- 09 May, 2022 2 commits
-
-
Andy Shevchenko authored
GPIO library now accepts fwnode as a firmware node, so switch the driver to use it. Signed-off-by:
Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by:
Kalle Valo <kvalo@kernel.org> Link: https://lore.kernel.org/r/20220504092525.71668-1-andriy.shevchenko@linux.intel.com
-
Chin-Yen Lee authored
Fine tune algorithm of adaptivity sensitivity to avoid disconnecting from AP suddenly in field. Signed-off-by:
Chin-Yen Lee <timlee@realtek.com> Signed-off-by:
Ping-Ke Shih <pkshih@realtek.com> Signed-off-by:
-
- 06 May, 2022 4 commits
-
-
H. Nikolaus Schaller authored
With introduction of vmap'ed stacks, stack parameters can no longer be used for DMA and now leads to kernel panic. It happens at several places for the wl1251 (e.g. when accessed through SDIO) making it unuseable on e.g. the OpenPandora. We solve this by allocating temporary buffers or use wl1251_read32(). Tested on v5.18-rc5 with OpenPandora. Fixes: a1c510d0 ("ARM: implement support for vmap'ed stacks") Signed-off-by:
H. Nikolaus Schaller <hns@goldelico.com> Signed-off-by:
-
Jakub Kicinski authored
qtnfmac chooses its own magic NAPI weight so switch to the new API created for those who don't use NAPI_POLL_WEIGHT. Signed-off-by:
Jakub Kicinski <kuba@kernel.org> Signed-off-by:
-
Jakub Kicinski authored
Switch to the new API not requiring passing in NAPI_POLL_WEIGHT. Signed-off-by:
-
Jakub Kicinski authored
Switch to the new API not requiring passing in NAPI_POLL_WEIGHT. Signed-off-by:
-
- 04 May, 2022 21 commits
-
-
Muna Sinada authored
Allow userspace to disable EHT mode. This forces EHT capable interfaces to disable during association. Signed-off-by:
Muna Sinada <quic_msinada@quicinc.com> Signed-off-by:
Aloka Dixit <quic_alokad@quicinc.com> Link: https://lore.kernel.org/r/20220323224636.20211-2-quic_alokad@quicinc.com [remove stray message change] Signed-off-by:
Johannes Berg <johannes.berg@intel.com>
-
Muna Sinada authored
Allow userspace to disable EHT mode during association. Signed-off-by:
-
Felix Fietkau authored
In client mode, we can't connect to hidden SSID APs or SSIDs not advertised in beacons on DFS channels, since we're forced to passive scan. Fix this by sending out a probe request immediately after the first beacon, if active scan was requested by the user. Cc: stable@vger.kernel.org Reported-by:
Catrinel Catrinescu <cc@80211.de> Signed-off-by:
Felix Fietkau <nbd@nbd.name> Link: https://lore.kernel.org/r/20220420104907.36275-1-nbd@nbd.nameSigned-off-by:
-
Hangyu Hua authored
If statement is meaningless because the code will goto out regardless of whether fast_tx is NULL or not. Signed-off-by:
Hangyu Hua <hbh25y@gmail.com> Link: https://lore.kernel.org/r/20220413091902.27438-1-hbh25y@gmail.comSigned-off-by:
-
Ping-Ke Shih authored
Follow IEEE 802.11-21 that HTC subfield masked to 0 for all data frames containing a QoS Control field. It also defines the AAD length depends on QC and A4 fields, so change logic to determine length accordingly. Signed-off-by:
-
Dimitri John Ledkov authored
Add MODULE_FIRMWARE declarations for regulatory.db and regulatory.db.p7s such that userspace tooling can discover and include these files. Cc: stable@vger.kernel.org Signed-off-by:
Dimitri John Ledkov <dimitri.ledkov@canonical.com> Link: https://lore.kernel.org/r/20220414125004.267819-1-dimitri.ledkov@canonical.comSigned-off-by:
-
Johannes Berg authored
Since internal_flags is only 8 bits, we can only have one more internal flag. However, we can obviously never use all of possible the combinations, in fact, we only use 14 of them (including no flags). Since we want more flags for MLO (multi-link operation) in the future, refactor the code to use a flags selector, so wrap all of the .internal_flags assignments in a IFLAGS() macro which selects the combination according to the pre- defined list of combinations. When we need a new combination, we'll have to add it, but again we will never use all possible combinations. Signed-off-by:
-
Johannes Berg authored
It's not necessary to hold the RTNL across color change requests, since all the inner locking needs only the wiphy mutex which we already hold as well. Fixes: 0d2ab3ae ("nl80211: add support for BSS coloring") Link: https://lore.kernel.org/r/20220414140402.32e03e8c261b.I5e7dc6bc563a129b938c43298da6bb4e812400a5@changeidSigned-off-by:
-
Johannes Berg authored
We haven't used this function for years, since commit c781944b ("cfg80211: Remove unused cfg80211_can_use_iftype_chan()") which itself removed a function unused since commit 97dc94f1 ("cfg80211: remove channel_switch combination check"), almost eight years ago. Also remove the now unused enum cfg80211_chan_mode and some struct members that were only used for this function. Link: https://lore.kernel.org/r/20220412220958.1a191dca19d7.Ide4448f02d0e2f1ca2992971421ffc1933a5370a@changeidSigned-off-by:
-
Ping-Ke Shih authored
This initial vesion is usable now. It can support STA, AP and monitor modes, so we can add 8852ce to Kconfig and Makefile. We are still working on some features, such as deep power save, and BT coexistence. But, this version still can have a good WiFi-only performance already, and will continue to fine tune power consumption. Signed-off-by:
-
Ping-Ke Shih authored
To fix the compiler warning of clang with i386 config, but not complain by gcc: __write_ctrl(R_AX_PWR_RATE_CTRL, B_AX_FORCE_PWR_BY_RATE_VALUE_MASK, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/net/wireless/realtek/rtw89/rtw8852c.c:2621:13: note: expanded from macro '__write_ctrl' u32 _wrt = FIELD_PREP(__msk, _val); \ ^~~~~~~~~~~~~~~~~~~~~~~ include/linux/bitfield.h:114:3: note: expanded from macro 'FIELD_PREP' __BF_FIELD_CHECK(_mask, 0ULL, _val, "FIELD_PREP: "); \ ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/bitfield.h:71:53: note: expanded from macro '__BF_FIELD_CHECK' BUILD_BUG_ON_MSG(__bf_cast_unsigned(_mask, _mask) > \ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~ note: (skipping 1 expansions in backtrace; use -fmacro-backtrace-limit=0 to see all) include/linux/compiler_types.h:352:22: note: expanded from macro 'compiletime_assert' _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__) ~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/compiler_types.h:340:23: note: expanded from macro '_compiletime_assert' __compiletime_assert(condition, msg, prefix, suffix) ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/compiler_types.h:332:9: note: expanded from macro '__compiletime_assert' if (!(condition)) \ ^~~~~~~~~ Reported-by:kernel test robot <lkp@intel.com> Signed-off-by:
-
Ping-Ke Shih authored
First one could affect SER because of false alarm event. Second one can affect spur elimination. Signed-off-by:
-
Ping-Ke Shih authored
Without this fix, it would cause IOT issue due to AID mismatch. Signed-off-by:
-
Ping-Ke Shih authored
By experiments, it spends ~45/1090~2480us to enter/leave low power mode, so the old polling interval 1000us can waste time. Use smaller polling interval depends on experimental results to reduce the time to transition state. Signed-off-by:
-
Ping-Ke Shih authored
The chip_info include BT coexistence tables, size and number of hardware components, and supported functions. Signed-off-by:
-
Ping-Ke Shih authored
Add to configure BT share RX path and related settings. Signed-off-by:
-
Ping-Ke Shih authored
The commit 16b44ed0 ("rtw89: add RF H2C to notify firmware") is to add firmware command, and this commit is to prepare the channels. Then, firmware can get proper channels. Signed-off-by:
-
Tetsuo Handa authored
Flushing system-wide workqueues is dangerous and will be forbidden. Replace system_highpri_wq with per "struct wfx_dev" bh_wq. Signed-off-by:
Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Acked-by:
Jérôme Pouiller <jerome.pouiller@silabs.com> Signed-off-by:
-
Jakub Kicinski authored
Merge tag 'wireless-next-2022-05-03' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next Kalle Valo says: ==================== wireless-next patches for v5.19 First set of patches for v5.19 and this is a big one. We have two new drivers, a change in mac80211 STA API affecting most drivers and ath11k getting support for WCN6750. And as usual lots of fixes and cleanups all over. Major changes: new drivers - wfx: silicon labs devices - plfxlc: pureLiFi X, XL, XC devices mac80211 - host based BSS color collision detection - prepare sta handling for IEEE 802.11be Multi-Link Operation (MLO) support rtw88 - support TP-Link T2E devices rtw89 - support firmware crash simulation - preparation for 8852ce hardware support ath11k - Wake-on-WLAN support for QCA6390 and WCN6855 - device recovery (firmware restart) support for QCA6390 and WCN6855 - support setting Specific Absorption Rate (SAR) for WCN6855 - read country code from SMBIOS for WCN6855/QCA6390 - support for WCN6750 wcn36xx - support for transmit rate reporting to user space * tag 'wireless-next-2022-05-03' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next: (228 commits) rtw89: 8852c: rfk: add DPK rtw89: 8852c: rfk: add IQK rtw89: 8852c: rfk: add RX DCK rtw89: 8852c: rfk: add RCK rtw89: 8852c: rfk: add TSSI rtw89: 8852c: rfk: add LCK rtw89: 8852c: rfk: add DACK rtw89: 8852c: rfk: add RFK tables plfxlc: fix le16_to_cpu warning for beacon_interval rtw88: remove a copy of the NAPI_POLL_WEIGHT define carl9170: tx: fix an incorrect use of list iterator wil6210: use NAPI_POLL_WEIGHT for napi budget ath10k: remove a copy of the NAPI_POLL_WEIGHT define ath11k: Add support for WCN6750 device ath11k: Datapath changes to support WCN6750 ath11k: HAL changes to support WCN6750 ath11k: Add QMI changes for WCN6750 ath11k: Fetch device information via QMI for WCN6750 ath11k: Add register access logic for WCN6750 ath11k: Add HW params for WCN6750 ... ==================== Link: https://lore.kernel.org/r/20220503153622.C1671C385A4@smtp.kernel.orgSigned-off-by:
-
Jakub Kicinski authored
Most drivers should not have to worry about selecting the right weight for their NAPI instances and pass NAPI_POLL_WEIGHT. It'd be best if we didn't require the argument at all and selected the default internally. This change prepares the ground for such reshuffling, allowing for a smooth transition. The following API should remain after the next release cycle: netif_napi_add() netif_napi_add_weight() netif_napi_add_tx() netif_napi_add_tx_weight() Where the _weight() variants take an explicit weight argument. I opted for a _weight() suffix rather than a __ prefix, because we use __ in places to mean that caller needs to also issue a synchronize_net() call. Signed-off-by:
-
Vladimir Oltean authored
Test basic (port-default, VLAN PCP and IP DSCP) QoS classification for Ocelot switches. Advanced QoS classification using tc filters is covered by tc_flower_chains.sh in the same directory. Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://lore.kernel.org/r/20220502155424.4098917-1-vladimir.oltean@nxp.comSigned-off-by:
-
- 03 May, 2022 13 commits
-
-
Jakub Kicinski authored
Mat Martineau says: ==================== mptcp: Userspace path manager prerequisites This series builds upon the path manager mode selection changes merged in 4994d4fa ("Merge branch 'mptcp-path-manager-mode-selection'") to further modify the path manager code in preparation for adding the new netlink commands to announce/remove advertised addresses and create/destroy subflows of an MPTCP connection. The third and final patch series for the userspace path manager will implement those commands as discussed in https://lore.kernel.org/netdev/23ff3b49-2563-1874-fa35-3af55d3088e7@linux.intel.com/#r Patches 1, 5, and 7 remove some internal constraints on path managers (in general) without changing in-kernel PM behavior. Patch 2 adds a self test to validate MPTCP address advertisement ack behavior. Patches 3, 4, and 6 add new attributes to existing MPTCP netlink events and track internal state for populating those attributes. ==================== Link: https://lore.kernel.org/r/20220502205237.129297-1-mathew.j.martineau@linux.intel.comSigned-off-by:
-
Kishen Maloor authored
This change allows userspace PM implementations to reissue ADD_ADDR announcements (if necessary) based on their chosen policy. Acked-by:
Paolo Abeni <pabeni@redhat.com> Signed-off-by:
Kishen Maloor <kishen.maloor@intel.com> Signed-off-by:
Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by:
-
Kishen Maloor authored
This change records the 'server_side' attribute of MPTCP_EVENT_CREATED and MPTCP_EVENT_ESTABLISHED events to inform their recipient about the Client/Server role of the running MPTCP application. Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/246Acked-by:
-
Kishen Maloor authored
This change updates internal logic to permit subflows to be established from either the client or server ends of MPTCP connections. This symmetry and added flexibility may be harnessed by PM implementations running on either end in creating new subflows. The essence of this change lies in not relying on the "server_side" flag (which continues to be available if needed). Acked-by:
-
Kishen Maloor authored
Per RFC 8684, if no port is specified in an ADD_ADDR message, MPTCP SHOULD attempt to connect to the specified address on the same port as the port that is already in use by the subflow on which the ADD_ADDR signal was sent. To facilitate that, this change reflects the specific remote port in use by that subflow in MPTCP_EVENT_ANNOUNCED events. Acked-by:
-
Kishen Maloor authored
This change reads the addr id assigned to the remote endpoint of a subflow from the MP_JOIN SYN/ACK message and stores it in the related subflow context. The remote id was not being captured prior to this change, and will now provide a consistent view of remote endpoints and their ids as seen through netlink events. Acked-by:
-
Mat Martineau authored
Check userspace PM behavior to ensure ADD_ADDR echoes are only sent when there is an active userspace daemon. If the daemon is restarting or hasn't loaded yet, the missing echo will cause the peer to retransmit the ADD_ADDR - and hopefully the daemon will be ready to receive it at that later time. Reviewed-by:
Matthieu Baerts <matthieu.baerts@tessares.net> Signed-off-by:
-
Kishen Maloor authored
Current limits on the # of addresses/subflows must apply only to in-kernel PM managed sockets. Thus this change removes such restrictions on connections overseen by non-kernel (e.g. userspace) PMs. This change also ensures that the kernel does not record stats inside struct mptcp_pm_data updated along kernel code paths when exercised via non-kernel PMs. Additionally, address announcements are acknolwedged and subflow requests are honored only when it's deemed that a userspace path manager is active at the time. Acked-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linuxPaolo Abeni authored
Saeed Mahameed says: ==================== mlx5-updates-2022-05-02 1) Trivial Misc updates to mlx5 driver 2) From Mark Bloch: Flow steering, general steering refactoring/cleaning An issue with flow steering deletion flow (when creating a rule without dests) turned out to be easy to fix but during the fix some issue with the flow steering creation/deletion flows have been found. The following patch series tries to fix long standing issues with flow steering code and hopefully preventing silly future bugs. A) Fix an issue where a proper dest type wasn't assigned. B) Refactor and fix dests enums values, refactor deletion function and do proper bookkeeping of dests. C) Change mlx5_del_flow_rules() to delete rules when there are no no more rules attached associated with an FTE. D) Don't call hard coded deletion function but use the node's defined one. E) Add a WARN_ON() to catch future bugs when an FTE with dests is deleted. * tag 'mlx5-updates-2022-05-02' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux: net/mlx5: fs, an FTE should have no dests when deleted net/mlx5: fs, call the deletion function of the node net/mlx5: fs, delete the FTE when there are no rules attached to it net/mlx5: fs, do proper bookkeeping for forward destinations net/mlx5: fs, add unused destination type