wikisrc/ports/xen/howto.mdwn - diff

Return to howto.mdwn CVS log

Up to [NetBSD Developer Wiki] / wikisrc / ports / xen

Diff for /wikisrc/ports/xen/howto.mdwn between versions 1.19 and 1.141

-version 1.19, 2014/12/24 00:41:04
+version 1.141, 2017/12/15 16:40:59
  Line 2  Introduction
  ============
  [![[Xen
- screenshot]](http://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](../../gallery/in-Action/hubertf-xen.png)
+ screenshot]](https://www.netbsd.org/gallery/in-Action/hubertf-xens.png)](https://www.netbsd.org/gallery/in-Action/hubertf-xen.png)
- Xen is a virtual machine monitor or hypervisor for x86 hardware
+ Xen is a hypervisor (or virtual machine monitor) for x86 hardware
  (i686-class or higher), which supports running multiple guest
- operating systems on a single physical machine.  With Xen, one uses
+ operating systems on a single physical machine.  Xen is a Type 1 or
- the Xen kernel to control the CPU, memory and console, a dom0
+ bare-metal hypervisor; one uses the Xen kernel to control the CPU,
- operating system which mediates access to other hardware (e.g., disks,
+ memory and console, a dom0 operating system which mediates access to
- network, USB), and one or more domU operating systems which operate in
+ other hardware (e.g., disks, network, USB), and one or more domU
- an unprivileged virtualized environment.  IO requests from the domU
+ operating systems which operate in an unprivileged virtualized
- systems are forwarded by the hypervisor (Xen) to the dom0 to be
+ environment.  IO requests from the domU systems are forwarded by the
- fulfilled.
+ hypervisor (Xen) to the dom0 to be fulfilled.
  Xen supports two styles of guests.  The original is Para-Virtualized
  (PV) which means that the guest OS does not attempt to access hardware
  Line 27  code for Xen and need not be aware that
  Attempts to access hardware registers are trapped and emulated.  This
  style is less efficient but can run unmodified guests.
- Generally any amd64 machine will work with Xen and PV guests.  For HVM
+ Generally any machine that runs NetBSD/amd64 will work with Xen and PV
- guests, the VMX cpu feature (Intel) or VT?? (amd64) is needed.  TODO:
+ guests.  In theory i386 computers (without x86_64/amd64 support) can
- Explain if i386 (non-amd64) machines can still be used - I think that
+ be used for Xen <= 4.2, but we have no recent reports of this working
- witthe requirement to use PAE kernels is about the hypervisor being
+ (this is a hint).  For HVM guests, hardware support is needed, but it
- amd64 only.
+ is common on recent machines.  For Intel CPUs, one needs the VT-x
+ extension, shown in "cpuctl identify 0" as VMX.  For AMD CPUs, one
+ needs the AMD-V extensions, shown in "cpuctl identify 0" as SVM.
+ There are further features for IOMMU virtualization, Intel's VT-d and
+ AMD's AMD-Vi.  TODO: Explain whether Xen on NetBSD makes use of these
+ features.  TODO: Review by someone who really understands this.
- At boot, the dom0 kernel is loaded as module with Xen as the kernel.
+ Note that a FreeBSD dom0 requires VT-x and VT-d (or equivalent); this
+ is because the FreeBSD dom0 does not run in PV mode.
+ At boot, the dom0 kernel is loaded as a module with Xen as the kernel.
  The dom0 can start one or more domUs.  (Booting is explained in detail
  in the dom0 section.)
  NetBSD supports Xen in that it can serve as dom0, be used as a domU,
  and that Xen kernels and tools are available in pkgsrc.  This HOWTO
  attempts to address both the case of running a NetBSD dom0 on hardware
- and running NetBSD as a domU in a VPS.
+ and running domUs under it (NetBSD and other), and also running NetBSD
+ as a domU in a VPS.
+ Xen 3.1 in pkgsrc supports "PCI passthrough", which means that
+ specific PCI devices can be made available to a specific domU instead
+ of the dom0.  This can be useful to let a domU run X11, or access some
+ network interface or other peripheral.
+ NetBSD 6 and earlier supported Xen 2; support was removed from NetBSD
+.  Xen 2 has been removed from pkgsrc.
  Prerequisites
  -------------
- Line 52  things must be done, guiding the reader
+ Line 69  things must be done, guiding the reader
  path when there are no known good reasons to stray.
  This HOWTO presumes a basic familiarity with the Xen system
- architecture.  This HOWTO presumes familiarity with installing NetBSD
+ architecture, with installing NetBSD on i386/amd64 hardware, and with
- on i386/amd64 hardware and installing software from pkgsrc.
+ installing software from pkgsrc.  See also the [Xen
- See also the [Xen website](http://www.xen.org/).
+ website](http://www.xenproject.org/).
- History
- -------
- NetBSD used to support Xen2; this has been removed.
- Before NetBSD's native bootloader could support Xen, the use of
- grub was recommended.  If necessary, see the
- [old grub information](/xen/howto-grub/).
  Versions of Xen and NetBSD
  ==========================
- Most of the installation concepts and instructions are independent of
+ Most of the installation concepts and instructions are independent
- Xen version.  This section gives advice on which version to choose.
+ of Xen version and NetBSD version.  This section gives advice on
- Versions not in pkgsrc and older unsupported versions of NetBSD are
+ which version to choose.  Versions not in pkgsrc and older unsupported
- inentionally ignored.
+ versions of NetBSD are intentionally ignored.
+ The term "amd64" is used to refer to both the NetBSD port and to the
+ hardware architecture on which it runs.  (Such hardware is made by
+ both Intel and AMD, and in 2016 a normal PC has this CPU
+ architecture.)
  Xen
  ---
- In NetBSD, xen is provided in pkgsrc, via matching pairs of packages
+ In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages
  xenkernel and xentools.  We will refer only to the kernel versions,
  but note that both packages must be installed together and must have
  matching versions.
- xenkernel3 and xenkernel33 provide Xen 3.1 and 3.3.  These no longer
+ xenkernel3 provides Xen 3.1.  It is no longer maintained by Xen, and
- receive security patches and should not be used.
+ the last applied security patch was in 2011. Thus, it should not be
+ used.  It supports PCI passthrough, which is why people use it anyway.
+ Xen 3.1 runs on i386 (both non-PAE and PAE) and amd64 hardware.
+ xenkernel33 provides Xen 3.3.  It is no longer maintained by Xen, and
+ the last applied security patch was in 2012.  Thus, it should not be
+ used.  Xen 3.3 runs on i386 PAE and amd64 hardware.  There are no good
+ reasons to run this version.
+ xenkernel41 provides Xen 4.1.  It is no longer maintained by Xen, but
+ as of 2016-12 received backported security patches.  Xen 4.1 runs on
+ i386 PAE and amd64 hardware.  There are no good reasons to run this
+ version.
+ Note that 3.1, 3.3 and 4.1 have been removed from pkgsrc-current, but
+ are in 2016Q4.  They will be removed from this HOWTO sometime after
+Q1.
+ xenkernel42 provides Xen 4.2.  It is no longer maintained by Xen, but
+ as of 2016-12 received backported security patches.  Xen 4.2 runs on
+ i386 PAE and amd64 hardware.  The only reason to run this is if you
+ need to use xm instead of xl, or if you need to run on hardware that
+ supports i386 but not amd64.  (This might also be useful if you need
+ an i386 dom0, if it turns out that an amd64 Xen kernel and an i386
+ dom0 is problematic.)
+ xenkernel45 provides Xen 4.5.  As of 2016-12, security patches were
+ released by Xen and applied to pkgsrc.  Xen 4.5 runs on amd64 hardware
+ only.  While slightly old, 4.5 has been tested and run by others, so
+ it is the conservative choice.
+ xenkernel46 provides Xen 4.6.  It is new to pkgsrc as of 2016-05.  As
+ of 2016-12, security patches were released by Xen and applied to
+ pkgsrc.  Xen 4.6 runs on amd64 hardware only For new installations,
+.6 is probably the appropriate choice and it will likely soon be the
+ standard approach.  (If using Ubuntu guests, be sure to have the
+ xentools46 from December, 2016).
- xenkernel41 provides Xen 4.1.  This is no longer maintained by Xen,
+ Xen 4.7 (released 2016-06) and 4.8 (released 2016-12) are not yet in
- but as of 2014-12 receives backported security patches.  It is a
+ pkgsrc.
- reasonable although trailing-edge choice.
- xenkernel42 provides Xen 4.2.  This is maintained by Xen, but old as
+ See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/).
- of 2014-12.
- Ideally newer versions of Xen will be added to pkgsrc.
+ Note that NetBSD support is called XEN3.  It works with Xen 3 and Xen
+because the hypercall interface has been stable.
  Xen command program
  -------------------
- Early Xen used a program called "xm" to manipulate the system from the
+ Early Xen used a program called xm to manipulate the system from the
  dom0.  Starting in 4.1, a replacement program with similar behavior
- called "xl" is provided.  In 4.2, "xm" is no longer available.
+ called xl is provided, but it does not work well in 4.1.  In 4.2, both
+ xm and xl work fine.  4.4 is the last version that has xm.
+ You must make a global choice to use xm or xl, because it affects not
+ only which command you use, but the command used by rc.d scripts
+ (specifically xendomains) and which daemons should be run.  The
+ xentools packages provide xm for 3.1, 3.3 and 4.1 and xl for 4.2 and up.
+ In 4.2, you can choose to use xm by simply changing the ctl_command
+ variable and setting xend=YES in rc.conf.
+ With xl, virtual devices are configured in parallel, which can cause
+ problems if they are written assuming serial operation (e.g., updating
+ firewall rules without explicit locking).  There is now locking for
+ the provided scripts, which works for normal casses (e.g, file-backed
+ xbd, where a vnd must be allocated).  But, as of 201612, it has not
+ been adequately tested for a complex custom setup with a large number
+ of interfaces.
  NetBSD
  ------
- The netbsd-5, netbsd-6, netbsd-7, and -current branches are all
+ The netbsd-6, netbsd-7, and -current branches are all reasonable
- reasonable choices, with more or less the same considerations for
+ choices, with more or less the same considerations for non-Xen use.
- non-Xen use.  Therefore, netbsd-6 is recommended as the stable version
+ Therefore, netbsd-7 is recommended as the stable version of the most
- of the most recent release.
+ recent release for production use.  In addition, netbsd-7 and -current
+ have a important scheduler fix (in November of 2015) affecting
+ contention between dom0 and domUs; see
+ https://releng.netbsd.org/cgi-bin/req-7.cgi?show=1040 for a
+ description.  For those wanting to learn Xen or without production
+ stability concerns, netbsd-7 is still likely most appropriate, but
+ -current is also a reasonable choice.  (Xen runs ok on netbsd-5, but
+ the xentools packages are likely difficult to build, and netbsd-5 is
+ not supported.)
  As of NetBSD 6, a NetBSD domU will support multiple vcpus.  There is
  no SMP support for NetBSD as dom0.  (The dom0 itself doesn't really
- need SMP; the lack of support is really a problem when using a dom0 as
+ need SMP for dom0 functions; the lack of support is really a problem
- a normal computer.)
+ when using a dom0 as a normal computer.)
  Architecture
  ------------
- Xen is basically amd64 only at this point.  One can either run i386
+ Xen itself can run on i386 (Xen < 4.2) or amd64 hardware (all Xen
- domains or amd64 domains.  If running i386, PAE versions are required,
+ versions).  (Practically, almost any computer where one would want to
- for both dom0 and domU.  These versions are built by default in NetBSD
+ run Xen today supports amd64.)
- releases.  While i386 dom0 works fine, amd64 is recommended as more
- normal.  (Note that emacs (at least) fails if run on i386 with PAE when
+ Xen, the dom0 system, and each domU system can be either i386 or
- built without, and vice versa, presumably due to bugs in the undump
+ amd64.  When building a xenkernel package, one obtains an i386 Xen
- code.)
+ kernel on an i386 host, and an amd64 Xen kernel on an amd64 host.  If
+ the Xen kernel is i386, then the dom0 kernel and all domU kernels must
+ be i386.  With an amd64 Xen kernel, an amd64 dom0 kernel is known to
+ work, and an i386 dom0 kernel should in theory work.  An amd64
+ Xen/dom0 is known to support both i386 and amd64 domUs.
+ i386 dom0 and domU kernels must be PAE (except for an i386 Xen 3.1
+ kernel, where one can use non-PAE for dom0 and all domUs); PAE kernels
+ are included in the NetBSD default build.  (Note that emacs (at least)
+ fails if run on i386 with PAE when built without, and vice versa,
+ presumably due to bugs in the undump code.)
+ Because of the above, the standard approach is to use an amd64 Xen
+ kernel and NetBSD/amd64 for the dom0.  For domUs, NetBSD/i386 (with
+ the PAE kernel) and NetBSD/amd64 are in widespread use, and there is
+ little to no Xen-specific reason to prefer one over the other.
+ Note that to use an i386 dom0 with Xen 4.5 or higher, one must build
+ (or obtain from pre-built packages) an amd64 Xen kernel and install
+ that on the system.  (One must also use a PAE i386 kernel, but this is
+ also required with an i386 Xen kernel.).  Almost no one in the
+ NetBSD/Xen community does this, and the standard, well-tested,
+ approach is to use an amd64 dom0.
+ A [posting on
+ xen-devel](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html)
+ explained that PV system call overhead was higher on amd64, and thus
+ there is some notion that i386 guests are faster.  It goes on to
+ caution that the total situation is complex and not entirely
+ understood. On top of that caution, the post is about Linux, not
+ NetBSD.  TODO: Include link to benchmarks, if someone posts them.
+ Stability
+ ---------
+ Mostly, NetBSD as a dom0 or domU is quite stable.
+ However, there are some open PRs indicating problems.
+  - [PR 48125](http://gnats.netbsd.org/48125)
+ Note also that there are issues with sparse vnd(4) instances, but
+ these are not about Xen -- they just are noticed with sparse vnd(4)
+ instances in support of virtual disks in a dom0.
  Recommendation
  --------------
- Therefore, this HOWTO recommends running xenkernel42 (and xentools42),
+ Therefore, this HOWTO recommends running xenkernel45 or xenkernel46,
- xl, the NetBSD 6 stable branch, and to use amd64 as the dom0.  Either
+ xl, the NetBSD 7 stable branch, and to use an amd64 kernel as the
- the i386 or amd64 of NetBSD may be used as domUs.
+ dom0.  Either the i386PAE or amd64 version of NetBSD may be used as
+ domUs.
+ Because bugs are fixed quite often, and because of Xen security
+ advisories, it is good to stay up to date with NetBSD (tracking a
+ stable branch), with the Xen kernel (tracking a Xen version via
+ pkgsrc), and with the Xen tools.  Specifically, NetBSD (-7 and
+ -current) got an important fix affecting dom0/domU timesharing in
+ November, 2015, and xentools46 got a fix to enable Ubuntu guests to
+ boot in December, 2016.
+ Status
+ ------
+ Ideally, all versions of Xen in pkgsrc would build on all supported
+ versions of NetBSD/amd64, to the point where this section would be
+ silly.  However, that has not always been the case.  Besides aging
+ code and aging compilers, qemu (included in xentools for HVM support)
+ is difficult to build.  Note that there is intentionally no data for
+.5+ up for i386, and often omits xentools info if the corresponding
+ kernel fails.
+ The following table gives status, with the date last checked
+ (generally on the most recent quarterly branch).  The first code is
+ "builds" if it builds ok, and "FAIL" for a failure to build.  The
+ second code/date only appears for xenkernel* and is "works" if it runs
+ ok as a dom0 and can support a domU, and "FAIL" if it won't boot or
+ run a domU.
+         xenkernel3 netbsd-6 i386 FAIL 201612
+         xenkernel33 netbsd-6 i386 FAIL 201612
+         xenkernel41 netbsd-6 i386 builds 201612
+         xenkernel42 netbsd-6 i386 builds 201612
+         xentools3 netbsd-6 i386 FAIL 201612
+         xentools33 netbsd-6 i386 FAIL 201612
+         xentools41 netbsd-6 i386 builds 201612
+         xentools42 netbsd-6 i386 FAIL 201612
+         xenkernel3 netbsd-7 i386 FAIL 201412
+         xenkernel33 netbsd-7 i386 FAIL 201412
+         xenkernel41 netbsd-7 i386 builds 201412
+         xenkernel42 netbsd-7 i386 builds 201412
+         xentools41 netbsd-7 i386 builds 201412
+         xentools42 netbsd-7 i386 ??FAIL 201412
+         xenkernel3 netbsd-6 amd64 FAIL 201612
+         xenkernel33 netbsd-6 amd64 FAIL 201612
+         xenkernel41 netbsd-6 amd64 builds 201612 works 201612
+         xenkernel42 netbsd-6 amd64 builds 201612 works 201612
+         xenkernel45 netbsd-6 amd64 builds 201612
+         xenkernel46 netbsd-6 amd64 builds 201612
+         xentools41 netbsd-6 amd64 builds 201612
+         xentools42 netbsd-6 amd64 builds 201612
+         xentools45 netbsd-6 amd64 builds 201612
+         xentools46 netbsd-6 amd64 FAIL 201612
+         xenkernel3 netbsd-7 amd64 builds 201612
+         xenkernel33 netbsd-7 amd64 builds 201612
+         xenkernel41 netbsd-7 amd64 builds 201612
+         xenkernel42 netbsd-7 amd64 builds 201612
+         xenkernel45 netbsd-7 amd64 builds 201612
+         xenkernel46 netbsd-7 amd64 builds 201612
+         xentools3 netbsd-7 amd64 builds 201612
+         xentools3-hvm netbsd-7 amd64 builds 201612
+         xentools33 netbsd-7 amd64 FAIL 201612
+         xentools41 netbsd-7 amd64 builds 201612
+         xentools42 netbsd-7 amd64 builds 201612
+         xentools45 netbsd-7 amd64 builds 201612
+         xentools46 netbsd-7 amd64 builds 201612
  NetBSD as a dom0
  ================
- Line 142  NetBSD, which is not yet a dom0, and the
+ Line 325  NetBSD, which is not yet a dom0, and the
  NetBSD install to a dom0 install by just changing the kernel and boot
  configuration.
+ For experimenting with Xen, a machine with as little as 1G of RAM and
+G of disk can work.  For running many domUs in productions, far
+ more will be needed; e.g. 4-8G and 1T of disk is reasonable for a
+ half-dozen domUs of 512M and 32G each.  Basically, the RAM and disk
+ have to be bigger than the sum of the RAM/disk needs of the dom0 and
+ all the domUs.
  Styles of dom0 operation
  ------------------------
- Line 155  dom0 is what the computer would have bee
+ Line 345  dom0 is what the computer would have bee
  desktop or laptop.  Then, one can run domUs at will.  Purists will
  deride this as less secure than the previous approach, and for a
  computer whose purpose is to run domUs, they are right.  But Xen and a
- dom0 (without domUs) is not meaingfully less secure than the same
+ dom0 (without domUs) is not meaningfully less secure than the same
  things running without Xen.  One can boot Xen or boot regular NetBSD
  alternately with little problems, simply refraining from starting the
  Xen daemons when not running Xen.
  Note that NetBSD as dom0 does not support multiple CPUs.  This will
- limit the performance of the Xen/dom0 workstation approach.
+ limit the performance of the Xen/dom0 workstation approach.  In theory
+ the only issue is that the "backend drivers" are not yet MPSAFE:
+   https://mail-index.netbsd.org/netbsd-users/2014/08/29/msg015195.html
  Installation of NetBSD
  ----------------------
  First,
- [install NetBSD/amd64](../../docs/guide/en/chap-inst.html)
+ [install NetBSD/amd64](/guide/inst/)
  just as you would if you were not using Xen.
  However, the partitioning approach is very important.
  If you want to use RAIDframe for the dom0, there are no special issues
  for Xen.  Typically one provides RAID storage for the dom0, and the
- domU systems are unaware of RAID.
+ domU systems are unaware of RAID.  The 2nd-stage loader bootxx_* skips
+ over a RAID1 header to find /boot from a file system within a RAID
+ partition; this is no different when booting Xen.
  There are 4 styles of providing backing storage for the virtual disks
- used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN,
+ used by domUs: raw partitions, LVM, file-backed vnd(4), and SAN.
  With raw partitions, one has a disklabel (or gpt) partition sized for
  each virtual disk to be used by the domU.  (If you are able to predict
  how domU usage will evolve, please add an explanation to the HOWTO.
  Seriously, needs tend to change over time.)
- One can use lvm(8) to create logical devices to use for domU disks.
+ One can use [lvm(8)](/guide/lvm/) to create logical devices to use
- This is almost as efficient sa raw disk partitions and more flexible.
+ for domU disks.  This is almost as efficient as raw disk partitions
- Hence raw disk partitions should typically not be used.
+ and more flexible.  Hence raw disk partitions should typically not
+ be used.
- One can use files in the dom0 filesystem, typically created by dd'ing
+ One can use files in the dom0 file system, typically created by dd'ing
  /dev/zero to create a specific size.  This is somewhat less efficient,
  but very convenient, as one can cp the files for backup, or move them
  between dom0 hosts.
- Line 199  HOWTO page.)
+ Line 394  HOWTO page.)
  Installation of Xen
  -------------------
- Next step is to install the Xen packages via pkgsrc or from binary
+ In the dom0, install sysutils/xenkernel42 and sysutils/xentools42 from
- packages. See [the pkgsrc
+ pkgsrc (or another matching pair).  See [the pkgsrc
- documentation](http://www.NetBSD.org/docs/pkgsrc/) if you are unfamiliar
+ documentation](https://www.NetBSD.org/docs/pkgsrc/) for help with
- with pkgsrc and/or handling of binary packages. Xen 3.1, 3.3, 4.1 and
+ pkgsrc.  Ensure that your packages are recent; the HOWTO does not
-.2 are available. 3.1 supports PCI pass-through while other versions do
+ contemplate old builds.
- not. You'll need either `sysutils/xentools3` and `sysutils/xenkernel3`
- for Xen 3.1, `sysutils/xentools33` and `sysutils/xenkernel33` for Xen
-.3, `sysutils/xentools41` and `sysutils/xenkernel41` for Xen 4.1. or
+ For Xen 3.1, support for HVM guests is in sysutils/xentool3-hvm.  More
- `sysutils/xentools42` and `sysutils/xenkernel42` for Xen 4.2. You'll
+ recent versions have HVM support integrated in the main xentools
- also need `sysutils/grub` if you plan do use the grub boot loader. If
+ package.  It is entirely reasonable to run only PV guests.
- using Xen 3.1, you may also want to install `sysutils/xentools3-hvm`
- which contains the utilities to run unmodified guests OSes using the
+ Next you need to install the selected Xen kernel itself, which is
- *HVM* support (for later versions this is included in
+ installed by pkgsrc as "/usr/pkg/xen*-kernel/xen.gz".  Copy it to /.
- `sysutils/xentools`). Note that your CPU needs to support this. Intel
+ For debugging, one may copy xen-debug.gz; this is conceptually similar
- CPUs must have the 'VT' instruction, AMD CPUs the 'SVM' instruction. You
+ to DIAGNOSTIC and DEBUG in NetBSD.  xen-debug.gz is basically only
- can easily find out if your CPU support HVM by using NetBSD's cpuctl
+ useful with a serial console.  Then, place a NetBSD XEN3_DOM0 kernel
- command:
+ in /, copied from releasedir/amd64/binary/kernel/netbsd-XEN3_DOM0.gz
+ of a NetBSD build.  If using i386, use
-     # cpuctl identify 0
+ releasedir/i386/binary/kernel/netbsd-XEN3PAE_DOM0.gz.  (If using Xen
-     cpu0: Intel Core 2 (Merom) (686-class), id 0x6f6
+.1 and i386, you may use XEN3_DOM0 with the non-PAE Xen.  But you
-     cpu0: features 0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR>
+ should not use Xen 3.1.)  Both xen and the NetBSD kernel may be (and
-     cpu0: features 0xbfebfbff<PGE,MCA,CMOV,PAT,PSE36,CFLUSH,DS,ACPI,MMX>
+ typically are) left compressed.
-     cpu0: features 0xbfebfbff<FXSR,SSE,SSE2,SS,HTT,TM,SBF>
-     cpu0: features2 0x4e33d<SSE3,DTES64,MONITOR,DS-CPL,,TM2,SSSE3,CX16,xTPR,PDCM,DCA>
+ In a dom0, kernfs is mandatory for xend to communicate with the
-     cpu0: features3 0x20100800<SYSCALL/SYSRET,XD,EM64T>
+ kernel, so ensure that /kern is in fstab.  (A standard NetBSD install
-     cpu0: "Intel(R) Xeon(R) CPU            5130  @ 2.00GHz"
+ should already mount /kern.)
-     cpu0: I-cache 32KB 64B/line 8-way, D-cache 32KB 64B/line 8-way
-     cpu0: L2 cache 4MB 64B/line 16-way
+ Because you already installed NetBSD, you have a working boot setup
-     cpu0: ITLB 128 4KB entries 4-way
+ with an MBR bootblock, either bootxx_ffsv1 or bootxx_ffsv2 at the
-     cpu0: DTLB 256 4KB entries 4-way, 32 4MB entries 4-way
+ beginning of your root file system, have /boot, and likely also
-     cpu0: Initial APIC ID 0
+ /boot.cfg.  (If not, fix before continuing!)
-     cpu0: Cluster/Package ID 0
-     cpu0: Core ID 0
+ Add a line to to /boot.cfg to boot Xen.  See boot.cfg(5) for an
-     cpu0: family 06 model 0f extfamily 00 extmodel 00
+ example.  The basic line is
- Depending on your CPU, the feature you are looking for is called HVM,
+         menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
- SVM or VMX.
+ which specifies that the dom0 should have 256M, leaving the rest to be
- Next you need to copy the selected Xen kernel itself. pkgsrc installed
+ allocated for domUs.  To use a serial console, use
- them under `/usr/pkg/xen*-kernel/`. The file you're looking for is
- `xen.gz`. Copy it to your root file system. `xen-debug.gz` is a kernel
+         menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=256M console=com1 com1=9600,8n1
- with more consistency checks and more details printed on the serial
- console. It is useful for debugging crashing guests if you use a serial
+ which will use the first serial port for Xen (which counts starting
- console. It is not useful with a VGA console.
+ from 1, unlike NetBSD which counts starting from 0), forcing
+ speed/parity.  Because the NetBSD command line lacks a
- You'll then need a NetBSD/Xen kernel for *domain0* on your root file
+ "console=pc" argument, it will use the default "xencons" console device,
- system. The XEN3PAE\_DOM0 kernel or XEN3\_DOM0 provided as part of the
+ which directs the console I/O through Xen to the same console device Xen
- i386 or amd64 binaries is suitable for this, but you may want to
+ itself uses (in this case, the serial port).
- customize it. Keep your native kernel around, as it can be useful for
- recovery. *Note:* the *domain0* kernel must support KERNFS and `/kern`
+ In an attempt to add performance, one can also add
- must be mounted because *xend* needs access to `/kern/xen/privcmd`.
+         dom0_max_vcpus=1 dom0_vcpus_pin
- Next you need to get a bootloader to load the `xen.gz` kernel, and the
- NetBSD *domain0* kernel as a module. This can be `grub` or NetBSD's boot
+ to force only one vcpu to be provided (since NetBSD dom0 can't use
- loader. Below is a detailled example for grub, see the boot.cfg(5)
+ more) and to pin that vcpu to a physical CPU.  TODO: benchmark this.
- manual page for an example using the latter.
+ Xen has [many boot
- This is also where you'll specify the memory allocated to *domain0*, the
+ options](http://xenbits.xenproject.org/docs/4.5-testing/misc/xen-command-line.html),
- console to use, etc ...
+ and other than dom0 memory and max_vcpus, they are generally not
+ necessary.
- Here is a commented `/grub/menu.lst` file:
+ As with non-Xen systems, you should have a line to boot /netbsd (a
-     #Grub config file for NetBSD/xen. Copy as /grub/menu.lst and run
+ kernel that works without Xen).  Consider a line to boot /netbsd.ok (a
-     # grub-install /dev/rwd0d (assuming your boot device is wd0).
+ fallback version of the non-Xen kernel, updated manually when you are
-     #
+ sure /netbsd is ok).  Consider also a line to boot fallback versions
-     # The default entry to load will be the first one
+ of Xen and the dom0 kernel, but note that non-Xen NetBSD can be used
-     default=0
+ to resolve Xen booting issues.
-     # boot the default entry after 10s if the user didn't hit keyboard
+ Probably you want a default=N line to choose Xen in the absence of
-     timeout=10
+ intervention.
-     # Configure serial port to use as console. Ignore if you'll use VGA only
+ Now, reboot so that you are running a DOM0 kernel under Xen, rather
-     serial --unit=0 --speed=115200 --word=8 --parity=no --stop=1
+ than GENERIC without Xen.
-     # Let the user select which console to use (serial or VGA), default
+ Using grub (historic)
-     # to serial after 10s
+ ---------------------
-     terminal --timeout=10 serial console
+ Before NetBSD's native bootloader could support Xen, the use of
-     # An entry for NetBSD/xen, using /netbsd as the domain0 kernel, and serial
+ grub was recommended.  If necessary, see the
-     # console. Domain0 will have 64MB RAM allocated.
+ [old grub information](/ports/xen/howto-grub).
-     # Assume NetBSD is installed in the first MBR partition.
