On Tue, Sep 30, 2008 at 04:54:16PM -0700, Dan Price wrote:

> So when you profile this code, where are the hotspots?

Haven't done that yet; I'll take a look tomorrow.

> I'm slightly shocked at the lack of speedup, but I'm wondering if
> the issue is that we're having to burn a lot of cycles on malloc
> inside of the various Py_XYZ routines.

I'd be surprised.  I only call *alloc() twice, and that's in very rare
cases -- ones that aren't tested by the performance benchmarks.  So I'm
pretty sure it's all inside the Py_* routines (though perhaps it's all in
memory allocation therein).

Danek
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to