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 Items affecting the ports and packages system can be found in
10 /usr/ports/UPDATING. Please read that file before running portupgrade.
12 20141022: p21 FreeBSD-EN-14:10.tzdata
14 Time zone data file update. [EN-14:10]
16 20141021: p20 FreeBSD-SA-14:20.rtsold
17 FreeBSD-SA-14:21.routed
18 FreeBSD-SA-14:22.namei
19 FreeBSD-SA-14:23.openssl
21 Fix rtsold(8) remote buffer overflow vulnerability. [SA-14:20]
23 Fix routed(8) remote denial of service vulnerability. [SA-14:21]
25 Fix memory leak in sandboxed namei lookup. [SA-14:22]
27 Fix OpenSSL multiple vulnerabilities. [SA-14:23]
29 20140916: p19 FreeBSD-SA-14:19.tcp
30 Fix Denial of Service in TCP packet processing. [SA-14:19]
32 20140909: p18 FreeBSD-SA-14:18.openssl
33 Fix OpenSSL multiple vulnerabilities. [SA-14:18]
35 20140708: p17 FreeBSD-SA-14:17.kmem
36 Fix kernel memory disclosure in control messages and SCTP
37 notifications. [SA-14:17]
39 20140624: p16 FreeBSD-SA-14:16.file
40 FreeBSD-EN-14:08.heimdal
42 Fix multiple vulnerabilities in file(1) and libmagic(3).
45 Fix gss_pseudo_random interoperability issue. [EN-14:08]
47 20140605: p15 FreeBSD-SA-14:14.openssl
48 Fix OpenSSL multiple vulnerabilities. [SA-14:14]
50 20140603: p14 FreeBSD-SA-14:11.sendmail
51 FreeBSD-SA-14:12.ktrace
55 Fix sendmail improper close-on-exec flag handling. [SA-14:11]
57 Fix ktrace memory disclosure. [SA-14:12]
59 Fix incorrect error handling in PAM policy parser. [SA-14:13]
61 Fix triple-fault when executing from a threaded process.
64 20140513: p13 FreeBSD-EN-14:03.pkg
65 FreeBSD-EN-14:04.kldxref
68 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
70 Improve build repeatability for kldxref(8). [EN-14:04]
72 Fix data corruption with ciss(4). [EN-14:05]
74 20140430: p12 FreeBSD-SA-14:08.tcp
76 Fix TCP reassembly vulnerability. [SA-14:08]
78 20140408: p11 FreeBSD-SA-14:05.nfsserver
79 FreeBSD-SA-14:06.openssl
80 Fix deadlock in the NFS server. [SA-14:05]
82 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
84 20140114: p10 FreeBSD-SA-14:01.bsnmpd
87 FreeBSD-EN-14:01.random
89 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
91 Fix ntpd distributed reflection Denial of Service
92 vulnerability. [SA-14:02]
94 Fix BIND remote denial of service vulnerability. [SA-14:04]
96 Disable hardware RNGs by default. [EN-14:01]
98 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
100 20131128: p9 FreeBSD-EN-13:05.freebsd-update
101 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
103 20131026: p8 FreeBSD-EN-13:04.freebsd-update
104 Fix multiple freebsd-update bugs that break upgrading to
107 20130910: p7 FreeBSD-SA-13:12.ifioctl
108 FreeBSD-SA-13:13.nullfs
110 In IPv6 and NetATM, stop SIOCSIFADDR, SIOCSIFBRDADDR,
111 SIOCSIFDSTADDR and SIOCSIFNETMASK at the socket layer rather
112 than pass them on to the link layer without validation or
113 credential checks. [SA-13:12]
115 Prevent cross-mount hardlinks between different nullfs mounts
116 of the same underlying filesystem. [SA-13:13]
118 20130822: p6 FreeBSD-SA-13:09.ip_multicast
119 FreeBSD-SA-13:10.sctp
121 Fix an integer overflow in computing the size of a temporary buffer
122 can result in a buffer which is too small for the requested
125 Fix a bug that could lead to kernel memory disclosure with
126 SCTP state cookie. [13:10]
128 Fix a data corruption problem with mfi(4) operating on > 2TB
129 disks in a JBOD. [EN-13:03]
131 20130726: p5 FreeBSD-SA-13:07.bind FreeBSD-SA-13:08.nfsserver
132 Fix Denial of Service vulnerability in named(8). [13:07]
134 Fix a bug that allows remote client bypass the normal
135 access checks when when -network or -host restrictions are
136 used at the same time with -mapall. [13:08]
138 20130618: p4 FreeBSD-SA-13:06.mmap
139 Fix a bug that allowed a tracing process (e.g. gdb) to write
140 to a memory-mapped file in the traced process's address space
141 even if neither the traced process nor the tracing process had
142 write access to that file.
144 20130429: p3 FreeBSD-SA-13:05.nfsserver
145 Fix a bug that allows NFS clients to issue READDIR on files.
147 20130402: p2 FreeBSD-SA-13:03.openssl FreeBSD-SA-13:04.bind
148 Fix multiple vulnerabilities in OpenSSL.
150 Fix Denial of Service vulnerability in named(8).
152 20130218: p1 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
153 Fix Denial of Service vulnerability in named(8) with DNS64.
155 Fix Denial of Service vulnerability in libc's glob(3) functionality.
161 The sparc64 ZFS loader has been changed to no longer try to auto-
162 detect ZFS providers based on diskN aliases but now requires these
163 to be explicitly listed in the OFW boot-device environment variable.
166 Now unix domain sockets behave "as expected" on nullfs(5). Previously
167 nullfs(5) did not pass through all behaviours to the underlying layer,
168 as a result if we bound to a socket on the lower layer we could connect
169 only to the lower path; if we bound to the upper layer we could connect
170 only to the upper path. The new behavior is one can connect to both the
171 lower and the upper paths regardless what layer path one binds to.
174 The acpi_wmi(4) status device /dev/wmistat has been renamed to
178 A new VOP_ADVISE() was added to support posix_fadvise(2). All
179 filesystem modules must be recompiled.
182 The interface of the VOP_VPTOCNP(9) changed, now the returned
183 vnode shall be referenced, previously it was required to be
184 only held. All in-tree filesystems are converted.
190 The broken amd(4) driver has been replaced with esp(4) in the amd64,
191 i386 and pc98 GENERIC kernel configuration files.
194 This commit modifies vfs_register() so that it uses a hash
195 calculation to set vfc_typenum, which is enabled by default.
196 The first time a system is booted after this change, the
197 vfc_typenum values will change for all file systems. The
198 main effect of this is a change to the NFS server file handles
199 for file systems that use vfc_typenum in their fsid, such as ZFS.
200 It will, however, prevent vfc_typenum from changing when file
201 systems are loaded in a different order for subsequent reboots.
202 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
203 until you are ready to remount all NFS clients after a reboot.
206 Bump the shared library version numbers for libraries that
207 do not use symbol versioning, have changed the ABI compared
208 to stable/8 and which shared library version was not bumped.
209 Done as part of 9.0-RELEASE cycle.
212 During the merge of Capsicum features, the fget(9) KPI was modified.
213 This may require the rebuilding of out-of-tree device drivers --
214 issues have been reported specifically with the nVidia device driver.
215 __FreeBSD_version is bumped to 900041.
217 Also, there is a period between 20110811 and 20110814 where the
218 special devices /dev/{stdin,stdout,stderr} did not work correctly.
219 Building world from a kernel during that window may not work.
222 The packet filter (pf) code has been updated to OpenBSD 4.5.
223 You need to update userland tools to be in sync with kernel.
224 This update breaks backward compatibility with earlier pfsync(4)
225 versions. Care must be taken when updating redundant firewall setups.
228 The following sysctls and tunables are retired on x86 platforms:
230 machdep.hlt_logical_cpus
231 The following sysctl is retired:
232 machdep.hyperthreading_allowed
233 The sysctls were supposed to provide a way to dynamically offline and
234 online selected CPUs on x86 platforms, but the implementation has not
235 been reliable especially with SCHED_ULE scheduler.
236 machdep.hyperthreading_allowed tunable is still available to ignore
237 hyperthreading CPUs at OS level.
238 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
239 where X is an APIC ID of a CPU. Be advised, though, that disabling
240 CPUs in non-uniform fashion will result in non-uniform topology and
241 may lead to sub-optimal system performance with SCHED_ULE, which is
245 cpumask_t type is retired and cpuset_t is used in order to describe
249 Changes to ifconfig(8) for dynamic address family detection mandate
250 that you are running a kernel of 20110525 or later. Make sure to
251 follow the update procedure to boot a new kernel before installing
255 Support for sun4v architecture is officially dropped
258 Several KPI breaking changes have been committed to the mii(4) layer,
259 the PHY drivers and consequently some Ethernet drivers using mii(4).
260 This means that miibus.ko and the modules of the affected Ethernet
261 drivers need to be recompiled.
263 Note to kernel developers: Given that the OUI bit reversion problem
264 was fixed as part of these changes all mii(4) commits related to OUIs,
265 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
266 handling, no longer can be merged verbatim to stable/8 and previous
270 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
271 into their kernel configurations along with 'device pci'.
274 The default NFS client is now the new NFS client, so fstype "newnfs"
275 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
276 Although mounts via fstype "nfs" will usually work without userland
277 changes, it is recommended that the mount(8) and mount_nfs(8)
278 commands be rebuilt from sources and that a link to mount_nfs called
279 mount_oldnfs be created. The new client is compiled into the
280 kernel with "options NFSCL" and this is needed for diskless root
281 file systems. The GENERIC kernel configs have been changed to use
282 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
283 To use the regular/old client, you can "mount -t oldnfs ...". For
284 a diskless root file system, you must also include a line like:
286 vfs.root.mountfrom="oldnfs:"
288 in the boot/loader.conf on the root fs on the NFS server to make
289 a diskless root fs use the old client.
292 The GENERIC kernels for all architectures now default to the new
293 CAM-based ATA stack. It means that all legacy ATA drivers were
294 removed and replaced by respective CAM drivers. If you are using
295 ATA device names in /etc/fstab or other places, make sure to update
296 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
297 where 'Y's are the sequential numbers starting from zero for each type
298 in order of detection, unless configured otherwise with tunables,
299 see cam(4)). There will be symbolic links created in /dev/ to map
300 old adX devices to the respective adaY. They should provide basic
301 compatibility for file systems mounting in most cases, but they do
302 not support old user-level APIs and do not have respective providers
303 in GEOM. Consider using updated management tools with new device names.
305 It is possible to load devices ahci, ata, siis and mvs as modules,
306 but option ATA_CAM should remain in kernel configuration to make ata
307 module work as CAM driver supporting legacy ATA controllers. Device ata
308 still can be used in modular fashion (atacore + ...). Modules atadisk
309 and atapi* are not used and won't affect operation in ATA_CAM mode.
310 Note that to use CAM-based ATA kernel should include CAM devices
311 scbus, pass, da (or explicitly ada), cd and optionally others. All of
312 them are parts of the cam module.
314 ataraid(4) functionality is now supported by the RAID GEOM class.
315 To use it you can load geom_raid kernel module and use graid(8) tool
316 for management. Instead of /dev/arX device names, use /dev/raid/rX.
318 No kernel config options or code have been removed, so if a problem
319 arises, please report it and optionally revert to the old ATA stack.
320 In order to do it you can remove from the kernel config:
325 , and instead add back:
326 device atadisk # ATA disk drives
327 device ataraid # ATA RAID drives
328 device atapicd # ATAPI CDROM drives
329 device atapifd # ATAPI floppy drives
330 device atapist # ATAPI tape drives
333 The default NFS server has been changed to the new server, which
334 was referred to as the experimental server. If you need to switch
335 back to the old NFS server, you must now put the "-o" option on
336 both the mountd and nfsd commands. This can be done using the
337 mountd_flags and nfs_server_flags rc.conf variables until an
338 update to the rc scripts is committed, which is coming soon.
341 The GNU Objective-C runtime library (libobjc), and other Objective-C
342 related components have been removed from the base system. If you
343 require an Objective-C library, please use one of the available ports.
346 ath(4) has been split into bus- and device- modules. if_ath contains
347 the HAL, the TX rate control and the network device code. if_ath_pci
348 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
349 contains the AHB glue. Users need to load both if_ath_pci and if_ath
350 in order to use ath on everything else.
352 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
353 need to load if_ath and if_ath_pci for ath(4) operation.
356 As part of the replacement of sysinstall, the process of building
357 release media has changed significantly. For details, please re-read
358 release(7), which has been updated to reflect the new build process.
361 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
362 is the last available version under GPLv2. It brings a number of new
363 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
364 4.1 and SSE 4.2), better support for powerpc64, a number of new
365 directives, and lots of other small improvements. See the ChangeLog
366 file in contrib/binutils for the full details.
369 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
370 compliant, and will now use half of hash for authentication.
371 This will break interoperability with all stacks (including all
372 actual FreeBSD versions) who implement
373 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
375 The only workaround with such peers is to use another HMAC
376 algorithm for IPsec ("phase 2") authentication.
379 Remove the uio_yield prototype and symbol. This function has
380 been misnamed since it was introduced and should not be
381 globally exposed with this name. The equivalent functionality
382 is now available using kern_yield(curthread->td_user_pri).
383 The function remains undocumented.
386 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
387 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
388 for scalar sysctls is defined but disabled. Code that needs
389 UQUAD to pass the type checking that must compile on older
390 systems where the define is not present can check against
391 __FreeBSD_version >= 900030.
393 The system dialog(1) has been replaced with a new version previously
394 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
395 with the previous version, but the libdialog associated with it has
396 a largely incompatible API. As such, the original version of libdialog
397 will be kept temporarily as libodialog, until its base system consumers
398 are replaced or updated. Bump __FreeBSD_version to 900030.
401 If you are trying to run make universe on a -stable system, and you get
402 the following warning:
403 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
404 unknown. config(8) likely too old."
405 or something similar to it, then you must upgrade your -stable system
406 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
407 stable/8) or build the config from the latest stable/8 branch and
408 install it on your system.
410 Prior to this date, building a current universe on 8-stable system from
411 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
412 error in the first kernel build phase. A new config on those old
413 systems will fix that problem for older versions of -current.
416 The TCP stack has been modified to allow Khelp modules to interact with
417 it via helper hook points and store per-connection data in the TCP
418 control block. Bump __FreeBSD_version to 900029. User space tools that
419 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
423 Generic IEEE 802.3 annex 31B full duplex flow control support has been
424 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
425 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
426 to take advantage of it instead of using custom implementations. This
427 means that these drivers now no longer unconditionally advertise
428 support for flow control but only do so if flow control is a selected
429 media option. This was implemented in the generic support that way in
430 order to allow flow control to be switched on and off via ifconfig(8)
431 with the PHY specific default to typically off in order to protect
432 from unwanted effects. Consequently, if you used flow control with
433 one of the above mentioned drivers you now need to explicitly enable
435 ifconfig bge0 media auto mediaopt flowcontrol
437 Along with the above mentioned changes generic support for setting
438 1000baseT master mode also has been added and brgphy(4), ciphy(4),
439 e1000phy(4) as well as ip1000phy(4) have been converted to take
440 advantage of it. This means that these drivers now no longer take the
441 link0 parameter for selecting master mode but the master media option
442 has to be used instead, for example like in the following:
443 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
445 Selection of master mode now is also available with all other PHY
446 drivers supporting 1000baseT.
449 The TCP stack has received a significant update to add support for
450 modularised congestion control and generally improve the clarity of
451 congestion control decisions. Bump __FreeBSD_version to 900025. User
452 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
453 sockstat) need to be recompiled.
456 The man(1) utility has been replaced by a new version that no longer
457 uses /etc/manpath.config. Please consult man.conf(5) for how to
458 migrate local entries to the new format.
461 The copyright strings printed by login(1) and sshd(8) at the time of a
462 new connection have been removed to follow other operating systems and
466 A workaround for a fixed ld bug has been removed in kernel code,
467 so make sure that your system ld is built from sources after
468 revision 210245 from 2010-07-19 (r211583 if building head kernel
469 on stable/8, r211584 for stable/7; both from 2010-08-21).
470 A symptom of incorrect ld version is different addresses for
471 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
474 The $ipv6_prefer variable in rc.conf(5) has been split into
475 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
477 The $ip6addrctl_policy is a variable to choose a pre-defined
478 address selection policy set by ip6addrctl(8). A value
479 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
482 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
483 flag (see an entry of 20090926) is set on an interface with no
484 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
485 security reason. If you want IPv6 link-local address on all
486 interfaces by default, set this to "YES".
488 The old ipv6_prefer="YES" is equivalent to
489 ipv6_activate_all_interfaces="YES" and
490 ip6addrctl_policy="ipv6_prefer".
493 DTrace has grown support for userland tracing. Due to this, DTrace is
494 now i386 and amd64 only.
495 dtruss(1) is now installed by default on those systems and a new
496 kernel module is needed for userland tracing: fasttrap.
497 No changes to your kernel config file are necessary to enable
498 userland tracing, but you might consider adding 'STRIP=' and
499 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
500 to have informative userland stack traces in DTrace (ustack).
503 The acpi_aiboost(4) driver has been removed in favor of the new
504 aibs(4) driver. You should update your kernel configuration file.
507 BSD grep has been imported to the base system and it is built by
508 default. It is completely BSD licensed, highly GNU-compatible, uses
509 less memory than its GNU counterpart and has a small codebase.
510 However, it is slower than its GNU counterpart, which is mostly
511 noticeable for larger searches, for smaller ones it is measurable
512 but not significant. The reason is complex, the most important factor
513 is that we lack a modern and efficient regex library and GNU
514 overcomes this by optimizing the searches internally. Future work
515 on improving the regex performance is planned, for the meantime,
516 users that need better performance, can build GNU grep instead by
517 setting the WITH_GNU_GREP knob.
520 Due to the import of powerpc64 support, all existing powerpc kernel
521 configuration files must be updated with a machine directive like this:
522 machine powerpc powerpc
524 In addition, an updated config(8) is required to build powerpc kernels
528 A new version of ZFS (version 15) has been merged to -HEAD.
529 This version uses a python library for the following subcommands:
530 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
531 For full functionality of these commands the following port must
532 be installed: sysutils/py-zfs
535 'vm_page's are now hashed by physical address to an array of mutexes.
536 Currently this is only used to serialize access to hold_count. Over
537 time the page queue mutex will be peeled away. This changes the size
538 of pmap on every architecture. And requires all callers of vm_page_hold
539 and vm_page_unhold to be updated.
542 WITH_CTF can now be specified in src.conf (not recommended, there
543 are some problems with static executables), make.conf (would also
544 affect ports which do not use GNU make and do not override the
545 compile targets) or in the kernel config (via "makeoptions
547 When WITH_CTF was specified there before this was silently ignored,
548 so make sure that WITH_CTF is not used in places which could lead
549 to unwanted behavior.
552 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
553 to allow 32-bit compatibility on non-x86 platforms. All kernel
554 configurations on amd64 and ia64 platforms using these options must
555 be modified accordingly.
558 The utmp user accounting database has been replaced with utmpx,
559 the user accounting interface standardized by POSIX.
560 Unfortunately the semantics of utmp and utmpx don't match,
561 making it practically impossible to support both interfaces.
562 The user accounting database is used by tools like finger(1),
563 last(1), talk(1), w(1) and ac(8).
565 All applications in the base system use utmpx. This means only
566 local binaries (e.g. from the ports tree) may still use these
567 utmp database files. These applications must be rebuilt to make
570 After the system has been upgraded, it is safe to remove the old
571 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
572 assuming their contents is of no importance anymore. Old wtmp
573 databases can only be used by last(1) and ac(8) after they have
574 been converted to the new format using wtmpcvt(1).
577 Introduce the kernel thread "deadlock resolver" (which can be enabled
578 via the DEADLKRES option, see NOTES for more details) and the
579 sleepq_type() function for sleepqueues.
582 The rc.firewall and rc.firewall6 were unified, and
583 rc.firewall6 and rc.d/ip6fw were removed.
584 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
585 variables are obsoleted. Instead, the following new rc
586 variables are added to rc.d/ipfw:
588 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
589 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
590 firewall_simple_onet_ipv6, firewall_trusted_ipv6
592 The meanings correspond to the relevant IPv4 variables.
598 The default terminal emulation for syscons(4) has been changed
599 from cons25 to xterm on all platforms except pc98. This means
600 that the /etc/ttys file needs to be updated to ensure correct
601 operation of applications on the console.
603 The terminal emulation style can be toggled per window by using
604 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
605 options can be used to change the compile-time default back to
608 To prevent graphical artifacts, make sure the TERM environment
609 variable is set to match the terminal emulation that is being
610 performed by syscons(4).
613 The layout of the structure ieee80211req_scan_result has changed.
614 Applications that require wireless scan results (e.g. ifconfig(8))
615 from net80211 need to be recompiled.
617 Applications such as wpa_supplicant(8) may require a full world
618 build without using NO_CLEAN in order to get synchronized with the
622 The iwn(4) driver has been updated to support the 5000 and 5150 series.
623 There's one kernel module for each firmware. Adding "device iwnfw"
624 to the kernel configuration file means including all three firmware
625 images inside the kernel. If you want to include just the one for
626 your wireless card, use the devices iwn4965fw, iwn5000fw or
630 The rc.d/network_ipv6, IPv6 configuration script has been integrated
631 into rc.d/netif. The changes are the following:
633 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
634 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
635 Note that both variables need the "inet6" keyword at the head.
637 Do not set $ipv6_network_interfaces manually if you do not
638 understand what you are doing. It is not needed in most cases.
640 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
643 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
644 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
646 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
647 all configured interfaces have "inet6 accept_rtadv" in the
648 $ifconfig_IF_ipv6. These are for backward compatibility.
650 3. A new variable $ipv6_prefer has been added. If NO, IPv6
651 functionality of interfaces with no corresponding
652 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
653 and the default address selection policy of ip6addrctl(8)
654 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
655 Note that if you want to configure IPv6 functionality on the
656 disabled interfaces after boot, first you need to clear the flag by
657 using ifconfig(8) like:
659 ifconfig em0 inet6 -ifdisabled
661 If YES, the default address selection policy is set as
664 The default value of $ipv6_prefer is NO.
666 4. If your system need to receive Router Advertisement messages,
667 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
668 scripts automatically invoke rtsol(8) when the interface becomes
669 UP. The Router Advertisement messages are used for SLAAC
670 (State-Less Address AutoConfiguration).
673 802.11s D3.03 support was committed. This is incompatible with the
674 previous code, which was based on D3.0.
677 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
678 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
679 control whether accepting Router Advertisement messages or not.
680 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
681 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
682 The ifconfig(8) utility now supports these flags.
685 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
686 mount(8) and -a option for df(1) to see them.
689 The old tunable hw.bus.devctl_disable has been superseded by
690 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
691 replaced by hw.bus.devctl_queue=0. The default for this new tunable
695 Remove the option STOP_NMI. The default action is now to use NMI only
696 for KDB via the newly introduced function stop_cpus_hard() and
697 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
700 The stable/8 branch created in subversion. This corresponds to the
701 RELENG_8 branch in CVS.
704 Bump the shared library version numbers for all libraries that do not
705 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
706 __FreeBSD_version to 800105.
709 Due to changes in the implementation of virtual network stack support,
710 all network-related kernel modules must be recompiled. As this change
711 breaks the ABI, bump __FreeBSD_version to 800104.
714 The TOE interface to the TCP syncache has been modified to remove
715 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
716 The cxgb driver is the only TOE consumer affected by this change, and
717 needs to be recompiled along with the kernel. As this change breaks
718 the ABI, bump __FreeBSD_version to 800103.
721 Padding has been added to struct tcpcb, sackhint and tcpstat in
722 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
723 maintaining the ABI. However, this change breaks the ABI, so bump
724 __FreeBSD_version to 800102. User space tools that rely on the size of
725 any of these structs (e.g. sockstat) need to be recompiled.
728 The NFS_LEGACYRPC option has been removed along with the old kernel
729 RPC implementation that this option selected. Kernel configurations
730 may need to be adjusted.
733 The network interface device nodes at /dev/net/<interface> have been
734 removed. All ioctl operations can be performed the normal way using
735 routing sockets. The kqueue functionality can generally be replaced
736 with routing sockets.
739 The documentation from the FreeBSD Documentation Project (Handbook,
740 FAQ, etc.) is now installed via packages by sysinstall(8) and under
741 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
744 The ABI of various structures related to the SYSV IPC API have been
745 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
746 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
750 Layout of struct vnet has changed as routing related variables were
751 moved to their own Vimage module. Modules need to be recompiled. Bump
752 __FreeBSD_version to 800099.
755 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
756 respectively. As long as no more than 16 groups per process are used,
757 no changes should be visible. When more than 16 groups are used, old
758 binaries may fail if they call getgroups() or getgrouplist() with
759 statically sized storage. Recompiling will work around this, but
760 applications should be modified to use dynamically allocated storage
761 for group arrays as POSIX.1-2008 does not cap an implementation's
762 number of supported groups at NGROUPS_MAX+1 as previous versions did.
764 NFS and portalfs mounts may also be affected as the list of groups is
765 truncated to 16. Users of NFS who use more than 16 groups, should
766 take care that negative group permissions are not used on the exported
767 file systems as they will not be reliable unless a GSSAPI based
768 authentication method is used.
771 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
772 option compiles in the support for adaptive spinning for lockmgrs
773 which want to enable it. The lockinit() function now accepts the flag
774 LK_ADAPTIVE in order to make the lock object subject to adaptive
775 spinning when both held in write and read mode.
778 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
779 changed. User applications that use this ioctl need to be rebuilt.
782 The layout of struct thread has changed. Kernel and modules need to
786 The layout of structs ifnet, domain, protosw and vnet_net has changed.
787 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
790 window(1) has been removed from the base system. It can now be
791 installed from ports. The port is called misc/window.
794 The way we are storing and accessing `routing table' entries has
795 changed. Programs reading the FIB, like netstat, need to be
799 A new netisr implementation has been added for FreeBSD 8. Network
800 file system modules, such as igmp, ipdivert, and others, should be
802 Bump __FreeBSD_version to 800096.
805 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
809 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
810 Bump __FreeBSD_version to 800094.
813 Add mnt_xflag field to 'struct mount'. File system modules need to be
815 Bump __FreeBSD_version to 800093.
818 The compiling option ADAPTIVE_SX has been retired while it has been
819 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
820 The KPI for sx_init_flags() changes as accepting flags:
821 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
822 been introduced in order to handle the reversed logic.
823 Bump __FreeBSD_version to 800092.
826 Add support for hierarchical jails. Remove global securelevel.
827 Bump __FreeBSD_version to 800091.
830 The layout of struct vnet_net has changed, therefore modules
832 Bump __FreeBSD_version to 800090.
835 The newly imported zic(8) produces a new format in the output. Please
836 run tzsetup(8) to install the newly created data to /etc/localtime.
839 The sysctl tree for the usb stack has renamed from hw.usb2.* to
840 hw.usb.* and is now consistent again with previous releases.
843 802.11 monitor mode support was revised and driver api's were changed.
844 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
845 of DLT_IEEE802_11. No user-visible data structures were changed but
846 applications that use DLT_IEEE802_11 may require changes.
847 Bump __FreeBSD_version to 800088.
850 The layout of the following structs has changed: sysctl_oid,
851 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
852 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
853 panics may be experienced. World rebuild is required for
854 correctly checking networking state from userland.
855 Bump __FreeBSD_version to 800085.
858 MLDv2 and Source-Specific Multicast (SSM) have been merged
859 to the IPv6 stack. VIMAGE hooks are in but not yet used.
860 The implementation of SSM within FreeBSD's IPv6 stack closely
861 follows the IPv4 implementation.
863 For kernel developers:
865 * The most important changes are that the ip6_output() and
866 ip6_input() paths no longer take the IN6_MULTI_LOCK,
867 and this lock has been downgraded to a non-recursive mutex.
869 * As with the changes to the IPv4 stack to support SSM, filtering
870 of inbound multicast traffic must now be performed by transport
871 protocols within the IPv6 stack. This does not apply to TCP and
872 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
874 * The KPIs used by IPv6 multicast are similar to those used by
875 the IPv4 stack, with the following differences:
876 * im6o_mc_filter() is analogous to imo_multicast_filter().
877 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
878 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
879 * IN6_LOOKUP_MULTI() has been deprecated and removed.
880 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
881 for MLDv1 have an additional 'timer' argument which is used to
882 jitter the initial membership report for the solicited-node
883 multicast membership on-link.
884 * This is not strictly needed for MLDv2, which already jitters
885 its report transmissions. However, the 'timer' argument is
886 preserved in case MLDv1 is active on the interface.
888 * The KAME linked-list based IPv6 membership implementation has
889 been refactored to use a vector similar to that used by the IPv4
891 Code which maintains a list of its own multicast memberships
892 internally, e.g. carp, has been updated to reflect the new
895 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
896 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
897 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
898 implementation constraint which needs to be addressed in HEAD.
900 For application developers:
902 * The changes are broadly similar to those made for the IPv4
905 * The use of IPv4 and IPv6 multicast socket options on the same
906 socket, using mapped addresses, HAS NOT been tested or supported.
908 * There are a number of issues with the implementation of various
909 IPv6 multicast APIs which need to be resolved in the API surface
910 before the implementation is fully compatible with KAME userland
911 use, and these are mostly to do with interface index treatment.
913 * The literature available discusses the use of either the delta / ASM
914 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
915 using setsourcefilter(3)/getsourcefilter(3). For more information
916 please refer to RFC 3768, 'Socket Interface Extensions for
917 Multicast Source Filters'.
919 * Applications which use the published RFC 3678 APIs should be fine.
921 For systems administrators:
923 * The mtest(8) utility has been refactored to support IPv6, in
924 addition to IPv4. Interface addresses are no longer accepted
925 as arguments, their names must be used instead. The utility
926 will map the interface name to its first IPv4 address as
927 returned by getifaddrs(3).
929 * The ifmcstat(8) utility has also been updated to print the MLDv2
930 endpoint state and source filter lists via sysctl(3).
932 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
933 loopback of IPv6 multicast datagrams by default; it defaults to 1
934 to preserve the existing behaviour. Disabling multicast loopback is
935 recommended for optimal system performance.
937 * The IPv6 MROUTING code has been changed to examine this sysctl
938 instead of attempting to perform a group lookup before looping
939 back forwarded datagrams.
941 Bump __FreeBSD_version to 800084.
944 Implement low-level Bluetooth HCI API.
945 Bump __FreeBSD_version to 800083.
948 The layout of struct malloc_type, used by modules to register new
949 memory allocation types, has changed. Most modules will need to
950 be rebuilt or panics may be experienced.
951 Bump __FreeBSD_version to 800081.
954 Anticipate overflowing inp_flags - add inp_flags2.
955 This changes most offsets in inpcb, so checking v4 connection
956 state will require a world rebuild.
957 Bump __FreeBSD_version to 800080.
960 Add an llentry to struct route and struct route_in6. Modules
961 embedding a struct route will need to be recompiled.
962 Bump __FreeBSD_version to 800079.
965 The size of rt_metrics_lite and by extension rtentry has changed.
966 Networking administration apps will need to be recompiled.
967 The route command now supports show as an alias for get, weighting
968 of routes, sticky and nostick flags to alter the behavior of stateful
970 Bump __FreeBSD_version to 800078.
973 Do not use Giant for kbdmux(4) locking. This is wrong and
974 apparently causing more problems than it solves. This will
975 re-open the issue where interrupt handlers may race with
976 kbdmux(4) in polling mode. Typical symptoms include (but
977 not limited to) duplicated and/or missing characters when
978 low level console functions (such as gets) are used while
979 interrupts are enabled (for example geli password prompt,
980 mountroot prompt etc.). Disabling kbdmux(4) may help.
983 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
984 kernel modules referencing any of the above need to be recompiled.
985 Bump __FreeBSD_version to 800075.
988 GEOM_PART has become the default partition slicer for storage devices,
989 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
990 introduces some changes:
992 MSDOS/EBR: the devices created from MSDOS extended partition entries
993 (EBR) can be named differently than with GEOM_MBR and are now symlinks
994 to devices with offset-based names. fstabs may need to be modified.
996 BSD: the "geometry does not match label" warning is harmless in most
997 cases but it points to problems in file system misalignment with
998 disk geometry. The "c" partition is now implicit, covers the whole
999 top-level drive and cannot be (mis)used by users.
1001 General: Kernel dumps are now not allowed to be written to devices
1002 whose partition types indicate they are meant to be used for file
1003 systems (or, in case of MSDOS partitions, as something else than
1006 Most of these changes date approximately from 200812.
1009 The uscanner(4) driver has been removed from the kernel. This follows
1010 Linux removing theirs in 2.6 and making libusb the default interface
1011 (supported by sane).
1014 The multicast forwarding code has been cleaned up. netstat(1)
1015 only relies on KVM now for printing bandwidth upcall meters.
1016 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1017 ip6_mroute_mod respectively. The config(5) options for statically
1018 compiling this code remain the same, i.e. 'options MROUTING'.
1021 Support for the IFF_NEEDSGIANT network interface flag has been
1022 removed, which means that non-MPSAFE network device drivers are no
1023 longer supported. In particular, if_ar, if_sr, and network device
1024 drivers from the old (legacy) USB stack can no longer be built or
1028 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1029 a bunch of new language catalog files have also been added.
1030 This means that some common libc messages are now localized and
1031 they depend on the LC_MESSAGES environmental variable.
1034 The k8temp(4) driver has been renamed to amdtemp(4) since
1035 support for Family 10 and Family 11 CPU families was added.
1038 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1039 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1041 For kernel developers, the most important changes are that the
1042 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1043 and this lock has been downgraded to a non-recursive mutex.
1045 Transport protocols (UDP, Raw IP) are now responsible for filtering
1046 inbound multicast traffic according to group membership and source
1047 filters. The imo_multicast_filter() KPI exists for this purpose.
1048 Transports which do not use multicast (SCTP, TCP) already reject
1049 multicast by default. Forwarding and receive performance may improve
1050 as a mutex acquisition is no longer needed in the ip_input()
1051 low-level input path. in_addmulti() and in_delmulti() are shimmed
1052 to new KPIs which exist to support SSM in-kernel.
1054 For application developers, it is recommended that loopback of
1055 multicast datagrams be disabled for best performance, as this
1056 will still cause the lock to be taken for each looped-back
1057 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1058 be tuned to 0 to disable loopback by default; it defaults to 1
1059 to preserve the existing behaviour.
1061 For systems administrators, to obtain best performance with
1062 multicast reception and multiple groups, it is always recommended
1063 that a card with a suitably precise hash filter is used. Hash
1064 collisions will still result in the lock being taken within the
1065 transport protocol input path to check group membership.
1067 If deploying FreeBSD in an environment with IGMP snooping switches,
1068 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1069 enabled; this forces 224.0.0.0/24 group membership to be announced
1072 The size of 'struct igmpstat' has changed; netstat needs to be
1073 recompiled to reflect this.
1074 Bump __FreeBSD_version to 800070.
1077 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1078 updated to use it. This requires a buildworld/installworld in order to
1079 update the library and dependencies (usbconfig, etc). Its advisable to
1080 rebuild all ports which uses libusb. More specific directions are given
1081 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1082 libusb are no longer required and can be removed.
1085 A workaround is committed to allow the creation of System V shared
1086 memory segment of size > 2 GB on the 64-bit architectures.
1087 Due to a limitation of the existing ABI, the shm_segsz member
1088 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1089 wrong for large segments. Note that limits must be explicitly
1090 raised to allow such segments to be created.
1093 The layout of struct ifnet has changed, requiring a rebuild of all
1094 network device driver modules.
1097 The /dev handling for the new USB stack has changed, a
1098 buildworld/installworld is required for libusb20.
1101 The new USB2 stack has now been permanently moved in and all kernel and
1102 module names reverted to their previous values (eg, usb, ehci, ohci,
1103 ums, ...). The old usb stack can be compiled in by prefixing the name
1104 with the letter 'o', the old usb modules have been removed.
1105 Updating entry 20090216 for xorg and 20090215 for libmap may still
1109 The rc.conf(5) option if_up_delay has been renamed to
1110 defaultroute_delay to better reflect its purpose. If you have
1111 customized this setting in /etc/rc.conf you need to update it to
1115 xorg 7.4 wants to configure its input devices via hald which does not
1116 yet work with USB2. If the keyboard/mouse does not work in xorg then
1118 Option "AllowEmptyInput" "off"
1119 to your ServerLayout section. This will cause X to use the configured
1120 kbd and mouse sections from your xorg.conf.
1123 The GENERIC kernels for all architectures now default to the new USB2
1124 stack. No kernel config options or code have been removed so if a
1125 problem arises please report it and optionally revert to the old USB
1126 stack. If you are loading USB kernel modules or have a custom kernel
1127 that includes GENERIC then ensure that usb names are also changed over,
1128 eg uftdi -> usb2_serial_ftdi.
1130 Older programs linked against the ports libusb 0.1 need to be
1131 redirected to the new stack's libusb20. /etc/libmap.conf can
1133 # Map old usb library to new one for usb2 stack
1134 libusb-0.1.so.8 libusb20.so.1
1137 All USB ethernet devices now attach as interfaces under the name ueN
1138 (eg. ue0). This is to provide a predictable name as vendors often
1139 change usb chipsets in a product without notice.
1142 The ichsmb(4) driver has been changed to require SMBus slave
1143 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1144 All of the other SMBus controller drivers require left-justified
1145 slave addresses, so this change makes all the drivers provide the
1149 INET6 statistics (struct ip6stat) was updated.
1150 netstat(1) needs to be recompiled.
1153 NTFS has been removed from GENERIC kernel on amd64 to match
1154 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1155 will load ntfs.ko module automatically when NTFS support is
1156 actually needed, unless ntfs.ko is not installed or security
1157 level prohibits loading kernel modules. If either is the case,
1158 "options NTFS" has to be added into kernel config.
1161 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1162 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1163 800061. User space tools that rely on the size of struct tcpcb in
1164 tcp_var.h (e.g. sockstat) need to be recompiled.
1167 ng_tty(4) module updated to match the new TTY subsystem.
1168 Due to API change, user-level applications must be updated.
1169 New API support added to mpd5 CVS and expected to be present
1170 in next mpd5.3 release.
1173 With __FreeBSD_version 800060 the makefs tool is part of
1174 the base system (it was a port).
1177 The afdata and ifnet locks have been changed from mutexes to
1178 rwlocks, network modules will need to be re-compiled.
1181 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1182 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1183 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1184 architecture and 40 bytes on 64-bit architecture. The userland
1185 applications "arp" and "ndp" have been updated accordingly.
1186 The output from "netstat -r" shows only routing entries and
1187 none of the L2 information.
1190 __FreeBSD_version 800057 marks the switchover from the
1191 binary ath hal to source code. Users must add the line:
1193 options AH_SUPPORT_AR5416
1195 to their kernel config files when specifying:
1199 The ath_hal module no longer exists; the code is now compiled
1200 together with the driver in the ath module. It is now
1201 possible to tailor chip support (i.e. reduce the set of chips
1202 and thereby the code size); consult ath_hal(4) for details.
1205 __FreeBSD_version 800054 adds memory barriers to
1206 <machine/atomic.h>, new interfaces to ifnet to facilitate
1207 multiple hardware transmit queues for cards that support
1208 them, and a lock-less ring-buffer implementation to
1209 enable drivers to more efficiently manage queueing of
1213 A new version of ZFS (version 13) has been merged to -HEAD.
1214 This version has zpool attribute "listsnapshots" off by
1215 default, which means "zfs list" does not show snapshots,
1216 and is the same as Solaris behavior.
1219 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1222 The uhci, ohci, ehci and slhci USB Host controller drivers have
1223 been put into separate modules. If you load the usb module
1224 separately through loader.conf you will need to load the
1225 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1226 controller add the following to loader.conf:
1232 The ABI used by the PMC toolset has changed. Please keep
1233 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1237 atapci kernel module now includes only generic PCI ATA
1238 driver. AHCI driver moved to ataahci kernel module.
1239 All vendor-specific code moved into separate kernel modules:
1240 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1241 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1242 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1243 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1246 The TTY subsystem of the kernel has been replaced by a new
1247 implementation, which provides better scalability and an
1248 improved driver model. Most common drivers have been migrated to
1249 the new TTY subsystem, while others have not. The following
1250 drivers have not yet been ported to the new TTY layer:
1253 cy, digi, rc, rp, sio
1259 ng_h4, ng_tty, ppp, sl, snp
1261 Adding these drivers to your kernel configuration file shall
1262 cause compilation to fail.
1265 ntpd has been upgraded to 4.2.4p5.
1268 OpenSSH has been upgraded to 5.1p1.
1270 For many years, FreeBSD's version of OpenSSH preferred DSA
1271 over RSA for host and user authentication keys. With this
1272 upgrade, we've switched to the vendor's default of RSA over
1273 DSA. This may cause upgraded clients to warn about unknown
1274 host keys even for previously known hosts. Users should
1275 follow the usual procedure for verifying host keys before
1276 accepting the RSA key.
1278 This can be circumvented by setting the "HostKeyAlgorithms"
1279 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1282 Please note that the sequence of keys offered for
1283 authentication has been changed as well. You may want to
1284 specify IdentityFile in a different order to revert this
1288 The sio(4) driver has been removed from the i386 and amd64
1289 kernel configuration files. This means uart(4) is now the
1290 default serial port driver on those platforms as well.
1292 To prevent collisions with the sio(4) driver, the uart(4) driver
1293 uses different names for its device nodes. This means the
1294 onboard serial port will now most likely be called "ttyu0"
1295 instead of "ttyd0". You may need to reconfigure applications to
1296 use the new device names.
1298 When using the serial port as a boot console, be sure to update
1299 /boot/device.hints and /etc/ttys before booting the new kernel.
1300 If you forget to do so, you can still manually specify the hints
1301 at the loader prompt:
1303 set hint.uart.0.at="isa"
1304 set hint.uart.0.port="0x3F8"
1305 set hint.uart.0.flags="0x10"
1306 set hint.uart.0.irq="4"
1310 The gpt(8) utility has been removed. Use gpart(8) to partition
1314 The version that Linuxulator emulates was changed from 2.4.2
1315 to 2.6.16. If you experience any problems with Linux binaries
1316 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1317 if it fixes the problem contact emulation mailing list.
1320 ISDN4BSD (I4B) was removed from the src tree. You may need to
1321 update a your kernel configuration and remove relevant entries.
1324 I have checked in code to support multiple routing tables.
1325 See the man pages setfib(1) and setfib(2).
1326 This is a hopefully backwards compatible version,
1327 but to make use of it you need to compile your kernel
1328 with options ROUTETABLES=2 (or more up to 16).
1331 The 802.11 wireless support was redone to enable multi-bss
1332 operation on devices that are capable. The underlying device
1333 is no longer used directly but instead wlanX devices are
1334 cloned with ifconfig. This requires changes to rc.conf files.
1335 For example, change:
1336 ifconfig_ath0="WPA DHCP"
1339 ifconfig_wlan0="WPA DHCP"
1340 see rc.conf(5) for more details. In addition, mergemaster of
1341 /etc/rc.d is highly recommended. Simultaneous update of userland
1342 and kernel wouldn't hurt either.
1344 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1345 modules were merged into the base wlan module. All references
1346 to these modules (e.g. in kernel config files) must be removed.
1349 psm(4) has gained write(2) support in native operation level.
1350 Arbitrary commands can be written to /dev/psm%d and status can
1351 be read back from it. Therefore, an application is responsible
1352 for status validation and error recovery. It is a no-op in
1353 other operation levels.
1356 Support for KSE threading has been removed from the kernel. To
1357 run legacy applications linked against KSE libmap.conf may
1358 be used. The following libmap.conf may be used to ensure
1359 compatibility with any prior release:
1361 libpthread.so.1 libthr.so.1
1362 libpthread.so.2 libthr.so.2
1363 libkse.so.3 libthr.so.3
1366 The layout of struct vmspace has changed. This affects libkvm
1367 and any executables that link against libkvm and use the
1368 kvm_getprocs() function. In particular, but not exclusively,
1369 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1370 The effects are minimal, but it's advisable to upgrade world
1374 The latest em driver no longer has support in it for the
1375 82575 adapter, this is now moved to the igb driver. The
1376 split was done to make new features that are incompatible
1377 with older hardware easier to do.
1380 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1381 likewise the kernel option is now GEOM_LINUX_LVM.
1384 The default NFS mount mode has changed from UDP to TCP for
1385 increased reliability. If you rely on (insecurely) NFS
1386 mounting across a firewall you may need to update your
1390 Belatedly note the addition of m_collapse for compacting
1394 The fts(3) structures have been changed to use adequate
1395 integer types for their members and so to be able to cope
1396 with huge file trees. The old fts(3) ABI is preserved
1397 through symbol versioning in libc, so third-party binaries
1398 using fts(3) should still work, although they will not take
1399 advantage of the extended types. At the same time, some
1400 third-party software might fail to build after this change
1401 due to unportable assumptions made in its source code about
1402 fts(3) structure members. Such software should be fixed
1403 by its vendor or, in the worst case, in the ports tree.
1404 FreeBSD_version 800015 marks this change for the unlikely
1405 case that a portable fix is impossible.
1408 To upgrade to -current after this date, you must be running
1409 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1410 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1413 The ADAPTIVE_GIANT kernel option has been retired because its
1414 functionality is the default now.
1417 The AT keyboard emulation of sunkbd(4) has been turned on
1418 by default. In order to make the special symbols of the Sun
1419 keyboards driven by sunkbd(4) work under X these now have
1420 to be configured the same way as Sun USB keyboards driven
1421 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1423 Option "XkbLayout" "us"
1424 Option "XkbRules" "xorg"
1425 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1428 It has been decided that it is desirable to provide ABI
1429 backwards compatibility to the FreeBSD 4/5/6 versions of the
1430 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1431 broken with the introduction of PCI domain support (see the
1432 20070930 entry). Unfortunately, this required the ABI of
1433 PCIOCGETCONF to be broken again in order to be able to
1434 provide backwards compatibility to the old version of that
1435 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1436 again. As for prominent ports this affects neither pciutils
1437 nor xorg-server this time, the hal port needs to be rebuilt
1441 The misnamed kthread_create() and friends have been renamed
1442 to kproc_create() etc. Many of the callers already
1443 used kproc_start()..
1444 I will return kthread_create() and friends in a while
1445 with implementations that actually create threads, not procs.
1446 Renaming corresponds with version 800002.
1455 Avoid using make -j when upgrading. While generally safe, there are
1456 sometimes problems using -j to upgrade. If your upgrade fails with
1457 -j, please try again without -j. From time to time in the past there
1458 have been problems using -j with buildworld and/or installworld. This
1459 is especially true when upgrading between "distant" versions (eg one
1460 that cross a major release boundary or several minor releases, or when
1461 several months have passed on the -current branch).
1463 Sometimes, obscure build problems are the result of environment
1464 poisoning. This can happen because the make utility reads its
1465 environment when searching for values for global variables. To run
1466 your build attempts in an "environmental clean room", prefix all make
1467 commands with 'env -i '. See the env(1) manual page for more details.
1469 When upgrading from one major version to another it is generally best
1470 to upgrade to the latest code in the currently installed branch first,
1471 then do an upgrade to the new branch. This is the best-tested upgrade
1472 path, and has the highest probability of being successful. Please try
1473 this approach before reporting problems with a major version upgrade.
1477 When upgrading the boot ZFS pool to a new version, always follow
1480 1.) recompile and reinstall the ZFS boot loader and boot block
1481 (this is part of "make buildworld" and "make installworld")
1483 2.) update the ZFS boot block on your boot drive
1485 The following example updates the ZFS boot block on the first
1486 partition (freebsd-boot) of a GPT partitioned drive ad0:
1487 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1489 Non-boot pools do not need these updates.
1493 If you are updating from a prior version of FreeBSD (even one just
1494 a few days old), you should follow this procedure. It is the most
1495 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1497 make kernel-toolchain
1498 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1499 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1501 To test a kernel once
1502 ---------------------
1503 If you just want to boot a kernel once (because you are not sure
1504 if it works, or if you want to boot a known bad kernel to provide
1505 debugging information) run
1506 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1507 nextboot -k testkernel
1509 To just build a kernel when you know that it won't mess you up
1510 --------------------------------------------------------------
1511 This assumes you are already running a CURRENT system. Replace
1512 ${arch} with the architecture of your machine (e.g. "i386",
1513 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1515 cd src/sys/${arch}/conf
1516 config KERNEL_NAME_HERE
1517 cd ../compile/KERNEL_NAME_HERE
1522 If this fails, go to the "To build a kernel" section.
1524 To rebuild everything and install it on the current system.
1525 -----------------------------------------------------------
1526 # Note: sometimes if you are running current you gotta do more than
1527 # is listed here if you are upgrading from a really old current.
1529 <make sure you have good level 0 dumps>
1531 make kernel KERNCONF=YOUR_KERNEL_HERE
1533 <reboot in single user> [3]
1541 To cross-install current onto a separate partition
1542 --------------------------------------------------
1543 # In this approach we use a separate partition to hold
1544 # current's root, 'usr', and 'var' directories. A partition
1545 # holding "/", "/usr" and "/var" should be about 2GB in
1548 <make sure you have good level 0 dumps>
1551 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1552 <maybe newfs current's root partition>
1553 <mount current's root partition on directory ${CURRENT_ROOT}>
1554 make installworld DESTDIR=${CURRENT_ROOT}
1555 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1556 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1557 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1558 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1559 <reboot into current>
1560 <do a "native" rebuild/install as described in the previous section>
1561 <maybe install compatibility libraries from ports/misc/compat*>
1565 To upgrade in-place from 8.x-stable to current
1566 ----------------------------------------------
1567 <make sure you have good level 0 dumps>
1569 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1571 <reboot in single user> [3]
1578 Make sure that you've read the UPDATING file to understand the
1579 tweaks to various things you need. At this point in the life
1580 cycle of current, things change often and you are on your own
1581 to cope. The defaults can also change, so please read ALL of
1582 the UPDATING entries.
1584 Also, if you are tracking -current, you must be subscribed to
1585 freebsd-current@freebsd.org. Make sure that before you update
1586 your sources that you have read and understood all the recent
1587 messages there. If in doubt, please track -stable which has
1588 much fewer pitfalls.
1590 [1] If you have third party modules, such as vmware, you
1591 should disable them at this point so they don't crash your
1594 [3] From the bootblocks, boot -s, and then do
1599 adjkerntz -i # if CMOS is wall time
1600 Also, when doing a major release upgrade, it is required that
1601 you boot into single user mode to do the installworld.
1603 [4] Note: This step is non-optional. Failure to do this step
1604 can result in a significant reduction in the functionality of the
1605 system. Attempting to do it by hand is not recommended and those
1606 that pursue this avenue should read this file carefully, as well
1607 as the archives of freebsd-current and freebsd-hackers mailing lists
1608 for potential gotchas. The -U option is also useful to consider.
1609 See mergemaster(8) for more information.
1611 [5] Usually this step is a noop. However, from time to time
1612 you may need to do this if you get unknown user in the following
1613 step. It never hurts to do it all the time. You may need to
1614 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1615 install) after the buildworld before this step if you last updated
1616 from current before 20020224 or from -stable before 20020408.
1618 [6] This only deletes old files and directories. Old libraries
1619 can be deleted by "make delete-old-libs", but you have to make
1620 sure that no program is using those libraries anymore.
1622 [8] In order to have a kernel that can run the 4.x binaries needed to
1623 do an installworld, you must include the COMPAT_FREEBSD4 option in
1624 your kernel. Failure to do so may leave you with a system that is
1625 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1626 required to run the 5.x binaries on more recent kernels. And so on
1627 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1629 Make sure that you merge any new devices from GENERIC since the
1630 last time you updated your kernel config file.
1632 [9] When checking out sources, you must include the -P flag to have
1633 cvs prune empty directories.
1635 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1636 "?=" instead of the "=" assignment operator, so that buildworld can
1637 override the CPUTYPE if it needs to.
1639 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1640 not on the command line, or in /etc/make.conf. buildworld will
1641 warn if it is improperly defined.
1644 This file contains a list, in reverse chronological order, of major
1645 breakages in tracking -current. It is not guaranteed to be a complete
1646 list of such breakages, and only contains entries since October 10, 2007.
1647 If you need to see UPDATING entries from before that date, you will need
1648 to fetch an UPDATING file from an older FreeBSD release.
1650 Copyright information:
1652 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1654 Redistribution, publication, translation and use, with or without
1655 modification, in full or in part, in any form or format of this
1656 document are permitted without further permission from the author.
1658 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1659 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1660 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1661 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1662 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1663 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1664 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1665 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1666 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1667 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1668 POSSIBILITY OF SUCH DAMAGE.
1670 Contact Warner Losh if you have any questions about your use of