Table of Contents
Advertisement
To capture a continuous stream of program execution no matter how large your program
Note that "trng addr=<addr>..<addr>" means define an address range for the
analyzer; "tsto !r" means store all trace activity except activity occurring in
the defined address range.
4
Start the analyzer trace with the command, Trace→Again
5
Start your program running using Execution→Run→from(), from
Transfer Address, or from Reset, as appropriate.
The Trace→Again (or trace again) command starts the analyzer trace
with the most recent trace specifications (including the pod_command
specifications you entered). The trace command cannot be used by itself
because it defaults the "bc -e trig1", "trng addr=...", and "tsto !r"
specifications, returning them to their default values before the trace begins.
You can see the progress of your trace with the command, Display→Status.
A line in the Trace Status listing will show how many states have been
captured.
The notation "trig1 break" usually followed by "Emulation trace
6
complete" will appear on the status line. If "trig1 break" remains on
the status line without "Emulation trace complete", manually stop the
trace with the command:
Trace→Stop
You must wait for the notation "trig1 break" and/or "Emulation trace
complete" to appear on the status line; this ensures the trace memory is filled
during the trace (except for the unfilled space you specified in Step 2 above).
Note that when you set a delay specification using tgout -c or tgout -t (trigger
output delay before trace complete/after trigger), the trace will indicate
complete as soon as the analyzer has captured the state specified, even
though the entire trace memory has not been filled.
If the notation "trig1 break" remains on the status line without being replaced
by "Emulation trace complete", it indicates the trace memory is not
completely filled, and no more states are being captured.
Store the entire trace memory content in a file with a command like:
7
wait measurement_complete ; copy trace to <directory/filename>
The "wait" command is inserted ahead of the "copy" command to ensure that
the unload of trace data is complete before you try to store it. Without "wait",
you will get an ERROR message warning that the unload is still in process.
The <filename> is an ASCII filename for a binary file that can be viewed using
the load trace command.
Start a new trace with the command: trace again
8
Using the Emulation Analyzer
221
Advertisement
Table of Contents
