec_binary, + "triplets", + "--mode=mir", + str(input_file), + "-o", + "-", + ], + stdout=subprocess.PIPE, + stderr=subprocess.PIPE, + text=True, + check=False, + ) + + if result.returncode != 0: + return set(), 1 + + return self._parse_triplet_output(result.stdout) + except (subprocess.SubprocessError, OSError): + return set(), 1 + def _parse_triplet_output(self, output: str) -> Tuple[Set[str], int]: """Parse triplet output and extract max relation""" if not output.strip(): @@ -160,12 +206,21 @@ class IR2VecTripletGenerator: return set(lines), max_relation def generate_triplets(self, file_list: Path) -> None: - """Main method to generate triplets from a list of LLVM IR files""" + """Main method to generate triplets from a list of LLVM IR or MIR files""" + # Store file_list_path for later use in entity generation + self.file_list_path = file_list + input_files = self._read_file_list(file_list) - logger.info( - f"Processing {len(input_files)} files with {self.num_optimizations} " - f"optimization levels using {self.max_workers} workers" - ) + + if self.mode == "mir": + logger.info( + f"Processing {len(input_files)} MIR files using {self.max_workers} workers" + ) + else: + logger.info( + f"Processing {len(input_files)} files with {self.num_optimizations} " + f"optimization levels using {self.max_workers} workers" + ) all_triplets = set() global_max_relation = 1 @@ -222,28 +277,60 @@ class IR2VecTripletGenerator: def _generate_entity2id(self, output_file: Path) -> None: """Generate entity2id.txt using llvm-ir2vec""" - subprocess.run( - [str(self.ir2vec_binary), "entities", "-o", str(output_file)], - check=True, - capture_output=True, - ) + if self.mode == "mir": + # For MIR mode, we need to provide a sample MIR file to determine target + # Use the first file from the processed list + input_files = self._read_file_list(self.file_list_path) + if not input_files: + raise ValueError("No input files available for entity generation") + + subprocess.run( + [ + str(self.ir2vec_binary), + "entities", + "--mode=mir", + str(input_files[0]), + "-o", + str(output_file), + ], + check=True, + capture_output=True, + ) + else: + subprocess.run( + [ + str(self.ir2vec_binary), + "entities", + "--mode=llvm", + "-o", + str(output_file), + ], + check=True, + capture_output=True, + ) def _generate_relation2id(self, output_file: Path, max_relation: int) -> None: """Generate relation2id.txt from max relation""" - max_relation = max(max_relation, 1) # At least Type and Next relations + max_relation = max(max_relation, 1) # At least Next relation num_relations = max_relation + 1 with open(output_file, "w") as f: f.write(f"{num_relations}\n") - f.write("Type\t0\n") - f.write("Next\t1\n") - f.writelines(f"Arg{i-2}\t{i}\n" for i in range(2, num_relations)) + if self.mode == "llvm": + # LLVM IR has Type relation at 0 + f.write("Type\t0\n") + f.write("Next\t1\n") + f.writelines(f"Arg{i-2}\t{i}\n" for i in range(2, num_relations)) + else: + # MIR doesn't have Type relation, starts with Next at 0 + f.write("Next\t0\n") + f.writelines(f"Arg{i-1}\t{i}\n" for i in range(1, num_relations)) def main(): """Main entry point""" parser = argparse.ArgumentParser( - description="Generate IR2Vec triplets from LLVM IR files", + description="Generate IR2Vec or MIR2Vec triplets from LLVM IR or Machine IR files", formatter_class=argparse.RawDescriptionHelpFormatter, ) @@ -253,17 +340,26 @@ def main(): parser.add_argument( "num_optimizations", type=int, - help="Number of optimization levels to apply (1-6)", + nargs="?", + default=1, + help="Number of optimization levels to apply (1-6) for LLVM IR mode", ) parser.add_argument( - "ll_file_list", + "input_file_list", type=Path, - help="File containing list of LLVM IR files to process", + help="File containing list of LLVM IR or MIR files to process", ) parser.add_argument( "output_dir", type=Path, help="Output directory for generated files" ) parser.add_argument( + "--mode", + type=str, + choices=["llvm", "mir"], + default="llvm", + help="Operation mode: 'llvm' for LLVM IR (default) or 'mir' for Machine IR", + ) + parser.add_argument( "-j", "--max-workers", type=int, @@ -296,8 +392,9 @@ def main(): args.num_optimizations, args.output_dir, args.max_workers, + args.mode, ) - generator.generate_triplets(args.ll_file_list) + generator.generate_triplets(args.input_file_list) if __name__ == "__main__": diff --git a/llvm/utils/update_mc_test_checks.py b/llvm/utils/update_mc_test_checks.py index ab7fe19..67fff56 100755 --- a/llvm/utils/update_mc_test_checks.py +++ b/llvm/utils/update_mc_test_checks.py @@ -290,11 +290,9 @@ def update_test(ti: common.TestInfo): # prefix is selected and generated with most shared output lines # each run_id can only be used once - gen_prefix = "" used_runid = set() - # line number diff between generated prefix and testline - line_offset = 1 + selected_prefixes = set() for prefix, tup in p_dict_sorted.items(): o, run_ids = tup @@ -308,18 +306,24 @@ def update_test(ti: common.TestInfo): else: used_runid.add(i) if not skip: - used_prefixes.add(prefix) + selected_prefixes.add(prefix) - if hasErr(o): - newline = getErrCheckLine(prefix, o, mc_mode, line_offset) - else: - newline = getStdCheckLine(prefix, o, mc_mode) + # Generate check lines in alphabetical order. + check_lines = [] + for prefix in sorted(selected_prefixes): + o, run_ids = p_dict[prefix] + used_prefixes.add(prefix) + + if hasErr(o): + line_offset = len(check_lines) + 1 + check = getErrCheckLine(prefix, o, mc_mode, line_offset) + else: + check = getStdCheckLine(prefix, o, mc_mode) - if newline: - gen_prefix += newline - line_offset += 1 + if check: + check_lines.append(check.strip()) - generated_prefixes[input_line] = gen_prefix.rstrip("\n") + generated_prefixes[input_line] = "\n".join(check_lines) # write output for input_info in ti.iterlines(output_lines): |