OpenBSD src changes summary for 2017-02-19 to 2017-02-26 inclusive ==================================================================
bin/cat bin/ksh bin/md5 distrib/miniroot distrib/octeon distrib/sets etc/Makefile etc/examples/pkg.conf etc/mtree/BSD.x11.dist etc/rc etc/rc.conf etc/rc.d/xdm etc/rc.d/xenodm gnu gnu/usr.bin/binutils-2.17 gnu/usr.bin/clang lib/csu lib/libc lib/libcrypto lib/libcxxabi lib/libssl lib/libtls regress/bin regress/lib regress/libexec regress/misc regress/sys regress/usr.bin regress/usr.sbin sbin/iked sbin/ping share/man sys/arch/amd64/amd64 sys/arch/amd64/conf sys/arch/amd64/include sys/arch/arm64/arm64 sys/arch/arm64/conf sys/arch/arm64/dev sys/arch/arm64/include sys/arch/armv7/broadcom sys/arch/i386/i386 sys/arch/loongson/dev sys/arch/loongson/include sys/arch/loongson/loongson sys/arch/mips64/mips64 sys/arch/octeon/include sys/conf sys/dev/acpi sys/dev/fdt sys/dev/pci sys/dev/pv sys/dev/rasops sys/kern sys/net sys/netinet sys/netinet6 sys/nfs sys/sys usr.bin/bc usr.bin/dc usr.bin/file usr.bin/fmt usr.bin/make usr.bin/mandoc usr.bin/netstat usr.bin/sed usr.bin/ssh usr.bin/tmux usr.sbin/bgpd usr.sbin/eigrpctl usr.sbin/eigrpd usr.sbin/ldapd usr.sbin/npppd usr.sbin/ocspcheck usr.sbin/pcidump usr.sbin/radiusd usr.sbin/unbound == bin =============================================================== 01/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/bin cat ~ Makefile > More whitespace, helps reading > ok ajacoutot@ (jca@) ksh ~ ksh.1 > Assigning to RANDOM calls srand_deterministic(3), not srand(3), > and simplify the next sentence; triggered by a different patch > from Anton dot Lindqvist at gmail dot com. > OK millert@ tb@ (schwarze@) md5 ~ md5.1 > SHA-1 collisions have now been found so update some man pages. As noted by > minek van on misc@. Thanks. > ok tb@, tj@ > (and jmc@ is ok with the diff but can't ok the technical content). > (daniel@) == distrib =========================================================== 02/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib miniroot ~ install.sub > Remove an unfortunate significant trailing whitespace that ended up passing > a bogus argument to ftp(1), disabling redirection and running ftp(1) in the > background. > ok rpe (tb@) ~ dot.profile > Zap the extra echo and show the prompt using the var?'prompt' feature > of read. (rpe@) ~ install.sub > Localize _OPT as it is not used as a global variable. > OK tb@ (rpe@) ~ install.sub > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) octeon ~ ramdisk/install.md > Bump the default size of the MSDOS partition from 16MB to 32MB. > 16MB is too small to hold bsd, bsd.mp and bsd.rd simultaneously. > Noticed by and OK stsp@ > OK deraadt@ (visa@) sets ~ lists/comp/mi > sync (tb@) ~ lists/comp/mi > sync (deraadt@) ~ lists/base/md.arm64 > sync (jsg@) ~ lists/base/mi > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) == etc =============================================================== 03/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/etc Makefile ~ Makefile > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) examples/pkg.conf ~ examples/pkg.conf > sync (sthen@) ~ examples/pkg.conf > sync (sthen@) ~ examples/pkg.conf > sync (sthen@) mtree/BSD.x11.dist ~ mtree/BSD.x11.dist > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) rc ~ rc > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) rc.conf ~ rc.conf > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) rc.d/xdm - rc.d/xdm > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) rc.d/xenodm + rc.d/xenodm > Switch to xenodm(1). > Do it now deraadt@ (matthieu@) == gnu =============================================================== 04/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/gnu gnu ~ usr.bin/Makefile > Skip building gdb on aarch64 there is no support for it. (jsg@) ~ llvm/tools/lld/ELF/OutputSections.cpp > Make sure the .eh_frame ends with a terminator. This avoids creating a > zero-sized .eh_frame section, which confuses the strip from our > frankenbinutils. > With this diff lld seems to be