This is the mail archive of the
mailing list for the GCC project.
Re: [dataflow]: PATCH COMMITTED
- From: Richard Sandiford <richard at codesourcery dot com>
- To: Kenneth Zadeck <zadeck at naturalbridge dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>, "Berlin, Daniel" <dberlin at dberlin dot org>, Seongbae Park <seongbae dot park at gmail dot com>, Ian Lance Taylor <iant at google dot com>
- Date: Wed, 15 Nov 2006 16:26:28 +0000
- Subject: Re: [dataflow]: PATCH COMMITTED
- References: <455922F5.email@example.com> <firstname.lastname@example.org> <455B382B.email@example.com>
Kenneth Zadeck <firstname.lastname@example.org> writes:
> The actual change is what Iant told me to do after a long conversation
> with many people on irc. Unfortunately this was several months ago when
> I was trying to get the ia-64 working and I have forgotten many of the
> The fix is to deal with some hackery in the ia-64. There is a failure
> that occurs when dce/dse is run after reload (which it never is on the
Are you sure? I thought we ran PROP_KILL_DEAD_CODE passes after reload,
which is why we have things like REG_MAYBE_DEAD.
> I believe, but I have forgotten all of the details, that an
> unspec is used rather than a use to keep something alive because a uses
> are not defined after reload and what we really wanted was a use. There
> was a lot of chatting about this on irc as to whether to change this to
> an unspec volatile but in the end this is what ian told me to do.
> The conversation was made on or right before sept 21. Maybe danny can
> find the irc log.
Thanks, that'd be great. It'd be a shame to have mysterious stuff
like this in the shiny new infrastructure ;)