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 hastctl(8)'s `status' command output changed to terse one-liner format.
16 Scripts using this should switch to `list' command or be rewritten.
19 Fix a bug that allowed a tracing process (e.g. gdb) to write
20 to a memory-mapped file in the traced process's address space
21 even if neither the traced process nor the tracing process had
22 write access to that file.
25 Added ZFS TRIM support which is enabled by default. To disable
26 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
28 Creating new ZFS pools and adding new devices to existing pools
29 first performs a full device level TRIM which can take a significant
30 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
31 to disable this behaviour.
33 ZFS TRIM requires the underlying device support BIO_DELETE which
34 is currently provided by methods such as ATA TRIM and SCSI UNMAP
35 via CAM, which are typically supported by SSD's.
37 Stats for ZFS TRIM can be monitored by looking at the sysctl's
38 under kstat.zfs.misc.zio_trim.
41 `list' command has been added to hastctl(8). For now, it is full
42 equivalent of `status' command.
43 WARNING: in the near future the output of hastctl's status command
44 will change to more terse format. If you use `hastctl status'
45 for parsing in your scripts, switch to `hastctl list'.
48 The mergemaster command now uses the default MAKEOBJDIRPREFIX
49 rather than creating it's own in the temporary directory in
50 order allow access to bootstrapped versions of tools such as
51 install and mtree. When upgrading from version of FreeBSD where
52 the install command does not support -l, you will need to
53 install a new mergemaster command if mergemaster -p is required.
54 This can be accomplished with the command (cd src/usr.sbin/mergemaster
57 Due to the use of the new -l option to install(1) during build
58 and install, you must take care not to directly set the INSTALL
59 make variable in your /etc/make.conf, /etc/src.conf, or on the
60 command line. If you wish to use the -C flag for all installs
61 you may be able to add INSTALL+=-C to /etc/make.conf or
65 Fix a bug that allows NFS clients to issue READDIR on files.
68 The install(1) option -M has changed meaning and now takes an
69 argument that is a file or path to append logs to. In the
70 unlikely event that -M was the last option on the command line
71 and the command line contained at least two files and a target
72 directory the first file will have logs appended to it. The -M
73 option served little practical purpose in the last decade so it's
74 used expected to be extremely rare.
77 A new compression method (lz4) has been merged to. Please refer to
78 zpool-features(7) for more information.
80 Please refer to the "ZFS notes" section of this file for information
81 on upgrading boot ZFS pools.
84 The VFS KBI was changed with the merge of several nullfs
85 optimizations and fixes. All filesystem modules must be
89 With the addition of auditdistd(8), a new auditdistd user is now
90 depended on during installworld. "mergemaster -p" can be used to add
91 the user prior to installworld, as documented in the handbook.
97 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
98 Starting with this version the old system of ZFS pool versioning
99 is superseded by "feature flags". This concept enables forward
100 compatibility against certain future changes in functionality of ZFS
101 pools. The first two read-only compatible "feature flags" for ZFS
102 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
103 For more information read the new zpool-features(7) manual page.
104 Please refer to the "ZFS notes" section of this file for information
105 on upgrading boot ZFS pools.
108 The commit introducing bsd.compiler.mk breaks the traditional
109 building of kernels before this point. Add -m ${SRC}/share/mk
110 (for the right value of SRC) to your command lines to work
111 around; update your useland to a point after this; or use the
112 buildkernel/installkernel top-level targets. See also 20120829.
115 The IPFIREWALL_FORWARD kernel option has been removed. Its
116 functionality now turned on by default.
119 The random(4) support for the VIA hardware random number
120 generator (`PADLOCK') is no longer enabled unconditionally.
121 Add the PADLOCK_RNG option in the custom kernel config if
122 needed. The GENERIC kernels on i386 and amd64 do include the
123 option, so the change only affects the custom kernel
127 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
128 the traditional method, you must update your system with an installworld
129 before the kernel will build. The documented make buildkernel/installkernel
130 interfaces (coupled with fresh make kernel-toolchain) continue to work.
133 The sparc64 ZFS loader has been changed to no longer try to auto-
134 detect ZFS providers based on diskN aliases but now requires these
135 to be explicitly listed in the OFW boot-device environment variable.
138 Now unix domain sockets behave "as expected" on nullfs(5). Previously
139 nullfs(5) did not pass through all behaviours to the underlying layer,
140 as a result if we bound to a socket on the lower layer we could connect
141 only to the lower path; if we bound to the upper layer we could connect
142 only to the upper path. The new behavior is one can connect to both the
143 lower and the upper paths regardless what layer path one binds to.
146 The acpi_wmi(4) status device /dev/wmistat has been renamed to
150 A new VOP_ADVISE() was added to support posix_fadvise(2). All
151 filesystem modules must be recompiled.
154 The interface of the VOP_VPTOCNP(9) changed, now the returned
155 vnode shall be referenced, previously it was required to be
156 only held. All in-tree filesystems are converted.
162 The broken amd(4) driver has been replaced with esp(4) in the amd64,
163 i386 and pc98 GENERIC kernel configuration files.
166 This commit modifies vfs_register() so that it uses a hash
167 calculation to set vfc_typenum, which is enabled by default.
168 The first time a system is booted after this change, the
169 vfc_typenum values will change for all file systems. The
170 main effect of this is a change to the NFS server file handles
171 for file systems that use vfc_typenum in their fsid, such as ZFS.
172 It will, however, prevent vfc_typenum from changing when file
173 systems are loaded in a different order for subsequent reboots.
174 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
175 until you are ready to remount all NFS clients after a reboot.
178 Bump the shared library version numbers for libraries that
179 do not use symbol versioning, have changed the ABI compared
180 to stable/8 and which shared library version was not bumped.
181 Done as part of 9.0-RELEASE cycle.
184 During the merge of Capsicum features, the fget(9) KPI was modified.
185 This may require the rebuilding of out-of-tree device drivers --
186 issues have been reported specifically with the nVidia device driver.
187 __FreeBSD_version is bumped to 900041.
189 Also, there is a period between 20110811 and 20110814 where the
190 special devices /dev/{stdin,stdout,stderr} did not work correctly.
191 Building world from a kernel during that window may not work.
194 The packet filter (pf) code has been updated to OpenBSD 4.5.
195 You need to update userland tools to be in sync with kernel.
196 This update breaks backward compatibility with earlier pfsync(4)
197 versions. Care must be taken when updating redundant firewall setups.
200 The following sysctls and tunables are retired on x86 platforms:
202 machdep.hlt_logical_cpus
203 The following sysctl is retired:
204 machdep.hyperthreading_allowed
205 The sysctls were supposed to provide a way to dynamically offline and
206 online selected CPUs on x86 platforms, but the implementation has not
207 been reliable especially with SCHED_ULE scheduler.
208 machdep.hyperthreading_allowed tunable is still available to ignore
209 hyperthreading CPUs at OS level.
210 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
211 where X is an APIC ID of a CPU. Be advised, though, that disabling
212 CPUs in non-uniform fashion will result in non-uniform topology and
213 may lead to sub-optimal system performance with SCHED_ULE, which is
217 cpumask_t type is retired and cpuset_t is used in order to describe
221 Changes to ifconfig(8) for dynamic address family detection mandate
222 that you are running a kernel of 20110525 or later. Make sure to
223 follow the update procedure to boot a new kernel before installing
227 Support for sun4v architecture is officially dropped
230 Several KPI breaking changes have been committed to the mii(4) layer,
231 the PHY drivers and consequently some Ethernet drivers using mii(4).
232 This means that miibus.ko and the modules of the affected Ethernet
233 drivers need to be recompiled.
235 Note to kernel developers: Given that the OUI bit reversion problem
236 was fixed as part of these changes all mii(4) commits related to OUIs,
237 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
238 handling, no longer can be merged verbatim to stable/8 and previous
242 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
243 into their kernel configurations along with 'device pci'.
246 The default NFS client is now the new NFS client, so fstype "newnfs"
247 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
248 Although mounts via fstype "nfs" will usually work without userland
249 changes, it is recommended that the mount(8) and mount_nfs(8)
250 commands be rebuilt from sources and that a link to mount_nfs called
251 mount_oldnfs be created. The new client is compiled into the
252 kernel with "options NFSCL" and this is needed for diskless root
253 file systems. The GENERIC kernel configs have been changed to use
254 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
255 To use the regular/old client, you can "mount -t oldnfs ...". For
256 a diskless root file system, you must also include a line like:
258 vfs.root.mountfrom="oldnfs:"
260 in the boot/loader.conf on the root fs on the NFS server to make
261 a diskless root fs use the old client.
264 The GENERIC kernels for all architectures now default to the new
265 CAM-based ATA stack. It means that all legacy ATA drivers were
266 removed and replaced by respective CAM drivers. If you are using
267 ATA device names in /etc/fstab or other places, make sure to update
268 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
269 where 'Y's are the sequential numbers starting from zero for each type
270 in order of detection, unless configured otherwise with tunables,
271 see cam(4)). There will be symbolic links created in /dev/ to map
272 old adX devices to the respective adaY. They should provide basic
273 compatibility for file systems mounting in most cases, but they do
274 not support old user-level APIs and do not have respective providers
275 in GEOM. Consider using updated management tools with new device names.
277 It is possible to load devices ahci, ata, siis and mvs as modules,
278 but option ATA_CAM should remain in kernel configuration to make ata
279 module work as CAM driver supporting legacy ATA controllers. Device ata
280 still can be used in modular fashion (atacore + ...). Modules atadisk
281 and atapi* are not used and won't affect operation in ATA_CAM mode.
282 Note that to use CAM-based ATA kernel should include CAM devices
283 scbus, pass, da (or explicitly ada), cd and optionally others. All of
284 them are parts of the cam module.
286 ataraid(4) functionality is now supported by the RAID GEOM class.
287 To use it you can load geom_raid kernel module and use graid(8) tool
288 for management. Instead of /dev/arX device names, use /dev/raid/rX.
290 No kernel config options or code have been removed, so if a problem
291 arises, please report it and optionally revert to the old ATA stack.
292 In order to do it you can remove from the kernel config:
297 , and instead add back:
298 device atadisk # ATA disk drives
299 device ataraid # ATA RAID drives
300 device atapicd # ATAPI CDROM drives
301 device atapifd # ATAPI floppy drives
302 device atapist # ATAPI tape drives
305 The default NFS server has been changed to the new server, which
306 was referred to as the experimental server. If you need to switch
307 back to the old NFS server, you must now put the "-o" option on
308 both the mountd and nfsd commands. This can be done using the
309 mountd_flags and nfs_server_flags rc.conf variables until an
310 update to the rc scripts is committed, which is coming soon.
313 The GNU Objective-C runtime library (libobjc), and other Objective-C
314 related components have been removed from the base system. If you
315 require an Objective-C library, please use one of the available ports.
318 ath(4) has been split into bus- and device- modules. if_ath contains
319 the HAL, the TX rate control and the network device code. if_ath_pci
320 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
321 contains the AHB glue. Users need to load both if_ath_pci and if_ath
322 in order to use ath on everything else.
324 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
325 need to load if_ath and if_ath_pci for ath(4) operation.
328 As part of the replacement of sysinstall, the process of building
329 release media has changed significantly. For details, please re-read
330 release(7), which has been updated to reflect the new build process.
333 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
334 is the last available version under GPLv2. It brings a number of new
335 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
336 4.1 and SSE 4.2), better support for powerpc64, a number of new
337 directives, and lots of other small improvements. See the ChangeLog
338 file in contrib/binutils for the full details.
341 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
342 compliant, and will now use half of hash for authentication.
343 This will break interoperability with all stacks (including all
344 actual FreeBSD versions) who implement
345 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
347 The only workaround with such peers is to use another HMAC
348 algorithm for IPsec ("phase 2") authentication.
351 Remove the uio_yield prototype and symbol. This function has
352 been misnamed since it was introduced and should not be
353 globally exposed with this name. The equivalent functionality
354 is now available using kern_yield(curthread->td_user_pri).
355 The function remains undocumented.
358 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
359 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
360 for scalar sysctls is defined but disabled. Code that needs
361 UQUAD to pass the type checking that must compile on older
362 systems where the define is not present can check against
363 __FreeBSD_version >= 900030.
365 The system dialog(1) has been replaced with a new version previously
366 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
367 with the previous version, but the libdialog associated with it has
368 a largely incompatible API. As such, the original version of libdialog
369 will be kept temporarily as libodialog, until its base system consumers
370 are replaced or updated. Bump __FreeBSD_version to 900030.
373 If you are trying to run make universe on a -stable system, and you get
374 the following warning:
375 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
376 unknown. config(8) likely too old."
377 or something similar to it, then you must upgrade your -stable system
378 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
379 stable/8) or build the config from the latest stable/8 branch and
380 install it on your system.
382 Prior to this date, building a current universe on 8-stable system from
383 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
384 error in the first kernel build phase. A new config on those old
385 systems will fix that problem for older versions of -current.
388 The TCP stack has been modified to allow Khelp modules to interact with
389 it via helper hook points and store per-connection data in the TCP
390 control block. Bump __FreeBSD_version to 900029. User space tools that
391 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
395 Generic IEEE 802.3 annex 31B full duplex flow control support has been
396 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
397 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
398 to take advantage of it instead of using custom implementations. This
399 means that these drivers now no longer unconditionally advertise
400 support for flow control but only do so if flow control is a selected
401 media option. This was implemented in the generic support that way in
402 order to allow flow control to be switched on and off via ifconfig(8)
403 with the PHY specific default to typically off in order to protect
404 from unwanted effects. Consequently, if you used flow control with
405 one of the above mentioned drivers you now need to explicitly enable
407 ifconfig bge0 media auto mediaopt flowcontrol
409 Along with the above mentioned changes generic support for setting
410 1000baseT master mode also has been added and brgphy(4), ciphy(4),
411 e1000phy(4) as well as ip1000phy(4) have been converted to take
412 advantage of it. This means that these drivers now no longer take the
413 link0 parameter for selecting master mode but the master media option
414 has to be used instead, for example like in the following:
415 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
417 Selection of master mode now is also available with all other PHY
418 drivers supporting 1000baseT.
421 The TCP stack has received a significant update to add support for
422 modularised congestion control and generally improve the clarity of
423 congestion control decisions. Bump __FreeBSD_version to 900025. User
424 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
425 sockstat) need to be recompiled.
428 The man(1) utility has been replaced by a new version that no longer
429 uses /etc/manpath.config. Please consult man.conf(5) for how to
430 migrate local entries to the new format.
433 The copyright strings printed by login(1) and sshd(8) at the time of a
434 new connection have been removed to follow other operating systems and
438 A workaround for a fixed ld bug has been removed in kernel code,
439 so make sure that your system ld is built from sources after
440 revision 210245 from 2010-07-19 (r211583 if building head kernel
441 on stable/8, r211584 for stable/7; both from 2010-08-21).
442 A symptom of incorrect ld version is different addresses for
443 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
446 The $ipv6_prefer variable in rc.conf(5) has been split into
447 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
449 The $ip6addrctl_policy is a variable to choose a pre-defined
450 address selection policy set by ip6addrctl(8). A value
451 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
454 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
455 flag (see an entry of 20090926) is set on an interface with no
456 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
457 security reason. If you want IPv6 link-local address on all
458 interfaces by default, set this to "YES".
460 The old ipv6_prefer="YES" is equivalent to
461 ipv6_activate_all_interfaces="YES" and
462 ip6addrctl_policy="ipv6_prefer".
465 DTrace has grown support for userland tracing. Due to this, DTrace is
466 now i386 and amd64 only.
467 dtruss(1) is now installed by default on those systems and a new
468 kernel module is needed for userland tracing: fasttrap.
469 No changes to your kernel config file are necessary to enable
470 userland tracing, but you might consider adding 'STRIP=' and
471 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
472 to have informative userland stack traces in DTrace (ustack).
475 The acpi_aiboost(4) driver has been removed in favor of the new
476 aibs(4) driver. You should update your kernel configuration file.
479 BSD grep has been imported to the base system and it is built by
480 default. It is completely BSD licensed, highly GNU-compatible, uses
481 less memory than its GNU counterpart and has a small codebase.
482 However, it is slower than its GNU counterpart, which is mostly
483 noticeable for larger searches, for smaller ones it is measurable
484 but not significant. The reason is complex, the most important factor
485 is that we lack a modern and efficient regex library and GNU
486 overcomes this by optimizing the searches internally. Future work
487 on improving the regex performance is planned, for the meantime,
488 users that need better performance, can build GNU grep instead by
489 setting the WITH_GNU_GREP knob.
492 Due to the import of powerpc64 support, all existing powerpc kernel
493 configuration files must be updated with a machine directive like this:
494 machine powerpc powerpc
496 In addition, an updated config(8) is required to build powerpc kernels
500 A new version of ZFS (version 15) has been merged to -HEAD.
501 This version uses a python library for the following subcommands:
502 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
503 For full functionality of these commands the following port must
504 be installed: sysutils/py-zfs
507 'vm_page's are now hashed by physical address to an array of mutexes.
508 Currently this is only used to serialize access to hold_count. Over
509 time the page queue mutex will be peeled away. This changes the size
510 of pmap on every architecture. And requires all callers of vm_page_hold
511 and vm_page_unhold to be updated.
514 WITH_CTF can now be specified in src.conf (not recommended, there
515 are some problems with static executables), make.conf (would also
516 affect ports which do not use GNU make and do not override the
517 compile targets) or in the kernel config (via "makeoptions
519 When WITH_CTF was specified there before this was silently ignored,
520 so make sure that WITH_CTF is not used in places which could lead
521 to unwanted behavior.
524 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
525 to allow 32-bit compatibility on non-x86 platforms. All kernel
526 configurations on amd64 and ia64 platforms using these options must
527 be modified accordingly.
530 The utmp user accounting database has been replaced with utmpx,
531 the user accounting interface standardized by POSIX.
532 Unfortunately the semantics of utmp and utmpx don't match,
533 making it practically impossible to support both interfaces.
534 The user accounting database is used by tools like finger(1),
535 last(1), talk(1), w(1) and ac(8).
537 All applications in the base system use utmpx. This means only
538 local binaries (e.g. from the ports tree) may still use these
539 utmp database files. These applications must be rebuilt to make
542 After the system has been upgraded, it is safe to remove the old
543 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
544 assuming their contents is of no importance anymore. Old wtmp
545 databases can only be used by last(1) and ac(8) after they have
546 been converted to the new format using wtmpcvt(1).
549 Introduce the kernel thread "deadlock resolver" (which can be enabled
550 via the DEADLKRES option, see NOTES for more details) and the
551 sleepq_type() function for sleepqueues.
554 The rc.firewall and rc.firewall6 were unified, and
555 rc.firewall6 and rc.d/ip6fw were removed.
556 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
557 variables are obsoleted. Instead, the following new rc
558 variables are added to rc.d/ipfw:
560 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
561 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
562 firewall_simple_onet_ipv6, firewall_trusted_ipv6
564 The meanings correspond to the relevant IPv4 variables.
570 The default terminal emulation for syscons(4) has been changed
571 from cons25 to xterm on all platforms except pc98. This means
572 that the /etc/ttys file needs to be updated to ensure correct
573 operation of applications on the console.
575 The terminal emulation style can be toggled per window by using
576 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
577 options can be used to change the compile-time default back to
580 To prevent graphical artifacts, make sure the TERM environment
581 variable is set to match the terminal emulation that is being
582 performed by syscons(4).
585 The layout of the structure ieee80211req_scan_result has changed.
586 Applications that require wireless scan results (e.g. ifconfig(8))
587 from net80211 need to be recompiled.
589 Applications such as wpa_supplicant(8) may require a full world
590 build without using NO_CLEAN in order to get synchronized with the
594 The iwn(4) driver has been updated to support the 5000 and 5150 series.
595 There's one kernel module for each firmware. Adding "device iwnfw"
596 to the kernel configuration file means including all three firmware
597 images inside the kernel. If you want to include just the one for
598 your wireless card, use the devices iwn4965fw, iwn5000fw or
602 The rc.d/network_ipv6, IPv6 configuration script has been integrated
603 into rc.d/netif. The changes are the following:
605 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
606 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
607 Note that both variables need the "inet6" keyword at the head.
609 Do not set $ipv6_network_interfaces manually if you do not
610 understand what you are doing. It is not needed in most cases.
612 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
615 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
616 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
618 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
619 all configured interfaces have "inet6 accept_rtadv" in the
620 $ifconfig_IF_ipv6. These are for backward compatibility.
622 3. A new variable $ipv6_prefer has been added. If NO, IPv6
623 functionality of interfaces with no corresponding
624 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
625 and the default address selection policy of ip6addrctl(8)
626 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
627 Note that if you want to configure IPv6 functionality on the
628 disabled interfaces after boot, first you need to clear the flag by
629 using ifconfig(8) like:
631 ifconfig em0 inet6 -ifdisabled
633 If YES, the default address selection policy is set as
636 The default value of $ipv6_prefer is NO.
638 4. If your system need to receive Router Advertisement messages,
639 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
640 scripts automatically invoke rtsol(8) when the interface becomes
641 UP. The Router Advertisement messages are used for SLAAC
642 (State-Less Address AutoConfiguration).
645 802.11s D3.03 support was committed. This is incompatible with the
646 previous code, which was based on D3.0.
649 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
650 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
651 control whether accepting Router Advertisement messages or not.
652 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
653 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
654 The ifconfig(8) utility now supports these flags.
657 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
658 mount(8) and -a option for df(1) to see them.
661 The old tunable hw.bus.devctl_disable has been superseded by
662 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
663 replaced by hw.bus.devctl_queue=0. The default for this new tunable
667 Remove the option STOP_NMI. The default action is now to use NMI only
668 for KDB via the newly introduced function stop_cpus_hard() and
669 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
672 The stable/8 branch created in subversion. This corresponds to the
673 RELENG_8 branch in CVS.
676 Bump the shared library version numbers for all libraries that do not
677 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
678 __FreeBSD_version to 800105.
681 Due to changes in the implementation of virtual network stack support,
682 all network-related kernel modules must be recompiled. As this change
683 breaks the ABI, bump __FreeBSD_version to 800104.
686 The TOE interface to the TCP syncache has been modified to remove
687 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
688 The cxgb driver is the only TOE consumer affected by this change, and
689 needs to be recompiled along with the kernel. As this change breaks
690 the ABI, bump __FreeBSD_version to 800103.
693 Padding has been added to struct tcpcb, sackhint and tcpstat in
694 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
695 maintaining the ABI. However, this change breaks the ABI, so bump
696 __FreeBSD_version to 800102. User space tools that rely on the size of
697 any of these structs (e.g. sockstat) need to be recompiled.
700 The NFS_LEGACYRPC option has been removed along with the old kernel
701 RPC implementation that this option selected. Kernel configurations
702 may need to be adjusted.
705 The network interface device nodes at /dev/net/<interface> have been
706 removed. All ioctl operations can be performed the normal way using
707 routing sockets. The kqueue functionality can generally be replaced
708 with routing sockets.
711 The documentation from the FreeBSD Documentation Project (Handbook,
712 FAQ, etc.) is now installed via packages by sysinstall(8) and under
713 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
716 The ABI of various structures related to the SYSV IPC API have been
717 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
718 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
722 Layout of struct vnet has changed as routing related variables were
723 moved to their own Vimage module. Modules need to be recompiled. Bump
724 __FreeBSD_version to 800099.
727 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
728 respectively. As long as no more than 16 groups per process are used,
729 no changes should be visible. When more than 16 groups are used, old
730 binaries may fail if they call getgroups() or getgrouplist() with
731 statically sized storage. Recompiling will work around this, but
732 applications should be modified to use dynamically allocated storage
733 for group arrays as POSIX.1-2008 does not cap an implementation's
734 number of supported groups at NGROUPS_MAX+1 as previous versions did.
736 NFS and portalfs mounts may also be affected as the list of groups is
737 truncated to 16. Users of NFS who use more than 16 groups, should
738 take care that negative group permissions are not used on the exported
739 file systems as they will not be reliable unless a GSSAPI based
740 authentication method is used.
743 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
744 option compiles in the support for adaptive spinning for lockmgrs
745 which want to enable it. The lockinit() function now accepts the flag
746 LK_ADAPTIVE in order to make the lock object subject to adaptive
747 spinning when both held in write and read mode.
750 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
751 changed. User applications that use this ioctl need to be rebuilt.
754 The layout of struct thread has changed. Kernel and modules need to
758 The layout of structs ifnet, domain, protosw and vnet_net has changed.
759 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
762 window(1) has been removed from the base system. It can now be
763 installed from ports. The port is called misc/window.
766 The way we are storing and accessing `routing table' entries has
767 changed. Programs reading the FIB, like netstat, need to be
771 A new netisr implementation has been added for FreeBSD 8. Network
772 file system modules, such as igmp, ipdivert, and others, should be
774 Bump __FreeBSD_version to 800096.
777 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
781 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
782 Bump __FreeBSD_version to 800094.
785 Add mnt_xflag field to 'struct mount'. File system modules need to be
787 Bump __FreeBSD_version to 800093.
790 The compiling option ADAPTIVE_SX has been retired while it has been
791 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
792 The KPI for sx_init_flags() changes as accepting flags:
793 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
794 been introduced in order to handle the reversed logic.
795 Bump __FreeBSD_version to 800092.
798 Add support for hierarchical jails. Remove global securelevel.
799 Bump __FreeBSD_version to 800091.
802 The layout of struct vnet_net has changed, therefore modules
804 Bump __FreeBSD_version to 800090.
807 The newly imported zic(8) produces a new format in the output. Please
808 run tzsetup(8) to install the newly created data to /etc/localtime.
811 The sysctl tree for the usb stack has renamed from hw.usb2.* to
812 hw.usb.* and is now consistent again with previous releases.
815 802.11 monitor mode support was revised and driver api's were changed.
816 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
817 of DLT_IEEE802_11. No user-visible data structures were changed but
818 applications that use DLT_IEEE802_11 may require changes.
819 Bump __FreeBSD_version to 800088.
822 The layout of the following structs has changed: sysctl_oid,
823 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
824 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
825 panics may be experienced. World rebuild is required for
826 correctly checking networking state from userland.
827 Bump __FreeBSD_version to 800085.
830 MLDv2 and Source-Specific Multicast (SSM) have been merged
831 to the IPv6 stack. VIMAGE hooks are in but not yet used.
832 The implementation of SSM within FreeBSD's IPv6 stack closely
833 follows the IPv4 implementation.
835 For kernel developers:
837 * The most important changes are that the ip6_output() and
838 ip6_input() paths no longer take the IN6_MULTI_LOCK,
839 and this lock has been downgraded to a non-recursive mutex.
841 * As with the changes to the IPv4 stack to support SSM, filtering
842 of inbound multicast traffic must now be performed by transport
843 protocols within the IPv6 stack. This does not apply to TCP and
844 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
846 * The KPIs used by IPv6 multicast are similar to those used by
847 the IPv4 stack, with the following differences:
848 * im6o_mc_filter() is analogous to imo_multicast_filter().
849 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
850 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
851 * IN6_LOOKUP_MULTI() has been deprecated and removed.
852 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
853 for MLDv1 have an additional 'timer' argument which is used to
854 jitter the initial membership report for the solicited-node
855 multicast membership on-link.
856 * This is not strictly needed for MLDv2, which already jitters
857 its report transmissions. However, the 'timer' argument is
858 preserved in case MLDv1 is active on the interface.
860 * The KAME linked-list based IPv6 membership implementation has
861 been refactored to use a vector similar to that used by the IPv4
863 Code which maintains a list of its own multicast memberships
864 internally, e.g. carp, has been updated to reflect the new
867 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
868 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
869 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
870 implementation constraint which needs to be addressed in HEAD.
872 For application developers:
874 * The changes are broadly similar to those made for the IPv4
877 * The use of IPv4 and IPv6 multicast socket options on the same
878 socket, using mapped addresses, HAS NOT been tested or supported.
880 * There are a number of issues with the implementation of various
881 IPv6 multicast APIs which need to be resolved in the API surface
882 before the implementation is fully compatible with KAME userland
883 use, and these are mostly to do with interface index treatment.
885 * The literature available discusses the use of either the delta / ASM
886 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
887 using setsourcefilter(3)/getsourcefilter(3). For more information
888 please refer to RFC 3768, 'Socket Interface Extensions for
889 Multicast Source Filters'.
891 * Applications which use the published RFC 3678 APIs should be fine.
893 For systems administrators:
895 * The mtest(8) utility has been refactored to support IPv6, in
896 addition to IPv4. Interface addresses are no longer accepted
897 as arguments, their names must be used instead. The utility
898 will map the interface name to its first IPv4 address as
899 returned by getifaddrs(3).
901 * The ifmcstat(8) utility has also been updated to print the MLDv2
902 endpoint state and source filter lists via sysctl(3).
904 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
905 loopback of IPv6 multicast datagrams by default; it defaults to 1
906 to preserve the existing behaviour. Disabling multicast loopback is
907 recommended for optimal system performance.
909 * The IPv6 MROUTING code has been changed to examine this sysctl
910 instead of attempting to perform a group lookup before looping
911 back forwarded datagrams.
913 Bump __FreeBSD_version to 800084.
916 Implement low-level Bluetooth HCI API.
917 Bump __FreeBSD_version to 800083.
920 The layout of struct malloc_type, used by modules to register new
921 memory allocation types, has changed. Most modules will need to
922 be rebuilt or panics may be experienced.
923 Bump __FreeBSD_version to 800081.
926 Anticipate overflowing inp_flags - add inp_flags2.
927 This changes most offsets in inpcb, so checking v4 connection
928 state will require a world rebuild.
929 Bump __FreeBSD_version to 800080.
932 Add an llentry to struct route and struct route_in6. Modules
933 embedding a struct route will need to be recompiled.
934 Bump __FreeBSD_version to 800079.
937 The size of rt_metrics_lite and by extension rtentry has changed.
938 Networking administration apps will need to be recompiled.
939 The route command now supports show as an alias for get, weighting
940 of routes, sticky and nostick flags to alter the behavior of stateful
942 Bump __FreeBSD_version to 800078.
945 Do not use Giant for kbdmux(4) locking. This is wrong and
946 apparently causing more problems than it solves. This will
947 re-open the issue where interrupt handlers may race with
948 kbdmux(4) in polling mode. Typical symptoms include (but
949 not limited to) duplicated and/or missing characters when
950 low level console functions (such as gets) are used while
951 interrupts are enabled (for example geli password prompt,
952 mountroot prompt etc.). Disabling kbdmux(4) may help.
955 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
956 kernel modules referencing any of the above need to be recompiled.
957 Bump __FreeBSD_version to 800075.
960 GEOM_PART has become the default partition slicer for storage devices,
961 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
962 introduces some changes:
964 MSDOS/EBR: the devices created from MSDOS extended partition entries
965 (EBR) can be named differently than with GEOM_MBR and are now symlinks
966 to devices with offset-based names. fstabs may need to be modified.
968 BSD: the "geometry does not match label" warning is harmless in most
969 cases but it points to problems in file system misalignment with
970 disk geometry. The "c" partition is now implicit, covers the whole
971 top-level drive and cannot be (mis)used by users.
973 General: Kernel dumps are now not allowed to be written to devices
974 whose partition types indicate they are meant to be used for file
975 systems (or, in case of MSDOS partitions, as something else than
978 Most of these changes date approximately from 200812.
981 The uscanner(4) driver has been removed from the kernel. This follows
982 Linux removing theirs in 2.6 and making libusb the default interface
986 The multicast forwarding code has been cleaned up. netstat(1)
987 only relies on KVM now for printing bandwidth upcall meters.
988 The IPv4 and IPv6 modules are split into ip_mroute_mod and
989 ip6_mroute_mod respectively. The config(5) options for statically
990 compiling this code remain the same, i.e. 'options MROUTING'.
993 Support for the IFF_NEEDSGIANT network interface flag has been
994 removed, which means that non-MPSAFE network device drivers are no
995 longer supported. In particular, if_ar, if_sr, and network device
996 drivers from the old (legacy) USB stack can no longer be built or
1000 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1001 a bunch of new language catalog files have also been added.
1002 This means that some common libc messages are now localized and
1003 they depend on the LC_MESSAGES environmental variable.
1006 The k8temp(4) driver has been renamed to amdtemp(4) since
1007 support for Family 10 and Family 11 CPU families was added.
1010 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1011 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1013 For kernel developers, the most important changes are that the
1014 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1015 and this lock has been downgraded to a non-recursive mutex.
1017 Transport protocols (UDP, Raw IP) are now responsible for filtering
1018 inbound multicast traffic according to group membership and source
1019 filters. The imo_multicast_filter() KPI exists for this purpose.
1020 Transports which do not use multicast (SCTP, TCP) already reject
1021 multicast by default. Forwarding and receive performance may improve
1022 as a mutex acquisition is no longer needed in the ip_input()
1023 low-level input path. in_addmulti() and in_delmulti() are shimmed
1024 to new KPIs which exist to support SSM in-kernel.
1026 For application developers, it is recommended that loopback of
1027 multicast datagrams be disabled for best performance, as this
1028 will still cause the lock to be taken for each looped-back
1029 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1030 be tuned to 0 to disable loopback by default; it defaults to 1
1031 to preserve the existing behaviour.
1033 For systems administrators, to obtain best performance with
1034 multicast reception and multiple groups, it is always recommended
1035 that a card with a suitably precise hash filter is used. Hash
1036 collisions will still result in the lock being taken within the
1037 transport protocol input path to check group membership.
1039 If deploying FreeBSD in an environment with IGMP snooping switches,
1040 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1041 enabled; this forces 224.0.0.0/24 group membership to be announced
1044 The size of 'struct igmpstat' has changed; netstat needs to be
1045 recompiled to reflect this.
1046 Bump __FreeBSD_version to 800070.
1049 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1050 updated to use it. This requires a buildworld/installworld in order to
1051 update the library and dependencies (usbconfig, etc). Its advisable to
1052 rebuild all ports which uses libusb. More specific directions are given
1053 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1054 libusb are no longer required and can be removed.
1057 A workaround is committed to allow the creation of System V shared
1058 memory segment of size > 2 GB on the 64-bit architectures.
1059 Due to a limitation of the existing ABI, the shm_segsz member
1060 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1061 wrong for large segments. Note that limits must be explicitly
1062 raised to allow such segments to be created.
1065 The layout of struct ifnet has changed, requiring a rebuild of all
1066 network device driver modules.
1069 The /dev handling for the new USB stack has changed, a
1070 buildworld/installworld is required for libusb20.
1073 The new USB2 stack has now been permanently moved in and all kernel and
1074 module names reverted to their previous values (eg, usb, ehci, ohci,
1075 ums, ...). The old usb stack can be compiled in by prefixing the name
1076 with the letter 'o', the old usb modules have been removed.
1077 Updating entry 20090216 for xorg and 20090215 for libmap may still
1081 The rc.conf(5) option if_up_delay has been renamed to
1082 defaultroute_delay to better reflect its purpose. If you have
1083 customized this setting in /etc/rc.conf you need to update it to
1087 xorg 7.4 wants to configure its input devices via hald which does not
1088 yet work with USB2. If the keyboard/mouse does not work in xorg then
1090 Option "AllowEmptyInput" "off"
1091 to your ServerLayout section. This will cause X to use the configured
1092 kbd and mouse sections from your xorg.conf.
1095 The GENERIC kernels for all architectures now default to the new USB2
1096 stack. No kernel config options or code have been removed so if a
1097 problem arises please report it and optionally revert to the old USB
1098 stack. If you are loading USB kernel modules or have a custom kernel
1099 that includes GENERIC then ensure that usb names are also changed over,
1100 eg uftdi -> usb2_serial_ftdi.
1102 Older programs linked against the ports libusb 0.1 need to be
1103 redirected to the new stack's libusb20. /etc/libmap.conf can
1105 # Map old usb library to new one for usb2 stack
1106 libusb-0.1.so.8 libusb20.so.1
1109 All USB ethernet devices now attach as interfaces under the name ueN
1110 (eg. ue0). This is to provide a predictable name as vendors often
1111 change usb chipsets in a product without notice.
1114 The ichsmb(4) driver has been changed to require SMBus slave
1115 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1116 All of the other SMBus controller drivers require left-justified
1117 slave addresses, so this change makes all the drivers provide the
1121 INET6 statistics (struct ip6stat) was updated.
1122 netstat(1) needs to be recompiled.
1125 NTFS has been removed from GENERIC kernel on amd64 to match
1126 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1127 will load ntfs.ko module automatically when NTFS support is
1128 actually needed, unless ntfs.ko is not installed or security
1129 level prohibits loading kernel modules. If either is the case,
1130 "options NTFS" has to be added into kernel config.
1133 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1134 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1135 800061. User space tools that rely on the size of struct tcpcb in
1136 tcp_var.h (e.g. sockstat) need to be recompiled.
1139 ng_tty(4) module updated to match the new TTY subsystem.
1140 Due to API change, user-level applications must be updated.
1141 New API support added to mpd5 CVS and expected to be present
1142 in next mpd5.3 release.
1145 With __FreeBSD_version 800060 the makefs tool is part of
1146 the base system (it was a port).
1149 The afdata and ifnet locks have been changed from mutexes to
1150 rwlocks, network modules will need to be re-compiled.
1153 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1154 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1155 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1156 architecture and 40 bytes on 64-bit architecture. The userland
1157 applications "arp" and "ndp" have been updated accordingly.
1158 The output from "netstat -r" shows only routing entries and
1159 none of the L2 information.
1162 __FreeBSD_version 800057 marks the switchover from the
1163 binary ath hal to source code. Users must add the line:
1165 options AH_SUPPORT_AR5416
1167 to their kernel config files when specifying:
1171 The ath_hal module no longer exists; the code is now compiled
1172 together with the driver in the ath module. It is now
1173 possible to tailor chip support (i.e. reduce the set of chips
1174 and thereby the code size); consult ath_hal(4) for details.
1177 __FreeBSD_version 800054 adds memory barriers to
1178 <machine/atomic.h>, new interfaces to ifnet to facilitate
1179 multiple hardware transmit queues for cards that support
1180 them, and a lock-less ring-buffer implementation to
1181 enable drivers to more efficiently manage queueing of
1185 A new version of ZFS (version 13) has been merged to -HEAD.
1186 This version has zpool attribute "listsnapshots" off by
1187 default, which means "zfs list" does not show snapshots,
1188 and is the same as Solaris behavior.
1191 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1194 The uhci, ohci, ehci and slhci USB Host controller drivers have
1195 been put into separate modules. If you load the usb module
1196 separately through loader.conf you will need to load the
1197 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1198 controller add the following to loader.conf:
1204 The ABI used by the PMC toolset has changed. Please keep
1205 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1209 atapci kernel module now includes only generic PCI ATA
1210 driver. AHCI driver moved to ataahci kernel module.
1211 All vendor-specific code moved into separate kernel modules:
1212 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1213 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1214 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1215 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1218 The TTY subsystem of the kernel has been replaced by a new
1219 implementation, which provides better scalability and an
1220 improved driver model. Most common drivers have been migrated to
1221 the new TTY subsystem, while others have not. The following
1222 drivers have not yet been ported to the new TTY layer:
1225 cy, digi, rc, rp, sio
1231 ng_h4, ng_tty, ppp, sl, snp
1233 Adding these drivers to your kernel configuration file shall
1234 cause compilation to fail.
1237 ntpd has been upgraded to 4.2.4p5.
1240 OpenSSH has been upgraded to 5.1p1.
1242 For many years, FreeBSD's version of OpenSSH preferred DSA
1243 over RSA for host and user authentication keys. With this
1244 upgrade, we've switched to the vendor's default of RSA over
1245 DSA. This may cause upgraded clients to warn about unknown
1246 host keys even for previously known hosts. Users should
1247 follow the usual procedure for verifying host keys before
1248 accepting the RSA key.
1250 This can be circumvented by setting the "HostKeyAlgorithms"
1251 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1254 Please note that the sequence of keys offered for
1255 authentication has been changed as well. You may want to
1256 specify IdentityFile in a different order to revert this
1260 The sio(4) driver has been removed from the i386 and amd64
1261 kernel configuration files. This means uart(4) is now the
1262 default serial port driver on those platforms as well.
1264 To prevent collisions with the sio(4) driver, the uart(4) driver
1265 uses different names for its device nodes. This means the
1266 onboard serial port will now most likely be called "ttyu0"
1267 instead of "ttyd0". You may need to reconfigure applications to
1268 use the new device names.
1270 When using the serial port as a boot console, be sure to update
1271 /boot/device.hints and /etc/ttys before booting the new kernel.
1272 If you forget to do so, you can still manually specify the hints
1273 at the loader prompt:
1275 set hint.uart.0.at="isa"
1276 set hint.uart.0.port="0x3F8"
1277 set hint.uart.0.flags="0x10"
1278 set hint.uart.0.irq="4"
1282 The gpt(8) utility has been removed. Use gpart(8) to partition
1286 The version that Linuxulator emulates was changed from 2.4.2
1287 to 2.6.16. If you experience any problems with Linux binaries
1288 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1289 if it fixes the problem contact emulation mailing list.
1292 ISDN4BSD (I4B) was removed from the src tree. You may need to
1293 update a your kernel configuration and remove relevant entries.
1296 I have checked in code to support multiple routing tables.
1297 See the man pages setfib(1) and setfib(2).
1298 This is a hopefully backwards compatible version,
1299 but to make use of it you need to compile your kernel
1300 with options ROUTETABLES=2 (or more up to 16).
1303 The 802.11 wireless support was redone to enable multi-bss
1304 operation on devices that are capable. The underlying device
1305 is no longer used directly but instead wlanX devices are
1306 cloned with ifconfig. This requires changes to rc.conf files.
1307 For example, change:
1308 ifconfig_ath0="WPA DHCP"
1311 ifconfig_wlan0="WPA DHCP"
1312 see rc.conf(5) for more details. In addition, mergemaster of
1313 /etc/rc.d is highly recommended. Simultaneous update of userland
1314 and kernel wouldn't hurt either.
1316 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1317 modules were merged into the base wlan module. All references
1318 to these modules (e.g. in kernel config files) must be removed.
1321 psm(4) has gained write(2) support in native operation level.
1322 Arbitrary commands can be written to /dev/psm%d and status can
1323 be read back from it. Therefore, an application is responsible
1324 for status validation and error recovery. It is a no-op in
1325 other operation levels.
1328 Support for KSE threading has been removed from the kernel. To
1329 run legacy applications linked against KSE libmap.conf may
1330 be used. The following libmap.conf may be used to ensure
1331 compatibility with any prior release:
1333 libpthread.so.1 libthr.so.1
1334 libpthread.so.2 libthr.so.2
1335 libkse.so.3 libthr.so.3
1338 The layout of struct vmspace has changed. This affects libkvm
1339 and any executables that link against libkvm and use the
1340 kvm_getprocs() function. In particular, but not exclusively,
1341 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1342 The effects are minimal, but it's advisable to upgrade world
1346 The latest em driver no longer has support in it for the
1347 82575 adapter, this is now moved to the igb driver. The
1348 split was done to make new features that are incompatible
1349 with older hardware easier to do.
1352 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1353 likewise the kernel option is now GEOM_LINUX_LVM.
1356 The default NFS mount mode has changed from UDP to TCP for
1357 increased reliability. If you rely on (insecurely) NFS
1358 mounting across a firewall you may need to update your
1362 Belatedly note the addition of m_collapse for compacting
1366 The fts(3) structures have been changed to use adequate
1367 integer types for their members and so to be able to cope
1368 with huge file trees. The old fts(3) ABI is preserved
1369 through symbol versioning in libc, so third-party binaries
1370 using fts(3) should still work, although they will not take
1371 advantage of the extended types. At the same time, some
1372 third-party software might fail to build after this change
1373 due to unportable assumptions made in its source code about
1374 fts(3) structure members. Such software should be fixed
1375 by its vendor or, in the worst case, in the ports tree.
1376 FreeBSD_version 800015 marks this change for the unlikely
1377 case that a portable fix is impossible.
1380 To upgrade to -current after this date, you must be running
1381 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1382 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1385 The ADAPTIVE_GIANT kernel option has been retired because its
1386 functionality is the default now.
1389 The AT keyboard emulation of sunkbd(4) has been turned on
1390 by default. In order to make the special symbols of the Sun
1391 keyboards driven by sunkbd(4) work under X these now have
1392 to be configured the same way as Sun USB keyboards driven
1393 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1395 Option "XkbLayout" "us"
1396 Option "XkbRules" "xorg"
1397 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1400 It has been decided that it is desirable to provide ABI
1401 backwards compatibility to the FreeBSD 4/5/6 versions of the
1402 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1403 broken with the introduction of PCI domain support (see the
1404 20070930 entry). Unfortunately, this required the ABI of
1405 PCIOCGETCONF to be broken again in order to be able to
1406 provide backwards compatibility to the old version of that
1407 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1408 again. As for prominent ports this affects neither pciutils
1409 nor xorg-server this time, the hal port needs to be rebuilt
1413 The misnamed kthread_create() and friends have been renamed
1414 to kproc_create() etc. Many of the callers already
1415 used kproc_start()..
1416 I will return kthread_create() and friends in a while
1417 with implementations that actually create threads, not procs.
1418 Renaming corresponds with version 800002.
1427 Avoid using make -j when upgrading. While generally safe, there are
1428 sometimes problems using -j to upgrade. If your upgrade fails with
1429 -j, please try again without -j. From time to time in the past there
1430 have been problems using -j with buildworld and/or installworld. This
1431 is especially true when upgrading between "distant" versions (eg one
1432 that cross a major release boundary or several minor releases, or when
1433 several months have passed on the -current branch).
1435 Sometimes, obscure build problems are the result of environment
1436 poisoning. This can happen because the make utility reads its
1437 environment when searching for values for global variables. To run
1438 your build attempts in an "environmental clean room", prefix all make
1439 commands with 'env -i '. See the env(1) manual page for more details.
1441 When upgrading from one major version to another it is generally best
1442 to upgrade to the latest code in the currently installed branch first,
1443 then do an upgrade to the new branch. This is the best-tested upgrade
1444 path, and has the highest probability of being successful. Please try
1445 this approach before reporting problems with a major version upgrade.
1449 When upgrading the boot ZFS pool to a new version, always follow
1452 1.) recompile and reinstall the ZFS boot loader and boot block
1453 (this is part of "make buildworld" and "make installworld")
1455 2.) update the ZFS boot block on your boot drive
1457 The following example updates the ZFS boot block on the first
1458 partition (freebsd-boot) of a GPT partitioned drive ad0:
1459 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1461 Non-boot pools do not need these updates.
1465 If you are updating from a prior version of FreeBSD (even one just
1466 a few days old), you should follow this procedure. It is the most
1467 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1469 make kernel-toolchain
1470 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1471 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1473 To test a kernel once
1474 ---------------------
1475 If you just want to boot a kernel once (because you are not sure
1476 if it works, or if you want to boot a known bad kernel to provide
1477 debugging information) run
1478 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1479 nextboot -k testkernel
1481 To just build a kernel when you know that it won't mess you up
1482 --------------------------------------------------------------
1483 This assumes you are already running a CURRENT system. Replace
1484 ${arch} with the architecture of your machine (e.g. "i386",
1485 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1487 cd src/sys/${arch}/conf
1488 config KERNEL_NAME_HERE
1489 cd ../compile/KERNEL_NAME_HERE
1494 If this fails, go to the "To build a kernel" section.
1496 To rebuild everything and install it on the current system.
1497 -----------------------------------------------------------
1498 # Note: sometimes if you are running current you gotta do more than
1499 # is listed here if you are upgrading from a really old current.
1501 <make sure you have good level 0 dumps>
1503 make kernel KERNCONF=YOUR_KERNEL_HERE
1505 <reboot in single user> [3]
1513 To cross-install current onto a separate partition
1514 --------------------------------------------------
1515 # In this approach we use a separate partition to hold
1516 # current's root, 'usr', and 'var' directories. A partition
1517 # holding "/", "/usr" and "/var" should be about 2GB in
1520 <make sure you have good level 0 dumps>
1523 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1524 <maybe newfs current's root partition>
1525 <mount current's root partition on directory ${CURRENT_ROOT}>
1526 make installworld DESTDIR=${CURRENT_ROOT}
1527 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1528 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1529 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1530 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1531 <reboot into current>
1532 <do a "native" rebuild/install as described in the previous section>
1533 <maybe install compatibility libraries from ports/misc/compat*>
1537 To upgrade in-place from stable to current
1538 ----------------------------------------------
1539 <make sure you have good level 0 dumps>
1541 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1543 <reboot in single user> [3]
1550 Make sure that you've read the UPDATING file to understand the
1551 tweaks to various things you need. At this point in the life
1552 cycle of current, things change often and you are on your own
1553 to cope. The defaults can also change, so please read ALL of
1554 the UPDATING entries.
1556 Also, if you are tracking -current, you must be subscribed to
1557 freebsd-current@freebsd.org. Make sure that before you update
1558 your sources that you have read and understood all the recent
1559 messages there. If in doubt, please track -stable which has
1560 much fewer pitfalls.
1562 [1] If you have third party modules, such as vmware, you
1563 should disable them at this point so they don't crash your
1566 [3] From the bootblocks, boot -s, and then do
1571 adjkerntz -i # if CMOS is wall time
1572 Also, when doing a major release upgrade, it is required that
1573 you boot into single user mode to do the installworld.
1575 [4] Note: This step is non-optional. Failure to do this step
1576 can result in a significant reduction in the functionality of the
1577 system. Attempting to do it by hand is not recommended and those
1578 that pursue this avenue should read this file carefully, as well
1579 as the archives of freebsd-current and freebsd-hackers mailing lists
1580 for potential gotchas. The -U option is also useful to consider.
1581 See mergemaster(8) for more information.
1583 [5] Usually this step is a noop. However, from time to time
1584 you may need to do this if you get unknown user in the following
1585 step. It never hurts to do it all the time. You may need to
1586 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1587 install) after the buildworld before this step if you last updated
1588 from [78]-stable or 9-stable before 20130430.
1590 [6] This only deletes old files and directories. Old libraries
1591 can be deleted by "make delete-old-libs", but you have to make
1592 sure that no program is using those libraries anymore.
1594 [8] In order to have a kernel that can run the 4.x binaries needed to
1595 do an installworld, you must include the COMPAT_FREEBSD4 option in
1596 your kernel. Failure to do so may leave you with a system that is
1597 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1598 required to run the 5.x binaries on more recent kernels. And so on
1599 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1601 Make sure that you merge any new devices from GENERIC since the
1602 last time you updated your kernel config file.
1604 [9] When checking out sources, you must include the -P flag to have
1605 cvs prune empty directories.
1607 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1608 "?=" instead of the "=" assignment operator, so that buildworld can
1609 override the CPUTYPE if it needs to.
1611 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1612 not on the command line, or in /etc/make.conf. buildworld will
1613 warn if it is improperly defined.
1616 This file contains a list, in reverse chronological order, of major
1617 breakages in tracking -current. It is not guaranteed to be a complete
1618 list of such breakages, and only contains entries since October 10, 2007.
1619 If you need to see UPDATING entries from before that date, you will need
1620 to fetch an UPDATING file from an older FreeBSD release.
1622 Copyright information:
1624 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1626 Redistribution, publication, translation and use, with or without
1627 modification, in full or in part, in any form or format of this
1628 document are permitted without further permission from the author.
1630 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1631 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1632 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1633 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1634 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1635 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1636 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1637 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1638 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1639 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1640 POSSIBILITY OF SUCH DAMAGE.
1642 Contact Warner Losh if you have any questions about your use of