Boost 1.33.1 Release Candidate #1 is available at: http://www.osl.iu.edu/~dgregor/boost-1.33.1-rc1/ If all goes well, this will become the Boost 1.33.1 release on Monday, December 5. Doug
At 2:00 PM -0500 12/3/05, Douglas Gregor wrote:
Boost 1.33.1 Release Candidate #1 is available at:
http://www.osl.iu.edu/~dgregor/boost-1.33.1-rc1/
If all goes well, this will become the Boost 1.33.1 release on Monday, December 5.
I downloaded and attempted to build this on my Mac. (Mac OS X 10.4.3, gcc 4.0.1): I used the command bjam -sTOOLS=darwin and received six link errors: ...failed darwin-Link-DyLib-action bin/boost/libs/test/build/libboost_unit_test_framework.dylib/darwin/release/shared-linkable-true/libboost_unit_test_framework-1_33_1.dylib... More specifically: /usr/bin/libtool: internal link edit command failed ld: Undefined symbols: __Z9test_mainiPPc --- and --- /usr/bin/libtool: internal link edit command failed ld: Undefined symbols: __Z20init_unit_test_suiteiPPc FYI: c++filt _Z20init_unit_test_suiteiPPc init_unit_test_suite(int, char**) c++filt _Z8cpp_mainiPPc cpp_main(int, char**) Running regression tests now...... -- -- Marshall Marshall Clow Idio Software <mailto:marshall@idio.com> It is by caffeine alone I set my mind in motion. It is by the beans of Java that thoughts acquire speed, the hands acquire shaking, the shaking becomes a warning. It is by caffeine alone I set my mind in motion.
and received six link errors:
...failed darwin-Link-DyLib-action bin/boost/libs/test/build/libboost_unit_test_framework.dylib/darwin/release/shared-linkable-true/libboost_unit_test_framework-1_33_1.dylib...
More specifically:
/usr/bin/libtool: internal link edit command failed ld: Undefined symbols: __Z9test_mainiPPc
I suspect those libraries can't be build as shared libraries on Darwin for the same reason that they can't be on Windows: they contain unresolved references to proceedures in the application that will link to them, this is allowed on some Unixes but not others. John.
On Sat, 03 Dec 2005 17:00:51 -0200, Douglas Gregor <doug.gregor@gmail.com> wrote:
Boost 1.33.1 Release Candidate #1 is available at:
http://www.osl.iu.edu/~dgregor/boost-1.33.1-rc1/
If all goes well, this will become the Boost 1.33.1 release on Monday, December 5.
The problem mentioned in http://lists.boost.org/boost-users/2005/08/13646.php http://lists.boost.org/Archives/boost/2005/11/97353.php persist. The documentation online on boost's website is alright. This seems like a documentation build error. Regards, Bruno
On Dec 3, 2005, at 7:31 PM, Bruno Martínez wrote:
On Sat, 03 Dec 2005 17:00:51 -0200, Douglas Gregor <doug.gregor@gmail.com> wrote: The problem mentioned in
http://lists.boost.org/boost-users/2005/08/13646.php http://lists.boost.org/Archives/boost/2005/11/97353.php
persist. The documentation online on boost's website is alright. This seems like a documentation build error.
Thanks for reporting this! I've now tracked it down: it seems that there was an entity resolution bug in the version of xsltproc used to generate the 1.33.1 documentation. Newer versions of xsltproc don't have this problem. I've regenerated the documentation and placed the results in CVS. The Boost.Threads docs read much better now. Doug
Douglas Gregor wrote:
Boost 1.33.1 Release Candidate #1 is available at:
I'm testing now, but I think I already see one error. IIRC, -I/usr/include shouldn't be used. -I should not be used with a path that is already standard for the system, as it can cause problems. Do I recall correctly that this was mentioned here before? [...] gcc-C++-action bin/boost/libs/regex/build/libboost_regex.so/gcc/release/shared-linkable-true/regex.o set -e "g++" -c -Wall -ftemplate-depth-255 -DNDEBUG -DNDEBUG -DBOOST_REGEX_DYN_LINK=1 -DBOOST_HAS_ICU=1 -O3 -finline-functions -Wno-inline -fPIC -I"bin/boost/libs/regex/build" -I "/home/nbecker/RPM/BUILD/boost_1_33_1" -I "/usr/include" -o "bin/boost/libs/regex/build/libboost_regex.so/gcc/release/shared-linkable-true/regex.o" "/home/nbecker/RPM/BUILD/boost_1_33_1/libs/regex/build/../src/regex.cpp"
I'm testing now, but I think I already see one error. IIRC, -I/usr/include shouldn't be used. -I should not be used with a path that is already standard for the system, as it can cause problems. Do I recall correctly that this was mentioned here before?
