[lld] add context-sensitive PGO options for COFF.

[YolandaCY]

Summary

Add lld CSPGO (Contex-Sensitive PGO) options for COFF target.

Reference the ELF options from https://reviews.llvm.org/D56675

[YolandaCY]

This patch is to enable CSPGO for Thinlto in lld on Windows. Could you please help review?

[xur]

This looks good to me. But I'll let other approve this.

[YolandaCY]

Thank you Rong. Any feedback from other reviewers?

I have verified this works for both old and new pass manager.

Regards
Yolanda

[pengfei]

I don't have much experience on lld. No objections.

[tejohnson]

lgtm

lld/COFF/Config.h

struct Configuration {
   std::pair<llvm::StringRef, llvm::StringRef> thinLTOPrefixReplace;
 
   // Used for /thinlto-object-suffix-replace:
   std::pair<llvm::StringRef, llvm::StringRef> thinLTOObjectSuffixReplace;
 
   // Used for /lto-obj-path:
   llvm::StringRef ltoObjPath;
 
+  // Used for /lto-cs-profile-generate:
+  bool ltoCSProfileGenerate = false;
+
+  // Used for /lto-cs-profile-path
+  llvm::StringRef ltoCSProfileFile;
+
   // Used for /call-graph-ordering-file:
   llvm::MapVector<std::pair<const SectionChunk *, const SectionChunk *>,
                   uint64_t>
       callGraphProfile;
   bool callGraphProfileSort = false;
 
   // Used for /print-symbol-order:
   StringRef printSymbolOrder;

lld/COFF/Driver.cpp

config->thinLTOIndexOnly = args.hasArg(OPT_thinlto_index_only) ||
                              args.hasArg(OPT_thinlto_index_only_arg);
   config->thinLTOIndexOnlyArg =
       args.getLastArgValue(OPT_thinlto_index_only_arg);
   config->thinLTOPrefixReplace =
       getOldNewOptions(args, OPT_thinlto_prefix_replace);
   config->thinLTOObjectSuffixReplace =
       getOldNewOptions(args, OPT_thinlto_object_suffix_replace);
   config->ltoObjPath = args.getLastArgValue(OPT_lto_obj_path);
+  config->ltoCSProfileGenerate = args.hasArg(OPT_lto_cs_profile_generate);
+  config->ltoCSProfileFile = args.getLastArgValue(OPT_lto_cs_profile_file);
   // Handle miscellaneous boolean flags.
   config->allowBind = args.hasFlag(OPT_allowbind, OPT_allowbind_no, true);
   config->allowIsolation =
       args.hasFlag(OPT_allowisolation, OPT_allowisolation_no, true);
   config->incremental =
       args.hasFlag(OPT_incremental, OPT_incremental_no,
                    !config->doGC && config->doICF == ICFLevel::None &&
                        !args.hasArg(OPT_order) && !args.hasArg(OPT_profile));

lld/COFF/LTO.cpp

static lto::Config createConfig() {
   c.DiagHandler = diagnosticHandler;
   c.OptLevel = config->ltoo;
   c.CPU = getCPUStr();
   c.MAttrs = getMAttrs();
   c.CGOptLevel = args::getCGOptLevel(config->ltoo);
   c.AlwaysEmitRegularLTOObj = !config->ltoObjPath.empty();
   c.UseNewPM = config->ltoNewPassManager;
   c.DebugPassManager = config->ltoDebugPassManager;
+  c.CSIRProfile = std::string(config->ltoCSProfileFile);
+  c.RunCSIRInstr = config->ltoCSProfileGenerate;
 
   if (config->saveTemps)
     checkError(c.addSaveTemps(std::string(config->outputFile) + ".",
                               /*UseInputModulePath*/ true));
   return c;
 }
 
 BitcodeCompiler::BitcodeCompiler() {

lld/COFF/Options.td

def thinlto_object_suffix_replace : P<
     "thinlto-object-suffix-replace",
     "'old;new' replace old suffix with new suffix in ThinLTO index">;
 def thinlto_prefix_replace: P<
     "thinlto-prefix-replace",
     "'old;new' replace old prefix with new prefix in ThinLTO outputs">;
 def lto_obj_path : P<
     "lto-obj-path",
     "output native object for merged LTO unit to this path">;
+def lto_cs_profile_generate: F<"lto-cs-profile-generate">,
+    HelpText<"Perform context sensitive PGO instrumentation">;
+def lto_cs_profile_file : P<"lto-cs-profile-file",
+    "Context sensitive profile file path">;
 def dash_dash_version : Flag<["--"], "version">,
   HelpText<"Display the version number and exit">;
 def threads
     : P<"threads", "Number of threads. '1' disables multi-threading. By "
                    "default all available hardware threads are used">;
 def call_graph_ordering_file: P<
     "call-graph-ordering-file", 
     "Layout sections to optimize the given callgraph">;