- 01 Oct, 2020 40 commits
-
-
Mordechay Goodstein authored
Driver doesn't support fw without wide_cmd so driver always sets it to true. instead of setting it always to true just remove the field. Signed-off-by:
Mordechay Goodstein <mordechay.goodstein@intel.com> Signed-off-by:
Luca Coelho <luciano.coelho@intel.com> Link: https://lore.kernel.org/r/iwlwifi.20200928121852.e6137861d917.I93405604eb503568688b28d3169fea7fbb88ed7e@changeidSigned-off-by:
-
Mordechay Goodstein authored
For new APIs this avoids checking every return if it's IWL_FW_CMD_VER_UNKNOWN (99) or it's lower than the new API value Done with spatch: -iwl_fw_lookup_cmd_ver(E1, E2, E3) +iwl_fw_lookup_cmd_ver(E1, E2, E3, IWL_FW_CMD_VER_UNKNOWN) Signed-off-by:
-
Mordechay Goodstein authored
In non acpi compilations iwl_sar_geo_init isn't called which results with compilation warning so just remove the declaration. Signed-off-by:
-
Mordechay Goodstein authored
This helps with congested environments reducing the conflict cost to RTS retries only, instead of the entire BA packet. Signed-off-by:
-
Nathan Errera authored
Add a new version to iwl_wowlan_kek_kck_material_cmd struct to support akm and send the new cmd if supported. Signed-off-by:
Nathan Errera <nathan.errera@intel.com> Signed-off-by:
-
Luca Coelho authored
Add support for the new version of the GEO_TX_POWER_LIMIT command. This new version includes UHB values in the table, but for now, since we don't have the ACPI values yet, we support the API, but don't set the extra values. Signed-off-by:
-
Luca Coelho authored
Add the version number to the iwl_geo_tx_power_profile_cmd structs and move the union into a common place. Additionally, reuse the code that sets elements that are at the same place in the struct across different versions. While at it remove an unused variable, add a comment and move the v2 specific element setting to inside the if statement. Additionally, invert the if for slightly more readability. Signed-off-by:
-
Luca Coelho authored
The new version of the command can support more subbands and CDB, so it can contain more data than earlier versions. Implement support for the new version of the command, even though we don't have more data to write to it yet. Signed-off-by:
-
Luca Coelho authored
The SAR profile tables will be larger in the next version, so prepare the iwl_sar_select_profile() function to handle multiple sizes and update the relevant callers. Signed-off-by:
-
Luca Coelho authored
Create a common structure to contain all different versions of the tx_power_cmd instead of making a union of the different structs everywhere we need them. Also move the common part of these structs into a separate structure (instead of reusing v3) and leave the per_chain_restriction part out of the common part, because this will change in version 6 of the command (which will be added soon). While at it, rename per_chain_restriction to per_chain to shorten it. Signed-off-by:
-
Roee Goldfiner authored
added ID for ma in AX family. Signed-off-by:
Roee Goldfiner <roee.h.goldfiner@intel.com> Signed-off-by:
-
Luca Coelho authored
This function is only called from acpi.c, which is only included in the makefile if CONFIG_ACPI is set. So it doesn't make sense to declare a dummy version of it when CONFIG_ACPI is not defined. Signed-off-by:
-
Luca Coelho authored
Only iwlmvm uses this function and it's so simple that it's clearer if it's spelled out in the code anyway, so remove it and add the check where needed. Signed-off-by:
-
Johannes Berg authored
On newer hardware, we have the full checksum, so use it to report CHECKSUM_COMPLETE and avoid the protocol specific hardware parsing. Note that the hardware already parses/removes the SNAP header, so we actually literally get what we need to report to the stack, as we're expected to checksum everything after the L2 header (which is translated/added by mac80211). Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
-
Mordechay Goodstein authored
Currently we have the same info in two variables, If umac_error_event_table is 0, we know that UMAC log is not supported, so we don't need the support_umac_log field. Signed-off-by:
-
Dan Halperin authored
- Change the iwl_all_tsc_rsc struct to hold a sta_id (__le32) field, while preserving the union, used in the older version. - Adjust the use of this command according to the TLV. Signed-off-by:
Dan Halperin <Dan1.Halperin@intel.com> Signed-off-by:
-
Luca Coelho authored
A couple of functions were exported from the acpi.c file unnecessarily, since they are only used internally in that file. Make them static. Signed-off-by:
-
Naftali Goldstein authored
The the driver prevents new Tx from being sent during the remove-station flow is by invalidating the fw_id_to_mac_id rcu of that station. However, if there was any Tx still in-flight (tx-cmd was sent but the ba-notif wasn't received yet) the ba-response to those frames is simply ignored without actually reclaiming anything. This later causes the driver to think that that some of the station's queues aren't empty when in practice they are which causes errors in the station remove flow. Fix this by performing the tx-reclaim also if the rcu is invalid. any DB that can't be updated due to this is not very important at this stage since the station is about to be removed soon anyways. Signed-off-by:
