[flang][driver] add command line arguments for alias tags pass

The ultimate intention is to have this pass enabled by default whenever
we are optimizing for speed. But for now, just add the arguments so this
can be more easily tested.

PR: #68595

Author: tblah

clang/include/clang/Driver/Options.td

@@ -6264,6 +6264,9 @@ defm stack_arrays : BoolOptionWithoutMarshalling<"f", "stack-arrays",
 defm loop_versioning : BoolOptionWithoutMarshalling<"f", "version-loops-for-stride",
   PosFlag<SetTrue, [], [ClangOption], "Create unit-strided versions of loops">,
    NegFlag<SetFalse, [], [ClangOption], "Do not create unit-strided loops (default)">>;
+defm alias_analysis : BoolOptionWithoutMarshalling<"f", "alias-analysis",
+  PosFlag<SetTrue, [], [], "Pass alias information on to LLVM (default when optimizing for speed)">,
+  NegFlag<SetFalse, [], [], "Do not pass alias information on to LLVM (default for unoptimized builds)">>;
 } // let Visibility = [FC1Option, FlangOption]
 
 def J : JoinedOrSeparate<["-"], "J">,

clang/lib/Driver/ToolChains/Flang.cpp

@@ -145,7 +145,9 @@ void Flang::addCodegenOptions(const ArgList &Args,
   Args.addAllArgs(CmdArgs, {options::OPT_flang_experimental_hlfir,
                             options::OPT_flang_experimental_polymorphism,
                             options::OPT_fno_ppc_native_vec_elem_order,
-                            options::OPT_fppc_native_vec_elem_order});
+                            options::OPT_fppc_native_vec_elem_order,
+                            options::OPT_falias_analysis,
+                            options::OPT_fno_alias_analysis});
 }
 
 void Flang::addPicOptions(const ArgList &Args, ArgStringList &CmdArgs) const {

flang/include/flang/Frontend/CodeGenOptions.def

@@ -32,6 +32,7 @@ CODEGENOPT(PrepareForThinLTO , 1, 0) ///< Set when -flto=thin is enabled on the
                                      ///< compile step.
 CODEGENOPT(StackArrays, 1, 0) ///< -fstack-arrays (enable the stack-arrays pass)
 CODEGENOPT(LoopVersioning, 1, 0) ///< Enable loop versioning.
+CODEGENOPT(AliasAnalysis, 1, 0) ///< Enable alias analysis pass
 
 CODEGENOPT(Underscoring, 1, 1)
 ENUM_CODEGENOPT(RelocationModel, llvm::Reloc::Model, 3, llvm::Reloc::PIC_) ///< Name of the relocation model to use.

flang/include/flang/Optimizer/CodeGen/CodeGen.h

@@ -54,6 +54,9 @@ struct FIRToLLVMPassOptions {
 
   // Generate TBAA information for FIR types and memory accessing operations.
   bool applyTBAA = false;
+
+  // Force the usage of a unified tbaa tree in TBAABuilder.
+  bool forceUnifiedTBAATree = false;
 };
 
 /// Convert FIR to the LLVM IR dialect with default options.

flang/include/flang/Optimizer/CodeGen/TypeConverter.h

@@ -45,7 +45,8 @@ namespace fir {
 /// This converts FIR types to LLVM types (for now)
 class LLVMTypeConverter : public mlir::LLVMTypeConverter {
 public:
-  LLVMTypeConverter(mlir::ModuleOp module, bool applyTBAA);
+  LLVMTypeConverter(mlir::ModuleOp module, bool applyTBAA,
+                    bool forceUnifiedTBAATree);
 
   // i32 is used here because LLVM wants i32 constants when indexing into struct
   // types. Indexing into other aggregate types is more flexible.

flang/include/flang/Tools/CLOptions.inc

@@ -65,6 +65,12 @@ DisableOption(FirAvc, "avc", "array value copy analysis and transformation");
 DisableOption(
     FirMao, "memory-allocation-opt", "memory allocation optimization");
 
