Re: [Plplot-devel] Results of comprehensive PLplot builds and tests for cmake-2.8.11-rc3

[Doug H. <dh...@uc...>]

Hi Alan:  I noticed that you did not run the perl/PDL tests.  Are they 
broken?

Thanks,

   Doug

dh...@uc...
Software Engineer
UCAR - COSMIC, Tel. (303) 497-2611

On Wed, 24 Apr 2013, Alan W. Irwin wrote:

> Today I decided to comprehensively test PLplot because it has been a
> long time since I did that, and there is a new CMake version
> cmake-2.8.11-rc3 that is about to get released as cmake-2.8.11 unless
> cmake users (such as us) can find any issues.
>
> After cmake-2.8.11-rc3 was downloaded, built (using the bootstrap
> method), and installed, I did the comprehensive test as follows:
>
> time (nice -19 scripts/comprehensive_test.sh --do_test_interactive no)
>
> (I turned off the interactive testing since otherwise I would have
> had to babysit the test clicking on many GUI's to get the interactive
> tests completed.)
>
> That comprehensive test required a number of minor PLplot fixups to be
> made (see my recent commits) which reaffirms that these comprehensive
> tests are worth doing.
>
> After those changes, the above run of comprehensive_test.sh took
> 51 minutes to complete (and from
> prior notes) on my entry-level PC and requires about 4GB of disk space
> as a high-water mark to store the results of each of the many
> components of the tests that are done before the resulting files are
> cleaned.  There were about 290M of files left at the end of the
> comprehensive test.
>
> To look for meaningful error messages in those tests, I ran
>
> find ../comprehensive_test_disposeable/ -name "*.out*" | \
> xargs grep -i error |grep -v test.error
>
> That shows there are linking errors for the case of the static library
> build + traditional (Make + pkg-config) build of the installed
> examples.  This used to work fine, but I guess the Debian wheezy
> toolchain just doesn't handle this case properly any more (or takes
> more care with linking of code generated from a variety of languages).
> Note, I got the same bad results for this case when using
> cmake-2.8.10.2 so this is not a cmake regression.
>
> Andrew, would you be willing to take a look?
>
> In contrast to the traditional build of the installed examples, the
> CMake build system for these installed examples works fine for the
> static library build.  So it appears we could fix up these
> linking issues for the traditional build by following what
> is done for the cmake-based build of the installed examples.
>
> I also looked for warnings in the comprehensive test results and
> got the following results:
>
> software@raven> find ../comprehensive_test_disposeable/ -name "*.out*" | \
> xargs grep -i warn |grep -v imultilib |grep -v  "PLPLOT WARNING" | \
> grep -v perl |grep -v AGG
> ../comprehensive_test_disposeable/shared/output_tree/installed_make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/shared/output_tree/make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/shared/output_tree/ctest.out:25: WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/shared/output_tree/make.out:sip: Deprecation warning: /home/software/plplot_svn/HEAD/plplot_allura/bindings/qt_gui/pyqt4/plplot_pyqt4.sip:25: %Module version number should be specified using the 'version' argument
> ../comprehensive_test_disposeable/shared/output_tree/make.out:/home/software/plplot_svn/HEAD/comprehensive_test_disposeable/shared/build_tree/config.h:55:0: warning: "HAVE_FREETYPE" redefined [enabled by default]
> ../comprehensive_test_disposeable/shared/output_tree/make.out:/home/software/plplot_svn/HEAD/plplot_allura/bindings/tk/plserver.c:134:9: warning: passing argument 2 of ?Tcl_SetResult? discards ?const? qualifier from pointer target type [enabled by default]
> ../comprehensive_test_disposeable/shared/output_tree/traditional_make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/nondynamic/output_tree/installed_make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/nondynamic/output_tree/make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/nondynamic/output_tree/cmake.out:-- WARNING: ENABLE_DYNDRIVERS is OFF so force PLD_extqt to be OFF
> ../comprehensive_test_disposeable/nondynamic/output_tree/cmake.out:-- WARNING: PLD_extqt is OFF so setting ENABLE_qt to OFF.
> ../comprehensive_test_disposeable/nondynamic/output_tree/cmake.out:-- WARNING: ENABLE_qt is OFF so setting ENABLE_pyqt4 to OFF.
> ../comprehensive_test_disposeable/nondynamic/output_tree/ctest.out:25: WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/nondynamic/output_tree/make.out:/home/software/plplot_svn/HEAD/comprehensive_test_disposeable/nondynamic/build_tree/config.h:55:0: warning: "HAVE_FREETYPE" redefined [enabled by default]
