1 Updating Information for FreeBSD current users
3 This file is maintained and copyrighted by M. Warner Losh <imp@freebsd.org>.
4 See end of file for further details. For commonly done items, please see the
5 COMMON ITEMS: section later in the file. These instructions assume that you
6 basically know what you are doing. If not, then please consult the FreeBSD
9 http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/makeworld.html
11 Items affecting the ports and packages system can be found in
12 /usr/ports/UPDATING. Please read that file before running portupgrade.
14 20140513: p6 FreeBSD-EN-14:03.pkg
15 FreeBSD-EN-14:04.kldxref
18 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
20 Improve build repeatability for kldxref(8). [EN-14:04]
22 Fix data corruption with ciss(4). [EN-14:05]
24 20140430: p5 FreeBSD-SA-14:08.tcp
26 Fix TCP reassembly vulnerability. [SA-14:08]
28 20140408: p4 FreeBSD-SA-14:05.nfsserver
29 FreeBSD-SA-14:06.openssl
30 Fix deadlock in the NFS server. [SA-14:05]
32 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
34 20140114: p3 FreeBSD-SA-14:01.bsnmpd
37 FreeBSD-EN-14:01.random
39 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
41 Fix ntpd distributed reflection Denial of Service
42 vulnerability. [SA-14:02]
44 Fix BIND remote denial of service vulnerability. [SA-14:04]
46 Disable hardware RNGs by default. [EN-14:01]
48 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
50 20131128: p2 FreeBSD-EN-13:05.freebsd-update
51 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
53 20131026: p1 FreeBSD-EN-13:04.freebsd-update
54 Fix multiple freebsd-update bugs that break upgrading to
58 hastctl(8)'s `status' command output changed to terse one-liner format.
59 Scripts using this should switch to `list' command or be rewritten.
62 Fix a bug that allowed a tracing process (e.g. gdb) to write
63 to a memory-mapped file in the traced process's address space
64 even if neither the traced process nor the tracing process had
65 write access to that file.
68 Added ZFS TRIM support which is enabled by default. To disable
69 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
71 Creating new ZFS pools and adding new devices to existing pools
72 first performs a full device level TRIM which can take a significant
73 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
74 to disable this behaviour.
76 ZFS TRIM requires the underlying device support BIO_DELETE which
77 is currently provided by methods such as ATA TRIM and SCSI UNMAP
78 via CAM, which are typically supported by SSD's.
80 Stats for ZFS TRIM can be monitored by looking at the sysctl's
81 under kstat.zfs.misc.zio_trim.
84 `list' command has been added to hastctl(8). For now, it is full
85 equivalent of `status' command.
86 WARNING: in the near future the output of hastctl's status command
87 will change to more terse format. If you use `hastctl status'
88 for parsing in your scripts, switch to `hastctl list'.
91 The mergemaster command now uses the default MAKEOBJDIRPREFIX
92 rather than creating it's own in the temporary directory in
93 order allow access to bootstrapped versions of tools such as
94 install and mtree. When upgrading from version of FreeBSD where
95 the install command does not support -l, you will need to
96 install a new mergemaster command if mergemaster -p is required.
97 This can be accomplished with the command (cd src/usr.sbin/mergemaster
100 Due to the use of the new -l option to install(1) during build
101 and install, you must take care not to directly set the INSTALL
102 make variable in your /etc/make.conf, /etc/src.conf, or on the
103 command line. If you wish to use the -C flag for all installs
104 you may be able to add INSTALL+=-C to /etc/make.conf or
108 Fix a bug that allows NFS clients to issue READDIR on files.
111 The install(1) option -M has changed meaning and now takes an
112 argument that is a file or path to append logs to. In the
113 unlikely event that -M was the last option on the command line
114 and the command line contained at least two files and a target
115 directory the first file will have logs appended to it. The -M
116 option served little practical purpose in the last decade so it's
117 used expected to be extremely rare.
120 A new compression method (lz4) has been merged to. Please refer to
121 zpool-features(7) for more information.
123 Please refer to the "ZFS notes" section of this file for information
124 on upgrading boot ZFS pools.
127 The VFS KBI was changed with the merge of several nullfs
128 optimizations and fixes. All filesystem modules must be
132 With the addition of auditdistd(8), a new auditdistd user is now
133 depended on during installworld. "mergemaster -p" can be used to add
134 the user prior to installworld, as documented in the handbook.
140 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
141 Starting with this version the old system of ZFS pool versioning
142 is superseded by "feature flags". This concept enables forward
143 compatibility against certain future changes in functionality of ZFS
144 pools. The first two read-only compatible "feature flags" for ZFS
145 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
146 For more information read the new zpool-features(7) manual page.
147 Please refer to the "ZFS notes" section of this file for information
148 on upgrading boot ZFS pools.
151 The commit introducing bsd.compiler.mk breaks the traditional
152 building of kernels before this point. Add -m ${SRC}/share/mk
153 (for the right value of SRC) to your command lines to work
154 around; update your useland to a point after this; or use the
155 buildkernel/installkernel top-level targets. See also 20120829.
158 The IPFIREWALL_FORWARD kernel option has been removed. Its
159 functionality now turned on by default.
162 The random(4) support for the VIA hardware random number
163 generator (`PADLOCK') is no longer enabled unconditionally.
164 Add the PADLOCK_RNG option in the custom kernel config if
165 needed. The GENERIC kernels on i386 and amd64 do include the
166 option, so the change only affects the custom kernel
170 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
171 the traditional method, you must update your system with an installworld
172 before the kernel will build. The documented make buildkernel/installkernel
173 interfaces (coupled with fresh make kernel-toolchain) continue to work.
176 The sparc64 ZFS loader has been changed to no longer try to auto-
177 detect ZFS providers based on diskN aliases but now requires these
178 to be explicitly listed in the OFW boot-device environment variable.
181 Now unix domain sockets behave "as expected" on nullfs(5). Previously
182 nullfs(5) did not pass through all behaviours to the underlying layer,
183 as a result if we bound to a socket on the lower layer we could connect
184 only to the lower path; if we bound to the upper layer we could connect
185 only to the upper path. The new behavior is one can connect to both the
186 lower and the upper paths regardless what layer path one binds to.
189 The acpi_wmi(4) status device /dev/wmistat has been renamed to
193 A new VOP_ADVISE() was added to support posix_fadvise(2). All
194 filesystem modules must be recompiled.
197 The interface of the VOP_VPTOCNP(9) changed, now the returned
198 vnode shall be referenced, previously it was required to be
199 only held. All in-tree filesystems are converted.
205 The broken amd(4) driver has been replaced with esp(4) in the amd64,
206 i386 and pc98 GENERIC kernel configuration files.
209 This commit modifies vfs_register() so that it uses a hash
210 calculation to set vfc_typenum, which is enabled by default.
211 The first time a system is booted after this change, the
212 vfc_typenum values will change for all file systems. The
213 main effect of this is a change to the NFS server file handles
214 for file systems that use vfc_typenum in their fsid, such as ZFS.
215 It will, however, prevent vfc_typenum from changing when file
216 systems are loaded in a different order for subsequent reboots.
217 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
218 until you are ready to remount all NFS clients after a reboot.
221 Bump the shared library version numbers for libraries that
222 do not use symbol versioning, have changed the ABI compared
223 to stable/8 and which shared library version was not bumped.
224 Done as part of 9.0-RELEASE cycle.
227 During the merge of Capsicum features, the fget(9) KPI was modified.
228 This may require the rebuilding of out-of-tree device drivers --
229 issues have been reported specifically with the nVidia device driver.
230 __FreeBSD_version is bumped to 900041.
232 Also, there is a period between 20110811 and 20110814 where the
233 special devices /dev/{stdin,stdout,stderr} did not work correctly.
234 Building world from a kernel during that window may not work.
237 The packet filter (pf) code has been updated to OpenBSD 4.5.
238 You need to update userland tools to be in sync with kernel.
239 This update breaks backward compatibility with earlier pfsync(4)
240 versions. Care must be taken when updating redundant firewall setups.
243 The following sysctls and tunables are retired on x86 platforms:
245 machdep.hlt_logical_cpus
246 The following sysctl is retired:
247 machdep.hyperthreading_allowed
248 The sysctls were supposed to provide a way to dynamically offline and
249 online selected CPUs on x86 platforms, but the implementation has not
250 been reliable especially with SCHED_ULE scheduler.
251 machdep.hyperthreading_allowed tunable is still available to ignore
252 hyperthreading CPUs at OS level.
253 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
254 where X is an APIC ID of a CPU. Be advised, though, that disabling
255 CPUs in non-uniform fashion will result in non-uniform topology and
256 may lead to sub-optimal system performance with SCHED_ULE, which is
260 cpumask_t type is retired and cpuset_t is used in order to describe
264 Changes to ifconfig(8) for dynamic address family detection mandate
265 that you are running a kernel of 20110525 or later. Make sure to
266 follow the update procedure to boot a new kernel before installing
270 Support for sun4v architecture is officially dropped
273 Several KPI breaking changes have been committed to the mii(4) layer,
274 the PHY drivers and consequently some Ethernet drivers using mii(4).
275 This means that miibus.ko and the modules of the affected Ethernet
276 drivers need to be recompiled.
278 Note to kernel developers: Given that the OUI bit reversion problem
279 was fixed as part of these changes all mii(4) commits related to OUIs,
280 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
281 handling, no longer can be merged verbatim to stable/8 and previous
285 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
286 into their kernel configurations along with 'device pci'.
289 The default NFS client is now the new NFS client, so fstype "newnfs"
290 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
291 Although mounts via fstype "nfs" will usually work without userland
292 changes, it is recommended that the mount(8) and mount_nfs(8)
293 commands be rebuilt from sources and that a link to mount_nfs called
294 mount_oldnfs be created. The new client is compiled into the
295 kernel with "options NFSCL" and this is needed for diskless root
296 file systems. The GENERIC kernel configs have been changed to use
297 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
298 To use the regular/old client, you can "mount -t oldnfs ...". For
299 a diskless root file system, you must also include a line like:
301 vfs.root.mountfrom="oldnfs:"
303 in the boot/loader.conf on the root fs on the NFS server to make
304 a diskless root fs use the old client.
307 The GENERIC kernels for all architectures now default to the new
308 CAM-based ATA stack. It means that all legacy ATA drivers were
309 removed and replaced by respective CAM drivers. If you are using
310 ATA device names in /etc/fstab or other places, make sure to update
311 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
312 where 'Y's are the sequential numbers starting from zero for each type
313 in order of detection, unless configured otherwise with tunables,
314 see cam(4)). There will be symbolic links created in /dev/ to map
315 old adX devices to the respective adaY. They should provide basic
316 compatibility for file systems mounting in most cases, but they do
317 not support old user-level APIs and do not have respective providers
318 in GEOM. Consider using updated management tools with new device names.
320 It is possible to load devices ahci, ata, siis and mvs as modules,
321 but option ATA_CAM should remain in kernel configuration to make ata
322 module work as CAM driver supporting legacy ATA controllers. Device ata
323 still can be used in modular fashion (atacore + ...). Modules atadisk
324 and atapi* are not used and won't affect operation in ATA_CAM mode.
325 Note that to use CAM-based ATA kernel should include CAM devices
326 scbus, pass, da (or explicitly ada), cd and optionally others. All of
327 them are parts of the cam module.
329 ataraid(4) functionality is now supported by the RAID GEOM class.
330 To use it you can load geom_raid kernel module and use graid(8) tool
331 for management. Instead of /dev/arX device names, use /dev/raid/rX.
333 No kernel config options or code have been removed, so if a problem
334 arises, please report it and optionally revert to the old ATA stack.
335 In order to do it you can remove from the kernel config:
340 , and instead add back:
341 device atadisk # ATA disk drives
342 device ataraid # ATA RAID drives
343 device atapicd # ATAPI CDROM drives
344 device atapifd # ATAPI floppy drives
345 device atapist # ATAPI tape drives
348 The default NFS server has been changed to the new server, which
349 was referred to as the experimental server. If you need to switch
350 back to the old NFS server, you must now put the "-o" option on
351 both the mountd and nfsd commands. This can be done using the
352 mountd_flags and nfs_server_flags rc.conf variables until an
353 update to the rc scripts is committed, which is coming soon.
356 The GNU Objective-C runtime library (libobjc), and other Objective-C
357 related components have been removed from the base system. If you
358 require an Objective-C library, please use one of the available ports.
361 ath(4) has been split into bus- and device- modules. if_ath contains
362 the HAL, the TX rate control and the network device code. if_ath_pci
363 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
364 contains the AHB glue. Users need to load both if_ath_pci and if_ath
365 in order to use ath on everything else.
367 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
368 need to load if_ath and if_ath_pci for ath(4) operation.
371 As part of the replacement of sysinstall, the process of building
372 release media has changed significantly. For details, please re-read
373 release(7), which has been updated to reflect the new build process.
376 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
377 is the last available version under GPLv2. It brings a number of new
378 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
379 4.1 and SSE 4.2), better support for powerpc64, a number of new
380 directives, and lots of other small improvements. See the ChangeLog
381 file in contrib/binutils for the full details.
384 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
385 compliant, and will now use half of hash for authentication.
386 This will break interoperability with all stacks (including all
387 actual FreeBSD versions) who implement
388 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
390 The only workaround with such peers is to use another HMAC
391 algorithm for IPsec ("phase 2") authentication.
394 Remove the uio_yield prototype and symbol. This function has
395 been misnamed since it was introduced and should not be
396 globally exposed with this name. The equivalent functionality
397 is now available using kern_yield(curthread->td_user_pri).
398 The function remains undocumented.
401 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
402 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
403 for scalar sysctls is defined but disabled. Code that needs
404 UQUAD to pass the type checking that must compile on older
405 systems where the define is not present can check against
406 __FreeBSD_version >= 900030.
408 The system dialog(1) has been replaced with a new version previously
409 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
410 with the previous version, but the libdialog associated with it has
411 a largely incompatible API. As such, the original version of libdialog
412 will be kept temporarily as libodialog, until its base system consumers
413 are replaced or updated. Bump __FreeBSD_version to 900030.
416 If you are trying to run make universe on a -stable system, and you get
417 the following warning:
418 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
419 unknown. config(8) likely too old."
420 or something similar to it, then you must upgrade your -stable system
421 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
422 stable/8) or build the config from the latest stable/8 branch and
423 install it on your system.
425 Prior to this date, building a current universe on 8-stable system from
426 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
427 error in the first kernel build phase. A new config on those old
428 systems will fix that problem for older versions of -current.
431 The TCP stack has been modified to allow Khelp modules to interact with
432 it via helper hook points and store per-connection data in the TCP
433 control block. Bump __FreeBSD_version to 900029. User space tools that
434 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
438 Generic IEEE 802.3 annex 31B full duplex flow control support has been
439 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
440 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
441 to take advantage of it instead of using custom implementations. This
442 means that these drivers now no longer unconditionally advertise
443 support for flow control but only do so if flow control is a selected
444 media option. This was implemented in the generic support that way in
445 order to allow flow control to be switched on and off via ifconfig(8)
446 with the PHY specific default to typically off in order to protect
447 from unwanted effects. Consequently, if you used flow control with
448 one of the above mentioned drivers you now need to explicitly enable
450 ifconfig bge0 media auto mediaopt flowcontrol
452 Along with the above mentioned changes generic support for setting
453 1000baseT master mode also has been added and brgphy(4), ciphy(4),
454 e1000phy(4) as well as ip1000phy(4) have been converted to take
455 advantage of it. This means that these drivers now no longer take the
456 link0 parameter for selecting master mode but the master media option
457 has to be used instead, for example like in the following:
458 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
460 Selection of master mode now is also available with all other PHY
461 drivers supporting 1000baseT.
464 The TCP stack has received a significant update to add support for
465 modularised congestion control and generally improve the clarity of
466 congestion control decisions. Bump __FreeBSD_version to 900025. User
467 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
468 sockstat) need to be recompiled.
471 The man(1) utility has been replaced by a new version that no longer
472 uses /etc/manpath.config. Please consult man.conf(5) for how to
473 migrate local entries to the new format.
476 The copyright strings printed by login(1) and sshd(8) at the time of a
477 new connection have been removed to follow other operating systems and
481 A workaround for a fixed ld bug has been removed in kernel code,
482 so make sure that your system ld is built from sources after
483 revision 210245 from 2010-07-19 (r211583 if building head kernel
484 on stable/8, r211584 for stable/7; both from 2010-08-21).
485 A symptom of incorrect ld version is different addresses for
486 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
489 The $ipv6_prefer variable in rc.conf(5) has been split into
490 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
492 The $ip6addrctl_policy is a variable to choose a pre-defined
493 address selection policy set by ip6addrctl(8). A value
494 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
497 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
498 flag (see an entry of 20090926) is set on an interface with no
499 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
500 security reason. If you want IPv6 link-local address on all
501 interfaces by default, set this to "YES".
503 The old ipv6_prefer="YES" is equivalent to
504 ipv6_activate_all_interfaces="YES" and
505 ip6addrctl_policy="ipv6_prefer".
508 DTrace has grown support for userland tracing. Due to this, DTrace is
509 now i386 and amd64 only.
510 dtruss(1) is now installed by default on those systems and a new
511 kernel module is needed for userland tracing: fasttrap.
512 No changes to your kernel config file are necessary to enable
513 userland tracing, but you might consider adding 'STRIP=' and
514 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
515 to have informative userland stack traces in DTrace (ustack).
518 The acpi_aiboost(4) driver has been removed in favor of the new
519 aibs(4) driver. You should update your kernel configuration file.
522 BSD grep has been imported to the base system and it is built by
523 default. It is completely BSD licensed, highly GNU-compatible, uses
524 less memory than its GNU counterpart and has a small codebase.
525 However, it is slower than its GNU counterpart, which is mostly
526 noticeable for larger searches, for smaller ones it is measurable
527 but not significant. The reason is complex, the most important factor
528 is that we lack a modern and efficient regex library and GNU
529 overcomes this by optimizing the searches internally. Future work
530 on improving the regex performance is planned, for the meantime,
531 users that need better performance, can build GNU grep instead by
532 setting the WITH_GNU_GREP knob.
535 Due to the import of powerpc64 support, all existing powerpc kernel
536 configuration files must be updated with a machine directive like this:
537 machine powerpc powerpc
539 In addition, an updated config(8) is required to build powerpc kernels
543 A new version of ZFS (version 15) has been merged to -HEAD.
544 This version uses a python library for the following subcommands:
545 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
546 For full functionality of these commands the following port must
547 be installed: sysutils/py-zfs
550 'vm_page's are now hashed by physical address to an array of mutexes.
551 Currently this is only used to serialize access to hold_count. Over
552 time the page queue mutex will be peeled away. This changes the size
553 of pmap on every architecture. And requires all callers of vm_page_hold
554 and vm_page_unhold to be updated.
557 WITH_CTF can now be specified in src.conf (not recommended, there
558 are some problems with static executables), make.conf (would also
559 affect ports which do not use GNU make and do not override the
560 compile targets) or in the kernel config (via "makeoptions
562 When WITH_CTF was specified there before this was silently ignored,
563 so make sure that WITH_CTF is not used in places which could lead
564 to unwanted behavior.
567 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
568 to allow 32-bit compatibility on non-x86 platforms. All kernel
569 configurations on amd64 and ia64 platforms using these options must
570 be modified accordingly.
573 The utmp user accounting database has been replaced with utmpx,
574 the user accounting interface standardized by POSIX.
575 Unfortunately the semantics of utmp and utmpx don't match,
576 making it practically impossible to support both interfaces.
577 The user accounting database is used by tools like finger(1),
578 last(1), talk(1), w(1) and ac(8).
580 All applications in the base system use utmpx. This means only
581 local binaries (e.g. from the ports tree) may still use these
582 utmp database files. These applications must be rebuilt to make
585 After the system has been upgraded, it is safe to remove the old
586 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
587 assuming their contents is of no importance anymore. Old wtmp
588 databases can only be used by last(1) and ac(8) after they have
589 been converted to the new format using wtmpcvt(1).
592 Introduce the kernel thread "deadlock resolver" (which can be enabled
593 via the DEADLKRES option, see NOTES for more details) and the
594 sleepq_type() function for sleepqueues.
597 The rc.firewall and rc.firewall6 were unified, and
598 rc.firewall6 and rc.d/ip6fw were removed.
599 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
600 variables are obsoleted. Instead, the following new rc
601 variables are added to rc.d/ipfw:
603 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
604 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
605 firewall_simple_onet_ipv6, firewall_trusted_ipv6
607 The meanings correspond to the relevant IPv4 variables.
613 The default terminal emulation for syscons(4) has been changed
614 from cons25 to xterm on all platforms except pc98. This means
615 that the /etc/ttys file needs to be updated to ensure correct
616 operation of applications on the console.
618 The terminal emulation style can be toggled per window by using
619 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
620 options can be used to change the compile-time default back to
623 To prevent graphical artifacts, make sure the TERM environment
624 variable is set to match the terminal emulation that is being
625 performed by syscons(4).
628 The layout of the structure ieee80211req_scan_result has changed.
629 Applications that require wireless scan results (e.g. ifconfig(8))
630 from net80211 need to be recompiled.
632 Applications such as wpa_supplicant(8) may require a full world
633 build without using NO_CLEAN in order to get synchronized with the
637 The iwn(4) driver has been updated to support the 5000 and 5150 series.
638 There's one kernel module for each firmware. Adding "device iwnfw"
639 to the kernel configuration file means including all three firmware
640 images inside the kernel. If you want to include just the one for
641 your wireless card, use the devices iwn4965fw, iwn5000fw or
645 The rc.d/network_ipv6, IPv6 configuration script has been integrated
646 into rc.d/netif. The changes are the following:
648 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
649 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
650 Note that both variables need the "inet6" keyword at the head.
652 Do not set $ipv6_network_interfaces manually if you do not
653 understand what you are doing. It is not needed in most cases.
655 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
658 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
659 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
661 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
662 all configured interfaces have "inet6 accept_rtadv" in the
663 $ifconfig_IF_ipv6. These are for backward compatibility.
665 3. A new variable $ipv6_prefer has been added. If NO, IPv6
666 functionality of interfaces with no corresponding
667 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
668 and the default address selection policy of ip6addrctl(8)
669 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
670 Note that if you want to configure IPv6 functionality on the
671 disabled interfaces after boot, first you need to clear the flag by
672 using ifconfig(8) like:
674 ifconfig em0 inet6 -ifdisabled
676 If YES, the default address selection policy is set as
679 The default value of $ipv6_prefer is NO.
681 4. If your system need to receive Router Advertisement messages,
682 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
683 scripts automatically invoke rtsol(8) when the interface becomes
684 UP. The Router Advertisement messages are used for SLAAC
685 (State-Less Address AutoConfiguration).
688 802.11s D3.03 support was committed. This is incompatible with the
689 previous code, which was based on D3.0.
692 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
693 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
694 control whether accepting Router Advertisement messages or not.
695 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
696 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
697 The ifconfig(8) utility now supports these flags.
700 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
701 mount(8) and -a option for df(1) to see them.
704 The old tunable hw.bus.devctl_disable has been superseded by
705 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
706 replaced by hw.bus.devctl_queue=0. The default for this new tunable
710 Remove the option STOP_NMI. The default action is now to use NMI only
711 for KDB via the newly introduced function stop_cpus_hard() and
712 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
715 The stable/8 branch created in subversion. This corresponds to the
716 RELENG_8 branch in CVS.
719 Bump the shared library version numbers for all libraries that do not
720 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
721 __FreeBSD_version to 800105.
724 Due to changes in the implementation of virtual network stack support,
725 all network-related kernel modules must be recompiled. As this change
726 breaks the ABI, bump __FreeBSD_version to 800104.
729 The TOE interface to the TCP syncache has been modified to remove
730 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
731 The cxgb driver is the only TOE consumer affected by this change, and
732 needs to be recompiled along with the kernel. As this change breaks
733 the ABI, bump __FreeBSD_version to 800103.
736 Padding has been added to struct tcpcb, sackhint and tcpstat in
737 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
738 maintaining the ABI. However, this change breaks the ABI, so bump
739 __FreeBSD_version to 800102. User space tools that rely on the size of
740 any of these structs (e.g. sockstat) need to be recompiled.
743 The NFS_LEGACYRPC option has been removed along with the old kernel
744 RPC implementation that this option selected. Kernel configurations
745 may need to be adjusted.
748 The network interface device nodes at /dev/net/<interface> have been
749 removed. All ioctl operations can be performed the normal way using
750 routing sockets. The kqueue functionality can generally be replaced
751 with routing sockets.
754 The documentation from the FreeBSD Documentation Project (Handbook,
755 FAQ, etc.) is now installed via packages by sysinstall(8) and under
756 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
759 The ABI of various structures related to the SYSV IPC API have been
760 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
761 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
765 Layout of struct vnet has changed as routing related variables were
766 moved to their own Vimage module. Modules need to be recompiled. Bump
767 __FreeBSD_version to 800099.
770 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
771 respectively. As long as no more than 16 groups per process are used,
772 no changes should be visible. When more than 16 groups are used, old
773 binaries may fail if they call getgroups() or getgrouplist() with
774 statically sized storage. Recompiling will work around this, but
775 applications should be modified to use dynamically allocated storage
776 for group arrays as POSIX.1-2008 does not cap an implementation's
777 number of supported groups at NGROUPS_MAX+1 as previous versions did.
779 NFS and portalfs mounts may also be affected as the list of groups is
780 truncated to 16. Users of NFS who use more than 16 groups, should
781 take care that negative group permissions are not used on the exported
782 file systems as they will not be reliable unless a GSSAPI based
783 authentication method is used.
786 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
787 option compiles in the support for adaptive spinning for lockmgrs
788 which want to enable it. The lockinit() function now accepts the flag
789 LK_ADAPTIVE in order to make the lock object subject to adaptive
790 spinning when both held in write and read mode.
793 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
794 changed. User applications that use this ioctl need to be rebuilt.
797 The layout of struct thread has changed. Kernel and modules need to
801 The layout of structs ifnet, domain, protosw and vnet_net has changed.
802 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
805 window(1) has been removed from the base system. It can now be
806 installed from ports. The port is called misc/window.
809 The way we are storing and accessing `routing table' entries has
810 changed. Programs reading the FIB, like netstat, need to be
814 A new netisr implementation has been added for FreeBSD 8. Network
815 file system modules, such as igmp, ipdivert, and others, should be
817 Bump __FreeBSD_version to 800096.
820 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
824 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
825 Bump __FreeBSD_version to 800094.
828 Add mnt_xflag field to 'struct mount'. File system modules need to be
830 Bump __FreeBSD_version to 800093.
833 The compiling option ADAPTIVE_SX has been retired while it has been
834 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
835 The KPI for sx_init_flags() changes as accepting flags:
836 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
837 been introduced in order to handle the reversed logic.
838 Bump __FreeBSD_version to 800092.
841 Add support for hierarchical jails. Remove global securelevel.
842 Bump __FreeBSD_version to 800091.
845 The layout of struct vnet_net has changed, therefore modules
847 Bump __FreeBSD_version to 800090.
850 The newly imported zic(8) produces a new format in the output. Please
851 run tzsetup(8) to install the newly created data to /etc/localtime.
854 The sysctl tree for the usb stack has renamed from hw.usb2.* to
855 hw.usb.* and is now consistent again with previous releases.
858 802.11 monitor mode support was revised and driver api's were changed.
859 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
860 of DLT_IEEE802_11. No user-visible data structures were changed but
861 applications that use DLT_IEEE802_11 may require changes.
862 Bump __FreeBSD_version to 800088.
865 The layout of the following structs has changed: sysctl_oid,
866 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
867 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
868 panics may be experienced. World rebuild is required for
869 correctly checking networking state from userland.
870 Bump __FreeBSD_version to 800085.
873 MLDv2 and Source-Specific Multicast (SSM) have been merged
874 to the IPv6 stack. VIMAGE hooks are in but not yet used.
875 The implementation of SSM within FreeBSD's IPv6 stack closely
876 follows the IPv4 implementation.
878 For kernel developers:
880 * The most important changes are that the ip6_output() and
881 ip6_input() paths no longer take the IN6_MULTI_LOCK,
882 and this lock has been downgraded to a non-recursive mutex.
884 * As with the changes to the IPv4 stack to support SSM, filtering
885 of inbound multicast traffic must now be performed by transport
886 protocols within the IPv6 stack. This does not apply to TCP and
887 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
889 * The KPIs used by IPv6 multicast are similar to those used by
890 the IPv4 stack, with the following differences:
891 * im6o_mc_filter() is analogous to imo_multicast_filter().
892 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
893 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
894 * IN6_LOOKUP_MULTI() has been deprecated and removed.
895 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
896 for MLDv1 have an additional 'timer' argument which is used to
897 jitter the initial membership report for the solicited-node
898 multicast membership on-link.
899 * This is not strictly needed for MLDv2, which already jitters
900 its report transmissions. However, the 'timer' argument is
901 preserved in case MLDv1 is active on the interface.
903 * The KAME linked-list based IPv6 membership implementation has
904 been refactored to use a vector similar to that used by the IPv4
906 Code which maintains a list of its own multicast memberships
907 internally, e.g. carp, has been updated to reflect the new
910 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
911 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
912 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
913 implementation constraint which needs to be addressed in HEAD.
915 For application developers:
917 * The changes are broadly similar to those made for the IPv4
920 * The use of IPv4 and IPv6 multicast socket options on the same
921 socket, using mapped addresses, HAS NOT been tested or supported.
923 * There are a number of issues with the implementation of various
924 IPv6 multicast APIs which need to be resolved in the API surface
925 before the implementation is fully compatible with KAME userland
926 use, and these are mostly to do with interface index treatment.
928 * The literature available discusses the use of either the delta / ASM
929 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
930 using setsourcefilter(3)/getsourcefilter(3). For more information
931 please refer to RFC 3768, 'Socket Interface Extensions for
932 Multicast Source Filters'.
934 * Applications which use the published RFC 3678 APIs should be fine.
936 For systems administrators:
938 * The mtest(8) utility has been refactored to support IPv6, in
939 addition to IPv4. Interface addresses are no longer accepted
940 as arguments, their names must be used instead. The utility
941 will map the interface name to its first IPv4 address as
942 returned by getifaddrs(3).
944 * The ifmcstat(8) utility has also been updated to print the MLDv2
945 endpoint state and source filter lists via sysctl(3).
947 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
948 loopback of IPv6 multicast datagrams by default; it defaults to 1
949 to preserve the existing behaviour. Disabling multicast loopback is
950 recommended for optimal system performance.
952 * The IPv6 MROUTING code has been changed to examine this sysctl
953 instead of attempting to perform a group lookup before looping
954 back forwarded datagrams.
956 Bump __FreeBSD_version to 800084.
959 Implement low-level Bluetooth HCI API.
960 Bump __FreeBSD_version to 800083.
963 The layout of struct malloc_type, used by modules to register new
964 memory allocation types, has changed. Most modules will need to
965 be rebuilt or panics may be experienced.
966 Bump __FreeBSD_version to 800081.
969 Anticipate overflowing inp_flags - add inp_flags2.
970 This changes most offsets in inpcb, so checking v4 connection
971 state will require a world rebuild.
972 Bump __FreeBSD_version to 800080.
975 Add an llentry to struct route and struct route_in6. Modules
976 embedding a struct route will need to be recompiled.
977 Bump __FreeBSD_version to 800079.
980 The size of rt_metrics_lite and by extension rtentry has changed.
981 Networking administration apps will need to be recompiled.
982 The route command now supports show as an alias for get, weighting
983 of routes, sticky and nostick flags to alter the behavior of stateful
985 Bump __FreeBSD_version to 800078.
988 Do not use Giant for kbdmux(4) locking. This is wrong and
989 apparently causing more problems than it solves. This will
990 re-open the issue where interrupt handlers may race with
991 kbdmux(4) in polling mode. Typical symptoms include (but
992 not limited to) duplicated and/or missing characters when
993 low level console functions (such as gets) are used while
994 interrupts are enabled (for example geli password prompt,
995 mountroot prompt etc.). Disabling kbdmux(4) may help.
998 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
999 kernel modules referencing any of the above need to be recompiled.
1000 Bump __FreeBSD_version to 800075.
1003 GEOM_PART has become the default partition slicer for storage devices,
1004 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
1005 introduces some changes:
1007 MSDOS/EBR: the devices created from MSDOS extended partition entries
1008 (EBR) can be named differently than with GEOM_MBR and are now symlinks
1009 to devices with offset-based names. fstabs may need to be modified.
1011 BSD: the "geometry does not match label" warning is harmless in most
1012 cases but it points to problems in file system misalignment with
1013 disk geometry. The "c" partition is now implicit, covers the whole
1014 top-level drive and cannot be (mis)used by users.
1016 General: Kernel dumps are now not allowed to be written to devices
1017 whose partition types indicate they are meant to be used for file
1018 systems (or, in case of MSDOS partitions, as something else than
1021 Most of these changes date approximately from 200812.
1024 The uscanner(4) driver has been removed from the kernel. This follows
1025 Linux removing theirs in 2.6 and making libusb the default interface
1026 (supported by sane).
1029 The multicast forwarding code has been cleaned up. netstat(1)
1030 only relies on KVM now for printing bandwidth upcall meters.
1031 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1032 ip6_mroute_mod respectively. The config(5) options for statically
1033 compiling this code remain the same, i.e. 'options MROUTING'.
1036 Support for the IFF_NEEDSGIANT network interface flag has been
1037 removed, which means that non-MPSAFE network device drivers are no
1038 longer supported. In particular, if_ar, if_sr, and network device
1039 drivers from the old (legacy) USB stack can no longer be built or
1043 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1044 a bunch of new language catalog files have also been added.
1045 This means that some common libc messages are now localized and
1046 they depend on the LC_MESSAGES environmental variable.
1049 The k8temp(4) driver has been renamed to amdtemp(4) since
1050 support for Family 10 and Family 11 CPU families was added.
1053 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1054 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1056 For kernel developers, the most important changes are that the
1057 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1058 and this lock has been downgraded to a non-recursive mutex.
1060 Transport protocols (UDP, Raw IP) are now responsible for filtering
1061 inbound multicast traffic according to group membership and source
1062 filters. The imo_multicast_filter() KPI exists for this purpose.
1063 Transports which do not use multicast (SCTP, TCP) already reject
1064 multicast by default. Forwarding and receive performance may improve
1065 as a mutex acquisition is no longer needed in the ip_input()
1066 low-level input path. in_addmulti() and in_delmulti() are shimmed
1067 to new KPIs which exist to support SSM in-kernel.
1069 For application developers, it is recommended that loopback of
1070 multicast datagrams be disabled for best performance, as this
1071 will still cause the lock to be taken for each looped-back
1072 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1073 be tuned to 0 to disable loopback by default; it defaults to 1
1074 to preserve the existing behaviour.
1076 For systems administrators, to obtain best performance with
1077 multicast reception and multiple groups, it is always recommended
1078 that a card with a suitably precise hash filter is used. Hash
1079 collisions will still result in the lock being taken within the
1080 transport protocol input path to check group membership.
1082 If deploying FreeBSD in an environment with IGMP snooping switches,
1083 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1084 enabled; this forces 224.0.0.0/24 group membership to be announced
1087 The size of 'struct igmpstat' has changed; netstat needs to be
1088 recompiled to reflect this.
1089 Bump __FreeBSD_version to 800070.
1092 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1093 updated to use it. This requires a buildworld/installworld in order to
1094 update the library and dependencies (usbconfig, etc). Its advisable to
1095 rebuild all ports which uses libusb. More specific directions are given
1096 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1097 libusb are no longer required and can be removed.
1100 A workaround is committed to allow the creation of System V shared
1101 memory segment of size > 2 GB on the 64-bit architectures.
1102 Due to a limitation of the existing ABI, the shm_segsz member
1103 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1104 wrong for large segments. Note that limits must be explicitly
1105 raised to allow such segments to be created.
1108 The layout of struct ifnet has changed, requiring a rebuild of all
1109 network device driver modules.
1112 The /dev handling for the new USB stack has changed, a
1113 buildworld/installworld is required for libusb20.
1116 The new USB2 stack has now been permanently moved in and all kernel and
1117 module names reverted to their previous values (eg, usb, ehci, ohci,
1118 ums, ...). The old usb stack can be compiled in by prefixing the name
1119 with the letter 'o', the old usb modules have been removed.
1120 Updating entry 20090216 for xorg and 20090215 for libmap may still
1124 The rc.conf(5) option if_up_delay has been renamed to
1125 defaultroute_delay to better reflect its purpose. If you have
1126 customized this setting in /etc/rc.conf you need to update it to
1130 xorg 7.4 wants to configure its input devices via hald which does not
1131 yet work with USB2. If the keyboard/mouse does not work in xorg then
1133 Option "AllowEmptyInput" "off"
1134 to your ServerLayout section. This will cause X to use the configured
1135 kbd and mouse sections from your xorg.conf.
1138 The GENERIC kernels for all architectures now default to the new USB2
1139 stack. No kernel config options or code have been removed so if a
1140 problem arises please report it and optionally revert to the old USB
1141 stack. If you are loading USB kernel modules or have a custom kernel
1142 that includes GENERIC then ensure that usb names are also changed over,
1143 eg uftdi -> usb2_serial_ftdi.
1145 Older programs linked against the ports libusb 0.1 need to be
1146 redirected to the new stack's libusb20. /etc/libmap.conf can
1148 # Map old usb library to new one for usb2 stack
1149 libusb-0.1.so.8 libusb20.so.1
1152 All USB ethernet devices now attach as interfaces under the name ueN
1153 (eg. ue0). This is to provide a predictable name as vendors often
1154 change usb chipsets in a product without notice.
1157 The ichsmb(4) driver has been changed to require SMBus slave
1158 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1159 All of the other SMBus controller drivers require left-justified
1160 slave addresses, so this change makes all the drivers provide the
1164 INET6 statistics (struct ip6stat) was updated.
1165 netstat(1) needs to be recompiled.
1168 NTFS has been removed from GENERIC kernel on amd64 to match
1169 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1170 will load ntfs.ko module automatically when NTFS support is
1171 actually needed, unless ntfs.ko is not installed or security
1172 level prohibits loading kernel modules. If either is the case,
1173 "options NTFS" has to be added into kernel config.
1176 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1177 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1178 800061. User space tools that rely on the size of struct tcpcb in
1179 tcp_var.h (e.g. sockstat) need to be recompiled.
1182 ng_tty(4) module updated to match the new TTY subsystem.
1183 Due to API change, user-level applications must be updated.
1184 New API support added to mpd5 CVS and expected to be present
1185 in next mpd5.3 release.
1188 With __FreeBSD_version 800060 the makefs tool is part of
1189 the base system (it was a port).
1192 The afdata and ifnet locks have been changed from mutexes to
1193 rwlocks, network modules will need to be re-compiled.
1196 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1197 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1198 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1199 architecture and 40 bytes on 64-bit architecture. The userland
1200 applications "arp" and "ndp" have been updated accordingly.
1201 The output from "netstat -r" shows only routing entries and
1202 none of the L2 information.
1205 __FreeBSD_version 800057 marks the switchover from the
1206 binary ath hal to source code. Users must add the line:
1208 options AH_SUPPORT_AR5416
1210 to their kernel config files when specifying:
1214 The ath_hal module no longer exists; the code is now compiled
1215 together with the driver in the ath module. It is now
1216 possible to tailor chip support (i.e. reduce the set of chips
1217 and thereby the code size); consult ath_hal(4) for details.
1220 __FreeBSD_version 800054 adds memory barriers to
1221 <machine/atomic.h>, new interfaces to ifnet to facilitate
1222 multiple hardware transmit queues for cards that support
1223 them, and a lock-less ring-buffer implementation to
1224 enable drivers to more efficiently manage queueing of
1228 A new version of ZFS (version 13) has been merged to -HEAD.
1229 This version has zpool attribute "listsnapshots" off by
1230 default, which means "zfs list" does not show snapshots,
1231 and is the same as Solaris behavior.
1234 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1237 The uhci, ohci, ehci and slhci USB Host controller drivers have
1238 been put into separate modules. If you load the usb module
1239 separately through loader.conf you will need to load the
1240 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1241 controller add the following to loader.conf:
1247 The ABI used by the PMC toolset has changed. Please keep
1248 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1252 atapci kernel module now includes only generic PCI ATA
1253 driver. AHCI driver moved to ataahci kernel module.
1254 All vendor-specific code moved into separate kernel modules:
1255 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1256 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1257 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1258 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1261 The TTY subsystem of the kernel has been replaced by a new
1262 implementation, which provides better scalability and an
1263 improved driver model. Most common drivers have been migrated to
1264 the new TTY subsystem, while others have not. The following
1265 drivers have not yet been ported to the new TTY layer:
1268 cy, digi, rc, rp, sio
1274 ng_h4, ng_tty, ppp, sl, snp
1276 Adding these drivers to your kernel configuration file shall
1277 cause compilation to fail.
1280 ntpd has been upgraded to 4.2.4p5.
1283 OpenSSH has been upgraded to 5.1p1.
1285 For many years, FreeBSD's version of OpenSSH preferred DSA
1286 over RSA for host and user authentication keys. With this
1287 upgrade, we've switched to the vendor's default of RSA over
1288 DSA. This may cause upgraded clients to warn about unknown
1289 host keys even for previously known hosts. Users should
1290 follow the usual procedure for verifying host keys before
1291 accepting the RSA key.
1293 This can be circumvented by setting the "HostKeyAlgorithms"
1294 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1297 Please note that the sequence of keys offered for
1298 authentication has been changed as well. You may want to
1299 specify IdentityFile in a different order to revert this
1303 The sio(4) driver has been removed from the i386 and amd64
1304 kernel configuration files. This means uart(4) is now the
1305 default serial port driver on those platforms as well.
1307 To prevent collisions with the sio(4) driver, the uart(4) driver
1308 uses different names for its device nodes. This means the
1309 onboard serial port will now most likely be called "ttyu0"
1310 instead of "ttyd0". You may need to reconfigure applications to
1311 use the new device names.
1313 When using the serial port as a boot console, be sure to update
1314 /boot/device.hints and /etc/ttys before booting the new kernel.
1315 If you forget to do so, you can still manually specify the hints
1316 at the loader prompt:
1318 set hint.uart.0.at="isa"
1319 set hint.uart.0.port="0x3F8"
1320 set hint.uart.0.flags="0x10"
1321 set hint.uart.0.irq="4"
1325 The gpt(8) utility has been removed. Use gpart(8) to partition
1329 The version that Linuxulator emulates was changed from 2.4.2
1330 to 2.6.16. If you experience any problems with Linux binaries
1331 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1332 if it fixes the problem contact emulation mailing list.
1335 ISDN4BSD (I4B) was removed from the src tree. You may need to
1336 update a your kernel configuration and remove relevant entries.
1339 I have checked in code to support multiple routing tables.
1340 See the man pages setfib(1) and setfib(2).
1341 This is a hopefully backwards compatible version,
1342 but to make use of it you need to compile your kernel
1343 with options ROUTETABLES=2 (or more up to 16).
1346 The 802.11 wireless support was redone to enable multi-bss
1347 operation on devices that are capable. The underlying device
1348 is no longer used directly but instead wlanX devices are
1349 cloned with ifconfig. This requires changes to rc.conf files.
1350 For example, change:
1351 ifconfig_ath0="WPA DHCP"
1354 ifconfig_wlan0="WPA DHCP"
1355 see rc.conf(5) for more details. In addition, mergemaster of
1356 /etc/rc.d is highly recommended. Simultaneous update of userland
1357 and kernel wouldn't hurt either.
1359 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1360 modules were merged into the base wlan module. All references
1361 to these modules (e.g. in kernel config files) must be removed.
1364 psm(4) has gained write(2) support in native operation level.
1365 Arbitrary commands can be written to /dev/psm%d and status can
1366 be read back from it. Therefore, an application is responsible
1367 for status validation and error recovery. It is a no-op in
1368 other operation levels.
1371 Support for KSE threading has been removed from the kernel. To
1372 run legacy applications linked against KSE libmap.conf may
1373 be used. The following libmap.conf may be used to ensure
1374 compatibility with any prior release:
1376 libpthread.so.1 libthr.so.1
1377 libpthread.so.2 libthr.so.2
1378 libkse.so.3 libthr.so.3
1381 The layout of struct vmspace has changed. This affects libkvm
1382 and any executables that link against libkvm and use the
1383 kvm_getprocs() function. In particular, but not exclusively,
1384 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1385 The effects are minimal, but it's advisable to upgrade world
1389 The latest em driver no longer has support in it for the
1390 82575 adapter, this is now moved to the igb driver. The
1391 split was done to make new features that are incompatible
1392 with older hardware easier to do.
1395 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1396 likewise the kernel option is now GEOM_LINUX_LVM.
1399 The default NFS mount mode has changed from UDP to TCP for
1400 increased reliability. If you rely on (insecurely) NFS
1401 mounting across a firewall you may need to update your
1405 Belatedly note the addition of m_collapse for compacting
1409 The fts(3) structures have been changed to use adequate
1410 integer types for their members and so to be able to cope
1411 with huge file trees. The old fts(3) ABI is preserved
1412 through symbol versioning in libc, so third-party binaries
1413 using fts(3) should still work, although they will not take
1414 advantage of the extended types. At the same time, some
1415 third-party software might fail to build after this change
1416 due to unportable assumptions made in its source code about
1417 fts(3) structure members. Such software should be fixed
1418 by its vendor or, in the worst case, in the ports tree.
1419 FreeBSD_version 800015 marks this change for the unlikely
1420 case that a portable fix is impossible.
1423 To upgrade to -current after this date, you must be running
1424 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1425 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1428 The ADAPTIVE_GIANT kernel option has been retired because its
1429 functionality is the default now.
1432 The AT keyboard emulation of sunkbd(4) has been turned on
1433 by default. In order to make the special symbols of the Sun
1434 keyboards driven by sunkbd(4) work under X these now have
1435 to be configured the same way as Sun USB keyboards driven
1436 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1438 Option "XkbLayout" "us"
1439 Option "XkbRules" "xorg"
1440 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1443 It has been decided that it is desirable to provide ABI
1444 backwards compatibility to the FreeBSD 4/5/6 versions of the
1445 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1446 broken with the introduction of PCI domain support (see the
1447 20070930 entry). Unfortunately, this required the ABI of
1448 PCIOCGETCONF to be broken again in order to be able to
1449 provide backwards compatibility to the old version of that
1450 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1451 again. As for prominent ports this affects neither pciutils
1452 nor xorg-server this time, the hal port needs to be rebuilt
1456 The misnamed kthread_create() and friends have been renamed
1457 to kproc_create() etc. Many of the callers already
1458 used kproc_start()..
1459 I will return kthread_create() and friends in a while
1460 with implementations that actually create threads, not procs.
1461 Renaming corresponds with version 800002.
1470 Avoid using make -j when upgrading. While generally safe, there are
1471 sometimes problems using -j to upgrade. If your upgrade fails with
1472 -j, please try again without -j. From time to time in the past there
1473 have been problems using -j with buildworld and/or installworld. This
1474 is especially true when upgrading between "distant" versions (eg one
1475 that cross a major release boundary or several minor releases, or when
1476 several months have passed on the -current branch).
1478 Sometimes, obscure build problems are the result of environment
1479 poisoning. This can happen because the make utility reads its
1480 environment when searching for values for global variables. To run
1481 your build attempts in an "environmental clean room", prefix all make
1482 commands with 'env -i '. See the env(1) manual page for more details.
1484 When upgrading from one major version to another it is generally best
1485 to upgrade to the latest code in the currently installed branch first,
1486 then do an upgrade to the new branch. This is the best-tested upgrade
1487 path, and has the highest probability of being successful. Please try
1488 this approach before reporting problems with a major version upgrade.
1492 When upgrading the boot ZFS pool to a new version, always follow
1495 1.) recompile and reinstall the ZFS boot loader and boot block
1496 (this is part of "make buildworld" and "make installworld")
1498 2.) update the ZFS boot block on your boot drive
1500 The following example updates the ZFS boot block on the first
1501 partition (freebsd-boot) of a GPT partitioned drive ad0:
1502 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1504 Non-boot pools do not need these updates.
1508 If you are updating from a prior version of FreeBSD (even one just
1509 a few days old), you should follow this procedure. It is the most
1510 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1512 make kernel-toolchain
1513 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1514 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1516 To test a kernel once
1517 ---------------------
1518 If you just want to boot a kernel once (because you are not sure
1519 if it works, or if you want to boot a known bad kernel to provide
1520 debugging information) run
1521 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1522 nextboot -k testkernel
1524 To just build a kernel when you know that it won't mess you up
1525 --------------------------------------------------------------
1526 This assumes you are already running a CURRENT system. Replace
1527 ${arch} with the architecture of your machine (e.g. "i386",
1528 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1530 cd src/sys/${arch}/conf
1531 config KERNEL_NAME_HERE
1532 cd ../compile/KERNEL_NAME_HERE
1537 If this fails, go to the "To build a kernel" section.
1539 To rebuild everything and install it on the current system.
1540 -----------------------------------------------------------
1541 # Note: sometimes if you are running current you gotta do more than
1542 # is listed here if you are upgrading from a really old current.
1544 <make sure you have good level 0 dumps>
1546 make kernel KERNCONF=YOUR_KERNEL_HERE
1548 <reboot in single user> [3]
1556 To cross-install current onto a separate partition
1557 --------------------------------------------------
1558 # In this approach we use a separate partition to hold
1559 # current's root, 'usr', and 'var' directories. A partition
1560 # holding "/", "/usr" and "/var" should be about 2GB in
1563 <make sure you have good level 0 dumps>
1566 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1567 <maybe newfs current's root partition>
1568 <mount current's root partition on directory ${CURRENT_ROOT}>
1569 make installworld DESTDIR=${CURRENT_ROOT}
1570 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1571 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1572 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1573 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1574 <reboot into current>
1575 <do a "native" rebuild/install as described in the previous section>
1576 <maybe install compatibility libraries from ports/misc/compat*>
1580 To upgrade in-place from stable to current
1581 ----------------------------------------------
1582 <make sure you have good level 0 dumps>
1584 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1586 <reboot in single user> [3]
1593 Make sure that you've read the UPDATING file to understand the
1594 tweaks to various things you need. At this point in the life
1595 cycle of current, things change often and you are on your own
1596 to cope. The defaults can also change, so please read ALL of
1597 the UPDATING entries.
1599 Also, if you are tracking -current, you must be subscribed to
1600 freebsd-current@freebsd.org. Make sure that before you update
1601 your sources that you have read and understood all the recent
1602 messages there. If in doubt, please track -stable which has
1603 much fewer pitfalls.
1605 [1] If you have third party modules, such as vmware, you
1606 should disable them at this point so they don't crash your
1609 [3] From the bootblocks, boot -s, and then do
1614 adjkerntz -i # if CMOS is wall time
1615 Also, when doing a major release upgrade, it is required that
1616 you boot into single user mode to do the installworld.
1618 [4] Note: This step is non-optional. Failure to do this step
1619 can result in a significant reduction in the functionality of the
1620 system. Attempting to do it by hand is not recommended and those
1621 that pursue this avenue should read this file carefully, as well
1622 as the archives of freebsd-current and freebsd-hackers mailing lists
1623 for potential gotchas. The -U option is also useful to consider.
1624 See mergemaster(8) for more information.
1626 [5] Usually this step is a noop. However, from time to time
1627 you may need to do this if you get unknown user in the following
1628 step. It never hurts to do it all the time. You may need to
1629 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1630 install) after the buildworld before this step if you last updated
1631 from [78]-stable or 9-stable before 20130430.
1633 [6] This only deletes old files and directories. Old libraries
1634 can be deleted by "make delete-old-libs", but you have to make
1635 sure that no program is using those libraries anymore.
1637 [8] In order to have a kernel that can run the 4.x binaries needed to
1638 do an installworld, you must include the COMPAT_FREEBSD4 option in
1639 your kernel. Failure to do so may leave you with a system that is
1640 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1641 required to run the 5.x binaries on more recent kernels. And so on
1642 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1644 Make sure that you merge any new devices from GENERIC since the
1645 last time you updated your kernel config file.
1647 [9] When checking out sources, you must include the -P flag to have
1648 cvs prune empty directories.
1650 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1651 "?=" instead of the "=" assignment operator, so that buildworld can
1652 override the CPUTYPE if it needs to.
1654 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1655 not on the command line, or in /etc/make.conf. buildworld will
1656 warn if it is improperly defined.
1659 This file contains a list, in reverse chronological order, of major
1660 breakages in tracking -current. It is not guaranteed to be a complete
1661 list of such breakages, and only contains entries since October 10, 2007.
1662 If you need to see UPDATING entries from before that date, you will need
1663 to fetch an UPDATING file from an older FreeBSD release.
1665 Copyright information:
1667 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1669 Redistribution, publication, translation and use, with or without
1670 modification, in full or in part, in any form or format of this
1671 document are permitted without further permission from the author.
1673 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1674 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1675 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1676 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1677 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1678 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1679 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1680 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1681 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1682 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1683 POSSIBILITY OF SUCH DAMAGE.
1685 Contact Warner Losh if you have any questions about your use of