Merge branch 'main' into python-limited-api-typemaps

Author: JDevlieghere

clang/lib/CIR/CodeGen/CIRGenBuilder.h

@@ -315,8 +315,10 @@ class CIRGenBuilderTy : public cir::CIRBaseBuilderTy {
     return getConstantInt(loc, getUInt32Ty(), c);
   }
   cir::ConstantOp getSInt64(uint64_t c, mlir::Location loc) {
-    cir::IntType sInt64Ty = getSInt64Ty();
-    return cir::ConstantOp::create(*this, loc, cir::IntAttr::get(sInt64Ty, c));
+    return getConstantInt(loc, getSInt64Ty(), c);
+  }
+  cir::ConstantOp getUInt64(uint64_t c, mlir::Location loc) {
+    return getConstantInt(loc, getUInt64Ty(), c);
   }
 
   mlir::Value createNeg(mlir::Value value) {

clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp

@@ -630,6 +630,22 @@ CIRGenFunction::emitTargetBuiltinExpr(unsigned builtinID, const CallExpr *e,
                                    getTarget().getTriple().getArch());
 }
 
+mlir::Value CIRGenFunction::emitScalarOrConstFoldImmArg(
+    const unsigned iceArguments, const unsigned idx, const Expr *argExpr) {
+  mlir::Value arg = {};
+  if ((iceArguments & (1 << idx)) == 0) {
+    arg = emitScalarExpr(argExpr);
+  } else {
+    // If this is required to be a constant, constant fold it so that we
+    // know that the generated intrinsic gets a ConstantInt.
+    const std::optional<llvm::APSInt> result =
+        argExpr->getIntegerConstantExpr(getContext());
+    assert(result && "Expected argument to be a constant");
+    arg = builder.getConstInt(getLoc(argExpr->getSourceRange()), *result);
+  }
+  return arg;
+}
+
 /// Given a builtin id for a function like "__builtin_fabsf", return a Function*
 /// for "fabsf".
 cir::FuncOp CIRGenModule::getBuiltinLibFunction(const FunctionDecl *fd,

clang/lib/CIR/CodeGen/CIRGenBuiltinX86.cpp

@@ -16,7 +16,6 @@
 #include "clang/Basic/Builtins.h"
 #include "clang/Basic/TargetBuiltins.h"
 #include "clang/CIR/MissingFeatures.h"
-#include "llvm/IR/IntrinsicsX86.h"
 
 using namespace clang;
 using namespace clang::CIRGen;
@@ -66,9 +65,8 @@ mlir::Value CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID,
   getContext().GetBuiltinType(builtinID, error, &iceArguments);
   assert(error == ASTContext::GE_None && "Error while getting builtin type.");
 
-  for (auto [idx, arg] : llvm::enumerate(e->arguments())) {
+  for (auto [idx, arg] : llvm::enumerate(e->arguments()))
     ops.push_back(emitScalarOrConstFoldImmArg(iceArguments, idx, arg));
-  }
 
