We do have a CL for support with 4.7: https://reviews.apache.org/r/6988
But we don't currently compile our changes with 4.7 during development. I think we're all ready and willing to make the switch soon. We don't have 4.7 on our CentOS installs currently so we'll have to deal with that first I imagine. On Tue, Oct 9, 2012 at 1:47 PM, Florian Leibert <[email protected]> wrote: > That's a bummer - none of the current Ubuntu versions ship with 4.1. Do > you have a list of versions you guys support? > > > On Tue, Oct 9, 2012 at 12:42 PM, Benjamin Mahler <[email protected]>wrote: > >> Until we make the switch to gcc 4.7 you'll likely continue to run into >> some issues building off trunk. We use gcc 4.1.2 on centos currently. >> >> >> On Tue, Oct 9, 2012 at 12:24 PM, Florian Leibert <[email protected]> wrote: >> >>> root@i-f090d68d:/srv/mesos/src/build# g++ --version >>> g++ (Ubuntu/Linaro 4.6.1-9ubuntu3) 4.6.1 >>> Copyright (C) 2011 Free Software Foundation, Inc. >>> This is free software; see the source for copying conditions. There is >>> NO >>> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR >>> PURPOSE. >>> >>> >>> On Tue, Oct 9, 2012 at 12:14 PM, Benjamin Mahler <[email protected]>wrote: >>> >>>> Vinod is working on it now. >>>> >>>> This is definitely in the standard: >>>> "If a program calls for the default initialization of an object of a >>>> const-qualified type T, T shall be a class type with a user-provided >>>> default constructor." >>>> >>>> I'm curious what version of gcc are you using? >>>> >>>> On Tue, Oct 9, 2012 at 11:49 AM, Florian Leibert <[email protected]>wrote: >>>> >>>>> >>>>> >>>>> > On Oct. 9, 2012, 5:04 p.m., Florian Leibert wrote: >>>>> > > I just tried to apply the patch from this review and compile >>>>> failed - haven't had a chance to dig yet but it seems related: >>>>> > > >>>>> > > \" -DBUILD_DIR=\"/srv/mesos/src/build\" >>>>> -I../third_party/gmock-1.6.0/gtest/include >>>>> -I../third_party/gmock-1.6.0/include -I/usr/lib/jvm/java-7-oracle/include >>>>> -I/usr/lib/jvm/java-7-oracle/include/linux -DZOOKEEPER_VERSION=\"3.3.4\" >>>>> -pthread -g2 -O2 -MT tests/mesos_tests-stout_tests.o -MD -MP -MF >>>>> tests/.deps/mesos_tests-stout_tests.Tpo -c -o >>>>> tests/mesos_tests-stout_tests.o `test -f 'tests/stout_tests.cpp' || echo >>>>> '/srv/mesos/src/src/'`tests/stout_tests.cpp >>>>> > > /srv/mesos/src/src/tests/stout_tests.cpp: In member function >>>>> 'virtual void StoutUtilsTest_rmdir_Test::TestBody()': >>>>> > > /srv/mesos/src/src/tests/stout_tests.cpp:316:30: error: >>>>> uninitialized const 'EMPTY' [-fpermissive] >>>>> > > >>>>> /srv/mesos/src/third_party/libprocess/include/stout/hashset.hpp:17:7: >>>>> note: >>>>> 'const class hashset<std::basic_string<char> >' has no user-provided >>>>> default constructor >>>>> > > make[3]: *** [tests/mesos_tests-stout_tests.o] Error 1 >>>>> > > make[3]: Leaving directory `/srv/mesos/src/build/src' >>>>> > > make[2]: *** [check-am] Error 2 >>>>> > > make[2]: Leaving directory `/srv/mesos/src/build/src' >>>>> > > make[1]: *** [check] Error 2 >>>>> > > make[1]: Leaving directory `/srv/mesos/src/build/src' >>>>> > > make: *** [check-recursive] Error 1 >>>>> > > run-parts: /var/lib/cloud/instance/scripts/part-018 exited with >>>>> return code 2 >>>>> > > >>>>> > >>>>> > Florian Leibert wrote: >>>>> > Sorry for the extra output - this comes from my cloud-init >>>>> scripts and is mixed into the compile output... >>>>> > >>>>> > Vinod Kone wrote: >>>>> > this doesnt look related! surprised that you got this error only >>>>> after applying the above patch. let me try to repro... >>>>> > >>>>> > Vinod Kone wrote: >>>>> > couldn't repro this on osx (10.6) or centos (5.5). can you let >>>>> me know what's your environment and the commands that you used to build >>>>> mesos? >>>>> >>>>> I'm on mesos trunk and Ubuntu 12.04 but you're right - I reverted the >>>>> patch and still get the error. I must have been on a different git >>>>> revision >>>>> last week... >>>>> But either way - it would be great if you guys could look into this >>>>> bug I filed as it's a blocker for us. >>>>> >>>>> >>>>> - Florian >>>>> >>>>> >>>>> ----------------------------------------------------------- >>>>> >>>>> This is an automatically generated e-mail. To reply, visit: >>>>> https://reviews.apache.org/r/7454/#review12277 >>>>> ----------------------------------------------------------- >>>>> >>>>> >>>>> On Oct. 8, 2012, 7:54 p.m., Vinod Kone wrote: >>>>> > >>>>> > ----------------------------------------------------------- >>>>> >>>>> > This is an automatically generated e-mail. To reply, visit: >>>>> > https://reviews.apache.org/r/7454/ >>>>> > ----------------------------------------------------------- >>>>> > >>>>> > (Updated Oct. 8, 2012, 7:54 p.m.) >>>>> >>>>> > >>>>> > >>>>> > Review request for mesos, Benjamin Hindman and Ben Mahler. >>>>> > >>>>> > >>>>> > Description >>>>> > ------- >>>>> >>>>> > >>>>> > Better error messages >>>>> > >>>>> > >>>>> > Fix for curl download >>>>> > >>>>> > >>>>> > Diffs >>>>> > ----- >>>>> > >>>>> > src/launcher/launcher.cpp 5de1f479cf03182a1c12bdd951b5bd57c00ee2ee >>>>> > third_party/libprocess/include/stout/net.hpp >>>>> f6b770c8ca7d21e0aca0f614168a0985c77046b0 >>>>> > third_party/libprocess/include/stout/os.hpp >>>>> 13dbc715ed08cfe6b24ee20744d427dac1104694 >>>>> > >>>>> > Diff: https://reviews.apache.org/r/7454/diff/ >>>>> > >>>>> > >>>>> > Testing >>>>> > ------- >>>>> > >>>>> > Tested locally >>>>> > >>>>> > >>>>> > Thanks, >>>>> > >>>>> > Vinod Kone >>>>> > >>>>> > >>>>> >>>>> >>>> >>> >>> >>> -- >>> Best regards, >>> >>> Florian >>> http://twitter.com/flo <http://twitter.com/floleibert> >>> http://flori.posterous.com/ >>> >>> >> > > > -- > Best regards, > > Florian > http://twitter.com/flo <http://twitter.com/floleibert> > http://flori.posterous.com/ > >
