[owc] Weekly src changes ending 2006-10-15
OWC
auto at squish.net
Mon Oct 16 08:08:12 BST 2006
OpenBSD src changes summary for 2006-10-08 to 2006-10-15 inclusive
==================================================================
distrib/common distrib/crunch
distrib/notes distrib/sets
etc/MAKEDEV.common etc/Makefile
etc/etc.alpha/MAKEDEV etc/etc.amd64/MAKEDEV
etc/etc.armish/MAKEDEV etc/etc.cats/MAKEDEV
etc/etc.hppa/MAKEDEV etc/etc.i386/MAKEDEV
etc/etc.macppc/MAKEDEV etc/etc.sparc64/MAKEDEV
etc/etc.zaurus/MAKEDEV etc/security
games/battlestar games/hack
gnu/usr.bin/binutils gnu/usr.bin/gcc
gnu/usr.bin/groff gnu/usr.bin/lynx
lib/csu lib/libc
lib/libcurses lib/libkvm
lib/libpthread lib/libssl
regress/lib sbin/disklabel
sbin/fdisk sbin/pfctl
sbin/route sbin/scsi
share/man share/tmac
sys/arch/alpha/alpha sys/arch/amd64/amd64
sys/arch/amd64/conf sys/arch/amd64/stand/boot
sys/arch/amd64/stand/cdboot sys/arch/amd64/stand/libsa
sys/arch/amd64/stand/pxeboot sys/arch/arm/arm
sys/arch/hppa/hppa sys/arch/hppa64/hppa64
sys/arch/i386/i386 sys/arch/i386/stand/boot
sys/arch/i386/stand/cdboot sys/arch/i386/stand/libsa
sys/arch/i386/stand/pxeboot sys/arch/landisk/dev
sys/arch/landisk/include sys/arch/landisk/landisk
sys/arch/landisk/stand sys/arch/landisk/stand/boot
sys/arch/landisk/stand/bootxx sys/arch/landisk/stand/mbr
sys/arch/macppc/dev sys/arch/macppc/macppc
sys/arch/macppc/stand sys/arch/mips64/mips64
sys/arch/mvmeppc/mvmeppc sys/arch/mvmeppc/stand/libsa
sys/arch/sh/include sys/arch/zaurus/stand/zboot
sys/compat/linux sys/conf
sys/dev/acpi sys/dev/cardbus
sys/dev/ic sys/dev/mii
sys/dev/onewire sys/dev/pci
sys/dev/raidframe sys/dev/usb
sys/dev/wscons sys/kern
sys/lib/libsa sys/net
sys/netinet sys/scsi
usr.bin/calendar usr.bin/cdio
usr.bin/cvs usr.bin/file
usr.bin/find usr.bin/gprof
usr.bin/head usr.bin/less
usr.bin/mail usr.bin/rcs
usr.bin/sed usr.bin/ssh
usr.bin/systat usr.bin/tic
usr.bin/tset usr.sbin/bind
usr.sbin/ftp-proxy usr.sbin/hostapd
usr.sbin/hotplugd usr.sbin/lpr
usr.sbin/mopd usr.sbin/ntpd
usr.sbin/pkg_add
== distrib =========================================================== 01/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/distrib
common
~ elfrdsetroot.c
> use SEEK_SET; chl at tuxfamily.orgeraadt@)
crunch
~ crunchgen/crunchgen.1 ~ crunchgen/crunchgen.c
> Add a new option to crunchgen '-O objdir-name' which allow for a object
> directory other than the previously hardcoded 'obj' to be specified.
> Allows for cross building of ramdisks (down the road). (drahn@)
~ crunchgen/crunchgen.c
> tidy usage(); (jmc@)
~ crunchgen/crunchgen.1
> tweaks; ok drahn (jmc@)
notes
~ armish/install ~ armish/xfer
> No need to tape about tape creation on armish.
> "kill" deraadt@ "probably not" drahn@ (krw@)
sets
~ lists/base/md.alpha ~ lists/base/md.amd64
~ lists/base/md.armish ~ lists/base/md.cats
~ lists/base/md.hp300 ~ lists/base/md.hppa
~ lists/base/md.hppa64 ~ lists/base/md.i386
~ lists/base/md.mac68k ~ lists/base/md.macppc
~ lists/base/md.mvme68k ~ lists/base/md.mvmeppc
~ lists/base/md.sgi ~ lists/base/md.sparc
~ lists/base/md.sparc64 ~ lists/base/md.zaurus
~ lists/comp/mi
> sync (deraadt@)
~ lists/man/mi
> sync (deraadt@)
~ lists/man/mi
> sync (deraadt@)
~ lists/man/mi
> sync (deraadt@)
== etc =============================================================== 02/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/etc
MAKEDEV.common
~ MAKEDEV.common
> 4 usb tty devices by default; ok msf@ marco@ (todd@)
Makefile
~ Makefile
> etc.landisk does not exist yet, dont try to build MAKEDEV for it yet either
> (todd@)
etc.alpha/MAKEDEV
~ etc.alpha/MAKEDEV
> sync (todd@)
etc.amd64/MAKEDEV
~ etc.amd64/MAKEDEV
> sync (todd@)
etc.armish/MAKEDEV
~ etc.armish/MAKEDEV
> sync (todd@)
etc.cats/MAKEDEV
~ etc.cats/MAKEDEV
> sync (todd@)
etc.hppa/MAKEDEV
~ etc.hppa/MAKEDEV
> sync (todd@)
etc.i386/MAKEDEV
~ etc.i386/MAKEDEV
> sync (todd@)
etc.macppc/MAKEDEV
~ etc.macppc/MAKEDEV
> sync (todd@)
etc.sparc64/MAKEDEV
~ etc.sparc64/MAKEDEV
> sync (todd@)
etc.zaurus/MAKEDEV
~ etc.zaurus/MAKEDEV
> sync (todd@)
security
~ security
> changelist files are ascii, even if they contain a few funny chars.
