OpenBSD src changes summary for 2017-01-04 ==========================================
Makefile.cross distrib/miniroot distrib/sets gnu lib/libc lib/libcompiler_rt lib/libcrypto lib/libfuse regress/lib sbin/iked sbin/isakmpd share/man sys/arch/arm/include sys/arch/armv7/sunxi sys/dev/ic sys/dev/sbus sys/net sys/netinet usr.bin/find usr.bin/less usr.bin/sort usr.bin/ssh usr.sbin/pcidump == Makefile.cross ==================================================== 01/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/Makefile.cross Makefile.cross > With clang as cross-compiler we need to tell it where our cross > directory is so that the correct include and library paths are used. > For this we create a simple wrapper that calls clang in the correct > mode and also passes the sysroot. > ok kettenis@ (patrick@) == distrib =========================================================== 02/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib miniroot ~ install.sub > Enforce https for connections to ftplist.cgi and ftpinstall.cgi on > platforms that have tls enabled ftp(1). > OK tb@ (rpe@) ~ install.sub > Support https for the url to the autopartitioning template for > disklabel on platforms that have tls enabled ftp(1) including > error handling in case ftp(1) doesn't have it. > OK tb@ halex@ (rpe@) sets ~ lists/man/mi > sync (deraadt@) == gnu =============================================================== 03/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/gnu gnu ~ llvm/tools/clang/lib/Driver/ToolChains.cpp ~ llvm/tools/clang/lib/Driver/Tools.cpp > With clang as cross-compiler we need to make it pass the library path > to our linker. Additionally, we need to make sure the library path is > prefixed with the cross directory, which is sysroot. > ok kettenis@ (patrick@) == lib =============================================================== 04/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib libc ~ stdlib/qsort.c > Remove unnecessary casts of 'a' to char * since 'a' is already char *. > This is a remnant from the original 4.4BSD code that had 'a' as > void * in the function args. No binary change. OK bluhm@ (millert@) libcompiler_rt ~ gcc_personality_v0.c + unwind-ehabi-helpers.h > Update compiler-rt to version 3.9.1. There has only been a single > change in the builtins that fixes the build on ARM with LLVM 3.8. > ok kettenis@ (patrick@) libcrypto ~ man/ASN1_OBJECT_new.3 ~ man/OBJ_nid2obj.3 > Describe what ASN1_OBJECT_new(3), ASN1_OBJECT_free(3), OBJ_dup(3), > and OBJ_create(3) really do rather than making broad and incomplete > statements that are only true in some cases. > Improve the one-line descriptions. > Some minor wording improvements while here. > There is obviously more work to do in the vicinity... (schwarze@) ~ man/d2i_ASN1_OBJECT.3 > Complete rewrite: > Better one-line description. > Specify the correct header file. > Same parameter names as in ASN1_item_d2i(3). > Lots of new information. > The ASN1_OBJECT interfaces appear specifically designed to maximize > the number and subtlety of traps, maybe in order to trap the wary > along with the unwary. All the quirks, caveats, and bugs of > ASN1_item_d2i(3) apply, and there are three additional ones on top > in this page. > It looks like that design approach was so successful that the designers > managed to trap even themselves: see the new BUGS section. (schwarze@) ~ bn/asm/armv4-gf2m.pl > Remove superfluous datatype that is 32 by default. Clang complains > about it and it's ok to remove it. This only came up as our clang > is targeted at armv7 which enables the NEON instructions. > ok kettenis@ (patrick@) ~ man/Makefile + man/d2i_ASN1_SEQUENCE_ANY.3 > Write new d2i_ASN1_SEQUENCE_ANY(3) manual page from scratch. > All four functions are listed in <openssl/asn1.h> > and in OpenSSL doc/man3/d2i_X509.pod. > Note that in the OpenSSL documentation, > three of the four prototypes are incorrect. (schwarze@) ~ modes/asm/ghash-armv4.pl > Convert ARM assembly to unified syntax. Clang demands it, binutils > supports it as long as it's marked as unified syntax. > ok bcook@ kettenis@ (patrick@) libfuse ~ fuse_opt.c > Revert fuse_opt.c r1.17. It broke exfat-fuse and perhaps other plugins. > http://marc.info/?l=openbsd-ports&m=148170738917809&w=2 > reported by yasuoka@ (stsp@) == regress =========================================================== 05/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress lib ~ libfuse/fuse-opt-match.c > Remove a check for empty option strings from our fuse regress tests. > Some fuse plugins rely on empty options (exfat-fuse, ntfs-3g) so this part > of the regress test was bogus. It ended up misleading mpi@ and myself when > we made a change in libfuse/fuse_opt.c r1.17 which broke some fuse plugins. > Regression with broken exfat-fuse reported by Helg via yasuoka@. > This commit just fixes our regress test. libfuse is not yet fixed. > ok mpi@ (stsp@) == sbin ============================================================== 06/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin iked ~ dh.c ~ iked.conf.5 ~ ikev2.h ~ parse.y > Remove modular exponential groups specified in RFC5114 > Brought up by doug@, ok reyk, djm, doug (mikeb@) isakmpd ~ dh.c > Remove modular exponential groups specified in RFC5114 > Brought up by doug@, ok reyk, djm, doug (mikeb@) == share ============================================================= 07/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share man ~ man5/pf.conf.5 > modfier -> modifier > From Dimitris Papastamos. (tb@) == sys =============================================================== 08/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/arm/include ~ armreg.h ~ cpuconf.h ~ cpufunc.h ~ pmap.h ~ pte.h > unifdef CPU_XSCALE_PXA2X0, ARM_MMU_XSCALE, ARM_MMU_GENERIC (armv3) > and remove some xscale definitions. > ok kettenis@ (jsg@) arch/armv7/sunxi ~ sxipio.c ~ sxipiovar.h > Cleanup gpio(4) support code. > ok jsg@ (kettenis@) dev/ic ~ re.c > better handle empty or full rx rings. > currently re appears to end up spinning on landisk. the most obvious > cause of this would be an inability to add entries on the rx ring > when we run out of mbufs. cope with this by calling rxeof from the > isr when the FIFO_OVERFLOW status bit is set. rxeof in turn > unconditionally tries to fill the rx ring, even if it doesnt dequeue > any packets off the ring. > while here limit the number of entries that can go on the rx ring > to 1 less than the size of the ring. a full ring looks the same as > an empty ring (cos the producer and consumer are the same in both > cases), so avoid letting this happen. > ok deraadt@ who hit problems on his landisk and tested this diff (dlg@) dev/sbus ~ cs4231.c > In the set_params() routine, remove the AUDIO_ENCODING_ULINEAR case > as upper layers calls it with AUDIO_ENCODING_ULINEAR_{LE,BE} only. > (ratchov@) net ~ if.c > make if_input return early if the mbuf list is empty. > this saves doing wakeups of the softnet thread. > ok deraadt@ (dlg@) ~ if.c > shuffle the last change slightly. > only check ml_empty the second time if bpf was run and may have > filtered the list. > only call task_add if mq_enlist worked. (dlg@) ~ if.c > dont assume setting IFF_UP will succeed. > run a drivers ioctl handler and check if it worked before calling > if_up or if_down to report the change. propagate that error up to > userland so ifconfig can report what happened. > ok mpi@ (dlg@) ~ if_vxlan.c > As noticed by bluhm@ the netlock is required for the multicast cleanup > While here, remove unnecessary splnet's. > ok mpi (mikeb@) netinet ~ igmp.c ~ igmp_var.h > convert igmp_stat into per-cpu counters > it follows the same pattern used in previous conversions. > from Dimitris Papastamos (thank you). > ok mpi@ (dlg@) == usr.bin =========================================================== 09/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin find ~ find.c > If find(1) may end up calling execve(2), it needs "proc exec" in addition > to its other promises. Previously, find ran unpledged if the -exec{,dir} > or -ok primaries were specified: the tame(2) calls for find(1) were added > a few days before the "exec" promise was implemented and these code paths > were never revisited. > ok millert (tb@) less ~ charset.c > Avoid an out of bounds read when the environment variable LESSBINFMT > is set to "*". Patch from Tobias Stoeckmann. OK tb@ (millert@) sort ~ sort.c > Fix a crash when "sort -m" is given no files. From Julien Ramseier. > (millert@) ssh ~ ssh-agent.c > relax PKCS#11 whitelist a bit to allow libexec as well as lib > directories. (djm@) ~ serverloop.c > unbreak Unix domain socket forwarding for root; ok markus@ (djm@) == usr.sbin ========================================================== 10/10 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin pcidump ~ pcidump.c > s/Signaled/Signalled/ (dlg@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
