Thanks for your suggestions. They did not do the trick though.
Some more detective work got me down to
../base/bin/python ../modules/cctbx_project/libtbx/configure.py dials
failing and both
libtbx.python ../modules/cctbx_project/libtbx/configure.py dials
and
LD_LIBRARY_PATH=lib ../base/bin/python ../modules/cctbx_project/libtbx/configure.py dials
succeeding.
A run with LD_DEBUG=files identified the problem:
/lib64/libboost_python.so: error: symbol lookup error: undefined symbol: PyUnicodeUCS4_FromEncodedObject (fatal)
CentOS7 comes with its own set of incompatible boost libraries.
I've put a fix in: http://sourceforge.net/p/cctbx/code/22986/
Installation on CentOS7 now works for me. As far as I could tell this does not break other things.
-Markus
________________________________
From: [email protected] [[email protected]] on behalf of Nigel Moriarty [[email protected]]
Sent: Tuesday, September 29, 2015 18:36
To: cctbx mailing list
Subject: Re: [cctbxbb] CentOS7 bugs (cont.)
libtbx.configure --scan-boost phenix
Cheers
Nigel
---
Nigel W. Moriarty
Building 33R0349, Physical Biosciences Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8235
Phone : 510-486-5709 Email : [email protected]
Fax : 510-486-5909 Web : CCI.LBL.govhttp://CCI.LBL.gov
On Tue, Sep 29, 2015 at 10:20 AM, Nicholas Sauter mailto:[email protected]> wrote:
Nigel,
Do remember how to force libtbx.scons to scan the boost directory?
Nick
On Tue, Sep 29, 2015 at 9:25 AM, Nigel Moriarty mailto:[email protected]> wrote:
Markus
The default usage of libtbx.scons in the build directory does not scan the boost directory so maybe that has something to do with the .so not behaving on Centos7 - because it was built on Centos5/6?
Cheers
Nigel
---
Nigel W. Moriarty
Building 33R0349, Physical Biosciences Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8235
Phone : 510-486-5709tel:510-486-5709 Email : [email protected]
Fax : 510-486-5909tel:510-486-5909 Web : CCI.LBL.govhttp://CCI.LBL.gov
On Tue, Sep 29, 2015 at 8:29 AM, mailto:[email protected]> wrote:
Hello cctbx folks,
We are having problems running, building and/or installing DIALS on CentOS7. Running the linux installer unceremoniously ends in a segfault.
The installer is even broken when it is build on the same system using bootstrapped cctbx/dials.
The broken installer can be traced to the DIALS extension loader, which uses python’s pkgutil loader, which triggers some library scanning routine within python, which in turn falls over boost_python_meta_ext.so and dies with the strace shown in the email below.
The installer can be successfully “fixed” by deleting build/lib/boost_python_meta_ext.so during the installation.
So I see two issues here:
1. What is broken with build/lib/boost_python_meta_ext.so that it causes segfaults on CentOS7?
2. Why is this not covered by any test? Neither of the following caught anything:
a. python bootstrap.py builder=dials tests
b. libtbx.run_tests_parallel module=cctbx module=dials module=boost_adaptbx
[NB: The previously reported 3 broken tests on CentOS7 are red herrings and are all fixed by running ‘yum update’]
I don’t know how to proceed from here.
-Markus
From: [email protected]mailto:[email protected] [mailto:[email protected]mailto:[email protected]]
Sent: 29 September 2015 13:29
To: [email protected]mailto:[email protected]
Subject: Re: [Dials-support] CentOS7 bugs (cont.)
Hello everyone,
(..)
Sorry, I hit ctrl+enter by accident…
In addition to these failed tests the installer dies a boost-related death according to strace:
(..)
poll([{fd=13, events=POLLIN|POLLPRI}, {fd=15, events=POLLIN|POLLPRI}], 2, 4294967295) = 1 ([{fd=13, revents=POLLIN}]) read(13, "\tlinux-vdso.so.1 => (0x00007fff"..., 4096) = 486
--- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_EXITED, si_pid=7648, si_status=0, si_utime=0, si_stime=0} --- poll([{fd=13, events=POLLIN|POLLPRI}, {fd=15, events=POLLIN|POLLPRI}], 2, 4294967295) = 2 ([{fd=13, revents=POLLHUP}, {fd=15, revents=POLLHUP}])
close(13) = 0
close(15) = 0
wait4(7648, [{WIFEXITED(s) && WEXITSTATUS(s) == 0}], 0, NULL) = 7648 stat("$prefix/dials-dev20150928/build/lib/boost_python_meta_ext", 0x7fff6421d3f0) = -1 ENOENT (No such file or directory) open("$prefix/dials-dev20150928/build/lib/boost_python_meta_ext.so", O_RDONLY) = 11 fstat(11, {st_mode=S_IFREG|0775, st_size=200624, ...}) = 0
--- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x18a86} ---
+++ killed by SIGSEGV (core dumped) +++
Segmentation fault (core dumped)
Not sure if relevant, but CentOS7 compiler is
GCC 4.8.3 20140911 (Red Hat 4.8.3-9)
compared to RHEL6
GCC 4.4.7 20120313 (Red Hat 4.4.7-11)
Anybody have any ideas?
-Markus
--
This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail.
Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd.
Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message.
Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom
--
This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail.
Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd.
Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message.
Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom
_______________________________________________
cctbxbb mailing list
[email protected]mailto:[email protected]
http://phenix-online.org/mailman/listinfo/cctbxbb
_______________________________________________
cctbxbb mailing list
[email protected]mailto:[email protected]
http://phenix-online.org/mailman/listinfo/cctbxbb
--
Nicholas K. Sauter, Ph. D.
Computer Staff Scientist, Physical Biosciences Division
Lawrence Berkeley National Laboratory
1 Cyclotron Rd., Bldg. 64R0121
Berkeley, CA 94720-8118
(510) 486-5713tel:%28510%29%20486-5713
_______________________________________________
cctbxbb mailing list
[email protected]mailto:[email protected]
http://phenix-online.org/mailman/listinfo/cctbxbb