-     title Xen 3 / NetBSD (hda0, serial)
+ The [HowTo on Installing into
-       root(hd0,0)
+ RAID-1](https://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
-       kernel (hd0,a)/xen.gz dom0_mem=65536 com1=115200,8n1
+ explains how to set up booting a dom0 with Xen using grub with
-       module (hd0,a)/netbsd bootdev=wd0a ro console=ttyS0
+ NetBSD's RAIDframe.  (This is obsolete with the use of NetBSD's native
+ boot.  Now, just create a system with RAID-1, and alter /boot.cfg as
-     # Same as above, but using VGA console
+ described above.)
-     # We can use console=tty0 (Linux syntax) or console=pc (NetBSD syntax)
-     title Xen 3 / NetBSD (hda0, vga)
+ Configuring Xen
-       root(hd0,0)
+ ---------------
-       kernel (hd0,a)/xen.gz dom0_mem=65536
-       module (hd0,a)/netbsd bootdev=wd0a ro console=tty0
+ Xen logs will be in /var/log/xen.
-     # NetBSD/xen using a backup domain0 kernel (in case you installed a
+ Now, you have a system that will boot Xen and the dom0 kernel, but not
-     # nonworking kernel as /netbsd
+ do anything else special.  Make sure that you have rebooted into Xen.
-     title Xen 3 / NetBSD (hda0, backup, serial)
+ There will be no domUs, and none can be started because you still have
-       root(hd0,0)
+ to configure the dom0 daemons.
-       kernel (hd0,a)/xen.gz dom0_mem=65536 com1=115200,8n1
-       module (hd0,a)/netbsd.backup bootdev=wd0a ro console=ttyS0
+ The daemons which should be run vary with Xen version and with whether
-     title Xen 3 / NetBSD (hda0, backup, VGA)
+ one is using xm or xl.  The Xen 3.1, 3.3 and 4.1 packages use xm.  Xen
-       root(hd0,0)
+.2 and up packages use xl.  To use xm with 4.2, edit xendomains to
-       kernel (hd0,a)/xen.gz dom0_mem=65536
+ use xm instead.
-       module (hd0,a)/netbsd.backup bootdev=wd0a ro console=tty0
+ For 3.1 and 3.3, you should enable xend and xenbackendd:
-     #Load a regular NetBSD/i386 kernel. Can be useful if you end up with a
-     #nonworking /xen.gz
+         xend=YES
-     title NetBSD 5.1
+         xenbackendd=YES
-       root (hd0,a)
-       kernel --type=netbsd /netbsd-GENERIC
+ For 4.1 and up, you should enable xencommons.  Not enabling xencommons
+ will result in a hang; it is necessary to hit ^C on the console to let
-     #Load the NetBSD bootloader, letting it load the NetBSD/i386 kernel.
+ the machine finish booting.  If you are using xm (default in 4.1, or
-     #May be better than the above, as grub can't pass all required infos
+ if you changed xendomains in 4.2), you should also enable xend:
-     #to the NetBSD/i386 kernel (e.g. console, root device, ...)
-     title NetBSD chain
+         xend=YES # only if using xm, and only installed <= 4.2
-       root        (hd0,0)
+         xencommons=YES
-       chainloader +1
+ TODO: Recommend for/against xen-watchdog.
-     ## end of grub config file.
+ After you have configured the daemons and either started them (in the
+ order given) or rebooted, use xm or xl to inspect Xen's boot messages,
- Install grub with the following command:
+ available resources, and running domains.  An example with xl follows:
-     # grub --no-floppy
+         # xl dmesg
+         [xen's boot info]
-     grub> root (hd0,a)
+         # xl info
-      Filesystem type is ffs, partition type 0xa9
+         [available memory, etc.]
+         # xl list
-     grub> setup (hd0)
+         Name              Id  Mem(MB)  CPU  State  Time(s)  Console
-      Checking if "/boot/grub/stage1" exists... no
+         Domain-0           0       64    0  r----     58.1
-      Checking if "/grub/stage1" exists... yes
-      Checking if "/grub/stage2" exists... yes
+ ### Issues with xencommons
-      Checking if "/grub/ffs_stage1_5" exists... yes
-      Running "embed /grub/ffs_stage1_5 (hd0)"...  14 sectors are embedded.
+ xencommons starts xenstored, which stores data on behalf of dom0 and
-     succeeded
+ domUs.  It does not currently work to stop and start xenstored.
-      Running "install /grub/stage1 (hd0) (hd0)1+14 p (hd0,0,a)/grub/stage2 /grub/menu.lst"...
+ Certainly all domUs should be shutdown first, following the sort order
-      succeeded
+ of the rc.d scripts.  However, the dom0 sets up state with xenstored,
-     Done.
+ and is not notified when xenstored exits, leading to not recreating
+ the state when the new xenstored starts.  Until there's a mechanism to
+ make this work, one should not expect to be able to restart xenstored
+ (and thus xencommons).  There is currently no reason to expect that
+ this will get fixed any time soon.
+ ### No-longer needed advice about devices
+ The installation of NetBSD should already have created devices for xen
+ (xencons, xenevt, xsd_kva), but if they are not present, create them:
+         cd /dev && sh MAKEDEV xen
+ anita (for testing NetBSD)
+ --------------------------
+ With the setup so far (assuming 4.2/xl), one should be able to run
+ anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as
+ root, because anita must create a domU):
+         anita --vmm=xl test file:///usr/obj/i386/
+ Alternatively, one can use --vmm=xm to use xm-based domU creation
+ instead (and must, on Xen <= 4.1).   TODO: confirm that anita xl really works.
+ Xen-specific NetBSD issues
+ --------------------------
+ There are (at least) two additional things different about NetBSD as a
+ dom0 kernel compared to hardware.
+ One is that the module ABI is different because some of the #defines
+ change, so one must build modules for Xen.  As of netbsd-7, the build
+ system does this automatically.  TODO: check this.  (Before building
+ Xen modules was added, it was awkward to use modules to the point
+ where it was considered that it did not work.)
+ The other difference is that XEN3_DOM0 does not have exactly the same
+ options as GENERIC.  While it is debatable whether or not this is a
+ bug, users should be aware of this and can simply add missing config
+ items if desired.
  Updating NetBSD in a dom0
  -------------------------
- Line 349  and adjusts /etc.
+ Line 583  and adjusts /etc.
  Note that one must update both the non-Xen kernel typically used for
  rescue purposes and the DOM0 kernel used with Xen.
- Updating Xen versions
+ Converting from grub to /boot
+ -----------------------------
+ These instructions were [TODO: will be] used to convert a system from
+ grub to /boot.  The system was originally installed in February of
+with a RAID1 setup and grub to boot Xen 2, and has been updated
+ over time.  Before these commands, it was running NetBSD 6 i386, Xen
+.1 and grub, much like the message linked earlier in the grub
+ section.
+         # Install MBR bootblocks on both disks.
+         fdisk -i /dev/rwd0d
+         fdisk -i /dev/rwd1d
+         # Install NetBSD primary boot loader (/ is FFSv1) into RAID1 components.
+         installboot -v /dev/rwd0d /usr/mdec/bootxx_ffsv1
+         installboot -v /dev/rwd1d /usr/mdec/bootxx_ffsv1
+         # Install secondary boot loader
+         cp -p /usr/mdec/boot /
+         # Create boot.cfg following earlier guidance:
+         menu=Xen:load /netbsd-XEN3PAE_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=256M
+         menu=Xen.ok:load /netbsd-XEN3PAE_DOM0.ok.gz console=pc;multiboot /xen.ok.gz dom0_mem=256M
+         menu=GENERIC:boot
+         menu=GENERIC single-user:boot -s
+         menu=GENERIC.ok:boot netbsd.ok
+         menu=GENERIC.ok single-user:boot netbsd.ok -s
+         menu=Drop to boot prompt:prompt
+         default=1
+         timeout=30
+ TODO: actually do this and fix it if necessary.
+ Upgrading Xen versions
  ---------------------
- TODO: write
+ Minor version upgrades are trivial.  Just rebuild/replace the
+ xenkernel version and copy the new xen.gz to / (where /boot.cfg
+ references it), and reboot.
+ Major version upgrades are conceptually not difficult, but can run
+ into all the issues found when installing Xen.  Assuming migration
+ from 4.1 to 4.2, remove the xenkernel41 and xentools41 packages and
+ install the xenkernel42 and xentools42 packages.  Copy the 4.2 xen.gz
+ to /.
+ Ensure that the contents of /etc/rc.d/xen* are correct.  Specifically,
+ they must match the package you just installed and not be left over
+ from some previous installation.
+ Enable the correct set of daemons; see the configuring section above.
+ (Upgrading from 3.x to 4.x without doing this will result in a hang.)
+ Ensure that the domU config files are valid for the new version.
+ Specifically, for 4.x remove autorestart=True, and ensure that disks
+ are specified with numbers as the second argument, as the examples
+ above show, and not NetBSD device names.
+ Hardware known to work
+ ----------------------
+ Arguably, this section is misplaced, and there should be a page of
+ hardware that runs NetBSD/amd64 well, with the mostly-well-founded
+ assumption that NetBSD/xen runs fine on any modern hardware that
+ NetBSD/amd64 runs well on.  Until then, we give motherboard/CPU (and
+ sometimes RAM) pairs/triples to aid those choosing a motherboard.
+ Note that Xen systems usually do not run X, so a listing here does not
+ imply that X works at all.
+         Supermicro X9SRL-F, Xeon E5-1650 v2, 96 GiB ECC
+         Supermicro ??, Atom C2758 (8 core), 32 GiB ECC
+         ASUS M5A78L-M/USB3 AM3+ microATX, AMD Piledriver X8 4000MHz, 16 GiB ECC
+ Older hardware:
+         Intel D915GEV, Pentium4 CPU 3.40GHz, 4GB 533MHz Synchronous DDR2
+         INTEL DG33FB, "Intel(R) Core(TM)2 Duo CPU     E6850  @ 3.00GHz"
+         INTEL DG33FB, "Intel(R) Core(TM)2 Duo CPU     E8400  @ 3.00GHz"
+ Running Xen under qemu
+ ----------------------
+ The astute reader will note that this section is somewhat twisted.
