summaryrefslogtreecommitdiff
path: root/sys/netinet/in_pcb.c
AgeCommit message (Collapse)Author
2017-08-11Validate sockaddr from userland in central functions. This resultsAlexander Bluhm
in common checks for unix, inet, inet6 instead of partial checks here and there. Some checks are already done at a higher layer, but better be paranoid with user input. OK claudio@ millert@
2017-08-04The in_pcbhashlookup() in in_pcbconnect() enforces that the 4 tupelAlexander Bluhm
of src/dst ip/port is unique for TCP. But if the socket is not bound, the automatic bind by connect happens after the check. If the socket has the SO_REUSEADDR flag, in_pcbbind() may select an existing local port. Then we had two colliding TCP PCBs. This resulted in a packet storm of ACK packets on loopback. The softnet task was constantly holding the netlock and has a high priority, so the system hung. Do the in_pcbhashlookup() again after in_pcbbind(). This creates sporadic "connect: Address already in use" errors instead of a hang. bug report and testing Olivier Antoine; OK mpi@
2017-06-09Replace rtrequest(RTM_DELETE...) rtrequest_delete() and do not evenMartin Pieuchot
try to remove a route from the table if it is and invalid cache. This is a step towards decoupling code dealing with userland and kernel inserted routes. ok bluhm@
2017-06-07Grab the KERNEL_LOCK() around rtm*() functions. Routing sockets globalsMartin Pieuchot
aren't protected by the NET_LOCK(). While here change lock assertions in rt_{set,put}gwroute(), the NET_LOCK() is enough. Tested by Hrvoje Popovski. ok jmatthew@, claudio@
2017-03-07When the inpcb queue and hash lists are traversed or modified weAlexander Bluhm
need netlock. Remove the obsolete splnet. OK mpi@
2017-03-07Initially in_pcballoc() hooked all new inpcb, including the IPv6Alexander Bluhm
ones, into the IPv4 hash. They cannot be used before bind(2) anyway and then they are rehashed and rehooked, so this was not noticed. Nevertheless put IPv6 PCBs into the IPv6 hash from the beginning. OK jca@ mpi@
2017-03-06Kill global list of IPv6 addresses.Martin Pieuchot
ok bluhm@
2017-03-06Prefix functions dealing with routing messages with 'rtm_' and keepMartin Pieuchot
them all in net/rtsock.c. This allows to easily spot which functions are doing a copyout(9) when dealing with the routing midlayer. ok phessler@, bluhm@, dhill@, krw@, claudio@
2016-10-06Remove redundant comments that say a function must be called atAlexander Bluhm
splsoftnet() if the function does a splsoftassert(IPL_SOFTNET) anyway.
2016-09-15all pools have their ipl set via pool_setipl, so fold it into pool_init.David Gwynne
the ioff argument to pool_init() is unused and has been for many years, so this replaces it with an ipl argument. because the ipl will be set on init we no longer need pool_setipl. most of these changes have been done with coccinelle using the spatch below. cocci sucks at formatting code though, so i fixed that by hand. the manpage and subr_pool.c bits i did myself. ok tedu@ jmatthew@ @ipl@ expression pp; expression ipl; expression s, a, o, f, m, p; @@ -pool_init(pp, s, a, o, f, m, p); -pool_setipl(pp, ipl); +pool_init(pp, s, a, ipl, f, m, p);
2016-09-06pool_setipl for various netinet and netinet6 bitsDavid Gwynne
thank you to everyone who helped reviewed these diffs ok mpi@
2016-08-04Commit in6_selectsrc() split again, with missing assignment fixed.Vincent Gross
2016-07-22Revert in_selectsrc() refactoring, it breaks IPv6.Martin Pieuchot
Reported by Heiko on bugs@. ok stsp@, claudio@
2016-07-20Split in6_selectsrc() into a low-level part and a pcb-level part, andVincent Gross
convert in_selectsrc() prototype to match. Ok bluhm@ mpi@.
2016-07-19NULLify a route pointer after calling rtfree(9).Martin Pieuchot
This should theoretically be a no-op because we're freeing the PCB right after, but it helps us debug a reference count problem found by otto@. ok mikeb@
2016-07-05Do not use ``rt_addr'' in in{6,}_selectsrc() it doesn't work with magicMartin Pieuchot
addresses set on p2p interfaces. Found the hardway by naddy@
2016-06-30Use ``rt_addr'' rather than ``rt_ifa'' to get the source addressMartin Pieuchot
corresponding to a route. ok florian@ on a previous version, input and ok bluhm@
2016-06-18Add net.inet.{tcp,udp}.rootonly sysctl, to mark which portsVincent Gross
cannot be bound to by non-root users. Ok millert@ bluhm@
2016-04-19Use the correct byte-order when checking against baddynamic ports,Stuart Henderson
restoring pre-1.190 behaviour. ok millert@ mpi@
2016-04-13Remove extra parenthesis around comparison.Vincent Gross
Found by David Hill <dhill@mindcry.org> with clang.
