Problem compiling sage 6.2.rc0-develop with macosx/mavericks

[Andrew]

Hello,

I pulled the latest development version (branch `develop=6.2.rc0`) and tried to compile it using mavericks and xcode 5.1.1. Now even

make distclean

is failing. Similarly, using `make -k` compiles a binary that crashes on startup (error message at end).

Although it is may well be some issue with my local configuration (or xcode/mavericks) , it is also an issue with `6.2.rc0` because if I change my git branch back to `6.1.1-master` then sage compiles without complaint -- well, at least it gets pass all of the problems encountered by 6.2.rc0 (compile still in progress).

Any ideas what's going wrong?
Andrew

--
The complete log file from `make distclean` is attached but the end of it is:

gcc -std=gnu99 -dynamiclib -Wl,-headerpad_max_install_names -undefined dynamic_lookup -single_module -multiply_defined suppress -L/usr/local/src/sage/local/lib/ -o graphics.so init.o base.o graphics.o par.o plot.o plot3d.o stem.o -L../../../../lib -lR -dylib_file libRblas.dylib:../../../../lib/libRblas.dylib -Wl,-framework -Wl,CoreFoundation
make[9]: `Makedeps' is up to date.
mkdir ../../../../library/graphics/libs
byte-compiling package 'graphics'
Error in dyn.load(file, DLLpath = DLLpath, ...) :
  unable to load shared object '/usr/local/src/sage/local/var/tmp/sage/build/r-3.0.2.p1/src/library/grDevices/libs/grDevices.so':
  dlopen(/usr/local/src/sage/local/var/tmp/sage/build/r-3.0.2.p1/src/library/grDevices/libs/grDevices.so, 6): Symbol not found: _sqlite3_intarray_bind
  Referenced from: /System/Library/Frameworks/CoreData.framework/Versions/A/CoreData
  Expected in: /usr/local/src/sage/local/lib/libsqlite3.dylib
 in /System/Library/Frameworks/CoreData.framework/Versions/A/CoreData
Calls: <Anonymous> ... namespaceImport -> loadNamespace -> library.dynam -> dyn.load
Execution halted
make[7]: *** [../../../library/graphics/R/graphics.rdb] Error 1
make[6]: *** [all] Error 2
make[5]: *** [R] Error 1
make[4]: *** [R] Error 1
make[3]: *** [R] Error 1
Error building R.

real 224.327    user 159.984    sys 29.426    pcpu 84.43

************************************************************************
Error installing package r-3.0.2.p1
************************************************************************
Please email sage-devel (http://groups.google.com/group/sage-devel)
explaining the problem and including the relevant part of the log file
  /usr/local/src/sage/logs/pkgs/r-3.0.2.p1.log
Describe your computer, operating system, etc.
If you want to try to fix the problem yourself, *don't* just cd to
/usr/local/src/sage/local/var/tmp/sage/build/r-3.0.2.p1 and type 'make' or whatever is appropriate.
Instead, the following commands setup all environment variables
correctly and load a subshell for you to debug the error:
  (cd '/usr/local/src/sage/local/var/tmp/sage/build/r-3.0.2.p1' && '/usr/local/src/sage/sage' --sh)
When you are done debugging, you can type "exit" to leave the subshell.
************************************************************************
make[2]: *** [/usr/local/src/sage/local/var/lib/sage/installed/r-3.0.2.p1] Error 1
make[1]: *** [all] Error 2

real 232.272    user 164.476    sys 30.688    pcpu 84.02

***************************************************************
Error building Sage.

The following package(s) may have failed to build:

package: finite_word.py
log file: /usr/local/src/sage/logs/pkgs/finite_word.py.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/finite_word.py

package: mpir-2.6.0.p3
log file: /usr/local/src/sage/logs/pkgs/mpir-2.6.0.p3.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/mpir-2.6.0.p3

package: r-3.0.2.p1
log file: /usr/local/src/sage/logs/pkgs/r-3.0.2.p1.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/r-3.0.2.p1

The build directory may contain configuration files and other potentially
helpful information. WARNING: if you now run 'make' again, the build
directory will, by default, be deleted. Set the environment variable
SAGE_KEEP_BUILT_SPKGS to 'yes' to prevent this.

make: *** [build] Error 1

Here's the crash messages from the version I compiled with `make -k`:

6.2.rc0-develop-sage: ./sage
┌────────────────────────────────────────────────────────────────────┐
│ Sage Version 6.2.rc0, Release Date: 2014-04-22                     │
│ Type "notebook()" for the browser-based notebook interface.        │
│ Type "help()" for help.                                            │
└────────────────────────────────────────────────────────────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ Warning: this is a prerelease version, and it may be unstable.     ┃
┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛
-------------------------------------------------------------------------------
Resolving lazy import FacadeSets during startup
Calling stack:
  File "/usr/local/src/sage/local/lib/python2.7/site-packages/IPython/core/application.py", line 207, in excepthook
    return self.crash_handler(etype, evalue, tb)
  File "/usr/local/src/sage/local/lib/python2.7/site-packages/IPython/core/crashhandler.py", line 159, in __call__
    traceback = TBhandler.text(etype,evalue,etb,context=31)
  File "/usr/local/src/sage/local/lib/python2.7/site-packages/IPython/core/ultratb.py", line 407, in text
    tb_offset, context)
  File "/usr/local/src/sage/local/lib/python2.7/site-packages/IPython/core/ultratb.py", line 815, in structured_traceback
    locals,formatvalue=var_repr))
  File "/usr/local/src/sage/local/lib/python/inspect.py", line 887, in formatargvalues
    specs.append(strseq(args[i], convert, join))
  File "/usr/local/src/sage/local/lib/python/inspect.py", line 842, in strseq
    return convert(object)
  File "/usr/local/src/sage/local/lib/python/inspect.py", line 884, in convert
    return formatarg(name) + formatvalue(locals[name])
  File "/usr/local/src/sage/local/lib/python2.7/site-packages/IPython/core/ultratb.py", line 724, in eqrepr
    def eqrepr(value, repr=text_repr): return '=%s' % repr(value)
  File "/usr/local/src/sage/local/lib/python2.7/site-packages/IPython/core/ultratb.py", line 701, in text_repr
    return pydoc.text.repr(value)
  File "/usr/local/src/sage/local/lib/python/repr.py", line 24, in repr
    return self.repr1(x, self.maxlevel)
  File "/usr/local/src/sage/local/lib/python/pydoc.py", line 970, in repr1
    return cram(stripid(repr(x)), self.maxother)
  File "/usr/local/src/sage/local/lib/python2.7/site-packages/sage/categories/category.py", line 2293, in _repr_
    if len(categories) == 2 and Sets().Facades() in categories:
-------------------------------------------------------------------------------

**********************************************************************

Oops, Sage crashed. We do our best to make it stable, but...

A crash report was automatically generated with the following information:
  - A verbatim copy of the crash traceback.
  - A copy of your input history during this session.
  - Data on your current Sage configuration.

It was left in the file named:
    '/Users/andrew/.sage/ipython-1.2.1/Sage_crash_report.txt'
If you can email this file to the developers, the information in it will help
them in understanding and correcting the problem.

You can mail it to: sage-support at sage-support@googlegroups.com
with the subject 'Sage Crash Report'.

If you want to do it now, the following command will work (under Unix):
mail -s 'Sage Crash Report' sage-support@googlegroups.com < /Users/andrew/.sage/ipython-1.2.1/Sage_crash_report.txt

To ensure accurate tracking of this issue, please file a report about it at:
http://trac.sagemath.org

Hit <Enter> to quit (your terminal may close):

[Andrew]

Update:

It's not a 6.2.rc0 issue after all. My compile of 6.1.1 failed with:

...
sh: line 1: 45940 Trace/BPT trap: 5       /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild -sdk / -find strip 2> /dev/null
xcrun: error: unable to find utility "strip", not a developer tool or in PATH
install: child process failed: xcrun strip - /usr/local/src/sage/local/bin/solve_IP
make[4]: *** [install] Error 70
make[3]: *** [install-nolns] Error 1
Unable to build and install Singular
Error building Singular (error in build_singular).

real 436.530    user 370.404    sys 40.987    pcpu 94.24

************************************************************************
Error installing package singular-3.1.5.p9

************************************************************************
Please email sage-devel (http://groups.google.com/group/sage-devel)
explaining the problem and

including the relevant part of the log file
  /usr/local/src/sage/logs/pkgs/singular-3.1.5.p9.log

Describe your computer, operating system, etc.
If you want to try to fix the problem yourself, *don

't* just cd to
/usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9 and type 'make' or whatever is appropriate.

Instead, the following commands setup all environment variables
correctly and load a subshell for you to debug the error:

  (cd '/usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9' && '/usr/local/src/sage/sage' --sh)

When you are done debugging, you can type "exit" to leave the subshell.
************************************************************************

make[2]: *** [/usr/local/src/sage/local/var/lib/sage/installed/singular-3.1.5.p9] Error 1

make[1]: *** [all] Error 2

real 3662.925    user 2655.310    sys 428.853    pcpu 84.20

***************************************************************
Error building Sage.

The following package(s) may have failed to build:

package: finite_word.py
log file: /usr/local/src/sage/logs/pkgs/finite_word.py.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/finite_word.py

package: mpir-2.6.0.p3
log file: /usr/local/src/sage/logs/pkgs/mpir-2.6.0.p3.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/mpir-2.6.0.p3

package: r-3.0.2.p1
log file: /usr/local/src/sage/logs/pkgs/r-3.0.2.p1.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/r-3.0.2.p1

package: singular-3.1.5.p9
log file: /usr/local/src/sage/logs/pkgs/singular-3.1.5.p9.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9

The build directory may contain configuration files and other potentially
helpful information. WARNING: if you now run '

make' again, the build

directory will, by default, be deleted. Set the environment variable
SAGE_KEEP_BUILT_SPKGS to '

yes' to prevent this.

[Michael Welsh]

On 28/04/2014, at 1809, Andrew <andrew...@gmail.com> wrote:
>
> sh: line 1: 45940 Trace/BPT trap: 5 /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild -sdk / -find strip 2> /dev/null
> xcrun: error: unable to find utility "strip", not a developer tool or in PATH
> install: child process failed: xcrun strip - /usr/local/src/sage/local/bin/solve_IP

You should have strip in /usr/bin. Have you tried (re)installing the developer tools?

[Andrew]

Yes I have strip and the command line tools:

533-6.1.1-master-sage:  which strip
strip is /Library/Developer/CommandLineTools/usr/bin/strip
strip is /usr/bin/strip

[Andrew]

Here's another bizarre feature:

538-6.1.1-master-sage: /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild -sdk / -find strip 2> /dev/null
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip\539-6.1.1-master-sage: ls -l /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip
-rwxr-xr-x  1 root  wheel   187K 27 Mar 11:17 /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip*

That is, xcodebuild actually find strip but still reports an error.
A.

[Volker Braun]

Whats xcodebuild -version

[Andrew]

On Monday, 28 April 2014 20:09:53 UTC+10, Volker Braun wrote:

Whats xcodebuild -version

It's:

Xcode 5.1.1
Build version 5B1008

Andrew

[Volker Braun]

On Monday, April 28, 2014 10:53:53 AM UTC+1, Andrew wrote:

Here's another bizarre feature:

538-6.1.1-master-sage: /Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild -sdk / -find strip 2> /dev/null
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip\539-6.1.1-master-sage: ls -l /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip
-rwxr-xr-x  1 root  wheel   187K 27 Mar 11:17 /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip*

Wtf is  "/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip\539-6.1.1-master-sage: ls -l". Its first in the dirent and shadows the real strip.

[Andrew]

Sorry, it's a bad line break: my prompt is:

 <stuff deleted>-<history number>-<git branch>-<working directory>+":"

The xcodebuild call is the first reported error in my makefile log and `ls -l` was to verify that `find`, as found by `xcodebuild`, actually exists.

It's bizarre because the error reported in the make log file is not an error when exactly the same command is run by hand.

