[owc] Weekly src changes ending 2006-08-13
OWC
auto at squish.net
Mon Aug 14 08:06:22 BST 2006
OpenBSD src changes summary for 2006-08-06 to 2006-08-13 inclusive
==================================================================
bin/mv distrib/sets
gnu/usr.sbin/sendmail lib/libc
libexec/ld.so regress/bin
regress/usr.bin sbin/mount_ffs
sbin/pfctl sbin/ping6
share/man sys/arch/alpha/alpha
sys/arch/alpha/conf sys/arch/alpha/include
sys/arch/amd64/amd64 sys/arch/amd64/conf
sys/arch/arm/arm sys/arch/armish/conf
sys/arch/armish/stand/boot sys/arch/aviion/aviion
sys/arch/hp300/dev sys/arch/hp300/stand/common
sys/arch/hppa/conf sys/arch/hppa/hppa
sys/arch/hppa/include sys/arch/hppa64/hppa64
sys/arch/hppa64/include sys/arch/i386/conf
sys/arch/i386/i386 sys/arch/i386/pci
sys/arch/luna88k/conf sys/arch/luna88k/dev
sys/arch/macppc/conf sys/arch/macppc/macppc
sys/arch/mips64/mips64 sys/arch/mvme88k/stand/tftpboot
sys/arch/mvmeppc/mvmeppc sys/arch/mvmeppc/stand/libsa
sys/arch/sgi/sgi sys/arch/solbourne/include
sys/arch/sparc/dev sys/arch/sparc/include
sys/arch/sparc/sparc sys/arch/sparc64/conf
sys/arch/sparc64/dev sys/arch/sparc64/include
sys/arch/sparc64/sparc64 sys/arch/vax/vax
sys/arch/vax/vsa sys/arch/zaurus/conf
sys/conf sys/dev
sys/dev/ata sys/dev/cardbus
sys/dev/hil sys/dev/ic
sys/dev/isa sys/dev/microcode
sys/dev/mii sys/dev/pci
sys/dev/rcons sys/dev/usb
sys/dev/wscons sys/dev/wsfont
sys/isofs/cd9660 sys/isofs/udf
sys/kern sys/msdosfs
sys/ufs/ffs usr.bin/calendar
usr.bin/newsyslog usr.bin/rcs
usr.bin/split usr.bin/ssh
usr.bin/tsort usr.bin/xargs
usr.sbin/authpf usr.sbin/cron
usr.sbin/dhcpd usr.sbin/fdformat
usr.sbin/ndp usr.sbin/ospfd
usr.sbin/pkg_add usr.sbin/route6d
usr.sbin/traceroute6 usr.sbin/wsfontload
== bin =============================================================== 01/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/bin
mv
~ mv.1
> better explanation of second synopsis form; add a caveat also
> started by freebsd pr #101330, from peter gildea
> diff from otto and myself (jmc@)
~ mv.1
> - use .Ex
> - bump .St
> - standard EXAMPLES
> - macro cleanup (jmc@)
== 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.cats
~ lists/base/md.hp300 ~ lists/base/md.hppa
~ lists/base/md.hppa64 ~ lists/base/md.i386
~ lists/base/md.mac68k ~ lists/base/md.macppc
~ lists/base/md.mvme68k ~ lists/base/md.mvmeppc
~ lists/base/md.sgi ~ lists/base/md.sparc
~ lists/base/md.sparc64 ~ lists/base/md.zaurus
> sets change for libc bump
> "be careful" deraadt@, ok drahn@, miod@ (beck@)
~ lists/base/md.alpha ~ lists/base/md.amd64
~ lists/base/md.armish ~ lists/base/md.aviion
~ lists/base/md.cats ~ lists/base/md.hp300
~ lists/base/md.hppa ~ lists/base/md.hppa64
~ lists/base/md.i386 ~ lists/base/md.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.cats ~ lists/base/md.hp300
~ lists/base/md.hppa ~ lists/base/md.hppa64
~ lists/base/md.i386 ~ 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@)
== gnu =============================================================== 03/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/gnu
usr.sbin/sendmail
~ sendmail/main.c
> Fix use-after-free bug, which was triggered with too
> long header lines. Diff from Claus Assmann.
> ok thib@ millert@ (moritz@)
~ sendmail/main.c
TAGGED OPENBSD_3_9
> MFC:
> Fix by moritz@
> Fix use-after-free bug, which was triggered with too
> long header lines. Diff from Claus Assmann. (brad@)
~ sendmail/main.c
TAGGED OPENBSD_3_8
> MFC:
> Fix by moritz@
> Fix use-after-free bug, which was triggered with too
> long header lines. Diff from Claus Assmann. (brad@)
~ RELEASE_NOTES ~ cf/README
~ cf/cf/submit.mc ~ cf/feature/dnsbl.m4
~ cf/feature/enhdnsbl.m4 ~ cf/m4/proto.m4
~ cf/m4/version.m4 ~ contrib/cidrexpand
~ include/libmilter/mfapi.h ~ libmilter/engine.c
~ libmilter/docs/xxfi_header.html ~ libsm/t-memstat.c
~ libsm/vfprintf.c ~ sendmail/README
~ sendmail/bf.c ~ sendmail/collect.c
~ sendmail/conf.c ~ sendmail/daemon.c
~ sendmail/deliver.c ~ sendmail/domain.c
~ sendmail/envelope.c ~ sendmail/headers.c
~ sendmail/helpfile ~ sendmail/main.c
~ sendmail/map.c ~ sendmail/mci.c
~ sendmail/mime.c ~ sendmail/parseaddr.c
~ sendmail/queue.c ~ sendmail/savemail.c
~ sendmail/sendmail.h ~ sendmail/sfsasl.c
~ sendmail/sfsasl.h ~ sendmail/srvrsmtp.c
~ sendmail/tls.c ~ sendmail/util.c
~ sendmail/version.c
TAGGED OPENBSD_3_8
> Update to sendmail-8.13.8 (millert@)
== lib =============================================================== 04/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/lib
libc
~ sys/msgsnd.2
> - fix a sentence
> - some indent (jmc@)
~ sys/msgctl.2
> correction from vasil dimov;
> tweak whilst in here (jmc@)
~ gen/gethostname.3
> Document error return when namelen is 0. OK jmc@ (millert@)
== libexec =========================================================== 05/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/libexec
ld.so
~ alpha/rtld_machine.c
> Work around an issue found when working with gcc3 on alpha,
> gcc2 nomally calls functions thru the gp (basically a function pointer),
> but
> gcc3 frequently calls directly to the plt. gcc2 does occasionally directly
> call to the plt. Currently how the plt/pltgot is fixed up on lazy
> binding, this will cause the dl_bind code to be called on each plt call
> causing a symbol resolution and plt fixup. This diff works around the
> problem partially by recognizing that a pltgot entry is already fixed
> and skips the symbol resolution. Since the pltgot is what is modified
> do not add PROT_EXEC permission to the mapping.
> While this is not an ideal fix, kettenis and I think it is better than
> before. (drahn@)
~ alpha/rtld_machine.c
> Do not take shortcuts when the PLT size is 0; makes emacs work again on
> alpha.
> ok drahn@ (kettenis@)
~ ldconfig/ldconfig.8
> spelling: elminate (stevesk@)
== regress =========================================================== 06/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/regress
bin
~ systrace/Makefile + systrace/symlink/Makefile
+ systrace/symlink/ln.policy
> failing qt4 test case. (espie@)
usr.bin
~ rcs/Makefile + rcs/merge-eflag.out
> add tests for merge -e; OK ray at .sa@)
== sbin ============================================================== 07/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
mount_ffs
~ mount_ffs.c
> Give a less cryptic error message when trying to mount invalid file
> systems, okay thib@ ian@ millert@ beck@ (pedro@)
pfctl
~ pfctl_parser.c
> properly join host lists in ifa_grouplookup(), closes PR 5195,
> reported by andrew fresh (dhartmei@)
ping6
~ ping6.8
> sort options; (jmc@)
~ ping6.8
> document -m; (jmc@)
~ ping6.8
> oops, put -m in the right place; (jmc@)
~ ping6.c
> sync usage(); (jmc@)
~ ping6.8
> comment out -R:
> ping6: -R is not supported in this configuration (jmc@)
~ ping6.8
> kill a truckload of useless macros; (jmc@)
~ ping6.8
> kill some .Tn; (jmc@)
~ ping6.8
> text improvements; (jmc@)
== share ============================================================= 08/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man4/pciide.4
> add the Intel 6321ESB (brad@)
~ man4/man4.i386/ichpcib.4
> add the Intel 6321ESB (brad@)
~ man4/mfi.4
> add the Intel RAID Controller SRCSAS18E and SRCSAS144E (brad@)
~ man4/rln.4
> Don't claim rln is 802.11 as it isn't, and don't Xr other drivers. (jsg@)
~ man4/cnw.4 ~ man4/pcmcia.4
> Consistently refer to rln and cnw as wireless devices. (jsg@)
~ man4/acx.4
> ACX100 does hardware WEP, acx111 software WEP.
> ok jsg@ (mglocker@)
~ man4/ubsa.4
> add HUAWEI Mobile E618 to the list of supported adapters (fkr@)
~ man4/acx.4
> Pretty printing of supported cards, similar as in ath(4). (mglocker@)
~ man4/acx.4
> Point to new firmware package -> More firmware images included.
> ok jsg@ (mglocker@)
~ man4/acx.4
> Oops. Also list the new firmware images in the man page. (mglocker@)
~ man4/em.4
> add 1000BaseLX (brad@)
~ man4/rum.4
> remove a diagnostic message that does not exist. (damien@)
~ man4/rum.4
> remove the MIMO keyword. RT2571W adapters aren't MIMO.
> + some cleanup while i'm here. (damien@)
~ man4/usb.4
> sync the rum(4) entry; (jmc@)
~ man4/em.4
> mention the 82562V chipset. (brad@)
~ man4/rum.4
> mention that rum(4) should also work with RT2671 (MIMO XR) chipsets.
