OpenBSD src changes summary for 2016-10-16 to 2016-10-23 inclusive ==================================================================
Makefile bin/cat bin/chmod bin/date bin/expr bin/ksh bin/mkdir bin/ps bin/rmdir distrib/amd64 distrib/miniroot distrib/notes distrib/sets gnu gnu/usr.bin/cc include/Makefile lib/libc lib/libcrypto lib/libpcap lib/librpcsvc lib/libssl libexec/security libexec/spamd regress/lib regress/sys regress/usr.bin sbin/bioctl sbin/ifconfig sbin/mknod sbin/ping sbin/sysctl share/man share/zoneinfo sys/arch/alpha/alpha sys/arch/amd64/amd64 sys/arch/arm/arm sys/arch/arm/include sys/arch/arm/simplebus sys/arch/armv7/armv7 sys/arch/armv7/conf sys/arch/armv7/imx sys/arch/armv7/marvell sys/arch/armv7/omap sys/arch/armv7/stand/efiboot sys/arch/armv7/sunxi sys/arch/hppa/hppa sys/arch/i386/conf sys/arch/i386/i386 sys/arch/i386/include sys/arch/m88k/m88k sys/arch/mips64/mips64 sys/arch/octeon/dev sys/arch/powerpc/powerpc sys/arch/sh/sh sys/arch/sparc64/sparc64 sys/conf sys/ddb sys/dev sys/dev/acpi sys/dev/pci sys/dev/pckbc sys/dev/wscons sys/kern sys/net sys/sys sys/uvm usr.bin/column usr.bin/cvs usr.bin/diff3 usr.bin/make usr.bin/mandoc usr.bin/rcs usr.bin/sendbug usr.bin/sndiod usr.bin/sort usr.bin/ssh usr.bin/tmux usr.bin/ul usr.sbin usr.sbin/bgpd usr.sbin/config usr.sbin/dhcpd usr.sbin/ldapd usr.sbin/makefs usr.sbin/ntpd usr.sbin/pstat usr.sbin/rebound usr.sbin/rpc.bootparamd usr.sbin/smtpd usr.sbin/switchctl usr.sbin/switchd usr.sbin/syslogd usr.sbin/tcpdump usr.sbin/traceroute usr.sbin/vmd usr.sbin/ypldap == Makefile ========================================================== 01/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/Makefile Makefile > Swap 'make cleandir' and 'make includes' for real. > Should have been part of commitid uHRFWpwUBjIz0UuR. > Original message: > Move libcrypto, librpcsvc and gnu/usr.bin/cc/include from RDIRS to PRDIRS, > and add prereq targets, so some header files are generated by BUILDUSER > during 'make prereq' instead of by root during 'make includes'. > Switch the order of 'make cleandir' and 'make includes' during 'make build' > so we don't generate many files twice. > Except for some machine@ symlinks from ${MACHINE}/stand, /usr/obj is now > clean from files generated by root during 'make build'. Those will be > cleaned up in a second step. > help, testing & ok deraadt, input from natano, further testing rpe (tb@) == bin =============================================================== 02/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/bin cat ~ cat.c > Delete useless setlocale(3) in src/bin/. > Minor cleanup while here: return from main, static void __dead usage, etc. > Based on a patch from Jan Stary <hans at stare dot cz>. > Feedback and OK tb@, OK millert@. (schwarze@) chmod ~ chmod.c > Delete useless setlocale(3) in src/bin/. > Minor cleanup while here: return from main, static void __dead usage, etc. > Based on a patch from Jan Stary <hans at stare dot cz>. > Feedback and OK tb@, OK millert@. (schwarze@) date ~ date.c > Delete useless setlocale(3) in src/bin/. > Minor cleanup while here: return from main, static void __dead usage, etc. > Based on a patch from Jan Stary <hans at stare dot cz>. > Feedback and OK tb@, OK millert@. (schwarze@) expr ~ expr.c > Delete useless setlocale(3) in src/bin/. > Minor cleanup while here: return from main, static void __dead usage, etc. > Based on a patch from Jan Stary <hans at stare dot cz>. > Feedback and OK tb@, OK millert@. (schwarze@) ksh ~ vi.c > Make sure the cursor goes to the first byte of a UTF-8 character > and not to a continuation byte for these commands: b B e E w W | > Let {e,E}ndword return the position after the word because that is > easier to handle in the caller. > OK tb@ czarkoff@ (schwarze@) ~ main.c ~ vi.c > No need to inspect LC_CTYPE: > * It was only used in vi input mode. > * No functional change in case of set -o vi-show8. > * No functional change if LC_CTYPE is set to UTF-8. > * More robust for the default of LC_CTYPE=C on a UTF-8 xterm. > Behaviour changes for the combination LC_CTYPE=C with set +o vi-show8 > on non-UTF-8 terminals, but that combination is useless and dangerous > with or without this patch. If you want to edit individual raw non-ASCII > non-UTF-8 bytes on the shell command line, always use set -o vi-show8. > (Besides, i doubt that you actually want to do that; better use a real > hex editor in the first place.) > OK czarkoff@. (schwarze@) ~ vi.c > resolve code duplication in domove() for the commands: b B w W > no functional change; suggested by tb@ (schwarze@) ~ vi.c > simplify implementation of the '$' command, no functional change; > OK nicm@ tb@ czarkoff@ (schwarze@) mkdir ~ mkdir.c > Delete useless setlocale(3) in src/bin/. > Minor cleanup while here: return from main, static void __dead usage, etc. > Based on a patch from Jan Stary <hans at stare dot cz>. > Feedback and OK tb@, OK millert@. (schwarze@) ps ~ ps.1 > Make the description of COLUMNS a bit more readable. > Triggered by a different patch from Michal Mazurek. > Joint work by tb@, jmc@, and tedu@, but no one ever committed it. > (schwarze@) rmdir ~ rmdir.c > Delete useless setlocale(3) in src/bin/. > Minor cleanup while here: return from main, static void __dead usage, etc. > Based on a patch from Jan Stary <hans at stare dot cz>. > Feedback and OK tb@, OK millert@. (schwarze@) == distrib =========================================================== 03/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib amd64 - common/Makefile.inc ~ ramdiskA/Makefile ~ ramdisk_cd/Makefile + ramdiskA/Makefile.inc + ramdisk_cd/Makefile.inc > Unshare Makefile.inc so we can determine how to squeeze makefs > into the picture. > ok natano (deraadt@) ~ ramdiskA/Makefile.inc > Inline IMAGE, CBIN, CRUNCHCONF and XNAME. > ok deraadt (natano@) ~ ramdisk_cd/Makefile ~ ramdisk_cd/Makefile.inc > Inline IMAGE, CBIN, CRUNCHCONF and XNAME. > ok deraadt (natano@) miniroot ~ install.sub > Exit autoinstall in case of missing / or duplicate mountpoints. > This also covers the case when a template is rejected by disklabel. > OK krw (rpe@) notes ~ m4.common > typo fix and zap some trailing eol whitespace; > from claus assman (jmc@) ~ m4.common > spelling: > initialising -> initializing from Claus Assmann > artitions -> partitions from spell(1) (tb@) sets ~ lists/base/mi ~ lists/man/mi > sync (deraadt@) ~ lists/base/md.alpha ~ lists/base/md.amd64 ~ lists/base/md.armv7 ~ lists/base/md.hppa ~ lists/base/md.i386 ~ lists/base/md.landisk ~ lists/base/md.loongson ~ lists/base/md.luna88k ~ lists/base/md.macppc ~ lists/base/md.octeon ~ lists/base/md.sgi ~ lists/base/md.socppc ~ lists/base/md.sparc64 ~ lists/comp/mi ~ lists/man/mi > sync (deraadt@) ~ lists/base/md.i386 ~ lists/comp/md.i386 ~ lists/comp/mi > sync (deraadt@) == gnu =============================================================== 04/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/gnu gnu ~ gcc/gcc/c-decl.c > Don't warn about shadowing a global function unless the local > variable or parameter is a pointer to a function. > ok kettenis@ (guenther@) usr.bin/cc ~ include/Makefile > Move libcrypto, librpcsvc and gnu/usr.bin/cc/include from RDIRS to PRDIRS, > and add prereq targets, so some header files are generated by BUILDUSER > during 'make prereq' instead of by root during 'make includes'. > Switch the order of 'make cleandir' and 'make includes' during 'make build' > so we don't generate many files twice. > Except for some machine@ symlinks from ${MACHINE}/stand, /usr/obj is now > clean from files generated by root during 'make build'. Those will be > cleaned up in a second step. > help, testing & ok deraadt, input from natano, further testing rpe (tb@) == include =========================================================== 05/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/include Makefile ~ Makefile > Move libcrypto, librpcsvc and gnu/usr.bin/cc/include from RDIRS to PRDIRS, > and add prereq targets, so some header files are generated by BUILDUSER > during 'make