[llvm-objdump] [debuginfod] Fetch for very-stripped binaries.

[mysterymath]

Summary

When a binary is missing section headers or symbols, objdump can't
provide as good of a disassembly. This change makes objdump try to fetch
a better verion of the binary by its build ID.

[mysterymath]

Integrate.

[mysterymath]

Correct clang-format formatting.

[jhenderson]

llvm/tools/llvm-objdump/llvm-objdump.cpp

1976

There's a lot of if clauses in this new area of code, so I'd expect to see a corresponding number of new test cases, but there's only one. Please make sure there's a test case that exercises each of the different possible code paths.

2096–2098

Might want to hold off on this usage here: there's an active discussion on whether we a) should allow it (due to potential footguns), and b) what the style of the C++17 if with initializer should be. See https://discourse.llvm.org/t/rfc-code-style-of-initializer-in-if-statement/65357/1.

[mysterymath]

Add tests; address review comments.

[jhenderson]

LGTM, with small nit.

llvm/test/tools/llvm-objdump/debuginfod.test

87–89

Small nit: I'd reorder these to match the order they're used in.

[mysterymath]

Reorder checks.

[mysterymath]

Update the no-section dump syntax.

llvm/test/tools/llvm-objdump/debuginfod.test

 BADBINARYERROR: warning: '[[PATH]]/broken{{[/\\]}}llvmcache-7361776989772977641': The file was not recognized as a valid object file
 
 # Use the original binary if the fetched binary has no debug info.
 RUN: env DEBUGINFOD_CACHE_PATH=%t/stripped-cache llvm-objdump -d --source \
 RUN:   --debuginfod %t/stripped 2> %t.err | \
 RUN:   FileCheck %s --check-prefix=NOTFOUND
 RUN: count 0 < %t.err
 
+# Use debuginfod to recover symbols.
+RUN: llvm-strip --strip-sections %t/stripped
+RUN: env DEBUGINFOD_CACHE_PATH=%t llvm-objdump -d --debuginfod \
+RUN:   %t/stripped | \
+RUN:   FileCheck %s --check-prefix=SYMBOLS
+
+# Use debuginfod to recover section headers, but not symbols.
+RUN: mkdir %t/stripped-symbols
+RUN: cp %p/Inputs/embedded-source %t/stripped-symbols/llvmcache-7361776989772977641
+RUN: llvm-strip %t/stripped-symbols/llvmcache-7361776989772977641
+RUN: env DEBUGINFOD_CACHE_PATH=%t/stripped-symbols llvm-objdump -d \
+RUN:   --debuginfod %t/stripped | \
+RUN:   FileCheck %s --check-prefix=SECTIONS
+
+# Don't use debuginfod if neither section headers nor symbols can be recovered.
+RUN: mkdir %t/stripped-sections
+RUN: echo "" | llvm-mc -filetype=obj -triple x86_64 > \
+RUN:   %t/stripped-sections/llvmcache-7361776989772977641
+RUN: llvm-strip --strip-sections %t/stripped-sections/llvmcache-7361776989772977641
+RUN: env DEBUGINFOD_CACHE_PATH=%t/stripped-sections llvm-objdump -d \
+RUN:   --debuginfod %t/stripped | \
+RUN:   FileCheck %s --check-prefix=NOSECTIONS
+
 NOTFOUND-NOT: int main(int argc, char *argv[]) {
 FOUND: int main(int argc, char *argv[]) {
+
+SYMBOLS: <main>:
+SECTIONS: <.text>:
+NOSECTIONS: <PT_LOAD#{{[0-9]+}}>:

llvm/tools/llvm-objdump/llvm-objdump.cpp

static void disassembleObject(const Target *TheTarget, ObjectFile &Obj,
   }
   StringSet<> MissingDisasmSymbolSet =
       set_difference(DisasmSymbolSet, FoundDisasmSymbolSet);
   for (StringRef Sym : MissingDisasmSymbolSet.keys())
     reportWarning("failed to disassemble missing symbol " + Sym, FileName);
 }
 
 static void disassembleObject(ObjectFile *Obj, bool InlineRelocs) {
+  // If information useful for showing the disassembly is missing, try to find a
+  // more complete binary and disassemble that instead.
+  OwningBinary<Binary> FetchedBinary;
+  if (Obj->symbols().empty()) {
+    if (Optional<OwningBinary<Binary>> FetchedBinaryOpt =
+            fetchBinaryByBuildID(*Obj)) {
+      if (auto *O = dyn_cast<ObjectFile>(FetchedBinaryOpt->getBinary())) {
+        if (!O->symbols().empty() ||
+            (!O->sections().empty() && Obj->sections().empty())) {
+          FetchedBinary = std::move(*FetchedBinaryOpt);
+          Obj = O;
+        }
+      }
+    }
+  }
+
   const Target *TheTarget = getTarget(Obj);
 
   // Package up features to be passed to target/subtarget
   SubtargetFeatures Features = Obj->getFeatures();
   if (!MAttrs.empty()) {
     for (unsigned I = 0; I != MAttrs.size(); ++I)
       Features.AddFeature(MAttrs[I]);
   } else if (MCPU.empty() && Obj->getArch() == llvm::Triple::aarch64) {
static void disassembleObject(ObjectFile *Obj, bool InlineRelocs) {
   IP->setPrintImmHex(PrintImmHex);
   IP->setPrintBranchImmAsAddress(true);
   IP->setSymbolizeOperands(SymbolizeOperands);
   IP->setMCInstrAnalysis(MIA.get());
 
   PrettyPrinter &PIP = selectPrettyPrinter(Triple(TripleName));
 
   const ObjectFile *DbgObj = Obj;
-  OwningBinary<Binary> DebugBinary;
-  if (!DbgObj->hasDebugInfo()) {
+  if (!FetchedBinary.getBinary() && !Obj->hasDebugInfo()) {
     if (Optional<OwningBinary<Binary>> DebugBinaryOpt =
             fetchBinaryByBuildID(*Obj)) {
       if (auto *FetchedObj =
               dyn_cast<const ObjectFile>(DebugBinaryOpt->getBinary())) {
         if (FetchedObj->hasDebugInfo()) {
-          DebugBinary = std::move(*DebugBinaryOpt);
+          FetchedBinary = std::move(*DebugBinaryOpt);
           DbgObj = FetchedObj;
         }
       }
     }
   }
 
   std::unique_ptr<object::Binary> DSYMBinary;
   std::unique_ptr<MemoryBuffer> DSYMBuf;