Armv8-A架构RAS特性详解

需积分: 0 164 浏览量更新于2024-07-10 收藏 1.52MB PDF 举报

"DDI0587D_b_RAS_Supplement.pdf 是一份关于 Armv8-A 架构的可靠性、可用性和服务性（RAS）的技术文档，由 Arm Limited 或其关联公司版权所有。该文档是 Arm DDI0587D 的第 .b 版本，详细介绍了 Armv8-A 架构在 RAS 方面的设计和功能，旨在提高处理器系统的稳定性和效率。文档经历了多次更新，最近的一次是在2021年1月25日，加入了 v8.6 Beta 版本的内容。" 《Arm RAS 补充指南》是针对 Armv8-A 架构的可靠性、可用性和服务性的详细技术参考，涵盖了以下关键知识点： 1. **可靠性（Reliability）**：这是指系统在运行过程中保持正常工作的能力。在 Armv8-A 中，RAS 功能可能包括错误检测、错误校正码（ECC）、故障隔离和故障恢复机制等，以确保即使在硬件故障发生时，系统仍能持续运行。 2. **可用性（Availability）**：指的是系统能够持续提供服务的时间比例。这涉及到系统设计中的故障预防、快速故障检测和快速恢复机制。例如，通过热插拔支持、冗余组件和故障切换策略来最大化系统的无故障运行时间。 3. **服务性（Serviceability）**：是指系统在出现问题时进行维护和修复的难易程度。这通常涉及到诊断工具、远程管理接口以及便于更换故障组件的设计。Arm RAS 提供了增强的服务性功能，以便于系统管理员进行故障排查和维护。 4. **错误处理与报告**：文档详细描述了 Armv8-A 处理器如何处理硬件错误，如内存错误、指令错误等，并提供错误报告机制，帮助开发者和系统管理员识别和解决问题。 5. **ECC 支持**：在 Armv8-A 架构中，ECC（Error Correction Code）被用来检测和纠正内存中的数据错误，提高了数据的完整性。 6. **硬件故障管理**：RAS 设计包括了对硬件故障的管理和响应机制，例如故障隔离、故障透明切换等，以确保系统在单个组件失败时不会整体崩溃。 7. **软件接口**：文档还可能包含与 RAS 相关的软件接口规范，如系统管理中断（SMI）、故障报告结构和控制寄存器的访问等，这些接口允许软件参与 RAS 机制，提高整个系统的可靠性。 8. **版本更新**：每个版本的更新都带来了新的特性和改进，比如 v8.6 Beta 版本可能引入了新的 RAS 功能或优化了现有机制。这份文档对于设计和维护基于 Armv8-A 架构的高可用性和高可靠性的服务器、嵌入式系统或者数据中心解决方案的工程师来说，是非常宝贵的参考资料。它提供了深入理解 Armv8-A RAS 功能的途径，有助于提升系统整体性能和稳定性。

Chapter 1. Introduction to RAS

1.1. Faults, errors, and failures

1.1 Faults, errors, and failures

NVNNC

Correct service is delivered when the service implements the system function.

QWSVK

Correct service might include:

• Producing correct results.

• Producing results within the time allotted to the task.

• Not divulging secret or secure information.

For the purpose of describing the RAS Extension and RAS System Architecture, deviation from correct service

is deﬁned using the following terms:

NSPJY

•

A failure is the event of deviation from correct service. This includes data corruption, data loss, and

service loss.

SCKWX

• An error is the deviation from correct service. An incorrect value that has an error is corrupt.

YRDDR

• A fault is the cause of the error.

JNBDX

Errors that are present but not detected are latent errors or undetected errors.

TNQPK

In a system with no error detection, all errors are latent errors and are silently propagated by components until

they are either masked or cause failure.

GRYKV

The severity of a failure can range from minor to catastrophic:

•

The harmful consequences of a minor failure are of a similar cost to the beneﬁts provided by correct

service delivery.

•

The harmful consequences of a catastrophic failure are orders of magnitude, or even incommensurably,

higher than the beneﬁt provided by correct service delivery.

NMGPQ

There are many sources of faults in a system, including both software and hardware faults:

• Hardware faults originate in, or affect, hardware.

• Software faults affect software, that is programs or data.

The RAS Extension and RAS System Architecture primarily address errors produced from hardware faults.

These fall into two main areas:

• Transient faults.

• Non-transient or persistent faults.

ARM DDI 0587

D.b-00bet1

Non-conﬁdential

Chapter 1. Introduction to RAS

1.2. General taxonomy of errors

1.2 General taxonomy of errors

1.2.1 Error detection

FHXWP

When a component accesses memory or other state, an error might be detected in that memory or state.

WKPVR

The error might be corrected or deferred by the component, or signaled to another component as either a deferred

error or a detected error.

1.2.2 Error propagation

LRZDN

A transaction occurs when a producer of the transaction passes a value or other signal to a consumer of the

transaction.

VYCCX

Transactions are part of the service provided by the producer for the consumer.

RHGDL

In many protocols and service interface deﬁnitions, a high-level transaction consists of a sequence of operations,

for instance between a Requester and a Completer.

For the purposes of this manual, the most basic form of a unidirectional transfer between a producer and

consumer is considered as a transaction.

That is, each one of the sequence of operations is considered a separate transaction. For some operations, such

as a request, the Requester is producer and the Completer is the consumer. For other operations, such as a

response, the Completer is producer and the Requester is the consumer.

SKZZG

An error is propagated by the producer of a transaction when the service interface is incorrect because of the

error. The error is propagated to the consumer.

An error is propagated by deviations from correct service, including when any of the following occurs that

would not have been permitted to occur had the fault not been activated:

XDHGD

• A corrupt value is passed from producer to consumer.

ZCNXB

• A transaction or other operation occurs that should not have occurred.

CFZKP

• A transaction or other operation that should have occurred does not occur.

MZVNK

•

A loss of uniprocessor semantics or any other loss of coherency in a multiprocessor coherent system is

observed.

KCDXV

•

Changing the timing and/or order of transactions or other operations such that the timing and/or order of

those transactions or operations is incorrect. In this case the service interface deﬁnes acceptable timings

and/or orders for transactions and other operations.

The service interface for a transaction might include means to signal that the transaction is propagating:

VVFYS

• A detected error.

CSVRC

• A deferred error.

BHWVX

An error is silently propagated by the producer of a transaction if the consumer of the transaction cannot detect

the error and consumes an undetected error because of the transaction. This might be because of one of the

following:

•

The error is present on the transaction, but was not detected by the producer. The error is silently

propagated by the producer.

•

The error is present on the transaction, but was not signaled to the consumer as an error. For example, a

corrupt value was passed in the transaction with no indication that it was corrupt. The error is silently

propagated by the producer.

ARM DDI 0587

D.b-00bet1

Non-conﬁdential

Chapter 1. Introduction to RAS

1.2. General taxonomy of errors

FPBYS

A latent, possibly detectable, error is silently propagated by the consumer of an otherwise correct transaction if

the transaction causes the error to become undetectable.

XYVCP

For example, a partial write to a protection granule that removes poison, leaving the unchanged portion of the

location corrupt. To implement a partial write, the consumer logically reads the current value of the location,

modiﬁes the value, and then writes the modiﬁed value back. These are internal transactions in the consumer that

silently propagate the error. In this example there was no error at the producer nor on the transaction.

Errors might be propagated by components in a system until one of the following occurs:

YZTDY

• They are masked and do not affect the outcome of the system.

The error might be masked because a corrupt value is discarded or overwritten, or the error is detected and

removed.

VQZPT

•

They affect the service interface of the system and possibly cause failure. If the error has been silently

propagated to the service interface then:

– This is a Silent Data Corruption (SDC).

–

The rate of such failures, measured as the number of failures per billion device-hours of operation, is

called the SDC Failure-in-Time (FIT) rate.

Alternatively, the error might have been detected, causing the system to invoke error handling and recovery.

See Error handling and recovery.

1.2.3 Infected and poisoned

KNHWB

The state of a component becomes infected when the component consumes an uncorrected error that updates

the state.

TZBSW

A value is poisoned in the state of a component if it is marked as being in error, such that a subsequent access of

the state will detect the value is so marked and is treated as a detected error.

YBMFK

Poison is used to defer an error.

1.2.4 Containable and uncontainable

DXQRD

An undetected error is uncontained at the component that failed to detect it.

RJYRQ

A silently propagated error is uncontained at the component that silently propagated it.

GJQNR

A detected uncorrected error is uncontainable at the component if it might be uncontained at the component.

A detected uncorrected error is containable at the component if it is not uncontainable at the component. If

the component cannot determine whether a detected uncorrected error is uncontainable or containable at the

component, it treats it as uncontainable at the component.

MRDMR

An error that is uncontainable at a component might be containable at the system level.

Note

Reporting an error as containable allows software to contain the error. It does not mean that hardware has

contained the error.

ARM DDI 0587

D.b-00bet1

Non-conﬁdential

Chapter 1. Introduction to RAS

1.3. Techniques for improving reliability, availability, and serviceability

1.3 Techniques for improving reliability, availability, and serviceability

TPGKF

Each device sets its own targets for reliability, availability, and serviceability, using various techniques to achieve

these targets, including:

• Fault prevention and fault removal.

• Error handling and recovery.

• Fault handling.

DMKGY

The level of reliability, availability, and serviceability in any implementation, and which parts of the system

include RAS, are IMPLEMENTATION DEFINED. The RAS Extension and RAS System Architecture do not

prescribe the level of reliability, availability, and serviceability in any implementation, or which parts of the

system include RAS.

1.3.1 Fault prevention and fault removal

YLVTS

Fault prevention and fault removal are two techniques for handling faults. Fault prevention and fault removal

mechanisms are IMPLEMENTATION DEFINED.

WZTKF

Fault prevention techniques are outside the scope of the architecture.

JVLNC

A fault that is removed is a corrected error and might be recorded and generate a fault handling interrupt, but it

is not propagated. This means that it is not consumed and does not cause service failure.

WSPBC

A common technique to detect and correct errors is the use of an Error Detection and Correction Code (EDAC),

more commonly referred to as simply an Error Correction Code (ECC). ECC schemes use mathematical codes

to detect and correct an error in a value in memory. The size of the value is the protection granule for the ECC

scheme.

PBJLC

The RAS Extension and RAS System Architecture do not require implementation any fault removal schemes,

including ECC.

1.3.2 Error handling and recovery

XPLVT

A fault that is not removed gives rise to an uncorrected error.

VTXYY

Error recovery is the process by which software and hardware minimize the impact of an uncorrected error.

Error recovery methods include:

DCGYX

•

Deferring an error from a fault. An error is deferred by hardware if hardware can make forward progress

without consuming the error. Deferring the error means:

–

The fault might become masked later (fault removal). For example, because the corrupt value is

overwritten before it is consumed.

–

If the deferred error is later consumed, then the error is reported at the point of consumption. For

example, if the deferred error is consumed by a Processing element (PE) then the consumer PE

generates an error exception. This can give better results in terms of error recovery in the case where

the original producer of the data is not known when the error was deferred. For example because a

latent error was detected.

A common technique to defer an error is to replace the corrupt value with a poisoned value, for example in

memory or in a transaction.

YLMTV

•

Preventing further propagation of the error, that is containing the error. In particular, preventing silent

propagation of the error.

RFLCM

• Reducing the severity of a failure by invoking a service failure mode:

ARM DDI 0587

D.b-00bet1

Non-conﬁdential

Chapter 1. Introduction to RAS

1.3. Techniques for improving reliability, availability, and serviceability

– This is a Detected Uncorrected Error (DUE).

– The rate of such failures gives the DUE FIT rate.

– The type of service failure mode depends on what is acceptable to the service.

BRDMK

A software error recovery agent is typically invoked when hardware detects an error it cannot correct, defer, or

remove.

PGXFK

An error recovery agent also provides information to the operator through error logs to improve serviceability,

for example to help with the identiﬁcation of a Field Replaceable Unit (FRU).

MFPRY

The RAS Extension and RAS System Architecture provide optional common programmers’ models to record

information about an error in an error record.

CVFFN

The RAS Extension describes the behavior of a PE when an error is signaled to it by the system, including

invoking a service failure mode by taking an error exception, and optional mechanisms to limit propagation of

an error.

TLDCY

The RAS Extension and RAS System Architecture do not require systems to implement error recovery

mechanisms, including poison, and do not require systems to limit the silent propagation of errors.

1.3.3 Fault handling

SWFLQ

Fault handling by software is the process by which software diagnoses and responds to faults to improve

availability.

Fault handling methods include:

GGCDN

•

Predictive Failure Analysis (PFA), using information recorded by hardware to trigger pre-emptive action.

WNHJF

The RAS Extension and RAS System Architecture provide optional mechanisms to allow the reporting of errors

and warnings to a fault handling agent, and to record information about the fault in an error record. It is the

responsibility of the error recovery and fault handling processes to collate the error record data and write it to an

error log.

FQRSQ

The detailed nature of the fault handling agent is outside the scope of this architecture. Fault handling and error

recovery might be independent agents.

Armv8-A架构RAS特性详解

ArmARM v8-A Supplement - RAS

DDI0487F_a_armv8_arm.pdf

print("\t", len(set(self.data.ddi_edge_classes.numpy())), "cell-lines")解释一下

DDI0243C_adk_r3p0_trm.pdf

ddi0424d_dma330_r1p2_trm

DXGK_DDI_NOTIFY_ACPI_EVENT

数据路径为"数据0507/DDI_network.txt"、“数据0507/DDI_pos.txt”和“数据0507/DDI_neg.txt”.完善这3个数据路径按要求完成题目

这些作为函数整合到一个程序中，数据存放在文本ddi_with_type_latest.txt中，类似与这样的格式：0 1583 6 1 43 20 1 97 20 1 113 20

ddi0484c_cortex_m0p_r0p1_trm

最新资源