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 http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/makeworld.html
11 Items affecting the ports and packages system can be found in
12 /usr/ports/UPDATING. Please read that file before running portupgrade.
14 20150319: p11 FreeBSD-SA-15:06.openssl
15 Fix multiple vulnerabilities in OpenSSL. [SA-15:06]
17 20150225: p10 FreeBSD-SA-15:04.igmp
20 FreeBSD-EN-15:02.openssl
21 FreeBSD-EN-15:03.freebsd-update
23 Fix integer overflow in IGMP protocol. [SA-15:04]
25 Fix BIND remote denial of service vulnerability. [SA-15:05]
27 Fix vt(4) crash with improper ioctl parameters. [EN-15:01]
29 Updated base system OpenSSL to 0.9.8zd. [EN-15:02]
31 Fix freebsd-update libraries update ordering issue. [EN-15:03]
33 20150127: p9 FreeBSD-SA-15:02.kmem
36 Fix SCTP SCTP_SS_VALUE kernel memory corruption and disclosure
37 vulnerability. [SA-15:02]
39 Fix SCTP stream reset vulnerability. [SA-15:03]
41 20150114: p8 FreeBSD-SA-15:01.openssl
42 Fix multiple vulnerabilities in OpenSSL. [SA-15:01]
44 20141223: p7 FreeBSD-SA-14:31.ntp
45 FreeBSD-EN-14:13.freebsd-update
47 Fix multiple vulnerabilities in NTP suite. [SA-14:31]
48 Fix directory deletion issue in freebsd-update. [EN-14:13]
50 20141210: p6 FreeBSD-SA-14:28.file
53 Fix multiple vulnerabilities in file(1) and libmagic(3).
56 Fix BIND remote denial of service vulnerability. [SA-14:29]
58 20141104: p5 FreeBSD-SA-14:25.setlogin
62 Fix kernel stack disclosure in setlogin(2) / getlogin(2).
65 Fix remote command execution in ftp(1). [SA-14:26]
67 Fix NFSv4 and ZFS cache consistency issue. [EN-14:12]
69 20141022: p4 FreeBSD-EN-14:10.tzdata
70 FreeBSD-EN-14:11.crypt
72 Time zone data file update. [EN-14:10]
74 Change crypt(3) default hashing algorithm back to DES. [EN-14:11]
76 20141021: p3 FreeBSD-SA-14:20.rtsold
77 FreeBSD-SA-14:21.routed
78 FreeBSD-SA-14:22.namei
79 FreeBSD-SA-14:23.openssl
81 Fix rtsold(8) remote buffer overflow vulnerability. [SA-14:20]
83 Fix routed(8) remote denial of service vulnerability. [SA-14:21]
85 Fix memory leak in sandboxed namei lookup. [SA-14:22]
87 Fix OpenSSL multiple vulnerabilities. [SA-14:23]
89 20140916: p2 FreeBSD-SA-14:19.tcp
90 Fix Denial of Service in TCP packet processing. [SA-14:19]
92 20140909: p1 FreeBSD-SA-14:18.openssl
93 Fix OpenSSL multiple vulnerabilities. [SA-14:18]
99 On i386 and amd64 systems, the onifconsole flag is now set by default
100 in /etc/ttys for ttyu0. This causes ttyu0 to be automatically enabled
101 as a login TTY if it is set in the bootloader as an active kernel
102 console. No changes in behavior should result otherwise. To revert to
103 the previous behavior, set ttyu0 to "off" in /etc/ttys.
106 Clang and llvm have been upgraded to 3.4.1 release.
109 Clang and llvm have been upgraded to 3.4 release.
112 The nve(4) driver for NVIDIA nForce MCP Ethernet adapters has
113 been deprecated and will not be part of FreeBSD 11.0 and later
114 releases. If you use this driver, please consider switching to
115 the nfe(4) driver instead.
118 The behavior of gss_pseudo_random() for the krb5 mechanism
119 has changed, for applications requesting a longer random string
120 than produced by the underlying enctype's pseudo-random() function.
121 In particular, the random string produced from a session key of
122 enctype aes256-cts-hmac-sha1-96 or aes256-cts-hmac-sha1-96 will
123 be different at the 17th octet and later, after this change.
124 The counter used in the PRF+ construction is now encoded as a
125 big-endian integer in accordance with RFC 4402.
126 __FreeBSD_version is bumped to 902505.
132 Behavior of devfs rules path matching has been changed.
133 Pattern is now always matched against fully qualified devfs
134 path and slash characters must be explicitly matched by
135 slashes in pattern (FNM_PATHNAME). Rulesets involving devfs
136 subdirectories must be reviewed.
139 hastctl(8)'s `status' command output changed to terse one-liner format.
140 Scripts using this should switch to `list' command or be rewritten.
143 Fix a bug that allowed a tracing process (e.g. gdb) to write
144 to a memory-mapped file in the traced process's address space
145 even if neither the traced process nor the tracing process had
146 write access to that file.
149 Added ZFS TRIM support which is enabled by default. To disable
150 ZFS TRIM support set vfs.zfs.trim.enabled=0 in loader.conf.
152 Creating new ZFS pools and adding new devices to existing pools
153 first performs a full device level TRIM which can take a significant
154 amount of time. The sysctl vfs.zfs.vdev.trim_on_init can be set to 0
155 to disable this behaviour.
157 ZFS TRIM requires the underlying device support BIO_DELETE which
158 is currently provided by methods such as ATA TRIM and SCSI UNMAP
159 via CAM, which are typically supported by SSD's.
161 Stats for ZFS TRIM can be monitored by looking at the sysctl's
162 under kstat.zfs.misc.zio_trim.
165 `list' command has been added to hastctl(8). For now, it is full
166 equivalent of `status' command.
167 WARNING: in the near future the output of hastctl's status command
168 will change to more terse format. If you use `hastctl status'
169 for parsing in your scripts, switch to `hastctl list'.
172 The mergemaster command now uses the default MAKEOBJDIRPREFIX
173 rather than creating it's own in the temporary directory in
174 order allow access to bootstrapped versions of tools such as
175 install and mtree. When upgrading from version of FreeBSD where
176 the install command does not support -l, you will need to
177 install a new mergemaster command if mergemaster -p is required.
178 This can be accomplished with the command (cd src/usr.sbin/mergemaster
181 Due to the use of the new -l option to install(1) during build
182 and install, you must take care not to directly set the INSTALL
183 make variable in your /etc/make.conf, /etc/src.conf, or on the
184 command line. If you wish to use the -C flag for all installs
185 you may be able to add INSTALL+=-C to /etc/make.conf or
189 Fix a bug that allows NFS clients to issue READDIR on files.
192 The install(1) option -M has changed meaning and now takes an
193 argument that is a file or path to append logs to. In the
194 unlikely event that -M was the last option on the command line
195 and the command line contained at least two files and a target
196 directory the first file will have logs appended to it. The -M
197 option served little practical purpose in the last decade so it's
198 used expected to be extremely rare.
201 A new compression method (lz4) has been merged to. Please refer to
202 zpool-features(7) for more information.
204 Please refer to the "ZFS notes" section of this file for information
205 on upgrading boot ZFS pools.
208 The VFS KBI was changed with the merge of several nullfs
209 optimizations and fixes. All filesystem modules must be
213 With the addition of auditdistd(8), a new auditdistd user is now
214 depended on during installworld. "mergemaster -p" can be used to add
215 the user prior to installworld, as documented in the handbook.
221 A new version of ZFS (pool version 5000) has been merged to 9-STABLE.
222 Starting with this version the old system of ZFS pool versioning
223 is superseded by "feature flags". This concept enables forward
224 compatibility against certain future changes in functionality of ZFS
225 pools. The first two read-only compatible "feature flags" for ZFS
226 pools are "com.delphix:async_destroy" and "com.delphix:empty_bpobj".
227 For more information read the new zpool-features(7) manual page.
228 Please refer to the "ZFS notes" section of this file for information
229 on upgrading boot ZFS pools.
232 The commit introducing bsd.compiler.mk breaks the traditional
233 building of kernels before this point. Add -m ${SRC}/share/mk
234 (for the right value of SRC) to your command lines to work
235 around; update your useland to a point after this; or use the
236 buildkernel/installkernel top-level targets. See also 20120829.
239 The IPFIREWALL_FORWARD kernel option has been removed. Its
240 functionality now turned on by default.
243 The random(4) support for the VIA hardware random number
244 generator (`PADLOCK') is no longer enabled unconditionally.
245 Add the PADLOCK_RNG option in the custom kernel config if
246 needed. The GENERIC kernels on i386 and amd64 do include the
247 option, so the change only affects the custom kernel
251 The amd64 kernel now uses xsetbv, xrstor instructions. To compile with
252 the traditional method, you must update your system with an installworld
253 before the kernel will build. The documented make buildkernel/installkernel
254 interfaces (coupled with fresh make kernel-toolchain) continue to work.
257 The sparc64 ZFS loader has been changed to no longer try to auto-
258 detect ZFS providers based on diskN aliases but now requires these
259 to be explicitly listed in the OFW boot-device environment variable.
262 Now unix domain sockets behave "as expected" on nullfs(5). Previously
263 nullfs(5) did not pass through all behaviours to the underlying layer,
264 as a result if we bound to a socket on the lower layer we could connect
265 only to the lower path; if we bound to the upper layer we could connect
266 only to the upper path. The new behavior is one can connect to both the
267 lower and the upper paths regardless what layer path one binds to.
270 The acpi_wmi(4) status device /dev/wmistat has been renamed to
274 A new VOP_ADVISE() was added to support posix_fadvise(2). All
275 filesystem modules must be recompiled.
278 The interface of the VOP_VPTOCNP(9) changed, now the returned
279 vnode shall be referenced, previously it was required to be
280 only held. All in-tree filesystems are converted.
286 The broken amd(4) driver has been replaced with esp(4) in the amd64,
287 i386 and pc98 GENERIC kernel configuration files.
290 This commit modifies vfs_register() so that it uses a hash
291 calculation to set vfc_typenum, which is enabled by default.
292 The first time a system is booted after this change, the
293 vfc_typenum values will change for all file systems. The
294 main effect of this is a change to the NFS server file handles
295 for file systems that use vfc_typenum in their fsid, such as ZFS.
296 It will, however, prevent vfc_typenum from changing when file
297 systems are loaded in a different order for subsequent reboots.
298 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf
299 until you are ready to remount all NFS clients after a reboot.
302 Bump the shared library version numbers for libraries that
303 do not use symbol versioning, have changed the ABI compared
304 to stable/8 and which shared library version was not bumped.
305 Done as part of 9.0-RELEASE cycle.
308 During the merge of Capsicum features, the fget(9) KPI was modified.
309 This may require the rebuilding of out-of-tree device drivers --
310 issues have been reported specifically with the nVidia device driver.
311 __FreeBSD_version is bumped to 900041.
313 Also, there is a period between 20110811 and 20110814 where the
314 special devices /dev/{stdin,stdout,stderr} did not work correctly.
315 Building world from a kernel during that window may not work.
318 The packet filter (pf) code has been updated to OpenBSD 4.5.
319 You need to update userland tools to be in sync with kernel.
320 This update breaks backward compatibility with earlier pfsync(4)
321 versions. Care must be taken when updating redundant firewall setups.
324 The following sysctls and tunables are retired on x86 platforms:
326 machdep.hlt_logical_cpus
327 The following sysctl is retired:
328 machdep.hyperthreading_allowed
329 The sysctls were supposed to provide a way to dynamically offline and
330 online selected CPUs on x86 platforms, but the implementation has not
331 been reliable especially with SCHED_ULE scheduler.
332 machdep.hyperthreading_allowed tunable is still available to ignore
333 hyperthreading CPUs at OS level.
334 Individual CPUs can be disabled using hint.lapic.X.disabled tunable,
335 where X is an APIC ID of a CPU. Be advised, though, that disabling
336 CPUs in non-uniform fashion will result in non-uniform topology and
337 may lead to sub-optimal system performance with SCHED_ULE, which is
341 cpumask_t type is retired and cpuset_t is used in order to describe
345 Changes to ifconfig(8) for dynamic address family detection mandate
346 that you are running a kernel of 20110525 or later. Make sure to
347 follow the update procedure to boot a new kernel before installing
351 Support for sun4v architecture is officially dropped
354 Several KPI breaking changes have been committed to the mii(4) layer,
355 the PHY drivers and consequently some Ethernet drivers using mii(4).
356 This means that miibus.ko and the modules of the affected Ethernet
357 drivers need to be recompiled.
359 Note to kernel developers: Given that the OUI bit reversion problem
360 was fixed as part of these changes all mii(4) commits related to OUIs,
361 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific
362 handling, no longer can be merged verbatim to stable/8 and previous
366 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci'
367 into their kernel configurations along with 'device pci'.
370 The default NFS client is now the new NFS client, so fstype "newnfs"
371 is now "nfs" and the regular/old NFS client is now fstype "oldnfs".
372 Although mounts via fstype "nfs" will usually work without userland
373 changes, it is recommended that the mount(8) and mount_nfs(8)
374 commands be rebuilt from sources and that a link to mount_nfs called
375 mount_oldnfs be created. The new client is compiled into the
376 kernel with "options NFSCL" and this is needed for diskless root
377 file systems. The GENERIC kernel configs have been changed to use
378 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER.
379 To use the regular/old client, you can "mount -t oldnfs ...". For
380 a diskless root file system, you must also include a line like:
382 vfs.root.mountfrom="oldnfs:"
384 in the boot/loader.conf on the root fs on the NFS server to make
385 a diskless root fs use the old client.
388 The GENERIC kernels for all architectures now default to the new
389 CAM-based ATA stack. It means that all legacy ATA drivers were
390 removed and replaced by respective CAM drivers. If you are using
391 ATA device names in /etc/fstab or other places, make sure to update
392 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY,
393 where 'Y's are the sequential numbers starting from zero for each type
394 in order of detection, unless configured otherwise with tunables,
395 see cam(4)). There will be symbolic links created in /dev/ to map
396 old adX devices to the respective adaY. They should provide basic
397 compatibility for file systems mounting in most cases, but they do
398 not support old user-level APIs and do not have respective providers
399 in GEOM. Consider using updated management tools with new device names.
401 It is possible to load devices ahci, ata, siis and mvs as modules,
402 but option ATA_CAM should remain in kernel configuration to make ata
403 module work as CAM driver supporting legacy ATA controllers. Device ata
404 still can be used in modular fashion (atacore + ...). Modules atadisk
405 and atapi* are not used and won't affect operation in ATA_CAM mode.
406 Note that to use CAM-based ATA kernel should include CAM devices
407 scbus, pass, da (or explicitly ada), cd and optionally others. All of
408 them are parts of the cam module.
410 ataraid(4) functionality is now supported by the RAID GEOM class.
411 To use it you can load geom_raid kernel module and use graid(8) tool
412 for management. Instead of /dev/arX device names, use /dev/raid/rX.
414 No kernel config options or code have been removed, so if a problem
415 arises, please report it and optionally revert to the old ATA stack.
