
October 13 1995, Draft 1
392
Addendum to –– Evaluating and Programming the 29K RISC Family
hexadecimal format. Use of the ASMDATA label indicates that the DATA label
information should be disassembled –– when the corresponding address is known to
lie in a TEXT region.
The “l(fā)atd start, end” command is used to display processed trace information
based on stored lines. Processed line and raw line numbers are the same with regard
to the processor status during the traced processor cycle, but only
valid
lines appear in
the processed trace listing. A valid line is one in which
useful
processor activity was
performed. For valid lines, the ADDR, DATA, R/_W and I/_D labels are reevaluated
to correspond with the associated processor status value.
MonDFE Trace Access Commands
This section briefly describes MonDFE commands relating to displaying trace
information. A complete list of MonDFE commands is obtained by entering the
command “”. MonDFE supports command files with the “zc file” command. It is
useful to place a list of “tip latadd SYMADD”–type commands in a command file
such as la.rc. This enables the “zc la.rc” command to initialize MonTIP trace
processing. The MonDFE command “ze file” can be used to record displayed
information into a log file.
Before a trigger can occur, trigger conditions must be installed in the analyzer.
The command “l(fā)atrig term, label=pattern” can be used when setting trigger patterns
in the logic analyzer. Specifying trigger logic and sequence control must be specified
using the analyzer window. Once the trigger has been established (and the
POD_040._D setup may be adequate) it is usually only necessary to use “l(fā)atrig”
commands such as “tip latrig a, ADDR=10004”. Trigger patterns can be entered
directly using the analyzer window, this also requires the hexadecimal patten (rather
than symbol) value be known. If the
swaf
program is used to build symbol table
information in Hewlett–Packard’s GPA format, it can be directly loaded into the
analyzer. This enables trigger labels to be set using symbols when directly using the
analyzer window.
The Analyzer can be triggered at each breakpoint by setting the break–address
to the illegal op–code vector address (vector 0). This technique is useful when
breakpoints are implemented by temporarily replacing instructions with illegal
op–code instructions. The MiniMON29K DebugCore uses this technique when
on–chip breakpoint registers are not available.
UDB Commands
UDB has been enhanced by CaseTools to support displaying trace information
in source format. This makes UDB a preferred tool for use with a logic analyzer.
Additionally, UDB, like other non enhanced source level debuggers, can also be used
in transparent–mode. With transparent–mode operation, it is possible to issue
commands for MonTIP processing. Given that UDB supports source level tracing, it
is unlikely that transparent–mode operation would be selected for use with UDB.