[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-users] R-Car M3 using Xen 4.12 with two DomU having a shared pinned pCPU


  • To: "xen-users@xxxxxxxxxxxxxxxxxxxx" <xen-users@xxxxxxxxxxxxxxxxxxxx>
  • From: LOPEZJAIRO / LOPEZ,JAIRO <jairo.lopez.uh@xxxxxxxxxxx>
  • Date: Fri, 22 Feb 2019 11:46:27 +0000
  • Accept-language: ja-JP, en-US
  • Deferred-delivery: Fri, 22 Feb 2019 11:48:00 +0000
  • Delivery-date: Fri, 22 Feb 2019 11:48:51 +0000
  • List-id: Xen user discussion <xen-users.lists.xenproject.org>
  • Thread-index: AdTKpDyibgesk/iSTlulc7ZZK1QGiw==
  • Thread-topic: R-Car M3 using Xen 4.12 with two DomU having a shared pinned pCPU

Hello all,

 

I am using Xen 4.12.0-rc3 commit 365aabb6e5023cee476adf81106729efd49c644f on an R-Car M3 to boot two DomU that have two vCPUs each.

 

The DomU are loaded on a CPU pool with 3 Arm Cortex-A53 pCPUs using the Credit2 scheduler.

 

The first DomU is hard pinning vCPU0 and vCPU1 on CPUs 3 and 4 of the R-Car M3 while the second DomU is pinning vCPU0 and vCPU1 on CPUs 5 and 4 respectively.

 

After reading the documentation, I thought this sort of configuration was possible and that the CPU weight would determine which DomU gets more time on pCPU 4.

 

Much to my surprise, when I load 4 memory read intensive tasks on each vCPU, both my DomU are killed without any apparent sort of error.

 

I am new to Xen, so there is clearly something I am not understanding.

 

The main question would be if having two DomU both pin a vCPU to the same pCPU is possible. If it is possible is there any way I can debug as to why my DomU are being scrapped?

 

I thank everyone for their time.

 

Best regards,

 

Jairo

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-users

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.