[owc] Weekly src changes ending 2007-09-30
OWC
auto at squish.net
Mon Oct 1 07:00:01 BST 2007
OpenBSD src changes summary for 2007-09-23 to 2007-09-30 inclusive
==================================================================
bin/ed lib/libc
lib/libossaudio lib/libssl
libexec/fingerd libexec/identd
regress/sbin regress/usr.bin
sbin/mount_vnd sbin/newfs_msdos
sbin/pfctl sbin/restore
sbin/route share/man
sys/arch/amd64/conf sys/arch/i386/conf
sys/arch/i386/i386 sys/arch/i386/include
sys/arch/macppc/conf sys/arch/sparc64/include
sys/arch/sparc64/sparc64 sys/dev
sys/dev/cardbus sys/dev/ic
sys/dev/microcode sys/dev/pci
sys/dev/usb sys/kern
sys/net sys/netinet
sys/nfs sys/sys
sys/ufs/ext2fs sys/ufs/ffs
sys/ufs/ufs usr.bin/audioctl
usr.bin/cvs usr.bin/make
usr.bin/mixerctl usr.bin/oldrdist
usr.bin/pcc usr.bin/radioctl
usr.bin/split usr.bin/ssh
usr.bin/tail usr.bin/tn3270
usr.bin/top usr.bin/xlint
usr.sbin/authpf usr.sbin/hoststatectl
usr.sbin/hoststated usr.sbin/ospfd
usr.sbin/pkg_add usr.sbin/ppp
== bin =============================================================== 01/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/bin
ed
~ POSIX
> typos; from Pierre Riteau (jmc@)
== lib =============================================================== 02/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
libc
~ gen/sysctl.3
> updatecomment and manpage toemtion optional seventh level (rtable id)
> From: Pierre Riteau <pierre.riteau at free.fr> (henning@)
~ gen/sysctl.3
> tweak previous; ok claudio henning (jmc@)
~ gen/fstab.c
> strtol -> strtonum
> assign values to struct members directly
> remove now-unused variable
> ok otto@
> ok & input millert@ (blambert@)
libossaudio
~ ossaudio.c
> implement SNDCTL_DSP_[G|S]ETTRIGGER
> SNDCTL_DSP_SETTRIGGER is used to pause and unpause recording and
> playback data transfers. SNDCTL_DSP_GETTRIGGER returns the paused
> states.
> this implementation of SNDCTL_DSP_SETTRIGGER also sets the audio(4)
> mode (AUMODE_[PLAY|RECORD]), depending on which direction (play, record
> or both) is to be triggered.
> this can be used allow OSS audio programs to work in full-duplex
> mode without changing the audio(4) layer itself. according to
> OSS docs, SNDCTL_DSP_SETTRIGGER is a fairly important ioctl, and
> should be used "when even the first read() cannot block."
> tested to not break any ports currently using SNDCTL_DSP_SETTRIGGER
> which has been a no-op for quite some time.
> ok ratchov (jakemsr@)
libssl
~ src/ssl/ssl_lib.c
> Fix off-by-one buffer overflow in SSL_get_shared_ciphers().
> From OpenSSL_0_9_8-stable branch.
> ok djm@ (moritz@)
== libexec =========================================================== 03/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/libexec
fingerd
~ fingerd.8
> better wording; from Tamas TEVESZ (jmc@)
identd
~ identd.8
> more accurate description of -U; from Tamas TEVESZ
> ok millert (jmc@)
~ identd.8 ~ identd.c
> - sort options and sync usage()
> - kill some .Tn whilst here (jmc@)
~ identd.c
> From Tamas TEVESZ:
> - atoi -> strtonum with proper bounds check
> - Use vfprintf and vsyslog in the appropriate places.
> OK millert. (ray@)
~ parse.c
> From Tamas TEVESZ:
> - Use strcspn to strip \r and \n.
> OK chl. (ray@)
== regress =========================================================== 04/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/regress
sbin
~ pfctl/Makefile + pfctl/pf95.in
+ pfctl/pf95.ok
> Allow numbers to be used as unquoted strings again.
> While there, also restrict the use of concatenated, unquoted
> strings for variable assignments only.
> Eyeballed by markus@, OK henning@ (mpf@)
usr.bin
+ pcc/cc/ccom/init002.c
> newly discovered init problem (otto@)
+ pcc/cc/ccom/init003.c
> another init problem (otto@)
+ pcc/cc/ccom/darray001.c + pcc/cc/ccom/init004.c
> two more test cases (otto@)
~ pcc/cc/ccom/init004.c
> more minimal example (otto@)
+ pcc/cc/ccom/enum001.c
> test case involving superfluous type warnings involving enums (otto@)
+ pcc/cc/ccom/const001.c
> compare of constants outside type range (otto@)
== sbin ============================================================== 05/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
mount_vnd
~ mount_vnd.8 ~ mount_vnd.c
> default to svnds now to accomodate change in the vnd driver.
> from Piotr Durlej with man-page nits from me
> tested henning@, krw@, thib@ and myself.
> ok thib@, grunk@, krw@ (fkr@)
newfs_msdos
~ newfs_msdos.c
> some cleanup. remove #ifdef __FreeBSD__ code, last sync to FreeBSD was
> 7 years ago. this improves readability.
> provide a standard usage message while there (synced with manpage).
> ok millert@, ``looks fine'' krw at . (pyr@)
~ newfs_msdos.c ~ newfs_msdos.8
> sort options; (jmc@)
pfctl
~ parse.y
> Allow numbers to be used as unquoted strings again.
> While there, also restrict the use of concatenated, unquoted
> strings for variable assignments only.
> Eyeballed by markus@, OK henning@ (mpf@)
~ parse.y
> Add loginterface support for groups.
> Using a group sums up the statistics of all members.
> Modify pfctl(1) slightly to allow a groupname "all",
> which gives us an overall pf(4) statistic.
> OK henning@, markus@ (mpf@)
restore
~ interactive.c
> check fgets return value
> use sizeof input intead of BUFSIZ
> remove unneeded feof
> handle empty strings returned by fgets
> properly remove newline and white space at the end of buffer
> with help and ok ray@ (chl@)
~ tape.c
> remove unneeded feof
> with help and ok ray@ (chl@)
route
~ route.c
> make route flush obey the rtable ID
> From: Pierre Riteau <pierre.riteau at free.fr> (henning@)
== share ============================================================= 06/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man4/bwi.4
> Add cards I've tested. (jsg@)
~ man4/bwi.4
> Add my working Compaq R4035 onboard BCM4306. ok mglocker@ (ian@)
~ man4/bwi.4
> Add new device reported by brad at . Remove dummy device. (mglocker@)
~ man4/bwi.4
> keep the hardware list sorted; (jmc@)
~ man4/bwi.4 ~ man4/cardbus.4
~ man4/pci.4
> Broadcom -> Broadcom AirForce. Add new device to HW list.
> From brad@ (mglocker@)
~ man4/bwi.4
> Add monitor mediaopt. Diff from brad@
> Disable HostAP and IBSS mediaopt while there. (mglocker@)
~ man4/bwi.4
> Zap HostAP examples for now.
> Pointed out by jmc@ (mglocker@)
~ man5/pf.conf.5
> Mention "set loginterface <ifgroup>" (mpf@)
~ man9/style.9
> make clear that the usage and synopsis take the same form
> problem noticed by jakemsr@; rewording suggested by jmc@
> ok jakemsr@, jmc@ (sobrado@)
~ man4/audio.4
> be explicit about configuring audio devices for full-duplex mode.
> OK jmc@ (a while back) (jakemsr@)
~ man5/pf.conf.5
> while rdr'ing port spamd to portsmtpis perfectly valid, it is at least
> a bit confuzzling, so swap.
> From: Olli Hauer <ohauer at gmx.de> (henning@)
== sys =============================================================== 07/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/amd64/conf
~ GENERIC
> Add bwi PCI and Cardbus GENERIC entries for i386 and amd64, but still
> disabled for now.
