[odc] Daily src changes for 2007-05-28
ODC
auto at squish.net
Tue May 29 07:00:01 BST 2007
OpenBSD src changes summary for 2007-05-28
==========================================
distrib/miniroot distrib/notes
distrib/sets distrib/vax
etc/MAKEDEV.common etc/etc.alpha/MAKEDEV
etc/etc.alpha/MAKEDEV.md etc/etc.alpha/sysctl.conf
etc/etc.amd64/MAKEDEV etc/etc.amd64/MAKEDEV.md
etc/etc.armish/MAKEDEV etc/etc.armish/MAKEDEV.md
etc/etc.aviion/MAKEDEV.md etc/etc.cats/MAKEDEV
etc/etc.cats/MAKEDEV.md etc/etc.hp300/MAKEDEV
etc/etc.hp300/MAKEDEV.md etc/etc.hppa/MAKEDEV
etc/etc.hppa/MAKEDEV.md etc/etc.hppa64/MAKEDEV
etc/etc.hppa64/MAKEDEV.md etc/etc.i386/MAKEDEV
etc/etc.i386/MAKEDEV.md etc/etc.landisk/MAKEDEV
etc/etc.landisk/MAKEDEV.md etc/etc.luna88k/MAKEDEV
etc/etc.luna88k/MAKEDEV.md etc/etc.mac68k/MAKEDEV
etc/etc.mac68k/MAKEDEV.md etc/etc.mvme68k/MAKEDEV
etc/etc.mvme68k/MAKEDEV.md etc/etc.mvme88k/MAKEDEV
etc/etc.mvme88k/MAKEDEV.md etc/etc.mvmeppc/MAKEDEV.md
etc/etc.sgi/MAKEDEV etc/etc.sgi/MAKEDEV.md
etc/etc.sparc/MAKEDEV etc/etc.sparc/MAKEDEV.md
etc/etc.sparc64/MAKEDEV etc/etc.sparc64/MAKEDEV.md
etc/etc.vax/MAKEDEV etc/etc.vax/MAKEDEV.md
etc/etc.zaurus/MAKEDEV etc/etc.zaurus/MAKEDEV.md
lib/libc sbin/bioctl
share/man sys/adosfs
sys/altq sys/arch/alpha/alpha
sys/arch/alpha/conf sys/arch/alpha/include
sys/arch/amd64/amd64 sys/arch/amd64/conf
sys/arch/amd64/include sys/arch/arm/arm
sys/arch/arm/include sys/arch/aviion/aviion
sys/arch/hp300/hp300 sys/arch/hppa/hppa
sys/arch/hppa64/hppa64 sys/arch/hppa64/include
sys/arch/i386/conf sys/arch/i386/i386
sys/arch/i386/include sys/arch/landisk/landisk
sys/arch/luna88k/luna88k sys/arch/m68k/include
sys/arch/m88k/include sys/arch/mac68k/mac68k
sys/arch/macppc/conf sys/arch/macppc/stand
sys/arch/mips64/include sys/arch/mvme68k/dev
sys/arch/mvme68k/mvme68k sys/arch/mvme88k/mvme88k
sys/arch/mvmeppc/mvmeppc sys/arch/powerpc/include
sys/arch/sgi/sgi sys/arch/sh/include
sys/arch/solbourne/include sys/arch/sparc/include
sys/arch/sparc/sparc sys/arch/sparc64/conf
sys/arch/sparc64/dev sys/arch/sparc64/include
sys/arch/sparc64/sparc64 sys/arch/sparc64/stand/ofwboot
sys/arch/vax/include sys/arch/vax/vax
sys/compat/netbsd sys/conf
sys/dev sys/dev/acpi
sys/dev/ic sys/dev/isa
sys/dev/microcode sys/dev/ofw
sys/dev/pci sys/dev/pcmcia
sys/dev/usb sys/kern
sys/net sys/netinet
sys/netinet6 sys/nfs
sys/sys sys/xfs
usr.bin/cdio usr.bin/cvs
usr.bin/mg usr.bin/rcs
usr.sbin/bgpctl usr.sbin/bgpd
usr.sbin/hoststatectl usr.sbin/hoststated
usr.sbin/pkg_add
== distrib =========================================================== 01/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/distrib
miniroot
~ install.sub
> Don't look for wt tape devices while installing. (krw@)
notes
~ i386/hardware
> Remove references to wt(4). Don't install wt.4 anymore. (krw@)
~ hppa/hardware
> sync for xl(4) (martin@)
~ macppc/hardware
> mention pcn(4) (martin@)
~ i386/hardware
> sync with i386.html; 'of course' deraadt (cnst@)
~ amd64/hardware
> major update of hardware monitoring section (cnst@)
~ sparc64/hardware
> create and populate a section on hardware monitoring sensors (cnst@)
~ alpha/hardware
> create and populate a section on hardware monitoring sensors (cnst@)
~ macppc/hardware
> create and populate a section on hardware monitoring sensors (cnst@)
sets
~ lists/man/mi
> sync (deraadt@)
~ lists/base/md.alpha ~ lists/base/md.sparc64
~ lists/comp/md.alpha ~ lists/comp/md.sparc64
> sync (deraadt@)
vax
~ install.md
> Don't look for wt tape devices while installing. (krw@)
== etc =============================================================== 02/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/etc
MAKEDEV.common
~ MAKEDEV.common
> remove wt tape drives. prodded by krw@ (todd@)
etc.alpha/MAKEDEV
~ etc.alpha/MAKEDEV
> sync (todd@)
etc.alpha/MAKEDEV.md
~ etc.alpha/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.alpha/sysctl.conf
~ etc.alpha/sysctl.conf
> zap kern.emul.netbsd=1; ok beck (jmc@)
etc.amd64/MAKEDEV
~ etc.amd64/MAKEDEV
> sync (todd@)
etc.amd64/MAKEDEV.md
~ etc.amd64/MAKEDEV.md
> remove wt tape drives. prodded by krw@ (todd@)
etc.armish/MAKEDEV
~ etc.armish/MAKEDEV
> sync (todd@)
etc.armish/MAKEDEV.md
~ etc.armish/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.aviion/MAKEDEV.md
~ etc.aviion/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.cats/MAKEDEV
~ etc.cats/MAKEDEV
> sync (todd@)
etc.cats/MAKEDEV.md
~ etc.cats/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.hp300/MAKEDEV
~ etc.hp300/MAKEDEV
> sync (todd@)
etc.hp300/MAKEDEV.md
~ etc.hp300/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.hppa/MAKEDEV
~ etc.hppa/MAKEDEV
> sync (todd@)
etc.hppa/MAKEDEV.md
~ etc.hppa/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.hppa64/MAKEDEV
~ etc.hppa64/MAKEDEV
> sync (todd@)
etc.hppa64/MAKEDEV.md
~ etc.hppa64/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.i386/MAKEDEV
~ etc.i386/MAKEDEV
> sync (todd@)
etc.i386/MAKEDEV.md
~ etc.i386/MAKEDEV.md
> remove wt tape drives. prodded by krw@ (todd@)
etc.landisk/MAKEDEV
~ etc.landisk/MAKEDEV
> sync (todd@)
etc.landisk/MAKEDEV.md
~ etc.landisk/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.luna88k/MAKEDEV
~ etc.luna88k/MAKEDEV
> sync (todd@)
etc.luna88k/MAKEDEV.md
~ etc.luna88k/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.mac68k/MAKEDEV
~ etc.mac68k/MAKEDEV
> sync (todd@)
etc.mac68k/MAKEDEV.md
~ etc.mac68k/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.mvme68k/MAKEDEV
~ etc.mvme68k/MAKEDEV
> sync (todd@)
etc.mvme68k/MAKEDEV.md
~ etc.mvme68k/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.mvme88k/MAKEDEV
~ etc.mvme88k/MAKEDEV
> sync (todd@)
etc.mvme88k/MAKEDEV.md
~ etc.mvme88k/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.mvmeppc/MAKEDEV.md
~ etc.mvmeppc/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.sgi/MAKEDEV
~ etc.sgi/MAKEDEV
> sync (todd@)
etc.sgi/MAKEDEV.md
~ etc.sgi/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.sparc/MAKEDEV
~ etc.sparc/MAKEDEV
> sync (todd@)
etc.sparc/MAKEDEV.md
~ etc.sparc/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.sparc64/MAKEDEV
~ etc.sparc64/MAKEDEV
> sync (todd@)
etc.sparc64/MAKEDEV.md
~ etc.sparc64/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.vax/MAKEDEV
~ etc.vax/MAKEDEV
> sync (todd@)
etc.vax/MAKEDEV.md
~ etc.vax/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
etc.zaurus/MAKEDEV
~ etc.zaurus/MAKEDEV
> sync (todd@)
etc.zaurus/MAKEDEV.md
~ etc.zaurus/MAKEDEV.md
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
== lib =============================================================== 03/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
libc
~ sys/nfssvc.2
> tweaks; ok thib (jmc@)
== sbin ============================================================== 04/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
bioctl
~ bioctl.c
> Make disk assembly smarter and add two qualifiers to it:
> 1) noautoassemble; when set the softraid volume will not be assembled
> during
> autoconf.
> 2) force; when set it will overwrite metadata on disk
> While writing this I ran into 3 bugs that were fixed along the way
> 1) bcopy in sr_read_meta was copying data to the wrong pointer
> 2) in sr_read_meta the wrong metadata was coppied into the chunk
> 3) sr_free_discipline was freing a pointer that wasn't malloc'd
> ok dlg (marco@)
== share ============================================================= 05/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man4/isa.4 ~ man4/mtio.4
~ man4/st.4 ~ man4/man4.i386/Makefile
~ man8/man8.amd64/MAKEDEV.8 ~ man8/man8.i386/MAKEDEV.8
> Remove references to wt(4). Don't install wt.4 anymore. (krw@)
~ man9/mbuf_tags.9
> pf does not use mbuf tags any more (henning@)
~ man8/release.8
> xenocara build no longer overwrite /etc/X11/xdm. (matthieu@)
~ man4/enc.4
> some tcpdump examples; from todd and myself (jmc@)
~ man4/re.4
> remove a duplicated section; from Janne Johansson (jmc@)
~ man4/re.4
> pval convinced me to reinstate previous. (jmc@)
- man4/man4.i386/wt.4
> Delete wt(4) man page. (krw@)
~ man8/man8.alpha/MAKEDEV.8 ~ man8/man8.amd64/MAKEDEV.8
~ man8/man8.armish/MAKEDEV.8 ~ man8/man8.cats/MAKEDEV.8
~ man8/man8.hp300/MAKEDEV.8 ~ man8/man8.hppa/MAKEDEV.8
~ man8/man8.hppa64/MAKEDEV.8 ~ man8/man8.i386/MAKEDEV.8
~ man8/man8.landisk/MAKEDEV.8 ~ man8/man8.luna88k/MAKEDEV.8
~ man8/man8.mac68k/MAKEDEV.8 ~ man8/man8.macppc/MAKEDEV.8
~ man8/man8.mvme68k/MAKEDEV.8 ~ man8/man8.mvme88k/MAKEDEV.8
~ man8/man8.sgi/MAKEDEV.8 ~ man8/man8.sparc/MAKEDEV.8
~ man8/man8.sparc64/MAKEDEV.8 ~ man8/man8.vax/MAKEDEV.8
~ man8/man8.zaurus/MAKEDEV.8
> sync (todd@)
~ man4/options.4
> COMPAT_NETBSD is no more. ok jmc@ (beck@)
== sys =============================================================== 06/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
adosfs
~ adutil.c
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
altq
~ altq_cbq.c ~ altq_hfsc.c
~ altq_priq.c ~ altq_red.c
> double pf performance.
> boring details:
> pf used to use an mbuf tag to keep track of route-to etc, altq, tags,
> routing table IDs, packets redirected to localhost etc. so each and every
> packet going through pf got an mbuf tag. mbuf tags use malloc'd memory,
> and that is knda slow.
> instead, stuff the information into the mbuf header directly.
> bridging soekris with just "pass" as ruleset went from 29 MBit/s to
> 58 MBit/s with that (before ryan's randomness fix, now it is even betterer)
> thanks to chris for the test setup!
> ok ryan ryan ckuethe reyk (henning@)
arch/alpha/alpha
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
- netbsd_machdep.c ~ genassym.cf
~ locore.s
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
arch/alpha/conf
~ GENERIC ~ files.alpha
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
arch/alpha/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
- netbsd_machdep.h
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
arch/amd64/amd64
~ conf.c
> Remove the Archive/Wangtek cartidge tape driver, wt(4). All tapes are
> st(4).
> ok tom@ deraadt@ jason@ (krw@)
arch/amd64/conf
~ files.amd64
> Remove the Archive/Wangtek cartidge tape driver, wt(4). All tapes are
> st(4).
> ok tom@ deraadt@ jason@ (krw@)
~ GENERIC
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/amd64/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/arm/arm
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/arm/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/aviion/aviion
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/hp300/hp300
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/hppa/hppa
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/hppa64/hppa64
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/hppa64/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/i386/conf
~ files.i386
> Remove the Archive/Wangtek cartidge tape driver, wt(4). All tapes are
> st(4).
> ok tom@ deraadt@ jason@ (krw@)
~ GENERIC ~ RAMDISK
~ RAMDISKB ~ RAMDISKC
~ RAMDISK_CD
> Remove #wt0 lines. (krw@)
~ GENERIC
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
~ GENERIC
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
~ GENERIC
> Revert accidental commit of acpi enabling. don't do that yet. (beck@)
arch/i386/i386
~ conf.c
> Remove the Archive/Wangtek cartidge tape driver, wt(4). All tapes are
> st(4).
> ok tom@ deraadt@ jason@ (krw@)
~ pmap.c
> Flush pmap from tlb before freeing it. Makes Core2Duo boxes more
> stable, but is not a fully-understood or final fix.
> From and ok art@ (krw@)
arch/i386/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/landisk/landisk
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/luna88k/luna88k
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/m68k/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/m88k/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/mac68k/mac68k
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/macppc/conf
~ GENERIC ~ RAMDISK
> pcn(4) works here (martin@)
arch/macppc/stand
~ alloc.c
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
arch/mips64/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/mvme68k/dev
~ sbic.c ~ ssh.c
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
arch/mvme68k/mvme68k
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/mvme88k/mvme88k
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/mvmeppc/mvmeppc
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/powerpc/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/sgi/sgi
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/sh/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/solbourne/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/sparc/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/sparc/sparc
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
arch/sparc64/conf
~ files.sparc64
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
arch/sparc64/dev
~ pci_machdep.c
> Fix comparison between signed and unsigned integer. Makes the random
> panics during autoconf disappear on the v215.
> ok dlg@, otto@ (kettenis@)
~ fd.c
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
arch/sparc64/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
- netbsd_machdep.h
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
arch/sparc64/sparc64
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
- netbsd_machdep.c ~ genassym.cf
~ locore.s
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
- sigcode_netbsd.s
> bob missed a file, zap zap (tedu@)
arch/sparc64/stand/ofwboot
~ alloc.c
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
arch/vax/include
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
arch/vax/vax
~ conf.c
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
compat/netbsd
- Makefile - files.netbsd
- netbsd_exec.c - netbsd_exec.h
- netbsd_file.c - netbsd_getcwd.c
- netbsd_getdents.c - netbsd_misc.c
- netbsd_signal.c - netbsd_signal.h
- netbsd_stat.c - netbsd_stat.h
- netbsd_syscall.h - netbsd_syscallargs.h
- netbsd_syscalls.c - netbsd_sysent.c
- netbsd_types.h - netbsd_util.h
- syscalls.conf - syscalls.master
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
conf
~ GENERIC
> bio for all archs, ok krw@ kettenis@ sounds good marco@ go for it deraadt@
> (todd@)
dev
~ biovar.h ~ softraid.c
~ softraidvar.h
> Make disk assembly smarter and add two qualifiers to it:
> 1) noautoassemble; when set the softraid volume will not be assembled
> during
> autoconf.
> 2) force; when set it will overwrite metadata on disk
> While writing this I ran into 3 bugs that were fixed along the way
> 1) bcopy in sr_read_meta was copying data to the wrong pointer
> 2) in sr_read_meta the wrong metadata was coppied into the chunk
> 3) sr_free_discipline was freing a pointer that wasn't malloc'd
> ok dlg (marco@)
dev/acpi
~ acpicpu.c
> Eliminate unsightly blank lines in dmesg.
> ok marco@ (krw@)
~ acpicpu.c
> fix build when ACPI_DEBUG is defined (robert@)
dev/ic
~ fxpvar.h
> Gently nudge up the interrupt coalescing timeout to 128 microseconds. This
> gives the firmware time to do its job. Makes my fxps go from 12000int/sec
> while moving 100Mbps to 3000int/sec
> ok beck claudio (ckuethe@)
dev/isa
~ files.isa
> Remove the Archive/Wangtek cartidge tape driver, wt(4). All tapes are
> st(4).
