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.
14 20140408: p4 FreeBSD-SA-14:05.nfsserver
15 FreeBSD-SA-14:06.openssl
16 Fix deadlock in the NFS server. [SA-14:05]
18 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
20 20140114: p3 FreeBSD-SA-14:01.bsnmpd
23 FreeBSD-EN-14:01.random
25 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
27 Fix ntpd distributed reflection Denial of Service
28 vulnerability. [SA-14:02]
30 Fix BIND remote denial of service vulnerability. [SA-14:04]
32 Disable hardware RNGs by default. [EN-14:01]
34 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
36 20131128: p2 FreeBSD-EN-13:05.freebsd-update
37 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
39 20131026: p1 FreeBSD-EN-13:04.freebsd-update
40 Fix multiple freebsd-update bugs that break upgrading to
44 hastctl(8)'s `status' command output changed to terse one-liner format.
45 Scripts using this should switch to `list' command or be rewritten.
48 Fix a bug that allowed a tracing process (e.g. gdb) to write
49 to a memory-mapped file in the traced process's address space
50 even if neither the traced process nor the tracing process had
51 write access to that file.
54 Added ZFS TRIM support which is enabled by default. To disable
55 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
57 Creating new ZFS pools and adding new devices to existing pools
58 first performs a full device level TRIM which can take a significant
59 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
60 to disable this behaviour.
62 ZFS TRIM requires the underlying device support BIO_DELETE which
63 is currently provided by methods such as ATA TRIM and SCSI UNMAP
64 via CAM, which are typically supported by SSD's.
66 Stats for ZFS TRIM can be monitored by looking at the sysctl's
67 under kstat.zfs.misc.zio_trim.
70 `list' command has been added to hastctl(8). For now, it is full
71 equivalent of `status' command.
72 WARNING: in the near future the output of hastctl's status command
73 will change to more terse format. If you use `hastctl status'
74 for parsing in your scripts, switch to `hastctl list'.
77 The mergemaster command now uses the default MAKEOBJDIRPREFIX
78 rather than creating it's own in the temporary directory in
79 order allow access to bootstrapped versions of tools such as
80 install and mtree. When upgrading from version of FreeBSD where
81 the install command does not support -l, you will need to
82 install a new mergemaster command if mergemaster -p is required.
83 This can be accomplished with the command (cd src/usr.sbin/mergemaster
86 Due to the use of the new -l option to install(1) during build
87 and install, you must take care not to directly set the INSTALL
88 make variable in your /etc/make.conf, /etc/src.conf, or on the
89 command line. If you wish to use the -C flag for all installs
90 you may be able to add INSTALL+=-C to /etc/make.conf or
94 Fix a bug that allows NFS clients to issue READDIR on files.
97 The install(1) option -M has changed meaning and now takes an
98 argument that is a file or path to append logs to. In the
99 unlikely event that -M was the last option on the command line
100 and the command line contained at least two files and a target
101 directory the first file will have logs appended to it. The -M
102 option served little practical purpose in the last decade so it's
103 used expected to be extremely rare.
106 A new compression method (lz4) has been merged to. Please refer to
107 zpool-features(7) for more information.
109 Please refer to the "ZFS notes" section of this file for information
110 on upgrading boot ZFS pools.
113 The VFS KBI was changed with the merge of several nullfs
114 optimizations and fixes. All filesystem modules must be
118 With the addition of auditdistd(8), a new auditdistd user is now
119 depended on during installworld. "mergemaster -p" can be used to add
120 the user prior to installworld, as documented in the handbook.
126 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
127 Starting with this version the old system of ZFS pool versioning
128 is superseded by "feature flags". This concept enables forward
129 compatibility against certain future changes in functionality of ZFS
130 pools. The first two read-only compatible "feature flags" for ZFS
131 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
132 For more information read the new zpool-features(7) manual page.
133 Please refer to the "ZFS notes" section of this file for information
134 on upgrading boot ZFS pools.
137 The commit introducing bsd.compiler.mk breaks the traditional
138 building of kernels before this point. Add -m ${SRC}/share/mk
139 (for the right value of SRC) to your command lines to work
140 around; update your useland to a point after this; or use the
141 buildkernel/installkernel top-level targets. See also 20120829.
144 The IPFIREWALL_FORWARD kernel option has been removed. Its
145 functionality now turned on by default.
148 The random(4) support for the VIA hardware random number
149 generator (`PADLOCK') is no longer enabled unconditionally.
150 Add the PADLOCK_RNG option in the custom kernel config if
151 needed. The GENERIC kernels on i386 and amd64 do include the
152 option, so the change only affects the custom kernel
156 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
157 the traditional method, you must update your system with an installworld
158 before the kernel will build. The documented make buildkernel/installkernel
159 interfaces (coupled with fresh make kernel-toolchain) continue to work.
162 The sparc64 ZFS loader has been changed to no longer try to auto-
163 detect ZFS providers based on diskN aliases but now requires these
164 to be explicitly listed in the OFW boot-device environment variable.
167 Now unix domain sockets behave "as expected" on nullfs(5). Previously
168 nullfs(5) did not pass through all behaviours to the underlying layer,
169 as a result if we bound to a socket on the lower layer we could connect
170 only to the lower path; if we bound to the upper layer we could connect
171 only to the upper path. The new behavior is one can connect to both the
172 lower and the upper paths regardless what layer path one binds to.
175 The acpi_wmi(4) status device /dev/wmistat has been renamed to
179 A new VOP_ADVISE() was added to support posix_fadvise(2). All
180 filesystem modules must be recompiled.
183 The interface of the VOP_VPTOCNP(9) changed, now the returned
184 vnode shall be referenced, previously it was required to be
185 only held. All in-tree filesystems are converted.
191 The broken amd(4) driver has been replaced with esp(4) in the amd64,
192 i386 and pc98 GENERIC kernel configuration files.
195 This commit modifies vfs_register() so that it uses a hash
196 calculation to set vfc_typenum, which is enabled by default.
197 The first time a system is booted after this change, the
198 vfc_typenum values will change for all file systems. The
199 main effect of this is a change to the NFS server file handles
200 for file systems that use vfc_typenum in their fsid, such as ZFS.
201 It will, however, prevent vfc_typenum from changing when file
202 systems are loaded in a different order for subsequent reboots.
203 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
204 until you are ready to remount all NFS clients after a reboot.
207 Bump the shared library version numbers for libraries that
208 do not use symbol versioning, have changed the ABI compared
209 to stable/8 and which shared library version was not bumped.
210 Done as part of 9.0-RELEASE cycle.
213 During the merge of Capsicum features, the fget(9) KPI was modified.
214 This may require the rebuilding of out-of-tree device drivers --
215 issues have been reported specifically with the nVidia device driver.
216 __FreeBSD_version is bumped to 900041.
218 Also, there is a period between 20110811 and 20110814 where the
219 special devices /dev/{stdin,stdout,stderr} did not work correctly.
220 Building world from a kernel during that window may not work.
223 The packet filter (pf) code has been updated to OpenBSD 4.5.
224 You need to update userland tools to be in sync with kernel.
225 This update breaks backward compatibility with earlier pfsync(4)
226 versions. Care must be taken when updating redundant firewall setups.
229 The following sysctls and tunables are retired on x86 platforms:
231 machdep.hlt_logical_cpus
232 The following sysctl is retired:
233 machdep.hyperthreading_allowed
234 The sysctls were supposed to provide a way to dynamically offline and
235 online selected CPUs on x86 platforms, but the implementation has not
236 been reliable especially with SCHED_ULE scheduler.
237 machdep.hyperthreading_allowed tunable is still available to ignore
238 hyperthreading CPUs at OS level.
239 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
240 where X is an APIC ID of a CPU. Be advised, though, that disabling
241 CPUs in non-uniform fashion will result in non-uniform topology and
242 may lead to sub-optimal system performance with SCHED_ULE, which is
246 cpumask_t type is retired and cpuset_t is used in order to describe
250 Changes to ifconfig(8) for dynamic address family detection mandate
251 that you are running a kernel of 20110525 or later. Make sure to
252 follow the update procedure to boot a new kernel before installing
256 Support for sun4v architecture is officially dropped
259 Several KPI breaking changes have been committed to the mii(4) layer,
260 the PHY drivers and consequently some Ethernet drivers using mii(4).
261 This means that miibus.ko and the modules of the affected Ethernet
262 drivers need to be recompiled.
264 Note to kernel developers: Given that the OUI bit reversion problem
265 was fixed as part of these changes all mii(4) commits related to OUIs,
266 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
267 handling, no longer can be merged verbatim to stable/8 and previous
271 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
272 into their kernel configurations along with 'device pci'.
275 The default NFS client is now the new NFS client, so fstype "newnfs"
276 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
277 Although mounts via fstype "nfs" will usually work without userland
278 changes, it is recommended that the mount(8) and mount_nfs(8)
279 commands be rebuilt from sources and that a link to mount_nfs called
280 mount_oldnfs be created. The new client is compiled into the
281 kernel with "options NFSCL" and this is needed for diskless root
282 file systems. The GENERIC kernel configs have been changed to use
283 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
284 To use the regular/old client, you can "mount -t oldnfs ...". For
285 a diskless root file system, you must also include a line like:
287 vfs.root.mountfrom="oldnfs:"
289 in the boot/loader.conf on the root fs on the NFS server to make
290 a diskless root fs use the old client.
293 The GENERIC kernels for all architectures now default to the new
294 CAM-based ATA stack. It means that all legacy ATA drivers were
295 removed and replaced by respective CAM drivers. If you are using
296 ATA device names in /etc/fstab or other places, make sure to update
297 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
298 where 'Y's are the sequential numbers starting from zero for each type
299 in order of detection, unless configured otherwise with tunables,
300 see cam(4)). There will be symbolic links created in /dev/ to map
301 old adX devices to the respective adaY. They should provide basic
302 compatibility for file systems mounting in most cases, but they do
303 not support old user-level APIs and do not have respective providers
304 in GEOM. Consider using updated management tools with new device names.
306 It is possible to load devices ahci, ata, siis and mvs as modules,
307 but option ATA_CAM should remain in kernel configuration to make ata
308 module work as CAM driver supporting legacy ATA controllers. Device ata
309 still can be used in modular fashion (atacore + ...). Modules atadisk
310 and atapi* are not used and won't affect operation in ATA_CAM mode.
311 Note that to use CAM-based ATA kernel should include CAM devices
312 scbus, pass, da (or explicitly ada), cd and optionally others. All of
313 them are parts of the cam module.
315 ataraid(4) functionality is now supported by the RAID GEOM class.
316 To use it you can load geom_raid kernel module and use graid(8) tool
317 for management. Instead of /dev/arX device names, use /dev/raid/rX.
319 No kernel config options or code have been removed, so if a problem
320 arises, please report it and optionally revert to the old ATA stack.
321 In order to do it you can remove from the kernel config:
326 , and instead add back:
327 device atadisk # ATA disk drives
328 device ataraid # ATA RAID drives
329 device atapicd # ATAPI CDROM drives
330 device atapifd # ATAPI floppy drives
331 device atapist # ATAPI tape drives
334 The default NFS server has been changed to the new server, which
335 was referred to as the experimental server. If you need to switch
336 back to the old NFS server, you must now put the "-o" option on
337 both the mountd and nfsd commands. This can be done using the
338 mountd_flags and nfs_server_flags rc.conf variables until an
339 update to the rc scripts is committed, which is coming soon.
342 The GNU Objective-C runtime library (libobjc), and other Objective-C
343 related components have been removed from the base system. If you
344 require an Objective-C library, please use one of the available ports.
347 ath(4) has been split into bus- and device- modules. if_ath contains
348 the HAL, the TX rate control and the network device code. if_ath_pci
349 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
350 contains the AHB glue. Users need to load both if_ath_pci and if_ath
351 in order to use ath on everything else.
353 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
354 need to load if_ath and if_ath_pci for ath(4) operation.
357 As part of the replacement of sysinstall, the process of building
358 release media has changed significantly. For details, please re-read
359 release(7), which has been updated to reflect the new build process.
362 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
363 is the last available version under GPLv2. It brings a number of new
364 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
365 4.1 and SSE 4.2), better support for powerpc64, a number of new
366 directives, and lots of other small improvements. See the ChangeLog
367 file in contrib/binutils for the full details.
370 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
371 compliant, and will now use half of hash for authentication.
372 This will break interoperability with all stacks (including all
373 actual FreeBSD versions) who implement
374 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
376 The only workaround with such peers is to use another HMAC
377 algorithm for IPsec ("phase 2") authentication.
380 Remove the uio_yield prototype and symbol. This function has
381 been misnamed since it was introduced and should not be
382 globally exposed with this name. The equivalent functionality
383 is now available using kern_yield(curthread->td_user_pri).
384 The function remains undocumented.
387 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
388 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
389 for scalar sysctls is defined but disabled. Code that needs
390 UQUAD to pass the type checking that must compile on older
391 systems where the define is not present can check against
392 __FreeBSD_version >= 900030.
394 The system dialog(1) has been replaced with a new version previously
395 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
396 with the previous version, but the libdialog associated with it has
397 a largely incompatible API. As such, the original version of libdialog
398 will be kept temporarily as libodialog, until its base system consumers
399 are replaced or updated. Bump __FreeBSD_version to 900030.
402 If you are trying to run make universe on a -stable system, and you get
403 the following warning:
404 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
405 unknown. config(8) likely too old."
406 or something similar to it, then you must upgrade your -stable system
407 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
408 stable/8) or build the config from the latest stable/8 branch and
409 install it on your system.
411 Prior to this date, building a current universe on 8-stable system from
412 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
413 error in the first kernel build phase. A new config on those old
414 systems will fix that problem for older versions of -current.
417 The TCP stack has been modified to allow Khelp modules to interact with
418 it via helper hook points and store per-connection data in the TCP
419 control block. Bump __FreeBSD_version to 900029. User space tools that
420 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
424 Generic IEEE 802.3 annex 31B full duplex flow control support has been
425 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
426 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
427 to take advantage of it instead of using custom implementations. This
428 means that these drivers now no longer unconditionally advertise
429 support for flow control but only do so if flow control is a selected
430 media option. This was implemented in the generic support that way in
431 order to allow flow control to be switched on and off via ifconfig(8)
432 with the PHY specific default to typically off in order to protect
433 from unwanted effects. Consequently, if you used flow control with
434 one of the above mentioned drivers you now need to explicitly enable
436 ifconfig bge0 media auto mediaopt flowcontrol
438 Along with the above mentioned changes generic support for setting
439 1000baseT master mode also has been added and brgphy(4), ciphy(4),
440 e1000phy(4) as well as ip1000phy(4) have been converted to take
441 advantage of it. This means that these drivers now no longer take the
442 link0 parameter for selecting master mode but the master media option
443 has to be used instead, for example like in the following:
444 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
446 Selection of master mode now is also available with all other PHY
447 drivers supporting 1000baseT.
450 The TCP stack has received a significant update to add support for
451 modularised congestion control and generally improve the clarity of
452 congestion control decisions. Bump __FreeBSD_version to 900025. User
453 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
454 sockstat) need to be recompiled.
457 The man(1) utility has been replaced by a new version that no longer
458 uses /etc/manpath.config. Please consult man.conf(5) for how to
459 migrate local entries to the new format.
462 The copyright strings printed by login(1) and sshd(8) at the time of a
463 new connection have been removed to follow other operating systems and
467 A workaround for a fixed ld bug has been removed in kernel code,
468 so make sure that your system ld is built from sources after
469 revision 210245 from 2010-07-19 (r211583 if building head kernel
470 on stable/8, r211584 for stable/7; both from 2010-08-21).
471 A symptom of incorrect ld version is different addresses for
472 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
475 The $ipv6_prefer variable in rc.conf(5) has been split into
476 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
478 The $ip6addrctl_policy is a variable to choose a pre-defined
479 address selection policy set by ip6addrctl(8). A value
480 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
483 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
484 flag (see an entry of 20090926) is set on an interface with no
485 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
486 security reason. If you want IPv6 link-local address on all
487 interfaces by default, set this to "YES".
489 The old ipv6_prefer="YES" is equivalent to
490 ipv6_activate_all_interfaces="YES" and
491 ip6addrctl_policy="ipv6_prefer".
494 DTrace has grown support for userland tracing. Due to this, DTrace is
495 now i386 and amd64 only.
496 dtruss(1) is now installed by default on those systems and a new
497 kernel module is needed for userland tracing: fasttrap.
498 No changes to your kernel config file are necessary to enable
499 userland tracing, but you might consider adding 'STRIP=' and
500 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
501 to have informative userland stack traces in DTrace (ustack).
504 The acpi_aiboost(4) driver has been removed in favor of the new
505 aibs(4) driver. You should update your kernel configuration file.
508 BSD grep has been imported to the base system and it is built by
509 default. It is completely BSD licensed, highly GNU-compatible, uses
510 less memory than its GNU counterpart and has a small codebase.
511 However, it is slower than its GNU counterpart, which is mostly
512 noticeable for larger searches, for smaller ones it is measurable
513 but not significant. The reason is complex, the most important factor
514 is that we lack a modern and efficient regex library and GNU
515 overcomes this by optimizing the searches internally. Future work
516 on improving the regex performance is planned, for the meantime,
517 users that need better performance, can build GNU grep instead by
518 setting the WITH_GNU_GREP knob.
521 Due to the import of powerpc64 support, all existing powerpc kernel
522 configuration files must be updated with a machine directive like this:
523 machine powerpc powerpc
525 In addition, an updated config(8) is required to build powerpc kernels
529 A new version of ZFS (version 15) has been merged to -HEAD.
530 This version uses a python library for the following subcommands:
531 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
532 For full functionality of these commands the following port must
533 be installed: sysutils/py-zfs
536 'vm_page's are now hashed by physical address to an array of mutexes.
537 Currently this is only used to serialize access to hold_count. Over
538 time the page queue mutex will be peeled away. This changes the size
539 of pmap on every architecture. And requires all callers of vm_page_hold
540 and vm_page_unhold to be updated.
543 WITH_CTF can now be specified in src.conf (not recommended, there
544 are some problems with static executables), make.conf (would also
545 affect ports which do not use GNU make and do not override the
546 compile targets) or in the kernel config (via "makeoptions
548 When WITH_CTF was specified there before this was silently ignored,
549 so make sure that WITH_CTF is not used in places which could lead
550 to unwanted behavior.
553 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
554 to allow 32-bit compatibility on non-x86 platforms. All kernel
555 configurations on amd64 and ia64 platforms using these options must
556 be modified accordingly.
559 The utmp user accounting database has been replaced with utmpx,
560 the user accounting interface standardized by POSIX.
561 Unfortunately the semantics of utmp and utmpx don't match,
562 making it practically impossible to support both interfaces.
563 The user accounting database is used by tools like finger(1),
564 last(1), talk(1), w(1) and ac(8).
566 All applications in the base system use utmpx. This means only
567 local binaries (e.g. from the ports tree) may still use these
568 utmp database files. These applications must be rebuilt to make
571 After the system has been upgraded, it is safe to remove the old
572 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
573 assuming their contents is of no importance anymore. Old wtmp
574 databases can only be used by last(1) and ac(8) after they have
575 been converted to the new format using wtmpcvt(1).
578 Introduce the kernel thread "deadlock resolver" (which can be enabled
579 via the DEADLKRES option, see NOTES for more details) and the
580 sleepq_type() function for sleepqueues.
583 The rc.firewall and rc.firewall6 were unified, and
584 rc.firewall6 and rc.d/ip6fw were removed.
585 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
586 variables are obsoleted. Instead, the following new rc
587 variables are added to rc.d/ipfw:
589 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
590 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
591 firewall_simple_onet_ipv6, firewall_trusted_ipv6
593 The meanings correspond to the relevant IPv4 variables.
599 The default terminal emulation for syscons(4) has been changed
600 from cons25 to xterm on all platforms except pc98. This means
601 that the /etc/ttys file needs to be updated to ensure correct
602 operation of applications on the console.
604 The terminal emulation style can be toggled per window by using
605 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
606 options can be used to change the compile-time default back to
609 To prevent graphical artifacts, make sure the TERM environment
610 variable is set to match the terminal emulation that is being
611 performed by syscons(4).
614 The layout of the structure ieee80211req_scan_result has changed.
615 Applications that require wireless scan results (e.g. ifconfig(8))
616 from net80211 need to be recompiled.
618 Applications such as wpa_supplicant(8) may require a full world
619 build without using NO_CLEAN in order to get synchronized with the
623 The iwn(4) driver has been updated to support the 5000 and 5150 series.
624 There's one kernel module for each firmware. Adding "device iwnfw"
625 to the kernel configuration file means including all three firmware
626 images inside the kernel. If you want to include just the one for
627 your wireless card, use the devices iwn4965fw, iwn5000fw or
631 The rc.d/network_ipv6, IPv6 configuration script has been integrated
632 into rc.d/netif. The changes are the following:
634 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
635 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
636 Note that both variables need the "inet6" keyword at the head.
638 Do not set $ipv6_network_interfaces manually if you do not
639 understand what you are doing. It is not needed in most cases.
641 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
644 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
645 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
647 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
648 all configured interfaces have "inet6 accept_rtadv" in the
649 $ifconfig_IF_ipv6. These are for backward compatibility.
651 3. A new variable $ipv6_prefer has been added. If NO, IPv6
652 functionality of interfaces with no corresponding
653 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
654 and the default address selection policy of ip6addrctl(8)
655 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
656 Note that if you want to configure IPv6 functionality on the
657 disabled interfaces after boot, first you need to clear the flag by
658 using ifconfig(8) like:
660 ifconfig em0 inet6 -ifdisabled
662 If YES, the default address selection policy is set as
665 The default value of $ipv6_prefer is NO.
667 4. If your system need to receive Router Advertisement messages,
668 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
669 scripts automatically invoke rtsol(8) when the interface becomes
670 UP. The Router Advertisement messages are used for SLAAC
671 (State-Less Address AutoConfiguration).
674 802.11s D3.03 support was committed. This is incompatible with the
675 previous code, which was based on D3.0.
678 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
679 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
680 control whether accepting Router Advertisement messages or not.
681 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
682 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
683 The ifconfig(8) utility now supports these flags.
686 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
687 mount(8) and -a option for df(1) to see them.
690 The old tunable hw.bus.devctl_disable has been superseded by
691 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
692 replaced by hw.bus.devctl_queue=0. The default for this new tunable
696 Remove the option STOP_NMI. The default action is now to use NMI only
697 for KDB via the newly introduced function stop_cpus_hard() and
698 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
701 The stable/8 branch created in subversion. This corresponds to the
702 RELENG_8 branch in CVS.
705 Bump the shared library version numbers for all libraries that do not
706 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
707 __FreeBSD_version to 800105.
710 Due to changes in the implementation of virtual network stack support,
711 all network-related kernel modules must be recompiled. As this change
712 breaks the ABI, bump __FreeBSD_version to 800104.
715 The TOE interface to the TCP syncache has been modified to remove
716 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
717 The cxgb driver is the only TOE consumer affected by this change, and
718 needs to be recompiled along with the kernel. As this change breaks
719 the ABI, bump __FreeBSD_version to 800103.
722 Padding has been added to struct tcpcb, sackhint and tcpstat in
723 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
724 maintaining the ABI. However, this change breaks the ABI, so bump
725 __FreeBSD_version to 800102. User space tools that rely on the size of
726 any of these structs (e.g. sockstat) need to be recompiled.
729 The NFS_LEGACYRPC option has been removed along with the old kernel
730 RPC implementation that this option selected. Kernel configurations
731 may need to be adjusted.
734 The network interface device nodes at /dev/net/<interface> have been
735 removed. All ioctl operations can be performed the normal way using
736 routing sockets. The kqueue functionality can generally be replaced
737 with routing sockets.
740 The documentation from the FreeBSD Documentation Project (Handbook,
741 FAQ, etc.) is now installed via packages by sysinstall(8) and under
742 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
745 The ABI of various structures related to the SYSV IPC API have been
746 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
747 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
751 Layout of struct vnet has changed as routing related variables were
752 moved to their own Vimage module. Modules need to be recompiled. Bump
753 __FreeBSD_version to 800099.
756 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
757 respectively. As long as no more than 16 groups per process are used,
758 no changes should be visible. When more than 16 groups are used, old
759 binaries may fail if they call getgroups() or getgrouplist() with
760 statically sized storage. Recompiling will work around this, but
761 applications should be modified to use dynamically allocated storage
762 for group arrays as POSIX.1-2008 does not cap an implementation's
763 number of supported groups at NGROUPS_MAX+1 as previous versions did.
765 NFS and portalfs mounts may also be affected as the list of groups is
766 truncated to 16. Users of NFS who use more than 16 groups, should
767 take care that negative group permissions are not used on the exported
768 file systems as they will not be reliable unless a GSSAPI based
769 authentication method is used.
772 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
773 option compiles in the support for adaptive spinning for lockmgrs
774 which want to enable it. The lockinit() function now accepts the flag
775 LK_ADAPTIVE in order to make the lock object subject to adaptive
776 spinning when both held in write and read mode.
779 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
780 changed. User applications that use this ioctl need to be rebuilt.
783 The layout of struct thread has changed. Kernel and modules need to
787 The layout of structs ifnet, domain, protosw and vnet_net has changed.
788 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
791 window(1) has been removed from the base system. It can now be
792 installed from ports. The port is called misc/window.
795 The way we are storing and accessing `routing table' entries has
796 changed. Programs reading the FIB, like netstat, need to be
800 A new netisr implementation has been added for FreeBSD 8. Network
801 file system modules, such as igmp, ipdivert, and others, should be
803 Bump __FreeBSD_version to 800096.
806 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
810 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
811 Bump __FreeBSD_version to 800094.
814 Add mnt_xflag field to 'struct mount'. File system modules need to be
816 Bump __FreeBSD_version to 800093.
819 The compiling option ADAPTIVE_SX has been retired while it has been
820 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
821 The KPI for sx_init_flags() changes as accepting flags:
822 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
823 been introduced in order to handle the reversed logic.
824 Bump __FreeBSD_version to 800092.
827 Add support for hierarchical jails. Remove global securelevel.
828 Bump __FreeBSD_version to 800091.
831 The layout of struct vnet_net has changed, therefore modules
833 Bump __FreeBSD_version to 800090.
836 The newly imported zic(8) produces a new format in the output. Please
837 run tzsetup(8) to install the newly created data to /etc/localtime.
840 The sysctl tree for the usb stack has renamed from hw.usb2.* to
841 hw.usb.* and is now consistent again with previous releases.
844 802.11 monitor mode support was revised and driver api's were changed.
845 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
846 of DLT_IEEE802_11. No user-visible data structures were changed but
847 applications that use DLT_IEEE802_11 may require changes.
848 Bump __FreeBSD_version to 800088.
851 The layout of the following structs has changed: sysctl_oid,
852 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
853 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
854 panics may be experienced. World rebuild is required for
855 correctly checking networking state from userland.
856 Bump __FreeBSD_version to 800085.
859 MLDv2 and Source-Specific Multicast (SSM) have been merged
860 to the IPv6 stack. VIMAGE hooks are in but not yet used.
861 The implementation of SSM within FreeBSD's IPv6 stack closely
862 follows the IPv4 implementation.
864 For kernel developers:
866 * The most important changes are that the ip6_output() and
867 ip6_input() paths no longer take the IN6_MULTI_LOCK,
868 and this lock has been downgraded to a non-recursive mutex.
870 * As with the changes to the IPv4 stack to support SSM, filtering
871 of inbound multicast traffic must now be performed by transport
872 protocols within the IPv6 stack. This does not apply to TCP and
873 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
875 * The KPIs used by IPv6 multicast are similar to those used by
876 the IPv4 stack, with the following differences:
877 * im6o_mc_filter() is analogous to imo_multicast_filter().
878 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
879 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
880 * IN6_LOOKUP_MULTI() has been deprecated and removed.
881 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
882 for MLDv1 have an additional 'timer' argument which is used to
883 jitter the initial membership report for the solicited-node
884 multicast membership on-link.
885 * This is not strictly needed for MLDv2, which already jitters
886 its report transmissions. However, the 'timer' argument is
887 preserved in case MLDv1 is active on the interface.
889 * The KAME linked-list based IPv6 membership implementation has
890 been refactored to use a vector similar to that used by the IPv4
892 Code which maintains a list of its own multicast memberships
893 internally, e.g. carp, has been updated to reflect the new
896 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
897 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
898 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
899 implementation constraint which needs to be addressed in HEAD.
901 For application developers:
903 * The changes are broadly similar to those made for the IPv4
906 * The use of IPv4 and IPv6 multicast socket options on the same
907 socket, using mapped addresses, HAS NOT been tested or supported.
909 * There are a number of issues with the implementation of various
910 IPv6 multicast APIs which need to be resolved in the API surface
911 before the implementation is fully compatible with KAME userland
912 use, and these are mostly to do with interface index treatment.
914 * The literature available discusses the use of either the delta / ASM
915 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
916 using setsourcefilter(3)/getsourcefilter(3). For more information
917 please refer to RFC 3768, 'Socket Interface Extensions for
918 Multicast Source Filters'.
920 * Applications which use the published RFC 3678 APIs should be fine.
922 For systems administrators:
924 * The mtest(8) utility has been refactored to support IPv6, in
925 addition to IPv4. Interface addresses are no longer accepted
926 as arguments, their names must be used instead. The utility
927 will map the interface name to its first IPv4 address as
928 returned by getifaddrs(3).
930 * The ifmcstat(8) utility has also been updated to print the MLDv2
931 endpoint state and source filter lists via sysctl(3).
933 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
934 loopback of IPv6 multicast datagrams by default; it defaults to 1
935 to preserve the existing behaviour. Disabling multicast loopback is
936 recommended for optimal system performance.
938 * The IPv6 MROUTING code has been changed to examine this sysctl
939 instead of attempting to perform a group lookup before looping
940 back forwarded datagrams.
942 Bump __FreeBSD_version to 800084.
945 Implement low-level Bluetooth HCI API.
946 Bump __FreeBSD_version to 800083.
949 The layout of struct malloc_type, used by modules to register new
950 memory allocation types, has changed. Most modules will need to
951 be rebuilt or panics may be experienced.
952 Bump __FreeBSD_version to 800081.
955 Anticipate overflowing inp_flags - add inp_flags2.
956 This changes most offsets in inpcb, so checking v4 connection
957 state will require a world rebuild.
958 Bump __FreeBSD_version to 800080.
961 Add an llentry to struct route and struct route_in6. Modules
962 embedding a struct route will need to be recompiled.
963 Bump __FreeBSD_version to 800079.
966 The size of rt_metrics_lite and by extension rtentry has changed.
967 Networking administration apps will need to be recompiled.
968 The route command now supports show as an alias for get, weighting
969 of routes, sticky and nostick flags to alter the behavior of stateful
971 Bump __FreeBSD_version to 800078.
974 Do not use Giant for kbdmux(4) locking. This is wrong and
975 apparently causing more problems than it solves. This will
976 re-open the issue where interrupt handlers may race with
977 kbdmux(4) in polling mode. Typical symptoms include (but
978 not limited to) duplicated and/or missing characters when
979 low level console functions (such as gets) are used while
980 interrupts are enabled (for example geli password prompt,
981 mountroot prompt etc.). Disabling kbdmux(4) may help.
984 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
985 kernel modules referencing any of the above need to be recompiled.
986 Bump __FreeBSD_version to 800075.
989 GEOM_PART has become the default partition slicer for storage devices,
990 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
991 introduces some changes:
993 MSDOS/EBR: the devices created from MSDOS extended partition entries
994 (EBR) can be named differently than with GEOM_MBR and are now symlinks
995 to devices with offset-based names. fstabs may need to be modified.
997 BSD: the "geometry does not match label" warning is harmless in most
998 cases but it points to problems in file system misalignment with
999 disk geometry. The "c" partition is now implicit, covers the whole
1000 top-level drive and cannot be (mis)used by users.
1002 General: Kernel dumps are now not allowed to be written to devices
1003 whose partition types indicate they are meant to be used for file
1004 systems (or, in case of MSDOS partitions, as something else than
1007 Most of these changes date approximately from 200812.
1010 The uscanner(4) driver has been removed from the kernel. This follows
1011 Linux removing theirs in 2.6 and making libusb the default interface
1012 (supported by sane).
1015 The multicast forwarding code has been cleaned up. netstat(1)
1016 only relies on KVM now for printing bandwidth upcall meters.
1017 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1018 ip6_mroute_mod respectively. The config(5) options for statically
1019 compiling this code remain the same, i.e. 'options MROUTING'.
1022 Support for the IFF_NEEDSGIANT network interface flag has been
1023 removed, which means that non-MPSAFE network device drivers are no
1024 longer supported. In particular, if_ar, if_sr, and network device
1025 drivers from the old (legacy) USB stack can no longer be built or
1029 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1030 a bunch of new language catalog files have also been added.
1031 This means that some common libc messages are now localized and
1032 they depend on the LC_MESSAGES environmental variable.
1035 The k8temp(4) driver has been renamed to amdtemp(4) since
1036 support for Family 10 and Family 11 CPU families was added.
1039 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1040 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1042 For kernel developers, the most important changes are that the
1043 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1044 and this lock has been downgraded to a non-recursive mutex.
1046 Transport protocols (UDP, Raw IP) are now responsible for filtering
1047 inbound multicast traffic according to group membership and source
1048 filters. The imo_multicast_filter() KPI exists for this purpose.
1049 Transports which do not use multicast (SCTP, TCP) already reject
1050 multicast by default. Forwarding and receive performance may improve
1051 as a mutex acquisition is no longer needed in the ip_input()
1052 low-level input path. in_addmulti() and in_delmulti() are shimmed
1053 to new KPIs which exist to support SSM in-kernel.
1055 For application developers, it is recommended that loopback of
1056 multicast datagrams be disabled for best performance, as this
1057 will still cause the lock to be taken for each looped-back
1058 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1059 be tuned to 0 to disable loopback by default; it defaults to 1
1060 to preserve the existing behaviour.
1062 For systems administrators, to obtain best performance with
1063 multicast reception and multiple groups, it is always recommended
1064 that a card with a suitably precise hash filter is used. Hash
1065 collisions will still result in the lock being taken within the
1066 transport protocol input path to check group membership.
1068 If deploying FreeBSD in an environment with IGMP snooping switches,
1069 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1070 enabled; this forces 224.0.0.0/24 group membership to be announced
1073 The size of 'struct igmpstat' has changed; netstat needs to be
1074 recompiled to reflect this.
1075 Bump __FreeBSD_version to 800070.
1078 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1079 updated to use it. This requires a buildworld/installworld in order to
1080 update the library and dependencies (usbconfig, etc). Its advisable to
1081 rebuild all ports which uses libusb. More specific directions are given
1082 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1083 libusb are no longer required and can be removed.
1086 A workaround is committed to allow the creation of System V shared
1087 memory segment of size > 2 GB on the 64-bit architectures.
1088 Due to a limitation of the existing ABI, the shm_segsz member
1089 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1090 wrong for large segments. Note that limits must be explicitly
1091 raised to allow such segments to be created.
1094 The layout of struct ifnet has changed, requiring a rebuild of all
1095 network device driver modules.
1098 The /dev handling for the new USB stack has changed, a
1099 buildworld/installworld is required for libusb20.
1102 The new USB2 stack has now been permanently moved in and all kernel and
1103 module names reverted to their previous values (eg, usb, ehci, ohci,
1104 ums, ...). The old usb stack can be compiled in by prefixing the name
1105 with the letter 'o', the old usb modules have been removed.
1106 Updating entry 20090216 for xorg and 20090215 for libmap may still
1110 The rc.conf(5) option if_up_delay has been renamed to
1111 defaultroute_delay to better reflect its purpose. If you have
1112 customized this setting in /etc/rc.conf you need to update it to
1116 xorg 7.4 wants to configure its input devices via hald which does not
1117 yet work with USB2. If the keyboard/mouse does not work in xorg then
1119 Option "AllowEmptyInput" "off"
1120 to your ServerLayout section. This will cause X to use the configured
1121 kbd and mouse sections from your xorg.conf.
1124 The GENERIC kernels for all architectures now default to the new USB2
1125 stack. No kernel config options or code have been removed so if a
1126 problem arises please report it and optionally revert to the old USB
1127 stack. If you are loading USB kernel modules or have a custom kernel
1128 that includes GENERIC then ensure that usb names are also changed over,
1129 eg uftdi -> usb2_serial_ftdi.
1131 Older programs linked against the ports libusb 0.1 need to be
1132 redirected to the new stack's libusb20. /etc/libmap.conf can
1134 # Map old usb library to new one for usb2 stack
1135 libusb-0.1.so.8 libusb20.so.1
1138 All USB ethernet devices now attach as interfaces under the name ueN
1139 (eg. ue0). This is to provide a predictable name as vendors often
1140 change usb chipsets in a product without notice.
1143 The ichsmb(4) driver has been changed to require SMBus slave
1144 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1145 All of the other SMBus controller drivers require left-justified
1146 slave addresses, so this change makes all the drivers provide the
1150 INET6 statistics (struct ip6stat) was updated.
1151 netstat(1) needs to be recompiled.
1154 NTFS has been removed from GENERIC kernel on amd64 to match
1155 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1156 will load ntfs.ko module automatically when NTFS support is
1157 actually needed, unless ntfs.ko is not installed or security
1158 level prohibits loading kernel modules. If either is the case,
1159 "options NTFS" has to be added into kernel config.
1162 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1163 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1164 800061. User space tools that rely on the size of struct tcpcb in
1165 tcp_var.h (e.g. sockstat) need to be recompiled.
1168 ng_tty(4) module updated to match the new TTY subsystem.
1169 Due to API change, user-level applications must be updated.
1170 New API support added to mpd5 CVS and expected to be present
1171 in next mpd5.3 release.
1174 With __FreeBSD_version 800060 the makefs tool is part of
1175 the base system (it was a port).
1178 The afdata and ifnet locks have been changed from mutexes to
1179 rwlocks, network modules will need to be re-compiled.
1182 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1183 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1184 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1185 architecture and 40 bytes on 64-bit architecture. The userland
1186 applications "arp" and "ndp" have been updated accordingly.
1187 The output from "netstat -r" shows only routing entries and
1188 none of the L2 information.
1191 __FreeBSD_version 800057 marks the switchover from the
1192 binary ath hal to source code. Users must add the line:
1194 options AH_SUPPORT_AR5416
1196 to their kernel config files when specifying:
1200 The ath_hal module no longer exists; the code is now compiled
1201 together with the driver in the ath module. It is now
1202 possible to tailor chip support (i.e. reduce the set of chips
1203 and thereby the code size); consult ath_hal(4) for details.
1206 __FreeBSD_version 800054 adds memory barriers to
1207 <machine/atomic.h>, new interfaces to ifnet to facilitate
1208 multiple hardware transmit queues for cards that support
1209 them, and a lock-less ring-buffer implementation to
1210 enable drivers to more efficiently manage queueing of
1214 A new version of ZFS (version 13) has been merged to -HEAD.
1215 This version has zpool attribute "listsnapshots" off by
1216 default, which means "zfs list" does not show snapshots,
1217 and is the same as Solaris behavior.
1220 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1223 The uhci, ohci, ehci and slhci USB Host controller drivers have
1224 been put into separate modules. If you load the usb module
1225 separately through loader.conf you will need to load the
1226 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1227 controller add the following to loader.conf:
1233 The ABI used by the PMC toolset has changed. Please keep
1234 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1238 atapci kernel module now includes only generic PCI ATA
1239 driver. AHCI driver moved to ataahci kernel module.
1240 All vendor-specific code moved into separate kernel modules:
1241 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1242 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1243 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1244 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1247 The TTY subsystem of the kernel has been replaced by a new
1248 implementation, which provides better scalability and an
1249 improved driver model. Most common drivers have been migrated to
1250 the new TTY subsystem, while others have not. The following
1251 drivers have not yet been ported to the new TTY layer:
1254 cy, digi, rc, rp, sio
1260 ng_h4, ng_tty, ppp, sl, snp
1262 Adding these drivers to your kernel configuration file shall
1263 cause compilation to fail.
1266 ntpd has been upgraded to 4.2.4p5.
1269 OpenSSH has been upgraded to 5.1p1.
1271 For many years, FreeBSD's version of OpenSSH preferred DSA
1272 over RSA for host and user authentication keys. With this
1273 upgrade, we've switched to the vendor's default of RSA over
1274 DSA. This may cause upgraded clients to warn about unknown
1275 host keys even for previously known hosts. Users should
1276 follow the usual procedure for verifying host keys before
1277 accepting the RSA key.
1279 This can be circumvented by setting the "HostKeyAlgorithms"
1280 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1283 Please note that the sequence of keys offered for
1284 authentication has been changed as well. You may want to
1285 specify IdentityFile in a different order to revert this
1289 The sio(4) driver has been removed from the i386 and amd64
1290 kernel configuration files. This means uart(4) is now the
1291 default serial port driver on those platforms as well.
1293 To prevent collisions with the sio(4) driver, the uart(4) driver
1294 uses different names for its device nodes. This means the
1295 onboard serial port will now most likely be called "ttyu0"
1296 instead of "ttyd0". You may need to reconfigure applications to
1297 use the new device names.
1299 When using the serial port as a boot console, be sure to update
1300 /boot/device.hints and /etc/ttys before booting the new kernel.
1301 If you forget to do so, you can still manually specify the hints
1302 at the loader prompt:
1304 set hint.uart.0.at="isa"
1305 set hint.uart.0.port="0x3F8"
1306 set hint.uart.0.flags="0x10"
1307 set hint.uart.0.irq="4"
1311 The gpt(8) utility has been removed. Use gpart(8) to partition
1315 The version that Linuxulator emulates was changed from 2.4.2
1316 to 2.6.16. If you experience any problems with Linux binaries
1317 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1318 if it fixes the problem contact emulation mailing list.
1321 ISDN4BSD (I4B) was removed from the src tree. You may need to
1322 update a your kernel configuration and remove relevant entries.
1325 I have checked in code to support multiple routing tables.
1326 See the man pages setfib(1) and setfib(2).
1327 This is a hopefully backwards compatible version,
1328 but to make use of it you need to compile your kernel
1329 with options ROUTETABLES=2 (or more up to 16).
1332 The 802.11 wireless support was redone to enable multi-bss
1333 operation on devices that are capable. The underlying device
1334 is no longer used directly but instead wlanX devices are
1335 cloned with ifconfig. This requires changes to rc.conf files.
1336 For example, change:
1337 ifconfig_ath0="WPA DHCP"
1340 ifconfig_wlan0="WPA DHCP"
1341 see rc.conf(5) for more details. In addition, mergemaster of
1342 /etc/rc.d is highly recommended. Simultaneous update of userland
1343 and kernel wouldn't hurt either.
1345 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1346 modules were merged into the base wlan module. All references
1347 to these modules (e.g. in kernel config files) must be removed.
1350 psm(4) has gained write(2) support in native operation level.
1351 Arbitrary commands can be written to /dev/psm%d and status can
1352 be read back from it. Therefore, an application is responsible
1353 for status validation and error recovery. It is a no-op in
1354 other operation levels.
1357 Support for KSE threading has been removed from the kernel. To
1358 run legacy applications linked against KSE libmap.conf may
1359 be used. The following libmap.conf may be used to ensure
1360 compatibility with any prior release:
1362 libpthread.so.1 libthr.so.1
1363 libpthread.so.2 libthr.so.2
1364 libkse.so.3 libthr.so.3
1367 The layout of struct vmspace has changed. This affects libkvm
1368 and any executables that link against libkvm and use the
1369 kvm_getprocs() function. In particular, but not exclusively,
1370 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1371 The effects are minimal, but it's advisable to upgrade world
1375 The latest em driver no longer has support in it for the
1376 82575 adapter, this is now moved to the igb driver. The
1377 split was done to make new features that are incompatible
1378 with older hardware easier to do.
1381 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1382 likewise the kernel option is now GEOM_LINUX_LVM.
1385 The default NFS mount mode has changed from UDP to TCP for
1386 increased reliability. If you rely on (insecurely) NFS
1387 mounting across a firewall you may need to update your
1391 Belatedly note the addition of m_collapse for compacting
1395 The fts(3) structures have been changed to use adequate
1396 integer types for their members and so to be able to cope
1397 with huge file trees. The old fts(3) ABI is preserved
1398 through symbol versioning in libc, so third-party binaries
1399 using fts(3) should still work, although they will not take
1400 advantage of the extended types. At the same time, some
1401 third-party software might fail to build after this change
1402 due to unportable assumptions made in its source code about
1403 fts(3) structure members. Such software should be fixed
1404 by its vendor or, in the worst case, in the ports tree.
1405 FreeBSD_version 800015 marks this change for the unlikely
1406 case that a portable fix is impossible.
1409 To upgrade to -current after this date, you must be running
1410 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1411 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1414 The ADAPTIVE_GIANT kernel option has been retired because its
1415 functionality is the default now.
1418 The AT keyboard emulation of sunkbd(4) has been turned on
1419 by default. In order to make the special symbols of the Sun
1420 keyboards driven by sunkbd(4) work under X these now have
1421 to be configured the same way as Sun USB keyboards driven
1422 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1424 Option "XkbLayout" "us"
1425 Option "XkbRules" "xorg"
1426 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1429 It has been decided that it is desirable to provide ABI
1430 backwards compatibility to the FreeBSD 4/5/6 versions of the
1431 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1432 broken with the introduction of PCI domain support (see the
1433 20070930 entry). Unfortunately, this required the ABI of
1434 PCIOCGETCONF to be broken again in order to be able to
1435 provide backwards compatibility to the old version of that
1436 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1437 again. As for prominent ports this affects neither pciutils
1438 nor xorg-server this time, the hal port needs to be rebuilt
1442 The misnamed kthread_create() and friends have been renamed
1443 to kproc_create() etc. Many of the callers already
1444 used kproc_start()..
1445 I will return kthread_create() and friends in a while
1446 with implementations that actually create threads, not procs.
1447 Renaming corresponds with version 800002.
1456 Avoid using make -j when upgrading. While generally safe, there are
1457 sometimes problems using -j to upgrade. If your upgrade fails with
1458 -j, please try again without -j. From time to time in the past there
1459 have been problems using -j with buildworld and/or installworld. This
1460 is especially true when upgrading between "distant" versions (eg one
1461 that cross a major release boundary or several minor releases, or when
1462 several months have passed on the -current branch).
1464 Sometimes, obscure build problems are the result of environment
1465 poisoning. This can happen because the make utility reads its
1466 environment when searching for values for global variables. To run
1467 your build attempts in an "environmental clean room", prefix all make
1468 commands with 'env -i '. See the env(1) manual page for more details.
1470 When upgrading from one major version to another it is generally best
1471 to upgrade to the latest code in the currently installed branch first,
1472 then do an upgrade to the new branch. This is the best-tested upgrade
1473 path, and has the highest probability of being successful. Please try
1474 this approach before reporting problems with a major version upgrade.
1478 When upgrading the boot ZFS pool to a new version, always follow
1481 1.) recompile and reinstall the ZFS boot loader and boot block
1482 (this is part of "make buildworld" and "make installworld")
1484 2.) update the ZFS boot block on your boot drive
1486 The following example updates the ZFS boot block on the first
1487 partition (freebsd-boot) of a GPT partitioned drive ad0:
1488 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1490 Non-boot pools do not need these updates.
1494 If you are updating from a prior version of FreeBSD (even one just
1495 a few days old), you should follow this procedure. It is the most
1496 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1498 make kernel-toolchain
1499 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1500 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1502 To test a kernel once
1503 ---------------------
1504 If you just want to boot a kernel once (because you are not sure
1505 if it works, or if you want to boot a known bad kernel to provide
1506 debugging information) run
1507 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1508 nextboot -k testkernel
1510 To just build a kernel when you know that it won't mess you up
1511 --------------------------------------------------------------
1512 This assumes you are already running a CURRENT system. Replace
1513 ${arch} with the architecture of your machine (e.g. "i386",
1514 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1516 cd src/sys/${arch}/conf
1517 config KERNEL_NAME_HERE
1518 cd ../compile/KERNEL_NAME_HERE
1523 If this fails, go to the "To build a kernel" section.
1525 To rebuild everything and install it on the current system.
1526 -----------------------------------------------------------
1527 # Note: sometimes if you are running current you gotta do more than
1528 # is listed here if you are upgrading from a really old current.
1530 <make sure you have good level 0 dumps>
1532 make kernel KERNCONF=YOUR_KERNEL_HERE
1534 <reboot in single user> [3]
1542 To cross-install current onto a separate partition
1543 --------------------------------------------------
1544 # In this approach we use a separate partition to hold
1545 # current's root, 'usr', and 'var' directories. A partition
1546 # holding "/", "/usr" and "/var" should be about 2GB in
1549 <make sure you have good level 0 dumps>
1552 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1553 <maybe newfs current's root partition>
1554 <mount current's root partition on directory ${CURRENT_ROOT}>
1555 make installworld DESTDIR=${CURRENT_ROOT}
1556 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1557 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1558 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1559 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1560 <reboot into current>
1561 <do a "native" rebuild/install as described in the previous section>
1562 <maybe install compatibility libraries from ports/misc/compat*>
1566 To upgrade in-place from stable to current
1567 ----------------------------------------------
1568 <make sure you have good level 0 dumps>
1570 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1572 <reboot in single user> [3]
1579 Make sure that you've read the UPDATING file to understand the
1580 tweaks to various things you need. At this point in the life
1581 cycle of current, things change often and you are on your own
1582 to cope. The defaults can also change, so please read ALL of
1583 the UPDATING entries.
1585 Also, if you are tracking -current, you must be subscribed to
1586 freebsd-current@freebsd.org. Make sure that before you update
1587 your sources that you have read and understood all the recent
1588 messages there. If in doubt, please track -stable which has
1589 much fewer pitfalls.
1591 [1] If you have third party modules, such as vmware, you
1592 should disable them at this point so they don't crash your
1595 [3] From the bootblocks, boot -s, and then do
1600 adjkerntz -i # if CMOS is wall time
1601 Also, when doing a major release upgrade, it is required that
1602 you boot into single user mode to do the installworld.
1604 [4] Note: This step is non-optional. Failure to do this step
1605 can result in a significant reduction in the functionality of the
1606 system. Attempting to do it by hand is not recommended and those
1607 that pursue this avenue should read this file carefully, as well
1608 as the archives of freebsd-current and freebsd-hackers mailing lists
1609 for potential gotchas. The -U option is also useful to consider.
1610 See mergemaster(8) for more information.
1612 [5] Usually this step is a noop. However, from time to time
1613 you may need to do this if you get unknown user in the following
1614 step. It never hurts to do it all the time. You may need to
1615 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1616 install) after the buildworld before this step if you last updated
1617 from [78]-stable or 9-stable before 20130430.
1619 [6] This only deletes old files and directories. Old libraries
1620 can be deleted by "make delete-old-libs", but you have to make
1621 sure that no program is using those libraries anymore.
1623 [8] In order to have a kernel that can run the 4.x binaries needed to
1624 do an installworld, you must include the COMPAT_FREEBSD4 option in
1625 your kernel. Failure to do so may leave you with a system that is
1626 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1627 required to run the 5.x binaries on more recent kernels. And so on
1628 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1630 Make sure that you merge any new devices from GENERIC since the
1631 last time you updated your kernel config file.
1633 [9] When checking out sources, you must include the -P flag to have
1634 cvs prune empty directories.
1636 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1637 "?=" instead of the "=" assignment operator, so that buildworld can
1638 override the CPUTYPE if it needs to.
1640 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1641 not on the command line, or in /etc/make.conf. buildworld will
1642 warn if it is improperly defined.
1645 This file contains a list, in reverse chronological order, of major
1646 breakages in tracking -current. It is not guaranteed to be a complete
1647 list of such breakages, and only contains entries since October 10, 2007.
1648 If you need to see UPDATING entries from before that date, you will need
1649 to fetch an UPDATING file from an older FreeBSD release.
1651 Copyright information:
1653 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1655 Redistribution, publication, translation and use, with or without
1656 modification, in full or in part, in any form or format of this
1657 document are permitted without further permission from the author.
1659 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1660 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1661 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1662 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1663 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1664 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1665 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1666 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1667 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1668 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1669 POSSIBILITY OF SUCH DAMAGE.
1671 Contact Warner Losh if you have any questions about your use of