net/mlx5: fs, jump to exit point and don't fall through net/mlx5: fs, refactor software deletion rule net/mlx5: fs, split software and IFC flow destination definitions net/mlx5e: TC, set proper dest type net/mlx5e: Remove unused mlx5e_dcbnl_build_rep_netdev function net/mlx5e: Drop error CQE handling from the XSK RX handler net/mlx5: Print initializing field in case of timeout net/mlx5: Delete redundant default assignment of runtime devlink params net/mlx5: Remove useless kfree net/mlx5: use kvfree() for kvzalloc() in mlx5_ct_fs_smfs_matcher_create ==================== Link: https://lore.kernel.org/r/Signed-off-by: -
Paolo Abeni authored
Ido Schimmel says: ==================== mlxsw: Remove size limitations on egress descriptor buffer Petr says: Spectrum machines have two resources related to keeping packets in an internal buffer: bytes (allocated in cell-sized units) for packet payload, and descriptors, for keeping headers. Currently, mlxsw only configures the bytes part of the resource management. Spectrum switches permit a full parallel configuration for the descriptor resources, including port-pool and port-TC-pool quotas. By default, these are all configured to use pool 14, with an infinite quota. The ingress pool 14 is then infinite in size. However, egress pool 14 has finite size by default. The size is chip dependent, but always much lower than what the chip actually permits. As a result, we can easily construct workloads that exhaust the configured descriptor limit. Going forward, mlxsw will have to fix this issue properly by maintaining descriptor buffer sizes, TC bindings, and quotas that match the architecture recommendation. Short term, fix the issue by configuring the egress descriptor pool to be infinite in size as well. This will maintain the same configuration philosophy, but will unlock all chip resources to be usable. In this patchset, patch #1 first adds the "desc" field into the pool configuration register. Then in patch #2, the new field is used to configure both ingress and egress pool 14 as infinite. In patches #3 and #4, add a selftest that verifies that a large burst can be absorbed by the shared buffer. This test specifically exercises a scenario where descriptor buffer is the limiting factor and the test fails without the above patches. ==================== Link: https://lore.kernel.org/r/20220502084926.365268-1-idosch@nvidia.comSigned-off-by:
-
Petr Machata authored
Add a test that sends 1Gbps of traffic through the switch, into which it then injects a burst of traffic and tests that there are no drops. Signed-off-by:
Petr Machata <petrm@nvidia.com> Signed-off-by:
Ido Schimmel <idosch@nvidia.com> Signed-off-by:
-
Petr Machata authored
Add two helpers, start_traffic_pktsize() and start_tcp_traffic_pktsize(), that allow explicit overriding of packet size. Change start_traffic() and start_tcp_traffic() to dispatch through these helpers with the default packet size. Signed-off-by:
-
Petr Machata authored
Spectrum machines have two resources related to keeping packets in an internal buffer: bytes (allocated in cell-sized units) for packet payload, and descriptors, for keeping metadata. Currently, mlxsw only configures the bytes part of the resource management. Spectrum switches permit a full parallel configuration for the descriptor resources, including port-pool and port-TC-pool quotas. By default, these are all configured to use pool 14, with an infinite quota. The ingress pool 14 is then infinite in size. However, egress pool 14 has finite size by default. The size is chip dependent, but always much lower than what the chip actually permits. As a result, we can easily construct workloads that exhaust the configured descriptor limit. Fix the issue by configuring the egress descriptor pool to be infinite in size as well. This will maintain the configuration philosophy of the default configuration, but will unlock all chip resources to be usable. In the code, include both the configuration of ingress and ingress, mostly for clarity. Signed-off-by:
-
