From: Naresh Rapolu <nrapolu@purdue.edu>
To: xen-devel@lists.xensource.com
Subject: Re: boot_cpu_data differs between Xen hypervisor boot and normal pv_ops kernel boot
Date: Wed, 31 Mar 2010 13:51:34 -0400 [thread overview]
Message-ID: <4BB38BA6.9070705@purdue.edu> (raw)
In-Reply-To: <20100331154340.GS1878@reaktio.net>
Thanks Pasi,
But I am concerned about the Performance Monitoring Unit (PMU). Not
sure if it has any link to power mgmt unit. Xen while booting believes
Xeon 5530 is in the P6 category and hence does not detect the Intel PMU
driver, thereby hindering Oprofile to use the performance counters.
This does not happen if I boot the kernel normally without Xen.
In 2.6.31.12 , the file for the same is
arch/x86/kernel/cpu/perf_counters.c , but dmesg doesnt show any
printk messages from this file when I boot with xen-3.4.2. Does Xen
even consider setting up the performance counters at boot time ? Did
anyone successfully use Oprofile in 2.6.31 kernels for hardware
performance events ( cache-misses ) etc ?
Iam using ubuntu 8.04 LTS... Does it depend on the OS ? Can someone
familiar with the boot code help me please !
Thanks,
Naresh Rapolu.
Pasi Kärkkäinen wrote:
> On Wed, Mar 31, 2010 at 11:29:04AM -0400, Naresh Rapolu wrote:
>
>> Hello,
>>
>> After booting the xen-stable 2.6.32.10 pv_ops kernel from jeremy git
>> tree with Xen-4.0.0-rc8 installed,
>> dmesg | grep PMU shows :
>>
>> Performance Events : unsupported p6 CPU model 26 no PMU driver, software
>> events only.
>> PMU:oprofile: found i386/core_i7.
>>
>> If I boot the same compiled kernel, without Xen installed, dmesg | grep
>> PMU shows :
>> Performance Events : Nehalem/Corei7 events, Intel PMU driver.
>> PMU:oprofile: found i386/core_i7.
>>
>>
>> When I looked into the source code, arch/x86/kernel/cpu/perf_event.c,
>> at run-time boot_cpu_data structure seems to differ in these two forms
>> of booting.
>> Due to this, Iam not able to use the hardware performance counters in
>> my Xeon 5530 processor as hardware events(cache-misses etc) are being
>> disabled.
>>
>> Is there any reason for this ?
>>
>>
>
> I think only xen/stable-2.6.31.x tree currently has power mgmt stuff included.
>
> -- Pasi
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
>
next prev parent reply other threads:[~2010-03-31 17:51 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-03-31 15:29 boot_cpu_data differs between Xen hypervisor boot and normal pv_ops kernel boot Naresh Rapolu
2010-03-31 15:43 ` Pasi Kärkkäinen
2010-03-31 17:51 ` Naresh Rapolu [this message]
2010-03-31 18:14 ` Jeremy Fitzhardinge
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4BB38BA6.9070705@purdue.edu \
--to=nrapolu@purdue.edu \
--cc=xen-devel@lists.xensource.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.