1 Updating Information for FreeBSD current users
3 This file is maintained and copyrighted by M. Warner Losh <imp@freebsd.org>.
4 See end of file for further details. For commonly done items, please see the
5 COMMON ITEMS: section later in the file. These instructions assume that you
6 basically know what you are doing. If not, then please consult the FreeBSD
9 Items affecting the ports and packages system can be found in
10 /usr/ports/UPDATING. Please read that file before running portupgrade.
12 20141223: p24 FreeBSD-SA-14:31.ntp
13 FreeBSD-EN-14:13.freebsd-update
15 Fix multiple vulnerabilities in NTP suite. [SA-14:31]
16 Fix directory deletion issue in freebsd-update. [EN-14:13]
18 20141210: p23 FreeBSD-SA-14:28.file
21 Fix multiple vulnerabilities in file(1) and libmagic(3).
24 Fix BIND remote denial of service vulnerability. [SA-14:29]
26 20141104: p22 FreeBSD-SA-14:24.sshd
27 FreeBSD-SA-14:25.setlogin
31 Fix denial of service attack against sshd(8). [SA-14:24]
33 Fix kernel stack disclosure in setlogin(2) / getlogin(2).
36 Fix remote command execution in ftp(1). [SA-14:26]
38 Fix NFSv4 and ZFS cache consistency issue. [EN-14:12]
40 20141022: p21 FreeBSD-EN-14:10.tzdata
42 Time zone data file update. [EN-14:10]
44 20141021: p20 FreeBSD-SA-14:20.rtsold
45 FreeBSD-SA-14:21.routed
46 FreeBSD-SA-14:22.namei
47 FreeBSD-SA-14:23.openssl
49 Fix rtsold(8) remote buffer overflow vulnerability. [SA-14:20]
51 Fix routed(8) remote denial of service vulnerability. [SA-14:21]
53 Fix memory leak in sandboxed namei lookup. [SA-14:22]
55 Fix OpenSSL multiple vulnerabilities. [SA-14:23]
57 20140916: p19 FreeBSD-SA-14:19.tcp
58 Fix Denial of Service in TCP packet processing. [SA-14:19]
60 20140909: p18 FreeBSD-SA-14:18.openssl
61 Fix OpenSSL multiple vulnerabilities. [SA-14:18]
63 20140708: p17 FreeBSD-SA-14:17.kmem
64 Fix kernel memory disclosure in control messages and SCTP
65 notifications. [SA-14:17]
67 20140624: p16 FreeBSD-SA-14:16.file
68 FreeBSD-EN-14:08.heimdal
70 Fix multiple vulnerabilities in file(1) and libmagic(3).
73 Fix gss_pseudo_random interoperability issue. [EN-14:08]
75 20140605: p15 FreeBSD-SA-14:14.openssl
76 Fix OpenSSL multiple vulnerabilities. [SA-14:14]
78 20140603: p14 FreeBSD-SA-14:11.sendmail
79 FreeBSD-SA-14:12.ktrace
83 Fix sendmail improper close-on-exec flag handling. [SA-14:11]
85 Fix ktrace memory disclosure. [SA-14:12]
87 Fix incorrect error handling in PAM policy parser. [SA-14:13]
89 Fix triple-fault when executing from a threaded process.
92 20140513: p13 FreeBSD-EN-14:03.pkg
93 FreeBSD-EN-14:04.kldxref
96 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
98 Improve build repeatability for kldxref(8). [EN-14:04]
100 Fix data corruption with ciss(4). [EN-14:05]
102 20140430: p12 FreeBSD-SA-14:08.tcp
104 Fix TCP reassembly vulnerability. [SA-14:08]
106 20140408: p11 FreeBSD-SA-14:05.nfsserver
107 FreeBSD-SA-14:06.openssl
108 Fix deadlock in the NFS server. [SA-14:05]
110 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
112 20140114: p10 FreeBSD-SA-14:01.bsnmpd
113 FreeBSD-SA-14:02.ntpd
114 FreeBSD-SA-14:04.bind
115 FreeBSD-EN-14:01.random
116 FreeBSD-EN-14:02.mmap
117 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
119 Fix ntpd distributed reflection Denial of Service
120 vulnerability. [SA-14:02]
122 Fix BIND remote denial of service vulnerability. [SA-14:04]
124 Disable hardware RNGs by default. [EN-14:01]
126 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
128 20131128: p9 FreeBSD-EN-13:05.freebsd-update
129 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
131 20131026: p8 FreeBSD-EN-13:04.freebsd-update
132 Fix multiple freebsd-update bugs that break upgrading to
135 20130910: p7 FreeBSD-SA-13:12.ifioctl
136 FreeBSD-SA-13:13.nullfs
138 In IPv6 and NetATM, stop SIOCSIFADDR, SIOCSIFBRDADDR,
139 SIOCSIFDSTADDR and SIOCSIFNETMASK at the socket layer rather
140 than pass them on to the link layer without validation or
141 credential checks. [SA-13:12]
143 Prevent cross-mount hardlinks between different nullfs mounts
144 of the same underlying filesystem. [SA-13:13]
146 20130822: p6 FreeBSD-SA-13:09.ip_multicast
147 FreeBSD-SA-13:10.sctp
149 Fix an integer overflow in computing the size of a temporary buffer
150 can result in a buffer which is too small for the requested
153 Fix a bug that could lead to kernel memory disclosure with
154 SCTP state cookie. [13:10]
156 Fix a data corruption problem with mfi(4) operating on > 2TB
157 disks in a JBOD. [EN-13:03]
159 20130726: p5 FreeBSD-SA-13:07.bind FreeBSD-SA-13:08.nfsserver
160 Fix Denial of Service vulnerability in named(8). [13:07]
162 Fix a bug that allows remote client bypass the normal
163 access checks when when -network or -host restrictions are
164 used at the same time with -mapall. [13:08]
166 20130618: p4 FreeBSD-SA-13:06.mmap
167 Fix a bug that allowed a tracing process (e.g. gdb) to write
168 to a memory-mapped file in the traced process's address space
169 even if neither the traced process nor the tracing process had
170 write access to that file.
172 20130429: p3 FreeBSD-SA-13:05.nfsserver
173 Fix a bug that allows NFS clients to issue READDIR on files.
175 20130402: p2 FreeBSD-SA-13:03.openssl FreeBSD-SA-13:04.bind
176 Fix multiple vulnerabilities in OpenSSL.
178 Fix Denial of Service vulnerability in named(8).
180 20130218: p1 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
181 Fix Denial of Service vulnerability in named(8) with DNS64.
183 Fix Denial of Service vulnerability in libc's glob(3) functionality.
189 The sparc64 ZFS loader has been changed to no longer try to auto-
190 detect ZFS providers based on diskN aliases but now requires these
191 to be explicitly listed in the OFW boot-device environment variable.
194 Now unix domain sockets behave "as expected" on nullfs(5). Previously
195 nullfs(5) did not pass through all behaviours to the underlying layer,
196 as a result if we bound to a socket on the lower layer we could connect
197 only to the lower path; if we bound to the upper layer we could connect
198 only to the upper path. The new behavior is one can connect to both the
199 lower and the upper paths regardless what layer path one binds to.
202 The acpi_wmi(4) status device /dev/wmistat has been renamed to
206 A new VOP_ADVISE() was added to support posix_fadvise(2). All
207 filesystem modules must be recompiled.
210 The interface of the VOP_VPTOCNP(9) changed, now the returned
211 vnode shall be referenced, previously it was required to be
212 only held. All in-tree filesystems are converted.
218 The broken amd(4) driver has been replaced with esp(4) in the amd64,
219 i386 and pc98 GENERIC kernel configuration files.
222 This commit modifies vfs_register() so that it uses a hash
223 calculation to set vfc_typenum, which is enabled by default.
224 The first time a system is booted after this change, the
225 vfc_typenum values will change for all file systems. The
226 main effect of this is a change to the NFS server file handles
227 for file systems that use vfc_typenum in their fsid, such as ZFS.
228 It will, however, prevent vfc_typenum from changing when file
229 systems are loaded in a different order for subsequent reboots.
230 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
231 until you are ready to remount all NFS clients after a reboot.
234 Bump the shared library version numbers for libraries that
235 do not use symbol versioning, have changed the ABI compared
236 to stable/8 and which shared library version was not bumped.
237 Done as part of 9.0-RELEASE cycle.
240 During the merge of Capsicum features, the fget(9) KPI was modified.
241 This may require the rebuilding of out-of-tree device drivers --
242 issues have been reported specifically with the nVidia device driver.
243 __FreeBSD_version is bumped to 900041.
245 Also, there is a period between 20110811 and 20110814 where the
246 special devices /dev/{stdin,stdout,stderr} did not work correctly.
247 Building world from a kernel during that window may not work.
250 The packet filter (pf) code has been updated to OpenBSD 4.5.
251 You need to update userland tools to be in sync with kernel.
252 This update breaks backward compatibility with earlier pfsync(4)
253 versions. Care must be taken when updating redundant firewall setups.
256 The following sysctls and tunables are retired on x86 platforms:
258 machdep.hlt_logical_cpus
259 The following sysctl is retired:
260 machdep.hyperthreading_allowed
261 The sysctls were supposed to provide a way to dynamically offline and
262 online selected CPUs on x86 platforms, but the implementation has not
263 been reliable especially with SCHED_ULE scheduler.
264 machdep.hyperthreading_allowed tunable is still available to ignore
265 hyperthreading CPUs at OS level.
266 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
267 where X is an APIC ID of a CPU. Be advised, though, that disabling
268 CPUs in non-uniform fashion will result in non-uniform topology and
269 may lead to sub-optimal system performance with SCHED_ULE, which is
273 cpumask_t type is retired and cpuset_t is used in order to describe
277 Changes to ifconfig(8) for dynamic address family detection mandate
278 that you are running a kernel of 20110525 or later. Make sure to
279 follow the update procedure to boot a new kernel before installing
283 Support for sun4v architecture is officially dropped
286 Several KPI breaking changes have been committed to the mii(4) layer,
287 the PHY drivers and consequently some Ethernet drivers using mii(4).
288 This means that miibus.ko and the modules of the affected Ethernet
289 drivers need to be recompiled.
291 Note to kernel developers: Given that the OUI bit reversion problem
292 was fixed as part of these changes all mii(4) commits related to OUIs,
293 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
294 handling, no longer can be merged verbatim to stable/8 and previous
298 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
299 into their kernel configurations along with 'device pci'.
302 The default NFS client is now the new NFS client, so fstype "newnfs"
303 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
304 Although mounts via fstype "nfs" will usually work without userland
305 changes, it is recommended that the mount(8) and mount_nfs(8)
306 commands be rebuilt from sources and that a link to mount_nfs called
307 mount_oldnfs be created. The new client is compiled into the
308 kernel with "options NFSCL" and this is needed for diskless root
309 file systems. The GENERIC kernel configs have been changed to use
310 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
311 To use the regular/old client, you can "mount -t oldnfs ...". For
312 a diskless root file system, you must also include a line like:
314 vfs.root.mountfrom="oldnfs:"
316 in the boot/loader.conf on the root fs on the NFS server to make
317 a diskless root fs use the old client.
320 The GENERIC kernels for all architectures now default to the new
321 CAM-based ATA stack. It means that all legacy ATA drivers were
322 removed and replaced by respective CAM drivers. If you are using
323 ATA device names in /etc/fstab or other places, make sure to update
324 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
325 where 'Y's are the sequential numbers starting from zero for each type
326 in order of detection, unless configured otherwise with tunables,
327 see cam(4)). There will be symbolic links created in /dev/ to map
328 old adX devices to the respective adaY. They should provide basic
329 compatibility for file systems mounting in most cases, but they do
330 not support old user-level APIs and do not have respective providers
331 in GEOM. Consider using updated management tools with new device names.
333 It is possible to load devices ahci, ata, siis and mvs as modules,
334 but option ATA_CAM should remain in kernel configuration to make ata
335 module work as CAM driver supporting legacy ATA controllers. Device ata
336 still can be used in modular fashion (atacore + ...). Modules atadisk
337 and atapi* are not used and won't affect operation in ATA_CAM mode.
338 Note that to use CAM-based ATA kernel should include CAM devices
339 scbus, pass, da (or explicitly ada), cd and optionally others. All of
340 them are parts of the cam module.
342 ataraid(4) functionality is now supported by the RAID GEOM class.
343 To use it you can load geom_raid kernel module and use graid(8) tool
344 for management. Instead of /dev/arX device names, use /dev/raid/rX.
346 No kernel config options or code have been removed, so if a problem
347 arises, please report it and optionally revert to the old ATA stack.
348 In order to do it you can remove from the kernel config:
353 , and instead add back:
354 device atadisk # ATA disk drives
355 device ataraid # ATA RAID drives
356 device atapicd # ATAPI CDROM drives
357 device atapifd # ATAPI floppy drives
358 device atapist # ATAPI tape drives
361 The default NFS server has been changed to the new server, which
362 was referred to as the experimental server. If you need to switch
363 back to the old NFS server, you must now put the "-o" option on
364 both the mountd and nfsd commands. This can be done using the
365 mountd_flags and nfs_server_flags rc.conf variables until an
366 update to the rc scripts is committed, which is coming soon.
369 The GNU Objective-C runtime library (libobjc), and other Objective-C
370 related components have been removed from the base system. If you
371 require an Objective-C library, please use one of the available ports.
374 ath(4) has been split into bus- and device- modules. if_ath contains
375 the HAL, the TX rate control and the network device code. if_ath_pci
376 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
377 contains the AHB glue. Users need to load both if_ath_pci and if_ath
378 in order to use ath on everything else.
380 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
381 need to load if_ath and if_ath_pci for ath(4) operation.
