Understanding the Diagnostic Subsystem for AIX

Test Unit Definition

Fundamental to the Test Unit methodology is a basic, modular building block that is referred to as a Test Unit. A test unit is a single operation performed on the system or subsystem under test. Most often this is an individual function test, such as a register read/write test. Several basic assumptions are made for the test units:

• Only one modular test function is performed in each individual test unit.
• Test units are numbered, and the calling application specifies the number of the test unit it wishes to execute.
• No environmental specific code is allowed in a test unit. This specifically includes user interface calls. Also, device-access methods such as reads or writes are done with generic function calls, which can then be defined in a different source file and coded, if necessary, to meet the specific requirements of the target environments.
• Test units are grouped appropriately in source files. This allows custom building of executable libraries to meet the requirements of the target environments.
• In cases where the same test unit may be used to test hardware in different ways based on some control variables (for example, speed or mode settings), that test unit may be used to represent several "logical" test units, each with a different test unit number. When the test unit is called, it would interpret the test unit number requested and set the control variables appropriately.