OpenBSD src changes summary for 2015-10-26
==========================================

bin/csh                                 distrib/sets
etc/netstart                            lib/libc
regress/sys                             regress/usr.bin
sbin/dhclient                           sbin/fdisk
sbin/ifconfig                           share/man
share/mk                                sys/crypto
sys/kern                                sys/lib/libsa
sys/netinet                             sys/sys
usr.bin/diff                            usr.bin/finger
usr.bin/kdump                           usr.bin/login
usr.bin/pkg-config                      usr.bin/sed
usr.bin/sendbug                         usr.bin/telnet
usr.bin/tmux                            usr.bin/top
usr.bin/usbhidaction                    usr.sbin
usr.sbin/cron                           usr.sbin/httpd
usr.sbin/mrouted                        usr.sbin/ospfd
usr.sbin/pppd                           usr.sbin/radiusd
usr.sbin/rarpd                          usr.sbin/rbootd
usr.sbin/rebound                        usr.sbin/repquota
usr.sbin/rip6query                      usr.sbin/route6d
usr.sbin/smtpd                          usr.sbin/wsmoused
usr.sbin/ypserv                         

== bin =============================================================== 01/10 ==

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

csh

  ~ Makefile                              ~ char.c
  ~ char.h                                ~ csh.c
  ~ csh.h                                 ~ exp.c
  ~ extern.h                              ~ file.c
  ~ func.c                                ~ glob.c
  ~ misc.c                                ~ set.c
  ~ str.c                                 

  > unifdef BUILTIN (unused), NLS (always set), SHORT_STRINGS (always set)
  > and remove a few unused defines; no binary changes; ok tedu@ zhuk@ (naddy@)

  ~ char.c                                ~ const.c
  ~ csh.c                                 ~ csh.h
  ~ func.c                                ~ str.c

  > Clean up cruft made visible by the NLS unifdefing:
  > * remove setlocale() calls
  > * remove write-only variable AsciiOnly
  > * remove now unused string constants STRLANG, STRLC_CTYPE
  > * remove hardcoded support for ISO8859-1
  > ok millert@ (naddy@)

  ~ csh.h                                 ~ extern.h
  ~ str.c                                 

  > Remove the #define Strfoo() s_strfoo() redirection and move everything
  > directly to Strfoo().  Purely mechanical.  ok deraadt@ (naddy@)

  ~ csh.1                                 

  > document the -p and -l file enquiries and explicitly do not document -s
  > with/ok schwarze@ (naddy@)

  ~ Makefile                              ~ csh.c
  ~ csh.h                                 ~ exp.c
  ~ extern.h                              ~ proc.c

  > remove the profiling and debugging ifdefs; ok zhuk@ (naddy@)

  ~ csh.c                                 ~ exp.c
  ~ glob.c                                ~ lex.c
  ~ misc.c                                ~ proc.c
  ~ sem.c                                 ~ set.c

  > miscellaneous cruft:
  > * remove lint
  > * unifdef S_IFLNK, S_ISFIFO, S_ISLNK, S_ISSOCK, O_APPEND
  > * remove 0, NOTUSED, notdef -- untouched since original import
  > * remove IIASA (naddy@)

== distrib =========================================================== 02/10 ==

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

sets

  ~ lists/base/mi                         ~ lists/man/mi

  > sync (deraadt@)

== etc =============================================================== 03/10 ==

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

netstart

  ~ netstart                              

  > The hostname variable is not used since r1.99. Remove it and use
  > stripcom() output directly with the hostname command.
  > OK deraadt@ krw@ (rpe@)

== lib =============================================================== 04/10 ==

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

libc

  ~ gen/pwcache.c                         

  > don't keep the passwd database open; callers do not expect that they will
  > become responsible for calling endpwent.
  > ok deraadt (tedu@)

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

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

sys

  ~ kern/Makefile                         

  > temporary unhook regress/sys/kern/pledge from regress (semarie@)

  ~ net/Makefile                          + net/rdomains/Makefile
  + net/rdomains/ifalocal.sh              

  > Add regress test on local route insertions (vgross@)

  ~ netinet6/Makefile                     

  > Add autoport to netinet6 regress tests (vgross@)

  ~ kern/pledge/generic/Makefile          ~ kern/pledge/generic/main.c
  ~ kern/pledge/generic/test_tty.c        ~ kern/pledge/generic/tests.out
  + kern/pledge/generic/test_stdio.c      

  > update regress suite for pledge (semarie@)

usr.bin

  ~ ssh/limit-keytype.sh                  

  > Fix typo certopt->certopts in shell variable.  This would cause the test to
  > hang at a host key prompt if you have an A or CNAME for "proxy" in your
  > local
  > domain. (dtucker@)

  ~ pkg-config/Makefile                   
  + pkg-config/pcdir/whitespace-prop-1.pc
  + pkg-config/pcdir/whitespace-prop-2.pc
  + pkg-config/pcdir/whitespace-prop-3.pc

  > add tests for whitespace surrounding properties (jasper@)

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

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

dhclient

  ~ clparse.c                             ~ dhclient.c
  ~ dhcp.h                                ~ dhcpd.h
  ~ options.c                             ~ tables.c

  > Give dhclient(8) the ability to use option 119, a.k.a. "Domain
  > Search" if supplied by the server.
  > Requested by a few. Original diff from Ray Lai via tech@.
  > Tested & ok claudio@ (krw@)

fdisk

  ~ fdisk.8                               

  > Flesh out GPT info and tweak other verbiage to be less MBR specific. (krw@)

  ~ cmd.c                                 ~ cmd.h
  ~ fdisk.c                               ~ gpt.c
  ~ gpt.h                                 ~ mbr.c
  ~ mbr.h                                 ~ misc.c
  ~ misc.h                                ~ part.c
  ~ part.h                                ~ user.c

  > Add GPT editing. Based on GSOC 2014 work by Markus Muller. (krw@)

  ~ fdisk.8                               

  > tweak previous; (jmc@)

ifconfig

  ~ ifconfig.8                            

  > Sync "ifconfig create" interface list with ifconfig -C output.
  > Adds pair, svlan, tap and vxlan. (jsg@)

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

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

man

  ~ man5/pf.conf.5                        

  > adjust macro usage to the usual conventions (schwarze@)

  ~ man5/mk.conf.5                        

  > Nothing in base uses native language support anymore.  Remove NLS
  > support from the makefiles.
  > OK naddy@ (bluhm@)

mk

  - bsd.nls.mk                            ~ Makefile
  ~ bsd.README                            ~ bsd.lib.mk
  ~ bsd.own.mk                            ~ bsd.prog.mk

  > Nothing in base uses native language support anymore.  Remove NLS
  > support from the makefiles.
  > OK naddy@ (bluhm@)

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

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

crypto

  ~ cryptosoft.c                          

  > Use axf's hashsize as a block size in the authenticated encryption routine.
  > No change for GCM, however upcoming changes will rely on this. (mikeb@)

kern

  ~ kern_pledge.c                         

  > make pledge_check(), used for syscall check with pledge, returns an error
  > and
  > provide the required pledge request for pledge_fail().
  > ok deraadt@ (semarie@)

  ~ kern_pledge.c                         

  > change some pledge_fail() error/code
  > - for PLEDGE_FATTR: the code is more accurate to be PLEDGE_FATTR, has
  > p->p_pledgenote could have multiple bits inside, and it is the lake of
  > PLEDGE_FATTR that make the check fail.
  > - for PLEDGE_RECVFD and PLEDGE_SENDFD, change the error to be EINVAL: we
  > fail,
  > not because of lake of PLEDGE_RECVFD / PLEDGE_SENDFD permission, but
  > because
  > the value passed is invalid in pledged program (for example trying to send
  > a
  > VDIR).
  > "go ahead" deraadt@ (semarie@)

  ~ kern_pledge.c                         

  > add setreuid/setregid to "id" (deraadt@)

  ~ kern_pledge.c                         

  > If the system call is entirely unpermitted, code will be 0, and there is
  > no pledge to recommend. (deraadt@)

  ~ kern_pledge.c                         

  > Allow NET_RT_IFLIST in pledge "dns" as well (deraadt@)

  ~ uipc_syscalls.c                       

  > dns check needs to be done on the kernel address after copyin (tedu@)

  ~ kern_pledge.c                         

  > Add ppoll() to "stdio" (deraadt@)

lib/libsa

  ~ net.c                                 

  > Cast isdigit()'s argument to unsigned char.
  > ok guenther@ (mmcc@)

  ~ net.c                                 

  > (char *)0 -> NULL
  > ok tedu@ (mmcc@)

netinet

  ~ ip_input.c                            

  > Use rt_ifidx rather than rt_ifp.
  > ok bluhm@ (mpi@)

sys

  ~ pledge.h                              ~ syscall_mi.h

  > make pledge_check(), used for syscall check with pledge, returns an error
  > and
  > provide the required pledge request for pledge_fail().
  > ok deraadt@ (semarie@)

  ~ queue.h                               

  > Let SLIST_REMOVE invalidate the pointer in more cases.
  > OK jsing@ (bluhm@)

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

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

diff

  ~ diff.1                                ~ diff.c
  ~ diffreg.c                             

  > remove TMPDIR support. ok millert (tedu@)

  ~ diff.1                                

  > no more TMPDIR; (jmc@)

finger

  ~ finger.c                              

  > Rather than setpassent(1) for the lifetime of the program, use two smaller
  > windows of setpassent(1) + entpwent()
  > ok tedu millert (deraadt@)

kdump

  ~ kdump.c                               

  > If the system call is entirely unpermitted, code will be 0, and there is
  > no pledge to recommend. (deraadt@)

login

  ~ login.c                               

  > Put off some "notyet" dust from login.c.
  > okay millert@ and deraadt@ (zhuk@)

pkg-config

  ~ OpenBSD/PkgConfig.pm                  

  > Don't reject properties with no whitespace after the colon; it's perfectly
  > valid to have a line like 'Requires:requiree'.
  > Confirmed with fd.o and this fixes using the libczmq pc file (jasper@)

sed

  ~ compile.c                             ~ extern.h
  ~ main.c                                ~ misc.c
  ~ process.c                             

  > Rename err() to error() to prevent confusion with the stdlib function.
  > Discussed with jca@. (mmcc@)

  ~ main.c                                

  > Use error(FATAL) instead of what was previously err(1,...)
  > ok tobias@ (jca@)

  ~ compile.c                             ~ defs.h

  > Tidy up error() levels: merge COMPILE and COMPILE2, and kill ERROR (unused)
  > ok tobias@ (jca@)

sendbug

  ~ sendbug.c                             

  > remove TMPDIR support (deraadt@)

telnet

  ~ commands.c                            

  > Set low-delay traffic class for IPv6 connections as well
  > While here, there is no option to pass a different tos, and no other
  > use of the "tos" variable out of tn(), so move the "tos" variable in tn()
  > and assign it the right value from the start.
  > ok millert@ (jca@)

tmux

  ~ tmux.1                                

  > space before punctuation; (jmc@)

  ~ input-keys.c                          ~ server-client.c

  > Some extra logging of where keys are actually going. (nicm@)

  ~ cmd-list-keys.c                       ~ key-string.c

  > Handle unknown keys more gracefully, return a string instead of NULL.
  > (nicm@)

  ~ key-bindings.c                        

  > Default bindings for mouse wheel on status line to change window (like
  > we had before), from Patrick Palka. (nicm@)

  ~ server-client.c                       

  > If a mouse event has no key binding, pass it through to the pane it
  > happened in, not the active pane like normal key presses. Fixes problems
  > seen by Enrico Ghirardi. (nicm@)

top

  ~ display.c                             

  > print uptime too. ok deraadt (tedu@)

  ~ username.c                            

  > remove the silly username cache here and use libc user_from_uid.
  > libc isn't (yet) any better, but it can improve with time. (tedu@)

usbhidaction

  ~ usbhidaction.c                        

  > stop using pidfile()
  > ok benno jung jca (deraadt@)

  ~ Makefile                              ~ usbhidaction.c

  > remove util.h include and -lutil after pidfile removal
  > ok deraadt (jung@)

  ~ Makefile                              

  > remove libutil from dpadd too (tedu@)

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

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

usr.sbin

  ~ Makefile                              

  > rip6query goes to the attic.  Hopefully route6d (rip6d?) will get some
  > love instead, prod prod.
  > ok claudio jca mpi (deraadt@)

cron

  ~ cron.c                                ~ crontab.c
  ~ database.c                            ~ entry.c
  ~ macros.h                              ~ misc.c

