1999-01-14 05:59:52 +00:00
|
|
|
Updating Information for FreeBSD current users
|
|
|
|
|
2001-06-29 06:00:44 +00:00
|
|
|
This file is maintained and copyrighted by M. Warner Losh
|
2002-04-11 05:45:17 +00:00
|
|
|
<imp@village.org>. See end of file for further details. For commonly
|
|
|
|
done items, please see the COMMON ITEMS: section later in the file.
|
2001-06-29 06:00:44 +00:00
|
|
|
|
2004-07-29 15:06:19 +00:00
|
|
|
Items affecting the ports and packages system can be found in
|
|
|
|
/usr/ports/UPDATING. Please read that file before running
|
2005-03-03 08:44:33 +00:00
|
|
|
portupgrade.
|
2004-07-29 15:06:19 +00:00
|
|
|
|
2007-10-11 04:28:08 +00:00
|
|
|
NOTE TO PEOPLE WHO THINK THAT FreeBSD 8.x IS SLOW:
|
|
|
|
FreeBSD 8.x has many debugging features turned on, in
|
2002-02-28 05:31:02 +00:00
|
|
|
both the kernel and userland. These features attempt to detect
|
|
|
|
incorrect use of system primitives, and encourage loud failure
|
|
|
|
through extra sanity checking and fail stop semantics. They
|
|
|
|
also substantially impact system performance. If you want to
|
|
|
|
do performance measurement, benchmarking, and optimization,
|
|
|
|
you'll want to turn them off. This includes various WITNESS-
|
|
|
|
related kernel options, INVARIANTS, malloc debugging flags
|
|
|
|
in userland, and various verbose features in the kernel. Many
|
|
|
|
developers choose to disable these features on build machines
|
2008-01-22 15:05:22 +00:00
|
|
|
to maximize performance. (To disable malloc debugging, run
|
|
|
|
ln -s aj /etc/malloc.conf.)
|
2008-12-19 23:10:55 +00:00
|
|
|
|
2009-07-19 17:25:24 +00:00
|
|
|
20090719:
|
|
|
|
Bump the shared library version numbers for all libraries that
|
|
|
|
do not use symbol versioning as part of the 8.0-RELEASE cycle.
|
|
|
|
Bump __FreeBSD_version to 800105.
|
|
|
|
|
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator
(DPCPU), as suggested by Peter Wemm, and implement a new per-virtual
network stack memory allocator. Modify vnet to use the allocator
instead of monolithic global container structures (vinet, ...). This
change solves many binary compatibility problems associated with
VIMAGE, and restores ELF symbols for virtualized global variables.
Each virtualized global variable exists as a "reference copy", and also
once per virtual network stack. Virtualized global variables are
tagged at compile-time, placing the in a special linker set, which is
loaded into a contiguous region of kernel memory. Virtualized global
variables in the base kernel are linked as normal, but those in modules
are copied and relocated to a reserved portion of the kernel's vnet
region with the help of a the kernel linker.
Virtualized global variables exist in per-vnet memory set up when the
network stack instance is created, and are initialized statically from
the reference copy. Run-time access occurs via an accessor macro, which
converts from the current vnet and requested symbol to a per-vnet
address. When "options VIMAGE" is not compiled into the kernel, normal
global ELF symbols will be used instead and indirection is avoided.
This change restores static initialization for network stack global
variables, restores support for non-global symbols and types, eliminates
the need for many subsystem constructors, eliminates large per-subsystem
structures that caused many binary compatibility issues both for
monitoring applications (netstat) and kernel modules, removes the
per-function INIT_VNET_*() macros throughout the stack, eliminates the
need for vnet_symmap ksym(2) munging, and eliminates duplicate
definitions of virtualized globals under VIMAGE_GLOBALS.
Bump __FreeBSD_version and update UPDATING.
Portions submitted by: bz
Reviewed by: bz, zec
Discussed with: gnn, jamie, jeff, jhb, julian, sam
Suggested by: peter
Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
|
|
|
20090714:
|
|
|
|
Due to changes in the implementation of virtual network stack
|
|
|
|
support, all network-related kernel modules must be recompiled.
|
|
|
|
As this change breaks the ABI, bump __FreeBSD_version to 800104.
|
|
|
|
|
2009-07-13 11:51:02 +00:00
|
|
|
20090713:
|
|
|
|
The TOE interface to the TCP syncache has been modified to remove struct
|
|
|
|
tcpopt (<netinet/tcp_var.h>) from the ABI of the network stack. The
|
|
|
|
cxgb driver is the only TOE consumer affected by this change, and needs
|
|
|
|
to be recompiled along with the kernel. As this change breaks the ABI,
|
|
|
|
bump __FreeBSD_version to 800103.
|
|
|
|
|
2009-07-12 09:14:28 +00:00
|
|
|
20090712:
|
|
|
|
Padding has been added to struct tcpcb, sackhint and tcpstat in
|
|
|
|
<netinet/tcp_var.h> to facilitate future MFCs and bug fixes whilst
|
|
|
|
maintainig the ABI. However, this change breaks the ABI, so bump
|
|
|
|
__FreeBSD_version to 800102. User space tools that rely on the size of
|
|
|
|
any of these structs (e.g. sockstat) need to be recompiled.
|
|
|
|
|
2009-07-01 07:35:57 +00:00
|
|
|
20090630:
|
2009-07-01 18:12:50 +00:00
|
|
|
The NFS_LEGACYRPC option has been removed along with the old
|
|
|
|
kernel RPC implementation that this option selected. Kernel
|
|
|
|
configurations may need to be adjusted.
|
2009-07-01 07:35:57 +00:00
|
|
|
|
2009-06-29 19:46:29 +00:00
|
|
|
20090629:
|
|
|
|
The network interface device nodes at /dev/net/<interface> have
|
|
|
|
been removed. All ioctl operations can be performed the normal
|
|
|
|
way using routing sockets. The kqueue functionality can
|
|
|
|
generally be replaced with routing sockets.
|
|
|
|
|
2009-06-28 08:59:46 +00:00
|
|
|
20090628:
|
|
|
|
The documentation from the FreeBSD Documentation Project
|
|
|
|
(Handbook, FAQ, etc.) is now installed via packages by
|
|
|
|
sysinstall(8) and under the /usr/local/share/doc/freebsd
|
|
|
|
directory instead of /usr/share/doc.
|
|
|
|
|
2009-06-26 17:50:52 +00:00
|
|
|
20090624:
|
|
|
|
The ABI of various structures related to the SYSV IPC API have
|
|
|
|
been changed. As a result, the COMPAT_FREEBSD[456] kernel
|
|
|
|
options now all require COMPAT_FREEBSD7. Bump __FreeBSD_version
|
|
|
|
to 800100.
|
|
|
|
|
2009-06-22 17:48:16 +00:00
|
|
|
20090622:
|
|
|
|
Layout of struct vnet has changed as routing related variables
|
|
|
|
were moved to their own Vimage module. Modules need to be
|
|
|
|
recompiled. Bump __FreeBSD_version to 800099.
|
|
|
|
|
2009-06-19 17:10:35 +00:00
|
|
|
20090619:
|
|
|
|
NGROUPS_MAX and NGROUPS have been increased from 16 to 1023
|
|
|
|
and 1024 respectively. As long as no more than 16 groups per
|
|
|
|
process are used, no changes should be visible. When more
|
|
|
|
than 16 groups are used, old binaries may fail if they call
|
|
|
|
getgroups() or getgrouplist() with statically sized storage.
|
|
|
|
Recompiling will work around this, but applications should be
|
|
|
|
modified to use dynamically allocated storage for group arrays
|
|
|
|
as POSIX.1-2008 does not cap an implementation's number of
|
|
|
|
supported groups at NGROUPS_MAX+1 as previous versions did.
|
|
|
|
|
|
|
|
NFS and portalfs mounts may also be affected as the list of
|
|
|
|
groups is truncated to 16. Users of NFS who use more than 16
|
|
|
|
groups, should take care that negative group permissions are not
|
|
|
|
used on the exported file systems as they will not be reliable
|
|
|
|
unless a GSSAPI based authentication method is used.
|
|
|
|
|
Introduce support for adaptive spinning in lockmgr.
Actually, as it did receive few tuning, the support is disabled by
default, but it can opt-in with the option ADAPTIVE_LOCKMGRS.
Due to the nature of lockmgrs, adaptive spinning needs to be
selectively enabled for any interested lockmgr.
The support is bi-directional, or, in other ways, it will work in both
cases if the lock is held in read or write way. In particular, the
read path is passible of further tunning using the sysctls
debug.lockmgr.retries and debug.lockmgr.loops . Ideally, such sysctls
should be axed or compiled out before release.
Addictionally note that adaptive spinning doesn't cope well with
LK_SLEEPFAIL. The reason is that many (and probabilly all) consumers
of LK_SLEEPFAIL are mainly interested in knowing if the interlock was
dropped or not in order to reacquire it and re-test initial conditions.
This directly interacts with adaptive spinning because lockmgr needs
to drop the interlock while spinning in order to avoid a deadlock
(further details in the comments inside the patch).
Final note: finding someone willing to help on tuning this with
relevant workloads would be either very important and appreciated.
Tested by: jeff, pho
Requested by: many
2009-06-17 01:55:42 +00:00
|
|
|
20090616:
|
|
|
|
The compiling option ADAPTIVE_LOCKMGRS has been introduced.
|
|
|
|
This option compiles in the support for adaptive spinning for lockmgrs
|
|
|
|
which want to enable it. The lockinit() function now accepts the
|
|
|
|
flag LK_ADAPTIVE in order to make the lock object subject to
|
|
|
|
adaptive spinning when both held in write and read mode.
|
|
|
|
|
2009-06-13 23:44:56 +00:00
|
|
|
20090613:
|
|
|
|
The layout of the structure returned by IEEE80211_IOC_STA_INFO
|
|
|
|
has changed. User applications that use this ioctl need to be
|
|
|
|
rebuilt.
|
|
|
|
|
Introduce a mechanism for detecting calls from outbound path of the
network stack when reentering the inbound path from netgraph, and
force queueing of mbufs at the outbound netgraph node.
The mechanism relies on two components. First, in netgraph nodes
where outbound path of the network stack calls into netgraph, the
current thread has to be appropriately marked using the new
NG_OUTBOUND_THREAD_REF() macro before proceeding to call further
into the netgraph topology, and unmarked using the
NG_OUTBOUND_THREAD_UNREF() macro before returning to the caller.
Second, netgraph nodes which can potentially reenter the network
stack in the inbound path have to mark their inbound hooks using
NG_HOOK_SET_TO_INBOUND() macro. The netgraph framework will then
detect when there is a danger of a call graph looping back from
outbound to inbound path via netgraph, and defer handing off the
mbufs to the "inbound" node to a worker thread with a clean stack.
In this first pass only the most obvious netgraph nodes have been
updated to ensure no outbound to inbound calls can occur. Nodes
such as ng_ipfw, ng_gif etc. should be further examined whether a
potential for outbound to inbound call looping exists.
This commit changes the layout of struct thread, but due to
__FreeBSD_version number shortage a version bump has been omitted
at this time, nevertheless kernel and modules have to be rebuilt.
Reviewed by: julian, rwatson, bz
Approved by: julian (mentor)
2009-06-11 16:50:49 +00:00
|
|
|
20090611:
|
|
|
|
The layout of struct thread has changed. Kernel and modules
|
|
|
|
need to be rebuilt.
|
|
|
|
|
Introduce an infrastructure for dismantling vnet instances.
Vnet modules and protocol domains may now register destructor
functions to clean up and release per-module state. The destructor
mechanisms can be triggered by invoking "vimage -d", or a future
equivalent command which will be provided via the new jail framework.
While this patch introduces numerous placeholder destructor functions,
many of those are currently incomplete, thus leaking memory or (even
worse) failing to stop all running timers. Many of such issues are
already known and will be incrementaly fixed over the next weeks in
smaller incremental commits.
Apart from introducing new fields in structs ifnet, domain, protosw
and vnet_net, which requires the kernel and modules to be rebuilt, this
change should have no impact on nooptions VIMAGE builds, since vnet
destructors can only be called in VIMAGE kernels. Moreover,
destructor functions should be in general compiled in only in
options VIMAGE builds, except for kernel modules which can be safely
kldunloaded at run time.
Bump __FreeBSD_version to 800097.
Reviewed by: bz, julian
Approved by: rwatson, kib (re), julian (mentor)
2009-06-08 17:15:40 +00:00
|
|
|
20090608:
|
|
|
|
The layout of structs ifnet, domain, protosw and vnet_net has
|
|
|
|
changed. Kernel modules need to be rebuilt.
|
|
|
|
Bump __FreeBSD_version to 800097.
|
|
|
|
|
2009-06-02 13:44:36 +00:00
|
|
|
20090602:
|
|
|
|
window(1) has been removed from the base system. It can now be
|
|
|
|
installed from ports. The port is called misc/window.
|
|
|
|
|
2009-06-01 15:49:42 +00:00
|
|
|
20090601:
|
2009-06-01 16:00:36 +00:00
|
|
|
The way we are storing and accessing `routing table' entries
|
2009-06-01 15:49:42 +00:00
|
|
|
has changed. Programs reading the FIB, like netstat, need to
|
|
|
|
be re-compiled.
|
|
|
|
|
2009-06-01 16:00:36 +00:00
|
|
|
20090601:
|
|
|
|
A new netisr implementation has been added for FreeBSD 8. Network
|
|
|
|
file system modules, such as igmp, ipdivert, and others, should be
|
|
|
|
rebuilt.
|
|
|
|
Bump __FreeBSD_version to 800096.
|
|
|
|
|
2009-05-30 23:52:23 +00:00
|
|
|
20090530:
|
|
|
|
Remove the tunable/sysctl debug.mpsafevfs as its initial purpose
|
|
|
|
is no more valid.
|
|
|
|
|
2009-05-30 14:01:01 +00:00
|
|
|
20090530:
|
|
|
|
Add VOP_ACCESSX(9). File system modules need to be rebuilt.
|
|
|
|
Bump __FreeBSD_version to 800094.
|
|
|
|
|
2009-05-29 18:50:27 +00:00
|
|
|
20090529:
|
2009-05-29 19:45:39 +00:00
|
|
|
Add mnt_xflag field to 'struct mount'. File system modules
|
2009-05-29 18:50:27 +00:00
|
|
|
need to be rebuilt.
|
|
|
|
Bump __FreeBSD_version to 800093.
|
|
|
|
|
2009-05-29 01:49:27 +00:00
|
|
|
20090528:
|
|
|
|
The compiling option ADAPTIVE_SX has been retired while it has been
|
|
|
|
introduced the option NO_ADAPTIVE_SX which handles the reversed logic.
|
|
|
|
The KPI for sx_init_flags() changes as accepting flags:
|
|
|
|
SX_ADAPTIVESPIN flag has been retired while the SX_NOADAPTIVE flag
|
|
|
|
has been introduced in order to handle the reversed logic.
|
|
|
|
Bump __FreeBSD_version to 800092.
|
|
|
|
|
2009-05-27 14:11:23 +00:00
|
|
|
20090527:
|
|
|
|
Add support for hierarchical jails. Remove global securelevel.
|
|
|
|
Bump __FreeBSD_version to 800091.
|
|
|
|
|
2009-05-23 21:43:44 +00:00
|
|
|
20090523:
|
|
|
|
The layout of struct vnet_net has changed, therefore modules
|
|
|
|
need to be rebuilt.
|
|
|
|
Bump __FreeBSD_version to 800090.
|
|
|
|
|
2009-05-23 06:31:50 +00:00
|
|
|
20090523:
|
|
|
|
The newly imported zic(8) produces a new format in the
|
2009-05-23 08:49:55 +00:00
|
|
|
output. Please run tzsetup(8) to install the newly created
|
|
|
|
data to /etc/localtime.
|
2009-05-23 06:31:50 +00:00
|
|
|
|
2009-05-21 01:48:42 +00:00
|
|
|
20090520:
|
|
|
|
The sysctl tree for the usb stack has renamed from hw.usb2.* to
|
|
|
|
hw.usb.* and is now consistent again with previous releases.
|
|
|
|
|
2009-05-20 20:05:56 +00:00
|
|
|
20090520:
|
|
|
|
802.11 monitor mode support was revised and driver api's
|
|
|
|
were changed. Drivers dependent on net80211 now support
|
|
|
|
DLT_IEEE802_11_RADIO instead of DLT_IEEE802_11. No
|
|
|
|
user-visible data structures were changed but applications
|
|
|
|
that use DLT_IEEE802_11 may require changes.
|
|
|
|
Bump __FreeBSD_version to 800088.
|
|
|
|
|
Permit buiding kernels with options VIMAGE, restricted to only a single
active network stack instance. Turning on options VIMAGE at compile
time yields the following changes relative to default kernel build:
1) V_ accessor macros for virtualized variables resolve to structure
fields via base pointers, instead of being resolved as fields in global
structs or plain global variables. As an example, V_ifnet becomes:
options VIMAGE: ((struct vnet_net *) vnet_net)->_ifnet
default build: vnet_net_0._ifnet
options VIMAGE_GLOBALS: ifnet
2) INIT_VNET_* macros will declare and set up base pointers to be used
by V_ accessor macros, instead of resolving to whitespace:
INIT_VNET_NET(ifp->if_vnet); becomes
struct vnet_net *vnet_net = (ifp->if_vnet)->mod_data[VNET_MOD_NET];
3) Memory for vnet modules registered via vnet_mod_register() is now
allocated at run time in sys/kern/kern_vimage.c, instead of per vnet
module structs being declared as globals. If required, vnet modules
can now request the framework to provide them with allocated bzeroed
memory by filling in the vmi_size field in their vmi_modinfo structures.
4) structs socket, ifnet, inpcbinfo, tcpcb and syncache_head are
extended to hold a pointer to the parent vnet. options VIMAGE builds
will fill in those fields as required.
5) curvnet is introduced as a new global variable in options VIMAGE
builds, always pointing to the default and only struct vnet.
6) struct sysctl_oid has been extended with additional two fields to
store major and minor virtualization module identifiers, oid_v_subs and
oid_v_mod. SYSCTL_V_* family of macros will fill in those fields
accordingly, and store the offset in the appropriate vnet container
struct in oid_arg1.
In sysctl handlers dealing with virtualized sysctls, the
SYSCTL_RESOLVE_V_ARG1() macro will compute the address of the target
variable and make it available in arg1 variable for further processing.
Unused fields in structs vnet_inet, vnet_inet6 and vnet_ipfw have
been deleted.
Reviewed by: bz, rwatson
Approved by: julian (mentor)
2009-04-30 13:36:26 +00:00
|
|
|
20090430:
|
|
|
|
The layout of the following structs has changed: sysctl_oid,
|
|
|
|
socket, ifnet, inpcbinfo, tcpcb, syncache_head, vnet_inet,
|
|
|
|
vnet_inet6 and vnet_ipfw. Most modules need to be rebuild or
|
|
|
|
panics may be experienced. World rebuild is required for
|
|
|
|
correctly checking networking state from userland.
|
|
|
|
Bump __FreeBSD_version to 800085.
|
|
|
|
|
Bite the bullet, and make the IPv6 SSM and MLDv2 mega-commit:
import from p4 bms_netdev. Summary of changes:
* Connect netinet6/in6_mcast.c to build.
The legacy KAME KPIs are mostly preserved.
* Eliminate now dead code from ip6_output.c.
Don't do mbuf bingo, we are not going to do RFC 2292 style
CMSG tricks for multicast options as they are not required
by any current IPv6 normative reference.
* Refactor transports (UDP, raw_ip6) to do own mcast filtering.
SCTP, TCP unaffected by this change.
* Add ip6_msource, in6_msource structs to in6_var.h.
* Hookup mld_ifinfo state to in6_ifextra, allocate from
domifattach path.
* Eliminate IN6_LOOKUP_MULTI(), it is no longer referenced.
Kernel consumers which need this should use in6m_lookup().
* Refactor IPv6 socket group memberships to use a vector (like IPv4).
* Update ifmcstat(8) for IPv6 SSM.
* Add witness lock order for IN6_MULTI_LOCK.
* Move IN6_MULTI_LOCK out of lower ip6_output()/ip6_input() paths.
* Introduce IP6STAT_ADD/SUB/INC/DEC as per rwatson's IPv4 cleanup.
* Update carp(4) for new IPv6 SSM KPIs.
* Virtualize ip6_mrouter socket.
Changes mostly localized to IPv6 MROUTING.
* Don't do a local group lookup in MROUTING.
* Kill unused KAME prototypes in6_purgemkludge(), in6_restoremkludge().
* Preserve KAME DAD timer jitter behaviour in MLDv1 compatibility mode.
* Bump __FreeBSD_version to 800084.
* Update UPDATING.
NOTE WELL:
* This code hasn't been tested against real MLDv2 queriers
(yet), although the on-wire protocol has been verified in Wireshark.
* There are a few unresolved issues in the socket layer APIs to
do with scope ID propagation.
* There is a LOR present in ip6_output()'s use of
in6_setscope() which needs to be resolved. See comments in mld6.c.
This is believed to be benign and can't be avoided for the moment
without re-introducing an indirect netisr.
This work was mostly derived from the IGMPv3 implementation, and
has been sponsored by a third party.
2009-04-29 19:19:13 +00:00
|
|
|
20090429:
|
|
|
|
MLDv2 and Source-Specific Multicast (SSM) have been merged
|
|
|
|
to the IPv6 stack. VIMAGE hooks are in but not yet used.
|
|
|
|
The implementation of SSM within FreeBSD's IPv6 stack closely
|
|
|
|
follows the IPv4 implementation.
|
|
|
|
|
|
|
|
For kernel developers:
|
|
|
|
|
|
|
|
* The most important changes are that the ip6_output() and
|
|
|
|
ip6_input() paths no longer take the IN6_MULTI_LOCK,
|
|
|
|
and this lock has been downgraded to a non-recursive mutex.
|
|
|
|
|
|
|
|
* As with the changes to the IPv4 stack to support SSM, filtering
|
|
|
|
of inbound multicast traffic must now be performed by transport
|
|
|
|
protocols within the IPv6 stack. This does not apply to TCP and
|
|
|
|
SCTP, however, it does apply to UDP in IPv6 and raw IPv6.
|
|
|
|
|
|
|
|
* The KPIs used by IPv6 multicast are similar to those used by
|
|
|
|
the IPv4 stack, with the following differences:
|
|
|
|
* im6o_mc_filter() is analogous to imo_multicast_filter().
|
|
|
|
* The legacy KAME entry points in6_joingroup and in6_leavegroup()
|
|
|
|
are shimmed to in6_mc_join() and in6_mc_leave() respectively.
|
|
|
|
* IN6_LOOKUP_MULTI() has been deprecated and removed.
|
|
|
|
* IPv6 relies on MLD for the DAD mechanism. KAME's internal KPIs
|
|
|
|
for MLDv1 have an additional 'timer' argument which is used to
|
|
|
|
jitter the initial membership report for the solicited-node
|
|
|
|
multicast membership on-link.
|
|
|
|
* This is not strictly needed for MLDv2, which already jitters
|
|
|
|
its report transmissions. However, the 'timer' argument is
|
|
|
|
preserved in case MLDv1 is active on the interface.
|
|
|
|
|
|
|
|
* The KAME linked-list based IPv6 membership implementation has
|
|
|
|
been refactored to use a vector similar to that used by the IPv4
|
|
|
|
stack.
|
|
|
|
Code which maintains a list of its own multicast memberships
|
|
|
|
internally, e.g. carp, has been updated to reflect the new
|
|
|
|
semantics.
|
|
|
|
|
|
|
|
* There is a known Lock Order Reversal (LOR) due to in6_setscope()
|
|
|
|
acquiring the IF_AFDATA_LOCK and being called within ip6_output().
|
|
|
|
Whilst MLDv2 tries to avoid this otherwise benign LOR, it is an
|
|
|
|
implementation constraint which needs to be addressed in HEAD.
|
|
|
|
|
|
|
|
For application developers:
|
|
|
|
|
|
|
|
* The changes are broadly similar to those made for the IPv4
|
|
|
|
stack.
|
|
|
|
|
|
|
|
* The use of IPv4 and IPv6 multicast socket options on the same
|
|
|
|
socket, using mapped addresses, HAS NOT been tested or supported.
|
|
|
|
|
|
|
|
* There are a number of issues with the implementation of various
|
|
|
|
IPv6 multicast APIs which need to be resolved in the API surface
|
|
|
|
before the implementation is fully compatible with KAME userland
|
|
|
|
use, and these are mostly to do with interface index treatment.
|
|
|
|
|
|
|
|
* The literature available discusses the use of either the delta / ASM
|
|
|
|
API with setsockopt(2)/getsockopt(2), or the full-state / ASM API
|
|
|
|
using setsourcefilter(3)/getsourcefilter(3). For more information
|
|
|
|
please refer to RFC 3768, 'Socket Interface Extensions for
|
|
|
|
Multicast Source Filters'.
|
|
|
|
|
|
|
|
* Applications which use the published RFC 3678 APIs should be fine.
|
|
|
|
|
|
|
|
For systems administrators:
|
|
|
|
|
|
|
|
* The mtest(8) utility has been refactored to support IPv6, in
|
|
|
|
addition to IPv4. Interface addresses are no longer accepted
|
|
|
|
as arguments, their names must be used instead. The utility
|
|
|
|
will map the interface name to its first IPv4 address as
|
|
|
|
returned by getifaddrs(3).
|
|
|
|
|
|
|
|
* The ifmcstat(8) utility has also been updated to print the MLDv2
|
|
|
|
endpoint state and source filter lists via sysctl(3).
|
|
|
|
|
|
|
|
* The net.inet6.ip6.mcast.loop sysctl may be tuned to 0 to disable
|
|
|
|
loopback of IPv6 multicast datagrams by default; it defaults to 1
|
|
|
|
to preserve the existing behaviour. Disabling multicast loopback is
|
|
|
|
recommended for optimal system performance.
|
|
|
|
|
|
|
|
* The IPv6 MROUTING code has been changed to examine this sysctl
|
|
|
|
instead of attempting to perform a group lookup before looping
|
|
|
|
back forwarded datagrams.
|
|
|
|
|
|
|
|
Bump __FreeBSD_version to 800084.
|
|
|
|
|
2009-04-24 09:58:50 +00:00
|
|
|
20090422:
|
2009-04-22 15:54:27 +00:00
|
|
|
Implement low-level Bluetooth HCI API.
|
|
|
|
Bump __FreeBSD_version to 800083.
|
|
|
|
|
2009-04-24 09:58:50 +00:00
|
|
|
20090419:
|
|
|
|
The layout of struct malloc_type, used by modules to register new
|
|
|
|
memory allocation types, has changed. Most modules will need to
|
|
|
|
be rebuilt or panics may be experienced.
|
|
|
|
Bump __FreeBSD_version to 800081.
|
|
|
|
|
2009-04-15 22:09:42 +00:00
|
|
|
20090415:
|
2009-04-24 15:38:13 +00:00
|
|
|
Anticipate overflowing inp_flags - add inp_flags2.
|
2009-04-15 22:09:42 +00:00
|
|
|
This changes most offsets in inpcb, so checking v4 connection
|
2009-04-24 15:38:13 +00:00
|
|
|
state will require a world rebuild.
|
2009-04-15 22:09:42 +00:00
|
|
|
Bump __FreeBSD_version to 800080.
|
2009-04-24 15:38:13 +00:00
|
|
|
|
2009-04-15 20:34:19 +00:00
|
|
|
20090415:
|
|
|
|
Add an llentry to struct route and struct route_in6. Modules
|
2009-04-24 15:38:13 +00:00
|
|
|
embedding a struct route will need to be recompiled.
|
2009-04-15 20:34:19 +00:00
|
|
|
Bump __FreeBSD_version to 800079.
|
|
|
|
|
2009-04-14 23:05:36 +00:00
|
|
|
20090414:
|
|
|
|
The size of rt_metrics_lite and by extension rtentry has changed.
|
|
|
|
Networking administration apps will need to be recompiled.
|
2009-04-24 15:38:13 +00:00
|
|
|
The route command now supports show as an alias for get, weighting
|
2009-04-14 23:05:36 +00:00
|
|
|
of routes, sticky and nostick flags to alter the behavior of stateful
|
|
|
|
load balancing.
|
|
|
|
Bump __FreeBSD_version to 800078.
|
2009-04-24 11:34:59 +00:00
|
|
|
|
2009-04-12 19:42:25 +00:00
|
|
|
20090408:
|
|
|
|
Do not use Giant for kbdmux(4) locking. This is wrong and
|
|
|
|
apparently causing more problems than it solves. This will
|
|
|
|
re-open the issue where interrupt handlers may race with
|
|
|
|
kbdmux(4) in polling mode. Typical symptoms include (but
|
|
|
|
not limited to) duplicated and/or missing characters when
|
|
|
|
low level console functions (such as gets) are used while
|
|
|
|
interrupts are enabled (for example geli password prompt,
|
|
|
|
mountroot prompt etc.). Disabling kbdmux(4) may help.
|
|
|
|
|
2009-04-08 18:30:42 +00:00
|
|
|
20090407:
|
|
|
|
The size of structs vnet_net, vnet_inet and vnet_ipfw has changed;
|
|
|
|
kernel modules referencing any of the above need to be recompiled.
|
|
|
|
Bump __FreeBSD_version to 800075.
|
|
|
|
|
2009-03-20 21:51:27 +00:00
|
|
|
20090320:
|
|
|
|
GEOM_PART has become the default partition slicer for storage devices,
|
|
|
|
replacing GEOM_MBR, GEOM_BSD, GEOM_PC98 and GEOM_GPT slicers. It
|
2009-03-20 23:13:32 +00:00
|
|
|
introduces some changes:
|
2009-04-24 15:38:13 +00:00
|
|
|
|
2009-03-20 23:13:32 +00:00
|
|
|
MSDOS/EBR: the devices created from MSDOS extended partition entries
|
|
|
|
(EBR) can be named differently than with GEOM_MBR and are now symlinks
|
|
|
|
to devices with offset-based names. fstabs may need to be modified.
|
|
|
|
|
|
|
|
BSD: the "geometry does not match label" warning is harmless in most
|
|
|
|
cases but it points to problems in file system misalignment with
|
|
|
|
disk geometry. The "c" partition is now implicit, covers the whole
|
|
|
|
top-level drive and cannot be (mis)used by users.
|
2009-04-24 15:38:13 +00:00
|
|
|
|
2009-03-20 23:13:32 +00:00
|
|
|
General: Kernel dumps are now not allowed to be written to devices
|
|
|
|
whose partition types indicate they are meant to be used for file
|
|
|
|
systems (or, in case of MSDOS partitions, as something else than
|
|
|
|
the "386BSD" type).
|
|
|
|
|
|
|
|
Most of these changes date approximately from 200812.
|
2009-03-20 21:51:27 +00:00
|
|
|
|
2009-03-19 20:33:26 +00:00
|
|
|
20090319:
|
|
|
|
The uscanner(4) driver has been removed from the kernel. This follows
|
|
|
|
Linux removing theirs in 2.6 and making libusb the default interface
|
|
|
|
(supported by sane).
|
|
|
|
|
2009-03-19 01:43:03 +00:00
|
|
|
20090319:
|
|
|
|
The multicast forwarding code has been cleaned up. netstat(1)
|
|
|
|
only relies on KVM now for printing bandwidth upcall meters.
|
|
|
|
The IPv4 and IPv6 modules are split into ip_mroute_mod and
|
|
|
|
ip6_mroute_mod respectively. The config(5) options for statically
|
|
|
|
compiling this code remain the same, i.e. 'options MROUTING'.
|
|
|
|
|
2009-03-15 16:12:50 +00:00
|
|
|
20090315:
|
|
|
|
Support for the IFF_NEEDSGIANT network interface flag has been
|
|
|
|
removed, which means that non-MPSAFE network device drivers are no
|
|
|
|
longer supported. In particular, if_ar, if_sr, and network device
|
|
|
|
drivers from the old (legacy) USB stack can no longer be built or
|
|
|
|
used.
|
|
|
|
|
2009-03-13 16:40:56 +00:00
|
|
|
20090313:
|
|
|
|
POSIX.1 Native Language Support (NLS) has been enabled in libc and
|
|
|
|
a bunch of new language catalog files have also been added.
|
|
|
|
This means that some common libc messages are now localized and
|
|
|
|
they depend on the LC_MESSAGES environmental variable.
|
|
|
|
|
2009-03-13 16:30:33 +00:00
|
|
|
20090313:
|
|
|
|
The k8temp(4) driver has been renamed to amdtemp(4) since
|
|
|
|
support for K10 and K11 CPU families was added.
|
|
|
|
|
2009-03-09 17:53:05 +00:00
|
|
|
20090309:
|
|
|
|
IGMPv3 and Source-Specific Multicast (SSM) have been merged
|
|
|
|
to the IPv4 stack. VIMAGE hooks are in but not yet used.
|
|
|
|
|
|
|
|
For kernel developers, the most important changes are that the
|
|
|
|
ip_output() and ip_input() paths no longer take the IN_MULTI_LOCK(),
|
|
|
|
and this lock has been downgraded to a non-recursive mutex.
|
|
|
|
|
|
|
|
Transport protocols (UDP, Raw IP) are now responsible for filtering
|
|
|
|
inbound multicast traffic according to group membership and source
|
|
|
|
filters. The imo_multicast_filter() KPI exists for this purpose.
|
|
|
|
Transports which do not use multicast (SCTP, TCP) already reject
|
|
|
|
multicast by default. Forwarding and receive performance may improve
|
|
|
|
as a mutex acquisition is no longer needed in the ip_input()
|
|
|
|
low-level input path. in_addmulti() and in_delmulti() are shimmed
|
|
|
|
to new KPIs which exist to support SSM in-kernel.
|
|
|
|
|
|
|
|
For application developers, it is recommended that loopback of
|
|
|
|
multicast datagrams be disabled for best performance, as this
|
|
|
|
will still cause the lock to be taken for each looped-back
|
|
|
|
datagram transmission. The net.inet.ip.mcast.loop sysctl may
|
|
|
|
be tuned to 0 to disable loopback by default; it defaults to 1
|
|
|
|
to preserve the existing behaviour.
|
|
|
|
|
|
|
|
For systems administrators, to obtain best performance with
|
|
|
|
multicast reception and multiple groups, it is always recommended
|
|
|
|
that a card with a suitably precise hash filter is used. Hash
|
|
|
|
collisions will still result in the lock being taken within the
|
|
|
|
transport protocol input path to check group membership.
|
|
|
|
|
|
|
|
If deploying FreeBSD in an environment with IGMP snooping switches,
|
|
|
|
it is recommended that the net.inet.igmp.sendlocal sysctl remain
|
|
|
|
enabled; this forces 224.0.0.0/24 group membership to be announced
|
|
|
|
via IGMP.
|
|
|
|
|
|
|
|
The size of 'struct igmpstat' has changed; netstat needs to be
|
|
|
|
recompiled to reflect this.
|
|
|
|
Bump __FreeBSD_version to 800070.
|
|
|
|
|
2009-03-09 17:05:31 +00:00
|
|
|
20090309:
|
|
|
|
libusb20.so.1 is now installed as libusb.so.1 and the ports system
|
|
|
|
updated to use it. This requires a buildworld/installworld in order to
|
|
|
|
update the library and dependencies (usbconfig, etc). Its advisable to
|
2009-03-09 19:56:37 +00:00
|
|
|
rebuild all ports which uses libusb. More specific directions are given
|
2009-03-09 22:42:01 +00:00
|
|
|
in the ports collection UPDATING file. Any /etc/libmap.conf entries for
|
2009-03-09 22:43:00 +00:00
|
|
|
libusb are no longer required and can be removed.
|
2009-03-09 17:05:31 +00:00
|
|
|
|
2009-03-02 18:53:30 +00:00
|
|
|
20090302:
|
2009-03-05 12:04:42 +00:00
|
|
|
A workaround is committed to allow the creation of System V shared
|
|
|
|
memory segment of size > 2 GB on the 64-bit architectures.
|
|
|
|
Due to a limitation of the existing ABI, the shm_segsz member
|
2009-03-02 18:53:30 +00:00
|
|
|
of the struct shmid_ds, returned by shmctl(IPC_STAT) call is
|
2009-05-23 09:24:07 +00:00
|
|
|
wrong for large segments. Note that limits must be explicitly
|
2009-03-02 18:53:30 +00:00
|
|
|
raised to allow such segments to be created.
|
|
|
|
|
2009-03-01 12:44:33 +00:00
|
|
|
20090301:
|
|
|
|
The layout of struct ifnet has changed, requiring a rebuild of all
|
|
|
|
network device driver modules.
|
|
|
|
|
2009-02-27 17:32:49 +00:00
|
|
|
20090227:
|
|
|
|
The /dev handling for the new USB stack has changed, a
|
|
|
|
buildworld/installworld is required for libusb20.
|
|
|
|
|
2009-02-23 18:56:49 +00:00
|
|
|
20090223:
|
|
|
|
The new USB2 stack has now been permanently moved in and all kernel and
|
2009-02-23 19:30:00 +00:00
|
|
|
module names reverted to their previous values (eg, usb, ehci, ohci,
|
2009-02-23 19:28:29 +00:00
|
|
|
ums, ...). The old usb stack can be compiled in by prefixing the name
|
2009-02-23 18:56:49 +00:00
|
|
|
with the letter 'o', the old usb modules have been removed.
|
2009-03-01 18:57:59 +00:00
|
|
|
Updating entry 20090216 for xorg and 20090215 for libmap may still
|
|
|
|
apply.
|
2009-02-23 18:56:49 +00:00
|
|
|
|
2009-02-17 11:55:50 +00:00
|
|
|
20090217:
|
|
|
|
The rc.conf(5) option if_up_delay has been renamed to
|
|
|
|
defaultroute_delay to better reflect its purpose. If you have
|
|
|
|
customized this setting in /etc/rc.conf you need to update it to
|
|
|
|
use the new name.
|
|
|
|
|
2009-02-16 18:59:18 +00:00
|
|
|
20090216:
|
|
|
|
xorg 7.4 wants to configure its input devices via hald which does not
|
|
|
|
yet work with USB2. If the keyboard/mouse does not work in xorg then
|
|
|
|
add
|
|
|
|
Option "AllowEmptyInput" "off"
|
|
|
|
to your ServerLayout section. This will cause X to use the configured
|
2009-02-17 10:49:36 +00:00
|
|
|
kbd and mouse sections from your xorg.conf.
|
2009-02-16 18:59:18 +00:00
|
|
|
|
2009-02-15 22:33:44 +00:00
|
|
|
20090215:
|
|
|
|
The GENERIC kernels for all architectures now default to the new USB2
|
|
|
|
stack. No kernel config options or code have been removed so if a
|
|
|
|
problem arises please report it and optionally revert to the old USB
|
|
|
|
stack. If you are loading USB kernel modules or have a custom kernel
|
|
|
|
that includes GENERIC then ensure that usb names are also changed over,
|
2009-02-17 10:49:36 +00:00
|
|
|
eg uftdi -> usb2_serial_ftdi.
|
2009-02-15 22:33:44 +00:00
|
|
|
|
2009-03-01 18:57:59 +00:00
|
|
|
Older programs linked against the ports libusb 0.1 need to be
|
|
|
|
redirected to the new stack's libusb20. /etc/libmap.conf can
|
|
|
|
be used for this:
|
|
|
|
# Map old usb library to new one for usb2 stack
|
|
|
|
libusb-0.1.so.8 libusb20.so.1
|
|
|
|
|
2009-02-06 15:03:14 +00:00
|
|
|
20090203:
|
|
|
|
The ichsmb(4) driver has been changed to require SMBus slave
|
|
|
|
addresses be left-justified (xxxxxxx0b) rather than right-justified.
|
|
|
|
All of the other SMBus controller drivers require left-justified
|
|
|
|
slave addresses, so this change makes all the drivers provide the
|
|
|
|
same interface.
|
|
|
|
|
2009-02-01 21:11:08 +00:00
|
|
|
20090201:
|
|
|
|
INET6 statistics (struct ip6stat) was updated.
|
|
|
|
netstat(1) needs to be recompiled.
|
|
|
|
|
2009-01-19 17:00:42 +00:00
|
|
|
20090119:
|
|
|
|
NTFS has been removed from GENERIC kernel on amd64 to match
|
|
|
|
GENERIC on i386. Should not cause any issues since mount_ntfs(8)
|
|
|
|
will load ntfs.ko module automatically when NTFS support is
|
|
|
|
actually needed, unless ntfs.ko is not installed or security
|
|
|
|
level prohibits loading kernel modules. If either is the case,
|
|
|
|
"options NTFS" has to be added into kernel config.
|
|
|
|
|
2009-01-15 06:44:22 +00:00
|
|
|
20090115:
|
|
|
|
TCP Appropriate Byte Counting (RFC 3465) support added to kernel.
|
|
|
|
New field in struct tcpcb breaks ABI, so bump __FreeBSD_version to
|
|
|
|
800061. User space tools that rely on the size of struct tcpcb in
|
|
|
|
tcp_var.h (e.g. sockstat) need to be recompiled.
|
|
|
|
|
2008-12-25 10:05:00 +00:00
|
|
|
20081225:
|
|
|
|
ng_tty(4) module updated to match the new TTY subsystem.
|
|
|
|
Due to API change, user-level applications must be updated.
|
|
|
|
New API support added to mpd5 CVS and expected to be present
|
|
|
|
in next mpd5.3 release.
|
|
|
|
|
2008-12-19 23:10:55 +00:00
|
|
|
20081219:
|
2008-12-19 23:12:14 +00:00
|
|
|
With __FreeBSD_version 800060 the makefs tool is part of
|
|
|
|
the base system (it was a port).
|
2008-12-19 23:10:55 +00:00
|
|
|
|
2008-12-17 00:11:56 +00:00
|
|
|
20081216:
|
2009-02-17 10:50:18 +00:00
|
|
|
The afdata and ifnet locks have been changed from mutexes to
|
2008-12-17 00:11:56 +00:00
|
|
|
rwlocks, network modules will need to be re-compiled.
|
2006-11-15 20:02:20 +00:00
|
|
|
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
20081214:
|
|
|
|
__FreeBSD_version 800059 incorporates the new arp-v2 rewrite.
|
|
|
|
RTF_CLONING, RTF_LLINFO and RTF_WASCLONED flags are eliminated.
|
2009-02-17 10:50:18 +00:00
|
|
|
The new code reduced struct rtentry{} by 16 bytes on 32-bit
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
architecture and 40 bytes on 64-bit architecture. The userland
|
|
|
|
applications "arp" and "ndp" have been updated accordingly.
|
|
|
|
The output from "netstat -r" shows only routing entries and
|
|
|
|
none of the L2 information.
|
|
|
|
|
2008-12-01 16:53:01 +00:00
|
|
|
20081130:
|
|
|
|
__FreeBSD_version 800057 marks the switchover from the
|
|
|
|
binary ath hal to source code. Users must add the line:
|
|
|
|
|
2008-12-01 23:09:58 +00:00
|
|
|
options AH_SUPPORT_AR5416
|
2008-12-01 16:53:01 +00:00
|
|
|
|
|
|
|
to their kernel config files when specifying:
|
|
|
|
|
|
|
|
device ath_hal
|
|
|
|
|
|
|
|
The ath_hal module no longer exists; the code is now compiled
|
|
|
|
together with the driver in the ath module. It is now
|
|
|
|
possible to tailor chip support (i.e. reduce the set of chips
|
|
|
|
and thereby the code size); consult ath_hal(4) for details.
|
|
|
|
|
2008-11-22 05:55:56 +00:00
|
|
|
20081121:
|
2008-12-08 17:12:40 +00:00
|
|
|
__FreeBSD_version 800054 adds memory barriers to
|
|
|
|
<machine/atomic.h>, new interfaces to ifnet to facilitate
|
2008-11-22 05:55:56 +00:00
|
|
|
multiple hardware transmit queues for cards that support
|
|
|
|
them, and a lock-less ring-buffer implementation to
|
2008-12-08 17:12:40 +00:00
|
|
|
enable drivers to more efficiently manage queueing of
|
2008-11-22 05:55:56 +00:00
|
|
|
packets.
|
|
|
|
|
2008-11-18 21:41:09 +00:00
|
|
|
20081117:
|
|
|
|
A new version of ZFS (version 13) has been merged to -HEAD.
|
2008-11-19 00:25:15 +00:00
|
|
|
This version has zpool attribute "listsnapshots" off by
|
|
|
|
default, which means "zfs list" does not show snapshots,
|
2008-11-18 21:41:09 +00:00
|
|
|
and is the same as Solaris behavior.
|
|
|
|
|
2008-10-28 14:14:57 +00:00
|
|
|
20081028:
|
|
|
|
dummynet(4) ABI has changed. ipfw(8) needs to be recompiled.
|
|
|
|
|
2008-10-10 06:37:51 +00:00
|
|
|
20081009:
|
|
|
|
The uhci, ohci, ehci and slhci USB Host controller drivers have
|
|
|
|
been put into separate modules. If you load the usb module
|
|
|
|
separately through loader.conf you will need to load the
|
|
|
|
appropriate *hci module as well. E.g. for a UHCI-based USB 2.0
|
|
|
|
controller add the following to loader.conf:
|
|
|
|
|
|
|
|
uhci_load="YES"
|
|
|
|
ehci_load="YES"
|
|
|
|
|
2008-10-10 04:23:40 +00:00
|
|
|
20081009:
|
|
|
|
The ABI used by the PMC toolset has changed. Please keep
|
|
|
|
userland (libpmc(3)) and the kernel module (hwpmc(4)) in
|
|
|
|
sync.
|
|
|
|
|
Integrate the new MPSAFE TTY layer to the FreeBSD operating system.
The last half year I've been working on a replacement TTY layer for the
FreeBSD kernel. The new TTY layer was designed to improve the following:
- Improved driver model:
The old TTY layer has a driver model that is not abstract enough to
make it friendly to use. A good example is the output path, where the
device drivers directly access the output buffers. This means that an
in-kernel PPP implementation must always convert network buffers into
TTY buffers.
If a PPP implementation would be built on top of the new TTY layer
(still needs a hooks layer, though), it would allow the PPP
implementation to directly hand the data to the TTY driver.
- Improved hotplugging:
With the old TTY layer, it isn't entirely safe to destroy TTY's from
the system. This implementation has a two-step destructing design,
where the driver first abandons the TTY. After all threads have left
the TTY, the TTY layer calls a routine in the driver, which can be
used to free resources (unit numbers, etc).
The pts(4) driver also implements this feature, which means
posix_openpt() will now return PTY's that are created on the fly.
- Improved performance:
One of the major improvements is the per-TTY mutex, which is expected
to improve scalability when compared to the old Giant locking.
Another change is the unbuffered copying to userspace, which is both
used on TTY device nodes and PTY masters.
Upgrading should be quite straightforward. Unlike previous versions,
existing kernel configuration files do not need to be changed, except
when they reference device drivers that are listed in UPDATING.
Obtained from: //depot/projects/mpsafetty/...
Approved by: philip (ex-mentor)
Discussed: on the lists, at BSDCan, at the DevSummit
Sponsored by: Snow B.V., the Netherlands
dcons(4) fixed by: kan
2008-08-20 08:31:58 +00:00
|
|
|
20080820:
|
|
|
|
The TTY subsystem of the kernel has been replaced by a new
|
|
|
|
implementation, which provides better scalability and an
|
|
|
|
improved driver model. Most common drivers have been migrated to
|
|
|
|
the new TTY subsystem, while others have not. The following
|
|
|
|
drivers have not yet been ported to the new TTY layer:
|
|
|
|
|
|
|
|
PCI/ISA:
|
2008-09-14 19:25:57 +00:00
|
|
|
cy, digi, rc, rp, sio
|
Integrate the new MPSAFE TTY layer to the FreeBSD operating system.
The last half year I've been working on a replacement TTY layer for the
FreeBSD kernel. The new TTY layer was designed to improve the following:
- Improved driver model:
The old TTY layer has a driver model that is not abstract enough to
make it friendly to use. A good example is the output path, where the
device drivers directly access the output buffers. This means that an
in-kernel PPP implementation must always convert network buffers into
TTY buffers.
If a PPP implementation would be built on top of the new TTY layer
(still needs a hooks layer, though), it would allow the PPP
implementation to directly hand the data to the TTY driver.
- Improved hotplugging:
With the old TTY layer, it isn't entirely safe to destroy TTY's from
the system. This implementation has a two-step destructing design,
where the driver first abandons the TTY. After all threads have left
the TTY, the TTY layer calls a routine in the driver, which can be
used to free resources (unit numbers, etc).
The pts(4) driver also implements this feature, which means
posix_openpt() will now return PTY's that are created on the fly.
- Improved performance:
One of the major improvements is the per-TTY mutex, which is expected
to improve scalability when compared to the old Giant locking.
Another change is the unbuffered copying to userspace, which is both
used on TTY device nodes and PTY masters.
Upgrading should be quite straightforward. Unlike previous versions,
existing kernel configuration files do not need to be changed, except
when they reference device drivers that are listed in UPDATING.
Obtained from: //depot/projects/mpsafetty/...
Approved by: philip (ex-mentor)
Discussed: on the lists, at BSDCan, at the DevSummit
Sponsored by: Snow B.V., the Netherlands
dcons(4) fixed by: kan
2008-08-20 08:31:58 +00:00
|
|
|
|
|
|
|
USB:
|
2008-09-14 19:25:57 +00:00
|
|
|
ubser, ucycom
|
Integrate the new MPSAFE TTY layer to the FreeBSD operating system.
The last half year I've been working on a replacement TTY layer for the
FreeBSD kernel. The new TTY layer was designed to improve the following:
- Improved driver model:
The old TTY layer has a driver model that is not abstract enough to
make it friendly to use. A good example is the output path, where the
device drivers directly access the output buffers. This means that an
in-kernel PPP implementation must always convert network buffers into
TTY buffers.
If a PPP implementation would be built on top of the new TTY layer
(still needs a hooks layer, though), it would allow the PPP
implementation to directly hand the data to the TTY driver.
- Improved hotplugging:
With the old TTY layer, it isn't entirely safe to destroy TTY's from
the system. This implementation has a two-step destructing design,
where the driver first abandons the TTY. After all threads have left
the TTY, the TTY layer calls a routine in the driver, which can be
used to free resources (unit numbers, etc).
The pts(4) driver also implements this feature, which means
posix_openpt() will now return PTY's that are created on the fly.
- Improved performance:
One of the major improvements is the per-TTY mutex, which is expected
to improve scalability when compared to the old Giant locking.
Another change is the unbuffered copying to userspace, which is both
used on TTY device nodes and PTY masters.
Upgrading should be quite straightforward. Unlike previous versions,
existing kernel configuration files do not need to be changed, except
when they reference device drivers that are listed in UPDATING.
Obtained from: //depot/projects/mpsafetty/...
Approved by: philip (ex-mentor)
Discussed: on the lists, at BSDCan, at the DevSummit
Sponsored by: Snow B.V., the Netherlands
dcons(4) fixed by: kan
2008-08-20 08:31:58 +00:00
|
|
|
|
|
|
|
Line disciplines:
|
|
|
|
ng_h4, ng_tty, ppp, sl, snp
|
|
|
|
|
|
|
|
Adding these drivers to your kernel configuration file shall
|
|
|
|
cause compilation to fail.
|
|
|
|
|
2008-09-03 08:30:17 +00:00
|
|
|
20080818:
|
|
|
|
ntpd has been upgraded to 4.2.4p5.
|
|
|
|
|
2008-09-01 23:50:56 +00:00
|
|
|
20080801:
|
|
|
|
OpenSSH has been upgraded to 5.1p1.
|
|
|
|
|
|
|
|
For many years, FreeBSD's version of OpenSSH preferred DSA
|
|
|
|
over RSA for host and user authentication keys. With this
|
|
|
|
upgrade, we've switched to the vendor's default of RSA over
|
|
|
|
DSA. This may cause upgraded clients to warn about unknown
|
|
|
|
host keys even for previously known hosts. Users should
|
|
|
|
follow the usual procedure for verifying host keys before
|
|
|
|
accepting the RSA key.
|
|
|
|
|
|
|
|
This can be circumvented by setting the "HostKeyAlgorithms"
|
|
|
|
option to "ssh-dss,ssh-rsa" in ~/.ssh/config or on the ssh
|
|
|
|
command line.
|
|
|
|
|
2008-10-08 01:31:00 +00:00
|
|
|
Please note that the sequence of keys offered for
|
|
|
|
authentication has been changed as well. You may want to
|
|
|
|
specify IdentityFile in a different order to revert this
|
|
|
|
behavior.
|
|
|
|
|
2008-07-13 07:20:14 +00:00
|
|
|
20080713:
|
|
|
|
The sio(4) driver has been removed from the i386 and amd64
|
|
|
|
kernel configuration files. This means uart(4) is now the
|
|
|
|
default serial port driver on those platforms as well.
|
2008-07-25 09:30:53 +00:00
|
|
|
|
2008-07-13 07:20:14 +00:00
|
|
|
To prevent collisions with the sio(4) driver, the uart(4) driver
|
|
|
|
uses different names for its device nodes. This means the
|
|
|
|
onboard serial port will now most likely be called "ttyu0"
|
|
|
|
instead of "ttyd0". You may need to reconfigure applications to
|
|
|
|
use the new device names.
|
|
|
|
|
2008-08-18 10:38:16 +00:00
|
|
|
When using the serial port as a boot console, be sure to update
|
|
|
|
/boot/device.hints and /etc/ttys before booting the new kernel.
|
|
|
|
If you forget to do so, you can still manually specify the hints
|
|
|
|
at the loader prompt:
|
|
|
|
|
|
|
|
set hint.uart.0.at="isa"
|
|
|
|
set hint.uart.0.port="0x3F8"
|
|
|
|
set hint.uart.0.flags="0x10"
|
|
|
|
set hint.uart.0.irq="4"
|
|
|
|
boot -s
|
|
|
|
|
2008-06-09 21:33:57 +00:00
|
|
|
20080609:
|
|
|
|
The gpt(8) utility has been removed. Use gpart(8) to partition
|
|
|
|
disks instead.
|
|
|
|
|
2008-06-03 18:09:10 +00:00
|
|
|
20080603:
|
2008-06-03 17:50:13 +00:00
|
|
|
The version that Linuxulator emulates was changed from 2.4.2
|
|
|
|
to 2.6.16. If you experience any problems with Linux binaries
|
2008-07-25 09:30:53 +00:00
|
|
|
please try to set sysctl compat.linux.osrelease to 2.4.2 and
|
2008-06-03 17:50:13 +00:00
|
|
|
if it fixes the problem contact emulation mailing list.
|
|
|
|
|
2008-05-26 10:40:09 +00:00
|
|
|
20080525:
|
|
|
|
ISDN4BSD (I4B) was removed from the src tree. You may need to
|
|
|
|
update a your kernel configuration and remove relevant entries.
|
|
|
|
|
2008-05-09 23:14:01 +00:00
|
|
|
20080509:
|
|
|
|
I have checked in code to support multiple routing tables.
|
2008-07-25 09:13:18 +00:00
|
|
|
See the man pages setfib(1) and setfib(2).
|
2008-05-09 23:14:01 +00:00
|
|
|
This is a hopefully backwards compatible version,
|
|
|
|
but to make use of it you need to compile your kernel
|
|
|
|
with options ROUTETABLES=2 (or more up to 16).
|
|
|
|
|
2008-04-20 21:25:37 +00:00
|
|
|
20080420:
|
|
|
|
The 802.11 wireless support was redone to enable multi-bss
|
|
|
|
operation on devices that are capable. The underlying device
|
|
|
|
is no longer used directly but instead wlanX devices are
|
|
|
|
cloned with ifconfig. This requires changes to rc.conf files.
|
2008-04-27 04:07:36 +00:00
|
|
|
For example, change:
|
|
|
|
ifconfig_ath0="WPA DHCP"
|
|
|
|
to
|
|
|
|
wlans_ath0=wlan0
|
|
|
|
ifconfig_wlan0="WPA DHCP"
|
2008-04-29 19:55:18 +00:00
|
|
|
see rc.conf(5) for more details. In addition, mergemaster of
|
|
|
|
/etc/rc.d is highly recommended. Simultaneous update of userland
|
|
|
|
and kernel wouldn't hurt either.
|
2008-04-20 21:25:37 +00:00
|
|
|
|
|
|
|
As part of the multi-bss changes the wlan_scan_ap and wlan_scan_sta
|
|
|
|
modules were merged into the base wlan module. All references
|
|
|
|
to these modules (e.g. in kernel config files) must be removed.
|
|
|
|
|
2008-04-08 17:55:26 +00:00
|
|
|
20080408:
|
|
|
|
psm(4) has gained write(2) support in native operation level.
|
|
|
|
Arbitrary commands can be written to /dev/psm%d and status can
|
|
|
|
be read back from it. Therefore, an application is responsible
|
|
|
|
for status validation and error recovery. It is a no-op in
|
|
|
|
other operation levels.
|
|
|
|
|
2008-03-12 09:48:42 +00:00
|
|
|
20080312:
|
|
|
|
Support for KSE threading has been removed from the kernel. To
|
|
|
|
run legacy applications linked against KSE libmap.conf may
|
|
|
|
be used. The following libmap.conf may be used to ensure
|
|
|
|
compatibility with any prior release:
|
|
|
|
|
|
|
|
libpthread.so.1 libthr.so.1
|
|
|
|
libpthread.so.2 libthr.so.2
|
|
|
|
libkse.so.3 libthr.so.3
|
|
|
|
|
2008-03-01 22:54:42 +00:00
|
|
|
20080301:
|
|
|
|
The layout of struct vmspace has changed. This affects libkvm
|
|
|
|
and any executables that link against libkvm and use the
|
|
|
|
kvm_getprocs() function. In particular, but not exclusively,
|
|
|
|
it affects ps(1), fstat(1), pkill(1), systat(1), top(1) and w(1).
|
|
|
|
The effects are minimal, but it's advisable to upgrade world
|
|
|
|
nonetheless.
|
|
|
|
|
2008-02-29 22:08:49 +00:00
|
|
|
20080229:
|
|
|
|
The latest em driver no longer has support in it for the
|
|
|
|
82575 adapter, this is now moved to the igb driver. The
|
|
|
|
split was done to make new features that are incompatible
|
|
|
|
with older hardware easier to do.
|
|
|
|
|
2008-02-20 07:50:13 +00:00
|
|
|
20080220:
|
|
|
|
The new geom_lvm(4) geom class has been renamed to geom_linux_lvm(4),
|
|
|
|
likewise the kernel option is now GEOM_LINUX_LVM.
|
|
|
|
|
2008-02-11 23:23:21 +00:00
|
|
|
20080211:
|
|
|
|
The default NFS mount mode has changed from UDP to TCP for
|
|
|
|
increased reliability. If you rely on (insecurely) NFS
|
|
|
|
mounting across a firewall you may need to update your
|
|
|
|
firewall rules.
|
|
|
|
|
2008-02-08 21:24:58 +00:00
|
|
|
20080208:
|
|
|
|
Belatedly note the addition of m_collapse for compacting
|
|
|
|
mbuf chains.
|
|
|
|
|
Our fts(3) API, as inherited from 4.4BSD, suffers from integer
fields in FTS and FTSENT structs being too narrow. In addition,
the narrow types creep from there into fts.c. As a result, fts(3)
consumers, e.g., find(1) or rm(1), can't handle file trees an ordinary
user can create, which can have security implications.
To fix the historic implementation of fts(3), OpenBSD and NetBSD
have already changed <fts.h> in somewhat incompatible ways, so we
are free to do so, too. This change is a superset of changes from
the other BSDs with a few more improvements. It doesn't touch
fts(3) functionality; it just extends integer types used by it to
match modern reality and the C standard.
Here are its points:
o For C object sizes, use size_t unless it's 100% certain that
the object will be really small. (Note that fts(3) can construct
pathnames _much_ longer than PATH_MAX for its consumers.)
o Avoid the short types because on modern platforms using them
results in larger and slower code. Change shorts to ints as
follows:
- For variables than count simple, limited things like states,
use plain vanilla `int' as it's the type of choice in C.
- For a limited number of bit flags use `unsigned' because signed
bit-wise operations are implementation-defined, i.e., unportable,
in C.
o For things that should be at least 64 bits wide, use long long
and not int64_t, as the latter is an optional type. See
FTSENT.fts_number aka FTS.fts_bignum. Extending fts_number `to
satisfy future needs' is pointless because there is fts_pointer,
which can be used to link to arbitrary data from an FTSENT.
However, there already are fts(3) consumers that require fts_number,
or fts_bignum, have at least 64 bits in it, so we must allow for them.
o For the tree depth, use `long'. This is a trade-off between making
this field too wide and allowing for 64-bit inode numbers and/or
chain-mounted filesystems. On the one hand, `long' is almost
enough for 32-bit filesystems on a 32-bit platform (our ino_t is
uint32_t now). On the other hand, platforms with a 64-bit (or
wider) `long' will be ready for 64-bit inode numbers, as well as
for several 32-bit filesystems mounted one under another. Note
that fts_level has to be signed because -1 is a magic value for it,
FTS_ROOTPARENTLEVEL.
o For the `nlinks' local var in fts_build(), use `long'. The logic
in fts_build() requires that `nlinks' be signed, but our nlink_t
currently is uint16_t. Therefore let's make the signed var wide
enough to be able to represent 2^16-1 in pure C99, and even 2^32-1
on a 64-bit platform. Perhaps the logic should be changed just
to use nlink_t, but it can be done later w/o breaking fts(3) ABI
any more because `nlinks' is just a local var.
This commit also inludes supporting stuff for the fts change:
o Preserve the old versions of fts(3) functions through libc symbol
versioning because the old versions appeared in all our former releases.
o Bump __FreeBSD_version just in case. There is a small chance that
some ill-written 3-rd party apps may fail to build or work correctly
if compiled after this change.
o Update the fts(3) manpage accordingly. In particular, remove
references to fts_bignum, which was a FreeBSD-specific hack to work
around the too narrow types of FTSENT members. Now fts_number is
at least 64 bits wide (long long) and fts_bignum is an undocumented
alias for fts_number kept around for compatibility reasons. According
to Google Code Search, the only big consumers of fts_bignum are in
our own source tree, so they can be fixed easily to use fts_number.
o Mention the change in src/UPDATING.
PR: bin/104458
Approved by: re (quite a while ago)
Discussed with: deischen (the symbol versioning part)
Reviewed by: -arch (mostly silence); das (generally OK, but we didn't
agree on some types used; assuming that no objections on
-arch let me to stick to my opinion)
2008-01-26 17:09:40 +00:00
|
|
|
20080126:
|
|
|
|
The fts(3) structures have been changed to use adequate
|
|
|
|
integer types for their members and so to be able to cope
|
|
|
|
with huge file trees. The old fts(3) ABI is preserved
|
|
|
|
through symbol versioning in libc, so third-party binaries
|
|
|
|
using fts(3) should still work, although they will not take
|
|
|
|
advantage of the extended types. At the same time, some
|
|
|
|
third-party software might fail to build after this change
|
|
|
|
due to unportable assumptions made in its source code about
|
|
|
|
fts(3) structure members. Such software should be fixed
|
|
|
|
by its vendor or, in the worst case, in the ports tree.
|
|
|
|
FreeBSD_version 800015 marks this change for the unlikely
|
|
|
|
case that a portable fix is impossible.
|
|
|
|
|
2008-01-23 22:21:36 +00:00
|
|
|
20080123:
|
2008-01-21 22:04:37 +00:00
|
|
|
To upgrade to -current after this date, you must be running
|
2008-01-23 22:21:36 +00:00
|
|
|
FreeBSD not older than 6.0-RELEASE. Upgrading to -current
|
2008-01-21 22:04:37 +00:00
|
|
|
from 5.x now requires a stop over at RELENG_6 or RELENG_7 systems.
|
|
|
|
|
2007-11-28 13:04:11 +00:00
|
|
|
20071128:
|
|
|
|
The ADAPTIVE_GIANT kernel option has been retired because its
|
|
|
|
functionality is the default now.
|
|
|
|
|
2007-11-18 18:11:16 +00:00
|
|
|
20071118:
|
|
|
|
The AT keyboard emulation of sunkbd(4) has been turned on
|
|
|
|
by default. In order to make the special symbols of the Sun
|
|
|
|
keyboards driven by sunkbd(4) work under X these now have
|
|
|
|
to be configured the same way as Sun USB keyboards driven
|
|
|
|
by ukbd(4) (which also does AT keyboard emulation), f.e.:
|
|
|
|
|
|
|
|
Option "XkbLayout" "us"
|
|
|
|
Option "XkbRules" "xorg"
|
|
|
|
Option "XkbSymbols" "pc(pc105)+sun_vndr/usb(sun_usb)+us"
|
|
|
|
|
2007-10-24 20:51:44 +00:00
|
|
|
20071024:
|
|
|
|
It has been decided that it is desirable to provide ABI
|
|
|
|
backwards compatibility to the FreeBSD 4/5/6 versions of the
|
|
|
|
PCIOCGETCONF, PCIOCREAD and PCIOCWRITE IOCTLs, which was
|
|
|
|
broken with the introduction of PCI domain support (see the
|
|
|
|
20070930 entry). Unfortunately, this required the ABI of
|
|
|
|
PCIOCGETCONF to be broken again in order to be able to
|
|
|
|
provide backwards compatibility to the old version of that
|
|
|
|
IOCTL. Thus consumers of PCIOCGETCONF have to be recompiled
|
|
|
|
again. As for prominent ports this affects neither pciutils
|
|
|
|
nor xorg-server this time, the hal port needs to be rebuilt
|
|
|
|
however.
|
|
|
|
|
2007-10-21 04:27:07 +00:00
|
|
|
20071020:
|
|
|
|
The misnamed kthread_create() and friends have been renamed
|
|
|
|
to kproc_create() etc. Many of the callers already
|
|
|
|
used kproc_start()..
|
|
|
|
I will return kthread_create() and friends in a while
|
|
|
|
with implementations that actually create threads, not procs.
|
|
|
|
Renaming corresponds with version 800002.
|
|
|
|
|
2007-10-11 04:28:08 +00:00
|
|
|
20071010:
|
|
|
|
RELENG_7 branched.
|
|
|
|
|
2007-10-09 17:51:11 +00:00
|
|
|
20071009:
|
2007-10-09 13:42:34 +00:00
|
|
|
Setting WITHOUT_LIBPTHREAD now means WITHOUT_LIBKSE and
|
|
|
|
WITHOUT_LIBTHR are set.
|
|
|
|
|
2007-09-30 11:05:18 +00:00
|
|
|
20070930:
|
|
|
|
The PCI code has been made aware of PCI domains. This means that
|
|
|
|
the location strings as used by pciconf(8) etc are now in the
|
|
|
|
following format: pci<domain>:<bus>:<device>[:<function>]. It
|
|
|
|
also means that consumers of <sys/pciio.h> potentially need to
|
|
|
|
be recompiled; this includes the hal and xorg-server ports.
|
|
|
|
|
2007-09-28 10:38:08 +00:00
|
|
|
20070928:
|
2008-12-08 17:12:40 +00:00
|
|
|
The caching daemon (cached) was renamed to nscd. nscd.conf
|
|
|
|
configuration file should be used instead of cached.conf and
|
|
|
|
nscd_enable, nscd_pidfile and nscd_flags options should be used
|
|
|
|
instead of cached_enable, cached_pidfile and cached_flags in
|
|
|
|
rc.conf.
|
2007-09-28 10:38:08 +00:00
|
|
|
|
2007-10-12 04:48:58 +00:00
|
|
|
20070921:
|
|
|
|
The getfacl(1) utility now prints owning user and group name
|
|
|
|
instead of owning uid and gid in the three line comment header.
|
|
|
|
This is the same behavior as getfacl(1) on Solaris and Linux.
|
|
|
|
|
2007-07-04 21:47:23 +00:00
|
|
|
20070704:
|
2008-12-08 17:12:40 +00:00
|
|
|
The new IPsec code is now compiled in using the IPSEC option. The
|
2007-07-04 21:47:23 +00:00
|
|
|
IPSEC option now requires "device crypto" be defined in your kernel
|
|
|
|
configuration. The FAST_IPSEC kernel option is now deprecated.
|
|
|
|
|
2007-07-03 13:06:45 +00:00
|
|
|
20070702:
|
|
|
|
The packet filter (pf) code has been updated to OpenBSD 4.1 Please
|
|
|
|
note the changed syntax - keep state is now on by default. Also
|
|
|
|
note the fact that ftp-proxy(8) has been changed from bottom up and
|
|
|
|
has been moved from libexec to usr/sbin. Changes in the ALTQ
|
|
|
|
handling also affect users of IPFW's ALTQ capabilities.
|
|
|
|
|
2007-07-01 11:41:27 +00:00
|
|
|
20070701:
|
|
|
|
Remove KAME IPsec in favor of FAST_IPSEC, which is now the
|
|
|
|
only IPsec supported by FreeBSD. The new IPsec stack
|
|
|
|
supports both IPv4 and IPv6. The kernel option will change
|
|
|
|
after the code changes have settled in. For now the kernel
|
|
|
|
option IPSEC is deprecated and FAST_IPSEC is the only option, that
|
|
|
|
will change after some settling time.
|
|
|
|
|
2007-07-01 10:25:07 +00:00
|
|
|
20070701:
|
|
|
|
The wicontrol(8) utility has been removed from the base system. wi(4)
|
|
|
|
cards should be configured using ifconfig(8), see the man page for more
|
|
|
|
information.
|
|
|
|
|
2007-07-09 01:13:00 +00:00
|
|
|
20070612:
|
2007-07-06 06:35:50 +00:00
|
|
|
The i386/amd64 GENERIC kernel now defaults to the nfe(4) driver
|
|
|
|
instead of the nve(4) driver. Please update your configuration
|
|
|
|
accordingly.
|
|
|
|
|
2007-06-12 17:33:23 +00:00
|
|
|
20070612:
|
|
|
|
By default, /etc/rc.d/sendmail no longer rebuilds the aliases
|
|
|
|
database if it is missing or older than the aliases file. If
|
|
|
|
desired, set the new rc.conf option sendmail_rebuild_aliases
|
|
|
|
to "YES" to restore that functionality.
|
|
|
|
|
Import rewrite of IPv4 socket multicast layer to support source-specific
and protocol-independent host mode multicast. The code is written to
accomodate IPv6, IGMPv3 and MLDv2 with only a little additional work.
This change only pertains to FreeBSD's use as a multicast end-station and
does not concern multicast routing; for an IGMPv3/MLDv2 router
implementation, consider the XORP project.
The work is based on Wilbert de Graaf's IGMPv3 code drop for FreeBSD 4.6,
which is available at: http://www.kloosterhof.com/wilbert/igmpv3.html
Summary
* IPv4 multicast socket processing is now moved out of ip_output.c
into a new module, in_mcast.c.
* The in_mcast.c module implements the IPv4 legacy any-source API in
terms of the protocol-independent source-specific API.
* Source filters are lazy allocated as the common case does not use them.
They are part of per inpcb state and are covered by the inpcb lock.
* struct ip_mreqn is now supported to allow applications to specify
multicast joins by interface index in the legacy IPv4 any-source API.
* In UDP, an incoming multicast datagram only requires that the source
port matches the 4-tuple if the socket was already bound by source port.
An unbound socket SHOULD be able to receive multicasts sent from an
ephemeral source port.
* The UDP socket multicast filter mode defaults to exclusive, that is,
sources present in the per-socket list will be blocked from delivery.
* The RFC 3678 userland functions have been added to libc: setsourcefilter,
getsourcefilter, setipv4sourcefilter, getipv4sourcefilter.
* Definitions for IGMPv3 are merged but not yet used.
* struct sockaddr_storage is now referenced from <netinet/in.h>. It
is therefore defined there if not already declared in the same way
as for the C99 types.
* The RFC 1724 hack (specify 0.0.0.0/8 addresses to IP_MULTICAST_IF
which are then interpreted as interface indexes) is now deprecated.
* A patch for the Rhyolite.com routed in the FreeBSD base system
is available in the -net archives. This only affects individuals
running RIPv1 or RIPv2 via point-to-point and/or unnumbered interfaces.
* Make IPv6 detach path similar to IPv4's in code flow; functionally same.
* Bump __FreeBSD_version to 700048; see UPDATING.
This work was financially supported by another FreeBSD committer.
Obtained from: p4://bms_netdev
Submitted by: Wilbert de Graaf (original work)
Reviewed by: rwatson (locking), silence from fenner,
net@ (but with encouragement)
2007-06-12 16:24:56 +00:00
|
|
|
20070612:
|
|
|
|
The IPv4 multicast socket code has been considerably modified, and
|
|
|
|
moved to the file sys/netinet/in_mcast.c. Initial support for the
|
|
|
|
RFC 3678 Source-Specific Multicast Socket API has been added to
|
|
|
|
the IPv4 network stack.
|
|
|
|
|
|
|
|
Strict multicast and broadcast reception is now the default for
|
|
|
|
UDP/IPv4 sockets; the net.inet.udp.strict_mcast_mship sysctl variable
|
|
|
|
has now been removed.
|
|
|
|
|
|
|
|
The RFC 1724 hack for interface selection has been removed; the use
|
|
|
|
of the Linux-derived ip_mreqn structure with IP_MULTICAST_IF has
|
|
|
|
been added to replace it. Consumers such as routed will soon be
|
|
|
|
updated to reflect this.
|
|
|
|
|
|
|
|
These changes affect users who are running routed(8) or rdisc(8)
|
|
|
|
from the FreeBSD base system on point-to-point or unnumbered
|
|
|
|
interfaces.
|
|
|
|
|
2007-06-11 04:06:50 +00:00
|
|
|
20070610:
|
|
|
|
The net80211 layer has changed significantly and all wireless
|
|
|
|
drivers that depend on it need to be recompiled. Further these
|
|
|
|
changes require that any program that interacts with the wireless
|
|
|
|
support in the kernel be recompiled; this includes: ifconfig,
|
|
|
|
wpa_supplicant, hostapd, and wlanstats. Users must also, for
|
|
|
|
the moment, kldload the wlan_scan_sta and/or wlan_scan_ap modules
|
|
|
|
if they use modules for wireless support. These modules implement
|
|
|
|
scanning support for station and ap modes, respectively. Failure
|
|
|
|
to load the appropriate module before marking a wireless interface
|
|
|
|
up will result in a message to the console and the device not
|
|
|
|
operating properly.
|
|
|
|
|
2007-06-10 18:57:20 +00:00
|
|
|
20070610:
|
|
|
|
The pam_nologin(8) module ceases to provide an authentication
|
|
|
|
function and starts providing an account management function.
|
|
|
|
Consequent changes to /etc/pam.d should be brought in using
|
|
|
|
mergemaster(8). Third-party files in /usr/local/etc/pam.d may
|
|
|
|
need manual editing as follows. Locate this line (or similar):
|
|
|
|
|
|
|
|
auth required pam_nologin.so no_warn
|
|
|
|
|
|
|
|
and change it according to this example:
|
|
|
|
|
|
|
|
account required pam_nologin.so no_warn
|
|
|
|
|
|
|
|
That is, the first word needs to be changed from "auth" to
|
|
|
|
"account". The new line can be moved to the account section
|
|
|
|
within the file for clarity. Not updating pam.conf(5) files
|
|
|
|
will result in nologin(5) ignored by the respective services.
|
|
|
|
|
2007-05-29 12:40:45 +00:00
|
|
|
20070529:
|
|
|
|
The ether_ioctl() function has been synchronized with ioctl(2)
|
|
|
|
and ifnet.if_ioctl. Due to that, the size of one of its arguments
|
|
|
|
has changed on 64-bit architectures. All kernel modules using
|
|
|
|
ether_ioctl() need to be rebuilt on such architectures.
|
|
|
|
|
2007-05-16 17:23:54 +00:00
|
|
|
20070516:
|
|
|
|
Improved INCLUDE_CONFIG_FILE support has been introduced to the
|
|
|
|
config(8) utility. In order to take advantage of this new
|
|
|
|
functionality, you are expected to recompile and install
|
|
|
|
src/usr.sbin/config. If you don't rebuild config(8), and your
|
|
|
|
kernel configuration depends on INCLUDE_CONFIG_FILE, the kernel
|
|
|
|
build will be broken because of a missing "kernconfstring"
|
|
|
|
symbol.
|
|
|
|
|
2007-05-13 14:41:48 +00:00
|
|
|
20070513:
|
|
|
|
Symbol versioning is enabled by default. To disable it, use
|
|
|
|
option WITHOUT_SYMVER. It is not advisable to attempt to
|
|
|
|
disable symbol versioning once it is enabled; your installworld
|
|
|
|
will break because a symbol version-less libc will get installed
|
|
|
|
before the install tools. As a result, the old install tools,
|
|
|
|
which previously had symbol dependencies to FBSD_1.0, will fail
|
|
|
|
because the freshly installed libc will not have them.
|
|
|
|
|
2007-10-09 13:42:34 +00:00
|
|
|
The default threading library (providing "libpthread") has been
|
|
|
|
changed to libthr. If you wish to have libkse as your default,
|
|
|
|
use option DEFAULT_THREAD_LIB=libkse for the buildworld.
|
2007-05-13 14:41:48 +00:00
|
|
|
|
2007-04-23 22:15:07 +00:00
|
|
|
20070423:
|
2007-06-12 17:33:56 +00:00
|
|
|
The ABI breakage in sendmail(8)'s libmilter has been repaired
|
|
|
|
so it is no longer necessary to recompile mail filters (aka,
|
|
|
|
milters). If you recompiled mail filters after the 20070408
|
|
|
|
note, it is not necessary to recompile them again.
|
2007-04-23 22:15:07 +00:00
|
|
|
|
2007-04-17 00:41:59 +00:00
|
|
|
20070417:
|
|
|
|
The new trunk(4) driver has been renamed to lagg(4) as it better
|
|
|
|
reflects its purpose. ifconfig will need to be recompiled.
|
|
|
|
|
2007-04-14 16:26:01 +00:00
|
|
|
20070408:
|
|
|
|
sendmail(8) has been updated to version 8.14.1. Mail filters
|
|
|
|
(aka, milters) compiled against the libmilter included in the
|
|
|
|
base operating system should be recompiled.
|
|
|
|
|
2007-03-02 14:56:15 +00:00
|
|
|
20070302:
|
|
|
|
Firmwares for ipw(4) and iwi(4) are now included in the base tree.
|
|
|
|
In order to use them one must agree to the respective LICENSE in
|
2007-04-03 10:04:54 +00:00
|
|
|
share/doc/legal and define legal.intel_<name>.license_ack=1 via
|
2007-04-01 17:49:27 +00:00
|
|
|
loader.conf(5) or kenv(1). Make sure to deinstall the now
|
|
|
|
deprecated modules from the respective firmware ports.
|
2007-03-02 14:56:15 +00:00
|
|
|
|
2007-02-28 21:33:40 +00:00
|
|
|
20070228:
|
|
|
|
The name resolution/mapping functions addr2ascii(3) and ascii2addr(3)
|
|
|
|
were removed from FreeBSD's libc. These originally came from INRIA
|
|
|
|
IPv6. Nothing in FreeBSD ever used them. They may be regarded as
|
|
|
|
deprecated in previous releases.
|
|
|
|
The AF_LINK support for getnameinfo(3) was merged from NetBSD to
|
|
|
|
replace it as a more portable (and re-entrant) API.
|
|
|
|
|
2007-02-24 19:45:09 +00:00
|
|
|
20070224:
|
2008-07-25 09:30:53 +00:00
|
|
|
To support interrupt filtering a modification to the newbus API
|
2007-02-24 20:15:04 +00:00
|
|
|
has occurred, ABI was broken and __FreeBSD_version was bumped
|
2008-07-25 09:30:53 +00:00
|
|
|
to 700031. Please make sure that your kernel and modules are in
|
2007-02-24 21:59:18 +00:00
|
|
|
sync. For more info:
|
|
|
|
http://docs.freebsd.org/cgi/mid.cgi?20070221233124.GA13941
|
2007-02-24 19:45:09 +00:00
|
|
|
|
2007-02-24 11:41:05 +00:00
|
|
|
20070224:
|
|
|
|
The IPv6 multicast forwarding code may now be loaded into GENERIC
|
2007-02-24 21:21:53 +00:00
|
|
|
kernels by loading the ip_mroute.ko module. This is built into the
|
|
|
|
module unless WITHOUT_INET6 or WITHOUT_INET6_SUPPORT options are
|
|
|
|
set; see src.conf(5) for more information.
|
2007-02-24 11:41:05 +00:00
|
|
|
|
2007-02-14 14:17:01 +00:00
|
|
|
20070214:
|
|
|
|
The output of netstat -r has changed. Without -n, we now only
|
|
|
|
print a "network name" without the prefix length if the network
|
|
|
|
address and mask exactly match a Class A/B/C network, and an entry
|
|
|
|
exists in the nsswitch "networks" map.
|
|
|
|
With -n, we print the full unabbreviated CIDR network prefix in
|
|
|
|
the form "a.b.c.d/p". 0.0.0.0/0 is always printed as "default".
|
|
|
|
This change is in preparation for changes such as equal-cost
|
|
|
|
multipath, and to more generally assist operational deployment
|
|
|
|
of FreeBSD as a modern IPv4 router.
|
|
|
|
|
2007-02-10 13:59:13 +00:00
|
|
|
20070210:
|
|
|
|
PIM has been turned on by default in the IPv4 multicast
|
|
|
|
routing code. The kernel option 'PIM' has now been removed.
|
|
|
|
PIM is now built by default if option 'MROUTING' is specified.
|
|
|
|
It may now be loaded into GENERIC kernels by loading the
|
|
|
|
ip_mroute.ko module.
|
|
|
|
|
2007-02-07 16:04:13 +00:00
|
|
|
20070207:
|
|
|
|
Support for IPIP tunnels (VIFF_TUNNEL) in IPv4 multicast routing
|
|
|
|
has been removed. Its functionality may be achieved by explicitly
|
|
|
|
configuring gif(4) interfaces and using the 'phyint' keyword in
|
|
|
|
mrouted.conf.
|
|
|
|
XORP does not support source-routed IPv4 multicast tunnels nor the
|
|
|
|
integrated IPIP tunneling, therefore it is not affected by this
|
|
|
|
change. The __FreeBSD_version macro has been bumped to 700030.
|
|
|
|
|
2006-12-22 03:03:31 +00:00
|
|
|
20061221:
|
|
|
|
Support for PCI Message Signalled Interrupts has been
|
|
|
|
re-enabled in the bge driver, only for those chips which are
|
|
|
|
believed to support it properly. If there are any problems,
|
|
|
|
MSI can be disabled completely by setting the
|
|
|
|
'hw.pci.enable_msi' and 'hw.pci.enable_msix' tunables to 0
|
|
|
|
in the loader.
|
|
|
|
|
2006-12-15 00:30:37 +00:00
|
|
|
20061214:
|
|
|
|
Support for PCI Message Signalled Interrupts has been
|
|
|
|
disabled again in the bge driver. Many revisions of the
|
|
|
|
hardware fail to support it properly. Support can be
|
|
|
|
re-enabled by removing the #define of BGE_DISABLE_MSI in
|
|
|
|
"src/sys/dev/bge/if_bge.c".
|
|
|
|
|
2006-12-14 23:10:59 +00:00
|
|
|
20061214:
|
|
|
|
Support for PCI Message Signalled Interrupts has been added
|
|
|
|
to the bge driver. If there are any problems, MSI can be
|
|
|
|
disabled completely by setting the 'hw.pci.enable_msi' and
|
|
|
|
'hw.pci.enable_msix' tunables to 0 in the loader.
|
|
|
|
|
2006-12-06 06:39:47 +00:00
|
|
|
20061205:
|
2008-07-25 09:30:53 +00:00
|
|
|
The removal of several facets of the experimental Threading
|
|
|
|
system from the kernel means that the proc and thread structures
|
2006-12-06 06:39:47 +00:00
|
|
|
have changed quite a bit. I suggest all kernel modules that might
|
|
|
|
reference these structures be recompiled.. Especially the
|
|
|
|
linux module.
|
|
|
|
|
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes
in every sense.
General
-------
- Multichannel safe, endian safe, format safe
* Large part of critical pcm filters such as vchan.c, feeder_rate.c,
feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that
using them does not cause the pcm data to be converted to 16bit little
endian.
* Macrosses for accessing pcm data safely are defined within sound.h in
the form of PCM_READ_* / PCM_WRITE_*
* Currently, most of them are probably limited for mono/stereo handling,
but the future addition of true multichannel will be much easier.
- Low latency operation
* Well, this require lot more works to do not just within sound driver,
but we're heading towards right direction. Buffer/block sizing within
channel.c is rewritten to calculate precise allocation for various
combination of sample/data/rate size. As a result, applying correct
SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar
to what commercial 4front driver do.
* Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not
result long delay.
* Eliminate sound truncation if the sound data is too small.
DIY:
1) Download / extract
http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz
2) Do a comparison between "cat state*.au > /dev/dsp" and
"for x in state*.au ; do cat $x > /dev/dsp ; done"
- there should be no "perceivable" differences.
Double close for PR kern/31445.
CAVEAT: Low latency come with (unbearable) price especially for poorly
written applications. Applications that trying to act smarter
by requesting (wrong) blocksize/blockcount will suffer the most.
Fixup samples/patches can be found at:
http://people.freebsd.org/~ariff/ports/
- Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42)
due to closer compatibility with 4front driver.
Discussed with: marcus@ (long time ago?)
- All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been
moved to their own dev sysctl nodes, notably:
hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans
Bump __FreeBSD_version.
Driver specific
---------------
- Ditto for sysctls.
- snd_atiixp, snd_es137x, snd_via8233, snd_hda
* Numerous cleanups and fixes.
* _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme.
This was intended for pure debugging and latency measurement, but proven
good enough in few unexpected and rare cases (such as problematic shared
IRQ with GIANT devices - USB). Polling can be enabled/disabled through
dev.pcm.0.polling. Disabled by default.
- snd_ich
* Fix possible overflow during speed calibration. Delay final
initialization (pcm_setstatus) after calibration finished.
PR: kern/100169
Tested by: Kevin Overman <oberman@es.net>
* Inverted EAPD for few Nec VersaPro.
PR: kern/104715
Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp>
Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman,
those at #freebsd-azalia @ freenode and others for testing.
Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
|
|
|
20061126:
|
|
|
|
Sound infrastructure has been updated with various fixes and
|
|
|
|
improvements. Most of the changes are pretty much transparent,
|
|
|
|
with exceptions of followings:
|
|
|
|
1) All sound driver specific sysctls (hw.snd.pcm%d.*) have been
|
|
|
|
moved to their own dev sysctl nodes, for example:
|
|
|
|
hw.snd.pcm0.vchans -> dev.pcm.0.vchans
|
|
|
|
2) /dev/dspr%d.%d has been deprecated. Each channel now has its
|
|
|
|
own chardev in the form of "dsp%d.<function>%d", where <function>
|
|
|
|
is p = playback, r = record and v = virtual, respectively. Users
|
|
|
|
are encouraged to use these devs instead of (old) "/dev/dsp%d.%d".
|
|
|
|
This does not affect those who are using "/dev/dsp".
|
|
|
|
|
2008-07-07 11:44:57 +00:00
|
|
|
20061122:
|
|
|
|
geom(4)'s gmirror(8) class metadata structure has been
|
|
|
|
rev'd from v3 to v4. If you update across this point and
|
|
|
|
your metadata is converted for you, you will not be easily
|
|
|
|
able to downgrade since the /boot/kernel.old/geom_mirror.ko
|
|
|
|
kernel module will be unable to read the v4 metadata. You
|
|
|
|
can resolve this by doing from the loader(8) prompt:
|
|
|
|
|
|
|
|
set vfs.root.mountfrom="ufs:/dev/XXX"
|
|
|
|
|
|
|
|
where XXX is the root slice of one of the disks that composed
|
2008-07-07 13:08:30 +00:00
|
|
|
the mirror (i.e.: /dev/ad0s1a). You can then rebuild
|
2008-07-07 11:44:57 +00:00
|
|
|
the array the same way you built it originally.
|
|
|
|
|
2006-11-22 23:01:40 +00:00
|
|
|
20061122:
|
|
|
|
The following binaries have been disconnected from the build:
|
|
|
|
mount_devfs, mount_ext2fs, mount_fdescfs, mount_procfs, mount_linprocfs,
|
|
|
|
and mount_std. The functionality of these programs has been
|
|
|
|
moved into the mount program. For example, to mount a devfs
|
|
|
|
filesystem, instead of using mount_devfs, use: "mount -t devfs".
|
|
|
|
This does not affect entries in /etc/fstab, since entries in
|
|
|
|
/etc/fstab are always processed with "mount -t fstype".
|
|
|
|
|
2006-11-15 20:02:20 +00:00
|
|
|
20061113:
|
|
|
|
Support for PCI Message Signalled Interrupts on i386 and amd64
|
|
|
|
has been added to the kernel and various drivers will soon be
|
|
|
|
updated to use MSI when it is available. If there are any problems,
|
|
|
|
MSI can be disabled completely by setting the 'hw.pci.enable_msi'
|
|
|
|
and 'hw.pci.enable_msix' tunables to 0 in the loader.
|
|
|
|
|
2006-11-11 03:18:07 +00:00
|
|
|
20061110:
|
|
|
|
The MUTEX_PROFILING option has been renamed to LOCK_PROFILING.
|
|
|
|
The lockmgr object layout has been changed as a result of having
|
|
|
|
a lock_object embedded in it. As a consequence all file system
|
2008-07-25 09:30:53 +00:00
|
|
|
kernel modules must be re-compiled. The mutex profiling man page
|
2006-11-11 03:18:07 +00:00
|
|
|
has not yet been updated to reflect this change.
|
2003-08-20 01:52:30 +00:00
|
|
|
|
2006-10-26 22:05:25 +00:00
|
|
|
20061026:
|
2006-10-26 21:42:22 +00:00
|
|
|
KSE in the kernel has now been made optional and turned on by
|
2006-10-26 22:05:25 +00:00
|
|
|
default. Use 'nooption KSE' in your kernel config to turn it
|
|
|
|
off. All kernel modules *must* be recompiled after this change.
|
|
|
|
There-after, modules from a KSE kernel should be compatible with
|
|
|
|
modules from a NOKSE kernel due to the temporary padding fields
|
|
|
|
added to 'struct proc'.
|
2006-10-26 21:42:22 +00:00
|
|
|
|
2006-09-29 10:39:23 +00:00
|
|
|
20060929:
|
|
|
|
mrouted and its utilities have been removed from the base system.
|
|
|
|
|
2006-09-30 20:01:15 +00:00
|
|
|
20060927:
|
|
|
|
Some ioctl(2) command codes have changed. Full backward ABI
|
|
|
|
compatibility is provided if the "options COMPAT_FREEBSD6" is
|
|
|
|
present in the kernel configuration file. Make sure to add
|
|
|
|
this option to your kernel config file, or recompile X.Org
|
|
|
|
and the rest of ports; otherwise they may refuse to work.
|
|
|
|
|
2006-09-28 13:00:50 +00:00
|
|
|
20060924:
|
|
|
|
tcpslice has been removed from the base system.
|
|
|
|
|
2006-09-29 10:14:37 +00:00
|
|
|
20060913:
|
|
|
|
The sizes of struct tcpcb (and struct xtcpcb) have changed due to
|
|
|
|
the rewrite of TCP syncookies. Tools like netstat, sockstat, and
|
|
|
|
systat needs to be rebuilt.
|
|
|
|
|
2006-09-04 21:49:31 +00:00
|
|
|
20060903:
|
|
|
|
libpcap updated to v0.9.4 and tcpdump to v3.9.4
|
|
|
|
|
2006-08-17 00:41:05 +00:00
|
|
|
20060816:
|
|
|
|
The IPFIREWALL_FORWARD_EXTENDED option is gone and the behaviour
|
2006-09-18 11:24:25 +00:00
|
|
|
for IPFIREWALL_FORWARD is now as it was before when it was first
|
2006-08-17 00:41:05 +00:00
|
|
|
committed and for years after. The behaviour is now ON.
|
|
|
|
|
2006-07-26 16:31:10 +00:00
|
|
|
20060725:
|
|
|
|
enigma(1)/crypt(1) utility has been changed on 64 bit architectures.
|
|
|
|
Now it can decrypt files created from different architectures.
|
|
|
|
Unfortunately, it is no longer able to decrypt a cipher text
|
|
|
|
generated with an older version on 64 bit architectures.
|
|
|
|
If you have such a file, you need old utility to decrypt it.
|
|
|
|
|
2006-07-09 21:16:06 +00:00
|
|
|
20060709:
|
2008-12-08 17:12:40 +00:00
|
|
|
The interface version of the i4b kernel part has changed. So
|
|
|
|
after updating the kernel sources and compiling a new kernel,
|
|
|
|
the i4b user space tools in "/usr/src/usr.sbin/i4b" must also
|
|
|
|
be rebuilt, and vice versa.
|
2008-07-25 09:30:53 +00:00
|
|
|
|
2006-06-27 20:22:32 +00:00
|
|
|
20060627:
|
|
|
|
The XBOX kernel now defaults to the nfe(4) driver instead of
|
|
|
|
the nve(4) driver. Please update your configuration
|
|
|
|
accordingly.
|
|
|
|
|
2006-09-02 22:17:15 +00:00
|
|
|
20060514:
|
|
|
|
The i386-only lnc(4) driver for the AMD Am7900 LANCE and Am79C9xx
|
|
|
|
PCnet family of NICs has been removed. The new le(4) driver serves
|
|
|
|
as an equivalent but cross-platform replacement with the pcn(4)
|
|
|
|
driver still providing performance-optimized support for the subset
|
|
|
|
of AMD Am79C971 PCnet-FAST and greater chips as before.
|
|
|
|
|
2006-05-15 15:47:45 +00:00
|
|
|
20060511:
|
|
|
|
The machdep.* sysctls and the adjkerntz utility have been
|
|
|
|
modified a bit. The new adjkerntz utility uses the new
|
|
|
|
sysctl names and sysctlbyname() calls, so it may be impossible
|
|
|
|
to run an old /sbin/adjkerntz utility in single-user mode
|
|
|
|
with a new kernel. Replace the `adjkerntz -i' step before
|
|
|
|
`make installworld' with:
|
|
|
|
|
|
|
|
/usr/obj/usr/src/sbin/adjkerntz/adjkerntz -i
|
|
|
|
|
|
|
|
and proceed as usual with the rest of the installworld-stage
|
|
|
|
steps. Otherwise, you risk installing binaries with their
|
|
|
|
timestamp set several hours in the future, especially if
|
|
|
|
you are running with local time set to GMT+X hours.
|
|
|
|
|
2006-05-13 06:08:25 +00:00
|
|
|
20060412:
|
|
|
|
The ip6fw utility has been removed. The behavior provided by
|
|
|
|
ip6fw has been in ipfw2 for a good while and the rc.d scripts
|
2006-05-14 20:54:45 +00:00
|
|
|
have been updated to deal with it. There are some rules that
|
2006-05-13 06:08:25 +00:00
|
|
|
might not migrate cleanly. Use rc.firewall6 as a template to
|
|
|
|
rewrite rules.
|
|
|
|
|
2006-05-04 03:48:06 +00:00
|
|
|
20060428:
|
|
|
|
The puc(4) driver has been overhauled. The ebus(4) and sbus(4)
|
|
|
|
attachments have been removed. Make sure to configure scc(4)
|
|
|
|
on sparc64. Note also that by default puc(4) will use uart(4)
|
|
|
|
and not sio(4) for serial ports because interrupt handling has
|
|
|
|
been optimized for multi-port serial cards and only uart(4)
|
|
|
|
implements the interface to support it.
|
|
|
|
|
|
|
|
20060330:
|
|
|
|
The scc(4) driver replaces puc(4) for Serial Communications
|
|
|
|
Controllers (SCCs) like the Siemens SAB82532 and the Zilog
|
|
|
|
Z8530. On sparc64, it is advised to add scc(4) to the kernel
|
|
|
|
configuration to make sure that the serial ports remain
|
|
|
|
functional.
|
|
|
|
|
2006-03-21 10:10:05 +00:00
|
|
|
20060317:
|
|
|
|
Most world/kernel related NO_* build options changed names.
|
|
|
|
New knobs have common prefixes WITHOUT_*/WITH_* (modelled
|
|
|
|
after FreeBSD ports) and should be set in /etc/src.conf
|
|
|
|
(the src.conf(5) manpage is provided). Full backwards
|
|
|
|
compatibility is maintained for the time being though it's
|
|
|
|
highly recommended to start moving old options out of the
|
|
|
|
system-wide /etc/make.conf file into the new /etc/src.conf
|
|
|
|
while also properly renaming them. More conversions will
|
|
|
|
likely follow. Posting to current@:
|
|
|
|
|
|
|
|
http://lists.freebsd.org/pipermail/freebsd-current/2006-March/061725.html
|
|
|
|
|
2006-03-05 22:52:17 +00:00
|
|
|
20060305:
|
|
|
|
The NETSMBCRYPTO kernel option has been retired because its
|
|
|
|
functionality is always included in NETSMB and smbfs.ko now.
|
|
|
|
|
2006-03-03 21:37:38 +00:00
|
|
|
20060303:
|
|
|
|
The TDFX_LINUX kernel option was retired and replaced by the
|
2006-03-09 17:50:01 +00:00
|
|
|
tdfx_linux device. The latter can be loaded as the 3dfx_linux.ko
|
|
|
|
kernel module. Loading it alone should suffice to get 3dfx support
|
|
|
|
for Linux apps because it will pull in 3dfx.ko and linux.ko through
|
|
|
|
its dependencies.
|
2006-03-03 21:37:38 +00:00
|
|
|
|
2006-03-10 18:40:31 +00:00
|
|
|
20060204:
|
|
|
|
The 'audit' group was added to support the new auditing functionality
|
|
|
|
in the base system. Be sure to follow the directions for updating,
|
|
|
|
including the requirement to run mergemaster -p.
|
|
|
|
|
2006-02-01 13:04:52 +00:00
|
|
|
20060201:
|
|
|
|
The kernel ABI to file system modules was changed on i386.
|
|
|
|
Please make sure that your kernel and modules are in sync.
|
2008-07-25 09:30:53 +00:00
|
|
|
|
2006-01-18 19:54:51 +00:00
|
|
|
20060118:
|
|
|
|
This actually occured some time ago, but installing the kernel
|
|
|
|
now also installs a bunch of symbol files for the kernel modules.
|
2006-01-18 20:36:58 +00:00
|
|
|
This increases the size of /boot/kernel to about 67Mbytes. You
|
2006-01-18 19:54:51 +00:00
|
|
|
will need twice this if you will eventually back this up to kernel.old
|
|
|
|
on your next install.
|
|
|
|
If you have a shortage of room in your root partition, you should add
|
|
|
|
-DINSTALL_NODEBUG to your make arguments or add INSTALL_NODEBUG="yes"
|
|
|
|
to your /etc/make.conf.
|
|
|
|
|
2006-01-13 22:37:48 +00:00
|
|
|
20060113:
|
|
|
|
libc's malloc implementation has been replaced. This change has the
|
|
|
|
potential to uncover application bugs that previously went unnoticed.
|
|
|
|
See the malloc(3) manual page for more details.
|
|
|
|
|
2006-01-13 17:32:22 +00:00
|
|
|
20060112:
|
|
|
|
The generic netgraph(4) cookie has been changed. If you upgrade
|
|
|
|
kernel passing this point, you also need to upgrade userland
|
|
|
|
and netgraph(4) utilities like ports/net/mpd or ports/net/mpd4.
|
|
|
|
|
2006-01-31 20:29:04 +00:00
|
|
|
20060106:
|
|
|
|
si(4)'s device files now contain the unit number.
|
|
|
|
Uses of {cua,tty}A[0-9a-f] should be replaced by {cua,tty}A0[0-9a-f].
|
|
|
|
|
2006-01-06 19:04:39 +00:00
|
|
|
20060106:
|
|
|
|
The kernel ABI was mostly destroyed due to a change in the size
|
|
|
|
of struct lock_object which is nested in other structures such
|
|
|
|
as mutexes which are nested in all sorts of other structures.
|
|
|
|
Make sure your kernel and modules are in sync.
|
|
|
|
|
2005-12-31 14:44:49 +00:00
|
|
|
20051231:
|
|
|
|
The page coloring algorithm in the VM subsystem was converted
|
|
|
|
from tuning with kernel options to autotuning. Please remove
|
|
|
|
any PQ_* option except PQ_NOOPT from your kernel config.
|
|
|
|
|
2005-12-11 23:18:58 +00:00
|
|
|
20051211:
|
|
|
|
The net80211-related tools in the tools/tools/ath directory
|
|
|
|
have been moved to tools/tools/net80211 and renamed with a
|
|
|
|
"wlan" prefix. Scripts that use them should be adjusted
|
|
|
|
accordingly.
|
|
|
|
|
2005-12-03 07:51:07 +00:00
|
|
|
20051202:
|
|
|
|
Scripts in the local_startup directories (as defined in
|
|
|
|
/etc/defaults/rc.conf) that have the new rc.d semantics will
|
|
|
|
now be run as part of the base system rcorder. If there are
|
|
|
|
errors or problems with one of these local scripts, it could
|
|
|
|
cause boot problems. If you encounter such problems, boot in
|
|
|
|
single user mode, remove that script from the */rc.d directory.
|
|
|
|
Please report the problem to the port's maintainer, and the
|
|
|
|
freebsd-ports@freebsd.org mailing list.
|
|
|
|
|
2005-11-29 19:13:28 +00:00
|
|
|
20051129:
|
|
|
|
The nodev mount option was deprecated in RELENG_6 (where it
|
|
|
|
was a no-op), and is now unsupported. If you have nodev or dev listed
|
|
|
|
in /etc/fstab, remove it, otherwise it will result in a mount error.
|
|
|
|
|
2005-11-29 08:59:41 +00:00
|
|
|
20051129:
|
|
|
|
ABI between ipfw(4) and ipfw(8) has been changed. You need
|
|
|
|
to rebuild ipfw(8) when rebuilding kernel.
|
|
|
|
|
2005-12-19 03:15:49 +00:00
|
|
|
20051108:
|
|
|
|
rp(4)'s device files now contain the unit number.
|
|
|
|
Uses of {cua,tty}R[0-9a-f] should be replaced by {cua,tty}R0[0-9a-f].
|
|
|
|
|
2005-10-29 05:27:32 +00:00
|
|
|
20051029:
|
|
|
|
/etc/rc.d/ppp-user has been renamed to /etc/rc.d/ppp.
|
|
|
|
Its /etc/rc.conf.d configuration file has been `ppp' from
|
|
|
|
the beginning, and hence there is no need to touch it.
|
|
|
|
|
2006-02-19 01:05:57 +00:00
|
|
|
20051014:
|
|
|
|
Now most modules get their build-time options from the kernel
|
|
|
|
configuration file. A few modules still have fixed options
|
|
|
|
due to their non-conformant implementation, but they will be
|
|
|
|
corrected eventually. You may need to review the options of
|
|
|
|
the modules in use, explicitly specify the non-default options
|
|
|
|
in the kernel configuration file, and rebuild the kernel and
|
|
|
|
modules afterwards.
|
|
|
|
|
2005-10-01 20:53:51 +00:00
|
|
|
20051001:
|
|
|
|
kern.polling.enable sysctl MIB is now deprecated. Use ifconfig(8)
|
|
|
|
to turn polling(4) on your interfaces.
|
|
|
|
|
2005-09-27 18:10:43 +00:00
|
|
|
20050927:
|
|
|
|
The old bridge(4) implementation was retired. The new
|
|
|
|
if_bridge(4) serves as a full functional replacement.
|
|
|
|
|
2005-07-22 18:51:36 +00:00
|
|
|
20050722:
|
|
|
|
The ai_addrlen of a struct addrinfo was changed to a socklen_t
|
|
|
|
to conform to POSIX-2001. This change broke an ABI
|
|
|
|
compatibility on 64 bit architecture. You have to recompile
|
|
|
|
userland programs that use getaddrinfo(3) on 64 bit
|
|
|
|
architecture.
|
|
|
|
|
2005-07-11 15:46:45 +00:00
|
|
|
20050711:
|
|
|
|
RELENG_6 branched here.
|
|
|
|
|
2005-06-30 05:02:34 +00:00
|
|
|
20050629:
|
|
|
|
The pccard_ifconfig rc.conf variable has been removed and a new
|
|
|
|
variable, ifconfig_DEFAULT has been introduced. Unlike
|
|
|
|
pccard_ifconfig, ifconfig_DEFAULT applies to ALL interfaces that
|
|
|
|
do not have ifconfig_ifn entries rather than just those in
|
|
|
|
removable_interfaces.
|
|
|
|
|
2005-07-01 15:12:09 +00:00
|
|
|
20050616:
|
|
|
|
Some previous versions of PAM have permitted the use of
|
|
|
|
non-absolute paths in /etc/pam.conf or /etc/pam.d/* when referring
|
|
|
|
to third party PAM modules in /usr/local/lib. A change has been
|
|
|
|
made to require the use of absolute paths in order to avoid
|
|
|
|
ambiguity and dependence on library path configuration, which may
|
|
|
|
affect existing configurations.
|
|
|
|
|
2005-06-10 16:51:49 +00:00
|
|
|
20050610:
|
|
|
|
Major changes to network interface API. All drivers must be
|
|
|
|
recompiled. Drivers not in the base system will need to be
|
|
|
|
updated to the new APIs.
|
|
|
|
|
2005-06-09 19:09:38 +00:00
|
|
|
20050609:
|
|
|
|
Changes were made to kinfo_proc in sys/user.h. Please recompile
|
|
|
|
userland, or commands like `fstat', `pkill', `ps', `top' and `w'
|
|
|
|
will not behave correctly.
|
|
|
|
|
2005-06-10 19:59:26 +00:00
|
|
|
The API and ABI for hwpmc(4) have changed with the addition
|
|
|
|
of sampling support. Please recompile lib/libpmc(3) and
|
|
|
|
usr.sbin/{pmcstat,pmccontrol}.
|
|
|
|
|
2005-06-07 18:07:47 +00:00
|
|
|
20050606:
|
|
|
|
The OpenBSD dhclient was imported in place of the ISC dhclient
|
|
|
|
and the network interface configuration scripts were updated
|
|
|
|
accordingly. If you use DHCP to configure your interfaces, you
|
|
|
|
must now run devd. Also, DNS updating was lost so you will need
|
|
|
|
to find a workaround if you use this feature.
|
|
|
|
|
2006-03-10 18:40:31 +00:00
|
|
|
The '_dhcp' user was added to support the OpenBSD dhclient. Be
|
|
|
|
sure to run mergemaster -p (like you are supposed to do every time
|
|
|
|
anyway).
|
|
|
|
|
2005-06-05 03:38:03 +00:00
|
|
|
20050605:
|
|
|
|
if_bridge was added to the tree. This has changed struct ifnet.
|
|
|
|
Please recompile userland and all network related modules.
|
|
|
|
|
2005-06-03 03:34:21 +00:00
|
|
|
20050603:
|
|
|
|
The n_net of a struct netent was changed to an uint32_t, and
|
|
|
|
1st argument of getnetbyaddr() was changed to an uint32_t, to
|
2005-06-03 09:15:17 +00:00
|
|
|
conform to POSIX-2001. These changes broke an ABI
|
2005-06-03 10:19:26 +00:00
|
|
|
compatibility on 64 bit architecture. With these changes,
|
|
|
|
shlib major of libpcap was bumped. You have to recompile
|
|
|
|
userland programs that use getnetbyaddr(3), getnetbyname(3),
|
|
|
|
getnetent(3) and/or libpcap on 64 bit architecture.
|
2005-06-03 03:34:21 +00:00
|
|
|
|
2005-05-28 22:45:31 +00:00
|
|
|
20050528:
|
|
|
|
Kernel parsing of extra options on '#!' first lines of shell
|
|
|
|
scripts has changed. Lines with multiple options likely will
|
|
|
|
fail after this date. For full details, please see
|
|
|
|
http://people.freebsd.org/~gad/Updating-20050528.txt
|
|
|
|
|
2005-05-03 17:43:14 +00:00
|
|
|
20050503:
|
|
|
|
The packet filter (pf) code has been updated to OpenBSD 3.7
|
|
|
|
Please note the changed anchor syntax and the fact that
|
|
|
|
authpf(8) now needs a mounted fdescfs(5) to function.
|
|
|
|
|
2005-04-18 14:33:18 +00:00
|
|
|
20050415:
|
|
|
|
The NO_MIXED_MODE kernel option has been removed from the i386
|
|
|
|
amd64 platforms as its use has been superceded by the new local
|
|
|
|
APIC timer code. Any kernel config files containing this option
|
|
|
|
should be updated.
|
|
|
|
|
2005-02-26 22:26:10 +00:00
|
|
|
20050227:
|
|
|
|
The on-disk format of LC_CTYPE files was changed to be machine
|
|
|
|
independent. Please make sure NOT to use NO_CLEAN buildworld
|
2005-12-06 10:39:14 +00:00
|
|
|
when crossing this point. Crossing this point also requires
|
|
|
|
recompile or reinstall of all locale depended packages.
|
2005-02-26 22:26:10 +00:00
|
|
|
|
2005-02-25 19:46:41 +00:00
|
|
|
20050225:
|
|
|
|
The ifi_epoch member of struct if_data has been changed to
|
|
|
|
contain the uptime at which the interface was created or the
|
|
|
|
statistics zeroed rather then the wall clock time because
|
|
|
|
wallclock time may go backwards. This should have no impact
|
|
|
|
unless an snmp implementation is using this value (I know of
|
|
|
|
none at this point.)
|
|
|
|
|
2005-02-25 20:10:38 +00:00
|
|
|
20050224:
|
|
|
|
The acpi_perf and acpi_throttle drivers are now part of the
|
|
|
|
acpi(4) main module. They are no longer built separately.
|
|
|
|
|
2005-02-25 11:49:42 +00:00
|
|
|
20050223:
|
|
|
|
The layout of struct image_params has changed. You have to
|
|
|
|
recompile all compatibility modules (linux, svr4, etc) for use
|
|
|
|
with the new kernel.
|
|
|
|
|
2005-02-23 16:52:55 +00:00
|
|
|
20050223:
|
|
|
|
The p4tcc driver has been merged into cpufreq(4). This makes
|
|
|
|
"options CPU_ENABLE_TCC" obsolete. Please load cpufreq.ko or
|
|
|
|
compile in "device cpufreq" to restore this functionality.
|
|
|
|
|
2005-03-01 02:33:34 +00:00
|
|
|
20050220:
|
|
|
|
The responsibility of recomputing the file system summary of
|
|
|
|
a SoftUpdates-enabled dirty volume has been transferred to the
|
|
|
|
background fsck. A rebuild of fsck(8) utility is recommended
|
|
|
|
if you have updated the kernel.
|
|
|
|
|
|
|
|
To get the old behavior (recompute file system summary at mount
|
|
|
|
time), you can set vfs.ffs.compute_summary_at_mount=1 before
|
|
|
|
mounting the new volume.
|
|
|
|
|
2005-02-06 21:24:50 +00:00
|
|
|
20050206:
|
|
|
|
The cpufreq import is complete. As part of this, the sysctls for
|
2005-02-25 20:10:38 +00:00
|
|
|
acpi(4) throttling have been removed. The power_profile script
|
2005-02-06 21:24:50 +00:00
|
|
|
has been updated, so you can use performance/economy_cpu_freq in
|
2005-02-25 20:10:38 +00:00
|
|
|
rc.conf(5) to set AC on/offline cpu frequencies.
|
2005-02-06 21:24:50 +00:00
|
|
|
|
2005-02-05 23:25:59 +00:00
|
|
|
20050206:
|
|
|
|
NG_VERSION has been increased. Recompiling kernel (or ng_socket.ko)
|
|
|
|
requires recompiling libnetgraph and userland netgraph utilities.
|
2005-01-15 01:53:49 +00:00
|
|
|
|
|
|
|
20050114:
|
|
|
|
Support for abbreviated forms of a number of ipfw options is
|
|
|
|
now deprecated. Warnings are printed to stderr indicating the
|
|
|
|
correct full form when a match occurs. Some abbreviations may
|
|
|
|
be supported at a later date based on user feedback. To be
|
|
|
|
considered for support, abbreviations must be in use prior to
|
|
|
|
this commit and unlikely to be confused with current key words.
|
|
|
|
|
2004-12-23 16:03:08 +00:00
|
|
|
20041221:
|
|
|
|
By a popular demand, a lot of NOFOO options were renamed
|
|
|
|
to NO_FOO (see bsd.compat.mk for a full list). The old
|
|
|
|
spellings are still supported, but will cause annoying
|
|
|
|
warnings on stderr. Make sure you upgrade properly (see
|
|
|
|
the COMMON ITEMS: section later in this file).
|
|
|
|
|
2004-12-20 04:27:23 +00:00
|
|
|
20041219:
|
|
|
|
Auto-loading of ancillary wlan modules such as wlan_wep has
|
|
|
|
been temporarily disabled; you need to statically configure
|
|
|
|
the modules you need into your kernel or explicitly load them
|
|
|
|
prior to use. Specifically, if you intend to use WEP encryption
|
|
|
|
with an 802.11 device load/configure wlan_wep; if you want to
|
|
|
|
use WPA with the ath driver load/configure wlan_tkip, wlan_ccmp,
|
|
|
|
and wlan_xauth as required.
|
|
|
|
|
2004-12-13 17:52:10 +00:00
|
|
|
20041213:
|
|
|
|
The behaviour of ppp(8) has changed slightly. If lqr is enabled
|
|
|
|
(``enable lqr''), older versions would revert to LCP ECHO mode on
|
|
|
|
negotiation failure. Now, ``enable echo'' is required for this
|
|
|
|
behaviour. The ppp version number has been bumped to 3.4.2 to
|
|
|
|
reflect the change.
|
|
|
|
|
2004-12-11 23:21:31 +00:00
|
|
|
20041201:
|
|
|
|
The wlan support has been updated to split the crypto support
|
|
|
|
into separate modules. For static WEP you must configure the
|
|
|
|
wlan_wep module in your system or build and install the module
|
|
|
|
in place where it can be loaded (the kernel will auto-load
|
|
|
|
the module when a wep key is configured).
|
|
|
|
|
|
|
|
20041201:
|
|
|
|
The ath driver has been updated to split the tx rate control
|
|
|
|
algorithm into a separate module. You need to include either
|
|
|
|
ath_rate_onoe or ath_rate_amrr when configuring the kernel.
|
|
|
|
|
2004-11-16 21:18:41 +00:00
|
|
|
20041116:
|
|
|
|
Support for systems with an 80386 CPU has been removed. Please
|
|
|
|
use FreeBSD 5.x or earlier on systems with an 80386.
|
|
|
|
|
2004-11-10 07:39:27 +00:00
|
|
|
20041110:
|
|
|
|
We have had a hack which would mount the root filesystem
|
|
|
|
R/W if the device were named 'md*'. As part of the vnode
|
|
|
|
work I'm doing I have had to remove this hack. People
|
|
|
|
building systems which use preloaded MD root filesystems
|
|
|
|
may need to insert a "/sbin/mount -u -o rw /dev/md0 /" in
|
|
|
|
their /etc/rc scripts.
|
|
|
|
|
2005-02-07 09:15:52 +00:00
|
|
|
20041104:
|
|
|
|
FreeBSD 5.3 shipped here.
|
|
|
|
|
2004-11-02 22:22:22 +00:00
|
|
|
20041102:
|
|
|
|
The size of struct tcpcb has changed again due to the removal
|
|
|
|
of RFC1644 T/TCP. You have to recompile userland programs that
|
|
|
|
read kmem for tcp sockets directly (netstat, sockstat, etc.)
|
|
|
|
|
2004-10-22 19:55:04 +00:00
|
|
|
20041022:
|
|
|
|
The size of struct tcpcb has changed. You have to recompile
|
|
|
|
userland programs that read kmem for tcp sockets directly
|
|
|
|
(netstat, sockstat, etc.)
|
|
|
|
|
2005-02-04 21:22:06 +00:00
|
|
|
20041016:
|
|
|
|
RELENG_5 branched here. For older entries, please see updating
|
|
|
|
in the RELENG_5 branch.
|
1999-01-14 05:59:52 +00:00
|
|
|
|
2000-01-30 23:15:21 +00:00
|
|
|
COMMON ITEMS:
|
|
|
|
|
2000-09-04 03:18:01 +00:00
|
|
|
General Notes
|
|
|
|
-------------
|
|
|
|
Avoid using make -j when upgrading. From time to time in the
|
|
|
|
past there have been problems using -j with buildworld and/or
|
|
|
|
installworld. This is especially true when upgrading between
|
|
|
|
"distant" versions (eg one that cross a major release boundary
|
|
|
|
or several minor releases, or when several months have passed
|
|
|
|
on the -current branch).
|
|
|
|
|
2002-01-26 21:33:07 +00:00
|
|
|
Sometimes, obscure build problems are the result of environment
|
|
|
|
poisoning. This can happen because the make utility reads its
|
|
|
|
environment when searching for values for global variables.
|
|
|
|
To run your build attempts in an "environmental clean room",
|
|
|
|
prefix all make commands with 'env -i '. See the env(1) manual
|
|
|
|
page for more details.
|
|
|
|
|
2005-12-07 20:49:42 +00:00
|
|
|
When upgrading from one major version to another it is generally
|
|
|
|
best to upgrade to the latest code in the currently installed branch
|
|
|
|
first, then do an upgrade to the new branch. This is the best-tested
|
|
|
|
upgrade path, and has the highest probability of being successful.
|
|
|
|
Please try this approach before reporting problems with a major
|
|
|
|
version upgrade.
|
|
|
|
|
2000-01-30 23:15:21 +00:00
|
|
|
To build a kernel
|
|
|
|
-----------------
|
2000-08-10 05:14:31 +00:00
|
|
|
If you are updating from a prior version of FreeBSD (even one just
|
2005-09-09 15:59:17 +00:00
|
|
|
a few days old), you should follow this procedure. It is the most
|
|
|
|
failsafe as it uses a /usr/obj tree with a fresh mini-buildworld,
|
|
|
|
|
|
|
|
make kernel-toolchain
|
2003-10-03 18:29:06 +00:00
|
|
|
make -DALWAYS_CHECK_MAKE buildkernel KERNCONF=YOUR_KERNEL_HERE
|
|
|
|
make -DALWAYS_CHECK_MAKE installkernel KERNCONF=YOUR_KERNEL_HERE
|
2000-08-10 05:14:31 +00:00
|
|
|
|
2005-05-25 21:03:13 +00:00
|
|
|
To test a kernel once
|
|
|
|
---------------------
|
|
|
|
If you just want to boot a kernel once (because you are not sure
|
|
|
|
if it works, or if you want to boot a known bad kernel to provide
|
|
|
|
debugging information) run
|
|
|
|
make installkernel KERNCONF=YOUR_KERNEL_HERE KODIR=/boot/testkernel
|
|
|
|
nextboot -k testkernel
|
|
|
|
|
2000-08-10 05:14:31 +00:00
|
|
|
To just build a kernel when you know that it won't mess you up
|
|
|
|
--------------------------------------------------------------
|
2004-04-11 03:30:09 +00:00
|
|
|
This assumes you are already running a 5.X system. Replace
|
|
|
|
${arch} with the architecture of your machine (e.g. "i386",
|
|
|
|
"alpha", "amd64", "ia64", "pc98", "sparc64", etc).
|
|
|
|
|
|
|
|
cd src/sys/${arch}/conf
|
2002-01-06 20:18:13 +00:00
|
|
|
config KERNEL_NAME_HERE
|
2004-04-11 03:30:09 +00:00
|
|
|
cd ../compile/KERNEL_NAME_HERE
|
2000-08-10 05:14:31 +00:00
|
|
|
make depend
|
|
|
|
make
|
2004-06-22 10:33:58 +00:00
|
|
|
make install
|
2000-08-10 05:14:31 +00:00
|
|
|
|
|
|
|
If this fails, go to the "To build a kernel" section.
|
|
|
|
|
|
|
|
To rebuild everything and install it on the current system.
|
|
|
|
-----------------------------------------------------------
|
2002-10-26 06:25:11 +00:00
|
|
|
# Note: sometimes if you are running current you gotta do more than
|
|
|
|
# is listed here if you are upgrading from a really old current.
|
|
|
|
|
2002-10-26 22:55:43 +00:00
|
|
|
<make sure you have good level 0 dumps>
|
2002-10-26 06:25:11 +00:00
|
|
|
make buildworld
|
2004-08-25 19:39:13 +00:00
|
|
|
make kernel KERNCONF=YOUR_KERNEL_HERE
|
2002-10-26 06:25:11 +00:00
|
|
|
[1]
|
|
|
|
<reboot in single user> [3]
|
|
|
|
mergemaster -p [5]
|
|
|
|
make installworld
|
2005-07-23 14:23:30 +00:00
|
|
|
make delete-old
|
2002-10-26 06:25:11 +00:00
|
|
|
mergemaster [4]
|
|
|
|
<reboot>
|
|
|
|
|
2000-02-06 04:07:11 +00:00
|
|
|
|
2003-02-07 08:26:23 +00:00
|
|
|
To cross-install current onto a separate partition
|
|
|
|
--------------------------------------------------
|
|
|
|
# In this approach we use a separate partition to hold
|
|
|
|
# current's root, 'usr', and 'var' directories. A partition
|
|
|
|
# holding "/", "/usr" and "/var" should be about 2GB in
|
|
|
|
# size.
|
|
|
|
|
|
|
|
<make sure you have good level 0 dumps>
|
|
|
|
<boot into -stable>
|
|
|
|
make buildworld
|
2005-02-23 20:37:11 +00:00
|
|
|
make buildkernel KERNCONF=YOUR_KERNEL_HERE
|
2003-02-07 08:26:23 +00:00
|
|
|
<maybe newfs current's root partition>
|
|
|
|
<mount current's root partition on directory ${CURRENT_ROOT}>
|
|
|
|
make installworld DESTDIR=${CURRENT_ROOT}
|
2006-04-07 11:36:25 +00:00
|
|
|
make distribution DESTDIR=${CURRENT_ROOT} # if newfs'd
|
2005-02-23 20:37:11 +00:00
|
|
|
make installkernel KERNCONF=YOUR_KERNEL_HERE DESTDIR=${CURRENT_ROOT}
|
2003-02-07 08:26:23 +00:00
|
|
|
cp /etc/fstab ${CURRENT_ROOT}/etc/fstab # if newfs'd
|
|
|
|
<edit ${CURRENT_ROOT}/etc/fstab to mount "/" from the correct partition>
|
|
|
|
<reboot into current>
|
|
|
|
<do a "native" rebuild/install as described in the previous section>
|
2008-10-08 01:31:00 +00:00
|
|
|
<maybe install compatibility libraries from ports/misc/compat*>
|
2008-12-08 17:12:40 +00:00
|
|
|
<reboot>
|
2003-02-07 08:26:23 +00:00
|
|
|
|
|
|
|
|
2005-03-03 08:43:33 +00:00
|
|
|
To upgrade in-place from 5.x-stable to current
|
2003-02-07 08:26:23 +00:00
|
|
|
----------------------------------------------
|
2002-10-26 22:55:43 +00:00
|
|
|
<make sure you have good level 0 dumps>
|
2002-10-28 21:33:10 +00:00
|
|
|
make buildworld [9]
|
2004-09-04 21:03:10 +00:00
|
|
|
make kernel KERNCONF=YOUR_KERNEL_HERE [8]
|
2002-10-26 06:21:00 +00:00
|
|
|
[1]
|
|
|
|
<reboot in single user> [3]
|
|
|
|
mergemaster -p [5]
|
2000-06-14 15:42:50 +00:00
|
|
|
make installworld
|
2005-07-23 14:23:30 +00:00
|
|
|
make delete-old
|
2004-06-26 21:58:56 +00:00
|
|
|
mergemaster -i [4]
|
2000-06-14 15:42:50 +00:00
|
|
|
<reboot>
|
|
|
|
|
2000-08-28 03:54:51 +00:00
|
|
|
Make sure that you've read the UPDATING file to understand the
|
|
|
|
tweaks to various things you need. At this point in the life
|
|
|
|
cycle of current, things change often and you are on your own
|
|
|
|
to cope. The defaults can also change, so please read ALL of
|
|
|
|
the UPDATING entries.
|
2000-06-14 15:42:50 +00:00
|
|
|
|
2000-08-06 22:16:34 +00:00
|
|
|
Also, if you are tracking -current, you must be subscribed to
|
|
|
|
freebsd-current@freebsd.org. Make sure that before you update
|
|
|
|
your sources that you have read and understood all the recent
|
|
|
|
messages there. If in doubt, please track -stable which has
|
|
|
|
much fewer pitfalls.
|
|
|
|
|
2000-08-10 05:03:49 +00:00
|
|
|
[1] If you have third party modules, such as vmware, you
|
|
|
|
should disable them at this point so they don't crash your
|
|
|
|
system on reboot.
|
|
|
|
|
2001-06-04 16:17:06 +00:00
|
|
|
[3] From the bootblocks, boot -s, and then do
|
|
|
|
fsck -p
|
|
|
|
mount -u /
|
|
|
|
mount -a
|
2004-08-25 19:39:13 +00:00
|
|
|
cd src
|
2002-01-06 20:18:13 +00:00
|
|
|
adjkerntz -i # if CMOS is wall time
|
2002-04-25 02:14:48 +00:00
|
|
|
Also, when doing a major release upgrade, it is required that
|
|
|
|
you boot into single user mode to do the installworld.
|
2001-06-04 16:17:06 +00:00
|
|
|
|
2001-08-11 02:03:51 +00:00
|
|
|
[4] Note: This step is non-optional. Failure to do this step
|
|
|
|
can result in a significant reduction in the functionality of the
|
|
|
|
system. Attempting to do it by hand is not recommended and those
|
|
|
|
that pursue this avenue should read this file carefully, as well
|
|
|
|
as the archives of freebsd-current and freebsd-hackers mailing lists
|
|
|
|
for potential gotchas.
|
|
|
|
|
2002-04-19 04:22:35 +00:00
|
|
|
[5] Usually this step is a noop. However, from time to time
|
|
|
|
you may need to do this if you get unknown user in the following
|
|
|
|
step. It never hurts to do it all the time. You may need to
|
|
|
|
install a new mergemaster (cd src/usr.sbin/mergemaster && make
|
|
|
|
install) after the buildworld before this step if you last updated
|
|
|
|
from current before 20020224 or from -stable before 20020408.
|
|
|
|
|
2002-10-27 04:48:31 +00:00
|
|
|
[8] In order to have a kernel that can run the 4.x binaries
|
|
|
|
needed to do an installworld, you must include the COMPAT_FREEBSD4
|
|
|
|
option in your kernel. Failure to do so may leave you with a system
|
2005-12-06 10:39:14 +00:00
|
|
|
that is hard to boot to recover. A similar kernel option COMPAT_FREEBSD5
|
|
|
|
is required to run the 5.x binaries on more recent kernels.
|
2002-10-27 04:48:31 +00:00
|
|
|
|
2004-09-04 21:03:10 +00:00
|
|
|
Make sure that you merge any new devices from GENERIC since the
|
|
|
|
last time you updated your kernel config file.
|
|
|
|
|
2002-10-28 21:33:10 +00:00
|
|
|
[9] When checking out sources, you must include the -P flag to have
|
2004-09-04 21:03:10 +00:00
|
|
|
cvs prune empty directories.
|
|
|
|
|
|
|
|
If CPUTYPE is defined in your /etc/make.conf, make sure to use the
|
|
|
|
"?=" instead of the "=" assignment operator, so that buildworld can
|
|
|
|
override the CPUTYPE if it needs to.
|
|
|
|
|
|
|
|
MAKEOBJDIRPREFIX must be defined in an environment variable, and
|
|
|
|
not on the command line, or in /etc/make.conf. buildworld will
|
|
|
|
warn if it is improperly defined.
|
2000-01-30 23:15:21 +00:00
|
|
|
FORMAT:
|
|
|
|
|
2000-09-04 16:59:32 +00:00
|
|
|
This file contains a list, in reverse chronological order, of major
|
1999-02-14 05:18:35 +00:00
|
|
|
breakages in tracking -current. Not all things will be listed here,
|
2005-03-03 08:43:33 +00:00
|
|
|
and it only starts on October 16, 2004. Updating files can found in
|
2000-03-23 05:57:10 +00:00
|
|
|
previous releases if your system is older than this.
|
1999-02-14 05:18:35 +00:00
|
|
|
|
2001-06-29 06:00:44 +00:00
|
|
|
Copyright information:
|
|
|
|
|
2005-03-03 08:43:33 +00:00
|
|
|
Copyright 1998-2005 M. Warner Losh. All Rights Reserved.
|
2001-06-29 06:00:44 +00:00
|
|
|
|
2001-09-23 06:36:41 +00:00
|
|
|
Redistribution, publication, translation and use, with or without
|
|
|
|
modification, in full or in part, in any form or format of this
|
2002-04-11 05:45:17 +00:00
|
|
|
document are permitted without further permission from the author.
|
2001-06-29 06:00:44 +00:00
|
|
|
|
|
|
|
THIS DOCUMENT IS PROVIDED BY WARNER LOSH ``AS IS'' AND ANY EXPRESS OR
|
|
|
|
IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
|
|
|
|
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
|
|
|
|
DISCLAIMED. IN NO EVENT SHALL WARNER LOSH BE LIABLE FOR ANY DIRECT,
|
|
|
|
INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
|
|
|
|
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
|
|
|
|
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
|
|
HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
|
|
|
|
STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
|
|
|
|
IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
|
|
|
POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
|
|
|
|
If you find this document useful, and you want to, you may buy the
|
|
|
|
author a beer.
|
|
|
|
|
2001-09-23 06:44:07 +00:00
|
|
|
Contact Warner Losh if you have any questions about your use of
|
2001-09-23 06:36:41 +00:00
|
|
|
this document.
|
|
|
|
|
1999-08-28 01:35:59 +00:00
|
|
|
$FreeBSD$
|