[odc] Daily src changes for 2006-03-10

ODC auto at squish.net
Sat Mar 11 07:00:45 GMT 2006


OpenBSD src changes summary for 2006-03-10
==========================================

bin/ps                                  etc/hostapd.conf
lib/libc                                regress/lib
regress/usr.bin                         share/man
sys/arch/amd64/amd64                    sys/arch/i386/i386
sys/dev/adb                             sys/dev/isa
sys/dev/mii                             sys/dev/onewire
sys/dev/pci                             sys/dev/pcmcia
sys/dev/rasops                          usr.bin/cap_mkdb
usr.bin/column                          usr.bin/cvs
usr.bin/mg                              usr.bin/rcs
usr.sbin/config                         usr.sbin/hostapd
usr.sbin/tcpdump                        

== bin =============================================================== 01/08 ==

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

ps

  ~ print.c                               

  > 1.36 and onwards did not print extra spaces, but sometimes skipped spaces.
  > we need a proper flag to keep track of space insert requirements
  > thanks for further testing by millert (deraadt@)

  ~ print.c                               

  > add missing break; found by lint
  > ok deraadt@ (moritz@)

== etc =============================================================== 02/08 ==

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

hostapd.conf

  ~ hostapd.conf                          

  > some IAPP implementations still use the pre-standard port 2313 in broadcast
  > mode (tested with the HP ProCurve 420wl). add a note to the manual page and
  > the example configuration. (reyk@)

== lib =============================================================== 03/08 ==

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

libc

  ~ gen/getcap.c                          

  > Don't dereference NULL pointer if file is truncated.
  > Sync with cap_mkdb(1).
  > ok otto and millert (ray@)

== regress =========================================================== 04/08 ==

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

lib

  + libc/getcap/Makefile                  + libc/getcap/getcaptest.c
  + libc/getcap/shortcap.in               

  > Regression test for truncated file.
  > ok otto and millert (ray@)

  ~ libc/Makefile                         

  > Hook up getcap test. (ray@)

usr.bin

  + cap_mkdb/Makefile                     + cap_mkdb/shortname.in

  > Regression test for truncated file.
  > ok otto and millert (ray@)

  ~ Makefile                              

  > Hook up cap_mkdb test. (ray@)

== share ============================================================= 05/08 ==

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

man

  ~ man8/vpn.8                            

  > remove default Transport=udp and DOI=IPSEC to simplify example isakmpd
  > config
  > ok jmc@ hshoexer@ (david@)

== sys =============================================================== 06/08 ==

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

arch/amd64/amd64

  ~ ipi.c                                 ~ lapic.c

  > provide a counter for ipis (mickey@)

  ~ ipi.c                                 

  > oops; from otto (mickey@)

  ~ ipi.c                                 ~ lapic.c

  > - Wrap IPI sending in splclock(), since an interrupt at
  > IPL_CLOCK or lower may cause IPIs.
  > - Make broadcast IPIs go through x86_ipi() as well, so that
  > they wait for the APIC to be ready too.
  > - Introduce a function, i82489_icr_wait, which waits for
  > LAPIC_DLSTAT_BUSY cleared, and use it where appropriate.
  > - Panic if lapic's busy too long and DIAGNOSTIC.
  > From NetBSD
  > Tested on a few amd64 MP systems. (brad@)

arch/i386/i386

  ~ ipifuncs.c                            ~ lapic.c

  > provide a counter for ipis (mickey@)

dev/adb

  ~ keyboard.h                            

  > Switch to RAWKEY_xxx constants. (miod@)

dev/isa

  ~ if_an_isapnp.c                        

  > ansi. no binary change. (jsg@)

  ~ mcd.c                                 

  > Silence mcd during probe, as in 1.13 but accidentally lost in 1.24. (miod@)

dev/mii

  ~ ciphy.c                               

  > Only do the forcing rgmii for nfe game after a reset not
  > after every link state change in the fixup function.
  > Suggested by brad@, tested by otto@ (jsg@)

  ~ ciphy.c                               

  > Last revision apparently does not work when cold booting so revert.
  > Noticed by otto@ (jsg@)

dev/onewire

  ~ devlist2h.awk                         

  > Oops. (grange@)

  ~ onewiredevs_data.h                    

  > sync (grange@)

dev/pci

  ~ if_an_pci.c                           

  > ansi. no binary change. (jsg@)

  ~ if_bge.c                              ~ if_bgereg.h

  > in bge_intr() also check the PCI State register to see if there
  > are really any interrupts to service.
  > Based on the Linux tg3 driver though the Broadcom bcm5700 driver
  > does the same thing. (brad@)

  ~ agpvar.h                              

  > Clean up unused definitions. ok miod@ (matthieu@)

dev/pcmcia

  ~ if_an_pcmcia.c                        

  > ansi. no binary change. (jsg@)

dev/rasops

  ~ rasops.c                              

  > Omit specific depth in rasops_init_devcmap() if not compiling for them.
  > (miod@)

== usr.bin =========================================================== 07/08 ==

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

cap_mkdb

  ~ getinfo.c                             

  > Don't ferror() after fclose().
  > Don't dereference NULL pointer if file is truncated.
  > Sync with cgetnext(3).
  > ok otto and millert (ray@)

column

  ~ column.c                              

  > realloc & size_t police; simplify some pointer arithmetic; ok ray@ (otto@)

cvs

  ~ worklist.c                            

  > make sure cvs_worklist_clean() is safe, since we are
  > calling it from inside a signal handler.
  > from deraadt@, ok niallo@ (joris@)

  ~ rcs.c                                 

  > Only write a log message if the log message is not empty.
  > Also prevents rfp->rf_desc[-1] from happening.
  > ok joris, niallo, otto, and xsa (ray@)

  ~ worklist.c                            

  > fix a stupid mistake in cvs_worklist_clean() while loop.
  > ok joris@ (niallo@)

mg

  ~ theo.c                                

  > i am going to be hated for including this one (djm@)

rcs

  ~ ci.c                                  

  > use same date format as gnu/usr.bin/rcs in error message. OK niallo at .  > (xsa@)

== usr.sbin ========================================================== 08/08 ==

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

config

  ~ scan.l                                

  > Avoid printing the whole file in the error message on missing trailing "
  > Found by by mpech@; ok mpech@ ray@ miod@ (otto@)

hostapd

  ~ hostapd.conf.5                        ~ iapp.h

  > some IAPP implementations still use the pre-standard port 2313 in broadcast
  > mode (tested with the HP ProCurve 420wl). add a note to the manual page and
  > the example configuration. (reyk@)

tcpdump

  ~ print-iapp.c                          

  > add whitespace (reyk@)

  ~ print-udp.c                           

  > print IAPP messages for the old IAPP port 2313. (reyk@)

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



More information about the odc mailing list