2018-07-27 21:36:29 +00:00
|
|
|
Updating Information for FreeBSD current users.
|
1999-01-14 05:59:52 +00:00
|
|
|
|
2009-09-03 17:04:42 +00:00
|
|
|
This file is maintained and copyrighted by M. Warner Losh <imp@freebsd.org>.
|
|
|
|
See end of file for further details. For commonly done items, please see the
|
|
|
|
COMMON ITEMS: section later in the file. These instructions assume that you
|
|
|
|
basically know what you are doing. If not, then please consult the FreeBSD
|
2012-12-02 22:09:16 +00:00
|
|
|
handbook:
|
|
|
|
|
2018-06-26 14:30:33 +00:00
|
|
|
https://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/makeworld.html
|
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
|
2009-09-03 17:04:42 +00:00
|
|
|
/usr/ports/UPDATING. Please read that file before running portupgrade.
|
2004-07-29 15:06:19 +00:00
|
|
|
|
2013-04-20 01:12:23 +00:00
|
|
|
NOTE: FreeBSD has switched from gcc to clang. If you have trouble bootstrapping
|
2014-05-06 04:22:29 +00:00
|
|
|
from older versions of FreeBSD, try WITHOUT_CLANG and WITH_GCC to bootstrap to
|
2015-10-13 12:51:44 +00:00
|
|
|
the tip of head, and then rebuild without this option. The bootstrap process
|
|
|
|
from older version of current across the gcc/clang cutover is a bit fragile.
|
2013-04-20 01:12:23 +00:00
|
|
|
|
2018-10-19 00:37:47 +00:00
|
|
|
NOTE TO PEOPLE WHO THINK THAT FreeBSD 13.x IS SLOW:
|
|
|
|
FreeBSD 13.x has many debugging features turned on, in both the kernel
|
2009-09-03 17:04:42 +00:00
|
|
|
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
|
2012-04-17 20:35:54 +00:00
|
|
|
machines to maximize performance. (To completely disable malloc
|
|
|
|
debugging, define MALLOC_PRODUCTION in /etc/make.conf, or to merely
|
|
|
|
disable the most expensive debugging functionality run
|
2016-10-19 15:13:08 +00:00
|
|
|
"ln -s 'abort:false,junk:false' /etc/malloc.conf".)
|
2016-10-19 13:45:51 +00:00
|
|
|
|
2019-06-21 00:16:30 +00:00
|
|
|
20190620:
|
2019-06-21 00:33:45 +00:00
|
|
|
Entropy collection and the /dev/random device are no longer optional
|
|
|
|
components. The "device random" option has been removed.
|
2019-06-21 00:16:30 +00:00
|
|
|
Implementations of distilling algorithms can still be made loadable
|
|
|
|
with "options RANDOM_LOADABLE" (e.g., random_fortuna.ko).
|
|
|
|
|
2019-06-12 21:10:37 +00:00
|
|
|
20190612:
|
|
|
|
Clang, llvm, lld, lldb, compiler-rt, libc++, libunwind and openmp have
|
|
|
|
been upgraded to 8.0.1. Please see the 20141231 entry below for
|
|
|
|
information about prerequisites and upgrading, if you are not already
|
|
|
|
using clang 3.5.0 or higher.
|
|
|
|
|
Fix dpcpu and vnet panics with complex types at the end of the section.
Apply a linker script when linking i386 kernel modules to apply padding
to a set_pcpu or set_vnet section. The padding value is kind-of random
and is used to catch modules not compiled with the linker-script, so
possibly still having problems leading to kernel panics.
This is needed as the code generated on certain architectures for
non-simple-types, e.g., an array can generate an absolute relocation
on the edge (just outside) the section and thus will not be properly
relocated. Adding the padding to the end of the section will ensure
that even absolute relocations of complex types will be inside the
section, if they are the last object in there and hence relocation will
work properly and avoid panics such as observed with carp.ko or ipsec.ko.
There is a rather lengthy discussion of various options to apply in
the mentioned PRs and their depends/blocks, and the review.
There seems no best solution working across multiple toolchains and
multiple version of them, so I took the liberty of taking one,
as currently our users (and our CI system) are hitting this on
just i386 and we need some solution. I wish we would have a proper
fix rather than another "hack".
Also backout r340009 which manually, temporarily fixed CARP before 12.0-R
"by chance" after a lead-up of various other link-elf.c and related fixes.
PR: 230857,238012
With suggestions from: arichardson (originally last year)
Tested by: lwhsu
Event: Waterloo Hackathon 2019
Reported by: lwhsu, olivier
MFC after: 6 weeks
Differential Revision: https://reviews.freebsd.org/D17512
2019-06-08 17:44:42 +00:00
|
|
|
20190608:
|
|
|
|
A fix was applied to i386 kernel modules to avoid panics with
|
|
|
|
dpcpu or vnet. Users need to recompile i386 kernel modules
|
|
|
|
having pcpu or vnet sections or they will refuse to load.
|
|
|
|
|
2019-05-13 18:48:08 +00:00
|
|
|
20190513:
|
|
|
|
User-wired pages now have their own counter,
|
|
|
|
vm.stats.vm.v_user_wire_count. The vm.max_wired sysctl was renamed
|
|
|
|
to vm.max_user_wired and changed from an unsigned int to an unsigned
|
|
|
|
long. bhyve VMs wired with the -S are now subject to the user
|
|
|
|
wiring limit; the vm.max_user_wired sysctl may need to be tuned to
|
|
|
|
avoid running into the limit.
|
|
|
|
|
Remove IPSEC from GENERIC due to performance issues
Having IPSEC compiled into the kernel imposes a non-trivial
performance penalty on multi-threaded workloads due to IPSEC
refcounting. In my benchmarks of multi-threaded UDP
transmit (connected sockets), I've seen a roughly 20% performance
penalty when the IPSEC option is included in the kernel (16.8Mpps
vs 13.8Mpps with 32 senders on a 14 core / 28 HTT Xeon
2697v3)). This is largely due to key_addref() incrementing and
decrementing an atomic reference count on the default
policy. This cause all CPUs to stall on the same cacheline, as it
bounces between different CPUs.
Given that relatively few users use ipsec, and that it can be
loaded as a module, it seems reasonable to ask those users to
load the ipsec module so as to avoid imposing this penalty on the
GENERIC kernel. Its my hope that this will make FreeBSD look
better in "out of the box" benchmark comparisons with other
operating systems.
Many thanks to ae for fixing auto-loading of ipsec.ko when
ifconfig tries to configure ipsec, and to cy for volunteering
to ensure the the racoon ports will load the ipsec.ko module
Reviewed by: cem, cy, delphij, gnn, jhb, jpaetzel
Differential Revision: https://reviews.freebsd.org/D20163
2019-05-09 22:38:15 +00:00
|
|
|
20190507:
|
|
|
|
The IPSEC option has been removed from GENERIC. Users requiring
|
|
|
|
ipsec(4) must now load the ipsec(4) kernel module.
|
|
|
|
|
tun/tap: merge and rename to `tuntap`
tun(4) and tap(4) share the same general management interface and have a lot
in common. Bugs exist in tap(4) that have been fixed in tun(4), and
vice-versa. Let's reduce the maintenance requirements by merging them
together and using flags to differentiate between the three interface types
(tun, tap, vmnet).
This fixes a couple of tap(4)/vmnet(4) issues right out of the gate:
- tap devices may no longer be destroyed while they're open [0]
- VIMAGE issues already addressed in tun by kp
[0] emaste had removed an easy-panic-button in r240938 due to devdrn
blocking. A naive glance over this leads me to believe that this isn't quite
complete -- destroy_devl will only block while executing d_* functions, but
doesn't block the device from being destroyed while a process has it open.
The latter is the intent of the condvar in tun, so this is "fixed" (for
certain definitions of the word -- it wasn't really broken in tap, it just
wasn't quite ideal).
ifconfig(8) also grew the ability to map an interface name to a kld, so
that `ifconfig {tun,tap}0` can continue to autoload the correct module, and
`ifconfig vmnet0 create` will now autoload the correct module. This is a
low overhead addition.
(MFC commentary)
This may get MFC'd if many bugs in tun(4)/tap(4) are discovered after this,
and how critical they are. Changes after this are likely easily MFC'd
without taking this merge, but the merge will be easier.
I have no plans to do this MFC as of now.
Reviewed by: bcr (manpages), tuexen (testing, syzkaller/packetdrill)
Input also from: melifaro
Relnotes: yes
Differential Revision: https://reviews.freebsd.org/D20044
2019-05-08 02:32:11 +00:00
|
|
|
20190507:
|
|
|
|
The tap(4) driver has been folded into tun(4), and the module has been
|
|
|
|
renamed to tuntap. You should update any kld_load="if_tap" or
|
|
|
|
kld_load="if_tun" entries in /etc/rc.conf, if_tap_load="YES" or
|
|
|
|
if_tun_load="YES" entries in /boot/loader.conf to load the if_tuntap
|
|
|
|
module instead, and "device tap" or "device tun" entries in kernel
|
|
|
|
config files to select the tuntap device instead.
|
|
|
|
|
random(4): Restore availability tradeoff prior to r346250
As discussed in that commit message, it is a dangerous default. But the
safe default causes enough pain on a variety of platforms that for now,
restore the prior default.
Some of this is self-induced pain we should/could do better about; for
example, programmatic CI systems and VM managers should introduce entropy
from the host for individual VM instances. This is considered a future work
item.
On modern x86 and Power9 systems, this may be wholly unnecessary after
D19928 lands (even in the non-ideal case where early /boot/entropy is
unavailable), because they have fast hardware random sources available early
in boot. But D19928 is not yet landed and we have a host of architectures
which do not provide fast random sources.
This change adds several tunables and diagnostic sysctls, documented
thoroughly in UPDATING and sys/dev/random/random_infra.c.
PR: 230875 (reopens)
Reported by: adrian, jhb, imp, and probably others
Reviewed by: delphij, imp (earlier version), markm (earlier version)
Discussed with: adrian
Approved by: secteam(delphij)
Relnotes: yeah
Security: related
Differential Revision: https://reviews.freebsd.org/D19944
2019-04-18 20:48:54 +00:00
|
|
|
20190418:
|
|
|
|
The following knobs have been added related to tradeoffs between
|
|
|
|
safe use of the random device and availability in the absence of
|
|
|
|
entropy:
|
|
|
|
|
|
|
|
kern.random.initial_seeding.bypass_before_seeding: tunable; set
|
|
|
|
non-zero to bypass the random device prior to seeding, or zero to
|
|
|
|
block random requests until the random device is initially seeded.
|
|
|
|
For now, set to 1 (unsafe) by default to restore pre-r346250 boot
|
|
|
|
availability properties.
|
|
|
|
|
|
|
|
kern.random.initial_seeding.read_random_bypassed_before_seeding:
|
|
|
|
read-only diagnostic sysctl that is set when bypass is enabled and
|
|
|
|
read_random(9) is bypassed, to enable programmatic handling of this
|
|
|
|
initial condition, if desired.
|
|
|
|
|
|
|
|
kern.random.initial_seeding.arc4random_bypassed_before_seeding:
|
|
|
|
Similar to the above, but for for arc4random(9) initial seeding.
|
|
|
|
|
|
|
|
kern.random.initial_seeding.disable_bypass_warnings: tunable; set
|
|
|
|
non-zero to disable warnings in dmesg when the same conditions are
|
|
|
|
met as for the diagnostic sysctls above. Defaults to zero, i.e.,
|
|
|
|
produce warnings in dmesg when the conditions are met.
|
|
|
|
|
2019-04-16 17:12:17 +00:00
|
|
|
20190416:
|
|
|
|
The loadable random module KPI has changed; the random_infra_init()
|
|
|
|
routine now requires a 3rd function pointer for a bool (*)(void)
|
|
|
|
method that returns true if the random device is seeded (and
|
|
|
|
therefore unblocked).
|
|
|
|
|
2019-04-04 23:40:30 +00:00
|
|
|
20190404:
|
|
|
|
r345895 reverts r320698. This implies that an nfsuserd(8) daemon
|
|
|
|
built from head sources between r320757 (July 6, 2017) and
|
|
|
|
r338192 (Aug. 22, 2018) will not work unless the "-use-udpsock"
|
|
|
|
is added to the command line.
|
|
|
|
nfsuserd daemons built from head sources that are post-r338192 are
|
|
|
|
not affected and should continue to work.
|
|
|
|
|
2019-03-20 21:48:43 +00:00
|
|
|
20190320:
|
|
|
|
The fuse(4) module has been renamed to fusefs(4) for consistency with
|
|
|
|
other filesystems. You should update any kld_load="fuse" entries in
|
|
|
|
/etc/rc.conf, fuse_load="YES" entries in /boot/loader.conf, and
|
2019-05-11 19:31:54 +00:00
|
|
|
"options FUSE" entries in kernel config files.
|
2019-03-20 21:48:43 +00:00
|
|
|
|
2019-03-04 19:23:11 +00:00
|
|
|
20190304:
|
2019-01-20 18:34:30 +00:00
|
|
|
Clang, llvm, lld, lldb, compiler-rt and libc++ have been upgraded to
|
|
|
|
8.0.0. Please see the 20141231 entry below for information about
|
|
|
|
prerequisites and upgrading, if you are not already using clang 3.5.0
|
|
|
|
or higher.
|
|
|
|
|
2019-03-23 10:13:01 +00:00
|
|
|
20190226:
|
|
|
|
geom_uzip(4) depends on the new module xz. If geom_uzip is statically
|
|
|
|
compiled into your custom kernel, add 'device xz' statement to the
|
|
|
|
kernel config.
|
|
|
|
|
2019-02-19 19:37:09 +00:00
|
|
|
20190219:
|
|
|
|
drm and drm2 have been removed from the tree. Please see
|
|
|
|
https://wiki.freebsd.org/Graphics for the latest information on
|
|
|
|
migrating to the drm ports.
|
|
|
|
|
2019-01-31 19:05:56 +00:00
|
|
|
20190131:
|
|
|
|
Iflib is no longer unconditionally compiled into the kernel. Drivers
|
|
|
|
using iflib and statically compiled into the kernel, now require
|
|
|
|
the 'device iflib' config option. For the same drivers loaded as
|
|
|
|
modules on kernels not having 'device iflib', the iflib.ko module
|
|
|
|
is loaded automatically.
|
|
|
|
|
2019-02-14 09:21:19 +00:00
|
|
|
20190125:
|
|
|
|
The IEEE80211_AMPDU_AGE and AH_SUPPORT_AR5416 kernel configuration
|
|
|
|
options no longer exist since r343219 and r343427 respectively;
|
|
|
|
nothing uses them, so they should be just removed from custom
|
|
|
|
kernel config files.
|
|
|
|
|
2018-12-31 00:15:05 +00:00
|
|
|
20181230:
|
|
|
|
r342635 changes the way efibootmgr(8) works by requiring users to add
|
|
|
|
the -b (bootnum) parameter for commands where the bootnum was previously
|
|
|
|
specified with each option. For example 'efibootmgr -B 0001' is now
|
|
|
|
'efibootmgr -B -b 0001'.
|
|
|
|
|
2018-12-20 22:26:54 +00:00
|
|
|
20181220:
|
|
|
|
r342286 modifies the NFSv4 server so that it obeys vfs.nfsd.nfs_privport
|
|
|
|
in the same as it is applied to NFSv2 and 3. This implies that NFSv4
|
|
|
|
servers that have vfs.nfsd.nfs_privport set will only allow mounts
|
|
|
|
from clients using a reserved port#. Since both the FreeBSD and Linux
|
|
|
|
NFSv4 clients use reserved port#s by default, this should not affect
|
|
|
|
most NFSv4 mounts.
|
|
|
|
|
2018-12-19 22:56:31 +00:00
|
|
|
20181219:
|
2018-12-19 22:54:29 +00:00
|
|
|
The XLP config has been removed. We can't support 64-bit atomics in this
|
|
|
|
kernel because it is running in 32-bit mode. XLP users must transition
|
|
|
|
to running a 64-bit kernel (XLP64 or XLPN32).
|
|
|
|
|
|
|
|
The mips GXEMUL support has been removed from FreeBSD. MALTA* + qemu is
|
|
|
|
the preferred emulator today and we don't need two different ones.
|
|
|
|
|
2018-12-19 22:54:03 +00:00
|
|
|
The old sibyte / swarm / Broadcom BCM1250 support has been
|
|
|
|
removed from the mips port.
|
|
|
|
|
2018-12-13 06:59:55 +00:00
|
|
|
20181211:
|
2018-12-11 17:39:49 +00:00
|
|
|
Clang, llvm, lld, lldb, compiler-rt and libc++ have been upgraded to
|
|
|
|
7.0.1. Please see the 20141231 entry below for information about
|
|
|
|
prerequisites and upgrading, if you are not already using clang 3.5.0
|
|
|
|
or higher.
|
|
|
|
|
2018-12-15 21:34:40 +00:00
|
|
|
20181211:
|
|
|
|
Remove the timed and netdate programs from the base tree. Setting
|
2019-05-11 19:31:54 +00:00
|
|
|
the time with these daemons has been obsolete for over a decade.
|
2018-12-15 21:34:40 +00:00
|
|
|
|
2018-11-26 17:22:23 +00:00
|
|
|
20181126:
|
|
|
|
On amd64, arm64 and armv7 (architectures that install LLVM's ld.lld
|
|
|
|
linker as /usr/bin/ld) GNU ld is no longer installed as ld.bfd, as
|
|
|
|
it produces broken binaries when ifuncs are in use. Users needing
|
|
|
|
GNU ld should install the binutils port or package.
|
|
|
|
|
2018-11-23 16:45:07 +00:00
|
|
|
20181123:
|
|
|
|
The BSD crtbegin and crtend code has been enabled by default. It has
|
|
|
|
had extensive testing on amd64, arm64, and i386. It can be disabled
|
|
|
|
by building a world with -DWITHOUT_BSD_CRTBEGIN.
|
|
|
|
|
2018-11-14 08:45:48 +00:00
|
|
|
20181115:
|
|
|
|
The set of CTM commands (ctm, ctm_smail, ctm_rmail, ctm_dequeue)
|
|
|
|
has been converted to a port (misc/ctm) and will be removed from
|
|
|
|
FreeBSD-13. It is available as a package (ctm) for all supported
|
|
|
|
FreeBSD versions.
|
|
|
|
|
2018-11-10 16:58:07 +00:00
|
|
|
20181110:
|
|
|
|
The default newsyslog.conf(5) file has been changed to only include
|
|
|
|
files in /etc/newsyslog.conf.d/ and /usr/local/etc/newsyslog.conf.d/ if
|
|
|
|
the filenames end in '.conf' and do not begin with a '.'.
|
|
|
|
|
|
|
|
You should check the configuration files in these two directories match
|
|
|
|
this naming convention. You can verify which configuration files are
|
|
|
|
being included using the command:
|
|
|
|
$ newsyslog -Nrv
|
|
|
|
|
2018-10-20 19:14:46 +00:00
|
|
|
20181015:
|
|
|
|
Ports for the DRM modules have been simplified. Now, amd64 users should
|
|
|
|
just install the drm-kmod port. All others should install
|
|
|
|
drm-legacy-kmod.
|
|
|
|
|
|
|
|
Graphics hardware that's newer than about 2010 usually works with
|
|
|
|
drm-kmod. For hardware older than 2013, however, some users will need
|
|
|
|
to use drm-legacy-kmod if drm-kmod doesn't work for them. Hardware older
|
|
|
|
than 2008 usually only works in drm-legacy-kmod. The graphics team can
|
|
|
|
only commit to hardware made since 2013 due to the complexity of the
|
|
|
|
market and difficulty to test all the older cards effectively. If you
|
|
|
|
have hardware supported by drm-kmod, you are strongly encouraged to use
|
|
|
|
that as you will get better support.
|
|
|
|
|
|
|
|
Other than KPI chasing, drm-legacy-kmod will not be updated. As outlined
|
|
|
|
elsewhere, the drm and drm2 modules will be eliminated from the src base
|
2019-03-09 17:17:55 +00:00
|
|
|
soon (with a limited exception for arm). Please update to the package
|
|
|
|
asap and report any issues to x11@freebsd.org.
|
2018-10-20 19:14:46 +00:00
|
|
|
|
|
|
|
Generally, anybody using the drm*-kmod packages should add
|
|
|
|
WITHOUT_DRM_MODULE=t and WITHOUT_DRM2_MODULE=t to avoid nasty
|
|
|
|
cross-threading surprises, especially with automatic driver
|
|
|
|
loading from X11 startup. These will become the defaults in 13-current
|
|
|
|
shortly.
|
|
|
|
|
2018-11-27 12:11:16 +00:00
|
|
|
20181012:
|
|
|
|
The ixlv(4) driver has been renamed to iavf(4). As a consequence,
|
|
|
|
custom kernel and module loading configuration files must be updated
|
|
|
|
accordingly. Moreover, interfaces previous presented as ixlvN to the
|
|
|
|
system are now exposed as iavfN and network configuration files must
|
|
|
|
be adjusted as necessary.
|
|
|
|
|
2018-10-09 21:28:26 +00:00
|
|
|
20181009:
|
|
|
|
OpenSSL has been updated to version 1.1.1. This update included
|
2019-05-11 19:31:54 +00:00
|
|
|
additional various API changes throughout the base system. It is
|
2018-10-09 21:28:26 +00:00
|
|
|
important to rebuild third-party software after upgrading. The value
|
|
|
|
of __FreeBSD_version has been bumped accordingly.
|
|
|
|
|
2018-10-07 01:53:43 +00:00
|
|
|
20181006:
|
|
|
|
The legacy DRM modules and drivers have now been added to the loader's
|
|
|
|
module blacklist, in favor of loading them with kld_list in rc.conf(5).
|
|
|
|
The module blacklist may be overridden with the loader.conf(5)
|
|
|
|
'module_blacklist' variable, but loading them via rc.conf(5) is strongly
|
|
|
|
encouraged.
|
|
|
|
|
2018-10-02 21:36:00 +00:00
|
|
|
20181002:
|
|
|
|
The cam(4) based nda(4) driver will be used over nvd(4) by default on
|
|
|
|
powerpc64. You may set 'options NVME_USE_NVD=1' in your kernel conf or
|
|
|
|
loader tunable 'hw.nvme.use_nvd=1' if you wish to use the existing
|
|
|
|
driver. Make sure to edit /boot/etc/kboot.conf and fstab to use the
|
|
|
|
nda device name.
|
|
|
|
|
2018-09-26 15:07:06 +00:00
|
|
|
20180913:
|
|
|
|
Reproducible build mode is now on by default, in preparation for
|
|
|
|
FreeBSD 12.0. This eliminates build metadata such as the user,
|
|
|
|
host, and time from the kernel (and uname), unless the working tree
|
|
|
|
corresponds to a modified checkout from a version control system.
|
|
|
|
The previous behavior can be obtained by setting the /etc/src.conf
|
|
|
|
knob WITHOUT_REPRODUCIBLE_BUILD.
|
|
|
|
|
2018-08-26 12:51:46 +00:00
|
|
|
20180826:
|
|
|
|
The Yarrow CSPRNG has been removed from the kernel as it has not been
|
|
|
|
supported by its designers since at least 2003. Fortuna has been the
|
|
|
|
default since FreeBSD-11.
|
|
|
|
|
2018-08-29 17:37:23 +00:00
|
|
|
20180822:
|
|
|
|
devctl freeze/thaw have gone into the tree, the rc scripts have been
|
2018-08-23 05:06:31 +00:00
|
|
|
updated to use them and devmatch has been changed. You should update
|
|
|
|
kernel, userland and rc scripts all at the same time.
|
|
|
|
|
2018-08-19 14:26:33 +00:00
|
|
|
20180818:
|
|
|
|
The default interpreter has been switched from 4th to Lua.
|
|
|
|
LOADER_DEFAULT_INTERP, documented in build(7), will override the default
|
|
|
|
interpreter. If you have custom FORTH code you will need to set
|
|
|
|
LOADER_DEFAULT_INTERP=4th (valid values are 4th, lua or simp) in
|
|
|
|
src.conf for the build. This will create default hard links between
|
|
|
|
loader and loader_4th instead of loader and loader_lua, the new default.
|
|
|
|
If you are using UEFI it will create the proper hard link to loader.efi.
|
|
|
|
|
2018-08-19 18:18:19 +00:00
|
|
|
bhyve uses userboot.so. It remains 4th-only until some issues are solved
|
|
|
|
regarding coexisting with multiple versions of FreeBSD are resolved.
|
|
|
|
|
2018-08-15 19:28:48 +00:00
|
|
|
20180815:
|
|
|
|
ls(1) now respects the COLORTERM environment variable used in other
|
|
|
|
systems and software to indicate that a colored terminal is both
|
|
|
|
supported and desired. If ls(1) is suddenly emitting colors, they may
|
2018-08-17 04:15:51 +00:00
|
|
|
be disabled again by either removing the unwanted COLORTERM from your
|
|
|
|
environment, or using `ls --color=never`. The ls(1) specific CLICOLOR
|
|
|
|
may not be observed in a future release.
|
2018-08-15 19:28:48 +00:00
|
|
|
|
2018-08-08 19:24:20 +00:00
|
|
|
20180808:
|
|
|
|
The default pager for most commands has been changed to "less". To
|
|
|
|
restore the old behavior, set PAGER="more" and MANPAGER="more -s" in
|
|
|
|
your environment.
|
|
|
|
|
2018-08-01 08:24:34 +00:00
|
|
|
20180731:
|
|
|
|
The jedec_ts(4) driver has been removed. A superset of its functionality
|
|
|
|
is available in the jedec_dimm(4) driver, and the manpage for that
|
|
|
|
driver includes migration instructions. If you have "device jedec_ts"
|
|
|
|
in your kernel configuration file, it must be removed.
|
|
|
|
|
2018-07-30 17:54:18 +00:00
|
|
|
20180730:
|
|
|
|
amd64/GENERIC now has EFI runtime services, EFIRT, enabled by default.
|
2018-07-30 17:59:18 +00:00
|
|
|
This should have no effect if the kernel is booted via BIOS/legacy boot.
|
2018-07-30 18:13:20 +00:00
|
|
|
EFIRT may be disabled via a loader tunable, efi.rt.disabled, if a system
|
2018-07-30 17:54:18 +00:00
|
|
|
has a buggy firmware that prevents a successful boot due to use of
|
|
|
|
runtime services.
|
|
|
|
|
2018-07-27 21:40:05 +00:00
|
|
|
20180727:
|
|
|
|
Atmel AT91RM9200 and AT91SAM9, Cavium CNS 11xx and XScale
|
|
|
|
support has been removed from the tree. These ports were
|
|
|
|
obsolete and/or known to be broken for many years.
|
|
|
|
|
Finalize the boot manager protocol support for next-stage boot
loading.
If we are booting in a conforming UEFI Boot Manager Environment, then
use the BootCurrent variable to find the BootXXXX we're using. Once we
find that, then if it contains more than one EFI_DEVICE_PATH in its
what to boot section, try to use the last one as the kernel to
load. This will also set the default root partition as well. If
there's only one path, or if there's an error along the way, assume
that nothing specific was specified and revert to the old
algorithm. If something was specified, but not found, then fail the
boot. Otherwise you that, specific thing. On FreeBSD, this can be set
using efibootmgr -l <loader> -k <kernel>. We try a few variations of
kernel to cope with the fact that UEFI comes from a DOS world where
paths might be upper case and/or contain back-slashes.
Note: In an ideal world, we'd work out where we are in chain loading
by looking at the passed-in image handle and doing name
matching. However, that's unreliable since at least boot1.efi booted
images don't have that, hence the assumption that loader.efi needs to
load the last thing on the list, if possible.
The reason we fail for something specific is so that we can fully
participate in the UEFI Boot Manager Protocol and fail over to the
next item in the list of BootOrder choices when something goes wrong
at this stage.
This implements was was talked about in freebsd-arch@ last year
https://docs.freebsd.org/cgi/getmsg.cgi?fetch=3576+0+archive/2017/freebsd-arch/20171022.freebsd-arch
and documented in full (after changed resulting from the discussion) in
https://docs.google.com/document/d/1aK9IqF-60JPEbUeSAUAkYjF2W_8EnmczFs6RqCT90Jg/edit#
although one or two minor details may have been modified in this
implementation to make it work, and the ZFS MEDIA PATH extension isn't
implemented. This does not yet move things to ESP:\efi\freebsd\loader.efi.
RelNotes: Yes
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D16403
2018-07-23 20:36:59 +00:00
|
|
|
20180723:
|
|
|
|
loader.efi has been augmented to participate more fully in the
|
2018-07-23 22:15:36 +00:00
|
|
|
UEFI boot manager protocol. loader.efi will now look at the
|
|
|
|
BootXXXX environment variable to determine if a specific kernel
|
|
|
|
or root partition was specified. XXXX is derived from BootCurrent.
|
|
|
|
efibootmgr(8) manages these standard UEFI variables.
|
Finalize the boot manager protocol support for next-stage boot
loading.
If we are booting in a conforming UEFI Boot Manager Environment, then
use the BootCurrent variable to find the BootXXXX we're using. Once we
find that, then if it contains more than one EFI_DEVICE_PATH in its
what to boot section, try to use the last one as the kernel to
load. This will also set the default root partition as well. If
there's only one path, or if there's an error along the way, assume
that nothing specific was specified and revert to the old
algorithm. If something was specified, but not found, then fail the
boot. Otherwise you that, specific thing. On FreeBSD, this can be set
using efibootmgr -l <loader> -k <kernel>. We try a few variations of
kernel to cope with the fact that UEFI comes from a DOS world where
paths might be upper case and/or contain back-slashes.
Note: In an ideal world, we'd work out where we are in chain loading
by looking at the passed-in image handle and doing name
matching. However, that's unreliable since at least boot1.efi booted
images don't have that, hence the assumption that loader.efi needs to
load the last thing on the list, if possible.
The reason we fail for something specific is so that we can fully
participate in the UEFI Boot Manager Protocol and fail over to the
next item in the list of BootOrder choices when something goes wrong
at this stage.
This implements was was talked about in freebsd-arch@ last year
https://docs.freebsd.org/cgi/getmsg.cgi?fetch=3576+0+archive/2017/freebsd-arch/20171022.freebsd-arch
and documented in full (after changed resulting from the discussion) in
https://docs.google.com/document/d/1aK9IqF-60JPEbUeSAUAkYjF2W_8EnmczFs6RqCT90Jg/edit#
although one or two minor details may have been modified in this
implementation to make it work, and the ZFS MEDIA PATH extension isn't
implemented. This does not yet move things to ESP:\efi\freebsd\loader.efi.
RelNotes: Yes
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D16403
2018-07-23 20:36:59 +00:00
|
|
|
|
2018-07-20 05:26:36 +00:00
|
|
|
20180720:
|
|
|
|
zfsloader's functionality has now been folded into loader.
|
2018-07-30 00:04:42 +00:00
|
|
|
zfsloader is no longer necessary once you've updated your
|
2018-07-23 06:04:05 +00:00
|
|
|
boot blocks. For a transition period, we will install a
|
|
|
|
hardlink for zfsloader to loader to allow a smooth transition
|
|
|
|
until the boot blocks can be updated (hard link because old
|
|
|
|
zfs boot blocks don't understand symlinks).
|
2018-07-20 05:26:36 +00:00
|
|
|
|
2018-07-28 20:33:08 +00:00
|
|
|
20180719:
|
|
|
|
ARM64 now have efifb support, if you want to have serial console
|
|
|
|
on your arm64 board when an screen is connected and the bootloader
|
2019-05-11 19:31:54 +00:00
|
|
|
setup a frame buffer for us to use, just add :
|
2018-07-28 20:33:08 +00:00
|
|
|
boot_serial=YES
|
|
|
|
boot_multicons=YES
|
|
|
|
in /boot/loader.conf
|
|
|
|
For Raspberry Pi 3 (RPI) users, this is needed even if you don't have
|
2019-05-11 19:31:54 +00:00
|
|
|
an screen connected as the firmware will setup a frame buffer are that
|
|
|
|
u-boot will expose as an EFI frame buffer.
|
2018-07-28 20:33:08 +00:00
|
|
|
|
2018-07-20 18:02:52 +00:00
|
|
|
20180719:
|
|
|
|
New uid:gid added, ntpd:ntpd (123:123). Be sure to run mergemaster
|
|
|
|
or take steps to update /etc/passwd before doing installworld on
|
2018-09-13 15:16:05 +00:00
|
|
|
existing systems. Do not skip the "mergemaster -Fp" step before
|
|
|
|
installworld, as described in the update procedures near the bottom
|
|
|
|
of this document. Also, rc.d/ntpd now starts ntpd(8) as user ntpd
|
2018-07-20 18:02:52 +00:00
|
|
|
if the new mac_ntpd(4) policy is available, unless ntpd_flags or
|
|
|
|
the ntp config file contain options that change file/dir locations.
|
|
|
|
When such options (e.g., "statsdir" or "crypto") are used, ntpd can
|
|
|
|
still be run as non-root by setting ntpd_user=ntpd in rc.conf, after
|
|
|
|
taking steps to ensure that all required files/dirs are accessible
|
|
|
|
by the ntpd user.
|
|
|
|
|
2018-07-19 23:54:18 +00:00
|
|
|
20180717:
|
|
|
|
Big endian arm support has been removed.
|
|
|
|
|
2018-07-12 02:51:50 +00:00
|
|
|
20180711:
|
|
|
|
The static environment setup in kernel configs is no longer mutually
|
|
|
|
exclusive with the loader(8) environment by default. In order to
|
|
|
|
restore the previous default behavior of disabling the loader(8)
|
|
|
|
environment if a static environment is present, you must specify
|
|
|
|
loader_env.disabled=1 in the static environment.
|
|
|
|
|
2018-07-05 13:13:48 +00:00
|
|
|
20180705:
|
|
|
|
The ABI of syscalls used by management tools like sockstat and
|
|
|
|
netstat has been broken to allow 32-bit binaries to work on
|
|
|
|
64-bit kernels without modification. These programs will need
|
|
|
|
to match the kernel in order to function. External programs may
|
|
|
|
require minor modifications to accommodate a change of type in
|
|
|
|
structures from pointers to 64-bit virtual addresses.
|
|
|
|
|
2018-07-02 19:48:38 +00:00
|
|
|
20180702:
|
|
|
|
On i386 and amd64 atomics are now inlined. Out of tree modules using
|
|
|
|
atomics will need to be rebuilt.
|
|
|
|
|
2018-07-01 17:28:46 +00:00
|
|
|
20180701:
|
|
|
|
The '%I' format in the kern.corefile sysctl limits the number of
|
|
|
|
core files that a process can generate to the number stored in the
|
|
|
|
debug.ncores sysctl. The '%I' format is replaced by the single digit
|
|
|
|
index. Previously, if all indexes were taken the kernel would overwrite
|
|
|
|
only a core file with the highest index in a filename.
|
|
|
|
Currently the system will create a new core file if there is a free
|
|
|
|
index or if all slots are taken it will overwrite the oldest one.
|
2016-10-19 13:45:51 +00:00
|
|
|
|
2018-06-30 15:03:22 +00:00
|
|
|
20180630:
|
|
|
|
Clang, llvm, lld, lldb, compiler-rt and libc++ have been upgraded to
|
|
|
|
6.0.1. Please see the 20141231 entry below for information about
|
|
|
|
prerequisites and upgrading, if you are not already using clang 3.5.0
|
|
|
|
or higher.
|
|
|
|
|
2018-07-01 01:45:41 +00:00
|
|
|
20180628:
|
|
|
|
r335753 introduced a new quoting method. However, etc/devd/devmatch.conf
|
|
|
|
needed to be changed to work with it. This change was made with r335763
|
2019-03-09 17:17:55 +00:00
|
|
|
and requires a mergemaster / etcupdate / etc to update the installed
|
|
|
|
file.
|
2018-07-01 01:45:41 +00:00
|
|
|
|
2018-06-12 22:52:27 +00:00
|
|
|
20180612:
|
|
|
|
r334930 changed the interface between the NFS modules, so they all
|
|
|
|
need to be rebuilt. r335018 did a __FreeBSD_version bump for this.
|
|
|
|
|
2018-09-26 14:32:36 +00:00
|
|
|
20180530:
|
|
|
|
As of r334391 lld is the default amd64 system linker; it is installed
|
|
|
|
as /usr/bin/ld. Kernel build workarounds (see 20180510 entry) are no
|
|
|
|
longer necessary.
|
|
|
|
|
2018-05-31 02:58:08 +00:00
|
|
|
20180530:
|
|
|
|
The kernel / userland interface for devinfo changed, so you'll
|
|
|
|
need a new kernel and userland as a pair for it to work (rebuilding
|
|
|
|
lib/libdevinfo is all that's required). devinfo and devmatch will
|
|
|
|
not work, but everything else will when there's a mismatch.
|
|
|
|
|
2018-05-23 17:30:23 +00:00
|
|
|
20180523:
|
|
|
|
The on-disk format for hwpmc callchain records has changed to include
|
2018-05-24 17:08:55 +00:00
|
|
|
threadid corresponding to a given record. This changes the field offsets
|
2018-05-23 17:30:23 +00:00
|
|
|
and thus requires that libpmcstat be rebuilt before using a kernel
|
|
|
|
later than r334108.
|
|
|
|
|
2018-05-17 14:55:41 +00:00
|
|
|
20180517:
|
|
|
|
The vxge(4) driver has been removed. This driver was introduced into
|
|
|
|
HEAD one week before the Exar left the Ethernet market and is not
|
|
|
|
known to be used. If you have device vxge in your kernel config file
|
|
|
|
it must be removed.
|
|
|
|
|
2018-05-16 21:07:12 +00:00
|
|
|
20180510:
|
2018-05-16 13:52:24 +00:00
|
|
|
The amd64 kernel now requires a ld that supports ifunc to produce a
|
|
|
|
working kernel, either lld or a newer binutils. lld is built by default
|
|
|
|
on amd64, and the 'buildkernel' target uses it automatically. However,
|
|
|
|
it is not the default linker, so building the kernel the traditional
|
|
|
|
way requires LD=ld.lld on the command line (or LD=/usr/local/bin/ld for
|
|
|
|
binutils port/package). lld will soon be default, and this requirement
|
|
|
|
will go away.
|
|
|
|
|
2018-09-26 14:32:36 +00:00
|
|
|
NOTE: As of r334391 lld is the default system linker on amd64, and no
|
|
|
|
workaround is necessary.
|
|
|
|
|
2018-05-08 21:14:29 +00:00
|
|
|
20180508:
|
|
|
|
The nxge(4) driver has been removed. This driver was for PCI-X 10g
|
2019-05-11 19:31:54 +00:00
|
|
|
cards made by s2io/Neterion. The company was acquired by Exar and
|
2018-05-08 21:14:29 +00:00
|
|
|
no longer sells or supports Ethernet products. If you have device
|
|
|
|
nxge in your kernel config file it must be removed.
|
|
|
|
|
2018-05-04 10:52:17 +00:00
|
|
|
20180504:
|
|
|
|
The tz database (tzdb) has been updated to 2018e. This version more
|
|
|
|
correctly models time stamps in time zones with negative DST such as
|
|
|
|
Europe/Dublin (from 1971 on), Europe/Prague (1946/7), and
|
|
|
|
Africa/Windhoek (1994/2017). This does not affect the UT offsets, only
|
|
|
|
time zone abbreviations and the tm_isdst flag.
|
|
|
|
|
2018-05-02 15:59:15 +00:00
|
|
|
20180502:
|
|
|
|
The ixgb(4) driver has been removed. This driver was for an early and
|
|
|
|
uncommon legacy PCI 10GbE for a single ASIC, Intel 82597EX. Intel
|
|
|
|
quickly shifted to the long lived ixgbe family. If you have device
|
|
|
|
ixgb in your kernel config file it must be removed.
|
|
|
|
|
2018-05-01 16:30:48 +00:00
|
|
|
20180501:
|
|
|
|
The lmc(4) driver has been removed. This was a WAN interface
|
|
|
|
card that was already reportedly rare in 2003, and had an ambiguous
|
|
|
|
license. If you have device lmc in your kernel config file it must
|
|
|
|
be removed.
|
|
|
|
|
2018-04-13 21:18:04 +00:00
|
|
|
20180413:
|
|
|
|
Support for Arcnet networks has been removed. If you have device
|
|
|
|
arcnet or device cm in your kernel config file they must be
|
|
|
|
removed.
|
|
|
|
|
2018-04-11 17:28:24 +00:00
|
|
|
20180411:
|
|
|
|
Support for FDDI networks has been removed. If you have device
|
|
|
|
fddi or device fpa in your kernel config file they must be
|
|
|
|
removed.
|
2018-07-30 00:04:42 +00:00
|
|
|
|
2018-04-06 13:00:45 +00:00
|
|
|
20180406:
|
|
|
|
In addition to supporting RFC 3164 formatted messages, the
|
|
|
|
syslogd(8) service is now capable of parsing RFC 5424 formatted
|
|
|
|
log messages. The main benefit of using RFC 5424 is that clients
|
|
|
|
may now send log messages with timestamps containing year numbers,
|
|
|
|
microseconds and time zone offsets.
|
|
|
|
|
|
|
|
Similarly, the syslog(3) C library function has been altered to
|
|
|
|
send RFC 5424 formatted messages to the local system logging
|
|
|
|
daemon. On systems using syslogd(8), this change should have no
|
|
|
|
negative impact, as long as syslogd(8) and the C library are
|
|
|
|
updated at the same time. On systems using a different system
|
|
|
|
logging daemon, it may be necessary to make configuration
|
|
|
|
adjustments, depending on the software used.
|
|
|
|
|
|
|
|
When using syslog-ng, add the 'syslog-protocol' flag to local
|
|
|
|
input sources to enable parsing of RFC 5424 formatted messages:
|
|
|
|
|
|
|
|
source src {
|
|
|
|
unix-dgram("/var/run/log" flags(syslog-protocol));
|
|
|
|
}
|
|
|
|
|
|
|
|
When using rsyslog, disable the 'SysSock.UseSpecialParser' option
|
|
|
|
of the 'imuxsock' module to let messages be processed by the
|
|
|
|
regular RFC 3164/5424 parsing pipeline:
|
|
|
|
|
|
|
|
module(load="imuxsock" SysSock.UseSpecialParser="off")
|
|
|
|
|
|
|
|
Do note that these changes only affect communication between local
|
|
|
|
applications and syslogd(8). The format that syslogd(8) uses to
|
|
|
|
store messages on disk or forward messages to other systems
|
|
|
|
remains unchanged. syslogd(8) still uses RFC 3164 for these
|
|
|
|
purposes. Options to customize this behaviour will be added in the
|
|
|
|
future. Utilities that process log files stored in /var/log are
|
|
|
|
thus expected to continue to function as before.
|
|
|
|
|
|
|
|
__FreeBSD_version has been incremented to 1200061 to denote this
|
|
|
|
change.
|
|
|
|
|
2018-03-29 17:44:55 +00:00
|
|
|
20180328:
|
2018-03-28 23:33:26 +00:00
|
|
|
Support for token ring networks has been removed. If you
|
|
|
|
have "device token" in your kernel config you should remove
|
|
|
|
it. No device drivers supported token ring.
|
|
|
|
|
2018-03-25 21:07:59 +00:00
|
|
|
20180323:
|
|
|
|
makefs was modified to be able to tag ISO9660 El Torito boot catalog
|
|
|
|
entries as EFI instead of overloading the i386 tag as done previously.
|
|
|
|
The amd64 mkisoimages.sh script used to build amd64 ISO images for
|
|
|
|
release was updated to use this. This may mean that makefs must be
|
|
|
|
updated before "make cdrom" can be run in the release directory. This
|
|
|
|
should be as simple as:
|
|
|
|
|
|
|
|
$ cd $SRCDIR/usr.sbin/makefs
|
|
|
|
$ make depend all install
|
|
|
|
|
Add Lua as a scripting langauge to /boot/loader
liblua glues the lua run time into the boot loader. It implements all
the runtime routines that lua expects. In addition, it has a few
standard 'C' headers that nueter various aspects of the LUA build that
are too specific to lua to be in libsa. Many refinements from the
original code to improve implementation and the number of included lua
libraries. Use int64_t for lua_Number. Have "/boot/lua" be the default
module path. Numerous cleanups from the original GSoC project,
including hacking libsa to allow lua to be built with only one change
outside luaconf.h.
Add the final bit of lua glue to bring in liblua and plug into the
multiple interpreter framework, previously committed.
Add LOADER_LUA option, currently off by default.
Presently, this is an experimental option. One must opt-in to using
this by defining WITH_LOADER_LUA and WITHOUT_FORTH. It's been
lightly tested, so keep a backup copy of your old loader handy.
The menu code, coming in the next commit, hasn't been exhaustively
tested. A LUA boot loader is 60k larger than a FORTH one, which is
80k larger than a no-interpreter one. Subtle changes in size
may tip things past some subtle limit (the binary is ~430k now
when built with LUA). A future version may offer coexistance.
Bump FreeBSD version to 1200058 to mark the milestone.
Pedro Souza's 2014 Summer of Code project. Rui Paulo, Pedro Arthur,
Zakary Nafziger and Wojciech A. Koszek also contributed. Warner Losh
reworked it extensively into its current form.
Obtained from: https://wiki.freebsd.org/SummerOfCode2014/LuaLoader
Sponsored by: Google Summer of Code
Relnotes: Yes
MFC After: 1 month
Differential Review: https://reviews.freebsd.org/D14295
2018-02-12 15:31:53 +00:00
|
|
|
20180212:
|
|
|
|
FreeBSD boot loader enhanced with Lua scripting. It's purely opt-in for
|
|
|
|
now by building WITH_LOADER_LUA and WITHOUT_FORTH in /etc/src.conf.
|
2019-05-11 19:31:54 +00:00
|
|
|
Co-existence for the transition period will come shortly. Booting is a
|
Add Lua as a scripting langauge to /boot/loader
liblua glues the lua run time into the boot loader. It implements all
the runtime routines that lua expects. In addition, it has a few
standard 'C' headers that nueter various aspects of the LUA build that
are too specific to lua to be in libsa. Many refinements from the
original code to improve implementation and the number of included lua
libraries. Use int64_t for lua_Number. Have "/boot/lua" be the default
module path. Numerous cleanups from the original GSoC project,
including hacking libsa to allow lua to be built with only one change
outside luaconf.h.
Add the final bit of lua glue to bring in liblua and plug into the
multiple interpreter framework, previously committed.
Add LOADER_LUA option, currently off by default.
Presently, this is an experimental option. One must opt-in to using
this by defining WITH_LOADER_LUA and WITHOUT_FORTH. It's been
lightly tested, so keep a backup copy of your old loader handy.
The menu code, coming in the next commit, hasn't been exhaustively
tested. A LUA boot loader is 60k larger than a FORTH one, which is
80k larger than a no-interpreter one. Subtle changes in size
may tip things past some subtle limit (the binary is ~430k now
when built with LUA). A future version may offer coexistance.
Bump FreeBSD version to 1200058 to mark the milestone.
Pedro Souza's 2014 Summer of Code project. Rui Paulo, Pedro Arthur,
Zakary Nafziger and Wojciech A. Koszek also contributed. Warner Losh
reworked it extensively into its current form.
Obtained from: https://wiki.freebsd.org/SummerOfCode2014/LuaLoader
Sponsored by: Google Summer of Code
Relnotes: Yes
MFC After: 1 month
Differential Review: https://reviews.freebsd.org/D14295
2018-02-12 15:31:53 +00:00
|
|
|
complex environment and test coverage for Lua-enabled loaders has been
|
|
|
|
thin, so it would be prudent to assume it might not work and make
|
|
|
|
provisions for backup boot methods.
|
|
|
|
|
2018-02-12 06:42:38 +00:00
|
|
|
20180211:
|
|
|
|
devmatch functionality has been turned on in devd. It will automatically
|
2019-03-09 17:17:55 +00:00
|
|
|
load drivers for unattached devices. This may cause unexpected drivers
|
|
|
|
to be loaded. Please report any problems to current@ and
|
|
|
|
imp@freebsd.org.
|
2018-02-12 06:42:38 +00:00
|
|
|
|
2018-01-14 00:08:34 +00:00
|
|
|
20180114:
|
|
|
|
Clang, llvm, lld, lldb, compiler-rt and libc++ have been upgraded to
|
|
|
|
6.0.0. Please see the 20141231 entry below for information about
|
|
|
|
prerequisites and upgrading, if you are not already using clang 3.5.0
|
|
|
|
or higher.
|
|
|
|
|
2018-01-10 20:49:01 +00:00
|
|
|
20180110:
|
|
|
|
LLVM's lld linker is now used as the FreeBSD/amd64 bootstrap linker.
|
|
|
|
This means it is used to link the kernel and userland libraries and
|
|
|
|
executables, but is not yet installed as /usr/bin/ld by default.
|
|
|
|
|
|
|
|
To revert to ld.bfd as the bootstrap linker, in /etc/src.conf set
|
|
|
|
WITHOUT_LLD_BOOTSTRAP=yes
|
|
|
|
|
2018-01-10 16:52:00 +00:00
|
|
|
20180110:
|
2018-01-12 17:33:20 +00:00
|
|
|
On i386, pmtimer has been removed. Its functionality has been folded
|
2019-03-09 17:17:55 +00:00
|
|
|
into apm. It was a no-op on ACPI in current for a while now (but was
|
|
|
|
still needed on i386 in FreeBSD 11 and earlier). Users may need to
|
|
|
|
remove it from kernel config files.
|
2018-01-10 16:52:00 +00:00
|
|
|
|
2018-01-04 20:05:47 +00:00
|
|
|
20180104:
|
|
|
|
The use of RSS hash from the network card aka flowid has been
|
|
|
|
disabled by default for lagg(4) as it's currently incompatible with
|
|
|
|
the lacp and loadbalance protocols.
|
|
|
|
|
|
|
|
This can be re-enabled by setting the following in loader.conf:
|
|
|
|
net.link.lagg.default_use_flowid="1"
|
|
|
|
|
2018-01-03 06:41:39 +00:00
|
|
|
20180102:
|
|
|
|
The SW_WATCHDOG option is no longer necessary to enable the
|
|
|
|
hardclock-based software watchdog if no hardware watchdog is
|
|
|
|
configured. As before, SW_WATCHDOG will cause the software
|
|
|
|
watchdog to be enabled even if a hardware watchdog is configured.
|
|
|
|
|
2017-12-15 23:19:49 +00:00
|
|
|
20171215:
|
|
|
|
r326887 fixes the issue described in the 20171214 UPDATING entry.
|
|
|
|
r326888 flips the switch back to building GELI support always.
|
|
|
|
|
2017-12-14 17:00:24 +00:00
|
|
|
20171214:
|
|
|
|
r362593 broke ZFS + GELI support for reasons unknown. However,
|
|
|
|
it also broke ZFS support generally, so GELI has been turned off
|
2017-12-16 11:49:30 +00:00
|
|
|
by default as the lesser evil in r326857. If you boot off ZFS and/or
|
2017-12-14 17:00:24 +00:00
|
|
|
GELI, it might not be a good time to update.
|
|
|
|
|
2017-11-26 03:53:20 +00:00
|
|
|
20171125:
|
|
|
|
PowerPC users must update loader(8) by rebuilding world before
|
|
|
|
installing a new kernel, as the protocol connecting them has
|
|
|
|
changed. Without the update, loader metadata will not be passed
|
|
|
|
successfully to the kernel and users will have to enter their
|
|
|
|
root partition at the kernel mountroot prompt to continue booting.
|
|
|
|
Newer versions of loader can boot old kernels without issue.
|
|
|
|
|
2017-11-26 04:55:23 +00:00
|
|
|
20171110:
|
2017-11-10 23:54:41 +00:00
|
|
|
The LOADER_FIREWIRE_SUPPORT build variable as been renamed to
|
2017-11-10 23:54:48 +00:00
|
|
|
WITH/OUT_LOADER_FIREWIRE. LOADER_{NO_,}GELI_SUPPORT has been renamed
|
|
|
|
to WITH/OUT_LOADER_GELI.
|
2017-11-10 23:54:41 +00:00
|
|
|
|
2017-11-06 15:29:33 +00:00
|
|
|
20171106:
|
|
|
|
The naive and non-compliant support of posix_fallocate(2) in ZFS
|
|
|
|
has been removed as of r325320. The system call now returns EINVAL
|
|
|
|
when used on a ZFS file. Although the new behavior complies with the
|
|
|
|
standard, some consumers are not prepared to cope with it.
|
|
|
|
One known victim is lld prior to r325420.
|
|
|
|
|
2017-11-02 18:09:07 +00:00
|
|
|
20171102:
|
|
|
|
Building in a FreeBSD src checkout will automatically create object
|
|
|
|
directories now rather than store files in the current directory if
|
2017-11-17 15:46:19 +00:00
|
|
|
'make obj' was not ran. Calling 'make obj' is no longer necessary.
|
2017-11-02 18:09:07 +00:00
|
|
|
This feature can be disabled by setting WITHOUT_AUTO_OBJ=yes in
|
|
|
|
/etc/src-env.conf (not /etc/src.conf), or passing the option in the
|
|
|
|
environment.
|
|
|
|
|
2017-11-01 21:22:05 +00:00
|
|
|
20171101:
|
|
|
|
The default MAKEOBJDIR has changed from /usr/obj/<srcdir> for native
|
|
|
|
builds, and /usr/obj/<arch>/<srcdir> for cross-builds, to a unified
|
|
|
|
/usr/obj/<srcdir>/<arch>. This behavior can be changed to the old
|
|
|
|
format by setting WITHOUT_UNIFIED_OBJDIR=yes in /etc/src-env.conf,
|
|
|
|
the environment, or with -DWITHOUT_UNIFIED_OBJDIR when building.
|
|
|
|
The UNIFIED_OBJDIR option is a transitional feature that will be
|
|
|
|
removed for 12.0 release; please migrate to the new format for any
|
|
|
|
tools by looking up the OBJDIR used by 'make -V .OBJDIR' means rather
|
|
|
|
than hardcoding paths.
|
|
|
|
|
2017-10-29 01:21:29 +00:00
|
|
|
20171028:
|
|
|
|
The native-xtools target no longer installs the files by default to the
|
|
|
|
OBJDIR. Use the native-xtools-install target with a DESTDIR to install
|
|
|
|
to ${DESTDIR}/${NXTP} where NXTP defaults to /nxb-bin.
|
|
|
|
|
2017-10-22 03:52:26 +00:00
|
|
|
20171021:
|
|
|
|
As part of the boot loader infrastructure cleanup, LOADER_*_SUPPORT
|
|
|
|
options are changing from controlling the build if defined / undefined
|
|
|
|
to controlling the build with explicit 'yes' or 'no' values. They will
|
|
|
|
shift to WITH/WITHOUT options to match other options in the system.
|
|
|
|
|
2017-10-09 22:17:18 +00:00
|
|
|
20171010:
|
|
|
|
libstand has turned into a private library for sys/boot use only.
|
|
|
|
It is no longer supported as a public interface outside of sys/boot.
|
|
|
|
|
2017-10-05 23:05:56 +00:00
|
|
|
20171005:
|
|
|
|
The arm port has split armv6 into armv6 and armv7. armv7 is now
|
2017-10-06 13:48:38 +00:00
|
|
|
a valid TARGET_ARCH/MACHINE_ARCH setting. If you have an armv7 system
|
|
|
|
and are running a kernel from before r324363, you will need to add
|
|
|
|
MACHINE_ARCH=armv7 to 'make buildworld' to do a native build.
|
2017-10-05 23:05:56 +00:00
|
|
|
|
2017-10-04 09:54:55 +00:00
|
|
|
20171003:
|
2017-10-04 03:39:54 +00:00
|
|
|
When building multiple kernels using KERNCONF, non-existent KERNCONF
|
|
|
|
files will produce an error and buildkernel will fail. Previously
|
|
|
|
missing KERNCONF files silently failed giving no indication as to
|
|
|
|
why, only to subsequently discover during installkernel that the
|
|
|
|
desired kernel was never built in the first place.
|
|
|
|
|
2017-09-12 19:36:24 +00:00
|
|
|
20170912:
|
|
|
|
The default serial number format for CTL LUNs has changed. This will
|
|
|
|
affect users who use /dev/diskid/* device nodes, or whose FibreChannel
|
|
|
|
or iSCSI clients care about their LUNs' serial numbers. Users who
|
|
|
|
require serial number stability should hardcode serial numbers in
|
|
|
|
/etc/ctl.conf .
|
|
|
|
|
2017-09-12 17:06:35 +00:00
|
|
|
20170912:
|
|
|
|
For 32-bit arm compiled for hard-float support, soft-floating point
|
|
|
|
binaries now always get their shared libraries from
|
|
|
|
LD_SOFT_LIBRARY_PATH (in the past, this was only used if
|
|
|
|
/usr/libsoft also existed). Only users with a hard-float ld.so, but
|
|
|
|
soft-float everything else should be affected.
|
|
|
|
|
2017-08-26 14:07:24 +00:00
|
|
|
20170826:
|
2017-09-17 10:28:45 +00:00
|
|
|
The geli password typed at boot is now hidden. To restore the previous
|
|
|
|
behavior, see geli(8) for configuration options.
|
2017-08-26 14:07:24 +00:00
|
|
|
|
2017-08-25 19:41:38 +00:00
|
|
|
20170825:
|
|
|
|
Move PMTUD blackhole counters to TCPSTATS and remove them from bare
|
|
|
|
sysctl values. Minor nit, but requires a rebuild of both world/kernel
|
|
|
|
to complete.
|
|
|
|
|
2017-08-14 19:18:50 +00:00
|
|
|
20170814:
|
|
|
|
"make check" behavior (made in ^/head@r295380) has been changed to
|
|
|
|
execute from a limited sandbox, as opposed to executing from
|
|
|
|
${TESTSDIR}.
|
|
|
|
|
|
|
|
Behavioral changes:
|
|
|
|
- The "beforecheck" and "aftercheck" targets are now specified.
|
|
|
|
- ${CHECKDIR} (added in commit noted above) has been removed.
|
|
|
|
- Legacy behavior can be enabled by setting
|
|
|
|
WITHOUT_MAKE_CHECK_USE_SANDBOX in src.conf(5) or the environment.
|
|
|
|
|
|
|
|
If the limited sandbox mode is enabled, "make check" will execute
|
|
|
|
"make distribution", then install, execute the tests, and clean up the
|
|
|
|
sandbox if successful.
|
|
|
|
|
|
|
|
The "make distribution" and "make install" targets are typically run as
|
|
|
|
root to set appropriate permissions and ownership at installation time.
|
|
|
|
The end-user should set "WITH_INSTALL_AS_USER" in src.conf(5) or the
|
|
|
|
environment if executing "make check" with limited sandbox mode using
|
|
|
|
an unprivileged user.
|
|
|
|
|
2017-08-09 05:21:57 +00:00
|
|
|
20170808:
|
|
|
|
Since the switch to GPT disk labels, fsck for UFS/FFS has been
|
|
|
|
unable to automatically find alternate superblocks. As of r322297,
|
|
|
|
the information needed to find alternate superblocks has been
|
|
|
|
moved to the end of the area reserved for the boot block.
|
|
|
|
Filesystems created with a newfs of this vintage or later
|
|
|
|
will create the recovery information. If you have a filesystem
|
|
|
|
created prior to this change and wish to have a recovery block
|
|
|
|
created for your filesystem, you can do so by running fsck in
|
2018-01-12 17:33:20 +00:00
|
|
|
foreground mode (i.e., do not use the -p or -y options). As it
|
2017-08-09 05:21:57 +00:00
|
|
|
starts, fsck will ask ``SAVE DATA TO FIND ALTERNATE SUPERBLOCKS''
|
|
|
|
to which you should answer yes.
|
|
|
|
|
2017-07-28 21:14:28 +00:00
|
|
|
20170728:
|
|
|
|
As of r321665, an NFSv4 server configuration that services
|
|
|
|
Kerberos mounts or clients that do not support the uid/gid in
|
|
|
|
owner/owner_group string capability, must explicitly enable
|
|
|
|
the nfsuserd daemon by adding nfsuserd_enable="YES" to the
|
|
|
|
machine's /etc/rc.conf file.
|
|
|
|
|
2017-07-21 18:53:43 +00:00
|
|
|
20170722:
|
2017-04-17 12:42:58 +00:00
|
|
|
Clang, llvm, lldb, compiler-rt and libc++ have been upgraded to 5.0.0.
|
|
|
|
Please see the 20141231 entry below for information about prerequisites
|
|
|
|
and upgrading, if you are not already using clang 3.5.0 or higher.
|
|
|
|
|
2017-07-01 10:04:42 +00:00
|
|
|
20170701:
|
2017-08-07 21:29:55 +00:00
|
|
|
WITHOUT_RCMDS is now the default. Set WITH_RCMDS if you need the
|
|
|
|
r-commands (rlogin, rsh, etc.) to be built with the base system.
|
2017-07-01 10:04:42 +00:00
|
|
|
|
2017-06-26 02:25:19 +00:00
|
|
|
20170625:
|
|
|
|
The FreeBSD/powerpc platform now uses a 64-bit type for time_t. This is
|
|
|
|
a very major ABI incompatible change, so users of FreeBSD/powerpc must
|
|
|
|
be careful when performing source upgrades. It is best to run
|
|
|
|
'make installworld' from an alternate root system, either a live
|
|
|
|
CD/memory stick, or a temporary root partition. Additionally, all ports
|
|
|
|
must be recompiled. powerpc64 is largely unaffected, except in the case
|
|
|
|
of 32-bit compatibility. All 32-bit binaries will be affected.
|
|
|
|
|
2017-06-24 14:32:57 +00:00
|
|
|
20170623:
|
2017-06-23 18:06:20 +00:00
|
|
|
Forward compatibility for the "ino64" project have been committed. This
|
|
|
|
will allow most new binaries to run on older kernels in a limited
|
|
|
|
fashion. This prevents many of the common foot-shooting actions in the
|
|
|
|
upgrade as well as the limited ability to roll back the kernel across
|
|
|
|
the ino64 upgrade. Complicated use cases may not work properly, though
|
2017-06-24 18:37:51 +00:00
|
|
|
enough simpler ones work to allow recovery in most situations.
|
2017-06-24 14:32:57 +00:00
|
|
|
|
|
|
|
20170620:
|
|
|
|
Switch back to the BSDL dtc (Device Tree Compiler). Set WITH_GPL_DTC
|
|
|
|
if you require the GPL compiler.
|
2017-06-23 18:06:20 +00:00
|
|
|
|
2017-06-18 21:53:17 +00:00
|
|
|
20170618:
|
|
|
|
The internal ABI used for communication between the NFS kernel modules
|
|
|
|
was changed by r320085, so __FreeBSD_version was bumped to
|
|
|
|
ensure all the NFS related modules are updated together.
|
|
|
|
|
2017-06-17 01:09:18 +00:00
|
|
|
20170617:
|
|
|
|
The ABI of struct event was changed by extending the data
|
|
|
|
member to 64bit and adding ext fields. For upgrade, same
|
|
|
|
precautions as for the entry 20170523 "ino64" must be
|
|
|
|
followed.
|
|
|
|
|
2017-06-17 01:27:15 +00:00
|
|
|
20170531:
|
|
|
|
The GNU roff toolchain has been removed from base. To render manpages
|
|
|
|
which are not supported by mandoc(1), man(1) can fallback on GNU roff
|
|
|
|
from ports (and recommends to install it).
|
|
|
|
To render roff(7) documents, consider using GNU roff from ports or the
|
|
|
|
heirloom doctools roff toolchain from ports via pkg install groff or
|
|
|
|
via pkg install heirloom-doctools.
|
|
|
|
|
2017-05-25 05:01:44 +00:00
|
|
|
20170524:
|
|
|
|
The ath(4) and ath_hal(4) modules now build piecemeal to allow for
|
|
|
|
smaller runtime footprint builds. This is useful for embedded systems
|
|
|
|
which only require one chipset support.
|
|
|
|
|
|
|
|
If you load it as a module, make sure this is in /boot/loader.conf:
|
|
|
|
|
|
|
|
if_ath_load="YES"
|
|
|
|
|
|
|
|
This will load the HAL, all chip/RF backends and if_ath_pci.
|
|
|
|
If you have if_ath_pci in /boot/loader.conf, ensure it is after
|
|
|
|
if_ath or it will not load any HAL chipset support.
|
|
|
|
|
|
|
|
If you want to selectively load things (eg on ye cheape ARM/MIPS
|
|
|
|
platforms where RAM is at a premium) you should:
|
|
|
|
|
|
|
|
* load ath_hal
|
|
|
|
* load the chip modules in question
|
|
|
|
* load ath_rate, ath_dfs
|
|
|
|
* load ath_main
|
|
|
|
* load if_ath_pci and/or if_ath_ahb depending upon your particular
|
|
|
|
bus bind type - this is where probe/attach is done.
|
|
|
|
|
|
|
|
For further comments/feedback, poke adrian@ .
|
|
|
|
|
2017-05-23 20:25:49 +00:00
|
|
|
20170523:
|
|
|
|
The "ino64" 64-bit inode project has been committed, which extends
|
2017-05-24 17:11:10 +00:00
|
|
|
a number of types to 64 bits. Upgrading in place requires care and
|
|
|
|
adherence to the documented upgrade procedure.
|
|
|
|
|
|
|
|
If using a custom kernel configuration ensure that the
|
|
|
|
COMPAT_FREEBSD11 option is included (as during the upgrade the
|
|
|
|
system will be running the ino64 kernel with the existing world).
|
|
|
|
|
|
|
|
For the safest in-place upgrade begin by removing previous build
|
2019-03-09 17:17:59 +00:00
|
|
|
artifacts via "rm -rf /usr/obj/*". Then, carefully follow the full
|
|
|
|
procedure documented below under the heading "To rebuild everything and
|
|
|
|
install it on the current system." Specifically, a reboot is required
|
|
|
|
after installing the new kernel before installing world. While an
|
|
|
|
installworld normally works by accident from multiuser after rebooting
|
|
|
|
the proper kernel, there are many cases where this will fail across this
|
|
|
|
upgrade and installworld from single user is required.
|
2017-05-23 20:25:49 +00:00
|
|
|
|
2017-04-24 21:27:37 +00:00
|
|
|
20170424:
|
|
|
|
The NATM framework including the en(4), fatm(4), hatm(4), and
|
|
|
|
patm(4) devices has been removed. Consumers should plan a
|
|
|
|
migration before the end-of-life date for FreeBSD 11.
|
|
|
|
|
2017-04-20 19:24:51 +00:00
|
|
|
20170420:
|
|
|
|
GNU diff has been replaced by a BSD licensed diff. Some features of GNU
|
|
|
|
diff has not been implemented, if those are needed a newer version of
|
|
|
|
GNU diff is available via the diffutils package under the gdiff name.
|
|
|
|
|
2017-04-14 05:25:40 +00:00
|
|
|
20170413:
|
|
|
|
As of r316810 for ipfilter, keep frags is no longer assumed when
|
|
|
|
keep state is specified in a rule. r316810 aligns ipfilter with
|
|
|
|
documentation in man pages separating keep frags from keep state.
|
2017-04-14 12:54:39 +00:00
|
|
|
This allows keep state to be specified without forcing keep frags
|
2017-04-14 05:25:40 +00:00
|
|
|
and allows keep frags to be specified independently of keep state.
|
|
|
|
To maintain previous behaviour, also specify keep frags with
|
2017-04-14 05:32:03 +00:00
|
|
|
keep state (as documented in ipf.conf.5).
|
2017-04-14 05:25:40 +00:00
|
|
|
|
2017-04-07 23:37:22 +00:00
|
|
|
20170407:
|
|
|
|
arm64 builds now use the base system LLD 4.0.0 linker by default,
|
|
|
|
instead of requiring that the aarch64-binutils port or package be
|
|
|
|
installed. To continue using aarch64-binutils, set
|
|
|
|
CROSS_BINUTILS_PREFIX=/usr/local/aarch64-freebsd/bin .
|
|
|
|
|
2017-04-26 16:50:54 +00:00
|
|
|
20170405:
|
|
|
|
The UDP optimization in entry 20160818 that added the sysctl
|
|
|
|
net.inet.udp.require_l2_bcast has been reverted. L2 broadcast
|
|
|
|
packets will no longer be treated as L3 broadcast packets.
|
|
|
|
|
2017-03-31 00:41:54 +00:00
|
|
|
20170331:
|
2017-03-31 09:10:05 +00:00
|
|
|
Binds and sends to the loopback addresses, IPv6 and IPv4, will now
|
2017-03-31 00:41:54 +00:00
|
|
|
use any explicitly assigned loopback address available in the jail
|
2017-03-31 09:10:05 +00:00
|
|
|
instead of using the first assigned address of the jail.
|
2017-03-31 00:41:54 +00:00
|
|
|
|
2017-03-30 05:11:58 +00:00
|
|
|
20170329:
|
|
|
|
The ctl.ko module no longer implements the iSCSI target frontend:
|
|
|
|
cfiscsi.ko does instead.
|
|
|
|
|
|
|
|
If building cfiscsi.ko as a kernel module, the module can be loaded
|
|
|
|
via one of the following methods:
|
|
|
|
- `cfiscsi_load="YES"` in loader.conf(5).
|
|
|
|
- Add `cfiscsi` to `$kld_list` in rc.conf(5).
|
|
|
|
- ctladm(8)/ctld(8), when compiled with iSCSI support
|
|
|
|
(`WITH_ISCSI=yes` in src.conf(5))
|
|
|
|
|
|
|
|
Please see cfiscsi(4) for more details.
|
|
|
|
|
- Add support for eMMC "partitions". Besides the user data area, i. e.
the default partition, eMMC v4.41 and later devices can additionally
provide up to:
1 enhanced user data area partition
2 boot partitions
1 RPMB (Replay Protected Memory Block) partition
4 general purpose partitions (optionally with a enhanced or extended
attribute)
Of these "partitions", only the enhanced user data area one actually
slices the user data area partition and, thus, gets handled with the
help of geom_flashmap(4). The other types of partitions have address
space independent from the default partition and need to be switched
to via CMD6 (SWITCH), i. e. constitute a set of additional "disks".
The second kind of these "partitions" doesn't fit that well into the
design of mmc(4) and mmcsd(4). I've decided to let mmcsd(4) hook all
of these "partitions" up as disk(9)'s (except for the RPMB partition
as it didn't seem to make much sense to be able to put a file-system
there and may require authentication; therefore, RPMB partitions are
solely accessible via the newly added IOCTL interface currently; see
also below). This approach for one resulted in cleaner code. Second,
it retains the notion of mmcsd(4) children corresponding to a single
physical device each. With the addition of some layering violations,
it also would have been possible for mmc(4) to add separate mmcsd(4)
instances with one disk each for all of these "partitions", however.
Still, both mmc(4) and mmcsd(4) share some common code now e. g. for
issuing CMD6, which has been factored out into mmc_subr.c.
Besides simply subdividing eMMC devices, some Intel NUCs having UEFI
code in the boot partitions etc., another use case for the partition
support is the activation of pseudo-SLC mode, which manufacturers of
eMMC chips typically associate with the enhanced user data area and/
or the enhanced attribute of general purpose partitions.
CAVEAT EMPTOR: Partitioning eMMC devices is a one-time operation.
- Now that properly issuing CMD6 is crucial (so data isn't written to
the wrong partition for example), make a step into the direction of
correctly handling the timeout for these commands in the MMC layer.
Also, do a SEND_STATUS when CMD6 is invoked with an R1B response as
recommended by relevant specifications. However, quite some work is
left to be done in this regard; all other R1B-type commands done by
the MMC layer also should be followed by a SEND_STATUS (CMD13), the
erase timeout calculations/handling as documented in specifications
are entirely ignored so far, the MMC layer doesn't provide timeouts
applicable up to the bridge drivers and at least sdhci(4) currently
is hardcoding 1 s as timeout for all command types unconditionally.
Let alone already available return codes often not being checked in
the MMC layer ...
- Add an IOCTL interface to mmcsd(4); this is sufficiently compatible
with Linux so that the GNU mmc-utils can be ported to and used with
FreeBSD (note that due to the remaining deficiencies outlined above
SANITIZE operations issued by/with `mmc` currently most likely will
fail). These latter will be added to ports as sysutils/mmc-utils in
a bit. Among others, the `mmc` tool of the GNU mmc-utils allows for
partitioning eMMC devices (tested working).
- For devices following the eMMC specification v4.41 or later, year 0
is 2013 rather than 1997; so correct this for assembling the device
ID string properly.
- Let mmcsd.ko depend on mmc.ko. Additionally, bump MMC_VERSION as at
least for some of the above a matching pair is required.
- In the ACPI front-end of sdhci(4) describe the Intel eMMC and SDXC
controllers as such in order to match the PCI one.
Additionally, in the entry for the 80860F14 SDXC controller remove
the eMMC-only SDHCI_QUIRK_INTEL_POWER_UP_RESET.
OKed by: imp
Submitted by: ian (mmc_switch_status() implementation)
2017-03-16 22:23:04 +00:00
|
|
|
20170316:
|
|
|
|
The mmcsd.ko module now additionally depends on geom_flashmap.ko.
|
|
|
|
Also, mmc.ko and mmcsd.ko need to be a matching pair built from the
|
|
|
|
same source (previously, the dependency of mmcsd.ko on mmc.ko was
|
|
|
|
missing, but mmcsd.ko now will refuse to load if it is incompatible
|
|
|
|
with mmc.ko).
|
|
|
|
|
2017-03-15 13:36:35 +00:00
|
|
|
20170315:
|
|
|
|
The syntax of ipfw(8) named states was changed to avoid ambiguity.
|
|
|
|
If you have used named states in the firewall rules, you need to modify
|
|
|
|
them after installworld and before rebooting. Now named states must
|
|
|
|
be prefixed with colon.
|
|
|
|
|
2017-03-11 05:12:39 +00:00
|
|
|
20170311:
|
|
|
|
The old drm (sys/dev/drm/) drivers for i915 and radeon have been
|
|
|
|
removed as the userland we provide cannot use them. The KMS version
|
2017-04-14 12:54:39 +00:00
|
|
|
(sys/dev/drm2) supports the same hardware.
|
2017-03-11 05:12:39 +00:00
|
|
|
|
2017-03-01 21:58:26 +00:00
|
|
|
20170302:
|
2017-01-04 21:58:33 +00:00
|
|
|
Clang, llvm, lldb, compiler-rt and libc++ have been upgraded to 4.0.0.
|
|
|
|
Please see the 20141231 entry below for information about prerequisites
|
|
|
|
and upgrading, if you are not already using clang 3.5.0 or higher.
|
|
|
|
|
2017-02-21 17:54:38 +00:00
|
|
|
20170221:
|
|
|
|
The code that provides support for ZFS .zfs/ directory functionality
|
|
|
|
has been reimplemented. It's not possible now to create a snapshot
|
|
|
|
by mkdir under .zfs/snapshot/. That should be the only user visible
|
|
|
|
change.
|
|
|
|
|
2017-02-17 06:22:00 +00:00
|
|
|
20170216:
|
|
|
|
EISA bus support has been removed. The WITH_EISA option is no longer
|
|
|
|
valid.
|
|
|
|
|
|
|
|
20170215:
|
|
|
|
MCA bus support has been removed.
|
|
|
|
|
2017-01-27 01:59:12 +00:00
|
|
|
20170127:
|
|
|
|
The WITH_LLD_AS_LD / WITHOUT_LLD_AS_LD build knobs have been renamed
|
|
|
|
WITH_LLD_IS_LD / WITHOUT_LLD_IS_LD, for consistency with CLANG_IS_CC.
|
|
|
|
|
2017-01-12 14:38:18 +00:00
|
|
|
20170112:
|
|
|
|
The EM_MULTIQUEUE kernel configuration option is deprecated now that
|
|
|
|
the em(4) driver conforms to iflib specifications.
|
|
|
|
|
2017-01-10 03:23:22 +00:00
|
|
|
20170109:
|
|
|
|
The igb(4), em(4) and lem(4) ethernet drivers are now implemented via
|
|
|
|
IFLIB. If you have a custom kernel configuration that excludes em(4)
|
2019-03-09 17:17:55 +00:00
|
|
|
but you use igb(4), you need to re-add em(4) to your custom
|
|
|
|
configuration.
|
2017-01-10 03:23:22 +00:00
|
|
|
|
2016-12-17 18:35:17 +00:00
|
|
|
20161217:
|
2016-11-26 01:17:03 +00:00
|
|
|
Clang, llvm, lldb, compiler-rt and libc++ have been upgraded to 3.9.1.
|
|
|
|
Please see the 20141231 entry below for information about prerequisites
|
|
|
|
and upgrading, if you are not already using clang 3.5.0 or higher.
|
|
|
|
|
2016-11-24 22:18:55 +00:00
|
|
|
20161124:
|
|
|
|
Clang, llvm, lldb, compiler-rt and libc++ have been upgraded to 3.9.0.
|
|
|
|
Please see the 20141231 entry below for information about prerequisites
|
|
|
|
and upgrading, if you are not already using clang 3.5.0 or higher.
|
|
|
|
|
2016-11-20 06:10:12 +00:00
|
|
|
20161119:
|
|
|
|
The layout of the pmap structure has changed for powerpc to put the pmap
|
|
|
|
statistics at the front for all CPU variations. libkvm(3) and all tools
|
|
|
|
that link against it need to be recompiled.
|
|
|
|
|
add iic interface to ig4 driver, move isl and cyapa to iicbus
Summary:
The hardware does not expose a classic SMBus interface.
Instead it has a lower level interface that can express a far richer
I2C protocol than what smbus offers. However, the interface does not
provide a way to explicitly generate the I2C stop and start conditions.
It's only possible to request that the stop condition is generated
after transferring the next byte in either direction. So, at least
one data byte must always be transferred.
Thus, some I2C sequences are impossible to generate, e.g., an equivalent
of smbus quick command (<start>-<slave addr>-<r/w bit>-<stop>).
At the same time isl(4) and cyapa(4) are moved to iicbus and now they use
iicbus_transfer for communication. Previously they used smbus_trans()
interface that is not defined by the SMBus protocol and was implemented
only by ig4(4). In fact, that interface was impossible to implement
for the typical SMBus controllers like intpm(4) or ichsmb(4) where
a type of the SMBus command must be programmed.
The plan is to remove smbus_trans() and all its uses.
As an aside, the smbus_trans() method deviates from the standard,
but perhaps backwards, FreeBSD convention of using 8-bit slave
addresses (shifted by 1 bit to the left). The method expects
7-bit addresses.
There is a user facing consequence of this change.
A user must now provide device hints for isl and cyapa that specify an iicbus to use
and a slave address on it.
On Chromebook hardware where isl and cyapa devices are commonly found
it is also possible to use a new chromebook_platform(4) driver that
automatically configures isl and cyapa devices. There is no need to
provide the device hints in that case,
Right now smbus(4) driver tries to discover all slaves on the bus.
That is very dangerous. Fortunately, the probing code uses smbus_trans()
to do its job, so it is really enabled for ig4 only.
The plan is to remove that auto-probing code and smbus_trans().
Tested by: grembo, Matthias Apitz <guru@unixarea.de> (w/o
chromebook_platform)
Discussed with: grembo, imp
Reviewed by: wblock (docs)
MFC after: 1 month
Relnotes: yes
Differential Revision: https://reviews.freebsd.org/D8172
2016-10-30 12:15:33 +00:00
|
|
|
20161030:
|
|
|
|
isl(4) and cyapa(4) drivers now require a new driver,
|
|
|
|
chromebook_platform(4), to work properly on Chromebook-class hardware.
|
|
|
|
On other types of hardware the drivers may need to be configured using
|
|
|
|
device hints. Please see the corresponding manual pages for details.
|
|
|
|
|
2016-10-17 21:35:13 +00:00
|
|
|
20161017:
|
|
|
|
The urtwn(4) driver was merged into rtwn(4) and now consists of
|
|
|
|
rtwn(4) main module + rtwn_usb(4) and rtwn_pci(4) bus-specific
|
|
|
|
parts.
|
|
|
|
Also, firmware for RTL8188CE was renamed due to possible name
|
|
|
|
conflict (rtwnrtl8192cU(B) -> rtwnrtl8192cE(B))
|
|
|
|
|
2016-10-15 12:07:37 +00:00
|
|
|
20161015:
|
2016-10-15 12:32:59 +00:00
|
|
|
GNU rcs has been removed from base. It is available as packages:
|
2016-10-15 12:07:37 +00:00
|
|
|
- rcs: Latest GPLv3 GNU rcs version.
|
2016-10-15 15:55:04 +00:00
|
|
|
- rcs57: Copy of the latest version of GNU rcs (GPLv2) before it was
|
|
|
|
removed from base.
|
2016-10-15 12:07:37 +00:00
|
|
|
|
In the TCP stack, the hhook(9) framework provides hooks for kernel modules
to add actions that run when a TCP frame is sent or received on a TCP
session in the ESTABLISHED state. In the base tree, this functionality is
only used for the h_ertt module, which is used by the cc_cdg, cc_chd, cc_hd,
and cc_vegas congestion control modules.
Presently, we incur overhead to check for hooks each time a TCP frame is
sent or received on an ESTABLISHED TCP session.
This change adds a new compile-time option (TCP_HHOOK) to determine whether
to include the hhook(9) framework for TCP. To retain backwards
compatibility, I added the TCP_HHOOK option to every configuration file that
already defined "options INET". (Therefore, this patch introduces no
functional change. In order to see a functional difference, you need to
compile a custom kernel without the TCP_HHOOK option.) This change will
allow users to easily exclude this functionality from their kernel, should
they wish to do so.
Note that any users who use a custom kernel configuration and use one of the
congestion control modules listed above will need to add the TCP_HHOOK
option to their kernel configuration.
Reviewed by: rrs, lstewart, hiren (previous version), sjg (makefiles only)
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D8185
2016-10-12 02:16:42 +00:00
|
|
|
20161008:
|
|
|
|
Use of the cc_cdg, cc_chd, cc_hd, or cc_vegas congestion control
|
|
|
|
modules now requires that the kernel configuration contain the
|
|
|
|
TCP_HHOOK option. (This option is included in the GENERIC kernel.)
|
|
|
|
|
2016-10-03 17:49:26 +00:00
|
|
|
20161003:
|
|
|
|
The WITHOUT_ELFCOPY_AS_OBJCOPY src.conf(5) knob has been retired.
|
|
|
|
ELF Tool Chain's elfcopy is always installed as /usr/bin/objcopy.
|
|
|
|
|
2016-09-25 16:39:18 +00:00
|
|
|
20160924:
|
|
|
|
Relocatable object files with the extension of .So have been renamed
|
|
|
|
to use an extension of .pico instead. The purpose of this change is
|
|
|
|
to avoid a name clash with shared libraries on case-insensitive file
|
|
|
|
systems. On those file systems, foo.So is the same file as foo.so.
|
|
|
|
|
2016-09-18 15:01:11 +00:00
|
|
|
20160918:
|
2016-09-18 15:55:57 +00:00
|
|
|
GNU rcs has been turned off by default. It can (temporarily) be built
|
2016-09-18 15:01:11 +00:00
|
|
|
again by adding WITH_RCS knob in src.conf.
|
2016-09-18 18:27:28 +00:00
|
|
|
Otherwise, GNU rcs is available from packages:
|
|
|
|
- rcs: Latest GPLv3 GNU rcs version.
|
|
|
|
- rcs57: Copy of the latest version of GNU rcs (GPLv2) from base.
|
2016-09-18 15:01:11 +00:00
|
|
|
|
2016-09-18 12:49:23 +00:00
|
|
|
20160918:
|
|
|
|
The backup_uses_rcs functionality has been removed from rc.subr.
|
|
|
|
|
2016-09-08 21:20:01 +00:00
|
|
|
20160908:
|
|
|
|
The queue(3) debugging macro, QUEUE_MACRO_DEBUG, has been split into
|
|
|
|
two separate components, QUEUE_MACRO_DEBUG_TRACE and
|
|
|
|
QUEUE_MACRO_DEBUG_TRASH. Define both for the original
|
|
|
|
QUEUE_MACRO_DEBUG behavior.
|
|
|
|
|
2016-08-25 05:28:02 +00:00
|
|
|
20160824:
|
|
|
|
r304787 changed some ioctl interfaces between the iSCSI userspace
|
|
|
|
programs and the kernel. ctladm, ctld, iscsictl, and iscsid must be
|
|
|
|
rebuilt to work with new kernels. __FreeBSD_version has been bumped
|
|
|
|
to 1200005.
|
|
|
|
|
2016-08-18 22:59:05 +00:00
|
|
|
20160818:
|
|
|
|
The UDP receive code has been updated to only treat incoming UDP
|
|
|
|
packets that were addressed to an L2 broadcast address as L3
|
|
|
|
broadcast packets. It is not expected that this will affect any
|
|
|
|
standards-conforming UDP application. The new behaviour can be
|
|
|
|
disabled by setting the sysctl net.inet.udp.require_l2_bcast to
|
|
|
|
0.
|
|
|
|
|
2016-08-18 10:54:39 +00:00
|
|
|
20160818:
|
|
|
|
Remove the openbsd_poll system call.
|
2017-04-14 12:54:39 +00:00
|
|
|
__FreeBSD_version has been bumped because of this.
|
2016-08-18 10:54:39 +00:00
|
|
|
|
2018-07-17 11:35:40 +00:00
|
|
|
20160708:
|
|
|
|
The stable/11 branch has been created from head@r302406.
|
|
|
|
|
2016-06-23 21:02:05 +00:00
|
|
|
20160622:
|
2016-06-23 21:09:07 +00:00
|
|
|
The libc stub for the pipe(2) system call has been replaced with
|
2016-07-20 18:00:22 +00:00
|
|
|
a wrapper that calls the pipe2(2) system call and the pipe(2)
|
|
|
|
system call is now only implemented by the kernels that include
|
|
|
|
"options COMPAT_FREEBSD10" in their config file (this is the
|
|
|
|
default). Users should ensure that this option is enabled in
|
|
|
|
their kernel or upgrade userspace to r302092 before upgrading their
|
|
|
|
kernel.
|
2016-06-23 21:02:05 +00:00
|
|
|
|
2016-05-27 22:26:43 +00:00
|
|
|
20160527:
|
|
|
|
CAM will now strip leading spaces from SCSI disks' serial numbers.
|
2016-11-17 00:21:55 +00:00
|
|
|
This will affect users who create UFS filesystems on SCSI disks using
|
2016-05-27 22:26:43 +00:00
|
|
|
those disk's diskid device nodes. For example, if /etc/fstab
|
|
|
|
previously contained a line like
|
|
|
|
"/dev/diskid/DISK-%20%20%20%20%20%20%20ABCDEFG0123456", you should
|
|
|
|
change it to "/dev/diskid/DISK-ABCDEFG0123456". Users of geom
|
2016-05-27 22:44:33 +00:00
|
|
|
transforms like gmirror may also be affected. ZFS users should
|
2016-05-27 22:26:43 +00:00
|
|
|
generally be fine.
|
|
|
|
|
2016-05-23 20:29:18 +00:00
|
|
|
20160523:
|
|
|
|
The bitstring(3) API has been updated with new functionality and
|
|
|
|
improved performance. But it is binary-incompatible with the old API.
|
|
|
|
Objects built with the new headers may not be linked against objects
|
|
|
|
built with the old headers.
|
|
|
|
|
2016-05-20 15:04:48 +00:00
|
|
|
20160520:
|
|
|
|
The brk and sbrk functions have been removed from libc on arm64.
|
|
|
|
Binutils from ports has been updated to not link to these
|
|
|
|
functions and should be updated to the latest version before
|
|
|
|
installing a new libc.
|
|
|
|
|
2016-05-18 06:01:18 +00:00
|
|
|
20160517:
|
|
|
|
The armv6 port now defaults to hard float ABI. Limited support
|
|
|
|
for running both hardfloat and soft float on the same system
|
|
|
|
is available using the libraries installed with -DWITH_LIBSOFT.
|
|
|
|
This has only been tested as an upgrade path for installworld
|
|
|
|
and packages may fail or need manual intervention to run. New
|
|
|
|
packages will be needed.
|
|
|
|
|
2016-05-23 04:07:41 +00:00
|
|
|
To update an existing self-hosted armv6hf system, you must add
|
2017-04-14 12:54:39 +00:00
|
|
|
TARGET_ARCH=armv6 on the make command line for both the build
|
2016-05-23 04:07:41 +00:00
|
|
|
and the install steps.
|
|
|
|
|
2016-05-10 22:32:23 +00:00
|
|
|
20160510:
|
|
|
|
Kernel modules compiled outside of a kernel build now default to
|
|
|
|
installing to /boot/modules instead of /boot/kernel. Many kernel
|
|
|
|
modules built this way (such as those in ports) already overrode
|
|
|
|
KMODDIR explicitly to install into /boot/modules. However,
|
|
|
|
manually building and installing a module from /sys/modules will
|
|
|
|
now install to /boot/modules instead of /boot/kernel.
|
|
|
|
|
2016-04-14 22:38:13 +00:00
|
|
|
20160414:
|
2016-04-15 04:45:37 +00:00
|
|
|
The CAM I/O scheduler has been committed to the kernel. There should be
|
|
|
|
no user visible impact. This does enable NCQ Trim on ada SSDs. While the
|
|
|
|
list of known rogues that claim support for this but actually corrupt
|
|
|
|
data is believed to be complete, be on the lookout for data
|
|
|
|
corruption. The known rogue list is believed to be complete:
|
|
|
|
|
|
|
|
o Crucial MX100, M550 drives with MU01 firmware.
|
|
|
|
o Micron M510 and M550 drives with MU01 firmware.
|
|
|
|
o Micron M500 prior to MU07 firmware
|
|
|
|
o Samsung 830, 840, and 850 all firmwares
|
|
|
|
o FCCT M500 all firmwares
|
|
|
|
|
|
|
|
Crucial has firmware http://www.crucial.com/usa/en/support-ssd-firmware
|
|
|
|
with working NCQ TRIM. For Micron branded drives, see your sales rep for
|
|
|
|
updated firmware. Black listed drives will work correctly because these
|
|
|
|
drives work correctly so long as no NCQ TRIMs are sent to them. Given
|
|
|
|
this list is the same as found in Linux, it's believed there are no
|
|
|
|
other rogues in the market place. All other models from the above
|
|
|
|
vendors work.
|
|
|
|
|
|
|
|
To be safe, if you are at all concerned, you can quirk each of your
|
|
|
|
drives to prevent NCQ from being sent by setting:
|
|
|
|
kern.cam.ada.X.quirks="0x2"
|
|
|
|
in loader.conf. If the drive requires the 4k sector quirk, set the
|
|
|
|
quirks entry to 0x3.
|
2016-04-14 22:38:13 +00:00
|
|
|
|
2016-03-30 23:50:23 +00:00
|
|
|
20160330:
|
|
|
|
The FAST_DEPEND build option has been removed and its functionality is
|
|
|
|
now the one true way. The old mkdep(1) style of 'make depend' has
|
|
|
|
been removed. See 20160311 for further details.
|
|
|
|
|
Use uintmax_t (typedef'd to rman_res_t type) for rman ranges.
On some architectures, u_long isn't large enough for resource definitions.
Particularly, powerpc and arm allow 36-bit (or larger) physical addresses, but
type `long' is only 32-bit. This extends rman's resources to uintmax_t. With
this change, any resource can feasibly be placed anywhere in physical memory
(within the constraints of the driver).
Why uintmax_t and not something machine dependent, or uint64_t? Though it's
possible for uintmax_t to grow, it's highly unlikely it will become 128-bit on
32-bit architectures. 64-bit architectures should have plenty of RAM to absorb
the increase on resource sizes if and when this occurs, and the number of
resources on memory-constrained systems should be sufficiently small as to not
pose a drastic overhead. That being said, uintmax_t was chosen for source
clarity. If it's specified as uint64_t, all printf()-like calls would either
need casts to uintmax_t, or be littered with PRI*64 macros. Casts to uintmax_t
aren't horrible, but it would also bake into the API for
resource_list_print_type() either a hidden assumption that entries get cast to
uintmax_t for printing, or these calls would need the PRI*64 macros. Since
source code is meant to be read more often than written, I chose the clearest
path of simply using uintmax_t.
Tested on a PowerPC p5020-based board, which places all device resources in
0xfxxxxxxxx, and has 8GB RAM.
Regression tested on qemu-system-i386
Regression tested on qemu-system-mips (malta profile)
Tested PAE and devinfo on virtualbox (live CD)
Special thanks to bz for his testing on ARM.
Reviewed By: bz, jhb (previous)
Relnotes: Yes
Sponsored by: Alex Perez/Inertial Computing
Differential Revision: https://reviews.freebsd.org/D4544
2016-03-18 01:28:41 +00:00
|
|
|
20160317:
|
|
|
|
Resource range types have grown from unsigned long to uintmax_t. All
|
|
|
|
drivers, and anything using libdevinfo, need to be recompiled.
|
|
|
|
|
2016-03-11 16:57:41 +00:00
|
|
|
20160311:
|
|
|
|
WITH_FAST_DEPEND is now enabled by default for in-tree and out-of-tree
|
|
|
|
builds. It no longer runs mkdep(1) during 'make depend', and the
|
|
|
|
'make depend' stage can safely be skipped now as it is auto ran
|
|
|
|
when building 'make all' and will generate all SRCS and DPSRCS before
|
|
|
|
building anything else. Dependencies are gathered at compile time with
|
|
|
|
-MF flags kept in separate .depend files per object file. Users should
|
|
|
|
run 'make cleandepend' once if using -DNO_CLEAN to clean out older
|
|
|
|
stale .depend files.
|
|
|
|
|
2016-03-06 17:34:21 +00:00
|
|
|
20160306:
|
|
|
|
On amd64, clang 3.8.0 can now insert sections of type AMD64_UNWIND into
|
|
|
|
kernel modules. Therefore, if you load any kernel modules at boot time,
|
|
|
|
please install the boot loaders after you install the kernel, but before
|
|
|
|
rebooting, e.g.:
|
|
|
|
|
|
|
|
make buildworld
|
2018-03-28 16:51:05 +00:00
|
|
|
make buildkernel KERNCONF=YOUR_KERNEL_HERE
|
|
|
|
make installkernel KERNCONF=YOUR_KERNEL_HERE
|
2016-03-06 17:34:21 +00:00
|
|
|
make -C sys/boot install
|
|
|
|
<reboot in single user>
|
|
|
|
|
|
|
|
Then follow the usual steps, described in the General Notes section,
|
|
|
|
below.
|
|
|
|
|
2016-03-05 20:24:31 +00:00
|
|
|
20160305:
|
|
|
|
Clang, llvm, lldb and compiler-rt have been upgraded to 3.8.0. Please
|
|
|
|
see the 20141231 entry below for information about prerequisites and
|
|
|
|
upgrading, if you are not already using clang 3.5.0 or higher.
|
|
|
|
|
2016-03-04 17:42:59 +00:00
|
|
|
20160301:
|
|
|
|
The AIO subsystem is now a standard part of the kernel. The
|
|
|
|
VFS_AIO kernel option and aio.ko kernel module have been removed.
|
|
|
|
Due to stability concerns, asynchronous I/O requests are only
|
|
|
|
permitted on sockets and raw disks by default. To enable
|
|
|
|
asynchronous I/O requests on all file types, set the
|
|
|
|
vfs.aio.enable_unsafe sysctl to a non-zero value.
|
|
|
|
|
2016-02-27 12:34:51 +00:00
|
|
|
20160226:
|
2016-02-26 15:46:14 +00:00
|
|
|
The ELF object manipulation tool objcopy is now provided by the
|
|
|
|
ELF Tool Chain project rather than by GNU binutils. It should be a
|
|
|
|
drop-in replacement, with the addition of arm64 support. The
|
|
|
|
(temporary) src.conf knob WITHOUT_ELFCOPY_AS_OBJCOPY knob may be set
|
|
|
|
to obtain the GNU version if necessary.
|
|
|
|
|
2016-01-29 17:08:26 +00:00
|
|
|
20160129:
|
|
|
|
Building ZFS pools on top of zvols is prohibited by default. That
|
|
|
|
feature has never worked safely; it's always been prone to deadlocks.
|
|
|
|
Using a zvol as the backing store for a VM guest's virtual disk will
|
|
|
|
still work, even if the guest is using ZFS. Legacy behavior can be
|
|
|
|
restored by setting vfs.zfs.vol.recursive=1.
|
|
|
|
|
2016-01-19 14:38:20 +00:00
|
|
|
20160119:
|
|
|
|
The NONE and HPN patches has been removed from OpenSSH. They are
|
|
|
|
still available in the security/openssh-portable port.
|
|
|
|
|
2016-01-13 01:49:35 +00:00
|
|
|
20160113:
|
|
|
|
With the addition of ypldap(8), a new _ypldap user is now required
|
|
|
|
during installworld. "mergemaster -p" can be used to add the user
|
|
|
|
prior to installworld, as documented in the handbook.
|
|
|
|
|
2015-12-16 17:13:09 +00:00
|
|
|
20151216:
|
|
|
|
The tftp loader (pxeboot) now uses the option root-path directive. As a
|
|
|
|
consequence it no longer looks for a pxeboot.4th file on the tftp
|
|
|
|
server. Instead it uses the regular /boot infrastructure as with the
|
|
|
|
other loaders.
|
|
|
|
|
2015-12-13 22:11:37 +00:00
|
|
|
20151211:
|
|
|
|
The code to start recording plug and play data into the modules has
|
|
|
|
been committed. While the old tools will properly build a new kernel,
|
|
|
|
a number of warnings about "unknown metadata record 4" will be produced
|
|
|
|
for an older kldxref. To avoid such warnings, make sure to rebuild
|
|
|
|
the kernel toolchain (or world). Make sure that you have r292078 or
|
|
|
|
later when trying to build 292077 or later before rebuilding.
|
|
|
|
|
2015-12-07 19:15:52 +00:00
|
|
|
20151207:
|
|
|
|
Debug data files are now built by default with 'make buildworld' and
|
|
|
|
installed with 'make installworld'. This facilitates debugging but
|
|
|
|
requires more disk space both during the build and for the installed
|
|
|
|
world. Debug files may be disabled by setting WITHOUT_DEBUG_FILES=yes
|
|
|
|
in src.conf(5).
|
|
|
|
|
2015-11-30 22:46:34 +00:00
|
|
|
20151130:
|
|
|
|
r291527 changed the internal interface between the nfsd.ko and
|
|
|
|
nfscommon.ko modules. As such, they must both be upgraded to-gether.
|
|
|
|
__FreeBSD_version has been bumped because of this.
|
|
|
|
|
2015-11-07 12:48:29 +00:00
|
|
|
20151108:
|
|
|
|
Add support for unicode collation strings leads to a change of
|
|
|
|
order of files listed by ls(1) for example. To get back to the old
|
|
|
|
behaviour, set LC_COLLATE environment variable to "C".
|
|
|
|
|
2015-11-07 17:54:51 +00:00
|
|
|
Databases administrators will need to reindex their databases given
|
|
|
|
collation results will be different.
|
|
|
|
|
2015-11-17 19:40:48 +00:00
|
|
|
Due to a bug in install(1) it is recommended to remove the ancient
|
2015-11-18 16:52:19 +00:00
|
|
|
locales before running make installworld.
|
2015-11-17 12:18:57 +00:00
|
|
|
|
2015-11-18 16:52:19 +00:00
|
|
|
rm -rf /usr/share/locale/*
|
2015-11-17 12:18:57 +00:00
|
|
|
|
2015-10-30 20:51:33 +00:00
|
|
|
20151030:
|
|
|
|
The OpenSSL has been upgraded to 1.0.2d. Any binaries requiring
|
|
|
|
libcrypto.so.7 or libssl.so.7 must be recompiled.
|
|
|
|
|
2015-10-20 19:04:13 +00:00
|
|
|
20151020:
|
|
|
|
Qlogic 24xx/25xx firmware images were updated from 5.5.0 to 7.3.0.
|
|
|
|
Kernel modules isp_2400_multi and isp_2500_multi were removed and
|
|
|
|
should be replaced with isp_2400 and isp_2500 modules respectively.
|
|
|
|
|
2015-10-17 16:42:54 +00:00
|
|
|
20151017:
|
|
|
|
The build previously allowed using 'make -n' to not recurse into
|
|
|
|
sub-directories while showing what commands would be executed, and
|
|
|
|
'make -n -n' to recursively show commands. Now 'make -n' will recurse
|
|
|
|
and 'make -N' will not.
|
|
|
|
|
2015-10-12 21:02:36 +00:00
|
|
|
20151012:
|
|
|
|
If you specify SENDMAIL_MC or SENDMAIL_CF in make.conf, mergemaster
|
|
|
|
and etcupdate will now use this file. A custom sendmail.cf is now
|
|
|
|
updated via this mechanism rather than via installworld. If you had
|
|
|
|
excluded sendmail.cf in mergemaster.rc or etcupdate.conf, you may
|
|
|
|
want to remove the exclusion or change it to "always install".
|
|
|
|
/etc/mail/sendmail.cf is now managed the same way regardless of
|
|
|
|
whether SENDMAIL_MC/SENDMAIL_CF is used. If you are not using
|
|
|
|
SENDMAIL_MC/SENDMAIL_CF there should be no change in behavior.
|
|
|
|
|
2015-10-11 13:01:51 +00:00
|
|
|
20151011:
|
|
|
|
Compatibility shims for legacy ATA device names have been removed.
|
|
|
|
It includes ATA_STATIC_ID kernel option, kern.cam.ada.legacy_aliases
|
|
|
|
and kern.geom.raid.legacy_aliases loader tunables, kern.devalias.*
|
|
|
|
environment variables, /dev/ad* and /dev/ar* symbolic links.
|
|
|
|
|
2015-10-06 16:26:07 +00:00
|
|
|
20151006:
|
2015-10-05 19:48:12 +00:00
|
|
|
Clang, llvm, lldb, compiler-rt and libc++ have been upgraded to 3.7.0.
|
|
|
|
Please see the 20141231 entry below for information about prerequisites
|
|
|
|
and upgrading, if you are not already using clang 3.5.0 or higher.
|
2015-05-30 15:39:49 +00:00
|
|
|
|
2015-09-24 16:56:44 +00:00
|
|
|
20150924:
|
2015-09-24 16:55:22 +00:00
|
|
|
Kernel debug files have been moved to /usr/lib/debug/boot/kernel/,
|
|
|
|
and renamed from .symbols to .debug. This reduces the size requirements
|
|
|
|
on the boot partition or file system and provides consistency with
|
|
|
|
userland debug files.
|
|
|
|
|
|
|
|
When using the supported kernel installation method the
|
|
|
|
/usr/lib/debug/boot/kernel directory will be renamed (to kernel.old)
|
|
|
|
as is done with /boot/kernel.
|
|
|
|
|
|
|
|
Developers wishing to maintain the historical behavior of installing
|
|
|
|
debug files in /boot/kernel/ can set KERN_DEBUGDIR="" in src.conf(5).
|
|
|
|
|
2015-08-29 07:59:31 +00:00
|
|
|
20150827:
|
|
|
|
The wireless drivers had undergone changes that remove the 'parent
|
|
|
|
interface' from the ifconfig -l output. The rc.d network scripts
|
|
|
|
used to check presence of a parent interface in the list, so old
|
|
|
|
scripts would fail to start wireless networking. Thus, etcupdate(3)
|
|
|
|
or mergemaster(8) run is required after kernel update, to update your
|
|
|
|
rc.d scripts in /etc.
|
|
|
|
|
2015-08-27 21:27:47 +00:00
|
|
|
20150827:
|
|
|
|
pf no longer supports 'scrub fragment crop' or 'scrub fragment drop-ovl'
|
|
|
|
These configurations are now automatically interpreted as
|
|
|
|
'scrub fragment reassemble'.
|
|
|
|
|
2015-08-17 07:36:12 +00:00
|
|
|
20150817:
|
|
|
|
Kernel-loadable modules for the random(4) device are back. To use
|
|
|
|
them, the kernel must have
|
|
|
|
|
|
|
|
device random
|
|
|
|
options RANDOM_LOADABLE
|
|
|
|
|
|
|
|
kldload(8) can then be used to load random_fortuna.ko
|
|
|
|
or random_yarrow.ko. Please note that due to the indirect
|
|
|
|
function calls that the loadable modules need to provide,
|
|
|
|
the build-in variants will be slightly more efficient.
|
|
|
|
|
|
|
|
The random(4) kernel option RANDOM_DUMMY has been retired due to
|
|
|
|
unpopularity. It was not all that useful anyway.
|
|
|
|
|
2015-08-13 17:50:47 +00:00
|
|
|
20150813:
|
|
|
|
The WITHOUT_ELFTOOLCHAIN_TOOLS src.conf(5) knob has been retired.
|
|
|
|
Control over building the ELF Tool Chain tools is now provided by
|
|
|
|
the WITHOUT_TOOLCHAIN knob.
|
|
|
|
|
2015-08-10 21:09:06 +00:00
|
|
|
20150810:
|
|
|
|
The polarity of Pulse Per Second (PPS) capture events with the
|
|
|
|
uart(4) driver has been corrected. Prior to this change the PPS
|
|
|
|
"assert" event corresponded to the trailing edge of a positive PPS
|
|
|
|
pulse and the "clear" event was the leading edge of the next pulse.
|
|
|
|
|
|
|
|
As the width of a PPS pulse in a typical GPS receiver is on the
|
|
|
|
order of 1 millisecond, most users will not notice any significant
|
|
|
|
difference with this change.
|
|
|
|
|
|
|
|
Anyone who has compensated for the historical polarity reversal by
|
|
|
|
configuring a negative offset equal to the pulse width will need to
|
|
|
|
remove that workaround.
|
|
|
|
|
2015-08-09 12:58:56 +00:00
|
|
|
20150809:
|
|
|
|
The default group assigned to /dev/dri entries has been changed
|
|
|
|
from 'wheel' to 'video' with the id of '44'. If you want to have
|
|
|
|
access to the dri devices please add yourself to the video group
|
|
|
|
with:
|
|
|
|
|
|
|
|
# pw groupmod video -m $USER
|
|
|
|
|
2015-08-06 16:07:27 +00:00
|
|
|
20150806:
|
2017-04-14 12:54:39 +00:00
|
|
|
The menu.rc and loader.rc files will now be replaced during
|
2015-08-06 16:07:27 +00:00
|
|
|
upgrades. Please migrate local changes to menu.rc.local and
|
|
|
|
loader.rc.local instead.
|
|
|
|
|
2015-08-05 18:30:00 +00:00
|
|
|
20150805:
|
|
|
|
GNU Binutils versions of addr2line, c++filt, nm, readelf, size,
|
|
|
|
strings and strip have been removed. The src.conf(5) knob
|
|
|
|
WITHOUT_ELFTOOLCHAIN_TOOLS no longer provides the binutils tools.
|
|
|
|
|
2015-07-28 11:19:38 +00:00
|
|
|
20150728:
|
2015-07-28 11:21:33 +00:00
|
|
|
As ZFS requires more kernel stack pages than is the default on some
|
2015-07-28 11:19:38 +00:00
|
|
|
architectures e.g. i386, it now warns if KSTACK_PAGES is less than
|
|
|
|
ZFS_MIN_KSTACK_PAGES (which is 4 at the time of writing).
|
|
|
|
|
|
|
|
Please consider using 'options KSTACK_PAGES=X' where X is greater
|
|
|
|
than or equal to ZFS_MIN_KSTACK_PAGES i.e. 4 in such configurations.
|
|
|
|
|
2015-07-07 03:03:17 +00:00
|
|
|
20150706:
|
|
|
|
sendmail has been updated to 8.15.2. Starting with FreeBSD 11.0
|
|
|
|
and sendmail 8.15, sendmail uses uncompressed IPv6 addresses by
|
|
|
|
default, i.e., they will not contain "::". For example, instead
|
|
|
|
of ::1, it will be 0:0:0:0:0:0:0:1. This permits a zero subnet
|
|
|
|
to have a more specific match, such as different map entries for
|
|
|
|
IPv6:0:0 vs IPv6:0. This change requires that configuration
|
|
|
|
data (including maps, files, classes, custom ruleset, etc.) must
|
|
|
|
use the same format, so make certain such configuration data is
|
|
|
|
upgrading. As a very simple check search for patterns like
|
|
|
|
'IPv6:[0-9a-fA-F:]*::' and 'IPv6::'. To return to the old
|
|
|
|
behavior, set the m4 option confUSE_COMPRESSED_IPV6_ADDRESSES or
|
|
|
|
the cf option UseCompressedIPv6Addresses.
|
|
|
|
|
Huge cleanup of random(4) code.
* GENERAL
- Update copyright.
- Make kernel options for RANDOM_YARROW and RANDOM_DUMMY. Set
neither to ON, which means we want Fortuna
- If there is no 'device random' in the kernel, there will be NO
random(4) device in the kernel, and the KERN_ARND sysctl will
return nothing. With RANDOM_DUMMY there will be a random(4) that
always blocks.
- Repair kern.arandom (KERN_ARND sysctl). The old version went
through arc4random(9) and was a bit weird.
- Adjust arc4random stirring a bit - the existing code looks a little
suspect.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Redo read_random(9) so as to duplicate random(4)'s read internals.
This makes it a first-class citizen rather than a hack.
- Move stuff out of locked regions when it does not need to be
there.
- Trim RANDOM_DEBUG printfs. Some are excess to requirement, some
behind boot verbose.
- Use SYSINIT to sequence the startup.
- Fix init/deinit sysctl stuff.
- Make relevant sysctls also tunables.
- Add different harvesting "styles" to allow for different requirements
(direct, queue, fast).
- Add harvesting of FFS atime events. This needs to be checked for
weighing down the FS code.
- Add harvesting of slab allocator events. This needs to be checked for
weighing down the allocator code.
- Fix the random(9) manpage.
- Loadable modules are not present for now. These will be re-engineered
when the dust settles.
- Use macros for locks.
- Fix comments.
* src/share/man/...
- Update the man pages.
* src/etc/...
- The startup/shutdown work is done in D2924.
* src/UPDATING
- Add UPDATING announcement.
* src/sys/dev/random/build.sh
- Add copyright.
- Add libz for unit tests.
* src/sys/dev/random/dummy.c
- Remove; no longer needed. Functionality incorporated into randomdev.*.
* live_entropy_sources.c live_entropy_sources.h
- Remove; content moved.
- move content to randomdev.[ch] and optimise.
* src/sys/dev/random/random_adaptors.c src/sys/dev/random/random_adaptors.h
- Remove; plugability is no longer used. Compile-time algorithm
selection is the way to go.
* src/sys/dev/random/random_harvestq.c src/sys/dev/random/random_harvestq.h
- Add early (re)boot-time randomness caching.
* src/sys/dev/random/randomdev_soft.c src/sys/dev/random/randomdev_soft.h
- Remove; no longer needed.
* src/sys/dev/random/uint128.h
- Provide a fake uint128_t; if a real one ever arrived, we can use
that instead. All that is needed here is N=0, N++, N==0, and some
localised trickery is used to manufacture a 128-bit 0ULLL.
* src/sys/dev/random/unit_test.c src/sys/dev/random/unit_test.h
- Improve unit tests; previously the testing human needed clairvoyance;
now the test will do a basic check of compressibility. Clairvoyant
talent is still a good idea.
- This is still a long way off a proper unit test.
* src/sys/dev/random/fortuna.c src/sys/dev/random/fortuna.h
- Improve messy union to just uint128_t.
- Remove unneeded 'static struct fortuna_start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
* src/sys/dev/random/yarrow.c src/sys/dev/random/yarrow.h
- Improve messy union to just uint128_t.
- Remove unneeded 'staic struct start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
- Fix some magic numbers elsewhere used as FAST and SLOW.
Differential Revision: https://reviews.freebsd.org/D2025
Reviewed by: vsevolod,delphij,rwatson,trasz,jmg
Approved by: so (delphij)
2015-06-30 17:00:45 +00:00
|
|
|
20150630:
|
|
|
|
The default kernel entropy-processing algorithm is now
|
|
|
|
Fortuna, replacing Yarrow.
|
2015-06-16 23:32:28 +00:00
|
|
|
|
Huge cleanup of random(4) code.
* GENERAL
- Update copyright.
- Make kernel options for RANDOM_YARROW and RANDOM_DUMMY. Set
neither to ON, which means we want Fortuna
- If there is no 'device random' in the kernel, there will be NO
random(4) device in the kernel, and the KERN_ARND sysctl will
return nothing. With RANDOM_DUMMY there will be a random(4) that
always blocks.
- Repair kern.arandom (KERN_ARND sysctl). The old version went
through arc4random(9) and was a bit weird.
- Adjust arc4random stirring a bit - the existing code looks a little
suspect.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Redo read_random(9) so as to duplicate random(4)'s read internals.
This makes it a first-class citizen rather than a hack.
- Move stuff out of locked regions when it does not need to be
there.
- Trim RANDOM_DEBUG printfs. Some are excess to requirement, some
behind boot verbose.
- Use SYSINIT to sequence the startup.
- Fix init/deinit sysctl stuff.
- Make relevant sysctls also tunables.
- Add different harvesting "styles" to allow for different requirements
(direct, queue, fast).
- Add harvesting of FFS atime events. This needs to be checked for
weighing down the FS code.
- Add harvesting of slab allocator events. This needs to be checked for
weighing down the allocator code.
- Fix the random(9) manpage.
- Loadable modules are not present for now. These will be re-engineered
when the dust settles.
- Use macros for locks.
- Fix comments.
* src/share/man/...
- Update the man pages.
* src/etc/...
- The startup/shutdown work is done in D2924.
* src/UPDATING
- Add UPDATING announcement.
* src/sys/dev/random/build.sh
- Add copyright.
- Add libz for unit tests.
* src/sys/dev/random/dummy.c
- Remove; no longer needed. Functionality incorporated into randomdev.*.
* live_entropy_sources.c live_entropy_sources.h
- Remove; content moved.
- move content to randomdev.[ch] and optimise.
* src/sys/dev/random/random_adaptors.c src/sys/dev/random/random_adaptors.h
- Remove; plugability is no longer used. Compile-time algorithm
selection is the way to go.
* src/sys/dev/random/random_harvestq.c src/sys/dev/random/random_harvestq.h
- Add early (re)boot-time randomness caching.
* src/sys/dev/random/randomdev_soft.c src/sys/dev/random/randomdev_soft.h
- Remove; no longer needed.
* src/sys/dev/random/uint128.h
- Provide a fake uint128_t; if a real one ever arrived, we can use
that instead. All that is needed here is N=0, N++, N==0, and some
localised trickery is used to manufacture a 128-bit 0ULLL.
* src/sys/dev/random/unit_test.c src/sys/dev/random/unit_test.h
- Improve unit tests; previously the testing human needed clairvoyance;
now the test will do a basic check of compressibility. Clairvoyant
talent is still a good idea.
- This is still a long way off a proper unit test.
* src/sys/dev/random/fortuna.c src/sys/dev/random/fortuna.h
- Improve messy union to just uint128_t.
- Remove unneeded 'static struct fortuna_start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
* src/sys/dev/random/yarrow.c src/sys/dev/random/yarrow.h
- Improve messy union to just uint128_t.
- Remove unneeded 'staic struct start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
- Fix some magic numbers elsewhere used as FAST and SLOW.
Differential Revision: https://reviews.freebsd.org/D2025
Reviewed by: vsevolod,delphij,rwatson,trasz,jmg
Approved by: so (delphij)
2015-06-30 17:00:45 +00:00
|
|
|
Assuming you have 'device random' in your kernel config
|
|
|
|
file, the configurations allow a kernel option to override
|
|
|
|
this default. You may choose *ONE* of:
|
2015-06-16 23:32:28 +00:00
|
|
|
|
Huge cleanup of random(4) code.
* GENERAL
- Update copyright.
- Make kernel options for RANDOM_YARROW and RANDOM_DUMMY. Set
neither to ON, which means we want Fortuna
- If there is no 'device random' in the kernel, there will be NO
random(4) device in the kernel, and the KERN_ARND sysctl will
return nothing. With RANDOM_DUMMY there will be a random(4) that
always blocks.
- Repair kern.arandom (KERN_ARND sysctl). The old version went
through arc4random(9) and was a bit weird.
- Adjust arc4random stirring a bit - the existing code looks a little
suspect.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Redo read_random(9) so as to duplicate random(4)'s read internals.
This makes it a first-class citizen rather than a hack.
- Move stuff out of locked regions when it does not need to be
there.
- Trim RANDOM_DEBUG printfs. Some are excess to requirement, some
behind boot verbose.
- Use SYSINIT to sequence the startup.
- Fix init/deinit sysctl stuff.
- Make relevant sysctls also tunables.
- Add different harvesting "styles" to allow for different requirements
(direct, queue, fast).
- Add harvesting of FFS atime events. This needs to be checked for
weighing down the FS code.
- Add harvesting of slab allocator events. This needs to be checked for
weighing down the allocator code.
- Fix the random(9) manpage.
- Loadable modules are not present for now. These will be re-engineered
when the dust settles.
- Use macros for locks.
- Fix comments.
* src/share/man/...
- Update the man pages.
* src/etc/...
- The startup/shutdown work is done in D2924.
* src/UPDATING
- Add UPDATING announcement.
* src/sys/dev/random/build.sh
- Add copyright.
- Add libz for unit tests.
* src/sys/dev/random/dummy.c
- Remove; no longer needed. Functionality incorporated into randomdev.*.
* live_entropy_sources.c live_entropy_sources.h
- Remove; content moved.
- move content to randomdev.[ch] and optimise.
* src/sys/dev/random/random_adaptors.c src/sys/dev/random/random_adaptors.h
- Remove; plugability is no longer used. Compile-time algorithm
selection is the way to go.
* src/sys/dev/random/random_harvestq.c src/sys/dev/random/random_harvestq.h
- Add early (re)boot-time randomness caching.
* src/sys/dev/random/randomdev_soft.c src/sys/dev/random/randomdev_soft.h
- Remove; no longer needed.
* src/sys/dev/random/uint128.h
- Provide a fake uint128_t; if a real one ever arrived, we can use
that instead. All that is needed here is N=0, N++, N==0, and some
localised trickery is used to manufacture a 128-bit 0ULLL.
* src/sys/dev/random/unit_test.c src/sys/dev/random/unit_test.h
- Improve unit tests; previously the testing human needed clairvoyance;
now the test will do a basic check of compressibility. Clairvoyant
talent is still a good idea.
- This is still a long way off a proper unit test.
* src/sys/dev/random/fortuna.c src/sys/dev/random/fortuna.h
- Improve messy union to just uint128_t.
- Remove unneeded 'static struct fortuna_start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
* src/sys/dev/random/yarrow.c src/sys/dev/random/yarrow.h
- Improve messy union to just uint128_t.
- Remove unneeded 'staic struct start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
- Fix some magic numbers elsewhere used as FAST and SLOW.
Differential Revision: https://reviews.freebsd.org/D2025
Reviewed by: vsevolod,delphij,rwatson,trasz,jmg
Approved by: so (delphij)
2015-06-30 17:00:45 +00:00
|
|
|
options RANDOM_YARROW # Legacy /dev/random algorithm.
|
|
|
|
options RANDOM_DUMMY # Blocking-only driver.
|
2015-06-16 23:32:28 +00:00
|
|
|
|
Huge cleanup of random(4) code.
* GENERAL
- Update copyright.
- Make kernel options for RANDOM_YARROW and RANDOM_DUMMY. Set
neither to ON, which means we want Fortuna
- If there is no 'device random' in the kernel, there will be NO
random(4) device in the kernel, and the KERN_ARND sysctl will
return nothing. With RANDOM_DUMMY there will be a random(4) that
always blocks.
- Repair kern.arandom (KERN_ARND sysctl). The old version went
through arc4random(9) and was a bit weird.
- Adjust arc4random stirring a bit - the existing code looks a little
suspect.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Redo read_random(9) so as to duplicate random(4)'s read internals.
This makes it a first-class citizen rather than a hack.
- Move stuff out of locked regions when it does not need to be
there.
- Trim RANDOM_DEBUG printfs. Some are excess to requirement, some
behind boot verbose.
- Use SYSINIT to sequence the startup.
- Fix init/deinit sysctl stuff.
- Make relevant sysctls also tunables.
- Add different harvesting "styles" to allow for different requirements
(direct, queue, fast).
- Add harvesting of FFS atime events. This needs to be checked for
weighing down the FS code.
- Add harvesting of slab allocator events. This needs to be checked for
weighing down the allocator code.
- Fix the random(9) manpage.
- Loadable modules are not present for now. These will be re-engineered
when the dust settles.
- Use macros for locks.
- Fix comments.
* src/share/man/...
- Update the man pages.
* src/etc/...
- The startup/shutdown work is done in D2924.
* src/UPDATING
- Add UPDATING announcement.
* src/sys/dev/random/build.sh
- Add copyright.
- Add libz for unit tests.
* src/sys/dev/random/dummy.c
- Remove; no longer needed. Functionality incorporated into randomdev.*.
* live_entropy_sources.c live_entropy_sources.h
- Remove; content moved.
- move content to randomdev.[ch] and optimise.
* src/sys/dev/random/random_adaptors.c src/sys/dev/random/random_adaptors.h
- Remove; plugability is no longer used. Compile-time algorithm
selection is the way to go.
* src/sys/dev/random/random_harvestq.c src/sys/dev/random/random_harvestq.h
- Add early (re)boot-time randomness caching.
* src/sys/dev/random/randomdev_soft.c src/sys/dev/random/randomdev_soft.h
- Remove; no longer needed.
* src/sys/dev/random/uint128.h
- Provide a fake uint128_t; if a real one ever arrived, we can use
that instead. All that is needed here is N=0, N++, N==0, and some
localised trickery is used to manufacture a 128-bit 0ULLL.
* src/sys/dev/random/unit_test.c src/sys/dev/random/unit_test.h
- Improve unit tests; previously the testing human needed clairvoyance;
now the test will do a basic check of compressibility. Clairvoyant
talent is still a good idea.
- This is still a long way off a proper unit test.
* src/sys/dev/random/fortuna.c src/sys/dev/random/fortuna.h
- Improve messy union to just uint128_t.
- Remove unneeded 'static struct fortuna_start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
* src/sys/dev/random/yarrow.c src/sys/dev/random/yarrow.h
- Improve messy union to just uint128_t.
- Remove unneeded 'staic struct start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
- Fix some magic numbers elsewhere used as FAST and SLOW.
Differential Revision: https://reviews.freebsd.org/D2025
Reviewed by: vsevolod,delphij,rwatson,trasz,jmg
Approved by: so (delphij)
2015-06-30 17:00:45 +00:00
|
|
|
If you have neither, you get Fortuna. For most people,
|
|
|
|
read no further, Fortuna will give a /dev/random that works
|
|
|
|
like it always used to, and the difference will be irrelevant.
|
|
|
|
|
|
|
|
If you remove 'device random', you get *NO* kernel-processed
|
2015-11-17 19:43:40 +00:00
|
|
|
entropy at all. This may be acceptable to folks building
|
Huge cleanup of random(4) code.
* GENERAL
- Update copyright.
- Make kernel options for RANDOM_YARROW and RANDOM_DUMMY. Set
neither to ON, which means we want Fortuna
- If there is no 'device random' in the kernel, there will be NO
random(4) device in the kernel, and the KERN_ARND sysctl will
return nothing. With RANDOM_DUMMY there will be a random(4) that
always blocks.
- Repair kern.arandom (KERN_ARND sysctl). The old version went
through arc4random(9) and was a bit weird.
- Adjust arc4random stirring a bit - the existing code looks a little
suspect.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Redo read_random(9) so as to duplicate random(4)'s read internals.
This makes it a first-class citizen rather than a hack.
- Move stuff out of locked regions when it does not need to be
there.
- Trim RANDOM_DEBUG printfs. Some are excess to requirement, some
behind boot verbose.
- Use SYSINIT to sequence the startup.
- Fix init/deinit sysctl stuff.
- Make relevant sysctls also tunables.
- Add different harvesting "styles" to allow for different requirements
(direct, queue, fast).
- Add harvesting of FFS atime events. This needs to be checked for
weighing down the FS code.
- Add harvesting of slab allocator events. This needs to be checked for
weighing down the allocator code.
- Fix the random(9) manpage.
- Loadable modules are not present for now. These will be re-engineered
when the dust settles.
- Use macros for locks.
- Fix comments.
* src/share/man/...
- Update the man pages.
* src/etc/...
- The startup/shutdown work is done in D2924.
* src/UPDATING
- Add UPDATING announcement.
* src/sys/dev/random/build.sh
- Add copyright.
- Add libz for unit tests.
* src/sys/dev/random/dummy.c
- Remove; no longer needed. Functionality incorporated into randomdev.*.
* live_entropy_sources.c live_entropy_sources.h
- Remove; content moved.
- move content to randomdev.[ch] and optimise.
* src/sys/dev/random/random_adaptors.c src/sys/dev/random/random_adaptors.h
- Remove; plugability is no longer used. Compile-time algorithm
selection is the way to go.
* src/sys/dev/random/random_harvestq.c src/sys/dev/random/random_harvestq.h
- Add early (re)boot-time randomness caching.
* src/sys/dev/random/randomdev_soft.c src/sys/dev/random/randomdev_soft.h
- Remove; no longer needed.
* src/sys/dev/random/uint128.h
- Provide a fake uint128_t; if a real one ever arrived, we can use
that instead. All that is needed here is N=0, N++, N==0, and some
localised trickery is used to manufacture a 128-bit 0ULLL.
* src/sys/dev/random/unit_test.c src/sys/dev/random/unit_test.h
- Improve unit tests; previously the testing human needed clairvoyance;
now the test will do a basic check of compressibility. Clairvoyant
talent is still a good idea.
- This is still a long way off a proper unit test.
* src/sys/dev/random/fortuna.c src/sys/dev/random/fortuna.h
- Improve messy union to just uint128_t.
- Remove unneeded 'static struct fortuna_start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
* src/sys/dev/random/yarrow.c src/sys/dev/random/yarrow.h
- Improve messy union to just uint128_t.
- Remove unneeded 'staic struct start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
- Fix some magic numbers elsewhere used as FAST and SLOW.
Differential Revision: https://reviews.freebsd.org/D2025
Reviewed by: vsevolod,delphij,rwatson,trasz,jmg
Approved by: so (delphij)
2015-06-30 17:00:45 +00:00
|
|
|
embedded systems, but has complications. Carry on reading,
|
|
|
|
and it is assumed you know what you need.
|
|
|
|
|
|
|
|
*PLEASE* read random(4) and random(9) if you are in the
|
2015-11-17 19:43:40 +00:00
|
|
|
habit of tweaking kernel configs, and/or if you are a member
|
Huge cleanup of random(4) code.
* GENERAL
- Update copyright.
- Make kernel options for RANDOM_YARROW and RANDOM_DUMMY. Set
neither to ON, which means we want Fortuna
- If there is no 'device random' in the kernel, there will be NO
random(4) device in the kernel, and the KERN_ARND sysctl will
return nothing. With RANDOM_DUMMY there will be a random(4) that
always blocks.
- Repair kern.arandom (KERN_ARND sysctl). The old version went
through arc4random(9) and was a bit weird.
- Adjust arc4random stirring a bit - the existing code looks a little
suspect.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Redo read_random(9) so as to duplicate random(4)'s read internals.
This makes it a first-class citizen rather than a hack.
- Move stuff out of locked regions when it does not need to be
there.
- Trim RANDOM_DEBUG printfs. Some are excess to requirement, some
behind boot verbose.
- Use SYSINIT to sequence the startup.
- Fix init/deinit sysctl stuff.
- Make relevant sysctls also tunables.
- Add different harvesting "styles" to allow for different requirements
(direct, queue, fast).
- Add harvesting of FFS atime events. This needs to be checked for
weighing down the FS code.
- Add harvesting of slab allocator events. This needs to be checked for
weighing down the allocator code.
- Fix the random(9) manpage.
- Loadable modules are not present for now. These will be re-engineered
when the dust settles.
- Use macros for locks.
- Fix comments.
* src/share/man/...
- Update the man pages.
* src/etc/...
- The startup/shutdown work is done in D2924.
* src/UPDATING
- Add UPDATING announcement.
* src/sys/dev/random/build.sh
- Add copyright.
- Add libz for unit tests.
* src/sys/dev/random/dummy.c
- Remove; no longer needed. Functionality incorporated into randomdev.*.
* live_entropy_sources.c live_entropy_sources.h
- Remove; content moved.
- move content to randomdev.[ch] and optimise.
* src/sys/dev/random/random_adaptors.c src/sys/dev/random/random_adaptors.h
- Remove; plugability is no longer used. Compile-time algorithm
selection is the way to go.
* src/sys/dev/random/random_harvestq.c src/sys/dev/random/random_harvestq.h
- Add early (re)boot-time randomness caching.
* src/sys/dev/random/randomdev_soft.c src/sys/dev/random/randomdev_soft.h
- Remove; no longer needed.
* src/sys/dev/random/uint128.h
- Provide a fake uint128_t; if a real one ever arrived, we can use
that instead. All that is needed here is N=0, N++, N==0, and some
localised trickery is used to manufacture a 128-bit 0ULLL.
* src/sys/dev/random/unit_test.c src/sys/dev/random/unit_test.h
- Improve unit tests; previously the testing human needed clairvoyance;
now the test will do a basic check of compressibility. Clairvoyant
talent is still a good idea.
- This is still a long way off a proper unit test.
* src/sys/dev/random/fortuna.c src/sys/dev/random/fortuna.h
- Improve messy union to just uint128_t.
- Remove unneeded 'static struct fortuna_start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
* src/sys/dev/random/yarrow.c src/sys/dev/random/yarrow.h
- Improve messy union to just uint128_t.
- Remove unneeded 'staic struct start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
- Fix some magic numbers elsewhere used as FAST and SLOW.
Differential Revision: https://reviews.freebsd.org/D2025
Reviewed by: vsevolod,delphij,rwatson,trasz,jmg
Approved by: so (delphij)
2015-06-30 17:00:45 +00:00
|
|
|
of the embedded community, wanting specific and not-usual
|
|
|
|
behaviour from your security subsystems.
|
|
|
|
|
|
|
|
NOTE!! If you use RANDOM_DUMMY and/or have no 'device
|
|
|
|
random', you will NOT have a functioning /dev/random, and
|
|
|
|
many cryptographic features will not work, including SSH.
|
|
|
|
You may also find strange behaviour from the random(3) set
|
|
|
|
of library functions, in particular sranddev(3), srandomdev(3)
|
|
|
|
and arc4random(3). The reason for this is that the KERN_ARND
|
|
|
|
sysctl only returns entropy if it thinks it has some to
|
|
|
|
share, and with RANDOM_DUMMY or no 'device random' this
|
|
|
|
will never happen.
|
|
|
|
|
2015-06-25 01:42:59 +00:00
|
|
|
20150623:
|
|
|
|
An additional fix for the issue described in the 20150614 sendmail
|
2017-12-27 03:23:58 +00:00
|
|
|
entry below has been committed in revision 284717.
|
2015-06-16 23:32:28 +00:00
|
|
|
|
2015-06-16 20:58:33 +00:00
|
|
|
20150616:
|
|
|
|
FreeBSD's old make (fmake) has been removed from the system. It is
|
|
|
|
available as the devel/fmake port or via pkg install fmake.
|
2015-07-19 14:34:35 +00:00
|
|
|
|
2015-06-16 03:03:26 +00:00
|
|
|
20150615:
|
|
|
|
The fix for the issue described in the 20150614 sendmail entry
|
2017-12-27 03:23:58 +00:00
|
|
|
below has been committed in revision 284436. The work
|
2015-06-16 03:03:26 +00:00
|
|
|
around described in that entry is no longer needed unless the
|
|
|
|
default setting is overridden by a confDH_PARAMETERS configuration
|
|
|
|
setting of '5' or pointing to a 512 bit DH parameter file.
|
|
|
|
|
2015-06-15 06:38:59 +00:00
|
|
|
20150614:
|
|
|
|
ALLOW_DEPRECATED_ATF_TOOLS/ATFFILE support has been removed from
|
|
|
|
atf.test.mk (included from bsd.test.mk). Please upgrade devel/atf
|
|
|
|
and devel/kyua to version 0.20+ and adjust any calling code to work
|
|
|
|
with Kyuafile and kyua.
|
|
|
|
|
2015-06-15 04:18:29 +00:00
|
|
|
20150614:
|
|
|
|
The import of openssl to address the FreeBSD-SA-15:10.openssl
|
|
|
|
security advisory includes a change which rejects handshakes
|
|
|
|
with DH parameters below 768 bits. sendmail releases prior
|
|
|
|
to 8.15.2 (not yet released), defaulted to a 512 bit
|
|
|
|
DH parameter setting for client connections. To work around
|
|
|
|
this interoperability, sendmail can be configured to use a
|
|
|
|
2048 bit DH parameter by:
|
|
|
|
|
2015-07-19 14:34:35 +00:00
|
|
|
1. Edit /etc/mail/`hostname`.mc
|
2015-06-15 04:18:29 +00:00
|
|
|
2. If a setting for confDH_PARAMETERS does not exist or
|
|
|
|
exists and is set to a string beginning with '5',
|
|
|
|
replace it with '2'.
|
|
|
|
3. If a setting for confDH_PARAMETERS exists and is set to
|
|
|
|
a file path, create a new file with:
|
|
|
|
openssl dhparam -out /path/to/file 2048
|
|
|
|
4. Rebuild the .cf file:
|
|
|
|
cd /etc/mail/; make; make install
|
|
|
|
5. Restart sendmail:
|
|
|
|
cd /etc/mail/; make restart
|
|
|
|
|
|
|
|
A sendmail patch is coming, at which time this file will be
|
|
|
|
updated.
|
|
|
|
|
2015-06-04 07:37:09 +00:00
|
|
|
20150604:
|
|
|
|
Generation of legacy formatted entries have been disabled by default
|
|
|
|
in pwd_mkdb(8), as all base system consumers of the legacy formatted
|
|
|
|
entries were converted to use the new format by default when the new,
|
|
|
|
machine independent format have been added and supported since FreeBSD
|
|
|
|
5.x.
|
|
|
|
|
|
|
|
Please see the pwd_mkdb(8) manual page for further details.
|
|
|
|
|
2015-05-25 13:43:03 +00:00
|
|
|
20150525:
|
|
|
|
Clang and llvm have been upgraded to 3.6.1 release. Please see the
|
|
|
|
20141231 entry below for information about prerequisites and upgrading,
|
|
|
|
if you are not already using 3.5.0 or higher.
|
|
|
|
|
2015-05-22 03:16:18 +00:00
|
|
|
20150521:
|
|
|
|
TI platform code switched to using vendor DTS files and this update
|
|
|
|
may break existing systems running on Beaglebone, Beaglebone Black,
|
|
|
|
and Pandaboard:
|
|
|
|
|
|
|
|
- dtb files should be regenerated/reinstalled. Filenames are the
|
|
|
|
same but content is different now
|
|
|
|
- GPIO addressing was changed, now each GPIO bank (32 pins per bank)
|
|
|
|
has its own /dev/gpiocX device, e.g. pin 121 on /dev/gpioc0 in old
|
|
|
|
addressing scheme is now pin 25 on /dev/gpioc3.
|
|
|
|
- Pandaboard: /etc/ttys should be updated, serial console device is
|
|
|
|
now /dev/ttyu2, not /dev/ttyu0
|
|
|
|
|
2015-05-01 21:26:40 +00:00
|
|
|
20150501:
|
|
|
|
soelim(1) from gnu/usr.bin/groff has been replaced by usr.bin/soelim.
|
|
|
|
If you need the GNU extension from groff soelim(1), install groff
|
|
|
|
from package: pkg install groff, or via ports: textproc/groff.
|
|
|
|
|
2015-04-29 14:17:40 +00:00
|
|
|
20150423:
|
2015-04-29 00:49:00 +00:00
|
|
|
chmod, chflags, chown and chgrp now affect symlinks in -R mode as
|
|
|
|
defined in symlink(7); previously symlinks were silently ignored.
|
|
|
|
|
2015-04-15 09:09:20 +00:00
|
|
|
20150415:
|
|
|
|
The const qualifier has been removed from iconv(3) to comply with
|
|
|
|
POSIX. The ports tree is aware of this from r384038 onwards.
|
|
|
|
|
2015-04-16 15:34:22 +00:00
|
|
|
20150416:
|
|
|
|
Libraries specified by LIBADD in Makefiles must have a corresponding
|
|
|
|
DPADD_<lib> variable to ensure correct dependencies. This is now
|
|
|
|
enforced in src.libnames.mk.
|
|
|
|
|
2015-03-24 18:09:07 +00:00
|
|
|
20150324:
|
|
|
|
From legacy ata(4) driver was removed support for SATA controllers
|
|
|
|
supported by more functional drivers ahci(4), siis(4) and mvs(4).
|
|
|
|
Kernel modules ataahci and ataadaptec were removed completely,
|
|
|
|
replaced by ahci and mvs modules respectively.
|
|
|
|
|
2015-03-15 11:56:43 +00:00
|
|
|
20150315:
|
2015-03-15 13:31:13 +00:00
|
|
|
Clang, llvm and lldb have been upgraded to 3.6.0 release. Please see
|
|
|
|
the 20141231 entry below for information about prerequisites and
|
|
|
|
upgrading, if you are not already using 3.5.0 or higher.
|
2015-01-26 19:41:26 +00:00
|
|
|
|
2015-03-07 20:14:46 +00:00
|
|
|
20150307:
|
|
|
|
The 32-bit PowerPC kernel has been changed to a position-independent
|
|
|
|
executable. This can only be booted with a version of loader(8)
|
|
|
|
newer than January 31, 2015, so make sure to update both world and
|
|
|
|
kernel before rebooting.
|
|
|
|
|
2015-02-17 21:18:17 +00:00
|
|
|
20150217:
|
|
|
|
If you are running a -CURRENT kernel since r273872 (Oct 30th, 2014),
|
2015-02-18 08:24:00 +00:00
|
|
|
but before r278950, the RNG was not seeded properly. Immediately
|
|
|
|
upgrade the kernel to r278950 or later and regenerate any keys (e.g.
|
2015-02-17 21:18:17 +00:00
|
|
|
ssh keys or openssl keys) that were generated w/ a kernel from that
|
2015-02-21 13:00:52 +00:00
|
|
|
range. This does not affect programs that directly used /dev/random
|
2015-02-19 05:20:59 +00:00
|
|
|
or /dev/urandom. All userland uses of arc4random(3) are affected.
|
2015-02-17 21:18:17 +00:00
|
|
|
|
2015-02-10 18:05:58 +00:00
|
|
|
20150210:
|
|
|
|
The autofs(4) ABI was changed in order to restore binary compatibility
|
|
|
|
with 10.1-RELEASE. The automountd(8) daemon needs to be rebuilt to work
|
|
|
|
with the new kernel.
|
|
|
|
|
2015-01-31 19:16:51 +00:00
|
|
|
20150131:
|
|
|
|
The powerpc64 kernel has been changed to a position-independent
|
|
|
|
executable. This can only be booted with a new version of loader(8),
|
|
|
|
so make sure to update both world and kernel before rebooting.
|
|
|
|
|
2015-01-18 14:14:47 +00:00
|
|
|
20150118:
|
|
|
|
Clang and llvm have been upgraded to 3.5.1 release. This is a bugfix
|
|
|
|
only release, no new features have been added. Please see the 20141231
|
|
|
|
entry below for information about prerequisites and upgrading, if you
|
|
|
|
are not already using 3.5.0.
|
|
|
|
|
2015-01-07 22:02:37 +00:00
|
|
|
20150107:
|
|
|
|
ELF tools addr2line, elfcopy (strip), nm, size, and strings are now
|
|
|
|
taken from the ELF Tool Chain project rather than GNU binutils. They
|
|
|
|
should be drop-in replacements, with the addition of arm64 support.
|
|
|
|
The WITHOUT_ELFTOOLCHAIN_TOOLS= knob may be used to obtain the
|
2015-08-05 18:30:00 +00:00
|
|
|
binutils tools, if necessary. See 20150805 for updated information.
|
2015-01-07 22:02:37 +00:00
|
|
|
|
2015-01-05 15:09:00 +00:00
|
|
|
20150105:
|
|
|
|
The default Unbound configuration now enables remote control
|
|
|
|
using a local socket. Users who have already enabled the
|
|
|
|
local_unbound service should regenerate their configuration
|
|
|
|
by running "service local_unbound setup" as root.
|
2015-07-19 14:34:35 +00:00
|
|
|
|
2015-01-02 18:45:03 +00:00
|
|
|
20150102:
|
|
|
|
The GNU texinfo and GNU info pages have been removed.
|
|
|
|
To be able to view GNU info pages please install texinfo from ports.
|
|
|
|
|
2014-12-30 22:46:20 +00:00
|
|
|
20141231:
|
|
|
|
Clang, llvm and lldb have been upgraded to 3.5.0 release.
|
|
|
|
|
2014-12-31 18:11:09 +00:00
|
|
|
As of this release, a prerequisite for building clang, llvm and lldb is
|
|
|
|
a C++11 capable compiler and C++11 standard library. This means that to
|
2014-12-30 22:46:20 +00:00
|
|
|
be able to successfully build the cross-tools stage of buildworld, with
|
|
|
|
clang as the bootstrap compiler, your system compiler or cross compiler
|
|
|
|
should either be clang 3.3 or later, or gcc 4.8 or later, and your
|
|
|
|
system C++ library should be libc++, or libdstdc++ from gcc 4.8 or
|
|
|
|
later.
|
|
|
|
|
2014-12-31 18:11:09 +00:00
|
|
|
On any standard FreeBSD 10.x or 11.x installation, where clang and
|
|
|
|
libc++ are on by default (that is, on x86 or arm), this should work out
|
|
|
|
of the box.
|
2014-12-30 22:46:20 +00:00
|
|
|
|
|
|
|
On 9.x installations where clang is enabled by default, e.g. on x86 and
|
|
|
|
powerpc, libc++ will not be enabled by default, so libc++ should be
|
|
|
|
built (with clang) and installed first. If both clang and libc++ are
|
|
|
|
missing, build clang first, then use it to build libc++.
|
|
|
|
|
|
|
|
On 8.x and earlier installations, upgrade to 9.x first, and then follow
|
|
|
|
the instructions for 9.x above.
|
|
|
|
|
2014-12-31 18:11:09 +00:00
|
|
|
Sparc64 and mips users are unaffected, as they still use gcc 4.2.1 by
|
|
|
|
default, and do not build clang.
|
|
|
|
|
|
|
|
Many embedded systems are resource constrained, and will not be able to
|
|
|
|
build clang in a reasonable time, or in some cases at all. In those
|
|
|
|
cases, cross building bootable systems on amd64 is a workaround.
|
|
|
|
|
2014-12-30 22:46:20 +00:00
|
|
|
This new version of clang introduces a number of new warnings, of which
|
|
|
|
the following are most likely to appear:
|
|
|
|
|
|
|
|
-Wabsolute-value
|
|
|
|
|
|
|
|
This warns in two cases, for both C and C++:
|
|
|
|
* When the code is trying to take the absolute value of an unsigned
|
|
|
|
quantity, which is effectively a no-op, and almost never what was
|
|
|
|
intended. The code should be fixed, if at all possible. If you are
|
|
|
|
sure that the unsigned quantity can be safely cast to signed, without
|
|
|
|
loss of information or undefined behavior, you can add an explicit
|
|
|
|
cast, or disable the warning.
|
|
|
|
|
|
|
|
* When the code is trying to take an absolute value, but the called
|
|
|
|
abs() variant is for the wrong type, which can lead to truncation.
|
|
|
|
If you want to disable the warning instead of fixing the code, please
|
|
|
|
make sure that truncation will not occur, or it might lead to unwanted
|
|
|
|
side-effects.
|
|
|
|
|
|
|
|
-Wtautological-undefined-compare and
|
|
|
|
-Wundefined-bool-conversion
|
|
|
|
|
|
|
|
These warn when C++ code is trying to compare 'this' against NULL, while
|
|
|
|
'this' should never be NULL in well-defined C++ code. However, there is
|
|
|
|
some legacy (pre C++11) code out there, which actively abuses this
|
|
|
|
feature, which was less strictly defined in previous C++ versions.
|
|
|
|
|
|
|
|
Squid and openjdk do this, for example. The warning can be turned off
|
|
|
|
for C++98 and earlier, but compiling the code in C++11 mode might result
|
|
|
|
in unexpected behavior; for example, the parts of the program that are
|
|
|
|
unreachable could be optimized away.
|
|
|
|
|
2014-12-23 01:32:18 +00:00
|
|
|
20141222:
|
|
|
|
The old NFS client and server (kernel options NFSCLIENT, NFSSERVER)
|
|
|
|
kernel sources have been removed. The .h files remain, since some
|
|
|
|
utilities include them. This will need to be fixed later.
|
|
|
|
If "mount -t oldnfs ..." is attempted, it will fail.
|
|
|
|
If the "-o" option on mountd(8), nfsd(8) or nfsstat(1) is used,
|
|
|
|
the utilities will report errors.
|
|
|
|
|
2014-12-16 20:45:17 +00:00
|
|
|
20141121:
|
|
|
|
The handling of LOCAL_LIB_DIRS has been altered to skip addition of
|
|
|
|
directories to top level SUBDIR variable when their parent
|
|
|
|
directory is included in LOCAL_DIRS. Users with build systems with
|
|
|
|
such hierarchies and without SUBDIR entries in the parent
|
|
|
|
directory Makefiles should add them or add the directories to
|
|
|
|
LOCAL_DIRS.
|
|
|
|
|
2014-11-09 21:33:01 +00:00
|
|
|
20141109:
|
2015-01-05 13:46:37 +00:00
|
|
|
faith(4) and faithd(8) have been removed from the base system. Faith
|
2014-11-09 21:33:01 +00:00
|
|
|
has been obsolete for a very long time.
|
|
|
|
|
2014-11-04 10:18:03 +00:00
|
|
|
20141104:
|
|
|
|
vt(4), the new console driver, is enabled by default. It brings
|
|
|
|
support for Unicode and double-width characters, as well as
|
|
|
|
support for UEFI and integration with the KMS kernel video
|
|
|
|
drivers.
|
|
|
|
|
|
|
|
You may need to update your console settings in /etc/rc.conf,
|
|
|
|
most probably the keymap. During boot, /etc/rc.d/syscons will
|
|
|
|
indicate what you need to do.
|
|
|
|
|
|
|
|
vt(4) still has issues and lacks some features compared to
|
|
|
|
syscons(4). See the wiki for up-to-date information:
|
|
|
|
https://wiki.freebsd.org/Newcons
|
|
|
|
|
|
|
|
If you want to keep using syscons(4), you can do so by adding
|
|
|
|
the following line to /boot/loader.conf:
|
|
|
|
kern.vty=sc
|
|
|
|
|
2014-11-03 07:18:42 +00:00
|
|
|
20141102:
|
|
|
|
pjdfstest has been integrated into kyua as an opt-in test suite.
|
2015-01-05 13:46:37 +00:00
|
|
|
Please see share/doc/pjdfstest/README for more details on how to
|
2014-11-03 07:18:42 +00:00
|
|
|
execute it.
|
|
|
|
|
2014-10-09 23:05:31 +00:00
|
|
|
20141009:
|
|
|
|
gperf has been removed from the base system for architectures
|
|
|
|
that use clang. Ports that require gperf will obtain it from the
|
|
|
|
devel/gperf port.
|
|
|
|
|
2014-09-24 07:00:35 +00:00
|
|
|
20140923:
|
|
|
|
pjdfstest has been moved from tools/regression/pjdfstest to
|
|
|
|
contrib/pjdfstest .
|
|
|
|
|
2014-09-22 19:07:27 +00:00
|
|
|
20140922:
|
|
|
|
At svn r271982, The default linux compat kernel ABI has been adjusted
|
|
|
|
to 2.6.18 in support of the linux-c6 compat ports infrastructure
|
|
|
|
update. If you wish to continue using the linux-f10 compat ports,
|
|
|
|
add compat.linux.osrelease=2.6.16 to your local sysctl.conf. Users are
|
|
|
|
encouraged to update their linux-compat packages to linux-c6 during
|
|
|
|
their next update cycle.
|
|
|
|
|
2014-07-29 23:11:05 +00:00
|
|
|
20140729:
|
|
|
|
The ofwfb driver, used to provide a graphics console on PowerPC when
|
2015-01-05 13:46:37 +00:00
|
|
|
using vt(4), no longer allows mmap() of all physical memory. This
|
2014-07-29 23:11:05 +00:00
|
|
|
will prevent Xorg on PowerPC with some ATI graphics cards from
|
|
|
|
initializing properly unless x11-servers/xorg-server is updated to
|
|
|
|
1.12.4_8 or newer.
|
|
|
|
|
2014-07-24 06:16:51 +00:00
|
|
|
20140723:
|
|
|
|
The xdev targets have been converted to using TARGET and
|
|
|
|
TARGET_ARCH instead of XDEV and XDEV_ARCH.
|
|
|
|
|
2014-07-19 00:11:51 +00:00
|
|
|
20140719:
|
|
|
|
The default unbound configuration has been modified to address
|
|
|
|
issues with reverse lookups on networks that use private
|
|
|
|
address ranges. If you use the local_unbound service, run
|
|
|
|
"service local_unbound setup" as root to regenerate your
|
|
|
|
configuration, then "service local_unbound reload" to load the
|
|
|
|
new configuration.
|
|
|
|
|
2014-07-10 15:05:41 +00:00
|
|
|
20140709:
|
|
|
|
The GNU texinfo and GNU info pages are not built and installed
|
|
|
|
anymore, WITH_INFO knob has been added to allow to built and install
|
|
|
|
them again.
|
2015-01-02 18:45:03 +00:00
|
|
|
UPDATE: see 20150102 entry on texinfo's removal
|
2014-07-10 15:05:41 +00:00
|
|
|
|
2014-07-09 15:52:30 +00:00
|
|
|
20140708:
|
|
|
|
The GNU readline library is now an INTERNALLIB - that is, it is
|
|
|
|
statically linked into consumers (GDB and variants) in the base
|
|
|
|
system, and the shared library is no longer installed. The
|
|
|
|
devel/readline port is available for third party software that
|
|
|
|
requires readline.
|
|
|
|
|
2014-07-02 22:34:06 +00:00
|
|
|
20140702:
|
|
|
|
The Itanium architecture (ia64) has been removed from the list of
|
|
|
|
known architectures. This is the first step in the removal of the
|
|
|
|
architecture.
|
|
|
|
|
2014-07-01 22:32:11 +00:00
|
|
|
20140701:
|
|
|
|
Commit r268115 has added NFSv4.1 server support, merged from
|
|
|
|
projects/nfsv4.1-server. Since this includes changes to the
|
|
|
|
internal interfaces between the NFS related modules, a full
|
|
|
|
build of the kernel and modules will be necessary.
|
|
|
|
__FreeBSD_version has been bumped.
|
|
|
|
|
2014-06-30 00:20:12 +00:00
|
|
|
20140629:
|
|
|
|
The WITHOUT_VT_SUPPORT kernel config knob has been renamed
|
|
|
|
WITHOUT_VT. (The other _SUPPORT knobs have a consistent meaning
|
|
|
|
which differs from the behaviour controlled by this knob.)
|
|
|
|
|
2014-06-24 18:56:45 +00:00
|
|
|
20140619:
|
|
|
|
Maximal length of the serial number in CTL was increased from 16 to
|
|
|
|
64 chars, that breaks ABI. All CTL-related tools, such as ctladm
|
|
|
|
and ctld, need to be rebuilt to work with a new kernel.
|
|
|
|
|
2014-06-06 17:05:43 +00:00
|
|
|
20140606:
|
|
|
|
The libatf-c and libatf-c++ major versions were downgraded to 0 and
|
|
|
|
1 respectively to match the upstream numbers. They were out of
|
|
|
|
sync because, when they were originally added to FreeBSD, the
|
|
|
|
upstream versions were not respected. These libraries are private
|
|
|
|
and not yet built by default, so renumbering them should be a
|
|
|
|
non-issue. However, unclean source trees will yield broken test
|
|
|
|
programs once the operator executes "make delete-old-libs" after a
|
|
|
|
"make installworld".
|
|
|
|
|
2014-06-06 18:58:06 +00:00
|
|
|
Additionally, the atf-sh binary was made private by moving it into
|
|
|
|
/usr/libexec/. Already-built shell test programs will keep the
|
|
|
|
path to the old binary so they will break after "make delete-old"
|
|
|
|
is run.
|
|
|
|
|
2014-06-06 17:05:43 +00:00
|
|
|
If you are using WITH_TESTS=yes (not the default), wipe the object
|
|
|
|
tree and rebuild from scratch to prevent spurious test failures.
|
2014-06-06 18:58:06 +00:00
|
|
|
This is only needed once: the misnumbered libraries and misplaced
|
|
|
|
binaries have been added to OptionalObsoleteFiles.inc so they will
|
|
|
|
be removed during a clean upgrade.
|
2014-06-06 17:05:43 +00:00
|
|
|
|
2014-05-12 18:45:56 +00:00
|
|
|
20140512:
|
|
|
|
Clang and llvm have been upgraded to 3.4.1 release.
|
|
|
|
|
2014-05-08 15:58:34 +00:00
|
|
|
20140508:
|
|
|
|
We bogusly installed src.opts.mk in /usr/share/mk. This file should
|
|
|
|
be removed to avoid issues in the future (and has been added to
|
|
|
|
ObsoleteFiles.inc).
|
|
|
|
|
2014-05-06 04:22:29 +00:00
|
|
|
20140505:
|
|
|
|
/etc/src.conf now affects only builds of the FreeBSD src tree. In the
|
|
|
|
past, it affected all builds that used the bsd.*.mk files. The old
|
|
|
|
behavior was a bug, but people may have relied upon it. To get this
|
|
|
|
behavior back, you can .include /etc/src.conf from /etc/make.conf
|
|
|
|
(which is still global and isn't changed). This also changes the
|
|
|
|
behavior of incremental builds inside the tree of individual
|
2014-06-24 16:10:48 +00:00
|
|
|
directories. Set MAKESYSPATH to ".../share/mk" to do that.
|
2014-05-06 04:22:29 +00:00
|
|
|
Although this has survived make universe and some upgrade scenarios,
|
2014-05-06 15:44:34 +00:00
|
|
|
other upgrade scenarios may have broken. At least one form of
|
|
|
|
temporary breakage was fixed with MAKESYSPATH settings for buildworld
|
2014-06-24 16:10:48 +00:00
|
|
|
as well... In cases where MAKESYSPATH isn't working with this
|
|
|
|
setting, you'll need to set it to the full path to your tree.
|
2014-05-06 04:22:29 +00:00
|
|
|
|
2014-05-07 18:15:02 +00:00
|
|
|
One side effect of all this cleaning up is that bsd.compiler.mk
|
|
|
|
is no longer implicitly included by bsd.own.mk. If you wish to
|
|
|
|
use COMPILER_TYPE, you must now explicitly include bsd.compiler.mk
|
|
|
|
as well.
|
|
|
|
|
2014-05-02 05:30:49 +00:00
|
|
|
20140430:
|
|
|
|
The lindev device has been removed since /dev/full has been made a
|
|
|
|
standard device. __FreeBSD_version has been bumped.
|
|
|
|
|
2015-02-11 07:49:00 +00:00
|
|
|
20140424:
|
|
|
|
The knob WITHOUT_VI was added to the base system, which controls
|
|
|
|
building ex(1), vi(1), etc. Older releases of FreeBSD required ex(1)
|
|
|
|
in order to reorder files share/termcap and didn't build ex(1) as a
|
|
|
|
build tool, so building/installing with WITH_VI is highly advised for
|
|
|
|
build hosts for older releases.
|
|
|
|
|
|
|
|
This issue has been fixed in stable/9 and stable/10 in r277022 and
|
|
|
|
r276991, respectively.
|
|
|
|
|
2014-04-18 17:06:18 +00:00
|
|
|
20140418:
|
|
|
|
The YES_HESIOD knob has been removed. It has been obsolete for
|
|
|
|
a decade. Please move to using WITH_HESIOD instead or your builds
|
|
|
|
will silently lack HESIOD.
|
|
|
|
|
2014-04-05 22:25:10 +00:00
|
|
|
20140405:
|
|
|
|
The uart(4) driver has been changed with respect to its handling
|
|
|
|
of the low-level console. Previously the uart(4) driver prevented
|
|
|
|
any process from changing the baudrate or the CLOCAL and HUPCL
|
|
|
|
control flags. By removing the restrictions, operators can make
|
|
|
|
changes to the serial console port without having to reboot.
|
|
|
|
However, when getty(8) is started on the serial device that is
|
|
|
|
associated with the low-level console, a misconfigured terminal
|
|
|
|
line in /etc/ttys will now have a real impact.
|
|
|
|
Before upgrading the kernel, make sure that /etc/ttys has the
|
|
|
|
serial console device configured as 3wire without baudrate to
|
2014-04-05 22:28:46 +00:00
|
|
|
preserve the previous behaviour. E.g:
|
2014-04-05 22:25:10 +00:00
|
|
|
ttyu0 "/usr/libexec/getty 3wire" vt100 on secure
|
|
|
|
|
2014-03-06 17:33:27 +00:00
|
|
|
20140306:
|
|
|
|
Support for libwrap (TCP wrappers) in rpcbind was disabled by default
|
|
|
|
to improve performance. To re-enable it, if needed, run rpcbind
|
|
|
|
with command line option -W.
|
|
|
|
|
2014-03-03 16:34:42 +00:00
|
|
|
20140226:
|
|
|
|
Switched back to the GPL dtc compiler due to updates in the upstream
|
|
|
|
dts files not being supported by the BSDL dtc compiler. You will need
|
|
|
|
to rebuild your kernel toolchain to pick up the new compiler. Core dumps
|
|
|
|
may result while building dtb files during a kernel build if you fail
|
|
|
|
to do so. Set WITHOUT_GPL_DTC if you require the BSDL compiler.
|
|
|
|
|
2014-02-16 19:44:07 +00:00
|
|
|
20140216:
|
2014-03-25 22:08:31 +00:00
|
|
|
Clang and llvm have been upgraded to 3.4 release.
|
|
|
|
|
2014-02-16 19:33:34 +00:00
|
|
|
20140216:
|
|
|
|
The nve(4) driver has been removed. Please use the nfe(4) driver
|
|
|
|
for NVIDIA nForce MCP Ethernet adapters instead.
|
|
|
|
|
2014-02-12 18:16:56 +00:00
|
|
|
20140212:
|
|
|
|
An ABI incompatibility crept into the libc++ 3.4 import in r261283.
|
|
|
|
This could cause certain C++ applications using shared libraries built
|
|
|
|
against the previous version of libc++ to crash. The incompatibility
|
|
|
|
has now been fixed, but any C++ applications or shared libraries built
|
|
|
|
between r261283 and r261801 should be recompiled.
|
|
|
|
|
2014-03-03 23:26:48 +00:00
|
|
|
20140204:
|
|
|
|
OpenSSH will now ignore errors caused by kernel lacking of Capsicum
|
|
|
|
capability mode support. Please note that enabling the feature in
|
|
|
|
kernel is still highly recommended.
|
|
|
|
|
|
|
|
20140131:
|
|
|
|
OpenSSH is now built with sandbox support, and will use sandbox as
|
|
|
|
the default privilege separation method. This requires Capsicum
|
|
|
|
capability mode support in kernel.
|
|
|
|
|
2014-01-28 19:07:28 +00:00
|
|
|
20140128:
|
2014-01-25 12:07:44 +00:00
|
|
|
The libelf and libdwarf libraries have been updated to newer
|
|
|
|
versions from upstream. Shared library version numbers for
|
|
|
|
these two libraries were bumped. Any ports or binaries
|
|
|
|
requiring these two libraries should be recompiled.
|
|
|
|
__FreeBSD_version is bumped to 1100006.
|
|
|
|
|
2014-01-10 10:36:14 +00:00
|
|
|
20140110:
|
|
|
|
If a Makefile in a tests/ directory was auto-generating a Kyuafile
|
|
|
|
instead of providing an explicit one, this would prevent such
|
|
|
|
Makefile from providing its own Kyuafile in the future during
|
|
|
|
NO_CLEAN builds. This has been fixed in the Makefiles but manual
|
|
|
|
intervention is needed to clean an objdir if you use NO_CLEAN:
|
|
|
|
# find /usr/obj -name Kyuafile | xargs rm -f
|
|
|
|
|
2013-12-15 19:18:18 +00:00
|
|
|
20131213:
|
|
|
|
The behavior of gss_pseudo_random() for the krb5 mechanism
|
|
|
|
has changed, for applications requesting a longer random string
|
|
|
|
than produced by the underlying enctype's pseudo-random() function.
|
|
|
|
In particular, the random string produced from a session key of
|
|
|
|
enctype aes256-cts-hmac-sha1-96 or aes256-cts-hmac-sha1-96 will
|
|
|
|
be different at the 17th octet and later, after this change.
|
|
|
|
The counter used in the PRF+ construction is now encoded as a
|
|
|
|
big-endian integer in accordance with RFC 4402.
|
|
|
|
__FreeBSD_version is bumped to 1100004.
|
|
|
|
|
2013-11-08 14:24:47 +00:00
|
|
|
20131108:
|
|
|
|
The WITHOUT_ATF build knob has been removed and its functionality
|
|
|
|
has been subsumed into the more generic WITHOUT_TESTS. If you were
|
|
|
|
using the former to disable the build of the ATF libraries, you
|
|
|
|
should change your settings to use the latter.
|
|
|
|
|
2013-10-25 22:45:18 +00:00
|
|
|
20131025:
|
|
|
|
The default version of mtree is nmtree which is obtained from
|
|
|
|
NetBSD. The output is generally the same, but may vary
|
|
|
|
slightly. If you found you need identical output adding
|
|
|
|
"-F freebsd9" to the command line should do the trick. For the
|
|
|
|
time being, the old mtree is available as fmtree.
|
|
|
|
|
2013-10-14 18:31:15 +00:00
|
|
|
20131014:
|
|
|
|
libbsdyml has been renamed to libyaml and moved to /usr/lib/private.
|
|
|
|
This will break ports-mgmt/pkg. Rebuild the port, or upgrade to pkg
|
|
|
|
1.1.4_8 and verify bsdyml not linked in, before running "make
|
|
|
|
delete-old-libs":
|
|
|
|
# make -C /usr/ports/ports-mgmt/pkg build deinstall install clean
|
|
|
|
or
|
|
|
|
# pkg install pkg; ldd /usr/local/sbin/pkg | grep bsdyml
|
|
|
|
|
2015-07-29 21:15:50 +00:00
|
|
|
20131010:
|
|
|
|
The stable/10 branch has been created in subversion from head
|
|
|
|
revision r256279.
|
|
|
|
|
2000-01-30 23:15:21 +00:00
|
|
|
COMMON ITEMS:
|
|
|
|
|
2000-09-04 03:18:01 +00:00
|
|
|
General Notes
|
|
|
|
-------------
|
2009-09-03 17:04:42 +00:00
|
|
|
Avoid using make -j when upgrading. While generally safe, there are
|
|
|
|
sometimes problems using -j to upgrade. If your upgrade fails with
|
2009-12-25 21:14:34 +00:00
|
|
|
-j, please try again without -j. From time to time in the past there
|
2009-09-03 17:04:42 +00:00
|
|
|
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).
|
2000-09-04 03:18:01 +00:00
|
|
|
|
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
|
2009-09-03 17:04:42 +00:00
|
|
|
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.
|
|
|
|
|
2016-10-19 13:45:51 +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 if you encounter problems with a major version upgrade. Since
|
2018-01-12 17:33:20 +00:00
|
|
|
the stable 4.x branch point, one has generally been able to upgrade from
|
2016-10-19 13:45:51 +00:00
|
|
|
anywhere in the most recent stable branch to head / current (or even the
|
|
|
|
last couple of stable branches). See the top of this file when there's
|
|
|
|
an exception.
|
2005-12-07 20:49:42 +00:00
|
|
|
|
2013-04-21 17:23:27 +00:00
|
|
|
When upgrading a live system, having a root shell around before
|
2013-04-20 23:33:51 +00:00
|
|
|
installing anything can help undo problems. Not having a root shell
|
|
|
|
around can lead to problems if pam has changed too much from your
|
|
|
|
starting point to allow continued authentication after the upgrade.
|
|
|
|
|
2015-07-29 21:15:50 +00:00
|
|
|
This file should be read as a log of events. When a later event changes
|
|
|
|
information of a prior event, the prior event should not be deleted.
|
|
|
|
Instead, a pointer to the entry with the new information should be
|
|
|
|
placed in the old entry. Readers of this file should also sanity check
|
|
|
|
older entries before relying on them blindly. Authors of new entries
|
|
|
|
should write them with this in mind.
|
|
|
|
|
Merge ZFS version 15 and almost all OpenSolaris bugfixes referenced
in Solaris 10 updates 141445-09 and 142901-14.
Detailed information:
(OpenSolaris revisions and Bug IDs, Solaris 10 patch numbers)
7844:effed23820ae
6755435 zfs_open() and zfs_close() needs to use ZFS_ENTER/ZFS_VERIFY_ZP (141445-01)
7897:e520d8258820
6748436 inconsistent zpool.cache in boot_archive could panic a zfs root filesystem upon boot-up (141445-01)
7965:b795da521357
6740164 zpool attach can create an illegal root pool (141909-02)
8084:b811cc60d650
6769612 zpool_import() will continue to write to cachefile even if altroot is set (N/A)
8121:7fd09d4ebd9c
6757430 want an option for zdb to disable space map loading and leak tracking (141445-01)
8129:e4f45a0bfbb0
6542860 ASSERT: reason != VDEV_LABEL_REMOVE||vdev_inuse(vd, crtxg, reason, 0) (141445-01)
8188:fd00c0a81e80
6761100 want zdb option to select older uberblocks (141445-01)
8190:6eeea43ced42
6774886 zfs_setattr() won't allow ndmp to restore SUNWattr_rw (141445-01)
8225:59a9961c2aeb
6737463 panic while trying to write out config file if root pool import fails (141445-01)
8227:f7d7be9b1f56
6765294 Refactor replay (141445-01)
8228:51e9ca9ee3a5
6572357 libzfs should do more to avoid mnttab lookups (141909-01)
6572376 zfs_iter_filesystems and zfs_iter_snapshots get objset stats twice (141909-01)
8241:5a60f16123ba
6328632 zpool offline is a bit too conservative (141445-01)
6739487 ASSERT: txg <= spa_final_txg due to scrub/export race (141445-01)
6767129 ASSERT: cvd->vdev_isspare, in spa_vdev_detach() (141445-01)
6747698 checksum failures after offline -t / export / import / scrub (141445-01)
6745863 ZFS writes to disk after it has been offlined (141445-01)
6722540 50% slowdown on scrub/resilver with certain vdev configurations (141445-01)
6759999 resilver logic rewrites ditto blocks on both source and destination (141445-01)
6758107 I/O should never suspend during spa_load() (141445-01)
6776548 codereview(1) runs off the page when faced with multi-line comments (N/A)
6761406 AMD errata 91 workaround doesn't work on 64-bit systems (141445-01)
8242:e46e4b2f0a03
6770866 GRUB/ZFS should require physical path or devid, but not both (141445-01)
8269:03a7e9050cfd
6674216 "zfs share" doesn't work, but "zfs set sharenfs=on" does (141445-01)
6621164 $SRC/cmd/zfs/zfs_main.c seems to have a syntax error in the translation note (141445-01)
6635482 i18n problems in libzfs_dataset.c and zfs_main.c (141445-01)
6595194 "zfs get" VALUE column is as wide as NAME (141445-01)
6722991 vdev_disk.c: error checking for ddi_pathname_to_dev_t() must test for NODEV (141445-01)
6396518 ASSERT strings shouldn't be pre-processed (141445-01)
8274:846b39508aff
6713916 scrub/resilver needlessly decompress data (141445-01)
8343:655db2375fed
6739553 libzfs_status msgid table is out of sync (141445-01)
6784104 libzfs unfairly rejects numerical values greater than 2^63 (141445-01)
6784108 zfs_realloc() should not free original memory on failure (141445-01)
8525:e0e0e525d0f8
6788830 set large value to reservation cause core dump (141445-01)
6791064 want sysevents for ZFS scrub (141445-01)
6791066 need to be able to set cachefile on faulted pools (141445-01)
6791071 zpool_do_import() should not enable datasets on faulted pools (141445-01)
6792134 getting multiple properties on a faulted pool leads to confusion (141445-01)
8547:bcc7b46e5ff7
6792884 Vista clients cannot access .zfs (141445-01)
8632:36ef517870a3
6798384 It can take a village to raise a zio (141445-01)
8636:7e4ce9158df3
6551866 deadlock between zfs_write(), zfs_freesp(), and zfs_putapage() (141909-01)
6504953 zfs_getpage() misunderstands VOP_GETPAGE() interface (141909-01)
6702206 ZFS read/writer lock contention throttles sendfile() benchmark (141445-01)
6780491 Zone on a ZFS filesystem has poor fork/exec performance (141445-01)
6747596 assertion failed: DVA_EQUAL(BP_IDENTITY(&zio->io_bp_orig), BP_IDENTITY(zio->io_bp))); (141445-01)
8692:692d4668b40d
6801507 ZFS read aggregation should not mind the gap (141445-01)
8697:e62d2612c14d
6633095 creating a filesystem with many properties set is slow (141445-01)
8768:dfecfdbb27ed
6775697 oracle crashes when overwriting after hitting quota on zfs (141909-01)
8811:f8deccf701cf
6790687 libzfs mnttab caching ignores external changes (141445-01)
6791101 memory leak from libzfs_mnttab_init (141445-01)
8845:91af0d9c0790
6800942 smb_session_create() incorrectly stores IP addresses (N/A)
6582163 Access Control List (ACL) for shares (141445-01)
6804954 smb_search - shortname field should be space padded following the NULL terminator (N/A)
6800184 Panic at smb_oplock_conflict+0x35() (N/A)
8876:59d2e67b4b65
6803822 Reboot after replacement of system disk in a ZFS mirror drops to grub> prompt (141445-01)
8924:5af812f84759
6789318 coredump when issue zdb -uuuu poolname/ (141445-01)
6790345 zdb -dddd -e poolname coredump (141445-01)
6797109 zdb: 'zdb -dddddd pool_name/fs_name inode' coredump if the file with inode was deleted (141445-01)
6797118 zdb: 'zdb -dddddd poolname inum' coredump if I miss the fs name (141445-01)
6803343 shareiscsi=on failed, iscsitgtd failed request to share (141445-01)
9030:243fd360d81f
6815893 hang mounting a dataset after booting into a new boot environment (141445-01)
9056:826e1858a846
6809691 'zpool create -f' no longer overwrites ufs infomation (141445-01)
9179:d8fbd96b79b3
6790064 zfs needs to determine uid and gid earlier in create process (141445-01)
9214:8d350e5d04aa
6604992 forced unmount + being in .zfs/snapshot/<snap1> = not happy (141909-01)
6810367 assertion failed: dvp->v_flag & VROOT, file: ../../common/fs/gfs.c, line: 426 (141909-01)
9229:e3f8b41e5db4
6807765 ztest_dsl_dataset_promote_busy needs to clean up after ENOSPC (141445-01)
9230:e4561e3eb1ef
6821169 offlining a device results in checksum errors (141445-01)
6821170 ZFS should not increment error stats for unavailable devices (141445-01)
6824006 need to increase issue and interrupt taskqs threads in zfs (141445-01)
9234:bffdc4fc05c4
6792139 recovering from a suspended pool needs some work (141445-01)
6794830 reboot command hangs on a failed zfs pool (141445-01)
9246:67c03c93c071
6824062 System panicked in zfs_mount due to NULL pointer dereference when running btts and svvs tests (141909-01)
9276:a8a7fc849933
6816124 System crash running zpool destroy on broken zpool (141445-03)
9355:09928982c591
6818183 zfs snapshot -r is slow due to set_snap_props() doing txg_wait_synced() for each new snapshot (141445-03)
9391:413d0661ef33
6710376 log device can show incorrect status when other parts of pool are degraded (141445-03)
9396:f41cf682d0d3 (part already merged)
6501037 want user/group quotas on ZFS (141445-03)
6827260 assertion failed in arc_read(): hdr == pbuf->b_hdr (141445-03)
6815592 panic: No such hold X on refcount Y from zfs_znode_move (141445-03)
6759986 zfs list shows temporary %clone when doing online zfs recv (141445-03)
9404:319573cd93f8
6774713 zfs ignores canmount=noauto when sharenfs property != off (141445-03)
9412:4aefd8704ce0
6717022 ZFS DMU needs zero-copy support (141445-03)
9425:e7ffacaec3a8
6799895 spa_add_spares() needs to be protected by config lock (141445-03)
6826466 want to post sysevents on hot spare activation (141445-03)
6826468 spa 'allowfaulted' needs some work (141445-03)
6826469 kernel support for storing vdev FRU information (141445-03)
6826470 skip posting checksum errors from DTL regions of leaf vdevs (141445-03)
6826471 I/O errors after device remove probe can confuse FMA (141445-03)
6826472 spares should enjoy some of the benefits of cache devices (141445-03)
9443:2a96d8478e95
6833711 gang leaders shouldn't have to be logical (141445-03)
9463:d0bd231c7518
6764124 want zdb to be able to checksum metadata blocks only (141445-03)
9465:8372081b8019
6830237 zfs panic in zfs_groupmember() (141445-03)
9466:1fdfd1fed9c4
6833162 phantom log device in zpool status (141445-03)
9469:4f68f041ddcd
6824968 add ZFS userquota support to rquotad (141445-03)
9470:6d827468d7b5
6834217 godfather I/O should reexecute (141445-03)
9480:fcff33da767f
6596237 Stop looking and start ganging (141909-02)
9493:9933d599bc93
6623978 lwb->lwb_buf != NULL, file ../../../uts/common/fs/zfs/zil.c, line 787, function zil_lwb_commit (141445-06)
9512:64cafcbcc337
6801810 Commit of aligned streaming rewrites to ZIL device causes unwanted disk reads (N/A)
9515:d3b739d9d043
6586537 async zio taskqs can block out userland commands (142901-09)
9554:787363635b6a
6836768 zfs_userspace() callback has no way to indicate failure (N/A)
9574:1eb6a6ab2c57
6838062 zfs panics when an error is encountered in space_map_load() (141909-02)
9583:b0696cd037cc
6794136 Panic BAD TRAP: type=e when importing degraded zraid pool. (141909-03)
9630:e25a03f552e0
6776104 "zfs import" deadlock between spa_unload() and spa_async_thread() (141445-06)
9653:a70048a304d1
6664765 Unable to remove files when using fat-zap and quota exceeded on ZFS filesystem (141445-06)
9688:127be1845343
6841321 zfs userspace / zfs get userused@ doesn't work on mounted snapshot (N/A)
6843069 zfs get userused@S-1-... doesn't work (N/A)
9873:8ddc892eca6e
6847229 assertion failed: refcount_count(&tx->tx_space_written) + delta <= tx->tx_space_towrite in dmu_tx.c (141445-06)
9904:d260bd3fd47c
6838344 kernel heap corruption detected on zil while stress testing (141445-06)
9951:a4895b3dd543
6844900 zfs_ioc_userspace_upgrade leaks (N/A)
10040:38b25aeeaf7a
6857012 zfs panics on zpool import (141445-06)
10000:241a51d8720c
6848242 zdb -e no longer works as expected (N/A)
10100:4a6965f6bef8
6856634 snv_117 not booting: zfs_parse_bootfs: error2 (141445-07)
10160:a45b03783d44
6861983 zfs should use new name <-> SID interfaces (N/A)
6862984 userquota commands can hang (141445-06)
10299:80845694147f
6696858 zfs receive of incremental replication stream can dereference NULL pointer and crash (N/A)
10302:a9e3d1987706
6696858 zfs receive of incremental replication stream can dereference NULL pointer and crash (fix lint) (N/A)
10575:2a8816c5173b (partial merge)
6882227 spa_async_remove() shouldn't do a full clear (142901-14)
10800:469478b180d9
6880764 fsync on zfs is broken if writes are greater than 32kb on a hard crash and no log attached (142901-09)
6793430 zdb -ivvvv assertion failure: bp->blk_cksum.zc_word[2] == dmu_objset_id(zilog->zl_os) (N/A)
10801:e0bf032e8673 (partial merge)
6822816 assertion failed: zap_remove_int(ds_next_clones_obj) returns ENOENT (142901-09)
10810:b6b161a6ae4a
6892298 buf->b_hdr->b_state != arc_anon, file: ../../common/fs/zfs/arc.c, line: 2849 (142901-09)
10890:499786962772
6807339 spurious checksum errors when replacing a vdev (142901-13)
11249:6c30f7dfc97b
6906110 bad trap panic in zil_replay_log_record (142901-13)
6906946 zfs replay isn't handling uid/gid correctly (142901-13)
11454:6e69bacc1a5a
6898245 suspended zpool should not cause rest of the zfs/zpool commands to hang (142901-10)
11546:42ea6be8961b (partial merge)
6833999 3-way deadlock in dsl_dataset_hold_ref() and dsl_sync_task_group_sync() (142901-09)
Discussed with: pjd
Approved by: delphij (mentor)
Obtained from: OpenSolaris (multiple Bug IDs)
MFC after: 2 months
2010-07-12 23:49:04 +00:00
|
|
|
ZFS notes
|
|
|
|
---------
|
|
|
|
When upgrading the boot ZFS pool to a new version, always follow
|
|
|
|
these two steps:
|
|
|
|
|
|
|
|
1.) recompile and reinstall the ZFS boot loader and boot block
|
|
|
|
(this is part of "make buildworld" and "make installworld")
|
|
|
|
|
|
|
|
2.) update the ZFS boot block on your boot drive
|
|
|
|
|
|
|
|
The following example updates the ZFS boot block on the first
|
2014-11-09 20:43:50 +00:00
|
|
|
partition (freebsd-boot) of a GPT partitioned drive ada0:
|
|
|
|
"gpart bootcode -p /boot/gptzfsboot -i 1 ada0"
|
Merge ZFS version 15 and almost all OpenSolaris bugfixes referenced
in Solaris 10 updates 141445-09 and 142901-14.
Detailed information:
(OpenSolaris revisions and Bug IDs, Solaris 10 patch numbers)
7844:effed23820ae
6755435 zfs_open() and zfs_close() needs to use ZFS_ENTER/ZFS_VERIFY_ZP (141445-01)
7897:e520d8258820
6748436 inconsistent zpool.cache in boot_archive could panic a zfs root filesystem upon boot-up (141445-01)
7965:b795da521357
6740164 zpool attach can create an illegal root pool (141909-02)
8084:b811cc60d650
6769612 zpool_import() will continue to write to cachefile even if altroot is set (N/A)
8121:7fd09d4ebd9c
6757430 want an option for zdb to disable space map loading and leak tracking (141445-01)
8129:e4f45a0bfbb0
6542860 ASSERT: reason != VDEV_LABEL_REMOVE||vdev_inuse(vd, crtxg, reason, 0) (141445-01)
8188:fd00c0a81e80
6761100 want zdb option to select older uberblocks (141445-01)
8190:6eeea43ced42
6774886 zfs_setattr() won't allow ndmp to restore SUNWattr_rw (141445-01)
8225:59a9961c2aeb
6737463 panic while trying to write out config file if root pool import fails (141445-01)
8227:f7d7be9b1f56
6765294 Refactor replay (141445-01)
8228:51e9ca9ee3a5
6572357 libzfs should do more to avoid mnttab lookups (141909-01)
6572376 zfs_iter_filesystems and zfs_iter_snapshots get objset stats twice (141909-01)
8241:5a60f16123ba
6328632 zpool offline is a bit too conservative (141445-01)
6739487 ASSERT: txg <= spa_final_txg due to scrub/export race (141445-01)
6767129 ASSERT: cvd->vdev_isspare, in spa_vdev_detach() (141445-01)
6747698 checksum failures after offline -t / export / import / scrub (141445-01)
6745863 ZFS writes to disk after it has been offlined (141445-01)
6722540 50% slowdown on scrub/resilver with certain vdev configurations (141445-01)
6759999 resilver logic rewrites ditto blocks on both source and destination (141445-01)
6758107 I/O should never suspend during spa_load() (141445-01)
6776548 codereview(1) runs off the page when faced with multi-line comments (N/A)
6761406 AMD errata 91 workaround doesn't work on 64-bit systems (141445-01)
8242:e46e4b2f0a03
6770866 GRUB/ZFS should require physical path or devid, but not both (141445-01)
8269:03a7e9050cfd
6674216 "zfs share" doesn't work, but "zfs set sharenfs=on" does (141445-01)
6621164 $SRC/cmd/zfs/zfs_main.c seems to have a syntax error in the translation note (141445-01)
6635482 i18n problems in libzfs_dataset.c and zfs_main.c (141445-01)
6595194 "zfs get" VALUE column is as wide as NAME (141445-01)
6722991 vdev_disk.c: error checking for ddi_pathname_to_dev_t() must test for NODEV (141445-01)
6396518 ASSERT strings shouldn't be pre-processed (141445-01)
8274:846b39508aff
6713916 scrub/resilver needlessly decompress data (141445-01)
8343:655db2375fed
6739553 libzfs_status msgid table is out of sync (141445-01)
6784104 libzfs unfairly rejects numerical values greater than 2^63 (141445-01)
6784108 zfs_realloc() should not free original memory on failure (141445-01)
8525:e0e0e525d0f8
6788830 set large value to reservation cause core dump (141445-01)
6791064 want sysevents for ZFS scrub (141445-01)
6791066 need to be able to set cachefile on faulted pools (141445-01)
6791071 zpool_do_import() should not enable datasets on faulted pools (141445-01)
6792134 getting multiple properties on a faulted pool leads to confusion (141445-01)
8547:bcc7b46e5ff7
6792884 Vista clients cannot access .zfs (141445-01)
8632:36ef517870a3
6798384 It can take a village to raise a zio (141445-01)
8636:7e4ce9158df3
6551866 deadlock between zfs_write(), zfs_freesp(), and zfs_putapage() (141909-01)
6504953 zfs_getpage() misunderstands VOP_GETPAGE() interface (141909-01)
6702206 ZFS read/writer lock contention throttles sendfile() benchmark (141445-01)
6780491 Zone on a ZFS filesystem has poor fork/exec performance (141445-01)
6747596 assertion failed: DVA_EQUAL(BP_IDENTITY(&zio->io_bp_orig), BP_IDENTITY(zio->io_bp))); (141445-01)
8692:692d4668b40d
6801507 ZFS read aggregation should not mind the gap (141445-01)
8697:e62d2612c14d
6633095 creating a filesystem with many properties set is slow (141445-01)
8768:dfecfdbb27ed
6775697 oracle crashes when overwriting after hitting quota on zfs (141909-01)
8811:f8deccf701cf
6790687 libzfs mnttab caching ignores external changes (141445-01)
6791101 memory leak from libzfs_mnttab_init (141445-01)
8845:91af0d9c0790
6800942 smb_session_create() incorrectly stores IP addresses (N/A)
6582163 Access Control List (ACL) for shares (141445-01)
6804954 smb_search - shortname field should be space padded following the NULL terminator (N/A)
6800184 Panic at smb_oplock_conflict+0x35() (N/A)
8876:59d2e67b4b65
6803822 Reboot after replacement of system disk in a ZFS mirror drops to grub> prompt (141445-01)
8924:5af812f84759
6789318 coredump when issue zdb -uuuu poolname/ (141445-01)
6790345 zdb -dddd -e poolname coredump (141445-01)
6797109 zdb: 'zdb -dddddd pool_name/fs_name inode' coredump if the file with inode was deleted (141445-01)
6797118 zdb: 'zdb -dddddd poolname inum' coredump if I miss the fs name (141445-01)
6803343 shareiscsi=on failed, iscsitgtd failed request to share (141445-01)
9030:243fd360d81f
6815893 hang mounting a dataset after booting into a new boot environment (141445-01)
9056:826e1858a846
6809691 'zpool create -f' no longer overwrites ufs infomation (141445-01)
9179:d8fbd96b79b3
6790064 zfs needs to determine uid and gid earlier in create process (141445-01)
9214:8d350e5d04aa
6604992 forced unmount + being in .zfs/snapshot/<snap1> = not happy (141909-01)
6810367 assertion failed: dvp->v_flag & VROOT, file: ../../common/fs/gfs.c, line: 426 (141909-01)
9229:e3f8b41e5db4
6807765 ztest_dsl_dataset_promote_busy needs to clean up after ENOSPC (141445-01)
9230:e4561e3eb1ef
6821169 offlining a device results in checksum errors (141445-01)
6821170 ZFS should not increment error stats for unavailable devices (141445-01)
6824006 need to increase issue and interrupt taskqs threads in zfs (141445-01)
9234:bffdc4fc05c4
6792139 recovering from a suspended pool needs some work (141445-01)
6794830 reboot command hangs on a failed zfs pool (141445-01)
9246:67c03c93c071
6824062 System panicked in zfs_mount due to NULL pointer dereference when running btts and svvs tests (141909-01)
9276:a8a7fc849933
6816124 System crash running zpool destroy on broken zpool (141445-03)
9355:09928982c591
6818183 zfs snapshot -r is slow due to set_snap_props() doing txg_wait_synced() for each new snapshot (141445-03)
9391:413d0661ef33
6710376 log device can show incorrect status when other parts of pool are degraded (141445-03)
9396:f41cf682d0d3 (part already merged)
6501037 want user/group quotas on ZFS (141445-03)
6827260 assertion failed in arc_read(): hdr == pbuf->b_hdr (141445-03)
6815592 panic: No such hold X on refcount Y from zfs_znode_move (141445-03)
6759986 zfs list shows temporary %clone when doing online zfs recv (141445-03)
9404:319573cd93f8
6774713 zfs ignores canmount=noauto when sharenfs property != off (141445-03)
9412:4aefd8704ce0
6717022 ZFS DMU needs zero-copy support (141445-03)
9425:e7ffacaec3a8
6799895 spa_add_spares() needs to be protected by config lock (141445-03)
6826466 want to post sysevents on hot spare activation (141445-03)
6826468 spa 'allowfaulted' needs some work (141445-03)
6826469 kernel support for storing vdev FRU information (141445-03)
6826470 skip posting checksum errors from DTL regions of leaf vdevs (141445-03)
6826471 I/O errors after device remove probe can confuse FMA (141445-03)
6826472 spares should enjoy some of the benefits of cache devices (141445-03)
9443:2a96d8478e95
6833711 gang leaders shouldn't have to be logical (141445-03)
9463:d0bd231c7518
6764124 want zdb to be able to checksum metadata blocks only (141445-03)
9465:8372081b8019
6830237 zfs panic in zfs_groupmember() (141445-03)
9466:1fdfd1fed9c4
6833162 phantom log device in zpool status (141445-03)
9469:4f68f041ddcd
6824968 add ZFS userquota support to rquotad (141445-03)
9470:6d827468d7b5
6834217 godfather I/O should reexecute (141445-03)
9480:fcff33da767f
6596237 Stop looking and start ganging (141909-02)
9493:9933d599bc93
6623978 lwb->lwb_buf != NULL, file ../../../uts/common/fs/zfs/zil.c, line 787, function zil_lwb_commit (141445-06)
9512:64cafcbcc337
6801810 Commit of aligned streaming rewrites to ZIL device causes unwanted disk reads (N/A)
9515:d3b739d9d043
6586537 async zio taskqs can block out userland commands (142901-09)
9554:787363635b6a
6836768 zfs_userspace() callback has no way to indicate failure (N/A)
9574:1eb6a6ab2c57
6838062 zfs panics when an error is encountered in space_map_load() (141909-02)
9583:b0696cd037cc
6794136 Panic BAD TRAP: type=e when importing degraded zraid pool. (141909-03)
9630:e25a03f552e0
6776104 "zfs import" deadlock between spa_unload() and spa_async_thread() (141445-06)
9653:a70048a304d1
6664765 Unable to remove files when using fat-zap and quota exceeded on ZFS filesystem (141445-06)
9688:127be1845343
6841321 zfs userspace / zfs get userused@ doesn't work on mounted snapshot (N/A)
6843069 zfs get userused@S-1-... doesn't work (N/A)
9873:8ddc892eca6e
6847229 assertion failed: refcount_count(&tx->tx_space_written) + delta <= tx->tx_space_towrite in dmu_tx.c (141445-06)
9904:d260bd3fd47c
6838344 kernel heap corruption detected on zil while stress testing (141445-06)
9951:a4895b3dd543
6844900 zfs_ioc_userspace_upgrade leaks (N/A)
10040:38b25aeeaf7a
6857012 zfs panics on zpool import (141445-06)
10000:241a51d8720c
6848242 zdb -e no longer works as expected (N/A)
10100:4a6965f6bef8
6856634 snv_117 not booting: zfs_parse_bootfs: error2 (141445-07)
10160:a45b03783d44
6861983 zfs should use new name <-> SID interfaces (N/A)
6862984 userquota commands can hang (141445-06)
10299:80845694147f
6696858 zfs receive of incremental replication stream can dereference NULL pointer and crash (N/A)
10302:a9e3d1987706
6696858 zfs receive of incremental replication stream can dereference NULL pointer and crash (fix lint) (N/A)
10575:2a8816c5173b (partial merge)
6882227 spa_async_remove() shouldn't do a full clear (142901-14)
10800:469478b180d9
6880764 fsync on zfs is broken if writes are greater than 32kb on a hard crash and no log attached (142901-09)
6793430 zdb -ivvvv assertion failure: bp->blk_cksum.zc_word[2] == dmu_objset_id(zilog->zl_os) (N/A)
10801:e0bf032e8673 (partial merge)
6822816 assertion failed: zap_remove_int(ds_next_clones_obj) returns ENOENT (142901-09)
10810:b6b161a6ae4a
6892298 buf->b_hdr->b_state != arc_anon, file: ../../common/fs/zfs/arc.c, line: 2849 (142901-09)
10890:499786962772
6807339 spurious checksum errors when replacing a vdev (142901-13)
11249:6c30f7dfc97b
6906110 bad trap panic in zil_replay_log_record (142901-13)
6906946 zfs replay isn't handling uid/gid correctly (142901-13)
11454:6e69bacc1a5a
6898245 suspended zpool should not cause rest of the zfs/zpool commands to hang (142901-10)
11546:42ea6be8961b (partial merge)
6833999 3-way deadlock in dsl_dataset_hold_ref() and dsl_sync_task_group_sync() (142901-09)
Discussed with: pjd
Approved by: delphij (mentor)
Obtained from: OpenSolaris (multiple Bug IDs)
MFC after: 2 months
2010-07-12 23:49:04 +00:00
|
|
|
|
|
|
|
Non-boot pools do not need these updates.
|
|
|
|
|
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 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
|
2017-10-29 08:25:31 +00:00
|
|
|
make buildkernel KERNCONF=YOUR_KERNEL_HERE
|
|
|
|
make installkernel KERNCONF=YOUR_KERNEL_HERE
|
2002-10-26 06:25:11 +00:00
|
|
|
[1]
|
|
|
|
<reboot in single user> [3]
|
2014-02-28 19:51:47 +00:00
|
|
|
mergemaster -Fp [5]
|
2002-10-26 06:25:11 +00:00
|
|
|
make installworld
|
2014-02-28 19:51:47 +00:00
|
|
|
mergemaster -Fi [4]
|
2010-12-20 15:17:34 +00:00
|
|
|
make delete-old [6]
|
2002-10-26 06:25:11 +00:00
|
|
|
<reboot>
|
|
|
|
|
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 ${CURRENT_ROOT}>
|
2013-12-20 23:34:10 +00:00
|
|
|
make installworld DESTDIR=${CURRENT_ROOT} -DDB_FROM_SRC
|
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
|
|
|
|
|
|
|
|
2012-10-31 13:52:03 +00:00
|
|
|
To upgrade in-place from 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]
|
2018-03-28 16:51:05 +00:00
|
|
|
make buildkernel KERNCONF=YOUR_KERNEL_HERE [8]
|
|
|
|
make installkernel KERNCONF=YOUR_KERNEL_HERE
|
2002-10-26 06:21:00 +00:00
|
|
|
[1]
|
|
|
|
<reboot in single user> [3]
|
2014-02-28 19:51:47 +00:00
|
|
|
mergemaster -Fp [5]
|
2000-06-14 15:42:50 +00:00
|
|
|
make installworld
|
2014-02-28 19:51:47 +00:00
|
|
|
mergemaster -Fi [4]
|
2010-12-20 15:17:34 +00:00
|
|
|
make delete-old [6]
|
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
|
2019-03-09 17:17:44 +00:00
|
|
|
sh /etc/rc.d/zfs start # mount zfs filesystem, if needed
|
|
|
|
cd src # full path to source
|
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
|
2010-03-10 05:44:57 +00:00
|
|
|
for potential gotchas. The -U option is also useful to consider.
|
|
|
|
See mergemaster(8) for more information.
|
2001-08-11 02:03:51 +00:00
|
|
|
|
2018-01-12 17:33:20 +00:00
|
|
|
[5] Usually this step is a no-op. However, from time to time
|
2002-04-19 04:22:35 +00:00
|
|
|
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
|
2013-04-25 21:19:50 +00:00
|
|
|
from current before 20130425 or from -stable before 20130430.
|
2002-04-19 04:22:35 +00:00
|
|
|
|
2010-12-20 15:17:34 +00:00
|
|
|
[6] This only deletes old files and directories. Old libraries
|
|
|
|
can be deleted by "make delete-old-libs", but you have to make
|
|
|
|
sure that no program is using those libraries anymore.
|
|
|
|
|
2019-03-12 06:01:43 +00:00
|
|
|
[8] The new kernel must be able to run existing binaries used by an
|
|
|
|
installworld. When upgrading across major versions, the new kernel's
|
|
|
|
configuration must include the correct COMPAT_FREEBSD<n> option for
|
|
|
|
existing binaries (e.g. COMPAT_FREEBSD11 to run 11.x binaries). Failure
|
|
|
|
to do so may leave you with a system that is hard to boot to recover. A
|
|
|
|
GENERIC kernel will include suitable compatibility options to run
|
|
|
|
binaries from older branches. Note that the ability to run binaries
|
|
|
|
from unsupported branches is not guaranteed.
|
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
|
2019-03-09 17:17:44 +00:00
|
|
|
last time you updated your kernel config file. Options also
|
|
|
|
change over time, so you may need to adjust your custom kernels
|
|
|
|
for these as well.
|
2004-09-04 21:03:10 +00:00
|
|
|
|
2017-10-29 08:25:31 +00:00
|
|
|
[9] If CPUTYPE is defined in your /etc/make.conf, make sure to use the
|
2004-09-04 21:03:10 +00:00
|
|
|
"?=" 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
|
2011-08-24 12:18:29 +00:00
|
|
|
breakages in tracking -current. It is not guaranteed to be a complete
|
2015-07-29 21:15:50 +00:00
|
|
|
list of such breakages, and only contains entries since September 23, 2011.
|
2011-08-24 12:18:29 +00:00
|
|
|
If you need to see UPDATING entries from before that date, you will need
|
|
|
|
to fetch an UPDATING file from an older FreeBSD release.
|
1999-02-14 05:18:35 +00:00
|
|
|
|
2001-06-29 06:00:44 +00:00
|
|
|
Copyright information:
|
|
|
|
|
2019-02-05 21:28:29 +00:00
|
|
|
Copyright 1998-2009 M. Warner Losh.
|
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.
|
|
|
|
|
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
|
|
|
$FreeBSD$
|