[Xen-devel] Want to Learn how Qemu works in Xen and dom0

["Hu Jia Yi" <jyhu@xxxxxxxxx>]

Hi, gurus, 

I want to have a clear picture of how xen/dom0 works with qemu and the
physical hardware  so I searched qemu website, but only found limited
documents. Xen doesn't include detailed docs either. It only says qemu
is modified to integrate in xen. What's the modification and why they
are modified remains only in the souce code files. 

I found it is quite complicated looking for clues in the source code
files to draw a big picture. Could anybody tell me where I should start?

Best regards,
Hu Jia Yi
Ext: 20430
Tel: 65-67510430

-----Original Message-----
From: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
[mailto:xen-devel-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of
xen-devel-request@xxxxxxxxxxxxxxxxxxx
Sent: Wednesday, September 26, 2007 5:35 AM
To: xen-devel@xxxxxxxxxxxxxxxxxxx
Subject: Xen-devel Digest, Vol 31, Issue 106

Send Xen-devel mailing list submissions to
        xen-devel@xxxxxxxxxxxxxxxxxxx

To subscribe or unsubscribe via the World Wide Web, visit
        http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel
or, via email, send a message with subject or body 'help' to
        xen-devel-request@xxxxxxxxxxxxxxxxxxx

You can reach the person managing the list at
        xen-devel-owner@xxxxxxxxxxxxxxxxxxx

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Xen-devel digest..."


Today's Topics:

   1. auto-translated-physmap resume (Andres Lagar-Cavilla)
   2. Xen Kernel Debug Tools (John Anderson)
   3. Re: Report Xen-3.1.1-rc1 make install does not    install
      /etc/hotplug (Keir Fraser)
   4. RE: auto-translated-physmap resume (Ian Pratt)
   5. How do I debug the hypervisor? (Roger Cruz)
   6. RE: Xen Kernel Debug Tools (Ian Pratt)


----------------------------------------------------------------------

Message: 1
Date: Tue, 25 Sep 2007 16:25:16 -0400
From: Andres Lagar-Cavilla <andreslc@xxxxxxxxxxxxxx>
Subject: [Xen-devel] auto-translated-physmap resume
To: xen-devel@xxxxxxxxxxxxxxxxxxx
Message-ID: <46F96EAC.8040101@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Hi,
I was wondering, perhaps in my ignorance, what prevents an 
auto-translated-physmap linux guest from being suspended/resumed.
As far as I understand, there would be no need for pagetable 
canonicalization code in xc_linux_save, as well as no need for those 
mfn_to_pfn translations in pre_suspend (and viceversa in post_suspend)
Anything I am missing?

Thanks!
Andres



------------------------------

Message: 2
Date: Tue, 25 Sep 2007 13:31:51 -0700
From: "John Anderson" <johnha@xxxxxxxxxx>
Subject: [Xen-devel] Xen Kernel Debug Tools
To: <xen-devel@xxxxxxxxxxxxxxxxxxx>
Message-ID:
        
<A7F39817EC2477418A3AA053E69F835A99BD0D@xxxxxxxxxxxxxxxxxxxxxxxx>
Content-Type: text/plain; charset="us-ascii"

Greetings,

 

I'm attempting to mash grsecurity-2.1.9 into Xen 3.1.1-rc1.  Last time I
did this, for Xen 3.02, things went off without much of a hitch.
However, this time I'm running into a snag in the Dom0 kernel when a
para-virtualized  DomU starts up.   The  page fault I'm seeing in the
Xen's dmesg (xm dmesg) is below, but as it is, its not very helpful.   

 

I'm looking for a Xen tutorial that lists the tools and techniques I
need to debug kernel code, and hopefully associate this stack trace with
a source file and line number.  Does anyone know if such a tutorial
exists?  If so, can you please point me in the right direction?  I've
found some dated documents about kdb patches being necessary or some
sort of thing.  I was hoping it would be as easy as turning on debug
support, getting a crash dump, and running it through gdb, but it
doesn't look like that is the case.

 

 

Here is the page fault exception I'm getting in xm dmesg.  The Dom0
Linux kernel's dmesg gets no messages.  This happens when I attempt to
start a grsecurity enabled DomU on a grsecurity enabled Dom0.   Starting
a grsecurity enabled DomU on a non-grsecurity Dom0 works just fine.

 

(XEN) Unhandled page fault in domain 14 on VCPU 0 (ec=0010)

(XEN) Pagetable walk from ffffffff80200000:

(XEN)  L4[0x1ff] = 0000000000000000 ffffffffffffffff

(XEN) domain_crash_sync called from entry.S

(XEN) Domain 14 (vcpu#0) crashed on cpu#0:

(XEN) ----[ Xen-3.1.0  x86_64  debug=n  Not tainted ]----

(XEN) CPU:    0

(XEN) RIP:    e033:[<ffffffff80200000>]

(XEN) RFLAGS: 0000000000010202   CONTEXT: guest

(XEN) rax: 0000000000000000   rbx: 0000000000000000   rcx:
0000000000000000

(XEN) rdx: 0000000000000000   rsi: ffffffff80569000   rdi:
0000000000000000

(XEN) rbp: 0000000000000000   rsp: ffffffff80574000   r8:
0000000000000000

(XEN) r9:  0000000000000000   r10: 0000000000000000   r11:
0000000000000000

(XEN) r12: 0000000000000000   r13: 0000000000000000   r14:
0000000000000000

(XEN) r15: 0000000000000000   cr0: 000000008005003b   cr4:
00000000000006f0

(XEN) cr3: 000000041c62a000   cr2: ffffffff80200000

(XEN) ds: 0000   es: 0000   fs: 0000   gs: 0000   ss: e02b   cs: e033

(XEN) Guest stack trace from rsp=ffffffff80574000:

(XEN)   Stack empty.

 

 

Thanks,

 

John Anderson

-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xensource.com/archives/html/xen-devel/attachments/20070925/
ab01e8fb/attachment.htm

------------------------------

Message: 3
Date: Tue, 25 Sep 2007 22:16:15 +0100
From: Keir Fraser <Keir.Fraser@xxxxxxxxxxxx>
Subject: Re: [Xen-devel] Report Xen-3.1.1-rc1 make install does not
        install /etc/hotplug
To: Teck Choon Giam <giamteckchoon@xxxxxxxxx>
Cc: xen-devel@xxxxxxxxxxxxxxxxxxx
Message-ID: <C31F392F.E070%Keir.Fraser@xxxxxxxxxxxx>
Content-Type: text/plain;       charset="US-ASCII"

On 25/9/07 19:27, "Teck Choon Giam" <giamteckchoon@xxxxxxxxx> wrote:

> Thanks for your information about udevinfo.  Yes, you are right as
> this is the output shown in my test CentOS 5.0:
> # udevinfo -V
> udevinfo, version 095
> 
> and in CentOS 4.5 it is:
> # udevinfo -V
> udevinfo, version 039
> 
> So would this means that for any udevinfo version greater than 059, I
> have to manually create /etc/hotplug folder and copy xen-backend.agent
> into it manually for future releases for Xen i.e. version 3.1.1?

The question is really why your system doesn't pick up the udev config
files
we install, and work properly with those. I wouldn't really expect it to
need the hotplug config file.

 -- Keir





------------------------------

Message: 4
Date: Tue, 25 Sep 2007 22:14:59 +0100
From: "Ian Pratt" <Ian.Pratt@xxxxxxxxxxxx>
Subject: RE: [Xen-devel] auto-translated-physmap resume
To: "Andres Lagar-Cavilla" <andreslc@xxxxxxxxxxxxxx>,
        <xen-devel@xxxxxxxxxxxxxxxxxxx>
Message-ID:
        
<8A87A9A84C201449A0C56B728ACF491E26083B@xxxxxxxxxxxxxxxxxxxxxxxxxxx>
Content-Type: text/plain;       charset="us-ascii"

> I was wondering, perhaps in my ignorance, what prevents an
> auto-translated-physmap linux guest from being suspended/resumed.
> As far as I understand, there would be no need for pagetable
> canonicalization code in xc_linux_save, as well as no need for those
> mfn_to_pfn translations in pre_suspend (and viceversa in post_suspend)
> Anything I am missing?

That's about it -- there's less to do if you're using shadow pagetables.
Since auto-translated-physmap is typically only used by various research
projects its not particularly well maintained, but its usually easy to
fix up.

Ian







------------------------------

Message: 5
Date: Tue, 25 Sep 2007 17:31:01 -0400
From: "Roger Cruz" <rcruz@xxxxxxxxxxxxxxxxxxxxxxxx>
Subject: [Xen-devel] How do I debug the hypervisor?
To: "xen-devel" <xen-devel@xxxxxxxxxxxxxxxxxxx>
Message-ID:
        
<40B551BEDC7945419A5897958AB3947C099207@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>
        
Content-Type: text/plain;       charset="us-ascii"


Just posted this in XenSource's Misc. forums, but I figure this may be a
more appropriate channel to ask the following question.

I'm using XenEnterprise 4.0 (hypervisor is based off 3.1.0) and would
like to make changes to the hypervisor.  I need to find out how to hook
up a debugger to it so I can step through the code and examine its
state.

I would appreciate a detail response on how to configure the hypervisor,
hook up GDB and have access to symbol information.

Thank you.
Roger Cruz




------------------------------

Message: 6
Date: Tue, 25 Sep 2007 22:33:11 +0100
From: "Ian Pratt" <Ian.Pratt@xxxxxxxxxxxx>
Subject: RE: [Xen-devel] Xen Kernel Debug Tools
To: "John Anderson" <johnha@xxxxxxxxxx>,
        <xen-devel@xxxxxxxxxxxxxxxxxxx>
Message-ID:
        
<8A87A9A84C201449A0C56B728ACF491E26083F@xxxxxxxxxxxxxxxxxxxxxxxxxxx>
Content-Type: text/plain;       charset="us-ascii"

> Here is the page fault exception I'm getting in xm dmesg.  The Dom0
> Linux kernel's dmesg gets no messages.  This happens when I attempt to
> start a grsecurity enabled DomU on a grsecurity enabled Dom0.
> Starting a grsecurity enabled DomU on a non-grsecurity Dom0 works just
> fine.

This is pretty surprising. When a domU is actually running, dom0 isn't
really involved (other than for IO), so its surprising grsec makes a
difference.

Do you get any console output from the guest before it crashes? I'm
wandering if its actually been built incorrectly by the domain builder
running in dom0.

It's looks like its dieing trying to do an instruction fetch from
ffffffff80200000 which is presumably the page fault handler. 

Are you using a debug build of Xen? You may get some more helpful
output.

Ian

 
> 
> 
> (XEN) Unhandled page fault in domain 14 on VCPU 0 (ec=0010)
> 
> (XEN) Pagetable walk from ffffffff80200000:
> 
> (XEN)  L4[0x1ff] = 0000000000000000 ffffffffffffffff
> 
> (XEN) domain_crash_sync called from entry.S
> 
> (XEN) Domain 14 (vcpu#0) crashed on cpu#0:
> 
> (XEN) ----[ Xen-3.1.0  x86_64  debug=n  Not tainted ]----
> 
> (XEN) CPU:    0
> 
> (XEN) RIP:    e033:[<ffffffff80200000>]
> 
> (XEN) RFLAGS: 0000000000010202   CONTEXT: guest
> 
> (XEN) rax: 0000000000000000   rbx: 0000000000000000   rcx:
> 0000000000000000
> 
> (XEN) rdx: 0000000000000000   rsi: ffffffff80569000   rdi:
> 0000000000000000
> 
> (XEN) rbp: 0000000000000000   rsp: ffffffff80574000   r8:
> 0000000000000000
> 
> (XEN) r9:  0000000000000000   r10: 0000000000000000   r11:
> 0000000000000000
> 
> (XEN) r12: 0000000000000000   r13: 0000000000000000   r14:
> 0000000000000000
> 
> (XEN) r15: 0000000000000000   cr0: 000000008005003b   cr4:
> 00000000000006f0
> 
> (XEN) cr3: 000000041c62a000   cr2: ffffffff80200000
> 
> (XEN) ds: 0000   es: 0000   fs: 0000   gs: 0000   ss: e02b   cs: e033
> 
> (XEN) Guest stack trace from rsp=ffffffff80574000:
> 
> (XEN)   Stack empty.
> 
> 
> 
> 
> 
> Thanks,
> 
> 
> 
> John Anderson




------------------------------

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


End of Xen-devel Digest, Vol 31, Issue 106
******************************************

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