Kernel/iwd
3
0
Fork 0
mirror of https://git.kernel.org/pub/scm/network/wireless/iwd.git synced 2024-11-05 11:39:24 +01:00
Code Releases Activity
2697af428e
iwd/src/ft.c

856 lines
20 KiB
C
Raw Normal View History

netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
/*
*
* Wireless daemon for Linux
*
treewide: Move the Intel copyright forward to 2019
2019-10-25 00:43:08 +02:00
* Copyright (C) 2017-2019 Intel Corporation. All rights reserved.
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
*
*/
#ifdef HAVE_CONFIG_H
#include <config.h>
#endif
#include <ell/ell.h>
#include "src/ie.h"
#include "src/handshake.h"
#include "src/crypto.h"
ft: rename ftutil to ft (prep for auth-proto) Now the 'ft' module, previously ftutil, will be used to drive FT via the auth-proto virtual class. This renaming is in preparation as ftutil will become obsolete since all the IE building/processing is going to be moved out of netdev. The new ft.c module will utilize the existing ftutil functionality, but since this is now a full blown auth protocol naming it 'ft' is better suited.
2019-05-07 20:53:41 +02:00
#include "src/ft.h"
ftutil: add associate parser Helper to obtain RSNE, MDE, and FTE from associate frame.
2019-03-05 18:12:30 +01:00
#include "src/mpdu.h"
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
#include "src/auth-proto.h"
struct ft_sm {
struct auth_proto ap;
struct handshake_state *hs;
ft_tx_authenticate_func_t tx_auth;
ft_tx_associate_func_t tx_assoc;
void *user_data;
};
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
/*
* Calculate the MIC field of the FTE and write it directly to that FTE,
* assuming it was all zeros before. See 12.8.4 and 12.8.5.
*/
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
static bool ft_calculate_fte_mic(struct handshake_state *hs, uint8_t seq_num,
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
const uint8_t *rsne, const uint8_t *fte,
const uint8_t *ric, uint8_t *out_mic)
{
struct iovec iov[10];
int iov_elems = 0;
struct l_checksum *checksum;
crypto/handshake/eapol: Allow other PTK lengths The crypto_ptk was hard coded for 16 byte KCK/KEK. Depending on the AKM these can be up to 32 bytes. This changes completely removes the crypto_ptk struct and adds getters to the handshake object for the kck and kek. Like before the PTK is derived into a continuous buffer, and the kck/kek getters take care of returning the proper key offset depending on AKM. To allow for larger than 16 byte keys aes_unwrap needed to be modified to take the kek length.
2019-01-17 21:25:30 +01:00
const uint8_t *kck = handshake_state_get_kck(hs);
ft: update for larger MIC/FILS-FT
2019-05-23 00:24:01 +02:00
size_t kck_len = handshake_state_get_kck_len(hs);
uint8_t zero_mic[24] = {};
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
iov[iov_elems].iov_base = hs->spa;
iov[iov_elems++].iov_len = 6;
iov[iov_elems].iov_base = hs->aa;
iov[iov_elems++].iov_len = 6;
iov[iov_elems].iov_base = &seq_num;
iov[iov_elems++].iov_len = 1;
if (rsne) {
iov[iov_elems].iov_base = (void *) rsne;
iov[iov_elems++].iov_len = rsne[1] + 2;
}
iov[iov_elems].iov_base = hs->mde;
iov[iov_elems++].iov_len = hs->mde[1] + 2;
if (fte) {
iov[iov_elems].iov_base = (void *) fte;
iov[iov_elems++].iov_len = 4;
iov[iov_elems].iov_base = zero_mic;
ft: update for larger MIC/FILS-FT
2019-05-23 00:24:01 +02:00
iov[iov_elems++].iov_len = kck_len;
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
ft: update for larger MIC/FILS-FT
2019-05-23 00:24:01 +02:00
iov[iov_elems].iov_base = (void *) (fte + 4 + kck_len);
iov[iov_elems++].iov_len = fte[1] + 2 - 4 - kck_len;
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
}
if (ric) {
iov[iov_elems].iov_base = (void *) ric;
iov[iov_elems++].iov_len = ric[1] + 2;
}
ft: update for larger MIC/FILS-FT
2019-05-23 00:24:01 +02:00
if (kck_len == 16)
checksum = l_checksum_new_cmac_aes(kck, kck_len);
else
checksum = l_checksum_new_hmac(L_CHECKSUM_SHA384, kck, kck_len);
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
if (!checksum)
return false;
l_checksum_updatev(checksum, iov, iov_elems);
ft: update for larger MIC/FILS-FT
2019-05-23 00:24:01 +02:00
l_checksum_get_digest(checksum, out_mic, kck_len);
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
l_checksum_free(checksum);
return true;
}
/*
* Validate the FC, the addresses, Auth Type and authentication sequence
* number of an FT Authentication Response frame, return status code, and
* the start of the IE array (RSN, MD, FT, TI and RIC).
* See 8.3.3.1 for the header and 8.3.3.11 for the body format.
*/
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
static bool ft_parse_authentication_resp_frame(const uint8_t *data, size_t len,
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
const uint8_t *addr1, const uint8_t *addr2,
const uint8_t *addr3, uint16_t auth_seq,
uint16_t *out_status, const uint8_t **out_ies,
size_t *out_ies_len)
{
const uint16_t frame_type = 0x00b0;
uint16_t status = 0;
if (len < 30)
return false;
/* Check FC == Management Frame -> Authentication */
if (l_get_le16(data + 0) != frame_type)
return false;
if (memcmp(data + 4, addr1, 6))
return false;
if (memcmp(data + 10, addr2, 6))
return false;
if (memcmp(data + 16, addr3, 6))
return false;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
/* Check Authentication algorithm number is FT (2) */
if (l_get_le16(data + 24) != 2)
netdev: Handle the FT Authentication Response message Parse the second message of the FT transition, validate and build the third message, the Reassociation Request.
2017-01-12 09:36:01 +01:00
return false;
if (l_get_le16(data + 26) != auth_seq)
return false;
if (auth_seq == 2 || auth_seq == 4)
status = l_get_le16(data + 28);
if (out_status)
*out_status = status;
if (status == 0 && out_ies) {
*out_ies = data + 28;
*out_ies_len = len - 28;
}
return true;
}
ftutil: add associate parser Helper to obtain RSNE, MDE, and FTE from associate frame.
2019-03-05 18:12:30 +01:00
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
static bool ft_parse_action_resp_frame(const uint8_t *frame, size_t frame_len,
const uint8_t *spa, const uint8_t *aa,
uint16_t *out_status,
const uint8_t **out_ies,
size_t *out_ies_len)
{
uint16_t status = 0;
/* Category FT */
if (frame[0] != 6)
return false;
/* FT Action */
if (frame[1] != 2)
return false;
if (memcmp(frame + 2, spa, 6))
return false;
if (memcmp(frame + 8, aa, 6))
return false;
status = l_get_le16(frame + 14);
if (out_status)
*out_status = status;
if (status == 0 && out_ies) {
*out_ies = frame + 16;
*out_ies_len = frame_len - 16;
}
return true;
}
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
static bool ft_parse_associate_resp_frame(const uint8_t *frame, size_t frame_len,
ftutil: add associate parser Helper to obtain RSNE, MDE, and FTE from associate frame.
2019-03-05 18:12:30 +01:00
uint16_t *out_status, const uint8_t **rsne,
const uint8_t **mde, const uint8_t **fte)
{
const struct mmpdu_header *mpdu;
const struct mmpdu_association_response *body;
struct ie_tlv_iter iter;
mpdu = mpdu_validate(frame, frame_len);
if (!mpdu)
return false;
body = mmpdu_body(mpdu);
ie_tlv_iter_init(&iter, body->ies, (const uint8_t *) mpdu + frame_len -
body->ies);
while (ie_tlv_iter_next(&iter)) {
switch (ie_tlv_iter_get_tag(&iter)) {
case IE_TYPE_RSN:
if (*rsne)
return false;
*rsne = ie_tlv_iter_get_data(&iter) - 2;
break;
case IE_TYPE_MOBILITY_DOMAIN:
if (*mde)
return false;
*mde = ie_tlv_iter_get_data(&iter) - 2;
break;
case IE_TYPE_FAST_BSS_TRANSITION:
if (*fte)
return false;
*fte = ie_tlv_iter_get_data(&iter) - 2;
break;
}
}
ft: Fix endianness issues The use of mmpdu_association_response members marked as __le requires the relevant endianness conversion.
2019-05-23 23:55:54 +02:00
*out_status = L_LE16_TO_CPU(body->status_code);
ftutil: add associate parser Helper to obtain RSNE, MDE, and FTE from associate frame.
2019-03-05 18:12:30 +01:00
return true;
}
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
static int ft_tx_reassociate(struct ft_sm *ft)
{
struct iovec iov[3];
int iov_elems = 0;
struct handshake_state *hs = ft->hs;
ie: allow 24 byte FTE MIC FT over FILS-SHA384 uses a 24 byte FT MIC rather than the 16 byte MIC used for all other AKMs. This change allows both the FT builder/parser to handle both lengths of MIC. The mic length is now passed directly into ie_parse_fast_bss_transition and ie_build_fast_bss_transition
2019-05-23 00:24:00 +02:00
uint32_t kck_len = handshake_state_get_kck_len(hs);
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
bool is_rsn = hs->supplicant_ie != NULL;
uint8_t *rsne = NULL;
if (is_rsn) {
struct ie_rsn_info rsn_info;
/*
* Rebuild the RSNE to include the PMKR1Name and append
* MDE + FTE.
*
* 12.8.4: "If present, the RSNE shall be set as follows:
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - Version field shall be set to 1.
* - PMKID Count field shall be set to 1.
* - PMKID field shall contain the PMKR1Name.
* - All other fields shall be as specified in 8.4.2.27
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* and 11.5.3."
*/
if (ie_parse_rsne_from_data(hs->supplicant_ie,
hs->supplicant_ie[1] + 2,
&rsn_info) < 0)
goto error;
rsn_info.num_pmkids = 1;
rsn_info.pmkids = hs->pmk_r1_name;
rsne = alloca(256);
ie_build_rsne(&rsn_info, rsne);
iov[iov_elems].iov_base = rsne;
iov[iov_elems].iov_len = rsne[1] + 2;
iov_elems += 1;
}
/* The MDE advertised by the BSS must be passed verbatim */
iov[iov_elems].iov_base = (void *) hs->mde;
iov[iov_elems].iov_len = hs->mde[1] + 2;
iov_elems += 1;
if (is_rsn) {
struct ie_ft_info ft_info;
uint8_t *fte;
/*
* 12.8.4: "If present, the FTE shall be set as follows:
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - ANonce, SNonce, R0KH-ID, and R1KH-ID shall be set to
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* the values contained in the second message of this
* sequence.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - The Element Count field of the MIC Control field shall
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* be set to the number of elements protected in this
* frame (variable).
* [...]
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - All other fields shall be set to 0."
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
*/
memset(&ft_info, 0, sizeof(ft_info));
ft_info.mic_element_count = 3;
memcpy(ft_info.r0khid, hs->r0khid, hs->r0khid_len);
ft_info.r0khid_len = hs->r0khid_len;
memcpy(ft_info.r1khid, hs->r1khid, 6);
ft_info.r1khid_present = true;
memcpy(ft_info.anonce, hs->anonce, 32);
memcpy(ft_info.snonce, hs->snonce, 32);
fte = alloca(256);
ie: allow 24 byte FTE MIC FT over FILS-SHA384 uses a 24 byte FT MIC rather than the 16 byte MIC used for all other AKMs. This change allows both the FT builder/parser to handle both lengths of MIC. The mic length is now passed directly into ie_parse_fast_bss_transition and ie_build_fast_bss_transition
2019-05-23 00:24:00 +02:00
ie_build_fast_bss_transition(&ft_info, kck_len, fte);
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
if (!ft_calculate_fte_mic(hs, 5, rsne, fte, NULL, ft_info.mic))
goto error;
/* Rebuild the FT IE now with the MIC included */
ie: allow 24 byte FTE MIC FT over FILS-SHA384 uses a 24 byte FT MIC rather than the 16 byte MIC used for all other AKMs. This change allows both the FT builder/parser to handle both lengths of MIC. The mic length is now passed directly into ie_parse_fast_bss_transition and ie_build_fast_bss_transition
2019-05-23 00:24:00 +02:00
ie_build_fast_bss_transition(&ft_info, kck_len, fte);
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
iov[iov_elems].iov_base = fte;
iov[iov_elems].iov_len = fte[1] + 2;
iov_elems += 1;
}
ft->tx_assoc(iov, iov_elems, ft->user_data);
return 0;
error:
return -EINVAL;
}
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
static int ft_parse_ies(const uint8_t *ies, size_t ies_len,
const uint8_t **rsne_out, const uint8_t **mde_out,
const uint8_t **fte_out)
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
{
struct ie_tlv_iter iter;
const uint8_t *rsne = NULL;
const uint8_t *mde = NULL;
const uint8_t *fte = NULL;
ie_tlv_iter_init(&iter, ies, ies_len);
while (ie_tlv_iter_next(&iter)) {
switch (ie_tlv_iter_get_tag(&iter)) {
case IE_TYPE_RSN:
if (rsne)
goto ft_error;
rsne = ie_tlv_iter_get_data(&iter) - 2;
break;
case IE_TYPE_MOBILITY_DOMAIN:
if (mde)
goto ft_error;
mde = ie_tlv_iter_get_data(&iter) - 2;
break;
case IE_TYPE_FAST_BSS_TRANSITION:
if (fte)
goto ft_error;
fte = ie_tlv_iter_get_data(&iter) - 2;
break;
}
}
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (rsne_out)
*rsne_out = rsne;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (mde_out)
*mde_out = mde;
if (fte_out)
*fte_out = fte;
return 0;
ft_error:
return -EINVAL;
}
static bool ft_verify_rsne(const uint8_t *rsne, const uint8_t *pmk_r0_name,
const uint8_t *authenticator_ie)
{
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
/*
* In an RSN, check for an RSNE containing the PMK-R0-Name and
* the remaining fields same as in the advertised RSNE.
*
* 12.8.3: "The RSNE shall be present only if dot11RSNAActivated
* is true. If present, the RSNE shall be set as follows:
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - Version field shall be set to 1.
* - PMKID Count field shall be set to 1.
* - PMKID List field shall be set to the value contained in the
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* first message of this sequence.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - All other fields shall be identical to the contents of the
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* RSNE advertised by the AP in Beacon and Probe Response frames."
*/
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
struct ie_rsn_info msg2_rsne;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (!rsne)
return false;
if (ie_parse_rsne_from_data(rsne, rsne[1] + 2,
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
&msg2_rsne) < 0)
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
return false;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (msg2_rsne.num_pmkids != 1 ||
memcmp(msg2_rsne.pmkids, pmk_r0_name, 16))
return false;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (!handshake_util_ap_ie_matches(rsne, authenticator_ie, false))
return false;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
return true;
}
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
static bool ft_parse_fte(struct handshake_state *hs,
const uint8_t *snonce,
const uint8_t *fte,
struct ie_ft_info *ft_info)
{
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
/*
* In an RSN, check for an FT IE with the same R0KH-ID and the same
* SNonce that we sent, and check that the R1KH-ID and the ANonce
* are present. Use them to generate new PMK-R1, PMK-R1-Name and PTK
* in handshake.c.
*
* 12.8.3: "The FTE shall be present only if dot11RSNAActivated is
* true. If present, the FTE shall be set as follows:
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - R0KH-ID shall be identical to the R0KH-ID provided by the FTO
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* in the first message.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - R1KH-ID shall be set to the R1KH-ID of the target AP, from
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* dot11FTR1KeyHolderID.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - ANonce shall be set to a value chosen randomly by the target AP,
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* following the recommendations of 11.6.5.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - SNonce shall be set to the value contained in the first message
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* of this sequence.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - All other fields shall be set to 0."
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
*/
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
uint8_t zeros[24] = {};
uint32_t kck_len = handshake_state_get_kck_len(hs);
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (!fte)
return false;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (ie_parse_fast_bss_transition_from_data(fte, fte[1] + 2,
kck_len, ft_info) < 0)
return false;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (ft_info->mic_element_count != 0 ||
memcmp(ft_info->mic, zeros, kck_len))
return false;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (hs->r0khid_len != ft_info->r0khid_len ||
memcmp(hs->r0khid, ft_info->r0khid,
hs->r0khid_len) ||
!ft_info->r1khid_present)
return false;
if (memcmp(ft_info->snonce, snonce, 32))
return false;
return true;
}
static int ft_process_ies(struct handshake_state *hs, const uint8_t *ies,
size_t ies_len)
{
const uint8_t *rsne = NULL;
const uint8_t *mde = NULL;
const uint8_t *fte = NULL;
bool is_rsn;
/* Check 802.11r IEs */
if (!ies)
goto ft_error;
if (ft_parse_ies(ies, ies_len, &rsne, &mde, &fte) < 0)
goto ft_error;
is_rsn = hs->supplicant_ie != NULL;
if (is_rsn) {
if (!ft_verify_rsne(rsne, hs->pmk_r0_name,
hs->authenticator_ie))
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
goto ft_error;
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
} else if (rsne)
goto ft_error;
/*
* Check for an MD IE identical to the one we sent in message 1
*
* 12.8.3: "The MDE shall contain the MDID and FT Capability and
* Policy fields. This element shall be the same as the MDE
* advertised by the target AP in Beacon and Probe Response frames."
*/
if (!mde || memcmp(hs->mde, mde, hs->mde[1] + 2))
goto ft_error;
if (is_rsn) {
struct ie_ft_info ft_info;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: factor out various parsing routines Break up the rather large code block which parses out IEs, verifies, and sets into the handshake. FT-over-DS needs these steps broken up in order to parse the action frame response without modifying the handshake.
2021-04-27 21:49:40 +02:00
if (!ft_parse_fte(hs, hs->snonce, fte, &ft_info))
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
goto ft_error;
handshake_state_set_fte(hs, fte);
handshake_state_set_anonce(hs, ft_info.anonce);
handshake_state_set_kh_ids(hs, ft_info.r0khid,
ft_info.r0khid_len,
ft_info.r1khid);
handshake_state_derive_ptk(hs);
} else if (fte)
goto ft_error;
ft: separate ft_sm from ft_process_ies To prepare for some refactoring remove the ft_sm dependency from ft_process_ies and instead only make it depend on the handshake_state object.
2021-04-16 00:45:01 +02:00
return 0;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
ft_error:
return -EBADMSG;
}
static int ft_rx_action(struct auth_proto *ap, const uint8_t *frame,
size_t frame_len)
{
struct ft_sm *ft = l_container_of(ap, struct ft_sm, ap);
uint16_t status_code = MMPDU_STATUS_CODE_UNSPECIFIED;
const uint8_t *ies = NULL;
size_t ies_len;
ft: separate ft_sm from ft_process_ies To prepare for some refactoring remove the ft_sm dependency from ft_process_ies and instead only make it depend on the handshake_state object.
2021-04-16 00:45:01 +02:00
int ret;
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
if (!ft_parse_action_resp_frame(frame, frame_len, ft->hs->spa,
ft->hs->aa, &status_code,
&ies, &ies_len))
return -EBADMSG;
/* AP Rejected the authenticate / associate */
if (status_code != 0)
goto auth_error;
ft: separate ft_sm from ft_process_ies To prepare for some refactoring remove the ft_sm dependency from ft_process_ies and instead only make it depend on the handshake_state object.
2021-04-16 00:45:01 +02:00
ret = ft_process_ies(ft->hs, ies, ies_len);
if (ret < 0)
goto auth_error;
return ft_tx_reassociate(ft);
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
auth_error:
return (int)status_code;
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
}
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
static int ft_rx_authenticate(struct auth_proto *ap, const uint8_t *frame,
size_t frame_len)
{
struct ft_sm *ft = l_container_of(ap, struct ft_sm, ap);
uint16_t status_code = MMPDU_STATUS_CODE_UNSPECIFIED;
const uint8_t *ies = NULL;
size_t ies_len;
ft: separate ft_sm from ft_process_ies To prepare for some refactoring remove the ft_sm dependency from ft_process_ies and instead only make it depend on the handshake_state object.
2021-04-16 00:45:01 +02:00
int ret;
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
/*
* Parse the Authentication Response and validate the contents
* according to 12.5.2 / 12.5.4: RSN or non-RSN Over-the-air
* FT Protocol.
*/
if (!ft_parse_authentication_resp_frame(frame, frame_len, ft->hs->spa,
ft->hs->aa, ft->hs->aa, 2, &status_code,
&ies, &ies_len))
goto auth_error;
/* AP Rejected the authenticate / associate */
if (status_code != 0)
goto auth_error;
ft: separate ft_sm from ft_process_ies To prepare for some refactoring remove the ft_sm dependency from ft_process_ies and instead only make it depend on the handshake_state object.
2021-04-16 00:45:01 +02:00
ret = ft_process_ies(ft->hs, ies, ies_len);
if (ret < 0)
goto auth_error;
return ft_tx_reassociate(ft);
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
auth_error:
return (int)status_code;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
}
static int ft_rx_associate(struct auth_proto *ap, const uint8_t *frame,
size_t frame_len)
{
struct ft_sm *ft = l_container_of(ap, struct ft_sm, ap);
struct handshake_state *hs = ft->hs;
ie: allow 24 byte FTE MIC FT over FILS-SHA384 uses a 24 byte FT MIC rather than the 16 byte MIC used for all other AKMs. This change allows both the FT builder/parser to handle both lengths of MIC. The mic length is now passed directly into ie_parse_fast_bss_transition and ie_build_fast_bss_transition
2019-05-23 00:24:00 +02:00
uint32_t kck_len = handshake_state_get_kck_len(hs);
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
const uint8_t *rsne = NULL;
const uint8_t *mde = NULL;
const uint8_t *fte = NULL;
const uint8_t *sent_mde = hs->mde;
bool is_rsn = hs->supplicant_ie != NULL;
uint16_t out_status;
if (!ft_parse_associate_resp_frame(frame, frame_len, &out_status, &rsne,
&mde, &fte))
return -EBADMSG;
/*
* During a transition in an RSN, check for an RSNE containing the
* PMK-R1-Name and the remaining fields same as in the advertised
* RSNE.
*
* 12.8.5: "The RSNE shall be present only if dot11RSNAActivated is
* true. If present, the RSNE shall be set as follows:
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - Version field shall be set to 1.
* - PMKID Count field shall be set to 1.
* - PMKID field shall contain the PMKR1Name
* - All other fields shall be identical to the contents of the RSNE
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
* advertised by the target AP in Beacon and Probe Response frames."
*/
if (is_rsn) {
struct ie_rsn_info msg4_rsne;
if (!rsne)
return -EBADMSG;
if (ie_parse_rsne_from_data(rsne, rsne[1] + 2,
&msg4_rsne) < 0)
return -EBADMSG;
if (msg4_rsne.num_pmkids != 1 ||
memcmp(msg4_rsne.pmkids, hs->pmk_r1_name, 16))
return -EBADMSG;
if (!handshake_util_ap_ie_matches(rsne, hs->authenticator_ie,
false))
return -EBADMSG;
} else {
if (rsne)
return -EBADMSG;
}
/* An MD IE identical to the one we sent must be present */
if (sent_mde && (!mde || memcmp(sent_mde, mde, sent_mde[1] + 2)))
return -EBADMSG;
/*
* An FT IE is required in an initial mobility domain
* association and re-associations in an RSN but not present
* in a non-RSN (12.4.2 vs. 12.4.3).
*/
if (sent_mde && is_rsn && !fte)
return -EBADMSG;
if (!(sent_mde && is_rsn) && fte)
return -EBADMSG;
if (fte) {
struct ie_ft_info ft_info;
ft: update for larger MIC/FILS-FT
2019-05-23 00:24:01 +02:00
uint8_t mic[24];
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
if (ie_parse_fast_bss_transition_from_data(fte, fte[1] + 2,
ie: allow 24 byte FTE MIC FT over FILS-SHA384 uses a 24 byte FT MIC rather than the 16 byte MIC used for all other AKMs. This change allows both the FT builder/parser to handle both lengths of MIC. The mic length is now passed directly into ie_parse_fast_bss_transition and ie_build_fast_bss_transition
2019-05-23 00:24:00 +02:00
kck_len, &ft_info) < 0)
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
return -EBADMSG;
/*
* In an RSN, check for an FT IE with the same
* R0KH-ID, R1KH-ID, ANonce and SNonce that we
* received in message 2, MIC Element Count
* of 6 and the correct MIC.
*/
if (!ft_calculate_fte_mic(hs, 6, rsne, fte, NULL, mic))
return -EBADMSG;
if (ft_info.mic_element_count != 3 ||
ft: update for larger MIC/FILS-FT
2019-05-23 00:24:01 +02:00
memcmp(ft_info.mic, mic, kck_len))
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
return -EBADMSG;
if (hs->r0khid_len != ft_info.r0khid_len ||
memcmp(hs->r0khid, ft_info.r0khid,
hs->r0khid_len) ||
!ft_info.r1khid_present ||
memcmp(hs->r1khid, ft_info.r1khid, 6))
return -EBADMSG;
if (memcmp(ft_info.anonce, hs->anonce, 32))
return -EBADMSG;
if (memcmp(ft_info.snonce, hs->snonce, 32))
return -EBADMSG;
if (ft_info.gtk_len) {
uint8_t gtk[32];
if (!handshake_decode_fte_key(hs, ft_info.gtk,
ft_info.gtk_len,
gtk))
return -EBADMSG;
if (ft_info.gtk_rsc[6] != 0x00 ||
ft_info.gtk_rsc[7] != 0x00)
return -EBADMSG;
handshake_state_install_gtk(hs, ft_info.gtk_key_id,
gtk, ft_info.gtk_len,
ft_info.gtk_rsc, 6);
}
if (ft_info.igtk_len) {
uint8_t igtk[16];
if (!handshake_decode_fte_key(hs, ft_info.igtk,
ft_info.igtk_len, igtk))
return -EBADMSG;
handshake_state_install_igtk(hs, ft_info.igtk_key_id,
igtk, ft_info.igtk_len,
ft_info.igtk_ipn);
}
handshake_state_install_ptk(ft->hs);
}
return 0;
}
static void ft_sm_free(struct auth_proto *ap)
{
struct ft_sm *ft = l_container_of(ap, struct ft_sm, ap);
l_free(ft);
}
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
bool ft_build_authenticate_ies(struct handshake_state *hs,
const uint8_t *new_snonce, uint8_t *buf,
size_t *len)
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
{
ie: allow 24 byte FTE MIC FT over FILS-SHA384 uses a 24 byte FT MIC rather than the 16 byte MIC used for all other AKMs. This change allows both the FT builder/parser to handle both lengths of MIC. The mic length is now passed directly into ie_parse_fast_bss_transition and ie_build_fast_bss_transition
2019-05-23 00:24:00 +02:00
uint32_t kck_len = handshake_state_get_kck_len(hs);
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
bool is_rsn = hs->supplicant_ie != NULL;
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
uint8_t *ptr = buf;
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
if (is_rsn) {
struct ie_rsn_info rsn_info;
/*
* Rebuild the RSNE to include the PMKR0Name and append
* MDE + FTE.
*
* 12.8.2: "If present, the RSNE shall be set as follows:
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - Version field shall be set to 1.
* - PMKID Count field shall be set to 1.
* - PMKID List field shall contain the PMKR0Name.
* - All other fields shall be as specified in 8.4.2.27
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
* and 11.5.3."
*/
if (ie_parse_rsne_from_data(hs->supplicant_ie,
hs->supplicant_ie[1] + 2,
&rsn_info) < 0)
return false;
rsn_info.num_pmkids = 1;
rsn_info.pmkids = hs->pmk_r0_name;
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
ie_build_rsne(&rsn_info, ptr);
ptr += ptr[1] + 2;
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
}
/* The MDE advertised by the BSS must be passed verbatim */
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
ptr[0] = IE_TYPE_MOBILITY_DOMAIN;
ptr[1] = 3;
memcpy(ptr + 2, hs->mde + 2, 3);
ptr += 5;
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
if (is_rsn) {
struct ie_ft_info ft_info;
/*
* 12.8.2: "If present, the FTE shall be set as follows:
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - R0KH-ID shall be the value of R0KH-ID obtained by the
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
* FTO during its FT initial mobility domain association
* exchange.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - SNonce shall be set to a value chosen randomly by the
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
* FTO, following the recommendations of 11.6.5.
treewide: Get rid of non-ASCII characters
2021-04-28 21:16:06 +02:00
* - All other fields shall be set to 0."
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
*/
memset(&ft_info, 0, sizeof(ft_info));
memcpy(ft_info.r0khid, hs->r0khid, hs->r0khid_len);
ft_info.r0khid_len = hs->r0khid_len;
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
memcpy(ft_info.snonce, new_snonce, 32);
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
ie_build_fast_bss_transition(&ft_info, kck_len, ptr);
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
ptr += ptr[1] + 2;
ft: netdev: move Authenticate IE building into ft.c Since FT already handles processing the FT IE's (and building for associate) it didn't make sense to have all the IE building inside netdev_build_cmd_ft_authenticate. Instead this logic was moved into ft.c, and an iovec is now passed from FT into netdev_ft_tx_authenticate. This leaves the netdev command builder unburdened by the details of FT, as well as prepares for FT-over-DS.
2019-05-09 19:14:12 +02:00
}
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft: expose ft_build_authenticate_ies The building of the FT IEs for Action/Authenticate frames will need to be shared between ft and netdev once FT-over-DS is refactored. The building was refactored to work off the callers buffer rather than internal stack buffers. An argument 'new_snonce' was included as FT-over-DS will generate a new snonce for the initial action frame, hence the handshakes snonce cannot be used.
2021-04-27 21:49:41 +02:00
if (len)
*len = ptr - buf;
return true;
}
static bool ft_start(struct auth_proto *ap)
{
struct ft_sm *ft = l_container_of(ap, struct ft_sm, ap);
struct handshake_state *hs = ft->hs;
struct iovec iov;
uint8_t buf[512];
size_t len;
if (!ft_build_authenticate_ies(hs, hs->snonce, buf, &len))
return false;
iov.iov_base = buf;
iov.iov_len = len;
ft->tx_auth(&iov, 1, ft->user_data);
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
return true;
}
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
static struct auth_proto *ft_sm_new(struct handshake_state *hs,
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft_tx_authenticate_func_t tx_auth,
ft_tx_associate_func_t tx_assoc,
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
bool over_air,
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
void *user_data)
{
struct ft_sm *ft = l_new(struct ft_sm, 1);
ft->tx_auth = tx_auth;
ft->tx_assoc = tx_assoc;
ft->hs = hs;
ft->user_data = user_data;
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
ft->ap.rx_authenticate = (over_air) ? ft_rx_authenticate : ft_rx_action;
netdev: ft: refactor FT into an auth-proto Since FT operates over Authenticate/Associate, it makes the most sense for it to behave like the other auth-protos. This change moves all the FT specific processing out of netdev and into ft.c. The bulk of the changes were strait copy-pastes from netdev into ft.c with minor API changes (e.g. remove struct netdev). The 'in_ft' boolean unforunately is still required for a few reasons: - netdev_disconnect_event relies on this flag so it can ignore the disconnect which comes in when doing a fast transition. We cannot simply check netdev->ap because this would cause the other auth-protos to not handle a disconnect correctly. - netdev_associate_event needs to correctly setup the eapol_sm when in FT mode by setting require_handshake and use_eapol_start to false. This cannot be handled inside eapol by checking the AKM because an AP may only advertise a FT AKM, and the initial mobility association does require the 4-way handshake.
2019-05-07 20:53:42 +02:00
ft->ap.rx_associate = ft_rx_associate;
ft->ap.start = ft_start;
ft->ap.free = ft_sm_free;
return &ft->ap;
}
ft: netdev: station: support FT-over-DS FT-over-DS is a way to do a Fast BSS Transition using action frames for the authenticate step. This allows a station to start a fast transition to a target AP while still being connected to the original AP. This, in theory, can result in less carrier downtime. The existing ft_sm_new was removed, and two new constructors were added; one for over-air, and another for over-ds. The internals of ft.c mostly remain the same. A flag to distinguish between air/ds was added along with a new parser to parse the action frames rather than authenticate frames. The IE parsing is identical. Netdev now just initializes the auth-proto differently depending on if its doing over-air or over-ds. A new TX authenticate function was added and used for over-ds. This will send out the IEs from ft.c with an FT Request action frame. The FT Response action frame is then recieved from the AP and fed into the auth-proto state machine. After this point ft-over-ds behaves the same as ft-over-air (associate to the target AP). Some simple code was added in station.c to determine if over-air or over-ds should be used. FT-over-DS can be beneficial in cases where the AP is directing us to roam, or if the RSSI falls below a threshold. It should not be used if we have lost communication to the AP all (beacon lost) as it only works while we can still talk to the original AP.
2019-05-09 20:16:28 +02:00
struct auth_proto *ft_over_air_sm_new(struct handshake_state *hs,
ft_tx_authenticate_func_t tx_auth,
ft_tx_associate_func_t tx_assoc,
void *user_data)
{
return ft_sm_new(hs, tx_auth, tx_assoc, true, user_data);
}
struct auth_proto *ft_over_ds_sm_new(struct handshake_state *hs,
ft_tx_authenticate_func_t tx_auth,
ft_tx_associate_func_t tx_assoc,
void *user_data)
{
return ft_sm_new(hs, tx_auth, tx_assoc, false, user_data);
}
Copy Permalink