   CIRGenBuilderTy &builder = getBuilder();
   mlir::Type voidTy = builder.getVoidTy();
@@ -98,6 +96,10 @@ mlir::Value CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID,
   case X86::BI__builtin_ia32_undef128:
   case X86::BI__builtin_ia32_undef256:
   case X86::BI__builtin_ia32_undef512:
+    cgm.errorNYI(e->getSourceRange(),
+                 std::string("unimplemented X86 builtin call: ") +
+                     getContext().BuiltinInfo.getName(builtinID));
+    return {};
   case X86::BI__builtin_ia32_vec_ext_v4hi:
   case X86::BI__builtin_ia32_vec_ext_v16qi:
   case X86::BI__builtin_ia32_vec_ext_v8hi:
@@ -107,7 +109,22 @@ mlir::Value CIRGenFunction::emitX86BuiltinExpr(unsigned builtinID,
   case X86::BI__builtin_ia32_vec_ext_v32qi:
   case X86::BI__builtin_ia32_vec_ext_v16hi:
   case X86::BI__builtin_ia32_vec_ext_v8si:
-  case X86::BI__builtin_ia32_vec_ext_v4di:
+  case X86::BI__builtin_ia32_vec_ext_v4di: {
+    unsigned numElts = cast<cir::VectorType>(ops[0].getType()).getSize();
+
+    uint64_t index =
+        ops[1].getDefiningOp<cir::ConstantOp>().getIntValue().getZExtValue();
+
+    index &= numElts - 1;
+
+    cir::ConstantOp indexVal =
+        builder.getUInt64(index, getLoc(e->getExprLoc()));
+
+    // These builtins exist so we can ensure the index is an ICE and in range.
+    // Otherwise we could just do this in the header file.
+    return cir::VecExtractOp::create(builder, getLoc(e->getExprLoc()), ops[0],
+                                     indexVal);
+  }
   case X86::BI__builtin_ia32_vec_set_v4hi:
   case X86::BI__builtin_ia32_vec_set_v16qi:
   case X86::BI__builtin_ia32_vec_set_v8hi:

clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp

@@ -1442,28 +1442,6 @@ mlir::Value CIRGenFunction::emitPromotedScalarExpr(const Expr *e,
   return ScalarExprEmitter(*this, builder).Visit(const_cast<Expr *>(e));
 }
 
-mlir::Value CIRGenFunction::emitScalarOrConstFoldImmArg(unsigned iceArguments,
-                                                        unsigned index,
-                                                        const Expr *arg) {
-  mlir::Value result{};
-
-  // The bit at the specified index indicates whether the argument is required
-  // to be a constant integer expression.
-  bool isArgRequiredToBeConstant = (iceArguments & (1 << index));
-
-  if (!isArgRequiredToBeConstant) {
-    result = emitScalarExpr(arg);
-  } else {
-    // If this is required to be a constant, constant fold it so that we
-    // know that the generated intrinsic gets a ConstantInt.
-    std::optional<llvm::APSInt> iceOpt =
-        arg->getIntegerConstantExpr(getContext());
-    assert(iceOpt && "Expected argument to be a constant");
-    result = builder.getConstInt(getLoc(arg->getSourceRange()), *iceOpt);
-  }
-  return result;
-}
-
 [[maybe_unused]] static bool mustVisitNullValue(const Expr *e) {
   // If a null pointer expression's type is the C++0x nullptr_t and
   // the expression is not a simple literal, it must be evaluated

clang/lib/CIR/CodeGen/CIRGenFunction.h

@@ -1546,9 +1546,6 @@ class CIRGenFunction : public CIRGenTypeCache {
   mlir::Value emitScalarExpr(const clang::Expr *e,
                              bool ignoreResultAssign = false);
 
-  mlir::Value emitScalarOrConstFoldImmArg(unsigned iceArguments, unsigned index,
-                                          const Expr *arg);
-
   mlir::Value emitScalarPrePostIncDec(const UnaryOperator *e, LValue lv,
                                       cir::UnaryOpKind kind, bool isPre);
 
@@ -1721,6 +1718,9 @@ class CIRGenFunction : public CIRGenTypeCache {
   void emitScalarInit(const clang::Expr *init, mlir::Location loc,
                       LValue lvalue, bool capturedByInit = false);
 