+DisableOption(FirAliasTags, "fir-alias-tags", "fir alias analysis");
+static llvm::cl::opt<bool> useOldAliasTags("use-old-alias-tags",
+    llvm::cl::desc("Use a single TBAA tree for all functions and do not use "
+                   "the FIR alias tags pass"),
+    llvm::cl::init(false), llvm::cl::Hidden);
+
 /// CodeGen Passes
 #if !defined(FLANG_EXCLUDE_CODEGEN)
 DisableOption(CodeGenRewrite, "codegen-rewrite", "rewrite FIR for codegen");
@@ -156,6 +162,7 @@ inline void addFIRToLLVMPass(
   fir::FIRToLLVMPassOptions options;
   options.ignoreMissingTypeDescriptors = ignoreMissingTypeDescriptors;
   options.applyTBAA = optLevel.isOptimizingForSpeed();
+  options.forceUnifiedTBAATree = useOldAliasTags;
   addPassConditionally(pm, disableFirToLlvmIr,
       [&]() { return fir::createFIRToLLVMPass(options); });
 }
@@ -220,6 +227,9 @@ inline void createDefaultFIROptimizerPassPipeline(
   // Polymorphic types
   pm.addPass(fir::createPolymorphicOpConversionPass());
 
+  if (pc.AliasAnalysis && !disableFirAliasTags && !useOldAliasTags)
+    pm.addPass(fir::createAliasTagsPass());
+
   // convert control flow to CFG form
   fir::addCfgConversionPass(pm);
   pm.addPass(mlir::createConvertSCFToCFPass());

flang/include/flang/Tools/CrossToolHelpers.h

@@ -34,12 +34,14 @@ struct MLIRToLLVMPassPipelineConfig {
     Underscoring = opts.Underscoring;
     LoopVersioning = opts.LoopVersioning;
     DebugInfo = opts.getDebugInfo();
+    AliasAnalysis = opts.AliasAnalysis;
   }
 
   llvm::OptimizationLevel OptLevel; ///< optimisation level
   bool StackArrays = false; ///< convert memory allocations to alloca.
   bool Underscoring = true; ///< add underscores to function names.
   bool LoopVersioning = false; ///< Run the version loop pass.
+  bool AliasAnalysis = false; ///< Add TBAA tags to generated LLVMIR
   llvm::codegenoptions::DebugInfoKind DebugInfo =
       llvm::codegenoptions::NoDebugInfo; ///< Debug info generation.
   unsigned VScaleMin = 0; ///< SVE vector range minimum.

flang/lib/Frontend/CompilerInvocation.cpp

@@ -214,6 +214,11 @@ static void parseCodeGenArgs(Fortran::frontend::CodeGenOptions &opts,
                    clang::driver::options::OPT_fno_loop_versioning, false))
     opts.LoopVersioning = 1;
 
+  opts.AliasAnalysis =
+      args.hasFlag(clang::driver::options::OPT_falias_analysis,
+                   clang::driver::options::OPT_fno_alias_analysis,
+                   /*default=*/false);
+
   for (auto *a : args.filtered(clang::driver::options::OPT_fpass_plugin_EQ))
     opts.LLVMPassPlugins.push_back(a->getValue());
 

flang/lib/Optimizer/CodeGen/CodeGen.cpp

@@ -3787,7 +3787,8 @@ class FIRToLLVMLowering
 
     auto *context = getModule().getContext();
     fir::LLVMTypeConverter typeConverter{getModule(),
-                                         options.applyTBAA || applyTBAA};
+                                         options.applyTBAA || applyTBAA,
+                                         options.forceUnifiedTBAATree};
     mlir::RewritePatternSet pattern(context);
     pattern.insert<
         AbsentOpConversion, AddcOpConversion, AddrOfOpConversion,

