File:  [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.174: download - view: text, annotated - select for diffs
Fri Nov 13 19:42:09 2020 UTC (2 months ago) by gdt
Branches: MAIN
CVS tags: HEAD
xen: markup

    1: [[!meta title="Xen HowTo"]]
    2: 
    3: Xen is a Type 1 hypervisor which supports running multiple guest operating
    4: systems on a single physical machine. One uses the Xen kernel to control the
    5: CPU, memory and console, a dom0 operating system which mediates access to
    6: other hardware (e.g., disks, network, USB), and one or more domU operating
    7: systems which operate in an unprivileged virtualized environment. IO requests
    8: from the domU systems are forwarded by the Xen hypervisor to the dom0 to be
    9: fulfilled.
   10: 
   11: Xen supports different styles of guests; see [PV on HVM](https://wiki.xen.org/wiki/PV_on_HVM) and [PVH(v2)](https://wiki.xenproject.org/wiki/PVH_(v2)_Domu) for upstream documentation.
   12: 
   13: [[!table data="""
   14: Style of guest	|Supported by NetBSD
   15: PV		|Yes (dom0, domU)
   16: HVM		|Yes (domU)
   17: PVH		|current-only (domU, dom0 not yet)
   18: PVHVM		|current-only (domU)
   19: """]]
   20: 
   21: In Para-Virtualized (PV) mode, the guest OS does not attempt to access
   22: hardware directly, but instead makes hypercalls to the hypervisor; PV
   23: guests must be specifically coded for Xen.
   24: 
   25: In HVM mode, no guest modification is required; however, hardware
   26: support is required, such as VT-x on Intel CPUs and SVM on AMD CPUs.
   27: 
   28: In PVH mode, the guest uses PV interfaces for IO and native interfaces for everything else.
   29: 
   30: In PVHVM mode, the guest mostly runs as HVM, but can use PV drivers for efficiency.
   31: 
   32: \todo Explain about "HVM with PV drivers".
   33: 
   34: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
   35: The dom0 can start one or more domUs.  (Booting is explained in detail
   36: in the dom0 section.)
   37: 
   38: This HOWTO presumes a basic familiarity with the Xen system
   39: architecture, with installing NetBSD on i386/amd64 hardware, and with
   40: installing software from pkgsrc.  See also the [Xen
   41: website](http://www.xenproject.org/).
   42: 
   43: [[!toc]]
   44: 
   45: # Versions and Support
   46: 
   47: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
   48: xenkernel and xentools.  We will refer only to the kernel versions,
   49: but note that both packages must be installed together and must have
   50: matching versions.
   51: 
   52: Versions available in pkgsrc:
   53: 
   54: [[!table data="""
   55: Xen Version	|Package Name	|Xen CPU Support	|xm?	|EOL'ed By Upstream
   56: 4.2		|xenkernel42	|i386 x86_64		|yes	|Yes
   57: 4.5		|xenkernel45	|x86_64			|	|Yes
   58: 4.6		|xenkernel46	|x86_64			|	|Yes
   59: 4.8		|xenkernel48	|x86_64			|	|Yes
   60: 4.11		|xenkernel411	|x86_64			|	|No
   61: """]]
   62: 
   63: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
   64: 
   65: Multiprocessor (SMP) support in NetBSD differs depending on the domain:
   66: 
   67: [[!table data="""
   68: Domain		|Supports SMP
   69: dom0		|No
   70: domU		|Yes
   71: """]]
   72: 
   73: Note: NetBSD support is called XEN3.  However, it does support Xen 4,
   74: because the hypercall interface has remained identical.
   75: 
   76: Older Xen had a python-based management tool called xm, now replaced
   77: by xl.  xm is obsolete, but 4.2 remains in pkgsrc because migrating
   78: from xm to xl is not always trivial, and because 4.2 is the last
   79: version to run on an i386 dom0.
   80: 
   81: Architecture
   82: ------------
   83: 
   84: Xen 4.5 and later runs on x86_64 hardware (the NetBSD amd64 port).
   85: Xen 4.2 can in theory use i386 hardware, but we do not have
   86: recent reports of success.
   87: 
   88: The dom0 system, plus each domU, can be either i386PAE or amd64.
   89: i386 without PAE is not supported.
   90: 
   91: The standard approach is to use NetBSD/amd64 for the dom0.
   92: 
   93: To use an i386PAE dom0 (other than on 4.2), one must build or obtain a
   94: 64bit Xen kernel and install it on the system.
   95: 
   96: For domUs, i386PAE is considered as
   97: [faster](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html)
   98: than amd64.
   99: 
  100: # Creating a dom0
  101: 
  102: In order to install a NetBSD as a dom0, one must first install a normal
  103: NetBSD system, and then pivot the install to a dom0 install by changing
  104: the kernel and boot configuration.
  105: 
  106: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with
  107: 512M it worked reliably.  This does not make sense, but if you see
  108: "not ELF" after Xen boots, try increasing dom0 RAM.
  109: 
  110: Installation of NetBSD
  111: ----------------------
  112: 
  113: [Install NetBSD/amd64](/guide/inst/)
  114: just as you would if you were not using Xen.
  115: 
  116: Installation of Xen
  117: -------------------
  118: 
  119: We will consider that you chose to use Xen 4.8, with NetBSD/amd64 as
  120: dom0. In the dom0, install xenkernel48 and xentools48 from pkgsrc.
  121: 
  122: Once this is done, install the Xen kernel itself:
  123: 
  124: [[!template id=programlisting text="""
  125: # cp /usr/pkg/xen48-kernel/xen.gz /
  126: """]]
  127: 
  128: Then, place a NetBSD XEN3_DOM0 kernel in the `/` directory. Such kernel
  129: can either be compiled manually, or downloaded from the NetBSD FTP, for
  130: example at:
  131: 
  132: [[!template id=programlisting text="""
  133: ftp.netbsd.org/pub/NetBSD/NetBSD-8.0/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
  134: """]]
  135: 
  136: Add a line to /boot.cfg to boot Xen:
  137: 
  138: [[!template id=filecontent name="/boot.cfg" text="""
  139: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
  140: """]]
  141: 
  142: This specifies that the dom0 should have 512MB of ram, leaving the rest
  143: to be allocated for domUs.  To use a serial console, use:
  144: 
  145: [[!template id=filecontent name="/boot.cfg" text="""
  146: menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1
  147: """]]
  148: 
  149: which will use the first serial port for Xen (which counts starting
  150: from 1, unlike NetBSD which counts starting from 0), forcing
  151: speed/parity.  Because the NetBSD command line lacks a
  152: "console=pc" argument, it will use the default "xencons" console device,
  153: which directs the console I/O through Xen to the same console device Xen
  154: itself uses (in this case, the serial port).
  155: 
  156: In an attempt to add performance, one can also add `dom0_max_vcpus=1 dom0_vcpus_pin`,
  157: to force only one vcpu to be provided (since NetBSD dom0 can't use
  158: more) and to pin that vcpu to a physical CPU. Xen has
  159: [many boot options](http://xenbits.xenproject.org/docs/4.8-testing/misc/xen-command-line.html),
  160: and other than dom0 memory and max_vcpus, they are generally not
  161: necessary.
  162: 
  163: Copy the boot scripts into `/etc/rc.d`:
  164: 
  165: [[!template id=programlisting text="""
  166: # cp /usr/pkg/share/examples/rc.d/xen* /etc/rc.d/
  167: """]]
  168: 
  169: Enable `xencommons`:
  170: 
  171: [[!template id=filecontent name="/etc/rc.conf" text="""
  172: xencommons=YES
  173: """]]
  174: 
  175: Now, reboot so that you are running a DOM0 kernel under Xen, rather
  176: than GENERIC without Xen.
  177: 
  178: TODO: Recommend for/against xen-watchdog.
  179: 
  180: Once the reboot is done, use `xl` to inspect Xen's boot messages,
  181: available resources, and running domains.  For example:
  182: 
  183: [[!template id=programlisting text="""
  184: # xl dmesg
  185: ... xen's boot info ...
  186: # xl info
  187: ... available memory, etc ...
  188: # xl list
  189: Name              Id  Mem(MB)  CPU  State  Time(s)  Console
  190: Domain-0           0       64    0  r----     58.1
  191: """]]
  192: 
  193: Xen logs will be in /var/log/xen.
