wireguard-openbsd - WireGuard implementation for the OpenBSD kernel

	Commit message (Collapse)	Author	Age	Files	Lines
*	Setup two pair(4) and one lo(4) interface in different routing	bluhm	2021-02-05	2	-0/+181
\| \| \| \| \|	domains and with different interface MTU. Test that packet flow and TCP path MTU discovery works.
*	Syntax of pf(4) route-to has changed. Adapt tests.	bluhm	2021-02-01	4	-24/+24
\|
*	Check the rewritten address output from tcpdump -e on pflog.	bluhm	2021-01-20	1	-15/+37
\|
*	If ip or ip6 forwarding is not set, expect af-to failure.	bluhm	2021-01-16	1	-1/+8
\|
*	IPv4 loopback addresses do not work with af-to as in_canforward()	bluhm	2021-01-14	2	-55/+92
\| \| \| \| \|	prevents forwarding such packets. Switch regress to IPv4 link-local addresses. Add tests for pflog(4) on af-to rules.
*	IPv6 link-local addresses are broken in pf(4) as scope ID is used	bluhm	2021-01-14	2	-28/+73
\| \| \| \| \|	inconsistently. Switch regress to unique-local addresses. Add tests for pflog(4) on rdr-to and nat-to rules.
*	Add pflog(4) tests for IPv6.	bluhm	2021-01-13	2	-8/+74
\|
*	Check that rule without logopt user has no uid in log entry.	bluhm	2021-01-12	1	-4/+3
\|
*	Add some basic tests for pflog(4) interface. Create special routing	bluhm	2021-01-11	3	-2/+225
\| \| \| \| \|	doamin, load rules into pf(4) regress anchor, tcpdump on pflog, send packets over lo(4), grep for expected result in tcpdump output.
*	py3-scapy is now scapy	kn	2020-12-30	3	-9/+9
\| \| \| \|	OK bluhm sthen
*	Switch to scapy with python 3.	bluhm	2020-12-25	18	-177/+177
\|
*	Switch to scapy with python 3.	bluhm	2020-12-24	3	-26/+26
\|
*	Switch to scapy with python 3.	bluhm	2020-12-24	3	-28/+28
\|
*	Fix a race where the ${SUDO} process was killed instead of tcpdump.	bluhm	2020-12-17	1	-2/+2
\|
*	Link pflow and wg regress to build.	bluhm	2020-12-17	1	-2/+2
\|
*	Remove echo headlines.	bluhm	2020-12-17	5	-44/+5
\|
*	Use setup and cleanup feature from regress framework.	bluhm	2020-12-17	3	-218/+170
\| \| \| \| \|	Remove echo headlines. Use consistent target names. Makefile cleanup.
*	Since rework of source IP selection, we can reverse the last changes	denis	2020-11-07	3	-14/+7
\|
*	Compiling rtable.c kernel source file needs satosin() and satosin6()	bluhm	2020-11-06	2	-6/+5
\| \| \| \|	now. Provide macros via CPPFLAGS on command line.
*	Set up two WireGuard wg(4) interfaces in different routing domains.	bluhm	2020-11-04	1	-0/+154
\| \| \| \| \| \| \| \|	Combinations of IPv4 and IPv6 are used for tunnel and interface addresses. Check that routes are created correctly by WireGuard setup. Ping addresses on the wg interfaces locally and through the VPN tunnel. Check with tcpdump that encrypted traffic appears on loopback. Configure wg interface with bad key and check that ping does not work.
*	The 'source IP address' commit (route.c r1.249) broke the rtable regress	tb	2020-11-03	1	-1/+9
\| \| \| \| \| \| \| \| \| \|	tests since it introduced use of ifa_ifwithaddr() from if.c, so the tests no longer link due to the missing symbol. Provide an ifa_ifwithaddr() shim that makes the test link again. Passes on amd64 and sparc64. ok bluhm
*	Since OpenBSD has switched to the strict host model, this regress	bluhm	2020-02-02	1	-7/+9
\| \| \| \| \|	needs IP forwarding enabled on the packet source machine. Otherwise the pf reply-to test fails.
*	Do not create files as dependency of .BEGIN rule. Noticed by	bluhm	2020-01-15	3	-9/+9
\| \| \| \|	deraadt@ with make obj in regress.
*	Rework when tests have to be skipped due to insufficient configuration	bluhm	2020-01-08	6	-83/+156
\| \| \| \|	of the regress machine.
*	Test should also run if there is no obj directory. Name regress	bluhm	2019-07-12	1	-5/+5
\| \| \| \|	targets consistently with a run- prefix.
*	Do not use a warning flag that base gcc does not understand,	visa	2019-06-24	2	-2/+4
\| \| \| \| \| \|	and fix the warnings. From Moritz Buhl
*	Fix test after kernel interface change.	visa	2019-06-24	2	-5/+5
\| \| \| \|	From Moritz Buhl
*	To avoid SYN retransmit, kill local tcp pf state that will be reused.	bluhm	2019-06-04	1	-5/+17
\| \| \| \|	This makes the state reuse tests faster and more reliable.
*	Simplify the detection for installed python and scapy. Refactor	bluhm	2019-05-10	3	-89/+69
\| \| \| \| \|	these tests to make them similar. Use the setup and cleanup variables.
*	Mutex functions have no __ prefix anymore. Adjust kernel compat	bluhm	2019-05-08	1	-3/+3
\| \| \| \|	macros to make art.c compile again in regress.
*	Make ART data structure definitions visible to userland, in order to fix	mpi	2019-03-31	1	-2/+2
\| \| \| \| \| \|	netstat(1) inspection of routing tables. From Naoki Fukaumi, ok yasuoka@
*	The field dom_rtkeylen has been removed from struct domain. Convert	bluhm	2018-12-03	1	-8/+23
\| \| \| \|	to designated initializers and adapt the test. Fix white space.
*	Before running the tests, consistently check whether interfaces or	bluhm	2018-06-25	3	-16/+16
\| \| \| \|	routing domains are busy.
*	During test setup, IPv6 duplicate address detection creates some	bluhm	2018-06-25	1	-15/+19
\| \| \| \| \|	bogus pf states. Flush all states after the pf rules have been set. Then the first test run will use the correct routing domains.
*	Fix a segfault when inserting the same prefix multiple times	denis	2018-06-23	1	-2/+2
\| \| \| \|	OK bluhm@, jca@, kn@
*	Add gif subdir.	bluhm	2018-04-19	1	-2/+2
\|
*	Create two gif(4) interfaces in different routing domains. Test	bluhm	2018-04-19	2	-0/+224
\| \| \| \|	encapsulating IPv4 and IPv6 packets in IPv4 and IPv6 gif tunnels.
*	typos	anton	2018-02-05	1	-4/+4
\|
*	Enable traceroute test for incoming reply-to rule.	bluhm	2018-01-16	1	-7/+1
\|
*	Enable traceroute test for incoming route-to rule.	bluhm	2018-01-15	1	-14/+7
\|
*	Fix after recent mutex change.	mpi	2018-01-14	1	-1/+4
\|
*	Add tests for incoming route-to and outgoing reply-to rules. Disable	bluhm	2018-01-11	2	-42/+82
\| \| \| \|	the cases that currently fail.
*	As the divert-to pf state is not connected, it is not linked to the	bluhm	2017-12-26	1	-9/+42
\| \| \| \| \| \| \|	incpb and so not removed when the socket is closed. Packets matching reply and reply-to states are linked to the sockets immediately. So these states must go away when the socket is closed. Check this with the output of "pfctl -s states" after running the Perl scripts.
*	Add some comments what the tests are doing.	bluhm	2017-12-21	13	-16/+50
\|
*	When setting KTRACE in the environment, client and server write	bluhm	2017-12-18	9	-57/+194
\| \| \| \| \|	ktrace files. This simplifies debugging as system calls can be inspected with kdump.
*	Do not run sysctl and pfctl by != during make clean and make obj.	bluhm	2017-11-24	1	-1/+3
\| \| \| \|	requested by miod@
*	Skip the test and print a warning if pf does not process packets	bluhm	2017-11-20	1	-11/+22
\| \| \| \|	on additional loopback devices.
*	Let packets jump between routing domains until the ttl is used up.	bluhm	2017-11-17	2	-12/+91
\| \| \| \|	Use tcpdump to check that the ttl decrements to one.
*	Link loopback test to build.	bluhm	2017-11-17	1	-2/+2
\|
*	Set up two loopback interfaces in different routing domains. Try	bluhm	2017-11-17	2	-0/+129
\| \| \| \| \| \|	to ping existing and non existing addresses in these domains. Also test pinging to different rdomains via pf. Create routing loops and check that the packet vanishes.