Hello all:

I am working on some benchmark stuff and would like to be able to start
execution via checkpoint restoration to avoid wasting unnecessary time
running the program from the start, and to ensure identical conditions at
the time that code modifications come into effect (after the checkpoint is
loaded).  In the end, the idea is to compare my execution output with the
'golden' output to see what the differences are.  However, "diff" does not
work since the golden output is complete and the output from my checkpoint
restoration is garbage up until the time the checkpointed code starts
running and producing normal output.

So, my question is, is there an easy way around this so that I can still
'diff' the part of my output that isn't garbage with the corresponding bit
of the 'golden' results?  I'm not well versed with various cat and sed
options, so a wee bit of guidance would be much appreciated.

Thanks,
Griffin Wright
_______________________________________________
m5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users

Reply via email to