linux.git/drivers/net/wireless, branch v5.16-rc4

mt76: fix key pointer overwrite in mt7921s_write_txwi/mt7663_usb_sdio_write_txwi

2021-11-29T17:33:33+00:00

Fix pointer overwrite in mt7921s_tx_prepare_skb and
mt7663_usb_sdio_tx_prepare_skb routines since in
commit '2a9e9857473b ("mt76: fix possible pktid leak")
mt76_tx_status_skb_add() has been moved out of
mt7921s_write_txwi()/mt7663_usb_sdio_write_txwi() overwriting
hw key pointer in ieee80211_tx_info structure. Fix the issue saving
key pointer before running mt76_tx_status_skb_add().

Fixes: 2a9e9857473b ("mt76: fix possible pktid leak")
Tested-by: Deren Wu 
Signed-off-by: Lorenzo Bianconi 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/eba40c84b6d114f618e2ae486cc6d0f2e9272cf9.1638193069.git.lorenzo@kernel.org

rt2x00: do not mark device gone on EPROTO errors during start

2021-11-29T10:53:50+00:00

As reported by Exuvo is possible that we have lot's of EPROTO errors
during device start i.e. firmware load. But after that device works
correctly. Hence marking device gone by few EPROTO errors done by
commit e383c70474db ("rt2x00: check number of EPROTO errors") caused
regression - Exuvo device stop working after kernel update. To fix
disable the check during device start.

Link: https://lore.kernel.org/linux-wireless/bff7d309-a816-6a75-51b6-5928ef4f7a8c@exuvo.se/
Reported-and-tested-by: Exuvo 
Fixes: e383c70474db ("rt2x00: check number of EPROTO errors")
Cc: stable@vger.kernel.org
Signed-off-by: Stanislaw Gruszka 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/20211111141003.GA134627@wp.pl

mt76: fix possible pktid leak

2021-11-24T17:21:50+00:00

Fix a possible idr pkt-id leak if the packet is dropped on tx side

Fixes: bd1e3e7b693c ("mt76: introduce packet_id idr")
Signed-off-by: Lorenzo Bianconi 
Acked-by: Felix Fietkau 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/a560caffcc24452fb48af53904bbe5c45ea5db93.1637602268.git.lorenzo@kernel.org

mt76: fix timestamp check in tx_status

2021-11-24T17:21:02+00:00

Should keep SKBs only if timeout timestamp is still after jiffies.
Otherwise, report tx status and drop it direclty.

Fixes: bd1e3e7b693c ("mt76: introduce packet_id idr")
Signed-off-by: Deren Wu 
Acked-by: Felix Fietkau 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/7e3784949c0b29a00465966b89fdb0192bd0298e.1637593492.git.deren.wu@mediatek.com

mt76: mt7915: fix NULL pointer dereference in mt7915_get_phy_mode

2021-11-22T18:43:09+00:00

Fix the following NULL pointer dereference in mt7915_get_phy_mode
routine adding an ibss interface to the mt7915 driver.

[  101.137097] wlan0: Trigger new scan to find an IBSS to join
[  102.827039] wlan0: Creating new IBSS network, BSSID 26:a4:50:1a:6e:69
[  103.064756] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000
[  103.073670] Mem abort info:
[  103.076520]   ESR = 0x96000005
[  103.079614]   EC = 0x25: DABT (current EL), IL = 32 bits
[  103.084934]   SET = 0, FnV = 0
[  103.088042]   EA = 0, S1PTW = 0
[  103.091215] Data abort info:
[  103.094104]   ISV = 0, ISS = 0x00000005
[  103.098041]   CM = 0, WnR = 0
[  103.101044] user pgtable: 4k pages, 39-bit VAs, pgdp=00000000460b1000
[  103.107565] [0000000000000000] pgd=0000000000000000, p4d=0000000000000000, pud=0000000000000000
[  103.116590] Internal error: Oops: 96000005 [#1] SMP
[  103.189066] CPU: 1 PID: 333 Comm: kworker/u4:3 Not tainted 5.10.75 #0
[  103.195498] Hardware name: MediaTek MT7622 RFB1 board (DT)
[  103.201124] Workqueue: phy0 ieee80211_iface_work [mac80211]
[  103.206695] pstate: 20000005 (nzCv daif -PAN -UAO -TCO BTYPE=--)
[  103.212705] pc : mt7915_get_phy_mode+0x68/0x120 [mt7915e]
[  103.218103] lr : mt7915_mcu_add_bss_info+0x11c/0x760 [mt7915e]
[  103.223927] sp : ffffffc011cdb9e0
[  103.227235] x29: ffffffc011cdb9e0 x28: ffffff8006563098
[  103.232545] x27: ffffff8005f4da22 x26: ffffff800685ac40
[  103.237855] x25: 0000000000000001 x24: 000000000000011f
[  103.243165] x23: ffffff8005f4e260 x22: ffffff8006567918
[  103.248475] x21: ffffff8005f4df80 x20: ffffff800685ac58
[  103.253785] x19: ffffff8006744400 x18: 0000000000000000
[  103.259094] x17: 0000000000000000 x16: 0000000000000001
[  103.264403] x15: 000899c3a2d9d2e4 x14: 000899bdc3c3a1c8
[  103.269713] x13: 0000000000000000 x12: 0000000000000000
[  103.275024] x11: ffffffc010e30c20 x10: 0000000000000000
[  103.280333] x9 : 0000000000000050 x8 : ffffff8006567d88
[  103.285642] x7 : ffffff8006563b5c x6 : ffffff8006563b44
[  103.290952] x5 : 0000000000000002 x4 : 0000000000000001
[  103.296262] x3 : 0000000000000001 x2 : 0000000000000001
[  103.301572] x1 : 0000000000000000 x0 : 0000000000000011
[  103.306882] Call trace:
[  103.309328]  mt7915_get_phy_mode+0x68/0x120 [mt7915e]
[  103.314378]  mt7915_bss_info_changed+0x198/0x200 [mt7915e]
[  103.319941]  ieee80211_bss_info_change_notify+0x128/0x290 [mac80211]
[  103.326360]  __ieee80211_sta_join_ibss+0x308/0x6c4 [mac80211]
[  103.332171]  ieee80211_sta_create_ibss+0x8c/0x10c [mac80211]
[  103.337895]  ieee80211_ibss_work+0x3dc/0x614 [mac80211]
[  103.343185]  ieee80211_iface_work+0x388/0x3f0 [mac80211]
[  103.348495]  process_one_work+0x288/0x690
[  103.352499]  worker_thread+0x70/0x464
[  103.356157]  kthread+0x144/0x150
[  103.359380]  ret_from_fork+0x10/0x18
[  103.362952] Code: 394008c3 52800220 394000e4 7100007f (39400023)

Fixes: 37f4ca907c46 ("mt76: mt7915: register per-phy HE capabilities for each interface")
Fixes: e57b7901469f ("mt76: add mac80211 driver for MT7915 PCIe-based chipsets")
Signed-off-by: Lorenzo Bianconi 
Acked-by: Felix Fietkau 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/ddae419a740f1fb9e48afd432035e9f394f512ee.1637239456.git.lorenzo@kernel.org

rtw89: update partition size of firmware header on skb->data

2021-11-22T18:42:15+00:00

The partition size is used to tell hardware the size of piece we are going
to send a firmware. The old code updates the size in constant buffer of
firmware, and leads system crash.

To fix this, update the size on skb->data after we copy the firmware data
into skb.

Buglink: https://bugzilla.opensuse.org/show_bug.cgi?id=1188303
Fixes: e3ec7017f6a2 ("rtw89: add Realtek 802.11ax driver")
Reported-by: Takashi Iwai 
Signed-off-by: Ping-Ke Shih 
Tested-by: Takashi Iwai 
Tested-by: Larry Finger 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/20211119054512.10620-2-pkshih@realtek.com

iwlwifi: Fix memory leaks in error handling path

2021-11-22T18:39:16+00:00

Should an error occur (invalid TLV len or memory allocation failure), the
memory already allocated in 'reduce_power_data' should be freed before
returning, otherwise it is leaking.

Fixes: 9dad325f9d57 ("iwlwifi: support loading the reduced power table from UEFI")
Signed-off-by: Christophe JAILLET 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/1504cd7d842d13ddb8244e18004523128d5c9523.1636615284.git.christophe.jaillet@wanadoo.fr

iwlwifi: fix warnings produced by kernel debug options

2021-11-22T18:38:46+00:00

Fix warnings produced by:
- lockdep_assert_wiphy() in function reg_process_self_managed_hint(),
- wiphy_dereference() in function iwl_mvm_init_fw_regd().
Both function are expected to be called in critical section.

The warnings were discovered when running v5.15 kernel
with debug options enabled:

1)
Hardware name: Google Delbin/Delbin
RIP: 0010:reg_process_self_managed_hint+0x254/0x347 [cfg80211]
...
Call Trace:
regulatory_set_wiphy_regd_sync+0x3d/0xb0
iwl_mvm_init_mcc+0x49d/0x5a2
iwl_op_mode_mvm_start+0x1b58/0x2507
? iwl_mvm_reprobe_wk+0x94/0x94
_iwl_op_mode_start+0x146/0x1a3
iwl_opmode_register+0xda/0x13d
init_module+0x28/0x1000

2)
drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c:263 suspicious rcu_dereference_protected() usage!
...
Hardware name: Google Delbin/Delbin, BIOS Google_Delbin
Call Trace:
dump_stack_lvl+0xb1/0xe6
iwl_mvm_init_fw_regd+0x2e7/0x379
iwl_mvm_init_mcc+0x2c6/0x5a2
iwl_op_mode_mvm_start+0x1b58/0x2507
? iwl_mvm_reprobe_wk+0x94/0x94
_iwl_op_mode_start+0x146/0x1a3
iwl_opmode_register+0xda/0x13d
init_module+0x28/0x100

Fixes: a05829a7222e ("cfg80211: avoid holding the RTNL when calling the driver")
Signed-off-by: Łukasz Bartosik 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/20211110215744.5487-1-lukasz.bartosik@semihalf.com

iwlwifi: mvm: retry init flow if failed

2021-11-22T18:38:17+00:00

In some very rare cases the init flow may fail.  In many cases, this is
recoverable, so we can retry.  Implement a loop to retry two more times
after the first attempt failed.

This can happen in two different situations, namely during probe and
during mac80211 start.  For the first case, a simple loop is enough.
For the second case, we need to add a flag to prevent mac80211 from
trying to restart it as well, leaving full control with the driver.

Cc: 
Signed-off-by: Mordechay Goodstein 
Signed-off-by: Luca Coelho 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/iwlwifi.20211110150132.57514296ecab.I52a0411774b700bdc7dedb124d8b59bf99456eb2@changeid

iwlwifi: Fix missing error code in iwl_pci_probe()

2021-11-22T18:37:25+00:00

The error code is missing in this code scenario, add the error code
'-EINVAL' to the return value 'ret'.

Eliminate the follow smatch warning:

drivers/net/wireless/intel/iwlwifi/pcie/drv.c:1376 iwl_pci_probe() warn:
missing error code 'ret'.

Reported-by: Abaci Robot 
Fixes: 1f171f4f1437 ("iwlwifi: Add support for getting rf id with blank otp")
Signed-off-by: chongjiapeng 
Acked-by: Luca Coelho 
Signed-off-by: Kalle Valo 
Link: https://lore.kernel.org/r/1635838727-128735-1-git-send-email-jiapeng.chong@linux.alibaba.com