[odc] Daily src changes for 2009-02-17

ODC auto at squish.net
Wed Feb 18 07:00:01 GMT 2009


OpenBSD src changes summary for 2009-02-17
==========================================

distrib/notes                           lib/libc
libexec/spamd                           regress/libexec
share/man                               sys/arch/i386/conf
sys/arch/m88k/include                   sys/arch/mvme68k/conf
sys/arch/mvme68k/dev                    sys/arch/mvme68k/include
sys/arch/mvme68k/mvme68k                sys/arch/mvme68k/stand/netboot
sys/arch/mvme88k/dev                    sys/arch/mvme88k/mvme88k
sys/arch/socppc/dev                     sys/arch/socppc/socppc
sys/arch/sparc64/sparc64                sys/crypto
sys/dev/pci                             sys/kern
sys/net                                 usr.bin/ssh
usr.sbin/bgpd                           usr.sbin/smtpd
usr.sbin/sysmerge                       

== distrib =========================================================== 01/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/distrib

notes

  ~ socppc/prep                           

  > - mention that you have to increase 'bootdelay' in some cases, as the
  > sata chip is slow to initialize. and otherwise openbsd won't find a root
  > discussed with kettenis@ (jasper@)

  ~ mvme68k/hardware                      ~ mvme68k/install

  > Sync with recent changes. (miod@)

== lib =============================================================== 02/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/lib

libc

  ~ stdio/mktemp.c                        

  > when the template is entirely XXX characters, would crash
  > from Vadim Zhukov <persgray at gmail
  > ok millert (deraadt@)

== libexec =========================================================== 03/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/libexec

spamd

  ~ spamd.8                               

  > clarification for the MX stuff; requested by Stephan A. Rickauer
  > ok beck (jmc@)

== regress =========================================================== 04/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/regress

libexec

  ~ ld.so/Makefile                        

  > Link lazy binding test into ld.so regress suite.
  > ok matthieu@ (jsing@)

== share ============================================================= 05/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/share

man

  ~ man4/pfsync.4                         

  > massage a bit. note that pfsync in openbsd 4.4 and 4.5 are incompatible.
  > requested by deraadt@ tweaks by jmc@ (dlg@)

  ~ man4/bce.4                            

  > Add a BUGS section warning that the device is likely not to work in systems
  > with more than 1GB of memory. (kettenis@)

  - man4/man4.mvme68k/ssh.4               ~ man4/man4.mvme68k/Makefile
  ~ man4/man4.mvme68k/intro.4             + man4/man4.mvme68k/vsbic.4

  > Sync with reality. (miod@)

  ~ man4/osiop.4                          

  > osiop synopsis lines for mvme68k. (miod@)

== sys =============================================================== 06/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/sys

arch/i386/conf

  ~ RAMDISKC                              

  > something has to go... because the piggy developers bloated the kernel
  > again (deraadt@)

arch/m88k/include

  ~ cpu.h                                 

  > Pass a cpu_info * to setsoftipi() so it does not need to curcpu(), which
  > synchronizes the pipeline on 88110. (miod@)

arch/mvme68k/conf

  ~ GENERIC                               ~ MVME147
  ~ MVME162                               ~ MVME167
  ~ MVME177                               ~ RAMDISK
  ~ files.mvme68k                         

  > Add real bus_space and bus_dma to mvme68k, so that it can benefit from
  > recent driver work; this gives us better osiop and vs drivers, vsbic
  > (although currently disabled on 68060 boards until more bugs are fixed),
  > and the ability to boot and root off vs and vsbic.
  > Existing code is not modified and still uses memory mapped structures;
  > this might change eventually as more code gets factorized and moved to
  > MI land. (miod@)

arch/mvme68k/dev

  - ssh.c                                 - ssh_script.out
  - ssh_script.ss                         - sshdma.c
  - sshreg.h                              - sshvar.h
  ~ if_le.c                               ~ if_lereg.h
  ~ mc.c                                  ~ pcc.c
  ~ pcctwo.c                              ~ vme.c
  ~ vme.h                                 ~ vs.c
  ~ vsreg.h                               ~ vsvar.h
  + bpp.c                                 + bppvar.h
  + mainbus.c                             + osiop_mvme68k.c
  + vsbic.c                               

  > Add real bus_space and bus_dma to mvme68k, so that it can benefit from
  > recent driver work; this gives us better osiop and vs drivers, vsbic
  > (although currently disabled on 68060 boards until more bugs are fixed),
  > and the ability to boot and root off vs and vsbic.
  > Existing code is not modified and still uses memory mapped structures;
  > this might change eventually as more code gets factorized and moved to
  > MI land. (miod@)

arch/mvme68k/include

  ~ autoconf.h                            ~ bus.h

  > Add real bus_space and bus_dma to mvme68k, so that it can benefit from
  > recent driver work; this gives us better osiop and vs drivers, vsbic
  > (although currently disabled on 68060 boards until more bugs are fixed),
  > and the ability to boot and root off vs and vsbic.
  > Existing code is not modified and still uses memory mapped structures;
  > this might change eventually as more code gets factorized and moved to
  > MI land. (miod@)

arch/mvme68k/mvme68k

  ~ autoconf.c                            + bus_dma.c

  > Add real bus_space and bus_dma to mvme68k, so that it can benefit from
  > recent driver work; this gives us better osiop and vs drivers, vsbic
  > (although currently disabled on 68060 boards until more bugs are fixed),
  > and the ability to boot and root off vs and vsbic.
  > Existing code is not modified and still uses memory mapped structures;
  > this might change eventually as more code gets factorized and moved to
  > MI land. (miod@)

arch/mvme68k/stand/netboot

  ~ version.c                             

  > Forgot to crank version after ethernet buffer address change. (miod@)

  ~ Makefile                              ~ if_ie.c
  ~ if_le.c                               ~ if_lereg.h

  > Move ethernet buffers higher in memory, just below our text, to allow
  > larger bsd.rd to load correctly. Tested on MVME147 (if_le) and MVME167
  > (if_ie). (miod@)

arch/mvme88k/dev

  ~ vme.c                                 

  > Be sure to pass a valid confargs * to child devices of vme0 when attaching,
  > for they will attempt to dereference it and having NULL readable while
  > in the kernel is just cheating. (miod@)

arch/mvme88k/mvme88k

  ~ m197_machdep.c                        

  > Pass a cpu_info * to setsoftipi() so it does not need to curcpu(), which
  > synchronizes the pipeline on 88110. (miod@)

arch/socppc/dev

  ~ if_tsec.c                             

  > - increase the delays in tsec_mii_{read,write}reg() , makes my n1200
  > happier
  > hints and ok kettenis@ (jasper@)

arch/socppc/socppc

  ~ machdep.c                             

  > Remove debug printf. (kettenis@)

  ~ cpu.c                                 

  > Declare cpu_model as an array instead of a pointer; makes sysctl hw.model
  > return a sensible value instead of an empty string.
  > tested by jasper@ (kettenis@)

arch/sparc64/sparc64

  ~ machdep.c                             

  > Remove an if 0ed out chunk for BUS_DMA_COHERENT in bus_dmamap_map.
  > sparc64 is dma coherent and won't be needing this.
  > ok kettenis at . (oga@)

crypto

  ~ cryptodev.c                           

  > The loop here preparing the uio always added the length of the first
  > iovec, not the correct one. It worked ok since iovcnt was always 1.
  > Since it's unlikely to be any other number, remove the loop and just add
  > the one length we care about.
  > "go ahead" deraadt at . (oga@)

  ~ cryptodev.c                           

  > variable no longer used (deraadt@)

dev/pci

  ~ agp_i810.c                            

  > Don't leak the dmamap on freeing memory. (oga@)

  ~ ips.c                                 

  > - add more bioctl support, now physical drives info available
  > - shrink dmesg strings (grange@)

kern

  ~ subr_pool.c                           

  > ensure all pi_magic checks are inside DIAGNOSTIC (deraadt@)

net

  ~ if_pfsync.c                           

  > fix uninitialized variable.
  > found by LLVM/Clang Static Analyzer.
  > ok dlg@ (chl@)

  ~ if_pfsync.c                           

  > init the tdb tailq. hopefully this fixes sthens crash. (dlg@)

  ~ if_pfsync.c                           

  > assert copyright over the changes i made.
  > reminded by deraadt@ (dlg@)

  ~ if_pfsync.c                           

  > // style comments shouldnt be in the tree.
  > reminded by deraadt@ (dlg@)

== usr.bin =========================================================== 07/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin

ssh

  ~ ssh_config                            

  > sync with revised default ciphers; pointed out by dkrause@ (djm@)

== usr.sbin ========================================================== 08/08 ==

  http://www.openbsd.org/cgi-bin/cvsweb/src/usr.sbin

bgpd

  ~ rde_attr.c                            

  > Fix aspath_prepend() in the case a AS-PATH has a sequence of 255 elements
  > and we try to prepend. The result was a corrupt AS-PATH and a RDE fatal
  > later on when some other operation was run on that path.
  > Found the hard way by henning@ and sthen at . OK henning@, sthen@ (claudio@)

smtpd

  ~ mfa.c                                 

  > - remove two debug messages (gilles@)

  ~ smtp.c                                ~ smtp_session.c

  > after a message has been commited to queue, do not incondtionnally clear
  > the message id and uid immediately. only do it if session has been flagged
  > with F_QUIT, otherwise session_pickup() will do it when in state S_DONE.
  > this fixes a bug reported by pea@ where the message id was not displayed
  > in the "message accepted for delivery" line. (gilles@)

  ~ lka.c                                 

  > kill unused func lka_resolve_rcpt() (jacekm@)

  ~ control.c                             ~ smtpctl.c

  > imsg_init depends on ibuf being calloc'd. fix few places where malloc
  > is used instead; ok gilles@
  > Problem made visible by malloc.conf = AFGJ, and pointed out by otto@
  > (jacekm@)

  ~ makemap.c                             ~ smtpd.h

  > make newaliases read aliases path from smtpd.conf; idea discussed with
  > gilles@, pyr@ and henning@, diff ok gilles at . (jacekm@)

  ~ makemap.8                             ~ makemap/Makefile
  + newaliases.8                          

  > sync makemap man page with the code. create separate page for newaliases
  > which points at makemap as the preferred utility; ok gilles@ (jacekm@)

  ~ makemap.c                             

  > - flock input file while the database is being built
  > - make newaliases finish with short summary
  > ok gilles@ (jacekm@)

sysmerge

  ~ sysmerge.sh                           

  > - better output
  > ok sthen@ (ajacoutot@)

  ~ sysmerge.8                            ~ sysmerge.sh

  > - always run pwd_mdb, newaliases, MAKEDEV and cap_mkdb even when not in
  > auto-mode so that the system is not left in an inconsistent state
  > - unify messages
  > - tweak manpage according to new behavior
  > based on an initial patch from matthieu@
  > ok matthieu@ (ajacoutot@)

===============================================================================


More information about the odc mailing list