[odc] Daily src changes for 2009-04-23
ODC
auto at squish.net
Fri Apr 24 07:00:01 BST 2009
OpenBSD src changes summary for 2009-04-23
==========================================
distrib/miniroot distrib/sets
etc/root/root.mail etc/skel/dot.mailrc
sbin/disklabel sbin/ping
share/man sys/arch/amd64/amd64
sys/arch/amd64/include sys/arch/amd64/stand/installboot
sys/arch/i386/stand/installboot sys/dev/ic
sys/dev/pci sys/net
usr.bin/mail usr.bin/systat
usr.sbin/bgpctl usr.sbin/bgpd
usr.sbin/smtpd
== distrib =========================================================== 01/07 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/distrib
miniroot
~ install.sh ~ install.sub
> Near the end of the installation, prompt for the setup of an initial
> user, who will also be placed in wheel. The setup of that user is
> expected to be mostly compatible with what useradd(8) would have done.
> krw will probably improve the ksh code in here after this :) (deraadt@)
~ install.sub
> Give the user a copy of root's (install time) mail (deraadt@)
~ install.sub
> Cut the ctime= message strings out of the mount -v output to make this
> a lot prettier. We still leave this output in -v mode because we want
> people to see how smart the install script is about making the different
> partitions have different nosetuid, nodev modes
> assistance with the right ksh code from drahn, beck, and espie (deraadt@)
~ install.sub
> simplify mount -v sed; from guenther (deraadt@)
~ install.sub
> The active mode ftp stuff is archaic, so delete it. People who still
> have that problem can use the plethora of www servers we have instead of
> ftp. Or as sthen pointed out, this is almost assuredly not biting anyone
> today since we have received no complaits of the ftplist fetching
> failing (that did not use the -A flag) (deraadt@)
~ install.sh
> configure the network unconditionally. You have an opportunity to bail
> out of this by saying "done" when you get to the first interface. Then
> it will handle the domain / DNS question, but since no interface is setup
> it cleverly already skips the default route question. Another useless
> question bites the dust.
> ok krw (deraadt@)
sets
~ lists/base/md.amd64 ~ lists/comp/md.amd64
> sync (deraadt@)
== etc =============================================================== 02/07 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/etc
root/root.mail
~ root/root.mail
> shorten subject to be prettier (deraadt@)
skel/dot.mailrc
~ skel/dot.mailrc
> the crt option is a good default for new users; ok millert (deraadt@)
== sbin ============================================================== 03/07 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sbin
disklabel
~ editor.c
> Unrevert reversion of r1.192. This time make sure 'aflag' is treated
> the same way 'dflag' is treated. i.e. reset it when the label is
> written out so a 'w' followed by a 'q' exits the editor without
> further questions. Also set it whenever the 'A' command is executed.
> Suggestions by Alexander Hall.
> ok deraadt@ (krw@)
ping
~ ping.c
> ping -v can segfault when displaying received icmp packets that aren't
> echo replies. it's because the pointer arithmetic is scaled to the size
> of struct icmp, so 28*28 bytes are added rather than just 28. fortunately
> a correct value was calculated 2 lines earlier, so we can just use that.
> "thats a cool fix" dlg, ok deraadt (sthen@)
== share ============================================================= 04/07 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/share
man
~ man7/mdoc.7
> improve the description for the .Nd macro in a way it is not only
> more accurate but also a good mnemonic.
> ok jmc@ (sobrado@)
== sys =============================================================== 05/07 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/sys
arch/amd64/amd64
~ est.c
> add missing newline in the unknown model case (jsg@)
~ cpu.c ~ genassym.cf
~ lapic.c ~ locore.S
~ pmap.c ~ vector.S
~ vm_machdep.c
> Make pmap_deactivate a NOP.
> Instead of keeping a bitmask of on which cpu the pmap might be active which
> we clear in pmap_deactivate, always keep a pointer to the currently loaded
> pmap in cpu_info. We can now optimize a context switch to the kernel pmap
> (idle and kernel threads) to keep the previously loaded pmap still loaded
> and then reuse that pmap if we context switch back to the same process.
> Introduce a new IPI to force a pmap reload before the pmap is destroyed.
> Clean up cpu_switchto.
> toby@ ok (art@)
arch/amd64/include
~ cpu.h ~ i82489var.h
~ pmap.h
> Make pmap_deactivate a NOP.
> Instead of keeping a bitmask of on which cpu the pmap might be active which
> we clear in pmap_deactivate, always keep a pointer to the currently loaded
> pmap in cpu_info. We can now optimize a context switch to the kernel pmap
> (idle and kernel threads) to keep the previously loaded pmap still loaded
> and then reuse that pmap if we context switch back to the same process.
> Introduce a new IPI to force a pmap reload before the pmap is destroyed.
> Clean up cpu_switchto.
> toby@ ok (art@)
arch/amd64/stand/installboot
~ installboot.c
> -v was printing some information which is irrelevant or duplicate. This
> shrinks the output of installboot by a few lines. (deraadt@)
arch/i386/stand/installboot
~ installboot.c
> -v was printing some information which is irrelevant or duplicate. This
> shrinks the output of installboot by a few lines. (deraadt@)
dev/ic
~ hme.c
> Clear IFF_RUNNING and IFF_OACTIVE flags and cancel watchdog timer in
> hme_stop(). From Brad. (kettenis@)
dev/pci
~ if_bge.c ~ if_bgereg.h
> Start using extended buffer descriptors on the jumbo ring and use MCLGETI
> to
> allocate mbufs for it. Another jumbo allocator bites the dust!
> ok dlg@ (kettenis@)
net
~ pf.c
> print the type of the icmp message we're bitching about when debugging is
> turned up in pf_icmp_state_lookup.
> ok sthen@ (dlg@)
== usr.bin =========================================================== 06/07 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin
mail
~ misc/mail.help
> document the "more" command in the help page; ok millert (deraadt@)
systat
~ pftop.c
> sync actiontypes (pass, block etc) with pfvar so that match rules are
> displayed correctly (actually thsi was out of sync for more than just
> match)
> From: Craig Barraclough <craigba at afnsecurity dot com> (henning@)
== usr.sbin ========================================================== 07/07 ==
http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin
bgpctl
~ parser.c
> allow bgpctl and bgpd.conf to contain 32-bit ASN written in ASPLAIN
> format (RFC5396). ok claudio@ henning@ (sthen@)
bgpd
~ bgpd.h ~ parse.y
> allow bgpctl and bgpd.conf to contain 32-bit ASN written in ASPLAIN
> format (RFC5396). ok claudio@ henning@ (sthen@)
~ rde.c ~ rde.h
~ rde_decide.c ~ rde_rib.c
> Rework the way we handle announced networks. Instead of two freak rde_peers
> use one that is less freaky. Merge bgpctl and config networks into one
> tree.
> First step of a larger change in the RDE and this goes now in to allow to
> move forward. (claudio@)
smtpd
~ smtpctl.c
> sort "show stats" output (jacekm@)
===============================================================================
More information about the odc
mailing list