similar to: [PATCH v2] xen: sched_credit: filter node-affinity mask against online cpus

Hello, Can anyone explain the need/purpose for the following variables in the structure csched_dom found in sched_credit.c file struct csched_dom { *struct list_head active_vcpu; struct list_head active_sdom_elem; struct domain *dom;* . . Thank you _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

At least to me, using ?: instead of the (a && ...) || (!a && ...) construct is far easier to grok with a single look. Signed-off-by: Jan Beulich <jbeulich@novell.com> --- a/xen/common/sched_credit.c +++ b/xen/common/sched_credit.c @@ -526,10 +526,9 @@ _csched_cpu_pick(const struct scheduler weight_cpu = cpus_weight(cpu_idlers); weight_nxt =

Hi, I was trying to add some printk in the credit scheduler to see how it works, but the printk seems not working. (it gives no output) But when I add printk in scheduler.c, it works pretty well. Does anyone know why? do I need to change some configuration file or something? Thanks very much! Best! Yours, Sisu _______________________________________________ Xen-devel mailing list

From: Nate Studer <nate.studer@dornerworks.com> Add a utility function to update the rest of the timeslice accounting fields when updating the timeslice of the credit scheduler, so that capped CPUs behave correctly. Before this patch changing the timeslice to a value higher than the default would result in a domain not utilizing its full capacity and changing the timeslice to a value lower

Hello, This small series deals with some weirdness in the mechanism with which the credit scheduler choses what PCPU to tickle upon a VCPU wake-up. Details are available in the changelog of the first patch. The new approach has been extensively benchmarked and proved itself either beneficial or harmless. That means it does not introduce any significant amount of overhead and/or performances

Add a xen command-line parameter, sched_credit_tslice_ms, to set the timeslice of the credit1 scheduler. Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com> diff -r 4a4882df5649 -r 782284c5b1bc xen/common/sched_credit.c --- a/xen/common/sched_credit.c Wed Aug 31 15:23:49 2011 +0100 +++ b/xen/common/sched_credit.c Thu Sep 01 16:29:50 2011 +0100 @@ -41,15 +41,9 @@ */ #define

Signed-off-by: Vincent Hanquez <vincent.hanquez@eu.citrix.com> --- tools/ocaml/Makefile | 2 +- tools/ocaml/common.make | 2 +- tools/ocaml/libs/xl/xl.ml | 209 ++++++++++++ tools/ocaml/libs/xl/xl.mli | 209 ++++++++++++ tools/ocaml/libs/xl/xl_stubs.c | 706 ++++++++++++++++++++++++++++++++++++++++ 5 files changed, 1126 insertions(+), 2 deletions(-)

Hi all, Since Xen 4.2.0, users can change time slice of the scheduler at runtime via xl command line, a very nice feature. However, it is not *correctly* implemented. Problem description -------------------- say you set the ''cap'' of one VM to 50 (a half core), -when setting the time slice to be *greater* than 30ms, the VM gets much *less* CPU cycles than its allocation -when

When in SYS_STATE_suspend, and going through the cpu_disable_scheduler path, save a copy of the current cpu affinity, and mark a flag to restore it later. Later, in the resume process, when enabling nonboot cpus restore these affinities. v2: Fix formatting problems. remove early return in cpu_disable_scheduler() path. v3: Fix remaining errant tab Move restore_vcpu_affinity() to thaw_domains(),

On Wed, Nov 23, 2011 at 2:55 PM, Dario Faggioli <raistlin@linux.it> wrote: > Hi everyone, > > This series changes how locks are dealt with while adjusting domains'' > scheduling parameters. > > I''ve done and am still doing tests for credit and credit2, and it''s > surviving to all I threw at it up to now. Unfortunately, I can''t test >

Hi all, I am trying to set processor affinity for a specific process using a shell script without result. Script: #!/bin/sh -x cpu_affinity_ok="2" cpu_affinity="taskset -p -c `cat /tmp/test.pid` | awk '{print $6}'" if [ -f /tmp/test.pid ]; then if [ "$cpu_affinity" == "$cpu_affinity_ok" ]; then exit 0 else taskset -p -c 2

================== STATUS ================== I do the heavy load test of create/destroy. CREDIT scheduler(cshed_schedule) checks BUG_ON(!vcpu_running) at the end of code. It makes error. The reason is that atomic_inc(&v->pausecnt)@vcpu_pause() is called without lock. (spin_lock(&schedule_data[cpu].schedule_lock)) This lock-less "pausecnt" makes vcpu_running state changing

As discussed in a previous e-mail, this patch series implements yield for the credit scheduler. This allows a VM to actually yield (give up the cpu to another VM) when it wants to. This has been shown to be effective when used in the spinlock code to avoid wasting time spinning when another vcpu is not currently scheduled. _______________________________________________ Xen-devel mailing list

Hi, I have the following errors: Panic on CPU0: FATAL PAGE FAULT [error_code=0002] Faulting linear address: fbfb1100 More details in snapshots as below as I don''t have serial console cables thus unable to get full details of the errors which I will try to get one and send the full error if needed: http://choon.net/xenerrors/Image000.jpg http://choon.net/xenerrors/Image001.jpg

Enable cpufreq support in Xen for AMD Operton processors by: 1) Allowing the PowerNow! driver in dom0 to write to the PowerNow! MSRs. 2) Adding the cpufreq notifier chain to time-xen.c in dom0. On a frequency change, a platform hypercall is performed to scale the frequency multiplier in the hypervisor. 3) Adding a platform hypercall to the hypervisor the scale the frequency multiplier and reset

Address some credit2 issues. This patch series, along with the recent changes to the cpupools interface, should address some of the strange credit2 instability. _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

Hi, I have some problems to configure the xen I''ve installed (3.3.1). The computer is a Intel Core 2 Duo, I''m using Ubuntu 8.10 and have linux in my dom0 and winxp pro in my domU. I have two cores and I''d like to set the affinity of dom0 to cpu0 and domU to cpu1 but I haven''t find the way of making this permanent. I''ve set cpus=1 in the domU config

Hi Everyone, I am running 3.2.1 on Centos 5.2 with HAP enabled, NUMA enabled, ACPI enabled and the dom0 allocated 512Mb. I have setup a single core 1Gb VM for performance testing under Windows 2008 Server. Most CPU results are within a few percent of theoretical max but Memory performance is about half what I expected. I get 3.22Gb/Sec Sandra 2009 Memory performance for a single Opteron 8350

Hi, This patch intends to accurate vcpu weighting for CPU intensive job. The reason of this problem is that vcpu round-robin queue blocks large weight vcpus by small weight vcpus. For example, we assume following case on 2pcpu environment. (with 4domains (each domain has 2vcpus)) dom1 vcpu0,1 w128 credit 4 dom2 vcpu0,1 w128 credit 4 dom3 vcpu0,1 w256 credit 8 dom4 vcpu0,1 w512 credit 15

Hi Keir, The CS# 17131 which I write for bind guest to NUMA node via cpu affinity missed one condition existing in some machines, where there aren''t any cpus but only memories. Under this condition it will fail to set cpu_affinity because of none parameter. I cope with this condition in the new patch and make a little change of the methods to find suitable node to bind guest. When