[Xen-devel] [9fans] running plan9 on xen. (fwd)

["Ronald G. Minnich" <rminnich@xxxxxxxx>]

here is the note I sent to 9fans. 

The '9fs sources' mounts a file system from a machine at bell labs. So the 
question is, how do I get these files to you all? 

You will need to get a program called 'drawterm' to talk to the xen 
instance -- get back to me if you don't know how to do this.

The plan 9 network comes up as 17.16.233.201, so to talk to it I do this:
ifconfig xen-br0 down
brctl del xen-br0
ifconfig vif1.0 172.16.233.200

and it all works.
ron

---------- Forwarded message ----------
Date: Mon, 27 Dec 2004 19:47:38 -0700 (MST)
From: Ronald G. Minnich <rminnich@xxxxxxxx>
Reply-To: Fans of the OS Plan 9 from Bell Labs <9fans@xxxxxxxxxxx>
To: 9fans@xxxxxxxxxxx
Subject: [9fans] running plan9 on xen. 

If you have gotten xen-2.0-testing, and installed it, you can now try out
plan 9 on xen.

9fs sources
cd /n/sources/xen
There are three files: 
9pccpu (the kernel you will boot)
kfs_root_image.gz (gzip'ed root image for kfs -- needs 4 GB)
etc_xen_plan9 (config file to copy to /etc/xen/plan9)

Boot the machine into xen. 
Find a way to get those files to your machine. 

copy etc_xen_plan9 to /etc/xen/plan9
copy 9pccpu and kfs_root_image.gz wherever you want. 
cd to that directory, gunzip kfs_root_image.gz

Here's what happens next.
(note: lines starting with q: are my linux command prompt, -- is my 
explanation)

-- start xen, do an xm list to make sure we're there.
q:~ # xend start
q:~ # xm list
Name              Id  Mem(MB)  CPU  State  Time(s)  Console
Domain-0           0      251    0  r----    384.2        

-- set up /dev/loop0 as the root file system for xen. 
q:~ # losetup /dev/loop0 kfs_root_image

-- start up xen, specify the config file (-f), tell it you want to 
-- auto-connect to the console (-c) and set the vmid (vmid=12)
q: # xm create -f /etc/xen/plan9 -c vmid=12
Using config file "/etc/xen/plan9".
Started domain VM12, console on port 9603
************ REMOTE CONSOLE: CTRL-] TO QUIT ********
root is from (local, tcp, il)[local]: XENFE:find_dev_by_handle 0 np 80260854 
np->dev 806c4630
There's still a leak in vif_disconnect
CONNECT: message sent. Set something to 0 just to see
CONNECT: set it
XENFE:find_dev_by_handle 0 np 80260854 np->dev 806c4630

-- note that the informational crap hides the 'root is from' prompt. 
-- note that tcp!whatever has been tested and works, even to u9fs, 
-- but just go ahead and hit return for 'local'. 

-- I have jiggered the boot process to start an rc instance. 

try /bin/rc
try /386/bin/rc
try /boot/rc

-- /boot/disco is an rc script that setups up partitions. Note, 
-- there is no load here!
-- So run the /boot/disco script with '. /boot/disco'
% # now set up the discs
% . /boot/disco

-- hit ^D to exit the rc and do the normal startup. 
-- I have a kfs which does some prints for my own info, this will go 
-- away soon.

% # now hit ^D
% kfs...f is 0
fileinit returns
DONE iobufinit
Done the for stuff ...
DONE the boottime
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
mkchallenge
done mkchallenge
call fileinit
f is 0
fileinit returns
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
DONE the consserve()
Done the superok
About to startproc ...
version...f is 0
fileinit returns
TIme ... not setting, hard-wired by Xen
ETNER /rc/bin/cpurc
DO IPCONFIG
SO NOW TRY DATE ...
sys=v9 dom=vm ip=172.16.233.201 auth=v9 fs=v9 cpu=v9 authdom=v9 
SYSNAME
v9SHIT
SYSNAME is v9

-- oops, my bad. Forgot to fix this stupid typo in my rc.
also: '/bin/also' directory entry not found
announce il!*!17008

init: starting /bin/rc

-- we're up. 

v9# # we're up, you can drawterm in
v9# # You will need to set your password for bootes
v9# # time to sut down
v9# disk/kfscmd sync
v9# disk/kfscmd halt
kfs: file system halted
v9# someone's exiting
cpu0: exiting

************ REMOTE CONSOLE EXITED *****************
q:/tmp # 
q:/tmp # exit

Script done on Tue 28 Dec 2004 08:57:02 AM MST

Here's the sequence minus any comments at all. 

q:~ # xend start
q:~ # xm list
Name              Id  Mem(MB)  CPU  State  Time(s)  Console
Domain-0           0      251    0  r----    384.2        
q: # losetup /dev/loop0 kfs_root_image
q: # xm create -f /etc/xen/plan9 -c vmid=12
Using config file "/etc/xen/plan9".
Started domain VM12, console on port 9603
************ REMOTE CONSOLE: CTRL-] TO QUIT ********
root is from (local, tcp, il)[local]: XENFE:find_dev_by_handle 0 np 80260854 
np->dev 806c4630
There's still a leak in vif_disconnect
CONNECT: message sent. Set something to 0 just to see
CONNECT: set it
XENFE:find_dev_by_handle 0 np 80260854 np->dev 806c4630

try /bin/rc
try /386/bin/rc
try /boot/rc
% # note that I hit return there.
% # now set up the discs
% . /boot/disco
% # now hit ^D
% kfs...f is 0
fileinit returns
DONE iobufinit
Done the for stuff ...
DONE the boottime
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
mkchallenge
done mkchallenge
call fileinit
f is 0
fileinit returns
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
fcall9p1: start
fcall9p1: lock it
fcall9p1: call it
fcall9p1: unlock it
fcall9p1: done
fcall9p1: end
DONE the consserve()
Done the superok
About to startproc ...
version...f is 0
fileinit returns
TIme ... not setting, hard-wired by Xen
ETNER /rc/bin/cpurc
DO IPCONFIG
SO NOW TRY DATE ...
sys=v9 dom=vm ip=172.16.233.201 auth=v9 fs=v9 cpu=v9 authdom=v9 
SYSNAME
v9SHIT
SYSNAME is v9
also: '/bin/also' directory entry not found
announce il!*!17008

init: starting /bin/rc
v9# # we're up, you can drawterm in
v9# # You will need to set your password for bootes
v9# # time to sut down
v9# disk/kfscmd sync
v9# disk/kfscmd halt
kfs: file system halted
v9# someone's exiting
cpu0: exiting

************ REMOTE CONSOLE EXITED *****************
q:/tmp # 
q:/tmp # exit

Script done on Tue 28 Dec 2004 08:57:02 AM MST

Somebody let me know if they get anywhere at all with this :-)

The boot process takes < 10 seconds, if it takes a really long time then 
something may be wrong.

ron


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/xen-devel