[llvm] [LoopInterchange] Add metadata to control loop-interchange (PR #127474)

[Michael Kruse via llvm-commits]

================
@@ -572,6 +578,70 @@ struct LoopInterchange {
 
 } // end anonymous namespace
 
+bool LoopInterchangeList::hasSupportedLoopDepth(
+    OptimizationRemarkEmitter &ORE) {
+  unsigned LoopNestDepth = ListEnd - ListBegin;
+  if (LoopNestDepth < MinLoopNestDepth || LoopNestDepth > MaxLoopNestDepth) {
+    LLVM_DEBUG(dbgs() << "Unsupported depth of loop nest " << LoopNestDepth
+                      << ", the supported range is [" << MinLoopNestDepth
+                      << ", " << MaxLoopNestDepth << "].\n");
+    Loop *OuterLoop = LoopList[ListBegin];
+    ORE.emit([&]() {
+      return OptimizationRemarkMissed(DEBUG_TYPE, "UnsupportedLoopNestDepth",
+                                      OuterLoop->getStartLoc(),
+                                      OuterLoop->getHeader())
+             << "Unsupported depth of loop nest, the supported range is ["
+             << std::to_string(MinLoopNestDepth) << ", "
+             << std::to_string(MaxLoopNestDepth) << "].\n";
+    });
+    return false;
+  }
+  return true;
+}
+
+// Check the metadata for interchange. The outermost one is taken into account
+// and nested ones are ignored. The metadata affects the entire loop nest such
+// that the outermost loop is the loop for which the metadata is specified. For
+// example, in the following example, the loop-interchange will be performed
+// only to the outermost two loops, and the second pragma is ignored.
+//
+// for (...)
+//   for (...)
+//     #pragma clang loop interchange(disable)
+//     for (...)
+//       #pragma clang loop interchange(enable)
+//       for (...)
+//         for (...)
+//           Stmt
+//
----------------
Meinersbur wrote:

If you look into `CGLoopInfo.cpp`, the metadata comonly have a `followup_*` property. This is so frontends can specify which other metadata/loop transformations to apply on the loops *after* unroll/vectorize/distribute/... has been apply. E.g. it could specify to apply `unroll_and_jam` on the i-loop, which becomes the outer loop after the interchange has been applied. For insterchange, this could be used to specify multiple pairwise exchanges, one following the other.

Unfortunately the system is only implemented incompletely. Only those series of loop transformations can be applied that follow the order of the pipeline. For instance, one cannot apply loop distribution after vectorization, because in LLVM's default pipeline, vectorize comes after loop distribution. This is an implementation detail that should not be visible to the user. The idea was once to apply passes until all such metadata has been consumed.




https://github.com/llvm/llvm-project/pull/127474