[PATCH v4 14/35] efi_loader: Allocate and write ACPI tables

[Tom Rini]

On Fri, Sep 27, 2024 at 04:14:55PM +0300, Ilias Apalodimas wrote:

[snip]
> > Well, I suppose so. But if it doesn't
> > matter, why is Sughosh spending all this time solving the problem?
> 
> First of all, we got multiple asks from Tom saying LMB wasn't doing
> all that it should for a long time and needed fixing regardless.
> LMB prior to the patches was calling efi_get_memory_map_alloc() on the
> reserve function and be aware of EFI mappings, but that wasn't the
> best solution. With the lmb created on the fly we also had to call
> that function every time before LMB could decide what memory is free.

On this point, to be clear, I asked Heinrich to tie EFI and LMB together
because I had missed that it wasn't doing anything before and wanted to
avoid a round of CVE assignments over "you can use EFI to overwrite
U-Boot" similar to the CVEs that caused us to start abusing the old LMB
mechanism too. Nobody liked that original change and it was intended to
be reworked.

-- 
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 659 bytes
Desc: not available
URL: <https://lists.denx.de/pipermail/u-boot/attachments/20240927/4fb57e99/attachment.sig>