|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Xen-devel] trying to trace xm create
sting sting wrote: Hello, I am trying ro trace and study a run of "xm create xmexample1" , after modifying xmexample1 a bit (to use ttylinux,etc). I have xen-2.0-testing. The stack I can see (by adding printing messages) is :main (xm/create.py) SrvDaemon is the xend main program. The HTTP server is started from code in SrvRoot, and resources are registered for various pieces of the xend functionality. Overall domain stuff is in SrvDomainDir, individual domains in SrvDomain. The implementation, below the HTTP api, is in XendDomain and XendDomainInfo. The devices have their own classes: blkif.py: block devices netif.py: network interfaces console.py: console usbif.py: usb devices. The 'segment not found' message is coming from inside the blkif code. Basically it means the disk you're trying to use doesn't exist. If you stop xend and restart with 'xend trace_start' it will trace its execution to /var/log/xend.trace. And more important: where does this create domain request reach ? I assume it does reach a "C" code and I assume it reaches xc_domain_create() method from tools\ libxc\xc_domain.c ; am I right ? Yes, domain create does call xc_domain_create(), but that's just the start of it. The main work in creating a domain is setting up its devices, which involves a lot of messaging between the device code in xen, the front-end device drivers in the domain and the back-end device drivers in backend domains. Regards, Mike _______________________________________________ Xen-devel mailing list Xen-devel@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-devel
|
![]() |
Lists.xenproject.org is hosted with RackSpace, monitoring our |