[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: getting warmer error messages
From: |
Philip Guenther |
Subject: |
Re: getting warmer error messages |
Date: |
Fri, 29 Dec 2006 13:56:19 -0700 |
On 12/28/06, Martin Dorey <address@hidden> wrote:
...
> there are usually many possible pattern rules
> which it MIGHT match, if it could build the
> prerequisites
An example might help persuade - % is often make-able from %.c - so any list of
potentially missing prerequisites would have to include %.c. I'm tempted to
suggest that I might sometimes find the list of missing prerequisites useful
when debugging but perhaps it would be so explosively large that it wouldn't be
useful even as an option. I do OK with make -p anyway.
The trick, of course, is that the list of missing prerequisities can
be quite large. For the first Makefile in the OP's message, the list
of possible prerequisites is:
RCS/x.aa
RCS/x.aa,v
RCS/x.bb
RCS/x.bb,v
SCCS/s.x.aa
SCCS/s.x.bb
s.x.aa
s.x.bb
x.aa,v
x.bb
x.bb,v
...plus 107** possible prerequistes for the Makefile itself! Don't
laugh: the problem may very well be that the Makefile is out of date.
If the target doesn't have a suffix, the list of possible
prerequisites is much larger. Listing them all is not actually likely
to be that useful, IMHO.
Philip Guenther
** Perhaps GNU make should include a pattern rule to prevent the
non-terminal pattern rules from applying to the Makefile itself. For
example, if you add the following line to your Makefile:
%akefile:
then there are only five prerequisites for the Makefile itself, from
the RCS and SCCS rules.