> add two products to the list of adapters that should work (pointed
> out by jsg@). (damien@)
~ man4/ral.4
> There are also dual-band RT2501 adapters.
> Sync wording w/ rum(4) while I'm here. (damien@)
~ man4/rum.4
> s/Airlink/Airlink101/ (damien@)
~ man4/rum.4
> bad comma; (jmc@)
~ man5/files.conf.5
> Fix a couple of typos: devices are declared with "device", not "define".
> ok miod@ (tom@)
~ man4/uftdi.4
> add Coastal/TNC-X (jason@)
~ man4/xge.4
> add 2 new IBM adapters and a new Neterion adapter. (brad@)
~ man4/scsi.4
> some ioctls were removed earlier, remove them here as well.
> go for it dlg@ (steven@)
~ man4/acx.4
> Uncomment IBSS mode and Host AP chunks, comment monitor mode for now.
> (mglocker@)
== sys =============================================================== 09/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/alpha/alpha
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
~ cpu.c
> Fix EV6 processor minor type: 0 is reserved and thus all entries were
> off by one.
> Also add a bunch of minor types for EV6 and tables for PCA57 and EV67.
> The present tables should now be complete as per Alpha Architecture
> Reference Manual v4.
> ok miod (mk@)
arch/alpha/conf
~ files.alpha
> rcons removal fallout, spotted by kettenis@ (miod@)
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
~ files.alpha
> Make sure we generate sfb.h and cfb.h, even though support is not there
> yet. (miod@)
arch/alpha/include
~ param.h
> Make all DEV_BSIZE definitions consistant by using (1 << DEV_BSHIFT)
> in the six cases using "512".
> As DEV_BSHIFT is always 9, this should be a no-op.
> "no objections" miod@ "I can't see any problem doing this." pedro@ (krw@)
arch/amd64/amd64
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/amd64/conf
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
arch/arm/arm
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/armish/conf
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
arch/armish/stand/boot
~ wd.c
> Setting d_secsize to DEV_BSIZE (or 1 << DEV_BSHIFT, or 512) and then
> setting RAW_PART's p_size to d_secperunit * (d_secsize / DEV_BSIZE) is
> a waste of a few ops. And p_size should be in sectors anyway.
> Just set RAW_PART's p_size to d_secperunit to make usage consistant
> across the tree.
> Should be a no-op. (krw@)
arch/aviion/aviion
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/hp300/dev
~ diofb.c ~ diofbvar.h
> Use the blitter again for the cursor on non-mono frame buffers. (miod@)
~ dnkbd.c
> Try and determine the keyboard layout from its identification string;
> no international keyboards to check I parse the right part of the string,
> at least domestic keyboards are not affected; hinted at by the old X11R5
> server. (miod@)
~ diofb.c ~ diofb_mono.c
~ diofbvar.h ~ dvbox.c
~ gbox.c ~ rbox.c
~ topcat.c
> Let the windowmove() callbacks return whether they could perform the
> operation
> or not, so that we can always feed them the complex operations before
> falling
> back to rasops; and then topcat can perform coloured solid fills in
> erase{cols,rows}. (miod@)
~ hd.c
> Setting d_secsize to DEV_BSIZE (or 1 << DEV_BSHIFT, or 512) and then
> setting RAW_PART's p_size to d_secperunit * (d_secsize / DEV_BSIZE) is
> a waste of a few ops. And p_size should be in sectors anyway.
> Just set RAW_PART's p_size to d_secperunit to make usage consistant
> across the tree.
> Should be a no-op. (krw@)
arch/hp300/stand/common
~ hil.c
> Oops, let HIL keyboards work again in the boot blocks. (miod@)
~ dnkbd.c
> Let the domain keyboard work on models 362 and 382 as well. (miod@)
arch/hppa/conf
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
arch/hppa/hppa
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/hppa/include
~ param.h
> Make all DEV_BSIZE definitions consistant by using (1 << DEV_BSHIFT)
> in the six cases using "512".
> As DEV_BSHIFT is always 9, this should be a no-op.
> "no objections" miod@ "I can't see any problem doing this." pedro@ (krw@)
arch/hppa64/hppa64
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/hppa64/include
~ param.h
> Make all DEV_BSIZE definitions consistant by using (1 << DEV_BSHIFT)
> in the six cases using "512".
> As DEV_BSHIFT is always 9, this should be a no-op.
