bug id=57303

[Dennis Clarke]

> > Fire off a rebuild and see : 
> > aster $ ./util/modular/build.sh --clone --autoresume built.modules /opt/xorg
> > Building to run Linux / x86_64 ()
> > Fri Jan 25 19:46:35 EST 2013
> > Skipping util module component macros...
> 
> skipping means it didn't build it, so judging by the output you 
> skipped all modules.
> 

Just my luck :-)

Well I have it down to the following very repeatable steps : 

aster $ rm -rf xorg/
aster $ rm -rf /opt/xorg/*
aster $ mkdir -p xorg/util
aster $ git clone git://anongit.freedesktop.org/git/xorg/util/modular xorg/util/modular
Cloning into xorg/util/modular...
remote: Counting objects: 2345, done.
remote: Compressing objects: 100% (1214/1214), done.
remote: Total 2345 (delta 1487), reused 1765 (delta 1125)
Receiving objects: 100% (2345/2345), 1.04 MiB | 517 KiB/s, done.
Resolving deltas: 100% (1487/1487), done.
aster $ cd $HOME/xorg
aster $ CONFFLAGS=\-\-with\-udev\-rules\-dir\=/opt/xorg/udev
aster $ export CONFFLAGS
aster $ cp -p  util/modular/build.sh util/modular/build.sh_backup
aster $ diff $HOME/build.sh util/modular/build.sh
858c858
< #                 build driver xf86-video-geode
---
>                   build driver xf86-video-geode
874c874
< #            build driver xf86-video-i740
---
>             build driver xf86-video-i740
879,880c879,880
< #    build driver xf86-video-apm
< #    build driver xf86-video-ark
---
>     build driver xf86-video-apm
>     build driver xf86-video-ark
893c893
< #     build driver xf86-video-newport
---
>     build driver xf86-video-newport
897c897
< #    build driver xf86-video-s3
---
>     build driver xf86-video-s3
907c907
< #    build driver xf86-video-vmware
---
>     build driver xf86-video-vmware
aster $ cp -p $HOME/build.sh util/modular/build.sh

Then fire away the build and await the results. I know that I am still 
compiling buckets of drivers I will never use, but I don't miss the cpu time
 at all .. so why not. 

On another note I am one of those poor sad souls that has piles of servers running 
all manner of revs of Solaris on all manner of weird hardware.  Even this : 

titan-i386-SunOS5.8 $ uname -a 
SunOS titan 5.8 Generic_127722-03 i86pc i386 i86pc
titan-i386-SunOS5.8 $ 
titan-i386-SunOS5.8 $ cat /etc/release 
                       Solaris 8 2/02 s28x_u7wos_08a INTEL
           Copyright 2002 Sun Microsystems, Inc.  All Rights Reserved.
                           Assembled 18 December 2001
titan-i386-SunOS5.8 $ 
titan-i386-SunOS5.8 $ psrinfo -v 
Status of virtual processor 0 as of: 01/26/13 05:55:55
  on-line since 04/28/11 17:39:44.
  The i386 processor operates at 400 MHz,
        and has an i387 compatible floating point processor.
Status of virtual processor 1 as of: 01/26/13 05:55:55
  on-line since 04/28/11 17:39:48.
  The i386 processor operates at 400 MHz,
        and has an i387 compatible floating point processor.


How is that for bizarre old hardware ?  I don't think anyone will ever 
know how many packages from Blastwave were compiled on that 
very machine and they went out to the world to run flawlessly all 
the way up to x86_64 based Solaris 10 servers.

I always had a morbid curiosity with compiling awesome software on
old bucket systems running the oldest UNIX kicking around.  

I just may try this process on old Solaris 8 sparc and i386 however I 
will need to get a version of git running first.  Not very likely.

Solaris 10 is far more likely to happen. 

In any case gents, this is moving along towards success wonderfully and 
maybe this weekend I will be able to fire up new X with a simple xterm 
and see it working. 

Dennis