[haiku-development] Re: Haiku SMP Status
- From: "Axel Dörfler" <axeld@xxxxxxxxxxxxxxxx>
- To: haiku-development@xxxxxxxxxxxxx
- Date: Wed, 12 Dec 2007 15:55:48 +0100 CET
"Rene Gollent" <anevilyak@xxxxxxxxx> wrote:
> > It is hit if you're BIOS sets up ACPI tables. But on a single CPU
> > system you shouldn't really notice anything. Even if the code
> > failed,
> > it normally just resorts back to 1 CPU. But if it hangs for you or
> > something like that since the changes, I'd obviously like to know ;
> > -)
> OK, will check in a few moments ; how can I determine if it used the
> ACPI detection code? Will it be written to syslog?
It should be in there, too, right - depends on how much info the boot
loader dumps into it on your system (currently, only 4096 bytes are
reserved for this, which is usually too little with all the output we
have now).
> > I think it does not consider it at all right now. But there's
> > supposed
> > to be a completely new scheduler implemented. I guess this one will
> > include affinity? Does anyone know of news on that one BTW?
> Ah right, the SoC project...I completely forgot about that. Last I
> checked it was still in a testing branch pending some bugfixes before
> being merged, but Andre or Axel would be best qualified to answer
> that.
I haven't heard from Andre in weeks, and the current test
implementation that I have from him (as a separate app), doesn't cover
all of the needed cases (he has revised his plans once more, and was
working on that last time I talked to him).
It's not really urgent, as the current scheduler works, too, but we'll
see when it happens :-)
Bye,
Axel.
Other related posts:
