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 20140603: p7 FreeBSD-SA-14:11.sendmail
15 FreeBSD-SA-14:12.ktrace
19 Fix sendmail improper close-on-exec flag handling. [SA-14:11]
21 Fix ktrace memory disclosure. [SA-14:12]
23 Fix incorrect error handling in PAM policy parser. [SA-14:13]
25 Fix triple-fault when executing from a threaded process.
28 20140513: p6 FreeBSD-EN-14:03.pkg
29 FreeBSD-EN-14:04.kldxref
32 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
34 Improve build repeatability for kldxref(8). [EN-14:04]
36 Fix data corruption with ciss(4). [EN-14:05]
38 20140430: p5 FreeBSD-SA-14:08.tcp
40 Fix TCP reassembly vulnerability. [SA-14:08]
42 20140408: p4 FreeBSD-SA-14:05.nfsserver
43 FreeBSD-SA-14:06.openssl
44 Fix deadlock in the NFS server. [SA-14:05]
46 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
48 20140114: p3 FreeBSD-SA-14:01.bsnmpd
51 FreeBSD-EN-14:01.random
53 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
55 Fix ntpd distributed reflection Denial of Service
56 vulnerability. [SA-14:02]
58 Fix BIND remote denial of service vulnerability. [SA-14:04]
60 Disable hardware RNGs by default. [EN-14:01]
62 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
64 20131128: p2 FreeBSD-EN-13:05.freebsd-update
65 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
67 20131026: p1 FreeBSD-EN-13:04.freebsd-update
68 Fix multiple freebsd-update bugs that break upgrading to
72 hastctl(8)'s `status' command output changed to terse one-liner format.
73 Scripts using this should switch to `list' command or be rewritten.
76 Fix a bug that allowed a tracing process (e.g. gdb) to write
77 to a memory-mapped file in the traced process's address space
78 even if neither the traced process nor the tracing process had
79 write access to that file.
82 Added ZFS TRIM support which is enabled by default. To disable
83 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
85 Creating new ZFS pools and adding new devices to existing pools
86 first performs a full device level TRIM which can take a significant
87 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
88 to disable this behaviour.
90 ZFS TRIM requires the underlying device support BIO_DELETE which
91 is currently provided by methods such as ATA TRIM and SCSI UNMAP
92 via CAM, which are typically supported by SSD's.
94 Stats for ZFS TRIM can be monitored by looking at the sysctl's
95 under kstat.zfs.misc.zio_trim.
98 `list' command has been added to hastctl(8). For now, it is full
99 equivalent of `status' command.
100 WARNING: in the near future the output of hastctl's status command
101 will change to more terse format. If you use `hastctl status'
102 for parsing in your scripts, switch to `hastctl list'.
105 The mergemaster command now uses the default MAKEOBJDIRPREFIX
106 rather than creating it's own in the temporary directory in
107 order allow access to bootstrapped versions of tools such as
108 install and mtree. When upgrading from version of FreeBSD where
109 the install command does not support -l, you will need to
110 install a new mergemaster command if mergemaster -p is required.
111 This can be accomplished with the command (cd src/usr.sbin/mergemaster
114 Due to the use of the new -l option to install(1) during build
115 and install, you must take care not to directly set the INSTALL
116 make variable in your /etc/make.conf, /etc/src.conf, or on the
117 command line. If you wish to use the -C flag for all installs
118 you may be able to add INSTALL+=-C to /etc/make.conf or
122 Fix a bug that allows NFS clients to issue READDIR on files.
125 The install(1) option -M has changed meaning and now takes an
126 argument that is a file or path to append logs to. In the
127 unlikely event that -M was the last option on the command line
128 and the command line contained at least two files and a target
129 directory the first file will have logs appended to it. The -M
130 option served little practical purpose in the last decade so it's
131 used expected to be extremely rare.
134 A new compression method (lz4) has been merged to. Please refer to
135 zpool-features(7) for more information.
137 Please refer to the "ZFS notes" section of this file for information
138 on upgrading boot ZFS pools.
141 The VFS KBI was changed with the merge of several nullfs
142 optimizations and fixes. All filesystem modules must be
146 With the addition of auditdistd(8), a new auditdistd user is now
147 depended on during installworld. "mergemaster -p" can be used to add
148 the user prior to installworld, as documented in the handbook.
154 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
155 Starting with this version the old system of ZFS pool versioning
156 is superseded by "feature flags". This concept enables forward
157 compatibility against certain future changes in functionality of ZFS
158 pools. The first two read-only compatible "feature flags" for ZFS
159 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
160 For more information read the new zpool-features(7) manual page.
161 Please refer to the "ZFS notes" section of this file for information
162 on upgrading boot ZFS pools.
165 The commit introducing bsd.compiler.mk breaks the traditional
166 building of kernels before this point. Add -m ${SRC}/share/mk
167 (for the right value of SRC) to your command lines to work
168 around; update your useland to a point after this; or use the
169 buildkernel/installkernel top-level targets. See also 20120829.
172 The IPFIREWALL_FORWARD kernel option has been removed. Its
173 functionality now turned on by default.
176 The random(4) support for the VIA hardware random number
177 generator (`PADLOCK') is no longer enabled unconditionally.
178 Add the PADLOCK_RNG option in the custom kernel config if
179 needed. The GENERIC kernels on i386 and amd64 do include the
180 option, so the change only affects the custom kernel
184 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
185 the traditional method, you must update your system with an installworld
186 before the kernel will build. The documented make buildkernel/installkernel
187 interfaces (coupled with fresh make kernel-toolchain) continue to work.
190 The sparc64 ZFS loader has been changed to no longer try to auto-
191 detect ZFS providers based on diskN aliases but now requires these
192 to be explicitly listed in the OFW boot-device environment variable.
195 Now unix domain sockets behave "as expected" on nullfs(5). Previously
196 nullfs(5) did not pass through all behaviours to the underlying layer,
197 as a result if we bound to a socket on the lower layer we could connect
198 only to the lower path; if we bound to the upper layer we could connect
199 only to the upper path. The new behavior is one can connect to both the
200 lower and the upper paths regardless what layer path one binds to.
203 The acpi_wmi(4) status device /dev/wmistat has been renamed to
207 A new VOP_ADVISE() was added to support posix_fadvise(2). All
208 filesystem modules must be recompiled.
211 The interface of the VOP_VPTOCNP(9) changed, now the returned
212 vnode shall be referenced, previously it was required to be
213 only held. All in-tree filesystems are converted.
219 The broken amd(4) driver has been replaced with esp(4) in the amd64,
220 i386 and pc98 GENERIC kernel configuration files.
223 This commit modifies vfs_register() so that it uses a hash
224 calculation to set vfc_typenum, which is enabled by default.
225 The first time a system is booted after this change, the
226 vfc_typenum values will change for all file systems. The
227 main effect of this is a change to the NFS server file handles
228 for file systems that use vfc_typenum in their fsid, such as ZFS.
229 It will, however, prevent vfc_typenum from changing when file
230 systems are loaded in a different order for subsequent reboots.
231 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
232 until you are ready to remount all NFS clients after a reboot.
235 Bump the shared library version numbers for libraries that
236 do not use symbol versioning, have changed the ABI compared
237 to stable/8 and which shared library version was not bumped.
238 Done as part of 9.0-RELEASE cycle.
241 During the merge of Capsicum features, the fget(9) KPI was modified.
242 This may require the rebuilding of out-of-tree device drivers --
243 issues have been reported specifically with the nVidia device driver.
244 __FreeBSD_version is bumped to 900041.
246 Also, there is a period between 20110811 and 20110814 where the
247 special devices /dev/{stdin,stdout,stderr} did not work correctly.
248 Building world from a kernel during that window may not work.
251 The packet filter (pf) code has been updated to OpenBSD 4.5.
252 You need to update userland tools to be in sync with kernel.
253 This update breaks backward compatibility with earlier pfsync(4)
254 versions. Care must be taken when updating redundant firewall setups.
257 The following sysctls and tunables are retired on x86 platforms:
259 machdep.hlt_logical_cpus
260 The following sysctl is retired:
261 machdep.hyperthreading_allowed
262 The sysctls were supposed to provide a way to dynamically offline and
263 online selected CPUs on x86 platforms, but the implementation has not
264 been reliable especially with SCHED_ULE scheduler.
265 machdep.hyperthreading_allowed tunable is still available to ignore
266 hyperthreading CPUs at OS level.
267 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
268 where X is an APIC ID of a CPU. Be advised, though, that disabling
269 CPUs in non-uniform fashion will result in non-uniform topology and
270 may lead to sub-optimal system performance with SCHED_ULE, which is
274 cpumask_t type is retired and cpuset_t is used in order to describe
278 Changes to ifconfig(8) for dynamic address family detection mandate
279 that you are running a kernel of 20110525 or later. Make sure to
280 follow the update procedure to boot a new kernel before installing
284 Support for sun4v architecture is officially dropped
287 Several KPI breaking changes have been committed to the mii(4) layer,
288 the PHY drivers and consequently some Ethernet drivers using mii(4).
289 This means that miibus.ko and the modules of the affected Ethernet
290 drivers need to be recompiled.
292 Note to kernel developers: Given that the OUI bit reversion problem
293 was fixed as part of these changes all mii(4) commits related to OUIs,
294 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
295 handling, no longer can be merged verbatim to stable/8 and previous
299 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
300 into their kernel configurations along with 'device pci'.
303 The default NFS client is now the new NFS client, so fstype "newnfs"
304 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
305 Although mounts via fstype "nfs" will usually work without userland
306 changes, it is recommended that the mount(8) and mount_nfs(8)
307 commands be rebuilt from sources and that a link to mount_nfs called
308 mount_oldnfs be created. The new client is compiled into the
309 kernel with "options NFSCL" and this is needed for diskless root
310 file systems. The GENERIC kernel configs have been changed to use
311 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
312 To use the regular/old client, you can "mount -t oldnfs ...". For
313 a diskless root file system, you must also include a line like:
315 vfs.root.mountfrom="oldnfs:"
317 in the boot/loader.conf on the root fs on the NFS server to make
318 a diskless root fs use the old client.
321 The GENERIC kernels for all architectures now default to the new
322 CAM-based ATA stack. It means that all legacy ATA drivers were
323 removed and replaced by respective CAM drivers. If you are using
324 ATA device names in /etc/fstab or other places, make sure to update
325 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
326 where 'Y's are the sequential numbers starting from zero for each type
327 in order of detection, unless configured otherwise with tunables,
328 see cam(4)). There will be symbolic links created in /dev/ to map
329 old adX devices to the respective adaY. They should provide basic
330 compatibility for file systems mounting in most cases, but they do
331 not support old user-level APIs and do not have respective providers
332 in GEOM. Consider using updated management tools with new device names.
334 It is possible to load devices ahci, ata, siis and mvs as modules,
335 but option ATA_CAM should remain in kernel configuration to make ata
336 module work as CAM driver supporting legacy ATA controllers. Device ata
337 still can be used in modular fashion (atacore + ...). Modules atadisk
338 and atapi* are not used and won't affect operation in ATA_CAM mode.
339 Note that to use CAM-based ATA kernel should include CAM devices
340 scbus, pass, da (or explicitly ada), cd and optionally others. All of
341 them are parts of the cam module.
343 ataraid(4) functionality is now supported by the RAID GEOM class.
344 To use it you can load geom_raid kernel module and use graid(8) tool
345 for management. Instead of /dev/arX device names, use /dev/raid/rX.
347 No kernel config options or code have been removed, so if a problem
348 arises, please report it and optionally revert to the old ATA stack.
349 In order to do it you can remove from the kernel config:
354 , and instead add back:
355 device atadisk # ATA disk drives
356 device ataraid # ATA RAID drives
357 device atapicd # ATAPI CDROM drives
358 device atapifd # ATAPI floppy drives
359 device atapist # ATAPI tape drives
362 The default NFS server has been changed to the new server, which
363 was referred to as the experimental server. If you need to switch
364 back to the old NFS server, you must now put the "-o" option on
365 both the mountd and nfsd commands. This can be done using the
366 mountd_flags and nfs_server_flags rc.conf variables until an
367 update to the rc scripts is committed, which is coming soon.
370 The GNU Objective-C runtime library (libobjc), and other Objective-C
371 related components have been removed from the base system. If you
372 require an Objective-C library, please use one of the available ports.
375 ath(4) has been split into bus- and device- modules. if_ath contains
376 the HAL, the TX rate control and the network device code. if_ath_pci
377 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
378 contains the AHB glue. Users need to load both if_ath_pci and if_ath
379 in order to use ath on everything else.
381 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
382 need to load if_ath and if_ath_pci for ath(4) operation.
385 As part of the replacement of sysinstall, the process of building
386 release media has changed significantly. For details, please re-read
387 release(7), which has been updated to reflect the new build process.
390 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
391 is the last available version under GPLv2. It brings a number of new
392 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
393 4.1 and SSE 4.2), better support for powerpc64, a number of new
394 directives, and lots of other small improvements. See the ChangeLog
395 file in contrib/binutils for the full details.
398 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
399 compliant, and will now use half of hash for authentication.
400 This will break interoperability with all stacks (including all
401 actual FreeBSD versions) who implement
402 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
404 The only workaround with such peers is to use another HMAC
405 algorithm for IPsec ("phase 2") authentication.
408 Remove the uio_yield prototype and symbol. This function has
409 been misnamed since it was introduced and should not be
410 globally exposed with this name. The equivalent functionality
411 is now available using kern_yield(curthread->td_user_pri).
