NAME

sample — Profile a process during a time interval

SYNOPSIS

DESCRIPTION

sample is a command-line tool for gathering data about the running behavior of a process. It suspends the process at specified intervals (by default, every 1 millisecond), records the call stacks of all threads in the process at that time, then resumes the process. The analysis done by sample is called ``sampling'' because it only checks the state of the program at the sampling points. The analysis may miss execution of some functions that are not executing during one of the samples, but sample still provides useful data about commonly executing functions.

At the end of the sampling duration, sample produces a report showing which functions were executing during the sampling. The data is condensed into a call tree, showing the functions seen on the stack and how they were called. (This tree is a subset of the actual call tree for the execution, since some functions may not have been executing during any of the sampling events.) The tree is displayed textually, with called functions indented one level to the right of the callee.

In the call tree, if a function calls more than one function then a vertical line is printed to visually connect those separate children functions, making it easier to see which functions are at the same level. The characters used to draw those lines, such as + | : ! are arbitrary and have no specific meaning.

ARGUMENTS

The user of sample specifies a target process (either by process id, or by name), the duration of the sampling run (in seconds), and a sampling rate (in milliseconds).

If the sampling duration is not specified, a default of 10 seconds is used. Longer sampling durations provide better data by collecting more samples, but could also be confusing if the target process performs many different types of operations during that period.

The default sampling rate is 1 millisecond. Fast sampling rates provide more samples and a better chance to capture all the functions that are executing.

-wait tells sample to wait for the process specified (usually as a partial name or hint) to exist, then start sampling that process. This option allows you to sample from an application's launch.

-mayDie tells sample to immediately grab the location of symbols from the application, on the assumption that the application may exit or crash at any point during the sampling. This ensures that sample can give information about the call stacks even if the process no longer exists.

-fullPaths tells sample to show the full path to the source code (rather than just the file name) for any symbol in a binary image for which debug information is available. The full path was the path to the source code when the binary image was built.

-e tells sample to open the output in TextEdit automatically when sampling completes.

-file filename tells sample the full path to where the output should be written. If this flag is not specified, results are written to a file in /tmp called <application name>_<date>_<time>.<XXXX>.sample.txt, where each 'X' is replaced by a random alphanumeric character.

If neither the -e nor -file flags are given, the output gets written to stdout as well as saved to the file in /tmp.

SEE ALSO

filtercalltree(1), spindump(8)

The Xcode developer tools also include Instruments, a graphical application that can give information similar to that provided by sample. The Time Profiler instrument graphically displays dynamic, real-time CPU sampling information.