File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.7: download - view: text, annotated - select for diffs
Mon Nov 4 02:43:35 2013 UTC (5 years, 11 months ago) by mspo
Branches: MAIN
CVS tags: HEAD
fix some issue sin the howto

    1: <table>
    2: <tbody>
    3: <tr class="odd">
    4: <td align="left"><a href="../../about/disclaimer.html#bsd-daemon"></a></td>
    5: <td align="left"><h1>Table Of Contents</h1>
    6: <ul>
    7: <li>Introduction</li>
    8: <li>Installing NetBSD as privileged domain (Dom0)</li>
    9: <li>Creating an unprivileged NetBSD domain (DomU)</li>
   10: <li>Creating an unprivileged Linux domain (DomU)</li>
   11: <li>Creating an unprivileged Solaris domain (DomU)</li>
   12: <li>Using PCI devices in guest domains</li>
   13: <li>Links and further information</li>
   14: </ul></td>
   15: </tr>
   16: </tbody>
   17: </table>
   18: 
   19: Introduction
   20: ------------
   21: 
   22: [![[Xen
   23: screenshot]](http://www.netbsd.org/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 “domains”) 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 [](http://www.xen.org/).
   38: 
   39: NetBSD can be used for both *domain0 (Dom0)* and further, unprivileged
   40: (DomU) domains. (Actually there can be multiple privileged domains
   41: accessing different parts of the hardware, all providing virtual devices
   42: to unprivileged domains. We will only talk about the case of a single
   43: privileged domain, *domain0*). *domain0* will see physical devices much
   44: like a regular i386 or amd64 kernel, and will own the physical console
   45: (VGA or serial). Unprivileged domains will only see a character-only
   46: virtual console, virtual disks (`xbd`) and virtual network interfaces
   47: (`xennet`) provided by a privileged domain (usually *domain0*). xbd
   48: devices are connected to a block device (i.e., a partition of a disk,
   49: raid, ccd, ... device) in the privileged domain. xennet devices are
   50: connected to virtual devices in the privileged domain, named
   51: xvif\<domain number\>.\<if number for this domain\>, e.g., xvif1.0. Both
   52: xennet and xvif devices are seen as regular Ethernet devices (they can
   53: be seen as a crossover cable between 2 PCs) and can be assigned
   54: addresses (and be routed or NATed, filtered using IPF, etc ...) or be
   55: added as part of a bridge.
   56: 
   57: Installing NetBSD as privileged domain (Dom0)
   58: ---------------------------------------------
   59: 
   60: First do a NetBSD/i386 or NetBSD/amd64
   61: [installation](../../docs/guide/en/chap-inst.html) of the 5.1 release
   62: (or newer) as you usually do on x86 hardware. The binary releases are
   63: available from [](ftp://ftp.NetBSD.org/pub/NetBSD/). Binary snapshots
   64: for current and the stable branches are available on daily autobuilds.
   65: If you plan to use the `grub` boot loader, when partitioning the disk
   66: you have to make the root partition smaller than 512Mb, and formatted as
   67: FFSv1 with 8k block/1k fragments. If the partition is larger than this,
   68: uses FFSv2 or has different block/fragment sizes, grub may fail to load
   69: some files. Also keep in mind that you'll probably want to provide
   70: virtual disks to other domains, so reserve some partitions for these
   71: virtual disks. Alternatively, you can create large files in the file
   72: system, map them to vnd(4) devices and export theses vnd devices to
   73: other domains.
   74: 
   75: Next step is to install the Xen packages via pkgsrc or from binary
   76: packages. See [the pkgsrc
   77: documentation](http://www.NetBSD.org/docs/pkgsrc/) if you are unfamiliar
   78: with pkgsrc and/or handling of binary packages. Xen 3.1, 3.3, 4.1 and
   79: 4.2 are available. 3.1 supports PCI pass-through while other versions do
   80: not. You'll need either `sysutils/xentools3` and `sysutils/xenkernel3`
   81: for Xen 3.1, `sysutils/xentools33` and `sysutils/xenkernel33` for Xen
   82: 3.3, `sysutils/xentools41` and `sysutils/xenkernel41` for Xen 4.1. or
   83: `sysutils/xentools42` and `sysutils/xenkernel42` for Xen 4.2. You'll
   84: also need `sysutils/grub` if you plan do use the grub boot loader. If
   85: using Xen 3.1, you may also want to install `sysutils/xentools3-hvm`
   86: which contains the utilities to run unmodified guests OSes using the
   87: *HVM* support (for later versions this is included in
   88: `sysutils/xentools`). Note that your CPU needs to support this. Intel
   89: CPUs must have the 'VT' instruction, AMD CPUs the 'SVM' instruction. You
   90: can easily find out if your CPU support HVM by using NetBSD's cpuctl
   91: command:
   92: 
   93:     # cpuctl identify 0
   94:     cpu0: Intel Core 2 (Merom) (686-class), id 0x6f6
   95:     cpu0: features 0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR>
   96:     cpu0: features 0xbfebfbff<PGE,MCA,CMOV,PAT,PSE36,CFLUSH,DS,ACPI,MMX>
   97:     cpu0: features 0xbfebfbff<FXSR,SSE,SSE2,SS,HTT,TM,SBF>
   98:     cpu0: features2 0x4e33d<SSE3,DTES64,MONITOR,DS-CPL,,TM2,SSSE3,CX16,xTPR,PDCM,DCA>
   99:     cpu0: features3 0x20100800<SYSCALL/SYSRET,XD,EM64T>
  100:     cpu0: "Intel(R) Xeon(R) CPU            5130  @ 2.00GHz"
  101:     cpu0: I-cache 32KB 64B/line 8-way, D-cache 32KB 64B/line 8-way
  102:     cpu0: L2 cache 4MB 64B/line 16-way
  103:     cpu0: ITLB 128 4KB entries 4-way
  104:     cpu0: DTLB 256 4KB entries 4-way, 32 4MB entries 4-way
  105:     cpu0: Initial APIC ID 0
  106:     cpu0: Cluster/Package ID 0
  107:     cpu0: Core ID 0
  108:     cpu0: family 06 model 0f extfamily 00 extmodel 00
  109: 
  110: Depending on your CPU, the feature you are looking for is called HVM,
  111: SVM or VMX.
  112: 
  113: Next you need to copy the selected Xen kernel itself. pkgsrc installed
  114: them under `/usr/pkg/xen*-kernel/`. The file you're looking for is
  115: `xen.gz`. Copy it to your root file system. `xen-debug.gz` is a kernel
  116: with more consistency checks and more details printed on the serial
  117: console. It is useful for debugging crashing guests if you use a serial
  118: console. It is not useful with a VGA console.
  119: 
  120: You'll then need a NetBSD/Xen kernel for *domain0* on your root file
  121: system. The XEN3PAE\_DOM0 kernel or XEN3\_DOM0 provided as part of the
  122: i386 or amd64 binaries is suitable for this, but you may want to
  123: customize it. Keep your native kernel around, as it can be useful for
  124: recovery. *Note:* the *domain0* kernel must support KERNFS and `/kern`
  125: must be mounted because *xend* needs access to `/kern/xen/privcmd`.
  126: 
  127: Next you need to get a bootloader to load the `xen.gz` kernel, and the
  128: NetBSD *domain0* kernel as a module. This can be `grub` or NetBSD's boot
  129: loader. Below is a detailled example for grub, see the boot.cfg(5)
  130: manual page for an example using the latter.
  131: 
  132: This is also where you'll specify the memory allocated to *domain0*, the
  133: console to use, etc ...
  134: 
  135: Here is a commented `/grub/menu.lst` file:
  136: 
  137:     #Grub config file for NetBSD/xen. Copy as /grub/menu.lst and run
  138:     # grub-install /dev/rwd0d (assuming your boot device is wd0).
  139:     #
  140:     # The default entry to load will be the first one
  141:     default=0
  142: 
  143:     # boot the default entry after 10s if the user didn't hit keyboard
  144:     timeout=10
  145: 
  146:     # Configure serial port to use as console. Ignore if you'll use VGA only
  147:     serial --unit=0 --speed=115200 --word=8 --parity=no --stop=1
  148: 
  149:     # Let the user select which console to use (serial or VGA), default
  150:     # to serial after 10s
  151:     terminal --timeout=10 serial console
  152: 
  153:     # An entry for NetBSD/xen, using /netbsd as the domain0 kernel, and serial
  154:     # console. Domain0 will have 64MB RAM allocated.
  155:     # Assume NetBSD is installed in the first MBR partition.
  156:     title Xen 3 / NetBSD (hda0, serial)
  157:       root(hd0,0)
  158:       kernel (hd0,a)/xen.gz dom0_mem=65536 com1=115200,8n1
  159:       module (hd0,a)/netbsd bootdev=wd0a ro console=ttyS0
  160: 
  161:     # Same as above, but using VGA console
  162:     # We can use console=tty0 (Linux syntax) or console=pc (NetBSD syntax)
  163:     title Xen 3 / NetBSD (hda0, vga)
  164:       root(hd0,0)
  165:       kernel (hd0,a)/xen.gz dom0_mem=65536
  166:       module (hd0,a)/netbsd bootdev=wd0a ro console=tty0
  167: 
  168:     # NetBSD/xen using a backup domain0 kernel (in case you installed a
  169:     # nonworking kernel as /netbsd
  170:     title Xen 3 / NetBSD (hda0, backup, serial)
  171:       root(hd0,0)
  172:       kernel (hd0,a)/xen.gz dom0_mem=65536 com1=115200,8n1
  173:       module (hd0,a)/netbsd.backup bootdev=wd0a ro console=ttyS0
  174:     title Xen 3 / NetBSD (hda0, backup, VGA)
  175:       root(hd0,0)
  176:       kernel (hd0,a)/xen.gz dom0_mem=65536
  177:       module (hd0,a)/netbsd.backup bootdev=wd0a ro console=tty0
  178: 
  179:     #Load a regular NetBSD/i386 kernel. Can be useful if you end up with a
  180:     #nonworking /xen.gz
  181:     title NetBSD 5.1
  182:       root (hd0,a)
  183:       kernel --type=netbsd /netbsd-GENERIC
  184: 
  185:     #Load the NetBSD bootloader, letting it load the NetBSD/i386 kernel.
  186:     #May be better than the above, as grub can't pass all required infos
  187:     #to the NetBSD/i386 kernel (e.g. console, root device, ...)
  188:     title NetBSD chain
  189:       root        (hd0,0)
  190:       chainloader +1
  191: 
  192:     ## end of grub config file.
  193:           
  194: 
  195: Install grub with the following command:
  196: 
  197:     # grub --no-floppy
  198: 
  199:     grub> root (hd0,a)
  200:      Filesystem type is ffs, partition type 0xa9
  201: 
  202:     grub> setup (hd0)
  203:      Checking if "/boot/grub/stage1" exists... no
  204:      Checking if "/grub/stage1" exists... yes
  205:      Checking if "/grub/stage2" exists... yes
  206:      Checking if "/grub/ffs_stage1_5" exists... yes
  207:      Running "embed /grub/ffs_stage1_5 (hd0)"...  14 sectors are embedded.
  208:     succeeded
  209:      Running "install /grub/stage1 (hd0) (hd0)1+14 p (hd0,0,a)/grub/stage2 /grub/menu.lst"...
  210:      succeeded
  211:     Done.
  212:           
  213: 
  214: Creating an unprivileged NetBSD domain (DomU)
  215: ---------------------------------------------
  216: 
  217: Once you have *domain0* running, you need to start the xen tool daemon
  218: (`/usr/pkg/share/examples/rc.d/xend start`) and the xen backend daemon
  219: (`/usr/pkg/share/examples/rc.d/xenbackendd start` for Xen3\*,
  220: `/usr/pkg/share/examples/rc.d/xencommons start` for Xen4.\*). Make sure
  221: that `/dev/xencons` and `/dev/xenevt` exist before starting `xend`. You
  222: can create them with this command:
  223: 
  224:     # cd /dev && sh MAKEDEV xen
  225: 
  226: xend will write logs to `/var/log/xend.log` and
  227: `/var/log/xend-debug.log`. You can then control xen with the xm tool.
  228: 'xm list' will show something like:
  229: 
  230:     # xm list
  231:     Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  232:     Domain-0           0       64    0  r----     58.1
  233: 
  234: 'xm create' allows you to create a new domain. It uses a config file in
  235: PKG\_SYSCONFDIR for its parameters. By default, this file will be in
  236: `/usr/pkg/etc/xen/`. On creation, a kernel has to be specified, which
  237: will be executed in the new domain (this kernel is in the *domain0* file
  238: system, not on the new domain virtual disk; but please note, you should
  239: install the same kernel into *domainU* as `/netbsd` in order to make
  240: your system tools, like MAN.SAVECORE.8, work). A suitable kernel is
  241: provided as part of the i386 and amd64 binary sets: XEN3\_DOMU.
  242: 
  243: Here is an /usr/pkg/etc/xen/nbsd example config file:
  244: 
  245:     #  -*- mode: python; -*-
  246:     #============================================================================
  247:     # Python defaults setup for 'xm create'.
  248:     # Edit this file to reflect the configuration of your system.
  249:     #============================================================================
  250: 
  251:     #----------------------------------------------------------------------------
  252:     # Kernel image file. This kernel will be loaded in the new domain.
  253:     kernel = "/home/bouyer/netbsd-XEN3_DOMU"
  254:     #kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  255: 
  256:     # Memory allocation (in megabytes) for the new domain.
  257:     memory = 128
  258: 
  259:     # A handy name for your new domain. This will appear in 'xm list',
  260:     # and you can use this as parameters for xm in place of the domain
  261:     # number. All domains must have different names.
  262:     #
  263:     name = "nbsd"
  264: 
  265:     # The number of virtual CPUs this domain has.
  266:     #
  267:     vcpus = 1
  268: 
  269:     #----------------------------------------------------------------------------
  270:     # Define network interfaces for the new domain.
  271: 
  272:     # Number of network interfaces (must be at least 1). Default is 1.
  273:     nics = 1
  274: 
  275:     # Define MAC and/or bridge for the network interfaces.
  276:     #
  277:     # The MAC address specified in ``mac'' is the one used for the interface
  278:     # in the new domain. The interface in domain0 will use this address XOR'd
  279:     # with 00:00:00:01:00:00 (i.e. aa:00:00:51:02:f0 in our example). Random
  280:     # MACs are assigned if not given.
  281:     #
  282:     # ``bridge'' is a required parameter, which will be passed to the
  283:     # vif-script called by xend(8) when a new domain is created to configure
  284:     # the new xvif interface in domain0.
  285:     #
  286:     # In this example, the xvif is added to bridge0, which should have been
  287:     # set up prior to the new domain being created -- either in the
  288:     # ``network'' script or using a /etc/ifconfig.bridge0 file.
  289:     #
  290:     vif = [ 'mac=aa:00:00:50:02:f0, bridge=bridge0' ]
  291: 
  292:     #----------------------------------------------------------------------------
  293:     # Define the disk devices you want the domain to have access to, and
  294:     # what you want them accessible as.
  295:     #
  296:     # Each disk entry is of the form:
  297:     #
  298:     #   phy:DEV,VDEV,MODE
  299:     #
  300:     # where DEV is the device, VDEV is the device name the domain will see,
  301:     # and MODE is r for read-only, w for read-write.  You can also create
  302:     # file-backed domains using disk entries of the form:
  303:     #
  304:     #   file:PATH,VDEV,MODE
  305:     #
  306:     # where PATH is the path to the file used as the virtual disk, and VDEV
  307:     # and MODE have the same meaning as for ``phy'' devices.
  308:     #
  309:     # VDEV doesn't really matter for a NetBSD guest OS (it's just used as an index),
  310:     # but it does for Linux.
  311:     # Worse, the device has to exist in /dev/ of domain0, because xm will
  312:     # try to stat() it. This means that in order to load a Linux guest OS
  313:     # from a NetBSD domain0, you'll have to create /dev/hda1, /dev/hda2, ...
  314:     # on domain0, with the major/minor from Linux :(
  315:     # Alternatively it's possible to specify the device number in hex,
  316:     # e.g. 0x301 for /dev/hda1, 0x302 for /dev/hda2, etc ...
  317: 
  318:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  319:     #disk = [ 'file:/var/xen/nbsd-disk,0x01,w' ]
  320:     #disk = [ 'file:/var/xen/nbsd-disk,0x301,w' ]
  321: 
  322:     #----------------------------------------------------------------------------
  323:     # Set the kernel command line for the new domain.
  324: 
  325:     # Set root device. This one does matter for NetBSD
  326:     root = "xbd0"
  327:     # extra parameters passed to the kernel
  328:     # this is where you can set boot flags like -s, -a, etc ...
  329:     #extra = ""
  330: 
  331:     #----------------------------------------------------------------------------
  332:     # Set according to whether you want the domain restarted when it exits.
  333:     # The default is False.
