Re: VIRT_ADDR_VARIES

[Paul Eggert]

On 11/10/11 03:06, Eli Zaretskii wrote:

> It could be that a test against my_edata was just a means,
> but we should be sure it was

Yes, I'm sure it was a means.  I've read the code carefully, and
remember this part well.  That's why the VIRT_ADDR_VARIES implementation
is also a valid one.

> isn't it dangerous to have an externally visible symbol with a
> name such as `pure'?

Emacs has been built with an extern symbol 'pure'
for decades, on many platforms.  And platform implementers
would be unlikely to usurp that particular name anyway.
So we should be safe.