librados-dev is not required by QEMU directly, only librbd-dev.
glusterfs-common is not directly needed by QEMU.
QEMU uses ncursesw only on non-Windows hosts.
The clang package is clang 10.
flex and bison are not required by QEMU.
Standardize on nmap ncat implementation to match Fedora/CentOS.
Remove vim since it is not a build pre-requisite and no other containers
include it.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20210623142245.307776-10-berrange@redhat.com>
Message-Id: <20210709143005.1554-18-alex.bennee@linaro.org>
Since Docker Hub has started to enforce pull rate limits on clients, it
is preferrable to use project specific container registries where they
are available. Fedora and OpenSUSE projects provide registries.
The images in these registries are also refreshed on a more regular
basis than the ones in docker hub, so the package update should
generally be faster.
While CentOS also has a registry it is considerably outdated compared
to docker.io, and also only provides x86 images, while docker.io images
are multi-arch.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20210623142245.307776-4-berrange@redhat.com>
Message-Id: <20210709143005.1554-12-alex.bennee@linaro.org>
The test aborts and error message as the following be throwed:
"No such file or directory: '/var/tmp/qemu-migrate-{pid}.migrate",
when the unix socket migration test nearly done. The reason is
qemu removes the unix socket file after migration before
guestperf.py script do it. So pre-check if the socket file exists
when removing it to prevent the guestperf program from aborting.
See also commit f9cc00346d ("tests/migration: fix unix socket batch
migration").
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Hyman <huangy81@chinatelecom.cn>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Currently the crypto layer exposes support for a 'des-rfb'
algorithm which is just normal single-DES, with the bits
in each key byte reversed. This special key munging is
required by the RFB protocol password authentication
mechanism.
Since the crypto layer is generic shared code, it makes
more sense to do the key byte munging in the VNC server
code, and expose normal single-DES support.
Replacing cipher 'des-rfb' by 'des' looks like an incompatible
interface change, but it doesn't matter. While the QMP schema
allows any QCryptoCipherAlgorithm for the 'cipher-alg' field
in QCryptoBlockCreateOptionsLUKS, the code restricts what can
be used at runtime. Thus the only effect is a change in error
message.
Original behaviour:
$ qemu-img create -f luks --object secret,id=sec0,data=123 -o cipher-alg=des-rfb,key-secret=sec0 demo.luks 1G
Formatting 'demo.luks', fmt=luks size=1073741824 key-secret=sec0 cipher-alg=des-rfb
qemu-img: demo.luks: Algorithm 'des-rfb' not supported
New behaviour:
$ qemu-img create -f luks --object secret,id=sec0,data=123 -o cipher-alg=des-rfb,key-secret=sec0 demo.luks 1G
Formatting 'demo.luks', fmt=luks size=1073741824 key-secret=sec0 cipher-alg=des-fish
qemu-img: demo.luks: Invalid parameter 'des-rfb'
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The GNUTLS crypto provider doesn't support DES-ECB, only DES-CBC.
We can use the latter to simulate the former, if we encrypt only
1 block (8 bytes) of data at a time, using an all-zeros IV. This
is a very inefficient way to use the QCryptoCipher APIs, but
since the VNC authentication challenge is only 16 bytes, this
is acceptable. No other part of QEMU should be using DES. This
test case demonstrates the equivalence of ECB and CBC for the
single-block case.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Using error_fatal provides better diagnostics when tests
failed, than using asserts, because we see the text of
the error message.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Since we now require gcrypt >= 1.8.0, there is no need
to exclude the pbkdf test case.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The main method checks whether the cipher choice is supported
at runtime, so there is no need for compile time conditions.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
These tests' setUp do not do anything beyong what their base class do.
And while they do decorate the setUp() we can decorate the classes
instead, so no functionality is lost here.
This is possible because since Avocado 76.0 we can decorate setUp()
directly.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210415215141.1865467-4-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
[PMD: added note to commit message about Avocado feature/version]
Signed-off-by: Cleber Rosa <crosa@redhat.com>
The premise behind the original behavior is that it would save people
from downloading Avocado (and other dependencies) if already installed
on the system. To be honest, I think it's extremely rare that the
same versions described as dependencies will be available on most
systems. But, the biggest motivations here are that:
1) Hacking on QEMU in the same system used to develop Avocado leads
to confusion with regards to the exact bits that are being used;
2) Not reusing Python packages from system wide installations gives
extra assurance that the same behavior will be seen from tests run
on different machines;
With regards to downloads, pip already caches the downloaded wheels
and tarballs under ~/.cache/pip, so there should not be more than
one download even if the venv is destroyed and recreated.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210415215141.1865467-3-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Some test cases on x86_cpu_model_versions.py are corner cases because they
need to pass extra options to the -cpu argument. Once the avocado_qemu
framework will set -cpu automatically, the value should be reset. This changed
those tests so to call set_vm_arg() to overwrite the -cpu value.
Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Message-Id: <20210430133414.39905-8-wainersm@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
This introduces a new feature to the functional tests: automatic setting of
the '-cpu VALUE' option to the created vm if the test is tagged with
'cpu:VALUE'. The 'cpu' property is made available to the test object as well.
For example, for a simple test as:
def test(self):
"""
🥑 tags=cpu:host
"""
self.assertEqual(self.cpu, "host")
self.vm.launch()
The resulting QEMU evocation will be like:
qemu-system-x86_64 -display none -vga none \
-chardev socket,id=mon,path=/var/tmp/avo_qemu_sock_pdgzbgd_/qemu-1135557-monitor.sock \
-mon chardev=mon,mode=control -cpu host
Reviewed-by: Cleber Rosa <crosa@redhat.com>
Tested-by: Cleber Rosa <crosa@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20210430133414.39905-2-wainersm@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Avocado allows us to select set of tests using tags.
When wanting to run all tests using a NetBSD guest OS,
it is convenient to have them tagged, add the 'os:netbsd'
tag.
It allows one to run the NetBSD tests with:
$ avocado --show=app,console run -t os:netbsd tests/acceptance/
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210623180021.898286-4-f4bug@amsat.org>
Reviewed-by: Niek Linnenbank <nieklinnenbank@gmail.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Reviewed-by: Cleber Rosa <crosa@redhat.com>
[PMD: ammend the commit message with example command]
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Add new tests checking the good behavior of the SMMUv3 protecting
2 virtio pci devices (block and net). We check the guest boots and
we are able to install a package. Different guest configs are tested:
standard, passthrough an strict=0. This is tested with both fedora 31 and
33. The former uses a 5.3 kernel without range invalidation whereas the
latter uses a 5.8 kernel that features range invalidation.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Tested-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20210706131729.30749-4-eric.auger@redhat.com>
[CR: split long lines]
[CR: added MAINTAINERS entry]
Signed-off-by: Cleber Rosa <crosa@redhat.com>
When running LinuxTests we may need to run the guest with
custom params. It is practical to store the pxeboot URL
and the default kernel params so that the
tests just need to fetch those and augment the kernel params.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Message-Id: <20210706131729.30749-3-eric.auger@redhat.com>
[CR: split long lines]
Signed-off-by: Cleber Rosa <crosa@redhat.com>
As the KNOWN_DISTROS grows, more loosely methods will be created in
the avocado_qemu/__init__.py file.
Let's refactor the code so that KNOWN_DISTROS and related methods are
packaged in a class
Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Message-Id: <20210706131729.30749-2-eric.auger@redhat.com>
[CR: moved aarch64 definition from patch 2 to 1]
[CR: protect get() when arch is not defined]
[CR: split long lines]
Acked-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
If a guest sends binary data on the serial console, we get:
File "tests/acceptance/avocado_qemu/__init__.py", line 92,
in _console_interaction msg = console.readline().strip()
File "/usr/lib64/python3.8/codecs.py", line 322,
in decode (result, consumed) = self._buffer_decode(data, self.errors, final)
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xa9 in position 2: invalid start byte
Since we use the console with readline(), fix it the easiest
way possible: ignore binary data (all current tests compare
text string anyway).
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210515134555.307404-2-f4bug@amsat.org>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Tested-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
The tests based on the LinuxTest class give the test writer a ready to
use guest operating system, currently pinned to Fedora 31.
With this change, it's now possible to choose different distros and
versions, similar to how other tags and parameter can be set for the
target arch, accelerator, etc.
One of the reasons for this work, is that some development features
depend on updates on the guest side. For instance the tests on
virtiofs_submounts.py, require newer kernels, and may benefit from
running, say on Fedora 34, without the need for a custom kernel.
Please notice that the pre-caching of the Fedora 31 images done during
the early stages of `make check-acceptance` (before the tests are
actually executed) are not expanded here to cover every new image
added. But, the tests will download other needed images (and cache
them) during the first execution.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210414221457.1653745-4-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Instead of having, by default, the checksum in the tests, and the
definition of tests in the framework, let's keep them together.
