This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: PATCH: PR target/35657: TDmode isn't aligned at 128bit when passing to a function
- From: "Uros Bizjak" <ubizjak at gmail dot com>
- To: "H.J. Lu" <hjl dot tools at gmail dot com>
- Cc: "Mark Mitchell" <mark at codesourcery dot com>, "Richard Guenther" <richard dot guenther at gmail dot com>, "Jan Hubicka" <jh at suse dot cz>, "GCC Patches" <gcc-patches at gcc dot gnu dot org>, "Girkar, Milind" <milind dot girkar at intel dot com>, "Jan Hubicka" <hubicka at ucw dot cz>
- Date: Tue, 29 Apr 2008 08:24:51 +0200
- Subject: Re: PATCH: PR target/35657: TDmode isn't aligned at 128bit when passing to a function
- References: <6dc9ffc80803270633v25be5051i90bf86a6a2182822@mail.gmail.com> <5787cf470803270702o76651c11gb9ef9eea338fe3e8@mail.gmail.com> <20080327143353.GR4732@kam.mff.cuni.cz> <84fc9c000803270940p54c47c3bg312311d5e5cb4f98@mail.gmail.com> <47EBDF4F.4050000@codesourcery.com> <6dc9ffc80803271103g72363d05se74349d083e3631e@mail.gmail.com>
On Thu, Mar 27, 2008 at 8:03 PM, H.J. Lu <hjl.tools@gmail.com> wrote:
> > > On Thu, Mar 27, 2008 at 3:33 PM, Jan Hubicka <jh@suse.cz> wrote:
> > >>> Regarding 4.3.1 support - since this is an ABI change w.r.t released
> > >> > compiler, RMs should have their last word. IMO, at least a notice in
> > >> > release notes is needed
> > >>
> > >> RM has last word here, for sure, but since the use of DFP currently is
> > >> minimal and is supposed to rise for next major release, I would preffer
> > >> fixing the issue in 4.3.1 rather than waiting for anohter half a year.
> > >> With news.html update for sure ;)
> > >
> > > I tend to agree. How often do you plan to change the DFP ABI? ;)
> >
> > I agree too, but unhappily. ABI changes really hurt GCC's acceptance;
> > they're one of the things that users most often complain about. :-(
> >
>
> The main problem is that the exiting ia32 psABI doesn't cover any new
> hardware/software features and there are no updates to support them.
> I am working on a solution.
Has the conclusion w.r.t. to TDmode alignment been reached? Looking
into the thread at [1], I don't think that we have authoritative
answer on this issue. I think that to avoid possible ABI
incompatibilities with future gcc versions, this should be finalized
before 4.3.1 is released.
IOW, This issue can be considered as blocker for 4.3.1 release as far
as x86 is concerned.
[1] http://groups.google.com/group/ia32-abi/browse_thread/thread/fc6db344b62dea76
Uros.