File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.3: download - view: text, annotated - select for diffs
Thu Oct 31 12:43:04 2013 UTC (6 years, 1 month ago) by mspo
Branches: MAIN
CVS tags: HEAD
fix up some code blocks, part 1

    1: NetBSD/xen Howto
    2: ================
    3: 
    4: [![BSD
    5: daemon](../../images/BSD-daemon.jpg)](../../about/disclaimer.html#bsd-daemon)
    6: 
    7: Table Of Contents
    8: -----------------
    9: 
   10: -   [Introduction](#introduction)
   11: -   [Installing NetBSD as privileged domain (Dom0)](#netbsd-dom0)
   12: -   [Creating an unprivileged NetBSD domain (DomU)](#netbsd-domU)
   13: -   [Creating an unprivileged Linux domain (DomU)](#linux-domU)
   14: -   [Creating an unprivileged Solaris domain (DomU)](#solaris-domU)
   15: -   [Using PCI devices in guest domains](#pci-pass-through)
   16: -   [Links and further information](#links-and-more)
   17: 
   18: * * * * *
   19: 
   20: ### Introduction
   21: 
   22: [![[Xen
   23: screenshot]](../../gallery/in-Action/hubertf-xens.png)](../../gallery/in-Action/hubertf-xen.png)
   24: 
   25: Xen is a virtual machine monitor for x86 hardware (requires i686-class
   26: CPUs), which supports running multiple guest operating systems on a
   27: single machine. Guest OSes (also called <E2><80><9C>domains<E2><80><9D>) require a modified
   28: kernel which supports Xen hypercalls in replacement to access to the
   29: physical hardware. At boot, the Xen kernel (also known as the Xen
   30: hypervisor) is loaded (via the bootloader) along with the guest kernel
   31: for the first domain (called *domain0*). The Xen kernel has to be loaded
   32: using the multiboot protocol. You would use the NetBSD boot loader for
   33: this, or alternatively the **grub** boot loader (**grub** has some
   34: limitations, detailed below). *domain0* has special privileges to access
   35: the physical hardware (PCI and ISA devices), administrate other domains
   36: and provide virtual devices (disks and network) to other domains that
   37: lack those privileges. For more details, see
   38: [http://www.xen.org/](http://www.xen.org/).
   39: 
   40: NetBSD can be used for both *domain0 (Dom0)* and further, unprivileged
   41: (DomU) domains. (Actually there can be multiple privileged domains
   42: accessing different parts of the hardware, all providing virtual devices
   43: to unprivileged domains. We will only talk about the case of a single
   44: privileged domain, *domain0*). *domain0* will see physical devices much
   45: like a regular i386 or amd64 kernel, and will own the physical console
   46: (VGA or serial). Unprivileged domains will only see a character-only
   47: virtual console, virtual disks (`xbd`{.code}) and virtual network
   48: interfaces (`xennet`{.code}) provided by a privileged domain (usually
   49: *domain0*). xbd devices are connected to a block device (i.e., a
   50: partition of a disk, raid, ccd, ... device) in the privileged domain.
   51: xennet devices are connected to virtual devices in the privileged
   52: domain, named xvif\<domain number\>.\<if number for this domain\>, e.g.,
   53: xvif1.0. Both xennet and xvif devices are seen as regular Ethernet
   54: devices (they can be seen as a crossover cable between 2 PCs) and can be
   55: assigned addresses (and be routed or NATed, filtered using IPF, etc ...)
   56: or be added as part of a bridge.
   57: 
   58: * * * * *
   59: ### Installing NetBSD as privileged domain (Dom0)
   60: 
   61: First do a NetBSD/i386 or NetBSD/amd64
   62: [installation](../../docs/guide/en/chap-inst.html) of the 5.1 release
   63: (or newer) as you usually do on x86 hardware. The binary releases are
   64: available from
   65: [ftp://ftp.NetBSD.org/pub/NetBSD/](ftp://ftp.NetBSD.org/pub/NetBSD/).
   66: Binary snapshots for current and the stable branches are available on
   67: [daily autobuilds](http://nyftp.NetBSD.org/pub/NetBSD-daily/). If you
   68: plan to use the **grub** boot loader, when partitioning the disk you
   69: have to make the root partition smaller than 512Mb, and formatted as
   70: FFSv1 with 8k block/1k fragments. If the partition is larger than this,
   71: uses FFSv2 or has different block/fragment sizes, grub may fail to load
   72: some files. Also keep in mind that you'll probably want to provide
   73: virtual disks to other domains, so reserve some partitions for these
   74: virtual disks. Alternatively, you can create large files in the file
   75: system, map them to vnd(4) devices and export theses vnd devices to
   76: other domains.
   77: 
   78: Next step is to install the Xen packages via pkgsrc or from binary
   79: packages. See [the pkgsrc
   80: documentation](http://www.NetBSD.org/docs/pkgsrc/) if you are unfamiliar
   81: with pkgsrc and/or handling of binary packages. Xen 3.1, 3.3, 4.1 and
   82: 4.2 are available. 3.1 supports PCI pass-through while other versions do
   83: not. You'll need either
   84: [`sysutils/xentools3`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xentools3/README.html)
   85: and
   86: [`sysutils/xenkernel3`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xenkernel3/README.html)
   87: for Xen 3.1,
   88: [`sysutils/xentools33`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xentools33/README.html)
   89: and
   90: [`sysutils/xenkernel33`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xenkernel33/README.html)
   91: for Xen 3.3,
   92: [`sysutils/xentools41`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xentools41/README.html)
   93: and
   94: [`sysutils/xenkernel41`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xenkernel41/README.html)
   95: for Xen 4.1. or
   96: [`sysutils/xentools42`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xentools42/README.html)
   97: and
   98: [`sysutils/xenkernel42`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xenkernel42/README.html)
   99: for Xen 4.2. You'll also need
  100: [`sysutils/grub`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/grub/README.html)
  101: if you plan do use the grub boot loader. If using Xen 3.1, you may also
  102: want to install
  103: [`sysutils/xentools3-hvm`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xentools3-hvm/README.html)
  104: which contains the utilities to run unmodified guests OSes using the
  105: *HVM* support (for later versions this is included in
  106: [`sysutils/xentools`{.filename}](http://ftp.NetBSD.org/pub/pkgsrc/current/pkgsrc/sysutils/xentools/README.html)).
  107: Note that your CPU needs to support this. Intel CPUs must have the 'VT'
  108: instruction, AMD CPUs the 'SVM' instruction. You can easily find out if
  109: your CPU support HVM by using NetBSD's cpuctl command:
  110: 
  111:     # cpuctl identify 0
  112:     cpu0: Intel Core 2 (Merom) (686-class), id 0x6f6
  113:     cpu0: features 0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR>
  114:     cpu0: features 0xbfebfbff<PGE,MCA,CMOV,PAT,PSE36,CFLUSH,DS,ACPI,MMX>
  115:     cpu0: features 0xbfebfbff<FXSR,SSE,SSE2,SS,HTT,TM,SBF>
  116:     cpu0: features2 0x4e33d<SSE3,DTES64,MONITOR,DS-CPL,VMX,TM2,SSSE3,CX16,xTPR,PDCM,DCA>
  117:     cpu0: features3 0x20100800<SYSCALL/SYSRET,XD,EM64T>
  118:     cpu0: "Intel(R) Xeon(R) CPU            5130  @ 2.00GHz"
  119:     cpu0: I-cache 32KB 64B/line 8-way, D-cache 32KB 64B/line 8-way
  120:     cpu0: L2 cache 4MB 64B/line 16-way
  121:     cpu0: ITLB 128 4KB entries 4-way
  122:     cpu0: DTLB 256 4KB entries 4-way, 32 4MB entries 4-way
  123:     cpu0: Initial APIC ID 0
  124:     cpu0: Cluster/Package ID 0
  125:     cpu0: Core ID 0
  126:     cpu0: family 06 model 0f extfamily 00 extmodel 00
  127: 
  128: Depending on your CPU, the feature you are looking for is called HVM,
  129: SVM or VMX.
  130: 
  131: Next you need to copy the selected Xen kernel itself. pkgsrc installed
  132: them under `/usr/pkg/xen*-kernel/`{.filename}. The file you're looking
  133: for is `xen.gz`{.filename}. Copy it to your root file system.
  134: `xen-debug.gz`{.filename} is a kernel with more consistency checks and
  135: more details printed on the serial console. It is useful for debugging
  136: crashing guests if you use a serial console. It is not useful with a VGA
  137: console.
  138: 
  139: You'll then need a NetBSD/Xen kernel for *domain0* on your root file
  140: system. The XEN3PAE\_DOM0 kernel or XEN3\_DOM0 provided as part of the
  141: i386 or amd64 binaries is suitable for this, but you may want to
  142: customize it. Keep your native kernel around, as it can be useful for
  143: recovery. *Note:* the *domain0* kernel must support KERNFS and
  144: `/kern`{.filename} must be mounted because *xend* needs access to
  145: `/kern/xen/privcmd`{.filename}.
  146: 
  147: Next you need to get a bootloader to load the `xen.gz`{.filename}
  148: kernel, and the NetBSD *domain0* kernel as a module. This can be
  149: **grub** or NetBSD's boot loader. Below is a detailled example for grub,
  150: see the boot.cfg(5) manual page for an example using the latter.
  151: 
  152: This is also where you'll specify the memory allocated to *domain0*, the
  153: console to use, etc ...
