xref: /src/doc/BRANCHES

#	$NetBSD: BRANCHES,v 1.187 2007/05/01 01:57:00 garbled Exp $
#
# This file contains a list of branches that exist in the NetBSD CVS
# tree and their current state.
#
# This list is necessarily incomplete.
# 
# Within reason, developers may create branch and version tags at any
# time for any purpose.  To avoid name collisions, private tags should
# have names which begin with the developer's NetBSD login name
# followed by a - or _ character (e.g., thorpej_scsipi,
# thorpej-signal)
# 
# Any branch or version tag not listed here should be assumed to be
# private to the developer who created it.  It is inappropriate for
# anyone other than that developer to commit, move tags, or otherwise
# modify the contents of the branch.
#
# Please update this file when a new branch is ready for consumption
# by folks other than the maintainer, or when the use or status of an
# existing branch changes significantly.
#
# Format:
#   Branch:	  name of branch
#   Description:  Purpose and intention of the branch
#   Status:	  Active/Terminated/Dormant
#   Start Date:	  date first instantiated
#   End Date:	  date it was Terminated/made_Dormant, if any
#   Base Tag:	  netbsd-1-5-base, etc.
#   Maintainer:	  Somebody to blame.
#   Scope:	  Portion of the tree covered.
#   Notes:	  Various other info, perhaps explanation of special tags,
#		  who-may-commit policies, etc.
#
# There are four sections to this file:
#   Release branches
#   Individual developers' branches (Active/Dormant)
#   Individual developers' branches (Terminated), and
#   other
#
# Entries within each section should be alphabetized.
#


# Release branches:

Branch:		comdex-fall-1999
Description:	Special release branch for pre-1.5 release at
		Fall 1999 COMDEX.
Status:		Terminated
Start Date:
End Date:
Base Tag:	comdex-fall-1999-base
Maintainer:	?
Scope:		Entire tree.
Notes:

Branch:		netbsd-0-9
Description:	The NetBSD 0.9 release branch
Status:		Terminated
Start Date:
End Date:
Base Tag:	netbsd-0-9-base
Maintainer:	Release Engineering <releng@NetBSD.org>
Scope:		Entire tree.
Notes:		Subsidiary tags of: netbsd-0-9-ALPHA, netbsd-0-9-ALPHA2,
		netbsd-0-9-BETA, netbsd-0-9-RELEASE.
		Commits restricted to release engineering.

Branch:		netbsd-1-0
Description:	The NetBSD 1.0 release branch
Status:		Terminated
Start Date:
End Date:
Base Tag:	netbsd-1-0-base
Maintainer:	Release Engineering <releng@NetBSD.org>
Scope:		Entire tree.
Notes:
		Commits restricted to release engineering.

Branch:		netbsd-1-1
Description:	The NetBSD 1.1 release branch
Status:		Terminated
Start Date:
End Date:
Base Tag:	netbsd-1-1-base
Maintainer:	Release Engineering <releng@NetBSD.org>
Scope:		Entire tree.
Notes:
		Commits restricted to release engineering.

Branch:		netbsd-1-2
Description:	The NetBSD 1.2 release branch
Status:		Terminated
Start Date:
End Date:
Base Tag:	netbsd-1-2-base
Maintainer:	Release Engineering <releng@NetBSD.org>
Scope:		Entire tree.
Notes:
		Commits restricted to release engineering.

Branch:		netbsd-1-3
Description:	The NetBSD 1.3 release branch
Status:		Terminated
Start Date:
End Date:
Base Tag:	netbsd-1-3-base
Maintainer:	Release Engineering <releng@NetBSD.org>
Scope:		Entire tree.
Notes:
		Commits restricted to release engineering.

Branch:		netbsd-1-4
Description:	The NetBSD 1.4 release branch
Status:		Terminated
Start Date:	1 Apr 2000
End Date:
Base Tag:	netbsd-1-4-base
Maintainer:	1.4 Release Engineering <releng-1-4@NetBSD.org>
Scope:		Entire tree.
Notes:
		Commits restricted to release engineering.

Branch:		netbsd-1-5
Description:	The NetBSD 1.5 release branch
Status:		Terminated
Start Date:	20 Jun 2000
End Date:	26 Jan 2005
Base Tag:	netbsd-1-5-base
Maintainer:	1.5 Release Engineering <releng-1-5@NetBSD.org>
Scope:		Entire tree.
Notes:
		Commits restricted to release engineering.

Branch:		netbsd-1-6
Description:	The NetBSD 1.6 release branch
Status:		Terminated
Start Date:	22 May 2002
End Date:
Base Tag:	netbsd-1-6-base
Maintainer:	1.6 Release Engineering <releng-1-6@NetBSD.org>
Scope:		Entire tree.
Notes:
		Commits restricted to Release Engineering.

Branch:		netbsd-2
Description:    The NetBSD 2 release branch
Status:		Active
Start Date:	23 Dec 2004
End Date:
Base Tag:	netbsd-2-base
Maintainer:	2 Release Engineering <releng-2@NetBSD.org>
Scope:		Entire tree. (src + xsrc)
Notes:
		Commits restricted to Release Engineering.

Branch:		netbsd-2-0
Description:	Originally the NetBSD 2.0 release branch and now the branch 
		tracking security/critical fixes for the NetBSD 2.0 series
Status:		Active
Start Date:	28 Mar 2004 and changed for security/critical function on
       		23 Dec 2004
End Date:
Base Tag:	netbsd-2-0-base
Maintainer:	2.0 Release Engineering <releng-2-0@NetBSD.org>
Scope:		Entire tree. (src + xsrc)
Notes:		Branch was re-purposed to security/critical change function
		once 2.0 released. See netbsd-2 for current branch to track
		NetBSD 2 tree.
		Commits restricted to Release Engineering.

Branch:		netbsd-2-1
Description:	Tracking security/critical fixes for NetBSD 2.1
Status:		Active
Start Date:     26 Oct 2005 
End Date:
Base Tag:	netbsd-2-1-RELEASE
Maintainer:	2 Release Engineering <releng-2@NetBSD.org>
Scope:		Entire tree. (src + xsrc)
Notes:		Branch contains only security/critical fixes to
		NetBSD 2.1.  For new features, use netbsd-2 branch.
		Commits restricted to Release Engineering.

