Re: [PATCHv4 net-next 3/3] tools: ynl: add YNL test framework

[Donald Hunter <donald.hunter@gmail.com>]

Matthieu Baerts <matttbe@kernel.org> writes:

> Hi Donald,
>
> On 14/11/2025 12:46, Donald Hunter wrote:
>> Hangbin Liu <liuhangbin@gmail.com> writes:
>>>
>>> +cleanup() {
>>> +	if [[ -n "$testns" ]]; then
>>> +		ip netns exec "$testns" bash -c "echo $NSIM_ID > /sys/bus/netdevsim/del_device" 2>/dev/null || true
>>> +		ip netns del "$testns" 2>/dev/null || true
>>> +	fi
>>> +}
>>> +
>>> +# Check if ynl command is available
>>> +if ! command -v $ynl &>/dev/null && [[ ! -x $ynl ]]; then
>>> +	ktap_skip_all "ynl command not found: $ynl"
>>> +	exit "$KSFT_SKIP"
>>> +fi
>>> +
>>> +trap cleanup EXIT
>>> +
>>> +ktap_print_header
>>> +ktap_set_plan 9>> +setup
>>> +
>>> +# Run all tests
>>> +cli_list_families
>>> +cli_netdev_ops
>>> +cli_ethtool_ops
>>> +cli_rt_route_ops
>>> +cli_rt_addr_ops
>>> +cli_rt_link_ops
>>> +cli_rt_neigh_ops
>>> +cli_rt_rule_ops
>>> +cli_nlctrl_ops
>>> +
>>> +ktap_finished
>> 
>> minor nit: ktap_finished should probably be in the 'cleanup' trap handler
>
> @Donald: I don't think 'ktap_finished' should be called there: in case
> of errors with an early exit during the setup phase, the two scripts
> will call 'ktap_skip_all', then 'exit "$KSFT_SKIP"'. If 'ktap_finished'
> is called in the 'cleanup' trap, it will print a total with everything
> set to 0 and call 'exit' again with other values (and no effects). So I
> think it is not supposed to be called from the exit trap.

Okay, fair. I thought the goal was to always output totals. Looking at
ktap_helpers.sh I see that it can't output a meaningful skip count for
the skip_call case.

>
> Cheers,
> Matt