Re: NForce2 pseudoscience stability testing (2.6.0-test11)

[bill davidsen]

In article <3FCD32F5.2050002@xxxxxx>,
Prakash K. Cheemplavam <prakashkc@xxxxxx> wrote:
| > smp off, preempt off. lapic on, apic on, acpi on
| 
| Why haven't you enabled preempt? Does it lock with preempt on?

I haven't found preempt to cause problems with my applications, but
neither have I seen any significant gain in throughput or responsiveness
in 2.6. I did see small gains in 2.4, but not enough to motivate me to
add the feature as a patch or even, in 2.6, enable it as an option.

I don't know about the original poster, but the reports of problems
discourage exploration unless the application is likely to benefit from
slight gains in response and to have multiple tasks of different
priority running.

Have you seen a benefit from preempt? And if so, what application and
how much? I may not be doing anything which benefits from preempt enough
to notice (news, mail and dns servers, desktops), but other than dns on
2.4 I haven't seen improvement.

enlightenmet, please?
-- 
bill davidsen <davidsen@xxxxxxx>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
-
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/