When inferring the pointer alignment, if the global doesn't have an initializer

and the alignment is 0 (i.e., it's defined globally in one file and declared in
another file) it could get an alignment which is larger than the ABI allows for
that type, resulting in aligned moves being used for unaligned loads.

For instance, in file A.c:

   struct S s;

In file B.c:
   struct {
     // something long
   };
   extern S s;

   void foo() {
     struct S p = s;
     // ...
   }

this copy is a 'memcpy' which is turned into a series of 'movaps' instructions
on X86. But this is wrong, because 'struct S' has alignment of 4, not 16.

llvm-svn: 140902

2 files changed, 28 insertions, 0 deletions

diff --git a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
index d617d6710c0..6a549e8760a 100644
--- a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
+++ b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
@@ -6537,6 +6537,8 @@ unsigned SelectionDAG::InferPtrAlignment(SDValue Ptr) const {
           Align = TD->getPreferredAlignment(GVar);
         }
       }
+      if (!Align)
+        Align = TLI.getTargetData()->getABITypeAlignment(GV->getType());
     }
     return MinAlign(Align, GVOffset);
   }
diff --git a/llvm/test/CodeGen/X86/alignment-2.ll b/llvm/test/CodeGen/X86/alignment-2.ll
new file mode 100644
index 00000000000..835cf6275b2
--- /dev/null
+++ b/llvm/test/CodeGen/X86/alignment-2.ll
@@ -0,0 +1,26 @@
+; RUN: llc < %s -mtriple i386-apple-darwin10 | not grep movaps
+; <rdar://problem/10058036>
+
+%struct._psqlSettings = type { %struct.pg_conn*, i32, %struct.__sFILE*, i8, %struct.printQueryOpt, i8*, i8, i32, %struct.__sFILE*, i8, i32, i8*, i8*, i8*, i64, i8, %struct.__sFILE*, %struct._variable*, i8, i8, i8, i8, i8, i32, i32, i32, i32, i32, i8*, i8*, i8*, i32 }
+%struct.pg_conn = type opaque
+%struct.__sFILE = type { i8*, i32, i32, i16, i16, %struct.__sbuf, i32, i8*, i32 (i8*)*, i32 (i8*, i8*, i32)*, i64 (i8*, i64, i32)*, i32 (i8*, i8*, i32)*, %struct.__sbuf, %struct.__sFILEX*, i32, [3 x i8], [1 x i8], %struct.__sbuf, i32, i64 }
+%struct.__sbuf = type { i8*, i32 }
+%struct.__sFILEX = type opaque
+%struct.printQueryOpt = type { %struct.printTableOpt, i8*, i8, i8*, i8**, i8, i8, i8* }
+%struct.printTableOpt = type { i32, i8, i16, i16, i8, i8, i8, i32, %struct.printTextFormat*, i8*, i8*, i8, i8*, i32, i32, i32 }
+%struct.printTextFormat = type { i8*, [4 x %struct.printTextLineFormat], i8*, i8*, i8*, i8*, i8*, i8*, i8*, i8*, i8*, i8 }
+%struct.printTextLineFormat = type { i8*, i8*, i8*, i8* }
+%struct._variable = type { i8*, i8*, void (i8*)*, %struct._variable* }
+%struct.pg_result = type opaque
+
+@pset = external global %struct._psqlSettings
+
+define signext i8 @do_lo_list() nounwind optsize ssp {
+bb:
+  %myopt = alloca %struct.printQueryOpt, align 4
+  %tmp = bitcast %struct.printQueryOpt* %myopt to i8*
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp, i8* bitcast (%struct.printQueryOpt* getelementptr inbounds (%struct._psqlSettings* @pset, i32 0, i32 4) to i8*), i32 76, i32 4, i1 false)
+  ret i8 0
+}
+
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind