segfault during configure, lang/python27 on armv6

[Kubilay Kocak]

On 10/09/2015 2:16 PM, Mike Brown wrote:
> I am trying to get Python installed from the ports collection on my BeagleBone 
> Black, but got a segmentation fault during the 'configure' phase of the build.
> 
> 
> [...]
> checking for setitimer... yes
> checking for getitimer... yes
> checking for bind_textdomain_codeset... yes
> checking for chown... (cached) yes
> checking for clock... (cached) yes
> Segmentation fault (core dumped)
> ===>  Script "configure" failed unexpectedly.
> Please report the problem to python at FreeBSD.org [maintainer] and attach the
> "/usr/ports/lang/python27/work/Python-2.7.10/config.log" including the output
> of the failure of your make command. Also, it might be a good idea to provide
> an overview of all packages installed on your system (e.g. a
> /usr/local/sbin/pkg-static info -g -Ea).
> *** Error code 1
> Stop.
> 
> 
> The OS is the snapshot FreeBSD-10.2-STABLE-arm-armv6-BEAGLEBONE-20150903-r287435
> from ftp://ftp.freebsd.org/pub/FreeBSD/snapshots/arm/armv6/ISO-IMAGES/10.2/
> 
> package list: http://pastebin.com/raw.php?i=SJExBxru
>   config.log: http://pastebin.com/raw.php?i=M7ERWKx1
> 
> It was /bin/sh that dumped core in the work/Python-2.7.10 directory. Nothing 
> in the gdb backtrace, just #0  0x00000000 in ?? ().
> 
> 
> Anything I should do before trying again?

A few things that may help to isolate:

 * With LIBFFI option on and off (off uses: bundled libffi, on: the
libbfi port)
 * Build with DEBUG option enabled
 * See if issue is reproducible with other Python ports
(lang/python{33,34,35})
 * Update to latest 10.2-STABLE
 * Update to latest 10-STABLE

Having said that, I don't like that it's dying @ configure, which may
indicate an autotools/armv6 issue. Having said that though I'd have
expected more reports if it were that widespread.

Let us know how you go.

./koobs