> "no objections" miod@ "I can't see any problem doing this." pedro@ (krw@)
arch/i386/conf
~ GENERIC
> enable arc so it can get some exposure to testing (dlg@)
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
arch/i386/i386
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/i386/pci
~ pci_intr_fixup.c
> add the Intel 82801DBM LPC. (brad@)
~ pci_intr_fixup.c
> add the Intel 82801GHM LPC and 6321ESB LPC (brad@)
~ ichpcib.c
> add the Intel 6321ESB LPC (brad@)
arch/luna88k/conf
~ files.luna88k
> Switch luna88k from rcons to rasops. Tested by aoyama@ (miod@)
arch/luna88k/dev
~ lunafb.c ~ omrasops.c
> Switch luna88k from rcons to rasops. Tested by aoyama@ (miod@)
~ siotty.c
> Fail TIOCSFLAGS if non-root, like all other serial drivers do. (miod@)
~ lunaws.c
> Missing rcsid (miod@)
arch/macppc/conf
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
arch/macppc/macppc
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/mips64/mips64
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/mvme88k/stand/tftpboot
~ tftpfs.c
> Remove remains of cut'n'paste from ffs code. (miod@)
arch/mvmeppc/mvmeppc
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/mvmeppc/stand/libsa
~ tftpfs.c
> Remove remains of cut'n'paste from ffs code. (miod@)
arch/sgi/sgi
~ machdep.c
> no O2's with a R14000 CPU were made, so remove superfluous code.
> ok pefo (jasper@)
arch/solbourne/include
~ param.h
> Make all DEV_BSIZE definitions consistant by using (1 << DEV_BSHIFT)
> in the six cases using "512".
> As DEV_BSHIFT is always 9, this should be a no-op.
> "no objections" miod@ "I can't see any problem doing this." pedro@ (krw@)
arch/sparc/dev
~ ts102.c
> An old dvorak typo was lurking here. (miod@)
~ presto.c
> Setting d_secsize to DEV_BSIZE (or 1 << DEV_BSHIFT, or 512) and then
> setting RAW_PART's p_size to d_secperunit * (d_secsize / DEV_BSIZE) is
> a waste of a few ops. And p_size should be in sectors anyway.
> Just set RAW_PART's p_size to d_secperunit to make usage consistant
> across the tree.
> Should be a no-op. (krw@)
arch/sparc/include
~ param.h
> Make all DEV_BSIZE definitions consistant by using (1 << DEV_BSHIFT)
> in the six cases using "512".
> As DEV_BSHIFT is always 9, this should be a no-op.
> "no objections" miod@ "I can't see any problem doing this." pedro@ (krw@)
arch/sparc/sparc
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/sparc64/conf
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
arch/sparc64/dev
~ vgafb.c
> For non-console vgafb, force rasops to initialize it as a monochrome
> resource,
> since we can't program the colormap anyway. (miod@)
arch/sparc64/include
~ param.h
> Make all DEV_BSIZE definitions consistant by using (1 << DEV_BSHIFT)
> in the six cases using "512".
> As DEV_BSHIFT is always 9, this should be a no-op.
> "no objections" miod@ "I can't see any problem doing this." pedro@ (krw@)
arch/sparc64/sparc64
~ autoconf.c
> Hookup hw.vendor and hw.product sysctls on sparc64.
> ok dlg@, "looks good" miod@ (gwk@)
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/vax/vax
~ disksubr.c
> Use d_secperunit rather than 0x1fffffff as the default value for
> RAW_PART's p_size. Since d_secperunit is initialized to 0x1fffffff if
> not specified there should be no need to use the magic number again.
> And if d_secperunit was specified then that value should be used
> instead of the magic number.
> This was already being done for hp300, luna88k, mac68k, mvme68k,
> mvme88k.
> Should be a no-op. (krw@)
arch/vax/vsa
~ lcg.c ~ lcgreg.h
> Replace magic numbers with adequate #defines, support vsync disabling in
> the
> burner routine, and do the colormap updates during vertical retrace,
> instead
> of using a fixed delay. (miod@)
arch/zaurus/conf
~ GENERIC
> add rum(4) on arches where ural(4) is enabled.
> rum(4) has been tested on i386 by niallo@ and me and on sparc64 by
> Pablo Mendez Hernandez so I assume it is endian safe and 64bit clean.
> (damien@)
conf
~ files
> Time for rcons code to retire. (miod@)
dev
~ ramdisk.c ~ vnd.c
> Setting d_secsize to DEV_BSIZE (or 1 << DEV_BSHIFT, or 512) and then
> setting RAW_PART's p_size to d_secperunit * (d_secsize / DEV_BSIZE) is
> a waste of a few ops. And p_size should be in sectors anyway.
> Just set RAW_PART's p_size to d_secperunit to make usage consistant
> across the tree.
> Should be a no-op. (krw@)
~ ccd.c
> Rather than using cs->sc_size to set d_secperunit and then RAW_PART's
> p_size, use d_secperunit to set RAW_PART's p_size. This is more
> consistant with other usage and makes the intent clearer.
> This should be a no-op. (krw@)
~ vnd.c
> fix a panic when trying to fdisk vnd devices (not svnd) wich
> are back by weird images (such as the ones qemu creates).
> proplem found the hard way by msf@
> ok mickey@, pedro@ (thib@)
dev/ata
~ wd.c
> Setting d_secsize to DEV_BSIZE (or 1 << DEV_BSHIFT, or 512) and then
> setting RAW_PART's p_size to d_secperunit * (d_secsize / DEV_BSIZE) is
> a waste of a few ops. And p_size should be in sectors anyway.
