Author: Yusuke MINATO Date: 2024-10-28T14:19:20Z New Revision: bd6ab32e6eb642f2b0b15be8c7c2a668192f07d8
URL: https://github.com/llvm/llvm-project/commit/bd6ab32e6eb642f2b0b15be8c7c2a668192f07d8 DIFF: https://github.com/llvm/llvm-project/commit/bd6ab32e6eb642f2b0b15be8c7c2a668192f07d8.diff LOG: Revert "[flang] Integrate the option -flang-experimental-integer-overflow into -fno-wrapv" (#113901) Reverts llvm/llvm-project#110063 due to the performance regression on 503.bwaves_r in SPEC2017. Added: Modified: clang/include/clang/Driver/Options.td clang/lib/Driver/ToolChains/Flang.cpp flang/include/flang/Lower/LoweringOptions.def flang/include/flang/Optimizer/Transforms/Passes.h flang/include/flang/Optimizer/Transforms/Passes.td flang/include/flang/Tools/CrossToolHelpers.h flang/lib/Frontend/CompilerInvocation.cpp flang/lib/Frontend/FrontendActions.cpp flang/lib/Lower/Bridge.cpp flang/lib/Lower/IO.cpp flang/lib/Optimizer/Passes/Pipelines.cpp flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp flang/test/Driver/frontend-forwarding.f90 flang/test/Fir/convert-to-llvm-openmp-and-fir.fir flang/test/Fir/loop01.fir flang/test/Fir/loop02.fir flang/test/Lower/HLFIR/goto-do-body.f90 flang/test/Lower/OpenMP/parallel-private-clause-fixes.f90 flang/test/Lower/OpenMP/wsloop-reduction-allocatable-array-minmax.f90 flang/test/Lower/OpenMP/wsloop-variable.f90 flang/test/Lower/array-character.f90 flang/test/Lower/array-derived-assignments.f90 flang/test/Lower/array-derived.f90 flang/test/Lower/array-elemental-calls-char-byval.f90 flang/test/Lower/array-elemental-calls-char.f90 flang/test/Lower/array-expression-assumed-size.f90 flang/test/Lower/array-expression-slice-1.f90 flang/test/Lower/array-substring.f90 flang/test/Lower/array-temp.f90 flang/test/Lower/components.f90 flang/test/Lower/do_loop.f90 flang/test/Lower/do_loop_unstructured.f90 flang/test/Lower/goto-do-body.f90 flang/test/Lower/host-associated.f90 flang/test/Lower/infinite_loop.f90 flang/test/Lower/io-implied-do-fixes.f90 flang/test/Lower/loops2.f90 flang/test/Lower/mixed_loops.f90 flang/test/Lower/vector-subscript-io.f90 flang/tools/bbc/bbc.cpp Removed: ################################################################################ diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td index 7b28e8b4c31ec1..f86e90e14c477b 100644 --- a/clang/include/clang/Driver/Options.td +++ b/clang/include/clang/Driver/Options.td @@ -6791,6 +6791,10 @@ def flang_deprecated_no_hlfir : Flag<["-"], "flang-deprecated-no-hlfir">, Flags<[HelpHidden]>, Visibility<[FlangOption, FC1Option]>, HelpText<"Do not use HLFIR lowering (deprecated)">; +def flang_experimental_integer_overflow : Flag<["-"], "flang-experimental-integer-overflow">, + Flags<[HelpHidden]>, Visibility<[FlangOption, FC1Option]>, + HelpText<"Add nsw flag to internal operations such as do-variable increment (experimental)">; + //===----------------------------------------------------------------------===// // FLangOption + CoreOption + NoXarchOption //===----------------------------------------------------------------------===// diff --git a/clang/lib/Driver/ToolChains/Flang.cpp b/clang/lib/Driver/ToolChains/Flang.cpp index 366cadc2e54775..a9d2b7a4dc48f9 100644 --- a/clang/lib/Driver/ToolChains/Flang.cpp +++ b/clang/lib/Driver/ToolChains/Flang.cpp @@ -148,6 +148,7 @@ void Flang::addCodegenOptions(const ArgList &Args, Args.addAllArgs(CmdArgs, {options::OPT_flang_experimental_hlfir, options::OPT_flang_deprecated_no_hlfir, + options::OPT_flang_experimental_integer_overflow, options::OPT_fno_ppc_native_vec_elem_order, options::OPT_fppc_native_vec_elem_order}); } diff --git a/flang/include/flang/Lower/LoweringOptions.def b/flang/include/flang/Lower/LoweringOptions.def index 0b22e54b648e94..231de533fbd30a 100644 --- a/flang/include/flang/Lower/LoweringOptions.def +++ b/flang/include/flang/Lower/LoweringOptions.def @@ -38,5 +38,10 @@ ENUM_LOWERINGOPT(Underscoring, unsigned, 1, 1) /// (i.e. wraps around as two's complement). Off by default. ENUM_LOWERINGOPT(IntegerWrapAround, unsigned, 1, 0) +/// If true, add nsw flags to loop variable increments. +/// Off by default. +/// TODO: integrate this option with the above +ENUM_LOWERINGOPT(NSWOnLoopVarInc, unsigned, 1, 0) + #undef LOWERINGOPT #undef ENUM_LOWERINGOPT diff --git a/flang/include/flang/Optimizer/Transforms/Passes.h b/flang/include/flang/Optimizer/Transforms/Passes.h index e1453cfa374bfc..e8f0a8444a31a1 100644 --- a/flang/include/flang/Optimizer/Transforms/Passes.h +++ b/flang/include/flang/Optimizer/Transforms/Passes.h @@ -66,7 +66,7 @@ namespace fir { std::unique_ptr<mlir::Pass> createAffineDemotionPass(); std::unique_ptr<mlir::Pass> createArrayValueCopyPass(fir::ArrayValueCopyOptions options = {}); -std::unique_ptr<mlir::Pass> createCFGConversionPassWithoutNSW(); +std::unique_ptr<mlir::Pass> createCFGConversionPassWithNSW(); std::unique_ptr<mlir::Pass> createMemDataFlowOptPass(); std::unique_ptr<mlir::Pass> createPromoteToAffinePass(); std::unique_ptr<mlir::Pass> @@ -83,7 +83,7 @@ createVScaleAttrPass(std::pair<unsigned, unsigned> vscaleAttr); void populateCfgConversionRewrites(mlir::RewritePatternSet &patterns, bool forceLoopToExecuteOnce = false, - bool setNSW = true); + bool setNSW = false); // declarative passes #define GEN_PASS_REGISTRATION diff --git a/flang/include/flang/Optimizer/Transforms/Passes.td b/flang/include/flang/Optimizer/Transforms/Passes.td index 54b43adbfc6bbf..a41f0f348f27a6 100644 --- a/flang/include/flang/Optimizer/Transforms/Passes.td +++ b/flang/include/flang/Optimizer/Transforms/Passes.td @@ -153,7 +153,7 @@ def CFGConversion : Pass<"cfg-conversion"> { /*default=*/"false", "force the body of a loop to execute at least once">, Option<"setNSW", "set-nsw", "bool", - /*default=*/"true", + /*default=*/"false", "set nsw on loop variable increment"> ]; } diff --git a/flang/include/flang/Tools/CrossToolHelpers.h b/flang/include/flang/Tools/CrossToolHelpers.h index 1626970600eff2..df4b21ada058fe 100644 --- a/flang/include/flang/Tools/CrossToolHelpers.h +++ b/flang/include/flang/Tools/CrossToolHelpers.h @@ -122,7 +122,7 @@ struct MLIRToLLVMPassPipelineConfig : public FlangEPCallBacks { bool NoSignedZerosFPMath = false; ///< Set no-signed-zeros-fp-math attribute for functions. bool UnsafeFPMath = false; ///< Set unsafe-fp-math attribute for functions. - bool NSWOnLoopVarInc = true; ///< Add nsw flag to loop variable increments. + bool NSWOnLoopVarInc = false; ///< Add nsw flag to loop variable increments. }; struct OffloadModuleOpts { diff --git a/flang/lib/Frontend/CompilerInvocation.cpp b/flang/lib/Frontend/CompilerInvocation.cpp index 5da5236af2b0e1..94d3d115417877 100644 --- a/flang/lib/Frontend/CompilerInvocation.cpp +++ b/flang/lib/Frontend/CompilerInvocation.cpp @@ -1350,6 +1350,12 @@ bool CompilerInvocation::createFromArgs( invoc.loweringOpts.setNoPPCNativeVecElemOrder(true); } + // -flang-experimental-integer-overflow + if (args.hasArg( + clang::driver::options::OPT_flang_experimental_integer_overflow)) { + invoc.loweringOpts.setNSWOnLoopVarInc(true); + } + // Preserve all the remark options requested, i.e. -Rpass, -Rpass-missed or // -Rpass-analysis. This will be used later when processing and outputting the // remarks generated by LLVM in ExecuteCompilerInvocation.cpp. diff --git a/flang/lib/Frontend/FrontendActions.cpp b/flang/lib/Frontend/FrontendActions.cpp index 3ea242315484fd..f2e460fc53a67f 100644 --- a/flang/lib/Frontend/FrontendActions.cpp +++ b/flang/lib/Frontend/FrontendActions.cpp @@ -828,8 +828,8 @@ void CodeGenAction::generateLLVMIR() { config.VScaleMax = vsr->second; } - if (ci.getInvocation().getLoweringOpts().getIntegerWrapAround()) - config.NSWOnLoopVarInc = false; + if (ci.getInvocation().getLoweringOpts().getNSWOnLoopVarInc()) + config.NSWOnLoopVarInc = true; // Create the pass pipeline fir::createMLIRToLLVMPassPipeline(pm, config, getCurrentFile()); diff --git a/flang/lib/Lower/Bridge.cpp b/flang/lib/Lower/Bridge.cpp index a3bd1ace11da21..877fe122265dd0 100644 --- a/flang/lib/Lower/Bridge.cpp +++ b/flang/lib/Lower/Bridge.cpp @@ -2271,7 +2271,7 @@ class FirConverter : public Fortran::lower::AbstractConverter { assert(!incrementLoopNestInfo.empty() && "empty loop nest"); mlir::Location loc = toLocation(); mlir::arith::IntegerOverflowFlags flags{}; - if (!getLoweringOptions().getIntegerWrapAround()) + if (getLoweringOptions().getNSWOnLoopVarInc()) flags = bitEnumSet(flags, mlir::arith::IntegerOverflowFlags::nsw); auto iofAttr = mlir::arith::IntegerOverflowFlagsAttr::get( builder->getContext(), flags); diff --git a/flang/lib/Lower/IO.cpp b/flang/lib/Lower/IO.cpp index b534c81a605a90..1894b0cfd1bec2 100644 --- a/flang/lib/Lower/IO.cpp +++ b/flang/lib/Lower/IO.cpp @@ -929,7 +929,7 @@ static void genIoLoop(Fortran::lower::AbstractConverter &converter, fir::FirOpBuilder &builder = converter.getFirOpBuilder(); mlir::Location loc = converter.getCurrentLocation(); mlir::arith::IntegerOverflowFlags flags{}; - if (!converter.getLoweringOptions().getIntegerWrapAround()) + if (converter.getLoweringOptions().getNSWOnLoopVarInc()) flags = bitEnumSet(flags, mlir::arith::IntegerOverflowFlags::nsw); auto iofAttr = mlir::arith::IntegerOverflowFlagsAttr::get(builder.getContext(), flags); diff --git a/flang/lib/Optimizer/Passes/Pipelines.cpp b/flang/lib/Optimizer/Passes/Pipelines.cpp index 091b7c4c164e40..3c139f7e93405c 100644 --- a/flang/lib/Optimizer/Passes/Pipelines.cpp +++ b/flang/lib/Optimizer/Passes/Pipelines.cpp @@ -35,11 +35,11 @@ void addCanonicalizerPassWithoutRegionSimplification(mlir::OpPassManager &pm) { void addCfgConversionPass(mlir::PassManager &pm, const MLIRToLLVMPassPipelineConfig &config) { if (config.NSWOnLoopVarInc) + addNestedPassToAllTopLevelOperationsConditionally( + pm, disableCfgConversion, fir::createCFGConversionPassWithNSW); + else addNestedPassToAllTopLevelOperationsConditionally(pm, disableCfgConversion, fir::createCFGConversion); - else - addNestedPassToAllTopLevelOperationsConditionally( - pm, disableCfgConversion, fir::createCFGConversionPassWithoutNSW); } void addAVC(mlir::PassManager &pm, const llvm::OptimizationLevel &optLevel) { diff --git a/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp b/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp index 411bf7f364a602..3b79d6d311b71c 100644 --- a/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp +++ b/flang/lib/Optimizer/Transforms/ControlFlowConverter.cpp @@ -332,6 +332,8 @@ class CfgConversion : public fir::impl::CFGConversionBase<CfgConversion> { public: using CFGConversionBase<CfgConversion>::CFGConversionBase; + CfgConversion(bool setNSW) { this->setNSW = setNSW; } + void runOnOperation() override { auto *context = &this->getContext(); mlir::RewritePatternSet patterns(context); @@ -364,8 +366,6 @@ void fir::populateCfgConversionRewrites(mlir::RewritePatternSet &patterns, patterns.getContext(), forceLoopToExecuteOnce, setNSW); } -std::unique_ptr<mlir::Pass> fir::createCFGConversionPassWithoutNSW() { - fir::CFGConversionOptions options; - options.setNSW = false; - return fir::createCFGConversion(options); +std::unique_ptr<mlir::Pass> fir::createCFGConversionPassWithNSW() { + return std::make_unique<CfgConversion>(true); } diff --git a/flang/test/Driver/frontend-forwarding.f90 b/flang/test/Driver/frontend-forwarding.f90 index 55a74ccf40467b..ff2d6609521464 100644 --- a/flang/test/Driver/frontend-forwarding.f90 +++ b/flang/test/Driver/frontend-forwarding.f90 @@ -20,6 +20,7 @@ ! RUN: -fversion-loops-for-stride \ ! RUN: -flang-experimental-hlfir \ ! RUN: -flang-deprecated-no-hlfir \ +! RUN: -flang-experimental-integer-overflow \ ! RUN: -fno-ppc-native-vector-element-order \ ! RUN: -fppc-native-vector-element-order \ ! RUN: -mllvm -print-before-all \ @@ -51,6 +52,7 @@ ! CHECK: "-fversion-loops-for-stride" ! CHECK: "-flang-experimental-hlfir" ! CHECK: "-flang-deprecated-no-hlfir" +! CHECK: "-flang-experimental-integer-overflow" ! CHECK: "-fno-ppc-native-vector-element-order" ! CHECK: "-fppc-native-vector-element-order" ! CHECK: "-Rpass" diff --git a/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir b/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir index 00f8e6e6cc9a6b..335877e7c9a872 100644 --- a/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir +++ b/flang/test/Fir/convert-to-llvm-openmp-and-fir.fir @@ -358,10 +358,10 @@ func.func @_QPopenmp_target_data_region() { %9 = arith.subi %8, %c1_i64 : i64 %10 = fir.coordinate_of %0, %9 : (!fir.ref<!fir.array<1024xi32>>, i64) -> !fir.ref<i32> fir.store %6 to %10 : !fir.ref<i32> - %11 = arith.addi %arg0, %c1 overflow<nsw> : index + %11 = arith.addi %arg0, %c1 : index %12 = fir.convert %c1 : (index) -> i32 %13 = fir.load %1 : !fir.ref<i32> - %14 = arith.addi %13, %12 overflow<nsw> : i32 + %14 = arith.addi %13, %12 : i32 fir.result %11, %14 : index, i32 } fir.store %5#1 to %1 : !fir.ref<i32> @@ -404,11 +404,11 @@ func.func @_QPopenmp_target_data_region() { // CHECK: %[[VAL_21:.*]] = llvm.sub %[[VAL_19]], %[[VAL_20]] : i64 // CHECK: %[[VAL_22:.*]] = llvm.getelementptr %[[VAL_1]][0, %[[VAL_21]]] : (!llvm.ptr, i64) -> !llvm.ptr // CHECK: llvm.store %[[VAL_17]], %[[VAL_22]] : i32, !llvm.ptr -// CHECK: %[[VAL_23:.*]] = llvm.add %[[VAL_12]], %[[VAL_8]] overflow<nsw> : i64 +// CHECK: %[[VAL_23:.*]] = llvm.add %[[VAL_12]], %[[VAL_8]] : i64 // CHECK: %[[VAL_24:.*]] = llvm.trunc %[[VAL_8]] : i64 to i32 // CHECK: %[[VAL_25:.*]] = llvm.load %[[VAL_3]] : !llvm.ptr -> i32 -// CHECK: %[[VAL_26:.*]] = llvm.add %[[VAL_25]], %[[VAL_24]] overflow<nsw> : i32 -// CHECK: %[[VAL_27:.*]] = llvm.add %[[VAL_12]], %[[VAL_8]] overflow<nsw> : i64 +// CHECK: %[[VAL_26:.*]] = llvm.add %[[VAL_25]], %[[VAL_24]] : i32 +// CHECK: %[[VAL_27:.*]] = llvm.add %[[VAL_12]], %[[VAL_8]] : i64 // CHECK: %[[VAL_28:.*]] = llvm.mlir.constant(1 : index) : i64 // CHECK: %[[VAL_29:.*]] = llvm.sub %[[VAL_14]], %[[VAL_28]] : i64 // CHECK: llvm.br ^bb1(%[[VAL_27]], %[[VAL_26]], %[[VAL_29]] : i64, i32, i64) diff --git a/flang/test/Fir/loop01.fir b/flang/test/Fir/loop01.fir index 30d10b9bbdb979..c1cbb522c378c0 100644 --- a/flang/test/Fir/loop01.fir +++ b/flang/test/Fir/loop01.fir @@ -1,7 +1,5 @@ // RUN: fir-opt --split-input-file --cfg-conversion %s | FileCheck %s -// RUN: fir-opt --split-input-file --cfg-conversion="set-nsw=false" %s | FileCheck %s --check-prefix=NO-NSW - -// NO-NSW-NOT: overflow<nsw> +// RUN: fir-opt --split-input-file --cfg-conversion="set-nsw=true" %s | FileCheck %s --check-prefix=NSW func.func @x(%lb : index, %ub : index, %step : index, %b : i1, %addr : !fir.ref<index>) { fir.do_loop %iv = %lb to %ub step %step unordered { @@ -37,7 +35,7 @@ func.func private @f2() -> i1 // CHECK: fir.store %[[VAL_12]] to %[[VAL_4]] : !fir.ref<index> // CHECK: br ^bb5 // CHECK: ^bb5: -// CHECK: %[[VAL_13:.*]] = arith.addi %[[VAL_8]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_13:.*]] = arith.addi %[[VAL_8]], %[[VAL_2]] : index // CHECK: %[[VAL_14:.*]] = arith.constant 1 : index // CHECK: %[[VAL_15:.*]] = arith.subi %[[VAL_9]], %[[VAL_14]] : index // CHECK: br ^bb1(%[[VAL_13]], %[[VAL_15]] : index, index) @@ -46,6 +44,34 @@ func.func private @f2() -> i1 // CHECK: } // CHECK: func private @f2() -> i1 +// NSW: func @x(%[[VAL_0:.*]]: index, %[[VAL_1:.*]]: index, %[[VAL_2:.*]]: index, %[[VAL_3:.*]]: i1, %[[VAL_4:.*]]: !fir.ref<index>) { +// NSW: %[[VAL_5:.*]] = arith.subi %[[VAL_1]], %[[VAL_0]] : index +// NSW: %[[VAL_6:.*]] = arith.addi %[[VAL_5]], %[[VAL_2]] : index +// NSW: %[[VAL_7:.*]] = arith.divsi %[[VAL_6]], %[[VAL_2]] : index +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_7]] : index, index) +// NSW: ^bb1(%[[VAL_8:.*]]: index, %[[VAL_9:.*]]: index): +// NSW: %[[VAL_10:.*]] = arith.constant 0 : index +// NSW: %[[VAL_11:.*]] = arith.cmpi sgt, %[[VAL_9]], %[[VAL_10]] : index +// NSW: cond_br %[[VAL_11]], ^bb2, ^bb6 +// NSW: ^bb2: +// NSW: cond_br %[[VAL_3]], ^bb3, ^bb4 +// NSW: ^bb3: +// NSW: fir.store %[[VAL_8]] to %[[VAL_4]] : !fir.ref<index> +// NSW: br ^bb5 +// NSW: ^bb4: +// NSW: %[[VAL_12:.*]] = arith.constant 0 : index +// NSW: fir.store %[[VAL_12]] to %[[VAL_4]] : !fir.ref<index> +// NSW: br ^bb5 +// NSW: ^bb5: +// NSW: %[[VAL_13:.*]] = arith.addi %[[VAL_8]], %[[VAL_2]] overflow<nsw> : index +// NSW: %[[VAL_14:.*]] = arith.constant 1 : index +// NSW: %[[VAL_15:.*]] = arith.subi %[[VAL_9]], %[[VAL_14]] : index +// NSW: br ^bb1(%[[VAL_13]], %[[VAL_15]] : index, index) +// NSW: ^bb6: +// NSW: return +// NSW: } +// NSW: func private @f2() -> i1 + // ----- func.func @x2(%lo : index, %up : index, %ok : i1) { @@ -75,13 +101,36 @@ func.func private @f3(i16) // CHECK: cond_br %[[VAL_14]], ^bb2, ^bb3 // CHECK: ^bb2: // CHECK: %[[VAL_15:.*]] = fir.call @f2() : () -> i1 -// CHECK: %[[VAL_16:.*]] = arith.addi %[[VAL_4]], %[[VAL_3]] overflow<nsw> : index +// CHECK: %[[VAL_16:.*]] = arith.addi %[[VAL_4]], %[[VAL_3]] : index // CHECK: br ^bb1(%[[VAL_16]], %[[VAL_15]] : index, i1) // CHECK: ^bb3: // CHECK: return // CHECK: } // CHECK: func private @f3(i16) +// NSW: func @x2(%[[VAL_0:.*]]: index, %[[VAL_1:.*]]: index, %[[VAL_2:.*]]: i1) { +// NSW: %[[VAL_3:.*]] = arith.constant 1 : index +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_2]] : index, i1) +// NSW: ^bb1(%[[VAL_4:.*]]: index, %[[VAL_5:.*]]: i1): +// NSW: %[[VAL_6:.*]] = arith.constant 0 : index +// NSW: %[[VAL_7:.*]] = arith.cmpi slt, %[[VAL_6]], %[[VAL_3]] : index +// NSW: %[[VAL_8:.*]] = arith.cmpi sle, %[[VAL_4]], %[[VAL_1]] : index +// NSW: %[[VAL_9:.*]] = arith.cmpi slt, %[[VAL_3]], %[[VAL_6]] : index +// NSW: %[[VAL_10:.*]] = arith.cmpi sle, %[[VAL_1]], %[[VAL_4]] : index +// NSW: %[[VAL_11:.*]] = arith.andi %[[VAL_7]], %[[VAL_8]] : i1 +// NSW: %[[VAL_12:.*]] = arith.andi %[[VAL_9]], %[[VAL_10]] : i1 +// NSW: %[[VAL_13:.*]] = arith.ori %[[VAL_11]], %[[VAL_12]] : i1 +// NSW: %[[VAL_14:.*]] = arith.andi %[[VAL_5]], %[[VAL_13]] : i1 +// NSW: cond_br %[[VAL_14]], ^bb2, ^bb3 +// NSW: ^bb2: +// NSW: %[[VAL_15:.*]] = fir.call @f2() : () -> i1 +// NSW: %[[VAL_16:.*]] = arith.addi %[[VAL_4]], %[[VAL_3]] overflow<nsw> : index +// NSW: br ^bb1(%[[VAL_16]], %[[VAL_15]] : index, i1) +// NSW: ^bb3: +// NSW: return +// NSW: } +// NSW: func private @f3(i16) + // ----- // do_loop with an extra loop-carried value @@ -110,7 +159,7 @@ func.func @x3(%lo : index, %up : index) -> i1 { // CHECK: cond_br %[[VAL_11]], ^bb2, ^bb3 // CHECK: ^bb2: // CHECK: %[[VAL_12:.*]] = fir.call @f2() : () -> i1 -// CHECK: %[[VAL_13:.*]] = arith.addi %[[VAL_7]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_13:.*]] = arith.addi %[[VAL_7]], %[[VAL_2]] : index // CHECK: %[[VAL_14:.*]] = arith.constant 1 : index // CHECK: %[[VAL_15:.*]] = arith.subi %[[VAL_9]], %[[VAL_14]] : index // CHECK: br ^bb1(%[[VAL_13]], %[[VAL_12]], %[[VAL_15]] : index, i1, index) @@ -118,6 +167,29 @@ func.func @x3(%lo : index, %up : index) -> i1 { // CHECK: return %[[VAL_8]] : i1 // CHECK: } +// NSW-LABEL: func @x3( +// NSW-SAME: %[[VAL_0:.*]]: index, +// NSW-SAME: %[[VAL_1:.*]]: index) -> i1 { +// NSW: %[[VAL_2:.*]] = arith.constant 1 : index +// NSW: %[[VAL_3:.*]] = arith.constant true +// NSW: %[[VAL_4:.*]] = arith.subi %[[VAL_1]], %[[VAL_0]] : index +// NSW: %[[VAL_5:.*]] = arith.addi %[[VAL_4]], %[[VAL_2]] : index +// NSW: %[[VAL_6:.*]] = arith.divsi %[[VAL_5]], %[[VAL_2]] : index +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_3]], %[[VAL_6]] : index, i1, index) +// NSW: ^bb1(%[[VAL_7:.*]]: index, %[[VAL_8:.*]]: i1, %[[VAL_9:.*]]: index): +// NSW: %[[VAL_10:.*]] = arith.constant 0 : index +// NSW: %[[VAL_11:.*]] = arith.cmpi sgt, %[[VAL_9]], %[[VAL_10]] : index +// NSW: cond_br %[[VAL_11]], ^bb2, ^bb3 +// NSW: ^bb2: +// NSW: %[[VAL_12:.*]] = fir.call @f2() : () -> i1 +// NSW: %[[VAL_13:.*]] = arith.addi %[[VAL_7]], %[[VAL_2]] overflow<nsw> : index +// NSW: %[[VAL_14:.*]] = arith.constant 1 : index +// NSW: %[[VAL_15:.*]] = arith.subi %[[VAL_9]], %[[VAL_14]] : index +// NSW: br ^bb1(%[[VAL_13]], %[[VAL_12]], %[[VAL_15]] : index, i1, index) +// NSW: ^bb3: +// NSW: return %[[VAL_8]] : i1 +// NSW: } + // ----- // iterate_while with an extra loop-carried value @@ -155,7 +227,7 @@ func.func private @f4(i32) -> i1 // CHECK: cond_br %[[VAL_16]], ^bb2, ^bb3 // CHECK: ^bb2: // CHECK: %[[VAL_17:.*]] = fir.call @f2() : () -> i1 -// CHECK: %[[VAL_18:.*]] = arith.addi %[[VAL_5]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_18:.*]] = arith.addi %[[VAL_5]], %[[VAL_2]] : index // CHECK: br ^bb1(%[[VAL_18]], %[[VAL_6]], %[[VAL_17]] : index, i1, i1) // CHECK: ^bb3: // CHECK: %[[VAL_19:.*]] = arith.andi %[[VAL_6]], %[[VAL_7]] : i1 @@ -163,6 +235,34 @@ func.func private @f4(i32) -> i1 // CHECK: } // CHECK: func private @f4(i32) -> i1 +// NSW-LABEL: func @y3( +// NSW-SAME: %[[VAL_0:.*]]: index, +// NSW-SAME: %[[VAL_1:.*]]: index) -> i1 { +// NSW: %[[VAL_2:.*]] = arith.constant 1 : index +// NSW: %[[VAL_3:.*]] = arith.constant true +// NSW: %[[VAL_4:.*]] = fir.call @f2() : () -> i1 +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_3]], %[[VAL_4]] : index, i1, i1) +// NSW: ^bb1(%[[VAL_5:.*]]: index, %[[VAL_6:.*]]: i1, %[[VAL_7:.*]]: i1): +// NSW: %[[VAL_8:.*]] = arith.constant 0 : index +// NSW: %[[VAL_9:.*]] = arith.cmpi slt, %[[VAL_8]], %[[VAL_2]] : index +// NSW: %[[VAL_10:.*]] = arith.cmpi sle, %[[VAL_5]], %[[VAL_1]] : index +// NSW: %[[VAL_11:.*]] = arith.cmpi slt, %[[VAL_2]], %[[VAL_8]] : index +// NSW: %[[VAL_12:.*]] = arith.cmpi sle, %[[VAL_1]], %[[VAL_5]] : index +// NSW: %[[VAL_13:.*]] = arith.andi %[[VAL_9]], %[[VAL_10]] : i1 +// NSW: %[[VAL_14:.*]] = arith.andi %[[VAL_11]], %[[VAL_12]] : i1 +// NSW: %[[VAL_15:.*]] = arith.ori %[[VAL_13]], %[[VAL_14]] : i1 +// NSW: %[[VAL_16:.*]] = arith.andi %[[VAL_6]], %[[VAL_15]] : i1 +// NSW: cond_br %[[VAL_16]], ^bb2, ^bb3 +// NSW: ^bb2: +// NSW: %[[VAL_17:.*]] = fir.call @f2() : () -> i1 +// NSW: %[[VAL_18:.*]] = arith.addi %[[VAL_5]], %[[VAL_2]] overflow<nsw> : index +// NSW: br ^bb1(%[[VAL_18]], %[[VAL_6]], %[[VAL_17]] : index, i1, i1) +// NSW: ^bb3: +// NSW: %[[VAL_19:.*]] = arith.andi %[[VAL_6]], %[[VAL_7]] : i1 +// NSW: return %[[VAL_19]] : i1 +// NSW: } +// NSW: func private @f4(i32) -> i1 + // ----- // do_loop that returns the final value of the induction @@ -191,7 +291,7 @@ func.func @x4(%lo : index, %up : index) -> index { // CHECK: ^bb2: // CHECK: %[[VAL_10:.*]] = fir.convert %[[VAL_6]] : (index) -> i32 // CHECK: %[[VAL_11:.*]] = fir.call @f4(%[[VAL_10]]) : (i32) -> i1 -// CHECK: %[[VAL_12:.*]] = arith.addi %[[VAL_6]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_12:.*]] = arith.addi %[[VAL_6]], %[[VAL_2]] : index // CHECK: %[[VAL_13:.*]] = arith.constant 1 : index // CHECK: %[[VAL_14:.*]] = arith.subi %[[VAL_7]], %[[VAL_13]] : index // CHECK: br ^bb1(%[[VAL_12]], %[[VAL_14]] : index, index) @@ -199,6 +299,29 @@ func.func @x4(%lo : index, %up : index) -> index { // CHECK: return %[[VAL_6]] : index // CHECK: } +// NSW-LABEL: func @x4( +// NSW-SAME: %[[VAL_0:.*]]: index, +// NSW-SAME: %[[VAL_1:.*]]: index) -> index { +// NSW: %[[VAL_2:.*]] = arith.constant 1 : index +// NSW: %[[VAL_3:.*]] = arith.subi %[[VAL_1]], %[[VAL_0]] : index +// NSW: %[[VAL_4:.*]] = arith.addi %[[VAL_3]], %[[VAL_2]] : index +// NSW: %[[VAL_5:.*]] = arith.divsi %[[VAL_4]], %[[VAL_2]] : index +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_5]] : index, index) +// NSW: ^bb1(%[[VAL_6:.*]]: index, %[[VAL_7:.*]]: index): +// NSW: %[[VAL_8:.*]] = arith.constant 0 : index +// NSW: %[[VAL_9:.*]] = arith.cmpi sgt, %[[VAL_7]], %[[VAL_8]] : index +// NSW: cond_br %[[VAL_9]], ^bb2, ^bb3 +// NSW: ^bb2: +// NSW: %[[VAL_10:.*]] = fir.convert %[[VAL_6]] : (index) -> i32 +// NSW: %[[VAL_11:.*]] = fir.call @f4(%[[VAL_10]]) : (i32) -> i1 +// NSW: %[[VAL_12:.*]] = arith.addi %[[VAL_6]], %[[VAL_2]] overflow<nsw> : index +// NSW: %[[VAL_13:.*]] = arith.constant 1 : index +// NSW: %[[VAL_14:.*]] = arith.subi %[[VAL_7]], %[[VAL_13]] : index +// NSW: br ^bb1(%[[VAL_12]], %[[VAL_14]] : index, index) +// NSW: ^bb3: +// NSW: return %[[VAL_6]] : index +// NSW: } + // ----- // iterate_while that returns the final value of both inductions @@ -233,12 +356,38 @@ func.func @y4(%lo : index, %up : index) -> index { // CHECK: ^bb2: // CHECK: %[[VAL_15:.*]] = fir.convert %[[VAL_4]] : (index) -> i32 // CHECK: %[[VAL_16:.*]] = fir.call @f4(%[[VAL_15]]) : (i32) -> i1 -// CHECK: %[[VAL_17:.*]] = arith.addi %[[VAL_4]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_17:.*]] = arith.addi %[[VAL_4]], %[[VAL_2]] : index // CHECK: br ^bb1(%[[VAL_17]], %[[VAL_16]] : index, i1) // CHECK: ^bb3: // CHECK: return %[[VAL_4]] : index // CHECK: } +// NSW-LABEL: func @y4( +// NSW-SAME: %[[VAL_0:.*]]: index, +// NSW-SAME: %[[VAL_1:.*]]: index) -> index { +// NSW: %[[VAL_2:.*]] = arith.constant 1 : index +// NSW: %[[VAL_3:.*]] = arith.constant true +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_3]] : index, i1) +// NSW: ^bb1(%[[VAL_4:.*]]: index, %[[VAL_5:.*]]: i1): +// NSW: %[[VAL_6:.*]] = arith.constant 0 : index +// NSW: %[[VAL_7:.*]] = arith.cmpi slt, %[[VAL_6]], %[[VAL_2]] : index +// NSW: %[[VAL_8:.*]] = arith.cmpi sle, %[[VAL_4]], %[[VAL_1]] : index +// NSW: %[[VAL_9:.*]] = arith.cmpi slt, %[[VAL_2]], %[[VAL_6]] : index +// NSW: %[[VAL_10:.*]] = arith.cmpi sle, %[[VAL_1]], %[[VAL_4]] : index +// NSW: %[[VAL_11:.*]] = arith.andi %[[VAL_7]], %[[VAL_8]] : i1 +// NSW: %[[VAL_12:.*]] = arith.andi %[[VAL_9]], %[[VAL_10]] : i1 +// NSW: %[[VAL_13:.*]] = arith.ori %[[VAL_11]], %[[VAL_12]] : i1 +// NSW: %[[VAL_14:.*]] = arith.andi %[[VAL_5]], %[[VAL_13]] : i1 +// NSW: cond_br %[[VAL_14]], ^bb2, ^bb3 +// NSW: ^bb2: +// NSW: %[[VAL_15:.*]] = fir.convert %[[VAL_4]] : (index) -> i32 +// NSW: %[[VAL_16:.*]] = fir.call @f4(%[[VAL_15]]) : (i32) -> i1 +// NSW: %[[VAL_17:.*]] = arith.addi %[[VAL_4]], %[[VAL_2]] overflow<nsw> : index +// NSW: br ^bb1(%[[VAL_17]], %[[VAL_16]] : index, i1) +// NSW: ^bb3: +// NSW: return %[[VAL_4]] : index +// NSW: } + // ----- // do_loop that returns the final induction value @@ -271,7 +420,7 @@ func.func @x5(%lo : index, %up : index) -> index { // CHECK: ^bb2: // CHECK: %[[VAL_12:.*]] = fir.call @f2() : () -> i1 // CHECK: %[[VAL_13:.*]] = fir.convert %[[VAL_12]] : (i1) -> i16 -// CHECK: %[[VAL_14:.*]] = arith.addi %[[VAL_7]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_14:.*]] = arith.addi %[[VAL_7]], %[[VAL_2]] : index // CHECK: %[[VAL_15:.*]] = arith.constant 1 : index // CHECK: %[[VAL_16:.*]] = arith.subi %[[VAL_9]], %[[VAL_15]] : index // CHECK: br ^bb1(%[[VAL_14]], %[[VAL_13]], %[[VAL_16]] : index, i16, index) @@ -280,6 +429,31 @@ func.func @x5(%lo : index, %up : index) -> index { // CHECK: return %[[VAL_7]] : index // CHECK: } +// NSW-LABEL: func @x5( +// NSW-SAME: %[[VAL_0:.*]]: index, +// NSW-SAME: %[[VAL_1:.*]]: index) -> index { +// NSW: %[[VAL_2:.*]] = arith.constant 1 : index +// NSW: %[[VAL_3:.*]] = arith.constant 42 : i16 +// NSW: %[[VAL_4:.*]] = arith.subi %[[VAL_1]], %[[VAL_0]] : index +// NSW: %[[VAL_5:.*]] = arith.addi %[[VAL_4]], %[[VAL_2]] : index +// NSW: %[[VAL_6:.*]] = arith.divsi %[[VAL_5]], %[[VAL_2]] : index +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_3]], %[[VAL_6]] : index, i16, index) +// NSW: ^bb1(%[[VAL_7:.*]]: index, %[[VAL_8:.*]]: i16, %[[VAL_9:.*]]: index): +// NSW: %[[VAL_10:.*]] = arith.constant 0 : index +// NSW: %[[VAL_11:.*]] = arith.cmpi sgt, %[[VAL_9]], %[[VAL_10]] : index +// NSW: cond_br %[[VAL_11]], ^bb2, ^bb3 +// NSW: ^bb2: +// NSW: %[[VAL_12:.*]] = fir.call @f2() : () -> i1 +// NSW: %[[VAL_13:.*]] = fir.convert %[[VAL_12]] : (i1) -> i16 +// NSW: %[[VAL_14:.*]] = arith.addi %[[VAL_7]], %[[VAL_2]] overflow<nsw> : index +// NSW: %[[VAL_15:.*]] = arith.constant 1 : index +// NSW: %[[VAL_16:.*]] = arith.subi %[[VAL_9]], %[[VAL_15]] : index +// NSW: br ^bb1(%[[VAL_14]], %[[VAL_13]], %[[VAL_16]] : index, i16, index) +// NSW: ^bb3: +// NSW: fir.call @f3(%[[VAL_8]]) : (i16) -> () +// NSW: return %[[VAL_7]] : index +// NSW: } + // ----- // iterate_while that returns the both induction values @@ -322,7 +496,7 @@ func.func @y5(%lo : index, %up : index) -> index { // CHECK: ^bb2: // CHECK: %[[VAL_17:.*]] = fir.call @f2() : () -> i1 // CHECK: %[[VAL_18:.*]] = fir.convert %[[VAL_17]] : (i1) -> i16 -// CHECK: %[[VAL_19:.*]] = arith.addi %[[VAL_5]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_19:.*]] = arith.addi %[[VAL_5]], %[[VAL_2]] : index // CHECK: br ^bb1(%[[VAL_19]], %[[VAL_17]], %[[VAL_18]] : index, i1, i16) // CHECK: ^bb3: // CHECK: cond_br %[[VAL_6]], ^bb4, ^bb5 @@ -334,3 +508,37 @@ func.func @y5(%lo : index, %up : index) -> index { // CHECK: fir.call @f3(%[[VAL_7]]) : (i16) -> () // CHECK: return %[[VAL_5]] : index // CHECK: } + +// NSW-LABEL: func @y5( +// NSW-SAME: %[[VAL_0:.*]]: index, +// NSW-SAME: %[[VAL_1:.*]]: index) -> index { +// NSW: %[[VAL_2:.*]] = arith.constant 1 : index +// NSW: %[[VAL_3:.*]] = arith.constant 42 : i16 +// NSW: %[[VAL_4:.*]] = arith.constant true +// NSW: br ^bb1(%[[VAL_0]], %[[VAL_4]], %[[VAL_3]] : index, i1, i16) +// NSW: ^bb1(%[[VAL_5:.*]]: index, %[[VAL_6:.*]]: i1, %[[VAL_7:.*]]: i16): +// NSW: %[[VAL_8:.*]] = arith.constant 0 : index +// NSW: %[[VAL_9:.*]] = arith.cmpi slt, %[[VAL_8]], %[[VAL_2]] : index +// NSW: %[[VAL_10:.*]] = arith.cmpi sle, %[[VAL_5]], %[[VAL_1]] : index +// NSW: %[[VAL_11:.*]] = arith.cmpi slt, %[[VAL_2]], %[[VAL_8]] : index +// NSW: %[[VAL_12:.*]] = arith.cmpi sle, %[[VAL_1]], %[[VAL_5]] : index +// NSW: %[[VAL_13:.*]] = arith.andi %[[VAL_9]], %[[VAL_10]] : i1 +// NSW: %[[VAL_14:.*]] = arith.andi %[[VAL_11]], %[[VAL_12]] : i1 +// NSW: %[[VAL_15:.*]] = arith.ori %[[VAL_13]], %[[VAL_14]] : i1 +// NSW: %[[VAL_16:.*]] = arith.andi %[[VAL_6]], %[[VAL_15]] : i1 +// NSW: cond_br %[[VAL_16]], ^bb2, ^bb3 +// NSW: ^bb2: +// NSW: %[[VAL_17:.*]] = fir.call @f2() : () -> i1 +// NSW: %[[VAL_18:.*]] = fir.convert %[[VAL_17]] : (i1) -> i16 +// NSW: %[[VAL_19:.*]] = arith.addi %[[VAL_5]], %[[VAL_2]] overflow<nsw> : index +// NSW: br ^bb1(%[[VAL_19]], %[[VAL_17]], %[[VAL_18]] : index, i1, i16) +// NSW: ^bb3: +// NSW: cond_br %[[VAL_6]], ^bb4, ^bb5 +// NSW: ^bb4: +// NSW: %[[VAL_20:.*]] = arith.constant 0 : i32 +// NSW: %[[VAL_21:.*]] = fir.call @f4(%[[VAL_20]]) : (i32) -> i1 +// NSW: br ^bb5 +// NSW: ^bb5: +// NSW: fir.call @f3(%[[VAL_7]]) : (i16) -> () +// NSW: return %[[VAL_5]] : index +// NSW: } diff --git a/flang/test/Fir/loop02.fir b/flang/test/Fir/loop02.fir index fb209a9dfeb426..50948e0e7aa6b5 100644 --- a/flang/test/Fir/loop02.fir +++ b/flang/test/Fir/loop02.fir @@ -31,7 +31,7 @@ func.func private @y(%addr : !fir.ref<index>) // CHECK: cond_br %[[VAL_13]], ^bb2, ^bb3 // CHECK: ^bb2: // CHECK: fir.call @y(%[[VAL_0]]) : (!fir.ref<index>) -> () -// CHECK: %[[VAL_14:.*]] = arith.addi %[[VAL_10]], %[[VAL_2]] overflow<nsw> : index +// CHECK: %[[VAL_14:.*]] = arith.addi %[[VAL_10]], %[[VAL_2]] : index // CHECK: %[[VAL_15:.*]] = arith.constant 1 : index // CHECK: %[[VAL_16:.*]] = arith.subi %[[VAL_11]], %[[VAL_15]] : index // CHECK: br ^bb1(%[[VAL_14]], %[[VAL_16]] : index, index) @@ -54,7 +54,7 @@ func.func private @y(%addr : !fir.ref<index>) // NOOPT: cond_br %[[VAL_9]], ^bb2, ^bb3 // NOOPT: ^bb2: // NOOPT: fir.call @y(%[[VAL_0]]) : (!fir.ref<index>) -> () -// NOOPT: %[[VAL_10:.*]] = arith.addi %[[VAL_6]], %[[VAL_2]] overflow<nsw> : index +// NOOPT: %[[VAL_10:.*]] = arith.addi %[[VAL_6]], %[[VAL_2]] : index // NOOPT: %[[VAL_11:.*]] = arith.constant 1 : index // NOOPT: %[[VAL_12:.*]] = arith.subi %[[VAL_7]], %[[VAL_11]] : index // NOOPT: br ^bb1(%[[VAL_10]], %[[VAL_12]] : index, index) diff --git a/flang/test/Lower/HLFIR/goto-do-body.f90 b/flang/test/Lower/HLFIR/goto-do-body.f90 index 5f5b09ccb8f7dc..383b839e591e33 100644 --- a/flang/test/Lower/HLFIR/goto-do-body.f90 +++ b/flang/test/Lower/HLFIR/goto-do-body.f90 @@ -40,7 +40,7 @@ subroutine sub1() ! CHECK: %[[TMP5:.*]] = arith.subi %[[TMP4]], %[[C1]] : i32 ! CHECK: fir.store %[[TMP5]] to %[[TRIP]] : !fir.ref<i32> ! CHECK: %[[TMP6:.*]] = fir.load %[[I]]#1 : !fir.ref<i32> -! CHECK: %[[TMP7:.*]] = arith.addi %[[TMP6]], %[[C1]] overflow<nsw> : i32 +! CHECK: %[[TMP7:.*]] = arith.addi %[[TMP6]], %[[C1]] : i32 ! CHECK: fir.store %[[TMP7]] to %[[I]]#1 : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER]] end do @@ -104,7 +104,7 @@ subroutine sub2() ! CHECK: fir.store %[[TMP9]] to %[[TRIP]] : !fir.ref<i32> ! CHECK: %[[TMP10:.*]] = fir.load %[[I]]#1 : !fir.ref<i32> ! CHECK: %[[STEP_VAL:.*]] = fir.load %[[STEP_VAR]] : !fir.ref<i32> -! CHECK: %[[TMP11:.*]] = arith.addi %[[TMP10]], %[[STEP_VAL]] overflow<nsw> : i32 +! CHECK: %[[TMP11:.*]] = arith.addi %[[TMP10]], %[[STEP_VAL]] : i32 ! CHECK: fir.store %[[TMP11]] to %[[I]]#1 : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER]] end do diff --git a/flang/test/Lower/OpenMP/parallel-private-clause-fixes.f90 b/flang/test/Lower/OpenMP/parallel-private-clause-fixes.f90 index 99323e69113bcc..5e76e8ff1663bf 100644 --- a/flang/test/Lower/OpenMP/parallel-private-clause-fixes.f90 +++ b/flang/test/Lower/OpenMP/parallel-private-clause-fixes.f90 @@ -82,10 +82,10 @@ ! CHECK: %[[VAL_15:.*]] = fir.load %[[PRIV_J_DECL]]#0 : !fir.ref<i32> ! CHECK: %[[VAL_16:.*]] = arith.addi %[[LOAD]], %[[VAL_15]] : i32 ! CHECK: hlfir.assign %[[VAL_16]] to %[[PRIV_X_DECL]]#0 : i32, !fir.ref<i32> -! CHECK: %[[VAL_17:.*]] = arith.addi %[[VAL_13]], %[[VAL_11]] overflow<nsw> : index +! CHECK: %[[VAL_17:.*]] = arith.addi %[[VAL_13]], %[[VAL_11]] : index ! CHECK: %[[STEPCAST:.*]] = fir.convert %[[VAL_11]] : (index) -> i32 ! CHECK: %[[IVLOAD:.*]] = fir.load %[[PRIV_J_DECL]]#1 : !fir.ref<i32> -! CHECK: %[[IVINC:.*]] = arith.addi %[[IVLOAD]], %[[STEPCAST]] overflow<nsw> : +! CHECK: %[[IVINC:.*]] = arith.addi %[[IVLOAD]], %[[STEPCAST]] ! CHECK: fir.result %[[VAL_17]], %[[IVINC]] : index, i32 ! CHECK: } ! CHECK: fir.store %[[VAL_12]]#1 to %[[PRIV_J_DECL]]#1 : !fir.ref<i32> diff --git a/flang/test/Lower/OpenMP/wsloop-reduction-allocatable-array-minmax.f90 b/flang/test/Lower/OpenMP/wsloop-reduction-allocatable-array-minmax.f90 index ce45d09d77a22a..a49eba69ff38cc 100644 --- a/flang/test/Lower/OpenMP/wsloop-reduction-allocatable-array-minmax.f90 +++ b/flang/test/Lower/OpenMP/wsloop-reduction-allocatable-array-minmax.f90 @@ -206,10 +206,10 @@ program reduce15 ! CHECK: %[[VAL_48:.*]] = fir.convert %[[VAL_47]] : (i32) -> i64 ! CHECK: %[[VAL_49:.*]] = hlfir.designate %[[VAL_46]] (%[[VAL_48]]) : (!fir.box<!fir.heap<!fir.array<?xi32>>>, i64) -> !fir.ref<i32> ! CHECK: hlfir.assign %[[VAL_45]] to %[[VAL_49]] : i32, !fir.ref<i32> -! CHECK: %[[VAL_50:.*]] = arith.addi %[[VAL_43]], %[[VAL_40]] overflow<nsw> : index +! CHECK: %[[VAL_50:.*]] = arith.addi %[[VAL_43]], %[[VAL_40]] : index ! CHECK: %[[VAL_51:.*]] = fir.convert %[[VAL_40]] : (index) -> i32 ! CHECK: %[[VAL_52:.*]] = fir.load %[[VAL_3]]#1 : !fir.ref<i32> -! CHECK: %[[VAL_53:.*]] = arith.addi %[[VAL_52]], %[[VAL_51]] overflow<nsw> : i32 +! CHECK: %[[VAL_53:.*]] = arith.addi %[[VAL_52]], %[[VAL_51]] : i32 ! CHECK: fir.result %[[VAL_50]], %[[VAL_53]] : index, i32 ! CHECK: } ! CHECK: fir.store %[[VAL_54:.*]]#1 to %[[VAL_3]]#1 : !fir.ref<i32> diff --git a/flang/test/Lower/OpenMP/wsloop-variable.f90 b/flang/test/Lower/OpenMP/wsloop-variable.f90 index cc77ce754d97e9..8d235c10fa1d60 100644 --- a/flang/test/Lower/OpenMP/wsloop-variable.f90 +++ b/flang/test/Lower/OpenMP/wsloop-variable.f90 @@ -150,10 +150,10 @@ subroutine wsloop_variable_sub !CHECK: %[[VAL_42:.*]] = arith.addi %[[VAL_40]], %[[VAL_41]] : i64 !CHECK: %[[VAL_43:.*]] = fir.convert %[[VAL_42]] : (i64) -> f32 !CHECK: hlfir.assign %[[VAL_43]] to %[[VAL_21]]#0 : f32, !fir.ref<f32> -!CHECK: %[[VAL_44:.*]] = arith.addi %[[VAL_37]], %[[VAL_34]] overflow<nsw> : index +!CHECK: %[[VAL_44:.*]] = arith.addi %[[VAL_37]], %[[VAL_34]] : index !CHECK: %[[VAL_45:.*]] = fir.convert %[[VAL_34]] : (index) -> i64 !CHECK: %[[VAL_46:.*]] = fir.load %[[VAL_17]]#1 : !fir.ref<i64> -!CHECK: %[[VAL_47:.*]] = arith.addi %[[VAL_46]], %[[VAL_45]] overflow<nsw> : i64 +!CHECK: %[[VAL_47:.*]] = arith.addi %[[VAL_46]], %[[VAL_45]] : i64 !CHECK: fir.result %[[VAL_44]], %[[VAL_47]] : index, i64 !CHECK: } !CHECK: fir.store %[[VAL_48:.*]]#1 to %[[VAL_17]]#1 : !fir.ref<i64> diff --git a/flang/test/Lower/array-character.f90 b/flang/test/Lower/array-character.f90 index 53adc5c02958c3..c93ef4be30823c 100644 --- a/flang/test/Lower/array-character.f90 +++ b/flang/test/Lower/array-character.f90 @@ -1,4 +1,4 @@ -! RUN: bbc -hlfir=false -fwrapv %s -o - | fir-opt --canonicalize --cse | FileCheck %s +! RUN: bbc -hlfir=false %s -o - | fir-opt --canonicalize --cse | FileCheck %s ! CHECK-LABEL: func @_QPissue( ! CHECK-SAME: %[[VAL_0:.*]]: !fir.boxchar<1>{{.*}}, %[[VAL_1:.*]]: !fir.boxchar<1>{{.