  154: 
  155: Here is a commented `/grub/menu.lst`{.filename} file:
  156: 
  157: #Grub config file for NetBSD/xen. Copy as /grub/menu.lst and run
  158:     # grub-install /dev/rwd0d (assuming your boot device is wd0).
  159:     #
  160:     # The default entry to load will be the first one
  161:     default=0
  162:     
  163:     # boot the default entry after 10s if the user didn't hit keyboard
  164:     timeout=10
  165:     
  166:     # Configure serial port to use as console. Ignore if you'll use VGA only
  167:     serial --unit=0 --speed=115200 --word=8 --parity=no --stop=1
  168:     
  169:     # Let the user select which console to use (serial or VGA), default
  170:     # to serial after 10s
  171:     terminal --timeout=10 serial console
  172:     
  173:     # An entry for NetBSD/xen, using /netbsd as the domain0 kernel, and serial
  174:     # console. Domain0 will have 64MB RAM allocated.
  175:     # Assume NetBSD is installed in the first MBR partition.
  176:     title Xen 3 / NetBSD (hda0, serial)
  177:       root(hd0,0)
  178:       kernel (hd0,a)/xen.gz dom0_mem=65536 com1=115200,8n1
  179:       module (hd0,a)/netbsd bootdev=wd0a ro console=ttyS0
  180:     
  181:     # Same as above, but using VGA console
  182:     # We can use console=tty0 (Linux syntax) or console=pc (NetBSD syntax)
  183:     title Xen 3 / NetBSD (hda0, vga)
  184:       root(hd0,0)
  185:       kernel (hd0,a)/xen.gz dom0_mem=65536
  186:       module (hd0,a)/netbsd bootdev=wd0a ro console=tty0
  187:     
  188:     # NetBSD/xen using a backup domain0 kernel (in case you installed a
  189:     # nonworking kernel as /netbsd
  190:     title Xen 3 / NetBSD (hda0, backup, serial)
  191:       root(hd0,0)
  192:       kernel (hd0,a)/xen.gz dom0_mem=65536 com1=115200,8n1
  193:       module (hd0,a)/netbsd.backup bootdev=wd0a ro console=ttyS0
  194:     title Xen 3 / NetBSD (hda0, backup, VGA)
  195:       root(hd0,0)
  196:       kernel (hd0,a)/xen.gz dom0_mem=65536
  197:       module (hd0,a)/netbsd.backup bootdev=wd0a ro console=tty0
  198:     
  199:     #Load a regular NetBSD/i386 kernel. Can be useful if you end up with a
  200:     #nonworking /xen.gz
  201:     title NetBSD 5.1
  202:       root (hd0,a)
  203:       kernel --type=netbsd /netbsd-GENERIC
  204:     
  205:     #Load the NetBSD bootloader, letting it load the NetBSD/i386 kernel.
  206:     #May be better than the above, as grub can't pass all required infos
  207:     #to the NetBSD/i386 kernel (e.g. console, root device, ...)
  208:     title NetBSD chain
  209:       root        (hd0,0)
  210:       chainloader +1
  211:     
  212:     ## end of grub config file.
  213:       
  214: Install grub with the following command:
  215: 
  216:     # grub --no-floppy
  217:     
  218:     grub> root (hd0,a)
  219:      Filesystem type is ffs, partition type 0xa9
  220:     
  221:     grub> setup (hd0)
  222:      Checking if "/boot/grub/stage1" exists... no
  223:      Checking if "/grub/stage1" exists... yes
  224:      Checking if "/grub/stage2" exists... yes
  225:      Checking if "/grub/ffs_stage1_5" exists... yes
  226:      Running "embed /grub/ffs_stage1_5 (hd0)"...  14 sectors are embedded.
  227:     succeeded
  228:      Running "install /grub/stage1 (hd0) (hd0)1+14 p (hd0,0,a)/grub/stage2 /grub/menu.lst"...
  229:      succeeded
  230:     Done.
  231: 
  232: * * * * *
  233: 
  234: ### Creating an unprivileged NetBSD domain (DomU)
  235: 
  236: Once you have *domain0* running, you need to start the xen tool daemon
  237: (**/usr/pkg/share/examples/rc.d/xend start**) and the xen backend daemon
  238: (**/usr/pkg/share/examples/rc.d/xenbackendd start** for Xen3\*,
  239: **/usr/pkg/share/examples/rc.d/xencommons start** for Xen4.\*). Make
  240: sure that `/dev/xencons`{.filename} and `/dev/xenevt`{.filename} exist
  241: before starting **xend**. You can create them with this command:
  242: 
  243:     # cd /dev && sh MAKEDEV xen
  244: 
  245: xend will write logs to `/var/log/xend.log`{.filename} and
  246: `/var/log/xend-debug.log`{.filename}. You can then control xen with the
  247: xm tool. 'xm list' will show something like:
  248: 
  249:     # xm list
  250:     Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  251:     Domain-0           0       64    0  r----     58.1
  252: 
  253: 'xm create' allows you to create a new domain. It uses a config file in
  254: PKG\_SYSCONFDIR for its parameters. By default, this file will be in
  255: `/usr/pkg/etc/xen/`{.filename}. On creation, a kernel has to be
  256: specified, which will be executed in the new domain (this kernel is in
  257: the *domain0* file system, not on the new domain virtual disk; but
  258: please note, you should install the same kernel into *domainU* as
  259: `/netbsd`{.filename} in order to make your system tools, like
  260: [savecore(8)](http://netbsd.gw.com/cgi-bin/man-cgi?savecore+8+NetBSD-6.0+i386),
  261: work). A suitable kernel is provided as part of the i386 and amd64
  262: binary sets: XEN3\_DOMU.
  263: 
  264: Here is an /usr/pkg/etc/xen/nbsd example config file:
  265: 
  266:     #  -*- mode: python; -*-
  267:     #============================================================================
  268:     # Python defaults setup for 'xm create'.
  269:     # Edit this file to reflect the configuration of your system.
  270:     #============================================================================
  271:     
  272:     #----------------------------------------------------------------------------
  273:     # Kernel image file. This kernel will be loaded in the new domain.
  274:     kernel = "/home/bouyer/netbsd-XEN3_DOMU"
  275:     #kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  276:     
  277:     # Memory allocation (in megabytes) for the new domain.
  278:     memory = 128
  279:     
  280:     # A handy name for your new domain. This will appear in 'xm list',
  281:     # and you can use this as parameters for xm in place of the domain
  282:     # number. All domains must have different names.
  283:     #
  284:     name = "nbsd"
  285:     
  286:     # The number of virtual CPUs this domain has.
  287:     #
  288:     vcpus = 1
  289:     
  290:     #----------------------------------------------------------------------------
  291:     # Define network interfaces for the new domain.
  292:     
  293:     # Number of network interfaces (must be at least 1). Default is 1.
  294:     nics = 1
  295:     
  296:     # Define MAC and/or bridge for the network interfaces.
  297:     #
  298:     # The MAC address specified in ``mac'' is the one used for the interface
  299:     # in the new domain. The interface in domain0 will use this address XOR'd
  300:     # with 00:00:00:01:00:00 (i.e. aa:00:00:51:02:f0 in our example). Random
  301:     # MACs are assigned if not given.
  302:     #
  303:     # ``bridge'' is a required parameter, which will be passed to the
  304:     # vif-script called by xend(8) when a new domain is created to configure
  305:     # the new xvif interface in domain0.
  306:     #
  307:     # In this example, the xvif is added to bridge0, which should have been
  308:     # set up prior to the new domain being created -- either in the
  309:     # ``network'' script or using a /etc/ifconfig.bridge0 file.
  310:     #
  311:     vif = [ 'mac=aa:00:00:50:02:f0, bridge=bridge0' ]
  312:     
  313:     #----------------------------------------------------------------------------
  314:     # Define the disk devices you want the domain to have access to, and
  315:     # what you want them accessible as.
  316:     #
  317:     # Each disk entry is of the form:
  318:     #
  319:     #       phy:DEV,VDEV,MODE
  320:     #
  321:     # where DEV is the device, VDEV is the device name the domain will see,
  322:     # and MODE is r for read-only, w for read-write.  You can also create
  323:     # file-backed domains using disk entries of the form:
  324:     #
  325:     #       file:PATH,VDEV,MODE
  326:     #
  327:     # where PATH is the path to the file used as the virtual disk, and VDEV
  328:     # and MODE have the same meaning as for ``phy'' devices.
  329:     #
  330:     # VDEV doesn't really matter for a NetBSD guest OS (it's just used as an index),
  331:     # but it does for Linux.
  332:     # Worse, the device has to exist in /dev/ of domain0, because xm will
  333:     # try to stat() it. This means that in order to load a Linux guest OS
  334:     # from a NetBSD domain0, you'll have to create /dev/hda1, /dev/hda2, ...
  335:     # on domain0, with the major/minor from Linux :(
  336:     # Alternatively it's possible to specify the device number in hex,
  337:     # e.g. 0x301 for /dev/hda1, 0x302 for /dev/hda2, etc ...
  338:     
  339:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  340:     #disk = [ 'file:/var/xen/nbsd-disk,0x01,w' ]
  341:     #disk = [ 'file:/var/xen/nbsd-disk,0x301,w' ]
  342:     
  343:     #----------------------------------------------------------------------------
  344:     # Set the kernel command line for the new domain.
  345:     
  346:     # Set root device. This one does matter for NetBSD
  347:     root = "xbd0"
  348:     # extra parameters passed to the kernel
  349:     # this is where you can set boot flags like -s, -a, etc ...
  350:     #extra = ""
  351:     
  352:     #----------------------------------------------------------------------------
  353:     # Set according to whether you want the domain restarted when it exits.
  354:     # The default is False.
  355:     #autorestart = True
  356:     
  357:     # end of nbsd config file ====================================================
  358: 
  359: When a new domain is created, xen calls the
  360: `/usr/pkg/etc/xen/vif-bridge`{.filename} script for each virtual network
  361: interface created in *domain0*. This can be used to automatically
  362: configure the xvif?.? interfaces in *domain0*. In our example, these
  363: will be bridged with the bridge0 device in *domain0*, but the bridge has
  364: to exist first. To do this, create the file
  365: `/etc/ifconfig.bridge0`{.filename} and make it look like this:
  366: 
  367:     create
  368:     !brconfig $int add ex0 up
  369: 
  370: (replace `ex0`{.literal} with the name of your physical interface). Then
  371: bridge0 will be created on boot. See the
  372: [bridge(4)](http://netbsd.gw.com/cgi-bin/man-cgi?bridge+4+NetBSD-6.0+i386)
  373: man page for details.
  374: 
  375: So, here is a suitable `/usr/pkg/etc/xen/vif-bridge`{.filename} for
  376: xvif?.? (a working vif-bridge is also provided with xentools20)
  377: configuring:
  378: 
  379: 
  380: #!/bin/sh
  381:     #============================================================================
  382:     # $NetBSD: howto.mdwn,v 1.3 2013/10/31 12:43:04 mspo Exp $
  383:     #
  384:     # /usr/pkg/etc/xen/vif-bridge
  385:     #
  386:     # Script for configuring a vif in bridged mode with a dom0 interface.
  387:     # The xend(8) daemon calls a vif script when bringing a vif up or down.
  388:     # The script name to use is defined in /usr/pkg/etc/xen/xend-config.sxp
  389:     # in the ``vif-script'' field.
  390:     #
  391:     # Usage: vif-bridge up|down [var=value ...]
  392:     #
  393:     # Actions:
  394:     #    up         Adds the vif interface to the bridge.
  395:     #    down       Removes the vif interface from the bridge.
  396:     #
  397:     # Variables:
  398:     #    domain     name of the domain the interface is on (required).
  399:     #    vifq       vif interface name (required).
  400:     #    mac        vif MAC address (required).
  401:     #    bridge     bridge to add the vif to (required).
  402:     #
  403:     # Example invocation:
  404:     #
  405:     # vif-bridge up domain=VM1 vif=xvif1.0 mac="ee:14:01:d0:ec:af" bridge=bridge0
  406:     #
  407:     #============================================================================
  408:     
  409:     # Exit if anything goes wrong
  410:     set -e
  411:     
  412:     echo "vif-bridge $*"
  413:     
  414:     # Operation name.
  415:     OP=$1; shift
  416:     
  417:     # Pull variables in args into environment
  418:     for arg ; do export "${arg}" ; done
  419:     
  420:     # Required parameters. Fail if not set.
  421:     domain=${domain:?}
  422:     vif=${vif:?}
  423:     mac=${mac:?}
  424:     bridge=${bridge:?}
  425:     
  426:     # Optional parameters. Set defaults.
  427:     ip=${ip:-''}   # default to null (do nothing)
  428:     
  429:     # Are we going up or down?
  430:     case $OP in
  431:     up)     brcmd='add' ;;
  432:     down)   brcmd='delete' ;;
  433:     *)
  434:             echo 'Invalid command: ' $OP
  435:             echo 'Valid commands are: up, down'
  436:             exit 1
  437:             ;;
  438:     esac
  439:     
  440:     # Don't do anything if the bridge is "null".
  441:     if [ "${bridge}" = "null" ] ; then
  442:             exit
  443:     fi
  444:     
  445:     # Don't do anything if the bridge doesn't exist.
  446:     if ! ifconfig -l | grep "${bridge}" >/dev/null; then
  447:             exit
  448:     fi
  449:     
  450:     # Add/remove vif to/from bridge.
  451:     ifconfig x${vif} $OP
  452:     brconfig ${bridge} ${brcmd} x${vif}
  453: 
  454: Now, running
  455: 
  456:     xm create -c /usr/pkg/etc/xen/nbsd
  457: 
  458: should create a domain and load a NetBSD kernel in it. (Note:
  459: `-c`{.code} causes xm to connect to the domain's console once created.)
  460: The kernel will try to find its root file system on xbd0 (i.e., wd0e)
  461: which hasn't been created yet. wd0e will be seen as a disk device in the
  462: new domain, so it will be 'sub-partitioned'. We could attach a ccd to
  463: wd0e in *domain0* and partition it, newfs and extract the NetBSD/i386 or
  464: amd64 tarballs there, but there's an easier way: load the
  465: `netbsd-INSTALL_XEN3_DOMU`{.filename} kernel provided in the NetBSD
  466: binary sets. Like other install kernels, it contains a ramdisk with
  467: sysinst, so you can install NetBSD using sysinst on your new domain.
