* qtest improvements (test for crash found with the fuzzer, increase
  downtime in migration test, less verbose output when running w/o KVM)
* Improve handling of acceptance tests in the Gitlab-CI
* Run checkpatch.pl in the Gitlab-CI
* Improve the gitlab-pipeline-status script
* Misc patches (mark 'moxie' as deprecated, remove stale .gitignore files, ...)

# gpg: Signature made Tue 13 Oct 2020 11:49:06 BST
# gpg:                using RSA key 27B88847EEE0250118F3EAB92ED9D774FE702DB5
# gpg:                issuer "thuth@redhat.com"
# gpg: Good signature from "Thomas Huth <th.huth@gmx.de>" [full]
# gpg:                 aka "Thomas Huth <thuth@redhat.com>" [full]
# gpg:                 aka "Thomas Huth <huth@tuxfamily.org>" [full]
# gpg:                 aka "Thomas Huth <th.huth@posteo.de>" [unknown]
# Primary key fingerprint: 27B8 8847 EEE0 2501 18F3  EAB9 2ED9 D774 FE70 2DB5

* remotes/huth-gitlab/tags/pull-request-2020-10-13: (23 commits)
  scripts/ci/gitlab-pipeline-status: wait for pipeline creation
  scripts/ci/gitlab-pipeline-status: use more descriptive exceptions
  scripts/ci/gitlab-pipeline-status: handle keyboard interrupts
  scripts/ci/gitlab-pipeline-status: refactor parser creation
  scripts/ci/gitlab-pipeline-status: give early feedback on running pipelines
  scripts/ci/gitlab-pipeline-status: improve message regarding timeout
  scripts/ci/gitlab-pipeline-status: make branch name configurable
  gitlab: assign python helper files to GitLab maintainers section
  gitlab: add a CI job to validate the DCO sign off
  gitlab: add a CI job for running checkpatch.pl
  configure: fixes indent of $meson setup
  docs/system/deprecated: Mark the 'moxie' CPU as deprecated
  Remove superfluous .gitignore files
  MAINTAINERS: Ignore bios-tables-test in the qtest section
  Add a comment in bios-tables-test.c to clarify the reason behind approach
  softmmu/vl: Be less verbose about missing KVM when running the qtests
  tests/migration: Allow longer timeouts
  qtest: add fuzz test case
  Acceptance tests: show test report on GitLab CI
  Acceptance tests: do not show canceled test logs on GitLab CI
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

When called in wait mode, this script will also wait for the pipeline
to be get to a "running" state.  Because many more status may be seen
until a pipeline gets to "running", and those need to be handle too.

Reference: https://docs.gitlab.com/ee/api/pipelines.html#list-project-pipelines


Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20200904164258.240278-8-crosa@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

For two very different error conditions.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20200904164258.240278-7-crosa@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

So that exits based on user requests are handled more gracefully.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20200904164258.240278-6-crosa@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Out of the main function.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20200904164258.240278-5-crosa@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

When waiting for a pipeline to run and finish, it's better to give
early feedback, and then sleep and wait, than the other wait around.

Specially for the first iteration, it's frustrating to see nothing
while the script is sleeping.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20200904164258.240278-4-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

The script has its own timeout, which is about how long the script
will wait (when called with --wait) for the pipeline to complete, and
not necessarily for the pipeline to complete.

Hopefully this new wording will be clearer.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20200904164258.240278-3-crosa@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

With the utility function `get_local_staging_branch_commit()`, the
name of the branch is hard coded (including in the function name).

For extensibility reasons, let's make that configurable.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20200904164258.240278-2-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20200918132903.1848939-4-berrange@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

While checkpatch.pl can validate DCO sign off that job must always be
advisory only since it is expected that certain patches will fail some
code style rules.

We require the DCO sign off to be mandatory for all commits though, so
it benefits from being validated in a standalone job.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20200918132903.1848939-3-berrange@redhat.com>
[thuth: Use "stage: build" to let it run earlier]
Signed-off-by: Thomas Huth <thuth@redhat.com>

This job is advisory since it is expected that certain patches will fail
the style checks and checkpatch.pl provides no way to mark exceptions to
the rules.

Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20200918132903.1848939-2-berrange@redhat.com>
[thuth: Use "stage: build" to let it run earlier]
Signed-off-by: Thomas Huth <thuth@redhat.com>

convert these line from tab to space

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20201012234348.1427-2-luoyonggang@gmail.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

It is currently unclear whether anybody is still using the 'moxie' CPU,
and there are no images for testing available this CPU, so the code has
likely bit-rotten in the course of time. When I asked the maintainer
for information, I did not get a reply within four weeks yet (see
https://lists.gnu.org/archive/html/qemu-devel/2020-08/msg07201.html

).
The last Signed-off-by line from Anthony in our repo is from 2013,
so it seems like this code is rather unmaintained. Time to put it onto
the deprecation list to see whether somebody is still interested in this
code or whether we could remove it in a couple of releases.

Message-Id: <20200923171815.97801-1-thuth@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Since we are now always doing out-of-tree builds, these gitignore
files should not be necessary anymore.

Message-Id: <20200919133637.72744-1-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

I'm very often getting CC: on rather large patch series that
modify the ACPI stuff of either ARM or x86, just because the
bios-table-test is often slightly involved here. I can't say
much about ACPI, and the bios-table-test is already covered
by the ACPI section in MAINTAINERS, so I'd rather prefer to
not getting automatically CC-ed on such patch series anymore.
If people want my opinion about qtest-related changes, they
can still put me on CC manually.

Message-Id: <20201001042717.136033-1-thuth@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

A comment is added in bios-tables-test.c that explains the reasoning
behind the process of updating the ACPI table blobs when new tests are added
or old tests are modified or code is committed that affect tests. The
explanation would help future contributors follow the correct process when
making code changes that affect ACPI tables.

Signed-off-by: Ani Sinha <ani@anisinha.ca>
Acked-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20200929142501.1057-1-ani@anisinha.ca>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Some of the qtests use "-accel kvm -accel tcg" to run real guest code.
This causes some error messages when kvm is not available. We do not
really care about these messages since the fallback to tcg is expected
here. So let's silence them to avoid that they spoil the output of
the tests.

Unfortunately, we can not use the qtest_enabled() wrapper in this case,
since the qtest accelerator itself is not initialized. Thus we have to
test for the qtest_chrdev variable instead.

Message-Id: <20200710085020.28222-1-thuth@redhat.com>
Reviewed-by: Alexander Bulekov <alxndr@bu.edu>
Signed-off-by: Thomas Huth <thuth@redhat.com>

In travis, with gcov and gprof we're seeing timeouts; hopefully fix
this by increasing the test timeouts a bit, but for xbzrle ensure it
really does get a couple of cycles through to test the cache.

I think the problem in travis is we have about 2 host CPU threads,
in the test we have at least 3:
   a) The vCPU thread (100% flat out)
   b) The source migration thread
   c) The destination migration thread

if (b) & (c) are slow for any reason - gcov+gperf or a slow host -
then they're sharing one host CPU thread so limit the migration
bandwidth.

Tested on my laptop with:
   taskset -c 0,1 ./tests/qtest/migration-test -p /x86_64/migration

Reported-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20201008160330.130431-1-dgilbert@redhat.com>
[thuth: Move the #define to the right location]
Signed-off-by: Thomas Huth <thuth@redhat.com>

Currently the device fuzzer finds more and more issues.
For every fuzz case, we need not only the fixes but also
the corresponding test case. We can analysis the reproducer
for every case and find what happened in where and write
a beautiful test case. However the raw data of reproducer is not
friendly to analysis. It will take a very long time, even far more
than the fixes itself. So let's create a new file to hold all of
the fuzz test cases and just use the raw data to act as the test
case. This way nobody will be afraid of writing a test case for
the fuzz reproducer.

This patch adds the issue LP#1878263 test case.

Signed-off-by: Li Qiang <liq3ea@163.com>
Message-Id: <20200921160605.19329-1-liq3ea@163.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alexander Bulekov <alxndr@bu.edu>
[thuth: Slightly adjusted commit message, removed empty lines]
Signed-off-by: Thomas Huth <thuth@redhat.com>

Avocado will, by default, produce JUnit files.  Let's ask GitLab
to present those in the web UI.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20201009205513.751968-4-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Tests resulting in "CANCEL" in Avocado are usually canceled on
purpose, and are almost identical to "SKIP".  The logs for canceled
tests are adding a lot of noise to the logs being shown on GitLab CI,
and causing distraction from real failures.

As a side note, this "after script" is scheduled for removal once the
feature is implemented within Avocado itself.

Reference: https://github.com/avocado-framework/avocado/issues/4266


Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20201009205513.751968-3-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

On with certain versions of "pip", package installations will attempt
to create wheels.  And, on environments without a "complete" Python
installation (as described in the acceptance tests requirements docs),
that will fail.

pycdlib, starting with version 1.11.0, is now being made available
as wheels, so its instalation on those constrained environments is
now possible.

Buglink: https://bugs.launchpad.net/qemu/+bug/1897783


Reported-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20201009205513.751968-2-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

With 1000 runs, there is a non-negligible chance that the fuzzer can
trigger a crash. With this CI job, we care about catching build/runtime
issues in the core fuzzing code. Actual device fuzzing takes place on
oss-fuzz. For these purposes, only running one input should be
sufficient.

Signed-off-by: Alexander Bulekov <alxndr@bu.edu>
Suggested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20201002143524.56930-1-alxndr@bu.edu>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Use self-explicit NANOSECONDS_PER_SECOND definition instead
of a magic value.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20201011194918.3219195-5-f4bug@amsat.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

* qtest documentation improvements (Eduardo, myself)
* libqtest event buffering (Maxim)
* use RCU for list of children of a bus (Maxim)
* move more files to softmmu/ (myself)
* meson.build cleanups, qemu-storage-daemon fix (Philippe)

# gpg: Signature made Mon 12 Oct 2020 16:55:19 BST
# gpg:                using RSA key F13338574B662389866C7682BFFBD25F78C7AE83
# gpg:                issuer "pbonzini@redhat.com"
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>" [full]
# gpg:                 aka "Paolo Bonzini <pbonzini@redhat.com>" [full]
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4  E2F7 7E15 100C CD36 69B1
#      Subkey fingerprint: F133 3857 4B66 2389 866C  7682 BFFB D25F 78C7 AE83

* remotes/bonzini-gitlab/tags/for-upstream: (38 commits)
  meson: identify more sections of meson.build
  scsi/scsi_bus: fix races in REPORT LUNS
  virtio-scsi: use scsi_device_get
  scsi/scsi_bus: Add scsi_device_get
  scsi/scsi-bus: scsi_device_find: don't return unrealized devices
  device-core: use atomic_set on .realized property
  scsi: switch to bus->check_address
  device-core: use RCU for list of children of a bus
  device_core: use drain_call_rcu in in qmp_device_add
  scsi/scsi_bus: switch search direction in scsi_device_find
  qdev: add "check if address free" callback for buses
  qemu-iotests, qtest: rewrite test 067 as a qtest
  qtest: check that drives are really appearing and disappearing
  qtest: switch users back to qtest_qmp_receive
  device-plug-test: use qtest_qmp to send the device_del command
  qtest: remove qtest_qmp_receive_success
  qtest: Reintroduce qtest_qmp_receive with QMP event buffering
  qtest: rename qtest_qmp_receive to qtest_qmp_receive_dict
  meson.build: Re-enable KVM support for MIPS
  build-sys: fix git version from -version
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Add more headers that clarify code organization.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Currently scsi_target_emulate_report_luns iterates over the child device list
twice, and there is no guarantee that this list is the same in both iterations.

