- 06 Jul, 2016 20 commits
-
-
Sara Sharon authored
For a000 device the FH was replaced by the TFH. This is the first patch in a series introducing the changes stemming from this change. This patch initializes the TFQ queue table with the new 64 bit register and the relevant TFH configuration registers. Signed-off-by:
Sara Sharon <sara.sharon@intel.com> Signed-off-by:
Luca Coelho <luciano.coelho@intel.com>
-
Sara Sharon authored
Move the write_prph_64 of pcie to be transport agnostic. Add direct write as well, as it is needed for a000 HW. Signed-off-by:
-
Sara Sharon authored
Currently the scratch buffer is set to 16 bytes and indicates the size of the bi-directional DMA. However, next HW generation will perform additional offloading, and will write the result in the key location of the TX command, so the size of the bi-directional consistent memory should grow accordingly - increase it to 40. Generalize the code to get rid of now irrelevant scratch references. Signed-off-by:
-
Sara Sharon authored
In MQ environment and new architecture in early stages we may encounter DMA issues. Track RXB status and bail out in case we receive index to an RXB that was not mapped and handed over to HW. Signed-off-by:
-
Emmanuel Grumbach authored
Upon firmware load interrupt (FH_TX), the ISR re-enables the firmware load interrupt only to avoid races with other flows as described in the commit below. When the firmware is completely loaded, the thread that is loading the firmware will enable all the interrupts to make sure that the driver gets the ALIVE interrupt. The problem with that is that the thread that is loading the firmware is actually racing against the ISR and we can get to the following situation: CPU0 CPU1 iwl_pcie_load_given_ucode ... iwl_pcie_load_firmware_chunk wait_for_interrupt <interrupt> ISR handles CSR_INT_BIT_FH_TX ISR wakes up the thread on CPU0 /* enable all the interrupts * to get the ALIVE interrupt */ iwl_enable_interrupts ISR re-enables CSR_INT_BIT_FH_TX only /* start the firmware */ iwl_write32(trans, CSR_RESET, 0); BUG! ALIVE interrupt will never arrive since it has been masked by CPU1. In order to fix that, change the ISR to first check if STATUS_INT_ENABLED is set. If so, re-enable all the interrupts. If STATUS_INT_ENABLED is clear, then we can check what specific interrupt happened and re-enable only that specific interrupt (RFKILL or FH_TX). All the credit for the analysis goes to Kirtika who did the actual debugging work. Cc: <stable@vger.kernel.org> [4.5+] Fixes: a6bd005f ("iwlwifi: pcie: fix RF-Kill vs. firmware load race") Signed-off-by:
-
Johannes Berg authored
The PCIe transport needs to store two pointers in each TX SKB, and currently assumes mac80211's ieee80211_tx_info is present in the CB to do that. In order to remove that assumption, have the opmodes pass in the offset to where the pointers can be stored in the CB and use the offset in the PCIe code. To make the disentanglement complete, remove mac80211.h includes from everywhere in the generic iwlwifi code. This required adding an include of cfg80211.h in one place. Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
-
Johannes Berg authored
In order to be able to properly record SKBs that didn't come through mac80211, don't rely on the IEEE80211_TX_CTRL_PORT_CTRL_PROTO flag but instead check for ETH_P_PAE directly. Signed-off-by:
-
Johannes Berg authored
In order to reduce reliance on mac80211 structs in the core iwlwifi code, store the cipher schemes in the format given by the firmware and convert it later, rather than storing it in the mac80211 format. Signed-off-by:
-
Sara Sharon authored
Restart flow zeroes the rx_ba_sessions counter. Mac80211 asks driver to tear down of the session only afterwards, and as a result driver didn't free the data. Fix it. Signed-off-by:
-
Sara Sharon authored
Add some prints to track BAID assignment. Signed-off-by:
-
Liad Kaufman authored
Make sure that in DQA mode, the SCD's configuration of a queue is redirected to the lower AC of the streams of the queue. Make sure that this queue is redirected to the lowest AC when adding a new RA/TID to an existing queue. If it isn't - redirect the queue. Also, as redirection revealed a bug in the marking of a shared queue, this patch contains a small fix to make sure a shared queue maintains the appropriate "shared queue marking". Signed-off-by:
Liad Kaufman <liad.kaufman@intel.com> Signed-off-by:
-
Liad Kaufman authored
Support TDLS when working in DQA mode. This is done mainly by NOT doing any special things for TDLS, as the queues are dynamically created anyway, so no need to allocate them ahead of time. Signed-off-by:
-
Sara Sharon authored
In cases of hardware or DMA error, the vid read from a zeroed location will be 0, and we will access the rxb at index 0 in the global table, while it may be NULL or owned by hardware. Invalidate vid 0 in order to detect the situation and bail out. Signed-off-by:
-
Emmanuel Grumbach authored
This led to a DMA splat. Fixes: a6c4fb44 ("iwlwifi: mvm: Add FW paging mechanism for the UMAC on PCI") Signed-off-by:
Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by:
-
Johannes Berg authored
This struct member is never set, so remove it. Since this is the last thing that needs mac80211.h, also change the includes to no longer use mac80211.h Signed-off-by:
-
Emmanuel Grumbach authored
The value for Channel 14 was wrong. Fix it. Signed-off-by:
-
Emmanuel Grumbach authored
Those comments were wrong, fix them. Signed-off-by:
-
Golan Ben-Ami authored
The TX fifos are arranged consecutively in the SMEM, beginning with the regular fifos, and tailed by the internal fifos. In the current code, while trying to read the internal fifos, we read the fifos beginning with the index zero. By doing this we actually re-read the regular fifos. In order to read the internal fifos, start the reading index from the number of regular fifos configured by the fw. Signed-off-by:
Golan Ben-Ami <golan.ben.ami@intel.com> Fixes: 39654cb3 ("iwlwifi: don't access a nonexistent register upon assert") Signed-off-by:
-
Sara Sharon authored
This fixes a pretty ancient bug that hasn't manifested itself until now. The scratchbuf for command queue is allocated only for 32 slots but is accessed with the queue write pointer - which can be up to 256. Since the scratch buf size was 16 and there are up to 256 TFDs we never passed a page boundary when accessing the scratch buffer, but when attempting to increase the size of the scratch buffer a panic was quick to follow when trying to access the address resulted in a page boundary. Signed-off-by:
-
Ido Yariv authored
On some of the chipsets MSI & INTA interrupts are disabled by default in the HW registers, and need to be explicitly enabled to be used. In case MSI-X isn't used, make sure MSI mode is enabled by setting the relevant HW register. Signed-off-by:
Ido Yariv <idox.yariv@intel.com> Signed-off-by:
-
- 05 Jul, 2016 20 commits
-
-
Luca Coelho authored
Read the SAR BIOS table from the ACPI and parse it into the iwl_mvm_sar_table structure. If the table is enabled, send it to the firmware via REDUCE_TX_POWER_CMD. Signed-off-by:
-
Emmanuel Grumbach authored
We removed support for old API for coexistence, but we forgot to remove defines and variable that are not needed anymore. Signed-off-by:
-
Sara Sharon authored
Our HW does not support checksum of fragmented packets. Fix code accordingly to checksum those packets in the driver. Signed-off-by:
-
striebit authored
Currently the state sent in SF configuration is always FULL_ON. This commit sets the correct state (e.g. INIT_OFF when station is not associated). Fixes: commit f4a3ee49 ("iwlwifi: mvm: Always enable the smart FIFO") Signed-off-by:
Shaul Triebitz <shaul.triebitz@intel.com> Signed-off-by:
-
Emmanuel Grumbach authored
The NIC's CPU gets started after the firmware has been written to its memory. The first thing it does is to send an interrupt to let the driver know that it is running. In order to get that interrupt, the driver needs to make sure it is not masked. Of course, the interrupt needs to be enabled in the driver before the CPU starts to run. I mistakenly inversed those two steps leading to races which prevented the driver from getting the alive interrupt from the firmware. Fix that. Cc: <stable@vger.kernel.org> [4.5+] Fixes: a6bd005f ("iwlwifi: pcie: fix RF-Kill vs. firmware load race") Signed-off-by:
-
Haim Dreyfuss authored
Add a000 family configuration to iwl-cfg struct Signed-off-by:
Haim Dreyfuss <haim.dreyfuss@intel.com> Signed-off-by:
-
Sara Sharon authored
For 9000 family we will get extended statistics notification with averaged data for RSSI, TCM and rogue AP detection. Support it. Future patches will added the required algorithms. Signed-off-by:
-
Ayala Beker authored
