On Sun, Mar 10, 2002 at 11:33:56AM +0100, Robert Schwebel wrote: > On Sat, 9 Mar 2002 [EMAIL PROTECTED] wrote: > > In any case, who cares? Latencies are latencies. The good thing about > > the RTLinux scheduler and timer tests are that they show delays that > > application code will actually experience. > > ... and that's what the RTAI tests do as well. Where's your problem?
If I recall correctly, you were telling me that these tests of yours detected something that the tests we've had forever did not test. Now you say they do the same thing. I'm lost. > > > If your robot tosses a 300Kilo steel bar through the wall, a "buslock" > > test that didn't account for such things as a stupid hardware design > > where video memory was put in main memory won't do you much good. > > Thanks for your very constructive comment. > > The idea behind the buslock test is that we observe modern x86 > derivatives¹ such as the Geodes that show extraordinary long single delays > which seem to be hardware introduced - but nevertheless the BIOS seems to > influence the performance, so there is a possibility to do something with > better initialisation of the chipset. However, what's wrong with one more > handy tool to find out what's going on with the system? It's no > replacement for the normal latency test... I'm fascinated to see you back on our list. --------------------------------------------------------- Victor Yodaiken Finite State Machine Labs: The RTLinux Company. www.fsmlabs.com www.rtlinux.com -- [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/
