On Mon, 29 Sep 2003 11:37:31 -0400 Ben Collins <[EMAIL PROTECTED]> wrote:
> Dave, I haven't looked into why gcc-3.3.1 is causing this, but maybe > I'll have time this week. Check for variables explicitly initialized to zero, gcc-3.x will put them into the BSS section whether you want this to happen or not. I wouldn't be surprised if explicitly initialized to zero variables are used by SILO so that it need not zero out the BSS section. There's some magic -f* option to gcc-3.x that shuts off this transformation.
