[owc] Weekly src changes ending 2007-12-30
OWC
auto at squish.net
Mon Dec 31 07:00:01 GMT 2007
OpenBSD src changes summary for 2007-12-23 to 2007-12-30 inclusive
==================================================================
bin/date bin/ps
distrib/sets gnu/usr.bin/binutils
lib/libc libexec/spamd
regress/sys sbin/disklabel
sbin/fdisk sbin/fsck_msdos
sbin/kbd sbin/mknod
sbin/mountd sbin/ping6
sbin/routed share/man
share/zoneinfo sys/arch/alpha/conf
sys/arch/alpha/include sys/arch/amd64/amd64
sys/arch/amd64/conf sys/arch/aviion/aviion
sys/arch/hp300/hp300 sys/arch/hp300/stand/libsa
sys/arch/hp300/stand/libsaz sys/arch/hppa/dev
sys/arch/hppa/hppa sys/arch/hppa/include
sys/arch/hppa/stand/boot sys/arch/hppa/stand/cdboot
sys/arch/hppa/stand/libsa sys/arch/i386/conf
sys/arch/i386/i386 sys/arch/m68k/include
sys/arch/m88k/conf sys/arch/m88k/fpu
sys/arch/m88k/include sys/arch/m88k/m88k
sys/arch/mac68k/dev sys/arch/mac68k/mac68k
sys/arch/mvme68k/stand/libsa sys/arch/mvme88k/dev
sys/arch/mvme88k/include sys/arch/mvme88k/mvme88k
sys/arch/sgi/dev sys/arch/sgi/sgi
sys/arch/sparc/dev sys/arch/sparc64/sparc64
sys/arch/vax/vax sys/arch/vax/vsa
sys/arch/zaurus/dev sys/dev
sys/dev/acpi sys/dev/ata
sys/dev/ic sys/dev/isa
sys/dev/pci sys/kern
sys/lib/libkern sys/miscfs/deadfs
sys/miscfs/fifofs sys/miscfs/specfs
sys/net sys/netinet
sys/nfs sys/scsi
sys/sys usr.bin/pcc
usr.bin/quota usr.bin/rcs
usr.bin/sectok usr.bin/ssh
usr.bin/vmstat usr.sbin/bgpctl
usr.sbin/bgpd usr.sbin/dhcpd
usr.sbin/ntpd usr.sbin/pkg_add
usr.sbin/pstat usr.sbin/snmpctl
usr.sbin/snmpd
== bin =============================================================== 01/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/bin
date
~ date.c
> printf field widths are always int, so add a cast to remove
> "warning: field width is not type int"
> "looks good" otto@ (chl@)
ps
~ ps.c
> printf field widths are always int, so add a cast to remove
> "warning: field width is not type int"
> "looks good" otto@ (chl@)
== distrib =========================================================== 02/11 ==
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.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/comp/mi
> sync (deraadt@)
~ lists/base/md.alpha ~ lists/base/md.amd64
~ lists/base/md.armish ~ lists/base/md.aviion
~ 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
> sync (deraadt@)
~ lists/base/mi
> sync (deraadt@)
== gnu =============================================================== 03/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/gnu
usr.bin/binutils
~ gas/config/tc-sparc.c
> Teach as(1) about %gl on sparc64.
> ok miod@ (kettenis@)
~ Makefile.bsd-wrapper
> enforce install -S, avoids races with make -j:
> libtool does relink some files during install, and it can happen that it
> relinks stuff at the exact same time that ld is being reinstalled.
> okay kettenis@ (espie@)
== lib =============================================================== 04/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
libc
~ time/localtime.c ~ time/private.h
~ time/strftime.c ~ time/tz-link.htm
~ time/zdump.c ~ time/zic.c
> update to tzcode2007j from elsie.nci.nih.gov; mostly ANSIfication
> (millert@)
== libexec =========================================================== 05/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/libexec
spamd
~ sync.c
> Add a missing length check of received spamd sync packet headers.
> From Diego Giagio (dgiagio at gmail)
> with input from tedu@ (reyk@)
== regress =========================================================== 06/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/regress
sys
~ kern/noexec/noexec.c
> -s uses optarg; Elad Efrat (deraadt@)
== sbin ============================================================== 07/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
disklabel
~ editor.c
> Setting p_fstype to the number of free sectors can't be right. Instead, set
> the partition size with DL_SETPSIZE(pp, *freep). (krw@)
~ editor.c
> Use zero_partitions() to implement the '-E' command 'd *', as it
> already is used to implement 'z'. Vocal support for both commands so
> keep both for now.
> 'sounds good' deraadt at . (krw@)
~ editor.c
> Some cosmetic code reorg to make things a bit more textually
> consistant. No functional change. (krw@)
~ editor.c
> Bring editor_add() code into line with the other editor_* functions.
> The only visible change is allowing only one chance to specify a valid
> partition letter before returning to the main edit prompt.
> "This looks OK to me" millert@ (krw@)
~ editor.c
> Since 'c' now always covers the entire disk and is always FS_UNUSED,
> there is no point in allowing the edit mode commands
> (a)add/(n)ame/(m)odify/(d)elete/(c)hange to pretend to fiddle with it.
> Some already checked for and avoided 'c', but do it in a consistant
> manner in all cases.
> ok millert@ otto@ marco@ (krw@)
~ editor.c
> Simplify code by avoiding manual manipulations of the free space
> counter. Call edit_countfree() instead.
> Feedback from otto at .
> ok millert@ "Go for it" deraadt@ (krw@)
~ editor.c
> Kill some now unneeded checks for 'c' partition manipulation. Change
> the variable 'c' to 'partno' in editor_delete() so all editor_*
> functions are even more consistant. No functional change. (krw@)
~ editor.c
> Use get_size() in editor_change() instead of hand-rolling identical
> code. Makes the size prompt/helpstring/looping the same for
> (a)dd/(m)odify/(c)hange. (krw@)
fdisk
~ fdisk.c
> sort flags in both synopsis and usage
> ok jmc@ (sobrado@)
~ fdisk.8
> sort flags in both synopsis and usage, file not commited in last diff
> ok jmc@ (sobrado@)
fsck_msdos
~ fsck_msdos.8
> sort flags in both synopsis and usage
> ok jmc@ (sobrado@)
kbd
~ main.c
> each usage requires its own line
> written with improvements by jmc@
> ok jmc@ (sobrado@)
mknod
~ mknod.c
> do not repeat "usage:" twice
> ok jmc@ (sobrado@)
mountd
~ mountd.c
> use the same argument name in synopsis and usage; lowercase "usage:"
> ok jmc@ (sobrado@)
ping6
~ ping6.c
> add missing space between argument names and ellipsis
> ok jmc@ (sobrado@)
routed
~ trace.c
> printf field widths are always int, so add a cast to remove
> "warning: field width is not type int"
> "looks good" otto@ (chl@)
~ rtquery/rtquery.c
> usage messages should have the form "usage: command [arguments]"
> ok jmc@ (sobrado@)
== share ============================================================= 08/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man9/pmap.9
> pmap_steal_memory() returns vaddr_t.
> ok miod@ (jsing@)
~ man4/it.4
> Look for environment controller first, and access ITE SuperIO address/data
> ports only if we found one. Fixes possible issues with some notebook
> chipsets.
> ok deraadt@
> tested by me, jcs@ and Rodolfo Gouveia (form@)
~ man4/aps.4 ~ man4/it.4
~ man4/nsclpcsio.4
> +.Xr isa 4 , (jmc@)
~ man4/man4.alpha/tga.4 ~ man4/man4.sparc/mgx.4
~ man4/man4.sparc/vigra.4 ~ man4/man4.sparc64/mgx.4
~ man4/man4.sparc64/vigra.4 ~ man4/pwdog.4
> {dip,DIP-} -> DIP; now they're all the same... (jmc@)
~ man3/queue.3
> some fixes from Girish Venkatachalam; as far as i can see, otto has already
> ok'd this, but is obviously away in some winter wonderland... (jmc@)
~ man3/tree.3
> bool -> int (otto@)
~ man9/microtime.9
> last item in NAME should not be followed by a comma; from Pierre Riteau
> (jmc@)
zoneinfo
~ datfiles/asia ~ datfiles/iso3166.tab
~ datfiles/northamerica ~ datfiles/southamerica
~ datfiles/zone.tab
> update to tzdata2007j from elsie.nci.nih.gov (millert@)
== sys =============================================================== 09/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/alpha/conf
~ files.alpha
> Unifdef NO_IEEE in the libkern softfloat code, and put an appropriate rule
> in files.alpha to compensate for NO_IEEE kernels.
> This will allow the softfloat code to be used by other platforms than
> alpha. (miod@)
arch/alpha/include
~ ieeefp.h
> Do not define _FP_DYNAMIC; no other platform defines this and nothing
> (either
> in kernel or in userland) uses it.
> ok deraadt@ (miod@)
arch/amd64/amd64
~ autoconf.c
> missing newline in autoconf pxe report; spotted by henning (deraadt@)
arch/amd64/conf
~ GENERIC
> Look for environment controller first, and access ITE SuperIO address/data
> ports only if we found one. Fixes possible issues with some notebook
> chipsets.
