From: Sebastien Boeuf <sebastien.boeuf@...> Instead of using the CMOS, the CloudHv platform relies on the list of memmap entries provided through the PVH boot protocol to determine the last RAM

From: Sebastien Boeuf <sebastien.boeuf@...> Instead of hardcoding the address of the RSDP in the firmware, let's rely on the PVH structure hvm_start_info to retrieve this

From: Sebastien Boeuf <sebastien.boeuf@...> Following the model from the Xen target, CloudHv is generated as a PVH ELF binary to take advantage of the PVH specification. Signed-off-by:

From: Sebastien Boeuf <sebastien.boeuf@...> Cloud Hypervisor aims at emulating the minimal amount of legacy devices and this is why the PVH boot specification is supported. The point is to be

Several comments as below: 1. Need to ensure sufficient space to save IDT The following will only reserve 8 bytes on stack to save IDT. It is fine for x86, but for x64 mode, the SIDT needs at

https://bugzilla.tianocore.org/show_bug.cgi?id=3D3844 This change added Standalone MM instance of DebugLib. Cc: Jian J Wang <jian.j.wang@...> Cc: Liming Gao

What is this library needed for? take care, Gerd

Acked-by: Gerd Hoffmann <kraxel@...>

.. and again. please quash with patch #23 (also #21 + #22). take care, Gerd

Same here, this patch should be squashed with #19 so git can detect the code motion and maintain history. take care, Gerd

Hi, That kills the git history. Better have adding the cmos function to PlatformInitLib and removing them from PlatformPei in the same patch, git will detect that you are effectively moving code

Is there some #define set when building with support for SEV and/or SNP? If not, should we add one? I think it would make sense to just not include the OvmfSevMetadata table (and other metadata) at

Build with BootLogo using default image by default. Cc: Ray Ni <ray.ni@...> Cc: Maurice Ma <maurice.ma@...> Cc: Benjamin You <benjamin.you@...> Signed-off-by: Sean Rhodes

We use CLANG38. Cheers, Ross

Reviewed-by: Ray Ni <ray.ni@...>

Grub runs later. Then what does “conflict” mean?

+ gUefiPayloadPkgTokenSpaceGuid.PcdBootSplashImage Is this needed?

Thanks Ray! @S, Ashraf Ali I think we need to return 64bit pointer in 64bit build, please help to verify and confirm. Thanks, Chasel

Thanks for good catch, Ray! @Desimone, Nathaniel L and @Ma, Maurice, please also share your comments for this change. I think we could directly change CoreStack size and implementation to support

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2668 WindowOpen will fail in some case. for example, without XServer. Shouldn't set ModeInfo in this case to avoid the caller use