------- Additional Comments From law at redhat dot com 2004-12-08 17:02 -------
Subject: Re: DSE is not doing its job for
global variables
Note that we're only fixing regressions, and I can't see a way to
classify this as a regression since the RTL DSE code ought to
catch this. This should be marked as postponed until 4.1.
This (untested) code ought to do the trick though:
------- Additional Comments From law at redhat dot com 2004-12-08 17:02 -------
Created an attachment (id=7706)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=7706&action=view)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18880
