This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug target/69421] [6 Regression] ICE in maybe_legitimize_operand, at optabs.c:6888 with -O3
- From: "jakub at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Fri, 22 Jan 2016 11:30:58 +0000
- Subject: [Bug target/69421] [6 Regression] ICE in maybe_legitimize_operand, at optabs.c:6888 with -O3
- Auto-submitted: auto-generated
- References: <bug-69421-4 at http dot gcc dot gnu dot org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69421
Jakub Jelinek <jakub at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |ienkovich at gcc dot gnu.org
--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I see there confusion on the sign of element's type of vector boolean types -
build_truth_vector_type / build_nonstandard_boolean_type creates a signed
element type, while get_vectype_for_scalar_type_and_size creates unsigned
element type. Plus for some strange reason the last argument is getting a V2DI
mode vector type instead of expected vector boolean V16QI.
Ilya, can you please have a look?