wikisrc/ports/xen/howto.mdwn - view

File: [NetBSD Developer Wiki] / wikisrc / ports / xen / howto.mdwn
Revision 1.190: download - view: text, annotated - select for diffs
Fri Feb 26 18:39:20 2021 UTC (7 weeks, 2 days ago) by gdt
Branches: MAIN
CVS tags: HEAD

xen: Demote info about migration from grub

At this point anything about grub is ancient history and there's no
reason to believe there are any maintained netbsd/xen systems with
grub.  We already have a place to store the old info, so simply demote
the conversion information to tidy the main HOWTO.

1: [[!meta title="Xen HowTo"]] 2: 3: Xen is a Type 1 hypervisor which supports running multiple guest operating 4: systems on a single physical machine. One uses the Xen kernel to control the 5: CPU, memory and console, a dom0 operating system which mediates access to 6: other hardware (e.g., disks, network, USB), and one or more domU operating 7: systems which operate in an unprivileged virtualized environment. IO requests 8: from the domU systems are forwarded by the Xen hypervisor to the dom0 to be 9: fulfilled. 10: 11: This HOWTO presumes a basic familiarity with the Xen system 12: architecture, with installing NetBSD on amd64 hardware, and with 13: installing software from pkgsrc. See also the [Xen 14: website](http://www.xenproject.org/). 15: 16: [[!toc]] 17: 18: # Overview 19: 20: The basic concept of Xen is that the hypervisor (xenkernel) runs on 21: the hardware, and runs a privileged domain ("dom0") that can access 22: disks/networking/etc. One then runs additonal unprivileged domains 23: (each a "domU"), presumably to do something useful. 24: 25: This HOWTO addresses how to run a NetBSD dom0 (and hence also build 26: xen itself). It also addresses how to run domUs in that environment, 27: and how to deal with having a domU in a Xen environment run by someone 28: else and/or not running NetBSD. 29: 30: There are many choices one can make; the HOWTO recommends the standard 31: approach and limits discussion of alternatives in many cases. 32: 33: ## Guest Styles 34: 35: Xen supports different styles of guests. 36: 37: [[!table data=""" 38: Style of guest |Supported by NetBSD 39: PV |Yes (dom0, domU) 40: HVM |Yes (domU) 41: PVHVM |current-only (domU) 42: PVH |current-only (domU, dom0 not yet) 43: """]] 44: 45: In Para-Virtualized (PV) mode, the guest OS does not attempt to access 46: hardware directly, but instead makes hypercalls to the hypervisor; PV 47: guests must be specifically coded for Xen. 48: See [PV](https://wiki.xen.org/wiki/Paravirtualization_(PV\)). 49: 50: In HVM mode, no guest modification is required; however, hardware 51: support is required, such as VT-x on Intel CPUs and SVM on AMD CPUs. 52: The dom0 runs qemu to emulate hardware. 53: 54: In PVHVM mode, the guest runs as HVM, but additionally can use PV 55: drivers for efficiency. 56: See [PV on HVM](https://wiki.xen.org/wiki/PV_on_HVM). 57: 58: There have been two PVH modes: original PVH and PVHv2. Original PVH 59: was based on PV mode and is no longer relevant at all. PVHv2 is 60: basically lightweight HVM with PV drivers. A critical feature of it 61: is that qemu is not needed; the hypervisor can do the emulation that 62: is required. Thus, a dom0 can be PVHv2. 63: The source code uses PVH and config files use pvh; this refers to PVHv2. 64: See [PVH(v2)](https://wiki.xenproject.org/wiki/PVH_(v2\)_Domu). 65: 66: At system boot, the dom0 kernel is loaded as a module with Xen as the kernel. 67: The dom0 can start one or more domUs. (Booting is explained in detail 68: in the dom0 section.) 69: 70: ## CPU Architecture 71: 72: Xen runs on x86_64 hardware (the NetBSD amd64 port). 73: 74: There is a concept of Xen running on ARM, but there are no reports of this working with NetBSD. 75: 76: The dom0 system should be amd64. (Instructions for i386PAE dom0 have been removed from the HOWTO.) 77: 78: The domU can be i386PAE or amd64. 79: i386PAE at one point was considered as [faster](https://lists.xen.org/archives/html/xen-devel/2012-07/msg00085.html) than amd64. 80: 81: ## Xen Versions 82: 83: In NetBSD, Xen is provided in pkgsrc, via matching pairs of packages 84: xenkernel and xentools. We will refer only to the kernel versions, 85: but note that both packages must be installed together and must have 86: matching versions. 87: 88: Versions available in pkgsrc: 89: 90: [[!table data=""" 91: Xen Version |Package Name |Xen CPU Support |EOL'ed By Upstream 92: 4.11 |xenkernel411 |x86_64 |No 93: 4.13 |xenkernel413 |x86_64 |No 94: """]] 95: 96: See also the [Xen Security Advisory page](http://xenbits.xen.org/xsa/). 97: 98: Older Xen had a python-based management tool called xm, now replaced 99: by xl. 100: 101: ## NetBSD versions 102: 103: Xen has been supported in NetBSD for a long time, at least since 2005. 104: Initially Xen was PV only. 105: 106: NetBSD 8 and up support PV and HVM modes. 107: 108: Support for PVHVM and PVH is available only in NetBSD-current. 109: 110: NetBSD as a dom0 does not run SMP, because some drivers are not yet 111: safe for this. \todo Link to more information about what needs work. 112: 113: NetBSD, when run as a domU, can and does typically run SMP. 114: 115: Note: NetBSD support is called XEN3. However, it does support Xen 4, 116: because the hypercall interface has remained identical. 117: 118: # Creating a NetBSD dom0 119: 120: In order to install a NetBSD as a dom0, one must first install a normal 121: NetBSD system, and then pivot the install to a dom0 install by changing 122: the kernel and boot configuration. 123: 124: In 2018-05, trouble booting a dom0 was reported with 256M of RAM: with 125: 512M it worked reliably. This does not make sense, but if you see 126: "not ELF" after Xen boots, try increasing dom0 RAM. 127: 128: ## Installation of NetBSD 129: 130: [Install NetBSD/amd64](/guide/inst/) 131: just as you would if you were not using Xen. 132: Therefore, use the most recent release, or a build from the most recent stable branch. 133: 134: ## Installation of Xen 135: 136: Use the most recent version of Xen in pkgsrc, unless the DESCR says that it is not suitable. 137: Therefore, choose 4.13. 138: In the dom0, install xenkernel413 and xentools413 from pkgsrc. 139: 140: Once this is done, copy the Xen kernel from where pkgsrc puts it to 141: where the boot process will be able to find it: 142: 143: [[!template id=programlisting text=""" 144: # cp -p /usr/pkg/xen413-kernel/xen.gz / 145: """]] 146: 147: Then, place a NetBSD XEN3_DOM0 kernel in the `/` directory. Such kernel 148: can either be compiled manually, or downloaded from the NetBSD FTP, for 149: example at: 150: 151: [[!template id=programlisting text=""" 152: ftp.netbsd.org/pub/NetBSD/NetBSD-9.1/amd64/binary/kernel/netbsd-XEN3_DOM0.gz 153: """]] 154: 155: Add a line to /boot.cfg to boot Xen: 156: 157: [[!template id=filecontent name="/boot.cfg" text=""" 158: menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz dom0_mem=512M 159: """]] 160: 161: This specifies that the dom0 should have 512MB of ram, leaving the rest 162: to be allocated for domUs. 163: 164: NB: This says add, not replace, so that you will be able to boot a 165: NetBSD kernel without Xen. Once Xen boots ok, you may want to set it 166: as default. 167: 168: To use a serial console, add settings as follows: 169: 170: [[!template id=filecontent name="/boot.cfg" text=""" 171: menu=Xen:load /netbsd-XEN3_DOM0.gz;multiboot /xen.gz dom0_mem=512M console=com1 com1=9600,8n1 172: """]] 173: 174: which will use the first serial port for Xen (which counts starting 175: from 1, unlike NetBSD which counts starting from 0), forcing 176: speed/parity. Because the NetBSD command line lacks a 177: "console=pc" argument, it will use the default "xencons" console device, 178: which directs the console I/O through Xen to the same console device Xen 179: itself uses (in this case, the serial port). 180: 181: In an attempt to add performance, one can also add `dom0_max_vcpus=1 dom0_vcpus_pin`, 182: to force only one vcpu to be provided (since NetBSD dom0 can't use 183: more) and to pin that vcpu to a physical CPU. Xen has 184: [many boot options](http://xenbits.xenproject.org/docs/4.13-testing/misc/xen-command-line.html), 185: and other than dom0 memory and max_vcpus, they are generally not 186: necessary. 187: 188: Ensure that the boot scripts installed in 189: `/usr/pkg/share/examples/rc.d` are in `/etc/rc.d`, either because you 190: have `PKG_RCD_SCRIPTS=yes`, or manually. (This is not special to Xen, 191: but a normal part of pkgsrc usage.) 192: 193: Set `xencommons=YES` in rc.conf: 194: 195: [[!template id=filecontent name="/etc/rc.conf" text=""" 196: xencommons=YES 197: """]] 198: 199: \todo Recommend for/against xen-watchdog. 200: 201: Now, reboot so that you are running a DOM0 kernel under Xen, rather 202: than GENERIC without Xen. 203: 204: Once the reboot is done, use `xl` to inspect Xen's boot messages, 205: available resources, and running domains. For example: 206: 207: [[!template id=programlisting text=""" 208: # xl dmesg 209: ... xen's boot info ... 210: # xl info 211: ... available memory, etc ... 212: # xl list 213: Name Id Mem(MB) CPU State Time(s) Console 214: Domain-0 0 64 0 r---- 58.1 215: """]] 216: 217: Xen logs will be in /var/log/xen. 218: 219: ### Issues with xencommons 220: 221: `xencommons` starts `xenstored`, which stores data on behalf of dom0 and 222: domUs. It does not currently work to stop and start xenstored. 223: Certainly all domUs should be shutdown first, following the sort order 224: of the rc.d scripts. However, the dom0 sets up state with xenstored, 225: and is not notified when xenstored exits, leading to not recreating 226: the state when the new xenstored starts. Until there's a mechanism to 227: make this work, one should not expect to be able to restart xenstored 228: (and thus xencommons). There is currently no reason to expect that 229: this will get fixed any time soon. 230: \todo Confirm if this is still true in 2020. 231: 232: ## Xen-specific NetBSD issues 233: 234: There are (at least) two additional things different about NetBSD as a 235: dom0 kernel compared to hardware. 236: 237: One is that the module ABI is different because some of the #defines 238: change, so one must build modules for Xen. As of netbsd-7, the build 239: system does this automatically. 240: 241: The other difference is that XEN3_DOM0 does not have exactly the same 242: options as GENERIC. While it is debatable whether or not this is a 243: bug, users should be aware of this and can simply add missing config 244: items if desired. 245: 246: Finally, there have been occasional reports of trouble with X11 247: servers in NetBSD as a dom0. 248: 249: ## Updating Xen in a dom0 250: 251: Basically, update the xenkernel and xentools packages and copy the new 252: Xen kernel into place, and reboot. This procedure should be usable to 253: update to a new Xen release, but the reader is reminded that having a 254: non-Xen boot methods was recommended earlier. 255: 256: ## Updating NetBSD in a dom0 257: 258: This is just like updating NetBSD on bare hardware, assuming the new 259: version supports the version of Xen you are running. Generally, one 260: replaces the kernel and reboots, and then overlays userland binaries 261: and adjusts `/etc`. 262: 263: Note that one must update both the non-Xen kernel typically used for 264: rescue purposes and the DOM0 kernel used with Xen. 265: 266: ## anita (for testing NetBSD) 267: 268: With a NetBSD dom0, even without any domUs, one should be able to run 269: anita (see pkgsrc/misc/py-anita) to test NetBSD releases, by doing (as 270: root, because anita must create a domU): 271: 272: [[!template id=programlisting text=""" 273: anita --vmm=xl test file:///usr/obj/i386/ 274: """]] 275: 276: # Unprivileged domains (domU) 277: 278: This section describes general concepts about domUs. It does not 279: address specific domU operating systems or how to install them. The 280: config files for domUs are typically in `/usr/pkg/etc/xen`, and are 281: typically named so that the file name, domU name and the domU's host 282: name match. 283: 284: The domU is provided with CPU and memory by Xen, configured by the 285: dom0. The domU is provided with disk and network by the dom0, 286: mediated by Xen, and configured in the dom0. 