> ../comprehensive_test_disposeable/nondynamic/output_tree/make.out:/home/software/plplot_svn/HEAD/plplot_allura/bindings/tk/plserver.c:134:9: warning: passing argument 2 of ?Tcl_SetResult? discards ?const? qualifier from pointer target type [enabled by default]
> ../comprehensive_test_disposeable/nondynamic/output_tree/traditional_make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/static/output_tree/installed_make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/static/output_tree/make_noninteractive.out:WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: Java requires shared libraries. Disabling java bindings
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: Python requires shared libraries. Disabling Python bindings
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: Octave requires shared libraries. Disabling octave bindings
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: Disabling Perl/PDL examples because BUILD_SHARED_LIBS=OFF
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: Lua requires shared libraries. Disabling Lua bindings
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: Shared libraries not built. Setting ENABLE_DYNDRIVERS OFF.
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: ENABLE_DYNDRIVERS is OFF so force PLD_extqt to be OFF
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: PLD_extqt is OFF so setting ENABLE_qt to OFF.
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING: ENABLE_python is OFF so setting ENABLE_pyqt4 to OFF.
> ../comprehensive_test_disposeable/static/output_tree/cmake.out:-- WARNING:Static build with ENABLE_ocaml_static false.  Therefore, disabling ocaml bindings
> ../comprehensive_test_disposeable/static/output_tree/ctest.out:20: WARNING: Some PostScript or stdout results were different
> ../comprehensive_test_disposeable/static/output_tree/make.out:/home/software/plplot_svn/HEAD/plplot_allura/bindings/tk/plserver.c:134:9: warning: passing argument 2 of ?Tcl_SetResult? discards ?const? qualifier from pointer target type [enabled by default]
>
> Most of these warnings are harmless, but it would be good to get rid of
> all of warnings associated with "sip: Deprecation", the HAVE_FREETYPE redefinition, the
> plserver.c build, and the "WARNING: Some PostScript or stdout results
> were different" message.  That latter is because the changes to
> examples/c/x16c.c have not been propagated only to examples/c++ so far.
> Here is the detailed report covering that:
>
> c++
>   Missing examples            :
>   Differing postscript output :
>   Missing stdout              :
>   Differing stdout            :
> f95
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> java
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> octave
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> python
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> tcl
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> ada
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> adathick
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> ocaml
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> lua
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
> d
>   Missing examples            :
>   Differing postscript output :  16
>   Missing stdout              :
>   Differing stdout            :
>
> In sum these comprehensive tests indicate no issues for
> cmake-2.8.11-rc3, serious linking issues for the
> traditional build of the installed examples for the static
> library case (these occur also for cmake-2.8.10.2); less serious
> warnings detailed above which should nevertheless be
> fixed, and example 16 propagation issues which should be fixed.
>
> Alan
> __________________________
> Alan W. Irwin
>
> Astronomical research affiliation with Department of Physics and Astronomy,
> University of Victoria (astrowww.phys.uvic.ca).
>
> Programming affiliations with the FreeEOS equation-of-state
> implementation for stellar interiors (freeeos.sf.net); the Time
> Ephemerides project (timeephem.sf.net); PLplot scientific plotting
> software package (plplot.sf.net); the libLASi project
> (unifont.org/lasi); the Loads of Linux Links project (loll.sf.net);
> and the Linux Brochure Project (lbproject.sf.net).
> __________________________
>
> Linux-powered Science
> __________________________
>
> ------------------------------------------------------------------------------
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> _______________________________________________
> Plplot-devel mailing list
> Plp...@li...
> https://lists.sourceforge.net/lists/listinfo/plplot-devel
>