[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Yet another bug (core dumped)
From: |
Dr. Jörn von Holten |
Subject: |
Yet another bug (core dumped) |
Date: |
Mon, 09 Feb 2004 12:26:11 +0100 |
I'm using make 3.81rc1 and got a core with the following Makefile (a.cdl
and b.idl exist)
---------------------
all: a.o b.o
a.o: a.cc
touch a.o
a.cc: a.idl
touch a.cc
a.idl: a.cdl
touch a.idl
$(eval a.o: b.cc)
b.o: b.cc
touch b.o
b.cc: b.idl
touch b.cc
------------------------
The position is in remake.c:435 while evaluating check_renamed (d->file)
as d->file is NULL.
You might now realize that I use the eval function at an "interesting"
location, but
as the manual reads: "it can be placed virtually anywhere".
I expect that the rule is added to the makefile at that moment, the eval
function
is evaluated as part of the a.idl rule and that Make takes it into
consideration as well
and at once.
Background: I want new dependencies added dynamically that result from some
actions. I do not want to solve this by the remaking of makefiles, as I
do not have rules
to generate missing makefiles (they shall be there or not), because I
suspect that
by doing this, make will regenerate unwanted things - actually I
generate the depend
files at the same moment I generate the targets.
Any suggestions? Maybe its possible to convince make to restart from
"scratch" with
some other trick
--Jörn
--
Dr. Jörn von Holten
Senior Consultant
Barco Orthogon AG
Hastedter Osterdeich 222
D-28207 Bremen
Tel +49-421-20122-426
Fax +49-421-20122-999
www.barco-orthogon.com
address@hidden
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- Yet another bug (core dumped),
Dr. Jörn von Holten <=