[ale] CentOS 7 install driving me mad

Chris Fowler cfowler at outpostsentinel.com
Fri Jun 3 14:49:34 EDT 2016


> From: "DJ-Pfulio" <DJPfulio at jdpfu.com>
> To: ale at ale.org
> Sent: Friday, June 3, 2016 2:11:11 PM
> Subject: Re: [ale] CentOS 7 install driving me mad

> Just because something **can** be done, doesn´t make it a good idea.

<disclaimer>drinking some moonshine on bday</disclaimer> 

Just because you think it may not be a good idea does not make it so. World is so much bigger than that :) 

The end game is this. I want to move my dev tree from CentOS 5 i386 (the real deal. Not some bullsh*t where someone has labeled i686 as i386). 
to CentOS 7 i586. This allows me to support this: 

vyos at vyos:~/nDPI-1.8$ cat /proc/cpuinfo 
processor : 0 
vendor_id : Genuine RDC 
cpu family : 5 
model : 8 
model name : 05/08 
stepping : 6 
cpu MHz : 999.982 
physical id : 0 
siblings : 1 
core id : 0 
cpu cores : 1 
apicid : 0 
initial apicid : 0 
fdiv_bug : no 
f00f_bug : no 
coma_bug : no 
fpu : yes 
fpu_exception : yes 
cpuid level : 1 
wp : yes 
flags : fpu tsc cx8 mmx 
bogomips : 1999.96 
clflush size : 32 
cache_alignment : 32 
address sizes : 32 bits physical, 32 bits virtual 
power management: 

vyos at vyos:~/nDPI-1.8$ 

And to support a Baytrail. 

While Joe blow may have only one RDC based device I have HUNDREDS. I have 300 J1900 BayTrail devices. 

If I can move to a base of CentOS 7 i586 then I may keep typing and split the SVN tree to support i586 and the BayTrail. This will give me optimization on both platforms for the CPU. 

I can no longer purchase the RDC hardware, but I have to support it. I also have a limited stock of it for internal use. Right now I'm using one as a pfsense router on my home network. I have another running Asterisk PBX. I am not running X. I'm running stuff where the load average is typically 0.25 or less! So why would it not be a good idea? 

It looks to me like 3.13 is happy. 

vyos at vyos:~/nDPI-1.8$ uname -a ;uptime 
Linux vyos 3.13.11-1-586-vyos #1 SMP Wed Aug 12 01:58:45 UTC 2015 i586 GNU/Linux 
18:38:08 up 44 min, 1 user, load average: 0.00, 0.04, 0.05 
vyos at vyos:~/nDPI-1.8$ 

This is one thing that grinds my gears. "I don't believe people should be using MP3 codecs because it is not open therefore I will not make it available. I also believe that everyone should be on 64bit systems so I will not make 32bit available" My original config program for Linux was VI so I'm capable of adding in the media repositories, but my gripe is based on me thinking of others who do not know how. They must search "how to play MP4 on Ubuntu". At least Ubuntu now allows you to make your own decisions. 

As a ham I can also break it down like this: 

"Just because you don't use a radio does not mean everyone else does not. Do not assume that something is obsolete." 

In the final I need to weight this: 

Do the benefits of compiling everything for the target CPU outweigh the management headache of two trees and two firmware packages? 

That does not address the fact that compiling modern packages for CentOS 5.X is getting harder. I guess I could recompile the whole dev environment and shove the packages of 7 into 5? 

What sucks is that this project is a real time sink. I get an idea and walk that path for 8+ hours. Change my mind and stop for a while. Repeat. 

One really great idea is Cross LFS. I worked that for 10 hours then my patience grew thin and I screwed it up. I even did a PI2 to x86-i586 so I'd learn more. CentOS 7 is just a target. Linux 4.X via i586 Cross-LFS is acceptable and I may try again next week. This time I'm going to write scripts for the stages. Put the RDC on gigabit for NFS instead of wireless bridge. 

Sure, I could do CentOS 6, but I'm burning up time and doing 7 gets me that much ahead. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ale.org/pipermail/ale/attachments/20160603/6d72cbb8/attachment.html>


More information about the Ale mailing list