Subzero: ARM32: lowering of vector insert and extract.

src/IceTypes.cpp

Index: src/IceTypes.cpp
diff --git a/src/IceTypes.cpp b/src/IceTypes.cpp
index 39eff2dfedaa2ea65ae80fbead1e579f648dbca3..b11313ae58f81086096544b718899d3c0433f8c4 100644
--- a/src/IceTypes.cpp
+++ b/src/IceTypes.cpp
@@ -145,6 +145,8 @@ size_t typeWidthInBytes(Type Ty) {
   return (Shift < 0) ? 0 : 1 << Shift;
 }
 
+size_t typeWidthInBits(Type Ty) { return typeWidthInBytes(Ty) * 8; }

[Jim Stichnoth, 2016/02/03 15:28:38]

Probably CHAR_BIT would be more appropriate than 8.

[Eric Holk, 2016/02/03 21:02:22]

Agreed. I removed this function and inlined it where needed, following John's
suggestion.

+
 int8_t typeWidthInBytesLog2(Type Ty) {
   size_t Index = static_cast<size_t>(Ty);
   if (Index < IceType_NUM)