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 20130218: p6 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
13 Fix Denial of Service vulnerability in named(8) with DNS64.
15 Fix Denial of Service vulnerability in libc's glob(3) functionality.
17 20121122: p5 FreeBSD-SA-12:06.bind FreeBSD-SA-12:07.hostapd
18 FreeBSD-SA-12:08.linux
19 Fix multiple Denial of Service vulnerabilities with named(8).
21 Fix insufficient message length validation for EAP-TLS messages.
23 Fix Linux compatibility layer input validation error.
25 20120806: p4 FreeBSD-SA-12:05.bind
26 Fix named(8) DNSSEC validation Denial of Service.
28 20120612: p3 FreeBSD-SA-12:03.bind
29 FreeBSD-SA-12:04.sysret
30 FreeBSD-EN-12:02.ipv6refcount
31 Fix a problem where zero-length RDATA fields can cause named to crash.
34 Correct a privilege escalation when returning from kernel if
35 running FreeBSD/amd64 on non-AMD processors. [12:04]
37 Fix reference count errors in IPv6 code. [EN-12:02]
39 20120530: p2 FreeBSD-SA-12:01.openssl (revised),
40 FreeBSD-SA-12:02.crypt
41 Update the previous openssl fix. [12:01]
43 Fix a bug in crypt(3) ignoring characters of a passphrase. [12:02]
45 20120503: p1 FreeBSD-SA-12:01.openssl
46 Fix multiple OpenSSL vulnerabilities.
52 The broken amd(4) driver has been replaced with esp(4) in the amd64,
53 i386 and pc98 GENERIC kernel configuration files.
56 This commit modifies vfs_register() so that it uses a hash
57 calculation to set vfc_typenum, which is enabled by default.
58 The first time a system is booted after this change, the
59 vfc_typenum values will change for all file systems. The
60 main effect of this is a change to the NFS server file handles
61 for file systems that use vfc_typenum in their fsid, such as ZFS.
62 It will, however, prevent vfc_typenum from changing when file
63 systems are loaded in a different order for subsequent reboots.
64 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
65 until you are ready to remount all NFS clients after a reboot.
68 Bump the shared library version numbers for libraries that
69 do not use symbol versioning, have changed the ABI compared
70 to stable/8 and which shared library version was not bumped.
71 Done as part of 9.0-RELEASE cycle.
74 During the merge of Capsicum features, the fget(9) KPI was modified.
75 This may require the rebuilding of out-of-tree device drivers --
76 issues have been reported specifically with the nVidia device driver.
77 __FreeBSD_version is bumped to 900041.
79 Also, there is a period between 20110811 and 20110814 where the
80 special devices /dev/{stdin,stdout,stderr} did not work correctly.
81 Building world from a kernel during that window may not work.
84 The packet filter (pf) code has been updated to OpenBSD 4.5.
85 You need to update userland tools to be in sync with kernel.
86 This update breaks backward compatibility with earlier pfsync(4)
87 versions. Care must be taken when updating redundant firewall setups.
90 The following sysctls and tunables are retired on x86 platforms:
92 machdep.hlt_logical_cpus
93 The following sysctl is retired:
94 machdep.hyperthreading_allowed
95 The sysctls were supposed to provide a way to dynamically offline and
96 online selected CPUs on x86 platforms, but the implementation has not
97 been reliable especially with SCHED_ULE scheduler.
98 machdep.hyperthreading_allowed tunable is still available to ignore
99 hyperthreading CPUs at OS level.
100 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
101 where X is an APIC ID of a CPU. Be advised, though, that disabling
102 CPUs in non-uniform fashion will result in non-uniform topology and
103 may lead to sub-optimal system performance with SCHED_ULE, which is
107 cpumask_t type is retired and cpuset_t is used in order to describe
111 Changes to ifconfig(8) for dynamic address family detection mandate
112 that you are running a kernel of 20110525 or later. Make sure to
113 follow the update procedure to boot a new kernel before installing
117 Support for sun4v architecture is officially dropped
120 Several KPI breaking changes have been committed to the mii(4) layer,
121 the PHY drivers and consequently some Ethernet drivers using mii(4).
122 This means that miibus.ko and the modules of the affected Ethernet
123 drivers need to be recompiled.
125 Note to kernel developers: Given that the OUI bit reversion problem
126 was fixed as part of these changes all mii(4) commits related to OUIs,
127 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
128 handling, no longer can be merged verbatim to stable/8 and previous
132 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
133 into their kernel configurations along with 'device pci'.
136 The default NFS client is now the new NFS client, so fstype "newnfs"
137 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
138 Although mounts via fstype "nfs" will usually work without userland
139 changes, it is recommended that the mount(8) and mount_nfs(8)
140 commands be rebuilt from sources and that a link to mount_nfs called
141 mount_oldnfs be created. The new client is compiled into the
142 kernel with "options NFSCL" and this is needed for diskless root
143 file systems. The GENERIC kernel configs have been changed to use
144 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
145 To use the regular/old client, you can "mount -t oldnfs ...". For
146 a diskless root file system, you must also include a line like:
148 vfs.root.mountfrom="oldnfs:"
150 in the boot/loader.conf on the root fs on the NFS server to make
151 a diskless root fs use the old client.
154 The GENERIC kernels for all architectures now default to the new
155 CAM-based ATA stack. It means that all legacy ATA drivers were
156 removed and replaced by respective CAM drivers. If you are using
157 ATA device names in /etc/fstab or other places, make sure to update
158 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
159 where 'Y's are the sequential numbers starting from zero for each type
160 in order of detection, unless configured otherwise with tunables,
161 see cam(4)). There will be symbolic links created in /dev/ to map
162 old adX devices to the respective adaY. They should provide basic
163 compatibility for file systems mounting in most cases, but they do
164 not support old user-level APIs and do not have respective providers
165 in GEOM. Consider using updated management tools with new device names.
167 It is possible to load devices ahci, ata, siis and mvs as modules,
168 but option ATA_CAM should remain in kernel configuration to make ata
169 module work as CAM driver supporting legacy ATA controllers. Device ata
170 still can be used in modular fashion (atacore + ...). Modules atadisk
171 and atapi* are not used and won't affect operation in ATA_CAM mode.
