aboutsummaryrefslogtreecommitdiff
path: root/net/wireless
AgeCommit message (Collapse)Author
2012-09-18cfg80211: fix possible circular lock on reg_regdb_search()Luis R. Rodriguez
When call_crda() is called we kick off a witch hunt search for the same regulatory domain on our internal regulatory database and that work gets kicked off on a workqueue, this is done while the cfg80211_mutex is held. If that workqueue kicks off it will first lock reg_regdb_search_mutex and later cfg80211_mutex but to ensure two CPUs will not contend against cfg80211_mutex the right thing to do is to have the reg_regdb_search() wait until the cfg80211_mutex is let go. The lockdep report is pasted below. cfg80211: Calling CRDA to update world regulatory domain ====================================================== [ INFO: possible circular locking dependency detected ] 3.3.8 #3 Tainted: G O ------------------------------------------------------- kworker/0:1/235 is trying to acquire lock: (cfg80211_mutex){+.+...}, at: [<816468a4>] set_regdom+0x78c/0x808 [cfg80211] but task is already holding lock: (reg_regdb_search_mutex){+.+...}, at: [<81646828>] set_regdom+0x710/0x808 [cfg80211] which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #2 (reg_regdb_search_mutex){+.+...}: [<800a8384>] lock_acquire+0x60/0x88 [<802950a8>] mutex_lock_nested+0x54/0x31c [<81645778>] is_world_regdom+0x9f8/0xc74 [cfg80211] -> #1 (reg_mutex#2){+.+...}: [<800a8384>] lock_acquire+0x60/0x88 [<802950a8>] mutex_lock_nested+0x54/0x31c [<8164539c>] is_world_regdom+0x61c/0xc74 [cfg80211] -> #0 (cfg80211_mutex){+.+...}: [<800a77b8>] __lock_acquire+0x10d4/0x17bc [<800a8384>] lock_acquire+0x60/0x88 [<802950a8>] mutex_lock_nested+0x54/0x31c [<816468a4>] set_regdom+0x78c/0x808 [cfg80211] other info that might help us debug this: Chain exists of: cfg80211_mutex --> reg_mutex#2 --> reg_regdb_search_mutex Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(reg_regdb_search_mutex); lock(reg_mutex#2); lock(reg_regdb_search_mutex); lock(cfg80211_mutex); *** DEADLOCK *** 3 locks held by kworker/0:1/235: #0: (events){.+.+..}, at: [<80089a00>] process_one_work+0x230/0x460 #1: (reg_regdb_work){+.+...}, at: [<80089a00>] process_one_work+0x230/0x460 #2: (reg_regdb_search_mutex){+.+...}, at: [<81646828>] set_regdom+0x710/0x808 [cfg80211] stack backtrace: Call Trace: [<80290fd4>] dump_stack+0x8/0x34 [<80291bc4>] print_circular_bug+0x2ac/0x2d8 [<800a77b8>] __lock_acquire+0x10d4/0x17bc [<800a8384>] lock_acquire+0x60/0x88 [<802950a8>] mutex_lock_nested+0x54/0x31c [<816468a4>] set_regdom+0x78c/0x808 [cfg80211] Reported-by: Felix Fietkau <nbd@openwrt.org> Tested-by: Felix Fietkau <nbd@openwrt.org> Cc: stable@vger.kernel.org Signed-off-by: Luis R. Rodriguez <mcgrof@do-not-panic.com> Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2012-09-05Merge branch 'for-john' of ↵John W. Linville
git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
2012-09-04nl80211: fix possible memory leak nl80211_connect()Wei Yongjun
connkeys is malloced in nl80211_parse_connkeys() and should be freed in the error handling case, otherwise it will cause memory leak. spatch with a semantic match is used to found this problem. (http://coccinelle.lip6.fr/) Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-08-06cfg80211: process pending events when unregistering net deviceDaniel Drake
libertas currently calls cfg80211_disconnected() when it is being brought down. This causes an event to be allocated, but since the wdev is already removed from the rdev by the time that the event processing work executes, the event is never processed or freed. http://article.gmane.org/gmane.linux.kernel.wireless.general/95666 Fix this leak, and other possible situations, by processing the event queue when a device is being unregistered. Thanks to Johannes Berg for the suggestion. Signed-off-by: Daniel Drake <dsd@laptop.org> Cc: stable@vger.kernel.org Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2012-08-02cfg80211: Clear "beacon_found" on regulatory restorePaul Stewart
Restore the default state to the "beacon_found" flag when the channel flags are restored. Otherwise, we can end up with a channel that we can no longer transmit on even when we can see beacons on that channel. Signed-off-by: Paul Stewart <pstew@chromium.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-08-02cfg80211: add channel flag to prohibit OFDM operationSeth Forshee
Currently the only way for wireless drivers to tell whether or not OFDM is allowed on the current channel is to check the regulatory information. However, this requires hodling cfg80211_mutex, which is not visible to the drivers. Other regulatory restrictions are provided as flags in the channel definition, so let's do similarly with OFDM. This patch adds a new flag, IEEE80211_CHAN_NO_OFDM, to tell drivers that OFDM on a channel is not allowed. This flag is set on any channels for which regulatory indicates that OFDM is prohibited. Signed-off-by: Seth Forshee <seth.forshee@canonical.com> Tested-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-30Merge remote-tracking branch 'wireless/master' into mac80211Johannes Berg
2012-07-25wireless: reg: restore previous behaviour of chan->max_power calculationsStanislaw Gruszka