flang/lib/Optimizer/CodeGen/TypeConverter.cpp

@@ -26,7 +26,8 @@
 
 namespace fir {
 
-LLVMTypeConverter::LLVMTypeConverter(mlir::ModuleOp module, bool applyTBAA)
+LLVMTypeConverter::LLVMTypeConverter(mlir::ModuleOp module, bool applyTBAA,
+                                     bool forceUnifiedTBAATree)
     : mlir::LLVMTypeConverter(module.getContext(),
                               [&] {
                                 mlir::LowerToLLVMOptions options(
@@ -38,8 +39,8 @@ LLVMTypeConverter::LLVMTypeConverter(mlir::ModuleOp module, bool applyTBAA)
       specifics(CodeGenSpecifics::get(module.getContext(),
                                       getTargetTriple(module),
                                       getKindMapping(module))),
-      tbaaBuilder(
-          std::make_unique<TBAABuilder>(module->getContext(), applyTBAA)) {
+      tbaaBuilder(std::make_unique<TBAABuilder>(module->getContext(), applyTBAA,
+                                                forceUnifiedTBAATree)) {
   LLVM_DEBUG(llvm::dbgs() << "FIR type converter\n");
 
   // Each conversion should return a value of type mlir::Type.

flang/test/Driver/driver-help-hidden.f90

@@ -24,6 +24,7 @@
 ! CHECK-NEXT: -D <macro>=<value>      Define <macro> to <value> (or 1 if <value> omitted)
 ! CHECK-NEXT: -emit-llvm              Use the LLVM representation for assembler and object files
 ! CHECK-NEXT: -E                      Only run the preprocessor
+! CHECK-NEXT: -falias-analysis        Pass alias information on to LLVM (default when optimizing for speed)
 ! CHECK-NEXT: -falternative-parameter-statement
 ! CHECK-NEXT:                         Enable the old style PARAMETER statement
 ! CHECK-NEXT: -fapprox-func           Allow certain math function calls to be replaced with an approximately equivalent calculation
@@ -56,6 +57,7 @@
 ! CHECK-NEXT: -flto=jobserver         Enable LTO in 'full' mode
 ! CHECK-NEXT: -flto=<value>           Set LTO mode
 ! CHECK-NEXT: -flto                   Enable LTO in 'full' mode
+! CHECK-NEXT: -fno-alias-analysis     Do not pass alias information on to LLVM (default for unoptimized builds)
 ! CHECK-NEXT: -fno-automatic          Implies the SAVE attribute for non-automatic local objects in subprograms unless RECURSIVE
 ! CHECK-NEXT: -fno-color-diagnostics  Disable colors in diagnostics
 ! CHECK-NEXT: -fno-integrated-as      Disable the integrated assembler

