------- Comment #1 from danglin at gcc dot gnu dot org 2007-09-15 22:48 -------
Regarding builtins-59.c, I believe it fails because expand_builtin_sincos()
returns NULL_RTX if the sincos insn is not available. Possibly, this is a
bit pessimistic but it looks like I should xfail the two failing checks.
Regarding builtins-62.c, I believe that it fails because execute_cse_sincos_1()
doesn't do the cexpi optimization:
/* Simply insert cexpi at the beginning of top_bb but not earlier than
the name def statement. */
fndecl = mathfn_built_in (type, BUILT_IN_CEXPI);
if (!fndecl)
return;
Same for builtins-61.c. fold_builtin_cexp() doesn't do its stuff because:
ifn = mathfn_built_in (rtype, BUILT_IN_CEXPI);
if (!ifn)
return NULL_TREE;
It looks like the failing checks in these two tests should also be xfailed.
Thoughts?
--
danglin at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |rguenther at suse dot de
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31497
