3.3 problem of -fzero-initialized-in-bss w/-ffreestanding

[Richard Henderson]

On Mon, Jul 14, 2003 at 01:51:44PM -0700, Geoff Keating wrote:
> > -no-zero-initialized-in-bss should be the default when -ffreestanding is
> > specified -- GCC should not^H^H^Hcannot make assumptions about when BSS
> > zeroing occurs in an -ffreestanding environment.
> > 
> > Can this change be made for 3.3.1 release?
> 
> Um, GCC already assumes that it happens before any code is run.  When
> does your system do it?

Frankly, I agree.  Zeroing of bss really ought to happen before you get
to C at all.  Anything else will confuse even kernel programmers.

And if you can supply -ffrestanding, I think you can just as well
supply -fno-zero-initialized-in-bss at the same time.

In short, I don't think this is an appropriate change for any release,
much less 3.3.1.


r~