Re: Context Switches
From: Nils Petter Vaskinn (no_at_spam.for.me.invalid)
Date: 02/05/04
- Previous message: Villy Kruse: "Re: Thoughts on Semaphores"
- In reply to: Brian Cummings: "Re: Context Switches"
- Next in thread: Barry Margolin: "Re: Context Switches"
- Reply: Barry Margolin: "Re: Context Switches"
- Reply: Brian Cummings: "Re: Context Switches"
- Reply: Nick Landsberg: "Re: Context Switches"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Thu, 05 Feb 2004 12:02:51 GMT
On Thu, 05 Feb 2004 11:12:54 +0000, Brian Cummings wrote:
> I see, but if I was preempted by a higher priority process, wouldn't
> that be an involuntary context switch? I say this because of the two
> fields specified below:
>
> I believe I forgot to mention that the context switches I am looking
> into are the counts returned by ru_nvcsw (voluntary).
Is the number constant?
Does removing the fprintf()-s affect it?
Can anyone tell me if there is a context switch involved in moving from
userland to kernel space? (as in when doing the write syscall)
--
NPV
"the large print giveth, and the small print taketh away"
Tom Waits - Step right up
- Previous message: Villy Kruse: "Re: Thoughts on Semaphores"
- In reply to: Brian Cummings: "Re: Context Switches"
- Next in thread: Barry Margolin: "Re: Context Switches"
- Reply: Barry Margolin: "Re: Context Switches"
- Reply: Brian Cummings: "Re: Context Switches"
- Reply: Nick Landsberg: "Re: Context Switches"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]