[DLCov 3/5] Implement DebugLoc origin-tracking

llvm/CMakeLists.txt

@@ -569,8 +569,8 @@ endif()
 option(LLVM_ENABLE_CRASH_DUMPS "Turn on memory dumps on crashes. Currently only implemented on Windows." OFF)
 
 set(LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING "DISABLED" CACHE STRING
-  "Enhance Debugify's line number coverage tracking; enabling this is ABI-breaking. Can be DISABLED, or COVERAGE.")
-set_property(CACHE LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING PROPERTY STRINGS DISABLED COVERAGE)
+  "Enhance debugify's line number coverage tracking; enabling this is abi-breaking. Can be DISABLED, COVERAGE, or COVERAGE_AND_ORIGIN.")
+set_property(CACHE LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING PROPERTY STRINGS DISABLED COVERAGE COVERAGE_AND_ORIGIN)
 
 option(LLVM_EXPERIMENTAL_KEY_INSTRUCTIONS
   "Add additional fields to DILocations to support Key Instructions" OFF)

llvm/cmake/modules/HandleLLVMOptions.cmake

@@ -200,6 +200,9 @@ string(TOUPPER "${LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING}" uppercase_LLVM_ENABLE
 
 if( uppercase_LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING STREQUAL "COVERAGE" )
   set( LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING 1 )
+elseif( uppercase_LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING STREQUAL "COVERAGE_AND_ORIGIN" )
+  set( ENABLE_DEBUGLOC_COVERAGE_TRACKING 1 )
+  set( LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING 1 )
 elseif( uppercase_LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING STREQUAL "DISABLED" OR NOT DEFINED LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING )
   # The DISABLED setting is default and requires no additional defines.
 else()

llvm/docs/CMake.rst

@@ -482,11 +482,14 @@ enabled sub-projects. Nearly all of these variable names begin with
 **LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING**:STRING
   Enhances Debugify's ability to detect line number errors by storing extra
   information inside Instructions, removing false positives from Debugify's
-  results at the cost of performance. Allowed values are `DISABLED` (default)
-  and `COVERAGE`. `COVERAGE` tracks whether and why a line number was
-  intentionally dropped or not generated for an instruction, allowing Debugify
-  to avoid reporting these as errors; this comes with a small performance cost
-  of ~0.1%. `COVERAGE` is an ABI-breaking option.
+  results at the cost of performance. Allowed values are `DISABLED` (default),
+  `COVERAGE`, and `COVERAGE_AND_ORIGIN`. `COVERAGE` tracks whether and why a
+  line number was intentionally dropped or not generated for an instruction,
+  allowing Debugify to avoid reporting these as errors; this comes with a small
+  performance cost of ~0.1%. `COVERAGE_AND_ORIGIN` additionally stores a
+  stacktrace of the point where each DebugLoc is unintentionally dropped,
+  allowing for much easier bug triaging at the cost of a ~10x performance
+  slowdown. `COVERAGE` and `COVERAGE_AND_ORIGIN` are ABI-breaking options.
 
 **LLVM_ENABLE_DIA_SDK**:BOOL
   Enable building with MSVC DIA SDK for PDB debugging support. Available

llvm/include/llvm/Config/config.h.cmake

@@ -19,6 +19,14 @@
 /* Define to 1 to enable crash memory dumps, and to 0 otherwise. */
 #cmakedefine01 LLVM_ENABLE_CRASH_DUMPS
 
+/* Define to 1 to enable expensive checks for debug location coverage checking,
+   and to 0 otherwise. */
+#cmakedefine01 ENABLE_DEBUGLOC_COVERAGE_TRACKING
+
+/* Define to 1 to enable expensive tracking of the origin of debug location
+   coverage bugs, and to 0 otherwise. */
+#cmakedefine01 LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING
+
 /* Define to 1 to prefer forward slashes on Windows, and to 0 prefer
    backslashes. */
 #cmakedefine01 LLVM_WINDOWS_PREFER_FORWARD_SLASH

llvm/include/llvm/IR/DebugLoc.h

@@ -27,6 +27,21 @@ namespace llvm {
   class Function;
 
 #if LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING
+#if LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING
+  struct DbgLocOrigin {
+    static constexpr unsigned long MaxDepth = 16;
+    using StackTracesTy =
+        SmallVector<std::pair<int, std::array<void *, MaxDepth>>, 0>;
+    StackTracesTy StackTraces;
+    DbgLocOrigin(bool ShouldCollectTrace);
+    void addTrace();
+    const StackTracesTy &getOriginStackTraces() const { return StackTraces; };
+  };
+#else
+  struct DbgLocOrigin {
+    DbgLocOrigin(bool) {}
+  };
+#endif
   // Used to represent different "kinds" of DebugLoc, expressing that the
   // instruction it is part of is either normal and should contain a valid
   // DILocation, or otherwise describing the reason why the instruction does
@@ -55,22 +70,29 @@ namespace llvm {
     Temporary
   };
 