287: 288: Entropy in domUs can be an issue; physical disks and network are on 289: the dom0. NetBSD's /dev/random system works, but is often challenged. 290: 291: ## Config files 292: 293: See /usr/pkg/share/examples/xen/xlexample* 294: for a small number of well-commented examples, mostly for running 295: GNU/Linux. 296: 297: The following is an example minimal domain configuration file. The domU 298: serves as a network file server. 299: 300: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text=""" 301: name = "domU-id" 302: kernel = "/netbsd-XEN3PAE_DOMU-i386-foo.gz" 303: memory = 1024 304: vif = [ 'mac=aa:00:00:d1:00:09,bridge=bridge0' ] 305: disk = [ 'file:/n0/xen/foo-wd0,0x0,w', 306: 'file:/n0/xen/foo-wd1,0x1,w' ] 307: """]] 308: 309: The domain will have name given in the `name` setting. The kernel has the 310: host/domU name in it, so that on the dom0 one can update the various 311: domUs independently. The `vif` line causes an interface to be provided, 312: with a specific mac address (do not reuse MAC addresses!), in bridge 313: mode. Two disks are provided, and they are both writable; the bits 314: are stored in files and Xen attaches them to a vnd(4) device in the 315: dom0 on domain creation. The system treats xbd0 as the boot device 316: without needing explicit configuration. 317: 318: There is not type line; that implicitly defines a pv domU. 319: 320: By convention, domain config files are kept in `/usr/pkg/etc/xen`. Note 321: that "xl create" takes the name of a config file, while other commands 322: take the name of a domain. 323: 324: Examples of commands: 325: 326: [[!template id=programlisting text=""" 327: xl create /usr/pkg/etc/xen/foo 328: xl console domU-id 329: xl create -c /usr/pkg/etc/xen/foo 330: xl shutdown domU-id 331: xl list 332: """]] 333: 334: Typing `^]` will exit the console session. Shutting down a domain is 335: equivalent to pushing the power button; a NetBSD domU will receive a 336: power-press event and do a clean shutdown. Shutting down the dom0 337: will trigger controlled shutdowns of all configured domUs. 338: 339: ## CPU and memory 340: 341: A domain is provided with some number of vcpus, up to the number 342: of CPUs seen by the hypervisor. For a domU, it is controlled 343: from the config file by the "vcpus = N" directive. 344: 345: A domain is provided with memory; this is controlled in the config 346: file by "memory = N" (in megabytes). In the straightforward case, the 347: sum of the the memory allocated to the dom0 and all domUs must be less 348: than the available memory. 349: 350: Xen also provides a "balloon" driver, which can be used to let domains 351: use more memory temporarily. 352: 353: ## Virtual disks 354: 355: In domU config files, the disks are defined as a sequence of 3-tuples: 356: 357: * The first element is "method:/path/to/disk". Common methods are 358: "file:" for a file-backed vnd, and "phy:" for something that is already 359: a device, such as an LVM logical volume. 360: 361: * The second element is an artifact of how virtual disks are passed to 362: Linux, and a source of confusion with NetBSD Xen usage. Linux domUs 363: are given a device name to associate with the disk, and values like 364: "hda1" or "sda1" are common. In a NetBSD domU, the first disk appears 365: as xbd0, the second as xbd1, and so on. However, xl demands a 366: second argument. The name given is converted to a major/minor by 367: calling stat(2) on the name in /dev and this is passed to the domU. 368: In the general case, the dom0 and domU can be different operating 369: systems, and it is an unwarranted assumption that they have consistent 370: numbering in /dev, or even that the dom0 OS has a /dev. With NetBSD 371: as both dom0 and domU, using values of 0x0 for the first disk and 0x1 372: for the second works fine and avoids this issue. For a GNU/Linux 373: guest, one can create /dev/hda1 in /dev, or to pass 0x301 for 374: /dev/hda1. 375: 376: * The third element is "w" for writable disks, and "r" for read-only 377: disks. 378: 379: Example: 380: [[!template id=filecontent name="/usr/pkg/etc/xen/foo" text=""" 381: disk = [ 'file:/n0/xen/foo-wd0,0x0,w' ] 382: """]] 383: 384: Note that NetBSD by default creates only vnd[0123]. If you need more 385: than 4 total virtual disks at a time, run e.g. "./MAKEDEV vnd4" in the 386: dom0. 387: 388: ## Virtual Networking 389: 390: Xen provides virtual Ethernets, each of which connects the dom0 and a 391: domU. For each virtual network, there is an interface "xvifN.M" in 392: the dom0, and a matching interface xennetM (NetBSD name) in domU index N. 393: The interfaces behave as if there is an Ethernet with two 394: adapters connected. From this primitive, one can construct various 395: configurations. We focus on two common and useful cases for which 396: there are existing scripts: bridging and NAT. 397: 398: With bridging (in the example above), the domU perceives itself to be 399: on the same network as the dom0. For server virtualization, this is 400: usually best. Bridging is accomplished by creating a bridge(4) device 401: and adding the dom0's physical interface and the various xvifN.0 402: interfaces to the bridge. One specifies "bridge=bridge0" in the domU 403: config file. The bridge must be set up already in the dom0; an 404: example /etc/ifconfig.bridge0 is: 405: 406: [[!template id=filecontent name="/etc/ifconfig.bridge0" text=""" 407: create 408: up 409: !brconfig bridge0 add wm0 410: """]] 411: 412: With NAT, the domU perceives itself to be behind a NAT running on the 413: dom0. This is often appropriate when running Xen on a workstation. 414: TODO: NAT appears to be configured by "vif = [ '' ]". 415: 416: The MAC address specified is the one used for the interface in the new 417: domain. The interface in dom0 will use this address XOR'd with 418: 00:00:00:01:00:00. Random MAC addresses are assigned if not given. 419: 420: ## Starting domains automatically 421: 422: To start domains `domU-netbsd` and `domU-linux` at boot and shut them 423: down cleanly on dom0 shutdown, add the following in rc.conf: 424: 425: [[!template id=filecontent name="/etc/rc.conf" text=""" 426: xendomains="domU-netbsd domU-linux" 427: """]] 428: 429: # domU setup for specific systems 430: 431: Creating domUs is almost entirely independent of operating system. We 432: have already presented the basics of config files in the previous system. 433: 434: Of course, this section presumes that you have a working dom0. 435: 436: ## Creating a NetBSD PV domU 437: 438: See the earlier config file, and adjust memory. Decide on how much 439: storage you will provide, and prepare it (file or LVM). 440: 441: While the kernel will be obtained from the dom0 file system, the same 442: file should be present in the domU as /netbsd so that tools like 443: savecore(8) can work. (This is helpful but not necessary.) 444: 445: The kernel must be specifically built for Xen, to use PV interfacesas 446: a domU. NetBSD release builds provide the following kernels: 447: 448: i386 XEN3PAE_DOMU 449: amd64 XEN3_DOMU 450: 451: This will boot NetBSD, but this is not that useful if the disk is 452: empty. One approach is to unpack sets onto the disk outside of Xen 453: (by mounting it, just as you would prepare a physical disk for a 454: system you can't run the installer on). 455: 456: A second approach is to run an INSTALL kernel, which has a miniroot 457: and can load sets from the network. To do this, copy the INSTALL 458: kernel to / and change the kernel line in the config file to: 459: 460: kernel = "/home/bouyer/netbsd-INSTALL_XEN3_DOMU" 461: 462: Then, start the domain as "xl create -c configfile". 463: 464: Alternatively, if you want to install NetBSD/Xen with a CDROM image, the following 465: line should be used in the config file. 466: 467: disk = [ 'phy:/dev/wd0e,0x1,w', 'phy:/dev/cd0a,0x2,r' ] 468: 469: After booting the domain, the option to install via CDROM may be 470: selected. The CDROM device should be changed to `xbd1d`. 471: 472: Once done installing, "halt -p" the new domain (don't reboot or halt: 473: it would reload the INSTALL_XEN3_DOMU kernel even if you changed the 474: config file), switch the config file back to the XEN3_DOMU kernel, 475: and start the new domain again. Now it should be able to use "root on 476: xbd0a" and you should have a functional NetBSD domU. 477: 478: TODO: check if this is still accurate. 479: When the new domain is booting you'll see some warnings about *wscons* 480: and the pseudo-terminals. These can be fixed by editing the files 481: `/etc/ttys` and `/etc/wscons.conf`. You must disable all terminals in 482: `/etc/ttys`, except *console*, like this: 483: 484: console "/usr/libexec/getty Pc" vt100 on secure 485: ttyE0 "/usr/libexec/getty Pc" vt220 off secure 486: ttyE1 "/usr/libexec/getty Pc" vt220 off secure 487: ttyE2 "/usr/libexec/getty Pc" vt220 off secure 488: ttyE3 "/usr/libexec/getty Pc" vt220 off secure 489: 490: Finally, all screens must be commented out from `/etc/wscons.conf`. 491: 492: It is also desirable to add 493: 494: powerd=YES 495: 496: in rc.conf. This way, the domain will be properly shut down if 497: `xl shutdown -R` or `xl shutdown -H` is used on the dom0. 498: \todo Check the translation to xl. 499: 500: It is not strictly necessary to have a kernel (as /netbsd) in the domU 501: file system. However, various programs (e.g. netstat) will use that 502: kernel to look up symbols to read from kernel virtual memory. If 503: /netbsd is not the running kernel, those lookups will fail. (This is 504: not really a Xen-specific issue, but because the domU kernel is 505: obtained from the dom0, it is far more likely to be out of sync or 506: missing with Xen.) 507: 508: Note that NetBSD by default creates only xbd[0123]. If you need more 509: virtual disks in a domU, run e.g. "./MAKEDEV xbd4" in the domU. 510: 511: ## Creating a Linux domU 512: 513: Creating unprivileged Linux domains isn't much different from 514: unprivileged NetBSD domains, but there are some details to know. 515: 516: First, the second parameter passed to the disk declaration (the '0x1' in 517: the example below) 518: 519: disk = [ 'phy:/dev/wd0e,0x1,w' ] 520: 521: does matter to Linux. It wants a Linux device number here (e.g. 0x300 522: for hda). Linux builds device numbers as: (major \<\< 8 + minor). 523: So, hda1 which has major 3 and minor 1 on a Linux system will have 524: device number 0x301. Alternatively, devices names can be used (hda, 525: hdb, ...) as xentools has a table to map these names to devices 526: numbers. To export a partition to a Linux guest we can use: 527: 528: disk = [ 'phy:/dev/wd0e,0x300,w' ] 529: root = "/dev/hda1 ro" 530: 531: and it will appear as /dev/hda on the Linux system, and be used as root 532: partition. 533: 534: To install the Linux system on the partition to be exported to the 535: guest domain, the following method can be used: install 536: sysutils/e2fsprogs from pkgsrc. Use mke2fs to format the partition 537: that will be the root partition of your Linux domain, and mount it. 538: Then copy the files from a working Linux system, make adjustments in 539: `/etc` (fstab, network config). It should also be possible to extract 540: binary packages such as .rpm or .deb directly to the mounted partition 541: using the appropriate tool, possibly running under NetBSD's Linux 542: emulation. Once the file system has been populated, umount it. If 543: desirable, the file system can be converted to ext3 using tune2fs -j. 544: It should now be possible to boot the Linux guest domain, using one of 545: the vmlinuz-\*-xenU kernels available in the Xen binary distribution. 546: 547: To get the Linux console right, you need to add: 548: 549: extra = "xencons=tty1" 550: 551: to your configuration since not all Linux distributions auto-attach a 552: tty to the xen console. 553: 554: ## Creating a NetBSD HVM domU 555: 556: Use type='hmv', probably. Use a GENERIC kernel within the disk image. 557: 558: ## Creating a NetBSD PVH domU 559: 560: Use type='pvh'. 561: 562: \todo Explain where the kernel comes from. 563: 564: 565: ## Creating a Solaris domU 566: 567: See possibly outdated 568: [Solaris domU instructions](/ports/xen/howto-solaris/). 569: 570: 571: ## PCI passthrough: Using PCI devices in guest domains 572: 573: NB: PCI passthrough only works on some Xen versions and as of 2020 it 574: is not clear that it works on any version in pkgsrc. Reports 575: confirming or denying this notion should be sent to port-xen@. 576: 577: The dom0 can give other domains access to selected PCI 578: devices. This can allow, for example, a non-privileged domain to have 579: access to a physical network interface or disk controller. However, 580: keep in mind that giving a domain access to a PCI device most likely 581: will give the domain read/write access to the whole physical memory, 582: as PCs don't have an IOMMU to restrict memory access to DMA-capable 583: device. Also, it's not possible to export ISA devices to non-dom0 584: domains, which means that the primary VGA adapter can't be exported. 585: A guest domain trying to access the VGA registers will panic. 586: 587: If the dom0 is NetBSD, it has to be running Xen 3.1, as support has 588: not been ported to later versions at this time. 589: 590: For a PCI device to be exported to a domU, is has to be attached to 591: the "pciback" driver in dom0. Devices passed to the dom0 via the 592: pciback.hide boot parameter will attach to "pciback" instead of the 593: usual driver. The list of devices is specified as "(bus:dev.func)", 594: where bus and dev are 2-digit hexadecimal numbers, and func a 595: single-digit number: 596: 597: pciback.hide=(00:0a.0)(00:06.0) 598: 599: pciback devices should show up in the dom0's boot messages, and the 600: devices should be listed in the `/kern/xen/pci` directory. 601: 602: PCI devices to be exported to a domU are listed in the "pci" array of 603: the domU's config file, with the format "0000:bus:dev.func". 604: 605: pci = [ '0000:00:06.0', '0000:00:0a.0' ] 606: 607: In the domU an "xpci" device will show up, to which one or more pci 608: buses will attach. Then the PCI drivers will attach to PCI buses as 609: usual. Note that the default NetBSD DOMU kernels do not have "xpci" 610: or any PCI drivers built in by default; you have to build your own 611: kernel to use PCI devices in a domU. Here's a kernel config example; 612: note that only the "xpci" lines are unusual. 613: 614: include "arch/i386/conf/XEN3_DOMU" 615: 616: # Add support for PCI buses to the XEN3_DOMU kernel 617: xpci* at xenbus ? 618: pci* at xpci ? 619: 620: # PCI USB controllers 621: uhci* at pci? dev ? function ? # Universal Host Controller (Intel) 622: 623: # USB bus support 624: usb* at uhci? 625: 626: # USB Hubs 627: uhub* at usb? 628: uhub* at uhub? port ? configuration ? interface ? 629: 630: # USB Mass Storage 631: umass* at uhub? port ? configuration ? interface ? 632: wd* at umass? 633: # SCSI controllers 634: ahc* at pci? dev ? function ? # Adaptec [23]94x, aic78x0 SCSI 635: 636: # SCSI bus support (for both ahc and umass) 637: scsibus* at scsi? 638: 639: # SCSI devices 640: sd* at scsibus? target ? lun ? # SCSI disk drives 641: cd* at scsibus? target ? lun ? # SCSI CD-ROM drives 642: 643: 644: # Miscellaneous Information 645: 646: ## Nesting under Linux KVM 647: 648: It is possible to run a Xen and a NetBSD dom0 under Linux KVM. One 649: can enable virtio in the dom0 for greater speed. 650: 651: ## Other nesting 652: 653: In theory, any full emulation should be able to run Xen and a NetBSD 654: dom0. The HOWTO does not currently have information about Xen XVM 655: mode, nvmm, qemu, Virtualbox, etc. 656: 657: ## NetBSD 5 as domU 658: 659: [NetBSD 5 is known to panic.](http://mail-index.netbsd.org/port-xen/2018/04/17/msg009181.html) 660: (However, NetBSD 5 systems should be updated to a supported version.) 661: 662: # NetBSD as a domU in a VPS 663: 664: The bulk of the HOWTO is about using NetBSD as a dom0 on your own 665: hardware. This section explains how to deal with Xen in a domU as a 666: virtual private server where you do not control or have access to the 667: dom0. This is not intended to be an exhaustive list of VPS providers; 668: only a few are mentioned that specifically support NetBSD. 