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 20140603: p14 FreeBSD-SA-14:11.sendmail
13 FreeBSD-SA-14:12.ktrace
17 Fix sendmail improper close-on-exec flag handling. [SA-14:11]
19 Fix ktrace memory disclosure. [SA-14:12]
21 Fix incorrect error handling in PAM policy parser. [SA-14:13]
23 Fix triple-fault when executing from a threaded process.
26 20140513: p13 FreeBSD-EN-14:03.pkg
27 FreeBSD-EN-14:04.kldxref
30 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
32 Improve build repeatability for kldxref(8). [EN-14:04]
34 Fix data corruption with ciss(4). [EN-14:05]
36 20140430: p12 FreeBSD-SA-14:08.tcp
38 Fix TCP reassembly vulnerability. [SA-14:08]
40 20140408: p11 FreeBSD-SA-14:05.nfsserver
41 FreeBSD-SA-14:06.openssl
42 Fix deadlock in the NFS server. [SA-14:05]
44 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
46 20140114: p10 FreeBSD-SA-14:01.bsnmpd
49 FreeBSD-EN-14:01.random
51 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
53 Fix ntpd distributed reflection Denial of Service
54 vulnerability. [SA-14:02]
56 Fix BIND remote denial of service vulnerability. [SA-14:04]
58 Disable hardware RNGs by default. [EN-14:01]
60 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
62 20131128: p9 FreeBSD-EN-13:05.freebsd-update
63 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
65 20131026: p8 FreeBSD-EN-13:04.freebsd-update
66 Fix multiple freebsd-update bugs that break upgrading to
69 20130910: p7 FreeBSD-SA-13:12.ifioctl
70 FreeBSD-SA-13:13.nullfs
72 In IPv6 and NetATM, stop SIOCSIFADDR, SIOCSIFBRDADDR,
73 SIOCSIFDSTADDR and SIOCSIFNETMASK at the socket layer rather
74 than pass them on to the link layer without validation or
75 credential checks. [SA-13:12]
77 Prevent cross-mount hardlinks between different nullfs mounts
78 of the same underlying filesystem. [SA-13:13]
80 20130822: p6 FreeBSD-SA-13:09.ip_multicast
83 Fix an integer overflow in computing the size of a temporary buffer
84 can result in a buffer which is too small for the requested
87 Fix a bug that could lead to kernel memory disclosure with
88 SCTP state cookie. [13:10]
90 Fix a data corruption problem with mfi(4) operating on > 2TB
91 disks in a JBOD. [EN-13:03]
93 20130726: p5 FreeBSD-SA-13:07.bind FreeBSD-SA-13:08.nfsserver
94 Fix Denial of Service vulnerability in named(8). [13:07]
96 Fix a bug that allows remote client bypass the normal
97 access checks when when -network or -host restrictions are
98 used at the same time with -mapall. [13:08]
100 20130618: p4 FreeBSD-SA-13:06.mmap
101 Fix a bug that allowed a tracing process (e.g. gdb) to write
102 to a memory-mapped file in the traced process's address space
103 even if neither the traced process nor the tracing process had
104 write access to that file.
106 20130429: p3 FreeBSD-SA-13:05.nfsserver
107 Fix a bug that allows NFS clients to issue READDIR on files.
109 20130402: p2 FreeBSD-SA-13:03.openssl FreeBSD-SA-13:04.bind
110 Fix multiple vulnerabilities in OpenSSL.
112 Fix Denial of Service vulnerability in named(8).
114 20130218: p1 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
115 Fix Denial of Service vulnerability in named(8) with DNS64.
117 Fix Denial of Service vulnerability in libc's glob(3) functionality.
123 The sparc64 ZFS loader has been changed to no longer try to auto-
124 detect ZFS providers based on diskN aliases but now requires these
125 to be explicitly listed in the OFW boot-device environment variable.
128 Now unix domain sockets behave "as expected" on nullfs(5). Previously
129 nullfs(5) did not pass through all behaviours to the underlying layer,
130 as a result if we bound to a socket on the lower layer we could connect
131 only to the lower path; if we bound to the upper layer we could connect
132 only to the upper path. The new behavior is one can connect to both the
133 lower and the upper paths regardless what layer path one binds to.
136 The acpi_wmi(4) status device /dev/wmistat has been renamed to
140 A new VOP_ADVISE() was added to support posix_fadvise(2). All
141 filesystem modules must be recompiled.
144 The interface of the VOP_VPTOCNP(9) changed, now the returned
145 vnode shall be referenced, previously it was required to be
146 only held. All in-tree filesystems are converted.
152 The broken amd(4) driver has been replaced with esp(4) in the amd64,
153 i386 and pc98 GENERIC kernel configuration files.
156 This commit modifies vfs_register() so that it uses a hash
157 calculation to set vfc_typenum, which is enabled by default.
158 The first time a system is booted after this change, the
159 vfc_typenum values will change for all file systems. The
160 main effect of this is a change to the NFS server file handles
161 for file systems that use vfc_typenum in their fsid, such as ZFS.
162 It will, however, prevent vfc_typenum from changing when file
163 systems are loaded in a different order for subsequent reboots.
164 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
165 until you are ready to remount all NFS clients after a reboot.
168 Bump the shared library version numbers for libraries that
169 do not use symbol versioning, have changed the ABI compared
170 to stable/8 and which shared library version was not bumped.
171 Done as part of 9.0-RELEASE cycle.
174 During the merge of Capsicum features, the fget(9) KPI was modified.
175 This may require the rebuilding of out-of-tree device drivers --
176 issues have been reported specifically with the nVidia device driver.
177 __FreeBSD_version is bumped to 900041.
179 Also, there is a period between 20110811 and 20110814 where the
180 special devices /dev/{stdin,stdout,stderr} did not work correctly.
181 Building world from a kernel during that window may not work.
184 The packet filter (pf) code has been updated to OpenBSD 4.5.
185 You need to update userland tools to be in sync with kernel.
186 This update breaks backward compatibility with earlier pfsync(4)
187 versions. Care must be taken when updating redundant firewall setups.
190 The following sysctls and tunables are retired on x86 platforms:
192 machdep.hlt_logical_cpus
193 The following sysctl is retired:
194 machdep.hyperthreading_allowed
195 The sysctls were supposed to provide a way to dynamically offline and
196 online selected CPUs on x86 platforms, but the implementation has not
197 been reliable especially with SCHED_ULE scheduler.
198 machdep.hyperthreading_allowed tunable is still available to ignore
199 hyperthreading CPUs at OS level.
200 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
201 where X is an APIC ID of a CPU. Be advised, though, that disabling
202 CPUs in non-uniform fashion will result in non-uniform topology and
203 may lead to sub-optimal system performance with SCHED_ULE, which is
207 cpumask_t type is retired and cpuset_t is used in order to describe
211 Changes to ifconfig(8) for dynamic address family detection mandate
212 that you are running a kernel of 20110525 or later. Make sure to
213 follow the update procedure to boot a new kernel before installing
217 Support for sun4v architecture is officially dropped
220 Several KPI breaking changes have been committed to the mii(4) layer,
221 the PHY drivers and consequently some Ethernet drivers using mii(4).
222 This means that miibus.ko and the modules of the affected Ethernet
223 drivers need to be recompiled.
225 Note to kernel developers: Given that the OUI bit reversion problem
226 was fixed as part of these changes all mii(4) commits related to OUIs,
227 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
228 handling, no longer can be merged verbatim to stable/8 and previous
232 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
233 into their kernel configurations along with 'device pci'.
236 The default NFS client is now the new NFS client, so fstype "newnfs"
237 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
238 Although mounts via fstype "nfs" will usually work without userland
239 changes, it is recommended that the mount(8) and mount_nfs(8)
240 commands be rebuilt from sources and that a link to mount_nfs called
241 mount_oldnfs be created. The new client is compiled into the
242 kernel with "options NFSCL" and this is needed for diskless root
243 file systems. The GENERIC kernel configs have been changed to use
244 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
245 To use the regular/old client, you can "mount -t oldnfs ...". For
246 a diskless root file system, you must also include a line like:
248 vfs.root.mountfrom="oldnfs:"
250 in the boot/loader.conf on the root fs on the NFS server to make
251 a diskless root fs use the old client.
254 The GENERIC kernels for all architectures now default to the new
255 CAM-based ATA stack. It means that all legacy ATA drivers were
256 removed and replaced by respective CAM drivers. If you are using
257 ATA device names in /etc/fstab or other places, make sure to update
258 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
259 where 'Y's are the sequential numbers starting from zero for each type
260 in order of detection, unless configured otherwise with tunables,
261 see cam(4)). There will be symbolic links created in /dev/ to map
262 old adX devices to the respective adaY. They should provide basic
263 compatibility for file systems mounting in most cases, but they do
264 not support old user-level APIs and do not have respective providers
265 in GEOM. Consider using updated management tools with new device names.
267 It is possible to load devices ahci, ata, siis and mvs as modules,
268 but option ATA_CAM should remain in kernel configuration to make ata
269 module work as CAM driver supporting legacy ATA controllers. Device ata
270 still can be used in modular fashion (atacore + ...). Modules atadisk
271 and atapi* are not used and won't affect operation in ATA_CAM mode.
272 Note that to use CAM-based ATA kernel should include CAM devices
273 scbus, pass, da (or explicitly ada), cd and optionally others. All of
274 them are parts of the cam module.
276 ataraid(4) functionality is now supported by the RAID GEOM class.
277 To use it you can load geom_raid kernel module and use graid(8) tool
278 for management. Instead of /dev/arX device names, use /dev/raid/rX.
280 No kernel config options or code have been removed, so if a problem
281 arises, please report it and optionally revert to the old ATA stack.
282 In order to do it you can remove from the kernel config:
287 , and instead add back:
288 device atadisk # ATA disk drives
289 device ataraid # ATA RAID drives
290 device atapicd # ATAPI CDROM drives
291 device atapifd # ATAPI floppy drives
292 device atapist # ATAPI tape drives
295 The default NFS server has been changed to the new server, which
296 was referred to as the experimental server. If you need to switch
297 back to the old NFS server, you must now put the "-o" option on
298 both the mountd and nfsd commands. This can be done using the
299 mountd_flags and nfs_server_flags rc.conf variables until an
300 update to the rc scripts is committed, which is coming soon.
303 The GNU Objective-C runtime library (libobjc), and other Objective-C
304 related components have been removed from the base system. If you
305 require an Objective-C library, please use one of the available ports.
308 ath(4) has been split into bus- and device- modules. if_ath contains
309 the HAL, the TX rate control and the network device code. if_ath_pci
310 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
311 contains the AHB glue. Users need to load both if_ath_pci and if_ath
312 in order to use ath on everything else.
314 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
315 need to load if_ath and if_ath_pci for ath(4) operation.
318 As part of the replacement of sysinstall, the process of building
319 release media has changed significantly. For details, please re-read
320 release(7), which has been updated to reflect the new build process.
323 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
324 is the last available version under GPLv2. It brings a number of new
325 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
326 4.1 and SSE 4.2), better support for powerpc64, a number of new
327 directives, and lots of other small improvements. See the ChangeLog
328 file in contrib/binutils for the full details.
331 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
332 compliant, and will now use half of hash for authentication.
333 This will break interoperability with all stacks (including all
334 actual FreeBSD versions) who implement
335 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
337 The only workaround with such peers is to use another HMAC
338 algorithm for IPsec ("phase 2") authentication.
341 Remove the uio_yield prototype and symbol. This function has
342 been misnamed since it was introduced and should not be
343 globally exposed with this name. The equivalent functionality
344 is now available using kern_yield(curthread->td_user_pri).
345 The function remains undocumented.
348 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
349 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
350 for scalar sysctls is defined but disabled. Code that needs
351 UQUAD to pass the type checking that must compile on older
352 systems where the define is not present can check against
353 __FreeBSD_version >= 900030.
355 The system dialog(1) has been replaced with a new version previously
356 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
357 with the previous version, but the libdialog associated with it has
358 a largely incompatible API. As such, the original version of libdialog
359 will be kept temporarily as libodialog, until its base system consumers
360 are replaced or updated. Bump __FreeBSD_version to 900030.
363 If you are trying to run make universe on a -stable system, and you get
364 the following warning:
365 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
366 unknown. config(8) likely too old."
367 or something similar to it, then you must upgrade your -stable system
368 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
369 stable/8) or build the config from the latest stable/8 branch and
370 install it on your system.
372 Prior to this date, building a current universe on 8-stable system from
373 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
374 error in the first kernel build phase. A new config on those old
375 systems will fix that problem for older versions of -current.
378 The TCP stack has been modified to allow Khelp modules to interact with
379 it via helper hook points and store per-connection data in the TCP
380 control block. Bump __FreeBSD_version to 900029. User space tools that
381 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
385 Generic IEEE 802.3 annex 31B full duplex flow control support has been
386 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
387 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
388 to take advantage of it instead of using custom implementations. This
389 means that these drivers now no longer unconditionally advertise
390 support for flow control but only do so if flow control is a selected
391 media option. This was implemented in the generic support that way in
392 order to allow flow control to be switched on and off via ifconfig(8)
393 with the PHY specific default to typically off in order to protect
394 from unwanted effects. Consequently, if you used flow control with
395 one of the above mentioned drivers you now need to explicitly enable
397 ifconfig bge0 media auto mediaopt flowcontrol
399 Along with the above mentioned changes generic support for setting
400 1000baseT master mode also has been added and brgphy(4), ciphy(4),
401 e1000phy(4) as well as ip1000phy(4) have been converted to take
402 advantage of it. This means that these drivers now no longer take the
403 link0 parameter for selecting master mode but the master media option
404 has to be used instead, for example like in the following:
405 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
407 Selection of master mode now is also available with all other PHY
408 drivers supporting 1000baseT.
411 The TCP stack has received a significant update to add support for
412 modularised congestion control and generally improve the clarity of
413 congestion control decisions. Bump __FreeBSD_version to 900025. User
414 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
415 sockstat) need to be recompiled.
418 The man(1) utility has been replaced by a new version that no longer
419 uses /etc/manpath.config. Please consult man.conf(5) for how to
420 migrate local entries to the new format.
423 The copyright strings printed by login(1) and sshd(8) at the time of a
424 new connection have been removed to follow other operating systems and
428 A workaround for a fixed ld bug has been removed in kernel code,
429 so make sure that your system ld is built from sources after
430 revision 210245 from 2010-07-19 (r211583 if building head kernel
431 on stable/8, r211584 for stable/7; both from 2010-08-21).
432 A symptom of incorrect ld version is different addresses for
433 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
436 The $ipv6_prefer variable in rc.conf(5) has been split into
437 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
439 The $ip6addrctl_policy is a variable to choose a pre-defined
440 address selection policy set by ip6addrctl(8). A value
441 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
444 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
445 flag (see an entry of 20090926) is set on an interface with no
446 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
447 security reason. If you want IPv6 link-local address on all
448 interfaces by default, set this to "YES".
450 The old ipv6_prefer="YES" is equivalent to
451 ipv6_activate_all_interfaces="YES" and
452 ip6addrctl_policy="ipv6_prefer".
455 DTrace has grown support for userland tracing. Due to this, DTrace is
456 now i386 and amd64 only.
457 dtruss(1) is now installed by default on those systems and a new
458 kernel module is needed for userland tracing: fasttrap.
459 No changes to your kernel config file are necessary to enable
460 userland tracing, but you might consider adding 'STRIP=' and
461 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
462 to have informative userland stack traces in DTrace (ustack).
465 The acpi_aiboost(4) driver has been removed in favor of the new
466 aibs(4) driver. You should update your kernel configuration file.
469 BSD grep has been imported to the base system and it is built by
470 default. It is completely BSD licensed, highly GNU-compatible, uses
471 less memory than its GNU counterpart and has a small codebase.
472 However, it is slower than its GNU counterpart, which is mostly
473 noticeable for larger searches, for smaller ones it is measurable
474 but not significant. The reason is complex, the most important factor
475 is that we lack a modern and efficient regex library and GNU
476 overcomes this by optimizing the searches internally. Future work
477 on improving the regex performance is planned, for the meantime,
478 users that need better performance, can build GNU grep instead by
479 setting the WITH_GNU_GREP knob.
482 Due to the import of powerpc64 support, all existing powerpc kernel
483 configuration files must be updated with a machine directive like this:
484 machine powerpc powerpc
486 In addition, an updated config(8) is required to build powerpc kernels
490 A new version of ZFS (version 15) has been merged to -HEAD.
491 This version uses a python library for the following subcommands:
492 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
493 For full functionality of these commands the following port must
494 be installed: sysutils/py-zfs
497 'vm_page's are now hashed by physical address to an array of mutexes.
498 Currently this is only used to serialize access to hold_count. Over
499 time the page queue mutex will be peeled away. This changes the size
500 of pmap on every architecture. And requires all callers of vm_page_hold
501 and vm_page_unhold to be updated.
504 WITH_CTF can now be specified in src.conf (not recommended, there
505 are some problems with static executables), make.conf (would also
506 affect ports which do not use GNU make and do not override the
507 compile targets) or in the kernel config (via "makeoptions
509 When WITH_CTF was specified there before this was silently ignored,
510 so make sure that WITH_CTF is not used in places which could lead
511 to unwanted behavior.
514 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
515 to allow 32-bit compatibility on non-x86 platforms. All kernel
516 configurations on amd64 and ia64 platforms using these options must
517 be modified accordingly.
520 The utmp user accounting database has been replaced with utmpx,
521 the user accounting interface standardized by POSIX.
522 Unfortunately the semantics of utmp and utmpx don't match,
523 making it practically impossible to support both interfaces.
524 The user accounting database is used by tools like finger(1),
525 last(1), talk(1), w(1) and ac(8).
527 All applications in the base system use utmpx. This means only
528 local binaries (e.g. from the ports tree) may still use these
529 utmp database files. These applications must be rebuilt to make
532 After the system has been upgraded, it is safe to remove the old
533 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
534 assuming their contents is of no importance anymore. Old wtmp
535 databases can only be used by last(1) and ac(8) after they have
536 been converted to the new format using wtmpcvt(1).
539 Introduce the kernel thread "deadlock resolver" (which can be enabled
540 via the DEADLKRES option, see NOTES for more details) and the
541 sleepq_type() function for sleepqueues.
544 The rc.firewall and rc.firewall6 were unified, and
545 rc.firewall6 and rc.d/ip6fw were removed.
546 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
547 variables are obsoleted. Instead, the following new rc
548 variables are added to rc.d/ipfw:
550 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
551 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
552 firewall_simple_onet_ipv6, firewall_trusted_ipv6
554 The meanings correspond to the relevant IPv4 variables.
560 The default terminal emulation for syscons(4) has been changed
561 from cons25 to xterm on all platforms except pc98. This means
562 that the /etc/ttys file needs to be updated to ensure correct
563 operation of applications on the console.
565 The terminal emulation style can be toggled per window by using
566 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
567 options can be used to change the compile-time default back to
570 To prevent graphical artifacts, make sure the TERM environment
571 variable is set to match the terminal emulation that is being
572 performed by syscons(4).
575 The layout of the structure ieee80211req_scan_result has changed.
576 Applications that require wireless scan results (e.g. ifconfig(8))
577 from net80211 need to be recompiled.
