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 20141104: p22 FreeBSD-SA-14:24.sshd
13 FreeBSD-SA-14:25.setlogin
17 Fix denial of service attack against sshd(8). [SA-14:24]
19 Fix kernel stack disclosure in setlogin(2) / getlogin(2).
22 Fix remote command execution in ftp(1). [SA-14:26]
24 Fix NFSv4 and ZFS cache consistency issue. [EN-14:12]
26 20141022: p21 FreeBSD-EN-14:10.tzdata
28 Time zone data file update. [EN-14:10]
30 20141021: p20 FreeBSD-SA-14:20.rtsold
31 FreeBSD-SA-14:21.routed
32 FreeBSD-SA-14:22.namei
33 FreeBSD-SA-14:23.openssl
35 Fix rtsold(8) remote buffer overflow vulnerability. [SA-14:20]
37 Fix routed(8) remote denial of service vulnerability. [SA-14:21]
39 Fix memory leak in sandboxed namei lookup. [SA-14:22]
41 Fix OpenSSL multiple vulnerabilities. [SA-14:23]
43 20140916: p19 FreeBSD-SA-14:19.tcp
44 Fix Denial of Service in TCP packet processing. [SA-14:19]
46 20140909: p18 FreeBSD-SA-14:18.openssl
47 Fix OpenSSL multiple vulnerabilities. [SA-14:18]
49 20140708: p17 FreeBSD-SA-14:17.kmem
50 Fix kernel memory disclosure in control messages and SCTP
51 notifications. [SA-14:17]
53 20140624: p16 FreeBSD-SA-14:16.file
54 FreeBSD-EN-14:08.heimdal
56 Fix multiple vulnerabilities in file(1) and libmagic(3).
59 Fix gss_pseudo_random interoperability issue. [EN-14:08]
61 20140605: p15 FreeBSD-SA-14:14.openssl
62 Fix OpenSSL multiple vulnerabilities. [SA-14:14]
64 20140603: p14 FreeBSD-SA-14:11.sendmail
65 FreeBSD-SA-14:12.ktrace
69 Fix sendmail improper close-on-exec flag handling. [SA-14:11]
71 Fix ktrace memory disclosure. [SA-14:12]
73 Fix incorrect error handling in PAM policy parser. [SA-14:13]
75 Fix triple-fault when executing from a threaded process.
78 20140513: p13 FreeBSD-EN-14:03.pkg
79 FreeBSD-EN-14:04.kldxref
82 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
84 Improve build repeatability for kldxref(8). [EN-14:04]
86 Fix data corruption with ciss(4). [EN-14:05]
88 20140430: p12 FreeBSD-SA-14:08.tcp
90 Fix TCP reassembly vulnerability. [SA-14:08]
92 20140408: p11 FreeBSD-SA-14:05.nfsserver
93 FreeBSD-SA-14:06.openssl
94 Fix deadlock in the NFS server. [SA-14:05]
96 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
98 20140114: p10 FreeBSD-SA-14:01.bsnmpd
100 FreeBSD-SA-14:04.bind
101 FreeBSD-EN-14:01.random
102 FreeBSD-EN-14:02.mmap
103 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
105 Fix ntpd distributed reflection Denial of Service
106 vulnerability. [SA-14:02]
108 Fix BIND remote denial of service vulnerability. [SA-14:04]
110 Disable hardware RNGs by default. [EN-14:01]
112 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
114 20131128: p9 FreeBSD-EN-13:05.freebsd-update
115 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
117 20131026: p8 FreeBSD-EN-13:04.freebsd-update
118 Fix multiple freebsd-update bugs that break upgrading to
121 20130910: p7 FreeBSD-SA-13:12.ifioctl
122 FreeBSD-SA-13:13.nullfs
124 In IPv6 and NetATM, stop SIOCSIFADDR, SIOCSIFBRDADDR,
125 SIOCSIFDSTADDR and SIOCSIFNETMASK at the socket layer rather
126 than pass them on to the link layer without validation or
127 credential checks. [SA-13:12]
129 Prevent cross-mount hardlinks between different nullfs mounts
130 of the same underlying filesystem. [SA-13:13]
132 20130822: p6 FreeBSD-SA-13:09.ip_multicast
133 FreeBSD-SA-13:10.sctp
135 Fix an integer overflow in computing the size of a temporary buffer
136 can result in a buffer which is too small for the requested
139 Fix a bug that could lead to kernel memory disclosure with
140 SCTP state cookie. [13:10]
142 Fix a data corruption problem with mfi(4) operating on > 2TB
143 disks in a JBOD. [EN-13:03]
145 20130726: p5 FreeBSD-SA-13:07.bind FreeBSD-SA-13:08.nfsserver
146 Fix Denial of Service vulnerability in named(8). [13:07]
148 Fix a bug that allows remote client bypass the normal
149 access checks when when -network or -host restrictions are
150 used at the same time with -mapall. [13:08]
152 20130618: p4 FreeBSD-SA-13:06.mmap
153 Fix a bug that allowed a tracing process (e.g. gdb) to write
154 to a memory-mapped file in the traced process's address space
155 even if neither the traced process nor the tracing process had
156 write access to that file.
158 20130429: p3 FreeBSD-SA-13:05.nfsserver
159 Fix a bug that allows NFS clients to issue READDIR on files.
161 20130402: p2 FreeBSD-SA-13:03.openssl FreeBSD-SA-13:04.bind
162 Fix multiple vulnerabilities in OpenSSL.
164 Fix Denial of Service vulnerability in named(8).
166 20130218: p1 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
167 Fix Denial of Service vulnerability in named(8) with DNS64.
169 Fix Denial of Service vulnerability in libc's glob(3) functionality.
175 The sparc64 ZFS loader has been changed to no longer try to auto-
176 detect ZFS providers based on diskN aliases but now requires these
177 to be explicitly listed in the OFW boot-device environment variable.
180 Now unix domain sockets behave "as expected" on nullfs(5). Previously
181 nullfs(5) did not pass through all behaviours to the underlying layer,
182 as a result if we bound to a socket on the lower layer we could connect
183 only to the lower path; if we bound to the upper layer we could connect
184 only to the upper path. The new behavior is one can connect to both the
185 lower and the upper paths regardless what layer path one binds to.
188 The acpi_wmi(4) status device /dev/wmistat has been renamed to
192 A new VOP_ADVISE() was added to support posix_fadvise(2). All
193 filesystem modules must be recompiled.
196 The interface of the VOP_VPTOCNP(9) changed, now the returned
197 vnode shall be referenced, previously it was required to be
198 only held. All in-tree filesystems are converted.
204 The broken amd(4) driver has been replaced with esp(4) in the amd64,
205 i386 and pc98 GENERIC kernel configuration files.
208 This commit modifies vfs_register() so that it uses a hash
209 calculation to set vfc_typenum, which is enabled by default.
210 The first time a system is booted after this change, the
211 vfc_typenum values will change for all file systems. The
212 main effect of this is a change to the NFS server file handles
213 for file systems that use vfc_typenum in their fsid, such as ZFS.
214 It will, however, prevent vfc_typenum from changing when file
215 systems are loaded in a different order for subsequent reboots.
216 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
217 until you are ready to remount all NFS clients after a reboot.
220 Bump the shared library version numbers for libraries that
221 do not use symbol versioning, have changed the ABI compared
222 to stable/8 and which shared library version was not bumped.
223 Done as part of 9.0-RELEASE cycle.
226 During the merge of Capsicum features, the fget(9) KPI was modified.
227 This may require the rebuilding of out-of-tree device drivers --
228 issues have been reported specifically with the nVidia device driver.
229 __FreeBSD_version is bumped to 900041.
231 Also, there is a period between 20110811 and 20110814 where the
232 special devices /dev/{stdin,stdout,stderr} did not work correctly.
233 Building world from a kernel during that window may not work.
236 The packet filter (pf) code has been updated to OpenBSD 4.5.
237 You need to update userland tools to be in sync with kernel.
238 This update breaks backward compatibility with earlier pfsync(4)
239 versions. Care must be taken when updating redundant firewall setups.
242 The following sysctls and tunables are retired on x86 platforms:
244 machdep.hlt_logical_cpus
245 The following sysctl is retired:
246 machdep.hyperthreading_allowed
247 The sysctls were supposed to provide a way to dynamically offline and
248 online selected CPUs on x86 platforms, but the implementation has not
249 been reliable especially with SCHED_ULE scheduler.
250 machdep.hyperthreading_allowed tunable is still available to ignore
251 hyperthreading CPUs at OS level.
252 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
253 where X is an APIC ID of a CPU. Be advised, though, that disabling
254 CPUs in non-uniform fashion will result in non-uniform topology and
255 may lead to sub-optimal system performance with SCHED_ULE, which is
259 cpumask_t type is retired and cpuset_t is used in order to describe
263 Changes to ifconfig(8) for dynamic address family detection mandate
264 that you are running a kernel of 20110525 or later. Make sure to
265 follow the update procedure to boot a new kernel before installing
269 Support for sun4v architecture is officially dropped
272 Several KPI breaking changes have been committed to the mii(4) layer,
273 the PHY drivers and consequently some Ethernet drivers using mii(4).
274 This means that miibus.ko and the modules of the affected Ethernet
275 drivers need to be recompiled.
277 Note to kernel developers: Given that the OUI bit reversion problem
278 was fixed as part of these changes all mii(4) commits related to OUIs,
279 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
280 handling, no longer can be merged verbatim to stable/8 and previous
284 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
285 into their kernel configurations along with 'device pci'.
288 The default NFS client is now the new NFS client, so fstype "newnfs"
289 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
290 Although mounts via fstype "nfs" will usually work without userland
291 changes, it is recommended that the mount(8) and mount_nfs(8)
292 commands be rebuilt from sources and that a link to mount_nfs called
293 mount_oldnfs be created. The new client is compiled into the
294 kernel with "options NFSCL" and this is needed for diskless root
295 file systems. The GENERIC kernel configs have been changed to use
296 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
297 To use the regular/old client, you can "mount -t oldnfs ...". For
298 a diskless root file system, you must also include a line like:
300 vfs.root.mountfrom="oldnfs:"
302 in the boot/loader.conf on the root fs on the NFS server to make
303 a diskless root fs use the old client.
306 The GENERIC kernels for all architectures now default to the new
307 CAM-based ATA stack. It means that all legacy ATA drivers were
308 removed and replaced by respective CAM drivers. If you are using
309 ATA device names in /etc/fstab or other places, make sure to update
310 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
311 where 'Y's are the sequential numbers starting from zero for each type
312 in order of detection, unless configured otherwise with tunables,
313 see cam(4)). There will be symbolic links created in /dev/ to map
314 old adX devices to the respective adaY. They should provide basic
315 compatibility for file systems mounting in most cases, but they do
316 not support old user-level APIs and do not have respective providers
317 in GEOM. Consider using updated management tools with new device names.
319 It is possible to load devices ahci, ata, siis and mvs as modules,
320 but option ATA_CAM should remain in kernel configuration to make ata
321 module work as CAM driver supporting legacy ATA controllers. Device ata
322 still can be used in modular fashion (atacore + ...). Modules atadisk
323 and atapi* are not used and won't affect operation in ATA_CAM mode.
324 Note that to use CAM-based ATA kernel should include CAM devices
325 scbus, pass, da (or explicitly ada), cd and optionally others. All of
326 them are parts of the cam module.
328 ataraid(4) functionality is now supported by the RAID GEOM class.
329 To use it you can load geom_raid kernel module and use graid(8) tool
330 for management. Instead of /dev/arX device names, use /dev/raid/rX.
332 No kernel config options or code have been removed, so if a problem
333 arises, please report it and optionally revert to the old ATA stack.
334 In order to do it you can remove from the kernel config:
339 , and instead add back:
340 device atadisk # ATA disk drives
341 device ataraid # ATA RAID drives
342 device atapicd # ATAPI CDROM drives
343 device atapifd # ATAPI floppy drives
344 device atapist # ATAPI tape drives
347 The default NFS server has been changed to the new server, which
348 was referred to as the experimental server. If you need to switch
349 back to the old NFS server, you must now put the "-o" option on
350 both the mountd and nfsd commands. This can be done using the
351 mountd_flags and nfs_server_flags rc.conf variables until an
352 update to the rc scripts is committed, which is coming soon.
355 The GNU Objective-C runtime library (libobjc), and other Objective-C
356 related components have been removed from the base system. If you
357 require an Objective-C library, please use one of the available ports.
360 ath(4) has been split into bus- and device- modules. if_ath contains
361 the HAL, the TX rate control and the network device code. if_ath_pci
362 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
363 contains the AHB glue. Users need to load both if_ath_pci and if_ath
364 in order to use ath on everything else.
366 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
367 need to load if_ath and if_ath_pci for ath(4) operation.
370 As part of the replacement of sysinstall, the process of building
371 release media has changed significantly. For details, please re-read
372 release(7), which has been updated to reflect the new build process.
375 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
376 is the last available version under GPLv2. It brings a number of new
377 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
378 4.1 and SSE 4.2), better support for powerpc64, a number of new
379 directives, and lots of other small improvements. See the ChangeLog
380 file in contrib/binutils for the full details.
383 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
384 compliant, and will now use half of hash for authentication.
385 This will break interoperability with all stacks (including all
386 actual FreeBSD versions) who implement
387 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
389 The only workaround with such peers is to use another HMAC
390 algorithm for IPsec ("phase 2") authentication.
393 Remove the uio_yield prototype and symbol. This function has
394 been misnamed since it was introduced and should not be
395 globally exposed with this name. The equivalent functionality
396 is now available using kern_yield(curthread->td_user_pri).
397 The function remains undocumented.
400 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
401 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
402 for scalar sysctls is defined but disabled. Code that needs
403 UQUAD to pass the type checking that must compile on older
404 systems where the define is not present can check against
405 __FreeBSD_version >= 900030.
407 The system dialog(1) has been replaced with a new version previously
408 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
409 with the previous version, but the libdialog associated with it has
410 a largely incompatible API. As such, the original version of libdialog
411 will be kept temporarily as libodialog, until its base system consumers
412 are replaced or updated. Bump __FreeBSD_version to 900030.
415 If you are trying to run make universe on a -stable system, and you get
416 the following warning:
417 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
418 unknown. config(8) likely too old."
419 or something similar to it, then you must upgrade your -stable system
420 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
421 stable/8) or build the config from the latest stable/8 branch and
422 install it on your system.
424 Prior to this date, building a current universe on 8-stable system from
425 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
426 error in the first kernel build phase. A new config on those old
427 systems will fix that problem for older versions of -current.
430 The TCP stack has been modified to allow Khelp modules to interact with
431 it via helper hook points and store per-connection data in the TCP
432 control block. Bump __FreeBSD_version to 900029. User space tools that
433 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
437 Generic IEEE 802.3 annex 31B full duplex flow control support has been
438 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
439 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
440 to take advantage of it instead of using custom implementations. This
441 means that these drivers now no longer unconditionally advertise
442 support for flow control but only do so if flow control is a selected
443 media option. This was implemented in the generic support that way in
444 order to allow flow control to be switched on and off via ifconfig(8)
445 with the PHY specific default to typically off in order to protect
446 from unwanted effects. Consequently, if you used flow control with
447 one of the above mentioned drivers you now need to explicitly enable
449 ifconfig bge0 media auto mediaopt flowcontrol
451 Along with the above mentioned changes generic support for setting
452 1000baseT master mode also has been added and brgphy(4), ciphy(4),
453 e1000phy(4) as well as ip1000phy(4) have been converted to take
454 advantage of it. This means that these drivers now no longer take the
455 link0 parameter for selecting master mode but the master media option
456 has to be used instead, for example like in the following:
457 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
459 Selection of master mode now is also available with all other PHY
460 drivers supporting 1000baseT.
463 The TCP stack has received a significant update to add support for
464 modularised congestion control and generally improve the clarity of
465 congestion control decisions. Bump __FreeBSD_version to 900025. User
466 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
467 sockstat) need to be recompiled.
470 The man(1) utility has been replaced by a new version that no longer
471 uses /etc/manpath.config. Please consult man.conf(5) for how to
472 migrate local entries to the new format.
475 The copyright strings printed by login(1) and sshd(8) at the time of a
476 new connection have been removed to follow other operating systems and
480 A workaround for a fixed ld bug has been removed in kernel code,
481 so make sure that your system ld is built from sources after
482 revision 210245 from 2010-07-19 (r211583 if building head kernel
483 on stable/8, r211584 for stable/7; both from 2010-08-21).
484 A symptom of incorrect ld version is different addresses for
485 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
488 The $ipv6_prefer variable in rc.conf(5) has been split into
489 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
491 The $ip6addrctl_policy is a variable to choose a pre-defined
492 address selection policy set by ip6addrctl(8). A value
493 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
496 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
497 flag (see an entry of 20090926) is set on an interface with no
498 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
499 security reason. If you want IPv6 link-local address on all
500 interfaces by default, set this to "YES".
502 The old ipv6_prefer="YES" is equivalent to
503 ipv6_activate_all_interfaces="YES" and
504 ip6addrctl_policy="ipv6_prefer".
507 DTrace has grown support for userland tracing. Due to this, DTrace is
508 now i386 and amd64 only.
509 dtruss(1) is now installed by default on those systems and a new
510 kernel module is needed for userland tracing: fasttrap.
511 No changes to your kernel config file are necessary to enable
512 userland tracing, but you might consider adding 'STRIP=' and
513 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
514 to have informative userland stack traces in DTrace (ustack).
517 The acpi_aiboost(4) driver has been removed in favor of the new
518 aibs(4) driver. You should update your kernel configuration file.
521 BSD grep has been imported to the base system and it is built by
522 default. It is completely BSD licensed, highly GNU-compatible, uses
523 less memory than its GNU counterpart and has a small codebase.
524 However, it is slower than its GNU counterpart, which is mostly
525 noticeable for larger searches, for smaller ones it is measurable
526 but not significant. The reason is complex, the most important factor
527 is that we lack a modern and efficient regex library and GNU
528 overcomes this by optimizing the searches internally. Future work
529 on improving the regex performance is planned, for the meantime,
530 users that need better performance, can build GNU grep instead by
531 setting the WITH_GNU_GREP knob.
534 Due to the import of powerpc64 support, all existing powerpc kernel
535 configuration files must be updated with a machine directive like this:
536 machine powerpc powerpc
538 In addition, an updated config(8) is required to build powerpc kernels
542 A new version of ZFS (version 15) has been merged to -HEAD.
543 This version uses a python library for the following subcommands:
544 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
545 For full functionality of these commands the following port must
546 be installed: sysutils/py-zfs
549 'vm_page's are now hashed by physical address to an array of mutexes.
550 Currently this is only used to serialize access to hold_count. Over
551 time the page queue mutex will be peeled away. This changes the size
552 of pmap on every architecture. And requires all callers of vm_page_hold
553 and vm_page_unhold to be updated.
556 WITH_CTF can now be specified in src.conf (not recommended, there
557 are some problems with static executables), make.conf (would also
558 affect ports which do not use GNU make and do not override the
559 compile targets) or in the kernel config (via "makeoptions
561 When WITH_CTF was specified there before this was silently ignored,
562 so make sure that WITH_CTF is not used in places which could lead
563 to unwanted behavior.
566 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
567 to allow 32-bit compatibility on non-x86 platforms. All kernel
568 configurations on amd64 and ia64 platforms using these options must
569 be modified accordingly.
572 The utmp user accounting database has been replaced with utmpx,
573 the user accounting interface standardized by POSIX.
574 Unfortunately the semantics of utmp and utmpx don't match,
575 making it practically impossible to support both interfaces.
576 The user accounting database is used by tools like finger(1),
577 last(1), talk(1), w(1) and ac(8).
579 All applications in the base system use utmpx. This means only
580 local binaries (e.g. from the ports tree) may still use these
581 utmp database files. These applications must be rebuilt to make
584 After the system has been upgraded, it is safe to remove the old
585 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
586 assuming their contents is of no importance anymore. Old wtmp
587 databases can only be used by last(1) and ac(8) after they have
588 been converted to the new format using wtmpcvt(1).