> ok deraadt@
> tested by me, jcs@ and Rodolfo Gouveia (form@)
arch/aviion/aviion
~ av400_machdep.c
> Fix the spurious or unclaimed interrupt messages, I swapped them by
> mistake. (miod@)
arch/hp300/hp300
~ locore.s
> Correctly handle non-null state frames on 68020 and 68030 in fpfault.
> (miod@)
arch/hp300/stand/libsa
~ Makefile
> Repair fallout from libkern changes. (miod@)
arch/hp300/stand/libsaz
~ Makefile
> Repair fallout from libkern changes. (miod@)
arch/hppa/dev
~ uturn.c
> Add a few more K-class models to the list of machines that have more
> devices
> than PDC tells us about. (kettenis@)
~ uturn.c
> Skip LBC on buswalk. Gets rid of unconfigured "Bus Converter Port" devices
> on machines like the C360. (kettenis@)
arch/hppa/hppa
~ mainbus.c
> Add a few more K-class models to the list of machines that have more
> devices
> than PDC tells us about. (kettenis@)
arch/hppa/include
~ loadfile_machdep.h
> Teach the hppa bootloader how to boot 64-bit kernels.
> ok miod@, deraadt@ (kettenis@)
arch/hppa/stand/boot
~ conf.c
> Teach the hppa bootloader how to boot 64-bit kernels.
> ok miod@, deraadt@ (kettenis@)
arch/hppa/stand/cdboot
~ Makefile
> Teach the hppa bootloader how to boot 64-bit kernels.
> ok miod@, deraadt@ (kettenis@)
arch/hppa/stand/libsa
~ Makefile + elf32.c
+ elf64.c
> Teach the hppa bootloader how to boot 64-bit kernels.
> ok miod@, deraadt@ (kettenis@)
arch/i386/conf
~ GENERIC
> Look for environment controller first, and access ITE SuperIO address/data
> ports only if we found one. Fixes possible issues with some notebook
> chipsets.
> ok deraadt@
> tested by me, jcs@ and Rodolfo Gouveia (form@)
arch/i386/i386
~ autoconf.c
> missing newline in autoconf pxe report; spotted by henning (deraadt@)
arch/m68k/include
~ pmap_motorola.h
> Do not expose struct pmap unless _KERNEL (miod@)
arch/m88k/conf
~ files.m88k
> Restart the m88110 floating-point trap code from scratch, basing it on the
> sparc{,64} floating-point emulator, adapted to the 88110 specifics.
> Handling
> of these traps is now entirely done in C for convenience.
> Although there are a few rounding issues to address, and the XRF is
> ignored,
> this allows all the ieee754 regression tests to pass (lib/libc/ieeefp,
> sys/kern/signal/fpsig, sys/kern/signal/sigfpe). (miod@)
~ files.m88k
> Had I known we had a kernel version of John Hauser's SoftFloat code, I
> would
> have jumped on it instead of basing the FPU completion work on the sparc
> FPU code.
> This is now repaired with this commit, and m88110_fp.c changes directory
> again, for the last time. (miod@)
arch/m88k/fpu
+ fpu_add.c + fpu_arith.h
+ fpu_compare.c + fpu_div.c
+ fpu_emu.h + fpu_explode.c
+ fpu_implode.c + fpu_mul.c
+ fpu_sqrt.c + fpu_subr.c
+ m88110_fp.c
> Restart the m88110 floating-point trap code from scratch, basing it on the
> sparc{,64} floating-point emulator, adapted to the 88110 specifics.
> Handling
> of these traps is now entirely done in C for convenience.
> Although there are a few rounding issues to address, and the XRF is
> ignored,
> this allows all the ieee754 regression tests to pass (lib/libc/ieeefp,
> sys/kern/signal/fpsig, sys/kern/signal/sigfpe). (miod@)
~ m88110_fp.c
> Never write boolean expressions which do not fit in one line, after
> midnight. (miod@)
~ m88110_fp.c
> Missing pointer initialization for int/nint/trnc exceptions. (miod@)
~ fpu_div.c
> Zero / Num should return a zero with the same sign as the divider. (miod@)
~ m88110_fp.c
> Accept valid forms of double-precision fmul/fadd/fsub/fdiv and reject
> invalid
> forms of fcvt where source and destination precision is the same. (miod@)
~ fpu_compare.c ~ fpu_implode.c
~ fpu_mul.c
> A few fixes in the exception bits reporting, spotted by John Hauser's
> TestFloat. (miod@)
~ fpu_arith.h
> Fix FPU_SET_CARRY() (miod@)
~ fpu_compare.c
> First try at getting the interval bits filled in fcmp{,u} results. (miod@)
~ fpu_implode.c
> Honour the rounding mode in fpu_ftoi(). (miod@)
- fpu_add.c - fpu_arith.h
- fpu_compare.c - fpu_div.c
- fpu_emu.h - fpu_explode.c
- fpu_implode.c - fpu_mul.c
- fpu_sqrt.c - fpu_subr.c
- m88110_fp.c
> Had I known we had a kernel version of John Hauser's SoftFloat code, I
> would
> have jumped on it instead of basing the FPU completion work on the sparc
> FPU code.
> This is now repaired with this commit, and m88110_fp.c changes directory
> again, for the last time. (miod@)
arch/m88k/include
~ trap.h
> Restart the m88110 floating-point trap code from scratch, basing it on the
> sparc{,64} floating-point emulator, adapted to the 88110 specifics.
> Handling
> of these traps is now entirely done in C for convenience.
> Although there are a few rounding issues to address, and the XRF is
> ignored,
> this allows all the ieee754 regression tests to pass (lib/libc/ieeefp,
> sys/kern/signal/fpsig, sys/kern/signal/sigfpe). (miod@)
~ pmap.h
> Do not make struct pmap visible unless _KERNEL. (miod@)
~ ieeefp.h
> Had I known we had a kernel version of John Hauser's SoftFloat code, I
> would
> have jumped on it instead of basing the FPU completion work on the sparc
> FPU code.
> This is now repaired with this commit, and m88110_fp.c changes directory
> again, for the last time. (miod@)
~ ieeefp.h
> Override SoftFloat's countLeadingZeros32() with a faster, ff1-based,
> version
> on m88k. (miod@)
arch/m88k/m88k
- m88110_fp.S ~ eh_common.S
~ trap.c
> Restart the m88110 floating-point trap code from scratch, basing it on the
> sparc{,64} floating-point emulator, adapted to the 88110 specifics.
> Handling
> of these traps is now entirely done in C for convenience.
> Although there are a few rounding issues to address, and the XRF is
> ignored,
> this allows all the ieee754 regression tests to pass (lib/libc/ieeefp,
> sys/kern/signal/fpsig, sys/kern/signal/sigfpe). (miod@)
~ m88k_machdep.c
> Remove the last debug bit from the PSR on 88110: do not force memory
> accesses
> instructions to be serialized (this defeats the purpose of having a
> superscalar
> processor, and accesses to volatile variables are done with explicit memory
> barriers anyway).
> This brings a HUGE speedup: openssl speed -elapsed shows AES is 90% faster,
> blowfish is 75% faster, and sha1 is 50% faster. Not so bad!
> However, doing this increases the pressure on the processor bus, so it is
> necessary to increase the processor bus timeout on 40MHz boards again (to
> 256
> usec). ``black cat'' 50MHz boards seem to be unaffected, so they remain at
> 64 usec. (miod@)
+ m88110_fp.c
> Had I known we had a kernel version of John Hauser's SoftFloat code, I
> would
> have jumped on it instead of basing the FPU completion work on the sparc
> FPU code.
> This is now repaired with this commit, and m88110_fp.c changes directory
> again, for the last time. (miod@)
arch/mac68k/dev
~ ncr5380.c ~ sbc.c
> fix comments after the members of scsi_adapter were changed.
> prompted by krw@ (dlg@)
arch/mac68k/mac68k
~ locore.s
> Correctly handle non-null state frames on 68020 and 68030 in fpfault.
> (miod@)
arch/mvme68k/stand/libsa
~ Makefile
> Repair fallout from libkern changes. (miod@)
arch/mvme88k/dev
~ busswreg.h
> Increase timeout on the processor bus when running without snooping
> enabled.
> This lets 40MHz MVME197LE boards run with instruction cache enabled, and
> also
> fixes random instruction faults occuring on the early 50MHz models. (miod@)
- sysconreg.h ~ dart.c
~ syscon.c ~ vme.c
+ sysconvar.h
> Make the mvme188 interrupt handling closer to the aviion interrupt
> handling:
> non-VME syscon interrupt sources will now use their own intrhand array,
> and interrupt sources will be enabled in the arbiter as interrupt handlers
> are registered. This allows VME devices to use the whole 256 interrupts
> range. (miod@)
arch/mvme88k/include
~ intr.h ~ mvme188.h
> Make the mvme188 interrupt handling closer to the aviion interrupt
> handling:
> non-VME syscon interrupt sources will now use their own intrhand array,
> and interrupt sources will be enabled in the arbiter as interrupt handlers
> are registered. This allows VME devices to use the whole 256 interrupts
> range. (miod@)
arch/mvme88k/mvme88k
~ m88110.c
> Increase timeout on the processor bus when running without snooping
> enabled.
> This lets 40MHz MVME197LE boards run with instruction cache enabled, and
> also
> fixes random instruction faults occuring on the early 50MHz models. (miod@)
~ m88110.c
> Tell 197SP and 197DP apart in MULTIPROCESSOR kernels. (miod@)
~ m197_machdep.c
> Limit physmem to 32MB on 01-W3869B02[EF] boards which shipped with 64MB of
> memory but a memory controller limited to 32MB.
> Not tested for lack of such crippled hardware, just the average
> once-per-leap-year act of niceness from me (a bit early though). (miod@)
~ m88110.c
> Since no cache flush/inval operation will occur before we enable D$, and
> we never disable it, it is not necessary to check for D$ to be enabled
> before acting. That's a few more cycles spared. (miod@)
~ m197_machdep.c ~ m88110.c
> Remove the last debug bit from the PSR on 88110: do not force memory
> accesses
> instructions to be serialized (this defeats the purpose of having a
> superscalar
> processor, and accesses to volatile variables are done with explicit memory
> barriers anyway).
> This brings a HUGE speedup: openssl speed -elapsed shows AES is 90% faster,
> blowfish is 75% faster, and sha1 is 50% faster. Not so bad!
> However, doing this increases the pressure on the processor bus, so it is
> necessary to increase the processor bus timeout on 40MHz boards again (to
> 256
> usec). ``black cat'' 50MHz boards seem to be unaffected, so they remain at
> 64 usec. (miod@)
~ m188_machdep.c ~ machdep.c
> Make the mvme188 interrupt handling closer to the aviion interrupt
> handling:
> non-VME syscon interrupt sources will now use their own intrhand array,
> and interrupt sources will be enabled in the arbiter as interrupt handlers
> are registered. This allows VME devices to use the whole 256 interrupts
> range. (miod@)
~ m197_machdep.c
> Remove leftover code in m197_machdep.c, back when I wanted to address the
> DCAM2 boards in a different way. (miod@)
arch/sgi/dev
~ power.c
> power(4) can now make use of the DS1687 register definitions header file.
> ok jasper@ (jsing@)
arch/sgi/sgi
~ clock_md.c
> Clean up the clock MD code. Move all of the DS1687 register definitions to
> the appropriate place and remove all of the magic numbers. Whilst we're
> here fix up some KNF issues. Also, the get/set functions used the wrong
> control registers, allowing the clock to update during operation.
> ok miod@, jasper@ (jsing@)
arch/sparc/dev
~ si.c
> fix comments after the members of scsi_adapter were changed.
> prompted by krw@ (dlg@)
arch/sparc64/sparc64
~ locore.s
> Remove DMMU_MISS_2; it's identical to DMMU_MISS. (kettenis@)
~ pmap.c
> In pmap_dumpmmu(), account for the fact that we use 4MB mappings for text
> and
> data segments. This makes the complete mappings available in kernel dumps.
> (kettenis@)
~ trap.c
> Use the TSTATE_xxx constants instead of (PSTATE_xxx << TSTATE_PSTATE_SHIFT)
> for readability; ok kettenis@ (miod@)
arch/vax/vax
~ machdep.c
> There is no need, and no point, to initialize panicstr here. (miod@)
~ wscons_machdep.c
> If initializing a frame buffer as glass console fails, instead of an
> invisible
> panic (for you ichc fans out there), disable the wscons console and force a
> console device reselection, which ends up in picking a serial console.
> This should not happen, but just in case, it's a less rude behaviour.
> (miod@)
arch/vax/vsa
~ dz_ibus.c
> Explicitely mention we are the console device, when applicable. (miod@)
~ gpx.c ~ lcg.c
~ lcspx.c ~ smg.c
> If initializing a frame buffer as glass console fails, instead of an
> invisible
> panic (for you ichc fans out there), disable the wscons console and force a
> console device reselection, which ends up in picking a serial console.
> This should not happen, but just in case, it's a less rude behaviour.
> (miod@)
~ ncr.c
> fix comments after the members of scsi_adapter were changed.
> prompted by krw@ (dlg@)
arch/zaurus/dev
~ zaurus_scoop.c
> use TAILQ_FOREACH when possible instead of manually
> rolling it.
> ok todd@, miod@; (thib@)
dev
~ bio.c
> use sizeof where needed (deraadt@)
~ rnd.c
> switch from spls to mutexes for protecting the random state.
> ok tedu@ djm@ (dlg@)
dev/acpi
~ acpicpu.c
> KNF, no functional change (marco@)
~ acpicpu.c
> Add _PPC support. Tested by several people. (marco@)
~ acpicpu.c
> Undo previous commit after debating a few corner cases with gwk. This
> needs to be redone in another way. (marco@)
~ acpicpu.c
> Remove debug message. Found by Mike Belopuhov, tsk tsk deraadt. (marco@)
dev/ata
~ atascsi.c
> implement emulation of vpd page 83, the device identification page. if the
> disk provides a wwn, use it, otherwise do our best to generate an id from
> the model and serial number. (dlg@)
~ atascsi.c ~ atascsi.h
> massage the way vpds are defined. rename the page used for the devices
> serial number and split the vpd header out for use in other places. while
> here define the device identification page bits too.
> ok krw@ marco@ (dlg@)
~ atascsi.c
> rename some functions to make it clear theyre used to emulate vpd responses
> (dlg@)
~ atascsi.c
> provide vpd page 0, which lists which vpd pages we provide. (dlg@)
~ atascsi.c
> the scsi layer always had function pointers for asking the hba about a
> device before issuing scsi commands to it, but it was never implemented,
> never used, and no hba actually filled them in. i came along and added
> another two function pointers for the same thing.
> this cleans up the extra pointers.
> ok krw@ marco@ miod@ deraadt@ (dlg@)
~ atascsi.c
> dont freeze lock or set caching modes on ata devices that arent disks, its
> not a good idea.
> found by simon@ who plugged a sata dvd drive into ahci. (dlg@)
dev/ic
~ mpi.c
> use the right flags when creating dmamaps for use in interrupt handlers.
> (dlg@)
~ mpi.c
> oops, there was other code in the previous commit that shouldnt have gone
> in. this reverts 1.90. (dlg@)
~ mpi.c
> 1.90 again. use the right flags when creating a dmamap for use during
> interrupts. (dlg@)
+ ds1687reg.h
> Clean up the clock MD code. Move all of the DS1687 register definitions to
> the appropriate place and remove all of the magic numbers. Whilst we're
> here fix up some KNF issues. Also, the get/set functions used the wrong
> control registers, allowing the clock to update during operation.
> ok miod@, jasper@ (jsing@)
~ ami.c
> massage the way vpds are defined. rename the page used for the devices
> serial number and split the vpd header out for use in other places. while
> here define the device identification page bits too.
> ok krw@ marco@ (dlg@)
~ iha.c
> fix comments after the members of scsi_adapter were changed.
> prompted by krw@ (dlg@)
~ pgt.c
> Only do pullups when necessary, m_pullup() always prepends an mbuf
> which is very bad if it is not necessary as it causes scrary mbuf
> fragmentation.
> tested and OK mglocker@ (claudio@)
dev/isa
~ it.c ~ itvar.h
> Look for environment controller first, and access ITE SuperIO address/data
> ports only if we found one. Fixes possible issues with some notebook
> chipsets.
> ok deraadt@
> tested by me, jcs@ and Rodolfo Gouveia (form@)
~ it.c
> alwais unmap i/o space in it_match (form@)
dev/pci
~ pcidevs
> add the PCI id for the Sun Crypto Accelerator 6000 PCIe board.
> ok dlg@ (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> - rename some SATA entries which are shared between the ATI SB700 and SB800
> chipsets
> - add some new SATA entries for the ATI SB700 and SB800 chipsets
> - add the ATI SB700 IDE PCI id
> From Linux
> ok dlg@ (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
kern
~ subr_disk.c ~ kern_sysctl.c
> protect the disk statistics with a mutex.
> ok tedu@ kettenis@ (dlg@)
~ vfs_default.c
> fifofs, deadfs and specfs all have the same "trivial lookup routine that
> always fails". Introduce vop_generic_lookup(), which is a trivial lookup
> routine that always fails and use that instead, zap the redundant copies.
> ok toby@, tedu@, art@ (thib@)
~ kern_tc.c
> allow for max 5000 uses/sec offset adjust, this makes it possible for
> clocks with drifts larger than 500ppm to be corrected. (otto@)
~ subr_prf.c
> Remove unused variable consintr. ok mbalmer (bluhm@)
lib/libkern
~ milieu.h ~ softfloat-macros.h
~ softfloat-specialize.h ~ softfloat.c
~ softfloat.h
> Unifdef NO_IEEE in the libkern softfloat code, and put an appropriate rule
> in files.alpha to compensate for NO_IEEE kernels.
> This will allow the softfloat code to be used by other platforms than
> alpha. (miod@)
~ softfloat-macros.h
> Override SoftFloat's countLeadingZeros32() with a faster, ff1-based,
> version
> on m88k. (miod@)
miscfs/deadfs
~ dead_vnops.c
> fifofs, deadfs and specfs all have the same "trivial lookup routine that
> always fails". Introduce vop_generic_lookup(), which is a trivial lookup
> routine that always fails and use that instead, zap the redundant copies.
> ok toby@, tedu@, art@ (thib@)
miscfs/fifofs
~ fifo.h ~ fifo_vnops.c
> fifofs, deadfs and specfs all have the same "trivial lookup routine that
> always fails". Introduce vop_generic_lookup(), which is a trivial lookup
> routine that always fails and use that instead, zap the redundant copies.
> ok toby@, tedu@, art@ (thib@)
miscfs/specfs
~ specdev.h ~ spec_vnops.c
> fifofs, deadfs and specfs all have the same "trivial lookup routine that
> always fails". Introduce vop_generic_lookup(), which is a trivial lookup
> routine that always fails and use that instead, zap the redundant copies.
> ok toby@, tedu@, art@ (thib@)
net
~ pf_norm.c
> Make "scrub max-mss" rule work correctly;
> In pf_normalize_tcpopt() pull the TCP options before processing them.
> This gets the correct TCP options even if an mbuf chain was used, instead
> like now pointing into an invalid mbuf data buffer.
> Will close PR 5623. Diff done together with dhartmei at .
> OK dhartmei@ (mglocker@)
~ pf_norm.c
> In pf_normalize_tcpopt() call pf_pull_hdr() address family safe.
> OK dhartmei@ (mglocker@)
netinet
~ in_pcb.c
> A struct rt_addrinfo is needed for rt_missmsg() so we can use rtrequest1()
> directly instead of takeing the detour via rtrequest().
> OK henning@ markus@ (claudio@)
nfs
~ nfs_serv.c
> Fix for PR5365.
> Use the va_blocksize, that is the blocksize of the underlying exported
> file system instead of the DIRBLKSIZ constant to size the chunks.
> Diff from Enache Adrian <3n4ch3 at gmail.com>
> Tested by myself and a few others. (thib@)
~ nfs_socket.c
> nfs_sndlock() is often called with its rep argument
> as NULL, so make sure its non-NULL before attempting
> to reference it.
> Fixes PR5694; Confirmed by submitter.
> OK art@, krw@ (thib@)
scsi
~ scsi_all.h ~ sd.c
> massage the way vpds are defined. rename the page used for the devices
> serial number and split the vpd header out for use in other places. while
> here define the device identification page bits too.
> ok krw@ marco@ (dlg@)
~ scsiconf.h
> the scsi layer always had function pointers for asking the hba about a
> device before issuing scsi commands to it, but it was never implemented,
> never used, and no hba actually filled them in. i came along and added
> another two function pointers for the same thing.
> this cleans up the extra pointers.
> ok krw@ marco@ miod@ deraadt@ (dlg@)
sys
~ disk.h
> protect the disk statistics with a mutex.
> ok tedu@ kettenis@ (dlg@)
~ vnode.h
> fifofs, deadfs and specfs all have the same "trivial lookup routine that
> always fails". Introduce vop_generic_lookup(), which is a trivial lookup
> routine that always fails and use that instead, zap the redundant copies.
> ok toby@, tedu@, art@ (thib@)
== usr.bin =========================================================== 10/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
pcc
~ ccom/pftn.c
> Pull from master repo:
> Set type to INT if it's still UNDEF upon return from typenode().
> ok ragge@ (stefan@)
~ powerpc/code.c
> Pull from master repo:
> Use sclass and soffset to access hdr.h_sclass and hdr.h_offset. (stefan@)
quota
~ quota.c
> lowercase "usage:"; align synopses
> ok jmc@ (sobrado@)
rcs
~ diff3.c
> increase the marker buffers, they are not big enough for hg merge (tedu@)
sectok
~ cyberflex.c ~ main.c
~ sectok.1
> add missing arguments to synopsis; fix brackets; lowercase "usage:"
> ok jmc@ (sobrado@)
~ sectok.1
> new sentence, new line; (jmc@)
ssh
~ canohost.c ~ channels.c
~ misc.c ~ misc.h
~ servconf.c ~ ssh-keyscan.c
~ sshconnect.c ~ sshd.c
> Add a small helper function to consistently handle the EAI_SYSTEM error
> code of getaddrinfo. Prompted by vgiffin at apple com via bz #1417.
> ok markus@ stevesk@ (dtucker@)
~ clientloop.c ~ packet.c
~ serverloop.c
> Make SSH2_MSG_UNIMPLEMENTED and SSH2_MSG_IGNORE messages reset the
> ServerAlive and ClientAlive timers. Prevents dropping a connection
> when these are enabled but the peer does not support our keepalives.
> bz #1307, ok djm at . (dtucker@)
~ clientloop.c
> Use the correct packet maximum sizes for remote port and agent forwarding.
