Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 1 | #include "oem_event_manager.hpp" |
| 2 | |
| 3 | #include "requester/handler.hpp" |
| 4 | #include "requester/request.hpp" |
| 5 | |
| 6 | #include <config.h> |
| 7 | #include <libpldm/pldm.h> |
| 8 | #include <libpldm/utils.h> |
| 9 | #include <systemd/sd-journal.h> |
| 10 | |
| 11 | #include <phosphor-logging/lg2.hpp> |
| 12 | #include <xyz/openbmc_project/Logging/Entry/server.hpp> |
| 13 | |
| 14 | #include <algorithm> |
| 15 | #include <map> |
| 16 | #include <sstream> |
| 17 | #include <string> |
| 18 | #include <unordered_map> |
| 19 | |
| 20 | namespace pldm |
| 21 | { |
| 22 | namespace oem_ampere |
| 23 | { |
| 24 | namespace boot_stage = boot::stage; |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 25 | namespace ddr_status = ddr::status; |
| 26 | namespace dimm_status = dimm::status; |
| 27 | namespace dimm_syndrome = dimm::training_failure::dimm_syndrome; |
| 28 | namespace phy_syndrome = dimm::training_failure::phy_syndrome; |
| 29 | namespace training_failure = dimm::training_failure; |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 30 | |
Chau Ly | 3de0d94 | 2024-10-03 08:57:11 +0000 | [diff] [blame] | 31 | constexpr const char* ampereEventRegistry = "OpenBMC.0.1.AmpereEvent.OK"; |
| 32 | constexpr const char* ampereWarningRegistry = |
| 33 | "OpenBMC.0.1.AmpereWarning.Warning"; |
| 34 | constexpr const char* ampereCriticalRegistry = |
| 35 | "OpenBMC.0.1.AmpereCritical.Critical"; |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 36 | constexpr const char* BIOSFWPanicRegistry = |
| 37 | "OpenBMC.0.1.BIOSFirmwarePanicReason.Warning"; |
| 38 | constexpr auto maxDIMMIdxBitNum = 24; |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 39 | constexpr auto maxDIMMInstantNum = 24; |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 40 | |
| 41 | /* |
| 42 | An array of possible boot status of a boot stage. |
| 43 | The index maps with byte 0 of boot code. |
| 44 | */ |
| 45 | std::array<std::string, 3> bootStatMsg = {" booting", " completed", " failed"}; |
| 46 | |
| 47 | /* |
| 48 | An array of possible boot status of DDR training stage. |
| 49 | The index maps with byte 0 of boot code. |
| 50 | */ |
| 51 | std::array<std::string, 3> ddrTrainingMsg = { |
| 52 | " progress started", " in-progress", " progress completed"}; |
| 53 | |
| 54 | /* |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 55 | A map between PMIC status and logging strings. |
| 56 | */ |
| 57 | std::array<std::string, 8> pmicTempAlertMsg = { |
| 58 | "Below 85°C", "85°C", "95°C", "105°C", |
| 59 | "115°C", "125°C", "135°C", "Equal or greater than 140°C"}; |
| 60 | |
| 61 | /* |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 62 | In Ampere systems, BMC only directly communicates with MCTP/PLDM SoC |
| 63 | EPs through SMBus and PCIe. When host boots up, SMBUS interface |
| 64 | comes up first. In this interface, BMC is bus owner. |
| 65 | |
| 66 | mctpd will set the EID 0x14 for S0 and 0x16 for S1 (if available). |
| 67 | pldmd will always use TID 1 for S0 and TID 2 for S1 (if available). |
| 68 | */ |
| 69 | EventToMsgMap_t tidToSocketNameMap = {{1, "SOCKET 0"}, {2, "SOCKET 1"}}; |
| 70 | |
| 71 | /* |
| 72 | A map between sensor IDs and their names in string. |
| 73 | Using pldm::oem::sensor_ids |
| 74 | */ |
Chau Ly | 4cca3dc | 2024-10-03 09:07:09 +0000 | [diff] [blame] | 75 | EventToMsgMap_t sensorIdToStrMap = { |
Chau Ly | ef214b5 | 2024-10-16 09:40:38 +0000 | [diff] [blame] | 76 | {DDR_STATUS, "DDR_STATUS"}, |
| 77 | {PCP_VR_STATE, "PCP_VR_STATE"}, |
| 78 | {SOC_VR_STATE, "SOC_VR_STATE"}, |
| 79 | {DPHY_VR1_STATE, "DPHY_VR1_STATE"}, |
| 80 | {DPHY_VR2_STATE, "DPHY_VR2_STATE"}, |
| 81 | {D2D_VR_STATE, "D2D_VR_STATE"}, |
| 82 | {IOC_VR1_STATE, "IOC_VR1_STATE"}, |
| 83 | {IOC_VR2_STATE, "IOC_VR2_STATE"}, |
| 84 | {PCI_D_VR_STATE, "PCI_D_VR_STATE"}, |
| 85 | {PCI_A_VR_STATE, "PCI_A_VR_STATE"}, |
| 86 | {PCIE_HOT_PLUG, "PCIE_HOT_PLUG"}, |
| 87 | {BOOT_OVERALL, "BOOT_OVERALL"}, |
Chau Ly | b01357f | 2024-10-17 09:18:01 +0000 | [diff] [blame] | 88 | {SOC_HEALTH_AVAILABILITY, "SOC_HEALTH_AVAILABILITY"}, |
| 89 | {WATCH_DOG, "WATCH_DOG"}}; |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 90 | |
| 91 | /* |
| 92 | A map between the boot stages and logging strings. |
| 93 | Using pldm::oem::boot::stage::boot_stage |
| 94 | */ |
| 95 | EventToMsgMap_t bootStageToMsgMap = { |
| 96 | {boot_stage::SECPRO, "SECpro"}, |
| 97 | {boot_stage::MPRO, "Mpro"}, |
| 98 | {boot_stage::ATF_BL1, "ATF BL1"}, |
| 99 | {boot_stage::ATF_BL2, "ATF BL2"}, |
| 100 | {boot_stage::DDR_INITIALIZATION, "DDR initialization"}, |
| 101 | {boot_stage::DDR_TRAINING, "DDR training"}, |
| 102 | {boot_stage::S0_DDR_TRAINING_FAILURE, "DDR training failure"}, |
| 103 | {boot_stage::ATF_BL31, "ATF BL31"}, |
| 104 | {boot_stage::ATF_BL32, "ATF BL32"}, |
| 105 | {boot_stage::S1_DDR_TRAINING_FAILURE, "DDR training failure"}, |
| 106 | {boot_stage::UEFI_STATUS_CLASS_CODE_MIN, |
| 107 | "ATF BL33 (UEFI) booting status = "}}; |
| 108 | |
| 109 | /* |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 110 | A map between DDR status and logging strings. |
| 111 | Using pldm::oem::ddr::status::ddr_status |
| 112 | */ |
| 113 | EventToMsgMap_t ddrStatusToMsgMap = { |
| 114 | {ddr_status::NO_SYSTEM_LEVEL_ERROR, "has no system level error"}, |
| 115 | {ddr_status::ECC_INITIALIZATION_FAILURE, "has ECC initialization failure"}, |
| 116 | {ddr_status::CONFIGURATION_FAILURE, "has configuration failure at DIMMs:"}, |
| 117 | {ddr_status::TRAINING_FAILURE, "has training failure at DIMMs:"}, |
| 118 | {ddr_status::OTHER_FAILURE, "has other failure"}, |
| 119 | {ddr_status::BOOT_FAILURE_NO_VALID_CONFIG, |
| 120 | "has boot failure due to no configuration"}, |
| 121 | {ddr_status::FAILSAFE_ACTIVATED_NEXT_BOOT_SUCCESS, |
| 122 | "failsafe activated but boot success with the next valid configuration"}}; |
| 123 | |
| 124 | /* |
| 125 | A map between DIMM status and logging strings. |
| 126 | Using pldm::oem::dimm::status::dimm_status |
| 127 | */ |
| 128 | EventToMsgMap_t dimmStatusToMsgMap = { |
| 129 | {dimm_status::INSTALLED_NO_ERROR, "is installed and no error"}, |
| 130 | {dimm_status::NOT_INSTALLED, "is not installed"}, |
| 131 | {dimm_status::OTHER_FAILURE, "has other failure"}, |
| 132 | {dimm_status::INSTALLED_BUT_DISABLED, "is installed but disabled"}, |
| 133 | {dimm_status::TRAINING_FAILURE, "has training failure; "}, |
| 134 | {dimm_status::PMIC_TEMP_ALERT, "has PMIC temperature alert"}}; |
| 135 | |
| 136 | /* |
| 137 | A map between PHY training failure syndrome and logging strings. |
| 138 | Using |
| 139 | pldm::oem::dimm::training_faillure::phy_syndrome::phy_training_failure_syndrome |
| 140 | */ |
| 141 | EventToMsgMap_t phyTrainingFailureSyndromeToMsgMap = { |
| 142 | {phy_syndrome::NA, "(N/A)"}, |
| 143 | {phy_syndrome::PHY_TRAINING_SETUP_FAILURE, "(PHY training setup failure)"}, |
| 144 | {phy_syndrome::CA_LEVELING, "(CA leveling)"}, |
| 145 | {phy_syndrome::PHY_WRITE_LEVEL_FAILURE, |
| 146 | "(PHY write level failure - see syndrome 1)"}, |
| 147 | {phy_syndrome::PHY_READ_GATE_LEVELING_FAILURE, |
| 148 | "(PHY read gate leveling failure)"}, |
| 149 | {phy_syndrome::PHY_READ_LEVEL_FAILURE, "(PHY read level failure)"}, |
| 150 | {phy_syndrome::WRITE_DQ_LEVELING, "(Write DQ leveling)"}, |
| 151 | {phy_syndrome::PHY_SW_TRAINING_FAILURE, "(PHY SW training failure)"}}; |
| 152 | |
| 153 | /* |
| 154 | A map between DIMM training failure syndrome and logging strings. |
| 155 | Using |
| 156 | pldm::oem::dimm::training_faillure::dimm_syndrome::dimm_training_failure_syndrome |
| 157 | */ |
| 158 | EventToMsgMap_t dimmTrainingFailureSyndromeToMsgMap = { |
| 159 | {dimm_syndrome::NA, "(N/A)"}, |
| 160 | {dimm_syndrome::DRAM_VREFDQ_TRAINING_FAILURE, |
| 161 | "(DRAM VREFDQ training failure)"}, |
| 162 | {dimm_syndrome::LRDIMM_DB_TRAINING_FAILURE, "(LRDIMM DB training failure)"}, |
| 163 | {dimm_syndrome::LRDRIMM_DB_SW_TRAINING_FAILURE, |
| 164 | "(LRDRIMM DB SW training failure)"}}; |
| 165 | |
| 166 | /* |
| 167 | A map between DIMM training failure type and a pair of <logging strings - |
| 168 | syndrome map>. Using |
| 169 | pldm::oem::dimm::training_faillure::dimm_training_failure_type |
| 170 | */ |
| 171 | std::unordered_map<uint8_t, std::pair<std::string, EventToMsgMap_t>> |
| 172 | dimmTrainingFailureTypeMap = { |
| 173 | {training_failure::PHY_TRAINING_FAILURE_TYPE, |
| 174 | std::make_pair("PHY training failure", |
| 175 | phyTrainingFailureSyndromeToMsgMap)}, |
| 176 | {training_failure::DIMM_TRAINING_FAILURE_TYPE, |
| 177 | std::make_pair("DIMM training failure", |
| 178 | dimmTrainingFailureSyndromeToMsgMap)}}; |
| 179 | |
| 180 | /* |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 181 | A map between log level and the registry used for Redfish SEL log |
| 182 | Using pldm::oem::log_level |
| 183 | */ |
| 184 | std::unordered_map<log_level, std::string> logLevelToRedfishMsgIdMap = { |
Chau Ly | 3de0d94 | 2024-10-03 08:57:11 +0000 | [diff] [blame] | 185 | {log_level::OK, ampereEventRegistry}, |
| 186 | {log_level::WARNING, ampereWarningRegistry}, |
| 187 | {log_level::CRITICAL, ampereCriticalRegistry}, |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 188 | {log_level::BIOSFWPANIC, BIOSFWPanicRegistry}}; |
| 189 | |
Chau Ly | ef214b5 | 2024-10-16 09:40:38 +0000 | [diff] [blame] | 190 | std::unordered_map< |
| 191 | uint16_t, |
| 192 | std::vector<std::pair< |
| 193 | std::string, |
| 194 | std::unordered_map<uint8_t, std::pair<log_level, std::string>>>>> |
| 195 | stateSensorToMsgMap = { |
| 196 | {SOC_HEALTH_AVAILABILITY, |
| 197 | {{"SoC Health", |
| 198 | {{1, {log_level::OK, "Normal"}}, |
| 199 | {2, {log_level::WARNING, "Non-Critical"}}, |
| 200 | {3, {log_level::CRITICAL, "Critical"}}, |
| 201 | {4, {log_level::CRITICAL, "Fatal"}}}}, |
| 202 | {"SoC Availability", |
| 203 | {{1, {log_level::OK, "Enabled"}}, |
| 204 | {2, {log_level::WARNING, "Disabled"}}, |
Chau Ly | b01357f | 2024-10-17 09:18:01 +0000 | [diff] [blame] | 205 | {3, {log_level::CRITICAL, "Shutdown"}}}}}}, |
| 206 | {WATCH_DOG, |
| 207 | {{"Global Watch Dog", |
| 208 | {{1, {log_level::OK, "Normal"}}, |
| 209 | {2, {log_level::CRITICAL, "Timer Expired"}}}}, |
| 210 | {"Secure Watch Dog", |
| 211 | {{1, {log_level::OK, "Normal"}}, |
| 212 | {2, {log_level::CRITICAL, "Timer Expired"}}}}, |
| 213 | {"Non-secure Watch Dog", |
| 214 | {{1, {log_level::OK, "Normal"}}, |
| 215 | {2, {log_level::CRITICAL, "Timer Expired"}}}}}}}; |
Chau Ly | ef214b5 | 2024-10-16 09:40:38 +0000 | [diff] [blame] | 216 | |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 217 | std::string |
| 218 | OemEventManager::prefixMsgStrCreation(pldm_tid_t tid, uint16_t sensorId) |
| 219 | { |
| 220 | std::string description; |
| 221 | if (!tidToSocketNameMap.contains(tid)) |
| 222 | { |
| 223 | description += "TID " + std::to_string(tid) + ": "; |
| 224 | } |
| 225 | else |
| 226 | { |
| 227 | description += tidToSocketNameMap[tid] + ": "; |
| 228 | } |
| 229 | |
| 230 | if (!sensorIdToStrMap.contains(sensorId)) |
| 231 | { |
| 232 | description += "Sensor ID " + std::to_string(sensorId) + ": "; |
| 233 | } |
| 234 | else |
| 235 | { |
| 236 | description += sensorIdToStrMap[sensorId] + ": "; |
| 237 | } |
| 238 | |
| 239 | return description; |
| 240 | } |
| 241 | |
| 242 | void OemEventManager::sendJournalRedfish(const std::string& description, |
| 243 | log_level& logLevel) |
| 244 | { |
| 245 | if (description.empty()) |
| 246 | { |
| 247 | return; |
| 248 | } |
| 249 | |
| 250 | if (!logLevelToRedfishMsgIdMap.contains(logLevel)) |
| 251 | { |
| 252 | lg2::error("Invalid {LEVEL} Description {DES}", "LEVEL", logLevel, |
| 253 | "DES", description); |
| 254 | return; |
| 255 | } |
| 256 | auto redfishMsgId = logLevelToRedfishMsgIdMap[logLevel]; |
| 257 | lg2::info("MESSAGE={DES}", "DES", description, "REDFISH_MESSAGE_ID", |
| 258 | redfishMsgId, "REDFISH_MESSAGE_ARGS", description); |
| 259 | } |
| 260 | |
| 261 | std::string OemEventManager::dimmIdxsToString(uint32_t dimmIdxs) |
| 262 | { |
| 263 | std::string description; |
| 264 | for (const auto bitIdx : std::views::iota(0, maxDIMMIdxBitNum)) |
| 265 | { |
| 266 | if (dimmIdxs & (static_cast<uint32_t>(1) << bitIdx)) |
| 267 | { |
| 268 | description += " #" + std::to_string(bitIdx); |
| 269 | } |
| 270 | } |
| 271 | return description; |
| 272 | } |
| 273 | |
Thu Nguyen | 93d0ca3 | 2024-11-14 23:46:40 +0000 | [diff] [blame^] | 274 | uint8_t OemEventManager::sensorIdToDIMMIdx(const uint16_t& sensorId) |
| 275 | { |
| 276 | uint8_t dimmIdx = maxDIMMInstantNum; |
| 277 | int sensorId_Off = sensorId - 4; |
| 278 | if ((sensorId_Off >= 0) && ((sensorId_Off % 2) == 0) && |
| 279 | ((sensorId_Off / 2) < maxDIMMInstantNum)) |
| 280 | { |
| 281 | dimmIdx = sensorId_Off / 2; |
| 282 | } |
| 283 | return dimmIdx; |
| 284 | } |
| 285 | |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 286 | void OemEventManager::handleBootOverallEvent( |
| 287 | pldm_tid_t /*tid*/, uint16_t /*sensorId*/, uint32_t presentReading) |
| 288 | { |
| 289 | log_level logLevel{log_level::OK}; |
| 290 | std::string description; |
| 291 | std::stringstream strStream; |
| 292 | |
| 293 | uint8_t byte0 = (presentReading & 0x000000ff); |
| 294 | uint8_t byte1 = (presentReading & 0x0000ff00) >> 8; |
| 295 | uint8_t byte2 = (presentReading & 0x00ff0000) >> 16; |
| 296 | uint8_t byte3 = (presentReading & 0xff000000) >> 24; |
| 297 | /* |
| 298 | * Handle SECpro, Mpro, ATF BL1, ATF BL2, ATF BL31, |
| 299 | * ATF BL32 and DDR initialization |
| 300 | */ |
| 301 | if (bootStageToMsgMap.contains(byte3)) |
| 302 | { |
| 303 | // Boot stage adding |
| 304 | description += bootStageToMsgMap[byte3]; |
| 305 | |
| 306 | switch (byte3) |
| 307 | { |
| 308 | case boot_stage::DDR_TRAINING: |
| 309 | if (byte0 >= ddrTrainingMsg.size()) |
| 310 | { |
| 311 | logLevel = log_level::BIOSFWPANIC; |
| 312 | description += " unknown status"; |
| 313 | } |
| 314 | else |
| 315 | { |
| 316 | description += ddrTrainingMsg[byte0]; |
| 317 | } |
| 318 | if (0x01 == byte0) |
| 319 | { |
| 320 | // Add complete percentage |
| 321 | description += " at " + std::to_string(byte1) + "%"; |
| 322 | } |
| 323 | break; |
| 324 | case boot_stage::S0_DDR_TRAINING_FAILURE: |
| 325 | case boot_stage::S1_DDR_TRAINING_FAILURE: |
| 326 | // ddr_training_status_msg() |
| 327 | logLevel = log_level::BIOSFWPANIC; |
| 328 | description += " at DIMMs:"; |
| 329 | // dimmIdxs = presentReading & 0x00ffffff; |
| 330 | description += dimmIdxsToString(presentReading & 0x00ffffff); |
| 331 | description += " of socket "; |
| 332 | description += |
| 333 | (boot_stage::S0_DDR_TRAINING_FAILURE == byte3) ? "0" : "1"; |
| 334 | break; |
| 335 | default: |
| 336 | if (byte0 >= bootStatMsg.size()) |
| 337 | { |
| 338 | logLevel = log_level::BIOSFWPANIC; |
| 339 | description += " unknown status"; |
| 340 | } |
| 341 | else |
| 342 | { |
| 343 | description += bootStatMsg[byte0]; |
| 344 | } |
| 345 | break; |
| 346 | } |
| 347 | |
| 348 | // Sensor report action is fail |
| 349 | if (boot::status::BOOT_STATUS_FAILURE == byte2) |
| 350 | { |
| 351 | logLevel = log_level::BIOSFWPANIC; |
| 352 | } |
| 353 | } |
| 354 | else |
| 355 | { |
| 356 | if (byte3 <= boot_stage::UEFI_STATUS_CLASS_CODE_MAX) |
| 357 | { |
| 358 | description += |
| 359 | bootStageToMsgMap[boot_stage::UEFI_STATUS_CLASS_CODE_MIN]; |
| 360 | |
| 361 | strStream |
| 362 | << "Segment (0x" << std::setfill('0') << std::hex |
| 363 | << std::setw(8) << static_cast<uint32_t>(presentReading) |
Chau Ly | 3de0d94 | 2024-10-03 08:57:11 +0000 | [diff] [blame] | 364 | << "); Status Class (0x" << std::setw(2) |
| 365 | << static_cast<uint32_t>(byte3) << "); Status SubClass (0x" |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 366 | << std::setw(2) << static_cast<uint32_t>(byte2) |
Chau Ly | 3de0d94 | 2024-10-03 08:57:11 +0000 | [diff] [blame] | 367 | << "); Operation Code (0x" << std::setw(4) |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 368 | << static_cast<uint32_t>((presentReading & 0xffff0000) >> 16) |
| 369 | << ")" << std::dec; |
| 370 | |
| 371 | description += strStream.str(); |
| 372 | } |
| 373 | } |
| 374 | |
| 375 | // Log to Redfish event |
| 376 | sendJournalRedfish(description, logLevel); |
| 377 | } |
| 378 | |
| 379 | int OemEventManager::processNumericSensorEvent( |
| 380 | pldm_tid_t tid, uint16_t sensorId, const uint8_t* sensorData, |
| 381 | size_t sensorDataLength) |
| 382 | { |
| 383 | uint8_t eventState = 0; |
| 384 | uint8_t previousEventState = 0; |
| 385 | uint8_t sensorDataSize = 0; |
| 386 | uint32_t presentReading; |
| 387 | auto rc = decode_numeric_sensor_data( |
| 388 | sensorData, sensorDataLength, &eventState, &previousEventState, |
| 389 | &sensorDataSize, &presentReading); |
| 390 | if (rc) |
| 391 | { |
| 392 | lg2::error( |
| 393 | "Failed to decode numericSensorState event for terminus ID {TID}, error {RC} ", |
| 394 | "TID", tid, "RC", rc); |
| 395 | return rc; |
| 396 | } |
| 397 | |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 398 | // DIMMx_Status sensorID 4+2*index (index 0 -> maxDIMMInstantNum-1) |
Thu Nguyen | 93d0ca3 | 2024-11-14 23:46:40 +0000 | [diff] [blame^] | 399 | if (auto dimmIdx = sensorIdToDIMMIdx(sensorId); dimmIdx < maxDIMMInstantNum) |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 400 | { |
| 401 | handleDIMMStatusEvent(tid, sensorId, presentReading); |
| 402 | return PLDM_SUCCESS; |
| 403 | } |
| 404 | |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 405 | switch (sensorId) |
| 406 | { |
| 407 | case BOOT_OVERALL: |
| 408 | handleBootOverallEvent(tid, sensorId, presentReading); |
| 409 | break; |
Chau Ly | 3de0d94 | 2024-10-03 08:57:11 +0000 | [diff] [blame] | 410 | case PCIE_HOT_PLUG: |
| 411 | handlePCIeHotPlugEvent(tid, sensorId, presentReading); |
| 412 | break; |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 413 | case DDR_STATUS: |
| 414 | handleDDRStatusEvent(tid, sensorId, presentReading); |
| 415 | break; |
Chau Ly | 4cca3dc | 2024-10-03 09:07:09 +0000 | [diff] [blame] | 416 | case PCP_VR_STATE: |
| 417 | case SOC_VR_STATE: |
| 418 | case DPHY_VR1_STATE: |
| 419 | case DPHY_VR2_STATE: |
| 420 | case D2D_VR_STATE: |
| 421 | case IOC_VR1_STATE: |
| 422 | case IOC_VR2_STATE: |
| 423 | case PCI_D_VR_STATE: |
| 424 | case PCI_A_VR_STATE: |
| 425 | handleVRDStatusEvent(tid, sensorId, presentReading); |
| 426 | break; |
Chau Ly | b01357f | 2024-10-17 09:18:01 +0000 | [diff] [blame] | 427 | case WATCH_DOG: |
| 428 | handleNumericWatchdogEvent(tid, sensorId, presentReading); |
| 429 | break; |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 430 | default: |
| 431 | std::string description; |
| 432 | std::stringstream strStream; |
| 433 | log_level logLevel = log_level::OK; |
| 434 | |
| 435 | description += "SENSOR_EVENT : NUMERIC_SENSOR_STATE: "; |
| 436 | description += prefixMsgStrCreation(tid, sensorId); |
| 437 | strStream << std::setfill('0') << std::hex << "eventState 0x" |
| 438 | << std::setw(2) << static_cast<uint32_t>(eventState) |
| 439 | << " previousEventState 0x" << std::setw(2) |
| 440 | << static_cast<uint32_t>(previousEventState) |
| 441 | << " sensorDataSize 0x" << std::setw(2) |
| 442 | << static_cast<uint32_t>(sensorDataSize) |
| 443 | << " presentReading 0x" << std::setw(8) |
| 444 | << static_cast<uint32_t>(presentReading) << std::dec; |
| 445 | description += strStream.str(); |
| 446 | |
| 447 | sendJournalRedfish(description, logLevel); |
| 448 | break; |
| 449 | } |
| 450 | return PLDM_SUCCESS; |
| 451 | } |
| 452 | |
| 453 | int OemEventManager::processStateSensorEvent(pldm_tid_t tid, uint16_t sensorId, |
| 454 | const uint8_t* sensorData, |
| 455 | size_t sensorDataLength) |
| 456 | { |
| 457 | uint8_t sensorOffset = 0; |
| 458 | uint8_t eventState = 0; |
| 459 | uint8_t previousEventState = 0; |
| 460 | |
| 461 | auto rc = |
| 462 | decode_state_sensor_data(sensorData, sensorDataLength, &sensorOffset, |
| 463 | &eventState, &previousEventState); |
| 464 | if (rc) |
| 465 | { |
| 466 | lg2::error( |
| 467 | "Failed to decode stateSensorState event for terminus ID {TID}, error {RC}", |
| 468 | "TID", tid, "RC", rc); |
| 469 | return rc; |
| 470 | } |
| 471 | |
| 472 | std::string description; |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 473 | log_level logLevel = log_level::OK; |
| 474 | |
Chau Ly | ef214b5 | 2024-10-16 09:40:38 +0000 | [diff] [blame] | 475 | if (stateSensorToMsgMap.contains(sensorId)) |
| 476 | { |
| 477 | description += prefixMsgStrCreation(tid, sensorId); |
| 478 | auto componentMap = stateSensorToMsgMap[sensorId]; |
| 479 | if (sensorOffset < componentMap.size()) |
| 480 | { |
| 481 | description += std::get<0>(componentMap[sensorOffset]); |
| 482 | auto stateMap = std::get<1>(componentMap[sensorOffset]); |
| 483 | if (stateMap.contains(eventState)) |
| 484 | { |
| 485 | logLevel = std::get<0>(stateMap[eventState]); |
| 486 | description += " state : " + std::get<1>(stateMap[eventState]); |
| 487 | if (stateMap.contains(previousEventState)) |
| 488 | { |
| 489 | description += "; previous state: " + |
| 490 | std::get<1>(stateMap[previousEventState]); |
| 491 | } |
| 492 | } |
| 493 | else |
| 494 | { |
| 495 | description += " sends unsupported event state: " + |
| 496 | std::to_string(eventState); |
| 497 | if (stateMap.contains(previousEventState)) |
| 498 | { |
| 499 | description += "; previous state: " + |
| 500 | std::get<1>(stateMap[previousEventState]); |
| 501 | } |
| 502 | } |
| 503 | } |
| 504 | else |
| 505 | { |
| 506 | description += "sends unsupported component sensor offset " + |
| 507 | std::to_string(sensorOffset); |
| 508 | } |
| 509 | } |
| 510 | else |
| 511 | { |
| 512 | std::stringstream strStream; |
| 513 | description += "SENSOR_EVENT : STATE_SENSOR_STATE: "; |
| 514 | description += prefixMsgStrCreation(tid, sensorId); |
| 515 | strStream << std::setfill('0') << std::hex << "sensorOffset 0x" |
| 516 | << std::setw(2) << static_cast<uint32_t>(sensorOffset) |
| 517 | << "eventState 0x" << std::setw(2) |
| 518 | << static_cast<uint32_t>(eventState) |
| 519 | << " previousEventState 0x" << std::setw(2) |
| 520 | << static_cast<uint32_t>(previousEventState) << std::dec; |
| 521 | description += strStream.str(); |
| 522 | } |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 523 | |
| 524 | sendJournalRedfish(description, logLevel); |
| 525 | |
| 526 | return PLDM_SUCCESS; |
| 527 | } |
| 528 | |
| 529 | int OemEventManager::processSensorOpStateEvent( |
| 530 | pldm_tid_t tid, uint16_t sensorId, const uint8_t* sensorData, |
| 531 | size_t sensorDataLength) |
| 532 | { |
| 533 | uint8_t present_op_state = 0; |
| 534 | uint8_t previous_op_state = 0; |
| 535 | |
| 536 | auto rc = decode_sensor_op_data(sensorData, sensorDataLength, |
| 537 | &present_op_state, &previous_op_state); |
| 538 | if (rc) |
| 539 | { |
| 540 | lg2::error( |
| 541 | "Failed to decode sensorOpState event for terminus ID {TID}, error {RC}", |
| 542 | "TID", tid, "RC", rc); |
| 543 | return rc; |
| 544 | } |
| 545 | |
| 546 | std::string description; |
| 547 | std::stringstream strStream; |
| 548 | log_level logLevel = log_level::OK; |
| 549 | |
| 550 | description += "SENSOR_EVENT : SENSOR_OP_STATE: "; |
| 551 | description += prefixMsgStrCreation(tid, sensorId); |
| 552 | strStream << std::setfill('0') << std::hex << "present_op_state 0x" |
| 553 | << std::setw(2) << static_cast<uint32_t>(present_op_state) |
| 554 | << "previous_op_state 0x" << std::setw(2) |
| 555 | << static_cast<uint32_t>(previous_op_state) << std::dec; |
| 556 | description += strStream.str(); |
| 557 | |
| 558 | sendJournalRedfish(description, logLevel); |
| 559 | |
| 560 | return PLDM_SUCCESS; |
| 561 | } |
| 562 | |
| 563 | int OemEventManager::handleSensorEvent( |
| 564 | const pldm_msg* request, size_t payloadLength, uint8_t /* formatVersion */, |
| 565 | pldm_tid_t tid, size_t eventDataOffset) |
| 566 | { |
| 567 | /* This OEM event handler is only used for SoC terminus*/ |
| 568 | if (!tidToSocketNameMap.contains(tid)) |
| 569 | { |
| 570 | return PLDM_SUCCESS; |
| 571 | } |
| 572 | auto eventData = |
| 573 | reinterpret_cast<const uint8_t*>(request->payload) + eventDataOffset; |
| 574 | auto eventDataSize = payloadLength - eventDataOffset; |
| 575 | |
| 576 | uint16_t sensorId = 0; |
| 577 | uint8_t sensorEventClassType = 0; |
| 578 | size_t eventClassDataOffset = 0; |
| 579 | auto rc = |
| 580 | decode_sensor_event_data(eventData, eventDataSize, &sensorId, |
| 581 | &sensorEventClassType, &eventClassDataOffset); |
| 582 | if (rc) |
| 583 | { |
| 584 | lg2::error("Failed to decode sensor event data return code {RC}.", "RC", |
| 585 | rc); |
| 586 | return rc; |
| 587 | } |
| 588 | const uint8_t* sensorData = eventData + eventClassDataOffset; |
| 589 | size_t sensorDataLength = eventDataSize - eventClassDataOffset; |
| 590 | |
| 591 | switch (sensorEventClassType) |
| 592 | { |
| 593 | case PLDM_NUMERIC_SENSOR_STATE: |
| 594 | { |
| 595 | return processNumericSensorEvent(tid, sensorId, sensorData, |
| 596 | sensorDataLength); |
| 597 | } |
| 598 | case PLDM_STATE_SENSOR_STATE: |
| 599 | { |
| 600 | return processStateSensorEvent(tid, sensorId, sensorData, |
| 601 | sensorDataLength); |
| 602 | } |
| 603 | case PLDM_SENSOR_OP_STATE: |
| 604 | { |
| 605 | return processSensorOpStateEvent(tid, sensorId, sensorData, |
| 606 | sensorDataLength); |
| 607 | } |
| 608 | default: |
| 609 | std::string description; |
| 610 | std::stringstream strStream; |
| 611 | log_level logLevel = log_level::OK; |
| 612 | |
| 613 | description += "SENSOR_EVENT : Unsupported Sensor Class " + |
| 614 | std::to_string(sensorEventClassType) + ": "; |
| 615 | description += prefixMsgStrCreation(tid, sensorId); |
| 616 | strStream << std::setfill('0') << std::hex |
| 617 | << std::setw(sizeof(sensorData) * 2) << "Sensor data: "; |
| 618 | |
| 619 | auto dataPtr = sensorData; |
| 620 | for ([[maybe_unused]] const auto& i : |
| 621 | std::views::iota(0, (int)sensorDataLength)) |
| 622 | { |
| 623 | strStream << "0x" << static_cast<uint32_t>(*dataPtr); |
| 624 | dataPtr += sizeof(sensorData); |
| 625 | } |
| 626 | |
| 627 | description += strStream.str(); |
| 628 | |
| 629 | sendJournalRedfish(description, logLevel); |
| 630 | } |
| 631 | lg2::info("Unsupported class type {CLASSTYPE}", "CLASSTYPE", |
| 632 | sensorEventClassType); |
| 633 | return PLDM_ERROR; |
| 634 | } |
| 635 | |
Chau Ly | 3de0d94 | 2024-10-03 08:57:11 +0000 | [diff] [blame] | 636 | void OemEventManager::handlePCIeHotPlugEvent(pldm_tid_t tid, uint16_t sensorId, |
| 637 | uint32_t presentReading) |
| 638 | { |
| 639 | std::string description; |
| 640 | std::stringstream strStream; |
| 641 | PCIeHotPlugEventRecord_t record{presentReading}; |
| 642 | |
| 643 | std::string sAction = (!record.bits.action) ? "Insertion" : "Removal"; |
| 644 | std::string sOpStatus = (!record.bits.opStatus) ? "Successful" : "Failed"; |
| 645 | log_level logLevel = |
| 646 | (!record.bits.opStatus) ? log_level::OK : log_level::WARNING; |