usable as the arm64 system linker. > ok patrick@, jsg@ (kettenis@) ~ llvm/tools/clang/lib/Basic/Targets.cpp > Fix the types used on aarch64 to match what we have in <sys/types.h>. > From Brad Smith. (kettenis@) ~ llvm/lib/Transforms/Scalar/LoopIdiomRecognize.cpp > Disable loop idiom recognition for _libc_memset and _libc_memcpy. These > are > the internal names we use in libc for memset and memcpy and having the > compiler optimize them as calls to memset and memcpy will lead to infinite > recursion. > This is a temporary solution while guenther@ tries to figure out a better > way to force calls from inside libc via identifiers that are of hidden > visibility. > ok jsg@, patrick@ (kettenis@) usr.bin/binutils-2.17 ~ bfd/elf64-aarch64.c > elf64-aarc64 -> elf64-aarch64 in TARGET_LITTLE_NAME > ok patrick@ (jsg@) ~ binutils/readelf.c > Add an ELF machine number to string mapping for AArch64. (jsg@) ~ bfd/elf64-aarch64.c > Set ELF_MAXPAGESIZE to 64k, to match lld(1). Also define ELF_MINPAGESIZE. > ok patrick@, jsg@ (kettenis@) ~ bfd/bfd-in2.h ~ bfd/elf32-arm.c ~ bfd/libbfd.h ~ bfd/reloc.c ~ gas/config/tc-arm.c > Implement support for generating movw/movt relocations on arm in gas(1). > Ported from FreeBSD's GPLv2 version of binutils. > ok guenther@ (kettenis@) usr.bin/clang ~ clang/clang.1 > sync clang.1 with clang.rst > Generated with gmake and py-sphinx installed via > cd /usr/src/gnu/llvm/tools/clang/docs && gmake -f Makefile.sphinx man > ok patrick@ (jsg@) == lib =============================================================== 05/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib csu ~ crtbegin.c ~ crtbeginS.c > Move static variables from .data to .bss by not initializing them to zero > ok kettenis@ (guenther@) ~ arm/md_init.h > Use a long branch (using movw/movt) to implement MD_SECT_CALL_FUNC. > Necessary (but perhaps not suffcient) to build large binaries on arm. > ok guenther@ (kettenis@) ~ aarch64/md_init.h ~ hppa/md_init.h > Our certbegin/certend is always PIC, so remove the non-PIC > MD_SECT_CALL_FUNC() > implementations. > ok guenther@ (kettenis@) libc ~ asr/asr_utils.c > Use T_OPT instead of literal "41". > Requested earlier by eric@ (jca@) ~ arch/aarch64/sys/tfork_thread.S > __tfork(3) returns the thread ID of the new thread in x0, not x1. The > value returned in x1 on __tfork(3) is always zero, which made the code > always fall into the child case. (patrick@) ~ asr/getaddrinfo_async.c > Report the errno set by getifaddrs(3) if the setup for AI_ADDRCONFIG fails, > rather than a non-informative EAI_FAIL. Compare to -1 for error detection > while here. > ok jca@ gilles@ (eric@) ~ asr/getaddrinfo_async.c ~ asr/getnetnamadr_async.c > missing breaks > ok jca@ krw@ gilles@ (eric@) ~ arch/aarch64/sys/sigprocmask.S > The assembly for sigprocmask(2) had a tiny typo which made it jump to > the wrong location, thus not making it set SIG_BLOCK with a zero mask > when attempting to read the signal mask (if set ptr is NULL). Instead > an attempt to read the mask actually told the syscall to set the mask > to zero. (patrick@) ~ asr/asr.c ~ asr/asr_private.h ~ asr/getaddrinfo_async.c ~ asr/gethostnamadr_async.c ~ asr/getnameinfo_async.c ~ asr/getnetnamadr_async.c ~ asr/getrrsetbyname_async.c ~ asr/res_search_async.c > Put a common subq pointer in the query struct, rather than one in each > element of the union. > ok gilles@ millert@ krw@ (eric@) ~ hash/md5.3 > Replace use of mathematical minus with an em-dash (from tb@) and break up > a long sentence (from tj@). (daniel@) ~ hash/sha1.3 > SHA-1 collisions have now been found so update some man pages. As noted by > minek van on misc@. Thanks. > ok tb@, tj@ > (and jmc@ is ok with the diff but can't ok the technical content). > (daniel@) ~ arch/aarch64/sys/sigpending.S > The assembly for sigpending(2) wasn't correctly implemented either. I > suppose this still stems from the first steps where we tried to get at > least some userland stuff to compile. This change properly saves the > pointer before doing the syscall and stores the return value in the > pointer on return. This makes sigpending regress pass. (patrick@) libcrypto ~ err/err.c ~ err/err.h ~ man/ERR_put_error.3 ~ man/Makefile + man/ERR_asprintf_error_data.3 > Mark ERR_add_error_data and ERR_add_error_vdata as not for internal use, > and document ERR_asprintf_error_data as their replacement. > ok jsing@, ingo@ (beck@) ~ man/ERR_asprintf_error_data.3 > don;t end Nd with a full stop; (jmc@) ~ cert.pem > Add the following root CAs, from SECOM Trust Systems / Security > Communication > of Japan, they are present in Mozilla's CA store. OK ajacoutot@ > /C=JP/O=SECOM Trust Systems CO.,LTD./OU=Security Communication EV RootCA1 > /C=JP/O=SECOM Trust Systems CO.,LTD./OU=Security Communication RootCA2 > /C=JP/O=SECOM Trust.net/OU=Security Communication RootCA1 (sthen@) libcxxabi + shlib_version > Provide libc++abi as shared library as well. Discussed and requested > due to ports build failures. > ok kettenis@ (patrick@) libssl ~ ssl_ciph.c ~ ssl_locl.h > Remove STREEBOG 512 as a TLS MAC since there are currently no cipher suites > that make use of it. > ok bcook@ inoguchi@ (jsing@) libtls ~ man/tls_init.3 ~ man/tls_read.3 > document tls_config_error(3) RETURN VALUES, improve what is said > about tls_error(3), and warn about some traps; > OK jmc@ (schwarze@) ~ man/tls_init.3 > spelling fix; (jmc@) == regress =========================================================== 06/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress bin ~ ksh/vi/sh.h > Add missing includes to avoid implicit function declarations. (jsg@) lib ~ libcrypto/bn/general/bntest.c > Check return value of every BN_* functions in bntest > - add macro CHECK_GOTO > - unify function return code to rc > - add err: label for error goto > ok bcook@ (inoguchi@) ~ libc/dirname/dirname_test.c ~ libc/ifnameindex/ifnitest.c ~ libc/locale/setlocale/setlocale.c ~ libc/strtod/strtodtest.c ~ libedit/chared/test_gets.c ~ libutil/fmt_scaled/fmt_test.c > Add missing includes to avoid implicit function declarations. (jsg@) libexec ~ ld.so/constructor/libab/ab.C ~ ld.so/lazy/libfoo/foo.c ~ ld.so/lazy/prog/prog.c > Add missing includes to avoid implicit function declarations. (jsg@) misc ~ X11/blt/blt.c > Add missing includes to avoid implicit function declarations. (jsg@) sys ~ kern/pledge/generic/main.c ~ kern/pledge/generic/manager.c ~ kern/pledge/generic/tests.out > Clean up regress/sys/kern/pledge/generic tests: many tests failed because > wlpaths is disabled in GENERIC kernels: remove or disable failing test > cases. Also remove the test_mmap() function testing the prot_exec promise > since it would need to be run on a wxallowed partition. > from semarie (tb@) ~ kern/pledge/Makefile > re-enable generic pledge tests (tb@) ~ kern/pledge/generic/Makefile ~ kern/pledge/generic/test_tty.c > Make test_request_tty() also pass in environments without tty by opening a > pseudo-tty and testing on the slave instead of stderr. > nicm confirms that this approach is fine > bluhm tested in his environment and the tests now pass > from semarie (tb@) ~ kern/pledge/ioctl/Makefile > Remove SUDO=doas in regress Makefile. The test developer must not > set the SUDO variable. The tester may decide wether he runs doas, > sudo or the whole test as root. > discussed with tb@ semarie@ benno@ (bluhm@) + kern/pledge/sendrecvfd/Makefile + kern/pledge/sendrecvfd/sendrecvfd.c > Add regress tests for fd passing and pledge. Check sending and receiving > all 7 types of vnodes as follows: > - nopledge: no pledge - verify that send/recv work as expected > - sendfd: pledge the sender with "stdio sendfd" > - recvfd: pledge the receiver with "stdio recvfd" > - nosendfd: pledge the sender with "stdio" (it is expected to fail) > - norecvfd: pledge the receiver with "stdio" (it is expected to fail) > This results in 35 possible combinations with different semantics depending > on the vnode type. > From semarie, many thanks! (tb@) ~ kern/pledge/Makefile > Enable semarie's sendrecvfd tests (tb@) ~ kern/bind/bind.c > Add missing includes to avoid implicit function declarations. (jsg@) usr.bin ~ ssh/unittests/utf8/tests.c > add test cases for C locale; ok schwarze@ (djm@) usr.sbin ~ ldapd/Makefile > ldapd regress tests depends on openldap-client, check > if that package is installed, otherwise print > a warning and skip them but not fail. > requested by/ok mpi@ (gsoares@) == sbin ============================================================== 07/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin iked ~ ikev2.c > In a scenario where a config reload happens during an IKE_AUTH exchange, > and we move an SA from one to another policy, we need to make sure to do > refcounting if the policies involved are already in the garbage collect > phase. > ok markus@ mikeb@ (patrick@) ping ~ ping.c > Unbreak "ping6 -L" > $ ping6 -L ff02::1%em1 > ping6: setsockopt IP6_MULTICAST_LOOP: Invalid argument > Unfortunately we have this inconsistency in the sockets API where we need > to pass an unsigned char for IP_MULTICAST_LOOP but we need an unsigned > int for IPV6_MULTICAST_LOOP. > While here, fix a typo in the error message. > OK millert@ florian@ (renato@) == share ============================================================= 08/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share man ~ man5/bsd.port.mk.5 > trailing space (espie@) ~ man5/bsd.port.mk.5 > more removing / simplifying of fetch stuff (espie@) ~ man5/bsd.port.mk.5 > more old mirror targets (espie@) ~ man5/bsd.port.mk.5 > tighter documentation for dependency handling (espie@) ~ man5/bsd.port.mk.5 > install-depends works as expected for multi-package setups. > it's prepare that's special. (espie@) ~ man7/roff.7 > fix spelling error; > found by Michael <Stapelberg at debian dot org> with Lintian (schwarze@) == sys =============================================================== 09/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/amd64/amd64 ~ vmm.c > SVM: fix segment A/R bits formatting (mlarkin@) ~ vmm.c > SVM: fix wrong treatment of MSRs, especially EFER. (mlarkin@) ~ vmm.c > typo in comment (mlarkin@) ~ vmm_support.S > SVM: asm support for SVM/RVI (mlarkin@) ~ vmm.c > VMX: assert that the supplied instruction length matches what is expected > for > HLT exits (mlarkin@) arch/amd64/conf ~ GENERIC > add acpisbs, an acpi smart battery subsystem driver reading data > over smbus > currently disabled because it conflicts with acpibat > ok deraadt, kettenis (jcs@) arch/amd64/include ~ vmmvar.h > SVM: asm support for SVM/RVI (mlarkin@) arch/arm64/arm64 ~ ast.c ~ exception.S ~ genassym.cf > Fix asynchronous system traps so that they actually work. Fixes at > least one regression test and allows processes to be killed using > Ctrl-C. Discussed with Dale Rahn. (patrick@) ~ vm_machdep.c > STACKALIGN() already does the right thing by casting the given argument > to unsigned long, so explicitly casting the address to unsigned int only > leads to truncation. (patrick@) ~ bus_dma.c > The AMD Seattle SoC incorporates DMA coherent controllers, especially > AHCI, PCIe and Ethernet. Since these systems rely on cache snooping, > we must not map pages that are supposed to be coherent as uncacheable. > Instead if we have a cache coherent device, create a bus tag and mark > that tag as coherent. For users of that tag we can simply use normal, > cached memory. > ok kettenis@ (patrick@) ~ intr.c > Implement an API for establishing legacy PCI interrupts. This specific > establish function parses the device tree's interrupt map to discover > the correct interrupt controller node and interrupt cells for the given > PCI node. After retrieving that information we can do the same the > normal FDT establish API already does. > ok kettenis@ (patrick@) ~ intr.c > Retire the global interrupt establish and disestablish API. It has been > of good use for supporting multiple SoCs with each their own interrupt > controllers, but as of now we have replaced it wit the FDT-enabled API. > ok kettenis@ (patrick@) ~ intr.c > MSI interrupts are established in a different way as well. Instead of > simply hooking up the interrupt handler and returning an MD cookie, we > need to pass back information for the PCI controller to configure its > interrupt correctly. For this, add another establish routine into the > FDT-based interrupt API which looks up msi-controller nodes and calls > their MSI-specific establish function if requested. > ok kettenis@ (patrick@) arch/arm64/conf ~ GENERIC ~ RAMDISK > Enable ahci(4). > ok kettenis@ (patrick@) ~ GENERIC ~ RAMDISK ~ files.arm64 > Hook up pciecam(4), but don't compile it yet. (patrick@) ~ files.arm64 > Implement ampintcmsi(4) in ampintc(4) to support MSI. The GICv2M is an > extension to the GIC controller, which is represented as subnode in the > device tree. There can be multiple GICv2Ms, so it makes sense to attach > those to ampintc(4) as some kind of simplebus. The GICv2M is simply an > interrupt generator that can be used by PCIe devices to ring the door > bell. There is no need for further configuration, we only need to find > out which SPIs we are allowed to use for MSI and to register an edge > triggered interrupt on a (randomly) allocated SPI. > ok kettenis@ (patrick@) ~ GENERIC ~ RAMDISK > Enable ampintcmsi(4), pciecam(4), ppb(4) and virtio(4). (patrick@) arch/arm64/dev ~ bcm2835_dwctwo.c > Set sc_vendor in bcm2835_dwctwo before calling dwc2_init(). > This changes the root hub dmesg output from > "vendor 0x0000 DWC2 root hub" to "Broadcom DWC2 root hub". > ok kettenis@ (jsg@) ~ mainbus.c ~ simplebus.c > The AMD Seattle SoC incorporates DMA coherent controllers, especially > AHCI, PCIe and Ethernet. Since these systems rely on cache snooping, > we must not map pages that are supposed to be coherent as uncacheable. > Instead if we have a cache coherent device, create a bus tag and mark > that tag as coherent. For users of that tag we can simply use normal, > cached memory. > ok kettenis@ (patrick@) + pciecam.c > Introduce pciecam(4), a driver for generic ECAM compatible PCI host > controllers. Support for legacy interrupts and MSI is implemented > but pending on further improvements to arm64's interrupt API. > ok kettenis@ (patrick@) ~ ampintc.c ~ bcm2836_intr.c > Retire the global interrupt establish and disestablish API. It has been > of good use for supporting multiple SoCs with each their own interrupt > controllers, but as of now we have replaced it wit the FDT-enabled API. > ok kettenis@ (patrick@) ~ ampintc.c > Implement support for interrupt types. The GIC only seems to support > level triggered active-high or egdge triggered low-to-high interrupts. > We currently always configure them to be level triggered, which is a > sane default for most controllers. Since MSI interupts on the GIC are > edge triggered, we need to be able to parse the type information and to > configure the interrupt correspondingly. > ok kettenis@ (patrick@) ~ ampintc.c > Implement ampintcmsi(4) in ampintc(4) to support MSI. The GICv2M is an > extension to the GIC controller, which is represented as subnode in the > device tree. There can be multiple GICv2Ms, so it makes sense to attach > those to ampintc(4) as some kind of simplebus. The GICv2M is simply an > interrupt generator that can be used by PCIe devices to ring the door > bell. There is no need for further configuration, we only need to find > out which SPIs we are allowed to use for MSI and to register an edge > triggered interrupt on a (randomly) allocated SPI. > ok kettenis@ (patrick@) arch/arm64/include ~ bus.h > The AMD Seattle SoC incorporates DMA coherent controllers, especially > AHCI, PCIe and Ethernet. Since these systems rely on cache snooping, > we must not map pages that are supposed to be coherent as uncacheable. > Instead if we have a cache coherent device, create a bus tag and mark > that tag as coherent. For users of that tag we can simply use normal, > cached memory. > ok kettenis@ (patrick@) ~ intr.h > Implement an API for establishing legacy PCI interrupts. This specific > establish function parses the device tree's interrupt map to discover > the correct interrupt controller node and interrupt cells for the given > PCI node. After retrieving that information we can do the same the > normal FDT establish API already does. > ok kettenis@ (patrick@) ~ intr.h > Retire the global interrupt establish and disestablish API. It has been > of good use for supporting multiple SoCs with each their own interrupt > controllers, but as of now we have replaced it wit the FDT-enabled API. > ok kettenis@ (patrick@) ~ intr.h > MSI interrupts are established in a different way as well. Instead of > simply hooking up the interrupt handler and returning an MD cookie, we > need to pass back information for the PCI controller to configure its > interrupt correctly. For this, add another establish routine into the > FDT-based interrupt API which looks up msi-controller nodes and calls > their MSI-specific establish function if requested. > ok kettenis@ (patrick@) arch/armv7/broadcom ~ bcm2835_dwctwo.c > Set sc_vendor in bcm2835_dwctwo before calling dwc2_init(). > This changes the root hub dmesg output from > "vendor 0x0000 DWC2 root hub" to "Broadcom DWC2 root hub". > ok kettenis@ (jsg@) arch/i386/i386 ~ vmm.c > same fix as amd64 (typo in comment) (mlarkin@) ~ vmm.c > VMX: assert that the supplied instruction length matches what is expected > for > HLT exits. Matches previous diff from amd64 (mlarkin@) arch/loongson/dev ~ com_leioc.c > Suppress messages about spurious UART interrupts on MP systems. > A similar workaround was added in r1.8 of cn30xxuart.c because > of the same root cause. (visa@) arch/loongson/include ~ autoconf.h ~ intr.h > Add SMP config and IPI control logic for Loongson 3A. (visa@) arch/loongson/loongson ~ generic3a_machdep.c > Add SMP config and IPI control logic for Loongson 3A. (visa@) ~ loongson3_intr.c > Skip IO interrupt mask update on secondary CPUs when restoring IPL. > The mask is node-wide, and only the primary CPU handles IO interrupts. > (visa@) arch/mips64/mips64 ~ ipifuncs.c > Add IPI counter. (visa@) arch/octeon/include ~ cpu.h > Define cache line size for the per-cpu API. (visa@) conf ~ GENERIC > temporarily disable BFD, to not block work on unlocking the socket layer. > requested by deraadt@ (phessler@) dev/acpi ~ acpi.c ~ acpidev.h ~ acpivar.h ~ files.acpi + acpisbs.c + smbus.h > add acpisbs, an acpi smart battery subsystem driver reading data > over smbus > currently disabled because it conflicts with acpibat > ok deraadt, kettenis (jcs@) ~ acpisbs.c > fix timeout loop > from Anton Lindqvist (jcs@) ~ acpireg.h ~ acpithinkpad.c > attach acpithinkpad to LEN0268 HID found on newer thinkpads (jcs@) ~ acpiasus.c ~ acpireg.h ~ acpithinkpad.c ~ acpitoshiba.c ~ atk0110.c > move vendor HIDs into their drivers, clean up some unused defines > ok kettenis (jcs@) dev/fdt ~ files.fdt + ahci_fdt.c > Add a generic ahci(4) frontend for FDT. > ok kettenis@ (patrick@) ~ ahci_fdt.c ~ sxiehci.c ~ virtio_mmio.c > The interrupts have been established using the FDT interrupt API, so we > need to use the FDT interrupt API do disestablish them. > ok kettenis@ (patrick@) dev/pci ~ if_iwm.c ~ if_iwn.c > Remove hack to release/re-acquire netlock in _ioctl(). > A short hang in X11 during an ifconfig scan is better than making > tcpdump(8) an express ticket to the ddb prompt. > ok mpi@ (krw@) dev/pv ~ hypervic.c > Add missing mtx_leave() calls in two hypervic error paths > ok mikeb@ (jsg@) ~ if_xnf.c ~ xbf.c ~ xen.c > Update license > Some final touches before the release, increase the maximum > number of CAS iterations before we declare the grant table > entry lost forever. This happens on older Xen 3.x versions > as reported by Kirill Miazine. (mikeb@) dev/rasops ~ rasops32.c > memcpy a character row at a time to the framebuffer, rather than > sending pixel by pixel > greatly speeds up the text framebuffer on some EFI implementations, > such as shaving 10 seconds off a boot under VMWare Fusion > ok deraadt (jcs@) kern ~ subr_pool.c > revert 1.206 because it allows deadlocks. > if the gc task is running on a cpu that handles interrupts it is > possible to allow a deadlock. the gc task my be cleaning up a pool > and holding its mutex when an non-MPSAFE interrupt arrives and tries > to take the kernel lock. another cpu may already be holding the > kernel lock when it then tries use the same pool thats the pool GC > is currently processing. > thanks to sthen@ and mpi@ for chasing this down. (dlg@) ~ sys_socket.c > Do not grab the NET_LOCK() when poll(2)ing on unix domain sockets. > Fix the 'X freeze' while scanning with wireless interfaces. Problem > reported by pirofti@. > ok tb@, bluhm@ (mpi@) ~ uipc_proto.c > Use c99 struct initialization with protosw. > This makes it easier to grep for a member, such as .pr_usrreq, and know > which functions to review. > ok mpi@ bluhm@ jca@ (dhill@) net ~ if_pfsync.c ~ if_pfsync.h > pfsync(4) percpu counters > ok florian@ (jca@) ~ bpf.h > The pcap people gave us ID 267 for DLT_OPENFLOW > Via https://github.com/the-tcpdump-group/libpcap/issues/542 > OK sthen@ jsg@ (reyk@) ~ pfvar.h > Amend a misleading comment mentioning ``ipintrq''. (mpi@) ~ pfkey.c ~ rtsock.c > Use c99 struct initialization with protosw. > This makes it easier to grep for a member, such as .pr_usrreq, and know > which functions to review. > ok mpi@ bluhm@ jca@ (dhill@) ~ pfkeyv2.c > Avoid some casts by using the sa field of union sockaddr_union. > OK millert@ (bluhm@) netinet ~ ip_carp.c ~ ip_carp.h > carp(4) percpu counters > ok florian@ (jca@) ~ ip_ipip.h > Crank all members of struct ipipstat to 64 bits > Will make conversion to percpu counters easier. ok bluhm@ (jca@) ~ in_proto.c > Use c99 struct initialization with protosw. > This makes it easier to grep for a member, such as .pr_usrreq, and know > which functions to review. > ok mpi@ bluhm@ jca@ (dhill@) netinet6 ~ ip6_output.c > In ip6_ctloutput, check for an invalid level early like the other > *ctloutput functions. This also reduces a level of indentation. > ok mpi@ (dhill@) ~ in6.h > Fix comments about a few ipv6 sockoptions > OK millert@ florian@ (renato@) ~ in6_proto.c > Use c99 struct initialization with protosw. > This makes it easier to grep for a member, such as .pr_usrreq, and know > which functions to review. > ok mpi@ bluhm@ jca@ (dhill@) nfs ~ nfs.h ~ nfs_bio.c ~ nfs_serv.c ~ nfs_socket.c ~ nfs_subs.c ~ nfs_syscalls.c ~ nfs_var.h ~ nfs_vfsops.c ~ nfs_vnops.c ~ nfsmount.h > Keep local definitions local. > "good work" deraadt@, ok visa@ (mpi@) sys ~ percpu.h > be nicer to 32bit strict alignment archs that can do 64bit loads/stores. > ie, align boot memory to the size of uint64_ts. > there's an argument to be made that we should align this to 16 bytes > to be consistent with malloc and pools. (dlg@) == usr.bin =========================================================== 10/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin bc ~ bc.y > - write parse errors to stderr, prompted by Martijn Dekker > - we're only interactive if stdout en stderr are a tty as well as stdin > (otto@) dc ~ bcode.c ~ dc.1 > Introduce e command, equivalent to p, but writes to stderr (otto@) ~ bcode.c ~ inout.c > Move away from BN_zero, the one returning an int is deprecated and the > new one is a void function. From Daniel Cegielka. (otto@) file ~ magdir/elf > Recognise AArch64 binaries and core dumps. (jsg@) fmt ~ fmt.c > Do not break the line at U+00A0 NO-BREAK SPACE. > Bug pointed out by Eric dot Pruitt at gmail dot com on misc@. > OK millert@ bentley@ (schwarze@) ~ fmt.c > * Hoist Copyright notice to the top, don't hide it in the middle. > * Add Copyright and license for my changes, and for pjanzen@'s in 2000. > * Comply with the obnoxious license condition that we must list changes > in the source code. > * Delete a fragment of an ancient RCS change log listing changes by > Gareth McCaughan himself. The license does NOT require us to keep that, > and it does not contain interesting information, not even for historians. > OK pjanzen@ deraadt@ (schwarze@) make ~ varmodifiers.c > add support for variables in the lhs of SYSV modifiers. > lack of support noted by naddy@ > source, xenocara, and ports still build without a hitch. > okay naddy@ (espie@) mandoc ~ mdoc_html.c > Handle an odd edge case where .It is preceded by .Sm. > NULL dereference in man.cgi reported by Gabriel Guzman <gabe at > guzman dash nunez dot com> on misc@. (schwarze@) ~ main.c > Since SQLite is gone, we no longer need the "flock" pledge. > Patch from semarie@, OK tb@. (schwarze@) ~ cgi.c > Pledge man.cgi(8). > Based on a more complicated patch from semarie@. > Sebastien and tb@ both agree with the simplification. (schwarze@) netstat ~ inet.c > Crank all members of struct ipipstat to 64 bits > Will make conversion to percpu counters easier. ok bluhm@ (jca@) sed ~ process.c > Remove unused-with-dead-store variable oldpsanl. > From Daniel Cegielka - thanks > ok jca@ (tom@) ssh ~ utf8.c > Add a common nl_langinfo(CODESET) alias for US-ASCII "ANSI_X3.4-1968" > that is used by Linux. Fixes mprintf output truncation for non-UTF-8 > locales on Linux spotted by dtucker@; ok deraadt@ schwarze@ (djm@) ~ sshd.c > might as well set the listener socket CLOEXEC (djm@) tmux ~ input.c > Add SGR 28 to clear hidden flag. (nicm@) ~ screen-write.c > Scrolling at least needs to be flushed before sending EL to the terminal > (but it is simpler to flush everything, so do that instead). (nicm@) ~ tmux.h ~ tty-term.c > Don't need is1,is2,is3 so remove them. (nicm@) ~ tty.c > Change pane redraw to collect cells up as well, and simplify it a bit. > (nicm@) ~ grid.c ~ key-string.c ~ window.c > Minor bits: fix an array size, add comment, make grid_cell_entry static. > (nicm@) == usr.sbin ========================================================== 11/11 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin bgpd ~ bgpd.8 > draft-ietf-idr-large-community has been published as RFC 8092 (phessler@) ~ parse.y ~ pfkey.c > Add missing htonl for IPsec SPI. > Also, do not allow to configure SPI values in the 0..255 range. RFC 4302 > and RFC 4303 say the following: > "The set of SPI values in the range 1 through 255 are reserved by the > Internet Assigned Numbers Authority (IANA) for future use; a reserved > SPI value will not normally be assigned by IANA unless the use of the > assigned SPI value is specified in an RFC. The SPI value of zero (0) > is reserved for local, implementation-specific use and MUST NOT be > sent on the wire". > ok and tweak benno@ (renato@) eigrpctl ~ eigrpctl.c ~ eigrpctl.8 > Allow specifying an alternate socket path. (renato@) eigrpd ~ eigrpd.h ~ eigrpe.h ~ interface.c ~ kroute.c ~ parse.y > No need to use SIOCGIFRDOMAIN anymore. (renato@) ldapd ~ btest.c > print a formatted errx() string instead of segfault, > by checking argc correctly. > ok guenther@ (gsoares@) ~ ldapd.c ~ ldapd.h ~ ldape.c > Implement fork+exec model > OK jmatthew@ (gsoares@) npppd ~ npppd/Makefile > Stop using "DEBUG" make variable for extra debugs. (yasuoka@) ~ npppd/npppd_auth.c > Delete lines which are misleading and unneeded. > spotted patrick (yasuoka@) ~ npppd/npppd.conf.5 > The default for the strip-nt-domain attribute is "yes", so fix up > the manpage to not tell lies. > ok yasuoka@ (patrick@) ocspcheck ~ ocspcheck.c > Add missing $OpenBSD$ (beck@) ~ ocspcheck.c > pledge stdio before parsing the http response > ok tb@ (beck@) pcidump ~ Makefile > build pcidump on arm64 (jsg@) radiusd ~ Makefile.inc > Stop using "DEBUG" make variable for extra debugs. (yasuoka@) unbound ~ configure ~ configure.ac ~ doc/README ~ doc/example.conf.in ~ doc/libunbound.3.in ~ doc/unbound-anchor.8.in ~ doc/unbound-checkconf.8.in ~ doc/unbound-control.8.in ~ doc/unbound-host.1.in ~ doc/unbound.8.in ~ doc/unbound.conf.5.in > update to unbound-1.6.1 release; only version string changes compared to > the release candidate we had. (sthen@) ~ Makefile.in > Use mkdir -p instead of install -d to create the obj/util directory. > The latter defaults to 755, which doesn't play nicely with BUILDUSER. > Problem reported by jmc and Jan Stary; tested by myself and jmc. > ok ajacoutot (tb@) =============================================================================== _______________________________________________ owc mailing list [email protected] http://www.squish.net/mailman/listinfo/owc
