Do you have gcc configured on the command-line? I now need to add --compiler=gcc to the python ../blah/libtbx/configure.py command when building from source, to avoid clang. Best wishes, Graeme ________________________________________ From: [email protected] [[email protected]] on behalf of Jeffrey Van Voorst [[email protected]] Sent: 01 August 2012 17:26 To: cctbx mailing list Subject: Re: [cctbxbb] Problems compiling cctbx on OS X (10.8.0) Apple has been "messing around" with the Xcode and the OS to prefer CLang over gcc (at least that is what I have inferred). It could be that on Mac OS X 10.8 that scons chooses a different compiler, etc. than on Mac OS X 10.6.8 (where cctbx compiles fine). Hopefully scons has a method to print the toolchain or config to see what differs from previous versions of Mac OS X (e.g. 10.6.8 and 10.8.0) Xcode 3.* or Xcode 4.*? --Jeff _______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb -- 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

Out of curiosity: i remember reading somewhere that cctbx can be compiled with clang (and that it took a lot of work). Do you plan to support this compiler? Marcin On 1 August 2012 17:26, Jeffrey Van Voorst wrote:

Apple has been "messing around" with the Xcode and the OS to prefer CLang over gcc (at least that is what I have inferred). It could be that on Mac OS X 10.8 that scons chooses a different compiler, etc. than on Mac OS X 10.6.8 (where cctbx compiles fine).

Hopefully scons has a method to print the toolchain or config to see what differs from previous versions of Mac OS X (e.g. 10.6.8 and 10.8.0) Xcode 3.* or Xcode 4.*?

--Jeff _______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

On Aug 1, 2012, at 10:40 AM, Richard Gildea wrote:

As Graeme mentioned you currently need to use --compiler=gcc to force compilation with gcc over clang for the latest Xcode/OS X (we should probably try and figure out how to the configure command use gcc by default). The cctbx should build with the more recent versions of clang (perhaps Luc could comment on this as I think he uses clang regularly), although depending on what version of Xcode you are using, Apple used to ship a relatively old version of clang/llvm with Xcode, so clang compilation might not necessarily work straight out the box.

What output do you get after running "clang --version"?

After upgrading to OS X 10.8/Xcode 4.4 I get:

Apple clang version 4.0 (tags/Apple/clang-421.0.57) (based on LLVM 3.1svn)

which is more recent than with older versions of Xcode.

We should probably add a clang build to our nightly builds at some point in the near future.

Cheers,

Richard

This is getting progressively more confusing, doubtlessly compounded by the fact that I lose 5 points of IQ for every year I spend in Santa Cruz… cctbx appears to be using c++, which, for me, is sym linked to clang++ (g++ sym links to llvm-g++-4.2) If I cut and paste the failing command directly into my shell command-line, and manually reissue either the very same command, or use the full compiler path, be it /usr/bin/c++ or /usr/bin/clang++, or /usr/bin/g++, it works in each case.

The weirdest thing is that I can issue the commands that fail in any of the build scripts manually, and they work. i.e., if I re-issue c++ -o mmtbx/ncs/restraints_ext.o -c -DBOOST_ALL_NO_LIB -DBOOST_DISABLE_THREADS -DBOOST_PYTHON_MAX_BASES=2 -I/Users/wgscott/src/newcctbx/cctbx_sources/boost -I/Users/wgscott/src/newcctbx/cctbx_sources/boost -fPIC -fno-strict-aliasing -w -no-cpp-precomp -DNDEBUG -O3 -ffast-math -I/Users/wgscott/src/newcctbx/cctbx_build/include -I/Users/wgscott/src/newcctbx/cctbx_sources -I/System/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -I/Users/wgscott/src/newcctbx/cctbx_build/include -I/Users/wgscott/src/newcctbx/cctbx_sources /Users/wgscott/src/newcctbx/cctbx_sources/mmtbx/ncs/restraints_ext.cpp which is the line that failed in the error log I posted. This happens even if I invoke a new shell with /bin/tcsh -f , or if I create a new shell script executable file with any of the shells (ksh, zsh, bash, sh, tcsh). So it must be something within the environment that cctbx is creating. On Aug 2, 2012, at 5:29 AM, Luc Bourhis wrote:

...

I but I can't reproduce the problem with the cctbx trunk as of a few minutes ago.

Aha. I guess the bug you stumbled upon was fixed by rev 15462.

Best wishes,

Luc

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Hi Luc: I don't know how to find the version of Boost, but everything else is here: http://fennario.ucsc.edu/~wgscott/temp/everything.log I've hidden all "third-party" stuff like /usr/local and /sw, so the only (unmodified) python is what comes with 10.8, the only compilers are those supplied with the latest Xcode, and so on. This time I used the supplied cctbx_install_script.csh and put a -v in the shell invocation to provide verbose output. I invoked this from zsh, from bash, and from tcsh. Only the first of these has significant customizations, so I don't think it is my shell environment. I can't think of anything else. Since g++ works, I can't see how it could be anything other than clang. Bill William G. Scott Professor Department of Chemistry and Biochemistry and The Center for the Molecular Biology of RNA 228 Sinsheimer Laboratories University of California at Santa Cruz Santa Cruz, California 95064 USA On Aug 2, 2012, at 5:26 AM, Luc Bourhis wrote:

Hi William,

...

