This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] PR middle-end/14289: TREE_STATIC global reg vars
- From: Mark Mitchell <mark at codesourcery dot com>
- To: Roger Sayle <roger at eyesopen dot com>
- Cc: Richard Henderson <rth at redhat dot com>, gcc-patches at gcc dot gnu dot org, Zack Weinberg <zack at codesourcery dot com>
- Date: Mon, 08 Mar 2004 12:44:47 -0800
- Subject: Re: [PATCH] PR middle-end/14289: TREE_STATIC global reg vars
- Organization: CodeSourcery, LLC
- References: <Pine.LNX.firstname.lastname@example.org>
Is this version OK? This mostly makes the middle-end more robust,The mental model we should have is that the front end passes the middle
end only either (a) valid code, or (b) code with error_mark_node in the
so perhaps additional front-end patches should also be investigated.
Obviously, criterion (b) is not optimially precise...
Your patch is OK for 3.4. However, I think that for the mainline, we
should try to arrange things so that the bogus register declaration
results in the thing not being put in a register at all. We should
issue an error at the point of declaration if at all possible, and then
unset DECL_REGISTER on the variable, or make its type error_mark_node,
or some such.