public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: "Laszlo Ersek" <lersek@redhat.com>
To: Benjamin Doron <benjamin.doron00@gmail.com>, devel@edk2.groups.io
Subject: Re: [edk2-devel] [Patch V2 0/2] Let AcpiTableDxe driver install Acpi table from Hob
Date: Wed, 24 Mar 2021 19:33:08 +0100	[thread overview]
Message-ID: <2e602df8-b1fc-193b-f61a-86802ace9108@redhat.com> (raw)
In-Reply-To: <8402.1616604928827766426@groups.io>

On 03/24/21 17:55, Benjamin Doron wrote:
>>
>>
>>> Hi all,
>>> Would it be acceptable/feasible for AcpiTableDxe or AcpiPlatformDxe (in
>>> MdeModulePkg) to use `EfiGetSystemConfigurationTable` to get the RSDP
>>> and then install the tables? It's a solution that uses the regular
>>> UefiLib, so it avoids platform-specific quirks (and as I see it, if RSDP
>>> is in the configuration table, we probably always want those tables).
>>
>> I'm sorry, I don't understand how this would help.
> 
> As I understand it, the issue is that this patchset changes MdeModulePkg to do platform-specific parsing.
> 
> Perhaps it would be an acceptable solution for platforms to retrieve the tables, then add
> RSDP/them to the configuration table to be installed by AcpiTableDxe/AcpiPlatformDxe.
> This allows MdeModulePkg to abstract away the parsing, only installing tables
> available to it.

But this is already the best approach, and already what's happening --
when you call EFI_ACPI_TABLE_PROTOCOL.InstallAcpiTable() from the
platform's AcpiPlatformDxe, that's *how* you tell AcpiTableDxe in
MdeModulePkg to pick up the table and hook it into the RSDT / XSDT /
wherever, and also to manage RSD PTR as a UEFI configuration table.

Are you (more or less) proposing a new EFI_ACPI_TABLE_PROTOCOL member
function for taking a forest of ACPI tables, passed in by RSD PTR?

Sorry about being dense. :)

> (Currently, UefiPayloadPkg's BlSupportDxe retrieves the data from a HOB and calls
> `gBS->InstallConfigurationTable` with the address of RSDP).
> 
> I understand that this may not work for OVMF if tables are located differently in memory.
> 
>>
>>
>>> Regarding UefiPayloadPkg: AcpiTableDxe is currently compiled (listed in
>>> DSC) but not added to a FV (not listed in FDF). So, how has this been
>>> tested?
>>
>> I assume through an out-of-tree platform. Many such core modules exist
>> in edk2 that are not consumed by any of the virtual platforms in the
>> edk2 repo itself (EmulatorPkg, ArmVirtPkg, OvmfPkg).
> 
> This makes sense, but AcpiTableDxe must be added to UefiPayloadPkg's FDF
> if patch 2/2 is merged. Otherwise, ACPI tables will not be advertised.
> 

Good point; I missed that. (I'm not familiar with the UefiPayloadPkg FDF
at all.)

Laszlo


  reply	other threads:[~2021-03-24 18:33 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-23  3:24 [Patch V2 1/2] MdeModulePkg/ACPI: Install ACPI table from HOB Zhiguang Liu
2021-03-23  3:24 ` [Patch V2 2/2] UefiPayloadPkg: Remove code that installs APCI Zhiguang Liu
2021-03-23  3:44   ` Ni, Ray
2021-03-23  5:19   ` Guo Dong
2021-03-23  3:24 ` [Patch V2 0/2] Let AcpiTableDxe driver install Acpi table from Hob Zhiguang Liu
2021-03-23 12:40   ` [edk2-devel] " Laszlo Ersek
2021-03-23 15:45     ` Guo Dong
2021-03-23 16:12       ` Andrew Fish
2021-03-23 17:29         ` Guo Dong
2021-03-24  5:30         ` Ni, Ray
2021-03-23 16:48       ` Laszlo Ersek
2021-03-23 17:15         ` Guo Dong
2021-03-24  9:50           ` Laszlo Ersek
2021-03-24  4:09         ` Ni, Ray
2021-03-24 10:29           ` Laszlo Ersek
2021-03-23 23:52     ` Benjamin Doron
2021-03-24  9:53       ` Laszlo Ersek
2021-03-24 16:55         ` Benjamin Doron
2021-03-24 18:33           ` Laszlo Ersek [this message]
2021-03-25  1:10             ` Ni, Ray
2021-03-25  3:55               ` Andrew Fish
2021-03-25 17:35                 ` Laszlo Ersek
2021-03-25 17:33               ` Laszlo Ersek
2021-03-25  1:39             ` Benjamin Doron
2021-03-23  3:44 ` [Patch V2 1/2] MdeModulePkg/ACPI: Install ACPI table from HOB Ni, Ray

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=2e602df8-b1fc-193b-f61a-86802ace9108@redhat.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