[odc] Daily src changes for 2007-06-01
ODC
auto at squish.net
Sat Jun 2 07:00:01 BST 2007
OpenBSD src changes summary for 2007-06-01
==========================================
bin/chio distrib/sets
lib/libc lib/libm
sbin sbin/bioctl
sbin/disklabel sbin/fsck_ffs
sbin/fsirand sbin/growfs
sbin/isakmpd sbin/mount
sbin/mount_ados sbin/mount_vnd
sbin/mountd sbin/ncheck_ffs
sbin/newfs share/man
sys/adosfs sys/arch
sys/arch/alpha/alpha sys/arch/amd64/amd64
sys/arch/amd64/include sys/arch/amd64/pci
sys/arch/armish/armish sys/arch/aviion/aviion
sys/arch/cats/cats sys/arch/hp300/dev
sys/arch/hp300/hp300 sys/arch/hppa/conf
sys/arch/hppa/hppa sys/arch/hppa64/hppa64
sys/arch/i386/i386 sys/arch/i386/include
sys/arch/i386/pci sys/arch/landisk/landisk
sys/arch/luna88k/luna88k sys/arch/mac68k/mac68k
sys/arch/macppc/conf sys/arch/macppc/dev
sys/arch/macppc/macppc sys/arch/mvme68k/mvme68k
sys/arch/mvme88k/mvme88k sys/arch/mvmeppc/mvmeppc
sys/arch/sgi/sgi sys/arch/solbourne/solbourne
sys/arch/sparc/dev sys/arch/sparc/sparc
sys/arch/sparc64/dev sys/arch/sparc64/sparc64
sys/arch/vax/conf sys/arch/vax/vax
sys/arch/zaurus/zaurus sys/conf
sys/dev sys/dev/ata
sys/dev/isa sys/dev/ofw
sys/dev/onewire sys/dev/pci
sys/dev/pcmcia sys/dev/usb
sys/isofs/cd9660 sys/isofs/udf
sys/kern sys/miscfs/deadfs
sys/miscfs/fifofs sys/miscfs/portal
sys/miscfs/procfs sys/miscfs/specfs
sys/msdosfs sys/net
sys/netbt sys/netinet
sys/netinet6 sys/nfs
sys/ntfs sys/scsi
sys/sys sys/ufs/ext2fs
sys/ufs/ffs sys/ufs/mfs
sys/ufs/ufs sys/uvm
sys/xfs usr.bin/cvs
usr.bin/man usr.sbin/bgpd
usr.sbin/bind usr.sbin/httpd
usr.sbin/pkg_add usr.sbin/sensorsd
usr.sbin/tcpdump
== bin =============================================================== 01/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/bin
chio
~ parse.y
> fix -Wall 'implicit declaration' of err & errx: s/errno.h/err.h/; ok
> deraadt, henning (cnst@)
== distrib =========================================================== 02/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/distrib
sets
~ lists/base/md.alpha ~ lists/base/md.amd64
~ lists/base/md.armish ~ lists/base/md.aviion
~ lists/base/md.cats ~ lists/base/md.hp300
~ lists/base/md.hppa ~ lists/base/md.hppa64
~ lists/base/md.i386 ~ lists/base/md.landisk
~ lists/base/md.luna88k ~ lists/base/md.mac68k
~ lists/base/md.macppc ~ lists/base/md.mvme68k
~ lists/base/md.mvme88k ~ lists/base/md.mvmeppc
~ lists/base/md.sgi ~ lists/base/md.sparc
~ lists/base/md.sparc64 ~ lists/base/md.vax
~ lists/base/md.zaurus ~ lists/base/mi
~ lists/comp/mi ~ lists/man/mi
> sync (deraadt@)
== lib =============================================================== 03/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
libc
~ gen/disklabel.c
> mark the non-handling of 48 bits units with XXX (otto@)
~ sys/mount.2
> more adosfs removal; (jmc@)
libm
~ src/math_private.h
> add little endian and big endian shapes for extended and quad float
> (jason@)
+ arch/i387/s_signbitl_e.c
> add __signbitl for i387 (not used yet) (jason@)
+ arch/mc68881/s_signbitl_e.c
> add __signbitl() for 68881 (not used yet) (jason@)
+ arch/amd64/s_signbitl_e.c
> amd64 version of __signbit() (pedro sized quad floats... really just IEEE
> extended padded to 16 bytes) (jason@)
== sbin ============================================================== 04/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
sbin
~ Makefile
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
bioctl
~ bioctl.c
> support for softraid level 'c'. (tedu@)
disklabel
~ editor.c
> add RAID to the list of partition types for which we do not prompt for a
> filesystem
> ok krw@ (todd@)
fsck_ffs
~ fsck.h ~ pass1.c
~ setup.c
> convert ufs2_daddr_t -> daddr64_t for greater clarity; ok pedro otto thib
> (deraadt@)
~ pass5.c
> Don't use cg.cg_niblk but fs.fs_ipg; the former has no meaning any more.
