[owc] Weekly src changes ending 2008-10-12
OWC
auto at squish.net
Mon Oct 13 07:00:01 BST 2008
OpenBSD src changes summary for 2008-10-05 to 2008-10-12 inclusive
==================================================================
bin/dd distrib/sets
etc/MAKEDEV.common etc/etc.alpha/MAKEDEV
etc/etc.amd64/MAKEDEV etc/etc.armish/MAKEDEV
etc/etc.aviion/MAKEDEV etc/etc.hp300/MAKEDEV
etc/etc.hppa/MAKEDEV etc/etc.hppa64/MAKEDEV
etc/etc.i386/MAKEDEV etc/etc.landisk/MAKEDEV
etc/etc.luna88k/MAKEDEV etc/etc.mac68k/MAKEDEV
etc/etc.macppc/MAKEDEV etc/etc.mvme68k/MAKEDEV
etc/etc.mvme88k/MAKEDEV etc/etc.sgi/MAKEDEV
etc/etc.socppc/MAKEDEV etc/etc.sparc/MAKEDEV
etc/etc.sparc64/MAKEDEV etc/etc.vax/MAKEDEV
etc/etc.zaurus/MAKEDEV etc/ntpd.conf
gnu/usr.bin/perl gnu/usr.bin/texinfo
lib/csu lib/libc
lib/libm regress/usr.bin
sbin/dhclient share/man
share/zoneinfo sys/arch/alpha/alpha
sys/arch/alpha/include sys/arch/amd64/amd64
sys/arch/amd64/conf sys/arch/arm/include
sys/arch/hppa/include sys/arch/hppa64/include
sys/arch/i386/conf sys/arch/i386/i386
sys/arch/i386/include sys/arch/m68k/include
sys/arch/m88k/include sys/arch/mac68k/dev
sys/arch/macppc/dev sys/arch/mips64/include
sys/arch/powerpc/include sys/arch/sgi/conf
sys/arch/sh/include sys/arch/socppc/dev
sys/arch/socppc/socppc sys/arch/solbourne/solbourne
sys/arch/sparc/sparc sys/arch/sparc64/dev
sys/arch/sparc64/include sys/arch/vax/bi
sys/arch/vax/conf sys/arch/vax/if
sys/arch/vax/include sys/arch/vax/vax
sys/conf sys/dev
sys/dev/acpi sys/dev/cardbus
sys/dev/ic sys/dev/isa
sys/dev/mii sys/dev/pci
sys/dev/usb sys/dev/wscons
sys/kern sys/lib/libkern
sys/net sys/netinet
sys/sys sys/uvm
usr.bin/awk usr.bin/chpass
usr.bin/compress usr.bin/fstat
usr.bin/mail usr.bin/rcs
usr.bin/sed usr.bin/sendbug
usr.bin/ssh usr.bin/stat
usr.bin/systat usr.bin/vmstat
usr.sbin/authpf usr.sbin/httpd
usr.sbin/mtree usr.sbin/ntpd
usr.sbin/pcidump usr.sbin/pkg_add
usr.sbin/pppd usr.sbin/pstat
usr.sbin/relayd usr.sbin/snmpd
usr.sbin/tcpdump usr.sbin/user
usr.sbin/ypldap
== bin =============================================================== 01/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/bin
dd
~ dd.1
> the ellipsis allow more than one operand being specified.
> ok jmc@ (sobrado@)
== distrib =========================================================== 02/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/distrib
sets
~ lists/man/mi
> sync (deraadt@)
== etc =============================================================== 03/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/etc
MAKEDEV.common
~ MAKEDEV.common
> skip making /dev/prandom. as a side effect, /dev/arandom becomes minor 3.
> but that is OK because both minor 3 and 4 now do the same thing.
> eventually,
> 2 releases from now, everyone will have forgotten about minor 4 and we can
> delete the support code for it from sys/dev/rnd.c
> various versions of this looked at by millert and todd (deraadt@)
etc.alpha/MAKEDEV
~ etc.alpha/MAKEDEV
> sync (deraadt@)
etc.amd64/MAKEDEV
~ etc.amd64/MAKEDEV
> sync (deraadt@)
etc.armish/MAKEDEV
~ etc.armish/MAKEDEV
> sync (deraadt@)
etc.aviion/MAKEDEV
~ etc.aviion/MAKEDEV
> sync (deraadt@)
etc.hp300/MAKEDEV
~ etc.hp300/MAKEDEV
> sync (deraadt@)
etc.hppa/MAKEDEV
~ etc.hppa/MAKEDEV
> sync (deraadt@)
etc.hppa64/MAKEDEV
~ etc.hppa64/MAKEDEV
> sync (deraadt@)
etc.i386/MAKEDEV
~ etc.i386/MAKEDEV
> sync (deraadt@)
etc.landisk/MAKEDEV
~ etc.landisk/MAKEDEV
> sync (deraadt@)
etc.luna88k/MAKEDEV
~ etc.luna88k/MAKEDEV
> sync (deraadt@)
etc.mac68k/MAKEDEV
~ etc.mac68k/MAKEDEV
> sync (deraadt@)
etc.macppc/MAKEDEV
~ etc.macppc/MAKEDEV
> sync (deraadt@)
etc.mvme68k/MAKEDEV
~ etc.mvme68k/MAKEDEV
> sync (deraadt@)
etc.mvme88k/MAKEDEV
~ etc.mvme88k/MAKEDEV
> sync (deraadt@)
etc.sgi/MAKEDEV
~ etc.sgi/MAKEDEV
> sync (deraadt@)
etc.socppc/MAKEDEV
~ etc.socppc/MAKEDEV
> sync (deraadt@)
etc.sparc/MAKEDEV
~ etc.sparc/MAKEDEV
> sync (deraadt@)
etc.sparc64/MAKEDEV
~ etc.sparc64/MAKEDEV
> sync (deraadt@)
etc.vax/MAKEDEV
~ etc.vax/MAKEDEV
> sync (deraadt@)
etc.zaurus/MAKEDEV
~ etc.zaurus/MAKEDEV
> sync (deraadt@)
ntpd.conf
~ ntpd.conf
> show examples for sensors. use nmea as suggested by mbalmer, henning
> agrees. (sthen@)
== gnu =============================================================== 04/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/gnu
usr.bin/perl
~ lib/ExtUtils/t/parse_version.t
> Prevent cvs keyword expansion from causing a test failure. (millert@)
~ Makefile.bsd-wrapper
> re-enable ieeefp mode on alpha again, now that the kernel code is
> fixed. tested by me and naddy at . ok theo (martynas@)
~ config.over ~ installperl
~ lib/ExtUtils/MM_Unix.pm
> Remove hack from MM_Unix.pm and simply set man1dir to 'none' in
> installperl to prevent man page installation of the perl man pages.
> (millert@)
usr.bin/texinfo
~ makeinfo/xml.c
> prevent access to output_paragraph with negative offset; PR 5518; ok
> deraadt@ (otto@)
== lib =============================================================== 05/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
csu
~ alpha/Makefile ~ amd64/Makefile
~ arm/Makefile ~ hppa/Makefile
~ hppa64/Makefile ~ i386/Makefile
~ powerpc/Makefile ~ sparc/Makefile
~ sparc64/Makefile
> The recent change to bsd.own.mk to allow PICFLAG to be overridden made
> these previously ignored PICFLAG settings become active. Remove PICFLAG
> overrides to restore -fpic/-fPIC modes. In snaps for a week.
> okay drahn@ (kurt@)
libc
~ locale/wcstod.c
> Fix conversion of numbers without an exponent.
> ok martynas@ (kettenis@)
~ time/tz-art.htm
> update to tzcode2008g from elsie.nci.nih.gov (millert@)
libm
- arch/vax/n_cabs.S ~ Makefile
~ noieee_src/n_cabs.c + noieee_src/n_cacos.c
+ noieee_src/n_cacosh.c + noieee_src/n_carg.c
+ noieee_src/n_casin.c + noieee_src/n_casinh.c
+ noieee_src/n_catan.c + noieee_src/n_catanh.c
+ noieee_src/n_ccos.c + noieee_src/n_ccosh.c
+ noieee_src/n_cexp.c + noieee_src/n_cimag.c
+ noieee_src/n_cimagf.c + noieee_src/n_clog.c
+ noieee_src/n_conj.c + noieee_src/n_conjf.c
+ noieee_src/n_cpow.c + noieee_src/n_cproj.c
+ noieee_src/n_cprojf.c + noieee_src/n_creal.c
+ noieee_src/n_crealf.c + noieee_src/n_csin.c
+ noieee_src/n_csinh.c + noieee_src/n_csqrt.c
+ noieee_src/n_ctan.c + noieee_src/n_ctanh.c
+ noieee_src/n_hypot.c + arch/vax/n_hypot.S
> - noieee_src: adapt complex versions of the functions it already
> supports (which is all, except the float ones)
> ok millert@ (martynas@)
== regress =========================================================== 06/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/regress
usr.bin
+ sed/Makefile + sed/hanoi.expected
+ sed/hanoi.sed + sed/math.expected
+ sed/math.in + sed/math.sed
+ sed/sedtest.expected + sed/sedtest.sh
+ sed/sierpinski.expected + sed/sierpinski.sed
> Adapt sed test suite to regress framework.
> Add sierpinski triangle script from
> http://sed.sourceforge.net/grabbag/scripts (millert@)
~ Makefile
> hook up sed regress (millert@)
== sbin ============================================================== 07/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
dhclient
~ dhclient.conf.5
> Clarify discussion of option modifiers. They only affect values passed
> to dhclient-script. Our default script only uses a few options so
> others are unaffected by option modifiers. PR#5834 is thus rendered
> moot.
> Feedback & suggestions from jmc@, ok deraadt at . (krw@)
~ dhclient.conf.5
> Oops. Missed one typo jmc@ spotted. (krw@)
~ dhclient.conf.5
> - talk about the "client configuration script" consistenly
> - Xr dhclient-script on first mention, which is section 8 not 5
> ok krw (jmc@)
== share ============================================================= 08/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man8/release.8
> Mention that /usr/src is now needed to build xenocara.
> ok jmc@, oga at . (matthieu@)
~ man4/umsm.4
> sync man page
> "Option N.V. Globetrotter HSDPA Modem" is moved to umsm(4) from ubsa(4)
> (yuo@)
~ man4/pflow.4
> write "pflow", "flowsrc" and "flowdst" using the right macros. (sobrado@)
~ man4/Makefile ~ man4/random.4
> prandom dies (deraadt@)
~ man5/port-modules.5
> sync tcl/tk description with reality (steven@)
zoneinfo
~ datfiles/southamerica
> update to tzdata2008g from elsie.nci.nih.gov (millert@)
== sys =============================================================== 09/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/alpha/alpha
~ fp_complete.c
> unbreak ieeefp emulation code wrt converting double to unsigned
> long ints for alpha. we've got only one instruction (cvttq) to
> convert double-t to quadword, and float64_to_int64 did not take
> into account the unsigned conversions
> therefore, overflow always occured, and half of the unsigned range
> (LONG_MAX .. ULONG_MAX) was broken
> introduce roundAndPackInt64NoOverflow and float64_to_int64_no_overflow
> for softfloat, that works with unsigned integers as well. note
> that this will return zero for nan/inf/oflow/uflow, raising exception
> flag
> perl is happy now
> looked over by miod@
> tested by naddy@, and by me on nick@'s alpha (martynas@)
arch/alpha/include
~ vmparam.h
> more readable virtual memory size values.
> ok deraadt@ (brad@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/amd64/amd64
~ cpu.c ~ db_disasm.c
> remove dead stores and newly created unused variables.
> Found by LLVM/Clang Static Analyzer.
> ok weingart@ (chl@)
~ machdep.c
> Remove useless comment. (kettenis@)
~ db_memrw.c
> remove a dead store.
> Found by LLVM/Clang Static Analyzer.
> ok weingart@ (chl@)
~ intr.c
> replace M_WAITOK with M_NOWAIT
> ok krw@ kettenis@ (chl@)
~ machdep.c
> Skip memory areas smaller than 1 MB above the 16 MB boundary to deal with
> buggy ACP BIOSes. Based on a diff from weingart at .
> ok deraadt@, oga@ (kettenis@)
arch/amd64/conf
~ GENERIC
> uhci(4) at cardbus(4) seems to work (to the same extent that ohci(4) and
> ehci(4) work; ie. it has the same variety of unplug bugs) (deraadt@)
arch/arm/include
~ cpu.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/hppa/include
~ cpu.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/hppa64/include
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/i386/conf
~ GENERIC
> uhci(4) at cardbus(4) seems to work (to the same extent that ohci(4) and
> ehci(4) work; ie. it has the same variety of unplug bugs) (deraadt@)
~ files.i386
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
arch/i386/i386
~ cpu.c ~ ioapic.c
~ mpbios.c
> Remove IOAPIC ID remapping code. It should not be necessary to avoid
> reusing
> LAPIC IDs and if there are MPBIOSes we handle them with ACPI now.
> ok deraadt@ (kettenis@)
~ machdep.c
> Unconditionally compile idt_vec_xxx functions. (kettenis@)
~ machdep.c
> Skip memory areas smaller than 1 MB above the 16 MB boundary to deal with
> buggy ACP BIOSes. Based on a diff from weingart at .
> ok deraadt@, oga@ (kettenis@)
- random.s
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
arch/i386/include
~ i82093var.h
> Remove IOAPIC ID remapping code. It should not be necessary to avoid
> reusing
> LAPIC IDs and if there are MPBIOSes we handle them with ACPI now.
> ok deraadt@ (kettenis@)
arch/m68k/include
~ cpu.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/m88k/include
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/mac68k/dev
~ if_mc.c ~ if_sn.c
> cosmetic change for ioctl funtions.. move splnet out from variable
> declaration. (brad@)
arch/macppc/dev
~ if_mc.c
> cosmetic change for ioctl funtions.. move splnet out from variable
> declaration. (brad@)
arch/mips64/include
~ cpu.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/powerpc/include
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
arch/sgi/conf
~ GENERIC
> - add and enable rum(4)
> committed over a rum(4) donated by Maxim Belooussov, thanks!
> ok jsing@ (jasper@)
arch/sh/include
~ cpu.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/socppc/dev
~ ipic.c
> Adjust for redesigned powerpc interrupt architecture. (kettenis@)
arch/socppc/socppc
~ machdep.c
> Adjust for redesigned powerpc interrupt architecture. (kettenis@)
~ clock.c
> Reduce diff to macppc version. (kettenis@)
arch/solbourne/solbourne
~ locore.s
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
arch/sparc/sparc
~ cpuvar.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpuvar.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpuvar.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
~ locore.s
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
arch/sparc64/dev
~ sbbc.c
> Add missing argument to printf. (kettenis@)
~ vbus.c ~ vbusvar.h
> Add interrupt support. (kettenis@)
~ vcons.c
> Make vcons(4) interrupt driven. (kettenis@)
arch/sparc64/include
~ cpu.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
arch/vax/bi
~ if_ni.c
> cosmetic change for ioctl funtions.. move splnet out from variable
> declaration. (brad@)
arch/vax/conf
~ files.vax
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
arch/vax/if
~ if_qe.c
> cosmetic change for ioctl funtions.. move splnet out from variable
> declaration. (brad@)
~ if_de.c ~ sgec.c
> cosmetic change for ioctl funtions.. move splnet out from variable
> declaration. (brad@)
~ if_de.c
> Use ether_ioctl(). (brad@)
arch/vax/include
~ cpu.h
> Implement CPU_INFO_UNIT for everyone, not just MP kernels.
> ok miod@ (art@)
~ cpu.h
> Define MAXCPUS on all architectures.
> For now, sparc64 is arbitrarily set to 256 (only architecture that didn't
> have
> a practical limit in the code on the number of cpus). (art@)
~ cpu.h
> Add empty cpu_unidle() macros for architectures that currently don't do
> anything special to prod a cpu to leave the idle loop in signotify.
> powerpc, i386, amd64 and sparc64 will follow soon so that everyone has
> the same interface to wake an idling cpu. (art@)
arch/vax/vax
- random.s
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
conf
~ files
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
dev
~ rnd.c ~ rndvar.h
> minor 3 was /dev/prandom, which used random(). make that use the
> arandom code, just like minor 4. in fact, make them act exactly
> the same. (deraadt@)
~ softraid.c
> Fix various printf constructs either missing arguments or format
> specifiers. (miod@)
dev/acpi
~ dsdt.c ~ dsdt.h
> Fix memory leak bug during scope evaluation (now without introducing a bug
> in evaluation of the Return() operator). Based jordan@'s diff from rev.
> 1.133.
> ok jordan@, marco@ (kettenis@)
~ acpicpu.c
> Fix various printf constructs either missing arguments or format
> specifiers. (miod@)
dev/cardbus
~ if_re_cardbus.c
> If re_attach() fails at any point make sure to disestablish various
> resources setup by the bus front ends, especially the shutdown hook.
> Found while testing some other changes with a CardBus adapter and
> re_attach() was failing early before the interface pointer was
> assigned. Then the system was rebooted, the shutdown hook was called
> and *boom* in re_stop(). (brad@)
dev/ic
~ rtl81x9reg.h
> Detect if the adapter is a PCIe adapter and set the RL_FLAG_PCIE flag
> if so. (brad@)
~ rtl81x9reg.h
> Add a flag to indicate a 64-bit PCI bus is present. (brad@)
~ re.c ~ rtl81x9reg.h
> Detect and store the PCI/PCIe bus speed.
> From DragonFly (brad@)
~ mpi.c
> if fetching a config page for a sas target doesnt work then let the scsi
> midlayer try to probe it anyway. this lets raid devices configured on an
> mpi to work again.
> reported by djm@ (dlg@)
~ re.c
> Set hardware timer according to bus clock speed.
> From sephe at DragonFly (brad@)
~ re.c ~ rtl81x9reg.h
> Use hardware timer to simulate interrupt moderation. Old devices will no
> longer be livelocked when they are receiving on GigE line. Newer devices
> also gain well controlled interrupt rate.
> If hardware supports interrupt moderation (e.g. 8168B, 8168C), you could
> also use hardware based interrupt moderation, however, due to lack of
> necessary information it does not work as reliably as simulated interrupt
> moderation. It is _not_ recommended currently.
> By default, PCIe devices' simulated interrupt moderation timer is set to
> 75us, while PCI devices' is set to 125us.
> From sephe at DragonFly (brad@)
dev/isa
~ it.c ~ itvar.h
> o chip revision is 4-bit value
> o some #define's instead of digits (form@)
~ it.c
> Older IT8712F chips have 8-bit watchdog timeout counter.
> Use minutes for 16-bit values. (form@)
~ it.c
> better comments (form@)
~ it.c
> Fix various printf constructs either missing arguments or format
> specifiers. (miod@)
dev/mii
~ eephy.c
> Make sure page 0 is selected when we initialize the PHY. Fixes problems
> with the eephy(4) that attaches to nfe(4) on machines like the Sun Ultra
> 40.
