Thanks Paul! From the link you shared, looks like when building with runspec, a build dir is created that has the appropriate makefile and the source. So I think can make changes there, use makefile to build it and get the exe.
On Wed, Sep 12, 2012 at 12:48 AM, Paul Rosenfeld <[email protected]>wrote: > Is it the build step or the run step that does the checksum? > > specmake and specinvoke should both have flags that will generate the > commands for a build/run without actually invoking them. Maybe something > on this page http://www.spec.org/cpu2006/docs/runspec-avoidance.html will > help? (ex: the -n flag for specmake and specinvoke) > > On Wed, Sep 12, 2012 at 1:41 AM, Ankita (Garg) Goel > <[email protected]>wrote: > >> I specified check_md5=0 in the config file, plus --noreportable, but >> couldn't bypass the check *sigh* >> >> Has anyone done this before ? >> >> >> On Tue, Sep 11, 2012 at 11:49 PM, Paul Rosenfeld <[email protected]>wrote: >> >>> I have successfully attempted to avoid SPEC and all of it's baggage in >>> recent years. If I recall correctly, there are some flags you can pass to >>> runspec to tell it that you don't want to do a "reportable" run. Try to >>> add --noreportable (see 3.2.1 on this page: >>> http://www.spec.org/cpu2006/docs/runspec.html ) and see if that removes >>> the checksum check. >>> >>> >>> >>> On Wed, Sep 12, 2012 at 12:39 AM, Ankita (Garg) Goel < >>> [email protected]> wrote: >>> >>>> Thanks Paul for the quick response! I am new to using SPEC. I just >>>> tried to put some ptlcalls in the bzip2 source, but the compilation >>>> framework of SPEC (using runspec) seems to be checking for checksums. Any >>>> easy way to work around this ? >>>> >>>> >>>> On Tue, Sep 11, 2012 at 7:55 PM, Paul Rosenfeld >>>> <[email protected]>wrote: >>>> >>>>> If you can build your workload from source, the easiest approach to >>>>> doing this would be to add some region of interest hooks in the form of >>>>> ptlcalls to the source code. This way, you can create the checkpoint right >>>>> at the point of interest instead of running simulation execution from a >>>>> checkpoint of the shell. >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> On Tue, Sep 11, 2012 at 8:28 PM, Ankita (Garg) Goel < >>>>> [email protected]> wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> Is there a way I could fast-forward the execution of a benchmark for >>>>>> a few million of instructions before switching into the simulation mode ? >>>>>> For instance, I create the checkpoint like below: >>>>>> >>>>>> # ./create_checkpoint bb; ./bzip2 <param>; ./stop_sim; >>>>>> >>>>>> Now, once I start marss again, from the above checkpoint 'bb', I want >>>>>> to fast-forward about 1million user instructions and then run simulation >>>>>> for the next few million instructions. Can I do this ? I did look around >>>>>> in >>>>>> the archives, but did not find the answer there. Any suggestions/thoughts >>>>>> on this will greatly help me! >>>>>> >>>>>> Thanks! >>>>>> >>>>>> -- >>>>>> Regards, >>>>>> Ankita >>>>>> >>>>>> _______________________________________________ >>>>>> http://www.marss86.org >>>>>> Marss86-Devel mailing list >>>>>> [email protected] >>>>>> https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel >>>>>> >>>>>> >>>>> >>>> >>>> >>>> -- >>>> Regards, >>>> Ankita >>>> Graduate Student >>>> Department of Computer Science >>>> University of Texas at Austin >>>> >>>> >>>> >>> >> >> >> -- >> Regards, >> Ankita >> Graduate Student >> Department of Computer Science >> University of Texas at Austin >> >> >> > -- Regards, Ankita Graduate Student Department of Computer Science University of Texas at Austin
_______________________________________________ http://www.marss86.org Marss86-Devel mailing list [email protected] https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