669: 670: VPS operators provide varying degrees of access and mechanisms for 671: configuration. The big issue is usually how one controls which kernel 672: is booted, because the kernel is nominally in the dom0 file system (to 673: which VPS users do not normally have access). A second issue is how 674: to install NetBSD. 675: A VPS user may want to compile a kernel for security updates, to run 676: npf, run IPsec, or any other reason why someone would want to change 677: their kernel. 678: 679: One approach is to have an administrative interface to upload a kernel, 680: or to select from a prepopulated list. Other approaches are pygrub 681: (deprecated) and pvgrub, which are ways to have a bootloader obtain a 682: kernel from the domU file system. This is closer to a regular physical 683: computer, where someone who controls a machine can replace the kernel. 684: 685: A second issue is multiple CPUs. With NetBSD 6, domUs support 686: multiple vcpus, and it is typical for VPS providers to enable multiple 687: CPUs for NetBSD domUs. 688: 689: ## Complexities due to Xen changes 690: 691: Xen has many security advisories and people running Xen systems make 692: different choices. 693: 694: ### stub domains 695: 696: Some (Linux only?) dom0 systems use something called "stub domains" to 697: isolate qemu from the dom0 system, as a security and reliabilty 698: mechanism when running HVM domUs. Somehow, NetBSD's GENERIC kernel 699: ends up using PIO for disks rather than DMA. Of course, all of this 700: is emulated, but emulated PIO is unusably slow. This problem is not 701: currently understood. 702: 703: ### Grant tables 704: 705: There are multiple versions of using grant tables, and some security 706: advisories have suggested disabling some versions. Some versions of 707: NetBSD apparently only use specific versions and this can lead to 708: "NetBSD current doesn't run on hosting provider X" situations. 709: 710: \todo Explain better. 711: 712: ## Boot methods 713: 714: ### pvgrub 715: 716: pvgrub is a version of grub that uses PV operations instead of BIOS 717: calls. It is booted from the dom0 as the domU kernel, and then reads 718: /grub/menu.lst and loads a kernel from the domU file system. 719: 720: [Panix](http://www.panix.com/) lets users use pvgrub. Panix reports 721: that pvgrub works with FFsv2 with 16K/2K and 32K/4K block/frag sizes 722: (and hence with defaults from "newfs -O 2"). See [Panix's pvgrub 723: page](http://www.panix.com/v-colo/grub.html), which describes only 724: Linux but should be updated to cover NetBSD :-). 725: 726: [prgmr.com](http://prgmr.com/) also lets users with pvgrub to boot 727: their own kernel. See then [prgmr.com NetBSD 728: HOWTO](http://wiki.prgmr.com/mediawiki/index.php/NetBSD_as_a_DomU) 729: (which is in need of updating). 730: 731: It appears that [grub's FFS 732: code](http://xenbits.xensource.com/hg/xen-unstable.hg/file/bca284f67702/tools/libfsimage/ufs/fsys_ufs.c) 733: does not support all aspects of modern FFS, but there are also reports 734: that FFSv2 works fine. At prgmr, typically one has an ext2 or FAT 735: partition for the kernel with the intent that grub can understand it, 736: which leads to /netbsd not being the actual kernel. One must remember 737: to update the special boot partition. 738: 739: ### pygrub 740: 741: pygrub runs in the dom0 and looks into the domU file system. This 742: implies that the domU must have a kernel in a file system in a format 743: known to pygrub. 744: 745: pygrub doesn't seem to work to load Linux images under NetBSD dom0, 746: and is inherently less secure than pvgrub due to running inside dom0. For both these 747: reasons, pygrub should not be used, and is only still present so that 748: historical DomU images using it still work. 749: 750: As of 2014, pygrub seems to be of mostly historical 751: interest. New DomUs should use pvgrub. 752: 753: ## Specific Providers 754: 755: ### Amazon 756: 757: See the [Amazon EC2 page](/amazon_ec2/).