  > Remove some useless defines and the set_cron_uid() function.
  > OK jung@ (millert@)

  ~ cron.c                                ~ misc.c

  > Use SOCK_NONBLOCK and SOCK_CLOEXEC instead of fcntl() calls.
  > OK guenther@ (millert@)

  ~ crontab.1                             ~ crontab.c

  > Remove TMPDIR support, it is not really useful in crontab.
  > OK deraadt@ (millert@)

httpd

  ~ httpd.c                               

  > fix PATH_INFO for / requests
  > diff from Denis Fondras
  > ok reyk (jung@)

mrouted

  ~ main.c                                

  > stop using pidfile()
  > ok benno jung jca (deraadt@)

  ~ Makefile                              ~ main.c

  > remove util.h include and -lutil after pidfile removal
  > ok deraadt (jung@)

  ~ Makefile                              

  > remove libutil from dpadd too (tedu@)

ospfd

  ~ kroute.c                              

  > Kernel got fixed we no longer need to check if rtm_pid is 0 to filter out
  > route messages that are not from us. (claudio@)

pppd

  ~ main.c                                

  > stop saving a pidfile; ok jung zhuk (deraadt@)

radiusd

  ~ radiusd.conf.5                        

  > Using realm for bsdauth is misleading.  It is currently unusable.
  > (yasuoka@)

rarpd

  ~ rarpd.c                               

  > stop using pidfile()
  > ok benno jung jca (deraadt@)

  ~ Makefile                              ~ rarpd.c

  > remove util.h include and -lutil after pidfile removal
  > ok deraadt (jung@)

  ~ Makefile                              

  > remove libutil from dpadd too (tedu@)

rbootd

  ~ rbootd.c                              

  > stop using pidfile()
  > ok benno jung jca (deraadt@)

  ~ Makefile                              ~ rbootd.c

  > remove util.h include and -lutil after pidfile removal
  > ok deraadt (jung@)

  ~ Makefile                              

  > remove libutil from dpadd too (tedu@)

rebound

  ~ rebound.c                             

  > set listening sockets to nonblocking, then handle as many incoming requests
  > as we can in the loop instead of keventing per request. (tedu@)

  ~ rebound.c                             

  > one second amnesty for timeouts so we don't spin with short timeouts
  > (tedu@)

repquota

  ~ repquota.8                            ~ repquota.c

  > Put repquota in line with other utilities: you either wants list everyting,
  > effectively using -a, or you specify filesystems one-by-one. Doing both
  > resulted in a funny error messages, so just don't allow that silly
  > behaviour.
  > okay millert@ (zhuk@)

rip6query

  - Makefile                              - rip6query.8
  - rip6query.c                           

  > rip6query goes to the attic.  Hopefully route6d (rip6d?) will get some
  > love instead, prod prod.
  > ok claudio jca mpi (deraadt@)

route6d

  ~ route6d.c                             

  > Kill unused variable and function.
  > No functional change. (jca@)

  ~ route6d.c                             

  > Those variables should be local to the functions where they are used.
  > Gets rid of a lot of warnings with -Wshadow.  No functional change. (jca@)

  ~ route6d.c                             

  > Rename the struct pollfd array, "set" -> "pfd"
  > "set" reminds of select(2) fd_set, and is never used elsewhere in the
  > tree as a name for a pollfd array.  No functional change. (jca@)

  ~ route6d.c                             

  > Partial revert of r1.79, "np" handling is a bit trickier than expected.
  > (jca@)

smtpd

  ~ mda.c                                 

  > avoid a potential double free
  > ok gilles@ (jsg@)

  ~ smtpd.c                               

  > remove unneeded pidfile() from smtpd
  > ok gilles (jung@)

  ~ makemap.c                             

  > Fix makemap(8) for values containing a '#'.
  > Ok millert@ gilles@ (sunil@)

wsmoused

  ~ wsmoused.8                            ~ wsmoused.c
  ~ wsmoused.h                            

  > Stop writing a pidfile
  > ok benno jung (deraadt@)

ypserv

  ~ ypserv/Makefile                       ~ ypserv/ypserv.c

  > last consumer in base of pidfile()... (deraadt@)

===============================================================================
_______________________________________________
odc mailing list
[email protected]
http://www.squish.net/mailman/listinfo/odc

Reply via email to