summaryrefslogtreecommitdiff
path: root/sys/kern
AgeCommit message (Expand)Author
2015-11-14For pledge "stdio", allow the break(2) system call which backends the brk/sbrkTheo de Raadt
2015-11-13All setsockopt IPPROTO_IPV6 IPV6_TCLASS (v4 calls this IP_TOS)Theo de Raadt
2015-11-13Use ph_ prefix for tag-related fields.Martin Pieuchot
2015-11-12Prefix flowid with ph_ and print it in m_print().Martin Pieuchot
2015-11-11ktrace vnodes do not need to be opened with FREAD, as they areTheo de Raadt
2015-11-10regenPhilip Guenther
2015-11-10Split the intra-thread functionality from kill(2) into its own syscallPhilip Guenther
2015-11-08pull initialization up before poosible goto bad, from Mark LatimerTed Unangst
2015-11-08keep all the setperf timeout(9) handling in one place; ok tedu@Christian Weisgerber
2015-11-05revert sys/kern/kern_pledge.c 1.103 and reenable pledge in pwd_mkdbSebastien Marie
2015-11-04pledge_ioctl only takes files, adjust prototype. ok semarieTed Unangst
2015-11-04move /etc/spwd.db blacklist outside PLEDGE_GETPW check.Sebastien Marie
2015-11-03AF_UNIX connect is a "unix" operation, not "rpath wpath"Theo de Raadt
2015-11-03pledge_aftersyscall has been reduced to one case, "getpw", to open aTheo de Raadt
2015-11-02some tweaks to the signal code.Ted Unangst
2015-11-02use binary-search for pledge-requestSebastien Marie
2015-11-02move the pledgenote annotation from `struct proc' to `struct nameidata'Sebastien Marie
2015-11-02also handle the kill(self) case for threads; from Theo BuehlerTheo de Raadt
2015-11-02provide ml_purge and mq_purge.David Gwynne
2015-11-01refactor pledge_*_check and pledge_fail functionsSebastien Marie
2015-11-01Do not need to check the pledge control bits for system calls that areTheo de Raadt
2015-11-01bind() on AF_UNIX should set PLEDGE_UNIX not PLEDGE_CPATH; ok semarieTheo de Raadt
2015-11-01uniformize "always allowed syscalls" with pledgeSebastien Marie
2015-10-31oops, forgot pselect! crazy how many interface the kernel has here.Theo de Raadt
2015-10-31expose the sysctl backing getloadavg(3) all the time, now that moreTheo de Raadt
2015-10-30Let m_resethdr() clear the whole mbuf packet header, not only theAlexander Bluhm
2015-10-30Add m_resethdr() to clear any state (pf, tags, flags) of an mbuf packet.Reyk Floeter
2015-10-29Add some setsockopt hacks for libisc, aka John FrumTheo de Raadt
2015-10-29In knote(), use SLIST_FOREACH_SAFE when walking the klist since a call toJoel Sing
2015-10-29Reorder system call table into sequential blocks of alike-functionalityTheo de Raadt
2015-10-28mkdir is PLEDGE_CPATH, not PLEDGE_CPATH | PLEDGE_RPATH...Theo de Raadt
2015-10-28cleanup indentation and comments in sysctl whitelistTheo de Raadt
2015-10-28more accurate pledge_fail() error and code for sys_socketSebastien Marie
2015-10-28merge whitelisted r/w paths and rd paths switch to only one. It becomesSebastien Marie
2015-10-28remove duplicate setting of p_pledgenote:Sebastien Marie
2015-10-28make sys_chroot() only allowed to be used when pledged, with "rpath id proc".Sebastien Marie
2015-10-28refactor pledge_namei() a bitSebastien Marie
2015-10-28canonpath() error isn't related to p_pledgenote requirement (only possibleSebastien Marie
2015-10-28in pledge_namei(), move PLEDGE_EXEC check sooner: it doesn't depend of pathSebastien Marie
2015-10-28Prevent F_SETOWN, unless a "proc" pledge was made.Theo de Raadt
2015-10-28syncTheo de Raadt
2015-10-28The short-lived dnssocket/dnsconnect calls are being required because weTheo de Raadt
2015-10-28Paranoa: p_pledgenote the NAMEI for ld.so loadingTheo de Raadt
2015-10-28The short-lived dnssocket/dnsconnect calls are being required because weTheo de Raadt
2015-10-28There are three situations where pty ioctl's result in a NDINIT.Theo de Raadt
2015-10-28Set pledgenote to PLEDGE_RPATH in chdir & chrootTheo de Raadt
2015-10-28TIOCCONS will probably never be permitted, but it is good style to setTheo de Raadt
2015-10-28Though sys_ktrace is not yet pledge allowed, prepare by settingTheo de Raadt
2015-10-28move p_pledgenote setting next to NDINIT()Theo de Raadt
2015-10-28kern.cptime is length 2Theo de Raadt