For fans of the classics, you can look up my email with subject "Merging of
FS and SE mode" from 2011, but I suspect that's too old for most people to
have seen, and not completely the same as what I'm trying to do here. This
is the other half of the workload concept which is already applied to FS
mode, ie separating and modularizing the thing you're running from the
thing you're running on. The "hardware" like the CPU, memory, devices, etc,
is the same no matter if you're running SE mode programs, FS operating
systems, an OS from scratch, bare metal, etc. Then on top of that, you can
add a workload object which will load an OS for you like the system objects
used to do and the workload objects do now, or which fakes an OS kernel and
loads and coordinates SE mode processes which is what these new changes are
setting up, or sets up SE mode like hooks for a BIOS to implement to fake
power modes or BIOS interrupt routines for something like DOS, etc, etc.

A long time ago I managed to break down the barriers between SE and FS
modes to the point where they could be combined into a single build, but
there is still a big FS/SE mode switch which means you have to pick which
one you want, and a decent number of places where global behaviors or
mechanisms are selected with that switch. These changes are a big step
towards not having that switch at all, and for SE mode and FS mode to not
be modes, but to just be handy labels to put on configurations which tend
to fit in certain categories. Other than just making gem5 cleaner and
simpler, this will also enable some configurations which are just not
possible today, like systems which fall outside of the FS or SE dichotomy,
or which combine them and have, for instance, an FS mode server connected
to SE mode clients over a simulated network.

Gabe

On Thu, Oct 22, 2020 at 8:06 AM Jason Lowe-Power via gem5-dev <
[email protected]> wrote:

> Hi Gabe,
>
> Could you give us some context to the 16 changes that you are going to
> commit without review at the end of the week? Starting with
> https://gem5-review.googlesource.com/c/public/gem5/+/33902/
>
> I would like to review these, but I don't have any context as to the
> purpose. What problem are they solving and how do they solve the problem?
>
> Sorry that I missed these for two months. You've contributed over 116
> changes in the past two months. That requires others to review about two
> changes per day. There's bound to be things that fall through the cracks.
>
> Thanks,
> Jason
> _______________________________________________
> gem5-dev mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
> %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
_______________________________________________
gem5-dev mailing list -- [email protected]
To unsubscribe send an email to [email protected]
%(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s

Reply via email to