Native Inspect Manual (H06.04+)

Introducing Native Inspect

Native Inspect Manual—528122-005

1-23

Stopping Native Inspect

Native Inspect runs as a separate process from the current process being debugged

and stops:

•

When the current process stops, if Native Inspect was started automatically to

debug a process.

•

If you enter the exit command or the quit command to explicitly stop. Native

Inspect detaches from the current process and stops, leaving breakpoints in place.

Note, however, that if any of these breakpoints is subsequently hit, another

debugger instance is automatically started.

Differences Between Native Inspect and WDB

and GDB

•

WDB and GDB support a run command that is used to start a program from within

the debugger. Native Inspect, however, does not allow you to start a process from

within the debugger. You must start the process from a TACL or OSS prompt.

You can, however, use the attach command to attach an instance of Native Inspect

to a TNS/E native process. (Native Inspect also supports several commands, such

as the vector command, that are not supported by WDB or GDB.)

•

Native Inspect does not support deferred breakpoints. In WDB and GDB, deferred

breakpoints can be set on functions before a program or library is loaded.

•

You cannot call functions in the current process from Native Inspect.

•

Native Inspect does not support threading (such as Standard POSIX Threads).

•

Native Inspect does not support the WDB/GDB ability to suspend the debugging

target when a DLL is loaded or unloaded (catch load and catch unload). If

you need to debug static initialization code, use Visual Inspect.

•

Native Inspect supports debugging multiple processes, but WBD and GDB do not

support multiprocess debugging in the way that Native Inspect does.