You have a point there. I'll flesh out my idea:
The kernel distinguishes between realtime processes and "other" processes. When a processor quantum needs to be assigned to a process, realtime processes get the first chance to take it. The realtime scheduler can be choosen by a kernel configuration. If they don't need it the quantum is offered to the "other" processes. A kernel configuration defines a sequence of process categories, each with their own scheduler and a percentage of the available processor time for each category. When time quantums become available they are divided amongst the categories by the specified percentage. In the case a category does not use all of its quantums, the quantums are assigned to the first category in the sequence that has no quantums left. This way: 1. Realtime processes get absolute priority (like they should) 2. Processes can be scheduled using different schedulers, based on their needs 3. We can assign priorities to schedulers 4. Processor time not being used by a scheduler is reassigned to the highest scheduler that will make good use of it What do you think? -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Sander van Rossen Sent: dinsdag 11 september 2007 21:01 To: sharpos-developers@lists.sourceforge.net Subject: Re: [SharpOS Developers] Schedulers, threads,processes and interfaces On 9/11/07, Pieter Mees <[EMAIL PROTECTED]> wrote: > Well, I guess a scheduler should just do the following: <snip> Yes.. but what would schedule the schedulers for each "set of processes"? ------------------------------------------------------------------------ - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ SharpOS-Developers mailing list SharpOS-Developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sharpos-developers ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ SharpOS-Developers mailing list SharpOS-Developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sharpos-developers
