OpenBSD src changes summary for 2016-09-07 ==========================================
bin/ksh etc/mtree/4.4BSD.dist etc/rc.d/rc.subr gnu lib/libc lib/libfuse regress/sys sbin/ping sbin/ping6 sbin/sysctl share/man sys/arch/amd64/amd64 sys/dev/acpi sys/isofs/cd9660 sys/isofs/udf sys/kern sys/miscfs/fuse sys/msdosfs sys/net sys/netinet sys/ntfs sys/sys sys/ufs/ext2fs sys/ufs/ffs usr.bin/mg usr.bin/spell usr.bin/ssh usr.sbin/config usr.sbin/npppd usr.sbin/rcctl usr.sbin/syspatch == bin =============================================================== 01/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/bin ksh ~ edit.c > remove unused variable 'last' > ok nicm@ (dcoppa@) == etc =============================================================== 02/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/etc mtree/4.4BSD.dist ~ mtree/4.4BSD.dist > restore commenting scheme: > before descending, after ascending, but not around single-subdir dirs > (schwarze@) ~ mtree/4.4BSD.dist > remove unused MD man5 dirs; ok schwarze > while here, remove the comment for man6, according to ingo's system... > (jmc@) rc.d/rc.subr ~ rc.d/rc.subr > Introduce an _rc_check_name() function to check the input script name so > that > we don't end up with cryptic error messages. > Regex help from rpe@ > Issue reported by Anthony Coulter in rcctl(8), but better fix the root > cause. > Also clarify the mage page that rc.d script name must follow ksh(1) > variable > naming. > ok robert@ sthen@ (ajacoutot@) == gnu =============================================================== 03/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/gnu gnu ~ llvm/lib/CodeGen/StackProtector.cpp > Backport https://reviews.llvm.org/rL279449 from upstream > ok pascal@ > Original commit message: > [SSP] Do not set __guard_local to hidden for OpenBSD SSP > guard_local is defined as long on OpenBSD. If the source file contains > a definition of guard_local, it mismatches with the int8 pointer type > used in LLVM. In that case, Module::getOrInsertGlobal() returns a > cast operation instead of a GlobalVariable. Trying to set the > visibility on the cast operation leads to random segfaults (seen when > compiling the OpenBSD kernel, which also runs with stack > protection). > In the kernel, the hidden attribute does not matter. For userspace code, > guard_local is defined as hidden in the startup code. If a program > re-defines guard_local, the definition from the startup code will > either win or the linker complains about multiple definitions > (depending on whether the re-defined __guard_local is placed in the > common segment or not). > It also matches what gcc on OpenBSD does. (stefan@) == lib =============================================================== 04/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib libc ~ citrus/citrus_none.c ~ citrus/citrus_utf8.c > trivial cleanup: > - delete unused headers > - add missing function prototype > - delete needless casts of return values > - KNF: return is not a function > - KNF: do not use a pointer as a boolean > - consistent wording in comments: s/octets/bytes/ > OK gcc: no object change after strip -g (schwarze@) ~ citrus/citrus_none.c > trivial cleanup: delete inline attribute on static function, > leave that decision to the compiler; no functional change (schwarze@) ~ citrus/citrus_utf8.c > Trivial cleanup: Pass pointers of the correct types to the private > functions and delete a macro that is used in only one place; > no functional change. > This completes the audit of the citrus directory (only 3 files left). > (schwarze@) ~ gen/sysctl.3 > Remove usermount remnants. ok tedu (natano@) libfuse ~ fuse_ops.c ~ fuse_private.h ~ fuse_subr.c > Fix fuse node lookups. Currently fusefs nodes in the kernel remember the > parent inode number for ".." lookups. This only works until the kernel > starts to reuse vnodes and the parent's vnode is reclaimed and the ino > to path mapping is removed from the userland process by libfuse. Fix > this by using reference counting in libfuse, so that parent mapping are > retained as long as a child uses them. Also, don't free the root node. > This commit resolves following issue: > $ doas fuse-zip ~/Downloads/foo.zip /mnt > $ ls /mnt > openbsd-www > $ grep -IR foo /usr/src > /dev/null # force vfs to reclaim vnodes > $ ls /mnt > ls: /mnt: No such file or directory > $ > ok tedu (natano@) == regress =========================================================== 05/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress sys ~ net/rtable/delete/Makefile ~ net/rtable/fullfeed/Makefile > add objdir to fix regress test (vgross@) == sbin ============================================================== 06/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin ping ~ ping.c > move arc4random down; reduces diff to ping6 (florian@) ~ ping.c > move header printing down; reduces diff to ping6 (florian@) ~ ping.c > normalize code to send preload packets (florian@) ~ ping.c > whitespace (florian@) ping6 ~ ping6.c > normalize code to send preload packets (florian@) ~ ping6.c > whitespace (florian@) sysctl ~ sysctl.8 > Remove usermount remnants. ok tedu (natano@) == share ============================================================= 07/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share man ~ man5/elf.5 > Add missing word > from Tim Kuijsten (info (at) netsend.nl) (guenther@) ~ man5/elf.5 > Describe the PT_GNU_{EH_FRAME,RELRO} and PT_OPENBSD_{RANDOMIZE,WXNEEDED} > segment types. Tighten up the description of the PT_PHDR type and > elaborate the descriptions of the PT_DYNAMIC and PT_TLS types. (guenther@) ~ man4/Makefile > install switch.4 (yasuoka@) ~ man5/elf.5 > Clarify what it means when PT_OPENBSD_RANDOMIZE and PT_GNU_RELRO segments > overlap (guenther@) ~ man5/elf.5 > Tweak wording on previous (guenther@) ~ man8/rc.subr.8 > Introduce an _rc_check_name() function to check the input script name so > that > we don't end up with cryptic error messages. > Regex help from rpe@ > Issue reported by Anthony Coulter in rcctl(8), but better fix the root > cause. > Also clarify the mage page that rc.d script name must follow ksh(1) > variable > naming. > ok robert@ sthen@ (ajacoutot@) ~ man5/elf.5 > minor tweak; (jmc@) == sys =============================================================== 08/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/amd64/amd64 ~ vmm.c > fix a wrong printf in a dump/debug function (mlarkin@) ~ vmm.c > dump some extra vcpu state if failure to launch is detected (mlarkin@) dev/acpi ~ dwiic.c > revert part of 1.18 which caused a regression with at least one ikbd > device > reported by Callum R. Davies (jcs@) isofs/cd9660 ~ cd9660_vfsops.c > Remove usermount remnants. ok tedu (natano@) isofs/udf ~ udf_vfsops.c > Remove usermount remnants. ok tedu (natano@) kern ~ kern_sysctl.c ~ vfs_syscalls.c > Remove usermount remnants. ok tedu (natano@) miscfs/fuse ~ fuse_lookup.c ~ fuse_vfsops.c ~ fuse_vnops.c ~ fusefs_node.h > Fix fuse node lookups. Currently fusefs nodes in the kernel remember the > parent inode number for ".." lookups. This only works until the kernel > starts to reuse vnodes and the parent's vnode is reclaimed and the ino > to path mapping is removed from the userland process by libfuse. Fix > this by using reference counting in libfuse, so that parent mapping are > retained as long as a child uses them. Also, don't free the root node. > This commit resolves following issue: > $ doas fuse-zip ~/Downloads/foo.zip /mnt > $ ls /mnt > openbsd-www > $ grep -IR foo /usr/src > /dev/null # force vfs to reclaim vnodes > $ ls /mnt > ls: /mnt: No such file or directory > $ > ok tedu (natano@) msdosfs ~ msdosfs_vfsops.c > Remove usermount remnants. ok tedu (natano@) net ~ route.c > Only free the old cached next hop route if the new one is valid. > Leaving a NULL pointer on a RTF_GATEWAY is no longer supported, > and a KASSERT() triggers. > Found the hardway by and ok sthen@ (mpi@) ~ bfd.c ~ bfd.h > Use C99 types. > ok claudio@, phessler@ (mpi@) ~ bfd.c > Do not check for NULL after calling an allocator with WAITOK. > All allocations are done during ioctl() so it is safe to sleep. > ok claudio@, phessler@ (mpi@) ~ if.c ~ pf.c ~ rtable.c ~ rtable.h ~ rtsock.c > Rename rtable_mpath_next() into rtable_iterate() and make it do a proper > reference count. > rtable_iterate() frees the passed ``rt'' and returns the next one on the > multipath list or NULL if there's none. > ok dlg@ (mpi@) netinet ~ if_ether.c > Rename rtable_mpath_next() into rtable_iterate() and make it do a proper > reference count. > rtable_iterate() frees the passed ``rt'' and returns the next one on the > multipath list or NULL if there's none. > ok dlg@ (mpi@) ntfs ~ ntfs_vfsops.c > Remove usermount remnants. ok tedu (natano@) sys ~ sysctl.h > Remove usermount remnants. ok tedu (natano@) ~ exec_elf.h > - point to binutils 2.17 version of elf/mips.h > - fix trailing whitespace (jasper@) ufs/ext2fs ~ ext2fs_vfsops.c > Remove usermount remnants. ok tedu (natano@) ufs/ffs ~ ffs_vfsops.c > Remove usermount remnants. ok tedu (natano@) == usr.bin =========================================================== 09/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin mg ~ buffer.c > Source Joachim Nilsson: > Found by Coverity Scan. The popbuf() function iterated over a list to > find a wp pointer, then sent it to showbuffer() which immediately went > ahead and dereferenced it. This patch simply adds a NULL pointer check > before calling showbuffer(), if NULL then just return NULL to callee. > ok awolk@ millert@ (lum@) spell ~ look.c > Avoid using the obsolete _tolower() macro. (millert@) ~ look.c > Remove other instance of _tolower (millert@) ssh ~ sshd_config.5 > sort; from matthew martin (jmc@) == usr.sbin ========================================================== 10/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin config ~ util.c > Get rid of the static nomem() function and replace it with err(1, NULL); > Written with and ok tb@ (akfaew@) ~ util.c > strlen + emalloc + snprintf = asprintf > Written with and ok tb@ (akfaew@) npppd ~ npppd/npppd.conf.5 > Fix about the default value of pptp-hostname. It's an empty string. > Reported by Joe Holden. (yasuoka@) rcctl ~ rcctl.sh > Validate service names with _rc_check_name() from rc.subr. > ok sthen@ robert@ (ajacoutot@) syspatch ~ syspatch.8 ~ syspatch.sh > syspatch goal is not to act as a package manager but to sync the system to > the > corresponding stable. So prevent rollbacking a specific binpatch but > instead > always rollback the latest one. It's the only way we can know the system is > consistent. > discussed with halex@ jasper@ robert@ (ajacoutot@) ~ syspatch.sh > Don't put bsd.mp twice in the GENERIC.MP rollback tarball. (ajacoutot@) ~ syspatch.sh > Don't fetch all patches at once but instead fetch+verify+patch one by one. > (ajacoutot@) ~ syspatch.8 > Minimal man page. (ajacoutot@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