Newer hardware supports GCMP and GCMP 256-bit ciphers. Add support for adding/setting GCMP key for TX mode. In the TX command handling GCMP-256 is handled in a different way as the key size should be up to 128-bits: Set the key value to the key index in the key table, and specify that this key should be taken form the key table instead of from the TX command. While at it - convert security control flags to an enum. Signed-off-by:
Ayala Beker <ayala.beker@intel.com> Signed-off-by:
-
Golan Ben-Ami authored
Reserve a single iwl_fw_error_dump_type enum for external code utilities. Signed-off-by:
-
Sara Sharon authored
Add support of dumping new RFH instead of FH registers. Signed-off-by:
-
Guenter Roeck authored
gcc-6 reports the following error if -Werror=unused-const-variable is enabled. drivers/net/wireless/intel/iwlwifi/dvm/lib.c:210:21: error: 'iwlagn_loose_lookup' defined but not used Signed-off-by:
Guenter Roeck <linux@roeck-us.net> Signed-off-by:
-
Oren Givon authored
Fix an issue where nullfunc frames and block ack requests had the same tid as aggregation frames and were queued on a non aggregation queue. The pending frames counter included those frames but the check whether to decrement the pending frames counter relied on the tid status and not on the txq id. The result was an inconsistent state of the pending frames counter followed by a failure to remove the station. This failure triggered SYSASSERT 0x3421. In addition, fix a situation in DQA mode where the number of pending frames turned negative. This was due to the TX queue being on the IWL_EMPTYING_HW_QUEUE_DELBA state and its frames were still decremented. Even though the SYSASSERT issue is fixed when DQA is disabled, the issue is not completely solved when DQA is enabled and should still be fixed. Signed-off-by:
Oren Givon <oren.givon@intel.com> Fixes: cf961e16 ("iwlwifi: mvm: support dqa-mode agg on non-shared queue") Signed-off-by:
-
Arnd Bergmann authored
gcc is apparently unablel to track the state of the local 'resp_v2' variable across the kzalloc() function, and warns about the response variable being used without an initialization: drivers/net/wireless/intel/iwlwifi/mvm/nvm.c: In function ‘iwl_mvm_update_mcc’: drivers/net/wireless/intel/iwlwifi/mvm/nvm.c:727:36: warning: ‘mcc_resp_v1’ may be used uninitialized in this function [-Wmaybe-uninitialized] resp_cp->n_channels = mcc_resp_v1->n_channels; drivers/net/wireless/intel/iwlwifi/mvm/nvm.c:721:3: warning: ‘mcc_resp’ may be used uninitialized in this function [-Wmaybe-uninitialized] memcpy(resp_cp, mcc_resp, resp_len); The warning showed up in x86 allmodconfig after my patch to unhide -Wmaybe-uninitialized warnings by default was merged, though it always existed in randconfig builds. I did not catch the warning earlier because I was testing on ARM, which never produced the warning. This rearranges the code in a way that improves readability for both humans and the compiler, and that avoids the warning. Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Fixes: 6fa52430 ("iwlwifi: mvm: change mcc update API") Signed-off-by:
-
Gregory Greenman authored
Expand TLC to support 160MHz channels. Full support for A-MSDU case will be added separately. Signed-off-by:
Gregory Greenman <gregory.greenman@intel.com> Signed-off-by:
-
Ayala Beker authored
FW sets status for each RX packet. Enum in the driver doesn't match with FW definition - fix it. Signed-off-by:
-
Matti Gottlieb authored
Currently we try to open an aggregation for every packet (given that one is not already open). This causes redundant overhead (addba/delba) for null data packets. Do not open an aggregation for null data packets. Signed-off-by:
Matti Gottlieb <matti.gottlieb@intel.com> Signed-off-by:
-
Sara Sharon authored
Somehow we ended up stopping RX using legacy RX registers even for devices that support RFH. Fix it. Signed-off-by:
-
Luca Coelho authored
We convert the mvm device to a PCI device and then back again when trying to find the handle for the device's ACPI data. This is unnecessary, so it can be removed. Signed-off-by: -
Oren Givon authored
Add 6 new 8265 series PCI IDs: - (0x24FD, 0x1130) - (0x24FD, 0x0130) - (0x24FD, 0x0910) - (0x24FD, 0x0930) - (0x24FD, 0x0950) - (0x24FD, 0x0850) CC: <stable@vger.kernel.org> [4.6+] Signed-off-by:
David Spinadel <david.spinadel@intel.com> Signed-off-by:
-
Oren Givon authored
Add 3 new 8260 series PCI IDs: - (0x24F3, 0x10B0) - (0x24F3, 0xD0B0) - (0x24F3, 0xB0B0) CC: <stable@vger.kernel.org> [4.1+] Signed-off-by:
-
