Gene Heskett wrote: > Knowing that whatever we hide behind the isolcpus= argument can't be seen > by top or other so-called "system" monitors, and that they might decide to > power down the parport, or a pci-e card as the case may be while Stewarts > Cinci's x table is doing a rapid at 400 ipm, which would not allow a > graceful stop to be done, I objected. Then they changed the argument that > it was only to prevent meltdowns. I still asked, and got no response for > an advanced warning of perhaps .5 seconds so we could at least initiate a > stop. So the patches went in, but if we stay with an x86 version on linux, > and decide to see if we can RTAI or whatever the latest 3.12+ kernel is by > the time we've hit 3.1, that there will be that possibility, and that it > can be compiled out, but will be on by default. > This is why you should have a watchdog timer on hardware motion controllers. By 3.12, we may not be depending on X86 platforms as much. If this is all moderated by OS drivers, then we will have these features disabled in the distribution, if possible, or just mark those motherboards as non-working.
Jon ------------------------------------------------------------------------------ October Webinars: Code for Performance Free Intel webinars can help you accelerate application performance. Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from the latest Intel processors and coprocessors. See abstracts and register > http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
