public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: "Nhi Pham" <nhi@os.amperecomputing.com>
To: Rebecca Cran <rebecca@bsdio.com>,
	devel@edk2.groups.io, ardb@kernel.org,
	Rebecca Cran <rebecca@quicinc.com>
Cc: nhi@os.amperecomputing.com, patches@amperecomputing.com,
	quic_llindhol@quicinc.com, ardb+tianocore@kernel.org,
	Tinh Nguyen <tinhnguyen@os.amperecomputing.com>,
	harb@amperecomputing.com
Subject: Re: [edk2-devel] [edk2-platforms][PATCH 1/1] AmpereAltraPkg: Update ArmPlatformLib to work with changed ARM_CORE_INFO
Date: Tue, 31 Jan 2023 13:35:50 +0700	[thread overview]
Message-ID: <3b16465b-6f2a-8bda-98b5-ddfbef68aa41@amperemail.onmicrosoft.com> (raw)
In-Reply-To: <923d3222-775c-f0e8-b4c1-4c10f122ad0a@bsdio.com>

Hi Rebecca,

++ Harb who can give more insights on this. FYI, the original concern is 
https://edk2.groups.io/g/devel/message/98482

On 1/18/2023 1:21 AM, Rebecca Cran wrote:
> On 1/17/23 09:40, Ard Biesheuvel wrote:
>> On Tue, 17 Jan 2023 at 13:55, Rebecca Cran<rebecca@quicinc.com>  wrote:
>>> I was under the impression that this is becoming a more standard format?
>>>
>> If this is not defined in an ARM spec somewhere, we shouldn't add it
>> to ArmPkg at this point.
>
> From what I've found, the ARM specs such as the Arm Architecture 
> Reference Manual for A-profile architecture don't define the meaning 
> of the affinity fields? That appears to be left up to the individual 
> Arm core TRMs.
>
I think so. This might be silicon specific implementation.

Per Arm Armv8-A Architecture Registers 
(https://developer.arm.com/documentation/ddi0595/2021-12/AArch64-Registers/MPIDR-EL1--Multiprocessor-Affinity-Register), 
if I interpret correctly, the AFF0 will give core ID or thread ID based 
on the MT bit in the MPIDR register.

I think we should remove the following definitions particularly for 
getting core id and cluster in ArmPkg/Include/Library/ArmLib.h to avoid 
the confusion to others

#define ARM_CORE_MASK         ARM_CORE_AFF0
#define ARM_CLUSTER_MASK      ARM_CORE_AFF1
#define GET_CORE_ID(MpId)     ((MpId) & ARM_CORE_MASK)
#define GET_CLUSTER_ID(MpId)  (((MpId) & ARM_CLUSTER_MASK) >> 8)
#define GET_MPID(ClusterId, CoreId)   (((ClusterId) << 8) | (CoreId))
#define PRIMARY_CORE_ID       (PcdGet32(PcdArmPrimaryCore) & ARM_CORE_MASK)

And, should support GET_AFFx like

#define ARM_MPIDR_GET_AFF0(Mpid) ((Mpid) & ARM_CORE_AFF0)

For silicon specific usage, it can abstract them to proper IDs like

#define XXX_GET_CORE_ID(Mpid) ARM_MPIDR_GET_AFF0(Mpid)

Thanks,

Nhi

> For example, the Cortex-X2 TRM says:
>
> Affinity level 0. This is the affinity level that is most significant 
> for determining PE behavior. Higher affinity
> levels are increasingly less significant in determining PE behavior. 
> The assigned value of the MPIDR.{Aff2,
> Aff1, Aff0} or AArch64-MPIDR_EL1.{Aff3, Aff2, Aff1, Aff0} set of 
> fields of each PE must be unique within the
> system as a whole.
> 0b00000000
> Only one thread.
>
>
> Affinity level 1. See the description of Aff0 for more information.
> Value read from the CPUID configuration pins. Identification number 
> for each CPU in an cluster counting from
> zero.
>
>
> Affinity level 2. See the description of Aff0 for more information.
> The value will be determined by the CLUSTERIDAFF2 configuration pins.
>
>
> Affinity level 3. See the description of Aff0 for more information.
> The value will be determined by the CLUSTERIDAFF3 configuration pins.
>


  reply	other threads:[~2023-01-31  6:36 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-13  4:21 [edk2-platforms][PATCH 1/1] AmpereAltraPkg: Update ArmPlatformLib to work with changed ARM_CORE_INFO Nhi Pham
2023-01-13 14:40 ` [edk2-devel] " Rebecca Cran
2023-01-17  9:53   ` Nhi Pham
2023-01-17 12:55     ` Rebecca Cran
2023-01-17 16:40       ` Ard Biesheuvel
2023-01-17 18:21         ` Rebecca Cran
2023-01-31  6:35           ` Nhi Pham [this message]
2023-02-15 11:28             ` Leif Lindholm

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-list from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3b16465b-6f2a-8bda-98b5-ddfbef68aa41@amperemail.onmicrosoft.com \
    --to=devel@edk2.groups.io \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox