[odc] Daily src changes for 2002-07-30

ODC auto at squish.net
Wed Jul 31 08:00:16 BST 2002


OpenBSD src changes summary for 2002-07-30
==========================================

bin/systrace                            distrib/notes
games/backgammon                        games/wump
lib/libc                                lib/libssl
libexec/ld.so                           libexec/tcpd
regress/sbin                            sbin/pfctl
share/man                               sys/arch/alpha/include
sys/arch/hppa/include                   sys/arch/i386/include
sys/arch/m68k/include                   sys/arch/mvme88k/include
sys/arch/powerpc/include                sys/arch/sparc/include
sys/arch/sparc64/dev                    sys/arch/sparc64/include
sys/arch/vax/include                    sys/dev
sys/dev/sbus                            sys/kern
sys/net                                 sys/netinet
sys/netinet6                            sys/nfs
usr.bin/ssh                             usr.sbin/httpd
usr.sbin/named                          usr.sbin/pppd
usr.sbin/rdate                          

== bin =============================================================== 01/11 ==

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

systrace

  ~ intercept.c                           

  > solve a problem with realpath when the last component of the path is
  > a directory without S_IXUSR; tested by me and dugsong.

  ~ parse.y                               

  > extern decls should be outside of function.

  ~ parse.y                               

  > include filter.h, dont' duplicate prototype

  ~ lex.l                                 

  > sync prototype for yyerror().

  ~ alias.c                               ~ systrace.c

  > SPLAY_INSERT is a void function

  ~ alias.c                               

  > oops, i've been looking at older tree.h

  ~ register.c                            ~ systrace-translate.c
  ~ systrace.h                            

  > avoid using same variable name for global and auto variable.

  ~ openbsd-syscalls.c                    

  > sync function decl and prototype (static-ness)

  ~ intercept.c                           

  > sometimes no-return syscalls (execve) emit errno < 0.  ignore them.

  ~ systrace.1                            

  > use Nm

  ~ systrace.1                            

  > obey section order to mandoc.  sync with netbsd.

== distrib =========================================================== 02/11 ==

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

notes

  ~ sparc64/hardware                      

  > Mention PGX and PGX24 here too.

== games ============================================================= 03/11 ==

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

backgammon

  ~ common_source/back.h                  ~ common_source/init.c
  ~ common_source/table.c                 

  > Kill buffer overflow.

  + backgammon/pubeval.c                  ~ backgammon/Makefile
  ~ backgammon/backgammon.6               ~ backgammon/backlocal.h
  ~ backgammon/move.c                     

  > Replace the "blows chunks" algorithm with pubeval, a public domain algorith
  > m
  > which plays an acceptable, if not optimal, game.  pubeval author approves.

wump

  ~ wump.c                                

  > ansi wump, plus a more accurate comment

== lib =============================================================== 04/11 ==

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

libc

  ~ stdlib/calloc.c                       

  > return failure if integer overflow happens.  sigh; too people had to
  > help get this right.

  ~ gen/exec.c                            

  > be even more careful with strlcpy()

  ~ net/gethostnamadr.c                   

  > more strlcpy; itojun ok

  ~ rpc/xdr_array.c                       

  TAGGED OPENBSD_3_0
  > Errata #029 (deraadt):
  > careful malloc

  ~ stdlib/calloc.c                       

  TAGGED OPENBSD_3_0
  > MFC (deraadt):
  > return failure if integer overflow happens.  sigh; too people had to
  > help get this right.

  ~ rpc/xdr_array.c                       

  TAGGED OPENBSD_3_1
  > Pull in patch from current:
  > Fix (deraadt), Errata 012:
  > careful malloc

  ~ stdlib/calloc.c                       

  TAGGED OPENBSD_3_1
  > Pull in patch from current:
  > Fix (deraadt):
  > return failure if integer overflow happens.  sigh; too people had to
  > help get this right.

  ~ gen/scandir.c                         ~ gen/opendir.c

  TAGGED OPENBSD_3_1
  > malloc paranoia; it is unlikely that any filesystem will support
  > enough directory entries to cause a problem but it is good form anyway.
  > deraadt@ OK.

libssl

  ~ src/CHANGES                           ~ src/crypto/cryptlib.h
  ~ src/crypto/asn1/asn1_lib.c            ~ src/crypto/conf/conf_def.c
  ~ src/crypto/objects/obj_dat.c          ~ src/ssl/s2_clnt.c
  ~ src/ssl/s2_lib.c                      ~ src/ssl/s2_srvr.c
  ~ src/ssl/s3_clnt.c                     ~ src/ssl/s3_srvr.c
  ~ src/ssl/ssl.h                         ~ src/ssl/ssl_asn1.c
  ~ src/ssl/ssl_err.c                     ~ src/ssl/ssl_lib.c
  ~ src/ssl/ssl_locl.h                    ~ src/ssl/ssl_sess.c

  TAGGED OPENBSD_3_0
  > Errata #030 (markus):
  > Fixes for the "OpenSSL Security Advisory [30 July 2002]"

  ~ src/CHANGES                           ~ src/crypto/cryptlib.h
  ~ src/crypto/asn1/asn1_lib.c            ~ src/crypto/conf/conf_def.c
  ~ src/crypto/objects/obj_dat.c          ~ src/ssl/s2_clnt.c
  ~ src/ssl/s2_lib.c                      ~ src/ssl/s2_srvr.c
  ~ src/ssl/s3_clnt.c                     ~ src/ssl/s3_srvr.c
  ~ src/ssl/ssl.h                         ~ src/ssl/ssl_asn1.c
  ~ src/ssl/ssl_err.c                     ~ src/ssl/ssl_lib.c
  ~ src/ssl/ssl_locl.h                    ~ src/ssl/ssl_sess.c
  ~ ssl/Makefile                          

  TAGGED OPENBSD_3_0
  > apply patches from OpenSSL Security Advisory [30 July 2002],
  > http://marc.theaimsgroup.com/?l=openssl-dev&m=102802395104110&w=2

  ~ src/CHANGES                           ~ src/crypto/cryptlib.h
  ~ src/crypto/asn1/asn1_lib.c            ~ src/crypto/conf/conf_def.c
  ~ src/crypto/objects/obj_dat.c          ~ src/ssl/s2_clnt.c
  ~ src/ssl/s2_lib.c                      ~ src/ssl/s2_srvr.c
  ~ src/ssl/s3_clnt.c                     ~ src/ssl/s3_srvr.c
  ~ src/ssl/ssl.h                         ~ src/ssl/ssl_asn1.c
  ~ src/ssl/ssl_err.c                     ~ src/ssl/ssl_lib.c
  ~ src/ssl/ssl_locl.h                    ~ src/ssl/ssl_sess.c

  TAGGED OPENBSD_3_1
  > Pull in patch from current:
  > Fix (markus), errata 013:
  > apply patches from OpenSSL Security Advisory [30 July 2002],
  > http://marc.theaimsgroup.com/?l=openssl-dev&m=102802395104110&w=2

  ~ src/crypto/conf/conf_mod.c            ~ src/crypto/engine/hw_cswift.c
  ~ src/ssl/s3_clnt.c                     ~ src/ssl/s3_srvr.c

  TAGGED OPENBSD_3_1
  > sync with http://www.openssl.org/news/patch_20020730_0_9_7.txt
  > (adds fix for unused kerberos and engine code, and some more
  > assertions, as well as a 64bit integer string fix for conf_mod.c)

== libexec =========================================================== 05/11 ==

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

ld.so

  ~ ldconfig/ldconfig.c                   ~ ldconfig/shlib.c

  > knf & ansi

tcpd

  ~ tcpd/tcpd.c                           ~ tcpdchk/inetcf.c
  ~ tcpdchk/scaffold.c                    ~ tcpdchk/tcpdchk.c

  > some snprintf and strlcpy

== regress =========================================================== 06/11 ==

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

sbin

  ~ pfctl/pf3.in                          ~ pfctl/pf3.ok

  > positive regression test for flag handling

  + pfctl/pfail13.in                      ~ pfctl/Makefile

  > negative regression test for flags handling

== sbin ============================================================== 07/11 ==

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

pfctl

  ~ parse.y                               

  > allow to specify flags on all rules that include tcp.
  > these are valid:
  > pass in from any to any flags S
  > pass in proto { tcp, udp, icmp } from any to any flags S
  > pass in proto tcp from any to any flags S
  > these are invalid:
  > pass in proto { udp, icmp } from any to any flags S
  > pass in proto udp from any to any flags S
  > ok "I've lost my slacker status for at least a week" frantzen@
  > ok pb@, dhartmei@, deraadt@

  ~ parse.y                               

  > grmpf.
  > in some cases, on non-tcp rules flags weren't resetted. cosmetical only
  > problem. but, well, checking for r->flags and r->flagset if we could have
  > assigned them zero just one round ago is just stupid, and it's not needed t
  > o
  > check them at all.
  > ok pb@, dhartmei@

== share ============================================================= 08/11 ==

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

man

  ~ man7/securelevel.7                    

  > Clarify time handling at securelevel 2. Idea from mpech@ ok millert@

  ~ man5/pf.conf.5                        

  > backout, this will go in in little pieces
  > as advised by theo and henning

  ~ man5/pf.conf.5                        

  > 65335->65535 typo
  > henning ok@

  ~ man5/pf.conf.5                        

  > .Sh GRAMMAR moves to bottom, it's a reference and not readable
  > for the casual user in first place
  > ok henning@

  ~ man5/pf.conf.5                        

  > typo/pasto in route-to/dup-to syntax
  > ok henning@

  ~ man5/pf.conf.5                        

  > BNF catchup to reality:
  > - set loginterface none
  > - add "self" to hosts
  > ok henning@

  ~ man5/pf.conf.5                        

  > Merge filter and nat BNF for simplification:
  > - top of reduction is now 'line', better to add more keywords later on
  > - reorder, group
  > - remove double productions
  > ok dhartmei@, henning@

  ~ man5/pf.conf.5                        

  > BNF catchup and consolidation of interface name handling:
  > ok henning@, dhartmei@

  ~ man5/pf.conf.5                        

  > BNF is now in sync with reality:
  > - commas are optional in lists
  > ok henning@, dhartmei@

  ~ man4/man4.sparc64/creator.4           ~ man4/man4.sparc64/cgsix.4

  > add docs for flags to disable console acceleration

== sys =============================================================== 09/11 ==

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

arch/alpha/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/hppa/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/i386/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/m68k/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/mvme88k/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/powerpc/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/sparc/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/sparc64/dev

  ~ vgafb.c                               

  > simplify vgafb_mmap()

  ~ creator.c                             ~ creatorvar.h

  > allow cf_flags to determine whether console acceleration should be used (de
  > faults to yes)

arch/sparc64/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

arch/vax/include

  ~ limits.h                              

  > Add SIZE_MAX define.  This is the same as SIZE_T_MAX but more portable.
  > The only OSes I've seen that use SIZE_T_MAX are 4.4BSD-derived whereas
  > SYSV things seem to use SIZE_MAX.  It is also consistent with SSIZE_MAX
  > (which we already have).  deraadt@ OK

dev

  ~ systrace.c                            

  > whitespace at EOL

  ~ systrace.c                            

  > minor KNF.  pid_t is unsigned.

dev/sbus

  ~ cgthree.c                             

  > handle dumbfb and mapped mode requests differently

  ~ cgsix.c                               ~ cgsixreg.h

  > - Allow disabling of console acceleration with cf_flags
  > - handle native and dumb mappings separately

  ~ cgsix.c                               ~ cgsixreg.h

  > - Whoops, I had the mode mask and value bits backwards
  > - make sure FBC_MODE is reset when switching back to WSDISPLAYIO_MODE_EMUL

kern

  ~ tty.c                                 

  > Release sessions to avoid memory leak. From NetBSD. ok deraadt@

net

  ~ if_strip.c                            

  > no need for check if m0 is NULL, as we have touched it already.
  > From: tedu <grendel at zeitbombe.org
netinet

  ~ ip_esp.c                              

  > Be sure to check the integrity verifier for packets that didn't have it don
  > e
  > in hardware; from angelos

netinet6

  ~ icmp6.c                               ~ nd6_rtr.c

  > remove unneeded NULL pointer checks.
  > From: tedu <grendel at zeitbombe.org
nfs

  ~ nfs_socket.c                          

  > Bug/type found by tedu

== usr.bin =========================================================== 10/11 ==

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

ssh

  ~ auth-options.c                        ~ servconf.c
  ~ servconf.h                            ~ session.c
  ~ sshd_config                           ~ sshd_config.5

  > add PermitUserEnvironment (off by default!); from dot at dotat.atk provos, 
  > deraadt

== usr.sbin ========================================================== 11/11 ==

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

httpd

  ~ src/modules/standard/mod_rewrite.c    

  > two more strip_chroot
  > found by Steph <sengel at melshake.comwho also tested this.
  > I'm very happy with the way you help here. Thanks a lot.

  ~ src/modules/proxy/mod_proxy.c         

  > strip_chroot here as well.
  > pointed out by sengel at melshake dot com

named

  ~ named/storage.c                       

  > for the disabled DSTORAGE option, fix the local calloc() here in the same w
  > ay as libc; eugene at securityarchitects.com
pppd

  ~ main.c                                

  TAGGED OPENBSD_3_0
  > Errata #028 (millert):
  > Replace an instance of chmod() with fchmod()

  ~ main.c                                

  TAGGED OPENBSD_3_1
  > Pull in patch from current:
  > Fix (millert), errata 011:
  > Replace an instance of chmod() with fchmod()

rdate

  ~ ntpleaps.c                            

  > knf

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




More information about the odc mailing list