Move late partial-unrolling thresholds into the processor definitions

The old method used by X86TTI to determine partial-unrolling thresholds was
messy (because it worked by testing target features), and also would not
correctly identify the target CPU if certain target features were disabled.
After some discussions on IRC with Chandler et al., it was decided that the
processor scheduling models were the right containers for this information
(because it is often tied to special uop dispatch-buffer sizes).

This does represent a small functionality change:
 - For generic x86-64 (which uses the SB model and, thus, will get some
   unrolling).
 - For AMD cores (because they still currently use the SB scheduling model)
 - For Haswell (based on benchmarking by Louis Gerbarg, it was decided to bump
   the default threshold to 50; we're working on a test case for this).
Otherwise, nothing has changed for any other targets. The logic, however, has
been moved into BasicTTI, so other targets may now also opt-in to this
functionality simply by setting LoopMicroOpBufferSize in their processor
model definitions.

llvm-svn: 208289

1 files changed, 1 insertions, 0 deletions

diff --git a/llvm/utils/TableGen/SubtargetEmitter.cpp b/llvm/utils/TableGen/SubtargetEmitter.cpp
index 16bbdb7cd7e..06f869436f1 100644
--- a/llvm/utils/TableGen/SubtargetEmitter.cpp
+++ b/llvm/utils/TableGen/SubtargetEmitter.cpp
@@ -1195,6 +1195,7 @@ void SubtargetEmitter::EmitProcessorModels(raw_ostream &OS) {
     OS << "static const llvm::MCSchedModel " << PI->ModelName << "(\n";
     EmitProcessorProp(OS, PI->ModelDef, "IssueWidth", ',');
     EmitProcessorProp(OS, PI->ModelDef, "MicroOpBufferSize", ',');
+    EmitProcessorProp(OS, PI->ModelDef, "LoopMicroOpBufferSize", ',');
     EmitProcessorProp(OS, PI->ModelDef, "LoadLatency", ',');
     EmitProcessorProp(OS, PI->ModelDef, "HighLatency", ',');
     EmitProcessorProp(OS, PI->ModelDef, "MispredictPenalty", ',');