Dino Del Favero wrote: > On Mon, Jun 18, 2007 at 12:35:30PM -0500, Jon Elson wrote: > >>The kernel framebuffer sounds like it (previous message) but if >>that doesn't fix it, the Nvidia ought to work. Some S3 boards > > I have installed the nVidia card, loaded the rtai_smi module, make the test > and the ovl_max is 17321 with glxgears running and moving some window!!! > I think now it is all ok... but when run emc I have an "Unexpected realtime > delay"!!! > The base-thread is set to 20000 nsec. > [ 7939.458856] 288243: ERROR: Unexpected realtime delay: check dmesg for > details. > [ 7939.458879] > [ 7939.458881] In recent history there were > [ 7939.458893] 286620, 288548, 288640, 288060, and 289784 > [ 7939.458895] elapsed clocks between calls to the motion controller. > [ 7939.458911] This time, there were 345192 which is so anomalously This doesn't sound like so large a problem as before. It may be acceptable, depending on the motor drives, etc. that you will be using. This is a 20% worst-case fluctuation, which shouldn't be enough to mess up the real-time predictability. If you are dedicating 80% of your CPU cycles to base thread, you are too close to the edge. You either need a faster CPU or increase base thread to a safer value.
Jon ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
