Re: [PATCH RFC] soc: fujitsu: Add cache driver code

From: James Morse <james.morse@arm.com>
To: "tan.shaopeng@fujitsu.com" <tan.shaopeng@fujitsu.com>
Cc: 'Arnd Bergmann' <arnd@arndb.de>, Will Deacon <will@kernel.org>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Olof Johansson <olof@lixom.net>,
	Linux ARM <linux-arm-kernel@lists.infradead.org>,
	SoC Team <soc@kernel.org>,
	"misono.tomohiro@fujitsu.com" <misono.tomohiro@fujitsu.com>
Subject: Re: [PATCH RFC] soc: fujitsu: Add cache driver code
Date: Thu, 1 Apr 2021 17:15:09 +0100	[thread overview]
Message-ID: <edaea436-d94e-d862-1136-66e9f881e0cc@arm.com> (raw)
In-Reply-To: <OSAPR01MB21463C735D4A07344F35E7CC8B7C9@OSAPR01MB2146.jpnprd01.prod.outlook.com>

Hi Tan Shaopeng,

On 31/03/2021 09:52, tan.shaopeng@fujitsu.com wrote:
>>> Given that all of this is outside of the scope of the architecture, I
>>> don't think that arch/arm64/ is the right place for it. Perhaps this
>>> would fit into the resctrl rework that James has been doing for MPAM?
>>
>> Indeed, that sounds like a good starting point. I don't understand enough about
>> either of the two to be sure, but it sounds like there is some overlap in
>> functionality, and ideally we would have one user interface that can deal with
>> all the hardware implementations (intel, arm, fujitsu and any future ones).

> According to my study, it seems possible to add A64FX's sector cache function
> to resctrl.

I think this depends on whether it maps to one of Intel RDT's existing schema.
While adding new ones looks easy, it is bad for user-space as they are not portable
between machines that support resctrl.

> I heard you are working on resctrl rework for MPAM , 
> but I cannot find related patches on ML archive. Therefore, 
> I would like to know the status about MPAM support. 

Its about five or six series that refactor resctrl inside arch/x86 to have a clear
boundary between arch-specific code and the resctrl code that implements the filesystem,
then pull it out to /fs/ and build the MPAM driver to make it work for arm64.

Unfortunately, its rather large (>100 patches), so will take some time to be reviewed.

> (1) I think the first step is to support resctrl for ARM arch. 
>    Have you finished the work of arm support and when will you release it? 

No-one wants a second copy of the code to implement resctrl, as this will introduce subtle
bugs that user-space would have to work around, and make it harder to merge later.

The first step is to refactor the arch/x86 implementation of resctrl so that the parts
that are visible to user-space can be moved somewhere that is common to multiple
architectures.

I've posted the next chunk of that work here:
https://lore.kernel.org/lkml/20210312175849.8327-1-james.morse@arm.com/

> (2) When will you release MPAM patch?

The latest complete version of the tree is here:
https://git.kernel.org/pub/scm/linux/kernel/git/morse/linux.git/log/?h=mpam/snapshot/feb

It is over a year old, and has plenty of bugs.

I hope to push a newer version of the branch once I have a complete copy of the tree that
has been tested, and is based on the current version of the ACPI table.

Thanks,

James

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel