Re: [Boost-build] [patch review] Re: Help optimizing a slow BJAM build

From: Vladimir Prus (ghost_at_[hidden])
Date: 2008-04-11 05:30:15

On Friday 11 April 2008 04:52:14 Jurko Gospodnetiæ wrote:
>    Hi Igor.
> 
>  > I'm starting to think that perhaps I'm not building the same BJAM that
>  > you are building. Is there a way for me to check out exactly the same
>  > version of the trunk that you are using? Also, can you send the
>  > commands that you used to build BJAM?
> 
>    OK, it seems I must have done something wrong the first time. Sorry...
> 
>    I now deleted all the builds and rebuilt them again and now I get 
> timings matching your own - both with the original test Jamfile and the 
> new one you sent.
> 
>    Time to change my vote and say that the patch should be applied to 
> the trunk. :-) And possibly a test added based on the latest example to 
> guard against any regressions. Volodya, should I do this or should we 
> wait for more opinions?

I actually wanted to figure out why the current algorithm performs that
bad, but I'll have little time next week, and that's more of a theoretical
question. I think getting this in is best at this point. Thanks!

>    I retested the new builds with one of our own big projects but it 
> seems that we just do not hit this speedup scenario there and timings 
> remain the same. Ah well... :-)

Well, one cannot expect one patch to fix all issues :-)

- Volodya