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

RE: [Xen-ia64-devel] [PATCH] [Resend]Enable hash vtlb


  • To: "Alex Williamson" <alex.williamson@xxxxxx>
  • From: "Xu, Anthony" <anthony.xu@xxxxxxxxx>
  • Date: Mon, 10 Apr 2006 22:07:03 +0800
  • Cc: xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
  • Delivery-date: Mon, 10 Apr 2006 07:07:40 -0700
  • Importance: normal
  • List-id: Discussion of the ia64 port of Xen <xen-ia64-devel.lists.xensource.com>
  • Priority: normal
  • Thread-index: AcZcXTRVhw5xRYIbRDCN2tvx7vJz/wASGCCw
  • Thread-topic: [Xen-ia64-devel] [PATCH] [Resend]Enable hash vtlb

Hi Alex,

Below data is got based on changeset 8489.

System:
        Tiger 4
        4G RAM (2GB available to xen)
        Montecito 1.4GHz dual core dual thread.
        DomU 512M RAM

bare metal (UP):
Total TimeBuild Time 2  Build Time 1
=====================================================================
4008 Second  2004 Second  1995 Second
=====================================================================

domU w/o hash vtlb
Total TimeBuild Time 2  Build Time 1  
=====================================================================
3966 Second  1976 Second  1978 Second  
=====================================================================

domU w/ hash vtlb
Total Time      Build Time1  Build Time2
=====================================================================
3959 Second  1970 Second  1975 Second  
=====================================================================


DomU can still get better performance after applying hash_vtlb patch.
This performance gain is based on 2% degradation of Dom0.(I don't get 
performance
data on Dom0 this time)

The attachment is the script which I used to get kernel build performance.
Usage Example,
./make_kernel.sh    2    /root/linux-2.6.16.tar.bz2
                  (times of build)   (absolute path)  


>From: Alex Williamson 
>Sent: 2006?4?10? 13:10
>Hi Anthony,
>
>   Thanks, I applied all the patches to my test tree and am able to
>boot.  However, I'm not able to reproduce the performance increase for
>domU.  I see a performance decrease across the board (including a
>significant increase in system time).  Here are my results:
>
>System:
>   HP zx6000
>   10GB RAM (7GB available to xen using max_addr=10G)
>   2x1.3GHz Mad6M
>
>(results below are the best time from 2 runs)
>
>bare metal (UP):
>   1829s (56.6s system) baseline
>
>dom0 w/o hash vtlb:
>   1928s (64.4s system) +5.4% (+13.7% system)
>
>domU w/o hash vtlb
>   1877s (63.8s system) +2.6% (+12.7% system)
>
>dom0 w/ hash vtlb
>   1986s (86.3s system) +8.6% (+52.5% system)
>
>domU w/ hash vtlb
>   1932s (84.2s system) +5.6% (+48.8% system)
>
>Is the performance increase reproducible on your system?  Can you
>provide the system specs?  Can anyone else verify performance with and
>without this patch set on their system?  Thanks,
>
>       Alex
>
>--
>Alex Williamson                             HP Linux & Open Source Lab

Attachment: Deleted Attachment.txt
Description: Deleted Attachment.txt

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel

 


Rackspace

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