The problem is that some options should be passed to m5 and some to
the configuration script itself. The ones that come before the name of
the configuration script are interpreted by m5, and that's where you
want --trace-flags. To see what options m5 itself supports, put --help
on the command line before the script name. To see what options the
script provides, put it after.
Gabe
[EMAIL PROTECTED] wrote:
Hi,
I have tried to use --trace-flags into command line(written in python
file) to invoke new traceflags as follow:
cmd = m5 + ' -d ' + out_dir +\
' --stats-file=' + sta_file +\
' ' + current_dir + sim_dir + 'configs/' + cf + '/fs.py' +\
' -b ' + benchmark +\
' --trace-flags=' + 'DRAM' +\
' > ' + std_file + ' 2> ' + err_file
When I run the simulation,error message shows:
usage: fs.py [options]
fs.py: error: no such option: --trace-flags
Does anybody know what the problem is ?
Thanks,
Tracy
I added a new traceflag into traceflags.py. How could I invoke this
trace flag in the simulation execution.
In m5_2.0b1, I could invoke traceflags by using the following
definition:
root.trace.flags = "UART"
What should I do in m5_2.0b3?
--trace-flags= on the m5 command line.
m5 --help
will give you extensive help output.
Nate
_______________________________________________
m5-users mailing list
m5-users@m5sim.org
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
_______________________________________________
m5-users mailing list
m5-users@m5sim.org
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
_______________________________________________
m5-users mailing list
m5-users@m5sim.org
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
