[PATCH 11/35] RFC: efi: Drop code that doesn't work with driver model

[Mark Kettenis]

> Date: Thu, 9 Sep 2021 16:23:08 -0400
> From: Tom Rini <trini at konsulko.com>
> 
> On Thu, Sep 09, 2021 at 10:15:44PM +0200, Mark Kettenis wrote:
> > > From: Simon Glass <sjg at chromium.org>
> > > Date: Thu, 9 Sep 2021 13:57:39 -0600
> > > 
> > > Hi Heinrich,
> > > 
> > > On Thu, 9 Sept 2021 at 03:26, Heinrich Schuchardt <xypron.glpk at gmx.de> wrote:
> > > >
> > > >
> > > >
> > > > On 9/9/21 10:57 AM, Simon Glass wrote:
> > > > > Hi Heinrich,
> > > > >
> > > > > On Wed, 8 Sept 2021 at 11:44, Heinrich Schuchardt <xypron.glpk at gmx.de> wrote:
> > > > >>
> > > > >>
> > > > >>
> > > > >> On 9/8/21 3:33 PM, Simon Glass wrote:
> > > > >>> This code should never have been added as it builds a new feature on top
> > > > >>> of legacy code. Drop it and add a dependency on BLK for this feature.
> > > > >>>
> > > > >>> Boards which want EFI_LOADER should migrate to driver model first.
> > > > >>>
> > > > >>> Signed-off-by: Simon Glass <sjg at chromium.org>
> > > > >>
> > > > >> This patch is not related to the rest of the series and the code has a
> > > > >> different maintainer.
> > > > >>
> > > > >> So, please, separate it from the series.
> > > > >
> > > > > Who is the maintainer?
> > > >
> > > > Until 623b3a57976 ("efi_selftest: provide an EFI selftest application")
> > > > there was no official maintainer for lib/efi/ but you were the main
> > > > contributor.
> > > >
> > > > But with that patch directory lib/efi/ was assigned to EFI PAYLOAD.
> > > >
> > > > I am happy if you would continue to care about U-Boot on EFI.
> > > 
> > > OK.
> > > 
> > > >
> > > > >
> > > > > I need this patch for this series to work. You can still review things
> > > > > for other maintainers and in this case it is common for one maintainer
> > > > > to pick up the series once the others are happy.
> > > >
> > > > The direction of this patch is completely correct.
> > > >
> > > > There are some things that will have to be changed, e.g we should not
> > > > require CONFIG_DM_ETH=y. I will work on reviewing this patch in detail.
> > > 
> > > OK, but why not require DM_ETH? The deadline passed a year ago.
> > 
> > Because we support boards without network ports?
> 
> Boards without networking should disable the relevant code, and as
> needed the EFI code return the proper error code?

Yes, but it means you can't make DM_ETH a (hard) requirement for
EFI_LOADER support.  What I mean is that it should still be possible
to build U-Boot with EFI_LOADER support even if DM_EFI isn't set for a
board.  It should just result in a UEFI implementation with no network
support instead.