1 Updating Information for FreeBSD current users
3 This file is maintained and copyrighted by M. Warner Losh <imp@freebsd.org>.
4 See end of file for further details. For commonly done items, please see the
5 COMMON ITEMS: section later in the file. These instructions assume that you
6 basically know what you are doing. If not, then please consult the FreeBSD
9 http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/makeworld.html
11 Items affecting the ports and packages system can be found in
12 /usr/ports/UPDATING. Please read that file before running portupgrade.
15 Behavior of devfs rules path matching has been changed.
16 Pattern is now always matched against fully qualified devfs
17 path and slash characters must be explicitly matched by
18 slashes in pattern (FNM_PATHNAME). Rulesets involving devfs
19 subdirectories must be reviewed.
22 hastctl(8)'s `status' command output changed to terse one-liner format.
23 Scripts using this should switch to `list' command or be rewritten.
26 Fix a bug that allowed a tracing process (e.g. gdb) to write
27 to a memory-mapped file in the traced process's address space
28 even if neither the traced process nor the tracing process had
29 write access to that file.
32 Added ZFS TRIM support which is enabled by default. To disable
33 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
35 Creating new ZFS pools and adding new devices to existing pools
36 first performs a full device level TRIM which can take a significant
37 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
38 to disable this behaviour.
40 ZFS TRIM requires the underlying device support BIO_DELETE which
41 is currently provided by methods such as ATA TRIM and SCSI UNMAP
42 via CAM, which are typically supported by SSD's.
44 Stats for ZFS TRIM can be monitored by looking at the sysctl's
45 under kstat.zfs.misc.zio_trim.
48 `list' command has been added to hastctl(8). For now, it is full
49 equivalent of `status' command.
50 WARNING: in the near future the output of hastctl's status command
51 will change to more terse format. If you use `hastctl status'
52 for parsing in your scripts, switch to `hastctl list'.
55 The mergemaster command now uses the default MAKEOBJDIRPREFIX
56 rather than creating it's own in the temporary directory in
57 order allow access to bootstrapped versions of tools such as
58 install and mtree. When upgrading from version of FreeBSD where
59 the install command does not support -l, you will need to
60 install a new mergemaster command if mergemaster -p is required.
61 This can be accomplished with the command (cd src/usr.sbin/mergemaster
64 Due to the use of the new -l option to install(1) during build
65 and install, you must take care not to directly set the INSTALL
66 make variable in your /etc/make.conf, /etc/src.conf, or on the
67 command line. If you wish to use the -C flag for all installs
68 you may be able to add INSTALL+=-C to /etc/make.conf or
72 Fix a bug that allows NFS clients to issue READDIR on files.
75 The install(1) option -M has changed meaning and now takes an
76 argument that is a file or path to append logs to. In the
77 unlikely event that -M was the last option on the command line
78 and the command line contained at least two files and a target
79 directory the first file will have logs appended to it. The -M
80 option served little practical purpose in the last decade so it's
81 used expected to be extremely rare.
84 A new compression method (lz4) has been merged to. Please refer to
85 zpool-features(7) for more information.
87 Please refer to the "ZFS notes" section of this file for information
88 on upgrading boot ZFS pools.
91 The VFS KBI was changed with the merge of several nullfs
92 optimizations and fixes. All filesystem modules must be
96 With the addition of auditdistd(8), a new auditdistd user is now
97 depended on during installworld. "mergemaster -p" can be used to add
98 the user prior to installworld, as documented in the handbook.
104 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
105 Starting with this version the old system of ZFS pool versioning
106 is superseded by "feature flags". This concept enables forward
107 compatibility against certain future changes in functionality of ZFS
108 pools. The first two read-only compatible "feature flags" for ZFS
109 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
110 For more information read the new zpool-features(7) manual page.
111 Please refer to the "ZFS notes" section of this file for information
112 on upgrading boot ZFS pools.
115 The commit introducing bsd.compiler.mk breaks the traditional
116 building of kernels before this point. Add -m ${SRC}/share/mk
117 (for the right value of SRC) to your command lines to work
118 around; update your useland to a point after this; or use the
119 buildkernel/installkernel top-level targets. See also 20120829.
122 The IPFIREWALL_FORWARD kernel option has been removed. Its
123 functionality now turned on by default.
126 The random(4) support for the VIA hardware random number
127 generator (`PADLOCK') is no longer enabled unconditionally.
128 Add the PADLOCK_RNG option in the custom kernel config if
129 needed. The GENERIC kernels on i386 and amd64 do include the
130 option, so the change only affects the custom kernel
134 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
135 the traditional method, you must update your system with an installworld
136 before the kernel will build. The documented make buildkernel/installkernel
137 interfaces (coupled with fresh make kernel-toolchain) continue to work.
140 The sparc64 ZFS loader has been changed to no longer try to auto-
141 detect ZFS providers based on diskN aliases but now requires these
142 to be explicitly listed in the OFW boot-device environment variable.
145 Now unix domain sockets behave "as expected" on nullfs(5). Previously
146 nullfs(5) did not pass through all behaviours to the underlying layer,
147 as a result if we bound to a socket on the lower layer we could connect
148 only to the lower path; if we bound to the upper layer we could connect
149 only to the upper path. The new behavior is one can connect to both the
150 lower and the upper paths regardless what layer path one binds to.
153 The acpi_wmi(4) status device /dev/wmistat has been renamed to
157 A new VOP_ADVISE() was added to support posix_fadvise(2). All
158 filesystem modules must be recompiled.
161 The interface of the VOP_VPTOCNP(9) changed, now the returned
162 vnode shall be referenced, previously it was required to be
163 only held. All in-tree filesystems are converted.
169 The broken amd(4) driver has been replaced with esp(4) in the amd64,
170 i386 and pc98 GENERIC kernel configuration files.
173 This commit modifies vfs_register() so that it uses a hash
174 calculation to set vfc_typenum, which is enabled by default.
175 The first time a system is booted after this change, the
176 vfc_typenum values will change for all file systems. The
177 main effect of this is a change to the NFS server file handles
178 for file systems that use vfc_typenum in their fsid, such as ZFS.
179 It will, however, prevent vfc_typenum from changing when file
180 systems are loaded in a different order for subsequent reboots.
181 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
182 until you are ready to remount all NFS clients after a reboot.
185 Bump the shared library version numbers for libraries that
186 do not use symbol versioning, have changed the ABI compared
187 to stable/8 and which shared library version was not bumped.
188 Done as part of 9.0-RELEASE cycle.
191 During the merge of Capsicum features, the fget(9) KPI was modified.
192 This may require the rebuilding of out-of-tree device drivers --
193 issues have been reported specifically with the nVidia device driver.
194 __FreeBSD_version is bumped to 900041.
196 Also, there is a period between 20110811 and 20110814 where the
197 special devices /dev/{stdin,stdout,stderr} did not work correctly.
198 Building world from a kernel during that window may not work.
201 The packet filter (pf) code has been updated to OpenBSD 4.5.
202 You need to update userland tools to be in sync with kernel.
203 This update breaks backward compatibility with earlier pfsync(4)
204 versions. Care must be taken when updating redundant firewall setups.
207 The following sysctls and tunables are retired on x86 platforms:
209 machdep.hlt_logical_cpus
210 The following sysctl is retired:
211 machdep.hyperthreading_allowed
212 The sysctls were supposed to provide a way to dynamically offline and
213 online selected CPUs on x86 platforms, but the implementation has not
214 been reliable especially with SCHED_ULE scheduler.
215 machdep.hyperthreading_allowed tunable is still available to ignore
216 hyperthreading CPUs at OS level.
217 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
218 where X is an APIC ID of a CPU. Be advised, though, that disabling
219 CPUs in non-uniform fashion will result in non-uniform topology and
220 may lead to sub-optimal system performance with SCHED_ULE, which is
224 cpumask_t type is retired and cpuset_t is used in order to describe
228 Changes to ifconfig(8) for dynamic address family detection mandate
229 that you are running a kernel of 20110525 or later. Make sure to
230 follow the update procedure to boot a new kernel before installing
234 Support for sun4v architecture is officially dropped
237 Several KPI breaking changes have been committed to the mii(4) layer,
238 the PHY drivers and consequently some Ethernet drivers using mii(4).
239 This means that miibus.ko and the modules of the affected Ethernet
240 drivers need to be recompiled.
242 Note to kernel developers: Given that the OUI bit reversion problem
243 was fixed as part of these changes all mii(4) commits related to OUIs,
244 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
245 handling, no longer can be merged verbatim to stable/8 and previous
249 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
250 into their kernel configurations along with 'device pci'.
253 The default NFS client is now the new NFS client, so fstype "newnfs"
254 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
255 Although mounts via fstype "nfs" will usually work without userland
256 changes, it is recommended that the mount(8) and mount_nfs(8)
257 commands be rebuilt from sources and that a link to mount_nfs called
258 mount_oldnfs be created. The new client is compiled into the
259 kernel with "options NFSCL" and this is needed for diskless root
260 file systems. The GENERIC kernel configs have been changed to use
261 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
262 To use the regular/old client, you can "mount -t oldnfs ...". For
263 a diskless root file system, you must also include a line like:
265 vfs.root.mountfrom="oldnfs:"
267 in the boot/loader.conf on the root fs on the NFS server to make
268 a diskless root fs use the old client.
271 The GENERIC kernels for all architectures now default to the new
272 CAM-based ATA stack. It means that all legacy ATA drivers were
273 removed and replaced by respective CAM drivers. If you are using
274 ATA device names in /etc/fstab or other places, make sure to update
275 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
276 where 'Y's are the sequential numbers starting from zero for each type
277 in order of detection, unless configured otherwise with tunables,
278 see cam(4)). There will be symbolic links created in /dev/ to map
279 old adX devices to the respective adaY. They should provide basic
280 compatibility for file systems mounting in most cases, but they do
281 not support old user-level APIs and do not have respective providers
282 in GEOM. Consider using updated management tools with new device names.
284 It is possible to load devices ahci, ata, siis and mvs as modules,
285 but option ATA_CAM should remain in kernel configuration to make ata
286 module work as CAM driver supporting legacy ATA controllers. Device ata
287 still can be used in modular fashion (atacore + ...). Modules atadisk
288 and atapi* are not used and won't affect operation in ATA_CAM mode.
289 Note that to use CAM-based ATA kernel should include CAM devices
290 scbus, pass, da (or explicitly ada), cd and optionally others. All of
291 them are parts of the cam module.
293 ataraid(4) functionality is now supported by the RAID GEOM class.
294 To use it you can load geom_raid kernel module and use graid(8) tool
295 for management. Instead of /dev/arX device names, use /dev/raid/rX.
297 No kernel config options or code have been removed, so if a problem
298 arises, please report it and optionally revert to the old ATA stack.
299 In order to do it you can remove from the kernel config:
304 , and instead add back:
305 device atadisk # ATA disk drives
306 device ataraid # ATA RAID drives
307 device atapicd # ATAPI CDROM drives
308 device atapifd # ATAPI floppy drives
309 device atapist # ATAPI tape drives
312 The default NFS server has been changed to the new server, which
313 was referred to as the experimental server. If you need to switch
314 back to the old NFS server, you must now put the "-o" option on
315 both the mountd and nfsd commands. This can be done using the
316 mountd_flags and nfs_server_flags rc.conf variables until an
317 update to the rc scripts is committed, which is coming soon.
320 The GNU Objective-C runtime library (libobjc), and other Objective-C
321 related components have been removed from the base system. If you
322 require an Objective-C library, please use one of the available ports.
325 ath(4) has been split into bus- and device- modules. if_ath contains
326 the HAL, the TX rate control and the network device code. if_ath_pci
327 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
328 contains the AHB glue. Users need to load both if_ath_pci and if_ath
329 in order to use ath on everything else.
331 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
332 need to load if_ath and if_ath_pci for ath(4) operation.
335 As part of the replacement of sysinstall, the process of building
336 release media has changed significantly. For details, please re-read
337 release(7), which has been updated to reflect the new build process.
340 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
341 is the last available version under GPLv2. It brings a number of new
342 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
343 4.1 and SSE 4.2), better support for powerpc64, a number of new
344 directives, and lots of other small improvements. See the ChangeLog
345 file in contrib/binutils for the full details.
348 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
349 compliant, and will now use half of hash for authentication.
350 This will break interoperability with all stacks (including all
351 actual FreeBSD versions) who implement
352 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
354 The only workaround with such peers is to use another HMAC
355 algorithm for IPsec ("phase 2") authentication.
358 Remove the uio_yield prototype and symbol. This function has
359 been misnamed since it was introduced and should not be
360 globally exposed with this name. The equivalent functionality
361 is now available using kern_yield(curthread->td_user_pri).
362 The function remains undocumented.
365 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
366 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
367 for scalar sysctls is defined but disabled. Code that needs
368 UQUAD to pass the type checking that must compile on older
369 systems where the define is not present can check against
370 __FreeBSD_version >= 900030.
372 The system dialog(1) has been replaced with a new version previously
373 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
374 with the previous version, but the libdialog associated with it has
375 a largely incompatible API. As such, the original version of libdialog
376 will be kept temporarily as libodialog, until its base system consumers
377 are replaced or updated. Bump __FreeBSD_version to 900030.
380 If you are trying to run make universe on a -stable system, and you get
381 the following warning:
382 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
383 unknown. config(8) likely too old."
384 or something similar to it, then you must upgrade your -stable system
385 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
386 stable/8) or build the config from the latest stable/8 branch and
387 install it on your system.
389 Prior to this date, building a current universe on 8-stable system from
390 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
391 error in the first kernel build phase. A new config on those old
392 systems will fix that problem for older versions of -current.
395 The TCP stack has been modified to allow Khelp modules to interact with
396 it via helper hook points and store per-connection data in the TCP
397 control block. Bump __FreeBSD_version to 900029. User space tools that
398 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
402 Generic IEEE 802.3 annex 31B full duplex flow control support has been
403 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
404 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
405 to take advantage of it instead of using custom implementations. This
406 means that these drivers now no longer unconditionally advertise
407 support for flow control but only do so if flow control is a selected
408 media option. This was implemented in the generic support that way in
409 order to allow flow control to be switched on and off via ifconfig(8)
410 with the PHY specific default to typically off in order to protect
411 from unwanted effects. Consequently, if you used flow control with
412 one of the above mentioned drivers you now need to explicitly enable
414 ifconfig bge0 media auto mediaopt flowcontrol
416 Along with the above mentioned changes generic support for setting
417 1000baseT master mode also has been added and brgphy(4), ciphy(4),
418 e1000phy(4) as well as ip1000phy(4) have been converted to take
419 advantage of it. This means that these drivers now no longer take the
420 link0 parameter for selecting master mode but the master media option
421 has to be used instead, for example like in the following:
422 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
424 Selection of master mode now is also available with all other PHY
425 drivers supporting 1000baseT.
428 The TCP stack has received a significant update to add support for
429 modularised congestion control and generally improve the clarity of
430 congestion control decisions. Bump __FreeBSD_version to 900025. User
431 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
432 sockstat) need to be recompiled.
435 The man(1) utility has been replaced by a new version that no longer
436 uses /etc/manpath.config. Please consult man.conf(5) for how to
437 migrate local entries to the new format.
440 The copyright strings printed by login(1) and sshd(8) at the time of a
441 new connection have been removed to follow other operating systems and
445 A workaround for a fixed ld bug has been removed in kernel code,
446 so make sure that your system ld is built from sources after
447 revision 210245 from 2010-07-19 (r211583 if building head kernel
448 on stable/8, r211584 for stable/7; both from 2010-08-21).
449 A symptom of incorrect ld version is different addresses for
450 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
453 The $ipv6_prefer variable in rc.conf(5) has been split into
454 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
456 The $ip6addrctl_policy is a variable to choose a pre-defined
457 address selection policy set by ip6addrctl(8). A value
458 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
461 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
462 flag (see an entry of 20090926) is set on an interface with no
463 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
464 security reason. If you want IPv6 link-local address on all
465 interfaces by default, set this to "YES".
467 The old ipv6_prefer="YES" is equivalent to
468 ipv6_activate_all_interfaces="YES" and
469 ip6addrctl_policy="ipv6_prefer".
472 DTrace has grown support for userland tracing. Due to this, DTrace is
473 now i386 and amd64 only.
474 dtruss(1) is now installed by default on those systems and a new
475 kernel module is needed for userland tracing: fasttrap.
476 No changes to your kernel config file are necessary to enable
477 userland tracing, but you might consider adding 'STRIP=' and
478 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
479 to have informative userland stack traces in DTrace (ustack).
482 The acpi_aiboost(4) driver has been removed in favor of the new
483 aibs(4) driver. You should update your kernel configuration file.
486 BSD grep has been imported to the base system and it is built by
487 default. It is completely BSD licensed, highly GNU-compatible, uses
488 less memory than its GNU counterpart and has a small codebase.
489 However, it is slower than its GNU counterpart, which is mostly
490 noticeable for larger searches, for smaller ones it is measurable
491 but not significant. The reason is complex, the most important factor
492 is that we lack a modern and efficient regex library and GNU
493 overcomes this by optimizing the searches internally. Future work
494 on improving the regex performance is planned, for the meantime,
495 users that need better performance, can build GNU grep instead by
496 setting the WITH_GNU_GREP knob.
499 Due to the import of powerpc64 support, all existing powerpc kernel
500 configuration files must be updated with a machine directive like this:
501 machine powerpc powerpc
503 In addition, an updated config(8) is required to build powerpc kernels
507 A new version of ZFS (version 15) has been merged to -HEAD.
508 This version uses a python library for the following subcommands:
509 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
510 For full functionality of these commands the following port must
511 be installed: sysutils/py-zfs
514 'vm_page's are now hashed by physical address to an array of mutexes.
515 Currently this is only used to serialize access to hold_count. Over
516 time the page queue mutex will be peeled away. This changes the size
517 of pmap on every architecture. And requires all callers of vm_page_hold
518 and vm_page_unhold to be updated.
521 WITH_CTF can now be specified in src.conf (not recommended, there
522 are some problems with static executables), make.conf (would also
523 affect ports which do not use GNU make and do not override the
524 compile targets) or in the kernel config (via "makeoptions
526 When WITH_CTF was specified there before this was silently ignored,
527 so make sure that WITH_CTF is not used in places which could lead
528 to unwanted behavior.
531 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
532 to allow 32-bit compatibility on non-x86 platforms. All kernel
533 configurations on amd64 and ia64 platforms using these options must
534 be modified accordingly.
537 The utmp user accounting database has been replaced with utmpx,
538 the user accounting interface standardized by POSIX.
539 Unfortunately the semantics of utmp and utmpx don't match,
540 making it practically impossible to support both interfaces.
541 The user accounting database is used by tools like finger(1),
542 last(1), talk(1), w(1) and ac(8).
544 All applications in the base system use utmpx. This means only
545 local binaries (e.g. from the ports tree) may still use these
546 utmp database files. These applications must be rebuilt to make
549 After the system has been upgraded, it is safe to remove the old
550 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
551 assuming their contents is of no importance anymore. Old wtmp
552 databases can only be used by last(1) and ac(8) after they have
553 been converted to the new format using wtmpcvt(1).
556 Introduce the kernel thread "deadlock resolver" (which can be enabled
557 via the DEADLKRES option, see NOTES for more details) and the
558 sleepq_type() function for sleepqueues.
561 The rc.firewall and rc.firewall6 were unified, and
562 rc.firewall6 and rc.d/ip6fw were removed.
563 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
564 variables are obsoleted. Instead, the following new rc
565 variables are added to rc.d/ipfw:
567 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
568 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
569 firewall_simple_onet_ipv6, firewall_trusted_ipv6
571 The meanings correspond to the relevant IPv4 variables.
577 The default terminal emulation for syscons(4) has been changed
578 from cons25 to xterm on all platforms except pc98. This means
579 that the /etc/ttys file needs to be updated to ensure correct
580 operation of applications on the console.
582 The terminal emulation style can be toggled per window by using
583 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
584 options can be used to change the compile-time default back to
587 To prevent graphical artifacts, make sure the TERM environment
588 variable is set to match the terminal emulation that is being
589 performed by syscons(4).
592 The layout of the structure ieee80211req_scan_result has changed.
593 Applications that require wireless scan results (e.g. ifconfig(8))
594 from net80211 need to be recompiled.
596 Applications such as wpa_supplicant(8) may require a full world
597 build without using NO_CLEAN in order to get synchronized with the
601 The iwn(4) driver has been updated to support the 5000 and 5150 series.
602 There's one kernel module for each firmware. Adding "device iwnfw"
603 to the kernel configuration file means including all three firmware
604 images inside the kernel. If you want to include just the one for
605 your wireless card, use the devices iwn4965fw, iwn5000fw or
609 The rc.d/network_ipv6, IPv6 configuration script has been integrated
610 into rc.d/netif. The changes are the following:
612 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
613 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
614 Note that both variables need the "inet6" keyword at the head.
616 Do not set $ipv6_network_interfaces manually if you do not
617 understand what you are doing. It is not needed in most cases.
619 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
622 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
623 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
625 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
626 all configured interfaces have "inet6 accept_rtadv" in the
627 $ifconfig_IF_ipv6. These are for backward compatibility.
629 3. A new variable $ipv6_prefer has been added. If NO, IPv6
630 functionality of interfaces with no corresponding
631 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
632 and the default address selection policy of ip6addrctl(8)
633 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
634 Note that if you want to configure IPv6 functionality on the
635 disabled interfaces after boot, first you need to clear the flag by
636 using ifconfig(8) like:
638 ifconfig em0 inet6 -ifdisabled
640 If YES, the default address selection policy is set as
643 The default value of $ipv6_prefer is NO.
645 4. If your system need to receive Router Advertisement messages,
646 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
647 scripts automatically invoke rtsol(8) when the interface becomes
648 UP. The Router Advertisement messages are used for SLAAC
649 (State-Less Address AutoConfiguration).
652 802.11s D3.03 support was committed. This is incompatible with the
653 previous code, which was based on D3.0.
656 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
657 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
658 control whether accepting Router Advertisement messages or not.
659 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
660 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
661 The ifconfig(8) utility now supports these flags.
664 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
665 mount(8) and -a option for df(1) to see them.
668 The old tunable hw.bus.devctl_disable has been superseded by
669 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
670 replaced by hw.bus.devctl_queue=0. The default for this new tunable
674 Remove the option STOP_NMI. The default action is now to use NMI only
675 for KDB via the newly introduced function stop_cpus_hard() and
676 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
679 The stable/8 branch created in subversion. This corresponds to the
680 RELENG_8 branch in CVS.
683 Bump the shared library version numbers for all libraries that do not
684 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
685 __FreeBSD_version to 800105.
688 Due to changes in the implementation of virtual network stack support,
689 all network-related kernel modules must be recompiled. As this change
690 breaks the ABI, bump __FreeBSD_version to 800104.
693 The TOE interface to the TCP syncache has been modified to remove
694 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
695 The cxgb driver is the only TOE consumer affected by this change, and
696 needs to be recompiled along with the kernel. As this change breaks
697 the ABI, bump __FreeBSD_version to 800103.
700 Padding has been added to struct tcpcb, sackhint and tcpstat in
701 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
702 maintaining the ABI. However, this change breaks the ABI, so bump
703 __FreeBSD_version to 800102. User space tools that rely on the size of
704 any of these structs (e.g. sockstat) need to be recompiled.
707 The NFS_LEGACYRPC option has been removed along with the old kernel
708 RPC implementation that this option selected. Kernel configurations
709 may need to be adjusted.
712 The network interface device nodes at /dev/net/<interface> have been
713 removed. All ioctl operations can be performed the normal way using
714 routing sockets. The kqueue functionality can generally be replaced
715 with routing sockets.
718 The documentation from the FreeBSD Documentation Project (Handbook,
719 FAQ, etc.) is now installed via packages by sysinstall(8) and under
720 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
723 The ABI of various structures related to the SYSV IPC API have been
724 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
725 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
729 Layout of struct vnet has changed as routing related variables were
730 moved to their own Vimage module. Modules need to be recompiled. Bump
731 __FreeBSD_version to 800099.
734 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
735 respectively. As long as no more than 16 groups per process are used,
736 no changes should be visible. When more than 16 groups are used, old
737 binaries may fail if they call getgroups() or getgrouplist() with
738 statically sized storage. Recompiling will work around this, but
739 applications should be modified to use dynamically allocated storage
740 for group arrays as POSIX.1-2008 does not cap an implementation's
741 number of supported groups at NGROUPS_MAX+1 as previous versions did.
743 NFS and portalfs mounts may also be affected as the list of groups is
744 truncated to 16. Users of NFS who use more than 16 groups, should
745 take care that negative group permissions are not used on the exported
746 file systems as they will not be reliable unless a GSSAPI based
747 authentication method is used.
750 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
751 option compiles in the support for adaptive spinning for lockmgrs
752 which want to enable it. The lockinit() function now accepts the flag
753 LK_ADAPTIVE in order to make the lock object subject to adaptive
754 spinning when both held in write and read mode.
757 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
758 changed. User applications that use this ioctl need to be rebuilt.
761 The layout of struct thread has changed. Kernel and modules need to
765 The layout of structs ifnet, domain, protosw and vnet_net has changed.
766 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
769 window(1) has been removed from the base system. It can now be
770 installed from ports. The port is called misc/window.
773 The way we are storing and accessing `routing table' entries has
774 changed. Programs reading the FIB, like netstat, need to be
778 A new netisr implementation has been added for FreeBSD 8. Network
779 file system modules, such as igmp, ipdivert, and others, should be
781 Bump __FreeBSD_version to 800096.
784 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
788 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
789 Bump __FreeBSD_version to 800094.
792 Add mnt_xflag field to 'struct mount'. File system modules need to be
794 Bump __FreeBSD_version to 800093.
797 The compiling option ADAPTIVE_SX has been retired while it has been
798 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
799 The KPI for sx_init_flags() changes as accepting flags:
800 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
801 been introduced in order to handle the reversed logic.
802 Bump __FreeBSD_version to 800092.
805 Add support for hierarchical jails. Remove global securelevel.
806 Bump __FreeBSD_version to 800091.
809 The layout of struct vnet_net has changed, therefore modules
811 Bump __FreeBSD_version to 800090.
814 The newly imported zic(8) produces a new format in the output. Please
815 run tzsetup(8) to install the newly created data to /etc/localtime.
818 The sysctl tree for the usb stack has renamed from hw.usb2.* to
819 hw.usb.* and is now consistent again with previous releases.
822 802.11 monitor mode support was revised and driver api's were changed.
823 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
824 of DLT_IEEE802_11. No user-visible data structures were changed but
825 applications that use DLT_IEEE802_11 may require changes.
826 Bump __FreeBSD_version to 800088.
829 The layout of the following structs has changed: sysctl_oid,
830 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
831 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
832 panics may be experienced. World rebuild is required for
833 correctly checking networking state from userland.
834 Bump __FreeBSD_version to 800085.
837 MLDv2 and Source-Specific Multicast (SSM) have been merged
838 to the IPv6 stack. VIMAGE hooks are in but not yet used.
839 The implementation of SSM within FreeBSD's IPv6 stack closely
840 follows the IPv4 implementation.
842 For kernel developers:
844 * The most important changes are that the ip6_output() and
845 ip6_input() paths no longer take the IN6_MULTI_LOCK,
846 and this lock has been downgraded to a non-recursive mutex.
848 * As with the changes to the IPv4 stack to support SSM, filtering
849 of inbound multicast traffic must now be performed by transport
850 protocols within the IPv6 stack. This does not apply to TCP and
851 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
853 * The KPIs used by IPv6 multicast are similar to those used by
854 the IPv4 stack, with the following differences:
855 * im6o_mc_filter() is analogous to imo_multicast_filter().
856 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
857 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
858 * IN6_LOOKUP_MULTI() has been deprecated and removed.
859 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
860 for MLDv1 have an additional 'timer' argument which is used to
861 jitter the initial membership report for the solicited-node
862 multicast membership on-link.
863 * This is not strictly needed for MLDv2, which already jitters
864 its report transmissions. However, the 'timer' argument is
865 preserved in case MLDv1 is active on the interface.
867 * The KAME linked-list based IPv6 membership implementation has
868 been refactored to use a vector similar to that used by the IPv4
870 Code which maintains a list of its own multicast memberships
871 internally, e.g. carp, has been updated to reflect the new
874 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
875 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
876 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
877 implementation constraint which needs to be addressed in HEAD.
879 For application developers:
881 * The changes are broadly similar to those made for the IPv4
884 * The use of IPv4 and IPv6 multicast socket options on the same
885 socket, using mapped addresses, HAS NOT been tested or supported.
887 * There are a number of issues with the implementation of various
888 IPv6 multicast APIs which need to be resolved in the API surface
889 before the implementation is fully compatible with KAME userland
890 use, and these are mostly to do with interface index treatment.
892 * The literature available discusses the use of either the delta / ASM
893 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
894 using setsourcefilter(3)/getsourcefilter(3). For more information
895 please refer to RFC 3768, 'Socket Interface Extensions for
896 Multicast Source Filters'.
898 * Applications which use the published RFC 3678 APIs should be fine.
900 For systems administrators:
902 * The mtest(8) utility has been refactored to support IPv6, in
903 addition to IPv4. Interface addresses are no longer accepted
904 as arguments, their names must be used instead. The utility
905 will map the interface name to its first IPv4 address as
906 returned by getifaddrs(3).
908 * The ifmcstat(8) utility has also been updated to print the MLDv2
909 endpoint state and source filter lists via sysctl(3).
911 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
912 loopback of IPv6 multicast datagrams by default; it defaults to 1
913 to preserve the existing behaviour. Disabling multicast loopback is
914 recommended for optimal system performance.
916 * The IPv6 MROUTING code has been changed to examine this sysctl
917 instead of attempting to perform a group lookup before looping
918 back forwarded datagrams.
920 Bump __FreeBSD_version to 800084.
923 Implement low-level Bluetooth HCI API.
924 Bump __FreeBSD_version to 800083.
927 The layout of struct malloc_type, used by modules to register new
928 memory allocation types, has changed. Most modules will need to
929 be rebuilt or panics may be experienced.
930 Bump __FreeBSD_version to 800081.
933 Anticipate overflowing inp_flags - add inp_flags2.
934 This changes most offsets in inpcb, so checking v4 connection
935 state will require a world rebuild.
936 Bump __FreeBSD_version to 800080.
939 Add an llentry to struct route and struct route_in6. Modules
940 embedding a struct route will need to be recompiled.
941 Bump __FreeBSD_version to 800079.
944 The size of rt_metrics_lite and by extension rtentry has changed.
945 Networking administration apps will need to be recompiled.
946 The route command now supports show as an alias for get, weighting
947 of routes, sticky and nostick flags to alter the behavior of stateful
949 Bump __FreeBSD_version to 800078.
952 Do not use Giant for kbdmux(4) locking. This is wrong and
953 apparently causing more problems than it solves. This will
954 re-open the issue where interrupt handlers may race with
955 kbdmux(4) in polling mode. Typical symptoms include (but
956 not limited to) duplicated and/or missing characters when
957 low level console functions (such as gets) are used while
958 interrupts are enabled (for example geli password prompt,
959 mountroot prompt etc.). Disabling kbdmux(4) may help.
962 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
963 kernel modules referencing any of the above need to be recompiled.
964 Bump __FreeBSD_version to 800075.
967 GEOM_PART has become the default partition slicer for storage devices,
968 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
969 introduces some changes:
971 MSDOS/EBR: the devices created from MSDOS extended partition entries
972 (EBR) can be named differently than with GEOM_MBR and are now symlinks
973 to devices with offset-based names. fstabs may need to be modified.
975 BSD: the "geometry does not match label" warning is harmless in most
976 cases but it points to problems in file system misalignment with
977 disk geometry. The "c" partition is now implicit, covers the whole
978 top-level drive and cannot be (mis)used by users.
980 General: Kernel dumps are now not allowed to be written to devices
981 whose partition types indicate they are meant to be used for file
982 systems (or, in case of MSDOS partitions, as something else than
985 Most of these changes date approximately from 200812.
988 The uscanner(4) driver has been removed from the kernel. This follows
989 Linux removing theirs in 2.6 and making libusb the default interface
993 The multicast forwarding code has been cleaned up. netstat(1)
994 only relies on KVM now for printing bandwidth upcall meters.
995 The IPv4 and IPv6 modules are split into ip_mroute_mod and
996 ip6_mroute_mod respectively. The config(5) options for statically
997 compiling this code remain the same, i.e. 'options MROUTING'.
1000 Support for the IFF_NEEDSGIANT network interface flag has been
1001 removed, which means that non-MPSAFE network device drivers are no
1002 longer supported. In particular, if_ar, if_sr, and network device
1003 drivers from the old (legacy) USB stack can no longer be built or
1007 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1008 a bunch of new language catalog files have also been added.
1009 This means that some common libc messages are now localized and
1010 they depend on the LC_MESSAGES environmental variable.
1013 The k8temp(4) driver has been renamed to amdtemp(4) since
1014 support for Family 10 and Family 11 CPU families was added.
1017 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1018 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1020 For kernel developers, the most important changes are that the
1021 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1022 and this lock has been downgraded to a non-recursive mutex.
1024 Transport protocols (UDP, Raw IP) are now responsible for filtering
1025 inbound multicast traffic according to group membership and source
1026 filters. The imo_multicast_filter() KPI exists for this purpose.
1027 Transports which do not use multicast (SCTP, TCP) already reject
1028 multicast by default. Forwarding and receive performance may improve
1029 as a mutex acquisition is no longer needed in the ip_input()
1030 low-level input path. in_addmulti() and in_delmulti() are shimmed
1031 to new KPIs which exist to support SSM in-kernel.
1033 For application developers, it is recommended that loopback of
1034 multicast datagrams be disabled for best performance, as this
1035 will still cause the lock to be taken for each looped-back
1036 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1037 be tuned to 0 to disable loopback by default; it defaults to 1
1038 to preserve the existing behaviour.
1040 For systems administrators, to obtain best performance with
1041 multicast reception and multiple groups, it is always recommended
1042 that a card with a suitably precise hash filter is used. Hash
1043 collisions will still result in the lock being taken within the
1044 transport protocol input path to check group membership.
1046 If deploying FreeBSD in an environment with IGMP snooping switches,
1047 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1048 enabled; this forces 224.0.0.0/24 group membership to be announced
1051 The size of 'struct igmpstat' has changed; netstat needs to be
1052 recompiled to reflect this.
1053 Bump __FreeBSD_version to 800070.
1056 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1057 updated to use it. This requires a buildworld/installworld in order to
1058 update the library and dependencies (usbconfig, etc). Its advisable to
1059 rebuild all ports which uses libusb. More specific directions are given
1060 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1061 libusb are no longer required and can be removed.
1064 A workaround is committed to allow the creation of System V shared
1065 memory segment of size > 2 GB on the 64-bit architectures.
1066 Due to a limitation of the existing ABI, the shm_segsz member
1067 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1068 wrong for large segments. Note that limits must be explicitly
1069 raised to allow such segments to be created.
1072 The layout of struct ifnet has changed, requiring a rebuild of all
1073 network device driver modules.
1076 The /dev handling for the new USB stack has changed, a
1077 buildworld/installworld is required for libusb20.
1080 The new USB2 stack has now been permanently moved in and all kernel and
1081 module names reverted to their previous values (eg, usb, ehci, ohci,
1082 ums, ...). The old usb stack can be compiled in by prefixing the name
1083 with the letter 'o', the old usb modules have been removed.
1084 Updating entry 20090216 for xorg and 20090215 for libmap may still
1088 The rc.conf(5) option if_up_delay has been renamed to
1089 defaultroute_delay to better reflect its purpose. If you have
1090 customized this setting in /etc/rc.conf you need to update it to
1094 xorg 7.4 wants to configure its input devices via hald which does not
1095 yet work with USB2. If the keyboard/mouse does not work in xorg then
1097 Option "AllowEmptyInput" "off"
1098 to your ServerLayout section. This will cause X to use the configured
1099 kbd and mouse sections from your xorg.conf.
1102 The GENERIC kernels for all architectures now default to the new USB2
1103 stack. No kernel config options or code have been removed so if a
1104 problem arises please report it and optionally revert to the old USB
1105 stack. If you are loading USB kernel modules or have a custom kernel
1106 that includes GENERIC then ensure that usb names are also changed over,
1107 eg uftdi -> usb2_serial_ftdi.
1109 Older programs linked against the ports libusb 0.1 need to be
1110 redirected to the new stack's libusb20. /etc/libmap.conf can
1112 # Map old usb library to new one for usb2 stack
1113 libusb-0.1.so.8 libusb20.so.1
1116 All USB ethernet devices now attach as interfaces under the name ueN
1117 (eg. ue0). This is to provide a predictable name as vendors often
1118 change usb chipsets in a product without notice.
1121 The ichsmb(4) driver has been changed to require SMBus slave
1122 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1123 All of the other SMBus controller drivers require left-justified
1124 slave addresses, so this change makes all the drivers provide the
1128 INET6 statistics (struct ip6stat) was updated.
1129 netstat(1) needs to be recompiled.
1132 NTFS has been removed from GENERIC kernel on amd64 to match
1133 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1134 will load ntfs.ko module automatically when NTFS support is
1135 actually needed, unless ntfs.ko is not installed or security
1136 level prohibits loading kernel modules. If either is the case,
1137 "options NTFS" has to be added into kernel config.
1140 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1141 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1142 800061. User space tools that rely on the size of struct tcpcb in
1143 tcp_var.h (e.g. sockstat) need to be recompiled.
1146 ng_tty(4) module updated to match the new TTY subsystem.
1147 Due to API change, user-level applications must be updated.
1148 New API support added to mpd5 CVS and expected to be present
1149 in next mpd5.3 release.
1152 With __FreeBSD_version 800060 the makefs tool is part of
1153 the base system (it was a port).
1156 The afdata and ifnet locks have been changed from mutexes to
1157 rwlocks, network modules will need to be re-compiled.
1160 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1161 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1162 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1163 architecture and 40 bytes on 64-bit architecture. The userland
1164 applications "arp" and "ndp" have been updated accordingly.
1165 The output from "netstat -r" shows only routing entries and
1166 none of the L2 information.
1169 __FreeBSD_version 800057 marks the switchover from the
1170 binary ath hal to source code. Users must add the line:
1172 options AH_SUPPORT_AR5416
1174 to their kernel config files when specifying:
1178 The ath_hal module no longer exists; the code is now compiled
1179 together with the driver in the ath module. It is now
1180 possible to tailor chip support (i.e. reduce the set of chips
1181 and thereby the code size); consult ath_hal(4) for details.
1184 __FreeBSD_version 800054 adds memory barriers to
1185 <machine/atomic.h>, new interfaces to ifnet to facilitate
1186 multiple hardware transmit queues for cards that support
1187 them, and a lock-less ring-buffer implementation to
1188 enable drivers to more efficiently manage queueing of
1192 A new version of ZFS (version 13) has been merged to -HEAD.
1193 This version has zpool attribute "listsnapshots" off by
1194 default, which means "zfs list" does not show snapshots,
1195 and is the same as Solaris behavior.
