lkml.org 
[lkml]   [2018]   [Oct]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [RFC PATCH 00/10] arch/x86: AMD QoS support
From
Date
Hi guys,

On 03/10/18 21:09, Moger, Babu wrote:
> On 10/03/2018 01:52 PM, Fenghua Yu wrote:
>> On Tue, Oct 02, 2018 at 07:16:23PM +0000, Moger, Babu wrote:
>>> On 10/02/2018 01:46 PM, Fenghua Yu wrote:
>>>> I don't see the name 'RDT' (Resource Director Technology) in the
>>>> spec "AMD64 Technology Platform Quality of Service Extensions".
>>>> Does AMD use the same name 'RDT' as Intel?
>>>
>>> No. AMD uses the word "Platform Quality of Service"(or in short QoS) to
>>> refer this feature.
>>
>> Hi, Thomas, Babu, and James,
>>
>> Can we use "resctrl" to replace "intel_rdt" and "rdt" in kernel? "resctrl"
>> is a neutral name and has been used in user interface already. Hopefully
>> "resctrl" can be acceptable by ARM, AMD, and Intel. We only use "intel_rdt",
>> "amd_qos", or "arm mpam" for vendor specific code.
>
> Yes. resctrl seems like a reasonable name. >
>> Can we move all of Intel RDT, AMD QoS, and ARM MPAM code into a generic

all? Hmmm... I wanted to keep anything with an ABI implication under /fs/ so its
definitely the same. Obviously some things are definitely arch specific, e.g.
writing msr-s in intel_rdt_sched_in().


>> place fs/resctrl where both ARM and X86 code can stay?
>
> Looking at the James patches, I thought he is going in the same(or
> similar) direction eventually.

Yes, I was trying to name stuff 'resctrl' if its moving to /fs/, and
resctrl_arch if its arch specific, thus stays where it is. (and the existing
names if I didn't need to touch it).

The struct names all have rdt in them, I thought it too noisy to change them.



Thanks,

James

\
 
 \ /
  Last update: 2018-10-05 18:20    [W:0.075 / U:0.780 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site