  468: 
  469: If you want to install NetBSD/Xen with a CDROM image, the following line
  470: should be used in the `/usr/pkg/etc/xen/nbsd`{.filename} file:
  471: 
  472:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  473: 
  474: After booting the domain, the option to install via CDROM may be
  475: selected. The CDROM device should be changed to **xbd1d**.
  476: 
  477: Once done installing, **halt -p** the new domain (don't reboot or halt,
  478: it would reload the INSTALL\_XEN3\_DOMU kernel even if you changed the
  479: config file), switch the config file back to the XEN3\_DOMU kernel, and
  480: start the new domain again. Now it should be able to use **root on
  481: xbd0a** and you should have a second, functional NetBSD system on your
  482: xen installation.
  483: 
  484: When the new domain is booting you'll see some warnings about *wscons*
  485: and the pseudo-terminals. These can be fixed by editing the files
  486: `/etc/ttys`{.filename} and `/etc/wscons.conf`{.filename}. You must
  487: disable all terminals in `/etc/ttys`{.filename}, except *console*, like
  488: this:
  489: 
  490:     console "/usr/libexec/getty Pc"         vt100   on secure
  491:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  492:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  493:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  494:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  495: 
  496: Finally, all screens must be commented out from
  497: `/etc/wscons.conf`{.filename}.
  498: 
  499: It is also desirable to add
  500: 
  501:     powerd=YES
  502: 
  503: in rc.conf. This way, the domain will be properly shut down if **xm
  504: shutdown -R** or **xm shutdown -H** is used on the domain0.
  505: 
  506: Your domain should be now ready to work, enjoy.
  507: 
  508: * * * * *
  509: 
  510: ### Creating an unprivileged Linux domain (DomU)
  511: 
  512: Creating unprivileged Linux domains isn't much different from
  513: unprivileged NetBSD domains, but there are some details to know.
  514: 
  515: First, the second parameter passed to the disk declaration (the '0x1' in
  516: the example below)
  517: 
  518:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  519: 
  520: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  521: for hda). Linux builds device numbers as: (major \<\< 8 + minor). So,
  522: hda1 which has major 3 and minor 1 on a Linux system will have device
  523: number 0x301. Alternatively, devices names can be used (hda, hdb, ...)
  524: as xentools has a table to map these names to devices numbers. To export
  525: a partition to a Linux guest we can use:
  526: 
  527:     disk = [ 'phy:/dev/wd0e,0x300,w' ]
  528:     root = "/dev/hda1 ro"
  529: 
  530: and it will appear as /dev/hda on the Linux system, and be used as root
  531: partition.
  532: 
  533: To install the Linux system on the partition to be exported to the guest
  534: domain, the following method can be used: install sysutils/e2fsprogs
  535: from pkgsrc. Use mke2fs to format the partition that will be the root
  536: partition of your Linux domain, and mount it. Then copy the files from a
  537: working Linux system, make adjustments in `/etc`{.filename} (fstab,
  538: network config). It should also be possible to extract binary packages
  539: such as .rpm or .deb directly to the mounted partition using the
  540: appropriate tool, possibly running under NetBSD's Linux emulation. Once
  541: the filesystem has been populated, umount it. If desirable, the
  542: filesystem can be converted to ext3 using tune2fs -j. It should now be
  543: possible to boot the Linux guest domain, using one of the
  544: vmlinuz-\*-xenU kernels available in the Xen binary distribution.
  545: 
  546: To get the linux console right, you need to add:
  547: 
  548:     extra = "xencons=tty1"
  549: 
  550: to your configuration since not all linux distributions auto-attach a
  551: tty to the xen console.
  552: 
  553: * * * * *
  554: 
  555: ### Creating an unprivileged Solaris domain (DomU)
  556: 
  557: Download an Opensolaris [release](http://opensolaris.org/os/downloads/)
  558: or [development snapshot](http://genunix.org/) DVD image. Attach the DVD
  559: image to a
  560: [vnd(4)](http://netbsd.gw.com/cgi-bin/man-cgi?vnd+4+NetBSD-6.0+i386)
  561: device. Copy the kernel and ramdisk filesystem image to your dom0
  562: filesystem.
  563: 
  564:     dom0# mkdir /root/solaris
  565:     dom0# vnconfig vnd0 osol-1002-124-x86.iso
  566:     dom0# mount /dev/vnd0a /mnt
  567:     
  568:     ## for a 64-bit guest
  569:     dom0# cp /mnt/boot/amd64/x86.microroot /root/solaris
  570:     dom0# cp /mnt/platform/i86xpv/kernel/amd64/unix /root/solaris
  571:     
  572:     ## for a 32-bit guest
  573:     dom0# cp /mnt/boot/x86.microroot /root/solaris
  574:     dom0# cp /mnt/platform/i86xpv/kernel/unix /root/solaris
  575:     
  576:     dom0# umount /mnt
  577:       
  578: Keep the
  579: [vnd(4)](http://netbsd.gw.com/cgi-bin/man-cgi?vnd+4+NetBSD-6.0+i386)
  580: configured. For some reason the boot process stalls unless the DVD image
  581: is attached to the guest as a "phy" device. Create an initial
  582: configuration file with the following contents. Substitute */dev/wd0k*
  583: with an empty partition at least 8 GB large.
