# US 2024/0004744 A1 · Scalable machine check architecture

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2024/0004744 A1
- Document ID: US20240004744A1
- Status: Published application
- Kind: A1
- Application number: 17/854,710
- Filed: 2022-06-30
- Published: 2024-01-04
- Earliest claimed: 2022-06-30
- Assignees: Advanced Micro Devices, Inc.
- Applicants: Advanced Micro Devices, Inc.
- Inventors: Vilas K. Sridharan; Dean A. Liberty; Magiting Talisayon; Srikanth Masanam
- Web page: https://patentgrep.com/patent/US20240004744A1

## CPC

- G06F 11/0772: {Means for error signaling, e.g. using interrupts, exception flags, dedicated error registers}
- G06F 11/0787: {Storage of error reports, e.g. persistent data storage, storage using memory protection}
- G06F 11/0793: {Remedial or corrective actions (recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40)}
- G06F 11/1405: {at machine instruction level}
- G06F 12/0284: {Multiple user address space allocation, e.g. using different base addresses (interprocessor communication G06F15/163)}
- G06F 12/0292: {using tables or multilevel address translation means (G06F12/023 takes precedence; address translation in virtual memory systems G06F12/10)}
- G06F 2212/1008: Correctness of operation, e.g. memory ordering
- G06F 2212/1016: Performance improvement
- G06F 2212/154: Networked environment
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

An apparatus and method for supporting communication during error handling in a computing system. A computing system includes a first partition and a second partition, each capable of performing error management based on a respective machine check architecture (MCA). When a host processor in the first partition detects an error that requires information from processor cores of the second partition, the host processor generates an access request with a target address pointing to a storage location in a memory of the second partition, not the first partition. When the host processor receives the requested error log information from the second partition, the host processor completes processing of the error. To support the host processor in generating the target address for the access request, during an earlier bootup operation, the second partition communicates the hardware topology of the second partition to the host processor.

## Claims

### Claim 1

An apparatus comprising:
a given processor core; and
circuitry configured to perform, based on a first machine check architecture, error management together with the given processor core; and
wherein, in response to processing an error, the given processor core is configured to generate a target address of an access request pointing to a storage location in a memory of a remote partition configured to perform error management based on a second machine check architecture different from the first machine check architecture.

### Claim 2 (depends on 1)

The apparatus as recited in claim 1, wherein the given processor core is further configured to send the access request with the target address to the remote partition.

### Claim 3 (depends on 2)

The apparatus as recited in claim 2, wherein the given processor core is further configured to complete processing of the error, in response to receiving error log information from the remote partition.

### Claim 4 (depends on 2)

The apparatus as recited in claim 2, wherein the given processor core is further configured to assign parallel data tasks to the remote partition that comprises a plurality of processor cores of a compute accelerator.

### Claim 5 (depends on 2)

The apparatus as recited in claim 2, wherein in response to performing a bootup operation, the given processor core is configured to generate a mapping table that stores address mappings between local addresses pointing to storage locations that are accessible by the given processor core and remote addresses pointing to storage locations that are accessible by processor cores of the remote partition.

### Claim 6 (depends on 5)

The apparatus as recited in claim 5, wherein the given processor core is further configured to execute a kernel of a base operating system of the apparatus.

### Claim 7 (depends on 5)

The apparatus as recited in claim 5, wherein the mapping table stores a connector identifier along with the address mappings, wherein the connector identifier identifies a connector that couples the given processor core to the remote partition.

### Claim 8

A method, comprising:
processing tasks of a given application by a given processor core;
performing, based on a first machine check architecture, error management by circuitry together with the given processor core; and
in response to processing an error, generating, by the given processor core, a target address of an access request pointing to a storage location in a memory of a remote partition configured to perform error management based on a second machine check architecture different from the first machine check architecture.

### Claim 9 (depends on 8)

The method as recited in claim 8, further comprising sending, by the given processor core, the access request with the target address to the remote partition.

### Claim 10 (depends on 9)

The method as recited in claim 9, further comprising completing processing of the error, by the given processor core, in response to receiving error log information from the remote partition.

### Claim 11 (depends on 9)

The method as recited in claim 9, further comprising assigning parallel data tasks to the remote partition that comprises a plurality of processor cores of a compute accelerator.

### Claim 12 (depends on 9)

The method as recited in claim 9, wherein in response to performing a bootup operation, the method further comprises generating, by the given processor core, a mapping table that stores address mappings between local addresses pointing to storage locations that are accessible by the given processor core and remote addresses pointing to storage locations that are accessible by processor cores of the remote partition.

### Claim 13 (depends on 12)

The method as recited in claim 12, further comprising executing, by the given processor core, a kernel of a base operating system.

### Claim 14 (depends on 12)

The method as recited in claim 12, further comprising storing, by the mapping table, a connector identifier along with the address mappings, wherein the connector identifier identifies a connector that couples the given processor core to the remote partition.

### Claim 15

A computing system comprising:
a server comprising:
a first partition configured to perform error management based on a first machine check architecture; and
a second partition configured to perform error management based on a second machine check architecture different from the first machine check architecture; and
wherein each of the first partition and the second partition is configured to process tasks of a given application; and
wherein, in response to processing an error detected in the first partition, a given processor core of the first partition is configured to generate a target address of an access request pointing to a storage location in a memory of the second partition.

### Claim 16 (depends on 15)

The computing system as recited in claim 15, wherein the given processor core is further configured to send the access request with the target address to the second partition.

### Claim 17 (depends on 16)

The computing system as recited in claim 16, wherein the given processor core is further configured to complete processing of the error, in response to receiving error log information from the second partition.

### Claim 18 (depends on 16)

The computing system as recited in claim 16, wherein the second partition comprises a plurality of processor cores of a compute accelerator.

### Claim 19 (depends on 16)

The computing system as recited in claim 16, wherein in response to performing a bootup operation, circuitry of the first partition is configured to generate a mapping table that stores address mappings between local addresses accessible by the given processor core of the first partition and remote addresses that are accessible by processor cores of the second partition.

### Claim 20 (depends on 19)

The computing system as recited in claim 19, wherein the given processor core is further configured to execute a kernel of a base operating system.
