[PATCH] D108470: [OpenCL] Fix as_type3 invalid store creation

[Sven van Haastregt via Phabricator via cfe-commits]

svenvh created this revision.
svenvh added reviewers: Anastasia, jaykang10.
svenvh added a project: clang.
Herald added subscribers: ldrumm, yaxunl.
svenvh requested review of this revision.
Herald added a subscriber: cfe-commits.

With -fpreserve-vec3-type enabled, a cast was not created when
converting from a non-vec3 type to a vec3 type, even though a
conversion to vec3 was performed.  This resulted in creation of
invalid store instructions.

Related to D107963 <https://reviews.llvm.org/D107963>, but now for the to-vec3 
case.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D108470

Files:
  clang/lib/CodeGen/CGExprScalar.cpp
  clang/test/CodeGenOpenCL/preserve_vec3.cl


Index: clang/test/CodeGenOpenCL/preserve_vec3.cl
===================================================================
--- clang/test/CodeGenOpenCL/preserve_vec3.cl
+++ clang/test/CodeGenOpenCL/preserve_vec3.cl
@@ -53,3 +53,11 @@
   // CHECK: store <4 x i16> %[[ASTYPE]], <4 x i16> addrspace(1)* %[[OUT_BC]]
   *out = __builtin_astype(a, long);
 }
+
+void to_char3(int a, global char3 *out) {
+  // CHECK-LABEL: void @to_char3
+  // CHECK: %[[IN_BC:.*]] = bitcast i32 %a to <4 x i8>
+  // CHECK: %[[ASTYPE:.*]] = shufflevector <4 x i8> %[[IN_BC]], <4 x i8> 
poison, <3 x i32> <i32 0, i32 1, i32 2>
+  // CHECK: store <3 x i8> %[[ASTYPE]], <3 x i8> addrspace(1)* %out
+  *out = __builtin_astype(a, char3);
+}
Index: clang/lib/CodeGen/CGExprScalar.cpp
===================================================================
--- clang/lib/CodeGen/CGExprScalar.cpp
+++ clang/lib/CodeGen/CGExprScalar.cpp
@@ -4796,12 +4796,10 @@
   // to vec4 if the original type is not vec4, then a shuffle vector to
   // get a vec3.
   if (NumElementsSrc != 3 && NumElementsDst == 3) {
-    if (!CGF.CGM.getCodeGenOpts().PreserveVec3Type) {
-      auto *Vec4Ty = llvm::FixedVectorType::get(
-          cast<llvm::VectorType>(DstTy)->getElementType(), 4);
-      Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
-                                         Vec4Ty);
-    }
+    auto *Vec4Ty = llvm::FixedVectorType::get(
+        cast<llvm::VectorType>(DstTy)->getElementType(), 4);
+    Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
+                                       Vec4Ty);
 
     Src = ConvertVec3AndVec4(Builder, CGF, Src, 3);
     Src->setName("astype");



Index: clang/test/CodeGenOpenCL/preserve_vec3.cl
===================================================================
--- clang/test/CodeGenOpenCL/preserve_vec3.cl
+++ clang/test/CodeGenOpenCL/preserve_vec3.cl
@@ -53,3 +53,11 @@
   // CHECK: store <4 x i16> %[[ASTYPE]], <4 x i16> addrspace(1)* %[[OUT_BC]]
   *out = __builtin_astype(a, long);
 }
+
+void to_char3(int a, global char3 *out) {
+  // CHECK-LABEL: void @to_char3
+  // CHECK: %[[IN_BC:.*]] = bitcast i32 %a to <4 x i8>
+  // CHECK: %[[ASTYPE:.*]] = shufflevector <4 x i8> %[[IN_BC]], <4 x i8> poison, <3 x i32> <i32 0, i32 1, i32 2>
+  // CHECK: store <3 x i8> %[[ASTYPE]], <3 x i8> addrspace(1)* %out
+  *out = __builtin_astype(a, char3);
+}
Index: clang/lib/CodeGen/CGExprScalar.cpp
===================================================================
--- clang/lib/CodeGen/CGExprScalar.cpp
+++ clang/lib/CodeGen/CGExprScalar.cpp
@@ -4796,12 +4796,10 @@
   // to vec4 if the original type is not vec4, then a shuffle vector to
   // get a vec3.
   if (NumElementsSrc != 3 && NumElementsDst == 3) {
-    if (!CGF.CGM.getCodeGenOpts().PreserveVec3Type) {
-      auto *Vec4Ty = llvm::FixedVectorType::get(
-          cast<llvm::VectorType>(DstTy)->getElementType(), 4);
-      Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
-                                         Vec4Ty);
-    }
+    auto *Vec4Ty = llvm::FixedVectorType::get(
+        cast<llvm::VectorType>(DstTy)->getElementType(), 4);
+    Src = createCastsForTypeOfSameSize(Builder, CGF.CGM.getDataLayout(), Src,
+                                       Vec4Ty);
 
     Src = ConvertVec3AndVec4(Builder, CGF, Src, 3);
     Src->setName("astype");

_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits