[odc] Daily src changes for 2004-06-21

ODC auto at squish.net
Tue Jun 22 08:00:52 BST 2004


OpenBSD src changes summary for 2004-06-21
==========================================

distrib/sets                            gnu/usr.bin/cvs
gnu/usr.bin/lynx                        lib/libc
libexec/spamd                           regress/lib
regress/sys                             sbin/ifconfig
sbin/isakmpd                            sbin/pfctl
sbin/raidctl                            share/man
sys/arch/amd64/amd64                    sys/arch/i386/i386
sys/arch/i386/include                   sys/arch/sparc64/conf
sys/arch/sparc64/include                sys/compat/common
sys/compat/hpux                         sys/ddb
sys/dev/ic                              sys/dev/pci
sys/dev/sbus                            sys/kern
sys/lib/libkern                         sys/net
sys/netatalk                            sys/netccitt
sys/netinet                             sys/netinet6
sys/netiso                              sys/nfs
sys/sys                                 sys/ufs/ext2fs
sys/ufs/ffs                             sys/ufs/lfs
sys/ufs/ufs                             sys/uvm
usr.bin/du                              usr.bin/pr
usr.bin/ssh                             usr.bin/xstr
usr.sbin/pkg_add                        usr.sbin/tokenadm

== distrib =========================================================== 01/10 ==

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

sets

  ~ lists/base/md.alpha                   ~ lists/base/md.amd64
  ~ lists/base/md.cats                    ~ lists/base/md.hp300
  ~ lists/base/md.hppa                    ~ lists/base/md.i386
  ~ lists/base/md.luna88k                 ~ lists/base/md.mac68k
  ~ lists/base/md.macppc                  ~ lists/base/md.mvme68k
  ~ lists/base/md.mvme88k                 ~ lists/base/md.mvmeppc
  ~ lists/base/md.sparc                   ~ lists/base/md.sparc64
  ~ lists/base/md.vax                     ~ lists/comp/mi
  ~ lists/man/mi                          

  > sync (deraadt@)

  ~ lists/man/mi                          

  > sync (deraadt@)

== gnu =============================================================== 02/10 ==

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

usr.bin/cvs

  ~ Makefile.bsd-wrapper                  

  > propagate CPP here as well (avsm@)

usr.bin/lynx

  ~ Makefile.bsd-wrapper                  

  > propagate CPP to build system as well, millert@ ok (avsm@)

  ~ Makefile.bsd-wrapper                  

  > back out CPP change; obviously not tested.	On at least arm, kills at
  > least ANSI_VARARGS deep inside the configure.  Sorry -- try again.
  > (deraadt@)

  ~ config.guess                          

  > Recognize cats as an arm machine. Dunno how this was missed before.
  > (drahn@)

== lib =============================================================== 03/10 ==

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

libc

  ~ arch/i386/sys/cerror.S                

  > Use _ENTRY(CERROR) instead of CERROR so correct type info is generated.
  > ok marc@ (drahn@)

  ~ stdlib/strtonum.c                     

  > errno -> err in a structure.  OK deraadt@ (marc@)

  ~ db/hash/hash.c                        ~ db/hash/hash.h
  ~ db/hash/ndbm.c                        

  > another errno -> err in a structure.  OK deraadt@ (marc@)

== libexec =========================================================== 04/10 ==

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

spamd

  ~ grey.c                                ~ spamd.c

  > use getaddr/nameinfo for address resolution.  beck, henning ok (itojun@)

  ~ grey.c                                

  > restrict address resolution to AF_INET for now (itojun@)

== regress =========================================================== 05/10 ==

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

lib

  ~ libc/getaddrinfo/answer               ~ libc/getaddrinfo/gaitest.c

  > make test successful even when /etc/services is changed (itojun@)

sys

  ~ arch/i386/ldt/Makefile                

  > always run userldt regress, but enable the sysctl if needed
  > ok millert@ (david@)

== sbin ============================================================== 06/10 ==

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

ifconfig

  ~ ifconfig.c                            

  > Make printing of 802.11 fields consistent with the rest by printing
  > a colon (':') after the field name.  Noticed by markus@, OK deraadt@
  > (millert@)

isakmpd

  ~ exchange.c                            ~ sa.h
  ~ transport.c                           ~ udp.c
  ~ udp_encap.c                           ~ virtual.c

  > Port floating (500->4500) for p1 and p2 exchanges. (ho@)

  ~ pf_key_v2.c                           

  > Tell the kernel to enable ESP-in-UDP encapsulation when we have
  > SAs negotiated with NAT-T. (ho@)

  ~ log.c                                 ~ message.c

  > Packet capture should add the ESP-marker when NAT-T is active. (ho@)

  ~ log.c                                 

  > Don't write too much IKE data in packet capture (ho@)

  ~ features/nat_traversal                

  > undo double-patch; Dries Schellekens (markus@)

  ~ ipsec.c                               

  > style nit (ho@)

  ~ monitor.c                             

  > Strip away umask bits in monitor_fopen(). hshoexer@ ok. (ho@)

  ~ virtual.c                             

  > When switching from main to encap transport, copy dst port if
  > translated (NAT). (ho@)

  ~ pf_key_v2.c                           

  > udpencap_port should be taken from dst transport (ho@)

  ~ ipsec.c                               ~ nat_traversal.c
  ~ nat_traversal.h                       ~ sa.c
  ~ sa.h                                  ~ udp_encap.c

  > Implement NAT-T keepalive messages. (ho@)

  ~ Makefile                              

  > Enable NAT-T support. (ho@)

  ~ isakmpd.conf.5                        

  > Describe the [Default]:NAT-T-Keepalive configuration parameter. (ho@)

pfctl

  ~ pfctl.8                               

  > Update manpage to reflect changes in anchor naming
  > ok beck@ claudio@ (msf@)

raidctl

  ~ raidctl.c                             

  > dont compare int with NULL, compare to 0 instead, otto@ ok (avsm@)

== share ============================================================= 07/10 ==

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

man

  ~ man4/man4.sparc64/Makefile            ~ man4/man4.sparc64/bwtwo.4
  ~ man4/man4.sparc64/cgsix.4             ~ man4/man4.sparc64/cgthree.4
  ~ man4/man4.sparc64/creator.4           ~ man4/man4.sparc64/intro.4
  ~ man4/man4.sparc64/mgx.4               ~ man4/man4.sparc64/rfx.4
  ~ man4/man4.sparc64/vgafb.4             ~ man4/man4.sparc64/zx.4
  + man4/man4.sparc64/agten.4             

  > agten(4) documentations updates. (miod@)

== sys =============================================================== 08/10 ==

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

arch/amd64/amd64

  ~ locore.S                              

  > Pure luck has protected us from this bug until now: locore.S
  > %r9 are not saved over function calls
  > and more we did not even want &proc0 as the old process  in switch_search,
  > but zero.  Fixes bsd.rd. (niklas@)

arch/i386/i386

  ~ db_interface.c                        ~ db_mp.c

  > Make 'machine ddbcpu' work, and use CPU device number instead of APIC id
  > KNF lesson and ok from niklas@ (andreas@)

arch/i386/include

  ~ db_machdep.h                          

  > Make 'machine ddbcpu' work, and use CPU device number instead of APIC id
  > KNF lesson and ok from niklas@ (andreas@)

  ~ cpu.h                                 

  > Put back the moving of schedstate_percpu into sched.h. This time expose
  > it to userland so that i386 builds (other architectures didn't show the
  > problem).
  > deraadt@ ok (art@)

arch/sparc64/conf

  ~ GENERIC                               

  > Add support for agten(4) - on Andrey's U2 the PROM is not able to
  > initialize
  > the board correctly, though. (miod@)

arch/sparc64/include

  ~ cpu.h                                 

  > Put back the moving of schedstate_percpu into sched.h. This time expose
  > it to userland so that i386 builds (other architectures didn't show the
  > problem).
  > deraadt@ ok (art@)

compat/common

  ~ kern_ipc_23.c                         

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

compat/hpux

  ~ hpux_compat.c                         

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

ddb

  ~ db_command.c                          

  > Make 'machine ddbcpu' work, and use CPU device number instead of APIC id
  > KNF lesson and ok from niklas@ (andreas@)

dev/ic

  ~ aic79xx.c                             ~ aic79xx_openbsd.c

  > Fix ahd_[get|set]_transaction_status vs ahd_[get|set]_scsi_status
  > confusion.
  > Allocate max number of SCB's at initialization.
  > Display negotiated transfer parameters.
  > ok marco@ (krw@)

  ~ ncr53c9x.c                            ~ smc90cx6.c
  ~ z8530tty.c                            

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

dev/pci

  ~ ncr.c                                 

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

dev/sbus

  ~ files.sbus                            + agten.c

  > Add support for agten(4) - on Andrey's U2 the PROM is not able to
  > initialize
  > the board correctly, though. (miod@)

kern

  ~ kern_clock.c                          ~ kern_synch.c
  ~ kern_time.c                           ~ sysv_msg.c
  ~ sysv_sem.c                            ~ sysv_shm.c
  ~ vfs_sync.c                            ~ vfs_syscalls.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

lib/libkern

  ~ funcs.h                               ~ libkern.h

  > add bounded attribute to some functions.  anil tested and ok'ed (itojun@)

net

  ~ if_arcsubr.c                          

  > Don't use time for random starting value... "love it" deraadt@ (tholo@)

  ~ if_bridge.c                           ~ pf.c
  ~ pfvar.h                               

  > Get rid of pf_test_eh() wrapper.
  > ok cedric@ henning@ (mcbride@)

  ~ pfkeyv2_convert.c                     

  > make it possble to use IPsec over link-local address (policy table uses
  > sin6_scope_id, IPsec porion uses embedded form).  beck ok (itojun@)

  ~ bpf.c                                 

  > move the IFF_UP check to bpfwrite; ok canacar@ (markus@)

  ~ pfkeyv2.c                             

  > don't accept SADB_X_EXT_UDPENCAP if encapsulation is disabled; ok ho@
  > (markus@)

  ~ if_arcsubr.c                          ~ if_bridge.c
  ~ if_ethersubr.c                        ~ if_fddisubr.c
  ~ if_pfsync.c                           ~ if_ppp.c
  ~ if_sl.c                               ~ if_strip.c
  ~ if_tokensubr.c                        ~ pf.c
  ~ pf_if.c                               ~ pf_ioctl.c
  ~ pf_norm.c                             ~ pf_table.c
  ~ route.c                               

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

netatalk

  ~ at_control.c                          

  > Use arc4random() instead of time-based node start id (tholo@)

netccitt

  ~ if_x25subr.c                          ~ pk_acct.c
  ~ pk_subr.c                             

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

netinet

  ~ ip_output.c                           

  > don't leak ipsec pmtu routes; with mpf@ (markus@)

  ~ ip_input.c                            ~ ip_output.c

  > Get rid of pf_test_eh() wrapper.
  > ok cedric@ henning@ (mcbride@)

  ~ ip_ipip.c                             ~ ip_spd.c
  ~ ipsec_input.c                         

  > make it possble to use IPsec over link-local address (policy table uses
  > sin6_scope_id, IPsec porion uses embedded form).  beck ok (itojun@)

  ~ ipsec_output.c                        

  > don't send UDP encapsulated packets w/o UDP header if encap is disabled; ok
  > ho@ (markus@)

  ~ in.c                                  

  > When in_ifinit fails and we've created a new address, clean it up before
  > returning.
  > ok markus@ (mcbride@)

  ~ if_ether.c                            ~ ip_id.c
  ~ ip_ipsp.c                             ~ ip_output.c
  ~ ip_spd.c                              ~ ipsec_input.c
  ~ ipsec_output.c                        ~ tcp_subr.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

netinet6

  ~ ip6_forward.c                         ~ ip6_input.c
  ~ ip6_output.c                          

  > Get rid of pf_test_eh() wrapper.
  > ok cedric@ henning@ (mcbride@)

  ~ icmp6.c                               ~ in6.c
  ~ in6.h                                 ~ ip6_forward.c
  ~ ip6_id.c                              ~ ip6_mroute.c
  ~ nd6.c                                 ~ nd6_rtr.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

netiso

  ~ if_eon.c                              ~ iso_snpac.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

nfs

  ~ nfs.h                                 ~ nfs_socket.c
  ~ nfs_subs.c                            ~ nfs_syscalls.c
  ~ nfs_vfsops.c                          ~ nfsm_subs.h

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

sys

  ~ sched.h                               ~ proc.h

  > Put back the moving of schedstate_percpu into sched.h. This time expose
  > it to userland so that i386 builds (other architectures didn't show the
  > problem).
  > deraadt@ ok (art@)

  ~ time.h                                

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

ufs/ext2fs

  ~ ext2fs_alloc.c                        ~ ext2fs_vfsops.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

ufs/ffs

  ~ ffs_alloc.c                           ~ ffs_vfsops.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

ufs/lfs

  ~ lfs_alloc.c                           ~ lfs_bio.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

ufs/ufs

  ~ inode.h                               ~ ufs_quota.c

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

uvm

  ~ uvm_meter.c                           

  > First step towards more sane time handling in the kernel -- this changes
  > things such that code that only need a second-resolution uptime or wall
  > time, and used to get that from time.tv_secs or mono_time.tv_secs now get
  > this from separate time_t globals time_second and time_uptime.
  > ok art@ niklas@ nordin@ (tholo@)

== usr.bin =========================================================== 09/10 ==

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

du

  ~ du.c                                  

  > Actually commit right (r&b tree) version, instead of hash table
  > one.  Previous commit from wrong tree. Spotted by Andre Lucas <andre
  > at ae-35 dot com> ok millert@ canacar@ (otto@)

pr

  ~ pr.c                                  

  > dont compare int with NULL, compare to 0 instead, otto@ ok (avsm@)

ssh

  ~ Makefile.inc                          ~ auth-rsa.c
  ~ auth2-gss.c                           ~ auth2-pubkey.c
  ~ authfile.c                            ~ canohost.c
  ~ channels.c                            ~ cipher.c
  ~ dns.c                                 ~ kex.c
  ~ monitor.c                             ~ monitor_fdpass.c
  ~ monitor_wrap.c                        ~ monitor_wrap.h
  ~ nchan.c                               ~ packet.c
  ~ progressmeter.c                       ~ scp.c
  ~ sftp-server.c                         ~ sftp.c
  ~ ssh-gss.h                             ~ ssh-keygen.c
  ~ ssh.c                                 ~ sshconnect.c
  ~ sshconnect1.c                         ~ sshlogin.c
  ~ sshpty.c                              

  > make ssh -Wshadow clean, no functional changes
  > markus@ ok (avsm@)

  ~ session.c                             

  > fix fd leak for multiple subsystem connections; with markus@ (djm@)

  ~ Makefile.inc                          

  > comment out -Wshadow for now, gcc2 sucks. noticed by markus (avsm@)

  ~ log.h                                 

  > mark fatal and cleanup exit as __dead; ok markus@ (djm@)

  ~ sftp.c                                

  > introduce sorting for ls, same options as /bin/ls; ok markus@ (djm@)

  ~ sftp.c                                

  > prefix ls option flags with LS_ (djm@)

  ~ sftp.1                                

  > document sort options (djm@)

xstr

  ~ xstr.c                                

  > dont compare int with NULL, compare to 0 instead, otto@ ok (avsm@)

== usr.sbin ========================================================== 10/10 ==

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

pkg_add

  ~ pkg_add                               

  > Implement pkg_add -I for real; ok espie@ (miod@)

  ~ Makefile                              ~ pkg_delete
  ~ pkg_info                              ~ OpenBSD/PackingElement.pm
  + OpenBSD/PackingOld.pm                 

  > support for old stuff in pkg_info/pkg_delete, e.g., @src
  > prompted by miod@, okay miod@ millert at .spie@)

tokenadm

  ~ tokenadm.c                            

  > dont compare int with NULL, compare to 0 instead, otto@ ok (avsm@)

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



More information about the odc mailing list