Release branch regression testing failures
Stale tests have been cleared, and there is a fresh release branch regression test report up at: http://beta.boost.org/development/tests/release/developer/summary.html Libraries reporting errors on more than one platform are: bimap - all platforms failing build - all non-Linux platforms failing math - all platforms failing numeric/interval - two platforms failing optional - two platforms failing python - Linux platforms failing spirit - all platforms failing type_traits - all platforms failing units - two platforms failing I'd appreciate it if the developers involved would clear these tests in the next few days. Thanks, --Beman
On Wed, Jul 22, 2009 at 4:20 PM, Beman Dawes<bdawes@acm.org> wrote:
Stale tests have been cleared, and there is a fresh release branch regression test report up at:
http://beta.boost.org/development/tests/release/developer/summary.html
Libraries reporting errors on more than one platform are:
bimap - all platforms failing
Everything is fine except for two tests that involves the Property Map library. I have read that Jeremiah Willcock move the headers up to a separate folder, but leaving a forward header to maintain everything working. Has the library change in some other way Jeremiah? I think this test should work with out modifications.
I'd appreciate it if the developers involved would clear these tests in the next few days.
I will try to get all that boxes green as soon as possible. Best regards Matias
On Wed, Jul 22, 2009 at 5:40 PM, Matias Capeletto<matias.capeletto@gmail.com> wrote:
On Wed, Jul 22, 2009 at 4:20 PM, Beman Dawes<bdawes@acm.org> wrote:
Stale tests have been cleared, and there is a fresh release branch regression test report up at:
http://beta.boost.org/development/tests/release/developer/summary.html
Libraries reporting errors on more than one platform are:
bimap - all platforms failing
Everything is fine except for two tests that involves the Property Map library. I have read that Jeremiah Willcock move the headers up to a separate folder, but leaving a forward header to maintain everything working. Has the library change in some other way Jeremiah? I think this test should work with out modifications.
Jeremiah's commits fix everything in Bimap tests. Thanks a lot. They are all green now. Regards Matias
-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Beman Dawes Sent: 22 July 2009 20:20 To: Boost Developers list Subject: [boost] Release branch regression testing failures
Stale tests have been cleared, and there is a fresh release branch regression test report up at:
http://beta.boost.org/development/tests/release/developer/summary.html
Libraries reporting errors on more than one platform are:
math - all platforms failing
The math failures on all platforms are caused by a missing file. /home/beman/boost/release-test/boost/libs/math/config/has_mpfr_class.cpp:13: 21: error: gmpfrxx.h: No such file or directory The failure will not affect user of the code at all (who would normally provide the latest version of this GMP file anyway), but I am sure John Maddock will be on the case soon ;-) Paul --- Paul A. Bristow Prizet Farmhouse Kendal, UK LA8 8AB +44 1539 561830, mobile +44 7714330204 pbristow@hetp.u-net.com
John,
math - all platforms failing
Hopefully fixed now.
type_traits - all platforms failing
Should also be fixed - I forgot to merge one file :-(
We have a single failing use case in Spirit (here: http://tinyurl.com/lyz5wo) occurring only on a ppc system running darwin-4.0.1, which is related to math::concepts::real_concept. Are you doing something special there? Or is it a compiler issue? Thanks! Regards Hartmut
On Jul 23, 2009, at 9:40 AM, Hartmut Kaiser wrote:
John,
math - all platforms failing
Hopefully fixed now.
type_traits - all platforms failing
Should also be fixed - I forgot to merge one file :-(
We have a single failing use case in Spirit (here: http://tinyurl.com/lyz5wo) occurring only on a ppc system running darwin-4.0.1, which is related to math::concepts::real_concept. Are you doing something special there? Or is it a compiler issue?
