From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============3664835549443056974==" MIME-Version: 1.0 From: Denis Kenzior Subject: Re: [PATCH] eapol: do not parse RSN for WPA1 in 1 of 4 Date: Thu, 17 Oct 2019 18:48:51 -0500 Message-ID: <53ce7dee-5075-02b3-61e4-e5d472e17767@gmail.com> In-Reply-To: <20191017234316.4271-1-prestwoj@gmail.com> List-Id: To: iwd@lists.01.org --===============3664835549443056974== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi James, On 10/17/19 6:43 PM, James Prestwood wrote: > A recent change checked the return value of ie_parse_rsne_from_data > inside the ptk 1/4 handler. This seemed safe, but actually caused > the eapol unit test to fail. > = > The reason was because eapol was parsing the IEs assuming they were > an RSN, when they could be a WPA IE (WPA1 not WPA2). The WPA case > does not end up using the rsn_info at all, so having rsn_info > uninitialized did not pose a problem. After adding the return value > check it was found this fails every time for WPA1. > = > Since the rsn_info is not needed for WPA1 we can only do the RSN > parse for WPA2 and leave rsn_info uninitialized. > --- > src/eapol.c | 7 +++++-- > 1 file changed, 5 insertions(+), 2 deletions(-) > = applied, thanks. Regards, -Denis --===============3664835549443056974==--