> Hi Thomas,
>
> Afraid I can't help much, except to confirm this. (And, to note that gmake
> -j works;
> so this problem is most likely something with make.)
>
> I have no idea why it doesn't work. I may write a letter to freebsd-ports
> sometime
> later this week, if I don't see one from you sooner :). (School limits time
> right
> now...)
I managed to spot that one now.
The FreeBSD version of "make" normally uses the -B flag when -j is not
used:
-B Try to be backwards compatible by executing a single shell per
command and by executing the commands to make the sources of a
dependency line in sequence. This is turned on by default unless
-j is used.
What this means is that each line of a make target is executed in its
own shell serially. When -j is used, the lines of the make target are
all executed in the same shell (where the commands are of course also
executed serially). But then, the "exit 0" in the second line of the
make target is fatal: it makes the shell quit. That obviously is not a
problem in the per-line shell execution (-B), but it is with -j,
because the following lines of the target aren't executed.
If the line in the Makefile is changed from
if test -f $@; then $(MV) $@ [EMAIL PROTECTED]; else exit 0; fi
to
if test -f $@; then $(MV) $@ [EMAIL PROTECTED]; fi
the -j option works as expected. I don't know why the "exit 0" is
present at all (is this autogenerated in the Configure script or
something?). This occurs at another position in the Makefile, too, but
has no negative impact here. Can both "exit 0" statements safely be
removed?
Greetings,
Thomas.
-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
_______________________________________________
Gtk-gnutella-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/gtk-gnutella-devel
