From: David Abrahams (gclbb-jamboost_at_[hidden])
Date: 2003-08-01 11:45:53
Vladimir Prus <ghost_at_[hidden]> writes:
> David Abrahams wrote:
>> Vladimir Prus <ghost_at_[hidden]> writes:
>> > Hmm... the net time for this rule alone is 1second. Pretty high.
>> Does that make it worthwhile to optimize?
> IMO, yes.
>> Is that a significant percentage of the total?
> The problem is that there's no single rule which takes 50% of the running
> time. It's a bunch of rules which take too much time, which, in total, makes
> V2 slow.
OK. I've had success chipping away at 1% slownesses, but it's
tedious. It would be nice to find some big wins.
> Concerning this rule: it's some internal bookkeeping that V2 does and yet it
> takes 1 second on a pretty fast machine. Seems a good candidate for
> optimization. Or do you have other ideas?
-- Dave Abrahams Boost Consulting www.boost-consulting.com
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk