[PATCH v2 02/15] mesh: factor out rsn initialization

[Peter Oh]

On 04/13/2018 02:39 AM, Jouni Malinen wrote:
> On Thu, Apr 12, 2018 at 07:32:26PM -0700, Peter Oh wrote:
>> I prefer you rebase your change after the series applied if you don't mind.
> I do mind.. There is no point in introducing new segmentation faults for
> a known issue and it is clear that this mesh patchset needs additional
> work regardless taken into account the fixes that Masashi sent.
>
>> Hence you may send your change separately and if your patch is merged before
>> the series, then I'll rebase my patchset.
> I applied this patch from Daniel now to get sae_password working. Please
> rebase the mesh DFS set on top of the updated hostap.git and pull in the
> fixes from the two patches that Masashi sent to the appropriate commits
> in the set. Each new patch should apply cleanly and not introduce
> compiler warnings.
I'll do rebase.
>>> From: Daniel Golle <daniel at makrotopia.org>
>>> Date: Fri, 13 Apr 2018 00:42:10 +0200
>>> Subject: [PATCH] mesh: properly handle sae_password
>>>
>>> The recently introduced sae_password parameter is only handled properly
>>> in wpa_supplicant/sme.c while wpa_supplicant/mesh.c assumed that
>>> ssid->passphrase exclusively holds the secret.
>>> Import the logic from sme.c to mesh.c to allow having only sae_password
>>> set which otherwise throws this error:
>>> AP-ENABLED
>>> mesh: Passphrase for SAE not configured
>>> Init RSN failed. Deinit mesh...
>>> wlan1: interface state ENABLED->DISABLED
>>> AP-DISABLED
>>> Segmentation fault
> This segmentation fault part was introduced in these new mesh DFS
> patches. Without them (i.e., the hostap.git master branch before today),
> I saw no segmentation fault here -- the setup just failed cleanly with
> that "passphrase for SAE not configured" message.
What is the compiling option generating segmentation fault? I want to try.

Thanks,
Peter