> [ sorry for the very late reply but I am been overloaded ]
>
> On Thu, 9 Dec 1999, Peter Rival wrote:
>
> >it (correct me if I paraphrased wrong :). If there are no violent
> >objections, I'd like to change this in the base tree as soon as possible
> >so that it gets as much soak time as we can before 2.4 is a reality.
>
> In general it's always safe and a bit more performant to remove it.
>
> But on alpha we don't have an NMI watchdog so it would be better to left
> them enabled at least until the 2.4.x branch will take place...
Yes, but... :) If we wait until the 2.4.x branch, our opportunity to "soak" this
relatively major change without production impact will have passed us by. The last thing
I want to do is add performance that impacts stability in, or too close to, a "stable"
kernel series. I'd like to get a few weeks or more of time with this change. I'm doing
it myself in all my kernels, but I know that I'm not running anywhere near the expanse of
programs and stresses on the different types of systems that the Linux community is.
> Is it
> possible to generate real NMI on alpha? I think it would be possible to
> use the RTC to emulate an nmi changing __cli() to raise the masking of
> irqs to 6.
>
Jay or Richard would be the best ones to ask, not me (yet (unfortunately) ;).
>
> About the kernel debugging tools I want to make the IKD stuff running on
> alpha too (as first kernel tracer and then memleak detector). kdb port
> will be not trivial, I'll start hacking with:
>
> mv linux/arch/i386/kdb linux/
>
Kewl. I'm hoping that we can move as many of the well-liked tools that are Intel-only
over to Alpha. I've started one, you're starting another, and someone else is doing the
LKCD, IIRC. The snowball is growing :)
- Pete
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/