> On June 27, 2013, 4:57 p.m., Nathan Binkert wrote:
> > I'm curious what the intentions of this are?  Is this simply a mechanism to 
> > provide voltage information for functional blocks so that it can more 
> > easily be paired with the stats to feed into a power simulator?  Do you 
> > plan to implement power management instructions, etc?  Do you plan to 
> > integrate a power model?  Inquiring minds want to know!
> > 
> > BTW: you spelled "volate" wrong in the commit message.
> > 
> > This commit makes me wonder if the name "ClockedObject" is right anymore.  
> > Seems like we need to separate the notion of PhysicalObject (which all have 
> > power and frequency) and VirtualObject (which don't).  Are there any other 
> > subdivisions?  I don't want to change the name for the sake of changing a 
> > name if this doesn't make sense.  (And of course all that could be a 
> > separate commit).

In short the goal is to enable modelling of DVFS (regulation). This involves 
expressing VF (i.e. the current patch), controlling VF, and also reacting to 
changes (evaluating the outcome using power/thermal models). So, yes, yes and 
yes.

Thanks for the spell check :-)


- Andreas


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviews.gem5.org/r/1937/#review4475
-----------------------------------------------------------


On June 27, 2013, 4:40 p.m., Andreas Hansson wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviews.gem5.org/r/1937/
> -----------------------------------------------------------
> 
> (Updated June 27, 2013, 4:40 p.m.)
> 
> 
> Review request for Default.
> 
> 
> Repository: gem5
> 
> 
> Description
> -------
> 
> Changeset 9801:9b1a216bdc52
> ---------------------------
> power: Add voltage domains to the clock domains
> 
> This patch adds the notion of voltage domains, and groups clock
> domains that operate under the same voltage (i.e. power supply) into
> domains. Each clock domain is required to be associated with a voltage
> domain, and the latter requires the voltage to be explicitly set.
> 
> A voltage domain is an independently controllable voltage supply being
> provided to section of the design. Thus, if you wish to perform
> dynamic voltage scaling on a CPU, its clock domain should be
> associated with a separate voltage domain.
> 
> The current implementation of the volate domain does not take into
> consideration cases where there are derived voltage domains running at
> ratio of native voltage domains, as with the case where there can be
> on-chip buck/boost (charge pumps) voltage regulation logic.
> 
> The regression and configuration scripts are updated with a generic
> voltage domain for the system, and one for the CPUs.
> 
> 
> Diffs
> -----
> 
>   configs/common/Options.py 9cd5f91e7a79 
>   configs/example/fs.py 9cd5f91e7a79 
>   configs/example/se.py 9cd5f91e7a79 
>   src/python/m5/params.py 9cd5f91e7a79 
>   src/python/m5/util/convert.py 9cd5f91e7a79 
>   src/sim/ClockDomain.py 9cd5f91e7a79 
>   src/sim/SConscript 9cd5f91e7a79 
>   src/sim/VoltageDomain.py PRE-CREATION 
>   src/sim/clock_domain.hh 9cd5f91e7a79 
>   src/sim/clock_domain.cc 9cd5f91e7a79 
>   src/sim/voltage_domain.hh PRE-CREATION 
>   src/sim/voltage_domain.cc PRE-CREATION 
>   tests/configs/base_config.py 9cd5f91e7a79 
>   tests/configs/memtest-ruby.py 9cd5f91e7a79 
>   tests/configs/memtest.py 9cd5f91e7a79 
>   tests/configs/rubytest-ruby.py 9cd5f91e7a79 
>   tests/configs/simple-timing-ruby.py 9cd5f91e7a79 
>   tests/configs/tgen-simple-dram.py 9cd5f91e7a79 
>   tests/configs/tgen-simple-mem.py 9cd5f91e7a79 
>   tests/configs/twosys-tsunami-simple-atomic.py 9cd5f91e7a79 
> 
> Diff: http://reviews.gem5.org/r/1937/diff/
> 
> 
> Testing
> -------
> 
> All regressions pass. Ran a number of the example scripts.
> 
> 
> Thanks,
> 
> Andreas Hansson
> 
>

_______________________________________________
gem5-dev mailing list
[email protected]
http://m5sim.org/mailman/listinfo/gem5-dev

Reply via email to