Get rid of static Atffiles and let bsd.test.mk generate them on the fly.

tp_glob -> tp-glob (i.e. better use correct keywords ..)

Add simple etfs test which almost tests for bug fixed yesterday
(the yesterday bug was against an emulated block device, so slighlty
different). But at least this test uncovered a few bugs already.

Add a test which checks autoloading modules from the host's
/stand/arch/vers/kmods works in rump (and that the result is usable ;).

On i386 this "just works". For amd64, due to -mcmodel=kernel,
things are a little more complicated. We must have the entire rump
kernel loaded in the lower 2GB. Currently, this can be done either
by using the non-PIC version for the rump kernel compiled with
-mcmodel=small, or, as njoly pointed out, using netbsd32, which
causes vm_default_addr() to give something in the lower 2GB and
therefore shared libs "magically" getting loaded there. I guess
it would be possible to put a suggested vaddr into the rump kernel
libs and make ld.elf_so map memory from the suggested address if
present ... but that's another show.

Also thanks to tron for access to an amd64 so that I could verify
the test works.

Add some kernel namespace (i.e. _KERNEL) helpers for regression tests.

branches: 1.1.2;
add regression test for module init/fini in rump

Add a test checks that 16 rump kernels on one shmif bus can ping
each other.

Get rid of static Atffiles and let bsd.test.mk generate them on the fly.

Add simple etfs test which almost tests for bug fixed yesterday
(the yesterday bug was against an emulated block device, so slighlty
different). But at least this test uncovered a few bugs already.

Add a test which checks autoloading modules from the host's
/stand/arch/vers/kmods works in rump (and that the result is usable ;).

On i386 this "just works". For amd64, due to -mcmodel=kernel,
things are a little more complicated. We must have the entire rump
kernel loaded in the lower 2GB. Currently, this can be done either
by using the non-PIC version for the rump kernel compiled with
-mcmodel=small, or, as njoly pointed out, using netbsd32, which
causes vm_default_addr() to give something in the lower 2GB and
therefore shared libs "magically" getting loaded there. I guess
it would be possible to put a suggested vaddr into the rump kernel
libs and make ld.elf_so map memory from the suggested address if
present ... but that's another show.

Also thanks to tron for access to an amd64 so that I could verify
the test works.

Add some kernel namespace (i.e. _KERNEL) helpers for regression tests.

branches: 1.1.2;
add regression test for module init/fini in rump

Define _KERNTYPES for things that need it.

add Makefile.inc everywhere so that we can set WARNS=4 by default. Amazing
how many bugs this found :-)

Add rump-based test cases for threadpool(9).

branches: 1.6.2; 1.6.4;
Add simple test for workqueue(9)

branches: 1.5.38;
elementary tests for rump sigmodels

Add some tests for lock errors. Notably, some of them fare nicely
with just the pthread consistency checks (which are enabled by
default in rumpuser), but others require real LOCKDEBUG.

add a vm allocator test which tests that:
* RUMP_MEMLIMIT works
* allocating memory with PR_NOWAIT will fail immediately if no memory
is available and PR_WAITOK will wait for memory to be available

need bsd.own.mk to avoid the "NETBSDSRCDIR is nothing" syndrome

Add some kernel namespace (i.e. _KERNEL) helpers for regression tests.

add a vm allocator test which tests that:
* RUMP_MEMLIMIT works
* allocating memory with PR_NOWAIT will fail immediately if no memory
is available and PR_WAITOK will wait for memory to be available

Tweak the March 14th change to make page waits interlocked by pg->interlock.
Remove unneeded changes and only deal with the PQ_WANTED flag, to exclude
possible bugs.

Catch up with reality.

UVM locking changes, proposed on tech-kern:

- Change the lock on uvm_object, vm_amap and vm_anon to be a RW lock.
- Break v_interlock and vmobjlock apart. v_interlock remains a mutex.
- Do partial PV list locking in the x86 pmap. Others to follow later.

branches: 1.5.42;
Rename slightly misleading KTHREAD_JOINABLE to KTHREAD_MUSTJOIN.

vmobjlock is now a pointer to kmutex_t. use it as one.

branches: 1.3.4;
hold object lock across page unbusy

vm object must be locked during page allocation.

Add some kernel namespace (i.e. _KERNEL) helpers for regression tests.

workqueue(9) tests: Add test for PR kern/57574.

XXX pullup-10
XXX pullup-9
XXX pullup-8

branches: 1.8.2; 1.8.10;
Add a test case that exercises repeated sceduling and cancelling of a job,
with periodic dropping of the interlock.

Add rump-based test cases for threadpool(9).

branches: 1.6.2; 1.6.4;
Add a test case for workqueue_wait

Add declaration. build fix

sorry, I forgot to commit this file.

branches: 1.4.38;
elementary tests for rump sigmodels

Add some tests for lock errors. Notably, some of them fare nicely
with just the pthread consistency checks (which are enabled by
default in rumpuser), but others require real LOCKDEBUG.

add a vm allocator test which tests that:
* RUMP_MEMLIMIT works
* allocating memory with PR_NOWAIT will fail immediately if no memory
is available and PR_WAITOK will wait for memory to be available

Add some kernel namespace (i.e. _KERNEL) helpers for regression tests.

Add some tests for lock errors. Notably, some of them fare nicely
with just the pthread consistency checks (which are enabled by
default in rumpuser), but others require real LOCKDEBUG.

Move proc_lock into the data segment. It was dynamically allocated because
at the time we had mutex_obj_alloc() but not __cacheline_aligned.

elementary tests for rump sigmodels

Rename slightly misleading KTHREAD_JOINABLE to KTHREAD_MUSTJOIN.

Add some kernel namespace (i.e. _KERNEL) helpers for regression tests.

s/shold/should/

branches: 1.5.2;
Fix a silly bug the "cancel thrash" test: the exit condition for the
job was never set, and so if the job actually got started before the
cancellation request came in, it would never finish and the cancellation
request would just hang forever.

Should address a spurious automated test failure reported by kre@.

Add a test case that exercises repeated sceduling and cancelling of a job,
with periodic dropping of the interlock.

- De-opaque'ify struct threadpool_job.
- De-_t'ify all of the structure types.

No functional chage, no ABI change (verified with old rump unit test
before and after new librump.so).

Per Taylor's request.

branches: 1.2.2;
Include some non-critical-but-still-for-correctness lines of code
in the cancellation unit test. (I could have sworn I actually typed
these lines, no really...)

Add rump-based test cases for threadpool(9).

remove sys/simplelock.h (straggler)

branches: 1.3.6;
tsleep(9) no more.

branches: 1.2.2;
Rename slightly misleading KTHREAD_JOINABLE to KTHREAD_MUSTJOIN.

Add some kernel namespace (i.e. _KERNEL) helpers for regression tests.

workqueue(9) tests: Fix mistake in rev. 1.9.

Somehow, despite manually verifying a build/install/test of every
revision in my patch series, I managed to commit the wrong version of
the file for what became rev. 1.9, so the test was just broken when
it went in and remained broken in the commit where I fixed the real
bug and removed the xfail marker on the test.

PR kern/5757

XXX pullup-10
XXX pullup-9
XXX pullup-8

workqueue(9) tests: Add test for PR kern/57574.

XXX pullup-10
XXX pullup-9
XXX pullup-8

workqueue(9) tests: Destroy struct work immediately on entry.

workqueue(9) tests: Nix trailing whitespace.

branches: 1.6.2; 1.6.8; 1.6.16;
Fix build

Add a test case for workqueue_wait

Functionalize some routines to add new tests easily (NFC)

Fix a race condition on taking the mutex

The workqueue worker can take the mutex before the tester tries to take it after
calling workqueue_enqueue. If it happens, the worker calls cv_broadcast before
the tester calls cv_timedwait and the tester will wait until the cv timed out

Take the mutex before calling workqueue_enqueue so that the tester surely calls
cv_timedwait before the worker calls cv_broadcast.

The fix stabilizes the test, t_workqueue/workqueue1.

Tweak use of cv_timedwait

- Handle its return value
- Specify more appropriate time-out periods (2 ticks is too short)

Add simple test for workqueue(9)

Get rid of static Atffiles and let bsd.test.mk generate them on the fly.

Add a test which checks autoloading modules from the host's
/stand/arch/vers/kmods works in rump (and that the result is usable ;).

On i386 this "just works". For amd64, due to -mcmodel=kernel,
things are a little more complicated. We must have the entire rump
kernel loaded in the lower 2GB. Currently, this can be done either
by using the non-PIC version for the rump kernel compiled with
-mcmodel=small, or, as njoly pointed out, using netbsd32, which
causes vm_default_addr() to give something in the lower 2GB and
therefore shared libs "magically" getting loaded there. I guess
it would be possible to put a suggested vaddr into the rump kernel
libs and make ld.elf_so map memory from the suggested address if
present ... but that's another show.

Also thanks to tron for access to an amd64 so that I could verify
the test works.

tests/rump/modautoload: reorder -lrumpvfs_nofifofs

Make the way we spell "-lrumpvfs -lrumpvfs_nofifofs" hack consistent.
The order doesn't matter here because of --whole-archive.

Allow to set sanitizer flags for the rump build from the
build.sh command line, like: -V RUMP_SANITIZE=address

Fix alpha build (relocation truncated to fit: GPREL16 against symbol ...)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=47230

Do the same thing with linker flags instead of directly specifying the archives.

more nofifofs

Avoid symbol clashes in test/rump/modautoload/t_modautoload with sanitizers

Set SANITIZER_RENAME_SYMBOL.t_modautoload to:

* sysctlbyname
* sysctlgetmibinfo

branches: 1.8.14; 1.8.16;
Disable aslr too since we are playing address space range tricks.

need bsd.init.mk

don't disable PIE

needs mprotect disable since it loads kernel code and relocates it.

add missing library dependencies.

branches: 1.3.2;
When MODULAR kernels grew a MODULAR_DEFAULT_AUTOLOAD option, (in rev 1.98
of src/sys/kern/kern_module.c), the default was "off" for all kernels
including rump kernels. While many (most?) kernel config files were
updated to include the new option, rump kernels weren't so lucky. Thus,
rump kernels still had autoload disabled.

This commit uses rump_sysctl to change the module_autoload_on value to
true (ie, enabled) before trying to test if autoloading actually works.

For now, I am _not_ changing the default for all rump kernels. I'll
leave that for another day, after all appropriate discussion has occurred.

Move the "is arch capable of loading native kernel modules into
rump kernel" clauses from bsd.own.mk to Makefile.rump. Also,
add a rump_nativeabi_p() call to determine if rump kernel is
compiled with native ABI support.

branches: 1.1.6; 1.1.12;
Add a test which checks autoloading modules from the host's
/stand/arch/vers/kmods works in rump (and that the result is usable ;).

On i386 this "just works". For amd64, due to -mcmodel=kernel,
things are a little more complicated. We must have the entire rump
kernel loaded in the lower 2GB. Currently, this can be done either
by using the non-PIC version for the rump kernel compiled with
-mcmodel=small, or, as njoly pointed out, using netbsd32, which
causes vm_default_addr() to give something in the lower 2GB and
therefore shared libs "magically" getting loaded there. I guess
it would be possible to put a suggested vaddr into the rump kernel
libs and make ld.elf_so map memory from the suggested address if
present ... but that's another show.

Also thanks to tron for access to an amd64 so that I could verify
the test works.

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

Fill old_len which is required by sysctlbyname

branches: 1.4.2;
Remove a left-over debugging variable. Fix the build.

When MODULAR kernels grew a MODULAR_DEFAULT_AUTOLOAD option, (in rev 1.98
of src/sys/kern/kern_module.c), the default was "off" for all kernels
including rump kernels. While many (most?) kernel config files were
updated to include the new option, rump kernels weren't so lucky. Thus,
rump kernels still had autoload disabled.

This commit uses rump_sysctl to change the module_autoload_on value to
true (ie, enabled) before trying to test if autoloading actually works.

For now, I am _not_ changing the default for all rump kernels. I'll
leave that for another day, after all appropriate discussion has occurred.

Move the "is arch capable of loading native kernel modules into
rump kernel" clauses from bsd.own.mk to Makefile.rump. Also,
add a rump_nativeabi_p() call to determine if rump kernel is
compiled with native ABI support.

branches: 1.1.6; 1.1.12;
Add a test which checks autoloading modules from the host's
/stand/arch/vers/kmods works in rump (and that the result is usable ;).