> Encountered by henning@; ok millert@ robert@ deraadt@ (otto@)
== games ============================================================= 03/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/games
battlestar
~ getcom.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
hack
~ hack.save.c
> use SEEK_SET; chl at tuxfamily.orgeraadt@)
== gnu =============================================================== 04/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/gnu
usr.bin/binutils
~ bfd/config.bfd ~ bfd/configure
~ bfd/configure.in ~ bfd/elf32-sh.c
~ bfd/targets.c ~ gas/configure
~ gas/configure.in ~ gas/config/tc-sh.h
~ ld/Makefile.am ~ ld/Makefile.in
~ ld/configure.tgt + ld/emulparams/shelf_obsd.sh
+ ld/emulparams/shlelf_obsd.sh
> Define superH obsd-specific emulations for ld, and use it by default.
> (miod@)
usr.bin/gcc
~ gcc/config/sh/openbsd.h ~ gcc/config/sh/t-openbsd
> Define superH obsd-specific emulations for ld, and use it by default.
> (miod@)
~ gcc/config/sh/openbsd.h
> Default to -m4. (miod@)
usr.bin/groff
~ psbb/psbb.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
usr.bin/lynx
~ src/LYStrings.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
~ src/LYMain.c
> Add check for mkdtemp(3) failure. Prevents segfaults when TMPDIR
> or TMP are set to nonexistent directories.
> Fixes PR 5247, reported by Marcus Popp <mpopp at paranoidbsd dot
> org>
> OK millert@, jaredy@, and improvement by moritz at .ay@)
== lib =============================================================== 05/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
csu
+ sh/Makefile + sh/md_init.h
+ sh/crt0.c
> New import:
> Preliminary userland bits for OpenBSD/landisk, many things coming from
libc
~ shlib_version
> system call changes (the file says "changes", leaving it purposely
> vague like that) REQUIRE a major crank. Yet it was not done for
> getcwd. Now a lot of people are going to get screwed (deraadt@)
~ sys/Makefile.inc
> MLINKS to macros described in stat(2) (deraadt@)
~ rpc/xdr_float.c ~ stdlib/strtod.c
> Teach'em about cpus in spandex (miod@)
+ arch/sh/Makefile.inc + arch/sh/SYS.h
+ arch/sh/gen/Makefile.inc + arch/sh/gen/_setjmp.S
+ arch/sh/gen/infinity.c + arch/sh/gen/isnan.c
+ arch/sh/gen/isinf.c + arch/sh/gen/fabs.c
+ arch/sh/gen/setjmp.S + arch/sh/gen/sigsetjmp.S
+ arch/sh/gen/fpsetround.c + arch/sh/gen/fpsetmask.c
+ arch/sh/gen/fpgetsticky.c + arch/sh/gen/fpgetround.c
+ arch/sh/gen/fpgetmask.c + arch/sh/gen/flt_rounds.c
+ arch/sh/gen/fpsetsticky.c + arch/sh/gen/modf.c
+ arch/sh/gen/ldexp.c + arch/sh/net/Makefile.inc
+ arch/sh/net/htonl.c + arch/sh/net/htons.c
+ arch/sh/net/ntohl.c + arch/sh/net/ntohs.c
+ arch/sh/string/Makefile.inc + arch/sh/string/bcopy.S
+ arch/sh/string/bzero.S + arch/sh/string/ffs.S
+ arch/sh/string/memcpy.S + arch/sh/string/memmove.S
+ arch/sh/string/memset.S + arch/sh/sys/Ovfork.S
+ arch/sh/sys/exect.S + arch/sh/sys/brk.S
+ arch/sh/sys/cerror.S + arch/sh/sys/fork.S
+ arch/sh/sys/sbrk.S + arch/sh/sys/rfork.S
+ arch/sh/sys/syscall.S + arch/sh/sys/sigpending.S
+ arch/sh/sys/sigreturn.S + arch/sh/sys/sigprocmask.S
+ arch/sh/sys/sigsuspend.S
> New import:
> Preliminary userland bits for OpenBSD/landisk, many things coming from
~ arch//sh/gen/Makefile.inc
> build modf.c (drahn@)
~ arch//sh/gen/Makefile.inc
> and I forgot ldexp.c (drahn@)
~ sys/read.2
> be specific about what we mean by a "slow device";
> kick started by netbsd pr 19603 from kevin p. neal.
> rillig at netbsdems to have adapted that, and i have in turn
> adapted it again;
> ok otto (jmc@)
~ compat-43/killpg.3 ~ sys/kill.2
> SIGCONT may be sent to any process with the same session ID as the caller,
> not to any process that is a descendant of the current process.
> from ryan younce (freebsd pr #58710);
> confirmed by otto and ok (jmc@)
~ stdlib/strtod.c
> reorder #if tests because apparently on amd64 cpp proceeds to blow it's
> own brains out, and then takes lint down with it. To be investigated...
> (deraadt@)
~ stdlib/strtod.c
> use __amd64__ instead, until lint is fixed (deraadt@)
~ stdlib/strtod.c
> sparc64 should be in the #ifdef (deraadt@)
~ stdlib/strtod.c
> __mips64__ handling too.. sigh (deraadt@)
~ stdio/fgets.3
> Make fgets description much clearer with respect to its exact behavior.
> OK and input from deraadt, jmc, millert. (cloder@)
~ stdio/fgets.3
> Add a CAVEAT section that describes how to properly trim a newline from
> a string, being careful to deal with 0-length strings. Help and input
> from jmc, ok deraadt (cloder@)
~ sys/stat.2
> sync with <sys/stat.h>;
> from simon barner (freebsd pr #57153);
> ok millert deraadt (jmc@)
~ sys/msgsnd.2
> freebsd -r1.21:
> Document what "msgsz" argument really denotes;
> from brian candler (freebsd pr #101145)
> Don't say that "msgsz" can be less than 0, it cannot
> as it's unsigned (POSIX has the same bug);
> from ru at freebsd > ok millert deraadt (jmc@)
~ stdio/fgets.3
> mark up `>'; (jmc@)
libcurses
~ tinfo/captoinfo.c ~ tinfo/make_keys.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
libkvm
+ kvm_sh.c
> New import:
> Preliminary userland bits for OpenBSD/landisk, many things coming from
libpthread
~ shlib_version
> libc/shlib_version (deraadt@)
+ arch/sh/uthread_machdep.h + arch/sh/_atomic_lock.c
> New import:
> Preliminary userland bits for OpenBSD/landisk, many things coming from
libssl
+ crypto/arch/sh/opensslconf.h
> Initial config for sh, copied from arm.
> ok miod@ drahn@ (jsg@)
~ crypto/Makefile
> architecture sh (landisk) is little endian. (drahn@)
== regress =========================================================== 06/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/regress
lib
~ libpthread/Makefile
> add blocked_shutdown (kurt@)
+ libpthread/blocked_shutdown/Makefile
+ libpthread/blocked_shutdown/blocked_shutdown.c
> Test shutdown() racing with other threads using the same file
> descriptor, with some of them blocking on data that will never
> arrive. (kurt@)
== sbin ============================================================== 07/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
disklabel
~ Makefile
> Should be correct for landisk. (drahn@)
~ disklabel.c ~ editor.c
> Friends don't let friends look for disklabels in Free/NetBSD MBR
> partitions since the kernel doesn't put them or look for them there
> anymore. (krw@)
~ disklabel.c
> Friends don't let friends look for disklabels in Free/NetBSD MBR
> partitions since the kernel doesn't put them or look for them there
> anymore. (krw@)
fdisk
~ Makefile
> build fdisk on landisk (drahn@)
pfctl
~ parse.y
> Allow the 'quick' keyword on an anchor. IFF there is a matching rule inside
> the anchor, terminate ruleset evaluation when stepping out of the anchor.
> This means that if you absolutely want the anchor to be terminal, you
> probably want to use a 'block all' or 'pass all' rule at the start of the
> anchor.
> ok dhartmei@ henning@ deraadt@ (mcbride@)
route
~ route.c
> Proper link state descriptions in print_rtmsg() RTM_IFINFO case.
> ok deraadt henning (mcbride@)
scsi
~ scsi.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
~ scsi.c
> Fix another potentially unsafe instance of foo[strlen(foo) - 1] = '\0'.
> OK krw (cloder@)
== share ============================================================= 08/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man4/bktr.4
> fix minor typo (ian@)
~ man4/mpi.4
> Also attach to boards with the PCI Express Fibre Channel FC949E chipset.
> "go go go" dlg@ (brad@)
~ man7/mdoc.samples.7
> landisk bits; ok mickey (jmc@)
~ man9/sensor_add.9
> state when the sensor api can be used, and that care must be taken with
> sensor_task_unregister.
> tweaks from jmc@ (dlg@)
~ man4/xge.4
> - add the Hitachi adapter
> - remove the BUGS section (brad@)
~ man5/pf.conf.5
> Document 'anchor "foo" quick'. (mcbride@)
~ man4/ip.4
> document IP_RECVTTL and IP_MINTTL (henning@)
~ man4/ip.4
> tweaks; (jmc@)
~ man4/udcf.4
> Reflect recent changes in the driver. (mbalmer@)
~ man5/pf.conf.5
> fix mark up mistake; (jmc@)
~ man9/mbuf.9
> Remove unused variable and simplify m_copym0(). Diff from bret.lambert at
> gmail.com.
> Kill another unused variable in m_devget(). Pointed out by mcbride.
> Rename all offset variables from off0 to off.
> OK markus@, deraadt@ (mpf@)
~ man9/mbuf.9
> - give smm/18.net a title
> - mark up `>' (jmc@)
~ man4/acpi.4
> Cleanup and additional verbiage in BUGS to delinate batteries that
> misbehave. (marco@)
~ man4/acpi.4
> Add language for events. (marco@)
~ man4/acpi.4
> One more tweat in authors. (marco@)
~ man4/man4.macppc/macobio.4 + man4/man4.macppc/mc.4
+ man4/man4.macppc/mesh.4
> Add manpages for mace and mesh, document them in macobio.4.
> ok jmc (gwk@)
~ man4/man4.macppc/Makefile
> I forgot to add mc.4 and mesh.4 to the Makefile.
> Spotted by and ok deraadt. (gwk@)
~ man4/acpi.4
> tweaks; (jmc@)
~ man4/Makefile ~ man4/cardbus.4
~ man4/pci.4 + man4/malo.4
> initial man page for malo.4 (deraadt@)
tmac
~ mdoc/doc-common
> landisk bits; ok mickey (jmc@)
== sys =============================================================== 09/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/alpha/alpha
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
arch/amd64/amd64
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
~ disksubr.c
> Since we no longer look in FreeBSD or NetBSD partitions for the
> disklabel, we shouldn't write the disklabel in such partitions either.
> Brings amd64, arm, i386, landisk, macppc, mvmeppc (which roll their
> own MBR search in writedisklabel()) into line with alpha, hppa,
> hppa64, mips64 (which use readdoslabel() in both reading and writing
> the disklabel). If all goes well, soon everyone will be using
> readdoslabel(). (krw@)
arch/amd64/conf
~ GENERIC
> udcf works here, so add it, ok mblamer (henning@)
~ GENERIC ~ RAMDISK_CD
> Enable bce.
> Sortof ok with grumblings deraadt@ (marco@)
arch/amd64/stand/boot
~ conf.c
> Bump versions to note behaviour change of no longer trying
> to boot from NetBSD partitions. Requested by tom at .rw@)
arch/amd64/stand/cdboot
~ conf.c
> Bump versions to note behaviour change of no longer trying
> to boot from NetBSD partitions. Requested by tom at .rw@)
arch/amd64/stand/libsa
~ biosdev.c
> Don't try to boot from a NetBSD MBR partition when no OpenBSD MBR
> partition is found. Change error message to 'no OpenBSD partition'. (krw@)
arch/amd64/stand/pxeboot
~ conf.c
> Bump versions to note behaviour change of no longer trying
> to boot from NetBSD partitions. Requested by tom at .rw@)
arch/arm/arm
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
~ disksubr.c
> Since we no longer look in FreeBSD or NetBSD partitions for the
> disklabel, we shouldn't write the disklabel in such partitions either.
> Brings amd64, arm, i386, landisk, macppc, mvmeppc (which roll their
> own MBR search in writedisklabel()) into line with alpha, hppa,
> hppa64, mips64 (which use readdoslabel() in both reading and writing
> the disklabel). If all goes well, soon everyone will be using
> readdoslabel(). (krw@)
arch/hppa/hppa
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
arch/hppa64/hppa64
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
arch/i386/i386
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
~ disksubr.c
> Since we no longer look in FreeBSD or NetBSD partitions for the
> disklabel, we shouldn't write the disklabel in such partitions either.
> Brings amd64, arm, i386, landisk, macppc, mvmeppc (which roll their
> own MBR search in writedisklabel()) into line with alpha, hppa,
> hppa64, mips64 (which use readdoslabel() in both reading and writing
> the disklabel). If all goes well, soon everyone will be using
> readdoslabel(). (krw@)
arch/i386/stand/boot
~ conf.c
> Bump versions to note behaviour change of no longer trying
> to boot from NetBSD partitions. Requested by tom at .rw@)
arch/i386/stand/cdboot
~ conf.c
> Bump versions to note behaviour change of no longer trying
> to boot from NetBSD partitions. Requested by tom at .rw@)
arch/i386/stand/libsa
~ biosdev.c
> Don't try to boot from a NetBSD MBR partition when no OpenBSD MBR
> partition is found. Change error message to 'no OpenBSD partition'. (krw@)
arch/i386/stand/pxeboot
~ conf.c
> Bump versions to note behaviour change of no longer trying
> to boot from NetBSD partitions. Requested by tom at .rw@)
arch/landisk/dev
~ obio.c
> Oops, wrong size in struct cfattach. (miod@)
arch/landisk/include
+ loadfile_machdep.h
> Minor tweaks. (miod@)
arch/landisk/landisk
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
~ disksubr.c
> Since we no longer look in FreeBSD or NetBSD partitions for the
> disklabel, we shouldn't write the disklabel in such partitions either.
> Brings amd64, arm, i386, landisk, macppc, mvmeppc (which roll their
> own MBR search in writedisklabel()) into line with alpha, hppa,
> hppa64, mips64 (which use readdoslabel() in both reading and writing
> the disklabel). If all goes well, soon everyone will be using
> readdoslabel(). (krw@)
arch/landisk/stand
~ Makefile
> Minor tweaks. (miod@)
arch/landisk/stand/boot
~ Makefile ~ getsecs.c
~ scifcons.c
> Minor tweaks. (miod@)
~ Makefile
> Build after compiler changed to hardfloat, and link against libgcc. ok
> miod@ (drahn@)
~ devs.c ~ libsa.h
~ srt0.S
> flush the cache before jumping into the just loaded kernel.
> Now it succesfully gets into the kernel 100% (drahn@)
arch/landisk/stand/bootxx
~ pbr.S
> Minor tweaks. (miod@)
~ boot1.c
> define some functions to go here, quiet warnings (drahn@)
~ Makefile
> Build after compiler changed to hardfloat, and link against libgcc. ok
> miod@ (drahn@)
arch/landisk/stand/mbr
~ mbr.8
> tweaks; ok mickey (jmc@)
arch/macppc/dev
~ if_mc.c
> Execute the Enforce in-order Execution of I/O (eieio) instruction in
> mc_dmaintr, its a little bit of voodoo, I dont understand why its necessary
> for mace versus the near identical dma isr in if_bm.c, but the system no
> longer panics under network load. (gwk@)
arch/macppc/macppc
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
~ disksubr.c
> Since we no longer look in FreeBSD or NetBSD partitions for the
> disklabel, we shouldn't write the disklabel in such partitions either.
> Brings amd64, arm, i386, landisk, macppc, mvmeppc (which roll their
> own MBR search in writedisklabel()) into line with alpha, hppa,
> hppa64, mips64 (which use readdoslabel() in both reading and writing
> the disklabel). If all goes well, soon everyone will be using
> readdoslabel(). (krw@)
arch/macppc/stand
~ Locore.c
> Typo in a comment. ok kettenis. (mbalmer@)
~ ofdev.c
> Don't try to boot from a NetBSD MBR partition when no OpenBSD MBR
> partition is found. Change error message to 'no OpenBSD partition'. (krw@)
arch/mips64/mips64
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
arch/mvmeppc/mvmeppc
~ disksubr.c
> Don't spoof a Free/NetBSD MBR partition as the 'a' partition in a
> disklabel. Fixes 'double spoofing' of these partitions as both 'a' and
> something in the 'i' to 'p' range.
> OpenBSD changed its MBR partition type to 'A6' eight years ago. As a
> backward compatibility measure NetBSD and FreeBSD MBR partitions were
> still spoofed as 'a' partitions when no OpenBSD partition could be
> found. This bit of backward compatibility is no longer required.
> 'we can get rid of it' deraadt@ (krw@)
~ disksubr.c
> Since we no longer look in FreeBSD or NetBSD partitions for the
> disklabel, we shouldn't write the disklabel in such partitions either.
> Brings amd64, arm, i386, landisk, macppc, mvmeppc (which roll their
> own MBR search in writedisklabel()) into line with alpha, hppa,
> hppa64, mips64 (which use readdoslabel() in both reading and writing
> the disklabel). If all goes well, soon everyone will be using
> readdoslabel(). (krw@)
arch/mvmeppc/stand/libsa
~ bugdev.c
> Don't try to boot from a NetBSD MBR partition when no OpenBSD MBR
> partition is found. Change error message to 'no OpenBSD partition'. (krw@)
arch/sh/include
~ ieeefp.h
> Match the SH4 FPSCR bits. (miod@)
~ ieeefp.h
> tsk, tsk (drahn@)
~ vmparam.h
> ps uses KERNBASE so define it to the correct value (pointed out by miod)
> (drahn@)
arch/zaurus/stand/zboot
~ diskprobe.c
> 'no BSD partition' -> 'no OpenBSD partition' for error message.
> ok deraadt@ (krw@)
compat/linux
~ linux_misc.c
> Pass linux_sys_waitpid() on to linux_sys_wait4() in order to properly
> check and translate options. from netbsd
> ok deraadt, kettenis (sturm@)
conf
~ files
> First stubs for a new driver for the Marvell Libertas chips.
> The card correctly loads the firmware and it is possible to send a few
> simple commands to the card but that's it. No packet are sent or received.
> Only the Netgear WG511v2 cardbus card is tested.
> A cheese fondue and a bigger amount of white wine was needed to make the
> interrupts work -- until last night the driver was uploading the firmware
> to fast and the card garbled the image without moaning. It took us a full
> day and hundreds of test kernels to figure that out.
> Joint work with mglocker@, fondue by mbalmer@
> OK mglocker@, get it in deraadt@ (claudio@)
dev/acpi
~ acpireg.h
> add a few more ACPI device ids, as found on a Acer Aspire 1360.
> From Maxim Bourmistrov <maxim at student dot chalmers dot se>
> via marco@, ok marco@ (brad@)
~ acpi.c ~ acpiac.c
~ acpibat.c ~ acpibtn.c
~ acpicpu.c ~ acpidebug.c
~ acpiec.c ~ acpitz.c
~ dsdt.c ~ acpidev.h
~ amltypes.h ~ dsdt.h
> Massive ACPI changes
> Fixed ACPI leaks
> new ACPI API
> ok marco@ (jordan@)
~ dsdt.c
> fix rcsid (marco@)
~ acpiec.c
> Remove debug print. Found by Peter Hessler. (marco@)
~ acpireg.h
> Fix typo. Found by: Janek Richter <janek at openbug dot org> (marco@)
~ dsdt.h ~ dsdt.c
> Added exportable aml_getpciaddr
> Added printing args for called functions
> ok marco@ (jordan@)
~ acpiec.c
> Fix the checks broken during conversion to new code. The Embedded
> Controller is now detected correctly. ok marco@ (canacar@)
~ dsdt.c
> Remove extra increment of bitmask. ok jordan@ (canacar@)
~ acpi.c ~ acpivar.h
> Removed old GPE handler code remnants
> ok marco@ (jordan@)
~ acpiec.c ~ dsdt.c
> Fixed buggy if-else end scope
> Fixed multi-level deref/index for thinkpads
> Some KNF
> ok marco@ (jordan@)
~ dsdt.c
> fix broken scope braces
> added debug print to gasio
> ok marco@ (jordan@)
dev/cardbus
~ if_pgt_cardbus.c
> Add powerhooks. Tested with a Netgear WG511v1 on my X40.
