Annotation of wikisrc/ports/xen/howto.mdwn, revision 1.104

1.5       mspo        1: Introduction
1.13      gdt         2: ============
1.1       mspo        3: 
                      4: [![[Xen
1.57      gdt         5: screenshot]](http://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](http://www.netbsd.org/gallery/in-Action/hubertf-xen.png)
1.1       mspo        6: 
1.58      gdt         7: Xen is a hypervisor (or virtual machine monitor) for x86 hardware
1.12      gdt         8: (i686-class or higher), which supports running multiple guest
1.58      gdt         9: operating systems on a single physical machine.  Xen is a Type 1 or
                     10: bare-metal hypervisor; one uses the Xen kernel to control the CPU,
                     11: memory and console, a dom0 operating system which mediates access to
                     12: other hardware (e.g., disks, network, USB), and one or more domU
                     13: operating systems which operate in an unprivileged virtualized
                     14: environment.  IO requests from the domU systems are forwarded by the
                     15: hypervisor (Xen) to the dom0 to be fulfilled.
1.12      gdt        16: 
                     17: Xen supports two styles of guests.  The original is Para-Virtualized
                     18: (PV) which means that the guest OS does not attempt to access hardware
                     19: directly, but instead makes hypercalls to the hypervisor.  This is
                     20: analogous to a user-space program making system calls.  (The dom0
                     21: operating system uses PV calls for some functions, such as updating
                     22: memory mapping page tables, but has direct hardware access for disk
                     23: and network.)   PV guests must be specifically coded for Xen.
                     24: 
                     25: The more recent style is HVM, which means that the guest does not have
                     26: code for Xen and need not be aware that it is running under Xen.
                     27: Attempts to access hardware registers are trapped and emulated.  This
                     28: style is less efficient but can run unmodified guests.
                     29: 
1.29      gdt        30: Generally any amd64 machine will work with Xen and PV guests.  In
                     31: theory i386 computers without amd64 support can be used for Xen <=
                     32: 4.2, but we have no recent reports of this working (this is a hint).
                     33: For HVM guests, the VT or VMX cpu feature (Intel) or SVM/HVM/VT
                     34: (amd64) is needed; "cpuctl identify 0" will show this.  TODO: Clean up
                     35: and check the above features.
1.19      gdt        36: 
1.27      jnemeth    37: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
1.12      gdt        38: The dom0 can start one or more domUs.  (Booting is explained in detail
                     39: in the dom0 section.)
                     40: 
                     41: NetBSD supports Xen in that it can serve as dom0, be used as a domU,
                     42: and that Xen kernels and tools are available in pkgsrc.  This HOWTO
                     43: attempts to address both the case of running a NetBSD dom0 on hardware
1.24      gdt        44: and running domUs under it (NetBSD and other), and also running NetBSD
                     45: as a domU in a VPS.
1.12      gdt        46: 
1.20      gdt        47: Some versions of Xen support "PCI passthrough", which means that
                     48: specific PCI devices can be made available to a specific domU instead
                     49: of the dom0.  This can be useful to let a domU run X11, or access some
                     50: network interface or other peripheral.
                     51: 
1.104   ! gdt        52: NetBSD 6 and earlier supported Xen 2; support was removed fro NetBSD
        !            53: 7.  Xen 2 has been removed from pkgsrc.
1.54      gdt        54: 
1.12      gdt        55: Prerequisites
1.13      gdt        56: -------------
1.12      gdt        57: 
                     58: Installing NetBSD/Xen is not extremely difficult, but it is more
                     59: complex than a normal installation of NetBSD.
1.15      gdt        60: In general, this HOWTO is occasionally overly restrictive about how
                     61: things must be done, guiding the reader to stay on the established
                     62: path when there are no known good reasons to stray.
1.12      gdt        63: 
                     64: This HOWTO presumes a basic familiarity with the Xen system
1.16      gdt        65: architecture.  This HOWTO presumes familiarity with installing NetBSD
                     66: on i386/amd64 hardware and installing software from pkgsrc.
1.27      jnemeth    67: See also the [Xen website](http://www.xenproject.org/).
1.1       mspo       68: 
1.15      gdt        69: Versions of Xen and NetBSD
                     70: ==========================
                     71: 
1.27      jnemeth    72: Most of the installation concepts and instructions are independent
                     73: of Xen version and NetBSD version.  This section gives advice on
                     74: which version to choose.  Versions not in pkgsrc and older unsupported
                     75: versions of NetBSD are intentionally ignored.
1.15      gdt        76: 
                     77: Xen
                     78: ---
                     79: 
                     80: In NetBSD, xen is provided in pkgsrc, via matching pairs of packages
                     81: xenkernel and xentools.  We will refer only to the kernel versions,
                     82: but note that both packages must be installed together and must have
                     83: matching versions.
                     84: 
1.104   ! gdt        85: xenkernel3 provides Xen 3.1.  It is no longer maintained by Xen, and the last applied security patch was in
        !            86: 2011. Thus, it should not be used.  It supports PCI passthrough,
        !            87: which is why people use it anyway. Xen 3.1 supports i386, both PAE and
        !            88: non-PAE.
        !            89: 
        !            90: xenkernel33 provides Xen 3.3.  It is no longer maintained by Xen, and
        !            91: the last applied security patch was in 2012.  Thus, it should not be
        !            92: used.  Xen 3.3 supports i386, but only in PAE mode.  There are no good
        !            93: reasons to run this version.
        !            94: 
        !            95: xenkernel41 provides Xen 4.1.  It is no longer maintained by Xen, but
        !            96: as of 2016-11 received backported security patches.  Xen 4.1 supports
        !            97: i386, but only in PAE mode.  There are no good reasons to run this
        !            98: version.
1.102     gdt        99: 
1.104   ! gdt       100: xenkernel42 provides Xen 4.2.  It is no longer maintained by Xen, but
        !           101: as of 2016-11 received backported security patches.  Xen 4.2 supports
        !           102: i386, but only in PAE mode.  The only reason to run this is if you
        !           103: need to use xm instead of xl, or if you need to run an i386 dom0
        !           104: (because your hardware is i386 only).
        !           105: 
        !           106: xenkernel45 provides Xen 4.5.  It is no longer maintained by Xen, but
        !           107: as of 2016-11 it received security patches.  Xen 4.5 requires an amd64
        !           108: dom0, but domUs can be amd64 or i386 PAE.  TODO: It is either a
        !           109: conservative choice or somewhat old.
        !           110: 
        !           111: xenkernel45 provides Xen 4.6.  It is new to pkgsrc in 2016-05.  It is
        !           112: no longer maintained by Xen, but as of 2016-11 it received security
        !           113: patches.  Xen 4.6 requires an amd64 dom0, but domUs can be amd64 or
        !           114: i386 PAE.  TODO: It is either a somewhat aggressive choice or the
        !           115: standard choice
1.85      gdt       116: 
1.96      gdt       117: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
                    118: 
1.15      gdt       119: Ideally newer versions of Xen will be added to pkgsrc.
                    120: 
1.85      gdt       121: Note that NetBSD support is called XEN3.  It works with Xen 3 and Xen
                    122: 4 because the hypercall interface has been stable.
1.20      gdt       123: 
1.19      gdt       124: Xen command program
                    125: -------------------
                    126: 
1.79      gdt       127: Early Xen used a program called xm to manipulate the system from the
1.19      gdt       128: dom0.  Starting in 4.1, a replacement program with similar behavior
1.79      gdt       129: called xl is provided, but it does not work well in 4.1.  In 4.2, both
                    130: xm and xl work fine.  4.4 is the last version that has xm.  You must
                    131: choose one or the other, because it affects which daemons you run.
1.19      gdt       132: 
1.15      gdt       133: NetBSD
                    134: ------
                    135: 
                    136: The netbsd-5, netbsd-6, netbsd-7, and -current branches are all
                    137: reasonable choices, with more or less the same considerations for
                    138: non-Xen use.  Therefore, netbsd-6 is recommended as the stable version
1.29      gdt       139: of the most recent release for production use.  For those wanting to
                    140: learn Xen or without production stability concerns, netbsd-7 is likely
                    141: most appropriate.
1.15      gdt       142: 
                    143: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
                    144: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
                    145: need SMP; the lack of support is really a problem when using a dom0 as
                    146: a normal computer.)
                    147: 
1.18      gdt       148: Architecture
                    149: ------------
                    150: 
1.29      gdt       151: Xen itself can run on i386 or amd64 machines.  (Practically, almost
1.99      gdt       152: any computer where one would want to run Xen today supports amd64.)
                    153: 
                    154: Xen, the dom0 kernel, and each domU kernel can be either i386 or
                    155: amd64.  When building a xenkernel package, one obtains i386 on an i386
                    156: host, and amd64 on an amd64 host.  If the xen kernel is i386, then the
                    157: dom0 kernel and all domU kernels must be i386.  With an amd64 xen
                    158: kernel, an amd64 dom0 kernel is known to work, and an i386 dom0 kernel
                    159: should in theory work.  An amd64 xen/dom0 is known to support both
                    160: i386 and amd64 domUs.
                    161: 
1.101     gdt       162: i386 dom0 and domU kernels must be PAE (except for Xen 3.1); these are
                    163: built by default.  (Note that emacs (at least) fails if run on i386
                    164: with PAE when built without, and vice versa, presumably due to bugs in
                    165: the undump code.)
1.99      gdt       166: 
                    167: Because of the above, the standard approach is to use amd64 for the
                    168: dom0.
1.29      gdt       169: 
                    170: Xen 4.2 is the last version to support i386 as a host.  TODO: Clarify
1.100     gdt       171: if this is about the CPU, the xen kernel, or the dom0 kernel having to
                    172: be amd64.
                    173: 
1.29      gdt       174: 
1.89      gdt       175: Stability
                    176: ---------
                    177: 
                    178: Mostly, NetBSD as a dom0 or domU is quite stable.
                    179: However, there are some open PRs indicating problems.
                    180: 
1.91      gdt       181:  - [PR 48125](http://gnats.netbsd.org/48125)
                    182:  - [PR 47720](http://gnats.netbsd.org/47720)
1.89      gdt       183: 
                    184: Note also that there are issues with sparse vnd(4) instances, but
                    185: these are not about Xen.
                    186: 
1.15      gdt       187: Recommendation
                    188: --------------
                    189: 
1.18      gdt       190: Therefore, this HOWTO recommends running xenkernel42 (and xentools42),
1.30      gdt       191: xl, the NetBSD 6 stable branch, and to use an amd64 kernel as the
                    192: dom0.  Either the i386 or amd64 of NetBSD may be used as domUs.
1.15      gdt       193: 
1.36      gdt       194: Build problems
                    195: --------------
                    196: 
                    197: Ideally, all versions of Xen in pkgsrc would build on all versions of
                    198: NetBSD on both i386 and amd64.  However, that isn't the case.  Besides
                    199: aging code and aging compilers, qemu (included in xentools for HVM
1.62      gdt       200: support) is difficult to build.  The following are known to work or FAIL:
1.36      gdt       201: 
1.62      gdt       202:         xenkernel3 netbsd-5 amd64
1.63      gdt       203:         xentools3 netbsd-5 amd64
1.64      gdt       204:         xentools3=hvm netbsd-5 amd64 ????
1.62      gdt       205:         xenkernel33 netbsd-5 amd64
1.63      gdt       206:         xentools33 netbsd-5 amd64
1.36      gdt       207:         xenkernel41 netbsd-5 amd64
                    208:         xentools41 netbsd-5 amd64
1.62      gdt       209:         xenkernel42 netbsd-5 amd64
1.64      gdt       210:         xentools42 netbsd-5 amd64
1.62      gdt       211: 
                    212:         xenkernel3 netbsd-6 i386 FAIL
                    213:         xentools3 netbsd-6 i386
1.63      gdt       214:         xentools3-hvm netbsd-6 i386 FAIL (dependencies fail)
                    215:         xenkernel33 netbsd-6 i386
                    216:         xentools33 netbsd-6 i386
1.36      gdt       217:         xenkernel41 netbsd-6 i386
                    218:         xentools41 netbsd-6 i386
1.63      gdt       219:         xenkernel42 netbsd-6 i386
1.64      gdt       220:         xentools42 netbsd-6 i386 *MIXED
                    221: 
1.69      gdt       222:        (all 3 and 33 seem to FAIL)
                    223:         xenkernel41 netbsd-7 i386
                    224:         xentools41 netbsd-7 i386
                    225:         xenkernel42 netbsd-7 i386
                    226:         xentools42 netbsd-7 i386 ??FAIL
                    227: 
1.64      gdt       228: (*On netbsd-6 i386, there is a xentools42 in the 2014Q3 official builds,
                    229: but it does not build for gdt.)
1.36      gdt       230: 
1.15      gdt       231: NetBSD as a dom0
                    232: ================
                    233: 
                    234: NetBSD can be used as a dom0 and works very well.  The following
                    235: sections address installation, updating NetBSD, and updating Xen.
1.19      gdt       236: Note that it doesn't make sense to talk about installing a dom0 OS
                    237: without also installing Xen itself.  We first address installing
                    238: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
                    239: NetBSD install to a dom0 install by just changing the kernel and boot
                    240: configuration.
1.15      gdt       241: 
1.45      gdt       242: For experimenting with Xen, a machine with as little as 1G of RAM and
                    243: 100G of disk can work.  For running many domUs in productions, far
                    244: more will be needed.
                    245: 
1.15      gdt       246: Styles of dom0 operation
                    247: ------------------------
                    248: 
                    249: There are two basic ways to use Xen.  The traditional method is for
                    250: the dom0 to do absolutely nothing other than providing support to some
                    251: number of domUs.  Such a system was probably installed for the sole
                    252: purpose of hosting domUs, and sits in a server room on a UPS.
                    253: 
                    254: The other way is to put Xen under a normal-usage computer, so that the
                    255: dom0 is what the computer would have been without Xen, perhaps a
                    256: desktop or laptop.  Then, one can run domUs at will.  Purists will
                    257: deride this as less secure than the previous approach, and for a
                    258: computer whose purpose is to run domUs, they are right.  But Xen and a
1.93      gdt       259: dom0 (without domUs) is not meaningfully less secure than the same
1.15      gdt       260: things running without Xen.  One can boot Xen or boot regular NetBSD
                    261: alternately with little problems, simply refraining from starting the
                    262: Xen daemons when not running Xen.
                    263: 
                    264: Note that NetBSD as dom0 does not support multiple CPUs.  This will
1.51      gdt       265: limit the performance of the Xen/dom0 workstation approach.  In theory
                    266: the only issue is that the "backend drivers" are not yet MPSAFE:
                    267:   http://mail-index.netbsd.org/netbsd-users/2014/08/29/msg015195.html
1.15      gdt       268: 
1.19      gdt       269: Installation of NetBSD
                    270: ----------------------
1.13      gdt       271: 
1.19      gdt       272: First,
1.27      jnemeth   273: [install NetBSD/amd64](/guide/inst/)
1.19      gdt       274: just as you would if you were not using Xen.
                    275: However, the partitioning approach is very important.
                    276: 
                    277: If you want to use RAIDframe for the dom0, there are no special issues
                    278: for Xen.  Typically one provides RAID storage for the dom0, and the
1.22      gdt       279: domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
                    280: over a RAID1 header to find /boot from a filesystem within a RAID
                    281: partition; this is no different when booting Xen.
1.19      gdt       282: 
                    283: There are 4 styles of providing backing storage for the virtual disks
1.93      gdt       284: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
1.19      gdt       285: 
                    286: With raw partitions, one has a disklabel (or gpt) partition sized for
                    287: each virtual disk to be used by the domU.  (If you are able to predict
                    288: how domU usage will evolve, please add an explanation to the HOWTO.
                    289: Seriously, needs tend to change over time.)
                    290: 
1.27      jnemeth   291: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
                    292: for domU disks.  This is almost as efficient as raw disk partitions
                    293: and more flexible.  Hence raw disk partitions should typically not
                    294: be used.
1.19      gdt       295: 
                    296: One can use files in the dom0 filesystem, typically created by dd'ing
                    297: /dev/zero to create a specific size.  This is somewhat less efficient,
                    298: but very convenient, as one can cp the files for backup, or move them
                    299: between dom0 hosts.
                    300: 
                    301: Finally, in theory one can place the files backing the domU disks in a
                    302: SAN.  (This is an invitation for someone who has done this to add a
                    303: HOWTO page.)
1.1       mspo      304: 
1.19      gdt       305: Installation of Xen
                    306: -------------------
1.1       mspo      307: 
1.20      gdt       308: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
                    309: pkgsrc (or another matching pair).
                    310: See [the pkgsrc
                    311: documentation](http://www.NetBSD.org/docs/pkgsrc/) for help with pkgsrc.
                    312: 
                    313: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm.  More
                    314: recent versions have HVM support integrated in the main xentools
                    315: package.  It is entirely reasonable to run only PV guests.
                    316: 
                    317: Next you need to install the selected Xen kernel itself, which is
                    318: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
                    319: For debugging, one may copy xen-debug.gz; this is conceptually similar
                    320: to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
                    321: useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
                    322: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
1.75      gdt       323: of a NetBSD build.  If using i386, use
                    324: releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.  (If using Xen
                    325: 3.1 and i386, you may use XEN3_DOM0 with the non-PAE Xen.  But you
                    326: should not use Xen 3.1.)  Both xen and the NetBSD kernel may be (and
                    327: typically are) left compressed.
                    328: 
                    329: In a dom0 kernel, kernfs is mandatory for xend to comunicate with the
                    330: kernel, so ensure that /kern is in fstab.  TODO: Say this is default,
                    331: or file a PR and give a reference.
1.20      gdt       332: 
                    333: Because you already installed NetBSD, you have a working boot setup
                    334: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
                    335: beginning of your root filesystem, /boot present, and likely
                    336: /boot.cfg.  (If not, fix before continuing!)
                    337: 
1.76      gdt       338: Add a line to to /boot.cfg to boot Xen.  See boot.cfg(5) for an
                    339: example.  The basic line is
1.20      gdt       340: 
1.37      gdt       341:         menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
1.20      gdt       342: 
                    343: which specifies that the dom0 should have 256M, leaving the rest to be
1.77      gdt       344: allocated for domUs.  To use a serial console, use
                    345: 
                    346:         menu=Xen:load /netbsd-XEN3_DOM0.gz console=com0;multiboot /xen.gz dom0_mem=256M console=com1 com1=9600,8n1
                    347: 
                    348: which will use the first serial port for Xen (which counts starting
                    349: from 1), forcing speed/parity, and also for NetBSD (which counts
                    350: starting at 0).  In an attempt to add performance, one can also add
1.37      gdt       351: 
                    352:         dom0_max_vcpus=1 dom0_vcpus_pin
                    353: 
                    354: to force only one vcpu to be provided (since NetBSD dom0 can't use
                    355: more) and to pin that vcpu to a physical cpu.  TODO: benchmark this.
1.20      gdt       356: 
1.93      gdt       357: Xen has [many boot
                    358: options](http://xenbits.xenproject.org/docs/4.5-testing/misc/xen-command-line.html),
                    359: and other tham dom0 memory and max_vcpus, they are generally not
                    360: necessary.
                    361: 
1.20      gdt       362: As with non-Xen systems, you should have a line to boot /netbsd (a
                    363: kernel that works without Xen) and fallback versions of the non-Xen
                    364: kernel, Xen, and the dom0 kernel.
1.1       mspo      365: 
1.76      gdt       366: Now, reboot so that you are running a DOM0 kernel under Xen, rather
                    367: than GENERIC without Xen.
                    368: 
1.54      gdt       369: Using grub (historic)
                    370: ---------------------
                    371: 
                    372: Before NetBSD's native bootloader could support Xen, the use of
                    373: grub was recommended.  If necessary, see the
                    374: [old grub information](/ports/xen/howto-grub/).
                    375: 
1.28      gdt       376: The [HowTo on Installing into
                    377: RAID-1](http://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
                    378: explains how to set up booting a dom0 with Xen using grub with
                    379: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
                    380: boot.)
                    381: 
1.21      gdt       382: Configuring Xen
                    383: ---------------
                    384: 
1.53      gdt       385: Xen logs will be in /var/log/xen.
                    386: 
1.76      gdt       387: Now, you have a system that will boot Xen and the dom0 kernel, but not
                    388: do anything else special.  Make sure that you have rebooted into Xen.
                    389: There will be no domUs, and none can be started because you still have
1.102     gdt       390: to configure the dom0 daemons.
1.21      gdt       391: 
1.102     gdt       392: The daemons which should be run vary with Xen version and with whether
                    393: one is using xm or xl.  The Xen 3.1 and 3.3 packages use xm.  Xen 4.1
                    394: and higher packages use xl.  While is is possible to use xm with some
                    395: 4.x versions (TODO: 4.1 and 4.2?), the pkgsrc-provided rc.d scripts do
                    396: not support this as of 2014-12-26, and thus the HOWTO does not support
                    397: it either.  (Make sure your packages are reasonably recent.)
1.43      gdt       398: 
1.102     gdt       399: For "xm" (3.1 and 3.3), you should enable xend and xenbackendd (but
                    400: note that you should be using 4.x):
1.31      gdt       401: 
1.32      gdt       402:         xend=YES
                    403:         xenbackendd=YES
1.31      gdt       404: 
1.102     gdt       405: For "xl" (4.x), you should enabled xend and xencommons (xenstored).
                    406: Trying to boot 4.x without xencommons=YES will result in a hang; it is
                    407: necessary to hig ^C on the console to let the machine finish booting.
                    408: TODO: explain why xend is installed by the package.
1.31      gdt       409: 
1.53      gdt       410:         xencommons=YES
1.31      gdt       411: 
1.102     gdt       412: The installation of NetBSD should already have created devices for xen
                    413: (xencons, xenevt), but if they are not present, create them:
1.31      gdt       414: 
1.102     gdt       415:         cd /dev && sh MAKEDEV xen
1.86      gdt       416: 
1.31      gdt       417: TODO: Recommend for/against xen-watchdog.
1.27      jnemeth   418: 
1.53      gdt       419: After you have configured the daemons and either started them (in the
1.79      gdt       420: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
1.102     gdt       421: available resources, and running domains.  An example with xl follows:
1.34      gdt       422: 
1.102     gdt       423:         # xl dmesg
1.43      gdt       424:        [xen's boot info]
1.102     gdt       425:         # xl info
1.43      gdt       426:        [available memory, etc.]
1.102     gdt       427:         # xl list
1.43      gdt       428:         Name              Id  Mem(MB)  CPU  State  Time(s)  Console
                    429:         Domain-0           0       64    0  r----     58.1
1.33      gdt       430: 
1.88      gdt       431: ### Issues with xencommons
                    432: 
                    433: xencommons starts xenstored, which stores data on behalf of dom0 and
                    434: domUs.  It does not currently work to stop and start xenstored.
                    435: Certainly all domUs should be shutdown first, following the sort order
                    436: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
                    437: and is not notified when xenstored exits, leading to not recreating
                    438: the state when the new xenstored starts.  Until there's a mechanism to
                    439: make this work, one should not expect to be able to restart xenstored
                    440: (and thus xencommons).  There is currently no reason to expect that
                    441: this will get fixed any time soon.
                    442: 
1.41      gdt       443: anita (for testing NetBSD)
                    444: --------------------------
                    445: 
1.82      gdt       446: With the setup so far (assuming 4.2/xl), one should be able to run
                    447: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
                    448: root, because anita must create a domU):
                    449: 
                    450:         anita --vmm=xl test file:///usr/obj/i386/
                    451: 
                    452: Alternatively, one can use --vmm=xm to use xm-based domU creation
                    453: instead (and must, on Xen <= 4.1).   TODO: confirm that anita xl really works.
                    454:     
1.40      gdt       455: Xen-specific NetBSD issues
                    456: --------------------------
                    457: 
                    458: There are (at least) two additional things different about NetBSD as a
                    459: dom0 kernel compared to hardware.
                    460: 
                    461: One is that modules are not usable in DOM0 kernels, so one must
                    462: compile in what's needed.  It's not really that modules cannot work,
                    463: but that modules must be built for XEN3_DOM0 because some of the
                    464: defines change and the normal module builds don't do this.  Basically,
                    465: enabling Xen changes the kernel ABI, and the module build system
                    466: doesn't cope with this.
                    467: 
                    468: The other difference is that XEN3_DOM0 does not have exactly the same
                    469: options as GENERIC.  While it is debatable whether or not this is a
                    470: bug, users should be aware of this and can simply add missing config
                    471: items if desired.
                    472: 
1.15      gdt       473: Updating NetBSD in a dom0
                    474: -------------------------
                    475: 
                    476: This is just like updating NetBSD on bare hardware, assuming the new
                    477: version supports the version of Xen you are running.  Generally, one
                    478: replaces the kernel and reboots, and then overlays userland binaries
                    479: and adjusts /etc.
                    480: 
                    481: Note that one must update both the non-Xen kernel typically used for
                    482: rescue purposes and the DOM0 kernel used with Xen.
                    483: 
1.55      gdt       484: Converting from grub to /boot
                    485: -----------------------------
                    486: 
                    487: These instructions were [TODO: will be] used to convert a system from
                    488: grub to /boot.  The system was originally installed in February of
                    489: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
                    490: over time.  Before these commands, it was running NetBSD 6 i386, Xen
                    491: 4.1 and grub, much like the message linked earlier in the grub
                    492: section.
                    493: 
                    494:         # Install mbr bootblocks on both disks. 
                    495:         fdisk -i /dev/rwd0d
                    496:         fdisk -i /dev/rwd1d
                    497:         # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
                    498:         installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
                    499:         installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
                    500:         # Install secondary boot loader
                    501:         cp -p /usr/mdec/boot /
                    502:         # Create boog.cfg following earlier guidance:
                    503:         menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
                    504:         menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=256M
                    505:         menu=GENERIC:boot
                    506:         menu=GENERIC single-user:boot -s
                    507:         menu=GENERIC.ok:boot netbsd.ok
                    508:         menu=GENERIC.ok single-user:boot netbsd.ok -s
                    509:         menu=Drop to boot prompt:prompt
                    510:         default=1
                    511:         timeout=30
                    512: 
                    513: TODO: actually do this and fix it if necessary.
1.22      gdt       514: 
1.102     gdt       515: Upgrading Xen versions
1.15      gdt       516: ---------------------
                    517: 
1.21      gdt       518: Updating Xen is conceptually not difficult, but can run into all the
                    519: issues found when installing Xen.  Assuming migration from 4.1 to 4.2,
                    520: remove the xenkernel41 and xentools41 packages and install the
                    521: xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz to /.
                    522: 
1.102     gdt       523: Ensure that the contents of /etc/rc.d/xen* are correct.  Specifically,
                    524: they must match the package you just installed and not be left over
                    525: from some previous installation.
                    526: 
                    527: Enable the correct set of daemons; see the configuring section above.
                    528: (Upgrading from 3.x to 4.x without doing this will result in a hang.)
                    529: 
                    530: Ensure that the domU config files are valid for the new version.
                    531: Specifically: remove autorestart=True, and ensure that disks are
                    532: specified with numbers as the second argument, as the examples above
                    533: show, and not NetBSD device names.
1.15      gdt       534: 
1.97      gdt       535: Hardware known to work
                    536: ----------------------
                    537: 
                    538: Arguably, this section is misplaced, and there should be a page of
                    539: hardware that runs NetBSD/amd64 well, with the mostly-well-founded
                    540: assumption that NetBSD/xen runs fine on any modern hardware that
                    541: NetBSD/amd64 runs well on.  Until then, we give motherboard/CPU/RAM
                    542: triples to aid those choosing a motherboard.  Note that Xen systems
                    543: usually do not run X, so a listing here does not imply that X works at
                    544: all.
                    545: 
                    546:         Supermicro X9SRL-F, Xeon E5-1650 v2, 96 GiB ECC
                    547:         Supermicro ??, Atom C2758 (8 core), 32 GiB ECC
                    548:         ASUS M5A78L-M/USB3 AM3+ microATX, AMD Piledriver X8 4000MHz, 16 GiB ECC
                    549: 
                    550: Older hardware:
                    551: 
1.98      gdt       552:         Intel D915GEV, Pentium4 CPU 3.40GHz, 4GB 533MHz Synchronous DDR2
1.28      gdt       553: 
1.82      gdt       554: Running Xen under qemu
                    555: ----------------------
                    556: 
                    557: The astute reader will note that this section is somewhat twisted.
                    558: However, it can be useful to run Xen under qemu either because the
                    559: version of NetBSD as a dom0 does not run on the hardware in use, or to
                    560: generate automated test cases involving Xen.
                    561: 
1.84      gdt       562: In 2015-01, the following combination was reported to mostly work:
1.82      gdt       563: 
                    564:         host OS: NetBSD/amd64 6.1.4
                    565:         qemu: 2.2.0 from pkgsrc
                    566:         Xen kernel: xenkernel42-4.2.5nb1 from pkgsrc
                    567:         dom0 kernel: NetBSD/amd64 6.1.5
                    568:         Xen tools: xentools42-4.2.5 from pkgsrc
                    569: 
1.91      gdt       570: See [PR 47720](http://gnats.netbsd.org/47720) for a problem with dom0
                    571: shutdown.
1.84      gdt       572: 
1.28      gdt       573: Unprivileged domains (domU)
                    574: ===========================
                    575: 
                    576: This section describes general concepts about domUs.  It does not
1.33      gdt       577: address specific domU operating systems or how to install them.  The
                    578: config files for domUs are typically in /usr/pkg/etc/xen, and are
1.60      wiki      579: typically named so that the file name, domU name and the domU's host
1.33      gdt       580: name match.
                    581: 
                    582: The domU is provided with cpu and memory by Xen, configured by the
                    583: dom0.  The domU is provided with disk and network by the dom0,
                    584: mediated by Xen, and configured in the dom0.
                    585: 
                    586: Entropy in domUs can be an issue; physical disks and network are on
                    587: the dom0.  NetBSD's /dev/random system works, but is often challenged.
                    588: 
1.48      gdt       589: Config files
                    590: ------------
                    591: 
                    592: There is no good order to present config files and the concepts
                    593: surrounding what is being configured.  We first show an example config
                    594: file, and then in the various sections give details.
                    595: 
                    596: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
                    597: for a large number of well-commented examples, mostly for running
                    598: GNU/Linux.
                    599: 
                    600: The following is an example minimal domain configuration file
                    601: "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
                    602: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
                    603: i386 domU).  The domU serves as a network file server.
                    604: 
                    605:         # -*- mode: python; -*-
                    606: 
                    607:         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
                    608:         memory = 1024
                    609:         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
                    610:         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
                    611:                  'file:/n0/xen/foo-wd1,0x1,w' ]
                    612: 
                    613: The domain will have the same name as the file.  The kernel has the
                    614: host/domU name in it, so that on the dom0 one can update the various
                    615: domUs independently.  The vif line causes an interface to be provided,
                    616: with a specific mac address (do not reuse MAC addresses!), in bridge
                    617: mode.  Two disks are provided, and they are both writable; the bits
                    618: are stored in files and Xen attaches them to a vnd(4) device in the
                    619: dom0 on domain creation.  The system treates xbd0 as the boot device
                    620: without needing explicit configuration.
                    621: 
                    622: By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
                    623: that "xm create" takes the name of a config file, while other commands
                    624: take the name of a domain.  To create the domain, connect to the
                    625: console, create the domain while attaching the console, shutdown the
                    626: domain, and see if it has finished stopping, do (or xl with Xen >=
                    627: 4.2):
                    628: 
                    629:         xm create foo
                    630:         xm console foo
                    631:         xm create -c foo
                    632:         xm shutdown foo
1.90      gdt       633:         xm list
1.48      gdt       634: 
                    635: Typing ^] will exit the console session.  Shutting down a domain is
                    636: equivalent to pushing the power button; a NetBSD domU will receive a
                    637: power-press event and do a clean shutdown.  Shutting down the dom0
                    638: will trigger controlled shutdowns of all configured domUs.
                    639: 
                    640: domU kernels
                    641: ------------
                    642: 
                    643: On a physical computer, the BIOS reads sector 0, and a chain of boot
                    644: loaders finds and loads a kernel.  Normally this comes from the root
                    645: filesystem.  With Xen domUs, the process is totally different.  The
                    646: normal path is for the domU kernel to be a file in the dom0's
                    647: filesystem.  At the request of the dom0, Xen loads that kernel into a
                    648: new domU instance and starts execution.  While domU kernels can be
                    649: anyplace, reasonable places to store domU kernels on the dom0 are in /
                    650: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
                    651: config files), or in /u0/xen (where the vdisks are).
                    652: 
1.59      gdt       653: Note that loading the domU kernel from the dom0 implies that boot
                    654: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
1.48      gdt       655: See the VPS section near the end for discussion of alternate ways to
                    656: obtain domU kernels.
                    657: 
1.33      gdt       658: CPU and memory
                    659: --------------
                    660: 
1.48      gdt       661: A domain is provided with some number of vcpus, less than the number
                    662: of cpus seen by the hypervisor.  (For a dom0, this is controlled by
                    663: the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
                    664: from the config file by the "vcpus = N" directive.
                    665: 
                    666: A domain is provided with memory; this is controlled in the config
                    667: file by "memory = N" (in megabytes).  In the straightforward case, the
                    668: sum of the the memory allocated to the dom0 and all domUs must be less
1.33      gdt       669: than the available memory.
                    670: 
                    671: Xen also provides a "balloon" driver, which can be used to let domains
                    672: use more memory temporarily.  TODO: Explain better, and explain how
                    673: well it works with NetBSD.
1.28      gdt       674: 
                    675: Virtual disks
                    676: -------------
                    677: 
1.33      gdt       678: With the file/vnd style, typically one creates a directory,
                    679: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
                    680: domUs.  Then, for each domU disk, one writes zeros to a file that then
                    681: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
                    682: for the first virtual disk for the domU called foo.  Writing zeros to
                    683: the file serves two purposes.  One is that preallocating the contents
                    684: improves performance.  The other is that vnd on sparse files has
                    685: failed to work.  TODO: give working/notworking NetBSD versions for
                    686: sparse vnd.  Note that the use of file/vnd for Xen is not really
                    687: different than creating a file-backed virtual disk for some other
1.39      gdt       688: purpose, except that xentools handles the vnconfig commands.  To
                    689: create an empty 4G virtual disk, simply do
                    690: 
                    691:         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
1.33      gdt       692: 
1.89      gdt       693: Do not use qemu-img-xen, because this will create sparse file.  There
                    694: have been recent (2015) reports of sparse vnd(4) devices causing
                    695: lockups, but there is apparently no PR.
                    696: 
1.33      gdt       697: With the lvm style, one creates logical devices.  They are then used
1.48      gdt       698: similarly to vnds.  TODO: Add an example with lvm.
                    699: 
                    700: In domU config files, the disks are defined as a sequence of 3-tuples.
                    701: The first element is "method:/path/to/disk".  Common methods are
                    702: "file:" for file-backed vnd. and "phy:" for something that is already
                    703: a (TODO: character or block) device.
                    704: 
                    705: The second element is an artifact of how virtual disks are passed to
                    706: Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
                    707: are given a device name to associate with the disk, and values like
                    708: "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
                    709: as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
                    710: second argument.  The name given is converted to a major/minor by
1.49      gdt       711: calling stat(2) on the name in /dev and this is passed to the domU.
                    712: In the general case, the dom0 and domU can be different operating
1.48      gdt       713: systems, and it is an unwarranted assumption that they have consistent
                    714: numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
                    715: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
1.49      gdt       716: for the second works fine and avoids this issue.  For a GNU/Linux
                    717: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
                    718: /dev/hda1.
1.48      gdt       719: 
                    720: The third element is "w" for writable disks, and "r" for read-only
                    721: disks.
1.28      gdt       722: 
                    723: Virtual Networking
                    724: ------------------
                    725: 
1.46      gdt       726: Xen provides virtual ethernets, each of which connects the dom0 and a
                    727: domU.  For each virtual network, there is an interface "xvifN.M" in
                    728: the dom0, and in domU index N, a matching interface xennetM (NetBSD
                    729: name).  The interfaces behave as if there is an Ethernet with two
                    730: adaptors connected.  From this primitive, one can construct various
                    731: configurations.  We focus on two common and useful cases for which
                    732: there are existing scripts: bridging and NAT.
1.28      gdt       733: 
1.48      gdt       734: With bridging (in the example above), the domU perceives itself to be
                    735: on the same network as the dom0.  For server virtualization, this is
                    736: usually best.  Bridging is accomplished by creating a bridge(4) device
                    737: and adding the dom0's physical interface and the various xvifN.0
                    738: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
                    739: config file.  The bridge must be set up already in the dom0; an
                    740: example /etc/ifconfig.bridge0 is:
1.46      gdt       741: 
                    742:         create
                    743:         up
                    744:         !brconfig bridge0 add wm0
1.28      gdt       745: 
                    746: With NAT, the domU perceives itself to be behind a NAT running on the
                    747: dom0.  This is often appropriate when running Xen on a workstation.
1.48      gdt       748: TODO: NAT appears to be configured by "vif = [ '' ]".
1.28      gdt       749: 
1.49      gdt       750: The MAC address specified is the one used for the interface in the new
1.53      gdt       751: domain.  The interface in dom0 will use this address XOR'd with
1.49      gdt       752: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
                    753: 
1.33      gdt       754: Sizing domains
                    755: --------------
                    756: 
                    757: Modern x86 hardware has vast amounts of resources.  However, many
                    758: virtual servers can function just fine on far less.  A system with
                    759: 256M of RAM and a 4G disk can be a reasonable choice.  Note that it is
                    760: far easier to adjust virtual resources than physical ones.  For
                    761: memory, it's just a config file edit and a reboot.  For disk, one can
                    762: create a new file and vnconfig it (or lvm), and then dump/restore,
                    763: just like updating physical disks, but without having to be there and
                    764: without those pesky connectors.
                    765: 
1.48      gdt       766: Starting domains automatically
                    767: ------------------------------
1.28      gdt       768: 
1.48      gdt       769: To start domains foo at bar at boot and shut them down cleanly on dom0
                    770: shutdown, in rc.conf add:
1.28      gdt       771: 
1.48      gdt       772:         xendomains="foo bar"
1.28      gdt       773: 
1.86      gdt       774: Note that earlier versions of the xentools41 xendomains rc.d scripth
                    775: usd xl, when one should use xm with 4.1.
1.28      gdt       776: 
                    777: Creating specific unprivileged domains (domU)
                    778: =============================================
1.14      gdt       779: 
                    780: Creating domUs is almost entirely independent of operating system.  We
1.49      gdt       781: have already presented the basics of config files.  Note that you must
                    782: have already completed the dom0 setup so that "xl list" (or "xm list")
                    783: works.
1.14      gdt       784: 
                    785: Creating an unprivileged NetBSD domain (domU)
                    786: ---------------------------------------------
1.1       mspo      787: 
1.49      gdt       788: See the earlier config file, and adjust memory.  Decide on how much
                    789: storage you will provide, and prepare it (file or lvm).
                    790: 
                    791: While the kernel will be obtained from the dom0 filesystem, the same
                    792: file should be present in the domU as /netbsd so that tools like
                    793: savecore(8) can work.   (This is helpful but not necessary.)
                    794: 
                    795: The kernel must be specifically for Xen and for use as a domU.  The
                    796: i386 and amd64 provide the following kernels:
                    797: 
                    798:         i386 XEN3_DOMU
                    799:         i386 XEN3PAE_DOMU
1.95      gdt       800:         amd64 XEN3_DOMU
1.5       mspo      801: 
1.49      gdt       802: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
                    803: use the PAE version of the i386 kernel.
                    804: 
                    805: This will boot NetBSD, but this is not that useful if the disk is
                    806: empty.  One approach is to unpack sets onto the disk outside of xen
                    807: (by mounting it, just as you would prepare a physical disk for a
                    808: system you can't run the installer on).
                    809: 
                    810: A second approach is to run an INSTALL kernel, which has a miniroot
                    811: and can load sets from the network.  To do this, copy the INSTALL
                    812: kernel to / and change the kernel line in the config file to:
1.5       mspo      813: 
1.49      gdt       814:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
1.5       mspo      815: 
1.49      gdt       816: Then, start the domain as "xl create -c configname".
1.1       mspo      817: 
1.49      gdt       818: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
                    819: line should be used in the config file.
1.1       mspo      820: 
1.3       mspo      821:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
1.1       mspo      822: 
                    823: After booting the domain, the option to install via CDROM may be
1.49      gdt       824: selected.  The CDROM device should be changed to `xbd1d`.
1.1       mspo      825: 
1.49      gdt       826: Once done installing, "halt -p" the new domain (don't reboot or halt,
                    827: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
                    828: config file), switch the config file back to the XEN3_DOMU kernel,
                    829: and start the new domain again. Now it should be able to use "root on
                    830: xbd0a" and you should have a, functional NetBSD domU.
1.1       mspo      831: 
1.49      gdt       832: TODO: check if this is still accurate.
1.1       mspo      833: When the new domain is booting you'll see some warnings about *wscons*
                    834: and the pseudo-terminals. These can be fixed by editing the files
1.5       mspo      835: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
                    836: `/etc/ttys`, except *console*, like this:
1.1       mspo      837: 
1.3       mspo      838:     console "/usr/libexec/getty Pc"         vt100   on secure
                    839:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
                    840:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
                    841:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
                    842:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
1.1       mspo      843: 
1.5       mspo      844: Finally, all screens must be commented out from `/etc/wscons.conf`.
1.1       mspo      845: 
                    846: It is also desirable to add
                    847: 
1.49      gdt       848:         powerd=YES
1.1       mspo      849: 
1.5       mspo      850: in rc.conf. This way, the domain will be properly shut down if
1.53      gdt       851: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
1.1       mspo      852: 
1.92      gdt       853: It is not strictly necessary to have a kernel (as /netbsd) in the domU
                    854: filesystem.  However, various programs (e.g. netstat) will use that
                    855: kernel to look up symbols to read from kernel virtual memory.  If
                    856: /netbsd is not the running kernel, those lookups will fail.  (This is
                    857: not really a Xen-specific issue, but because the domU kernel is
                    858: obtained from the dom0, it is far more likely to be out of sync or
                    859: missing with Xen.)
                    860: 
1.14      gdt       861: Creating an unprivileged Linux domain (domU)
1.5       mspo      862: --------------------------------------------
1.1       mspo      863: 
                    864: Creating unprivileged Linux domains isn't much different from
                    865: unprivileged NetBSD domains, but there are some details to know.
                    866: 
                    867: First, the second parameter passed to the disk declaration (the '0x1' in
                    868: the example below)
                    869: 
1.3       mspo      870:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
1.1       mspo      871: 
                    872: does matter to Linux. It wants a Linux device number here (e.g. 0x300
1.49      gdt       873: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
                    874: So, hda1 which has major 3 and minor 1 on a Linux system will have
                    875: device number 0x301.  Alternatively, devices names can be used (hda,
                    876: hdb, ...)  as xentools has a table to map these names to devices
                    877: numbers.  To export a partition to a Linux guest we can use:
1.1       mspo      878: 
1.49      gdt       879:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
                    880:         root = "/dev/hda1 ro"
1.1       mspo      881: 
                    882: and it will appear as /dev/hda on the Linux system, and be used as root
                    883: partition.
                    884: 
1.49      gdt       885: To install the Linux system on the partition to be exported to the
                    886: guest domain, the following method can be used: install
                    887: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
                    888: that will be the root partition of your Linux domain, and mount it.
                    889: Then copy the files from a working Linux system, make adjustments in
                    890: `/etc` (fstab, network config).  It should also be possible to extract
                    891: binary packages such as .rpm or .deb directly to the mounted partition
                    892: using the appropriate tool, possibly running under NetBSD's Linux
                    893: emulation.  Once the filesystem has been populated, umount it.  If
                    894: desirable, the filesystem can be converted to ext3 using tune2fs -j.
                    895: It should now be possible to boot the Linux guest domain, using one of
                    896: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
1.1       mspo      897: 
                    898: To get the linux console right, you need to add:
                    899: 
1.3       mspo      900:     extra = "xencons=tty1"
1.1       mspo      901: 
                    902: to your configuration since not all linux distributions auto-attach a
                    903: tty to the xen console.
                    904: 
1.14      gdt       905: Creating an unprivileged Solaris domain (domU)
1.5       mspo      906: ----------------------------------------------
1.1       mspo      907: 
1.50      gdt       908: See possibly outdated
                    909: [Solaris domU instructions](/ports/xen/howto-solaris/).
1.5       mspo      910: 
1.1       mspo      911: 
1.52      gdt       912: PCI passthrough: Using PCI devices in guest domains
                    913: ---------------------------------------------------
1.1       mspo      914: 
1.53      gdt       915: The dom0 can give other domains access to selected PCI
1.52      gdt       916: devices. This can allow, for example, a non-privileged domain to have
                    917: access to a physical network interface or disk controller.  However,
                    918: keep in mind that giving a domain access to a PCI device most likely
                    919: will give the domain read/write access to the whole physical memory,
                    920: as PCs don't have an IOMMU to restrict memory access to DMA-capable
1.53      gdt       921: device.  Also, it's not possible to export ISA devices to non-dom0
1.52      gdt       922: domains, which means that the primary VGA adapter can't be exported.
                    923: A guest domain trying to access the VGA registers will panic.
                    924: 
1.53      gdt       925: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
1.52      gdt       926: not been ported to later versions at this time.
                    927: 
                    928: For a PCI device to be exported to a domU, is has to be attached to
                    929: the "pciback" driver in dom0.  Devices passed to the dom0 via the
                    930: pciback.hide boot parameter will attach to "pciback" instead of the
                    931: usual driver.  The list of devices is specified as "(bus:dev.func)",
1.5       mspo      932: where bus and dev are 2-digit hexadecimal numbers, and func a
                    933: single-digit number:
1.1       mspo      934: 
1.52      gdt       935:         pciback.hide=(00:0a.0)(00:06.0)
1.1       mspo      936: 
1.52      gdt       937: pciback devices should show up in the dom0's boot messages, and the
1.5       mspo      938: devices should be listed in the `/kern/xen/pci` directory.
1.1       mspo      939: 
1.52      gdt       940: PCI devices to be exported to a domU are listed in the "pci" array of
                    941: the domU's config file, with the format "0000:bus:dev.func".
1.1       mspo      942: 
1.52      gdt       943:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
1.1       mspo      944: 
1.52      gdt       945: In the domU an "xpci" device will show up, to which one or more pci
                    946: busses will attach.  Then the PCI drivers will attach to PCI busses as
                    947: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
                    948: or any PCI drivers built in by default; you have to build your own
                    949: kernel to use PCI devices in a domU.  Here's a kernel config example;
                    950: note that only the "xpci" lines are unusual.
                    951: 
                    952:         include         "arch/i386/conf/XEN3_DOMU"
                    953: 
                    954:         # Add support for PCI busses to the XEN3_DOMU kernel
                    955:         xpci* at xenbus ?
                    956:         pci* at xpci ?
                    957: 
                    958:         # PCI USB controllers
                    959:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
                    960: 
                    961:         # USB bus support
                    962:         usb*    at uhci?
                    963: 
                    964:         # USB Hubs
                    965:         uhub*   at usb?
                    966:         uhub*   at uhub? port ? configuration ? interface ?
                    967: 
                    968:         # USB Mass Storage
                    969:         umass*  at uhub? port ? configuration ? interface ?
                    970:         wd*     at umass?
                    971:         # SCSI controllers
                    972:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
                    973: 
                    974:         # SCSI bus support (for both ahc and umass)
                    975:         scsibus* at scsi?
                    976: 
                    977:         # SCSI devices
                    978:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
                    979:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
1.1       mspo      980: 
                    981: 
1.28      gdt       982: NetBSD as a domU in a VPS
                    983: =========================
                    984: 
                    985: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
                    986: hardware.  This section explains how to deal with Xen in a domU as a
                    987: virtual private server where you do not control or have access to the
1.70      gdt       988: dom0.  This is not intended to be an exhaustive list of VPS providers;
                    989: only a few are mentioned that specifically support NetBSD.
1.28      gdt       990: 
1.52      gdt       991: VPS operators provide varying degrees of access and mechanisms for
                    992: configuration.  The big issue is usually how one controls which kernel
                    993: is booted, because the kernel is nominally in the dom0 filesystem (to
1.70      gdt       994: which VPS users do not normally have acesss).  A second issue is how
                    995: to install NetBSD.
1.52      gdt       996: A VPS user may want to compile a kernel for security updates, to run
                    997: npf, run IPsec, or any other reason why someone would want to change
                    998: their kernel.
                    999: 
                   1000: One approach is to have an adminstrative interface to upload a kernel,
1.68      gdt      1001: or to select from a prepopulated list.  Other approaches are pygrub
1.59      gdt      1002: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
                   1003: kernel from the domU filesystem.  This is closer to a regular physical
                   1004: computer, where someone who controls a machine can replace the kernel.
1.52      gdt      1005: 
1.74      gdt      1006: A second issue is multiple CPUs.  With NetBSD 6, domUs support
                   1007: multiple vcpus, and it is typical for VPS providers to enable multiple
                   1008: CPUs for NetBSD domUs.
                   1009: 
1.68      gdt      1010: pygrub
1.59      gdt      1011: -------
1.52      gdt      1012: 
1.68      gdt      1013: pygrub runs in the dom0 and looks into the domU filesystem.  This
1.59      gdt      1014: implies that the domU must have a kernel in a filesystem in a format
1.68      gdt      1015: known to pygrub.  As of 2014, pygrub seems to be of mostly historical
                   1016: interest.
1.52      gdt      1017: 
1.59      gdt      1018: pvgrub
                   1019: ------
                   1020: 
                   1021: pvgrub is a version of grub that uses PV operations instead of BIOS
                   1022: calls.  It is booted from the dom0 as the domU kernel, and then reads
                   1023: /grub/menu.lst and loads a kernel from the domU filesystem.
                   1024: 
1.70      gdt      1025: [Panix](http://www.panix.com/) lets users use pvgrub.  Panix reports
1.71      gdt      1026: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
                   1027: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
1.70      gdt      1028: page](http://www.panix.com/v-colo/grub.html), which describes only
1.74      gdt      1029: Linux but should be updated to cover NetBSD :-).
1.70      gdt      1030: 
                   1031: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
                   1032: their own kernel.  See then [prgmr.com NetBSD
1.74      gdt      1033: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
                   1034: (which is in need of updating).
1.59      gdt      1035: 
1.70      gdt      1036: It appears that [grub's FFS
                   1037: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
                   1038: does not support all aspects of modern FFS, but there are also reports
1.72      gdt      1039: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
1.70      gdt      1040: partition for the kernel with the intent that grub can understand it,
                   1041: which leads to /netbsd not being the actual kernel.  One must remember
                   1042: to update the special boot partiion.
1.59      gdt      1043: 
                   1044: Amazon
                   1045: ------
                   1046: 
1.95      gdt      1047: See the [Amazon EC2 page](../amazon_ec2/).
1.44      gdt      1048: 
                   1049: Using npf
                   1050: ---------
                   1051: 
1.81      gdt      1052: In standard kernels, npf is a module, and thus cannot be loaded in a
1.44      gdt      1053: DOMU kernel.
                   1054: 
1.95      gdt      1055: TODO: Explain how to compile npf into a custom kernel, answering (but
                   1056: note that the problem was caused by not booting the right kernel)
                   1057: [this email to
                   1058: netbsd-users](http://mail-index.netbsd.org/netbsd-users/2014/12/26/msg015576.html).
1.65      gdt      1059: 
                   1060: TODO items for improving NetBSD/xen
                   1061: ===================================
                   1062: 
1.93      gdt      1063: * Make the NetBSD dom0 kernel work with SMP.
                   1064: * Test the Xen 4.5 packages adequately to be able to recommend them as
                   1065:   the standard approach.
                   1066: * Get PCI passthrough working on Xen 4.5
1.65      gdt      1067: * Get pvgrub into pkgsrc, either via xentools or separately.
                   1068: * grub
1.70      gdt      1069:   * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
1.66      gdt      1070:     fragsize/blocksize (UFS2 support may be present; the point is to
                   1071:     make it so that with any UFS1/UFS2 filesystem setup that works
                   1072:     with NetBSD grub will also work).
1.70      gdt      1073:     See [pkg/40258](http://gnats.netbsd.org/40258).
1.65      gdt      1074:   * Push patches upstream.
                   1075:   * Get UFS2 patches into pvgrub.
                   1076: * Add support for PV ops to a version of /boot, and make it usable as
                   1077:   a kernel in Xen, similar to pvgrub.
1.93      gdt      1078: * Solve somehow the issue with modules for GENERIC not being loadable
                   1079:   in a Xen dom0 or domU kernel.
                   1080: 
                   1081: Random pointers
                   1082: ===============
                   1083: 
                   1084: TODO: This section contains links from elsewhere not yet integrated
                   1085: into the HOWTO.
                   1086: 
                   1087: * http://www.lumbercartel.ca/library/xen/
                   1088: * http://pbraun.nethence.com/doc/sysutils/xen_netbsd_dom0.html

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