> ok tom@ deraadt@ jason@ (krw@)
- wt.c - wtreg.h
> Delete wt code files. The end of wt(4). (krw@)
dev/microcode
~ aic7xxx/aicasm_gram.y ~ aic7xxx/aicasm_scan.l
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
dev/ofw
~ ofdisk.c
> Art left a trailing quote as an easter egg in this file five years ago. Can
> I have a cookie? (miod@)
dev/pci
~ ips.c
> Release CCB before calling scsi_done(). Thanks dlg@ for the tip. (grange@)
~ ips.c
> Sprincle some more scsi_done(). (grange@)
~ ips.c
> Provide a separate DMA-able command frame buffer for each CCB to prevent
> a race when the new command is posting while the target DMA engine is still
> reading the previous command. Now I can finaly remove the openings=1 hack.
> Many thanks to dlg@ for analizing the problem. (grange@)
~ if_tht.c
> use memcpy when copying to/from the fifos, apparently gcc can do a better
> job of optimising it. (dlg@)
~ if_che.c
> Initial bits of if_media and mii support. (claudio@)
~ ips.c
> Better debug printf. (grange@)
~ ips.c
> Fix timeout loop in ips_poll(). (grange@)
~ if_che.c
> Add more MII glue. Still not working but I want to get this into the tree.
> (claudio@)
~ ips.c
> Start adding copperhead chipset support. Not tested at all. (grange@)
~ ips.c
> Decrease max number of commands to 128. (grange@)
~ ips.c
> Bump timeout. (grange@)
~ if_sisreg.h
> Enable interrupt holdoff on DP83816 chips. This significantly improves
> performance of devices such as the soekris under load. (ckuethe@)
~ if_nx.c ~ if_nxreg.h
> make it compile on i386 (reyk@)
~ if_xge.c
> fix C++ style comment
> from brad, ok deraadt (ckuethe@)
dev/pcmcia
~ if_malo.c ~ if_maloreg.h
~ if_malovar.h
> Add a bunch of new firmware commands to configure MAC/PHY/RF, and do so.
> Simplify command interface. (mglocker@)
dev/usb
~ if_ubt.c
> unbreak, and remove the "goto bad's".
> discussed with and ok uwe@, jsg@, grange@ (fkr@)
~ usbdevs.h ~ usbdevs_data.h
> sync (tom@)
~ usbdevs
> Vendor name shouldn't be in product string; this one was mis-spelt
> anyway.
> ok jsg@ (tom@)
kern
~ uipc_mbuf.c
> double pf performance.
> boring details:
> pf used to use an mbuf tag to keep track of route-to etc, altq, tags,
> routing table IDs, packets redirected to localhost etc. so each and every
> packet going through pf got an mbuf tag. mbuf tags use malloc'd memory,
> and that is knda slow.
> instead, stuff the information into the mbuf header directly.
> bridging soekris with just "pass" as ruleset went from 29 MBit/s to
> 58 MBit/s with that (before ryan's randomness fix, now it is even betterer)
> thanks to chris for the test setup!
> ok ryan ryan ckuethe reyk (henning@)
~ subr_pool.c
> add a pool_setipl function, which allows setting an appropriate ipl
> for splassert inside pool_get and pool_put (DIAGNOSTIC only)
> ok miod pedro thib (tedu@)
~ uipc_mbuf.c
> set a hiwat mark for mbpool. we spend quite a bit of time bouncing pages
> in and out with the very low default.
> ok dlg henning ryan (tedu@)
~ subr_pool.c
> remove time from pool header. it slows us down quite a bit, and it's
> probably a better idea to just let reclaim have the emptypages. we can
> still use the partial pages.
> this lets dlg sling many many more packets
> ok dlg henning miod pedro ryan (tedu@)
~ vfs_subr.c
> de-inline vref();
> ok pedro@ (thib@)
~ vfs_bio.c
> pool_setipl() on the bufpool, to make sure that
> every get/put is at IPL_BIO.
> ok pedro@ (thib@)
~ exec_elf.c ~ exec_conf.c
> Maintaining a broken compatibility layer for a broken OS is not a
> productive
> activity for anyone. Bye bye COMPAT_NETBSD. ok tedu@, deraadt@, and many
> others
> in the hackathon room. (beck@)
~ subr_pool.c
> some remnants of the timestamping code i missed (tedu@)
net
~ if_bridge.c ~ if_pppoe.c
~ if_spppsubr.c
> Users of arc4random() should include dev/rndvar.h directly.
> ok dlg claudio (mcbride@)
~ if.c ~ netisr.h
~ netisr_dispatch.h
> Only call add_net_randomness() once per interrupt instead of once per
> packet. If multiple packets come in on a single interrupt the times mixed
> into
> the randomness pool will be identical or predictably close anyways, and
> nanotime() is expensive.
> ok toby jason miod claudio (mcbride@)
~ if_bridge.c ~ pf.c
~ pf_norm.c ~ pfvar.h
> double pf performance.
> boring details:
> pf used to use an mbuf tag to keep track of route-to etc, altq, tags,
> routing table IDs, packets redirected to localhost etc. so each and every
> packet going through pf got an mbuf tag. mbuf tags use malloc'd memory,
> and that is knda slow.
> instead, stuff the information into the mbuf header directly.
> bridging soekris with just "pass" as ruleset went from 29 MBit/s to
> 58 MBit/s with that (before ryan's randomness fix, now it is even betterer)
> thanks to chris for the test setup!
> ok ryan ryan ckuethe reyk (henning@)
netinet
~ ip_carp.c
> Users of arc4random() should include dev/rndvar.h directly.
> ok dlg claudio (mcbride@)
~ ip_input.c ~ ipsec_input.c
~ ipsec_output.c
> double pf performance.
> boring details:
> pf used to use an mbuf tag to keep track of route-to etc, altq, tags,
> routing table IDs, packets redirected to localhost etc. so each and every
> packet going through pf got an mbuf tag. mbuf tags use malloc'd memory,
> and that is knda slow.
> instead, stuff the information into the mbuf header directly.
> bridging soekris with just "pass" as ruleset went from 29 MBit/s to
> 58 MBit/s with that (before ryan's randomness fix, now it is even betterer)
> thanks to chris for the test setup!
> ok ryan ryan ckuethe reyk (henning@)
netinet6
~ ip6_forward.c ~ ip6_input.c
> double pf performance.
> boring details:
> pf used to use an mbuf tag to keep track of route-to etc, altq, tags,
> routing table IDs, packets redirected to localhost etc. so each and every
> packet going through pf got an mbuf tag. mbuf tags use malloc'd memory,
> and that is knda slow.
> instead, stuff the information into the mbuf header directly.
> bridging soekris with just "pass" as ruleset went from 29 MBit/s to
> 58 MBit/s with that (before ryan's randomness fix, now it is even betterer)
> thanks to chris for the test setup!
> ok ryan ryan ckuethe reyk (henning@)
~ nd6.c ~ nd6_rtr.c
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
~ nd6_rtr.c
> typo slipped in, unbreak. (pyr@)
nfs
~ nfs_syscalls.c
> Fix tree breakage. Variables only used inside #ifdef NFSSERVER should
> be defined inside same. ramdisk_cd (has nfs, but is not a server)
> compiles again. (krw@)
~ nfs_node.c
> lockmgr -> rwlock for the nfs_hashlock.
> ok tedu@ (thib@)
~ nfs_kq.c
> zap annoying __KERNEL_RCSID().
> ok krw@ (thib@)
sys
~ mbuf.h
> double pf performance.
> boring details:
> pf used to use an mbuf tag to keep track of route-to etc, altq, tags,
> routing table IDs, packets redirected to localhost etc. so each and every
> packet going through pf got an mbuf tag. mbuf tags use malloc'd memory,
> and that is knda slow.
> instead, stuff the information into the mbuf header directly.
> bridging soekris with just "pass" as ruleset went from 29 MBit/s to
> 58 MBit/s with that (before ryan's randomness fix, now it is even betterer)
> thanks to chris for the test setup!
> ok ryan ryan ckuethe reyk (henning@)
~ pool.h
> add a pool_setipl function, which allows setting an appropriate ipl
> for splassert inside pool_get and pool_put (DIAGNOSTIC only)
> ok miod pedro thib (tedu@)
~ buf.h
> cluster_read() is gone, so remove its prototype, spotted by mickey@
> okay tedu@ (pedro@)
~ param.h
> Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
> mbuf constants from MD param.h to MI param.h.
> Besides being the same on every arch, things will
> most probly break if any arch has different values
> then the others.
> The NMBCLUSTERS constants needs to be MD though;
> ok miod@,krw@,claudio@ (thib@)
~ vnode.h
> de-inline vref();
> ok pedro@ (thib@)
xfs
~ xfs_node-bsd.c
> avoid bypassing sys/queue.h in many places in the kernel.
> many assumptions were made about the way the various list types are
> implemented.
> lots of suggestions and help from otto and miod.
> ok otto@ (pyr@)
== usr.bin =========================================================== 07/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
cdio
~ rip.c
> the device fd is global. no need to pass it between functions. (mjc@)
cvs
~ diff_internals.c
> Line up comments with the rest of the file, like in diff(1). (ray@)
~ diff_internals.c
> "in formation" -> "information"
> OK niallo@ (ray@)
~ diff_internals.c
> Some spacing syncs with diff(1). (ray@)
mg
~ buffer.c ~ def.h
~ dir.c ~ display.c
~ funmap.c ~ grep.c
> Add a global-wd-mode command, which toggles between the current
> behavior (every buffer maintains its own cwd) and the old behavior of
> one global working directory. This makes it slightly easier to hack
> on things like kernel code, where compilation, etc, are initiated from
> a different directory than you are working in.
> While here, fix setting/handling of global wd. (kjell@)
rcs
~ diff.c
> Line up comments with the rest of the file, like in diff(1). (ray@)
~ diff.c
> "in formation" -> "information"
> OK niallo@ (ray@)
~ diff.c
> Some spacing syncs with diff(1). (ray@)
== usr.sbin ========================================================== 08/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
bgpctl
~ parser.c
> parse_asnum must return 0 (no match) if word is not an as number instead
> of bitching and erroring out.
> worked so far because at the position it kicked in, only asnumor nothing
> were allowed (henning@)
~ irrfilter.c
> on tsartup, along with saying which AS number we are building filters for,
> give the output dir (henning@)
bgpd
~ bgpd.h ~ parse.y
~ printconf.c ~ rde_filter.c
> allow matching on communities using 0 in the AS part, that is in use.
> that unfortunately means we cannot use 0 for "unset".
> ok claudio (henning@)
hoststatectl
~ hoststatectl.c
> Reflect recent hoststated changes into hoststatectl
> This unbreaks the tree.
> Fixed from the depth of the OpenBSD lair. (pyr@)
hoststated
~ hoststated.h ~ parse.y
> store the configuration file's path, this will be useful when reloading.
> (pyr@)
~ hce.c ~ hoststated.c
~ hoststated.h ~ pfe.c
~ relay.c
> another small step towards hoststated reloading.
> allow purging of parts of the hoststated environment structure.
> start using this function now to only keep vital information in
> hoststated children processes.
> ok reyk@ (pyr@)
pkg_add
~ pkg_add ~ OpenBSD/Add.pm
~ OpenBSD/Delete.pm
> clean up packing-list validation, call the methods `prepare_for_addition'
> and `prepare_for_deletion', that's clearer... (espie@)
~ pkg_add
> split off handle creation (for added packages) into its own routine.
> (espie@)
~ pkg_create ~ OpenBSD/PackingList.pm
> always create packing-lists with a `contents' entry. (espie@)
~ pkg_create
> oops, can't copy CONTENTS twice (espie@)
~ OpenBSD/Add.pm ~ OpenBSD/Delete.pm
~ OpenBSD/Vstat.pm
> lift up more of the dev checking code into Vstat.
> un-special case SpecialFiles on deletion. Now that CONTENTS is part of the
> plist, it's checked as well on addition. (espie@)
~ pkg_add
> store skipupdatedeps in the set, not the plist (espie@)
~ OpenBSD/Add.pm ~ OpenBSD/Delete.pm
~ OpenBSD/Vstat.pm
> delay error reporting a bit more: try installing and deinstalling
> everything
> and report more issues. (espie@)
~ pkg_add
> move `interrupted' stuff inside state: removes one local variable, allows
> me to cut up big function into smaller ones. (espie@)
===============================================================================
More information about the odc
mailing list