591 Introduce the kernel thread "deadlock resolver" (which can be enabled
592 via the DEADLKRES option, see NOTES for more details) and the
593 sleepq_type() function for sleepqueues.
596 The rc.firewall and rc.firewall6 were unified, and
597 rc.firewall6 and rc.d/ip6fw were removed.
598 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
599 variables are obsoleted. Instead, the following new rc
600 variables are added to rc.d/ipfw:
602 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
603 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
604 firewall_simple_onet_ipv6, firewall_trusted_ipv6
606 The meanings correspond to the relevant IPv4 variables.
612 The default terminal emulation for syscons(4) has been changed
613 from cons25 to xterm on all platforms except pc98. This means
614 that the /etc/ttys file needs to be updated to ensure correct
615 operation of applications on the console.
617 The terminal emulation style can be toggled per window by using
618 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
619 options can be used to change the compile-time default back to
622 To prevent graphical artifacts, make sure the TERM environment
623 variable is set to match the terminal emulation that is being
624 performed by syscons(4).
627 The layout of the structure ieee80211req_scan_result has changed.
628 Applications that require wireless scan results (e.g. ifconfig(8))
629 from net80211 need to be recompiled.
631 Applications such as wpa_supplicant(8) may require a full world
632 build without using NO_CLEAN in order to get synchronized with the
636 The iwn(4) driver has been updated to support the 5000 and 5150 series.
637 There's one kernel module for each firmware. Adding "device iwnfw"
638 to the kernel configuration file means including all three firmware
639 images inside the kernel. If you want to include just the one for
640 your wireless card, use the devices iwn4965fw, iwn5000fw or
644 The rc.d/network_ipv6, IPv6 configuration script has been integrated
645 into rc.d/netif. The changes are the following:
647 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
648 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
649 Note that both variables need the "inet6" keyword at the head.
651 Do not set $ipv6_network_interfaces manually if you do not
652 understand what you are doing. It is not needed in most cases.
654 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
657 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
658 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
660 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
661 all configured interfaces have "inet6 accept_rtadv" in the
662 $ifconfig_IF_ipv6. These are for backward compatibility.
664 3. A new variable $ipv6_prefer has been added. If NO, IPv6
665 functionality of interfaces with no corresponding
666 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
667 and the default address selection policy of ip6addrctl(8)
668 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
669 Note that if you want to configure IPv6 functionality on the
670 disabled interfaces after boot, first you need to clear the flag by
671 using ifconfig(8) like:
673 ifconfig em0 inet6 -ifdisabled
675 If YES, the default address selection policy is set as
678 The default value of $ipv6_prefer is NO.
680 4. If your system need to receive Router Advertisement messages,
681 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
682 scripts automatically invoke rtsol(8) when the interface becomes
683 UP. The Router Advertisement messages are used for SLAAC
684 (State-Less Address AutoConfiguration).
687 802.11s D3.03 support was committed. This is incompatible with the
688 previous code, which was based on D3.0.
691 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
692 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
693 control whether accepting Router Advertisement messages or not.
694 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
695 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
696 The ifconfig(8) utility now supports these flags.
699 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
700 mount(8) and -a option for df(1) to see them.
703 The old tunable hw.bus.devctl_disable has been superseded by
704 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
705 replaced by hw.bus.devctl_queue=0. The default for this new tunable
709 Remove the option STOP_NMI. The default action is now to use NMI only
710 for KDB via the newly introduced function stop_cpus_hard() and
711 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
714 The stable/8 branch created in subversion. This corresponds to the
715 RELENG_8 branch in CVS.
718 Bump the shared library version numbers for all libraries that do not
719 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
720 __FreeBSD_version to 800105.
723 Due to changes in the implementation of virtual network stack support,
724 all network-related kernel modules must be recompiled. As this change
725 breaks the ABI, bump __FreeBSD_version to 800104.
728 The TOE interface to the TCP syncache has been modified to remove
729 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
730 The cxgb driver is the only TOE consumer affected by this change, and
731 needs to be recompiled along with the kernel. As this change breaks
732 the ABI, bump __FreeBSD_version to 800103.
735 Padding has been added to struct tcpcb, sackhint and tcpstat in
736 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
737 maintaining the ABI. However, this change breaks the ABI, so bump
738 __FreeBSD_version to 800102. User space tools that rely on the size of
739 any of these structs (e.g. sockstat) need to be recompiled.
742 The NFS_LEGACYRPC option has been removed along with the old kernel
743 RPC implementation that this option selected. Kernel configurations
744 may need to be adjusted.
747 The network interface device nodes at /dev/net/<interface> have been
748 removed. All ioctl operations can be performed the normal way using
749 routing sockets. The kqueue functionality can generally be replaced
750 with routing sockets.
753 The documentation from the FreeBSD Documentation Project (Handbook,
754 FAQ, etc.) is now installed via packages by sysinstall(8) and under
755 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
758 The ABI of various structures related to the SYSV IPC API have been
759 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
760 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
764 Layout of struct vnet has changed as routing related variables were
765 moved to their own Vimage module. Modules need to be recompiled. Bump
766 __FreeBSD_version to 800099.
769 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
770 respectively. As long as no more than 16 groups per process are used,
771 no changes should be visible. When more than 16 groups are used, old
772 binaries may fail if they call getgroups() or getgrouplist() with
773 statically sized storage. Recompiling will work around this, but
774 applications should be modified to use dynamically allocated storage
775 for group arrays as POSIX.1-2008 does not cap an implementation's
776 number of supported groups at NGROUPS_MAX+1 as previous versions did.
778 NFS and portalfs mounts may also be affected as the list of groups is
779 truncated to 16. Users of NFS who use more than 16 groups, should
780 take care that negative group permissions are not used on the exported
781 file systems as they will not be reliable unless a GSSAPI based
782 authentication method is used.
785 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
786 option compiles in the support for adaptive spinning for lockmgrs
787 which want to enable it. The lockinit() function now accepts the flag
788 LK_ADAPTIVE in order to make the lock object subject to adaptive
789 spinning when both held in write and read mode.
792 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
793 changed. User applications that use this ioctl need to be rebuilt.
796 The layout of struct thread has changed. Kernel and modules need to
800 The layout of structs ifnet, domain, protosw and vnet_net has changed.
801 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
804 window(1) has been removed from the base system. It can now be
805 installed from ports. The port is called misc/window.
808 The way we are storing and accessing `routing table' entries has
809 changed. Programs reading the FIB, like netstat, need to be
813 A new netisr implementation has been added for FreeBSD 8. Network
814 file system modules, such as igmp, ipdivert, and others, should be
816 Bump __FreeBSD_version to 800096.
819 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
823 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
824 Bump __FreeBSD_version to 800094.
827 Add mnt_xflag field to 'struct mount'. File system modules need to be
829 Bump __FreeBSD_version to 800093.
832 The compiling option ADAPTIVE_SX has been retired while it has been
833 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
834 The KPI for sx_init_flags() changes as accepting flags:
835 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
836 been introduced in order to handle the reversed logic.
837 Bump __FreeBSD_version to 800092.
840 Add support for hierarchical jails. Remove global securelevel.
841 Bump __FreeBSD_version to 800091.
844 The layout of struct vnet_net has changed, therefore modules
846 Bump __FreeBSD_version to 800090.
849 The newly imported zic(8) produces a new format in the output. Please
850 run tzsetup(8) to install the newly created data to /etc/localtime.
853 The sysctl tree for the usb stack has renamed from hw.usb2.* to
854 hw.usb.* and is now consistent again with previous releases.
857 802.11 monitor mode support was revised and driver api's were changed.
858 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
859 of DLT_IEEE802_11. No user-visible data structures were changed but
860 applications that use DLT_IEEE802_11 may require changes.
861 Bump __FreeBSD_version to 800088.
864 The layout of the following structs has changed: sysctl_oid,
865 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
866 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
867 panics may be experienced. World rebuild is required for
868 correctly checking networking state from userland.
869 Bump __FreeBSD_version to 800085.
872 MLDv2 and Source-Specific Multicast (SSM) have been merged
873 to the IPv6 stack. VIMAGE hooks are in but not yet used.
874 The implementation of SSM within FreeBSD's IPv6 stack closely
875 follows the IPv4 implementation.
877 For kernel developers:
879 * The most important changes are that the ip6_output() and
880 ip6_input() paths no longer take the IN6_MULTI_LOCK,
881 and this lock has been downgraded to a non-recursive mutex.
883 * As with the changes to the IPv4 stack to support SSM, filtering
884 of inbound multicast traffic must now be performed by transport
885 protocols within the IPv6 stack. This does not apply to TCP and
886 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
888 * The KPIs used by IPv6 multicast are similar to those used by
889 the IPv4 stack, with the following differences:
890 * im6o_mc_filter() is analogous to imo_multicast_filter().
891 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
892 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
893 * IN6_LOOKUP_MULTI() has been deprecated and removed.
894 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
895 for MLDv1 have an additional 'timer' argument which is used to
896 jitter the initial membership report for the solicited-node
897 multicast membership on-link.
898 * This is not strictly needed for MLDv2, which already jitters
899 its report transmissions. However, the 'timer' argument is
900 preserved in case MLDv1 is active on the interface.
902 * The KAME linked-list based IPv6 membership implementation has
903 been refactored to use a vector similar to that used by the IPv4
905 Code which maintains a list of its own multicast memberships
906 internally, e.g. carp, has been updated to reflect the new
909 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
910 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
911 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
912 implementation constraint which needs to be addressed in HEAD.
914 For application developers:
916 * The changes are broadly similar to those made for the IPv4
919 * The use of IPv4 and IPv6 multicast socket options on the same
920 socket, using mapped addresses, HAS NOT been tested or supported.
922 * There are a number of issues with the implementation of various
923 IPv6 multicast APIs which need to be resolved in the API surface
924 before the implementation is fully compatible with KAME userland
925 use, and these are mostly to do with interface index treatment.
927 * The literature available discusses the use of either the delta / ASM
928 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
929 using setsourcefilter(3)/getsourcefilter(3). For more information
930 please refer to RFC 3768, 'Socket Interface Extensions for
931 Multicast Source Filters'.
933 * Applications which use the published RFC 3678 APIs should be fine.
935 For systems administrators:
937 * The mtest(8) utility has been refactored to support IPv6, in
938 addition to IPv4. Interface addresses are no longer accepted
939 as arguments, their names must be used instead. The utility
940 will map the interface name to its first IPv4 address as
941 returned by getifaddrs(3).
943 * The ifmcstat(8) utility has also been updated to print the MLDv2
944 endpoint state and source filter lists via sysctl(3).
946 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
947 loopback of IPv6 multicast datagrams by default; it defaults to 1
948 to preserve the existing behaviour. Disabling multicast loopback is
949 recommended for optimal system performance.
951 * The IPv6 MROUTING code has been changed to examine this sysctl
952 instead of attempting to perform a group lookup before looping
953 back forwarded datagrams.
955 Bump __FreeBSD_version to 800084.
958 Implement low-level Bluetooth HCI API.
959 Bump __FreeBSD_version to 800083.
962 The layout of struct malloc_type, used by modules to register new
963 memory allocation types, has changed. Most modules will need to
964 be rebuilt or panics may be experienced.
965 Bump __FreeBSD_version to 800081.
968 Anticipate overflowing inp_flags - add inp_flags2.
969 This changes most offsets in inpcb, so checking v4 connection
970 state will require a world rebuild.
971 Bump __FreeBSD_version to 800080.
974 Add an llentry to struct route and struct route_in6. Modules
975 embedding a struct route will need to be recompiled.
976 Bump __FreeBSD_version to 800079.
979 The size of rt_metrics_lite and by extension rtentry has changed.
980 Networking administration apps will need to be recompiled.
981 The route command now supports show as an alias for get, weighting
982 of routes, sticky and nostick flags to alter the behavior of stateful
984 Bump __FreeBSD_version to 800078.
987 Do not use Giant for kbdmux(4) locking. This is wrong and
988 apparently causing more problems than it solves. This will
989 re-open the issue where interrupt handlers may race with
990 kbdmux(4) in polling mode. Typical symptoms include (but
991 not limited to) duplicated and/or missing characters when
992 low level console functions (such as gets) are used while
993 interrupts are enabled (for example geli password prompt,
994 mountroot prompt etc.). Disabling kbdmux(4) may help.
997 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
998 kernel modules referencing any of the above need to be recompiled.
999 Bump __FreeBSD_version to 800075.
1002 GEOM_PART has become the default partition slicer for storage devices,
1003 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
1004 introduces some changes:
1006 MSDOS/EBR: the devices created from MSDOS extended partition entries
1007 (EBR) can be named differently than with GEOM_MBR and are now symlinks
1008 to devices with offset-based names. fstabs may need to be modified.
1010 BSD: the "geometry does not match label" warning is harmless in most
1011 cases but it points to problems in file system misalignment with
1012 disk geometry. The "c" partition is now implicit, covers the whole
1013 top-level drive and cannot be (mis)used by users.
1015 General: Kernel dumps are now not allowed to be written to devices
1016 whose partition types indicate they are meant to be used for file
1017 systems (or, in case of MSDOS partitions, as something else than
1020 Most of these changes date approximately from 200812.
1023 The uscanner(4) driver has been removed from the kernel. This follows
1024 Linux removing theirs in 2.6 and making libusb the default interface
1025 (supported by sane).
1028 The multicast forwarding code has been cleaned up. netstat(1)
1029 only relies on KVM now for printing bandwidth upcall meters.
1030 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1031 ip6_mroute_mod respectively. The config(5) options for statically
1032 compiling this code remain the same, i.e. 'options MROUTING'.
1035 Support for the IFF_NEEDSGIANT network interface flag has been
1036 removed, which means that non-MPSAFE network device drivers are no
1037 longer supported. In particular, if_ar, if_sr, and network device
1038 drivers from the old (legacy) USB stack can no longer be built or
1042 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1043 a bunch of new language catalog files have also been added.
1044 This means that some common libc messages are now localized and
1045 they depend on the LC_MESSAGES environmental variable.
1048 The k8temp(4) driver has been renamed to amdtemp(4) since
1049 support for Family 10 and Family 11 CPU families was added.
1052 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1053 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1055 For kernel developers, the most important changes are that the
1056 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1057 and this lock has been downgraded to a non-recursive mutex.
1059 Transport protocols (UDP, Raw IP) are now responsible for filtering
1060 inbound multicast traffic according to group membership and source
1061 filters. The imo_multicast_filter() KPI exists for this purpose.
1062 Transports which do not use multicast (SCTP, TCP) already reject
1063 multicast by default. Forwarding and receive performance may improve
1064 as a mutex acquisition is no longer needed in the ip_input()
1065 low-level input path. in_addmulti() and in_delmulti() are shimmed
1066 to new KPIs which exist to support SSM in-kernel.
1068 For application developers, it is recommended that loopback of
1069 multicast datagrams be disabled for best performance, as this
1070 will still cause the lock to be taken for each looped-back
1071 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1072 be tuned to 0 to disable loopback by default; it defaults to 1
1073 to preserve the existing behaviour.
1075 For systems administrators, to obtain best performance with
1076 multicast reception and multiple groups, it is always recommended
1077 that a card with a suitably precise hash filter is used. Hash
1078 collisions will still result in the lock being taken within the
1079 transport protocol input path to check group membership.
1081 If deploying FreeBSD in an environment with IGMP snooping switches,
1082 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1083 enabled; this forces 224.0.0.0/24 group membership to be announced
1086 The size of 'struct igmpstat' has changed; netstat needs to be
1087 recompiled to reflect this.
1088 Bump __FreeBSD_version to 800070.
1091 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1092 updated to use it. This requires a buildworld/installworld in order to
1093 update the library and dependencies (usbconfig, etc). Its advisable to
1094 rebuild all ports which uses libusb. More specific directions are given
1095 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1096 libusb are no longer required and can be removed.
1099 A workaround is committed to allow the creation of System V shared
1100 memory segment of size > 2 GB on the 64-bit architectures.
1101 Due to a limitation of the existing ABI, the shm_segsz member
1102 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1103 wrong for large segments. Note that limits must be explicitly
1104 raised to allow such segments to be created.
1107 The layout of struct ifnet has changed, requiring a rebuild of all
1108 network device driver modules.
1111 The /dev handling for the new USB stack has changed, a
1112 buildworld/installworld is required for libusb20.
1115 The new USB2 stack has now been permanently moved in and all kernel and
1116 module names reverted to their previous values (eg, usb, ehci, ohci,
1117 ums, ...). The old usb stack can be compiled in by prefixing the name
1118 with the letter 'o', the old usb modules have been removed.
1119 Updating entry 20090216 for xorg and 20090215 for libmap may still
1123 The rc.conf(5) option if_up_delay has been renamed to
1124 defaultroute_delay to better reflect its purpose. If you have
1125 customized this setting in /etc/rc.conf you need to update it to
1129 xorg 7.4 wants to configure its input devices via hald which does not
1130 yet work with USB2. If the keyboard/mouse does not work in xorg then
1132 Option "AllowEmptyInput" "off"
1133 to your ServerLayout section. This will cause X to use the configured
1134 kbd and mouse sections from your xorg.conf.
1137 The GENERIC kernels for all architectures now default to the new USB2
1138 stack. No kernel config options or code have been removed so if a
1139 problem arises please report it and optionally revert to the old USB
1140 stack. If you are loading USB kernel modules or have a custom kernel
1141 that includes GENERIC then ensure that usb names are also changed over,
1142 eg uftdi -> usb2_serial_ftdi.
1144 Older programs linked against the ports libusb 0.1 need to be
1145 redirected to the new stack's libusb20. /etc/libmap.conf can
1147 # Map old usb library to new one for usb2 stack
1148 libusb-0.1.so.8 libusb20.so.1
1151 All USB ethernet devices now attach as interfaces under the name ueN
1152 (eg. ue0). This is to provide a predictable name as vendors often
1153 change usb chipsets in a product without notice.
1156 The ichsmb(4) driver has been changed to require SMBus slave
1157 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1158 All of the other SMBus controller drivers require left-justified
1159 slave addresses, so this change makes all the drivers provide the
1163 INET6 statistics (struct ip6stat) was updated.
1164 netstat(1) needs to be recompiled.
1167 NTFS has been removed from GENERIC kernel on amd64 to match
1168 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1169 will load ntfs.ko module automatically when NTFS support is
1170 actually needed, unless ntfs.ko is not installed or security
1171 level prohibits loading kernel modules. If either is the case,
1172 "options NTFS" has to be added into kernel config.
1175 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1176 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1177 800061. User space tools that rely on the size of struct tcpcb in
1178 tcp_var.h (e.g. sockstat) need to be recompiled.
1181 ng_tty(4) module updated to match the new TTY subsystem.
1182 Due to API change, user-level applications must be updated.
1183 New API support added to mpd5 CVS and expected to be present
1184 in next mpd5.3 release.
1187 With __FreeBSD_version 800060 the makefs tool is part of
1188 the base system (it was a port).
1191 The afdata and ifnet locks have been changed from mutexes to
1192 rwlocks, network modules will need to be re-compiled.
1195 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1196 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1197 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1198 architecture and 40 bytes on 64-bit architecture. The userland
1199 applications "arp" and "ndp" have been updated accordingly.
1200 The output from "netstat -r" shows only routing entries and
1201 none of the L2 information.
1204 __FreeBSD_version 800057 marks the switchover from the
1205 binary ath hal to source code. Users must add the line:
1207 options AH_SUPPORT_AR5416
1209 to their kernel config files when specifying:
1213 The ath_hal module no longer exists; the code is now compiled
1214 together with the driver in the ath module. It is now
1215 possible to tailor chip support (i.e. reduce the set of chips
1216 and thereby the code size); consult ath_hal(4) for details.
1219 __FreeBSD_version 800054 adds memory barriers to
1220 <machine/atomic.h>, new interfaces to ifnet to facilitate
1221 multiple hardware transmit queues for cards that support
1222 them, and a lock-less ring-buffer implementation to
1223 enable drivers to more efficiently manage queueing of
1227 A new version of ZFS (version 13) has been merged to -HEAD.
1228 This version has zpool attribute "listsnapshots" off by
1229 default, which means "zfs list" does not show snapshots,
1230 and is the same as Solaris behavior.
1233 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1236 The uhci, ohci, ehci and slhci USB Host controller drivers have
1237 been put into separate modules. If you load the usb module
1238 separately through loader.conf you will need to load the
1239 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1240 controller add the following to loader.conf:
1246 The ABI used by the PMC toolset has changed. Please keep
1247 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1251 atapci kernel module now includes only generic PCI ATA
1252 driver. AHCI driver moved to ataahci kernel module.
1253 All vendor-specific code moved into separate kernel modules:
1254 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1255 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1256 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1257 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1260 The TTY subsystem of the kernel has been replaced by a new
1261 implementation, which provides better scalability and an
1262 improved driver model. Most common drivers have been migrated to
1263 the new TTY subsystem, while others have not. The following
1264 drivers have not yet been ported to the new TTY layer:
1267 cy, digi, rc, rp, sio
1273 ng_h4, ng_tty, ppp, sl, snp
1275 Adding these drivers to your kernel configuration file shall
1276 cause compilation to fail.
1279 ntpd has been upgraded to 4.2.4p5.
1282 OpenSSH has been upgraded to 5.1p1.
1284 For many years, FreeBSD's version of OpenSSH preferred DSA
1285 over RSA for host and user authentication keys. With this
1286 upgrade, we've switched to the vendor's default of RSA over
1287 DSA. This may cause upgraded clients to warn about unknown
1288 host keys even for previously known hosts. Users should
1289 follow the usual procedure for verifying host keys before
1290 accepting the RSA key.
1292 This can be circumvented by setting the "HostKeyAlgorithms"
1293 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1296 Please note that the sequence of keys offered for
1297 authentication has been changed as well. You may want to
1298 specify IdentityFile in a different order to revert this
1302 The sio(4) driver has been removed from the i386 and amd64
1303 kernel configuration files. This means uart(4) is now the
1304 default serial port driver on those platforms as well.
1306 To prevent collisions with the sio(4) driver, the uart(4) driver
1307 uses different names for its device nodes. This means the
1308 onboard serial port will now most likely be called "ttyu0"
1309 instead of "ttyd0". You may need to reconfigure applications to
1310 use the new device names.
1312 When using the serial port as a boot console, be sure to update
1313 /boot/device.hints and /etc/ttys before booting the new kernel.
1314 If you forget to do so, you can still manually specify the hints
1315 at the loader prompt:
1317 set hint.uart.0.at="isa"
1318 set hint.uart.0.port="0x3F8"
1319 set hint.uart.0.flags="0x10"
1320 set hint.uart.0.irq="4"
1324 The gpt(8) utility has been removed. Use gpart(8) to partition
1328 The version that Linuxulator emulates was changed from 2.4.2
1329 to 2.6.16. If you experience any problems with Linux binaries
1330 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1331 if it fixes the problem contact emulation mailing list.
1334 ISDN4BSD (I4B) was removed from the src tree. You may need to
1335 update a your kernel configuration and remove relevant entries.
1338 I have checked in code to support multiple routing tables.
1339 See the man pages setfib(1) and setfib(2).
1340 This is a hopefully backwards compatible version,
1341 but to make use of it you need to compile your kernel
1342 with options ROUTETABLES=2 (or more up to 16).
1345 The 802.11 wireless support was redone to enable multi-bss
1346 operation on devices that are capable. The underlying device
1347 is no longer used directly but instead wlanX devices are
1348 cloned with ifconfig. This requires changes to rc.conf files.
1349 For example, change:
1350 ifconfig_ath0="WPA DHCP"
1353 ifconfig_wlan0="WPA DHCP"
1354 see rc.conf(5) for more details. In addition, mergemaster of
1355 /etc/rc.d is highly recommended. Simultaneous update of userland
1356 and kernel wouldn't hurt either.
1358 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1359 modules were merged into the base wlan module. All references
1360 to these modules (e.g. in kernel config files) must be removed.
1363 psm(4) has gained write(2) support in native operation level.
1364 Arbitrary commands can be written to /dev/psm%d and status can
1365 be read back from it. Therefore, an application is responsible
1366 for status validation and error recovery. It is a no-op in
1367 other operation levels.
1370 Support for KSE threading has been removed from the kernel. To
1371 run legacy applications linked against KSE libmap.conf may
1372 be used. The following libmap.conf may be used to ensure
1373 compatibility with any prior release:
1375 libpthread.so.1 libthr.so.1
1376 libpthread.so.2 libthr.so.2
1377 libkse.so.3 libthr.so.3
1380 The layout of struct vmspace has changed. This affects libkvm
1381 and any executables that link against libkvm and use the
1382 kvm_getprocs() function. In particular, but not exclusively,
1383 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1384 The effects are minimal, but it's advisable to upgrade world
1388 The latest em driver no longer has support in it for the
1389 82575 adapter, this is now moved to the igb driver. The
1390 split was done to make new features that are incompatible
1391 with older hardware easier to do.
1394 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1395 likewise the kernel option is now GEOM_LINUX_LVM.
1398 The default NFS mount mode has changed from UDP to TCP for
1399 increased reliability. If you rely on (insecurely) NFS
1400 mounting across a firewall you may need to update your
1404 Belatedly note the addition of m_collapse for compacting
1408 The fts(3) structures have been changed to use adequate
1409 integer types for their members and so to be able to cope
1410 with huge file trees. The old fts(3) ABI is preserved
1411 through symbol versioning in libc, so third-party binaries
1412 using fts(3) should still work, although they will not take
1413 advantage of the extended types. At the same time, some
1414 third-party software might fail to build after this change
1415 due to unportable assumptions made in its source code about
1416 fts(3) structure members. Such software should be fixed
1417 by its vendor or, in the worst case, in the ports tree.
1418 FreeBSD_version 800015 marks this change for the unlikely
1419 case that a portable fix is impossible.
1422 To upgrade to -current after this date, you must be running
1423 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1424 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1427 The ADAPTIVE_GIANT kernel option has been retired because its
1428 functionality is the default now.
1431 The AT keyboard emulation of sunkbd(4) has been turned on
1432 by default. In order to make the special symbols of the Sun
1433 keyboards driven by sunkbd(4) work under X these now have
1434 to be configured the same way as Sun USB keyboards driven
1435 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1437 Option "XkbLayout" "us"
1438 Option "XkbRules" "xorg"
1439 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1442 It has been decided that it is desirable to provide ABI
1443 backwards compatibility to the FreeBSD 4/5/6 versions of the
1444 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1445 broken with the introduction of PCI domain support (see the
1446 20070930 entry). Unfortunately, this required the ABI of
1447 PCIOCGETCONF to be broken again in order to be able to
1448 provide backwards compatibility to the old version of that
1449 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1450 again. As for prominent ports this affects neither pciutils
1451 nor xorg-server this time, the hal port needs to be rebuilt
1455 The misnamed kthread_create() and friends have been renamed
1456 to kproc_create() etc. Many of the callers already
1457 used kproc_start()..
1458 I will return kthread_create() and friends in a while
1459 with implementations that actually create threads, not procs.
1460 Renaming corresponds with version 800002.
1469 Avoid using make -j when upgrading. While generally safe, there are
1470 sometimes problems using -j to upgrade. If your upgrade fails with
1471 -j, please try again without -j. From time to time in the past there
1472 have been problems using -j with buildworld and/or installworld. This
1473 is especially true when upgrading between "distant" versions (eg one
1474 that cross a major release boundary or several minor releases, or when
1475 several months have passed on the -current branch).
1477 Sometimes, obscure build problems are the result of environment
1478 poisoning. This can happen because the make utility reads its
1479 environment when searching for values for global variables. To run
1480 your build attempts in an "environmental clean room", prefix all make
1481 commands with 'env -i '. See the env(1) manual page for more details.
1483 When upgrading from one major version to another it is generally best
1484 to upgrade to the latest code in the currently installed branch first,
1485 then do an upgrade to the new branch. This is the best-tested upgrade
1486 path, and has the highest probability of being successful. Please try
1487 this approach before reporting problems with a major version upgrade.
1491 When upgrading the boot ZFS pool to a new version, always follow
1494 1.) recompile and reinstall the ZFS boot loader and boot block
1495 (this is part of "make buildworld" and "make installworld")
1497 2.) update the ZFS boot block on your boot drive
1499 The following example updates the ZFS boot block on the first
1500 partition (freebsd-boot) of a GPT partitioned drive ad0:
1501 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1503 Non-boot pools do not need these updates.
1507 If you are updating from a prior version of FreeBSD (even one just
1508 a few days old), you should follow this procedure. It is the most
1509 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1511 make kernel-toolchain
1512 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1513 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1515 To test a kernel once
1516 ---------------------
1517 If you just want to boot a kernel once (because you are not sure
1518 if it works, or if you want to boot a known bad kernel to provide
1519 debugging information) run
1520 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1521 nextboot -k testkernel
1523 To just build a kernel when you know that it won't mess you up
1524 --------------------------------------------------------------
1525 This assumes you are already running a CURRENT system. Replace
1526 ${arch} with the architecture of your machine (e.g. "i386",
1527 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1529 cd src/sys/${arch}/conf
1530 config KERNEL_NAME_HERE
1531 cd ../compile/KERNEL_NAME_HERE
1536 If this fails, go to the "To build a kernel" section.
1538 To rebuild everything and install it on the current system.
1539 -----------------------------------------------------------
1540 # Note: sometimes if you are running current you gotta do more than
1541 # is listed here if you are upgrading from a really old current.
1543 <make sure you have good level 0 dumps>
1545 make kernel KERNCONF=YOUR_KERNEL_HERE
1547 <reboot in single user> [3]
1555 To cross-install current onto a separate partition
1556 --------------------------------------------------
1557 # In this approach we use a separate partition to hold
1558 # current's root, 'usr', and 'var' directories. A partition
1559 # holding "/", "/usr" and "/var" should be about 2GB in
1562 <make sure you have good level 0 dumps>
1565 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1566 <maybe newfs current's root partition>
1567 <mount current's root partition on directory ${CURRENT_ROOT}>
1568 make installworld DESTDIR=${CURRENT_ROOT}
1569 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1570 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1571 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1572 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1573 <reboot into current>
1574 <do a "native" rebuild/install as described in the previous section>
1575 <maybe install compatibility libraries from ports/misc/compat*>
1579 To upgrade in-place from 8.x-stable to current
1580 ----------------------------------------------
1581 <make sure you have good level 0 dumps>
1583 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1585 <reboot in single user> [3]
1592 Make sure that you've read the UPDATING file to understand the
1593 tweaks to various things you need. At this point in the life
1594 cycle of current, things change often and you are on your own
1595 to cope. The defaults can also change, so please read ALL of
1596 the UPDATING entries.
1598 Also, if you are tracking -current, you must be subscribed to
1599 freebsd-current@freebsd.org. Make sure that before you update
1600 your sources that you have read and understood all the recent
1601 messages there. If in doubt, please track -stable which has
1602 much fewer pitfalls.
1604 [1] If you have third party modules, such as vmware, you
1605 should disable them at this point so they don't crash your
1608 [3] From the bootblocks, boot -s, and then do
1613 adjkerntz -i # if CMOS is wall time
1614 Also, when doing a major release upgrade, it is required that
1615 you boot into single user mode to do the installworld.
1617 [4] Note: This step is non-optional. Failure to do this step
1618 can result in a significant reduction in the functionality of the
1619 system. Attempting to do it by hand is not recommended and those
1620 that pursue this avenue should read this file carefully, as well
1621 as the archives of freebsd-current and freebsd-hackers mailing lists
1622 for potential gotchas. The -U option is also useful to consider.
1623 See mergemaster(8) for more information.
1625 [5] Usually this step is a noop. However, from time to time
1626 you may need to do this if you get unknown user in the following
1627 step. It never hurts to do it all the time. You may need to
1628 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1629 install) after the buildworld before this step if you last updated
1630 from current before 20020224 or from -stable before 20020408.
1632 [6] This only deletes old files and directories. Old libraries
1633 can be deleted by "make delete-old-libs", but you have to make
1634 sure that no program is using those libraries anymore.
1636 [8] In order to have a kernel that can run the 4.x binaries needed to
1637 do an installworld, you must include the COMPAT_FREEBSD4 option in
1638 your kernel. Failure to do so may leave you with a system that is
1639 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1640 required to run the 5.x binaries on more recent kernels. And so on
1641 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1643 Make sure that you merge any new devices from GENERIC since the
1644 last time you updated your kernel config file.
1646 [9] When checking out sources, you must include the -P flag to have
1647 cvs prune empty directories.
1649 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1650 "?=" instead of the "=" assignment operator, so that buildworld can
1651 override the CPUTYPE if it needs to.
1653 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1654 not on the command line, or in /etc/make.conf. buildworld will
1655 warn if it is improperly defined.
1658 This file contains a list, in reverse chronological order, of major
1659 breakages in tracking -current. It is not guaranteed to be a complete
1660 list of such breakages, and only contains entries since October 10, 2007.
1661 If you need to see UPDATING entries from before that date, you will need
1662 to fetch an UPDATING file from an older FreeBSD release.
1664 Copyright information:
1666 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1668 Redistribution, publication, translation and use, with or without
1669 modification, in full or in part, in any form or format of this
1670 document are permitted without further permission from the author.
1672 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1673 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1674 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1675 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1676 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1677 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1678 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1679 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1680 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1681 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1682 POSSIBILITY OF SUCH DAMAGE.
1684 Contact Warner Losh if you have any questions about your use of