public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: "Ard Biesheuvel" <ard.biesheuvel@arm.com>
To: Laszlo Ersek <lersek@redhat.com>,
	devel@edk2.groups.io, guomin.jiang@intel.com
Subject: Re: [edk2-devel] [PATCH 00/18] Remove All UGA Support
Date: Fri, 8 May 2020 13:00:20 +0200	[thread overview]
Message-ID: <0cfa0d84-8637-73ec-5d36-50182169cd67@arm.com> (raw)
In-Reply-To: <a45b1be3-2b73-4138-3795-4dd724fa72ca@redhat.com>

On 5/8/20 12:09 PM, Laszlo Ersek wrote:
> Hello Guomin,
> 
> On 05/08/20 10:38, Guomin Jiang wrote:
>> REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2368
>>
>> UGA is replaced by GOP and remove all related code.
> 
> I'm responding under the cover letter.
> 
> 
> (1) You should have CC'd the cover letter to each person that is CC'd on
> at least one patch in the series. Please do that in the future.
> 
> I'm CC'ing Ard now, at least.
> 

Thanks Laszlo

> 
> (2) The series currently has identical subjects for some of the patches.
> For example:
> 
> [PATCH 08/18] OvmfPkg: Remove All UGA Support
> [PATCH 15/18] OvmfPkg: Remove All UGA Support
> 
> [PATCH 10/18] ArmVirtPkg: Remove All UGA Support
> [PATCH 17/18] ArmVirtPkg: Remove All UGA Support
> 
> This is extremely bad practice. Please don't do this.
> 
> 
> (3) The commit messages do a bad job explaining why the PCDs are being
> removed.
> 
> If I understand correctly, the core modules are updated in this series
> as follows:
> 
> - PcdConOutUgaSupport is assumed constant FALSE
> - PcdConOutGopSupport is assumed constant TRUE
> - conditions are simplified and dead code is eliminated
> 
> 
> The proper approach is therefore the following:
> 
> - In patch#1, change the DEC default value of PcdConOutUgaSupport from
> TRUE to FALSE.
> 
> - In patches #2 .. #n, remove the PCD settins from all the DSC files.
> Just one patch per package is sufficient. (No need for a separate patch
> per PCD per Package.)
> 
> The commit messages on these patches should explain that the PCDs are
> going away, and by deleting the DSC settings, the platforms in question
> inherit the DEC defaults. And, at this point the DEC defaults are:
> 
> PcdConOutUgaSupport = FALSE
> PcdConOutGopSupport = TRUE
> 
> which is going to be the only configuration supported by edk2 in the future.
> 
> For example, patch#2 could be for ArmVirtPkg, patch#3 could be for OvmfPkg.
> 
> - In the rest of the patches, simplify code / eliminate dead code by
> substituting FALSE for PcdConOutUgaSupport, and TRUE for
> PcdConOutGopSupport. If you want, you can keep the code changes for each
> PCD separate, in every module. (This kind of separation may be a good idea.)
> 
> - In the last patch in the series, remove both PCDs from the DEC file.
> 
> 
> Before posting v2, please verify that the series (including the
> platforms in edk2) build fine at every stage (at every patch in the series).
> 

Also, looking at edk2-platforms

$ git grep -l PcdUga
Platform/AMD/OverdriveBoard/OverdriveBoard.dsc
Platform/Comcast/RDKQemu/RDKQemu.dsc
Platform/Intel/SimicsOpenBoardPkg/BoardX58Ich10/OpenBoardPkgPcd.dsc
Platform/Intel/SimicsOpenBoardPkg/Library/DxeLogoLib/DxeLogoLib.inf
Platform/Intel/SimicsOpenBoardPkg/Library/DxeLogoLib/Logo.c
Platform/RaspberryPi/Library/PlatformBootManagerLib/PlatformBootManagerLib.inf
Platform/RaspberryPi/RPi3/RPi3.dsc
Platform/RaspberryPi/RPi4/RPi4.dsc
Platform/Socionext/DeveloperBox/DeveloperBox.dsc.inc
Platform/Socionext/SynQuacerEvalBoard/SynQuacerEvalBoard.dsc
Silicon/Marvell/Armada7k8k/Armada7k8k.dsc.inc
Silicon/NXP/NxpQoriqLs.dsc.inc

Those platforms will all be broken as soon as you drop the PCD definitions.


  reply	other threads:[~2020-05-08 11:00 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-08  8:38 [PATCH 00/18] Remove All UGA Support Guomin Jiang
2020-05-08  8:38 ` [PATCH 01/18] BaseTools: " Guomin Jiang
2020-05-09  1:47   ` [edk2-devel] " Ni, Ray
2020-05-08  8:38 ` [PATCH 02/18] UefiPayloadPkg: " Guomin Jiang
2020-05-08 14:57   ` Ma, Maurice
2020-05-08  8:38 ` [PATCH 03/18] ShellPkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 04/18] MdeModulePkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 05/18] MdeModulePkg/ConSplitterDxe: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 06/18] MdeModulePkg/GraphicsConsoleDxe: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 07/18] EmulatorPkg: " Guomin Jiang
2020-05-09  3:02   ` Ni, Ray
2020-05-08  8:38 ` [PATCH 08/18] OvmfPkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 09/18] ArmPkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 10/18] ArmVirtPkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 11/18] MdePkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 12/18] MdeModulePkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 13/18] MdeModulePkg/ConSplitterDxe: " Guomin Jiang
2020-05-09  3:03   ` Ni, Ray
2020-05-08  8:38 ` [PATCH 14/18] " Guomin Jiang
2020-05-08  8:38 ` [PATCH 15/18] OvmfPkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 16/18] UefiPayloadPkg: " Guomin Jiang
2020-05-08 14:58   ` Ma, Maurice
2020-05-08  8:38 ` [PATCH 17/18] ArmVirtPkg: " Guomin Jiang
2020-05-08  8:38 ` [PATCH 18/18] MdeModulePkg: " Guomin Jiang
2020-05-08 10:09 ` [edk2-devel] [PATCH 00/18] " Laszlo Ersek
2020-05-08 11:00   ` Ard Biesheuvel [this message]
2020-05-08 13:04     ` Guomin Jiang

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=0cfa0d84-8637-73ec-5d36-50182169cd67@arm.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