> From brad@ (mglocker@)
~ GENERIC
> Enable Broadcom wireless, bwi(4).
> If people testing could send mail to mglocker and myself we'd appreciate
> it.
> ok mglocker@ (jsg@)
arch/i386/conf
~ GENERIC
> Add bwi PCI and Cardbus GENERIC entries for i386 and amd64, but still
> disabled for now.
> From brad@ (mglocker@)
~ GENERIC
> Ooops, i386 entry was enabled by mistake. Disable it. Sorry. (mglocker@)
~ GENERIC
> Enable Broadcom wireless, bwi(4).
> If people testing could send mail to mglocker and myself we'd appreciate
> it.
> ok mglocker@ (jsg@)
arch/i386/i386
~ machdep.c
> On i386, only call amd64_errata() on amd64 processors (family 0xf),
> tested by krw at . This diff by itself is insufficient to ensure that
> we are only running on an amd64: todd@'s machine is a mobile Athlon XP
> that has a cpuid family of 0xf but is not an amd64 processor so more
> must be done but this should prevent the vast majority of cases.
> ok tom@ (gwk@)
arch/i386/include
~ stdarg.h
> At the requst of deraadt@, backpout the temporary hack for pcc.
> Paraphrasing, "make it clear this should be fixed in pcc asap". (otto@)
arch/macppc/conf
~ GENERIC
> Enable Broadcom wireless, bwi(4).
> If people testing could send mail to mglocker and myself we'd appreciate
> it.
> ok mglocker@ (jsg@)
arch/sparc64/include
~ cpu.h
> Move intrpending array into 'struct cpu_info'. (kettenis@)
arch/sparc64/sparc64
~ genassym.cf ~ locore.s
> Move intrpending array into 'struct cpu_info'. (kettenis@)
dev
~ audio.c
> fix audioctl(1) returning bugus values of 'output_muted' because of
> uninitialized variable.
> found by deanna@, ok jackmsr@ (ratchov@)
~ audio.c
> accept values > 1 for "output_muted". That's what SunOS do. Suggested
> by deanna@
> ok jakemsr@ (ratchov@)
~ vnd.c
> make safe vnds work on block devices.
> from Piotr Durlej, tested by henning@, krw@, thib@ and myself
> ok thib@, grunk@, krw@ (fkr@)
dev/cardbus
~ if_bwi_cardbus.c
> Remove BCM4311 and BCM4312 which are PCIe chipsets and therefore will not
> come in Cardbus devices.
> From brad@ (mglocker@)
~ if_bwi_cardbus.c
> Add automatic rate control (AMRR). (mglocker@)
dev/ic
~ bwi.c
> Make sure to set physical address in bwi_newbuf() (jsg@)
~ bwi.c
> Properly handle the physical address in bwi_encap() as well.
> With this change I can now scan, associate and ssh via bwi(4) (jsg@)
~ bwi.c
> Correctly clear PCI_STATUS_TARGET_TARGET_ABORT bit. (mglocker@)
~ bwi.c
> More debug ouput cleanup. (mglocker@)
~ bwi.c
> Add missing device name printf arguments for firmware load error output.
> (mglocker@)
~ bwi.c
> Fix page fault kernel crash when changing media and device is up.
> (mglocker@)
~ bwi.c ~ bwivar.h
> Add radio tap support. We still need to calculate correct RX rate and
> RX signal strength. (mglocker@)
~ elink3.c
> ep(4) is totally weird. due to hardware design limitations it plays
> wierd games with caching mbufs. as part of that it used regular mbufs
> and makes them M_PKTHDR mbufs on the fly, instead of calling m_gethdr.
> of course it did it hackish and wrong and did not properly initialize
> the pkthdr fields, which stated to cause problems as soon as we started
> to stuff more into the packet header.
> so use the freshly introduced m_inithdr() to make the mbuf a pktdr mbuf.
> problem found + analyzed by janjaap at stack.nl PR5563, input theo, ok claudio
> (henning@)
~ bwi.c ~ bwireg.h
~ bwivar.h
> Kill __BIT* macros.
> Help and OK jsg@ (mglocker@)
~ ami.c ~ mfi.c
> M_ZERO changes
> ok dlg@ (chl@)
~ bwi.c ~ bwireg.h
~ bwivar.h
> Spacing. (mglocker@)
~ bwi.c ~ bwivar.h
> Add automatic rate control (AMRR). (mglocker@)
~ bwi.c ~ bwivar.h
> Spacing. (mglocker@)
~ bwi.c
> Fix two panic's on macppc. Allows me first network communication over
> my PowerBook G4 internal BCM4306 device. (mglocker@)
~ an.c ~ atw.c
~ gem.c ~ rtw.c
> s/NPBFILTER/NBPFILTER/ in #endif comment. No functional change. (kettenis@)
~ bwi.c
> BWI_BBP_ATTEN is a 16-bit register, so use CSR_WRITE_2 to access it.
> Fixes unaligned access on strict-alignment architectures.
> ok mglocker@ (kettenis@)
~ bwi.c
> Don't count interrupts with state "not of interest".
> From Oleg Safiullin. OK jsg@ (mglocker@)
~ bwireg.h
> Shows the real size of a register in the hex define.
> OK jsg@ (mglocker@)
~ bwi.c
> Spacing. (mglocker@)
dev/microcode
~ tht/microcode.h
> Upgrade the firmware to version SNIC 0xf.
> Here's the issue this fixes:
> Problem description: Tehuti NIC generated RxD descriptor with UDP
> checksum error bit raised for incoming UDP packet with UDP checksum
> field value zero. This behavior is not in accordance with UDP RFC
> which defines that packets with zero UDP checksum should not be
> verified by receiver.
> Solution description: Never raise RxD UDP checksum error bit raised
> for incoming UDP packet with UDP checksum field value zero.
> Many thanks to Alexander Indenbaum and the rest of Tehuti generally
> for notifying us and providing the new firmware under the same
> license as the previous revision. (dlg@)
dev/pci
~ azalia_codec.c
> Clean up STAC7661 codec: sort mixer items, add prev and next pointers
> for master, remove useless pin direction calls, add mixer item for mic
> boost, remove useless item for the volume knob. (deanna@)
~ pcidevs
> Add Sigmatel to vendors.
> ok brad, dlg. (deanna@)
~ pcidevs.h ~ pcidevs_data.h
> regen (deanna@)
~ if_bwi_pci.c
> Add automatic rate control (AMRR). (mglocker@)
~ qli_pci.c ~ qlireg.h
> Add first part of queue magic to attach. (marco@)
~ mbg.c
> Add support for the high-resulution timer found in the GPS170PCI card.
> Thanks for otto@ for looking at the time calculations (mbalmer@)
dev/usb
~ uaudio.c
> fix emulated encodings
> ok jakemsr@ (ratchov@)
~ uaudio.c
> no longer used macros
> spotted by brad
> ok ratchov (jakemsr@)
kern
~ uipc_mbuf.c
> provide m_inithdr(), which takes an mbuf and gives an initialized M_PKTHDR
> mbuf back.
> for fixing PR5563 in a few, tested janjaap at stack.nl, ok claudio (henning@)
net
~ if_spppsubr.c
> Handle incoming protocol reject packets now that we have
> more than one protocol. From FreeBSD, reported and tested by jmc@
> (canacar@)
~ pf_if.c ~ pf_ioctl.c
~ pfvar.h
> Add loginterface support for groups.
> Using a group sums up the statistics of all members.
> Modify pfctl(1) slightly to allow a groupname "all",
> which gives us an overall pf(4) statistic.
> OK henning@, markus@ (mpf@)
netinet
~ ip_carp.c
> Fix carp(4) after the switch to dynamic imo_membership allocation.
> The multicast code is still a mess but will no longer crash the box.
> With input from pascoe@ who hit the same bug.
> OK markus@, henning@, mpf@ (on a previous version) (claudio@)
nfs
~ nfsm_subs.h ~ nfs_vnops.c
> nfsm_uiotom macro does nothing but hide error-handling code; replace
> inline in the one place that it's called
> ok thib@, weingart@
> "I don't see why not" fgsch@ (blambert@)
sys
~ socket.h
> updatecomment and manpage toemtion optional seventh level (rtable id)
> From: Pierre Riteau <pierre.riteau at free.fr> (henning@)
~ mbuf.h
> provide m_inithdr(), which takes an mbuf and gives an initialized M_PKTHDR
> mbuf back.
> for fixing PR5563 in a few, tested janjaap at stack.nl, ok claudio (henning@)
ufs/ext2fs
~ ext2fs_vnops.c
> Move notatime check from ffs_update() to ffs_read() and ufs_setattr().
> Fixes a couple cases where atime was being updated even when the
> filesystem was mounted with the noatime option.
> Adapted from similar placement in FreeBSD and NetBSD.
> Tested by myself and naddy@ (millert@)
ufs/ffs
~ ffs_inode.c ~ ffs_vnops.c
> Move notatime check from ffs_update() to ffs_read() and ufs_setattr().
> Fixes a couple cases where atime was being updated even when the
> filesystem was mounted with the noatime option.
> Adapted from similar placement in FreeBSD and NetBSD.
> Tested by myself and naddy@ (millert@)
ufs/ufs
~ ufs_vnops.c
> Move notatime check from ffs_update() to ffs_read() and ufs_setattr().
> Fixes a couple cases where atime was being updated even when the
> filesystem was mounted with the noatime option.
> Adapted from similar placement in FreeBSD and NetBSD.
> Tested by myself and naddy@ (millert@)
== usr.bin =========================================================== 08/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
audioctl
~ audioctl.1 ~ audioctl.c
> style(9) fixes, optional flags without arguments come first;
> while here, some spacing fixes
> ok deanna@, jakemsr@ (sobrado@)
cvs
~ file.c ~ file.h
~ update.c
> pass FILE_HAS_FLAG towards the update code so it can identify
> if a given tag is present in the RCSfile or not. (joris@)
~ add.c ~ checkout.c
~ commit.c ~ cvs.h
~ file.c ~ update.c
> make sure when running update -r on existing files to overwrite
> the sticky directory tag with it. since -r gets priority. (joris@)
~ rcs.c
> zap unused stuff, cleans it up a bit. (joris@)
~ commit.c ~ file.c
~ rcs.c ~ rcs.h
~ rcsnum.c
> add support to commit modified files to branches. (joris@)
~ commit.c
> unused vars (joris@)
~ edit.c ~ getlog.c
> oops i was suppose to pass cvs_directory_tag to cvs_file_classify().
> (joris@)
~ buf.c
> missing header for mkstemp
> ok joris@ (chl@)
~ logmsg.c
> missing header for mkstemp
> check mkstemp return value against -1 instead of NULL
> ok joris@ (chl@)
~ cmd.c ~ release.c
> remove unused extern variable
> ok joris@ (chl@)
~ entries.c
> missing header for strptime
> ok joris@ (chl@)
make
~ job.c
> zap sunos bug work-around (espie@)
~ job.c
> more specific error messages if dup2 fails (espie@)
~ job.c ~ job.h
~ main.c
> kill local/jobs distinction. Correctly this time... (espie@)
~ job.c
> replace if (DEBUG(job)) { fprintf(stdout,...); fflush(stdout); }
> with a debug_printf() function.
> Maybe slightly slower, but it's ways easier to read. (espie@)
~ gnode.h ~ parse.c
> put parsing of operator (:, ::, !) into its own function (espie@)
~ PSD.doc/tutorial.ms
> zap keywords we don't have. remove exercise that says we don't have $$V,
> since we do ;) (espie@)
~ job.c
> compatMake is never true in there, so kill dead code. (espie@)
~ job.c
> last parameter to JobStart is always NULL, so nuke it. (espie@)
~ job.c
> now, JOB_FIRST is always set, so kill more conditional code.
> (taken from NetBSD) (espie@)
~ job.c
> don't even attempt to reuse the same temp file name.
