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 Fix a bug that allowed a tracing process (e.g. gdb) to write
16 to a memory-mapped file in the traced process's address space
17 even if neither the traced process nor the tracing process had
18 write access to that file.
21 Added ZFS TRIM support which is enabled by default. To disable
22 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
24 Creating new ZFS pools and adding new devices to existing pools
25 first performs a full device level TRIM which can take a significant
26 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
27 to disable this behaviour.
29 ZFS TRIM requires the underlying device support BIO_DELETE which
30 is currently provided by methods such as ATA TRIM and SCSI UNMAP
31 via CAM, which are typically supported by SSD's.
33 Stats for ZFS TRIM can be monitored by looking at the sysctl's
34 under kstat.zfs.misc.zio_trim.
37 `list' command has been added to hastctl(8). For now, it is full
38 equivalent of `status' command.
39 WARNING: in the near future the output of hastctl's status command
40 will change to more terse format. If you use `hastctl status'
41 for parsing in your scripts, switch to `hastctl list'.
44 The mergemaster command now uses the default MAKEOBJDIRPREFIX
45 rather than creating it's own in the temporary directory in
46 order allow access to bootstrapped versions of tools such as
47 install and mtree. When upgrading from version of FreeBSD where
48 the install command does not support -l, you will need to
49 install a new mergemaster command if mergemaster -p is required.
50 This can be accomplished with the command (cd src/usr.sbin/mergemaster
53 Due to the use of the new -l option to install(1) during build
54 and install, you must take care not to directly set the INSTALL
55 make variable in your /etc/make.conf, /etc/src.conf, or on the
56 command line. If you wish to use the -C flag for all installs
57 you may be able to add INSTALL+=-C to /etc/make.conf or
61 Fix a bug that allows NFS clients to issue READDIR on files.
64 The install(1) option -M has changed meaning and now takes an
65 argument that is a file or path to append logs to. In the
66 unlikely event that -M was the last option on the command line
67 and the command line contained at least two files and a target
68 directory the first file will have logs appended to it. The -M
69 option served little practical purpose in the last decade so it's
70 used expected to be extremely rare.
73 A new compression method (lz4) has been merged to. Please refer to
74 zpool-features(7) for more information.
76 Please refer to the "ZFS notes" section of this file for information
77 on upgrading boot ZFS pools.
80 The VFS KBI was changed with the merge of several nullfs
81 optimizations and fixes. All filesystem modules must be
85 With the addition of auditdistd(8), a new auditdistd user is now
86 depended on during installworld. "mergemaster -p" can be used to add
87 the user prior to installworld, as documented in the handbook.
93 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
94 Starting with this version the old system of ZFS pool versioning
95 is superseded by "feature flags". This concept enables forward
96 compatibility against certain future changes in functionality of ZFS
97 pools. The first two read-only compatible "feature flags" for ZFS
98 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
99 For more information read the new zpool-features(7) manual page.
100 Please refer to the "ZFS notes" section of this file for information
101 on upgrading boot ZFS pools.
104 The commit introducing bsd.compiler.mk breaks the traditional
105 building of kernels before this point. Add -m ${SRC}/share/mk
106 (for the right value of SRC) to your command lines to work
107 around; update your useland to a point after this; or use the
108 buildkernel/installkernel top-level targets. See also 20120829.
111 The IPFIREWALL_FORWARD kernel option has been removed. Its
112 functionality now turned on by default.
115 The random(4) support for the VIA hardware random number
116 generator (`PADLOCK') is no longer enabled unconditionally.
117 Add the PADLOCK_RNG option in the custom kernel config if
118 needed. The GENERIC kernels on i386 and amd64 do include the
119 option, so the change only affects the custom kernel
123 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
124 the traditional method, you must update your system with an installworld
125 before the kernel will build. The documented make buildkernel/installkernel
126 interfaces (coupled with fresh make kernel-toolchain) continue to work.
129 The sparc64 ZFS loader has been changed to no longer try to auto-
130 detect ZFS providers based on diskN aliases but now requires these
131 to be explicitly listed in the OFW boot-device environment variable.
134 Now unix domain sockets behave "as expected" on nullfs(5). Previously
135 nullfs(5) did not pass through all behaviours to the underlying layer,
136 as a result if we bound to a socket on the lower layer we could connect
137 only to the lower path; if we bound to the upper layer we could connect
138 only to the upper path. The new behavior is one can connect to both the
139 lower and the upper paths regardless what layer path one binds to.
142 The acpi_wmi(4) status device /dev/wmistat has been renamed to
146 A new VOP_ADVISE() was added to support posix_fadvise(2). All
147 filesystem modules must be recompiled.
150 The interface of the VOP_VPTOCNP(9) changed, now the returned
151 vnode shall be referenced, previously it was required to be
152 only held. All in-tree filesystems are converted.
158 The broken amd(4) driver has been replaced with esp(4) in the amd64,
159 i386 and pc98 GENERIC kernel configuration files.
162 This commit modifies vfs_register() so that it uses a hash
163 calculation to set vfc_typenum, which is enabled by default.
164 The first time a system is booted after this change, the
165 vfc_typenum values will change for all file systems. The
166 main effect of this is a change to the NFS server file handles
167 for file systems that use vfc_typenum in their fsid, such as ZFS.
168 It will, however, prevent vfc_typenum from changing when file
169 systems are loaded in a different order for subsequent reboots.
170 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
171 until you are ready to remount all NFS clients after a reboot.
174 Bump the shared library version numbers for libraries that
175 do not use symbol versioning, have changed the ABI compared
176 to stable/8 and which shared library version was not bumped.
177 Done as part of 9.0-RELEASE cycle.
180 During the merge of Capsicum features, the fget(9) KPI was modified.
181 This may require the rebuilding of out-of-tree device drivers --
182 issues have been reported specifically with the nVidia device driver.
183 __FreeBSD_version is bumped to 900041.
185 Also, there is a period between 20110811 and 20110814 where the
186 special devices /dev/{stdin,stdout,stderr} did not work correctly.
187 Building world from a kernel during that window may not work.
190 The packet filter (pf) code has been updated to OpenBSD 4.5.
191 You need to update userland tools to be in sync with kernel.
192 This update breaks backward compatibility with earlier pfsync(4)
193 versions. Care must be taken when updating redundant firewall setups.
196 The following sysctls and tunables are retired on x86 platforms:
198 machdep.hlt_logical_cpus
199 The following sysctl is retired:
200 machdep.hyperthreading_allowed
201 The sysctls were supposed to provide a way to dynamically offline and
202 online selected CPUs on x86 platforms, but the implementation has not
203 been reliable especially with SCHED_ULE scheduler.
204 machdep.hyperthreading_allowed tunable is still available to ignore
205 hyperthreading CPUs at OS level.
206 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
207 where X is an APIC ID of a CPU. Be advised, though, that disabling
208 CPUs in non-uniform fashion will result in non-uniform topology and
209 may lead to sub-optimal system performance with SCHED_ULE, which is
213 cpumask_t type is retired and cpuset_t is used in order to describe
217 Changes to ifconfig(8) for dynamic address family detection mandate
218 that you are running a kernel of 20110525 or later. Make sure to
219 follow the update procedure to boot a new kernel before installing
223 Support for sun4v architecture is officially dropped
226 Several KPI breaking changes have been committed to the mii(4) layer,
227 the PHY drivers and consequently some Ethernet drivers using mii(4).
228 This means that miibus.ko and the modules of the affected Ethernet
229 drivers need to be recompiled.
231 Note to kernel developers: Given that the OUI bit reversion problem
232 was fixed as part of these changes all mii(4) commits related to OUIs,
233 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
234 handling, no longer can be merged verbatim to stable/8 and previous
238 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
239 into their kernel configurations along with 'device pci'.
242 The default NFS client is now the new NFS client, so fstype "newnfs"
243 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
244 Although mounts via fstype "nfs" will usually work without userland
245 changes, it is recommended that the mount(8) and mount_nfs(8)
246 commands be rebuilt from sources and that a link to mount_nfs called
247 mount_oldnfs be created. The new client is compiled into the
248 kernel with "options NFSCL" and this is needed for diskless root
249 file systems. The GENERIC kernel configs have been changed to use
250 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
251 To use the regular/old client, you can "mount -t oldnfs ...". For
252 a diskless root file system, you must also include a line like:
254 vfs.root.mountfrom="oldnfs:"
256 in the boot/loader.conf on the root fs on the NFS server to make
257 a diskless root fs use the old client.
260 The GENERIC kernels for all architectures now default to the new
261 CAM-based ATA stack. It means that all legacy ATA drivers were
262 removed and replaced by respective CAM drivers. If you are using
263 ATA device names in /etc/fstab or other places, make sure to update
264 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
265 where 'Y's are the sequential numbers starting from zero for each type
266 in order of detection, unless configured otherwise with tunables,
267 see cam(4)). There will be symbolic links created in /dev/ to map
268 old adX devices to the respective adaY. They should provide basic
269 compatibility for file systems mounting in most cases, but they do
270 not support old user-level APIs and do not have respective providers
271 in GEOM. Consider using updated management tools with new device names.
273 It is possible to load devices ahci, ata, siis and mvs as modules,
274 but option ATA_CAM should remain in kernel configuration to make ata
275 module work as CAM driver supporting legacy ATA controllers. Device ata
276 still can be used in modular fashion (atacore + ...). Modules atadisk
277 and atapi* are not used and won't affect operation in ATA_CAM mode.
278 Note that to use CAM-based ATA kernel should include CAM devices
279 scbus, pass, da (or explicitly ada), cd and optionally others. All of
280 them are parts of the cam module.
282 ataraid(4) functionality is now supported by the RAID GEOM class.
283 To use it you can load geom_raid kernel module and use graid(8) tool
284 for management. Instead of /dev/arX device names, use /dev/raid/rX.
286 No kernel config options or code have been removed, so if a problem
287 arises, please report it and optionally revert to the old ATA stack.
288 In order to do it you can remove from the kernel config:
293 , and instead add back:
294 device atadisk # ATA disk drives
295 device ataraid # ATA RAID drives
296 device atapicd # ATAPI CDROM drives
297 device atapifd # ATAPI floppy drives
298 device atapist # ATAPI tape drives
301 The default NFS server has been changed to the new server, which
302 was referred to as the experimental server. If you need to switch
303 back to the old NFS server, you must now put the "-o" option on
304 both the mountd and nfsd commands. This can be done using the
305 mountd_flags and nfs_server_flags rc.conf variables until an
306 update to the rc scripts is committed, which is coming soon.
309 The GNU Objective-C runtime library (libobjc), and other Objective-C
310 related components have been removed from the base system. If you
311 require an Objective-C library, please use one of the available ports.
314 ath(4) has been split into bus- and device- modules. if_ath contains
315 the HAL, the TX rate control and the network device code. if_ath_pci
316 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
317 contains the AHB glue. Users need to load both if_ath_pci and if_ath
318 in order to use ath on everything else.
320 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
321 need to load if_ath and if_ath_pci for ath(4) operation.
324 As part of the replacement of sysinstall, the process of building
325 release media has changed significantly. For details, please re-read
326 release(7), which has been updated to reflect the new build process.
329 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
330 is the last available version under GPLv2. It brings a number of new
331 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
332 4.1 and SSE 4.2), better support for powerpc64, a number of new
333 directives, and lots of other small improvements. See the ChangeLog
334 file in contrib/binutils for the full details.
337 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
338 compliant, and will now use half of hash for authentication.
339 This will break interoperability with all stacks (including all
340 actual FreeBSD versions) who implement
341 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
343 The only workaround with such peers is to use another HMAC
344 algorithm for IPsec ("phase 2") authentication.
347 Remove the uio_yield prototype and symbol. This function has
348 been misnamed since it was introduced and should not be
349 globally exposed with this name. The equivalent functionality
350 is now available using kern_yield(curthread->td_user_pri).
351 The function remains undocumented.
354 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
355 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
356 for scalar sysctls is defined but disabled. Code that needs
357 UQUAD to pass the type checking that must compile on older
358 systems where the define is not present can check against
359 __FreeBSD_version >= 900030.
361 The system dialog(1) has been replaced with a new version previously
362 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
363 with the previous version, but the libdialog associated with it has
364 a largely incompatible API. As such, the original version of libdialog
365 will be kept temporarily as libodialog, until its base system consumers
366 are replaced or updated. Bump __FreeBSD_version to 900030.
369 If you are trying to run make universe on a -stable system, and you get
370 the following warning:
371 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
372 unknown. config(8) likely too old."
373 or something similar to it, then you must upgrade your -stable system
374 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
375 stable/8) or build the config from the latest stable/8 branch and
376 install it on your system.
378 Prior to this date, building a current universe on 8-stable system from
379 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
380 error in the first kernel build phase. A new config on those old
381 systems will fix that problem for older versions of -current.
384 The TCP stack has been modified to allow Khelp modules to interact with
385 it via helper hook points and store per-connection data in the TCP
386 control block. Bump __FreeBSD_version to 900029. User space tools that
387 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
391 Generic IEEE 802.3 annex 31B full duplex flow control support has been
392 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
393 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
394 to take advantage of it instead of using custom implementations. This
395 means that these drivers now no longer unconditionally advertise
396 support for flow control but only do so if flow control is a selected
397 media option. This was implemented in the generic support that way in
398 order to allow flow control to be switched on and off via ifconfig(8)
399 with the PHY specific default to typically off in order to protect
400 from unwanted effects. Consequently, if you used flow control with
401 one of the above mentioned drivers you now need to explicitly enable
403 ifconfig bge0 media auto mediaopt flowcontrol
405 Along with the above mentioned changes generic support for setting
406 1000baseT master mode also has been added and brgphy(4), ciphy(4),
407 e1000phy(4) as well as ip1000phy(4) have been converted to take
408 advantage of it. This means that these drivers now no longer take the
409 link0 parameter for selecting master mode but the master media option
410 has to be used instead, for example like in the following:
411 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
413 Selection of master mode now is also available with all other PHY
414 drivers supporting 1000baseT.
417 The TCP stack has received a significant update to add support for
418 modularised congestion control and generally improve the clarity of
419 congestion control decisions. Bump __FreeBSD_version to 900025. User
420 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
421 sockstat) need to be recompiled.
424 The man(1) utility has been replaced by a new version that no longer
425 uses /etc/manpath.config. Please consult man.conf(5) for how to
426 migrate local entries to the new format.
429 The copyright strings printed by login(1) and sshd(8) at the time of a
430 new connection have been removed to follow other operating systems and
434 A workaround for a fixed ld bug has been removed in kernel code,
435 so make sure that your system ld is built from sources after
436 revision 210245 from 2010-07-19 (r211583 if building head kernel
437 on stable/8, r211584 for stable/7; both from 2010-08-21).
438 A symptom of incorrect ld version is different addresses for
439 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
442 The $ipv6_prefer variable in rc.conf(5) has been split into
443 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
445 The $ip6addrctl_policy is a variable to choose a pre-defined
446 address selection policy set by ip6addrctl(8). A value
447 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
450 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
451 flag (see an entry of 20090926) is set on an interface with no
452 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
453 security reason. If you want IPv6 link-local address on all
454 interfaces by default, set this to "YES".
456 The old ipv6_prefer="YES" is equivalent to
457 ipv6_activate_all_interfaces="YES" and
458 ip6addrctl_policy="ipv6_prefer".
461 DTrace has grown support for userland tracing. Due to this, DTrace is
462 now i386 and amd64 only.
463 dtruss(1) is now installed by default on those systems and a new
464 kernel module is needed for userland tracing: fasttrap.
465 No changes to your kernel config file are necessary to enable
466 userland tracing, but you might consider adding 'STRIP=' and
467 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
468 to have informative userland stack traces in DTrace (ustack).
471 The acpi_aiboost(4) driver has been removed in favor of the new
472 aibs(4) driver. You should update your kernel configuration file.
475 BSD grep has been imported to the base system and it is built by
476 default. It is completely BSD licensed, highly GNU-compatible, uses
477 less memory than its GNU counterpart and has a small codebase.
478 However, it is slower than its GNU counterpart, which is mostly
479 noticeable for larger searches, for smaller ones it is measurable
480 but not significant. The reason is complex, the most important factor
481 is that we lack a modern and efficient regex library and GNU
482 overcomes this by optimizing the searches internally. Future work
483 on improving the regex performance is planned, for the meantime,
484 users that need better performance, can build GNU grep instead by
485 setting the WITH_GNU_GREP knob.
488 Due to the import of powerpc64 support, all existing powerpc kernel
489 configuration files must be updated with a machine directive like this:
490 machine powerpc powerpc
492 In addition, an updated config(8) is required to build powerpc kernels
496 A new version of ZFS (version 15) has been merged to -HEAD.
497 This version uses a python library for the following subcommands:
498 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
499 For full functionality of these commands the following port must
500 be installed: sysutils/py-zfs
503 'vm_page's are now hashed by physical address to an array of mutexes.
504 Currently this is only used to serialize access to hold_count. Over
505 time the page queue mutex will be peeled away. This changes the size
506 of pmap on every architecture. And requires all callers of vm_page_hold
507 and vm_page_unhold to be updated.
510 WITH_CTF can now be specified in src.conf (not recommended, there
511 are some problems with static executables), make.conf (would also
512 affect ports which do not use GNU make and do not override the
513 compile targets) or in the kernel config (via "makeoptions
515 When WITH_CTF was specified there before this was silently ignored,
516 so make sure that WITH_CTF is not used in places which could lead
517 to unwanted behavior.
520 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
521 to allow 32-bit compatibility on non-x86 platforms. All kernel
522 configurations on amd64 and ia64 platforms using these options must
523 be modified accordingly.
526 The utmp user accounting database has been replaced with utmpx,
527 the user accounting interface standardized by POSIX.
528 Unfortunately the semantics of utmp and utmpx don't match,
529 making it practically impossible to support both interfaces.
530 The user accounting database is used by tools like finger(1),
531 last(1), talk(1), w(1) and ac(8).
533 All applications in the base system use utmpx. This means only
534 local binaries (e.g. from the ports tree) may still use these
535 utmp database files. These applications must be rebuilt to make
538 After the system has been upgraded, it is safe to remove the old
539 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
540 assuming their contents is of no importance anymore. Old wtmp
541 databases can only be used by last(1) and ac(8) after they have
542 been converted to the new format using wtmpcvt(1).
545 Introduce the kernel thread "deadlock resolver" (which can be enabled
546 via the DEADLKRES option, see NOTES for more details) and the
547 sleepq_type() function for sleepqueues.
550 The rc.firewall and rc.firewall6 were unified, and
551 rc.firewall6 and rc.d/ip6fw were removed.
552 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
553 variables are obsoleted. Instead, the following new rc
554 variables are added to rc.d/ipfw:
556 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
557 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
558 firewall_simple_onet_ipv6, firewall_trusted_ipv6
560 The meanings correspond to the relevant IPv4 variables.
566 The default terminal emulation for syscons(4) has been changed
567 from cons25 to xterm on all platforms except pc98. This means
568 that the /etc/ttys file needs to be updated to ensure correct
569 operation of applications on the console.
571 The terminal emulation style can be toggled per window by using
572 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
573 options can be used to change the compile-time default back to
576 To prevent graphical artifacts, make sure the TERM environment
577 variable is set to match the terminal emulation that is being
578 performed by syscons(4).
581 The layout of the structure ieee80211req_scan_result has changed.
582 Applications that require wireless scan results (e.g. ifconfig(8))
583 from net80211 need to be recompiled.
585 Applications such as wpa_supplicant(8) may require a full world
586 build without using NO_CLEAN in order to get synchronized with the
590 The iwn(4) driver has been updated to support the 5000 and 5150 series.
591 There's one kernel module for each firmware. Adding "device iwnfw"
592 to the kernel configuration file means including all three firmware
593 images inside the kernel. If you want to include just the one for
594 your wireless card, use the devices iwn4965fw, iwn5000fw or
598 The rc.d/network_ipv6, IPv6 configuration script has been integrated
599 into rc.d/netif. The changes are the following:
601 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
602 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
603 Note that both variables need the "inet6" keyword at the head.
605 Do not set $ipv6_network_interfaces manually if you do not
606 understand what you are doing. It is not needed in most cases.
608 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
611 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
612 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
614 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
615 all configured interfaces have "inet6 accept_rtadv" in the
616 $ifconfig_IF_ipv6. These are for backward compatibility.
618 3. A new variable $ipv6_prefer has been added. If NO, IPv6
619 functionality of interfaces with no corresponding
620 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
621 and the default address selection policy of ip6addrctl(8)
622 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
623 Note that if you want to configure IPv6 functionality on the
624 disabled interfaces after boot, first you need to clear the flag by
625 using ifconfig(8) like:
627 ifconfig em0 inet6 -ifdisabled
629 If YES, the default address selection policy is set as
632 The default value of $ipv6_prefer is NO.
634 4. If your system need to receive Router Advertisement messages,
635 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
636 scripts automatically invoke rtsol(8) when the interface becomes
637 UP. The Router Advertisement messages are used for SLAAC
638 (State-Less Address AutoConfiguration).
641 802.11s D3.03 support was committed. This is incompatible with the
642 previous code, which was based on D3.0.
645 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
646 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
647 control whether accepting Router Advertisement messages or not.
648 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
649 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
650 The ifconfig(8) utility now supports these flags.
653 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
654 mount(8) and -a option for df(1) to see them.
657 The old tunable hw.bus.devctl_disable has been superseded by
658 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
659 replaced by hw.bus.devctl_queue=0. The default for this new tunable
663 Remove the option STOP_NMI. The default action is now to use NMI only
664 for KDB via the newly introduced function stop_cpus_hard() and
665 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
668 The stable/8 branch created in subversion. This corresponds to the
669 RELENG_8 branch in CVS.
672 Bump the shared library version numbers for all libraries that do not
673 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
674 __FreeBSD_version to 800105.
677 Due to changes in the implementation of virtual network stack support,
678 all network-related kernel modules must be recompiled. As this change
679 breaks the ABI, bump __FreeBSD_version to 800104.
682 The TOE interface to the TCP syncache has been modified to remove
683 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
684 The cxgb driver is the only TOE consumer affected by this change, and
685 needs to be recompiled along with the kernel. As this change breaks
686 the ABI, bump __FreeBSD_version to 800103.
689 Padding has been added to struct tcpcb, sackhint and tcpstat in
690 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
691 maintaining the ABI. However, this change breaks the ABI, so bump
692 __FreeBSD_version to 800102. User space tools that rely on the size of
693 any of these structs (e.g. sockstat) need to be recompiled.
696 The NFS_LEGACYRPC option has been removed along with the old kernel
697 RPC implementation that this option selected. Kernel configurations
698 may need to be adjusted.
701 The network interface device nodes at /dev/net/<interface> have been
702 removed. All ioctl operations can be performed the normal way using
703 routing sockets. The kqueue functionality can generally be replaced
704 with routing sockets.
707 The documentation from the FreeBSD Documentation Project (Handbook,
708 FAQ, etc.) is now installed via packages by sysinstall(8) and under
709 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
712 The ABI of various structures related to the SYSV IPC API have been
713 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
714 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
718 Layout of struct vnet has changed as routing related variables were
719 moved to their own Vimage module. Modules need to be recompiled. Bump
720 __FreeBSD_version to 800099.
723 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
724 respectively. As long as no more than 16 groups per process are used,
725 no changes should be visible. When more than 16 groups are used, old
726 binaries may fail if they call getgroups() or getgrouplist() with
727 statically sized storage. Recompiling will work around this, but
728 applications should be modified to use dynamically allocated storage
729 for group arrays as POSIX.1-2008 does not cap an implementation's
730 number of supported groups at NGROUPS_MAX+1 as previous versions did.
732 NFS and portalfs mounts may also be affected as the list of groups is
733 truncated to 16. Users of NFS who use more than 16 groups, should
734 take care that negative group permissions are not used on the exported
735 file systems as they will not be reliable unless a GSSAPI based
736 authentication method is used.
739 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
740 option compiles in the support for adaptive spinning for lockmgrs
741 which want to enable it. The lockinit() function now accepts the flag
742 LK_ADAPTIVE in order to make the lock object subject to adaptive
743 spinning when both held in write and read mode.
746 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
747 changed. User applications that use this ioctl need to be rebuilt.
750 The layout of struct thread has changed. Kernel and modules need to
754 The layout of structs ifnet, domain, protosw and vnet_net has changed.
755 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
758 window(1) has been removed from the base system. It can now be
759 installed from ports. The port is called misc/window.
762 The way we are storing and accessing `routing table' entries has
763 changed. Programs reading the FIB, like netstat, need to be
767 A new netisr implementation has been added for FreeBSD 8. Network
768 file system modules, such as igmp, ipdivert, and others, should be
770 Bump __FreeBSD_version to 800096.
773 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
777 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
778 Bump __FreeBSD_version to 800094.
781 Add mnt_xflag field to 'struct mount'. File system modules need to be
783 Bump __FreeBSD_version to 800093.
786 The compiling option ADAPTIVE_SX has been retired while it has been
787 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
788 The KPI for sx_init_flags() changes as accepting flags:
789 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
790 been introduced in order to handle the reversed logic.
791 Bump __FreeBSD_version to 800092.
794 Add support for hierarchical jails. Remove global securelevel.
795 Bump __FreeBSD_version to 800091.
798 The layout of struct vnet_net has changed, therefore modules
800 Bump __FreeBSD_version to 800090.
803 The newly imported zic(8) produces a new format in the output. Please
804 run tzsetup(8) to install the newly created data to /etc/localtime.
807 The sysctl tree for the usb stack has renamed from hw.usb2.* to
808 hw.usb.* and is now consistent again with previous releases.
811 802.11 monitor mode support was revised and driver api's were changed.
812 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
813 of DLT_IEEE802_11. No user-visible data structures were changed but
814 applications that use DLT_IEEE802_11 may require changes.
815 Bump __FreeBSD_version to 800088.
818 The layout of the following structs has changed: sysctl_oid,
819 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
820 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
821 panics may be experienced. World rebuild is required for
822 correctly checking networking state from userland.
823 Bump __FreeBSD_version to 800085.
826 MLDv2 and Source-Specific Multicast (SSM) have been merged
827 to the IPv6 stack. VIMAGE hooks are in but not yet used.
828 The implementation of SSM within FreeBSD's IPv6 stack closely
829 follows the IPv4 implementation.
831 For kernel developers:
833 * The most important changes are that the ip6_output() and
834 ip6_input() paths no longer take the IN6_MULTI_LOCK,
835 and this lock has been downgraded to a non-recursive mutex.
837 * As with the changes to the IPv4 stack to support SSM, filtering
838 of inbound multicast traffic must now be performed by transport
839 protocols within the IPv6 stack. This does not apply to TCP and
840 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
842 * The KPIs used by IPv6 multicast are similar to those used by
843 the IPv4 stack, with the following differences:
844 * im6o_mc_filter() is analogous to imo_multicast_filter().
845 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
846 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
847 * IN6_LOOKUP_MULTI() has been deprecated and removed.
848 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
849 for MLDv1 have an additional 'timer' argument which is used to
850 jitter the initial membership report for the solicited-node
851 multicast membership on-link.
852 * This is not strictly needed for MLDv2, which already jitters
853 its report transmissions. However, the 'timer' argument is
854 preserved in case MLDv1 is active on the interface.
856 * The KAME linked-list based IPv6 membership implementation has
857 been refactored to use a vector similar to that used by the IPv4
859 Code which maintains a list of its own multicast memberships
860 internally, e.g. carp, has been updated to reflect the new
863 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
864 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
865 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
866 implementation constraint which needs to be addressed in HEAD.
868 For application developers:
870 * The changes are broadly similar to those made for the IPv4
873 * The use of IPv4 and IPv6 multicast socket options on the same
874 socket, using mapped addresses, HAS NOT been tested or supported.
876 * There are a number of issues with the implementation of various
877 IPv6 multicast APIs which need to be resolved in the API surface
878 before the implementation is fully compatible with KAME userland
879 use, and these are mostly to do with interface index treatment.
881 * The literature available discusses the use of either the delta / ASM
882 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
883 using setsourcefilter(3)/getsourcefilter(3). For more information
884 please refer to RFC 3768, 'Socket Interface Extensions for
885 Multicast Source Filters'.
887 * Applications which use the published RFC 3678 APIs should be fine.
889 For systems administrators:
891 * The mtest(8) utility has been refactored to support IPv6, in
892 addition to IPv4. Interface addresses are no longer accepted
893 as arguments, their names must be used instead. The utility
894 will map the interface name to its first IPv4 address as
895 returned by getifaddrs(3).
897 * The ifmcstat(8) utility has also been updated to print the MLDv2
898 endpoint state and source filter lists via sysctl(3).
900 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
901 loopback of IPv6 multicast datagrams by default; it defaults to 1
902 to preserve the existing behaviour. Disabling multicast loopback is
903 recommended for optimal system performance.
905 * The IPv6 MROUTING code has been changed to examine this sysctl
906 instead of attempting to perform a group lookup before looping
907 back forwarded datagrams.
909 Bump __FreeBSD_version to 800084.
912 Implement low-level Bluetooth HCI API.
913 Bump __FreeBSD_version to 800083.
916 The layout of struct malloc_type, used by modules to register new
917 memory allocation types, has changed. Most modules will need to
918 be rebuilt or panics may be experienced.
919 Bump __FreeBSD_version to 800081.
922 Anticipate overflowing inp_flags - add inp_flags2.
923 This changes most offsets in inpcb, so checking v4 connection
924 state will require a world rebuild.
925 Bump __FreeBSD_version to 800080.
928 Add an llentry to struct route and struct route_in6. Modules
929 embedding a struct route will need to be recompiled.
930 Bump __FreeBSD_version to 800079.
933 The size of rt_metrics_lite and by extension rtentry has changed.
934 Networking administration apps will need to be recompiled.
935 The route command now supports show as an alias for get, weighting
936 of routes, sticky and nostick flags to alter the behavior of stateful
938 Bump __FreeBSD_version to 800078.
941 Do not use Giant for kbdmux(4) locking. This is wrong and
942 apparently causing more problems than it solves. This will
943 re-open the issue where interrupt handlers may race with
944 kbdmux(4) in polling mode. Typical symptoms include (but
945 not limited to) duplicated and/or missing characters when
946 low level console functions (such as gets) are used while
947 interrupts are enabled (for example geli password prompt,
948 mountroot prompt etc.). Disabling kbdmux(4) may help.
951 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
952 kernel modules referencing any of the above need to be recompiled.
953 Bump __FreeBSD_version to 800075.
956 GEOM_PART has become the default partition slicer for storage devices,
957 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
958 introduces some changes:
960 MSDOS/EBR: the devices created from MSDOS extended partition entries
961 (EBR) can be named differently than with GEOM_MBR and are now symlinks
962 to devices with offset-based names. fstabs may need to be modified.
964 BSD: the "geometry does not match label" warning is harmless in most
965 cases but it points to problems in file system misalignment with
966 disk geometry. The "c" partition is now implicit, covers the whole
967 top-level drive and cannot be (mis)used by users.
969 General: Kernel dumps are now not allowed to be written to devices
970 whose partition types indicate they are meant to be used for file
971 systems (or, in case of MSDOS partitions, as something else than
974 Most of these changes date approximately from 200812.
977 The uscanner(4) driver has been removed from the kernel. This follows
978 Linux removing theirs in 2.6 and making libusb the default interface
982 The multicast forwarding code has been cleaned up. netstat(1)
983 only relies on KVM now for printing bandwidth upcall meters.
984 The IPv4 and IPv6 modules are split into ip_mroute_mod and
985 ip6_mroute_mod respectively. The config(5) options for statically
986 compiling this code remain the same, i.e. 'options MROUTING'.
989 Support for the IFF_NEEDSGIANT network interface flag has been
990 removed, which means that non-MPSAFE network device drivers are no
991 longer supported. In particular, if_ar, if_sr, and network device
992 drivers from the old (legacy) USB stack can no longer be built or
996 POSIX.1 Native Language Support (NLS) has been enabled in libc and
997 a bunch of new language catalog files have also been added.
998 This means that some common libc messages are now localized and
999 they depend on the LC_MESSAGES environmental variable.
1002 The k8temp(4) driver has been renamed to amdtemp(4) since
1003 support for Family 10 and Family 11 CPU families was added.