> Just set RAW_PART's p_size to d_secperunit to make usage consistant
> across the tree.
> Should be a no-op. (krw@)
dev/cardbus
~ if_acx_cardbus.c
> Mention which MAC we are using (ACX100 or ACX111) and fix
> newline in cardbus attachment. (jsg@)
~ if_acx_cardbus.c
> Some KNF. (mglocker@)
~ if_acx_cardbus.c
> On detach the bus space was not unmapped correctly which could lead to
> a panic. Fix it.
> Diff received by Marc Winiger <mw at msys.ch thanks!
> ok claudio@ (mglocker@)
~ if_xl_cardbus.c
> - ANSI functions
> - de-register
> - remove return at the end of void functions
> - some cosmetic tweaking (brad@)
dev/hil
~ hildevs_data.h
> Regen (miod@)
~ files.hil
> No need for needs-flag for hil and hilms. (miod@)
~ devlist2h.awk ~ hildevs
> Embed device list comments in generated hildevs_data.h file; no functional
> change. (miod@)
~ hilkbd.c ~ hilreg.h
> Handle the ``arrow repeat'' special scancode on button boxes; from the old
> X11R5 server, untested. (miod@)
dev/ic
~ re.c
> With the PCIe devices, it looks as if issuing a TX command while
> there is already a transmission in progress has no effect. In other
> words, if you send two packets in rapid succession, the second one may
> end up sitting in the TX DMA ring until another transmit command is
> issued later in the future. Basically, if re_txeof() sees that there
> are still descriptors outstanding, it needs to manually resume the
> TX DMA channel by issuing another TX command to make sure all
> transmissions are flushed out. (The PCI devices seem to keep the
> TX channel moving until all descriptors have been consumed. I'm not
> sure why the PCIe devices behave differently.)
> From wpaul at FreeBSDrad@)
~ re.c
> fix the config space macro used with the TX start command.
> From wpaul at FreeBSDrad@)
~ re.c
> fix swapping of the station address for big-endian archs.
> From yongari at FreeBSDrad@)
~ acx.c ~ acx100.c
~ acx111.c ~ acxvar.h
> Add bpf hooks. Based on slightly modified diff from jsg.
> ok jsg@ (mglocker@)
~ acx.c
> use the same naming conventions than net80211.
> set IEEE80211_FC0_VERSION_0 and IEEE80211_FC1_DIR_NODS instead of
> blindly assuming these constants are zero.
> ok mglocker@ (damien@)
~ acx.c
> fix a comment. (damien@)
~ acx.c
> more s/f/wh/ that I forgot in rev 1.25 (damien@)
~ acx111.c
> Some KNF. (mglocker@)
~ acx.c
> remove a dragonflybsd'ism (damien@)
~ acx.c
> Radio type is in hex not decimal. Therefore decide correctly which
> radio firmware to load.
> Spotted by Marc Winiger <mw at msys.chmglocker@)
~ acx.c ~ acx100.c
~ acxvar.h
> Enable WEP.
> ok jsg@ (mglocker@)
~ acx.c
> Remove some unused/uneeded code. (jsg@)
~ acx.c
> Make radio type 0x11 work. Radio firmware needs longer to initialize
> for those types. Increase tsleep() from 150ms to 300ms.
> Tested by Marc Winiger <mw at msys.chith 0x11 ACX100 CardBus & ACX100 PCI.
> ok claudio@ (mglocker@)
~ acx.c
> Print a string for radio type. (jsg@)
~ acx.c
> Mention which MAC we are using (ACX100 or ACX111) and fix
> newline in cardbus attachment. (jsg@)
~ acx.c
> Don't yell about failed channel enabling in none-debug mode, because we
> use a static number of channels.
> ok jsg@ (mglocker@)
~ acx.c
> A more flexible method for deciding what firmware filenames to load.
> ok mglocker@ (jsg@)
~ rt2661.c ~ rt2661reg.h
> flags for register TXRX_CSR4 were shifted one bit too much on the left.
> this has probably affected short preamble support as well as hardware
> multi-rate retries settings. (damien@)
~ xl.c
> - move the promiscuous mode handling code to xl_setpromisc() and simplify
> the ioctl handler.
> - eliminate re-initialization's when adding IP addresses. (brad@)
~ xl.c
> - ANSI functions
> - de-register
> - remove return at the end of void functions
> - some cosmetic tweaking (brad@)
~ dc.c
> - eliminate re-initialization's when adding IP addresses.
> - simplify ioctl handler. (brad@)
~ acx.c ~ acxvar.h
> Fill ic_max_rssi with values depending on the radio type. Makes signal
> strength in percentage work.
> ok jsg@ (mglocker@)
~ acx.c
> KNF tweak. (mglocker@)
~ acx.c
> Enable hostap mode. Tested with acx100 and acx111. (mglocker@)
dev/isa
~ fd.c ~ fdreg.h
> Define and use FD_BSIZE, similar to FD_BSIZE from sparc and sparc64
> fd. Eliminate unused variable and FDC_BSIZE which was only used to
> set value of the unused variable.
> No change to object file. (krw@)
~ fd.c ~ mcd.c
> RAW_PART p_size on floppies and mcd's should be in sectors.
> Only 'Japanese 1.2MB' floppies have sectors that are not 512 bytes,
> according
> to isa/fd.c's fd_types[].
> SCSI/ATAPI cd's already use sectors, and mcd cd's should be no different.
> So there should be no functional change. (krw@)
dev/microcode
~ bnx/bnxfw.h
> remove typedef's. (brad@)
dev/mii
~ rgephy.c ~ ipgphy.c
> typo, settning -> setting (brad@)
dev/pci
~ arc.c
> io commands are the only ones every posted on the queues, so hardcode the
> type into the ccb rather than use void *s and insane casting to get at bits
> of the command. (dlg@)
~ arc.c
> less magic (dlg@)
~ if_re_pci.c
> fix message printing when resetting the power state. (brad@)
~ pciide.c
> add the Intel 6321ESB IDE controller. (brad@)
~ pcidevs
> add vendor Mitac and Intel 6321ESB SATA, AHCI SATA and SATA RAID PCI ids.
> (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> add another MegaRAID PCI id. (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ mfi_pci.c
> add another MegaRAID PCI id. (brad@)
~ pcidevs
> be more consistent in naming MegaRAID devices. (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ ami_pci.c ~ mfi_pci.c
> sync PCI ids. (brad@)
~ arc.c
> we can support more than one entry in an sgl, so stop lying to bus_dma
> (dlg@)
~ arc.c
> fill out the interrupt handler. io is now no longer polled, and as a bonus
> it's fairly fast :) (dlg@)
~ if_wpi.c
> cosmetic tweaks. (damien@)
~ if_wpi.c
> endianness. (damien@)
~ arc.c
> endian fixes. this works on sparc64 now. (dlg@)
~ arc.c
> from the drivers side of the iop331 there is only one set of interrupt
> registers to deal with, the other set are used on the firmware side of the
> device. shrink a bit by removing the definitions of the firmwares
> registers. (dlg@)
~ arc.c
> whitespace fixes (dlg@)
~ pcidevs
> add Conexant CX2388x Video and Audio PCI decoder ids. (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> add another Intel 5000 PCIe bridge and another Intel 6321ESB USB port.
> (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs.h ~ pcidevs_data.h
> regen (brad@)
~ pcidevs
> add Intel 82571EB PCI Express quad port PCI id.
> From Intel's FreeBSD driver. (brad@)
~ if_em.c ~ if_ixgb.c
> Use the DMA map size from the DMA map instead of the dma_size field with
> bus_dma sync's.
> ok dlg@ marco@ (brad@)
~ if_em.c ~ if_em_hw.c
~ if_em_hw.h
> Sync up to Intel's latest FreeBSD em driver (6.1.4). Adds PCI id for the
> PCIe
> quad port copper adapter, improvements for media support with fiber
> adapters,
> and some fixes for the ICH8 support. (brad@)
~ if_sk.c
> revert rev 1.119 as this seems to cause problems with some PHY. (brad@)
~ if_ixgb.c
> update the version I am tracking. (brad@)
~ arc.c
> define the size of regions in the register space in terms of bytes, not
> dwords (dlg@)
~ if_bnx.c ~ if_bnxreg.h
> Reorder dmamap & dmamem to match man page.
> Redo detection of _LO & _HI macro; help from miod and jordan.
> ok beck brad (marco@)
~ if_bnxreg.h
> Change #if __LP64__ to #ifdef __LP64__ as requested by brad (marco@)
~ if_bnxreg.h
> add sys/timeout.h (brad@)
~ if_em.c
> cosmetic tweaks. (brad@)
~ mpi_pci.c
> fix an obvious typo. found by Quentin Garnier. (dlg@)
~ if_bnxreg.h
> just use BYTE_ORDER directly instead of setting the drivers endian flags.
> (brad@)
~ if_bnx.c
> cosmetic tweaking. (brad@)
~ if_bnx.c ~ if_bnxreg.h
> remove typedef's. (brad@)
~ if_bnx.c ~ if_bnxreg.h
> unmap memory address space in bnx_release_resources(). (brad@)
~ if_wpi.c
> fix a use-after-free (read) of a mbuf in wpi_tx_intr(). (damien@)
~ if_xge.c
> - Use BUS_DMA_NOWAIT flag with bus_dmamap_create() in xge_attach() for the
> creation of transmit and receive DMA maps.
> - Fix typos in xge_alloc_rxmem(), make sure to destroy and unmap the RX
> descriptor map instead of the TX descriptor map when encountering an error.
> From Gireesh Nagabhushana @ Neterion (brad@)
~ if_xge.c
> - xge_mcast_filter() -> xge_setmulti().
> - move the promiscuous mode handling code to xge_setpromisc() and simplify
> the ioctl handler. (brad@)
~ if_sis.c
> move the promiscuous mode handling code to sis_setpromisc() and simplify
> the ioctl handler. (brad@)
~ if_xl_pci.c
> - ANSI functions
> - de-register
> - remove return at the end of void functions
> - some cosmetic tweaking (brad@)
~ if_bnx.c
> fix a typo, BNX_DRBUG -> BNX_DEBUG (brad@)
~ if_bnx.c ~ if_bnxreg.h
> Get rid of _HI & _LO macros altogether since they used a wrong idiom.
> This was pointed out by mickey The driver now uses the same idiom as mpi.
> help from miod, mickey and kettenis
> ok brad (marco@)
dev/rcons
- raster.h - raster_op.c
- raster_text.c
> Time for rcons code to retire. (miod@)
dev/usb
~ ehci.c
> Minor spelling and whitespace fixes - for diff reduction. (pascoe@)
~ usbdevs ~ usbdevs.h
~ usbdevs_data.h
> add HUAWEI technologies along with their Mobile E618 umts adapter
> ok mbalmer@ (fkr@)
~ ubsa.c
> attach the HUAWEI Mobile E618 umts adapter (fkr@)
~ usbdevs.h ~ usbdevs_data.h
> regen again for proper tags
> reminded by martin@ (fkr@)
~ ubsa.c
> small whitespace nit (fkr@)
~ if_ral.c ~ if_ralreg.h
> fix endianness. this should give rate adaptation a better chance to work
> on big endian architectures. (damien@)
~ if_rum.c ~ if_rumvar.h
~ if_rumreg.h
> mega-commit of unrelated changes to bring rum(4) into a working state.
> more to come but works well enough for me to commit this over a rum(4).
> it shares a lot of code with ral(4) rt2661.c.
> thanks to jsg@ for sending me a rum(4). (damien@)
~ if_rum.c
> tell if chipset supports MIMO in dmesg output. (damien@)
~ usbdevs
> yet another uftdi: Coastal ChipWorks TNC-X (jason@)
~ uftdi.c ~ usbdevs.h
~ usbdevs_data.h
> match the Coastal ChipWorks TNC-X (yet another uftdi) (jason@)
~ usbdevs.h ~ usbdevs_data.h
> sync (jason@)
~ usbdevs
> another delorme earthmate-alike (jason@)
~ usb_quirks.c
> the delorme earthmate is NOT a HID (UQ_BAD_HID) (jason@)
dev/wscons
~ wskbd.c
> Do not try and mate wskbd0 and wsdisplay0 if there is now wsdisplay in the
> kernel; spotted by kettenis. (miod@)
- wscons_font8x16.c - wscons_gallant19.c
- wscons_rinit.c - wscons_rops.c
~ files.wscons
> Time for rcons code to retire. (miod@)
~ wsconsio.h
> Convert the Sony wsfonts to ISO-Latin1 encoding, and drop the Sony encoding
> conversion tables; from NetBSD. (miod@)
dev/wsfont
~ wsfont.c
> Switch luna88k from rcons to rasops. Tested by aoyama@ (miod@)
~ sony12x24.h ~ sony8x16.h
~ wsfont.c
> Convert the Sony wsfonts to ISO-Latin1 encoding, and drop the Sony encoding
> conversion tables; from NetBSD. (miod@)
isofs/cd9660
~ cd9660_vfsops.c
> obey the locking disaplince wrt to VOP_CLOSE during umounts
> and mount error paths.
> ok sturm@ pedro@ (thib@)
~ cd9660_vfsops.c
> Revert last commit (pedro@)
isofs/udf
~ udf_vfsops.c
> obey the locking disaplince wrt to VOP_CLOSE during umounts
> and mount error paths.
> ok sturm@ pedro@ (thib@)
~ udf_vfsops.c
> Revert last commit (pedro@)
kern
~ vfs_bio.c
> Propagate error on getnewbuf(), effectively allowing callers to be
> notified of posted signals for the current process, okay tedu at . > Based on initial diagnosis by R. Tyler Ballance <tyler at bleepsoft.com > (pedro@)
~ sysv_sem.c
> Fix a panic when more than the default number of sempahores are
> allocated. Problem debugged by thib@ (millert@)
~ sysv_sem.c
TAGGED OPENBSD_3_9
> MFC:
> Fix by millert@
> Fix a panic when more than the default number of sempahores are
> allocated. Problem debugged by thib@ (brad@)
msdosfs
~ msdosfs_vfsops.c
> obey the locking disaplince wrt to VOP_CLOSE during umounts
> and mount error paths.
> ok sturm@ pedro@ (thib@)
~ msdosfs_vfsops.c
> Revert last commit (pedro@)
ufs/ffs
~ ffs_vfsops.c
> obey the locking disaplince wrt to VOP_CLOSE during umounts
> and mount error paths.
> ok sturm@ pedro@ (thib@)
~ ffs_vfsops.c
> Revert last commit (pedro@)
== usr.bin =========================================================== 10/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
calendar
~ calendars/calendar.birthday
> fix fidel bday as pt out by some revolutionary comarades (mickey@)
newsyslog
~ newsyslog.8
> missing full stop; from matt jibson (jmc@)
rcs
~ ident.c
> Flush printf buffer before directly writing to STDIN_FILENO. Prevents
> the printf output from printing after the rcs_buf_write_fd output.
