[odc] Daily src changes for 2007-10-05

[ODC]

OpenBSD src changes summary for 2007-10-05
==========================================

regress/usr.bin                         sbin/growfs
sys/dev/ic                              sys/dev/mii
sys/dev/pci                             sys/net
sys/ufs/ufs                             usr.bin/cvs
usr.bin/file                            usr.bin/learn
usr.bin/mail                            usr.bin/pcc
usr.sbin/hoststated                     

== regress =========================================================== 01/05 ==

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

usr.bin

  ~ pcc/cc/cpp/Makefile                   ~ pcc/cc/cpp/skip001.out

  > fix by actually using pcc (otto@)

  + pcc/Makefile                          + pcc/cc/Makefile
  + pcc/cc/ccom/Makefile                  

  > more regress scaffolding (otto@)

  ~ pcc/cc/cpp/Makefile                   ~ pcc/cc/cpp/skip001.out

  > work in obj dir (otto@)

  ~ bc/Makefile                           

  > typo (otto@)

  ~ pcc/cc/ccom/Makefile                  ~ pcc/cc/ccom/enum001.c
  ~ pcc/cc/ccom/init004.c                 ~ pcc/cc/ccom/tmpalloc001.c
  ~ pcc/cc/cpp/Makefile                   

  > basic tests now work, including those that should fail (otto@)

  ~ pcc/cc/ccom/darray001.c               

  > typo (otto@)

== sbin ============================================================== 02/05 ==

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

growfs

  ~ growfs.c                              

  > check fgets return value
  > use strncmp instead of strcmp
  > with help of ray@ "Looks OK" millert@ (chl@)

== sys =============================================================== 03/05 ==

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

dev/ic

  ~ anreg.h                               

  > gilles missed one of the __attribute__((__packed__)) here in
  > the __attribute__((__packed__)) -> __packed conversion. (jsg@)

dev/mii

  ~ eephy.c                               ~ eephyreg.h

  > Disable fiber/copper auto-selection on the 88E1111 if it is in RGMII mode,
  > to
  > work around the fact that the onboard PHYs attached to nfe(4) on the Sun
  > X4100 M2 have fiber/copper auto-selection enabled even though the
  > interfaces
  > are clearly copper-only.
  > Fixes PR system/5587.
  > ok jsg@, claudio@ (kettenis@)

dev/pci

  ~ azalia_codec.c                        

  > The Acer TravelMate 3010 (ALC883) has a couple of gpio pins that need
  > to be enabled in order for it to produce sound.  It can use the same
  > gpio function as the macs, so rename that to azalia_gpio_unmute.
  > Tested by damien@, no regressions noticed by jakemsr@ (deanna@)

net

  ~ if_media.h                            

  > add a 2.5 Gbps fiber media type.
  > From FreeBSD
  > ok claudio@ dlg@ (brad@)

ufs/ufs

  ~ ufs_dirhash.c                         

  > MALLOC/FREE -> malloc/free + M_ZERO.
  > As a side effect, this probably fixes PR5596, if the allocation
  > of dh_hash succeeds and the dh_blkfree fails, we jump into the
  > fail case, but we haven't initialized dh_hash properly, that is
  > filling the array with memory from the dirhash pool, but the
  > != NULL check holds, since the memory hasn't been zeroed and
  > so we start pool_put()'ing, causing the crash in PR5596.
  > PR5596 debugging by pedro.
  > ok art@, krw@ (thib@)

== usr.bin =========================================================== 04/05 ==

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

cvs

  ~ edit.c                                ~ entries.c
  ~ file.c                                ~ root.c
  ~ util.c                                

  > strcspn() change
  > was okay'd by pyr@ and ok by millert@ (gilles@)

file

  ~ apprentice.c                          

  > use strcspn instead of strchr to properly overwrite '\n' in fgets returned
  > buffer
  > check if the first byte is NUL instead of invoking strlen()
  > with help of ray@ "Looks OK" millert@ (chl@)

learn

  ~ src/learn.c                           

  > use strcspn to properly overwrite '\n' in fgets returned buffer
  > check fgets return value
  > use sizeof (buf) instead of hardcoded value
  > with help of moritz@ and ray@
  > "Those seem OK" moritz@ "OK" ray@ "Looks OK" millert@ (chl@)

mail

  ~ fio.c                                 

  > check if it's really a newline before removing it
  > with the help of ray@ "Looks OK" millert@ (chl@)

pcc

  ~ cc/ccom/init.c                        

  > optimize a "not so bright" piece of code. Reduces compilation time on my
  > evil test case from > 3m to < 1s. ok ragge@ (otto@)

  ~ arch/x86/local.c                      ~ arch/x86/macdefs.h

  > From master repository log:
  > Do not optimize away a cast if constant too large for target type.
  > Should remove branch but need to check for side effects in that case.
  > Also correct possible incorrect min constant. (ragge@)

== usr.sbin ========================================================== 05/05 ==

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

hoststated

  ~ hoststated.h                          

  > cast to an int (reyk@)

  ~ parse.y                               

  > cast to an int (reyk@)

  ~ relay.c                               

  > unbreak non-SSL relays by calling the ssl context init only if the SSL
  > flag is present... (reyk@)

  ~ hoststated.c                          ~ hoststated.h
  ~ relay.c                               

  > stylistic changes in the relay/relay_config structure. (reyk@)

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