This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Crashes on PPC VMX tests
- From: Mark Mitchell <mark at codesourcery dot com>
- To: Nathan Sidwell <nathan at codesourcery dot com>
- Cc: gcc at gcc dot gnu dot org, Andrew Pinski <pinskia at physics dot uc dot edu>
- Date: Fri, 20 Aug 2004 09:49:20 -0700
- Subject: Re: Crashes on PPC VMX tests
- Organization: CodeSourcery, LLC
- References: <4125A31B.30003@codesourcery.com> <412608E2.501@codesourcery.com>
Nathan Sidwell wrote:
Mark Mitchell wrote:
Nathan --
I'm seeing build_int_cst assertion failures on gcc.dg/vmx/debug-3.c
and a bunch of other tests. These are on powerpc64-none-linux-gnu;
probably reproducable with a cross-compiler.
this fixes this rs6000 specific breakage. we were using copy_node to
create
a new type, because build_type_copy is not applicable here. I'm guessing
copy_node worked ok because TYPE_NEXT_VARIANT, TYPE_POINTER_TO and
TYPE_REFERENCE_TO happened to be NULL at this point.
Makes sense.
I decided to create build_distinct_type_copy and rename
build_type_copy to
build_variant_type_copy, to make the distinction between these to
functions,
er, distinct.
Andrew, it might be that the patch for the objc failure you told me about
will want to use one of these new functions.
booted and tested on i686-pc-linux-gnu, and built and tested on
powerpc64-unknown-linux-gnu, ok?
Yes.
--
Mark Mitchell
CodeSourcery, LLC
(916) 791-8304
mark@codesourcery.com