[llvm] [DebugInfo] Separate error generation from reporting in DWARFHeaderUnit::extract (PR #68242)

[David Blaikie via llvm-commits]

================
@@ -89,9 +89,12 @@ void fixupIndexV4(DWARFContext &C, DWARFUnitIndex &Index) {
     DWARFDataExtractor Data(DObj, S, C.isLittleEndian(), 0);
     while (Data.isValidOffset(Offset)) {
       DWARFUnitHeader Header;
-      if (!Header.extract(C, Data, &Offset, DWARFSectionKind::DW_SECT_INFO)) {
+      if (Error ExtractionErr = Header.extract(
+              C, Data, &Offset, DWARFSectionKind::DW_SECT_INFO)) {
         logAllUnhandledErrors(
-            createError("Failed to parse CU header in DWP file"), errs());
+            joinErrors(createError("Failed to parse CU header in DWP file"),
+                       std::move(ExtractionErr)),
+            errs());
----------------
dwblaikie wrote:

ah, `fixupIndex` should be reachable (simplest way to find it would be to `assert(false);` in there and you'd find which tests reach the code) via tests using `-manaully-generate-unit-index` - though they might not be testing invalid headers, and such - and could be improved to do so, that'd involve some manually crafted DWARF (probably written in assembly and assembled with `llvm-mc` in the test) that has a corrupted header.

It's a bit of a chore to create a hand-crafted dwp file, though... 

https://github.com/llvm/llvm-project/pull/68242