Go back –> Atheros Linux wireless drivers
ath5k is a completely FOSS wireless driver for Atheros WLAN based chipsets.
ath5k is currently available through different places.
Developers are encouraged to work using the git repository. If you are not familiar with git please check out our Linux wireless git-guide. Alternatively you can use the Linux wireless compatibility package but please be sure to post patches in unified diff format (diff -u). To learn how to submit patches please read our Submitting patches guideline.
Read this section on Atheros specifications and documentation.
To enable ath5k in the kernel configuration, you must first enable mac80211:
Networking ---> Wireless ---> <M> Improved wireless configuration API <M> Generic IEEE 802.11 Networking Stack (mac80211)
Please note that there is another 802.11 networking stack:
<M> Generic IEEE 802.11 Networking Stack
You do not need this. This option enables the old SoftMAC stack we hope to kill one of these days. You can still safely enable this though.
You can then enable ath5k in the kernel configuration under
Device Drivers ---> [*] Network device support ---> Wireless LAN ---> <M> Atheros 5xxx wireless cards support
AR5210 - 802.11a (Crete/fez) AR5211 - 802.11ab (Oahu) (draft g -OFDM only- supported by hw but not by ath5k) AR5212 - 802.11abg (Venice) AR5213 - 802.11abg (Hainan) AR2413/4 - 802.11bg (Griffin) AR5413/4 - 802.11abg (Eagle) AR5423/4 - 802.11abg (Condor) (PCI-E) AR2425 - 802.11bg (Swan) (PCI-E) AR2417 - 802.11bg (Nala) (PCI-E)
To try the driver you can do this:
modprobe ath5k sudo ip link set wlan%d up sudo iwconfig wlan%d essid any # Make sure you get auth'd and then assoc'd # Then either set an IP manually or get it via DHCP ping gw
You'll probably see an immediate rate drop to 1M, this is because of how we currently handle rate control. You should be able to keep at least at 11M for 802.11b/g but for now set this manually:
iwconfig wlan%d rate 11M
Currently supported PCI ID list with respective status report on basic-testing as defined above.
Vendor:device Type Name Basic-testing 168c:0207 AR5210 AR5210 No tx - working on it 168c:0007 << << No tx - working on it 168c:0011 AR5211 AR5211 Should work 168c:0012 << << OK 168c:0013 AR5212 AR5212 OK a727:0013 << << Should Work 10b7:0013 << << << 168c:0014 << << << 168c:0015 << << << 168c:0016 << << << 168c:0017 << << << 168c:0018 << << << 168c:0019 << << << 168c:001a << AR2413 OK 168c:001b << AR5413 OK 168c:001c << AR5424 OK (*) 168c:001d << AR2417 OK
Notes on supported devices
"OK" means that card operates almost as good as with binary drivers (performance is around 25Mbits max depending on environment -that's because we still don't support ANI or tx power calibration)
"Should work" means that card hasn't been tested (we haven't received any reports) but since chipset is the same with a tested, working card, it should work as well.
(*) 001c is used also for RF2425 parts.
MadWifi and ndis drivers perform better on noisy environments than ath5k because we don't support Adaptive Noise Immunity (ANI) yet.
Supports 802.11abg, depending on the chipset. This driver requires no firmware or binary-only HAL !
Mesh Point Mode
Access Point Mode (enabled in Linux 2.6.31 and newer and in compat-wireless, can also be enabled by patching an older kernel.)
not working yet
* We don't plan on supporting XR mode nor dynamic 20/40MHz turbo mode supported by hw.
Atheros common module
ath5k uses the common shared ath.ko module.
We have kerneldocs available for ath5k:
- General driver structure:
- Main PCI and mac80211 driver interface:
- Debug utilities:
- Initialization and PHY control:
- Hardware registers:
Reported bugs on ath5k
This is a collection of bug reports both unresolved and resolved to help users track issues and to find patches for fixes which have not yet been merged.
Things ath5k developers are currently working on.
- Automatic Noise Immunity (Nick -code needs review)
- Tx power support (setting tx power) (Nick/Felix -works but experiments show that the card transmits only on some standard power levels instead of a power range as expected)
- Rework phy calibration (I/Q, noise floor etc) (Nick)
- 802.11d/Regulatory enhancements (Bob)
802.11e/WME/QoS support - see Notes on QoS
- Rfkill (pending patch by Tobias Doerffel)
Jumbo frame processing (these appear to just be errors)
- CCMP hardware support (Bob - have a patch, but it's broken)