OpenBSD src changes summary for 2015-11-06 ==========================================
bin/test distrib/sets lib/libc lib/libm lib/libssl libexec/getty libexec/ld.so regress/usr.bin sys/arch/i386/i386 sys/arch/socppc/dev sys/arch/sparc/dev sys/arch/sparc/sparc sys/arch/sparc64/sparc64 sys/arch/vax/if sys/conf sys/crypto sys/net sys/netinet sys/netinet6 usr.bin/at usr.bin/less usr.bin/make usr.bin/mandoc usr.bin/ssh usr.sbin/bgpd usr.sbin/cron usr.sbin/ikectl usr.sbin/netgroup_mkdb usr.sbin/pkg_add usr.sbin/relayd usr.sbin/traceroute == bin =============================================================== 01/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/bin test ~ test.1 > add [ to NAME, and document it; the NAME addition was > proposed by ingo a while back but never got committed; > ok schwarze (jmc@) == distrib =========================================================== 02/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib sets ~ lists/base/mi > sync (deraadt@) == lib =============================================================== 03/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib libc ~ sys/pledge.2 > Remove needless comma. (mmcc@) libm ~ src/e_hypotf.c > From FreeBSD 23397: > Fixed wrong magic numbers in scaling. hypotf() was very broken for large > and small values: > hypotf(2.3819765e+38, 2.0416943e+38) was NaN instead of 3.1372484e+38 > hypotf(-3.4028235e+38, 3.3886450e+38) was NaN instead of Inf > hypotf(-2.8025969e-45, -2.8025969e-45) was 0 instead of 4.2038954e-45 > Found by: ucbtest > ok miod@ (shadchin@) libssl ~ src/crypto/bn/bn_lcl.h > Fix gcc version preprocessor checks to cope with gcc 5.x and beyond; > reported by Ruslan Babayev. (miod@) == libexec =========================================================== 04/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/libexec getty ~ main.c > remove ppplogin support. ok sthen (tedu@) ~ gettytab.5 ~ gettytab.h ~ init.c ~ main.c > further PPP reduction, hint from sthen (tedu@) ld.so ~ library.c ~ library_mquery.c > Fix unloading of load groups when the last reference wasn't on the > load_object but rather some descendent. Detect that case in > _dl_unload_shlib() and switch to unloading the entire group. > Based on partial analyses by Henri Kemppainen (duclare (at) guu.fi) > and Peter Hajdu (peter.ferenc.hajdu (at) gmail.com) > ok millert@ (guenther@) == regress =========================================================== 05/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress usr.bin ~ mandoc/char/unicode/input.out_utf8 ~ mandoc/char/unicode/nogroff.out_utf8 > The recent update to /usr/share/locale/UTF-8/LC_CTYPE by afresh1@ > fixed wcwidth(3) for various unusual characters. (schwarze@) == sys =============================================================== 06/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/i386/i386 ~ cpu.c > Only read/write cr4 if we have a bit to set after testing the various > cpuid based flags. > Unlike amd64 there is no CR4_DEFAULT used here, cr4 does not exist > on most/all? 486s and cr4 access traps on Cyrix 5x86 processors. > Should fix a regression introduced in rev 1.65 encountered > by 'Ppractivus' on a 5x86. > "looks good" guenther@ (jsg@) arch/socppc/dev ~ if_tsec.c > Use m_defrag(9) when we fail to load fragmented mbuf chains. > From Richard Procter <richard DOT n DOT procter AT gmail DOT com>. > ok dlg@ (mpi@) arch/sparc/dev ~ if_ie.c > make hfsc work on ie(4/sparc) by using IFQ_DEQUEUE instead of IF_DEQUEUE. > (dlg@) arch/sparc/sparc ~ trap.c ~ vm_machdep.c > Move the logic for adjusting userspace registers in the child after fork > from cpu_fork() to child_return(), putting all the SYSCALL_G2RFLAG logic > in trap.c > sparc testing by sebastia@ and miod@ (guenther@) arch/sparc64/sparc64 ~ trap.c ~ vm_machdep.c > Move the logic for adjusting userspace registers in the child after fork > from cpu_fork() to child_return(), putting all the SYSCALL_G2RFLAG logic > in trap.c > sparc testing by sebastia@ and miod@ (guenther@) arch/vax/if ~ if_de.c > use IFQ_DEQUEUE, not IF_DEQUEUE, to pull off the send queue. > this is more correct, as it supports hfsc use on these interfaces. (dlg@) conf ~ files > Do not compile net/radix_mpath.c in ART-enabled kernels. > In such kernels multipath routing support is independant from the data > structure used to perform the binary lookup. Is is implemented as a > singly-listed list of route entries. This is possible because a "struct > rtentry" is no longer mapped 1:1 to a node in the tree. (mpi@) crypto ~ gmac.c > Instead of multiplying with 0..1, extend the bit into a mask and do an AND. > The same technique was already used a few lines above. ok mikeb@ (naddy@) net ~ if_ppp.c > replace sc_npqueue and sc_npqtail with a simple mbuf_list. (dlg@) ~ if_pppvar.h > unbreak tree after src/sys/net/if_ppp.c r1.95. > i forgot to commit the change of the npqueue type in the header. > noticed by benoit lecocq (dlg@) ~ if_mpw.c > Do not use rt_ifp directly. > ok rzalamena@ (mpi@) ~ if_mpe.c > Do not use rt_ifp directly. > ok claudio@ (mpi@) ~ rtable.c > In ART separate the MPATH delete case to properly recover if art_delete() > does not find a matching node. > This currently never happens because we always do a route lookup before > calling rtable_delete(). Yes this is odd & due to the way multipath is > implemented in the radix tree. (mpi@) ~ art.h ~ route.h ~ rtable.c > Use a SLIST instead of a LIST for MPATH route entries with ART. (mpi@) ~ if.c ~ pf.c ~ radix_mpath.c ~ route.h ~ rtable.c ~ rtable.h > Rename rt_mpath_next() into rtable_mpath_next() and provide an > implementation for ART based on the singly-linked list of route > entries. (mpi@) ~ radix.c > Do not compile net/radix_mpath.c in ART-enabled kernels. > In such kernels multipath routing support is independant from the data > structure used to perform the binary lookup. Is is implemented as a > singly-listed list of route entries. This is possible because a "struct > rtentry" is no longer mapped 1:1 to a node in the tree. (mpi@) netinet ~ tcp_input.c > Change nd6_nud_hint() to no longer manipulate rt_ifp directly. > While here remove unused argument and convert the route check to > rtisvalid(9). > ok bluhm@ (mpi@) ~ if_ether.c > In in_arpinput() do not make an extra route lookup to check wether > the sender address is a local IP. The arplookup() does this route > lookup anyway, so just check its result. > OK mpi@ (bluhm@) netinet6 ~ nd6.c ~ nd6.h > Change nd6_nud_hint() to no longer manipulate rt_ifp directly. > While here remove unused argument and convert the route check to > rtisvalid(9). > ok bluhm@ (mpi@) == usr.bin =========================================================== 07/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin at ~ at.c ~ parsetime.c > Use __progname instead of the homegrown ProgramName. (millert@) less ~ less.h ~ lsystem.c ~ signal.c > Remove LSIGNAL which is just the same as lsignal() now. (nicm@) ~ funcs.h ~ linenum.c ~ os.c > time_t is not a long. remove incorrect abstraction around time(). > ok nicm (tedu@) ~ brac.c ~ ch.c ~ charset.c ~ charset.h ~ cmd.h ~ cmdbuf.c ~ command.c ~ cvt.c ~ decode.c ~ edit.c ~ filename.c ~ forwback.c ~ ifile.c ~ input.c ~ jump.c ~ less.h ~ lesskey.c ~ lesskey.h ~ line.c ~ linenum.c ~ lsystem.c ~ main.c ~ mark.c ~ optfunc.c ~ option.c ~ option.h ~ opttbl.c ~ os.c ~ output.c ~ pattern.c ~ pattern.h ~ position.c ~ position.h ~ prompt.c ~ screen.c ~ search.c ~ signal.c ~ tags.c ~ ttyin.c ~ version.c > Garrett D'Amore has agreed to moving his copyright line up above the > terms with the existing copyright, so it is clear it applies the same > terms. (nicm@) ~ cmd.h ~ command.c ~ decode.c ~ less.1 ~ lesskey.c > Remove support for ! to run a shell command, we have ^Z around these > parts. ok ratchov jung millert (nicm@) ~ charset.c ~ less.1 > remove rarely used LESSCHARDEF support. ok nicm (tedu@) make ~ parse.c > I was very optimistic in groupling creation. Turns out lists of targets > can have duplicates, or overlap, or even be empty thanks to fnmatch. > So use the big guns to build the circular list correctly, namely > actual lists of targets that are made together will be registered in a > hash first, then we recreate the full list from there. > (merging lists is not an issue, since groupling links are only used to > temporarily lock targets in parallel mode). > Issue noticed by guenther@, okay guenther@ (espie@) mandoc ~ main.c ~ mandocdb.c > Use getprogname(3) rather than __progname. > Suggested by Joerg@ Sonnenberger (NetBSD). > Last year, deraadt@ confirmed on tech@ that this "has the potential > to be more portable", and micro-optimizing for speed is not relevant > here. Also gets rid of one global variable. (schwarze@) ~ main.c > correct exit status on pledge(2) errors (schwarze@) ssh ~ ssh.1 > 1. rlogin and rsh are long gone > 2. protocol version isn't of core relevance here, and v1 is going away > ok markus@, deraadt@ (mmcc@) == usr.sbin ========================================================== 08/08 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin bgpd ~ bgpd.h ~ rde.c ~ rde.h ~ rde_filter.c > Radically improve the performance of bgpd filters. Based on PF's skip > steps (and uses much of the same code). > In a torture test of ~600k prefix filters and 65k prefixes, convergance > time goes from 35 minutes to 30 seconds. > Many thanks to LONAP for providing a base configuration for torture > testing. > many discussions with claudio@, benno@, sthen@ and the rest of the bgpd > crowd > OK sthen@ benno@ (phessler@) cron ~ client.c ~ cron.c ~ crontab.c ~ globals.h ~ misc.c > Use __progname instead of the homegrown ProgramName. (millert@) ikectl ~ ikectl.c > Use pledge in ikectl. For now one request for sending imsgs to iked > another request for the ca portion. > ok deraadt@ (jsg@) netgroup_mkdb ~ util.c > err(1, "%s", "") -> err(1, NULL) > Discussed with schwarze@ (mmcc@) pkg_add ~ OpenBSD/PkgCreate.pm > the Solver uses kept_names, so we need these there too > found by aja@ (espie@) relayd ~ relayd.conf.5 > relayd.conf(5) macro cleanup. > - use <> instead of \*(Lt and \*(Gt > - use <> instead of Aq (Aq is not the same as <> in a UTF-8 locale) > - replace Ar usage when appropriate > - mark up RTP_STATIC with Dv > with input from jmc@ schwarze@, ok schwarze@ (bentley@) traceroute ~ traceroute.c > "source" doesn't need to be a global. While here, the correct spelling > is NULL. > Noticed while trying to hoist the source address selection from > traceroute into ping / ping6. (florian@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
