summaryrefslogtreecommitdiff
path: root/drivers/platform/msm/mhi (follow)
Commit message (Collapse)AuthorAge
* mhi: core: fix sequence for new transferjiad2017-11-22
| | | | | | | | | | | | | | | | | sequence for both rddm and fw download are initialized to be 0. They are incremented by 1 in bhi_probe() and upon real transfer takes place. Issue is in SSR case, sequence is incrementing in a way like 1, 3, 5 and etc as observed. The increment in bhi_probe() seems unnecessay as there's no real transfer. Fix is to initialize sequence to 1 and increment by 1 when real transfer takes place in bhi_bhie_transfer(). Change-Id: I210840cd1f7fecc300ecb69da8a50fbb1d200562 Signed-off-by: Jia Ding <jiad@codeaurora.org> CRs-Fixed: 2147433
* mhi: core: move mhi_states_transition_str out of stackSujeev Dias2017-09-22
| | | | | | | | | | To prevent un-initialized address returning from MHI state transition lookup table, move mhi_states_transition_str array out of stack CRs-Fixed: 2111049 Change-Id: Ife5f907a711cc0c615161f2b2a677275d6773686 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* msm: mhi: initialize variables to default valuesRama Krishna Phani A2017-08-03
| | | | | | | | | Few functions have variables which might get used with out proper initialization. Initialize variables to default values. CRs-Fixed: 2087109 Change-Id: I7645940e8d466e0ef67a5b8b7702b18b160cc10f Signed-off-by: Rama Krishna Phani A <rphani@codeaurora.org>
* mhi: core: add a timeout when waiting for MHI state BHI event.Sujeev Dias2017-06-16
| | | | | | | | | | It's possible for POR to fail and endpoint not enter correct reset state. Add a timeout when waiting for MHI state BHI transition event during firmware download. CRs-Fixed: 2062442 Change-Id: I6af16ec1a514d50b9e984e3fb16696d23e605ba0 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Add support for host triggered device ram dumpSujeev Dias2017-06-06
| | | | | | | | | | Add support for bus master to trigger PCIe device into ram dump mode to collect device ram dump. RDDM capabilities are extended to support under kernel panic as well. CRs-Fixed: 2055981 Change-Id: I18f7d5784992df70aafc6e41d248ced3fac37181 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Add support for PCIe linkdown abortSujeev Dias2017-06-06
| | | | | | | | | PCIe linkdown event can trigger anytime. During linkdown event abort any work MHI host doing that require PCIE link. CRs-Fixed: 2055975 Change-Id: Idba907c7977594dc08575e0020a7afa4868bc0e4 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Using usleep_range instead of msleep for sleepSujeev Dias2017-06-03
| | | | | | | | | msleep is not accurate for small delays. Use usleep_range for accuracy. CRs-Fixed: 1117121 Change-Id: I9a076f843875bd83a21c96d4e4ea0510e82ce352 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Bind devices with variable PCIe device IDs with same device nodeSujeev Dias2017-06-01
| | | | | | | | | | | In order to bind devices with variable PCIe device IDs with same device tree node, make PCIe device ID optional parameter in device tree node. If PCIe device ID not specified in device tree, any device can be bind with the node. CRs-Fixed: 2040899 Change-Id: Iaedcb40db15fa0e8c5c30eb147c904a10cd7569d Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Only allow registration if channel is supported by endpoint.Sujeev Dias2017-05-31
| | | | | | | | | Not all endpoints support all channels supported by MHI host. Only allow clients to register if requested channel is supported by endpoint. CRs-Fixed: 2053250 Change-Id: Ib4bd8d21b8f43d693da30b8fafe2cb9d2e128a17 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Fix memory leak in MHI DE_INITSujeev Dias2017-05-31
| | | | | | | | | | During MHI INIT, memory is allocated for firmware and rddm vector table. This memory is not freed when DE_INIT is called. CRs-Fixed: 2053206 Change-Id: I85b2644189915dd25d8ac31190f16799f86cbd12 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: fix potential buffer overflowSujeev Dias2017-05-24
| | | | | | | | | | Fix a potential buffer overflow that could occur during firmware download by storing buffer address locally instead of inside MHI context. CRs-Fixed: 1109397 Change-Id: If27ba602cdafdc8d25a94fdc0d74e8970bf0b0f4 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* Merge "mhi: core: add support for 64bit doorbell address"Linux Build Service Account2017-04-27
|\
| * mhi: core: add support for 64bit doorbell addressSujeev Dias2017-04-25
| | | | | | | | | | | | | | | | | | Doorbell pointers are 64bit addresses. Use correct data type when passing the address. CRs-Fixed: 2038601 Change-Id: Ie8e42d38944c0529386b90f5a6d37dcd77c19184 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* | mhi: core: allow communication during shutdownSujeev Dias2017-04-24
|/ | | | | | | | | | MHI clients allowed to communicate with external soc via MHI during shutdown. Only disable MHI communication after receiving shutdown notification. CRs-Fixed: 2037184 Change-Id: If69f61a244a6175256753dea361085fa5f9afeac Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add support for subsystem restart and shutdownSujeev Dias2017-03-31
| | | | | | | | | Add support for subsystem restart and shutdown processing for slave mode and master mode. CRs-Fixed: 1097560 Change-Id: I8c60605d202d38a7a8f8fb587a179f95ec402692 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: process incoming data based on event ring typeSujeev Dias2017-03-17
| | | | | | | | | | mhi clients can manage both hardware and software event rings. Process incoming data based on event ring type instead of channel type. CRs-Fixed: 2018326 Change-Id: Ifa2de76034224c19a846f7be153400888f8c66de Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: only accumulate transfer complete events as processed eventsSujeev Dias2017-03-17
| | | | | | | | | | When a client ask MHI host to processed X number of events, only count transfer events as part of processed events. All other events are internal events. CRs-Fixed: 2018324 Change-Id: I4cfcae0ba187348442c3fc45b76d7c4b01f687bd Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: mhi_core: move mhi from module_Init to subsys_initSujeev Dias2017-03-11
| | | | | | | | | | MHI clients expects MHI to finish initializing before clients probe. Move MHI to subsys init to guarantee MHI will finish executing before device drivers with module_init starts. CRs-Fixed: 2006891 Change-Id: I44e3612721fdc247c8c083f74736a1b1bcfc962e Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add support for priority based event processingSujeev Dias2017-03-11
| | | | | | | | | | | Not all events require same priority to process. Process high priority events using a tasklet while processing low priority events using a kworker. Priority for events can be configured by device tree. CRs-Fixed: 2004759 Change-Id: I6bf04624c2fd92afd7c698d6a57f66619ea13b5b Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: confirm MHI offset is valid before using itSujeev Dias2017-03-11
| | | | | | | | | | | PCIe link can go down anytime asynchronously without host knowing. In event of a link-down, MMIO reads will always return 0xffffffff. Confirm the read value is valid prior to executing further. CRs-Fixed: 1030692 Change-Id: If052a1a31c62a81f380cec3f523b605b38a33847 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: fix memory alignment calculationSujeev Dias2017-03-11
| | | | | | | | | | Fix issue of incorrectly truncating 64bit address as a 32bit address during memory allocation for MHI. CRs-Fixed: 1113884 Change-Id: Iba5123b3710c805cabf63bb6a2958d8218d642e8 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add support for bounce buffer transferSujeev Dias2017-03-11
| | | | | | | | | | | Not all MHI capable devices can access entire host DDR. If an MHI client provide a transfer packet that’s outside of supported address range, copy the packet to a local bounce buffer prior to transfer. CRs-Fixed 1110280 Change-Id: Ie7292a2c89cc6608d5360f8a330d78635d913c29 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add support for MHI to operate on slave modeSujeev Dias2017-03-11
| | | | | | | | | MHI host is not always the bus master. Add support to operate on slave mode. CRs-Fixed: 1095436 Change-Id: I4af92c8171cf9ba38267afa753cec33f291d55eb Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add support for MHI host managing firmware uploadSujeev Dias2017-03-11
| | | | | | | | | Add support for MHI host to directly upload firmware to compatible devices using MHI protocol. CRs-Fixed: 1095436 Change-Id: Iff7043f1f9afc4824edeaeccc46ed427ce7ee291 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Add support for multiple MHI devicesSujeev Dias2017-03-11
| | | | | | | | | | | | | In order to support multiple MHI devices connected to same host we have, re-organize struct mhi_device_ctxt such that all information related to device is included inside struct mhi_device_ctxt. Created an abstract power management layer to operate MHI in bus master, and slave mode. PCIe bus, device, slot are extracted from DT to identify physical device. CRs-Fixed: 1086301 Change-Id: I27a12b7412db75e4843794d80146ca7475b02fdc Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Mark unprocessed events as STALESujeev Dias2017-02-24
| | | | | | | | | | | | MHI clients permitted to reset channels asynchronous to pending data. It's possible after channel reset there are pending events for that particular channel. Mark all those events as stale, such that MHI event process thread will ignore them. CRs-Fixed: 1074226 Change-Id: I86547aa2c5d7ad5095807cfe87378f5540c05a66 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: power management redesignSujeev Dias2017-02-24
| | | | | | | | | | In order to support subsystem restart, and link down recovery, redesign mhi power management state machine. CRs-Fixed: 1081654 Change-Id: I3005b829bfdea1d3e9f086676c446b62b5d30d0b Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add support to enable MHI burst mode per channel basisSujeev Dias2017-02-22
| | | | | | | | | Not all MHI hardware channels support burst mode, add support to enable burst mode per channel basis. CRs-Fixed: 1027069 Change-Id: Icd3061f1a1a4b4b3dab8f9fe3b989a8afefd18b2 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Add support for new MHI hardware channelSujeev Dias2017-02-22
| | | | | | | | | Add support for new MHI hardware channel 102 to be use by MHI clients as ADPL channel. CRs-Fixed: 1027069 Change-Id: Ib3c2019fc269064d097bb7f40f01d4580e63a603 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Do not clear transaction statusSujeev Dias2017-02-22
| | | | | | | | | | | MHI transaction status stores the OVERFLOW status received from device. MHI clients uses this status to determine overflow buffers, do not clear the status. CRs-Fixed: 1042516 Change-Id: Iaaff06c1c39775d6a33ca17851f1e1579b2a2ecb Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* msm: mhi: Check bb ring and transfer ring when checking for spaceSujeev Dias2017-02-22
| | | | | | | | | When checking for available spaces, check available spaces on both bounce buffer ring and transfer ring and return min. Change-Id: I9208b46c32821de3f5d9e3d828087d7bc29b9546 CRs-Fixed: 1055681 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add missing MHI stateSujeev Dias2017-02-22
| | | | | | | | | Add missing state MHI_STATE_RESERVE to MHI states look up table. CRs-Fixed: 1049595 Change-Id: I9a6bd2750f81f6cabc1e7b5aff488b4a01f7897d Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Preserve DB Mode state during lpmSujeev Dias2017-02-21
| | | | | | | | | | When exiting low power modes (M3) do not reset the DB Mode state if DB mode preserve flag is set for channel. CRs-Fixed: 1022868 Change-Id: I6557d28afe9d0ac11b76c683ffba76d7d6ffd377 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Do not reset doorbell mode during MHI_M3 abortSujeev Dias2017-02-21
| | | | | | | | | | In order to avoid being out of sync between MHI client and MHI host, host shall not reset the doorbell modes for hardware channels during MHI_M3 state transition abort. CRs-Fixed: 1023725 Change-Id: I6c742fc968fd57d71a86039bf1f3f65b1362bc90 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Lock MHI_M3 state transitionTony Truong2017-02-21
| | | | | | | | | | | To avoid any race conditions between MHI_M2 state transition and MHI_M3 state transition lock the entire MHI_M3 transition using xfer_lock. CRs-Fixed: 972390 Change-Id: I7c2e1b7b3966dc5fb8bf2f91bce734bbc58c6fd7 Signed-off-by: Tony Truong <truong@codeaurora.org> Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: add memory barrier after ring operationsSujeev Dias2017-02-21
| | | | | | | | | | | | There are instances when MHI channel context read pointer can be accessed simultaneously by different CPU cores. To make sure read pointer updates visible to all cores, add a memory barrier after completion of MHI ring operation. CRs-Fixed: 966338 Change-Id: Ifc8c4cd7595fed9049009c42420a665fb170079f Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* msm: mhi: fix error handlingSujeev Dias2017-02-21
| | | | | | | | | Validate the channel prior to proceeding further. Unlock spin lock before jumping to error handler. CRs-Fixed: 1016969 Change-Id: Ie3328f878b582a333ae15f3b950c258ec42fd768 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: change logs to print MHI_STATEs in textSujeev Dias2017-02-21
| | | | | | | | | | Instead of printing enum index convert MHI_STATE enum to text representation for easier interpretation of debug logs. CRs-Fixed: 1012249 Change-Id: I97a9a7ff293c739531d8197334a0f0a35bf20419 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* mhi: core: Protect DB register when handling DB MODESujeev Dias2017-02-21
| | | | | | | | | | | | Possibility exist when handling a DB MODE event another core to queue a TRE to same channel. During that time CC ctxt WP may get updated, however DB MODE event thread still be using a stale WP. Add a lock to synchronize DB MODE event thread and queue TRE thread. CRs-Fixed: 1005752 Change-Id: I7f285da8751a867a1c3d651466537368799eb657 Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
* msm: mhi: fix latent issues in MHI driversTony Truong2017-02-10
| | | | | | | | | Incorrect checks in MHI drivers may crash the system. Fix these checks so that MHI drivers will handle the errors gracefully. Change-Id: I462263799c2e76573300ed43b51e39ecba42914b Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Enable Runtime PM by defaultTony Truong2016-03-23
| | | | | | | | | Enable MHI runtime PM by default to enable additional power savings. Change-Id: I115af9e823ed685d7bf6e2ebaf18ad526256dc98 Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Fix race condition on poll_inboundTony Truong2016-03-23
| | | | | | | | | Fix race condition in mhi_poll_inbound which could leave the bounce buffer list and TRE list out of sync. Change-Id: I76c7de75256b59f8a0dff2a85af733c441761385 Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Read the EXEC_ENV from deviceTony Truong2016-03-23
| | | | | | | | | Read the EXEC_ENV register from device to determine the MHI base state. Change-Id: Id70a360cfa3775a7186e5059306385a37960df2d Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Enable dynamic event ring allocationTony Truong2016-03-23
| | | | | | | | | Enable dynamic allocation of event rings from the coherent pool to ensure appropriate page alignment with device. Change-Id: Id57ccd889cf96aad73f4ada01405e2d734351c8c Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Set correct DMA mask on deviceTony Truong2016-03-23
| | | | | | | | | Set the correct DMA mask on the platform device to avoid bounce buffering from the default 32 bit mask. Change-Id: I7e201e83fd21b0f0a6f57754a3b6f3eded9be58f Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: Use standard kernel return codesTony Truong2016-03-23
| | | | | | | | | | Replace the MHI defined enum for return codes with kernel standard return codes. This greatly improves readability and parsing of errors by userspace clients. Change-Id: Ica47f8c3a149f23243b5fb6a4aabcdc0a36b1535 Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Track number of failed M2 wakeupsTony Truong2016-03-23
| | | | | | | | | | | | Failed M2 wakeups may represent a serious hardware problem. The device is refusing to wake up from low power states and all modem functionality is now stalled. Track these wakeup attempts. Change-Id: I3115a4416f85290c5faead0e2d9f280d8cc3b662 Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Enable runtime PM on pcie deviceTony Truong2016-03-23
| | | | | | | | | Enable runtime PM on PCIe device to ensure sync with PCIe framework. Change-Id: Ied432440df0dc788642fe1a1a8d9195b6cc1cf7c Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Remove incorrect macro for tre sizesTony Truong2016-03-23
| | | | | | | | | Remove unused macros for TRE sizes. This information is now present in device tree. Change-Id: Id3ca59c62d66ce8913e83fcba60e1a912cf146ea Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
* mhi: core: Add mdmcalifornium as a supported MDM by MHITony Truong2016-03-23
| | | | | | | | | Add mdmcalifornium to the list of supported PCIe devices by MHI. Change-Id: I34e846caceb8fd2808594470cccbb7e318011cda Signed-off-by: Andrei Danaila <adanaila@codeaurora.org> Signed-off-by: Tony Truong <truong@codeaurora.org>
generated by cgit v1.2.3 (git 2.25.1) at 2026-01-28 11:05:35 +0000