ELF: Add ThinLTO caching support.

[pcc]

Summary

This patch adds an option named --thinlto-cache-dir, which specifies the
path to a directory in which to cache native object files for ThinLTO
incremental builds.

[ruiu]

I can make a guess, but can you explain what this is for? (and add that to the code as comments?)

[davide]

The logic looks good. I'll let Rui review the style.

lld/ELF/LTO.cpp

177–178

explicit type, maybe?

[pcc]

Add comment explaining the --thinlto-cache-dir option

[pcc]

Done, I've also updated the commit message with an explanation.

[ruiu]

LGTM

[mehdi_amini]

We should llvm/cmake/modules/HandleLLVMOptions.cmake to add this by default when LLVM_ENABLE_LTO=Thin to match Darwin.

[pcc]

In D30509#690142, @mehdi_amini wrote:

We should llvm/cmake/modules/HandleLLVMOptions.cmake to add this by default when LLVM_ENABLE_LTO=Thin to match Darwin.

Makes sense, D30522.

lld/ELF/Config.h

struct Configuration {
   llvm::StringRef Init;
   llvm::StringRef LTOAAPipeline;
   llvm::StringRef LTONewPmPasses;
   llvm::StringRef MapFile;
   llvm::StringRef OutputFile;
   llvm::StringRef OptRemarksFilename;
   llvm::StringRef SoName;
   llvm::StringRef Sysroot;
+  llvm::StringRef ThinLTOCacheDir;
   std::string RPath;
   std::vector<VersionDefinition> VersionDefinitions;
   std::vector<llvm::StringRef> AuxiliaryList;
   std::vector<llvm::StringRef> SearchPaths;
   std::vector<llvm::StringRef> SymbolOrderingFile;
   std::vector<llvm::StringRef> Undefined;
   std::vector<SymbolVersion> VersionScriptGlobals;
   std::vector<SymbolVersion> VersionScriptLocals;

lld/ELF/Driver.cpp

void LinkerDriver::readConfigs(opt::InputArgList &Args) {
   Config->Shared = Args.hasArg(OPT_shared);
   Config->SingleRoRx = Args.hasArg(OPT_no_rosegment);
   Config->SoName = getString(Args, OPT_soname);
   Config->SortSection = getSortSection(Args);
   Config->Strip = getStrip(Args);
   Config->Sysroot = getString(Args, OPT_sysroot);
   Config->Target1Rel = getArg(Args, OPT_target1_rel, OPT_target1_abs, false);
   Config->Target2 = getTarget2(Args);
+  Config->ThinLTOCacheDir = getString(Args, OPT_thinlto_cache_dir);
   Config->ThinLTOJobs = getInteger(Args, OPT_thinlto_jobs, -1u);
   Config->Threads = getArg(Args, OPT_threads, OPT_no_threads, true);
   Config->Trace = Args.hasArg(OPT_trace);
   Config->Undefined = getArgs(Args, OPT_undefined);
   Config->UnresolvedSymbols = getUnresolvedSymbolPolicy(Args);
   Config->Verbose = Args.hasArg(OPT_verbose);
   Config->WarnCommon = Args.hasArg(OPT_warn_common);
   Config->ZCombreloc = !hasZOption(Args, "nocombreloc");

lld/ELF/LTO.h

public:
   ~BitcodeCompiler();
 
   void add(BitcodeFile &F);
   std::vector<InputFile *> compile();
 
 private:
   std::unique_ptr<llvm::lto::LTO> LTOObj;
   std::vector<SmallString<0>> Buff;
+  std::vector<std::unique_ptr<MemoryBuffer>> Files;
 };
 }
 }
 