1198 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1201 The uhci, ohci, ehci and slhci USB Host controller drivers have
1202 been put into separate modules. If you load the usb module
1203 separately through loader.conf you will need to load the
1204 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1205 controller add the following to loader.conf:
1211 The ABI used by the PMC toolset has changed. Please keep
1212 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1216 atapci kernel module now includes only generic PCI ATA
1217 driver. AHCI driver moved to ataahci kernel module.
1218 All vendor-specific code moved into separate kernel modules:
1219 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1220 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1221 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1222 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1225 The TTY subsystem of the kernel has been replaced by a new
1226 implementation, which provides better scalability and an
1227 improved driver model. Most common drivers have been migrated to
1228 the new TTY subsystem, while others have not. The following
1229 drivers have not yet been ported to the new TTY layer:
1232 cy, digi, rc, rp, sio
1238 ng_h4, ng_tty, ppp, sl, snp
1240 Adding these drivers to your kernel configuration file shall
1241 cause compilation to fail.
1244 ntpd has been upgraded to 4.2.4p5.
1247 OpenSSH has been upgraded to 5.1p1.
1249 For many years, FreeBSD's version of OpenSSH preferred DSA
1250 over RSA for host and user authentication keys. With this
1251 upgrade, we've switched to the vendor's default of RSA over
1252 DSA. This may cause upgraded clients to warn about unknown
1253 host keys even for previously known hosts. Users should
1254 follow the usual procedure for verifying host keys before
1255 accepting the RSA key.
1257 This can be circumvented by setting the "HostKeyAlgorithms"
1258 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1261 Please note that the sequence of keys offered for
1262 authentication has been changed as well. You may want to
1263 specify IdentityFile in a different order to revert this
1267 The sio(4) driver has been removed from the i386 and amd64
1268 kernel configuration files. This means uart(4) is now the
1269 default serial port driver on those platforms as well.
1271 To prevent collisions with the sio(4) driver, the uart(4) driver
1272 uses different names for its device nodes. This means the
1273 onboard serial port will now most likely be called "ttyu0"
1274 instead of "ttyd0". You may need to reconfigure applications to
1275 use the new device names.
1277 When using the serial port as a boot console, be sure to update
1278 /boot/device.hints and /etc/ttys before booting the new kernel.
1279 If you forget to do so, you can still manually specify the hints
1280 at the loader prompt:
1282 set hint.uart.0.at="isa"
1283 set hint.uart.0.port="0x3F8"
1284 set hint.uart.0.flags="0x10"
1285 set hint.uart.0.irq="4"
1289 The gpt(8) utility has been removed. Use gpart(8) to partition
1293 The version that Linuxulator emulates was changed from 2.4.2
1294 to 2.6.16. If you experience any problems with Linux binaries
1295 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1296 if it fixes the problem contact emulation mailing list.
1299 ISDN4BSD (I4B) was removed from the src tree. You may need to
1300 update a your kernel configuration and remove relevant entries.
1303 I have checked in code to support multiple routing tables.
1304 See the man pages setfib(1) and setfib(2).
1305 This is a hopefully backwards compatible version,
1306 but to make use of it you need to compile your kernel
1307 with options ROUTETABLES=2 (or more up to 16).
1310 The 802.11 wireless support was redone to enable multi-bss
1311 operation on devices that are capable. The underlying device
1312 is no longer used directly but instead wlanX devices are
1313 cloned with ifconfig. This requires changes to rc.conf files.
1314 For example, change:
1315 ifconfig_ath0="WPA DHCP"
1318 ifconfig_wlan0="WPA DHCP"
1319 see rc.conf(5) for more details. In addition, mergemaster of
1320 /etc/rc.d is highly recommended. Simultaneous update of userland
1321 and kernel wouldn't hurt either.
1323 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1324 modules were merged into the base wlan module. All references
1325 to these modules (e.g. in kernel config files) must be removed.
1328 psm(4) has gained write(2) support in native operation level.
1329 Arbitrary commands can be written to /dev/psm%d and status can
1330 be read back from it. Therefore, an application is responsible
1331 for status validation and error recovery. It is a no-op in
1332 other operation levels.
1335 Support for KSE threading has been removed from the kernel. To
1336 run legacy applications linked against KSE libmap.conf may
1337 be used. The following libmap.conf may be used to ensure
1338 compatibility with any prior release:
1340 libpthread.so.1 libthr.so.1
1341 libpthread.so.2 libthr.so.2
1342 libkse.so.3 libthr.so.3
1345 The layout of struct vmspace has changed. This affects libkvm
1346 and any executables that link against libkvm and use the
1347 kvm_getprocs() function. In particular, but not exclusively,
1348 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1349 The effects are minimal, but it's advisable to upgrade world
1353 The latest em driver no longer has support in it for the
1354 82575 adapter, this is now moved to the igb driver. The
1355 split was done to make new features that are incompatible
1356 with older hardware easier to do.
1359 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1360 likewise the kernel option is now GEOM_LINUX_LVM.
1363 The default NFS mount mode has changed from UDP to TCP for
1364 increased reliability. If you rely on (insecurely) NFS
1365 mounting across a firewall you may need to update your
1369 Belatedly note the addition of m_collapse for compacting
1373 The fts(3) structures have been changed to use adequate
1374 integer types for their members and so to be able to cope
1375 with huge file trees. The old fts(3) ABI is preserved
1376 through symbol versioning in libc, so third-party binaries
1377 using fts(3) should still work, although they will not take
1378 advantage of the extended types. At the same time, some
1379 third-party software might fail to build after this change
1380 due to unportable assumptions made in its source code about
1381 fts(3) structure members. Such software should be fixed
1382 by its vendor or, in the worst case, in the ports tree.
1383 FreeBSD_version 800015 marks this change for the unlikely
1384 case that a portable fix is impossible.
1387 To upgrade to -current after this date, you must be running
1388 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1389 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1392 The ADAPTIVE_GIANT kernel option has been retired because its
1393 functionality is the default now.
1396 The AT keyboard emulation of sunkbd(4) has been turned on
1397 by default. In order to make the special symbols of the Sun
1398 keyboards driven by sunkbd(4) work under X these now have
1399 to be configured the same way as Sun USB keyboards driven
1400 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1402 Option "XkbLayout" "us"
1403 Option "XkbRules" "xorg"
1404 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1407 It has been decided that it is desirable to provide ABI
1408 backwards compatibility to the FreeBSD 4/5/6 versions of the
1409 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1410 broken with the introduction of PCI domain support (see the
1411 20070930 entry). Unfortunately, this required the ABI of
1412 PCIOCGETCONF to be broken again in order to be able to
1413 provide backwards compatibility to the old version of that
1414 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1415 again. As for prominent ports this affects neither pciutils
1416 nor xorg-server this time, the hal port needs to be rebuilt
1420 The misnamed kthread_create() and friends have been renamed
1421 to kproc_create() etc. Many of the callers already
1422 used kproc_start()..
