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 20140513: p13 FreeBSD-EN-14:03.pkg
13 FreeBSD-EN-14:04.kldxref
16 Add pkg bootstrapping, configuration and public keys. [EN-14:03]
18 Improve build repeatability for kldxref(8). [EN-14:04]
20 Fix data corruption with ciss(4). [EN-14:05]
22 20140430: p12 FreeBSD-SA-14:08.tcp
24 Fix TCP reassembly vulnerability. [SA-14:08]
26 20140408: p11 FreeBSD-SA-14:05.nfsserver
27 FreeBSD-SA-14:06.openssl
28 Fix deadlock in the NFS server. [SA-14:05]
30 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
32 20140114: p10 FreeBSD-SA-14:01.bsnmpd
35 FreeBSD-EN-14:01.random
37 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
39 Fix ntpd distributed reflection Denial of Service
40 vulnerability. [SA-14:02]
42 Fix BIND remote denial of service vulnerability. [SA-14:04]
44 Disable hardware RNGs by default. [EN-14:01]
46 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
48 20131128: p9 FreeBSD-EN-13:05.freebsd-update
49 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
51 20131026: p8 FreeBSD-EN-13:04.freebsd-update
52 Fix multiple freebsd-update bugs that break upgrading to
55 20130910: p7 FreeBSD-SA-13:12.ifioctl
56 FreeBSD-SA-13:13.nullfs
58 In IPv6 and NetATM, stop SIOCSIFADDR, SIOCSIFBRDADDR,
59 SIOCSIFDSTADDR and SIOCSIFNETMASK at the socket layer rather
60 than pass them on to the link layer without validation or
61 credential checks. [SA-13:12]
63 Prevent cross-mount hardlinks between different nullfs mounts
64 of the same underlying filesystem. [SA-13:13]
66 20130822: p6 FreeBSD-SA-13:09.ip_multicast
69 Fix an integer overflow in computing the size of a temporary buffer
70 can result in a buffer which is too small for the requested
73 Fix a bug that could lead to kernel memory disclosure with
74 SCTP state cookie. [13:10]
76 Fix a data corruption problem with mfi(4) operating on > 2TB
77 disks in a JBOD. [EN-13:03]
79 20130726: p5 FreeBSD-SA-13:07.bind FreeBSD-SA-13:08.nfsserver
80 Fix Denial of Service vulnerability in named(8). [13:07]
82 Fix a bug that allows remote client bypass the normal
83 access checks when when -network or -host restrictions are
84 used at the same time with -mapall. [13:08]
86 20130618: p4 FreeBSD-SA-13:06.mmap
87 Fix a bug that allowed a tracing process (e.g. gdb) to write
88 to a memory-mapped file in the traced process's address space
89 even if neither the traced process nor the tracing process had
90 write access to that file.
92 20130429: p3 FreeBSD-SA-13:05.nfsserver
93 Fix a bug that allows NFS clients to issue READDIR on files.
95 20130402: p2 FreeBSD-SA-13:03.openssl FreeBSD-SA-13:04.bind
96 Fix multiple vulnerabilities in OpenSSL.
98 Fix Denial of Service vulnerability in named(8).
100 20130218: p1 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
101 Fix Denial of Service vulnerability in named(8) with DNS64.
103 Fix Denial of Service vulnerability in libc's glob(3) functionality.
109 The sparc64 ZFS loader has been changed to no longer try to auto-
110 detect ZFS providers based on diskN aliases but now requires these
111 to be explicitly listed in the OFW boot-device environment variable.
114 Now unix domain sockets behave "as expected" on nullfs(5). Previously
115 nullfs(5) did not pass through all behaviours to the underlying layer,
116 as a result if we bound to a socket on the lower layer we could connect
117 only to the lower path; if we bound to the upper layer we could connect
118 only to the upper path. The new behavior is one can connect to both the
119 lower and the upper paths regardless what layer path one binds to.
122 The acpi_wmi(4) status device /dev/wmistat has been renamed to
126 A new VOP_ADVISE() was added to support posix_fadvise(2). All
127 filesystem modules must be recompiled.
130 The interface of the VOP_VPTOCNP(9) changed, now the returned
131 vnode shall be referenced, previously it was required to be
132 only held. All in-tree filesystems are converted.
138 The broken amd(4) driver has been replaced with esp(4) in the amd64,
139 i386 and pc98 GENERIC kernel configuration files.
142 This commit modifies vfs_register() so that it uses a hash
143 calculation to set vfc_typenum, which is enabled by default.
144 The first time a system is booted after this change, the
145 vfc_typenum values will change for all file systems. The
146 main effect of this is a change to the NFS server file handles
147 for file systems that use vfc_typenum in their fsid, such as ZFS.
148 It will, however, prevent vfc_typenum from changing when file
149 systems are loaded in a different order for subsequent reboots.
150 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
151 until you are ready to remount all NFS clients after a reboot.
154 Bump the shared library version numbers for libraries that
155 do not use symbol versioning, have changed the ABI compared
156 to stable/8 and which shared library version was not bumped.
157 Done as part of 9.0-RELEASE cycle.
160 During the merge of Capsicum features, the fget(9) KPI was modified.
161 This may require the rebuilding of out-of-tree device drivers --
162 issues have been reported specifically with the nVidia device driver.
163 __FreeBSD_version is bumped to 900041.
165 Also, there is a period between 20110811 and 20110814 where the
166 special devices /dev/{stdin,stdout,stderr} did not work correctly.
167 Building world from a kernel during that window may not work.
170 The packet filter (pf) code has been updated to OpenBSD 4.5.
171 You need to update userland tools to be in sync with kernel.
172 This update breaks backward compatibility with earlier pfsync(4)
173 versions. Care must be taken when updating redundant firewall setups.
176 The following sysctls and tunables are retired on x86 platforms:
178 machdep.hlt_logical_cpus
179 The following sysctl is retired:
180 machdep.hyperthreading_allowed
181 The sysctls were supposed to provide a way to dynamically offline and
182 online selected CPUs on x86 platforms, but the implementation has not
183 been reliable especially with SCHED_ULE scheduler.
184 machdep.hyperthreading_allowed tunable is still available to ignore
185 hyperthreading CPUs at OS level.
186 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
187 where X is an APIC ID of a CPU. Be advised, though, that disabling
188 CPUs in non-uniform fashion will result in non-uniform topology and
189 may lead to sub-optimal system performance with SCHED_ULE, which is
193 cpumask_t type is retired and cpuset_t is used in order to describe
197 Changes to ifconfig(8) for dynamic address family detection mandate
198 that you are running a kernel of 20110525 or later. Make sure to
199 follow the update procedure to boot a new kernel before installing
203 Support for sun4v architecture is officially dropped
206 Several KPI breaking changes have been committed to the mii(4) layer,
207 the PHY drivers and consequently some Ethernet drivers using mii(4).
208 This means that miibus.ko and the modules of the affected Ethernet
209 drivers need to be recompiled.
211 Note to kernel developers: Given that the OUI bit reversion problem
212 was fixed as part of these changes all mii(4) commits related to OUIs,
213 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
214 handling, no longer can be merged verbatim to stable/8 and previous
218 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
219 into their kernel configurations along with 'device pci'.
222 The default NFS client is now the new NFS client, so fstype "newnfs"
223 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
224 Although mounts via fstype "nfs" will usually work without userland
225 changes, it is recommended that the mount(8) and mount_nfs(8)
226 commands be rebuilt from sources and that a link to mount_nfs called
227 mount_oldnfs be created. The new client is compiled into the
228 kernel with "options NFSCL" and this is needed for diskless root
229 file systems. The GENERIC kernel configs have been changed to use
230 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
231 To use the regular/old client, you can "mount -t oldnfs ...". For
232 a diskless root file system, you must also include a line like:
234 vfs.root.mountfrom="oldnfs:"
236 in the boot/loader.conf on the root fs on the NFS server to make
237 a diskless root fs use the old client.
240 The GENERIC kernels for all architectures now default to the new
241 CAM-based ATA stack. It means that all legacy ATA drivers were
242 removed and replaced by respective CAM drivers. If you are using
243 ATA device names in /etc/fstab or other places, make sure to update
244 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
245 where 'Y's are the sequential numbers starting from zero for each type
246 in order of detection, unless configured otherwise with tunables,
247 see cam(4)). There will be symbolic links created in /dev/ to map
248 old adX devices to the respective adaY. They should provide basic
249 compatibility for file systems mounting in most cases, but they do
250 not support old user-level APIs and do not have respective providers
251 in GEOM. Consider using updated management tools with new device names.
253 It is possible to load devices ahci, ata, siis and mvs as modules,
254 but option ATA_CAM should remain in kernel configuration to make ata
255 module work as CAM driver supporting legacy ATA controllers. Device ata
256 still can be used in modular fashion (atacore + ...). Modules atadisk
257 and atapi* are not used and won't affect operation in ATA_CAM mode.
258 Note that to use CAM-based ATA kernel should include CAM devices
259 scbus, pass, da (or explicitly ada), cd and optionally others. All of
260 them are parts of the cam module.
262 ataraid(4) functionality is now supported by the RAID GEOM class.
263 To use it you can load geom_raid kernel module and use graid(8) tool
264 for management. Instead of /dev/arX device names, use /dev/raid/rX.
266 No kernel config options or code have been removed, so if a problem
267 arises, please report it and optionally revert to the old ATA stack.
268 In order to do it you can remove from the kernel config:
273 , and instead add back:
274 device atadisk # ATA disk drives
275 device ataraid # ATA RAID drives
276 device atapicd # ATAPI CDROM drives
277 device atapifd # ATAPI floppy drives
278 device atapist # ATAPI tape drives
281 The default NFS server has been changed to the new server, which
282 was referred to as the experimental server. If you need to switch
283 back to the old NFS server, you must now put the "-o" option on
284 both the mountd and nfsd commands. This can be done using the
285 mountd_flags and nfs_server_flags rc.conf variables until an
286 update to the rc scripts is committed, which is coming soon.
289 The GNU Objective-C runtime library (libobjc), and other Objective-C
290 related components have been removed from the base system. If you
291 require an Objective-C library, please use one of the available ports.
294 ath(4) has been split into bus- and device- modules. if_ath contains
295 the HAL, the TX rate control and the network device code. if_ath_pci
296 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
297 contains the AHB glue. Users need to load both if_ath_pci and if_ath
298 in order to use ath on everything else.
300 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
301 need to load if_ath and if_ath_pci for ath(4) operation.
304 As part of the replacement of sysinstall, the process of building
305 release media has changed significantly. For details, please re-read
306 release(7), which has been updated to reflect the new build process.
309 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
310 is the last available version under GPLv2. It brings a number of new
311 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
312 4.1 and SSE 4.2), better support for powerpc64, a number of new
313 directives, and lots of other small improvements. See the ChangeLog
314 file in contrib/binutils for the full details.
317 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
318 compliant, and will now use half of hash for authentication.
319 This will break interoperability with all stacks (including all
320 actual FreeBSD versions) who implement
321 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
323 The only workaround with such peers is to use another HMAC
324 algorithm for IPsec ("phase 2") authentication.
327 Remove the uio_yield prototype and symbol. This function has
328 been misnamed since it was introduced and should not be
329 globally exposed with this name. The equivalent functionality
330 is now available using kern_yield(curthread->td_user_pri).
331 The function remains undocumented.
334 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
335 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
336 for scalar sysctls is defined but disabled. Code that needs
337 UQUAD to pass the type checking that must compile on older
338 systems where the define is not present can check against
339 __FreeBSD_version >= 900030.
341 The system dialog(1) has been replaced with a new version previously
342 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
343 with the previous version, but the libdialog associated with it has
344 a largely incompatible API. As such, the original version of libdialog
345 will be kept temporarily as libodialog, until its base system consumers
346 are replaced or updated. Bump __FreeBSD_version to 900030.
349 If you are trying to run make universe on a -stable system, and you get
350 the following warning:
351 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
352 unknown. config(8) likely too old."
353 or something similar to it, then you must upgrade your -stable system
354 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
355 stable/8) or build the config from the latest stable/8 branch and
356 install it on your system.
358 Prior to this date, building a current universe on 8-stable system from
359 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
360 error in the first kernel build phase. A new config on those old
361 systems will fix that problem for older versions of -current.
364 The TCP stack has been modified to allow Khelp modules to interact with
365 it via helper hook points and store per-connection data in the TCP
366 control block. Bump __FreeBSD_version to 900029. User space tools that
367 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
371 Generic IEEE 802.3 annex 31B full duplex flow control support has been
372 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
373 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
374 to take advantage of it instead of using custom implementations. This
375 means that these drivers now no longer unconditionally advertise
376 support for flow control but only do so if flow control is a selected
377 media option. This was implemented in the generic support that way in
378 order to allow flow control to be switched on and off via ifconfig(8)
379 with the PHY specific default to typically off in order to protect
380 from unwanted effects. Consequently, if you used flow control with
381 one of the above mentioned drivers you now need to explicitly enable
383 ifconfig bge0 media auto mediaopt flowcontrol
385 Along with the above mentioned changes generic support for setting
386 1000baseT master mode also has been added and brgphy(4), ciphy(4),
387 e1000phy(4) as well as ip1000phy(4) have been converted to take
388 advantage of it. This means that these drivers now no longer take the
389 link0 parameter for selecting master mode but the master media option
390 has to be used instead, for example like in the following:
391 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
393 Selection of master mode now is also available with all other PHY
394 drivers supporting 1000baseT.
397 The TCP stack has received a significant update to add support for
398 modularised congestion control and generally improve the clarity of
399 congestion control decisions. Bump __FreeBSD_version to 900025. User
400 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
401 sockstat) need to be recompiled.
404 The man(1) utility has been replaced by a new version that no longer
405 uses /etc/manpath.config. Please consult man.conf(5) for how to
406 migrate local entries to the new format.
409 The copyright strings printed by login(1) and sshd(8) at the time of a
410 new connection have been removed to follow other operating systems and
414 A workaround for a fixed ld bug has been removed in kernel code,
415 so make sure that your system ld is built from sources after
416 revision 210245 from 2010-07-19 (r211583 if building head kernel
417 on stable/8, r211584 for stable/7; both from 2010-08-21).
418 A symptom of incorrect ld version is different addresses for
419 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
422 The $ipv6_prefer variable in rc.conf(5) has been split into
423 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
425 The $ip6addrctl_policy is a variable to choose a pre-defined
426 address selection policy set by ip6addrctl(8). A value
427 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
430 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
431 flag (see an entry of 20090926) is set on an interface with no
432 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
433 security reason. If you want IPv6 link-local address on all
434 interfaces by default, set this to "YES".
436 The old ipv6_prefer="YES" is equivalent to
437 ipv6_activate_all_interfaces="YES" and
438 ip6addrctl_policy="ipv6_prefer".
441 DTrace has grown support for userland tracing. Due to this, DTrace is
442 now i386 and amd64 only.
443 dtruss(1) is now installed by default on those systems and a new
444 kernel module is needed for userland tracing: fasttrap.
445 No changes to your kernel config file are necessary to enable
446 userland tracing, but you might consider adding 'STRIP=' and
447 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
448 to have informative userland stack traces in DTrace (ustack).
451 The acpi_aiboost(4) driver has been removed in favor of the new
452 aibs(4) driver. You should update your kernel configuration file.
455 BSD grep has been imported to the base system and it is built by
456 default. It is completely BSD licensed, highly GNU-compatible, uses
457 less memory than its GNU counterpart and has a small codebase.
458 However, it is slower than its GNU counterpart, which is mostly
459 noticeable for larger searches, for smaller ones it is measurable
460 but not significant. The reason is complex, the most important factor
461 is that we lack a modern and efficient regex library and GNU
462 overcomes this by optimizing the searches internally. Future work
463 on improving the regex performance is planned, for the meantime,
464 users that need better performance, can build GNU grep instead by
465 setting the WITH_GNU_GREP knob.
468 Due to the import of powerpc64 support, all existing powerpc kernel
469 configuration files must be updated with a machine directive like this:
470 machine powerpc powerpc
472 In addition, an updated config(8) is required to build powerpc kernels
476 A new version of ZFS (version 15) has been merged to -HEAD.
477 This version uses a python library for the following subcommands:
478 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
479 For full functionality of these commands the following port must
480 be installed: sysutils/py-zfs
483 'vm_page's are now hashed by physical address to an array of mutexes.
484 Currently this is only used to serialize access to hold_count. Over
485 time the page queue mutex will be peeled away. This changes the size
486 of pmap on every architecture. And requires all callers of vm_page_hold
487 and vm_page_unhold to be updated.
490 WITH_CTF can now be specified in src.conf (not recommended, there
491 are some problems with static executables), make.conf (would also
492 affect ports which do not use GNU make and do not override the
493 compile targets) or in the kernel config (via "makeoptions
495 When WITH_CTF was specified there before this was silently ignored,
496 so make sure that WITH_CTF is not used in places which could lead
497 to unwanted behavior.
500 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
501 to allow 32-bit compatibility on non-x86 platforms. All kernel
502 configurations on amd64 and ia64 platforms using these options must
503 be modified accordingly.
506 The utmp user accounting database has been replaced with utmpx,
507 the user accounting interface standardized by POSIX.
508 Unfortunately the semantics of utmp and utmpx don't match,
509 making it practically impossible to support both interfaces.
510 The user accounting database is used by tools like finger(1),
511 last(1), talk(1), w(1) and ac(8).
513 All applications in the base system use utmpx. This means only
514 local binaries (e.g. from the ports tree) may still use these
515 utmp database files. These applications must be rebuilt to make
518 After the system has been upgraded, it is safe to remove the old
519 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
520 assuming their contents is of no importance anymore. Old wtmp
521 databases can only be used by last(1) and ac(8) after they have
522 been converted to the new format using wtmpcvt(1).
525 Introduce the kernel thread "deadlock resolver" (which can be enabled
526 via the DEADLKRES option, see NOTES for more details) and the
527 sleepq_type() function for sleepqueues.
530 The rc.firewall and rc.firewall6 were unified, and
531 rc.firewall6 and rc.d/ip6fw were removed.
532 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
533 variables are obsoleted. Instead, the following new rc
534 variables are added to rc.d/ipfw:
536 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
537 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
538 firewall_simple_onet_ipv6, firewall_trusted_ipv6
540 The meanings correspond to the relevant IPv4 variables.
546 The default terminal emulation for syscons(4) has been changed
547 from cons25 to xterm on all platforms except pc98. This means
548 that the /etc/ttys file needs to be updated to ensure correct
549 operation of applications on the console.
551 The terminal emulation style can be toggled per window by using
552 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
553 options can be used to change the compile-time default back to
556 To prevent graphical artifacts, make sure the TERM environment
557 variable is set to match the terminal emulation that is being
558 performed by syscons(4).
561 The layout of the structure ieee80211req_scan_result has changed.
562 Applications that require wireless scan results (e.g. ifconfig(8))
563 from net80211 need to be recompiled.
565 Applications such as wpa_supplicant(8) may require a full world
566 build without using NO_CLEAN in order to get synchronized with the
570 The iwn(4) driver has been updated to support the 5000 and 5150 series.
571 There's one kernel module for each firmware. Adding "device iwnfw"
572 to the kernel configuration file means including all three firmware
573 images inside the kernel. If you want to include just the one for
574 your wireless card, use the devices iwn4965fw, iwn5000fw or
578 The rc.d/network_ipv6, IPv6 configuration script has been integrated
579 into rc.d/netif. The changes are the following:
581 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
582 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
583 Note that both variables need the "inet6" keyword at the head.
585 Do not set $ipv6_network_interfaces manually if you do not
586 understand what you are doing. It is not needed in most cases.
588 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
591 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
592 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
594 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
595 all configured interfaces have "inet6 accept_rtadv" in the
596 $ifconfig_IF_ipv6. These are for backward compatibility.
598 3. A new variable $ipv6_prefer has been added. If NO, IPv6
599 functionality of interfaces with no corresponding
600 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
601 and the default address selection policy of ip6addrctl(8)
602 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
603 Note that if you want to configure IPv6 functionality on the
604 disabled interfaces after boot, first you need to clear the flag by
605 using ifconfig(8) like:
607 ifconfig em0 inet6 -ifdisabled
609 If YES, the default address selection policy is set as
612 The default value of $ipv6_prefer is NO.
614 4. If your system need to receive Router Advertisement messages,
615 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
616 scripts automatically invoke rtsol(8) when the interface becomes
617 UP. The Router Advertisement messages are used for SLAAC
618 (State-Less Address AutoConfiguration).
621 802.11s D3.03 support was committed. This is incompatible with the
622 previous code, which was based on D3.0.
625 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
626 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
627 control whether accepting Router Advertisement messages or not.
628 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
629 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
630 The ifconfig(8) utility now supports these flags.
633 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
634 mount(8) and -a option for df(1) to see them.
637 The old tunable hw.bus.devctl_disable has been superseded by
638 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
639 replaced by hw.bus.devctl_queue=0. The default for this new tunable
643 Remove the option STOP_NMI. The default action is now to use NMI only
644 for KDB via the newly introduced function stop_cpus_hard() and
645 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
648 The stable/8 branch created in subversion. This corresponds to the
649 RELENG_8 branch in CVS.
652 Bump the shared library version numbers for all libraries that do not
653 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
654 __FreeBSD_version to 800105.
657 Due to changes in the implementation of virtual network stack support,
658 all network-related kernel modules must be recompiled. As this change
659 breaks the ABI, bump __FreeBSD_version to 800104.
662 The TOE interface to the TCP syncache has been modified to remove
663 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
664 The cxgb driver is the only TOE consumer affected by this change, and
665 needs to be recompiled along with the kernel. As this change breaks
666 the ABI, bump __FreeBSD_version to 800103.
669 Padding has been added to struct tcpcb, sackhint and tcpstat in
670 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
671 maintaining the ABI. However, this change breaks the ABI, so bump
672 __FreeBSD_version to 800102. User space tools that rely on the size of
673 any of these structs (e.g. sockstat) need to be recompiled.
676 The NFS_LEGACYRPC option has been removed along with the old kernel
677 RPC implementation that this option selected. Kernel configurations
678 may need to be adjusted.
681 The network interface device nodes at /dev/net/<interface> have been
682 removed. All ioctl operations can be performed the normal way using
683 routing sockets. The kqueue functionality can generally be replaced
684 with routing sockets.
687 The documentation from the FreeBSD Documentation Project (Handbook,
688 FAQ, etc.) is now installed via packages by sysinstall(8) and under
689 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
692 The ABI of various structures related to the SYSV IPC API have been
693 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
694 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
698 Layout of struct vnet has changed as routing related variables were
699 moved to their own Vimage module. Modules need to be recompiled. Bump
700 __FreeBSD_version to 800099.
703 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
704 respectively. As long as no more than 16 groups per process are used,
705 no changes should be visible. When more than 16 groups are used, old
706 binaries may fail if they call getgroups() or getgrouplist() with
707 statically sized storage. Recompiling will work around this, but
708 applications should be modified to use dynamically allocated storage
709 for group arrays as POSIX.1-2008 does not cap an implementation's
710 number of supported groups at NGROUPS_MAX+1 as previous versions did.
712 NFS and portalfs mounts may also be affected as the list of groups is
713 truncated to 16. Users of NFS who use more than 16 groups, should
714 take care that negative group permissions are not used on the exported
715 file systems as they will not be reliable unless a GSSAPI based
716 authentication method is used.
719 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
720 option compiles in the support for adaptive spinning for lockmgrs
721 which want to enable it. The lockinit() function now accepts the flag
722 LK_ADAPTIVE in order to make the lock object subject to adaptive
723 spinning when both held in write and read mode.
726 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
727 changed. User applications that use this ioctl need to be rebuilt.
730 The layout of struct thread has changed. Kernel and modules need to
734 The layout of structs ifnet, domain, protosw and vnet_net has changed.
735 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
738 window(1) has been removed from the base system. It can now be
739 installed from ports. The port is called misc/window.
742 The way we are storing and accessing `routing table' entries has
743 changed. Programs reading the FIB, like netstat, need to be
747 A new netisr implementation has been added for FreeBSD 8. Network
748 file system modules, such as igmp, ipdivert, and others, should be
750 Bump __FreeBSD_version to 800096.
753 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
757 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
758 Bump __FreeBSD_version to 800094.
761 Add mnt_xflag field to 'struct mount'. File system modules need to be
763 Bump __FreeBSD_version to 800093.
766 The compiling option ADAPTIVE_SX has been retired while it has been
767 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
768 The KPI for sx_init_flags() changes as accepting flags:
769 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
770 been introduced in order to handle the reversed logic.
771 Bump __FreeBSD_version to 800092.
774 Add support for hierarchical jails. Remove global securelevel.
775 Bump __FreeBSD_version to 800091.
778 The layout of struct vnet_net has changed, therefore modules
780 Bump __FreeBSD_version to 800090.
783 The newly imported zic(8) produces a new format in the output. Please
784 run tzsetup(8) to install the newly created data to /etc/localtime.
787 The sysctl tree for the usb stack has renamed from hw.usb2.* to
788 hw.usb.* and is now consistent again with previous releases.
791 802.11 monitor mode support was revised and driver api's were changed.
792 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
793 of DLT_IEEE802_11. No user-visible data structures were changed but
794 applications that use DLT_IEEE802_11 may require changes.
795 Bump __FreeBSD_version to 800088.
798 The layout of the following structs has changed: sysctl_oid,
799 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
800 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
801 panics may be experienced. World rebuild is required for
802 correctly checking networking state from userland.
803 Bump __FreeBSD_version to 800085.
806 MLDv2 and Source-Specific Multicast (SSM) have been merged
807 to the IPv6 stack. VIMAGE hooks are in but not yet used.
808 The implementation of SSM within FreeBSD's IPv6 stack closely
809 follows the IPv4 implementation.
811 For kernel developers:
813 * The most important changes are that the ip6_output() and
814 ip6_input() paths no longer take the IN6_MULTI_LOCK,
815 and this lock has been downgraded to a non-recursive mutex.
817 * As with the changes to the IPv4 stack to support SSM, filtering
818 of inbound multicast traffic must now be performed by transport
819 protocols within the IPv6 stack. This does not apply to TCP and
820 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
822 * The KPIs used by IPv6 multicast are similar to those used by
823 the IPv4 stack, with the following differences:
824 * im6o_mc_filter() is analogous to imo_multicast_filter().
825 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
826 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
827 * IN6_LOOKUP_MULTI() has been deprecated and removed.
828 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
829 for MLDv1 have an additional 'timer' argument which is used to
830 jitter the initial membership report for the solicited-node
831 multicast membership on-link.
832 * This is not strictly needed for MLDv2, which already jitters
833 its report transmissions. However, the 'timer' argument is
834 preserved in case MLDv1 is active on the interface.
836 * The KAME linked-list based IPv6 membership implementation has
837 been refactored to use a vector similar to that used by the IPv4
839 Code which maintains a list of its own multicast memberships
840 internally, e.g. carp, has been updated to reflect the new
843 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
844 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
845 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
846 implementation constraint which needs to be addressed in HEAD.
848 For application developers:
850 * The changes are broadly similar to those made for the IPv4
853 * The use of IPv4 and IPv6 multicast socket options on the same
854 socket, using mapped addresses, HAS NOT been tested or supported.
856 * There are a number of issues with the implementation of various
857 IPv6 multicast APIs which need to be resolved in the API surface
858 before the implementation is fully compatible with KAME userland
859 use, and these are mostly to do with interface index treatment.
861 * The literature available discusses the use of either the delta / ASM
862 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
863 using setsourcefilter(3)/getsourcefilter(3). For more information
864 please refer to RFC 3768, 'Socket Interface Extensions for
865 Multicast Source Filters'.
867 * Applications which use the published RFC 3678 APIs should be fine.
869 For systems administrators:
871 * The mtest(8) utility has been refactored to support IPv6, in
872 addition to IPv4. Interface addresses are no longer accepted
873 as arguments, their names must be used instead. The utility
874 will map the interface name to its first IPv4 address as
875 returned by getifaddrs(3).
877 * The ifmcstat(8) utility has also been updated to print the MLDv2
878 endpoint state and source filter lists via sysctl(3).
880 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
881 loopback of IPv6 multicast datagrams by default; it defaults to 1
882 to preserve the existing behaviour. Disabling multicast loopback is
883 recommended for optimal system performance.
885 * The IPv6 MROUTING code has been changed to examine this sysctl
886 instead of attempting to perform a group lookup before looping
887 back forwarded datagrams.
889 Bump __FreeBSD_version to 800084.
892 Implement low-level Bluetooth HCI API.
893 Bump __FreeBSD_version to 800083.
896 The layout of struct malloc_type, used by modules to register new
897 memory allocation types, has changed. Most modules will need to
898 be rebuilt or panics may be experienced.
899 Bump __FreeBSD_version to 800081.
902 Anticipate overflowing inp_flags - add inp_flags2.
903 This changes most offsets in inpcb, so checking v4 connection
904 state will require a world rebuild.
905 Bump __FreeBSD_version to 800080.
908 Add an llentry to struct route and struct route_in6. Modules
909 embedding a struct route will need to be recompiled.
910 Bump __FreeBSD_version to 800079.
913 The size of rt_metrics_lite and by extension rtentry has changed.
914 Networking administration apps will need to be recompiled.
915 The route command now supports show as an alias for get, weighting
916 of routes, sticky and nostick flags to alter the behavior of stateful
918 Bump __FreeBSD_version to 800078.
921 Do not use Giant for kbdmux(4) locking. This is wrong and
922 apparently causing more problems than it solves. This will
923 re-open the issue where interrupt handlers may race with
924 kbdmux(4) in polling mode. Typical symptoms include (but
925 not limited to) duplicated and/or missing characters when
926 low level console functions (such as gets) are used while
927 interrupts are enabled (for example geli password prompt,
928 mountroot prompt etc.). Disabling kbdmux(4) may help.
931 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
932 kernel modules referencing any of the above need to be recompiled.
933 Bump __FreeBSD_version to 800075.
936 GEOM_PART has become the default partition slicer for storage devices,
937 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
938 introduces some changes:
940 MSDOS/EBR: the devices created from MSDOS extended partition entries
941 (EBR) can be named differently than with GEOM_MBR and are now symlinks
942 to devices with offset-based names. fstabs may need to be modified.
944 BSD: the "geometry does not match label" warning is harmless in most
945 cases but it points to problems in file system misalignment with
946 disk geometry. The "c" partition is now implicit, covers the whole
947 top-level drive and cannot be (mis)used by users.
949 General: Kernel dumps are now not allowed to be written to devices
950 whose partition types indicate they are meant to be used for file
951 systems (or, in case of MSDOS partitions, as something else than
954 Most of these changes date approximately from 200812.
957 The uscanner(4) driver has been removed from the kernel. This follows
958 Linux removing theirs in 2.6 and making libusb the default interface
962 The multicast forwarding code has been cleaned up. netstat(1)
963 only relies on KVM now for printing bandwidth upcall meters.
964 The IPv4 and IPv6 modules are split into ip_mroute_mod and
965 ip6_mroute_mod respectively. The config(5) options for statically
966 compiling this code remain the same, i.e. 'options MROUTING'.
969 Support for the IFF_NEEDSGIANT network interface flag has been
970 removed, which means that non-MPSAFE network device drivers are no
971 longer supported. In particular, if_ar, if_sr, and network device
972 drivers from the old (legacy) USB stack can no longer be built or
976 POSIX.1 Native Language Support (NLS) has been enabled in libc and
977 a bunch of new language catalog files have also been added.
978 This means that some common libc messages are now localized and
979 they depend on the LC_MESSAGES environmental variable.
982 The k8temp(4) driver has been renamed to amdtemp(4) since
983 support for Family 10 and Family 11 CPU families was added.
986 IGMPv3 and Source-Specific Multicast (SSM) have been merged
987 to the IPv4 stack. VIMAGE hooks are in but not yet used.
989 For kernel developers, the most important changes are that the
990 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
991 and this lock has been downgraded to a non-recursive mutex.
993 Transport protocols (UDP, Raw IP) are now responsible for filtering
994 inbound multicast traffic according to group membership and source
995 filters. The imo_multicast_filter() KPI exists for this purpose.
996 Transports which do not use multicast (SCTP, TCP) already reject
997 multicast by default. Forwarding and receive performance may improve
998 as a mutex acquisition is no longer needed in the ip_input()
999 low-level input path. in_addmulti() and in_delmulti() are shimmed
1000 to new KPIs which exist to support SSM in-kernel.
1002 For application developers, it is recommended that loopback of
1003 multicast datagrams be disabled for best performance, as this
1004 will still cause the lock to be taken for each looped-back
1005 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1006 be tuned to 0 to disable loopback by default; it defaults to 1
1007 to preserve the existing behaviour.
1009 For systems administrators, to obtain best performance with
1010 multicast reception and multiple groups, it is always recommended
1011 that a card with a suitably precise hash filter is used. Hash
1012 collisions will still result in the lock being taken within the
1013 transport protocol input path to check group membership.
1015 If deploying FreeBSD in an environment with IGMP snooping switches,
1016 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1017 enabled; this forces 224.0.0.0/24 group membership to be announced
1020 The size of 'struct igmpstat' has changed; netstat needs to be
1021 recompiled to reflect this.
1022 Bump __FreeBSD_version to 800070.
1025 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1026 updated to use it. This requires a buildworld/installworld in order to
1027 update the library and dependencies (usbconfig, etc). Its advisable to
1028 rebuild all ports which uses libusb. More specific directions are given
1029 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1030 libusb are no longer required and can be removed.
1033 A workaround is committed to allow the creation of System V shared
1034 memory segment of size > 2 GB on the 64-bit architectures.
1035 Due to a limitation of the existing ABI, the shm_segsz member
1036 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1037 wrong for large segments. Note that limits must be explicitly
1038 raised to allow such segments to be created.
1041 The layout of struct ifnet has changed, requiring a rebuild of all
1042 network device driver modules.
1045 The /dev handling for the new USB stack has changed, a
1046 buildworld/installworld is required for libusb20.
1049 The new USB2 stack has now been permanently moved in and all kernel and
1050 module names reverted to their previous values (eg, usb, ehci, ohci,
1051 ums, ...). The old usb stack can be compiled in by prefixing the name
1052 with the letter 'o', the old usb modules have been removed.
1053 Updating entry 20090216 for xorg and 20090215 for libmap may still
1057 The rc.conf(5) option if_up_delay has been renamed to
1058 defaultroute_delay to better reflect its purpose. If you have
1059 customized this setting in /etc/rc.conf you need to update it to
1063 xorg 7.4 wants to configure its input devices via hald which does not
1064 yet work with USB2. If the keyboard/mouse does not work in xorg then
1066 Option "AllowEmptyInput" "off"
1067 to your ServerLayout section. This will cause X to use the configured
1068 kbd and mouse sections from your xorg.conf.
1071 The GENERIC kernels for all architectures now default to the new USB2
1072 stack. No kernel config options or code have been removed so if a
1073 problem arises please report it and optionally revert to the old USB
1074 stack. If you are loading USB kernel modules or have a custom kernel
1075 that includes GENERIC then ensure that usb names are also changed over,
1076 eg uftdi -> usb2_serial_ftdi.
1078 Older programs linked against the ports libusb 0.1 need to be
1079 redirected to the new stack's libusb20. /etc/libmap.conf can
1081 # Map old usb library to new one for usb2 stack
1082 libusb-0.1.so.8 libusb20.so.1
1085 All USB ethernet devices now attach as interfaces under the name ueN
1086 (eg. ue0). This is to provide a predictable name as vendors often
1087 change usb chipsets in a product without notice.
1090 The ichsmb(4) driver has been changed to require SMBus slave
1091 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1092 All of the other SMBus controller drivers require left-justified
1093 slave addresses, so this change makes all the drivers provide the
1097 INET6 statistics (struct ip6stat) was updated.
1098 netstat(1) needs to be recompiled.
1101 NTFS has been removed from GENERIC kernel on amd64 to match
1102 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1103 will load ntfs.ko module automatically when NTFS support is
1104 actually needed, unless ntfs.ko is not installed or security
1105 level prohibits loading kernel modules. If either is the case,
1106 "options NTFS" has to be added into kernel config.
1109 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1110 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1111 800061. User space tools that rely on the size of struct tcpcb in
1112 tcp_var.h (e.g. sockstat) need to be recompiled.
1115 ng_tty(4) module updated to match the new TTY subsystem.
1116 Due to API change, user-level applications must be updated.
1117 New API support added to mpd5 CVS and expected to be present
1118 in next mpd5.3 release.
1121 With __FreeBSD_version 800060 the makefs tool is part of
1122 the base system (it was a port).
1125 The afdata and ifnet locks have been changed from mutexes to
1126 rwlocks, network modules will need to be re-compiled.
1129 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1130 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1131 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1132 architecture and 40 bytes on 64-bit architecture. The userland
1133 applications "arp" and "ndp" have been updated accordingly.
1134 The output from "netstat -r" shows only routing entries and
1135 none of the L2 information.
1138 __FreeBSD_version 800057 marks the switchover from the
1139 binary ath hal to source code. Users must add the line:
1141 options AH_SUPPORT_AR5416
1143 to their kernel config files when specifying:
1147 The ath_hal module no longer exists; the code is now compiled
1148 together with the driver in the ath module. It is now
1149 possible to tailor chip support (i.e. reduce the set of chips
1150 and thereby the code size); consult ath_hal(4) for details.
1153 __FreeBSD_version 800054 adds memory barriers to
1154 <machine/atomic.h>, new interfaces to ifnet to facilitate
1155 multiple hardware transmit queues for cards that support
1156 them, and a lock-less ring-buffer implementation to
1157 enable drivers to more efficiently manage queueing of
1161 A new version of ZFS (version 13) has been merged to -HEAD.
1162 This version has zpool attribute "listsnapshots" off by
1163 default, which means "zfs list" does not show snapshots,
1164 and is the same as Solaris behavior.
1167 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1170 The uhci, ohci, ehci and slhci USB Host controller drivers have
1171 been put into separate modules. If you load the usb module
1172 separately through loader.conf you will need to load the
1173 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1174 controller add the following to loader.conf:
1180 The ABI used by the PMC toolset has changed. Please keep
1181 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1185 atapci kernel module now includes only generic PCI ATA
1186 driver. AHCI driver moved to ataahci kernel module.
1187 All vendor-specific code moved into separate kernel modules:
1188 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1189 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1190 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1191 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1194 The TTY subsystem of the kernel has been replaced by a new
1195 implementation, which provides better scalability and an
1196 improved driver model. Most common drivers have been migrated to
1197 the new TTY subsystem, while others have not. The following
1198 drivers have not yet been ported to the new TTY layer:
1201 cy, digi, rc, rp, sio
1207 ng_h4, ng_tty, ppp, sl, snp
1209 Adding these drivers to your kernel configuration file shall
1210 cause compilation to fail.
1213 ntpd has been upgraded to 4.2.4p5.
1216 OpenSSH has been upgraded to 5.1p1.
1218 For many years, FreeBSD's version of OpenSSH preferred DSA
1219 over RSA for host and user authentication keys. With this
1220 upgrade, we've switched to the vendor's default of RSA over
1221 DSA. This may cause upgraded clients to warn about unknown
1222 host keys even for previously known hosts. Users should
1223 follow the usual procedure for verifying host keys before
1224 accepting the RSA key.
1226 This can be circumvented by setting the "HostKeyAlgorithms"
1227 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1230 Please note that the sequence of keys offered for
1231 authentication has been changed as well. You may want to
1232 specify IdentityFile in a different order to revert this
1236 The sio(4) driver has been removed from the i386 and amd64
1237 kernel configuration files. This means uart(4) is now the
1238 default serial port driver on those platforms as well.
1240 To prevent collisions with the sio(4) driver, the uart(4) driver
1241 uses different names for its device nodes. This means the
1242 onboard serial port will now most likely be called "ttyu0"
1243 instead of "ttyd0". You may need to reconfigure applications to
1244 use the new device names.
1246 When using the serial port as a boot console, be sure to update
1247 /boot/device.hints and /etc/ttys before booting the new kernel.
1248 If you forget to do so, you can still manually specify the hints
1249 at the loader prompt:
1251 set hint.uart.0.at="isa"
1252 set hint.uart.0.port="0x3F8"
1253 set hint.uart.0.flags="0x10"
1254 set hint.uart.0.irq="4"
1258 The gpt(8) utility has been removed. Use gpart(8) to partition
1262 The version that Linuxulator emulates was changed from 2.4.2
1263 to 2.6.16. If you experience any problems with Linux binaries
1264 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1265 if it fixes the problem contact emulation mailing list.
1268 ISDN4BSD (I4B) was removed from the src tree. You may need to
1269 update a your kernel configuration and remove relevant entries.
1272 I have checked in code to support multiple routing tables.
1273 See the man pages setfib(1) and setfib(2).
1274 This is a hopefully backwards compatible version,
1275 but to make use of it you need to compile your kernel
1276 with options ROUTETABLES=2 (or more up to 16).
1279 The 802.11 wireless support was redone to enable multi-bss
1280 operation on devices that are capable. The underlying device
1281 is no longer used directly but instead wlanX devices are
1282 cloned with ifconfig. This requires changes to rc.conf files.
1283 For example, change:
1284 ifconfig_ath0="WPA DHCP"
1287 ifconfig_wlan0="WPA DHCP"
1288 see rc.conf(5) for more details. In addition, mergemaster of
1289 /etc/rc.d is highly recommended. Simultaneous update of userland
1290 and kernel wouldn't hurt either.
1292 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1293 modules were merged into the base wlan module. All references
1294 to these modules (e.g. in kernel config files) must be removed.
1297 psm(4) has gained write(2) support in native operation level.
1298 Arbitrary commands can be written to /dev/psm%d and status can
1299 be read back from it. Therefore, an application is responsible
1300 for status validation and error recovery. It is a no-op in
1301 other operation levels.
1304 Support for KSE threading has been removed from the kernel. To
1305 run legacy applications linked against KSE libmap.conf may
1306 be used. The following libmap.conf may be used to ensure
1307 compatibility with any prior release:
1309 libpthread.so.1 libthr.so.1
1310 libpthread.so.2 libthr.so.2
1311 libkse.so.3 libthr.so.3
1314 The layout of struct vmspace has changed. This affects libkvm
1315 and any executables that link against libkvm and use the
1316 kvm_getprocs() function. In particular, but not exclusively,
1317 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1318 The effects are minimal, but it's advisable to upgrade world
1322 The latest em driver no longer has support in it for the
1323 82575 adapter, this is now moved to the igb driver. The
1324 split was done to make new features that are incompatible
1325 with older hardware easier to do.
1328 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1329 likewise the kernel option is now GEOM_LINUX_LVM.
1332 The default NFS mount mode has changed from UDP to TCP for
1333 increased reliability. If you rely on (insecurely) NFS
1334 mounting across a firewall you may need to update your
1338 Belatedly note the addition of m_collapse for compacting
1342 The fts(3) structures have been changed to use adequate
1343 integer types for their members and so to be able to cope
1344 with huge file trees. The old fts(3) ABI is preserved
1345 through symbol versioning in libc, so third-party binaries
1346 using fts(3) should still work, although they will not take
1347 advantage of the extended types. At the same time, some
1348 third-party software might fail to build after this change
1349 due to unportable assumptions made in its source code about
1350 fts(3) structure members. Such software should be fixed
1351 by its vendor or, in the worst case, in the ports tree.
1352 FreeBSD_version 800015 marks this change for the unlikely
1353 case that a portable fix is impossible.
1356 To upgrade to -current after this date, you must be running
1357 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1358 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1361 The ADAPTIVE_GIANT kernel option has been retired because its
1362 functionality is the default now.
1365 The AT keyboard emulation of sunkbd(4) has been turned on
1366 by default. In order to make the special symbols of the Sun
1367 keyboards driven by sunkbd(4) work under X these now have
1368 to be configured the same way as Sun USB keyboards driven
1369 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1371 Option "XkbLayout" "us"
1372 Option "XkbRules" "xorg"
1373 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1376 It has been decided that it is desirable to provide ABI
1377 backwards compatibility to the FreeBSD 4/5/6 versions of the
1378 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1379 broken with the introduction of PCI domain support (see the
1380 20070930 entry). Unfortunately, this required the ABI of
1381 PCIOCGETCONF to be broken again in order to be able to
1382 provide backwards compatibility to the old version of that
1383 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1384 again. As for prominent ports this affects neither pciutils
1385 nor xorg-server this time, the hal port needs to be rebuilt
1389 The misnamed kthread_create() and friends have been renamed
1390 to kproc_create() etc. Many of the callers already
1391 used kproc_start()..
1392 I will return kthread_create() and friends in a while
1393 with implementations that actually create threads, not procs.
1394 Renaming corresponds with version 800002.
1403 Avoid using make -j when upgrading. While generally safe, there are
1404 sometimes problems using -j to upgrade. If your upgrade fails with
1405 -j, please try again without -j. From time to time in the past there
1406 have been problems using -j with buildworld and/or installworld. This
1407 is especially true when upgrading between "distant" versions (eg one
1408 that cross a major release boundary or several minor releases, or when
1409 several months have passed on the -current branch).
1411 Sometimes, obscure build problems are the result of environment
1412 poisoning. This can happen because the make utility reads its
1413 environment when searching for values for global variables. To run
1414 your build attempts in an "environmental clean room", prefix all make
1415 commands with 'env -i '. See the env(1) manual page for more details.
1417 When upgrading from one major version to another it is generally best
1418 to upgrade to the latest code in the currently installed branch first,
1419 then do an upgrade to the new branch. This is the best-tested upgrade
1420 path, and has the highest probability of being successful. Please try
1421 this approach before reporting problems with a major version upgrade.
1425 When upgrading the boot ZFS pool to a new version, always follow
1428 1.) recompile and reinstall the ZFS boot loader and boot block
1429 (this is part of "make buildworld" and "make installworld")
1431 2.) update the ZFS boot block on your boot drive
1433 The following example updates the ZFS boot block on the first
1434 partition (freebsd-boot) of a GPT partitioned drive ad0:
1435 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1437 Non-boot pools do not need these updates.
1441 If you are updating from a prior version of FreeBSD (even one just
1442 a few days old), you should follow this procedure. It is the most
1443 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1445 make kernel-toolchain
1446 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1447 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1449 To test a kernel once
1450 ---------------------
1451 If you just want to boot a kernel once (because you are not sure
1452 if it works, or if you want to boot a known bad kernel to provide
1453 debugging information) run
1454 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1455 nextboot -k testkernel
1457 To just build a kernel when you know that it won't mess you up
1458 --------------------------------------------------------------
1459 This assumes you are already running a CURRENT system. Replace
1460 ${arch} with the architecture of your machine (e.g. "i386",
1461 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1463 cd src/sys/${arch}/conf
1464 config KERNEL_NAME_HERE
1465 cd ../compile/KERNEL_NAME_HERE
1470 If this fails, go to the "To build a kernel" section.
1472 To rebuild everything and install it on the current system.
1473 -----------------------------------------------------------
1474 # Note: sometimes if you are running current you gotta do more than
1475 # is listed here if you are upgrading from a really old current.
1477 <make sure you have good level 0 dumps>
1479 make kernel KERNCONF=YOUR_KERNEL_HERE
1481 <reboot in single user> [3]
1489 To cross-install current onto a separate partition
1490 --------------------------------------------------
1491 # In this approach we use a separate partition to hold
1492 # current's root, 'usr', and 'var' directories. A partition
1493 # holding "/", "/usr" and "/var" should be about 2GB in
1496 <make sure you have good level 0 dumps>
1499 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1500 <maybe newfs current's root partition>
1501 <mount current's root partition on directory ${CURRENT_ROOT}>
1502 make installworld DESTDIR=${CURRENT_ROOT}
1503 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1504 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1505 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1506 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1507 <reboot into current>
1508 <do a "native" rebuild/install as described in the previous section>
1509 <maybe install compatibility libraries from ports/misc/compat*>
1513 To upgrade in-place from 8.x-stable to current
1514 ----------------------------------------------
1515 <make sure you have good level 0 dumps>
1517 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1519 <reboot in single user> [3]
1526 Make sure that you've read the UPDATING file to understand the
1527 tweaks to various things you need. At this point in the life
1528 cycle of current, things change often and you are on your own
1529 to cope. The defaults can also change, so please read ALL of
1530 the UPDATING entries.
1532 Also, if you are tracking -current, you must be subscribed to
1533 freebsd-current@freebsd.org. Make sure that before you update
1534 your sources that you have read and understood all the recent
1535 messages there. If in doubt, please track -stable which has
1536 much fewer pitfalls.
1538 [1] If you have third party modules, such as vmware, you
1539 should disable them at this point so they don't crash your
1542 [3] From the bootblocks, boot -s, and then do
1547 adjkerntz -i # if CMOS is wall time
1548 Also, when doing a major release upgrade, it is required that
1549 you boot into single user mode to do the installworld.
1551 [4] Note: This step is non-optional. Failure to do this step
1552 can result in a significant reduction in the functionality of the
1553 system. Attempting to do it by hand is not recommended and those
1554 that pursue this avenue should read this file carefully, as well
1555 as the archives of freebsd-current and freebsd-hackers mailing lists
1556 for potential gotchas. The -U option is also useful to consider.
1557 See mergemaster(8) for more information.
1559 [5] Usually this step is a noop. However, from time to time
1560 you may need to do this if you get unknown user in the following
1561 step. It never hurts to do it all the time. You may need to
1562 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1563 install) after the buildworld before this step if you last updated
1564 from current before 20020224 or from -stable before 20020408.
1566 [6] This only deletes old files and directories. Old libraries
1567 can be deleted by "make delete-old-libs", but you have to make
1568 sure that no program is using those libraries anymore.
1570 [8] In order to have a kernel that can run the 4.x binaries needed to
1571 do an installworld, you must include the COMPAT_FREEBSD4 option in
1572 your kernel. Failure to do so may leave you with a system that is
1573 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1574 required to run the 5.x binaries on more recent kernels. And so on
1575 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1577 Make sure that you merge any new devices from GENERIC since the
1578 last time you updated your kernel config file.
1580 [9] When checking out sources, you must include the -P flag to have
1581 cvs prune empty directories.
1583 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1584 "?=" instead of the "=" assignment operator, so that buildworld can
1585 override the CPUTYPE if it needs to.
1587 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1588 not on the command line, or in /etc/make.conf. buildworld will
1589 warn if it is improperly defined.
1592 This file contains a list, in reverse chronological order, of major
1593 breakages in tracking -current. It is not guaranteed to be a complete
1594 list of such breakages, and only contains entries since October 10, 2007.
1595 If you need to see UPDATING entries from before that date, you will need
1596 to fetch an UPDATING file from an older FreeBSD release.
1598 Copyright information:
1600 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1602 Redistribution, publication, translation and use, with or without
1603 modification, in full or in part, in any form or format of this
1604 document are permitted without further permission from the author.
1606 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1607 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1608 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1609 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1610 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1611 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1612 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1613 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1614 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1615 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1616 POSSIBILITY OF SUCH DAMAGE.
1618 Contact Warner Losh if you have any questions about your use of