- 19 Apr, 2021 13 commits
-
-
Johannes Berg authored
Some names were changed, align that with the spec as of 802.11ax-D6.1. Signed-off-by:
Luca Coelho <luciano.coelho@intel.com> Link: https://lore.kernel.org/r/iwlwifi.20210409123755.b1e5fbab0d8c.I3eb6076cb0714ec6aec6b8f9dee613ce4a05d825@changeidSigned-off-by:
Johannes Berg <johannes.berg@intel.com>
-
Naftali Goldstein authored
Don't bother keeping the link in that case. It is way too complicated to keep the connection. Signed-off-by:
Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by:
Naftali Goldstein <naftali.goldstein@intel.com> Signed-off-by:
-
Ilan Peer authored
Draft P802.11ax_D2.5 defines the following capabilities that can be negotiated using RSNXE capabilities: - Secure LTF measurement exchange protocol. - Secure RTT measurement exchange protocol. - Management frame protection for all management frames exchanged during the negotiation and range measurement procedure. Extend the nl80211 API to allow drivers to declare support for these new capabilities as part of extended feature. Signed-off-by:
Ilan Peer <ilan.peer@intel.com> Signed-off-by:
-
Emmanuel Grumbach authored
In case the frequency is invalid, ieee80211_parse_ch_switch_ie will fail and we may not even reach the check in ieee80211_sta_process_chanswitch. Drop the connection in case ieee80211_parse_ch_switch_ie failed, but still take into account the CSA mode to remember not to send a deauth frame in case if it is forbidden to. Signed-off-by:
-
Emmanuel Grumbach authored
This will allow the low level driver to get the wdev during the add_interface flow. In order to do that, remove a few checks from there and do not return NULL for vifs that were not yet added to the driver. Note that all the current callers of this helper function assume that the vif already exists: - The callers from the drivers already have a vif pointer. Before this change, ieee80211_vif_to_wdev would return NULL in some cases, but those callers don't even check they get a non-NULL pointer from ieee80211_vif_to_wdev. - The callers from net/mac80211/cfg.c assume the vif is already added to the driver as well. So, this change has no impact on existing callers of this helper function. Signed-off-by:
-
Avraham Stern authored
Add a flag that indicates that the ISTA shall indicate support for LMR feedback in NDP ranging negotiation. Signed-off-by:
Avraham Stern <avraham.stern@intel.com> Signed-off-by:
-
Avraham Stern authored
Add an enum with the values of the ranging parameters max LTF total field, as defined in IEEE802.11az_D2.6, table Table 9-322h23fc. Signed-off-by:
-
Johannes Berg authored
As crypto_shash_setkey() can fail, we should check the return value. Addresses-Coverity-ID: 1401813 ("Unchecked return value") Signed-off-by: -
Colin Ian King authored
The increment of idx is indented one level too deeply, clean up the code by removing the extraneous tab. Signed-off-by:
Colin Ian King <colin.king@canonical.com> Link: https://lore.kernel.org/r/20210416095137.2033469-1-colin.king@canonical.comSigned-off-by:
-
Johannes Berg authored
My previous commit to not apply flow control to management frames that are going over TXQs (which is currently only the case for iwlwifi, I think) broke things, with iwlwifi firmware crashing on certain frames. As it turns out, that was due to the frame being too short: space for the MIC wasn't added at the end of encrypted management frames. Clearly, this is due to using the 'frags' queue - this is meant only for frames that have already been processed for TX, and the code in ieee80211_tx_dequeue() just returns them. This caused all management frames to now not get any TX processing. To fix this, use IEEE80211_TX_INTCFL_NEED_TXPROCESSING (which is currently used only in other circumstances) to indicate that the frames need processing, and clear it immediately after so that, at least in theory, MMPDUs can be fragmented. Fixes: 73bc9e0a ("mac80211: don't apply flow control on management frames") Acked-by:
Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/r/20210416134702.ef8486a64293.If0a9025b39c71bb91b11dd6ac45547aba682df34@changeidSigned-off-by:
-
Joe Perches authored
It's not modified so make it const with the eventual goal of moving data to text for various static struct ieee80211_rate arrays. Signed-off-by:
Joe Perches <joe@perches.com> Link: https://lore.kernel.org/r/8b210b5f5972e39eded269b35a1297cf824c4181.camel@perches.comSigned-off-by:
-
Randy Dunlap authored
Fix multiple kernel-doc warnings in cfg80211.h. cfg80211.h:363: warning: missing initial short description on line: * struct ieee80211_sband_iftype_data cfg80211.h:6743: warning: missing initial short description on line: * cfg80211_vendor_cmd_get_sender Signed-off-by:
Randy Dunlap <rdunlap@infradead.org> Link: https://lore.kernel.org/r/20210417060142.1648-1-rdunlap@infradead.orgSigned-off-by:
-
Emmanuel Grumbach authored
