Profiling Tools Report Column Definitions
The following table contains the definitions of values that are available in performance report views.
Column Name |
Description |
---|---|
% Application Exclusive Time |
The percentage of the total time (processor cycles), or performance counter values of the specified data segment that was spent in the application exclusive time of all instances of the function in this context. |
% Application Inclusive Time |
The percentage of the total time (processor cycles), or performance counter values of the specified data range that was spent in the application inclusive time of all instances of the function in this context. |
% Elapsed Exclusive Time |
The percentage of the total time (processor cycles), or performance counter values of the specified data segment that was spent in the elapsed exclusive time of all instances of the function in this context. |
% Elapsed Inclusive Time |
The percentage of the total time (processor cycles), or performance counter values of the specified data segment that was spent in the elapsed inclusive time of all instances of the function in this context. |
Application |
The file name of the application that created the process or thread. |
Application Etl File |
Location of the file that contains ETW data. |
Application Exclusive Time |
Calculated time for a function or summary item that excludes performance data that contains operating system events, time from collection probes, and the time of subroutines called by the function. |
Application Inclusive Time |
Calculated time for a function or summary item that excludes performance data that contains operating system events and time from collection probes, but includes the application time of subroutines called by the function. |
AVG Application Exclusive Time |
The average application exclusive clock time (processor cycles), P6 performance, or user-defined counter value for all instances of this function in the specified data range. |
AVG Application Inclusive Time |
The average application inclusive clock time (processor cycles), P6 performance, or user-defined counter value for all instances of the function in the specified data range. |
AVG Elapsed Exclusive Time |
The average elapsed exclusive clock time (processor cycles), P6 performance or user-defined counter value for all instances of the function in the specified data range. |
AVG Elapsed Inclusive Time |
The average elapsed inclusive clock time (processor cycles), P6 performance or user-defined counter value for all instances of the function in the specified data range. |
Base Address |
The memory address of the loaded module. |
Call Depth |
The depth of the call tree. |
Clock Frequency |
The processor clock speed. |
Command Line |
Command that is used to create the performance report. |
Counter 1 - 16 |
During instrumentation, you can collect data from many different counters. By default, only the TimeStamp counter is used. |
CPU ID |
CPU manufacturer and type information. |
Creation Time |
This is the date/time that the .vsp file was created. |
Dirty Shutdown |
Indicates that the profiler was abruptly exited. |
Elapsed Exclusive Time |
Calculated time for a function or summary item that excludes time from collection probes and the elapsed time of subroutines called by the function. |
Elapsed Inclusive Time |
Calculated time for a function or summary item that includes intervals from subroutines called by the function, and excludes time from collection probes. |
Event Action |
The type of event action that occurred. Its Event actions can be Entry or Exit. It only appears in the Call Tree report. |
Event Type |
Event Type, such as event action, appears only in Call Tree report. The following list explains different event types: L - Loaded module M - Mark E - End Explicit - refers to explicitly receiving an exit from a function. Implicit - refers to inferring function exit because of an exception. Additionally, if ETW events are enabled, event type also lists them. |
Exclusive Allocations |
Allocations in a function excluding all other allocations in subroutines that it called. |
Exclusive Allocations Percent |
([Allocations in a function excluding all other allocations in subroutines that it called] / [Total number of exclusive allocation that occurred during profiling]) * 100 |
Exclusive Bytes Allocated |
Bytes allocated in a function excluding all other byte allocations in subroutines that it called. |
Exclusive Bytes Percent |
(Bytes allocated in a function / total bytes ) * 100 |
Exclusive Percent |
([Exclusive samples for the function] / [Total exclusive samples taken during profiling]) * 100 |
Exclusive Samples |
Total number of performance data samples collected for the function excluding performance data for other functions that were called by it. |
Exclusive Transitions |
The number of transition (operating system) events that occurred in all instance of this function, excluding functions called by the function. |
Exclusive Transitions Percentage |
The percentage of the total number of operating system (transition) events that occurred in the inclusive time of the instances of this function in the data range that also occurred in the exclusive time of the function. |
Final Processes |
Number of processes alive at the end of the profiling run, typically only nonzero when detaching with sampling. |
Final Threads |
Number of active profiled threads that were running when the application terminated. |
Function Address |
The address of the function in hexadecimal format. |
Function Name |
Name of the function. |
I/O Buffers |
Count of buffers used for storing data during profiling. |
ID |
The system-defined numeric identifier assigned to the process or thread. |
Inclusive Allocations |
Allocations in a function including all allocations in subroutines that it called |
Inclusive Allocations Percent |
([Allocations in a function including all allocations in subroutines that it called] / [Total number of inclusive allocation that occurred during profiling]) * 100 |
Inclusive Allocs |
Allocations in a function including all allocations in subroutines that it called. |
Inclusive Byte Percent |
(Bytes allocated in a function and all other subroutines that it called / total bytes ) * 100 |
Inclusive Bytes Allocated |
Bytes allocated in a function including all other byte allocations in subroutines that it called. |
Inclusive Bytes Percent |
(Bytes allocated in a function and all other subroutines that it called / total bytes ) * 100 |
Inclusive Percent |
([Inclusive samples for the function] / [Total inclusive samples taken during profiling]) * 100 |
Inclusive Samples |
Total number of performance data collected for the function including performance data for other functions that were called by it. |
Inclusive Transitions |
The number of operating system (transition) events that occurred in all instances of this function, including functions called by the function. |
Inclusive Transitions Percentage |
The percentage of operating (transition) system events that occurred in the inclusive time of the instances of this function that were called by the parent function in the call tree compared to the total number of inclusive transitions in the data range. |
Instruction Address |
The address of the instruction in hexadecimal format. |
Kernel Etl File |
If ETW data was collected, a separate file is created by using that data and this is the location of the file with kernel events. |
Level |
The call depth in call tree. |
Line Number |
Location of the start of the function in the source file. |
Machine Name |
Computer that is used for profiling. |
Mark |
Data inserted by the user as a marker in the code to help document performance issues. |
MAX Application Exclusive Time |
The maximum single application exclusive clock time (processor cycles), or performance counter value for any one instance of this function that was called by the parent function in the call tree. |
MAX Application Inclusive Time |
The greatest application inclusive clock time (processor cycles), or performance counter value for any one instance of this function that was called by the parent function in the call tree. |
MAX Elapsed Exclusive Time |
The greatest elapsed exclusive clock time (processor cycles), or performance counter value for any one instance of this function that was called by the parent function in the call tree. |
Maximum Processes |
Maximum number of processes that the profiler allows to be profiled at the same time, beyond which an error is reported. This can be controlled by a registry key. |
Maximum Threads |
Maximum number of threads that the profiler allows to be profiled at the same time, beyond which an error is reported. This can be controlled by a registry key. |
MIN Application Exclusive Time |
The minimum application exclusive clock time (processor cycles), or performance counter value for any one instance of this function that was called by the parent function in the call tree. |
MIN Application Inclusive Time |
The least application inclusive clock time (processor cycles), or performance counter value for any one instance of this function that was called by the parent function in the call tree. |
MIN Elapsed Exclusive Time |
The least elapsed exclusive clock time (processor cycles), or performance counter value for any one instance of this function that was called by the parent function in the call tree. |
MIN Elapsed Inclusive Time |
The least elapsed inclusive clock time (processor cycles), or performance counter value for any one instance of this function that was called by the parent function in the call tree. |
Module Identifier |
An unsigned integer used to track the modules. It is assigned to a module based on the order in which it was loaded into the process. For example, the first module is module 0 and the second is module 1, and so on. |
Module Name |
Name of the module that contains the function. |
Module Path |
Directory location of the module. |
Module Size |
The size of the module in hexadecimal. |
Name |
The string assigned to the process or thread with the NameProfile API function. If no name was assigned to the item, the type of item (Process, Thread) is pre-pended to the item ID. |
Num. CPUs |
Total number of CPUs in the computer that is used for profiling. |
Number of Calls |
The number of calls made to the instances of this function that were called by the parent function in the call tree. |
Operating System |
Operating system version information. |
Parent Function Address |
Address in memory of the parent function that called another function. |
Percentage of calls |
The number of instances of this function that were called by the parent function in the call tree as a percentage of the total number of calls to all functions in the specified data range. |
Process Highwater |
Maximum number of process profiled at the same time. |
Process ID |
Numeric identifier of the process. |
Process Name |
Name of the process. |
Report Creation Time |
The date and time of report creation. |
Root Node Recursive |
Indicates if this function was directly or indirectly called recursively in this context. |
Sample Interval |
Average of number of clock cycles between samples or the event used. |
Samples (Aborted) |
Samples taken while the application's stack could not be walked. |
Samples (Application) |
Samples that were taken while the application was in application mode. |
Samples (Broken) |
Incomplete stack because the call stack walk failed to walk up to the thread start address. |
Samples (Kernel) |
Samples that were taken while the application was in kernel mode. These are thrown out. |
Samples (Overhead) |
Samples that occurred while a call stack walk was occurring. These are thrown out. |
Samples (Total) |
Total samples that occurred. |
Scenario Name |
Default value is ProfileRun. Can be set in process using the NameProfile API. |
Size of Buffers |
Size of the I/O Buffers. See I/O Buffers. |
Source File |
The full path of the .vsp report file. |
Source File Name |
Name of the source file that contains this function. |
Stack Pointer |
Stack pointer value for the function. |
Source Character Begin |
Column number at the beginning of the reported code sample. |
Source Character End |
Column number at the end of the reported code sample. |
Source Line Begin |
Line number at the beginning of the reported code sample. |
Source Line End |
Line number at the end of the reported code sample. This is the same as Source Line Begin except for samples taken on multiline statements, for example: y = myFunction(x + 2, y * 3, x / y ) – x; |
Thread Highwater |
Maximum number of threads profiled at a single time. |
Thread ID |
The system-defined numeric identifier assigned to the thread. |
Thread Name |
The string assigned to the process or thread with the NameProfile API function. If no name was assigned to the item, the type of item (Process, Thread) is pre-pended to the item ID. |
Time Delta |
The difference between the timestamp at which this event occurred and the previous event's timestamp. |
Time Exclusive Probe Overhead |
The total time attributed to profiling probes in the exclusive time of the instances of this function that were called by the parent function in the call tree. |
Time Inclusive Probe Overhead |
The total time attributed to profiling probes in the inclusive time of the instances of this function that were called by the parent function in the call tree. |
Timestamp |
The time at which an event occurred. |
Tool Name and Full Version |
The name and version of VSPerfReport. |
Total Processes |
Total number of processes profiled over the lifetime of the profiling session. |
Total Threads |
Total number of threads created during profiling. |
Type |
Represents the type of row in a call tree. A value of: 0 represents the root function 1 represents the caller 2 represents the callee For example, if you have the following call tree: main->funcA->funcB Type column will have the following rows: 0 - main 2 - funcA 0 - funcA 1 - main 2 - funcB 0 - funcB 1 - funcA |
Unique ID |
A hexadecimal number that identifies the function. |
Unique Process ID |
An unsigned integer used to track the process. It is assigned to a process based on the order in which it was loaded. For example, the first process is 0 and the second is 1, and so on. |
Unique Process ID |
Unsigned integer to track the process. |
VSPerf90.DLL File Version |
The version of the data collection dynamic link library, named VSPerf90.DLL, that created the .vsp file. |