OpenBSD src changes summary for 2017-06-21 ==========================================
bin/csh distrib/miniroot distrib/sets etc/Makefile gnu gnu/usr.bin/binutils-2.17 gnu/usr.bin/cc regress/bin sbin/dhclient share/man sys/arch/amd64/conf sys/arch/hppa/conf sys/arch/i386/conf sys/dev/pv sys/kern sys/net usr.bin/make usr.bin/mandoc == bin =============================================================== 01/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/bin csh ~ file.c > Replace usage of TIOCSTI in csh using a more common IO-loop where ICANON is > disabled and a single char of input is read at a time. This requires the > line > editing capabilities provided when ICANON is enabled to be implemented. > With help and feedback from deraadt@ (anton@) ~ file.c > Minor style(9) corrections and whitespace cleanup to csh. > ok deraadt@ tb@ (anton@) == distrib =========================================================== 02/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/distrib miniroot ~ install.sub > /var/db/kernel.SHA256 contains the hash of the /bsd kernel. Upon > install or upgrade, set it correctly. > diff from rpe, also thanks tb for comments and review (deraadt@) sets ~ lists/base/mi > sync (deraadt@) == etc =============================================================== 03/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/etc Makefile ~ Makefile > When building a snapshot, install the compile.tgz file to builder's > /usr/share/ so that next boot will find it and perhaps use it. > ok tb rpe (deraadt@) == gnu =============================================================== 04/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/gnu gnu ~ llvm/lib/MC/MCAsmInfoELF.cpp > For a decade or more, we've patched gcc to make -fno-ident the default. > clang doesn't seem to have a command line option for the feature, > so just hard wire it off. We don't need advertisements in our binaries. > ok deraadt (tedu@) usr.bin/binutils-2.17 ~ gas/config/tc-i386.c > Large nopsleds make it easier for ROP machinery to sloppily reach targets. > Convert these to "trapsleds' instead. Since these can occur inside > fall-through blocks of code (and gcc sure isn't helping us by identifying > those), convert most to "short jump over" followed a sequence of int3. > This works for sled sizes 3-15, for i386 and amd64. Work by Todd Mortimer > ok mlarkin (deraadt@) ~ ld/emulparams/elf_i386_obsd.sh ~ ld/emulparams/elf_x86_64_obsd.sh > Intra-segment traps instead of nops. Similar to kernel change, further > reduction in nopsleds in code and data segments (if those data segments > ever become X, that is) > ok mlarkin (deraadt@) ~ bfd/elf64-x86-64.c > elf64_x86_64_plt0_entry had a hardcoded nop-sled, before the subsequent > plt entry. Since previous effective instruction is a a jmp, this can be > a sequence of traps. > ok mlarkin (deraadt@) usr.bin/cc ~ cc_tools/Makefile > DEPENDSFILE comes from FreeBSD, we don't have it. replace it with its > value .depend, so that the conditional is really used. > okay guenther@ (espie@) == regress =========================================================== 05/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/regress bin ~ Makefile + csh/Makefile + csh/filec.sh > Add regression tests for the filec feature in csh. (anton@) == sbin ============================================================== 06/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sbin dhclient ~ dhclient.c > Allow familiar idiom free_client_lease(NULL). (krw@) ~ dhclient.c ~ dhcpd.h ~ dispatch.c > Consistently use time(&var) idiom rather than tossing in a few > var = time(NULL). Flip "struct interface *" to be the first > parameter in set_timeout(), as in other functions taking the > state info. (krw@) ~ dhclient.c > No need to manually check validity of ifi->active when > REBOOT'ing. It will be pointing to a recorded lease if > anything. So rely on get_recorded_lease() to return the > next valid candidate lease, which might be the lease > ifi->active was pointing at. (krw@) ~ dhclient.c > Teach get_recorded_lease() that static leases do not > expire. (krw@) == share ============================================================= 07/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/share man ~ man4/viocon.4 > List all five devices, not just the first two. > OK jmc@ sf@ (akfaew@) == sys =============================================================== 08/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys arch/amd64/conf ~ ld.script > Fill space between sections with traps, rather than nops. (deraadt@) arch/hppa/conf ~ ld.script > Better off padding with 0x0. 0x0 = "break 0,0". > (BTW, the binutils disasm cannot handle that instruction with those > parameters, because the decoder logic is wrong..) (deraadt@) arch/i386/conf ~ ld.script > Fill space between sections with traps, rather than nops. (deraadt@) dev/pv ~ hvs.c > Include the request id in the I/O debug output (mikeb@) ~ if_hvn.c > Return early from NVS command submission if no response is required > (mikeb@) ~ hvs.c > Don't confuse SCSI command status and transfer error code. > While here, make SCSI command completion routine more robust. (mikeb@) kern ~ kern_pledge.c > Permit TIOCSTAT on a tty. (deraadt@) net ~ pf.c > Fix kernel diagnostic assertion "(sk->inp == NULL) || (sk->inp->inp_pf_sk > == NULL)". The problem was that setting the inp pointer in the > statekey to NULL was delayed until the statekey refcounter reached > 0. So the inp could get linked to another statekey while an mbuf > in the socket buffer was keeping the refcounter at 1. Set the > statekey inp pointer to NULL in pf_state_key_detach() immediately, > then the kassert can be even stricter. > OK sashan@ (bluhm@) == usr.bin =========================================================== 09/09 == http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin make ~ cond.c > error out if an exists condition tests an empty path, because that makes > zero sense. > this would have caught DEPENDSFILE in gnu/cc/cc_tools > okay guenther@, also tested by krw@ (espie@) mandoc ~ eqn.c > Recognize well-known functions names (the same that Heirloom recognizes, > which includes those recognized by groff) and wrap them in a roman box > unless they already are in roman context. > Missing feature reported by bentley@. (schwarze@) ~ eqn.c > quoted words are not parsed for defined keys (schwarze@) ~ eqn.c > Outside explicit font context, give every letter its own box. > The formatters need this to correctly select fonts. > Missing feature reported by bentley@. (schwarze@) =============================================================================== _______________________________________________ odc mailing list [email protected] http://www.squish.net/mailman/listinfo/odc