1006 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1007 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1009 For kernel developers, the most important changes are that the
1010 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1011 and this lock has been downgraded to a non-recursive mutex.
1013 Transport protocols (UDP, Raw IP) are now responsible for filtering
1014 inbound multicast traffic according to group membership and source
1015 filters. The imo_multicast_filter() KPI exists for this purpose.
1016 Transports which do not use multicast (SCTP, TCP) already reject
1017 multicast by default. Forwarding and receive performance may improve
1018 as a mutex acquisition is no longer needed in the ip_input()
1019 low-level input path. in_addmulti() and in_delmulti() are shimmed
1020 to new KPIs which exist to support SSM in-kernel.
1022 For application developers, it is recommended that loopback of
1023 multicast datagrams be disabled for best performance, as this
1024 will still cause the lock to be taken for each looped-back
1025 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1026 be tuned to 0 to disable loopback by default; it defaults to 1
1027 to preserve the existing behaviour.
1029 For systems administrators, to obtain best performance with
1030 multicast reception and multiple groups, it is always recommended
1031 that a card with a suitably precise hash filter is used. Hash
1032 collisions will still result in the lock being taken within the
1033 transport protocol input path to check group membership.
1035 If deploying FreeBSD in an environment with IGMP snooping switches,
1036 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1037 enabled; this forces 224.0.0.0/24 group membership to be announced
1040 The size of 'struct igmpstat' has changed; netstat needs to be
1041 recompiled to reflect this.
1042 Bump __FreeBSD_version to 800070.
1045 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1046 updated to use it. This requires a buildworld/installworld in order to
1047 update the library and dependencies (usbconfig, etc). Its advisable to
1048 rebuild all ports which uses libusb. More specific directions are given
1049 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1050 libusb are no longer required and can be removed.
1053 A workaround is committed to allow the creation of System V shared
1054 memory segment of size > 2 GB on the 64-bit architectures.
1055 Due to a limitation of the existing ABI, the shm_segsz member
1056 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1057 wrong for large segments. Note that limits must be explicitly
1058 raised to allow such segments to be created.
1061 The layout of struct ifnet has changed, requiring a rebuild of all
1062 network device driver modules.
1065 The /dev handling for the new USB stack has changed, a
1066 buildworld/installworld is required for libusb20.
1069 The new USB2 stack has now been permanently moved in and all kernel and
1070 module names reverted to their previous values (eg, usb, ehci, ohci,
1071 ums, ...). The old usb stack can be compiled in by prefixing the name
1072 with the letter 'o', the old usb modules have been removed.
1073 Updating entry 20090216 for xorg and 20090215 for libmap may still
1077 The rc.conf(5) option if_up_delay has been renamed to
1078 defaultroute_delay to better reflect its purpose. If you have
1079 customized this setting in /etc/rc.conf you need to update it to
1083 xorg 7.4 wants to configure its input devices via hald which does not
1084 yet work with USB2. If the keyboard/mouse does not work in xorg then
1086 Option "AllowEmptyInput" "off"
1087 to your ServerLayout section. This will cause X to use the configured
1088 kbd and mouse sections from your xorg.conf.
1091 The GENERIC kernels for all architectures now default to the new USB2
1092 stack. No kernel config options or code have been removed so if a
1093 problem arises please report it and optionally revert to the old USB
1094 stack. If you are loading USB kernel modules or have a custom kernel
1095 that includes GENERIC then ensure that usb names are also changed over,
1096 eg uftdi -> usb2_serial_ftdi.
1098 Older programs linked against the ports libusb 0.1 need to be
1099 redirected to the new stack's libusb20. /etc/libmap.conf can
1101 # Map old usb library to new one for usb2 stack
1102 libusb-0.1.so.8 libusb20.so.1
1105 All USB ethernet devices now attach as interfaces under the name ueN
1106 (eg. ue0). This is to provide a predictable name as vendors often
1107 change usb chipsets in a product without notice.
1110 The ichsmb(4) driver has been changed to require SMBus slave
1111 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1112 All of the other SMBus controller drivers require left-justified
1113 slave addresses, so this change makes all the drivers provide the
1117 INET6 statistics (struct ip6stat) was updated.
1118 netstat(1) needs to be recompiled.
1121 NTFS has been removed from GENERIC kernel on amd64 to match
1122 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1123 will load ntfs.ko module automatically when NTFS support is
1124 actually needed, unless ntfs.ko is not installed or security
1125 level prohibits loading kernel modules. If either is the case,
1126 "options NTFS" has to be added into kernel config.
1129 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1130 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1131 800061. User space tools that rely on the size of struct tcpcb in
1132 tcp_var.h (e.g. sockstat) need to be recompiled.
1135 ng_tty(4) module updated to match the new TTY subsystem.
1136 Due to API change, user-level applications must be updated.
1137 New API support added to mpd5 CVS and expected to be present
1138 in next mpd5.3 release.
1141 With __FreeBSD_version 800060 the makefs tool is part of
1142 the base system (it was a port).
1145 The afdata and ifnet locks have been changed from mutexes to
1146 rwlocks, network modules will need to be re-compiled.
1149 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1150 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1151 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1152 architecture and 40 bytes on 64-bit architecture. The userland
1153 applications "arp" and "ndp" have been updated accordingly.
1154 The output from "netstat -r" shows only routing entries and
1155 none of the L2 information.
1158 __FreeBSD_version 800057 marks the switchover from the
1159 binary ath hal to source code. Users must add the line:
1161 options AH_SUPPORT_AR5416
1163 to their kernel config files when specifying:
1167 The ath_hal module no longer exists; the code is now compiled
1168 together with the driver in the ath module. It is now
1169 possible to tailor chip support (i.e. reduce the set of chips
1170 and thereby the code size); consult ath_hal(4) for details.
1173 __FreeBSD_version 800054 adds memory barriers to
1174 <machine/atomic.h>, new interfaces to ifnet to facilitate
1175 multiple hardware transmit queues for cards that support
1176 them, and a lock-less ring-buffer implementation to
1177 enable drivers to more efficiently manage queueing of
1181 A new version of ZFS (version 13) has been merged to -HEAD.
1182 This version has zpool attribute "listsnapshots" off by
1183 default, which means "zfs list" does not show snapshots,
1184 and is the same as Solaris behavior.
1187 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1190 The uhci, ohci, ehci and slhci USB Host controller drivers have
1191 been put into separate modules. If you load the usb module
1192 separately through loader.conf you will need to load the
1193 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1194 controller add the following to loader.conf:
1200 The ABI used by the PMC toolset has changed. Please keep
1201 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1205 atapci kernel module now includes only generic PCI ATA
1206 driver. AHCI driver moved to ataahci kernel module.
1207 All vendor-specific code moved into separate kernel modules:
1208 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1209 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1210 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1211 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1214 The TTY subsystem of the kernel has been replaced by a new
1215 implementation, which provides better scalability and an
1216 improved driver model. Most common drivers have been migrated to
1217 the new TTY subsystem, while others have not. The following
1218 drivers have not yet been ported to the new TTY layer:
1221 cy, digi, rc, rp, sio
1227 ng_h4, ng_tty, ppp, sl, snp
1229 Adding these drivers to your kernel configuration file shall
1230 cause compilation to fail.
1233 ntpd has been upgraded to 4.2.4p5.
1236 OpenSSH has been upgraded to 5.1p1.
1238 For many years, FreeBSD's version of OpenSSH preferred DSA
1239 over RSA for host and user authentication keys. With this
1240 upgrade, we've switched to the vendor's default of RSA over
1241 DSA. This may cause upgraded clients to warn about unknown
1242 host keys even for previously known hosts. Users should
1243 follow the usual procedure for verifying host keys before
1244 accepting the RSA key.
1246 This can be circumvented by setting the "HostKeyAlgorithms"
1247 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1250 Please note that the sequence of keys offered for
1251 authentication has been changed as well. You may want to
1252 specify IdentityFile in a different order to revert this
1256 The sio(4) driver has been removed from the i386 and amd64
1257 kernel configuration files. This means uart(4) is now the
1258 default serial port driver on those platforms as well.
1260 To prevent collisions with the sio(4) driver, the uart(4) driver
1261 uses different names for its device nodes. This means the
1262 onboard serial port will now most likely be called "ttyu0"
1263 instead of "ttyd0". You may need to reconfigure applications to
1264 use the new device names.
1266 When using the serial port as a boot console, be sure to update
1267 /boot/device.hints and /etc/ttys before booting the new kernel.
1268 If you forget to do so, you can still manually specify the hints
1269 at the loader prompt:
1271 set hint.uart.0.at="isa"
1272 set hint.uart.0.port="0x3F8"
1273 set hint.uart.0.flags="0x10"
1274 set hint.uart.0.irq="4"
1278 The gpt(8) utility has been removed. Use gpart(8) to partition
1282 The version that Linuxulator emulates was changed from 2.4.2
1283 to 2.6.16. If you experience any problems with Linux binaries
1284 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1285 if it fixes the problem contact emulation mailing list.
1288 ISDN4BSD (I4B) was removed from the src tree. You may need to
1289 update a your kernel configuration and remove relevant entries.
1292 I have checked in code to support multiple routing tables.
1293 See the man pages setfib(1) and setfib(2).
1294 This is a hopefully backwards compatible version,
1295 but to make use of it you need to compile your kernel
1296 with options ROUTETABLES=2 (or more up to 16).
1299 The 802.11 wireless support was redone to enable multi-bss
1300 operation on devices that are capable. The underlying device
1301 is no longer used directly but instead wlanX devices are
1302 cloned with ifconfig. This requires changes to rc.conf files.
1303 For example, change:
1304 ifconfig_ath0="WPA DHCP"
1307 ifconfig_wlan0="WPA DHCP"
1308 see rc.conf(5) for more details. In addition, mergemaster of
1309 /etc/rc.d is highly recommended. Simultaneous update of userland
1310 and kernel wouldn't hurt either.
1312 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1313 modules were merged into the base wlan module. All references
1314 to these modules (e.g. in kernel config files) must be removed.
1317 psm(4) has gained write(2) support in native operation level.
1318 Arbitrary commands can be written to /dev/psm%d and status can
1319 be read back from it. Therefore, an application is responsible
1320 for status validation and error recovery. It is a no-op in
1321 other operation levels.
1324 Support for KSE threading has been removed from the kernel. To
1325 run legacy applications linked against KSE libmap.conf may
1326 be used. The following libmap.conf may be used to ensure
1327 compatibility with any prior release:
1329 libpthread.so.1 libthr.so.1
1330 libpthread.so.2 libthr.so.2
1331 libkse.so.3 libthr.so.3
1334 The layout of struct vmspace has changed. This affects libkvm
1335 and any executables that link against libkvm and use the
1336 kvm_getprocs() function. In particular, but not exclusively,
1337 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1338 The effects are minimal, but it's advisable to upgrade world
1342 The latest em driver no longer has support in it for the
1343 82575 adapter, this is now moved to the igb driver. The
1344 split was done to make new features that are incompatible
1345 with older hardware easier to do.
1348 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1349 likewise the kernel option is now GEOM_LINUX_LVM.
1352 The default NFS mount mode has changed from UDP to TCP for
1353 increased reliability. If you rely on (insecurely) NFS
1354 mounting across a firewall you may need to update your
1358 Belatedly note the addition of m_collapse for compacting
1362 The fts(3) structures have been changed to use adequate
1363 integer types for their members and so to be able to cope
1364 with huge file trees. The old fts(3) ABI is preserved
1365 through symbol versioning in libc, so third-party binaries
1366 using fts(3) should still work, although they will not take
1367 advantage of the extended types. At the same time, some
1368 third-party software might fail to build after this change
1369 due to unportable assumptions made in its source code about
1370 fts(3) structure members. Such software should be fixed
1371 by its vendor or, in the worst case, in the ports tree.
1372 FreeBSD_version 800015 marks this change for the unlikely
1373 case that a portable fix is impossible.
1376 To upgrade to -current after this date, you must be running
1377 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1378 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1381 The ADAPTIVE_GIANT kernel option has been retired because its
1382 functionality is the default now.
1385 The AT keyboard emulation of sunkbd(4) has been turned on
1386 by default. In order to make the special symbols of the Sun
1387 keyboards driven by sunkbd(4) work under X these now have
1388 to be configured the same way as Sun USB keyboards driven
1389 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1391 Option "XkbLayout" "us"
1392 Option "XkbRules" "xorg"
1393 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1396 It has been decided that it is desirable to provide ABI
1397 backwards compatibility to the FreeBSD 4/5/6 versions of the
1398 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1399 broken with the introduction of PCI domain support (see the
1400 20070930 entry). Unfortunately, this required the ABI of
1401 PCIOCGETCONF to be broken again in order to be able to
1402 provide backwards compatibility to the old version of that
1403 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1404 again. As for prominent ports this affects neither pciutils
1405 nor xorg-server this time, the hal port needs to be rebuilt
1409 The misnamed kthread_create() and friends have been renamed
1410 to kproc_create() etc. Many of the callers already
1411 used kproc_start()..
1412 I will return kthread_create() and friends in a while
1413 with implementations that actually create threads, not procs.
1414 Renaming corresponds with version 800002.
1423 Avoid using make -j when upgrading. While generally safe, there are
1424 sometimes problems using -j to upgrade. If your upgrade fails with
1425 -j, please try again without -j. From time to time in the past there
1426 have been problems using -j with buildworld and/or installworld. This
1427 is especially true when upgrading between "distant" versions (eg one
1428 that cross a major release boundary or several minor releases, or when
1429 several months have passed on the -current branch).
1431 Sometimes, obscure build problems are the result of environment
1432 poisoning. This can happen because the make utility reads its
1433 environment when searching for values for global variables. To run
1434 your build attempts in an "environmental clean room", prefix all make
1435 commands with 'env -i '. See the env(1) manual page for more details.
1437 When upgrading from one major version to another it is generally best
1438 to upgrade to the latest code in the currently installed branch first,
1439 then do an upgrade to the new branch. This is the best-tested upgrade
1440 path, and has the highest probability of being successful. Please try
1441 this approach before reporting problems with a major version upgrade.
1445 When upgrading the boot ZFS pool to a new version, always follow
1448 1.) recompile and reinstall the ZFS boot loader and boot block
1449 (this is part of "make buildworld" and "make installworld")
1451 2.) update the ZFS boot block on your boot drive
1453 The following example updates the ZFS boot block on the first
1454 partition (freebsd-boot) of a GPT partitioned drive ad0:
1455 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1457 Non-boot pools do not need these updates.
1461 If you are updating from a prior version of FreeBSD (even one just
1462 a few days old), you should follow this procedure. It is the most
1463 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1465 make kernel-toolchain
1466 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1467 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1469 To test a kernel once
1470 ---------------------
1471 If you just want to boot a kernel once (because you are not sure
1472 if it works, or if you want to boot a known bad kernel to provide
1473 debugging information) run
1474 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1475 nextboot -k testkernel
1477 To just build a kernel when you know that it won't mess you up
1478 --------------------------------------------------------------
1479 This assumes you are already running a CURRENT system. Replace
1480 ${arch} with the architecture of your machine (e.g. "i386",
1481 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1483 cd src/sys/${arch}/conf
1484 config KERNEL_NAME_HERE
1485 cd ../compile/KERNEL_NAME_HERE
1490 If this fails, go to the "To build a kernel" section.
1492 To rebuild everything and install it on the current system.
1493 -----------------------------------------------------------
1494 # Note: sometimes if you are running current you gotta do more than
1495 # is listed here if you are upgrading from a really old current.
1497 <make sure you have good level 0 dumps>
1499 make kernel KERNCONF=YOUR_KERNEL_HERE
1501 <reboot in single user> [3]
1509 To cross-install current onto a separate partition
1510 --------------------------------------------------
1511 # In this approach we use a separate partition to hold
1512 # current's root, 'usr', and 'var' directories. A partition
1513 # holding "/", "/usr" and "/var" should be about 2GB in
1516 <make sure you have good level 0 dumps>
1519 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1520 <maybe newfs current's root partition>
1521 <mount current's root partition on directory ${CURRENT_ROOT}>
1522 make installworld DESTDIR=${CURRENT_ROOT}
1523 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1524 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1525 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1526 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1527 <reboot into current>
1528 <do a "native" rebuild/install as described in the previous section>
1529 <maybe install compatibility libraries from ports/misc/compat*>
1533 To upgrade in-place from stable to current
1534 ----------------------------------------------
1535 <make sure you have good level 0 dumps>
1537 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1539 <reboot in single user> [3]
1546 Make sure that you've read the UPDATING file to understand the
1547 tweaks to various things you need. At this point in the life
1548 cycle of current, things change often and you are on your own
1549 to cope. The defaults can also change, so please read ALL of
1550 the UPDATING entries.
1552 Also, if you are tracking -current, you must be subscribed to
1553 freebsd-current@freebsd.org. Make sure that before you update
1554 your sources that you have read and understood all the recent
1555 messages there. If in doubt, please track -stable which has
1556 much fewer pitfalls.
1558 [1] If you have third party modules, such as vmware, you
1559 should disable them at this point so they don't crash your
1562 [3] From the bootblocks, boot -s, and then do
1567 adjkerntz -i # if CMOS is wall time
1568 Also, when doing a major release upgrade, it is required that
1569 you boot into single user mode to do the installworld.
1571 [4] Note: This step is non-optional. Failure to do this step
1572 can result in a significant reduction in the functionality of the
1573 system. Attempting to do it by hand is not recommended and those
1574 that pursue this avenue should read this file carefully, as well
1575 as the archives of freebsd-current and freebsd-hackers mailing lists
1576 for potential gotchas. The -U option is also useful to consider.
1577 See mergemaster(8) for more information.
1579 [5] Usually this step is a noop. However, from time to time
1580 you may need to do this if you get unknown user in the following
1581 step. It never hurts to do it all the time. You may need to
1582 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1583 install) after the buildworld before this step if you last updated
1584 from [78]-stable or 9-stable before 20130430.
1586 [6] This only deletes old files and directories. Old libraries
1587 can be deleted by "make delete-old-libs", but you have to make
1588 sure that no program is using those libraries anymore.
1590 [8] In order to have a kernel that can run the 4.x binaries needed to
1591 do an installworld, you must include the COMPAT_FREEBSD4 option in
1592 your kernel. Failure to do so may leave you with a system that is
1593 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1594 required to run the 5.x binaries on more recent kernels. And so on
1595 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1597 Make sure that you merge any new devices from GENERIC since the
1598 last time you updated your kernel config file.
1600 [9] When checking out sources, you must include the -P flag to have
1601 cvs prune empty directories.
1603 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1604 "?=" instead of the "=" assignment operator, so that buildworld can
1605 override the CPUTYPE if it needs to.
1607 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1608 not on the command line, or in /etc/make.conf. buildworld will
1609 warn if it is improperly defined.
1612 This file contains a list, in reverse chronological order, of major
1613 breakages in tracking -current. It is not guaranteed to be a complete
1614 list of such breakages, and only contains entries since October 10, 2007.
1615 If you need to see UPDATING entries from before that date, you will need
1616 to fetch an UPDATING file from an older FreeBSD release.
1618 Copyright information:
1620 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1622 Redistribution, publication, translation and use, with or without
1623 modification, in full or in part, in any form or format of this
1624 document are permitted without further permission from the author.
1626 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1627 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1628 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1629 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1630 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1631 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1632 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1633 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1634 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1635 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1636 POSSIBILITY OF SUCH DAMAGE.
1638 Contact Warner Losh if you have any questions about your use of