Op een zonnige herfstdag (Sunday 03 October 2004 19:26),schreef Craig A.
Berry:
> At 3:53 PM +0200 10/3/04, Abe Timmerman wrote:
> >Op een grimmige herfstdag (Sunday 03 October 2004 03:29),schreef Craig A.
> >
> >Berry:
> > > >No change in the buildprocess with the newer mmk however :-(
[snip]
> You can use MMK/DUMP and/or MMK/LOG to try to diagnose exactly what's
> happening, but it must have something to do with an implicit
> dependency getting missed when going from Encode.xs to Encode.c to
> Encode.obj. I would try making the dependencies more explicit,
> perhaps with a hack like the following to Encode's Makefile.PL.
>
> --- Makefile.PL;-0 Thu Apr 24 15:24:20 2003
> +++ Makefile.PL Sun Oct 3 12:02:36 2004
> @@ -83,8 +83,9 @@
> {
> my $self = shift;
> my $dir = $self->catdir($self->curdir,'ucm');
> - my $str = "# Encode\$(OBJ_EXT) depends on .h and .exh files not .c
> files - but all written by enc2xs\n";
> + my $str = "# In addition to Encode.c, Encode\$(OBJ_EXT) depends on .h
> and .exh files written by enc2xs\n";
> $str .= $^O eq 'MacOS' ?
> 'Encode.c.{$(MACPERL_BUILD_EXT_STATIC)}.o :' : 'Encode$(OBJ_EXT) :';
> + $str .= ' Encode.c';
> foreach my $table (keys %tables)
> {
> $str .= " $table.c";
> [end]
I'll try that and see what happens... (things do not go fast on this machine,
so it may take some time)
> > > DISK$USER_2:[TIMMERMAN.KLAD.PERL-CURRENT.LIB.AUTO.FILTER.UTIL.CALL]
> > >
> > > That's 8 levels, which is the maximum allowed on OpenVMS VAX 7.2.
> >
> >I was affraid of that. CONFIGURE.COM is currently checking the depth of
> > the source-tree directory and bails out if it is too deep. Should this
> > max_allowed_dir_depth be set to 0 and hint for this rooted logical way.
>
> I thought the level we had it at was ok, but perhaps something has
> gotten deeper. Have you tried building one level higher? A hint
> about the rooted logical solution is a good idea, though.
The current max_allowed_dir_depth = 3 (which is where my source tree is). That
will always trigger an error for Filter::Util::Call as it needs another 2
levels to build (on systems where the limit applies). So currently the depth
of one (1) is maximum. (On this system that means peer to my "home directory"
I guess most people in such systems *must* use the rooted logical.
> >Perhaps even set it automatically from DESCRIP.MMS?
>
> Once MMK has started processing the Descrip.MMS, I don't think you
> can change default directory in the main process (only in
> subprocesses).
Perhaps configure.com can leave a DCL-thingy to set it up, so you can run that
before you run MM[KS]
> > > The typical way around this is to create a rooted logical name
> >
> >How can I set this rooted logical from perl?
>
> There's an example in MakeMaker. See
>
> http://www.makemaker.org/src/ExtUtils-MakeMaker-snap/t/lib/MakeMaker/Test/S
>etup/BFD.pm
>
> and look for the _setup_bfd_test_root subroutine.
okay, thanks!
Thanks for your help +
Good luck,
Abe
--
Benjamin Goldberg> chr(0x20_0000+number) has that property,
Benjamin Goldberg> unless space aliens land on earth.
These are the kinds of statements that make the snowballs in hell look
expectantly at the thermometer. Or, in this case, the aliens look at
their calendar whether they can fit in an invasion.
-- Jarkko Hietaniemi on p5p @ 2002-02-03
