Re: [PATCH 6/7] [RFC] Support multiply-bindable cgroup subsystems

[Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>]

Paul Menage wrote:
> On Tue, Mar 17, 2009 at 7:09 PM, Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org> wrote:
>> Ah, I just notice debug_subsys will never be in rootnode.
> 
> Right - this is the first of the "Open Issues" that I mentioned in the
> patch. The question is how to display the multiply-bindable subsystems
> in /proc/cgroup in a way that won't break backward-compatibility.
> Options include:
> 
> - always display one line for each subsystem; if a subsystem is
> multiply-bindable then the "hierarchy id" and "num cgroups" columns
> may be empty or have multiple (comma or slash-separated?) values
> 

Then "hierarchy id" is no long a single number..

> - for a multiply-bindable subsystem, have a header line to indicate
> that the subsystem exists, and then a separate line for each bound
> instance of the subsystem.
> 

I think it's better to show every subsystems including debug_subsys in
/proc/cgroups, and show exactly n lines of debug_susys if we have n
hierarcies with debug_subsys binded, but no header line.

> Either of these could confuse existing userspace parsers, but I'm not

And some cgroup test programs, like controller tests in ltp and some
test programs that I wrote, but they are ok for this change.

> sure that any exist other than libcgroup, and I don't think it's in
> wide use yet.
>