172 Note that to use CAM-based ATA kernel should include CAM devices
173 scbus, pass, da (or explicitly ada), cd and optionally others. All of
174 them are parts of the cam module.
176 ataraid(4) functionality is now supported by the RAID GEOM class.
177 To use it you can load geom_raid kernel module and use graid(8) tool
178 for management. Instead of /dev/arX device names, use /dev/raid/rX.
180 No kernel config options or code have been removed, so if a problem
181 arises, please report it and optionally revert to the old ATA stack.
182 In order to do it you can remove from the kernel config:
187 , and instead add back:
188 device atadisk # ATA disk drives
189 device ataraid # ATA RAID drives
190 device atapicd # ATAPI CDROM drives
191 device atapifd # ATAPI floppy drives
192 device atapist # ATAPI tape drives
195 The default NFS server has been changed to the new server, which
196 was referred to as the experimental server. If you need to switch
197 back to the old NFS server, you must now put the "-o" option on
198 both the mountd and nfsd commands. This can be done using the
199 mountd_flags and nfs_server_flags rc.conf variables until an
200 update to the rc scripts is committed, which is coming soon.
203 The GNU Objective-C runtime library (libobjc), and other Objective-C
204 related components have been removed from the base system. If you
205 require an Objective-C library, please use one of the available ports.
208 ath(4) has been split into bus- and device- modules. if_ath contains
209 the HAL, the TX rate control and the network device code. if_ath_pci
210 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
211 contains the AHB glue. Users need to load both if_ath_pci and if_ath
212 in order to use ath on everything else.
214 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
215 need to load if_ath and if_ath_pci for ath(4) operation.
218 As part of the replacement of sysinstall, the process of building
219 release media has changed significantly. For details, please re-read
220 release(7), which has been updated to reflect the new build process.
223 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
224 is the last available version under GPLv2. It brings a number of new
225 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
226 4.1 and SSE 4.2), better support for powerpc64, a number of new
227 directives, and lots of other small improvements. See the ChangeLog
228 file in contrib/binutils for the full details.
231 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
232 compliant, and will now use half of hash for authentication.
233 This will break interoperability with all stacks (including all
234 actual FreeBSD versions) who implement
235 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
237 The only workaround with such peers is to use another HMAC
238 algorithm for IPsec ("phase 2") authentication.
241 Remove the uio_yield prototype and symbol. This function has
242 been misnamed since it was introduced and should not be
243 globally exposed with this name. The equivalent functionality
244 is now available using kern_yield(curthread->td_user_pri).
245 The function remains undocumented.
248 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
249 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
250 for scalar sysctls is defined but disabled. Code that needs
251 UQUAD to pass the type checking that must compile on older
252 systems where the define is not present can check against
253 __FreeBSD_version >= 900030.
255 The system dialog(1) has been replaced with a new version previously
256 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
257 with the previous version, but the libdialog associated with it has
258 a largely incompatible API. As such, the original version of libdialog
259 will be kept temporarily as libodialog, until its base system consumers
260 are replaced or updated. Bump __FreeBSD_version to 900030.
263 If you are trying to run make universe on a -stable system, and you get
264 the following warning:
265 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
266 unknown. config(8) likely too old."
267 or something similar to it, then you must upgrade your -stable system
268 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
269 stable/8) or build the config from the latest stable/8 branch and
270 install it on your system.
272 Prior to this date, building a current universe on 8-stable system from
273 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
274 error in the first kernel build phase. A new config on those old
275 systems will fix that problem for older versions of -current.
278 The TCP stack has been modified to allow Khelp modules to interact with
279 it via helper hook points and store per-connection data in the TCP
280 control block. Bump __FreeBSD_version to 900029. User space tools that
281 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
285 Generic IEEE 802.3 annex 31B full duplex flow control support has been
286 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
287 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
288 to take advantage of it instead of using custom implementations. This
289 means that these drivers now no longer unconditionally advertise
290 support for flow control but only do so if flow control is a selected
291 media option. This was implemented in the generic support that way in
292 order to allow flow control to be switched on and off via ifconfig(8)
293 with the PHY specific default to typically off in order to protect
294 from unwanted effects. Consequently, if you used flow control with
295 one of the above mentioned drivers you now need to explicitly enable
297 ifconfig bge0 media auto mediaopt flowcontrol
299 Along with the above mentioned changes generic support for setting
300 1000baseT master mode also has been added and brgphy(4), ciphy(4),
301 e1000phy(4) as well as ip1000phy(4) have been converted to take
302 advantage of it. This means that these drivers now no longer take the
303 link0 parameter for selecting master mode but the master media option
304 has to be used instead, for example like in the following:
305 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
307 Selection of master mode now is also available with all other PHY
308 drivers supporting 1000baseT.
311 The TCP stack has received a significant update to add support for
312 modularised congestion control and generally improve the clarity of
313 congestion control decisions. Bump __FreeBSD_version to 900025. User
314 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
315 sockstat) need to be recompiled.
318 The man(1) utility has been replaced by a new version that no longer
319 uses /etc/manpath.config. Please consult man.conf(5) for how to
320 migrate local entries to the new format.
323 The copyright strings printed by login(1) and sshd(8) at the time of a
324 new connection have been removed to follow other operating systems and
328 A workaround for a fixed ld bug has been removed in kernel code,
329 so make sure that your system ld is built from sources after
330 revision 210245 from 2010-07-19 (r211583 if building head kernel
331 on stable/8, r211584 for stable/7; both from 2010-08-21).
332 A symptom of incorrect ld version is different addresses for
333 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
336 The $ipv6_prefer variable in rc.conf(5) has been split into
337 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
339 The $ip6addrctl_policy is a variable to choose a pre-defined
340 address selection policy set by ip6addrctl(8). A value
341 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
344 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
345 flag (see an entry of 20090926) is set on an interface with no
346 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
347 security reason. If you want IPv6 link-local address on all
348 interfaces by default, set this to "YES".
350 The old ipv6_prefer="YES" is equivalent to
351 ipv6_activate_all_interfaces="YES" and
352 ip6addrctl_policy="ipv6_prefer".
355 DTrace has grown support for userland tracing. Due to this, DTrace is
356 now i386 and amd64 only.
357 dtruss(1) is now installed by default on those systems and a new
358 kernel module is needed for userland tracing: fasttrap.
359 No changes to your kernel config file are necessary to enable
360 userland tracing, but you might consider adding 'STRIP=' and
361 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
362 to have informative userland stack traces in DTrace (ustack).
365 The acpi_aiboost(4) driver has been removed in favor of the new
366 aibs(4) driver. You should update your kernel configuration file.
369 BSD grep has been imported to the base system and it is built by
370 default. It is completely BSD licensed, highly GNU-compatible, uses
371 less memory than its GNU counterpart and has a small codebase.
372 However, it is slower than its GNU counterpart, which is mostly
373 noticeable for larger searches, for smaller ones it is measurable
374 but not significant. The reason is complex, the most important factor
375 is that we lack a modern and efficient regex library and GNU
376 overcomes this by optimizing the searches internally. Future work
377 on improving the regex performance is planned, for the meantime,
378 users that need better performance, can build GNU grep instead by
379 setting the WITH_GNU_GREP knob.
382 Due to the import of powerpc64 support, all existing powerpc kernel
383 configuration files must be updated with a machine directive like this:
384 machine powerpc powerpc
386 In addition, an updated config(8) is required to build powerpc kernels
390 A new version of ZFS (version 15) has been merged to -HEAD.
391 This version uses a python library for the following subcommands:
392 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
393 For full functionality of these commands the following port must
394 be installed: sysutils/py-zfs
397 'vm_page's are now hashed by physical address to an array of mutexes.
398 Currently this is only used to serialize access to hold_count. Over
399 time the page queue mutex will be peeled away. This changes the size
400 of pmap on every architecture. And requires all callers of vm_page_hold
401 and vm_page_unhold to be updated.
404 WITH_CTF can now be specified in src.conf (not recommended, there
405 are some problems with static executables), make.conf (would also
406 affect ports which do not use GNU make and do not override the
407 compile targets) or in the kernel config (via "makeoptions
409 When WITH_CTF was specified there before this was silently ignored,
410 so make sure that WITH_CTF is not used in places which could lead
411 to unwanted behavior.
414 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
415 to allow 32-bit compatibility on non-x86 platforms. All kernel
416 configurations on amd64 and ia64 platforms using these options must
417 be modified accordingly.
420 The utmp user accounting database has been replaced with utmpx,
421 the user accounting interface standardized by POSIX.
422 Unfortunately the semantics of utmp and utmpx don't match,
423 making it practically impossible to support both interfaces.
424 The user accounting database is used by tools like finger(1),
425 last(1), talk(1), w(1) and ac(8).
427 All applications in the base system use utmpx. This means only
428 local binaries (e.g. from the ports tree) may still use these
429 utmp database files. These applications must be rebuilt to make
432 After the system has been upgraded, it is safe to remove the old
433 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
434 assuming their contents is of no importance anymore. Old wtmp
435 databases can only be used by last(1) and ac(8) after they have
436 been converted to the new format using wtmpcvt(1).
439 Introduce the kernel thread "deadlock resolver" (which can be enabled
440 via the DEADLKRES option, see NOTES for more details) and the
441 sleepq_type() function for sleepqueues.
444 The rc.firewall and rc.firewall6 were unified, and
445 rc.firewall6 and rc.d/ip6fw were removed.
446 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
447 variables are obsoleted. Instead, the following new rc
448 variables are added to rc.d/ipfw:
450 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
451 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
452 firewall_simple_onet_ipv6, firewall_trusted_ipv6
454 The meanings correspond to the relevant IPv4 variables.
460 The default terminal emulation for syscons(4) has been changed
461 from cons25 to xterm on all platforms except pc98. This means
462 that the /etc/ttys file needs to be updated to ensure correct
463 operation of applications on the console.
465 The terminal emulation style can be toggled per window by using
466 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
467 options can be used to change the compile-time default back to
470 To prevent graphical artifacts, make sure the TERM environment
471 variable is set to match the terminal emulation that is being
472 performed by syscons(4).
475 The layout of the structure ieee80211req_scan_result has changed.
476 Applications that require wireless scan results (e.g. ifconfig(8))
477 from net80211 need to be recompiled.
479 Applications such as wpa_supplicant(8) may require a full world
480 build without using NO_CLEAN in order to get synchronized with the
484 The iwn(4) driver has been updated to support the 5000 and 5150 series.
485 There's one kernel module for each firmware. Adding "device iwnfw"
486 to the kernel configuration file means including all three firmware
487 images inside the kernel. If you want to include just the one for
488 your wireless card, use the the devices iwn4965fw, iwn5000fw or
492 The rc.d/network_ipv6, IPv6 configuration script has been integrated
493 into rc.d/netif. The changes are the following:
495 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
496 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
497 Note that both variables need the "inet6" keyword at the head.
499 Do not set $ipv6_network_interfaces manually if you do not
500 understand what you are doing. It is not needed in most cases.
502 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
505 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
506 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
508 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
509 all configured interfaces have "inet6 accept_rtadv" in the
510 $ifconfig_IF_ipv6. These are for backward compatibility.
512 3. A new variable $ipv6_prefer has been added. If NO, IPv6
513 functionality of interfaces with no corresponding
514 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
515 and the default address selection policy of ip6addrctl(8)
516 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
517 Note that if you want to configure IPv6 functionality on the
518 disabled interfaces after boot, first you need to clear the flag by
519 using ifconfig(8) like:
521 ifconfig em0 inet6 -ifdisabled
523 If YES, the default address selection policy is set as
526 The default value of $ipv6_prefer is NO.
528 4. If your system need to receive Router Advertisement messages,
529 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
530 scripts automatically invoke rtsol(8) when the interface becomes
531 UP. The Router Advertisement messages are used for SLAAC
532 (State-Less Address AutoConfiguration).
535 802.11s D3.03 support was committed. This is incompatible with the
536 previous code, which was based on D3.0.
539 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
540 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
541 control whether accepting Router Advertisement messages or not.
542 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
543 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
544 The ifconfig(8) utility now supports these flags.
547 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
548 mount(8) and -a option for df(1) to see them.
551 The old tunable hw.bus.devctl_disable has been superseded by
552 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
553 replaced by hw.bus.devctl_queue=0. The default for this new tunable
557 Remove the option STOP_NMI. The default action is now to use NMI only
558 for KDB via the newly introduced function stop_cpus_hard() and
559 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
562 The stable/8 branch created in subversion. This corresponds to the
563 RELENG_8 branch in CVS.
566 Bump the shared library version numbers for all libraries that do not
567 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
568 __FreeBSD_version to 800105.
571 Due to changes in the implementation of virtual network stack support,
572 all network-related kernel modules must be recompiled. As this change
573 breaks the ABI, bump __FreeBSD_version to 800104.
576 The TOE interface to the TCP syncache has been modified to remove
577 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
578 The cxgb driver is the only TOE consumer affected by this change, and
579 needs to be recompiled along with the kernel. As this change breaks
580 the ABI, bump __FreeBSD_version to 800103.
583 Padding has been added to struct tcpcb, sackhint and tcpstat in
584 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
585 maintaining the ABI. However, this change breaks the ABI, so bump
586 __FreeBSD_version to 800102. User space tools that rely on the size of
587 any of these structs (e.g. sockstat) need to be recompiled.
590 The NFS_LEGACYRPC option has been removed along with the old kernel
591 RPC implementation that this option selected. Kernel configurations
592 may need to be adjusted.
595 The network interface device nodes at /dev/net/<interface> have been
596 removed. All ioctl operations can be performed the normal way using
597 routing sockets. The kqueue functionality can generally be replaced
598 with routing sockets.
601 The documentation from the FreeBSD Documentation Project (Handbook,
602 FAQ, etc.) is now installed via packages by sysinstall(8) and under
603 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
606 The ABI of various structures related to the SYSV IPC API have been
607 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
608 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
612 Layout of struct vnet has changed as routing related variables were
613 moved to their own Vimage module. Modules need to be recompiled. Bump
614 __FreeBSD_version to 800099.
617 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
618 respectively. As long as no more than 16 groups per process are used,
619 no changes should be visible. When more than 16 groups are used, old
620 binaries may fail if they call getgroups() or getgrouplist() with
621 statically sized storage. Recompiling will work around this, but
622 applications should be modified to use dynamically allocated storage
623 for group arrays as POSIX.1-2008 does not cap an implementation's
624 number of supported groups at NGROUPS_MAX+1 as previous versions did.
626 NFS and portalfs mounts may also be affected as the list of groups is
627 truncated to 16. Users of NFS who use more than 16 groups, should
628 take care that negative group permissions are not used on the exported
629 file systems as they will not be reliable unless a GSSAPI based
630 authentication method is used.
633 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
634 option compiles in the support for adaptive spinning for lockmgrs
635 which want to enable it. The lockinit() function now accepts the flag
636 LK_ADAPTIVE in order to make the lock object subject to adaptive
637 spinning when both held in write and read mode.
640 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
641 changed. User applications that use this ioctl need to be rebuilt.
644 The layout of struct thread has changed. Kernel and modules need to
648 The layout of structs ifnet, domain, protosw and vnet_net has changed.
649 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
652 window(1) has been removed from the base system. It can now be
653 installed from ports. The port is called misc/window.
656 The way we are storing and accessing `routing table' entries has
657 changed. Programs reading the FIB, like netstat, need to be
661 A new netisr implementation has been added for FreeBSD 8. Network
662 file system modules, such as igmp, ipdivert, and others, should be
664 Bump __FreeBSD_version to 800096.
667 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
671 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
672 Bump __FreeBSD_version to 800094.
675 Add mnt_xflag field to 'struct mount'. File system modules need to be
677 Bump __FreeBSD_version to 800093.
680 The compiling option ADAPTIVE_SX has been retired while it has been
681 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
682 The KPI for sx_init_flags() changes as accepting flags:
683 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
684 been introduced in order to handle the reversed logic.
685 Bump __FreeBSD_version to 800092.
688 Add support for hierarchical jails. Remove global securelevel.
689 Bump __FreeBSD_version to 800091.
692 The layout of struct vnet_net has changed, therefore modules
694 Bump __FreeBSD_version to 800090.
697 The newly imported zic(8) produces a new format in the output. Please
698 run tzsetup(8) to install the newly created data to /etc/localtime.
701 The sysctl tree for the usb stack has renamed from hw.usb2.* to
702 hw.usb.* and is now consistent again with previous releases.
705 802.11 monitor mode support was revised and driver api's were changed.
706 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
707 of DLT_IEEE802_11. No user-visible data structures were changed but
708 applications that use DLT_IEEE802_11 may require changes.
709 Bump __FreeBSD_version to 800088.
712 The layout of the following structs has changed: sysctl_oid,
713 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
714 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
715 panics may be experienced. World rebuild is required for
716 correctly checking networking state from userland.
717 Bump __FreeBSD_version to 800085.
720 MLDv2 and Source-Specific Multicast (SSM) have been merged
721 to the IPv6 stack. VIMAGE hooks are in but not yet used.
722 The implementation of SSM within FreeBSD's IPv6 stack closely
723 follows the IPv4 implementation.
725 For kernel developers:
727 * The most important changes are that the ip6_output() and
728 ip6_input() paths no longer take the IN6_MULTI_LOCK,
729 and this lock has been downgraded to a non-recursive mutex.
731 * As with the changes to the IPv4 stack to support SSM, filtering
732 of inbound multicast traffic must now be performed by transport
733 protocols within the IPv6 stack. This does not apply to TCP and
734 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
736 * The KPIs used by IPv6 multicast are similar to those used by
737 the IPv4 stack, with the following differences:
738 * im6o_mc_filter() is analogous to imo_multicast_filter().
739 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
740 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
741 * IN6_LOOKUP_MULTI() has been deprecated and removed.
742 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
743 for MLDv1 have an additional 'timer' argument which is used to
744 jitter the initial membership report for the solicited-node
745 multicast membership on-link.
746 * This is not strictly needed for MLDv2, which already jitters
747 its report transmissions. However, the 'timer' argument is
748 preserved in case MLDv1 is active on the interface.
750 * The KAME linked-list based IPv6 membership implementation has
751 been refactored to use a vector similar to that used by the IPv4
753 Code which maintains a list of its own multicast memberships
754 internally, e.g. carp, has been updated to reflect the new
757 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
758 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
759 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
760 implementation constraint which needs to be addressed in HEAD.
762 For application developers:
764 * The changes are broadly similar to those made for the IPv4
767 * The use of IPv4 and IPv6 multicast socket options on the same
768 socket, using mapped addresses, HAS NOT been tested or supported.
770 * There are a number of issues with the implementation of various
771 IPv6 multicast APIs which need to be resolved in the API surface
772 before the implementation is fully compatible with KAME userland
773 use, and these are mostly to do with interface index treatment.
775 * The literature available discusses the use of either the delta / ASM
776 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
777 using setsourcefilter(3)/getsourcefilter(3). For more information
778 please refer to RFC 3768, 'Socket Interface Extensions for
779 Multicast Source Filters'.
781 * Applications which use the published RFC 3678 APIs should be fine.
783 For systems administrators:
785 * The mtest(8) utility has been refactored to support IPv6, in
786 addition to IPv4. Interface addresses are no longer accepted
787 as arguments, their names must be used instead. The utility
788 will map the interface name to its first IPv4 address as
789 returned by getifaddrs(3).
791 * The ifmcstat(8) utility has also been updated to print the MLDv2
792 endpoint state and source filter lists via sysctl(3).
794 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
795 loopback of IPv6 multicast datagrams by default; it defaults to 1
796 to preserve the existing behaviour. Disabling multicast loopback is
797 recommended for optimal system performance.
799 * The IPv6 MROUTING code has been changed to examine this sysctl
800 instead of attempting to perform a group lookup before looping
801 back forwarded datagrams.
803 Bump __FreeBSD_version to 800084.
806 Implement low-level Bluetooth HCI API.
807 Bump __FreeBSD_version to 800083.
810 The layout of struct malloc_type, used by modules to register new
811 memory allocation types, has changed. Most modules will need to
812 be rebuilt or panics may be experienced.
813 Bump __FreeBSD_version to 800081.
816 Anticipate overflowing inp_flags - add inp_flags2.
817 This changes most offsets in inpcb, so checking v4 connection
818 state will require a world rebuild.
819 Bump __FreeBSD_version to 800080.
822 Add an llentry to struct route and struct route_in6. Modules
823 embedding a struct route will need to be recompiled.
824 Bump __FreeBSD_version to 800079.
827 The size of rt_metrics_lite and by extension rtentry has changed.
828 Networking administration apps will need to be recompiled.
829 The route command now supports show as an alias for get, weighting
830 of routes, sticky and nostick flags to alter the behavior of stateful
832 Bump __FreeBSD_version to 800078.
835 Do not use Giant for kbdmux(4) locking. This is wrong and
836 apparently causing more problems than it solves. This will
837 re-open the issue where interrupt handlers may race with
838 kbdmux(4) in polling mode. Typical symptoms include (but
839 not limited to) duplicated and/or missing characters when
840 low level console functions (such as gets) are used while
841 interrupts are enabled (for example geli password prompt,
842 mountroot prompt etc.). Disabling kbdmux(4) may help.
845 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
846 kernel modules referencing any of the above need to be recompiled.
847 Bump __FreeBSD_version to 800075.
850 GEOM_PART has become the default partition slicer for storage devices,
851 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
852 introduces some changes:
854 MSDOS/EBR: the devices created from MSDOS extended partition entries
855 (EBR) can be named differently than with GEOM_MBR and are now symlinks
856 to devices with offset-based names. fstabs may need to be modified.
858 BSD: the "geometry does not match label" warning is harmless in most
859 cases but it points to problems in file system misalignment with
860 disk geometry. The "c" partition is now implicit, covers the whole
861 top-level drive and cannot be (mis)used by users.
863 General: Kernel dumps are now not allowed to be written to devices
864 whose partition types indicate they are meant to be used for file
865 systems (or, in case of MSDOS partitions, as something else than
868 Most of these changes date approximately from 200812.
871 The uscanner(4) driver has been removed from the kernel. This follows
872 Linux removing theirs in 2.6 and making libusb the default interface
876 The multicast forwarding code has been cleaned up. netstat(1)
877 only relies on KVM now for printing bandwidth upcall meters.
878 The IPv4 and IPv6 modules are split into ip_mroute_mod and
879 ip6_mroute_mod respectively. The config(5) options for statically
880 compiling this code remain the same, i.e. 'options MROUTING'.
883 Support for the IFF_NEEDSGIANT network interface flag has been
884 removed, which means that non-MPSAFE network device drivers are no
885 longer supported. In particular, if_ar, if_sr, and network device
886 drivers from the old (legacy) USB stack can no longer be built or
890 POSIX.1 Native Language Support (NLS) has been enabled in libc and
891 a bunch of new language catalog files have also been added.
892 This means that some common libc messages are now localized and
893 they depend on the LC_MESSAGES environmental variable.
896 The k8temp(4) driver has been renamed to amdtemp(4) since
897 support for Family 10 and Family 11 CPU families was added.
900 IGMPv3 and Source-Specific Multicast (SSM) have been merged
901 to the IPv4 stack. VIMAGE hooks are in but not yet used.
903 For kernel developers, the most important changes are that the
904 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
905 and this lock has been downgraded to a non-recursive mutex.
907 Transport protocols (UDP, Raw IP) are now responsible for filtering
908 inbound multicast traffic according to group membership and source
909 filters. The imo_multicast_filter() KPI exists for this purpose.
910 Transports which do not use multicast (SCTP, TCP) already reject
911 multicast by default. Forwarding and receive performance may improve
912 as a mutex acquisition is no longer needed in the ip_input()
913 low-level input path. in_addmulti() and in_delmulti() are shimmed
914 to new KPIs which exist to support SSM in-kernel.
916 For application developers, it is recommended that loopback of
917 multicast datagrams be disabled for best performance, as this
918 will still cause the lock to be taken for each looped-back
919 datagram transmission. The net.inet.ip.mcast.loop sysctl may
920 be tuned to 0 to disable loopback by default; it defaults to 1
921 to preserve the existing behaviour.
923 For systems administrators, to obtain best performance with
924 multicast reception and multiple groups, it is always recommended
925 that a card with a suitably precise hash filter is used. Hash
926 collisions will still result in the lock being taken within the
927 transport protocol input path to check group membership.
929 If deploying FreeBSD in an environment with IGMP snooping switches,
930 it is recommended that the net.inet.igmp.sendlocal sysctl remain
931 enabled; this forces 224.0.0.0/24 group membership to be announced
934 The size of 'struct igmpstat' has changed; netstat needs to be
935 recompiled to reflect this.
936 Bump __FreeBSD_version to 800070.
939 libusb20.so.1 is now installed as libusb.so.1 and the ports system
940 updated to use it. This requires a buildworld/installworld in order to
941 update the library and dependencies (usbconfig, etc). Its advisable to
942 rebuild all ports which uses libusb. More specific directions are given
943 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
944 libusb are no longer required and can be removed.
947 A workaround is committed to allow the creation of System V shared
948 memory segment of size > 2 GB on the 64-bit architectures.
949 Due to a limitation of the existing ABI, the shm_segsz member
950 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
951 wrong for large segments. Note that limits must be explicitly
952 raised to allow such segments to be created.
955 The layout of struct ifnet has changed, requiring a rebuild of all
956 network device driver modules.
959 The /dev handling for the new USB stack has changed, a
960 buildworld/installworld is required for libusb20.
963 The new USB2 stack has now been permanently moved in and all kernel and
964 module names reverted to their previous values (eg, usb, ehci, ohci,
965 ums, ...). The old usb stack can be compiled in by prefixing the name
966 with the letter 'o', the old usb modules have been removed.
967 Updating entry 20090216 for xorg and 20090215 for libmap may still
971 The rc.conf(5) option if_up_delay has been renamed to
972 defaultroute_delay to better reflect its purpose. If you have
973 customized this setting in /etc/rc.conf you need to update it to
977 xorg 7.4 wants to configure its input devices via hald which does not
978 yet work with USB2. If the keyboard/mouse does not work in xorg then
980 Option "AllowEmptyInput" "off"
981 to your ServerLayout section. This will cause X to use the configured
982 kbd and mouse sections from your xorg.conf.
985 The GENERIC kernels for all architectures now default to the new USB2
986 stack. No kernel config options or code have been removed so if a
987 problem arises please report it and optionally revert to the old USB
988 stack. If you are loading USB kernel modules or have a custom kernel
989 that includes GENERIC then ensure that usb names are also changed over,
990 eg uftdi -> usb2_serial_ftdi.
992 Older programs linked against the ports libusb 0.1 need to be
993 redirected to the new stack's libusb20. /etc/libmap.conf can
995 # Map old usb library to new one for usb2 stack
996 libusb-0.1.so.8 libusb20.so.1
999 All USB ethernet devices now attach as interfaces under the name ueN
1000 (eg. ue0). This is to provide a predictable name as vendors often
1001 change usb chipsets in a product without notice.
1004 The ichsmb(4) driver has been changed to require SMBus slave
1005 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1006 All of the other SMBus controller drivers require left-justified
1007 slave addresses, so this change makes all the drivers provide the
1011 INET6 statistics (struct ip6stat) was updated.
1012 netstat(1) needs to be recompiled.
1015 NTFS has been removed from GENERIC kernel on amd64 to match
1016 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1017 will load ntfs.ko module automatically when NTFS support is
1018 actually needed, unless ntfs.ko is not installed or security
1019 level prohibits loading kernel modules. If either is the case,
1020 "options NTFS" has to be added into kernel config.
1023 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1024 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1025 800061. User space tools that rely on the size of struct tcpcb in
1026 tcp_var.h (e.g. sockstat) need to be recompiled.
1029 ng_tty(4) module updated to match the new TTY subsystem.
1030 Due to API change, user-level applications must be updated.
1031 New API support added to mpd5 CVS and expected to be present
1032 in next mpd5.3 release.
1035 With __FreeBSD_version 800060 the makefs tool is part of
1036 the base system (it was a port).
1039 The afdata and ifnet locks have been changed from mutexes to
1040 rwlocks, network modules will need to be re-compiled.
1043 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1044 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1045 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1046 architecture and 40 bytes on 64-bit architecture. The userland
1047 applications "arp" and "ndp" have been updated accordingly.
1048 The output from "netstat -r" shows only routing entries and
1049 none of the L2 information.
1052 __FreeBSD_version 800057 marks the switchover from the
1053 binary ath hal to source code. Users must add the line:
1055 options AH_SUPPORT_AR5416
1057 to their kernel config files when specifying:
1061 The ath_hal module no longer exists; the code is now compiled
1062 together with the driver in the ath module. It is now
1063 possible to tailor chip support (i.e. reduce the set of chips
1064 and thereby the code size); consult ath_hal(4) for details.
1067 __FreeBSD_version 800054 adds memory barriers to
1068 <machine/atomic.h>, new interfaces to ifnet to facilitate
1069 multiple hardware transmit queues for cards that support
1070 them, and a lock-less ring-buffer implementation to
1071 enable drivers to more efficiently manage queueing of
1075 A new version of ZFS (version 13) has been merged to -HEAD.
1076 This version has zpool attribute "listsnapshots" off by
1077 default, which means "zfs list" does not show snapshots,
1078 and is the same as Solaris behavior.
1081 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1084 The uhci, ohci, ehci and slhci USB Host controller drivers have
1085 been put into separate modules. If you load the usb module
1086 separately through loader.conf you will need to load the
1087 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1088 controller add the following to loader.conf:
1094 The ABI used by the PMC toolset has changed. Please keep
1095 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1099 atapci kernel module now includes only generic PCI ATA
1100 driver. AHCI driver moved to ataahci kernel module.
1101 All vendor-specific code moved into separate kernel modules:
1102 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1103 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1104 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1105 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1108 The TTY subsystem of the kernel has been replaced by a new
1109 implementation, which provides better scalability and an
1110 improved driver model. Most common drivers have been migrated to
1111 the new TTY subsystem, while others have not. The following
1112 drivers have not yet been ported to the new TTY layer:
1115 cy, digi, rc, rp, sio
1121 ng_h4, ng_tty, ppp, sl, snp
1123 Adding these drivers to your kernel configuration file shall
1124 cause compilation to fail.
1127 ntpd has been upgraded to 4.2.4p5.
1130 OpenSSH has been upgraded to 5.1p1.
1132 For many years, FreeBSD's version of OpenSSH preferred DSA
1133 over RSA for host and user authentication keys. With this
1134 upgrade, we've switched to the vendor's default of RSA over
1135 DSA. This may cause upgraded clients to warn about unknown
1136 host keys even for previously known hosts. Users should
1137 follow the usual procedure for verifying host keys before
1138 accepting the RSA key.
1140 This can be circumvented by setting the "HostKeyAlgorithms"
1141 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1144 Please note that the sequence of keys offered for
1145 authentication has been changed as well. You may want to
1146 specify IdentityFile in a different order to revert this
1150 The sio(4) driver has been removed from the i386 and amd64
1151 kernel configuration files. This means uart(4) is now the
1152 default serial port driver on those platforms as well.
1154 To prevent collisions with the sio(4) driver, the uart(4) driver
1155 uses different names for its device nodes. This means the
1156 onboard serial port will now most likely be called "ttyu0"
1157 instead of "ttyd0". You may need to reconfigure applications to
1158 use the new device names.
1160 When using the serial port as a boot console, be sure to update
1161 /boot/device.hints and /etc/ttys before booting the new kernel.
1162 If you forget to do so, you can still manually specify the hints
1163 at the loader prompt:
1165 set hint.uart.0.at="isa"
1166 set hint.uart.0.port="0x3F8"
1167 set hint.uart.0.flags="0x10"
1168 set hint.uart.0.irq="4"
1172 The gpt(8) utility has been removed. Use gpart(8) to partition
1176 The version that Linuxulator emulates was changed from 2.4.2
1177 to 2.6.16. If you experience any problems with Linux binaries
1178 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1179 if it fixes the problem contact emulation mailing list.
1182 ISDN4BSD (I4B) was removed from the src tree. You may need to
1183 update a your kernel configuration and remove relevant entries.
1186 I have checked in code to support multiple routing tables.
1187 See the man pages setfib(1) and setfib(2).
1188 This is a hopefully backwards compatible version,
1189 but to make use of it you need to compile your kernel
1190 with options ROUTETABLES=2 (or more up to 16).
1193 The 802.11 wireless support was redone to enable multi-bss
1194 operation on devices that are capable. The underlying device
1195 is no longer used directly but instead wlanX devices are
1196 cloned with ifconfig. This requires changes to rc.conf files.
1197 For example, change:
1198 ifconfig_ath0="WPA DHCP"
1201 ifconfig_wlan0="WPA DHCP"
1202 see rc.conf(5) for more details. In addition, mergemaster of
1203 /etc/rc.d is highly recommended. Simultaneous update of userland
1204 and kernel wouldn't hurt either.
1206 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1207 modules were merged into the base wlan module. All references
1208 to these modules (e.g. in kernel config files) must be removed.
1211 psm(4) has gained write(2) support in native operation level.
1212 Arbitrary commands can be written to /dev/psm%d and status can
1213 be read back from it. Therefore, an application is responsible
1214 for status validation and error recovery. It is a no-op in
1215 other operation levels.
1218 Support for KSE threading has been removed from the kernel. To
1219 run legacy applications linked against KSE libmap.conf may
1220 be used. The following libmap.conf may be used to ensure
1221 compatibility with any prior release:
1223 libpthread.so.1 libthr.so.1
1224 libpthread.so.2 libthr.so.2
1225 libkse.so.3 libthr.so.3
1228 The layout of struct vmspace has changed. This affects libkvm
1229 and any executables that link against libkvm and use the
1230 kvm_getprocs() function. In particular, but not exclusively,
1231 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1232 The effects are minimal, but it's advisable to upgrade world
1236 The latest em driver no longer has support in it for the
1237 82575 adapter, this is now moved to the igb driver. The
1238 split was done to make new features that are incompatible
1239 with older hardware easier to do.
1242 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1243 likewise the kernel option is now GEOM_LINUX_LVM.
1246 The default NFS mount mode has changed from UDP to TCP for
1247 increased reliability. If you rely on (insecurely) NFS
1248 mounting across a firewall you may need to update your
1252 Belatedly note the addition of m_collapse for compacting
1256 The fts(3) structures have been changed to use adequate
1257 integer types for their members and so to be able to cope
1258 with huge file trees. The old fts(3) ABI is preserved
1259 through symbol versioning in libc, so third-party binaries
1260 using fts(3) should still work, although they will not take
1261 advantage of the extended types. At the same time, some
1262 third-party software might fail to build after this change
1263 due to unportable assumptions made in its source code about
1264 fts(3) structure members. Such software should be fixed
1265 by its vendor or, in the worst case, in the ports tree.
1266 FreeBSD_version 800015 marks this change for the unlikely
1267 case that a portable fix is impossible.
1270 To upgrade to -current after this date, you must be running
1271 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1272 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1275 The ADAPTIVE_GIANT kernel option has been retired because its
1276 functionality is the default now.
1279 The AT keyboard emulation of sunkbd(4) has been turned on
1280 by default. In order to make the special symbols of the Sun
1281 keyboards driven by sunkbd(4) work under X these now have
1282 to be configured the same way as Sun USB keyboards driven
1283 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1285 Option "XkbLayout" "us"
1286 Option "XkbRules" "xorg"
1287 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1290 It has been decided that it is desirable to provide ABI
1291 backwards compatibility to the FreeBSD 4/5/6 versions of the
1292 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1293 broken with the introduction of PCI domain support (see the
1294 20070930 entry). Unfortunately, this required the ABI of
1295 PCIOCGETCONF to be broken again in order to be able to
1296 provide backwards compatibility to the old version of that
1297 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1298 again. As for prominent ports this affects neither pciutils
1299 nor xorg-server this time, the hal port needs to be rebuilt
1303 The misnamed kthread_create() and friends have been renamed
1304 to kproc_create() etc. Many of the callers already
1305 used kproc_start()..
1306 I will return kthread_create() and friends in a while
1307 with implementations that actually create threads, not procs.
1308 Renaming corresponds with version 800002.
1317 Avoid using make -j when upgrading. While generally safe, there are
1318 sometimes problems using -j to upgrade. If your upgrade fails with
1319 -j, please try again without -j. From time to time in the past there
1320 have been problems using -j with buildworld and/or installworld. This
1321 is especially true when upgrading between "distant" versions (eg one
1322 that cross a major release boundary or several minor releases, or when
1323 several months have passed on the -current branch).
1325 Sometimes, obscure build problems are the result of environment
1326 poisoning. This can happen because the make utility reads its
1327 environment when searching for values for global variables. To run
1328 your build attempts in an "environmental clean room", prefix all make
1329 commands with 'env -i '. See the env(1) manual page for more details.
1331 When upgrading from one major version to another it is generally best
1332 to upgrade to the latest code in the currently installed branch first,
1333 then do an upgrade to the new branch. This is the best-tested upgrade
1334 path, and has the highest probability of being successful. Please try
1335 this approach before reporting problems with a major version upgrade.
1339 When upgrading the boot ZFS pool to a new version, always follow
1342 1.) recompile and reinstall the ZFS boot loader and boot block
1343 (this is part of "make buildworld" and "make installworld")
1345 2.) update the ZFS boot block on your boot drive
1347 The following example updates the ZFS boot block on the first
1348 partition (freebsd-boot) of a GPT partitioned drive ad0:
1349 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1351 Non-boot pools do not need these updates.
1355 If you are updating from a prior version of FreeBSD (even one just
1356 a few days old), you should follow this procedure. It is the most
1357 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1359 make kernel-toolchain
1360 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1361 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1363 To test a kernel once
1364 ---------------------
1365 If you just want to boot a kernel once (because you are not sure
1366 if it works, or if you want to boot a known bad kernel to provide
1367 debugging information) run
1368 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1369 nextboot -k testkernel
1371 To just build a kernel when you know that it won't mess you up
1372 --------------------------------------------------------------
1373 This assumes you are already running a CURRENT system. Replace
1374 ${arch} with the architecture of your machine (e.g. "i386",
1375 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1377 cd src/sys/${arch}/conf
1378 config KERNEL_NAME_HERE
1379 cd ../compile/KERNEL_NAME_HERE
1384 If this fails, go to the "To build a kernel" section.
1386 To rebuild everything and install it on the current system.
1387 -----------------------------------------------------------
1388 # Note: sometimes if you are running current you gotta do more than
1389 # is listed here if you are upgrading from a really old current.
1391 <make sure you have good level 0 dumps>
1393 make kernel KERNCONF=YOUR_KERNEL_HERE
1395 <reboot in single user> [3]
1403 To cross-install current onto a separate partition
1404 --------------------------------------------------
1405 # In this approach we use a separate partition to hold
1406 # current's root, 'usr', and 'var' directories. A partition
1407 # holding "/", "/usr" and "/var" should be about 2GB in
1410 <make sure you have good level 0 dumps>
1413 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1414 <maybe newfs current's root partition>
1415 <mount current's root partition on directory ${CURRENT_ROOT}>
1416 make installworld DESTDIR=${CURRENT_ROOT}
1417 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1418 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1419 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1420 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1421 <reboot into current>
1422 <do a "native" rebuild/install as described in the previous section>
1423 <maybe install compatibility libraries from ports/misc/compat*>
1427 To upgrade in-place from 8.x-stable to current
1428 ----------------------------------------------
1429 <make sure you have good level 0 dumps>
1431 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1433 <reboot in single user> [3]
1440 Make sure that you've read the UPDATING file to understand the
1441 tweaks to various things you need. At this point in the life
1442 cycle of current, things change often and you are on your own
1443 to cope. The defaults can also change, so please read ALL of
1444 the UPDATING entries.
1446 Also, if you are tracking -current, you must be subscribed to
1447 freebsd-current@freebsd.org. Make sure that before you update
1448 your sources that you have read and understood all the recent
1449 messages there. If in doubt, please track -stable which has
1450 much fewer pitfalls.
1452 [1] If you have third party modules, such as vmware, you
1453 should disable them at this point so they don't crash your
1456 [3] From the bootblocks, boot -s, and then do
1461 adjkerntz -i # if CMOS is wall time
1462 Also, when doing a major release upgrade, it is required that
1463 you boot into single user mode to do the installworld.
1465 [4] Note: This step is non-optional. Failure to do this step
1466 can result in a significant reduction in the functionality of the
1467 system. Attempting to do it by hand is not recommended and those
1468 that pursue this avenue should read this file carefully, as well
1469 as the archives of freebsd-current and freebsd-hackers mailing lists
1470 for potential gotchas. The -U option is also useful to consider.
1471 See mergemaster(8) for more information.
1473 [5] Usually this step is a noop. However, from time to time
1474 you may need to do this if you get unknown user in the following
1475 step. It never hurts to do it all the time. You may need to
1476 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1477 install) after the buildworld before this step if you last updated
1478 from current before 20020224 or from -stable before 20020408.
1480 [6] This only deletes old files and directories. Old libraries
1481 can be deleted by "make delete-old-libs", but you have to make
1482 sure that no program is using those libraries anymore.
1484 [8] In order to have a kernel that can run the 4.x binaries needed to
1485 do an installworld, you must include the COMPAT_FREEBSD4 option in
1486 your kernel. Failure to do so may leave you with a system that is
1487 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1488 required to run the 5.x binaries on more recent kernels. And so on
1489 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1491 Make sure that you merge any new devices from GENERIC since the
1492 last time you updated your kernel config file.
1494 [9] When checking out sources, you must include the -P flag to have
1495 cvs prune empty directories.
1497 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1498 "?=" instead of the "=" assignment operator, so that buildworld can
1499 override the CPUTYPE if it needs to.
1501 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1502 not on the command line, or in /etc/make.conf. buildworld will
1503 warn if it is improperly defined.
1506 This file contains a list, in reverse chronological order, of major
1507 breakages in tracking -current. It is not guaranteed to be a complete
1508 list of such breakages, and only contains entries since October 10, 2007.
1509 If you need to see UPDATING entries from before that date, you will need
1510 to fetch an UPDATING file from an older FreeBSD release.
1512 Copyright information:
1514 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1516 Redistribution, publication, translation and use, with or without
1517 modification, in full or in part, in any form or format of this
1518 document are permitted without further permission from the author.
1520 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1521 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1522 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1523 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1524 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1525 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1526 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1527 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1528 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1529 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1530 POSSIBILITY OF SUCH DAMAGE.
1532 Contact Warner Losh if you have any questions about your use of