Andrew

[Volker Braun]

whats the output of "xcrun -v strip"

[Andrew]

On Monday, 28 April 2014 23:49:47 UTC+10, Volker Braun wrote:

whats the output of "xcrun -v strip"

xcrun: note: PATH = '/Users/andrew/bin:/Library/Developer/CommandLineTools/usr/bin:/bin:/sbin:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/usr/texbin:/usr/local/bin'
xcrun: note: SDKROOT = '/'
xcrun: note: TOOLCHAINS = ''
xcrun: note: DEVELOPER_DIR = '/Applications/Xcode.app/Contents/Developer'
xcrun: note: XCODE_DEVELOPER_USR_PATH = ''
xcrun: note: xcrun_db = '/var/folders/pz/2_c5gpmx3q90fwldvvcfc_x0000325/T/xcrun_db'
xcrun: note: xcrun via strip (xcrun)
xcrun: note: database key is: strip|/||/Applications/Xcode.app/Contents/Developer|
xcrun: note: looking up with '/Applications/Xcode.app/Contents/Developer/usr/bin/xcodebuild -sdk / -find strip 2> /dev/null'
xcrun: note: lookup resolved with 'xcodebuild -find' to '/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip'
fatal error: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/strip: no files specified

[Volker Braun]

So it works. Your real error is elsewhere. Run "make distclean && make -j1"

[Andrew]

On Tuesday, 29 April 2014 00:02:12 UTC+10, Volker Braun wrote:

So it works. Your real error is elsewhere. Run "make distclean && make -j1"

Pretty much as above, but log file attached.

Thanks, btw, for looking at this.
Andrew

[Volker Braun]

ld: symbol dyld_stub_binding_helper not found, normally in crt1.o/dylib1.o/bundle1.o for architecture x86_64

fatal error: libtool: internal link edit command failed

sounds like your gcc install is hosed. Post the full log, not just the singular log.

[Volker Braun]

Does your compile always fail in different places with different errors? Maybe its more relevant to look at the output of "diskutil info /" and/or install smartmontools...

[kcrisman]

  File "/usr/local/src/sage/local/lib/python2.7/site-packages/sage/categories/category.py", line 2293, in _repr_
    if len(categories) == 2 and Sets().Facades() in categories:
-------------------------------------------------------------------------------

**********************************************************************

Oops, Sage crashed. We do our best to make it stable, but...

A crash report was automatically generated with the following information:
  - A verbatim copy of the crash traceback.
  - A copy of your input history during this session.
  - Data on your current Sage configuration.

It was left in the file named:
    '/Users/andrew/.sage/ipython-1.2.1/Sage_crash_report.txt'
If you can email this file to the developers, the information in it will help
them in understanding and correcting the problem.

You can mail it to: sage-support at sage-support@googlegroups.com
with the subject 'Sage Crash Report'.

If you want to do it now, the following command will work (under Unix):
mail -s 'Sage Crash Report' sage-support@googlegroups.com < /Users/andrew/.sage/ipython-1.2.1/Sage_crash_report.txt

To ensure accurate tracking of this issue, please file a report about it at:
http://trac.sagemath.org

Hit <Enter> to quit (your terminal may close):

This is exactly what I get when switching between the latest develop and a branch based on beta6 or so.   The actual error is that the wrong version of libpng is linked in libm4ri, and what is even weirder is that the version of libm4ri isn't even the right current one! 

ImportError: dlopen(/Users/.../sage/local/lib/python2.7/site-packages/sage/matrix/matrix_mod2_dense.so, 2): Library not loaded: /Users$

  Referenced from: /Users/.../sage/local/lib/libm4ri-0.0.20121224.dylib

  Reason: Incompatible library version: libm4ri-0.0.20121224.dylib requires version 52.0.0 or later, but libpng12.0.dylib provides version 36.0.0

So currently my install is completely destroyed, and there is no point in trying again because I'm sure that trying to switch back and forth among branches with different upstream will cause this again.  

And why do I have both .tar.gz and .tar.bz2 files in upstream?  Is that right?  Especially of the same thing?  (The fact that libpng is one of them is particularly suspicious.)