| 647 | |
| 648 | description += prefixMsgStrCreation(tid, sensorId); |
| 649 | |
| 650 | strStream << "Segment (0x" << std::setfill('0') << std::hex << std::setw(2) |
| 651 | << static_cast<uint32_t>(record.bits.segment) << "); Bus (0x" |
| 652 | << std::setw(2) << static_cast<uint32_t>(record.bits.bus) |
| 653 | << "); Device (0x" << std::setw(2) |
| 654 | << static_cast<uint32_t>(record.bits.device) << "); Function (0x" |
| 655 | << std::setw(2) << static_cast<uint32_t>(record.bits.function) |
| 656 | << "); Action (" << sAction << "); Operation status (" |
| 657 | << sOpStatus << "); Media slot number (" << std::dec |
| 658 | << static_cast<uint32_t>(record.bits.mediaSlot) << ")"; |
| 659 | |
| 660 | description += strStream.str(); |
| 661 | |
| 662 | // Log to Redfish event |
| 663 | sendJournalRedfish(description, logLevel); |
| 664 | } |
| 665 | |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 666 | std::string OemEventManager::dimmTrainingFailureToMsg(uint32_t failureInfo) |
| 667 | { |
| 668 | std::string description; |
| 669 | DIMMTrainingFailure_t failure{failureInfo}; |
| 670 | |
| 671 | if (dimmTrainingFailureTypeMap.contains(failure.bits.type)) |
| 672 | { |
| 673 | auto failureInfoMap = dimmTrainingFailureTypeMap[failure.bits.type]; |
| 674 | |
| 675 | description += std::get<0>(failureInfoMap); |
| 676 | |
| 677 | description += "; MCU rank index " + |
| 678 | std::to_string(failure.bits.mcuRankIdx); |
| 679 | |
| 680 | description += "; Slice number " + |
| 681 | std::to_string(failure.bits.sliceNum); |
| 682 | |
| 683 | description += "; Upper nibble error status: "; |
| 684 | description += (!failure.bits.upperNibbStatErr) |
| 685 | ? "No error" |
| 686 | : "Found no rising edge"; |
| 687 | |
| 688 | description += "; Lower nibble error status: "; |
| 689 | description += (!failure.bits.lowerNibbStatErr) |
| 690 | ? "No error" |
| 691 | : "Found no rising edge"; |
| 692 | |
| 693 | description += "; Failure syndrome 0: "; |
| 694 | |
| 695 | auto& syndromeMap = std::get<1>(failureInfoMap); |
| 696 | if (syndromeMap.contains(failure.bits.syndrome)) |
| 697 | { |
| 698 | description += syndromeMap[failure.bits.syndrome]; |
| 699 | } |
| 700 | else |
| 701 | { |
| 702 | description += "(Unknown syndrome)"; |
| 703 | } |
| 704 | } |
| 705 | else |
| 706 | { |
| 707 | description += "Unknown training failure type " + |
| 708 | std::to_string(failure.bits.type); |
| 709 | } |
| 710 | |
| 711 | return description; |
| 712 | } |
| 713 | |
| 714 | void OemEventManager::handleDIMMStatusEvent(pldm_tid_t tid, uint16_t sensorId, |
| 715 | uint32_t presentReading) |
| 716 | { |
| 717 | log_level logLevel{log_level::WARNING}; |
| 718 | std::string description; |
| 719 | uint8_t byte3 = (presentReading & 0xff000000) >> 24; |
| 720 | uint32_t byte012 = presentReading & 0xffffff; |
| 721 | |
| 722 | description += prefixMsgStrCreation(tid, sensorId); |
| 723 | |
Thu Nguyen | 93d0ca3 | 2024-11-14 23:46:40 +0000 | [diff] [blame^] | 724 | // DIMMx_Status sensorID 4+2*index (index 0 -> maxDIMMInstantNum-1) |
| 725 | auto dimmIdx = sensorIdToDIMMIdx(sensorId); |
| 726 | if (dimmIdx >= maxDIMMIdxBitNum) |
| 727 | { |
| 728 | return; |
| 729 | } |
Chau Ly | cebf476 | 2024-10-03 09:02:54 +0000 | [diff] [blame] | 730 | |
| 731 | description += "DIMM " + std::to_string(dimmIdx) + " "; |
| 732 | |
| 733 | if (dimmStatusToMsgMap.contains(byte3)) |
| 734 | { |
| 735 | if (byte3 == dimm_status::INSTALLED_NO_ERROR || |
| 736 | byte3 == dimm_status::INSTALLED_BUT_DISABLED) |
| 737 | { |
| 738 | logLevel = log_level::OK; |
| 739 | } |
| 740 | |
| 741 | description += dimmStatusToMsgMap[byte3]; |
| 742 | |
| 743 | if (byte3 == dimm_status::TRAINING_FAILURE) |
| 744 | { |
| 745 | description += "; " + dimmTrainingFailureToMsg(byte012); |
| 746 | } |
| 747 | else if (byte3 == dimm_status::PMIC_TEMP_ALERT) |
| 748 | { |
| 749 | uint8_t byte0 = (byte012 & 0xff); |
| 750 | if (byte0 < pmicTempAlertMsg.size()) |
| 751 | { |
| 752 | description += ": " + pmicTempAlertMsg[byte0]; |
| 753 | } |
| 754 | } |
| 755 | } |
| 756 | else |
| 757 | { |
| 758 | switch (byte3) |
| 759 | { |
| 760 | case dimm_status::PMIC_HIGH_TEMP: |
| 761 | if (byte012 == 0x01) |
| 762 | { |
| 763 | description += "has PMIC high temp condition"; |
| 764 | } |
| 765 | break; |
| 766 | case dimm_status::TSx_HIGH_TEMP: |
| 767 | switch (byte012) |
| 768 | { |
| 769 | case 0x01: |
| 770 | description += "has TS0"; |
| 771 | break; |
| 772 | case 0x02: |
| 773 | description += "has TS1"; |
| 774 | break; |
| 775 | case 0x03: |
| 776 | description += "has TS0 and TS1"; |
| 777 | break; |
| 778 | } |
| 779 | description += " exceeding their high temperature threshold"; |
| 780 | break; |
| 781 | case dimm_status::SPD_HUB_HIGH_TEMP: |
| 782 | if (byte012 == 0x01) |
| 783 | { |
| 784 | description += "has SPD/HUB high temp condition"; |
| 785 | } |
| 786 | break; |
| 787 | default: |
| 788 | description += "has unsupported status " + |
| 789 | std::to_string(byte3); |
| 790 | break; |
| 791 | } |
| 792 | } |
| 793 | |
| 794 | // Log to Redfish event |
| 795 | sendJournalRedfish(description, logLevel); |
| 796 | } |
| 797 | |
| 798 | void OemEventManager::handleDDRStatusEvent(pldm_tid_t tid, uint16_t sensorId, |
| 799 | uint32_t presentReading) |
| 800 | { |
| 801 | log_level logLevel{log_level::WARNING}; |
| 802 | std::string description; |
| 803 | uint8_t byte3 = (presentReading & 0xff000000) >> 24; |
| 804 | uint32_t byte012 = presentReading & 0xffffff; |
| 805 | |
| 806 | description += prefixMsgStrCreation(tid, sensorId); |
| 807 | |
| 808 | description += "DDR "; |
| 809 | if (ddrStatusToMsgMap.contains(byte3)) |
| 810 | { |
| 811 | if (byte3 == ddr_status::NO_SYSTEM_LEVEL_ERROR) |
| 812 | { |
| 813 | logLevel = log_level::OK; |
| 814 | } |
| 815 | |
| 816 | description += ddrStatusToMsgMap[byte3]; |
| 817 | |
| 818 | if (byte3 == ddr_status::CONFIGURATION_FAILURE || |
| 819 | byte3 == ddr_status::TRAINING_FAILURE) |
| 820 | { |
| 821 | // List out failed DIMMs |
| 822 | description += dimmIdxsToString(byte012); |
| 823 | } |
| 824 | } |
| 825 | else |
| 826 | { |
| 827 | description += "has unsupported status " + std::to_string(byte3); |
| 828 | } |
| 829 | |
| 830 | // Log to Redfish event |
| 831 | sendJournalRedfish(description, logLevel); |
| 832 | } |
| 833 | |
Chau Ly | 4cca3dc | 2024-10-03 09:07:09 +0000 | [diff] [blame] | 834 | void OemEventManager::handleVRDStatusEvent(pldm_tid_t tid, uint16_t sensorId, |
| 835 | uint32_t presentReading) |
| 836 | { |
| 837 | log_level logLevel{log_level::WARNING}; |
| 838 | std::string description; |
| 839 | std::stringstream strStream; |
| 840 | |
| 841 | description += prefixMsgStrCreation(tid, sensorId); |
| 842 | |
| 843 | VRDStatus_t status{presentReading}; |
| 844 | |
| 845 | if (status.bits.warning && status.bits.critical) |
| 846 | { |
| 847 | description += "A VR warning and a VR critical"; |
| 848 | logLevel = log_level::CRITICAL; |
| 849 | } |
| 850 | else |
| 851 | { |
| 852 | if (status.bits.warning) |
| 853 | { |
| 854 | description += "A VR warning"; |
| 855 | } |
| 856 | else if (status.bits.critical) |
| 857 | { |
| 858 | description += "A VR critical"; |
| 859 | logLevel = log_level::CRITICAL; |
| 860 | } |
| 861 | else |
| 862 | { |
| 863 | description += "No VR warning or critical"; |
| 864 | logLevel = log_level::OK; |
| 865 | } |
| 866 | } |
| 867 | description += " condition observed"; |
| 868 | |
| 869 | strStream << "; VR status byte high is 0x" << std::setfill('0') << std::hex |
| 870 | << std::setw(2) |
| 871 | << static_cast<uint32_t>(status.bits.vr_status_byte_high) |
| 872 | << "; VR status byte low is 0x" << std::setw(2) |
| 873 | << static_cast<uint32_t>(status.bits.vr_status_byte_low) |
| 874 | << "; Reading is 0x" << std::setw(2) |
| 875 | << static_cast<uint32_t>(presentReading) << ";"; |
| 876 | |
| 877 | description += strStream.str(); |
| 878 | |
| 879 | // Log to Redfish event |
| 880 | sendJournalRedfish(description, logLevel); |
| 881 | } |
| 882 | |
Chau Ly | b01357f | 2024-10-17 09:18:01 +0000 | [diff] [blame] | 883 | void OemEventManager::handleNumericWatchdogEvent( |
| 884 | pldm_tid_t tid, uint16_t sensorId, uint32_t presentReading) |
| 885 | { |
| 886 | std::string description; |
| 887 | log_level logLevel = log_level::CRITICAL; |
| 888 | |
| 889 | description += prefixMsgStrCreation(tid, sensorId); |
| 890 | |
| 891 | if (presentReading & 0x01) |
| 892 | { |
| 893 | description += "Global watchdog expired;"; |
| 894 | } |
| 895 | if (presentReading & 0x02) |
| 896 | { |
| 897 | description += "Secure watchdog expired;"; |
| 898 | } |
| 899 | if (presentReading & 0x04) |
| 900 | { |
| 901 | description += "Non-secure watchdog expired;"; |
| 902 | } |
| 903 | |
| 904 | // Log to Redfish event |
| 905 | sendJournalRedfish(description, logLevel); |
| 906 | } |
| 907 | |
Chau Ly | a743e38 | 2024-10-26 11:12:22 +0000 | [diff] [blame] | 908 | } // namespace oem_ampere |
| 909 | } // namespace pldm |