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 20140408: p11 FreeBSD-SA-14:05.nfsserver
13 FreeBSD-SA-14:06.openssl
14 Fix deadlock in the NFS server. [SA-14:05]
16 Fix for ECDSA Cache Side-channel Attack in OpenSSL. [SA-14:06]
18 20140114: p10 FreeBSD-SA-14:01.bsnmpd
21 FreeBSD-EN-14:01.random
23 Fix bsnmpd remote denial of service vulnerability. [SA-14:01]
25 Fix ntpd distributed reflection Denial of Service
26 vulnerability. [SA-14:02]
28 Fix BIND remote denial of service vulnerability. [SA-14:04]
30 Disable hardware RNGs by default. [EN-14:01]
32 Fix incorrect coalescing of stack entry with mmap. [EN-14:02]
34 20131128: p9 FreeBSD-EN-13:05.freebsd-update
35 Fix error in patch for FreeBSD-EN-13:04.freebsd-update.
37 20131026: p8 FreeBSD-EN-13:04.freebsd-update
38 Fix multiple freebsd-update bugs that break upgrading to
41 20130910: p7 FreeBSD-SA-13:12.ifioctl
42 FreeBSD-SA-13:13.nullfs
44 In IPv6 and NetATM, stop SIOCSIFADDR, SIOCSIFBRDADDR,
45 SIOCSIFDSTADDR and SIOCSIFNETMASK at the socket layer rather
46 than pass them on to the link layer without validation or
47 credential checks. [SA-13:12]
49 Prevent cross-mount hardlinks between different nullfs mounts
50 of the same underlying filesystem. [SA-13:13]
52 20130822: p6 FreeBSD-SA-13:09.ip_multicast
55 Fix an integer overflow in computing the size of a temporary buffer
56 can result in a buffer which is too small for the requested
59 Fix a bug that could lead to kernel memory disclosure with
60 SCTP state cookie. [13:10]
62 Fix a data corruption problem with mfi(4) operating on > 2TB
63 disks in a JBOD. [EN-13:03]
65 20130726: p5 FreeBSD-SA-13:07.bind FreeBSD-SA-13:08.nfsserver
66 Fix Denial of Service vulnerability in named(8). [13:07]
68 Fix a bug that allows remote client bypass the normal
69 access checks when when -network or -host restrictions are
70 used at the same time with -mapall. [13:08]
72 20130618: p4 FreeBSD-SA-13:06.mmap
73 Fix a bug that allowed a tracing process (e.g. gdb) to write
74 to a memory-mapped file in the traced process's address space
75 even if neither the traced process nor the tracing process had
76 write access to that file.
78 20130429: p3 FreeBSD-SA-13:05.nfsserver
79 Fix a bug that allows NFS clients to issue READDIR on files.
81 20130402: p2 FreeBSD-SA-13:03.openssl FreeBSD-SA-13:04.bind
82 Fix multiple vulnerabilities in OpenSSL.
84 Fix Denial of Service vulnerability in named(8).
86 20130218: p1 FreeBSD-SA-13:01.bind FreeBSD-SA-13:02.libc
87 Fix Denial of Service vulnerability in named(8) with DNS64.
89 Fix Denial of Service vulnerability in libc's glob(3) functionality.
95 The sparc64 ZFS loader has been changed to no longer try to auto-
96 detect ZFS providers based on diskN aliases but now requires these
97 to be explicitly listed in the OFW boot-device environment variable.
100 Now unix domain sockets behave "as expected" on nullfs(5). Previously
101 nullfs(5) did not pass through all behaviours to the underlying layer,
102 as a result if we bound to a socket on the lower layer we could connect
103 only to the lower path; if we bound to the upper layer we could connect
104 only to the upper path. The new behavior is one can connect to both the
105 lower and the upper paths regardless what layer path one binds to.
108 The acpi_wmi(4) status device /dev/wmistat has been renamed to
112 A new VOP_ADVISE() was added to support posix_fadvise(2). All
113 filesystem modules must be recompiled.
116 The interface of the VOP_VPTOCNP(9) changed, now the returned
117 vnode shall be referenced, previously it was required to be
118 only held. All in-tree filesystems are converted.
124 The broken amd(4) driver has been replaced with esp(4) in the amd64,
125 i386 and pc98 GENERIC kernel configuration files.
128 This commit modifies vfs_register() so that it uses a hash
129 calculation to set vfc_typenum, which is enabled by default.
130 The first time a system is booted after this change, the
131 vfc_typenum values will change for all file systems. The
132 main effect of this is a change to the NFS server file handles
133 for file systems that use vfc_typenum in their fsid, such as ZFS.
134 It will, however, prevent vfc_typenum from changing when file
135 systems are loaded in a different order for subsequent reboots.
136 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
137 until you are ready to remount all NFS clients after a reboot.
140 Bump the shared library version numbers for libraries that
141 do not use symbol versioning, have changed the ABI compared
142 to stable/8 and which shared library version was not bumped.
143 Done as part of 9.0-RELEASE cycle.
146 During the merge of Capsicum features, the fget(9) KPI was modified.
147 This may require the rebuilding of out-of-tree device drivers --
148 issues have been reported specifically with the nVidia device driver.
149 __FreeBSD_version is bumped to 900041.
151 Also, there is a period between 20110811 and 20110814 where the
152 special devices /dev/{stdin,stdout,stderr} did not work correctly.
153 Building world from a kernel during that window may not work.
156 The packet filter (pf) code has been updated to OpenBSD 4.5.
157 You need to update userland tools to be in sync with kernel.
158 This update breaks backward compatibility with earlier pfsync(4)
159 versions. Care must be taken when updating redundant firewall setups.
162 The following sysctls and tunables are retired on x86 platforms:
164 machdep.hlt_logical_cpus
165 The following sysctl is retired:
166 machdep.hyperthreading_allowed
167 The sysctls were supposed to provide a way to dynamically offline and
168 online selected CPUs on x86 platforms, but the implementation has not
169 been reliable especially with SCHED_ULE scheduler.
170 machdep.hyperthreading_allowed tunable is still available to ignore
171 hyperthreading CPUs at OS level.
172 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
173 where X is an APIC ID of a CPU. Be advised, though, that disabling
174 CPUs in non-uniform fashion will result in non-uniform topology and
175 may lead to sub-optimal system performance with SCHED_ULE, which is
179 cpumask_t type is retired and cpuset_t is used in order to describe
183 Changes to ifconfig(8) for dynamic address family detection mandate
184 that you are running a kernel of 20110525 or later. Make sure to
185 follow the update procedure to boot a new kernel before installing
189 Support for sun4v architecture is officially dropped
192 Several KPI breaking changes have been committed to the mii(4) layer,
193 the PHY drivers and consequently some Ethernet drivers using mii(4).
194 This means that miibus.ko and the modules of the affected Ethernet
195 drivers need to be recompiled.
197 Note to kernel developers: Given that the OUI bit reversion problem
198 was fixed as part of these changes all mii(4) commits related to OUIs,
199 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
200 handling, no longer can be merged verbatim to stable/8 and previous
204 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
205 into their kernel configurations along with 'device pci'.
208 The default NFS client is now the new NFS client, so fstype "newnfs"
209 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
210 Although mounts via fstype "nfs" will usually work without userland
211 changes, it is recommended that the mount(8) and mount_nfs(8)
212 commands be rebuilt from sources and that a link to mount_nfs called
213 mount_oldnfs be created. The new client is compiled into the
214 kernel with "options NFSCL" and this is needed for diskless root
215 file systems. The GENERIC kernel configs have been changed to use
216 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
217 To use the regular/old client, you can "mount -t oldnfs ...". For
218 a diskless root file system, you must also include a line like:
220 vfs.root.mountfrom="oldnfs:"
222 in the boot/loader.conf on the root fs on the NFS server to make
223 a diskless root fs use the old client.
226 The GENERIC kernels for all architectures now default to the new
227 CAM-based ATA stack. It means that all legacy ATA drivers were
228 removed and replaced by respective CAM drivers. If you are using
229 ATA device names in /etc/fstab or other places, make sure to update
230 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
231 where 'Y's are the sequential numbers starting from zero for each type
232 in order of detection, unless configured otherwise with tunables,
233 see cam(4)). There will be symbolic links created in /dev/ to map
234 old adX devices to the respective adaY. They should provide basic
235 compatibility for file systems mounting in most cases, but they do
236 not support old user-level APIs and do not have respective providers
237 in GEOM. Consider using updated management tools with new device names.
239 It is possible to load devices ahci, ata, siis and mvs as modules,
240 but option ATA_CAM should remain in kernel configuration to make ata
241 module work as CAM driver supporting legacy ATA controllers. Device ata
242 still can be used in modular fashion (atacore + ...). Modules atadisk
243 and atapi* are not used and won't affect operation in ATA_CAM mode.
244 Note that to use CAM-based ATA kernel should include CAM devices
245 scbus, pass, da (or explicitly ada), cd and optionally others. All of
246 them are parts of the cam module.
248 ataraid(4) functionality is now supported by the RAID GEOM class.
249 To use it you can load geom_raid kernel module and use graid(8) tool
250 for management. Instead of /dev/arX device names, use /dev/raid/rX.
252 No kernel config options or code have been removed, so if a problem
253 arises, please report it and optionally revert to the old ATA stack.
254 In order to do it you can remove from the kernel config:
259 , and instead add back:
260 device atadisk # ATA disk drives
261 device ataraid # ATA RAID drives
262 device atapicd # ATAPI CDROM drives
263 device atapifd # ATAPI floppy drives
264 device atapist # ATAPI tape drives
267 The default NFS server has been changed to the new server, which
268 was referred to as the experimental server. If you need to switch
269 back to the old NFS server, you must now put the "-o" option on
270 both the mountd and nfsd commands. This can be done using the
271 mountd_flags and nfs_server_flags rc.conf variables until an
272 update to the rc scripts is committed, which is coming soon.
275 The GNU Objective-C runtime library (libobjc), and other Objective-C
276 related components have been removed from the base system. If you
277 require an Objective-C library, please use one of the available ports.
280 ath(4) has been split into bus- and device- modules. if_ath contains
281 the HAL, the TX rate control and the network device code. if_ath_pci
282 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
283 contains the AHB glue. Users need to load both if_ath_pci and if_ath
284 in order to use ath on everything else.
286 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
287 need to load if_ath and if_ath_pci for ath(4) operation.
290 As part of the replacement of sysinstall, the process of building
291 release media has changed significantly. For details, please re-read
292 release(7), which has been updated to reflect the new build process.
295 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
296 is the last available version under GPLv2. It brings a number of new
297 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
298 4.1 and SSE 4.2), better support for powerpc64, a number of new
299 directives, and lots of other small improvements. See the ChangeLog
300 file in contrib/binutils for the full details.
303 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
304 compliant, and will now use half of hash for authentication.
305 This will break interoperability with all stacks (including all
306 actual FreeBSD versions) who implement
307 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
309 The only workaround with such peers is to use another HMAC
310 algorithm for IPsec ("phase 2") authentication.
313 Remove the uio_yield prototype and symbol. This function has
314 been misnamed since it was introduced and should not be
315 globally exposed with this name. The equivalent functionality
316 is now available using kern_yield(curthread->td_user_pri).
317 The function remains undocumented.
320 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
321 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
322 for scalar sysctls is defined but disabled. Code that needs
323 UQUAD to pass the type checking that must compile on older
324 systems where the define is not present can check against
325 __FreeBSD_version >= 900030.
327 The system dialog(1) has been replaced with a new version previously
328 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
329 with the previous version, but the libdialog associated with it has
330 a largely incompatible API. As such, the original version of libdialog
331 will be kept temporarily as libodialog, until its base system consumers
332 are replaced or updated. Bump __FreeBSD_version to 900030.
335 If you are trying to run make universe on a -stable system, and you get
336 the following warning:
337 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
338 unknown. config(8) likely too old."
339 or something similar to it, then you must upgrade your -stable system
340 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
341 stable/8) or build the config from the latest stable/8 branch and
342 install it on your system.
344 Prior to this date, building a current universe on 8-stable system from
345 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
346 error in the first kernel build phase. A new config on those old
347 systems will fix that problem for older versions of -current.
350 The TCP stack has been modified to allow Khelp modules to interact with
351 it via helper hook points and store per-connection data in the TCP
352 control block. Bump __FreeBSD_version to 900029. User space tools that
353 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
357 Generic IEEE 802.3 annex 31B full duplex flow control support has been
358 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
359 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
360 to take advantage of it instead of using custom implementations. This
361 means that these drivers now no longer unconditionally advertise
362 support for flow control but only do so if flow control is a selected
363 media option. This was implemented in the generic support that way in
364 order to allow flow control to be switched on and off via ifconfig(8)
365 with the PHY specific default to typically off in order to protect
366 from unwanted effects. Consequently, if you used flow control with
367 one of the above mentioned drivers you now need to explicitly enable
369 ifconfig bge0 media auto mediaopt flowcontrol
371 Along with the above mentioned changes generic support for setting
372 1000baseT master mode also has been added and brgphy(4), ciphy(4),
373 e1000phy(4) as well as ip1000phy(4) have been converted to take
374 advantage of it. This means that these drivers now no longer take the
375 link0 parameter for selecting master mode but the master media option
376 has to be used instead, for example like in the following:
377 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
379 Selection of master mode now is also available with all other PHY
380 drivers supporting 1000baseT.
383 The TCP stack has received a significant update to add support for
384 modularised congestion control and generally improve the clarity of
385 congestion control decisions. Bump __FreeBSD_version to 900025. User
386 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
387 sockstat) need to be recompiled.
390 The man(1) utility has been replaced by a new version that no longer
391 uses /etc/manpath.config. Please consult man.conf(5) for how to
392 migrate local entries to the new format.
395 The copyright strings printed by login(1) and sshd(8) at the time of a
396 new connection have been removed to follow other operating systems and
400 A workaround for a fixed ld bug has been removed in kernel code,
401 so make sure that your system ld is built from sources after
402 revision 210245 from 2010-07-19 (r211583 if building head kernel
403 on stable/8, r211584 for stable/7; both from 2010-08-21).
404 A symptom of incorrect ld version is different addresses for
405 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
408 The $ipv6_prefer variable in rc.conf(5) has been split into
409 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
411 The $ip6addrctl_policy is a variable to choose a pre-defined
412 address selection policy set by ip6addrctl(8). A value
413 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
416 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
417 flag (see an entry of 20090926) is set on an interface with no
418 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
419 security reason. If you want IPv6 link-local address on all
420 interfaces by default, set this to "YES".
422 The old ipv6_prefer="YES" is equivalent to
423 ipv6_activate_all_interfaces="YES" and
424 ip6addrctl_policy="ipv6_prefer".
427 DTrace has grown support for userland tracing. Due to this, DTrace is
428 now i386 and amd64 only.
429 dtruss(1) is now installed by default on those systems and a new
430 kernel module is needed for userland tracing: fasttrap.
431 No changes to your kernel config file are necessary to enable
432 userland tracing, but you might consider adding 'STRIP=' and
433 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
434 to have informative userland stack traces in DTrace (ustack).
437 The acpi_aiboost(4) driver has been removed in favor of the new
438 aibs(4) driver. You should update your kernel configuration file.
441 BSD grep has been imported to the base system and it is built by
442 default. It is completely BSD licensed, highly GNU-compatible, uses
443 less memory than its GNU counterpart and has a small codebase.
444 However, it is slower than its GNU counterpart, which is mostly
445 noticeable for larger searches, for smaller ones it is measurable
446 but not significant. The reason is complex, the most important factor
447 is that we lack a modern and efficient regex library and GNU
448 overcomes this by optimizing the searches internally. Future work
449 on improving the regex performance is planned, for the meantime,
450 users that need better performance, can build GNU grep instead by
451 setting the WITH_GNU_GREP knob.
454 Due to the import of powerpc64 support, all existing powerpc kernel
455 configuration files must be updated with a machine directive like this:
456 machine powerpc powerpc
458 In addition, an updated config(8) is required to build powerpc kernels
462 A new version of ZFS (version 15) has been merged to -HEAD.
463 This version uses a python library for the following subcommands:
464 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
465 For full functionality of these commands the following port must
466 be installed: sysutils/py-zfs
469 'vm_page's are now hashed by physical address to an array of mutexes.
470 Currently this is only used to serialize access to hold_count. Over
471 time the page queue mutex will be peeled away. This changes the size
472 of pmap on every architecture. And requires all callers of vm_page_hold
473 and vm_page_unhold to be updated.
476 WITH_CTF can now be specified in src.conf (not recommended, there
477 are some problems with static executables), make.conf (would also
478 affect ports which do not use GNU make and do not override the
479 compile targets) or in the kernel config (via "makeoptions
481 When WITH_CTF was specified there before this was silently ignored,
482 so make sure that WITH_CTF is not used in places which could lead
483 to unwanted behavior.
486 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
487 to allow 32-bit compatibility on non-x86 platforms. All kernel
488 configurations on amd64 and ia64 platforms using these options must
489 be modified accordingly.
492 The utmp user accounting database has been replaced with utmpx,
493 the user accounting interface standardized by POSIX.
494 Unfortunately the semantics of utmp and utmpx don't match,
495 making it practically impossible to support both interfaces.
496 The user accounting database is used by tools like finger(1),
497 last(1), talk(1), w(1) and ac(8).
499 All applications in the base system use utmpx. This means only
500 local binaries (e.g. from the ports tree) may still use these
501 utmp database files. These applications must be rebuilt to make
504 After the system has been upgraded, it is safe to remove the old
505 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
506 assuming their contents is of no importance anymore. Old wtmp
507 databases can only be used by last(1) and ac(8) after they have
508 been converted to the new format using wtmpcvt(1).
511 Introduce the kernel thread "deadlock resolver" (which can be enabled
512 via the DEADLKRES option, see NOTES for more details) and the
513 sleepq_type() function for sleepqueues.
516 The rc.firewall and rc.firewall6 were unified, and
517 rc.firewall6 and rc.d/ip6fw were removed.
518 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
519 variables are obsoleted. Instead, the following new rc
520 variables are added to rc.d/ipfw:
522 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
523 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
524 firewall_simple_onet_ipv6, firewall_trusted_ipv6
526 The meanings correspond to the relevant IPv4 variables.
532 The default terminal emulation for syscons(4) has been changed
533 from cons25 to xterm on all platforms except pc98. This means
534 that the /etc/ttys file needs to be updated to ensure correct
535 operation of applications on the console.
537 The terminal emulation style can be toggled per window by using
538 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
539 options can be used to change the compile-time default back to
542 To prevent graphical artifacts, make sure the TERM environment
543 variable is set to match the terminal emulation that is being
544 performed by syscons(4).
547 The layout of the structure ieee80211req_scan_result has changed.
548 Applications that require wireless scan results (e.g. ifconfig(8))
549 from net80211 need to be recompiled.
551 Applications such as wpa_supplicant(8) may require a full world
552 build without using NO_CLEAN in order to get synchronized with the
556 The iwn(4) driver has been updated to support the 5000 and 5150 series.
557 There's one kernel module for each firmware. Adding "device iwnfw"
558 to the kernel configuration file means including all three firmware
559 images inside the kernel. If you want to include just the one for
560 your wireless card, use the devices iwn4965fw, iwn5000fw or
564 The rc.d/network_ipv6, IPv6 configuration script has been integrated
565 into rc.d/netif. The changes are the following:
567 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
568 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
569 Note that both variables need the "inet6" keyword at the head.
571 Do not set $ipv6_network_interfaces manually if you do not
572 understand what you are doing. It is not needed in most cases.
574 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
577 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
578 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
580 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
581 all configured interfaces have "inet6 accept_rtadv" in the
582 $ifconfig_IF_ipv6. These are for backward compatibility.
584 3. A new variable $ipv6_prefer has been added. If NO, IPv6
585 functionality of interfaces with no corresponding
586 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
587 and the default address selection policy of ip6addrctl(8)
588 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
589 Note that if you want to configure IPv6 functionality on the
590 disabled interfaces after boot, first you need to clear the flag by
591 using ifconfig(8) like:
593 ifconfig em0 inet6 -ifdisabled
595 If YES, the default address selection policy is set as
598 The default value of $ipv6_prefer is NO.
600 4. If your system need to receive Router Advertisement messages,
601 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
602 scripts automatically invoke rtsol(8) when the interface becomes
603 UP. The Router Advertisement messages are used for SLAAC
604 (State-Less Address AutoConfiguration).
607 802.11s D3.03 support was committed. This is incompatible with the
608 previous code, which was based on D3.0.
611 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
612 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
613 control whether accepting Router Advertisement messages or not.
614 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
615 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
616 The ifconfig(8) utility now supports these flags.
619 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
620 mount(8) and -a option for df(1) to see them.
623 The old tunable hw.bus.devctl_disable has been superseded by
624 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
625 replaced by hw.bus.devctl_queue=0. The default for this new tunable
629 Remove the option STOP_NMI. The default action is now to use NMI only
630 for KDB via the newly introduced function stop_cpus_hard() and
631 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
634 The stable/8 branch created in subversion. This corresponds to the
635 RELENG_8 branch in CVS.
638 Bump the shared library version numbers for all libraries that do not
639 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
640 __FreeBSD_version to 800105.
643 Due to changes in the implementation of virtual network stack support,
644 all network-related kernel modules must be recompiled. As this change
645 breaks the ABI, bump __FreeBSD_version to 800104.
648 The TOE interface to the TCP syncache has been modified to remove
649 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
650 The cxgb driver is the only TOE consumer affected by this change, and
651 needs to be recompiled along with the kernel. As this change breaks
652 the ABI, bump __FreeBSD_version to 800103.
655 Padding has been added to struct tcpcb, sackhint and tcpstat in
656 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
657 maintaining the ABI. However, this change breaks the ABI, so bump
658 __FreeBSD_version to 800102. User space tools that rely on the size of
659 any of these structs (e.g. sockstat) need to be recompiled.
662 The NFS_LEGACYRPC option has been removed along with the old kernel
663 RPC implementation that this option selected. Kernel configurations
664 may need to be adjusted.
667 The network interface device nodes at /dev/net/<interface> have been
668 removed. All ioctl operations can be performed the normal way using
669 routing sockets. The kqueue functionality can generally be replaced
670 with routing sockets.
673 The documentation from the FreeBSD Documentation Project (Handbook,
674 FAQ, etc.) is now installed via packages by sysinstall(8) and under
675 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
678 The ABI of various structures related to the SYSV IPC API have been
679 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
680 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
684 Layout of struct vnet has changed as routing related variables were
685 moved to their own Vimage module. Modules need to be recompiled. Bump
686 __FreeBSD_version to 800099.
689 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
690 respectively. As long as no more than 16 groups per process are used,
691 no changes should be visible. When more than 16 groups are used, old
692 binaries may fail if they call getgroups() or getgrouplist() with
693 statically sized storage. Recompiling will work around this, but
694 applications should be modified to use dynamically allocated storage
695 for group arrays as POSIX.1-2008 does not cap an implementation's
696 number of supported groups at NGROUPS_MAX+1 as previous versions did.
698 NFS and portalfs mounts may also be affected as the list of groups is
699 truncated to 16. Users of NFS who use more than 16 groups, should
700 take care that negative group permissions are not used on the exported
701 file systems as they will not be reliable unless a GSSAPI based
702 authentication method is used.
705 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
706 option compiles in the support for adaptive spinning for lockmgrs
707 which want to enable it. The lockinit() function now accepts the flag
708 LK_ADAPTIVE in order to make the lock object subject to adaptive
709 spinning when both held in write and read mode.
712 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
713 changed. User applications that use this ioctl need to be rebuilt.
716 The layout of struct thread has changed. Kernel and modules need to
720 The layout of structs ifnet, domain, protosw and vnet_net has changed.
721 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
724 window(1) has been removed from the base system. It can now be
725 installed from ports. The port is called misc/window.
728 The way we are storing and accessing `routing table' entries has
729 changed. Programs reading the FIB, like netstat, need to be
733 A new netisr implementation has been added for FreeBSD 8. Network
734 file system modules, such as igmp, ipdivert, and others, should be
736 Bump __FreeBSD_version to 800096.
739 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
743 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
744 Bump __FreeBSD_version to 800094.
747 Add mnt_xflag field to 'struct mount'. File system modules need to be
749 Bump __FreeBSD_version to 800093.
752 The compiling option ADAPTIVE_SX has been retired while it has been
753 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
754 The KPI for sx_init_flags() changes as accepting flags:
755 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
756 been introduced in order to handle the reversed logic.
757 Bump __FreeBSD_version to 800092.
760 Add support for hierarchical jails. Remove global securelevel.
761 Bump __FreeBSD_version to 800091.
764 The layout of struct vnet_net has changed, therefore modules
766 Bump __FreeBSD_version to 800090.
769 The newly imported zic(8) produces a new format in the output. Please
770 run tzsetup(8) to install the newly created data to /etc/localtime.
773 The sysctl tree for the usb stack has renamed from hw.usb2.* to
774 hw.usb.* and is now consistent again with previous releases.
777 802.11 monitor mode support was revised and driver api's were changed.
778 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
779 of DLT_IEEE802_11. No user-visible data structures were changed but
780 applications that use DLT_IEEE802_11 may require changes.
781 Bump __FreeBSD_version to 800088.
784 The layout of the following structs has changed: sysctl_oid,
785 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
786 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
787 panics may be experienced. World rebuild is required for
788 correctly checking networking state from userland.
789 Bump __FreeBSD_version to 800085.
792 MLDv2 and Source-Specific Multicast (SSM) have been merged
793 to the IPv6 stack. VIMAGE hooks are in but not yet used.
794 The implementation of SSM within FreeBSD's IPv6 stack closely
795 follows the IPv4 implementation.
797 For kernel developers:
799 * The most important changes are that the ip6_output() and
800 ip6_input() paths no longer take the IN6_MULTI_LOCK,
801 and this lock has been downgraded to a non-recursive mutex.
803 * As with the changes to the IPv4 stack to support SSM, filtering
804 of inbound multicast traffic must now be performed by transport
805 protocols within the IPv6 stack. This does not apply to TCP and
806 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
808 * The KPIs used by IPv6 multicast are similar to those used by
809 the IPv4 stack, with the following differences:
810 * im6o_mc_filter() is analogous to imo_multicast_filter().
811 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
812 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
813 * IN6_LOOKUP_MULTI() has been deprecated and removed.
814 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
815 for MLDv1 have an additional 'timer' argument which is used to
816 jitter the initial membership report for the solicited-node
817 multicast membership on-link.
818 * This is not strictly needed for MLDv2, which already jitters
819 its report transmissions. However, the 'timer' argument is
820 preserved in case MLDv1 is active on the interface.
822 * The KAME linked-list based IPv6 membership implementation has
823 been refactored to use a vector similar to that used by the IPv4
825 Code which maintains a list of its own multicast memberships
826 internally, e.g. carp, has been updated to reflect the new
829 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
830 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
831 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
832 implementation constraint which needs to be addressed in HEAD.
834 For application developers:
836 * The changes are broadly similar to those made for the IPv4
839 * The use of IPv4 and IPv6 multicast socket options on the same
840 socket, using mapped addresses, HAS NOT been tested or supported.
842 * There are a number of issues with the implementation of various
843 IPv6 multicast APIs which need to be resolved in the API surface
844 before the implementation is fully compatible with KAME userland
845 use, and these are mostly to do with interface index treatment.
847 * The literature available discusses the use of either the delta / ASM
848 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
849 using setsourcefilter(3)/getsourcefilter(3). For more information
850 please refer to RFC 3768, 'Socket Interface Extensions for
851 Multicast Source Filters'.
853 * Applications which use the published RFC 3678 APIs should be fine.
855 For systems administrators:
857 * The mtest(8) utility has been refactored to support IPv6, in
858 addition to IPv4. Interface addresses are no longer accepted
859 as arguments, their names must be used instead. The utility
860 will map the interface name to its first IPv4 address as
861 returned by getifaddrs(3).
863 * The ifmcstat(8) utility has also been updated to print the MLDv2
864 endpoint state and source filter lists via sysctl(3).
866 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
867 loopback of IPv6 multicast datagrams by default; it defaults to 1
868 to preserve the existing behaviour. Disabling multicast loopback is
869 recommended for optimal system performance.
871 * The IPv6 MROUTING code has been changed to examine this sysctl
872 instead of attempting to perform a group lookup before looping
873 back forwarded datagrams.
875 Bump __FreeBSD_version to 800084.
878 Implement low-level Bluetooth HCI API.
879 Bump __FreeBSD_version to 800083.
882 The layout of struct malloc_type, used by modules to register new
883 memory allocation types, has changed. Most modules will need to
884 be rebuilt or panics may be experienced.
885 Bump __FreeBSD_version to 800081.
888 Anticipate overflowing inp_flags - add inp_flags2.
889 This changes most offsets in inpcb, so checking v4 connection
890 state will require a world rebuild.
891 Bump __FreeBSD_version to 800080.
894 Add an llentry to struct route and struct route_in6. Modules
895 embedding a struct route will need to be recompiled.
896 Bump __FreeBSD_version to 800079.
899 The size of rt_metrics_lite and by extension rtentry has changed.
900 Networking administration apps will need to be recompiled.
901 The route command now supports show as an alias for get, weighting
902 of routes, sticky and nostick flags to alter the behavior of stateful
904 Bump __FreeBSD_version to 800078.
907 Do not use Giant for kbdmux(4) locking. This is wrong and
908 apparently causing more problems than it solves. This will
909 re-open the issue where interrupt handlers may race with
910 kbdmux(4) in polling mode. Typical symptoms include (but
911 not limited to) duplicated and/or missing characters when
912 low level console functions (such as gets) are used while
913 interrupts are enabled (for example geli password prompt,
914 mountroot prompt etc.). Disabling kbdmux(4) may help.
917 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
918 kernel modules referencing any of the above need to be recompiled.
919 Bump __FreeBSD_version to 800075.
922 GEOM_PART has become the default partition slicer for storage devices,
923 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
924 introduces some changes:
926 MSDOS/EBR: the devices created from MSDOS extended partition entries
927 (EBR) can be named differently than with GEOM_MBR and are now symlinks
928 to devices with offset-based names. fstabs may need to be modified.
930 BSD: the "geometry does not match label" warning is harmless in most
931 cases but it points to problems in file system misalignment with
932 disk geometry. The "c" partition is now implicit, covers the whole
933 top-level drive and cannot be (mis)used by users.
935 General: Kernel dumps are now not allowed to be written to devices
936 whose partition types indicate they are meant to be used for file
937 systems (or, in case of MSDOS partitions, as something else than
940 Most of these changes date approximately from 200812.
943 The uscanner(4) driver has been removed from the kernel. This follows
944 Linux removing theirs in 2.6 and making libusb the default interface
948 The multicast forwarding code has been cleaned up. netstat(1)
949 only relies on KVM now for printing bandwidth upcall meters.
950 The IPv4 and IPv6 modules are split into ip_mroute_mod and
951 ip6_mroute_mod respectively. The config(5) options for statically
952 compiling this code remain the same, i.e. 'options MROUTING'.
955 Support for the IFF_NEEDSGIANT network interface flag has been
956 removed, which means that non-MPSAFE network device drivers are no
957 longer supported. In particular, if_ar, if_sr, and network device
958 drivers from the old (legacy) USB stack can no longer be built or
962 POSIX.1 Native Language Support (NLS) has been enabled in libc and
963 a bunch of new language catalog files have also been added.
964 This means that some common libc messages are now localized and
965 they depend on the LC_MESSAGES environmental variable.
968 The k8temp(4) driver has been renamed to amdtemp(4) since
969 support for Family 10 and Family 11 CPU families was added.
972 IGMPv3 and Source-Specific Multicast (SSM) have been merged
973 to the IPv4 stack. VIMAGE hooks are in but not yet used.
975 For kernel developers, the most important changes are that the
976 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
977 and this lock has been downgraded to a non-recursive mutex.
979 Transport protocols (UDP, Raw IP) are now responsible for filtering
980 inbound multicast traffic according to group membership and source
981 filters. The imo_multicast_filter() KPI exists for this purpose.
982 Transports which do not use multicast (SCTP, TCP) already reject
983 multicast by default. Forwarding and receive performance may improve
984 as a mutex acquisition is no longer needed in the ip_input()
985 low-level input path. in_addmulti() and in_delmulti() are shimmed
986 to new KPIs which exist to support SSM in-kernel.
988 For application developers, it is recommended that loopback of
989 multicast datagrams be disabled for best performance, as this
990 will still cause the lock to be taken for each looped-back
991 datagram transmission. The net.inet.ip.mcast.loop sysctl may
992 be tuned to 0 to disable loopback by default; it defaults to 1
993 to preserve the existing behaviour.
995 For systems administrators, to obtain best performance with
996 multicast reception and multiple groups, it is always recommended
997 that a card with a suitably precise hash filter is used. Hash
998 collisions will still result in the lock being taken within the
999 transport protocol input path to check group membership.
1001 If deploying FreeBSD in an environment with IGMP snooping switches,
1002 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1003 enabled; this forces 224.0.0.0/24 group membership to be announced
1006 The size of 'struct igmpstat' has changed; netstat needs to be
1007 recompiled to reflect this.
1008 Bump __FreeBSD_version to 800070.
1011 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1012 updated to use it. This requires a buildworld/installworld in order to
1013 update the library and dependencies (usbconfig, etc). Its advisable to
1014 rebuild all ports which uses libusb. More specific directions are given
1015 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1016 libusb are no longer required and can be removed.
1019 A workaround is committed to allow the creation of System V shared
1020 memory segment of size > 2 GB on the 64-bit architectures.
1021 Due to a limitation of the existing ABI, the shm_segsz member
1022 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1023 wrong for large segments. Note that limits must be explicitly
1024 raised to allow such segments to be created.
1027 The layout of struct ifnet has changed, requiring a rebuild of all
1028 network device driver modules.
1031 The /dev handling for the new USB stack has changed, a
1032 buildworld/installworld is required for libusb20.
1035 The new USB2 stack has now been permanently moved in and all kernel and
1036 module names reverted to their previous values (eg, usb, ehci, ohci,
1037 ums, ...). The old usb stack can be compiled in by prefixing the name
1038 with the letter 'o', the old usb modules have been removed.
1039 Updating entry 20090216 for xorg and 20090215 for libmap may still
1043 The rc.conf(5) option if_up_delay has been renamed to
1044 defaultroute_delay to better reflect its purpose. If you have
1045 customized this setting in /etc/rc.conf you need to update it to
1049 xorg 7.4 wants to configure its input devices via hald which does not
1050 yet work with USB2. If the keyboard/mouse does not work in xorg then
1052 Option "AllowEmptyInput" "off"
1053 to your ServerLayout section. This will cause X to use the configured
1054 kbd and mouse sections from your xorg.conf.
1057 The GENERIC kernels for all architectures now default to the new USB2
1058 stack. No kernel config options or code have been removed so if a
1059 problem arises please report it and optionally revert to the old USB
1060 stack. If you are loading USB kernel modules or have a custom kernel
1061 that includes GENERIC then ensure that usb names are also changed over,
1062 eg uftdi -> usb2_serial_ftdi.
1064 Older programs linked against the ports libusb 0.1 need to be
1065 redirected to the new stack's libusb20. /etc/libmap.conf can
1067 # Map old usb library to new one for usb2 stack
1068 libusb-0.1.so.8 libusb20.so.1
1071 All USB ethernet devices now attach as interfaces under the name ueN
1072 (eg. ue0). This is to provide a predictable name as vendors often
1073 change usb chipsets in a product without notice.
1076 The ichsmb(4) driver has been changed to require SMBus slave
1077 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1078 All of the other SMBus controller drivers require left-justified
1079 slave addresses, so this change makes all the drivers provide the
1083 INET6 statistics (struct ip6stat) was updated.
1084 netstat(1) needs to be recompiled.
1087 NTFS has been removed from GENERIC kernel on amd64 to match
1088 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1089 will load ntfs.ko module automatically when NTFS support is
1090 actually needed, unless ntfs.ko is not installed or security
1091 level prohibits loading kernel modules. If either is the case,
1092 "options NTFS" has to be added into kernel config.
1095 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1096 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1097 800061. User space tools that rely on the size of struct tcpcb in
1098 tcp_var.h (e.g. sockstat) need to be recompiled.
1101 ng_tty(4) module updated to match the new TTY subsystem.
1102 Due to API change, user-level applications must be updated.
1103 New API support added to mpd5 CVS and expected to be present
1104 in next mpd5.3 release.
1107 With __FreeBSD_version 800060 the makefs tool is part of
1108 the base system (it was a port).
1111 The afdata and ifnet locks have been changed from mutexes to
1112 rwlocks, network modules will need to be re-compiled.
1115 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1116 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1117 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1118 architecture and 40 bytes on 64-bit architecture. The userland
1119 applications "arp" and "ndp" have been updated accordingly.
1120 The output from "netstat -r" shows only routing entries and
1121 none of the L2 information.
1124 __FreeBSD_version 800057 marks the switchover from the
1125 binary ath hal to source code. Users must add the line:
1127 options AH_SUPPORT_AR5416
1129 to their kernel config files when specifying:
1133 The ath_hal module no longer exists; the code is now compiled
1134 together with the driver in the ath module. It is now
1135 possible to tailor chip support (i.e. reduce the set of chips
1136 and thereby the code size); consult ath_hal(4) for details.
1139 __FreeBSD_version 800054 adds memory barriers to
1140 <machine/atomic.h>, new interfaces to ifnet to facilitate
1141 multiple hardware transmit queues for cards that support
1142 them, and a lock-less ring-buffer implementation to
1143 enable drivers to more efficiently manage queueing of
1147 A new version of ZFS (version 13) has been merged to -HEAD.
1148 This version has zpool attribute "listsnapshots" off by
1149 default, which means "zfs list" does not show snapshots,
1150 and is the same as Solaris behavior.
1153 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1156 The uhci, ohci, ehci and slhci USB Host controller drivers have
1157 been put into separate modules. If you load the usb module
1158 separately through loader.conf you will need to load the
1159 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1160 controller add the following to loader.conf:
1166 The ABI used by the PMC toolset has changed. Please keep
1167 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1171 atapci kernel module now includes only generic PCI ATA
1172 driver. AHCI driver moved to ataahci kernel module.
1173 All vendor-specific code moved into separate kernel modules:
1174 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1175 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1176 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1177 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1180 The TTY subsystem of the kernel has been replaced by a new
1181 implementation, which provides better scalability and an
1182 improved driver model. Most common drivers have been migrated to
1183 the new TTY subsystem, while others have not. The following
1184 drivers have not yet been ported to the new TTY layer:
1187 cy, digi, rc, rp, sio
1193 ng_h4, ng_tty, ppp, sl, snp
1195 Adding these drivers to your kernel configuration file shall
1196 cause compilation to fail.
1199 ntpd has been upgraded to 4.2.4p5.
1202 OpenSSH has been upgraded to 5.1p1.
1204 For many years, FreeBSD's version of OpenSSH preferred DSA
1205 over RSA for host and user authentication keys. With this
1206 upgrade, we've switched to the vendor's default of RSA over
1207 DSA. This may cause upgraded clients to warn about unknown
1208 host keys even for previously known hosts. Users should
1209 follow the usual procedure for verifying host keys before
1210 accepting the RSA key.
1212 This can be circumvented by setting the "HostKeyAlgorithms"
1213 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1216 Please note that the sequence of keys offered for
1217 authentication has been changed as well. You may want to
1218 specify IdentityFile in a different order to revert this
1222 The sio(4) driver has been removed from the i386 and amd64
1223 kernel configuration files. This means uart(4) is now the
1224 default serial port driver on those platforms as well.
1226 To prevent collisions with the sio(4) driver, the uart(4) driver
1227 uses different names for its device nodes. This means the
1228 onboard serial port will now most likely be called "ttyu0"
1229 instead of "ttyd0". You may need to reconfigure applications to
1230 use the new device names.
1232 When using the serial port as a boot console, be sure to update
1233 /boot/device.hints and /etc/ttys before booting the new kernel.
1234 If you forget to do so, you can still manually specify the hints
1235 at the loader prompt:
1237 set hint.uart.0.at="isa"
1238 set hint.uart.0.port="0x3F8"
1239 set hint.uart.0.flags="0x10"
1240 set hint.uart.0.irq="4"
1244 The gpt(8) utility has been removed. Use gpart(8) to partition
1248 The version that Linuxulator emulates was changed from 2.4.2
1249 to 2.6.16. If you experience any problems with Linux binaries
1250 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1251 if it fixes the problem contact emulation mailing list.
1254 ISDN4BSD (I4B) was removed from the src tree. You may need to
1255 update a your kernel configuration and remove relevant entries.
1258 I have checked in code to support multiple routing tables.
1259 See the man pages setfib(1) and setfib(2).
1260 This is a hopefully backwards compatible version,
1261 but to make use of it you need to compile your kernel
1262 with options ROUTETABLES=2 (or more up to 16).
1265 The 802.11 wireless support was redone to enable multi-bss
1266 operation on devices that are capable. The underlying device
1267 is no longer used directly but instead wlanX devices are
1268 cloned with ifconfig. This requires changes to rc.conf files.
1269 For example, change:
1270 ifconfig_ath0="WPA DHCP"
1273 ifconfig_wlan0="WPA DHCP"
1274 see rc.conf(5) for more details. In addition, mergemaster of
1275 /etc/rc.d is highly recommended. Simultaneous update of userland
1276 and kernel wouldn't hurt either.
1278 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1279 modules were merged into the base wlan module. All references
1280 to these modules (e.g. in kernel config files) must be removed.
1283 psm(4) has gained write(2) support in native operation level.
1284 Arbitrary commands can be written to /dev/psm%d and status can
1285 be read back from it. Therefore, an application is responsible
1286 for status validation and error recovery. It is a no-op in
1287 other operation levels.
1290 Support for KSE threading has been removed from the kernel. To
1291 run legacy applications linked against KSE libmap.conf may
1292 be used. The following libmap.conf may be used to ensure
1293 compatibility with any prior release:
1295 libpthread.so.1 libthr.so.1
1296 libpthread.so.2 libthr.so.2
1297 libkse.so.3 libthr.so.3
1300 The layout of struct vmspace has changed. This affects libkvm
1301 and any executables that link against libkvm and use the
1302 kvm_getprocs() function. In particular, but not exclusively,
1303 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1304 The effects are minimal, but it's advisable to upgrade world
1308 The latest em driver no longer has support in it for the
1309 82575 adapter, this is now moved to the igb driver. The
1310 split was done to make new features that are incompatible
1311 with older hardware easier to do.
1314 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1315 likewise the kernel option is now GEOM_LINUX_LVM.
1318 The default NFS mount mode has changed from UDP to TCP for
1319 increased reliability. If you rely on (insecurely) NFS
1320 mounting across a firewall you may need to update your
1324 Belatedly note the addition of m_collapse for compacting
1328 The fts(3) structures have been changed to use adequate
1329 integer types for their members and so to be able to cope
1330 with huge file trees. The old fts(3) ABI is preserved
1331 through symbol versioning in libc, so third-party binaries
1332 using fts(3) should still work, although they will not take
1333 advantage of the extended types. At the same time, some
1334 third-party software might fail to build after this change
1335 due to unportable assumptions made in its source code about
1336 fts(3) structure members. Such software should be fixed
1337 by its vendor or, in the worst case, in the ports tree.
1338 FreeBSD_version 800015 marks this change for the unlikely
1339 case that a portable fix is impossible.
1342 To upgrade to -current after this date, you must be running
1343 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1344 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1347 The ADAPTIVE_GIANT kernel option has been retired because its
1348 functionality is the default now.
1351 The AT keyboard emulation of sunkbd(4) has been turned on
1352 by default. In order to make the special symbols of the Sun
1353 keyboards driven by sunkbd(4) work under X these now have
1354 to be configured the same way as Sun USB keyboards driven
1355 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1357 Option "XkbLayout" "us"
1358 Option "XkbRules" "xorg"
1359 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1362 It has been decided that it is desirable to provide ABI
1363 backwards compatibility to the FreeBSD 4/5/6 versions of the
1364 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1365 broken with the introduction of PCI domain support (see the
1366 20070930 entry). Unfortunately, this required the ABI of
1367 PCIOCGETCONF to be broken again in order to be able to
1368 provide backwards compatibility to the old version of that
1369 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1370 again. As for prominent ports this affects neither pciutils
1371 nor xorg-server this time, the hal port needs to be rebuilt
1375 The misnamed kthread_create() and friends have been renamed
1376 to kproc_create() etc. Many of the callers already
1377 used kproc_start()..
1378 I will return kthread_create() and friends in a while
1379 with implementations that actually create threads, not procs.
1380 Renaming corresponds with version 800002.
1389 Avoid using make -j when upgrading. While generally safe, there are
1390 sometimes problems using -j to upgrade. If your upgrade fails with
1391 -j, please try again without -j. From time to time in the past there
1392 have been problems using -j with buildworld and/or installworld. This
1393 is especially true when upgrading between "distant" versions (eg one
1394 that cross a major release boundary or several minor releases, or when
1395 several months have passed on the -current branch).
1397 Sometimes, obscure build problems are the result of environment
1398 poisoning. This can happen because the make utility reads its
1399 environment when searching for values for global variables. To run
1400 your build attempts in an "environmental clean room", prefix all make
1401 commands with 'env -i '. See the env(1) manual page for more details.
1403 When upgrading from one major version to another it is generally best
1404 to upgrade to the latest code in the currently installed branch first,
1405 then do an upgrade to the new branch. This is the best-tested upgrade
1406 path, and has the highest probability of being successful. Please try
1407 this approach before reporting problems with a major version upgrade.
1411 When upgrading the boot ZFS pool to a new version, always follow
1414 1.) recompile and reinstall the ZFS boot loader and boot block
1415 (this is part of "make buildworld" and "make installworld")
1417 2.) update the ZFS boot block on your boot drive
1419 The following example updates the ZFS boot block on the first
1420 partition (freebsd-boot) of a GPT partitioned drive ad0:
1421 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1423 Non-boot pools do not need these updates.
1427 If you are updating from a prior version of FreeBSD (even one just
1428 a few days old), you should follow this procedure. It is the most
1429 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1431 make kernel-toolchain
1432 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1433 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1435 To test a kernel once
1436 ---------------------
1437 If you just want to boot a kernel once (because you are not sure
1438 if it works, or if you want to boot a known bad kernel to provide
1439 debugging information) run
1440 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1441 nextboot -k testkernel
1443 To just build a kernel when you know that it won't mess you up
1444 --------------------------------------------------------------
1445 This assumes you are already running a CURRENT system. Replace
1446 ${arch} with the architecture of your machine (e.g. "i386",
1447 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1449 cd src/sys/${arch}/conf
1450 config KERNEL_NAME_HERE
1451 cd ../compile/KERNEL_NAME_HERE
1456 If this fails, go to the "To build a kernel" section.
1458 To rebuild everything and install it on the current system.
1459 -----------------------------------------------------------
1460 # Note: sometimes if you are running current you gotta do more than
1461 # is listed here if you are upgrading from a really old current.
1463 <make sure you have good level 0 dumps>
1465 make kernel KERNCONF=YOUR_KERNEL_HERE
1467 <reboot in single user> [3]
1475 To cross-install current onto a separate partition
1476 --------------------------------------------------
1477 # In this approach we use a separate partition to hold
1478 # current's root, 'usr', and 'var' directories. A partition
1479 # holding "/", "/usr" and "/var" should be about 2GB in
1482 <make sure you have good level 0 dumps>
1485 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1486 <maybe newfs current's root partition>
1487 <mount current's root partition on directory ${CURRENT_ROOT}>
1488 make installworld DESTDIR=${CURRENT_ROOT}
1489 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1490 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1491 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1492 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1493 <reboot into current>
1494 <do a "native" rebuild/install as described in the previous section>
1495 <maybe install compatibility libraries from ports/misc/compat*>
1499 To upgrade in-place from 8.x-stable to current
1500 ----------------------------------------------
1501 <make sure you have good level 0 dumps>
1503 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1505 <reboot in single user> [3]
1512 Make sure that you've read the UPDATING file to understand the
1513 tweaks to various things you need. At this point in the life
1514 cycle of current, things change often and you are on your own
1515 to cope. The defaults can also change, so please read ALL of
1516 the UPDATING entries.
1518 Also, if you are tracking -current, you must be subscribed to
1519 freebsd-current@freebsd.org. Make sure that before you update
1520 your sources that you have read and understood all the recent
1521 messages there. If in doubt, please track -stable which has
1522 much fewer pitfalls.
1524 [1] If you have third party modules, such as vmware, you
1525 should disable them at this point so they don't crash your
1528 [3] From the bootblocks, boot -s, and then do
1533 adjkerntz -i # if CMOS is wall time
1534 Also, when doing a major release upgrade, it is required that
1535 you boot into single user mode to do the installworld.
1537 [4] Note: This step is non-optional. Failure to do this step
1538 can result in a significant reduction in the functionality of the
1539 system. Attempting to do it by hand is not recommended and those
1540 that pursue this avenue should read this file carefully, as well
1541 as the archives of freebsd-current and freebsd-hackers mailing lists
1542 for potential gotchas. The -U option is also useful to consider.
1543 See mergemaster(8) for more information.
1545 [5] Usually this step is a noop. However, from time to time
1546 you may need to do this if you get unknown user in the following
1547 step. It never hurts to do it all the time. You may need to
1548 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1549 install) after the buildworld before this step if you last updated
1550 from current before 20020224 or from -stable before 20020408.
1552 [6] This only deletes old files and directories. Old libraries
1553 can be deleted by "make delete-old-libs", but you have to make
1554 sure that no program is using those libraries anymore.
1556 [8] In order to have a kernel that can run the 4.x binaries needed to
1557 do an installworld, you must include the COMPAT_FREEBSD4 option in
1558 your kernel. Failure to do so may leave you with a system that is
1559 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1560 required to run the 5.x binaries on more recent kernels. And so on
1561 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1563 Make sure that you merge any new devices from GENERIC since the
1564 last time you updated your kernel config file.
1566 [9] When checking out sources, you must include the -P flag to have
1567 cvs prune empty directories.
1569 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1570 "?=" instead of the "=" assignment operator, so that buildworld can
1571 override the CPUTYPE if it needs to.
1573 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1574 not on the command line, or in /etc/make.conf. buildworld will
1575 warn if it is improperly defined.
1578 This file contains a list, in reverse chronological order, of major
1579 breakages in tracking -current. It is not guaranteed to be a complete
1580 list of such breakages, and only contains entries since October 10, 2007.
1581 If you need to see UPDATING entries from before that date, you will need
1582 to fetch an UPDATING file from an older FreeBSD release.
1584 Copyright information:
1586 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1588 Redistribution, publication, translation and use, with or without
1589 modification, in full or in part, in any form or format of this
1590 document are permitted without further permission from the author.
1592 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1593 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1594 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1595 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1596 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1597 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1598 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1599 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1600 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1601 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1602 POSSIBILITY OF SUCH DAMAGE.
1604 Contact Warner Losh if you have any questions about your use of