Hartmut -- I have a Darwin PPC system with gcc 4.0.1 installed on it. Let me know if you need any specific tests run.... - Marshall
on Wed Jul 22 2009, Beman Dawes <bdawes-AT-acm.org> wrote:
python - Linux platforms failing
These have the usual cause: improperly/incompletely configured test machine. There's no information about the BGD-Ubuntu tester available by click-through, so I don't know what to tell you. -- Dave Abrahams BoostPro Computing http://www.boostpro.com
On Thu, Jul 23, 2009 at 09:30:14AM -0400, David Abrahams wrote:
on Wed Jul 22 2009, Beman Dawes <bdawes-AT-acm.org> wrote:
python - Linux platforms failing
These have the usual cause: improperly/incompletely configured test machine.
OK. I run one of the test machines ("Debian-Sid"), can you help me completely and properly configure it? I'm testing with python 2.5.4. There is one python test failing for me (http://tinyurl.com/lswbtc) which produces an exception: ********************************************************************** File "../libs/python/test/slice.py", line 23, in __main__ Failed example: try: from numarray import array, all except: print 1 else: check_numeric_array_rich_slice('numarray', 'NDArray', all) Exception raised: Traceback (most recent call last): File "/usr/lib/python2.5/doctest.py", line 1228, in __run compileflags, 1) in test.globs File "<doctest __main__[4]>", line 6, in <module> check_numeric_array_rich_slice('numarray', 'NDArray', all) File "/usr/lib/python2.5/site-packages/numarray/generic.py", line 652, in _dualbroadcast raise ValueError("Arrays have incompatible shapes"); ValueError: Arrays have incompatible shapes ********************************************************************** I'm willing to help debug this, but I don't know where to begin. Thanks, -Steve
On Thu, Jul 23, 2009 at 9:30 AM, David Abrahams<dave@boostpro.com> wrote:
on Wed Jul 22 2009, Beman Dawes <bdawes-AT-acm.org> wrote:
python - Linux platforms failing
These have the usual cause: improperly/incompletely configured test machine. There's no information about the BGD-Ubuntu tester available by click-through, so I don't know what to tell you.
Hum... user-config.jam contains: using python ; Looking at the log from running bjam in boost-root/libs/python/example/quickstart: ...patience... ...patience... ...found 1510 targets... ...updating 37 targets... gcc.compile.c++ bin/gcc-4.3.3/debug/extending.o In file included from ../../../../boost/python/detail/prefix.hpp:13, from ../../../../boost/python/class.hpp:8, from extending.cpp:5: ../../../../boost/python/detail/wrap_python.hpp:75:24: error: patchlevel.h: No such file or directory ../../../../boost/python/detail/wrap_python.hpp:78:2: error: #error Python 2.2 or higher is required for this version of Boost.Python. ../../../../boost/python/detail/wrap_python.hpp:142:21: error: Python.h: No such file or directory ... The "Python 2.2 or higher is required for this version of Boost.Python" message is somewhat confusing since Python 2.6.2 is in fact installed. It would be clearer if it said "Python 2.2 or higher headers are required for this version of Boost.Python; have you installed the Python development system?" Once the development package was installed the quickstart tests passed, so that's probably what was causing the regressions to fail, too. Thanks, --Beman
On Thu, Jul 23, 2009 at 11:11 PM, Steven Watanabe<watanabesj@gmail.com> wrote:
AMDG
Beman Dawes wrote:
Libraries reporting errors on more than one platform are:
units - two platforms failing
These should be fixed in the trunk now. I'll merge as soon as the tests cycle.
Thanks! And thanks to everyone else who has been clearing release regression test problems. The failures are already much reduced. I take this to mean most of the failures are minor snafus, and that release branch testing will soon be good enough for a beta. --Beman
Beman Dawes wrote:
Stale tests have been cleared, and there is a fresh release branch regression test report up at:
http://beta.boost.org/development/tests/release/developer/summary.html
Libraries reporting errors on more than one platform are:
bimap - all platforms failing build - all non-Linux platforms failing
Windows is now all clear. OSX still have from 2 to 3 failures depending on the version, I hope to fix that up during this week. - Volodya
participants (10)
-
Beman Dawes -
David Abrahams -
Hartmut Kaiser -
John Maddock -
Marshall Clow -
Matias Capeletto -
Paul A. Bristow -
Steve M. Robbins -
Steven Watanabe -
Vladimir Prus