Re: boundary crossing latencies
Re: boundary crossing latencies
- Subject: Re: boundary crossing latencies
- From: Andrew Gallatin <email@hidden>
- Date: Fri, 25 Jun 2004 09:05:39 -0400 (EDT)
Andrew Gallatin writes:
>
>
Platform null (us/call) ioctl(us/call) user client(us/call)
>
-------- ----------- -------- -------------------
>
G5/linux2.6.5 0.2112 0.6679 N/A
>
G5/OSX 10.3.3 1.4568 2.5152 4.7201
In single user-mode (nvram 'boot-args= -s'), the null syscall latency
remains about the same. The ioctl latency actually does drop a tiny
bit to a little more than 2.40us. I forgot the copy the
SimleUserClient kext to a local disk, so I did not test that..
I first entered single user mode via kill -TERM init. This led
to a doubling of the null and ioctl latencies. Its not important
at all, but I'm curious why this was. Some problem caused by
mach_init going away?
Drew
_______________________________________________
darwin-kernel mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/darwin-kernel
Do not post admin requests to the list. They will be ignored.