Re: MATH_EMULATE

[Heikki Suonsivu <hsu%hsu.bbnetworks.net@localhost>]

Joerg Sonnenberger <joerg%britannica.bec.de@localhost> writes:

> On Mon, Jan 21, 2008 at 05:31:50PM +0200, Heikki Suonsivu wrote:
> (a) MATH_EMULATE is slow.
> (b) Keeping all the overhead in userland is not as bad.

My point is avoiding dual ports.  If a specific application is time
critical and too slow with MATH_EMULATE, one could always compile that
specific package using soft-float, as that does not conflict with
having MATH_EMULATE handling the default case?

I am putting convenience here over performance, it is better if it
first works at all, and then works fast if you tune the knobs.  These
won't be number-crunching things anyway, more like small applications
like web servers, wireless APs, thin clients.  Or maybe I should not
be saying that, the next thing from Taiwan may be 4096 486 cores on
single chip, with no fpu :)

> (c) Serious, you are talking about hardware with limited computation
> power -- binary compatibility is not such a big deal for that.

It is very convenient to install binary packages to slow computers
instead of compiling it locally.  But this is variation of multiple
ports theme :)

> That said, I have to take a look at how much work -msoft-float really
> is. I would be willing to fix it.
> 
> Joerg