[PATCH] D104689: [LoopDeletion] Benefit from branches by undef conditions when symbolically executing 1st iteration

[Max Kazantsev via Phabricator via llvm-commits]

mkazantsev created this revision.
mkazantsev added reviewers: lebedev.ri, nikic, reames, aqjune.
Herald added a subscriber: hiraditya.
mkazantsev requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

We can exploit branches by `undef` condition. Frankly, the LangRef says that
such branches are UB, so we can assume that all outgoing edges of such blocks
are dead.

However, from practical perspective, we know that this is not supported correctly
in some other places. So we are being conservative about it.

Branch by undef is treated in the following way:

- If it is a loop-exiting branch, we always assume it exits the loop;
- If not, we arbitrarily assume it takes `true` value.


https://reviews.llvm.org/D104689

Files:
  llvm/lib/Transforms/Scalar/LoopDeletion.cpp
  llvm/test/Transforms/LoopDeletion/eval_first_iteration.ll


Index: llvm/test/Transforms/LoopDeletion/eval_first_iteration.ll
===================================================================
--- llvm/test/Transforms/LoopDeletion/eval_first_iteration.ll
+++ llvm/test/Transforms/LoopDeletion/eval_first_iteration.ll
@@ -887,14 +887,14 @@
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    br label [[LOOP:%.*]]
 ; CHECK:       loop:
-; CHECK-NEXT:    [[SUM:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[SUM_NEXT:%.*]], [[BACKEDGE:%.*]] ]
+; CHECK-NEXT:    [[SUM:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    [[SUB:%.*]] = sub i32 4, [[SUM]]
 ; CHECK-NEXT:    [[IS_POSITIVE:%.*]] = icmp sgt i32 [[SUB]], 0
 ; CHECK-NEXT:    br i1 [[IS_POSITIVE]], label [[IF_TRUE:%.*]], label [[IF_FALSE:%.*]]
 ; CHECK:       if.true:
 ; CHECK-NEXT:    br i1 undef, label [[IF_TRUE_1:%.*]], label [[IF_TRUE_2:%.*]]
 ; CHECK:       if.true.1:
-; CHECK-NEXT:    br label [[BACKEDGE]]
+; CHECK-NEXT:    br label [[BACKEDGE:%.*]]
 ; CHECK:       if.true.2:
 ; CHECK-NEXT:    br label [[BACKEDGE]]
 ; CHECK:       if.false:
@@ -905,9 +905,11 @@
 ; CHECK-NEXT:    br label [[BACKEDGE]]
 ; CHECK:       backedge:
 ; CHECK-NEXT:    [[MERGE_PHI:%.*]] = phi i32 [ 0, [[IF_FALSE_1]] ], [ 0, [[IF_FALSE_2]] ], [ undef, [[IF_TRUE_1]] ], [ undef, [[IF_TRUE_2]] ]
-; CHECK-NEXT:    [[SUM_NEXT]] = add i32 [[SUM]], [[MERGE_PHI]]
+; CHECK-NEXT:    [[SUM_NEXT:%.*]] = add i32 [[SUM]], [[MERGE_PHI]]
 ; CHECK-NEXT:    [[LOOP_COND:%.*]] = icmp ne i32 [[SUM_NEXT]], 4
-; CHECK-NEXT:    br i1 [[LOOP_COND]], label [[LOOP]], label [[DONE:%.*]]
+; CHECK-NEXT:    br i1 [[LOOP_COND]], label [[BACKEDGE_LOOP_CRIT_EDGE:%.*]], label [[DONE:%.*]]
+; CHECK:       backedge.loop_crit_edge:
+; CHECK-NEXT:    unreachable
 ; CHECK:       done:
 ; CHECK-NEXT:    [[SUM_NEXT_LCSSA:%.*]] = phi i32 [ [[SUM_NEXT]], [[BACKEDGE]] ]
 ; CHECK-NEXT:    ret i32 [[SUM_NEXT_LCSSA]]
Index: llvm/lib/Transforms/Scalar/LoopDeletion.cpp
===================================================================
--- llvm/lib/Transforms/Scalar/LoopDeletion.cpp
+++ llvm/lib/Transforms/Scalar/LoopDeletion.cpp
@@ -333,13 +333,35 @@
     // Can we prove constant true or false for this condition?
     LHS = getValueOnFirstIteration(LHS, FirstIterValue, SQ);
     RHS = getValueOnFirstIteration(RHS, FirstIterValue, SQ);
-    auto *KnownCondition =
-        dyn_cast_or_null<ConstantInt>(SimplifyICmpInst(Pred, LHS, RHS, SQ));
+    auto *KnownCondition = SimplifyICmpInst(Pred, LHS, RHS, SQ);
     if (!KnownCondition) {
+      // Failed to simplify.
       MarkAllSuccessorsLive(BB);
       continue;
     }
-    if (KnownCondition->isAllOnesValue())
+    if (isa<UndefValue>(KnownCondition)) {
+      // TODO: According to langref, branching by undef is undefined behavior.
+      // It means that, theoretically, we should be able to just continue
+      // without marking any successors as live. However, we are not certain
+      // how correct our compiler is at handling such cases. So we are being
+      // very conservative here.
+      //
+      // If there is a non-loop successor, always assume this branch leaves the
+      // loop. Otherwise, arbitrarily take IfTrue.
+      //
+      // Once we are certain that branching by undef is handled correctly by
+      // other transforms, we should not mark any successors live here.
+      if (L->contains(IfTrue) && L->contains(IfFalse))
+        MarkLiveEdge(BB, IfTrue);
+      continue;
+    }
+    auto *ConstCondition = dyn_cast<ConstantInt>(KnownCondition);
+    if (!ConstCondition) {
+      // Non-constant condition, cannot analyze any further.
+      MarkAllSuccessorsLive(BB);
+      continue;
+    }
+    if (ConstCondition->isAllOnesValue())
       MarkLiveEdge(BB, IfTrue);
     else
       MarkLiveEdge(BB, IfFalse);


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D104689.353555.patch
Type: text/x-patch
Size: 3755 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210622/1546c282/attachment.bin>