> Fixes PR 5494; with pedro@; ok pedro@ millert@ (otto@)
~ fsck.h ~ pass1.c
~ setup.c
> ufs1_daddr_t cleanup, okay otto@ (pedro@)
fsirand
~ fsirand.c
> convert ufs2_daddr_t -> daddr64_t for greater clarity; ok pedro otto thib
> (deraadt@)
growfs
~ debug.c ~ growfs.c
> ufs_daddr_t is a dead type. growfs is the only program that uses it, yet
> it is ffs1-aware only. therefore all the types are on disk, and thus,
> they are actually int32_t. ok pedro (deraadt@)
isakmpd
~ conf.c
> Let conf_trans_node() set all parts of the node, so that we don't
> have to expose the node to the outside.
> Without this, conf_trans_node() created a node, linked it into the
> conf_trans queue and returned it to the caller. If something failed
> in one of the callers, the half-initialized node would still be
> linked in the queue and could get accessed later on.
> ok hshoexer@ (moritz@)
mount
~ mount.8 ~ mount.c
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
mount_ados
~ Makefile ~ mount_ados.8
~ mount_ados.c
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
- Makefile - mount_ados.8
- mount_ados.c
> really gone (deraadt@)
mount_vnd
~ mount_vnd.8
> since real boottime vnd support would have caused many problems,
> do a poor man's (but yet working) solution and provide the user with
> instructions to enhance fstab(5) and rc.local(8) with own boot time
> support.
> tweaks and ok jmc@ (grunk@)
mountd
~ mountd.c
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
ncheck_ffs
~ ncheck_ffs.c
> convert ufs2_daddr_t -> daddr64_t for greater clarity; ok pedro otto thib
> (deraadt@)
~ ncheck_ffs.c
> ufs1_daddr_t cleanup, okay otto@ (pedro@)
newfs
~ mkfs.c ~ newfs.c
> convert ufs2_daddr_t -> daddr64_t for greater clarity; ok pedro otto thib
> (deraadt@)
~ mkfs.c
> daddr64_t changes; not fssize is still an int; ok pedro@ (otto@)
~ mkfs.c ~ newfs.c
> ufs1_daddr_t cleanup, okay otto@ (pedro@)
== share ============================================================= 05/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man9/workq_add_task.9
> start new sentence w/ capital letter; (jmc@)
~ man4/options.4 ~ man5/fstab.5
~ man9/malloc.9
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
~ man4/gre.4 ~ man4/ubsa.4
~ man4/uipaq.4 ~ man4/umct.4
~ man4/uplcom.4 ~ man7/securelevel.7
> enforce DESCRIPTION-after-SYNOPSIS order, to appease man -h; (jmc@)
~ man4/ixgb.4
> Intel(R) -> Intel (jmc@)
~ man4/man4.hppa/pdc.4
> s/endianess/endianness/
> ok jmc@ (aanriot@)
~ man9/printf.9
> Sync return value section with reality.
> ok jmc@, "love it" deraadt@ (moritz@)
~ man4/cd.4
> FIBMAP is used by nothing; checked by miod and pvalchev (deraadt@)
== sys =============================================================== 06/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
adosfs
~ adlookup.c ~ adosfs.h
~ adutil.c ~ advfsops.c
~ advnops.c
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
- adlookup.c - adosfs.h
- adutil.c - advfsops.c
- advnops.c
> really gone (deraadt@)
arch
~ macppc/Makefile
> Fix macppc tags.
> ok drahn@ (grange@)
arch/alpha/alpha
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/amd64/amd64
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
~ pmap.c
> set hiwat mark for some of the more popular pools to reduce bouncing
> ok art bob (tedu@)
~ vector.S ~ genassym.cf
> Instead of reexporting PAGE_SIZE from assym.h which causes warnings
> for locore, just include machine/param.h in vector.S (art@)
~ est.c
> some pentium 4 machines are 64-bit and have EST, but a different msr to
> get bus clock. copy in some more code from i386 to deal with both
> families,
> and be more watchful for unknown models.
> fixes a panic reported by johan lindman. ok gwk (tedu@)
arch/amd64/include
~ specialreg.h
> some pentium 4 machines are 64-bit and have EST, but a different msr to
> get bus clock. copy in some more code from i386 to deal with both
> families,
> and be more watchful for unknown models.
> fixes a panic reported by johan lindman. ok gwk (tedu@)
arch/amd64/pci
~ pchb.c
> never attempt to tweak a file just before commit (tedu@)
arch/armish/armish
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/aviion/aviion
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/cats/cats
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/hp300/dev
~ hd.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
arch/hp300/hp300
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/hppa/conf
~ files.hppa
> Infrastructure to allow root on wd0. (miod@)
arch/hppa/hppa
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
~ autoconf.c
> Infrastructure to allow root on wd0. (miod@)
arch/hppa64/hppa64
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/i386/i386
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/i386/include
~ vmparam.h
> Provide carefully crafted DEADBEEF* values, this was initially forgotten in
> the commit which enabled DEADBEEF* overrides, then values have been
> discussed a bit more and improved; reminded by and ok art@ (miod@)
arch/i386/pci
~ pchb.c
> Revert INTEL_82915GM part of tedu's commit as it hung my T43
> ok tedu@ (biorn@)
arch/landisk/landisk
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/luna88k/luna88k
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/mac68k/mac68k
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/macppc/conf
~ files.macppc
> remove ofw devices which we don't use; that way of mixing devices is crazy
> (means i also don not have to fix the daddr_t problems)
> ok miod (deraadt@)
arch/macppc/dev
~ xlights.c
> Establish interrupt after checking the i2s subsystem, fixes a problem with
> xserve G5.
> ok drahn (gwk@)
arch/macppc/macppc
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
~ autoconf.c ~ conf.c
> remove ofw devices which we don't use; that way of mixing devices is crazy
> (means i also don not have to fix the daddr_t problems)
> ok miod (deraadt@)
arch/mvme68k/mvme68k
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/mvme88k/mvme88k
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/mvmeppc/mvmeppc
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
~ autoconf.c ~ conf.c
> remove ofw devices which we don't use; that way of mixing devices is crazy
> (means i also don not have to fix the daddr_t problems)
> ok miod (deraadt@)
arch/sgi/sgi
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/solbourne/solbourne
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/sparc/dev
~ fd.c ~ presto.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
arch/sparc/sparc
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/sparc64/dev
~ fd.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
arch/sparc64/sparc64
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/vax/conf
~ files.vax
> Replace comment in Swedish. -moj (maja@)
arch/vax/vax
~ autoconf.c
> Fix cd block major number (miod@)
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
arch/zaurus/zaurus
~ autoconf.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
conf
~ files
> complete the sys/netbt import and adopt some more code to our interfaces
> (uwe@)
~ files
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
dev
~ ccd.c ~ flash.c
~ ramdisk.c ~ vnd.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
~ softraid.c
> save metdata in context using dlg and tedu's shiny new workq api.
> detect stale metadata during bringup. (marco@)
~ softraid.c
> Check return code for resource allocation.
> prompted by tedu (marco@)
~ softraid.c ~ softraidvar.h
> Add additional flag to indicate that we have a metdadata io pending; this
> way if we shut down the system we can make sure that it makes it onto disk
> before scsibus detach.
> ok dlg (marco@)
~ softraid.c
> check return value of findblkmajor; this fixes a crash on vax during boot.
> add check for failed geteblk.
> exclude rx during boot probe.
> help miod todd
> ok todd (marco@)
dev/ata
~ wd.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
dev/isa
~ mcd.c ~ fd.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
~ nsclpcsio_isa.c
> remove strings from sensor.desc which are now present in sysctl variable
> names; ok grange@ (cnst@)
dev/ofw
~ ofdisk.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
- files.ofw - ofbus.c
- ofcons.c - ofdisk.c
- ofnet.c - ofrtc.c
> remove ofw devices which we don't use; that way of mixing devices is crazy
> (means i also don not have to fix the daddr_t problems)
> ok miod (deraadt@)
dev/onewire
~ owid.c
> remove strings from sensor.desc which are now present in sysctl variable
> names; ok grange@ (cnst@)
~ owsbm.c
> remove strings from sensor.desc which are now present in sysctl variable
> names; ok grange@ (cnst@)
dev/pci
~ if_msk.c
> Only probe phy's at address 0; gets rid of bogus phy's on the
> Marvell Yukon 88E8036.
> ok jsg@ (kettenis@)
~ if_myx.c ~ if_myxreg.h
> initialize the rings (reyk@)
dev/pcmcia
~ if_malo.c ~ if_maloreg.h
~ if_malovar.h
> Add RX path and enable monitor mode. (mglocker@)
dev/usb
~ ubt.c
> Remove the hci_enable now done in a higher layer.
> ok uwe (gwk@)
~ if_upl.c ~ ubsa.c
~ uhub.c ~ ulpt.c
~ umass.c
> Remove more Free(?)BSD and NetBSD #ifdef/#endifs. No functional nor
> binary change.
> ok jsg (mbalmer@)
isofs/cd9660
~ cd9660_vnops.c
> FIBMAP is used by nothing; checked by miod and pvalchev (deraadt@)
~ cd9660_bmap.c ~ cd9660_lookup.c
~ cd9660_node.c ~ cd9660_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
isofs/udf
~ udf_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
kern
~ kern_sensors.c
> rework the sensor task handling to run in the kernels generic workq.
> apologies to art for abusing timeouts so badly. apologies to tedu for
> making his head hurt.
> tested by and ok tedu@ (dlg@)
~ Make.tags.inc ~ Makefile
~ vfs_conf.c
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
~ kern_sysctl.c ~ vfs_subr.c
> decouple the allocated number of vnodes from the "desiredvnodes" variable
> which is used to size a zillion other things that increasing excessively
> has been shown to cause problems - so that we may incrementally look at
> increasing those other things without making the kernel unusable.
> This diff effectivly increases the number of vnodes back to the number
> of buffers, as in the earlier dynamic buffer cache commits, without
> increasing anything else (namecache, softdeps, etc. etc.)
> ok pedro@ tedu@ art@ thib@ (beck@)
~ vfs_bio.c
> dont request zeroed memory when we allocate data regions for buffers. this
> moves memset from the 20th most expensive function in the kernel to the
> 331st when doing heavy io.
> ok tedu@ thib@ pedro@ beck@ art@ (dlg@)
~ subr_disk.c
> Add two constraints to V1 disklabels:
> 1) All partitions must start before the end of the disk. 2) All
> partitions must end at or before the end of the disk.
> Partitions not satisfying these constraints will be truncated and
> marked FS_UNUSED.
> ok otto@ deraadt@ (krw@)
~ init_main.c
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
~ vfs_vnops.c
> FIBMAP is used by nothing; checked by miod and pvalchev (deraadt@)
~ vfs_bio.c
> Uninline bio_doread(), okay art@ (pedro@)
~ vfs_default.c ~ vfs_sync.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
miscfs/deadfs
~ dead_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
miscfs/fifofs
~ fifo_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
miscfs/portal
~ portal_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
miscfs/procfs
~ procfs_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
miscfs/specfs
~ spec_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
msdosfs
~ msdosfs_denode.c ~ msdosfs_lookup.c
~ msdosfs_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
net
~ pf.c
> apply the "skip ipsec if there are no flows" speedup diff to IPv6 too.
> we need a pointer to the inpcb to decide, which was not previously
> passed to ip6_output, so this diff is a little bigger.
> from itojun, ok ryan (henning@)
~ pf.c
> fold pf_test_tcp(), pf_test_udp(), pf_test_icmp(), pf_test_other() into
> one - pf_test_rule(). now we have one place to make things clearer and
> maybe find another few performance bits :)
> shrinks i386 GENERIC by 11K, no measurable performance impact or gain.
> lots lots lots lots lots of testing and headbanging with ryan,
> performance testing ckuethe. ok ryan (henning@)
~ if_pfsync.c ~ pf.c
~ pf_ioctl.c ~ pfvar.h
> factor out duplicated code to allocate state key and cross-reference it
> with a state entry into a new pf_alloc_state_key() function and use it
> everywhere. makes upcoming changes way easier and is cleaner anyway.
> conceptually agreed by ryan, but he's on the road now ;( (henning@)
netbt
~ bt_proto.c ~ hci.h
~ hci_event.c ~ hci_link.c
~ hci_socket.c ~ hci_unit.c
~ l2cap.h ~ l2cap_misc.c
~ sco.h + hci_ioctl.c
+ l2cap_socket.c + l2cap_upper.c
+ rfcomm.h + rfcomm_dlc.c
+ rfcomm_session.c + rfcomm_socket.c
+ rfcomm_upper.c + sco_socket.c
+ sco_upper.c
> complete the sys/netbt import and adopt some more code to our interfaces
> (uwe@)
~ hci_event.c
> set debug level to 0 (uwe@)
~ hci_unit.c
> workq_add_task() flags argument position changed (uwe@)
~ hci_unit.c
> Set BTF_UP before hci_enable() to make hci_unit_lookup() work; ok gwk
> (uwe@)
netinet
~ ip_carp.c
> Remove code to create a carp(4) Token Ring lladdr.
> OK mcbride@ (mpf@)
~ ip_carp.c ~ ipsec_output.c
~ tcp_input.c ~ tcp_output.c
~ tcp_subr.c
> apply the "skip ipsec if there are no flows" speedup diff to IPv6 too.
> we need a pointer to the inpcb to decide, which was not previously
> passed to ip6_output, so this diff is a little bigger.
> from itojun, ok ryan (henning@)
~ ip_carp.c
> sc_carplladdr does not need to be dereferenced.
> OK mcbride@ (mpf@)
netinet6
~ icmp6.c ~ in6_gif.c
~ ip6_forward.c ~ ip6_mroute.c
~ ip6_output.c ~ ip6_var.h
~ mld6.c ~ nd6_nbr.c
~ raw_ip6.c ~ udp6_output.c
> apply the "skip ipsec if there are no flows" speedup diff to IPv6 too.
> we need a pointer to the inpcb to decide, which was not previously
> passed to ip6_output, so this diff is a little bigger.
> from itojun, ok ryan (henning@)
nfs
~ nfs_bio.c ~ nfs_var.h
~ nfs_vnops.c
> daddr_t -> daddr64_t;
> Basically the usage of daddr_t was to math out arguments to
> nfs_getcacheblk, wich calls getblk();
> ok deraadt@ (thib@)
~ nfs_bio.c ~ nfs_kq.c
~ nfs_node.c ~ nfs_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
ntfs
~ ntfs_vnops.c
> more vop argument "ap = v" comment removal; ok pedro (deraadt@)
scsi
~ cd.c ~ sd.c
> Don't initialize d_partitions[RAW_DISK] just before calling
> readdisklabel(), since all readdisklabel()'s do that already.
> ok deraadt@ (krw@)
~ st.c
> No need for tape fields to be daddr_t. int will do.
> ok beck@ deraadt@ (krw@)
~ ss_scanjet.c
> Fix snprintf(9) truncation check and skip the check for -1 because
> only the libc snprintf(3) could return this.
> ok krw@ (moritz@)
~ safte.c
> backout r1.33: temperature sensor description was removed on purpose,
> because it started to be part of the sysctl variable name
> ok deanna@, dlg@ (cnst@)
sys
~ malloc.h ~ mount.h
> remove crufty adosfs. everytime we try to change a vfs layer interface,
> adosfs has the issue and in an entirely different way. noone uses this
> code so this is hte best way to handle the problem
> diff from pedro who went to sleep but we need to move ahead; ok thib
> (deraadt@)
~ vnode.h
> decouple the allocated number of vnodes from the "desiredvnodes" variable
> which is used to size a zillion other things that increasing excessively
> has been shown to cause problems - so that we may incrementally look at
> increasing those other things without making the kernel unusable.
> This diff effectivly increases the number of vnodes back to the number
> of buffers, as in the earlier dynamic buffer cache commits, without
> increasing anything else (namecache, softdeps, etc. etc.)
> ok pedro@ tedu@ art@ thib@ (beck@)
~ disklabel.h
> format_op df_startblk can be a daddr64_t too (deraadt@)
~ disk.h
> more completely safe daddr_t -> daddr64_t (deraadt@)
~ mtio.h
> No need for tape fields to be daddr_t. int will do.
> ok beck@ deraadt@ (krw@)
~ systm.h
> some architectures called setroot() from cpu_configure(), *way* before some
> subsystems were enabled. others used a *md_diskconf -> diskconf() method
> to
> make sure init_main could "do late setroot". Change all architectures to
> have diskconf(), use it directly & late. tested by todd and myself on most
> architectures, ok miod too (deraadt@)
- tablet.h
> wipe tablet.h that was used by tb(4) that was
> removed months ago; ok deraadt@ (robert@)
~ filio.h
> FIBMAP is used by nothing; checked by miod and pvalchev (deraadt@)
ufs/ext2fs
~ ext2fs_bmap.c ~ ext2fs_inode.c
~ ext2fs_lookup.c ~ ext2fs_readwrite.c
~ ext2fs_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
ufs/ffs
~ ffs_alloc.c ~ ffs_balloc.c
~ ffs_extern.h ~ ffs_inode.c
~ ffs_softdep.c ~ ffs_vfsops.c
> convert ufs2_daddr_t -> daddr64_t for greater clarity; ok pedro otto thib
> (deraadt@)
~ fs.h
> fix comment, fs_ipg is inode *blocks* per cylinder group (otto@)
~ fs.h
> revert previous comment "fix", must have been sleeping (otto@)
~ ffs_alloc.c ~ ffs_balloc.c
~ ffs_extern.h ~ ffs_inode.c
~ ffs_softdep.c
> cleanup of 'ufs1_daddr_t', first round, okay deraadt@ (pedro@)
~ ffs_alloc.c ~ ffs_extern.h
~ ffs_softdep.c ~ ffs_softdep_stub.c
~ softdep.h
> Nuke 'ufs_lbn_t', okay otto@ deraadt@ krw@ beck@ (pedro@)
~ ffs_alloc.c ~ ffs_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
ufs/mfs
~ mfs_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
ufs/ufs
~ dinode.h ~ ufs_bmap.c
> convert ufs2_daddr_t -> daddr64_t for greater clarity; ok pedro otto thib
> (deraadt@)
~ ufs_bmap.c
> fix struct vop_bmap_args comment. ok deraadt@ (otto@)
~ ufs_bmap.c
> and don't forget the other daddr var in the comment (otto@)
~ dinode.h
> Nuke ufs_time_t, okay beck@ krw@ marco@ (pedro@)
~ dinode.h
> cleanup of 'ufs1_daddr_t', first round, okay deraadt@ (pedro@)
~ dinode.h
> ufs_daddr_t is a dead type. growfs is the only program that uses it, yet
> it is ffs1-aware only. therefore all the types are on disk, and thus,
> they are actually int32_t. ok pedro (deraadt@)
~ dinode.h
> Nuke 'ufs_lbn_t', okay otto@ deraadt@ krw@ beck@ (pedro@)
~ ufs_bmap.c ~ ufs_inode.c
~ ufs_lookup.c ~ ufs_vnops.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
uvm
~ uvm_swap.c
> wrong cast checking for VOP_BMAP return value; ok miod (deraadt@)
~ uvm_amap.c ~ uvm_map.c
> set hiwat mark for some of the more popular pools to reduce bouncing
> ok art bob (tedu@)
xfs
~ xfs_vnodeops-bsd.c
> pedro ok'd this ~3500 line diff which removes the vop argument
> "ap = v" comments in under 8 seconds, so it must be ok. and it compiles
> too. (deraadt@)
== usr.bin =========================================================== 07/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
cvs
~ checkout.c ~ commit.c
~ file.c ~ file.h
~ import.c ~ rcs.c
~ rcs.h ~ repository.c
~ repository.h ~ status.c
~ update.c
> add support for local and remote branch checkout. for example,
> cvs co -rOPENBSD_3_0 works now.
