OpenBSD src changes summary for 2015-11-28
==========================================

gnu/usr.bin/binutils-2.17               sbin/growfs
sbin/newfs_ext2fs                       sys/arch/amd64/stand/efiboot
sys/dev/ic                              sys/dev/pci
sys/dev/sbus                            sys/kern
sys/net                                 sys/sys
sys/ufs/ffs                             usr.bin/grep
usr.bin/ssh                             usr.sbin/relayctl
usr.sbin/relayd                         usr.sbin/sysmerge

== gnu =============================================================== 01/05 ==

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

usr.bin/binutils-2.17

  ~ bfd/elflink.c                         

  > Back out previous commit.  Not emitting DT_RPATH isn't helpful if our ld.so
  > doesn't grok DT_RUNPATH yet. (kettenis@)

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

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

growfs

  ~ growfs.c                              

  > hoist TIOCGWINSZ code upwards, and pledge "stdio disklabel" at the
  > right place.
  > tested by mmcc (deraadt@)

newfs_ext2fs

  ~ newfs_ext2fs.c                        

  > Oops, left the pledge "tty" active in the previous commited
  > pointed out by Chris Hettrick (deraadt@)

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

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

arch/amd64/stand/efiboot

  ~ efidev.c                              

  > Check the protective MBR stricter not to choose a broken GPT.
  > test and ok gonzalo (yasuoka@)

  ~ efidev.c                              

  > Use the same function as kern/subr_disk.c and fdisk/mbr.c to check
  > for a GPT protective MBR. Better than mangling findopenbsd() to do
  > the check.
  > Tested & ok yasuoka@ (krw@)

dev/ic

  ~ gem.c                                 ~ gemvar.h

  > Rework gem_start to check that there's enough space in the ring before
  > dequeueing a packet, then drop the packet if it can't be sent, rather than
  > using two phase dequeue then commit.
  > Adjust sc_tx_cnt using atomic operations, take the kernel lock before
  > calling gem_start or gem_init from the interrupt handler, and use an
  > interrupt barrier when taking the interface down.
  > With all this done, we can mark the interrupt handler for pci (but not
  > sbus)
  > attached gem(4) as mpsafe.
  > mpi@ wrote this initially, I fixed bugs to keep sparc64 happy and kept it
  > up
  > to date with ifq changes.
  > ok mpi@ dlg@ (jmatthew@)

  ~ dc.c                                  

  > rework dc_start and dc_encap to take advantage of m_defrag.
  > if the chip needs coalesced packages in tx, set the tx dmamaps up
  > to only use a single dma descriptor. use m_defrag when bus_dmamap_load_mbuf
  > returns EFBIG rather than copying the packet to a separate mbuf
  > with hand rolled code in dc_coal. that in turn makes the ifq_deq_begin,
  > ifq_deq_commit, and ifq_deq_rollback handling more straightforward.
  > tested by me on a hppa a180c with a "DEC 21142/3", and fred on bugs@
  > with a sparc64 netra x1 "Davicom DM9102". (dlg@)

dev/pci

  ~ if_gem_pci.c                          

  > Rework gem_start to check that there's enough space in the ring before
  > dequeueing a packet, then drop the packet if it can't be sent, rather than
  > using two phase dequeue then commit.
  > Adjust sc_tx_cnt using atomic operations, take the kernel lock before
  > calling gem_start or gem_init from the interrupt handler, and use an
  > interrupt barrier when taking the interface down.
  > With all this done, we can mark the interrupt handler for pci (but not
  > sbus)
  > attached gem(4) as mpsafe.
  > mpi@ wrote this initially, I fixed bugs to keep sparc64 happy and kept it
  > up
  > to date with ifq changes.
  > ok mpi@ dlg@ (jmatthew@)

  ~ pcidevs                               

  > Add a few devices found on the MacBookPro12,1 (kettenis@)

  ~ pcidevs.h                             ~ pcidevs_data.h

  > regen (kettenis@)

dev/sbus

  ~ if_gem_sbus.c                         

  > Rework gem_start to check that there's enough space in the ring before
  > dequeueing a packet, then drop the packet if it can't be sent, rather than
  > using two phase dequeue then commit.
  > Adjust sc_tx_cnt using atomic operations, take the kernel lock before
  > calling gem_start or gem_init from the interrupt handler, and use an
  > interrupt barrier when taking the interface down.
  > With all this done, we can mark the interrupt handler for pci (but not
  > sbus)
  > attached gem(4) as mpsafe.
  > mpi@ wrote this initially, I fixed bugs to keep sparc64 happy and kept it
  > up
  > to date with ifq changes.
  > ok mpi@ dlg@ (jmatthew@)

kern

  ~ kern_pledge.c                         

  > pledge: allow getsockopt IP_IPDEFTTL with promise inet
  > then relayd's host check engine can be pledged.
  > ok reyk@, approach suggested by deraadt@ weeks ago. (benno@)

  ~ vfs_bio.c                             

  > move buffer size adjustment to buf_adjcnt - from Walter Neto
  > ok mpi@ (beck@)

net

  ~ if_bridge.h                           

  > Delete sc_root_port which became unused when RSTP is merged.
  > ok reyk mpi (yasuoka@)

  ~ if_pppoe.c                            

  > Convert pppoe(4) to use if_get()/if_put(): instead of storing a
  > pointer to the parent "pppoedev", it now only stores an interface
  > index.  This also fixes a potential NULL pointer dereference that
  > could happen in pppoe_find_softc_by_session() when the parent got
  > deconfigured but the session was still active.
  > Found the hard way with pppoe(4) on vlan7.
  > OK mpi@, with debugging help from mikeb@ (reyk@)

sys

  ~ buf.h                                 

  > move buffer size adjustment to buf_adjcnt - from Walter Neto
  > ok mpi@ (beck@)

ufs/ffs

  ~ ffs_alloc.c                           ~ ffs_balloc.c
  ~ ffs_inode.c                           ~ ffs_subr.c

  > move buffer size adjustment to buf_adjcnt - from Walter Neto
  > ok mpi@ (beck@)

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

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

grep

  ~ grep.c                                

  > fix exit status on pledge(2) error, where it should be >1
  > OK millert@ deraadt@ (gsoares@)

ssh

  ~ serverloop.c                          

  > don't include port number in tcpip-forward replies for requests
  > that don't allocate a port; bz#2509 diagnosed by Ron Frederick
  > ok markus (djm@)

  ~ ssh-keygen.c                          

  > do not leak temp file if there is no known_hosts file
  > from craig leres, ok djm (deraadt@)

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

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

relayctl

  ~ relayctl.8                            

  > add missing "log" bits
  > spotted out by Luiz Gustavo Costa < luizgustavo at mundounix.com.br >,
  > Thanks!
  > OK jmc@ jsg@ (gsoares@)

relayd

  ~ check_icmp.c                          ~ check_tcp.c
  ~ control.c                             ~ proc.c
  ~ relay.c                               ~ relay_udp.c
  ~ relayd.c                              ~ relayd.h
  ~ snmp.c                                

  > Use SOCK_NONBLOCK in relayd as well.
  > OK benno@ (reyk@)

  ~ hce.c                                 

  > pledge: allow getsockopt IP_IPDEFTTL with promise inet
  > then relayd's host check engine can be pledged.
  > ok reyk@, approach suggested by deraadt@ weeks ago. (benno@)

sysmerge

  ~ sysmerge.sh                           

  > chflags -l flag went away.  Noticed this myself 15 minutes
  > before Chris Hettrick did. (deraadt@)

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

Reply via email to