Jim Cromie wrote:
On Wed, Jun 17, 2009 at 11:17 PM, Bruno Wolff III<bruno@xxxxxxxx> wrote:
On Wed, Jun 17, 2009 at 08:25:27 -0400,
Bill Davidsen <davidsen@xxxxxxx> wrote:
I believe you need i686 or better to run f11. I suppose you could
recompile everything for [345]86 processor, but I don't think it comes
that way. And there may be some assumptions in scripts and commands that
certain features are present.
F11 runs on 586 processors.
Indeed. In fact, many packages came compiled that way.
[jimc@harpo dev]$ rpm -qa |grep 586 |wc
1335 1335 40840
[jimc@harpo dev]$ rpm -qa |wc
1604 1604 50709
Whats not clear is why theyre not i686, even for the kernel itself.
IIRC - i386 packages used to be default, but i686 kernels were always available,
and they were installed by default on my boxes.
I have the vague uneasy feeling that im losing 3-6 / PPM (parts per
million) in performance ;-)
The tech was discussed elsewhere, the "run/no run" is the discussion. The i586
used to mean the Intel Pentium, the next chip after the 486, but more recently
it has been applied to "compatible" chips, like the Via and AMD (K5?) chips. And
there were three flavors from Intel as well, I ordered processors and a
motherboard "back when" and the chips they sent wouldn't work in an SMP system.
Number which come to mind are P53, P54, and P54C, but if someone has the info at
hand good for then. Definitely won't boot on my old K5 in the corner, I intend
to make no effort to see why, but my old Slackware still boots. There were also
the models with the "f00f bug" which also had a few other firmware issues I
totally don't recall other than there were some, because after the upgrade we
didn't have to load the firmware at boot time.
I doubt you're losing performance except on a few models with certain pipeline
issues, and even then I doubt you can measure the drop. Google for details if
you care.
--
Bill Davidsen <davidsen@xxxxxxx>
"We have more to fear from the bungling of the incompetent than from
the machinations of the wicked." - from Slashdot
--
fedora-list mailing list
fedora-list@xxxxxxxxxx
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
Guidelines: http://fedoraproject.org/wiki/Communicate/MailingListGuidelines
