linux-stable.git/drivers/net/wireless/intel, branch v6.7.2

wifi: iwlwifi: assign phy_ctxt before eSR activation

2024-01-25T23:44:55+00:00

[ Upstream commit 9b6614e5ead5d19a71893bcca3f1a6569ca0c456 ]

eSR is activated when a chanctx is assigned to more than one link.
During eSR activation we should disable RLC for both phys, and configure
the FW with a special phy command for both phys.
Currently we assign the phy_ctxt to the link only after eSR activation,
so RLC is not disabled for the new phy_ctxt, and a cmd is not sent to FW.
Fix this by first assigning the new phy_ctxt to the link and then
doing the eSR activation.

Fixes: 12bacfc2c065 ("wifi: iwlwifi: handle eSR transitions")
Reviewed-by: Gregory Greenman 
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231219215605.3d94507f5d9a.I537fcd73aedf94c7348c03157e486f24301fef14@changeid
Signed-off-by: Johannes Berg 
Signed-off-by: Sasha Levin

wifi: iwlwifi: fix out of bound copy_from_user

2024-01-25T23:44:55+00:00

[ Upstream commit cb2dfacb197bed0241fbb4f84bd0995a47f4465e ]

The driver copies the userspace buffer into an internal NUL
byte terminated buffer. While doing so, it was reading beyond
the end of the userspace buffer, overwriting its own NUL
termination in the process.

Fix this by only copying the correct number of bytes.

Fixes: 3f244876ef73 ("wifi: iwlwifi: make debugfs entries link specific")
Signed-off-by: Anjaneyulu 
Reviewed-by: Gregory Greenman 
Reviewed-by: Benjamin Berg 
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231219215605.e4913deb2ad4.Idcf6a7e909ff4b7801cd49c2f691f84a2f68eff9@changeid
Signed-off-by: Johannes Berg 
Signed-off-by: Sasha Levin

wifi: iwlwifi: mvm: Do not warn if valid link pair was not found

2024-01-25T23:44:55+00:00

[ Upstream commit c5bfdb46636a2ea7f0678243c6d3e9f8d26b027a ]

It is possible that though multiple links are enabled we cannot enabled
EMLSR enable more than a single link, e.g., all valid links are on the
same band etc. Thus, do not warn in case no valid link pair is found.

Fixes: b9be67fb4207 ("wifi: iwlwifi: mvm: Add basic link selection logic")
Signed-off-by: Ilan Peer 
Reviewed-by: Gregory Greenman 
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231219215605.142e57a05230.I7cfe78c94c3d15c4c744bccadd8f187e43594932@changeid
Signed-off-by: Johannes Berg 
Signed-off-by: Sasha Levin

wifi: iwlwifi: mvm: send TX path flush in rfkill

2024-01-25T23:44:55+00:00

[ Upstream commit 2afc3dad39ea84a072d04ff40a417234326adc47 ]

If we want to drop packets, that's surely a good thing to
do when we want to enter rfkill. Send this command despite
rfkill so we can successfully clean up everything, we need
to handle it separately since it has CMD_WANT_SKB, so it's
not going to automatically return success when in rfkill.

Fixes: d4e3a341b87b ("iwlwifi: mvm: add support for new flush queue response")
Signed-off-by: Johannes Berg 
Reviewed-by: Gregory Greenman 
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231219215605.c528a6fa6cec.Ibe5e9560359ccc0fba60c35e01de285c376748a2@changeid
Signed-off-by: Johannes Berg 
Signed-off-by: Sasha Levin

wifi: iwlwifi: mvm: set siso/mimo chains to 1 in FW SMPS request

2024-01-25T23:44:55+00:00

[ Upstream commit b1a2e5c310e063560760806d2cc5d2233c596067 ]

The firmware changed their mind, don't set the chains to zero,
instead set them to 1 as we normally would for connections to
APs that don't use MIMO.

Fixes: 2a7ce54ccc23 ("iwlwifi: mvm: honour firmware SMPS requests")
Signed-off-by: Johannes Berg 
Reviewed-by: Luciano Coelho 
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231219215605.7f031f1a127f.Idc816e0f604b07d22a9d5352bc23c445512fad14@changeid
Signed-off-by: Johannes Berg 
Signed-off-by: Sasha Levin

wifi: iwlwifi: don't support triggered EHT CQI feedback

2024-01-25T23:44:51+00:00

[ Upstream commit 637bbd5b3cbd0fc6945ebd2e311315b6cca1f9c5 ]

EHT CQI is one of the EHT PHY capabilities. We don't support EHT CQI.
The non-triggered CQI feedback bit was unset in a previous patch,
but the triggered CQI feedback bit wasn't. Unset it.

Fixes: 0e21ec6edbb5 ("wifi: iwlwifi: nvm: Update EHT capabilities for GL device")
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231207044813.092528daf59e.I5715769490835819beddb00c91bbc9e806e170cb@changeid
Signed-off-by: Johannes Berg 
Signed-off-by: Sasha Levin

wifi: iwlwifi: pcie: don't synchronize IRQs from IRQ

2023-12-18T16:41:48+00:00

On older devices (before unified image!) we can end up calling
stop_device from an rfkill interrupt. However, in stop_device
we attempt to synchronize IRQs, which then of course deadlocks.

Avoid this by checking the context, if running from the IRQ
thread then don't synchronize. This wouldn't be correct on a
new device since RSS is supported, but older devices only have
a single interrupt/queue.

Fixes: 37fb29bd1f90 ("wifi: iwlwifi: pcie: synchronize IRQs before NAPI")
Reviewed-by: Miri Korenblit 
Reviewed-by: Emmanuel Grumbach 
Signed-off-by: Johannes Berg 
Signed-off-by: Kalle Valo 
Link: https://msgid.link/20231215111335.59aab00baed7.Iadfe154d6248e7f9dfd69522e5429dbbd72925d7@changeid

wifi: iwlwifi: pcie: avoid a NULL pointer dereference

2023-12-12T09:14:57+00:00

It possible that while the rx rb is being handled, the transport has
been stopped and re-started. In this case the tx queue pointer is not
yet initialized, which will lead to a NULL pointer dereference.
Fix it.

Signed-off-by: Avraham Stern 
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231207044813.cd0898cafd89.I0b84daae753ba9612092bf383f5c6f761446e964@changeid
Signed-off-by: Johannes Berg

wifi: iwlwifi: pcie: add another missing bh-disable for rxq->lock

2023-12-12T09:14:48+00:00

Evidently I had only looked at all the ones in rx.c, and missed this.
Add bh-disable to this use of the rxq->lock as well.

Fixes: 25edc8f259c7 ("iwlwifi: pcie: properly implement NAPI")
Reported-by: Brian Norris 
Signed-off-by: Johannes Berg 
Signed-off-by: Miri Korenblit 
Link: https://msgid.link/20231208183100.e79ad3dae649.I8f19713c4383707f8be7fc20ff5cc1ecf12429bb@changeid
Signed-off-by: Johannes Berg

Merge tag 'wireless-2023-11-29' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless

2023-11-30T03:43:34+00:00

Johannes Berg says:

====================
wireless fixes:
 - debugfs had a deadlock (removal vs. use of files),
   fixes going through wireless ACKed by Greg
 - support for HT STAs on 320 MHz channels, even if it's
   not clear that should ever happen (that's 6 GHz), best
   not to WARN()
 - fix for the previous CQM fix that broke most cases
 - various wiphy locking fixes
 - various small driver fixes

* tag 'wireless-2023-11-29' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
  wifi: mac80211: use wiphy locked debugfs for sdata/link
  wifi: mac80211: use wiphy locked debugfs helpers for agg_status
  wifi: cfg80211: add locked debugfs wrappers
  debugfs: add API to allow debugfs operations cancellation
  debugfs: annotate debugfs handlers vs. removal with lockdep
  debugfs: fix automount d_fsdata usage
  wifi: mac80211: handle 320 MHz in ieee80211_ht_cap_ie_to_sta_ht_cap
  wifi: avoid offset calculation on NULL pointer
  wifi: cfg80211: hold wiphy mutex for send_interface
  wifi: cfg80211: lock wiphy mutex for rfkill poll
  wifi: cfg80211: fix CQM for non-range use
  wifi: mac80211: do not pass AP_VLAN vif pointer to drivers during flush
  wifi: iwlwifi: mvm: fix an error code in iwl_mvm_mld_add_sta()
  wifi: mt76: mt7925: fix typo in mt7925_init_he_caps
  wifi: mt76: mt7921: fix 6GHz disabled by the missing default CLC config
====================

Link: https://lore.kernel.org/r/20231129150809.31083-3-johannes@sipsolutions.net
Signed-off-by: Jakub Kicinski