devel/m4 broken?

[Anonymous]

"R.Mahmatkhanov" <cvs-src at yandex.ru> writes:

> Anonymous пишет:
>> cvs-src at yandex.ru writes:
>> 
>>> Hi there!
>> 
>>> devel/m4 is broken for me for a long time. i use 7.2-STABLE, all ports
>> 
>> RELENG_7 (and RELENG_7_2) is a moving target. You're missing either date
>> of the build or exact revision number.
>
> RELENG_7 from July 28. kern.osreldate: 702104
>
>>> checking for posix_spawn... no
>> 
>> Can you show config.log?
>
> Sure. Attached.

> configure:6078: cc -std=gnu99 -o conftest -O2 -fno-strict-aliasing -pipe   conftest.c  >&5
> /var/tmp//ccK26sgd.o(.text+0x12): In function `main':
> : undefined reference to `posix_spawn'
> configure:6078: $? = 1
> configure: failed program was:
[...]
> configure:6078: result: no

Looks like you have no posix_spawn in your libc. I wonder where did you
get /usr/include/spawn.h that gcc complained about.

> ./spawn.h:113: error: conflicting types for 'posix_spawnattr_t'
> /usr/include/spawn.h:53: error: previous declaration of 'posix_spawnattr_t' was here

It doesn't look like spawn.h in RELENG_7 branch yet. But it *is* in HEAD.

$ svn log svn://svn.freebsd.org/base/stable/7/include/spawn.h
svn: File not found: revision 195974, path '/stable/7/include/spawn.h'

>
>> 
>> And try to add ac_cv_func_posix_spawn=yes to CONFIGURE_ENV in the port
>> Makefile to force using posix_spawn from /usr/include/spawn.h
>> 
>
> It fixed this. Thank you much!

I think the correct fix would to delete /usr/include/spawn.h so it
doesn't confuse configure script.