[edk2-devel] [PATCH v2 2/6] OvmfPkg/AmdSev: add Grub Firmware Volume Package

[James Bottomley]

On Wed, 2020-11-25 at 00:22 +0100, Laszlo Ersek wrote:
> On 11/24/20 17:22, James Bottomley wrote:
> > On Tue, 2020-11-24 at 16:58 +0100, Laszlo Ersek wrote:
> > > On 11/24/20 15:54, Laszlo Ersek wrote:
> > [...]
> > > Perhaps I can make that recommendation / request more detailed
> > > too:
> > > 
> > > 
> > > * drop:
> > > 
> > > - UefiRuntimeServicesTableLib: no use of "gRT"
> > > 
> > > - ReportStatusCodeLib: commit 0a0566d5edad is not relevant,
> > > because
> > > we justifiedly removed TryRunningQemuKernel()
> > > 
> > > - XenPlatformLib: as discussed before; substitute FALSE for each
> > > XenDetected() call, and compress the resultant code
> > 
> > All dropped.
> > 
> > > * keep:
> > > 
> > > - BaseLib: for CpuDeadLoop()
> > > 
> > > - MemoryAllocationLib: for FreePool()
> > > 
> > > - UefiBootServicesTableLib: for gBS->xxx()
> > > 
> > > - BaseMemoryLib: for CompareMem()
> > > 
> > > - DebugLib: for DEBUG() and ASSERT()
> > > 
> > > - PcdLib: for PcdGet16 (PcdOvmfHostBridgePciDevId)
> > > 
> > > - UefiBootManagerLib: for APIs central to the functionality of
> > > PlatformBootManagerLibGrub
> > > 
> > > - BootLogoLib: for BootLogoEnableLogo()
> > > 
> > > - DevicePathLib: for a bunch of device path manipulation
> > > 
> > > - PciLib: mainly for the functions called in
> > > PciAcpiInitialization()
> > > 
> > > - UefiLib: for EfiEventGroupSignal() etc
> > > 
> > > - PlatformBmPrintScLib: for PlatformBmPrintScRegisterHandler() --
> > > this is responsible for printing the boot option processing steps
> > > to
> > > the UEFI console
> > > 
> > > - Tcg2PhysicalPresenceLib: for
> > > Tcg2PhysicalItPresenceLibProcessRequest() -- we preserve TPM
> > > support
> > > 
> > > 
> > > Then leaving the trimming of [Packages] to the end makes sense --
> > > after trimming everything else, try to remove each package DEC in
> > > isolation, and see if the lib instance continues to build.
> > 
> > I can't seem to remove any packages.  Even the source debug agent
> > one
> > is used by the uart devices.
> 
> OK! Thanks for checking.
> 
> > It seems I can get rid of PcdFSBClock
> > 
> > and in the protocols section (with some header file removal)
> > 
> > gEfiDecompressProtocolGuid
> > gEfiS3SaveStateProtocolGuid
> 
> Yes, I agree removing these is correct; they were also included in my
> other email at <https://edk2.groups.io/g/devel/message/67892>;.
> 
> There are some others that should be possible to remove (pls. refer
> to the rest of that email).

Heh, well, I spoke too soon.  Even though the OVMF this produces boots
to grub and decrypts an encrypted volume, the kernel boot panics
because of something missing in the runtime ... it looks like it's
tripping over gRT->SetVariable, so I'm going to have to start putting
some stuff back again ...

James




-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#67913): https://edk2.groups.io/g/devel/message/67913
Mute This Topic: https://groups.io/mt/78455898/1813853
Group Owner: devel+owner at edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [edk2-devel-archive at redhat.com]
-=-=-=-=-=-=-=-=-=-=-=-