ruby: memory controllers now inherit from an abstract "MemoryControl" class

Information
Submitter:	Brad Beckmann
Repository:	gem5
Branch:	default
Bugs:
Depends On:
Reviewers
Groups:	Default
People:

Description

Changeset 9075:3b3c9db76ebc
---------------------------
ruby: memory controllers now inherit from an abstract "MemoryControl" class

Testing Done

Issue Summary

Description	From	Last Updated	Status
MemoryControlParams?	Andreas Hansson	July 10, 2012, 9:21 a.m.	Resolved
Is init, drain, wakeup etc, not implicit from the SimObject?	Andreas Hansson	July 10, 2012, 9:22 a.m.	Resolved
I thought this was a single channel controller?	Andreas Hansson	July 10, 2012, 9:30 a.m.	Dropped
Why not constructor initialisation list?	Andreas Hansson	July 10, 2012, 9:30 a.m.	Resolved
If it always has one, why a heap allocated object and not just a member?	Andreas Hansson	July 10, 2012, 9:30 a.m.	Resolved
whitespace	Andreas Hansson	July 10, 2012, 9:30 a.m.	Resolved
whitespace	Andreas Hansson	July 10, 2012, 9:30 a.m.	Resolved
whitespace	Andreas Hansson	July 10, 2012, 9:31 a.m.	Resolved
Could this possibly work? What if there are message in the queues?	Andreas Hansson	July 10, 2012, 9:47 a.m.	Dropped
-2012?	Andreas Hansson	July 10, 2012, 9:47 a.m.	Resolved
Why?	Andreas Hansson	July 10, 2012, 9:47 a.m.	Resolved
Some comments would be great, but I understand this is just a rename/copy-paste.	Andreas Hansson	July 10, 2012, 9:48 a.m.	Dropped

A little off topic, but this makes me wonder... are we ever going to be able to merge ruby and classic memory controller models?  It seems like that would be useful.

Andreas Hansson July 4, 2012, 5:56 p.m. (July 4, 2012, 5:56 p.m.)

I think the first and most important step towards this is to split the functionality that relates to the cache maintenance, and the "pure" memory controller functionality. If that is done, I do not see why the memory controller themselves could not be living outside of Ruby and be generic enough to fit in both worlds.

The question is, who could do it, and when?

src/mem/ruby/system/MemoryControl.hh (Diff revision 1)

MemoryControlParams?

Show all issues

src/mem/ruby/system/MemoryControl.hh (Diff revision 1)

Is init, drain, wakeup etc, not implicit from the SimObject?

Show all issues

src/mem/ruby/system/MemoryControl.hh (Diff revision 1)

I thought this was a single channel controller?

Show all issues

Brad Beckmann July 10, 2012, 9:48 a.m. (July 10, 2012, 9:48 a.m.)

There are other memory controllers that inherit from this base class that implement multi-channel controllers.

src/mem/ruby/system/RubyMemoryControl.cc (Diff revision 1)

Why not constructor initialisation list?

Show all issues

src/mem/ruby/system/RubyMemoryControl.cc (Diff revision 1)

If it always has one, why a heap allocated object and not just a member?

Show all issues

src/mem/ruby/system/RubyMemoryControl.cc (Diff revision 1)

whitespace

Show all issues

src/mem/ruby/system/RubyMemoryControl.cc (Diff revision 1)

whitespace

Show all issues

src/mem/ruby/system/RubyMemoryControl.cc (Diff revision 1)

whitespace

Show all issues

src/mem/ruby/system/RubyMemoryControl.cc (Diff revision 1)

Could this possibly work? What if there are message in the queues?

Show all issues

Brad Beckmann July 10, 2012, 9:48 a.m. (July 10, 2012, 9:48 a.m.)

I actually added this drain function to the memory controll several months ago (8937:225590437eb2).  This function was just copied from that patch.  The reason for immediate descheduling of the mem cntrl event is that the cntrl is clock driven and uses the event as a repeatable tick.  If I didn't deschedule it, it would continue forever.

That being said, your concerns are justified.  The RubyPort is responsible for ensuring ruby drains properly, but it is not clear to me that there could be a corner case where ruby's drain can complete because the mem cntrl already went to sleep.  In practice I haven't seen that problem and it is a situation that should be obvious when it occurs.  When it does, we'll fix it appropriately.  However, I don't want to fix a bug that I have yet to see.  Especially because the memory controller code is not my speciality.

src/mem/ruby/system/RubyMemoryControl.py (Diff revision 1)

-2012?

Show all issues

src/mem/ruby/system/RubyMemoryControl.py (Diff revision 1)

Why?

Show all issues

src/mem/ruby/system/RubyMemoryControl.py (Diff revision 1)

Some comments would be great, but I understand this is just a rename/copy-paste.

Show all issues

Diff:

Revision 2 (+976 -752)

Show changes

	src/mem/ruby/system/MemoryControl.py
	src/mem/ruby/system/MemoryControl.hh
	src/mem/ruby/system/MemoryControl.cc
	src/mem/ruby/system/RubyMemoryControl.py
	src/mem/ruby/system/RubyMemoryControl.hh
	src/mem/ruby/system/RubyMemoryControl.cc
	src/mem/ruby/system/SConscript
	src/mem/slicc/symbols/StateMachine.py

Ship It!

You have a pending review.

Review Board 2.0.15

This change has been marked as submitted.

Screenshots

Files

Issue Summary

Diff:

Status: Closed (submitted)