> along the way, simplify and rationalise code and fix a few nits.
> ok joris@ xsa@ ray@ (niallo@)
man
~ man.1
> more adosfs removal; (jmc@)
== usr.sbin ========================================================== 08/08 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
bgpd
~ rde.h ~ rde_rib.c
> Remove a stupid wrapper function that does nothing more then calling
> another
> function with the same arguments. (claudio@)
bind
~ lib/dns/name.c
> avoid potentially bad snprintf construct; ok henning millert otto
> (pvalchev@)
httpd
~ src/modules/ssl/ssl_util_ssl.c
> paranoid snprintf length calculation to avoid a possible buffer
> overflow in SSL session id parsing (by reaching a negative size arg)
> ok henning (pvalchev@)
pkg_add
~ OpenBSD/Replace.pm
> no need to require ProgressMeter twice. (espie@)
~ pkg_add ~ pkg_delete
~ OpenBSD/Add.pm ~ OpenBSD/Delete.pm
~ OpenBSD/PackageInfo.pm ~ OpenBSD/ProgressMeter.pm
~ OpenBSD/Replace.pm ~ OpenBSD/SharedItems.pm
~ OpenBSD/Update.pm ~ OpenBSD/Vstat.pm
> refactor ProgressMeter code to create objects, put these into state.
> Create a common state class for delete and add, put the progressmeter
> creation in there.
> neat effect: we don't have a progressmeter, we use a stub class instead
> and *never* pull the code in at all. (espie@)
~ OpenBSD/PackingElement.pm ~ OpenBSD/PackingList.pm
> micro-optimizations (espie@)
~ pkg_add
> compute real list. (espie@)
~ pkg_add ~ OpenBSD/Replace.pm
> don't store stuff in the plist, use the handle which is there for that.
> so loop directly over $set->older for replacements. (espie@)
~ OpenBSD/Add.pm ~ OpenBSD/Replace.pm
> apply the same checks on extraction that are done during pkg_create
> (modes and owners without annotation WILL be refused).
> Somewhat pedantic: skip archive entry during extraction of links/symlinks.
> Doesn't change a thing, except in the very rare case where a package would
> end with a link in the packing-list. Now we're able to detect a premature
> end of the tarball.
> (okay, it's an obscure bug, I haven't even met it for real. So what ?)
> (espie@)
~ pkg_add ~ OpenBSD/Dependencies.pm
~ OpenBSD/Replace.pm ~ OpenBSD/RequiredBy.pm
~ pod/OpenBSD::RequiredBy.pod
> very small clean-up of dependencies list handling. (espie@)
sensorsd
~ sensorsd.c
> Convert sensorsd to use two-level data-structures. This makes sensorsd be
> more consistent with the current sensors framework, conserves some memory,
> and will make it easier to implement hotplugging and other nifty features
> in the future. This does not change any other functionality ATM.
> OK henning@, beck at . (cnst@)
tcpdump
~ tcpdump.c
> Move a useless warning to the only case that uses it (-f) from canacar@
> ok beck@ (todd@)
===============================================================================
More information about the odc
mailing list