$ ls upstream/

Cython-0.20.1.tar.gz gsl-1.15.tar.bz2 pil-1.1.6.tar.bz2

Pillow-2.2.2.tar.gz iconv-1.14.tar.bz2 pkgconf-0.9.4.tar.bz2

atlas-3.10.1.20140128.tar.bz2 iml-1.0.3.tar.bz2 pkgconfig-1.1.0.tar.gz

atlas-3.10.1.20140210.tar.bz2 ipython-0.13.2.tar.bz2 polybori-0.8.3.tar.bz2

atlas-3.10.1.tar.bz2 ipython-1.2.1.tar.gz polytopes_db-20120220.tar.bz2

boehm_gc-7.2d.tar.bz2 jinja2-2.5.5.tar.bz2 ppl-1.1.tar.bz2

boost_cropped-1.52.0.tar.bz2 jmol-12.3.27.tar.bz2 ppl-1.1pre9.tar.bz2

bzip2-1.0.6.20140317.tar.gz jmol-14.0.13.tar.gz pycrypto-2.1.0.tar.bz2

bzip2-1.0.6.tar.gz lcalc-1.23.tar.bz2 pycrypto-2.6.1.tar.gz

cddlib-094f.tar.bz2 libfplll-4.0.4.tar.bz2 pygments-1.3.1.tar.bz2

cddlib-094g.tar.bz2 libgap-4.6.4.tar.bz2 pynac-0.3.0.tar.bz2

cephes-2.8.tar.bz2 libgap-4.7.4.tar.gz pynac-0.3.1.tar.bz2

cliquer-1.21.tar.bz2 libm4ri-20130416.tar.bz2 pynac-0.3.2.tar.bz2

conway_polynomials-0.4.tar.bz2 libm4rie-20130416.tar.bz2 pyparsing-2.0.1.tar.gz

cvxopt-1.1.6.tar.bz2 libpng-1.2.35.tar.bz2 python-2.7.5.tar.bz2

cython-0.19.1.tar.bz2 libpng-1.2.51.tar.gz r-3.0.2.tar.bz2

dateutil-2.2.tar.gz linbox-1.3.2.tar.bz2 ratpoints-2.1.3.tar.bz2

docutils-0.7.tar.bz2 lrcalc-1.1.6.tar.bz2 readline-6.2.tar.bz2

ecl-12.12.1.tar.bz2 matplotlib-1.2.1.tar.bz2 rpy2-2.0.8.tar.bz2

eclib-20120830.tar.bz2 matplotlib-1.3.1.tar.gz rpy2-2.3.8.tar.gz

eclib-20140128.tar.bz2 maxima-5.29.1.tar.bz2 rubiks-20070912.tar.bz2

ecm-6.4.4.tar.bz2 mpc-1.0.2.tar.gz sagenb-0.10.7.2.tar.bz2

elliptic_curves-0.7.tar.bz2 mpc-1.0.tar.bz2 sagenb-0.10.8.2.tar

fflas_ffpack-1.6.0.tar.bz2 mpfi-1.5.1.tar.bz2 sagetex-2.3.4.tar.bz2

flint-2.3.tar.bz2 mpfr-3.1.2.tar.bz2 scipy-0.12.0.tar.bz2

flint-2.4.1.tar.gz mpir-2.6.0.tar.bz2 scons-1.2.0.tar.bz2

flint-2.4.2.tar.gz mpmath-0.17.tar.bz2 setuptools-0.6.16.tar.bz2

flint-2.4.3.tar.gz mpmath-0.18.tar.gz setuptools-2.0.1.tar.gz

flintqs-20070817.tar.bz2 ncurses-5.9.20131221.tar.bz2 singular-3.1.5.tar.bz2

freetype-2.3.5.tar.bz2 ncurses-5.9.tar.bz2 singular-3.1.6.tar.bz2

freetype-2.5.2.tar.bz2 networkx-1.6.tar.bz2 six-1.4.1.tar.gz

gap-4.6.4.tar.bz2 networkx-1.8.1.tar.gz sphinx-1.1.2.tar.bz2

gap-4.7.4.tar.bz2 ntl-5.5.2.tar.bz2 sqlalchemy-0.5.8.tar.bz2

gcc-4.7.3.tar.bz2 ntl-6.0.0.tar.bz2 sqlite-3.7.17.tar.bz2

gd-2.0.35.tar.bz2 ntl-6.1.0.tar.bz2 sqlite-3.8.4.3.tar.bz2

gdmodule-0.56.tar.bz2 numpy-1.7.0.tar.bz2 symmetrica-2.0.tar.bz2

genus2reduction-0.3.tar.bz2 palp-2.1.tar.bz2 sympow-1.018.1.tar.bz2

gf2x-1.1.tar.bz2 pari-2.5.5.tar.bz2 sympy-0.7.3.tar.bz2

gfan-0.5.tar.bz2 pari-2.5.5b.tar.bz2 sympy-0.7.4.tar.gz

git-1.8.4.4.tar.bz2 pari_galdata-20080411.tar.bz2 tachyon-0.98.9.tar.bz2

givaro-3.7.1.tar.bz2 pari_seadata_small-20090618.tar.bz2 tornado-3.1.1.tar.gz

glpk-4.44.tar.bz2 patch-2.5.9.tar.bz2 zlib-1.2.8.tar.bz2

graphs-20120404.tar.bz2 pexpect-2.0.tar.bz2 zn_poly-0.9.tar.bz2

[Volker Braun]

libpng was upgraded recently, and if you switch between different versions then you have to rebuild (in the worst case: make distclean && make will always work).

[kcrisman]

On Tuesday, April 29, 2014 9:44:54 AM UTC-4, Volker Braun wrote:

libpng was upgraded recently, and if you switch between different versions then you have to rebuild (in the worst case: make distclean && make will always work).

I figured, but that does take some time.  And more to the point:

$ ./sage -f libpng

Found local metadata for libpng-1.2.35.p6

...

And "Reason: Incompatible library version: libm4ri-0.0.20121224.dylib" makes no sense; wasn't this updated to 20130416 in Sage 5.10???  I don't even have that version listed in upstream/.

$ ./sage -version

Sage Version 6.2.beta6, Release Date: 2014-04-02

By the way, it's really annoying that the "Merged in:" field is no longer used.  It's much easier to do a quick search for "sage update package_name" and click on the ticket than sift through sage-release to find where it might have happened.

[Volker Braun]

The issue is that libpng was updated, not libm4ri

There is still something wonky because the two libpng versions ought to be compatible. There is something wrong in how m4ri links to libpng on OSX, causing it to depend on the version patchlevel.

On Tuesday, April 29, 2014 4:47:37 PM UTC+1, kcrisman wrote:

By the way, it's really annoying that the "Merged in:" field is no longer used.

Not needed: 

$ git log --oneline  build/pkgs/libpng 

b42c581 Update libpng to version 1.2.51.

f570a62 Fixed patching of libpng.

de6e2ee Patching libpng to allow compilation of libfreetype-2.5.2

19382e1 Add missing tarballs, use VERSION in tarball path.

8a79b12 Store actual tarball name in package metadata.

...

$ git trac find b42c581

Commit has been merged by the release manager into your current branch.

commit e2ff979af93ea5f118e46621768efd923deaf2e5

Merge: bf56254 b42c581

Author: Release Manager <rel...@sagemath.org>

Date:   Mon Apr 14 00:00:20 2014 +0200

    Trac #16099: Update libpng to 1.2.51

    

    Straightforward update.

    Works on Cygwin64.

    

    Use tarball at:

    * http://boxen.math.washington.edu/home/jpflori/upstream/libpng-1.2.51.t

    ar.gz

    

    URL: http://trac.sagemath.org/16099

    Reported by: jpflori

    Ticket author(s): Jean-Pierre Flori

    Reviewer(s): Volker Braun

[leif]

Volker Braun wrote:
> The issue is that libpng was updated, not libm4ri
>
> There is still something wonky because the two libpng versions ought to
> be compatible. There is something wrong in how m4ri links to libpng on
> OSX, causing it to depend on the version patchlevel.
>
> On Tuesday, April 29, 2014 4:47:37 PM UTC+1, kcrisman wrote:
>
> By the way, it's really annoying that the "Merged in:" field is no
> longer used.
>
>
> Not needed:
>
> $ git log --oneline build/pkgs/libpng
> b42c581 Update libpng to version 1.2.51.
> f570a62 Fixed patching of libpng.
> de6e2ee Patching libpng to allow compilation of libfreetype-2.5.2
> 19382e1 Add missing tarballs, use VERSION in tarball path.
> 8a79b12 Store actual tarball name in package metadata.
> ...

Or look at:

http://git.sagemath.org/sage.git/log/build/pkgs/libpng?h=develop&id=9d7c869f4040ede16f1c8f2079508a78d9dfa247

(Navigating in the tree to build/pkgs/, then clicking "log" next to
"libpng".)

But while from there you can figure out which release the branch /
commit was /based on/, you can't immediately tell into which release it
finally got merged.

> $ git trac find b42c581
> Commit has been merged by the release manager into your current branch.
> commit e2ff979af93ea5f118e46621768efd923deaf2e5
> Merge: bf56254 b42c581
> Author: Release Manager <rel...@sagemath.org>
> Date: Mon Apr 14 00:00:20 2014 +0200
>
> Trac #16099: Update libpng to 1.2.51
> Straightforward update.
> Works on Cygwin64.
> Use tarball at:
> *
> http://boxen.math.washington.edu/home/jpflori/upstream/libpng-1.2.51.t
> ar.gz
> URL: http://trac.sagemath.org/16099
> Reported by: jpflori
> Ticket author(s): Jean-Pierre Flori
> Reviewer(s): Volker Braun
>

But that's certainly not the same as having the (IMHO useful) "Merged:"
field filled.

Similar for the changelog entries in SPKG.txt files; they provide[d]
different (namely usually higher-level), and easier to access
information than commit logs (and diffs) do.


-leif

--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail

[kcrisman]

On Tuesday, April 29, 2014 1:44:18 PM UTC-4, leif wrote:

Volker Braun wrote:
> The issue is that libpng was updated, not libm4ri
>
> There is still something wonky because the two libpng versions ought to
> be compatible. There is something wrong in how m4ri links to libpng on
> OSX, causing it to depend on the version patchlevel.
>

Well, starting from scratch did fix it, as always...

 

But that's certainly not the same as having the (IMHO useful) "Merged:"
field filled.

Similar for the changelog entries in SPKG.txt files; they provide[d]
different (namely usually higher-level), and easier to access
information than commit logs (and diffs) do.

EXACTLY.   Your syntax doesn't actually tell me what I want to know, which is the specific beta.  

"Commit has been merged by the release manager into your current branch."

is not helpful.  As Leif says, "you can't immediately tell into which release it finally got merged."

[Volker Braun]

Ok, I changed it to:

$ git trac find b42c581

Commit has been merged in 6.2.beta8.       <--------------------- new addition

commit e2ff979af93ea5f118e46621768efd923deaf2e5

Merge: bf56254 b42c581

Author: Release Manager <rel...@sagemath.org>

Date:   Mon Apr 14 00:00:20 2014 +0200

    Trac #16099: Update libpng to 1.2.51

    

    Straightforward update.

    Works on Cygwin64.

    

    Use tarball at:

    * http://boxen.math.washington.edu/home/jpflori/upstream/libpng-1.2.51.t

    ar.gz

    

    URL: http://trac.sagemath.org/16099

    Reported by: jpflori

    Ticket author(s): Jean-Pierre Flori

    Reviewer(s): Volker Braun

[Andrew]

On Tuesday, 29 April 2014 01:06:35 UTC+10, Volker Braun wrote:

Does your compile always fail in different places with different errors? Maybe its more relevant to look at the output of "diskutil info /" and/or install smartmontools...

Sorry, I am still having issues. I tried re-installing the xcode command line toolsbut this didn't help. Is it possible some of the code I installed with brew is causing a conflict?

As Volker pointed out the log I attached was incomplete. This is because `make distclean` ends with an error (so the second make in `make distclean && make` never happened).

Regarding the errors, they are not always the same. I've had `xcrun` not finding strip and clang. On the other hand, I  have been trying to fix the issue in various ways, so the state of my system when I run make is not stable either...

The packages that are failing to build are always:

• finite_word.py
• mpir-2.6.0.p3
• r-3.0.0.2.pl
• singular-3.1.5.p9

Finally, you asked about diskutil and smartmontools. The output from these are:

smartctl 6.2 2013-07-26 r3841 [x86_64-apple-darwin13.1.0] (local build)
Copyright (C) 2002-13, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF READ SMART DATA SECTION ===
SMART Self-test log structure revision number 1
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed without error       00%     14811         -
# 2  Short offline       Completed without error       00%     14808         -

and

sage: diskutil info /
   Device Identifier:        disk0s2
   Device Node:              /dev/disk0s2
   Part of Whole:            disk0
   Device / Media Name:      Customer

   Volume Name:              Mac
   Escaped with Unicode:     Mac

   Mounted:                  Yes
   Mount Point:              /
   Escaped with Unicode:     /

   File System Personality:  Journaled HFS+
   Type (Bundle):            hfs
   Name (User Visible):      Mac OS Extended (Journaled)
   Journal:                  Journal size 40960 KB at offset 0x1638e000
   Owners:                   Enabled

   Partition Type:           Apple_HFS
   OS Can Be Installed:      Yes
   Recovery Disk:            disk0s3
   Media Type:               Generic
   Protocol:                 SATA
   SMART Status:             Verified
   Volume UUID:              5FD5CA9F-2845-3735-9024-426A671CE311

   Total Size:               499.2 GB (499248099328 Bytes) (exactly 975093944 512-Byte-Units)
   Volume Free Space:        60.4 GB (60374343680 Bytes) (exactly 117918640 512-Byte-Units)
   Device Block Size:        512 Bytes

   Read-Only Media:          No
   Read-Only Volume:         No
   Ejectable:                No

   Whole:                    No
   Internal:                 Yes
   Solid State:              No

[Volker Braun]

Homebrew/Macports/Fink/etc are not supported and likely to conflict with Sage. What kind of error do you get from "make distclean"? You can try to exclude non-system stuff from the PATH to ignore 3rd party software:

export PATH=/usr/bin:/bin:/usr/sbin:/sbin

make distclean

make

[Andrew]

On Wednesday, 30 April 2014 18:21:13 UTC+10, Volker Braun wrote:

Homebrew/Macports/Fink/etc are not supported and likely to conflict with Sage. What kind of error do you get from "make distclean"? You can try to exclude non-system stuff from the PATH to ignore 3rd party software:

export PATH=/usr/bin:/bin:/usr/sbin:/sbin

make distclean

Thanks. This was clearly masking the problem. Changing path and `make distclean` show that your previous guess about my gcc being "hosed" is right. The log from `make distclean` is below but the important bit is probably:

checking for gcc... gcc
checking whether the C compiler (gcc -O2 -g  -fPIC ) works... no
configure: error: installation or configuration problem: C compiler cannot create executables.