Are you building with -sICU_PATH=/usr ? This will inject that include path. If the ICU libraries are already installed in the usual places just use -sHAVE_ICU=1 instead, this is the prefered option and it's documented in the regex build instructions in libs/regex/doc/install.html#unicode and also in the getting started guide. Any help? John.
John Maddock wrote:
I'm testing now, but I think I already see one error. IIRC, -I/usr/include shouldn't be used. -I should not be used with a path that is already standard for the system, as it can cause problems. Do I recall correctly that this was mentioned here before?
Are you building with -sICU_PATH=/usr ? This will inject that include path.
If the ICU libraries are already installed in the usual places just use -sHAVE_ICU=1 instead, this is the prefered option and it's documented in the regex build instructions in libs/regex/doc/install.html#unicode and also in the getting started guide.
Any help?
Yes, you are correct. Thanks.
Neal Becker wrote:
Douglas Gregor wrote:
Boost 1.33.1 Release Candidate #1 is available at:
I'm testing now, but I think I already see one error. IIRC, -I/usr/include shouldn't be used. -I should not be used with a path that is already standard for the system, as it can cause problems. Do I recall correctly that this was mentioned here before?
Right, this is very dangerous, as it will make g++'s use of wrapper headers in conjunction with #include_next fail. Regards, Stefan
Douglas Gregor wrote:
Boost 1.33.1 Release Candidate #1 is available at:
http://www.osl.iu.edu/~dgregor/boost-1.33.1-rc1/
If all goes well, this will become the Boost 1.33.1 release on Monday, December 5.
I ran the tests on the "boost_1_33_1.exe" archive with some rather disastrous results. Something is broken with the MinGW toolset <http://tinyurl.com/bw3lz> that AFAIK isn't broken when testing the CVS version. Still investigating. -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - Grafik/jabber.org
On Dec 4, 2005, at 11:47 AM, Rene Rivera wrote:
Douglas Gregor wrote:
Boost 1.33.1 Release Candidate #1 is available at:
http://www.osl.iu.edu/~dgregor/boost-1.33.1-rc1/
If all goes well, this will become the Boost 1.33.1 release on Monday, December 5.
I ran the tests on the "boost_1_33_1.exe" archive with some rather disastrous results. Something is broken with the MinGW toolset <http://tinyurl.com/bw3lz> that AFAIK isn't broken when testing the CVS version. Still investigating.
Thanks for looking into this. Doug
Douglas Gregor wrote:
On Dec 4, 2005, at 11:47 AM, Rene Rivera wrote:
Douglas Gregor wrote:
Boost 1.33.1 Release Candidate #1 is available at:
http://www.osl.iu.edu/~dgregor/boost-1.33.1-rc1/
If all goes well, this will become the Boost 1.33.1 release on Monday, December 5.
I ran the tests on the "boost_1_33_1.exe" archive with some rather disastrous results. Something is broken with the MinGW toolset <http://tinyurl.com/bw3lz> that AFAIK isn't broken when testing the CVS version. Still investigating.
Thanks for looking into this.
Welcome... I think I found the problem, a confusion between two different mingw-3_4_2-tools.jam files. I'm rerunning the tests now so I'll find out shortly if it's resolved. -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - Grafik/jabber.org
Rene Rivera wrote:
Welcome... I think I found the problem, a confusion between two different mingw-3_4_2-tools.jam files. I'm rerunning the tests now so I'll find out shortly if it's resolved.
And a follow up... Looking at the running log it looks like it's working fine now. Results will show up eventually. -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - Grafik/jabber.org
participants (2)
-
Bruno Martínez -
Douglas Gregor -
John Maddock -
Marshall Clow -
Neal Becker -
Rene Rivera -
Stefan Seefeld