 #endif

lld/ELF/LTO.cpp

 #include "InputFiles.h"
 #include "Symbols.h"
 #include "lld/Core/TargetOptionsCommandFlags.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/StringRef.h"
 #include "llvm/ADT/Twine.h"
 #include "llvm/IR/DiagnosticPrinter.h"
+#include "llvm/LTO/Caching.h"
 #include "llvm/LTO/Config.h"
 #include "llvm/LTO/LTO.h"
 #include "llvm/Object/SymbolicFile.h"
 #include "llvm/Support/CodeGen.h"
 #include "llvm/Support/ELF.h"
 #include "llvm/Support/Error.h"
 #include "llvm/Support/FileSystem.h"
 #include "llvm/Support/MemoryBuffer.h"

 }
 
 // Merge all the bitcode files we have seen, codegen the result
 // and return the resulting ObjectFile(s).
 std::vector<InputFile *> BitcodeCompiler::compile() {
   std::vector<InputFile *> Ret;
   unsigned MaxTasks = LTOObj->getMaxTasks();
   Buff.resize(MaxTasks);
+  Files.resize(MaxTasks);
 
-  checkError(LTOObj->run([&](size_t Task) {
+  // The --thinlto-cache-dir option specifies the path to a directory in which
+  // to cache native object files for ThinLTO incremental builds. If a path was
+  // specified, configure LTO to use it as the cache directory.
+  lto::NativeObjectCache Cache;
+  if (!Config->ThinLTOCacheDir.empty())
+    Cache = lto::localCache(Config->ThinLTOCacheDir,
+                            [&](size_t Task, StringRef Path) {
+                              Files[Task] = check(MemoryBuffer::getFile(Path));
+                            });
+
+  checkError(LTOObj->run(
+      [&](size_t Task) {
         return llvm::make_unique<lto::NativeObjectStream>(
             llvm::make_unique<raw_svector_ostream>(Buff[Task]));
-  }));
+      },
+      Cache));
 
   for (unsigned I = 0; I != MaxTasks; ++I) {
     if (Buff[I].empty())
       continue;
     if (Config->SaveTemps) {
       if (I == 0)
         saveBuffer(Buff[I], Config->OutputFile + ".lto.o");
       else
         saveBuffer(Buff[I], Config->OutputFile + Twine(I) + ".lto.o");
     }
     InputFile *Obj = createObjectFile(MemoryBufferRef(Buff[I], "lto.tmp"));
     Ret.push_back(Obj);
   }
+
+  for (std::unique_ptr<MemoryBuffer> &File : Files)
+    if (File)
+      Ret.push_back(createObjectFile(*File));
+
   return Ret;
 }

lld/ELF/Options.td

def lto_partitions: J<"lto-partitions=">,
   HelpText<"Number of LTO codegen partitions">;
 def disable_verify: F<"disable-verify">;
 def mllvm: S<"mllvm">;
 def opt_remarks_filename: S<"opt-remarks-filename">,
   HelpText<"YAML output file for optimization remarks">;
 def opt_remarks_with_hotness: F<"opt-remarks-with-hotness">,
   HelpText<"Include hotness informations in the optimization remarks file">;
 def save_temps: F<"save-temps">;
+def thinlto_cache_dir: J<"thinlto-cache-dir=">,
+  HelpText<"Path to ThinLTO cached object file directory">;
 def thinlto_jobs: J<"thinlto-jobs=">, HelpText<"Number of ThinLTO jobs">;

lld/test/ELF/lto/Inputs/cache.ll

+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define i32 @main() {
+entry:
+  call void (...) @globalfunc()
+  ret i32 0
+}
+
+declare void @globalfunc(...)

lld/test/ELF/lto/cache.ll

+; RUN: opt -module-hash -module-summary %s -o %t.o
+; RUN: opt -module-hash -module-summary %p/Inputs/cache.ll -o %t2.o
+
+; RUN: rm -Rf %t.cache && mkdir %t.cache
+; RUN: ld.lld --thinlto-cache-dir=%t.cache -o %t3.o %t2.o %t.o
+
+; RUN: ls %t.cache | count 2
+
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define void @globalfunc() #0 {
+entry:
+  ret void
+}