> ok deraadt@ (kettenis@)
dev/pci
~ if_em.c
> Always update published link state even if the internal link state doesn't
> change. Prevents us from getting stuck in LINK_STATE_UNKNOWN. Fixes PR
> 5914.
> tested by deraadt@, sthen@
> ok deraadt@ (kettenis@)
~ if_re_pci.c
> Detect if the adapter is a PCIe adapter and set the RL_FLAG_PCIE flag
> if so. (brad@)
~ pcidevs
> another geforce (deraadt@)
~ pcidevs.h ~ pcidevs_data.h
> sync (deraadt@)
~ pcidevs.h ~ pcidevs_data.h
> regen (jsg@)
~ pcidevs
> Add some devices from submitted dmesgs. (jsg@)
~ if_lii.c
> If we fail to read the MAC address from the eeprom, fallback to reading
> the MAC address registers. In most cases these already have the correct
> MAC address. This should address PR5743. Thanks to Rodolfo Gouveia for
> testing earlier diffs.
> ok dlg@ (jsing@)
~ drm/drmP.h ~ drm/drm_agpsupport.c
~ drm/drm_bufs.c ~ drm/drm_context.c
~ drm/drm_dma.c ~ drm/drm_drv.c
~ drm/drm_ioctl.c ~ drm/drm_irq.c
~ drm/drm_lock.c ~ drm/i915_dma.c
~ drm/i915_drv.c ~ drm/i915_drv.h
~ drm/mach64_drv.c ~ drm/mach64_drv.h
~ drm/mach64_state.c ~ drm/mga_drv.c
~ drm/mga_drv.h ~ drm/mga_state.c
~ drm/r128_drv.c ~ drm/r128_drv.h
~ drm/r128_state.c ~ drm/radeon_drv.c
~ drm/radeon_drv.h ~ drm/radeon_state.c
~ drm/savage_bci.c ~ drm/savage_drv.c
~ drm/savage_drv.h ~ drm/sis_drv.c
~ drm/sis_drv.h ~ drm/sis_mm.c
~ drm/tdfx_drv.c ~ drm/via_dma.c
~ drm/via_drv.c ~ drm/via_drv.h
> Move dev->driver over to being a pointer to a const struct, instead of
> stupidly
> filling in a pre-allocated one on each attach.
> Makes the code a bunch nicer, shrinks a kernel by about 1.5k on amd64,
> helps with my sanity, and paves way for later changes.
> Tested by a few for a couple of weeks now. (oga@)
~ drm/drmP.h ~ drm/drm_dma.c
~ drm/drm_drv.c ~ drm/i915_dma.c
~ drm/i915_irq.c ~ drm/i915_mem.c
~ drm/mach64_dma.c ~ drm/mach64_drv.h
~ drm/mach64_irq.c ~ drm/mach64_state.c
~ drm/mga_dma.c ~ drm/mga_drv.h
~ drm/mga_state.c ~ drm/mga_warp.c
~ drm/r128_cce.c ~ drm/r128_drv.h
~ drm/r128_state.c ~ drm/r300_cmdbuf.c
~ drm/radeon_cp.c ~ drm/radeon_irq.c
~ drm/radeon_mem.c ~ drm/radeon_state.c
~ drm/savage_bci.c ~ drm/savage_state.c
~ drm/sis_mm.c ~ drm/via_dma.c
~ drm/via_irq.c ~ drm/via_map.c
~ drm/via_mm.c ~ drm/via_verifier.c
~ drm/via_video.c
> Kill the linux-ready negative return codes in ``shared'' code. We handle
> them wrong in several cases that i've noticed and Merging when needed is
> still fairly simple, anyway. This shaves another 500 bytes from an amd64
> kernel due to not having to flip the sign on some things. It also stops
> my eyes bleeding.
> Tested by a few along with the last diff that went in. (oga@)
~ drm/i915_drv.c ~ drm/mach64_drv.c
~ drm/mga_drv.c ~ drm/r128_drv.c
~ drm/radeon_drv.c ~ drm/savage_drv.c
~ drm/sis_drv.c ~ drm/tdfx_drv.c
~ drm/via_drv.c
> Kill a bunch of #ifdef freebsd/netbsd code in *_drv.c. No binary change.
> (oga@)
~ pcidevs
> new PCI ID for Intel WiFi Link 5350. (damien@)
~ pcidevs.h ~ pcidevs_data.h
> sync (damien@)
~ azalia.c
> missing spl() for codec command setting/retrieval
> from NetBSD via Alexey Suslikov, thanks
> "looks good" ratchov@ (jakemsr@)
~ azalia.c
> Enable PCIe snooping on the remaining ATI/NVIDIA chipsets.
> Fixes clicking and other noise related issues.
> ok ratchov@ (brad@)
~ if_ix.c ~ drm/drm_drv.c
> Fix various printf constructs either missing arguments or format
> specifiers. (miod@)
~ if_re_pci.c
> If re_attach() fails at any point make sure to disestablish various
> resources setup by the bus front ends, especially the shutdown hook.
> Found while testing some other changes with a CardBus adapter and
> re_attach() was failing early before the interface pointer was
> assigned. Then the system was rebooted, the shutdown hook was called
> and *boom* in re_stop(). (brad@)
~ azalia.c
> add azalia_codec_disconnect_stream(), a function to direct all
> converters in a group to use HDA link stream #0. use it when
> halting playback/recording.
> from NetBSD via Alexey Suslikov, thanks
> tested by several, thanks also (jakemsr@)
~ azalia.c
> if converters in a group are not needed to satisfy the requirements
> of a stream, direct the unneeded converters use HDA link stream #0
> from NetBSD via Alexey Suslikov, thanks
> tested by several, thanks also (jakemsr@)
dev/usb
~ ehci.c
> Don't force multi transaction endpoints (>1024 bytes) down to single
> transactions. From NetBSD. (mglocker@)
~ ehci.c
> In ehci_free_itd() replace LIST_INSERT_AFTER(LIST_FIRST(..), ..) with
> LIST_INSERT_HEAD(..) to prevent a crash when the freelist is empty. From
> NetBSD. (mglocker@)
~ umsm.c ~ ubsa.c
> As the Option N.V. Globetrotter HSDPA Modem is not compatible
> with ubsa(4), it is moved to umsm(4).
> ok jsg@, reported by Daniele Pilenga (yuo@)
~ ulpt.c
> stop waiting for the printer to become ready, see PR 5427 for more
> information. Also explains PR 4124 and 5318 (deraadt@)
~ usbdi.c
> For none interrupt endpoints call the xfer done function before
> calling the xfer callback function to allow a device driver to
> immediately reschedule the same xfer.
> This e.g. improves performance for isoc high- fullspeed xfers (ehci).
> From NetBSD. OK phessler@ (mglocker@)
dev/wscons
~ wsdisplay.c
> Add missing checks for sc_focus being NULL. Found the hard way by
> Jesus Sanchez. (miod@)
kern
~ kern_malloc.c
> In malloc_page_free(), restore the correct wire_count value. (miod@)
~ kern_sysctl.c
> Do not display file offsets and a few other pieces of information, except
> to the user or the superuser. Display * for those fields instead. From
> PR 5113, but modified to use copyout correctly.
> comments from tedu, ok from others (deraadt@)
~ kern_prot.c
> Use the reference count in struct process to update the proc count
> for users by the correct amount when changing real UID on (r)threaded
> processes.
> ok art@ tedu@ (guenther@)
~ init_main.c ~ kern_exit.c
~ kern_fork.c
> Put a reference count in struct process to prevent use-after-free
> if the main thread reaches the reaper ahead of some other thread
> in the process.
> ok art@ tedu@ (guenther@)
~ kern_sysctl.c
> pass NULL rather than 0 for interrupt time usage to calcru()
> ok art@ (kevlo@)
~ uipc_socket.c
> Change sb_timeo to unsigned, so that even if some calculation (ie. n * HZ)
> becomes a very large number it will not wrap the short into a negative
> number and screw up timeouts. It will simply become a max of 65535. Since
> this happens when HZ is cranked to a high number, this will still only take
> n seconds, or less. Safer than crashing.
> Prompted by PR 5511
> ok guenther (deraadt@)
~ init_main.c ~ kern_exit.c
~ kern_fork.c ~ kern_prot.c
> backout; is causing some people difficulty (deraadt@)
~ kern_malloc.c
> Since malloc_page_alloc() is a pool allocator it should check for PR_WAITOK
> instead of M_NOWAIT. Checking for M_NOWAIT made many malloc calls that
> used
> that flag actually wait. This probably explains many if the strange hangs
> people have seen recently.
> ok miod@ (kettenis@)
lib/libkern
~ softfloat.c ~ softfloat.h
> unbreak ieeefp emulation code wrt converting double to unsigned
> long ints for alpha. we've got only one instruction (cvttq) to
> convert double-t to quadword, and float64_to_int64 did not take
> into account the unsigned conversions
> therefore, overflow always occured, and half of the unsigned range
> (LONG_MAX .. ULONG_MAX) was broken
> introduce roundAndPackInt64NoOverflow and float64_to_int64_no_overflow
> for softfloat, that works with unsigned integers as well. note
> that this will return zero for nan/inf/oflow/uflow, raising exception
> flag
> perl is happy now
> looked over by miod@
> tested by naddy@, and by me on nick@'s alpha (martynas@)
- arch/sparc/random.S
> Use the MI random() for all cases; no need for a MD version since it
> has never been performance sensitive.