A central definition for distributions is available, and it should
allow other known distros to be added more easily.
No behavior change is expected here, and tests can still define
a distro_checksum value if for some reason they want to override
the known distribution information.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210414221457.1653745-3-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Eric Auger <eric.auger@redhat.com>
[CR: split long lines]
Signed-off-by: Cleber Rosa <crosa@redhat.com>
This renames the attribute that holds the checksum for the image Linux
distribution image used.
The current name of the attribute is not very descriptive. Also, in
preparation for making the distribution used configurable, which will
add distro related parameters, attributes and tags, let's make the
naming of those more uniform.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210414221457.1653745-2-crosa@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
[CR: split long lines]
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Logs can be very important to debug issues, and currently QEMUMachine
instances will remove logs that are created under the temporary
directories.
With this change, the stdout and stderr generated by the QEMU process
started by QEMUMachine will always be kept along the test results
directory.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210211220146.2525771-6-crosa@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>
The recently-added NBD context qemu:allocation-depth is able to
distinguish between locally-present data (even when that data is
sparse) [shown as depth 1 over NBD], and data that could not be found
anywhere in the backing chain [shown as depth 0]; and the libnbd
project was recently patched to give the human-readable name "absent"
to an allocation-depth of 0. But qemu-img map --output=json predates
that addition, and has the unfortunate behavior that all portions of
the backing chain that resolve without finding a hit in any backing
layer report the same depth as the final backing layer. This makes it
harder to reconstruct a qcow2 backing chain using just 'qemu-img map'
output, especially when using "backing":null to artificially limit a
backing chain, because it is impossible to distinguish between a
QCOW2_CLUSTER_UNALLOCATED (which defers to a [missing] backing file)
and a QCOW2_CLUSTER_ZERO_PLAIN cluster (which would override any
backing file), since both types of clusters otherwise show as
"data":false,"zero":true" (but note that we can distinguish a
QCOW2_CLUSTER_ZERO_ALLOCATED, which would also have an "offset":
listing).
The task of reconstructing a qcow2 chain was made harder in commit
0da9856851 (nbd: server: Report holes for raw images), because prior
to that point, it was possible to abuse NBD's block status command to
see which portions of a qcow2 file resulted in BDRV_BLOCK_ALLOCATED
(showing up as NBD_STATE_ZERO in isolation) vs. missing from the chain
(showing up as NBD_STATE_ZERO|NBD_STATE_HOLE); but now qemu reports
more accurate sparseness information over NBD.
An obvious solution is to make 'qemu-img map --output=json' add an
additional "present":false designation to any cluster lacking an
allocation anywhere in the chain, without any change to the "depth"
parameter to avoid breaking existing clients. The iotests have
several examples where this distinction demonstrates the additional
accuracy.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210701190655.2131223-3-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
[eblake: fix more iotest fallout]
Signed-off-by: Eric Blake <eblake@redhat.com>
OSS-Fuzz found sending illegal addresses when querying the write
protection bits triggers an assertion:
qemu-fuzz-i386: hw/sd/sd.c:824: uint32_t sd_wpbits(SDState *, uint64_t): Assertion `wpnum < sd->wpgrps_size' failed.
==11578== ERROR: libFuzzer: deadly signal
#8 0x7ffff628e091 in __assert_fail
#9 0x5555588f1a3c in sd_wpbits hw/sd/sd.c:824:9
#10 0x5555588dd271 in sd_normal_command hw/sd/sd.c:1383:38
#11 0x5555588d777c in sd_do_command hw/sd/sd.c
#12 0x555558cb25a0 in sdbus_do_command hw/sd/core.c💯16
#13 0x555558e02a9a in sdhci_send_command hw/sd/sdhci.c:337:12
#14 0x555558dffa46 in sdhci_write hw/sd/sdhci.c:1187:9
#15 0x5555598b9d76 in memory_region_write_accessor softmmu/memory.c:489:5
Similarly to commit 8573378e62 ("hw/sd: fix out-of-bounds check
for multi block reads"), check the address range before sending
the status of the write protection bits.
Include the qtest reproducer provided by Alexander Bulekov:
$ make check-qtest-i386
...
Running test qtest-i386/fuzz-sdcard-test
qemu-system-i386: ../hw/sd/sd.c:824: sd_wpbits: Assertion `wpnum < sd->wpgrps_size' failed.
Reported-by: OSS-Fuzz (Issue 29225)
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/450
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210702155900.148665-4-f4bug@amsat.org>
Some qemu updates for IPMI and I2C
Move some ADC file to where they belong and move some sensors to a
sensor directory, since with new BMCs coming in lots of different
sensors should be coming in. Keep from cluttering things up.
Add support for I2C PMBus devices.
Replace the confusing and error-prone i2c_send_recv and i2c_transfer with
specific send and receive functions. Several errors have already been
made with these, avoid any new errors.
Fix the watchdog_expired field in the IPMI watchdog, it's not a bool,
it's a u8. After a vmstate transfer, the new value could be wrong.
# gpg: Signature made Fri 09 Jul 2021 17:25:04 BST
# gpg: using RSA key FD0D5CE67CE0F59A6688268661F38C90919BFF81
# gpg: Good signature from "Corey Minyard <cminyard@mvista.com>" [unknown]
# gpg: aka "Corey Minyard <minyard@acm.org>" [unknown]
# gpg: aka "Corey Minyard <corey@minyard.net>" [unknown]
# gpg: aka "Corey Minyard <minyard@mvista.com>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg: There is no indication that the signature belongs to the owner.
# Primary key fingerprint: FD0D 5CE6 7CE0 F59A 6688 2686 61F3 8C90 919B FF81
* remotes/cminyard/tags/for-qemu-6.1-2: (24 commits)
tests/qtest: add tests for MAX34451 device model
hw/misc: add MAX34451 device
tests/qtest: add tests for ADM1272 device model
hw/misc: add ADM1272 device
hw/i2c: add support for PMBus
ipmi/sim: fix watchdog_expired data type error in IPMIBmcSim struct
hw/i2c: Introduce i2c_start_recv() and i2c_start_send()
hw/i2c: Extract i2c_do_start_transfer() from i2c_start_transfer()
hw/i2c: Make i2c_start_transfer() direction argument a boolean
hw/i2c: Rename i2c_set_slave_address() -> i2c_slave_set_address()
hw/i2c: Remove confusing i2c_send_recv()
hw/misc/auxbus: Replace i2c_send_recv() by i2c_recv() & i2c_send()
hw/misc/auxbus: Replace 'is_write' boolean by its value
hw/misc/auxbus: Explode READ_I2C / WRITE_I2C_MOT cases
hw/misc/auxbus: Fix MOT/classic I2C mode
hw/i2c/ppc4xx_i2c: Replace i2c_send_recv() by i2c_recv() & i2c_send()
hw/i2c/ppc4xx_i2c: Add reference to datasheet
hw/display/sm501: Replace i2c_send_recv() by i2c_recv() & i2c_send()
hw/display/sm501: Simplify sm501_i2c_write() logic
hw/input/lm832x: Define TYPE_LM8323 in public header
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Block layer patches
- Make blockdev-reopen stable
- Remove deprecated qemu-img backing file without format
- rbd: Convert to coroutines and add write zeroes support
- rbd: Updated MAINTAINERS
- export/fuse: Allow other users access to the export
- vhost-user: Fix backends without multiqueue support
- Fix drive-backup transaction endless drained section
# gpg: Signature made Fri 09 Jul 2021 13:49:22 BST
# gpg: using RSA key DC3DEB159A9AF95D3D7456FE7F09B272C88F2FD6
# gpg: issuer "kwolf@redhat.com"
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>" [full]
# Primary key fingerprint: DC3D EB15 9A9A F95D 3D74 56FE 7F09 B272 C88F 2FD6
* remotes/kevin/tags/for-upstream: (28 commits)
block: Make blockdev-reopen stable API
iotests: Test reopening multiple devices at the same time
block: Support multiple reopening with x-blockdev-reopen
block: Acquire AioContexts during bdrv_reopen_multiple()
block: Add bdrv_reopen_queue_free()
qcow2: Fix dangling pointer after reopen for 'file'
qemu-img: Improve error for rebase without backing format
qemu-img: Require -F with -b backing image
qcow2: Prohibit backing file changes in 'qemu-img amend'
blockdev: fix drive-backup transaction endless drained section
vhost-user: Fix backends without multiqueue support
MAINTAINERS: add block/rbd.c reviewer
block/rbd: fix type of task->complete
iotests/fuse-allow-other: Test allow-other
iotests/308: Test +w on read-only FUSE exports
export/fuse: Let permissions be adjustable
export/fuse: Give SET_ATTR_SIZE its own branch
export/fuse: Add allow-other option
export/fuse: Pass default_permissions for mount
util/uri: do not check argument of uri_free()
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
