[sldev] [VWR] Vectorization on i386?

[Robin Cornelius]

On Dec 3, 2007 8:13 PM, Tofu Linden <tofu.linden at lindenlab.com> wrote:
> Callum Lerwick wrote:
> >> What testing?  I thought you said you couldn't get this enabled on i386?
> >> Which part of the problem are you failing to reproduce?
> >
> > The viewer runs, and always has run on my Athlon T-bird just fine. And
> > I've run it on my dual PII 333 system. If it were leaking SSE code into
> > places it shouldn't, I would have run into it by now.
>
> It must be a bad brain day for me, because I still don't understand;
> the viewer runs and has always run on the Athlon because SSE code isn't
> enabled on i386 builds by default (and we happen to prefer GCC 3.4 for
> now which doesn't leak SSE code), which I thought was the whole point
> of this thread.


I think Callum's point is the SSE code is all there in the binary, its
only the initial test in viewer.cpp that is hidden in a #define.  As
this test works on the linux cpu string anyway it might just as well
be tested.

Right i'm going to upset the Debian users now too!

Robin