*}}) { diff --git a/flang/test/Lower/array-derived-assignments.f90 b/flang/test/Lower/array-derived-assignments.f90 index f4e51271d59368..71e61f651302a3 100644 --- a/flang/test/Lower/array-derived-assignments.f90 +++ b/flang/test/Lower/array-derived-assignments.f90 @@ -1,5 +1,5 @@ ! Test derived type assignment lowering inside array expression -! RUN: bbc -hlfir=false -fwrapv %s -o - | FileCheck %s +! RUN: bbc -hlfir=false %s -o - | FileCheck %s module array_derived_assign type simple_copy diff --git a/flang/test/Lower/array-derived.f90 b/flang/test/Lower/array-derived.f90 index a0c55f5d88255f..b5eb7621c90f10 100644 --- a/flang/test/Lower/array-derived.f90 +++ b/flang/test/Lower/array-derived.f90 @@ -1,4 +1,4 @@ -! RUN: bbc -hlfir=false -fwrapv %s -o - | FileCheck %s +! RUN: bbc -hlfir=false %s -o - | FileCheck %s module cs type r diff --git a/flang/test/Lower/array-elemental-calls-char-byval.f90 b/flang/test/Lower/array-elemental-calls-char-byval.f90 index 682191fc789562..c321614e7fc5b5 100644 --- a/flang/test/Lower/array-elemental-calls-char-byval.f90 +++ b/flang/test/Lower/array-elemental-calls-char-byval.f90 @@ -1,6 +1,6 @@ ! Test lowering of elemental calls with character argument ! with the VALUE attribute. -! RUN: bbc -hlfir=false -fwrapv -o - %s | FileCheck %s +! RUN: bbc -hlfir=false -o - %s | FileCheck %s module char_elem_byval diff --git a/flang/test/Lower/array-elemental-calls-char.f90 b/flang/test/Lower/array-elemental-calls-char.f90 index 00e2f8e8f9c127..603cc677805fc9 100644 --- a/flang/test/Lower/array-elemental-calls-char.f90 +++ b/flang/test/Lower/array-elemental-calls-char.f90 @@ -1,6 +1,6 @@ ! Test lowering of elemental calls with character argument ! without the VALUE attribute. -! RUN: bbc -hlfir=false -fwrapv -o - %s | FileCheck %s +! RUN: bbc -hlfir=false -o - %s | FileCheck %s module char_elem diff --git a/flang/test/Lower/array-expression-assumed-size.f90 b/flang/test/Lower/array-expression-assumed-size.f90 index 2fbf315aff1148..ae35da951538b8 100644 --- a/flang/test/Lower/array-expression-assumed-size.f90 +++ b/flang/test/Lower/array-expression-assumed-size.f90 @@ -1,5 +1,5 @@ ! RUN: bbc --emit-fir -hlfir=false %s -o - | FileCheck %s -! RUN: bbc -hlfir=false -fwrapv %s -o - | FileCheck --check-prefix=PostOpt %s +! RUN: bbc -hlfir=false %s -o - | FileCheck --check-prefix=PostOpt %s subroutine assumed_size_test(a) diff --git a/flang/test/Lower/array-expression-slice-1.f90 b/flang/test/Lower/array-expression-slice-1.f90 index b597814bc0d9f1..15245090243290 100644 --- a/flang/test/Lower/array-expression-slice-1.f90 +++ b/flang/test/Lower/array-expression-slice-1.f90 @@ -1,4 +1,4 @@ -! RUN: bbc -hlfir=false -fwrapv -o - --outline-intrinsics %s | FileCheck %s +! RUN: bbc -hlfir=false -o - --outline-intrinsics %s | FileCheck %s ! CHECK-LABEL: func @_QQmain() attributes {fir.bindc_name = "p"} { ! CHECK-DAG: %[[VAL_0:.*]] = arith.constant 10 : index diff --git a/flang/test/Lower/array-substring.f90 b/flang/test/Lower/array-substring.f90 index 02101039120e9f..2e283997e3e003 100644 --- a/flang/test/Lower/array-substring.f90 +++ b/flang/test/Lower/array-substring.f90 @@ -1,7 +1,5 @@ ! RUN: bbc -hlfir=false %s -o - | FileCheck %s -! RUN: bbc -hlfir=false -fwrapv %s -o - | FileCheck %s --check-prefix=NO-NSW - -! NO-NSW-NOT: overflow<nsw> +! RUN: bbc -hlfir=false -integer-overflow %s -o - | FileCheck %s --check-prefix=NSW ! CHECK-LABEL: func @_QPtest( ! CHECK-SAME: %[[VAL_0:.*]]: !fir.boxchar<1>{{.*}}) -> !fir.array<1x!fir.logical<4>> { @@ -34,9 +32,8 @@ ! CHECK: %[[VAL_26:.*]] = fir.convert %[[VAL_25]] : (i1) -> !fir.logical<4> ! CHECK: %[[VAL_27:.*]] = fir.array_coor %[[VAL_8]](%[[VAL_9]]) %[[VAL_15]] : (!fir.ref<!fir.array<1x!fir.logical<4>>>, !fir.shape<1>, index) -> !fir.ref<!fir.logical<4>> ! CHECK: fir.store %[[VAL_26]] to %[[VAL_27]] : !fir.ref<!fir.logical<4>> -! CHECK: %[[VAL_15_NSW:.*]] = arith.addi %[[VAL_12]], %[[VAL_1]] overflow<nsw> : index ! CHECK: %[[VAL_28:.*]] = arith.subi %[[VAL_13]], %[[VAL_1]] : index -! CHECK: br ^bb1(%[[VAL_15_NSW]], %[[VAL_28]] : index, index) +! CHECK: br ^bb1(%[[VAL_15]], %[[VAL_28]] : index, index) ! CHECK: ^bb3: ! CHECK: %[[VAL_29:.*]] = fir.load %[[VAL_8]] : !fir.ref<!fir.array<1x!fir.logical<4>>> ! CHECK: return %[[VAL_29]] : !fir.array<1x!fir.logical<4>> @@ -49,3 +46,42 @@ function test(C) test = C(1:1)(1:8) == (/'ABCDabcd'/) end function test + +! NSW-LABEL: func @_QPtest( +! NSW-SAME: %[[VAL_0:.*]]: !fir.boxchar<1>{{.*}}) -> !fir.array<1x!fir.logical<4>> { +! NSW-DAG: %[[VAL_1:.*]] = arith.constant 1 : index +! NSW-DAG: %[[VAL_2:.*]] = arith.constant 0 : index +! NSW-DAG: %[[VAL_3:.*]] = arith.constant 0 : i32 +! NSW-DAG: %[[VAL_4:.*]] = arith.constant 8 : index +! NSW: %[[VAL_6:.*]]:2 = fir.unboxchar %[[VAL_0]] : (!fir.boxchar<1>) -> (!fir.ref<!fir.char<1,?>>, index) +! NSW: %[[VAL_7:.*]] = fir.convert %[[VAL_6]]#0 : (!fir.ref<!fir.char<1,?>>) -> !fir.ref<!fir.array<1x!fir.char<1,12>>> +! NSW: %[[VAL_8:.*]] = fir.alloca !fir.array<1x!fir.logical<4>> {bindc_name = "test", uniq_name = "_QFtestEtest"} +! NSW: %[[VAL_9:.*]] = fir.shape %[[VAL_1]] : (index) -> !fir.shape<1> +! NSW: %[[VAL_10:.*]] = fir.slice %[[VAL_1]], %[[VAL_1]], %[[VAL_1]] : (index, index, index) -> !fir.slice<1> +! NSW: %[[VAL_11:.*]] = fir.address_of(@_QQ{{.*}}) : !fir.ref<!fir.array<1x!fir.char<1,8>>> +! NSW: br ^bb1(%[[VAL_2]], %[[VAL_1]] : index, index) +! NSW: ^bb1(%[[VAL_12:.*]]: index, %[[VAL_13:.*]]: index): +! NSW: %[[VAL_14:.*]] = arith.cmpi sgt, %[[VAL_13]], %[[VAL_2]] : index +! NSW: cond_br %[[VAL_14]], ^bb2, ^bb3 +! NSW: ^bb2: +! NSW: %[[VAL_15:.*]] = arith.addi %[[VAL_12]], %[[VAL_1]] : index +! NSW: %[[VAL_16:.*]] = fir.array_coor %[[VAL_7]](%[[VAL_9]]) {{\[}}%[[VAL_10]]] %[[VAL_15]] : (!fir.ref<!fir.array<1x!fir.char<1,12>>>, !fir.shape<1>, !fir.slice<1>, index) -> !fir.ref<!fir.char<1,12>> +! NSW: %[[VAL_17:.*]] = fir.convert %[[VAL_16]] : (!fir.ref<!fir.char<1,12>>) -> !fir.ref<!fir.array<12x!fir.char<1>>> +! NSW: %[[VAL_18:.*]] = fir.coordinate_of %[[VAL_17]], %[[VAL_2]] : (!fir.ref<!fir.array<12x!fir.char<1>>>, index) -> !fir.ref<!fir.char<1>> +! NSW: %[[VAL_19:.*]] = fir.convert %[[VAL_18]] : (!fir.ref<!fir.char<1>>) -> !fir.ref<!fir.char<1,?>> +! NSW: %[[VAL_20:.*]] = fir.array_coor %[[VAL_11]](%[[VAL_9]]) %[[VAL_15]] : (!fir.ref<!fir.array<1x!fir.char<1,8>>>, !fir.shape<1>, index) -> !fir.ref<!fir.char<1,8>> +! NSW: %[[VAL_21:.*]] = fir.convert %[[VAL_19]] : (!fir.ref<!fir.char<1,?>>) -> !fir.ref<i8> +! NSW: %[[VAL_22:.*]] = fir.convert %[[VAL_20]] : (!fir.ref<!fir.char<1,8>>) -> !fir.ref<i8> +! NSW: %[[VAL_23:.*]] = fir.convert %[[VAL_4]] : (index) -> i64 +! NSW: %[[VAL_24:.*]] = fir.call @_FortranACharacterCompareScalar1(%[[VAL_21]], %[[VAL_22]], %[[VAL_23]], %[[VAL_23]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i8>, i64, i64) -> i32 +! NSW: %[[VAL_25:.*]] = arith.cmpi eq, %[[VAL_24]], %[[VAL_3]] : i32 +! NSW: %[[VAL_26:.*]] = fir.convert %[[VAL_25]] : (i1) -> !fir.logical<4> +! NSW: %[[VAL_27:.*]] = fir.array_coor %[[VAL_8]](%[[VAL_9]]) %[[VAL_15]] : (!fir.ref<!fir.array<1x!fir.logical<4>>>, !fir.shape<1>, index) -> !fir.ref<!fir.logical<4>> +! NSW: fir.store %[[VAL_26]] to %[[VAL_27]] : !fir.ref<!fir.logical<4>> +! NSW: %[[VAL_15_NSW:.*]] = arith.addi %[[VAL_12]], %[[VAL_1]] overflow<nsw> : index +! NSW: %[[VAL_28:.*]] = arith.subi %[[VAL_13]], %[[VAL_1]] : index +! NSW: br ^bb1(%[[VAL_15_NSW]], %[[VAL_28]] : index, index) +! NSW: ^bb3: +! NSW: %[[VAL_29:.*]] = fir.load %[[VAL_8]] : !fir.ref<!fir.array<1x!fir.logical<4>>> +! NSW: return %[[VAL_29]] : !fir.array<1x!fir.logical<4>> +! NSW: } diff --git a/flang/test/Lower/array-temp.f90 b/flang/test/Lower/array-temp.f90 index 718aef84a4e853..10c5ee91d44bda 100644 --- a/flang/test/Lower/array-temp.f90 +++ b/flang/test/Lower/array-temp.f90 @@ -1,4 +1,4 @@ -! RUN: bbc -hlfir=false -fwrapv %s -o - | FileCheck %s +! RUN: bbc -hlfir=false %s -o - | FileCheck %s ! CHECK-LABEL: func @_QPss1() subroutine ss1 diff --git a/flang/test/Lower/components.f90 b/flang/test/Lower/components.f90 index 28e836c5d10456..e1582a8a31e0d4 100644 --- a/flang/test/Lower/components.f90 +++ b/flang/test/Lower/components.f90 @@ -1,4 +1,4 @@ -! RUN: bbc -hlfir=false -fwrapv %s -o - | FileCheck %s +! RUN: bbc -hlfir=false %s -o - | FileCheck %s module components_test type t1 diff --git a/flang/test/Lower/do_loop.f90 b/flang/test/Lower/do_loop.f90 index 5d8343b8d68a45..a46e6c947391b7 100644 --- a/flang/test/Lower/do_loop.f90 +++ b/flang/test/Lower/do_loop.f90 @@ -1,17 +1,17 @@ ! RUN: bbc --use-desc-for-alloc=false -emit-fir -hlfir=false -o - %s | FileCheck %s ! RUN: %flang_fc1 -mllvm --use-desc-for-alloc=false -emit-fir -flang-deprecated-no-hlfir -o - %s | FileCheck %s -! RUN: %flang_fc1 -mllvm --use-desc-for-alloc=false -emit-fir -flang-deprecated-no-hlfir -fwrapv -o - %s | FileCheck %s --check-prefix=NO-NSW +! RUN: %flang_fc1 -mllvm --use-desc-for-alloc=false -emit-fir -flang-deprecated-no-hlfir -flang-experimental-integer-overflow -o - %s | FileCheck %s --check-prefix=NSW ! Simple tests for structured ordered loops with loop-control. ! Tests the structure of the loop, storage to index variable and return and ! storage of the final value of the index variable. -! NO-NSW-NOT: overflow<nsw> - ! Test a simple loop with the final value of the index variable read outside the loop ! CHECK-LABEL: simple_loop +! NSW-LABEL: simple_loop subroutine simple_loop ! CHECK: %[[I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFsimple_loopEi"} + ! NSW: %[[I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFsimple_loopEi"} integer :: i ! CHECK: %[[C1:.*]] = arith.constant 1 : i32 @@ -21,14 +21,18 @@ subroutine simple_loop ! CHECK: %[[C1:.*]] = arith.constant 1 : index ! CHECK: %[[LB:.*]] = fir.convert %[[C1_CVT]] : (index) -> i32 ! CHECK: %[[LI_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = + ! NSW: %[[LI_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = ! CHECK-SAME: %[[C1_CVT]] to %[[C5_CVT]] step %[[C1]] ! CHECK-SAME: iter_args(%[[IV:.*]] = %[[LB]]) -> (index, i32) { do i=1,5 ! CHECK: fir.store %[[IV]] to %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[C1]] overflow<nsw> : index + ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[C1]] : index + ! NSW: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[C1:.*]] overflow<nsw> : index ! CHECK: %[[STEPCAST:.*]] = fir.convert %[[C1]] : (index) -> i32 ! CHECK: %[[IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[IVINC:.*]] = arith.addi %[[IVLOAD]], %[[STEPCAST]] overflow<nsw> : i32 + ! NSW: %[[IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> + ! CHECK: %[[IVINC:.*]] = arith.addi %[[IVLOAD]], %[[STEPCAST]] : i32 + ! NSW: %[[IVINC:.*]] = arith.addi %[[IVLOAD]], %[[STEPCAST:.*]] overflow<nsw> : i32 ! CHECK: fir.result %[[LI_NEXT]], %[[IVINC]] : index, i32 ! CHECK: } end do @@ -40,11 +44,14 @@ subroutine simple_loop ! Test a 2-nested loop with a body composed of a reduction. Values are read from a 2d array. ! CHECK-LABEL: nested_loop +! NSW-LABEL: nested_loop subroutine nested_loop ! CHECK: %[[ARR_REF:.*]] = fir.alloca !fir.array<5x5xi32> {bindc_name = "arr", uniq_name = "_QFnested_loopEarr"} ! CHECK: %[[ASUM_REF:.*]] = fir.alloca i32 {bindc_name = "asum", uniq_name = "_QFnested_loopEasum"} ! CHECK: %[[I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFnested_loopEi"} + ! NSW: %[[I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFnested_loopEi"} ! CHECK: %[[J_REF:.*]] = fir.alloca i32 {bindc_name = "j", uniq_name = "_QFnested_loopEj"} + ! NSW: %[[J_REF:.*]] = fir.alloca i32 {bindc_name = "j", uniq_name = "_QFnested_loopEj"} integer :: asum, arr(5,5) integer :: i, j asum = 0 @@ -55,6 +62,7 @@ subroutine nested_loop ! CHECK: %[[ST_I:.*]] = arith.constant 1 : index ! CHECK: %[[I_LB:.*]] = fir.convert %[[S_I_CVT]] : (index) -> i32 ! CHECK: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = + ! NSW: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = ! CHECK-SAME: %[[S_I_CVT]] to %[[E_I_CVT]] step %[[ST_I]] ! CHECK-SAME: iter_args(%[[I_IV:.*]] = %[[I_LB]]) -> (index, i32) { do i=1,5 @@ -66,6 +74,7 @@ subroutine nested_loop ! CHECK: %[[ST_J:.*]] = arith.constant 1 : index ! CHECK: %[[J_LB:.*]] = fir.convert %[[S_J_CVT]] : (index) -> i32 ! CHECK: %[[J_RES:.*]]:2 = fir.do_loop %[[LJ:[^ ]*]] = + ! NSW: %[[J_RES:.*]]:2 = fir.do_loop %[[LJ:[^ ]*]] = ! CHECK-SAME: %[[S_J_CVT]] to %[[E_J_CVT]] step %[[ST_J]] ! CHECK-SAME: iter_args(%[[J_IV:.*]] = %[[J_LB]]) -> (index, i32) { do j=1,5 @@ -84,18 +93,24 @@ subroutine nested_loop ! CHECK: %[[ASUM_NEW:.*]] = arith.addi %[[ASUM]], %[[ARR_VAL]] : i32 ! CHECK: fir.store %[[ASUM_NEW]] to %[[ASUM_REF]] : !fir.ref<i32> asum = asum + arr(i,j) - ! CHECK: %[[LJ_NEXT:.*]] = arith.addi %[[LJ]], %[[ST_J]] overflow<nsw> : index + ! CHECK: %[[LJ_NEXT:.*]] = arith.addi %[[LJ]], %[[ST_J]] : index + ! NSW: %[[LJ_NEXT:.*]] = arith.addi %[[LJ]], %[[ST_J:.*]] overflow<nsw> : index ! CHECK: %[[J_STEPCAST:.*]] = fir.convert %[[ST_J]] : (index) -> i32 ! CHECK: %[[J_IVLOAD:.*]] = fir.load %[[J_REF]] : !fir.ref<i32> - ! CHECK: %[[J_IVINC:.*]] = arith.addi %[[J_IVLOAD]], %[[J_STEPCAST]] overflow<nsw> : i32 + ! NSW: %[[J_IVLOAD:.*]] = fir.load %[[J_REF]] : !fir.ref<i32> + ! CHECK: %[[J_IVINC:.*]] = arith.addi %[[J_IVLOAD]], %[[J_STEPCAST]] : i32 + ! NSW: %[[J_IVINC:.*]] = arith.addi %[[J_IVLOAD]], %[[J_STEPCAST:.*]] overflow<nsw> : i32 ! CHECK: fir.result %[[LJ_NEXT]], %[[J_IVINC]] : index, i32 ! CHECK: } end do ! CHECK: fir.store %[[J_RES]]#1 to %[[J_REF]] : !fir.ref<i32> - ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_I]] overflow<nsw> : index + ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_I]] : index + ! NSW: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_I:.*]] overflow<nsw> : index ! CHECK: %[[I_STEPCAST:.*]] = fir.convert %[[ST_I]] : (index) -> i32 ! CHECK: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] overflow<nsw> : i32 + ! NSW: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> + ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] : i32 + ! NSW: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST:.*]] overflow<nsw> : i32 ! CHECK: fir.result %[[LI_NEXT]], %[[I_IVINC]] : index, i32 ! CHECK: } end do @@ -104,9 +119,11 @@ subroutine nested_loop ! Test a downcounting loop ! CHECK-LABEL: down_counting_loop +! NSW-LABEL: down_counting_loop subroutine down_counting_loop() integer :: i ! CHECK: %[[I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFdown_counting_loopEi"} + ! NSW: %[[I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFdown_counting_loopEi"} ! CHECK: %[[C5:.*]] = arith.constant 5 : i32 ! CHECK: %[[C5_CVT:.*]] = fir.convert %[[C5]] : (i32) -> index @@ -116,14 +133,18 @@ subroutine down_counting_loop() ! CHECK: %[[CMINUS1_STEP_CVT:.*]] = fir.convert %[[CMINUS1]] : (i32) -> index ! CHECK: %[[I_LB:.*]] = fir.convert %[[C5_CVT]] : (index) -> i32 ! CHECK: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = + ! NSW: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = ! CHECK-SAME: %[[C5_CVT]] to %[[C1_CVT]] step %[[CMINUS1_STEP_CVT]] ! CHECK-SAME: iter_args(%[[I_IV:.*]] = %[[I_LB]]) -> (index, i32) { do i=5,1,-1 ! CHECK: fir.store %[[I_IV]] to %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[CMINUS1_STEP_CVT]] overflow<nsw> : index + ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[CMINUS1_STEP_CVT]] : index + ! NSW: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[CMINUS1_STEP_CVT:.*]] overflow<nsw> : index ! CHECK: %[[I_STEPCAST:.*]] = fir.convert %[[CMINUS1_STEP_CVT]] : (index) -> i32 ! CHECK: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] overflow<nsw> : i32 + ! NSW: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> + ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] : i32 + ! NSW: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST:.*]] overflow<nsw> : i32 ! CHECK: fir.result %[[LI_NEXT]], %[[I_IVINC]] : index, i32 ! CHECK: } end do @@ -132,6 +153,7 @@ subroutine down_counting_loop() ! Test a general loop with a variable step ! CHECK-LABEL: loop_with_variable_step +! NSW-LABEL: loop_with_variable_step ! CHECK-SAME: (%[[S_REF:.*]]: !fir.ref<i32> {fir.bindc_name = "s"}, %[[E_REF:.*]]: !fir.ref<i32> {fir.bindc_name = "e"}, %[[ST_REF:.*]]: !fir.ref<i32> {fir.bindc_name = "st"}) { subroutine loop_with_variable_step(s,e,st) integer :: s, e, st @@ -144,14 +166,18 @@ subroutine loop_with_variable_step(s,e,st) ! CHECK: %[[ST_CVT:.*]] = fir.convert %[[ST]] : (i32) -> index ! CHECK: %[[I_LB:.*]] = fir.convert %[[S_CVT]] : (index) -> i32 ! CHECK: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = + ! NSW: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = ! CHECK-SAME: %[[S_CVT]] to %[[E_CVT]] step %[[ST_CVT]] ! CHECK-SAME: iter_args(%[[I_IV:.*]] = %[[I_LB]]) -> (index, i32) { do i=s,e,st ! CHECK: fir.store %[[I_IV]] to %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT]] overflow<nsw> : index + ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT]] : index + ! NSW: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT:.*]] overflow<nsw> : index ! CHECK: %[[I_STEPCAST:.*]] = fir.convert %[[ST_CVT]] : (index) -> i32 ! CHECK: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] overflow<nsw> : i32 + ! NSW: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> + ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] : i32 + ! NSW: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST:.*]] overflow<nsw> : i32 ! CHECK: fir.result %[[LI_NEXT]], %[[I_IVINC]] : index, i32 ! CHECK: } end do @@ -160,11 +186,13 @@ subroutine loop_with_variable_step(s,e,st) ! Test usage of pointer variables as index, start, end and step variables ! CHECK-LABEL: loop_with_pointer_variables +! NSW-LABEL: loop_with_pointer_variables ! CHECK-SAME: (%[[S_REF:.*]]: !fir.ref<i32> {fir.bindc_name = "s", fir.target}, %[[E_REF:.*]]: !fir.ref<i32> {fir.bindc_name = "e", fir.target}, %[[ST_REF:.*]]: !fir.ref<i32> {fir.bindc_name = "st", fir.target}) { subroutine loop_with_pointer_variables(s,e,st) ! CHECK: %[[E_PTR_REF:.*]] = fir.alloca !fir.ptr<i32> {uniq_name = "_QFloop_with_pointer_variablesEeptr.addr"} ! CHECK: %[[I_REF:.*]] = fir.alloca i32 {bindc_name = "i", fir.target, uniq_name = "_QFloop_with_pointer_variablesEi"} ! CHECK: %[[I_PTR_REF:.*]] = fir.alloca !fir.ptr<i32> {uniq_name = "_QFloop_with_pointer_variablesEiptr.addr"} +! NSW: %[[I_PTR_REF:.*]] = fir.alloca !fir.ptr<i32> {uniq_name = "_QFloop_with_pointer_variablesEiptr.addr"} ! CHECK: %[[S_PTR_REF:.*]] = fir.alloca !fir.ptr<i32> {uniq_name = "_QFloop_with_pointer_variablesEsptr.addr"} ! CHECK: %[[ST_PTR_REF:.*]] = fir.alloca !fir.ptr<i32> {uniq_name = "_QFloop_with_pointer_variablesEstptr.addr"} integer, target :: i @@ -185,6 +213,7 @@ subroutine loop_with_pointer_variables(s,e,st) stptr => st ! CHECK: %[[I_PTR:.*]] = fir.load %[[I_PTR_REF]] : !fir.ref<!fir.ptr<i32>> +! NSW: %[[I_PTR:.*]] = fir.load %[[I_PTR_REF]] : !fir.ref<!fir.ptr<i32>> ! CHECK: %[[S_PTR:.*]] = fir.load %[[S_PTR_REF]] : !fir.ref<!fir.ptr<i32>> ! CHECK: %[[S:.*]] = fir.load %[[S_PTR]] : !fir.ptr<i32> ! CHECK: %[[S_CVT:.*]] = fir.convert %[[S]] : (i32) -> index @@ -196,14 +225,18 @@ subroutine loop_with_pointer_variables(s,e,st) ! CHECK: %[[ST_CVT:.*]] = fir.convert %[[ST]] : (i32) -> index ! CHECK: %[[I_LB:.*]] = fir.convert %[[S_CVT]] : (index) -> i32 ! CHECK: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = +! NSW: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = ! CHECK-SAME: %[[S_CVT]] to %[[E_CVT]] step %[[ST_CVT]] ! CHECK-SAME: iter_args(%[[I_IV:.*]] = %[[I_LB]]) -> (index, i32) { do iptr=sptr,eptr,stptr ! CHECK: fir.store %[[I_IV]] to %[[I_PTR]] : !fir.ptr<i32> -! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT]] overflow<nsw> : index +! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT]] : index +! NSW: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT:.*]] overflow<nsw> : index ! CHECK: %[[I_STEPCAST:.*]] = fir.convert %[[ST_CVT]] : (index) -> i32 ! CHECK: %[[I_IVLOAD:.*]] = fir.load %[[I_PTR]] : !fir.ptr<i32> -! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] overflow<nsw> : i32 +! NSW: %[[I_IVLOAD:.*]] = fir.load %[[I_PTR]] : !fir.ptr<i32> +! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] : i32 +! NSW: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST:.*]] overflow<nsw> : i32 ! CHECK: fir.result %[[LI_NEXT]], %[[I_IVINC]] : index, i32 end do ! CHECK: } @@ -212,9 +245,11 @@ subroutine loop_with_pointer_variables(s,e,st) ! Test usage of non-default integer kind for loop control and loop index variable ! CHECK-LABEL: loop_with_non_default_integer +! NSW-LABEL: loop_with_non_default_integer ! CHECK-SAME: (%[[S_REF:.*]]: !fir.ref<i64> {fir.bindc_name = "s"}, %[[E_REF:.*]]: !fir.ref<i64> {fir.bindc_name = "e"}, %[[ST_REF:.*]]: !fir.ref<i64> {fir.bindc_name = "st"}) { subroutine loop_with_non_default_integer(s,e,st) ! CHECK: %[[I_REF:.*]] = fir.alloca i64 {bindc_name = "i", uniq_name = "_QFloop_with_non_default_integerEi"} + ! NSW: %[[I_REF:.*]] = fir.alloca i64 {bindc_name = "i", uniq_name = "_QFloop_with_non_default_integerEi"} integer(kind=8):: i ! CHECK: %[[S:.*]] = fir.load %[[S_REF]] : !fir.ref<i64> ! CHECK: %[[S_CVT:.*]] = fir.convert %[[S]] : (i64) -> index @@ -226,14 +261,18 @@ subroutine loop_with_non_default_integer(s,e,st) ! CHECK: %[[I_LB:.*]] = fir.convert %[[S_CVT]] : (index) -> i64 ! CHECK: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = + ! NSW: %[[I_RES:.*]]:2 = fir.do_loop %[[LI:[^ ]*]] = ! CHECK-SAME: %[[S_CVT]] to %[[E_CVT]] step %[[ST_CVT]] ! CHECK-SAME: iter_args(%[[I_IV:.*]] = %[[I_LB]]) -> (index, i64) { do i=s,e,st ! CHECK: fir.store %[[I_IV]] to %[[I_REF]] : !fir.ref<i64> - ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT]] overflow<nsw> : index + ! CHECK: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT]] : index + ! NSW: %[[LI_NEXT:.*]] = arith.addi %[[LI]], %[[ST_CVT:.*]] overflow<nsw> : index ! CHECK: %[[I_STEPCAST:.*]] = fir.convert %[[ST_CVT]] : (index) -> i64 ! CHECK: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i64> - ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] overflow<nsw> : i64 + ! NSW: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i64> + ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] : i64 + ! NSW: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST:.*]] overflow<nsw> : i64 ! CHECK: fir.result %[[LI_NEXT]], %[[I_IVINC]] : index, i64 end do ! CHECK: } diff --git a/flang/test/Lower/do_loop_unstructured.f90 b/flang/test/Lower/do_loop_unstructured.f90 index d8890b2d0926ea..e1a669e09c9a89 100644 --- a/flang/test/Lower/do_loop_unstructured.f90 +++ b/flang/test/Lower/do_loop_unstructured.f90 @@ -1,11 +1,9 @@ ! RUN: bbc -emit-fir -hlfir=false -o - %s | FileCheck %s ! RUN: %flang_fc1 -emit-fir -flang-deprecated-no-hlfir -o - %s | FileCheck %s -! RUN: %flang_fc1 -emit-fir -flang-deprecated-no-hlfir -fwrapv -o - %s | FileCheck %s --check-prefix=NO-NSW +! RUN: %flang_fc1 -emit-fir -flang-deprecated-no-hlfir -flang-experimental-integer-overflow -o - %s | FileCheck %s --check-prefix=NSW ! Tests for unstructured loops. -! NO-NSW-NOT: overflow<nsw> - ! Test a simple unstructured loop. Test for the existence of, ! -> The initialization of the trip-count and loop-variable ! -> The branch to the body or the exit inside the header @@ -41,12 +39,42 @@ subroutine simple_unstructured() ! CHECK: fir.store %[[TRIP_VAR_NEXT]] to %[[TRIP_VAR_REF]] : !fir.ref<i32> ! CHECK: %[[LOOP_VAR:.*]] = fir.load %[[LOOP_VAR_REF]] : !fir.ref<i32> ! CHECK: %[[STEP_ONE_2:.*]] = arith.constant 1 : i32 -! CHECK: %[[LOOP_VAR_NEXT:.*]] = arith.addi %[[LOOP_VAR]], %[[STEP_ONE_2]] overflow<nsw> : i32 +! CHECK: %[[LOOP_VAR_NEXT:.*]] = arith.addi %[[LOOP_VAR]], %[[STEP_ONE_2]] : i32 ! CHECK: fir.store %[[LOOP_VAR_NEXT]] to %[[LOOP_VAR_REF]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER]] ! CHECK: ^[[EXIT]]: ! CHECK: return +! NSW-LABEL: simple_unstructured +! NSW: %[[TRIP_VAR_REF:.*]] = fir.alloca i32 +! NSW: %[[LOOP_VAR_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFsimple_unstructuredEi"} +! NSW: %[[ONE:.*]] = arith.constant 1 : i32 +! NSW: %[[HUNDRED:.*]] = arith.constant 100 : i32 +! NSW: %[[STEP_ONE:.*]] = arith.constant 1 : i32 +! NSW: %[[TMP1:.*]] = arith.subi %[[HUNDRED]], %[[ONE]] : i32 +! NSW: %[[TMP2:.*]] = arith.addi %[[TMP1]], %[[STEP_ONE]] : i32 +! NSW: %[[TRIP_COUNT:.*]] = arith.divsi %[[TMP2]], %[[STEP_ONE]] : i32 +! NSW: fir.store %[[TRIP_COUNT]] to %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: fir.store %[[ONE]] to %[[LOOP_VAR_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER:.*]] +! NSW: ^[[HEADER]]: +! NSW: %[[TRIP_VAR:.*]] = fir.load %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[ZERO:.*]] = arith.constant 0 : i32 +! NSW: %[[COND:.*]] = arith.cmpi sgt, %[[TRIP_VAR]], %[[ZERO]] : i32 +! NSW: cf.cond_br %[[COND]], ^[[BODY:.*]], ^[[EXIT:.*]] +! NSW: ^[[BODY]]: +! NSW: %[[TRIP_VAR:.*]] = fir.load %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[ONE_1:.*]] = arith.constant 1 : i32 +! NSW: %[[TRIP_VAR_NEXT:.*]] = arith.subi %[[TRIP_VAR]], %[[ONE_1]] : i32 +! NSW: fir.store %[[TRIP_VAR_NEXT]] to %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[LOOP_VAR:.*]] = fir.load %[[LOOP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[STEP_ONE_2:.*]] = arith.constant 1 : i32 +! NSW: %[[LOOP_VAR_NEXT:.*]] = arith.addi %[[LOOP_VAR]], %[[STEP_ONE_2]] overflow<nsw> : i32 +! NSW: fir.store %[[LOOP_VAR_NEXT]] to %[[LOOP_VAR_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER]] +! NSW: ^[[EXIT]]: +! NSW: return + ! Test an unstructured loop with a step. Mostly similar to the previous one. ! Only diff erence is a non-unit step. subroutine simple_unstructured_with_step() @@ -80,12 +108,42 @@ subroutine simple_unstructured_with_step() ! CHECK: fir.store %[[TRIP_VAR_NEXT]] to %[[TRIP_VAR_REF]] : !fir.ref<i32> ! CHECK: %[[LOOP_VAR:.*]] = fir.load %[[LOOP_VAR_REF]] : !fir.ref<i32> ! CHECK: %[[STEP_2:.*]] = arith.constant 2 : i32 -! CHECK: %[[LOOP_VAR_NEXT:.*]] = arith.addi %[[LOOP_VAR]], %[[STEP_2]] overflow<nsw> : i32 +! CHECK: %[[LOOP_VAR_NEXT:.*]] = arith.addi %[[LOOP_VAR]], %[[STEP_2]] : i32 ! CHECK: fir.store %[[LOOP_VAR_NEXT]] to %[[LOOP_VAR_REF]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER]] ! CHECK: ^[[EXIT]]: ! CHECK: return +! NSW-LABEL: simple_unstructured_with_step +! NSW: %[[TRIP_VAR_REF:.