summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
...
| | * | | | Replace work queue related functions with wrapper APIs.vinothkumar2014-01-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Replace the work queue related functions with vos wrapper APIs and handle for both CONFIG_CNSS and WLAN_OPEN_SOURCE Change-Id: I617191bb4ba952a8bfc2158d6d5a1babb346e7bc CRs-Fixed: 583276
| | * | | | Delete vdev up from vdev restart response handler.Rakesh Sunki2014-01-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In SAP mode issuing vdev up command prematurely in vdev restart response handler, causing beacons to send out with stale channel IE's. Fix to start beacon transmission and do vdev up with new channel IE's on a Non-DFS channel after channel change . Also, set the vdev_up flag in wma interface to FALSE, after vdev restart which will allow vdev up to happen on the new channel. . CRs-Fixed: 599208 Change-Id: Ia61142ffee1dd66949c1e9a7528c7f02f4c6799a
| | * | | | Fix the CAC timeout value for ETSI reg domainKiran Kumar Lokere2014-01-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix the CAC timeout value for ETSI domain wether channels. Change-Id: I07529f2b8c463bb7e55c266d6a2117f8fcacb87c CRs-Fixed: 600104
| | * | | | qcacld/txrx: Process TKIP MIC failure Rx indicationVasanthakumar Thiagarajan2014-01-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pass the necessary information for MIC failure indication from the erroneous msdu to UMAC in SIR_HAL_MIC_FAILURE_IND. This patch extends ol_rx_err by adding keyid and TKIP sequence number. Also defines an API to access key index from Rx descriptor. . Change-Id: Ic1895a92537d073b20e9d5c471d3c00b0eae3223 CRs-Fixed: 584815
| | * | | | wlan: bap: fix unsafe use of assertPraveen Kumar Sirisilla2014-01-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The CR identifies improper use of vos assert. At some places NULL pointers are asserted but no action is taken. Patch fix such issues. . Change-Id: I8a278fe019948630f629d8e76abd8f262ab5aa2c CRs-Fixed: 589661
| | * | | | qcacld: update fw common interface files as per CL 814912Rajesh Chauhan2014-01-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add define WMI_VDEV_PARAM_MAX_RATE in wmi header file. . Change-Id: Ib01f33d52195af045df3b2cfc866c1bcbf80c662 CRs-Fixed: 599151
| * | | | | Merge "Merge remote-tracking branch 'origin/caf/caf-wlan/master'"Linux Build Service Account2014-01-25
| |\ \ \ \ \
| | * | | | | Merge remote-tracking branch 'origin/caf/caf-wlan/master'Prakash Dhavali2014-01-24
| |/| | | | | | | |/ / / /
| | * | | | Clean un-needed content and match master copyPrakash Dhavali2014-01-23
| | | | | |
| | * | | | Release 4.0.0.72 FinalPrakash Dhavali2014-01-23
| | | | | |
| | * | | | qcacld: Use correct values for power limitsAmar Singhal2014-01-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In function create_linux_regulatory_entry, the power values sent by the kernel already are in DBM. Therefore, remove the division by 100. Also remove the unused function create_linux_regulatory_entry_from_regd. Change-Id: I658ea9ff78c0d3aea4efd224b11ff9f217871190 CRs-Fixed: 597106
| | * | | | qcacld: hdd: fix logging format/argument mismatchesJeff Johnson2014-01-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Source code analysis uncovered mismatches between format and argument in invocations of logging APIs in the hdd code, so fix them. Change-Id: I97b98163136e55ef6f56c0ea5ea6d7ec12ba1821 CRs-fixed: 570511
| | * | | | Revert "qcacld: Use correct values for power limits"Prakash Dhavali2014-01-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This reverts commit c24a3d08b5cec6c4d3cf9e2a97ab7e348accbee3 Issue with git patching seems to have resulted in change id mismatch. Revert this to resubmit a new change to match https://review-android.quicinc.com/#/c/577609/ Change-Id: Ia17a9012da01f49cd05de8a41f03280b765e2826
| | * | | | Change LF copyright year to match othersPrakash Dhavali2014-01-18
| | | | | |
| | * | | | Release 4.0.0.72 InitialPrakash Dhavali2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | Release 4.0.0.72 Initial
| | * | | | Remove left-over trailing internal flags in endifPrakash Dhavali2014-01-18
| | | | | |
| | * | | | qcacld: Use correct values for power limitsAmar Singhal2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In function create_linux_regulatory_entry, the power values sent by the kernel already are in DBM. Therefore, remove the division by 100. Also remove the unused function create_linux_regulatory_entry_from_regd. Change-Id: I8a4de075553e440c2304ac8d7ae65ce62081ae76 CRs-Fixed: 597106
| | * | | | wlan: pmc: fix additional logging format/argument mismatchJeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Patch "CLD-P2P: Pass the correct session id for p2p session" introduced a mismatch between argument and format string in a logging API, so fix the format to match the argument. Change-Id: I189b05bb7670a0b4acee0281b60702095c4ff2e3 CRs-fixed: 570511
| | * | | | cld: restrict the userspace from adding more vdevs than allowed by the HostChandrasekaran, Manishekar2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The Host by default configures the number of vdevs in the target.For Rome, this default value is 3 (5 for HL). In the issue scenario, the Host tries to configure more than 3 vdevs. This results in assert. Changes are done to restrict the userspace from adding more than the configured number of vdevs Change-Id: I91d7e54d0d86b59703de57c34fc6ff5c2d2d509c CRs-Fixed: 598220
| | * | | | qcacld:ipa: WLAN-IPA TX OptimizationYun Park2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Removing dev-id and sta_id from the Tx partial header - WLAN-IPA TX-optimization (Removing DMA address map/unmap) - Fix IPA RX header format from 802.3 to Ethernet-II - Fix ipa_tx_dp_mul() failure by making total # of desc 1 less than the IPA descriptor FIFO size - Fix the crash due to invaid sta_id due to IPA Rx header is not retained when using IPA default routing table CRs-fixed: 595907
| | * | | | Correct copyright year to match it was beforePrakash Dhavali2014-01-18
| | | | | |
| | * | | | wlan: Fix 64-bit type conversion issue in vos_crypto_init()Jeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When built for a 64-bit architecture vos_crypto_init() generates a "pointer-to-int-cast" warning. The cause is that a 64-bit pointer is being assigned to a 32-bit crypto handle. Note that if the crypto handle was actually being used by the crypto routines, the appropriate fix would be to change the type of the crypto handle from (v_U32_t) to (void *) so that it could represent an opaque memory pointer. But since such a change would require modification to all of the VOSS crypto APIs as well as changes to all clients of these APIs, and since in reality the crypto handle is not actually being used, the less intrusive change is to simply modify vos_crypto_init() to set the crypto context to 0 instead of NULL. Change-Id: I73a9c6f0f0361eec1bf49d14a1cf6f109acd3432 CRs-fixed: 567258
| | * | | | wlan: Add adapter MAGIC checking in *_tx_fetch_packet_cbk()Jeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A kernel panic in _list_del_entry() was observed when SoftAP was disabled. Root cause is that when the SoftAP function was disabled and the pAdapter was freed, the TX Thread was blocked. When the TX Thread subsequently unblocked, a stale pAdapter was dereferenced. Add adapter MAGIC checking to the *_tx_fetch_packet_cbk() functions to prevent dereference of a stale pAdapter. Change-Id: I7d7436bf0433e86833230b8bfa620a8fa9c7f538 CRs-fixed: 580814
| | * | | | wlan: csr: Fix logging format/argument mismatchJeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The change "wlan: Select best RSSI with SSID & Sec match for 5G Preference" introduced a mismatch between format and argument in one instance of an invocation of a logging API, so correct the mismatch. Change-Id: Ida2cccd9421b2d694dbd9a96e2f89dbd1d0fb42e CRs-fixed: 570511
| | * | | | qcacld: SAP not working in 2.4G VHT modeMushtaq A Mujale2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Configure proper mode for self vdev in case of vht. Change-Id: I2c3d705ff3fe9ccb803ce2d0882533eb9277a891 CRs-Fixed:589609
| | * | | | qcacld: Disable asynchronous suspend capability during pci probeSundar Subramaniyan2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When WLAN driver refuses to suspend, the PCIe bus driver should not suspend as the WLAN driver might access target address space or the target might be accessing APSS DDR. If the bus driver still suspends when the endpoint is accessing target space over the momentarily broken PCIe link it could cause L2 errors in host or AXI bus error in target. Also, due to this incorrect suspend sequence, after the ongoing suspend process is unwinded by PM, any target register access becomes invalid as target returns unsupported request (-1) after D3 handshake is completed by now. Here the target may refuse to suspend for multiple reasons including but not limited to host DDR access by not sending the HTC endpoint acknowledgement for the SUSPEND_TARGET or WOW_ENABLE WMI commands. The issue is that asynchronous suspend capability flag is getting set by default when pci device is probed. If this flag is set runtime PM will schedule the device suspend function to asynchronously run independent of parent device. If the flag is not set, runtime PM will do suspend in order one by one and maintains any parent-child relationship. To fix this, disable the asynchronous suspend during pci probe, so that the runtime PM perform suspend of PCI endpoint and PCI bus drivers serially and if endpoint fails to suspend the return value is checked immediately and aborts the suspend and starts recovering the suspended devices. Change-Id: Ia88448094209048c4bb52af30c3404e53de5a2a1 CRs-Fixed: 594454
| | * | | | qcacld: Use copy_from_user() in iw_setchar_getnone()Raja Mani2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The private ioctl and sub ioctl registered from the driver using WLAN_PRIV_SET_CHAR_GET_NONE expose only SET attributes and sets nothing in GET attributes. iwpriv ioctl code used for WLAN_PRIV_SET_CHAR_GET_NONE is not in even number (SIOCIWFIRSTPRIV + 3). So, WEXT layer in kernel will not call copy_from_user() and provide local copy in kernel space. iw_setchar_getnone() must take care of calling copy_from_user() and making local copy before using buffer which comes from the user space. Change-Id: Ie6347b2a3f84e4fc34b41fe53083f5162516dd7c CRs-Fixed: 594921
| | * | | | P2P: Check remaining RoC only in case of GO_NEG_CNF.Nirav Shah2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Now check remaining RoC only in case of GO_NEG_CNF. and requesting new RoC for all other Resp type action packets. CRs-Fixed: 596484 Change-Id: If83d03fb0738de02f7990293d5a7e96d5464167f
| | * | | | CLD_P2P: Improve the logging for beacon miss indicationGanesh Kondabattini2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Increasing the log level for beacon indication debug logs, so that these logs will be printed always. This will make easy to debug the disconnection issues in P2P, SCC and MCC cases. Change-Id: I318d4c6cf57ae5fd7b55aeece1ec7f1b97bff0c6 CRs-Fixed: 592966
| | * | | | DroidSec: Multiple issues in file csrApiRoam.cPraveen Kumar Sirisilla2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In file csrApiRoam.c, identified lines of code which probably can cause kernel panic,memory leaks,etc.This patch fixes these lines of code. Change-Id: I5ac62a0f5f34555324555f5e4e3da315b8e04dfb CRs-Fixed: 569380
| | * | | | Enable CCX and CCX uploadNaveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | CCX right now is optionally compiled with a flag change in Kbuild. This gerrit will enable CCX Change-Id: Ib1d2fcd90b60f94d33f26254f3f9112c6bdf8721 CRs-Fixed: 596020
| | * | | | wlan: btamp: fix logging format/argument mismatchJeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Source code analysis uncovered a mismatch between format and argument in an invocation of VOS_TRACE() in the btamp code, so fix it. Change-Id: I428fb3bd9d7f72dd59663a845853ace8e4f5d59e CRs-fixed: 570328
| | * | | | wlan: svc: fix logging format/argument mismatchJeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Source code analysis uncovered mismatches between format and argument in invocations of VOS_TRACE() in the svc code, so fix them. Change-Id: Ia84f7d288fbd1a711d9628c81c29b14b5879b47e CRs-fixed: 570511
| | * | | | wlan: increase MAX_CFG_INI_ITEMS to 512Jeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently when the driver loads it logs an error that MAX_CFG_INI_ITEMS is too small. So increase the value to compensate for the configuration items that have been recently added and to allow some future growth as well. Change-Id: Ib3c7d117ee489eaa1843c3fc4b9f68c2f8d758aa CRs-fixed: 591510
| | * | | | wlan: wda: fix logging format/argument mismatchJeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Source code analysis uncovered a mismatch between format and argument in an invocation of VOS_TRACE() in the wda code, so fix it. Change-Id: I6ebed2c3011fece30db88bdac272755462c15bfd CRs-fixed: 570511
| | * | | | CLD-P2P: Use additional IE from scan_info structGanesh Kondabattini2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | After 8-way handshake supplicant should give the scan command in that it update the additional IEs, But because of scan enhancements, the supplicant is not issuing the scan command now. So the unicast frames which are sent from the host are not having the additional IEs. If it is P2P CLIENT and there is no additional IE present in roamProfile, then use the addtional IE form scan_info. Change-Id: I935622ddf668f85309dd0f27532b8a2ea4b3d142 CRs-Fixed: 596322
| | * | | | qcacld: fix for SENDACTIONFRAME dwell time checkNaveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | in SENDACTIONFRAME dwell time can be zero. Due to wrong sanity check correct SENDACTIONFRAME command was failing Change-Id: Idb4d242a76c48043ac0dcd0ca85da5eb7fce96d6 CRs-Fixed: 597551
| | * | | | qcacld: Add knob to disable DFS channel switchKiran Kumar Lokere2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Disable channel switch using config parametet to support CST tests. When enabled, SAP will remain on the same channel even after radar is found. Change-Id: Ib5ecfdea07f0a4f3832b85dcde0ef4248bfd1828 CRs-Fixed: 598216
| | * | | | wlan: qcacld-new: update dpu feedback macroSrinivas Girigowda2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change got missed in the original PMF change Change-Id: I6696e3cb755339b3be18e3bc3425f669dfa338c7 CRs-Fixed: 590924
| | * | | | qcacld: DroidSec string validation issuesNaveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 1. the total_len is a user controlled parameter, an attacker could pass total_len as a huge number and driver may result in heap based overflow. Fix is to add MAX_LEN check and the value of 4096 is based on wpa_supplicant max data 2. while reading strings using sscanf buffers are allocated with 32 bytes and sscanf %32s could write 32 bytes and automatically add a NUL termination. which could result in buffer overflow. Fix is to read only 31 bytes and keep 1 byte for '\0' Change-Id: I91ecb416e0e5ce6300a8c08a977032cfd0116a1c CRs-Fixed: 598730
| | * | | | Enhancement for SCANHOMEAWAYTIME, SETROAMSCANCHANNELS and SETBANDNaveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Retain the values of SCANHOMEAWAYTIME and ROAMSCANCHANNELS If the input does not meet certain criteria, then modify it only while sending down the parameters to the firmware and retain the original values in the global configuration parameters.The retained values would be again used in future for any further runtime config operations. Change-Id: I78507d30af3d4f355963ff9e28ccf8f7354d5dbd CRs-Fixed: 596021
| | * | | | Fix of: Enabling fast roam while connected does not work.Naveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Disable fast roaming in .ini file and connect station to AP. Enable fast roaming using SETFASTROAM=1 command. Change attenuation of AP to force smooth roaming. Station gets beacon miss and moves to the new AP. Change-Id: I310b9a4589f8171855a3665d7431ac748b3f24d8 CRs-Fixed: 595949
| | * | | | NCHO SET API don't take effect immediately.Naveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In wma.c sending roaming params at start and at update cfg had different behavior. Also for new params to take effect, roaming engine needs to be stopped and restarted. Change-Id: Ie7e45f95ce4f011e1dd338fac74117fcc14d43e6 CRs-Fixed: 595326
| | * | | | qcacld: fix for SETSCANCHANNELTIME value is not taking effectNaveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Roaming firmware was uses single value for dwell time and not max and min. Host was passing two values and this was causing wrong value to be taken and correct value to be dropped. Change-Id: Idad06453a221d70c6e37638315bb7d072e1bc6a6 CRs-Fixed: 595262
| | * | | | qcacld: fix for Manual REASSOC from NCHO APINaveen Rawat2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | REASSOC request is processed in Roam_Scan_Offload_Rsp, hence posting a dummy rsp msg from WMA back to SME with proper reason code. Change-Id: I3bc5098816fa7997a3da9cfdc7ac222593571a48 CRs-Fixed: 596319
| | * | | | qcacld-new: Fix of Batch scan RSSI valueRajeev Kumar2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In CLD ROME FW returns raw RSSI value and driver need to convert it into proper RSSI value Change-Id: If2996144fa460d32725c9115a46cd9f5acd02372 CRs-Fixed: 596379
| | * | | | qcacld : Configure CSA wake up event during suspendAtul Mittal2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Firmware sends wow wakeup reason as CSA only when WOW_CSA_IE_EVENT is registered by the host before enabling wow Change-Id: Ie40c26a577be44ac424a94f0fb3af2de00e4fe80 CRs-fixed: 594942
| | * | | | qcacld : Enabling NS offload by default in ini fileAtul Mittal2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | NS offload feature should be enabled by default in ini file Change-Id: I6ae928e5d2b8727555766192f89531dd9bc99be8 CRs-fixed: 594452
| | * | | | qcacld: Move one more bank from DRAM to IRAMSameer Thalappil2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For Rome 2.0, move one more bank from DRAM to IRAM. Change-Id: Ief5744deade564df81fba22787e4d4b6547a3ab4 CRs-Fixed: 596501
| | * | | | wlan: Fix architecture issue with voss messaging and timersJeff Johnson2014-01-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The current usage of vos messaging does not work correctly on all software architectures due to the overloading of certain message fields. Correct this by adding an explicit 'callback' field to the vos message structure, and modify the vos timer messaging to use this new field. .. Change-Id: I5fb3d30f32dee7dbdc300637dcf16c190392a280 CRs-fixed: 567258
generated by cgit v1.2.3 (git 2.25.1) at 2026-04-23 04:11:31 +0000