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 20130429: p3 FreeBSD-SA-13:05.nfsserver
13 Fix a bug that allows NFS clients to issue READDIR on files.
15 20130402: p2 FreeBSD-SA-13:03.openssl FreeBSD-SA-13:04.bind
16 Fix multiple vulnerabilities in OpenSSL.
18 Fix Denial of Service vulnerability in named(8).
20 20130218: p1 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
21 Fix Denial of Service vulnerability in named(8) with DNS64.
23 Fix Denial of Service vulnerability in libc's glob(3) functionality.
29 The sparc64 ZFS loader has been changed to no longer try to auto-
30 detect ZFS providers based on diskN aliases but now requires these
31 to be explicitly listed in the OFW boot-device environment variable.
34 Now unix domain sockets behave "as expected" on nullfs(5). Previously
35 nullfs(5) did not pass through all behaviours to the underlying layer,
36 as a result if we bound to a socket on the lower layer we could connect
37 only to the lower path; if we bound to the upper layer we could connect
38 only to the upper path. The new behavior is one can connect to both the
39 lower and the upper paths regardless what layer path one binds to.
42 The acpi_wmi(4) status device /dev/wmistat has been renamed to
46 A new VOP_ADVISE() was added to support posix_fadvise(2). All
47 filesystem modules must be recompiled.
50 The interface of the VOP_VPTOCNP(9) changed, now the returned
51 vnode shall be referenced, previously it was required to be
52 only held. All in-tree filesystems are converted.
58 The broken amd(4) driver has been replaced with esp(4) in the amd64,
59 i386 and pc98 GENERIC kernel configuration files.
62 This commit modifies vfs_register() so that it uses a hash
63 calculation to set vfc_typenum, which is enabled by default.
64 The first time a system is booted after this change, the
65 vfc_typenum values will change for all file systems. The
66 main effect of this is a change to the NFS server file handles
67 for file systems that use vfc_typenum in their fsid, such as ZFS.
68 It will, however, prevent vfc_typenum from changing when file
69 systems are loaded in a different order for subsequent reboots.
70 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
71 until you are ready to remount all NFS clients after a reboot.
74 Bump the shared library version numbers for libraries that
75 do not use symbol versioning, have changed the ABI compared
76 to stable/8 and which shared library version was not bumped.
77 Done as part of 9.0-RELEASE cycle.
80 During the merge of Capsicum features, the fget(9) KPI was modified.
81 This may require the rebuilding of out-of-tree device drivers --
82 issues have been reported specifically with the nVidia device driver.
83 __FreeBSD_version is bumped to 900041.
85 Also, there is a period between 20110811 and 20110814 where the
86 special devices /dev/{stdin,stdout,stderr} did not work correctly.
87 Building world from a kernel during that window may not work.
90 The packet filter (pf) code has been updated to OpenBSD 4.5.
91 You need to update userland tools to be in sync with kernel.
92 This update breaks backward compatibility with earlier pfsync(4)
93 versions. Care must be taken when updating redundant firewall setups.
96 The following sysctls and tunables are retired on x86 platforms:
98 machdep.hlt_logical_cpus
99 The following sysctl is retired:
100 machdep.hyperthreading_allowed
101 The sysctls were supposed to provide a way to dynamically offline and
102 online selected CPUs on x86 platforms, but the implementation has not
103 been reliable especially with SCHED_ULE scheduler.
104 machdep.hyperthreading_allowed tunable is still available to ignore
105 hyperthreading CPUs at OS level.
106 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
107 where X is an APIC ID of a CPU. Be advised, though, that disabling
108 CPUs in non-uniform fashion will result in non-uniform topology and
109 may lead to sub-optimal system performance with SCHED_ULE, which is
113 cpumask_t type is retired and cpuset_t is used in order to describe
117 Changes to ifconfig(8) for dynamic address family detection mandate
118 that you are running a kernel of 20110525 or later. Make sure to
119 follow the update procedure to boot a new kernel before installing
123 Support for sun4v architecture is officially dropped
126 Several KPI breaking changes have been committed to the mii(4) layer,
127 the PHY drivers and consequently some Ethernet drivers using mii(4).
128 This means that miibus.ko and the modules of the affected Ethernet
129 drivers need to be recompiled.
131 Note to kernel developers: Given that the OUI bit reversion problem
132 was fixed as part of these changes all mii(4) commits related to OUIs,
133 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
134 handling, no longer can be merged verbatim to stable/8 and previous
138 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
139 into their kernel configurations along with 'device pci'.
142 The default NFS client is now the new NFS client, so fstype "newnfs"
143 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
144 Although mounts via fstype "nfs" will usually work without userland
145 changes, it is recommended that the mount(8) and mount_nfs(8)
146 commands be rebuilt from sources and that a link to mount_nfs called
147 mount_oldnfs be created. The new client is compiled into the
148 kernel with "options NFSCL" and this is needed for diskless root
149 file systems. The GENERIC kernel configs have been changed to use
150 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
151 To use the regular/old client, you can "mount -t oldnfs ...". For
152 a diskless root file system, you must also include a line like:
154 vfs.root.mountfrom="oldnfs:"
156 in the boot/loader.conf on the root fs on the NFS server to make
157 a diskless root fs use the old client.
160 The GENERIC kernels for all architectures now default to the new
161 CAM-based ATA stack. It means that all legacy ATA drivers were
162 removed and replaced by respective CAM drivers. If you are using
163 ATA device names in /etc/fstab or other places, make sure to update
164 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
165 where 'Y's are the sequential numbers starting from zero for each type
166 in order of detection, unless configured otherwise with tunables,
167 see cam(4)). There will be symbolic links created in /dev/ to map
168 old adX devices to the respective adaY. They should provide basic
169 compatibility for file systems mounting in most cases, but they do
170 not support old user-level APIs and do not have respective providers
171 in GEOM. Consider using updated management tools with new device names.
173 It is possible to load devices ahci, ata, siis and mvs as modules,
174 but option ATA_CAM should remain in kernel configuration to make ata
175 module work as CAM driver supporting legacy ATA controllers. Device ata
176 still can be used in modular fashion (atacore + ...). Modules atadisk
177 and atapi* are not used and won't affect operation in ATA_CAM mode.
178 Note that to use CAM-based ATA kernel should include CAM devices
179 scbus, pass, da (or explicitly ada), cd and optionally others. All of
180 them are parts of the cam module.
182 ataraid(4) functionality is now supported by the RAID GEOM class.
183 To use it you can load geom_raid kernel module and use graid(8) tool
184 for management. Instead of /dev/arX device names, use /dev/raid/rX.
186 No kernel config options or code have been removed, so if a problem
187 arises, please report it and optionally revert to the old ATA stack.
188 In order to do it you can remove from the kernel config:
193 , and instead add back:
194 device atadisk # ATA disk drives
195 device ataraid # ATA RAID drives
196 device atapicd # ATAPI CDROM drives
197 device atapifd # ATAPI floppy drives
198 device atapist # ATAPI tape drives
201 The default NFS server has been changed to the new server, which
202 was referred to as the experimental server. If you need to switch
203 back to the old NFS server, you must now put the "-o" option on
204 both the mountd and nfsd commands. This can be done using the
205 mountd_flags and nfs_server_flags rc.conf variables until an
206 update to the rc scripts is committed, which is coming soon.
209 The GNU Objective-C runtime library (libobjc), and other Objective-C
210 related components have been removed from the base system. If you
211 require an Objective-C library, please use one of the available ports.
214 ath(4) has been split into bus- and device- modules. if_ath contains
215 the HAL, the TX rate control and the network device code. if_ath_pci
216 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
217 contains the AHB glue. Users need to load both if_ath_pci and if_ath
218 in order to use ath on everything else.
220 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
221 need to load if_ath and if_ath_pci for ath(4) operation.
224 As part of the replacement of sysinstall, the process of building
225 release media has changed significantly. For details, please re-read
226 release(7), which has been updated to reflect the new build process.
229 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
230 is the last available version under GPLv2. It brings a number of new
231 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
232 4.1 and SSE 4.2), better support for powerpc64, a number of new
233 directives, and lots of other small improvements. See the ChangeLog
234 file in contrib/binutils for the full details.
237 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
238 compliant, and will now use half of hash for authentication.
239 This will break interoperability with all stacks (including all
240 actual FreeBSD versions) who implement
241 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
243 The only workaround with such peers is to use another HMAC
244 algorithm for IPsec ("phase 2") authentication.
247 Remove the uio_yield prototype and symbol. This function has
248 been misnamed since it was introduced and should not be
249 globally exposed with this name. The equivalent functionality
250 is now available using kern_yield(curthread->td_user_pri).
251 The function remains undocumented.
254 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
255 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
256 for scalar sysctls is defined but disabled. Code that needs
257 UQUAD to pass the type checking that must compile on older
258 systems where the define is not present can check against
259 __FreeBSD_version >= 900030.
261 The system dialog(1) has been replaced with a new version previously
262 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
263 with the previous version, but the libdialog associated with it has
264 a largely incompatible API. As such, the original version of libdialog
265 will be kept temporarily as libodialog, until its base system consumers
266 are replaced or updated. Bump __FreeBSD_version to 900030.
269 If you are trying to run make universe on a -stable system, and you get
270 the following warning:
271 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
272 unknown. config(8) likely too old."
273 or something similar to it, then you must upgrade your -stable system
274 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
275 stable/8) or build the config from the latest stable/8 branch and
276 install it on your system.
278 Prior to this date, building a current universe on 8-stable system from
279 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
280 error in the first kernel build phase. A new config on those old
281 systems will fix that problem for older versions of -current.
284 The TCP stack has been modified to allow Khelp modules to interact with
285 it via helper hook points and store per-connection data in the TCP
286 control block. Bump __FreeBSD_version to 900029. User space tools that
287 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
291 Generic IEEE 802.3 annex 31B full duplex flow control support has been
292 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
293 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
294 to take advantage of it instead of using custom implementations. This
295 means that these drivers now no longer unconditionally advertise
296 support for flow control but only do so if flow control is a selected
297 media option. This was implemented in the generic support that way in
298 order to allow flow control to be switched on and off via ifconfig(8)
299 with the PHY specific default to typically off in order to protect
300 from unwanted effects. Consequently, if you used flow control with
301 one of the above mentioned drivers you now need to explicitly enable
303 ifconfig bge0 media auto mediaopt flowcontrol
305 Along with the above mentioned changes generic support for setting
306 1000baseT master mode also has been added and brgphy(4), ciphy(4),
307 e1000phy(4) as well as ip1000phy(4) have been converted to take
308 advantage of it. This means that these drivers now no longer take the
309 link0 parameter for selecting master mode but the master media option
310 has to be used instead, for example like in the following:
311 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
313 Selection of master mode now is also available with all other PHY
314 drivers supporting 1000baseT.
317 The TCP stack has received a significant update to add support for
318 modularised congestion control and generally improve the clarity of
319 congestion control decisions. Bump __FreeBSD_version to 900025. User
320 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
321 sockstat) need to be recompiled.
324 The man(1) utility has been replaced by a new version that no longer
325 uses /etc/manpath.config. Please consult man.conf(5) for how to
326 migrate local entries to the new format.
329 The copyright strings printed by login(1) and sshd(8) at the time of a
330 new connection have been removed to follow other operating systems and
334 A workaround for a fixed ld bug has been removed in kernel code,
335 so make sure that your system ld is built from sources after
336 revision 210245 from 2010-07-19 (r211583 if building head kernel
337 on stable/8, r211584 for stable/7; both from 2010-08-21).
338 A symptom of incorrect ld version is different addresses for
339 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
342 The $ipv6_prefer variable in rc.conf(5) has been split into
343 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
345 The $ip6addrctl_policy is a variable to choose a pre-defined
346 address selection policy set by ip6addrctl(8). A value
347 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
350 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
351 flag (see an entry of 20090926) is set on an interface with no
352 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
353 security reason. If you want IPv6 link-local address on all
354 interfaces by default, set this to "YES".
356 The old ipv6_prefer="YES" is equivalent to
357 ipv6_activate_all_interfaces="YES" and
358 ip6addrctl_policy="ipv6_prefer".
361 DTrace has grown support for userland tracing. Due to this, DTrace is
362 now i386 and amd64 only.
363 dtruss(1) is now installed by default on those systems and a new
364 kernel module is needed for userland tracing: fasttrap.
365 No changes to your kernel config file are necessary to enable
366 userland tracing, but you might consider adding 'STRIP=' and
367 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
368 to have informative userland stack traces in DTrace (ustack).
371 The acpi_aiboost(4) driver has been removed in favor of the new
372 aibs(4) driver. You should update your kernel configuration file.
375 BSD grep has been imported to the base system and it is built by
376 default. It is completely BSD licensed, highly GNU-compatible, uses
377 less memory than its GNU counterpart and has a small codebase.
378 However, it is slower than its GNU counterpart, which is mostly
379 noticeable for larger searches, for smaller ones it is measurable
380 but not significant. The reason is complex, the most important factor
381 is that we lack a modern and efficient regex library and GNU
382 overcomes this by optimizing the searches internally. Future work
383 on improving the regex performance is planned, for the meantime,
384 users that need better performance, can build GNU grep instead by
385 setting the WITH_GNU_GREP knob.
388 Due to the import of powerpc64 support, all existing powerpc kernel
389 configuration files must be updated with a machine directive like this:
390 machine powerpc powerpc
392 In addition, an updated config(8) is required to build powerpc kernels
396 A new version of ZFS (version 15) has been merged to -HEAD.
397 This version uses a python library for the following subcommands:
398 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
399 For full functionality of these commands the following port must
400 be installed: sysutils/py-zfs
403 'vm_page's are now hashed by physical address to an array of mutexes.
404 Currently this is only used to serialize access to hold_count. Over
405 time the page queue mutex will be peeled away. This changes the size
406 of pmap on every architecture. And requires all callers of vm_page_hold
407 and vm_page_unhold to be updated.
410 WITH_CTF can now be specified in src.conf (not recommended, there
411 are some problems with static executables), make.conf (would also
412 affect ports which do not use GNU make and do not override the
413 compile targets) or in the kernel config (via "makeoptions
415 When WITH_CTF was specified there before this was silently ignored,
416 so make sure that WITH_CTF is not used in places which could lead
417 to unwanted behavior.
420 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
421 to allow 32-bit compatibility on non-x86 platforms. All kernel
422 configurations on amd64 and ia64 platforms using these options must
423 be modified accordingly.
426 The utmp user accounting database has been replaced with utmpx,
427 the user accounting interface standardized by POSIX.
428 Unfortunately the semantics of utmp and utmpx don't match,
429 making it practically impossible to support both interfaces.
430 The user accounting database is used by tools like finger(1),
431 last(1), talk(1), w(1) and ac(8).
433 All applications in the base system use utmpx. This means only
434 local binaries (e.g. from the ports tree) may still use these
435 utmp database files. These applications must be rebuilt to make
438 After the system has been upgraded, it is safe to remove the old
439 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
440 assuming their contents is of no importance anymore. Old wtmp
441 databases can only be used by last(1) and ac(8) after they have
442 been converted to the new format using wtmpcvt(1).
445 Introduce the kernel thread "deadlock resolver" (which can be enabled
446 via the DEADLKRES option, see NOTES for more details) and the
447 sleepq_type() function for sleepqueues.
450 The rc.firewall and rc.firewall6 were unified, and
451 rc.firewall6 and rc.d/ip6fw were removed.
452 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
453 variables are obsoleted. Instead, the following new rc
454 variables are added to rc.d/ipfw:
456 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
457 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
458 firewall_simple_onet_ipv6, firewall_trusted_ipv6
460 The meanings correspond to the relevant IPv4 variables.
466 The default terminal emulation for syscons(4) has been changed
467 from cons25 to xterm on all platforms except pc98. This means
468 that the /etc/ttys file needs to be updated to ensure correct
469 operation of applications on the console.
471 The terminal emulation style can be toggled per window by using
472 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
473 options can be used to change the compile-time default back to
476 To prevent graphical artifacts, make sure the TERM environment
477 variable is set to match the terminal emulation that is being
478 performed by syscons(4).
481 The layout of the structure ieee80211req_scan_result has changed.
482 Applications that require wireless scan results (e.g. ifconfig(8))
483 from net80211 need to be recompiled.
485 Applications such as wpa_supplicant(8) may require a full world
486 build without using NO_CLEAN in order to get synchronized with the
490 The iwn(4) driver has been updated to support the 5000 and 5150 series.
491 There's one kernel module for each firmware. Adding "device iwnfw"
492 to the kernel configuration file means including all three firmware
493 images inside the kernel. If you want to include just the one for
494 your wireless card, use the devices iwn4965fw, iwn5000fw or
498 The rc.d/network_ipv6, IPv6 configuration script has been integrated
499 into rc.d/netif. The changes are the following:
501 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
502 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
503 Note that both variables need the "inet6" keyword at the head.
505 Do not set $ipv6_network_interfaces manually if you do not
506 understand what you are doing. It is not needed in most cases.
508 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
511 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
512 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
514 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
515 all configured interfaces have "inet6 accept_rtadv" in the
516 $ifconfig_IF_ipv6. These are for backward compatibility.
518 3. A new variable $ipv6_prefer has been added. If NO, IPv6
519 functionality of interfaces with no corresponding
520 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
521 and the default address selection policy of ip6addrctl(8)
522 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
523 Note that if you want to configure IPv6 functionality on the
524 disabled interfaces after boot, first you need to clear the flag by
525 using ifconfig(8) like:
527 ifconfig em0 inet6 -ifdisabled
529 If YES, the default address selection policy is set as
532 The default value of $ipv6_prefer is NO.
534 4. If your system need to receive Router Advertisement messages,
535 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
536 scripts automatically invoke rtsol(8) when the interface becomes
537 UP. The Router Advertisement messages are used for SLAAC
538 (State-Less Address AutoConfiguration).
541 802.11s D3.03 support was committed. This is incompatible with the
542 previous code, which was based on D3.0.
545 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
546 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
547 control whether accepting Router Advertisement messages or not.
548 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
549 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
550 The ifconfig(8) utility now supports these flags.
553 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
554 mount(8) and -a option for df(1) to see them.
557 The old tunable hw.bus.devctl_disable has been superseded by
558 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
559 replaced by hw.bus.devctl_queue=0. The default for this new tunable
563 Remove the option STOP_NMI. The default action is now to use NMI only
564 for KDB via the newly introduced function stop_cpus_hard() and
565 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
568 The stable/8 branch created in subversion. This corresponds to the
569 RELENG_8 branch in CVS.
572 Bump the shared library version numbers for all libraries that do not
573 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
574 __FreeBSD_version to 800105.
577 Due to changes in the implementation of virtual network stack support,
578 all network-related kernel modules must be recompiled. As this change
579 breaks the ABI, bump __FreeBSD_version to 800104.
582 The TOE interface to the TCP syncache has been modified to remove
583 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
584 The cxgb driver is the only TOE consumer affected by this change, and
585 needs to be recompiled along with the kernel. As this change breaks
586 the ABI, bump __FreeBSD_version to 800103.
589 Padding has been added to struct tcpcb, sackhint and tcpstat in
590 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
591 maintaining the ABI. However, this change breaks the ABI, so bump
592 __FreeBSD_version to 800102. User space tools that rely on the size of
593 any of these structs (e.g. sockstat) need to be recompiled.
596 The NFS_LEGACYRPC option has been removed along with the old kernel
597 RPC implementation that this option selected. Kernel configurations
598 may need to be adjusted.
601 The network interface device nodes at /dev/net/<interface> have been
602 removed. All ioctl operations can be performed the normal way using
603 routing sockets. The kqueue functionality can generally be replaced
604 with routing sockets.
607 The documentation from the FreeBSD Documentation Project (Handbook,
608 FAQ, etc.) is now installed via packages by sysinstall(8) and under
609 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
612 The ABI of various structures related to the SYSV IPC API have been
613 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
614 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
618 Layout of struct vnet has changed as routing related variables were
619 moved to their own Vimage module. Modules need to be recompiled. Bump
620 __FreeBSD_version to 800099.
623 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
624 respectively. As long as no more than 16 groups per process are used,
625 no changes should be visible. When more than 16 groups are used, old
626 binaries may fail if they call getgroups() or getgrouplist() with
627 statically sized storage. Recompiling will work around this, but
628 applications should be modified to use dynamically allocated storage
629 for group arrays as POSIX.1-2008 does not cap an implementation's
630 number of supported groups at NGROUPS_MAX+1 as previous versions did.
632 NFS and portalfs mounts may also be affected as the list of groups is
633 truncated to 16. Users of NFS who use more than 16 groups, should
634 take care that negative group permissions are not used on the exported
635 file systems as they will not be reliable unless a GSSAPI based
636 authentication method is used.
639 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
640 option compiles in the support for adaptive spinning for lockmgrs
641 which want to enable it. The lockinit() function now accepts the flag
642 LK_ADAPTIVE in order to make the lock object subject to adaptive
643 spinning when both held in write and read mode.
646 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
647 changed. User applications that use this ioctl need to be rebuilt.
650 The layout of struct thread has changed. Kernel and modules need to
654 The layout of structs ifnet, domain, protosw and vnet_net has changed.
655 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
658 window(1) has been removed from the base system. It can now be
659 installed from ports. The port is called misc/window.
662 The way we are storing and accessing `routing table' entries has
663 changed. Programs reading the FIB, like netstat, need to be
667 A new netisr implementation has been added for FreeBSD 8. Network
668 file system modules, such as igmp, ipdivert, and others, should be
670 Bump __FreeBSD_version to 800096.
673 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
677 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
678 Bump __FreeBSD_version to 800094.
681 Add mnt_xflag field to 'struct mount'. File system modules need to be
683 Bump __FreeBSD_version to 800093.
686 The compiling option ADAPTIVE_SX has been retired while it has been
687 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
688 The KPI for sx_init_flags() changes as accepting flags:
689 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
690 been introduced in order to handle the reversed logic.
691 Bump __FreeBSD_version to 800092.
694 Add support for hierarchical jails. Remove global securelevel.
695 Bump __FreeBSD_version to 800091.
698 The layout of struct vnet_net has changed, therefore modules
700 Bump __FreeBSD_version to 800090.
703 The newly imported zic(8) produces a new format in the output. Please
704 run tzsetup(8) to install the newly created data to /etc/localtime.
707 The sysctl tree for the usb stack has renamed from hw.usb2.* to
708 hw.usb.* and is now consistent again with previous releases.
711 802.11 monitor mode support was revised and driver api's were changed.
712 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
713 of DLT_IEEE802_11. No user-visible data structures were changed but
714 applications that use DLT_IEEE802_11 may require changes.
715 Bump __FreeBSD_version to 800088.
718 The layout of the following structs has changed: sysctl_oid,
719 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
720 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
721 panics may be experienced. World rebuild is required for
722 correctly checking networking state from userland.
723 Bump __FreeBSD_version to 800085.
726 MLDv2 and Source-Specific Multicast (SSM) have been merged
727 to the IPv6 stack. VIMAGE hooks are in but not yet used.
728 The implementation of SSM within FreeBSD's IPv6 stack closely
729 follows the IPv4 implementation.
731 For kernel developers:
733 * The most important changes are that the ip6_output() and
734 ip6_input() paths no longer take the IN6_MULTI_LOCK,
735 and this lock has been downgraded to a non-recursive mutex.
737 * As with the changes to the IPv4 stack to support SSM, filtering
738 of inbound multicast traffic must now be performed by transport
739 protocols within the IPv6 stack. This does not apply to TCP and
740 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
742 * The KPIs used by IPv6 multicast are similar to those used by
743 the IPv4 stack, with the following differences:
744 * im6o_mc_filter() is analogous to imo_multicast_filter().
745 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
746 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
747 * IN6_LOOKUP_MULTI() has been deprecated and removed.
748 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
749 for MLDv1 have an additional 'timer' argument which is used to
750 jitter the initial membership report for the solicited-node
751 multicast membership on-link.
752 * This is not strictly needed for MLDv2, which already jitters
753 its report transmissions. However, the 'timer' argument is
754 preserved in case MLDv1 is active on the interface.
756 * The KAME linked-list based IPv6 membership implementation has
757 been refactored to use a vector similar to that used by the IPv4
759 Code which maintains a list of its own multicast memberships
760 internally, e.g. carp, has been updated to reflect the new
763 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
764 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
765 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
766 implementation constraint which needs to be addressed in HEAD.
768 For application developers:
770 * The changes are broadly similar to those made for the IPv4
773 * The use of IPv4 and IPv6 multicast socket options on the same
774 socket, using mapped addresses, HAS NOT been tested or supported.
776 * There are a number of issues with the implementation of various
777 IPv6 multicast APIs which need to be resolved in the API surface
778 before the implementation is fully compatible with KAME userland
779 use, and these are mostly to do with interface index treatment.
781 * The literature available discusses the use of either the delta / ASM
782 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
783 using setsourcefilter(3)/getsourcefilter(3). For more information
784 please refer to RFC 3768, 'Socket Interface Extensions for
785 Multicast Source Filters'.
787 * Applications which use the published RFC 3678 APIs should be fine.
789 For systems administrators:
791 * The mtest(8) utility has been refactored to support IPv6, in
792 addition to IPv4. Interface addresses are no longer accepted
793 as arguments, their names must be used instead. The utility
794 will map the interface name to its first IPv4 address as
795 returned by getifaddrs(3).
797 * The ifmcstat(8) utility has also been updated to print the MLDv2
798 endpoint state and source filter lists via sysctl(3).
800 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
801 loopback of IPv6 multicast datagrams by default; it defaults to 1
802 to preserve the existing behaviour. Disabling multicast loopback is
803 recommended for optimal system performance.
805 * The IPv6 MROUTING code has been changed to examine this sysctl
806 instead of attempting to perform a group lookup before looping
807 back forwarded datagrams.
809 Bump __FreeBSD_version to 800084.
812 Implement low-level Bluetooth HCI API.
813 Bump __FreeBSD_version to 800083.
816 The layout of struct malloc_type, used by modules to register new
817 memory allocation types, has changed. Most modules will need to
818 be rebuilt or panics may be experienced.
819 Bump __FreeBSD_version to 800081.
822 Anticipate overflowing inp_flags - add inp_flags2.
823 This changes most offsets in inpcb, so checking v4 connection
824 state will require a world rebuild.
825 Bump __FreeBSD_version to 800080.
828 Add an llentry to struct route and struct route_in6. Modules
829 embedding a struct route will need to be recompiled.
830 Bump __FreeBSD_version to 800079.
833 The size of rt_metrics_lite and by extension rtentry has changed.
834 Networking administration apps will need to be recompiled.
835 The route command now supports show as an alias for get, weighting
836 of routes, sticky and nostick flags to alter the behavior of stateful
838 Bump __FreeBSD_version to 800078.
841 Do not use Giant for kbdmux(4) locking. This is wrong and
842 apparently causing more problems than it solves. This will
843 re-open the issue where interrupt handlers may race with
844 kbdmux(4) in polling mode. Typical symptoms include (but
845 not limited to) duplicated and/or missing characters when
846 low level console functions (such as gets) are used while
847 interrupts are enabled (for example geli password prompt,
848 mountroot prompt etc.). Disabling kbdmux(4) may help.
851 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
852 kernel modules referencing any of the above need to be recompiled.
853 Bump __FreeBSD_version to 800075.
856 GEOM_PART has become the default partition slicer for storage devices,
857 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
858 introduces some changes:
860 MSDOS/EBR: the devices created from MSDOS extended partition entries
861 (EBR) can be named differently than with GEOM_MBR and are now symlinks
862 to devices with offset-based names. fstabs may need to be modified.
864 BSD: the "geometry does not match label" warning is harmless in most
865 cases but it points to problems in file system misalignment with
866 disk geometry. The "c" partition is now implicit, covers the whole
867 top-level drive and cannot be (mis)used by users.
869 General: Kernel dumps are now not allowed to be written to devices
870 whose partition types indicate they are meant to be used for file
871 systems (or, in case of MSDOS partitions, as something else than
874 Most of these changes date approximately from 200812.
877 The uscanner(4) driver has been removed from the kernel. This follows
878 Linux removing theirs in 2.6 and making libusb the default interface
882 The multicast forwarding code has been cleaned up. netstat(1)
883 only relies on KVM now for printing bandwidth upcall meters.
884 The IPv4 and IPv6 modules are split into ip_mroute_mod and
885 ip6_mroute_mod respectively. The config(5) options for statically
886 compiling this code remain the same, i.e. 'options MROUTING'.
889 Support for the IFF_NEEDSGIANT network interface flag has been
890 removed, which means that non-MPSAFE network device drivers are no
891 longer supported. In particular, if_ar, if_sr, and network device
892 drivers from the old (legacy) USB stack can no longer be built or
896 POSIX.1 Native Language Support (NLS) has been enabled in libc and
897 a bunch of new language catalog files have also been added.
898 This means that some common libc messages are now localized and
899 they depend on the LC_MESSAGES environmental variable.
902 The k8temp(4) driver has been renamed to amdtemp(4) since
903 support for Family 10 and Family 11 CPU families was added.
906 IGMPv3 and Source-Specific Multicast (SSM) have been merged
907 to the IPv4 stack. VIMAGE hooks are in but not yet used.
909 For kernel developers, the most important changes are that the
910 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
911 and this lock has been downgraded to a non-recursive mutex.
913 Transport protocols (UDP, Raw IP) are now responsible for filtering
914 inbound multicast traffic according to group membership and source
915 filters. The imo_multicast_filter() KPI exists for this purpose.
916 Transports which do not use multicast (SCTP, TCP) already reject
917 multicast by default. Forwarding and receive performance may improve
918 as a mutex acquisition is no longer needed in the ip_input()
919 low-level input path. in_addmulti() and in_delmulti() are shimmed
920 to new KPIs which exist to support SSM in-kernel.
922 For application developers, it is recommended that loopback of
923 multicast datagrams be disabled for best performance, as this
924 will still cause the lock to be taken for each looped-back
925 datagram transmission. The net.inet.ip.mcast.loop sysctl may
926 be tuned to 0 to disable loopback by default; it defaults to 1
927 to preserve the existing behaviour.
929 For systems administrators, to obtain best performance with
930 multicast reception and multiple groups, it is always recommended
931 that a card with a suitably precise hash filter is used. Hash
932 collisions will still result in the lock being taken within the
933 transport protocol input path to check group membership.
935 If deploying FreeBSD in an environment with IGMP snooping switches,
936 it is recommended that the net.inet.igmp.sendlocal sysctl remain
937 enabled; this forces 224.0.0.0/24 group membership to be announced
940 The size of 'struct igmpstat' has changed; netstat needs to be
941 recompiled to reflect this.
942 Bump __FreeBSD_version to 800070.
945 libusb20.so.1 is now installed as libusb.so.1 and the ports system
946 updated to use it. This requires a buildworld/installworld in order to
947 update the library and dependencies (usbconfig, etc). Its advisable to
948 rebuild all ports which uses libusb. More specific directions are given
949 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
950 libusb are no longer required and can be removed.
953 A workaround is committed to allow the creation of System V shared
954 memory segment of size > 2 GB on the 64-bit architectures.
955 Due to a limitation of the existing ABI, the shm_segsz member
956 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
957 wrong for large segments. Note that limits must be explicitly
958 raised to allow such segments to be created.
961 The layout of struct ifnet has changed, requiring a rebuild of all
962 network device driver modules.
965 The /dev handling for the new USB stack has changed, a
966 buildworld/installworld is required for libusb20.
969 The new USB2 stack has now been permanently moved in and all kernel and
970 module names reverted to their previous values (eg, usb, ehci, ohci,
971 ums, ...). The old usb stack can be compiled in by prefixing the name
972 with the letter 'o', the old usb modules have been removed.
973 Updating entry 20090216 for xorg and 20090215 for libmap may still
977 The rc.conf(5) option if_up_delay has been renamed to
978 defaultroute_delay to better reflect its purpose. If you have
979 customized this setting in /etc/rc.conf you need to update it to
983 xorg 7.4 wants to configure its input devices via hald which does not
984 yet work with USB2. If the keyboard/mouse does not work in xorg then
986 Option "AllowEmptyInput" "off"
987 to your ServerLayout section. This will cause X to use the configured
988 kbd and mouse sections from your xorg.conf.
991 The GENERIC kernels for all architectures now default to the new USB2
992 stack. No kernel config options or code have been removed so if a
993 problem arises please report it and optionally revert to the old USB
994 stack. If you are loading USB kernel modules or have a custom kernel
995 that includes GENERIC then ensure that usb names are also changed over,
996 eg uftdi -> usb2_serial_ftdi.
998 Older programs linked against the ports libusb 0.1 need to be
999 redirected to the new stack's libusb20. /etc/libmap.conf can
1001 # Map old usb library to new one for usb2 stack
1002 libusb-0.1.so.8 libusb20.so.1
1005 All USB ethernet devices now attach as interfaces under the name ueN
1006 (eg. ue0). This is to provide a predictable name as vendors often
1007 change usb chipsets in a product without notice.
1010 The ichsmb(4) driver has been changed to require SMBus slave
1011 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1012 All of the other SMBus controller drivers require left-justified
1013 slave addresses, so this change makes all the drivers provide the
1017 INET6 statistics (struct ip6stat) was updated.
1018 netstat(1) needs to be recompiled.
1021 NTFS has been removed from GENERIC kernel on amd64 to match
1022 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1023 will load ntfs.ko module automatically when NTFS support is
1024 actually needed, unless ntfs.ko is not installed or security
1025 level prohibits loading kernel modules. If either is the case,
1026 "options NTFS" has to be added into kernel config.
1029 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1030 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1031 800061. User space tools that rely on the size of struct tcpcb in
1032 tcp_var.h (e.g. sockstat) need to be recompiled.
1035 ng_tty(4) module updated to match the new TTY subsystem.
1036 Due to API change, user-level applications must be updated.
1037 New API support added to mpd5 CVS and expected to be present
1038 in next mpd5.3 release.
1041 With __FreeBSD_version 800060 the makefs tool is part of
1042 the base system (it was a port).
1045 The afdata and ifnet locks have been changed from mutexes to
1046 rwlocks, network modules will need to be re-compiled.
1049 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1050 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1051 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1052 architecture and 40 bytes on 64-bit architecture. The userland
1053 applications "arp" and "ndp" have been updated accordingly.
1054 The output from "netstat -r" shows only routing entries and
1055 none of the L2 information.
1058 __FreeBSD_version 800057 marks the switchover from the
1059 binary ath hal to source code. Users must add the line:
1061 options AH_SUPPORT_AR5416
1063 to their kernel config files when specifying:
1067 The ath_hal module no longer exists; the code is now compiled
1068 together with the driver in the ath module. It is now
1069 possible to tailor chip support (i.e. reduce the set of chips
1070 and thereby the code size); consult ath_hal(4) for details.
1073 __FreeBSD_version 800054 adds memory barriers to
1074 <machine/atomic.h>, new interfaces to ifnet to facilitate
1075 multiple hardware transmit queues for cards that support
1076 them, and a lock-less ring-buffer implementation to
1077 enable drivers to more efficiently manage queueing of
1081 A new version of ZFS (version 13) has been merged to -HEAD.
1082 This version has zpool attribute "listsnapshots" off by
1083 default, which means "zfs list" does not show snapshots,
1084 and is the same as Solaris behavior.
1087 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1090 The uhci, ohci, ehci and slhci USB Host controller drivers have
1091 been put into separate modules. If you load the usb module
1092 separately through loader.conf you will need to load the
1093 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1094 controller add the following to loader.conf:
1100 The ABI used by the PMC toolset has changed. Please keep
1101 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1105 atapci kernel module now includes only generic PCI ATA
1106 driver. AHCI driver moved to ataahci kernel module.
1107 All vendor-specific code moved into separate kernel modules:
1108 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1109 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1110 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1111 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1114 The TTY subsystem of the kernel has been replaced by a new
1115 implementation, which provides better scalability and an
1116 improved driver model. Most common drivers have been migrated to
1117 the new TTY subsystem, while others have not. The following
1118 drivers have not yet been ported to the new TTY layer:
1121 cy, digi, rc, rp, sio
1127 ng_h4, ng_tty, ppp, sl, snp
1129 Adding these drivers to your kernel configuration file shall
1130 cause compilation to fail.
1133 ntpd has been upgraded to 4.2.4p5.
1136 OpenSSH has been upgraded to 5.1p1.
1138 For many years, FreeBSD's version of OpenSSH preferred DSA
1139 over RSA for host and user authentication keys. With this
1140 upgrade, we've switched to the vendor's default of RSA over
1141 DSA. This may cause upgraded clients to warn about unknown
1142 host keys even for previously known hosts. Users should
1143 follow the usual procedure for verifying host keys before
1144 accepting the RSA key.
1146 This can be circumvented by setting the "HostKeyAlgorithms"
1147 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1150 Please note that the sequence of keys offered for
1151 authentication has been changed as well. You may want to
1152 specify IdentityFile in a different order to revert this
1156 The sio(4) driver has been removed from the i386 and amd64
1157 kernel configuration files. This means uart(4) is now the
1158 default serial port driver on those platforms as well.
1160 To prevent collisions with the sio(4) driver, the uart(4) driver
1161 uses different names for its device nodes. This means the
1162 onboard serial port will now most likely be called "ttyu0"
1163 instead of "ttyd0". You may need to reconfigure applications to
1164 use the new device names.
1166 When using the serial port as a boot console, be sure to update
1167 /boot/device.hints and /etc/ttys before booting the new kernel.
1168 If you forget to do so, you can still manually specify the hints
1169 at the loader prompt:
1171 set hint.uart.0.at="isa"
1172 set hint.uart.0.port="0x3F8"
1173 set hint.uart.0.flags="0x10"
1174 set hint.uart.0.irq="4"
1178 The gpt(8) utility has been removed. Use gpart(8) to partition
1182 The version that Linuxulator emulates was changed from 2.4.2
1183 to 2.6.16. If you experience any problems with Linux binaries
1184 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1185 if it fixes the problem contact emulation mailing list.
1188 ISDN4BSD (I4B) was removed from the src tree. You may need to
1189 update a your kernel configuration and remove relevant entries.
1192 I have checked in code to support multiple routing tables.
1193 See the man pages setfib(1) and setfib(2).
1194 This is a hopefully backwards compatible version,
1195 but to make use of it you need to compile your kernel
1196 with options ROUTETABLES=2 (or more up to 16).
1199 The 802.11 wireless support was redone to enable multi-bss
1200 operation on devices that are capable. The underlying device
1201 is no longer used directly but instead wlanX devices are
1202 cloned with ifconfig. This requires changes to rc.conf files.
1203 For example, change:
1204 ifconfig_ath0="WPA DHCP"
1207 ifconfig_wlan0="WPA DHCP"
1208 see rc.conf(5) for more details. In addition, mergemaster of
1209 /etc/rc.d is highly recommended. Simultaneous update of userland
1210 and kernel wouldn't hurt either.
1212 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1213 modules were merged into the base wlan module. All references
1214 to these modules (e.g. in kernel config files) must be removed.
1217 psm(4) has gained write(2) support in native operation level.
1218 Arbitrary commands can be written to /dev/psm%d and status can
1219 be read back from it. Therefore, an application is responsible
1220 for status validation and error recovery. It is a no-op in
1221 other operation levels.
1224 Support for KSE threading has been removed from the kernel. To
1225 run legacy applications linked against KSE libmap.conf may
1226 be used. The following libmap.conf may be used to ensure
1227 compatibility with any prior release:
1229 libpthread.so.1 libthr.so.1
1230 libpthread.so.2 libthr.so.2
1231 libkse.so.3 libthr.so.3
1234 The layout of struct vmspace has changed. This affects libkvm
1235 and any executables that link against libkvm and use the
1236 kvm_getprocs() function. In particular, but not exclusively,
1237 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1238 The effects are minimal, but it's advisable to upgrade world
1242 The latest em driver no longer has support in it for the
1243 82575 adapter, this is now moved to the igb driver. The
1244 split was done to make new features that are incompatible
1245 with older hardware easier to do.
1248 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1249 likewise the kernel option is now GEOM_LINUX_LVM.
1252 The default NFS mount mode has changed from UDP to TCP for
1253 increased reliability. If you rely on (insecurely) NFS
1254 mounting across a firewall you may need to update your
1258 Belatedly note the addition of m_collapse for compacting
1262 The fts(3) structures have been changed to use adequate
1263 integer types for their members and so to be able to cope
1264 with huge file trees. The old fts(3) ABI is preserved
1265 through symbol versioning in libc, so third-party binaries
1266 using fts(3) should still work, although they will not take
1267 advantage of the extended types. At the same time, some
1268 third-party software might fail to build after this change
1269 due to unportable assumptions made in its source code about
1270 fts(3) structure members. Such software should be fixed
1271 by its vendor or, in the worst case, in the ports tree.
1272 FreeBSD_version 800015 marks this change for the unlikely
1273 case that a portable fix is impossible.
1276 To upgrade to -current after this date, you must be running
1277 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1278 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1281 The ADAPTIVE_GIANT kernel option has been retired because its
1282 functionality is the default now.
1285 The AT keyboard emulation of sunkbd(4) has been turned on
1286 by default. In order to make the special symbols of the Sun
1287 keyboards driven by sunkbd(4) work under X these now have
1288 to be configured the same way as Sun USB keyboards driven
1289 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1291 Option "XkbLayout" "us"
1292 Option "XkbRules" "xorg"
1293 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1296 It has been decided that it is desirable to provide ABI
1297 backwards compatibility to the FreeBSD 4/5/6 versions of the
1298 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1299 broken with the introduction of PCI domain support (see the
1300 20070930 entry). Unfortunately, this required the ABI of
1301 PCIOCGETCONF to be broken again in order to be able to
1302 provide backwards compatibility to the old version of that
1303 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1304 again. As for prominent ports this affects neither pciutils
1305 nor xorg-server this time, the hal port needs to be rebuilt
1309 The misnamed kthread_create() and friends have been renamed
1310 to kproc_create() etc. Many of the callers already
1311 used kproc_start()..
1312 I will return kthread_create() and friends in a while
1313 with implementations that actually create threads, not procs.
1314 Renaming corresponds with version 800002.
1323 Avoid using make -j when upgrading. While generally safe, there are
1324 sometimes problems using -j to upgrade. If your upgrade fails with
1325 -j, please try again without -j. From time to time in the past there
1326 have been problems using -j with buildworld and/or installworld. This
1327 is especially true when upgrading between "distant" versions (eg one
1328 that cross a major release boundary or several minor releases, or when
1329 several months have passed on the -current branch).
1331 Sometimes, obscure build problems are the result of environment
1332 poisoning. This can happen because the make utility reads its
1333 environment when searching for values for global variables. To run
1334 your build attempts in an "environmental clean room", prefix all make
1335 commands with 'env -i '. See the env(1) manual page for more details.
1337 When upgrading from one major version to another it is generally best
1338 to upgrade to the latest code in the currently installed branch first,
1339 then do an upgrade to the new branch. This is the best-tested upgrade
1340 path, and has the highest probability of being successful. Please try
1341 this approach before reporting problems with a major version upgrade.
1345 When upgrading the boot ZFS pool to a new version, always follow
1348 1.) recompile and reinstall the ZFS boot loader and boot block
1349 (this is part of "make buildworld" and "make installworld")
1351 2.) update the ZFS boot block on your boot drive
1353 The following example updates the ZFS boot block on the first
1354 partition (freebsd-boot) of a GPT partitioned drive ad0:
1355 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1357 Non-boot pools do not need these updates.
1361 If you are updating from a prior version of FreeBSD (even one just
1362 a few days old), you should follow this procedure. It is the most
1363 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1365 make kernel-toolchain
1366 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1367 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1369 To test a kernel once
1370 ---------------------
1371 If you just want to boot a kernel once (because you are not sure
1372 if it works, or if you want to boot a known bad kernel to provide
1373 debugging information) run
1374 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1375 nextboot -k testkernel
1377 To just build a kernel when you know that it won't mess you up
1378 --------------------------------------------------------------
1379 This assumes you are already running a CURRENT system. Replace
1380 ${arch} with the architecture of your machine (e.g. "i386",
1381 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1383 cd src/sys/${arch}/conf
1384 config KERNEL_NAME_HERE
1385 cd ../compile/KERNEL_NAME_HERE
1390 If this fails, go to the "To build a kernel" section.
1392 To rebuild everything and install it on the current system.
1393 -----------------------------------------------------------
1394 # Note: sometimes if you are running current you gotta do more than
1395 # is listed here if you are upgrading from a really old current.
1397 <make sure you have good level 0 dumps>
1399 make kernel KERNCONF=YOUR_KERNEL_HERE
1401 <reboot in single user> [3]
1409 To cross-install current onto a separate partition
1410 --------------------------------------------------
1411 # In this approach we use a separate partition to hold
1412 # current's root, 'usr', and 'var' directories. A partition
1413 # holding "/", "/usr" and "/var" should be about 2GB in
1416 <make sure you have good level 0 dumps>
1419 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1420 <maybe newfs current's root partition>
1421 <mount current's root partition on directory ${CURRENT_ROOT}>
1422 make installworld DESTDIR=${CURRENT_ROOT}
1423 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1424 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1425 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1426 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1427 <reboot into current>
1428 <do a "native" rebuild/install as described in the previous section>
1429 <maybe install compatibility libraries from ports/misc/compat*>
1433 To upgrade in-place from 8.x-stable to current
1434 ----------------------------------------------
1435 <make sure you have good level 0 dumps>
1437 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1439 <reboot in single user> [3]
1446 Make sure that you've read the UPDATING file to understand the
1447 tweaks to various things you need. At this point in the life
1448 cycle of current, things change often and you are on your own
1449 to cope. The defaults can also change, so please read ALL of
1450 the UPDATING entries.
1452 Also, if you are tracking -current, you must be subscribed to
1453 freebsd-current@freebsd.org. Make sure that before you update
1454 your sources that you have read and understood all the recent
1455 messages there. If in doubt, please track -stable which has
1456 much fewer pitfalls.
1458 [1] If you have third party modules, such as vmware, you
1459 should disable them at this point so they don't crash your
1462 [3] From the bootblocks, boot -s, and then do
1467 adjkerntz -i # if CMOS is wall time
1468 Also, when doing a major release upgrade, it is required that
1469 you boot into single user mode to do the installworld.
1471 [4] Note: This step is non-optional. Failure to do this step
1472 can result in a significant reduction in the functionality of the
1473 system. Attempting to do it by hand is not recommended and those
1474 that pursue this avenue should read this file carefully, as well
1475 as the archives of freebsd-current and freebsd-hackers mailing lists
1476 for potential gotchas. The -U option is also useful to consider.
1477 See mergemaster(8) for more information.
1479 [5] Usually this step is a noop. However, from time to time
1480 you may need to do this if you get unknown user in the following
1481 step. It never hurts to do it all the time. You may need to
1482 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1483 install) after the buildworld before this step if you last updated
1484 from current before 20020224 or from -stable before 20020408.
1486 [6] This only deletes old files and directories. Old libraries
1487 can be deleted by "make delete-old-libs", but you have to make
1488 sure that no program is using those libraries anymore.
1490 [8] In order to have a kernel that can run the 4.x binaries needed to
1491 do an installworld, you must include the COMPAT_FREEBSD4 option in
1492 your kernel. Failure to do so may leave you with a system that is
1493 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1494 required to run the 5.x binaries on more recent kernels. And so on
1495 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1497 Make sure that you merge any new devices from GENERIC since the
1498 last time you updated your kernel config file.
1500 [9] When checking out sources, you must include the -P flag to have
1501 cvs prune empty directories.
1503 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1504 "?=" instead of the "=" assignment operator, so that buildworld can
1505 override the CPUTYPE if it needs to.
1507 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1508 not on the command line, or in /etc/make.conf. buildworld will
1509 warn if it is improperly defined.
1512 This file contains a list, in reverse chronological order, of major
1513 breakages in tracking -current. It is not guaranteed to be a complete
1514 list of such breakages, and only contains entries since October 10, 2007.
1515 If you need to see UPDATING entries from before that date, you will need
1516 to fetch an UPDATING file from an older FreeBSD release.
1518 Copyright information:
1520 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1522 Redistribution, publication, translation and use, with or without
1523 modification, in full or in part, in any form or format of this
1524 document are permitted without further permission from the author.
1526 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1527 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1528 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1529 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1530 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1531 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1532 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1533 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1534 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1535 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1536 POSSIBILITY OF SUCH DAMAGE.
1538 Contact Warner Losh if you have any questions about your use of