Re: EFI boot failing on new machine

[Paul Ripke <stix%stix.id.au@localhost>]

On Sun, Feb 23, 2025 at 07:34:02AM +0000, RVP wrote:
> On Sun, 23 Feb 2025, Paul Ripke wrote:
> 
> > Doing some Print debugging, I'm finding the first AllocatePages in
> > efi_md_init hangs. Which is a little interesting, since an AllocatePages
> > in efi_heap_init has previously succeeded? My current debug log:
> > 
> > efi_main starting...
> > efi_heap_init...
> > AllocatePages addr: 0000000100000000 sz: 256
> > heap_start: 00000000&E66F000
> > efi_md_init...
> > startprog64_size: 8480
> > AllocatePages addr: 0000000100000000 sz: 3
> > 
> 
> So, this is trying to allocate (any) memory below 4GB and succeeds the first
> time, but, not the second...

Agreed - but hangs rather than failing?

> - What does the memory map look like?

I found an EFI shell which was able to dump it out - it's 100's of lines :/

> - Is there sufficient memory left below 4GB, or is all the memory mapped above
>   4GB on this board?

Great question! Without capturing the memory map in full, it's hard to tell,
except...

> - What happens if you change AllocateMaxAddress to AllocateAnyPages in
>   efi_md_init()?

I tried both switching AllocateMaxAddress to AllocateAnyPages, and also
tried shifting EFI_ALLOCATE_MAX_ADDRESS by 4 bits (x16); same result in
both cases, the same hang at the same place. 

-- 
Paul Ripke
"Great minds discuss ideas, average minds discuss events, small minds
 discuss people."
-- Disputed: Often attributed to Eleanor Roosevelt. 1948.