Messages in this thread | | | Subject | Re: interrupt latency | Date | Tue, 11 Aug 1998 10:01:43 +0100 | From | Philip Blundell <> |
| |
>I reall wish this wasn't the case because I would dearly love to be able to >have interrupt latency as a benchmark test in lmbench. Are you sure there >is no way to do this in software? What if you had some way to generate >interrupts essentially continuously?
I can't think of any way it could be done, certainly not sufficiently portably to be of use to you in lmbench. There are two components to the interrupt latency that the hardware sees; firstly the latency inherent in the interrupt-handling code (how long does it take to get in, service the IRQ and get out again) and secondly the extra latency induced by IRQs having been disabled at the instant the interrupt occurred.
The former might be measurable in software and is certainly an interesting statistic in its own right, but for people interested in real-time work (or even device drivers under some circumstances) the second measurement is important also; possibly more so because it's a lot harder to put a bound on it. And without using special hardware or instrumenting the kernel, I can't think how you could measure this part.
p.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html
| |