[PATCH 0/2] efi: fix and test hii keyboard layout pointer computation

[Michael Brown]

On 18/05/2026 10:48, Vincent Stehlé wrote:
> On Tue, May 12, 2026 at 09:05:29PM +0200, Heinrich Schuchardt wrote:
>> Did you find any real use for HII protocols in U-Boot except for running the SCT?
>>
>> Are HII based EFI applications a thing?
> 
> Hi Heinrich,
> 
> A fair question indeed, as the HII protocols do not seem to be used that much in
> practice.
> 
> The UEFI Shell seems to be using them, though, which should be a sufficient
> reason to support them, I think.
> 
> Those protocols are used as well a little in systemd-boot and iPXE for example,
> but as far as I can tell, they are not used at all in Shim, GRUB, Linux, Xen,
> the BSDs or Android GBL.

For the record: iPXE has zero interest in almost all of the HII 
protocols.  We provide an HII interface to allow external utilities to 
read and modify iPXE settings, but I don't believe anything actually 
uses this interface.  (There are some comments about the HII config 
protocols at 
https://github.com/ipxe/ipxe/blob/ca85200809eb0bf5571869264cb2bf7f0548180a/src/interface/efi/efi_snp_hii.c#L30-L49)

The only use case that iPXE itself has for any part of HII is the 
EFI_HII_FONT_PROTOCOL, which we use to obtain font glyphs in the cases 
when we are drawing our own framebuffer console instead of using ConOut.

Thanks,

Michael