+  mlir::Value emitScalarOrConstFoldImmArg(unsigned iceArguments, unsigned idx,
+                                          const Expr *argExpr);
+
   void emitStaticVarDecl(const VarDecl &d, cir::GlobalLinkageKind linkage);
 
   void emitStoreOfComplex(mlir::Location loc, mlir::Value v, LValue dest,

clang/test/CIR/CodeGen/X86/sse2-builtins.c

@@ -16,6 +16,21 @@
 
 #include <immintrin.h>
 
+// Lowering to pextrw requires optimization.
+int test_mm_extract_epi16(__m128i A) {
+  // CIR-LABEL: test_mm_extract_epi16
+  // CIR %{{.*}} = cir.vec.extract %{{.*}}[%{{.*}} : {{!u32i|!u64i}}] : !cir.vector<!s16i x 8>
+  // CIR %{{.*}} = cir.cast integral %{{.*}} : !u16i -> !s32i
+
+  // LLVM-LABEL: test_mm_extract_epi16
+  // LLVM: extractelement <8 x i16> %{{.*}}, {{i32|i64}} 1
+  // LLVM: zext i16 %{{.*}} to i32
+
+  // OGCG-LABEL: test_mm_extract_epi16
+  // OGCG: extractelement <8 x i16> %{{.*}}, {{i32|i64}} 1
+  // OGCG: zext i16 %{{.*}} to i32
+  return _mm_extract_epi16(A, 1);
+}
 
 void test_mm_clflush(void* A) {
   // CIR-LABEL: test_mm_clflush

lldb/bindings/python/python-typemaps.swig

@@ -628,61 +628,34 @@ template <> bool SetNumberFromPyObject<double>(double &number, PyObject *obj) {
   }
 }
 
-// These two pybuffer macros are copied out of swig/Lib/python/pybuffer.i,
-// and fixed so they will not crash if PyObject_GetBuffer fails.
-// https://github.com/swig/swig/issues/1640
-//
-// I've also moved the call to PyBuffer_Release to the end of the SWIG wrapper,
-// doing it right away is not legal according to the python buffer protocol.
-%inline %{
-struct Py_buffer_RAII {
-  Py_buffer buffer = {};
-  Py_buffer_RAII(){};
-  Py_buffer &operator=(const Py_buffer_RAII &) = delete;
-  Py_buffer_RAII(const Py_buffer_RAII &) = delete;
-  ~Py_buffer_RAII() {
-    if (buffer.obj)
-      PyBuffer_Release(&buffer);
-  }
-};
-%}
 
-%define %pybuffer_mutable_binary(TYPEMAP, SIZE)
-%typemap(in) (TYPEMAP, SIZE) (Py_buffer_RAII view) {
-  int res;
-  Py_ssize_t size = 0;
-  void *buf = 0;
-  res = PyObject_GetBuffer($input, &view.buffer, PyBUF_WRITABLE);
-  if (res < 0) {
-    PyErr_Clear();
-    %argument_fail(res, "(TYPEMAP, SIZE)", $symname, $argnum);
-  }
-  size = view.buffer.len;
-  buf = view.buffer.buf;
-  $1 = ($1_ltype)buf;
-  $2 = ($2_ltype)(size / sizeof($*1_type));
-}
-%enddef
-
-%define %pybuffer_binary(TYPEMAP, SIZE)
-%typemap(in) (TYPEMAP, SIZE) (Py_buffer_RAII view) {
-  int res;
-  Py_ssize_t size = 0;
-  const void *buf = 0;
-  res = PyObject_GetBuffer($input, &view.buffer, PyBUF_CONTIG_RO);
-  if (res < 0) {
-    PyErr_Clear();
-    %argument_fail(res, "(TYPEMAP, SIZE)", $symname, $argnum);
+// Typemap for SBFile::Write.
+%typemap(in) (const uint8_t *buf, size_t num_bytes) {
+  if (PythonByteArray::Check($input)) {
+    PythonByteArray bytearray(PyRefType::Borrowed, $input);
+    $1 = (uint8_t *)bytearray.GetBytes().data();
+    $2 = bytearray.GetSize();
+  } else if (PythonBytes::Check($input)) {
+    PythonBytes bytes(PyRefType::Borrowed, $input);
+    $1 = (uint8_t *)bytes.GetBytes().data();
+    $2 = bytes.GetSize();
+  } else {
+    PyErr_SetString(PyExc_ValueError, "Expecting a bytes or bytearray object");
+    SWIG_fail;
   }
-  size = view.buffer.len;
-  buf = view.buffer.buf;
-  $1 = ($1_ltype)buf;
-  $2 = ($2_ltype)(size / sizeof($*1_type));
 }
-%enddef
 
