Max,
What is the rational for using...
SetCPPFLAGS: -mdynamic-no-pic
in the fink make 4.1 build? I commented that line out and so far the
rebuilt fink make isn't showing those parallel make failures on 10.11.
Perhaps that is tickling some obscure code generation bug in clang 7.
Jack
ps Two packages that exhibit the parallel build failures against stock fink
make 4.1 on 10.11 are cmake and texlive-base.
On Sun, Sep 13, 2015 at 11:38 AM, Jack Howarth <howarth.at.f...@gmail.com>
wrote:
> I also see a similar failure with fink make doing a parallel build of
> texlive-0.20140523-2...
>
> mv -f
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-start.Tpo
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-start.Po
> gcc -DHAVE_CONFIG_H -I. -I../../../texk/web2c -I./w2c -I/sw/include
> -I/sw/src/fink.build/texlive-0.20140523-2/texlive/Work/libs/luajit/include
> -DLUA_FF_LIB=1 -D_NO_PYTHON=1 -DX_DISPLAY_MISSING=1 -DUSE_OUR_MEMORY=1
> -I../../../texk/web2c/w2c -I../../../texk/web2c/luatexdir/unilib
> -I../../../texk/web2c/luatexdir/luafontloader/fontforge/inc
> -I../../../texk/web2c/luatexdir/luafontloader/fontforge/fontforge
> -I/sw/include -g -O2 -MT
> luatexdir/luafontloader/fontforge/fontforge/libjitff_a-stemdb.o -MD -MP -MF
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-stemdb.Tpo -c
> -o luatexdir/luafontloader/fontforge/fontforge/libjitff_a-stemdb.o `test -f
> 'luatexdir/luafontloader/fontforge/fontforge/stemdb.c' || echo
> '../../../texk/web2c/'`luatexdir/luafontloader/fontforge/fontforge/stemdb.c
> mv -f
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-stemdb.Tpo
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-stemdb.Po
> gcc -DHAVE_CONFIG_H -I. -I../../../texk/web2c -I./w2c -I/sw/include
> -I/sw/src/fink.build/texlive-0.20140523-2/texlive/Work/libs/luajit/include
> -DLUA_FF_LIB=1 -D_NO_PYTHON=1 -DX_DISPLAY_MISSING=1 -DUSE_OUR_MEMORY=1
> -I../../../texk/web2c/w2c -I../../../texk/web2c/luatexdir/unilib
> -I../../../texk/web2c/luatexdir/luafontloader/fontforge/inc
> -I../../../texk/web2c/luatexdir/luafontloader/fontforge/fontforge
> -I/sw/include -g -O2 -MT
> luatexdir/luafontloader/fontforge/fontforge/libjitff_a-tottf.o -MD -MP -MF
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-tottf.Tpo -c
> -o luatexdir/luafontloader/fontforge/fontforge/libjitff_a-tottf.o `test -f
> 'luatexdir/luafontloader/fontforge/fontforge/tottf.c' || echo
> '../../../texk/web2c/'`luatexdir/luafontloader/fontforge/fontforge/tottf.c
> mv -f
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-splinesave.Tpo
> luatexdir/luafontloader/fontforge/fontforge/.deps/libjitff_a-splinesave.Po
> Makefile:13666: recipe for target 'all-recursive' failed
> make[4]: *** [all-recursive] Error 1
> make[4]: Leaving directory
> '/sw/src/fink.build/texlive-0.20140523-2/texlive/Work/texk/web2c'
> Makefile:3580: recipe for target 'all' failed
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
> '/sw/src/fink.build/texlive-0.20140523-2/texlive/Work/texk/web2c'
> Makefile:878: recipe for target 'recurse' failed
> make[2]: *** [recurse] Error 1
> make[2]: Leaving directory
> '/sw/src/fink.build/texlive-0.20140523-2/texlive/Work/texk'
> Makefile:458: recipe for target 'all-recursive' failed
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory
> '/sw/src/fink.build/texlive-0.20140523-2/texlive/Work/texk'
> Makefile:544: recipe for target 'all-recursive' failed
> make: *** [all-recursive] Error 1
> make: INTERNAL: Exiting with 1 jobserver tokens available; should be 8!
> ### execution of /tmp/fink.1e959 failed, exit code 2
> ### execution of /tmp/fink.RIzpl failed, exit code 2
>
>
> On Sun, Sep 13, 2015 at 11:11 AM, Jack Howarth <howarth.at.f...@gmail.com>
> wrote:
>
>> Martin,
>> There is a regression in fink make under by El Capitan which I
>> haven't been able to pinpoint the origin of yet. It is commonly reproduced
>> here by trying to build fink cmake with fink make installed. The error
>> appears on my MacPro 3,1 as...
>>
>> [ 69%] Building CXX object
>> Source/CMakeFiles/CMakeLib.dir/cmDocumentation.cxx.o
>>
>> CMakeFiles/Makefile2:1972: recipe for target
>> 'Source/CMakeFiles/CMakeLib.dir/all' failed
>>
>> make[1]: *** [Source/CMakeFiles/CMakeLib.dir/all] Broken pipe: 13
>>
>> Makefile:160: recipe for target 'all' failed
>>
>> make: *** [all] Error 2
>>
>> make: INTERNAL: Exiting with 1 jobserver tokens available; should be 8!
>>
>> ### execution of /tmp/fink.ceCyn failed, exit code 2
>>
>> ### execution of /tmp/fink.eNsNl failed, exit code 2
>>
>> It can be avoided by hard coding the usage of system make into the cmake
>> build with...
>>
>> Index: cmake.info
>>
>> ===================================================================
>>
>> RCS file: /cvsroot/fink/dists/10.9-libcxx/stable/main/finkinfo/devel/
>> cmake.info,v
>>
>> retrieving revision 1.1
>>
>> diff -r1.1 cmake.info
>>
>> 2c2
>>
>> < Version: 3.3.0
>>
>> ---
>>
>> > Version: 3.3.1
>>
>> 5c5
>>
>> < Source-MD5: b1d4823e3f74c961486d9f253da9825d
>>
>> ---
>>
>> > Source-MD5: 52638576f4e1e621fed6c3410d3a1b12
>>
>> 22c22
>>
>> < make
>>
>> ---
>>
>> > /usr/bin/make
>>
>> Oddly building cmake under MacPorts with their make 4.1 doesn't trigger
>> the glitch under 10.11 here.
>> I haven't been able to reproduce the problem with the cmake build, using
>> fink make, outside of fink (even when running under 'sudo -u fink-bld
>> csh'). I wonder if there is something else set in the fink build
>> environment that could be that trigger to reproduce it outside of fink?
>> Jack
>> ps So far I have been unable to trigger any InfoTest failures in the fink
>> make build itself under El Capitan so the make testsuite seems unable to
>> capture the problem.
>>
>
>
------------------------------------------------------------------------------
_______________________________________________
Fink-devel mailing list
Fink-devel@lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel
