Re: [PATCH 1/2] x86: mce init optimization and signedness fixup

[Alan Cox]

On Thu, 11 Oct 2007 16:55:36 +0200 (CEST)
Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:

> On Thu, 11 Oct 2007, Christoph Egger wrote:
> 
> > On Thursday 11 October 2007 15:51:49 Thomas Gleixner wrote:
> > > On Thu, 11 Oct 2007, Joerg Roedel wrote:
> > > > MCG_CAP never reports a negative count of available error-reporting
> > > > banks. Therefore, make nr_mce_banks unsigned.
> > > > Check for MCA/MCE feature bits as early as possible.
> > > >
> > > > +
> > > > +	if (!cpu_has(c, X86_FEATURE_MCA) || !cpu_has(c, X86_FEATURE_MCE)) {
> > > > +		printk(KERN_INFO "CPU%i: No machine check support available\n",
> > > > +			smp_processor_id());
> > > > +		return;
> > >
> > > This breaks winchip MCE support.
> > 
> > First, what is a winchip? It sounds to be something windows specific. ;)
> > Second, can you explain in which way MCE support gets broken, please?
> 
> First, winchip is the code name of Centaurs early x86 cpus.
> 
> Second, those beasts do not have FEATURE_MCA, but they have FEATURE_MCE, 
> so they support the fatal exception, but not the non fatal check.

Ditto some of the original Intel Preventium processors which have limited
MC support hidden in the magic 'Appendix H'

Alan
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/