commit eccc068e8e84c8fe997115629925e0422a98e4de Author: Hong Wu <Hong.Wu@dspg.com> Date: Wed Jan 11 20:33:39 2012 +0200 wireless: Save original maximum regulatory transmission power for the calucation of the local maximum transmit pow changed the way we calculate chan->max_power as min(chan->max_power, chan->max_reg_power). That broke rt2x00 (and perhaps some other drivers) that do not set chan->max_power. It is not so easy to fix this problem correctly in rt2x00. According to commit eccc068e8 changelog, change claim only to save maximum regulatory power - changing setting of chan->max_power was side effect. This patch restore previous calculations of chan->max_power and do not touch chan->max_reg_power. Cc: stable@vger.kernel.org # 3.4+ Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Acked-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-20Merge branch 'master' of ↵John W. Linville
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next into for-davem
2012-07-18cfg80211: Fix mutex locking in reg_last_request_cell_baseMohammed Shafi Shajakhan
should fix the following issue [ 3229.815012] [ BUG: lock held when returning to user space! ] [ 3229.815016] 3.5.0-rc7-wl #28 Tainted: G W O [ 3229.815017] ------------------------------------------------ [ 3229.815019] wpa_supplicant/5783 is leaving the kernel with locks still held! [ 3229.815022] 1 lock held by wpa_supplicant/5783: [ 3229.815023] #0: (reg_mutex){+.+.+.}, at: [<fa65834d>] reg_last_request_cell_base+0x1d/0x60 [cfg80211] Cc: Luis Rodriguez <mcgrof@gmail.com> Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com> Tested-by: Luciano Coelho <coelho@ti.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17cfg80211: remove regulatory_update()Luis R. Rodriguez
regulatory_update() just calls wiphy_update_regulatory(). wiphy_update_regulatory() assumes you already have the reg_mutex held so just move the call within locking context and kill the superfluous regulatory_update(). Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17cfg80211: make regulatory_update() staticLuis R. Rodriguez
Now that we have wiphy_regulatory_register() we can tuck away the core's regulatory_update() call there and make it static. Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17cfg80211: rename reg_device_remove() to wiphy_regulatory_deregister()Luis R. Rodriguez
This makes it clearer what we're doing. This now makes a bit more sense given that regardless of the wiphy if the cell base station hint feature is supported we will be modifying the way the regulatory core behaves. Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17cfg80211: add cellular base station regulatory hint supportLuis R. Rodriguez
Cellular base stations can provide hints to cfg80211 about where they think we are. This can be done for example on a cell phone. To enable these hints we simply allow them through as user regulatory hints but we allow userspace to clasify the hint as either coming directly from the user or coming from a cellular base station. This option is only available when you enable CONFIG_CFG80211_CERTIFICATION_ONUS. The base station hints themselves will not be processed by the core unless at least one device on the system supports this feature. Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17cfg80211: add CONFIG_CFG80211_CERTIFICATION_ONUSLuis R. Rodriguez
This adds CONFIG_CFG80211_CERTIFICATION_ONUS which is to be used for features / code which require a bit of work on the system integrator's part to ensure that the system will still pass 802.11 regulatory certification. This option is also usable for researchers and experimenters looking to add code in the kernel without impacting compliant code. We'd use CONFIG_EXPERT alone but it seems that most standard Linux distributions are enabling CONFIG_EXPERT already. This allows us to define 802.11 specific kernel features under a flag that is intended by design to be disabled by standard Linux distributions, and only enabled by system integrators or distributions that have done work to ensure regulatory certification on the system with the enabled features. Signed-off-by: Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17cfg80211: fix set_regdom() to cancel requests with same alpha2Kalle Valo
While adding regulatory support to ath6kl I noticed that I easily got the regulatory code confused. The way to reproduce the bug was: 1. iw reg set FI (in userspace) 2. cfg80211 calls ath6kl_reg_notify(FI) 3. ath6kl sets regdomain in firmware 4. firmware sends regdomain event to notify about the new regdomain (FI) 5. ath6kl calls regulatory_hint(FI) And this (from FI to FI transition) confuses cfg80211 and after that I only get "Pending regulatory request, waiting for it to be processed...." messages and regdomain changes won't work anymore. The reason why ath6kl calls regulatory_hint() is that firmware can change the regulatory domain by it's own, for example due to 11d IEs. I could of course workaround this in ath6kl but I think it's better to handle the case in cfg80211. The fix is pretty simple, use a different error code if the regdomain is same and then just set the request processed so that it doesn't block new requests. Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17cfg80211: support TX error rate CQMThomas Pedersen
Let the user configure serveral TX error conection quality monitoring parameters: % error rate, survey interval, and # of attempted packets. On exceeding the TX failure rate over the given interval, the driver will send a CQM notify event with the actual TX failure rate and packets attempted. Signed-off-by: Thomas Pedersen <c_tpeder@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-17nl80211: add wdev ID as u64 as it shouldJohannes Berg
In one of my previous patches I erroneously used nla_put_u32 for the wdev_id, fix that to use nla_put_u64. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-13cfg80211: reduce monitor interface trackingJohannes Berg
Revert commit b78e8ceac23655e1e06b30aa95ab11742d1ac7c0 ("cfg80211: track monitor channel") and remove the set_monitor_enabled() callback. Due to the tracking happening in NETDEV_PRE_UP, it had introduced bugs because the monitor interface callback would be called before the device was started. It looks like there's no way to fix this, and using NETDEV_PRE_UP is broken anyway (since there's no NETDEV_UP_FAIL), so remove all that code, track interfaces in NETDEV_UP and also stop tracking the monitor channel in cfg80211. This mostly reverts to before the tracking, except that we keep the interface count tracking so that setting the monitor channel can be rejected properly. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-13cfg80211/mac80211: re-add get_channel operationJohannes Berg
This essentially reverts commit 2e165b818456 but introduces the get_channel operation with a new wireless_dev argument so that you can retrieve the channel per interface. This is necessary as even though we can track all interface channels (except monitor) we can't track the channel type used. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-12nl80211: allow enabling WoWLAN without triggersJohannes Berg
It may be desirable to use WoWLAN without triggers to keep the connection alive to the AP while suspended. Allow this use by enabling WoWLAN without triggers if no triggers were requested. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-12Merge branch 'master' of ↵John W. Linville
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next into for-davem
2012-07-12cfg80211: ignore channel state for stopped AP/mesh interfacesFelix Fietkau
Without this change, running AP + station on the same wiphy does not work since the commit "cfg80211: add channel checking for iface combinations". The stopped AP prevents the client from connecting to an AP on a different channel. Signed-off-by: Felix Fietkau <nbd@openwrt.org> [line-break commit message to < 72 chars] Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-12cfg80211: fix locking and lockdep complaintsJohannes Berg
To call cfg80211_get_chan_state() we need to lock the wdev, so we need to lock the wdev_iter mutex in cfg80211_can_use_iftype_chan(). This needs to use nested locking for lockdep. Also, cfg80211_get_chan_state() doesn't actually use the rdev, so remove that completely including the lock assertion that isn't needed. Reported-by: Eliad Peller <eliad@wizery.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-12nl80211: move scan API to wdevJohannes Berg
The new P2P Device will have to be able to scan for P2P search, so move scanning to use struct wireless_dev instead of struct net_device. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-12nl80211: send interface after creationJohannes Berg
After a new virtual interface is created, reply to userspace with a message detailing it so it knows the new wdev identifier. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-12cfg80211: use wireless_dev for interface managementJohannes Berg
In order to be able to create P2P Device wdevs, move the virtual interface management over to wireless_dev structures. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-09cfg80211: use wdev in mgmt-tx/ROC APIsJohannes Berg
The management frame and remain-on-channel APIs will be needed in the P2P device abstraction, so move them over to the new wdev-based APIs. Userspace can still use both the interface index and wdev identifier for them so it's backward compatible, but for the P2P Device wdev it will be able to use the wdev identifier only. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-09nl80211: don't assume wdev->netdev existsJohannes Berg
There are a few places that iterate the wdev list and assume wdev->netdev exists, check there. The rfkill one has to be extended for each non-netdev type later. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-09nl80211: retrieve interface data by wdevJohannes Berg
Since soon there will be virtual interfaces that don't have a netdev, use the wdev identifier for the API to retrieve interface data. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-09nl80211: add NL80211_FLAG_NEED_WDEVJohannes Berg
Some nl80211 callbacks will soon need the wdev instead of the netdev, so add NL80211_FLAG_NEED_WDEV to allow them to request that. Add NL80211_FLAG_NEED_WDEV_UP as well which checks the netdev is UP if one exists. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-09nl80211: prepare for non-netdev wireless devsJohannes Berg
In order to support a P2P device abstraction and Bluetooth high-speed AMPs, we need to have a way to identify virtual interfaces that don't have a netdev associated. Do this by adding a NL80211_ATTR_WDEV attribute to identify a wdev which may or may not also be a netdev. To simplify things, use a 64-bit value with the high 32 bits being the wiphy index for this new wdev identifier in the nl80211 API. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-08cfg80211: fix oops due to unassigned set_monitor_enabled callbackVladimir Kondratiev
Quick fix for method being invoked without checking its existence. Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-05cfg80211: bitrate calculation for 60gVladimir Kondratiev
60g band uses different from .11n MCS scheme, so bitrate should be calculated differently Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-05{nl,cfg}80211: support high bitratesVladimir Kondratiev
Until now, a u16 value was used to represent bitrate value. With VHT bitrates this becomes too small. Introduce a new 32-bit bitrate attribute. nl80211 will report both the new and the old attribute, unless the bitrate doesn't fit into the old u16 attribute in which case only the new one will be reported. User space tools encouraged to prefer the 32-bit attribute, if available (since it won't be available on older kernels.) Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> [reword commit message and comments a bit] Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-04cfg80211: fix locking regression in monitor channel trackingJohannes Berg
Michal's monitor channel tracking introduce a locking problem as it locked the rdev lock inside the netdev notifier which isn't allowed as we might already hold it if we get there by removing an interface that is up. Fix this by relying only on the RTNL to protect the interface counters, the RTNL is always held in these code paths anyway. Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-02wireless: regulatory for 60gVladimir Kondratiev
Add regulatory rule for the 60g band Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-07-02cfg80211: add 802.11ad (60gHz band) supportVladimir Kondratiev
Add enumerations for both cfg80211 and nl80211. This expands wiphy.bands etc. arrays. Extend channel <-> frequency translation to cover 60g band and modify the rate check logic since there are no legacy mandatory rates (only MCS is used.) Signed-off-by: Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29Merge branch 'master' of ↵John W. Linville
git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next into for-davem Conflicts: drivers/net/wireless/brcm80211/brcmfmac/dhd_sdio.c
2012-06-29cfg80211: respect iface combinations when starting operationMichal Kazior
devlist_mtx locking is changed to accomodate changes. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: add channel checking for iface combinationsMichal Kazior
.connect cannot be handled since the driver scans and connects on its own. It is up to the driver then to refuse a connection (with -EBUSY for example). Non-fixed channel IBSSes always take a single channel resource. For example two non-fixed channel IBSSes always take up 2 num_different_channels, even if they operate on the same channel at a given point of time. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211/mac80211: remove .get_channelMichal Kazior
We do not need it anymore since cfg80211 tracks monitor channel and monitor channel type. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: set initial monitor channelMichal Kazior
Implements behaviour seen in mac80211. A running monitor always has a channel - even before .set_channel. This way we won't break current behaviour. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: track monitor channelMichal Kazior
Make it even more obvious we support single monitor channel. This will allow us to remove .get_channel. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: refuse to .set_monitor_channel when non-monitors are presentMichal Kazior
Having .set_monitor_channel work with non-monitor interfaces running would make interface combinations accounting ambiguous. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: track monitor interfaces countMichal Kazior
Implements .set_monitor_enabled(wiphy, enabled). Notifies driver upon change of interface layout. If only monitor interfaces become present it is called with 2nd argument being true. If non-monitor interface appears then 2nd argument is false. Driver is notified only upon change. This makes it more obvious about the fact that cfg80211 supports single monitor channel. Once we implement multi-channel we don't want to allow setting monitor channel while other interface types are running. Otherwise it would be ambiguous once we start considering num_different_channels. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: introduce cfg80211_get_chan_stateMichal Kazior
Helper function for finding out which channel is used by a given interface. An exclusive channel can be used only by a single interface. This is mainly for non-fixed channel IBSS handling. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: track ibss fixed channelMichal Kazior
IBSS may hop between channels. It is necessary to account this special case when considering interface combinations. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: add channel tracking for AP and meshMichal Kazior
We need to know which channel is used by a running AP and mesh for channel context accounting and finding matching/active interface combination. STA/IBSS have current_bss already which allows us to check which channel a vif is tuned to. Non-fixed channel IBSS can be handled with additional changes. Monitor mode is going to be handled differently. Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2012-06-29cfg80211: .stop_ap when interface is going downMichal Kazior
We'll need this for proper channel tracking (which is going to be needed for channel context accounting and finding matching/active interface combination). Signed-off-by: Michal Kazior <michal.kazior@tieto.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-14 11:47:59 +0000