On Sun, 29 Jul 2001, Jim Pazarena wrote:
> I find that during a "make", every script flying by complains about
> a re-definition of "ERRNO_QUOTA".
>
> The warning indicates:
> in config.h has a redefinition of "ERRNO_QUOTA"
>
> macros.h was the previous definition
>
> ---
>
> when you examine macros.h, you see a:
> #ifndef ERRNO_QUOTA
>
> but if you examine config.h there is no such #ifndef
>
> It appears that the definition is being presented my macros.h
> prior to config.h.
>
> Is this a buglet of some sort?
Yes, but I'm also confused by something else.
I see that the SCO-specific file (OS/Makefile-SCO) contains the line
ERRNO_QUOTA=0
Now, that looks like a ridiculous thing to have there. (This
configuration came from the first person who ported Exim to a SCO
system). Does SCO have a file quota system? If so, it should either have
the EDQUOT or ENOSPC errors, and Exim should work without setting
anything in OS/Makefile-SCO.
The effect of the current setting in config.h is to disable any
detection of quota errors. Is this what is really wanted?
[The bug is that the code that is in macros.h should really be in
exim.h, _after_ it has included config.h. However, this is such a minor
thing that I'm not going to bother to do anything about it for Exim 3.
At least, not unless lots of people mutter at me.]
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
