OpenBSD src changes summary for 2016-04-18 ==========================================
distrib/sets lib/libedit sbin/ifconfig share/zoneinfo sys/arch/amd64/amd64 sys/conf sys/crypto sys/dev/ic sys/net sys/netinet usr.sbin/eigrpd usr.sbin/smtpd == distrib =========================================================== 01/06 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib sets ~ lists/base/mi > sync (deraadt@) == lib =============================================================== 02/06 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib libedit ~ common.c ~ editrc.5 ~ map.c > Delete redundant, empty callbacks; > no functional change, minus hundred lines of code. > OK martijn@; > also proof-read by Christian Heckendorf <mbie at ulmus dot me>. (schwarze@) == sbin ============================================================== 03/06 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin ifconfig ~ ifconfig.c > Print interface index after priority. > Suggestion from claudio@, ok benno@, sthen@ (mpi@) == share ============================================================= 04/06 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share zoneinfo ~ datfiles/asia ~ datfiles/europe ~ datfiles/northamerica ~ datfiles/southamerica ~ datfiles/zone.tab ~ datfiles/zone1970.tab > Update to tzdata2016d from from ftp.iana.org. (millert@) == sys =============================================================== 05/06 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/amd64/amd64 ~ aesni.c > Make the aesni crypto implementation mpsafe. In order to do so this moves > the working buffer from the softc into session struct. The list of > sessions > is protected by a mutex. > There has been some discussion about what IPL should be used for the mutex. > Initially my code used IPL_HIGH since that is the default to be used for > subsystems that can be called from any other subsystem. But since the > crypto > code may call malloc/free, I settled on IPL_VM. Calling it from an > interrupt > handler that runs at a higher level is unsafe. Sorry, can't encrypt audio! > ok mikeb@ (kettenis@) conf ~ param.c > Bump the default of cachepercent to 90 to see if we can find problems > before > we try to remove it entirely > ok deraadt@ (beck@) crypto ~ crypto.c ~ cryptodev.h > Add a mechanism for dispatching mpsafe crypto operations. This adds a new > CRYPTOCAP_F_MPSAFE flag that crypto implementations can set to indicate > that > their cc_process() implementation can safely run without holding the kernel > lock. > ok mikeb@ (kettenis@) dev/ic ~ nvme.c > allocate an array of entries, not pointers for the queues > this solves my memory corruption problem with a samsung sm951 in a > particular slot on a dell 2950. > hilariously, i had picked values which masked this problem on > sparc64. i randomly picked 128 as the number of entries on the > queues, and dmamem allocs get rounded up to PAGE_SIZE. on amd64 and > sparc64 this meant i was asking for 128 * 8 (sizeof pointer), or > 1024 bytes, which got rounded up to 4096 and 8192 on each arch > respectively. 128 * 64 (the size of a submission queue entry) is > 8192, so it worked fine on sparc64 for that reason, but randomly > blows up on amd64. the 2950 above allocated mbufs out of the page > after the submission queue, which i ended over overwriting. > anyway. let's move on. (dlg@) net ~ if_pppoe.c ~ if_spppsubr.c > Remove the hack that prevents changing pppoe params at runtime. > The EBUSY hack imposes an order on the ifconfig commands issued > against the pppoe interface used to configure the sppp layer below. > To counter this we use the ENETRESET trick that other drivers use > to tell the pppoe layer that sppp has requested a stop/init reset > sequence to proceed which we oblige with in case pppoe is UP and > RUNNING. > Tested by semarie@ and Jan Schreiber <[email protected]>, thanks! (mikeb@) netinet ~ in.c ~ ip_input.c ~ ip_output.c > Put a KERNEL_LOCK/UNLOCK dance around sections that still need some > work in the forwarding path. > Tested by Hrvoje Popovski, ok dlg@ (mpi@) ~ ip_input.c ~ ip_output.c > Unbreak RAMDISK, found by deraadt@ (mpi@) == usr.sbin ========================================================== 06/06 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin eigrpd ~ eigrpd.c ~ eigrpe.c ~ rde.c > Revert previous commit. > When a process receives an EOF on a socketpair, it removes the event > handler for that fd and then calls event_loopexit(). Once the process > leaves the main loop, it calls its shutdown function. With that said, > revert the previous patch because it was unnecessary. > Pointed out and ok by deraadt@ (renato@) smtpd ~ smtpd.8 > document the -T flag > original diff from Bernard Spil > ok millert (jung@) ~ smtpd.8 > add -T to SYNOPSIS; (jmc@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
