Re: [Boost-build] Is there any way to prevent Boost.Buildfromrecursively scanning header files for #include directives?

Subject: Re: [Boost-build] Is there any way to prevent Boost.Buildfromrecursively scanning header files for #include directives?
From: Konstantin Litvinenko (to.darkangel_at_[hidden])
Date: 2009-04-28 04:20:27

Johan Nilsson пишет:
> 
> There is (was?) something called "header cache" support in bjam 
> (originating from FTJam or Perforce Jam).
> 
> IIRC, this simply stored the names of all header files in a file 
> together with their timestamps and used this as an optimization. Don't 
> remember the details and thus don't know if this is something 
> conceptually similar to PSH. It might also not be supported no more.
> 
> Someone else (Rene/Volodya) probably remember the details.
> 

Yes it is, but not saved into file because HCACHEFILE variable not 
initialized anywhere. But anyway this doesn't help. On my projects 
setting it to '.bjam_hcache' does generate cache file and reduce no-op 
build from 32 s. to 31 s. - not much help :(. If some body more 
experienced in bjam codebase will look why bjam doing so many directory 
querying and looks for many strange named files (according to FileMon 
from SysInternals on windows machine) that would be great and may have 
positive changes in bjam performance.

 From my experience header scanning is high cost process but not 
measured in minutes. My own c++ scanner doing header scanning in less 
then a second on a project where bjam doing this 31 seconds. Maybe the 
algorithm it uses is not optimal or somewhat else prevent it from being 
fast.