[rescue] Spark10 CPU question
mcguire at neurotica.com
Thu Dec 15 00:32:15 CST 2016
On 12/14/2016 09:27 PM, Sandwich Maker wrote:
> the giant spinlock seems like rather a crude hack for multiproc
> support. surely sun could've done better -
There was pretty significant time pressure to get multiprocessor
support...ANY multiprocessor support...shipped. And it did get a lot of
work done, that's for sure, but yes, multiple processes in kernel space
simultaneously sure would've been nice.
> but they didn't have to; i
> was at another job that had a solbourne about the time they were going
> under, and i thought it was a shame that their work on finer-grain
> spinlocks in the sunos kernel was just landfilled [along with other
> interesting things].
Yeah. Solbourne did amazingly good work.
> but sun was understandably not interested in
> anything which would slow customers in moving from sunos to solaris,
> especially when they were moving much slower than they wanted anyway.
...because it SUCKED for so many years! Lots of SunOS4 shops, mine
included, were just plain disinterested. They forced our hand by
introducing new machines, UltraSPARCs, that weren't supported by SunOS4.
We were still a mostly-SunOS4 shop until I left that place in 1997. It
wasn't for another year or two after that I considered Solaris fit for
use in a production environment.
Nowadays, I'd run it on everything if I could, but Oracle has their
ties tied too tightly. So I only run it on the servers that do the
heavy lifting and face the outside world. Best OS I've ever run, and
I've run quite a few.
Dave McGuire, AK4HZ
New Kensington, PA
More information about the rescue