Re: [PATCH v2 3/4] nvme: Expose the tls_configured sysfs for secure concat connections

[Hannes Reinecke <hare@suse.de>]

On 11/13/25 13:22, Alistair Francis wrote:
> On Thu, Nov 13, 2025 at 5:21 PM Hannes Reinecke <hare@suse.de> wrote:
>>
>> On 11/13/25 03:08, Alistair Francis wrote:
>>> On Wed, Nov 12, 2025 at 5:08 PM Hannes Reinecke <hare@suse.de> wrote:
>>>>
>>>> On 11/12/25 00:45, alistair23@gmail.com wrote:
>>>>> From: Alistair Francis <alistair.francis@wdc.com>
>>>>>
>>>>> Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
>>>>> ---
>>>>> v2:
>>>>>     - New patch
>>>>>
>>>>>     drivers/nvme/host/sysfs.c | 2 +-
>>>>>     1 file changed, 1 insertion(+), 1 deletion(-)
>>>>>
>>>>> diff --git a/drivers/nvme/host/sysfs.c b/drivers/nvme/host/sysfs.c
>>>>> index 29430949ce2f..6d10e12136d0 100644
>>>>> --- a/drivers/nvme/host/sysfs.c
>>>>> +++ b/drivers/nvme/host/sysfs.c
>>>>> @@ -838,7 +838,7 @@ static umode_t nvme_tls_attrs_are_visible(struct kobject *kobj,
>>>>>             !ctrl->opts->tls && !ctrl->opts->concat)
>>>>>                 return 0;
>>>>>         if (a == &dev_attr_tls_configured_key.attr &&
>>>>> -         (!ctrl->opts->tls_key || ctrl->opts->concat))
>>>>> +         !ctrl->opts->concat)
>>>>>                 return 0;
>>>>>         if (a == &dev_attr_tls_keyring.attr &&
>>>>>             !ctrl->opts->keyring)
>>>>
>>>> ??
>>>>
>>>> How can you have a configured TLS Key for secure concatenation?
>>>
>>> I'm not sure I follow
>>>
>>> `ctrl->opts->tls_key` is directly set at the end of the
>>> `nvme_auth_secure_concat()` function, so it will be set for secure
>>> concatenation.
>>>
>> Right, sorry. Of course you are right.
>>
>> But I'm still a bit unsure about the interface here; just writing
>> anything into it doesn't feel like the correct way of doing it.
> 
> I'm happy to change that.
> 
> I was thinking a bool, so basically write 1 to trigger
> 

I would like to keep the interface consistent, so as it's displaying
a key serial number it should also accept a key serial number on writing.
(Or '0', which is 'invalid key serial number');

>>
>> I would rather modify the interface to allow a key serial number (or 0).
>> That would allow us to modify the configured key, which currently is
>> fixed for the lifetime of the connection.
> 
> Wouldn't a configured key (`tls_key`) be changed via the tlshd
> interface, not writing a key serial to sysfs?
> 

We would need to figure out the effects of REPLACETLSPSK with secure
concatenation.
Clearly for secure concatenation the resulting key is generated from
the secure key material negotiated during authentication. So the only
valid keys here should be generated keys, not persistent ones.
But in order to have _new_ generated keys we need to run authentication.
Consequently the key serial number can't be known when starting a
REPLACETLSPSK operation, and we need to write '0' into the
tls_configured_key attribute to a) reset the generated key and b)
start the authentication protocol.
That will then generate a new key, which will be visible via
'tls_configured_key'.

So all good, methinks.

>> And writing '0' would reset the configured key, reverting to automatic
>> key selection.
>> Having such an interface would actually be beneficial, as it would
>> remove some limitations from the current interface.
> 
> Anyway, I'm not sure your vision conflicts with allowing "write 1 to
> `tls_configured_key` to trigger a REPLACETLSPSK"
> 

Please don't use '1'. As mentioned earlier, I would like the sysfs
attribute to represent entries in the keystore (by displaying the key
serial number). '1' _is_ a valid number, and so the interface will try
to lookup a key with serial number '1'.
(And failing to do so :-)
'0' is _not_ valid key serial number, so writing '0' is equivalent to
'disable that key'.

Cheers,

Hannes--
Dr. Hannes Reinecke                  Kernel Storage Architect
hare@suse.de                                +49 911 74053 688
SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich