[odc] Daily src changes for 2005-12-30

ODC auto at squish.net
Sat Dec 31 07:00:33 GMT 2005


OpenBSD src changes summary for 2005-12-30
==========================================

Makefile.cross                          lib/libc
lib/librthread                          sbin/umount
share/mk                                sys/arch/alpha/stand
sys/arch/amd64/stand                    sys/arch/hp300/dev
sys/arch/hp300/hp300                    sys/arch/hp300/include
sys/arch/hp300/stand                    sys/arch/hppa/stand
sys/arch/hppa64/stand                   sys/arch/i386/conf
sys/arch/i386/stand                     sys/arch/macppc/stand
sys/arch/mvme68k/stand                  sys/arch/mvme88k/stand
sys/arch/mvmeppc/stand                  sys/arch/sgi/stand
sys/arch/sparc/sparc                    sys/arch/sparc/stand
sys/arch/sparc/stand/libsa              sys/arch/sparc64/include
sys/arch/sparc64/sparc64                sys/arch/sparc64/stand
sys/compat/hpux                         sys/compat/linux
sys/compat/sunos                        sys/dev
sys/dev/acpi                            sys/dev/cardbus
sys/dev/i2c                             sys/dev/ic
sys/dev/pci                             sys/kern
sys/sys                                 usr.bin/cvs
usr.bin/ssh                             usr.bin/units
usr.sbin/bgpctl                         usr.sbin/bgpd

== Makefile.cross ==================================================== 01/07 ==

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

Makefile.cross


  > Disable lint for libraries.
  > ok deraadt@ espie@ (grange@)

== lib =============================================================== 02/07 ==

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

libc

  ~ net/ethers.3                          ~ net/if_indextoname.3
  ~ net/inet_net.3                        

  > Currently sys/socket.h needs sys/types.h so add #include <sys/types.h> to
  > the list of necessary includes. Now the manpages match reality.
  > OK mpf@ millert@ (claudio@)

librthread

  ~ rthread.c                             ~ rthread.h
  ~ rthread_sched.c                       ~ rthread_sig.c
  ~ rthread_sync.c                        

  > prototype all the thread syscalls in rthread.h for now.
  > update for new thrwakeup that takes a count argument (tedu@)

  ~ rthread.c                             

  > __aligned__ isn't really necessary on the struct, an aligned member
  > will make the right thing happen.  verified by kettenis and drahn (tedu@)

  ~ rthread.c                             ~ rthread.h
  ~ rthread_np.c                          

  > use queue.h macros for thread house keeping; make some vars non-static
  > and fix a bug in thread creation error path. ok tedu@ (otto@)

  ~ rthread.c                             

  > Remove exitinng thread from the list of threads. ok tedu@ (otto@)

== sbin ============================================================== 03/07 ==

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

umount

  ~ umount.8                              

  > better wording, based on a diff from daniel matic; (jmc@)

== share ============================================================= 04/07 ==

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

mk

  ~ bsd.lib.mk                            

  > Allow to disable lint by setting WANTLIB=no
  > ok deraadt@ espie@ (grange@)

== sys =============================================================== 05/07 ==

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

arch/alpha/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/amd64/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/hp300/dev

  ~ dvbox.c                               ~ gbox.c
  ~ rbox.c                                ~ topcat.c

  > Introduce symbolic constants for the magic conscode values.
  > No functional change. (miod@)

  ~ apci.c                                ~ dca.c
  ~ dcm.c                                 ~ dvbox.c
  ~ gbox.c                                ~ hyper.c
  ~ rbox.c                                ~ sti_sgc.c
  ~ topcat.c                              

  > When a driver claims console, let it set conscode to itself, instead of
  > letting the bus walkers do this for it. This is necessary since apci does
  > not use the bus walkers, now that we do not force console probes to stop
  > when apci claims console. (miod@)

  ~ apci.c                                ~ dca.c
  ~ dcm.c                                 ~ dvbox.c
  ~ gbox.c                                ~ hyper.c
  ~ rbox.c                                ~ sti_sgc.c
  ~ topcat.c                              

  > Kill conforced, and have console devices use CN_FORCED priority instead.
  > This means we are no longer aborting the console scans even if the final
  > console has been found, but on the other hand makes some logic simpler;
  > plus it will be necessary for upcoming scode vs device loop order reversal.
  > No user-visible change. (miod@)

arch/hp300/hp300

  ~ machdep.c                             

  > Introduce symbolic constants for the magic conscode values.
  > No functional change. (miod@)

  ~ autoconf.c                            

  > When a driver claims console, let it set conscode to itself, instead of
  > letting the bus walkers do this for it. This is necessary since apci does
  > not use the bus walkers, now that we do not force console probes to stop
  > when apci claims console. (miod@)

  ~ autoconf.c                            ~ machdep.c

  > Kill conforced, and have console devices use CN_FORCED priority instead.
  > This means we are no longer aborting the console scans even if the final
  > console has been found, but on the other hand makes some logic simpler;
  > plus it will be necessary for upcoming scode vs device loop order reversal.
  > No user-visible change. (miod@)

arch/hp300/include

  ~ autoconf.h                            

  > Introduce symbolic constants for the magic conscode values.
  > No functional change. (miod@)

  ~ autoconf.h                            

  > Kill conforced, and have console devices use CN_FORCED priority instead.
  > This means we are no longer aborting the console scans even if the final
  > console has been found, but on the other hand makes some logic simpler;
  > plus it will be necessary for upcoming scode vs device loop order reversal.
  > No user-visible change. (miod@)

arch/hp300/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/hppa/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/hppa64/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/i386/conf

  ~ GENERIC                               

  > add asbtm iic sensor; ok kettenis@ (djm@)

arch/i386/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/macppc/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/mvme68k/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/mvme88k/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/mvmeppc/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/sgi/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/sparc/sparc

  ~ process_machdep.c                     

  > Make PT_SETFPREGS work even if the process didn't use the fpu yet.
  > ok miod@ (kettenis@)

arch/sparc/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/sparc/stand/libsa

  ~ Makefile                              

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

arch/sparc64/include

  ~ types.h                               

  > Remove unused COMPAT_NETBSD32 block and get rid of register{32,64}_t
  > which are now unused.  Tested and OK miod@ (millert@)

arch/sparc64/sparc64

  ~ trap.c                                

  > Remove unused COMPAT_NETBSD32 block and get rid of register{32,64}_t
  > which are now unused.  Tested and OK miod@ (millert@)

arch/sparc64/stand

  ~ Makefile.inc                          

  > Now that errno.h and sys/errno.h are separate header files there is no
  > need to use the -Derrno=errno workaround added when __errno() was
  > introduced.  OK deraadt@ (millert@)

compat/hpux

  ~ hppa/hpux_exec.c                      ~ m68k/hpux_exec.c

  > Missing or incorrect header sizes bounds check; ``looks ok'' mickey@
  > (miod@)

compat/linux

  ~ linux_exec.c                          

  > Missing or incorrect header sizes bounds check; ``looks ok'' mickey@
  > (miod@)

compat/sunos

  ~ sunos_exec.c                          

  > Missing or incorrect header sizes bounds check; ``looks ok'' mickey@
  > (miod@)

dev

  ~ cons.h                                

  > Introduce another console device priority level, CN_FORCED, which wins over
  > CN_REMOTE. (miod@)

  ~ cons.c                                

  > Stop TIOCCONS from silently doing nothing if the console device has
  > CN_REMOTE
  > priority; this might have been bright 12 years ago, but this is wrong (and
  > does not rely upon any decent semantic). agreed krw@ (miod@)

dev/acpi

  ~ acpibat.c                             ~ acpivar.h
  ~ dsdt.c                                

  > make ACPI_DEBUG compile (tedu@)

dev/cardbus

  ~ ehci_cardbus.c                        

  > while looking for device_ptr_t typedef cruft i noticed #if __NetBSD__. why
  > did i put that in? this cleans up the if __NetBSD__ goo (dlg@)

  ~ ohci_cardbus.c                        ~ uhci_cardbus.c

  > remove #if __NetBSD__ and device_ptr_t. its more readable and less
  > confusing now. (dlg@)

  ~ ehci_cardbus.c                        

  > remove the tailq used to locate ehci's companion controllers. i have
  > absolutely no idea how this compiled since the usb_cardbus struct (which is
  > what the list was made of) was defined in usb_cardbus.h. usb_cardbus.h was
  > removed from the tree over 8 months ago. gcc must be crap. (dlg@)

dev/i2c

  ~ i2c_scan.c                            

  > check for another sign that a lm75 is not there (deraadt@)

  ~ i2c_scan.c                            ~ lm_i2c.c

  > Initial W83791D support for lm(4).
  > ok krw@ (kettenis@)

  ~ i2c_scan.c                            

  > knf (replace lines of c++/c99 '//' comments with #if 0) (reyk@)

  ~ files.i2c                             ~ i2c_scan.c
  + asb100.c                              

  > add i2c driver for Asus ASB100 sensors using register info from lm_sensors
  > driver (Asus believe this chip is such critically important IP they won't
  > release docs for it); feedback deraadt@ and kettenis@, ok kettenis@ (djm@)

dev/ic

  ~ nslm7x.c                              ~ nslm7xvar.h

  > Initial W83791D support for lm(4).
  > ok krw@ (kettenis@)

  ~ nslm7x.c                              ~ nslm7xvar.h

  > Properly detect AS99127F rev 1.
  > From Jonathan Fromer <jf at daimi.au.dk(kettenis@)

dev/pci

  ~ viapm.c                               

  > Some spaces->tab indentation fixes. (jsg@)

  ~ ehci_pci.c                            ~ ohci_pci.c
  ~ uhci_pci.c                            

  > device_ptr_t -> struct device *
  > i hate typedefs (dlg@)

  ~ pcidevs                               

  > typo (jolan@)

  ~ pcidevs.h                             

  > regen (jolan@)

kern

  ~ kern_descrip.c                        ~ kern_exec.c

  TAGGED OPENBSD_3_7
  > MFC:
  > Fix by deraadt@
  > do not allow setugid processes to use /dev/fd/#, unless they are a
  > setuid-script and are attempting to dup is the specific setuid-script
  > fd via such a pathname;
  > ok deraadt@ (brad@)

  ~ init_sysent.c                         ~ kern_synch.c
  ~ syscalls.c                            ~ syscalls.master

  TAGGED OPENBSD_3_7
  > change thrwakeup to take an argument which specifies how many threads
  > to wakeup. (tedu@)

  ~ init_sysent.c                         ~ syscalls.c

  TAGGED OPENBSD_3_7
  > rebuild (tedu@)

sys

  ~ proc.h                                

  TAGGED OPENBSD_3_7
  > MFC:
  > Fix by deraadt@
  > do not allow setugid processes to use /dev/fd/#, unless they are a
  > setuid-script and are attempting to dup is the specific setuid-script
  > fd via such a pathname;
  > ok deraadt@ (brad@)

  ~ syscall.h                             ~ syscallargs.h

  TAGGED OPENBSD_3_7
  > change thrwakeup to take an argument which specifies how many threads
  > to wakeup. (tedu@)

  ~ syscall.h                             ~ syscallargs.h

  TAGGED OPENBSD_3_7
  > rebuild (tedu@)

== usr.bin =========================================================== 06/07 ==

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

cvs

  ~ fatal.c                               

  > pass LP_ABORT to cvs_vlog() instead of LP_ERR. (joris@)

  ~ buf.c                                 

  > spacing (joris@)

  ~ add.c                                 ~ admin.c
  ~ annotate.c                            ~ checkout.c
  ~ cmd.c                                 ~ commit.c
  ~ cvs.h                                 ~ diff.c
  ~ edit.c                                ~ getlog.c
  ~ history.c                             ~ import.c
  ~ logmsg.c                              ~ proto.c
  ~ proto.h                               ~ release.c
  ~ remove.c                              ~ status.c
  ~ tag.c                                 ~ update.c
  ~ watch.c                               

  > major cleanup of the functions handling the remote cvs protocol.
  > makes the code a lot more readable and understandable.
  > ok xsa@ and niallo@ (joris@)

  ~ buf.c                                 

  > adjust some thruth in a comment. (joris@)

  ~ annotate.c                            ~ getlog.c
  ~ release.c                             ~ update.c

  > minor style nits; (xsa@)

  ~ rcs.c                                 

  > - realloc() -> xrealloc(), was missed in the original sweep. (niallo@)

  ~ cvs.h                                 ~ proto.c
  ~ proto.h                               ~ req.c
  ~ resp.c                                ~ util.c

  > more code cleanup, ok niallo@ and xsa@ (joris@)

  ~ rcs.c                                 

  > - simplify an unecessary if/else construct in rcs_open()
  > spotted by joris@ (niallo@)

  ~ rcs.h                                 ~ rcs.c

  > - zap unused 'rf_ref' member of RCSFILE structure.
  > spotted by joris@ (niallo@)

  ~ checkout.c                            ~ cmd.c
  ~ diff3.c                               ~ proto.c
  ~ rcs.c                                 ~ rcsnum.c
  ~ xmalloc.c                             

  > knf (reyk@)

ssh

  ~ channels.c                            ~ channels.h
  ~ clientloop.c                          

  > add channel output filter interface.
  > ok djm@, suggested by markus@ (reyk@)

  ~ sftp.1                                

  > do not suggest that interactive authentication will work
  > with the -b flag;
  > based on a diff from john l. scarfone;
  > ok djm (jmc@)

units

  ~ units.1                               

  > consistency tweak, from daniel matic; (jmc@)

  ~ units.1                               

  > do not use .Pp before displays; (jmc@)

== usr.sbin ========================================================== 07/07 ==

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

bgpctl

  ~ parser.c                              

  > Fix parse_prefix() for IPv4 addresses. Looks like fallout after IPv6
  > addition.
  > Now bgpctl show rib 129.128.5.191 works again. (claudio@)

bgpd

  ~ rde.c                                 

  > Merge rde_attr_error() into rde_attr_parse() it is no longer necessary to
  > have a separate error function. It is no porblem to call rde_update_err()
  > directly. (claudio@)

  ~ rde.h                                 ~ rde_attr.c
  ~ rde_rib.c                             ~ rde_update.c

  > Use sys/hash.h instead of own built functions that work similar.
  > While there reorder some structs to help with alignment. (claudio@)

  ~ rde_attr.c                            ~ rde.h

  > Remove unused attr_optlen() (claudio@)

  ~ rde.h                                 

  > There is no attr_mp_nexthop() function, remove the prototype. (claudio@)

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



More information about the odc mailing list