[odc] Daily src changes for 2006-02-27

[ODC]

OpenBSD src changes summary for 2006-02-27
==========================================

share/man                               sys/arch/i386/i386
sys/arch/mips64/include                 sys/conf
sys/dev/pci                             sys/kern
sys/net                                 

== share ============================================================= 01/02 ==

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

man

  ~ man4/nfe.4                            

  > There apears to be a problem with nfe+ciphy that prevents
  > the interface from being useable.  Mention this a known problem.
  > commit allowed by deraadt@ (jsg@)

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

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

arch/i386/i386

  ~ est.c                                 

  > Add 2 additional CPUs and uncomment a 3rd.	Based on this intel doc:
  > download.intel.com/design/mobile/datashts/30218908.pdf
  > cpu_id part of the diff was done by Dimitry Andric <dimitry at andric dot
  > com>
  > Thanks :-)
  > Pointed to the doc and ok deraadt@ (marco@)

  ~ est.c                                 

  > Add the 770 model.	From NetBSD through Gregory Steuck
  > <greg at y2006 dot nest dot cx>
  > Suggestions and ok deraadt@
  > ok mbalmer@ (marco@)

arch/mips64/include

  ~ endian.h                              

  > mips requires strict alignment; spotted after a discussion on undeadly;
  > ok deraadt@ kettenis@ (miod@)

conf

  ~ newvers.sh                            

  > stop being -beta (deraadt@)

dev/pci

  ~ ppb.c                                 

  > Workaround for bridge attachment where nothing lives behind these 'rare'
  > bridges and attaching them causes the machine to hang.
  > A better fix needs to be found after release.
  > ok brad@ kettenis@ deraadt@ (drahn@)

  ~ piixpm.c                              

  > Increase delay value, fixes pr5018.
  > With a help from kettenis@, ok deraadt at .range@)

kern

  ~ kern_ktrace.c                         

  > When checking syscall numbers, qualify the test with the emulation as well.
  > fixes potential panics on emulated binaries. ok deraadt@ (niklas@)

  ~ uipc_usrreq.c                         

  > Do not stupidly panic but return ENOTCONN when trying to pass fds on an
  > unconnected socket; reported by Christian Biere <christianbiere at gmx dot
  > de>
  > ok claudio@ deraadt@ (miod@)

net

  ~ bpf_filter.c                          

  > add missing break; now filters containing a division can pass
  > validation; from NetBSD; ok deraadt@ (otto@)

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