summaryrefslogtreecommitdiff
path: root/sys/net80211
AgeCommit message (Collapse)Author
2018-02-19Remove almost unused `flags' argument of suser().Martin Pieuchot
The account flag `ASU' will no longer be set but that makes suser() mpsafe since it no longer mess with a per-process field. No objection from millert@, ok tedu@, bluhm@
2018-02-06convert __inline to inlinePeter Hessler
no binary change on amd64
2018-02-06convert the test if a node support HT (aka, 11n) into an inline function,Peter Hessler
instead of testing some un-obvious bitfield OK stsp@
2018-02-06remove extern from function prototypes, remove some variable names, and aPeter Hessler
left-over kernel malloc feature we don't use. OK deraadt@
2018-02-05Add a new function hook to struct ieee80211com which wireless driversStefan Sperling
can use to process, and then acknowledge or reject, incoming AUTH requests in hostap mode. net80211 accepts an AUTH request from any STA which fits into the node cache. This behaviour doesn't work for devices which have a lower limit on concurrent STAs they can serve, so such drivers need an override. This will be used by our athn(4) USB driver soon. ok kevlo@
2017-12-21Call if_start() directly to send the EAPOL key.Martin Pieuchot
Using if_enqueue() here, from interrupt context, might result in the packet beeing enqueued, incorrectly encrypted, on the TX ring. This race has been recently exposed by the re-introduction of the TX mitigation. It exists because the net80211 stack sets IEEE80211_NODE_TXPROT on the node while processing the 3rd message, assuming the answer has already been transmitted. However a CPU returns from if_enqueue() it cannot assume that the send queue is empty. So call if_start() to flush this queue. Encrypting the 4th message of the 4way handshake with the new key breaks WPA handshake as found the hardway by anton@. Race analysed by dlg@, a lot of net80211 inputs and suggetions from stsp@. ok stsp@, dlg@
2017-12-14My previous commit broke the ramdisk build; fix itStefan Sperling
2017-12-14Stop printing debug info about stations leaving a wifi network whenStefan Sperling
we aren't running in hostap or ibss mode.
2017-12-12Only trigger background scans while we are in RUN state, and do not updateStefan Sperling
a node's RSSI info while we are still in INIT state. ok phessler@
2017-12-12Make ieee80211_chan2ieee() panic if a bogus channel pointer is detected.Stefan Sperling
ok patrick@
2017-12-12Make IEEE80211_CHAN_ANYC be a NULL pointer instead of a pointer to 0xffff.Stefan Sperling
Should speed up debugging. ok phessler patrick
2017-12-12Don't dereference IEEE80211_CHAN_ANYC pointer in ieee80211_node_checkrssi().Stefan Sperling
Problem reported by Gregoire Jadi on bugs@
2017-12-12Allow interfaces that aren't SCANNALLBAND drivers to benefit from thePeter Hessler
strongest received signal. OK stsp@
2017-12-12Fix the behavior of preferring weaker-but-still-good 5Ghz APs overJoshua Stein
2Ghz APs because the 5Ghz band is generally less saturated. The previous implementation was dependent upon the order of walking APs. ok stsp
2017-12-11Omit the PMKID count field from RSN information elements (IE) if the PMKIDStefan Sperling
list has zero elements and PMKID would be the last field in the RSN IE. This is correct as per 802.11-2012 8.4.2.27.1 and aligns net80211 code with behaviour of bwfm(4) firmware, unblocking further progress in that driver. ok patrick@ phessler@
2017-12-08Add support for background scanning to net80211 and iwm(4).Stefan Sperling
The iwm(4) driver will now roam between access points which share an SSID. Use 'ifconfig iwm0 debug' and 'tail -f /var/log/messages' to watch it do so. Tested by several people in various iterations. As usual, let me know if you run into issues. ok phessler deraadt
2017-11-27Stop reporting WPA and WEP keys back to userland.Stefan Sperling
The kernel is not a password database; look your wifi keys up elsewhere. Discussed with several. ok phessler@ jca@
2017-11-06move a function declaration, so the whole net80211 stack can disable wep or wpaPeter Hessler
OK stsp@
2017-11-05Changing nwid on a wifi network means it is a new network, so clear thePeter Hessler
WPA and WEP configuration. OK pirofti@ stsp@ sthen@
2017-10-29Kill unused members using forward declaration only types.Martin Pieuchot
Found with ctfconv(1). ok jsg@, guenther@
2017-10-27Remove 80211WMMPARMS ioctls. Last used in ifconfig in 2009.Jonathan Gray
ok stsp@ kevlo@ jca@
2017-10-26Move common code to add/remove multicast filters to ieee80211_ioctl(9).Martin Pieuchot
ok jsg@, stsp@
2017-10-24remove defines for ioctls the kernel doesn't recogniseJonathan Gray
ok mpi@
2017-10-21Make ieee80211_classify() available in a header so we can make thePatrick Wildt
priority visible to underlying bus protocols like bwfm(4)'s bcdc.
2017-10-16Add comments which document already fixed WPA attack vectors.Stefan Sperling
2017-09-05When starting a new scan always set the mode to AUTO if the driver scansStefan Sperling
all bands at once. Fixes a problem where e.g. 5GHz APs were filtered out if we were previously associated to an 11g-only AP. ok mpi@ phessler@
2017-09-05When selecting the next wireless mode during the scan loop, always selectStefan Sperling
AUTO mode if the driver scans all bands at once. Otherwise the net80211 layer unnecessarily filters out some of the beacons received by the device. ok phessler@ mpi@ kevlo@
2017-09-04If a beacon is received in RUN state, reset the management timer.Stefan Sperling
Some wifi drivers send a probe request if the hardware reports "missed beacon" events. If the AP replies with a probe response it is still servicing us and there is no need to search for a new AP. However, the management timer was not reset if a beacon was received while in RUN state. So the interface watchdog always ended up putting the driver into SCAN state after a missed beacon event, even if the AP did respond to our probe request. Under some conditions this bug would cause spurious disconnects. Problem reported and fix tested by mlarkin@ (Using the management timer in RUN state is a new convention. Before support for missed beacons was added, this timer was only used during the association sequence to handle APs which don't respond to our assoc requests and such.)
2017-08-18Clear WPA group keys from memory before initiating a key exchangeStefan Sperling
with an access point. Prevents false positive 'reused group key' warnings in dmesg when re-associating to the same access point. Problem reported by tb@ ok tb@
2017-08-17Add an entry to dmesg if pairwise WPA keys arrive unexpectedly or if WPAStefan Sperling
group keys are being reused. OpenBSD wireless clients will now leave a trail of such events in their message log. There has been increased public scrutiny of WPA's security recently, so I am curious to see if anyone is attempting replay attacks in the wild. ok deraadt
2017-08-04Compile a debug printf in ieee80211_match_bss() by default, previouslyStefan Sperling
guarded by the IEEE80211_DEBUG preprocessor flag. This shows one line per detected AP after a scan, and indicates which APs are considered candidates for association. Shorten the output a bit to fit into 80 columns more likely. ok sthen@
2017-08-04Remove ieee80211_input_print() which printed information about receivedStefan Sperling
frames to dmesg, if debug mode was enabled with ifconfig. This debug output was much too verbose and not actually useful for debugging. tcpdump -y IEEE802_11_RADIO will show the same information. ok sthen@
2017-08-04Show net80211 state transitions in desmg if 'ifconfig debug' was used.Stefan Sperling
This information is needed in bug reports. Convert the invalid state transitions from panic() to a printf() which is also guarded by ifconfig debug. There are many races exposed by these panics which should all be fixed. But that will surely take some time, and the panics have now served their purpose. Thanks to everyone who reported these panics being triggered, your help is appreciated.
2017-07-22Make the kernel panic if an invalid state transition occurs in net80211.Stefan Sperling
Triggers on driver bugs such as those which were fixed in rsu(4) recently. ok kevlo@
2017-07-22Fix length checks in EAPOL key frame parsing.Stefan Sperling
Problem reported by Ilja Van Sprundel. ok tb@ kevlo@
2017-07-19Plug an information leak in ieee80211_node2req().Stefan Sperling
Problem reported by Ilja Van Sprundel. ok tb@
2017-07-19Improve the heuristic which selects 5GHz APs over 2GHz APs.Stefan Sperling
The previous code wasn't quite right: it didn't account for the fact that some drivers don't set ic_max_rssi, and it compared 5GHz APs to a threshold relative to the max RSSI, rather than comparing RSSI on 5GHz relative to RSSI on 2GHz. This heuristic is only used by SCANNALLBAND drivers: iwn(4), iwm(4), wpi(4) In the future the AP selection heuristic should be made more intelligent, e.g. it should take BSS load information in beacons into account. Another open problem is inconsistent representation of RSSI measurement values throughout our drivers and stack. Help is welcome! For now, this hopefully improves AP selection at busy airports. ok sthen@ deraadt@
2017-07-02Add the definition of IEEE80211_DUR_DS_SHSLOT.Kevin Lo
From IEEE Std. 802.11-2016, Table 18-5 "ERP characteristics", p. 2332: aSlotTime characteristic: If dont11OperatingClassesRequired is false: Long = 20 us Short = 9 us ok stsp@
2017-06-20Initialize the link state of a wireless interface to DOWN when theStefan Sperling
interface is attached to the net80211 layer. Prevents confusion in cases where drivers forget to initialize the link state. ok mpi@ kettenis@
2017-06-04Add sizes for free for the RSN IEs. Rewrite ieee80211_save_ie() slightlyTheo Buehler
to make it more readable. help, many explanations and ok stsp
2017-06-03Add a few sizes to free().Theo Buehler
Input, help & ok stsp
2017-06-03Explicitly zero out the wepseed for TKIP and WEP.Theo Buehler
ok stsp
2017-06-02Scale the missed beacon counter threshold to the AP's beacon interval.Stefan Sperling
This should make fading APs time out consistently regardless of what the beacon interval is set to (range is 1 to 2^16 TU, though in practice 100 TU seems to be a common value). Print the beacon interval and missed beacon counter threshold to dmesg if the DEBUG flag was set on the wireless interface with ifconfig(8). This should help with diagnosing any issues that pop up. Requested and diff eye-balled by kettenis@ help & ok tb@ phessler@
2017-05-31The net80211 stack was providing a 'beacon miss timeout' value (in ms)Stefan Sperling
which specified how much time may elapse without beacons before drivers begin searching for a new AP. Drivers convert this timeout value into the amount of beacons they're allowed to miss. Having the stack provide this number upfront simplifies things. ok mpi@
2017-05-30Improve the new ieee80211_{min,max}_rates() APIs and fix regressions.Stefan Sperling
Instead of returning an index into ni_rates, return the RVAL of the basic rate we want to use. This allows a driver to unambiguously map the basic rate to the corresponding hardware-specific rate value, and reduces the possibility of bugs where indices are used with arrays they weren't intended for. Adjust iwn(4) accordingly, and use the lowest instead of the highest basic rate in iwn_tx() to cope better in noisy environments. Fixes association problems on 5GHz reported by tb@
2017-05-30Introduce ieee80211_min_basic_rate() and ieee80211_max_basic_rate().Stefan Sperling
These helpers can be used by drivers to improve compatibility with APs that disable some mandatory PHY rates in the basic rate set. For instance, many of our drivers hard-code 11b rates on 2 Ghz and run into problems when APs disable them. Since 11b rates are being disabled by default by some vendors, hardcoding them is not a good idea anymore. ok mpi@ phessler@
2017-05-30Always set the link state DOWN once we enter ieee80211_newstate(), regardlessStefan Sperling
of whether the wifi interface happens to be leaving RUN state. The interface is never usable during state transitions so setting the link DOWN is the only reasonable option when any transition is triggered. Fixes a problem where, at boot time, the link state of wifi interfaces was reported to userland as UNKNOWN (which, curiously, has value 0). dhclient's link detection logic was recently changed from ifmedia to getifaddrs which exposed the UNKOWN link state. Since dhclient assumes an UNKNOWN link state means UP it would start trying to negotiate a lease too early during boot. Problem reported by tb@ ok krw@
2017-05-02Switch 802.11 crypto over to the new AESMike Belopuhov
OK stsp@
2017-05-02Fix a problem with associating to wifi networks with a hidden SSID.Stefan Sperling
If an AP is configured to hide its SSID it sends a non-zero length SSID which contains only zeroes. The AP sends its actual SSID only in probe responses after a client includes this SSID in a probe request. If we happened to receive a beacon before the probe response we stored a non-zero-length SSID of zeroes and never updated the SSID when the probe response arrived. The client was then unable to find the AP. test & ok jung@
2017-04-23Handle unequal numbers of Tx and Rx streams in MiRA.Stefan Sperling
Problem reported by Colton Lewis on misc@ ok tb@
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-30 21:01:10 +0000