5 guestfs - Library for accessing and modifying virtual machine images
11 guestfs_h *g = guestfs_create ();
12 guestfs_add_drive (g, "guest.img");
14 guestfs_mount (g, "/dev/sda1", "/");
15 guestfs_touch (g, "/hello");
16 guestfs_umount (g, "/");
19 cc prog.c -o prog -lguestfs
21 cc prog.c -o prog `pkg-config libguestfs --cflags --libs`
25 Libguestfs is a library for accessing and modifying guest disk images.
26 Amongst the things this is good for: making batch configuration
27 changes to guests, getting disk used/free statistics (see also:
28 virt-df), migrating between virtualization systems (see also:
29 virt-p2v), performing partial backups, performing partial guest
30 clones, cloning guests and changing registry/UUID/hostname info, and
33 Libguestfs uses Linux kernel and qemu code, and can access any type of
34 guest filesystem that Linux and qemu can, including but not limited
35 to: ext2/3/4, btrfs, FAT and NTFS, LVM, many different disk partition
36 schemes, qcow, qcow2, vmdk.
38 Libguestfs provides ways to enumerate guest storage (eg. partitions,
39 LVs, what filesystem is in each LV, etc.). It can also run commands
40 in the context of the guest. Also you can access filesystems over
43 Libguestfs is a library that can be linked with C and C++ management
44 programs (or management programs written in OCaml, Perl, Python, Ruby,
45 Java, PHP, Haskell or C#). You can also use it from shell scripts or the
48 You don't need to be root to use libguestfs, although obviously you do
49 need enough permissions to access the disk images.
51 Libguestfs is a large API because it can do many things. For a gentle
52 introduction, please read the L</API OVERVIEW> section next.
54 There are also some example programs in the L<guestfs-examples(3)>
59 This section provides a gentler overview of the libguestfs API. We
60 also try to group API calls together, where that may not be obvious
61 from reading about the individual calls in the main section of this
66 Before you can use libguestfs calls, you have to create a handle.
67 Then you must add at least one disk image to the handle, followed by
68 launching the handle, then performing whatever operations you want,
69 and finally closing the handle. By convention we use the single
70 letter C<g> for the name of the handle variable, although of course
71 you can use any name you want.
73 The general structure of all libguestfs-using programs looks like
76 guestfs_h *g = guestfs_create ();
78 /* Call guestfs_add_drive additional times if there are
79 * multiple disk images.
81 guestfs_add_drive (g, "guest.img");
83 /* Most manipulation calls won't work until you've launched
84 * the handle 'g'. You have to do this _after_ adding drives
85 * and _before_ other commands.
89 /* Now you can examine what partitions, LVs etc are available.
91 char **partitions = guestfs_list_partitions (g);
92 char **logvols = guestfs_lvs (g);
94 /* To access a filesystem in the image, you must mount it.
96 guestfs_mount (g, "/dev/sda1", "/");
98 /* Now you can perform filesystem actions on the guest
101 guestfs_touch (g, "/hello");
103 /* This is only needed for libguestfs < 1.5.24. Since then
104 * it is done automatically when you close the handle. See
105 * discussion of autosync in this page.
109 /* Close the handle 'g'. */
112 The code above doesn't include any error checking. In real code you
113 should check return values carefully for errors. In general all
114 functions that return integers return C<-1> on error, and all
115 functions that return pointers return C<NULL> on error. See section
116 L</ERROR HANDLING> below for how to handle errors, and consult the
117 documentation for each function call below to see precisely how they
118 return error indications. See L<guestfs-examples(3)> for fully worked
123 The image filename (C<"guest.img"> in the example above) could be a
124 disk image from a virtual machine, a L<dd(1)> copy of a physical hard
125 disk, an actual block device, or simply an empty file of zeroes that
126 you have created through L<posix_fallocate(3)>. Libguestfs lets you
127 do useful things to all of these.
129 The call you should use in modern code for adding drives is
130 L</guestfs_add_drive_opts>. To add a disk image, allowing writes, and
131 specifying that the format is raw, do:
133 guestfs_add_drive_opts (g, filename,
134 GUESTFS_ADD_DRIVE_OPTS_FORMAT, "raw",
137 You can add a disk read-only using:
139 guestfs_add_drive_opts (g, filename,
140 GUESTFS_ADD_DRIVE_OPTS_FORMAT, "raw",
141 GUESTFS_ADD_DRIVE_OPTS_READONLY, 1,
144 or by calling the older function L</guestfs_add_drive_ro>. In either
145 case libguestfs won't modify the file.
147 Be extremely cautious if the disk image is in use, eg. if it is being
148 used by a virtual machine. Adding it read-write will almost certainly
149 cause disk corruption, but adding it read-only is safe.
151 You must add at least one disk image, and you may add multiple disk
152 images. In the API, the disk images are usually referred to as
153 C</dev/sda> (for the first one you added), C</dev/sdb> (for the second
156 Once L</guestfs_launch> has been called you cannot add any more images.
157 You can call L</guestfs_list_devices> to get a list of the device
158 names, in the order that you added them. See also L</BLOCK DEVICE
163 Before you can read or write files, create directories and so on in a
164 disk image that contains filesystems, you have to mount those
165 filesystems using L</guestfs_mount>. If you already know that a disk
166 image contains (for example) one partition with a filesystem on that
167 partition, then you can mount it directly:
169 guestfs_mount (g, "/dev/sda1", "/");
171 where C</dev/sda1> means literally the first partition (C<1>) of the
172 first disk image that we added (C</dev/sda>). If the disk contains
173 Linux LVM2 logical volumes you could refer to those instead (eg. C</dev/VG/LV>).
175 If you are given a disk image and you don't know what it contains then
176 you have to find out. Libguestfs can do that too: use
177 L</guestfs_list_partitions> and L</guestfs_lvs> to list possible
178 partitions and LVs, and either try mounting each to see what is
179 mountable, or else examine them with L</guestfs_vfs_type> or
180 L</guestfs_file>. Libguestfs also has a set of APIs for inspection of
181 disk images (see L</INSPECTION> below). But you might find it easier
182 to look at higher level programs built on top of libguestfs, in
183 particular L<virt-inspector(1)>.
185 To mount a disk image read-only, use L</guestfs_mount_ro>. There are
186 several other variations of the C<guestfs_mount_*> call.
188 =head2 FILESYSTEM ACCESS AND MODIFICATION
190 The majority of the libguestfs API consists of fairly low-level calls
191 for accessing and modifying the files, directories, symlinks etc on
192 mounted filesystems. There are over a hundred such calls which you
193 can find listed in detail below in this man page, and we don't even
194 pretend to cover them all in this overview.
196 Specify filenames as full paths, starting with C<"/"> and including
199 For example, if you mounted a filesystem at C<"/"> and you want to
200 read the file called C<"etc/passwd"> then you could do:
202 char *data = guestfs_cat (g, "/etc/passwd");
204 This would return C<data> as a newly allocated buffer containing the
205 full content of that file (with some conditions: see also
206 L</DOWNLOADING> below), or C<NULL> if there was an error.
208 As another example, to create a top-level directory on that filesystem
209 called C<"var"> you would do:
211 guestfs_mkdir (g, "/var");
213 To create a symlink you could do:
215 guestfs_ln_s (g, "/etc/init.d/portmap",
216 "/etc/rc3.d/S30portmap");
218 Libguestfs will reject attempts to use relative paths and there is no
219 concept of a current working directory.
221 Libguestfs can return errors in many situations: for example if the
222 filesystem isn't writable, or if a file or directory that you
223 requested doesn't exist. If you are using the C API (documented here)
224 you have to check for those error conditions after each call. (Other
225 language bindings turn these errors into exceptions).
227 File writes are affected by the per-handle umask, set by calling
228 L</guestfs_umask> and defaulting to 022. See L</UMASK>.
232 Libguestfs contains API calls to read, create and modify partition
233 tables on disk images.
235 In the common case where you want to create a single partition
236 covering the whole disk, you should use the L</guestfs_part_disk>
239 const char *parttype = "mbr";
240 if (disk_is_larger_than_2TB)
242 guestfs_part_disk (g, "/dev/sda", parttype);
244 Obviously this effectively wipes anything that was on that disk image
249 Libguestfs provides access to a large part of the LVM2 API, such as
250 L</guestfs_lvcreate> and L</guestfs_vgremove>. It won't make much sense
251 unless you familiarize yourself with the concepts of physical volumes,
252 volume groups and logical volumes.
254 This author strongly recommends reading the LVM HOWTO, online at
255 L<http://tldp.org/HOWTO/LVM-HOWTO/>.
259 Use L</guestfs_cat> to download small, text only files. This call
260 is limited to files which are less than 2 MB and which cannot contain
261 any ASCII NUL (C<\0>) characters. However it has a very simple
264 L</guestfs_read_file> can be used to read files which contain
265 arbitrary 8 bit data, since it returns a (pointer, size) pair.
266 However it is still limited to "small" files, less than 2 MB.
268 L</guestfs_download> can be used to download any file, with no
269 limits on content or size (even files larger than 4 GB).
271 To download multiple files, see L</guestfs_tar_out> and
276 It's often the case that you want to write a file or files to the disk
279 To write a small file with fixed content, use L</guestfs_write>. To
280 create a file of all zeroes, use L</guestfs_truncate_size> (sparse) or
281 L</guestfs_fallocate64> (with all disk blocks allocated). There are a
282 variety of other functions for creating test files, for example
283 L</guestfs_fill> and L</guestfs_fill_pattern>.
285 To upload a single file, use L</guestfs_upload>. This call has no
286 limits on file content or size (even files larger than 4 GB).
288 To upload multiple files, see L</guestfs_tar_in> and L</guestfs_tgz_in>.
290 However the fastest way to upload I<large numbers of arbitrary files>
291 is to turn them into a squashfs or CD ISO (see L<mksquashfs(8)> and
292 L<mkisofs(8)>), then attach this using L</guestfs_add_drive_ro>. If
293 you add the drive in a predictable way (eg. adding it last after all
294 other drives) then you can get the device name from
295 L</guestfs_list_devices> and mount it directly using
296 L</guestfs_mount_ro>. Note that squashfs images are sometimes
297 non-portable between kernel versions, and they don't support labels or
298 UUIDs. If you want to pre-build an image or you need to mount it
299 using a label or UUID, use an ISO image instead.
303 There are various different commands for copying between files and
304 devices and in and out of the guest filesystem. These are summarised
309 =item B<file> to B<file>
311 Use L</guestfs_cp> to copy a single file, or
312 L</guestfs_cp_a> to copy directories recursively.
314 =item B<file or device> to B<file or device>
316 Use L</guestfs_dd> which efficiently uses L<dd(1)>
317 to copy between files and devices in the guest.
319 Example: duplicate the contents of an LV:
321 guestfs_dd (g, "/dev/VG/Original", "/dev/VG/Copy");
323 The destination (C</dev/VG/Copy>) must be at least as large as the
324 source (C</dev/VG/Original>). To copy less than the whole
325 source device, use L</guestfs_copy_size>.
327 =item B<file on the host> to B<file or device>
329 Use L</guestfs_upload>. See L</UPLOADING> above.
331 =item B<file or device> to B<file on the host>
333 Use L</guestfs_download>. See L</DOWNLOADING> above.
337 =head2 UPLOADING AND DOWNLOADING TO PIPES AND FILE DESCRIPTORS
339 Calls like L</guestfs_upload>, L</guestfs_download>,
340 L</guestfs_tar_in>, L</guestfs_tar_out> etc appear to only take
341 filenames as arguments, so it appears you can only upload and download
342 to files. However many Un*x-like hosts let you use the special device
343 files C</dev/stdin>, C</dev/stdout>, C</dev/stderr> and C</dev/fd/N>
344 to read and write from stdin, stdout, stderr, and arbitrary file
347 For example, L<virt-cat(1)> writes its output to stdout by
350 guestfs_download (filename, "/dev/stdout");
352 and you can write tar output to a pipe C<fd> by doing:
355 snprintf (devfd, sizeof devfd, "/dev/fd/%d", fd);
356 guestfs_tar_out ("/", devfd);
360 L</guestfs_ll> is just designed for humans to read (mainly when using
361 the L<guestfish(1)>-equivalent command C<ll>).
363 L</guestfs_ls> is a quick way to get a list of files in a directory
364 from programs, as a flat list of strings.
366 L</guestfs_readdir> is a programmatic way to get a list of files in a
367 directory, plus additional information about each one. It is more
368 equivalent to using the L<readdir(3)> call on a local filesystem.
370 L</guestfs_find> and L</guestfs_find0> can be used to recursively list
373 =head2 RUNNING COMMANDS
375 Although libguestfs is primarily an API for manipulating files
376 inside guest images, we also provide some limited facilities for
377 running commands inside guests.
379 There are many limitations to this:
385 The kernel version that the command runs under will be different
386 from what it expects.
390 If the command needs to communicate with daemons, then most likely
391 they won't be running.
395 The command will be running in limited memory.
399 The network may not be available unless you enable it
400 (see L</guestfs_set_network>).
404 Only supports Linux guests (not Windows, BSD, etc).
408 Architecture limitations (eg. won't work for a PPC guest on
413 For SELinux guests, you may need to enable SELinux and load policy
414 first. See L</SELINUX> in this manpage.
418 I<Security:> It is not safe to run commands from untrusted, possibly
419 malicious guests. These commands may attempt to exploit your program
420 by sending unexpected output. They could also try to exploit the
421 Linux kernel or qemu provided by the libguestfs appliance. They could
422 use the network provided by the libguestfs appliance to bypass
423 ordinary network partitions and firewalls. They could use the
424 elevated privileges or different SELinux context of your program
427 A secure alternative is to use libguestfs to install a "firstboot"
428 script (a script which runs when the guest next boots normally), and
429 to have this script run the commands you want in the normal context of
430 the running guest, network security and so on. For information about
431 other security issues, see L</SECURITY>.
435 The two main API calls to run commands are L</guestfs_command> and
436 L</guestfs_sh> (there are also variations).
438 The difference is that L</guestfs_sh> runs commands using the shell, so
439 any shell globs, redirections, etc will work.
441 =head2 CONFIGURATION FILES
443 To read and write configuration files in Linux guest filesystems, we
444 strongly recommend using Augeas. For example, Augeas understands how
445 to read and write, say, a Linux shadow password file or X.org
446 configuration file, and so avoids you having to write that code.
448 The main Augeas calls are bound through the C<guestfs_aug_*> APIs. We
449 don't document Augeas itself here because there is excellent
450 documentation on the L<http://augeas.net/> website.
452 If you don't want to use Augeas (you fool!) then try calling
453 L</guestfs_read_lines> to get the file as a list of lines which
454 you can iterate over.
458 We support SELinux guests. To ensure that labeling happens correctly
459 in SELinux guests, you need to enable SELinux and load the guest's
466 Before launching, do:
468 guestfs_set_selinux (g, 1);
472 After mounting the guest's filesystem(s), load the policy. This
473 is best done by running the L<load_policy(8)> command in the
476 guestfs_sh (g, "/usr/sbin/load_policy");
478 (Older versions of C<load_policy> require you to specify the
479 name of the policy file).
483 Optionally, set the security context for the API. The correct
484 security context to use can only be known by inspecting the
485 guest. As an example:
487 guestfs_setcon (g, "unconfined_u:unconfined_r:unconfined_t:s0");
491 This will work for running commands and editing existing files.
493 When new files are created, you may need to label them explicitly,
494 for example by running the external command
495 C<restorecon pathname>.
499 Certain calls are affected by the current file mode creation mask (the
500 "umask"). In particular ones which create files or directories, such
501 as L</guestfs_touch>, L</guestfs_mknod> or L</guestfs_mkdir>. This
502 affects either the default mode that the file is created with or
503 modifies the mode that you supply.
505 The default umask is C<022>, so files are created with modes such as
506 C<0644> and directories with C<0755>.
508 There are two ways to avoid being affected by umask. Either set umask
509 to 0 (call C<guestfs_umask (g, 0)> early after launching). Or call
510 L</guestfs_chmod> after creating each file or directory.
512 For more information about umask, see L<umask(2)>.
514 =head2 ENCRYPTED DISKS
516 Libguestfs allows you to access Linux guests which have been
517 encrypted using whole disk encryption that conforms to the
518 Linux Unified Key Setup (LUKS) standard. This includes
519 nearly all whole disk encryption systems used by modern
522 Use L</guestfs_vfs_type> to identify LUKS-encrypted block
523 devices (it returns the string C<crypto_LUKS>).
525 Then open these devices by calling L</guestfs_luks_open>.
526 Obviously you will require the passphrase!
528 Opening a LUKS device creates a new device mapper device
529 called C</dev/mapper/mapname> (where C<mapname> is the
530 string you supply to L</guestfs_luks_open>).
531 Reads and writes to this mapper device are decrypted from and
532 encrypted to the underlying block device respectively.
534 LVM volume groups on the device can be made visible by calling
535 L</guestfs_vgscan> followed by L</guestfs_vg_activate_all>.
536 The logical volume(s) can now be mounted in the usual way.
538 Use the reverse process to close a LUKS device. Unmount
539 any logical volumes on it, deactivate the volume groups
540 by caling C<guestfs_vg_activate (g, 0, ["/dev/VG"])>.
541 Then close the mapper device by calling
542 L</guestfs_luks_close> on the C</dev/mapper/mapname>
543 device (I<not> the underlying encrypted block device).
547 Libguestfs has APIs for inspecting an unknown disk image to find out
548 if it contains operating systems. (These APIs used to be in a
549 separate Perl-only library called L<Sys::Guestfs::Lib(3)> but since
550 version 1.5.3 the most frequently used part of this library has been
551 rewritten in C and moved into the core code).
553 Add all disks belonging to the unknown virtual machine and call
554 L</guestfs_launch> in the usual way.
556 Then call L</guestfs_inspect_os>. This function uses other libguestfs
557 calls and certain heuristics, and returns a list of operating systems
558 that were found. An empty list means none were found. A single
559 element is the root filesystem of the operating system. For dual- or
560 multi-boot guests, multiple roots can be returned, each one
561 corresponding to a separate operating system. (Multi-boot virtual
562 machines are extremely rare in the world of virtualization, but since
563 this scenario can happen, we have built libguestfs to deal with it.)
565 For each root, you can then call various C<guestfs_inspect_get_*>
566 functions to get additional details about that operating system. For
567 example, call L</guestfs_inspect_get_type> to return the string
568 C<windows> or C<linux> for Windows and Linux-based operating systems
571 Un*x-like and Linux-based operating systems usually consist of several
572 filesystems which are mounted at boot time (for example, a separate
573 boot partition mounted on C</boot>). The inspection rules are able to
574 detect how filesystems correspond to mount points. Call
575 C<guestfs_inspect_get_mountpoints> to get this mapping. It might
576 return a hash table like this example:
579 / => /dev/vg_guest/lv_root
580 /usr => /dev/vg_guest/lv_usr
582 The caller can then make calls to L</guestfs_mount_options> to
583 mount the filesystems as suggested.
585 Be careful to mount filesystems in the right order (eg. C</> before
586 C</usr>). Sorting the keys of the hash by length, shortest first,
589 Inspection currently only works for some common operating systems.
590 Contributors are welcome to send patches for other operating systems
591 that we currently cannot detect.
593 Encrypted disks must be opened before inspection. See
594 L</ENCRYPTED DISKS> for more details. The L</guestfs_inspect_os>
595 function just ignores any encrypted devices.
597 A note on the implementation: The call L</guestfs_inspect_os> performs
598 inspection and caches the results in the guest handle. Subsequent
599 calls to C<guestfs_inspect_get_*> return this cached information, but
600 I<do not> re-read the disks. If you change the content of the guest
601 disks, you can redo inspection by calling L</guestfs_inspect_os>
602 again. (L</guestfs_inspect_list_applications> works a little
603 differently from the other calls and does read the disks. See
604 documentation for that function for details).
606 =head2 SPECIAL CONSIDERATIONS FOR WINDOWS GUESTS
608 Libguestfs can mount NTFS partitions. It does this using the
609 L<http://www.ntfs-3g.org/> driver.
611 =head3 DRIVE LETTERS AND PATHS
613 DOS and Windows still use drive letters, and the filesystems are
614 always treated as case insensitive by Windows itself, and therefore
615 you might find a Windows configuration file referring to a path like
616 C<c:\windows\system32>. When the filesystem is mounted in libguestfs,
617 that directory might be referred to as C</WINDOWS/System32>.
619 Drive letter mappings are outside the scope of libguestfs. You have
620 to use libguestfs to read the appropriate Windows Registry and
621 configuration files, to determine yourself how drives are mapped (see
622 also L<hivex(3)> and L<virt-inspector(1)>).
624 Replacing backslash characters with forward slash characters is also
625 outside the scope of libguestfs, but something that you can easily do.
627 Where we can help is in resolving the case insensitivity of paths.
628 For this, call L</guestfs_case_sensitive_path>.
630 =head3 ACCESSING THE WINDOWS REGISTRY
632 Libguestfs also provides some help for decoding Windows Registry
633 "hive" files, through the library C<hivex> which is part of the
634 libguestfs project although ships as a separate tarball. You have to
635 locate and download the hive file(s) yourself, and then pass them to
636 C<hivex> functions. See also the programs L<hivexml(1)>,
637 L<hivexsh(1)>, L<hivexregedit(1)> and L<virt-win-reg(1)> for more help
640 =head3 SYMLINKS ON NTFS-3G FILESYSTEMS
642 Ntfs-3g tries to rewrite "Junction Points" and NTFS "symbolic links"
643 to provide something which looks like a Linux symlink. The way it
644 tries to do the rewriting is described here:
646 L<http://www.tuxera.com/community/ntfs-3g-advanced/junction-points-and-symbolic-links/>
648 The essential problem is that ntfs-3g simply does not have enough
649 information to do a correct job. NTFS links can contain drive letters
650 and references to external device GUIDs that ntfs-3g has no way of
651 resolving. It is almost certainly the case that libguestfs callers
652 should ignore what ntfs-3g does (ie. don't use L</guestfs_readlink> on
655 Instead if you encounter a symbolic link on an ntfs-3g filesystem, use
656 L</guestfs_lgetxattr> to read the C<system.ntfs_reparse_data> extended
657 attribute, and read the raw reparse data from that (you can find the
658 format documented in various places around the web).
660 =head3 EXTENDED ATTRIBUTES ON NTFS-3G FILESYSTEMS
662 There are other useful extended attributes that can be read from
663 ntfs-3g filesystems (using L</guestfs_getxattr>). See:
665 L<http://www.tuxera.com/community/ntfs-3g-advanced/extended-attributes/>
667 =head2 USING LIBGUESTFS WITH OTHER PROGRAMMING LANGUAGES
669 Although we don't want to discourage you from using the C API, we will
670 mention here that the same API is also available in other languages.
672 The API is broadly identical in all supported languages. This means
673 that the C call C<guestfs_mount(g,path)> is
674 C<$g-E<gt>mount($path)> in Perl, C<g.mount(path)> in Python,
675 and C<Guestfs.mount g path> in OCaml. In other words, a
676 straightforward, predictable isomorphism between each language.
678 Error messages are automatically transformed
679 into exceptions if the language supports it.
681 We don't try to "object orientify" parts of the API in OO languages,
682 although contributors are welcome to write higher level APIs above
683 what we provide in their favourite languages if they wish.
689 You can use the I<guestfs.h> header file from C++ programs. The C++
690 API is identical to the C API. C++ classes and exceptions are not
695 The C# bindings are highly experimental. Please read the warnings
696 at the top of C<csharp/Libguestfs.cs>.
700 This is the only language binding that is working but incomplete.
701 Only calls which return simple integers have been bound in Haskell,
702 and we are looking for help to complete this binding.
706 Full documentation is contained in the Javadoc which is distributed
711 For documentation see L<guestfs-ocaml(3)>.
715 For documentation see L<Sys::Guestfs(3)>.
719 For documentation see C<README-PHP> supplied with libguestfs
720 sources or in the php-libguestfs package for your distribution.
722 The PHP binding only works correctly on 64 bit machines.
726 For documentation see L<guestfs-python(3)>.
730 For documentation see L<guestfs-ruby(3)>.
732 =item B<shell scripts>
734 For documentation see L<guestfish(1)>.
738 =head2 LIBGUESTFS GOTCHAS
740 L<http://en.wikipedia.org/wiki/Gotcha_(programming)>: "A feature of a
741 system [...] that works in the way it is documented but is
742 counterintuitive and almost invites mistakes."
744 Since we developed libguestfs and the associated tools, there are
745 several things we would have designed differently, but are now stuck
746 with for backwards compatibility or other reasons. If there is ever a
747 libguestfs 2.0 release, you can expect these to change. Beware of
752 =item Autosync / forgetting to sync.
754 When modifying a filesystem from C or another language, you B<must>
755 unmount all filesystems and call L</guestfs_sync> explicitly before
756 you close the libguestfs handle. You can also call:
758 guestfs_set_autosync (g, 1);
760 to have the unmount/sync done automatically for you when the handle 'g'
761 is closed. (This feature is called "autosync", L</guestfs_set_autosync>
764 If you forget to do this, then it is entirely possible that your
765 changes won't be written out, or will be partially written, or (very
766 rarely) that you'll get disk corruption.
768 Note that in L<guestfish(3)> autosync is the default. So quick and
769 dirty guestfish scripts that forget to sync will work just fine, which
770 can make this very puzzling if you are trying to debug a problem.
772 Update: Autosync is enabled by default for all API users starting from
775 =item Mount option C<-o sync> should not be the default.
777 If you use L</guestfs_mount>, then C<-o sync,noatime> are added
778 implicitly. However C<-o sync> does not add any reliability benefit,
779 but does have a very large performance impact.
781 The work around is to use L</guestfs_mount_options> and set the mount
782 options that you actually want to use.
784 =item Read-only should be the default.
786 In L<guestfish(3)>, I<--ro> should be the default, and you should
787 have to specify I<--rw> if you want to make changes to the image.
789 This would reduce the potential to corrupt live VM images.
791 Note that many filesystems change the disk when you just mount and
792 unmount, even if you didn't perform any writes. You need to use
793 L</guestfs_add_drive_ro> to guarantee that the disk is not changed.
795 =item guestfish command line is hard to use.
797 C<guestfish disk.img> doesn't do what people expect (open C<disk.img>
798 for examination). It tries to run a guestfish command C<disk.img>
799 which doesn't exist, so it fails. In earlier versions of guestfish
800 the error message was also unintuitive, but we have corrected this
801 since. Like the Bourne shell, we should have used C<guestfish -c
802 command> to run commands.
804 =item guestfish megabyte modifiers don't work right on all commands
806 In recent guestfish you can use C<1M> to mean 1 megabyte (and
807 similarly for other modifiers). What guestfish actually does is to
808 multiply the number part by the modifier part and pass the result to
809 the C API. However this doesn't work for a few APIs which aren't
810 expecting bytes, but are already expecting some other unit
813 The most common is L</guestfs_lvcreate>. The guestfish command:
817 does not do what you might expect. Instead because
818 L</guestfs_lvcreate> is already expecting megabytes, this tries to
819 create a 100 I<terabyte> (100 megabytes * megabytes) logical volume.
820 The error message you get from this is also a little obscure.
822 This could be fixed in the generator by specially marking parameters
823 and return values which take bytes or other units.
825 =item Ambiguity between devices and paths
827 There is a subtle ambiguity in the API between a device name
828 (eg. C</dev/sdb2>) and a similar pathname. A file might just happen
829 to be called C<sdb2> in the directory C</dev> (consider some non-Unix
832 In the current API we usually resolve this ambiguity by having two
833 separate calls, for example L</guestfs_checksum> and
834 L</guestfs_checksum_device>. Some API calls are ambiguous and
835 (incorrectly) resolve the problem by detecting if the path supplied
836 begins with C</dev/>.
838 To avoid both the ambiguity and the need to duplicate some calls, we
839 could make paths/devices into structured names. One way to do this
840 would be to use a notation like grub (C<hd(0,0)>), although nobody
841 really likes this aspect of grub. Another way would be to use a
842 structured type, equivalent to this OCaml type:
844 type path = Path of string | Device of int | Partition of int * int
846 which would allow you to pass arguments like:
849 Device 1 (* /dev/sdb, or perhaps /dev/sda *)
850 Partition (1, 2) (* /dev/sdb2 (or is it /dev/sda2 or /dev/sdb3?) *)
851 Path "/dev/sdb2" (* not a device *)
853 As you can see there are still problems to resolve even with this
854 representation. Also consider how it might work in guestfish.
858 =head2 PROTOCOL LIMITS
860 Internally libguestfs uses a message-based protocol to pass API calls
861 and their responses to and from a small "appliance" (see L</INTERNALS>
862 for plenty more detail about this). The maximum message size used by
863 the protocol is slightly less than 4 MB. For some API calls you may
864 need to be aware of this limit. The API calls which may be affected
865 are individually documented, with a link back to this section of the
868 A simple call such as L</guestfs_cat> returns its result (the file
869 data) in a simple string. Because this string is at some point
870 internally encoded as a message, the maximum size that it can return
871 is slightly under 4 MB. If the requested file is larger than this
872 then you will get an error.
874 In order to transfer large files into and out of the guest filesystem,
875 you need to use particular calls that support this. The sections
876 L</UPLOADING> and L</DOWNLOADING> document how to do this.
878 You might also consider mounting the disk image using our FUSE
879 filesystem support (L<guestmount(1)>).
881 =head2 KEYS AND PASSPHRASES
883 Certain libguestfs calls take a parameter that contains sensitive key
884 material, passed in as a C string.
886 In the future we would hope to change the libguestfs implementation so
887 that keys are L<mlock(2)>-ed into physical RAM, and thus can never end
888 up in swap. However this is I<not> done at the moment, because of the
889 complexity of such an implementation.
891 Therefore you should be aware that any key parameter you pass to
892 libguestfs might end up being written out to the swap partition. If
893 this is a concern, scrub the swap partition or don't use libguestfs on
896 =head2 MULTIPLE HANDLES AND MULTIPLE THREADS
898 All high-level libguestfs actions are synchronous. If you want
899 to use libguestfs asynchronously then you must create a thread.
901 Only use the handle from a single thread. Either use the handle
902 exclusively from one thread, or provide your own mutex so that two
903 threads cannot issue calls on the same handle at the same time.
905 See the graphical program guestfs-browser for one possible
906 architecture for multithreaded programs using libvirt and libguestfs.
910 Libguestfs needs a kernel and initrd.img, which it finds by looking
911 along an internal path.
913 By default it looks for these in the directory C<$libdir/guestfs>
914 (eg. C</usr/local/lib/guestfs> or C</usr/lib64/guestfs>).
916 Use L</guestfs_set_path> or set the environment variable
917 L</LIBGUESTFS_PATH> to change the directories that libguestfs will
918 search in. The value is a colon-separated list of paths. The current
919 directory is I<not> searched unless the path contains an empty element
920 or C<.>. For example C<LIBGUESTFS_PATH=:/usr/lib/guestfs> would
921 search the current directory and then C</usr/lib/guestfs>.
925 If you want to compile your own qemu, run qemu from a non-standard
926 location, or pass extra arguments to qemu, then you can write a
927 shell-script wrapper around qemu.
929 There is one important rule to remember: you I<must C<exec qemu>> as
930 the last command in the shell script (so that qemu replaces the shell
931 and becomes the direct child of the libguestfs-using program). If you
932 don't do this, then the qemu process won't be cleaned up correctly.
934 Here is an example of a wrapper, where I have built my own copy of
938 qemudir=/home/rjones/d/qemu
939 exec $qemudir/x86_64-softmmu/qemu-system-x86_64 -L $qemudir/pc-bios "$@"
941 Save this script as C</tmp/qemu.wrapper> (or wherever), C<chmod +x>,
942 and then use it by setting the LIBGUESTFS_QEMU environment variable.
945 LIBGUESTFS_QEMU=/tmp/qemu.wrapper guestfish
947 Note that libguestfs also calls qemu with the -help and -version
948 options in order to determine features.
952 We guarantee the libguestfs ABI (binary interface), for public,
953 high-level actions as outlined in this section. Although we will
954 deprecate some actions, for example if they get replaced by newer
955 calls, we will keep the old actions forever. This allows you the
956 developer to program in confidence against the libguestfs API.
958 =head2 BLOCK DEVICE NAMING
960 In the kernel there is now quite a profusion of schemata for naming
961 block devices (in this context, by I<block device> I mean a physical
962 or virtual hard drive). The original Linux IDE driver used names
963 starting with C</dev/hd*>. SCSI devices have historically used a
964 different naming scheme, C</dev/sd*>. When the Linux kernel I<libata>
965 driver became a popular replacement for the old IDE driver
966 (particularly for SATA devices) those devices also used the
967 C</dev/sd*> scheme. Additionally we now have virtual machines with
968 paravirtualized drivers. This has created several different naming
969 systems, such as C</dev/vd*> for virtio disks and C</dev/xvd*> for Xen
972 As discussed above, libguestfs uses a qemu appliance running an
973 embedded Linux kernel to access block devices. We can run a variety
974 of appliances based on a variety of Linux kernels.
976 This causes a problem for libguestfs because many API calls use device
977 or partition names. Working scripts and the recipe (example) scripts
978 that we make available over the internet could fail if the naming
981 Therefore libguestfs defines C</dev/sd*> as the I<standard naming
982 scheme>. Internally C</dev/sd*> names are translated, if necessary,
983 to other names as required. For example, under RHEL 5 which uses the
984 C</dev/hd*> scheme, any device parameter C</dev/sda2> is translated to
985 C</dev/hda2> transparently.
987 Note that this I<only> applies to parameters. The
988 L</guestfs_list_devices>, L</guestfs_list_partitions> and similar calls
989 return the true names of the devices and partitions as known to the
992 =head3 ALGORITHM FOR BLOCK DEVICE NAME TRANSLATION
994 Usually this translation is transparent. However in some (very rare)
995 cases you may need to know the exact algorithm. Such cases include
996 where you use L</guestfs_config> to add a mixture of virtio and IDE
997 devices to the qemu-based appliance, so have a mixture of C</dev/sd*>
998 and C</dev/vd*> devices.
1000 The algorithm is applied only to I<parameters> which are known to be
1001 either device or partition names. Return values from functions such
1002 as L</guestfs_list_devices> are never changed.
1008 Is the string a parameter which is a device or partition name?
1012 Does the string begin with C</dev/sd>?
1016 Does the named device exist? If so, we use that device.
1017 However if I<not> then we continue with this algorithm.
1021 Replace initial C</dev/sd> string with C</dev/hd>.
1023 For example, change C</dev/sda2> to C</dev/hda2>.
1025 If that named device exists, use it. If not, continue.
1029 Replace initial C</dev/sd> string with C</dev/vd>.
1031 If that named device exists, use it. If not, return an error.
1035 =head3 PORTABILITY CONCERNS WITH BLOCK DEVICE NAMING
1037 Although the standard naming scheme and automatic translation is
1038 useful for simple programs and guestfish scripts, for larger programs
1039 it is best not to rely on this mechanism.
1041 Where possible for maximum future portability programs using
1042 libguestfs should use these future-proof techniques:
1048 Use L</guestfs_list_devices> or L</guestfs_list_partitions> to list
1049 actual device names, and then use those names directly.
1051 Since those device names exist by definition, they will never be
1056 Use higher level ways to identify filesystems, such as LVM names,
1057 UUIDs and filesystem labels.
1063 This section discusses security implications of using libguestfs,
1064 particularly with untrusted or malicious guests or disk images.
1066 =head2 GENERAL SECURITY CONSIDERATIONS
1068 Be careful with any files or data that you download from a guest (by
1069 "download" we mean not just the L</guestfs_download> command but any
1070 command that reads files, filenames, directories or anything else from
1071 a disk image). An attacker could manipulate the data to fool your
1072 program into doing the wrong thing. Consider cases such as:
1078 the data (file etc) not being present
1082 being present but empty
1086 being much larger than normal
1090 containing arbitrary 8 bit data
1094 being in an unexpected character encoding
1098 containing homoglyphs.
1102 =head2 SECURITY OF MOUNTING FILESYSTEMS
1104 When you mount a filesystem under Linux, mistakes in the kernel
1105 filesystem (VFS) module can sometimes be escalated into exploits by
1106 deliberately creating a malicious, malformed filesystem. These
1107 exploits are very severe for two reasons. Firstly there are very many
1108 filesystem drivers in the kernel, and many of them are infrequently
1109 used and not much developer attention has been paid to the code.
1110 Linux userspace helps potential crackers by detecting the filesystem
1111 type and automatically choosing the right VFS driver, even if that
1112 filesystem type is obscure or unexpected for the administrator.
1113 Secondly, a kernel-level exploit is like a local root exploit (worse
1114 in some ways), giving immediate and total access to the system right
1115 down to the hardware level.
1117 That explains why you should never mount a filesystem from an
1118 untrusted guest on your host kernel. How about libguestfs? We run a
1119 Linux kernel inside a qemu virtual machine, usually running as a
1120 non-root user. The attacker would need to write a filesystem which
1121 first exploited the kernel, and then exploited either qemu
1122 virtualization (eg. a faulty qemu driver) or the libguestfs protocol,
1123 and finally to be as serious as the host kernel exploit it would need
1124 to escalate its privileges to root. This multi-step escalation,
1125 performed by a static piece of data, is thought to be extremely hard
1126 to do, although we never say 'never' about security issues.
1128 In any case callers can reduce the attack surface by forcing the
1129 filesystem type when mounting (use L</guestfs_mount_vfs>).
1131 =head2 PROTOCOL SECURITY
1133 The protocol is designed to be secure, being based on RFC 4506 (XDR)
1134 with a defined upper message size. However a program that uses
1135 libguestfs must also take care - for example you can write a program
1136 that downloads a binary from a disk image and executes it locally, and
1137 no amount of protocol security will save you from the consequences.
1139 =head2 INSPECTION SECURITY
1141 Parts of the inspection API (see L</INSPECTION>) return untrusted
1142 strings directly from the guest, and these could contain any 8 bit
1143 data. Callers should be careful to escape these before printing them
1144 to a structured file (for example, use HTML escaping if creating a web
1147 Guest configuration may be altered in unusual ways by the
1148 administrator of the virtual machine, and may not reflect reality
1149 (particularly for untrusted or actively malicious guests). For
1150 example we parse the hostname from configuration files like
1151 C</etc/sysconfig/network> that we find in the guest, but the guest
1152 administrator can easily manipulate these files to provide the wrong
1155 The inspection API parses guest configuration using two external
1156 libraries: Augeas (Linux configuration) and hivex (Windows Registry).
1157 Both are designed to be robust in the face of malicious data, although
1158 denial of service attacks are still possible, for example with
1159 oversized configuration files.
1161 =head2 RUNNING UNTRUSTED GUEST COMMANDS
1163 Be very cautious about running commands from the guest. By running a
1164 command in the guest, you are giving CPU time to a binary that you do
1165 not control, under the same user account as the library, albeit
1166 wrapped in qemu virtualization. More information and alternatives can
1167 be found in the section L</RUNNING COMMANDS>.
1169 =head2 CVE-2010-3851
1171 https://bugzilla.redhat.com/642934
1173 This security bug concerns the automatic disk format detection that
1174 qemu does on disk images.
1176 A raw disk image is just the raw bytes, there is no header. Other
1177 disk images like qcow2 contain a special header. Qemu deals with this
1178 by looking for one of the known headers, and if none is found then
1179 assuming the disk image must be raw.
1181 This allows a guest which has been given a raw disk image to write
1182 some other header. At next boot (or when the disk image is accessed
1183 by libguestfs) qemu would do autodetection and think the disk image
1184 format was, say, qcow2 based on the header written by the guest.
1186 This in itself would not be a problem, but qcow2 offers many features,
1187 one of which is to allow a disk image to refer to another image
1188 (called the "backing disk"). It does this by placing the path to the
1189 backing disk into the qcow2 header. This path is not validated and
1190 could point to any host file (eg. "/etc/passwd"). The backing disk is
1191 then exposed through "holes" in the qcow2 disk image, which of course
1192 is completely under the control of the attacker.
1194 In libguestfs this is rather hard to exploit except under two
1201 You have enabled the network or have opened the disk in write mode.
1205 You are also running untrusted code from the guest (see
1206 L</RUNNING COMMANDS>).
1210 The way to avoid this is to specify the expected disk format when
1211 adding disks (the optional C<format> option to
1212 L</guestfs_add_drive_opts>). You should always do this if the disk is
1213 raw format, and it's a good idea for other cases too.
1215 For disks added from libvirt using calls like L</guestfs_add_domain>,
1216 the format is fetched from libvirt and passed through.
1218 For libguestfs tools, use the I<--format> command line parameter as
1221 =head1 CONNECTION MANAGEMENT
1225 C<guestfs_h> is the opaque type representing a connection handle.
1226 Create a handle by calling L</guestfs_create>. Call L</guestfs_close>
1227 to free the handle and release all resources used.
1229 For information on using multiple handles and threads, see the section
1230 L</MULTIPLE HANDLES AND MULTIPLE THREADS> below.
1232 =head2 guestfs_create
1234 guestfs_h *guestfs_create (void);
1236 Create a connection handle.
1238 You have to call L</guestfs_add_drive_opts> (or one of the equivalent
1239 calls) on the handle at least once.
1241 This function returns a non-NULL pointer to a handle on success or
1244 After configuring the handle, you have to call L</guestfs_launch>.
1246 You may also want to configure error handling for the handle. See
1247 L</ERROR HANDLING> section below.
1249 =head2 guestfs_close
1251 void guestfs_close (guestfs_h *g);
1253 This closes the connection handle and frees up all resources used.
1255 =head1 ERROR HANDLING
1257 API functions can return errors. For example, almost all functions
1258 that return C<int> will return C<-1> to indicate an error.
1260 Additional information is available for errors: an error message
1261 string and optionally an error number (errno) if the thing that failed
1264 You can get at the additional information about the last error on the
1265 handle by calling L</guestfs_last_error>, L</guestfs_last_errno>,
1266 and/or by setting up an error handler with
1267 L</guestfs_set_error_handler>.
1269 When the handle is created, a default error handler is installed which
1270 prints the error message string to C<stderr>. For small short-running
1271 command line programs it is sufficient to do:
1273 if (guestfs_launch (g) == -1)
1274 exit (EXIT_FAILURE);
1276 since the default error handler will ensure that an error message has
1277 been printed to C<stderr> before the program exits.
1279 For other programs the caller will almost certainly want to install an
1280 alternate error handler or do error handling in-line like this:
1282 g = guestfs_create ();
1284 /* This disables the default behaviour of printing errors
1286 guestfs_set_error_handler (g, NULL, NULL);
1288 if (guestfs_launch (g) == -1) {
1289 /* Examine the error message and print it etc. */
1290 char *msg = guestfs_last_error (g);
1291 int errnum = guestfs_last_errno (g);
1292 fprintf (stderr, "%s\n", msg);
1296 Out of memory errors are handled differently. The default action is
1297 to call L<abort(3)>. If this is undesirable, then you can set a
1298 handler using L</guestfs_set_out_of_memory_handler>.
1300 L</guestfs_create> returns C<NULL> if the handle cannot be created,
1301 and because there is no handle if this happens there is no way to get
1302 additional error information. However L</guestfs_create> is supposed
1303 to be a lightweight operation which can only fail because of
1304 insufficient memory (it returns NULL in this case).
1306 =head2 guestfs_last_error
1308 const char *guestfs_last_error (guestfs_h *g);
1310 This returns the last error message that happened on C<g>. If
1311 there has not been an error since the handle was created, then this
1314 The lifetime of the returned string is until the next error occurs, or
1315 L</guestfs_close> is called.
1317 =head2 guestfs_last_errno
1319 int guestfs_last_errno (guestfs_h *g);
1321 This returns the last error number (errno) that happened on C<g>.
1323 If successful, an errno integer not equal to zero is returned.
1325 If no error, this returns 0. This call can return 0 in three
1332 There has not been any error on the handle.
1336 There has been an error but the errno was meaningless. This
1337 corresponds to the case where the error did not come from a
1338 failed system call, but for some other reason.
1342 There was an error from a failed system call, but for some
1343 reason the errno was not captured and returned. This usually
1344 indicates a bug in libguestfs.
1348 Libguestfs tries to convert the errno from inside the applicance into
1349 a corresponding errno for the caller (not entirely trivial: the
1350 appliance might be running a completely different operating system
1351 from the library and error numbers are not standardized across
1352 Un*xen). If this could not be done, then the error is translated to
1353 C<EINVAL>. In practice this should only happen in very rare
1356 =head2 guestfs_set_error_handler
1358 typedef void (*guestfs_error_handler_cb) (guestfs_h *g,
1361 void guestfs_set_error_handler (guestfs_h *g,
1362 guestfs_error_handler_cb cb,
1365 The callback C<cb> will be called if there is an error. The
1366 parameters passed to the callback are an opaque data pointer and the
1367 error message string.
1369 C<errno> is not passed to the callback. To get that the callback must
1370 call L</guestfs_last_errno>.
1372 Note that the message string C<msg> is freed as soon as the callback
1373 function returns, so if you want to stash it somewhere you must make
1376 The default handler prints messages on C<stderr>.
1378 If you set C<cb> to C<NULL> then I<no> handler is called.
1380 =head2 guestfs_get_error_handler
1382 guestfs_error_handler_cb guestfs_get_error_handler (guestfs_h *g,
1385 Returns the current error handler callback.
1387 =head2 guestfs_set_out_of_memory_handler
1389 typedef void (*guestfs_abort_cb) (void);
1390 int guestfs_set_out_of_memory_handler (guestfs_h *g,
1393 The callback C<cb> will be called if there is an out of memory
1394 situation. I<Note this callback must not return>.
1396 The default is to call L<abort(3)>.
1398 You cannot set C<cb> to C<NULL>. You can't ignore out of memory
1401 =head2 guestfs_get_out_of_memory_handler
1403 guestfs_abort_fn guestfs_get_out_of_memory_handler (guestfs_h *g);
1405 This returns the current out of memory handler.
1417 =head2 GROUPS OF FUNCTIONALITY IN THE APPLIANCE
1419 Using L</guestfs_available> you can test availability of
1420 the following groups of functions. This test queries the
1421 appliance to see if the appliance you are currently using
1422 supports the functionality.
1426 =head2 GUESTFISH supported COMMAND
1428 In L<guestfish(3)> there is a handy interactive command
1429 C<supported> which prints out the available groups and
1430 whether they are supported by this build of libguestfs.
1431 Note however that you have to do C<run> first.
1433 =head2 SINGLE CALLS AT COMPILE TIME
1435 Since version 1.5.8, C<E<lt>guestfs.hE<gt>> defines symbols
1436 for each C API function, such as:
1438 #define LIBGUESTFS_HAVE_DD 1
1440 if L</guestfs_dd> is available.
1442 Before version 1.5.8, if you needed to test whether a single
1443 libguestfs function is available at compile time, we recommended using
1444 build tools such as autoconf or cmake. For example in autotools you
1447 AC_CHECK_LIB([guestfs],[guestfs_create])
1448 AC_CHECK_FUNCS([guestfs_dd])
1450 which would result in C<HAVE_GUESTFS_DD> being either defined
1451 or not defined in your program.
1453 =head2 SINGLE CALLS AT RUN TIME
1455 Testing at compile time doesn't guarantee that a function really
1456 exists in the library. The reason is that you might be dynamically
1457 linked against a previous I<libguestfs.so> (dynamic library)
1458 which doesn't have the call. This situation unfortunately results
1459 in a segmentation fault, which is a shortcoming of the C dynamic
1460 linking system itself.
1462 You can use L<dlopen(3)> to test if a function is available
1463 at run time, as in this example program (note that you still
1464 need the compile time check as well):
1470 #include <guestfs.h>
1474 #ifdef LIBGUESTFS_HAVE_DD
1478 /* Test if the function guestfs_dd is really available. */
1479 dl = dlopen (NULL, RTLD_LAZY);
1481 fprintf (stderr, "dlopen: %s\n", dlerror ());
1482 exit (EXIT_FAILURE);
1484 has_function = dlsym (dl, "guestfs_dd") != NULL;
1488 printf ("this libguestfs.so does NOT have guestfs_dd function\n");
1490 printf ("this libguestfs.so has guestfs_dd function\n");
1491 /* Now it's safe to call
1492 guestfs_dd (g, "foo", "bar");
1496 printf ("guestfs_dd function was not found at compile time\n");
1500 You may think the above is an awful lot of hassle, and it is.
1501 There are other ways outside of the C linking system to ensure
1502 that this kind of incompatibility never arises, such as using
1505 Requires: libguestfs >= 1.0.80
1507 =head1 CALLS WITH OPTIONAL ARGUMENTS
1509 A recent feature of the API is the introduction of calls which take
1510 optional arguments. In C these are declared 3 ways. The main way is
1511 as a call which takes variable arguments (ie. C<...>), as in this
1514 int guestfs_add_drive_opts (guestfs_h *g, const char *filename, ...);
1516 Call this with a list of optional arguments, terminated by C<-1>.
1517 So to call with no optional arguments specified:
1519 guestfs_add_drive_opts (g, filename, -1);
1521 With a single optional argument:
1523 guestfs_add_drive_opts (g, filename,
1524 GUESTFS_ADD_DRIVE_OPTS_FORMAT, "qcow2",
1529 guestfs_add_drive_opts (g, filename,
1530 GUESTFS_ADD_DRIVE_OPTS_FORMAT, "qcow2",
1531 GUESTFS_ADD_DRIVE_OPTS_READONLY, 1,
1534 and so forth. Don't forget the terminating C<-1> otherwise
1535 Bad Things will happen!
1537 =head2 USING va_list FOR OPTIONAL ARGUMENTS
1539 The second variant has the same name with the suffix C<_va>, which
1540 works the same way but takes a C<va_list>. See the C manual for
1541 details. For the example function, this is declared:
1543 int guestfs_add_drive_opts_va (guestfs_h *g, const char *filename,
1546 =head2 CONSTRUCTING OPTIONAL ARGUMENTS
1548 The third variant is useful where you need to construct these
1549 calls. You pass in a structure where you fill in the optional
1550 fields. The structure has a bitmask as the first element which
1551 you must set to indicate which fields you have filled in. For
1552 our example function the structure and call are declared:
1554 struct guestfs_add_drive_opts_argv {
1560 int guestfs_add_drive_opts_argv (guestfs_h *g, const char *filename,
1561 const struct guestfs_add_drive_opts_argv *optargs);
1563 You could call it like this:
1565 struct guestfs_add_drive_opts_argv optargs = {
1566 .bitmask = GUESTFS_ADD_DRIVE_OPTS_READONLY_BITMASK |
1567 GUESTFS_ADD_DRIVE_OPTS_FORMAT_BITMASK,
1572 guestfs_add_drive_opts_argv (g, filename, &optargs);
1580 The C<_BITMASK> suffix on each option name when specifying the
1585 You do not need to fill in all fields of the structure.
1589 There must be a one-to-one correspondence between fields of the
1590 structure that are filled in, and bits set in the bitmask.
1594 =head2 OPTIONAL ARGUMENTS IN OTHER LANGUAGES
1596 In other languages, optional arguments are expressed in the
1597 way that is natural for that language. We refer you to the
1598 language-specific documentation for more details on that.
1600 For guestfish, see L<guestfish(1)/OPTIONAL ARGUMENTS>.
1602 =head2 SETTING CALLBACKS TO HANDLE EVENTS
1604 The child process generates events in some situations. Current events
1605 include: receiving a log message, the child process exits.
1607 Use the C<guestfs_set_*_callback> functions to set a callback for
1608 different types of events.
1610 Only I<one callback of each type> can be registered for each handle.
1611 Calling C<guestfs_set_*_callback> again overwrites the previous
1612 callback of that type. Cancel all callbacks of this type by calling
1613 this function with C<cb> set to C<NULL>.
1615 =head2 guestfs_set_log_message_callback
1617 typedef void (*guestfs_log_message_cb) (guestfs_h *g, void *opaque,
1618 char *buf, int len);
1619 void guestfs_set_log_message_callback (guestfs_h *g,
1620 guestfs_log_message_cb cb,
1623 The callback function C<cb> will be called whenever qemu or the guest
1624 writes anything to the console.
1626 Use this function to capture kernel messages and similar.
1628 Normally there is no log message handler, and log messages are just
1631 =head2 guestfs_set_subprocess_quit_callback
1633 typedef void (*guestfs_subprocess_quit_cb) (guestfs_h *g, void *opaque);
1634 void guestfs_set_subprocess_quit_callback (guestfs_h *g,
1635 guestfs_subprocess_quit_cb cb,
1638 The callback function C<cb> will be called when the child process
1639 quits, either asynchronously or if killed by
1640 L</guestfs_kill_subprocess>. (This corresponds to a transition from
1641 any state to the CONFIG state).
1643 =head2 guestfs_set_launch_done_callback
1645 typedef void (*guestfs_launch_done_cb) (guestfs_h *g, void *opaque);
1646 void guestfs_set_launch_done_callback (guestfs_h *g,
1647 guestfs_launch_done_cb cb,
1650 The callback function C<cb> will be called when the child process
1651 becomes ready first time after it has been launched. (This
1652 corresponds to a transition from LAUNCHING to the READY state).
1654 =head2 guestfs_set_close_callback
1656 typedef void (*guestfs_close_cb) (guestfs_h *g, void *opaque);
1657 void guestfs_set_close_callback (guestfs_h *g,
1658 guestfs_close_cb cb,
1661 The callback function C<cb> will be called while the handle
1662 is being closed (synchronously from L</guestfs_close>).
1664 Note that libguestfs installs an L<atexit(3)> handler to try to
1665 clean up handles that are open when the program exits. This
1666 means that this callback might be called indirectly from
1667 L<exit(3)>, which can cause unexpected problems in higher-level
1668 languages (eg. if your HLL interpreter has already been cleaned
1669 up by the time this is called, and if your callback then jumps
1670 into some HLL function).
1672 =head2 guestfs_set_progress_callback
1674 typedef void (*guestfs_progress_cb) (guestfs_h *g, void *opaque,
1675 int proc_nr, int serial,
1676 uint64_t position, uint64_t total);
1677 void guestfs_set_progress_callback (guestfs_h *g,
1678 guestfs_progress_cb cb,
1681 Some long-running operations can generate progress messages. If
1682 this callback is registered, then it will be called each time a
1683 progress message is generated (usually two seconds after the
1684 operation started, and three times per second thereafter until
1685 it completes, although the frequency may change in future versions).
1687 The callback receives two numbers: C<position> and C<total>.
1688 The units of C<total> are not defined, although for some
1689 operations C<total> may relate in some way to the amount of
1690 data to be transferred (eg. in bytes or megabytes), and
1691 C<position> may be the portion which has been transferred.
1693 The only defined and stable parts of the API are:
1699 The callback can display to the user some type of progress bar or
1700 indicator which shows the ratio of C<position>:C<total>.
1704 0 E<lt>= C<position> E<lt>= C<total>
1708 If any progress notification is sent during a call, then a final
1709 progress notification is always sent when C<position> = C<total>.
1711 This is to simplify caller code, so callers can easily set the
1712 progress indicator to "100%" at the end of the operation, without
1713 requiring special code to detect this case.
1717 The callback also receives the procedure number and serial number of
1718 the call. These are only useful for debugging protocol issues, and
1719 the callback can normally ignore them. The callback may want to
1720 print these numbers in error messages or debugging messages.
1722 =head1 PRIVATE DATA AREA
1724 You can attach named pieces of private data to the libguestfs handle,
1725 and fetch them by name for the lifetime of the handle. This is called
1726 the private data area and is only available from the C API.
1728 To attach a named piece of data, use the following call:
1730 void guestfs_set_private (guestfs_h *g, const char *key, void *data);
1732 C<key> is the name to associate with this data, and C<data> is an
1733 arbitrary pointer (which can be C<NULL>). Any previous item with the
1734 same name is overwritten.
1736 You can use any C<key> you want, but names beginning with an
1737 underscore character are reserved for internal libguestfs purposes
1738 (for implementing language bindings). It is recommended to prefix the
1739 name with some unique string to avoid collisions with other users.
1741 To retrieve the pointer, use:
1743 void *guestfs_get_private (guestfs_h *g, const char *key);
1745 This function returns C<NULL> if either no data is found associated
1746 with C<key>, or if the user previously set the C<key>'s C<data>
1749 Libguestfs does not try to look at or interpret the C<data> pointer in
1750 any way. As far as libguestfs is concerned, it need not be a valid
1751 pointer at all. In particular, libguestfs does I<not> try to free the
1752 data when the handle is closed. If the data must be freed, then the
1753 caller must either free it before calling L</guestfs_close> or must
1754 set up a close callback to do it (see L</guestfs_set_close_callback>,
1755 and note that only one callback can be registered for a handle).
1757 The private data area is implemented using a hash table, and should be
1758 reasonably efficient for moderate numbers of keys.
1762 <!-- old anchor for the next section -->
1763 <a name="state_machine_and_low_level_event_api"/>
1769 Internally, libguestfs is implemented by running an appliance (a
1770 special type of small virtual machine) using L<qemu(1)>. Qemu runs as
1771 a child process of the main program.
1777 | | child process / appliance
1778 | | __________________________
1780 +-------------------+ RPC | +-----------------+ |
1781 | libguestfs <--------------------> guestfsd | |
1782 | | | +-----------------+ |
1783 \___________________/ | | Linux kernel | |
1784 | +--^--------------+ |
1785 \_________|________________/
1793 The library, linked to the main program, creates the child process and
1794 hence the appliance in the L</guestfs_launch> function.
1796 Inside the appliance is a Linux kernel and a complete stack of
1797 userspace tools (such as LVM and ext2 programs) and a small
1798 controlling daemon called L</guestfsd>. The library talks to
1799 L</guestfsd> using remote procedure calls (RPC). There is a mostly
1800 one-to-one correspondence between libguestfs API calls and RPC calls
1801 to the daemon. Lastly the disk image(s) are attached to the qemu
1802 process which translates device access by the appliance's Linux kernel
1803 into accesses to the image.
1805 A common misunderstanding is that the appliance "is" the virtual
1806 machine. Although the disk image you are attached to might also be
1807 used by some virtual machine, libguestfs doesn't know or care about
1808 this. (But you will care if both libguestfs's qemu process and your
1809 virtual machine are trying to update the disk image at the same time,
1810 since these usually results in massive disk corruption).
1812 =head1 STATE MACHINE
1814 libguestfs uses a state machine to model the child process:
1825 / | \ \ guestfs_launch
1836 \______/ <------ \________/
1838 The normal transitions are (1) CONFIG (when the handle is created, but
1839 there is no child process), (2) LAUNCHING (when the child process is
1840 booting up), (3) alternating between READY and BUSY as commands are
1841 issued to, and carried out by, the child process.
1843 The guest may be killed by L</guestfs_kill_subprocess>, or may die
1844 asynchronously at any time (eg. due to some internal error), and that
1845 causes the state to transition back to CONFIG.
1847 Configuration commands for qemu such as L</guestfs_add_drive> can only
1848 be issued when in the CONFIG state.
1850 The API offers one call that goes from CONFIG through LAUNCHING to
1851 READY. L</guestfs_launch> blocks until the child process is READY to
1852 accept commands (or until some failure or timeout).
1853 L</guestfs_launch> internally moves the state from CONFIG to LAUNCHING
1854 while it is running.
1856 API actions such as L</guestfs_mount> can only be issued when in the
1857 READY state. These API calls block waiting for the command to be
1858 carried out (ie. the state to transition to BUSY and then back to
1859 READY). There are no non-blocking versions, and no way to issue more
1860 than one command per handle at the same time.
1862 Finally, the child process sends asynchronous messages back to the
1863 main program, such as kernel log messages. You can register a
1864 callback to receive these messages.
1868 =head2 COMMUNICATION PROTOCOL
1870 Don't rely on using this protocol directly. This section documents
1871 how it currently works, but it may change at any time.
1873 The protocol used to talk between the library and the daemon running
1874 inside the qemu virtual machine is a simple RPC mechanism built on top
1875 of XDR (RFC 1014, RFC 1832, RFC 4506).
1877 The detailed format of structures is in C<src/guestfs_protocol.x>
1878 (note: this file is automatically generated).
1880 There are two broad cases, ordinary functions that don't have any
1881 C<FileIn> and C<FileOut> parameters, which are handled with very
1882 simple request/reply messages. Then there are functions that have any
1883 C<FileIn> or C<FileOut> parameters, which use the same request and
1884 reply messages, but they may also be followed by files sent using a
1887 =head3 ORDINARY FUNCTIONS (NO FILEIN/FILEOUT PARAMS)
1889 For ordinary functions, the request message is:
1891 total length (header + arguments,
1892 but not including the length word itself)
1893 struct guestfs_message_header (encoded as XDR)
1894 struct guestfs_<foo>_args (encoded as XDR)
1896 The total length field allows the daemon to allocate a fixed size
1897 buffer into which it slurps the rest of the message. As a result, the
1898 total length is limited to C<GUESTFS_MESSAGE_MAX> bytes (currently
1899 4MB), which means the effective size of any request is limited to
1900 somewhere under this size.
1902 Note also that many functions don't take any arguments, in which case
1903 the C<guestfs_I<foo>_args> is completely omitted.
1905 The header contains the procedure number (C<guestfs_proc>) which is
1906 how the receiver knows what type of args structure to expect, or none
1909 For functions that take optional arguments, the optional arguments are
1910 encoded in the C<guestfs_I<foo>_args> structure in the same way as
1911 ordinary arguments. A bitmask in the header indicates which optional
1912 arguments are meaningful. The bitmask is also checked to see if it
1913 contains bits set which the daemon does not know about (eg. if more
1914 optional arguments were added in a later version of the library), and
1915 this causes the call to be rejected.
1917 The reply message for ordinary functions is:
1919 total length (header + ret,
1920 but not including the length word itself)
1921 struct guestfs_message_header (encoded as XDR)
1922 struct guestfs_<foo>_ret (encoded as XDR)
1924 As above the C<guestfs_I<foo>_ret> structure may be completely omitted
1925 for functions that return no formal return values.
1927 As above the total length of the reply is limited to
1928 C<GUESTFS_MESSAGE_MAX>.
1930 In the case of an error, a flag is set in the header, and the reply
1931 message is slightly changed:
1933 total length (header + error,
1934 but not including the length word itself)
1935 struct guestfs_message_header (encoded as XDR)
1936 struct guestfs_message_error (encoded as XDR)
1938 The C<guestfs_message_error> structure contains the error message as a
1941 =head3 FUNCTIONS THAT HAVE FILEIN PARAMETERS
1943 A C<FileIn> parameter indicates that we transfer a file I<into> the
1944 guest. The normal request message is sent (see above). However this
1945 is followed by a sequence of file chunks.
1947 total length (header + arguments,
1948 but not including the length word itself,
1949 and not including the chunks)
1950 struct guestfs_message_header (encoded as XDR)
1951 struct guestfs_<foo>_args (encoded as XDR)
1952 sequence of chunks for FileIn param #0
1953 sequence of chunks for FileIn param #1 etc.
1955 The "sequence of chunks" is:
1957 length of chunk (not including length word itself)
1958 struct guestfs_chunk (encoded as XDR)
1960 struct guestfs_chunk (encoded as XDR)
1963 struct guestfs_chunk (with data.data_len == 0)
1965 The final chunk has the C<data_len> field set to zero. Additionally a
1966 flag is set in the final chunk to indicate either successful
1967 completion or early cancellation.
1969 At time of writing there are no functions that have more than one
1970 FileIn parameter. However this is (theoretically) supported, by
1971 sending the sequence of chunks for each FileIn parameter one after
1972 another (from left to right).
1974 Both the library (sender) I<and> the daemon (receiver) may cancel the
1975 transfer. The library does this by sending a chunk with a special
1976 flag set to indicate cancellation. When the daemon sees this, it
1977 cancels the whole RPC, does I<not> send any reply, and goes back to
1978 reading the next request.
1980 The daemon may also cancel. It does this by writing a special word
1981 C<GUESTFS_CANCEL_FLAG> to the socket. The library listens for this
1982 during the transfer, and if it gets it, it will cancel the transfer
1983 (it sends a cancel chunk). The special word is chosen so that even if
1984 cancellation happens right at the end of the transfer (after the
1985 library has finished writing and has started listening for the reply),
1986 the "spurious" cancel flag will not be confused with the reply
1989 This protocol allows the transfer of arbitrary sized files (no 32 bit
1990 limit), and also files where the size is not known in advance
1991 (eg. from pipes or sockets). However the chunks are rather small
1992 (C<GUESTFS_MAX_CHUNK_SIZE>), so that neither the library nor the
1993 daemon need to keep much in memory.
1995 =head3 FUNCTIONS THAT HAVE FILEOUT PARAMETERS
1997 The protocol for FileOut parameters is exactly the same as for FileIn
1998 parameters, but with the roles of daemon and library reversed.
2000 total length (header + ret,
2001 but not including the length word itself,
2002 and not including the chunks)
2003 struct guestfs_message_header (encoded as XDR)
2004 struct guestfs_<foo>_ret (encoded as XDR)
2005 sequence of chunks for FileOut param #0
2006 sequence of chunks for FileOut param #1 etc.
2008 =head3 INITIAL MESSAGE
2010 When the daemon launches it sends an initial word
2011 (C<GUESTFS_LAUNCH_FLAG>) which indicates that the guest and daemon is
2012 alive. This is what L</guestfs_launch> waits for.
2014 =head3 PROGRESS NOTIFICATION MESSAGES
2016 The daemon may send progress notification messages at any time. These
2017 are distinguished by the normal length word being replaced by
2018 C<GUESTFS_PROGRESS_FLAG>, followed by a fixed size progress message.
2020 The library turns them into progress callbacks (see
2021 C<guestfs_set_progress_callback>) if there is a callback registered,
2022 or discards them if not.
2024 The daemon self-limits the frequency of progress messages it sends
2025 (see C<daemon/proto.c:notify_progress>). Not all calls generate
2028 =head1 LIBGUESTFS VERSION NUMBERS
2030 Since April 2010, libguestfs has started to make separate development
2031 and stable releases, along with corresponding branches in our git
2032 repository. These separate releases can be identified by version
2035 even numbers for stable: 1.2.x, 1.4.x, ...
2036 .-------- odd numbers for development: 1.3.x, 1.5.x, ...
2042 | `-------- sub-version
2044 `------ always '1' because we don't change the ABI
2046 Thus "1.3.5" is the 5th update to the development branch "1.3".
2048 As time passes we cherry pick fixes from the development branch and
2049 backport those into the stable branch, the effect being that the
2050 stable branch should get more stable and less buggy over time. So the
2051 stable releases are ideal for people who don't need new features but
2052 would just like the software to work.
2054 Our criteria for backporting changes are:
2060 Documentation changes which don't affect any code are
2061 backported unless the documentation refers to a future feature
2062 which is not in stable.
2066 Bug fixes which are not controversial, fix obvious problems, and
2067 have been well tested are backported.
2071 Simple rearrangements of code which shouldn't affect how it works get
2072 backported. This is so that the code in the two branches doesn't get
2073 too far out of step, allowing us to backport future fixes more easily.
2077 We I<don't> backport new features, new APIs, new tools etc, except in
2078 one exceptional case: the new feature is required in order to
2079 implement an important bug fix.
2083 A new stable branch starts when we think the new features in
2084 development are substantial and compelling enough over the current
2085 stable branch to warrant it. When that happens we create new stable
2086 and development versions 1.N.0 and 1.(N+1).0 [N is even]. The new
2087 dot-oh release won't necessarily be so stable at this point, but by
2088 backporting fixes from development, that branch will stabilize over
2091 =head1 ENVIRONMENT VARIABLES
2095 =item LIBGUESTFS_APPEND
2097 Pass additional options to the guest kernel.
2099 =item LIBGUESTFS_DEBUG
2101 Set C<LIBGUESTFS_DEBUG=1> to enable verbose messages. This
2102 has the same effect as calling C<guestfs_set_verbose (g, 1)>.
2104 =item LIBGUESTFS_MEMSIZE
2106 Set the memory allocated to the qemu process, in megabytes. For
2109 LIBGUESTFS_MEMSIZE=700
2111 =item LIBGUESTFS_PATH
2113 Set the path that libguestfs uses to search for kernel and initrd.img.
2114 See the discussion of paths in section PATH above.
2116 =item LIBGUESTFS_QEMU
2118 Set the default qemu binary that libguestfs uses. If not set, then
2119 the qemu which was found at compile time by the configure script is
2122 See also L</QEMU WRAPPERS> above.
2124 =item LIBGUESTFS_TRACE
2126 Set C<LIBGUESTFS_TRACE=1> to enable command traces. This
2127 has the same effect as calling C<guestfs_set_trace (g, 1)>.
2131 Location of temporary directory, defaults to C</tmp>.
2133 If libguestfs was compiled to use the supermin appliance then the
2134 real appliance is cached in this directory, shared between all
2135 handles belonging to the same EUID. You can use C<$TMPDIR> to
2136 configure another directory to use in case C</tmp> is not large
2143 L<guestfs-examples(3)>,
2144 L<guestfs-ocaml(3)>,
2145 L<guestfs-python(3)>,
2152 L<virt-filesystems(1)>,
2153 L<virt-inspector(1)>,
2154 L<virt-list-filesystems(1)>,
2155 L<virt-list-partitions(1)>,
2164 L<http://libguestfs.org/>.
2166 Tools with a similar purpose:
2175 To get a list of bugs against libguestfs use this link:
2177 L<https://bugzilla.redhat.com/buglist.cgi?component=libguestfs&product=Virtualization+Tools>
2179 To report a new bug against libguestfs use this link:
2181 L<https://bugzilla.redhat.com/enter_bug.cgi?component=libguestfs&product=Virtualization+Tools>
2183 When reporting a bug, please check:
2189 That the bug hasn't been reported already.
2193 That you are testing a recent version.
2197 Describe the bug accurately, and give a way to reproduce it.
2201 Run libguestfs-test-tool and paste the B<complete, unedited>
2202 output into the bug report.
2208 Richard W.M. Jones (C<rjones at redhat dot com>)
2212 Copyright (C) 2009-2010 Red Hat Inc.
2213 L<http://libguestfs.org/>
2215 This library is free software; you can redistribute it and/or
2216 modify it under the terms of the GNU Lesser General Public
2217 License as published by the Free Software Foundation; either
2218 version 2 of the License, or (at your option) any later version.
2220 This library is distributed in the hope that it will be useful,
2221 but WITHOUT ANY WARRANTY; without even the implied warranty of
2222 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
2223 Lesser General Public License for more details.
2225 You should have received a copy of the GNU Lesser General Public
2226 License along with this library; if not, write to the Free Software
2227 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA