[owc] Weekly src changes ending 2006-06-04
OWC
auto at squish.net
Mon Jun 5 08:13:30 BST 2006
OpenBSD src changes summary for 2006-05-28 to 2006-06-04 inclusive
==================================================================
bin/chio bin/dd
bin/ksh bin/mt
bin/stty distrib/notes
distrib/sets etc/Makefile
etc/changelist etc/chio.conf
etc/dvmrpd.conf etc/ftpusers
etc/group etc/mail/aliases
etc/master.passwd etc/mtree/special
etc/pf.os etc/rc
etc/rc.conf etc/services
gnu/usr.bin/binutils lib/libc
lib/libpthread regress/sbin
regress/usr.bin sbin/brconfig
sbin/dhclient sbin/dump
sbin/fdisk sbin/fsck
sbin/ifconfig sbin/ipsecctl
sbin/isakmpd sbin/lmccontrol
sbin/mountd sbin/pdisk
sbin/pfctl sbin/reboot
sbin/route sbin/sysctl
share/man share/mk
sys sys/arch
sys/arch/alpha/conf sys/arch/amd64/amd64
sys/arch/amd64/conf sys/arch/amd64/pci
sys/arch/arm/arm sys/arch/arm/footbridge
sys/arch/arm/include sys/arch/arm/xscale
sys/arch/armish/armish sys/arch/armish/compile
sys/arch/armish/conf sys/arch/armish/dev
sys/arch/armish/include sys/arch/armish/stand
sys/arch/cats/conf sys/arch/hp300/include
sys/arch/hppa/conf sys/arch/hppa/dev
sys/arch/hppa/gsc sys/arch/hppa/hppa
sys/arch/hppa/include sys/arch/hppa64/conf
sys/arch/hppa64/include sys/arch/i386/conf
sys/arch/i386/i386 sys/arch/i386/include
sys/arch/i386/isa sys/arch/i386/pci
sys/arch/m88k/include sys/arch/m88k/m88k
sys/arch/macppc/conf sys/arch/macppc/pci
sys/arch/mips64/include sys/arch/mips64/mips64
sys/arch/mvme68k/dev sys/arch/mvme68k/include
sys/arch/mvme88k/conf sys/arch/mvmeppc/conf
sys/arch/powerpc/include sys/arch/sgi/conf
sys/arch/solbourne/conf sys/arch/solbourne/include
sys/arch/sparc/conf sys/arch/sparc/dev
sys/arch/sparc/sparc sys/arch/sparc64/conf
sys/arch/sparc64/dev sys/arch/sparc64/include
sys/arch/sparc64/sparc64 sys/arch/vax/include
sys/arch/vax/vax sys/arch/zaurus/conf
sys/arch/zaurus/include sys/arch/zaurus/zaurus
sys/conf sys/crypto
sys/dev sys/dev/acpi
sys/dev/cardbus sys/dev/eisa
sys/dev/ic sys/dev/isa
sys/dev/mii sys/dev/pci
sys/dev/pckbc sys/dev/pcmcia
sys/dev/puc sys/dev/sbus
sys/dev/sdmmc sys/dev/usb
sys/dev/wscons sys/kern
sys/miscfs/specfs sys/net
sys/netinet sys/nfs
sys/ntfs sys/scsi
sys/sys sys/ufs/ufs
sys/uvm usr.bin/cdio
usr.bin/cvs usr.bin/env
usr.bin/file usr.bin/ftp
usr.bin/mg usr.bin/nc
usr.bin/netstat usr.bin/rcs
usr.bin/ssh usr.bin/stat
usr.bin/systat usr.bin/tip
usr.bin/vi usr.bin/whois
usr.bin/xlint usr.sbin
usr.sbin/bgpd usr.sbin/dhcpd
usr.sbin/dvmrpctl usr.sbin/dvmrpd
usr.sbin/hostapd usr.sbin/hotplugd
usr.sbin/ntpd usr.sbin/openssl
usr.sbin/ospfd usr.sbin/pstat
usr.sbin/sasyncd usr.sbin/tcpdump
== bin =============================================================== 01/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/bin
chio
~ Makefile ~ chio.c
+ parse.y
> Make chio have a connection between st(4) devices and "drives" in a
> changer - chio will then attempt to open the /dev/rstX device when doing
> a move from a source of drive X, and will attempt to unload the tape.
> this avoids issues where pickers madly attempt to grab a busy tape on
> completely decoupled libraries, or fail to grab an unloaded tape
> on tightly coupled libraries, the extra unload being harmless if the
> media has already been ejected.
> The mapping between st(4) devices and ch drives is by default a
> simple mapping between picker drive X being mapped to /dev/rstX, however
> for non-obvious or complicated configurations, we support a /etc/chio.conf
> file in which the drives for a changer may have their corresponding
> st(4) devices defined individually. chio will use the default
> mapping if the /etc/chio.conf file is not present, or does not define
> a st(4) device for a changer drive.
> (example chio.conf and man page changes to come)
> yacc parser for chio.conf written by henning@,
> ok henning@, krw@ (beck@)
~ parse.y
> de-beckify (KNF) (henning@)
~ chio.c
> KNF (henning@)
~ chio.c
> spacing (deraadt@)
~ parse.y
> oups, left one strdup unchecked, and some more error handling smallies
> (henning@)
~ chio.1
> document the st(4) device to chio drive mapping, and the existence of
> the /etc/chio.conf file.
> "You commit, then jmc cleans" deraadt@ (beck@)
~ chio.1
> tweaks; (jmc@)
~ chio.c
> This adds support for retrieving volume tags in chio(8) with the
> status command and options -v to request the primary volume tag, -V for
> the alternate volume tag. Man page will follow shortly, along with move
> support by volume tag. (beck@)
~ chio.c
> prettier (deraadt@)
~ chio.1 ~ chio.c
> Move mtio drive prep to a function, and makes the changer
> check it's notion of access, avoiding problems on more tightly
> coupled changers if the tape is already ejected - currently
> these stupid devices return "Initialization Command Required"
> sense codes but there is (as yet) no way to tell userland
> that. checking the ACCESS changer flag appears safe in both
> cases.
> Adds support for "voltag" type for the source for a move, so
> you can do "chio move voltag 0003B72 drive 0" to move a tape
> with the corresponding bar code.
> Documents the above, as well as status -vVa in the man page.
> partly from freebsd, man page tweaks by jmc@
> ok krw@ (beck@)
~ chio.1
> fix stupid typo. mea culpa (beck@)
~ Makefile + chio.conf
> Add sample chio.conf to /etc
> ok deraadt@ (beck@)
- chio.conf ~ Makefile
> do chio.conf install from right place (deraadt@)
dd
~ dd.1
> - use .Ex
> - standard EXAMPLES (jmc@)
~ dd.1
> add another exmaple; stolen from a nick holland email... (jmc@)
ksh
~ history.c ~ ksh.1
~ lex.c ~ lex.h
> Implement \$ and \# expansion for PS1. Whoever thought it a clever
> idea to assign a special meaning to "\$" -- two chars that are already
> (very) special -- deserves a spanking. (otto@)
~ ksh.1
> document some advice when \$ is part of your prompt;
> from otto and myself (jmc@)
~ ksh.1
> simplify the PS1 section somewhat by moving the ! and \! stuff together;
> ok otto (jmc@)
mt
~ mt.c
> We do not create st devices with different density modes. i.e. with device
> minor numbers having bits 2 or 3 set.
> Eliminate the quirks, etc. used to store info on these non-existant modes.
> Also eliminate a couple of 'unimplemented' fields in the ioctl request
> structure.
> ok beck@ deraadt@ (krw@)
stty
~ print.c
> teach stty about NMEADISC (deraadt@)
== distrib =========================================================== 02/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/distrib
notes
~ alpha/hardware ~ amd64/hardware
~ i386/hardware
> sync, enable and move twe(4) out from untested on alpha (martin@)
sets
~ lists/man/mi
> sync (deraadt@)
~ lists/base/md.alpha ~ lists/base/md.amd64
~ lists/base/md.aviion ~ lists/base/md.cats
~ lists/base/md.hp300 ~ lists/base/md.hppa
~ lists/base/md.hppa64 ~ lists/base/md.i386
~ lists/base/md.luna88k ~ lists/base/md.mac68k
~ lists/base/md.macppc ~ lists/base/md.mvme68k
~ lists/base/md.mvme88k ~ lists/base/md.mvmeppc
~ lists/base/md.sgi ~ lists/base/md.sparc
~ lists/base/md.sparc64 ~ lists/base/md.vax
~ lists/base/md.zaurus ~ lists/comp/mi
> sync (deraadt@)
~ lists/man/mi
> sync (deraadt@)
~ lists/base/md.cats ~ lists/base/md.zaurus
~ lists/comp/md.cats ~ lists/comp/md.zaurus
> sync (deraadt@)
~ lists/etc/mi
> sync (deraadt@)
~ lists/base/mi ~ lists/man/mi
> sync (deraadt@)
~ lists/base/md.alpha ~ lists/base/md.amd64
~ lists/base/md.aviion ~ lists/base/md.cats
~ lists/base/md.hp300 ~ lists/base/md.hppa
~ lists/base/md.hppa64 ~ lists/base/md.i386
~ lists/base/md.luna88k ~ lists/base/md.mac68k
~ lists/base/md.macppc ~ lists/base/md.mvme68k
~ lists/base/md.mvme88k ~ lists/base/md.mvmeppc
~ lists/base/md.sgi ~ lists/base/md.sparc
~ lists/base/md.sparc64 ~ lists/base/md.vax
~ lists/base/md.zaurus ~ lists/comp/mi
~ lists/man/mi
> sync (deraadt@)
~ lists/man/mi
> sync (deraadt@)
~ lists/etc/mi
> sync (deraadt@)
== etc =============================================================== 03/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/etc
Makefile
~ Makefile
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
~ Makefile
> do chio.conf install from right place (deraadt@)
changelist
~ changelist
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
chio.conf
+ chio.conf
> do chio.conf install from right place (deraadt@)
dvmrpd.conf
+ dvmrpd.conf
> add a sample dvmrpd.conf
> ok deraadt@ (norby@)
ftpusers
~ ftpusers
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
group
~ group
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
mail/aliases
~ mail/aliases
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
master.passwd
~ master.passwd
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
mtree/special
~ mtree/special
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
~ mtree/special
> fix location of dvmrpd.conf; ok norby@ (david@)
pf.os
~ pf.os
> crank the OpenBSD version to 3.9; ok deraadt@
> Opera on OpenBSD verified by sturm@ (david@)
rc
~ rc
> Add support for NFS mounts to be from non-reserved ports:
> - new sysctl vfs.nfs.privport to require NFS mount requests to be on
> reserved ports when set to 1 (the default).
> - mountd now automatically sets the sysctl depending on the -n flag.
> - add mountd_flags to rc.conf to enable the -n flag at boot.
> deraadt@ ok (avsm@)
~ rc
> revert vfs.nfs.privport sysctl, broke a few architectures
> requested by deraadt@ (avsm@)
~ rc
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
~ rc
> Hook dhcrelay(8) into the startup process.
> ok henning@ (norby@)
~ rc
> Carp demotion interlock. Prevents carp from preempting until the system is
> booted, allowing for daemons to sync with peers before we take over.
> ok deraadt@ mpf@ moritz@ (mcbride@)
~ rc
> Add sasyncd to rc / rc.conf.
> ok deraadt@ cloder@ (mcbride@)
rc.conf
~ rc.conf
> Add support for NFS mounts to be from non-reserved ports:
> - new sysctl vfs.nfs.privport to require NFS mount requests to be on
> reserved ports when set to 1 (the default).
> - mountd now automatically sets the sysctl depending on the -n flag.
> - add mountd_flags to rc.conf to enable the -n flag at boot.
> deraadt@ ok (avsm@)
~ rc.conf
> revert vfs.nfs.privport sysctl, broke a few architectures
> requested by deraadt@ (avsm@)
~ rc.conf
> add all the goo to hook dvmrp into the system
> ok derradt@ (norby@)
~ rc.conf
> Hook dhcrelay(8) into the startup process.
> ok henning@ (norby@)
~ rc.conf
> Add sasyncd to rc / rc.conf.
> ok deraadt@ cloder@ (mcbride@)
services
~ services
> asf/ipmi balony at udp 623; sthen at bootes.spacehopper.orgeraadt@)
== gnu =============================================================== 04/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/gnu
usr.bin/binutils
~ bfd/tekhex.c
> Avoid stack oflow on bad data. PR 5132. Backport from binutils main
> repo rev 1.26.22.1; ok miod@ (otto@)
== lib =============================================================== 05/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
libc
~ gen/sysctl.3
> VFS_GENERIC returns VFS info, not VM info (pedro@)
~ gen/sysctl.3
> Add support for NFS mounts to be from non-reserved ports:
> - new sysctl vfs.nfs.privport to require NFS mount requests to be on
> reserved ports when set to 1 (the default).
> - mountd now automatically sets the sysctl depending on the -n flag.
> - add mountd_flags to rc.conf to enable the -n flag at boot.
> deraadt@ ok (avsm@)
~ gen/sysctl.3
> revert vfs.nfs.privport sysctl, broke a few architectures
> requested by deraadt@ (avsm@)
~ gen/sysctl.3
> document NFS_NIOTHREADS, with some pointers from mickey; (jmc@)
~ sys/adjtime.2
> Allow any user to request the current adjustment; ok henning@ deraadt@
> (otto@)
~ sys/adjtime.2
> Add a CAVEATS section: other systems might have more restrictive
> adjtime(2).
> With help from jmc@; ok deraadt@ jmc@ (otto@)
libpthread
~ include/pthread.h
> Bump minimum stack size; current setting causes hangs or illegal
> instructions once every few runs. Problem noted by Botond Botyanszki.
> ok marc@ (otto@)
== regress =========================================================== 06/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/regress
sbin
~ ipsecctl/ike11.ok
> Fix test including "Local-address" (hshoexer@)
~ ipsecctl/Makefile + ipsecctl/ipsec22.in
+ ipsecctl/ipsec22.ok + ipsecctl/ipsec23.in
+ ipsecctl/ipsec23.ok + ipsecctl/ipsec24.in
+ ipsecctl/ipsec24.ok
> tests for rule expansion (hshoexer@)
~ ipsecctl/Makefile + ipsecctl/ike12.in
+ ipsecctl/ike12.ok + ipsecctl/ike13.in
+ ipsecctl/ike13.ok + ipsecctl/ike14.in
+ ipsecctl/ike14.ok
> tests for rule expansion and ike (hshoexer@)
~ ipsecctl/Makefile + ipsecctl/ike15.in
+ ipsecctl/ike15.ok + ipsecctl/ike16.in
+ ipsecctl/ike16.ok + ipsecctl/ike17.in
+ ipsecctl/ike17.ok + ipsecctl/ike18.in
+ ipsecctl/ike18.ok + ipsecctl/ike19.in
+ ipsecctl/ike19.ok + ipsecctl/ike20.in
+ ipsecctl/ike20.ok + ipsecctl/ike21.in
+ ipsecctl/ike21.ok + ipsecctl/ike22.in
+ ipsecctl/ike22.ok + ipsecctl/ike23.in
+ ipsecctl/ike23.ok + ipsecctl/ike29.in
+ ipsecctl/ike29.ok + ipsecctl/ike30.in
+ ipsecctl/ike30.ok + ipsecctl/ike31.in
+ ipsecctl/ike31.ok + ipsecctl/ike32.in
+ ipsecctl/ike32.ok + ipsecctl/ike33.in
+ ipsecctl/ike33.ok + ipsecctl/ike34.in
+ ipsecctl/ike34.ok + ipsecctl/ike35.in
+ ipsecctl/ike35.ok + ipsecctl/ike36.in
+ ipsecctl/ike36.ok + ipsecctl/ike37.in
+ ipsecctl/ike37.ok + ipsecctl/ike38.in
+ ipsecctl/ike38.ok + ipsecctl/ike39.in
+ ipsecctl/ike39.ok + ipsecctl/ike40.in
+ ipsecctl/ike40.ok + ipsecctl/ikefail1.in
+ ipsecctl/ikefail1.ok + ipsecctl/ikefail2.in
+ ipsecctl/ikefail2.ok + ipsecctl/ipsec25.in
+ ipsecctl/ipsec25.ok + ipsecctl/ipsec26.in
+ ipsecctl/ipsec26.ok + ipsecctl/ipsec27.in
+ ipsecctl/ipsec27.ok + ipsecctl/ipsec28.in
+ ipsecctl/ipsec28.ok + ipsecctl/ipsec29.in
+ ipsecctl/ipsec29.ok + ipsecctl/ipsec30.in
+ ipsecctl/ipsec30.ok + ipsecctl/ipsec31.in
+ ipsecctl/ipsec31.ok + ipsecctl/ipsec32.in
+ ipsecctl/ipsec32.ok + ipsecctl/ipsec33.in
+ ipsecctl/ipsec33.ok + ipsecctl/ipsec34.in
+ ipsecctl/ipsec34.ok + ipsecctl/ipsec35.in
+ ipsecctl/ipsec35.ok + ipsecctl/ipsec36.in
+ ipsecctl/ipsec36.ok + ipsecctl/ipsec37.in
+ ipsecctl/ipsec37.ok + ipsecctl/ipsec38.in
+ ipsecctl/ipsec38.ok + ipsecctl/ipsec39.in
+ ipsecctl/ipsec39.ok + ipsecctl/ipsec40.in
+ ipsecctl/ipsec40.ok + ipsecctl/ipsec41.in
+ ipsecctl/ipsec41.ok + ipsecctl/ipsec42.in
+ ipsecctl/ipsec42.ok + ipsecctl/ipsec43.in
+ ipsecctl/ipsec43.ok + ipsecctl/ipsec44.in
+ ipsecctl/ipsec44.ok + ipsecctl/ipsec50.in
+ ipsecctl/ipsec50.ok + ipsecctl/ipsecfail1.in
+ ipsecctl/ipsecfail1.ok + ipsecctl/ipsecfail2.in
+ ipsecctl/ipsecfail2.ok
> add some regress for v6; ok hshoexer@ (todd@)
+ ipsecctl/sa13.in + ipsecctl/sa13.ok
+ ipsecctl/sa14.in + ipsecctl/sa14.ok
+ ipsecctl/sa15.in + ipsecctl/sa15.ok
+ ipsecctl/sa16.in + ipsecctl/sa16.ok
+ ipsecctl/sa17.in + ipsecctl/sa17.ok
+ ipsecctl/sa18.in + ipsecctl/sa18.ok
+ ipsecctl/sa19.in + ipsecctl/sa19.ok
+ ipsecctl/sa20.in + ipsecctl/sa20.ok
> prepare some more tests for v6 (hshoexer@)
+ ipsecctl/sa21.in + ipsecctl/sa21.ok
+ ipsecctl/sa22.in + ipsecctl/sa22.ok
+ ipsecctl/sa23.in + ipsecctl/sa23.ok
+ ipsecctl/sa24.in + ipsecctl/sa24.ok
> some more, not hooked up yet (hshoexer@)
~ ipsecctl/ike15.ok ~ ipsecctl/ike3.ok
~ ipsecctl/ike4.ok ~ ipsecctl/ike5.ok
~ ipsecctl/ike9.ok
> change the ike regression tests to use the new local-ID format
> ok hshoexer@ (msf@)
~ ipsecctl/ike16.ok ~ ipsecctl/ike23.ok
~ ipsecctl/ike29.ok ~ ipsecctl/ike37.ok
~ ipsecctl/ike38.ok
> modify local-ID sections for disabled regress tests too
> ok hshoexer@ (msf@)
- ipsecctl/ipsecfail2.in - ipsecctl/ipsecfail2.ok
~ ipsecctl/Makefile ~ ipsecctl/ipsecfail1.in
> fix this tests. Enable a bunch of new sa tests (hshoexer@)
~ ipsecctl/Makefile ~ ipsecctl/ipsec39.ok
> fix this test, too. (hshoexer@)
~ ipsecctl/Makefile + ipsecctl/ipsec51.in
+ ipsecctl/ipsec51.ok + ipsecctl/ipsec52.in
+ ipsecctl/ipsec52.ok + ipsecctl/ipsecfail2.in
+ ipsecctl/ipsecfail2.ok
> check port modifiers in flow rules; ok hshoexer@ (naddy@)
~ pfctl/pf56.loaded ~ pfctl/pf56.ok
~ pfctl/pf56.optimized
> Fix regress tests for automatic adaptive timeouts. (mcbride@)
~ ipsecctl/Makefile + ipsecctl/ike41.in
+ ipsecctl/ike41.ok
> new test for setting lifetimes (hshoexer@)
~ ipsecctl/Makefile
> enable all ike tests, even though some of them are currently broken.
> (hshoexer@)
~ ipsecctl/sa1.ok ~ ipsecctl/sa11.ok
~ ipsecctl/sa12.ok ~ ipsecctl/sa13.ok
~ ipsecctl/sa14.ok ~ ipsecctl/sa15.ok
~ ipsecctl/sa16.ok ~ ipsecctl/sa17.ok
~ ipsecctl/sa18.ok ~ ipsecctl/sa19.ok
~ ipsecctl/sa2.ok ~ ipsecctl/sa21.ok
~ ipsecctl/sa23.ok ~ ipsecctl/sa24.ok
~ ipsecctl/sa3.ok ~ ipsecctl/sa4.ok
~ ipsecctl/sa5.ok ~ ipsecctl/sa6.ok
~ ipsecctl/sa7.ok ~ ipsecctl/sa9.ok
> add trailing \ (markus@)
~ ipsecctl/Makefile + ipsecctl/ike42.in
+ ipsecctl/ike42.ok + ipsecctl/ike43.in
+ ipsecctl/ike43.ok + ipsecctl/ikefail3.in
+ ipsecctl/ikefail3.ok
> check port modifiers in ike rules (naddy@)
usr.bin
~ xlint/Makefile
> We pull in LINTFLAGS from sys.mk, no need to reproduce it here (cloder@)
~ xlint/test-23.c.exp
> The '-c' option is no longer the default in LINTFLAGS, so update the
> output of this test (cloder@)
+ file/Makefile + file/t0.in
+ file/t0.out + file/t1.in
+ file/t1.out + file/t10.in
+ file/t10.out + file/t15.in
+ file/t15.out + file/t17.in
+ file/t17.out + file/t18.in
+ file/t18.out + file/t2.in
+ file/t2.out + file/t20.in
+ file/t20.out + file/t21.in
+ file/t21.out + file/t22.in
+ file/t22.out + file/t23.in
+ file/t23.out + file/t3.in
+ file/t3.out + file/t30.in
+ file/t30.out + file/t4.in
+ file/t4.out + file/t5.in
+ file/t5.out + file/t6.in
+ file/t6.out + file/t7.in
+ file/t7.out + file/t8.in
+ file/t8.out + file/t9.in
+ file/t9.out
> Initial regress for file(1); not hooked in yet; looked at by cloder. (ian@)
~ xlint/test-16.c.exp ~ xlint/test-19.c.exp
> Reflect -r1.43 of lint1/tree.c (cloder@)
== sbin ============================================================== 07/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
brconfig
~ brconfig.8
> apropriate -> appropriate (jmc@)
dhclient
~ dhclient.c
> Fix 'special value 0 requests that dhclient not wait for a link state
> change before timing out' behaviour of the man page for link-timeout;
> ok deraadt@ henning@ (stevesk@)
~ dhclient.conf.5
> clarify link-timeout default as seconds; ok jmc@ (stevesk@)
~ dhclient.c
> Don't treat the addition of a dhclient.conf 'alias{}' address as an
> external modification when the RTM_NEWADDR message arrives from the
> routing socket. Now dhclient will not exit if 'alias{}' is specified
> in dhclient.conf.
> Mentioned many times, most recently by Matthias Bertschy on misc at . > ok henning@ beck@ (krw@)
~ dhclient-script
> '-alias' is deprecated. Use 'delete' in ifconfig statements.
> ok beck@ (krw@)
~ dhclient.conf.5
> backoff-cutoff defaults to 15 seconds; ok henning@ (stevesk@)
~ dhclient-script
> Use ifconfig delete/alias conforming with the man page.
> OK krw@ (mpf@)
dump
~ tape.c
> Don't double-eject tapes. Once is enough. Closes PR#1282.
> ok miod@ beck@, functionally equivalent diff ok deraadt@ a while ago (krw@)
~ tape.c
> Fix comment to reflect updated st(4) verbiage. (krw@)
fdisk
~ mbr.c
> Change MBR_read() and MBR_write() to return -1 on error and set
> errno = EIO for short reads. This makes it easier to check for
> read and write errors. Much patient tutoring by weingart@, deraadt@,
> and cloder at . > Correctly use ssize_t to store read(2) and write(2) return values.
> No functional change, since MBR_read() and MBR_write() return values
> aren't currently checked.
> OK deraadt@ (ray@)
fsck
~ fsutil.c
> revert a part of last commit causes newline
> proplems during boot.
> pointed out by sturm@ (thib@)
ifconfig
~ ifconfig.c
> ifconfig should be either deleting an interface address, or adding one in
> one
> invocation, not both.
> This change ensures that a delete does not also do an add.
> Fixes stupid problem where deleting the last address with
> ifconfig delete addr
> worked differently than
> ifconfig addr delete
> where the first way would re-add an address of 0.0.0.0/0 after deleting
> the address.
> ok claudio@, krw@ (beck@)
~ ifconfig.c
> Revert last commit. Modifing a interface does a remove and then an add.
> We need to figure out a better way to fix this.
> Brought up by markus@ OK beck@ (claudio@)
~ ifconfig.8 ~ ifconfig.c
> Introduce attributes to interface groups.
> As a first user, move the global carp(4) demotion counter
> into the interface group. Thus we have the possibility
> to define which carp interfaces are demoted together.
> Put the demotion counter into the reserved field of the carp header.
> With this, we can have carp act smarter if multiple errors occur.
> It now always takes over other carp peers, that are advertising
> with a higher demote count. As a side effect, we can also have
> group failovers without the need of running in preempt mode.
> The protocol change does not break compability with older
> implementations.
> Collaborative work with mcbride@
> OK mcbride@, henning@ (mpf@)
~ ifconfig.8
> kill trailing whitespace; (jmc@)
~ ifconfig.c
> Do not set newaddr to 1 if "delete" was specified beforhands. In this case
> doalias is < 0. This fixes the problem where ifconfig em0 delete 10.0.0.1
> created a 0.0.0.0/0 route entry and created a total mess because of that.
> Diff from markus@ OK beck@ markus@ (claudio@)
~ ifconfig.8
> groups are specified using -g, not -m; ok mpf (jmc@)
~ ifconfig.8
> better synopsis for -g; ok mpf (jmc@)
ipsecctl
~ parse.y
> add ERANGE error detection, found when looking at bgpd's parse.y
> ok hshoexer@ (todd@)
~ ike.c
> matching brackets are useful
> ok dlg@ (todd@)
~ ipsec.conf.5
> missing `Ar'; (jmc@)
~ ipsecctl.h
> this one not needed yet. (hshoexer@)
~ ipsecctl.h ~ pfkey.c
> fill in AF_INET6 cases
> this does not complete v6 support for ipsecctl, but makes progress
> ok hshoexer@ (todd@)
~ pfkey.c
> kill trainling whitespace (hshoexer@)
~ parse.y
> prepare for rule expansion. Get rid of addr_node, link struct
> ipsec_addr_wrap directly. (hshoexer@)
~ ipsecctl.h
> whoops, undo last commit. Of course, set_ipmask() is needed... (hshoexer@)
~ parse.y
> when parsing host specifications, initialize host address queue pointers,
> not
> used yet. (hshoexer@)
~ parse.y
> unify expansion of SA rules. Needed for general rule expansion.
> (hshoexer@)
~ parse.y
> move generation of reverse flow rules to seperat function. (hshoexer@)
~ parse.y
> merge expand_sa() and expand_rule(). (hshoexer@)
~ parse.y
> fix rule numbering (for -vv)
> late ikerule also use expand_rule (hshoexer@)
~ parse.y
> teach expand_rule() to iterate over host lists, not used yet. (hshoexer@)
~ parse.y
> As the rule expansion is now aware of host lists, host_if() has to
> return a proper list of addresses bound to an interface. (hshoexer@)
~ parse.y
> Also return proper list of addresses for interface groups.
> As usual, this and the previous commit reused suitable code from
> the tree (pfctl). (hshoexer@)
~ parse.y
> unify code a little bit (consistent variable names). (hshoexer@)
~ ipsecctl.c
> add ipsecctl_free_rule() for cleaning up rules. (hshoexer@)
~ parse.y
> Provide functions for copying members of rules. Implement copyrule()
> function to copy a single rule. Use that for rule expansion. (hshoexer@)
~ ipsecctl.h
> Need protoype for ipsecctl_free_rule(). While around clean up
> prototype for ipsecctl_add_rule. (hshoexer@)
~ parse.y
> enable lists.
> This allows rules like:
> ike from em0 to { 192.168.7.0/24, 192.168.9.0/24 } peer 1.2.3.4
> This will setup two tunnels to the networks 192.168.7.0/24 and
> 192.168.9.0/24. (hshoexer@)
~ ipsecctl.8 ~ ipsecctl.c
~ ipsecctl.h ~ pfkdump.c
~ pfkey.c ~ pfkey.h
> implement monitor mode for ipsecctl. worked on with markus@
> ok hshoexer@ (msf@)
~ ike.c
> Prepare for handling unnamed remote peers. (hshoexer@)
~ ike.c
> Small function header knf. (hshoexer@)
~ pfkdump.c
> shuffle some newlines to make all the formatting work again
> ok hshoexer@ (msf@)
~ pfkdump.c ~ pfkey.c
> whitespace cleanup (hshoexer@)
~ ipsecctl.8
> add a little markup; (jmc@)
~ parse.y
> add basic
> - IPv6 parsing for only v6 host addresses
> - checks for dst <-> src address family sanity
> ok hshoexer@ (todd@)
~ parse.y ~ ipsecctl.h
> white spaces (hshoexer@)
~ ike.c
> Generate correct configuration for default peers. (hshoexer@)
~ parse.y
> When no peer is specified, make this rule a "catch-all" rule for any remote
> peer. Similar to isakmpd(8)s "Default=" tag. (hshoexer@)
~ ipsecctl.c ~ ipsecctl.h
> rename list link for ipsec_rule structures from "entries" to "rule_entry".
> (hshoexer@)
~ parse.y
> add more v6 support, this round `any' expands additionally to ::/0
> skip link-locals for now, to be handled separately later
> ok hshoexer@ (todd@)
~ parse.y
> knf (hshoexer@)
~ ipsecctl.c ~ pfkey.c
> correct error messages to match calloc where appropriate
> ok hshoexer@ (todd@)
~ ike.c
> permit feeding isakmpd.fifo IPv6 addresses
> ok hshoexer@ (todd@)
~ ike.c
> knf (hshoexer@)
~ parse.y ~ pfkdump.c
~ pfkey.h
> spacing (deraadt@)
~ pfkey.c
> read the full reply from PFKEY even if sadb_errno is set; ok hshoexer
> (markus@)
~ parse.y
> print actual key size when warning about the wrong key size; ok hshoexer
> (markus@)
~ ipsecctl.h ~ ipsecctl.c
> Prepare for SA grouping. (hshoexer@)
~ ipsecctl.h
> Add members dst2, proto2 and spi2 to struct ipsec_rule and define
> rule type "group". Needed for grouping. (hshoexer@)
~ ipsecctl.c ~ ipsecctl.h
~ pfkdump.c
> convert pfkey to ipsec_rule and use ipsecctl_print_rule() when dumping
> the in-kernel SAs. this way we produce the same output as rule loading
> ok hshoexer (markus@)
~ ipsecctl.c
> more to free, needed for SA grouping. (hshoexer@)
~ ipsecctl.c ~ ipsecctl.h
~ parse.y ~ pfkey.c
> Support flows with port modifiers for proto tcp/udp, e.g.
> flow proto udp from 1.2.3.4 port ntp to 5.6.7.8
> ok hshoexer@ msf@ (naddy@)
~ ike.c
> change the local-ID section name to always be unique as we may want to use
> more than one ISAKMP ID on the local peer.
> ok hshoexer@ (msf@)
~ ipsec.conf.5
> document port matching in flows; ok hshoexer@ (naddy@)
~ ipsec.conf.5
> address has two `d', and i had to use a dictionary to check ;) (jmc@)
~ pfkey.c
> pfkey bits needed for SA grouping (hshoexer@)
~ parse.y
> Final bits for SA grouping. (hshoexer@)
~ ipsecctl.h ~ parse.y
> Prepare for parsing lifetimes for ike main and quick mode. Not enabled
> yet. (hshoexer@)
~ ipsecctl.h ~ parse.y
> Generalize parsing of main/quick mode specification. Preparation
> for lifetime support. (hshoexer@)
~ parse.y ~ ike.c
> Simplify main/quick mode parsing and generation of the actual ike config.
> (hshoexer@)
~ pfkdump.c
> fix the formatting for sadb_register messages in monitor mode.
> put back one mistakenly deleted newline.
> ok hshoexer@ (msf@)
~ pfkdump.c
> tiny style cleanup and white spaces (hshoexer@)
~ parse.y ~ ipsecctl.h
> put src and dst host in dedicated structure. Make the API more
> compact which will soon simplify my life. (hshoexer@)
~ parse.y
> some more cleanup and simplification, no functional change. (hshoexer@)
~ parse.y
> simplify handling of peers. (hshoexer@)
~ ike.c ~ ipsecctl.c
~ ipsecctl.h ~ parse.y
> allow to specify phase 1 and 2 lifetimes. Right now, these values
> can only be set globally (ie. Default-phase-[12]-lifetime). (hshoexer@)
~ ipsec.conf.5
> mark up keywords using .Ic; ok hshoexer (jmc@)
~ ipsecctl.c
> add trailing \ when printing multiple lines for an SA, this way
> the output of ispecctl matches its input (markus@)
~ ipsecctl.h ~ pfkdump.c
> print full information about tcpmd5 and ipcomp SAs, too (markus@)
~ ike.c ~ parse.y
> support tcp/udp port modifiers in ike rules
> "put it in if it doesn't break regress" hshoexer@ (naddy@)
~ ipsec.conf.5
> document port modifiers in ike rules (naddy@)
~ ipsecctl.c
> exit(2) when loading of rules did work partially. ok markus@ (hshoexer@)
~ ipsec.conf.5
> correct spelling of specified (david@)
isakmpd
~ ipsec.c
> Change the default replay window for SAs created by the isakmpd responder
> to be DEFAULT_REPLAY_WINDOW instead of zero. The default replay window is
> then the same for both initiator and receiver.
> Fix from nathanael at polymorpheous dot com.
> ok hshoexer@ (mcbride@)
~ sa.c
> also report SA flags. (hshoexer@)
~ sa.c
> Assign a finalization event to the exchange initiated on soft expiry.
> If the exchange fails, the existing phase 1 SA is invalidated and the
> exchange is retried at the Exchange-Max-Time interval until the SA hard
> timeout expires.
> Another sasyncd-related fix from nathanael at polymorpheous dot com
> ok ho@ hshoexer@ (mcbride@)
~ ipsec.c
> Fix broken merge of patch. Pointed out by nathanael at polymorpheus dot
> com. (mcbride@)
~ sa.c
> Oops, return after calling sa_release() (mcbride@)
~ pf_key_v2.c ~ pf_key_v2.h
> export pf_key_v2_disable_sa() (unbreaks build) (markus@)
~ sa.c
> Do not use C++ comments. Noticed by markus. (hshoexer@)
~ message.c
> fix SA grouping. Now, esp+ah and ah+esp works again.
> ok markus@ (hshoexer@)
~ pf_key_v2.c
> tiny KNF (hshoexer@)
~ exchange.c ~ exchange.h
~ message.c ~ pf_key_v2.c
~ sa.c
> Make sure, that phase 1 SAs of active connections stay alive. Fixes a DPD
> breakage noticed and reported by Mitja Muzenic.
> ok markus@ ho@, testing by Mitja and cloder@, discussed with Nathanael.
> (hshoexer@)
~ isakmpd.c
> Fix a comment (hshoexer@)
~ ipsec.h
> Big whitespace cleanup. (hshoexer@)
~ BUGS ~ DESIGN-NOTES
~ crypto.h ~ dpd.c
~ exchange.c ~ exchange_num.cst
~ ike_quick_mode.c ~ isakmpd.policy.5
~ math_2n.c ~ math_group.c
~ math_group.h ~ message.c
~ sa.c ~ sa.h
~ transport.h ~ util.c
~ virtual.c ~ x509.c
> Big spelling cleanup, no binary change. From david@ (hshoexer@)
lmccontrol
~ lmccontrol.c
> Replace atoi with strtonum.
> Use __progname in usage()
> ok otto jaredy (dhill@)
~ lmccontrol.8
> Change LMCCTL to LMCCONTROL. It is lmccontrol(8), after all.
> ``yeah!'' jmc@ (ray@)
~ lmccontrol.8 ~ lmccontrol.c
> - sort options
> - sync usage()
> - clean up (jmc@)
mountd
~ mountd.c
> Add support for NFS mounts to be from non-reserved ports:
> - new sysctl vfs.nfs.privport to require NFS mount requests to be on
> reserved ports when set to 1 (the default).
> - mountd now automatically sets the sysctl depending on the -n flag.
> - add mountd_flags to rc.conf to enable the -n flag at boot.
> deraadt@ ok (avsm@)
~ mountd.c
> revert vfs.nfs.privport sysctl, broke a few architectures
> requested by deraadt@ (avsm@)
pdisk
~ convert.c
> use <machine/endian.h> to figure what we are running on instead of defining
> our own macros (martin@)
pfctl
~ pfctl.c
> Enable adaptive timeouts by default, with adaptive.start of 60% of the
> state limit and adaptive.end of 120% of the state limit.
> Explicitly setting the adaptive timeouts will override the default,
> and it can be disabled by setting both adaptive.start and adaptive.end to
> 0.
> ok henning@ (mcbride@)
~ parse.y
> Make per-rule adaptive timeouts behave the same way as the global adaptive
> timeouts. (mcbride@)
reboot
~ reboot.c
> delint
> ok moritz (dhill@)
route
~ show.c
> this fixes a crash observed by Naddy on amd64, and permits the src/dst ip's
> to
> be visible .. when IPv6 sa flows are present
> ok claudio@ (todd@)
sysctl
~ sysctl.8
> Add support for NFS mounts to be from non-reserved ports:
> - new sysctl vfs.nfs.privport to require NFS mount requests to be on
> reserved ports when set to 1 (the default).
> - mountd now automatically sets the sysctl depending on the -n flag.
> - add mountd_flags to rc.conf to enable the -n flag at boot.
> deraadt@ ok (avsm@)
~ sysctl.8
> revert vfs.nfs.privport sysctl, broke a few architectures
> requested by deraadt@ (avsm@)
~ sysctl.c
> print time offsets much nicer (deraadt@)
== share ============================================================= 08/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man5/pf.conf.5
> Adaptive timeouts are now on by default. (mcbride@)
~ man5/pf.conf.5
> put previous in the correct place; ok mcbride (jmc@)
~ man4/Makefile ~ man4/pci.4
+ man4/sdhc.4 + man4/sdmmc.4
> man pages for the SD subsystem (deraadt@)
~ man4/udcf.4
> +.Xr ntpd 8 ,
> ok henning mbalmer (jmc@)
~ man4/scsi.4
> Xr sdmmc 4 (uwe@)
~ man4/sdmmc.4
> complete sentence (uwe@)
~ man4/trunk.4
> in the second example, only the trunk interface should have an IP address,
> not the trunk ports.
> figured out by jmc@ (reyk@)
~ man4/udcf.4
> Change the timedelta sensor descritpion so that the first four bytes
> contain
> the NTP identifier (filled with spaces), followed by a space and the
> station
> name (which might be the same as the NTP identifier).
> All timedelta sensor have to follow this scheme as OpenNTPD needs it.
> Requested by henning. (mbalmer@)
~ man4/sdmmc.4
> - scsi -> SCSI
> - fix SEE ALSO (jmc@)
+ man9/sensor_add.9
> Add docs for kernel sensor API.
> ok jmc dlg (mk@)
~ man9/Makefile
> Install and MLINK sensor function man pages. (mk@)
~ man4/man4.sparc64/Makefile ~ man4/man4.sparc64/beeper.4
> document beep(4) (jason@)
~ man9/sensor_add.9
> small tweaks; (jmc@)
~ man4/wskbd.4
> Add support for the audio volume keys found on many laptops' builtin
> keyboard. These specific keys are posted to a kernel thread which will
> issue mixer commands if an audio device exists.
> Written by Alexey Vatchenko <avv , mail zp ua> with tweaks by deraadt@ and
> I. (miod@)
~ man4/mpi.4
> typo. fixed by Srebrenko Sehic. (dlg@)
~ man4/mpi.4
> a small start at documenting which boards are supported by my driver. i
> found these on the lsi site, but there are many more out there.
> im not happy with how this is formatted, if someone wants to make it better
> then send diffs. please. (dlg@)
~ man4/man4.i386/wdt.4
> Sync with reality. ok jmc (mk@)
~ man4/st.4
> Bring man page into closer correspondance with current reality. Tweaks by
> jmc at . > ok beck@ jmc@ (krw@)
~ man4/st.4
> Clean out more crap. Tweaks by jmc at . > ok beck@ jmc@ (krw@)
~ man4/Makefile ~ man4/tty.4
+ man4/nmea.4
> Add basic NMEA0183 support as a tty line discipline. The line discipline
> decodes NMEA messages completely transparent to userland applications, i.e.
> userland can still use the NMEA stream and talk to the device. If data
> is received, a timedelta sensor suitable for ntpd is created. The
> timestamp
> is not very precise at the moment, use of this is experimental at best.
> "get it in or we see how well the swiss people can swim" deraadt@
> (mbalmer@)
~ man4/tty.4
> missing full stop; (jmc@)
~ man4/nmea.4
> tweaks; (jmc@)
~ man4/man4.sparc64/Makefile + man4/man4.sparc64/ppm.4
> manpage for ppm driver (jason@)
~ man4/com.4
> mention com* at cardbus? for macppc. (fkr@)
~ man4/udcf.4
> - ntpd no longer needs the NTP identifier as part of a timedelta sensor
> description.
> - create the timedelta sensors as early as possible, but mark them invalid
> as long as there is no real data.
> - update docs accordingly (mbalmer@)
mk
~ sys.mk
> Change LINTFLAGS to '-hx', now that 'a' and 'b' are gone and 'c' really
> only produces a 'pointer casts may be troublesome' warning that everyone
> ignores. (cloder@)
== sys =============================================================== 09/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
sys
~ Makefile
> Link armish in to the builds. (drahn@)
arch
+ armish/Makefile
> Link armish in to the builds. (drahn@)
arch/alpha/conf
~ GENERIC ~ RAMDISKBIG
> sync, enable and move twe(4) out from untested on alpha (martin@)
~ Makefile.alpha
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
~ GENERIC ~ RAMDISKBIG
> add xge(4). (brad@)
arch/amd64/amd64
~ autoconf.c
> do not print "root on ..." twice for the -a case, also cope with systems
> configured for 'root on nfs' (deraadt@)
~ apic.c ~ mpbios.c
> Remove unused NetBSD new-style %b strings. (miod@)
~ machdep.c
> do not call acpi_powerdown() unless acpi has successfully initialized
> ok gwk@ canacar@ (todd@)
arch/amd64/conf
~ GENERIC
> Add ACPI Thermal zone and Embedded controller devices
> commented out. (canacar@)
~ GENERIC ~ files.amd64
> Enable sdmmc on amd64 too. ok brad@, deraadt at .atthieu@)
~ GENERIC
> enable mpi in GENERIC on amd64, i386, and macppc. jason already enabled it
> on sparc64 last night. macppc gets it in RAMDISK as well. (dlg@)
~ RAMDISK_CD
> enable mpi
> ok deraadt@ (dlg@)
~ GENERIC
> Disable ipmi in GENERIC for the moment - it breaks too much stuff
> on any of my sun amd64's as well as other ones (hp). We can re-enable
> it again when marco fixes it to be stable.
> ok deraadt@ and the crocodile hunter.. (beck@)
~ RAMDISK_CD
> enable xge(4) on the CD RAMDISKs. (brad@)
~ GENERIC
> Compile broken ipmi into GENERIC, so it's easy for marco to fix
> it by enabling it, but leave it disabled by default so it doesn't
> screw people over until it is fixed.
> ok dlg@, asked for by deraadt@ (beck@)
~ Makefile.amd64
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/amd64/pci
~ pci_machdep.c
> Reorg the code to be more sane. More coming.
> ok brad@ jason@ (weingart@)
arch/arm/arm
~ conf.c ~ cpu.c
~ cpufunc.c
> Add support for i80321 based systems. (drahn@)
arch/arm/footbridge
- genassym.cf
> Remove dead toes. (miod@)
arch/arm/include
~ armreg.h
> Add support for i80321 based systems. (drahn@)
arch/arm/xscale
+ files.i80321 + i80321.c
+ i80321_intr.c + i80321_mcu.c
+ i80321_pci.c + i80321_space.c
+ i80321_timer.c + i80321reg.h
+ i80321var.h + xscalereg.h
> Add support for i80321 based systems. (drahn@)
~ i80321.c ~ i80321_intr.c
~ i80321_mcu.c ~ i80321_pci.c
~ i80321_space.c ~ i80321_timer.c
~ i80321reg.h ~ i80321var.h
~ xscalereg.h
> Tags. (drahn@)
~ i80321_space.c
> correctly perform bus base/physical addr translation. (drahn@)
~ i80321.c
> Just say NO to noisy dmesg. (drahn@)
~ i80321_space.c
> IO addresses will always be in the right range, this check is wrong.
> (drahn@)
~ i80321_intr.c
> current_ipl_level is a value, not a mask, do not '|=' into it. (drahn@)
~ i80321_intr.c
> Splraise is strictly to increase spl, do not lower. (drahn@)
~ i80321_timer.c
> After changing J3 to J4 on the thecus board, serial interrupts work just
> fine (on XINT1), so remove the poll console hacks and rewire the com
> irq to the correct value for thecus. (drahn@)
~ i80321.c
> cleanup, move thecus board_reset from MI arm/xscale into armish dir.
> (drahn@)
arch/armish/armish
+ armish_machdep.c + armish_start.S
+ autoconf.c
> Add a new port for ARM based products from a variety of vendors, typically
> using Redboot as a loader and using pci, that have a variety of fairly
> minor
> hardware differences between them. (drahn@)
~ armish_start.S
> Tag. (drahn@)
~ autoconf.c
> Remove unused cfdriver field, pointed out by miod@ (drahn@)
~ armish_machdep.c
> Use same defines as on zaurus, need to verify the correct settings here.
> again, pointed out by miod@ (drahn@)
arch/armish/compile
+ .cvsignore
> Add a new port for ARM based products from a variety of vendors, typically
> using Redboot as a loader and using pci, that have a variety of fairly
> minor
> hardware differences between them. (drahn@)
arch/armish/conf
+ GENERIC + Makefile.armish
+ RAMDISK + files.armish
> Add a new port for ARM based products from a variety of vendors, typically
> using Redboot as a loader and using pci, that have a variety of fairly
> minor
> hardware differences between them. (drahn@)
~ files.armish
> Tag. (drahn@)
~ files.armish
> while $ is right next to #, it is not a comment character. (drahn@)
~ Makefile.armish
> Remove dead toes. (miod@)
~ RAMDISK
> Consistantly map com0 irq. (drahn@)
~ files.armish
> Better to attach obio devices after pci. (drahn@)
arch/armish/dev
+ com_obio.c + i80321_mainbus.c
+ iq80321_pci.c + iq80321reg.h
+ iq80321var.h + obio.c
+ obio_space.c + obiovar.h
+ pci_addr_fixup.c + pciide_machdep.c
> Add a new port for ARM based products from a variety of vendors, typically
> using Redboot as a loader and using pci, that have a variety of fairly
> minor
> hardware differences between them. (drahn@)
~ com_obio.c ~ i80321_mainbus.c
~ iq80321_pci.c ~ iq80321reg.h
~ iq80321var.h ~ obio.c
~ obio_space.c ~ obiovar.h
> Tag. (drahn@)
~ pci_addr_fixup.c
> do not commit debug. (drahn@)
~ i80321_mainbus.c ~ iq80321_pci.c
~ iq80321var.h ~ pci_addr_fixup.c
> Preserve ATU translations instead of forcing the values to something.
> Detect device type using some simple checks and use table driven
> interrupt maps for known devices.
> IO space appears to not want to be at 0, but at some bus offset
> (0x90000000)
> so skew mappings into this io bus base.
> Thecus N2100 and I/O Data Landisk (HDL-G) are now partially working.
> (drahn@)
~ i80321_mainbus.c
> Just say no to noisy dmesg. (drahn@)
~ pci_addr_fixup.c
> Properly remap I/O Addresses for theucs/IOData boxes. (drahn@)
~ iq80321_pci.c
> Now that this maps correctly, allow the interrupt to map properly. (drahn@)
~ iq80321_pci.c
> Correct IRQ mappings for thecus devices. (drahn@)
~ i80321_mainbus.c ~ obio.c
> Proper config attach, not always return 1 from probe. (drahn@)
~ com_obio.c ~ iq80321_pci.c
> After changing J3 to J4 on the thecus board, serial interrupts work just
> fine (on XINT1), so remove the poll console hacks and rewire the com
> irq to the correct value for thecus. (drahn@)
~ i80321_mainbus.c
> cleanup, move thecus board_reset from MI arm/xscale into armish dir.
> (drahn@)
arch/armish/include
+ _types.h + apmvar.h
+ armish_intr.h + asm.h
+ atomic.h + bootconfig.h
+ bus.h + cdefs.h
+ conf.h + cpu.h
+ db_machdep.h + disklabel.h
+ elf_abi.h + endian.h
+ exec.h + float.h
+ fp.h + frame.h
+ ieee.h + ieeefp.h
+ internal_types.h + intr.h
+ irqhandler.h + isa_machdep.h
+ limits.h + loadfile_machdep.h
+ lock.h + math.h
+ param.h + pcb.h
+ pci_machdep.h + pio.h
+ pmap.h + proc.h
+ profile.h + psl.h
+ ptrace.h + reg.h
+ reloc.h + rtc.h
+ setjmp.h + signal.h
+ spinlock.h + stdarg.h
+ sysarch.h + trap.h
+ vmparam.h
> Add a new port for ARM based products from a variety of vendors, typically
> using Redboot as a loader and using pci, that have a variety of fairly
> minor
> hardware differences between them. (drahn@)
- irqhandler.h - isa_machdep.h
- math.h - psl.h
- rtc.h ~ vmparam.h
> Several pieces were copied over from previous ports that are unnecessary,
> pointed out by miod@ (drahn@)
~ intr.h
> Ding dong, __NEWINTR is dead (miod@)
~ vmparam.h
> Shrink VM_PHYSSEG_MAX to 1 on platforms where physical memory is contiguous
> and there is only one call to uvm_page_physload(). (miod@)
arch/armish/stand
+ Makefile
> Link armish in to the builds. (drahn@)
arch/cats/conf
~ Makefile.cats
> Remove dead toes. (miod@)
~ Makefile.cats
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/hp300/include
~ vmparam.h
> When VM_PHYSSEG_MAX is 1, use VM_PSTRAT_RANDOM for VM_PHYSSEG_STRAT, this
> makes uvm_page.o smaller. (miod@)
arch/hppa/conf
~ Makefile.hppa
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/hppa/dev
~ mem.c
> allow ints for older asp; found on 710 @session; tested on 710, 715/33
> (miod), 735/125 (graham gower) (mickey@)
arch/hppa/gsc
~ gsckbdmap.c
> Regen (although this does not change anything for these keyboards). (miod@)
arch/hppa/hppa
~ machdep.c
> minor nit, use CPU ID instead of FPU ID for BTLB parameters after passing
> all kludges (martin@)
arch/hppa/include
~ vmparam.h
> Shrink VM_PHYSSEG_MAX to 1 on platforms where physical memory is contiguous
> and there is only one call to uvm_page_physload(). (miod@)
arch/hppa64/conf
~ Makefile.hppa64
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/hppa64/include
~ vmparam.h
> Shrink VM_PHYSSEG_MAX to 1 on platforms where physical memory is contiguous
> and there is only one call to uvm_page_physload(). (miod@)
arch/i386/conf
~ GENERIC ~ files.i386
> Support for standard SD host controllers like the Ricoh 5C822, a small
> generic bus layer, and SCSI emulation for SD/MMC memory cards. (uwe@)
~ GENERIC
> GENERIC (canacar@)
~ GENERIC
> Enable wdt(4) by default. ok deraadt (mk@)
~ GENERIC
> enable mpi in GENERIC on amd64, i386, and macppc. jason already enabled it
> on sparc64 last night. macppc gets it in RAMDISK as well. (dlg@)
~ RAMDISK_CD
> enable mpi
> ok deraadt@ (dlg@)
~ GENERIC
> Disable ipmi in GENERIC for the moment - it breaks too much stuff
> on any of my sun amd64's as well as other ones (hp). We can re-enable
> it again when marco fixes it to be stable.
> ok deraadt@ and the crocodile hunter.. (beck@)
~ RAMDISK_CD
> enable xge(4) on the CD RAMDISKs. (brad@)
~ GENERIC
> Compile broken ipmi into GENERIC, so it's easy for marco to fix
> it by enabling it, but leave it disabled by default so it doesn't
> screw people over until it is fixed.
> ok dlg@, asked for by deraadt@ (beck@)
~ Makefile.i386
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/i386/i386
~ autoconf.c
> do not print "root on ..." twice for the -a case, also cope with systems
> configured for 'root on nfs' (deraadt@)
~ est.c
> Remove obvious EST table dupe: Pentium M 733 and Pentium M 773 (VID #H),
> and zap trailing spaces. No functional change. (dim@)
~ ipifuncs.c ~ lapic.c
> implement reliable microtime on smp (joint work w/ otto):
> send a bcast ipi from the cpu0 to all others to timestamp tsc
> on every clock intr. this way using local tsc provides accurate
> microtime() measurements. thus gettimeofday test passes now.
> otto@ deraadt@ ok (mickey@)
~ apic.c ~ mpbios.c
> Remove unused NetBSD new-style %b strings. (miod@)
~ machdep.c
> do not call acpi_powerdown() unless acpi has successfully initialized
> ok gwk@ canacar@ (todd@)
arch/i386/include
~ cpu.h ~ intr.h
> implement reliable microtime on smp (joint work w/ otto):
> send a bcast ipi from the cpu0 to all others to timestamp tsc
> on every clock intr. this way using local tsc provides accurate
> microtime() measurements. thus gettimeofday test passes now.
> otto@ deraadt@ ok (mickey@)
arch/i386/isa
~ pccomvar.h
> remove sc_ic member from struct com_softc as it is not used anywhere
> ok miod@, martin@ (fkr@)
arch/i386/pci
~ pci_machdep.c
> Spelling. (pascoe@)
arch/m88k/include
~ vmparam.h
> When VM_PHYSSEG_MAX is 1, use VM_PSTRAT_RANDOM for VM_PHYSSEG_STRAT, this
> makes uvm_page.o smaller. (miod@)
arch/m88k/m88k
~ pmap.c
> Better test for cacheability of the previous mapping in pmap_cache_ctrl()
> (miod@)
~ pmap.c
> CACHE_GLOBAL is enforced at the APR level, no need to propagate it into
> the individual PTEs. No functional change. (miod@)
arch/macppc/conf
~ GENERIC ~ RAMDISK
> enable mpi in GENERIC on amd64, i386, and macppc. jason already enabled it
> on sparc64 last night. macppc gets it in RAMDISK as well. (dlg@)
~ Makefile.macppc
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
~ GENERIC ~ RAMDISK
> add xge(4). (brad@)
~ GENERIC
> now that com_cardbus works on macppc, add 'com* at cardbus?' to GENERIC
> ok martin@ (fkr@)
arch/macppc/pci
~ vgafb.c
> The aperture that is needed in legacy vga space is even larger for
> the mac mini's ATI graphics card. The full 0xa0000-0xfffff range is
> used here. (matthieu@)
arch/mips64/include
~ pmap.h
> clean userland namespace a bit more; ok miod (deraadt@)
arch/mips64/mips64
~ trap.c
> Reset pcb_onfault before invoking uvm_fault(). (miod@)
arch/mvme68k/dev
~ vs.c
> feild -> field (miod@)
arch/mvme68k/include
~ vmparam.h
> When VM_PHYSSEG_MAX is 1, use VM_PSTRAT_RANDOM for VM_PHYSSEG_STRAT, this
> makes uvm_page.o smaller. (miod@)
arch/mvme88k/conf
~ GENERIC ~ M187
~ M188 ~ M197
~ RAMDISK
> Configure network devices at ipl 3, not ipl 1. (miod@)
arch/mvmeppc/conf
~ Makefile.mvmeppc
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/powerpc/include
~ pmap.h
> clean userland namespace a bit more; ok miod (deraadt@)
arch/sgi/conf
~ Makefile.sgi
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/solbourne/conf
~ Makefile.solbourne
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/solbourne/include
~ vmparam.h
> Shrink VM_PHYSSEG_MAX to 1 on platforms where physical memory is contiguous
> and there is only one call to uvm_page_physload(). (miod@)
arch/sparc/conf
~ Makefile.sparc
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/sparc/dev
~ if_le.c
> When trying to find leaf lebuffer devices, to work around broken old PROMs,
> walk the lebuffer_cd device list, instead of the sbus children list.
> (miod@)
~ agten.c ~ bevar.h
~ bwtwo.c ~ cgsix.c
~ cgthree.c ~ cgtwelve.c
~ cs4231.c ~ cs4231var.h
~ dma.c ~ dmavar.h
~ esp.c ~ espvar.h
~ hme.c ~ hmevar.h
~ if_en_sbus.c ~ if_le.c
~ if_levar.h ~ lebuffer.c
~ lebuffervar.h ~ mgx.c
~ p9000.c ~ p9100.c
~ presto.c ~ qec.c
~ qecvar.h ~ qevar.h
~ rfx.c ~ sbus.c
~ sbusvar.h ~ spif.c
~ spifvar.h ~ stp_sbus.c
~ tcx.c ~ ts102.c
~ tvtwo.c ~ vigra.c
~ xbox.c ~ xboxvar.h
~ zx.c
> sbus_establish() and the associated linked list in the sbus softc is now
> only
> used to store a per-device reset callback, for use in sbusreset(). Except
> sbusreset() has never, ever, been used since Torek's sbus code went in.
> Time to recycle those wasted bits. (miod@)
arch/sparc/sparc
~ intreg.h
> Convert NetBSD new-style %b strings to old-style. (miod@)
arch/sparc64/conf
~ files.sparc64
> placeholder driver for the bbc,beep device. should be straightforward
> to fill in later. (jason@)
~ GENERIC
> mpi works on sparc64 (running a U3 booted/rooted from one to commit this).
> dlg can fix the distrib/www stuff to doc it. (jason@)
~ Makefile.sparc64
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
~ GENERIC ~ files.sparc64
> ppm driver. We don't use it for much... except to blink the power switch
> led. woot! (jason@)
arch/sparc64/dev
~ iommureg.h ~ schizo.c
> don't hardcode the iova... it uses the same scheme as the sbus machines
> (cd0 at scsibus0 at siop0 now attaches on the blade2k) (jason@)
~ sab.c
> make the ISR bits visible so we can poll them in cngetc (w/out cnpollc).
> I still hate this chip. (jason@)
~ schizo.c
> oops, nuke debugging stuff (jason@)
~ schizo.c ~ schizoreg.h
> - Implement schizo_intr_map (mainly to | in the agentid).
> - get correct IGN for intr_establish
> all of this is from kettenis and henric, I'm just cleaning it up and
> committing
> the easy bits (jason@)
~ sab.c
> make sure the IMR values are visible when we drop back into the prom
> (jason@)
+ beep.c
> placeholder driver for the bbc,beep device. should be straightforward
> to fill in later. (jason@)
~ beep.c
> fetch the beep clock frequency (we'll need it to compute the audio
> frequency later) (jason@)
~ beep.c
> implement frequency setting for the beep. It also beeps for one second
> during boot (this is annoying, I know... I'll hook it up to the keyboard
> handler later). (jason@)
~ vgafb.c
> Sync region detection algorithm with macppc, helps some recent cards, such
> as 3DLabs Permedia and Intergraph Expert3D.
> Also, if no mmio region is found, attach anyway, this only prevents
> accelerated
> X11 from starting. (miod@)
~ schizo.c ~ schizoreg.h
~ schizovar.h
> enable error interrupts and install handlers for same (the handlers don't
> do much yet) (jason@)
~ schizoreg.h
> add bit defns for [UE]FAR (jason@)
~ schizo.c ~ schizoreg.h
~ schizovar.h
> - stop using magic constants for the onboard error INOs (one was wrong
> anyway)
> - setup pci error interrupts and do an interpretive dance with the
> registers (jason@)
~ schizo.c ~ schizoreg.h
> do a better job of setting up the PBM. Doesn't fix the PIO faults we're
> seeing, I"m not convinced they're fixable any more. (jason@)
~ schizo.c
> latch the AFSR/AFAR -before- reading the PCI control/status register
> (jason@)
~ schizoreg.h
> bits bits bits, who got dat bits (jason@)
+ ppm.c
> ppm driver. We don't use it for much... except to blink the power switch
> led. woot! (jason@)
~ schizoreg.h
> correct bits for bus type in PCIAFSR (jason@)
~ schizo.c
> - clear the PCI_AFSR during init
> - set up the interrupts more correctly
> - print errors nicer (jason@)
~ pci_machdep.c
> turn the non-DEBUG case of DPRINTF into a do {} while (0) so it works
> more or less as expected (and removes the need to a stupid #ifdef) (jason@)
~ ce4231var.h ~ sbus.c
~ sbusvar.h ~ stp_sbus.c
> sbus_establish() and the associated linked list in the sbus softc is now
> only
> used to store a per-device reset callback, for use in sbusreset(). Except
> sbusreset() has never, ever, been used since Torek's sbus code went in.
> Time to recycle those wasted bits. (miod@)
arch/sparc64/include
~ autoconf.h
> extend the bootpath values to 64bits, and deal with fallout
> extend str2hex to handle long's, and teach it that strings starting with
> 'w',
> consisting of hex digits are really just hex digits
> match fp to scsibus (fiber port)
> match fiber channel portwwn/lun to it's bootpath variables
> result: with a small change to isp(mailed to mjacob) we can correctly
> autoconf with a boot path of:
> /pci at 80000/SUNW,qlc at 4@0,0/disk at w21000004cf948498a/bsd
> ie. schizo/ isp/ scsisbus/disk with port WWN of 21...98, lun 0,
> partition a (jason@)
~ pmap.h
> It's bit 13 that's causing the address aliasing in the cache, not bit 14.
> ok jason@ (kettenis@)
~ pte.h
> feild -> field (miod@)
~ pte.h
> All I wanted was to fix the bogus PG_SZ macro, but this is better:
> Do not try to map the message buffer with large pages, it's just as easy to
> simply use as many 8KB pages as necessary. From NetBSD. (miod@)
arch/sparc64/sparc64
~ autoconf.c
> more hotplug fallout: add DV_SENSOR to the device type switch (jason@)
~ autoconf.c
> intrducing device type DV_SENSOR turned out to be a bad idea as it
> affects way more than just hotplug messages. remove DV_SENSOR and use
> DV_DULL for the moment. proper solution coming.
> discussed with deraadt mk jason (henning@)
~ autoconf.c
> formatting (jason@)
~ autoconf.c
> - if the "upa-portid" property doesn't exist, try "portid" (schizo)
> - hack enough of autoconf to get psycho and schizo working with
> device_register
> (autoconf appears to work now for schizo machines). (jason@)
~ autoconf.c
> Can't call directly back into the prom if we "exit" at the root
> device prompt. We need to call shutdownhooks which will set the sab
> back into the mode the prom likes (I hate this chip) (jason@)
~ vm_machdep.c
> Remove dead extern declarations. (miod@)
~ autoconf.c
> rework the schizo attach to match the upaid and paddr regs
> (/pci at 80000/, 8 == upaid, 600000 == paddr of registers)
> Also, match the mpi (LSILogic,sas) (jason@)
~ autoconf.c
> how about we -correctly- match the schizo. When I get home, I'm going
> to make this code cleaner, but I'm bloody tired of typing the root
> device name (jason@)
~ autoconf.c
> match SUNW,qlc->isp (one step away from root on isp/fibre channel on US3)
> (jason@)
~ locore.s
> Use symbolic constants for TLB demapping operations.
> This fixes a few cases where the hardcoded constant indicated a reserved
> operation instead of the intended secundary context flush.
> ok jason@ (kettenis@)
~ locore.s
> The bright people at Sun removed the secondary context from the IMMU in the
> UltraSPARC III. Hack around it by temporarily switching to nucleus context
> when flushing TLBs. Put this hack under HORRID_III_HACK for now.
> Compiling with HORRID_III_HACK results in a kernel that runs on the Blade
> 2000 "at the speed of a 140MHz UltraSPARC I".
> ok jason@ (kettenis@)
~ autoconf.c
> extend the bootpath values to 64bits, and deal with fallout
> extend str2hex to handle long's, and teach it that strings starting with
> 'w',
> consisting of hex digits are really just hex digits
> match fp to scsibus (fiber port)
> match fiber channel portwwn/lun to it's bootpath variables
> result: with a small change to isp(mailed to mjacob) we can correctly
> autoconf with a boot path of:
> /pci at 80000/SUNW,qlc at 4@0,0/disk at w21000004cf948498a/bsd
> ie. schizo/ isp/ scsisbus/disk with port WWN of 21...98, lun 0,
> partition a (jason@)
~ autoconf.c
> oops, remove debugging cruft (jason@)
~ pmap.c
> Remove unused local variable from debug code. (kettenis@)
~ locore.s
> The I-cache seems to work fine on the UltraSPARC III, so never disable it,
> even if HORRID_III_HACK is defined.
> ok jason@ (kettenis@)
~ pmap.c
> All I wanted was to fix the bogus PG_SZ macro, but this is better:
> Do not try to map the message buffer with large pages, it's just as easy to
> simply use as many 8KB pages as necessary. From NetBSD. (miod@)
arch/vax/include
~ intr.h ~ param.h
> Define meaningful values for IPL_ levels, which directly map to the spl
> operation. Will allow a vax implementation of splassert(). (miod@)
~ intr.h
> Change spl4() to spl7() to raise-only semantics. (miod@)
~ intr.h
> Implement splassert() on vax. (miod@)
~ vmparam.h
> When VM_PHYSSEG_MAX is 1, use VM_PSTRAT_RANDOM for VM_PHYSSEG_STRAT, this
> makes uvm_page.o smaller. (miod@)
arch/vax/vax
~ machdep.c
> Map buffers with pmap_kenter_pa() instead of pmap_enter(). (miod@)
~ machdep.c
> Implement splassert() on vax. (miod@)
arch/zaurus/conf
~ Makefile.zaurus
> Remove dead toes. (miod@)
~ Makefile.zaurus
> Right at the end of a debug kernel build, the current Makefile copies bsd
> (with symbols) to bsd.gdb, then strips bsd. This diff makes it rename the
> file with symbols, and then asks strip to output the stripped version as
> bsd. Avoids a potentially lengthy wait when we would otherwise copy the
> large (50MB+) kernel with debug symbols.
> ok reyk@ ketennis@ (pascoe@)
arch/zaurus/include
- irqhandler.h - isa_machdep.h
> Not necessary here either. (miod@)
~ intr.h
> Ding dong, __NEWINTR is dead (miod@)
~ vmparam.h
> Shrink VM_PHYSSEG_MAX to 1 on platforms where physical memory is contiguous
> and there is only one call to uvm_page_physload(). (miod@)
arch/zaurus/zaurus
~ zaurus_machdep.c
> Map the vector page cacheable during startup. (miod@)
~ zaurus_lcd.c
> Fix name of lcd(4) in dmesg. (miod@)
conf
~ files
> Support for standard SD host controllers like the Ricoh 5C822, a small
> generic bus layer, and SCSI emulation for SD/MMC memory cards. (uwe@)
~ files
> Add basic NMEA0183 support as a tty line discipline. The line discipline
> decodes NMEA messages completely transparent to userland applications, i.e.
> userland can still use the NMEA stream and talk to the device. If data
> is received, a timedelta sensor suitable for ntpd is created. The
> timestamp
> is not very precise at the moment, use of this is experimental at best.
> "get it in or we see how well the swiss people can swim" deraadt@
> (mbalmer@)
~ files
> Add a clonable devices implementation. Hacked along with thib@, input
> from krw@ and toby@, subliminal prodding from dlg@, okay deraadt at .edro@)
crypto
~ crypto.c ~ cryptodev.c
~ cryptodev.h
> remove some silly casts. put spl calls after all declarations.
> put one splx in a better spot. make a variable size MALLOC use malloc.
> remove null test after malloc(M_WAITOK).
> add PR_NOWAIT flag to pool_get instead of 0. change callbacks to correct
> type.
> ok brad deraadt markus mickey (tedu@)
dev
~ hotplug.c
> Include device id in hotplug events. This will be used by ntpd to check
> sensors on attach/detach. hotplugd changes following in a minute.
> ok henning. (mk@)
~ hotplug.c
> Remove device id from hotplug events.
> ok henning (mk@)
~ systrace.c
> Allow len == 1 in systrace_fname(), so systraced processes can proceed
> their *stat() calls, for example, and get ENOENT, which is expected for
> "", instead of a forced EINVAL.
> Fixes the spurious 'rm: : Invalid argument' warnings when building with
> option USE_SYSTRACE, reported on ports at . > Okay provos@ sturm@ deraadt@ (pedro@)
~ bio.c
> Nuke unused local variable (ok beck deraadt) and unused struct
> bio_softc declaration (ok and testing dlg). (mk@)
~ audio.c
> Add support for the audio volume keys found on many laptops' builtin
> keyboard. These specific keys are posted to a kernel thread which will
> issue mixer commands if an audio device exists.
> Written by Alexey Vatchenko <avv , mail zp ua> with tweaks by deraadt@ and
> I. (miod@)
dev/acpi
~ dsdt.c
> Fix a memory overrun in the aml parser code.
> ok deraadt@, gwk@ (canacar@)
~ acpi.c ~ acpivar.h
~ files.acpi + acpiec.c
> Add embedded controller support.
> ok marco@ deraadt@ testing & ok gwk@ (canacar@)
~ dsdt.c
> Check the return value of aml_ederef isnt NULL before use.
> Fixes a few panics.
> ok canacar@ (gwk@)
~ dsdt.c
> Check for possible NULL dereference in aml_showvalue
> tested & ok gwk@ dim@ (canacar@)
~ dsdt.c
> When allocating a buffer, use the supplied length if the requested
> length is smaller. Tested by joris@, ok gwk@ (canacar@)
~ acpibtn.c
> Match the specs and only shut down if notify_type is 0x80
> Reported and tested by todd@, ok gwk@ (canacar@)
~ acpi.c
> do not call acpi_powerdown() unless acpi has successfully initialized
> ok gwk@ canacar@ (todd@)
~ dsdt.c
> Do not put _OS_ value on the stack.
> excellent bug report and testing by
> Berk Demir < bdd at mindcast dot org > (canacar@)
dev/cardbus
~ com_cardbus.c
> cleanup whitespaces, tabs are so much nicer
> ok mbalmer@ (fkr@)
~ com_cardbus.c
> remove the dev/isa/isavar.h include, as it is not needed
> ok miod@, martin@ (fkr@)
~ if_rl_cardbus.c
> minor space nit (pvalchev@)
~ com_cardbus.c
> add Oxford CB950 to the list of devices having more than one BAR, so
> that it can be attached properly (fkr@)
~ com_cardbus.c
> order cdevs alphabetically by vendor
> ok brad@ (fkr@)
~ com_cardbus.c
> remove a comma from the attach line, makes dmesg prettier (fkr@)
~ com_cardbus.c
> use the cardbus_conf_read/cardbus_conf_write macros instead of the
> Cardbus_conf_read/Cardbus_conf_write ones, since the latters cause
> us to panic on macppc.
> ok pascoe@, uwe@, martin@ (fkr@)
~ if_fxp_cardbus.c
> KNF and whitespace cleanup, no functional change
> ok mbalmer@ (fkr@)
dev/eisa
~ if_ep_eisa.c
> Do not reset the board, for it will not keep the interrupt configuration
> the
> BIOS configured it in; from NetBSD via Andrew Daugherity. (miod@)
dev/ic
~ mpi.c
> rename mpi_complete to mpi_poll and make it call mpi_start itself. (dlg@)
~ mpi.c ~ mpivar.h
> pass the reply from the hardware via the ccb rather than on the stack to
> the function handling the completion. this means that the completion
> function can do nothing with the reply, instead leaving it up to the
> requester to deal with. (dlg@)
~ ami.c
> Remove dead code; free commit on behalf of dlg@ (uwe@)
~ mpi.c ~ mpireg.h
> junk to read and write configuration pages. (dlg@)
~ re.c
> Only claim an interrupt if a real interrupt event occured, prevents false
> counts on shared interrupts.
> Flush the tx data buffer before handing the buffer over to the hardware,
> fixes tx on arm. ok pval@ looked at by jsg@ (drahn@)
~ mpi.c ~ mpireg.h
~ mpivar.h
> bump mpi up to using 64bit for all dva (dlg@)
~ mpi.c
> remove some if 0 code (dlg@)
~ mpi.c
> split mpi_poll out into mpi_complete, and rewrite poll to use start and
> complete to do its job. enforce the use of a timeout on polled commands
> and make the callers check if the timeout happened. (dlg@)
~ mpi.c
> put the 64bit flag in the right place on the scatter gather list. (dlg@)
~ mpireg.h
> add more mpi goo for scsi task management and port config (dlg@)
~ mpi.c ~ mpivar.h
> limit the number of scatter gather entries sent with the scsi_io commands
> so it fits in the maximum request frame size. this will do until i can
> write sgl chaining in a nice way. (dlg@)
~ lsi64854reg.h
> Convert NetBSD new-style %b strings to old-style. (miod@)
~ mpi.c
> increase the timeout on the enabling commands during attach so that sas
> controllers have a chance to respond. (dlg@)
~ mpi.c
> hush; ok dlg (jason@)
~ mpi.c
> raise the timeout on the init commands (especially port enable) to give
> the sas controller currently being used in jasons blade 2000 a chance to
> respond. (dlg@)
~ com_subr.c
> remove dev/isa/isavar.h include, not needed anymore
> ok miod (fkr@)
~ pdq_ifsubr.c
> removes the isavar.h include and the ifdef bsdi cruft
> ok miod (fkr@)
~ mpi.c ~ mpivar.h
> implement chained scatter gather lists. (dlg@)
~ mpi.c
> int32 i >> 32 is undefined, so cast to unsigned long long first, ok dlg
> (deraadt@)
~ mpi.c
> c++ style comments shouldnt be in the tree (dlg@)
~ mpi.c
> dont display the first manufacturing page (dlg@)
~ mpi.c
> mpi hardware uses an 8 bit field to describe the number of devices it has
> on a port. since 256 wont fit into 8 bits they say 0 means 256. this diff
> does the appropriate interpretation. it also avoids a divide by zero when
> we figure the openings out by dividing the number of commands the
> controller can support by the number of devices it supports.
> panic found while testing the fc controller at home. (dlg@)
~ mpi.c
> remove the fetching of the manufacturing page. it was just there to see if
> i got the page fetching right, its not really useful for anything in the
> real world. (dlg@)
~ mpi.c
> byteswap the ioc_status field so we can respond to scsi things properly on
> bigendian archs. this lets ses attach now. (dlg@)
~ isp.c
> Capture the port_wwn and node_wwn corresponding to a target id every
> time a scsi_request is processed. Crude but effective until a more
> elegant solution is found. This allows the boot drive to be found on
> fibre channel booting machines.
> ok jason@ deraadt@ (krw@)
~ mpi.c
> 64 bit dva addresses so we can >> 32 later; ok dlg (deraadt@)
~ mpi.c ~ mpivar.h
> fix the reply handling on crazy machines that give me bits in the high part
> of the address. (dlg@)
~ mpireg.h
> add spi device configuration pages 0 and 1 (dlg@)
~ mpireg.h
> oops, i dont know how to count bits. fix the OFFSET bitmasks in the device
> config pages. (dlg@)
~ rt2661.c ~ rt2661var.h
> Add powerhooks for ral(4) cards with RT2561 chipset, and also fix
> builds with RAL_DEBUG, so we have debug information now.
> ok damien@ (robert@)
~ rt2560.c
> Make the debug printfs work if RAL_DEBUG is defined. (robert@)
~ mpireg.h
> fix some macros (dlg@)
~ mpi.c ~ mpivar.h
> leave the reply_dva address alone so we can post it back to the ioc rather
> than posting back the offset of the reply frame to the start of the reply
> space.
> nobody likes panics, even if theyre free. (dlg@)
~ rt2560.c ~ rt2560var.h
> add powerhooks for ral(4) cards with the RT2560 chipset;
> ok damien@ (robert@)
~ aic6360var.h
> ditch unused struct isadev and in turn the isavar.h include (martin@)
~ aic6360.c ~ aic6360var.h
> make detachable, inspired by netbsd (martin@)
dev/isa
~ wt.c
> We do not create st devices with different density modes. i.e. with device
> minor numbers having bits 2 or 3 set.
> Eliminate the quirks, etc. used to store info on these non-existant modes.
> Also eliminate a couple of 'unimplemented' fields in the ioctl request
> structure.
> ok beck@ deraadt@ (krw@)
dev/mii
~ eephy.c
> formatting (brad@)
~ eephy.c
> Add the MIIF_AUTOTSLEEP flag, needed by gigabit axe(4) devices.
> Cloned from mii_physubr.c.
> ok brad@ (pascoe@)
dev/pci
~ if_bce.c ~ if_bge.c
~ if_de.c ~ if_em.c
~ if_ixgb.c ~ if_lge.c
~ if_nfe.c ~ if_nge.c
~ if_pcn.c ~ if_sf.c
~ if_sis.c ~ if_sk.c
~ if_ste.c ~ if_stge.c
~ if_ti.c ~ if_tl.c
~ if_txp.c ~ if_vge.c
~ if_vic.c ~ if_vr.c
~ if_wb.c ~ if_xge.c
> unknown ioctl is ENOTTY not EINVAL (jason@)
~ if_bge.c ~ if_bgereg.h
~ if_lge.c ~ if_lgereg.h
~ if_nfe.c ~ if_nfereg.h
~ if_nge.c ~ if_ngereg.h
~ if_sk.c ~ if_skreg.h
~ if_stge.c ~ if_stgereg.h
~ if_ti.c ~ if_tireg.h
~ if_vge.c ~ if_vgevar.h
~ if_vic.c ~ if_vicreg.h
~ if_xge.c
> - remove ETHER_MAX_LEN_JUMBO and ETHERMTU_JUMBO.
> - use if_hardmtu for MTU ioctl handlers.
> ok reyk@ (brad@)
~ if_em.c
> use if_hardmtu for MTU ioctl handler. (brad@)
~ pcidevs
> Add a few more devices from submitted dmesgs. (jsg@)
~ pcidevs.h ~ pcidevs_data.h
> regen (jsg@)
~ if_em.c
> always set if_hardmtu. (brad@)
~ files.pci ~ pcireg.h
+ sdhc_pci.c
> Support for standard SD host controllers like the Ricoh 5C822, a small
> generic bus layer, and SCSI emulation for SD/MMC memory cards. (uwe@)
~ if_em.c
> - force the maximum receivable frame size down to 1518 bytes for
> strict alignment architectures for the time being.
> - remove the m_adj() on non-strict alignment architectures as this
> seems to resolve the Jumbo crashing issue.
> tested by a few developers. ok reyk@ (brad@)
~ if_nfe.c
> garbage collect NFE_NO_JUMBO. (brad@)
~ auich.c
> NVIDIA nForce MCP04 (some kind of nForce4) AC97 should work here
> as well going by the Linux driver.
> "fine by me" jason@ (jsg@)
~ if_wpi.c
> Do not cast a bus_addr_t back in to a pointer as these types may not be
> compatible; plus this makes a computation clearer. ok damien@ (miod@)
~ mpi_pci.c
> add the sas1064 as a supported device (dlg@)
~ wdt.c
> Use the watchdog(4) framework for watchdog services instead of rolling
> more or less the same code. This means that all our watchdog drivers
> use the framework, and that they all can be configured using sysctl
> kern.watchdog.{period,auto}, and they can all use watchdogd.
> At least for now we still support the /dev/wdt0 char device and
> userland poking using ioctl.
> Testing by < jg ! rilk , com >, many thanks!
> ok deraadt (mk@)
~ if_em.c
> fix Jumbo frames on strict alignment architectures by allocating a new mbuf
> and
> copying the Ethernet header to the new mbuf. The new mbuf is then prepended
> into the existing mbuf chain.
> From FreeBSD
> ok reyk@ pascoe@ jason@ (brad@)
~ if_devar.h
> allow TULIP_TXTIMER to fit in the softc field tulip_txtimer.
> From NetBSD (brad@)
~ mpi_pci.c
> we dont do domain validation on real scsi controllers (as opposed to sas
> and fc controllers), so mpi will be slower than mpt for io on those hbas.
> for now.
> ifdef the scsi stuff out till we make it go fast, and so we can enable mpi
> everywhere. (dlg@)
~ mpt_pci.c ~ mpi_pci.c
> allow mpi to attach to the scsi controllers again. however, mpt will attach
> to them by default, so disable mpt to get mpi to attach.
> requested by deraadt@ (dlg@)
~ mpi_pci.c
> we should support all these sas controllers (dlg@)
~ if_vic.c
> - stop counting input/output bytes in the driver as this is taken care of
> in ether_input/ether_output.
> - only count input packets when sure the packet can be received as opposed
> to counting it unconditionally, then potentially running into an error
> and then dropping the packet.
> ok reyk@ (brad@)
~ pci_map.c
> Damnit! no static functions in the kernel (jason@)
~ if_bge.c
> xcs -> xsc (brad@)
~ if_xge.c
> formatting (brad@)
~ if_xge.c
> enable hardware RX checksum offload. (brad@)
~ if_de.c
> de-static and formatting. (brad@)
~ pcireg.h ~ pci_subr.c
> add 2 subclasses, subclass storage SAS and subclass system SD Host
> Controller.
> From NetBSD (brad@)
~ if_xge.c
> These magics -> This magic (brad@)
~ if_xge.c
> do not enable promiscous mode by default. (brad@)
~ if_xge.c
> hide link up/down messages under XGE_DEBUG. (brad@)
~ if_bge.c
> Deja-vu - again I make bge stop claiming all interrupts for it's own.
> This appears to have been due to driver changes and the fact that
> the status word seems to not be cleared. linux may have the same bug.
> freebsd does clear the status word with and atomic read and clear
> ops. we instead do best effort here to check for the interrupt being
> ours by reading the status from the status block to a local copy,
> then clearing the status word once we enter the interrupt handler
> and using the local copy for checking link status
> ok jason@ (beck@)
~ pcidevs
> Add pci vendor id for Equinox systems, and several device ids for
> their multiport serial boards (SST series).
> ok deraadt@ (gwk@)
~ pcidevs.h ~ pcidevs_data.h
> regen (gwk@)
dev/pckbc
~ wskbdmap_mfii.c
> Add support for the audio volume keys found on many laptops' builtin
> keyboard. These specific keys are posted to a kernel thread which will
> issue mixer commands if an audio device exists.
> Written by Alexey Vatchenko <avv , mail zp ua> with tweaks by deraadt@ and
> I. (miod@)
dev/pcmcia
~ com_pcmcia.c
> remove the dev/isa/isavar.h include, as it is not needed
> ok miod@, martin@ (fkr@)
~ aic_pcmcia.c
> ditch unused struct isadev and in turn the isavar.h include (martin@)
~ aic_pcmcia.c
> make detachable, inspired by netbsd (martin@)
dev/puc
~ com_puc.c
> remove another useless isavar.h include
> ok miod@ (fkr@)
dev/sbus
~ bereg.h ~ qereg.h
> Convert NetBSD new-style %b strings to old-style. (miod@)
~ if_le.c
> When trying to find leaf lebuffer devices, to work around broken old PROMs,
> walk the lebuffer_cd device list, instead of the sbus children list.
> (miod@)
~ agten.c ~ be.c
~ bwtwo.c ~ cgsix.c
~ cgsixreg.h ~ cgthree.c
~ cgtwelve.c ~ cs4231.c
~ cs4231var.h ~ dma_sbus.c
~ esp_sbus.c ~ if_hme_sbus.c
~ if_le.c ~ if_le_lebuffer.c
~ if_le_ledma.c ~ isp_sbus.c
~ lebuffer.c ~ lebuffervar.h
~ mgx.c ~ qe.c
~ qec.c ~ qecvar.h
~ rfx.c ~ sbusvar.h
~ spifvar.h ~ tvtwo.c
~ vigra.c ~ xbox.c
~ xboxvar.h ~ zx.c
> sbus_establish() and the associated linked list in the sbus softc is now
> only
> used to store a per-device reset callback, for use in sbusreset(). Except
> sbusreset() has never, ever, been used since Torek's sbus code went in.
> Time to recycle those wasted bits. (miod@)
dev/sdmmc
+ files.sdmmc + sdhc.c
+ sdhcreg.h + sdhcvar.h
+ sdmmc.c + sdmmc_io.c
+ sdmmc_mem.c + sdmmc_scsi.c
+ sdmmc_scsi.h + sdmmcchip.h
+ sdmmcreg.h + sdmmcvar.h
> Support for standard SD host controllers like the Ricoh 5C822, a small
> generic bus layer, and SCSI emulation for SD/MMC memory cards. (uwe@)
~ sdhc.c ~ sdmmc.c
~ sdmmc_io.c ~ sdmmc_mem.c
~ sdmmc_scsi.c
> don't #undef SDMMC_DEBUG (uwe@)
+ Makefile + devlist2h.awk
+ sdmmcdevs
> Start a list of SD vendors and products. (uwe@)
+ sdmmcdevs.h
> regen (uwe@)
~ sdmmcdevs.h
> regen (uwe@)
~ sdmmcdevs
> The Spectec device is probably a SDW-820. Keep product names short
> and descriptive. (uwe@)
~ sdhc.c
> fewer debug messages (uwe@)
~ files.sdmmc ~ sdmmc.c
~ sdmmc_io.c ~ sdmmc_mem.c
~ sdmmc_scsi.c ~ sdmmc_scsi.h
~ sdmmcreg.h ~ sdmmcvar.h
+ sdmmc_cis.c + sdmmc_ioreg.h
> SDIO card identification (uwe@)
~ sdmmc.c
> Multiply the capacity for cards with block size >512 bytes (e.g. 2G cards).
> (uwe@)
~ sdhc.c
> Implement power hook (without looking at the capabilities yet). (uwe@)
~ sdmmcdevs.h
> regen (uwe@)
~ devlist2h.awk
> s/CID/CIS/ (uwe@)
~ sdmmcdevs
> more products from http://mmc.drzeus.cx/wiki/SDIO (uwe@)
~ sdmmcdevs
> correct product id for the SyChip device (uwe@)
~ sdmmcdevs.h
> regen (uwe@)
~ sdmmc_io.c
> - only call config_found for function number > 0
> - always display the product information in sdmmc_print() (uwe@)
~ sdmmc.c
> prettier function declaration (martin@)
dev/usb
~ udcf.c
> Change the timedelta sensor descritpion so that the first four bytes
> contain
> the NTP identifier (filled with spaces), followed by a space and the
> station
> name (which might be the same as the NTP identifier).
> All timedelta sensor have to follow this scheme as OpenNTPD needs it.
> Requested by henning. (mbalmer@)
~ ehci.c
> Make sure to honor the USBD_FORCE_SHORT_XFER flag and generate an empty
> transfer when necessary.
> From Dan Ellis via NetBSD rev 1.102 (pascoe@)
~ ehci.c
> Make sure all physical addresses are filled in the qTD even when we have
> a page offset.
> From Dan Ellis via NetBSD rev 1.103 (pascoe@)
~ uhci.c
> Only do the software part of an abort if we are dying. (pascoe@)
~ ohci.c
> Only do the software part of an abort if we are dying. (pascoe@)
~ ohci.c
> Only check USBD_FORCE_SHORT_XFER for writes.
> From NetBSD rev1.164 (pascoe@)
~ ohci.c
> Correct status value check for OHCI isoc transfer;
> the spec says that "not accessed" is 111x, not 1111.
> From NetBSD rev1.158 (pascoe@)
~ ohcireg.h
> Missed header file change in previous commit. (pascoe@)
~ ehci.c
> Eliminate variable shadowing.
> From NetBSD r1.104 (pascoe@)
~ uhci.c
> Fix a comment, remove a stray variable assignment. (pascoe@)
~ ukbdmap.c
> Regen (although this does not change anything for these keyboards). (miod@)
~ ehci.c
> Handle 0 length bulk and interrupt transfers correctly.
> Based on changes in FreeBSD rev1.47 (pascoe@)
~ FILES
> sync with reality. (pascoe@)
~ if_cdce.c
> Give up trying to communicate after 10 consecutive errors - the device
> is most likely not coming back. (pascoe@)
~ ehci.c ~ ohci.c
~ uhci.c ~ usb_subr.c
~ usbdivar.h
> Save and restore the data toggle value when a pipe to an endpoint
> is closed and then reopened. This may be necessary now that
> we no longer clear endpoint stalls every time a pipe is opened.
> Previously we could assume an initial toggle value of zero because
> a clear-stall operation resets the device's toggle state.
> Derived from work in FreeBSD.
> This is most likely to affect devices like printers, which open
> pipes for short periods of time and close them again.
> ok dlg@ (pascoe@)
~ ehci.c
> Let the EHCI hardware track the toggle state for bulk and interrupt
> transfers. This fixes some cases where the software toggle tracking
> was not doing the right thing. For example, a short transfer that
> transferred 0 bytes of the requested qTD transfer size does cause
> a toggle change, but the existing code was assuming it didn't.
> Derived from work in FreeBSD.
> Fixes at least one USB2 cdce device, which would otherwise drop
> every second packet due to incorrect toggle state.
> commit it, dlg@ (pascoe@)
~ ehci.c
> make non debug kernels compile again (dlg@)
~ if_axe.c
> Whitespace cleanup. (pascoe@)
~ if_axe.c
> KNF. (pascoe@)
~ if_axe.c
> Pass the value of IPG2 to the AX88178 in the index register. (pascoe@)
~ if_axereg.h
> Remove unused macros, improve readability. (pascoe@)
~ udcf.c
> - ntpd no longer needs the NTP identifier as part of a timedelta sensor
> description.
> - create the timedelta sensors as early as possible, but mark them invalid
> as long as there is no real data.
> - update docs accordingly (mbalmer@)
dev/wscons
~ files.wscons ~ wskbd.c
~ wsksymdef.h ~ wsksymvar.h
+ wskbd_hotkey.c
> Add support for the audio volume keys found on many laptops' builtin
> keyboard. These specific keys are posted to a kernel thread which will
> issue mixer commands if an audio device exists.
> Written by Alexey Vatchenko <avv , mail zp ua> with tweaks by deraadt@ and
> I. (miod@)
kern
~ kern_sensors.c
> unbreak, of course this needs to included sys/hotplug.h too (henning@)
~ kern_sensors.c ~ subr_autoconf.c
> Include device id in hotplug events. This will be used by ntpd to check
> sensors on attach/detach. hotplugd changes following in a minute.
> ok henning. (mk@)
~ kern_sensors.c
> need <sys/device.h> or it doesn't compile on sparc64 (jason@)
~ kern_sensors.c
> intrducing device type DV_SENSOR turned out to be a bad idea as it
> affects way more than just hotplug messages. remove DV_SENSOR and use
> DV_DULL for the moment. proper solution coming.
> discussed with deraadt mk jason (henning@)
~ vfs_subr.c
> Spacing in vfs_sysctl() (pedro@)
~ subr_autoconf.c
> more fallout from the sensors: Every single ramdisk got broken. (deraadt@)
~ kern_sensors.c ~ subr_autoconf.c
> Remove device id from hotplug events.
> ok henning (mk@)
~ kern_sysctl.c
> back out my last commit. art pointed out that sysctl cannot sleep when
> copying out to userland, so my fix was useless bloat that didnt really do
> anything. the problem instead appears to be a use after free in a driver
> rather than bad interactions with sysctl. (dlg@)
~ subr_prf.c
> extend %b support so that it will allow for arguments with more than 32
> bits. (jason@)
~ tty_conf.c + tty_nmea.c
> Add basic NMEA0183 support as a tty line discipline. The line discipline
> decodes NMEA messages completely transparent to userland applications, i.e.
> userland can still use the NMEA stream and talk to the device. If data
> is received, a timedelta sensor suitable for ntpd is created. The
> timestamp
> is not very precise at the moment, use of this is experimental at best.
> "get it in or we see how well the swiss people can swim" deraadt@
> (mbalmer@)
~ tty_nmea.c
> mbalmer says the swiss have learned how to swim.
> - fixed debug printfs
> - cope with nmea0183 or nmea 2.3
> - allow multiple receivers at the same time
> Hacked in the park in downtown...
> "commit this right away" mbalmer (ckuethe@)
~ tty_nmea.c
> Avoid buffer overflow in case of oversized input.
> ok mbalmber (ckuethe@)
~ subr_prf.c
> the bit shifted should be a type at least as wide as the comparison
> (u_quad_t) (jason@)
~ kern_rwlock.c
> remove upgrade/downgrade, they are useless.
> rework the main entry points to just use rw_test_and_set.
> rework exit paths to be more like enter paths.
> add sleepfail so more lockmgr can be replaced.
> some from art, ok sturm (tedu@)
~ vfs_subr.c ~ vfs_vnops.c
> Add a clonable devices implementation. Hacked along with thib@, input
> from krw@ and toby@, subliminal prodding from dlg@, okay deraadt at .edro@)
~ tty_nmea.c
> - ntpd no longer needs the NTP identifier as part of a timedelta sensor
> description.
> - create the timedelta sensors as early as possible, but mark them invalid
> as long as there is no real data.
> - update docs accordingly (mbalmer@)
~ kern_time.c
> Allow any user to request the current adjustment; ok henning@ deraadt@
> (otto@)
~ subr_mutex.c
> Missing RCS Id (miod@)
~ subr_extent.c
> In extent_destroy(), only remove the extent from the global extent list if
> the list is really used, i.e. if option DIAGNOSTIC || option DDB. (miod@)
miscfs/specfs
~ spec_vnops.c ~ specdev.h
+ spec_subr.c
> Add a clonable devices implementation. Hacked along with thib@, input
> from krw@ and toby@, subliminal prodding from dlg@, okay deraadt at .edro@)
~ spec_subr.c
> Add some comments in the return endpoints (pedro@)
net
~ if_bridge.h
> fix a whitespace error causing confusion (henning@)
~ if_trunk.c ~ if_trunk.h
> check if the interface is active and UP. some, but not all, network
> drivers report an active link state even if the interface is DOWN.
> this should fix trunk with various ethernet devices.
> ok brad@ (reyk@)
~ if_pfsync.c ~ if_pfsync.h
> Only preemptively increase the replay counter for outbound TDBs.
> Another ipsec failover fix from nathanael at polymorpheus dot com.
> ok hshoexer@ (mcbride@)
~ pfvar.h ~ pf_ioctl.c
> Enable adaptive timeouts by default, with adaptive.start of 60% of the
> state limit and adaptive.end of 120% of the state limit.
> Explicitly setting the adaptive timeouts will override the default,
> and it can be disabled by setting both adaptive.start and adaptive.end to
> 0.
> ok henning@ (mcbride@)
~ if_gre.c
> Fix multicast and broadcast over gre(4) tunnels. Clear the M_MCAST and
> M_BCAST
> mbuf flags so that the packet is sent out with a normal link local address
> instead of a multicast or broadcast address. OK brad@ (claudio@)
~ pfkeyv2_convert.c
> with Naddy prodding, claudio's peril sensitive glasses turning black, and
> msf@'s assistance with hexdumping, this bug was found.
> do NOT access the wrong union member (sockaddr_in) when using v6 addresses
> this fixes output errors when displaying v6 encap routes for ipsecctl,
> route,
> and recently netstat
> ok hshoexer@ markus@ claudio@ naddy@ (todd@)
~ rtsock.c
> Export the route labels via sysctl interface not only via RTM_GET.
> OK henning@ norby@ (claudio@)
~ route.c ~ route.h
> move us to multiple routing table heads.
> -instead of staticaly declaring the heads, allocate memory for them
> -add second dimension to the routing table head array, table ID
> -space saver: do not use the AF as array index directly, but have an
> AF->index translation array and only account for AFs actually attaching
> a routing table
> note that whil ethat allows for multiple routing table heads, nothing uses
> taht yet.
> tested by many, ok claudio reyk norby mcbride (henning@)
~ route.c
> add comment indictation that rtable_add must be called at splsoftnet
> (henning@)
~ route.c ~ route.h
> use the table ID as array index directly, faster in the forwarding path
> we trade higher memory consumption if the user doesn't use continous table
> IDs, but in the worst case (table IDs 0 and 255, 64bit machine) that is 2KB
> ok claudio ryan (henning@)
~ pfkeyv2_convert.c
> observed by Naddy, found by claudio, strange constructs require strange
> defines
> . fixes proto display for the non zero case
> ok claudio@ (todd@)
~ if.c ~ if.h
~ if_pfsync.c
> Introduce attributes to interface groups.
> As a first user, move the global carp(4) demotion counter
> into the interface group. Thus we have the possibility
> to define which carp interfaces are demoted together.
> Put the demotion counter into the reserved field of the carp header.
> With this, we can have carp act smarter if multiple errors occur.
> It now always takes over other carp peers, that are advertising
> with a higher demote count. As a side effect, we can also have
> group failovers without the need of running in preempt mode.
> The protocol change does not break compability with older
> implementations.
> Collaborative work with mcbride@
> OK mcbride@, henning@ (mpf@)
netinet
~ if_ether.h
> - remove ETHER_MAX_LEN_JUMBO and ETHERMTU_JUMBO.
> - use if_hardmtu for MTU ioctl handlers.
> ok reyk@ (brad@)
~ ip_ah.c ~ ip_esp.c
> Only preemptively increase the replay counter for outbound TDBs.
> Another ipsec failover fix from nathanael at polymorpheus dot com.
> ok hshoexer@ (mcbride@)
~ in.h ~ in_pcb.h
~ ip_input.c ~ ip_output.c
~ ip_var.h ~ raw_ip.c
~ udp_usrreq.c
> Make savecontrol functions more generic and use them now for raw IP too.
> Additionally add the IP_RECVIF option which returns the interface a packet
> was received on. OK markus@ norby@ (claudio@)
~ in_pcb.h
> Put ASF/IPMI port 623 into the bad dynamic udp table, because otherwise
> we will randomly choose that stupid port, which NIC's are sometimes
> programmed to eat invisibly; sthen at bootes.spacehopper.orgr5139
> (deraadt@)
~ ip_carp.c ~ ip_carp.h
> Introduce attributes to interface groups.
> As a first user, move the global carp(4) demotion counter
> into the interface group. Thus we have the possibility
> to define which carp interfaces are demoted together.
> Put the demotion counter into the reserved field of the carp header.
> With this, we can have carp act smarter if multiple errors occur.
> It now always takes over other carp peers, that are advertising
> with a higher demote count. As a side effect, we can also have
> group failovers without the need of running in preempt mode.
> The protocol change does not break compability with older
> implementations.
> Collaborative work with mcbride@
> OK mcbride@, henning@ (mpf@)
nfs
~ nfs.h ~ nfs_subs.c
~ nfs_vfsops.c
> Add support for NFS mounts to be from non-reserved ports:
> - new sysctl vfs.nfs.privport to require NFS mount requests to be on
> reserved ports when set to 1 (the default).
> - mountd now automatically sets the sysctl depending on the -n flag.
> - add mountd_flags to rc.conf to enable the -n flag at boot.
> deraadt@ ok (avsm@)
~ nfs.h ~ nfs_subs.c
~ nfs_vfsops.c
> revert vfs.nfs.privport sysctl, broke a few architectures
> requested by deraadt@ (avsm@)
ntfs
~ ntfsmount.h
> Remove unused NetBSD new-style %b strings. (miod@)
scsi
~ st.c
> Don't print tape media information during probe. This is transitory
> information that doesn't have anything to do with the device. We don't
> say 'cd drive empty or not ready."
> ok beck@ (krw@)
~ st.c
> Always allow ioctl's to work, even if there is no media loaded.
> Analogous to what sd does for the raw partition. Eliminates useless
> error messages and premature command completion. e.g. 'mt status' will
> wait for media to to finish loading.
> ok beck@ (krw@)
~ st.c
> Use bit 23 of the device minor to indicate a control mode device. A
> control mode device can be opened but does not permit any i/o.
> Currently no such devices are created by MAKEDEV. However the previous
> code used bits 0 and 1 of the minor being set to indicate a control
> device. This conflicted with other uses of device minors to indicate
> things like rewind and eject. This behaviour would result in an open
> of a newly loaded tape drive failing until the tape was successfully
> mounted as a side effect of another command like 'mt status'.
> ok beck@ (krw@)
~ st.c
> Use TEST_READY_RETRIES_TAPE instead of TEST_READY_RETRIES_DEFAULT,
> thus allowing tape drives more time to become ready.
> ok beck@ (krw@)
~ scsi_base.c
> Add PCATCH to tsleep flags when a scsi command is waiting for an opening
> or is waiting to be retried, abort it when a signal interrupts the
> tsleep().
> This fixes what was commited in version 1.92 to work properly.
> ok krw@ (beck@)
~ ch.c
> Fix open(2) semantics for ch(4) to not be completely on fresh minty crack:
> 1) Don't use SCSI_IGNORE_NOT_READY when doing test_unit ready to wait for
> the device to come up.
> 2) use SCSI_IGNORE_ILLEGAL_REQUEST when doing test_unit_ready to wait for
> the device to come up (just as is done in st(4))
> 3) add a ch