Hi, > >i tried to measure the interrupt response time. > >i am getting best case is 30microseconds and worst case is > >150microsenconds. > >couldn't understand why this much difference. > >can anybody explain? or is there any thing wrong in my port? > > With RTAI I measured sometime ago on a TQM860L, 50MHz, > 8KB data cache: > Min Max Average: > No load : 13us 46us 14us > while : ; do a=a ; done: 13us 46us 18us > ping -f to target : 13us 46us 19us > > If I remember correctly, rtlinux has bigger overhead (more code).
The documentation on RTLinux says that RTAI is the more bigger/slower/advanced one - but this may be pure propaganda? > BTW: Denx has dropped support for rtlinux quite a while ago > because of fsmlabs license policy (AFAIK, there is no open rtlinux > for the MPC 8xx). "open" = there is no possibility to use RTLinux on an MPC8xx with the GNU GPL license? As I understood I could choose to use the GPL for my programs, OR, make my program as a module, and thus choose the license I want (fx. proprietary if I want). Further more, I was under the impression that RTLinux was released under the GNU GPL and that this mean that IF I can get RTLinux to work with MPC823 I am free to do so? FMLabs, please clear this up, it seems kinda messy to me - is your code GPL or not, can I use it with the processor (and kernel) I want to or can't I? > Hope it helps, Thanks for your timing-results :) Did you get RTLinux to work with MPC823? Have you continued to support RTAI, and is it working for you? Regards, David Marqvar -- [rtl] --- To unsubscribe: echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED] -- For more information on Real-Time Linux see: http://www.rtlinux.org/
