public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Anthony Liguori <anthony@codemonkey.ws>
To: Lukas Kolbe <l-lists@einfachkaffee.de>
Cc: Avi Kivity <avi@redhat.com>, kvm@vger.kernel.org
Subject: Re: kvm-84 and guests with more than 3536 MB Ram?
Date: Tue, 24 Mar 2009 10:45:22 -0500	[thread overview]
Message-ID: <49C90012.1020700@codemonkey.ws> (raw)
In-Reply-To: <1237906054.18072.8.camel@vesuv.TechFak.Uni-Bielefeld.DE>

Lukas Kolbe wrote:
> On Di, 2009-03-24 at 14:59 +0200, Avi Kivity wrote:
>   
>> Lukas Kolbe wrote:
>>     
>>> Hi!
>>>
>>> This is my first post here so please bear with me; we have a Debian
>>> Lenny system with kernel 2.6.28 and kvm-84, and can't start a guest with
>>> more than 3536 MB Ram. With kvm-72 (the version lenny released with) we
>>> can use all 7GB that is intended for that guest.
>>>
>>> Is this a known bug? Someone on freenode #kvm said he'd seen something
>>> like this but couldn't remember where or when.
>>>
>>>   
>>>       
>> Can you describe the failure?  I just booted a guest (FC6) with 6GB, 
>> without issue (admittedly the guest is only using 300MB or so).
>>     
>
> Well, when I start the guest like this:
>
> kvm -smp 8 -drive if=virtio,file=/dev/system/test_root,boot=on -m 3585 \
> -nographic -name test -kernel /boot/kvm/test/vmlinuz-2.6.28-1-amd64 \
> -initrd /boot/kvm/test/initrd.img-2.6.28-1-amd64 \
> -append 'root=/dev/vda ro console=ttyS0,115200' \
> -serial mon:unix:/etc/kvm/consoles/test.sock,server,nowait \
>
> I get the following output:
>
> qemu: loading initrd (0x781b93 bytes) at 0x000000007f87e000
> create_userspace_phys_mem: Invalid argument
> kvm_cpu_register_physical_memory: failed
>
> And back to the console. When I try the same with 3584MB, I can boot
> into the machine flawlessly.
>
> Sorry for getting the numbers wrong in the first mail - the actual
> problem starts at 3585MB Ram for the guest.
>
> If you can't reproduce it with yout 2.6.28 and kvm-84, I should possibly
> take this to the debian bugtracker ...
>   

kvm-72 is pretty old.  It used to be that we used phys_ram_base for 
loading kernel/initrds which would break when using > 3.5GB of memory.  
I wouldn't be surprised if that fix happened post kvm-72.

Regards,

Anthony Liguori


  reply	other threads:[~2009-03-24 15:45 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-24 12:22 kvm-84 and guests with more than 3536 MB Ram? Lukas Kolbe
2009-03-24 12:59 ` Avi Kivity
2009-03-24 14:47   ` Lukas Kolbe
2009-03-24 15:45     ` Anthony Liguori [this message]
2009-03-24 16:04       ` Tomasz Chmielewski
2009-03-24 18:07         ` Lukas Kolbe
2009-03-24 20:40           ` Ryan Harper
2009-03-24 21:02             ` Lukas Kolbe
2009-03-24 21:27               ` Ryan Harper
2009-03-25 19:25                 ` Lukas Kolbe

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=49C90012.1020700@codemonkey.ws \
    --to=anthony@codemonkey.ws \
    --cc=avi@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=l-lists@einfachkaffee.de \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox