Apparatus and method for profiling system events in a fine grain multi-threaded multi-core processor

• US 8,762,951 B1
• Filed: 03/21/2007
• Issued: 06/24/2014
• Est. Priority Date: 03/21/2007
• Status: Active Grant

• Current Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Original Assignee: Oracle America, Inc. (Oracle Corporation)
• Inventors: Kosche, Nicolai, Laudon, James P., Talcott, Adam R., Patel, Sanjay, Sajjadian, Farnad
• Primary Examiner(s): LOUIE, JUE WANG
• Application Number:
  US11/689,359
• Time in Patent Office: 2,652 Days
• Field of Search: None
• US Class Current: 717/127
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 11/3476 Data logging G06F11/14, G06F11/2205 take precedence
• CPC Class Codes: G06F 2201/86 Event-based monitoring
• CPC Class Codes: G06F 2201/865 Monitoring of software
• CPC Class Codes: G06F 2201/87 Monitoring of transactions
• CPC Class Codes: G06F 2201/88 Monitoring involving counting
• CPC Class Codes: G06F 8/443 Optimisation

Abstract

A system and method for profiling runtime system events of a computer system may include associating a data source type with detected system events. The system events may be detected dependent on information included in a reply message received by a processor in response to a data request or other transaction request message. The reply message may include information characterizing a source type of a source of data included in the reply message. The source type information may indicate that the source is remote or local; that it is a shared or a private storage location; that the data is supplied via a cache-to-cache transfer; or that the data is sourced from a coherency domain other than that of the requesting process. Instructions, events, messages, and replies may be sampled, and extended address information corresponding to the samples may be stored in an event set database for performance analysis.

Claims

1. A computer-implemented method, comprising:

a hardware processor sending a message to one or more system components in a computing system;

an event agent of the hardware processor receiving a reply to the message from one of the one or more system components, wherein the reply comprises data supplied by a source in response to the message and a source type indicator for the source of the data, and wherein the source type indicator is set to one of a plurality of values, wherein the plurality of values comprises two or more of;

a value that characterizes the source of the data as being a remote source, relative to the hardware processor,

a value that characterizes the source of the data as being a local source, relative to the hardware processor,

a value that characterizes the source of the data as being a private storage location,

a value that characterizes the source of the data as being a shared storage location,

a value that characterizes the source of the data as being in a different coherency domain than an originator of the message,

a value that characterizes the source of the data as being in a same coherency domain as an originator of the message, or

a value that indicates the data was supplied via a cache-to-cache transfer;

the event agent detecting a runtime system event in the computing system dependent on the reply, wherein the runtime system event is a system event that occurs as a result of executing a particular software application whose behavior during execution is being profiled and on whose behalf the message was sent; and

the event agent associating an indication of the source type with the detected event in an event set, wherein the indication is dependent on the source type indicator that was included in the reply.

2. The method of claim 1, further comprising the event agent sampling the event, wherein said sampling comprises:

capturing extended address information corresponding to an instruction on whose behalf the message was sent, wherein the extended address information comprises one or more of;

a time, an identifier of a hardware component, an identifier of a software construct, an identifier of a data allocation, an identifier of a protocol, a physical address, a virtual address, a cost value, or a register value; and

associating the extended address information corresponding to the instruction with the detected event in the event set.

3. The method of claim 1, wherein the message comprises an identifier of a thread requesting data, further comprising the event agent associating the identifier with the detected event and the indication of the source type in the event set.

4. The method of claim 1, wherein the message comprises a transaction request and the reply comprises a transaction response, further comprising the event agent associating extended address information corresponding to the transaction request with the detected event and the indication of the source type in the event set.

5. The method of claim 1, wherein the detected event comprises one of:

a specified data reference, a reference to a storage location in a specified hierarchy level, a hit in a cache, a miss in a cache, a coherency domain crossing, a remote memory access, a cache-to-cache transfer, or a hardware event counter condition.

6. A computer-implemented method, comprising:

a system component receiving a message comprising a data request from a hardware processor; and

the system component sending a reply message to the hardware processor in response to the data request, wherein the reply message comprises data supplied by a source satisfying the data request and a source type indicator for the source of the data, and wherein the source type indicator is set to one of a plurality of values, wherein the plurality of values comprises two or more of;

a value that characterizes the source of the data as being a remote source, relative to the hardware processor,

a value that characterizes the source of the data as being a local source, relative to the hardware processor,

a value that characterizes the source of the data as being a private storage location,

a value that characterizes the source of the data as being a shared storage location,

a value that characterizes the source of the data as being in a different coherency domain than an originator of the data request,

a value that characterizes the source of the data as being in a same coherency domain as an originator of the data request, or

a value that indicates the data was supplied via a cache-to-cache transfer;

wherein information included in the reply message is to be used by the hardware processor in detecting, in response to receiving the reply message, a system event that occurs as a result of executing a particular software application whose behavior during execution is being profiled and on whose behalf the message was sent; and

wherein the source type indicator is to be used by the hardware processor for storing in association with the detected event in an event set.

7. The method of claim 6, wherein the reply message further comprises an identifier of the system component.

8. The method of claim 6, further comprising an intermediate component relaying the reply to the hardware processor, wherein relaying the reply comprises:

receiving the reply;

modifying the reply to indicate that the reply was relayed via the intermediate component; and

sending the modified reply to the hardware processor.

9. The method of claim 6, further comprising:

the hardware processor receiving the reply message;

the hardware processor detecting a runtime system event in response to receiving the reply message and dependent at least in part on information comprised in the reply message, wherein the runtime system event is a system event that occurs as a result of executing a particular software application on whose behalf the message was sent; and

the hardware processor associating the source type indicator with the detected event in an event set.

10. A computing system, comprising:

a hardware processor; and

a memory comprising program instructions executable by the hardware processor to implement a particular software application whose behavior during execution is to be profiled;

wherein the program instructions are configured to implement;

sending a message to one or more system components;

receiving a reply to the message, wherein the reply comprises data supplied by a source in response to the message and a source type indicator for the source of the data, and wherein the source type indicator is set to one of a plurality of values, wherein the plurality of values comprises two or more of;

a value that characterizes the source of the data as being a remote source, relative to the hardware processor,

a value that characterizes the source of the data as being a local source, relative to the hardware processor,

a value that characterizes the source of the data as being a private storage location,

a value that characterizes the source of the data as being a shared storage location,

a value that characterizes the source of the data as being in a different coherency domain than an originator of the message,

a value that characterizes the source of the data as being in a same coherency domain as an originator of the message, or

a value that indicates the data was supplied via a cache-to-cache transfer;

detecting a runtime system event dependent on the reply, wherein the runtime system event is a system event that occurs as a result of executing the program instructions that implement the particular software application on whose behalf the message was sent; and

associating an indication of the source type with the detected event in an event set, wherein the indication is dependent on the source type indicator that was included in the reply.

11. The computing system of claim 10,

wherein the program instructions are further configured to implement sampling an instruction on whose behalf the message was sent, wherein said sampling comprises:

capturing extended address information corresponding to the instruction, wherein the extended address information comprises one or more of;

a time, an identifier of a hardware component, an identifier of a software construct, an identifier of a data allocation, an identifier of a protocol, a physical address, a virtual address, a cost value, or a register value; and

associating the extended address information with the detected event and the indication of the source type in the event set.

12. The computing system of claim 10, wherein the message comprises an identifier of a thread requesting data;

wherein in response to detecting the event, the program instructions are further configured to associate the thread identifier with the detected event and the indication of the source type in the event set.

13. The computing system of claim 10,

wherein the message comprises a transaction request;

wherein the reply comprises a corresponding transaction response;

wherein in response to detecting the event, the program instructions are further executable to associate extended address information corresponding to the transaction request with the detected event and the indication of the source type in the event set.

14. The computing system of claim 10, wherein the detected event comprises one of:

a specified data reference, a reference to a storage location in a specified hierarchy level, a hit in a cache, a miss in a cache, a coherency domain crossing, a remote memory access, a cache-to-cache transfer, or a hardware event counter condition.