Branch:		netbsd-3
Description:	The NetBSD 3 release branch
Status:		Active
Start Date:	16 Mar 2005
End Date:
Base Tag:	netbsd-3-base
Maintainer:	3 Release Engineering <releng-3@NetBSD.org>
Scope:		Entire tree. (src + xsrc)
Notes:
		Commits restricted to Release Engineering.

Branch:		netbsd-4
Description:	The NetBSD 4 release branch
Status:		Active
Start Date:	8 Aug 2006
End Date:
Base Tag:	netbsd-4-base
Maintainer:	4 Release Engineering <releng-4@NetBSD.org>
Scope:		Entire tree. (src + xsrc)
Notes:
		Commits restricted to Release Engineering.

########################################################################
# Individual developers' branches (Active/Dormant):
########################################################################

Branch:		ad-audiomp
Description:	Add MP locking to the audio drivers
Status:		Active
Start Date:	28 Febuary 2007
End Date:
Base Tag:	ad-audiomp-base
Maintainer:	Andrew Doran <ad@NetBSD.org>
Scope:		kernel
Notes:		Making the audio drivers MP safe is necessary before efforts
		can be made to make the VM system MP safe. About 1/2 of the
		drivers are converted, the remaining 1/2 need more changes.

		As of import this is snapshot of work in progress and so
		may not compile; in particular the midi changes are very
		much "in progress". The audio component has been tested
		and verified to work with emuxki and auvia. There may
		be a locking issue in the ac97 code at boot.

		The are two locks per device instance, an ISR lock and
		a character device lock. The ISR lock replaces calls to
		splaudio()/splx(), and will be held across calls to device
		methods which were called at splaudio() before (e.g.
		trigger_output). The character device lock is held across
		calls to nearly all of the methods, excluding some only
		used for initializtion, e.g. get_locks. This needs to be
		documented at merge time.

		To test on a multi way x86 system, search for "mpsafe" in
		x86/x86/intr.c and change IPL_SCHED to IPL_AUDIO. Run a job
		that will acquire the ISR lock (depends on the hardware, but
		"while true; do mixerctl -a > /dev/null; done" may suffice).
		Play audio, run lockstat and verify that that there is
		contention on the ISR lock.

Branch:		bjh21-hydra
Description:	Simtec Hydra support and ARM SMP in general
Status:		Dormant
Start Date:	18 Oct 2002
End Date:
Base Tag:	bjh21-hydra-base
Maintainer:	Ben Harris <bjh21@NetBSD.org>
Scope:		sys/arch/arm sys/arch/acorn32
Notes:		This code was on a branch because it involves changes to
		cpuswitch.S which would otherwise make merging the
		nathanw_sa branch difficult.

Branch:		gmcgarry_ctxsw
Description:	Separate scheduler from context-switch code
Status:		Dormant
Start Date:	18 December 2002
End Date:
Base Tag:	gmcgarry_ctxsw_base
Maintainer:	Gregory McGarry <gmcgarry@NetBSD.org>
Scope:		kernel
Notes:		Removing knowledge of the schedular and its run queues
		from the machine-dependent context switch code is the
		first step in cleaning up the scheduler.  It is necessary
		for the newlock branch, making the scheduler SMP friendly,
		adding real-time scheduler extensions, and reducing the
		number of code paths for LWP and proc switching.

Branch:		gmcgarry_ucred
Description:	Integrate and encapsulate user credentials
Status:		Dormant
Start Date:	18 December 2002
End Date:
Base Tag:	gmcgarry_ucred_base
Maintainer:	Gregory McGarry <gmcgarry@NetBSD.org>
Scope:		kernel
Notes:		merge pcred and ucred and poolify

Branch:		kame
Description:	KAME Project
Status:		Dormant
Start Date:	28 Jun 1999
End Date:
Base Tag:
Maintainer:	Jun-ichiro itojun Hagino <itojun@NetBSD.org>
Scope:		kernel
Notes:		http://www.kame.net
		Used for "reference" purposes for early part of KAME
		integration effort.  The branch is not actively used, as I
		(itojun) use patch(1) and diff(1) for the KAME syncs.

Branch:		kent-audio2
Description:	In-kernel audio mixer
Status:		Active
Start Date:	Sun Jan 16 2005
End Date:	
Base Tag:	kent-audio2-base
Maintainer:	TAMURA Kent <kent@NetBSD.org>
Scope:		kernel (audio device drivers)
Notes:		http://mail-index.netbsd.org/tech-kern/2004/12/03/0007.html
		Please consult with the maintainer before committing
		to this branch.
		This branch is not expected to be compilable yet.

Branch:		matt-timespec
Description:	Convert the kernel to struct timespec as its primary time
		storage mechanism.
Status:		Active
Start Date:	2005-02-19
End Date:
Base Tag:	
Maintainer:	Matt Thomas <matt@NetBSD.org>
Scope:		kernel
Notes:	

Branch:		newlock
Description:	New locking primitives, based on Solaris's
Status:		Terminated
Start Date:	March 9, 2002
End Date:	Febuary 12, 2007
Base Tag:	newlock-base
Maintainer:	Jason R. Thorpe <thorpej@NetBSD.org>
Scope:		Kernel
Notes:		The goal is to replace the existing lockmgr() and
		simple_lock() based locking mechanisms with the
		primitives found in Solaris: mutexes and rwlocks.
		Superseeded by the newlock2 branch.

Branch:		newlock2
Description:	New locking primitives, based on Solaris's
Status:		Terminated
Start Date:	September 9, 2006
End Date:	Febuary 12, 2007
Base Tag:	newlock2-base
Maintainer:	Andrew Doran <ad@NetBSD.org>
Scope:		Kernel
Notes:		As per the newlock branch, however given the amount of time
		the original branch has lain dormant, a new one has been
		created.  Goals:
		- Implement new locking primitives on most-used ports.
		- Minor locking improvements.
		- Remove spinlockmgr().

Branch:	  	pkgviews
Description:  	package views, enabling multiple conflicting packages to co-exist
Status:		Active
Start Date:	22 July 2002
End Date:
Base Tag:	pkgviews
Maintainer:	Alistair Crooks <agc@NetBSD.org>
Scope:		pkgsrc, basesrc/usr.sbin/pkg_install
Notes:		A more flexible infrastructure for third-party packages by
		allowing multiple conflicting packages and versions to co-exist
		within the same tree

Branch:		ppcoea-renovation
Description:	General renovation and cleanup of all OEA PowerPC ports.
Status:		Active
Start Date:	30 April 2007
End Date:
Base Tag:	ppcoea-renovation-base
Maintainer:	Tim Rightnour <garbled@netbsd.org>
Scope:		src/sys/arch
Notes:		This branch is for a general cleanup of all OEA-based PowerPC
		ports.  The general concept is to get them all compiling, and
		sharing as much code as possible.  All developers are welcome
		to commit to this branch.

Branch:		rpaulo-netinet-merge-pcb
Description:	merge in6pcb with inpcb
Status:		Dormant
Start Date:	Wed Feb 01 2006
End Date:	
Base Tag:	rpaulo-netinet-merge-pcb-base
Maintainer:	Rui Paulo <rpaulo@NetBSD.org>
Scope:		src/sys
Notes:		No longer active due to the lack of time. If you plan to
		continue this branch, sys/netinet/in_pcb.h already contains
		the proposed structure layout (which was roughly discussed
		in the tech-net mailing list).
		I also defined INP_*() macros to mimicate FreeBSD's inpcb
		locking style, but currently they do nothing and are not
		yet called at the correct places (needs discussion about
		kernel fine-grained locking).
		To continue this branch, one needs to read all the relevant
		inpcb/in6pcb source files and change the function names,
		structure names, structure fileds according to the new
		layout (in6pcb is gone).
		Don't try to build a kernel from this branch because it won't
		work. The missing pieces required to do a complete build
		were not committed because they were incomplete and generated
		panics.
		You shouldn't need to worry about KAME syncs because
		they did most of them by now.
		After the work is done, you are required to test the branch
		(before the merge to -current) with an interop IPv6 test.
		More info at: http://www.tahi.org/

Branch:		rtr-xorg-branch
Description:	xorg & static build alterations to src/x11
Status:		Active
Start Date:	Nov 15 2004
End Date:	
Base tag:	rtr-xorg-branch-base
Maintainer:	Tyler R. Retzlaff <rtr@NetBSD.org>
Scope:		x11
		distrib/sets/lists
		share/mk/bsd.x11.mk
		share/mk/bsd.own.mk
		etc/mtree/NetBSD.dist
Notes:		Requires parts of xsrc/xfree (HEAD)
		xsrc/local, xsrc/xfree/xc/programs/Xserver/hw/netbsd
		Requires xorg (XORG-6_8_1)

Branch:		vmlocking
Description:	Make memory allocators and trap handling MP safe
Status:		Active
Start Date:	13 March 2007
End Date:
Base Tag:	vmlocking-base
Maintainer:	Andrew Doran <ad@NetBSD.org>
Scope:		kernel
Notes:
		- Replace all simplelocks with mutexes
		- Remove spinlockmgr
		- Make memory allocators MP safe
		- Make trap handling MP safe: push kernel lock acquire
		  back into pager ops.

		Only i386 is currently update.  Most per architecture
		changes need to be made in the pmap module.  LFS is not
		currently stable due interaction between the pagedaemon (?)
		and the HOLDRELE() in lfs_segment.c.  It may trash your
		filesystem..

		Most of the locks replaced are now sleep/adaptive locks or
		spin locks at IPL_VM.  Assuming audio interrupt handlers
		are made MP safe there are far fewer potential deadlocks
		against the kernel lock.

		v_interlock becomes an adaptive mutex so now acquring
		it can potentially block, similarly for other locks
		that were before simplelocks.  This needs to be audited
		carefully.

		Use vnodes in general needs to be audited, as vnodes can
		now be inspected without the kernel lock held.

		Acquiring locks in the pmap module can now block, meaning
		that the current pmap, cpu or lazy pmap status can change
		out from underneath the requestor.  This also needs to be
		audited.

		Currently, i386 has the necessary modifications, and has
		trap handling and a few calls to allocators out from under
		the kernel lock.  There are still problems that need to be
		debugged.

		Need to check that whenever a lock is initialized, it is
		also destroyed.

		Lots of testing required.

Branch:		yamt-idlelwp
Description:	idle lwp, and some changes depending on it.

		1. separate context switching and thread scheduling.
		   (cf. gmcgarry_ctxsw)
		2. implement idle lwp.
		3. clean up related MD/MI interfaces.
		4. make scheduler(s) modular.

Status:		Active
Start Date:	Fri Feb 17 2007
End Date:	
Base Tag:	yamt-idlelwp-base5
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		src/sys  (src/common is tagged but not branched)
Notes:		
		Status of ports (minimal goals):
		  alpha			[ Done,
					  UP tested, MP not tested ]
		  i386			[ Done, tested ]
		  amd64			[ Done, tested ]
		  mips			[ Done, MIPS1 tested, MIPSX needs
					  further testing + fixes ]
		  sparc			[ WIP, commited known broken ]
		  sparc64		[ WIP, commited known broken ]
		  arm			[ Done, SA1 tested. ]
		  macppc/powerpc	[ Done, OEA tested ]
		  hppa			[ WIP. SYSCALLGATE trap issues remain. ]
		  sh3			[ Done, tested ]
		  vax			[ Done, UP tested, MP untested ]

		How to adapt a port?

		- unify cpu_switch, cpu_exit, cpu_switchto into
		  a single cpu_switchto.  see cpu_switchto.9.

		- implement cpu_idle.  see cpu_idle.9.

		- add a flag argument to cpu_need_resched.
		  handle RESCHED_IMMED flag, which means "cause a preemption
		  as soon as possible."  eg. in the case of MULTIPROCESSOR,
		  send an IPI to the target cpu.

		- "curlwp == NULL" is not a valid check for idle state anymore.
		  use "(curlwp->l_flag & LW_IDLE) != 0" instead.  curlwp is
		  almost always non-NULL, with exceptions like early on boot.

		- remove idle pcbs.

		- (if MULTIPROCESSOR) tweak processor trampoline code.

		- (if MULTIPROCESSOR) call mi_cpu_attach when attaching
		  secondary processors.

		- (if schedhz!=0) call schedclock regardless of the value of
		  curlwp.

		- of course, update MD users of scheduler/switching API.

Branch:		yamt-lazymbuf
Description:	implement and evaluate lazy mapping of mbuf external data pages
		it's intended as the first step of larger changes.
			1. support operations (esp. bus_dma load) on
			   unmapped mbuf.
			2. unify mbuf and buf.
			   (or, at least some parts of them)
			3. then it should be trivial to implement
			   i/o coalescing, etc.
Status:		Active
Start Date:	Wed Jul  6 2005
End Date:	
Base Tag:	yamt-lazymbuf-base4
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		src/sys  (src/common is tagged but not branched)
Notes:		

Branch:		reinoud-bufcleanup
Description:	implement and evaluate struct buf usage cleanup strategies.
		Ideas currently in mind (preference for b):
		a)
			1. use of bio_ops per buffer.
			2. allow chaining/overloading of functions in bio_ops
			   and provide some private context.
			3. extend the bio_ops with commenly used functionality
			   like pending action counting, custom context passing,
			   last minute processing of buffer data, buffer cache
			   etc.
			4. investigate the use of the kcont()
			   continuation-passing framework for async io
			   completion notification for bufs
			...
		b)
			1. all sequencing use of B_CALL replaced by
			   device-threads, if possible generic code around BUFQ
			2. possibly use kcont() for async IO finished signalling
			3. counter of pending and queued actions requestable
			   by IOCTL interface?
			4. implement `last minute buffer fixup' calls and maybe
			   even finished I/O calls as vnode operations
			...
Status:		Active
Start Date:	Thu Mar 15 2007
End Date:
Base Tag:	reinoud-bufcleanup-base
Maintainer:	Reinoud Zandijk <reinoud@NetBSD.org>
Scope:		src/sys and src/common
Notes:		Used as expiriment sandbox and if successful hopefully one day
		to be intergrated.

Branch:		thorpej-atomic
Description:	Implementation of an atomic memory operations API modeled
		after the one in Solaris 10.
Status:		Active
Start Date:	Apr 11 2007
End Date:
Base tag:	thorpej-atomic-base
Maintainer:	Jason Thorpe <thorpej@netbsd.org>
Scope:		src/common
		src/sys
Notes:		Base atomic primitives need to be implemented for
		all of the supported architectures:

		alpha		DONE
		arm		DONE
		hppa
		i386		DONE
		ia64
		m68k		DONE (except for 68010)
		mips
		ns32k
		powerpc		DONE except for membar_*()
		powerpc64	DONE except for membar_*()
		sh3
		sparc		DONE except for membar_*()
		sparc64		DONE except for membar_*()
		vax
		x86_64		DONE

		Tested:

		alpha		NO
		arm		NO
		hppa		NO
		i386		NO
		ia64		NO
		m68k		NO
		mips		NO
		ns32k		NO
		powerpc		NO
		powerpc64	NO
		sh3		NO
		sparc		NO
		sparc64		NO
		vax		NO
		x86_64		NO

		TODO:

		all		- Need to replace mb_*() with membar_*().

				- Audit MUTEX_CAS() and RWLOCK_CAS() uses
				  to see of membar_enter() is required here.

				- Talk to ad@ -- figure out of non-atomic
				  mutex / rwlock release is actually safe.

		arm		- Consider using a RAS technique like
				  MIPS uses for _lock_cas() and the mutex
				  stubs.  Possibly do more than one atomic
				  op as a RAS.  Possibly implement mutex
				  stubs using the technique.

		m68k		- Need to do 68010 support.
		
		i386		- Need to change when x86_patch() is called
				  so we can test ncpus.

		powerpc		- Figure out if we need barriers in the
				  atomic ops themselves (hopefully not!)

		x86_64		- Need to change when x86_patch() is called
				  so we can test ncpus.

########################################################################
# Individual developers' branches (Terminated):
########################################################################

Branch:		bouyer-scsipi
Description:	Integration of atapi support
Status:		Terminated
Start Date:
End Date:
Base Tag:
Maintainer:	Manuel Bouyer <bouyer@NetBSD.org>
Scope:		kernel
Notes:		Superseded by thorpej_scsipi

Branch:		bouyer-xen2
Description:	update xen port to xen 2.0, with support for domain0 operations
Status:		Terminated
Start Date:	Dec 13 2004
End Date:	Mar 10 2005
Base Tag:	netbsd-2-base
Maintainer:	Manuel Bouyer <bouyer@NetBSD.org>
Scope:		src/sys/arch/xen
Notes:

Branch: 	chap-midi
Description:	Renovation of MIDI support
Status: 	Terminated; merged to -current
Start Date:	19 May 2006
End Date:	30 Jun 2006
Base Tag:	chap-midi-base
Maintainer:	Chapman Flack <chap@NetBSD.org>
Scope:		sys, share/man, usr.bin
Notes:

Branch:		chs-ubc
Description:	Unified Buffer Cache
Status:		Terminated
Start Date:
End Date:
Base Tag:	chs-ubc-base
Maintainer:	Chuck Silvers <chs@NetBSD.org>
Scope:		kernel
Notes:

Branch:		chs-ubc2
Description:	 Unified Buffer Cache
Status:		Terminated; merged to -current
Start Date:
End Date:
Base Tag:	chs-ubc2-base, chs-ubc2-newbase
Maintainer:	Chuck Silvers <chs@NetBSD.org>
Scope:		kernel
Notes:

Branch:		cjs-build-against-obj-dir
Description:	?
Status:		?
Start Date:
End Date:
Base Tag:
Maintainer:	Curt Sampson <cjs@NetBSD.org>
Scope:		Entire tree.
Notes:

Branch:		eeh-paddr_t
Description:	use separate types for virtual vs. physical addresses
Status:		Terminated; merged to -current
Start Date:
End Date:
Base Tag:	eeh-paddr_t-base
Maintainer:	Eduardo Horvath <eeh@NetBSD.org>
Scope:		kernel
Notes:		This is needed to support 32-bit kernels on sparc64.

