Intel® Trace Analyzer and Collector

Recording Source Location Information

Intel® Trace Collector can automatically record the locations of subroutine calls in the source code. Compile the relevant application modules with support for debugging by using these compiler flags that enable the generation of debug information for Intel® Trace Collector:

mpicc -g -c ctest.c
mpif77 -g -c ftest.c

If your compiler does not support a flag, then search for a similar one.

Allinea* Distributed Debugging Tool

Allinea*'s Distributed Debugging Tool (DDT) must be configured to run the user's application with the necessary Intel libraries preloaded.

Take the following steps:

  • Go to the Run dialog box

  • Select the Session/Options menu

  • In the Session/Options menu, choose Intel MPI Library and the Submit through queue or configure own mpirun command option

Datatype Mismatches


Data type mismatches are detected by calculating a hash value of the data type signature and comparing that hash value: if the hash values are different, the type signatures must have been different too and an error is reported. Because the information about the full type signature at the sender is not available, it has to be deduced from the function call parameters and/or source code locations where the data is transmitted.

Using VTserver

This is how the application starts, collects trace data and terminates:

  1. The application initializes itself and its communication.

  2. The application initializes communication between VTserver and processes.

  3. Trace data is collected locally by each process.

  4. VT data collection is finalized, which moves the data from the processes to the VTserver, where it is written into a file.

  5. The application terminates.



--sumdata <output trace name>




Forces creation of additional Summary Data files for the given trace.

You can use the --sumdat option with or without --merge option. Thus, there can be the following three scenarios:

Recording Communication Events

These are API calls that allow logging of message send and receive and MPI-style collective operations. Because they are modeled after MPI operations, they use the same kind of communicator to define the context for the operation.

enum _VT_CommIDs

Logging send/receive events evaluates the rank local within the active communicator, and matches events only if they take place in the same communicator (in other words, it is the same behavior as in MPI).

Subscribe to Intel® Trace Analyzer and Collector