-%pybuffer_binary(const uint8_t *buf, size_t num_bytes);
-%pybuffer_mutable_binary(uint8_t *buf, size_t num_bytes);
+// Typemap for SBFile::Read.
+%typemap(in) (uint8_t *buf, size_t num_bytes) {
+  if (PythonByteArray::Check($input)) {
+    PythonByteArray bytearray(PyRefType::Borrowed, $input);
+    $1 = (uint8_t *)bytearray.GetBytes().data();
+    $2 = bytearray.GetSize();
+  } else {
+    PyErr_SetString(PyExc_ValueError, "Expecting a bytearray");
+    SWIG_fail;
+  }
+}
 
 %typemap(in) (const char **symbol_name, uint32_t num_names) {
   using namespace lldb_private;

lldb/cmake/modules/LLDBConfig.cmake

@@ -180,13 +180,25 @@ if (LLDB_ENABLE_PYTHON)
       "Path to use as PYTHONHOME in lldb. If a relative path is specified, it will be resolved at runtime relative to liblldb directory.")
   endif()
 
-  if (SWIG_VERSION VERSION_GREATER_EQUAL "4.2" AND NOT LLDB_EMBED_PYTHON_HOME)
+  # Enable targeting the Python Limited C API.
+  set(PYTHON_LIMITED_API_MIN_SWIG_VERSION "4.2")
+  if (SWIG_VERSION VERSION_GREATER_EQUAL PYTHON_LIMITED_API_MIN_SWIG_VERSION
+      AND NOT LLDB_EMBED_PYTHON_HOME)
     set(default_enable_python_limited_api ON)
   else()
     set(default_enable_python_limited_api OFF)
   endif()
   option(LLDB_ENABLE_PYTHON_LIMITED_API "Force LLDB to only use the Python Limited API (requires SWIG 4.2 or later)"
     ${default_enable_python_limited_api})
+
+  # Diagnose unsupported configurations.
+  if (LLDB_ENABLE_PYTHON_LIMITED_API AND LLDB_EMBED_PYTHON_HOME)
+    message(SEND_ERROR "LLDB_ENABLE_PYTHON_LIMITED_API is not compatible with LLDB_EMBED_PYTHON_HOME")
+  endif()
+  if (LLDB_ENABLE_PYTHON_LIMITED_API AND SWIG_VERSION VERSION_LESS PYTHON_LIMITED_API_MIN_SWIG_VERSION)
+    message(SEND_ERROR "LLDB_ENABLE_PYTHON_LIMITED_API is not compatible with SWIG ${SWIG_VERSION} (requires SWIG ${PYTHON_LIMITED_API_MIN_SWIG_VERSION})")
+  endif()
+
 else()
   # Even if Python scripting is disabled, we still need a Python interpreter to
   # build, for example to generate SBLanguages.h.

lldb/include/lldb/Core/ModuleList.h

@@ -511,6 +511,12 @@ class ModuleList {
   /// Atomically swaps the contents of this module list with \a other.
   void Swap(ModuleList &other);
 
+  /// For each module in this ModuleList, preload its symbols.
+  ///
+  /// \param[in] parallelize
+  ///     If true, all modules will be preloaded in parallel.
+  void PreloadSymbols(bool parallelize) const;
+
 protected:
   // Class typedefs.
   typedef std::vector<lldb::ModuleSP>

lldb/include/lldb/Target/DynamicLoader.h

@@ -352,6 +352,7 @@ class DynamicLoader : public PluginInterface {
 protected:
   // Utility methods for derived classes
 
+  /// Find a module in the target that matches the given file.
   lldb::ModuleSP FindModuleViaTarget(const FileSpec &file);
 
   /// Checks to see if the target module has changed, updates the target