Branch:		elad-kernelauth
Description:	Native NetBSD implementation of Apple's kernel authorization
		based on TN2127. This is a "clean-room" implementation to
		avoid any licensing issues.
Status:		Terminated; merged to -current
Start Date:	Tue Mar	 7 2006
End Date:	Mon May 15 2006
Base Tag:	elad-kernelauth-base
Maintainer:	Elad Efrat <elad@NetBSD.org>
Scope:		src/sys, src/share/man/man9/Makefile,
		src/share/man/man9/kauth.h, src/distrib/sets/lists/comp/mi
Notes:

Branch:		fvdl-softdep
Description:	FFS soft dependencies
Status:		Terminated; merged to -current
Start Date:
End Date:
Base Tag:	fvdl-softdep-base
Maintainer:	Frank van der Linden <fvdl@NetBSD.org>
Scope:		kernel
Notes:

Branch:		gehenna-devsw
Description:	auto-generation of devsw table
Status:		Terminated
Start Date:	16 May 2002
End Date:	6 Sep 2002
Base Tag:	gehenna-devsw-base
Maintainer:	MAEKAWA Masahide <gehenna@NetBSD.org>
Scope:		syssrc
Notes:		The goal is to get rid of the existing device switch tables in
		port-dependent conf.c and to generate them dynamically by
		config(8).

Branch:		is-newarp
Description:	ARP for non-Ethernet
Status:		Terminated; merged before NetBSD 1.3
Start Date:	1996/10/13
End Date:	1997/03/15
Base Tag:	is-newarp-base
Maintainer:	Ignatios Souvatzis <is@NetBSD.org>
Scope:		kernel
Notes:

Branch:		kenh-if-detach
Description:	Interface detach support.
Status:		?
Start Date:
End Date:
Base Tag:	kenh-if-detach-base
Maintainer:	Ken Hornstein <kenh@NetBSD.org>
Scope:		kernel
Notes:

Branch:		kent-audio1
Description:	Audio converter pipeline
Status:		Terminated; merged to -current
Start Date:	Tue Dec  7 2004
End Date:	Tue Jan 11 2005
Base Tag:	kent-audio1-base
Maintainer:	TAMURA Kent <kent@NetBSD.org>
Scope:		kernel (audio device drivers)
Notes:		http://mail-index.netbsd.org/tech-kern/2004/12/03/0007.html
		Please consult with the maintainer before committing
		to this branch.

Branch:		kqueue
Description:	kqueue kernel event notification mechanism (from FreeBSD)
Status:		Terminated; merged to -current
Start Date:	Tue Jul 10 23:08:38 EST 2001
End Date:	Tue Oct 24 09:14:14 GMT 2002
Base Tag:	kqueue-base
Maintainer:	Jaromir Dolecek <jdolecek@NetBSD.org>
Scope:		kernel, libc/sys, share/man/man9, usr.bin/fstat
Notes:		Integration of FreeBSD's kqueue(2) mechanism, with various
		local internal changes to better suit NetBSD, and
		enhancements such as looking up filters by name or
		number, and dynamic registration of "user" filters.
		The work is complete as is, but is awaiting more
		feedback from various kernel gurus before going mainline.

Branch:		ktrace-lwp
Description:	ktrace support for lwp ids.
Status:		Terminated; merged to -current
Start Date:	Wed Jul  2 15:27:30 UTC 2003
End Date:	Sun Dec 11 12:25:29 UTC 2005
Base Tag:	ktrace-lwp-base
Maintainer:	Darren Reed <darrenr@NetBSD.org>
Scope:		sys
Notes:		

Branch:		marc-pcmcia
Description:	PCMCIA support
Status:		Terminated; merged to -current
Start Date:
End Date:
Base Tag:	marc-pcmcia-base, marc-pcmcia-bp
Maintainer:	Marc Horowitz <marc@NetBSD.org>
Scope:		kernel
Notes:

Branch:		minoura_x68k_bus_h
Description:	bus_space(9) and bus_dma(9) implementation for x68k
Status:		Terminated; merged to -current
Start Date:	23 Dec 1998
End Date:	16 Mar 1999
Base Tag:	minoura_x68k_bus_h-base
Maintainer:	Minoura Makoto <minoura@NetBSD.org>
Scope:		sys/arch/x68k
Notes:

Branch:		minoura-xpg4dl
Description:	Citrus Project locale subsystem
Status:		Terminated; (partly) merged to -current
Start Date:	28 March 2000
End Date:
Base Tag:	minoura-xpg4dl-base
Maintainer:	Minoura Makoto <minoura@NetBSD.org>
Scope:		basesrc, syssrc, sharesrc
Notes:		Citrus Project is developing dynamically-loadable, codeset-
		independent locale subsystem.  Part of the implementation
		(LC_CTYPE) called XPG4DL is now integrated to -current.

Branch:		mjf-ufs-trans
Description:	filesystem transactions for ufs
Status:		Active
Start Date:	12 March 2007
End Date:
Base Tag:	mjf-ufs-trans-base
Maintainer:	Matt Fleming <mjf@NetBSD.org>
Scope:		kernel
Notes:

Branch:		mrg-vm-swap
Description:	dynamic swap allocation, swap-to-files support
Status:		Terminated; abondoned for mrg-vm-swap2
Start Date:	12 February 1997
End Date:	15 March 1997
Base Tag:	[none]
Maintainer:	Matthew R. Green <mrg@NetBSD.org>
Scope:		kernel
Notes:

Branch:		mrg-vm-swap2
Description:	dynamic swap allocation, swap-to-files support
Status:		Terminated; merged to -current
Start Date:	4 May 1997
End Date:	12 June 1997
Base Tag:	mrg-vm-swap2-base
Maintainer:	Matthew R. Green <mrg@NetBSD.org>
Scope:		kernel
Notes:

Branch:		nathanw_sa
Description:	Scheduler activations
Status:		Terminated; merged to -current
Start Date:	5 March 2001
End Date:	18 January 2003
Base Tag:	nathanw_sa_base
Maintainer:	Nathan Williams <nathanw@NetBSD.org>
Scope:		kernel, libc, libkvm, libpthread, libpthread_dbg, include, 
		bin/ps usr.bin/systat usr.bin/top usr.bin/w
		gnu/dist/toolchain/gdb gnu/usr.bin/gdb
Notes:		The idea of scheduler activations is described in the
		classic paper by Anderson et al., in ACM Transactions
		a on Computer Systems volume 10 issue 1 (1992),which
		can be found at
		http://www.acm.org/pubs/citations/journals/tocs/1992-10-1/p53-anderson/
		Many ideas in the implementation are based on the
		implementation of scheduler activations added to Mach
		3.0 and described by Davis et al. in the University of
		Washington CS tech report 92-08-93, which can be found
		at
		ftp://ftp.cs.washington.edu/tr/1992/08/UW-CSE-92-08-03.PS.Z

		My USENIX paper on the subject can be found at
		http://web.mit.edu/nathanw/www/usenix/

		Please consult with the maintainer before committing
		to this branch.
		Other tags starting with nathanw_sa reserved for
		branch management and are not guaranteed to be stable.

		Trunk tagged with "nathanw_sa_before_merge" prior to
		merging branch down to trunk.  Branch tagged with
		"nathanw_sa_end" to indicate the ending revision of
		the branch.

Branch:		perseant-lfsv2
Description:	LFS incompatible on-disk data structure changes
Status:		Terminated
Start Date:	27 June 2001
End Date:	16 July 2001
Base Tag:	perseant-lfsv2-base
Maintainer:	Konrad Schroder <perseant@NetBSD.org>
Scope:		basesrc/{sbin/*lfs,libexec/lfs_cleanerd,usr.sbin/dumplfs}, syssrc/sys/ufs
Notes:		This branch should be short-lived, I expect to merge as soon
		as the data structures required for unremove are in and tested.

Branch:		thorpej-devvp
Description:	replace dev_t with struct vnode * in devsw entry points
Status:		Terminated
Start Date:	Sep 6 2001
End Date:	???
Base Tag:	thorpej-devvp-base
Maintainer:	Jason Thorpe <thorpej@NetBSD.org>
Scope:		syssrc
Notes:		Kernels built from this branch are not expected
		to function correctly at this time.

		Ports known to compile on this branch:
			alpha
			i386

Branch:		peter-altq
Description:	New ALTQ framework:
		- sync with KAME sources
		- share mtag/name mapping functions
		- separate altq + pf attachment
Status:		Terminated; merged to -current
Start Date:	March 16, 2006
End Date:	October 12, 2006
Base Tag:	peter-altq-base
Maintainer:	Peter Postma <peter@NetBSD.org>,
		Martin Husemann <martin@NetBSD.org>,
		Matthias Scheler <tron@NetBSD.org>
Scope:		dist/pf dist/ipf distrib/sets/lists sys usr.sbin/altq
Notes:		Only the first item in the description was finished
		in this branch.

Branch:		sommerfeld_i386mp_1
Description:	Multiprocessor support for i386 using Intel MP BIOS.
Status:		Terminated; merged to -current
Start Date:	21 Feb 2000
End Date:	01 Oct 2002
Base Tag:	sommerfeld_i386mpbase_1
Maintainer:	Bill Sommerfeld <sommerfeld@NetBSD.org>
Scope:		sys/arch/i386
Notes:
		Code committed to the branch has booted to multiuser 
		on at least one system.

		Not guaranteed to compile against mainline since -current
		may have changed incompatibly with the last branch update;
		use "cvs update -D" to roll-back the rest of the tree to match.

		Please get explicit permission from the maintainer 
		before each commit, or your changes will be backed out.
		Other tags starting with sommerfeld_ are private to 
		the developer and are guaranteed to be unstable.
		Only the maintainer should move tags.

		Note that this branch uses a DIFFERENT scheme for 
		synching with the mainline.  the maintainer uses some private
		scripts for keeping them in synch; if you make commits 
		to the branch to "resynchronize" things, you'll
		probably do it wrong; instead, ask the maintainer to
		resynchronize things.

		See sys/arch/i386/TODO on the branch for a list of known
		issues with the branch.

Branch:		thorpej-mips-cache
Description:	overhaul MIPS cache support code
Status:		Terminated; merged to -current
Start Date:	Oct 23 2001
End Date:	Nov 14 2001
Base Tag:	thorpej-mips-cache-base
Scope:		syssrc
Notes:		This branch needs more debugging, testing, and more
		ports need to be converted to the new world order.

		The following processors have had cache ops
		written:

			* R2000/R3000 (cache_r3k) -- light testing,
			  needs to be beat on a lot more to make
			  sure it's working.

			* R4000/R4400 (cache_r4k) -- mhitch@NetBSD.org
			  tracked down a bug, which has been fixed.

			  Confirmed working on R4000-with-L2.

			  Confirmed working on R4400-no-L2 and
			  R4400-with-L2.

			* R4600/R5000 (cache_r5k) -- shin@NetBSD.org
			  has committed fixes to this code.

			  Confirmed working on R4600 v2 (SGI IP-22 with
			  no SysAD L2 cache).

			  Confirmed working on RM5260 (Algorithmics P-5064).

			* TX39 (cache_tx39) -- uch@NetBSD.org has
			  committed fixes to this code.  Awaiting
			  confirmation that it is working properly.

			* R5900 (cache_r5900) -- uch@NetBSD.org has
			  written this code and tested it on a
			  PlayStation 2.

		The following ports have been updated to compile (and
		use optimized-for-processor bus_dma routines).  Those
		that have been tested and work are marked with [WORKING].

			* algor		[WORKING]

			* arc		[WORKING]

			* pmax		[WORKING]

			* playstation2	[WORKING]
					Needs optimized bus_dmamap_sync().

			* hpcmips	[need verification that this is working]
					Needs optimized bus_dmamap_sync().

			* sgimips	[WORKING]
					Needs optimized bus_dmamap_sync().

			* cobalt	[need testing]

			* mipsco	[need testing]

			* newsmips	[need testing]

Branch:		thorpej_scsipi
Description:	SCSI/ATAPI midlayer rewrite.
Status:		Terminated; merged to -current
Start Date:
End Date:	Apr 25 2001
Base Tag:	thorpej_scsipi_base, thorpej_scsipi_nbase
Maintainer:	Manuel Bouyer <bouyer@NetBSD.org>,
		Jason Thorpe <thorpej@NetBSD.org>
Scope:		syssrc
Notes:		When a change from -current is pulled up please move the
		thorpej_scsipi_base tag to the revision you're syncing with.
		thorpej_scsipi_nbase is private to maintainer (only
		used when syncing the whole branch with HEAD).
		Just before the merge, syssrc has been tagged with
		thorpej_scsipi_beforemerge

Branch:		thorpej-signal
Description:	?
Status:		?
Start Date:
End Date:
Base Tag:	thorpej-signal-base
Maintainer:	Jason Thorpe <thorpej@NetBSD.org>
Scope:		kernel
Notes:

Branch:		thorpej-setroot
Description:	?
Status:		Terminated; merged to -current
Start Date:
End Date:
Base Tag:
Maintainer:	Jason Thorpe <thorpej@NetBSD.org>
Scope:		kernel
Notes:

Branch:		simonb-timecounters
Description:	Port FreeBSD timecounters to NetBSD
Status:		Terminated
Start Date:	Sat Feb 04 2006
End Date:	Wed Jun 07 2006
Base Tag:	simonb-timecounters-base
Maintainer:	Simon Burge <simonb@NetBSD.org>,
		Frank Kardel <kardel@NetBSD.org>
Scope:		sys, sbin/savecore, usr.bin/vmstat, usr.sbin/ntp
Notes:		early (partial conversion) commit to -current
		ok'ed by core@ on 20060605 for testing and further
		simplified conversion

Branch:		wrstuden-devbsize
Description:	Modify buffer cache to deal with different devices having
		different block sizes. Also DEV_BSIZE would go away.
Status:		Terminated
Start Date:	Aug or Sept 1999
End Date:	1 Apr 2000
Base tag:	wrstuden-devbsize-base
Maintainer:	Bill Studenmund <wrstuden@NetBSD.org>
Scope:		kernel
Notes:		No longer active due to lack of time, and introduction
		of UBC. In terms of UBC, the primary cache for
		file data is the VM cache, which needs to operate
		in terms of VM pages. As no existing device has native
		pages the same size as our VM pages, there already is
		a mapping going on between the VM system and the
		underlying blocks. So it would be easier and cleaner
		to adjust that to deal w/ different block sizes.
		Also with UBC, all i/o is in terms of bytes at an offset
		with a certain length, so the exact block size isn't a big
		deal.

		Any other wrstuden-devbsize tag may (and should) go away.

Branch:		yamt-km, yamt-km-doc
Description:	simplify/improve kernel memory management and bootstrap
Status:		Terminated; merged to -current
Start Date:	Tue Jan 25 2005
End Date:	Fri Apr  1 2005
Base Tag:	yamt-km-base{,2,3,4}, yamt-km-doc-base
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		yamt-km: src/sys
		yamt-km-doc: src/share/man/man9/uvm.9
Notes:		http://mail-index.NetBSD.org/tech-kern/2004/12/25/0005.html

		the followings are per-port status.
		(account names mean "tested by".)

		works enough to go to multiuser:
			alpha, i386, xen (yamt@)
			pc532 (simonb@)
			alpha, i386, mac68k, macppc, pmax, shark,
			    sparc, sparc64, sun3, x68k (chs@)
			cobalt, dreamcast, hp300, news68k, newsmips,
			    sun3x (tsutsui@)
			amiga (mhitch@)
			cats, hp700 (skrll@)
			xen (bouyer@)

		known to have the same problem as trunk:
			vax (chs@, stuck while running kvm_mkdb)

		known to have a problem:
			luna68k (and, if any, other m68k ports which use TTR)
			needs to be revisited.
			-> for luna68k, although not tested,
			   Sysmap has been moved to avoid kva conflict with
			   I/O ranges.

		buildable:
			src/sys/arch/acorn26/conf/GENERIC
			src/sys/arch/acorn32/conf/GENERIC
			src/sys/arch/algor/conf/P6032
			src/sys/arch/alpha/conf/GENERIC
			src/sys/arch/alpha/conf/GENERIC.MP
			src/sys/arch/amd64/conf/GENERIC
			src/sys/arch/amd64/conf/GENERIC.MP
			src/sys/arch/amiga/conf/GENERIC
			src/sys/arch/arc/conf/GENERIC
			src/sys/arch/atari/conf/MILAN-PCIIDE
			src/sys/arch/bebox/conf/GENERIC
			src/sys/arch/cats/conf/GENERIC
			src/sys/arch/cesfic/conf/GENERIC
			src/sys/arch/cobalt/conf/GENERIC
			src/sys/arch/dreamcast/conf/GENERIC
			src/sys/arch/evbarm/conf/TS7200
			src/sys/arch/evbppc/conf/WALNUT
			src/sys/arch/evbsh3/conf/COMPUTEXEVB
			src/sys/arch/evbsh5/conf/CAYMAN64
			src/sys/arch/hp300/conf/GENERIC
			src/sys/arch/hp700/conf/GENERIC
			src/sys/arch/hpcarm/conf/JORNADA728
			src/sys/arch/hpcmips/conf/GENERIC
			src/sys/arch/hpcsh/conf/GENERIC
			src/sys/arch/hpcsh/conf/HPW650PA
			src/sys/arch/i386/conf/GENERIC
			src/sys/arch/i386/conf/GENERIC.MP
			src/sys/arch/i386/conf/XEN0
			src/sys/arch/i386/conf/XENU
			src/sys/arch/ibmnws/conf/GENERIC
			src/sys/arch/iyonix/conf/GENERIC
			src/sys/arch/luna68k/conf/GENERIC
			src/sys/arch/mac68k/conf/GENERIC
			src/sys/arch/macppc/conf/GENERIC
			src/sys/arch/macppc/conf/GENERIC.MP
			src/sys/arch/mipsco/conf/GENERIC
			src/sys/arch/mmeye/conf/GENERIC
			src/sys/arch/mvme68k/conf/GENERIC
			src/sys/arch/netwinder/conf/GENERIC
			src/sys/arch/news68k/conf/GENERIC
			src/sys/arch/newsmips/conf/GENERIC
			src/sys/arch/next68k/conf/GENERIC
			src/sys/arch/ofppc/conf/GENERIC
			src/sys/arch/pc532/conf/GENERIC
			src/sys/arch/pmax/conf/GENERIC
			src/sys/arch/prep/conf/GENERIC
			src/sys/arch/sandpoint/conf/GENERIC
			src/sys/arch/sbmips/conf/GENERIC
			src/sys/arch/sgimips/conf/GENERIC32_IP3x
			src/sys/arch/shark/conf/GENERIC
			src/sys/arch/sparc/conf/GENERIC
			src/sys/arch/sparc/conf/GENERIC.MP
			src/sys/arch/sparc64/conf/GENERIC
			src/sys/arch/sparc64/conf/GENERIC.MP
			src/sys/arch/sun2/conf/GENERIC
			src/sys/arch/sun3/conf/GENERIC
			src/sys/arch/vax/conf/GENERIC
			src/sys/arch/vax/conf/GENERIC.MP
			src/sys/arch/x68k/conf/GENERIC

			src/sys/arch/playstation2/conf/GENERIC

Branch:		yamt-pdpolicy
Description:	play with page replacement policy
		- separate page replacement policy from the rest of kernel
		- implement alternative replacement policy
		- related read-ahead adjustment?
Status:		Terminated; merged to -current
Start Date:	Sun Mar  5 2006
End Date:	Sat Sep 16 2006
Base Tag:	yamt-pdpolicy-base9
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		src/sys  (src/common is tagged but not branched)
Notes:		

Branch:		yamt-readahead
Description:	play with file readahead
Status:		Terminated; merged to -current
Start Date:	Mon Nov 14 2005
End Date:	Wed Nov 30 2005
Base Tag:	yamt-readahead-base3
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		src/sys
Notes:		

Branch:		yamt-uio_vmspace
Description:	uio_lwp/uio_segflg -> uio_vmspace
Status:		Terminated; merged to -current
Start Date:	Mon Dec 31 2005
End Date:	Wed Mar  1 2006
Base Tag:	yamt-uio_vmspace-base5
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		src/sys
Notes:		

Branch:		yamt-splraiseipl
Description:	finish implementing splraiseipl (and makeiplcookie).
		http://mail-index.NetBSD.org/tech-kern/2006/07/01/0000.html

		- completes workqueue(9) and netbt.
		- fix PR/33218.

Status:		Terminated; merged to -current
Start Date:	Mon Sep 18 2006
End Date:	Fri Dec 22 2006
Base Tag:	yamt-splraiseipl-base5
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		src/sys  (src/common is tagged but not branched)
Notes:		
		tested (by who):
			alpha		(yamt)
			alpha		(pavel)
			amd64		(reported privately)
			cobalt		(tsutsui)
			ews4800mips	(tsutsui)
			hp300		(tsutsui)
			hpcarm		(reported privately)
			i386		(yamt)
			macppc		(tsutsui)
			news68k		(tsutsui)
			sgimips(IP32)	(tsutsui)
			sparc(sun4c)	(tsutsui)
			sun2		(tsutsui)
			sun3		(tsutsui)
			sun3x		(tsutsui)
			xen		(yamt)

		compile-tested:
			acorn26
			acorn32
			algor
			amiga
			arc
			atari
			bebox
			cats
			cesfic
			dreamcast
			evbarm
			evbmips
			evbppc(marvell)
			evbppc(ibm4xx)
			evbsh3
			hp700
			hpcmips
			hpcsh
			ia64
			ibmnws
			iyonix
			landisk
			luna68k
			mac68k
			mmeye
			mvme68k
			mvmeppc
			mipsco
			netwinder
			newsmips
			next68k
			ofppc
			pc532
			pmax
			pmppc
			prep
			sandpoint
			sbmips
			shark
			sparc
			sparc64
			vax
			x68k
			zaurus

		not tested:
			pdp10 (i don't know how to build)
			sh5 (not buildable with an unrelated problem.)

    compile  CAYMAN64/netbsd32_sa.o
    cc1: warnings being treated as errors
    /exports/nbsd/src/sys/compat/netbsd32/netbsd32_sa.c: In function 'netbsd32_sa_ucsp':
    /exports/nbsd/src/sys/compat/netbsd32/netbsd32_sa.c:126: warning: implicit declaration of function '_UC_MACHINE32_SP'
    --- netbsd32_sa.o ---
    *** [netbsd32_sa.o] Error code 1
    1 error

		notyet:
			amigappc
			playstation2

Branch:		yamt-vop
Description:	remove several VOPs
Status:		Terminated; merged to -current
Start Date:	Wed Oct 19 2005
End Date:	Wed Nov  2 2005
Base Tag:	yamt-vop-base3
Maintainer:	YAMAMOTO Takashi <yamt@NetBSD.org>
Scope:		src/sys
Notes:		http://mail-index.NetBSD.org/tech-kern/2005/09/27/0000.html

Branch:		gdamore-uart
Description:	Add register table support to com driver.
Status:		Active
Start Date:	14 June 2006
End Date:	13 July 2006
Base Tag:	gdamore-uart-base
Maintainer:	Garrett D'Amore <gdamore@NetBSD.org>
Scope:		kernel
Notes:		Results in consolidation of Alchemy (mips/dev/alchemy/aucom.c)
		com support, and also removes a nasty hack from Atheros com.

########################################################################
# Other:
########################################################################

Branch:		magnum
Description:	?
Status:		?
Start Date:
End Date:
Base Tag:	magnum-base
Maintainer:	?
Scope:		kernel
Notes: