[edk2] [PATCH] ArmVirtPkg/ArmVirt.dsc.inc: define TcpIoLib resolution unconditionally

Ard Biesheuvel ard.biesheuvel at linaro.org
Fri Dec 14 06:03:18 PST 2018


On Fri, 14 Dec 2018 at 14:56, Laszlo Ersek <lersek at redhat.com> wrote:
>
> On 12/14/18 12:22, Ard Biesheuvel wrote:
> > Commit 9a67ba261fe9 ("ArmVirtPkg: Replace obsoleted network drivers
> > from platform DSC/FDF") failed to take into account that the now
> > unconditionally included IScsiDxe.inf from NetworkPkg requires a
> > resolution for TcpIoLib.
>
> I don't understand why this happened.
>
>
> (a) I warned *precisely* about this issue, when I reviewed the v2
> version of said commit. See bullet (5) in the following message:
>
> http://mid.mail-archive.com/cdd81f4c-1bdc-8bae-63a9-58eb4eb2afbd@redhat.com
>
>
> (b) What's more, my comments for the v3 version were summarily ignored
> as well. See bullet (2) in:
>
> http://mid.mail-archive.com/91d253ae-9d0c-e28b-1bda-1be98cee4340@redhat.com
>
> And now, the BaseCryptLib, OpensslLib and IntrinsicLib resolutions for
> [LibraryClasses.common.UEFI_DRIVER] have been added to
> "ArmVirtPkg/ArmVirtQemuKernel.dsc", despite their being redundant and my
> having pointed out that fact. Worse, "ArmVirtQemuKernel.dsc" uses
> "OpensslLib.inf", while "OpensslLibCrypto.inf" from "ArmVirt.dsc.inc" is
> perfectly sufficient.
>
>
> Commit 9a67ba261fe9 does not carry my R-b, and that's not a random fact.
> The v3 patch was *not* ready for being pushed, to my eyes. And I was
> pretty explicit about that.
>
>
> > Since specifying such a resolution is harmless
> > for platforms that have no networking enabled, let's just fix things
> > by dropping the conditionals around it.
> >
> > Contributed-under: TianoCore Contribution Agreement 1.1
> > Signed-off-by: Ard Biesheuvel <ard.biesheuvel at linaro.org>
> > ---
> >  ArmVirtPkg/ArmVirt.dsc.inc | 2 --
> >  1 file changed, 2 deletions(-)
> >
> > diff --git a/ArmVirtPkg/ArmVirt.dsc.inc b/ArmVirtPkg/ArmVirt.dsc.inc
> > index c3549c84d4c6..89c2db074711 100644
> > --- a/ArmVirtPkg/ArmVirt.dsc.inc
> > +++ b/ArmVirtPkg/ArmVirt.dsc.inc
> > @@ -80,9 +80,7 @@
> >    DpcLib|MdeModulePkg/Library/DxeDpcLib/DxeDpcLib.inf
> >    UdpIoLib|MdeModulePkg/Library/DxeUdpIoLib/DxeUdpIoLib.inf
> >    IpIoLib|MdeModulePkg/Library/DxeIpIoLib/DxeIpIoLib.inf
> > -!if $(NETWORK_IP6_ENABLE) == TRUE
> >    TcpIoLib|MdeModulePkg/Library/DxeTcpIoLib/DxeTcpIoLib.inf
> > -!endif
> >  !if $(HTTP_BOOT_ENABLE) == TRUE
> >    HttpLib|MdeModulePkg/Library/DxeHttpLib/DxeHttpLib.inf
> >  !endif
> >
>
> I'm *very strongly* tempted to simply revert 9a67ba261fe9, for blatantly
> ignoring my explicit requests for updates. However, that would only
> result in my having to review more (possibly incomplete) iterations of
> the patch.
>
> At least, this incremental fix is in line with my request in (a) -- "we
> should make the current TcpIoLib class resolution unconditional". Please
> go ahead and push it.
>
> Reviewed-by: Laszlo Ersek <lersek at redhat.com>
>

Thanks. Apologies for not looking more carefully whether the feedback
had in fact been incorporated before giving my R-b

Pushed as 9a67ba261fe9..a9ff32909b47

> I should really file a TianoCore BZ about the wrong / redundant
> OpensslLib resolution in ArmVirtQemuKernel.dsc too. It's difficult for
> me to find te motivation for that right now, seeing the disregard for my
> earlier reviews.
>

I assumed [incorrectly, I suppose] that people actually build test
platforms that they modify, but apparently not :-)


More information about the edk2-devel mailing list