All of lore.kernel.org
 help / color / mirror / Atom feed
From: James Prestwood <prestwoj@gmail.com>
To: Wen Gong <quic_wgong@quicinc.com>, iwd@lists.01.org, iwd@lists.linux.dev
Cc: Denis Kenzior <denkenz@gmail.com>
Subject: Re: [PATCH v3 3/3] owe: netdev: refactor to remove OWE as an auth-proto
Date: Wed, 12 Jul 2023 19:33:59 -0700	[thread overview]
Message-ID: <d2a490cf-ae06-ed26-33eb-5475616bd725@gmail.com> (raw)
In-Reply-To: <00246aa8-fe45-f5c2-cf2a-3450cee414f2@quicinc.com>

Hi Wen,

On 7/12/23 7:23 PM, Wen Gong wrote:
> On 7/12/2023 10:52 PM, James Prestwood wrote:
>> Hi Wen,
>>
>> On 7/12/23 4:17 AM, Wen Gong wrote:
>>> On 9/4/2021 3:35 AM, James Prestwood wrote:
>>>> ---
>>> ...
>>>> +
>>>> +            case IE_TYPE_RSN:
>>>> +                if (!netdev->owe_sm)
>>>> +                    continue;
>>>> +
>>>> +                if (ie_parse_rsne(&iter, &info) < 0) {
>>>> +                    l_error("could not parse RSN IE");
>>>> +                    goto error;
>>>> +                }
>>>> +
>>>> +                /*
>>>> +                 * RFC 8110 Section 4.2
>>>> +                 * An AP agreeing to do OWE MUST include the OWE 
>>>> AKM in
>>>> +                 * the RSN element portion of the 802.11 association
>>>> +                 * response.
>>>> +                 */
>>> Now it is happen connect to OWE AP fail with some APs, because the 
>>> assoc resp do not inclued RSNE.
>>>> +                if (info.akm_suites != IE_RSN_AKM_SUITE_OWE) {
>>>> +                    l_error("OWE AKM not included");
>>>> +                    goto deauth;
>>>> +                }
>>>> +
>>>> +                owe_akm_found = true;
>>>> +
>>>> +                break;
>>>> +            }
>>>> +        }
>>>> +
>>>> +        if (netdev->owe_sm) {
>>>> +            if (!owe_dh || !owe_akm_found) {
>>>> +                l_error("OWE DH element/RSN not found");
>>>
>>> It failed here.
>>>
>>> So is the check for owe_akm_found MUST added here if owe_dh is 
>>> existed in assoc resp?
>>
>> As the comment states, the OWE RFC says that the RSN element must be 
>> included in the associate response if we are to stay compliant with 
>> the spec.
>>
>> But as you may have noticed we don't actually do anything with this 
>> element besides parse it. I don't see that wpa_supplicant enforces 
>> this either, so it may be fine to relax this check.
> Thanks for you to relax the check.
>>
>> I would prefer to see iwmon logs when you connect to this AP, just to 
>> confirm that the AP isn't including the IE and not something else. 
>> Would you be able to get those?
>>
> I have collected iwd log with hexdump, it does not include RSNIE in 
> assoc resp ies:

Sorry, I should have been more specific. Using IWD_GENL_DEBUG does 
contain the information, but its not really a readable format. We have a 
tool called iwmon which you run in parallel to IWD. This parses the raw 
data and displays a human readable output:

$ sudo iwmon --nortnl --nowiphy --noscan

Denis,

Is relaxing the RSNE check in the associate response (for OWE only) 
something you'd be ok with? Obviously with an L_WARN_ON/l_warn message.

Thanks,
James

  parent reply	other threads:[~2023-07-13  2:34 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-03 19:35 [PATCH v3 1/3] netdev: factor out scan_bss from CMD_CONNECT builder James Prestwood
2021-09-03 19:35 ` [PATCH v3 2/3] wsc: set ssid in handshake James Prestwood
2021-09-03 19:35 ` [PATCH v3 3/3] owe: netdev: refactor to remove OWE as an auth-proto James Prestwood
2021-09-03 19:44   ` Denis Kenzior
2023-07-12 11:17   ` Wen Gong
2023-07-12 14:52     ` James Prestwood
     [not found]       ` <00246aa8-fe45-f5c2-cf2a-3450cee414f2@quicinc.com>
2023-07-13  2:33         ` James Prestwood [this message]
2023-07-13  2:52           ` Wen Gong
2023-07-13  3:04             ` James Prestwood
2023-07-13  3:07               ` Wen Gong
2023-07-13  3:56               ` Wen Gong
2023-07-13 14:14           ` Denis Kenzior

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=d2a490cf-ae06-ed26-33eb-5475616bd725@gmail.com \
    --to=prestwoj@gmail.com \
    --cc=denkenz@gmail.com \
    --cc=iwd@lists.01.org \
    --cc=iwd@lists.linux.dev \
    --cc=quic_wgong@quicinc.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.