-  // Extends TrackingMDNodeRef to also store a DebugLocKind, allowing Debugify
-  // to ignore intentionally-empty DebugLocs.
-  class DILocAndCoverageTracking : public TrackingMDNodeRef {
+  // Extends TrackingMDNodeRef to also store a DebugLocKind and Origin,
+  // allowing Debugify to ignore intentionally-empty DebugLocs and display the
+  // code responsible for generating unintentionally-empty DebugLocs.
+  // Currently we only need to track the Origin of this DILoc when using a
+  // DebugLoc that is not annotated (i.e. has DebugLocKind::Normal) and has a
+  // null DILocation, so only collect the origin stacktrace in those cases.
+  class DILocAndCoverageTracking : public TrackingMDNodeRef,
+                                   public DbgLocOrigin {
   public:
     DebugLocKind Kind;
     // Default constructor for empty DebugLocs.
     DILocAndCoverageTracking()
-        : TrackingMDNodeRef(nullptr), Kind(DebugLocKind::Normal) {}
-    // Valid or nullptr MDNode*, normal DebugLocKind.
+        : TrackingMDNodeRef(nullptr), DbgLocOrigin(true),
+          Kind(DebugLocKind::Normal) {}
+    // Valid or nullptr MDNode*, no annotative DebugLocKind.
     DILocAndCoverageTracking(const MDNode *Loc)
-        : TrackingMDNodeRef(const_cast<MDNode *>(Loc)),
+        : TrackingMDNodeRef(const_cast<MDNode *>(Loc)), DbgLocOrigin(!Loc),
           Kind(DebugLocKind::Normal) {}
     LLVM_ABI DILocAndCoverageTracking(const DILocation *Loc);
     // Explicit DebugLocKind, which always means a nullptr MDNode*.
     DILocAndCoverageTracking(DebugLocKind Kind)
-        : TrackingMDNodeRef(nullptr), Kind(Kind) {}
+        : TrackingMDNodeRef(nullptr),
+          DbgLocOrigin(Kind == DebugLocKind::Normal), Kind(Kind) {}
   };
   template <> struct simplify_type<DILocAndCoverageTracking> {
     using SimpleType = MDNode *;
@@ -142,6 +164,32 @@ namespace llvm {
     static inline DebugLoc getDropped() { return DebugLoc(); }
 #endif // LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING
 
+#if LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING
+    DebugLoc(DebugLocKind Kind) : Loc(Kind) {}
+    DebugLocKind getKind() const { return Loc.Kind; }
+#endif
+
+#if LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING
+#if !LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING
+#error Cannot enable DebugLoc origin-tracking without coverage-tracking!
+#endif
+
+    const DbgLocOrigin::StackTracesTy &getOriginStackTraces() const {
+      return Loc.getOriginStackTraces();
+    }
+    DebugLoc getCopied() const {
+      DebugLoc NewDL = *this;
+      NewDL.Loc.addTrace();
+      return NewDL;
+    }
+#else
+    DebugLoc getCopied() const { return *this; }
+#endif
+
+    static DebugLoc getTemporary();
+    static DebugLoc getUnknown();
+    static DebugLoc getLineZero();
+
     /// Get the underlying \a DILocation.
     ///
     /// \pre !*this or \c isa<DILocation>(getAsMDNode()).

llvm/include/llvm/IR/Instruction.h

@@ -507,7 +507,7 @@ class Instruction : public User,
   LLVM_ABI bool extractProfTotalWeight(uint64_t &TotalVal) const;
 
   /// Set the debug location information for this instruction.
-  void setDebugLoc(DebugLoc Loc) { DbgLoc = std::move(Loc); }
+  void setDebugLoc(DebugLoc Loc) { DbgLoc = std::move(Loc).getCopied(); }
 
   /// Return the debug location for this node as a DebugLoc.
   const DebugLoc &getDebugLoc() const { return DbgLoc; }

llvm/include/llvm/Support/Signals.h

@@ -14,14 +14,32 @@
 #ifndef LLVM_SUPPORT_SIGNALS_H
 #define LLVM_SUPPORT_SIGNALS_H
 
+#include "llvm/Config/llvm-config.h"
 #include "llvm/Support/Compiler.h"
+#include <array>
 #include <cstdint>
 #include <string>
 
 namespace llvm {
 class StringRef;
 class raw_ostream;
 
