Re: revamped gdb_mbuild.sh

[Richard Earnshaw <rearnsha at arm dot com>]

> > Richard,
> >> 
> >> Finally got to trying the gdb_mbuild.sh script and, in the process, 
> >> integrated some, er, `new features' from my old local script:
> >> 
> > 
> > 
> > Looks pretty good to me.  My only comment is that we loose the ability to 
> > run gnumake with -j <n> replacing it with simultaneous configures/builds 
> > of different targets.  That means a higher transient disk usage (storage) 
> > which would be a marginal problem for me due to lack of disk quota on the 
> > multi-way machines... ;-(  But that's me thinking about my environment, so 
> > it isn't a major objection
> 
> I was wondering about a schema where one resource was allocated to 
> configure while the remaining N-1 were allocated to a single build. 
> That would mean an even faster turnaround on the first build (which I've 
> found is what I'm after :-).
> 
> I guess we just wonder ...
> 
> Andrew
> 
> 


How about -c <x> -j <y>?  Ie x configures in parallel & y make jobs in 
parallel?

R.