[PATCH] D42946: Verify profile data confirms large loop trip counts.

[Mircea Trofin via Phabricator via llvm-commits]

mtrofin created this revision.
mtrofin added reviewers: twoh, mkuper.
Herald added a subscriber: llvm-commits.

Loops with inequality comparers, such as:

  // unsigned bound
  for (unsigned i = 1; i < bound; ++i) {...}

have getSmallConstantMaxTripCount report a large maximum static
trip count - in this case, 0xffff fffe. However, profiling info
may show that the trip count is much smaller, and thus
counter-recommend vectorization.

This change:

- flips loop-vectorize-with-block-frequency on by default.
- validates profiled loop frequency data supports vectorization, when static info appears to not counter-recommend it. Absence of profile data means we rely on static data, just as we've done so far.


Repository:
  rL LLVM

https://reviews.llvm.org/D42946

Files:
  lib/Transforms/Vectorize/LoopVectorize.cpp
  test/Transforms/LoopVectorize/tripcount.ll


Index: test/Transforms/LoopVectorize/tripcount.ll
===================================================================
--- test/Transforms/LoopVectorize/tripcount.ll
+++ test/Transforms/LoopVectorize/tripcount.ll
@@ -57,7 +57,7 @@
 }
 
 define i32 @foo_low_trip_count3(i1 %cond, i32 %bound) !prof !0 {
-; The loop has low invocation count compare to the function invocation count, 
+; The loop has low invocation count compare to the function invocation count,
 ; but has a high trip count per invocation. Vectorize it.
 
 ; CHECK-LABEL: @foo_low_trip_count3(
@@ -84,6 +84,30 @@
   ret i32 0
 }
 
+define i32 @foo_low_trip_count_icmp_sgt(i32 %bound) {
+; Simple loop with low tripcount and inequality test for exit.
+; Should not be vectorized.
+
+; CHECK-LABEL: @foo_low_trip_count_icmp_sgt(
+; CHECK-NOT: <{{[0-9]+}} x i8>
+
+entry:
+  br label %for.body
+
+for.body:                                         ; preds = %for.body, %entry
+  %i.08 = phi i32 [ 0, %entry ], [ %inc, %for.body ]
+  %arrayidx = getelementptr inbounds [32 x i8], [32 x i8]* @tab, i32 0, i32 %i.08
+  %0 = load i8, i8* %arrayidx, align 1
+  %cmp1 = icmp eq i8 %0, 0
+  %. = select i1 %cmp1, i8 2, i8 1
+  store i8 %., i8* %arrayidx, align 1
+  %inc = add nsw i32 %i.08, 1
+  %exitcond = icmp sgt i32 %i.08, %bound
+  br i1 %exitcond, label %for.end, label %for.body, !prof !1
+
+for.end:                                          ; preds = %for.body
+  ret i32 0
+}
 
 !0 = !{!"function_entry_count", i64 100}
 !1 = !{!"branch_weights", i32 100, i32 0}
Index: lib/Transforms/Vectorize/LoopVectorize.cpp
===================================================================
--- lib/Transforms/Vectorize/LoopVectorize.cpp
+++ lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -208,7 +208,7 @@
         "The cost of a loop that is considered 'small' by the interleaver."));
 
 static cl::opt<bool> LoopVectorizeWithBlockFrequency(
-    "loop-vectorize-with-block-frequency", cl::init(false), cl::Hidden,
+    "loop-vectorize-with-block-frequency", cl::init(true), cl::Hidden,
     cl::desc("Enable the use of the block frequency analysis to access PGO "
              "heuristics minimizing code growth in cold regions and being more "
              "aggressive in hot regions."));
@@ -8350,7 +8350,10 @@
   unsigned ExpectedTC = SE->getSmallConstantMaxTripCount(L);
   bool HasExpectedTC = (ExpectedTC > 0);
 
-  if (!HasExpectedTC && LoopVectorizeWithBlockFrequency) {
+  // ExpectedTC may be large because it's bound by a variable. Check
+  // profiling information to validate we should vectorize.
+  if ((!HasExpectedTC || ExpectedTC >= TinyTripCountVectorThreshold)
+      && LoopVectorizeWithBlockFrequency) {
     auto EstimatedTC = getLoopEstimatedTripCount(L);
     if (EstimatedTC) {
       ExpectedTC = *EstimatedTC;


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D42946.132934.patch
Type: text/x-patch
Size: 2795 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180206/a54d10bd/attachment.bin>