Subject: Re: [linux-audio-dev] Another Annoying "How Do I Get Started" Question
From: Christopher Lee (clee_AT_gnwy100.wuh.wustl.edu)
Date: Thu Jul 12 2001 - 23:58:12 EEST
>>>>> "Andre" == Andre Majorel <amajorel_AT_teaser.fr> writes:
Andre> On 2001-07-12 00:39 -0500, Christopher Lee wrote:
>> I've read that the fastest x86 intel architecture can do interrupts
>> is w/ max latencies of 40us using rtlinux. You might be able to do
>> at least the interrupt part with a PPC-based system (1us
>> interrupts).
Andre> Did you write 40 microseconds ? That's puzzling to say the
Andre> least. Even in the bad old times when all we had was one-MHz
Andre> 6502s, it was possible to achieve higher interrupt rates than
Andre> that.
Yes, microseconds. x86 isn't very fast when in comes to interrupts as
compared to many other architectures. One reason why other CPUs are more
popular in the embedded market. I may not have the number exactly
right though. Checking....
From the rtlinux.org FAQ
------------------------
6. Q: How well does it [rtlinux] work?
A: Quite well. Worst-case times are about 15microseconds between the
assertion of an interrupt and the starting of the realtime handler on a
generic x86 PC, and better on the Alpha and PowerPC platforms. These times
are close to the hardware limit, yet all of the power of Linux remains
easily accessible to the realtime programmer.
Note: When researching realtime operating systems, be sure to compare
"worst case" timings. It's tempting to quote "typical" times, but unless
you only need your RT application to "typically" meet timing requirements,
this is not particularly useful.
So my number might be a bit pessimistic, it's possible that the number 40us
stuck in my mind from an application they had on the site.
-chris
This archive was generated by hypermail 2b28 : Fri Jul 13 2001 - 02:32:53 EEST