Hello,

I cannot use m5_resetstats and m5_dumpstats if I understand well how
they work. Indeed, most of the time the section of code will omp
parallel section and if several threads clean the stats at the same time
I risk to have false values for some threads.

I take a look at the PMU support, I am working on ARM architecture, but
I don't understand how it could help me. Could you give me some
explanations, please?

Thank you

+-------------------------------+
| Louisa Bessad                 |
| PhD student - LIRMM - Sysmic  |
| Bâtiment 4 Bureau 2.92        |
+-------------------------------+

Le 19/03/2016 13:50, Fernando Endo a écrit :
> Hello,
> 
> If the section of code is executed only a few times, it is easier to
> call m5_resetstats and/or m5_dumpstats at the begining and the end of
> the section.
> 
> You're probably summing the number of issued uops (i.e.,
> system.cpu.iq...IntAlu, etc), the sim_insts refers to the committed
> instructions (committedInsts).
> 
> You can also check the current support of the PMU of your arch (it seems
> that for ARM it is working).
> 
> Regards,
> 
> --
> Fernando A. Endo, Post-doc
> 
> INRIA Rennes-Bretagne Atlantique
> France
> 
> 
> 2016-03-17 9:26 GMT+01:00 Louisa Bessad <[email protected]
> <mailto:[email protected]>>:
> 
>     Hi all,
> 
>     I am trying to count the number of load/store instructions and
>     arithmetic instructions in a section of code of an application.
> 
>     My idea is to modify the pipeline by adding a counter in the last stage
>     that will be activated at the beginning and deactivated at the end of
>     the section by a pseudo instructions added to m5. I am wondering if
>     there is a simple way to count the two kinds of instructions between
>     these two calls.
> 
>     Moreover, I saw in the stats file that there is the number of
>     arithmetic, floating, load and store instructions but when I sum these
>     values for all CPUs I don't get the number of simulated instruction
>     given at the beginning of the stats file.
> 
>     Thank you
> 
>     +-------------------------------+
>     | Louisa Bessad                 |
>     | PhD student - LIRMM - Sysmic  |
>     | Bâtiment 4 Bureau 2.92        |
>     +-------------------------------+
>     _______________________________________________
>     gem5-users mailing list
>     [email protected] <mailto:[email protected]>
>     http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> 
> 
> 
> 
> _______________________________________________
> gem5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> 
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to