[PATCH] D83408: [SVE] Disable some BUILD_VECTOR related code generator features.

[Paul Walker via Phabricator via llvm-commits]

paulwalker-arm created this revision.
Herald added subscribers: llvm-commits, psnobl, hiraditya, tschuett.
Herald added a reviewer: rengolin.
Herald added a reviewer: efriedma.
Herald added a project: LLVM.

Fixed length vector code generation for SVE does not yet custom
lower BUILD_VECTOR and instead relies on expansion.  At the same
time custom lowering for VECTOR_SHUFFLE is also not available so
this patch disables using VECTOR_SHUFFLE to expand BUILD_VECTOR.

Related to this it also prevents the merging of stores after
legalisation because this only works when BUILD_VECTOR is either
legal or can be elminated.  When this is not the case the code
generator enters an infinite legalisation loop.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D83408

Files:
  llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
  llvm/lib/Target/AArch64/AArch64ISelLowering.h
  llvm/test/CodeGen/AArch64/sve-fixed-length-shuffles.ll


Index: llvm/test/CodeGen/AArch64/sve-fixed-length-shuffles.ll
===================================================================
--- llvm/test/CodeGen/AArch64/sve-fixed-length-shuffles.ll
+++ llvm/test/CodeGen/AArch64/sve-fixed-length-shuffles.ll
@@ -3,6 +3,18 @@
 
 target triple = "aarch64-unknown-linux-gnu"
 
+; Currently there is no custom lowering for vector shuffles operating on types
+; bigger than NEON. However, having no support opens us up to a code generator
+; hang when expanding BUILD_VECTOR. Here we just validate the promblematic case
+; successfully exits code generation.
+define void @hang_when_merging_stores_after_legalisation(<8 x i32>* %a, <2 x i32> %b) #0 {
+; CHECK-LABEL: hang_when_merging_stores_after_legalisation:
+  %splat = shufflevector <2 x i32> %b, <2 x i32> undef, <8 x i32> zeroinitializer
+  %interleaved.vec = shufflevector <8 x i32> %splat, <8 x i32> undef, <8 x i32> <i32 0, i32 4, i32 1, i32 5, i32 2, i32 6, i32 3, i32 7>
+  store <8 x i32> %interleaved.vec, <8 x i32>* %a, align 4
+  ret void
+}
+
 ; NOTE: Currently all CONCAT_VECTORS get expanded so there's little point in
 ; validating all combinations of vector type.
 
Index: llvm/lib/Target/AArch64/AArch64ISelLowering.h
===================================================================
--- llvm/lib/Target/AArch64/AArch64ISelLowering.h
+++ llvm/lib/Target/AArch64/AArch64ISelLowering.h
@@ -734,6 +734,18 @@
 
   bool fallBackToDAGISel(const Instruction &Inst) const override;
 
+  bool
+  shouldExpandBuildVectorWithShuffles(EVT VT,
+                                      unsigned DefinedValues) const override;
+
+  /// SVE code generation for fixed length vectors does not custom lower
+  /// BUILD_VECTOR. This makes BUILD_VECTOR legalisation a source of stores to
+  /// merge. However, merging them creates a BUILD_VECTOR that is just as
+  /// illegal as the original, thus leading to an infinite legalisation loop.
+  bool mergeStoresAfterLegalization(EVT VT) const override {
+    return !useSVEForFixedLengthVectors();
+  }
+
 private:
   /// Keep a pointer to the AArch64Subtarget around so that we can
   /// make the right decision when generating code for different targets.
Index: llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
===================================================================
--- llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -3564,6 +3564,16 @@
   }
 }
 
+// VECTOR_SHUFFLE is not legal for vectors bigger than NEON, so we cannot use
+// them to expand BUILD_VECTOR.
+bool AArch64TargetLowering::shouldExpandBuildVectorWithShuffles(
+    EVT VT, unsigned DefinedValues) const {
+  if (useSVEForFixedLengthVectorVT(VT))
+    return false;
+
+  return TargetLowering::shouldExpandBuildVectorWithShuffles(VT, DefinedValues);
+}
+
 bool AArch64TargetLowering::useSVEForFixedLengthVectors() const {
   // Prefer NEON unless larger SVE registers are available.
   return Subtarget->hasSVE() && Subtarget->getMinSVEVectorSizeInBits() >= 256;


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D83408.276471.patch
Type: text/x-patch
Size: 3035 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200708/06305f9e/attachment.bin>