The htmldoc produces this warning which was introduced bu the commit below. include/net/cfg80211.h:6643: warning: expecting prototype for wiphy_rfkill_set_hw_state(). Prototype was for wiphy_rfkill_set_hw_state_reason() instead Fixes: 6f779a66 ("cfg80211: allow specifying a reason for hw_rfkill") Signed-off-by:
-
- 08 Apr, 2021 16 commits
-
-
Ilan Peer authored
Remove a wrong length check for RNR information element as it can have arbitrary length. Signed-off-by:
-
Johannes Berg authored
If any of the cipher schemes specified by the driver are invalid, bail out and fail the registration rather than just warning. Otherwise, we might later crash when we try to use the invalid cipher scheme, e.g. if the hdr_len is (significantly) less than the pn_offs + pn_len, we'd have an out-of-bounds access in RX validation. Fixes: 2475b1cc ("mac80211: add generic cipher scheme support") Link: https://lore.kernel.org/r/20210408143149.38a3a13a1b19.I6b7f5790fa0958ed8049cf02ac2a535c61e9bc96@changeidSigned-off-by:
-
Emmanuel Grumbach authored
After channel switch, we should consider any beacon with a CSA IE as a new switch. If the CSA IE is a leftover from before the switch that the AP forgot to remove, we'll get a CSA-to-Self. This caused issues in iwlwifi where the firmware saw a beacon with a CSA-to-Self with mode = 1 on the new channel after a switch. The firmware considered this a new switch and closed its queues. Since the beacon didn't change between before and after the switch, we wouldn't handle it (the CRC is the same) and we wouldn't let the firmware open its queues again or disconnect if the CSA IE stays for too long. Clear the CRC valid state after we switch to make sure that we handle the beacon and handle the CSA IE as required. Signed-off-by:
-
Johan Almbladh authored
Some drivers, for example mt76, use the skb priority field, and expects that to be consistent with the skb queue mapping. On some frame injection code paths that was not true, and it broke frame injection. Now the skb queue mapping is set according to the skb priority value when the frame is injected. The skb priority value is also derived from the frame data for all frame types, as it was done prior to commit dbd50a85 (only allocate one queue when using iTXQs). Fixes frame injection with the mt76 driver on MT7610E chipset. Signed-off-by:
Johan Almbladh <johan.almbladh@anyfinetworks.com> Link: https://lore.kernel.org/r/20210401164455.978245-1-johan.almbladh@anyfinetworks.comSigned-off-by:
-
Sriram R authored
Some HW/driver can support passing ethernet rx decap frames and raw 802.11 frames for the monitor interface concurrently and via separate RX calls to mac80211. Packets going to the monitor interface(s) would be in 802.11 format and thus not have the RX_FLAG_8023 set, and 802.11 format monitoring frames should have RX_FLAG_ONLY_MONITOR set. Drivers doing such can enable the SUPPORTS_CONC_MON_RX_DECAP to allow using ethernet decap offload while a monitor interface is active, currently RX decapsulation offload gets disabled when a monitor interface is added. Signed-off-by:
Sriram R <srirrama@codeaurora.org> Link: https://lore.kernel.org/r/1617068116-32253-1-git-send-email-srirrama@codeaurora.org [add proper documentation, rewrite commit message] Signed-off-by:
-
Vamsi Krishna authored
Enhance enum nl80211_tdls_peer_capability to configure TDLS peer's support for HE mode. Userspace decodes the TDLS setup response frame and confugures the HE mode support to driver if the peer has advertized HE mode support in TDLS setup response frame. The driver uses this information to decide whether to include HE operation IE in TDLS setup confirmation frame. Signed-off-by:
Vamsi Krishna <vamsin@codeaurora.org> Link: https://lore.kernel.org/r/1614696636-30144-1-git-send-email-vamsin@codeaurora.orgSigned-off-by:
-
Emmanuel Grumbach authored
rfkill now allows to report a reason for the hw_rfkill state. Allow cfg80211 drivers to specify this reason. Keep the current API to use the default reason (RFKILL_HARD_BLOCK_SIGNAL). Signed-off-by:
-
James Prestwood authored
The docs were very sparse with how exactly CMD_ROAM should be used. Specifically related to BSS information normally obtained through a user space scan. Signed-off-by:
James Prestwood <prestwoj@gmail.com> Link: https://lore.kernel.org/r/20210311230333.103934-1-prestwoj@gmail.comSigned-off-by:
-
Lorenzo Bianconi authored
Similar to 802.11 txpath, set socket sk_pacing_shift for 802.3 tx path. Signed-off-by:
Lorenzo Bianconi <lorenzo@kernel.org> Link: https://lore.kernel.org/r/7230abc48dcf940657838546cdaef7dce691ecdd.1615240733.git.lorenzo@kernel.orgSigned-off-by:
-
Johannes Berg authored
In some cases (depending on the driver, but it's true e.g. for iwlwifi) we're using an internal TXQ for management packets, mostly to simplify the code and to have a place to queue them. However, it appears that in certain cases we can confuse the code and management frames are dropped, which is certainly not what we want. Short-circuit the processing of management frames. To keep the impact minimal, only put them on the frags queue and check the tid == management only for doing that and to skip the airtime fairness checks, if applicable. Signed-off-by:
-
Aloka Dixit authored
Add NL80211_FILS_DISCOVERY_ATTR_TMPL explicitly in nl80211_fils_discovery_policy definition. Signed-off-by:
Aloka Dixit <alokad@codeaurora.org> Link: https://lore.kernel.org/r/20210222212059.22492-1-alokad@codeaurora.orgSigned-off-by:
-
Colin Ian King authored
The variable result is being assigned a value that is never read and it is being updated later with a new value. The initialization is redundant and can be removed. Addresses-Coverity: ("Unused value") Signed-off-by: -
Wei Yongjun authored
GCC reports the following warning with W=1: net/mac80211/rc80211_minstrel_ht.c:871:34: warning: variable 'mg' set but not used [-Wunused-but-set-variable] 871 | struct minstrel_mcs_group_data *mg; | ^~ This variable is not used in function , this commit remove it to fix the warning. Reported-by:Hulk Robot <hulkci@huawei.com> Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Link: https://lore.kernel.org/r/20210326024843.987941-1-weiyongjun1@huawei.comSigned-off-by:
-
Qiheng Lin authored
spinlock can be initialized automatically with DEFINE_SPINLOCK() rather than explicitly calling spin_lock_init(). Reported-by:
Qiheng Lin <linqiheng@huawei.com> Link: https://lore.kernel.org/r/20210325143854.13186-1-linqiheng@huawei.comSigned-off-by:
-
Guobin Huang authored
spinlock can be initialized automatically with DEFINE_SPINLOCK() rather than explicitly calling spin_lock_init(). Reported-by:
Guobin Huang <huangguobin4@huawei.com> Link: https://lore.kernel.org/r/1616839160-6654-1-git-send-email-huangguobin4@huawei.comSigned-off-by:
-
Guobin Huang authored
spinlock can be initialized automatically with DEFINE_SPINLOCK() rather than explicitly calling spin_lock_init(). Reported-by:
-
- 07 Apr, 2021 11 commits
-
-
Wei Yongjun authored
Use the module_pci_driver() macro to make the code simpler by eliminating module_init and module_exit calls. Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Wei Yongjun authored
Use the module_pci_driver() macro to make the code simpler by eliminating module_init and module_exit calls. Signed-off-by:
-
Wei Yongjun authored
Use the module_pci_driver() macro to make the code simpler by eliminating module_init and module_exit calls. Signed-off-by:
-
Wei Yongjun authored
Use the module_pci_driver() macro to make the code simpler by eliminating module_init and module_exit calls. Signed-off-by:
-
Wei Yongjun authored
Use the module_pci_driver() macro to make the code simpler by eliminating module_init and module_exit calls. Signed-off-by:
-
Wei Yongjun authored
Use the module_pci_driver() macro to make the code simpler by eliminating module_init and module_exit calls. Signed-off-by:
-
Wei Yongjun authored
module_spi_driver() makes the code simpler by eliminating boilerplate code. Signed-off-by:
-
Johan Hovold authored
According to the changelog, asynchronous mode was dropped sometime before v2.2. Let's get rid of the unused driver-specific async state as well so that it doesn't show up when doing tree-wide tty work. Signed-off-by:
Johan Hovold <johan@kernel.org> Signed-off-by:
-
Colin Ian King authored
The macro CN23XX_PEM_BAR1_INDEX_REG is being used to shift oct->pcie_port (a u16) left 24 places. There are two subtle issues here, first the shift gets promoted to an signed int and then sign extended to a u64. If oct->pcie_port is 0x80 or more then the upper bits get sign extended to 1. Secondly shfiting a u16 24 bits will lead to an overflow so it needs to be cast to a u64 for all the bits to not overflow. It is entirely possible that the u16 port value is never large enough for this to fail, but it is useful to fix unintended overflows such as this. Fix this by casting the port parameter to the macro to a u64 before the shift. Addresses-Coverity: ("Unintended sign extension") Fixes: 5bc67f58 ("liquidio: CN23XX register definitions") Signed-off-by: -
Colin Ian King authored
The error check on err is always false as err is always 0 at the port_found label. The code is redundant and can be removed. Addresses-Coverity: ("Logically dead code") Signed-off-by: -
David S. Miller authored
Merge tag 'linux-can-next-for-5.13-20210407' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next Marc Kleine-Budde says: ==================== pull-request: can-next 2021-04-07 this is a pull request of 6 patches for net-next/master. The first patch targets the CAN driver infrastructure, it improves the alloc_can{,fd}_skb() function to set the pointer to the CAN frame to NULL if skb allocation fails. The next patch adds missing error handling to the m_can driver's RX path (the code was introduced in -next, no need to backport). In the next patch an unused constant is removed from an enum in the c_can driver. The last 3 patches target the mcp251xfd driver. They add BQL support and try to work around a sometimes broken CRC when reading the TBC register. ==================== Signed-off-by:
-
