pkg/35015: devel/cmake needs a large datasize to compile

[kre%munnari.OZ.AU@localhost]

>Number:         35015
>Category:       pkg
>Synopsis:       devel/cmake needs a large datasize to compile
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    pkg-manager
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Nov 08 04:00:00 +0000 2006
>Originator:     Robert Elz
>Release:        NetBSD 3.99.15  (pkgsrc current in past 8 hours or so)
>Organization:
        Prince of Songkla University
>Environment:
        
        
System: NetBSD jade.coe.psu.ac.th 3.99.15 NetBSD 3.99.15 
(GENERIC-1.696-20060125) #8: Wed Jan 25 04:59:39 ICT 2006 
kre%jade.coe.psu.ac.th@localhost:/usr/obj/current/kernels/JADE_ASUS i386
Architecture: i386
Machine: i386
>Description:
        I'm not sure this should be a PR - please simply close it
        if there's nothing that can really be done about this...

        devel/cmake seems to need a very large datasize to successfully
        compile.   By default, more than what (I expect) most users get
        as their (NetBSD) data size limit.

>How-To-Repeat:
        Compile devel.cmake, expect to see ...

Building CXX object Source/CMakeFiles/CMakeLib.dir/cmDependsJavaParser.o
 
cc1plus: out of memory allocating 264177376 bytes after a total of 6699620 bytes
gmake[2]: *** [Source/CMakeFiles/CMakeLib.dir/cmDependsJavaParser.o] Error 1
gmake[1]: *** [Source/CMakeFiles/CMakeLib.dir/all] Error 2
gmake: *** [all] Error 2
*** Error code 2
 
Stop.
make: stopped in /usr/pkgsrc/devel/cmake
*** Error code 1

        If I read that message correctly, something had (almost) 7MB
        allocated, and needed 250+MB more ...

        Compiling again with ulimit -d 3145728 fixed the problem.
        (that number because that's what ulimit said was the hard limit).

>Fix:
        I have no idea whether pkgsrc has any mechanism to set resource
        limits when needed to compile particular packages, if it does,
        then perhaps this package could be taught how to use that to
        request a higher data segment limit.

        Otherwiase, perhaps at least a MESSAGE indicating how to
        fix the problem should it be observed.