The reason for iterating twice is that the first iteration calculates
how much memory to allocate.  However if we use a dynamic array we can
avoid iterating twice, and therefore we avoid this race.

Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1866707



Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200913160259.32145-10-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-14-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

This will help us to avoid the scsi device disappearing
after we took a reference to it.

It doesn't by itself forbid case when we try to access
an unrealized device

Suggested-by: Stefan Hajnoczi <stefanha@gmail.com>
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200913160259.32145-9-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-13-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Add scsi_device_get which finds the scsi device
and takes a reference to it.

Suggested-by: Stefan Hajnoczi <stefanha@gmail.com>
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20200913160259.32145-8-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-12-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

The device core first places a device on the bus and then realizes it.
Make scsi_device_find avoid returing such devices to avoid
races in drivers that use an iothread (currently virtio-scsi)

Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1812399



Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200913160259.32145-7-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-11-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Some code might race with placement of new devices on a bus.
We currently first place a (unrealized) device on the bus
and then realize it.

As a workaround, users that scan the child device list, can
check the realized property to see if it is safe to access such a device.
Use an atomic write here too to aid with this.

A separate discussion is what to do with devices that are unrealized:
It looks like for this case we only call the hotplug handler's unplug
callback and its up to it to unrealize the device.
An atomic operation doesn't cause harm for this code path though.

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200913160259.32145-6-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-10-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-6-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

This fixes the race between device emulation code that tries to find
a child device to dispatch the request to (e.g a scsi disk),
and hotplug of a new device to that bus.

Note that this doesn't convert all the readers of the list
but only these that might go over that list without BQL held.

This is a very small first step to make this code thread safe.

Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200913160259.32145-5-mlevitsk@redhat.com>
[Use RCU_READ_LOCK_GUARD in more places, adjust testcase now that
 the delay in DEVICE_DELETED due to RCU is more consistent. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-9-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Soon, a device removal might only happen on RCU callback execution.
This is okay for device-del which provides a DEVICE_DELETED event,
but not for the failure case of device-add.  To avoid changing
monitor semantics, just drain all pending RCU callbacks on error.

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@gmail.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200913160259.32145-4-mlevitsk@redhat.com>
[Don't use it in qmp_device_del. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

This change will allow us to convert the bus children list to RCU,
while not changing the logic of this function

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200913160259.32145-2-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Check if an address is free on the bus before plugging in the
device.  This makes it possible to do the check without any
side effects, and to detect the problem early without having
to do it in the realize callback.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20201006123904.610658-5-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Test 067 from qemu-iotests is executing QMP commands to hotplug
and hot-unplug disks, devices and blockdevs.  Because the power
of the text-based test harness is limited, it is actually limiting
the checks that it does, for example by skipping DEVICE_DELETED
events.

tests/qtest already has a similar test, drive_del-test.c.
We can merge them, and even reuse some of the existing code in
drive_del-test.c.  This will improve the quality of the test by
covering DEVICE_DELETED events and testing multiple architectures
(therefore covering multiple PCI hotplug mechanisms as well as s390x
virtio-ccw).

The only difference is that the new test will always use null-co:// for
the medium rather than qcow2 or raw, but this should be irrelevant for
what the test is covering.  For example there are no "qemu-img check"
runs in 067 that would check that the file is properly closed.

The new tests requires PCI hot-plug support, so drive_del-test
is moved from qemu-system-ppc to qemu-system-ppc64.

Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Do not just trust the HMP commands to create and delete the drive, use
query-block to check that this is actually the case.

Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Let test use the new functionality for buffering events.
The only remaining users of qtest_qmp_receive_dict are tests
that fuzz the QMP protocol.

Tested with 'make check-qtest'.

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20201006123904.610658-4-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Simplify the code now that events are buffered.  There is no need
anymore to separate sending the command and retrieving the response.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>