412 The function remains undocumented.
415 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
416 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
417 for scalar sysctls is defined but disabled. Code that needs
418 UQUAD to pass the type checking that must compile on older
419 systems where the define is not present can check against
420 __FreeBSD_version >= 900030.
422 The system dialog(1) has been replaced with a new version previously
423 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
424 with the previous version, but the libdialog associated with it has
425 a largely incompatible API. As such, the original version of libdialog
426 will be kept temporarily as libodialog, until its base system consumers
427 are replaced or updated. Bump __FreeBSD_version to 900030.
430 If you are trying to run make universe on a -stable system, and you get
431 the following warning:
432 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
433 unknown. config(8) likely too old."
434 or something similar to it, then you must upgrade your -stable system
435 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
436 stable/8) or build the config from the latest stable/8 branch and
437 install it on your system.
439 Prior to this date, building a current universe on 8-stable system from
440 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
441 error in the first kernel build phase. A new config on those old
442 systems will fix that problem for older versions of -current.
445 The TCP stack has been modified to allow Khelp modules to interact with
446 it via helper hook points and store per-connection data in the TCP
447 control block. Bump __FreeBSD_version to 900029. User space tools that
448 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
452 Generic IEEE 802.3 annex 31B full duplex flow control support has been
453 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
454 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
455 to take advantage of it instead of using custom implementations. This
456 means that these drivers now no longer unconditionally advertise
457 support for flow control but only do so if flow control is a selected
458 media option. This was implemented in the generic support that way in
459 order to allow flow control to be switched on and off via ifconfig(8)
460 with the PHY specific default to typically off in order to protect
461 from unwanted effects. Consequently, if you used flow control with
462 one of the above mentioned drivers you now need to explicitly enable
464 ifconfig bge0 media auto mediaopt flowcontrol
466 Along with the above mentioned changes generic support for setting
467 1000baseT master mode also has been added and brgphy(4), ciphy(4),
468 e1000phy(4) as well as ip1000phy(4) have been converted to take
469 advantage of it. This means that these drivers now no longer take the
470 link0 parameter for selecting master mode but the master media option
471 has to be used instead, for example like in the following:
472 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
474 Selection of master mode now is also available with all other PHY
475 drivers supporting 1000baseT.
478 The TCP stack has received a significant update to add support for
479 modularised congestion control and generally improve the clarity of
480 congestion control decisions. Bump __FreeBSD_version to 900025. User
481 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
482 sockstat) need to be recompiled.
485 The man(1) utility has been replaced by a new version that no longer
486 uses /etc/manpath.config. Please consult man.conf(5) for how to
487 migrate local entries to the new format.
490 The copyright strings printed by login(1) and sshd(8) at the time of a
491 new connection have been removed to follow other operating systems and
495 A workaround for a fixed ld bug has been removed in kernel code,
496 so make sure that your system ld is built from sources after
497 revision 210245 from 2010-07-19 (r211583 if building head kernel
498 on stable/8, r211584 for stable/7; both from 2010-08-21).
499 A symptom of incorrect ld version is different addresses for
500 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
503 The $ipv6_prefer variable in rc.conf(5) has been split into
504 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
506 The $ip6addrctl_policy is a variable to choose a pre-defined
507 address selection policy set by ip6addrctl(8). A value
508 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
511 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
512 flag (see an entry of 20090926) is set on an interface with no
513 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
514 security reason. If you want IPv6 link-local address on all
515 interfaces by default, set this to "YES".
517 The old ipv6_prefer="YES" is equivalent to
518 ipv6_activate_all_interfaces="YES" and
519 ip6addrctl_policy="ipv6_prefer".
522 DTrace has grown support for userland tracing. Due to this, DTrace is
523 now i386 and amd64 only.
524 dtruss(1) is now installed by default on those systems and a new
525 kernel module is needed for userland tracing: fasttrap.
526 No changes to your kernel config file are necessary to enable
527 userland tracing, but you might consider adding 'STRIP=' and
528 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
529 to have informative userland stack traces in DTrace (ustack).
532 The acpi_aiboost(4) driver has been removed in favor of the new
533 aibs(4) driver. You should update your kernel configuration file.
536 BSD grep has been imported to the base system and it is built by
537 default. It is completely BSD licensed, highly GNU-compatible, uses
538 less memory than its GNU counterpart and has a small codebase.
539 However, it is slower than its GNU counterpart, which is mostly
540 noticeable for larger searches, for smaller ones it is measurable
541 but not significant. The reason is complex, the most important factor
542 is that we lack a modern and efficient regex library and GNU
543 overcomes this by optimizing the searches internally. Future work
544 on improving the regex performance is planned, for the meantime,
545 users that need better performance, can build GNU grep instead by
546 setting the WITH_GNU_GREP knob.
549 Due to the import of powerpc64 support, all existing powerpc kernel
550 configuration files must be updated with a machine directive like this:
551 machine powerpc powerpc
553 In addition, an updated config(8) is required to build powerpc kernels
557 A new version of ZFS (version 15) has been merged to -HEAD.
558 This version uses a python library for the following subcommands:
559 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
560 For full functionality of these commands the following port must
561 be installed: sysutils/py-zfs
564 'vm_page's are now hashed by physical address to an array of mutexes.
565 Currently this is only used to serialize access to hold_count. Over
566 time the page queue mutex will be peeled away. This changes the size
567 of pmap on every architecture. And requires all callers of vm_page_hold
568 and vm_page_unhold to be updated.
571 WITH_CTF can now be specified in src.conf (not recommended, there
572 are some problems with static executables), make.conf (would also
573 affect ports which do not use GNU make and do not override the
574 compile targets) or in the kernel config (via "makeoptions
576 When WITH_CTF was specified there before this was silently ignored,
577 so make sure that WITH_CTF is not used in places which could lead
578 to unwanted behavior.
581 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
582 to allow 32-bit compatibility on non-x86 platforms. All kernel
583 configurations on amd64 and ia64 platforms using these options must
584 be modified accordingly.
587 The utmp user accounting database has been replaced with utmpx,
588 the user accounting interface standardized by POSIX.
589 Unfortunately the semantics of utmp and utmpx don't match,
590 making it practically impossible to support both interfaces.
591 The user accounting database is used by tools like finger(1),
592 last(1), talk(1), w(1) and ac(8).
594 All applications in the base system use utmpx. This means only
595 local binaries (e.g. from the ports tree) may still use these
596 utmp database files. These applications must be rebuilt to make
599 After the system has been upgraded, it is safe to remove the old
600 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
601 assuming their contents is of no importance anymore. Old wtmp
602 databases can only be used by last(1) and ac(8) after they have
603 been converted to the new format using wtmpcvt(1).
606 Introduce the kernel thread "deadlock resolver" (which can be enabled
607 via the DEADLKRES option, see NOTES for more details) and the
608 sleepq_type() function for sleepqueues.
611 The rc.firewall and rc.firewall6 were unified, and
612 rc.firewall6 and rc.d/ip6fw were removed.
613 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
614 variables are obsoleted. Instead, the following new rc
615 variables are added to rc.d/ipfw:
617 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
618 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
619 firewall_simple_onet_ipv6, firewall_trusted_ipv6
621 The meanings correspond to the relevant IPv4 variables.
627 The default terminal emulation for syscons(4) has been changed
628 from cons25 to xterm on all platforms except pc98. This means
629 that the /etc/ttys file needs to be updated to ensure correct
630 operation of applications on the console.
632 The terminal emulation style can be toggled per window by using
633 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
634 options can be used to change the compile-time default back to
637 To prevent graphical artifacts, make sure the TERM environment
638 variable is set to match the terminal emulation that is being
639 performed by syscons(4).
642 The layout of the structure ieee80211req_scan_result has changed.
643 Applications that require wireless scan results (e.g. ifconfig(8))
644 from net80211 need to be recompiled.
646 Applications such as wpa_supplicant(8) may require a full world
647 build without using NO_CLEAN in order to get synchronized with the
651 The iwn(4) driver has been updated to support the 5000 and 5150 series.
652 There's one kernel module for each firmware. Adding "device iwnfw"
653 to the kernel configuration file means including all three firmware
654 images inside the kernel. If you want to include just the one for
655 your wireless card, use the devices iwn4965fw, iwn5000fw or
659 The rc.d/network_ipv6, IPv6 configuration script has been integrated
660 into rc.d/netif. The changes are the following:
662 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
663 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
664 Note that both variables need the "inet6" keyword at the head.
666 Do not set $ipv6_network_interfaces manually if you do not
667 understand what you are doing. It is not needed in most cases.
669 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
672 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
673 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
675 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
676 all configured interfaces have "inet6 accept_rtadv" in the
677 $ifconfig_IF_ipv6. These are for backward compatibility.
679 3. A new variable $ipv6_prefer has been added. If NO, IPv6
680 functionality of interfaces with no corresponding
681 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
682 and the default address selection policy of ip6addrctl(8)
683 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
684 Note that if you want to configure IPv6 functionality on the
685 disabled interfaces after boot, first you need to clear the flag by
686 using ifconfig(8) like:
688 ifconfig em0 inet6 -ifdisabled
690 If YES, the default address selection policy is set as
693 The default value of $ipv6_prefer is NO.
695 4. If your system need to receive Router Advertisement messages,
696 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
697 scripts automatically invoke rtsol(8) when the interface becomes
698 UP. The Router Advertisement messages are used for SLAAC
699 (State-Less Address AutoConfiguration).
702 802.11s D3.03 support was committed. This is incompatible with the
703 previous code, which was based on D3.0.
706 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
707 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
708 control whether accepting Router Advertisement messages or not.
709 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
710 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
711 The ifconfig(8) utility now supports these flags.
714 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
715 mount(8) and -a option for df(1) to see them.
718 The old tunable hw.bus.devctl_disable has been superseded by
719 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
720 replaced by hw.bus.devctl_queue=0. The default for this new tunable
724 Remove the option STOP_NMI. The default action is now to use NMI only
725 for KDB via the newly introduced function stop_cpus_hard() and
726 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
729 The stable/8 branch created in subversion. This corresponds to the
730 RELENG_8 branch in CVS.
733 Bump the shared library version numbers for all libraries that do not
734 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
735 __FreeBSD_version to 800105.
738 Due to changes in the implementation of virtual network stack support,
739 all network-related kernel modules must be recompiled. As this change
740 breaks the ABI, bump __FreeBSD_version to 800104.
743 The TOE interface to the TCP syncache has been modified to remove
744 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
745 The cxgb driver is the only TOE consumer affected by this change, and
746 needs to be recompiled along with the kernel. As this change breaks
747 the ABI, bump __FreeBSD_version to 800103.
750 Padding has been added to struct tcpcb, sackhint and tcpstat in
751 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
752 maintaining the ABI. However, this change breaks the ABI, so bump
753 __FreeBSD_version to 800102. User space tools that rely on the size of
754 any of these structs (e.g. sockstat) need to be recompiled.
757 The NFS_LEGACYRPC option has been removed along with the old kernel
758 RPC implementation that this option selected. Kernel configurations
759 may need to be adjusted.
762 The network interface device nodes at /dev/net/<interface> have been
763 removed. All ioctl operations can be performed the normal way using
764 routing sockets. The kqueue functionality can generally be replaced
765 with routing sockets.
768 The documentation from the FreeBSD Documentation Project (Handbook,
769 FAQ, etc.) is now installed via packages by sysinstall(8) and under
770 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
773 The ABI of various structures related to the SYSV IPC API have been
774 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
775 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
779 Layout of struct vnet has changed as routing related variables were
780 moved to their own Vimage module. Modules need to be recompiled. Bump
781 __FreeBSD_version to 800099.
784 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
785 respectively. As long as no more than 16 groups per process are used,
786 no changes should be visible. When more than 16 groups are used, old
787 binaries may fail if they call getgroups() or getgrouplist() with
788 statically sized storage. Recompiling will work around this, but
789 applications should be modified to use dynamically allocated storage
790 for group arrays as POSIX.1-2008 does not cap an implementation's
791 number of supported groups at NGROUPS_MAX+1 as previous versions did.
793 NFS and portalfs mounts may also be affected as the list of groups is
794 truncated to 16. Users of NFS who use more than 16 groups, should
795 take care that negative group permissions are not used on the exported
796 file systems as they will not be reliable unless a GSSAPI based
797 authentication method is used.
800 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
801 option compiles in the support for adaptive spinning for lockmgrs
802 which want to enable it. The lockinit() function now accepts the flag
803 LK_ADAPTIVE in order to make the lock object subject to adaptive
804 spinning when both held in write and read mode.
807 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
808 changed. User applications that use this ioctl need to be rebuilt.
811 The layout of struct thread has changed. Kernel and modules need to
815 The layout of structs ifnet, domain, protosw and vnet_net has changed.
816 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
819 window(1) has been removed from the base system. It can now be
820 installed from ports. The port is called misc/window.
823 The way we are storing and accessing `routing table' entries has
824 changed. Programs reading the FIB, like netstat, need to be
828 A new netisr implementation has been added for FreeBSD 8. Network
829 file system modules, such as igmp, ipdivert, and others, should be
831 Bump __FreeBSD_version to 800096.
834 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
838 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
839 Bump __FreeBSD_version to 800094.
842 Add mnt_xflag field to 'struct mount'. File system modules need to be
844 Bump __FreeBSD_version to 800093.
847 The compiling option ADAPTIVE_SX has been retired while it has been
848 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
849 The KPI for sx_init_flags() changes as accepting flags:
850 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
851 been introduced in order to handle the reversed logic.
852 Bump __FreeBSD_version to 800092.
855 Add support for hierarchical jails. Remove global securelevel.
856 Bump __FreeBSD_version to 800091.
859 The layout of struct vnet_net has changed, therefore modules
861 Bump __FreeBSD_version to 800090.
864 The newly imported zic(8) produces a new format in the output. Please
865 run tzsetup(8) to install the newly created data to /etc/localtime.
868 The sysctl tree for the usb stack has renamed from hw.usb2.* to
869 hw.usb.* and is now consistent again with previous releases.
872 802.11 monitor mode support was revised and driver api's were changed.
873 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
874 of DLT_IEEE802_11. No user-visible data structures were changed but
875 applications that use DLT_IEEE802_11 may require changes.
876 Bump __FreeBSD_version to 800088.
879 The layout of the following structs has changed: sysctl_oid,
880 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
881 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
882 panics may be experienced. World rebuild is required for
883 correctly checking networking state from userland.
884 Bump __FreeBSD_version to 800085.
887 MLDv2 and Source-Specific Multicast (SSM) have been merged
888 to the IPv6 stack. VIMAGE hooks are in but not yet used.
889 The implementation of SSM within FreeBSD's IPv6 stack closely
890 follows the IPv4 implementation.
892 For kernel developers:
894 * The most important changes are that the ip6_output() and
895 ip6_input() paths no longer take the IN6_MULTI_LOCK,
896 and this lock has been downgraded to a non-recursive mutex.
898 * As with the changes to the IPv4 stack to support SSM, filtering
899 of inbound multicast traffic must now be performed by transport
900 protocols within the IPv6 stack. This does not apply to TCP and
901 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
903 * The KPIs used by IPv6 multicast are similar to those used by
904 the IPv4 stack, with the following differences:
905 * im6o_mc_filter() is analogous to imo_multicast_filter().
906 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
907 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
908 * IN6_LOOKUP_MULTI() has been deprecated and removed.
909 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
910 for MLDv1 have an additional 'timer' argument which is used to
911 jitter the initial membership report for the solicited-node
912 multicast membership on-link.
913 * This is not strictly needed for MLDv2, which already jitters
914 its report transmissions. However, the 'timer' argument is
915 preserved in case MLDv1 is active on the interface.
917 * The KAME linked-list based IPv6 membership implementation has
918 been refactored to use a vector similar to that used by the IPv4
920 Code which maintains a list of its own multicast memberships
921 internally, e.g. carp, has been updated to reflect the new
924 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
925 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
926 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
927 implementation constraint which needs to be addressed in HEAD.
929 For application developers:
931 * The changes are broadly similar to those made for the IPv4
934 * The use of IPv4 and IPv6 multicast socket options on the same
935 socket, using mapped addresses, HAS NOT been tested or supported.
937 * There are a number of issues with the implementation of various
938 IPv6 multicast APIs which need to be resolved in the API surface
939 before the implementation is fully compatible with KAME userland
940 use, and these are mostly to do with interface index treatment.
942 * The literature available discusses the use of either the delta / ASM
943 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
944 using setsourcefilter(3)/getsourcefilter(3). For more information
945 please refer to RFC 3768, 'Socket Interface Extensions for
946 Multicast Source Filters'.
948 * Applications which use the published RFC 3678 APIs should be fine.
950 For systems administrators:
952 * The mtest(8) utility has been refactored to support IPv6, in
953 addition to IPv4. Interface addresses are no longer accepted
954 as arguments, their names must be used instead. The utility
955 will map the interface name to its first IPv4 address as
956 returned by getifaddrs(3).
958 * The ifmcstat(8) utility has also been updated to print the MLDv2
959 endpoint state and source filter lists via sysctl(3).
961 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
962 loopback of IPv6 multicast datagrams by default; it defaults to 1
963 to preserve the existing behaviour. Disabling multicast loopback is
964 recommended for optimal system performance.
966 * The IPv6 MROUTING code has been changed to examine this sysctl
967 instead of attempting to perform a group lookup before looping
968 back forwarded datagrams.
970 Bump __FreeBSD_version to 800084.
973 Implement low-level Bluetooth HCI API.
974 Bump __FreeBSD_version to 800083.
977 The layout of struct malloc_type, used by modules to register new
978 memory allocation types, has changed. Most modules will need to
979 be rebuilt or panics may be experienced.
980 Bump __FreeBSD_version to 800081.
983 Anticipate overflowing inp_flags - add inp_flags2.
984 This changes most offsets in inpcb, so checking v4 connection
985 state will require a world rebuild.
986 Bump __FreeBSD_version to 800080.
989 Add an llentry to struct route and struct route_in6. Modules
990 embedding a struct route will need to be recompiled.
991 Bump __FreeBSD_version to 800079.
994 The size of rt_metrics_lite and by extension rtentry has changed.
995 Networking administration apps will need to be recompiled.
996 The route command now supports show as an alias for get, weighting
997 of routes, sticky and nostick flags to alter the behavior of stateful
999 Bump __FreeBSD_version to 800078.
1002 Do not use Giant for kbdmux(4) locking. This is wrong and
1003 apparently causing more problems than it solves. This will
1004 re-open the issue where interrupt handlers may race with
1005 kbdmux(4) in polling mode. Typical symptoms include (but
1006 not limited to) duplicated and/or missing characters when
1007 low level console functions (such as gets) are used while
1008 interrupts are enabled (for example geli password prompt,
1009 mountroot prompt etc.). Disabling kbdmux(4) may help.
1012 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
1013 kernel modules referencing any of the above need to be recompiled.
1014 Bump __FreeBSD_version to 800075.
1017 GEOM_PART has become the default partition slicer for storage devices,
1018 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
1019 introduces some changes:
1021 MSDOS/EBR: the devices created from MSDOS extended partition entries
1022 (EBR) can be named differently than with GEOM_MBR and are now symlinks
1023 to devices with offset-based names. fstabs may need to be modified.
1025 BSD: the "geometry does not match label" warning is harmless in most
1026 cases but it points to problems in file system misalignment with
1027 disk geometry. The "c" partition is now implicit, covers the whole
1028 top-level drive and cannot be (mis)used by users.
1030 General: Kernel dumps are now not allowed to be written to devices
1031 whose partition types indicate they are meant to be used for file
1032 systems (or, in case of MSDOS partitions, as something else than
1035 Most of these changes date approximately from 200812.
1038 The uscanner(4) driver has been removed from the kernel. This follows
1039 Linux removing theirs in 2.6 and making libusb the default interface
1040 (supported by sane).
1043 The multicast forwarding code has been cleaned up. netstat(1)
1044 only relies on KVM now for printing bandwidth upcall meters.
1045 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1046 ip6_mroute_mod respectively. The config(5) options for statically
1047 compiling this code remain the same, i.e. 'options MROUTING'.
1050 Support for the IFF_NEEDSGIANT network interface flag has been
1051 removed, which means that non-MPSAFE network device drivers are no
1052 longer supported. In particular, if_ar, if_sr, and network device
1053 drivers from the old (legacy) USB stack can no longer be built or
1057 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1058 a bunch of new language catalog files have also been added.
1059 This means that some common libc messages are now localized and
1060 they depend on the LC_MESSAGES environmental variable.
1063 The k8temp(4) driver has been renamed to amdtemp(4) since
1064 support for Family 10 and Family 11 CPU families was added.
1067 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1068 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1070 For kernel developers, the most important changes are that the
1071 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1072 and this lock has been downgraded to a non-recursive mutex.
1074 Transport protocols (UDP, Raw IP) are now responsible for filtering
1075 inbound multicast traffic according to group membership and source
1076 filters. The imo_multicast_filter() KPI exists for this purpose.
1077 Transports which do not use multicast (SCTP, TCP) already reject
1078 multicast by default. Forwarding and receive performance may improve
1079 as a mutex acquisition is no longer needed in the ip_input()
1080 low-level input path. in_addmulti() and in_delmulti() are shimmed
1081 to new KPIs which exist to support SSM in-kernel.
1083 For application developers, it is recommended that loopback of
1084 multicast datagrams be disabled for best performance, as this
1085 will still cause the lock to be taken for each looped-back
1086 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1087 be tuned to 0 to disable loopback by default; it defaults to 1
1088 to preserve the existing behaviour.
1090 For systems administrators, to obtain best performance with
1091 multicast reception and multiple groups, it is always recommended
1092 that a card with a suitably precise hash filter is used. Hash
1093 collisions will still result in the lock being taken within the
1094 transport protocol input path to check group membership.
1096 If deploying FreeBSD in an environment with IGMP snooping switches,
1097 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1098 enabled; this forces 224.0.0.0/24 group membership to be announced
1101 The size of 'struct igmpstat' has changed; netstat needs to be
1102 recompiled to reflect this.
1103 Bump __FreeBSD_version to 800070.
1106 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1107 updated to use it. This requires a buildworld/installworld in order to
1108 update the library and dependencies (usbconfig, etc). Its advisable to
1109 rebuild all ports which uses libusb. More specific directions are given
1110 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1111 libusb are no longer required and can be removed.
1114 A workaround is committed to allow the creation of System V shared
1115 memory segment of size > 2 GB on the 64-bit architectures.
1116 Due to a limitation of the existing ABI, the shm_segsz member
1117 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1118 wrong for large segments. Note that limits must be explicitly
1119 raised to allow such segments to be created.
1122 The layout of struct ifnet has changed, requiring a rebuild of all
1123 network device driver modules.
1126 The /dev handling for the new USB stack has changed, a
1127 buildworld/installworld is required for libusb20.
1130 The new USB2 stack has now been permanently moved in and all kernel and
1131 module names reverted to their previous values (eg, usb, ehci, ohci,
1132 ums, ...). The old usb stack can be compiled in by prefixing the name
1133 with the letter 'o', the old usb modules have been removed.
1134 Updating entry 20090216 for xorg and 20090215 for libmap may still
1138 The rc.conf(5) option if_up_delay has been renamed to
1139 defaultroute_delay to better reflect its purpose. If you have
1140 customized this setting in /etc/rc.conf you need to update it to
1144 xorg 7.4 wants to configure its input devices via hald which does not
1145 yet work with USB2. If the keyboard/mouse does not work in xorg then
1147 Option "AllowEmptyInput" "off"
1148 to your ServerLayout section. This will cause X to use the configured
1149 kbd and mouse sections from your xorg.conf.
1152 The GENERIC kernels for all architectures now default to the new USB2
1153 stack. No kernel config options or code have been removed so if a
1154 problem arises please report it and optionally revert to the old USB
1155 stack. If you are loading USB kernel modules or have a custom kernel
1156 that includes GENERIC then ensure that usb names are also changed over,
1157 eg uftdi -> usb2_serial_ftdi.
1159 Older programs linked against the ports libusb 0.1 need to be
1160 redirected to the new stack's libusb20. /etc/libmap.conf can
1162 # Map old usb library to new one for usb2 stack
1163 libusb-0.1.so.8 libusb20.so.1
1166 All USB ethernet devices now attach as interfaces under the name ueN
1167 (eg. ue0). This is to provide a predictable name as vendors often
1168 change usb chipsets in a product without notice.
1171 The ichsmb(4) driver has been changed to require SMBus slave
1172 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1173 All of the other SMBus controller drivers require left-justified
1174 slave addresses, so this change makes all the drivers provide the
1178 INET6 statistics (struct ip6stat) was updated.
1179 netstat(1) needs to be recompiled.
1182 NTFS has been removed from GENERIC kernel on amd64 to match
1183 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1184 will load ntfs.ko module automatically when NTFS support is
1185 actually needed, unless ntfs.ko is not installed or security
1186 level prohibits loading kernel modules. If either is the case,
1187 "options NTFS" has to be added into kernel config.
1190 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1191 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1192 800061. User space tools that rely on the size of struct tcpcb in
1193 tcp_var.h (e.g. sockstat) need to be recompiled.
1196 ng_tty(4) module updated to match the new TTY subsystem.
1197 Due to API change, user-level applications must be updated.
1198 New API support added to mpd5 CVS and expected to be present
1199 in next mpd5.3 release.
1202 With __FreeBSD_version 800060 the makefs tool is part of
1203 the base system (it was a port).
1206 The afdata and ifnet locks have been changed from mutexes to
1207 rwlocks, network modules will need to be re-compiled.
1210 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1211 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1212 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1213 architecture and 40 bytes on 64-bit architecture. The userland
1214 applications "arp" and "ndp" have been updated accordingly.
1215 The output from "netstat -r" shows only routing entries and
1216 none of the L2 information.
1219 __FreeBSD_version 800057 marks the switchover from the
1220 binary ath hal to source code. Users must add the line:
1222 options AH_SUPPORT_AR5416
1224 to their kernel config files when specifying:
1228 The ath_hal module no longer exists; the code is now compiled
1229 together with the driver in the ath module. It is now
1230 possible to tailor chip support (i.e. reduce the set of chips
1231 and thereby the code size); consult ath_hal(4) for details.
1234 __FreeBSD_version 800054 adds memory barriers to
1235 <machine/atomic.h>, new interfaces to ifnet to facilitate
1236 multiple hardware transmit queues for cards that support
1237 them, and a lock-less ring-buffer implementation to
1238 enable drivers to more efficiently manage queueing of
1242 A new version of ZFS (version 13) has been merged to -HEAD.
1243 This version has zpool attribute "listsnapshots" off by
1244 default, which means "zfs list" does not show snapshots,
1245 and is the same as Solaris behavior.
1248 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1251 The uhci, ohci, ehci and slhci USB Host controller drivers have
1252 been put into separate modules. If you load the usb module
1253 separately through loader.conf you will need to load the
1254 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1255 controller add the following to loader.conf:
1261 The ABI used by the PMC toolset has changed. Please keep
1262 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1266 atapci kernel module now includes only generic PCI ATA
1267 driver. AHCI driver moved to ataahci kernel module.
1268 All vendor-specific code moved into separate kernel modules:
1269 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1270 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1271 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1272 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1275 The TTY subsystem of the kernel has been replaced by a new
1276 implementation, which provides better scalability and an
1277 improved driver model. Most common drivers have been migrated to
1278 the new TTY subsystem, while others have not. The following
1279 drivers have not yet been ported to the new TTY layer:
1282 cy, digi, rc, rp, sio
1288 ng_h4, ng_tty, ppp, sl, snp
1290 Adding these drivers to your kernel configuration file shall
1291 cause compilation to fail.
1294 ntpd has been upgraded to 4.2.4p5.
1297 OpenSSH has been upgraded to 5.1p1.
1299 For many years, FreeBSD's version of OpenSSH preferred DSA
1300 over RSA for host and user authentication keys. With this
1301 upgrade, we've switched to the vendor's default of RSA over
1302 DSA. This may cause upgraded clients to warn about unknown
1303 host keys even for previously known hosts. Users should
1304 follow the usual procedure for verifying host keys before
1305 accepting the RSA key.
1307 This can be circumvented by setting the "HostKeyAlgorithms"
1308 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1311 Please note that the sequence of keys offered for
1312 authentication has been changed as well. You may want to
1313 specify IdentityFile in a different order to revert this
1317 The sio(4) driver has been removed from the i386 and amd64
1318 kernel configuration files. This means uart(4) is now the
1319 default serial port driver on those platforms as well.
1321 To prevent collisions with the sio(4) driver, the uart(4) driver
1322 uses different names for its device nodes. This means the
1323 onboard serial port will now most likely be called "ttyu0"
1324 instead of "ttyd0". You may need to reconfigure applications to
1325 use the new device names.
1327 When using the serial port as a boot console, be sure to update
1328 /boot/device.hints and /etc/ttys before booting the new kernel.
1329 If you forget to do so, you can still manually specify the hints
1330 at the loader prompt:
1332 set hint.uart.0.at="isa"
1333 set hint.uart.0.port="0x3F8"
1334 set hint.uart.0.flags="0x10"
1335 set hint.uart.0.irq="4"
1339 The gpt(8) utility has been removed. Use gpart(8) to partition
1343 The version that Linuxulator emulates was changed from 2.4.2
1344 to 2.6.16. If you experience any problems with Linux binaries
1345 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1346 if it fixes the problem contact emulation mailing list.
1349 ISDN4BSD (I4B) was removed from the src tree. You may need to
1350 update a your kernel configuration and remove relevant entries.
1353 I have checked in code to support multiple routing tables.
1354 See the man pages setfib(1) and setfib(2).
1355 This is a hopefully backwards compatible version,
1356 but to make use of it you need to compile your kernel
1357 with options ROUTETABLES=2 (or more up to 16).
1360 The 802.11 wireless support was redone to enable multi-bss
1361 operation on devices that are capable. The underlying device
1362 is no longer used directly but instead wlanX devices are
1363 cloned with ifconfig. This requires changes to rc.conf files.
1364 For example, change:
1365 ifconfig_ath0="WPA DHCP"
1368 ifconfig_wlan0="WPA DHCP"
1369 see rc.conf(5) for more details. In addition, mergemaster of
1370 /etc/rc.d is highly recommended. Simultaneous update of userland
1371 and kernel wouldn't hurt either.
1373 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1374 modules were merged into the base wlan module. All references
1375 to these modules (e.g. in kernel config files) must be removed.
1378 psm(4) has gained write(2) support in native operation level.
1379 Arbitrary commands can be written to /dev/psm%d and status can
1380 be read back from it. Therefore, an application is responsible
1381 for status validation and error recovery. It is a no-op in
1382 other operation levels.
1385 Support for KSE threading has been removed from the kernel. To
1386 run legacy applications linked against KSE libmap.conf may
1387 be used. The following libmap.conf may be used to ensure
1388 compatibility with any prior release:
1390 libpthread.so.1 libthr.so.1
1391 libpthread.so.2 libthr.so.2
1392 libkse.so.3 libthr.so.3
1395 The layout of struct vmspace has changed. This affects libkvm
1396 and any executables that link against libkvm and use the
1397 kvm_getprocs() function. In particular, but not exclusively,
1398 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1399 The effects are minimal, but it's advisable to upgrade world
1403 The latest em driver no longer has support in it for the
1404 82575 adapter, this is now moved to the igb driver. The
1405 split was done to make new features that are incompatible
1406 with older hardware easier to do.
1409 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1410 likewise the kernel option is now GEOM_LINUX_LVM.
1413 The default NFS mount mode has changed from UDP to TCP for
1414 increased reliability. If you rely on (insecurely) NFS
1415 mounting across a firewall you may need to update your
1419 Belatedly note the addition of m_collapse for compacting
1423 The fts(3) structures have been changed to use adequate
1424 integer types for their members and so to be able to cope
1425 with huge file trees. The old fts(3) ABI is preserved
1426 through symbol versioning in libc, so third-party binaries
1427 using fts(3) should still work, although they will not take
1428 advantage of the extended types. At the same time, some
1429 third-party software might fail to build after this change
1430 due to unportable assumptions made in its source code about
1431 fts(3) structure members. Such software should be fixed
1432 by its vendor or, in the worst case, in the ports tree.
1433 FreeBSD_version 800015 marks this change for the unlikely
1434 case that a portable fix is impossible.
1437 To upgrade to -current after this date, you must be running
1438 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1439 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1442 The ADAPTIVE_GIANT kernel option has been retired because its
1443 functionality is the default now.
1446 The AT keyboard emulation of sunkbd(4) has been turned on
1447 by default. In order to make the special symbols of the Sun
1448 keyboards driven by sunkbd(4) work under X these now have
1449 to be configured the same way as Sun USB keyboards driven
1450 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1452 Option "XkbLayout" "us"
1453 Option "XkbRules" "xorg"
1454 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1457 It has been decided that it is desirable to provide ABI
1458 backwards compatibility to the FreeBSD 4/5/6 versions of the
1459 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1460 broken with the introduction of PCI domain support (see the
1461 20070930 entry). Unfortunately, this required the ABI of
1462 PCIOCGETCONF to be broken again in order to be able to
1463 provide backwards compatibility to the old version of that
1464 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1465 again. As for prominent ports this affects neither pciutils
1466 nor xorg-server this time, the hal port needs to be rebuilt
1470 The misnamed kthread_create() and friends have been renamed
1471 to kproc_create() etc. Many of the callers already
1472 used kproc_start()..
1473 I will return kthread_create() and friends in a while
1474 with implementations that actually create threads, not procs.
1475 Renaming corresponds with version 800002.
1484 Avoid using make -j when upgrading. While generally safe, there are
1485 sometimes problems using -j to upgrade. If your upgrade fails with
1486 -j, please try again without -j. From time to time in the past there
1487 have been problems using -j with buildworld and/or installworld. This
1488 is especially true when upgrading between "distant" versions (eg one
1489 that cross a major release boundary or several minor releases, or when
1490 several months have passed on the -current branch).
1492 Sometimes, obscure build problems are the result of environment
1493 poisoning. This can happen because the make utility reads its
1494 environment when searching for values for global variables. To run
1495 your build attempts in an "environmental clean room", prefix all make
1496 commands with 'env -i '. See the env(1) manual page for more details.
1498 When upgrading from one major version to another it is generally best
1499 to upgrade to the latest code in the currently installed branch first,
1500 then do an upgrade to the new branch. This is the best-tested upgrade
1501 path, and has the highest probability of being successful. Please try
1502 this approach before reporting problems with a major version upgrade.
1506 When upgrading the boot ZFS pool to a new version, always follow
1509 1.) recompile and reinstall the ZFS boot loader and boot block
1510 (this is part of "make buildworld" and "make installworld")
1512 2.) update the ZFS boot block on your boot drive
1514 The following example updates the ZFS boot block on the first
1515 partition (freebsd-boot) of a GPT partitioned drive ad0:
1516 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1518 Non-boot pools do not need these updates.
1522 If you are updating from a prior version of FreeBSD (even one just
1523 a few days old), you should follow this procedure. It is the most
1524 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1526 make kernel-toolchain
1527 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1528 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1530 To test a kernel once
1531 ---------------------
1532 If you just want to boot a kernel once (because you are not sure
1533 if it works, or if you want to boot a known bad kernel to provide
1534 debugging information) run
1535 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1536 nextboot -k testkernel
1538 To just build a kernel when you know that it won't mess you up
1539 --------------------------------------------------------------
1540 This assumes you are already running a CURRENT system. Replace
1541 ${arch} with the architecture of your machine (e.g. "i386",
1542 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1544 cd src/sys/${arch}/conf
1545 config KERNEL_NAME_HERE
1546 cd ../compile/KERNEL_NAME_HERE
1551 If this fails, go to the "To build a kernel" section.
1553 To rebuild everything and install it on the current system.
1554 -----------------------------------------------------------
1555 # Note: sometimes if you are running current you gotta do more than
1556 # is listed here if you are upgrading from a really old current.
1558 <make sure you have good level 0 dumps>
1560 make kernel KERNCONF=YOUR_KERNEL_HERE
1562 <reboot in single user> [3]
1570 To cross-install current onto a separate partition
1571 --------------------------------------------------
1572 # In this approach we use a separate partition to hold
1573 # current's root, 'usr', and 'var' directories. A partition
1574 # holding "/", "/usr" and "/var" should be about 2GB in
1577 <make sure you have good level 0 dumps>
1580 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1581 <maybe newfs current's root partition>
1582 <mount current's root partition on directory ${CURRENT_ROOT}>
1583 make installworld DESTDIR=${CURRENT_ROOT}
1584 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1585 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1586 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1587 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1588 <reboot into current>
1589 <do a "native" rebuild/install as described in the previous section>
1590 <maybe install compatibility libraries from ports/misc/compat*>
1594 To upgrade in-place from stable to current
1595 ----------------------------------------------
1596 <make sure you have good level 0 dumps>
1598 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1600 <reboot in single user> [3]
1607 Make sure that you've read the UPDATING file to understand the
1608 tweaks to various things you need. At this point in the life
1609 cycle of current, things change often and you are on your own
1610 to cope. The defaults can also change, so please read ALL of
1611 the UPDATING entries.
1613 Also, if you are tracking -current, you must be subscribed to
1614 freebsd-current@freebsd.org. Make sure that before you update
1615 your sources that you have read and understood all the recent
1616 messages there. If in doubt, please track -stable which has
1617 much fewer pitfalls.
1619 [1] If you have third party modules, such as vmware, you
1620 should disable them at this point so they don't crash your
1623 [3] From the bootblocks, boot -s, and then do
1628 adjkerntz -i # if CMOS is wall time
1629 Also, when doing a major release upgrade, it is required that
1630 you boot into single user mode to do the installworld.
1632 [4] Note: This step is non-optional. Failure to do this step
1633 can result in a significant reduction in the functionality of the
1634 system. Attempting to do it by hand is not recommended and those
1635 that pursue this avenue should read this file carefully, as well
1636 as the archives of freebsd-current and freebsd-hackers mailing lists
1637 for potential gotchas. The -U option is also useful to consider.
1638 See mergemaster(8) for more information.
1640 [5] Usually this step is a noop. However, from time to time
1641 you may need to do this if you get unknown user in the following
1642 step. It never hurts to do it all the time. You may need to
1643 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1644 install) after the buildworld before this step if you last updated
1645 from [78]-stable or 9-stable before 20130430.
1647 [6] This only deletes old files and directories. Old libraries
1648 can be deleted by "make delete-old-libs", but you have to make
1649 sure that no program is using those libraries anymore.
1651 [8] In order to have a kernel that can run the 4.x binaries needed to
1652 do an installworld, you must include the COMPAT_FREEBSD4 option in
1653 your kernel. Failure to do so may leave you with a system that is
1654 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1655 required to run the 5.x binaries on more recent kernels. And so on
1656 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1658 Make sure that you merge any new devices from GENERIC since the
1659 last time you updated your kernel config file.
1661 [9] When checking out sources, you must include the -P flag to have
1662 cvs prune empty directories.
1664 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1665 "?=" instead of the "=" assignment operator, so that buildworld can
1666 override the CPUTYPE if it needs to.
1668 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1669 not on the command line, or in /etc/make.conf. buildworld will
1670 warn if it is improperly defined.
1673 This file contains a list, in reverse chronological order, of major
1674 breakages in tracking -current. It is not guaranteed to be a complete
1675 list of such breakages, and only contains entries since October 10, 2007.
1676 If you need to see UPDATING entries from before that date, you will need
1677 to fetch an UPDATING file from an older FreeBSD release.
1679 Copyright information:
1681 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1683 Redistribution, publication, translation and use, with or without
1684 modification, in full or in part, in any form or format of this
1685 document are permitted without further permission from the author.
1687 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1688 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1689 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1690 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1691 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1692 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1693 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1694 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1695 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1696 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1697 POSSIBILITY OF SUCH DAMAGE.
1699 Contact Warner Losh if you have any questions about your use of