summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
...
| * | | | qcacld-2.0: SSR-protect iw_get_mode()Jeff Johnson2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently iw_get_mode() does not use the SSR protection wrapper. This can lead to unexpected behavior if the function is executing concurrently with SSR, so add the SSR protection wrapper. Change-Id: I0bd055345e6fc22087f14e3d37b2b8c9e3239322 CRs-Fixed: 891029
| * | | | qcacld-2.0: SSR-protect iw_set_mode()Jeff Johnson2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently iw_set_mode() does not use the SSR protection wrapper. This can lead to unexpected behavior if the function is executing concurrently with SSR, so add the SSR protection wrapper. Change-Id: Ied574a68b5f9dfbf75ab6157b3e80b25b69bd54b CRs-Fixed: 891029
| * | | | qcacld-2.0: SSR-protect iw_get_freq()Jeff Johnson2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently iw_get_freq() does not use the SSR protection wrapper. This can lead to unexpected behavior if the function is executing concurrently with SSR, so add the SSR protection wrapper. Change-Id: I0321c47d270b5a5ae04ff3307071bb0df36fe8df CRs-Fixed: 891029
| * | | | qcacld-2.0: SSR-protect iw_set_freq()Jeff Johnson2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently iw_set_freq() does not use the SSR protection wrapper. This can lead to unexpected behavior if the function is executing concurrently with SSR, so add the SSR protection wrapper. Change-Id: I55d2449017015cf8dbd1de2ac972ef37e63fbcde CRs-Fixed: 891029
| * | | | qcacld: Use ini flag to control FW initiated self recoveryYue Ma2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use the same ini flag to enable/disable FW initiated self recovery. If the flag is disabled, host will treat it as a normal FW assert. This will also apply to host driver found PCIe linkdown or ROME SOC cannot wake up where it will trigger a VOS_BUG. Change-Id: Ida53fb834605750e503445cc0dbb9c570721f67b CRs-fixed: 894205
| * | | | qcacld-2.0: Fix issue in Tx LDPC enable/disable using INIKiran Kumar Lokere2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | FW checks the peer Rx LDPC capability in ht and vht capabilities to do Tx LPDC. But host sets the Tx LPDC flag in peer_flags. To disable Tx LDPC reset the Rx LDPC capability in peer ht and vht caps based on Tx LPDC configuration in INI. Change-Id: I211fd73fd70a35faaf4cfc2c5dc8d842c7ff1e5a CRs-Fixed: 894306
| * | | | qcacld-2.0: Remove mPnoScanPending and make scan_lock a timed oneSrinivas Girigowda2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | mPnoScanPending was added to skip normal scan when PNO scan is in progress. There is no need to skip the normal scan which is already offloaded to fw and fw takes care of scheduling both normal scan and PNO Scan. 1. Remove mPnoScanPending. 2. Make wake lock acquired during scan a timed wakelock for 5 sec. Change-Id: I61cb8888e23b405990b4fd2733aa04e8d93f0a1c CRs-Fixed: 894939
| * | | | qcacld-2.0: Extract the Adapter context from notifier block argumentVinay Krishna Eranna2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a prima to qcacld-2.0 propagation. Current implementation of IP notifier extracts adapter context from the net device structure and then interpret it. The notifier can be called for different interface other than the one the driver is interested in, which leads interpreting an invalid address as the valid adapter context. Avoid this by extracting the relevant information using the notifier block argument passed to the IP notifier callback. Change-Id: Iaf1529b393e5433790baf60985fed7662cf5e92f CRs-Fixed: 791657
| * | | | qcacld: Cancel pending work when exit runtime PM featureYue Ma2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Cancel pending runtime PM work when exit runtime PM feature. This can avoid race condtion between the work and driver unloading. Change-Id: Ie1f9a259f4f8e6fbe08c5ba4023ee456b5140087 CRs-fixed: 894804
| * | | | qcacld: Set PM state at proper place in runtime suspendYue Ma2015-08-24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Set PM state after load/unload/LOGP checks in runtime suspend. This can avoid crash when driver unloads while runtime suspend still comes. Change-Id: Ie48bcc837d2dc6e42dd75c81e7223f320fc38a99 CRs-fixed: 894109
| * | | | Release 4.0.10.165AAkash Patel2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Release 4.0.10.165A CRs-Fixed: 688141 Change-Id: Ib6b9b0b71e0575075f45fb8ed8146c90a48f549c
| * | | | qcacld-3.0: Prevent runtime PM for 30 seconds during connectionRajeev Kumar2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Prevent runtime PM for 30 seconds during connection Change-Id: Idc817a241da1f4d2db5a6282549e3e1b1e97a78e CRs-Fixed: 894404
| * | | | Release 4.0.10.165AnjaneeDevi Kapparapu2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Release 4.0.10.165 Change-Id: Ia7b03587998f2e7c24bc7ac49ef93f359ff557b5 CRs-Fixed: 688141
| * | | | qcacld: hif: Optimize runtime prevent suspend APIsPrashanth Bhatta2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Runtime prevent suspend APIs namely, prevent_suspend, allow_suspend and prevent_suspend_timeout were independently doing get/put. Since these APIs are also called with vos_wake_lock APIs, usage count on the kernel runtime PM side used to get messed up. Mainly in the scenarios like prevent_suspend_timeout followed by allow_suspend. Fix the issue by using internal reference count and handle all 3 APIs properly to not to mess-up the runtime PM usage count. Change-Id: I453f6c3858f6ae2e24adfd830ffb54036f23344a CRs-fixed: 894421
| * | | | qcacld-2.0: Prevent Runtime PM when connection in progressKomal Seelam2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Association may take few seconds in cases where OTA is noisy. During Association if runtime PM gets triggered, we do Pdev Suspend on the interface and remote wakeup capability on the interface is disabled, we may eventually lose all management/data(like DHCP) frames coming from AP which will result into assoc failure, key exchange timeout, DHCP timeout etc. Prevent Runtime PM to ensure Power Save is not affecting functionality in stress cases. Change-Id: I3048dcc0e61b2193094ff5d589e4afc355cf6341 CRs-Fixed: 894404
| * | | | qcacld-2.0: extscan: Return timestamp since boot instead of from frameSrinivas Girigowda2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently for extscan full scan results, scan results are forwarded to the user space and timestamp is set from the received frame. On android platforms, instead of timestamp from frame, return the timestamp as time since boot in micro seconds Change-Id: If34be897af1f24515ccd134ad369c976f10bcda9 CRs-Fixed: 894193
| * | | | qcacld-2.0: extscan: Modify repeat_probe_time based on dwell timeSrinivas Girigowda2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently repeat_probe_time is hardcoded to 10 ms, instead it should be variable depending on dwell time and number of probes configured. Change-Id: Ic176e12005a0cb6f2323ecf48785bba87e300b48 CRs-Fixed: 892740
| * | | | qcacld-2.0: Add missing MTRACE log in HDDKonamki, Sreelakshmi2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | prima to qcacld-2.0 propagation Currently, MTRACE logs are not captured for the events posted from SME to HDD. So, add missing MTRACE log in hdd_smeRoamCallback() and also add missing eRoamCmdStatus enums in get_eRoamCmdStatus_str(). Change-Id: I26ef8c651b7b873939341b643056ff977a40ea19 CRs-Fixed: 855875
| * | | | qcacld-2.0: Fix compilation error on x86 inNirav Shah2015-08-21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | wlan_hdd_display_tx_rx_histogram Fix compilation error on x86 in wlan_hdd_display_tx_rx_histogram. CRs-Fixed: 866743 Change-Id: I83f667386dd2c46fdd38647e436b92485754ae95
| * | | | qcacld-2.0: enhance ipa realtime debuggingLeo Chang2015-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | code review bug fix Enhance ipa uc realtime debug capability Internal counters may print whenever user required Previous change code review comments clean up Change-Id: I787cf0a9416faee69d99ee2989d0e2c9cfeb565e CRs-fixed: 892958
| * | | | Release 4.0.10.164AnjaneeDevi Kapparapu2015-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Release 4.0.10.164 Change-Id: If21760ed9f079a3610677d71a11a7b7fec022603 CRs-Fixed: 688141
| * | | | qcacld-2.0: Add vendor command support for setbandSrinivas Dasari2015-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a prima to qcacld-2.0 propagation. This commit allows the upper layers to send setband request to driver using vendor command. Change-Id: I3ce769e5426a5afb85979b4e19dbabd234f18e98 CRs-Fixed: 879527
| * | | | qcacld-2.0: Do not send cfg80211_scan_done at unload timemukul sharma2015-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As a part of NETDEV_GOING_DOWN, host try to abort the scan but time out (Fwr was in bad state) happens hence cfg80211_scan_done was not invoked. Since kernel 3.14, Kernel complete the pending scan on NETDEV_DOWN event, hence it is freeing the scan request. During driver unload, Driver release the command and invoke the callback.Hence cfg80211_scan_done will be invoked. But the scan req was freed by kernel earlier, Which lead to BUG_ON on NULL wiphy in kernel. Change-Id: Ie5f0a562a6f591fa6e5ae9084f03cf3b7b526e91 CRs-Fixed: 872270
| * | | | qcacld: Schedule any pending RoC after existing RoC completionChandrasekaran, Manishekar2015-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Schedule any queued, pending remain on channel request after the completion of any existing remain on channel (either p2p listen or off channel Tx) request. When supplicant issues an off channel Tx and runs into completion, and the processing of this RoC completion of off channel tx is in progress within the driver, if the supplicant sends a new RoC request, it will get queued. The queued RoC request needs to be dequeued and executed when the processing of the RoC completion is done. Change-Id: I3d195e43ffa960b30b9366b8eea78019e8593889 CRs-Fixed: 893143
| * | | | qcacld-2.0: enhance ipa realtime debuggingLeo Chang2015-08-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Enhance ipa uc realtime debug capability Internal counters may print whenever user required Previous change code review comments clean up Change-Id: I51bbed695a01c73a877fd892b360f5635ec599a1 CRs-fixed: 892958
| * | | | Release 4.0.10.163AnjaneeDevi Kapparapu2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Release 4.0.10.163 Change-Id: Ic3e542d421a73357612f834a4a789777dffcab3f CRs-Fixed: 688141
| * | | | qcacld-2.0: Change tx/rx histogram statisticsNirav Shah2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use bandwidth computation timer for calculating tx/rx histogram statistics and add more debugging information. CRs-Fixed: 866743 Change-Id: Iabd7b642e33824cccdbe667d4898582f517da871
| * | | | qcacld-2.0: Do not reset rx_buf_debug entry when in_use is falseNirav Shah2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Do not reset rx_buf_debug (Buffer used to keep history of RX Buffers) entry when we set in_use flag to false. CRs-Fixed: 887590 Change-Id: I03a19506027b9186e9ff17ff6883e1460e308a9a
| * | | | qcacld-2.0: wlan: Cleanup North-Bound load/unload logsHanumantha Reddy Pothula2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | prima to qcacld-2.0 propagation If IOCTLs or CFG callbacks are issued continiously during driver load/unload or during SSR, observing stability issues because of excessive logging. Hence move log level to INFO from ERROR in wlan_hdd_validate_context function and don't print return status in its caller functions. Change-Id: I8fd9b25e0c774e1e12b01128eb86783c9dec2ab7 CRs-Fixed: 800631
| * | | | qcacld: Fix race condition between runtime/system PMKomal Seelam2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | WMI_WOW_DEL_WAKE_PATTERN_CMDID is not properly tagged as runtime packet during some scenarios wherein runtime suspend coming immediately after runtime resume. Intially the design was not to tag the WOW_DEL commands as the FW won't operate on the wakeup patterns after wow is disabled. I5eae66e09d5fac4f913c0eaaf90ff600d8c7d626 added few optimizations not to delete the patterns if wow is not configured. But there can be cases where in few of WOW_DEL commands are getting tagged and few not due to changes in commit bbb44d9f23d868a2837c6b22b8dfb123d8e7800c in 3.18 kernel where runtime suspend is coming back to back after resume and when cfg80211 resume comes when device is runtime suspend, these WOW_DEL commands are going as runtime tagged and accessing the PCIe link which clocked down during runtime suspend. The commit Ibf7eca6c344d427e0ac05c1bd903bab1073236a5 fixes the synchronization issue between system PM and runtime PM, ensuring driver system suspend happens after device is runtime active till system resume completes. Ensure to tag all runtime PM control path commands by setting runtime_pm_inprogress flag to false after sending all the WOW_DEL commands. Since runtime/system suspend/resume all are serialized to MC thread, the packets will reach FW in the order we queued. Improve the logic in I5eae66e09d5fac4f913c0eaaf90ff600d8c7d626 to feed WoW config during system suspend by ensuring device runtime status is ACTIVE. Change-Id: I365a8a43e37278207a643058896913a2d60a4990 CRs-Fixed: 891485
| * | | | qcacld-2.0: add rt debug capabilityYun Park2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | new real time debug capability. proactively detect out of memory condition and added new packet routing debug log and histogram Change-Id: Idf23d072697572375fd187bc2996bc5dbb518c3d CRs-Fixed: 892958
| * | | | qcacld-2.0: Update MCS index appropriatelyHanumantha Reddy Pothula2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | prima to qcacld-2.0 propagation When AP is configured with MCS index less than 4(43Mbps) then Host reports invalid MCS Index to upper-layer, leading to display invalid link speed on GUI. Updating MCS index with valid value and thereby updating correct MCS index to upper-layer Change-Id: Ibd1503be3825c879b7fa52bdfb6ff07423c39023 CRs-Fixed: 855022
| * | | | qcacld-2.0: SSR-protect wlan_hdd_cfg80211_set_ap_channel_width()Jeff Johnson2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently wlan_hdd_cfg80211_set_ap_channel_width() does not use the SSR protection wrapper. This can lead to unexpected behavior if the function is executing concurrently with SSR, so add the SSR protection wrapper. Change-Id: I6d35881235463b492e0375c0d268653b25862f0f CRs-Fixed: 889507
| * | | | qcacld-2.0: Update file access permission for firmware dumpKrishna Kumaar Natarajan2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Update file access permission for firmware memory dump file so that android framework can access the binary file while generating bug report. Change-Id: I0cd35cd90657059c760f68d986abc10d12c91acc CRs-Fixed: 892026
| * | | | qcacld: Remove calls to runtime prevent/allow suspendPrashanth Bhatta2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | With the new change I846930213cfd28e8fbe172f4871a9cd0f99bc835 to not to allow runtime suspend whenever wake_locks are held, there is no need of explicitly calling prevent/allow runtime suspend APIs as they will be duplicate requests. Also do not call runtime prevent/allow suspend APIs during SSR as HIF context may not be available during SSR. Moreover, runtime PM will not be triggered till SSR completes so it is not necessary. Change-Id: Ida8fc93e90e28978ac11dacd1e571786c5475a29 CRs-fixed: 892870
| * | | | qcacld: wma: Fix the condition for runtime PMPrashanth Bhatta2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Condition to check if the suspend is because of the runtime is not right because of which broadcast ARP and NS offloads are not configured which can cause issues. Correct the condition check so that broadcast ARP and NS offloads are configured properly. Change-Id: I8455314bd95e150e6bace33efe7f53161958bc0f CRs-fixed: 892486
| * | | | qcacld: Optimize wlan wake lock durationsRajeev Kumar2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Remove obsolete RA match wake lock - Wake locks are acquired for ridiculously long time - Reduce wake lock time to match real world use case Change-Id: Ie8c5c416923ad38d448dc8c63e0787c0bc28940d CRs-Fixed: 892557
| * | | | qcacld-2.0: Change RA interval to 600 from 60Srinivas Girigowda2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Increase the RA rate limit interval to 600 seconds. This implies that if Router configuration from the same router does not change, we forward it only once in 10 minutes to conserve power. Change-Id: I547fa8c20889fd0401fbb5eca3a5d27f48cd0702 CRs-Fixed: 892695
| * | | | Revert "qcacld: Don't disconnect in case of HS2.0 for change in AP capability."Sandeep Puligilla2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This reverts commit d6241c9e5cc9a2a87826fc95b1c547e269676a7b Change-Id: Idf6b428e07cff6da55bf1e90e48d48d27cbc794c CRs-Fixed: 892921
| * | | | qcacld-2.0: CL 1363882 - update fw common interface filesSandeep Puligilla2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | wmi command to configure sap blacklist parameters Change-Id: I7a4fcbdcfcbe8bdc2f939b7ac6a6053f0ce88722 CRs-Fixed: 865207
| * | | | qcacld-2.0: CL 1362845 - update fw common interface filesSandeep Puligilla2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | define WMI_LPI_STATUS_EXTSCAN_CYCLE_AND_SCAN_REQ_COMPLETED Change-Id: Ideb4e555146c2b7ac75e72546429dd224173f183 CRs-Fixed: 865207
| * | | | qcacld-2.0: CL 1360810 - update fw common interface filesSandeep Puligilla2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Packet filter configure and enable commands Change-Id: I7ab7d4a98a25d0c4746f4524488449128701c310 CRs-Fixed: 865207
| * | | | cld-2.0: Fix memory leak in tx flow control timerMasti, Narayanraddi2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | TX flow control timer initialized when STATION interface is created. During IBSS startup, STATION interface is changed to ADHOC type. TX flow control timer is not required, so close tx flow control and destroy timer. Change-Id: I6ab6a91ed2bc54e00d00ede02b466d8ca294d1f3 CRs-Fixed: 882821
| * | | | cld-2.0: Fix memory leak in TDLS when interface is changed to ADHOCMasti, Narayanraddi2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | TDLS initialization is done when STATION interface is created. During IBSS startup, STATION interface is changed to ADHOC type. TDLS is not applicable for ADHOC interface, so call wlan_hdd_tdls_exit to perform the TDLS cleanup. Change-Id: I8d6df15004c1e2f7bf6900a14aa6c2abc848032f CRs-Fixed: 882821
| * | | | qcacld-2.0: TDLS: provide operating class and channel to HALMasti, Narayanraddi2015-08-19
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 1.Provide off channel and operating class info to user space in HAL API wifi_get_tdls_status. If off channel is not enabled then provide base channel info and in that case pass opclass as 0. 2.Provide correct tdls state based on tdls off channel configuration. Change-Id: Ic60500b95ebddcb88b4e03f29178068c5322b2b9 CRs-Fixed: 838167
| * | | | Release 4.0.10.162AnjaneeDevi Kapparapu2015-08-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Release 4.0.10.162 Change-Id: I67653d4b4fe22197570d8f72ffcb11cea9686d8e CRs-Fixed: 688141
| * | | | qcacld-2.0: Donot defer Rx frames processing if ADD TS request sentHema Aparna Medicharla2015-08-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | prima to qcacld-2.0 propagation Current implementation defers all the received frames when ADD TS request is sent out over the air. This, in addition, to check for pMac->lim.gLimAddtsSent incurs delay, till ADD TS response timeout, in processing of the ADD TS response received if SET_LIM_PROCESS_DEFD_MESGS is set to FALSE. As there is no rationale defer processing of Rx frames while waiting for ADDTS response, changes are made to not defer processing of RX frames for the same. Change-Id: I1d1a41a0e3555ae19548a0cabf3b9569af775f72 CRs-Fixed: 839546
| * | | | qcacld: voss: Tie wake_lock with runtime PMPrashanth Bhatta2015-08-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Prevent runtime suspend when wake_locks are held so that there is no bus suspend. Change-Id: I846930213cfd28e8fbe172f4871a9cd0f99bc835 CRs-fixed: 891862
| * | | | qcacld: voss: Add timeout prevent suspend supportPrashanth Bhatta2015-08-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add timeout prevent suspend API for runtime PM so that it can be used in cases where there is no deterministic way to allow the runtime suspend. This API uses single timer to allow the suspend and timer is modified if the timeout is changed before timer fires. Also added optimization such that if the timeout is less than auto_suspend_delay then don't start the timer instead just use mark_last_busy. It is wise to try not to use this API and correct the design if possible. Change-Id: Iee76e748c82e9e7b58ddcd5ac6f93bd8f2fff29e CRs-fixed: 891862
| * | | | qcacld: hif: Add timeout prevent suspend supportPrashanth Bhatta2015-08-18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add timeout prevent suspend API for runtime PM so that it can be used in cases where there is no deterministic way to allow the runtime suspend. This API uses single timer to allow the suspend and timer is modified if the timeout is changed before timer fires. Also added optimization such that if the timeout is less than auto_suspend_delay then don't start the timer instead just use mark_last_busy. It is wise to try not to use this API and correct the design if possible. Change-Id: I080250a2e4a5a023fde37a69b9e41c7302438f43 CRs-fixed: 891862
generated by cgit v1.2.3 (git 2.25.1) at 2026-01-31 13:39:13 +0000