[ModuloSchedule][AArch64] Implement modulo variable expansion for pipelining (#65609)

Modulo variable expansion is a technique that resolves overlap of
variable lifetimes by unrolling. The existing implementation solves it
by making a copy by move instruction for processors with ordinary
registers such as Arm and x86. This method may result in a very large
number of move instructions, which can cause performance problems.

Modulo variable expansion is enabled by specifying -pipeliner-mve-cg. A
backend must implement some newly defined interfaces in
PipelinerLoopInfo. They were implemented for AArch64.

Discourse thread:
https://discourse.llvm.org/t/implementing-modulo-variable-expansion-for-machinepipeliner

1 files changed, 9 insertions, 0 deletions

diff --git a/llvm/lib/CodeGen/MachinePipeliner.cpp b/llvm/lib/CodeGen/MachinePipeliner.cpp
index 32f65f0..6c24cfc 100644
--- a/llvm/lib/CodeGen/MachinePipeliner.cpp
+++ b/llvm/lib/CodeGen/MachinePipeliner.cpp
@@ -192,6 +192,10 @@ static cl::opt<int>
                       cl::desc("Margin representing the unused percentage of "
                                "the register pressure limit"));
 
+static cl::opt<bool>
+    MVECodeGen("pipeliner-mve-cg", cl::Hidden, cl::init(false),
+               cl::desc("Use the MVE code generator for software pipelining"));
+
 namespace llvm {
 
 // A command line option to enable the CopyToPhi DAG mutation.
@@ -677,6 +681,11 @@ void SwingSchedulerDAG::schedule() {
   if (ExperimentalCodeGen && NewInstrChanges.empty()) {
     PeelingModuloScheduleExpander MSE(MF, MS, &LIS);
     MSE.expand();
+  } else if (MVECodeGen && NewInstrChanges.empty() &&
+             LoopPipelinerInfo->isMVEExpanderSupported() &&
+             ModuloScheduleExpanderMVE::canApply(Loop)) {
+    ModuloScheduleExpanderMVE MSE(MF, MS, LIS);
+    MSE.expand();
   } else {
     ModuloScheduleExpander MSE(MF, MS, LIS, std::move(NewInstrChanges));
     MSE.expand();