> ``ok!'' joris@ (ray@)
~ co.c ~ diff3.c
~ rcs.c
> More strlcpy/strlcat -> snprintf/xasprintf cleanup.
> OK joris@ (ray@)
~ diff3.c
> let ed_patch_lines work with a c command that encompasses the last line
> of the file. The TAILQ structures specifically contain an extra `first
> line'
> to avoid this kind of issue, but no extra last line.
> Instead of walking beyond the end of the structure, and wondering at the
> null pointer, keep a pointer before the deleted block, as this one is
> guaranteed to exist.
> Fix Matthieu's testcase.
> okay joris@, xsa@ (espie@)
~ diff3.c ~ merge.c
~ rcsmerge.c ~ rcsprog.h
> correctly handle merge(1) and rcsmerge(1) -e and -E flags; tests and OK
> ray at .sa@)
split
~ split.1 ~ split.c
> - indicate that -b, -l, and -p are mutually exclusive
> - tidy up the text somewhat
> - make it clear that the original file is not changed
> mostly this is based on a freebsd pr (33852),
> submitted by gary w. swearingen (jmc@)
~ split.1 ~ split.c
> Add -a option as per single unix. Rewrote the code to generate
> file names to be less blecherous, though it is still not pretty.
> OK beck@, with man page help from jmc@ (millert@)
~ split.1
> use .Sq for single letters; (jmc@)
~ split.c
> Check for int oflow w/ -b option and unit suffixes, make usage() __dead,
> use ssize_t not long for read/write length. OK beck@ (millert@)
ssh
~ compress.c ~ monitor.c
~ monitor_wrap.c
> "zlib.h" can be <zlib.h>; ok djm@ markus@ (stevesk@)
~ monitor.c ~ monitor_wrap.c
> Revert previous include file ordering change, for ssh to compile under gcc2
> (or until openssl include files are cleaned of parameter names in function
> prototypes) (miod@)
tsort
~ tsort.1
> +.Sh STANDARDS
> from espie, tweaked by myself; (jmc@)
xargs
~ xargs.1
> a variation of netbsd -r1.18:
> PR/34182 from der Mouse -- clarify -0 option (jmc@)
== usr.sbin ========================================================== 11/11 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
authpf
~ authpf.c
> handle SIGQUIT instead of SIGSTOP, from Stefan Krah (dhartmei@)
cron
~ atrun.c ~ do_command.c
> Add Auto-Submitted header in generated mail as per RFC 3834.
> OK henning@, ckuethe@, krw@, ian at .om Tamas TEVESZ. (millert@)
dhcpd
~ dhcpd.8
> typo in example config; ok henning@ (stevesk@)
~ memory.c
> Fix a reliability issue in dhcpd where a DHCPDISCOVER packet with a 32-byte
> client identifier would cause dhcpd to mistakenly exit. From Debian bug
> 380273, affecting ISC dhcpd 2.x, upon which our dhcpd is based.
> OK beck@ (cloder@)
~ memory.c
TAGGED OPENBSD_3_9
> MFC:
> Fix by cloder@
> Fix a reliability issue in dhcpd where a DHCPDISCOVER packet with a 32-byte
> client identifier would cause dhcpd to mistakenly exit. From Debian bug
> 380273, affecting ISC dhcpd 2.x, upon which our dhcpd is based.
> ok cloder@ (brad@)
~ memory.c
TAGGED OPENBSD_3_8
> MFC:
> Fix by cloder@
> Fix a reliability issue in dhcpd where a DHCPDISCOVER packet with a 32-byte
> client identifier would cause dhcpd to mistakenly exit. From Debian bug
> 380273, affecting ISC dhcpd 2.x, upon which our dhcpd is based.
> ok cloder@ (brad@)
fdformat
~ fdformat.c
> Avoid divide by zero in visual effects; spotted by Tobias Stoeckmann.
> (miod@)
ndp
~ ndp.8
> various fixes; (jmc@)
~ ndp.c
> sync usage(); (jmc@)
ospfd
~ neighbor.c
> Remove a XXX in a comment. (claudio@)
~ ospfd.c
> There is no need to initialize the log framework twice. Especially because
> between the two calls no logging can happen. (claudio@)
pkg_add
~ OpenBSD/PackageName.pm
> typo in comment (bernd@)
route6d
~ route6d.8
> fixes up; (jmc@)
traceroute6
~ traceroute6.8
> - better .Nd
> - remove some unneccessary .Bk/.Ek
> - provide a DESCRIPTION (albeit brief)
> - text improvements
> - use .Ex
> - update SEE ALSO (jmc@)
~ traceroute6.c
> sync usage(); (jmc@)
wsfontload
~ wsfontload.c
> Remove the Sony wsfonts from the list of fonts with special encoding.
> requested by miod@ (kettenis@)
===============================================================================
More information about the owc
mailing list