| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <110ac63c-5d93-4f11-a064-b8b4a7994fec@bootlin.com>
Date: Thu, 4 Jan 2024 16:41:41 +0100
From: Alexis Lothoré <alexis.lothore@...tlin.com>
To: Christophe JAILLET <christophe.jaillet@...adoo.fr>
Cc: ajay.kathat@...rochip.com, claudiu.beznea@...on.dev, kvalo@...nel.org,
linux-kernel@...r.kernel.org, linux-wireless@...r.kernel.org,
thomas.petazzoni@...tlin.com
Subject: Re: [PATCH 2/2] wifi: wilc1000: fix RCU usage in connect path
On 1/4/24 16:30, Christophe JAILLET wrote:
> Le 04/01/2024 à 15:39, Alexis Lothoré a écrit :
>> From: Alexis Lothoré <alexis.lothore-LDxbnhwyfcJBDgjK7y7TUQ@...lic.gmane.org>
>>
>> With lockdep enabled, calls to the connect function from cfg802.11 layer
>> lead to the following warning:
>>
>> =============================
>> WARNING: suspicious RCU usage
>> 6.7.0-rc1-wt+ #333 Not tainted
>> -----------------------------
>> drivers/net/wireless/microchip/wilc1000/hif.c:386
>> suspicious rcu_dereference_check() usage!
>> [...]
>> stack backtrace:
>> CPU: 0 PID: 100 Comm: wpa_supplicant Not tainted 6.7.0-rc1-wt+ #333
>> Hardware name: Atmel SAMA5
>> unwind_backtrace from show_stack+0x18/0x1c
>> show_stack from dump_stack_lvl+0x34/0x48
>> dump_stack_lvl from wilc_parse_join_bss_param+0x7dc/0x7f4
>> wilc_parse_join_bss_param from connect+0x2c4/0x648
>> connect from cfg80211_connect+0x30c/0xb74
>> cfg80211_connect from nl80211_connect+0x860/0xa94
>> nl80211_connect from genl_rcv_msg+0x3fc/0x59c
>> genl_rcv_msg from netlink_rcv_skb+0xd0/0x1f8
>> netlink_rcv_skb from genl_rcv+0x2c/0x3c
>> genl_rcv from netlink_unicast+0x3b0/0x550
>> netlink_unicast from netlink_sendmsg+0x368/0x688
>> netlink_sendmsg from ____sys_sendmsg+0x190/0x430
>> ____sys_sendmsg from ___sys_sendmsg+0x110/0x158
>> ___sys_sendmsg from sys_sendmsg+0xe8/0x150
>> sys_sendmsg from ret_fast_syscall+0x0/0x1c
>>
>> This warning is emitted because in the connect path, when trying to parse
>> target BSS parameters, we dereference a RCU pointer whithout being in RCU
>> critical section.
>> Fix RCU dereference usage by moving it to a RCU read critical section. To
>> avoid wrapping the whole wilc_parse_join_bss_param under the critical
>> section, just use the critical section to copy ies data
>>
>> Fixes: c460495ee072 ("staging: wilc1000: fix incorrent type in initializer")
>> Signed-off-by: Alexis Lothoré
>> <alexis.lothore-LDxbnhwyfcJBDgjK7y7TUQ@...lic.gmane.org>
>> ---
>> drivers/net/wireless/microchip/wilc1000/hif.c | 35 ++++++++++++-------
>> 1 file changed, 23 insertions(+), 12 deletions(-)
>>
>> diff --git a/drivers/net/wireless/microchip/wilc1000/hif.c
>> b/drivers/net/wireless/microchip/wilc1000/hif.c
>> index 6f1218a51061..806b7337b5ae 100644
>> --- a/drivers/net/wireless/microchip/wilc1000/hif.c
>> +++ b/drivers/net/wireless/microchip/wilc1000/hif.c
>> @@ -377,38 +377,48 @@ struct wilc_join_bss_param *
>> wilc_parse_join_bss_param(struct cfg80211_bss *bss,
>> struct cfg80211_crypto_settings *crypto)
>> {
>> - const struct cfg80211_bss_ies *ies = rcu_dereference(bss->ies);
>> - const u8 *tim_elm, *ssid_elm, *rates_ie, *supp_rates_ie;
>> + const u8 *ies_data, *tim_elm, *ssid_elm, *rates_ie, *supp_rates_ie;
>> const u8 *ht_ie, *wpa_ie, *wmm_ie, *rsn_ie;
>> struct ieee80211_p2p_noa_attr noa_attr;
>> + const struct cfg80211_bss_ies *ies;
>> struct wilc_join_bss_param *param;
>> - u8 rates_len = 0;
>> + u8 rates_len = 0, ies_len;
>> int ret;
>> param = kzalloc(sizeof(*param), GFP_KERNEL);
>> if (!param)
>> return NULL;
>> + rcu_read_lock();
>> + ies = rcu_dereference(bss->ies);
>> + ies_data = kmemdup(ies->data, ies->len, GFP_ATOMIC);
>
> Is it mandatory to have GFP_ATOMIC here?
> If yes, then some other drivers may need to be fixed (at least [1]).
>From my understanding: yes, because if any sleep occurs in a section protected
by rcu_read_lock/rcu_read_unlock, it may mess with RCU grace period wait (and
so, rcu reclaim step may occur while this reader is still in the read section)
>
>> + if (!ies_data) {
>> + rcu_read_unlock();
>
> 'param' allocated just a few lines above is leaking now.
You are right, good catch, thanks. I will update accordingly.
>
> CJ
>
>
> [1]:
> https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/tree/drivers/pci/p2pdma.c#n660
>
>
>> + return NULL;
>> + }
>> + ies_len = ies->len;
>> + rcu_read_unlock();
>> +
>> param->beacon_period = cpu_to_le16(bss->beacon_interval);
>> param->cap_info = cpu_to_le16(bss->capability);
>> param->bss_type = WILC_FW_BSS_TYPE_INFRA;
>> param->ch = ieee80211_frequency_to_channel(bss->channel->center_freq);
>> ether_addr_copy(param->bssid, bss->bssid);
>> - ssid_elm = cfg80211_find_ie(WLAN_EID_SSID, ies->data, ies->len);
>> + ssid_elm = cfg80211_find_ie(WLAN_EID_SSID, ies_data, ies_len);
>> if (ssid_elm) {
>> if (ssid_elm[1] <= IEEE80211_MAX_SSID_LEN)
>> memcpy(param->ssid, ssid_elm + 2, ssid_elm[1]);
>> }
>> - tim_elm = cfg80211_find_ie(WLAN_EID_TIM, ies->data, ies->len);
>> + tim_elm = cfg80211_find_ie(WLAN_EID_TIM, ies_data, ies_len);
>> if (tim_elm && tim_elm[1] >= 2)
>> param->dtim_period = tim_elm[3];
>> memset(param->p_suites, 0xFF, 3);
>> memset(param->akm_suites, 0xFF, 3);
>> - rates_ie = cfg80211_find_ie(WLAN_EID_SUPP_RATES, ies->data, ies->len);
>> + rates_ie = cfg80211_find_ie(WLAN_EID_SUPP_RATES, ies_data, ies_len);
>> if (rates_ie) {
>> rates_len = rates_ie[1];
>> if (rates_len > WILC_MAX_RATES_SUPPORTED)
>> @@ -419,7 +429,7 @@ wilc_parse_join_bss_param(struct cfg80211_bss *bss,
>> if (rates_len < WILC_MAX_RATES_SUPPORTED) {
>> supp_rates_ie = cfg80211_find_ie(WLAN_EID_EXT_SUPP_RATES,
>> - ies->data, ies->len);
>> + ies_data, ies_len);
>> if (supp_rates_ie) {
>> u8 ext_rates = supp_rates_ie[1];
>> @@ -434,11 +444,11 @@ wilc_parse_join_bss_param(struct cfg80211_bss *bss,
>> }
>> }
>> - ht_ie = cfg80211_find_ie(WLAN_EID_HT_CAPABILITY, ies->data, ies->len);
>> + ht_ie = cfg80211_find_ie(WLAN_EID_HT_CAPABILITY, ies_data, ies_len);
>> if (ht_ie)
>> param->ht_capable = true;
>> - ret = cfg80211_get_p2p_attr(ies->data, ies->len,
>> + ret = cfg80211_get_p2p_attr(ies_data, ies_len,
>> IEEE80211_P2P_ATTR_ABSENCE_NOTICE,
>> (u8 *)&noa_attr, sizeof(noa_attr));
>> if (ret > 0) {
>> @@ -462,7 +472,7 @@ wilc_parse_join_bss_param(struct cfg80211_bss *bss,
>> }
>> wmm_ie = cfg80211_find_vendor_ie(WLAN_OUI_MICROSOFT,
>> WLAN_OUI_TYPE_MICROSOFT_WMM,
>> - ies->data, ies->len);
>> + ies_data, ies_len);
>> if (wmm_ie) {
>> struct ieee80211_wmm_param_ie *ie;
>> @@ -477,13 +487,13 @@ wilc_parse_join_bss_param(struct cfg80211_bss *bss,
>> wpa_ie = cfg80211_find_vendor_ie(WLAN_OUI_MICROSOFT,
>> WLAN_OUI_TYPE_MICROSOFT_WPA,
>> - ies->data, ies->len);
>> + ies_data, ies_len);
>> if (wpa_ie) {
>> param->mode_802_11i = 1;
>> param->rsn_found = true;
>> }
>> - rsn_ie = cfg80211_find_ie(WLAN_EID_RSN, ies->data, ies->len);
>> + rsn_ie = cfg80211_find_ie(WLAN_EID_RSN, ies_data, ies_len);
>> if (rsn_ie) {
>> int rsn_ie_len = sizeof(struct element) + rsn_ie[1];
>> int offset = 8;
>> @@ -517,6 +527,7 @@ wilc_parse_join_bss_param(struct cfg80211_bss *bss,
>> param->akm_suites[i] = crypto->akm_suites[i] & 0xFF;
>> }
>> + kfree(ies_data);
>> return (void *)param;
>> }
>>
>
--
Alexis Lothoré, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
Powered by blists - more mailing lists