Second attempt:

Massive check in. This changes the "-fast" flag to "-O#" in llc. If you want to
use the old behavior, the flag is -O0. This change allows for finer-grained
control over which optimizations are run at different -O levels.

Most of this work was pretty mechanical. The majority of the fixes came from
verifying that a "fast" variable wasn't used anymore. The JIT still uses a
"Fast" flag. I'll change the JIT with a follow-up patch.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@70343 91177308-0d34-0410-b5e6-96231b3b80d8

lib/Target/CBackend/CBackend.cpp

@@ -3587,7 +3587,7 @@ void CWriter::visitExtractValueInst(ExtractValueInst &EVI) {
 bool CTargetMachine::addPassesToEmitWholeFile(PassManager &PM,
                                               raw_ostream &o,
                                               CodeGenFileType FileType,
-                                              bool Fast) {
+                                              unsigned OptLevel) {
   if (FileType != TargetMachine::AssemblyFile) return true;
 
   PM.add(createGCLoweringPass());

lib/Target/CBackend/CTargetMachine.h

@@ -27,7 +27,8 @@ struct CTargetMachine : public TargetMachine {
 
   virtual bool WantsWholeFile() const { return true; }
   virtual bool addPassesToEmitWholeFile(PassManager &PM, raw_ostream &Out,
-                                        CodeGenFileType FileType, bool Fast);
+                                        CodeGenFileType FileType,
+                                        unsigned OptLevel);
 
   // This class always works, but must be requested explicitly on 
   // llc command line.