  584: 
  585: ~~~ {.programlisting}
  586: memory = 640
  587: name = 'solaris'
  588: disk = [ 'phy:/dev/wd0k,0,w' ]
  589: disk += [ 'phy:/dev/vnd0d,6:cdrom,r' ]
  590: vif = [ 'bridge=bridge0' ]
  591: kernel = '/root/solaris/unix'
  592: ramdisk = '/root/solaris/x86.microroot'
  593: # for a 64-bit guest
  594: extra = '/platform/i86xpv/kernel/amd64/unix - nowin -B install_media=cdrom'
  595: # for a 32-bit guest
  596: #extra = '/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'
  597:       
  598: ~~~
  599: 
  600: Start the guest.
  601: 
  602: ~~~ {.programlisting}
  603: dom0# xm create -c solaris.cfg
  604: Started domain solaris
  605:                       v3.3.2 chgset 'unavailable'
  606: SunOS Release 5.11 Version snv_124 64-bit
  607: Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  608: Use is subject to license terms.
  609: Hostname: opensolaris
  610: Remounting root read/write
  611: Probing for device nodes ...
  612: WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  613: Preparing live image for use
  614: Done mounting Live image
  615:       
  616: ~~~
  617: 
  618: Make sure the network is configured. Note that it can take a minute for
  619: the xnf0 interface to appear.
  620: 
  621: ~~~ {.programlisting}
  622: opensolaris console login: jack
  623: Password: jack
  624: Sun Microsystems Inc.   SunOS 5.11      snv_124 November 2008
  625: jack@opensolaris:~$ pfexec sh
  626: sh-3.2# ifconfig -a
  627: sh-3.2# exit
  628:       
  629: ~~~
  630: 
  631: Set a password for VNC and start the VNC server which provides the X11
  632: display where the installation program runs.
  633: 
  634: ~~~ {.programlisting}
  635: jack@opensolaris:~$ vncpasswd
  636: Password: solaris
  637: Verify: solaris
  638: jack@opensolaris:~$ cp .Xclients .vnc/xstartup
  639: jack@opensolaris:~$ vncserver :1
  640:       
  641: ~~~
  642: 
  643: From a remote machine connect to the VNC server. Use **ifconfig xnf0**
  644: on the guest to find the correct IP address to use.
  645: 
  646: ~~~ {.programlisting}
  647: remote$ vncviewer 172.18.2.99:1
  648:       
  649: ~~~
  650: 
  651: It is also possible to launch the installation on a remote X11 display.
  652: 
  653: ~~~ {.programlisting}
  654: jack@opensolaris:~$ export DISPLAY=172.18.1.1:0
  655: jack@opensolaris:~$ pfexec gui-install
  656:        
  657: ~~~
  658: 
  659: After the GUI installation is complete you will be asked to reboot.
  660: Before that you need to determine the ZFS ID for the new boot filesystem
  661: and update the configuration file accordingly. Return to the guest
  662: console.
  663: 
  664: ~~~ {.programlisting}
  665: jack@opensolaris:~$ pfexec zdb -vvv rpool | grep bootfs
  666:                 bootfs = 43
  667: ^C
  668: jack@opensolaris:~$
  669:        
  670: ~~~
  671: 
  672: The final configuration file should look like this. Note in particular
  673: the last line.
  674: 
  675: 
  676: ~~~ {.programlisting}
  677: memory = 640
  678: name = 'solaris'
  679: disk = [ 'phy:/dev/wd0k,0,w' ]
  680: vif = [ 'bridge=bridge0' ]
  681: kernel = '/root/solaris/unix'
  682: ramdisk = '/root/solaris/x86.microroot'
  683: extra = '/platform/i86xpv/kernel/amd64/unix -B zfs-bootfs=rpool/43,bootpath="/xpvd/xdf@0:a"'
  684:        
  685: ~~~
  686: 
  687: Restart the guest to verify it works correctly.
  688: 
  689: ~~~ {.programlisting}
  690: dom0# xm destroy solaris
  691: dom0# xm create -c solaris.cfg
  692: Using config file "./solaris.cfg".
  693: v3.3.2 chgset 'unavailable'
  694: Started domain solaris
  695: SunOS Release 5.11 Version snv_124 64-bit
  696: Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  697: Use is subject to license terms.
  698: WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  699: Hostname: osol
  700: Configuring devices.
  701: Loading smf(5) service descriptions: 160/160
  702: svccfg import warnings. See /var/svc/log/system-manifest-import:default.log .
  703: Reading ZFS config: done.
  704: Mounting ZFS filesystems: (6/6)
  705: Creating new rsa public/private host key pair
  706: Creating new dsa public/private host key pair
  707: 
  708: osol console login:
  709:        
  710: ~~~
  711: 
  712: Using PCI devices in guest domains
  713: ----------------------------------
  714: 
  715: The domain0 can give other domains access to selected PCI devices. This
  716: can allow, for example, a non-privileged domain to have access to a
  717: physical network interface or disk controller. However, keep in mind
  718: that giving a domain access to a PCI device most likely will give the
  719: domain read/write access to the whole physical memory, as PCs don't have
  720: an IOMMU to restrict memory access to DMA-capable device. Also, it's not
  721: possible to export ISA devices to non-domain0 domains (which means that
  722: the primary VGA adapter can't be exported. A guest domain trying to
  723: access the VGA registers will panic).
  724: 
  725: This functionality is only available in NetBSD-5.1 (and later) domain0
  726: and domU. If the domain0 is NetBSD, it has to be running Xen 3.1, as
  727: support has not been ported to later versions at this time.
  728: 
  729: For a PCI device to be exported to a domU, is has to be attached to the
  730: `pciback`{.literal} driver in domain0. Devices passed to the domain0 via
  731: the pciback.hide boot parameter will attach to `pciback`{.literal}
  732: instead of the usual driver. The list of devices is specified as
  733: `(bus:dev.func)`{.literal}, where bus and dev are 2-digit hexadecimal
  734: numbers, and func a single-digit number:
  735: 
  736: ~~~ {.programlisting}
  737: pciback.hide=(00:0a.0)(00:06.0)
  738: ~~~
  739: 
  740: pciback devices should show up in the domain0's boot messages, and the
  741: devices should be listed in the `/kern/xen/pci`{.filename} directory.
  742: 
  743: PCI devices to be exported to a domU are listed in the `pci`{.literal}
  744: array of the domU's config file, with the format
  745: `'0000:bus:dev.func'`{.literal}
  746: 
  747: ~~~ {.programlisting}
  748: pci = [ '0000:00:06.0', '0000:00:0a.0' ]
  749: ~~~
  750: 
  751: In the domU an `xpci`{.literal} device will show up, to which one or
  752: more pci busses will attach. Then the PCI drivers will attach to PCI
  753: busses as usual. Note that the default NetBSD DOMU kernels do not have
  754: `xpci`{.literal} or any PCI drivers built in by default; you have to
  755: build your own kernel to use PCI devices in a domU. Here's a kernel
  756: config example:
  757: 
  758: ~~~ {.programlisting}
  759: include         "arch/i386/conf/XEN3_DOMU"
  760: #include         "arch/i386/conf/XENU"           # in NetBSD 3.0
  761: 
  762: # Add support for PCI busses to the XEN3_DOMU kernel
  763: xpci* at xenbus ?
  764: pci* at xpci ?
  765: 
  766: # Now add PCI and related devices to be used by this domain
  767: # USB Controller and Devices
  768: 
  769: # PCI USB controllers
  770: uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
  771: 
  772: # USB bus support
  773: usb*    at uhci?
  774: 
  775: # USB Hubs
  776: uhub*   at usb?
  777: uhub*   at uhub? port ? configuration ? interface ?
  778: 
  779: # USB Mass Storage
  780: umass*  at uhub? port ? configuration ? interface ?
  781: wd*     at umass?
  782: # SCSI controllers
  783: ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
  784: 
  785: # SCSI bus support (for both ahc and umass)
  786: scsibus* at scsi?
  787: 
  788: # SCSI devices
  789: sd*     at scsibus? target ? lun ?      # SCSI disk drives
  790: cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
  791: ~~~
  792: 
  793: Links and further information
  794: -----------------------------
  795: 
  796: -   The HowTo on [Installing into
  797:     RAID-1](http://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
  798:     gives some hints on using Xen (grub) with NetBSD's RAIDframe
  799: -   Harold Gutch wrote documentation on [setting up a Linux DomU with a
  800:     NetBSD Dom0](http://www.gutch.de/NetBSD/docs/xen.html)
  801: -   An example of how to use NetBSD's native bootloader to load
  802:     NetBSD/Xen instead of Grub can be found in the i386/amd64
  803:     [boot(8)](http://netbsd.gw.com/cgi-bin/man-cgi?boot+8+NetBSD-6.0+i386)
  804:     and
  805:     [boot.cfg(5)](http://netbsd.gw.com/cgi-bin/man-cgi?boot.cfg+5+NetBSD-6.0+i386)
  806:     manpages.
  807: 
  808: 

CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb