Re: [openvrml-develop] Status update: breaking up large files

[Braden M. <br...@en...>]

On Sat, 2008-09-27 at 01:58 -0400, Braden McDaniel wrote:

[snip]

> So far I've pared browser.cpp down to a little more than 6000 lines.
> Compiling on my x86_64 Linux machine, its high-water mark in memory is
> around 1.3 GB.  If that sounds big, consider that before this surgery it
> was taking at least 2.2 GB to compile.  (And recall that for a 32-bit
> platform, you can expect to cut this memory footprint roughly in half.)
> 
> I suspect that the better part of that 1.3 GB footprint has to do with
> the fact that two big Spirit parsers get instantiated in browser.cpp.
> Pushing these instantiations out to different translation units would
> probably be a significant win; and that's something I'll probably pursue
> before releasing 0.18.

My most recent commit makes good on this threat.  I moved the parser
instantiations to src/libopenvrml/openvrml/local/parse_vrml.cpp.  The
parse actions classes were the biggest chunk of code directly involved
here; they're now in src/libopenvrml/openvrml/local/parse_vrml.h.  This
means the most expensive-to-compile file is now parse_vrml.cpp; what's
left of browser.cpp should be pretty trivial.

On my machine, parse_vrml.cpp's high water mark in memory seems to be
somewhere between 900 MB and 1 GB.  I think this means that OpenVRML
should compile comfortably in a non-parallel build on 64-bit machines
with at least 2 GB of memory; and on 32-bit machines with at least 1 GB
of memory.  For machines with more memory, OpenVRML's build will
parallelize pretty nicely.

I think the above minimums are pretty reasonable. So I'm inclined to
consider this pursuit done unless another problem emerges.

-- 
Braden McDaniel                           e-mail: <br...@en...>
<http://endoframe.com>                    Jabber: <br...@ja...>