[PATCH 09/10] nvme/042: test bi-directional authentication

[Sagi Grimberg]

On 11/28/21 7:34 PM, Hannes Reinecke wrote:
> On 11/28/21 2:36 PM, Sagi Grimberg wrote:
>>
>>
>> On 11/23/21 9:49 AM, Hannes Reinecke wrote:
>>> Signed-off-by: Hannes Reinecke <hare at suse.de>
>>> ---
>>>   tests/nvme/042     | 107 +++++++++++++++++++++++++++++++++++++++++++++
>>>   tests/nvme/042.out |   8 ++++
>>>   2 files changed, 115 insertions(+)
>>>   create mode 100644 tests/nvme/042
>>>   create mode 100644 tests/nvme/042.out
>>>
>>> diff --git a/tests/nvme/042 b/tests/nvme/042
>>> new file mode 100644
>>> index 0000000..f3954b4
>>> --- /dev/null
>>> +++ b/tests/nvme/042
>>> @@ -0,0 +1,107 @@
>>> +#!/bin/bash
>>> +# SPDX-License-Identifier: GPL-3.0+
>>> +# Copyright (C) 2021 Hannes Reinecke, SUSE Labs
>>> +#
>>> +# Test bi-directional authentication
>>> +
>>> +. tests/nvme/rc
>>> +
>>> +DESCRIPTION="Test bi-directional authentication"
>>> +QUICK=1
>>> +
>>> +requires() {
>>> +    _nvme_requires
>>> +    _have_modules loop
>>> +    _require_nvme_trtype_is_fabrics
>>> +}
>>> +
>>> +
>>> +test() {
>>> +    local port
>>> +    local subsys="blktests-subsystem-1"
>>> +    local hostid="$(uuidgen)"
>>> +    local hostnqn="nqn.2014-08.org.nvmexpress:uuid:${hostid}"
>>> +    local scratch="/tmp/blktest-ns1.img"
>>> +    local hostkey
>>> +    local ctrlkey
>>> +    local ctrldev
>>> +
>>> +    echo "Running ${TEST_NAME}"
>>> +
>>> +    hostkey="$(nvme gen-dhchap-key -n ${subsys} 2> /dev/null)"
>>> +    if [ $? -ne 0 ] ; then
>>> +        echo "nvme gen-dhchap-key command missing"
>>> +        return 1
>>> +    fi
>>> +
>>> +    ctrlkey="$(nvme gen-dhchap-key -n ${subsys} 2> /dev/null)"
>>> +    if [ $? -ne 0 ] ; then
>>> +        echo "nvme gen-dhchap-key command missing"
>>> +        return 1
>>> +    fi
>>> +
>>> +    _setup_nvmet
>>> +
>>> +    truncate -s 512M "${scratch}"
>>> +
>>> +    port="$(_create_nvmet_port "${nvme_trtype}")"
>>> +
>>> +    _create_nvmet_subsystem "${subsys}" "${scratch}"
>>> +    _add_nvmet_subsys_to_port "${port}" "${subsys}"
>>> +    _create_nvmet_host "${subsys}" "${hostnqn}" "${hostkey}" 
>>> "${ctrlkey}"
>>> +
>>> +    _set_nvmet_dhgroup "${hostnqn}" "ffdhe2048"
>>> +
>>> +    # Step 1: Connect with host authentication only
>>> +    echo "Test host authentication"
>>> +    _nvme_connect_subsys "${nvme_trtype}" "${subsys}" \
>>> +                 "${def_traddr}" "${def_trsvcid}" \
>>> +                 "${hostnqn}" "${hostid}" \
>>> +                 "${hostkey}"
>>> +
>>> +    ctrldev=$(_find_nvme_dev "${subsys}")
>>> +    if [ -z "$ctrldev" ] ; then
>>> +        echo "nvme controller not found"
>>> +    fi
>>> +
>>> +    _nvme_disconnect_subsys "${subsys}"
>>> +
>>> +    # Step 2: Connect with host authentication
>>> +    # and invalid ctrl authentication
>>> +    echo "Test host authentication and invalid ctrl authentication"
>>> +    _nvme_connect_subsys "${nvme_trtype}" "${subsys}" \
>>> +                 "${def_traddr}" "${def_trsvcid}" \
>>> +                 "${hostnqn}" "${hostid}" \
>>> +                 "${hostkey}" "${hostkey}"
>>> +    ctrldev=$(_find_nvme_dev "${subsys}1")
>>> +    if [ -n "${ctrldev}" ] ; then
>>> +        echo "nvme controller found!"
>>> +        _nvme_disconnect_subsys "${ctrldev}"
>>> +    fi
>>
>> This test case has a weird behavior. Why isn't _nvme_connect_subsys
>> failing?
> 
> Oh, but it is.
> The line 'nvme controller found' is not part of the .out file, hence the 
> test will fail.
> 
> Call it an obscure feature of blktests :-)
> 
> But yeah, I can make it more obvious that this is the failure case.

Hannes, my question is why isn't nvme connect not failing (i.e.
returning a non-zero status code)?