Naftali Goldstein <naftali.goldstein@intel.com> Signed-off-by:
-
Dan Halperin authored
Add a sta_id (__le32) to the cmd, in order to support CDB protocol. Signed-off-by:
-
Andrei Otcheretianski authored
Due to a missing break, the management multicast key was installed even though we don't really support it. Fix that, so mac80211 would know that it should protect frames in software. Signed-off-by:
Andrei Otcheretianski <andrei.otcheretianski@intel.com> Signed-off-by:
-
Emmanuel Grumbach authored
A print in the remain on channel code was too long and caused a WARNING, split it. Signed-off-by:
Emmanuel Grumbach <emmanuel.grumbach@intel.com> Fixes: dc28e12f ("iwlwifi: mvm: ROC: Extend the ROC max delay duration & limit ROC duration") Signed-off-by:
-
Mordechay Goodstein authored
dev_cmd_offs, page_offs field is not directly related to the PCIe transport, so move the structures it uses to the common iwl-trans.h header. Signed-off-by:
-
Gil Adam authored
Support v2 of regulatory capability flags parsed from the device NVM. New API support is determined by FW lookup of the MCC update command resposnse version, where version 6 supports the new API. Signed-off-by:
Gil Adam <gil.adam@intel.com> Signed-off-by:
-
Mordechay Goodstein authored
With the new region we can handle in user space understanding the struct type and version and driver doesn't need to be involved at all. Signed-off-by:
-
Luca Coelho authored
This timepoint was removed from the FW, so we remove it from here too for consistency. It's now marked as reserved to preserve the enumeration. Signed-off-by:
-
Gil Adam authored
Version 2 of the PPAG table command supports more sub-bands than previous. Change relevant command structs and the reading of the ACPI tables. Signed-off-by:
-
Ihab Zhaika authored
add few PCI ID'S for ma with gf and mr in AX family. Signed-off-by:
Ihab Zhaika <ihab.zhaika@intel.com> Signed-off-by:
-
Mordechay Goodstein authored
The bc_table_dword code is not directly related to the PCIe transport, so move the structures it uses to the common iwl-trans.h header. Signed-off-by:
-
Mordechay Goodstein authored
The tfd_* code is not directly related to the PCIe transport, so move the structures it uses to the common iwl-trans.h header. Signed-off-by:
-
Mordechay Goodstein authored
The bc_pool code is not directly related to the PCIe transport, so move the structures it uses to the common iwl-trans.h header. Signed-off-by:
-
Mordechay Goodstein authored
txcmd fields is not directly related to the PCIe transport, so move to the common iwl_trans_alloc function. Signed-off-by:
-
Golan Ben Ami authored
TWT has been disabled during stabilization and further testing of the FW code. It is now considered stable, so TWT can be enabled in the driver. Signed-off-by:
Golan Ben Ami <golan.ben.ami@intel.com> Signed-off-by:
-
Avraham Stern authored
A FTM responder may do PASN authentication with unassociated stations to allow secure ranging. In this case, the driver will add an internal station and install the TK so the FW will accept protected FTM request frames from this station and will send a protected FTM response frame. In addition, the driver needs to configure the HLTK to the FW so the FW can derive the secure LTF bits. This is left for a later patch since it is not yet supported by the FW. Signed-off-by:
Avraham Stern <avraham.stern@intel.com> Signed-off-by:
-
Ihab Zhaika authored
add few PCI ID'S for AX201 family including killer cards. Signed-off-by:
-
Mordechay Goodstein authored
Move this parser to a common place, instead of having it in code that is specific to mvm. Signed-off-by:
-
Mordechay Goodstein authored
In version 10 we actually go back to the struct size before 9 so only for version 9 we need the size fixup Signed-off-by:
-
Gil Adam authored
Evaluate the appropriate DSM from ACPI to set ETSI SRD 5.8GHz channels to passive or disabled, default behaviour is enabled. Add enums and refactor evaluation of DSM functions for better readablity and more informative debug prints. Signed-off-by:
-
Mordechay Goodstein authored
There is an issue in the HW DMA engine in the 9000 family of devices when more than 6 RX queues are used. The issue is that the FW may hang when IWL_MVM_RXQ_NSSN_SYNC notifications are sent. Fix this by limiting the number of RX queues to 6 in the 9000 family of devices. Signed-off-by:
-
Mordechay Goodstein authored
This avoids cases that we have a lot of packets in the queue and until we are able to transmit SSN packet the other side sends DELBA due to timeout. Signed-off-by:
-
Shaul Triebitz authored
Signed-off-by:
Shaul Triebitz <shaul.triebitz@intel.com> Signed-off-by:
-
