Annotation of wikisrc/ports/xen/howto.mdwn, revision 1.119

1.5       mspo        1: Introduction
1.13      gdt         2: ============
1.1       mspo        3: 
                      4: [![[Xen
1.57      gdt         5: screenshot]](](
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.
                     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.
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).
1.116     gdt        33: For HVM guests, hardware support is needed, but it is common on recent
                     34: machines.  For Intel CPUs, one needs the VT-x extension, shown in
                     35: "cpuctl identify 0" as VMX.  For AMD CPus, one needs the AMD-V
                     36: extensions, shown in "cpuctl identify 0" as SVM.  There are further
                     37: features for IOMMU virtualization, Intel's VT-d and AMD's AMD-Vi.
                     38: TODO: Explain whether Xen on NetBSD makes use of these features.
                     39: TODO: Review by someone who really understands this.
1.19      gdt        40: 
1.118     gdt        41: Note a FreeBSD dom0 requires VT-x and VT-d (or equivalent); this is
                     42: because the FreeBSD dom0 does not run in PV mode.
1.27      jnemeth    44: At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
1.12      gdt        45: The dom0 can start one or more domUs.  (Booting is explained in detail
                     46: in the dom0 section.)
                     48: NetBSD supports Xen in that it can serve as dom0, be used as a domU,
                     49: and that Xen kernels and tools are available in pkgsrc.  This HOWTO
                     50: attempts to address both the case of running a NetBSD dom0 on hardware
1.24      gdt        51: and running domUs under it (NetBSD and other), and also running NetBSD
                     52: as a domU in a VPS.
1.12      gdt        53: 
1.20      gdt        54: Some versions of Xen support "PCI passthrough", which means that
                     55: specific PCI devices can be made available to a specific domU instead
                     56: of the dom0.  This can be useful to let a domU run X11, or access some
                     57: network interface or other peripheral.
1.111     wiz        59: NetBSD 6 and earlier supported Xen 2; support was removed from NetBSD
1.104     gdt        60: 7.  Xen 2 has been removed from pkgsrc.
1.54      gdt        61: 
1.12      gdt        62: Prerequisites
1.13      gdt        63: -------------
1.12      gdt        64: 
                     65: Installing NetBSD/Xen is not extremely difficult, but it is more
                     66: complex than a normal installation of NetBSD.
1.15      gdt        67: In general, this HOWTO is occasionally overly restrictive about how
                     68: things must be done, guiding the reader to stay on the established
                     69: path when there are no known good reasons to stray.
1.12      gdt        70: 
                     71: This HOWTO presumes a basic familiarity with the Xen system
1.16      gdt        72: architecture.  This HOWTO presumes familiarity with installing NetBSD
                     73: on i386/amd64 hardware and installing software from pkgsrc.
1.27      jnemeth    74: See also the [Xen website](
1.1       mspo       75: 
1.15      gdt        76: Versions of Xen and NetBSD
                     77: ==========================
1.27      jnemeth    79: Most of the installation concepts and instructions are independent
                     80: of Xen version and NetBSD version.  This section gives advice on
                     81: which version to choose.  Versions not in pkgsrc and older unsupported
                     82: versions of NetBSD are intentionally ignored.
1.15      gdt        83: 
                     84: Xen
                     85: ---
1.111     wiz        87: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
1.15      gdt        88: xenkernel and xentools.  We will refer only to the kernel versions,
                     89: but note that both packages must be installed together and must have
                     90: matching versions.
1.104     gdt        92: xenkernel3 provides Xen 3.1.  It is no longer maintained by Xen, and the last applied security patch was in
                     93: 2011. Thus, it should not be used.  It supports PCI passthrough,
                     94: which is why people use it anyway. Xen 3.1 supports i386, both PAE and
                     95: non-PAE.
                     97: xenkernel33 provides Xen 3.3.  It is no longer maintained by Xen, and
                     98: the last applied security patch was in 2012.  Thus, it should not be
                     99: used.  Xen 3.3 supports i386, but only in PAE mode.  There are no good
                    100: reasons to run this version.
                    102: xenkernel41 provides Xen 4.1.  It is no longer maintained by Xen, but
1.112     gdt       103: as of 2016-12 received backported security patches.  Xen 4.1 supports
1.104     gdt       104: i386, but only in PAE mode.  There are no good reasons to run this
                    105: version.
1.102     gdt       106: 
1.104     gdt       107: xenkernel42 provides Xen 4.2.  It is no longer maintained by Xen, but
1.112     gdt       108: as of 2016-12 received backported security patches.  Xen 4.2 supports
1.104     gdt       109: i386, but only in PAE mode.  The only reason to run this is if you
                    110: need to use xm instead of xl, or if you need to run an i386 dom0
                    111: (because your hardware is i386 only).
1.114     gdt       113: xenkernel45 provides Xen 4.5.  As of 2016-12, security patches were
1.119   ! gdt       114: released by Xen and applied to pkgsrc.  Xen 4.5 requires using a dom0
        !           115: running NetBSD's amd64 port (Intel or AMD hardware is fine), but domUs
        !           116: can be amd64 or i386 PAE.  TODO: It is either a conservative choice or
        !           117: somewhat old.
1.104     gdt       118: 
1.115     gdt       119: xenkernel46 provides Xen 4.6.  It is new to pkgsrc as of 2016-05.  As
1.114     gdt       120: of 2016-12, security patches were released by Xen and applied to
1.119   ! gdt       121: pkgsrc.  Xen 4.6 similarly requires a NetBSD/amd64 dom0, but domUs can
        !           122: be amd64 or i386 PAE.  TODO: It is either a somewhat aggressive choice
        !           123: or the standard choice.
1.85      gdt       124: 
1.114     gdt       125: Xen 4.7 (released 2016-06) and 4.8 (released 2016-12) are not yet in
                    126: pkgsrc.
1.113     gdt       127: 
1.96      gdt       128: See also the [Xen Security Advisory page](
1.85      gdt       130: Note that NetBSD support is called XEN3.  It works with Xen 3 and Xen
                    131: 4 because the hypercall interface has been stable.
1.20      gdt       132: 
1.19      gdt       133: Xen command program
                    134: -------------------
1.79      gdt       136: Early Xen used a program called xm to manipulate the system from the
1.19      gdt       137: dom0.  Starting in 4.1, a replacement program with similar behavior
1.79      gdt       138: called xl is provided, but it does not work well in 4.1.  In 4.2, both
                    139: xm and xl work fine.  4.4 is the last version that has xm.  You must
                    140: choose one or the other, because it affects which daemons you run.
1.106     gdt       141: However, the rc.d scripts provided by xentools packages expect a
                    142: particular version, and you should use the version used by the
                    143: scripts.
1.19      gdt       144: 
1.15      gdt       145: NetBSD
                    146: ------
1.105     gdt       148: The netbsd-6, netbsd-7, and -current branches are all reasonable
                    149: choices, with more or less the same considerations for non-Xen use.
                    150: Therefore, netbsd-7 is recommended as the stable version of the most
1.117     gdt       151: recent release for production use.  In addition, netbsd-7 and -current
                    152: have a important scheduler fix (in November of 2015) affecting
                    153: contention between dom0 and domUs; see
                    154: for a
                    155: description.  For those wanting to learn Xen or without production
                    156: stability concerns, netbsd-7 is still likely most appropriate, but
                    157: -current is also a reasonable choice.  (Xen runs ok on netbsd-5, but
                    158: the xentools packages are likely difficult to build, and netbsd-5 is
                    159: not supported.)
1.15      gdt       160: 
                    161: As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
                    162: no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
1.105     gdt       163: need SMP for dom0 functions; the lack of support is really a problem
                    164: when using a dom0 as a normal computer.)
1.15      gdt       165: 
1.18      gdt       166: Architecture
                    167: ------------
1.108     gdt       169: Xen itself can run on i386 (Xen < 3.1) or amd64 machines (all Xen
1.105     gdt       170: versions).  (Practically, almost any computer where one would want to
                    171: run Xen today supports amd64.)
1.99      gdt       172: 
                    173: Xen, the dom0 kernel, and each domU kernel can be either i386 or
                    174: amd64.  When building a xenkernel package, one obtains i386 on an i386
1.111     wiz       175: host, and amd64 on an amd64 host.  If the Xen kernel is i386, then the
                    176: dom0 kernel and all domU kernels must be i386.  With an amd64 Xen
1.105     gdt       177: kernel, an amd64 dom0 kernel is known to work, and an i386PAE dom0
1.111     wiz       178: kernel should in theory work.  An amd64 Xen/dom0 is known to support
1.105     gdt       179: both i386PAE and amd64 domUs.
1.99      gdt       180: 
1.101     gdt       181: i386 dom0 and domU kernels must be PAE (except for Xen 3.1); these are
                    182: built by default.  (Note that emacs (at least) fails if run on i386
                    183: with PAE when built without, and vice versa, presumably due to bugs in
                    184: the undump code.)
1.99      gdt       185: 
                    186: Because of the above, the standard approach is to use amd64 for the
                    187: dom0.
1.29      gdt       188: 
                    189: Xen 4.2 is the last version to support i386 as a host.  TODO: Clarify
1.111     wiz       190: if this is about the CPU, the Xen kernel, or the dom0 kernel having to
1.100     gdt       191: be amd64.
1.29      gdt       193: 
1.89      gdt       194: Stability
                    195: ---------
                    197: Mostly, NetBSD as a dom0 or domU is quite stable.
                    198: However, there are some open PRs indicating problems.
1.91      gdt       200:  - [PR 48125](
                    201:  - [PR 47720](
1.89      gdt       202: 
                    203: Note also that there are issues with sparse vnd(4) instances, but
1.105     gdt       204: these are not about Xen -- they just are noticed with sparse vnd(4)
                    205: instances in support of virtual disks in a dom0.
1.89      gdt       206: 
1.15      gdt       207: Recommendation
                    208: --------------
1.105     gdt       210: Therefore, this HOWTO recommends running xenkernel45 or xenkernel46,
                    211: xl, the NetBSD 7 stable branch, and to use an amd64 kernel as the
                    212: dom0.  Either the i386PAE or amd64 version of NetBSD may be used as
                    213: domUs.
1.15      gdt       214: 
1.36      gdt       215: Build problems
                    216: --------------
                    218: Ideally, all versions of Xen in pkgsrc would build on all versions of
                    219: NetBSD on both i386 and amd64.  However, that isn't the case.  Besides
                    220: aging code and aging compilers, qemu (included in xentools for HVM
1.62      gdt       221: support) is difficult to build.  The following are known to work or FAIL:
1.36      gdt       222: 
1.62      gdt       223:         xenkernel3 netbsd-5 amd64
1.63      gdt       224:         xentools3 netbsd-5 amd64
1.64      gdt       225:         xentools3=hvm netbsd-5 amd64 ????
1.62      gdt       226:         xenkernel33 netbsd-5 amd64
1.63      gdt       227:         xentools33 netbsd-5 amd64
1.36      gdt       228:         xenkernel41 netbsd-5 amd64
                    229:         xentools41 netbsd-5 amd64
1.62      gdt       230:         xenkernel42 netbsd-5 amd64
1.64      gdt       231:         xentools42 netbsd-5 amd64
1.62      gdt       232: 
                    233:         xenkernel3 netbsd-6 i386 FAIL
                    234:         xentools3 netbsd-6 i386
1.63      gdt       235:         xentools3-hvm netbsd-6 i386 FAIL (dependencies fail)
                    236:         xenkernel33 netbsd-6 i386
                    237:         xentools33 netbsd-6 i386
1.36      gdt       238:         xenkernel41 netbsd-6 i386
                    239:         xentools41 netbsd-6 i386
1.63      gdt       240:         xenkernel42 netbsd-6 i386
1.64      gdt       241:         xentools42 netbsd-6 i386 *MIXED
1.69      gdt       243:        (all 3 and 33 seem to FAIL)
                    244:         xenkernel41 netbsd-7 i386
                    245:         xentools41 netbsd-7 i386
                    246:         xenkernel42 netbsd-7 i386
                    247:         xentools42 netbsd-7 i386 ??FAIL
1.64      gdt       249: (*On netbsd-6 i386, there is a xentools42 in the 2014Q3 official builds,
                    250: but it does not build for gdt.)
1.36      gdt       251: 
1.15      gdt       252: NetBSD as a dom0
                    253: ================
                    255: NetBSD can be used as a dom0 and works very well.  The following
                    256: sections address installation, updating NetBSD, and updating Xen.
1.19      gdt       257: Note that it doesn't make sense to talk about installing a dom0 OS
                    258: without also installing Xen itself.  We first address installing
                    259: NetBSD, which is not yet a dom0, and then adding Xen, pivoting the
                    260: NetBSD install to a dom0 install by just changing the kernel and boot
                    261: configuration.
1.15      gdt       262: 
1.45      gdt       263: For experimenting with Xen, a machine with as little as 1G of RAM and
                    264: 100G of disk can work.  For running many domUs in productions, far
                    265: more will be needed.
1.15      gdt       267: Styles of dom0 operation
                    268: ------------------------
                    270: There are two basic ways to use Xen.  The traditional method is for
                    271: the dom0 to do absolutely nothing other than providing support to some
                    272: number of domUs.  Such a system was probably installed for the sole
                    273: purpose of hosting domUs, and sits in a server room on a UPS.
                    275: The other way is to put Xen under a normal-usage computer, so that the
                    276: dom0 is what the computer would have been without Xen, perhaps a
                    277: desktop or laptop.  Then, one can run domUs at will.  Purists will
                    278: deride this as less secure than the previous approach, and for a
                    279: computer whose purpose is to run domUs, they are right.  But Xen and a
1.93      gdt       280: dom0 (without domUs) is not meaningfully less secure than the same
1.15      gdt       281: things running without Xen.  One can boot Xen or boot regular NetBSD
                    282: alternately with little problems, simply refraining from starting the
                    283: Xen daemons when not running Xen.
                    285: Note that NetBSD as dom0 does not support multiple CPUs.  This will
1.51      gdt       286: limit the performance of the Xen/dom0 workstation approach.  In theory
                    287: the only issue is that the "backend drivers" are not yet MPSAFE:
1.15      gdt       289: 
1.19      gdt       290: Installation of NetBSD
                    291: ----------------------
1.13      gdt       292: 
1.19      gdt       293: First,
1.27      jnemeth   294: [install NetBSD/amd64](/guide/inst/)
1.19      gdt       295: just as you would if you were not using Xen.
                    296: However, the partitioning approach is very important.
                    298: If you want to use RAIDframe for the dom0, there are no special issues
                    299: for Xen.  Typically one provides RAID storage for the dom0, and the
1.22      gdt       300: domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
1.111     wiz       301: over a RAID1 header to find /boot from a file system within a RAID
1.22      gdt       302: partition; this is no different when booting Xen.
1.19      gdt       303: 
                    304: There are 4 styles of providing backing storage for the virtual disks
1.93      gdt       305: used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
1.19      gdt       306: 
                    307: With raw partitions, one has a disklabel (or gpt) partition sized for
                    308: each virtual disk to be used by the domU.  (If you are able to predict
                    309: how domU usage will evolve, please add an explanation to the HOWTO.
                    310: Seriously, needs tend to change over time.)
1.27      jnemeth   312: One can use [lvm(8)](/guide/lvm/) to create logical devices to use
                    313: for domU disks.  This is almost as efficient as raw disk partitions
                    314: and more flexible.  Hence raw disk partitions should typically not
                    315: be used.
1.19      gdt       316: 
1.111     wiz       317: One can use files in the dom0 file system, typically created by dd'ing
1.19      gdt       318: /dev/zero to create a specific size.  This is somewhat less efficient,
                    319: but very convenient, as one can cp the files for backup, or move them
                    320: between dom0 hosts.
                    322: Finally, in theory one can place the files backing the domU disks in a
                    323: SAN.  (This is an invitation for someone who has done this to add a
                    324: HOWTO page.)
1.1       mspo      325: 
1.19      gdt       326: Installation of Xen
                    327: -------------------
1.1       mspo      328: 
1.20      gdt       329: In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
                    330: pkgsrc (or another matching pair).
                    331: See [the pkgsrc
                    332: documentation]( for help with pkgsrc.
                    334: For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm.  More
                    335: recent versions have HVM support integrated in the main xentools
                    336: package.  It is entirely reasonable to run only PV guests.
                    338: Next you need to install the selected Xen kernel itself, which is
                    339: installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
                    340: For debugging, one may copy xen-debug.gz; this is conceptually similar
                    341: to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
                    342: useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
                    343: in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
1.75      gdt       344: of a NetBSD build.  If using i386, use
                    345: releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.  (If using Xen
                    346: 3.1 and i386, you may use XEN3_DOM0 with the non-PAE Xen.  But you
                    347: should not use Xen 3.1.)  Both xen and the NetBSD kernel may be (and
                    348: typically are) left compressed.
1.111     wiz       350: In a dom0 kernel, kernfs is mandatory for xend to communicate with the
1.75      gdt       351: kernel, so ensure that /kern is in fstab.  TODO: Say this is default,
                    352: or file a PR and give a reference.
1.20      gdt       353: 
                    354: Because you already installed NetBSD, you have a working boot setup
                    355: with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
1.111     wiz       356: beginning of your root file system, /boot present, and likely
1.20      gdt       357: /boot.cfg.  (If not, fix before continuing!)
1.76      gdt       359: Add a line to to /boot.cfg to boot Xen.  See boot.cfg(5) for an
                    360: example.  The basic line is
1.20      gdt       361: 
1.37      gdt       362:         menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
1.20      gdt       363: 
                    364: which specifies that the dom0 should have 256M, leaving the rest to be
1.77      gdt       365: allocated for domUs.  To use a serial console, use
                    367:         menu=Xen:load /netbsd-XEN3_DOM0.gz console=com0;multiboot /xen.gz dom0_mem=256M console=com1 com1=9600,8n1
                    369: which will use the first serial port for Xen (which counts starting
                    370: from 1), forcing speed/parity, and also for NetBSD (which counts
                    371: starting at 0).  In an attempt to add performance, one can also add
1.37      gdt       372: 
                    373:         dom0_max_vcpus=1 dom0_vcpus_pin
                    375: to force only one vcpu to be provided (since NetBSD dom0 can't use
1.111     wiz       376: more) and to pin that vcpu to a physical CPU.  TODO: benchmark this.
1.20      gdt       377: 
1.93      gdt       378: Xen has [many boot
                    379: options](,
1.111     wiz       380: and other than dom0 memory and max_vcpus, they are generally not
1.93      gdt       381: necessary.
1.20      gdt       383: As with non-Xen systems, you should have a line to boot /netbsd (a
                    384: kernel that works without Xen) and fallback versions of the non-Xen
                    385: kernel, Xen, and the dom0 kernel.
1.1       mspo      386: 
1.76      gdt       387: Now, reboot so that you are running a DOM0 kernel under Xen, rather
                    388: than GENERIC without Xen.
1.54      gdt       390: Using grub (historic)
                    391: ---------------------
                    393: Before NetBSD's native bootloader could support Xen, the use of
                    394: grub was recommended.  If necessary, see the
                    395: [old grub information](/ports/xen/howto-grub/).
1.28      gdt       397: The [HowTo on Installing into
                    398: RAID-1](
                    399: explains how to set up booting a dom0 with Xen using grub with
                    400: NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
                    401: boot.)
1.21      gdt       403: Configuring Xen
                    404: ---------------
1.53      gdt       406: Xen logs will be in /var/log/xen.
1.76      gdt       408: Now, you have a system that will boot Xen and the dom0 kernel, but not
                    409: do anything else special.  Make sure that you have rebooted into Xen.
                    410: There will be no domUs, and none can be started because you still have
1.102     gdt       411: to configure the dom0 daemons.
1.21      gdt       412: 
1.102     gdt       413: The daemons which should be run vary with Xen version and with whether
                    414: one is using xm or xl.  The Xen 3.1 and 3.3 packages use xm.  Xen 4.1
                    415: and higher packages use xl.  While is is possible to use xm with some
                    416: 4.x versions (TODO: 4.1 and 4.2?), the pkgsrc-provided rc.d scripts do
                    417: not support this as of 2014-12-26, and thus the HOWTO does not support
                    418: it either.  (Make sure your packages are reasonably recent.)
1.43      gdt       419: 
1.102     gdt       420: For "xm" (3.1 and 3.3), you should enable xend and xenbackendd (but
                    421: note that you should be using 4.x):
1.31      gdt       422: 
1.32      gdt       423:         xend=YES
                    424:         xenbackendd=YES
1.31      gdt       425: 
1.102     gdt       426: For "xl" (4.x), you should enabled xend and xencommons (xenstored).
                    427: Trying to boot 4.x without xencommons=YES will result in a hang; it is
1.111     wiz       428: necessary to hit ^C on the console to let the machine finish booting.
1.102     gdt       429: TODO: explain why xend is installed by the package.
1.31      gdt       430: 
1.53      gdt       431:         xencommons=YES
1.31      gdt       432: 
1.102     gdt       433: The installation of NetBSD should already have created devices for xen
                    434: (xencons, xenevt), but if they are not present, create them:
1.31      gdt       435: 
1.102     gdt       436:         cd /dev && sh MAKEDEV xen
1.86      gdt       437: 
1.31      gdt       438: TODO: Recommend for/against xen-watchdog.
1.27      jnemeth   439: 
1.53      gdt       440: After you have configured the daemons and either started them (in the
1.79      gdt       441: order given) or rebooted, use xm or xl to inspect Xen's boot messages,
1.102     gdt       442: available resources, and running domains.  An example with xl follows:
1.34      gdt       443: 
1.102     gdt       444:         # xl dmesg
1.43      gdt       445:        [xen's boot info]
1.102     gdt       446:         # xl info
1.43      gdt       447:        [available memory, etc.]
1.102     gdt       448:         # xl list
1.43      gdt       449:         Name              Id  Mem(MB)  CPU  State  Time(s)  Console
                    450:         Domain-0           0       64    0  r----     58.1
1.33      gdt       451: 
1.88      gdt       452: ### Issues with xencommons
                    454: xencommons starts xenstored, which stores data on behalf of dom0 and
                    455: domUs.  It does not currently work to stop and start xenstored.
                    456: Certainly all domUs should be shutdown first, following the sort order
                    457: of the rc.d scripts.  However, the dom0 sets up state with xenstored,
                    458: and is not notified when xenstored exits, leading to not recreating
                    459: the state when the new xenstored starts.  Until there's a mechanism to
                    460: make this work, one should not expect to be able to restart xenstored
                    461: (and thus xencommons).  There is currently no reason to expect that
                    462: this will get fixed any time soon.
1.41      gdt       464: anita (for testing NetBSD)
                    465: --------------------------
1.82      gdt       467: With the setup so far (assuming 4.2/xl), one should be able to run
                    468: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
                    469: root, because anita must create a domU):
                    471:         anita --vmm=xl test file:///usr/obj/i386/
                    473: Alternatively, one can use --vmm=xm to use xm-based domU creation
                    474: instead (and must, on Xen <= 4.1).   TODO: confirm that anita xl really works.
1.40      gdt       476: Xen-specific NetBSD issues
                    477: --------------------------
                    479: There are (at least) two additional things different about NetBSD as a
                    480: dom0 kernel compared to hardware.
1.111     wiz       482: One is that the module ABI is different because some of the #defines
1.109     gdt       483: change, so one must build modules for Xen.  As of netbsd-7, the build
                    484: system does this automatically.  TODO: check this.  (Before building
                    485: Xen modules was added, it was awkward to use modules to the point
                    486: where it was considered that it did not work.)
1.40      gdt       487: 
                    488: The other difference is that XEN3_DOM0 does not have exactly the same
                    489: options as GENERIC.  While it is debatable whether or not this is a
                    490: bug, users should be aware of this and can simply add missing config
                    491: items if desired.
1.15      gdt       493: Updating NetBSD in a dom0
                    494: -------------------------
                    496: This is just like updating NetBSD on bare hardware, assuming the new
                    497: version supports the version of Xen you are running.  Generally, one
                    498: replaces the kernel and reboots, and then overlays userland binaries
                    499: and adjusts /etc.
                    501: Note that one must update both the non-Xen kernel typically used for
                    502: rescue purposes and the DOM0 kernel used with Xen.
1.55      gdt       504: Converting from grub to /boot
                    505: -----------------------------
                    507: These instructions were [TODO: will be] used to convert a system from
                    508: grub to /boot.  The system was originally installed in February of
                    509: 2006 with a RAID1 setup and grub to boot Xen 2, and has been updated
                    510: over time.  Before these commands, it was running NetBSD 6 i386, Xen
                    511: 4.1 and grub, much like the message linked earlier in the grub
                    512: section.
1.111     wiz       514:         # Install MBR bootblocks on both disks. 
1.55      gdt       515:         fdisk -i /dev/rwd0d
                    516:         fdisk -i /dev/rwd1d
                    517:         # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
                    518:         installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
                    519:         installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
                    520:         # Install secondary boot loader
                    521:         cp -p /usr/mdec/boot /
1.111     wiz       522:         # Create boot.cfg following earlier guidance:
1.55      gdt       523:         menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
                    524:         menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=256M
                    525:         menu=GENERIC:boot
                    526:         menu=GENERIC single-user:boot -s
                    527:         menu=GENERIC.ok:boot netbsd.ok
                    528:         menu=GENERIC.ok single-user:boot netbsd.ok -s
                    529:         menu=Drop to boot prompt:prompt
                    530:         default=1
                    531:         timeout=30
                    533: TODO: actually do this and fix it if necessary.
1.22      gdt       534: 
1.102     gdt       535: Upgrading Xen versions
1.15      gdt       536: ---------------------
1.110     gdt       538: Minor version upgrades are trivial.  Just rebuild/replace the
                    539: xenkernel version and copy the new xen.gz to / (where /boot.cfg
                    540: references it), and reboot.
                    542: Major version upgrades are conceptually not difficult, but can run
                    543: into all the issues found when installing Xen.  Assuming migration
                    544: from 4.1 to 4.2, remove the xenkernel41 and xentools41 packages and
                    545: install the xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz
                    546: to /.
1.21      gdt       547: 
1.102     gdt       548: Ensure that the contents of /etc/rc.d/xen* are correct.  Specifically,
                    549: they must match the package you just installed and not be left over
                    550: from some previous installation.
                    552: Enable the correct set of daemons; see the configuring section above.
                    553: (Upgrading from 3.x to 4.x without doing this will result in a hang.)
                    555: Ensure that the domU config files are valid for the new version.
1.110     gdt       556: Specifically, for 4.x remove autorestart=True, and ensure that disks
                    557: are specified with numbers as the second argument, as the examples
                    558: above show, and not NetBSD device names.
1.15      gdt       559: 
1.97      gdt       560: Hardware known to work
                    561: ----------------------
                    563: Arguably, this section is misplaced, and there should be a page of
                    564: hardware that runs NetBSD/amd64 well, with the mostly-well-founded
                    565: assumption that NetBSD/xen runs fine on any modern hardware that
                    566: NetBSD/amd64 runs well on.  Until then, we give motherboard/CPU/RAM
                    567: triples to aid those choosing a motherboard.  Note that Xen systems
                    568: usually do not run X, so a listing here does not imply that X works at
                    569: all.
                    571:         Supermicro X9SRL-F, Xeon E5-1650 v2, 96 GiB ECC
                    572:         Supermicro ??, Atom C2758 (8 core), 32 GiB ECC
                    573:         ASUS M5A78L-M/USB3 AM3+ microATX, AMD Piledriver X8 4000MHz, 16 GiB ECC
                    575: Older hardware:
1.98      gdt       577:         Intel D915GEV, Pentium4 CPU 3.40GHz, 4GB 533MHz Synchronous DDR2
1.28      gdt       578: 
1.82      gdt       579: Running Xen under qemu
                    580: ----------------------
                    582: The astute reader will note that this section is somewhat twisted.
                    583: However, it can be useful to run Xen under qemu either because the
                    584: version of NetBSD as a dom0 does not run on the hardware in use, or to
                    585: generate automated test cases involving Xen.
1.84      gdt       587: In 2015-01, the following combination was reported to mostly work:
1.82      gdt       588: 
                    589:         host OS: NetBSD/amd64 6.1.4
                    590:         qemu: 2.2.0 from pkgsrc
                    591:         Xen kernel: xenkernel42-4.2.5nb1 from pkgsrc
                    592:         dom0 kernel: NetBSD/amd64 6.1.5
                    593:         Xen tools: xentools42-4.2.5 from pkgsrc
1.91      gdt       595: See [PR 47720]( for a problem with dom0
                    596: shutdown.
1.84      gdt       597: 
1.28      gdt       598: Unprivileged domains (domU)
                    599: ===========================
                    601: This section describes general concepts about domUs.  It does not
1.33      gdt       602: address specific domU operating systems or how to install them.  The
                    603: config files for domUs are typically in /usr/pkg/etc/xen, and are
1.60      wiki      604: typically named so that the file name, domU name and the domU's host
1.33      gdt       605: name match.
1.111     wiz       607: The domU is provided with CPU and memory by Xen, configured by the
1.33      gdt       608: dom0.  The domU is provided with disk and network by the dom0,
                    609: mediated by Xen, and configured in the dom0.
                    611: Entropy in domUs can be an issue; physical disks and network are on
                    612: the dom0.  NetBSD's /dev/random system works, but is often challenged.
1.48      gdt       614: Config files
                    615: ------------
                    617: There is no good order to present config files and the concepts
                    618: surrounding what is being configured.  We first show an example config
                    619: file, and then in the various sections give details.
                    621: See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
                    622: for a large number of well-commented examples, mostly for running
                    623: GNU/Linux.
                    625: The following is an example minimal domain configuration file
                    626: "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
                    627: known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
                    628: i386 domU).  The domU serves as a network file server.
                    630:         # -*- mode: python; -*-
                    632:         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
                    633:         memory = 1024
                    634:         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
                    635:         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
                    636:                  'file:/n0/xen/foo-wd1,0x1,w' ]
                    638: The domain will have the same name as the file.  The kernel has the
                    639: host/domU name in it, so that on the dom0 one can update the various
                    640: domUs independently.  The vif line causes an interface to be provided,
                    641: with a specific mac address (do not reuse MAC addresses!), in bridge
                    642: mode.  Two disks are provided, and they are both writable; the bits
                    643: are stored in files and Xen attaches them to a vnd(4) device in the
1.111     wiz       644: dom0 on domain creation.  The system treats xbd0 as the boot device
1.48      gdt       645: without needing explicit configuration.
                    647: By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
                    648: that "xm create" takes the name of a config file, while other commands
                    649: take the name of a domain.  To create the domain, connect to the
                    650: console, create the domain while attaching the console, shutdown the
                    651: domain, and see if it has finished stopping, do (or xl with Xen >=
                    652: 4.2):
                    654:         xm create foo
                    655:         xm console foo
                    656:         xm create -c foo
                    657:         xm shutdown foo
1.90      gdt       658:         xm list
1.48      gdt       659: 
                    660: Typing ^] will exit the console session.  Shutting down a domain is
                    661: equivalent to pushing the power button; a NetBSD domU will receive a
                    662: power-press event and do a clean shutdown.  Shutting down the dom0
                    663: will trigger controlled shutdowns of all configured domUs.
                    665: domU kernels
                    666: ------------
                    668: On a physical computer, the BIOS reads sector 0, and a chain of boot
                    669: loaders finds and loads a kernel.  Normally this comes from the root
1.111     wiz       670: file system.  With Xen domUs, the process is totally different.  The
1.48      gdt       671: normal path is for the domU kernel to be a file in the dom0's
1.111     wiz       672: file system.  At the request of the dom0, Xen loads that kernel into a
1.48      gdt       673: new domU instance and starts execution.  While domU kernels can be
                    674: anyplace, reasonable places to store domU kernels on the dom0 are in /
                    675: (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
                    676: config files), or in /u0/xen (where the vdisks are).
1.59      gdt       678: Note that loading the domU kernel from the dom0 implies that boot
                    679: blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
1.48      gdt       680: See the VPS section near the end for discussion of alternate ways to
                    681: obtain domU kernels.
1.33      gdt       683: CPU and memory
                    684: --------------
1.48      gdt       686: A domain is provided with some number of vcpus, less than the number
1.111     wiz       687: of CPUs seen by the hypervisor.  (For a dom0, this is controlled by
1.48      gdt       688: the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
                    689: from the config file by the "vcpus = N" directive.
                    691: A domain is provided with memory; this is controlled in the config
                    692: file by "memory = N" (in megabytes).  In the straightforward case, the
                    693: sum of the the memory allocated to the dom0 and all domUs must be less
1.33      gdt       694: than the available memory.
                    696: Xen also provides a "balloon" driver, which can be used to let domains
                    697: use more memory temporarily.  TODO: Explain better, and explain how
                    698: well it works with NetBSD.
1.28      gdt       699: 
                    700: Virtual disks
                    701: -------------
1.33      gdt       703: With the file/vnd style, typically one creates a directory,
                    704: e.g. /u0/xen, on a disk large enough to hold virtual disks for all
                    705: domUs.  Then, for each domU disk, one writes zeros to a file that then
                    706: serves to hold the virtual disk's bits; a suggested name is foo-xbd0
                    707: for the first virtual disk for the domU called foo.  Writing zeros to
                    708: the file serves two purposes.  One is that preallocating the contents
                    709: improves performance.  The other is that vnd on sparse files has
                    710: failed to work.  TODO: give working/notworking NetBSD versions for
                    711: sparse vnd.  Note that the use of file/vnd for Xen is not really
                    712: different than creating a file-backed virtual disk for some other
1.39      gdt       713: purpose, except that xentools handles the vnconfig commands.  To
                    714: create an empty 4G virtual disk, simply do
                    716:         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
1.33      gdt       717: 
1.89      gdt       718: Do not use qemu-img-xen, because this will create sparse file.  There
                    719: have been recent (2015) reports of sparse vnd(4) devices causing
                    720: lockups, but there is apparently no PR.
1.33      gdt       722: With the lvm style, one creates logical devices.  They are then used
1.48      gdt       723: similarly to vnds.  TODO: Add an example with lvm.
                    725: In domU config files, the disks are defined as a sequence of 3-tuples.
                    726: The first element is "method:/path/to/disk".  Common methods are
                    727: "file:" for file-backed vnd. and "phy:" for something that is already
                    728: a (TODO: character or block) device.
                    730: The second element is an artifact of how virtual disks are passed to
                    731: Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
                    732: are given a device name to associate with the disk, and values like
                    733: "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
                    734: as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
                    735: second argument.  The name given is converted to a major/minor by
1.49      gdt       736: calling stat(2) on the name in /dev and this is passed to the domU.
                    737: In the general case, the dom0 and domU can be different operating
1.48      gdt       738: systems, and it is an unwarranted assumption that they have consistent
                    739: numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
                    740: as both dom0 and domU, using values of 0x0 for the first disk and 0x1
1.49      gdt       741: for the second works fine and avoids this issue.  For a GNU/Linux
                    742: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
                    743: /dev/hda1.
1.48      gdt       744: 
                    745: The third element is "w" for writable disks, and "r" for read-only
                    746: disks.
1.28      gdt       747: 
                    748: Virtual Networking
                    749: ------------------
1.111     wiz       751: Xen provides virtual Ethernets, each of which connects the dom0 and a
1.46      gdt       752: domU.  For each virtual network, there is an interface "xvifN.M" in
                    753: the dom0, and in domU index N, a matching interface xennetM (NetBSD
                    754: name).  The interfaces behave as if there is an Ethernet with two
1.111     wiz       755: adapters connected.  From this primitive, one can construct various
1.46      gdt       756: configurations.  We focus on two common and useful cases for which
                    757: there are existing scripts: bridging and NAT.
1.28      gdt       758: 
1.48      gdt       759: With bridging (in the example above), the domU perceives itself to be
                    760: on the same network as the dom0.  For server virtualization, this is
                    761: usually best.  Bridging is accomplished by creating a bridge(4) device
                    762: and adding the dom0's physical interface and the various xvifN.0
                    763: interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
                    764: config file.  The bridge must be set up already in the dom0; an
                    765: example /etc/ifconfig.bridge0 is:
1.46      gdt       766: 
                    767:         create
                    768:         up
                    769:         !brconfig bridge0 add wm0
1.28      gdt       770: 
                    771: With NAT, the domU perceives itself to be behind a NAT running on the
                    772: dom0.  This is often appropriate when running Xen on a workstation.
1.48      gdt       773: TODO: NAT appears to be configured by "vif = [ '' ]".
1.28      gdt       774: 
1.49      gdt       775: The MAC address specified is the one used for the interface in the new
1.53      gdt       776: domain.  The interface in dom0 will use this address XOR'd with
1.49      gdt       777: 00:00:00:01:00:00.  Random MAC addresses are assigned if not given.
1.33      gdt       779: Sizing domains
                    780: --------------
                    782: Modern x86 hardware has vast amounts of resources.  However, many
                    783: virtual servers can function just fine on far less.  A system with
                    784: 256M of RAM and a 4G disk can be a reasonable choice.  Note that it is
                    785: far easier to adjust virtual resources than physical ones.  For
                    786: memory, it's just a config file edit and a reboot.  For disk, one can
                    787: create a new file and vnconfig it (or lvm), and then dump/restore,
                    788: just like updating physical disks, but without having to be there and
                    789: without those pesky connectors.
1.48      gdt       791: Starting domains automatically
                    792: ------------------------------
1.28      gdt       793: 
1.48      gdt       794: To start domains foo at bar at boot and shut them down cleanly on dom0
                    795: shutdown, in rc.conf add:
1.28      gdt       796: 
1.48      gdt       797:         xendomains="foo bar"
1.28      gdt       798: 
1.111     wiz       799: Note that earlier versions of the xentools41 xendomains rc.d script
                    800: used xl, when one should use xm with 4.1.
1.28      gdt       801: 
                    802: Creating specific unprivileged domains (domU)
                    803: =============================================
1.14      gdt       804: 
                    805: Creating domUs is almost entirely independent of operating system.  We
1.49      gdt       806: have already presented the basics of config files.  Note that you must
                    807: have already completed the dom0 setup so that "xl list" (or "xm list")
                    808: works.
1.14      gdt       809: 
                    810: Creating an unprivileged NetBSD domain (domU)
                    811: ---------------------------------------------
1.1       mspo      812: 
1.49      gdt       813: See the earlier config file, and adjust memory.  Decide on how much
                    814: storage you will provide, and prepare it (file or lvm).
1.111     wiz       816: While the kernel will be obtained from the dom0 file system, the same
1.49      gdt       817: file should be present in the domU as /netbsd so that tools like
                    818: savecore(8) can work.   (This is helpful but not necessary.)
                    820: The kernel must be specifically for Xen and for use as a domU.  The
                    821: i386 and amd64 provide the following kernels:
                    823:         i386 XEN3_DOMU
                    824:         i386 XEN3PAE_DOMU
1.95      gdt       825:         amd64 XEN3_DOMU
1.5       mspo      826: 
1.49      gdt       827: Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
                    828: use the PAE version of the i386 kernel.
                    830: This will boot NetBSD, but this is not that useful if the disk is
                    831: empty.  One approach is to unpack sets onto the disk outside of xen
                    832: (by mounting it, just as you would prepare a physical disk for a
                    833: system you can't run the installer on).
                    835: A second approach is to run an INSTALL kernel, which has a miniroot
                    836: and can load sets from the network.  To do this, copy the INSTALL
                    837: kernel to / and change the kernel line in the config file to:
1.5       mspo      838: 
1.49      gdt       839:         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
1.5       mspo      840: 
1.49      gdt       841: Then, start the domain as "xl create -c configname".
1.1       mspo      842: 
1.49      gdt       843: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
                    844: line should be used in the config file.
1.1       mspo      845: 
1.3       mspo      846:     disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
1.1       mspo      847: 
                    848: After booting the domain, the option to install via CDROM may be
1.49      gdt       849: selected.  The CDROM device should be changed to `xbd1d`.
1.1       mspo      850: 
1.49      gdt       851: Once done installing, "halt -p" the new domain (don't reboot or halt,
                    852: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
                    853: config file), switch the config file back to the XEN3_DOMU kernel,
                    854: and start the new domain again. Now it should be able to use "root on
                    855: xbd0a" and you should have a, functional NetBSD domU.
1.1       mspo      856: 
1.49      gdt       857: TODO: check if this is still accurate.
1.1       mspo      858: When the new domain is booting you'll see some warnings about *wscons*
                    859: and the pseudo-terminals. These can be fixed by editing the files
1.5       mspo      860: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
                    861: `/etc/ttys`, except *console*, like this:
1.1       mspo      862: 
1.3       mspo      863:     console "/usr/libexec/getty Pc"         vt100   on secure
                    864:     ttyE0   "/usr/libexec/getty Pc"         vt220   off secure
                    865:     ttyE1   "/usr/libexec/getty Pc"         vt220   off secure
                    866:     ttyE2   "/usr/libexec/getty Pc"         vt220   off secure
                    867:     ttyE3   "/usr/libexec/getty Pc"         vt220   off secure
1.1       mspo      868: 
1.5       mspo      869: Finally, all screens must be commented out from `/etc/wscons.conf`.
1.1       mspo      870: 
                    871: It is also desirable to add
1.49      gdt       873:         powerd=YES
1.1       mspo      874: 
1.5       mspo      875: in rc.conf. This way, the domain will be properly shut down if
1.53      gdt       876: `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
1.1       mspo      877: 
1.92      gdt       878: It is not strictly necessary to have a kernel (as /netbsd) in the domU
1.111     wiz       879: file system.  However, various programs (e.g. netstat) will use that
1.92      gdt       880: kernel to look up symbols to read from kernel virtual memory.  If
                    881: /netbsd is not the running kernel, those lookups will fail.  (This is
                    882: not really a Xen-specific issue, but because the domU kernel is
                    883: obtained from the dom0, it is far more likely to be out of sync or
                    884: missing with Xen.)
1.14      gdt       886: Creating an unprivileged Linux domain (domU)
1.5       mspo      887: --------------------------------------------
1.1       mspo      888: 
                    889: Creating unprivileged Linux domains isn't much different from
                    890: unprivileged NetBSD domains, but there are some details to know.
                    892: First, the second parameter passed to the disk declaration (the '0x1' in
                    893: the example below)
1.3       mspo      895:     disk = [ 'phy:/dev/wd0e,0x1,w' ]
1.1       mspo      896: 
                    897: does matter to Linux. It wants a Linux device number here (e.g. 0x300
1.49      gdt       898: for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
                    899: So, hda1 which has major 3 and minor 1 on a Linux system will have
                    900: device number 0x301.  Alternatively, devices names can be used (hda,
                    901: hdb, ...)  as xentools has a table to map these names to devices
                    902: numbers.  To export a partition to a Linux guest we can use:
1.1       mspo      903: 
1.49      gdt       904:         disk = [ 'phy:/dev/wd0e,0x300,w' ]
                    905:         root = "/dev/hda1 ro"
1.1       mspo      906: 
                    907: and it will appear as /dev/hda on the Linux system, and be used as root
                    908: partition.
1.49      gdt       910: To install the Linux system on the partition to be exported to the
                    911: guest domain, the following method can be used: install
                    912: sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
                    913: that will be the root partition of your Linux domain, and mount it.
                    914: Then copy the files from a working Linux system, make adjustments in
                    915: `/etc` (fstab, network config).  It should also be possible to extract
                    916: binary packages such as .rpm or .deb directly to the mounted partition
                    917: using the appropriate tool, possibly running under NetBSD's Linux
1.111     wiz       918: emulation.  Once the file system has been populated, umount it.  If
                    919: desirable, the file system can be converted to ext3 using tune2fs -j.
1.49      gdt       920: It should now be possible to boot the Linux guest domain, using one of
                    921: the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
1.1       mspo      922: 
1.111     wiz       923: To get the Linux console right, you need to add:
1.1       mspo      924: 
1.3       mspo      925:     extra = "xencons=tty1"
1.1       mspo      926: 
1.111     wiz       927: to your configuration since not all Linux distributions auto-attach a
1.1       mspo      928: tty to the xen console.
1.14      gdt       930: Creating an unprivileged Solaris domain (domU)
1.5       mspo      931: ----------------------------------------------
1.1       mspo      932: 
1.50      gdt       933: See possibly outdated
                    934: [Solaris domU instructions](/ports/xen/howto-solaris/).
1.5       mspo      935: 
1.1       mspo      936: 
1.52      gdt       937: PCI passthrough: Using PCI devices in guest domains
                    938: ---------------------------------------------------
1.1       mspo      939: 
1.53      gdt       940: The dom0 can give other domains access to selected PCI
1.52      gdt       941: devices. This can allow, for example, a non-privileged domain to have
                    942: access to a physical network interface or disk controller.  However,
                    943: keep in mind that giving a domain access to a PCI device most likely
                    944: will give the domain read/write access to the whole physical memory,
                    945: as PCs don't have an IOMMU to restrict memory access to DMA-capable
1.53      gdt       946: device.  Also, it's not possible to export ISA devices to non-dom0
1.52      gdt       947: domains, which means that the primary VGA adapter can't be exported.
                    948: A guest domain trying to access the VGA registers will panic.
1.53      gdt       950: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
1.52      gdt       951: not been ported to later versions at this time.
                    953: For a PCI device to be exported to a domU, is has to be attached to
                    954: the "pciback" driver in dom0.  Devices passed to the dom0 via the
                    955: pciback.hide boot parameter will attach to "pciback" instead of the
                    956: usual driver.  The list of devices is specified as "(bus:dev.func)",
1.5       mspo      957: where bus and dev are 2-digit hexadecimal numbers, and func a
                    958: single-digit number:
1.1       mspo      959: 
1.52      gdt       960:         pciback.hide=(00:0a.0)(00:06.0)
1.1       mspo      961: 
1.52      gdt       962: pciback devices should show up in the dom0's boot messages, and the
1.5       mspo      963: devices should be listed in the `/kern/xen/pci` directory.
1.1       mspo      964: 
1.52      gdt       965: PCI devices to be exported to a domU are listed in the "pci" array of
                    966: the domU's config file, with the format "0000:bus:dev.func".
1.1       mspo      967: 
1.52      gdt       968:         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
1.1       mspo      969: 
1.52      gdt       970: In the domU an "xpci" device will show up, to which one or more pci
1.111     wiz       971: buses will attach.  Then the PCI drivers will attach to PCI buses as
1.52      gdt       972: usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
                    973: or any PCI drivers built in by default; you have to build your own
                    974: kernel to use PCI devices in a domU.  Here's a kernel config example;
                    975: note that only the "xpci" lines are unusual.
                    977:         include         "arch/i386/conf/XEN3_DOMU"
1.111     wiz       979:         # Add support for PCI buses to the XEN3_DOMU kernel
1.52      gdt       980:         xpci* at xenbus ?
                    981:         pci* at xpci ?
                    983:         # PCI USB controllers
                    984:         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
                    986:         # USB bus support
                    987:         usb*    at uhci?
                    989:         # USB Hubs
                    990:         uhub*   at usb?
                    991:         uhub*   at uhub? port ? configuration ? interface ?
                    993:         # USB Mass Storage
                    994:         umass*  at uhub? port ? configuration ? interface ?
                    995:         wd*     at umass?
                    996:         # SCSI controllers
                    997:         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
                    999:         # SCSI bus support (for both ahc and umass)
                   1000:         scsibus* at scsi?
                   1002:         # SCSI devices
                   1003:         sd*     at scsibus? target ? lun ?      # SCSI disk drives
                   1004:         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
1.1       mspo     1005: 
1.28      gdt      1007: NetBSD as a domU in a VPS
                   1008: =========================
                   1010: The bulk of the HOWTO is about using NetBSD as a dom0 on your own
                   1011: hardware.  This section explains how to deal with Xen in a domU as a
                   1012: virtual private server where you do not control or have access to the
1.70      gdt      1013: dom0.  This is not intended to be an exhaustive list of VPS providers;
                   1014: only a few are mentioned that specifically support NetBSD.
1.28      gdt      1015: 
1.52      gdt      1016: VPS operators provide varying degrees of access and mechanisms for
                   1017: configuration.  The big issue is usually how one controls which kernel
1.111     wiz      1018: is booted, because the kernel is nominally in the dom0 file system (to
                   1019: which VPS users do not normally have access).  A second issue is how
1.70      gdt      1020: to install NetBSD.
1.52      gdt      1021: A VPS user may want to compile a kernel for security updates, to run
                   1022: npf, run IPsec, or any other reason why someone would want to change
                   1023: their kernel.
1.111     wiz      1025: One approach is to have an administrative interface to upload a kernel,
1.68      gdt      1026: or to select from a prepopulated list.  Other approaches are pygrub
1.59      gdt      1027: (deprecated) and pvgrub, which are ways to have a bootloader obtain a
1.111     wiz      1028: kernel from the domU file system.  This is closer to a regular physical
1.59      gdt      1029: computer, where someone who controls a machine can replace the kernel.
1.52      gdt      1030: 
1.74      gdt      1031: A second issue is multiple CPUs.  With NetBSD 6, domUs support
                   1032: multiple vcpus, and it is typical for VPS providers to enable multiple
                   1033: CPUs for NetBSD domUs.
1.68      gdt      1035: pygrub
1.59      gdt      1036: -------
1.52      gdt      1037: 
1.111     wiz      1038: pygrub runs in the dom0 and looks into the domU file system.  This
                   1039: implies that the domU must have a kernel in a file system in a format
1.68      gdt      1040: known to pygrub.  As of 2014, pygrub seems to be of mostly historical
                   1041: interest.
1.52      gdt      1042: 
1.59      gdt      1043: pvgrub
                   1044: ------
                   1046: pvgrub is a version of grub that uses PV operations instead of BIOS
                   1047: calls.  It is booted from the dom0 as the domU kernel, and then reads
1.111     wiz      1048: /grub/menu.lst and loads a kernel from the domU file system.
1.59      gdt      1049: 
1.70      gdt      1050: [Panix]( lets users use pvgrub.  Panix reports
1.71      gdt      1051: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
                   1052: (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
1.70      gdt      1053: page](, which describes only
1.74      gdt      1054: Linux but should be updated to cover NetBSD :-).
1.70      gdt      1055: 
                   1056: []( also lets users with pvgrub to boot
                   1057: their own kernel.  See then [ NetBSD
1.74      gdt      1058: HOWTO](
                   1059: (which is in need of updating).
1.59      gdt      1060: 
1.70      gdt      1061: It appears that [grub's FFS
                   1062: code](
                   1063: does not support all aspects of modern FFS, but there are also reports
1.72      gdt      1064: that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
1.70      gdt      1065: partition for the kernel with the intent that grub can understand it,
                   1066: which leads to /netbsd not being the actual kernel.  One must remember
1.111     wiz      1067: to update the special boot partition.
1.59      gdt      1068: 
                   1069: Amazon
                   1070: ------
1.95      gdt      1072: See the [Amazon EC2 page](../amazon_ec2/).
1.44      gdt      1073: 
                   1074: Using npf
                   1075: ---------
1.81      gdt      1077: In standard kernels, npf is a module, and thus cannot be loaded in a
1.44      gdt      1078: DOMU kernel.
1.95      gdt      1080: TODO: Explain how to compile npf into a custom kernel, answering (but
                   1081: note that the problem was caused by not booting the right kernel)
                   1082: [this email to
                   1083: netbsd-users](
1.65      gdt      1084: 
                   1085: TODO items for improving NetBSD/xen
                   1086: ===================================
1.93      gdt      1088: * Make the NetBSD dom0 kernel work with SMP.
                   1089: * Test the Xen 4.5 packages adequately to be able to recommend them as
                   1090:   the standard approach.
                   1091: * Get PCI passthrough working on Xen 4.5
1.65      gdt      1092: * Get pvgrub into pkgsrc, either via xentools or separately.
                   1093: * grub
1.70      gdt      1094:   * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
1.66      gdt      1095:     fragsize/blocksize (UFS2 support may be present; the point is to
1.111     wiz      1096:     make it so that with any UFS1/UFS2 file system setup that works
1.66      gdt      1097:     with NetBSD grub will also work).
1.70      gdt      1098:     See [pkg/40258](
1.65      gdt      1099:   * Push patches upstream.
                   1100:   * Get UFS2 patches into pvgrub.
                   1101: * Add support for PV ops to a version of /boot, and make it usable as
                   1102:   a kernel in Xen, similar to pvgrub.
1.93      gdt      1103: * Solve somehow the issue with modules for GENERIC not being loadable
                   1104:   in a Xen dom0 or domU kernel.
                   1106: Random pointers
                   1107: ===============
1.117     gdt      1109: This section contains links from elsewhere not yet integrated into the
                   1110: HOWTO, and other guides.
1.93      gdt      1111: 
                   1112: *
                   1113: *
1.117     gdt      1114: *

CVSweb for NetBSD wikisrc <> software: FreeBSD-CVSweb