+ However, it can be useful to run Xen under qemu either because the
+ version of NetBSD as a dom0 does not run on the hardware in use, or to
+ generate automated test cases involving Xen.
+ In 2015-01, the following combination was reported to mostly work:
+         host OS: NetBSD/amd64 6.1.4
+         qemu: 2.2.0 from pkgsrc
+         Xen kernel: xenkernel42-4.2.5nb1 from pkgsrc
+         dom0 kernel: NetBSD/amd64 6.1.5
+         Xen tools: xentools42-4.2.5 from pkgsrc
+ See [PR 47720](https://gnats.netbsd.org/47720) for a problem with dom0
+ shutdown.
+ Unprivileged domains (domU)
+ ===========================
+ This section describes general concepts about domUs.  It does not
+ address specific domU operating systems or how to install them.  The
+ config files for domUs are typically in /usr/pkg/etc/xen, and are
+ typically named so that the file name, domU name and the domU's host
+ name match.
+ The domU is provided with CPU and memory by Xen, configured by the
+ dom0.  The domU is provided with disk and network by the dom0,
+ mediated by Xen, and configured in the dom0.
+ Entropy in domUs can be an issue; physical disks and network are on
+ the dom0.  NetBSD's /dev/random system works, but is often challenged.
+ Config files
+ ------------
+ There is no good order to present config files and the concepts
+ surrounding what is being configured.  We first show an example config
+ file, and then in the various sections give details.
+ See (at least in xentools41) /usr/pkg/share/examples/xen/xmexample*,
+ for a large number of well-commented examples, mostly for running
+ GNU/Linux.
+ The following is an example minimal domain configuration file
+ "/usr/pkg/etc/xen/foo".  It is (with only a name change) an actual
+ known working config file on Xen 4.1 (NetBSD 5 amd64 dom0 and NetBSD 5
+ i386 domU).  The domU serves as a network file server.
+         # -*- mode: python; -*-
+         kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz"
+         memory = 1024
+         vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ]
+         disk = [ 'file:/n0/xen/foo-wd0,0x0,w',
+                  'file:/n0/xen/foo-wd1,0x1,w' ]
+ The domain will have the same name as the file.  The kernel has the
+ host/domU name in it, so that on the dom0 one can update the various
+ domUs independently.  The vif line causes an interface to be provided,
+ with a specific mac address (do not reuse MAC addresses!), in bridge
+ mode.  Two disks are provided, and they are both writable; the bits
+ are stored in files and Xen attaches them to a vnd(4) device in the
+ dom0 on domain creation.  The system treats xbd0 as the boot device
+ without needing explicit configuration.
+ By default xm looks for domain config files in /usr/pkg/etc/xen.  Note
+ that "xm create" takes the name of a config file, while other commands
+ take the name of a domain.  To create the domain, connect to the
+ console, create the domain while attaching the console, shutdown the
+ domain, and see if it has finished stopping, do (or xl with Xen >=
+.2):
+         xm create foo
+         xm console foo
+         xm create -c foo
+         xm shutdown foo
+         xm list
+ Typing ^] will exit the console session.  Shutting down a domain is
+ equivalent to pushing the power button; a NetBSD domU will receive a
+ power-press event and do a clean shutdown.  Shutting down the dom0
+ will trigger controlled shutdowns of all configured domUs.
- Creating unprivileged domains (domU)
+ domU kernels
- ====================================
+ ------------
+ On a physical computer, the BIOS reads sector 0, and a chain of boot
+ loaders finds and loads a kernel.  Normally this comes from the root
+ file system.  With Xen domUs, the process is totally different.  The
+ normal path is for the domU kernel to be a file in the dom0's
+ file system.  At the request of the dom0, Xen loads that kernel into a
+ new domU instance and starts execution.  While domU kernels can be
+ anyplace, reasonable places to store domU kernels on the dom0 are in /
+ (so they are near the dom0 kernel), in /usr/pkg/etc/xen (near the
+ config files), or in /u0/xen (where the vdisks are).
+ Note that loading the domU kernel from the dom0 implies that boot
+ blocks, /boot, /boot.cfg, and so on are all ignored in the domU.
+ See the VPS section near the end for discussion of alternate ways to
+ obtain domU kernels.
+ CPU and memory
+ --------------
+ A domain is provided with some number of vcpus, less than the number
+ of CPUs seen by the hypervisor.  (For a dom0, this is controlled by
+ the boot argument "dom0_max_vcpus=1".)  For a domU, it is controlled
+ from the config file by the "vcpus = N" directive.
+ A domain is provided with memory; this is controlled in the config
+ file by "memory = N" (in megabytes).  In the straightforward case, the
+ sum of the the memory allocated to the dom0 and all domUs must be less
+ than the available memory.
+ Xen also provides a "balloon" driver, which can be used to let domains
+ use more memory temporarily.  TODO: Explain better, and explain how
+ well it works with NetBSD.
+ Virtual disks
+ -------------
+ With the file/vnd style, typically one creates a directory,
+ e.g. /u0/xen, on a disk large enough to hold virtual disks for all
+ domUs.  Then, for each domU disk, one writes zeros to a file that then
+ serves to hold the virtual disk's bits; a suggested name is foo-xbd0
+ for the first virtual disk for the domU called foo.  Writing zeros to
+ the file serves two purposes.  One is that preallocating the contents
+ improves performance.  The other is that vnd on sparse files has
+ failed to work.  TODO: give working/notworking NetBSD versions for
+ sparse vnd and gnats reference.  Note that the use of file/vnd for Xen
+ is not really different than creating a file-backed virtual disk for
+ some other purpose, except that xentools handles the vnconfig
+ commands.  To create an empty 4G virtual disk, simply do
+         dd if=/dev/zero of=foo-xbd0 bs=1m count=4096
+ Do not use qemu-img-xen, because this will create sparse file.  There
+ have been recent (2015) reports of sparse vnd(4) devices causing
+ lockups, but there is apparently no PR.
+ With the lvm style, one creates logical devices.  They are then used
+ similarly to vnds.  TODO: Add an example with lvm.
+ In domU config files, the disks are defined as a sequence of 3-tuples.
+ The first element is "method:/path/to/disk".  Common methods are
+ "file:" for file-backed vnd. and "phy:" for something that is already
+ a (TODO: character or block) device.
+ The second element is an artifact of how virtual disks are passed to
+ Linux, and a source of confusion with NetBSD Xen usage.  Linux domUs
+ are given a device name to associate with the disk, and values like
+ "hda1" or "sda1" are common.  In a NetBSD domU, the first disk appears
+ as xbd0, the second as xbd1, and so on.  However, xm/xl demand a
+ second argument.  The name given is converted to a major/minor by
+ calling stat(2) on the name in /dev and this is passed to the domU.
+ In the general case, the dom0 and domU can be different operating
+ systems, and it is an unwarranted assumption that they have consistent
+ numbering in /dev, or even that the dom0 OS has a /dev.  With NetBSD
+ as both dom0 and domU, using values of 0x0 for the first disk and 0x1
+ for the second works fine and avoids this issue.  For a GNU/Linux
+ guest, one can create /dev/hda1 in /dev, or to pass 0x301 for
+ /dev/hda1.
+ The third element is "w" for writable disks, and "r" for read-only
+ disks.
+ Note that NetBSD by default creates only vnd[0123].  If you need more
+ than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the
+ dom0.
+ Note that NetBSD by default creates only xbd[0123].  If you need more
+ virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU.
+ Virtual Networking
+ ------------------
+ Xen provides virtual Ethernets, each of which connects the dom0 and a
+ domU.  For each virtual network, there is an interface "xvifN.M" in
+ the dom0, and in domU index N, a matching interface xennetM (NetBSD
+ name).  The interfaces behave as if there is an Ethernet with two
+ adapters connected.  From this primitive, one can construct various
+ configurations.  We focus on two common and useful cases for which
+ there are existing scripts: bridging and NAT.
+ With bridging (in the example above), the domU perceives itself to be
+ on the same network as the dom0.  For server virtualization, this is
+ usually best.  Bridging is accomplished by creating a bridge(4) device
+ and adding the dom0's physical interface and the various xvifN.0
+ interfaces to the bridge.  One specifies "bridge=bridge0" in the domU
+ config file.  The bridge must be set up already in the dom0; an
+ example /etc/ifconfig.bridge0 is:
+         create
+         up
+         !brconfig bridge0 add wm0
+ With NAT, the domU perceives itself to be behind a NAT running on the
+ dom0.  This is often appropriate when running Xen on a workstation.
+ TODO: NAT appears to be configured by "vif = [ '' ]".
+ The MAC address specified is the one used for the interface in the new
+ domain.  The interface in dom0 will use this address XOR'd with
+:00:00:01:00:00.  Random MAC addresses are assigned if not given.
+ Sizing domains
+ --------------
+ Modern x86 hardware has vast amounts of resources.  However, many
+ virtual servers can function just fine on far less.  A system with
+M of RAM and a 4G disk can be a reasonable choice.  Note that it is
+ far easier to adjust virtual resources than physical ones.  For
+ memory, it's just a config file edit and a reboot.  For disk, one can
+ create a new file and vnconfig it (or lvm), and then dump/restore,
+ just like updating physical disks, but without having to be there and
+ without those pesky connectors.
+ Starting domains automatically
+ ------------------------------
+ To start domains foo at bar at boot and shut them down cleanly on dom0
+ shutdown, in rc.conf add:
+         xendomains="foo bar"
+ Note that earlier versions of the xentools41 xendomains rc.d script
+ used xl, when one should use xm with 4.1.
+ Creating specific unprivileged domains (domU)
+ =============================================
  Creating domUs is almost entirely independent of operating system.  We
- first explain NetBSD, and then differences for Linux and Solaris.
+ have already presented the basics of config files.  Note that you must
+ have already completed the dom0 setup so that "xl list" (or "xm list")
+ works.
  Creating an unprivileged NetBSD domain (domU)
  ---------------------------------------------
- Once you have *domain0* running, you need to start the xen tool daemon
+ See the earlier config file, and adjust memory.  Decide on how much
- (`/usr/pkg/share/examples/rc.d/xend start`) and the xen backend daemon
+ storage you will provide, and prepare it (file or lvm).
