This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Floating point -- fix H8/300 and other targets without XF/TFmode
- From: Zack Weinberg <zack at codesourcery dot com>
- To: Kazu Hirata <kazu at cs dot umass dot edu>
- Cc: gcc-patches at gcc dot gnu dot org, tm at kloo dot net
- Date: Sun, 20 Oct 2002 12:50:46 -0700
- Subject: Re: Floating point -- fix H8/300 and other targets without XF/TFmode
- References: <20020418180815.GI30952@codesourcery.com> <20021019.220421.58790856.kazu@cs.umass.edu>
On Sat, Oct 19, 2002 at 10:04:21PM -0400, Kazu Hirata wrote:
> Hi Zack,
>
> > Patch has bootstrapped on i686-linux (LONG_DOUBLE_TYPE_SIZE 96) with
> > no regressions. I attempted to test against the h8300 simulator but
> > got tons of spurious failures due to some problem with testglue.o
> > which I was unable to debug. I would appreciate testing on a real H8.
>
> I got a little bit of clue as to where spurious failures come from.
> After I clean-installed RedHat Linux 8.0, including dejagnu, the gcc
> test with the h8300 simulator caused spurious failures. After I
> removed the system dejagnu and started using dejagnu that I installed
> with h8300-hms-gdb, those failures disappeared. So I am guessing that
> there are some differences between the system dejagnu and local
> dejagnu. Anyway, I'll look into this.
Yes, I'm using plain old dejagnu 1.4.2. If there's some patch that
needs to make it into the official dejagnu, we need to find out what
it is.
zw