This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch] Merge loop-invariant changes from killloop branch
- From: Kenneth Zadeck <Kenneth dot Zadeck at NaturalBridge dot com>
- To: gcc-patches at gcc dot gnu dot org
- Date: Mon, 28 Nov 2005 14:44:39 -0500 (EST)
- Subject: Re: [patch] Merge loop-invariant changes from killloop branch
- Reply-to: Kenneth dot Zadeck at NaturalBridge dot com
> Steven Bosscher wrote:
>
> > But I do. The project to remove loop.c has been going on for _years_.
> > Now you're saying it will have to wait for another, totally unrelated
> > project that just happens to fix one tiny little buglet in df.c that
> > can easily be worked around?
>
> I wish we didn't have to make this contentious.
>
> No, I'm saying it's not unreasonable to consider waiting for the
> dataflow stuff.
>
Danny and myself are currently in the process of changing a large
number of the interfaces of df.c. Kill loop appears to use many of
these interfaces. We will update all of the uses and references to
the old interfaces of df.
>
> Please note that I also didn't say that killoop must wait for the
> dataflow stuff.
>
> It might be perfectly appropriate to check in killoop, with a
> work-around for this issue; then again, it might not. That's up to
> whoever reviews your patch.
>
> My statement was meant to be more general. If one patch has a
> dependency on another, then even if the first has been in progress for
> some length of time, it may be appropriate to wait until the second is
> ready, especially if, in the end, both will appear in the same release
> anyhow.
>
> --
> Mark Mitchell
> CodeSourcery, LLC
> mark@codesourcery.com
> (916) 791-8304
Kenny and Danny