Add support for missed and analysis optimization remarks.

Summary:
This adds two new diagnostics: -pass-remarks-missed and
-pass-remarks-analysis. They take the same values as -pass-remarks but
are intended to be triggered in different contexts.

-pass-remarks-missed is used by LLVMContext::emitOptimizationRemarkMissed,
which passes call when they tried to apply a transformation but
couldn't.

-pass-remarks-analysis is used by LLVMContext::emitOptimizationRemarkAnalysis,
which passes call when they want to inform the user about analysis
results.

The patch also:

1- Adds support in the inliner for the two new remarks and a
   test case.

2- Moves emitOptimizationRemark* functions to the llvm namespace.

3- Adds an LLVMContext argument instead of making them member functions
   of LLVMContext.

Reviewers: qcolombet

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D3682

llvm-svn: 209442

5 files changed, 67 insertions, 29 deletions

diff --git a/llvm/lib/Transforms/IPO/Inliner.cpp b/llvm/lib/Transforms/IPO/Inliner.cpp
index 10b20cfc917..9087ab23bb7 100644
--- a/llvm/lib/Transforms/IPO/Inliner.cpp
+++ b/llvm/lib/Transforms/IPO/Inliner.cpp
@@ -301,6 +301,13 @@ unsigned Inliner::getInlineThreshold(CallSite CS) const {
   return thres;
 }
 
+static void emitAnalysis(CallSite CS, const Twine &Msg) {
+  Function *Caller = CS.getCaller();
+  LLVMContext &Ctx = Caller->getContext();
+  DebugLoc DLoc = CS.getInstruction()->getDebugLoc();
+  emitOptimizationRemarkAnalysis(Ctx, DEBUG_TYPE, *Caller, DLoc, Msg);
+}
+
 /// shouldInline - Return true if the inliner should attempt to inline
 /// at the given CallSite.
 bool Inliner::shouldInline(CallSite CS) {
@@ -309,12 +316,16 @@ bool Inliner::shouldInline(CallSite CS) {
   if (IC.isAlways()) {
     DEBUG(dbgs() << "    Inlining: cost=always"
           << ", Call: " << *CS.getInstruction() << "\n");
+    emitAnalysis(CS, Twine(CS.getCalledFunction()->getName()) +
+                         " should always be inlined (cost=always)");
     return true;
   }
   
   if (IC.isNever()) {
     DEBUG(dbgs() << "    NOT Inlining: cost=never"
           << ", Call: " << *CS.getInstruction() << "\n");
+    emitAnalysis(CS, Twine(CS.getCalledFunction()->getName() +
+                           " should never be inlined (cost=never)"));
     return false;
   }
   
@@ -323,6 +334,10 @@ bool Inliner::shouldInline(CallSite CS) {
     DEBUG(dbgs() << "    NOT Inlining: cost=" << IC.getCost()
           << ", thres=" << (IC.getCostDelta() + IC.getCost())
           << ", Call: " << *CS.getInstruction() << "\n");
+    emitAnalysis(CS, Twine(CS.getCalledFunction()->getName() +
+                           " too costly to inline (cost=") +
+                         Twine(IC.getCost()) + ", threshold=" +
+                         Twine(IC.getCostDelta() + IC.getCost()) + ")");
     return false;
   }
   
@@ -390,6 +405,11 @@ bool Inliner::shouldInline(CallSite CS) {
       DEBUG(dbgs() << "    NOT Inlining: " << *CS.getInstruction() <<
            " Cost = " << IC.getCost() <<
            ", outer Cost = " << TotalSecondaryCost << '\n');
+      emitAnalysis(
+          CS, Twine("Not inlining. Cost of inlining " +
+                    CS.getCalledFunction()->getName() +
+                    " increases the cost of inlining " +
+                    CS.getCaller()->getName() + " in other contexts"));
       return false;
     }
   }
@@ -397,6 +417,10 @@ bool Inliner::shouldInline(CallSite CS) {
   DEBUG(dbgs() << "    Inlining: cost=" << IC.getCost()
         << ", thres=" << (IC.getCostDelta() + IC.getCost())
         << ", Call: " << *CS.getInstruction() << '\n');
+  emitAnalysis(
+      CS, CS.getCalledFunction()->getName() + Twine(" can be inlined into ") +
+              CS.getCaller()->getName() + " with cost=" + Twine(IC.getCost()) +
+              " (threshold=" + Twine(IC.getCostDelta() + IC.getCost()) + ")");
   return true;
 }
 
@@ -518,24 +542,35 @@ bool Inliner::runOnSCC(CallGraphSCC &SCC) {
             InlineHistoryIncludes(Callee, InlineHistoryID, InlineHistory))
           continue;
         
-        
-        // If the policy determines that we should inline this function,
-        // try to do so.
-        if (!shouldInline(CS))
-          continue;
+        LLVMContext &CallerCtx = Caller->getContext();
 
         // Get DebugLoc to report. CS will be invalid after Inliner.
         DebugLoc DLoc = CS.getInstruction()->getDebugLoc();
 
+        // If the policy determines that we should inline this function,
+        // try to do so.
+        if (!shouldInline(CS)) {
+          emitOptimizationRemarkMissed(CallerCtx, DEBUG_TYPE, *Caller, DLoc,
+                                       Twine(Callee->getName() +
+                                             " will not be inlined into " +
+                                             Caller->getName()));
+          continue;
+        }
+
         // Attempt to inline the function.
         if (!InlineCallIfPossible(CS, InlineInfo, InlinedArrayAllocas,
-                                  InlineHistoryID, InsertLifetime, DL))
+                                  InlineHistoryID, InsertLifetime, DL)) {
+          emitOptimizationRemarkMissed(CallerCtx, DEBUG_TYPE, *Caller, DLoc,
+                                       Twine(Callee->getName() +
+                                             " will not be inlined into " +
+                                             Caller->getName()));
           continue;
+        }
         ++NumInlined;
 
         // Report the inline decision.
-        Caller->getContext().emitOptimizationRemark(
-            DEBUG_TYPE, *Caller, DLoc,
+        emitOptimizationRemark(
+            CallerCtx, DEBUG_TYPE, *Caller, DLoc,
             Twine(Callee->getName() + " inlined into " + Caller->getName()));
 
         // If inlining this function gave us any new call sites, throw them
diff --git a/llvm/lib/Transforms/Scalar/TailRecursionElimination.cpp b/llvm/lib/Transforms/Scalar/TailRecursionElimination.cpp
index 946af800776..05b9892470b 100644
--- a/llvm/lib/Transforms/Scalar/TailRecursionElimination.cpp
+++ b/llvm/lib/Transforms/Scalar/TailRecursionElimination.cpp
@@ -64,6 +64,7 @@
 #include "llvm/IR/CallSite.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/DerivedTypes.h"
+#include "llvm/IR/DiagnosticInfo.h"
 #include "llvm/IR/Function.h"
 #include "llvm/IR/Instructions.h"
 #include "llvm/IR/IntrinsicInst.h"
@@ -318,8 +319,8 @@ bool TailCallElim::markTails(Function &F, bool &AllCallsAreTailCalls) {
           break;
         }
         if (SafeToTail) {
-          F.getContext().emitOptimizationRemark(
-              "tailcallelim", F, CI->getDebugLoc(),
+          emitOptimizationRemark(
+              F.getContext(), "tailcallelim", F, CI->getDebugLoc(),
               "marked this readnone call a tail call candidate");
           CI->setTailCall();
           Modified = true;
@@ -365,9 +366,9 @@ bool TailCallElim::markTails(Function &F, bool &AllCallsAreTailCalls) {
     if (Visited[CI->getParent()] != ESCAPED) {
       // If the escape point was part way through the block, calls after the
       // escape point wouldn't have been put into DeferredTails.
-      F.getContext().emitOptimizationRemark(
-          "tailcallelim", F, CI->getDebugLoc(),
-          "marked this call a tail call candidate");
+      emitOptimizationRemark(F.getContext(), "tailcallelim", F,
+                             CI->getDebugLoc(),
+                             "marked this call a tail call candidate");
       CI->setTailCall();
       Modified = true;
     } else {
@@ -678,9 +679,8 @@ bool TailCallElim::EliminateRecursiveTailCall(CallInst *CI, ReturnInst *Ret,
   BasicBlock *BB = Ret->getParent();
   Function *F = BB->getParent();
 
-  F->getContext().emitOptimizationRemark(
-      "tailcallelim", *F, CI->getDebugLoc(),
-      "transforming tail recursion to loop");
+  emitOptimizationRemark(F->getContext(), "tailcallelim", *F, CI->getDebugLoc(),
+                         "transforming tail recursion to loop");
 