> Running on all platforms, discussed with millert and kettenis, ok toby
> (deraadt@)
net
~ pf_table.c ~ pfvar.h
> Get rid of the second table entry pool (pfr_kentry_pl2); we're already
> using the default interrupt handler for both, so there's no need to keep
> table entries created in interrupt context separate.
> ok henning art (mcbride@)
~ if_ethersubr.c
> Add fallback ioctl handlers for MTU and multicast to ether_ioctl(). (brad@)
~ if_pppoe.c
> Make sure no two pppoe devices share the same sc_unique identifier. While
> there, fix some typos, and pass M_CANFAIL to all malloc() calls which use
> M_WAITOK but are tested for failure.
> test&ok brad@ (miod@)
~ trunklacp.c ~ trunklacp.h
> Some more cleanup. No binary change.
> OK reyk@ (mpf@)
netinet
~ tcp_input.c
> Comment out statekey code to stop 'panic: soreceive 3', which
> happens with IPv6 TCP traffic, until a better fix is found.
> patch from henning@
> proded by deraadt@ (dhill@)
~ tcp_input.c
> back out previous change. Another panic, not as frequent, and
> definitely not at will. (dhill@)
sys
~ sysctl.h
> Do not display file offsets and a few other pieces of information, except
> to the user or the superuser. Display * for those fields instead. From
> PR 5113, but modified to use copyout correctly.
> comments from tedu, ok from others (deraadt@)
~ proc.h
> Put a reference count in struct process to prevent use-after-free
> if the main thread reaches the reaper ahead of some other thread
> in the process.
> ok art@ tedu@ (guenther@)
~ socketvar.h
> Change sb_timeo to unsigned, so that even if some calculation (ie. n * HZ)
> becomes a very large number it will not wrap the short into a negative
> number and screw up timeouts. It will simply become a max of 65535. Since
> this happens when HZ is cranked to a high number, this will still only take
> n seconds, or less. Safer than crashing.
> Prompted by PR 5511
> ok guenther (deraadt@)
~ proc.h
> backout; is causing some people difficulty (deraadt@)
uvm
~ uvm_vnode.c
> uvn_attach message is purely diagnostic, not needed
> no ok's from anyone because they are all slacking (deraadt@)
~ uvm_amap.c ~ uvm_map.c
> Don't extend amaps beyond what their supposed maximum. This code path is
> not taken anymore, but it doesn't hurt to be correct.
> from NetBSD, through mickey in pr 5812
> prodded by otto@ (art@)
== usr.bin =========================================================== 10/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
awk
~ FIXES ~ awk.h
~ awkgram.y ~ b.c
~ lex.c ~ lib.c
~ main.c ~ maketab.c
~ run.c
> Update awk to version 20071023; OK deraadt@ (millert@)
chpass
~ util.c
> remove unused variable
> ok millert@ jsing@ (chl@)
compress
~ main.c
> For "gunzip -N", the name stored in the gzip header has no directory,
> so write the uncompressed file in the same directory we would use
> were the -N option not specified. Closes PR 5224 (millert@)
fstat
~ fstat.1 ~ fstat.c
> Do not display file offsets and a few other pieces of information, except
> to the user or the superuser. Display * for those fields instead. From
> PR 5113, but modified to use copyout correctly.
> comments from tedu, ok from others (deraadt@)
mail
~ mail.1
> document -d. ok millert@ (martynas@)
~ main.c
> - add -d to usage
> - sync -f arg name with man page synopsis (jmc@)
rcs
~ rcs.c
> cvs(1) uses a config file to know about the $OpenBSD$ expansion keyword,
> but unfortunately rcs is not so lucky, and it will have to encoded inside.
> A bit gross, but what can we do? PR 5953, ok tobias (deraadt@)
sed
~ TEST/sed.test
> There is no /usr/old/bin/sed so compare test results against /usr/bin/sed
> (millert@)
~ sed.1 ~ process.c
> The 'l' command should not encode newlines. Also document that
> backspace is escaped as \b. OK phessler@ and jmc@ (millert@)
~ compile.c ~ extern.h
~ main.c ~ misc.c
> Allow sed to handle arbitrarily long lines. Also plug a memory
> leak noticed in the process. Closes PR 5303. OK otto@ deraadt@ (millert@)
~ main.c
> Don't strip newline when reading lines; matches the original behavior.
> From otto@ (millert@)
~ compile.c
> In compile_subst(), adjust for the fact that the initial buffer
> that is passed in may now be larger than _POSIX2_LINE_MAX.
> Thanks to pedro@ for the test case. OK pedro@ (millert@)
~ compile.c
> Now that len is set later in compile_subst() we need to move the
> xmalloc call that goes with it. (millert@)
sendbug
~ sendbug.c
> parse all arguments before going and doing the job requested by -P
> (slight change made to check argc == 0)
> from Daniel Dickman (deraadt@)
ssh
~ ssh.1 ~ ssh.c
> Add -y option to force logging via syslog rather than stderr.
> Useful for daemonised ssh connection (ssh -f). Patch originally from
> and ok'd by markus@ (djm@)
~ servconf.c ~ sshd_config.5
> support setting PermitEmptyPasswords in a Match block
> requested in PR3891; ok dtucker@ (djm@)
~ ssh.c
> add -y to usage(); (jmc@)
~ scp.c
> spelling in comment; ok djm@ (stevesk@)
~ key.c
> typo in error message; ok djm@ (stevesk@)
~ ssh_config.5
> use 'Privileged ports can be forwarded only when logging in as root on
> the remote machine.' for RemoteForward just like ssh.1 -R.
> ok djm@ jmc@ (stevesk@)
stat
~ stat.1
> give working examples (s/timespec/time/ for variables exported)
> please commit from jmc@ (todd@)
systat
~ pftop.c
> Fix the order of traversal when printing the queues. Properly places
> child queues under their respective interface root queues. Reported
> by jared r r spiegel via PR 5861. (canacar@)
vmstat
~ vmstat.c
> use (unsigned) in rate calculation for > 2 billion interrupts. This will
> do until when we tweak struct uvmexp to have unsigned 64 bit counters,
> and when we do, it will not break.
> from PR 4046, ok millert (deraadt@)
== usr.sbin ========================================================== 11/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
authpf
~ authpf.c
> grammar; PR 5394 (deraadt@)
~ authpf.c
> protect better against races from incoming signals; slightly changed
> from 5394 by tracking the fd instead of the fp.
> ok beck (deraadt@)
httpd
~ src/support/rotatelogs.c
> Don't exit if the logfile specified on the command line can not be
> created due to a path component not existing, but log the error and
> retry. This will repeatetly log an error, like it does without the patch,
> but it will not slowly eat up memory. This fixes pr 4228. (mbalmer@)
mtree
~ mtree.h ~ verify.c
> Correct exit codes for extra files. PR 5072.
> Reminded by millert. ok millert, deraaadt (kili@)
ntpd
~ ntpd.conf.5
> - use nmea(4) for general sensor example rather than udcf(4)
> which only works in w.europe
> - use a more realistic value for udcf correction example
> - use a GPS-only receiver for "refid GPS" example
> suggested/ok mbalmer (sthen@)
pcidump
~ pcidump.8 ~ pcidump.c
> dump pci devices from all the domains available in the system. this gives
> a proper view of what is plugged into the box, previously only the first
> domain was displayed.
> from Paul de Weerd <weerd at weirdnet.nl> with help and feedback from
> kettenis at . thanks paul :) (dlg@)
pkg_add
~ pkg_create ~ pkg_merge
~ OpenBSD/Add.pm ~ OpenBSD/CollisionReport.pm
~ OpenBSD/Delete.pm ~ OpenBSD/PackingElement.pm
~ OpenBSD/PackingList.pm ~ OpenBSD/md5.pm
> turn the checksums stored in packing elements into real objects, so that
> pkg_add becomes mostly independent of the type of checksum stored.
> separate md5 into an abstract OpenBSD::digest class and an md5 subclass
> with specific methods to compute checksums and serialize results, and
> create
> an sha class that does the same thing with sha256 (with a base64
> serializer) (espie@)
~ OpenBSD/Replace.pm
> oops, missed one compute_md5
> noticed by djm@ (espie@)
~ pkg_merge ~ pkg_mklocatedb
> add rcs tags
> ok espie (steven@)
pppd
~ sys-bsd.c
> SIOCAIFADDR can fail with EEXIST either if the local or the destination
> address already exists so mention both addresses in the error message.
> From PR4584 (alexander at beard dot se) I just tuned the error message a
> bit. (claudio@)
pstat
~ pstat.8 ~ pstat.c
> pstat can hide information too; also based on PR 5113 (deraadt@)
relayd
~ relayd.conf.5
> tweak previous; (jmc@)
snmpd
~ trap.c
> plug memleak by calling ber_free() after sending a trap to cleanup the
> BER write buffer. (reyk@)
tcpdump
~ print-ppp.c
> Fix -X output of ppp packets. Inspired by PR4624 and especially the patch
> provided by canacar at . I just modified it a bit to skip the data link proto
> number as well. OK deraadt@ (claudio@)
user
~ user.c
> When adding a new user, do not move a potential yp line (+:*:::0:0::::) to
> the
> end of the file, for this would make logins coming after the yp line (such
> as nomadic or fallback accounts) to come back before the yp line and take
> precedence. Found the hard way installing packages needing a user to be
> created.
> ok deraadt@ (miod@)
ypldap
~ aldap.c
> use strtonum() instead of atoi() and remove an unnecessary bzero()
> ok pyr@ (aschrijver@)
~ ldapclient.c ~ aldap.h
> Oops, also commit the header file. Remove silly debug message.
> (aschrijver@)
===============================================================================
More information about the owc
mailing list