[PATCH] D140697: [MemCpyOpt] Merge succeeding undefs while attempting a `memset`

[Jamie Hill-Daniel via Phabricator via llvm-commits]

clubby789 updated this revision to Diff 485423.
clubby789 added a comment.

clang-format


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D140697/new/

https://reviews.llvm.org/D140697

Files:
  llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp


Index: llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp
===================================================================
--- llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp
+++ llvm/lib/Transforms/Scalar/MemCpyOptimizer.cpp
@@ -111,24 +111,26 @@
   bool no_undef = true;
   for (Instruction *SI : TheStores) {
     if (auto *Store = dyn_cast<StoreInst>(SI)) {
-        Value *StoredVal = Store->getValueOperand();
-        if (!isa<UndefValue>(StoredVal))
-            non_undef_stores++;
-        else
-            no_undef = false;
+      Value *StoredVal = Store->getValueOperand();
+      if (!isa<UndefValue>(StoredVal))
+        non_undef_stores++;
+      else
+        no_undef = false;
     } else {
-        // If any of the stores are a memset, then it is always good to extend the
-        // memset.
-        return true;
+      // If any of the stores are a memset, then it is always good to extend the
+      // memset.
+      return true;
     }
   }
   // If we found more than 4 stores to merge or 16 bytes, use memset.
-  if (non_undef_stores >= 4 || (no_undef && End-Start >= 16)) return true;
+  if (non_undef_stores >= 4 || (no_undef && End - Start >= 16))
+    return true;
 
   // If there is nothing to merge, don't do anything.
   // Assume that the code generator is capable of merging pairs of stores
   // together if it wants to.
-  if (non_undef_stores <= 2) return false;
+  if (non_undef_stores <= 2)
+    return false;
 
   // If we have fewer than 8 stores, it can still be worthwhile to do this.
   // For example, merging 4 i8 stores into an i32 store is useful almost always.
@@ -152,7 +154,7 @@
   // If we will reduce the # stores (according to this heuristic), do the
   // transformation.  This encourages merging 4 x i8 -> i32 and 2 x i16 -> i32
   // etc.
-  return non_undef_stores > NumPointerStores+NumByteStores;
+  return non_undef_stores > NumPointerStores + NumByteStores;
 }
 
 namespace {


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D140697.485423.patch
Type: text/x-patch
Size: 1941 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20221227/a7ff6688/attachment.bin>