> Just create a new one each time we need it, and kill the file right away.
> (espie@)
~ job.c
> remove comment that no longer apply.
> Replace a stupid comment with something that makes sense.
> (this function needs a rewrite, it's incredibly stupid anyways) (espie@)
~ job.c
> use constant FD_CLOEXEC, more readable than 1.
> zap comment that should have gone with setting use_pipes. (espie@)
~ job.c
> simple refactoring: replace convoluted code with simple equivalent test
> structure. (espie@)
mixerctl
~ mixerctl.1 ~ mixerctl.c
> style(9) fixes, optional flags without arguments come first;
> while here, some spacing fixes
> ok deanna@, jakemsr@ (sobrado@)
oldrdist
~ docmd.c ~ main.c
> "mailling" -> "mailing", from Jung's list; while here, some spacing fixes
> ok jmc@ (sobrado@)
pcc
~ cc/ccom/init.c
> Pull from ragge's repo:
> Fix for a bug found by Otto Muelbeek.
> I think ragge found at least 4 different ways to misspell my name ;-)
> (otto@)
~ cc/cc/cc.c
> Pull form ragge's repo:
> Patch to understand .i files (no preprocessing), from Arthur Loiret.
> (otto@)
~ cc/ccom/pftn.c
> Avoid potential segfault when debugging (otto@)
~ cc/ccom/init.c
> from ragge's repo:
> Fix init error to give an error message instead of segfault, as
> reported by Otto Muerbeek. (otto@)
~ cc/cc/cc.c ~ cc/ccom/main.c
~ cc/cpp/cpp.c ~ mip/mkext.c
> pull from ragge's repo: -v flag and reg class check (otto@)
~ arch/x86/local.c ~ mip/reader.c
> Pull from ragge's repo:
> Add support for signed bitfields, this has been missing ~forever.
> Bug reported by TAKAHASHI Tamotsu.
> Fix bitfield overflow bug, reported by TAKAHASHI Tamotsu. (otto@)
~ cc/ccom/init.c
> pull from ragge's repo:
> Give a better error message for too many {, "from otto". (otto@)
~ cc/ccom/main.c
> pull from ragge's repo:
> Do not try to print out common symbols if errors, from Stefan Kempf.
> (otto@)
~ cc/cpp/cpp.c ~ cc/cpp/scanner.l
> pull from ragge's repo:
> Handle #pragma and _Pragma() correct. (otto@)
~ cc/cc/cc.1 ~ cc/ccom/ccom.1
~ cc/cpp/cpp.1
> pullup of manual pages from ragge's repo (otto@)
~ cc/cc/cc.c
> flag usage checks and -E with outfile. from Jeremy C. Reed via ragge's
> repo. (otto@)
~ cc/cc/cc.c
> Wrap MAXFIL to 10000, so make depend works.
> This should be dynamic but, ... not now. (ragge@)
- cc/cpp/token.c
> unused file (otto@)
~ cc/cpp/scanner.l
> back to start state when pragma is done; ok ragge@ (otto@)
~ cc/ccom/pftn.c ~ cc/ccom/init.c
> Avoid a member array overflow in the init code; ok ragge@ (otto@)
~ cc/ccom/init.c
> check if initializer element is a constant in MI code. ok ragge@ (otto@)
radioctl
~ radioctl.1 ~ radioctl.c
> style(9) fixes, optional flags without arguments come first;
> while here, some spacing fixes
> ok deanna@, jakemsr@ (sobrado@)
split
~ split.c
> handle empty strings returned by fgets
> ok ray@ (chl@)
ssh
~ ssh-agent.c
> When adding a key that already exists, update the properties
> (time, confirm, comment) instead of discarding them. ok djm@ markus@
> (canacar@)
~ dh.c
> Don't return -1 on error in dh_pub_is_valid(), since it evaluates
> to true.
> Also fix a typo.
> Initial diff from Matthew Dempsky, input from djm.
> OK djm, markus. (ray@)
~ auth2.c
> Remove unused prototype. ok djm@ (dtucker@)
tail
~ forward.c ~ misc.c
~ read.c ~ reverse.c
> unneeded includes and whitespace police; from Igor Zinovik (otto@)
tn3270
~ tools/mkhits/dohits.c
> handle empty strings returned by fgets
> use strcspn to properly overwrite '\n' in fgets returned buffer
> ok ray@ (chl@)
top
- patchlevel.h - version.c
~ Makefile ~ top.h
> remove unneeded files, from Mark Lumsden (otto@)
~ top.c
> reuse available var, from Mark Lumsden (otto@)
~ top.c
> No need to use two buffers for reading data from the user. From Mark
> Lumsden; ok ray@ (otto@)
~ display.c
> unused static var; from Mark Lumsden (otto@)
xlint
~ README
> typos; from Pierre Riteau (jmc@)
== usr.sbin ========================================================== 09/09 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
authpf
~ authpf.c
> handle empty strings returned by fgets
> ok ray@ (chl@)
hoststatectl
~ hoststatectl.8 ~ hoststatectl.c
~ parser.c ~ parser.h
> Correct my mail address. (pyr@)
~ hoststatectl.c
> KNF. (pyr@)
hoststated
~ Makefile ~ hoststated.h
~ relay.c + ssl_privsep.c
> Introduce two new functions to be able to load certificates while
> already chrooted and with privileges dropped.
> This is the very first step in being able to reload a
> layer 7 configuration.
> not ok reyk who's away but should be glad to see this in. (pyr@)
~ pfe.c
> Do not clear the changed flag to early, which prevented a table
> from being used by several services.
> ``looks fine'' reyk@ (pyr@)
~ hoststated.h ~ ssl_privsep.c
~ relay.c
> Simplify ssl_privsep.c, since it won't need to remain synced with the
> equivalent openssl functions. (pyr@)
~ relay.c
> Move SSL context creation after privileges are dropped.
> This puts the ssl_privsep code to use.
> One more step towards graceful L7 reload. (pyr@)
~ hoststated.conf.5
> Add missing "s" to https check description.
> ok pyr@ (pascoe@)
~ hoststated.conf.5
> "require to +inf." is not a good verb pattern, so reword; (jmc@)
~ hoststated.h ~ relay.c
~ ssl_privsep.c
> Change the ssl_privsep code to work on char buffers.
> The fd based code introduced weirdness since all children were accessing
> the same fd at once. This will also greatly facilitate reloading, no
> fd-passing will be involved between the parent and relay children.
> While there, cleanup the code diverting from the original ssl_rsa.c code
> a bit more.
> Weird behavior discovery by pascoe at . (pyr@)
~ check_icmp.c ~ check_tcp.c
~ hce.c ~ hoststated.c
~ hoststated.conf.5 ~ hoststated.h
~ parse.y ~ pfe.c
~ pfe_filter.c ~ ssl.c
> Correct my mail address. (pyr@)
~ hoststated.8
> Better description of the daemon.
> Initially prompted by deraadt@, with much much help from jmc@ (as always).
> (pyr@)
~ hoststated.c
> KNF (pyr@)
ospfd
~ kroute.c ~ ospfd.c
~ ospfd.h ~ rde.c
> Last missing piece in the equal cost multipath support for ospfd.
> Send all possible nexthops to the parent process and correctly sync
> the RIB, FIB and kernel routing table. Based on initial work by pyr at .
> OK pyr@ norby@
> PS: don't forget that you need to enable multipath support via a sysctl
> (claudio@)
~ area.c
> Cleanup some #includes plus some minor other cleanup. OK norby@ (claudio@)
~ log.h
> spaces (claudio@)
~ printconf.c
> When printing the config always include the interface IP address.
> Without this multiple networks on one interface could not be distinguished.
> OK norby@ (claudio@)
pkg_add
~ OpenBSD/SharedLibs.pm
> nit
> from qgiovanni at gmail.com (espie@)
ppp
~ ppp/radlib.c
> handle empty strings returned by fgets
> ok ray@ (chl@)
===============================================================================
More information about the owc
mailing list