[Lldb-commits] [PATCH] Introduce DWARFDataExtractor for 64-Bit DWARF parsing

[Ed Maste]

================
Comment at: include/lldb/Core/MappedHash.h:17
@@ -16,2 +16,3 @@
 
+template <typename ExtractorT>
 class MappedHash
----------------
I'd like to avoid templatizing this if possible, but I don't see a straightforward way.

================
Comment at: source/Plugins/SymbolFile/DWARF/DWARFDebugArangeSet.h:44
@@ -43,3 +43,3 @@
         void        Compact();
-        bool        Extract(const lldb_private::DataExtractor &data, lldb::offset_t *offset_ptr);
+        bool        Extract(lldb_private::DWARFDataExtractor &data, lldb::offset_t *offset_ptr);
         void        Dump(lldb_private::Stream *s) const;
----------------
Dropping the consts is a consequence of setting the 64-bit flag in DWARFDataExtractor::GetDWARFInitialLength.

Would it be reasonable to use a const_cast in there instead?

================
Comment at: source/Plugins/SymbolFile/DWARF/DWARFDebugArangeSet.cpp:186
@@ -185,14 +185,3 @@
         // the size appropriate for an address on the target architecture.
-        size_t dwarf_offset_size = 4;
-        m_header.length     = data.GetU32(offset_ptr);
-        if (m_header.length == 0xffffffff)
-        {
-            dwarf_offset_size = 8;
-            m_header.length = data.GetU64(offset_ptr);
-        }
-        else if (m_header.length >= 0xffffff00)
-        {
-            // Reserved.
-            return false;
-        }
+        m_header.length     = data.GetDWARFInitialLength(offset_ptr);
         m_header.version    = data.GetU16(offset_ptr);
----------------
Here's the actual cleanup this permits


http://llvm-reviews.chandlerc.com/D2007