416 In order to do it you can remove from the kernel config:
421 , and instead add back:
422 device atadisk # ATA disk drives
423 device ataraid # ATA RAID drives
424 device atapicd # ATAPI CDROM drives
425 device atapifd # ATAPI floppy drives
426 device atapist # ATAPI tape drives
429 The default NFS server has been changed to the new server, which
430 was referred to as the experimental server. If you need to switch
431 back to the old NFS server, you must now put the "-o" option on
432 both the mountd and nfsd commands. This can be done using the
433 mountd_flags and nfs_server_flags rc.conf variables until an
434 update to the rc scripts is committed, which is coming soon.
437 The GNU Objective-C runtime library (libobjc), and other Objective-C
438 related components have been removed from the base system. If you
439 require an Objective-C library, please use one of the available ports.
442 ath(4) has been split into bus- and device- modules. if_ath contains
443 the HAL, the TX rate control and the network device code. if_ath_pci
444 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb
445 contains the AHB glue. Users need to load both if_ath_pci and if_ath
446 in order to use ath on everything else.
448 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only
449 need to load if_ath and if_ath_pci for ath(4) operation.
452 As part of the replacement of sysinstall, the process of building
453 release media has changed significantly. For details, please re-read
454 release(7), which has been updated to reflect the new build process.
457 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This
458 is the last available version under GPLv2. It brings a number of new
459 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE
460 4.1 and SSE 4.2), better support for powerpc64, a number of new
461 directives, and lots of other small improvements. See the ChangeLog
462 file in contrib/binutils for the full details.
465 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868
466 compliant, and will now use half of hash for authentication.
467 This will break interoperability with all stacks (including all
468 actual FreeBSD versions) who implement
469 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for
471 The only workaround with such peers is to use another HMAC
472 algorithm for IPsec ("phase 2") authentication.
475 Remove the uio_yield prototype and symbol. This function has
476 been misnamed since it was introduced and should not be
477 globally exposed with this name. The equivalent functionality
478 is now available using kern_yield(curthread->td_user_pri).
479 The function remains undocumented.
482 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers,
483 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking
484 for scalar sysctls is defined but disabled. Code that needs
485 UQUAD to pass the type checking that must compile on older
486 systems where the define is not present can check against
487 __FreeBSD_version >= 900030.
489 The system dialog(1) has been replaced with a new version previously
490 in ports as devel/cdialog. dialog(1) is mostly command-line compatible
491 with the previous version, but the libdialog associated with it has
492 a largely incompatible API. As such, the original version of libdialog
493 will be kept temporarily as libodialog, until its base system consumers
494 are replaced or updated. Bump __FreeBSD_version to 900030.
497 If you are trying to run make universe on a -stable system, and you get
498 the following warning:
499 "Makefile", line 356: "Target architecture for i386/conf/GENERIC
500 unknown. config(8) likely too old."
501 or something similar to it, then you must upgrade your -stable system
502 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in
503 stable/8) or build the config from the latest stable/8 branch and
504 install it on your system.
506 Prior to this date, building a current universe on 8-stable system from
507 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing
508 error in the first kernel build phase. A new config on those old
509 systems will fix that problem for older versions of -current.
512 The TCP stack has been modified to allow Khelp modules to interact with
513 it via helper hook points and store per-connection data in the TCP
514 control block. Bump __FreeBSD_version to 900029. User space tools that
515 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to
519 Generic IEEE 802.3 annex 31B full duplex flow control support has been
520 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along
521 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted
522 to take advantage of it instead of using custom implementations. This
523 means that these drivers now no longer unconditionally advertise
524 support for flow control but only do so if flow control is a selected
525 media option. This was implemented in the generic support that way in
526 order to allow flow control to be switched on and off via ifconfig(8)
527 with the PHY specific default to typically off in order to protect
528 from unwanted effects. Consequently, if you used flow control with
529 one of the above mentioned drivers you now need to explicitly enable
531 ifconfig bge0 media auto mediaopt flowcontrol
533 Along with the above mentioned changes generic support for setting
534 1000baseT master mode also has been added and brgphy(4), ciphy(4),
535 e1000phy(4) as well as ip1000phy(4) have been converted to take
536 advantage of it. This means that these drivers now no longer take the
537 link0 parameter for selecting master mode but the master media option
538 has to be used instead, for example like in the following:
539 ifconfig bge0 media 1000baseT mediaopt full-duplex,master
541 Selection of master mode now is also available with all other PHY
542 drivers supporting 1000baseT.
545 The TCP stack has received a significant update to add support for
546 modularised congestion control and generally improve the clarity of
547 congestion control decisions. Bump __FreeBSD_version to 900025. User
548 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g.
549 sockstat) need to be recompiled.
552 The man(1) utility has been replaced by a new version that no longer
553 uses /etc/manpath.config. Please consult man.conf(5) for how to
554 migrate local entries to the new format.
557 The copyright strings printed by login(1) and sshd(8) at the time of a
558 new connection have been removed to follow other operating systems and
562 A workaround for a fixed ld bug has been removed in kernel code,
563 so make sure that your system ld is built from sources after
564 revision 210245 from 2010-07-19 (r211583 if building head kernel
565 on stable/8, r211584 for stable/7; both from 2010-08-21).
566 A symptom of incorrect ld version is different addresses for
567 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules.
570 The $ipv6_prefer variable in rc.conf(5) has been split into
571 $ip6addrctl_policy and $ipv6_activate_all_interfaces.
573 The $ip6addrctl_policy is a variable to choose a pre-defined
574 address selection policy set by ip6addrctl(8). A value
575 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The
578 The $ipv6_activate_all_interfaces specifies whether IFDISABLED
579 flag (see an entry of 20090926) is set on an interface with no
580 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for
581 security reason. If you want IPv6 link-local address on all
582 interfaces by default, set this to "YES".
584 The old ipv6_prefer="YES" is equivalent to
585 ipv6_activate_all_interfaces="YES" and
586 ip6addrctl_policy="ipv6_prefer".
589 DTrace has grown support for userland tracing. Due to this, DTrace is
590 now i386 and amd64 only.
591 dtruss(1) is now installed by default on those systems and a new
592 kernel module is needed for userland tracing: fasttrap.
593 No changes to your kernel config file are necessary to enable
594 userland tracing, but you might consider adding 'STRIP=' and
595 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want
596 to have informative userland stack traces in DTrace (ustack).
599 The acpi_aiboost(4) driver has been removed in favor of the new
600 aibs(4) driver. You should update your kernel configuration file.
603 BSD grep has been imported to the base system and it is built by
604 default. It is completely BSD licensed, highly GNU-compatible, uses
605 less memory than its GNU counterpart and has a small codebase.
606 However, it is slower than its GNU counterpart, which is mostly
607 noticeable for larger searches, for smaller ones it is measurable
608 but not significant. The reason is complex, the most important factor
609 is that we lack a modern and efficient regex library and GNU
610 overcomes this by optimizing the searches internally. Future work
611 on improving the regex performance is planned, for the meantime,
612 users that need better performance, can build GNU grep instead by
613 setting the WITH_GNU_GREP knob.
616 Due to the import of powerpc64 support, all existing powerpc kernel
617 configuration files must be updated with a machine directive like this:
618 machine powerpc powerpc
620 In addition, an updated config(8) is required to build powerpc kernels
624 A new version of ZFS (version 15) has been merged to -HEAD.
625 This version uses a python library for the following subcommands:
626 zfs allow, zfs unallow, zfs groupspace, zfs userspace.
627 For full functionality of these commands the following port must
628 be installed: sysutils/py-zfs
631 'vm_page's are now hashed by physical address to an array of mutexes.
632 Currently this is only used to serialize access to hold_count. Over
633 time the page queue mutex will be peeled away. This changes the size
634 of pmap on every architecture. And requires all callers of vm_page_hold
635 and vm_page_unhold to be updated.
638 WITH_CTF can now be specified in src.conf (not recommended, there
639 are some problems with static executables), make.conf (would also
640 affect ports which do not use GNU make and do not override the
641 compile targets) or in the kernel config (via "makeoptions
643 When WITH_CTF was specified there before this was silently ignored,
644 so make sure that WITH_CTF is not used in places which could lead
645 to unwanted behavior.
648 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32
649 to allow 32-bit compatibility on non-x86 platforms. All kernel
650 configurations on amd64 and ia64 platforms using these options must
651 be modified accordingly.
654 The utmp user accounting database has been replaced with utmpx,
655 the user accounting interface standardized by POSIX.
656 Unfortunately the semantics of utmp and utmpx don't match,
657 making it practically impossible to support both interfaces.
658 The user accounting database is used by tools like finger(1),
659 last(1), talk(1), w(1) and ac(8).
661 All applications in the base system use utmpx. This means only
662 local binaries (e.g. from the ports tree) may still use these
663 utmp database files. These applications must be rebuilt to make
666 After the system has been upgraded, it is safe to remove the old
667 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*),
668 assuming their contents is of no importance anymore. Old wtmp
669 databases can only be used by last(1) and ac(8) after they have
670 been converted to the new format using wtmpcvt(1).
673 Introduce the kernel thread "deadlock resolver" (which can be enabled
674 via the DEADLKRES option, see NOTES for more details) and the
675 sleepq_type() function for sleepqueues.
678 The rc.firewall and rc.firewall6 were unified, and
679 rc.firewall6 and rc.d/ip6fw were removed.
680 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc
681 variables are obsoleted. Instead, the following new rc
682 variables are added to rc.d/ipfw:
684 firewall_client_net_ipv6, firewall_simple_iif_ipv6,
685 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6,
686 firewall_simple_onet_ipv6, firewall_trusted_ipv6
688 The meanings correspond to the relevant IPv4 variables.
694 The default terminal emulation for syscons(4) has been changed
695 from cons25 to xterm on all platforms except pc98. This means
696 that the /etc/ttys file needs to be updated to ensure correct
697 operation of applications on the console.
699 The terminal emulation style can be toggled per window by using
700 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration
701 options can be used to change the compile-time default back to
704 To prevent graphical artifacts, make sure the TERM environment
705 variable is set to match the terminal emulation that is being
706 performed by syscons(4).
709 The layout of the structure ieee80211req_scan_result has changed.
710 Applications that require wireless scan results (e.g. ifconfig(8))
711 from net80211 need to be recompiled.
713 Applications such as wpa_supplicant(8) may require a full world
714 build without using NO_CLEAN in order to get synchronized with the
718 The iwn(4) driver has been updated to support the 5000 and 5150 series.
719 There's one kernel module for each firmware. Adding "device iwnfw"
720 to the kernel configuration file means including all three firmware
721 images inside the kernel. If you want to include just the one for
722 your wireless card, use the devices iwn4965fw, iwn5000fw or
726 The rc.d/network_ipv6, IPv6 configuration script has been integrated
727 into rc.d/netif. The changes are the following:
729 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF
730 for IPv4. For aliases, $ifconfig_IF_aliasN should be used.
731 Note that both variables need the "inet6" keyword at the head.
733 Do not set $ipv6_network_interfaces manually if you do not
734 understand what you are doing. It is not needed in most cases.
736 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but
739 2. $ipv6_enable is obsolete. Use $ipv6_prefer and
740 "inet6 accept_rtadv" keyword in ifconfig(8) instead.
742 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and
743 all configured interfaces have "inet6 accept_rtadv" in the
744 $ifconfig_IF_ipv6. These are for backward compatibility.
746 3. A new variable $ipv6_prefer has been added. If NO, IPv6
747 functionality of interfaces with no corresponding
748 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag,
749 and the default address selection policy of ip6addrctl(8)
750 is the IPv4-preferred one (see rc.d/ip6addrctl for more details).
751 Note that if you want to configure IPv6 functionality on the
752 disabled interfaces after boot, first you need to clear the flag by
753 using ifconfig(8) like:
755 ifconfig em0 inet6 -ifdisabled
757 If YES, the default address selection policy is set as
760 The default value of $ipv6_prefer is NO.
762 4. If your system need to receive Router Advertisement messages,
763 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8)
764 scripts automatically invoke rtsol(8) when the interface becomes
765 UP. The Router Advertisement messages are used for SLAAC
766 (State-Less Address AutoConfiguration).
769 802.11s D3.03 support was committed. This is incompatible with the
770 previous code, which was based on D3.0.
773 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value
774 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to
775 control whether accepting Router Advertisement messages or not.
776 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and
777 a sysctl variable net.inet6.ip6.auto_linklocal is its default value.
778 The ifconfig(8) utility now supports these flags.
781 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for
782 mount(8) and -a option for df(1) to see them.
785 The old tunable hw.bus.devctl_disable has been superseded by
786 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be
787 replaced by hw.bus.devctl_queue=0. The default for this new tunable
791 Remove the option STOP_NMI. The default action is now to use NMI only
792 for KDB via the newly introduced function stop_cpus_hard() and
793 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64.
796 The stable/8 branch created in subversion. This corresponds to the
797 RELENG_8 branch in CVS.
800 Bump the shared library version numbers for all libraries that do not
801 use symbol versioning as part of the 8.0-RELEASE cycle. Bump
802 __FreeBSD_version to 800105.
805 Due to changes in the implementation of virtual network stack support,
806 all network-related kernel modules must be recompiled. As this change
807 breaks the ABI, bump __FreeBSD_version to 800104.
810 The TOE interface to the TCP syncache has been modified to remove
811 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack.
812 The cxgb driver is the only TOE consumer affected by this change, and
813 needs to be recompiled along with the kernel. As this change breaks
814 the ABI, bump __FreeBSD_version to 800103.
817 Padding has been added to struct tcpcb, sackhint and tcpstat in
818 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
819 maintaining the ABI. However, this change breaks the ABI, so bump
820 __FreeBSD_version to 800102. User space tools that rely on the size of
821 any of these structs (e.g. sockstat) need to be recompiled.
824 The NFS_LEGACYRPC option has been removed along with the old kernel
825 RPC implementation that this option selected. Kernel configurations
826 may need to be adjusted.
829 The network interface device nodes at /dev/net/<interface> have been
830 removed. All ioctl operations can be performed the normal way using
831 routing sockets. The kqueue functionality can generally be replaced
832 with routing sockets.
835 The documentation from the FreeBSD Documentation Project (Handbook,
836 FAQ, etc.) is now installed via packages by sysinstall(8) and under
837 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc.
840 The ABI of various structures related to the SYSV IPC API have been
841 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel
842 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to
846 Layout of struct vnet has changed as routing related variables were
847 moved to their own Vimage module. Modules need to be recompiled. Bump
848 __FreeBSD_version to 800099.
851 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024
852 respectively. As long as no more than 16 groups per process are used,
853 no changes should be visible. When more than 16 groups are used, old
854 binaries may fail if they call getgroups() or getgrouplist() with
855 statically sized storage. Recompiling will work around this, but
856 applications should be modified to use dynamically allocated storage
857 for group arrays as POSIX.1-2008 does not cap an implementation's
858 number of supported groups at NGROUPS_MAX+1 as previous versions did.
860 NFS and portalfs mounts may also be affected as the list of groups is
861 truncated to 16. Users of NFS who use more than 16 groups, should
862 take care that negative group permissions are not used on the exported
863 file systems as they will not be reliable unless a GSSAPI based
864 authentication method is used.
867 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This
868 option compiles in the support for adaptive spinning for lockmgrs
869 which want to enable it. The lockinit() function now accepts the flag
870 LK_ADAPTIVE in order to make the lock object subject to adaptive
871 spinning when both held in write and read mode.
874 The layout of the structure returned by IEEE80211_IOC_STA_INFO has
875 changed. User applications that use this ioctl need to be rebuilt.
878 The layout of struct thread has changed. Kernel and modules need to
882 The layout of structs ifnet, domain, protosw and vnet_net has changed.
883 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097.
886 window(1) has been removed from the base system. It can now be
887 installed from ports. The port is called misc/window.
890 The way we are storing and accessing `routing table' entries has
891 changed. Programs reading the FIB, like netstat, need to be
895 A new netisr implementation has been added for FreeBSD 8. Network
896 file system modules, such as igmp, ipdivert, and others, should be
898 Bump __FreeBSD_version to 800096.
901 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no
905 Add VOP_ACCESSX(9). File system modules need to be rebuilt.
906 Bump __FreeBSD_version to 800094.
909 Add mnt_xflag field to 'struct mount'. File system modules need to be
911 Bump __FreeBSD_version to 800093.
914 The compiling option ADAPTIVE_SX has been retired while it has been
915 introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
916 The KPI for sx_init_flags() changes as accepting flags:
917 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has
918 been introduced in order to handle the reversed logic.
919 Bump __FreeBSD_version to 800092.
922 Add support for hierarchical jails. Remove global securelevel.
923 Bump __FreeBSD_version to 800091.
926 The layout of struct vnet_net has changed, therefore modules
928 Bump __FreeBSD_version to 800090.
931 The newly imported zic(8) produces a new format in the output. Please
932 run tzsetup(8) to install the newly created data to /etc/localtime.
935 The sysctl tree for the usb stack has renamed from hw.usb2.* to
936 hw.usb.* and is now consistent again with previous releases.
939 802.11 monitor mode support was revised and driver api's were changed.
940 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead
941 of DLT_IEEE802_11. No user-visible data structures were changed but
942 applications that use DLT_IEEE802_11 may require changes.
943 Bump __FreeBSD_version to 800088.
946 The layout of the following structs has changed: sysctl_oid,
947 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
948 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
949 panics may be experienced. World rebuild is required for
950 correctly checking networking state from userland.
951 Bump __FreeBSD_version to 800085.
954 MLDv2 and Source-Specific Multicast (SSM) have been merged
955 to the IPv6 stack. VIMAGE hooks are in but not yet used.
956 The implementation of SSM within FreeBSD's IPv6 stack closely
957 follows the IPv4 implementation.
959 For kernel developers:
961 * The most important changes are that the ip6_output() and
962 ip6_input() paths no longer take the IN6_MULTI_LOCK,
963 and this lock has been downgraded to a non-recursive mutex.
965 * As with the changes to the IPv4 stack to support SSM, filtering
966 of inbound multicast traffic must now be performed by transport
967 protocols within the IPv6 stack. This does not apply to TCP and
968 SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
970 * The KPIs used by IPv6 multicast are similar to those used by
971 the IPv4 stack, with the following differences:
972 * im6o_mc_filter() is analogous to imo_multicast_filter().
973 * The legacy KAME entry points in6_joingroup and in6_leavegroup()
974 are shimmed to in6_mc_join() and in6_mc_leave() respectively.
975 * IN6_LOOKUP_MULTI() has been deprecated and removed.
976 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
977 for MLDv1 have an additional 'timer' argument which is used to
978 jitter the initial membership report for the solicited-node
979 multicast membership on-link.
980 * This is not strictly needed for MLDv2, which already jitters
981 its report transmissions. However, the 'timer' argument is
982 preserved in case MLDv1 is active on the interface.
984 * The KAME linked-list based IPv6 membership implementation has
985 been refactored to use a vector similar to that used by the IPv4
987 Code which maintains a list of its own multicast memberships
988 internally, e.g. carp, has been updated to reflect the new
991 * There is a known Lock Order Reversal (LOR) due to in6_setscope()
992 acquiring the IF_AFDATA_LOCK and being called within ip6_output().
993 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
994 implementation constraint which needs to be addressed in HEAD.
996 For application developers:
998 * The changes are broadly similar to those made for the IPv4
1001 * The use of IPv4 and IPv6 multicast socket options on the same
1002 socket, using mapped addresses, HAS NOT been tested or supported.
1004 * There are a number of issues with the implementation of various
1005 IPv6 multicast APIs which need to be resolved in the API surface
1006 before the implementation is fully compatible with KAME userland
1007 use, and these are mostly to do with interface index treatment.
1009 * The literature available discusses the use of either the delta / ASM
1010 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
1011 using setsourcefilter(3)/getsourcefilter(3). For more information
1012 please refer to RFC 3768, 'Socket Interface Extensions for
1013 Multicast Source Filters'.
1015 * Applications which use the published RFC 3678 APIs should be fine.
1017 For systems administrators:
1019 * The mtest(8) utility has been refactored to support IPv6, in
1020 addition to IPv4. Interface addresses are no longer accepted
1021 as arguments, their names must be used instead. The utility
1022 will map the interface name to its first IPv4 address as
1023 returned by getifaddrs(3).
1025 * The ifmcstat(8) utility has also been updated to print the MLDv2
1026 endpoint state and source filter lists via sysctl(3).
1028 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
1029 loopback of IPv6 multicast datagrams by default; it defaults to 1
1030 to preserve the existing behaviour. Disabling multicast loopback is
1031 recommended for optimal system performance.
1033 * The IPv6 MROUTING code has been changed to examine this sysctl
1034 instead of attempting to perform a group lookup before looping
1035 back forwarded datagrams.
1037 Bump __FreeBSD_version to 800084.
1040 Implement low-level Bluetooth HCI API.
1041 Bump __FreeBSD_version to 800083.
1044 The layout of struct malloc_type, used by modules to register new
1045 memory allocation types, has changed. Most modules will need to
1046 be rebuilt or panics may be experienced.
1047 Bump __FreeBSD_version to 800081.
1050 Anticipate overflowing inp_flags - add inp_flags2.
1051 This changes most offsets in inpcb, so checking v4 connection
1052 state will require a world rebuild.
1053 Bump __FreeBSD_version to 800080.
1056 Add an llentry to struct route and struct route_in6. Modules
1057 embedding a struct route will need to be recompiled.
1058 Bump __FreeBSD_version to 800079.
1061 The size of rt_metrics_lite and by extension rtentry has changed.
1062 Networking administration apps will need to be recompiled.
1063 The route command now supports show as an alias for get, weighting
1064 of routes, sticky and nostick flags to alter the behavior of stateful
1066 Bump __FreeBSD_version to 800078.
1069 Do not use Giant for kbdmux(4) locking. This is wrong and
1070 apparently causing more problems than it solves. This will
1071 re-open the issue where interrupt handlers may race with
1072 kbdmux(4) in polling mode. Typical symptoms include (but
1073 not limited to) duplicated and/or missing characters when
1074 low level console functions (such as gets) are used while
1075 interrupts are enabled (for example geli password prompt,
1076 mountroot prompt etc.). Disabling kbdmux(4) may help.
1079 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
1080 kernel modules referencing any of the above need to be recompiled.
1081 Bump __FreeBSD_version to 800075.
1084 GEOM_PART has become the default partition slicer for storage devices,
1085 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
1086 introduces some changes:
1088 MSDOS/EBR: the devices created from MSDOS extended partition entries
1089 (EBR) can be named differently than with GEOM_MBR and are now symlinks
1090 to devices with offset-based names. fstabs may need to be modified.
1092 BSD: the "geometry does not match label" warning is harmless in most
1093 cases but it points to problems in file system misalignment with
1094 disk geometry. The "c" partition is now implicit, covers the whole
1095 top-level drive and cannot be (mis)used by users.
1097 General: Kernel dumps are now not allowed to be written to devices
1098 whose partition types indicate they are meant to be used for file
1099 systems (or, in case of MSDOS partitions, as something else than
1102 Most of these changes date approximately from 200812.
1105 The uscanner(4) driver has been removed from the kernel. This follows
1106 Linux removing theirs in 2.6 and making libusb the default interface
1107 (supported by sane).
1110 The multicast forwarding code has been cleaned up. netstat(1)
1111 only relies on KVM now for printing bandwidth upcall meters.
1112 The IPv4 and IPv6 modules are split into ip_mroute_mod and
1113 ip6_mroute_mod respectively. The config(5) options for statically
1114 compiling this code remain the same, i.e. 'options MROUTING'.
1117 Support for the IFF_NEEDSGIANT network interface flag has been
1118 removed, which means that non-MPSAFE network device drivers are no
1119 longer supported. In particular, if_ar, if_sr, and network device
1120 drivers from the old (legacy) USB stack can no longer be built or
1124 POSIX.1 Native Language Support (NLS) has been enabled in libc and
1125 a bunch of new language catalog files have also been added.
1126 This means that some common libc messages are now localized and
1127 they depend on the LC_MESSAGES environmental variable.
1130 The k8temp(4) driver has been renamed to amdtemp(4) since
1131 support for Family 10 and Family 11 CPU families was added.
1134 IGMPv3 and Source-Specific Multicast (SSM) have been merged
1135 to the IPv4 stack. VIMAGE hooks are in but not yet used.
1137 For kernel developers, the most important changes are that the
1138 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
1139 and this lock has been downgraded to a non-recursive mutex.
1141 Transport protocols (UDP, Raw IP) are now responsible for filtering
1142 inbound multicast traffic according to group membership and source
1143 filters. The imo_multicast_filter() KPI exists for this purpose.
1144 Transports which do not use multicast (SCTP, TCP) already reject
1145 multicast by default. Forwarding and receive performance may improve
1146 as a mutex acquisition is no longer needed in the ip_input()
1147 low-level input path. in_addmulti() and in_delmulti() are shimmed
1148 to new KPIs which exist to support SSM in-kernel.
1150 For application developers, it is recommended that loopback of
1151 multicast datagrams be disabled for best performance, as this
1152 will still cause the lock to be taken for each looped-back
1153 datagram transmission. The net.inet.ip.mcast.loop sysctl may
1154 be tuned to 0 to disable loopback by default; it defaults to 1
1155 to preserve the existing behaviour.
1157 For systems administrators, to obtain best performance with
1158 multicast reception and multiple groups, it is always recommended
1159 that a card with a suitably precise hash filter is used. Hash
1160 collisions will still result in the lock being taken within the
1161 transport protocol input path to check group membership.
1163 If deploying FreeBSD in an environment with IGMP snooping switches,
1164 it is recommended that the net.inet.igmp.sendlocal sysctl remain
1165 enabled; this forces 224.0.0.0/24 group membership to be announced
1168 The size of 'struct igmpstat' has changed; netstat needs to be
1169 recompiled to reflect this.
1170 Bump __FreeBSD_version to 800070.
1173 libusb20.so.1 is now installed as libusb.so.1 and the ports system
1174 updated to use it. This requires a buildworld/installworld in order to
1175 update the library and dependencies (usbconfig, etc). Its advisable to
1176 rebuild all ports which uses libusb. More specific directions are given
1177 in the ports collection UPDATING file. Any /etc/libmap.conf entries for
1178 libusb are no longer required and can be removed.
1181 A workaround is committed to allow the creation of System V shared
1182 memory segment of size > 2 GB on the 64-bit architectures.
1183 Due to a limitation of the existing ABI, the shm_segsz member
1184 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
1185 wrong for large segments. Note that limits must be explicitly
1186 raised to allow such segments to be created.
1189 The layout of struct ifnet has changed, requiring a rebuild of all
1190 network device driver modules.
1193 The /dev handling for the new USB stack has changed, a
1194 buildworld/installworld is required for libusb20.
1197 The new USB2 stack has now been permanently moved in and all kernel and
1198 module names reverted to their previous values (eg, usb, ehci, ohci,
1199 ums, ...). The old usb stack can be compiled in by prefixing the name
1200 with the letter 'o', the old usb modules have been removed.
1201 Updating entry 20090216 for xorg and 20090215 for libmap may still
1205 The rc.conf(5) option if_up_delay has been renamed to
1206 defaultroute_delay to better reflect its purpose. If you have
1207 customized this setting in /etc/rc.conf you need to update it to
1211 xorg 7.4 wants to configure its input devices via hald which does not
1212 yet work with USB2. If the keyboard/mouse does not work in xorg then
1214 Option "AllowEmptyInput" "off"
1215 to your ServerLayout section. This will cause X to use the configured
1216 kbd and mouse sections from your xorg.conf.
1219 The GENERIC kernels for all architectures now default to the new USB2
1220 stack. No kernel config options or code have been removed so if a
1221 problem arises please report it and optionally revert to the old USB
1222 stack. If you are loading USB kernel modules or have a custom kernel
1223 that includes GENERIC then ensure that usb names are also changed over,
1224 eg uftdi -> usb2_serial_ftdi.
1226 Older programs linked against the ports libusb 0.1 need to be
1227 redirected to the new stack's libusb20. /etc/libmap.conf can
1229 # Map old usb library to new one for usb2 stack
1230 libusb-0.1.so.8 libusb20.so.1
1233 All USB ethernet devices now attach as interfaces under the name ueN
1234 (eg. ue0). This is to provide a predictable name as vendors often
1235 change usb chipsets in a product without notice.
1238 The ichsmb(4) driver has been changed to require SMBus slave
1239 addresses be left-justified (xxxxxxx0b) rather than right-justified.
1240 All of the other SMBus controller drivers require left-justified
1241 slave addresses, so this change makes all the drivers provide the
1245 INET6 statistics (struct ip6stat) was updated.
1246 netstat(1) needs to be recompiled.
1249 NTFS has been removed from GENERIC kernel on amd64 to match
1250 GENERIC on i386. Should not cause any issues since mount_ntfs(8)
1251 will load ntfs.ko module automatically when NTFS support is
1252 actually needed, unless ntfs.ko is not installed or security
1253 level prohibits loading kernel modules. If either is the case,
1254 "options NTFS" has to be added into kernel config.
1257 TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
1258 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
1259 800061. User space tools that rely on the size of struct tcpcb in
1260 tcp_var.h (e.g. sockstat) need to be recompiled.
1263 ng_tty(4) module updated to match the new TTY subsystem.
1264 Due to API change, user-level applications must be updated.
1265 New API support added to mpd5 CVS and expected to be present
1266 in next mpd5.3 release.
1269 With __FreeBSD_version 800060 the makefs tool is part of
1270 the base system (it was a port).
1273 The afdata and ifnet locks have been changed from mutexes to
1274 rwlocks, network modules will need to be re-compiled.
1277 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
1278 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
1279 The new code reduced struct rtentry{} by 16 bytes on 32-bit
1280 architecture and 40 bytes on 64-bit architecture. The userland
1281 applications "arp" and "ndp" have been updated accordingly.
1282 The output from "netstat -r" shows only routing entries and
1283 none of the L2 information.
1286 __FreeBSD_version 800057 marks the switchover from the
1287 binary ath hal to source code. Users must add the line:
1289 options AH_SUPPORT_AR5416
1291 to their kernel config files when specifying:
1295 The ath_hal module no longer exists; the code is now compiled
1296 together with the driver in the ath module. It is now
1297 possible to tailor chip support (i.e. reduce the set of chips
1298 and thereby the code size); consult ath_hal(4) for details.
1301 __FreeBSD_version 800054 adds memory barriers to
1302 <machine/atomic.h>, new interfaces to ifnet to facilitate
1303 multiple hardware transmit queues for cards that support
1304 them, and a lock-less ring-buffer implementation to
1305 enable drivers to more efficiently manage queueing of
1309 A new version of ZFS (version 13) has been merged to -HEAD.
1310 This version has zpool attribute "listsnapshots" off by
1311 default, which means "zfs list" does not show snapshots,
1312 and is the same as Solaris behavior.
1315 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
1318 The uhci, ohci, ehci and slhci USB Host controller drivers have
1319 been put into separate modules. If you load the usb module
1320 separately through loader.conf you will need to load the
1321 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
1322 controller add the following to loader.conf:
1328 The ABI used by the PMC toolset has changed. Please keep
1329 userland (libpmc(3)) and the kernel module (hwpmc(4)) in
1333 atapci kernel module now includes only generic PCI ATA
1334 driver. AHCI driver moved to ataahci kernel module.
1335 All vendor-specific code moved into separate kernel modules:
1336 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek,
1337 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron,
1338 atamarvell, atamicron, atanational, atanetcell, atanvidia,
1339 atapromise, ataserverworks, atasiliconimage, atasis, atavia
1342 The TTY subsystem of the kernel has been replaced by a new
1343 implementation, which provides better scalability and an
1344 improved driver model. Most common drivers have been migrated to
1345 the new TTY subsystem, while others have not. The following
1346 drivers have not yet been ported to the new TTY layer:
1349 cy, digi, rc, rp, sio
1355 ng_h4, ng_tty, ppp, sl, snp
1357 Adding these drivers to your kernel configuration file shall
1358 cause compilation to fail.
1361 ntpd has been upgraded to 4.2.4p5.
1364 OpenSSH has been upgraded to 5.1p1.
1366 For many years, FreeBSD's version of OpenSSH preferred DSA
1367 over RSA for host and user authentication keys. With this
1368 upgrade, we've switched to the vendor's default of RSA over
1369 DSA. This may cause upgraded clients to warn about unknown
1370 host keys even for previously known hosts. Users should
1371 follow the usual procedure for verifying host keys before
1372 accepting the RSA key.
1374 This can be circumvented by setting the "HostKeyAlgorithms"
1375 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
1378 Please note that the sequence of keys offered for
1379 authentication has been changed as well. You may want to
1380 specify IdentityFile in a different order to revert this
1384 The sio(4) driver has been removed from the i386 and amd64
1385 kernel configuration files. This means uart(4) is now the
1386 default serial port driver on those platforms as well.
1388 To prevent collisions with the sio(4) driver, the uart(4) driver
1389 uses different names for its device nodes. This means the
1390 onboard serial port will now most likely be called "ttyu0"
1391 instead of "ttyd0". You may need to reconfigure applications to
1392 use the new device names.
1394 When using the serial port as a boot console, be sure to update
1395 /boot/device.hints and /etc/ttys before booting the new kernel.
1396 If you forget to do so, you can still manually specify the hints
1397 at the loader prompt:
1399 set hint.uart.0.at="isa"
1400 set hint.uart.0.port="0x3F8"
1401 set hint.uart.0.flags="0x10"
1402 set hint.uart.0.irq="4"
1406 The gpt(8) utility has been removed. Use gpart(8) to partition
1410 The version that Linuxulator emulates was changed from 2.4.2
1411 to 2.6.16. If you experience any problems with Linux binaries
1412 please try to set sysctl compat.linux.osrelease to 2.4.2 and
1413 if it fixes the problem contact emulation mailing list.
1416 ISDN4BSD (I4B) was removed from the src tree. You may need to
1417 update a your kernel configuration and remove relevant entries.
1420 I have checked in code to support multiple routing tables.
1421 See the man pages setfib(1) and setfib(2).
1422 This is a hopefully backwards compatible version,
1423 but to make use of it you need to compile your kernel
1424 with options ROUTETABLES=2 (or more up to 16).
1427 The 802.11 wireless support was redone to enable multi-bss
1428 operation on devices that are capable. The underlying device
1429 is no longer used directly but instead wlanX devices are
1430 cloned with ifconfig. This requires changes to rc.conf files.
1431 For example, change:
1432 ifconfig_ath0="WPA DHCP"
1435 ifconfig_wlan0="WPA DHCP"
1436 see rc.conf(5) for more details. In addition, mergemaster of
1437 /etc/rc.d is highly recommended. Simultaneous update of userland
1438 and kernel wouldn't hurt either.
1440 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
1441 modules were merged into the base wlan module. All references
1442 to these modules (e.g. in kernel config files) must be removed.
1445 psm(4) has gained write(2) support in native operation level.
1446 Arbitrary commands can be written to /dev/psm%d and status can
1447 be read back from it. Therefore, an application is responsible
1448 for status validation and error recovery. It is a no-op in
1449 other operation levels.
1452 Support for KSE threading has been removed from the kernel. To
1453 run legacy applications linked against KSE libmap.conf may
1454 be used. The following libmap.conf may be used to ensure
1455 compatibility with any prior release:
1457 libpthread.so.1 libthr.so.1
1458 libpthread.so.2 libthr.so.2
1459 libkse.so.3 libthr.so.3
1462 The layout of struct vmspace has changed. This affects libkvm
1463 and any executables that link against libkvm and use the
1464 kvm_getprocs() function. In particular, but not exclusively,
1465 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
1466 The effects are minimal, but it's advisable to upgrade world
1470 The latest em driver no longer has support in it for the
1471 82575 adapter, this is now moved to the igb driver. The
1472 split was done to make new features that are incompatible
1473 with older hardware easier to do.
1476 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
1477 likewise the kernel option is now GEOM_LINUX_LVM.
1480 The default NFS mount mode has changed from UDP to TCP for
1481 increased reliability. If you rely on (insecurely) NFS
1482 mounting across a firewall you may need to update your
1486 Belatedly note the addition of m_collapse for compacting
1490 The fts(3) structures have been changed to use adequate
1491 integer types for their members and so to be able to cope
1492 with huge file trees. The old fts(3) ABI is preserved
1493 through symbol versioning in libc, so third-party binaries
1494 using fts(3) should still work, although they will not take
1495 advantage of the extended types. At the same time, some
1496 third-party software might fail to build after this change
1497 due to unportable assumptions made in its source code about
1498 fts(3) structure members. Such software should be fixed
1499 by its vendor or, in the worst case, in the ports tree.
1500 FreeBSD_version 800015 marks this change for the unlikely
1501 case that a portable fix is impossible.
1504 To upgrade to -current after this date, you must be running
1505 FreeBSD not older than 6.0-RELEASE. Upgrading to -current
1506 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
1509 The ADAPTIVE_GIANT kernel option has been retired because its
1510 functionality is the default now.
1513 The AT keyboard emulation of sunkbd(4) has been turned on
1514 by default. In order to make the special symbols of the Sun
1515 keyboards driven by sunkbd(4) work under X these now have
1516 to be configured the same way as Sun USB keyboards driven
1517 by ukbd(4) (which also does AT keyboard emulation), f.e.:
1519 Option "XkbLayout" "us"
1520 Option "XkbRules" "xorg"
1521 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
1524 It has been decided that it is desirable to provide ABI
1525 backwards compatibility to the FreeBSD 4/5/6 versions of the
1526 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
1527 broken with the introduction of PCI domain support (see the
1528 20070930 entry). Unfortunately, this required the ABI of
1529 PCIOCGETCONF to be broken again in order to be able to
1530 provide backwards compatibility to the old version of that
1531 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
1532 again. As for prominent ports this affects neither pciutils
1533 nor xorg-server this time, the hal port needs to be rebuilt
1537 The misnamed kthread_create() and friends have been renamed
1538 to kproc_create() etc. Many of the callers already
1539 used kproc_start()..
1540 I will return kthread_create() and friends in a while
1541 with implementations that actually create threads, not procs.
1542 Renaming corresponds with version 800002.
1551 Avoid using make -j when upgrading. While generally safe, there are
1552 sometimes problems using -j to upgrade. If your upgrade fails with
1553 -j, please try again without -j. From time to time in the past there
1554 have been problems using -j with buildworld and/or installworld. This
1555 is especially true when upgrading between "distant" versions (eg one
1556 that cross a major release boundary or several minor releases, or when
1557 several months have passed on the -current branch).
1559 Sometimes, obscure build problems are the result of environment
1560 poisoning. This can happen because the make utility reads its
1561 environment when searching for values for global variables. To run
1562 your build attempts in an "environmental clean room", prefix all make
1563 commands with 'env -i '. See the env(1) manual page for more details.
1565 When upgrading from one major version to another it is generally best
1566 to upgrade to the latest code in the currently installed branch first,
1567 then do an upgrade to the new branch. This is the best-tested upgrade
1568 path, and has the highest probability of being successful. Please try
1569 this approach before reporting problems with a major version upgrade.
1573 When upgrading the boot ZFS pool to a new version, always follow
1576 1.) recompile and reinstall the ZFS boot loader and boot block
1577 (this is part of "make buildworld" and "make installworld")
1579 2.) update the ZFS boot block on your boot drive
1581 The following example updates the ZFS boot block on the first
1582 partition (freebsd-boot) of a GPT partitioned drive ad0:
1583 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0"
1585 Non-boot pools do not need these updates.
1589 If you are updating from a prior version of FreeBSD (even one just
1590 a few days old), you should follow this procedure. It is the most
1591 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
1593 make kernel-toolchain
1594 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
1595 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
1597 To test a kernel once
1598 ---------------------
1599 If you just want to boot a kernel once (because you are not sure
1600 if it works, or if you want to boot a known bad kernel to provide
1601 debugging information) run
1602 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
1603 nextboot -k testkernel
1605 To just build a kernel when you know that it won't mess you up
1606 --------------------------------------------------------------
1607 This assumes you are already running a CURRENT system. Replace
1608 ${arch} with the architecture of your machine (e.g. "i386",
1609 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc).
1611 cd src/sys/${arch}/conf
1612 config KERNEL_NAME_HERE
1613 cd ../compile/KERNEL_NAME_HERE
1618 If this fails, go to the "To build a kernel" section.
1620 To rebuild everything and install it on the current system.
1621 -----------------------------------------------------------
1622 # Note: sometimes if you are running current you gotta do more than
1623 # is listed here if you are upgrading from a really old current.
1625 <make sure you have good level 0 dumps>
1627 make kernel KERNCONF=YOUR_KERNEL_HERE
1629 <reboot in single user> [3]
1637 To cross-install current onto a separate partition
1638 --------------------------------------------------
1639 # In this approach we use a separate partition to hold
1640 # current's root, 'usr', and 'var' directories. A partition
1641 # holding "/", "/usr" and "/var" should be about 2GB in
1644 <make sure you have good level 0 dumps>
1647 make buildkernel KERNCONF=YOUR_KERNEL_HERE
1648 <maybe newfs current's root partition>
1649 <mount current's root partition on directory ${CURRENT_ROOT}>
1650 make installworld DESTDIR=${CURRENT_ROOT}
1651 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
1652 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
1653 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
1654 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
1655 <reboot into current>
1656 <do a "native" rebuild/install as described in the previous section>
1657 <maybe install compatibility libraries from ports/misc/compat*>
1661 To upgrade in-place from stable to current
1662 ----------------------------------------------
1663 <make sure you have good level 0 dumps>
1665 make kernel KERNCONF=YOUR_KERNEL_HERE [8]
1667 <reboot in single user> [3]
1674 Make sure that you've read the UPDATING file to understand the
1675 tweaks to various things you need. At this point in the life
1676 cycle of current, things change often and you are on your own
1677 to cope. The defaults can also change, so please read ALL of
1678 the UPDATING entries.
1680 Also, if you are tracking -current, you must be subscribed to
1681 freebsd-current@freebsd.org. Make sure that before you update
1682 your sources that you have read and understood all the recent
1683 messages there. If in doubt, please track -stable which has
1684 much fewer pitfalls.
1686 [1] If you have third party modules, such as vmware, you
1687 should disable them at this point so they don't crash your
1690 [3] From the bootblocks, boot -s, and then do
1695 adjkerntz -i # if CMOS is wall time
1696 Also, when doing a major release upgrade, it is required that
1697 you boot into single user mode to do the installworld.
1699 [4] Note: This step is non-optional. Failure to do this step
1700 can result in a significant reduction in the functionality of the
1701 system. Attempting to do it by hand is not recommended and those
1702 that pursue this avenue should read this file carefully, as well
1703 as the archives of freebsd-current and freebsd-hackers mailing lists
1704 for potential gotchas. The -U option is also useful to consider.
1705 See mergemaster(8) for more information.
1707 [5] Usually this step is a noop. However, from time to time
1708 you may need to do this if you get unknown user in the following
1709 step. It never hurts to do it all the time. You may need to
1710 install a new mergemaster (cd src/usr.sbin/mergemaster && make
1711 install) after the buildworld before this step if you last updated
1712 from [78]-stable or 9-stable before 20130430.
1714 [6] This only deletes old files and directories. Old libraries
1715 can be deleted by "make delete-old-libs", but you have to make
1716 sure that no program is using those libraries anymore.
1718 [8] In order to have a kernel that can run the 4.x binaries needed to
1719 do an installworld, you must include the COMPAT_FREEBSD4 option in
1720 your kernel. Failure to do so may leave you with a system that is
1721 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is
1722 required to run the 5.x binaries on more recent kernels. And so on
1723 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7.
1725 Make sure that you merge any new devices from GENERIC since the
1726 last time you updated your kernel config file.
1728 [9] When checking out sources, you must include the -P flag to have
1729 cvs prune empty directories.
1731 If CPUTYPE is defined in your /etc/make.conf, make sure to use the
1732 "?=" instead of the "=" assignment operator, so that buildworld can
1733 override the CPUTYPE if it needs to.
1735 MAKEOBJDIRPREFIX must be defined in an environment variable, and
1736 not on the command line, or in /etc/make.conf. buildworld will
1737 warn if it is improperly defined.
1740 This file contains a list, in reverse chronological order, of major
1741 breakages in tracking -current. It is not guaranteed to be a complete
1742 list of such breakages, and only contains entries since October 10, 2007.
1743 If you need to see UPDATING entries from before that date, you will need
1744 to fetch an UPDATING file from an older FreeBSD release.
1746 Copyright information:
1748 Copyright 1998-2009 M. Warner Losh. All Rights Reserved.
1750 Redistribution, publication, translation and use, with or without
1751 modification, in full or in part, in any form or format of this
1752 document are permitted without further permission from the author.
1754 THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
1755 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1756 WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
1757 DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
1758 INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
1759 (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
1760 SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
1761 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
1762 STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
1763 IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
1764 POSSIBILITY OF SUCH DAMAGE.
1766 Contact Warner Losh if you have any questions about your use of