On i386 this "just works". For amd64, due to -mcmodel=kernel,
things are a little more complicated. We must have the entire rump
kernel loaded in the lower 2GB. Currently, this can be done either
by using the non-PIC version for the rump kernel compiled with
-mcmodel=small, or, as njoly pointed out, using netbsd32, which
causes vm_default_addr() to give something in the lower 2GB and
therefore shared libs "magically" getting loaded there. I guess
it would be possible to put a suggested vaddr into the rump kernel
libs and make ld.elf_so map memory from the suggested address if
present ... but that's another show.

Also thanks to tron for access to an amd64 so that I could verify
the test works.

Get rid of static Atffiles and let bsd.test.mk generate them on the fly.

Add regression test to verify that linksets work correctly with rump.

branches: 1.1.2;
add regression test for module init/fini in rump

tests/rump/rumpkern: Use PROGDPLIBS, not explicit -L/-l.

This way we relink the t_* test programs whenever changes under
tests/rump/kernspace change libkernspace.a.

branches: 1.19.8;
Centralize the base rump libraries into a variable used by all the other
Makefiles so that we can make changes to it centrally as needed and have
less mess. Fixes the sun2 build that needs rumpvfs after librump after
the latest changes.

branches: 1.18.2;
Add -lrump after -lkernspace, because kernspace.a references symbols
from librump and hooray for static linking semantics.

Fixes sun2 build issue reported by kre@.

(XXX WTF did this only start failing after the addition of t_threadpool?)

Add rump-based test cases for threadpool(9).

branches: 1.16.2; 1.16.4;
Add simple test for workqueue(9)

branches: 1.15.16;
Fix static linking for the tests: -lrump is also used by -lrumpuser,
so we also need -lrump after -lrumpuser. Fixes build for sun2.

branches: 1.14.12; 1.14.22;
test remove signal delivery

elementary tests for rump sigmodels

Add some tests for lock errors. Notably, some of them fare nicely
with just the pthread consistency checks (which are enabled by
default in rumpuser), but others require real LOCKDEBUG.

Retire h_simpleserver and use rump_server instead.

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

Test copystr/copyinstr/copyoutstr focusing on the off-by-one cases.

As a notable caveat, this does not test the remote RPC paths of
those routines, which are arguably a little more complex.

tests for rump_lwproc

add tests for:
* threading
* tsleep variants
* uvm page busying and wanting

msdosfs depends on disk devices now

It seems that LDADD.file gets added to the linker invocation after
the LDADD variable expansion. To support static linking, ensure
that the common libraries gets tacked on at the end as well (order
matters for static linking), by using a convenience variable and
add it to the end of LDADD.t_modlinkset, as well as when doing the
general LDADD+= setting.

WARNS=4

Add regression test to verify that linksets work correctly with rump.

Apparently, there's not a clear dependency ordering for the libraries,
this becomes apparent when linking statically (e.g. as for sun2).
Add an extra instance of -lrump to the library list so that it can be
linked statically as well.

branches: 1.1.2;
add regression test for module init/fini in rump

bridge with the higher Makefile.inc

New testcase which checks that rump copyin/copyout/copyinstr/copyoutstr
version return EFAULT for special NULL "user" address.
ok pooka@.

branches: 1.1.6; 1.1.12;
Test copystr/copyinstr/copyoutstr focusing on the off-by-one cases.

As a notable caveat, this does not test the remote RPC paths of
those routines, which are arguably a little more complex.

lockme_DOUBLEINIT:
The failure message differs slightly when using LOCKDEBUG, modify the
expected pattern to cover both variants.

Fix detection of expected results. The rump kernel code apparently
includes source-code line numbers in the messages, so rather than
hard-coding them in the test, just use a reg-ex to match the text.

Fixes PR bin/52207

branches: 1.4.4;
Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.3.22;
Adjust MEMFREE expected output to recent kmem(9) changes.

branches: 1.2.6;
In case of LOCKDEBUG, expect certain text in the kernel output.

Add some tests for lock errors. Notably, some of them fare nicely
with just the pthread consistency checks (which are enabled by
default in rumpuser), but others require real LOCKDEBUG.

Hopefully fix some problems seen with MP support on non-x86, in particular
where curcpu() is defined as curlwp->l_cpu:

- mi_switch(): undo the ~2007ish optimisation to unlock curlwp before
calling cpu_switchto(). It's not safe to let other actors mess with the
LWP (in particular l->l_cpu) while it's still context switching. This
removes l->l_ctxswtch.

- Move the LP_RUNNING flag into l->l_flag and rename to LW_RUNNING since
it's now covered by the LWP's lock.

- Ditch lwp_exit_switchaway() and just call mi_switch() instead. Everything
is in cache anyway so it wasn't buying much by trying to avoid saving old
state. This means cpu_switchto() will never be called with prevlwp ==
NULL.

- Remove some KERNEL_LOCK handling which hasn't been needed for years.

branches: 1.9.14;
Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

more tests needing <sys/stat.h>

branches: 1.7.2;
call previous with the correct args too

check that setuid() works for procs with >1 lwp

+ rump_lwproc_newproc -> rump_lwproc_rfork()
+ add a tess for rump_lwproc_rfork()

implicit pid is 1 instead of 0 now

+tc (turns out the bug was elsewhere, but a test is always a test)

check that curlwp for an implicit context is NULL

tests for rump_lwproc

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.9.28;
hmm, using unifdef for the previous might have been a good idea ...

Now that atf gdb/coredump conflicts have been solved (thanks jmmv!),
remove USE_ATF ifdefs.

convert to ifdef USE_ATF

Disable module autoload so that it won't foil our module unload test.

Update to use newstyle rump_module interfaces.

WARNS=4

rump_module -> rump_pub_module

branches: 1.2.2;
don't hardcode /usr/lib path

add regression test for module init/fini in rump

fix various typos in comments, docs and log messages.

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.2.28;
WARNS=4

Add regression test to verify that linksets work correctly with rump.

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.2.26;
explicitly set the signal model

branches: 1.1.2;
elementary tests for rump sigmodels

tests/rump/rumpkern/t_sp: Go fishing for a rump_server core dump.

librumpuser: New RUMPUSER_DEBUG env var enables debug output.

Use it in t_sp:sigsafe to see if we can get more diagnostics out.

branches: 1.19.2;
Temporarily make PR t_sp:sigsafe noisier.

This is an attempt to diagnose why it's flaky in the releng testbeds
but not when I run it. This change will be backed out once we get
new testbed output.

rumpkern/t_sp: Save stdout and stderr to report on failure.

This is not very tidy at the moment, but it's an experiment using the
approach in PR bin/58112, to diagnose why t_sp:sigsafe keeps failing
intermittently. If this works, perhaps can tidy it up and apply it
to all the other tests that run rump_servers.

Format PR references consistently so that they can be automatically
turned into links in HTML reports.

Also skip the stress_short test - it just needs way too long to timeout
and fail.

Skip a few tests with reference to already existing PRs as after "recent"
scheduler changes these tests now leave rump_server processes around that
eat CPU and disturb later tests.

Get rid of all the -lrumpdev and -lrumpvfs that are no longer needed
after moving rump's mainbus from rumpdev to rumpkern.

Produces the same atf-run results as before.

branches: 1.13.14;

+ -lrumpdev

On systems that have only 32M of memory, the stress_short and
stress_long tests thrash a lot, and the stress_long test sometimes
times out, so skip them if we don't have at least 64M.

in reconnect, ignore any all reconnect messages

test for rumpclient reconnect feature

branches: 1.9.2;
test remove signal delivery

Execute a variety of syscalls in the stress test, including opening
files and sockets.

* run the killer stress test
* reduce default runtime to 5s

Make sure stressclient worker threads complete their operation
instead of hanging.

test rumpclient syscalls from a signal handler

Some tests for rumpclient fork.

Retire h_simpleserver and use rump_server instead.

use rump.halt in cleanup

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

branches: 1.2.2;
Add a test case that exercises repeated sceduling and cancelling of a job,
with periodic dropping of the interlock.

branches: 1.1.2;
Add rump-based test cases for threadpool(9).

Run both tests (rather than rumptest_threadjoin twice)

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.1.28;
add tests for:
* threading
* tsleep variants
* uvm page busying and wanting

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.1.28;
add tests for:
* threading
* tsleep variants
* uvm page busying and wanting

tests/rump/rumpkern/t_vm: Re-enable test as xfail.

We do, after all, have a way to mark a test as expecting a crash,
namely, with atf_tc_expect_signal(SIGABRT, ...). Also cite the PR
while here.

It is unclear to me what the bug in the test is. Is it supposed to
get the page via uobj->uo_pager->pgo_get (perhaps in turn via
uvm_obj_wirepages) instead of uvm_pagealloc? This should be
explained more clearly in the message (or just fixed).

PR kern/55945: rump/rumpkern/t_vm:busypage fails since early December

branches: 1.7.2;

PR misc/57343 (Jim Spath) - fixed a typo in atf_tc_skip message.

for the busypage test, replace atf_tc_expect_fail() with atf_tc_skip()
because atf apparently has no way to expect a test program to crash.
fixes PR 55945.

the busypage test is buggy, expect it to fail.

branches: 1.4.16;
Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.3.14;
Starting with Rev. 1.191 of kern/subr_pool.c a pool has to be inactive
for at least 10 seconds before it can be reclaimed.

Change the uvmwait test timeout from 10 to 30 seconds so it has a chance
to reclaim memory and succeed.

branches: 1.2.6; 1.2.8;
add a vm allocator test which tests that:
* RUMP_MEMLIMIT works
* allocating memory with PR_NOWAIT will fail immediately if no memory
is available and PR_WAITOK will wait for memory to be available

add tests for:
* threading
* tsleep variants
* uvm page busying and wanting

workqueue(9): Avoid touching running work items in workqueue_wait.

As soon as the workqueue function has called, it is forbidden to
touch the struct work passed to it -- the function might free or
reuse the data structure it is embedded in.

So workqueue_wait is forbidden to search the queue for the batch of
running work items. Instead, use a generation number which is odd
while the thread is processing a batch of work and even when not.

There's still a small optimization available with the struct work
pointer to wait for: if we find the work item in one of the per-CPU
_pending_ queues, then after we wait for a batch of work to complete
on that CPU, we don't need to wait for work on any other CPUs.

PR kern/57574

XXX pullup-10
XXX pullup-9
XXX pullup-8

workqueue(9) tests: Add test for PR kern/57574.

XXX pullup-10
XXX pullup-9
XXX pullup-8

branches: 1.2.2; 1.2.8; 1.2.16;
Add a test case for workqueue_wait

Add simple test for workqueue(9)

test for rumpclient reconnect feature

branches: 1.4.2;
test rumpclient syscalls from a signal handler

Some tests for rumpclient fork.

h_reboot has been superceded by rump.halt

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

PR misc/54382: whenever open(2) is called with O_CREAT, make sure to
pass an open mode argument.

branches: 1.1.46;
Some tests for rumpclient fork.

h_reboot has been superceded by rump.halt

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

set reconnect property now that it's off by default

branches: 1.1.2;
test for rumpclient reconnect feature

Limit test duration based on wall time instead of loops
(took a long time on slow platforms).

branches: 1.2.2;
check that we actually go into the signal handler

test rumpclient syscalls from a signal handler

test remove signal delivery

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

fix fallout from including signal.h in rump_syscallargs.h

Execute a variety of syscalls in the stress test, including opening
files and sockets.

Give the stress queen two modes: one which nicely suggests an exit
and another one which keeps moet et chandon in her pretty cabinet.

Establish signal handler already in parent. Otherwise the child
might be killed before it has a chance to run.

Print some more info in an attempt to try to understand why the
test fails in automated test runs but not my desktop.

Make sure stressclient worker threads complete their operation
instead of hanging.

Don't use printf in a signal handler.

XXX: it would be nice if the deadlock with malloc were a little
more obvious, especially since gdb doesn't provide any clues unless
you compile libpthread with -g

check that child exits with an expected status

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

Centralize the base rump libraries into a variable used by all the other
Makefiles so that we can make changes to it centrally as needed and have
less mess. Fixes the sun2 build that needs rumpvfs after librump after
the latest changes.

Deal with fallout from the addition of
KERN_PROC_CWD in sysctl(3)
That is kern.proc.$$.KERN_PROC_CWD (I think - not that it matters here)

The effect is that -lrump now requires -lrumpvfs

This set of changes fixes (I believe) regular dynamic builds,
more might be required for static builds (will be verified soon).

branches: 1.5.16;
Move sysproxy support into a separate component, rumpkern_sysproxy,
instead of it being always provided by the rump kernel base. This
move accomplishes two things:

1) it is no longer necessary to provide sysproxy hypercall stubs for
platforms which do not want to use sysproxy
2) it is easier to reason about the security aspects, since configurations
not linking the sysproxy component simply do not support remote
system calls

discussed on rumpkernel-users

Fix static linking for the tests: -lrump is also used by -lrumpuser,
so we also need -lrump after -lrumpuser. Fixes build for sun2.

branches: 1.3.12; 1.3.22;
test remove signal delivery

Retire h_simpleserver and use rump_server instead.

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

include <stdbool.h> for bool

test remove signal delivery

Retire h_simpleserver and use rump_server instead.

Add a simple and stress test cases for the rump remote syscall code.
Fun fact: since the server daemonizes into its own process group,
it's not in atf's pgrp. Use rump_sys_reboot() in atf cleanup to
kill the server after the test.

Add a test checks that 16 rump kernels on one shmif bus can ping
each other.

Disable reverse DNS lookups in ping to avoid spurious test failures
due to unresponsive DNS servers.

Get rid of all the -lrumpdev and -lrumpvfs that are no longer needed
after moving rump's mainbus from rumpdev to rumpkern.

Produces the same atf-run results as before.

branches: 1.3.14;

+ -lrumpdev

Use 8 TCP/IP stacks instead of 16. That still gives us plenty crossping
testing for 1/4th of the cost.

branches: 1.1.4; 1.1.10;
Add a test checks that 16 rump kernels on one shmif bus can ping
each other.

Get rid of static Atffiles and let bsd.test.mk generate them on the fly.

Add simple etfs test which almost tests for bug fixed yesterday
(the yesterday bug was against an emulated block device, so slighlty
different). But at least this test uncovered a few bugs already.

Centralize the base rump libraries into a variable used by all the other
Makefiles so that we can make changes to it centrally as needed and have
less mess. Fixes the sun2 build that needs rumpvfs after librump after
the latest changes.

branches: 1.4.24;
Fix static linking for the tests: -lrump is also used by -lrumpuser,
so we also need -lrump after -lrumpuser. Fixes build for sun2.

branches: 1.3.10; 1.3.20;
test lseek return value (off_t, might be >register_t)

branches: 1.2.2;
test makecn/freecn

Add simple etfs test which almost tests for bug fixed yesterday
(the yesterday bug was against an emulated block device, so slighlty
different). But at least this test uncovered a few bugs already.

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.2.2; 1.2.28;
One more lseek test case: in case of an error, check rv == -1
(so that it's not e.g UINT_MAX).

test lseek return value (off_t, might be >register_t)

t_etfs: Skip large_blk for system with page size > 8192

Work around kernel freeze reported as PR kern/55658 for
powerpc/ibm4xx (16K pages)

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.10.6;
CID 978337: Fix resource leak

branches: 1.9.6; 1.9.12; 1.9.22;
one more test

check etfs key rules

Adjusts tests after import of atf-0.12:

- The use.fs property is gone.
- Mark the tests/fs/t_create:attrs test as broken when using the default
unprivileged-user:_atf setting. This probably deserves a fix somehow
but I'm not sure at this point.

check etfs mappings with offsets

Check that etfs block mapping works for >2TB devices.

... and remove the not-supposed-to-be-committed call to rump_sys_reboot()

fix pasto, sprinkle ATF_REQUIRE_EQ()

ATF_REQUIRE_ERRNO, oh ATF_REQUIRE_ERRNO, wherefore art thou ATF_REQUIRE_ERRNO?

add test for block device support

Add simple etfs test which almost tests for bug fixed yesterday
(the yesterday bug was against an emulated block device, so slighlty
different). But at least this test uncovered a few bugs already.

Lock vnode for VOP_LOOKUP().

Don't play with "../.." in includes for h_macros.h; deal with it centrally.
Minor fixes.

branches: 1.5.2;
Don't include <rump/rumpvnode_if.h> from rump.h. It's not needed
unless you're doing something special, but requires register_t.
Adjust the few places which actually need rumpvnode_if.h.

Change vnode operation lookup to return the resulting vnode *vpp unlocked.
Change cache_lookup() to return an unlocked vnode.

Discussed on tech-kern@

Welcome to 6.99.31

branches: 1.3.6; 1.3.12;
+ rump_lwproc_newproc -> rump_lwproc_rfork()
+ add a tess for rump_lwproc_rfork()

Use a larger array to avoid overflow.

from Piotr Meyer

test makecn/freecn