   // OK! We can transform this tail call.  If this is the first one found,
   // create the new entry block, allowing us to branch back to the old entry.
diff --git a/llvm/lib/Transforms/Utils/LoopUnroll.cpp b/llvm/lib/Transforms/Utils/LoopUnroll.cpp
index faaab5c12e4..d953e307310 100644
--- a/llvm/lib/Transforms/Utils/LoopUnroll.cpp
+++ b/llvm/lib/Transforms/Utils/LoopUnroll.cpp
@@ -24,6 +24,7 @@
 #include "llvm/Analysis/ScalarEvolution.h"
 #include "llvm/IR/BasicBlock.h"
 #include "llvm/IR/Dominators.h"
+#include "llvm/IR/DiagnosticInfo.h"
 #include "llvm/IR/LLVMContext.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/raw_ostream.h"
@@ -237,9 +238,9 @@ bool llvm::UnrollLoop(Loop *L, unsigned Count, unsigned TripCount,
   if (CompletelyUnroll) {
     DEBUG(dbgs() << "COMPLETELY UNROLLING loop %" << Header->getName()
           << " with trip count " << TripCount << "!\n");
-    Ctx.emitOptimizationRemark(DEBUG_TYPE, *F, LoopLoc,
-                               Twine("completely unrolled loop with ") +
-                                   Twine(TripCount) + " iterations");
+    emitOptimizationRemark(Ctx, DEBUG_TYPE, *F, LoopLoc,
+                           Twine("completely unrolled loop with ") +
+                               Twine(TripCount) + " iterations");
   } else {
     DEBUG(dbgs() << "UNROLLING loop %" << Header->getName()
           << " by " << Count);
@@ -255,7 +256,7 @@ bool llvm::UnrollLoop(Loop *L, unsigned Count, unsigned TripCount,
       DiagMsg.concat(" with run-time trip count");
     }
     DEBUG(dbgs() << "!\n");
-    Ctx.emitOptimizationRemark(DEBUG_TYPE, *F, LoopLoc, DiagMsg);
+    emitOptimizationRemark(Ctx, DEBUG_TYPE, *F, LoopLoc, DiagMsg);
   }
 
   bool ContinueOnTrue = L->contains(BI->getSuccessor(0));
diff --git a/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp b/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
index 0fd185816fd..3b61bb575a8 100644
--- a/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
+++ b/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
@@ -20,6 +20,7 @@
 #include "llvm/ADT/Triple.h"
 #include "llvm/Analysis/ValueTracking.h"
 #include "llvm/IR/DataLayout.h"
+#include "llvm/IR/DiagnosticInfo.h"
 #include "llvm/IR/Function.h"
 #include "llvm/IR/IRBuilder.h"
 #include "llvm/IR/IntrinsicInst.h"
@@ -789,9 +790,9 @@ struct StrLenOpt : public LibCallOptimization {
       uint64_t LenTrue = GetStringLength(SI->getTrueValue());
       uint64_t LenFalse = GetStringLength(SI->getFalseValue());
       if (LenTrue && LenFalse) {
-        Context->emitOptimizationRemark(
-            "simplify-libcalls", *Caller, SI->getDebugLoc(),
-            "folded strlen(select) to select of constants");
+        emitOptimizationRemark(*Context, "simplify-libcalls", *Caller,
+                               SI->getDebugLoc(),
+                               "folded strlen(select) to select of constants");
         return B.CreateSelect(SI->getCondition(),
                               ConstantInt::get(CI->getType(), LenTrue-1),
                               ConstantInt::get(CI->getType(), LenFalse-1));
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index d57fae34dd6..34d8a1053fa 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -67,6 +67,7 @@
 #include "llvm/IR/DataLayout.h"
 #include "llvm/IR/DebugInfo.h"
 #include "llvm/IR/DerivedTypes.h"
+#include "llvm/IR/DiagnosticInfo.h"
 #include "llvm/IR/Dominators.h"
 #include "llvm/IR/Function.h"
 #include "llvm/IR/IRBuilder.h"
@@ -1213,10 +1214,10 @@ struct LoopVectorize : public FunctionPass {
       DEBUG(dbgs() << "LV: Trying to at least unroll the loops.\n");
 
       // Report the unrolling decision.
-      F->getContext().emitOptimizationRemark(
-          DEBUG_TYPE, *F, L->getStartLoc(),
-          Twine("unrolled with interleaving factor " + Twine(UF) +
-                " (vectorization not beneficial)"));
+      emitOptimizationRemark(F->getContext(), DEBUG_TYPE, *F, L->getStartLoc(),
+                             Twine("unrolled with interleaving factor " +
+                                   Twine(UF) +
+                                   " (vectorization not beneficial)"));
 
       // We decided not to vectorize, but we may want to unroll.
       InnerLoopUnroller Unroller(L, SE, LI, DT, DL, TLI, UF);
@@ -1228,8 +1229,8 @@ struct LoopVectorize : public FunctionPass {
       ++LoopsVectorized;
 
       // Report the vectorization decision.
-      F->getContext().emitOptimizationRemark(
-          DEBUG_TYPE, *F, L->getStartLoc(),
+      emitOptimizationRemark(
+          F->getContext(), DEBUG_TYPE, *F, L->getStartLoc(),
           Twine("vectorized loop (vectorization factor: ") + Twine(VF.Width) +
               ", unrolling interleave factor: " + Twine(UF) + ")");
     }