*]] = fir.alloca i32 +! NSW: %[[LOOP_VAR_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFsimple_unstructured_with_stepEi"} +! NSW: %[[ONE:.*]] = arith.constant 1 : i32 +! NSW: %[[HUNDRED:.*]] = arith.constant 100 : i32 +! NSW: %[[STEP:.*]] = arith.constant 2 : i32 +! NSW: %[[TMP1:.*]] = arith.subi %[[HUNDRED]], %[[ONE]] : i32 +! NSW: %[[TMP2:.*]] = arith.addi %[[TMP1]], %[[STEP]] : i32 +! NSW: %[[TRIP_COUNT:.*]] = arith.divsi %[[TMP2]], %[[STEP]] : i32 +! NSW: fir.store %[[TRIP_COUNT]] to %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: fir.store %[[ONE]] to %[[LOOP_VAR_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER:.*]] +! NSW: ^[[HEADER]]: +! NSW: %[[TRIP_VAR:.*]] = fir.load %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[ZERO:.*]] = arith.constant 0 : i32 +! NSW: %[[COND:.*]] = arith.cmpi sgt, %[[TRIP_VAR]], %[[ZERO]] : i32 +! NSW: cf.cond_br %[[COND]], ^[[BODY:.*]], ^[[EXIT:.*]] +! NSW: ^[[BODY]]: +! NSW: %[[TRIP_VAR:.*]] = fir.load %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[ONE_1:.*]] = arith.constant 1 : i32 +! NSW: %[[TRIP_VAR_NEXT:.*]] = arith.subi %[[TRIP_VAR]], %[[ONE_1]] : i32 +! NSW: fir.store %[[TRIP_VAR_NEXT]] to %[[TRIP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[LOOP_VAR:.*]] = fir.load %[[LOOP_VAR_REF]] : !fir.ref<i32> +! NSW: %[[STEP_2:.*]] = arith.constant 2 : i32 +! NSW: %[[LOOP_VAR_NEXT:.*]] = arith.addi %[[LOOP_VAR]], %[[STEP_2]] overflow<nsw> : i32 +! NSW: fir.store %[[LOOP_VAR_NEXT]] to %[[LOOP_VAR_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER]] +! NSW: ^[[EXIT]]: +! NSW: return + ! Test a three nested unstructured loop. Three nesting is the basic case where ! we have loops that are neither innermost or outermost. subroutine nested_unstructured() @@ -157,7 +215,7 @@ subroutine nested_unstructured() ! CHECK: fir.store %[[TRIP_VAR_K_NEXT]] to %[[TRIP_VAR_K_REF]] : !fir.ref<i32> ! CHECK: %[[LOOP_VAR_K:.*]] = fir.load %[[LOOP_VAR_K_REF]] : !fir.ref<i32> ! CHECK: %[[K_STEP_2:.*]] = arith.constant 1 : i32 -! CHECK: %[[LOOP_VAR_K_NEXT:.*]] = arith.addi %[[LOOP_VAR_K]], %[[K_STEP_2]] overflow<nsw> : i32 +! CHECK: %[[LOOP_VAR_K_NEXT:.*]] = arith.addi %[[LOOP_VAR_K]], %[[K_STEP_2]] : i32 ! CHECK: fir.store %[[LOOP_VAR_K_NEXT]] to %[[LOOP_VAR_K_REF]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER_K]] ! CHECK: ^[[EXIT_K]]: @@ -167,7 +225,7 @@ subroutine nested_unstructured() ! CHECK: fir.store %[[TRIP_VAR_J_NEXT]] to %[[TRIP_VAR_J_REF]] : !fir.ref<i32> ! CHECK: %[[LOOP_VAR_J:.*]] = fir.load %[[LOOP_VAR_J_REF]] : !fir.ref<i32> ! CHECK: %[[J_STEP_2:.*]] = arith.constant 1 : i32 -! CHECK: %[[LOOP_VAR_J_NEXT:.*]] = arith.addi %[[LOOP_VAR_J]], %[[J_STEP_2]] overflow<nsw> : i32 +! CHECK: %[[LOOP_VAR_J_NEXT:.*]] = arith.addi %[[LOOP_VAR_J]], %[[J_STEP_2]] : i32 ! CHECK: fir.store %[[LOOP_VAR_J_NEXT]] to %[[LOOP_VAR_J_REF]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER_J]] ! CHECK: ^[[EXIT_J]]: @@ -177,12 +235,96 @@ subroutine nested_unstructured() ! CHECK: fir.store %[[TRIP_VAR_I_NEXT]] to %[[TRIP_VAR_I_REF]] : !fir.ref<i32> ! CHECK: %[[LOOP_VAR_I:.*]] = fir.load %[[LOOP_VAR_I_REF]] : !fir.ref<i32> ! CHECK: %[[I_STEP_2:.*]] = arith.constant 1 : i32 -! CHECK: %[[LOOP_VAR_I_NEXT:.*]] = arith.addi %[[LOOP_VAR_I]], %[[I_STEP_2]] overflow<nsw> : i32 +! CHECK: %[[LOOP_VAR_I_NEXT:.*]] = arith.addi %[[LOOP_VAR_I]], %[[I_STEP_2]] : i32 ! CHECK: fir.store %[[LOOP_VAR_I_NEXT]] to %[[LOOP_VAR_I_REF]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER_I]] ! CHECK: ^[[EXIT_I]]: ! CHECK: return +! NSW-LABEL: nested_unstructured +! NSW: %[[TRIP_VAR_K_REF:.*]] = fir.alloca i32 +! NSW: %[[TRIP_VAR_J_REF:.*]] = fir.alloca i32 +! NSW: %[[TRIP_VAR_I_REF:.*]] = fir.alloca i32 +! NSW: %[[LOOP_VAR_I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFnested_unstructuredEi"} +! NSW: %[[LOOP_VAR_J_REF:.*]] = fir.alloca i32 {bindc_name = "j", uniq_name = "_QFnested_unstructuredEj"} +! NSW: %[[LOOP_VAR_K_REF:.*]] = fir.alloca i32 {bindc_name = "k", uniq_name = "_QFnested_unstructuredEk"} +! NSW: %[[I_START:.*]] = arith.constant 1 : i32 +! NSW: %[[I_END:.*]] = arith.constant 100 : i32 +! NSW: %[[I_STEP:.*]] = arith.constant 1 : i32 +! NSW: %[[TMP1:.*]] = arith.subi %[[I_END]], %[[I_START]] : i32 +! NSW: %[[TMP2:.*]] = arith.addi %[[TMP1]], %[[I_STEP]] : i32 +! NSW: %[[TRIP_COUNT_I:.*]] = arith.divsi %[[TMP2]], %[[I_STEP]] : i32 +! NSW: fir.store %[[TRIP_COUNT_I]] to %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: fir.store %[[I_START]] to %[[LOOP_VAR_I_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER_I:.*]] +! NSW: ^[[HEADER_I]]: +! NSW: %[[TRIP_VAR_I:.*]] = fir.load %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[ZERO_1:.*]] = arith.constant 0 : i32 +! NSW: %[[COND_I:.*]] = arith.cmpi sgt, %[[TRIP_VAR_I]], %[[ZERO_1]] : i32 +! NSW: cf.cond_br %[[COND_I]], ^[[BODY_I:.*]], ^[[EXIT_I:.*]] +! NSW: ^[[BODY_I]]: +! NSW: %[[J_START:.*]] = arith.constant 1 : i32 +! NSW: %[[J_END:.*]] = arith.constant 200 : i32 +! NSW: %[[J_STEP:.*]] = arith.constant 1 : i32 +! NSW: %[[TMP3:.*]] = arith.subi %[[J_END]], %[[J_START]] : i32 +! NSW: %[[TMP4:.*]] = arith.addi %[[TMP3]], %[[J_STEP]] : i32 +! NSW: %[[TRIP_COUNT_J:.*]] = arith.divsi %[[TMP4]], %[[J_STEP]] : i32 +! NSW: fir.store %[[TRIP_COUNT_J]] to %[[TRIP_VAR_J_REF]] : !fir.ref<i32> +! NSW: fir.store %[[J_START]] to %[[LOOP_VAR_J_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER_J:.*]] +! NSW: ^[[HEADER_J]]: +! NSW: %[[TRIP_VAR_J:.*]] = fir.load %[[TRIP_VAR_J_REF]] : !fir.ref<i32> +! NSW: %[[ZERO_2:.*]] = arith.constant 0 : i32 +! NSW: %[[COND_J:.*]] = arith.cmpi sgt, %[[TRIP_VAR_J]], %[[ZERO_2]] : i32 +! NSW: cf.cond_br %[[COND_J]], ^[[BODY_J:.*]], ^[[EXIT_J:.*]] +! NSW: ^[[BODY_J]]: +! NSW: %[[K_START:.*]] = arith.constant 1 : i32 +! NSW: %[[K_END:.*]] = arith.constant 300 : i32 +! NSW: %[[K_STEP:.*]] = arith.constant 1 : i32 +! NSW: %[[TMP3:.*]] = arith.subi %[[K_END]], %[[K_START]] : i32 +! NSW: %[[TMP4:.*]] = arith.addi %[[TMP3]], %[[K_STEP]] : i32 +! NSW: %[[TRIP_COUNT_K:.*]] = arith.divsi %[[TMP4]], %[[K_STEP]] : i32 +! NSW: fir.store %[[TRIP_COUNT_K]] to %[[TRIP_VAR_K_REF]] : !fir.ref<i32> +! NSW: fir.store %[[K_START]] to %[[LOOP_VAR_K_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER_K:.*]] +! NSW: ^[[HEADER_K]]: +! NSW: %[[TRIP_VAR_K:.*]] = fir.load %[[TRIP_VAR_K_REF]] : !fir.ref<i32> +! NSW: %[[ZERO_2:.*]] = arith.constant 0 : i32 +! NSW: %[[COND_K:.*]] = arith.cmpi sgt, %[[TRIP_VAR_K]], %[[ZERO_2]] : i32 +! NSW: cf.cond_br %[[COND_K]], ^[[BODY_K:.*]], ^[[EXIT_K:.*]] +! NSW: ^[[BODY_K]]: +! NSW: %[[TRIP_VAR_K:.*]] = fir.load %[[TRIP_VAR_K_REF]] : !fir.ref<i32> +! NSW: %[[ONE_1:.*]] = arith.constant 1 : i32 +! NSW: %[[TRIP_VAR_K_NEXT:.*]] = arith.subi %[[TRIP_VAR_K]], %[[ONE_1]] : i32 +! NSW: fir.store %[[TRIP_VAR_K_NEXT]] to %[[TRIP_VAR_K_REF]] : !fir.ref<i32> +! NSW: %[[LOOP_VAR_K:.*]] = fir.load %[[LOOP_VAR_K_REF]] : !fir.ref<i32> +! NSW: %[[K_STEP_2:.*]] = arith.constant 1 : i32 +! NSW: %[[LOOP_VAR_K_NEXT:.*]] = arith.addi %[[LOOP_VAR_K]], %[[K_STEP_2]] overflow<nsw> : i32 +! NSW: fir.store %[[LOOP_VAR_K_NEXT]] to %[[LOOP_VAR_K_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER_K]] +! NSW: ^[[EXIT_K]]: +! NSW: %[[TRIP_VAR_J:.*]] = fir.load %[[TRIP_VAR_J_REF]] : !fir.ref<i32> +! NSW: %[[ONE_1:.*]] = arith.constant 1 : i32 +! NSW: %[[TRIP_VAR_J_NEXT:.*]] = arith.subi %[[TRIP_VAR_J]], %[[ONE_1]] : i32 +! NSW: fir.store %[[TRIP_VAR_J_NEXT]] to %[[TRIP_VAR_J_REF]] : !fir.ref<i32> +! NSW: %[[LOOP_VAR_J:.*]] = fir.load %[[LOOP_VAR_J_REF]] : !fir.ref<i32> +! NSW: %[[J_STEP_2:.*]] = arith.constant 1 : i32 +! NSW: %[[LOOP_VAR_J_NEXT:.*]] = arith.addi %[[LOOP_VAR_J]], %[[J_STEP_2]] overflow<nsw> : i32 +! NSW: fir.store %[[LOOP_VAR_J_NEXT]] to %[[LOOP_VAR_J_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER_J]] +! NSW: ^[[EXIT_J]]: +! NSW: %[[TRIP_VAR_I:.*]] = fir.load %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[ONE_1:.*]] = arith.constant 1 : i32 +! NSW: %[[TRIP_VAR_I_NEXT:.*]] = arith.subi %[[TRIP_VAR_I]], %[[ONE_1]] : i32 +! NSW: fir.store %[[TRIP_VAR_I_NEXT]] to %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[LOOP_VAR_I:.*]] = fir.load %[[LOOP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[I_STEP_2:.*]] = arith.constant 1 : i32 +! NSW: %[[LOOP_VAR_I_NEXT:.*]] = arith.addi %[[LOOP_VAR_I]], %[[I_STEP_2]] overflow<nsw> : i32 +! NSW: fir.store %[[LOOP_VAR_I_NEXT]] to %[[LOOP_VAR_I_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER_I]] +! NSW: ^[[EXIT_I]]: +! NSW: return + ! Test the existence of a structured loop inside an unstructured loop. ! Only minimal checks are inserted for the structured loop. subroutine nested_structured_in_unstructured() @@ -217,9 +359,9 @@ subroutine nested_structured_in_unstructured() ! CHECK-SAME: %{{.*}} to %{{.*}} step %[[ST:[^ ]*]] ! CHECK-SAME: iter_args(%[[J_IV:.*]] = %{{.*}}) -> (index, i32) { ! CHECK: fir.store %[[J_IV]] to %[[LOOP_VAR_J_REF]] : !fir.ref<i32> -! CHECK: %[[J_INDEX_NEXT:.*]] = arith.addi %[[J_INDEX]], %[[ST]] overflow<nsw> : index +! CHECK: %[[J_INDEX_NEXT:.*]] = arith.addi %[[J_INDEX]], %[[ST]] : index ! CHECK: %[[LOOP_VAR_J:.*]] = fir.load %[[LOOP_VAR_J_REF]] : !fir.ref<i32> -! CHECK: %[[LOOP_VAR_J_NEXT:.*]] = arith.addi %[[LOOP_VAR_J]], %{{[^ ]*}} overflow<nsw> : i32 +! CHECK: %[[LOOP_VAR_J_NEXT:.*]] = arith.addi %[[LOOP_VAR_J]], %{{[^ ]*}} : i32 ! CHECK: } ! CHECK: %[[TRIP_VAR_I:.*]] = fir.load %[[TRIP_VAR_I_REF]] : !fir.ref<i32> ! CHECK: %[[C1_3:.*]] = arith.constant 1 : i32 @@ -227,8 +369,47 @@ subroutine nested_structured_in_unstructured() ! CHECK: fir.store %[[TRIP_VAR_I_NEXT]] to %[[TRIP_VAR_I_REF]] : !fir.ref<i32> ! CHECK: %[[LOOP_VAR_I:.*]] = fir.load %[[LOOP_VAR_I_REF]] : !fir.ref<i32> ! CHECK: %[[I_STEP_2:.*]] = arith.constant 1 : i32 -! CHECK: %[[LOOP_VAR_I_NEXT:.*]] = arith.addi %[[LOOP_VAR_I]], %[[I_STEP_2]] overflow<nsw> : i32 +! CHECK: %[[LOOP_VAR_I_NEXT:.*]] = arith.addi %[[LOOP_VAR_I]], %[[I_STEP_2]] : i32 ! CHECK: fir.store %[[LOOP_VAR_I_NEXT]] to %[[LOOP_VAR_I_REF]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER]] ! CHECK: ^[[EXIT]]: ! CHECK: return + +! NSW-LABEL: nested_structured_in_unstructured +! NSW: %[[TRIP_VAR_I_REF:.*]] = fir.alloca i32 +! NSW: %[[LOOP_VAR_I_REF:.*]] = fir.alloca i32 {bindc_name = "i", uniq_name = "_QFnested_structured_in_unstructuredEi"} +! NSW: %[[LOOP_VAR_J_REF:.*]] = fir.alloca i32 {bindc_name = "j", uniq_name = "_QFnested_structured_in_unstructuredEj"} +! NSW: %[[I_START:.*]] = arith.constant 1 : i32 +! NSW: %[[I_END:.*]] = arith.constant 100 : i32 +! NSW: %[[I_STEP:.*]] = arith.constant 1 : i32 +! NSW: %[[TMP1:.*]] = arith.subi %[[I_END]], %[[I_START]] : i32 +! NSW: %[[TMP2:.*]] = arith.addi %[[TMP1]], %[[I_STEP]] : i32 +! NSW: %[[TRIP_COUNT:.*]] = arith.divsi %[[TMP2]], %[[I_STEP]] : i32 +! NSW: fir.store %[[TRIP_COUNT]] to %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: fir.store %[[I_START]] to %[[LOOP_VAR_I_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER:.*]] +! NSW: ^[[HEADER]]: +! NSW: %[[TRIP_VAR:.*]] = fir.load %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[ZERO:.*]] = arith.constant 0 : i32 +! NSW: %[[COND:.*]] = arith.cmpi sgt, %[[TRIP_VAR]], %[[ZERO]] : i32 +! NSW: cf.cond_br %[[COND]], ^[[BODY:.*]], ^[[EXIT:.*]] +! NSW: ^[[BODY]]: +! NSW: %{{.*}} = fir.do_loop %[[J_INDEX:[^ ]*]] = +! NSW-SAME: %{{.*}} to %{{.*}} step %[[ST:[^ ]*]] +! NSW-SAME: iter_args(%[[J_IV:.*]] = %{{.*}}) -> (index, i32) { +! NSW: fir.store %[[J_IV]] to %[[LOOP_VAR_J_REF]] : !fir.ref<i32> +! NSW: %[[J_INDEX_NEXT:.*]] = arith.addi %[[J_INDEX]], %[[ST]] overflow<nsw> : index +! NSW: %[[LOOP_VAR_J:.*]] = fir.load %[[LOOP_VAR_J_REF]] : !fir.ref<i32> +! NSW: %[[LOOP_VAR_J_NEXT:.*]] = arith.addi %[[LOOP_VAR_J]], %{{[^ ]*}} overflow<nsw> : i32 +! NSW: } +! NSW: %[[TRIP_VAR_I:.*]] = fir.load %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[C1_3:.*]] = arith.constant 1 : i32 +! NSW: %[[TRIP_VAR_I_NEXT:.*]] = arith.subi %[[TRIP_VAR_I]], %[[C1_3]] : i32 +! NSW: fir.store %[[TRIP_VAR_I_NEXT]] to %[[TRIP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[LOOP_VAR_I:.*]] = fir.load %[[LOOP_VAR_I_REF]] : !fir.ref<i32> +! NSW: %[[I_STEP_2:.*]] = arith.constant 1 : i32 +! NSW: %[[LOOP_VAR_I_NEXT:.*]] = arith.addi %[[LOOP_VAR_I]], %[[I_STEP_2]] overflow<nsw> : i32 +! NSW: fir.store %[[LOOP_VAR_I_NEXT]] to %[[LOOP_VAR_I_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[HEADER]] +! NSW: ^[[EXIT]]: +! NSW: return diff --git a/flang/test/Lower/goto-do-body.f90 b/flang/test/Lower/goto-do-body.f90 index 89e4a7a64a87ba..910e55f1839fd2 100644 --- a/flang/test/Lower/goto-do-body.f90 +++ b/flang/test/Lower/goto-do-body.f90 @@ -48,7 +48,7 @@ subroutine sub1() ! CHECK: fir.store %[[TMP8]] to %[[TRIP]] : !fir.ref<i32> ! CHECK: %[[TMP9:.*]] = fir.load %[[I]] : !fir.ref<i32> ! CHECK: %[[C1_4:.*]] = arith.constant 1 : i32 -! CHECK: %[[TMP10:.*]] = arith.addi %[[TMP9]], %[[C1_4]] overflow<nsw> : i32 +! CHECK: %[[TMP10:.*]] = arith.addi %[[TMP9]], %[[C1_4]] : i32 ! CHECK: fir.store %[[TMP10]] to %[[I]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER]] end do @@ -115,7 +115,7 @@ subroutine sub2() ! CHECK: fir.store %[[TMP10]] to %[[TRIP]] : !fir.ref<i32> ! CHECK: %[[TMP11:.*]] = fir.load %[[I]] : !fir.ref<i32> ! CHECK: %[[STEP_VAL:.*]] = fir.load %[[STEP_VAR]] : !fir.ref<i32> -! CHECK: %[[TMP12:.*]] = arith.addi %[[TMP11]], %[[STEP_VAL]] overflow<nsw> : i32 +! CHECK: %[[TMP12:.*]] = arith.addi %[[TMP11]], %[[STEP_VAL]] : i32 ! CHECK: fir.store %[[TMP12]] to %[[I]] : !fir.ref<i32> ! CHECK: cf.br ^[[HEADER]] end do diff --git a/flang/test/Lower/host-associated.f90 b/flang/test/Lower/host-associated.f90 index 33acdff1bb74cf..9b4269df7bfcb6 100644 --- a/flang/test/Lower/host-associated.f90 +++ b/flang/test/Lower/host-associated.f90 @@ -1,5 +1,5 @@ ! Test internal procedure host association lowering. -! RUN: bbc -hlfir=false -fwrapv %s -o - | FileCheck %s +! RUN: bbc -hlfir=false %s -o - | FileCheck %s ! ----------------------------------------------------------------------------- ! Test non character intrinsic scalars diff --git a/flang/test/Lower/infinite_loop.f90 b/flang/test/Lower/infinite_loop.f90 index de0bee779c5b62..6942dda8d7a23a 100644 --- a/flang/test/Lower/infinite_loop.f90 +++ b/flang/test/Lower/infinite_loop.f90 @@ -1,11 +1,9 @@ ! RUN: bbc -emit-fir -hlfir=false -o - %s | FileCheck %s ! RUN: %flang_fc1 -emit-fir -flang-deprecated-no-hlfir -o - %s | FileCheck %s -! RUN: %flang_fc1 -emit-fir -flang-deprecated-no-hlfir -fwrapv -o - %s | FileCheck %s --check-prefix=NO-NSW +! RUN: %flang_fc1 -emit-fir -flang-deprecated-no-hlfir -flang-experimental-integer-overflow -o - %s | FileCheck %s --check-prefix=NSW ! Tests for infinite loop. -! NO-NSW-NOT: overflow<nsw> - subroutine empty_infinite() do end do @@ -98,10 +96,10 @@ subroutine structured_loop_in_infinite(i) ! CHECK-SAME: %[[C1_INDEX]] to %[[C10_INDEX]] step %[[C1_1]] ! CHECK-SAME: iter_args(%[[J_IV:.*]] = %[[J_LB]]) -> (index, i32) { ! CHECK: fir.store %[[J_IV]] to %[[J_REF]] : !fir.ref<i32> -! CHECK: %[[J_NEXT:.*]] = arith.addi %[[J]], %[[C1_1]] overflow<nsw> : index +! CHECK: %[[J_NEXT:.*]] = arith.addi %[[J]], %[[C1_1]] : index ! CHECK: %[[J_STEPCAST:.*]] = fir.convert %[[C1_1]] : (index) -> i32 ! CHECK: %[[J_IVLOAD:.*]] = fir.load %[[J_REF]] : !fir.ref<i32> -! CHECK: %[[J_IVINC:.*]] = arith.addi %[[J_IVLOAD]], %[[J_STEPCAST]] overflow<nsw> : i32 +! CHECK: %[[J_IVINC:.*]] = arith.addi %[[J_IVLOAD]], %[[J_STEPCAST]] : i32 ! CHECK: fir.result %[[J_NEXT]], %[[J_IVINC]] : index, i32 ! CHECK: } ! CHECK: fir.store %[[J_FINAL]]#1 to %[[J_REF]] : !fir.ref<i32> @@ -109,6 +107,39 @@ subroutine structured_loop_in_infinite(i) ! CHECK: ^[[RETURN]]: ! CHECK: return +! NSW-LABEL: structured_loop_in_infinite +! NSW-SAME: %[[I_REF:.*]]: !fir.ref<i32> +! NSW: %[[J_REF:.*]] = fir.alloca i32 {bindc_name = "j", uniq_name = "_QFstructured_loop_in_infiniteEj"} +! NSW: cf.br ^[[BODY1:.*]] +! NSW: ^[[BODY1]]: +! NSW: %[[I:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> +! NSW: %[[C100:.*]] = arith.constant 100 : i32 +! NSW: %[[COND:.*]] = arith.cmpi sgt, %[[I]], %[[C100]] : i32 +! NSW: cf.cond_br %[[COND]], ^[[EXIT:.*]], ^[[BODY2:.*]] +! NSW: ^[[EXIT]]: +! NSW: cf.br ^[[RETURN:.*]] +! NSW: ^[[BODY2:.*]]: +! NSW: %[[C1:.*]] = arith.constant 1 : i32 +! NSW: %[[C1_INDEX:.*]] = fir.convert %[[C1]] : (i32) -> index +! NSW: %[[C10:.*]] = arith.constant 10 : i32 +! NSW: %[[C10_INDEX:.*]] = fir.convert %[[C10]] : (i32) -> index +! NSW: %[[C1_1:.*]] = arith.constant 1 : index +! NSW: %[[J_LB:.*]] = fir.convert %[[C1_INDEX]] : (index) -> i32 +! NSW: %[[J_FINAL:.*]]:2 = fir.do_loop %[[J:[^ ]*]] = +! NSW-SAME: %[[C1_INDEX]] to %[[C10_INDEX]] step %[[C1_1]] +! NSW-SAME: iter_args(%[[J_IV:.*]] = %[[J_LB]]) -> (index, i32) { +! NSW: fir.store %[[J_IV]] to %[[J_REF]] : !fir.ref<i32> +! NSW: %[[J_NEXT:.*]] = arith.addi %[[J]], %[[C1_1]] overflow<nsw> : index +! NSW: %[[J_STEPCAST:.*]] = fir.convert %[[C1_1]] : (index) -> i32 +! NSW: %[[J_IVLOAD:.*]] = fir.load %[[J_REF]] : !fir.ref<i32> +! NSW: %[[J_IVINC:.*]] = arith.addi %[[J_IVLOAD]], %[[J_STEPCAST]] overflow<nsw> : i32 +! NSW: fir.result %[[J_NEXT]], %[[J_IVINC]] : index, i32 +! NSW: } +! NSW: fir.store %[[J_FINAL]]#1 to %[[J_REF]] : !fir.ref<i32> +! NSW: cf.br ^[[BODY1]] +! NSW: ^[[RETURN]]: +! NSW: return + subroutine empty_infinite_in_while(i) integer :: i do while (i .gt. 50) diff --git a/flang/test/Lower/io-implied-do-fixes.f90 b/flang/test/Lower/io-implied-do-fixes.f90 index cd4fd43e051946..a6c115fa80ded1 100644 --- a/flang/test/Lower/io-implied-do-fixes.f90 +++ b/flang/test/Lower/io-implied-do-fixes.f90 @@ -1,20 +1,30 @@ ! RUN: bbc --use-desc-for-alloc=false -emit-fir -hlfir=false %s -o - | FileCheck %s -! RUN: bbc --use-desc-for-alloc=false -emit-fir -hlfir=false -fwrapv %s -o - | FileCheck %s --check-prefix=NO-NSW +! RUN: bbc --use-desc-for-alloc=false -emit-fir -hlfir=false -integer-overflow %s -o - | FileCheck %s --check-prefix=NSW ! UNSUPPORTED: system-windows -! NO-NSW-NOT: overflow<nsw> - ! CHECK-LABEL: func @_QPido1 ! CHECK: %[[J_REF_ADDR:.*]] = fir.alloca !fir.ptr<i32> {uniq_name = "_QFido1Eiptr.addr"} ! CHECK: %[[J_ADDR:.*]] = fir.load %[[J_REF_ADDR]] : !fir.ref<!fir.ptr<i32>> ! CHECK: %[[J_VAL_FINAL:.*]] = fir.do_loop %[[J_VAL:.*]] = %{{.*}} to %{{.*}} step %{{.*}} -> index { ! CHECK: %[[J_VAL_CVT1:.*]] = fir.convert %[[J_VAL]] : (index) -> i32 ! CHECK: fir.store %[[J_VAL_CVT1]] to %[[J_ADDR]] : !fir.ptr<i32> -! CHECK: %[[J_VAL_NEXT:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} overflow<nsw> : index +! CHECK: %[[J_VAL_NEXT:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} : index ! CHECK: fir.result %[[J_VAL_NEXT]] : index ! CHECK: } ! CHECK: %[[J_VAL_CVT2:.*]] = fir.convert %[[J_VAL_FINAL]] : (index) -> i32 ! CHECK: fir.store %[[J_VAL_CVT2]] to %[[J_ADDR]] : !fir.ptr<i32> + +! NSW-LABEL: func @_QPido1 +! NSW: %[[J_REF_ADDR:.*]] = fir.alloca !fir.ptr<i32> {uniq_name = "_QFido1Eiptr.addr"} +! NSW: %[[J_ADDR:.*]] = fir.load %[[J_REF_ADDR]] : !fir.ref<!fir.ptr<i32>> +! NSW: %[[J_VAL_FINAL:.*]] = fir.do_loop %[[J_VAL:.*]] = %{{.*}} to %{{.*}} step %{{.*}} -> index { +! NSW: %[[J_VAL_CVT1:.*]] = fir.convert %[[J_VAL]] : (index) -> i32 +! NSW: fir.store %[[J_VAL_CVT1]] to %[[J_ADDR]] : !fir.ptr<i32> +! NSW: %[[J_VAL_NEXT:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} overflow<nsw> : index +! NSW: fir.result %[[J_VAL_NEXT]] : index +! NSW: } +! NSW: %[[J_VAL_CVT2:.*]] = fir.convert %[[J_VAL_FINAL]] : (index) -> i32 +! NSW: fir.store %[[J_VAL_CVT2]] to %[[J_ADDR]] : !fir.ptr<i32> subroutine ido1 integer, pointer :: iptr integer, target :: itgt @@ -28,11 +38,23 @@ subroutine ido1 ! CHECK: %[[J_VAL_FINAL:.*]] = fir.do_loop %[[J_VAL:.*]] = %{{.*}} to %{{.*}} step %{{.*}} -> index { ! CHECK: %[[J_VAL_CVT1:.*]] = fir.convert %[[J_VAL]] : (index) -> i32 ! CHECK: fir.store %[[J_VAL_CVT1]] to %[[J_ADDR]] : !fir.heap<i32> -! CHECK: %[[J_VAL_NEXT:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} overflow<nsw> : index +! CHECK: %[[J_VAL_NEXT:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} : index ! CHECK: fir.result %[[J_VAL_NEXT]] : index ! CHECK: } ! CHECK: %[[J_VAL_CVT2:.*]] = fir.convert %[[J_VAL_FINAL]] : (index) -> i32 ! CHECK: fir.store %[[J_VAL_CVT2]] to %[[J_ADDR]] : !fir.heap<i32> + +! NSW-LABEL: func @_QPido2 +! NSW: %[[J_REF_ADDR:.*]] = fir.alloca !fir.heap<i32> {uniq_name = "_QFido2Eiptr.addr"} +! NSW: %[[J_ADDR:.*]] = fir.load %[[J_REF_ADDR]] : !fir.ref<!fir.heap<i32>> +! NSW: %[[J_VAL_FINAL:.*]] = fir.do_loop %[[J_VAL:.*]] = %{{.*}} to %{{.*}} step %{{.*}} -> index { +! NSW: %[[J_VAL_CVT1:.*]] = fir.convert %[[J_VAL]] : (index) -> i32 +! NSW: fir.store %[[J_VAL_CVT1]] to %[[J_ADDR]] : !fir.heap<i32> +! NSW: %[[J_VAL_NEXT:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} overflow<nsw> : index +! NSW: fir.result %[[J_VAL_NEXT]] : index +! NSW: } +! NSW: %[[J_VAL_CVT2:.*]] = fir.convert %[[J_VAL_FINAL]] : (index) -> i32 +! NSW: fir.store %[[J_VAL_CVT2]] to %[[J_ADDR]] : !fir.heap<i32> subroutine ido2 integer, allocatable :: iptr allocate(iptr) @@ -47,12 +69,27 @@ subroutine ido2 ! CHECK: fir.store %[[J_VAL_CVT1]] to %[[J_ADDR]] : !fir.heap<i32> ! CHECK: %[[RES:.*]] = fir.if %[[OK]] -> (i1) { ! CHECK: } -! CHECK: %[[J_VAL_INC:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} overflow<nsw> : index +! CHECK: %[[J_VAL_INC:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} : index ! CHECK: %[[J_VAL_NEXT:.*]] = arith.select %[[RES]], %[[J_VAL_INC]], %[[J_VAL]] : index ! CHECK: fir.result %[[J_VAL_NEXT]], %[[RES]] : index, i1 ! CHECK: } ! CHECK: %[[J_VAL_CVT2:.*]] = fir.convert %[[J_VAL_FINAL]]#0 : (index) -> i32 ! CHECK: fir.store %[[J_VAL_CVT2]] to %[[J_ADDR]] : !fir.heap<i32 + +! NSW-LABEL: func @_QPido3 +! NSW: %[[J_REF_ADDR:.*]] = fir.alloca !fir.heap<i32> {uniq_name = "_QFido3Ej.addr"} +! NSW: %[[J_ADDR:.*]] = fir.load %[[J_REF_ADDR]] : !fir.ref<!fir.heap<i32>> +! NSW: %[[J_VAL_FINAL:.*]]:2 = fir.iterate_while (%[[J_VAL:.*]] = %{{.*}} to %{{.*}} step %{{.*}}) and (%[[OK:.*]] = {{.*}}) -> (index, i1) { +! NSW: %[[J_VAL_CVT1:.*]] = fir.convert %[[J_VAL]] : (index) -> i32 +! NSW: fir.store %[[J_VAL_CVT1]] to %[[J_ADDR]] : !fir.heap<i32> +! NSW: %[[RES:.*]] = fir.if %[[OK]] -> (i1) { +! NSW: } +! NSW: %[[J_VAL_INC:.*]] = arith.addi %[[J_VAL]], %{{[^ ]*}} overflow<nsw> : index +! NSW: %[[J_VAL_NEXT:.*]] = arith.select %[[RES]], %[[J_VAL_INC]], %[[J_VAL]] : index +! NSW: fir.result %[[J_VAL_NEXT]], %[[RES]] : index, i1 +! NSW: } +! NSW: %[[J_VAL_CVT2:.*]] = fir.convert %[[J_VAL_FINAL]]#0 : (index) -> i32 +! NSW: fir.store %[[J_VAL_CVT2]] to %[[J_ADDR]] : !fir.heap<i32 subroutine ido3 integer, allocatable :: j allocate(j) diff --git a/flang/test/Lower/loops2.f90 b/flang/test/Lower/loops2.f90 index 60a6bf6c531197..0a587234a991b6 100644 --- a/flang/test/Lower/loops2.f90 +++ b/flang/test/Lower/loops2.f90 @@ -107,7 +107,7 @@ subroutine test_pointer_unstructured_loop() ! CHECK: ^bb4: ! CHECK: %[[VAL_20:.*]] = fir.load %[[VAL_3]] : !fir.ptr<i32> ! CHECK: %[[VAL_21:.*]] = arith.constant 1 : i32 -! CHECK: %[[VAL_22:.*]] = arith.addi %[[VAL_20]], %[[VAL_21]] overflow<nsw> : i32 +! CHECK: %[[VAL_22:.*]] = arith.addi %[[VAL_20]], %[[VAL_21]] : i32 ! CHECK: fir.store %[[VAL_22]] to %[[VAL_3]] : !fir.ptr<i32> ! CHECK: br ^bb1 ! CHECK: ^bb5: diff --git a/flang/test/Lower/mixed_loops.f90 b/flang/test/Lower/mixed_loops.f90 index 991fd7aa82bb95..1aa0225129bed6 100644 --- a/flang/test/Lower/mixed_loops.f90 +++ b/flang/test/Lower/mixed_loops.f90 @@ -53,7 +53,7 @@ subroutine while_inside_do_loop ! CHECK: fir.store %[[TDEC]] to %[[T_REF]] ! CHECK: %[[I3:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> ! CHECK: %[[C1_2:.*]] = arith.constant 1 : i32 - ! CHECK: %[[IINC:.*]] = arith.addi %[[I3]], %[[C1_2]] overflow<nsw> : i32 + ! CHECK: %[[IINC:.*]] = arith.addi %[[I3]], %[[C1_2]] : i32 ! CHECK: fir.store %[[IINC]] to %[[I_REF]] : !fir.ref<i32> ! CHECK: br ^[[HDR1]] end do @@ -100,10 +100,10 @@ subroutine do_inside_while_loop ! CHECK-DAG: %[[C2:.*]] = arith.constant 2 : i32 ! CHECK: %[[JINC:.*]] = arith.muli %[[C2]], %[[J2]] : i32 ! CHECK: fir.store %[[JINC]] to %[[J_REF]] : !fir.ref<i32> - ! CHECK: %[[IINC:.*]] = arith.addi %[[IDX]], %[[C1]] overflow<nsw> : index + ! CHECK: %[[IINC:.*]] = arith.addi %[[IDX]], %[[C1]] : index ! CHECK: %[[I_STEPCAST:.*]] = fir.convert %[[C1]] : (index) -> i32 ! CHECK: %[[I_IVLOAD:.*]] = fir.load %[[I_REF]] : !fir.ref<i32> - ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] overflow<nsw> : i32 + ! CHECK: %[[I_IVINC:.*]] = arith.addi %[[I_IVLOAD]], %[[I_STEPCAST]] : i32 ! CHECK: fir.result %[[IINC]], %[[I_IVINC]] : index, i32 do i=8,13 j=j*2 diff --git a/flang/test/Lower/vector-subscript-io.f90 b/flang/test/Lower/vector-subscript-io.f90 index 372130fd099074..129e3ee1206c09 100644 --- a/flang/test/Lower/vector-subscript-io.f90 +++ b/flang/test/Lower/vector-subscript-io.f90 @@ -30,7 +30,7 @@ subroutine simple(x, y) ! CHECK: %[[VAL_19:.*]] = fir.array_coor %[[VAL_20]](%[[VAL_10]]) {{\[}}%[[VAL_11]]] %[[VAL_18]] : (!fir.ref<!fir.array<10xi32>>, !fir.shape<1>, !fir.slice<1>, index) -> !fir.ref<i32> ! CHECK: %[[VAL_21:.*]] = fir.convert %[[VAL_19]] : (!fir.ref<i32>) -> !fir.ref<i64> ! CHECK: %[[VAL_22:.