[rescue] sparc10 cpu - what to do.
jjhudak at gmail.com
Tue Dec 20 13:34:44 CST 2016
Still a requirement in my book!
I once interviewed ppl for a sw engr position for RTE application. One
person was very proud of their position that they had to know *very little*
about the actual hardware (e.g. the machine) that they were dealing with.
He was well enamored with his ability to 'make his C code do anything and
not be concerned about HW 'details'....after giving him enough hints that
knowing some hw detail was *really* important for this application, he did
not take the hint. His abstraction level was permanently fixed on I can
make C do anything...After a bit more discussion, I politely ended the
interview, and put his resume in the circular file.
On Tue, Dec 20, 2016 at 1:31 PM, Dave McGuire <mcguire at neurotica.com> wrote:
> On 12/20/2016 01:29 PM, Mouse wrote:
> >> I've wondered what kind of performance boost we could get if we ran
> >> directly at the microcode layer (the one that pretends it's this
> >> obsolete junk).
> > Significant, probably, but also hellishly complicated to do anything
> > useful with. Risky, too; microcode tends to allow things like gating
> > more than one internal register onto an internal data bus at the same
> > time. And probably strictly Harvard, rather than von Neumann.
> > And very, very processor-rev-specific.
> That would very quickly bring back the now-obsolete requirement that
> programmers actually know what they're doing. ;)
> Dave McGuire, AK4HZ
> New Kensington, PA
> rescue list - http://www.sunhelp.org/mailman/listinfo/rescue
More information about the rescue