+#if LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING
+// Typedefs that are convenient but only used by the stack-trace-collection code
+// added if DebugLoc origin-tracking is enabled.
+template <typename T, typename Enable> struct DenseMapInfo;
+template <typename ValueT, typename ValueInfoT> class DenseSet;
+namespace detail {
+template <typename KeyT, typename ValueT> struct DenseMapPair;
+}
+template <typename KeyT, typename ValueT, typename KeyInfoT, typename BucketT>
+class DenseMap;
+using AddressSet = DenseSet<void *, DenseMapInfo<void *, void>>;
+using SymbolizedAddressMap =
+    DenseMap<void *, std::string, DenseMapInfo<void *, void>,
+             detail::DenseMapPair<void *, std::string>>;
+#endif
+
 namespace sys {
 
 /// This function runs all the registered interrupt handlers, including the
@@ -57,6 +75,28 @@ LLVM_ABI void DisableSystemDialogsOnCrash();
 ///        specified, the entire frame is printed.
 LLVM_ABI void PrintStackTrace(raw_ostream &OS, int Depth = 0);
 
+#if LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING
+#ifdef NDEBUG
+#error DebugLoc origin-tracking should not be enabled in Release builds.
+#endif
+  /// Populates the given array with a stack trace of the current program, up to
+  /// MaxDepth frames. Returns the number of frames returned, which will be
+  /// inserted into \p StackTrace from index 0. All entries after the returned
+  /// depth will be unmodified. NB: This is only intended to be used for
+  /// introspection of LLVM by Debugify, will not be enabled in release builds,
+  /// and should not be relied on for other purposes.
+  template <unsigned long MaxDepth>
+  int getStackTrace(std::array<void *, MaxDepth> &StackTrace);
+
+  /// Takes a set of \p Addresses, symbolizes them and stores the result in the
+  /// provided \p SymbolizedAddresses map.
+  /// NB: This is only intended to be used for introspection of LLVM by
+  /// Debugify, will not be enabled in release builds, and should not be relied
+  /// on for other purposes.
+  void symbolizeAddresses(AddressSet &Addresses,
+                          SymbolizedAddressMap &SymbolizedAddresses);
+#endif
+
 // Run all registered signal handlers.
 LLVM_ABI void RunSignalHandlers();
 

llvm/lib/CodeGen/BranchFolding.cpp

@@ -42,6 +42,7 @@
 #include "llvm/CodeGen/TargetPassConfig.h"
 #include "llvm/CodeGen/TargetRegisterInfo.h"
 #include "llvm/CodeGen/TargetSubtargetInfo.h"
+#include "llvm/Config/llvm-config.h"
 #include "llvm/IR/DebugInfoMetadata.h"
 #include "llvm/IR/DebugLoc.h"
 #include "llvm/IR/Function.h"
@@ -933,7 +934,13 @@ bool BranchFolder::TryTailMergeBlocks(MachineBasicBlock *SuccBB,
 
   // Sort by hash value so that blocks with identical end sequences sort
   // together.
+#if LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING
+  // If origin-tracking is enabled then MergePotentialElt is no longer a POD
+  // type, so we need std::sort instead.
+  std::sort(MergePotentials.begin(), MergePotentials.end());
+#else
   array_pod_sort(MergePotentials.begin(), MergePotentials.end());
+#endif
 
   // Walk through equivalence sets looking for actual exact matches.
   while (MergePotentials.size() > 1) {

llvm/lib/IR/DebugLoc.cpp

@@ -9,11 +9,31 @@
 #include "llvm/IR/DebugLoc.h"
 #include "llvm/Config/llvm-config.h"
 #include "llvm/IR/DebugInfo.h"
+
+#if LLVM_ENABLE_DEBUGLOC_ORIGIN_TRACKING
+#include "llvm/Support/Signals.h"
+
+namespace llvm {
+DbgLocOrigin::DbgLocOrigin(bool ShouldCollectTrace) {
+  if (ShouldCollectTrace) {
+    auto &[Depth, StackTrace] = StackTraces.emplace_back();
+    Depth = sys::getStackTrace(StackTrace);
+  }
+}
+void DbgLocOrigin::addTrace() {
+  if (StackTraces.empty())
+    return;
+  auto &[Depth, StackTrace] = StackTraces.emplace_back();
+  Depth = sys::getStackTrace(StackTrace);
+}
+} // namespace llvm
+#endif
+
 using namespace llvm;
 
 #if LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING
 DILocAndCoverageTracking::DILocAndCoverageTracking(const DILocation *L)
-    : TrackingMDNodeRef(const_cast<DILocation *>(L)),
+    : TrackingMDNodeRef(const_cast<DILocation *>(L)), DbgLocOrigin(!L),
       Kind(DebugLocKind::Normal) {}
 #endif // LLVM_ENABLE_DEBUGLOC_COVERAGE_TRACKING
 

llvm/lib/IR/Instruction.cpp

@@ -1393,5 +1393,6 @@ Instruction *Instruction::clone() const {
 
   New->SubclassOptionalData = SubclassOptionalData;
   New->copyMetadata(*this);
+  New->setDebugLoc(getDebugLoc().getCopied());
   return New;
 }