[ARM GlobalISel] Support G_FDIV for s32 and s64

TableGen already generates code for selecting a G_FDIV, so we only need
to add a test.

For the legalizer and reg bank select, we do the same thing as for the
other floating point binary operations: either mark as legal if we have
a FP unit or lower to a libcall, and map to the floating point
registers.

llvm-svn: 318915
diff --git a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
index c54c715..bb2e615 100644
--- a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
+++ b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
@@ -97,6 +97,9 @@
   case TargetOpcode::G_FMUL:
     assert((Size == 32 || Size == 64) && "Unsupported size");
     return Size == 64 ? RTLIB::MUL_F64 : RTLIB::MUL_F32;
+  case TargetOpcode::G_FDIV:
+    assert((Size == 32 || Size == 64) && "Unsupported size");
+    return Size == 64 ? RTLIB::DIV_F64 : RTLIB::DIV_F32;
   case TargetOpcode::G_FREM:
     return Size == 64 ? RTLIB::REM_F64 : RTLIB::REM_F32;
   case TargetOpcode::G_FPOW:
@@ -154,6 +157,7 @@
   case TargetOpcode::G_FADD:
   case TargetOpcode::G_FSUB:
   case TargetOpcode::G_FMUL:
+  case TargetOpcode::G_FDIV:
   case TargetOpcode::G_FPOW:
   case TargetOpcode::G_FREM: {
     Type *HLTy = Size == 64 ? Type::getDoubleTy(Ctx) : Type::getFloatTy(Ctx);