flang/test/Driver/driver-help.f90

@@ -20,6 +20,7 @@
 ! HELP-NEXT: -D <macro>=<value>      Define <macro> to <value> (or 1 if <value> omitted)
 ! HELP-NEXT: -emit-llvm              Use the LLVM representation for assembler and object files
 ! HELP-NEXT: -E                      Only run the preprocessor
+! HELP-NEXT: -falias-analysis        Pass alias information on to LLVM (default when optimizing for speed)
 ! HELP-NEXT: -falternative-parameter-statement
 ! HELP-NEXT:                         Enable the old style PARAMETER statement
 ! HELP-NEXT: -fapprox-func           Allow certain math function calls to be replaced with an approximately equivalent calculation
@@ -48,6 +49,7 @@
 ! HELP-NEXT: -flto=jobserver         Enable LTO in 'full' mode
 ! HELP-NEXT: -flto=<value>           Set LTO mode
 ! HELP-NEXT: -flto                   Enable LTO in 'full' mode
+! HELP-NEXT: -fno-alias-analysis     Do not pass alias information on to LLVM (default for unoptimized builds)
 ! HELP-NEXT: -fno-automatic          Implies the SAVE attribute for non-automatic local objects in subprograms unless RECURSIVE
 ! HELP-NEXT: -fno-color-diagnostics  Disable colors in diagnostics
 ! HELP-NEXT: -fno-integrated-as      Disable the integrated assembler
@@ -141,6 +143,7 @@
 ! HELP-FC1-NEXT: -emit-llvm              Use the LLVM representation for assembler and object files
 ! HELP-FC1-NEXT: -emit-obj               Emit native object files
 ! HELP-FC1-NEXT: -E                      Only run the preprocessor
+! HELP-FC1-NEXT: -falias-analysis        Pass alias information on to LLVM (default when optimizing for speed)
 ! HELP-FC1-NEXT: -falternative-parameter-statement
 ! HELP-FC1-NEXT:                         Enable the old style PARAMETER statement
 ! HELP-FC1-NEXT: -fapprox-func           Allow certain math function calls to be replaced with an approximately equivalent calculation
@@ -187,6 +190,7 @@
 ! HELP-FC1-NEXT: -flogical-abbreviations Enable logical abbreviations
 ! HELP-FC1-NEXT: -flto=<value>           Set LTO mode
 ! HELP-FC1-NEXT: -flto                   Enable LTO in 'full' mode
+! HELP-FC1-NEXT: -fno-alias-analysis     Do not pass alias information on to LLVM (default for unoptimized builds)
 ! HELP-FC1-NEXT: -fno-analyzed-objects-for-unparse
 ! HELP-FC1-NEXT:                         Do not use the analyzed objects when unparsing
 ! HELP-FC1-NEXT: -fno-automatic          Implies the SAVE attribute for non-automatic local objects in subprograms unless RECURSIVE

flang/test/Driver/falias-analysis.f90

@@ -0,0 +1,21 @@
+! Check that -falias-analysis and -fno-alias-analysis work as expected
+! See flang/test/Fir/tbaa-codegen.fir for a test that the output is correct
+
+! RUN: %flang -c -emit-llvm -falias-analysis %s -o - | llvm-dis | FileCheck %s --check-prefix=CHECK-AA --check-prefix=CHECK-ALL
+! RUN: %flang -c -emit-llvm -falias-analysis -fno-alias-analysis %s -o - | llvm-dis | FileCheck %s --check-prefix=CHECK-NOAA --check-prefix=CHECK-ALL
+! RUN: %flang -c -emit-llvm %s -o - | llvm-dis | FileCheck %s --check-prefix=CHECK-NOAA --check-prefix=CHECK-ALL
+
+! RUN: %flang -fc1 -emit-llvm -falias-analysis %s -o - | FileCheck %s --check-prefix=CHECK-AA --check-prefix=CHECK-ALL
+! RUN: %flang -fc1 -emit-llvm -falias-analysis -fno-alias-analysis %s -o - | FileCheck %s --check-prefix=CHECK-NOAA --check-prefix=CHECK-ALL
+! RUN: %flang -fc1 -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK-NOAA --check-prefix=CHECK-ALL
+
+subroutine simple(a)
+  integer, intent(inout) :: a(:)
+  a(1) = a(2)
+end subroutine
+! CHECK-ALL-LABEL: define void @simple
+! CHECK-ALL:       ret
+! CHECK-ALL:     }
+
+! CHECK-AA: ![[ROOT:.*]] = !{!"Flang function root _QPsimple"}
+! CHECK-NOAA-NOT: ![[ROOT:.*]] = !{!"Flang function root _QPsimple"}

flang/test/Fir/tbaa-codegen.fir

@@ -1,4 +1,4 @@
-// test that tbaa attributes can be added to fir.load and fir.store
+// Test that tbaa attributes can be added to fir.load and fir.store
 // and that these attributes are propagated to LLVMIR
 
 // RUN: tco %s | FileCheck %s