*]] = fir.call @_FortranAioInputInteger(%[[VAL_9]], %[[VAL_21]], %[[VAL_3]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i64>, i32) -> i1 -! CHECK: %[[VAL_23:.*]] = arith.addi %[[VAL_12]], %[[VAL_6]] overflow<nsw> : index +! CHECK: %[[VAL_23:.*]] = arith.addi %[[VAL_12]], %[[VAL_6]] : index ! CHECK: %[[VAL_24:.*]] = arith.subi %[[VAL_13]], %[[VAL_6]] : index ! CHECK: cf.br ^bb1(%[[VAL_23]], %[[VAL_24]] : index, index) ! CHECK: ^bb3: @@ -79,7 +79,7 @@ integer function get_substcript() ! CHECK: %[[VAL_49:.*]] = fir.convert %[[VAL_48]] : (i32) -> index ! CHECK: %[[VAL_50:.*]] = fir.array_coor %[[VAL_51]] {{\[}}%[[VAL_42]]] %[[VAL_46]], %[[VAL_49]] : (!fir.box<!fir.array<?x?xf32>>, !fir.slice<2>, index, index) -> !fir.ref<f32> ! CHECK: %[[VAL_52:.*]] = fir.call @_FortranAioInputReal32(%[[VAL_34]], %[[VAL_50]]) {{.*}}: (!fir.ref<i8>, !fir.ref<f32>) -> i1 -! CHECK: %[[VAL_53:.*]] = arith.addi %[[VAL_43]], %[[VAL_30]] overflow<nsw> : index +! CHECK: %[[VAL_53:.*]] = arith.addi %[[VAL_43]], %[[VAL_30]] : index ! CHECK: %[[VAL_54:.*]] = arith.subi %[[VAL_44]], %[[VAL_30]] : index ! CHECK: cf.br ^bb1(%[[VAL_53]], %[[VAL_54]] : index, index) ! CHECK: ^bb3: @@ -122,7 +122,7 @@ subroutine with_assumed_shapes(x, y) ! CHECK: %[[VAL_77:.*]] = fir.array_coor %[[VAL_78]] {{\[}}%[[VAL_70]]] %[[VAL_76]] : (!fir.box<!fir.array<?xi32>>, !fir.slice<1>, index) -> !fir.ref<i32> ! CHECK: %[[VAL_79:.*]] = fir.convert %[[VAL_77]] : (!fir.ref<i32>) -> !fir.ref<i64> ! CHECK: %[[VAL_80:.*]] = fir.call @_FortranAioInputInteger(%[[VAL_67]], %[[VAL_79]], %[[VAL_62]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i64>, i32) -> i1 -! CHECK: %[[VAL_81:.*]] = arith.addi %[[VAL_71]], %[[VAL_64]] overflow<nsw> : index +! CHECK: %[[VAL_81:.*]] = arith.addi %[[VAL_71]], %[[VAL_64]] : index ! CHECK: %[[VAL_82:.*]] = arith.subi %[[VAL_72]], %[[VAL_64]] : index ! CHECK: cf.br ^bb1(%[[VAL_81]], %[[VAL_82]] : index, index) ! CHECK: ^bb3: @@ -162,7 +162,7 @@ subroutine lower_bounds(x, y) ! CHECK: %[[VAL_107:.*]] = fir.array_coor %[[VAL_108]](%[[VAL_97]]) {{\[}}%[[VAL_99]]] %[[VAL_91]], %[[VAL_106]] : (!fir.ref<!fir.array<4x6xi32>>, !fir.shapeshift<2>, !fir.slice<2>, index, index) -> !fir.ref<i32> ! CHECK: %[[VAL_109:.*]] = fir.convert %[[VAL_107]] : (!fir.ref<i32>) -> !fir.ref<i64> ! CHECK: %[[VAL_110:.*]] = fir.call @_FortranAioInputInteger(%[[VAL_96]], %[[VAL_109]], %[[VAL_90]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i64>, i32) -> i1 -! CHECK: %[[VAL_111:.*]] = arith.addi %[[VAL_100]], %[[VAL_93]] overflow<nsw> : index +! CHECK: %[[VAL_111:.*]] = arith.addi %[[VAL_100]], %[[VAL_93]] : index ! CHECK: %[[VAL_112:.*]] = arith.subi %[[VAL_101]], %[[VAL_93]] : index ! CHECK: cf.br ^bb1(%[[VAL_111]], %[[VAL_112]] : index, index) ! CHECK: ^bb3: @@ -202,11 +202,11 @@ subroutine two_vectors(x, y1, y2) ! CHECK: %[[VAL_138:.*]] = fir.convert %[[VAL_137]] : (i32) -> index ! CHECK: %[[VAL_139:.*]] = fir.array_coor %[[VAL_140]](%[[VAL_123]]) {{\[}}%[[VAL_124]]] %[[VAL_134]], %[[VAL_138]] : (!fir.ref<!fir.array<4x4xf32>>, !fir.shape<2>, !fir.slice<2>, index, index) -> !fir.ref<f32> ! CHECK: %[[VAL_141:.*]] = fir.call @_FortranAioInputReal32(%[[VAL_122]], %[[VAL_139]]) {{.*}}: (!fir.ref<i8>, !fir.ref<f32>) -> i1 -! CHECK: %[[VAL_142:.*]] = arith.addi %[[VAL_128]], %[[VAL_119]] overflow<nsw> : index +! CHECK: %[[VAL_142:.*]] = arith.addi %[[VAL_128]], %[[VAL_119]] : index ! CHECK: %[[VAL_143:.*]] = arith.subi %[[VAL_129]], %[[VAL_119]] : index ! CHECK: cf.br ^bb2(%[[VAL_142]], %[[VAL_143]] : index, index) ! CHECK: ^bb4: -! CHECK: %[[VAL_144:.*]] = arith.addi %[[VAL_125]], %[[VAL_119]] overflow<nsw> : index +! CHECK: %[[VAL_144:.*]] = arith.addi %[[VAL_125]], %[[VAL_119]] : index ! CHECK: %[[VAL_145:.*]] = arith.subi %[[VAL_126]], %[[VAL_119]] : index ! CHECK: cf.br ^bb1(%[[VAL_144]], %[[VAL_145]] : index, index) ! CHECK: ^bb5: @@ -245,11 +245,11 @@ subroutine triplets_and_vector(x, y) ! CHECK: %[[VAL_169:.*]] = fir.array_coor %[[VAL_170]](%[[VAL_157]]) {{\[}}%[[VAL_158]]] %[[VAL_162]], %[[VAL_168]] : (!fir.ref<!fir.array<4x4xcomplex<f32>>>, !fir.shape<2>, !fir.slice<2>, index, index) -> !fir.ref<complex<f32>> ! CHECK: %[[VAL_171:.*]] = fir.convert %[[VAL_169]] : (!fir.ref<complex<f32>>) -> !fir.ref<f32> ! CHECK: %[[VAL_172:.*]] = fir.call @_FortranAioInputComplex32(%[[VAL_156]], %[[VAL_171]]) {{.*}}: (!fir.ref<i8>, !fir.ref<f32>) -> i1 -! CHECK: %[[VAL_173:.*]] = arith.addi %[[VAL_162]], %[[VAL_153]] overflow<nsw> : index +! CHECK: %[[VAL_173:.*]] = arith.addi %[[VAL_162]], %[[VAL_153]] : index ! CHECK: %[[VAL_174:.*]] = arith.subi %[[VAL_163]], %[[VAL_153]] : index ! CHECK: cf.br ^bb2(%[[VAL_173]], %[[VAL_174]] : index, index) ! CHECK: ^bb4: -! CHECK: %[[VAL_175:.*]] = arith.addi %[[VAL_159]], %[[VAL_153]] overflow<nsw> : index +! CHECK: %[[VAL_175:.*]] = arith.addi %[[VAL_159]], %[[VAL_153]] : index ! CHECK: %[[VAL_176:.*]] = arith.subi %[[VAL_160]], %[[VAL_153]] : index ! CHECK: cf.br ^bb1(%[[VAL_175]], %[[VAL_176]] : index, index) ! CHECK: ^bb5: @@ -287,7 +287,7 @@ subroutine simple_char(x, y) ! CHECK: %[[VAL_200:.*]] = fir.convert %[[VAL_199]] : (!fir.ref<!fir.char<1,?>>) -> !fir.ref<i8> ! CHECK: %[[VAL_201:.*]] = fir.convert %[[VAL_184]]#1 : (index) -> i64 ! CHECK: %[[VAL_202:.*]] = fir.call @_FortranAioInputAscii(%[[VAL_189]], %[[VAL_200]], %[[VAL_201]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i8>, i64) -> i1 -! CHECK: %[[VAL_203:.*]] = arith.addi %[[VAL_192]], %[[VAL_183]] overflow<nsw> : index +! CHECK: %[[VAL_203:.*]] = arith.addi %[[VAL_192]], %[[VAL_183]] : index ! CHECK: %[[VAL_204:.*]] = arith.subi %[[VAL_193]], %[[VAL_183]] : index ! CHECK: cf.br ^bb1(%[[VAL_203]], %[[VAL_204]] : index, index) ! CHECK: ^bb3: @@ -333,7 +333,7 @@ subroutine substring(x, y, i, j) ! CHECK: %[[VAL_238:.*]] = fir.convert %[[VAL_233]] : (!fir.ref<!fir.char<1,?>>) -> !fir.ref<i8> ! CHECK: %[[VAL_239:.*]] = fir.convert %[[VAL_237]] : (index) -> i64 ! CHECK: %[[VAL_240:.*]] = fir.call @_FortranAioInputAscii(%[[VAL_213]], %[[VAL_238]], %[[VAL_239]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i8>, i64) -> i1 -! CHECK: %[[VAL_241:.*]] = arith.addi %[[VAL_221]], %[[VAL_210]] overflow<nsw> : index +! CHECK: %[[VAL_241:.*]] = arith.addi %[[VAL_221]], %[[VAL_210]] : index ! CHECK: %[[VAL_242:.*]] = arith.subi %[[VAL_222]], %[[VAL_210]] : index ! CHECK: cf.br ^bb1(%[[VAL_241]], %[[VAL_242]] : index, index) ! CHECK: ^bb3: @@ -366,7 +366,7 @@ subroutine complex_part(z, y) ! CHECK: %[[VAL_260:.*]] = fir.convert %[[VAL_259]] : (i32) -> index ! CHECK: %[[VAL_261:.*]] = fir.array_coor %[[VAL_262]] {{\[}}%[[VAL_254]]] %[[VAL_260]] : (!fir.box<!fir.array<?xcomplex<f32>>>, !fir.slice<1>, index) -> !fir.ref<f32> ! CHECK: %[[VAL_263:.*]] = fir.call @_FortranAioInputReal32(%[[VAL_251]], %[[VAL_261]]) {{.*}}: (!fir.ref<i8>, !fir.ref<f32>) -> i1 -! CHECK: %[[VAL_264:.*]] = arith.addi %[[VAL_255]], %[[VAL_248]] overflow<nsw> : index +! CHECK: %[[VAL_264:.*]] = arith.addi %[[VAL_255]], %[[VAL_248]] : index ! CHECK: %[[VAL_265:.*]] = arith.subi %[[VAL_256]], %[[VAL_248]] : index ! CHECK: cf.br ^bb1(%[[VAL_264]], %[[VAL_265]] : index, index) ! CHECK: ^bb3: @@ -414,7 +414,7 @@ subroutine simple_derived(x, y) ! CHECK: %[[VAL_288:.*]] = fir.embox %[[VAL_286]] : (!fir.ref<!fir.type<_QMderived_typesTt{i:i32,c:!fir.char<1,2>}>>) -> !fir.box<!fir.type<_QMderived_typesTt{i:i32,c:!fir.char<1,2>}>> ! CHECK: %[[VAL_289:.*]] = fir.convert %[[VAL_288]] : (!fir.box<!fir.type<_QMderived_typesTt{i:i32,c:!fir.char<1,2>}>>) -> !fir.box<none> ! CHECK: %[[VAL_290:.*]] = fir.call @_FortranAioInputDerivedType(%[[VAL_276]], %[[VAL_289]], {{.*}}) {{.*}}: (!fir.ref<i8>, !fir.box<none>, !fir.ref<none>) -> i1 -! CHECK: %[[VAL_291:.*]] = arith.addi %[[VAL_279]], %[[VAL_273]] overflow<nsw> : index +! CHECK: %[[VAL_291:.*]] = arith.addi %[[VAL_279]], %[[VAL_273]] : index ! CHECK: %[[VAL_292:.*]] = arith.subi %[[VAL_280]], %[[VAL_273]] : index ! CHECK: cf.br ^bb1(%[[VAL_291]], %[[VAL_292]] : index, index) ! CHECK: ^bb3: @@ -463,11 +463,11 @@ subroutine with_path(b, i) ! CHECK: %[[VAL_325:.*]] = fir.array_coor %[[VAL_326:.*]](%[[VAL_313]]) {{\[}}%[[VAL_315]]] %[[VAL_301]], %[[VAL_324]], %[[VAL_316]] : (!fir.box<!fir.array<?x?x?x!fir.type<_QMderived_typesTt2{a:!fir.array<5x5x!fir.type<_QMderived_typesTt{i:i32,c:!fir.char<1,2>}>>}>>>, !fir.shift<3>, !fir.slice<3>, index, index, index) -> !fir.ref<i32> ! CHECK: %[[VAL_327:.*]] = fir.convert %[[VAL_325]] : (!fir.ref<i32>) -> !fir.ref<i64> ! CHECK: %[[VAL_328:.*]] = fir.call @_FortranAioInputInteger(%[[VAL_308]], %[[VAL_327]], %[[VAL_302]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i64>, i32) -> i1 -! CHECK: %[[VAL_329:.*]] = arith.addi %[[VAL_319]], %[[VAL_305]] overflow<nsw> : index +! CHECK: %[[VAL_329:.*]] = arith.addi %[[VAL_319]], %[[VAL_305]] : index ! CHECK: %[[VAL_330:.*]] = arith.subi %[[VAL_320]], %[[VAL_305]] : index ! CHECK: cf.br ^bb2(%[[VAL_329]], %[[VAL_330]] : index, index) ! CHECK: ^bb4: -! CHECK: %[[VAL_331:.*]] = arith.addi %[[VAL_316]], %[[VAL_305]] overflow<nsw> : index +! CHECK: %[[VAL_331:.*]] = arith.addi %[[VAL_316]], %[[VAL_305]] : index ! CHECK: %[[VAL_332:.*]] = arith.subi %[[VAL_317]], %[[VAL_305]] : index ! CHECK: cf.br ^bb1(%[[VAL_331]], %[[VAL_332]] : index, index) ! CHECK: ^bb5: @@ -505,7 +505,7 @@ subroutine simple_iostat(x, y, j, stat) ! CHECK: %[[VAL_355:.*]] = fir.convert %[[VAL_354]] : (i32) -> index ! CHECK: %[[VAL_356:.*]] = fir.array_coor %[[VAL_357]] {{\[}}%[[VAL_347]]] %[[VAL_355]] : (!fir.box<!fir.array<?xf32>>, !fir.slice<1>, index) -> !fir.ref<f32> ! CHECK: %[[VAL_358:.*]] = fir.call @_FortranAioInputReal32(%[[VAL_343]], %[[VAL_356]]) {{.*}}: (!fir.ref<i8>, !fir.ref<f32>) -> i1 -! CHECK: %[[VAL_359:.*]] = arith.addi %[[VAL_349]], %[[VAL_338]] overflow<nsw> : index +! CHECK: %[[VAL_359:.*]] = arith.addi %[[VAL_349]], %[[VAL_338]] : index ! CHECK: cf.br ^bb1(%[[VAL_359]], %[[VAL_358]] : index, i1) ! CHECK: ^bb3: ! CHECK: cf.cond_br %[[VAL_350]], ^bb4, ^bb5 @@ -568,10 +568,10 @@ subroutine iostat_in_io_loop(k, j, stat) ! CHECK: %[[VAL_399:.*]] = fir.array_coor %[[VAL_400]](%[[VAL_387]]) {{\[}}%[[VAL_389]]] %[[VAL_394]], %[[VAL_398]] : (!fir.ref<!fir.array<3x5xi32>>, !fir.shape<2>, !fir.slice<2>, index, index) -> !fir.ref<i32> ! CHECK: %[[VAL_401:.*]] = fir.convert %[[VAL_399]] : (!fir.ref<i32>) -> !fir.ref<i64> ! CHECK: %[[VAL_402:.*]] = fir.call @_FortranAioInputInteger(%[[VAL_378]], %[[VAL_401]], %[[VAL_374]]) {{.*}}: (!fir.ref<i8>, !fir.ref<i64>, i32) -> i1 -! CHECK: %[[VAL_403:.*]] = arith.addi %[[VAL_390]], %[[VAL_371]] overflow<nsw> : index +! CHECK: %[[VAL_403:.*]] = arith.addi %[[VAL_390]], %[[VAL_371]] : index ! CHECK: cf.br ^bb4(%[[VAL_403]], %[[VAL_402]] : index, i1) ! CHECK: ^bb6(%[[VAL_404:.*]]: i1): -! CHECK: %[[VAL_405:.*]] = arith.addi %[[VAL_380]], %[[VAL_371]] overflow<nsw> : index +! CHECK: %[[VAL_405:.*]] = arith.addi %[[VAL_380]], %[[VAL_371]] : index ! CHECK: cf.br ^bb1(%[[VAL_405]], %[[VAL_404]] : index, i1) ! CHECK: ^bb7: ! CHECK: %[[VAL_406:.*]] = fir.convert %[[VAL_380]] : (index) -> i32 diff --git a/flang/tools/bbc/bbc.cpp b/flang/tools/bbc/bbc.cpp index af38bf272b02b1..fe5e36f704c76c 100644 --- a/flang/tools/bbc/bbc.cpp +++ b/flang/tools/bbc/bbc.cpp @@ -234,6 +234,12 @@ static llvm::cl::opt<bool> integerWrapAround( llvm::cl::desc("Treat signed integer overflow as two's complement"), llvm::cl::init(false)); +// TODO: integrate this option with the above +static llvm::cl::opt<bool> + setNSW("integer-overflow", + llvm::cl::desc("add nsw flag to internal operations"), + llvm::cl::init(false)); + #define FLANG_EXCLUDE_CODEGEN #include "flang/Optimizer/Passes/CommandLineOpts.h" #include "flang/Optimizer/Passes/Pipelines.h" @@ -375,6 +381,7 @@ static llvm::LogicalResult convertFortranSourceToMLIR( loweringOptions.setNoPPCNativeVecElemOrder(enableNoPPCNativeVecElemOrder); loweringOptions.setLowerToHighLevelFIR(useHLFIR || emitHLFIR); loweringOptions.setIntegerWrapAround(integerWrapAround); + loweringOptions.setNSWOnLoopVarInc(setNSW); std::vector<Fortran::lower::EnvironmentDefault> envDefaults = {}; Fortran::frontend::TargetOptions targetOpts; Fortran::frontend::CodeGenOptions cgOpts; @@ -460,7 +467,7 @@ static llvm::LogicalResult convertFortranSourceToMLIR( // Add O2 optimizer pass pipeline. MLIRToLLVMPassPipelineConfig config(llvm::OptimizationLevel::O2); - config.NSWOnLoopVarInc = !integerWrapAround; + config.NSWOnLoopVarInc = setNSW; fir::registerDefaultInlinerPass(config); fir::createDefaultFIROptimizerPassPipeline(pm, config); } _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