  194: 
  195: ### Issues with xencommons
  196: 
  197: `xencommons` starts `xenstored`, which stores data on behalf of dom0 and
  198: domUs.  It does not currently work to stop and start xenstored.
  199: Certainly all domUs should be shutdown first, following the sort order
  200: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
  201: and is not notified when xenstored exits, leading to not recreating
  202: the state when the new xenstored starts.  Until there's a mechanism to
  203: make this work, one should not expect to be able to restart xenstored
  204: (and thus xencommons).  There is currently no reason to expect that
  205: this will get fixed any time soon.
  206: 
  207: anita (for testing NetBSD)
  208: --------------------------
  209: 
  210: With the setup so far (assuming 4.8/xl), one should be able to run
  211: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
  212: root, because anita must create a domU):
  213: 
  214: [[!template id=programlisting text="""
  215: anita --vmm=xl test file:///usr/obj/i386/
  216: """]]
  217: 
  218: Xen-specific NetBSD issues
  219: --------------------------
  220: 
  221: There are (at least) two additional things different about NetBSD as a
  222: dom0 kernel compared to hardware.
  223: 
  224: One is that the module ABI is different because some of the #defines
  225: change, so one must build modules for Xen.  As of netbsd-7, the build
  226: system does this automatically.
  227: 
  228: The other difference is that XEN3_DOM0 does not have exactly the same
  229: options as GENERIC.  While it is debatable whether or not this is a
  230: bug, users should be aware of this and can simply add missing config
  231: items if desired.
  232: 
  233: Updating NetBSD in a dom0
  234: -------------------------
  235: 
  236: This is just like updating NetBSD on bare hardware, assuming the new
  237: version supports the version of Xen you are running.  Generally, one
  238: replaces the kernel and reboots, and then overlays userland binaries
  239: and adjusts `/etc`.
  240: 
  241: Note that one must update both the non-Xen kernel typically used for
  242: rescue purposes and the DOM0 kernel used with Xen.
  243: 
  244: Converting from grub to /boot
  245: -----------------------------
  246: 
  247: These instructions were used to convert a system from
  248: grub to /boot.  The system was originally installed in February of
  249: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
  250: over time.  Before these commands, it was running NetBSD 6 i386, Xen
  251: 4.1 and grub, much like the message linked earlier in the grub
  252: section.
  253: 
  254: [[!template id=programlisting text="""
  255: # Install MBR bootblocks on both disks.
  256: fdisk -i /dev/rwd0d
  257: fdisk -i /dev/rwd1d
  258: # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
  259: installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
  260: installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
  261: # Install secondary boot loader
  262: cp -p /usr/mdec/boot /
  263: # Create boot.cfg following earlier guidance:
  264: menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M
  265: menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=512M
  266: menu=GENERIC:boot
  267: menu=GENERIC single-user:boot -s
  268: menu=GENERIC.ok:boot netbsd.ok
  269: menu=GENERIC.ok single-user:boot netbsd.ok -s
  270: menu=Drop to boot prompt:prompt
  271: default=1
  272: timeout=30
  273: """]]
  274: 
  275: Upgrading Xen versions
  276: ---------------------
  277: 
  278: Minor version upgrades are trivial.  Just rebuild/replace the
  279: xenkernel version and copy the new xen.gz to `/` (where `/boot.cfg`
  280: references it), and reboot.
  281: 
  282: #Unprivileged domains (domU)
  283: 
  284: This section describes general concepts about domUs.  It does not
  285: address specific domU operating systems or how to install them.  The
  286: config files for domUs are typically in `/usr/pkg/etc/xen`, and are
  287: typically named so that the file name, domU name and the domU's host
  288: name match.
  289: 
  290: The domU is provided with CPU and memory by Xen, configured by the
  291: dom0.  The domU is provided with disk and network by the dom0,
  292: mediated by Xen, and configured in the dom0.
  293: 
  294: Entropy in domUs can be an issue; physical disks and network are on
  295: the dom0.  NetBSD's /dev/random system works, but is often challenged.
  296: 
  297: Config files
  298: ------------
  299: 
  300: See /usr/pkg/share/examples/xen/xlexample*
  301: for a small number of well-commented examples, mostly for running
  302: GNU/Linux.
  303: 
  304: The following is an example minimal domain configuration file. The domU
  305: serves as a network file server.
  306: 
  307: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
  308: name = "domU-id"
  309: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
  310: memory = 1024
  311: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
  312: disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
  313:          'file:/n0/xen/foo-wd1,0x1,w' ]
  314: """]]
  315: 
  316: The domain will have name given in the `name` setting.  The kernel has the
  317: host/domU name in it, so that on the dom0 one can update the various
  318: domUs independently.  The `vif` line causes an interface to be provided,
  319: with a specific mac address (do not reuse MAC addresses!), in bridge
  320: mode.  Two disks are provided, and they are both writable; the bits
  321: are stored in files and Xen attaches them to a vnd(4) device in the
  322: dom0 on domain creation.  The system treats xbd0 as the boot device
  323: without needing explicit configuration.
  324: 
  325: By convention, domain config files are kept in `/usr/pkg/etc/xen`.  Note
  326: that "xl create" takes the name of a config file, while other commands
  327: take the name of a domain.
  328: 
  329: Examples of commands:
  330: 
  331: [[!template id=programlisting text="""
  332: xl create /usr/pkg/etc/xen/foo
  333: xl console domU-id
  334: xl create -c /usr/pkg/etc/xen/foo
  335: xl shutdown domU-id
  336: xl list
  337: """]]
  338: 
  339: Typing `^]` will exit the console session.  Shutting down a domain is
  340: equivalent to pushing the power button; a NetBSD domU will receive a
  341: power-press event and do a clean shutdown.  Shutting down the dom0
  342: will trigger controlled shutdowns of all configured domUs.
  343: 
  344: CPU and memory
  345: --------------
  346: 
  347: A domain is provided with some number of vcpus, up to the number
  348: of CPUs seen by the hypervisor. For a domU, it is controlled
  349: from the config file by the "vcpus = N" directive.
  350: 
  351: A domain is provided with memory; this is controlled in the config
  352: file by "memory = N" (in megabytes).  In the straightforward case, the
  353: sum of the the memory allocated to the dom0 and all domUs must be less
  354: than the available memory.
  355: 
  356: Xen also provides a "balloon" driver, which can be used to let domains
  357: use more memory temporarily.
  358: 
  359: Virtual disks
  360: -------------
  361: 
  362: In domU config files, the disks are defined as a sequence of 3-tuples:
  363: 
  364:  * The first element is "method:/path/to/disk". Common methods are
  365:    "file:" for a file-backed vnd, and "phy:" for something that is already
  366:    a device, such as an LVM logical volume.
  367: 
  368:  * The second element is an artifact of how virtual disks are passed to
  369:    Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
  370:    are given a device name to associate with the disk, and values like
  371:    "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
  372:    as xbd0, the second as xbd1, and so on.  However, xl demands a
  373:    second argument.  The name given is converted to a major/minor by
  374:    calling stat(2) on the name in /dev and this is passed to the domU.
  375:    In the general case, the dom0 and domU can be different operating
  376:    systems, and it is an unwarranted assumption that they have consistent
  377:    numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
  378:    as both dom0 and domU, using values of 0x0 for the first disk and 0x1
  379:    for the second works fine and avoids this issue.  For a GNU/Linux
  380:    guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
  381:    /dev/hda1.
  382: 
  383:  * The third element is "w" for writable disks, and "r" for read-only
  384:    disks.
  385: 
  386: Example:
  387: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text="""
  388: disk = [ 'file:/n0/xen/foo-wd0,0x0,w' ]
  389: """]]
  390: 
  391: Note that NetBSD by default creates only vnd[0123].  If you need more
  392: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
  393: dom0.
  394: 
  395: Note that NetBSD by default creates only xbd[0123].  If you need more
  396: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
  397: 
  398: Virtual Networking
  399: ------------------
  400: 
  401: Xen provides virtual Ethernets, each of which connects the dom0 and a
  402: domU.  For each virtual network, there is an interface "xvifN.M" in
  403: the dom0, and a matching interface xennetM (NetBSD name) in domU index N.
  404: The interfaces behave as if there is an Ethernet with two
  405: adapters connected.  From this primitive, one can construct various
  406: configurations.  We focus on two common and useful cases for which
  407: there are existing scripts: bridging and NAT.
  408: 
  409: With bridging (in the example above), the domU perceives itself to be
  410: on the same network as the dom0.  For server virtualization, this is
  411: usually best.  Bridging is accomplished by creating a bridge(4) device
  412: and adding the dom0's physical interface and the various xvifN.0
  413: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
  414: config file.  The bridge must be set up already in the dom0; an
  415: example /etc/ifconfig.bridge0 is:
  416: 
  417: [[!template id=filecontent name="/etc/ifconfig.bridge0" text="""
  418: create
  419: up
  420: !brconfig bridge0 add wm0
  421: """]]
  422: 
  423: With NAT, the domU perceives itself to be behind a NAT running on the
  424: dom0.  This is often appropriate when running Xen on a workstation.
  425: TODO: NAT appears to be configured by "vif = [ '' ]".
  426: 
  427: The MAC address specified is the one used for the interface in the new
  428: domain.  The interface in dom0 will use this address XOR'd with
  429: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
  430: 
  431: Starting domains automatically
  432: ------------------------------
  433: 
  434: To start domains `domU-netbsd` and `domU-linux` at boot and shut them
  435: down cleanly on dom0 shutdown, add the following in rc.conf:
  436: 
  437: [[!template id=filecontent name="/etc/rc.conf" text="""
  438: xendomains="domU-netbsd domU-linux"
  439: """]]
  440: 
  441: #Creating a domU
  442: 
  443: Creating domUs is almost entirely independent of operating system.  We
  444: have already presented the basics of config files.  Note that you must
  445: have already completed the dom0 setup so that "xl list" works.
  446: 
  447: Creating a NetBSD domU
  448: ----------------------
  449: 
  450: See the earlier config file, and adjust memory.  Decide on how much
  451: storage you will provide, and prepare it (file or LVM).
  452: 
  453: While the kernel will be obtained from the dom0 file system, the same
  454: file should be present in the domU as /netbsd so that tools like
  455: savecore(8) can work.   (This is helpful but not necessary.)
  456: 
  457: The kernel must be specifically for Xen and for use as a domU.  The
  458: i386 and amd64 provide the following kernels:
  459: 
  460:         i386 XEN3PAE_DOMU
  461:         amd64 XEN3_DOMU
  462: 
  463: This will boot NetBSD, but this is not that useful if the disk is
  464: empty.  One approach is to unpack sets onto the disk outside of xen
  465: (by mounting it, just as you would prepare a physical disk for a
  466: system you can't run the installer on).
  467: 
  468: A second approach is to run an INSTALL kernel, which has a miniroot
  469: and can load sets from the network.  To do this, copy the INSTALL
  470: kernel to / and change the kernel line in the config file to:
  471: 
  472:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
  473: 
  474: Then, start the domain as "xl create -c configfile".
  475: 
  476: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
  477: line should be used in the config file.
  478: 
  479:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  480: 
  481: After booting the domain, the option to install via CDROM may be
  482: selected.  The CDROM device should be changed to `xbd1d`.
  483: 
  484: Once done installing, "halt -p" the new domain (don't reboot or halt,
  485: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
  486: config file), switch the config file back to the XEN3_DOMU kernel,
  487: and start the new domain again. Now it should be able to use "root on
  488: xbd0a" and you should have a, functional NetBSD domU.
  489: 
  490: TODO: check if this is still accurate.
  491: When the new domain is booting you'll see some warnings about *wscons*
  492: and the pseudo-terminals. These can be fixed by editing the files
  493: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
  494: `/etc/ttys`, except *console*, like this:
  495: 
  496:     console "/usr/libexec/getty Pc"         vt100   on secure
  497:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
  498:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
  499:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
  500:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
  501: 
  502: Finally, all screens must be commented out from `/etc/wscons.conf`.
  503: 
  504: It is also desirable to add
  505: 
  506:         powerd=YES
  507: 
  508: in rc.conf. This way, the domain will be properly shut down if
  509: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
  510: 
  511: It is not strictly necessary to have a kernel (as /netbsd) in the domU
  512: file system.  However, various programs (e.g. netstat) will use that
  513: kernel to look up symbols to read from kernel virtual memory.  If
  514: /netbsd is not the running kernel, those lookups will fail.  (This is
  515: not really a Xen-specific issue, but because the domU kernel is
  516: obtained from the dom0, it is far more likely to be out of sync or
  517: missing with Xen.)
  518: 
  519: Creating a Linux domU
  520: ---------------------
  521: 
  522: Creating unprivileged Linux domains isn't much different from
  523: unprivileged NetBSD domains, but there are some details to know.
  524: 
  525: First, the second parameter passed to the disk declaration (the '0x1' in
  526: the example below)
  527: 
  528:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
  529: 
  530: does matter to Linux. It wants a Linux device number here (e.g. 0x300
  531: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
  532: So, hda1 which has major 3 and minor 1 on a Linux system will have
  533: device number 0x301.  Alternatively, devices names can be used (hda,
  534: hdb, ...)  as xentools has a table to map these names to devices
  535: numbers.  To export a partition to a Linux guest we can use:
  536: 
  537:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
  538:         root = "/dev/hda1 ro"
  539: 
  540: and it will appear as /dev/hda on the Linux system, and be used as root
  541: partition.
  542: 
  543: To install the Linux system on the partition to be exported to the
  544: guest domain, the following method can be used: install
  545: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
  546: that will be the root partition of your Linux domain, and mount it.
  547: Then copy the files from a working Linux system, make adjustments in
  548: `/etc` (fstab, network config).  It should also be possible to extract
  549: binary packages such as .rpm or .deb directly to the mounted partition
  550: using the appropriate tool, possibly running under NetBSD's Linux
  551: emulation.  Once the file system has been populated, umount it.  If
  552: desirable, the file system can be converted to ext3 using tune2fs -j.
  553: It should now be possible to boot the Linux guest domain, using one of
  554: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
  555: 
  556: To get the Linux console right, you need to add:
  557: 
  558:     extra = "xencons=tty1"
  559: 
  560: to your configuration since not all Linux distributions auto-attach a
  561: tty to the xen console.
  562: 
  563: Creating a Solaris domU
  564: -----------------------
  565: 
  566: See possibly outdated
  567: [Solaris domU instructions](/ports/xen/howto-solaris/).
  568: 
  569: 
  570: PCI passthrough: Using PCI devices in guest domains
  571: ---------------------------------------------------
  572: 
  573: The dom0 can give other domains access to selected PCI
  574: devices. This can allow, for example, a non-privileged domain to have
  575: access to a physical network interface or disk controller.  However,
  576: keep in mind that giving a domain access to a PCI device most likely
  577: will give the domain read/write access to the whole physical memory,
  578: as PCs don't have an IOMMU to restrict memory access to DMA-capable
  579: device.  Also, it's not possible to export ISA devices to non-dom0
  580: domains, which means that the primary VGA adapter can't be exported.
  581: A guest domain trying to access the VGA registers will panic.
  582: 
  583: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
  584: not been ported to later versions at this time.
  585: 
  586: For a PCI device to be exported to a domU, is has to be attached to
  587: the "pciback" driver in dom0.  Devices passed to the dom0 via the
  588: pciback.hide boot parameter will attach to "pciback" instead of the
  589: usual driver.  The list of devices is specified as "(bus:dev.func)",
  590: where bus and dev are 2-digit hexadecimal numbers, and func a
  591: single-digit number:
  592: 
  593:         pciback.hide=(00:0a.0)(00:06.0)
  594: 
  595: pciback devices should show up in the dom0's boot messages, and the
  596: devices should be listed in the `/kern/xen/pci` directory.
  597: 
  598: PCI devices to be exported to a domU are listed in the "pci" array of
  599: the domU's config file, with the format "0000:bus:dev.func".
  600: 
  601:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
  602: 
  603: In the domU an "xpci" device will show up, to which one or more pci
  604: buses will attach.  Then the PCI drivers will attach to PCI buses as
  605: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
  606: or any PCI drivers built in by default; you have to build your own
  607: kernel to use PCI devices in a domU.  Here's a kernel config example;
  608: note that only the "xpci" lines are unusual.
  609: 
  610:         include         "arch/i386/conf/XEN3_DOMU"
  611: 
  612:         # Add support for PCI buses to the XEN3_DOMU kernel
  613:         xpci* at xenbus ?
  614:         pci* at xpci ?
  615: 
  616:         # PCI USB controllers
  617:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
  618: 
  619:         # USB bus support
  620:         usb*    at uhci?
  621: 
  622:         # USB Hubs
  623:         uhub*   at usb?
  624:         uhub*   at uhub? port ? configuration ? interface ?
  625: 
  626:         # USB Mass Storage
  627:         umass*  at uhub? port ? configuration ? interface ?
  628:         wd*     at umass?
  629:         # SCSI controllers
  630:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
  631: 
  632:         # SCSI bus support (for both ahc and umass)
  633:         scsibus* at scsi?
  634: 
  635:         # SCSI devices
  636:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
  637:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
  638: 
  639: 
  640: # Specific Issues
  641: 
  642: ## domU
  643: 
  644: [NetBSD 5 is known to panic.]
  645: (http://mail-index.netbsd.org/port-xen/2018/04/17/msg009181.html)
  646: 
  647: # NetBSD as a domU in a VPS
  648: 
  649: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
  650: hardware.  This section explains how to deal with Xen in a domU as a
  651: virtual private server where you do not control or have access to the
  652: dom0.  This is not intended to be an exhaustive list of VPS providers;
  653: only a few are mentioned that specifically support NetBSD.
  654: 
  655: VPS operators provide varying degrees of access and mechanisms for
  656: configuration.  The big issue is usually how one controls which kernel
  657: is booted, because the kernel is nominally in the dom0 file system (to
  658: which VPS users do not normally have access).  A second issue is how
  659: to install NetBSD.
  660: A VPS user may want to compile a kernel for security updates, to run
  661: npf, run IPsec, or any other reason why someone would want to change
  662: their kernel.
  663: 
  664: One approach is to have an administrative interface to upload a kernel,
  665: or to select from a prepopulated list.  Other approaches are pygrub
  666: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
  667: kernel from the domU file system.  This is closer to a regular physical
  668: computer, where someone who controls a machine can replace the kernel.
  669: 
  670: A second issue is multiple CPUs.  With NetBSD 6, domUs support
  671: multiple vcpus, and it is typical for VPS providers to enable multiple
  672: CPUs for NetBSD domUs.
  673: 
  674: pvgrub
  675: ------
  676: 
  677: pvgrub is a version of grub that uses PV operations instead of BIOS
  678: calls.  It is booted from the dom0 as the domU kernel, and then reads
  679: /grub/menu.lst and loads a kernel from the domU file system.
  680: 
  681: [Panix](http://www.panix.com/) lets users use pvgrub.  Panix reports
  682: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
  683: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
  684: page](http://www.panix.com/v-colo/grub.html), which describes only
  685: Linux but should be updated to cover NetBSD :-).
  686: 
  687: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
  688: their own kernel.  See then [prgmr.com NetBSD
  689: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
  690: (which is in need of updating).
  691: 
  692: It appears that [grub's FFS
  693: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
  694: does not support all aspects of modern FFS, but there are also reports
  695: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
  696: partition for the kernel with the intent that grub can understand it,
  697: which leads to /netbsd not being the actual kernel.  One must remember
  698: to update the special boot partition.
  699: 
  700: pygrub
  701: -------
  702: 
  703: pygrub runs in the dom0 and looks into the domU file system.  This
  704: implies that the domU must have a kernel in a file system in a format
  705: known to pygrub.
  706: 
  707: pygrub doesn't seem to work to load Linux images under NetBSD dom0,
  708: and is inherently less secure than pvgrub due to running inside dom0. For both these
  709: reasons, pygrub should not be used, and is only still present so that
  710: historical DomU images using it still work.
  711: 
  712: As of 2014, pygrub seems to be of mostly historical
  713: interest. New DomUs should use pvgrub.
  714: 
  715: Amazon
  716: ------
  717: 
  718: See the [Amazon EC2 page](/amazon_ec2/).

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