2.5 The ‘watch’ and ‘clear’ Commands
One can also set breakpoint on a variable to monitor any read or write instructions
over it. This kind of breakpoints are called data breakpoint or ‘watchpoint’. If any
statement reads a watched variable or writes over it, the debugger paused and shows
the change and the statement accessed the variable. Likewise, any expression over the
variable can be watched. The ‘watch’ command is used to monitor any read or write
accesses over a variable. ‘rwatch’ and ‘awatch’ commands only intercept a read and
write access, respectively.
Watching commands in GDB
-
1.
- One way to put a watchpoint over a variable is to pause the execution
where it is declared, and then place a watchpoint. Breakpoints are used
to pause at the declaration statement. For example, if we are going to
watch any accesses to size variable in the Stack class, a breakpoint has
to be placed at line ‘main.cpp:11’, where the Stack class is instantiated.
After size is marked to be watched, any further access is reported, like the
default constructor and push function in the Stack class (Figure 23). After
any pause for a watchpoint, one can use ‘cont’ to resume the execution
until the next breakpoint.
-
2.
- Since watchpoints are like break points, one can see a list of all breakpoints
using ‘info breakpoints’ or ‘info watchpoints’. A break point can be
deleted using ‘delete’, or ‘d’, command. Figure 24 shows how one or all
breakpoints are cleared.
Watching in Visual Studio
-
1.
- Once a breakpoint pauses a program execution, one can add any live
variable around the breakpoint to the watch list. Later he can watch any
changes on the variable in the ‘watch’ window. By right clicking over the
variable name, once can add its name to the watch list. For example, the
variable size is added to watch list in Figure 25.
One can also add any expression to be watched
-
2.
- All the active breakpoints can be deleted or disabled from ‘Debugging’
menue.
