This is the mail archive of the
mailing list for the GCC project.
Re: [fortran, patch] Patch for PR18827
- From: Feng Wang <wf_cs at yahoo dot com>
- To: Feng Wang <wf_cs at yahoo dot com>, fortran at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
- Cc: Steven Bosscher <stevenb at suse dot de>
- Date: Mon, 28 Feb 2005 10:19:03 +0800 (CST)
- Subject: Re: [fortran, patch] Patch for PR18827
It seems that all the maintainers and reviewers do not care about it. If you
don't care, I don't care. I just feel sick about the increasing bugs in the
list. Isn't it a policy that all the patchs should be approval?
I have no write access. In fact, I am very glad to commit the patchs myself.
Can I apply this "privilege", maintainers?
--- Feng Wang <firstname.lastname@example.org> 的正文：
> Feng Wang
> --- Feng Wang <email@example.com> 的正文：
> > This patch fix pr18827. Allows variables in a common block to be assigned a
> > label. And do not set symbol's assign attribute when match assigned goto
> > statement and i/o statements using label assigned variable. We should set
> > only when match assign statment. And add checking this attribute when
> > resolving.
> > But this patch reveals another bug of the backend. The backend will remove
> > the
> > label unless we set !DECL_ARTIFICIAL flag on assigned label, even we set
> > TREE_ADDRESSABLE flag and !DECL_IGNORED_P and FORCED_LABEL flag. After
> > discussion with Steven, we reserve the setting of !DECL_ARTIFICIAL on
> > assigned
> > label to fix pr18827. And after committing this patch, I will file a BE bug
> > reporting.
> > Tested on i686 with no regression. If it is ok, please commit it.
Do You Yahoo!?