384 As part of the replacement of sysinstall, the process of building
385 release media has changed significantly. For details, please re-read
386 release(7), which has been updated to reflect the new build process.
389 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
390 is the last available version under GPLv2. It brings a number of new
391 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
392 4.1 and SSE 4.2), better support for powerpc64, a number of new
393 directives, and lots of other small improvements. See the ChangeLog
394 file in contrib/binutils for the full details.
397 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
398 compliant, and will now use half of hash for authentication.
399 This will break interoperability with all stacks (including all
400 actual FreeBSD versions) who implement
401 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
403 The only workaround with such peers is to use another HMAC
404 algorithm for IPsec ("phase 2") authentication.
407 Remove the uio_yield prototype and symbol. This function has
408 been misnamed since it was introduced and should not be
409 globally exposed with this name. The equivalent functionality
410 is now available using kern_yield(curthread->td_user_pri).
411 The function remains undocumented.
414 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
415 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
416 for scalar sysctls is defined but disabled. Code that needs
417 UQUAD to pass the type checking that must compile on older
418 systems where the define is not present can check against
419 __FreeBSD_version >= 900030.
421 The system dialog(1) has been replaced with a new version previously
422 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
423 with the previous version, but the libdialog associated with it has
424 a largely incompatible API. As such, the original version of libdialog
425 will be kept temporarily as libodialog, until its base system consumers
426 are replaced or updated. Bump __FreeBSD_version to 900030.
429 If you are trying to run make universe on a -stable system, and you get
430 the following warning:
431 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
432 unknown. config(8) likely too old."
433 or something similar to it, then you must upgrade your -stable system
434 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
435 stable/8) or build the config from the latest stable/8 branch and
436 install it on your system.
438 Prior to this date, building a current universe on 8-stable system from
439 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
440 error in the first kernel build phase. A new config on those old
441 systems will fix that problem for older versions of -current.
444 The TCP stack has been modified to allow Khelp modules to interact with
445 it via helper hook points and store per-connection data in the TCP
446 control block. Bump __FreeBSD_version to 900029. User space tools that
447 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
451 Generic IEEE 802.3 annex 31B full duplex flow control support has been
452 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
453 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
454 to take advantage of it instead of using custom implementations. This
455 means that these drivers now no longer unconditionally advertise
456 support for flow control but only do so if flow control is a selected
457 media option. This was implemented in the generic support that way in
458 order to allow flow control to be switched on and off via ifconfig(8)
459 with the PHY specific default to typically off in order to protect
460 from unwanted effects. Consequently, if you used flow control with
461 one of the above mentioned drivers you now need to explicitly enable
463 ifconfig bge0 media auto mediaopt flowcontrol
465 Along with the above mentioned changes generic support for setting
466 1000baseT master mode also has been added and brgphy(4), ciphy(4),
467 e1000phy(4) as well as ip1000phy(4) have been converted to take
468 advantage of it. This means that these drivers now no longer take the
469 link0 parameter for selecting master mode but the master media option
470 has to be used instead, for example like in the following:
471 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
473 Selection of master mode now is also available with all other PHY
474 drivers supporting 1000baseT.
477 The TCP stack has received a significant update to add support for
478 modularised congestion control and generally improve the clarity of
479 congestion control decisions. Bump __FreeBSD_version to 900025. User
480 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
481 sockstat) need to be recompiled.
484 The man(1) utility has been replaced by a new version that no longer
485 uses /etc/manpath.config. Please consult man.conf(5) for how to
486 migrate local entries to the new format.
489 The copyright strings printed by login(1) and sshd(8) at the time of a
490 new connection have been removed to follow other operating systems and
494 A workaround for a fixed ld bug has been removed in kernel code,
495 so make sure that your system ld is built from sources after
496 revision 210245 from 2010-07-19 (r211583 if building head kernel
497 on stable/8, r211584 for stable/7; both from 2010-08-21).
498 A symptom of incorrect ld version is different addresses for
499 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
502 The $ipv6_prefer variable in rc.conf(5) has been split into
503 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
505 The $ip6addrctl_policy is a variable to choose a pre-defined
506 address selection policy set by ip6addrctl(8). A value
507 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
510 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
511 flag (see an entry of 20090926) is set on an interface with no
512 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
513 security reason. If you want IPv6 link-local address on all
514 interfaces by default, set this to "YES".
516 The old ipv6_prefer="YES" is equivalent to
517 ipv6_activate_all_interfaces="YES" and
518 ip6addrctl_policy="ipv6_prefer".
521 DTrace has grown support for userland tracing. Due to this, DTrace is
522 now i386 and amd64 only.
523 dtruss(1) is now installed by default on those systems and a new
524 kernel module is needed for userland tracing: fasttrap.
525 No changes to your kernel config file are necessary to enable
526 userland tracing, but you might consider adding 'STRIP=' and
527 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
528 to have informative userland stack traces in DTrace (ustack).
531 The acpi_aiboost(4) driver has been removed in favor of the new
532 aibs(4) driver. You should update your kernel configuration file.
535 BSD grep has been imported to the base system and it is built by
536 default. It is completely BSD licensed, highly GNU-compatible, uses
537 less memory than its GNU counterpart and has a small codebase.
538 However, it is slower than its GNU counterpart, which is mostly
539 noticeable for larger searches, for smaller ones it is measurable
540 but not significant. The reason is complex, the most important factor
541 is that we lack a modern and efficient regex library and GNU
542 overcomes this by optimizing the searches internally. Future work
543 on improving the regex performance is planned, for the meantime,
544 users that need better performance, can build GNU grep instead by
545 setting the WITH_GNU_GREP knob.
548 Due to the import of powerpc64 support, all existing powerpc kernel
549 configuration files must be updated with a machine directive like this:
550 machine powerpc powerpc
552 In addition, an updated config(8) is required to build powerpc kernels
556 A new version of ZFS (version 15) has been merged to -HEAD.
557 This version uses a python library for the following subcommands:
558 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
559 For full functionality of these commands the following port must
560 be installed: sysutils/py-zfs
563 'vm_page's are now hashed by physical address to an array of mutexes.
564 Currently this is only used to serialize access to hold_count. Over
565 time the page queue mutex will be peeled away. This changes the size
566 of pmap on every architecture. And requires all callers of vm_page_hold
567 and vm_page_unhold to be updated.
570 WITH_CTF can now be specified in src.conf (not recommended, there
571 are some problems with static executables), make.conf (would also
572 affect ports which do not use GNU make and do not override the
573 compile targets) or in the kernel config (via "makeoptions
575 When WITH_CTF was specified there before this was silently ignored,
576 so make sure that WITH_CTF is not used in places which could lead
577 to unwanted behavior.
580 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
581 to allow 32-bit compatibility on non-x86 platforms. All kernel
582 configurations on amd64 and ia64 platforms using these options must
583 be modified accordingly.
586 The utmp user accounting database has been replaced with utmpx,
587 the user accounting interface standardized by POSIX.
588 Unfortunately the semantics of utmp and utmpx don't match,
589 making it practically impossible to support both interfaces.
590 The user accounting database is used by tools like finger(1),
591 last(1), talk(1), w(1) and ac(8).
593 All applications in the base system use utmpx. This means only
594 local binaries (e.g. from the ports tree) may still use these
595 utmp database files. These applications must be rebuilt to make
598 After the system has been upgraded, it is safe to remove the old
599 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
600 assuming their contents is of no importance anymore. Old wtmp
601 databases can only be used by last(1) and ac(8) after they have
602 been converted to the new format using wtmpcvt(1).
605 Introduce the kernel thread "deadlock resolver" (which can be enabled
606 via the DEADLKRES option, see NOTES for more details) and the
607 sleepq_type() function for sleepqueues.
610 The rc.firewall and rc.firewall6 were unified, and
611 rc.firewall6 and rc.d/ip6fw were removed.
612 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
613 variables are obsoleted. Instead, the following new rc
614 variables are added to rc.d/ipfw:
616 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
617 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
618 firewall_simple_onet_ipv6, firewall_trusted_ipv6
620 The meanings correspond to the relevant IPv4 variables.
626 The default terminal emulation for syscons(4) has been changed
627 from cons25 to xterm on all platforms except pc98. This means
628 that the /etc/ttys file needs to be updated to ensure correct
629 operation of applications on the console.
631 The terminal emulation style can be toggled per window by using
632 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
633 options can be used to change the compile-time default back to
636 To prevent graphical artifacts, make sure the TERM environment
637 variable is set to match the terminal emulation that is being
638 performed by syscons(4).
641 The layout of the structure ieee80211req_scan_result has changed.
642 Applications that require wireless scan results (e.g. ifconfig(8))
643 from net80211 need to be recompiled.
645 Applications such as wpa_supplicant(8) may require a full world
646 build without using NO_CLEAN in order to get synchronized with the
650 The iwn(4) driver has been updated to support the 5000 and 5150 series.
651 There's one kernel module for each firmware. Adding "device iwnfw"
652 to the kernel configuration file means including all three firmware
653 images inside the kernel. If you want to include just the one for
654 your wireless card, use the devices iwn4965fw, iwn5000fw or
658 The rc.d/network_ipv6, IPv6 configuration script has been integrated
659 into rc.d/netif. The changes are the following:
661 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
662 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
663 Note that both variables need the "inet6" keyword at the head.
665 Do not set $ipv6_network_interfaces manually if you do not
666 understand what you are doing. It is not needed in most cases.
668 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
671 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
672 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
674 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
675 all configured interfaces have "inet6 accept_rtadv" in the
676 $ifconfig_IF_ipv6. These are for backward compatibility.
678 3. A new variable $ipv6_prefer has been added. If NO, IPv6
679 functionality of interfaces with no corresponding
680 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
681 and the default address selection policy of ip6addrctl(8)
682 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
683 Note that if you want to configure IPv6 functionality on the
684 disabled interfaces after boot, first you need to clear the flag by
685 using ifconfig(8) like:
687 ifconfig em0 inet6 -ifdisabled
689 If YES, the default address selection policy is set as
692 The default value of $ipv6_prefer is NO.
694 4. If your system need to receive Router Advertisement messages,
695 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
696 scripts automatically invoke rtsol(8) when the interface becomes
697 UP. The Router Advertisement messages are used for SLAAC
698 (State-Less Address AutoConfiguration).
701 802.11s D3.03 support was committed. This is incompatible with the
702 previous code, which was based on D3.0.
705 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
706 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
707 control whether accepting Router Advertisement messages or not.
708 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
709 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
710 The ifconfig(8) utility now supports these flags.
713 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
714 mount(8) and -a option for df(1) to see them.
717 The old tunable hw.bus.devctl_disable has been superseded by
718 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
719 replaced by hw.bus.devctl_queue=0. The default for this new tunable
723 Remove the option STOP_NMI. The default action is now to use NMI only
724 for KDB via the newly introduced function stop_cpus_hard() and
725 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
728 The stable/8 branch created in subversion. This corresponds to the
729 RELENG_8 branch in CVS.
732 Bump the shared library version numbers for all libraries that do not
733 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
734 __FreeBSD_version to 800105.
737 Due to changes in the implementation of virtual network stack support,
738 all network-related kernel modules must be recompiled. As this change
739 breaks the ABI, bump __FreeBSD_version to 800104.
742 The TOE interface to the TCP syncache has been modified to remove
743 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
744 The cxgb driver is the only TOE consumer affected by this change, and
745 needs to be recompiled along with the kernel. As this change breaks
746 the ABI, bump __FreeBSD_version to 800103.
749 Padding has been added to struct tcpcb, sackhint and tcpstat in
750 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
751 maintaining the ABI. However, this change breaks the ABI, so bump
752 __FreeBSD_version to 800102. User space tools that rely on the size of
753 any of these structs (e.g. sockstat) need to be recompiled.
756 The NFS_LEGACYRPC option has been removed along with the old kernel
757 RPC implementation that this option selected. Kernel configurations
758 may need to be adjusted.
761 The network interface device nodes at /dev/net/<interface> have been
762 removed. All ioctl operations can be performed the normal way using
763 routing sockets. The kqueue functionality can generally be replaced
764 with routing sockets.
767 The documentation from the FreeBSD Documentation Project (Handbook,
768 FAQ, etc.) is now installed via packages by sysinstall(8) and under
769 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
772 The ABI of various structures related to the SYSV IPC API have been
773 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
774 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
778 Layout of struct vnet has changed as routing related variables were
779 moved to their own Vimage module. Modules need to be recompiled. Bump
780 __FreeBSD_version to 800099.
783 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
784 respectively. As long as no more than 16 groups per process are used,
785 no changes should be visible. When more than 16 groups are used, old
786 binaries may fail if they call getgroups() or getgrouplist() with
787 statically sized storage. Recompiling will work around this, but
788 applications should be modified to use dynamically allocated storage
789 for group arrays as POSIX.1-2008 does not cap an implementation's
790 number of supported groups at NGROUPS_MAX+1 as previous versions did.
792 NFS and portalfs mounts may also be affected as the list of groups is
793 truncated to 16. Users of NFS who use more than 16 groups, should
794 take care that negative group permissions are not used on the exported
795 file systems as they will not be reliable unless a GSSAPI based
796 authentication method is used.
799 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
800 option compiles in the support for adaptive spinning for lockmgrs
801 which want to enable it. The lockinit() function now accepts the flag
802 LK_ADAPTIVE in order to make the lock object subject to adaptive
803 spinning when both held in write and read mode.
806 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
807 changed. User applications that use this ioctl need to be rebuilt.
810 The layout of struct thread has changed. Kernel and modules need to
814 The layout of structs ifnet, domain, protosw and vnet_net has changed.
815 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
818 window(1) has been removed from the base system. It can now be
819 installed from ports. The port is called misc/window.
822 The way we are storing and accessing `routing table' entries has
823 changed. Programs reading the FIB, like netstat, need to be
827 A new netisr implementation has been added for FreeBSD 8. Network
828 file system modules, such as igmp, ipdivert, and others, should be
830 Bump __FreeBSD_version to 800096.
833 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
837 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
838 Bump __FreeBSD_version to 800094.
841 Add mnt_xflag field to 'struct mount'. File system modules need to be
843 Bump __FreeBSD_version to 800093.
846 The compiling option ADAPTIVE_SX has been retired while it has been
847 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
848 The KPI for sx_init_flags() changes as accepting flags:
849 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
850 been introduced in order to handle the reversed logic.
851 Bump __FreeBSD_version to 800092.
854 Add support for hierarchical jails. Remove global securelevel.
855 Bump __FreeBSD_version to 800091.
858 The layout of struct vnet_net has changed, therefore modules
860 Bump __FreeBSD_version to 800090.
863 The newly imported zic(8) produces a new format in the output. Please
864 run tzsetup(8) to install the newly created data to /etc/localtime.
867 The sysctl tree for the usb stack has renamed from hw.usb2.* to
868 hw.usb.* and is now consistent again with previous releases.
871 802.11 monitor mode support was revised and driver api's were changed.
872 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
873 of DLT_IEEE802_11. No user-visible data structures were changed but
874 applications that use DLT_IEEE802_11 may require changes.
875 Bump __FreeBSD_version to 800088.
878 The layout of the following structs has changed: sysctl_oid,
879 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
880 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
881 panics may be experienced. World rebuild is required for
882 correctly checking networking state from userland.
883 Bump __FreeBSD_version to 800085.
886 MLDv2 and Source-Specific Multicast (SSM) have been merged
887 to the IPv6 stack. VIMAGE hooks are in but not yet used.
888 The implementation of SSM within FreeBSD's IPv6 stack closely
889 follows the IPv4 implementation.
891 For kernel developers:
893 * The most important changes are that the ip6_output() and
894 ip6_input() paths no longer take the IN6_MULTI_LOCK,
895 and this lock has been downgraded to a non-recursive mutex.
897 * As with the changes to the IPv4 stack to support SSM, filtering
898 of inbound multicast traffic must now be performed by transport
899 protocols within the IPv6 stack. This does not apply to TCP and
900 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
902 * The KPIs used by IPv6 multicast are similar to those used by
903 the IPv4 stack, with the following differences:
904 * im6o_mc_filter() is analogous to imo_multicast_filter().
905 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
906 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
907 * IN6_LOOKUP_MULTI() has been deprecated and removed.
908 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
909 for MLDv1 have an additional 'timer' argument which is used to
910 jitter the initial membership report for the solicited-node
911 multicast membership on-link.
912 * This is not strictly needed for MLDv2, which already jitters
913 its report transmissions. However, the 'timer' argument is
914 preserved in case MLDv1 is active on the interface.
916 * The KAME linked-list based IPv6 membership implementation has
917 been refactored to use a vector similar to that used by the IPv4
919 Code which maintains a list of its own multicast memberships
920 internally, e.g. carp, has been updated to reflect the new
923 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
924 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
925 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
926 implementation constraint which needs to be addressed in HEAD.
928 For application developers:
930 * The changes are broadly similar to those made for the IPv4
933 * The use of IPv4 and IPv6 multicast socket options on the same
934 socket, using mapped addresses, HAS NOT been tested or supported.
936 * There are a number of issues with the implementation of various
937 IPv6 multicast APIs which need to be resolved in the API surface
938 before the implementation is fully compatible with KAME userland
939 use, and these are mostly to do with interface index treatment.
941 * The literature available discusses the use of either the delta / ASM
942 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
943 using setsourcefilter(3)/getsourcefilter(3). For more information
944 please refer to RFC 3768, 'Socket Interface Extensions for
945 Multicast Source Filters'.
947 * Applications which use the published RFC 3678 APIs should be fine.
949 For systems administrators:
951 * The mtest(8) utility has been refactored to support IPv6, in
952 addition to IPv4. Interface addresses are no longer accepted
953 as arguments, their names must be used instead. The utility
954 will map the interface name to its first IPv4 address as
955 returned by getifaddrs(3).
957 * The ifmcstat(8) utility has also been updated to print the MLDv2
958 endpoint state and source filter lists via sysctl(3).
960 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
961 loopback of IPv6 multicast datagrams by default; it defaults to 1
962 to preserve the existing behaviour. Disabling multicast loopback is
963 recommended for optimal system performance.
965 * The IPv6 MROUTING code has been changed to examine this sysctl
966 instead of attempting to perform a group lookup before looping
967 back forwarded datagrams.
969 Bump __FreeBSD_version to 800084.
972 Implement low-level Bluetooth HCI API.
973 Bump __FreeBSD_version to 800083.
976 The layout of struct malloc_type, used by modules to register new
977 memory allocation types, has changed. Most modules will need to
978 be rebuilt or panics may be experienced.
979 Bump __FreeBSD_version to 800081.
982 Anticipate overflowing inp_flags - add inp_flags2.
983 This changes most offsets in inpcb, so checking v4 connection
984 state will require a world rebuild.
985 Bump __FreeBSD_version to 800080.
988 Add an llentry to struct route and struct route_in6. Modules
989 embedding a struct route will need to be recompiled.
990 Bump __FreeBSD_version to 800079.
993 The size of rt_metrics_lite and by extension rtentry has changed.
994 Networking administration apps will need to be recompiled.
995 The route command now supports show as an alias for get, weighting
996 of routes, sticky and nostick flags to alter the behavior of stateful
998 Bump __FreeBSD_version to 800078.
1001 Do not use Giant for kbdmux(4) locking. This is wrong and
1002 apparently causing more problems than it solves. This will
1003 re-open the issue where interrupt handlers may race with
1004 kbdmux(4) in polling mode. Typical symptoms include (but
1005 not limited to) duplicated and/or missing characters when
1006 low level console functions (such as gets) are used while
1007 interrupts are enabled (for example geli password prompt,
1008 mountroot prompt etc.). Disabling kbdmux(4) may help.
1011 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
1012 kernel modules referencing any of the above need to be recompiled.
1013 Bump __FreeBSD_version to 800075.
1016 GEOM_PART has become the default partition slicer for storage devices,
1017 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
1018 introduces some changes:
1020 MSDOS/EBR: the devices created from MSDOS extended partition entries
1021 (EBR) can be named differently than with GEOM_MBR and are now symlinks
1022 to devices with offset-based names. fstabs may need to be modified.
1024 BSD: the "geometry does not match label" warning is harmless in most
1025 cases but it points to problems in file system misalignment with
1026 disk geometry. The "c" partition is now implicit, covers the whole
1027 top-level drive and cannot be (mis)used by users.
1029 General: Kernel dumps are now not allowed to be written to devices
1030 whose partition types indicate they are meant to be used for file
1031 systems (or, in case of MSDOS partitions, as something else than
1034 Most of these changes date approximately from 200812.
1037 The uscanner(4) driver has been removed from the kernel. This follows
1038 Linux removing theirs in 2.6 and making libusb the default interface
1039 (supported by sane).
1042 The multicast forwarding code has been cleaned up. netstat(1)
1043 only relies on KVM now for printing bandwidth upcall meters.
1044 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1045 ip6_mroute_mod respectively. The config(5) options for statically
1046 compiling this code remain the same, i.e. 'options MROUTING'.
1049 Support for the IFF_NEEDSGIANT network interface flag has been
1050 removed, which means that non-MPSAFE network device drivers are no
1051 longer supported. In particular, if_ar, if_sr, and network device
1052 drivers from the old (legacy) USB stack can no longer be built or
1056 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1057 a bunch of new language catalog files have also been added.
1058 This means that some common libc messages are now localized and
1059 they depend on the LC_MESSAGES environmental variable.
1062 The k8temp(4) driver has been renamed to amdtemp(4) since
1063 support for Family 10 and Family 11 CPU families was added.
1066 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1067 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1069 For kernel developers, the most important changes are that the
1070 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1071 and this lock has been downgraded to a non-recursive mutex.
1073 Transport protocols (UDP, Raw IP) are now responsible for filtering
1074 inbound multicast traffic according to group membership and source
1075 filters. The imo_multicast_filter() KPI exists for this purpose.
1076 Transports which do not use multicast (SCTP, TCP) already reject
1077 multicast by default. Forwarding and receive performance may improve
1078 as a mutex acquisition is no longer needed in the ip_input()
1079 low-level input path. in_addmulti() and in_delmulti() are shimmed
1080 to new KPIs which exist to support SSM in-kernel.
1082 For application developers, it is recommended that loopback of
1083 multicast datagrams be disabled for best performance, as this
1084 will still cause the lock to be taken for each looped-back
1085 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1086 be tuned to 0 to disable loopback by default; it defaults to 1
1087 to preserve the existing behaviour.
1089 For systems administrators, to obtain best performance with
1090 multicast reception and multiple groups, it is always recommended
1091 that a card with a suitably precise hash filter is used. Hash
1092 collisions will still result in the lock being taken within the
1093 transport protocol input path to check group membership.
1095 If deploying FreeBSD in an environment with IGMP snooping switches,
1096 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1097 enabled; this forces 224.0.0.0/24 group membership to be announced
1100 The size of 'struct igmpstat' has changed; netstat needs to be
1101 recompiled to reflect this.
1102 Bump __FreeBSD_version to 800070.
1105 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1106 updated to use it. This requires a buildworld/installworld in order to
1107 update the library and dependencies (usbconfig, etc). Its advisable to
1108 rebuild all ports which uses libusb. More specific directions are given
1109 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1110 libusb are no longer required and can be removed.
1113 A workaround is committed to allow the creation of System V shared
1114 memory segment of size > 2 GB on the 64-bit architectures.
1115 Due to a limitation of the existing ABI, the shm_segsz member
1116 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1117 wrong for large segments. Note that limits must be explicitly
1118 raised to allow such segments to be created.
1121 The layout of struct ifnet has changed, requiring a rebuild of all
1122 network device driver modules.
1125 The /dev handling for the new USB stack has changed, a
1126 buildworld/installworld is required for libusb20.
1129 The new USB2 stack has now been permanently moved in and all kernel and
1130 module names reverted to their previous values (eg, usb, ehci, ohci,
1131 ums, ...). The old usb stack can be compiled in by prefixing the name
1132 with the letter 'o', the old usb modules have been removed.
1133 Updating entry 20090216 for xorg and 20090215 for libmap may still
1137 The rc.conf(5) option if_up_delay has been renamed to
1138 defaultroute_delay to better reflect its purpose. If you have
1139 customized this setting in /etc/rc.conf you need to update it to
1143 xorg 7.4 wants to configure its input devices via hald which does not
1144 yet work with USB2. If the keyboard/mouse does not work in xorg then
1146 Option "AllowEmptyInput" "off"
1147 to your ServerLayout section. This will cause X to use the configured
1148 kbd and mouse sections from your xorg.conf.
1151 The GENERIC kernels for all architectures now default to the new USB2
1152 stack. No kernel config options or code have been removed so if a
1153 problem arises please report it and optionally revert to the old USB
1154 stack. If you are loading USB kernel modules or have a custom kernel
1155 that includes GENERIC then ensure that usb names are also changed over,
1156 eg uftdi -> usb2_serial_ftdi.
1158 Older programs linked against the ports libusb 0.1 need to be
1159 redirected to the new stack's libusb20. /etc/libmap.conf can
1161 # Map old usb library to new one for usb2 stack
1162 libusb-0.1.so.8 libusb20.so.1
1165 All USB ethernet devices now attach as interfaces under the name ueN
1166 (eg. ue0). This is to provide a predictable name as vendors often
1167 change usb chipsets in a product without notice.
1170 The ichsmb(4) driver has been changed to require SMBus slave
1171 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1172 All of the other SMBus controller drivers require left-justified
1173 slave addresses, so this change makes all the drivers provide the
1177 INET6 statistics (struct ip6stat) was updated.
1178 netstat(1) needs to be recompiled.
1181 NTFS has been removed from GENERIC kernel on amd64 to match
1182 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1183 will load ntfs.ko module automatically when NTFS support is
1184 actually needed, unless ntfs.ko is not installed or security
1185 level prohibits loading kernel modules. If either is the case,
1186 "options NTFS" has to be added into kernel config.
1189 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1190 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1191 800061. User space tools that rely on the size of struct tcpcb in
1192 tcp_var.h (e.g. sockstat) need to be recompiled.
1195 ng_tty(4) module updated to match the new TTY subsystem.
1196 Due to API change, user-level applications must be updated.
1197 New API support added to mpd5 CVS and expected to be present
1198 in next mpd5.3 release.
1201 With __FreeBSD_version 800060 the makefs tool is part of
1202 the base system (it was a port).
1205 The afdata and ifnet locks have been changed from mutexes to
1206 rwlocks, network modules will need to be re-compiled.
1209 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1210 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1211 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1212 architecture and 40 bytes on 64-bit architecture. The userland
1213 applications "arp" and "ndp" have been updated accordingly.
1214 The output from "netstat -r" shows only routing entries and
1215 none of the L2 information.
1218 __FreeBSD_version 800057 marks the switchover from the
1219 binary ath hal to source code. Users must add the line:
1221 options AH_SUPPORT_AR5416
1223 to their kernel config files when specifying:
1227 The ath_hal module no longer exists; the code is now compiled
1228 together with the driver in the ath module. It is now
1229 possible to tailor chip support (i.e. reduce the set of chips
1230 and thereby the code size); consult ath_hal(4) for details.
1233 __FreeBSD_version 800054 adds memory barriers to
1234 <machine/atomic.h>, new interfaces to ifnet to facilitate
1235 multiple hardware transmit queues for cards that support
1236 them, and a lock-less ring-buffer implementation to
1237 enable drivers to more efficiently manage queueing of
1241 A new version of ZFS (version 13) has been merged to -HEAD.
1242 This version has zpool attribute "listsnapshots" off by
1243 default, which means "zfs list" does not show snapshots,
1244 and is the same as Solaris behavior.
1247 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1250 The uhci, ohci, ehci and slhci USB Host controller drivers have
1251 been put into separate modules. If you load the usb module
1252 separately through loader.conf you will need to load the
1253 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1254 controller add the following to loader.conf:
1260 The ABI used by the PMC toolset has changed. Please keep
1261 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1265 atapci kernel module now includes only generic PCI ATA
1266 driver. AHCI driver moved to ataahci kernel module.
1267 All vendor-specific code moved into separate kernel modules:
1268 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1269 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1270 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1271 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1274 The TTY subsystem of the kernel has been replaced by a new
1275 implementation, which provides better scalability and an
1276 improved driver model. Most common drivers have been migrated to
1277 the new TTY subsystem, while others have not. The following
1278 drivers have not yet been ported to the new TTY layer:
1281 cy, digi, rc, rp, sio
1287 ng_h4, ng_tty, ppp, sl, snp
1289 Adding these drivers to your kernel configuration file shall
1290 cause compilation to fail.
1293 ntpd has been upgraded to 4.2.4p5.
1296 OpenSSH has been upgraded to 5.1p1.
1298 For many years, FreeBSD's version of OpenSSH preferred DSA
1299 over RSA for host and user authentication keys. With this
1300 upgrade, we've switched to the vendor's default of RSA over
1301 DSA. This may cause upgraded clients to warn about unknown
1302 host keys even for previously known hosts. Users should
1303 follow the usual procedure for verifying host keys before
1304 accepting the RSA key.
1306 This can be circumvented by setting the "HostKeyAlgorithms"
1307 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1310 Please note that the sequence of keys offered for
1311 authentication has been changed as well. You may want to
1312 specify IdentityFile in a different order to revert this
1316 The sio(4) driver has been removed from the i386 and amd64
1317 kernel configuration files. This means uart(4) is now the
1318 default serial port driver on those platforms as well.
1320 To prevent collisions with the sio(4) driver, the uart(4) driver
1321 uses different names for its device nodes. This means the
1322 onboard serial port will now most likely be called "ttyu0"
1323 instead of "ttyd0". You may need to reconfigure applications to
1324 use the new device names.
1326 When using the serial port as a boot console, be sure to update
1327 /boot/device.hints and /etc/ttys before booting the new kernel.
1328 If you forget to do so, you can still manually specify the hints
1329 at the loader prompt:
1331 set hint.uart.0.at="isa"
1332 set hint.uart.0.port="0x3F8"
1333 set hint.uart.0.flags="0x10"
1334 set hint.uart.0.irq="4"
1338 The gpt(8) utility has been removed. Use gpart(8) to partition
1342 The version that Linuxulator emulates was changed from 2.4.2
1343 to 2.6.16. If you experience any problems with Linux binaries
1344 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1345 if it fixes the problem contact emulation mailing list.
1348 ISDN4BSD (I4B) was removed from the src tree. You may need to
1349 update a your kernel configuration and remove relevant entries.
1352 I have checked in code to support multiple routing tables.
1353 See the man pages setfib(1) and setfib(2).
1354 This is a hopefully backwards compatible version,
1355 but to make use of it you need to compile your kernel
1356 with options ROUTETABLES=2 (or more up to 16).
1359 The 802.11 wireless support was redone to enable multi-bss
1360 operation on devices that are capable. The underlying device
1361 is no longer used directly but instead wlanX devices are
1362 cloned with ifconfig. This requires changes to rc.conf files.
1363 For example, change:
1364 ifconfig_ath0="WPA DHCP"
1367 ifconfig_wlan0="WPA DHCP"
1368 see rc.conf(5) for more details. In addition, mergemaster of
1369 /etc/rc.d is highly recommended. Simultaneous update of userland
1370 and kernel wouldn't hurt either.
1372 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1373 modules were merged into the base wlan module. All references
1374 to these modules (e.g. in kernel config files) must be removed.
1377 psm(4) has gained write(2) support in native operation level.
1378 Arbitrary commands can be written to /dev/psm%d and status can
1379 be read back from it. Therefore, an application is responsible
1380 for status validation and error recovery. It is a no-op in
1381 other operation levels.
1384 Support for KSE threading has been removed from the kernel. To
1385 run legacy applications linked against KSE libmap.conf may
1386 be used. The following libmap.conf may be used to ensure
1387 compatibility with any prior release:
1389 libpthread.so.1 libthr.so.1
1390 libpthread.so.2 libthr.so.2
1391 libkse.so.3 libthr.so.3
1394 The layout of struct vmspace has changed. This affects libkvm
1395 and any executables that link against libkvm and use the
1396 kvm_getprocs() function. In particular, but not exclusively,
1397 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1398 The effects are minimal, but it's advisable to upgrade world
1402 The latest em driver no longer has support in it for the
1403 82575 adapter, this is now moved to the igb driver. The
1404 split was done to make new features that are incompatible
1405 with older hardware easier to do.
1408 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1409 likewise the kernel option is now GEOM_LINUX_LVM.
1412 The default NFS mount mode has changed from UDP to TCP for
1413 increased reliability. If you rely on (insecurely) NFS
1414 mounting across a firewall you may need to update your
1418 Belatedly note the addition of m_collapse for compacting
1422 The fts(3) structures have been changed to use adequate
1423 integer types for their members and so to be able to cope
1424 with huge file trees. The old fts(3) ABI is preserved
1425 through symbol versioning in libc, so third-party binaries
1426 using fts(3) should still work, although they will not take
1427 advantage of the extended types. At the same time, some
1428 third-party software might fail to build after this change
1429 due to unportable assumptions made in its source code about
1430 fts(3) structure members. Such software should be fixed
1431 by its vendor or, in the worst case, in the ports tree.
1432 FreeBSD_version 800015 marks this change for the unlikely
1433 case that a portable fix is impossible.
1436 To upgrade to -current after this date, you must be running
1437 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1438 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1441 The ADAPTIVE_GIANT kernel option has been retired because its
1442 functionality is the default now.
1445 The AT keyboard emulation of sunkbd(4) has been turned on
1446 by default. In order to make the special symbols of the Sun
1447 keyboards driven by sunkbd(4) work under X these now have
1448 to be configured the same way as Sun USB keyboards driven
1449 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1451 Option "XkbLayout" "us"
1452 Option "XkbRules" "xorg"
1453 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1456 It has been decided that it is desirable to provide ABI
1457 backwards compatibility to the FreeBSD 4/5/6 versions of the
1458 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1459 broken with the introduction of PCI domain support (see the
1460 20070930 entry). Unfortunately, this required the ABI of
1461 PCIOCGETCONF to be broken again in order to be able to
1462 provide backwards compatibility to the old version of that
1463 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1464 again. As for prominent ports this affects neither pciutils
1465 nor xorg-server this time, the hal port needs to be rebuilt
1469 The misnamed kthread_create() and friends have been renamed
1470 to kproc_create() etc. Many of the callers already
1471 used kproc_start()..
1472 I will return kthread_create() and friends in a while
1473 with implementations that actually create threads, not procs.
1474 Renaming corresponds with version 800002.
1483 Avoid using make -j when upgrading. While generally safe, there are
1484 sometimes problems using -j to upgrade. If your upgrade fails with
1485 -j, please try again without -j. From time to time in the past there
1486 have been problems using -j with buildworld and/or installworld. This
1487 is especially true when upgrading between "distant" versions (eg one
1488 that cross a major release boundary or several minor releases, or when
1489 several months have passed on the -current branch).
1491 Sometimes, obscure build problems are the result of environment
1492 poisoning. This can happen because the make utility reads its
1493 environment when searching for values for global variables. To run
1494 your build attempts in an "environmental clean room", prefix all make
1495 commands with 'env -i '. See the env(1) manual page for more details.
1497 When upgrading from one major version to another it is generally best
1498 to upgrade to the latest code in the currently installed branch first,
1499 then do an upgrade to the new branch. This is the best-tested upgrade
1500 path, and has the highest probability of being successful. Please try
1501 this approach before reporting problems with a major version upgrade.
1505 When upgrading the boot ZFS pool to a new version, always follow
1508 1.) recompile and reinstall the ZFS boot loader and boot block
1509 (this is part of "make buildworld" and "make installworld")
1511 2.) update the ZFS boot block on your boot drive
1513 The following example updates the ZFS boot block on the first
1514 partition (freebsd-boot) of a GPT partitioned drive ad0:
1515 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1517 Non-boot pools do not need these updates.
1521 If you are updating from a prior version of FreeBSD (even one just
1522 a few days old), you should follow this procedure. It is the most
1523 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1525 make kernel-toolchain
1526 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1527 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1529 To test a kernel once
1530 ---------------------
1531 If you just want to boot a kernel once (because you are not sure
1532 if it works, or if you want to boot a known bad kernel to provide
1533 debugging information) run
1534 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1535 nextboot -k testkernel
1537 To just build a kernel when you know that it won't mess you up
1538 --------------------------------------------------------------
1539 This assumes you are already running a CURRENT system. Replace
1540 ${arch} with the architecture of your machine (e.g. "i386",
1541 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1543 cd src/sys/${arch}/conf
1544 config KERNEL_NAME_HERE
1545 cd ../compile/KERNEL_NAME_HERE
1550 If this fails, go to the "To build a kernel" section.
1552 To rebuild everything and install it on the current system.
1553 -----------------------------------------------------------
1554 # Note: sometimes if you are running current you gotta do more than
1555 # is listed here if you are upgrading from a really old current.
1557 <make sure you have good level 0 dumps>
1559 make kernel KERNCONF=YOUR_KERNEL_HERE
1561 <reboot in single user> [3]
1569 To cross-install current onto a separate partition
1570 --------------------------------------------------
1571 # In this approach we use a separate partition to hold
1572 # current's root, 'usr', and 'var' directories. A partition
1573 # holding "/", "/usr" and "/var" should be about 2GB in
1576 <make sure you have good level 0 dumps>
1579 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1580 <maybe newfs current's root partition>
1581 <mount current's root partition on directory ${CURRENT_ROOT}>
1582 make installworld DESTDIR=${CURRENT_ROOT}
1583 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1584 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1585 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1586 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1587 <reboot into current>
1588 <do a "native" rebuild/install as described in the previous section>
1589 <maybe install compatibility libraries from ports/misc/compat*>
1593 To upgrade in-place from 8.x-stable to current
1594 ----------------------------------------------
1595 <make sure you have good level 0 dumps>
1597 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1599 <reboot in single user> [3]
1606 Make sure that you've read the UPDATING file to understand the
1607 tweaks to various things you need. At this point in the life
1608 cycle of current, things change often and you are on your own
1609 to cope. The defaults can also change, so please read ALL of
1610 the UPDATING entries.
1612 Also, if you are tracking -current, you must be subscribed to
1613 freebsd-current@freebsd.org. Make sure that before you update
1614 your sources that you have read and understood all the recent
1615 messages there. If in doubt, please track -stable which has
1616 much fewer pitfalls.
1618 [1] If you have third party modules, such as vmware, you
1619 should disable them at this point so they don't crash your
1622 [3] From the bootblocks, boot -s, and then do
1627 adjkerntz -i # if CMOS is wall time
1628 Also, when doing a major release upgrade, it is required that
1629 you boot into single user mode to do the installworld.
1631 [4] Note: This step is non-optional. Failure to do this step
1632 can result in a significant reduction in the functionality of the
1633 system. Attempting to do it by hand is not recommended and those
1634 that pursue this avenue should read this file carefully, as well
1635 as the archives of freebsd-current and freebsd-hackers mailing lists
1636 for potential gotchas. The -U option is also useful to consider.
1637 See mergemaster(8) for more information.
1639 [5] Usually this step is a noop. However, from time to time
1640 you may need to do this if you get unknown user in the following
1641 step. It never hurts to do it all the time. You may need to
1642 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1643 install) after the buildworld before this step if you last updated
1644 from current before 20020224 or from -stable before 20020408.
1646 [6] This only deletes old files and directories. Old libraries
1647 can be deleted by "make delete-old-libs", but you have to make
1648 sure that no program is using those libraries anymore.
1650 [8] In order to have a kernel that can run the 4.x binaries needed to
1651 do an installworld, you must include the COMPAT_FREEBSD4 option in
1652 your kernel. Failure to do so may leave you with a system that is
1653 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1654 required to run the 5.x binaries on more recent kernels. And so on
1655 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1657 Make sure that you merge any new devices from GENERIC since the
1658 last time you updated your kernel config file.
1660 [9] When checking out sources, you must include the -P flag to have
1661 cvs prune empty directories.
1663 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1664 "?=" instead of the "=" assignment operator, so that buildworld can
1665 override the CPUTYPE if it needs to.
1667 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1668 not on the command line, or in /etc/make.conf. buildworld will
1669 warn if it is improperly defined.
1672 This file contains a list, in reverse chronological order, of major
1673 breakages in tracking -current. It is not guaranteed to be a complete
1674 list of such breakages, and only contains entries since October 10, 2007.
1675 If you need to see UPDATING entries from before that date, you will need
1676 to fetch an UPDATING file from an older FreeBSD release.
1678 Copyright information:
1680 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1682 Redistribution, publication, translation and use, with or without
1683 modification, in full or in part, in any form or format of this
1684 document are permitted without further permission from the author.
1686 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1687 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1688 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1689 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1690 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1691 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1692 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1693 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1694 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1695 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1696 POSSIBILITY OF SUCH DAMAGE.
1698 Contact Warner Losh if you have any questions about your use of