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 20140624: p9 FreeBSD-SA-14:16.file
15 FreeBSD-EN-14:08.heimdal
17 Fix multiple vulnerabilities in file(1) and libmagic(3).
20 Fix gss_pseudo_random interoperability issue. [EN-14:08]
22 20140605: p8 FreeBSD-SA-14:14.openssl
23 Fix OpenSSL multiple vulnerabilities. [SA-14:14]
25 20140603: p7 FreeBSD-SA-14:11.sendmail
26 FreeBSD-SA-14:12.ktrace
30 Fix sendmail improper close-on-exec flag handling. [SA-14:11]
32 Fix ktrace memory disclosure. [SA-14:12]
34 Fix incorrect error handling in PAM policy parser. [SA-14:13]
36 Fix triple-fault when executing from a threaded process.
39 20140513: p6 FreeBSD-EN-14:03.pkg
40 FreeBSD-EN-14:04.kldxref
43 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
45 Improve build repeatability for kldxref(8). [EN-14:04]
47 Fix data corruption with ciss(4). [EN-14:05]
49 20140430: p5 FreeBSD-SA-14:08.tcp
51 Fix TCP reassembly vulnerability. [SA-14:08]
53 20140408: p4 FreeBSD-SA-14:05.nfsserver
54 FreeBSD-SA-14:06.openssl
55 Fix deadlock in the NFS server. [SA-14:05]
57 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
59 20140114: p3 FreeBSD-SA-14:01.bsnmpd
62 FreeBSD-EN-14:01.random
64 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
66 Fix ntpd distributed reflection Denial of Service
67 vulnerability. [SA-14:02]
69 Fix BIND remote denial of service vulnerability. [SA-14:04]
71 Disable hardware RNGs by default. [EN-14:01]
73 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
75 20131128: p2 FreeBSD-EN-13:05.freebsd-update
76 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
78 20131026: p1 FreeBSD-EN-13:04.freebsd-update
79 Fix multiple freebsd-update bugs that break upgrading to
83 hastctl(8)'s `status' command output changed to terse one-liner format.
84 Scripts using this should switch to `list' command or be rewritten.
87 Fix a bug that allowed a tracing process (e.g. gdb) to write
88 to a memory-mapped file in the traced process's address space
89 even if neither the traced process nor the tracing process had
90 write access to that file.
93 Added ZFS TRIM support which is enabled by default. To disable
94 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
96 Creating new ZFS pools and adding new devices to existing pools
97 first performs a full device level TRIM which can take a significant
98 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
99 to disable this behaviour.
101 ZFS TRIM requires the underlying device support BIO_DELETE which
102 is currently provided by methods such as ATA TRIM and SCSI UNMAP
103 via CAM, which are typically supported by SSD's.
105 Stats for ZFS TRIM can be monitored by looking at the sysctl's
106 under kstat.zfs.misc.zio_trim.
109 `list' command has been added to hastctl(8). For now, it is full
110 equivalent of `status' command.
111 WARNING: in the near future the output of hastctl's status command
112 will change to more terse format. If you use `hastctl status'
113 for parsing in your scripts, switch to `hastctl list'.
116 The mergemaster command now uses the default MAKEOBJDIRPREFIX
117 rather than creating it's own in the temporary directory in
118 order allow access to bootstrapped versions of tools such as
119 install and mtree. When upgrading from version of FreeBSD where
120 the install command does not support -l, you will need to
121 install a new mergemaster command if mergemaster -p is required.
122 This can be accomplished with the command (cd src/usr.sbin/mergemaster
125 Due to the use of the new -l option to install(1) during build
126 and install, you must take care not to directly set the INSTALL
127 make variable in your /etc/make.conf, /etc/src.conf, or on the
128 command line. If you wish to use the -C flag for all installs
129 you may be able to add INSTALL+=-C to /etc/make.conf or
133 Fix a bug that allows NFS clients to issue READDIR on files.
136 The install(1) option -M has changed meaning and now takes an
137 argument that is a file or path to append logs to. In the
138 unlikely event that -M was the last option on the command line
139 and the command line contained at least two files and a target
140 directory the first file will have logs appended to it. The -M
141 option served little practical purpose in the last decade so it's
142 used expected to be extremely rare.
145 A new compression method (lz4) has been merged to. Please refer to
146 zpool-features(7) for more information.
148 Please refer to the "ZFS notes" section of this file for information
149 on upgrading boot ZFS pools.
152 The VFS KBI was changed with the merge of several nullfs
153 optimizations and fixes. All filesystem modules must be
157 With the addition of auditdistd(8), a new auditdistd user is now
158 depended on during installworld. "mergemaster -p" can be used to add
159 the user prior to installworld, as documented in the handbook.
165 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
166 Starting with this version the old system of ZFS pool versioning
167 is superseded by "feature flags". This concept enables forward
168 compatibility against certain future changes in functionality of ZFS
169 pools. The first two read-only compatible "feature flags" for ZFS
170 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
171 For more information read the new zpool-features(7) manual page.
172 Please refer to the "ZFS notes" section of this file for information
173 on upgrading boot ZFS pools.
176 The commit introducing bsd.compiler.mk breaks the traditional
177 building of kernels before this point. Add -m ${SRC}/share/mk
178 (for the right value of SRC) to your command lines to work
179 around; update your useland to a point after this; or use the
180 buildkernel/installkernel top-level targets. See also 20120829.
183 The IPFIREWALL_FORWARD kernel option has been removed. Its
184 functionality now turned on by default.
187 The random(4) support for the VIA hardware random number
188 generator (`PADLOCK') is no longer enabled unconditionally.
189 Add the PADLOCK_RNG option in the custom kernel config if
190 needed. The GENERIC kernels on i386 and amd64 do include the
191 option, so the change only affects the custom kernel
195 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
196 the traditional method, you must update your system with an installworld
197 before the kernel will build. The documented make buildkernel/installkernel
198 interfaces (coupled with fresh make kernel-toolchain) continue to work.
201 The sparc64 ZFS loader has been changed to no longer try to auto-
202 detect ZFS providers based on diskN aliases but now requires these
203 to be explicitly listed in the OFW boot-device environment variable.
206 Now unix domain sockets behave "as expected" on nullfs(5). Previously
207 nullfs(5) did not pass through all behaviours to the underlying layer,
208 as a result if we bound to a socket on the lower layer we could connect
209 only to the lower path; if we bound to the upper layer we could connect
210 only to the upper path. The new behavior is one can connect to both the
211 lower and the upper paths regardless what layer path one binds to.
214 The acpi_wmi(4) status device /dev/wmistat has been renamed to
218 A new VOP_ADVISE() was added to support posix_fadvise(2). All
219 filesystem modules must be recompiled.
222 The interface of the VOP_VPTOCNP(9) changed, now the returned
223 vnode shall be referenced, previously it was required to be
224 only held. All in-tree filesystems are converted.
230 The broken amd(4) driver has been replaced with esp(4) in the amd64,
231 i386 and pc98 GENERIC kernel configuration files.
234 This commit modifies vfs_register() so that it uses a hash
235 calculation to set vfc_typenum, which is enabled by default.
236 The first time a system is booted after this change, the
237 vfc_typenum values will change for all file systems. The
238 main effect of this is a change to the NFS server file handles
239 for file systems that use vfc_typenum in their fsid, such as ZFS.
240 It will, however, prevent vfc_typenum from changing when file
241 systems are loaded in a different order for subsequent reboots.
242 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
243 until you are ready to remount all NFS clients after a reboot.
246 Bump the shared library version numbers for libraries that
247 do not use symbol versioning, have changed the ABI compared
248 to stable/8 and which shared library version was not bumped.
249 Done as part of 9.0-RELEASE cycle.
252 During the merge of Capsicum features, the fget(9) KPI was modified.
253 This may require the rebuilding of out-of-tree device drivers --
254 issues have been reported specifically with the nVidia device driver.
255 __FreeBSD_version is bumped to 900041.
257 Also, there is a period between 20110811 and 20110814 where the
258 special devices /dev/{stdin,stdout,stderr} did not work correctly.
259 Building world from a kernel during that window may not work.
262 The packet filter (pf) code has been updated to OpenBSD 4.5.
263 You need to update userland tools to be in sync with kernel.
264 This update breaks backward compatibility with earlier pfsync(4)
265 versions. Care must be taken when updating redundant firewall setups.
268 The following sysctls and tunables are retired on x86 platforms:
270 machdep.hlt_logical_cpus
271 The following sysctl is retired:
272 machdep.hyperthreading_allowed
273 The sysctls were supposed to provide a way to dynamically offline and
274 online selected CPUs on x86 platforms, but the implementation has not
275 been reliable especially with SCHED_ULE scheduler.
276 machdep.hyperthreading_allowed tunable is still available to ignore
277 hyperthreading CPUs at OS level.
278 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
279 where X is an APIC ID of a CPU. Be advised, though, that disabling
280 CPUs in non-uniform fashion will result in non-uniform topology and
281 may lead to sub-optimal system performance with SCHED_ULE, which is
285 cpumask_t type is retired and cpuset_t is used in order to describe
289 Changes to ifconfig(8) for dynamic address family detection mandate
290 that you are running a kernel of 20110525 or later. Make sure to
291 follow the update procedure to boot a new kernel before installing
295 Support for sun4v architecture is officially dropped
298 Several KPI breaking changes have been committed to the mii(4) layer,
299 the PHY drivers and consequently some Ethernet drivers using mii(4).
300 This means that miibus.ko and the modules of the affected Ethernet
301 drivers need to be recompiled.
303 Note to kernel developers: Given that the OUI bit reversion problem
304 was fixed as part of these changes all mii(4) commits related to OUIs,
305 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
306 handling, no longer can be merged verbatim to stable/8 and previous
310 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
311 into their kernel configurations along with 'device pci'.
314 The default NFS client is now the new NFS client, so fstype "newnfs"
315 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
316 Although mounts via fstype "nfs" will usually work without userland
317 changes, it is recommended that the mount(8) and mount_nfs(8)
318 commands be rebuilt from sources and that a link to mount_nfs called
319 mount_oldnfs be created. The new client is compiled into the
320 kernel with "options NFSCL" and this is needed for diskless root
321 file systems. The GENERIC kernel configs have been changed to use
322 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
323 To use the regular/old client, you can "mount -t oldnfs ...". For
324 a diskless root file system, you must also include a line like:
326 vfs.root.mountfrom="oldnfs:"
328 in the boot/loader.conf on the root fs on the NFS server to make
329 a diskless root fs use the old client.
332 The GENERIC kernels for all architectures now default to the new
333 CAM-based ATA stack. It means that all legacy ATA drivers were
334 removed and replaced by respective CAM drivers. If you are using
335 ATA device names in /etc/fstab or other places, make sure to update
336 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
337 where 'Y's are the sequential numbers starting from zero for each type
338 in order of detection, unless configured otherwise with tunables,
339 see cam(4)). There will be symbolic links created in /dev/ to map
340 old adX devices to the respective adaY. They should provide basic
341 compatibility for file systems mounting in most cases, but they do
342 not support old user-level APIs and do not have respective providers
343 in GEOM. Consider using updated management tools with new device names.
345 It is possible to load devices ahci, ata, siis and mvs as modules,
346 but option ATA_CAM should remain in kernel configuration to make ata
347 module work as CAM driver supporting legacy ATA controllers. Device ata
348 still can be used in modular fashion (atacore + ...). Modules atadisk
349 and atapi* are not used and won't affect operation in ATA_CAM mode.
350 Note that to use CAM-based ATA kernel should include CAM devices
351 scbus, pass, da (or explicitly ada), cd and optionally others. All of
352 them are parts of the cam module.
354 ataraid(4) functionality is now supported by the RAID GEOM class.
355 To use it you can load geom_raid kernel module and use graid(8) tool
356 for management. Instead of /dev/arX device names, use /dev/raid/rX.
358 No kernel config options or code have been removed, so if a problem
359 arises, please report it and optionally revert to the old ATA stack.
360 In order to do it you can remove from the kernel config:
365 , and instead add back:
366 device atadisk # ATA disk drives
367 device ataraid # ATA RAID drives
368 device atapicd # ATAPI CDROM drives
369 device atapifd # ATAPI floppy drives
370 device atapist # ATAPI tape drives
373 The default NFS server has been changed to the new server, which
374 was referred to as the experimental server. If you need to switch
375 back to the old NFS server, you must now put the "-o" option on
376 both the mountd and nfsd commands. This can be done using the
377 mountd_flags and nfs_server_flags rc.conf variables until an
378 update to the rc scripts is committed, which is coming soon.
381 The GNU Objective-C runtime library (libobjc), and other Objective-C
382 related components have been removed from the base system. If you
383 require an Objective-C library, please use one of the available ports.
386 ath(4) has been split into bus- and device- modules. if_ath contains
387 the HAL, the TX rate control and the network device code. if_ath_pci
388 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
389 contains the AHB glue. Users need to load both if_ath_pci and if_ath
390 in order to use ath on everything else.
392 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
393 need to load if_ath and if_ath_pci for ath(4) operation.
396 As part of the replacement of sysinstall, the process of building
397 release media has changed significantly. For details, please re-read
398 release(7), which has been updated to reflect the new build process.
401 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
402 is the last available version under GPLv2. It brings a number of new
403 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
404 4.1 and SSE 4.2), better support for powerpc64, a number of new
405 directives, and lots of other small improvements. See the ChangeLog
406 file in contrib/binutils for the full details.
409 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
410 compliant, and will now use half of hash for authentication.
411 This will break interoperability with all stacks (including all
412 actual FreeBSD versions) who implement
413 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
415 The only workaround with such peers is to use another HMAC
416 algorithm for IPsec ("phase 2") authentication.
419 Remove the uio_yield prototype and symbol. This function has
420 been misnamed since it was introduced and should not be
421 globally exposed with this name. The equivalent functionality
422 is now available using kern_yield(curthread->td_user_pri).
423 The function remains undocumented.
426 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
427 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
428 for scalar sysctls is defined but disabled. Code that needs
429 UQUAD to pass the type checking that must compile on older
430 systems where the define is not present can check against
431 __FreeBSD_version >= 900030.
433 The system dialog(1) has been replaced with a new version previously
434 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
435 with the previous version, but the libdialog associated with it has
436 a largely incompatible API. As such, the original version of libdialog
437 will be kept temporarily as libodialog, until its base system consumers
438 are replaced or updated. Bump __FreeBSD_version to 900030.
441 If you are trying to run make universe on a -stable system, and you get
442 the following warning:
443 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
444 unknown. config(8) likely too old."
445 or something similar to it, then you must upgrade your -stable system
446 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
447 stable/8) or build the config from the latest stable/8 branch and
448 install it on your system.
450 Prior to this date, building a current universe on 8-stable system from
451 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
452 error in the first kernel build phase. A new config on those old
453 systems will fix that problem for older versions of -current.
456 The TCP stack has been modified to allow Khelp modules to interact with
457 it via helper hook points and store per-connection data in the TCP
458 control block. Bump __FreeBSD_version to 900029. User space tools that
459 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
463 Generic IEEE 802.3 annex 31B full duplex flow control support has been
464 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
465 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
466 to take advantage of it instead of using custom implementations. This
467 means that these drivers now no longer unconditionally advertise
468 support for flow control but only do so if flow control is a selected
469 media option. This was implemented in the generic support that way in
470 order to allow flow control to be switched on and off via ifconfig(8)
471 with the PHY specific default to typically off in order to protect
472 from unwanted effects. Consequently, if you used flow control with
473 one of the above mentioned drivers you now need to explicitly enable
475 ifconfig bge0 media auto mediaopt flowcontrol
477 Along with the above mentioned changes generic support for setting
478 1000baseT master mode also has been added and brgphy(4), ciphy(4),
479 e1000phy(4) as well as ip1000phy(4) have been converted to take
480 advantage of it. This means that these drivers now no longer take the
481 link0 parameter for selecting master mode but the master media option
482 has to be used instead, for example like in the following:
483 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
485 Selection of master mode now is also available with all other PHY
486 drivers supporting 1000baseT.
489 The TCP stack has received a significant update to add support for
490 modularised congestion control and generally improve the clarity of
491 congestion control decisions. Bump __FreeBSD_version to 900025. User
492 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
493 sockstat) need to be recompiled.
496 The man(1) utility has been replaced by a new version that no longer
497 uses /etc/manpath.config. Please consult man.conf(5) for how to
498 migrate local entries to the new format.
501 The copyright strings printed by login(1) and sshd(8) at the time of a
502 new connection have been removed to follow other operating systems and
506 A workaround for a fixed ld bug has been removed in kernel code,
507 so make sure that your system ld is built from sources after
508 revision 210245 from 2010-07-19 (r211583 if building head kernel
509 on stable/8, r211584 for stable/7; both from 2010-08-21).
510 A symptom of incorrect ld version is different addresses for
511 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
514 The $ipv6_prefer variable in rc.conf(5) has been split into
515 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
517 The $ip6addrctl_policy is a variable to choose a pre-defined
518 address selection policy set by ip6addrctl(8). A value
519 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
522 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
523 flag (see an entry of 20090926) is set on an interface with no
524 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
525 security reason. If you want IPv6 link-local address on all
526 interfaces by default, set this to "YES".
528 The old ipv6_prefer="YES" is equivalent to
529 ipv6_activate_all_interfaces="YES" and
530 ip6addrctl_policy="ipv6_prefer".
533 DTrace has grown support for userland tracing. Due to this, DTrace is
534 now i386 and amd64 only.
535 dtruss(1) is now installed by default on those systems and a new
536 kernel module is needed for userland tracing: fasttrap.
537 No changes to your kernel config file are necessary to enable
538 userland tracing, but you might consider adding 'STRIP=' and
539 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
540 to have informative userland stack traces in DTrace (ustack).
543 The acpi_aiboost(4) driver has been removed in favor of the new
544 aibs(4) driver. You should update your kernel configuration file.
547 BSD grep has been imported to the base system and it is built by
548 default. It is completely BSD licensed, highly GNU-compatible, uses
549 less memory than its GNU counterpart and has a small codebase.
550 However, it is slower than its GNU counterpart, which is mostly
551 noticeable for larger searches, for smaller ones it is measurable
552 but not significant. The reason is complex, the most important factor
553 is that we lack a modern and efficient regex library and GNU
554 overcomes this by optimizing the searches internally. Future work
555 on improving the regex performance is planned, for the meantime,
556 users that need better performance, can build GNU grep instead by
557 setting the WITH_GNU_GREP knob.
560 Due to the import of powerpc64 support, all existing powerpc kernel
561 configuration files must be updated with a machine directive like this:
562 machine powerpc powerpc
564 In addition, an updated config(8) is required to build powerpc kernels
568 A new version of ZFS (version 15) has been merged to -HEAD.
569 This version uses a python library for the following subcommands:
570 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
571 For full functionality of these commands the following port must
572 be installed: sysutils/py-zfs
575 'vm_page's are now hashed by physical address to an array of mutexes.
576 Currently this is only used to serialize access to hold_count. Over
577 time the page queue mutex will be peeled away. This changes the size
578 of pmap on every architecture. And requires all callers of vm_page_hold
579 and vm_page_unhold to be updated.
582 WITH_CTF can now be specified in src.conf (not recommended, there
583 are some problems with static executables), make.conf (would also
584 affect ports which do not use GNU make and do not override the
585 compile targets) or in the kernel config (via "makeoptions
587 When WITH_CTF was specified there before this was silently ignored,
588 so make sure that WITH_CTF is not used in places which could lead
589 to unwanted behavior.
592 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
593 to allow 32-bit compatibility on non-x86 platforms. All kernel
594 configurations on amd64 and ia64 platforms using these options must
595 be modified accordingly.
598 The utmp user accounting database has been replaced with utmpx,
599 the user accounting interface standardized by POSIX.
600 Unfortunately the semantics of utmp and utmpx don't match,
601 making it practically impossible to support both interfaces.
602 The user accounting database is used by tools like finger(1),
603 last(1), talk(1), w(1) and ac(8).
605 All applications in the base system use utmpx. This means only
606 local binaries (e.g. from the ports tree) may still use these
607 utmp database files. These applications must be rebuilt to make
610 After the system has been upgraded, it is safe to remove the old
611 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
612 assuming their contents is of no importance anymore. Old wtmp
613 databases can only be used by last(1) and ac(8) after they have
614 been converted to the new format using wtmpcvt(1).
617 Introduce the kernel thread "deadlock resolver" (which can be enabled
618 via the DEADLKRES option, see NOTES for more details) and the
619 sleepq_type() function for sleepqueues.
622 The rc.firewall and rc.firewall6 were unified, and
623 rc.firewall6 and rc.d/ip6fw were removed.
624 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
625 variables are obsoleted. Instead, the following new rc
626 variables are added to rc.d/ipfw:
628 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
629 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
630 firewall_simple_onet_ipv6, firewall_trusted_ipv6
632 The meanings correspond to the relevant IPv4 variables.
638 The default terminal emulation for syscons(4) has been changed
639 from cons25 to xterm on all platforms except pc98. This means
640 that the /etc/ttys file needs to be updated to ensure correct
641 operation of applications on the console.
643 The terminal emulation style can be toggled per window by using
644 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
645 options can be used to change the compile-time default back to
648 To prevent graphical artifacts, make sure the TERM environment
649 variable is set to match the terminal emulation that is being
650 performed by syscons(4).
653 The layout of the structure ieee80211req_scan_result has changed.
654 Applications that require wireless scan results (e.g. ifconfig(8))
655 from net80211 need to be recompiled.
657 Applications such as wpa_supplicant(8) may require a full world
658 build without using NO_CLEAN in order to get synchronized with the
662 The iwn(4) driver has been updated to support the 5000 and 5150 series.
663 There's one kernel module for each firmware. Adding "device iwnfw"
664 to the kernel configuration file means including all three firmware
665 images inside the kernel. If you want to include just the one for
666 your wireless card, use the devices iwn4965fw, iwn5000fw or
670 The rc.d/network_ipv6, IPv6 configuration script has been integrated
671 into rc.d/netif. The changes are the following:
673 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
674 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
675 Note that both variables need the "inet6" keyword at the head.
677 Do not set $ipv6_network_interfaces manually if you do not
678 understand what you are doing. It is not needed in most cases.
680 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
683 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
684 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
686 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
687 all configured interfaces have "inet6 accept_rtadv" in the
688 $ifconfig_IF_ipv6. These are for backward compatibility.
690 3. A new variable $ipv6_prefer has been added. If NO, IPv6
691 functionality of interfaces with no corresponding
692 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
693 and the default address selection policy of ip6addrctl(8)
694 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
695 Note that if you want to configure IPv6 functionality on the
696 disabled interfaces after boot, first you need to clear the flag by
697 using ifconfig(8) like:
699 ifconfig em0 inet6 -ifdisabled
701 If YES, the default address selection policy is set as
704 The default value of $ipv6_prefer is NO.
706 4. If your system need to receive Router Advertisement messages,
707 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
708 scripts automatically invoke rtsol(8) when the interface becomes
709 UP. The Router Advertisement messages are used for SLAAC
710 (State-Less Address AutoConfiguration).
713 802.11s D3.03 support was committed. This is incompatible with the
714 previous code, which was based on D3.0.
717 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
718 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
719 control whether accepting Router Advertisement messages or not.
720 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
721 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
722 The ifconfig(8) utility now supports these flags.
725 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
726 mount(8) and -a option for df(1) to see them.
729 The old tunable hw.bus.devctl_disable has been superseded by
730 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
731 replaced by hw.bus.devctl_queue=0. The default for this new tunable
735 Remove the option STOP_NMI. The default action is now to use NMI only
736 for KDB via the newly introduced function stop_cpus_hard() and
737 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
740 The stable/8 branch created in subversion. This corresponds to the
741 RELENG_8 branch in CVS.
744 Bump the shared library version numbers for all libraries that do not
745 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
746 __FreeBSD_version to 800105.
749 Due to changes in the implementation of virtual network stack support,
750 all network-related kernel modules must be recompiled. As this change
751 breaks the ABI, bump __FreeBSD_version to 800104.
754 The TOE interface to the TCP syncache has been modified to remove
755 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
756 The cxgb driver is the only TOE consumer affected by this change, and
757 needs to be recompiled along with the kernel. As this change breaks
758 the ABI, bump __FreeBSD_version to 800103.
761 Padding has been added to struct tcpcb, sackhint and tcpstat in
762 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
763 maintaining the ABI. However, this change breaks the ABI, so bump
764 __FreeBSD_version to 800102. User space tools that rely on the size of
765 any of these structs (e.g. sockstat) need to be recompiled.
768 The NFS_LEGACYRPC option has been removed along with the old kernel
769 RPC implementation that this option selected. Kernel configurations
770 may need to be adjusted.
773 The network interface device nodes at /dev/net/<interface> have been
774 removed. All ioctl operations can be performed the normal way using
775 routing sockets. The kqueue functionality can generally be replaced
776 with routing sockets.
779 The documentation from the FreeBSD Documentation Project (Handbook,
780 FAQ, etc.) is now installed via packages by sysinstall(8) and under
781 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
784 The ABI of various structures related to the SYSV IPC API have been
785 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
786 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
790 Layout of struct vnet has changed as routing related variables were
791 moved to their own Vimage module. Modules need to be recompiled. Bump
792 __FreeBSD_version to 800099.
795 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
796 respectively. As long as no more than 16 groups per process are used,
797 no changes should be visible. When more than 16 groups are used, old
798 binaries may fail if they call getgroups() or getgrouplist() with
799 statically sized storage. Recompiling will work around this, but
800 applications should be modified to use dynamically allocated storage
801 for group arrays as POSIX.1-2008 does not cap an implementation's
802 number of supported groups at NGROUPS_MAX+1 as previous versions did.
804 NFS and portalfs mounts may also be affected as the list of groups is
805 truncated to 16. Users of NFS who use more than 16 groups, should
806 take care that negative group permissions are not used on the exported
807 file systems as they will not be reliable unless a GSSAPI based
808 authentication method is used.
811 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
812 option compiles in the support for adaptive spinning for lockmgrs
813 which want to enable it. The lockinit() function now accepts the flag
814 LK_ADAPTIVE in order to make the lock object subject to adaptive
815 spinning when both held in write and read mode.
818 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
819 changed. User applications that use this ioctl need to be rebuilt.
822 The layout of struct thread has changed. Kernel and modules need to
826 The layout of structs ifnet, domain, protosw and vnet_net has changed.
827 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
830 window(1) has been removed from the base system. It can now be
831 installed from ports. The port is called misc/window.
834 The way we are storing and accessing `routing table' entries has
835 changed. Programs reading the FIB, like netstat, need to be
839 A new netisr implementation has been added for FreeBSD 8. Network
840 file system modules, such as igmp, ipdivert, and others, should be
842 Bump __FreeBSD_version to 800096.
845 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
849 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
850 Bump __FreeBSD_version to 800094.
853 Add mnt_xflag field to 'struct mount'. File system modules need to be
855 Bump __FreeBSD_version to 800093.
858 The compiling option ADAPTIVE_SX has been retired while it has been
859 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
860 The KPI for sx_init_flags() changes as accepting flags:
861 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
862 been introduced in order to handle the reversed logic.
863 Bump __FreeBSD_version to 800092.
866 Add support for hierarchical jails. Remove global securelevel.
867 Bump __FreeBSD_version to 800091.
870 The layout of struct vnet_net has changed, therefore modules
872 Bump __FreeBSD_version to 800090.
875 The newly imported zic(8) produces a new format in the output. Please
876 run tzsetup(8) to install the newly created data to /etc/localtime.
879 The sysctl tree for the usb stack has renamed from hw.usb2.* to
880 hw.usb.* and is now consistent again with previous releases.
883 802.11 monitor mode support was revised and driver api's were changed.
884 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
885 of DLT_IEEE802_11. No user-visible data structures were changed but
886 applications that use DLT_IEEE802_11 may require changes.
887 Bump __FreeBSD_version to 800088.
890 The layout of the following structs has changed: sysctl_oid,
891 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
892 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
893 panics may be experienced. World rebuild is required for
894 correctly checking networking state from userland.
895 Bump __FreeBSD_version to 800085.
898 MLDv2 and Source-Specific Multicast (SSM) have been merged
899 to the IPv6 stack. VIMAGE hooks are in but not yet used.
900 The implementation of SSM within FreeBSD's IPv6 stack closely
901 follows the IPv4 implementation.
903 For kernel developers:
905 * The most important changes are that the ip6_output() and
906 ip6_input() paths no longer take the IN6_MULTI_LOCK,
907 and this lock has been downgraded to a non-recursive mutex.
909 * As with the changes to the IPv4 stack to support SSM, filtering
910 of inbound multicast traffic must now be performed by transport
911 protocols within the IPv6 stack. This does not apply to TCP and
912 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
914 * The KPIs used by IPv6 multicast are similar to those used by
915 the IPv4 stack, with the following differences:
916 * im6o_mc_filter() is analogous to imo_multicast_filter().
917 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
918 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
919 * IN6_LOOKUP_MULTI() has been deprecated and removed.
920 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
921 for MLDv1 have an additional 'timer' argument which is used to
922 jitter the initial membership report for the solicited-node
923 multicast membership on-link.
924 * This is not strictly needed for MLDv2, which already jitters
925 its report transmissions. However, the 'timer' argument is
926 preserved in case MLDv1 is active on the interface.
928 * The KAME linked-list based IPv6 membership implementation has
929 been refactored to use a vector similar to that used by the IPv4
931 Code which maintains a list of its own multicast memberships
932 internally, e.g. carp, has been updated to reflect the new
935 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
936 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
937 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
938 implementation constraint which needs to be addressed in HEAD.
940 For application developers:
942 * The changes are broadly similar to those made for the IPv4
945 * The use of IPv4 and IPv6 multicast socket options on the same
946 socket, using mapped addresses, HAS NOT been tested or supported.
948 * There are a number of issues with the implementation of various
949 IPv6 multicast APIs which need to be resolved in the API surface
950 before the implementation is fully compatible with KAME userland
951 use, and these are mostly to do with interface index treatment.
953 * The literature available discusses the use of either the delta / ASM
954 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
955 using setsourcefilter(3)/getsourcefilter(3). For more information
956 please refer to RFC 3768, 'Socket Interface Extensions for
957 Multicast Source Filters'.
959 * Applications which use the published RFC 3678 APIs should be fine.
961 For systems administrators:
963 * The mtest(8) utility has been refactored to support IPv6, in
964 addition to IPv4. Interface addresses are no longer accepted
965 as arguments, their names must be used instead. The utility
966 will map the interface name to its first IPv4 address as
967 returned by getifaddrs(3).
969 * The ifmcstat(8) utility has also been updated to print the MLDv2
970 endpoint state and source filter lists via sysctl(3).
972 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
973 loopback of IPv6 multicast datagrams by default; it defaults to 1
974 to preserve the existing behaviour. Disabling multicast loopback is
975 recommended for optimal system performance.
977 * The IPv6 MROUTING code has been changed to examine this sysctl
978 instead of attempting to perform a group lookup before looping
979 back forwarded datagrams.
981 Bump __FreeBSD_version to 800084.
984 Implement low-level Bluetooth HCI API.
985 Bump __FreeBSD_version to 800083.
988 The layout of struct malloc_type, used by modules to register new
989 memory allocation types, has changed. Most modules will need to
990 be rebuilt or panics may be experienced.
991 Bump __FreeBSD_version to 800081.
994 Anticipate overflowing inp_flags - add inp_flags2.
995 This changes most offsets in inpcb, so checking v4 connection
996 state will require a world rebuild.
997 Bump __FreeBSD_version to 800080.
1000 Add an llentry to struct route and struct route_in6. Modules
1001 embedding a struct route will need to be recompiled.
1002 Bump __FreeBSD_version to 800079.
1005 The size of rt_metrics_lite and by extension rtentry has changed.
1006 Networking administration apps will need to be recompiled.
1007 The route command now supports show as an alias for get, weighting
1008 of routes, sticky and nostick flags to alter the behavior of stateful
1010 Bump __FreeBSD_version to 800078.
1013 Do not use Giant for kbdmux(4) locking. This is wrong and
1014 apparently causing more problems than it solves. This will
1015 re-open the issue where interrupt handlers may race with
1016 kbdmux(4) in polling mode. Typical symptoms include (but
1017 not limited to) duplicated and/or missing characters when
1018 low level console functions (such as gets) are used while
1019 interrupts are enabled (for example geli password prompt,
1020 mountroot prompt etc.). Disabling kbdmux(4) may help.
1023 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
1024 kernel modules referencing any of the above need to be recompiled.
1025 Bump __FreeBSD_version to 800075.
1028 GEOM_PART has become the default partition slicer for storage devices,
1029 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
1030 introduces some changes:
1032 MSDOS/EBR: the devices created from MSDOS extended partition entries
1033 (EBR) can be named differently than with GEOM_MBR and are now symlinks
1034 to devices with offset-based names. fstabs may need to be modified.
1036 BSD: the "geometry does not match label" warning is harmless in most
1037 cases but it points to problems in file system misalignment with
1038 disk geometry. The "c" partition is now implicit, covers the whole
1039 top-level drive and cannot be (mis)used by users.
1041 General: Kernel dumps are now not allowed to be written to devices
1042 whose partition types indicate they are meant to be used for file
1043 systems (or, in case of MSDOS partitions, as something else than
1046 Most of these changes date approximately from 200812.
1049 The uscanner(4) driver has been removed from the kernel. This follows
1050 Linux removing theirs in 2.6 and making libusb the default interface
1051 (supported by sane).
1054 The multicast forwarding code has been cleaned up. netstat(1)
1055 only relies on KVM now for printing bandwidth upcall meters.
1056 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1057 ip6_mroute_mod respectively. The config(5) options for statically
1058 compiling this code remain the same, i.e. 'options MROUTING'.
1061 Support for the IFF_NEEDSGIANT network interface flag has been
1062 removed, which means that non-MPSAFE network device drivers are no
1063 longer supported. In particular, if_ar, if_sr, and network device
1064 drivers from the old (legacy) USB stack can no longer be built or
1068 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1069 a bunch of new language catalog files have also been added.
1070 This means that some common libc messages are now localized and
1071 they depend on the LC_MESSAGES environmental variable.
1074 The k8temp(4) driver has been renamed to amdtemp(4) since
1075 support for Family 10 and Family 11 CPU families was added.
1078 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1079 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1081 For kernel developers, the most important changes are that the
1082 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1083 and this lock has been downgraded to a non-recursive mutex.
1085 Transport protocols (UDP, Raw IP) are now responsible for filtering
1086 inbound multicast traffic according to group membership and source
1087 filters. The imo_multicast_filter() KPI exists for this purpose.
1088 Transports which do not use multicast (SCTP, TCP) already reject
1089 multicast by default. Forwarding and receive performance may improve
1090 as a mutex acquisition is no longer needed in the ip_input()
1091 low-level input path. in_addmulti() and in_delmulti() are shimmed
1092 to new KPIs which exist to support SSM in-kernel.
1094 For application developers, it is recommended that loopback of
1095 multicast datagrams be disabled for best performance, as this
1096 will still cause the lock to be taken for each looped-back
1097 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1098 be tuned to 0 to disable loopback by default; it defaults to 1
1099 to preserve the existing behaviour.
1101 For systems administrators, to obtain best performance with
1102 multicast reception and multiple groups, it is always recommended
1103 that a card with a suitably precise hash filter is used. Hash
1104 collisions will still result in the lock being taken within the
1105 transport protocol input path to check group membership.
1107 If deploying FreeBSD in an environment with IGMP snooping switches,
1108 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1109 enabled; this forces 224.0.0.0/24 group membership to be announced
1112 The size of 'struct igmpstat' has changed; netstat needs to be
1113 recompiled to reflect this.
1114 Bump __FreeBSD_version to 800070.
1117 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1118 updated to use it. This requires a buildworld/installworld in order to
1119 update the library and dependencies (usbconfig, etc). Its advisable to
1120 rebuild all ports which uses libusb. More specific directions are given
1121 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1122 libusb are no longer required and can be removed.
1125 A workaround is committed to allow the creation of System V shared
1126 memory segment of size > 2 GB on the 64-bit architectures.
1127 Due to a limitation of the existing ABI, the shm_segsz member
1128 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1129 wrong for large segments. Note that limits must be explicitly
1130 raised to allow such segments to be created.
1133 The layout of struct ifnet has changed, requiring a rebuild of all
1134 network device driver modules.
1137 The /dev handling for the new USB stack has changed, a
1138 buildworld/installworld is required for libusb20.
1141 The new USB2 stack has now been permanently moved in and all kernel and
1142 module names reverted to their previous values (eg, usb, ehci, ohci,
1143 ums, ...). The old usb stack can be compiled in by prefixing the name
1144 with the letter 'o', the old usb modules have been removed.
1145 Updating entry 20090216 for xorg and 20090215 for libmap may still
1149 The rc.conf(5) option if_up_delay has been renamed to
1150 defaultroute_delay to better reflect its purpose. If you have
1151 customized this setting in /etc/rc.conf you need to update it to
1155 xorg 7.4 wants to configure its input devices via hald which does not
1156 yet work with USB2. If the keyboard/mouse does not work in xorg then
1158 Option "AllowEmptyInput" "off"
1159 to your ServerLayout section. This will cause X to use the configured
1160 kbd and mouse sections from your xorg.conf.
1163 The GENERIC kernels for all architectures now default to the new USB2
1164 stack. No kernel config options or code have been removed so if a
1165 problem arises please report it and optionally revert to the old USB
1166 stack. If you are loading USB kernel modules or have a custom kernel
1167 that includes GENERIC then ensure that usb names are also changed over,
1168 eg uftdi -> usb2_serial_ftdi.
1170 Older programs linked against the ports libusb 0.1 need to be
1171 redirected to the new stack's libusb20. /etc/libmap.conf can
1173 # Map old usb library to new one for usb2 stack
1174 libusb-0.1.so.8 libusb20.so.1
1177 All USB ethernet devices now attach as interfaces under the name ueN
1178 (eg. ue0). This is to provide a predictable name as vendors often
1179 change usb chipsets in a product without notice.
1182 The ichsmb(4) driver has been changed to require SMBus slave
1183 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1184 All of the other SMBus controller drivers require left-justified
1185 slave addresses, so this change makes all the drivers provide the
1189 INET6 statistics (struct ip6stat) was updated.
1190 netstat(1) needs to be recompiled.
1193 NTFS has been removed from GENERIC kernel on amd64 to match
1194 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1195 will load ntfs.ko module automatically when NTFS support is
1196 actually needed, unless ntfs.ko is not installed or security
1197 level prohibits loading kernel modules. If either is the case,
1198 "options NTFS" has to be added into kernel config.
1201 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1202 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1203 800061. User space tools that rely on the size of struct tcpcb in
1204 tcp_var.h (e.g. sockstat) need to be recompiled.
1207 ng_tty(4) module updated to match the new TTY subsystem.
1208 Due to API change, user-level applications must be updated.
1209 New API support added to mpd5 CVS and expected to be present
1210 in next mpd5.3 release.
1213 With __FreeBSD_version 800060 the makefs tool is part of
1214 the base system (it was a port).
1217 The afdata and ifnet locks have been changed from mutexes to
1218 rwlocks, network modules will need to be re-compiled.
1221 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1222 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1223 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1224 architecture and 40 bytes on 64-bit architecture. The userland
1225 applications "arp" and "ndp" have been updated accordingly.
1226 The output from "netstat -r" shows only routing entries and
1227 none of the L2 information.
1230 __FreeBSD_version 800057 marks the switchover from the
1231 binary ath hal to source code. Users must add the line:
1233 options AH_SUPPORT_AR5416
1235 to their kernel config files when specifying:
1239 The ath_hal module no longer exists; the code is now compiled
1240 together with the driver in the ath module. It is now
1241 possible to tailor chip support (i.e. reduce the set of chips
1242 and thereby the code size); consult ath_hal(4) for details.
1245 __FreeBSD_version 800054 adds memory barriers to
1246 <machine/atomic.h>, new interfaces to ifnet to facilitate
1247 multiple hardware transmit queues for cards that support
1248 them, and a lock-less ring-buffer implementation to
1249 enable drivers to more efficiently manage queueing of
1253 A new version of ZFS (version 13) has been merged to -HEAD.
1254 This version has zpool attribute "listsnapshots" off by
1255 default, which means "zfs list" does not show snapshots,
1256 and is the same as Solaris behavior.
1259 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1262 The uhci, ohci, ehci and slhci USB Host controller drivers have
1263 been put into separate modules. If you load the usb module
1264 separately through loader.conf you will need to load the
1265 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1266 controller add the following to loader.conf:
1272 The ABI used by the PMC toolset has changed. Please keep
1273 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1277 atapci kernel module now includes only generic PCI ATA
1278 driver. AHCI driver moved to ataahci kernel module.
1279 All vendor-specific code moved into separate kernel modules:
1280 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1281 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1282 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1283 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1286 The TTY subsystem of the kernel has been replaced by a new
1287 implementation, which provides better scalability and an
1288 improved driver model. Most common drivers have been migrated to
1289 the new TTY subsystem, while others have not. The following
1290 drivers have not yet been ported to the new TTY layer:
1293 cy, digi, rc, rp, sio
1299 ng_h4, ng_tty, ppp, sl, snp
1301 Adding these drivers to your kernel configuration file shall
1302 cause compilation to fail.
1305 ntpd has been upgraded to 4.2.4p5.
1308 OpenSSH has been upgraded to 5.1p1.
1310 For many years, FreeBSD's version of OpenSSH preferred DSA
1311 over RSA for host and user authentication keys. With this
1312 upgrade, we've switched to the vendor's default of RSA over
1313 DSA. This may cause upgraded clients to warn about unknown
1314 host keys even for previously known hosts. Users should
1315 follow the usual procedure for verifying host keys before
1316 accepting the RSA key.
1318 This can be circumvented by setting the "HostKeyAlgorithms"
1319 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1322 Please note that the sequence of keys offered for
1323 authentication has been changed as well. You may want to
1324 specify IdentityFile in a different order to revert this
1328 The sio(4) driver has been removed from the i386 and amd64
1329 kernel configuration files. This means uart(4) is now the
1330 default serial port driver on those platforms as well.
1332 To prevent collisions with the sio(4) driver, the uart(4) driver
1333 uses different names for its device nodes. This means the
1334 onboard serial port will now most likely be called "ttyu0"
1335 instead of "ttyd0". You may need to reconfigure applications to
1336 use the new device names.
1338 When using the serial port as a boot console, be sure to update
1339 /boot/device.hints and /etc/ttys before booting the new kernel.
1340 If you forget to do so, you can still manually specify the hints
1341 at the loader prompt:
1343 set hint.uart.0.at="isa"
1344 set hint.uart.0.port="0x3F8"
1345 set hint.uart.0.flags="0x10"
1346 set hint.uart.0.irq="4"
1350 The gpt(8) utility has been removed. Use gpart(8) to partition
1354 The version that Linuxulator emulates was changed from 2.4.2
1355 to 2.6.16. If you experience any problems with Linux binaries
1356 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1357 if it fixes the problem contact emulation mailing list.
1360 ISDN4BSD (I4B) was removed from the src tree. You may need to
1361 update a your kernel configuration and remove relevant entries.
1364 I have checked in code to support multiple routing tables.
1365 See the man pages setfib(1) and setfib(2).
1366 This is a hopefully backwards compatible version,
1367 but to make use of it you need to compile your kernel
1368 with options ROUTETABLES=2 (or more up to 16).
1371 The 802.11 wireless support was redone to enable multi-bss
1372 operation on devices that are capable. The underlying device
1373 is no longer used directly but instead wlanX devices are
1374 cloned with ifconfig. This requires changes to rc.conf files.
1375 For example, change:
1376 ifconfig_ath0="WPA DHCP"
1379 ifconfig_wlan0="WPA DHCP"
1380 see rc.conf(5) for more details. In addition, mergemaster of
1381 /etc/rc.d is highly recommended. Simultaneous update of userland
1382 and kernel wouldn't hurt either.
1384 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1385 modules were merged into the base wlan module. All references
1386 to these modules (e.g. in kernel config files) must be removed.
1389 psm(4) has gained write(2) support in native operation level.
1390 Arbitrary commands can be written to /dev/psm%d and status can
1391 be read back from it. Therefore, an application is responsible
1392 for status validation and error recovery. It is a no-op in
1393 other operation levels.
1396 Support for KSE threading has been removed from the kernel. To
1397 run legacy applications linked against KSE libmap.conf may
1398 be used. The following libmap.conf may be used to ensure
1399 compatibility with any prior release:
1401 libpthread.so.1 libthr.so.1
1402 libpthread.so.2 libthr.so.2
1403 libkse.so.3 libthr.so.3
1406 The layout of struct vmspace has changed. This affects libkvm
1407 and any executables that link against libkvm and use the
1408 kvm_getprocs() function. In particular, but not exclusively,
1409 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1410 The effects are minimal, but it's advisable to upgrade world
1414 The latest em driver no longer has support in it for the
1415 82575 adapter, this is now moved to the igb driver. The
1416 split was done to make new features that are incompatible
1417 with older hardware easier to do.
1420 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1421 likewise the kernel option is now GEOM_LINUX_LVM.
1424 The default NFS mount mode has changed from UDP to TCP for
1425 increased reliability. If you rely on (insecurely) NFS
1426 mounting across a firewall you may need to update your
1430 Belatedly note the addition of m_collapse for compacting
1434 The fts(3) structures have been changed to use adequate
1435 integer types for their members and so to be able to cope
1436 with huge file trees. The old fts(3) ABI is preserved
1437 through symbol versioning in libc, so third-party binaries
1438 using fts(3) should still work, although they will not take
1439 advantage of the extended types. At the same time, some
1440 third-party software might fail to build after this change
1441 due to unportable assumptions made in its source code about
1442 fts(3) structure members. Such software should be fixed
1443 by its vendor or, in the worst case, in the ports tree.
1444 FreeBSD_version 800015 marks this change for the unlikely
1445 case that a portable fix is impossible.
1448 To upgrade to -current after this date, you must be running
1449 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1450 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1453 The ADAPTIVE_GIANT kernel option has been retired because its
1454 functionality is the default now.
1457 The AT keyboard emulation of sunkbd(4) has been turned on
1458 by default. In order to make the special symbols of the Sun
1459 keyboards driven by sunkbd(4) work under X these now have
1460 to be configured the same way as Sun USB keyboards driven
1461 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1463 Option "XkbLayout" "us"
1464 Option "XkbRules" "xorg"
1465 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1468 It has been decided that it is desirable to provide ABI
1469 backwards compatibility to the FreeBSD 4/5/6 versions of the
1470 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1471 broken with the introduction of PCI domain support (see the
1472 20070930 entry). Unfortunately, this required the ABI of
1473 PCIOCGETCONF to be broken again in order to be able to
1474 provide backwards compatibility to the old version of that
1475 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1476 again. As for prominent ports this affects neither pciutils
1477 nor xorg-server this time, the hal port needs to be rebuilt
1481 The misnamed kthread_create() and friends have been renamed
1482 to kproc_create() etc. Many of the callers already
1483 used kproc_start()..
1484 I will return kthread_create() and friends in a while
1485 with implementations that actually create threads, not procs.
1486 Renaming corresponds with version 800002.
1495 Avoid using make -j when upgrading. While generally safe, there are
1496 sometimes problems using -j to upgrade. If your upgrade fails with
1497 -j, please try again without -j. From time to time in the past there
1498 have been problems using -j with buildworld and/or installworld. This
1499 is especially true when upgrading between "distant" versions (eg one
1500 that cross a major release boundary or several minor releases, or when
1501 several months have passed on the -current branch).
1503 Sometimes, obscure build problems are the result of environment
1504 poisoning. This can happen because the make utility reads its
1505 environment when searching for values for global variables. To run
1506 your build attempts in an "environmental clean room", prefix all make
1507 commands with 'env -i '. See the env(1) manual page for more details.
1509 When upgrading from one major version to another it is generally best
1510 to upgrade to the latest code in the currently installed branch first,
1511 then do an upgrade to the new branch. This is the best-tested upgrade
1512 path, and has the highest probability of being successful. Please try
1513 this approach before reporting problems with a major version upgrade.
1517 When upgrading the boot ZFS pool to a new version, always follow
1520 1.) recompile and reinstall the ZFS boot loader and boot block
1521 (this is part of "make buildworld" and "make installworld")
1523 2.) update the ZFS boot block on your boot drive
1525 The following example updates the ZFS boot block on the first
1526 partition (freebsd-boot) of a GPT partitioned drive ad0:
1527 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1529 Non-boot pools do not need these updates.
1533 If you are updating from a prior version of FreeBSD (even one just
1534 a few days old), you should follow this procedure. It is the most
1535 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1537 make kernel-toolchain
1538 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1539 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1541 To test a kernel once
1542 ---------------------
1543 If you just want to boot a kernel once (because you are not sure
1544 if it works, or if you want to boot a known bad kernel to provide
1545 debugging information) run
1546 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1547 nextboot -k testkernel
1549 To just build a kernel when you know that it won't mess you up
1550 --------------------------------------------------------------
1551 This assumes you are already running a CURRENT system. Replace
1552 ${arch} with the architecture of your machine (e.g. "i386",
1553 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1555 cd src/sys/${arch}/conf
1556 config KERNEL_NAME_HERE
1557 cd ../compile/KERNEL_NAME_HERE
1562 If this fails, go to the "To build a kernel" section.
1564 To rebuild everything and install it on the current system.
1565 -----------------------------------------------------------
1566 # Note: sometimes if you are running current you gotta do more than
1567 # is listed here if you are upgrading from a really old current.
1569 <make sure you have good level 0 dumps>
1571 make kernel KERNCONF=YOUR_KERNEL_HERE
1573 <reboot in single user> [3]
1581 To cross-install current onto a separate partition
1582 --------------------------------------------------
1583 # In this approach we use a separate partition to hold
1584 # current's root, 'usr', and 'var' directories. A partition
1585 # holding "/", "/usr" and "/var" should be about 2GB in
1588 <make sure you have good level 0 dumps>
1591 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1592 <maybe newfs current's root partition>
1593 <mount current's root partition on directory ${CURRENT_ROOT}>
1594 make installworld DESTDIR=${CURRENT_ROOT}
1595 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1596 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1597 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1598 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1599 <reboot into current>
1600 <do a "native" rebuild/install as described in the previous section>
1601 <maybe install compatibility libraries from ports/misc/compat*>
1605 To upgrade in-place from stable to current
1606 ----------------------------------------------
1607 <make sure you have good level 0 dumps>
1609 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1611 <reboot in single user> [3]
1618 Make sure that you've read the UPDATING file to understand the
1619 tweaks to various things you need. At this point in the life
1620 cycle of current, things change often and you are on your own
1621 to cope. The defaults can also change, so please read ALL of
1622 the UPDATING entries.
1624 Also, if you are tracking -current, you must be subscribed to
1625 freebsd-current@freebsd.org. Make sure that before you update
1626 your sources that you have read and understood all the recent
1627 messages there. If in doubt, please track -stable which has
1628 much fewer pitfalls.
1630 [1] If you have third party modules, such as vmware, you
1631 should disable them at this point so they don't crash your
1634 [3] From the bootblocks, boot -s, and then do
1639 adjkerntz -i # if CMOS is wall time
1640 Also, when doing a major release upgrade, it is required that
1641 you boot into single user mode to do the installworld.
1643 [4] Note: This step is non-optional. Failure to do this step
1644 can result in a significant reduction in the functionality of the
1645 system. Attempting to do it by hand is not recommended and those
1646 that pursue this avenue should read this file carefully, as well
1647 as the archives of freebsd-current and freebsd-hackers mailing lists
1648 for potential gotchas. The -U option is also useful to consider.
1649 See mergemaster(8) for more information.
1651 [5] Usually this step is a noop. However, from time to time
1652 you may need to do this if you get unknown user in the following
1653 step. It never hurts to do it all the time. You may need to
1654 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1655 install) after the buildworld before this step if you last updated
1656 from [78]-stable or 9-stable before 20130430.
1658 [6] This only deletes old files and directories. Old libraries
1659 can be deleted by "make delete-old-libs", but you have to make
1660 sure that no program is using those libraries anymore.
1662 [8] In order to have a kernel that can run the 4.x binaries needed to
1663 do an installworld, you must include the COMPAT_FREEBSD4 option in
1664 your kernel. Failure to do so may leave you with a system that is
1665 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1666 required to run the 5.x binaries on more recent kernels. And so on
1667 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1669 Make sure that you merge any new devices from GENERIC since the
1670 last time you updated your kernel config file.
1672 [9] When checking out sources, you must include the -P flag to have
1673 cvs prune empty directories.
1675 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1676 "?=" instead of the "=" assignment operator, so that buildworld can
1677 override the CPUTYPE if it needs to.
1679 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1680 not on the command line, or in /etc/make.conf. buildworld will
1681 warn if it is improperly defined.
1684 This file contains a list, in reverse chronological order, of major
1685 breakages in tracking -current. It is not guaranteed to be a complete
1686 list of such breakages, and only contains entries since October 10, 2007.
1687 If you need to see UPDATING entries from before that date, you will need
1688 to fetch an UPDATING file from an older FreeBSD release.
1690 Copyright information:
1692 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1694 Redistribution, publication, translation and use, with or without
1695 modification, in full or in part, in any form or format of this
1696 document are permitted without further permission from the author.
1698 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1699 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1700 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1701 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1702 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1703 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1704 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1705 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1706 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1707 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1708 POSSIBILITY OF SUCH DAMAGE.
1710 Contact Warner Losh if you have any questions about your use of