[cctbxbb] Fw: Build constraints in SConscript?

Graeme.Winter at diamond.ac.uk Graeme.Winter at diamond.ac.uk
Fri May 15 10:16:07 PDT 2015


Nigel

If you export the environment variable I sent before before kicking off a build you should make fully portable binaries. We've been doing this in dials for a while.

Best wishes Graeme

On 15 May 2015 18:00, Nigel Moriarty <nwmoriarty at lbl.gov> wrote:
Greame

Did you try replacing 10.8 in the SConscript file with 10.7? Please do so and let me know if this works.

I was working on getting the mechanism to be OS specific on the buildbot and bootstrap when Rob solved the problem independently. If you don't tell me if changing 10.8 to 10.7 works I will do something to get it working here and we can worry about your 2011 vintage OS next week.

Cheers

Nigel

---
Nigel W. Moriarty
Building 64R0246B, Physical Biosciences Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720-8235
Phone : 510-486-5709     Email : NWMoriarty at LBL.gov
Fax   : 510-486-5909       Web  : CCI.LBL.gov<http://CCI.LBL.gov>

On Fri, May 15, 2015 at 2:31 AM, <Graeme.Winter at diamond.ac.uk<mailto:Graeme.Winter at diamond.ac.uk>> wrote:
Dear Rob

Given that the tests are currently not working right anyway, and there are no build / deployment tests for testing builds on an older version of OS X than they are compiled on, this is an easy problem to get in to. It's also a collaborative project where we're all working hard on making things work. This will happen from time to time.

I had thought the same environment variable based mechanism for the builds was already in place for PHENIX - am surprised it is not!

So, no worries. Perhaps I came across more grumpy than I meant to below.

Occurs to me it would be very handy to have one email address for all cctbx related developers where (someone) could ask (something like this) and get an answer perhaps suggesting a more optimum solution? Thoughts anyone? (lots of mail lists in send: list!)

Thanks & best wishes Graeme

-----Original Message-----
From: R.D. Oeffner [mailto:rdo20 at cam.ac.uk<mailto:rdo20 at cam.ac.uk>]
Sent: 15 May 2015 10:26
To: Winter, Graeme (DLSLtd,RAL,LSCI); cctbxbb at phenix-online.org<mailto:cctbxbb at phenix-online.org>; dials-support at lists.sourceforge.net<mailto:dials-support at lists.sourceforge.net>; dials-dev at cci.lbl.gov<mailto:dials-dev at cci.lbl.gov>; Nigel W Moriarty
Subject: Re: Fw: Build constraints in SConscript?

Dear Graeme,

I'm sorry for the trouble my change caused and the time taken to identify it.
Ironically we had the same issue as you with Phenix not installing on anything lower than OSX 10.9. As nothing appeared to be specified in our build system to that effect my change was a quick fix to address the problem. It was genuinely well intended.

Nigel, I defer to you for implementing the minimum OSX platform for Phenix builds.


Rob






> FROM: Graeme.Winter at Diamond.ac.uk<mailto:Graeme.Winter at Diamond.ac.uk>
> SENT: Friday, May 15, 2015 9:09 AM
> TO: Graeme.Winter at Diamond.ac.uk<mailto:Graeme.Winter at Diamond.ac.uk> ; cctbxbb at phenix-online.org<mailto:cctbxbb at phenix-online.org> ;
> robert at oeffner.net<mailto:robert at oeffner.net>
> CC: dials-support at lists.sourceforge.net<mailto:dials-support at lists.sourceforge.net> ; dials-dev at cci.lbl.gov<mailto:dials-dev at cci.lbl.gov>
> SUBJECT: RE: Build constraints in SConscript?
>
> [gw56 at ws133 libtbx]$ svn commit -m "Reverting with extreme prejudice
> hard coded 10.8 requirement, build no longer works on 10.7 and
> earlier; please use ENV VAR method to achieve this kind of constraint;
> reverted revision=22202" SConscript
>
> Sending SConscript
>
> Transmitting file data .
>
> Committed revision 22223.
>
> Please use MACOSX_DEPLOYMENT_TARGET=(version) locally for your
> personal requirement, export this before kicking off bootstrap.py
>
> Best wishes Graeme
>
> ==================== DIFFS FOLLOW ====================
>
> Revision: 22223
>  http://sourceforge.net/p/cctbx/code/22223 [1]
> Author: graeme_winter
> Date: 2015-05-15 08:06:10 +0000 (Fri, 15 May 2015)
> Log Message:
> -----------
> Reverting with extreme prejudice hard coded 10.8 requirement, build no
> longer works on 10.7 and earlier; please use ENV VAR method to achieve
> this kind of constraint; reverted revision=22202
>
> Modified Paths:
> --------------
>  trunk/libtbx/SConscript
>
> Modified: trunk/libtbx/SConscript
> ===================================================================
> --- trunk/libtbx/SConscript 2015-05-15 02:13:42 UTC (rev 22222)
> +++ trunk/libtbx/SConscript 2015-05-15 08:06:10 UTC (rev 22223)
> @@ -803,7 +803,7 @@
>  env_etc.static_libraries = 1
>  env_etc.static_bpl = 0
>
> - link_flags = ["-w", "-mmacosx-version-min=10.8"] # suppress
> "source/lib does not exist" warning
> + link_flags = ["-w"] # suppress "source/lib does not exist" warning
>  if libtbx.env.build_options.force_32bit:
>  if (env_etc.gcc_version and env_etc.gcc_version >= 40201) or
> env_etc.clang_version:
>  link_flags.extend(["-arch", "i386"])
> @@ -811,8 +811,7 @@
>  env_etc.c_link = "cc"
>  env_etc.ccflags_base = [
>  "-fPIC",
> - "-fno-strict-aliasing",
> - "-mmacosx-version-min=10.8"]
> + "-fno-strict-aliasing"]
>  if env_etc.clang_version:
>  setup_clang_warning(env_etc, libtbx.env.build_options.warning_level)
>  else:
> @@ -828,7 +827,7 @@
>  env_etc.ccflags_base.extend(["-no-cpp-precomp"])
>  if (env_etc.gcc_version < 40201):
>  env_etc.ccflags_base.append("-Wno-long-double")
> - env_etc.cxxflags_base = ["-mmacosx-version-min=10.8"]
> + env_etc.cxxflags_base = []
>  if env_etc.clang_version is None and env_etc.gcc_version < 40000:
>  env_etc.cxxflags_base.append("-fcoalesce-templates")
>  env_etc.cxxflags_base.append("-DBOOST_ALL_NO_LIB")
> @@ -861,11 +860,10 @@
>  if env_etc.apple_gcc_builds.gcc < 5646:
>  # fixed on 10.6 (Snow Leopard) whose Dev Tools start with that build
>  opts.insert(0, "-gstabs+")
> - env_etc.shlinkflags = ["-mmacosx-version-min=10.8"]
> + env_etc.shlinkflags = []
>  env_etc.shlinkflags_bpl = [
>  "-w", # suppress "source/lib does not exist" warning
> - "-bundle", "-mmacosx-version-min=10.8",
> - "-undefined", "dynamic_lookup"]
> + "-bundle", "-undefined", "dynamic_lookup"]
>  if libtbx.env.build_options.force_32bit:
>  if (env_etc.gcc_version and env_etc.gcc_version >= 40201) or
> env_etc.clang_version:
>  env_etc.ccflags_base.extend(["-arch", "i386"])
>
> FROM: Graeme.Winter at diamond.ac.uk<mailto:Graeme.Winter at diamond.ac.uk> [mailto:Graeme.Winter at diamond.ac.uk<mailto:Graeme.Winter at diamond.ac.uk>]
>
> SENT: 15 May 2015 09:01
> TO: cctbxbb at phenix-online.org<mailto:cctbxbb at phenix-online.org>; robert at oeffner.net<mailto:robert at oeffner.net>
> CC: dials-support at lists.sourceforge.net<mailto:dials-support at lists.sourceforge.net>; dials-dev at cci.lbl.gov<mailto:dials-dev at cci.lbl.gov>
> SUBJECT: Build constraints in SConscript?
>
> Hi Folks
>
> A while back we worked on making DIALS work right built on 10.10 but
> executed on 10.6 or later – this is very useful as lots of people
> have 10.6/7 macs & are not keen to update. This is done by the very
> flexible mechanism of exporting an environment variable prior to
> compilation
>
> Then this commit appears
>
>  env_etc.static_bpl = 0
> @@ -802,7 +803,7 @@
>  env_etc.static_libraries = 1
>  env_etc.static_bpl = 0
>
> - link_flags = ["-w"] # suppress "source/lib does not exist" warning
> + link_flags = ["-w", "-mmacosx-version-min=10.8"] # suppress
> "source/lib does not exist" warning
>  if libtbx.env.build_options.force_32bit:
>  if (env_etc.gcc_version and env_etc.gcc_version >= 40201) or
> env_etc.clang_version:
>  link_flags.extend(["-arch", "i386"])
> @@ -810,7 +811,8 @@
>  env_etc.c_link = "cc"
>  env_etc.ccflags_base = [
>  "-fPIC",
> - "-fno-strict-aliasing"]
> + "-fno-strict-aliasing",
> + "-mmacosx-version-min=10.8"]
>  if env_etc.clang_version:
>  setup_clang_warning(env_etc, libtbx.env.build_options.warning_level)
>  else:
> @@ -826,7 +828,7 @@
>  env_etc.ccflags_base.extend(["-no-cpp-precomp"])
>  if (env_etc.gcc_version < 40201):
>  env_etc.ccflags_base.append("-Wno-long-double")
> - env_etc.cxxflags_base = []
> + env_etc.cxxflags_base = ["-mmacosx-version-min=10.8"]
>  if env_etc.clang_version is None and env_etc.gcc_version < 40000:
>  env_etc.cxxflags_base.append("-fcoalesce-templates")
>  env_etc.cxxflags_base.append("-DBOOST_ALL_NO_LIB")
> @@ -859,10 +861,10 @@
>
> to
>
> Modified Paths:
> --------------
>  trunk/libtbx/SConscript
>
> and now our build no longer works on 10.7. This is *REALLY* annoying.
> The head of the commit message follows below.
>
> Please could someone fix this? The environment variable method is
> *FAR* better, works for the dependencies and also allows dials to work
> on older machines. You broke our build. Before this it worked fine.
>
> (we can’t automate this test as we are not supposed to deploy out of
> support machines on lab network, so this was missed)
>
> Thanks Graeme
>
> EFFNER
>
>  to cctbx-cvs
>
> May 13
>
> Revision: 22202
>  http://sourceforge.net/p/cctbx/code/22202 [2]
> Author: oeffner
> Date: 2015-05-13 15:48:24 +0000 (Wed, 13 May 2015)
> Log Message:
> -----------
> ensure installers can run on MacOSX 10.8 or higher
>
> --
>
> 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
>
> No virus found in this message.
> Checked by AVG - www.avg.com<http://www.avg.com> [3]
> Version: 2015.0.5941 / Virus Database: 4342/9780 - Release Date:
> 05/15/15
>
> Links:
> ------
> [1] http://sourceforge.net/p/cctbx/code/22223
> [2] http://sourceforge.net/p/cctbx/code/22202
> [3] http://www.avg.com

--
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




More information about the cctbxbb mailing list