I am having the same problem compiling cctbx stand-alone, or within CCP4. I've posted the error log here:

http://fennario.ucsc.edu/~wgscott/temp/cctbx_compile_errors.log

From the information given in the discussion that followed, you have installed Xcode 4.4 command line tools. So do I but I can't reproduce the problem with the cctbx trunk as of a few minutes ago. From the error log you linked to, the compilation of mmtbx/bulk_solvent/bulk_solvent_ext.cpp fails. Could you copy me the exact compiler invocation? Moreover, although it should not matter considering the compiler errors at hand, could you tell me which version of Boost you used?

Finally, a general remark is in order: thou shall never try to work around such problems by playing with --compiler. Instead, please, loudly complain on this forum: the cctbx must compile with the latest Xcode tools and a build directory setup with simply "python configure.py mmtbx".

Best wishes,

Luc

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

Hi Richard: The tar ball isn't versioned, but has an md5sum of 9444c00f3c60ad855896e1032b2c5fd2. I vaguely recall the date being from about 3 months ago. This only started because I was trying to figure out why I couldn't get it to compile in ccp4. (It is the same problem. I am also having trouble compiling the ccp4 programs that rely on this, like phaser, aimless, pointless, even after I apply the previous fix. I haven't had a chance to check why yet, but it is a significant problem within the context the ccp4 6.3.0 released a couple of weeks ago.) Thanks. Bill William G. Scott Professor Department of Chemistry and Biochemistry and The Center for the Molecular Biology of RNA 228 Sinsheimer Laboratories University of California at Santa Cruz Santa Cruz, California 95064 USA On Aug 3, 2012, at 10:25 AM, Richard Gildea wrote:

Hi Bill,

What version of the cctbx sources are you using? As Luc mentioned, the first error from your log was resolved about 2 months ago with svn revision 15462 (http://cctbx.svn.sourceforge.net/viewvc/cctbx?view=revision&revision=15462). If you are using the latest "released" build (about 3 months old), then you should either update to latest nightly builds (http://cci.lbl.gov/builds/) or track the svn. It is known that the cctbx will not fully compile using clang with the older version of boost we currently distribute, however if you use a recent boost release it should be fine (it is also fine using --compiler=gcc). We are currently working on updating to the latest boost (the problem is with compatibility with some of the older platforms we support).

Hope that helps,

Richard

On 3 August 2012 08:53, William G. Scott wrote: Hi Luc:

I don't know how to find the version of Boost, but everything else is here:

http://fennario.ucsc.edu/~wgscott/temp/everything.log

I've hidden all "third-party" stuff like /usr/local and /sw, so the only (unmodified) python is what comes with 10.8, the only compilers are those supplied with the latest Xcode, and so on.

This time I used the supplied cctbx_install_script.csh and put a -v in the shell invocation to provide verbose output.

I invoked this from zsh, from bash, and from tcsh. Only the first of these has significant customizations, so I don't think it is my shell environment.

I can't think of anything else. Since g++ works, I can't see how it could be anything other than clang.

Bill

William G. Scott Professor Department of Chemistry and Biochemistry and The Center for the Molecular Biology of RNA 228 Sinsheimer Laboratories University of California at Santa Cruz Santa Cruz, California 95064 USA

On Aug 2, 2012, at 5:26 AM, Luc Bourhis wrote:

...

Hi William,

...

I am having the same problem compiling cctbx stand-alone, or within CCP4. I've posted the error log here:

http://fennario.ucsc.edu/~wgscott/temp/cctbx_compile_errors.log

From the information given in the discussion that followed, you have installed Xcode 4.4 command line tools. So do I but I can't reproduce the problem with the cctbx trunk as of a few minutes ago. From the error log you linked to, the compilation of mmtbx/bulk_solvent/bulk_solvent_ext.cpp fails. Could you copy me the exact compiler invocation? Moreover, although it should not matter considering the compiler errors at hand, could you tell me which version of Boost you used?

Finally, a general remark is in order: thou shall never try to work around such problems by playing with --compiler. Instead, please, loudly complain on this forum: the cctbx must compile with the latest Xcode tools and a build directory setup with simply "python configure.py mmtbx".

Best wishes,

Luc

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

_______________________________________________ cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb

On 8/3/12 3:27 PM, Jeffrey Van Voorst wrote:

On 8/3/12 3:07 PM, William G. Scott wrote:

...

Hi Richard:

The tar ball isn't versioned, but has an md5sum of 9444c00f3c60ad855896e1032b2c5fd2.

In the cctbx_source directory there is a TAG file that should give you the date of the tarball.

During my multitasking I failed to remember that the latest release doesn't have the fix. If you are using a release of cctbx, there will be a TAG file, but that is not relevant because the fixes have not been released. Sorry for the noise.