On 20/11/05, thewade <pdman@xxxxxxxxxxxxxxxx> wrote: > A (in no way) comprehensive search yields that rtirq is a patch, possibly only to > the 2.4 kernel. >From what I read, ccrma has a 2.6 kernel patched for realtime low latency. This patch makes the kernel's interrupt handlers into separate visible threads, which can then be given different scheduling priorities. rtirq is a script which automates interrupt handler prioritization optimised for audio use. Basically, this does in software what you want to do in hardware, it seems. (From reading http://lac.zkm.de/2005/papers/fernando_lopez_lezcano.pdf - I've only skimmed the surface of this, I'm sure there are people here far more knowledgable than I am on the subject! ;-) > 63-0100-000001-00101111-071504-SiS755-258KA000-Y2KC, Uniwill 258ka0 (aka fujitsu-siemens Amilo A 1630, and others) ? > If I cannot do this I might find where my modem is enabled or where > drivers are built for it in the kernel and rebuild the kernel without thoes > drivers. My own preference would be kernel tinkering over bios tinkering, *by far* because it's far safer - especially when the exact bios to use is in doubt (if indeed there is one with the capabilities you want). Loading the wrong bios can be a non-recoverable error! - Pete.