Not sure why this is. I tried compiling some old c-code of mine and they all failed (most complaining about not finding malloc.h). The gcc is really clang:

sage: gcc --version
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Apple LLVM version 5.1 (clang-503.0.40) (based on LLVM 3.4svn)
Target: x86_64-apple-darwin13.1.0
Thread model: posix

Andrew

sage: make distclean
cd build && \
    "../build/pipestatus" \
        "env SAGE_PARALLEL_SPKG_BUILD='' ./install all 2>&1" \
        "tee -a ../logs/install.log"
/usr/local/src/sage/build
*** ALL ENVIRONMENT VARIABLES BEFORE BUILD: ***

Apple_PubSub_Socket_Render=/tmp/launch-ambPxc/Render
BIBINPUTS=.:/Users/andrew/Papers:/Users/andrew/Grants:
BSTINPUTS=/Users/andrew/TeX/BibTeX:
CDPATH=.:..:~:~/Papers:~/Books:~/Preprints:~/TeX:~/Gap:~:~/Grants:~/Grants/Current:~/Courses:~/Documents:~/Documents/Home:~/Documents/AMAE:~/Documents/Home/Kiddies:~/Courses:~/Admin:~/Admin/Promotion/:~/Sage/:../..:~/Sites
DISPLAY=/tmp/launch-8ONZAE/org.macosforge.xquartz:0
HISTIGNORE=&:bg:fg:ll:h
HISTTIMEFORMAT=%H:%M >
HOME=/Users/andrew
HOSTFILE=/Users/andrew/.hosts
INPUTRC=/Users/andrew/.inputrc
ITERM_PROFILE=Default
ITERM_SESSION_ID=w6t1p0
LANG=en_AU.UTF-8
LOGNAME=andrew
MAKE=make
MAKEFLAGS=
MAKELEVEL=1
MFLAGS=
PATH=/usr/local/src/sage/src/bin:/usr/local/src/sage/local/bin:/usr/bin:/bin:/usr/sbin:/sbin
PWD=/usr/local/src/sage/build
PYTHONPATH=/usr/local/src/sage/local
SAGE_EXTCODE=/usr/local/src/sage/local/share/sage/ext
SAGE_LOCAL=/usr/local/src/sage/local
SAGE_LOGS=/usr/local/src/sage/logs/pkgs
SAGE_PARALLEL_SPKG_BUILD=
SAGE_ROOT=/usr/local/src/sage
SAGE_SPKG_INST=/usr/local/src/sage/local/var/lib/sage/installed
SAGE_SRC=/usr/local/src/sage/src
SAGE_VERSION=6.1.1
SECURITYSESSIONID=186a4
SHELL=/bin/bash
SHLVL=4
SSH_AUTH_SOCK=/tmp/launch-wOEVpB/Listeners
TERM=xterm-256color
TERM_PROGRAM=iTerm.app
TEXEDIT=/usr/bin/gvim +%d %s
TEXINPUTS=.:~/TeX/inputs//:!!/usr/local/src/sage/local/share/texmf//:
TIMEFORMAT=
TMPDIR=/var/folders/pz/2_c5gpmx3q90fwldvvcfc_x0000325/T/
USER=andrew
_=/usr/bin/env
__CF_USER_TEXT_ENCODING=0xC45:0:15
__CHECKFIX1436934=1
real %3R    user %3U    sys %3S    pcpu %P
***********************************************
make base
make[2]: Nothing to be done for `base'.
env SAGE_BUILD_TOOLCHAIN=yes make toolchain
make[2]: Nothing to be done for `toolchain'.
make toolchain-deps
make /usr/local/src/sage/local/var/lib/sage/installed/zlib-1.2.8.p0
make[3]: `/usr/local/src/sage/local/var/lib/sage/installed/zlib-1.2.8.p0' is up to date.
make /usr/local/src/sage/local/var/lib/sage/installed/mpir-2.6.0.p4
make[3]: `/usr/local/src/sage/local/var/lib/sage/installed/mpir-2.6.0.p4' is up to date.
make /usr/local/src/sage/local/var/lib/sage/installed/mpfr-3.1.2.p0
make[3]: `/usr/local/src/sage/local/var/lib/sage/installed/mpfr-3.1.2.p0' is up to date.
make /usr/local/src/sage/local/var/lib/sage/installed/mpc-1.0.p0
make[3]: `/usr/local/src/sage/local/var/lib/sage/installed/mpc-1.0.p0' is up to date.
make /usr/local/src/sage/local/var/lib/sage/installed/ppl-1.1pre9.p0
make[3]: `/usr/local/src/sage/local/var/lib/sage/installed/ppl-1.1pre9.p0' is up to date.
make all-sage
/usr/local/src/sage/build/pipestatus "sage-spkg ${SAGE_SPKG_OPTS} singular-3.1.5.p9 2>&1" "tee -a /usr/local/src/sage/logs/pkgs/singular-3.1.5.p9.log"
Found local metadata for singular-3.1.5.p9
Found local sources at /usr/local/src/sage/upstream/singular-3.1.5.tar.bz2
Checksum: b4834021a4b44f43806cca2561ec942dc2185c17 vs b4834021a4b44f43806cca2561ec942dc2185c17
singular-3.1.5.p9
====================================================
Setting up build directory for singular-3.1.5.p9
/usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9
Finished set up
/usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9
****************************************************
Host system:
Darwin pmathas.pc.maths.usyd.edu.au 13.1.0 Darwin Kernel Version 13.1.0: Wed Apr  2 23:52:02 PDT 2014; root:xnu-2422.92.1~2/RELEASE_X86_64 x86_64
****************************************************
C compiler: gcc
C compiler version:
Using built-in specs.
COLLECT_GCC=/usr/local/src/sage/local/bin/gcc
COLLECT_LTO_WRAPPER=/usr/local/src/sage/local/libexec/gcc/x86_64-apple-darwin13.1.0/4.7.3/lto-wrapper
Target: x86_64-apple-darwin13.1.0
Configured with: ../src/configure --prefix=/usr/local/src/sage/local --with-local-prefix=/usr/local/src/sage/local --with-gmp=/usr/local/src/sage/local --with-mpfr=/usr/local/src/sage/local --with-mpc=/usr/local/src/sage/local --with-system-zlib --disable-multilib --disable-nls  
Thread model: posix
gcc version 4.7.3 (GCC)
****************************************************
### Singular spkg-install: choose_patches ###
### Singular spkg-install: apply_patches ###
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/Minor.h.patch
patching file Singular/Minor.h
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/NTL_negate.patch
patching file factory/cf_chinese.cc
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/assert.patch
patching file factory/assert.h
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/cygwin-makefile.patch
patching file Singular/Makefile.in
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/install_table.patch
patching file Singular/Makefile.in
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/osx_link.patch
patching file Singular/Makefile.in
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/pullrequest215.patch
patching file factory/facFqBivar.cc
patching file factory/facFqBivarUtil.cc
patching file factory/facMul.cc
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/sage_trac_12089.patch
patching file Singular/Makefile.in
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/sage_trac_14295.patch
patching file kernel/bigintmat.cc
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/sanitize_gmp_header_hack.patch
patching file factory/gen_cf_gmp.cc
patching file factory/gen_cf_gmp.template
patching file factory/GNUmakefile.in
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/sing_win.patch
patching file Singular/sing_win.cc
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/singular_15435.patch
patching file kernel/gring.cc
patching file kernel/ring.cc
patching file kernel/ring.h
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/singular_part_of_changeset_baadc0f7.patch
patching file factory/facFactorize.cc
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/singular_trac_439.patch
patching file factory/facBivar.h
patching file factory/facFactorize.h
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/singular_trac_440.patch
patching file factory/facFqBivar.cc
Hunk #1 succeeded at 5944 (offset 16 lines).
patching file factory/cf_util.cc
patching file factory/cf_util.h
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/singular_trac_441.patch
patching file factory/facFqBivar.cc
Hunk #1 succeeded at 6118 (offset 16 lines).
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/singular_trac_443.patch
patching file Singular/configure
patching file Singular/configure.in
Applying /usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9/patches/slibdir.patch
patching file IntegerProgramming/Makefile.in
patching file Singular/Makefile.in
### Singular spkg-install: remove_old_version ###
### Singular spkg-install: config ###
creating cache ./config.cache
checking uname for singular... x86_64Mac-darwin
checking for gcc... gcc
checking whether the C compiler (gcc -O2 -g  -fPIC ) works... no
configure: error: installation or configuration problem: C compiler cannot create executables.
Unable to configure Singular.
Error building Singular (error in config).

real 1.110    user 0.279    sys 0.342    pcpu 55.89

************************************************************************
Error installing package singular-3.1.5.p9
************************************************************************
Please email sage-devel (http://groups.google.com/group/sage-devel)
explaining the problem and including the relevant part of the log file
  /usr/local/src/sage/logs/pkgs/singular-3.1.5.p9.log
Describe your computer, operating system, etc.
If you want to try to fix the problem yourself, *don't* just cd to
/usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9 and type 'make' or whatever is appropriate.
Instead, the following commands setup all environment variables
correctly and load a subshell for you to debug the error:
  (cd '/usr/local/src/sage/local/var/tmp/sage/build/singular-3.1.5.p9' && '/usr/local/src/sage/sage' --sh)
When you are done debugging, you can type "exit" to leave the subshell.
************************************************************************
make[2]: *** [/usr/local/src/sage/local/var/lib/sage/installed/singular-3.1.5.p9] Error 1
make[1]: *** [all] Error 2

real 5.429    user 2.523    sys 0.980    pcpu 64.54

***************************************************************
Error building Sage.

The following package(s) may have failed to build:

package: finite_word.py
log file: /usr/local/src/sage/logs/pkgs/finite_word.py.log
build directory: /usr/local/src/sage/local/var/tmp/sage/build/finite_word.py

[Volker Braun]

On Wednesday, April 30, 2014 9:50:27 AM UTC+1, Andrew wrote:

singular-3.1.5.p9

Sage-6.2.rc0 uses  Singular 3.1.6.p1, you are building an old branch (possibly older than when we fixed the OSX 10.9 build). Are you sure you didn't accidentally check out 6.2.beta0? Check "git log"

[Andrew]

Strange. In $SAGE_ROOT/upstream I have both

singular-3.1.5.tar.bz2
singular-3.1.6.tar.bz2

I jut tried deleting 3.1.5 and remaking....and make downloaded 3.1.5 again giving the messages

...

/usr/local/src/sage/build/pipestatus "sage-spkg ${SAGE_SPKG_OPTS} singular-3.1.5.p9 2>&1" "tee -a /usr/local/src/sage/logs/pkgs/singular-3.1.5.p9.log"
Found local metadata for singular-3.1.5.p9

Attempting to download package singular-3.1.5.p9
>>> Trying to download http://www.sagemath.org/packages/upstream/singular/singular-3.1.5.tar.bz2
[............................................................]

Checksum: b4834021a4b44f43806cca2561ec942dc2185c17 vs b4834021a4b44f43806cca2561ec942dc2185c17
singular-3.1.5.p9

....

Does anyone know where the 'local metadata for singular 3.1.5.p9" can be found?

Possibly there is also something wrong my how I have configured git. From $SAGE_ROOT I'm using:

git remote -v
trac    g...@trac.sagemath.org:sage.git (fetch)
trac    g...@trac.sagemath.org:sage.git (push)

git status -b
On branch master
Your branch is up-to-date with 'trac/master'.

My $SAGE_ROOT/.git/config contains, among other things, the following:

[core]
    repositoryformatversion = 0
    filemode = true
    bare = false
    logallrefupdates = true
    ignorecase = true
    precomposeunicode = false
[remote "trac"]
    url = g...@trac.sagemath.org:sage.git
    fetch = +refs/heads/*:refs/remotes/trac/*
[branch "master"]
    remote = trac
    merge = refs/heads/master

Is there anything obviously wrong here?

Andrew

[Andrew]

OK, so the singular version seems to be in

$SAGE_ROOT/build/pkgs/singular/package-version.txt

and it is indeed set to 3.1.5.p9. Updating this by hand to 3.1.6.p1 and then re-making gave a checksum error on

upstream/singular-3.1.6.p1

Deleting this and then re-making still gives a checksum error so I'm guessing that the heck sum that it is being compared against is for 3.1.5.p6.

I'm guessing that I have somehow tied git up in knots?
Andrew

[Jean-Pierre Flori]

To get the latest betas you should be tracking the "develop" branch (from trac or github, they are the same).
You can issue something like:
"""
git checkout -b develop trac/develop
"""
to make a local copy of it and directly swith to it.

[Volker Braun]

Or just do a "git pull", since you are on the master branch this will give you 6.1.1 which afaik also builds on OSX 10.9.

[Andrew]

On Thursday, 1 May 2014 00:47:20 UTC+10, Volker Braun wrote:

Or just do a "git pull", since you are on the master branch this will give you 6.1.1 which afaik also builds on OSX 10.9.

As detailed above, I have the latest master branch but for some reason it has the wrong version of singular:

6.1.1-master-sage: git pull
remote: Counting objects: 496, done.
remote: Compressing objects: 100% (208/208), done.
remote: Total 303 (delta 232), reused 145 (delta 92)
Receiving objects: 100% (303/303), 84.66 KiB | 0 bytes/s, done.
Resolving deltas: 100% (232/232), completed with 70 local objects.
From trac.sagemath.org:sage
   2001f33..fc52070  develop    -> trac/develop
   ad1c60b..ab269f4  public/algebras/smash_product_algebra-15874 -> trac/public/algebras/smash_product_algebra-15874
 * [new branch]      u/cnassau/ticket/16271 -> trac/u/cnassau/ticket/16271
 * [new branch]      u/dkrenn/fsm/accepting_show_empty_final_word_out_1 -> trac/u/dkrenn/fsm/accepting_show_empty_final_word_out_1
   3c275de..b944d26  u/mshimo/combinat/double_affine_hecke_algebra -> trac/u/mshimo/combinat/double_affine_hecke_algebra
   62f0158..d743414  u/ncohen/15310 -> trac/u/ncohen/15310
   aa07ba1..d22e245  u/ncohen/16269 -> trac/u/ncohen/16269
 * [new tag]         6.2.rc1    -> 6.2.rc1
Already up-to-date.   [my emphasis]

Note that the 6.1.1-master in my prompt comes from:

6.1.1-master+1-sage: echo "$(git describe --abbrev=0)-$(git rev-parse --abbrev-ref HEAD)"
6.1.1-master

The real issue, however, seems to be my configuration and gcc. [As I am doing `make distclean` I don't know why make is reporting packages that didn't build, because I am not asking it to make any -- probably the build system stores packages that failed previously and when the distclean fails it prints these without checking if they are relevant?]

From the `make distclean` log:

...
checking for gcc... gcc
checking whether the C compiler works... no
configure: error: in `/usr/local/src/sage':
configure: error: C compiler cannot create executables
See `config.log' for more details

and in config.log:

configure:3645: checking for gcc
configure:3672: result: gcc
configure:3901: checking for C compiler version
configure:3910: gcc --version >&5
gcc (GCC) 4.7.3
Copyright (C) 2012 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

configure:3921: $? = 0
configure:3910: gcc -v >&5
Using built-in specs.
COLLECT_GCC=/usr/local/src/sage/local/bin/gcc

So make seems to be using $SAGE_ROOT/local/bin/gcc rather than /usr/bin/gcc = clang-503.0.40. I have taken brew out of the equation by changing my PATH and moving the /usr/local/*bin directories.

[Volker Braun]

I think you half-built Sage gcc conflicts with Apple gcc. Just "rm -rf /usr/local/src/sage/local" and try again.

[kcrisman]

On Tuesday, April 29, 2014 3:56:10 PM UTC-4, Volker Braun wrote:

Ok, I changed it to:

$ git trac find b42c581

Commit has been merged in 6.2.beta8.       <--------------------- new addition

Yes, thanks!  That will help a lot down the way.

Though Merged in is still a useful field ;-) 

[leif]

kcrisman wrote:
> Though Merged in is still a useful field ;-)

Couldn't a script update all newly merged tickets after each release
(including devel releases)?

I have no idea, but recall that Minh used to run a script to update the
Milestone fields (everytime a new stable version had been released).

[Andrew]

I now have a working version sage! I should qualify this: I have a working version of 6.2.rc0 but I am still unable to compile 6.1.1. I think that this is because of #15937. This patch is included in 6.2.rc0 and it fixes issues between singular and the version of xcode that preceded the one that I am using.

Please correct me if I am wrong, but I don't think that 6.1.1 compiles with the latest version of xcode and the command line tools (5.1.1 build 5B1008).

As I'd previously compiled 6.1.1 I assumed that my problems were caused by 6.2.rc0. Then I got stuck, however, because I couldn't recompile 6.1.1. In the comments above there is some confusion about which version of singular sage should have been using and that there may have been git issues. In fact, git was happy and sage was always using the appropriate version of singular for the version of sage being compiled. Once I realised this and found #15937 I gave up trying to compile 6.1.1 and found that I was able to compile 6.2.rc0.

I have to say that I am a little confused about the build process. With most systems,

make distclean

will attempt to return the code to its pristine download state. It seems that sage is doing something different? If this is the case, then I think it would be good to rename this make flag to something more accurate like rebuild/reset/...?

Anyway, sage is now working again so I am very happy. Thank you everyone for your help, especially Volker.

Andrew

[John H Palmieri]

Can you clarify what you mean here? "make distclean" should indeed return the code to its pristine download state, except that it won't clean up files in the "upstream" directory. Are there other things it doesn't clean up which it should?

  John
 

[Andrew]

Can you clarify what you mean here? "make distclean" should indeed return the code to its pristine download state, except that it won't clean up files in the "upstream" directory. Are there other things it doesn't clean up which it should?

 It doesn't remove the $SAGE_ROOT/local subdirectory either. I had the impression that it was also running the configure script(s), but perhaps I was wrong.

Andrew

[Volker Braun]

From the Makefile:

distclean: clean doc-clean lib-clean bdist-clean

        @echo "Deleting all remaining output from build system ..."

        rm -rf local

[Andrew]

Well, I'm completely confused as to why one of your recommendations above was:

I think you half-built Sage gcc conflicts with Apple gcc. Just "rm -rf /usr/local/src/sage/local" and try again.

A.

[Volker Braun]

If "make distclean" fails then $SAGE_LOCAL is not being deleted (it is the last command in the "distclean" target)