This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: GDB/libiberty support for IBM long double
- From: Daniel Jacobowitz <drow at false dot org>
- To: Luis Machado <luisgpm at linux dot vnet dot ibm dot com>
- Cc: Andreas Schwab <schwab at suse dot de>, gdb-patches at sourceware dot org
- Date: Sat, 29 Dec 2007 23:38:45 -0500
- Subject: Re: GDB/libiberty support for IBM long double
- References: <200711090107.lA917ZGs027733@d12av02.megacenter.de.ibm.com> <Pine.LNX.4.64.0711130038330.29493@digraph.polyomino.org.uk> <1198783208.7822.51.camel@gargoyle> <jeprwrf6rp.fsf@sykes.suse.de> <1198852288.7822.56.camel@gargoyle>
On Fri, Dec 28, 2007 at 12:31:27PM -0200, Luis Machado wrote:
> On Fri, 2007-12-28 at 11:33 +0100, Andreas Schwab wrote:
> > That part of the function can be simplified quite a bit. For example,
> > NaNs and infinities are already handled at this point, thus
> > special_exponent can only be true when the biased exponent is zero.
>
> Thanks for the feedback Andreas. How should we proceed to get code
> included on libiberty?
Post a patch to gcc-patches. You may want to compare the libiberty
version with the gdb/doublest.c version, which is used otherwise;
I believe I introduced a use of the libiberty version for NaNs and
Joseph added some others for IBM split long double, but no other
platform uses them. So problems are not unlikely.
I wonder if we really need to use floatformat_to_double instead of
floatformat_to_doublest for the split_half cases. It seems like that
relies unnecessarily on the host double format. Do things work better
if you use floatformat_to_doublest and DOUBLEST variables, instead
of floatformat_to_double (likewise floatformat_from_doublest)?
--
Daniel Jacobowitz
CodeSourcery