  334:     #autorestart = True
  335: 
  336:     # end of nbsd config file ====================================================
  337: 
  338: When a new domain is created, xen calls the
  339: `/usr/pkg/etc/xen/vif-bridge` script for each virtual network interface
  340: created in *domain0*. This can be used to automatically configure the
  341: xvif?.? interfaces in *domain0*. In our example, these will be bridged
  342: with the bridge0 device in *domain0*, but the bridge has to exist first.
  343: To do this, create the file `/etc/ifconfig.bridge0` and make it look
  344: like this:
  345: 
  346:     create
  347:     !brconfig $int add ex0 up
  348: 
  349: (replace `ex0` with the name of your physical interface). Then bridge0
  350: will be created on boot. See the MAN.BRIDGE.4 man page for details.
  351: 
  352: So, here is a suitable `/usr/pkg/etc/xen/vif-bridge` for xvif?.? (a
  353: working vif-bridge is also provided with xentools20) configuring:
  354: 
  355:     #!/bin/sh
  356:     #============================================================================
  357:     # $NetBSD: howto.mdwn,v 1.7 2013/11/04 02:43:35 mspo Exp $
  358:     #
  359:     # /usr/pkg/etc/xen/vif-bridge
  360:     #
  361:     # Script for configuring a vif in bridged mode with a dom0 interface.
  362:     # The xend(8) daemon calls a vif script when bringing a vif up or down.
  363:     # The script name to use is defined in /usr/pkg/etc/xen/xend-config.sxp
  364:     # in the ``vif-script'' field.
  365:     #
  366:     # Usage: vif-bridge up|down [var=value ...]
  367:     #
  368:     # Actions:
  369:     #    up     Adds the vif interface to the bridge.
  370:     #    down   Removes the vif interface from the bridge.
  371:     #
  372:     # Variables:
  373:     #    domain name of the domain the interface is on (required).
  374:     #    vifq   vif interface name (required).
  375:     #    mac    vif MAC address (required).
  376:     #    bridge bridge to add the vif to (required).
  377:     #
  378:     # Example invocation:
  379:     #
  380:     # vif-bridge up domain=VM1 vif=xvif1.0 mac="ee:14:01:d0:ec:af" bridge=bridge0
  381:     #
  382:     #============================================================================
  383: 
  384:     # Exit if anything goes wrong
  385:     set -e
  386: 
  387:     echo "vif-bridge $*"
  388: 
  389:     # Operation name.
  390:     OP=$1; shift
  391: 
  392:     # Pull variables in args into environment
  393:     for arg ; do export "${arg}" ; done
  394: 
  395:     # Required parameters. Fail if not set.
  396:     domain=${domain:?}
  397:     vif=${vif:?}
  398:     mac=${mac:?}
  399:     bridge=${bridge:?}
  400: 
  401:     # Optional parameters. Set defaults.
  402:     ip=${ip:-''}   # default to null (do nothing)
  403: 
  404:     # Are we going up or down?
  405:     case $OP in
  406:     up) brcmd='add' ;;
  407:     down)   brcmd='delete' ;;
  408:     *)
  409:         echo 'Invalid command: ' $OP
  410:         echo 'Valid commands are: up, down'
  411:         exit 1
  412:         ;;
  413:     esac
  414: 
  415:     # Don't do anything if the bridge is "null".
  416:     if [ "${bridge}" = "null" ] ; then
  417:         exit
  418:     fi
  419: 
  420:     # Don't do anything if the bridge doesn't exist.
  421:     if ! ifconfig -l | grep "${bridge}" >/dev/null; then
  422:         exit
  423:     fi
  424: 
  425:     # Add/remove vif to/from bridge.
  426:     ifconfig x${vif} $OP
  427:     brconfig ${bridge} ${brcmd} x${vif}
  428: 
  429: Now, running
  430: 
  431:     xm create -c /usr/pkg/etc/xen/nbsd
  432: 
  433: should create a domain and load a NetBSD kernel in it. (Note: `-c`
  434: causes xm to connect to the domain's console once created.) The kernel
  435: will try to find its root file system on xbd0 (i.e., wd0e) which hasn't
  436: been created yet. wd0e will be seen as a disk device in the new domain,
  437: so it will be 'sub-partitioned'. We could attach a ccd to wd0e in
  438: *domain0* and partition it, newfs and extract the NetBSD/i386 or amd64
  439: tarballs there, but there's an easier way: load the
  440: `netbsd-INSTALL_XEN3_DOMU` kernel provided in the NetBSD binary sets.
  441: Like other install kernels, it contains a ramdisk with sysinst, so you
  442: can install NetBSD using sysinst on your new domain.
  443: 
  444: If you want to install NetBSD/Xen with a CDROM image, the following line
  445: should be used in the `/usr/pkg/etc/xen/nbsd` file:
  446: 
  447:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  448: 
  449: After booting the domain, the option to install via CDROM may be
  450: selected. The CDROM device should be changed to `xbd1d`.
  451: 
  452: Once done installing, `halt -p` the new domain (don't reboot or halt, it
  453: would reload the INSTALL\_XEN3\_DOMU kernel even if you changed the
  454: config file), switch the config file back to the XEN3\_DOMU kernel, and
  455: start the new domain again. Now it should be able to use `root on xbd0a`
  456: and you should have a second, functional NetBSD system on your xen
  457: installation.
  458: 
  459: When the new domain is booting you'll see some warnings about *wscons*
  460: and the pseudo-terminals. These can be fixed by editing the files
  461: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
  462: `/etc/ttys`, except *console*, like this:
  463: 
  464:     console "/usr/libexec/getty Pc"         vt100   on secure
  465:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  466:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  467:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  468:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  469: 
  470: Finally, all screens must be commented out from `/etc/wscons.conf`.
  471: 
  472: It is also desirable to add
  473: 
  474:     powerd=YES
  475: 
  476: in rc.conf. This way, the domain will be properly shut down if
  477: `xm shutdown -R` or `xm shutdown -H` is used on the domain0.
  478: 
  479: Your domain should be now ready to work, enjoy.
  480: 
  481: Creating an unprivileged Linux domain (DomU)
  482: --------------------------------------------
  483: 
  484: Creating unprivileged Linux domains isn't much different from
  485: unprivileged NetBSD domains, but there are some details to know.
  486: 
  487: First, the second parameter passed to the disk declaration (the '0x1' in
  488: the example below)
  489: 
  490:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  491: 
  492: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  493: for hda). Linux builds device numbers as: (major \<\< 8 + minor). So,
  494: hda1 which has major 3 and minor 1 on a Linux system will have device
  495: number 0x301. Alternatively, devices names can be used (hda, hdb, ...)
  496: as xentools has a table to map these names to devices numbers. To export
  497: a partition to a Linux guest we can use:
  498: 
  499:     disk = [ 'phy:/dev/wd0e,0x300,w' ]
  500:     root = "/dev/hda1 ro"
  501: 
  502: and it will appear as /dev/hda on the Linux system, and be used as root
  503: partition.
  504: 
  505: To install the Linux system on the partition to be exported to the guest
  506: domain, the following method can be used: install sysutils/e2fsprogs
  507: from pkgsrc. Use mke2fs to format the partition that will be the root
  508: partition of your Linux domain, and mount it. Then copy the files from a
  509: working Linux system, make adjustments in `/etc` (fstab, network
  510: config). It should also be possible to extract binary packages such as
  511: .rpm or .deb directly to the mounted partition using the appropriate
  512: tool, possibly running under NetBSD's Linux emulation. Once the
  513: filesystem has been populated, umount it. If desirable, the filesystem
  514: can be converted to ext3 using tune2fs -j. It should now be possible to
  515: boot the Linux guest domain, using one of the vmlinuz-\*-xenU kernels
  516: available in the Xen binary distribution.
  517: 
  518: To get the linux console right, you need to add:
  519: 
  520:     extra = "xencons=tty1"
  521: 
  522: to your configuration since not all linux distributions auto-attach a
  523: tty to the xen console.
  524: 
  525: Creating an unprivileged Solaris domain (DomU)
  526: ----------------------------------------------
  527: 
  528: Download an Opensolaris [release](http://opensolaris.org/os/downloads/)
  529: or [development snapshot](http://genunix.org/) DVD image. Attach the DVD
  530: image to a MAN.VND.4 device. Copy the kernel and ramdisk filesystem
  531: image to your dom0 filesystem.
  532: 
  533:     dom0# mkdir /root/solaris
  534:     dom0# vnconfig vnd0 osol-1002-124-x86.iso
  535:     dom0# mount /dev/vnd0a /mnt
  536: 
  537:     ## for a 64-bit guest
  538:     dom0# cp /mnt/boot/amd64/x86.microroot /root/solaris
  539:     dom0# cp /mnt/platform/i86xpv/kernel/amd64/unix /root/solaris
  540: 
  541:     ## for a 32-bit guest
  542:     dom0# cp /mnt/boot/x86.microroot /root/solaris
  543:     dom0# cp /mnt/platform/i86xpv/kernel/unix /root/solaris
  544: 
  545:     dom0# umount /mnt
  546:           
  547: 
  548: Keep the MAN.VND.4 configured. For some reason the boot process stalls
  549: unless the DVD image is attached to the guest as a "phy" device. Create
  550: an initial configuration file with the following contents. Substitute
  551: */dev/wd0k* with an empty partition at least 8 GB large.
  552: 
  553:     memory = 640
  554:     name = 'solaris'
  555:     disk = [ 'phy:/dev/wd0k,0,w' ]
  556:     disk += [ 'phy:/dev/vnd0d,6:cdrom,r' ]
  557:     vif = [ 'bridge=bridge0' ]
  558:     kernel = '/root/solaris/unix'
  559:     ramdisk = '/root/solaris/x86.microroot'
  560:     # for a 64-bit guest
  561:     extra = '/platform/i86xpv/kernel/amd64/unix - nowin -B install_media=cdrom'
  562:     # for a 32-bit guest
  563:     #extra = '/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'
  564:           
  565: 
  566: Start the guest.
  567: 
  568:     dom0# xm create -c solaris.cfg
  569:     Started domain solaris
  570:                           v3.3.2 chgset 'unavailable'
  571:     SunOS Release 5.11 Version snv_124 64-bit
  572:     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  573:     Use is subject to license terms.
  574:     Hostname: opensolaris
  575:     Remounting root read/write
  576:     Probing for device nodes ...
  577:     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  578:     Preparing live image for use
  579:     Done mounting Live image
  580:           
  581: 
  582: Make sure the network is configured. Note that it can take a minute for
  583: the xnf0 interface to appear.
  584: 
  585:     opensolaris console login: jack
  586:     Password: jack
  587:     Sun Microsystems Inc.   SunOS 5.11      snv_124 November 2008
  588:     jack@opensolaris:~$ pfexec sh
  589:     sh-3.2# ifconfig -a
  590:     sh-3.2# exit
  591:           
  592: 
  593: Set a password for VNC and start the VNC server which provides the X11
  594: display where the installation program runs.
  595: 
  596:     jack@opensolaris:~$ vncpasswd
  597:     Password: solaris
  598:     Verify: solaris
  599:     jack@opensolaris:~$ cp .Xclients .vnc/xstartup
  600:     jack@opensolaris:~$ vncserver :1
  601:           
  602: 
  603: From a remote machine connect to the VNC server. Use `ifconfig xnf0` on
  604: the guest to find the correct IP address to use.
  605: 
  606:     remote$ vncviewer 172.18.2.99:1
  607:           
  608: 
  609: It is also possible to launch the installation on a remote X11 display.
  610: 
  611:     jack@opensolaris:~$ export DISPLAY=172.18.1.1:0
  612:     jack@opensolaris:~$ pfexec gui-install
  613:            
  614: 
  615: After the GUI installation is complete you will be asked to reboot.
  616: Before that you need to determine the ZFS ID for the new boot filesystem
  617: and update the configuration file accordingly. Return to the guest
  618: console.
  619: 
  620:     jack@opensolaris:~$ pfexec zdb -vvv rpool | grep bootfs
  621:                     bootfs = 43
  622:     ^C
  623:     jack@opensolaris:~$
  624:            
  625: 
  626: The final configuration file should look like this. Note in particular
  627: the last line.
  628: 
  629:     memory = 640
  630:     name = 'solaris'
  631:     disk = [ 'phy:/dev/wd0k,0,w' ]
  632:     vif = [ 'bridge=bridge0' ]
  633:     kernel = '/root/solaris/unix'
  634:     ramdisk = '/root/solaris/x86.microroot'
  635:     extra = '/platform/i86xpv/kernel/amd64/unix -B zfs-bootfs=rpool/43,bootpath="/xpvd/xdf@0:a"'
  636:            
  637: 
  638: Restart the guest to verify it works correctly.
  639: 
  640:     dom0# xm destroy solaris
  641:     dom0# xm create -c solaris.cfg
  642:     Using config file "./solaris.cfg".
  643:     v3.3.2 chgset 'unavailable'
  644:     Started domain solaris
  645:     SunOS Release 5.11 Version snv_124 64-bit
  646:     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
  647:     Use is subject to license terms.
  648:     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
  649:     Hostname: osol
  650:     Configuring devices.
  651:     Loading smf(5) service descriptions: 160/160
  652:     svccfg import warnings. See /var/svc/log/system-manifest-import:default.log .
  653:     Reading ZFS config: done.
  654:     Mounting ZFS filesystems: (6/6)
  655:     Creating new rsa public/private host key pair
  656:     Creating new dsa public/private host key pair
  657: 
  658:     osol console login:
  659:            
  660: 
  661: Using PCI devices in guest domains
  662: ==================================
  663: 
  664: The domain0 can give other domains access to selected PCI devices. This
  665: can allow, for example, a non-privileged domain to have access to a
  666: physical network interface or disk controller. However, keep in mind
  667: that giving a domain access to a PCI device most likely will give the
  668: domain read/write access to the whole physical memory, as PCs don't have
  669: an IOMMU to restrict memory access to DMA-capable device. Also, it's not
  670: possible to export ISA devices to non-domain0 domains (which means that
  671: the primary VGA adapter can't be exported. A guest domain trying to
  672: access the VGA registers will panic).
  673: 
  674: This functionality is only available in NetBSD-5.1 (and later) domain0
  675: and domU. If the domain0 is NetBSD, it has to be running Xen 3.1, as
  676: support has not been ported to later versions at this time.
  677: 
  678: For a PCI device to be exported to a domU, is has to be attached to the
  679: `pciback` driver in domain0. Devices passed to the domain0 via the
  680: pciback.hide boot parameter will attach to `pciback` instead of the
  681: usual driver. The list of devices is specified as `(bus:dev.func)`,
  682: where bus and dev are 2-digit hexadecimal numbers, and func a
  683: single-digit number:
  684: 
  685:     pciback.hide=(00:0a.0)(00:06.0)
  686: 
  687: pciback devices should show up in the domain0's boot messages, and the
  688: devices should be listed in the `/kern/xen/pci` directory.
  689: 
  690: PCI devices to be exported to a domU are listed in the `pci` array of
  691: the domU's config file, with the format `'0000:bus:dev.func'`
  692: 
  693:     pci = [ '0000:00:06.0', '0000:00:0a.0' ]
  694: 
  695: In the domU an `xpci` device will show up, to which one or more pci
  696: busses will attach. Then the PCI drivers will attach to PCI busses as
  697: usual. Note that the default NetBSD DOMU kernels do not have `xpci` or
  698: any PCI drivers built in by default; you have to build your own kernel
  699: to use PCI devices in a domU. Here's a kernel config example:
  700: 
  701:     include         "arch/i386/conf/XEN3_DOMU"
  702:     #include         "arch/i386/conf/XENU"           # in NetBSD 3.0
  703: 
  704:     # Add support for PCI busses to the XEN3_DOMU kernel
  705:     xpci* at xenbus ?
  706:     pci* at xpci ?
  707: 
  708:     # Now add PCI and related devices to be used by this domain
  709:     # USB Controller and Devices
  710: 
  711:     # PCI USB controllers
  712:     uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
  713: 
  714:     # USB bus support
  715:     usb*    at uhci?
  716: 
  717:     # USB Hubs
  718:     uhub*   at usb?
  719:     uhub*   at uhub? port ? configuration ? interface ?
  720: 
  721:     # USB Mass Storage
  722:     umass*  at uhub? port ? configuration ? interface ?
  723:     wd*     at umass?
  724:     # SCSI controllers
  725:     ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
  726: 
  727:     # SCSI bus support (for both ahc and umass)
  728:     scsibus* at scsi?
  729: 
  730:     # SCSI devices
  731:     sd*     at scsibus? target ? lun ?      # SCSI disk drives
  732:     cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
  733: 
  734: Links and further information
  735: =============================
  736: 
  737: -   The HowTo on
  738:     Installing into RAID-1
  739:     gives some hints on using Xen (grub) with NetBSD's RAIDframe
  740: -   Harold Gutch wrote documentation on
  741:     setting up a Linux DomU with a NetBSD Dom0
  742: -   An example of how to use NetBSD's native bootloader to load
  743:     NetBSD/Xen instead of Grub can be found in the i386/amd64 MAN.BOOT.8
  744:     and MAN.BOOT.CFG.5 manpages.
  745: 

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