Re: GNU make 3.99.91 release candidate is available

[Dagobert Michelsen]

Hi Paul,

Am 22.09.2013 um 18:57 schrieb Paul Smith <address@hidden>:
> I've done some testing and bug fixing using the Solaris build service,
> thanks very much Dago!
> 
> On Mon, 2013-09-16 at 20:52 +0200, Dagobert Michelsen wrote:
>> I'll get an error compiling it on Solaris 10 Sparc with Sun Studio 12:
> 
>>> /opt/SUNWspro/bin/cc -D_REENTRANT -pthreads
>> -I/opt/csw/include/guile/2.0 -I/opt/csw/include   -xO3 -m32
>> -xarch=sparc  -m32 -xarch=sparc -L/opt/csw/lib -o make ar.o arscan.o
>> commands.o default.o dir.o expand.o file.o function.o getopt.o
>> getopt1.o implicit.o job.o load.o loadapi.o main.o misc.o output.o
>> read.o remake.o rule.o signame.o strcache.o variable.o version.o
>> vpath.o hash.o remote-stub.o guile.o glob/libglob.a
>> -lkstat  /opt/csw/lib/libintl.so -R/opt/csw/lib -L/opt/csw/lib
>> -lguile-2.0 -lgc    -lrt 
>>> ld: fatal: soname option (-h, --soname) is incompatible with
>> building a dynamic executable
>>> ld: fatal: flags processing errors
>>> gmake[2]: *** [make] Error 1
> 
> Investigating this I've discovered the problem.  The Guile library
> provided on these systems was compiled with GCC, not the Sun Studio
> compilers.
> ...
> I don't actually have any good ideas about how to fix this problem.
> Most likely this is something that needs to be raised with the autoconf,
> pkgconfig, and/or Guile folks.

Ah, I know. That happened every once in a while in the past. This is solely
an issue about our distribution and we will switch to gcc completely
for the next release. I suggest you ignore the issue.

> I also found that for some reason configure is determining that libc
> contains a GNU gettext implementation which is clearly not true, and
> thus leaving off the -lintl library causing link errors.
> 
> Finally, because the GNU libraries are not installed in the standard
> locations you need to add "LDFLAGS=-R/opt/csw/lib" to the configure line
> so that GNU make will know where to look for the guile, etc. libraries
> at runtime.  Using LD_LIBRARY_FLAGS is not a good idea because it will
> break other programs.

Our build system adds LD_OPTIONS=-R/opt/csw/lib among other Solaris-specific
flags as they usually have problems with some libtool versions.

> After working around these issues and my regression test fixes (to be
> pushed shortly), I am able to build on Solaris 11 with the Sun Studio
> compiler and get 100% test coverage, including Guile support:
> 
>  540 Tests in 119 Categories Complete ... No Failures :-)
> 
>  
> ======================================================================================================
>   Regression PASSED: GNU Make 3.99.91 (sparc-sun-solaris2.11) built with 
> /opt/solarisstudio12.3/bin/cc 
>  
> ======================================================================================================

I still have two fails with SunStudio without Guile and with GCC and both 
with/without Guile:

targets/ONESHELL ........................................ Error running 
/home/dam/mgar/pkg/gmake/trunk/work/solaris10-sparc/build-isa-sparcv8plus/make-3.99.92/tests/../make
 (expected 0; got 512): 
/home/dam/mgar/pkg/gmake/trunk/work/solaris10-sparc/build-isa-sparcv8plus/make-3.99.92/tests/../make
 -f work/targets/ONESHELL.mk.1

address@hidden 
[unstable10s]:/home/dam/mgar/pkg/gmake/trunk/work/solaris10-sparc/build-isa-sparcv8plus/make-3.99.92
 > cat ./tests/work/targets/ONESHELL.diff.1
*** work/targets/ONESHELL.base.1        Mon Sep 23 13:42:39 2013
--- work/targets/ONESHELL.log.1 Mon Sep 23 13:42:39 2013
***************
*** 1,2 ****
--- 1,5 ----
  a=$$
  [ 0"$a" -eq "$$" ] || echo fail
+ -c: -c: cannot open
+ work/targets/ONESHELL.mk.1:5: recipe for target 'all' failed
+ make: *** [all] Error 1

variables/SHELL ......................................... Error running 
/home/dam/mgar/pkg/gmake/trunk/work/solaris10-sparc/build-isa-sparcv8plus/make-3.99.92/tests/../make
 (expected 0; got 512): 
/home/dam/mgar/pkg/gmake/trunk/work/solaris10-sparc/build-isa-sparcv8plus/make-3.99.92/tests/../make
 -f work/variables/SHELL.mk.6

address@hidden 
[unstable10s]:/home/dam/mgar/pkg/gmake/trunk/work/solaris10-sparc/build-isa-sparcv8plus/make-3.99.92
 > cat ./tests/work/variables/SHELL.diff.6
*** work/variables/SHELL.base.6 Mon Sep 23 13:42:44 2013
--- work/variables/SHELL.log.6  Mon Sep 23 13:42:44 2013
***************
*** 1,2 ****
! + true 
! + true 
--- 1,3 ----
! -c: -c: cannot open
! work/variables/SHELL.mk.6:3: recipe for target 'all' failed
! make: *** [all] Error 1

The tests are invoked with
  cd tests && perl ./run_make_tests.pl -srcdir 
/home/dam/mgar/pkg/gmake/trunk/work/solaris10-sparc/build-isa-sparcv8plus/make-3.99.92
 -make ../make 

Maybe there is some environmental $(MAKE) leaking from our enclosing package 
build system?


Best regards

  -- Dago

-- 
"You don't become great by trying to be great, you become great by wanting to 
do something,
and then doing it so hard that you become great in the process." - xkcd #896

smime.p7s
Description: S/MIME cryptographic signature