1Updating Information for FreeBSD current users 2 3This file is maintained and copyrighted by M. Warner Losh <imp@freebsd.org>. 4See end of file for further details. For commonly done items, please see the 5COMMON ITEMS: section later in the file. These instructions assume that you 6basically know what you are doing. If not, then please consult the FreeBSD 7handbook. 8 9Items affecting the ports and packages system can be found in 10/usr/ports/UPDATING. Please read that file before running portupgrade. 11 12NOTE TO PEOPLE WHO THINK THAT FreeBSD 10.x IS SLOW: 13 FreeBSD 10.x has many debugging features turned on, in both the kernel 14 and userland. These features attempt to detect incorrect use of 15 system primitives, and encourage loud failure through extra sanity 16 checking and fail stop semantics. They also substantially impact 17 system performance. If you want to do performance measurement, 18 benchmarking, and optimization, you'll want to turn them off. This 19 includes various WITNESS- related kernel options, INVARIANTS, malloc 20 debugging flags in userland, and various verbose features in the 21 kernel. Many developers choose to disable these features on build 22 machines to maximize performance. (To disable malloc debugging, run 23 ln -s aj /etc/malloc.conf.) 24 2520110930: 26 sysinstall has been removed 27 2820110923: 29 The stable/9 branch created in subversion. This corresponds to the 30 RELENG_9 branch in CVS. 31 3220110913: 33 This commit modifies vfs_register() so that it uses a hash 34 calculation to set vfc_typenum, which is enabled by default. 35 The first time a system is booted after this change, the 36 vfc_typenum values will change for all file systems. The 37 main effect of this is a change to the NFS server file handles 38 for file systems that use vfc_typenum in their fsid, such as ZFS. 39 It will, however, prevent vfc_typenum from changing when file 40 systems are loaded in a different order for subsequent reboots. 41 To disable this, you can set vfs.typenumhash=0 in /boot/loader.conf 42 until you are ready to remount all NFS clients after a reboot. 43 4420110828: 45 Bump the shared library version numbers for libraries that 46 do not use symbol versioning, have changed the ABI compared 47 to stable/8 and which shared library version was not bumped. 48 Done as part of 9.0-RELEASE cycle. 49 5020110815: 51 During the merge of Capsicum features, the fget(9) KPI was modified. 52 This may require the rebuilding of out-of-tree device drivers -- 53 issues have been reported specifically with the nVidia device driver. 54 __FreeBSD_version is bumped to 900041. 55 56 Also, there is a period between 20110811 and 20110814 where the 57 special devices /dev/{stdin,stdout,stderr} did not work correctly. 58 Building world from a kernel during that window may not work. 59 6020110628: 61 The packet filter (pf) code has been updated to OpenBSD 4.5. 62 You need to update userland tools to be in sync with kernel. 63 This update breaks backward compatibility with earlier pfsync(4) 64 versions. Care must be taken when updating redundant firewall setups. 65 6620110608: 67 The following sysctls and tunables are retired on x86 platforms: 68 machdep.hlt_cpus 69 machdep.hlt_logical_cpus 70 The following sysctl is retired: 71 machdep.hyperthreading_allowed 72 The sysctls were supposed to provide a way to dynamically offline and 73 online selected CPUs on x86 platforms, but the implementation has not 74 been reliable especially with SCHED_ULE scheduler. 75 machdep.hyperthreading_allowed tunable is still available to ignore 76 hyperthreading CPUs at OS level. 77 Individual CPUs can be disabled using hint.lapic.X.disabled tunable, 78 where X is an APIC ID of a CPU. Be advised, though, that disabling 79 CPUs in non-uniform fashion will result in non-uniform topology and 80 may lead to sub-optimal system performance with SCHED_ULE, which is 81 a default scheduler. 82 8320110607: 84 cpumask_t type is retired and cpuset_t is used in order to describe 85 a mask of CPUs. 86 8720110531: 88 Changes to ifconfig(8) for dynamic address family detection mandate 89 that you are running a kernel of 20110525 or later. Make sure to 90 follow the update procedure to boot a new kernel before installing 91 world. 92 9320110513: 94 Support for sun4v architecture is officially dropped 95 9620110503: 97 Several KPI breaking changes have been committed to the mii(4) layer, 98 the PHY drivers and consequently some Ethernet drivers using mii(4). 99 This means that miibus.ko and the modules of the affected Ethernet 100 drivers need to be recompiled. 101 102 Note to kernel developers: Given that the OUI bit reversion problem 103 was fixed as part of these changes all mii(4) commits related to OUIs, 104 i.e. to sys/dev/mii/miidevs, PHY driver probing and vendor specific 105 handling, no longer can be merged verbatim to stable/8 and previous 106 branches. 107 10820110430: 109 Users of the Atheros AR71xx SoC code now need to add 'device ar71xx_pci' 110 into their kernel configurations along with 'device pci'. 111 11220110427: 113 The default NFS client is now the new NFS client, so fstype "newnfs" 114 is now "nfs" and the regular/old NFS client is now fstype "oldnfs". 115 Although mounts via fstype "nfs" will usually work without userland 116 changes, it is recommended that the mount(8) and mount_nfs(8) 117 commands be rebuilt from sources and that a link to mount_nfs called 118 mount_oldnfs be created. The new client is compiled into the 119 kernel with "options NFSCL" and this is needed for diskless root 120 file systems. The GENERIC kernel configs have been changed to use 121 NFSCL and NFSD (the new server) instead of NFSCLIENT and NFSSERVER. 122 To use the regular/old client, you can "mount -t oldnfs ...". For 123 a diskless root file system, you must also include a line like: 124 125 vfs.root.mountfrom="oldnfs:" 126 127 in the boot/loader.conf on the root fs on the NFS server to make 128 a diskless root fs use the old client. 129 13020110424: 131 The GENERIC kernels for all architectures now default to the new 132 CAM-based ATA stack. It means that all legacy ATA drivers were 133 removed and replaced by respective CAM drivers. If you are using 134 ATA device names in /etc/fstab or other places, make sure to update 135 them respectively (adX -> adaY, acdX -> cdY, afdX -> daY, astX -> saY, 136 where 'Y's are the sequential numbers starting from zero for each type 137 in order of detection, unless configured otherwise with tunables, 138 see cam(4)). There will be symbolic links created in /dev/ to map 139 old adX devices to the respective adaY. They should provide basic 140 compatibility for file systems mounting in most cases, but they do 141 not support old user-level APIs and do not have respective providers 142 in GEOM. Consider using updated management tools with new device names. 143 144 It is possible to load devices ahci, ata, siis and mvs as modules, 145 but option ATA_CAM should remain in kernel configuration to make ata 146 module work as CAM driver supporting legacy ATA controllers. Device ata 147 still can be used in modular fashion (atacore + ...). Modules atadisk 148 and atapi* are not used and won't affect operation in ATA_CAM mode. 149 Note that to use CAM-based ATA kernel should include CAM devices 150 scbus, pass, da (or explicitly ada), cd and optionally others. All of 151 them are parts of the cam module. 152 153 ataraid(4) functionality is now supported by the RAID GEOM class. 154 To use it you can load geom_raid kernel module and use graid(8) tool 155 for management. Instead of /dev/arX device names, use /dev/raid/rX. 156 157 No kernel config options or code have been removed, so if a problem 158 arises, please report it and optionally revert to the old ATA stack. 159 In order to do it you can remove from the kernel config: 160 options ATA_CAM 161 device ahci 162 device mvs 163 device siis 164 , and instead add back: 165 device atadisk # ATA disk drives 166 device ataraid # ATA RAID drives 167 device atapicd # ATAPI CDROM drives 168 device atapifd # ATAPI floppy drives 169 device atapist # ATAPI tape drives 170 17120110423: 172 The default NFS server has been changed to the new server, which 173 was referred to as the experimental server. If you need to switch 174 back to the old NFS server, you must now put the "-o" option on 175 both the mountd and nfsd commands. This can be done using the 176 mountd_flags and nfs_server_flags rc.conf variables until an 177 update to the rc scripts is committed, which is coming soon. 178 17920110418: 180 The GNU Objective-C runtime library (libobjc), and other Objective-C 181 related components have been removed from the base system. If you 182 require an Objective-C library, please use one of the available ports. 183 18420110331: 185 ath(4) has been split into bus- and device- modules. if_ath contains 186 the HAL, the TX rate control and the network device code. if_ath_pci 187 contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_ahb 188 contains the AHB glue. Users need to load both if_ath_pci and if_ath 189 in order to use ath on everything else. 190 191 TO REPEAT: if_ath_ahb is not needed for normal users. Normal users only 192 need to load if_ath and if_ath_pci for ath(4) operation. 193 19420110314: 195 As part of the replacement of sysinstall, the process of building 196 release media has changed significantly. For details, please re-read 197 release(7), which has been updated to reflect the new build process. 198 19920110218: 200 GNU binutils 2.17.50 (as of 2007-07-03) has been merged to -HEAD. This 201 is the last available version under GPLv2. It brings a number of new 202 features, such as support for newer x86 CPU's (with SSE-3, SSSE-3, SSE 203 4.1 and SSE 4.2), better support for powerpc64, a number of new 204 directives, and lots of other small improvements. See the ChangeLog 205 file in contrib/binutils for the full details. 206 20720110218: 208 IPsec's HMAC_SHA256-512 support has been fixed to be RFC4868 209 compliant, and will now use half of hash for authentication. 210 This will break interoperability with all stacks (including all 211 actual FreeBSD versions) who implement 212 draft-ietf-ipsec-ciph-sha-256-00 (they use 96 bits of hash for 213 authentication). 214 The only workaround with such peers is to use another HMAC 215 algorithm for IPsec ("phase 2") authentication. 216 21720110207: 218 Remove the uio_yield prototype and symbol. This function has 219 been misnamed since it was introduced and should not be 220 globally exposed with this name. The equivalent functionality 221 is now available using kern_yield(curthread->td_user_pri). 222 The function remains undocumented. 223 22420110112: 225 A SYSCTL_[ADD_]UQUAD was added for unsigned uint64_t pointers, 226 symmetric with the existing SYSCTL_[ADD_]QUAD. Type checking 227 for scalar sysctls is defined but disabled. Code that needs 228 UQUAD to pass the type checking that must compile on older 229 systems where the define is not present can check against 230 __FreeBSD_version >= 900030. 231 232 The system dialog(1) has been replaced with a new version previously 233 in ports as devel/cdialog. dialog(1) is mostly command-line compatible 234 with the previous version, but the libdialog associated with it has 235 a largely incompatible API. As such, the original version of libdialog 236 will be kept temporarily as libodialog, until its base system consumers 237 are replaced or updated. Bump __FreeBSD_version to 900030. 238 23920110103: 240 If you are trying to run make universe on a -stable system, and you get 241 the following warning: 242 "Makefile", line 356: "Target architecture for i386/conf/GENERIC 243 unknown. config(8) likely too old." 244 or something similar to it, then you must upgrade your -stable system 245 to 8.2-Release or newer (really, any time after r210146 7/15/2010 in 246 stable/8) or build the config from the latest stable/8 branch and 247 install it on your system. 248 249 Prior to this date, building a current universe on 8-stable system from 250 between 7/15/2010 and 1/2/2011 would result in a weird shell parsing 251 error in the first kernel build phase. A new config on those old 252 systems will fix that problem for older versions of -current. 253 25420101228: 255 The TCP stack has been modified to allow Khelp modules to interact with 256 it via helper hook points and store per-connection data in the TCP 257 control block. Bump __FreeBSD_version to 900029. User space tools that 258 rely on the size of struct tcpcb in tcp_var.h (e.g. sockstat) need to 259 be recompiled. 260 26120101114: 262 Generic IEEE 802.3 annex 31B full duplex flow control support has been 263 added to mii(4) and bge(4), bce(4), msk(4), nfe(4) and stge(4) along 264 with brgphy(4), e1000phy(4) as well as ip1000phy() have been converted 265 to take advantage of it instead of using custom implementations. This 266 means that these drivers now no longer unconditionally advertise 267 support for flow control but only do so if flow control is a selected 268 media option. This was implemented in the generic support that way in 269 order to allow flow control to be switched on and off via ifconfig(8) 270 with the PHY specific default to typically off in order to protect 271 from unwanted effects. Consequently, if you used flow control with 272 one of the above mentioned drivers you now need to explicitly enable 273 it, for example via: 274 ifconfig bge0 media auto mediaopt flowcontrol 275 276 Along with the above mentioned changes generic support for setting 277 1000baseT master mode also has been added and brgphy(4), ciphy(4), 278 e1000phy(4) as well as ip1000phy(4) have been converted to take 279 advantage of it. This means that these drivers now no longer take the 280 link0 parameter for selecting master mode but the master media option 281 has to be used instead, for example like in the following: 282 ifconfig bge0 media 1000baseT mediaopt full-duplex,master 283 284 Selection of master mode now is also available with all other PHY 285 drivers supporting 1000baseT. 286 28720101111: 288 The TCP stack has received a significant update to add support for 289 modularised congestion control and generally improve the clarity of 290 congestion control decisions. Bump __FreeBSD_version to 900025. User 291 space tools that rely on the size of struct tcpcb in tcp_var.h (e.g. 292 sockstat) need to be recompiled. 293 29420101002: 295 The man(1) utility has been replaced by a new version that no longer 296 uses /etc/manpath.config. Please consult man.conf(5) for how to 297 migrate local entries to the new format. 298 29920100928: 300 The copyright strings printed by login(1) and sshd(8) at the time of a 301 new connection have been removed to follow other operating systems and 302 upstream sshd. 303 30420100915: 305 A workaround for a fixed ld bug has been removed in kernel code, 306 so make sure that your system ld is built from sources after 307 revision 210245 from 2010-07-19 (r211583 if building head kernel 308 on stable/8, r211584 for stable/7; both from 2010-08-21). 309 A symptom of incorrect ld version is different addresses for 310 set_pcpu section and __start_set_pcpu symbol in kernel and/or modules. 311 31220100913: 313 The $ipv6_prefer variable in rc.conf(5) has been split into 314 $ip6addrctl_policy and $ipv6_activate_all_interfaces. 315 316 The $ip6addrctl_policy is a variable to choose a pre-defined 317 address selection policy set by ip6addrctl(8). A value 318 "ipv4_prefer", "ipv6_prefer" or "AUTO" can be specified. The 319 default is "AUTO". 320 321 The $ipv6_activate_all_interfaces specifies whether IFDISABLED 322 flag (see an entry of 20090926) is set on an interface with no 323 corresponding $ifconfig_IF_ipv6 line. The default is "NO" for 324 security reason. If you want IPv6 link-local address on all 325 interfaces by default, set this to "YES". 326 327 The old ipv6_prefer="YES" is equivalent to 328 ipv6_activate_all_interfaces="YES" and 329 ip6addrctl_policy="ipv6_prefer". 330 33120100913: 332 DTrace has grown support for userland tracing. Due to this, DTrace is 333 now i386 and amd64 only. 334 dtruss(1) is now installed by default on those systems and a new 335 kernel module is needed for userland tracing: fasttrap. 336 No changes to your kernel config file are necessary to enable 337 userland tracing, but you might consider adding 'STRIP=' and 338 'CFLAGS+=-fno-omit-frame-pointer' to your make.conf if you want 339 to have informative userland stack traces in DTrace (ustack). 340 34120100725: 342 The acpi_aiboost(4) driver has been removed in favor of the new 343 aibs(4) driver. You should update your kernel configuration file. 344 34520100722: 346 BSD grep has been imported to the base system and it is built by 347 default. It is completely BSD licensed, highly GNU-compatible, uses 348 less memory than its GNU counterpart and has a small codebase. 349 However, it is slower than its GNU counterpart, which is mostly 350 noticeable for larger searches, for smaller ones it is measurable 351 but not significant. The reason is complex, the most important factor 352 is that we lack a modern and efficient regex library and GNU 353 overcomes this by optimizing the searches internally. Future work 354 on improving the regex performance is planned, for the meantime, 355 users that need better performance, can build GNU grep instead by 356 setting the WITH_GNU_GREP knob. 357 35820100713: 359 Due to the import of powerpc64 support, all existing powerpc kernel 360 configuration files must be updated with a machine directive like this: 361 machine powerpc powerpc 362 363 In addition, an updated config(8) is required to build powerpc kernels 364 after this change. 365 36620100713: 367 A new version of ZFS (version 15) has been merged to -HEAD. 368 This version uses a python library for the following subcommands: 369 zfs allow, zfs unallow, zfs groupspace, zfs userspace. 370 For full functionality of these commands the following port must 371 be installed: sysutils/py-zfs 372 37320100429: 374 'vm_page's are now hashed by physical address to an array of mutexes. 375 Currently this is only used to serialize access to hold_count. Over 376 time the page queue mutex will be peeled away. This changes the size 377 of pmap on every architecture. And requires all callers of vm_page_hold 378 and vm_page_unhold to be updated. 379 38020100402: 381 WITH_CTF can now be specified in src.conf (not recommended, there 382 are some problems with static executables), make.conf (would also 383 affect ports which do not use GNU make and do not override the 384 compile targets) or in the kernel config (via "makeoptions 385 WITH_CTF=yes"). 386 When WITH_CTF was specified there before this was silently ignored, 387 so make sure that WITH_CTF is not used in places which could lead 388 to unwanted behavior. 389 39020100311: 391 The kernel option COMPAT_IA32 has been replaced with COMPAT_FREEBSD32 392 to allow 32-bit compatibility on non-x86 platforms. All kernel 393 configurations on amd64 and ia64 platforms using these options must 394 be modified accordingly. 395 39620100113: 397 The utmp user accounting database has been replaced with utmpx, 398 the user accounting interface standardized by POSIX. 399 Unfortunately the semantics of utmp and utmpx don't match, 400 making it practically impossible to support both interfaces. 401 The user accounting database is used by tools like finger(1), 402 last(1), talk(1), w(1) and ac(8). 403 404 All applications in the base system use utmpx. This means only 405 local binaries (e.g. from the ports tree) may still use these 406 utmp database files. These applications must be rebuilt to make 407 use of utmpx. 408 409 After the system has been upgraded, it is safe to remove the old 410 log files (/var/run/utmp, /var/log/lastlog and /var/log/wtmp*), 411 assuming their contents is of no importance anymore. Old wtmp 412 databases can only be used by last(1) and ac(8) after they have 413 been converted to the new format using wtmpcvt(1). 414 41520100108: 416 Introduce the kernel thread "deadlock resolver" (which can be enabled 417 via the DEADLKRES option, see NOTES for more details) and the 418 sleepq_type() function for sleepqueues. 419 42020091202: 421 The rc.firewall and rc.firewall6 were unified, and 422 rc.firewall6 and rc.d/ip6fw were removed. 423 According to the removal of rc.d/ip6fw, ipv6_firewall_* rc 424 variables are obsoleted. Instead, the following new rc 425 variables are added to rc.d/ipfw: 426 427 firewall_client_net_ipv6, firewall_simple_iif_ipv6, 428 firewall_simple_inet_ipv6, firewall_simple_oif_ipv6, 429 firewall_simple_onet_ipv6, firewall_trusted_ipv6 430 431 The meanings correspond to the relevant IPv4 variables. 432 43320091125: 434 8.0-RELEASE. 435 43620091113: 437 The default terminal emulation for syscons(4) has been changed 438 from cons25 to xterm on all platforms except pc98. This means 439 that the /etc/ttys file needs to be updated to ensure correct 440 operation of applications on the console. 441 442 The terminal emulation style can be toggled per window by using 443 vidcontrol(1)'s -T flag. The TEKEN_CONS25 kernel configuration 444 options can be used to change the compile-time default back to 445 cons25. 446 447 To prevent graphical artifacts, make sure the TERM environment 448 variable is set to match the terminal emulation that is being 449 performed by syscons(4). 450 45120091109: 452 The layout of the structure ieee80211req_scan_result has changed. 453 Applications that require wireless scan results (e.g. ifconfig(8)) 454 from net80211 need to be recompiled. 455 456 Applications such as wpa_supplicant(8) may require a full world 457 build without using NO_CLEAN in order to get synchronized with the 458 new structure. 459 46020091025: 461 The iwn(4) driver has been updated to support the 5000 and 5150 series. 462 There's one kernel module for each firmware. Adding "device iwnfw" 463 to the kernel configuration file means including all three firmware 464 images inside the kernel. If you want to include just the one for 465 your wireless card, use the devices iwn4965fw, iwn5000fw or 466 iwn5150fw. 467 46820090926: 469 The rc.d/network_ipv6, IPv6 configuration script has been integrated 470 into rc.d/netif. The changes are the following: 471 472 1. To use IPv6, simply define $ifconfig_IF_ipv6 like $ifconfig_IF 473 for IPv4. For aliases, $ifconfig_IF_aliasN should be used. 474 Note that both variables need the "inet6" keyword at the head. 475 476 Do not set $ipv6_network_interfaces manually if you do not 477 understand what you are doing. It is not needed in most cases. 478 479 $ipv6_ifconfig_IF and $ipv6_ifconfig_IF_aliasN still work, but 480 they are obsolete. 481 482 2. $ipv6_enable is obsolete. Use $ipv6_prefer and 483 "inet6 accept_rtadv" keyword in ifconfig(8) instead. 484 485 If you define $ipv6_enable=YES, it means $ipv6_prefer=YES and 486 all configured interfaces have "inet6 accept_rtadv" in the 487 $ifconfig_IF_ipv6. These are for backward compatibility. 488 489 3. A new variable $ipv6_prefer has been added. If NO, IPv6 490 functionality of interfaces with no corresponding 491 $ifconfig_IF_ipv6 is disabled by using "inet6 ifdisabled" flag, 492 and the default address selection policy of ip6addrctl(8) 493 is the IPv4-preferred one (see rc.d/ip6addrctl for more details). 494 Note that if you want to configure IPv6 functionality on the 495 disabled interfaces after boot, first you need to clear the flag by 496 using ifconfig(8) like: 497 498 ifconfig em0 inet6 -ifdisabled 499 500 If YES, the default address selection policy is set as 501 IPv6-preferred. 502 503 The default value of $ipv6_prefer is NO. 504 505 4. If your system need to receive Router Advertisement messages, 506 define "inet6 accept_rtadv" in $ifconfig_IF_ipv6. The rc(8) 507 scripts automatically invoke rtsol(8) when the interface becomes 508 UP. The Router Advertisement messages are used for SLAAC 509 (State-Less Address AutoConfiguration). 510 51120090922: 512 802.11s D3.03 support was committed. This is incompatible with the 513 previous code, which was based on D3.0. 514 51520090912: 516 A sysctl variable net.inet6.ip6.accept_rtadv now sets the default value 517 of a per-interface flag ND6_IFF_ACCEPT_RTADV, not a global knob to 518 control whether accepting Router Advertisement messages or not. 519 Also, a per-interface flag ND6_IFF_AUTO_LINKLOCAL has been added and 520 a sysctl variable net.inet6.ip6.auto_linklocal is its default value. 521 The ifconfig(8) utility now supports these flags. 522 52320090910: 524 ZFS snapshots are now mounted with MNT_IGNORE flag. Use -v option for 525 mount(8) and -a option for df(1) to see them. 526 52720090825: 528 The old tunable hw.bus.devctl_disable has been superseded by 529 hw.bus.devctl_queue. hw.bus.devctl_disable=1 in loader.conf should be 530 replaced by hw.bus.devctl_queue=0. The default for this new tunable 531 is 1000. 532 53320090813: 534 Remove the option STOP_NMI. The default action is now to use NMI only 535 for KDB via the newly introduced function stop_cpus_hard() and 536 maintain stop_cpus() to just use a normal IPI_STOP on ia32 and amd64. 537 53820090803: 539 The stable/8 branch created in subversion. This corresponds to the 540 RELENG_8 branch in CVS. 541 54220090719: 543 Bump the shared library version numbers for all libraries that do not 544 use symbol versioning as part of the 8.0-RELEASE cycle. Bump 545 __FreeBSD_version to 800105. 546 54720090714: 548 Due to changes in the implementation of virtual network stack support, 549 all network-related kernel modules must be recompiled. As this change 550 breaks the ABI, bump __FreeBSD_version to 800104. 551 55220090713: 553 The TOE interface to the TCP syncache has been modified to remove 554 struct tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack. 555 The cxgb driver is the only TOE consumer affected by this change, and 556 needs to be recompiled along with the kernel. As this change breaks 557 the ABI, bump __FreeBSD_version to 800103. 558 55920090712: 560 Padding has been added to struct tcpcb, sackhint and tcpstat in 561 <netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst 562 maintaining the ABI. However, this change breaks the ABI, so bump 563 __FreeBSD_version to 800102. User space tools that rely on the size of 564 any of these structs (e.g. sockstat) need to be recompiled. 565 56620090630: 567 The NFS_LEGACYRPC option has been removed along with the old kernel 568 RPC implementation that this option selected. Kernel configurations 569 may need to be adjusted. 570 57120090629: 572 The network interface device nodes at /dev/net/<interface> have been 573 removed. All ioctl operations can be performed the normal way using 574 routing sockets. The kqueue functionality can generally be replaced 575 with routing sockets. 576 57720090628: 578 The documentation from the FreeBSD Documentation Project (Handbook, 579 FAQ, etc.) is now installed via packages by sysinstall(8) and under 580 the /usr/local/share/doc/freebsd directory instead of /usr/share/doc. 581 58220090624: 583 The ABI of various structures related to the SYSV IPC API have been 584 changed. As a result, the COMPAT_FREEBSD[456] and COMPAT_43 kernel 585 options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version to 586 800100. 587 58820090622: 589 Layout of struct vnet has changed as routing related variables were 590 moved to their own Vimage module. Modules need to be recompiled. Bump 591 __FreeBSD_version to 800099. 592 59320090619: 594 NGROUPS_MAX and NGROUPS have been increased from 16 to 1023 and 1024 595 respectively. As long as no more than 16 groups per process are used, 596 no changes should be visible. When more than 16 groups are used, old 597 binaries may fail if they call getgroups() or getgrouplist() with 598 statically sized storage. Recompiling will work around this, but 599 applications should be modified to use dynamically allocated storage 600 for group arrays as POSIX.1-2008 does not cap an implementation's 601 number of supported groups at NGROUPS_MAX+1 as previous versions did. 602 603 NFS and portalfs mounts may also be affected as the list of groups is 604 truncated to 16. Users of NFS who use more than 16 groups, should 605 take care that negative group permissions are not used on the exported 606 file systems as they will not be reliable unless a GSSAPI based 607 authentication method is used. 608 60920090616: 610 The compiling option ADAPTIVE_LOCKMGRS has been introduced. This 611 option compiles in the support for adaptive spinning for lockmgrs 612 which want to enable it. The lockinit() function now accepts the flag 613 LK_ADAPTIVE in order to make the lock object subject to adaptive 614 spinning when both held in write and read mode. 615 61620090613: 617 The layout of the structure returned by IEEE80211_IOC_STA_INFO has 618 changed. User applications that use this ioctl need to be rebuilt. 619 62020090611: 621 The layout of struct thread has changed. Kernel and modules need to 622 be rebuilt. 623 62420090608: 625 The layout of structs ifnet, domain, protosw and vnet_net has changed. 626 Kernel modules need to be rebuilt. Bump __FreeBSD_version to 800097. 627 62820090602: 629 window(1) has been removed from the base system. It can now be 630 installed from ports. The port is called misc/window. 631 63220090601: 633 The way we are storing and accessing `routing table' entries has 634 changed. Programs reading the FIB, like netstat, need to be 635 re-compiled. 636 63720090601: 638 A new netisr implementation has been added for FreeBSD 8. Network 639 file system modules, such as igmp, ipdivert, and others, should be 640 rebuilt. 641 Bump __FreeBSD_version to 800096. 642 64320090530: 644 Remove the tunable/sysctl debug.mpsafevfs as its initial purpose is no 645 more valid. 646 64720090530: 648 Add VOP_ACCESSX(9). File system modules need to be rebuilt. 649 Bump __FreeBSD_version to 800094. 650 65120090529: 652 Add mnt_xflag field to 'struct mount'. File system modules need to be 653 rebuilt. 654 Bump __FreeBSD_version to 800093. 655 65620090528: 657 The compiling option ADAPTIVE_SX has been retired while it has been 658 introduced the option NO_ADAPTIVE_SX which handles the reversed logic. 659 The KPI for sx_init_flags() changes as accepting flags: 660 SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag has 661 been introduced in order to handle the reversed logic. 662 Bump __FreeBSD_version to 800092. 663 66420090527: 665 Add support for hierarchical jails. Remove global securelevel. 666 Bump __FreeBSD_version to 800091. 667 66820090523: 669 The layout of struct vnet_net has changed, therefore modules 670 need to be rebuilt. 671 Bump __FreeBSD_version to 800090. 672 67320090523: 674 The newly imported zic(8) produces a new format in the output. Please 675 run tzsetup(8) to install the newly created data to /etc/localtime. 676 67720090520: 678 The sysctl tree for the usb stack has renamed from hw.usb2.* to 679 hw.usb.* and is now consistent again with previous releases. 680 68120090520: 682 802.11 monitor mode support was revised and driver api's were changed. 683 Drivers dependent on net80211 now support DLT_IEEE802_11_RADIO instead 684 of DLT_IEEE802_11. No user-visible data structures were changed but 685 applications that use DLT_IEEE802_11 may require changes. 686 Bump __FreeBSD_version to 800088. 687 68820090430: 689 The layout of the following structs has changed: sysctl_oid, 690 socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet, 691 vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or 692 panics may be experienced. World rebuild is required for 693 correctly checking networking state from userland. 694 Bump __FreeBSD_version to 800085. 695 69620090429: 697 MLDv2 and Source-Specific Multicast (SSM) have been merged 698 to the IPv6 stack. VIMAGE hooks are in but not yet used. 699 The implementation of SSM within FreeBSD's IPv6 stack closely 700 follows the IPv4 implementation. 701 702 For kernel developers: 703 704 * The most important changes are that the ip6_output() and 705 ip6_input() paths no longer take the IN6_MULTI_LOCK, 706 and this lock has been downgraded to a non-recursive mutex. 707 708 * As with the changes to the IPv4 stack to support SSM, filtering 709 of inbound multicast traffic must now be performed by transport 710 protocols within the IPv6 stack. This does not apply to TCP and 711 SCTP, however, it does apply to UDP in IPv6 and raw IPv6. 712 713 * The KPIs used by IPv6 multicast are similar to those used by 714 the IPv4 stack, with the following differences: 715 * im6o_mc_filter() is analogous to imo_multicast_filter(). 716 * The legacy KAME entry points in6_joingroup and in6_leavegroup() 717 are shimmed to in6_mc_join() and in6_mc_leave() respectively. 718 * IN6_LOOKUP_MULTI() has been deprecated and removed. 719 * IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs 720 for MLDv1 have an additional 'timer' argument which is used to 721 jitter the initial membership report for the solicited-node 722 multicast membership on-link. 723 * This is not strictly needed for MLDv2, which already jitters 724 its report transmissions. However, the 'timer' argument is 725 preserved in case MLDv1 is active on the interface. 726 727 * The KAME linked-list based IPv6 membership implementation has 728 been refactored to use a vector similar to that used by the IPv4 729 stack. 730 Code which maintains a list of its own multicast memberships 731 internally, e.g. carp, has been updated to reflect the new 732 semantics. 733 734 * There is a known Lock Order Reversal (LOR) due to in6_setscope() 735 acquiring the IF_AFDATA_LOCK and being called within ip6_output(). 736 Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an 737 implementation constraint which needs to be addressed in HEAD. 738 739 For application developers: 740 741 * The changes are broadly similar to those made for the IPv4 742 stack. 743 744 * The use of IPv4 and IPv6 multicast socket options on the same 745 socket, using mapped addresses, HAS NOT been tested or supported. 746 747 * There are a number of issues with the implementation of various 748 IPv6 multicast APIs which need to be resolved in the API surface 749 before the implementation is fully compatible with KAME userland 750 use, and these are mostly to do with interface index treatment. 751 752 * The literature available discusses the use of either the delta / ASM 753 API with setsockopt(2)/getsockopt(2), or the full-state / ASM API 754 using setsourcefilter(3)/getsourcefilter(3). For more information 755 please refer to RFC 3768, 'Socket Interface Extensions for 756 Multicast Source Filters'. 757 758 * Applications which use the published RFC 3678 APIs should be fine. 759 760 For systems administrators: 761 762 * The mtest(8) utility has been refactored to support IPv6, in 763 addition to IPv4. Interface addresses are no longer accepted 764 as arguments, their names must be used instead. The utility 765 will map the interface name to its first IPv4 address as 766 returned by getifaddrs(3). 767 768 * The ifmcstat(8) utility has also been updated to print the MLDv2 769 endpoint state and source filter lists via sysctl(3). 770 771 * The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable 772 loopback of IPv6 multicast datagrams by default; it defaults to 1 773 to preserve the existing behaviour. Disabling multicast loopback is 774 recommended for optimal system performance. 775 776 * The IPv6 MROUTING code has been changed to examine this sysctl 777 instead of attempting to perform a group lookup before looping 778 back forwarded datagrams. 779 780 Bump __FreeBSD_version to 800084. 781 78220090422: 783 Implement low-level Bluetooth HCI API. 784 Bump __FreeBSD_version to 800083. 785 78620090419: 787 The layout of struct malloc_type, used by modules to register new 788 memory allocation types, has changed. Most modules will need to 789 be rebuilt or panics may be experienced. 790 Bump __FreeBSD_version to 800081. 791 79220090415: 793 Anticipate overflowing inp_flags - add inp_flags2. 794 This changes most offsets in inpcb, so checking v4 connection 795 state will require a world rebuild. 796 Bump __FreeBSD_version to 800080. 797 79820090415: 799 Add an llentry to struct route and struct route_in6. Modules 800 embedding a struct route will need to be recompiled. 801 Bump __FreeBSD_version to 800079. 802 80320090414: 804 The size of rt_metrics_lite and by extension rtentry has changed. 805 Networking administration apps will need to be recompiled. 806 The route command now supports show as an alias for get, weighting 807 of routes, sticky and nostick flags to alter the behavior of stateful 808 load balancing. 809 Bump __FreeBSD_version to 800078. 810 81120090408: 812 Do not use Giant for kbdmux(4) locking. This is wrong and 813 apparently causing more problems than it solves. This will 814 re-open the issue where interrupt handlers may race with 815 kbdmux(4) in polling mode. Typical symptoms include (but 816 not limited to) duplicated and/or missing characters when 817 low level console functions (such as gets) are used while 818 interrupts are enabled (for example geli password prompt, 819 mountroot prompt etc.). Disabling kbdmux(4) may help. 820 82120090407: 822 The size of structs vnet_net, vnet_inet and vnet_ipfw has changed; 823 kernel modules referencing any of the above need to be recompiled. 824 Bump __FreeBSD_version to 800075. 825 82620090320: 827 GEOM_PART has become the default partition slicer for storage devices, 828 replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It 829 introduces some changes: 830 831 MSDOS/EBR: the devices created from MSDOS extended partition entries 832 (EBR) can be named differently than with GEOM_MBR and are now symlinks 833 to devices with offset-based names. fstabs may need to be modified. 834 835 BSD: the "geometry does not match label" warning is harmless in most 836 cases but it points to problems in file system misalignment with 837 disk geometry. The "c" partition is now implicit, covers the whole 838 top-level drive and cannot be (mis)used by users. 839 840 General: Kernel dumps are now not allowed to be written to devices 841 whose partition types indicate they are meant to be used for file 842 systems (or, in case of MSDOS partitions, as something else than 843 the "386BSD" type). 844 845 Most of these changes date approximately from 200812. 846 84720090319: 848 The uscanner(4) driver has been removed from the kernel. This follows 849 Linux removing theirs in 2.6 and making libusb the default interface 850 (supported by sane). 851 85220090319: 853 The multicast forwarding code has been cleaned up. netstat(1) 854 only relies on KVM now for printing bandwidth upcall meters. 855 The IPv4 and IPv6 modules are split into ip_mroute_mod and 856 ip6_mroute_mod respectively. The config(5) options for statically 857 compiling this code remain the same, i.e. 'options MROUTING'. 858 85920090315: 860 Support for the IFF_NEEDSGIANT network interface flag has been 861 removed, which means that non-MPSAFE network device drivers are no 862 longer supported. In particular, if_ar, if_sr, and network device 863 drivers from the old (legacy) USB stack can no longer be built or 864 used. 865 86620090313: 867 POSIX.1 Native Language Support (NLS) has been enabled in libc and 868 a bunch of new language catalog files have also been added. 869 This means that some common libc messages are now localized and 870 they depend on the LC_MESSAGES environmental variable. 871 87220090313: 873 The k8temp(4) driver has been renamed to amdtemp(4) since 874 support for Family 10 and Family 11 CPU families was added. 875 87620090309: 877 IGMPv3 and Source-Specific Multicast (SSM) have been merged 878 to the IPv4 stack. VIMAGE hooks are in but not yet used. 879 880 For kernel developers, the most important changes are that the 881 ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(), 882 and this lock has been downgraded to a non-recursive mutex. 883 884 Transport protocols (UDP, Raw IP) are now responsible for filtering 885 inbound multicast traffic according to group membership and source 886 filters. The imo_multicast_filter() KPI exists for this purpose. 887 Transports which do not use multicast (SCTP, TCP) already reject 888 multicast by default. Forwarding and receive performance may improve 889 as a mutex acquisition is no longer needed in the ip_input() 890 low-level input path. in_addmulti() and in_delmulti() are shimmed 891 to new KPIs which exist to support SSM in-kernel. 892 893 For application developers, it is recommended that loopback of 894 multicast datagrams be disabled for best performance, as this 895 will still cause the lock to be taken for each looped-back 896 datagram transmission. The net.inet.ip.mcast.loop sysctl may 897 be tuned to 0 to disable loopback by default; it defaults to 1 898 to preserve the existing behaviour. 899 900 For systems administrators, to obtain best performance with 901 multicast reception and multiple groups, it is always recommended 902 that a card with a suitably precise hash filter is used. Hash 903 collisions will still result in the lock being taken within the 904 transport protocol input path to check group membership. 905 906 If deploying FreeBSD in an environment with IGMP snooping switches, 907 it is recommended that the net.inet.igmp.sendlocal sysctl remain 908 enabled; this forces 224.0.0.0/24 group membership to be announced 909 via IGMP. 910 911 The size of 'struct igmpstat' has changed; netstat needs to be 912 recompiled to reflect this. 913 Bump __FreeBSD_version to 800070. 914 91520090309: 916 libusb20.so.1 is now installed as libusb.so.1 and the ports system 917 updated to use it. This requires a buildworld/installworld in order to 918 update the library and dependencies (usbconfig, etc). Its advisable to 919 rebuild all ports which uses libusb. More specific directions are given 920 in the ports collection UPDATING file. Any /etc/libmap.conf entries for 921 libusb are no longer required and can be removed. 922 92320090302: 924 A workaround is committed to allow the creation of System V shared 925 memory segment of size > 2 GB on the 64-bit architectures. 926 Due to a limitation of the existing ABI, the shm_segsz member 927 of the struct shmid_ds, returned by shmctl(IPC_STAT) call is 928 wrong for large segments. Note that limits must be explicitly 929 raised to allow such segments to be created. 930 93120090301: 932 The layout of struct ifnet has changed, requiring a rebuild of all 933 network device driver modules. 934 93520090227: 936 The /dev handling for the new USB stack has changed, a 937 buildworld/installworld is required for libusb20. 938 93920090223: 940 The new USB2 stack has now been permanently moved in and all kernel and 941 module names reverted to their previous values (eg, usb, ehci, ohci, 942 ums, ...). The old usb stack can be compiled in by prefixing the name 943 with the letter 'o', the old usb modules have been removed. 944 Updating entry 20090216 for xorg and 20090215 for libmap may still 945 apply. 946 94720090217: 948 The rc.conf(5) option if_up_delay has been renamed to 949 defaultroute_delay to better reflect its purpose. If you have 950 customized this setting in /etc/rc.conf you need to update it to 951 use the new name. 952 95320090216: 954 xorg 7.4 wants to configure its input devices via hald which does not 955 yet work with USB2. If the keyboard/mouse does not work in xorg then 956 add 957 Option "AllowEmptyInput" "off" 958 to your ServerLayout section. This will cause X to use the configured 959 kbd and mouse sections from your xorg.conf. 960 96120090215: 962 The GENERIC kernels for all architectures now default to the new USB2 963 stack. No kernel config options or code have been removed so if a 964 problem arises please report it and optionally revert to the old USB 965 stack. If you are loading USB kernel modules or have a custom kernel 966 that includes GENERIC then ensure that usb names are also changed over, 967 eg uftdi -> usb2_serial_ftdi. 968 969 Older programs linked against the ports libusb 0.1 need to be 970 redirected to the new stack's libusb20. /etc/libmap.conf can 971 be used for this: 972 # Map old usb library to new one for usb2 stack 973 libusb-0.1.so.8 libusb20.so.1 974 97520090209: 976 All USB ethernet devices now attach as interfaces under the name ueN 977 (eg. ue0). This is to provide a predictable name as vendors often 978 change usb chipsets in a product without notice. 979 98020090203: 981 The ichsmb(4) driver has been changed to require SMBus slave 982 addresses be left-justified (xxxxxxx0b) rather than right-justified. 983 All of the other SMBus controller drivers require left-justified 984 slave addresses, so this change makes all the drivers provide the 985 same interface. 986 98720090201: 988 INET6 statistics (struct ip6stat) was updated. 989 netstat(1) needs to be recompiled. 990 99120090119: 992 NTFS has been removed from GENERIC kernel on amd64 to match 993 GENERIC on i386. Should not cause any issues since mount_ntfs(8) 994 will load ntfs.ko module automatically when NTFS support is 995 actually needed, unless ntfs.ko is not installed or security 996 level prohibits loading kernel modules. If either is the case, 997 "options NTFS" has to be added into kernel config. 998 99920090115: 1000 TCP Appropriate Byte Counting (RFC 3465) support added to kernel. 1001 New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to 1002 800061. User space tools that rely on the size of struct tcpcb in 1003 tcp_var.h (e.g. sockstat) need to be recompiled. 1004 100520081225: 1006 ng_tty(4) module updated to match the new TTY subsystem. 1007 Due to API change, user-level applications must be updated. 1008 New API support added to mpd5 CVS and expected to be present 1009 in next mpd5.3 release. 1010 101120081219: 1012 With __FreeBSD_version 800060 the makefs tool is part of 1013 the base system (it was a port). 1014 101520081216: 1016 The afdata and ifnet locks have been changed from mutexes to 1017 rwlocks, network modules will need to be re-compiled. 1018 101920081214: 1020 __FreeBSD_version 800059 incorporates the new arp-v2 rewrite. 1021 RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated. 1022 The new code reduced struct rtentry{} by 16 bytes on 32-bit 1023 architecture and 40 bytes on 64-bit architecture. The userland 1024 applications "arp" and "ndp" have been updated accordingly. 1025 The output from "netstat -r" shows only routing entries and 1026 none of the L2 information. 1027 102820081130: 1029 __FreeBSD_version 800057 marks the switchover from the 1030 binary ath hal to source code. Users must add the line: 1031 1032 options AH_SUPPORT_AR5416 1033 1034 to their kernel config files when specifying: 1035 1036 device ath_hal 1037 1038 The ath_hal module no longer exists; the code is now compiled 1039 together with the driver in the ath module. It is now 1040 possible to tailor chip support (i.e. reduce the set of chips 1041 and thereby the code size); consult ath_hal(4) for details. 1042 104320081121: 1044 __FreeBSD_version 800054 adds memory barriers to 1045 <machine/atomic.h>, new interfaces to ifnet to facilitate 1046 multiple hardware transmit queues for cards that support 1047 them, and a lock-less ring-buffer implementation to 1048 enable drivers to more efficiently manage queueing of 1049 packets. 1050 105120081117: 1052 A new version of ZFS (version 13) has been merged to -HEAD. 1053 This version has zpool attribute "listsnapshots" off by 1054 default, which means "zfs list" does not show snapshots, 1055 and is the same as Solaris behavior. 1056 105720081028: 1058 dummynet(4) ABI has changed. ipfw(8) needs to be recompiled. 1059 106020081009: 1061 The uhci, ohci, ehci and slhci USB Host controller drivers have 1062 been put into separate modules. If you load the usb module 1063 separately through loader.conf you will need to load the 1064 appropriate *hci module as well. E.g. for a UHCI-based USB 2.0 1065 controller add the following to loader.conf: 1066 1067 uhci_load="YES" 1068 ehci_load="YES" 1069 107020081009: 1071 The ABI used by the PMC toolset has changed. Please keep 1072 userland (libpmc(3)) and the kernel module (hwpmc(4)) in 1073 sync. 1074 107520081009: 1076 atapci kernel module now includes only generic PCI ATA 1077 driver. AHCI driver moved to ataahci kernel module. 1078 All vendor-specific code moved into separate kernel modules: 1079 ataacard, ataacerlabs, ataadaptec, ataamd, ataati, atacenatek, 1080 atacypress, atacyrix, atahighpoint, ataintel, ataite, atajmicron, 1081 atamarvell, atamicron, atanational, atanetcell, atanvidia, 1082 atapromise, ataserverworks, atasiliconimage, atasis, atavia 1083 108420080820: 1085 The TTY subsystem of the kernel has been replaced by a new 1086 implementation, which provides better scalability and an 1087 improved driver model. Most common drivers have been migrated to 1088 the new TTY subsystem, while others have not. The following 1089 drivers have not yet been ported to the new TTY layer: 1090 1091 PCI/ISA: 1092 cy, digi, rc, rp, sio 1093 1094 USB: 1095 ubser, ucycom 1096 1097 Line disciplines: 1098 ng_h4, ng_tty, ppp, sl, snp 1099 1100 Adding these drivers to your kernel configuration file shall 1101 cause compilation to fail. 1102 110320080818: 1104 ntpd has been upgraded to 4.2.4p5. 1105 110620080801: 1107 OpenSSH has been upgraded to 5.1p1. 1108 1109 For many years, FreeBSD's version of OpenSSH preferred DSA 1110 over RSA for host and user authentication keys. With this 1111 upgrade, we've switched to the vendor's default of RSA over 1112 DSA. This may cause upgraded clients to warn about unknown 1113 host keys even for previously known hosts. Users should 1114 follow the usual procedure for verifying host keys before 1115 accepting the RSA key. 1116 1117 This can be circumvented by setting the "HostKeyAlgorithms" 1118 option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh 1119 command line. 1120 1121 Please note that the sequence of keys offered for 1122 authentication has been changed as well. You may want to 1123 specify IdentityFile in a different order to revert this 1124 behavior. 1125 112620080713: 1127 The sio(4) driver has been removed from the i386 and amd64 1128 kernel configuration files. This means uart(4) is now the 1129 default serial port driver on those platforms as well. 1130 1131 To prevent collisions with the sio(4) driver, the uart(4) driver 1132 uses different names for its device nodes. This means the 1133 onboard serial port will now most likely be called "ttyu0" 1134 instead of "ttyd0". You may need to reconfigure applications to 1135 use the new device names. 1136 1137 When using the serial port as a boot console, be sure to update 1138 /boot/device.hints and /etc/ttys before booting the new kernel. 1139 If you forget to do so, you can still manually specify the hints 1140 at the loader prompt: 1141 1142 set hint.uart.0.at="isa" 1143 set hint.uart.0.port="0x3F8" 1144 set hint.uart.0.flags="0x10" 1145 set hint.uart.0.irq="4" 1146 boot -s 1147 114820080609: 1149 The gpt(8) utility has been removed. Use gpart(8) to partition 1150 disks instead. 1151 115220080603: 1153 The version that Linuxulator emulates was changed from 2.4.2 1154 to 2.6.16. If you experience any problems with Linux binaries 1155 please try to set sysctl compat.linux.osrelease to 2.4.2 and 1156 if it fixes the problem contact emulation mailing list. 1157 115820080525: 1159 ISDN4BSD (I4B) was removed from the src tree. You may need to 1160 update a your kernel configuration and remove relevant entries. 1161 116220080509: 1163 I have checked in code to support multiple routing tables. 1164 See the man pages setfib(1) and setfib(2). 1165 This is a hopefully backwards compatible version, 1166 but to make use of it you need to compile your kernel 1167 with options ROUTETABLES=2 (or more up to 16). 1168 116920080420: 1170 The 802.11 wireless support was redone to enable multi-bss 1171 operation on devices that are capable. The underlying device 1172 is no longer used directly but instead wlanX devices are 1173 cloned with ifconfig. This requires changes to rc.conf files. 1174 For example, change: 1175 ifconfig_ath0="WPA DHCP" 1176 to 1177 wlans_ath0=wlan0 1178 ifconfig_wlan0="WPA DHCP" 1179 see rc.conf(5) for more details. In addition, mergemaster of 1180 /etc/rc.d is highly recommended. Simultaneous update of userland 1181 and kernel wouldn't hurt either. 1182 1183 As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta 1184 modules were merged into the base wlan module. All references 1185 to these modules (e.g. in kernel config files) must be removed. 1186 118720080408: 1188 psm(4) has gained write(2) support in native operation level. 1189 Arbitrary commands can be written to /dev/psm%d and status can 1190 be read back from it. Therefore, an application is responsible 1191 for status validation and error recovery. It is a no-op in 1192 other operation levels. 1193 119420080312: 1195 Support for KSE threading has been removed from the kernel. To 1196 run legacy applications linked against KSE libmap.conf may 1197 be used. The following libmap.conf may be used to ensure 1198 compatibility with any prior release: 1199 1200 libpthread.so.1 libthr.so.1 1201 libpthread.so.2 libthr.so.2 1202 libkse.so.3 libthr.so.3 1203 120420080301: 1205 The layout of struct vmspace has changed. This affects libkvm 1206 and any executables that link against libkvm and use the 1207 kvm_getprocs() function. In particular, but not exclusively, 1208 it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1). 1209 The effects are minimal, but it's advisable to upgrade world 1210 nonetheless. 1211 121220080229: 1213 The latest em driver no longer has support in it for the 1214 82575 adapter, this is now moved to the igb driver. The 1215 split was done to make new features that are incompatible 1216 with older hardware easier to do. 1217 121820080220: 1219 The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4), 1220 likewise the kernel option is now GEOM_LINUX_LVM. 1221 122220080211: 1223 The default NFS mount mode has changed from UDP to TCP for 1224 increased reliability. If you rely on (insecurely) NFS 1225 mounting across a firewall you may need to update your 1226 firewall rules. 1227 122820080208: 1229 Belatedly note the addition of m_collapse for compacting 1230 mbuf chains. 1231 123220080126: 1233 The fts(3) structures have been changed to use adequate 1234 integer types for their members and so to be able to cope 1235 with huge file trees. The old fts(3) ABI is preserved 1236 through symbol versioning in libc, so third-party binaries 1237 using fts(3) should still work, although they will not take 1238 advantage of the extended types. At the same time, some 1239 third-party software might fail to build after this change 1240 due to unportable assumptions made in its source code about 1241 fts(3) structure members. Such software should be fixed 1242 by its vendor or, in the worst case, in the ports tree. 1243 FreeBSD_version 800015 marks this change for the unlikely 1244 case that a portable fix is impossible. 1245 124620080123: 1247 To upgrade to -current after this date, you must be running 1248 FreeBSD not older than 6.0-RELEASE. Upgrading to -current 1249 from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems. 1250 125120071128: 1252 The ADAPTIVE_GIANT kernel option has been retired because its 1253 functionality is the default now. 1254 125520071118: 1256 The AT keyboard emulation of sunkbd(4) has been turned on 1257 by default. In order to make the special symbols of the Sun 1258 keyboards driven by sunkbd(4) work under X these now have 1259 to be configured the same way as Sun USB keyboards driven 1260 by ukbd(4) (which also does AT keyboard emulation), f.e.: 1261 1262 Option "XkbLayout" "us" 1263 Option "XkbRules" "xorg" 1264 Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us" 1265 126620071024: 1267 It has been decided that it is desirable to provide ABI 1268 backwards compatibility to the FreeBSD 4/5/6 versions of the 1269 PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was 1270 broken with the introduction of PCI domain support (see the 1271 20070930 entry). Unfortunately, this required the ABI of 1272 PCIOCGETCONF to be broken again in order to be able to 1273 provide backwards compatibility to the old version of that 1274 IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled 1275 again. As for prominent ports this affects neither pciutils 1276 nor xorg-server this time, the hal port needs to be rebuilt 1277 however. 1278 127920071020: 1280 The misnamed kthread_create() and friends have been renamed 1281 to kproc_create() etc. Many of the callers already 1282 used kproc_start().. 1283 I will return kthread_create() and friends in a while 1284 with implementations that actually create threads, not procs. 1285 Renaming corresponds with version 800002. 1286 128720071010: 1288 RELENG_7 branched. 1289 1290COMMON ITEMS: 1291 1292 General Notes 1293 ------------- 1294 Avoid using make -j when upgrading. While generally safe, there are 1295 sometimes problems using -j to upgrade. If your upgrade fails with 1296 -j, please try again without -j. From time to time in the past there 1297 have been problems using -j with buildworld and/or installworld. This 1298 is especially true when upgrading between "distant" versions (eg one 1299 that cross a major release boundary or several minor releases, or when 1300 several months have passed on the -current branch). 1301 1302 Sometimes, obscure build problems are the result of environment 1303 poisoning. This can happen because the make utility reads its 1304 environment when searching for values for global variables. To run 1305 your build attempts in an "environmental clean room", prefix all make 1306 commands with 'env -i '. See the env(1) manual page for more details. 1307 1308 When upgrading from one major version to another it is generally best 1309 to upgrade to the latest code in the currently installed branch first, 1310 then do an upgrade to the new branch. This is the best-tested upgrade 1311 path, and has the highest probability of being successful. Please try 1312 this approach before reporting problems with a major version upgrade. 1313 1314 ZFS notes 1315 --------- 1316 When upgrading the boot ZFS pool to a new version, always follow 1317 these two steps: 1318 1319 1.) recompile and reinstall the ZFS boot loader and boot block 1320 (this is part of "make buildworld" and "make installworld") 1321 1322 2.) update the ZFS boot block on your boot drive 1323 1324 The following example updates the ZFS boot block on the first 1325 partition (freebsd-boot) of a GPT partitioned drive ad0: 1326 "gpart bootcode -p /boot/gptzfsboot -i 1 ad0" 1327 1328 Non-boot pools do not need these updates. 1329 1330 To build a kernel 1331 ----------------- 1332 If you are updating from a prior version of FreeBSD (even one just 1333 a few days old), you should follow this procedure. It is the most 1334 failsafe as it uses a /usr/obj tree with a fresh mini-buildworld, 1335 1336 make kernel-toolchain 1337 make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE 1338 make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE 1339 1340 To test a kernel once 1341 --------------------- 1342 If you just want to boot a kernel once (because you are not sure 1343 if it works, or if you want to boot a known bad kernel to provide 1344 debugging information) run 1345 make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel 1346 nextboot -k testkernel 1347 1348 To just build a kernel when you know that it won't mess you up 1349 -------------------------------------------------------------- 1350 This assumes you are already running a CURRENT system. Replace 1351 ${arch} with the architecture of your machine (e.g. "i386", 1352 "arm", "amd64", "ia64", "pc98", "sparc64", "powerpc", "mips", etc). 1353 1354 cd src/sys/${arch}/conf 1355 config KERNEL_NAME_HERE 1356 cd ../compile/KERNEL_NAME_HERE 1357 make depend 1358 make 1359 make install 1360 1361 If this fails, go to the "To build a kernel" section. 1362 1363 To rebuild everything and install it on the current system. 1364 ----------------------------------------------------------- 1365 # Note: sometimes if you are running current you gotta do more than 1366 # is listed here if you are upgrading from a really old current. 1367 1368 <make sure you have good level 0 dumps> 1369 make buildworld 1370 make kernel KERNCONF=YOUR_KERNEL_HERE 1371 [1] 1372 <reboot in single user> [3] 1373 mergemaster -p [5] 1374 make installworld 1375 mergemaster -i [4] 1376 make delete-old [6] 1377 <reboot> 1378 1379 1380 To cross-install current onto a separate partition 1381 -------------------------------------------------- 1382 # In this approach we use a separate partition to hold 1383 # current's root, 'usr', and 'var' directories. A partition 1384 # holding "/", "/usr" and "/var" should be about 2GB in 1385 # size. 1386 1387 <make sure you have good level 0 dumps> 1388 <boot into -stable> 1389 make buildworld 1390 make buildkernel KERNCONF=YOUR_KERNEL_HERE 1391 <maybe newfs current's root partition> 1392 <mount current's root partition on directory ${CURRENT_ROOT}> 1393 make installworld DESTDIR=${CURRENT_ROOT} 1394 make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd 1395 make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT} 1396 cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd 1397 <edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition> 1398 <reboot into current> 1399 <do a "native" rebuild/install as described in the previous section> 1400 <maybe install compatibility libraries from ports/misc/compat*> 1401 <reboot> 1402 1403 1404 To upgrade in-place from 8.x-stable to current 1405 ---------------------------------------------- 1406 <make sure you have good level 0 dumps> 1407 make buildworld [9] 1408 make kernel KERNCONF=YOUR_KERNEL_HERE [8] 1409 [1] 1410 <reboot in single user> [3] 1411 mergemaster -p [5] 1412 make installworld 1413 mergemaster -i [4] 1414 make delete-old [6] 1415 <reboot> 1416 1417 Make sure that you've read the UPDATING file to understand the 1418 tweaks to various things you need. At this point in the life 1419 cycle of current, things change often and you are on your own 1420 to cope. The defaults can also change, so please read ALL of 1421 the UPDATING entries. 1422 1423 Also, if you are tracking -current, you must be subscribed to 1424 freebsd-current@freebsd.org. Make sure that before you update 1425 your sources that you have read and understood all the recent 1426 messages there. If in doubt, please track -stable which has 1427 much fewer pitfalls. 1428 1429 [1] If you have third party modules, such as vmware, you 1430 should disable them at this point so they don't crash your 1431 system on reboot. 1432 1433 [3] From the bootblocks, boot -s, and then do 1434 fsck -p 1435 mount -u / 1436 mount -a 1437 cd src 1438 adjkerntz -i # if CMOS is wall time 1439 Also, when doing a major release upgrade, it is required that 1440 you boot into single user mode to do the installworld. 1441 1442 [4] Note: This step is non-optional. Failure to do this step 1443 can result in a significant reduction in the functionality of the 1444 system. Attempting to do it by hand is not recommended and those 1445 that pursue this avenue should read this file carefully, as well 1446 as the archives of freebsd-current and freebsd-hackers mailing lists 1447 for potential gotchas. The -U option is also useful to consider. 1448 See mergemaster(8) for more information. 1449 1450 [5] Usually this step is a noop. However, from time to time 1451 you may need to do this if you get unknown user in the following 1452 step. It never hurts to do it all the time. You may need to 1453 install a new mergemaster (cd src/usr.sbin/mergemaster && make 1454 install) after the buildworld before this step if you last updated 1455 from current before 20020224 or from -stable before 20020408. 1456 1457 [6] This only deletes old files and directories. Old libraries 1458 can be deleted by "make delete-old-libs", but you have to make 1459 sure that no program is using those libraries anymore. 1460 1461 [8] In order to have a kernel that can run the 4.x binaries needed to 1462 do an installworld, you must include the COMPAT_FREEBSD4 option in 1463 your kernel. Failure to do so may leave you with a system that is 1464 hard to boot to recover. A similar kernel option COMPAT_FREEBSD5 is 1465 required to run the 5.x binaries on more recent kernels. And so on 1466 for COMPAT_FREEBSD6 and COMPAT_FREEBSD7. 1467 1468 Make sure that you merge any new devices from GENERIC since the 1469 last time you updated your kernel config file. 1470 1471 [9] When checking out sources, you must include the -P flag to have 1472 cvs prune empty directories. 1473 1474 If CPUTYPE is defined in your /etc/make.conf, make sure to use the 1475 "?=" instead of the "=" assignment operator, so that buildworld can 1476 override the CPUTYPE if it needs to. 1477 1478 MAKEOBJDIRPREFIX must be defined in an environment variable, and 1479 not on the command line, or in /etc/make.conf. buildworld will 1480 warn if it is improperly defined. 1481FORMAT: 1482 1483This file contains a list, in reverse chronological order, of major 1484breakages in tracking -current. It is not guaranteed to be a complete 1485list of such breakages, and only contains entries since October 10, 2007. 1486If you need to see UPDATING entries from before that date, you will need 1487to fetch an UPDATING file from an older FreeBSD release. 1488 1489Copyright information: 1490 1491Copyright 1998-2009 M. Warner Losh. All Rights Reserved. 1492 1493Redistribution, publication, translation and use, with or without 1494modification, in full or in part, in any form or format of this 1495document are permitted without further permission from the author. 1496 1497THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR 1498IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 1499WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 1500DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT, 1501INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 1502(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR 1503SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 1504HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, 1505STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 1506IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 1507POSSIBILITY OF SUCH DAMAGE. 1508 1509Contact Warner Losh if you have any questions about your use of 1510this document. 1511 1512$FreeBSD$ 1513