[5.9][Reflection] Improve ownsAddress and metadataIsActor logic.

Separately track text and data segments for ownsAddress. We were previously tracking one range per image, encompassing the range from the start of the image through the end of the data segment. This ends up including a lot of unwanted address space if the two aren't adjacent, as is the case for libraries in the shared cache on Darwin.

This makes metadataIsActor a lot more reliable, as it was previously identifying a lot of garbage as actor metadata due to the supposed descriptor pointer falling in this range.

rdar://113417637
(cherry picked from commit 2faa2d6)

Author: mikeash

include/swift/RemoteInspection/ReflectionContext.h

@@ -126,7 +126,8 @@ class ReflectionContext
   /// All buffers we need to keep around long term. This will automatically free them
   /// when this object is destroyed.
   std::vector<MemoryReader::ReadBytesResult> savedBuffers;
-  std::vector<std::tuple<RemoteAddress, RemoteAddress>> imageRanges;
+  std::vector<std::tuple<RemoteAddress, RemoteAddress>> textRanges;
+  std::vector<std::tuple<RemoteAddress, RemoteAddress>> dataRanges;
 
   bool setupTargetPointers = false;
   typename super::StoredPointer target_non_future_adapter = 0;
@@ -251,7 +252,7 @@ class ReflectionContext
     uint64_t Offset = 0;
 
     // Find the __TEXT segment.
-    typename T::SegmentCmd *Command = nullptr;
+    typename T::SegmentCmd *TextCommand = nullptr;
     for (unsigned I = 0; I < NumCommands; ++I) {
       auto CmdBuf = this->getReader().readBytes(
           RemoteAddress(CmdStartAddress.getAddressData() + Offset),
@@ -260,15 +261,15 @@ class ReflectionContext
         return false;
       auto CmdHdr = reinterpret_cast<typename T::SegmentCmd *>(CmdBuf.get());
       if (strncmp(CmdHdr->segname, "__TEXT", sizeof(CmdHdr->segname)) == 0) {
-        Command = CmdHdr;
+        TextCommand = CmdHdr;
         savedBuffers.push_back(std::move(CmdBuf));
         break;
       }
       Offset += CmdHdr->cmdsize;
     }
 
     // No __TEXT segment, bail out.
-    if (!Command)
+    if (!TextCommand)
       return false;
 
    // Find the load command offset.
@@ -291,7 +292,7 @@ class ReflectionContext
     if (!Sections)
       return false;
 
-    auto Slide = ImageStart.getAddressData() - Command->vmaddr;
+    auto Slide = ImageStart.getAddressData() - TextCommand->vmaddr;
     auto SectionsBuf = reinterpret_cast<const char *>(Sections.get());
 
     auto findMachOSectionByName = [&](llvm::StringRef Name)
@@ -355,22 +356,27 @@ class ReflectionContext
 
     auto InfoID = this->addReflectionInfo(info);
 
-    // Find the __DATA segment.
+    auto TextSegmentStart = Slide + TextCommand->vmaddr;
+    auto TextSegmentEnd = TextSegmentStart + TextCommand->vmsize;
+    textRanges.push_back(std::make_tuple(RemoteAddress(TextSegmentStart),
+                                         RemoteAddress(TextSegmentEnd)));
+
+    // Find the __DATA segments.
     for (unsigned I = 0; I < NumCommands; ++I) {
       auto CmdBuf = this->getReader().readBytes(
           RemoteAddress(CmdStartAddress.getAddressData() + Offset),
           SegmentCmdHdrSize);
       if (!CmdBuf)
         return false;
       auto CmdHdr = reinterpret_cast<typename T::SegmentCmd *>(CmdBuf.get());
-      if (strncmp(CmdHdr->segname, "__DATA", sizeof(CmdHdr->segname)) == 0) {
-        auto DataSegmentEnd =
-            ImageStart.getAddressData() + CmdHdr->vmaddr + CmdHdr->vmsize;
-        assert(DataSegmentEnd > ImageStart.getAddressData() &&
+      // Look for any segment name starting with __DATA.
+      if (strncmp(CmdHdr->segname, "__DATA", 6) == 0) {
+        auto DataSegmentStart = Slide + CmdHdr->vmaddr;
+        auto DataSegmentEnd = DataSegmentStart + CmdHdr->vmsize;
+        assert(DataSegmentStart > ImageStart.getAddressData() &&
                "invalid range for __DATA");
-        imageRanges.push_back(
-            std::make_tuple(ImageStart, RemoteAddress(DataSegmentEnd)));
-        break;
+        dataRanges.push_back(std::make_tuple(RemoteAddress(DataSegmentStart),
+                                             RemoteAddress(DataSegmentEnd)));
       }
       Offset += CmdHdr->cmdsize;
     }
@@ -829,9 +835,11 @@ class ReflectionContext
     return ownsAddress(RemoteAddress(*MetadataAddress));
   }
 
-  /// Returns true if the address falls within a registered image.
-  bool ownsAddressRaw(RemoteAddress Address) {
-    for (auto Range : imageRanges) {
+  /// Returns true if the address falls within the given address ranges.
+  bool ownsAddress(
+      RemoteAddress Address,
+      const std::vector<std::tuple<RemoteAddress, RemoteAddress>> &ranges) {
+    for (auto Range : ranges) {
       auto Start = std::get<0>(Range);
       auto End = std::get<1>(Range);
       if (Start.getAddressData() <= Address.getAddressData()
@@ -843,11 +851,14 @@ class ReflectionContext
   }
 
   /// Returns true if the address is known to the reflection context.
-  /// Currently, that means that either the address falls within a registered
-  /// image, or the address points to a Metadata whose type context descriptor
-  /// is within a registered image.
+  /// Currently, that means that either the address falls within the text or
+  /// data segments of a registered image, or the address points to a Metadata
+  /// whose type context descriptor is within the text segment of a registered
+  /// image.
   bool ownsAddress(RemoteAddress Address) {
-    if (ownsAddressRaw(Address))
+    if (ownsAddress(Address, textRanges))
+      return true;
+    if (ownsAddress(Address, dataRanges))
       return true;
 
     // This is usually called on a Metadata address which might have been
@@ -856,7 +867,7 @@ class ReflectionContext
     if (auto Metadata = readMetadata(Address.getAddressData()))
       if (auto DescriptorAddress =
           super::readAddressOfNominalTypeDescriptor(Metadata, true))
-        if (ownsAddressRaw(RemoteAddress(DescriptorAddress)))
+        if (ownsAddress(RemoteAddress(DescriptorAddress), textRanges))
           return true;
 
     return false;
@@ -1206,6 +1217,8 @@ class ReflectionContext
         super::readAddressOfNominalTypeDescriptor(Metadata);
     if (!DescriptorAddress)
       return false;
+    if (!ownsAddress(RemoteAddress(DescriptorAddress), textRanges))
+      return false;
 
     auto DescriptorBytes =
         getReader().readBytes(RemoteAddress(DescriptorAddress),