RAS--APEI 报错解析流程(1)
除了APEI 中除了GHES会记录错误，在Post过程中的错误通常是通过BERT Table汇报
1.BERT
Boot Error Record Table is used to report unhandled errors that occurred in a previous boot，it is reported as a ‘one-time polled’type error source.
Bert 用于记录post 过程中产生的 error 以及UCE hang 重启 BIOS 错误状态未进行清除在下次重启扫描出的error
具体结构如图所示：
使用BERT 的header 通过section type 区分错误类型 对应到不同的错误结构 都是通过block error status addrss 链接
内存 pcie cpu的错误汇报信息结构体
使用IASL 解析BERT table:
BIOS 在Post 过程中去扫描pcie error 内存 error CPU error
NBIOErrorDetection 检测到错误 addbert ->GENERIC_PCIE_AER_ERR_ENTRY
MCAErrorDetection 检测到MCA Bank UMC 错误 addbert ->GENERIC_MEM_ERR_ENTRY
MCAErrorDetection 检测到MCA Bank PIE 错误 addbert ->GENERIC_PRO_ERR_ENTRY
CPU BERT OS 解析
内存OS 解析
PCIE OS 解析
Post过程中检测到多个BERT ：有内存和CPU bank 的错误状态
BIOS 日志：两个MCA Bank 读取到错误MCA_Status
BERT Table: BIOS 汇报和OS 解析通过Boot Error Region Address 联系
Boot Error Region Address 对应到结构体
///
/// Generic Error Status Definition
///
typedef struct {
EFI_ACPI_6_2_ERROR_BLOCK_STATUS BlockStatus;
UINT32 RawDataOffset;
UINT32 RawDataLength;
UINT32 DataLength;
UINT32 ErrorSeverity;
} EFI_ACPI_6_2_GENERIC_ERROR_STATUS_STRUCTURE;
对于OS去打印HardWare error 只需要参考GHES 中的Error Block status 就会去打印 错误的GHES Table中的错误信息
typedef struct {
UINT32 UncorrectableErrorValid : 1;
UINT32 CorrectableErrorValid : 1;
UINT32 MultipleUncorrectableErrors : 1;
UINT32 MultipleCorrectableErrors : 1;
UINT32 ErrorDataEntryCount : 10;
UINT32 Reserved : 18;
} EFI_ACPI_6_2_ERROR_BLOCK_STATUS;
也就是当扫描到BlockStatus 存在错误状态 OS就会上报Hardware error 然后清除错误状态
后面接着结构体，后面的结构体会根据SectionType 接着内存/CPU/PCIE 的结构体
typedef struct {
UINT8 SectionType[16];
UINT32 ErrorSeverity;
UINT16 Revision;
UINT8 ValidationBits;
UINT8 Flags;
UINT32 ErrorDataLength;
UINT8 FruId[16];
UINT8 FruText[20];
UINT8 Timestamp[8];
} EFI_ACPI_6_2_GENERIC_ERROR_DATA_ENTRY_STRUCTURE;
/*
* Section type definitions, used in section_type field in struct
* cper_section_descriptor
*
* Processor Generic
*/
#define CPER_SEC_PROC_GENERIC \
GUID_INIT(0x9876CCAD, 0x47B4, 0x4bdb, 0xB6, 0x5E, 0x16, 0xF1, \
0x93, 0xC4, 0xF3, 0xDB)
/* Processor Specific: X86/X86_64 */
#define CPER_SEC_PROC_IA \
GUID_INIT(0xDC3EA0B0, 0xA144, 0x4797, 0xB9, 0x5B, 0x53, 0xFA, \
0x24, 0x2B, 0x6E, 0x1D)
/* Processor Specific: IA64 */
#define CPER_SEC_PROC_IPF \
GUID_INIT(0xE429FAF1, 0x3CB7, 0x11D4, 0x0B, 0xCA, 0x07, 0x00, \
0x80, 0xC7, 0x3C, 0x88, 0x81)
/* Processor Specific: ARM */
#define CPER_SEC_PROC_ARM \
GUID_INIT(0xE19E3D16, 0xBC11, 0x11E4, 0x9C, 0xAA, 0xC2, 0x05, \
0x1D, 0x5D, 0x46, 0xB0)
/* Platform Memory */
#define CPER_SEC_PLATFORM_MEM \
GUID_INIT(0xA5BC1114, 0x6F64, 0x4EDE, 0xB8, 0x63, 0x3E, 0x83, \
0xED, 0x7C, 0x83, 0xB1)
#define CPER_SEC_PCIE \
GUID_INIT(0xD995E954, 0xBBC1, 0x430F, 0xAD, 0x91, 0xB4, 0x4D, \
0xCB, 0x3C, 0x6F, 0x35)
/* Firmware Error Record Reference */
Section Type=CPER_SEC_PROC_GENERIC 对应到 CPU 的错误结构体
对于GHES 的错误OS需要使用定时器，BERT 只需要在Kernel 加载时跑一边即可。
OS 下错误解析 ghes.c bert.c cper.c
GHES 驱动：
static struct platform_driver ghes_platform_driver = {
.driver = {
.name = "GHES",
},
.probe = ghes_probe,
.remove = ghes_remove,
};
ghes_init 加载GHES的驱动 ，系统下的解析策略和Notify 的结构体相关联，BIOS中会设置Notify Type ,Pollinterval
系统下扫描GHES 的驱动 是通过定时器周期性去扫描错误状态，Pollinterval 是定时器的参考时间
switch (generic->notify.type) {
case ACPI_HEST_NOTIFY_POLLED:
case ACPI_HEST_NOTIFY_EXTERNAL:
case ACPI_HEST_NOTIFY_SCI:
case ACPI_HEST_NOTIFY_GSIV:
case ACPI_HEST_NOTIFY_GPIO:
break;
case ACPI_HEST_NOTIFY_SEA:
if (!IS_ENABLED(CONFIG_ACPI_APEI_SEA)) {
pr_warn(GHES_PFX "Generic hardware error source: %d notified via SEA is not supported\n",
generic->header.source_id);
rc = -ENOTSUPP;
goto err;
}
break;
case ACPI_HEST_NOTIFY_NMI:
if (!IS_ENABLED(CONFIG_HAVE_ACPI_APEI_NMI)) {
pr_warn(GHES_PFX "Generic hardware error source: %d notified via NMI interrupt is not supported!\n",
generic->header.source_id);
goto err;
}
break;
case ACPI_HEST_NOTIFY_LOCAL:
pr_warning(GHES_PFX "Generic hardware error source: %d notified via local interrupt is not supported!\n",
generic->header.source_id);
goto err;
timer_setup(&ghes->timer, ghes_poll_func, TIMER_DEFERRABLE);
ghes_proc(ghes);
ghes_read_estatus(ghes, 0);//-->apei_read(&buf_paddr, &g->error_status_address);
ghes_print_estatus
cper_estatus_print(pfx_seq, estatus);
这就对应到Dmesg 中的HardWare error 错误，就可以识别到错误的source id ，既可以大致定位错误信息来源 Source id = 512 对应到PCIE 错误
后续通过Section error type 定位到更加详细的信息
前面的信息都是来自于固定结构体
///
/// Generic Error Data Entry Definition
///
typedef struct {
UINT8 SectionType[16];
UINT32 ErrorSeverity;
UINT16 Revision;
UINT8 ValidationBits;
UINT8 Flags;
UINT32 ErrorDataLength;
UINT8 FruId[16];
UINT8 FruText[20];
UINT8 Timestamp[8];
} EFI_ACPI_6_2_GENERIC_ERROR_DATA_ENTRY_STRUCTURE;
static void
cper_estatus_print_section(const char *pfx, struct acpi_hest_generic_data *gdata,
int sec_no)
{
guid_t *sec_type = (guid_t *)gdata->section_type;
__u16 severity;
char newpfx[64];
if (acpi_hest_get_version(gdata) >= 3)
cper_print_tstamp(pfx, (struct acpi_hest_generic_data_v300 *)gdata);
severity = gdata->error_severity;
printk("%s""Error %d, type: %s\n", pfx, sec_no,
cper_severity_str(severity));
if (gdata->validation_bits & CPER_SEC_VALID_FRU_ID)
printk("%s""fru_id: %pUl\n", pfx, gdata->fru_id);
if (gdata->validation_bits & CPER_SEC_VALID_FRU_TEXT)
printk("%s""fru_text: %.20s\n", pfx, gdata->fru_text);
}
对于后续的结构体包括内存PCIE CPU 有着不同的结构体主要包括三个函数打印错误信息，通过匹配Section type Guid 判断
cper_print_proc_generic(); cper_print_mem(); cper_print_pcie()
if (guid_equal(sec_type, &CPER_SEC_PROC_GENERIC)) {
struct cper_sec_proc_generic *proc_err = acpi_hest_get_payload(gdata);
printk("%s""section_type: general processor error\n", newpfx);
if (gdata->error_data_length >= sizeof(*proc_err))
cper_print_proc_generic(newpfx, proc_err);
else
goto err_section_too_small;
} else if (guid_equal(sec_type, &CPER_SEC_PLATFORM_MEM)) {
struct cper_sec_mem_err *mem_err = acpi_hest_get_payload(gdata);
printk("%s""section_type: memory error\n", newpfx);
if (gdata->error_data_length >=
sizeof(struct cper_sec_mem_err_old))
cper_print_mem(newpfx, mem_err,
gdata->error_data_length);
else
goto err_section_too_small;
} else if (guid_equal(sec_type, &CPER_SEC_PCIE)) {
struct cper_sec_pcie *pcie = acpi_hest_get_payload(gdata);
printk("%s""section_type: PCIe error\n", newpfx);
if (gdata->error_data_length >= sizeof(*pcie))
cper_print_pcie(newpfx, pcie, gdata);
else
goto err_section_too_small;
#if defined(CONFIG_ARM64) || defined(CONFIG_ARM)
} else if (guid_equal(sec_type, &CPER_SEC_PROC_ARM)) {
struct cper_sec_proc_arm *arm_err = acpi_hest_get_payload(gdata);
printk("%ssection_type: ARM processor error\n", newpfx);
if (gdata->error_data_length >= sizeof(*arm_err))
cper_print_proc_arm(newpfx, arm_err);
else
goto err_section_too_small;
CPU错误信息结构体：
typedef struct _PLATFORM_PROC_ERR_SEC {
PROC_ERR_VALID_BIT ValidBits; ///< Validation Bits
UINT64 LocalApicID; ///< Processor APIC ID
UINT64 CpuIdInfo_EAX; ///< CPUID Information output value from EAX
UINT64 CpuIdInfo_EBX; ///< CPUID Information output value from EBX
UINT64 CpuIdInfo_ECX; ///< CPUID Information output value from ECX
UINT64 CpuIdInfo_EDX; ///< CPUID Information output value from EDX
UINT64 CpuIdInfo_PD1; ///< CPUID Information Padding 1
UINT64 CpuIdInfo_PD2; ///< CPUID Information Padding 2
} PLATFORM_PROC_ERR_SEC;
内存错误信息结构体：
typedef struct _PLATFORM_MEM_ERR_SEC {
MEM_ERR_VALID_BIT ValidBits; ///< Valid bits Bitmp
UINT64 ErrStatus; ///< Error Status
UINT64 PhyAddr; ///< Physical memory address of detected error
UINT64 PhyAddrMask; ///< Physical Error Address mask
UINT16 Node; ///< Node Number
UINT16 Card; ///< Card Number
UINT16 Module; ///< Module Number
UINT16 Bank; ///< Bank Number
UINT16 Device; ///< Device Number
UINT16 Row; ///< Row Number
UINT16 Column; ///< Column Number
UINT16 BitPosition; ///< Bit Position
UINT64 RequestorID; ///< Requestor ID
UINT64 ResponderID; ///< Responder ID
UINT64 TargetID; ///< Target ID
UINT8 MemErrType; ///< Memory Error Type
UINT8 Extend; ///< Extened
UINT16 RankNumber; ///< Rank Number
UINT16 CardHandle; ///< Card Number
UINT16 ModuleHandle;///< Module Number
} PLATFORM_MEM_ERR_SEC;
PCIE 错误信息结构体：
///
/// PCIE Error Section
///
typedef struct {
PCIE_ERR_VALID_BIT Validation; ///< Validation Bits
UINT32 PortType; ///< Port Type
UINT32 Revision; ///< Revision
UINT32 CommandStatus; ///< Command Status
UINT32 Reserved; ///< Reserved
DEVICE_ID DeviceId; ///< Device Id
UINT8 SerialNum[8]; ///< Serial Num
UINT32 BridgeCtrlStatus; ///< Bridge Control Status
CAP_STRUCTURE CapabilityStructure; ///< Capability Structure
AER_INFO AerInfo; ///< AER Info
} PCIE_ERROR_SECTION;
对于BERT 错误不需要使用定时器在Kernel 加载的时候会去初始化一遍BERT
BERT \drivers\acpi\apei\bert.c
bert_init -->
pr_info_once("Error records from previous boot:\n");
bert_print_all(boot_error_region, region_len);
cper_estatus_print(KERN_INFO HW_ERR, estatus);
错误解析和HEST GHES 类似