> Prevents the server from killing the connection if too much data is queued
> and an excessively large packet gets sent. bz #1360, ok djm at . (dtucker@)
vmstat
~ vmstat.c
> minor improvement on the diff to r1.106
> ok jsing@ (sobrado@)
== usr.sbin ========================================================== 11/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
bgpctl
~ bgpctl.c
> send timers for bgpctlshow neighbor foo timer in seperate messages after
> the peer data. makes bgpctl display code independent from timer
> implementation internals. only running timers are displayed now, stopped
> ones are skipped. (henning@)
bgpd
~ timer.c
> make all timer_get() callers (except timer_set which will need
> special handling) and onein session.c that will go away able to
> cope with timer_get() returning NULL (i.e.nosuch timer/not running)
> (henning@)
~ session.c
> use timer_running (with left = NULL, which the last commit made possible)
> instead of timer_get and checking manually (henning@)
~ bgpd.h ~ control.c
~ log.h ~ session.h
~ timer.c
> send timers for bgpctlshow neighbor foo timer in seperate messages after
> the peer data. makes bgpctl display code independent from timer
> implementation internals. only running timers are displayed now, stopped
> ones are skipped. (henning@)
~ session.c ~ session.h
~ timer.c
> provide timer_nextduein, which provides the number of seconds until the
> next timer expires.
> use that in the session engine's mainloop, which simplifies it and
> removes the last bits of timer internals knowledge from anywhere outside
> timer.c. (henning@)
dhcpd
~ pfutils.c
> remove double word in comment; (jmc@)
~ dhcpd.conf.5
> for "fixed-address", note that no lease is generated; after some
> discussion with krw, who hasn;t ok'd this diff, but we can change it if
> he doesn;t like it;
> started by mail from Max Hayden Chiz (jmc@)
~ dhcpd.c
> add missing space between argument names and ellipsis
> ok jmc@ (sobrado@)
ntpd
~ ntpd.h
> remove some unused #defines; ok henning@ (stevesk@)
~ ntpd.c
> log a warning one time when we can't open or write the drift file;
> also add "(no drift file)" to the adjfreq log message on failure;
> ok henning@ (stevesk@)
~ client.c ~ ntp.c
> some fatal() calls that should be fatalx(); ok henning@ (stevesk@)
pkg_add
~ OpenBSD/PkgCfl.pm ~ OpenBSD/Search.pm
> tweak the semantics so that no-default-conflict works.
> requested by mbalmer@,sturm at . okay winiger@ (espie@)
~ OpenBSD/PackageRepository/Source.pm
> allows src: urls to work better by using -F nolock during pkg builds.
> Also turns off ftp package fetching, as noticed by bernd@
> With this, src: urls should be mostly functional now. (espie@)
pstat
~ pstat.c
> printf field widths are always int, so add casts to remove
> "warning: field width is not type int"
> "looks good" otto@ (chl@)
snmpctl
~ parser.c ~ snmpctl.c
> use $OpenBSD$ tags.
> ok reyk@ (thib@)
~ Makefile ~ parser.c
~ parser.h ~ snmpctl.8
~ snmpctl.c
> add "show mib" command to dump compiled-in MIB names (reyk@)
snmpd
~ snmpe.c
> Make a snmpwalk against snmpd on hppa and macppc work. armish still dont
> work. -moj (maja@)
~ mib.c ~ mib.h
~ parse.y ~ snmpd.h
> use the snmp camelCase for MIB definitions, this allows to simplify
> the code a bit and is mostly a stylistic change. (reyk@)
~ kroute.c ~ parse.y
> spacing (reyk@)
~ mib.c ~ mib.h
~ mps.c ~ snmpd.h
> split the MIB declarations (names) from the definitions (implementation).
> (reyk@)
~ Makefile ~ kroute.c
~ mib.c ~ mps.c
~ parse.y ~ snmpd.c
~ snmpd.h ~ snmpe.c
+ smi.c
> split code handling the "structure of management information" and the
> "message processing subsystem" into separate files smi.c and mps.c (reyk@)
~ README
> sync (reyk@)
~ parse.y
> set the default port in host byte order to get the correct result (reyk@)
~ ber.h
> shrink the max number of supported OID elements from 128 to 32. (reyk@)
~ kroute.c ~ mib.c
~ mib.h ~ mps.c
~ smi.c ~ snmpd.h
> add the IP-MIB ipAddrTable. it requires to encode the ipv4 addresses
> in the OIDs as the table index. the next step is to simplify the
> common mib implementation regarding the special requirements of these
> strange snmp tables. (reyk@)
===============================================================================
More information about the owc
mailing list