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