2016-04-12Unbreak tree. Put 'laddr6' declaration inside #ifdef INET6/#endif.Kenneth R Westerback
2016-04-11Simplify in_pcblookup_local() logic.Vincent Gross
ok mpi@
2016-04-11Rename in_pcblookup() to in_pcblookup_local() and change its prototypeVincent Gross
to get rid of the now useless foreign address and ports parameters. ok mpi@
2016-04-08We must always check if the socket is already bound, and always setVincent Gross
INPLOOKUP_IPV6 when applicable. Bug introduced in revision 1.197 by yours truly. Ok bluhm@
2016-04-05Move reserved port checks from in(6)_pcbaddrisavail() to in_pcbbind().Vincent Gross
Kill old comments while at it. Ok mpi@ bluhm@
2016-04-05Move inp_laddr assignment after in_pcbpickport(), extend in_pcbpickport()Vincent Gross
as needed. Ok bluhm@
2016-03-26Always include the route priority in routing messages.Martin Pieuchot
From Florian Riehm, ok bluhm@
2016-03-23Merge in_pcbbind() and in6_pcbbind(), and change every call toVincent Gross
in6_pcbbind() into in_pcbbind(). Ok jca@ mpi@
2016-03-23Extract in_pcbaddrisavail() from in_pcbbind().Vincent Gross
ok jca@
2015-12-18Fix SO_REUSE* flags effects when binding multicast addresses. NoVincent Gross
regression observed on avahi. ok benno@
2015-12-03Get rid of rt_mask() and stop allocating a "struct sockaddr" for everyMartin Pieuchot
route entry in ART. rt_plen() now represents the prefix length of a route entry and should be used instead. For now use a "struct sockaddr_in6" to represent the mask when needed, this should be then replaced by the prefix length and RTA_NETMASK only used for compatibility with userland. ok claudio@
2015-12-03Remove broadcast matching from ifa_ifwithaddr(), use in_broadcast() whereVincent Gross
required. ok bluhm@ mpi@.
2015-12-03To avoid that the stack manipules the pf statekeys directly, introduceAlexander Bluhm
pf_inp_...() lookup, link and unlink functions as an interface. Locking can be added to them later. Remove the first linking at the beginning of tcp_input() and udp_input() as it is not necessary. It will be done later anyway. That code was a relict, from the time before I had added the second linking. Input from mikeb@ and sashan@; OK sashan@
2015-12-03Rename pf_unlink_state() to pf_remove_state() so the name does notAlexander Bluhm
collide with the statekey to inp unlinking. OK sashan@ mpi@
2015-12-02Move port picking away from in_pcbbind()Vincent Gross
ok sthen@
2015-12-02- hide PF internals to pf_unlink_divert_state() from in_pcb.cAlexandr Nedvedicky
OK mpi@, bluhm@
2015-10-30Rename rtrequest1() to rtrequest().Alexander Bluhm
OK mpi@
2015-10-25Instead of doing the the if_get() dance for rt_missmsg(), change theAlexander Bluhm
function to take an interface index. discussed with mpi@; OK claudio@
2015-10-23``rt_ifp'' cannot be NULL.Martin Pieuchot
ok claudio@
2015-10-20At guenther's suggestion replace dnssocket() with a SOCK_DNS flag onTheo de Raadt
socket(). Without pledge, all other socket behaviours become permitted, except this one case: connect/send* only works to *:53. In pledge mode, a very few are further restricted. Some backwards compatibility for the dnssocket/dnsconnect calls will remain in the tree temporarily so that people can build through the transition. ok tedu guenther semarie
2015-10-19Stop checking for RTF_UP directly, call rtisvalid(9) instead.Martin Pieuchot
While here add a missing ``rtableid'' check in in_selectsrc(). ok bluhm@
2015-10-19deduplicate in[6]_pcbbind() port scan loop.Vincent Gross
ok mpi@
2015-10-18Add two new system calls: dnssocket() and dnsconnect(). This creates aTheo de Raadt
SS_DNS tagged socket which has limited functionality (for example, you cannot accept on them...) The libc resolver will switch to using these, therefore pledge can identify a DNS transaction better. ok tedu guenther kettenis beck and others
2015-10-09Rename tame() to pledge(). This fairly interface has evolved to be moreTheo de Raadt
strict than anticipated. It allows a programmer to pledge/promise/covenant that their program will operate within an easily defined subset of the Unix environment, or it pays the price.
2015-09-22Remove inpt_lastport from struct inpcbtable, use local variablesVincent Gross
in in_pcbbind() and in6_pcbsetport() ok claudio@, with input from David Hill
2015-09-11Convert _TM_ flags to TAME_ flags, collapsing the entire mappingTheo de Raadt
layer because the strings select the right options. Mechanical conversion. ok guenther
2015-09-11Stricter checks on sockaddr content when binding PF_INET sockets.Vincent Gross
Ok mpi@
2015-09-11Only include <sys/tame.h> in the .c files that need itPhilip Guenther
ok deraadt@ miod@
2015-09-09if_put after you if_getDavid Gwynne
ok mpi@
2015-09-01Replace sockaddr casts with the proper satosin(), ... calls.Alexander Bluhm
From David Hill; OK mpi@; tested kspillner@; tweaks bluhm@
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-12 07:43:11 +0000