BlockPlacement: add a flag to force cold block outlining w/o a profile.

[iteratee]

Summary

Add a flag that forces cold blocks to be evicted from a loop. NFC.

[iteratee]

Known issues:
needs a test case, and needs an adjustment to the frequency for non-profile functions.

[rsmith]

LGTM

[rsmith]

FWIW, I see a 10% performance improvement on one of our internal benchmarks with this patch + force-precise-rotation-cost, compared to force-precise-rotation-cost alone.

I wrote a testcase that covers this flag in https://reviews.llvm.org/D36338

[iteratee]

Submitted in r310129

lib/CodeGen/MachineBlockPlacement.cpp

 // - Outlining: placement of a basic block outside the chain or hot path.
 
 static cl::opt<unsigned> LoopToColdBlockRatio(
     "loop-to-cold-block-ratio",
     cl::desc("Outline loop blocks from loop chain if (frequency of loop) / "
              "(frequency of block) is greater than this ratio"),
     cl::init(5), cl::Hidden);
 
+static cl::opt<bool> ForceLoopColdBlock(
+    "force-loop-cold-block",
+    cl::desc("Force outlining cold blocks from loops."),
+    cl::init(false), cl::Hidden);
+
 static cl::opt<bool>
     PreciseRotationCost("precise-rotation-cost",
                         cl::desc("Model the cost of loop rotation more "
                                  "precisely by using profile data."),
                         cl::init(false), cl::Hidden);
 static cl::opt<bool>
     ForcePreciseRotationCost("force-precise-rotation-cost",
                              cl::desc("Force the use of precise cost "
MachineBlockPlacement::collectLoopBlockSet(const MachineLoop &L) {
   // Collect the sum of frequencies of incoming edges to the loop header from
   // outside. If we treat the loop as a super block, this is the frequency of
   // the loop. Then for each block in the loop, we calculate the ratio between
   // its frequency and the frequency of the loop block. When it is too small,
   // don't add it to the loop chain. If there are outer loops, then this block
   // will be merged into the first outer loop chain for which this block is not
   // cold anymore. This needs precise profile data and we only do this when
   // profile data is available.
-  if (F->getFunction()->getEntryCount()) {
+  if (F->getFunction()->getEntryCount() || ForceLoopColdBlock) {
     BlockFrequency LoopFreq(0);
     for (auto LoopPred : L.getHeader()->predecessors())
       if (!L.contains(LoopPred))
         LoopFreq += MBFI->getBlockFreq(LoopPred) *
                     MBPI->getEdgeProbability(LoopPred, L.getHeader());
 
     for (MachineBasicBlock *LoopBB : L.getBlocks()) {
       auto Freq = MBFI->getBlockFreq(LoopBB).getFrequency();