Assume a SILModule is whole-module when SILGen-ing from a ModuleDecl

No functionality change. Unfortunately we still need the flag in
SILModule itself because of the ability to create an empty SILModule
and parse SIL into it incrementally, which can happen before there's
a FileUnit to use as the associated DeclContext instead of a
CompilerInstance's main module.

lib/FrontendTool/FrontendTool.cpp

@@ -819,7 +819,7 @@ generateSILModules(CompilerInvocation &Invocation, CompilerInstance &Instance) {
   if (!opts.InputsAndOutputs.hasPrimaryInputs()) {
     // If there are no primary inputs the compiler is in WMO mode and builds one
     // SILModule for the entire module.
-    auto SM = performSILGeneration(mod, SILOpts, true);
+    auto SM = performSILGeneration(mod, SILOpts);
     std::deque<PostSILGenInputs> PSGIs;
     const PrimarySpecificPaths PSPs =
         Instance.getPrimarySpecificPathsForWholeModuleOptimizationMode();