> ok claudio@ (mglocker@)
~ cardbus.c ~ cardbusvar.h
~ com_cardbus.c ~ ehci_cardbus.c
~ if_acx_cardbus.c ~ if_ath_cardbus.c
~ if_atw_cardbus.c ~ if_dc_cardbus.c
~ if_fxp_cardbus.c ~ if_pgt_cardbus.c
~ if_ral_cardbus.c ~ if_re_cardbus.c
~ if_rl_cardbus.c ~ if_rtw_cardbus.c
~ if_xl_cardbus.c ~ ohci_cardbus.c
~ puc_cardbus.c ~ uhci_cardbus.c
> Add a symbolic name parameter to cardbus_intr_establish().
> Count cardbus devices interrupts in cbb(4) using evcount(9).
> ok deraadt@ (grange@)
~ files.cardbus + if_malo_cardbus.c
> First stubs for a new driver for the Marvell Libertas chips.
> The card correctly loads the firmware and it is possible to send a few
> simple commands to the card but that's it. No packet are sent or received.
> Only the Netgear WG511v2 cardbus card is tested.
> A cheese fondue and a bigger amount of white wine was needed to make the
> interrupts work -- until last night the driver was uploading the firmware
> to fast and the card garbled the image without moaning. It took us a full
> day and hundreds of test kernels to figure that out.
> Joint work with mglocker@, fondue by mbalmer@
> OK mglocker@, get it in deraadt@ (claudio@)
~ if_malo_cardbus.c
> Add $OpenBSD$ tag (claudio@)
dev/ic
~ pgt.c ~ pgtvar.h
> Add powerhooks. Tested with a Netgear WG511v1 on my X40.
> ok claudio@ (mglocker@)
~ pgt.c
> Change historical leftovers (tsleep "pftres" to "pgtres"). (mglocker@)
~ pgtvar.h
> Kill unused kthread struct. (mglocker@)
~ pgt.c
> Kill unused debug flag. (mglocker@)
~ dp8390.c ~ dp8390var.h
~ ne2000.c
> Add bus_space_barrier.
> Add DELAY in some busy loops to allow the DMA engine to actually do
> something
> (neccessary on fast processors).
> From ws@ & mycroft@ NetBSD
> Testing by deraadt@ nick@, ok deraadt@ (brad@)
~ aic79xx.c
> fix a small typo (in a comment). ok kettenis. (mbalmer@)
~ acx.c
> bescon -> beacon, found by Maxim Bourmistrov, thanks. Since I am
> correcting
> typos today... (mbalmer@)
~ pgt.c
> Use correct arguments in printfs. OK mglocker@ (claudio@)
~ pgtreg.h
> Use __packed instead of __attribute__((packed)). OK mglocker@ (claudio@)
~ pgt.c
> fix a potential null deref and a memleak (both in error path).
> ok mglocker@ (damien@)
~ pgt.c
> more ether_sprintf()'s
> ok claudio@ (damien@)
~ pgt.c
> Fix kernel page fault when the device gets detached after the firmware
> failed to load.
> Spotted by bernd@, aaron@
> ok claudio@ (mglocker@)
~ ath.c
> fix the slightly different reset procedure of single chip variants.
> this should successfully initialize the ar5424 PCI-E chipset.
> figured out by kettenis@ (reyk@)
~ rtl81x9.c
> Yet another example of a non-OK'd diff by brad breaking someone's machine.
> Brad's bullshit process stops NOW. (deraadt@)
~ malo.c
> More serious firmware load error messages.
> ok claudio@ (mglocker@)
~ malo.c
> Place bus_dmamap_sync() around malo_send_cmd this should help porting to
> other archs later. OK mglocker@ (claudio@)
~ malo.c ~ malo.h
> Add $OpenBSD$ tag (claudio@)
+ malo.c + malo.h
> First stubs for a new driver for the Marvell Libertas chips.
> The card correctly loads the firmware and it is possible to send a few
> simple commands to the card but that's it. No packet are sent or received.
> Only the Netgear WG511v2 cardbus card is tested.
> A cheese fondue and a bigger amount of white wine was needed to make the
> interrupts work -- until last night the driver was uploading the firmware
> to fast and the card garbled the image without moaning. It took us a full
> day and hundreds of test kernels to figure that out.
> Joint work with mglocker@, fondue by mbalmer@
> OK mglocker@, get it in deraadt@ (claudio@)
dev/mii
~ brgphy.c
> - g/c unused code.
> - only call PHY_RESET if trying to set a valid media type. (brad@)
~ brgphy.c
> refactor the code a bit, use the bge(4) PHY flags and add another
> PHY workaround from the Linux tg3 driver. (brad@)
~ brgphy.c
> set Jumbo frame settings for Jumbo capable bge(4) adapters. (brad@)
~ brgphy.c
> revert the last two changes. the changes were causing brgphy at gem panic.
> (brad@)
~ brgphy.c
> revert rev 1.54. This causes breakage with the PHY on a 5704 A3 chipset;
> the link comes up, there is activity with tcpdump, but it doesn't work
> as expected.
> Issue reported by kettenis. ok kettenis@ (brad@)
~ miidevs.h
> regen (brad@)
~ miidevs
> add another OUI id for Broadcom and the BCM5787 PHY id. (brad@)
~ brgphy.c
> attach to the Broadcom BCM5787 PHY.
> Tested by Benjamin Black <ben at layer8 dot net> (brad@)
~ brgphy.c
> fix parent driver match in brgphy_reset(), simplify code a bit as well,
> don't
> need to bring in the bnx(4) softc.
> Tested on bge, bnx, gem, sk
> ok krw@ kettenis@ (brad@)
dev/onewire
~ onewire.c ~ onewirevar.h
> Rework bus scanning code so that most of work could be offloaded
> the hardware. And fix another bug in the search algorithm. (grange@)
dev/pci
~ mpi_pci.c
> Also attach to boards with the PCI Express Fibre Channel FC949E chipset.
> "go go go" dlg@ (brad@)
~ if_bge.c
> The mickey (/brad) jumbo allocation avoidance patch triggers the same
> ami bug "on a particular machine" as the uvm anon change that needed
> to be backed out on jul 13.
> at least now you are getting close to knowing whether this is a uvm
> bug or an ami bug or somewhere else; something starves or cuts memory
> up badly enough to cause a big problem. (deraadt@)
~ pcidevs
> add two Pericom PCI bridges. (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ if_bge.c
> revert the firmware synchronization change in rev 1.178, this is the cause
> of the first port of a dual port MAC to not work. (brad@)
~ if_xge.c ~ if_xgereg.h
> - Only use the MAC address workaround for the Xframe chipset.
> - For the Xfrme II chipset, bring EOI out of reset before XGXS.
> - Check for the adapter status bit RIC_RUNNING before trying to
> access any registers with the Xframe II chipset.
> - Add new DTX_Control values for the Xframe II chipset to
> initialize the XGXS transciever.
> - Make sure to clear IFF_RUNNING/IFF_OACTIVE in xge_stop().
> From Veena Parat @ Neterion (brad@)
~ if_xge.c
> call xge_stop() at the top of xge_init(). (brad@)
~ pcidevs
> add the NetXen vendor id. (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> add my Radeon X800
> ok brad (ckuethe@)
~ pcidevs.h ~ pcidevs_data.h
> regenerate. ok brad (ckuethe@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> add another ServerWorks HT-2100 PCIe bridge.
> dmesg from marco@ (brad@)
~ pccbb.c ~ pccbbvar.h
> Add a symbolic name parameter to cardbus_intr_establish().
> Count cardbus devices interrupts in cbb(4) using evcount(9).
> ok deraadt@ (grange@)
~ if_bnx.c ~ if_bnxreg.h
> - Simplify the arguments to bnx_tx_encap.
> - Don't copy the bd_chain head pointers into temporary objects, they are
> available globally.
> From scottl at FreeBSDrad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> use the correct PCI id for the Broadcom BCM5756 chipset. (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> add a few Intel 82Q963 (Q963) chipset ids. (brad@)
~ if_bge.c ~ if_bgereg.h
> recognize the BCM5787 A2 chipset. (brad@)
~ pcidevs
> Add some Intel 965 devices. (jsg@)
~ pcidevs.h ~ pcidevs_data.h
> regen (jsg@)
~ if_hme_pci.c
> Fix indentation. (kettenis@)
~ if_gem_pci.c
> Get MAC address from VPD, just like we already do for hme(4).
> ok brad@ (kettenis@)
~ if_gem_pci.c
> Fix last commit. (kettenis@)
dev/raidframe
~ rf_disks.c
> Get rid of another potentially unsafe pattern of foo[strlen(foo) - 1] = ...
> OK krw, deraadt (cloder@)
dev/usb
~ uowreg.h
> Add status register bits string for pretty %b printf. (grange@)
~ uow.c
> Implement bus searching. (grange@)
~ udcf.c
> Make the timedelta sensor appear right when a device is plugged-in, with a
> status of SENSOR_S_UNKNOWN. Do not set the SENSOR_FINVALID bit, sysctl
> hw.sensors does not list sensors with this flags set and we want to the
> sensor
> when the device is present.
> ok henning (mbalmer@)
~ udcf.c
> Apply some formatting. (mbalmer@)
dev/wscons
~ wsemul_vt100.c
> Fix jump scroll glitch, reported by dim@ and Paul Stoeber. (miod@)
kern
~ sched_bsd.c
> bret lambert sent a patch removing register. i made it ansi. (tedu@)
~ uipc_mbuf.c
> Remove unused variable and simplify m_copym0(). Diff from bret.lambert at
> gmail.com.
> Kill another unused variable in m_devget(). Pointed out by mcbride.
> Rename all offset variables from off0 to off.
> OK markus@, deraadt@ (mpf@)
lib/libsa
~ readdir.c
> use SEEK_SET; chl at tuxfamily.orgeraadt@)
net
~ pfvar.h ~ pf.c
> Allow the 'quick' keyword on an anchor. IFF there is a matching rule inside
> the anchor, terminate ruleset evaluation when stepping out of the anchor.
> This means that if you absolutely want the anchor to be terminal, you
> probably want to use a 'block all' or 'pass all' rule at the start of the
> anchor.
> ok dhartmei@ henning@ deraadt@ (mcbride@)
netinet
~ in.h ~ in_pcb.h
~ ip_input.c ~ ip_output.c
> implement IP_RECVTTL socket option.
> when set on raw or udp sockets, userland receives the incoming packet's TTL
> as ancillary data (cmsg shitz). modeled after the FreeBSD implementation.
> ok claudio djm deraadt (henning@)
~ in.h ~ in_pcb.h
~ ip_output.c ~ tcp_input.c
> implement IP_MINTTL socket option fo tcp sockets
> This is for RFC3682 aka the TTL security hack - sender sets TTL to 255,
> receiver checks no router on the way (or, no more than expected) reduced
> the TTL. carp uses that technique already.
> modeled after FreeBSD implementation.
> ok claudio djm deraadt (henning@)
scsi
~ cd.c
> Something more closely resembling english in this comment. (beck@)
== usr.bin =========================================================== 10/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
calendar
~ calendars/calendar.computer
> fix fortran man year (mickey@)
cdio
~ cdio.1
> document environment var CDROM; from ross l richardson (pr #5258)
> i used a different fix, and merged the CDROM stuff into DISC.
> whilst in there, document the default if neither is given.
> ok mjc (jmc@)
cvs
~ util.c
> check fgets(3) return value and also fix a
> buf[strlen(buf) - 1] = something; misuse.
> ok niallo@, cloder@ (thib@)
file
~ magdir/elf
> recognize 'sh' binaries. (drahn@)
find
~ find.1
> add an example of -exec {} and remove a less useful example; ok otto (jmc@)
~ find.1
> get the spacing right for -exec and -execdir; ok otto (jmc@)
~ find.1
> get the spacing right for -ok too; i missed this in my previous commit.
> (jmc@)
~ find.1
> for -depth, -follow, and -xdev, note that they always evaluate to true;
> ok otto (jmc@)
~ find.1
> in EXAMPLES:
> - make the first example positive (we have enough using `!')
> - remove second example (it is too similar to others)
> - remove `-print' from all examples
> - put the examples in a more logical order
> - no need to escape dot, as otto points out i have been escaping
> it needlessly for years ;(
> ok otto (jmc@)
~ find.1
> Suggest to use xargs(1) and give an example; started by jmc, ok jmc@
> (otto@)
~ find.1
> a little macro cleanup; (jmc@)
~ find.1
> improve markup a wee bit; (jmc@)
~ find.1
> some important information about how find(1) works was tucked away
> under -print, and slightly inaccurate anway; move it up into
> DESCRIPTION and correct it;
> from otto and myself (jmc@)
~ find.1
> hmm, slightly reword the previous to sound less like a susv3 cut/paste,
> thus also avoiding a line split that was annoying me; (jmc@)
gprof
+ sh.c + sh.h
> Extra files needed for 'sh' architecture, copied from arm. (drahn@)
head
~ head.c
> pretty (tedu@)
less
~ tags.c
> Fix an instance of foo[strlen(foo) - 1] = something, which is dangerous
> because strlen(foo) could be 0.
> OK beck@ (cloder@)
~ Makefile.bsd-wrapper
> Allow less to build in Makefile.cross environment. (drahn@)
mail
~ edit.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
rcs
~ co.c ~ rcsdiff.c
~ rcsmerge.c ~ rcsprog.c
~ rcsutil.c ~ rlog.c
> - convert some warnx() to warn(), which fixes a few stupidly un-informative
> error messages (found by tom@)
> - make rcs_choosefile() save wrt errno, input from ray@
> ok joris@ xsa@ (niallo@)
sed
~ compile.c ~ main.c
~ misc.c ~ process.c
> pretty. the comma operator should not be used to cram two statements into
> an if without braces, or worse, into a return. (tedu@)
ssh
~ session.c
> xmalloc -> xcalloc that was missed previously, from portable (djm@)
~ sshconnect.c
> sleep before retrying (not after) since sleep changes errno; fixes
> pr 5250; rad at twig.comk dtucker djm (markus@)
~ clientloop.c ~ serverloop.c
> exit instead of doing a blocking tcp send if we detect a client/server
> timeout,
> since the tcp sendqueue might be already full (of alive requests); ok
> dtucker, report mpf (markus@)
systat
~ systat.1
> vmstat memory stats are in kb, not number of pages;
> from jan srzednicki (freebsd pr #71690)
> ok mickey (jmc@)
~ systat.1
> Talk about "memory" not "pages in kilobytes", it is less awkward
> that way. OK jmc@ deraadt@ (millert@)
tic
~ tic.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
tset
~ tset.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
== usr.sbin ========================================================== 11/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
bind
~ bin/dig/dig.c
> fgets(3) returns NULL on error, not 0. No functional change, but it makes
> the code easier to read.
> OK deraadt (cloder@)
ftp-proxy
~ ftp-proxy.c
> Close file descriptors before cleaning up the events, might cause a race
> otherwise. (camield@)
hostapd
~ parse.y
> check error condition of strtonum (reyk@)
hotplugd
~ Makefile
> landisk will need this too (deraadt@)
lpr
~ lpr/lpr.c
> use SEEK_SET; chl at tuxfamily.orgeraadt@)
mopd
~ common/loop-bsd.c
> use SEEK_SET; chl at tuxfamily.orgeraadt@)
ntpd
~ ntpd.conf.5
> use 'weight-value' vs. 'offset' for the weight argument;
> ok jmc@ henning@ (stevesk@)
~ sensors.c
> need to call adjtime once in a while here too, otherwise sensor-only
> servers never update the system clock (henning@)
~ sensors.c
> internally, ntpd doesn't work with absolute offsets to system time, but
> takes the offset it adjtime() is already correcting for into account
> when taking the offset from a sensor, we need to correct it by the
> offset between system time and ntpd view. (henning@)
pkg_add
~ pkg_create ~ pkg_create.1
> Normalize query options: make them not stop the package generation unless
> -n is also mentioned. Do some devious speed tricks in case -n && -[qQ]
> (forgetting to checksum files, for instance), so that it can be used for
> batch jobs. (espie@)
~ pkg_create ~ pkg_create.1
> tweaks; (jmc@)
===============================================================================
More information about the owc
mailing list