[LSST|dm-users 665] Re: eups distrib: Failed to build fftw-3.3.3.eupspkg with error: member reference base type 'ticks' (aka 'int') is not a structure or union

Sat May 16 22:01:04 PDT 2015

Phil,

A really brute force (but effective and simple) way to figure out what's going on may be to grep all .h files on your system for the presence of timebasestruct_t. That may tell you what is the offending header file. Btw, this issue is not LSST specific -- you'd get the same error if you tried to build fftw 3.3.3 from source yourself.

But I agree with Robert -- if the stack is broken due to the presence of homebrew, it's probably a bug on out end (they're usually careful not to mess with the system).

Either way, let's try to figure it out.
-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

On May 16, 2015 3:39:44 AM PDT, Phil Marshall <dr.phil.marshall at gmail.com> wrote:
>Thanks all.
>
>I unsetenv-ed all compilers, and also unsetenv-ed my LD_LIBRARY_PATH
>and
>tried again. The build.log is much the same (and is attached).
>Along with Mario's flagging of read_real_time being found, I see this
>error:
>
>In file included from timer.c:29:
>./cycle.h:112:9: error: unknown type name 'timebasestruct_t'
>typedef timebasestruct_t ticks;
>
>So I guess I have installed a library somewhere for some other reason
>(with
>homebrew, probably - I don't use macports) and now the LSST installer
>is
>trying to use that one instead of its own. But, I don't know what that
>library is, and presumably the other software I installed that needs it
>will stop working if I were to remove it. So, isn't the right thing to
>do
>to somehow force the LSST installer to ignore read_real_time if it
>exists
>on my system, and carry on as if the test had returned
>
>   checking for read_real_time... no
>
>like Mario's did? I don't know how to make that change in the
>installation
>script.
>
>
>Alternatively, going back to Russell's note: I use homebrew, which is
>very
>useful. How do I "disable" this while I install the DM stack? Any
>disabling
>would need to be temporary, of course.
>
>Thanks for your help!
>
>Phil
>
>On Fri, May 15, 2015 at 9:47 PM, Mario Juric
><mjuric at astro.washington.edu>
>wrote:
>
>> On 5/15/15 12:33 , Phil Marshall wrote:
>> >
>> > This *looks* like a genuine bug,  or maybe an undocumented need for
>some
>> > compiler flags. I couldn't find any mention of this error in this
>email
>> > list archive.
>> >
>> > I am following Simon's instructions
>> > at https://confluence.lsstcorp.org/display/SIM/Catalogs+and+MAF
>(like so
>> > many of us MAF hopefuls) and running Mac OS X Yosemite.
>> >
>> > The build.log is attached. All suggestions welcome!
>> >
>>
>> Hi Phil,
>>         I also suspect something is interfering with the normal
>build. Do
>> you
>> have other compilers on your path? Special libraries? Do you have CC
>> and/or CXX defined?
>>
>> Here's some further evidence pointing to that:
>>
>> My Yosemite fftw build.log file:
>> ========
>> ....
>> checking for read_real_time... no
>> checking for time_base_to_time... no
>> ....
>> ========
>>
>> and the build.log that you attached:
>> ========
>> ....
>> checking for read_real_time... yes
>> checking for time_base_to_time... yes
>> ....
>> ========
>>
>> Based on some cursory reading on teh internets, read_real_time seems
>to
>> be an AIX thing (!?).
>>
>> -- Mario
>>
>> > Cheers
>> >
>> > Phil
>> >
>> >
>> > pjm% eups distrib install sims_maf -t sims
>> >   [  1/59 ]  cfitsio 3360.lsst1
>> > done.
>> >   [  2/59 ]  doxygen 1.8.5 (already installed)
>> >  done.
>> >   [  3/59 ]  eigen 3.2.0
>> >  done.
>> >   [  4/59 ]  fftw 3.3.3 ...
>> >
>> > ***** error: from
>> > /Users/pjm/lsst/EupsBuildDir/DarwinX86/fftw-3.3.3/build.log:
>> >      return (((double)t1.tb_high - (double)t0.tb_high) * 1.0e9 +
>> >                       ~~^~~~~~~~
>> > ./cycle.h:125:46: error: member reference base type 'ticks' (aka
>'int')
>> > is not a structure or union
>> >      return (((double)t1.tb_high - (double)t0.tb_high) * 1.0e9 +
>> >                                            ~~^~~~~~~~
>> > ./cycle.h:126:18: error: member reference base type 'ticks' (aka
>'int')
>> > is not a structure or union
>> >              ((double)t1.tb_low - (double)t0.tb_low));
>> >                       ~~^~~~~~~
>> > ./cycle.h:126:38: error: member reference base type 'ticks' (aka
>'int')
>> > is not a structure or union
>> >              ((double)t1.tb_low - (double)t0.tb_low));
>> >                                           ~~^~~~~~~
>> > 2 warnings and 8 errors generated.
>> > make[2]: *** [timer.lo] Error 1
>> > make[2]: *** Waiting for unfinished jobs....
>> > mv -f .deps/tensor8.Tpo .deps/tensor8.Plo
>> > mv -f .deps/tensor9.Tpo .deps/tensor9.Plo
>> > mv -f .deps/tile2d.Tpo .deps/tile2d.Plo
>> > make[1]: *** [all-recursive] Error 1
>> > make: *** [all] Error 2
>> > + exit -4
>> > eups distrib: Failed to build fftw-3.3.3.eupspkg: Command:
>> > source /Users/pjm/lsst/eups/bin/setups.sh; export
>> > EUPS_PATH=/Users/pjm/lsst;
>> > (/Users/pjm/lsst/EupsBuildDir/DarwinX86/fftw-3.3.3/build.sh) >>
>> > /Users/pjm/lsst/EupsBuildDir/DarwinX86/fftw-3.3.3/build.log 2>&1
>> > 4>/Users/pjm/lsst/EupsBuildDir/DarwinX86/fftw-3.3.3/build.msg
>> > exited with code 252
>> >
>> >
>> > _______________________________________________
>> > dm-users mailing list
>> > dm-users at lists.lsst.org
>> > https://lists.lsst.org/mailman/listinfo/dm-users
>> >
>>
>>
>> --
>> Mario Juric,
>> UW Astronomy Faculty | UW eScience | LSST DM Project Scientist
>> Web : http://research.majuric.org      Phone : +1 609 933 1033
>> _______________________________________________
>> dm-users mailing list
>> dm-users at lists.lsst.org
>> https://lists.lsst.org/mailman/listinfo/dm-users
>>
>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>dm-users mailing list
>dm-users at lists.lsst.org
>https://lists.lsst.org/mailman/listinfo/dm-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.lsst.org/pipermail/dm-users/attachments/20150516/19b9ea03/attachment.html>