[odc] Daily src changes for 2008-03-18

[ODC]

OpenBSD src changes summary for 2008-03-18
==========================================

share/man                               sys/arch/i386/i386
sys/arch/sparc64/sparc64                sys/dev/usb
sys/net                                 usr.bin/cvs
usr.bin/netstat                         usr.sbin/authpf
usr.sbin/rtadvd                         usr.sbin/rtsold
usr.sbin/snmpd                          

== share ============================================================= 01/04 ==

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

man

  ~ man4/uslcom.4                         

  > IRZ MC35pu GSM Terminal
  > From/tested by Sergey Prysiazhnyi <apelsin at atmnis.com> (jsg@)

== sys =============================================================== 02/04 ==

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

arch/i386/i386

  ~ freebsd_machdep.c                     ~ linux_machdep.c
  ~ svr4_machdep.c                        

  > Clear PSL_D when entering signal handlers; pointed out by dfly people;
  > ok miod@ (kettenis@)

arch/sparc64/sparc64

  ~ locore.s                              

  > Get rid of some dead wood. (kettenis@)

dev/usb

  ~ usbdevs                               ~ uftdi.c
  ~ if_aue.c                              

  > Bring in some changes from DragonFly:
  > - New matching logic so we don't need to check against vid/pid
  > to figure out which kind of chip we have
  > - Match list in a struct instead of messy if statement (hurrah!)
  > - More generic dual port handling
  > - Insane amount of device additions, mostly found in Linux driver
  > (aue included as a usb vendor got renamed)
  > Tested by ckuethe@, 'looks super sensible' deraadt@ (jsg@)

  ~ usbdevs.h                             ~ usbdevs_data.h

  > regen (jsg@)

  ~ usbdevs                               

  > IRZ MC35pu GSM Terminal
  > From Sergey Prysiazhnyi <apelsin at atmnis.com> (jsg@)

  ~ usbdevs.h                             ~ usbdevs_data.h

  > regen (jsg@)

  ~ uslcom.c                              

  > IRZ MC35pu GSM Terminal
  > From/tested by Sergey Prysiazhnyi <apelsin at atmnis.com> (jsg@)

net

  ~ if_pppoe.c                            

  > Don't use MBUFLOCK() for ifqueue protection use the correct splnet() dance
  > instead. OK canacar@ (claudio@)

  ~ if_pppoe.c                            

  > Don't use ph after the m_pulldown() calls because the mbuf chain could be
  > modified. Better save than sorry. OK canacar@ (claudio@)

== usr.bin =========================================================== 03/04 ==

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

cvs

  ~ tag.c                                 

  > sync the usage with the manpage, use "symbolic_tag" instead of "tag"
  > for the argument to distinguish it from the command keyword which is
  > less confusing.
  > ok joris@ (reyk@)

netstat

  ~ if.c                                  

  > Make -I work correctly. RTM_NEWADDR needs to filter the ifname as well.
  > Found by and tested jmc@ (claudio@)

== usr.sbin ========================================================== 04/04 ==

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

authpf

  ~ authpf.8                              

  > Fix mention of authpf_users table (s/authpf users/authpf_users/).
  > ok jmc@, mcbride@ (merdely@)

rtadvd

  ~ rtadvd.c                              

  > correct CMSG_SPACE and CMSG_LEN handling as done in other places.  These
  > are the complicated cases where multiple messages are sent.  discussed
  > with kettenis, hshoexer, and rtsol tested by jmc - thanks (deraadt@)

rtsold

  ~ probe.c                               ~ rtsol.c

  > correct CMSG_SPACE and CMSG_LEN handling as done in other places.  These
  > are the complicated cases where multiple messages are sent.  discussed
  > with kettenis, hshoexer, and rtsol tested by jmc - thanks (deraadt@)

snmpd

  ~ ber.c                                 

  > fix bug that ber_oid2ber skips the first byte when encoding powers of
  > 128.  E.g., it will encode 128 as 0x00 instead of 0x81 0x00.
  > From Matthew at Dempsky dot org (reyk@)

  ~ ber.3                                 ~ ber.c
  ~ ber.h                                 ~ mib.c
  ~ trap.c                                

  > remove ber_add_astring and make it the default behaviour in
  > ber_add_string, ber_add_nstring, and ber_add_bitstring to allocate and
  > copy the provided buffer instead of just attaching it to the ber
  > element.  this may add some overhead but fixes and avoids some bugs
  > with static buffers and simplifies the ber api again.
  > From Matthew at Dempsky dot org
  > With input from claudio@ (reyk@)

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