This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: RFC: TLS improvements for IA32 and AMD64/EM64T
On Sep 17, 2005, Richard Henderson <rth@redhat.com> wrote:
> On Sat, Sep 17, 2005 at 02:25:49PM -0300, Alexandre Oliva wrote:
>> Err... Mainly because that would make it easier for people (including
>> myself) to experiment with it. Get it tested by early birds by
>> including it disabled by default, and then eventually switch to it as
>> the default, perhaps as early as 4.2.
> Eh... I think this should wait until after the branch.
My turn to ask why :-)
If you're concerned about potential changes that would render the
implementation unusable, bear in mind that the details implemented by
GCC are the ones that are pretty much set in stone right now. What
remains to be figured out is some selection of `nop' instructions in
relaxations, and that doesn't affect GCC in any way. So it's really
safe to go with the patch.
And then, since it will not affect the generated code unless you use
the command-line option that enables this new feature, I feel it would
be perfectly safe and useful to have this in.
Of course, this is all said before listening to your reasons, so...
Your turn ;-)
--
Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/
Red Hat Compiler Engineer aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist oliva@{lsd.ic.unicamp.br, gnu.org}