SCIENTIFIC-LINUX-USERS Archives

April 2016

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Lamar Owen <[log in to unmask]>
Reply To:
Lamar Owen <[log in to unmask]>
Date:
Fri, 15 Apr 2016 10:58:11 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (34 lines)
On 04/14/2016 03:49 PM, R P Herrold wrote:
> Lamar -- could you please pull and build SRPMs out of the 'local' 
> path, and confirm that no gremlins have crept in, via a cross-check on 
> your buildsystem? I don't use 'mock' for scratch solves Thanks -- Russ 
Russ, I'll try to look at it later today.  While I have a full-blown  
mock buildsystem on the IA64 box, that's CentOS 5..... I haven't done 
the same on CentOS 7 as yet, as I haven't had the need.  Yeah, I've not 
been doing clean scratch builds on C7 yet, although I need to do so.  I 
need to add RAM to make it faster, and I need to add disk for the repo 
cache (I have 20GB of RAM now in my laptop (Dell M6500 Core i7), but I 
need 32GB to really be able to put the buildroot entirely in RAM during 
the build, and a second 1TB drive (putting the boot SSD on mSATA module 
for three drives).... Well, I don't *need* to do that, but I *want* to 
do that, as 'buildroot in RAMdisk' makes builds fly..... at least that 
has been my experience on the Altix IA64 boxen, where putting the 
buildroot in a RAMdisk cut build iteration times (especially for the 
kernel, glibc, and gcc) to 10% of what they had been, since the Altix is 
not an I/O monster, but the RAM bandwidth is very high.

Yasha, this is a classic automaton problem, and should be solved 
accordingly.  The state machine needs inputs of dependencies, and the 
build order for the dependencies will become clear.  It is also a 
*solved* problem, in the form of the 'smock' perl script, which you can 
find at https://github.com/richm/scripts/blob/master/smock.pl

But you may want to not pass up the teaching opportunity of applying 
'critical thinking' about 'what constitutes an automaton?' as it relates 
to depsolving.  Russ has done the legwork of getting all the inputs you 
need in the ftp area he posted.  Simple (and not so simple) depsolving 
is a really good way to get state machine theory into a hands-on 
learning activity to bring the abstract into the realm of the concrete.  
At least I think it is an interesting problem.  The next step is solving 
something like the fpc (Free Pascal Compiler) build dependencies.....

ATOM RSS1 RSS2