[prev in list] [next in list] [prev in thread] [next in thread] 

List:       boost-build
Subject:    Re: [Boost-build] bjam install
From:       "Belcourt, K. Noel" <kbelco () sandia ! gov>
Date:       2010-11-20 2:54:23
Message-ID: 3EDB2F59-1585-4E58-B707-EB32653A455D () sandia ! gov
[Download RAW message or body]

Hi Brent,

On Nov 19, 2010, at 7:30 PM, Bryce Lelbach wrote:

> I let one of the processes run in an idle shell, just for kicks.  
> Eventually, it
> did finish, but, clearly there's a problem if three copy commands  
> and a ln-UNIX,
> are taking 5+ minutes... is this likely an issue on my end?

No I don't think so.

> wash@Pegasus:~/boost$ time bjam install --prefix=/usr/local/

I suspect this is the well known bjam startup penalty.  If I  
understand it correctly, it amounts to the overhead involved in  
recreating the dependency graph and determining out of date targets  
each time bjam starts up (about 5 minutes on a slower system, around  
90 seconds on a faster one).

One way to test this is to cd to boost/status, edit the Jamfile to  
remove all libraries/test except for (in your case) system/test, and  
rerun bjam, does it still take 5+ minutes before patience ends and  
targets start building?

-- Noel


_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost-build
[prev in list] [next in list] [prev in thread] [next in thread]