- (`/usr/pkg/share/examples/rc.d/xenbackendd start` for Xen3\*,
- `/usr/pkg/share/examples/rc.d/xencommons start` for Xen4.\*). Make sure
- that `/dev/xencons` and `/dev/xenevt` exist before starting `xend`. You
- can create them with this command:
-     # cd /dev && sh MAKEDEV xen
- xend will write logs to `/var/log/xend.log` and
- `/var/log/xend-debug.log`. You can then control xen with the xm tool.
- 'xm list' will show something like:
-     # xm list
-     Name              Id  Mem(MB)  CPU  State  Time(s)  Console
-     Domain-0           0       64    0  r----     58.1
- 'xm create' allows you to create a new domain. It uses a config file in
- PKG\_SYSCONFDIR for its parameters. By default, this file will be in
- `/usr/pkg/etc/xen/`. On creation, a kernel has to be specified, which
- will be executed in the new domain (this kernel is in the *domain0* file
- system, not on the new domain virtual disk; but please note, you should
- install the same kernel into *domainU* as `/netbsd` in order to make
- your system tools, like MAN.SAVECORE.8, work). A suitable kernel is
- provided as part of the i386 and amd64 binary sets: XEN3\_DOMU.
- Here is an /usr/pkg/etc/xen/nbsd example config file:
-     #  -*- mode: python; -*-
-     #============================================================================
-     # Python defaults setup for 'xm create'.
-     # Edit this file to reflect the configuration of your system.
-     #============================================================================
-     #----------------------------------------------------------------------------
-     # Kernel image file. This kernel will be loaded in the new domain.
-     kernel = "/home/bouyer/netbsd-XEN3_DOMU"
-     #kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
-     # Memory allocation (in megabytes) for the new domain.
-     memory = 128
-     # A handy name for your new domain. This will appear in 'xm list',
-     # and you can use this as parameters for xm in place of the domain
-     # number. All domains must have different names.
-     #
-     name = "nbsd"
-     # The number of virtual CPUs this domain has.
-     #
-     vcpus = 1
-     #----------------------------------------------------------------------------
-     # Define network interfaces for the new domain.
-     # Number of network interfaces (must be at least 1). Default is 1.
-     nics = 1
-     # Define MAC and/or bridge for the network interfaces.
-     #
-     # The MAC address specified in ``mac'' is the one used for the interface
-     # in the new domain. The interface in domain0 will use this address XOR'd
-     # with 00:00:00:01:00:00 (i.e. aa:00:00:51:02:f0 in our example). Random
-     # MACs are assigned if not given.
-     #
-     # ``bridge'' is a required parameter, which will be passed to the
-     # vif-script called by xend(8) when a new domain is created to configure
-     # the new xvif interface in domain0.
-     #
-     # In this example, the xvif is added to bridge0, which should have been
-     # set up prior to the new domain being created -- either in the
-     # ``network'' script or using a /etc/ifconfig.bridge0 file.
-     #
-     vif = [ 'mac=aa:00:00:50:02:f0, bridge=bridge0' ]
-     #----------------------------------------------------------------------------
-     # Define the disk devices you want the domain to have access to, and
-     # what you want them accessible as.
-     #
-     # Each disk entry is of the form:
-     #
-     #   phy:DEV,VDEV,MODE
-     #
-     # where DEV is the device, VDEV is the device name the domain will see,
-     # and MODE is r for read-only, w for read-write.  You can also create
-     # file-backed domains using disk entries of the form:
-     #
-     #   file:PATH,VDEV,MODE
-     #
-     # where PATH is the path to the file used as the virtual disk, and VDEV
-     # and MODE have the same meaning as for ``phy'' devices.
-     #
-     # VDEV doesn't really matter for a NetBSD guest OS (it's just used as an index),
-     # but it does for Linux.
-     # Worse, the device has to exist in /dev/ of domain0, because xm will
-     # try to stat() it. This means that in order to load a Linux guest OS
-     # from a NetBSD domain0, you'll have to create /dev/hda1, /dev/hda2, ...
-     # on domain0, with the major/minor from Linux :(
-     # Alternatively it's possible to specify the device number in hex,
-     # e.g. 0x301 for /dev/hda1, 0x302 for /dev/hda2, etc ...
-     disk = [ 'phy:/dev/wd0e,0x1,w' ]
+ While the kernel will be obtained from the dom0 file system, the same
-     #disk = [ 'file:/var/xen/nbsd-disk,0x01,w' ]
+ file should be present in the domU as /netbsd so that tools like
-     #disk = [ 'file:/var/xen/nbsd-disk,0x301,w' ]
+ savecore(8) can work.   (This is helpful but not necessary.)
+ The kernel must be specifically for Xen and for use as a domU.  The
+ i386 and amd64 provide the following kernels:
+         i386 XEN3_DOMU
+         i386 XEN3PAE_DOMU
+         amd64 XEN3_DOMU
+ Unless using Xen 3.1 (and you shouldn't) with i386-mode Xen, you must
+ use the PAE version of the i386 kernel.
-     #----------------------------------------------------------------------------
+ This will boot NetBSD, but this is not that useful if the disk is
-     # Set the kernel command line for the new domain.
+ empty.  One approach is to unpack sets onto the disk outside of xen
+ (by mounting it, just as you would prepare a physical disk for a
+ system you can't run the installer on).
-     # Set root device. This one does matter for NetBSD
+ A second approach is to run an INSTALL kernel, which has a miniroot
-     root = "xbd0"
+ and can load sets from the network.  To do this, copy the INSTALL
-     # extra parameters passed to the kernel
+ kernel to / and change the kernel line in the config file to:
-     # this is where you can set boot flags like -s, -a, etc ...
-     #extra = ""
-     #----------------------------------------------------------------------------
-     # Set according to whether you want the domain restarted when it exits.
-     # The default is False.
-     #autorestart = True
-     # end of nbsd config file ====================================================
- When a new domain is created, xen calls the
- `/usr/pkg/etc/xen/vif-bridge` script for each virtual network interface
- created in *domain0*. This can be used to automatically configure the
- xvif?.? interfaces in *domain0*. In our example, these will be bridged
- with the bridge0 device in *domain0*, but the bridge has to exist first.
- To do this, create the file `/etc/ifconfig.bridge0` and make it look
- like this:
-     create
-     !brconfig $int add ex0 up
- (replace `ex0` with the name of your physical interface). Then bridge0
- will be created on boot. See the MAN.BRIDGE.4 man page for details.
- So, here is a suitable `/usr/pkg/etc/xen/vif-bridge` for xvif?.? (a
- working vif-bridge is also provided with xentools20) configuring:
-     #!/bin/sh
-     #============================================================================
-     # $NetBSD: howto.mdwn,v 1.18 2014/12/24 00:11:15 gdt Exp $
-     #
-     # /usr/pkg/etc/xen/vif-bridge
-     #
-     # Script for configuring a vif in bridged mode with a dom0 interface.
-     # The xend(8) daemon calls a vif script when bringing a vif up or down.
-     # The script name to use is defined in /usr/pkg/etc/xen/xend-config.sxp
-     # in the ``vif-script'' field.
-     #
-     # Usage: vif-bridge up|down [var=value ...]
-     #
-     # Actions:
-     #    up     Adds the vif interface to the bridge.
-     #    down   Removes the vif interface from the bridge.
-     #
-     # Variables:
-     #    domain name of the domain the interface is on (required).
-     #    vifq   vif interface name (required).
-     #    mac    vif MAC address (required).
-     #    bridge bridge to add the vif to (required).
-     #
-     # Example invocation:
-     #
-     # vif-bridge up domain=VM1 vif=xvif1.0 mac="ee:14:01:d0:ec:af" bridge=bridge0
-     #
-     #============================================================================
-     # Exit if anything goes wrong
-     set -e
-     echo "vif-bridge $*"
-     # Operation name.
-     OP=$1; shift
-     # Pull variables in args into environment
-     for arg ; do export "${arg}" ; done
-     # Required parameters. Fail if not set.
-     domain=${domain:?}
-     vif=${vif:?}
-     mac=${mac:?}
-     bridge=${bridge:?}
-     # Optional parameters. Set defaults.
-     ip=${ip:-''}   # default to null (do nothing)
-     # Are we going up or down?
-     case $OP in
-     up) brcmd='add' ;;
-     down)   brcmd='delete' ;;
-     *)
-         echo 'Invalid command: ' $OP
-         echo 'Valid commands are: up, down'
-         exit 1
-         ;;
-     esac
-     # Don't do anything if the bridge is "null".
-     if [ "${bridge}" = "null" ] ; then
-         exit
-     fi
-     # Don't do anything if the bridge doesn't exist.
-     if ! ifconfig -l | grep "${bridge}" >/dev/null; then
-         exit
-     fi
-     # Add/remove vif to/from bridge.
-     ifconfig x${vif} $OP
-     brconfig ${bridge} ${brcmd} x${vif}
- Now, running
-     xm create -c /usr/pkg/etc/xen/nbsd
- should create a domain and load a NetBSD kernel in it. (Note: `-c`
- causes xm to connect to the domain's console once created.) The kernel
- will try to find its root file system on xbd0 (i.e., wd0e) which hasn't
- been created yet. wd0e will be seen as a disk device in the new domain,
- so it will be 'sub-partitioned'. We could attach a ccd to wd0e in
- *domain0* and partition it, newfs and extract the NetBSD/i386 or amd64
- tarballs there, but there's an easier way: load the
- `netbsd-INSTALL_XEN3_DOMU` kernel provided in the NetBSD binary sets.
- Like other install kernels, it contains a ramdisk with sysinst, so you
- can install NetBSD using sysinst on your new domain.
- If you want to install NetBSD/Xen with a CDROM image, the following line
+         kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU"
- should be used in the `/usr/pkg/etc/xen/nbsd` file:
+ Then, start the domain as "xl create -c configname".
+ Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following
+ line should be used in the config file.
      disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ]
  After booting the domain, the option to install via CDROM may be
- selected. The CDROM device should be changed to `xbd1d`.
+ selected.  The CDROM device should be changed to `xbd1d`.
- Once done installing, `halt -p` the new domain (don't reboot or halt, it
+ Once done installing, "halt -p" the new domain (don't reboot or halt,
- would reload the INSTALL\_XEN3\_DOMU kernel even if you changed the
+ it would reload the INSTALL_XEN3_DOMU kernel even if you changed the
- config file), switch the config file back to the XEN3\_DOMU kernel, and
+ config file), switch the config file back to the XEN3_DOMU kernel,
- start the new domain again. Now it should be able to use `root on xbd0a`
+ and start the new domain again. Now it should be able to use "root on
- and you should have a second, functional NetBSD system on your xen
+ xbd0a" and you should have a, functional NetBSD domU.
- installation.
+ TODO: check if this is still accurate.
  When the new domain is booting you'll see some warnings about *wscons*
  and the pseudo-terminals. These can be fixed by editing the files
  `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in
- Line 620  Finally, all screens must be commented o
+ Line 961  Finally, all screens must be commented o
  It is also desirable to add
-     powerd=YES
+         powerd=YES
  in rc.conf. This way, the domain will be properly shut down if
- `xm shutdown -R` or `xm shutdown -H` is used on the domain0.
+ `xm shutdown -R` or `xm shutdown -H` is used on the dom0.
- Your domain should be now ready to work, enjoy.
+ It is not strictly necessary to have a kernel (as /netbsd) in the domU
+ file system.  However, various programs (e.g. netstat) will use that
+ kernel to look up symbols to read from kernel virtual memory.  If
+ /netbsd is not the running kernel, those lookups will fail.  (This is
+ not really a Xen-specific issue, but because the domU kernel is
+ obtained from the dom0, it is far more likely to be out of sync or
+ missing with Xen.)
  Creating an unprivileged Linux domain (domU)
  --------------------------------------------
- Line 639  the example below)
+ Line 986  the example below)
      disk = [ 'phy:/dev/wd0e,0x1,w' ]
  does matter to Linux. It wants a Linux device number here (e.g. 0x300
- for hda). Linux builds device numbers as: (major \<\< 8 + minor). So,
+ for hda).  Linux builds device numbers as: (major \<\< 8 + minor).
- hda1 which has major 3 and minor 1 on a Linux system will have device
+ So, hda1 which has major 3 and minor 1 on a Linux system will have
- number 0x301. Alternatively, devices names can be used (hda, hdb, ...)
+ device number 0x301.  Alternatively, devices names can be used (hda,
- as xentools has a table to map these names to devices numbers. To export
+ hdb, ...)  as xentools has a table to map these names to devices
- a partition to a Linux guest we can use:
+ numbers.  To export a partition to a Linux guest we can use:
-     disk = [ 'phy:/dev/wd0e,0x300,w' ]
+         disk = [ 'phy:/dev/wd0e,0x300,w' ]
-     root = "/dev/hda1 ro"
+         root = "/dev/hda1 ro"
  and it will appear as /dev/hda on the Linux system, and be used as root
  partition.
- To install the Linux system on the partition to be exported to the guest
+ To install the Linux system on the partition to be exported to the
- domain, the following method can be used: install sysutils/e2fsprogs
+ guest domain, the following method can be used: install
- from pkgsrc. Use mke2fs to format the partition that will be the root
+ sysutils/e2fsprogs from pkgsrc.  Use mke2fs to format the partition
- partition of your Linux domain, and mount it. Then copy the files from a
+ that will be the root partition of your Linux domain, and mount it.
- working Linux system, make adjustments in `/etc` (fstab, network
+ Then copy the files from a working Linux system, make adjustments in
- config). It should also be possible to extract binary packages such as
+ `/etc` (fstab, network config).  It should also be possible to extract
- .rpm or .deb directly to the mounted partition using the appropriate
+ binary packages such as .rpm or .deb directly to the mounted partition
- tool, possibly running under NetBSD's Linux emulation. Once the
+ using the appropriate tool, possibly running under NetBSD's Linux
- filesystem has been populated, umount it. If desirable, the filesystem
+ emulation.  Once the file system has been populated, umount it.  If
- can be converted to ext3 using tune2fs -j. It should now be possible to
+ desirable, the file system can be converted to ext3 using tune2fs -j.
- boot the Linux guest domain, using one of the vmlinuz-\*-xenU kernels
+ It should now be possible to boot the Linux guest domain, using one of
- available in the Xen binary distribution.
+ the vmlinuz-\*-xenU kernels available in the Xen binary distribution.
- To get the linux console right, you need to add:
+ To get the Linux console right, you need to add:
      extra = "xencons=tty1"
- to your configuration since not all linux distributions auto-attach a
+ to your configuration since not all Linux distributions auto-attach a
  tty to the xen console.
  Creating an unprivileged Solaris domain (domU)
  ----------------------------------------------
- Download an Opensolaris [release](http://opensolaris.org/os/downloads/)
+ See possibly outdated
- or [development snapshot](http://genunix.org/) DVD image. Attach the DVD
+ [Solaris domU instructions](/ports/xen/howto-solaris/).
- image to a MAN.VND.4 device. Copy the kernel and ramdisk filesystem
- image to your dom0 filesystem.
+ PCI passthrough: Using PCI devices in guest domains
-     dom0# mkdir /root/solaris
+ ---------------------------------------------------
-     dom0# vnconfig vnd0 osol-1002-124-x86.iso
-     dom0# mount /dev/vnd0a /mnt
+ The dom0 can give other domains access to selected PCI
+ devices. This can allow, for example, a non-privileged domain to have
-     ## for a 64-bit guest
+ access to a physical network interface or disk controller.  However,
-     dom0# cp /mnt/boot/amd64/x86.microroot /root/solaris
+ keep in mind that giving a domain access to a PCI device most likely
-     dom0# cp /mnt/platform/i86xpv/kernel/amd64/unix /root/solaris
+ will give the domain read/write access to the whole physical memory,
+ as PCs don't have an IOMMU to restrict memory access to DMA-capable
-     ## for a 32-bit guest
+ device.  Also, it's not possible to export ISA devices to non-dom0
-     dom0# cp /mnt/boot/x86.microroot /root/solaris
+ domains, which means that the primary VGA adapter can't be exported.
-     dom0# cp /mnt/platform/i86xpv/kernel/unix /root/solaris
+ A guest domain trying to access the VGA registers will panic.
-     dom0# umount /mnt
+ If the dom0 is NetBSD, it has to be running Xen 3.1, as support has
+ not been ported to later versions at this time.
- Keep the MAN.VND.4 configured. For some reason the boot process stalls
+ For a PCI device to be exported to a domU, is has to be attached to
- unless the DVD image is attached to the guest as a "phy" device. Create
+ the "pciback" driver in dom0.  Devices passed to the dom0 via the
- an initial configuration file with the following contents. Substitute
+ pciback.hide boot parameter will attach to "pciback" instead of the
- */dev/wd0k* with an empty partition at least 8 GB large.
+ usual driver.  The list of devices is specified as "(bus:dev.func)",
-     memory = 640
-     name = 'solaris'
-     disk = [ 'phy:/dev/wd0k,0,w' ]
-     disk += [ 'phy:/dev/vnd0d,6:cdrom,r' ]
-     vif = [ 'bridge=bridge0' ]
-     kernel = '/root/solaris/unix'
-     ramdisk = '/root/solaris/x86.microroot'
-     # for a 64-bit guest
-     extra = '/platform/i86xpv/kernel/amd64/unix - nowin -B install_media=cdrom'
-     # for a 32-bit guest
-     #extra = '/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'
- Start the guest.
-     dom0# xm create -c solaris.cfg
-     Started domain solaris
-                           v3.3.2 chgset 'unavailable'
-     SunOS Release 5.11 Version snv_124 64-bit
-     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
-     Use is subject to license terms.
-     Hostname: opensolaris
-     Remounting root read/write
-     Probing for device nodes ...
-     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
-     Preparing live image for use
-     Done mounting Live image
- Make sure the network is configured. Note that it can take a minute for
- the xnf0 interface to appear.
-     opensolaris console login: jack
-     Password: jack
-     Sun Microsystems Inc.   SunOS 5.11      snv_124 November 2008
-     jack@opensolaris:~$ pfexec sh
-     sh-3.2# ifconfig -a
-     sh-3.2# exit
- Set a password for VNC and start the VNC server which provides the X11
- display where the installation program runs.
-     jack@opensolaris:~$ vncpasswd
-     Password: solaris
-     Verify: solaris
-     jack@opensolaris:~$ cp .Xclients .vnc/xstartup
-     jack@opensolaris:~$ vncserver :1
- From a remote machine connect to the VNC server. Use `ifconfig xnf0` on
- the guest to find the correct IP address to use.
-     remote$ vncviewer 172.18.2.99:1
- It is also possible to launch the installation on a remote X11 display.
-     jack@opensolaris:~$ export DISPLAY=172.18.1.1:0
-     jack@opensolaris:~$ pfexec gui-install
- After the GUI installation is complete you will be asked to reboot.
- Before that you need to determine the ZFS ID for the new boot filesystem
- and update the configuration file accordingly. Return to the guest
- console.
-     jack@opensolaris:~$ pfexec zdb -vvv rpool | grep bootfs
-                     bootfs = 43
-     ^C
-     jack@opensolaris:~$
- The final configuration file should look like this. Note in particular
- the last line.
-     memory = 640
-     name = 'solaris'
-     disk = [ 'phy:/dev/wd0k,0,w' ]
-     vif = [ 'bridge=bridge0' ]
-     kernel = '/root/solaris/unix'
-     ramdisk = '/root/solaris/x86.microroot'
-     extra = '/platform/i86xpv/kernel/amd64/unix -B zfs-bootfs=rpool/43,bootpath="/xpvd/xdf@0:a"'
- Restart the guest to verify it works correctly.
-     dom0# xm destroy solaris
-     dom0# xm create -c solaris.cfg
-     Using config file "./solaris.cfg".
-     v3.3.2 chgset 'unavailable'
-     Started domain solaris
-     SunOS Release 5.11 Version snv_124 64-bit
-     Copyright 1983-2009 Sun Microsystems, Inc.  All rights reserved.
-     Use is subject to license terms.
-     WARNING: emlxs: ddi_modopen drv/fct failed: err 2
-     Hostname: osol
-     Configuring devices.
-     Loading smf(5) service descriptions: 160/160
-     svccfg import warnings. See /var/svc/log/system-manifest-import:default.log .
-     Reading ZFS config: done.
-     Mounting ZFS filesystems: (6/6)
-     Creating new rsa public/private host key pair
-     Creating new dsa public/private host key pair
-     osol console login:
- Using PCI devices in guest domains
- ----------------------------------
- The domain0 can give other domains access to selected PCI devices. This
- can allow, for example, a non-privileged domain to have access to a
- physical network interface or disk controller. However, keep in mind
- that giving a domain access to a PCI device most likely will give the
- domain read/write access to the whole physical memory, as PCs don't have
- an IOMMU to restrict memory access to DMA-capable device. Also, it's not
- possible to export ISA devices to non-domain0 domains (which means that
- the primary VGA adapter can't be exported. A guest domain trying to
- access the VGA registers will panic).
- This functionality is only available in NetBSD-5.1 (and later) domain0
- and domU. If the domain0 is NetBSD, it has to be running Xen 3.1, as
- support has not been ported to later versions at this time.
- For a PCI device to be exported to a domU, is has to be attached to the
- `pciback` driver in domain0. Devices passed to the domain0 via the
- pciback.hide boot parameter will attach to `pciback` instead of the
- usual driver. The list of devices is specified as `(bus:dev.func)`,
  where bus and dev are 2-digit hexadecimal numbers, and func a
  single-digit number:
-     pciback.hide=(00:0a.0)(00:06.0)
+         pciback.hide=(00:0a.0)(00:06.0)
- pciback devices should show up in the domain0's boot messages, and the
+ pciback devices should show up in the dom0's boot messages, and the
  devices should be listed in the `/kern/xen/pci` directory.
- PCI devices to be exported to a domU are listed in the `pci` array of
+ PCI devices to be exported to a domU are listed in the "pci" array of
- the domU's config file, with the format `'0000:bus:dev.func'`
+ the domU's config file, with the format "0000:bus:dev.func".
+         pci = [ '0000:00:06.0', '0000:00:0a.0' ]
+ In the domU an "xpci" device will show up, to which one or more pci
+ buses will attach.  Then the PCI drivers will attach to PCI buses as
+ usual.  Note that the default NetBSD DOMU kernels do not have "xpci"
+ or any PCI drivers built in by default; you have to build your own
+ kernel to use PCI devices in a domU.  Here's a kernel config example;
+ note that only the "xpci" lines are unusual.
+         include         "arch/i386/conf/XEN3_DOMU"
+         # Add support for PCI buses to the XEN3_DOMU kernel
+         xpci* at xenbus ?
+         pci* at xpci ?
+         # PCI USB controllers
+         uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
+         # USB bus support
+         usb*    at uhci?
+         # USB Hubs
+         uhub*   at usb?
+         uhub*   at uhub? port ? configuration ? interface ?
+         # USB Mass Storage
+         umass*  at uhub? port ? configuration ? interface ?
+         wd*     at umass?
+         # SCSI controllers
+         ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
+         # SCSI bus support (for both ahc and umass)
+         scsibus* at scsi?
+         # SCSI devices
+         sd*     at scsibus? target ? lun ?      # SCSI disk drives
+         cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
+ NetBSD as a domU in a VPS
+ =========================
+ The bulk of the HOWTO is about using NetBSD as a dom0 on your own
+ hardware.  This section explains how to deal with Xen in a domU as a
+ virtual private server where you do not control or have access to the
+ dom0.  This is not intended to be an exhaustive list of VPS providers;
+ only a few are mentioned that specifically support NetBSD.
+ VPS operators provide varying degrees of access and mechanisms for
+ configuration.  The big issue is usually how one controls which kernel
+ is booted, because the kernel is nominally in the dom0 file system (to
+ which VPS users do not normally have access).  A second issue is how
+ to install NetBSD.
+ A VPS user may want to compile a kernel for security updates, to run
+ npf, run IPsec, or any other reason why someone would want to change
+ their kernel.
+ One approach is to have an administrative interface to upload a kernel,
+ or to select from a prepopulated list.  Other approaches are pygrub
+ (deprecated) and pvgrub, which are ways to have a bootloader obtain a
+ kernel from the domU file system.  This is closer to a regular physical
+ computer, where someone who controls a machine can replace the kernel.
+ A second issue is multiple CPUs.  With NetBSD 6, domUs support
+ multiple vcpus, and it is typical for VPS providers to enable multiple
+ CPUs for NetBSD domUs.
+ pygrub
+ -------
+ pygrub runs in the dom0 and looks into the domU file system.  This
+ implies that the domU must have a kernel in a file system in a format
+ known to pygrub.  As of 2014, pygrub seems to be of mostly historical
+ interest.
+ pvgrub
+ ------
+ pvgrub is a version of grub that uses PV operations instead of BIOS
+ calls.  It is booted from the dom0 as the domU kernel, and then reads
+ /grub/menu.lst and loads a kernel from the domU file system.
+ [Panix](http://www.panix.com/) lets users use pvgrub.  Panix reports
+ that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes
+ (and hence with defaults from "newfs -O 2").  See [Panix's pvgrub
+ page](http://www.panix.com/v-colo/grub.html), which describes only
+ Linux but should be updated to cover NetBSD :-).
+ [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot
+ their own kernel.  See then [prgmr.com NetBSD
+ HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU)
+ (which is in need of updating).
+ It appears that [grub's FFS
+ code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c)
+ does not support all aspects of modern FFS, but there are also reports
+ that FFSv2 works fine.  At prgmr, typically one has an ext2 or FAT
+ partition for the kernel with the intent that grub can understand it,
+ which leads to /netbsd not being the actual kernel.  One must remember
+ to update the special boot partition.
+ Amazon
+ ------
+ See the [Amazon EC2 page](../amazon_ec2/).
+ Using npf
+ ---------
-     pci = [ '0000:00:06.0', '0000:00:0a.0' ]
+ In standard kernels, npf is a module, and thus cannot be loaded in a
+ DOMU kernel.
- In the domU an `xpci` device will show up, to which one or more pci
+ TODO: Explain how to compile npf into a custom kernel, answering (but
- busses will attach. Then the PCI drivers will attach to PCI busses as
+ note that the problem was caused by not booting the right kernel)
- usual. Note that the default NetBSD DOMU kernels do not have `xpci` or
+ [this email to
- any PCI drivers built in by default; you have to build your own kernel
+ netbsd-users](https://mail-index.netbsd.org/netbsd-users/2014/12/26/msg015576.html).
- to use PCI devices in a domU. Here's a kernel config example:
+ TODO items for improving NetBSD/xen
-     include         "arch/i386/conf/XEN3_DOMU"
+ ===================================
-     #include         "arch/i386/conf/XENU"           # in NetBSD 3.0
+ * Make the NetBSD dom0 kernel work with SMP.
-     # Add support for PCI busses to the XEN3_DOMU kernel
+ * Test the Xen 4.5 packages adequately to be able to recommend them as
-     xpci* at xenbus ?
+   the standard approach.
-     pci* at xpci ?
+ * Get PCI passthrough working on Xen 4.5
+ * Get pvgrub into pkgsrc, either via xentools or separately.
-     # Now add PCI and related devices to be used by this domain
+ * grub
-     # USB Controller and Devices
+   * Check/add support to pkgsrc grub2 for UFS2 and arbitrary
+     fragsize/blocksize (UFS2 support may be present; the point is to
-     # PCI USB controllers
+     make it so that with any UFS1/UFS2 file system setup that works
-     uhci*   at pci? dev ? function ?        # Universal Host Controller (Intel)
+     with NetBSD grub will also work).
+     See [pkg/40258](https://gnats.netbsd.org/40258).
-     # USB bus support
+   * Push patches upstream.
-     usb*    at uhci?
+   * Get UFS2 patches into pvgrub.
+ * Add support for PV ops to a version of /boot, and make it usable as
-     # USB Hubs
+   a kernel in Xen, similar to pvgrub.
-     uhub*   at usb?
+ * Solve somehow the issue with modules for GENERIC not being loadable
-     uhub*   at uhub? port ? configuration ? interface ?
+   in a Xen dom0 or domU kernel.
-     # USB Mass Storage
+ Random pointers
-     umass*  at uhub? port ? configuration ? interface ?
+ ===============
-     wd*     at umass?
-     # SCSI controllers
+ This section contains links from elsewhere not yet integrated into the
-     ahc*    at pci? dev ? function ?        # Adaptec [23]94x, aic78x0 SCSI
+ HOWTO, and other guides.
-     # SCSI bus support (for both ahc and umass)
+ * http://www.lumbercartel.ca/library/xen/
-     scsibus* at scsi?
+ * http://pbraun.nethence.com/doc/sysutils/xen_netbsd_dom0.html
+ * https://gmplib.org/~tege/xen.html
-     # SCSI devices
-     sd*     at scsibus? target ? lun ?      # SCSI disk drives
-     cd*     at scsibus? target ? lun ?      # SCSI CD-ROM drives
- Links and further information
- =============================
- -   The [HowTo on Installing into RAID-1](http://mail-index.NetBSD.org/port-xen/2006/03/01/0010.html)
-     explains how to set up booting a dom0 with Xen using grub
-     with NetBSD's RAIDframe.  (This is obsolete with the use of
-     NetBSD's native boot.)
- -   An example of how to use NetBSD's native bootloader to load
-     NetBSD/Xen instead of Grub can be found in the i386/amd64 boot(8)
-     and boot.cfg(5) manpages.

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

Removed from v.1.19
changed lines
	Added in v.1.141