Boost logo

Boost :

Subject: Re: [boost] [predef] The predef-check functionality is not working on the Android regression tests
From: Rene Rivera (grafikrobot_at_[hidden])
Date: 2015-07-12 23:15:20


On Sun, Jul 12, 2015 at 9:00 PM, Edward Diener <eldiener_at_[hidden]>
wrote:

> I do not believe that predef-check is working at all on the Android
> regression tests. Predef itself appears to be working as far as the predef
> macro values. But attempting to use predef-check to set a <cxxflags> value
> is just not happening. I can attest that just setting a <cxxflags> value,
> outside of using predef-check, does work correctly on the Android
> regression tests.
>
> The failure is occurring on my VMD tests. But you can also see the failure
> on the predef tests also. Looking at the info_as_cpp test output at
> http://www.boost.org/development/tests/develop/developer/CrystaX-NET-apilevel-21-armeabi-v7a-hard-predef-clang-linux-3-6-info_as_cpp-variants_.html
> you can see that BOOST_COMP_CLANG > 0 and BOOST_OS_LINUX = 0. Yet if you
> look at the corresponding check_value test at
> http://www.boost.org/development/tests/develop/developer/CrystaX-NET-apilevel-21-armeabi-v7a-hard-predef-clang-linux-3-6-check_value-variants_.html
> you will see that the output is 'CHECK_VALUE == undefined', whereas if
> predef-check were actually working the output should be 'CHECK_VALUE == 1'
> ( or 'true' ).
>
> In all other environments predef-check is working. Maybe this failure of
> predef-check somehow occurs because the Android tests are a cross-compile,
> being done on a Unix system but targeting the Android. But for whatever
> reason the inability of getting predef-check to work keeps my VMD tests
> from adding the '-std=c++0x' option and the VMD tests all fail without that
> option since the variadic macro support in VMD for gcc needs C++11 support
> or better.
>

Yes, I am fully aware of the issue. And I describe the basic problem in
another post. In which I mentioned the basic problem of the predef check
being "incorrect" in the cross-compile use case (which I posted from an
airport). But, I will not have time to attempt a fix for this until next
month. As I'm out of the country at the moment.

-- 
-- Rene Rivera
-- Grafik - Don't Assume Anything
-- Robot Dreams - http://robot-dreams.net
-- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk