Re: [PATCH 0/2] Modify the watchdog selftest for execution with

[Laura Nao <laura.nao@collabora.com>]

Hi Shuah,

On 6/7/24 01:03, Shuah Khan wrote:
> On 6/6/24 03:57, Laura Nao wrote:
>> Hi Shuah,
>>
>> On 5/6/24 13:13, Laura Nao wrote:
>>> The watchdog selftest script supports various parameters for testing
>>> different IOCTLs. The watchdog ping functionality is validated by
>>> starting
>>> a loop where the watchdog device is periodically pet, which can only
>>> be
>>> stopped by the user interrupting the script.
>>>
>>> This results in a timeout when running this test using the kselftest
>>> runner
>>> with no non-oneshot parameters (or no parameters at all):
> 
> 
> Sorry for the delay on this.
> 
> This test isn't include in the default kselftest run? How are you
> running this?
>

The goal of this series is to enable the test to be run using the
kselftest runner individually, not as part of the default run. So for
example w/out args:

make -C tools/testing/selftests TARGETS=watchdog run_tests

or with args:

KSELFTEST_WATCHDOG_TEST_ARGS='-b -d -e -s -t 12 -T 3 -n 7 -N -L' make -C
tools/testing/selftests TARGETS=watchdog run_tests
 
>>>
>>>    TAP version 13
>>>    1..1
>>>    # timeout set to 45
>>>    # selftests: watchdog: watchdog-test
>>>    # Watchdog Ticking Away!
>>>    # .............................................#
>>>    not ok 1 selftests: watchdog: watchdog-test # TIMEOUT 45 seconds
>>>
>>> To address this issue, the first patch in this series limits the
>>> loop
>>> to 5
>>> iterations by default and adds support for a new '-c' option to
>>> customize
>>> the number of pings as required.
>>>
>>> The second patch conforms the test output to the KTAP format.
>>>
>>
>> Gentle ping - any thoughts on this series? It would simplify running
>> the
>> watchdog kselftest in CI environments by leveraging the runner.
>>
> 
> This test isn't intended to be included in the default run. It
> requires
> loading a watchdog driver first. Do you load the driver from the
> runner?
>

I get this test requires watchdog drivers to be loaded (which in this
case can't be added to a config fragment that goes with the selftest, as
they are platform-specific) and therefore cannot be included in the
default run. However, having ktap output support and limiting the ping
loop would allow the test to be run individually in the same way as
other selftests (so through the kselftest runner). 

Naturally, driver dependencies must be met for the test to run and
produce valid results. From my understanding the runner itself cannot
ensure this, so in this case it would be up to the user or CI to
enable/load the appropriate drivers before running the test.
If these dependencies are not satisfied, the test could just exit
with a skip code.

Does this make sense to you? or is the kselftest runner intended to be
used to run exclusively a subset of tests in the selftests directory
(i.e. the ones that don't have platform-specific driver requirements)?

Thanks,

Laura