Test Action Reference

The pipeline jobs (LAVA v2) have retained compatibility with respect to the content of Lava-Test-Shell Test Definitions although the submission format has changed:

  1. The test action will never boot the device - a boot action must be specified. Multiple test operations need to be specified as multiple definitions listed within the same test block.

  2. The LAVA support scripts are prepared by the deploy action action and the same scripts will be used for all test definitions until next deploy block is encountered.

There are 3 types of test actions:

  • lava-test-shell definitions (YAML directive: definitions) are used for POSIX compliant operating systems on the DUT. The deployed system is expected to support a POSIX shell environment (/bin/ash, /bin/dash or /bin/bash are the most common) so that LAVA can execute the LAVA Test Shell Helper scripts.

  • Output monitors (YAML directive: monitors) are used for devices which have no POSIX shell and start the test (and corresponding output) immediately after booting, for example microcontroller/IoT boards.

  • Interactive tests (YAML directive: interactive) are further extension of “monitor” tests idea, allowing not just matching some output from a device, but also feeding some input. They are useful for non-POSIX shells like bootloaders (u-boot for instance) and other interactive command-line applications.

Definitions

repository

A publicly readable repository location.

from

The type of the repository is not guessed, it must be specified explicitly. Support is available for git. Support is planned for url and tar.

git

A remote git repository which needs to be cloned by the dispatcher.

inline

A simple test definition present in the same file as the job submission, instead of from a separate file or VCS repository. This allows tests to be run based on a single file. When combined with file:// URLs to the deploy parameters, this allows tests to run without needing external access. See Inline test definition example.

path

The path within that repository to the YAML file containing the test definition.

name

(required) - replaces the name from the YAML.

params

(optional): Pass parameters to the Lava Test Shell Definition. The format is a YAML dictionary - the key is the name of the variable to be made available to the test shell, the value is the value of that variable.

- test:
    definitions:
    - repository: https://gitlab.com/lava/hacking-session.git
      from: git
      path: hacking-session-debian.yaml
      name: hacking
      params:
        IRC_USER: ""
        PUB_KEY: ""
- test:
    definitions:
    - repository: git://git.linaro.org/lava-team/lava-functional-tests.git
      from: git
      path: lava-test-shell/smoke-tests-basic.yaml
      name: smoke-tests
    - repository: https://git.linaro.org/lava-team/lava-functional-tests.git
      from: git
      path: lava-test-shell/single-node/singlenode03.yaml
      name: singlenode-advanced

Skipping elements of test definitions

When a single test definition is to be used across multiple deployment types (e.g. Debian and OpenEmbedded), it may become necessary to only perform certain actions within that definition in specific jobs. The skip_install support has been migrated from V1 for compatibility. Other methods of optimizing test definitions for specific deployments may be implemented in V2 later.

The available steps which can be (individually) skipped are:

deps

skip running lava-install-packages for the deps: list of the install: section of the definition.

keys

skip running lava-add-keys for the keys: list of the install: section of the definition.

sources

skip running lava-add-sources for the sources: list of the install: section of the definition.

steps

skip running any of the steps: of the install: section of the definition.

all

identical to ['deps', 'keys', 'sources', 'steps']

Example syntax:

- test:
    failure_retry: 3
    name: kvm-basic-singlenode
    timeout:
      minutes: 5
    definitions:
    - repository: git://git.linaro.org/lava-team/lava-functional-tests.git
      from: git
      path: lava-test-shell/smoke-tests-basic.yaml
      name: smoke-tests
    - repository: http://git.linaro.org/lava-team/lava-functional-tests.git
      skip_install:
      - all
      from: git
      path: lava-test-shell/single-node/singlenode03.yaml
      name: singlenode-advanced

The following will skip dependency installation and key addition in the same definition:

- test:
    failure_retry: 3
    name: kvm-basic-singlenode
    timeout:
      minutes: 5
    definitions:
    - repository: git://git.linaro.org/lava-team/lava-functional-tests.git
      from: git
      path: lava-test-shell/smoke-tests-basic.yaml
      name: smoke-tests
    - repository: http://git.linaro.org/lava-team/lava-functional-tests.git
      skip_install:
      - deps
      - keys
      from: git
      path: lava-test-shell/single-node/singlenode03.yaml
      name: singlenode-advanced

Inline test definition example

https://gitlab.com/lava/lava/-/raw/master/tests/lava_dispatcher/sample_jobs/kvm-inline.yaml

- test:
    failure_retry: 3
    definitions:
    - repository:
        metadata:
          format: Lava-Test Test Definition 1.0
          name: smoke-tests-basic
          description: "Basic system test command for Linaro Ubuntu images"
          os:
          - ubuntu
          scope:
          - functional
          devices:
          - panda
          - panda-es
          - arndale
          - vexpress-a9
          - vexpress-tc2
        run:
          steps:
          - lava-test-case linux-INLINE-pwd --shell pwd
          - lava-test-case linux-INLINE-uname --shell uname -a
          - lava-test-case linux-INLINE-vmstat --shell vmstat
          - lava-test-case linux-INLINE-ifconfig --shell ifconfig -a
          - lava-test-case linux-INLINE-lscpu --shell lscpu
          - lava-test-case linux-INLINE-lsusb --shell lsusb
          - lava-test-case linux-INLINE-lsb_release --shell lsb_release -a
      from: inline
      name: smoke-tests-inline
      path: inline/smoke-tests-basic.yaml

Additional support

The V2 dispatcher supports some additional elements in Lava Test Shell which will not be supported in the older V1 dispatcher.

Result checks

LAVA collects results from internal operations, these form the lava test suite results as well as from the submitted test definitions. The full set of results for a job are available at:

results/1234

LAVA records when a submitted test definition starts execution on the test device. If the number of test definitions which started is not the same as the number of test definitions submitted (allowing for the lava test suite results), a warning will be displayed on this page.

TestSets

A TestSet is a group of lava test cases which will be collated within the LAVA Results. This allows queries to look at a set of related test cases within a single definition.

- test:
   definitions:
   - repository:
       run:
         steps:
         - lava-test-set start first_set
         - lava-test-case date --shell ntpdate-debian
         - ls /
         - lava-test-case mount --shell mount
         - lava-test-set stop
         - lava-test-case uname --shell uname -a

This results in the date and mount test cases being included into a first_set TestSet, independent of other test cases. The TestSet is concluded with the lava-test-set stop command, meaning that the uname test case has no test set, providing a structure like:

results:
  first_set:
    date: pass
    mount: pass
  uname: pass
{'results': {'first_set': {'date': 'pass', 'mount': 'pass'}, 'uname': 'pass'}}

Each TestSet name must be valid as a URL, which is consistent with the requirements for test definition names and test case names in the V1 dispatcher.

For TestJob 1234, the uname test case would appear as:

results/1234/testset-def/uname

The date and mount test cases are referenced via the TestSet:

results/1234/testset-def/first_set/date
results/1234/testset-def/first_set/mount

A single test definition can start and stop different TestSets in sequence, as long as the name of each TestSet is unique for that test definition.

Interactive

An interactive test action allows to interact with a non-POSIX shell or just arbitrary interactive application. For instance, the shell of u-boot bootloader.

The workflow of the interactive test action is:

  • send the command to the DUT, unless empty

  • if echo: discard is specified, discard next output line (assumed to be an echo of the command)

  • wait for the prompts, successes or failures

  • if a name is defined, log the result for this command (as soon as a prompt or a message is matched)

  • if a successes or failures was matched, wait for the prompts

Note

The interactive test action expects the prompt to be already matched before it starts. If this is not the case, then wait for the prompt by adding an empty command directive as described below.

A u-boot interactive test might look like:

- test:
    interactive:
    - name: network
      prompts: ["=>", "/ # "]
      echo: discard
      script:
      - name: dhcp
        command: dhcp
        successes:
        - message: "DHCP client bound to address"
        failures:
        - message: "TIMEOUT"
          exception: InfrastructureError
          error: "dhcp failed"
      - name: setenv
        command: "setenv serverip {SERVER_IP}"
      - name: wait for the prompt
        command:

name

The name of the test suite.

prompts

The list of possible prompts for the interactive session. In many cases, there is just one prompt, but if shell has different prompts for different states, it can be accommodated. (Prompts can also include regexps, as any other match strings).

echo

If set to discard, after each sent command of a script, discard the next output line (assumed to be an echo of the command). This option should be set when interacting with shell (like u-boot shell) that will echo the command, to avoid false positive matches. Note that this options applies to every command in the script. If you need different value of this option for different commands, you would need to group them in different script’s.

script

The list of commands to send and what kind of output to expect:

  • name: If present, log the result (pass/fail) of this command under the given name (as a testcase). If not present, and the command fails, the entire test will fail (with TestError).

  • command: The command (string) to send to device, followed by newline. The command can use variables that will be substituted with live data, like {SERVER_IP}. If value is empty (command: in YAML), nothing is sent, but output matching (prompts/successes/failures) will be performed as usual. (Note that empty command: is different from empty string command: "". In the latter case, just a newline will be sent to device.)

  • failures and successes: Each optional. If present, check the device output for the given patterns.

successes should be a list of dictionaries with just one key:

  • message: The string (or regexp) to match. Substring match is performed, so care should be taken to reliably encode the match pattern. (E.g. message: 4 would match “4” appearing anywhere in the output, e.g. “14” or “41”).

failures should be a list of dictionaries with:

  • message: The string (or regexp) to match. Substring match is performed.

  • exception (optional): If the message indicates a fatal problem, an exception can be raised, one of: InfrastructureError, JobError, TestError. If not present, the error is not fatal and will be recorded just as a failed testcase in test results. (If this is a named command; as mentioned above, failure of unnamed (“not a testcase”) command leads to implicit TestError).

  • error: if defined, the exception message which will appear in the job log

If successes is defined, but LAVA matches one of the prompts instead, an error will be recorded (following the logic that the lack of expected success output is an error). This means that in many cases you don’t need to specify failures - any output but the successes will be recorded as an error.

However, if successes is not defined, then matching a prompt will generate a passing result (this is useful for interactive commands which don’t generate any output on success; of course, in this case you would need to specify failures to catch them).

Monitors

Test jobs using Monitors must:

  1. Be carefully designed to automatically execute after boot.

  2. Emit a unique start string:

    1. Only once per boot operation.

    2. Before any test operation starts.

  3. Emit a unique end string:

    1. Only once per boot operation.

    2. After all test operations have completed.

  4. Provide a regular expression which matches all expected test output and maps the output to results without leading to excessively long test case names.

start and end strings will match part of a line but make sure that each string is long enough that it can only match once per boot.

If start does not match, the job will timeout with no results.

If end does not match, the job will timeout but the results (of the current boot) will already have been reported.

name

The name of the test suite.

- test:
    monitors:
    - name: tests
      start: BOOTING ZEPHYR
      end: PROJECT EXECUTION SUCCESSFUL
      pattern: '(?P<test_case_id>\d+ *- [^-]+) (?P<measurement>\d+) tcs = [0-9]+ nsec'
      fixupdict:
        PASS: pass
        FAIL: fail

If the device output is of the form:

***** BOOTING ZEPHYR OS v1.7.99 - BUILD: Apr 18 2018 10:00:55 *****
|-----------------------------------------------------------------------------|
|                            Latency Benchmark                                |
|-----------------------------------------------------------------------------|
|  tcs = timer clock cycles: 1 tcs is 12 nsec                                 |
|-----------------------------------------------------------------------------|
| 1 - Measure time to switch from ISR back to interrupted thread              |
| switching time is 107 tcs = 1337 nsec                                       |
|-----------------------------------------------------------------------------|

...

PROJECT EXECUTION SUCCESSFUL

The above regular expression can result in test case names like:

1_measure_time_to_switch_from_isr_back_to_interrupted_thread_switching_time_is

The raw data will be logged as:

test_case_id: 1 - Measure time to switch from ISR back to interrupted thread              |
| switching time is

Caution

Notice how the regular expression has not closed the match at the end of the “line” but has continued on to the first non-matching character. The test case name then concatenates all whitespace and invalid characters to a single underscore. LAVA uses pexpect to perform output parsing. pexpect docs explain how to find line ending strings: https://pexpect.readthedocs.io/en/stable/overview.html#find-the-end-of-line-cr-lf-conventions

r'(?P<test_case_id>\d+ *- [^-]+) (?P<measurement>\d+) tcs = [0-9]+ nsec'

The test_case_id will be formed from the match of the expression \d+ *- [^-]+ followed by a single space - but only if the rest of the expression matches as well.

The measurement will be taken from the match of the expression \d+ preceded by a single space and followed by the exact string tcs = `` which itself must be followed by a number of digits, then a single space and finally the **exact** string ``nsec - but only if the rest of the expression also matches.