kernel/linux.git/include/net/mac80211.h, branch v6.12.80

wifi: mac80211: avoid weird state in error path

2025-08-20T16:30:32+00:00

[ Upstream commit be1ba9ed221ffb95a8bb15f4c83d0694225ba808 ] If we get to the error path of ieee80211_prep_connection, for example because of a FW issue, then ieee80211_vif_set_links is called with 0. But the call to drv_change_vif_links from ieee80211_vif_update_links will probably fail as well, for the same reason. In this case, the valid_links and active_links bitmaps will be reverted to the value of the failing connection. Then, in the next connection, due to the logic of ieee80211_set_vif_links_bitmaps, valid_links will be set to the ID of the new connection assoc link, but the active_links will remain with the ID of the old connection's assoc link. If those IDs are different, we get into a weird state of valid_links and active_links being different. One of the consequences of this state is to call drv_change_vif_links with new_links as 0, since the & operation between the bitmaps will be 0. Since a removal of a link should always succeed, ignore the return value of drv_change_vif_links if it was called to only remove links, which is the case for the ieee80211_prep_connection's error path. That way, the bitmaps will not be reverted to have the value from the failing connection and will have 0, so the next connection will have a good state. Signed-off-by: Miri Korenblit Reviewed-by: Johannes Berg Link: https://patch.msgid.link/20250609213231.ba2011fb435f.Id87ff6dab5e1cf757b54094ac2d714c656165059@changeid Signed-off-by: Johannes Berg Signed-off-by: Sasha Levin

wifi: mac80211: don't complete management TX on SAE commit

2025-08-20T16:30:32+00:00

[ Upstream commit 6b04716cdcac37bdbacde34def08bc6fdb5fc4e2 ] When SAE commit is sent and received in response, there's no ordering for the SAE confirm messages. As such, don't call drivers to stop listening on the channel when the confirm message is still expected. This fixes an issue if the local confirm is transmitted later than the AP's confirm, for iwlwifi (and possibly mt76) the AP's confirm would then get lost since the device isn't on the channel at the time the AP transmit the confirm. For iwlwifi at least, this also improves the overall timing of the authentication handshake (by about 15ms according to the report), likely since the session protection won't be aborted and rescheduled. Note that even before this, mgd_complete_tx() wasn't always called for each call to mgd_prepare_tx() (e.g. in the case of WEP key shared authentication), and the current drivers that have the complete callback don't seem to mind. Document this as well though. Reported-by: Jan Hendrik Farr Closes: https://lore.kernel.org/all/aB30Ea2kRG24LINR@archlinux/ Signed-off-by: Johannes Berg Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20250609213232.12691580e140.I3f1d3127acabcd58348a110ab11044213cf147d3@changeid Signed-off-by: Johannes Berg Signed-off-by: Sasha Levin

Revert "mac80211: Dynamically set CoDel parameters per station"

2025-06-27T10:11:33+00:00

[ Upstream commit 4876376988081d636a4c4e5f03a5556386b49087 ] This reverts commit 484a54c2e597dbc4ace79c1687022282905afba0. The CoDel parameter change essentially disables CoDel on slow stations, with some questionable assumptions, as Dave pointed out in [0]. Quoting from there: But here are my pithy comments as to why this part of mac80211 is so wrong... static void sta_update_codel_params(struct sta_info *sta, u32 thr) { - if (thr && thr < STA_SLOW_THRESHOLD * sta->local->num_sta) { 1) sta->local->num_sta is the number of associated, rather than active, stations. "Active" stations in the last 50ms or so, might have been a better thing to use, but as most people have far more than that associated, we end up with really lousy codel parameters, all the time. Mistake numero uno! 2) The STA_SLOW_THRESHOLD was completely arbitrary in 2016. - sta->cparams.target = MS2TIME(50); This, by itself, was probably not too bad. 30ms might have been better, at the time, when we were battling powersave etc, but 20ms was enough, really, to cover most scenarios, even where we had low rate 2Ghz multicast to cope with. Even then, codel has a hard time finding any sane drop rate at all, with a target this high. - sta->cparams.interval = MS2TIME(300); But this was horrible, a total mistake, that is leading to codel being completely ineffective in almost any scenario on clients or APS. 100ms, even 80ms, here, would be vastly better than this insanity. I'm seeing 5+seconds of delay accumulated in a bunch of otherwise happily fq-ing APs.... 100ms of observed jitter during a flow is enough. Certainly (in 2016) there were interactions with powersave that I did not understand, and still don't, but if you are transmitting in the first place, powersave shouldn't be a problemmmm..... - sta->cparams.ecn = false; At the time we were pretty nervous about ecn, I'm kind of sanguine about it now, and reliably indicating ecn seems better than turning it off for any reason. [...] In production, on p2p wireless, I've had 8ms and 80ms for target and interval for years now, and it works great. I think Dave's arguments above are basically sound on the face of it, and various experimentation with tighter CoDel parameters in the OpenWrt community have show promising results[1]. So I don't think there's any reason to keep this parameter fiddling; hence this revert. [0] https://lore.kernel.org/linux-wireless/CAA93jw6NJ2cmLmMauz0xAgC2MGbBq6n0ZiZzAdkK0u4b+O2yXg@mail.gmail.com/ [1] https://forum.openwrt.org/t/reducing-multiplexing-latencies-still-further-in-wifi/133605/130 Suggested-By: Dave Taht In-memory-of: Dave Taht Signed-off-by: Toke Høiland-Jørgensen Link: https://patch.msgid.link/20250403183930.197716-1-toke@toke.dk Signed-off-by: Johannes Berg Signed-off-by: Sasha Levin

wifi: mac80211: set ieee80211_prep_tx_info::link_id upon Auth Rx

2025-05-29T09:02:56+00:00

[ Upstream commit 8c60179b64434894eac1ffab7396bac131bc8b6e ] This will be used by the low level driver. Note that link_id will be 0 in case of a non-MLO authentication. Also fix a call-site of mgd_prepare_tx() where the link_id was not populated. Update the documentation to reflect the current state ieee80211_prep_tx_info::link_id is also available in mgd_complete_tx(). Signed-off-by: Emmanuel Grumbach Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20250205110958.6a590f189ce5.I1fc5c0da26b143f5b07191eb592f01f7083d55ae@changeid Signed-off-by: Johannes Berg Signed-off-by: Sasha Levin

wifi: mac80211: add strict mode disabling workarounds

2025-04-20T08:15:17+00:00

[ Upstream commit 3ad4fce66e4f9d82abfc366707757e29cc14a9d2 ] Add a strict mode where we disable certain workarounds and have additional checks such as, for now, that VHT capabilities from association response match those from beacon/probe response. We can extend the checks in the future. Make it an opt-in setting by the driver so it can be set there in some driver-specific way, for example. Also allow setting this one hw flag through the hwflags debugfs, by writing a new strict=0 or strict=1 value. Signed-off-by: Johannes Berg Reviewed-by: Ilan Peer Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20250205110958.5cecb0469479.I4a69617dc60ba0d6308416ffbc3102cfd08ba068@changeid Signed-off-by: Johannes Berg Signed-off-by: Sasha Levin

wifi: mac80211: fix a queue stall in certain cases of CSA

2024-12-19T17:13:10+00:00

[ Upstream commit 11ac0d7c3b5ba58232fb7dacb54371cbe75ec183 ] If we got an unprotected action frame with CSA and then we heard the beacon with the CSA IE, we'll block the queues with the CSA reason twice. Since this reason is refcounted, we won't wake up the queues since we wake them up only once and the ref count will never reach 0. This led to blocked queues that prevented any activity (even disconnection wouldn't reset the queue state and the only way to recover would be to reload the kernel module. Fix this by not refcounting the CSA reason. It becomes now pointless to maintain the csa_blocked_queues state. Remove it. Signed-off-by: Emmanuel Grumbach Fixes: 414e090bc41d ("wifi: mac80211: restrict public action ECSA frame handling") Closes: https://bugzilla.kernel.org/show_bug.cgi?id=219447 Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20241119173108.5ea90828c2cc.I4f89e58572fb71ae48e47a81e74595cac410fbac@changeid Signed-off-by: Johannes Berg Signed-off-by: Sasha Levin

move asm/unaligned.h to linux/unaligned.h

2024-10-02T21:23:23+00:00

asm/unaligned.h is always an include of asm-generic/unaligned.h; might as well move that thing to linux/unaligned.h and include that - there's nothing arch-specific in that header. auto-generated by the following: for i in `git grep -l -w asm/unaligned.h`; do sed -i -e "s/asm\/unaligned.h/linux\/unaligned.h/" $i done for i in `git grep -l -w asm-generic/unaligned.h`; do sed -i -e "s/asm-generic\/unaligned.h/linux\/unaligned.h/" $i done git mv include/asm-generic/unaligned.h include/linux/unaligned.h git mv tools/include/asm-generic/unaligned.h tools/include/linux/unaligned.h sed -i -e "/unaligned.h/d" include/asm-generic/Kbuild sed -i -e "s/__ASM_GENERIC/__LINUX/" include/linux/unaligned.h tools/include/linux/unaligned.h

wifi: mac80211: handle ieee80211_radar_detected() for MLO

2024-09-06T11:01:05+00:00

Currently DFS works under assumption there could be only one channel context in the hardware. Hence, drivers just calls the function ieee80211_radar_detected() passing the hardware structure. However, with MLO, this obviously will not work since number of channel contexts will be more than one and hence drivers would need to pass the channel information as well on which the radar is detected. Also, when radar is detected in one of the links, other link's CAC should not be cancelled. Hence, in order to support DFS with MLO, do the following changes - * Add channel context conf pointer as an argument to the function ieee80211_radar_detected(). During MLO, drivers would have to pass on which channel context conf radar is detected. Otherwise, drivers could just pass NULL. * ieee80211_radar_detected() will iterate over all channel contexts present and * if channel context conf is passed, only mark that as radar detected * if NULL is passed, then mark all channel contexts as radar detected * Then as usual, schedule the radar detected work. * In the worker, go over all the contexts again and for all such context which is marked with radar detected, cancel the ongoing CAC by calling ieee80211_dfs_cac_cancel() and then notify cfg80211 via cfg80211_radar_event(). * To cancel the CAC, pass the channel context as well where radar is detected to ieee80211_dfs_cac_cancel(). This ensures that CAC is canceled only on the links using the provided context, leaving other links unaffected. This would also help in scenarios where there is split phy 5 GHz radio, which is capable of DFS channels in both lower and upper band. In this case, simultaneous radars can be detected. Signed-off-by: Aditya Kumar Singh Link: https://patch.msgid.link/20240906064426.2101315-9-quic_adisi@quicinc.com Signed-off-by: Johannes Berg

wifi: mac80211: export ieee80211_purge_tx_queue() for drivers

2024-08-27T08:28:55+00:00

Drivers need to purge TX SKB when stopping. Using skb_queue_purge() can't report TX status to mac80211, causing ieee80211_free_ack_frame() warns "Have pending ack frames!". Export ieee80211_purge_tx_queue() for drivers to not have to reimplement it. Signed-off-by: Ping-Ke Shih Link: https://patch.msgid.link/20240822014255.10211-1-pkshih@realtek.com Signed-off-by: Johannes Berg

wifi: mac80211: Add non-atomic station iterator

2024-08-27T08:28:53+00:00

Drivers may at times want to iterate their stations with a function which requires some non-atomic operations. ieee80211_iterate_stations_mtx() introduces an API to iterate stations while holding that wiphy's mutex. This allows the iterating function to do non-atomic operations safely. Signed-off-by: Rory Little Link: https://patch.msgid.link/20240806004024.2014080-2-rory@candelatech.com [unify internal list iteration functions] Signed-off-by: Johannes Berg