[COFF] RelocPtr is now a ulittle32_t pointer in LLVM

[jrtc27]

Summary

The type of RelocPtr has changed from uint32_t* to ulittle32_t*, as it
is pointing to little-endian data, so update our code to match. This
fixes COFF/pdb-framedata.yaml on big-endian hosts.

[jrtc27]

Ping?

[zturner]

I think the fix should in the api call which should probably return the
correct endian aware type

[ruiu]

I agree with Zach. If this API cannot be used correctly without reinterpret_cast, the return type needs fixing.

[jrtc27]

Remove reinterpret_cast and use D54421

[zturner]

Lgtm

[ruiu]

LGTM

[jrtc27]

Ping? Stack is now empty.

[jrtc27]

Ah, never mind, D54421 was updated to also patch lld at the same time, so this has been merged already.

COFF/PDB.cpp

if (!StringTableReferences.empty())
       warn("No StringTable subsection was encountered, but there are string "
            "table references");
     return;
   }
 
   // Rewrite string table indices in the Fpo Data and symbol records to refer to
   // the global PDB string table instead of the object file string table.
   for (DebugFrameDataSubsectionRef &FDS : NewFpoFrames) {
-    const uint32_t *Reloc = FDS.getRelocPtr();
+    const ulittle32_t *Reloc =
+      reinterpret_cast<const ulittle32_t *>(FDS.getRelocPtr());
     for (codeview::FrameData FD : FDS) {
       FD.RvaStart += *Reloc;
       FD.FrameFunc =
           translateStringTableIndex(FD.FrameFunc, CVStrTab, Linker.PDBStrTab);
       DbiBuilder.addNewFpoData(FD);
     }
   }