Hi all, A little note from the linux-kernel list. See http://kpreempt.sourceforge.net/ for description. This is the low-latency work that is most likely to bee included in linux-2.5 (Note the spikes some part of the code is holding a spin-lock for too long... But the nice part is that - to get good SMP performance they will have to go!) /RogerL ---------- Forwarded Message ---------- Subject: Updated Linux 2.4.9/2.4.10 kernel preemption patches Date: 29 Aug 2001 01:35:26 -0400 From: Robert Love <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED] Updated patches are at: http://tech9.net/rml/linux/patch-rml-2.4.9-ac3-preempt-kernel-1 and, http://tech9.net/rml/linux/patch-rml-2.4.10-pre2-preempt-kernel-1 for kernels 2.4.9-ac3 and 2.4.10-pre2. These are updates of Nigel Gamble's kernel preemption patches for recent kernels. See http://kpreempt.sourceforge.net/. These patches create a configure option to enable a preemptible kernel using SMP lock points. A preemptible kernel will yield control of execution to higher priority processes as needed. Ie, the process timeslice now applies to kernel space. Changes since my previous patch: * update for 2.4.9-ac3 and 2.4.10-pre2 * fix the compile bug (yay!) -- the linking dependency of dec_and_lock requires CONFIG_HAVE_DEC_LOCK which SMP sets in recent kernels. now CONFIG_PREEMPT sets, too So, yes, this should fix the kernel compile buggy. At least it did for me, after I was finally able to reproduce the problem. Enjoy and please comment, test, and benchmark. -- Robert M. Love rml at ufl.edu rml at tech9.net - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ ------------------------------------------------------- -- Roger Larsson Skellefte� Sweden
