summaryrefslogtreecommitdiff
path: root/regress/sys
AgeCommit message (Collapse)Author
2024-01-05Wait until the expected interface state is reached instead of relying onAnton Lindqvist
arbitrary sleeps. Should hopefully make these tests more stable.
2024-01-04Skip tests that use too many resources on armv7.Alexander Bluhm
2024-01-02Run bind(2) and connect(2) in parallel also for TCP and Raw IP.Alexander Bluhm
Before only UDP was tested.
2023-12-29more regress tests for RB_* and RBT_* macrosaisha
help, pointers and OK bluhm@
2023-12-07Run bind(2) and connect(2) stress test also with IPv6.Alexander Bluhm
2023-12-06Add tests that create and delete cloned routes during connect(2).Alexander Bluhm
2023-12-06Stress test bind(2) and connect(2) system calls in OpenBSD regress.Alexander Bluhm
2023-11-11Cope with recent rt_hash() const changes.Anton Lindqvist
2023-10-19Add missing removal of network interfaces in regress cleanup targets.Anton Lindqvist
Reducing the risk tests conflicting with each other. ok bluhm@
2023-10-14Suppress harmless EPIPE errors during test shutdown.Anton Lindqvist
2023-10-13Use a unique mount point rooted in /mnt in order to not conflict with otherAnton Lindqvist
tests.
2023-10-13Use the idiom in cleanup target like many othersAnton Lindqvist
in regress already does.
2023-10-12Let vnconfig select an unused device as opposed of unconditionallyAnton Lindqvist
assume vnd0 is available.
2023-10-11Check if loopback interfaces used are skipped by pf as opposed ofAnton Lindqvist
bailing out if any loopback interface is skipped other than lo0.
2023-10-11Infer rdomains from environment variables N1 and N2 like many others inAnton Lindqvist
regress/sys/net already does. No functional change as the defaults remains the same.
2023-10-10Fix timing issue in which one thread could be left hanging inAnton Lindqvist
pipe_read().
2023-10-10Make it possible to run fileops tests in parallel by making use of aAnton Lindqvist
unique mount point and vnd device.
2023-09-08Print a message when test has finished successfully.Alexander Bluhm
2023-09-08Test pf and stack with double atomic IPv6 fragments.Alexander Bluhm
That means the IPv6 header chain contains two fragment header that spawn the whole packet. Such packets are illegal and pf drops them. Otherwise they could bypass pf rules as described in CVE-2023-4809. OpenBSD is not affected as pf_walk_header6() drops them with "IPv6 multiple fragment" log message. This check exists since 2013 when special support for atomic fragments was added to pf. If pf is disabled, the IPv6 stack accepts such packets. I do not consider this a security issue.
2023-08-20Add initial regress tests for kqueue1(2)Visa Hankala
2023-08-14unreferneced -> unreferencedMiod Vallat
2023-08-13Ignore failure when deleting regress/ttest anchorAlexandr Nedvedicky
in 'stamp-setup' target. Found by anton@.
2023-08-13kevent: Add precision and abstimer flags for EVFILT_TIMERVisa Hankala
Add timer precision flags NOTE_SECONDS, NOTE_MSECONDS, NOTE_USECONDS and NOTE_NSECONDS for EVFILT_TIMER. Also, add an initial implementation of NOTE_ABSTIME timers. Similar kevent(2) flags exist on FreeBSD, NetBSD and XNU. Initial diff by and OK aisha@ OK mpi@
2023-08-10Table persistent flag (PFR_TFLAG_PERSIST) won't get setAlexandr Nedvedicky
by ioctl(2) operation if table exists already. The issue has been noticed by Giannis Kapetanakis (_at_) edu.physics.uoc.gr. Giannis noticed relayd(8) aborts unexpectedly when 'relay host' gets disabled by 'relayctl host dis ...' command. To understand what's going on we must look at the way how relayd(8) manages its firewall configuration. If relay-host 'example' is enabled relayd(8) inserts rule to anchor relayd/example. The rule looks somewhat like this: pass in from ... to ... rdr-to <example> When the rule gets inserted to pf(4) the table 'example' is created with flags: lumpy# pfctl -a relayd/example -vg -sT ----r-- example relayd/example r-flag indicates table is referred by rule only. In the next step relayd(8) creates and populates table example. It asks pf(4) to add a persistent flag (PFR_TFLAG_PERSIST), so table survives flush operation of relayd/example ruleset on behalf of 'relayctl host dis ...' command. relayd(8) exits via abort() when table is gone with disable operation. Giannis was patient enough so we could debug and fix issue. The committed change has been tested by Giannis too. OK kn@, bluhm@
2023-07-19Cope with LRO for TCP being enabled per default by now.Anton Lindqvist
2023-07-12Report fork errors as this test is likely to hit the defaultAnton Lindqvist
kern.maxproc limit.
2023-07-10Infer the timeout from the environment, with sane defaults. ShouldAnton Lindqvist
hopefully make these tests more stable on my slow^W regress machines. ok sashan@
2023-07-10Improve err/warn messages:Anton Lindqvist
* The colon space separator is already appended by err/warn. * Favor err(1, NULL) for malloc errors.
2023-07-09Allocate enough memory to pervent SIGSEGV or SIGABRT due to writeAlexander Bluhm
after free. There was confusion between size of struct and size of pointer.
2023-07-07Implement route reference counting mockup poperly in rtable regress.Alexander Bluhm
This allows to change rtable implementation in kernel without breaking test.
2023-07-06adding regression tests:Alexandr Nedvedicky
to verify limit on tickets progam can retrieve by DIOCXGETRULES. Ad tests which verify DIOCXEND works as expected, that program can release ticket obtained by earlier call to DIOCGETRULES improvements from anton@ OK anton@, bluhm@
2023-07-06sysctl(2) is now sysctl(3), in particular it does not access name[0] inTheo de Raadt
userland. This test should have been written with a signal handler to be accurate.. Anyways, stop using syscall() to work around that ok kettenis
2023-07-06syscall(2) regress will eventually be pointless, and untestable alsoTheo de Raadt
ok kettenis
2023-06-26On amd64, test whether PKU has been enabled and set our expectationPhilip Guenther
of the results based on that. Also, the system now enforces unreadability in copyin() of ld.so, libc, and application text, even when PKU isn't enabled, so adjust those results to match. ok deraadt@ anton@
2023-05-24Skip tests if p5-BSD-Socket-Splice is not installed.Alexander Bluhm
2023-05-24Default pf rule "block return; pass" interferes with test.Alexander Bluhm
Use "block; pass" instead.
2023-05-13vmm(4)/vmd(8): switch to anonymous shared mappings.Dave Voutila
While splitting out emulated virtio network and block devices into separate processes, I originally used named mappings via shm_mkstemp(3). While this functionally achieved the desired result, it had two unintended consequences: 1) tearing down a vm process and its child processes required excessive locking as the guest memory was tied into the VFS layer. 2) it was observed by mlarkin@ that actions in other parts of the VFS layer could cause some of the guest memory to flush to storage, possibly filling /tmp. This commit adds a new vmm(4) ioctl dedicated to allowing a process request the kernel share a mapping of guest memory into its own vm space. This requires an open fd to /dev/vmm (requiring root) and both the "vmm" and "proc" pledge(2) promises. In addition, the caller must know enough about the original memory ranges to reconstruct them to make the vm's ranges. Tested with help from Mischa Peters. ok mlarkin@
2023-05-11Teach dump_tables about PKs and display non-readable pages as suchPhilip Guenther
2023-04-27cope with recent vmm changesAnton Lindqvist
2023-04-23cope with more s/XCR0/XFEATURE renamesAnton Lindqvist
2023-04-18Define net lock assertions for building kernel code in regress.Alexander Bluhm
2023-03-08Delete obsolete /* ARGSUSED */ lint comments.Philip Guenther
ok miod@ millert@
2023-03-08Delete obsolete /* ARGSUSED */ lint comments.Philip Guenther
ok miod@ millert@
2023-02-19Disable tests using the preferred bsd.regress.mk technique.Anton Lindqvist
2023-02-07switch to new simplified disklabel template format.Kenneth R Westerback
ok kn@
2023-01-22Cope with xonly on amd64 by moving the testfly routine to the rodataAnton Lindqvist
section.
2023-01-18Add xonly tests, initially written by deraadt@. Tweaked by me to fit andAnton Lindqvist
the regress framework and allowing the expected outcome to be enumerated per architecture. Currently limited to amd64 and arm64.
2023-01-11make REGRESS_CLEANUP consistent with REGRESS_CLEANUP foundAlexandr Nedvedicky
in pf_state/Makefile. discussed with anton@ OK anton@
2023-01-10purge unhooked wx_syscall filesAnton Lindqvist
2023-01-10Retire wx_syscall regress tests, they have been failing since theAnton Lindqvist
introduction of mimmutable(2) and also needs to cope with xonly on some architectures by now. There's already the syscallwx target in regress/usr.bin/lastcomm which does the same thing but also ensures that the expected code path is reached by looking at accounting records.
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-27 02:58:07 +0000