Re: [Gems-users] infinite loop when trying to use con0.break

["Min Xu (Hsu)" <xu@xxxxxxxxxxx>]

Hi,

I don't think opal supports con0.break. Opal changes simics exception 
queues which can break how simics stops itself. We have write some code 
in opal to support Ctrl-C, you many need to write your own code to 
support con0.break as well.

Hope this helps.

-Min


Vince Weaver wrote:

> Hello,
>
> I am trying to use "con0.break" to help automate using gems.
>
> I'd think the script below should work, but when the second string appears
> it just prints seemingly infinite patch  PC / patch  NPC errors.
> I've commented out those errors, and re-tried and it just seems to get
> stuck in an inifinite loop rather than breaking.
>
> Is using con0.break supported when using GEMS?
>
> In the below script I'd hope that once the second string is printed it
> would break out and print the stats.
>
> Vince
>
>
>
> con0.break "572_START"
> con0.break "572_END"
> con0.input "echo 572_START && echo test && echo 572_END\n"
> run
> instruction-fetch-mode instruction-fetch-trace
> istc-disable
> dstc-disable
> load-module ruby
> load-module opal
> ruby0.setparam g_NUM_PROCESSORS 4
> ruby0.setparam g_MEMORY_SIZE_BYTES 536870912
> ruby0.init
> opal0.init
> opal0.sim-start "results.opal"
> opal0.sim-step 1000000000
> ruby0.dump-stats
> opal0.listparam
> opal0.stats
>
>
> _______________________________________________
> Gems-users mailing list
> Gems-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/gems-users
>  


--
That is what makes time travel into our past impossible and thereby
explains, as those who say it is possible cannot explain, why it will
never happen - as we know it will not, since we know it never has.
 -- D. H. Mellor, Emeritus Professor of Philosophy, University of Cambridge