1423 I will return kthread_create() and friends in a while
1424 with implementations that actually create threads, not procs.
1425 Renaming corresponds with version 800002.
1434 Avoid using make -j when upgrading. While generally safe, there are
1435 sometimes problems using -j to upgrade. If your upgrade fails with
1436 -j, please try again without -j. From time to time in the past there
1437 have been problems using -j with buildworld and/or installworld. This
1438 is especially true when upgrading between "distant" versions (eg one
1439 that cross a major release boundary or several minor releases, or when
1440 several months have passed on the -current branch).
1442 Sometimes, obscure build problems are the result of environment
1443 poisoning. This can happen because the make utility reads its
1444 environment when searching for values for global variables. To run
1445 your build attempts in an "environmental clean room", prefix all make
1446 commands with 'env -i '. See the env(1) manual page for more details.
1448 When upgrading from one major version to another it is generally best
1449 to upgrade to the latest code in the currently installed branch first,
1450 then do an upgrade to the new branch. This is the best-tested upgrade
1451 path, and has the highest probability of being successful. Please try
1452 this approach before reporting problems with a major version upgrade.
1456 When upgrading the boot ZFS pool to a new version, always follow
1459 1.) recompile and reinstall the ZFS boot loader and boot block
1460 (this is part of "make buildworld" and "make installworld")
1462 2.) update the ZFS boot block on your boot drive
1464 The following example updates the ZFS boot block on the first
1465 partition (freebsd-boot) of a GPT partitioned drive ad0:
1466 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1468 Non-boot pools do not need these updates.
1472 If you are updating from a prior version of FreeBSD (even one just
1473 a few days old), you should follow this procedure. It is the most
1474 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1476 make kernel-toolchain
1477 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1478 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1480 To test a kernel once
1481 ---------------------
1482 If you just want to boot a kernel once (because you are not sure
1483 if it works, or if you want to boot a known bad kernel to provide
1484 debugging information) run
1485 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1486 nextboot -k testkernel
1488 To just build a kernel when you know that it won't mess you up
1489 --------------------------------------------------------------
1490 This assumes you are already running a CURRENT system. Replace
1491 ${arch} with the architecture of your machine (e.g. "i386",
1492 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1494 cd src/sys/${arch}/conf
1495 config KERNEL_NAME_HERE
1496 cd ../compile/KERNEL_NAME_HERE
1501 If this fails, go to the "To build a kernel" section.
1503 To rebuild everything and install it on the current system.
1504 -----------------------------------------------------------
1505 # Note: sometimes if you are running current you gotta do more than
1506 # is listed here if you are upgrading from a really old current.
1508 <make sure you have good level 0 dumps>
1510 make kernel KERNCONF=YOUR_KERNEL_HERE
1512 <reboot in single user> [3]
1520 To cross-install current onto a separate partition
1521 --------------------------------------------------
1522 # In this approach we use a separate partition to hold
1523 # current's root, 'usr', and 'var' directories. A partition
1524 # holding "/", "/usr" and "/var" should be about 2GB in
1527 <make sure you have good level 0 dumps>
1530 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1531 <maybe newfs current's root partition>
1532 <mount current's root partition on directory ${CURRENT_ROOT}>
1533 make installworld DESTDIR=${CURRENT_ROOT}
1534 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1535 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1536 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1537 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1538 <reboot into current>
1539 <do a "native" rebuild/install as described in the previous section>
1540 <maybe install compatibility libraries from ports/misc/compat*>
1544 To upgrade in-place from stable to current
1545 ----------------------------------------------
1546 <make sure you have good level 0 dumps>
1548 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1550 <reboot in single user> [3]
1557 Make sure that you've read the UPDATING file to understand the
1558 tweaks to various things you need. At this point in the life
1559 cycle of current, things change often and you are on your own
1560 to cope. The defaults can also change, so please read ALL of
1561 the UPDATING entries.
1563 Also, if you are tracking -current, you must be subscribed to
1564 freebsd-current@freebsd.org. Make sure that before you update
1565 your sources that you have read and understood all the recent
1566 messages there. If in doubt, please track -stable which has
1567 much fewer pitfalls.
1569 [1] If you have third party modules, such as vmware, you
1570 should disable them at this point so they don't crash your
1573 [3] From the bootblocks, boot -s, and then do
1578 adjkerntz -i # if CMOS is wall time
1579 Also, when doing a major release upgrade, it is required that
1580 you boot into single user mode to do the installworld.
1582 [4] Note: This step is non-optional. Failure to do this step
1583 can result in a significant reduction in the functionality of the
1584 system. Attempting to do it by hand is not recommended and those
1585 that pursue this avenue should read this file carefully, as well
1586 as the archives of freebsd-current and freebsd-hackers mailing lists
1587 for potential gotchas. The -U option is also useful to consider.
1588 See mergemaster(8) for more information.
1590 [5] Usually this step is a noop. However, from time to time
1591 you may need to do this if you get unknown user in the following
1592 step. It never hurts to do it all the time. You may need to
1593 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1594 install) after the buildworld before this step if you last updated
1595 from [78]-stable or 9-stable before 20130430.
1597 [6] This only deletes old files and directories. Old libraries
1598 can be deleted by "make delete-old-libs", but you have to make
1599 sure that no program is using those libraries anymore.
1601 [8] In order to have a kernel that can run the 4.x binaries needed to
1602 do an installworld, you must include the COMPAT_FREEBSD4 option in
1603 your kernel. Failure to do so may leave you with a system that is
1604 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1605 required to run the 5.x binaries on more recent kernels. And so on
1606 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1608 Make sure that you merge any new devices from GENERIC since the
1609 last time you updated your kernel config file.
1611 [9] When checking out sources, you must include the -P flag to have
1612 cvs prune empty directories.
1614 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1615 "?=" instead of the "=" assignment operator, so that buildworld can
1616 override the CPUTYPE if it needs to.
1618 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1619 not on the command line, or in /etc/make.conf. buildworld will
1620 warn if it is improperly defined.
1623 This file contains a list, in reverse chronological order, of major
1624 breakages in tracking -current. It is not guaranteed to be a complete
1625 list of such breakages, and only contains entries since October 10, 2007.
1626 If you need to see UPDATING entries from before that date, you will need
1627 to fetch an UPDATING file from an older FreeBSD release.
1629 Copyright information:
1631 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1633 Redistribution, publication, translation and use, with or without
1634 modification, in full or in part, in any form or format of this
1635 document are permitted without further permission from the author.
1637 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1638 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1639 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1640 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1641 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1642 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1643 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1644 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1645 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1646 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1647 POSSIBILITY OF SUCH DAMAGE.
1649 Contact Warner Losh if you have any questions about your use of