PR bin/185052 filed (was: RE: nanobsd build failure 'WITHOUT_CASPER=YES' r259661 and earlier)

[Stefan Hegnauer]

Hi there, 

thanks for your support. PR 185052 is filed and should be visible soon -
please let me know if I can help in any way (more info, testing, whatever).

Thanks again
Stefan

> From: adrian.chadd at gmail.com [mailto:adrian.chadd at gmail.com] On Behalf
> Of Adrian Chadd
> Sent: Friday, December 20, 2013 9:52 PM
> To: Stefan Hegnauer
> Cc: freebsd-current
> Subject: Re: nanobsd build failure 'WITHOUT_CASPER=YES' r259661 and
> earlier
> 
> Hi,
> 
> Please file a PR and then ask the developer (pjd@) very nicely to take
> a look at it.
> 
> Thanks,
> 
> 
> -adrian
> 
> 
> On 20 December 2013 12:45, Stefan Hegnauer <stefan.hegnauer at gmx.ch>
> wrote:
> > When using 'WITHOUT_CAPSICUM=YES', 'WITHOUT_CASPER=YES' my nanobsd
> builds in
> > a Virtualbox VM (i386, march=geode, GENERIC without debug+Witness et.
> al.)
> > fail buildworld for any revision from at least r259518-r259661; like
> so
> > (this example is r259661):
> >
> >
> >
> > " ...
> >
> > ===> lib/clang/libllvmsupport (obj,depend,all,install)
> >
> >
> /usr/obj/nanobsd.sstream//usr/src/tmp/usr/src/lib/clang/libllvmsupport
> > created for /usr/src/lib/clang/libllvmsupport
> >


snip - see PR for details ...


> >
> > make: stopped in /usr/src"
> >
> >
> >
> > Note that this is with PMAKE="-j1", i.e. single threaded build (same
> happens
> > with standard PMAKE=-j3 but slightly less intuitive to see where it
> fails)
> >
> >
> >
> > Removing WITHOUT_CASPER=YES in the build instructions cures the
> problem (!),
> > however I fail to see why I should include it for an embedded device
> > (pcengines.ch Alix boards, several different versions).
> >
> > Also, with the error reported above I have the impression it is not
> exactly
> > intuitive that you have to include CASPER (and not CAPSICUM) to
> eliminate
> > the failure?
> >
> > Any pointers/hints/solutions?
> >
> >
> >
> > Sorry for the rant & thanks
> >
> > Stefan
> >
"