ath10k+hostapd: dynamic VLAN issue

Guenther Kelleter Guenther.Kelleter at devolo.de
Fri May 13 08:23:54 PDT 2016 

Hi

> -----Original Message-----
> From: ath10k [mailto:ath10k-bounces at lists.infradead.org] On Behalf Of Guenther
> Kelleter
> Sent: Monday, May 09, 2016 1:02 PM
> To: 'ath10k at lists.infradead.org'
> Subject: ath10k+hostapd: dynamic VLAN issue
> 
> Hi
> 
> With OpenWrt master branch (based on hostapd
> e15dcf6d1bc2725388555523effca75b1ffab735, 2016-01-15) dynamic vlans are not
> working anymore. There seems to be an Issue introduced with hostapd commit
> 7cebc8e2100626dc6981a3f53301058452953b9b (at least the error messages I see
> were introduced with this commit).
> On OpenWrt Chaos Calmer based on hostapd 2.4 dynamic VLANs are working
> properly.
> 
> But: Failure happens with ath10k driver.
> Ath9k looks good regarding this issue: vlan_bridge is created and the tagged
> wlan interface added to this bridge.
> I'm not very familiar with hostapd/mac80211 and don't know how to debug this
> further or whether this is a bug in hostapd, mac80211 or ath10k...
> 
> Following is a debug log of the failure:  look for "set_key failed"
> 
> 
> Configuration file: /var/run/hostapd-phy0.conf
> [...]
> wlan0-2: STA 78:7e:61:61:15:a3 RADIUS: VLAN ID 20
> VLAN: vlan_add_dynamic(vlan_id=20 ifname=wlan0-2.#)
> nl80211: Create interface iftype 4 (AP_VLAN)
> nl80211: Ignored event (cmd=7) for foreign interface (ifindex 60 wdev 0x0)
> nl80211: New interface wlan0-2.20 created: ifindex=60
> nl80211: Add own interface ifindex 60
> nl80211: if_indices[16]: 5 34 58 59 60
> nl80211: Add own interface ifindex 60
> nl80211: ifindex 60 already in the list
> VLAN: Set interface wlan0-2.20 up
> WPA: Add group state machine for VLAN-ID 20
> GMK - hexdump(len=32): [REMOVED]
> Key Counter - hexdump(len=32): [REMOVED]
> WPA: group state machine entering state GTK_INIT (VLAN-ID 20)
> GTK - hexdump(len=16): [REMOVED]
> IGTK - hexdump(len=16): [REMOVED]
> WPA: group state machine entering state SETKEYSDONE (VLAN-ID 20)
> wpa_driver_nl80211_set_key: ifindex=60 (wlan0-2.20) alg=3 addr=0x496be4
> key_idx=1 set_tx=1 seq_len=0 key_len=16
> nl80211: KEY_DATA - hexdump(len=16): [REMOVED]
>    broadcast key
> nl80211: set_key failed; err=-22 Invalid argument)


I could track this failure to function net/mac80211/key.c:ieee80211_key_enable_hw_accel() of mac-80211

static int ieee80211_key_enable_hw_accel(struct ieee80211_key *key)
{
[...]
        if (sdata->vif.type == NL80211_IFTYPE_AP_VLAN) {
                /*
                 * The driver doesn't know anything about VLAN interfaces.
                 * Hence, don't send GTKs for VLAN interfaces to the driver.
                 */
                if (!(key->conf.flags & IEEE80211_KEY_FLAG_PAIRWISE))         // ! This condition is true
                        goto out_unsupported;
        }
[...]
out_unsupported:
        switch (key->conf.cipher) {
        case WLAN_CIPHER_SUITE_WEP40:
        case WLAN_CIPHER_SUITE_WEP104:
        case WLAN_CIPHER_SUITE_TKIP:
        case WLAN_CIPHER_SUITE_CCMP:
        case WLAN_CIPHER_SUITE_CCMP_256:
        case WLAN_CIPHER_SUITE_AES_CMAC:
        case WLAN_CIPHER_SUITE_BIP_CMAC_256:
        case WLAN_CIPHER_SUITE_BIP_GMAC_128:
        case WLAN_CIPHER_SUITE_BIP_GMAC_256:
        case WLAN_CIPHER_SUITE_GCMP:
        case WLAN_CIPHER_SUITE_GCMP_256:
                /* all of these we can do in software - if driver can */
                if (ret == 1)
                        return 0;
                if (ieee80211_hw_check(&key->local->hw, SW_CRYPTO_CONTROL))    // ! this condition also true 
-------->               return -EINVAL;
                return 0;
        default:
                return -EINVAL;
        }
}


i.e. set_key fails because ath10k does support SW_CRYPTO_CONTROL?

ATH10k firmware is:
[   58.740000] ath10k_pci 0000:00:00.0: Direct firmware load for ath10k/cal-pci-0000:00:00.0.bin failed with error -2
[   58.750000] ath10k_pci 0000:00:00.0: Falling back to user helper
[   65.310000] ath10k_pci 0000:00:00.0: qca988x hw2.0 target 0x4100016c chip_id 0x043202ff sub 0000:0000
[   65.320000] ath10k_pci 0000:00:00.0: kconfig debug 0 debugfs 1 tracing 1 dfs 1 testmode 1
[   65.340000] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.70.42-2 api 5 features no-p2p,raw-mode,mfp crc32 44f66eae
[   65.390000] ath10k_pci 0000:00:00.0: Direct firmware load for ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
[   65.400000] ath10k_pci 0000:00:00.0: Falling back to user helper
[   65.470000] firmware ath10k!QCA988X!hw2.0!board-2.bin: firmware_loading_store: map pages failed
[   65.490000] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A crc32 bebc7c08
[   66.540000] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2 cal file max-sta 128 raw 0 hwcrypto 1


If I remove the following 2 lines from drivers/net/wireless/ath/ath10k/mac.c, set_key succeeds!

	if (!test_bit(ATH10K_FLAG_RAW_MODE, &ar->dev_flags))
		ieee80211_hw_set(ar->hw, SW_CRYPTO_CONTROL);

i.e. not setting flag SW_CRYPTO_CONTROL fixes it (apparently).

This was introduced in linux commit ccec9038c7217e537f5ae5e3ac5af8293a21bbfd "ath10k: enable raw encap mode and software crypto engine".


Can someone shed some light on this crazy stuff?
Is it bad to disable SW_CRYPTO_CONTROL for ath10k? Or does this break crypto?


Thanks
Günther


> WPA: group state machine entering state FATAL_FAILURE
> WPA: Ensure group state machine running for VLAN ID 20
> WPA initialization for VLAN 20 failed (-1)
> WPA: Try stopping group state machine for VLAN ID 20
> WPA: Remove group state machine for VLAN-ID 20
> WPA deinit of wlan0-2.20 failed
> nl80211: wpa_driver_nl80211_if_remove(type=1 ifname=wlan0-2.20) ifindex=60
> added_if=1
> nl80211: Remove interface ifindex=60
> nl80211: if_indices[16]: 5 34 58 59
>[...}
> 
> 
> 
> 
> Regards
> Günther
> 
> 
> 
> 
> _______________________________________________
> ath10k mailing list
> ath10k at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/ath10k

More information about the ath10k mailing list