Subzero: Crosstest test_arith properly tests i8/i16.

crosstest/crosstest.cfg

 [simple_loop]
 driver: simple_loop_main.c
 test: simple_loop.c
 
 [mem_intrin]
 driver: mem_intrin_main.cpp
 test: mem_intrin.cpp
 flags: --sz=-fmem-intrin-opt
 
 [test_arith]
 driver: test_arith_main.cpp
-test: test_arith.cpp test_arith_frem.ll test_arith_sqrt.ll test_arith_fabs.ll
+# Note: Originally test_arith.cpp was used to generate the bulk of the tests.
+# However, the C++ integral promotion rules mean that most of the time i8 and
+# i16 operations are not actually tested; instead, they are extended to i32 and
+# only the i32 operation is tested.  To help fix this, the relevant parts of
+# test_arith.cpp are commented out, and instead a python script auto-generates
+# test_arith_ll.ll which contains the proper functionality.
+#test: test_arith.cpp test_arith_frem.ll test_arith_sqrt.ll test_arith_fabs.ll
+test: test_arith_ll.ll test_arith.cpp test_arith_sqrt.ll test_arith_fabs.ll
 
 [test_bitmanip]
 driver: test_bitmanip_main.cpp
 test: test_bitmanip.cpp test_bitmanip_intrin.ll
 
 [test_calling_conv]
 driver: test_calling_conv_main.cpp
 test: test_calling_conv.cpp
 
 [test_cast]
@@ skipping 27 matching lines @@
 # bitcode patterns.
 flags: --clang-opt=0
 
 [test_sync_atomic]
 driver: test_sync_atomic_main.cpp
 test: test_sync_atomic.cpp
 
 [test_vector_ops]
 driver: test_vector_ops_main.cpp
 test: test_vector_ops.cpp test_vector_ops_ll.ll