prereq' instead of by root during 'make includes'. > Switch the order of 'make cleandir' and 'make includes' during 'make build' > so we don't generate many files twice. > Except for some machine@ symlinks from ${MACHINE}/stand, /usr/obj is now > clean from files generated by root during 'make build'. Those will be > cleaned up in a second step. > help, testing & ok deraadt, input from natano, further testing rpe (tb@) == lib =============================================================== 06/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib libc ~ string/strlcat.c ~ string/strlcpy.c ~ string/strnlen.c > Roll back uintptr_t cast changes after discussions with tedu, otto and > others. > C11 6.5.6.9 says: > When two pointers are subtracted, both shall point to elements of the > same array object, or one past the last element of the array object; the > result is the difference of the subscripts of the two array elements. > In these cases the objects are arrays of char so the result is defined, > and we believe that the report is based on a compiler incorrectly trapping > on defined behaviour. (dtucker@) ~ gen/readpassphrase.c > Avoid generate SIGTTOU when restoring the terminal mode. If we get > SIGTTOU it means the process is not in the foreground process group > which, in most cases, means that the shell has taken control of the > tty. Requiring the user the fg the process in this case doesn't > make sense and can result in both SIGTSTP and SIGTTOU being sent > which can lead to the process being suspended again immediately > after being brought into the foreground. (millert@) ~ sys/getsockopt.2 > note that SO_REUSEADDR is restricted to callers with the same uid or root; > while there, super-user -> superuser for internal consistency; > diff from kirill miazine, tweaked by myself; > ok deraadt millert guenther (jmc@) ~ gen/syslog_r.c > Use the reentrant strerror_r() instead of strerror() to expand %m. > Previously, syslog_r() would avoid calling strerror() since the latter > is not reentrant. OK bluhm@ (millert@) ~ string/strerror_r.c > Remove the save_errno dance inside strerror_r(3). It is from the > time when we had national language support. > OK millert@ (bluhm@) ~ stdlib/malloc.c > Also place canaries in > page sized objects (if C is in effect); ok tb@ > (otto@) ~ stdlib/malloc.3 > canary corruption message changed a bit (otto@) ~ stdlib/malloc.3 > avoid sentence splicing; (jmc@) ~ stdlib/malloc.c > backout for now; flag combination GC is not ok (otto@) ~ stdlib/malloc.c > and recommit with fixed GC (otto@) ~ stdlib/malloc.c > fix some void * arithmetic by casting (otto@) ~ stdlib/malloc.c > - fix MALLOC_STATS compile > - redundant cast is redundant (otto@) ~ stdlib/bsearch.c ~ stdlib/heapsort.c ~ stdlib/rand.c > $OpenBSD$ (tb@) libcrypto ~ Makefile > Move libcrypto, librpcsvc and gnu/usr.bin/cc/include from RDIRS to PRDIRS, > and add prereq targets, so some header files are generated by BUILDUSER > during 'make prereq' instead of by root during 'make includes'. > Switch the order of 'make cleandir' and 'make includes' during 'make build' > so we don't generate many files twice. > Except for some machine@ symlinks from ${MACHINE}/stand, /usr/obj is now > clean from files generated by root during 'make build'. Those will be > cleaned up in a second step. > help, testing & ok deraadt, input from natano, further testing rpe (tb@) ~ bn/bn_print.c > If BN_div_word() fails (by returning (BN_ULONG)-1) or if the division > fails to reduce the input in the expected space then fail out instead > of overflowing the allocated buffer. > combines openssl commits 28a89639da50b1caed4ff3015508f23173bf3e49 and > 3612ff6fcec0e3d1f2a598135fe12177c0419582 > ok doug@ beck@ (guenther@) ~ dsa/dsa_ameth.c ~ ec/ec_ameth.c ~ err/err_all.c ~ gost/gostr341001_ameth.c ~ gost/gostr341001_pmeth.c ~ pem/pem_lib.c ~ rsa/rsa_ameth.c ~ rsa/rsa_pmeth.c > unifdef OPENSSL_NO_CMS (jsing@) libpcap ~ pcap.3 > Add missing argument names for pcap_offline_filter. (lteo@) librpcsvc ~ Makefile > Move libcrypto, librpcsvc and gnu/usr.bin/cc/include from RDIRS to PRDIRS, > and add prereq targets, so some header files are generated by BUILDUSER > during 'make prereq' instead of by root during 'make includes'. > Switch the order of 'make cleandir' and 'make includes' during 'make build' > so we don't generate many files twice. > Except for some machine@ symlinks from ${MACHINE}/stand, /usr/obj is now > clean from files generated by root during 'make build'. Those will be > cleaned up in a second step. > help, testing & ok deraadt, input from natano, further testing rpe (tb@) libssl ~ s3_clnt.c ~ s3_lib.c ~ s3_srvr.c ~ ssl_ciph.c ~ ssl_lib.c ~ ssl_locl.h ~ t1_lib.c > Remove support for fixed ECDH cipher suites - these is not widely supported > and more importantly they do not provide PFS (if you want to use ECDH, use > ECDHE instead). > With input from guenther@. > ok deraadt@ guenther@ (jsing@) == libexec =========================================================== 07/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/libexec security ~ security > When checking ownership and modes of files in /var/mail/, > ignore *.lock files, to avoid pointless warning mails > reported by Philippe Meunier <meunier at ccs dot neu dot edu>; > OK florian@ jca@ (schwarze@) spamd ~ grey.c ~ sync.c > - Remove useless var assignment > - Use memset(*b, 0, len) instead of bzero(*b, len) > - Use memcpy(*dst, *src, len) instead of bcopy(*src, *dst, len) > - Use explicit_bzero(3) instead of bzero(3) to clean temporary HMAC > Remarks from deraadt@ (duly noted!!), reviewed and OK by henning@ and tb@ > (mestre@) == regress =========================================================== 08/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress lib ~ libssl/client/clienttest.c > Update client hello messages to follow the removal of fixed ECDH. (jsing@) ~ libm/lgamma/lgamma.c > fix lgamma test on i386 and use symbolic constant; ok martynas@ (otto@) sys ~ netinet/pmtu/Makefile ~ netinet/pmtu/README > Do not change routes during test run, but expect that everything > is setup correctly before. Use make check-setup to verify that. (bluhm@) ~ net/pf_forward/Makefile ~ net/pf_fragment/Makefile ~ netinet/arp/Makefile ~ netinet/pmtu/Makefile ~ netinet6/frag6/Makefile ~ netinet6/nd6/Makefile ~ netinet6/rh0/Makefile > The clean:, cleandir:, and obj: targets of src/Makefile descend into > regress/ and run python to set the IMPORT_PYTHON variable in several > test Makefiles. This results in W^X violations during 'make build'. > Avoid doing that. > ok bluhm (tb@) + net/pf_state/challenge_ack.py + net/pf_state/LICENSE + net/pf_state/Makefile + net/pf_state/pf.conf + net/pf_state/README > New import: > Add a test suite for pf state handling. It needs a remote machine ~ net/Makefile > Link pf_state test to build. (bluhm@) ~ net/pf_state/Makefile ~ net/pf_state/challenge_ack.py > Make the test faster. Move all the packet matching code into the > pcap filter. That means the first packet sniffed is the correct > one. In the success case, we can stop without waiting for a timeout. > OK sashan@ (bluhm@) ~ net/pf_forward/ping6_mtu.py > Replace fork() and sleep() with a Python thread for sniffing packets. > This reduces test execution time from 1m20.34s to 0m37.32s. (bluhm@) ~ net/pf_fragment/frag.py ~ net/pf_fragment/frag6.py ~ net/pf_fragment/frag6_cutnew.py ~ net/pf_fragment/frag6_cutold.py ~ net/pf_fragment/frag6_dropnew.py ~ net/pf_fragment/frag6_dropold.py ~ net/pf_fragment/frag6_ext.py ~ net/pf_fragment/frag_cutnew.py ~ net/pf_fragment/frag_cutold.py ~ net/pf_fragment/frag_dropnew.py ~ net/pf_fragment/frag_dropold.py ~ net/pf_fragment/ping6_mtu_1300.py > Replace fork() and sleep() with a Python thread for sniffing packets. > This reduces test execution time from 2m21.95s to 1m09.80s. (bluhm@) ~ net/pf_forward/ping6_mtu.py ~ net/pf_fragment/frag.py ~ net/pf_fragment/frag6.py ~ net/pf_fragment/frag6_cutnew.py ~ net/pf_fragment/frag6_cutold.py ~ net/pf_fragment/frag6_dropnew.py ~ net/pf_fragment/frag6_dropold.py ~ net/pf_fragment/frag6_ext.py ~ net/pf_fragment/frag_cutnew.py ~ net/pf_fragment/frag_cutold.py ~ net/pf_fragment/frag_dropnew.py ~ net/pf_fragment/frag_dropold.py ~ net/pf_fragment/ping6_mtu_1300.py ~ net/pf_state/challenge_ack.py > By removing all the sleeps I created a race in the pf tests. The > packet was sent and its reply received before the sniffer was up > and running. So sleep a second after starting the sniffer thread > but before sending the packet. With pf_forward 0m51.54s, pf_fragment > 1m33.59s, pf_state 0m03.64s the test are still faster than with > forking. (bluhm@) ~ net/pf_forward/ping6_mtu.py ~ net/pf_forward/ping_mtu.py ~ net/pf_fragment/ping6_mtu_1300.py ~ net/pf_fragment/ping_mtu_1300.py > During sniffing filter strictly on icmp6 "packet too big" to avoid > that "neighbor discovery" confuses the test. (bluhm@) usr.bin ~ mandoc/mdoc/Bl/Makefile ~ mandoc/mdoc/Bl/tag.in ~ mandoc/mdoc/Bl/tag.out_ascii + mandoc/mdoc/Bl/tag.out_lint > test .Bl -tag without -width; > related to mandoc/mdoc_term.c rev. 1.231 and to textproc/groff-1.22.3p3 > (schwarze@) ~ mandoc/db/out/twosect.dout > adjust after the ordering was made well-defined in mandocdb.c rev. 1.180 > (schwarze@) == sbin ============================================================== 09/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin bioctl ~ bioctl.c > optarg is declared in unistd.h and usage is __dead. From Jan Stary. > ok natano (tb@) ifconfig ~ brconfig.c ~ ifconfig.c > Use strtoull() to read the datapath id and expect "datapath" instead of > "datapathid" as stated by the man page. > ok reyk@ (rzalamena@) mknod ~ mknod.c > delete useless call to setlocale(3); > patch from Jan Stary <hans at stare dot cz>; > OK millert@ natano@ (schwarze@) ping ~ ping.c > Only drop to _ping user if invoked as root. If invoked as an > unprivileged user just shed the setuid privs as previously. > sthen@ pointed out that you cannot kill(2) your own ping. > With this we rely on pledge(2) for unprivileged users. > OK sthen@, deraadt@ is happy with the middle ground (florian@) sysctl ~ sysctl.c > Rewrite the kern.malloc.kmemstat in the same style as previous. This one > does > not overflow because the initial count includes empty slots; ok tb@ (otto@) == share ============================================================= 10/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share man ~ man1/gcc-local.1 > Delete URL to dead site. (deraadt@) ~ man9/mbuf.9 > "send/received" -> "sent/received" > ok jmc@ (lteo@) ~ man5/malloc.conf.5 > R option doesn't compact memory at all; flags are not only useful for > tetsing and debugging; mult-threaded programs use multiple pools (otto@) ~ man9/style.9 > Use "continue;" instead of a bare ";" in the for() loop example > with no body. (millert@) ~ man9/style.9 > Don't tell the programmer not to put variable declarations inside > blocks. OK guenther@ kettenis@ (millert@) ~ man5/mk.conf.5 > GLOBAL_AUTOCONF_CACHE is the name of a file, not 'yes'. Leftover > from KERBEROS5 removal, spotted by Raf Czlonka. (tb@) ~ man4/man4.i386/Makefile + man4/man4.i386/vmm.4 > man pages for vmm(4) i386 (mlarkin@) ~ man4/man4.i386/vmm.4 > add openbsd id; (jmc@) ~ man9/Makefile + man9/cpumem_get.9 > first cut at a man page for cpumem (dlg@) ~ man9/cpumem_get.9 > tweak previous; (jmc@) ~ man9/cpumem_get.9 > provos didnt write any of this (dlg@) ~ man9/Makefile + man9/counters_alloc.9 > first stab at describing the per cpu counters api (dlg@) ~ man9/counters_alloc.9 ~ man9/cpumem_get.9 > more tweaks; ok dlg (jmc@) ~ man9/cpumem_get.9 > Add missing dot at beginning of macro line. (bluhm@) zoneinfo ~ datfiles/asia ~ datfiles/australasia ~ datfiles/europe ~ datfiles/northamerica > Update to tzdata2016h from from ftp.iana.org. (millert@) == sys =============================================================== 11/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/alpha/alpha ~ process_machdep.c > The first argument to process_domem() and its uio_procp should be curproc > ok kettenis@ (guenther@) arch/amd64/amd64 ~ vmm.c > SVM on AMD CPUs is not supported yet, so don't attach there; ok mlarkin@ > (naddy@) arch/arm/arm ~ pmap.c ~ pmap7.c > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) ~ fault.c ~ pmap7.c > If an Access Flag fault happens while we were running the kernel and > it happened on a kernel page, we need to consult the kernel pmap > instead of the current proc's pmap. Fixes panic when using tmpfs. > ok kettenis@ (patrick@) arch/arm/include ~ frame.h > If an Access Flag fault happens while we were running the kernel and > it happened on a kernel page, we need to consult the kernel pmap > instead of the current proc's pmap. Fixes panic when using tmpfs. > ok kettenis@ (patrick@) arch/arm/simplebus ~ simplebus.c + simplebusvar.h > Outsource simplebus softc definition so that it can be used by > simplebus compatible bus drivers. > ok kettenis@ (patrick@) arch/armv7/armv7 ~ platform.c > De-platformize sunxi. This removes the board initialization code that sets > up > SRAM for the EMAC. Instead we rely on u-boot setting this up for us. > ok jsg@, patrick@ (kettenis@) ~ armv7var.h ~ intr.c > Remove sunxi board IDs. (kettenis@) arch/armv7/conf ~ GENERIC ~ RAMDISK > Implement a driver for Marvell's Mbus bridge. This is the component > that lets e.g. the CPU access a USB controller or the USB controller > access the system's RAM. The bridge needs to be configured for the > devices before we can access the controller or a controller our memory. > Since it otherwise acts like a simplebus, simply attach simplebus once > we are finished. > "looks good to me" kettenis@ (patrick@) ~ GENERIC ~ RAMDISK > Attach sxitimer(4) using the fdt. (kettenis@) ~ GENERIC ~ RAMDISK > De-platformize sunxi. This removes the board initialization code that sets > up > SRAM for the EMAC. Instead we rely on u-boot setting this up for us. > ok jsg@, patrick@ (kettenis@) arch/armv7/imx ~ if_fec.c > Stop setting MIIF_AUTOTSLEEP in fec(4) > This can lead to tsleep in an interrupt context, which resulted in the > following panic: > kernel diagnostic assertion "p->p_wchan == NULL" failed: file > "/usr/src/sys/kern/kern_sched.c", line 333 > TID PPID PGRP UID S FLAGS WAIT COMMAND > 86009 0 0 0 2 0x40014200 miiaut idle0 > Reported by matthieu@ with sabre lite, analysis by guenther@ > ok kettenis@ (jsg@) arch/armv7/marvell ~ files.marvell + mvmbus.c + mvmbusvar.h > Implement a driver for Marvell's Mbus bridge. This is the component > that lets e.g. the CPU access a USB controller or the USB controller > access the system's RAM. The bridge needs to be configured for the > devices before we can access the controller or a controller our memory. > Since it otherwise acts like a simplebus, simply attach simplebus once > we are finished. > "looks good to me" kettenis@ (patrick@) arch/armv7/omap ~ omap_machdep.c > stop disabling the external pl310 l2 cache on pandaboard > The consensus is that this was done before the bugs in the cache > flushing code were fixed. The pl310 has been enabled on imx for some > time now and it works fine here on pandaboard as well. > ok patrick@ kettenis@ bmercer@ (jsg@) arch/armv7/stand/efiboot ~ conf.c ~ efiboot.c > Remove sunxi board IDs. (kettenis@) arch/armv7/sunxi ~ sunxireg.h ~ sxitimer.c > Run timers from the 24MHz; it seems the A10s doesn't have a reliable 32kHz > clock. Fixes the clock on the Olimux A10s OLinuXino-Micro board. > (kettenis@) ~ sxie.c > Stop setting MIIF_AUTOTSLEEP in sxie(4) > This can lead to tsleep in an interrupt context. > Tested by and ok kettenis@ (jsg@) ~ files.sunxi ~ sunxi.c ~ sxitimer.c > Attach sxitimer(4) using the fdt. (kettenis@) ~ sxitimer.c > Only attach on sun4i and sun5i. The timer implemented on later SoCs lacks > the 64-bit counter that we use as timecounter. Those SoCs have a generic > timer so we don't need the SoC-specific one. (kettenis@) - sun4i.c - sun7i.c - sunxi.c - sunxi_machdep.c ~ files.sunxi > De-platformize sunxi. This removes the board initialization code that sets > up > SRAM for the EMAC. Instead we rely on u-boot setting this up for us. > ok jsg@, patrick@ (kettenis@) arch/hppa/hppa ~ pmap.c > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) ~ trap.c > Change process_{domem,auxv_offset}() to take a process instead of a proc. > Make process_auxv_offset() take and release a reference of the vmspace like > process_domem() does. > ok kettenis@ (guenther@) arch/i386/conf ~ ld.script > put .openbsd.randomdata into rodata, since i386 can handle it (deraadt@) ~ GENERIC ~ Makefile.i386 ~ files.i386 > vmm(4) for i386. Userland changes forthcoming. Note that for the time > being, > i386 hosts are limited to running only i386 guests, even if the underlying > hardware supports amd64. This is a restriction I hope to lift moving > forward, > but for now please don't report problems running amd64 guests on i386 > hosts. > This was a straightforward port of the in-tree amd64 code plus the old > rotted > tree I had from last year for i386 support. Changes included converting > 64-bit > VMREAD/VMWRITE ops to 2x32-bit ops, and fixing treatment of the TSS, which > differs on i386. > ok deraadt@ (mlarkin@) arch/i386/i386 ~ conf.c ~ cpu.c ~ ipifuncs.c ~ machdep.c ~ mainbus.c ~ pmap.c ~ pmapae.c + vmm.c + vmm_support.S > vmm(4) for i386. Userland changes forthcoming. Note that for the time > being, > i386 hosts are limited to running only i386 guests, even if the underlying > hardware supports amd64. This is a restriction I hope to lift moving > forward, > but for now please don't report problems running amd64 guests on i386 > hosts. > This was a straightforward port of the in-tree amd64 code plus the old > rotted > tree I had from last year for i386 support. Changes included converting > 64-bit > VMREAD/VMWRITE ops to 2x32-bit ops, and fixing treatment of the TSS, which > differs on i386. > ok deraadt@ (mlarkin@) ~ ipifuncs.c ~ vmm.c > Fix MP builds including vmm > ok mlarkin (deraadt@) arch/i386/include ~ cpu.h ~ intrdefs.h ~ pmap.h ~ pte.h ~ specialreg.h + vmmvar.h > vmm(4) for i386. Userland changes forthcoming. Note that for the time > being, > i386 hosts are limited to running only i386 guests, even if the underlying > hardware supports amd64. This is a restriction I hope to lift moving > forward, > but for now please don't report problems running amd64 guests on i386 > hosts. > This was a straightforward port of the in-tree amd64 code plus the old > rotted > tree I had from last year for i386 support. Changes included converting > 64-bit > VMREAD/VMWRITE ops to 2x32-bit ops, and fixing treatment of the TSS, which > differs on i386. > ok deraadt@ (mlarkin@) ~ conf.h > pledge changes needed to support pledging vmd(8) on i386, forthcoming. > ok deraadt@ (mlarkin@) arch/m88k/m88k ~ pmap.c > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) ~ trap.c > Change process_{domem,auxv_offset}() to take a process instead of a proc. > Make process_auxv_offset() take and release a reference of the vmspace like > process_domem() does. > ok kettenis@ (guenther@) arch/mips64/mips64 ~ trap.c > The first argument to process_domem() and its uio_procp should be curproc > ok kettenis@ (guenther@) ~ pmap.c ~ trap.c > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) arch/octeon/dev ~ octeon_pcibus.c > Fix build: declare the pcie read/write routines before using them. > (pirofti@) ~ octeon_pcibus.c > Revert previous commit. > I was building in the wrong tree and got confused. Sorry! (pirofti@) ~ cn30xxgmx.c > Fix grammar, and simplify a little. > Pointed out by pirofti@ (visa@) ~ cn30xxgmx.c > Check if memory allocation failed. (visa@) arch/powerpc/powerpc ~ pmap.c > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) arch/sh/sh ~ pmap.c > struct process has a pointer to the vmspace now, so simplify the ASID > search > ok deraadt@ jca@ (guenther@) ~ pmap.c > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) arch/sparc64/sparc64 ~ autoconf.c ~ db_interface.c ~ locore.s ~ trap.c > Delete remnants of "traptrace" support > ok deraadt@ (guenther@) conf ~ newvers.sh > Strip trailing obj/ from kernel build directories, so kernels > are again marked with GENERIC{,.MP} RAMDISK, etc. > Problem noticed by several (jsg, semarie, ...) > ok many (sthen, natano, millert, deraadt, ...) > Explanations why quotes aren't necessary by even more. Thanks! (tb@) ~ files > add generalised access to per cpu data structures and counters. > both the cpumem and counters api simply allocates memory for each cpu in > the system that can be used for arbitrary per cpu data (via cpumem), or > a versioned set of counters per cpu (counters). > there is an alternate backend for uniprocessor systems that basically > turns the percpu data access into an immediate access to a single > allocation. > there is also support for percpu data structures that are available at > boot time by providing an allocation for the boot cpu. after autoconf, > these allocations have to be resized to provide for all cpus that were > enumerated by boot. > ok mpi@ (dlg@) ddb ~ db_examine.c > Add the examine /m format, "display in unsigned hex with character > dump at the end of each line", which has always been documented but > not implemented. > Mostly from NetBSD, but I fixed the column alignment code. > ok mlarkin@, pirofti@ (naddy@) dev ~ rnd.c > when openbsd.randomdata was made readonly, the proto seed copy got done > too late, leading to poor rng in the kernel early on. a behavioural > artifact in vmm spotted the issue. > ok tedu guenther mlarkin (deraadt@) dev/acpi ~ dsdt.c > Add support for the ACPI timer opcode. > ok kettenis@ (joris@) dev/pci ~ if_iwm.c > Only force iwm's firmware to use RTS for all frames if the AP requires it. > RTS is still used when sending long frames regardless of this change. > Tested by myself and Jan Stary. (stsp@) ~ mpii.c > mitigate against mpii device attachment racing with mountroot > mpii devices get attached by processing sas discovery events after > threads are running, but if you have a lean (in terms of devices) > and fast machine, these events might run after mountroot. > this change has mpii use config_pending_incr() before it attaches > scsibus, and runs config_pending_decr() after the first scan of the > bus is complete. > to make sure config_pending_decr() is run after the scsi attach/detach > events are processed, the MPII_EVENT_SAS_DISCOVERY events are > deferred to the same context that the MPII_EVENT_SAS_TOPOLOGY_CHANGE_LIST > events run in. > ok jmatthew@ (dlg@) dev/pckbc ~ pms.c > Improved parameter handling in wsmouse and new ioctls for reading and > setting parameter values. > ok matthieu@ (bru@) dev/wscons ~ wsconsio.h ~ wsmouse.c ~ wsmouseinput.h ~ wsmousevar.h > Improved parameter handling in wsmouse and new ioctls for reading and > setting parameter values. > ok matthieu@ (bru@) kern ~ sys_process.c > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) ~ sys_process.c > Change process_{domem,auxv_offset}() to take a process instead of a proc. > Make process_auxv_offset() take and release a reference of the vmspace like > process_domem() does. > ok kettenis@ (guenther@) ~ init_main.c + subr_percpu.c > add generalised access to per cpu data structures and counters. > both the cpumem and counters api simply allocates memory for each cpu in > the system that can be used for arbitrary per cpu data (via cpumem), or > a versioned set of counters per cpu (counters). > there is an alternate backend for uniprocessor systems that basically > turns the percpu data access into an immediate access to a single > allocation. > there is also support for percpu data structures that are available at > boot time by providing an allocation for the boot cpu. after autoconf, > these allocations have to be resized to provide for all cpus that were > enumerated by boot. > ok mpi@ (dlg@) ~ subr_percpu.c > consistently zero the allocated memory in both the MP and UP cases. > from markus@ (dlg@) ~ kern_pledge.c > pledge changes needed to support pledging vmd(8) on i386, forthcoming. > ok deraadt@ (mlarkin@) ~ kern_exec.c > Delete dead copy of pr->ps_vmspace; uvmspace_exec() can change it anyway > ok kettenis@ jsing@ (guenther@) ~ kern_fork.c > Adjust allocpid() to take into account lastpid > ok jsing@ kettensi@ (guenther@) ~ kern_sysctl.c > Factor out pr->ps_vmspace into a local variable for fill_kproc() > ok jsing@ kettenis@ (guenther@) ~ uipc_syscalls.c > dns hijacking must be af specific. move it into the port check function, > and redirect inet6 sockets to the ::1 flavor of localhost. (tedu@) ~ uipc_syscalls.c > handle non-INET6 kernels in some way (deraadt@) ~ uipc_syscalls.c > unbreak by fixing obvious pastos (naddy@) net ~ bpf.c > Fix bpf_catchpacket comment. (jca@) ~ if.c > m_resethdr() clears information attached to a mbuf that has been > accumulated during processing, especially pf(4) data is removed. > When a packet is reinserted to local input processing, we have to > start from scratch with a clean mbuf header. Also the packet has > to be in the routing doamin of the interface where it is inserted. > Incoming packet with stale inp pointer seen by patrick@ on lo(4); > OK sashan@ mpi@ (bluhm@) ~ rtsock.c > Place splsoftnet() in route_input() to shut up splassert warnings > in sorwakeup(). > Report from egorenar-dev at posteo.net; OK henning@ (bluhm@) ~ pf.c > The checksum of a ICMP "need to frag" packet for TCP was wrong when > created from a ICMP6 "too big" packet. The function pf_change_icmp_af() > has code to adjust the pseudo-header checksum in the ICMP6 case, > but pf_test_state_icmp() changed the proto before the case was > entered. > So call pf_change_icmp_af() before the pd->proto is converted in > the TCP and UDP payload case like it was already done for ICMP and > ICMP6 payload. > Found by sys/net/pf_forward regress test; OK henning@ (bluhm@) ~ pf.c > The variable dlen is always positive and d may be negative. So > declare both with the correct sign in pf_change_icmp_af(). > OK henning@ (bluhm@) ~ pf.c ~ pfvar.h > split pf_send_tcp() into the part that builds the mbuf and the actual > sending, needed soon. ok sashan mikeb lteo (henning@) ~ if.c > Remove the if_input task from the correct taskq. > ok bluhm henning (yasuoka@) ~ switchofp.c > Lets just be safe and check M_PREPEND return values for functions that > madj() and injects back the layer 2/3 contents. > ok lteo@ (rzalamena@) ~ pf.c > - PF should send a challenge ACK as response to SYN, which matches existing > state. Extra thanks goes to bluhm@ for careful testing and fixing patch > I've > sent to tech@ > O.K. henning@ bluhm@ (sashan@) ~ switchofp.c > tag large constants as ULL to fix the build of switch(4) on ILP32 archs > ok jca@ deraadt@ (jsg@) ~ switchofp.c > Don't free mbuf on error during swofp_flow_entry_put_instructions() parse, > otherwise we won't have the request ofp_header to send in the error > message. > ok jca@ (rzalamena@) sys ~ mbuf.h > "send/received" -> "sent/received" > ok jmc@ (lteo@) ~ ptrace.h > Change process_{domem,auxv_offset}() to take a process instead of a proc. > Make process_auxv_offset() take and release a reference of the vmspace like > process_domem() does. > ok kettenis@ (guenther@) ~ srp.h + percpu.h > add generalised access to per cpu data structures and counters. > both the cpumem and counters api simply allocates memory for each cpu in > the system that can be used for arbitrary per cpu data (via cpumem), or > a versioned set of counters per cpu (counters). > there is an alternate backend for uniprocessor systems that basically > turns the percpu data access into an immediate access to a single > allocation. > there is also support for percpu data structures that are available at > boot time by providing an allocation for the boot cpu. after autoconf, > these allocations have to be resized to provide for all cpus that were > enumerated by boot. > ok mpi@ (dlg@) uvm ~ uvm_pmap.h > Change pmap_proc_iflush() to take a process instead of a proc > powerpc: rename second argument of pmap_proc_iflush() to match other archs > ok kettenis@ (guenther@) == usr.bin =========================================================== 12/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin column ~ column.1 > Add missing information about the handling of input, output, and > character encoding, clarify the description of table mode, and say > that columns are filled before rows by default. > Feedback an OK jmc@, OK martijn@. (schwarze@) cvs ~ diff3.c > Remove useless n=n assignment in the first parse of a for(). > It was already removed from the rcs version by nicm@. (millert@) ~ status.c > Add the branch revision to the status output much like GNU cvs does. > OK millert@ (joris@) ~ diff3.c > Using bitwise OR along with two assignments in the conditional of > a while() loop is a trap for the unwary programmer (albeit a clever > trap). Break this up into two separate assignments and using boolean > OR for clarity. OK otto@ (millert@) diff3 ~ diff3prog.c > Remove useless n=n assignment in the first parse of a for(). > It was already removed from the rcs version by nicm@. (millert@) ~ diff3prog.c > Using bitwise OR along with two assignments in the conditional of > a while() loop is a trap for the unwary programmer (albeit a clever > trap). Break this up into two separate assignments and using boolean > OR for clarity. OK otto@ (millert@) make ~ arch.c ~ compat.c ~ cond.c ~ dir.c ~ direxpand.c ~ dump.c ~ engine.c ~ lowparse.c ~ main.c ~ make.c ~ parse.c ~ suff.c ~ targ.c ~ targequiv.c > small obvious cleanups: > - remove a lot of unnecessary casts > - zap extra param that's no longer needed > - add proper prototype and make function static > okay natano@ (espie@) ~ make.1 ~ parse.c ~ parsevar.c ~ var.c ~ var.h > add !!= > feedback and okay millert@ (espie@) ~ make.1 > zap trailing whitespace; (jmc@) mandoc ~ mandocdb.c > Simplify and correct support for reproducible builds, such that database > entries come in a well-defined order even in the presence of MLINKS. > Do this by using the compar() argument of fts_open(3) rather than > trying to sort later, which missed some cases. > This also shortens the code by a few lines. > Diff from Ed Maste <emaste @ FreeBSD>, adapted to our tree > and tweaked a bit by me, final version confirmed by Ed. (schwarze@) ~ mandocdb.c > Fix a comment: We don't have manpath(1). No code change. (schwarze@) ~ dbm.c > The termination condition of the iteration logic in page_bymacro() > was overzealous. Consequently, macro=substr and macro~regexp searches > only returned all pages containing the first matching macro value, > rather than all pages containing any of the matching macro values. > Bug reported by tb@ - thanks! (schwarze@) ~ man.1 > Avoid the confusion that -a might require a section argument; it does not. > Triggered by a question from <jmates at ee dot washington at edu>. > OK jmc@. (schwarze@) ~ dbm_map.c > st_size is off_t, which is signed, so use %lld rather than %llu; > noticed by Ed Maste when compiling on FreeBSD (schwarze@) ~ mandocdb.c > 0x%llu is a bad idea, make that 0x%llx; > noticed while investigating a report from Ed Maste (schwarze@) ~ dbm_map.c ~ mandocdb.c > cast off_t and uint64_t for printf(3) > to make sure that widths match on all platforms; > from Ed Maste <emaste at freebsd dot org>; > OK guenther@ (schwarze@) rcs ~ buf.c ~ diff.c ~ diff3.c ~ rlog.c > Quiet compiler warnings; as applied to usr.bin/cvs by millert@ > ok millert@ (okan@) ~ diff3.c > Using bitwise OR along with two assignments in the conditional of > a while() loop is a trap for the unwary programmer (albeit a clever > trap). Break this up into two separate assignments and using boolean > OR for clarity. OK otto@ (millert@) sendbug ~ sendbug.c > Use the files in /var/db/acpi instead of trying to run acpidump. > ok brynet@ (kettenis@) sndiod ~ dsp.c ~ dsp.h > remove unused variables (ratchov@) ~ sndiod.c > Move initialization of the helper process in its own routine to make > code more readable. No bahavior change. (ratchov@) sort ~ file.c ~ sort.1 > /var/tmp -> /tmp > From Jan Stary. > "POSIX doesn't restrict us here" and ok schwarze@ (lteo@) ssh ~ sftp.c > Install a signal handler for tty-generated signals and wait for the > ssh child to suspend before suspending sftp. This lets ssh restore > the terminal mode as needed when it is suspended at the password > prompt. OK dtucker@ (millert@) ~ channels.c ~ channels.h > Remove channel_input_port_forward_request(); the only caller was the > recently-removed SSH1 server code so it's now dead code. ok markus@ > (dtucker@) ~ mux.c > When tearing down ControlMaster connecctions, don't pollute stderr when > LogLevel=quiet. Patch from Tim Kuijsten via tech@. (dtucker@) ~ misc.c ~ misc.h ~ readconf.c ~ serverloop.c > Factor out "can bind to low ports" check into its own function. This will > make > it easier for Portable to support platforms with permissions models other > than > uid==0 (eg bz#2625). ok djm@, "doesn't offend me too much" deraadt@. > (dtucker@) tmux ~ cfg.c ~ cmd-attach-session.c ~ cmd-command-prompt.c ~ cmd-confirm-before.c ~ cmd-copy-mode.c ~ cmd-display-panes.c ~ cmd-find.c ~ cmd-if-shell.c ~ cmd-load-buffer.c ~ cmd-new-session.c ~ cmd-new-window.c ~ cmd-queue.c ~ cmd-resize-pane.c ~ cmd-run-shell.c ~ cmd-send-keys.c ~ cmd-source-file.c ~ cmd-split-window.c ~ cmd-wait-for.c ~ cmd.c ~ control.c ~ format.c ~ hooks.c ~ key-bindings.c ~ notify.c ~ server-client.c ~ server.c ~ tmux.1 ~ tmux.h ~ window-choose.c > Rewrite command queue handling. Each client still has a command queue, > but there is also now a global command queue. Instead of command queues > being dispatched on demand from wherever the command happens to be > added, they are now all dispatched from the top level server > loop. Command queues may now also include callbacks as well as commands, > and items may be inserted after the current command as well as at the end. > This all makes command queues significantly more predictable and easier > to use, and avoids the complex multiple nested command queues used by > source-file, if-shell and friends. > A mass rename of struct cmdq to a better name (cmdq_item probably) is > coming. (nicm@) ~ cfg.c ~ cmd-attach-session.c ~ cmd-bind-key.c ~ cmd-break-pane.c ~ cmd-capture-pane.c ~ cmd-choose-buffer.c ~ cmd-choose-client.c ~ cmd-choose-tree.c ~ cmd-clear-history.c ~ cmd-command-prompt.c ~ cmd-confirm-before.c ~ cmd-copy-mode.c ~ cmd-detach-client.c ~ cmd-display-message.c ~ cmd-display-panes.c ~ cmd-find-window.c ~ cmd-find.c ~ cmd-if-shell.c ~ cmd-join-pane.c ~ cmd-kill-pane.c ~ cmd-kill-server.c ~ cmd-kill-session.c ~ cmd-kill-window.c ~ cmd-list-buffers.c ~ cmd-list-clients.c ~ cmd-list-keys.c ~ cmd-list-panes.c ~ cmd-list-sessions.c ~ cmd-list-windows.c ~ cmd-load-buffer.c ~ cmd-lock-server.c ~ cmd-move-window.c ~ cmd-new-session.c ~ cmd-new-window.c ~ cmd-paste-buffer.c ~ cmd-pipe-pane.c ~ cmd-queue.c ~ cmd-refresh-client.c ~ cmd-rename-session.c ~ cmd-rename-window.c ~ cmd-resize-pane.c ~ cmd-respawn-pane.c ~ cmd-respawn-window.c ~ cmd-rotate-window.c ~ cmd-run-shell.c ~ cmd-save-buffer.c ~ cmd-select-layout.c ~ cmd-select-pane.c ~ cmd-select-window.c ~ cmd-send-keys.c ~ cmd-set-buffer.c ~ cmd-set-environment.c ~ cmd-set-hook.c ~ cmd-set-option.c ~ cmd-show-environment.c ~ cmd-show-messages.c ~ cmd-show-options.c ~ cmd-source-file.c ~ cmd-split-window.c ~ cmd-swap-pane.c ~ cmd-swap-window.c ~ cmd-switch-client.c ~ cmd-unbind-key.c ~ cmd-wait-for.c ~ cmd.c ~ control.c ~ format.c ~ hooks.c ~ key-bindings.c ~ notify.c ~ server-client.c ~ tmux.h ~ window-choose.c > Mass rename struct cmd_q to struct cmdq_item and related. (nicm@) ~ notify.c ~ server.c ~ tmux.h > Notifys can go via the command queue instead of using their own queue. > (nicm@) ~ cmd-queue.c ~ format.c ~ hooks.c ~ notify.c ~ tmux.h > Provide a way for hooks to tag formats onto the commands they fire so > that the user can get at additional information - now used for the > "hook" format, more to come. (nicm@) ~ notify.c ~ tmux.1 > Add hook_session and hook_window formats to get information on the > affected session or window when a hook fires. Enable session-created and > session-closed hooks now that that is available. (nicm@) ~ cmd-attach-session.c ~ cmd-find.c ~ cmd-join-pane.c ~ cmd-new-session.c ~ cmd-new-window.c ~ cmd-rename-session.c ~ cmd-split-window.c ~ control-notify.c ~ layout-custom.c ~ layout.c ~ notify.c ~ resize.c ~ server-client.c ~ server-fn.c ~ session.c ~ tmux.1 ~ tmux.h ~ window.c > Use the notify name string instead of going via an enum and change > existing hooks to use notifys instead. (nicm@) ~ alerts.c ~ notify.c ~ tmux.h > Use notifys for alerts too. (nicm@) ~ window.c > Store the right size in the pipe offset for pipe-pane. (nicm@) ~ screen-write.c > Zero dirty count after flushing. (nicm@) ~ cmd-queue.c ~ tmux.h > Give each item on queue a name for better logging. (nicm@) ~ cmd-find.c > Tweak a couple of log statements. (nicm@) ~ grid.c > Make grid_clear_cell set up the entry properly for 256 and RGB cells. > (nicm@) ~ tmux.1 > xterm-keys was in the wrong place in the list; Dilyan Palauzov. (nicm@) ~ grid.c > Clear cell entry with grid_default_entry not grid_default_cell. (nicm@) ~ cmd-new-session.c ~ session.c > Move session-create hook out of session_create so it works with grouped > sessions. (nicm@) ~ alerts.c ~ server-client.c ~ session.c ~ tmux.h ~ window.c > Alerts are too slow, so rather than walking all sessions and windows, > add a link of winlinks to each window and a pointer to the session to > each winlink. Also rewrite the alerts processing to return to the old > behaviour (alert in any window sets the flag on any winlink). (nicm@) ~ key-bindings.c > Do not have a default binding for C-b in copy-mode-vi or it conflicts > with the default prefix. Reported by natano@. (nicm@) ~ cmd.c ~ key-bindings.c ~ tmux.1 > Add %%% to substitute with quotes escaped (convert " to \"). Use this > for the prompts in copy mode. Fixes problems with jumping to ' reported > by Theo Buehler. (nicm@) ul ~ ul.c > Zap unused variable. (jca@) == usr.sbin ========================================================== 13/13 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin usr.sbin ~ Makefile > enable makefs (deraadt@) bgpd ~ rde.c ~ bgpd.conf.5 > "Terminate the session after number prefixes have been received" means > that max-prefix 1 will close the session on the first prefix which is > off by one to what one would expect. Change it to No. received > > max-prefix, which is also in line with other implementations. > Found by jbg@ and deraadt@. > ok phessler@ deraadt@ henning@ claudio@ (benno@) config ~ main.c > Stop complaining about changed kernel options when called for an empty > directory. With the new kernel obj mechanism config always prints > Kernel options have changed -- you must run "make clean" > on the first run. The message is suppressed when the compile directory > didn't exist before, but now the directory always exists when config is > run, so we have to use something else to detect if this is a "fresh" run > or an update of an existing directory. Use the options file instead, > which is used to track the kernel options of the last config run. If the > file doesn't exist, it is safe to assume, that this is a "fresh" > directory. > ok tb deraadt (natano@) ~ main.c > Remove declarations for optarg and optind. They are declared in <unistd.h>. > ok tb (natano@) ~ config.h ~ util.c > Add printf format attributes to the custom error functions. > ok natano (tb@) ~ main.c > Remove the -g option. It is obsolete/undocumented since the initial > import from NetBSD in '95 and does nothing except print an error. > ok tb (natano@) ~ main.c ~ mkheaders.c ~ mkioconf.c ~ mkmakefile.c ~ mkswap.c ~ ukc.c > Use the err(3) family of functions more consistently. > ok jca, improvements and ok millert, natano (tb@) ~ main.c > Add a == EOF that was accidentally lost in previous commit. (tb@) ~ ukc.c > There used to be disabled code that used /dev/tty to determine the > terminal size. That code was removed a while ago, and now there is > no longer any reason to open /dev/tty right before exiting init(). > ok natano (tb@) dhcpd ~ sync.c > - Use memset(*b, 0, len) instead of bzero(*b, len) > - Use explicit_bzero(3) instead of bzero(3) to clean temporary HMAC > OK krw@ (mestre@) ldapd ~ ldapd.conf.5 > Fix a few mistakes and add a FILES entry for /etc/ldap/certs > From Rob Pierce, help & ok jmc@ (jca@) ~ ldapd.conf.5 > In FILES mention /etc/ldapd.conf first, then /etc/ldap/ entries (jca@) ~ ldapd.conf.5 > Tweak /etc/ldap/certs/ FILES entry. (jca@) makefs + Makefile + README + cd9660.c + cd9660.h + ffs.c + ffs.h + makefs.8 + makefs.c + makefs.h + msdos.c + msdos.h + walk.c + xmalloc.c + cd9660/cd9660_archimedes.c + cd9660/cd9660_archimedes.h + cd9660/cd9660_conversion.c + cd9660/cd9660_debug.c + cd9660/cd9660_eltorito.c + cd9660/cd9660_eltorito.h + cd9660/cd9660_strings.c + cd9660/cd9660_write.c + cd9660/iso9660_rrip.c + cd9660/iso9660_rrip.h + ffs/buf.c + ffs/buf.h + ffs/ffs_alloc.c + ffs/ffs_balloc.c + ffs/ffs_extern.h + ffs/mkfs.c + ffs/newfs_extern.h + ffs/ufs_bmap.c + ffs/ufs_inode.h + fs/cd9660/cd9660_rrip.h + fs/cd9660/iso.h + fs/cd9660/iso_rrip.h + fs/msdosfs/bootsect.h + fs/msdosfs/bpb.h + fs/msdosfs/clock_subr.c + fs/msdosfs/clock_subr.h + fs/msdosfs/denode.h + fs/msdosfs/direntry.h + fs/msdosfs/fat.h + fs/msdosfs/msdosfs_conv.c + fs/msdosfs/msdosfs_fat.c + fs/msdosfs/msdosfs_lookup.c + fs/msdosfs/msdosfs_unicode.c + fs/msdosfs/msdosfsmount.h + msdos/msdosfs_denode.c + msdos/msdosfs_vfsops.c + msdos/msdosfs_vnops.c + newfs_msdos/mkfs_msdos.c + newfs_msdos/mkfs_msdos.h + sys/bootblock.h + sys/clock.h + sys/quota.h + ufs/ffs/ffs_bswap.c + ufs/ffs/ffs_extern.h + ufs/ffs/ffs_subr.c + ufs/ffs/ffs_tables.c + ufs/ffs/fs.h + ufs/ufs/dinode.h + ufs/ufs/dir.h + ufs/ufs/quota.h + ufs/ufs/ufs_bswap.h > Import makefs - a tool to create filesystem images from a directory. > This is a rough port of the NetBSD tool with some features removed we > don't need. It compiles, but I don't promise anything more. Importing > now, so we can hack on it in tree. > The supported filesystem types are: cd9660, ffs and msdosfs. > ok deraadt (natano@) - fs/msdosfs/clock_subr.c - fs/msdosfs/clock_subr.h - sys/clock.h ~ Makefile ~ msdos.c ~ fs/msdosfs/denode.h ~ fs/msdosfs/direntry.h ~ fs/msdosfs/msdosfs_conv.c ~ fs/msdosfs/msdosfs_lookup.c ~ fs/msdosfs/msdosfsmount.h ~ msdos/msdosfs_vfsops.c ~ msdos/msdosfs_vnops.c > Replace fs/msdosfs/{msdosfs_conv.c,direntry.h} with stripped-down > versions from our own tree. This allows to remove the clock_subr.c > tentacle. (natano@) ~ ffs/ffs_alloc.c ~ ffs/ffs_extern.h ~ fs/msdosfs/denode.h ~ fs/msdosfs/msdosfs_lookup.c ~ ufs/ffs/ffs_extern.h ~ ufs/ffs/ffs_subr.c > Remove some unused functions. Found by cppcheck from ports. (natano@) ~ cd9660.c ~ ffs.c ~ makefs.c ~ msdos.c ~ walk.c ~ cd9660/cd9660_archimedes.c ~ cd9660/cd9660_debug.c ~ cd9660/cd9660_eltorito.c ~ cd9660/cd9660_strings.c ~ cd9660/cd9660_write.c ~ cd9660/iso9660_rrip.c ~ ffs/ffs_alloc.c ~ ffs/ffs_balloc.c > Unused includes, as reported by include-what-you-use from ports. (natano@) ~ Makefile ~ README ~ cd9660.c ~ cd9660.h ~ ffs.c ~ ffs.h ~ makefs.8 ~ makefs.c ~ makefs.h ~ msdos.c ~ msdos.h ~ walk.c ~ xmalloc.c ~ cd9660/cd9660_archimedes.c ~ cd9660/cd9660_archimedes.h ~ cd9660/cd9660_conversion.c ~ cd9660/cd9660_debug.c ~ cd9660/cd9660_eltorito.c ~ cd9660/cd9660_eltorito.h ~ cd9660/cd9660_strings.c ~ cd9660/cd9660_write.c ~ cd9660/iso9660_rrip.c ~ cd9660/iso9660_rrip.h ~ ffs/buf.c ~ ffs/buf.h ~ ffs/ffs_alloc.c ~ ffs/ffs_balloc.c ~ ffs/ffs_extern.h ~ ffs/mkfs.c ~ ffs/newfs_extern.h ~ ffs/ufs_bmap.c ~ ffs/ufs_inode.h ~ fs/cd9660/cd9660_rrip.h ~ fs/cd9660/iso.h ~ fs/cd9660/iso_rrip.h ~ fs/msdosfs/bootsect.h ~ fs/msdosfs/bpb.h ~ fs/msdosfs/denode.h ~ fs/msdosfs/fat.h ~ fs/msdosfs/msdosfs_fat.c ~ fs/msdosfs/msdosfs_lookup.c ~ fs/msdosfs/msdosfs_unicode.c ~ fs/msdosfs/msdosfsmount.h ~ msdos/msdosfs_denode.c ~ msdos/msdosfs_vfsops.c ~ msdos/msdosfs_vnops.c ~ newfs_msdos/mkfs_msdos.c ~ newfs_msdos/mkfs_msdos.h ~ sys/bootblock.h ~ sys/quota.h ~ ufs/ffs/ffs_bswap.c ~ ufs/ffs/ffs_extern.h ~ ufs/ffs/ffs_subr.c ~ ufs/ffs/ffs_tables.c ~ ufs/ffs/fs.h ~ ufs/ufs/dinode.h ~ ufs/ufs/dir.h ~ ufs/ufs/quota.h ~ ufs/ufs/ufs_bswap.h > Add OpenBSD RCS tags; reminded by tb (natano@) ~ ffs.c ~ makefs.8 ~ makefs.c ~ makefs.h > Remove the -Z (spare ffs image) option. (natano@) ~ makefs.8 ~ makefs.c ~ makefs.h ~ walk.c > Only allow one directory as argument and remove the -r option. > ok deraadt (on principle) (natano@) ~ Makefile > The MSDOS_EI define isn't used anywhere. (natano@) ~ fs/msdosfs/fat.h ~ fs/msdosfs/msdosfsmount.h ~ msdos/msdosfs_vfsops.c > remove GEMDOS code (tedu@) ~ cd9660.c ~ ffs.c > /* XXX bounds checking! */ > very crudely fix a few strcpy and sprintf warnings. > leave the quality warnings. (tedu@) ~ cd9660.c > there's no need for a special typedef and function pointer for picking > between two functions. (tedu@) ~ ffs.c ~ makefs.c ~ makefs.h ~ ffs/mkfs.c ~ ufs/ufs/dir.h > remove some byte swapping code from the top end (tedu@) ~ ffs/ffs_alloc.c ~ ffs/ffs_balloc.c ~ ufs/ffs/ffs_subr.c ~ ufs/ufs/ufs_bswap.h > another round of deswapping (tedu@) ~ ufs/ufs/ufs_bswap.h > simplify swapper functions a bit (tedu@) ~ ffs.c ~ ffs/ffs_alloc.c ~ ufs/ffs/ffs_subr.c ~ ufs/ufs/ufs_bswap.h > remove swapping add macros (tedu@) ~ ffs/buf.h ~ fs/msdosfs/denode.h ~ fs/msdosfs/msdosfs_lookup.c ~ msdos/msdosfs_denode.c ~ msdos/msdosfs_vnops.c > we're not going to be using kauth_creds (tedu@) - sys/bootblock.h ~ cd9660/cd9660_eltorito.c > Only a few apple-specific bits are needed by eltorito, so copy them local. > (deraadt@) ~ ffs.c ~ msdos.c ~ msdos.h ~ ffs/buf.c ~ ffs/buf.h ~ ffs/ffs_alloc.c ~ ffs/ffs_balloc.c ~ ffs/ffs_extern.h ~ ffs/ufs_inode.h ~ fs/msdosfs/denode.h ~ fs/msdosfs/fat.h ~ fs/msdosfs/msdosfs_conv.c ~ fs/msdosfs/msdosfs_fat.c ~ fs/msdosfs/msdosfs_lookup.c ~ fs/msdosfs/msdosfsmount.h ~ msdos/msdosfs_denode.c ~ msdos/msdosfs_vfsops.c ~ msdos/msdosfs_vnops.c ~ ufs/ffs/ffs_extern.h > mechanical rename of vnode to mkfsvnode and buf to mkfsbuf to avoid > collisions with the kernel structures of the same name. > caught one bug where the wrong header was being included. (tedu@) - ufs/ffs/ffs_bswap.c ~ Makefile > Everything in ffs_bswap.c is unused. (natano@) ~ makefs.8 ~ makefs.c > Remove the -d option. > -d debug-mask > Enable various levels of debugging, depending upon which bits are > set in debug-mask. XXX: document these > Useful, eh? Leaving the actual debug code in there for now, so debug > flags can be set at compile time. That might be removed in the future > too. (natano@) ~ ffs.c ~ makefs.h ~ walk.c > Remove "feature" defines. This code won't try to be portable when we are > done hacking it. (natano@) - cd9660/cd9660_archimedes.c - cd9660/cd9660_archimedes.h ~ Makefile ~ cd9660.c ~ cd9660.h ~ makefs.8 > Remove cd9660's archimedes option, we won't need RISC OS metadata on our > ISO images. (natano@) ~ cd9660.c ~ cd9660.h > Remove some more unused cd9660 options. (natano@) - ufs/ffs/ffs_extern.h ~ ffs/ffs_extern.h ~ ufs/ffs/ffs_subr.c > Dedup ffs_extern.h. (natano@) - sys/quota.h - ufs/ufs/quota.h ~ msdos/msdosfs_vfsops.c > makefs doesn't care about quotas. (natano@) ~ ufs/ffs/fs.h ~ ufs/ufs/dinode.h ~ ufs/ufs/dir.h > Remove apple ufs bits. (natano@) ~ makefs.c ~ msdos/msdosfs_vfsops.c > unused variables (natano@) - fs/cd9660/cd9660_rrip.h - fs/cd9660/iso.h - fs/cd9660/iso_rrip.h ~ Makefile ~ cd9660.h ~ cd9660/iso9660_rrip.h + cd9660/cd9660_rrip.h + cd9660/iso.h + cd9660/iso_rrip.h > Merge fs/cd9660/ into cd9660. (natano@) - fs/msdosfs/bootsect.h - fs/msdosfs/bpb.h - fs/msdosfs/denode.h - fs/msdosfs/direntry.h - fs/msdosfs/fat.h - fs/msdosfs/msdosfs_conv.c - fs/msdosfs/msdosfs_fat.c - fs/msdosfs/msdosfs_lookup.c - fs/msdosfs/msdosfs_unicode.c - fs/msdosfs/msdosfsmount.h - newfs_msdos/mkfs_msdos.c - newfs_msdos/mkfs_msdos.h ~ Makefile ~ msdos.c ~ msdos/msdosfs_denode.c ~ msdos/msdosfs_vfsops.c ~ msdos/msdosfs_vnops.c + msdos/bootsect.h + msdos/bpb.h + msdos/denode.h + msdos/direntry.h + msdos/fat.h + msdos/mkfs_msdos.c + msdos/mkfs_msdos.h + msdos/msdosfs_conv.c + msdos/msdosfs_fat.c + msdos/msdosfs_lookup.c + msdos/msdosfs_unicode.c + msdos/msdosfsmount.h > Merge fs/msdosfs/ and newfs_msdos/ into msdos/. (natano@) - ufs/ffs/ffs_subr.c - ufs/ffs/ffs_tables.c - ufs/ffs/fs.h - ufs/ufs/dinode.h - ufs/ufs/dir.h - ufs/ufs/ufs_bswap.h ~ Makefile ~ ffs.c ~ ffs.h ~ ffs/ffs_alloc.c ~ ffs/ffs_balloc.c ~ ffs/mkfs.c ~ ffs/ufs_bmap.c + ffs/dinode.h + ffs/dir.h + ffs/ffs_subr.c + ffs/ffs_tables.c + ffs/fs.h + ffs/ufs_bswap.h > Merge ufs/{ufs,ffs}/ into ffs/. (natano@) ~ makefs.8 ~ makefs.c > Remove the -B flag. Most of the byteswapping code was tedu'd already. > (natano@) ~ makefs.8 ~ makefs.c > Only allow epoch timestamps for -T. A shared option for either using the > timestamps from a file or using an epoch timestamp is ridiculous. (natano@) ~ makefs.c > Remove setprogname(). (natano@) ~ Makefile > cleanup Makefile (deraadt@) ~ ffs/mkfs.c > whitespace; no functional change (natano@) ~ ffs/mkfs.c > Remove the superfluous 'sbsize' variable & s/sbsize/SBLOCKSIZE/. (natano@) ~ ffs/mkfs.c > Use the superblock passed to ffs_write_superblock() instead of referring > to the global 'sblock' variable for some operations. (natano@) ~ ffs/mkfs.c > Avoid writing pointers to the disk (image). (natano@) ~ ffs.c > Use deterministic pseudo-random numbers when the -T flag is used; for > repeatable builds. (natano@) ~ ffs/mkfs.c > Remove unused code for 4.3BSD format filesystems. (natano@) - ffs/dinode.h - ffs/dir.h - ffs/fs.h ~ ffs.c ~ ffs.h ~ ffs/ffs_alloc.c ~ ffs/ffs_balloc.c ~ ffs/ffs_subr.c ~ ffs/ffs_tables.c ~ ffs/mkfs.c ~ ffs/ufs_bmap.c > Include our own <ufs/ufs/dinode.h>, <ufs/ufs/dir.h> and <ufs/ffs/fs.h> > header files instead of relying on copies of NetBSD's headers. This > required some changes in the .c files, due to renamed fields, functions > and preprocessor macros. Also pull in our own ffs_tables.c. As a bonus > this diff gets rid of layout compatibility issues in the superblock > (position of the fs_flags field). (natano@) ~ ffs.c ~ ffs/ffs_extern.h ~ ffs/ffs_subr.c > s/panic/errx/ This is userland. (natano@) ~ cd9660.c ~ ffs.c ~ makefs.c ~ makefs.h ~ msdos.c ~ walk.c ~ ffs/buf.c ~ ffs/ffs_alloc.c ~ msdos/msdosfs_fat.c > Remove huge amounts of debug code, that make the code nearly unreadable. > (natano@) ~ cd9660.c ~ ffs.c ~ makefs.c ~ makefs.h ~ walk.c > Clean up timestamp handling code to be less confusing. (natano@) ~ makefs.h ~ walk.c > More "debug" code that has to go. (natano@) - ffs/ufs_bswap.h ~ ffs.c ~ ffs/ffs_alloc.c ~ ffs/ffs_balloc.c ~ ffs/ffs_subr.c ~ ffs/mkfs.c ~ ffs/ufs_bmap.c > Get rid of remaining FFS byteswap function stubs. (natano@) ~ makefs.8 > various cleanup; ok natano (jmc@) ~ cd9660.c ~ cd9660.h ~ makefs.c ~ msdos/mkfs_msdos.h > -#include <stdbool.h> (natano@) - msdos/bootsect.h - msdos/bpb.h ~ msdos/msdosfs_denode.c ~ msdos/msdosfs_fat.c ~ msdos/msdosfs_lookup.c ~ msdos/msdosfs_vfsops.c ~ msdos/msdosfs_vnops.c > Dedup msdosfs/{bootsect.h,bpb.h}. (natano@) ~ cd9660.c ~ ffs.c ~ makefs.c ~ makefs.h ~ msdos.c ~ msdos/mkfs_msdos.h > Remove short option names for -o. (natano@) ~ cd9660.c ~ cd9660.h ~ makefs.8 ~ cd9660/cd9660_write.c > Remove cd9660's keep-bad-images option. (natano@) ~ msdos/mkfs_msdos.c ~ msdos/mkfs_msdos.h > Remove the msdos no_create option. (natano@) ~ cd9660.c ~ cd9660.h ~ cd9660/cd9660_eltorito.c ~ cd9660/cd9660_write.c ~ msdos/msdosfs_vfsops.c > No verbose flag for cd9660. (natano@) ntpd ~ constraint.c > Save the constraint process pid by getting the start_child() return value, > this should fix the problem with random ntpd(8) deaths. > ok deraadt@ (rzalamena@) ~ constraint.c > Check for EAGAIN on imsg_flush() return otherwise we might be failing > to send message to the child process. Do like we learned in httpd(8). > ok deraadt@ (rzalamena@) pstat ~ pstat.c > For -d, if there is no modifier, don't go down the "longformat" path. > Fixes cases like "pstat -d x ticks" on armv7. > ok jca@, deraadt@ (kettenis@) rebound ~ rebound.c > switch to a re-exec model instead of plain forking to reduce sharing. > this shuffles about some of the initialization code and consolidates all > the worker initialization in one place. > the parent process runs the monitor loop and execs workers via -W, which > then drop immediately into the worker loop. > file descriptors currently inherited across exec, which probably exceeds > safe magic levels, but fits the existing model without too many changes. > (tedu@) ~ rebound.c > listen on inet6 sockets as well. we need this because stolen inet6 sockets > can't be redirected to inet4 listeners. (tedu@) ~ rebound.c > unbreak by fixing obvious pastos (naddy@) rpc.bootparamd ~ bootparamd.c > Rename local variable 'err' to 'error', to avoid -Wshadow conflicts with > err.h (jca@) smtpd ~ smtp_session.c > Remove the "phase" state variable from the smtp session. Simply check > "helo[0]" to see if an HELO/EHLO command was received (reset after > starttls), and "tx" to know if we are in a transaction. > ok gilles@ millert@ (eric@) ~ smtpd.8 > document -F which somehow escaped documentation (gilles@) ~ smtpd.c > add -F to usage() too; (jmc@) switchctl ~ switchctl.8 > spread the Xr! as a bonus, typo fix in switchd.conf.5 too; > Xr diff from kapetanakis giannis (jmc@) switchd ~ switchd.8 ~ switchd.conf.5 > spread the Xr! as a bonus, typo fix in switchd.conf.5 too; > Xr diff from kapetanakis giannis (jmc@) syslogd ~ privsep.c ~ syslogd.c ~ syslogd.h > Use closefrom(4) in privsep parent and avoid some global file > descriptor variables in syslogd(8) this way. > OK rzalamena@ (bluhm@) ~ privsep.c ~ ringbuf.c ~ syslogd.c > Fix trailing whitespace and shorten long lines. > No binary change. (bluhm@) ~ syslogd.8 ~ syslogd.c ~ syslogd.h > Remove the artificial maximum number of unix domain sockets in > syslogd(8). Just malloc(3) them dynamically which also gives a > more random address space layout. > OK deraadt@ (bluhm@) tcpdump ~ Makefile ~ interface.h ~ print-tcp.c + print-ofp.c > Teach tcpdump(8) how to read OpenFlow packets. This initial implementation > supports the following message types: hello, error, echo request/reply, > feature request/reply, set config, packet-in, packet-out, flow removed and > flow mod. > We currently only support printing this messages for OpenFlow 1.3.5, > however > it is possible to reuse some functions and get other versions working too. > ok deraadt@ (rzalamena@) traceroute ~ traceroute.8 > try to make DESCRIPTION read like it was written for one program, not two; > ok florian (jmc@) ~ traceroute.8 > some macro fixes for the options list; (jmc@) ~ traceroute.8 > some text cleanup and remove an irrelevant note; ok florian (jmc@) vmd ~ virtio.c > Prefer memcpy/memmove over bcopy > ok mlarkin@ (guenther@) ~ parse.y ~ priv.c ~ vm.conf.5 ~ vmd.c ~ vmd.h > Add the option to specify an interface group per virtual switch as well; > this group will be added to all VM tap(4) interfaces in the switch. > Tested by martijn@ (reyk@) ~ virtio.c ~ vm.conf.5 > When the guest OS gets a zero MAC address on a network interface, it should > create a randomized locally administrated address. OpenBSD as a guest OS > does this correctly but we cannot rely on it for other guests, so randomize > the MAC address in vmd(8) on the host side if it is not specified by the > user. I incremented the prefix by one to differentiate from the ones > that are generated by OpenBSD in the kernel; fe:e1:bb:xx:xx:xx. > OK deraadt@ mlarkin@ (reyk@) ~ vm.conf.5 > some readability fixes for the SWITCH CONFIGURATION section, > from edd barrett; > while here, there seems to be three sections to the config file, not two; > (jmc@) ypldap ~ aldap.c > Fix copy-pasto in comment; from Rob Pierce (rob (at) 2keys.ca) (guenther@) ~ aldap.c > string terminators are called NUL, not NULL (deraadt@) =============================================================================== _______________________________________________ owc mailing list [email protected] http://www.squish.net/mailman/listinfo/owc
