Re: [PATCH v2] x86/mce: Distirbute the clear operation of mces_seen to Per-CPU rather than only monarch CPU

[Tony Luck]

On Thu, May 22, 2014 at 6:32 PM, Chen Yucong <slaoub@xxxxxxxxx> wrote:
> As Naoya Horiguchi says, this patch also have a small benefit that it
> can reduce the processing time of monarch CPU.

This is indeed a benefit - but I'm not super worried about performance
of machine check handler.

>        /*
>         * Now clear the mces_seen of current CPU -*final - so that it
> does not
>         * reappear on the next mce.
>         */
>        memset(final, 0, sizeof(struct mce));
>        mce_wrmsrl(MSR_IA32_MCG_STATUS, 0);

But if the monarch hasn't managed to clean mces_seen, then
it certainly hasn't cleared MCG_STATUS ... so there can't be
a "next" mce that would see these old values. Any extra MCE
will result in system reset.

So we are not arguing that your patch is wrong - it just doesn't seem
to be any better that what we have now (except for an unimportant
small performance improvement).

-Tony
--
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/