579 Applications such as wpa_supplicant(8) may require a full world
580 build without using NO_CLEAN in order to get synchronized with the
584 The iwn(4) driver has been updated to support the 5000 and 5150 series.
585 There's one kernel module for each firmware. Adding "device iwnfw"
586 to the kernel configuration file means including all three firmware
587 images inside the kernel. If you want to include just the one for
588 your wireless card, use the devices iwn4965fw, iwn5000fw or
592 The rc.d/network_ipv6, IPv6 configuration script has been integrated
593 into rc.d/netif. The changes are the following:
595 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
596 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
597 Note that both variables need the "inet6" keyword at the head.
599 Do not set $ipv6_network_interfaces manually if you do not
600 understand what you are doing. It is not needed in most cases.
602 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
605 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
606 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
608 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
609 all configured interfaces have "inet6 accept_rtadv" in the
610 $ifconfig_IF_ipv6. These are for backward compatibility.
612 3. A new variable $ipv6_prefer has been added. If NO, IPv6
613 functionality of interfaces with no corresponding
614 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
615 and the default address selection policy of ip6addrctl(8)
616 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
617 Note that if you want to configure IPv6 functionality on the
618 disabled interfaces after boot, first you need to clear the flag by
619 using ifconfig(8) like:
621 ifconfig em0 inet6 -ifdisabled
623 If YES, the default address selection policy is set as
626 The default value of $ipv6_prefer is NO.
628 4. If your system need to receive Router Advertisement messages,
629 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
630 scripts automatically invoke rtsol(8) when the interface becomes
631 UP. The Router Advertisement messages are used for SLAAC
632 (State-Less Address AutoConfiguration).
635 802.11s D3.03 support was committed. This is incompatible with the
636 previous code, which was based on D3.0.
639 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
640 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
641 control whether accepting Router Advertisement messages or not.
642 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
643 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
644 The ifconfig(8) utility now supports these flags.
647 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
648 mount(8) and -a option for df(1) to see them.
651 The old tunable hw.bus.devctl_disable has been superseded by
652 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
653 replaced by hw.bus.devctl_queue=0. The default for this new tunable
657 Remove the option STOP_NMI. The default action is now to use NMI only
658 for KDB via the newly introduced function stop_cpus_hard() and
659 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
662 The stable/8 branch created in subversion. This corresponds to the
663 RELENG_8 branch in CVS.
666 Bump the shared library version numbers for all libraries that do not
667 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
668 __FreeBSD_version to 800105.
671 Due to changes in the implementation of virtual network stack support,
672 all network-related kernel modules must be recompiled. As this change
673 breaks the ABI, bump __FreeBSD_version to 800104.
676 The TOE interface to the TCP syncache has been modified to remove
677 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
678 The cxgb driver is the only TOE consumer affected by this change, and
679 needs to be recompiled along with the kernel. As this change breaks
680 the ABI, bump __FreeBSD_version to 800103.
683 Padding has been added to struct tcpcb, sackhint and tcpstat in
684 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
685 maintaining the ABI. However, this change breaks the ABI, so bump
686 __FreeBSD_version to 800102. User space tools that rely on the size of
687 any of these structs (e.g. sockstat) need to be recompiled.
690 The NFS_LEGACYRPC option has been removed along with the old kernel
691 RPC implementation that this option selected. Kernel configurations
692 may need to be adjusted.
695 The network interface device nodes at /dev/net/<interface> have been
696 removed. All ioctl operations can be performed the normal way using
697 routing sockets. The kqueue functionality can generally be replaced
698 with routing sockets.
701 The documentation from the FreeBSD Documentation Project (Handbook,
702 FAQ, etc.) is now installed via packages by sysinstall(8) and under
703 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
706 The ABI of various structures related to the SYSV IPC API have been
707 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
708 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
712 Layout of struct vnet has changed as routing related variables were
713 moved to their own Vimage module. Modules need to be recompiled. Bump
714 __FreeBSD_version to 800099.
717 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
718 respectively. As long as no more than 16 groups per process are used,
719 no changes should be visible. When more than 16 groups are used, old
720 binaries may fail if they call getgroups() or getgrouplist() with
721 statically sized storage. Recompiling will work around this, but
722 applications should be modified to use dynamically allocated storage
723 for group arrays as POSIX.1-2008 does not cap an implementation's
724 number of supported groups at NGROUPS_MAX+1 as previous versions did.
726 NFS and portalfs mounts may also be affected as the list of groups is
727 truncated to 16. Users of NFS who use more than 16 groups, should
728 take care that negative group permissions are not used on the exported
729 file systems as they will not be reliable unless a GSSAPI based
730 authentication method is used.
733 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
734 option compiles in the support for adaptive spinning for lockmgrs
735 which want to enable it. The lockinit() function now accepts the flag
736 LK_ADAPTIVE in order to make the lock object subject to adaptive
737 spinning when both held in write and read mode.
740 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
741 changed. User applications that use this ioctl need to be rebuilt.
744 The layout of struct thread has changed. Kernel and modules need to
748 The layout of structs ifnet, domain, protosw and vnet_net has changed.
749 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
752 window(1) has been removed from the base system. It can now be
753 installed from ports. The port is called misc/window.
756 The way we are storing and accessing `routing table' entries has
757 changed. Programs reading the FIB, like netstat, need to be
761 A new netisr implementation has been added for FreeBSD 8. Network
762 file system modules, such as igmp, ipdivert, and others, should be
764 Bump __FreeBSD_version to 800096.
767 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
771 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
772 Bump __FreeBSD_version to 800094.
775 Add mnt_xflag field to 'struct mount'. File system modules need to be
777 Bump __FreeBSD_version to 800093.
780 The compiling option ADAPTIVE_SX has been retired while it has been
781 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
782 The KPI for sx_init_flags() changes as accepting flags:
783 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
784 been introduced in order to handle the reversed logic.
785 Bump __FreeBSD_version to 800092.
788 Add support for hierarchical jails. Remove global securelevel.
789 Bump __FreeBSD_version to 800091.
792 The layout of struct vnet_net has changed, therefore modules
794 Bump __FreeBSD_version to 800090.
797 The newly imported zic(8) produces a new format in the output. Please
798 run tzsetup(8) to install the newly created data to /etc/localtime.
801 The sysctl tree for the usb stack has renamed from hw.usb2.* to
802 hw.usb.* and is now consistent again with previous releases.
805 802.11 monitor mode support was revised and driver api's were changed.
806 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
807 of DLT_IEEE802_11. No user-visible data structures were changed but
808 applications that use DLT_IEEE802_11 may require changes.
809 Bump __FreeBSD_version to 800088.
812 The layout of the following structs has changed: sysctl_oid,
813 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
814 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
815 panics may be experienced. World rebuild is required for
816 correctly checking networking state from userland.
817 Bump __FreeBSD_version to 800085.
820 MLDv2 and Source-Specific Multicast (SSM) have been merged
821 to the IPv6 stack. VIMAGE hooks are in but not yet used.
822 The implementation of SSM within FreeBSD's IPv6 stack closely
823 follows the IPv4 implementation.
825 For kernel developers:
827 * The most important changes are that the ip6_output() and
828 ip6_input() paths no longer take the IN6_MULTI_LOCK,
829 and this lock has been downgraded to a non-recursive mutex.
831 * As with the changes to the IPv4 stack to support SSM, filtering
832 of inbound multicast traffic must now be performed by transport
833 protocols within the IPv6 stack. This does not apply to TCP and
834 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
836 * The KPIs used by IPv6 multicast are similar to those used by
837 the IPv4 stack, with the following differences:
838 * im6o_mc_filter() is analogous to imo_multicast_filter().
839 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
840 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
841 * IN6_LOOKUP_MULTI() has been deprecated and removed.
842 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
843 for MLDv1 have an additional 'timer' argument which is used to
844 jitter the initial membership report for the solicited-node
845 multicast membership on-link.
846 * This is not strictly needed for MLDv2, which already jitters
847 its report transmissions. However, the 'timer' argument is
848 preserved in case MLDv1 is active on the interface.
850 * The KAME linked-list based IPv6 membership implementation has
851 been refactored to use a vector similar to that used by the IPv4
853 Code which maintains a list of its own multicast memberships
854 internally, e.g. carp, has been updated to reflect the new
857 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
858 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
859 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
860 implementation constraint which needs to be addressed in HEAD.
862 For application developers:
864 * The changes are broadly similar to those made for the IPv4
867 * The use of IPv4 and IPv6 multicast socket options on the same
868 socket, using mapped addresses, HAS NOT been tested or supported.
870 * There are a number of issues with the implementation of various
871 IPv6 multicast APIs which need to be resolved in the API surface
872 before the implementation is fully compatible with KAME userland
873 use, and these are mostly to do with interface index treatment.
875 * The literature available discusses the use of either the delta / ASM
876 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
877 using setsourcefilter(3)/getsourcefilter(3). For more information
878 please refer to RFC 3768, 'Socket Interface Extensions for
879 Multicast Source Filters'.
881 * Applications which use the published RFC 3678 APIs should be fine.
883 For systems administrators:
885 * The mtest(8) utility has been refactored to support IPv6, in
886 addition to IPv4. Interface addresses are no longer accepted
887 as arguments, their names must be used instead. The utility
888 will map the interface name to its first IPv4 address as
889 returned by getifaddrs(3).
891 * The ifmcstat(8) utility has also been updated to print the MLDv2
892 endpoint state and source filter lists via sysctl(3).
894 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
895 loopback of IPv6 multicast datagrams by default; it defaults to 1
896 to preserve the existing behaviour. Disabling multicast loopback is
897 recommended for optimal system performance.
899 * The IPv6 MROUTING code has been changed to examine this sysctl
900 instead of attempting to perform a group lookup before looping
901 back forwarded datagrams.
903 Bump __FreeBSD_version to 800084.
906 Implement low-level Bluetooth HCI API.
907 Bump __FreeBSD_version to 800083.
910 The layout of struct malloc_type, used by modules to register new
911 memory allocation types, has changed. Most modules will need to
912 be rebuilt or panics may be experienced.
913 Bump __FreeBSD_version to 800081.
916 Anticipate overflowing inp_flags - add inp_flags2.
917 This changes most offsets in inpcb, so checking v4 connection
918 state will require a world rebuild.
919 Bump __FreeBSD_version to 800080.
922 Add an llentry to struct route and struct route_in6. Modules
923 embedding a struct route will need to be recompiled.
924 Bump __FreeBSD_version to 800079.
927 The size of rt_metrics_lite and by extension rtentry has changed.
928 Networking administration apps will need to be recompiled.
929 The route command now supports show as an alias for get, weighting
930 of routes, sticky and nostick flags to alter the behavior of stateful
932 Bump __FreeBSD_version to 800078.
935 Do not use Giant for kbdmux(4) locking. This is wrong and
936 apparently causing more problems than it solves. This will
937 re-open the issue where interrupt handlers may race with
938 kbdmux(4) in polling mode. Typical symptoms include (but
939 not limited to) duplicated and/or missing characters when
940 low level console functions (such as gets) are used while
941 interrupts are enabled (for example geli password prompt,
942 mountroot prompt etc.). Disabling kbdmux(4) may help.
945 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
946 kernel modules referencing any of the above need to be recompiled.
947 Bump __FreeBSD_version to 800075.
950 GEOM_PART has become the default partition slicer for storage devices,
951 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
952 introduces some changes:
954 MSDOS/EBR: the devices created from MSDOS extended partition entries
955 (EBR) can be named differently than with GEOM_MBR and are now symlinks
956 to devices with offset-based names. fstabs may need to be modified.
958 BSD: the "geometry does not match label" warning is harmless in most
959 cases but it points to problems in file system misalignment with
960 disk geometry. The "c" partition is now implicit, covers the whole
961 top-level drive and cannot be (mis)used by users.
963 General: Kernel dumps are now not allowed to be written to devices
964 whose partition types indicate they are meant to be used for file
965 systems (or, in case of MSDOS partitions, as something else than
968 Most of these changes date approximately from 200812.
971 The uscanner(4) driver has been removed from the kernel. This follows
972 Linux removing theirs in 2.6 and making libusb the default interface
976 The multicast forwarding code has been cleaned up. netstat(1)
977 only relies on KVM now for printing bandwidth upcall meters.
978 The IPv4 and IPv6 modules are split into ip_mroute_mod and
979 ip6_mroute_mod respectively. The config(5) options for statically
980 compiling this code remain the same, i.e. 'options MROUTING'.
983 Support for the IFF_NEEDSGIANT network interface flag has been
984 removed, which means that non-MPSAFE network device drivers are no
985 longer supported. In particular, if_ar, if_sr, and network device
986 drivers from the old (legacy) USB stack can no longer be built or
990 POSIX.1 Native Language Support (NLS) has been enabled in libc and
991 a bunch of new language catalog files have also been added.
992 This means that some common libc messages are now localized and
993 they depend on the LC_MESSAGES environmental variable.
996 The k8temp(4) driver has been renamed to amdtemp(4) since
997 support for Family 10 and Family 11 CPU families was added.
1000 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1001 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1003 For kernel developers, the most important changes are that the
1004 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1005 and this lock has been downgraded to a non-recursive mutex.
1007 Transport protocols (UDP, Raw IP) are now responsible for filtering
1008 inbound multicast traffic according to group membership and source
1009 filters. The imo_multicast_filter() KPI exists for this purpose.
1010 Transports which do not use multicast (SCTP, TCP) already reject
1011 multicast by default. Forwarding and receive performance may improve
1012 as a mutex acquisition is no longer needed in the ip_input()
1013 low-level input path. in_addmulti() and in_delmulti() are shimmed
1014 to new KPIs which exist to support SSM in-kernel.
1016 For application developers, it is recommended that loopback of
1017 multicast datagrams be disabled for best performance, as this
1018 will still cause the lock to be taken for each looped-back
1019 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1020 be tuned to 0 to disable loopback by default; it defaults to 1
1021 to preserve the existing behaviour.
1023 For systems administrators, to obtain best performance with
1024 multicast reception and multiple groups, it is always recommended
1025 that a card with a suitably precise hash filter is used. Hash
1026 collisions will still result in the lock being taken within the
1027 transport protocol input path to check group membership.
1029 If deploying FreeBSD in an environment with IGMP snooping switches,
1030 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1031 enabled; this forces 224.0.0.0/24 group membership to be announced
1034 The size of 'struct igmpstat' has changed; netstat needs to be
1035 recompiled to reflect this.
1036 Bump __FreeBSD_version to 800070.
1039 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1040 updated to use it. This requires a buildworld/installworld in order to
1041 update the library and dependencies (usbconfig, etc). Its advisable to
1042 rebuild all ports which uses libusb. More specific directions are given
1043 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1044 libusb are no longer required and can be removed.
1047 A workaround is committed to allow the creation of System V shared
1048 memory segment of size > 2 GB on the 64-bit architectures.
1049 Due to a limitation of the existing ABI, the shm_segsz member
1050 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1051 wrong for large segments. Note that limits must be explicitly
1052 raised to allow such segments to be created.
1055 The layout of struct ifnet has changed, requiring a rebuild of all
1056 network device driver modules.
1059 The /dev handling for the new USB stack has changed, a
1060 buildworld/installworld is required for libusb20.
1063 The new USB2 stack has now been permanently moved in and all kernel and
1064 module names reverted to their previous values (eg, usb, ehci, ohci,
1065 ums, ...). The old usb stack can be compiled in by prefixing the name
1066 with the letter 'o', the old usb modules have been removed.
1067 Updating entry 20090216 for xorg and 20090215 for libmap may still
1071 The rc.conf(5) option if_up_delay has been renamed to
1072 defaultroute_delay to better reflect its purpose. If you have
1073 customized this setting in /etc/rc.conf you need to update it to
1077 xorg 7.4 wants to configure its input devices via hald which does not
1078 yet work with USB2. If the keyboard/mouse does not work in xorg then
1080 Option "AllowEmptyInput" "off"
1081 to your ServerLayout section. This will cause X to use the configured
1082 kbd and mouse sections from your xorg.conf.
1085 The GENERIC kernels for all architectures now default to the new USB2
1086 stack. No kernel config options or code have been removed so if a
1087 problem arises please report it and optionally revert to the old USB
1088 stack. If you are loading USB kernel modules or have a custom kernel
1089 that includes GENERIC then ensure that usb names are also changed over,
1090 eg uftdi -> usb2_serial_ftdi.
1092 Older programs linked against the ports libusb 0.1 need to be
1093 redirected to the new stack's libusb20. /etc/libmap.conf can
1095 # Map old usb library to new one for usb2 stack
1096 libusb-0.1.so.8 libusb20.so.1
1099 All USB ethernet devices now attach as interfaces under the name ueN
1100 (eg. ue0). This is to provide a predictable name as vendors often
1101 change usb chipsets in a product without notice.
1104 The ichsmb(4) driver has been changed to require SMBus slave
1105 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1106 All of the other SMBus controller drivers require left-justified
1107 slave addresses, so this change makes all the drivers provide the
1111 INET6 statistics (struct ip6stat) was updated.
1112 netstat(1) needs to be recompiled.
1115 NTFS has been removed from GENERIC kernel on amd64 to match
1116 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1117 will load ntfs.ko module automatically when NTFS support is
1118 actually needed, unless ntfs.ko is not installed or security
1119 level prohibits loading kernel modules. If either is the case,
1120 "options NTFS" has to be added into kernel config.
1123 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1124 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1125 800061. User space tools that rely on the size of struct tcpcb in
1126 tcp_var.h (e.g. sockstat) need to be recompiled.
1129 ng_tty(4) module updated to match the new TTY subsystem.
1130 Due to API change, user-level applications must be updated.
1131 New API support added to mpd5 CVS and expected to be present
1132 in next mpd5.3 release.
1135 With __FreeBSD_version 800060 the makefs tool is part of
1136 the base system (it was a port).
1139 The afdata and ifnet locks have been changed from mutexes to
1140 rwlocks, network modules will need to be re-compiled.
1143 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1144 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1145 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1146 architecture and 40 bytes on 64-bit architecture. The userland
1147 applications "arp" and "ndp" have been updated accordingly.
1148 The output from "netstat -r" shows only routing entries and
1149 none of the L2 information.
1152 __FreeBSD_version 800057 marks the switchover from the
1153 binary ath hal to source code. Users must add the line:
1155 options AH_SUPPORT_AR5416
1157 to their kernel config files when specifying:
1161 The ath_hal module no longer exists; the code is now compiled
1162 together with the driver in the ath module. It is now
1163 possible to tailor chip support (i.e. reduce the set of chips
1164 and thereby the code size); consult ath_hal(4) for details.
1167 __FreeBSD_version 800054 adds memory barriers to
1168 <machine/atomic.h>, new interfaces to ifnet to facilitate
1169 multiple hardware transmit queues for cards that support
1170 them, and a lock-less ring-buffer implementation to
1171 enable drivers to more efficiently manage queueing of
1175 A new version of ZFS (version 13) has been merged to -HEAD.
1176 This version has zpool attribute "listsnapshots" off by
1177 default, which means "zfs list" does not show snapshots,
1178 and is the same as Solaris behavior.
1181 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1184 The uhci, ohci, ehci and slhci USB Host controller drivers have
1185 been put into separate modules. If you load the usb module
1186 separately through loader.conf you will need to load the
1187 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1188 controller add the following to loader.conf:
1194 The ABI used by the PMC toolset has changed. Please keep
1195 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1199 atapci kernel module now includes only generic PCI ATA
1200 driver. AHCI driver moved to ataahci kernel module.
1201 All vendor-specific code moved into separate kernel modules:
1202 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1203 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1204 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1205 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1208 The TTY subsystem of the kernel has been replaced by a new
1209 implementation, which provides better scalability and an
1210 improved driver model. Most common drivers have been migrated to
1211 the new TTY subsystem, while others have not. The following
1212 drivers have not yet been ported to the new TTY layer:
1215 cy, digi, rc, rp, sio
1221 ng_h4, ng_tty, ppp, sl, snp
1223 Adding these drivers to your kernel configuration file shall
1224 cause compilation to fail.
1227 ntpd has been upgraded to 4.2.4p5.
1230 OpenSSH has been upgraded to 5.1p1.
1232 For many years, FreeBSD's version of OpenSSH preferred DSA
1233 over RSA for host and user authentication keys. With this
1234 upgrade, we've switched to the vendor's default of RSA over
1235 DSA. This may cause upgraded clients to warn about unknown
1236 host keys even for previously known hosts. Users should
1237 follow the usual procedure for verifying host keys before
1238 accepting the RSA key.
1240 This can be circumvented by setting the "HostKeyAlgorithms"
1241 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1244 Please note that the sequence of keys offered for
1245 authentication has been changed as well. You may want to
1246 specify IdentityFile in a different order to revert this
1250 The sio(4) driver has been removed from the i386 and amd64
1251 kernel configuration files. This means uart(4) is now the
1252 default serial port driver on those platforms as well.
1254 To prevent collisions with the sio(4) driver, the uart(4) driver
1255 uses different names for its device nodes. This means the
1256 onboard serial port will now most likely be called "ttyu0"
1257 instead of "ttyd0". You may need to reconfigure applications to
1258 use the new device names.
1260 When using the serial port as a boot console, be sure to update
1261 /boot/device.hints and /etc/ttys before booting the new kernel.
1262 If you forget to do so, you can still manually specify the hints
1263 at the loader prompt:
1265 set hint.uart.0.at="isa"
1266 set hint.uart.0.port="0x3F8"
1267 set hint.uart.0.flags="0x10"
1268 set hint.uart.0.irq="4"
1272 The gpt(8) utility has been removed. Use gpart(8) to partition
1276 The version that Linuxulator emulates was changed from 2.4.2
1277 to 2.6.16. If you experience any problems with Linux binaries
1278 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1279 if it fixes the problem contact emulation mailing list.
1282 ISDN4BSD (I4B) was removed from the src tree. You may need to
1283 update a your kernel configuration and remove relevant entries.
1286 I have checked in code to support multiple routing tables.
1287 See the man pages setfib(1) and setfib(2).
1288 This is a hopefully backwards compatible version,
1289 but to make use of it you need to compile your kernel
1290 with options ROUTETABLES=2 (or more up to 16).
1293 The 802.11 wireless support was redone to enable multi-bss
1294 operation on devices that are capable. The underlying device
1295 is no longer used directly but instead wlanX devices are
1296 cloned with ifconfig. This requires changes to rc.conf files.
1297 For example, change:
1298 ifconfig_ath0="WPA DHCP"
1301 ifconfig_wlan0="WPA DHCP"
1302 see rc.conf(5) for more details. In addition, mergemaster of
1303 /etc/rc.d is highly recommended. Simultaneous update of userland
1304 and kernel wouldn't hurt either.
1306 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1307 modules were merged into the base wlan module. All references
1308 to these modules (e.g. in kernel config files) must be removed.
1311 psm(4) has gained write(2) support in native operation level.
1312 Arbitrary commands can be written to /dev/psm%d and status can
1313 be read back from it. Therefore, an application is responsible
1314 for status validation and error recovery. It is a no-op in
1315 other operation levels.
1318 Support for KSE threading has been removed from the kernel. To
1319 run legacy applications linked against KSE libmap.conf may
1320 be used. The following libmap.conf may be used to ensure
1321 compatibility with any prior release:
1323 libpthread.so.1 libthr.so.1
1324 libpthread.so.2 libthr.so.2
1325 libkse.so.3 libthr.so.3
1328 The layout of struct vmspace has changed. This affects libkvm
1329 and any executables that link against libkvm and use the
1330 kvm_getprocs() function. In particular, but not exclusively,
1331 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1332 The effects are minimal, but it's advisable to upgrade world
1336 The latest em driver no longer has support in it for the
1337 82575 adapter, this is now moved to the igb driver. The
1338 split was done to make new features that are incompatible
1339 with older hardware easier to do.
1342 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1343 likewise the kernel option is now GEOM_LINUX_LVM.
1346 The default NFS mount mode has changed from UDP to TCP for
1347 increased reliability. If you rely on (insecurely) NFS
1348 mounting across a firewall you may need to update your
1352 Belatedly note the addition of m_collapse for compacting
1356 The fts(3) structures have been changed to use adequate
1357 integer types for their members and so to be able to cope
1358 with huge file trees. The old fts(3) ABI is preserved
1359 through symbol versioning in libc, so third-party binaries
1360 using fts(3) should still work, although they will not take
1361 advantage of the extended types. At the same time, some
1362 third-party software might fail to build after this change
1363 due to unportable assumptions made in its source code about
1364 fts(3) structure members. Such software should be fixed
1365 by its vendor or, in the worst case, in the ports tree.
1366 FreeBSD_version 800015 marks this change for the unlikely
1367 case that a portable fix is impossible.
1370 To upgrade to -current after this date, you must be running
1371 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1372 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1375 The ADAPTIVE_GIANT kernel option has been retired because its
1376 functionality is the default now.
1379 The AT keyboard emulation of sunkbd(4) has been turned on
1380 by default. In order to make the special symbols of the Sun
1381 keyboards driven by sunkbd(4) work under X these now have
1382 to be configured the same way as Sun USB keyboards driven
1383 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1385 Option "XkbLayout" "us"
1386 Option "XkbRules" "xorg"
1387 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1390 It has been decided that it is desirable to provide ABI
1391 backwards compatibility to the FreeBSD 4/5/6 versions of the
1392 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1393 broken with the introduction of PCI domain support (see the
1394 20070930 entry). Unfortunately, this required the ABI of
1395 PCIOCGETCONF to be broken again in order to be able to
1396 provide backwards compatibility to the old version of that
1397 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1398 again. As for prominent ports this affects neither pciutils
1399 nor xorg-server this time, the hal port needs to be rebuilt
1403 The misnamed kthread_create() and friends have been renamed
1404 to kproc_create() etc. Many of the callers already
1405 used kproc_start()..
1406 I will return kthread_create() and friends in a while
1407 with implementations that actually create threads, not procs.
1408 Renaming corresponds with version 800002.
1417 Avoid using make -j when upgrading. While generally safe, there are
1418 sometimes problems using -j to upgrade. If your upgrade fails with
1419 -j, please try again without -j. From time to time in the past there
1420 have been problems using -j with buildworld and/or installworld. This
1421 is especially true when upgrading between "distant" versions (eg one
1422 that cross a major release boundary or several minor releases, or when
1423 several months have passed on the -current branch).
1425 Sometimes, obscure build problems are the result of environment
1426 poisoning. This can happen because the make utility reads its
1427 environment when searching for values for global variables. To run
1428 your build attempts in an "environmental clean room", prefix all make
1429 commands with 'env -i '. See the env(1) manual page for more details.
1431 When upgrading from one major version to another it is generally best
1432 to upgrade to the latest code in the currently installed branch first,
1433 then do an upgrade to the new branch. This is the best-tested upgrade
1434 path, and has the highest probability of being successful. Please try
1435 this approach before reporting problems with a major version upgrade.
1439 When upgrading the boot ZFS pool to a new version, always follow
1442 1.) recompile and reinstall the ZFS boot loader and boot block
1443 (this is part of "make buildworld" and "make installworld")
1445 2.) update the ZFS boot block on your boot drive
1447 The following example updates the ZFS boot block on the first
1448 partition (freebsd-boot) of a GPT partitioned drive ad0:
1449 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1451 Non-boot pools do not need these updates.
1455 If you are updating from a prior version of FreeBSD (even one just
1456 a few days old), you should follow this procedure. It is the most
1457 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1459 make kernel-toolchain
1460 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1461 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1463 To test a kernel once
1464 ---------------------
1465 If you just want to boot a kernel once (because you are not sure
1466 if it works, or if you want to boot a known bad kernel to provide
1467 debugging information) run
1468 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1469 nextboot -k testkernel
1471 To just build a kernel when you know that it won't mess you up
1472 --------------------------------------------------------------
1473 This assumes you are already running a CURRENT system. Replace
1474 ${arch} with the architecture of your machine (e.g. "i386",
1475 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1477 cd src/sys/${arch}/conf
1478 config KERNEL_NAME_HERE
1479 cd ../compile/KERNEL_NAME_HERE
1484 If this fails, go to the "To build a kernel" section.
1486 To rebuild everything and install it on the current system.
1487 -----------------------------------------------------------
1488 # Note: sometimes if you are running current you gotta do more than
1489 # is listed here if you are upgrading from a really old current.
1491 <make sure you have good level 0 dumps>
1493 make kernel KERNCONF=YOUR_KERNEL_HERE
1495 <reboot in single user> [3]
1503 To cross-install current onto a separate partition
1504 --------------------------------------------------
1505 # In this approach we use a separate partition to hold
1506 # current's root, 'usr', and 'var' directories. A partition
1507 # holding "/", "/usr" and "/var" should be about 2GB in
1510 <make sure you have good level 0 dumps>
1513 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1514 <maybe newfs current's root partition>
1515 <mount current's root partition on directory ${CURRENT_ROOT}>
1516 make installworld DESTDIR=${CURRENT_ROOT}
1517 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1518 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1519 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1520 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1521 <reboot into current>
1522 <do a "native" rebuild/install as described in the previous section>
1523 <maybe install compatibility libraries from ports/misc/compat*>
1527 To upgrade in-place from 8.x-stable to current
1528 ----------------------------------------------
1529 <make sure you have good level 0 dumps>
1531 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1533 <reboot in single user> [3]
1540 Make sure that you've read the UPDATING file to understand the
1541 tweaks to various things you need. At this point in the life
1542 cycle of current, things change often and you are on your own
1543 to cope. The defaults can also change, so please read ALL of
1544 the UPDATING entries.
1546 Also, if you are tracking -current, you must be subscribed to
1547 freebsd-current@freebsd.org. Make sure that before you update
1548 your sources that you have read and understood all the recent
1549 messages there. If in doubt, please track -stable which has
1550 much fewer pitfalls.
1552 [1] If you have third party modules, such as vmware, you
1553 should disable them at this point so they don't crash your
1556 [3] From the bootblocks, boot -s, and then do
1561 adjkerntz -i # if CMOS is wall time
1562 Also, when doing a major release upgrade, it is required that
1563 you boot into single user mode to do the installworld.
1565 [4] Note: This step is non-optional. Failure to do this step
1566 can result in a significant reduction in the functionality of the
1567 system. Attempting to do it by hand is not recommended and those
1568 that pursue this avenue should read this file carefully, as well
1569 as the archives of freebsd-current and freebsd-hackers mailing lists
1570 for potential gotchas. The -U option is also useful to consider.
1571 See mergemaster(8) for more information.
1573 [5] Usually this step is a noop. However, from time to time
1574 you may need to do this if you get unknown user in the following
1575 step. It never hurts to do it all the time. You may need to
1576 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1577 install) after the buildworld before this step if you last updated
1578 from current before 20020224 or from -stable before 20020408.
1580 [6] This only deletes old files and directories. Old libraries
1581 can be deleted by "make delete-old-libs", but you have to make
1582 sure that no program is using those libraries anymore.
1584 [8] In order to have a kernel that can run the 4.x binaries needed to
1585 do an installworld, you must include the COMPAT_FREEBSD4 option in
1586 your kernel. Failure to do so may leave you with a system that is
1587 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1588 required to run the 5.x binaries on more recent kernels. And so on
1589 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1591 Make sure that you merge any new devices from GENERIC since the
1592 last time you updated your kernel config file.
1594 [9] When checking out sources, you must include the -P flag to have
1595 cvs prune empty directories.
1597 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1598 "?=" instead of the "=" assignment operator, so that buildworld can
1599 override the CPUTYPE if it needs to.
1601 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1602 not on the command line, or in /etc/make.conf. buildworld will
1603 warn if it is improperly defined.
1606 This file contains a list, in reverse chronological order, of major
1607 breakages in tracking -current. It is not guaranteed to be a complete
1608 list of such breakages, and only contains entries since October 10, 2007.
1609 If you need to see UPDATING entries from before that date, you will need
1610 to fetch an UPDATING file from an older FreeBSD release.
1612 Copyright information:
1614 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1616 Redistribution, publication, translation and use, with or without
1617 modification, in full or in part, in any form or format of this
1618 document are permitted without further permission from the author.
1620 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1621 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1622 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1623 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1624 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1625 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1626 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1627 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1628 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1629 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1630 POSSIBILITY OF SUCH DAMAGE.
1632 Contact Warner Losh if you have any questions about your use of