OpenBSD src changes summary for 2016-02-09 ==========================================
etc/examples/pkg.conf games/snake sys/arch/hppa/include sys/arch/hppa64/include sys/net sys/net80211 sys/netinet sys/nfs usr.bin/hexdump usr.bin/ssh usr.bin/vi usr.sbin/pkg_add usr.sbin/radiusd usr.sbin/rtadvd usr.sbin/smtpd == etc =============================================================== 01/05 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/etc examples/pkg.conf ~ examples/pkg.conf > sync (sthen@) == games ============================================================= 02/05 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/games snake ~ snake.c > - Remove parameter fd from snscore() since it's never used > - And while here, fwrite(3) returns size_t whereas nscores is an int, so > cast > nscores to u_int. This is a false positive, but silences a compiler warning > with -Wextra -pedantic > tb@ : "Looks ok to me" after suggestion from him (mestre@) == sys =============================================================== 03/05 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/hppa/include ~ atomic.h > make __cpu_simple_lock provide serialisation of the critical section. > that in turn makes atomic sequences actually atomic, which in turn > means the refcnt api asserts wont fire erronously when if_get and > if_put are actually used correctly. > such embarrassment. > reported by landry@ who also let me debug on the affected machines > ok jmatthew@ (dlg@) arch/hppa64/include ~ atomic.h > bring in dlg's fix from rev 1.10 of the equivalent hppa file: > make __cpu_simple_lock provide serialisation of the critical section. > that in turn makes atomic sequences actually atomic, which in turn > means the refcnt api asserts wont fire erronously when if_get and > if_put are actually used correctly. (jsg@) net ~ if_vlan.c TAGGED OPENBSD_5_8 > There was another bug in vlan and carp ifih refcounting. An ifih > could be inserted twice into the the parent's list. Then the kernel > will panic if root does an ifconfig destroy of the parent interface. > This affects OpenBSD 5.8 only. > From markus@; OK mpi@ (bluhm@) net80211 ~ ieee80211_input.c > Log frames which fall outside the BlockAack window in dmesg if the > interface debug flag is set (enabled with: ifconfig iwn0 debug). > Shows the frame's sequence number and the current BA window. > I'm adding this for diagnosis, just in case it will be needed to make > future decisions about tuning the heuristic which works around network > stalls caused by such frames. (stsp@) netinet ~ ip_carp.c TAGGED OPENBSD_5_8 > There was another bug in vlan and carp ifih refcounting. An ifih > could be inserted twice into the the parent's list. Then the kernel > will panic if root does an ifconfig destroy of the parent interface. > This affects OpenBSD 5.8 only. > From markus@; OK mpi@ (bluhm@) nfs ~ nfs_node.c > sync a function's comment with its signature > ok tedu@, with a slight tweak (mmcc@) == usr.bin =========================================================== 04/05 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin hexdump ~ display.c ~ hexdump.h ~ parse.c > For a long time, nomem() has been a wrapper for err(1, NULL) and > meanwhile emalloc() has turned into calloc() by doing malloc + memset. > Delete them and Use the libc functions directly instead. I've been > running with this diff for ages and was reminded of it by a similar > suggestion from mmcc@ and by recent diffs from Michal Mazurek. > ok mmcc@ (tb@) ~ parse.c > Replace a three step string duplication with strndup(). > ok and with input from tb@ (mmcc@) ssh ~ myproposal.h > turn off more old crypto in the client: hmac-md5, ripemd, truncated > HMACs, RC4, blowfish. ok markus@ dtucker@ (djm@) vi ~ docs/USD.doc/vi.man/vi.1 > avoid special characters; from michael reed (jmc@) == usr.sbin ========================================================== 05/05 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin pkg_add ~ OpenBSD/PackageRepository.pm > drop privs prior to running ftp > okay sthen@ (espie@) ~ OpenBSD/PackageRepository.pm > tweak names > set LC_ALL everywhere since it's now easy (espie@) radiusd ~ radiusd.c > fix a use after free in an error path > ok yasuoka@ (jsg@) rtadvd ~ config.c ~ rtadvd.c ~ rtadvd.h > Kill code to send unicast packets which never got enabled. > and get rid of pointless memory allocations. > ok bluhm@ (jca@) ~ rtadvd.c > another typo (jca@) smtpd ~ control.c ~ smtpctl.c > rename variable sun to s_un to ease portability > diff from Freddy DISSAUX, ok gilles@, jung@, sunil@, millert@ (gilles@) ~ mproc.c > r1.16 accidentally masked retries on EAGAIN introduced in r1.15. > Fix and simplify imsg_read() return value checking into a single > if..else block. > Ok gilles@ jung@ (sunil@) ~ smtpd-api.h > add PROC_QUEUE_MESSAGE_UNCORRUPT to smtpd-api.h, needed by privsep queues > (gilles@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
