Re: [EXTERNAL] Re: F45 Change Proposal: Grub EFI For Confidential Computing (self-contained)

CS Sushi Man via devel <[email protected]>
Newsgroups gmane.linux.redhat.fedora.devel
Message-ID <kAdHZyIUsfbLDxNS7RYPZ4b2YvxyO5XCm1m8qDmFdkPMyTgq2ekps9n2tqLsYRzRzKVUzWumoilE4o3lXzBqPtvvSAmfJlX2h-aNAoesvbw=@protonmail.com>
On Wednesday, June 24th, 2026 at 15:23, Przemek Klosowski via devel <[email protected]> wrote:

> On 6/24/26 12:21 PM, CS Sushi Man via devel wrote:
> > write a simple boot loader, that loads the kernel
> > into RAM directly from the hard drive, *without using any
> > filesystem.* It would do this by baking an offset into the binary,
> > which the program can then use to load the kernel image from a
> > location on the hard drive.
> 
> Yeah but the kernel is likely to be written across non-contiguous blocks
> on the filesystem, and then you need to keep the ordered list of all the
> consecutive blocks. You also need the same for the initrd. And it would
> have to be done on every new kernel install.
> 
> Your proposal amounts to rewriting GRUB.
> 
> --

  Is it not possible to dedicate the first 100K or so odd blocks to
store a contiguous packed image (sort of like a SquashFS, but not
an FS)? And if need be, encrypt & sign it? I think a tool can be
written that will accomplish this task on the userspace side. This
will fix the problem of a non-contiguous kernel/initrd on the hard
drive, since everything will be stored without a filesystem to
scramble everything in the first place.

  Note that I do not know how the kernel loads an initrd image, or
whether that requires a FAT/EXT file system or not (although I'm
pretty sure the kernel loads the initrd by using the primary EXT
file system on the hard drive).

  I believe that most of the boot process can, and should be
handled by the kernel, since it contains so many drivers in one
package. I don't actually know how GRUB passes kernel parameters,
but if there isn't a way as of now to pass parameters via loading a
set of blocks off the hard drive, I'm sure a change in the kernel
can be made to address this. Again, I don't have much experience
with kernel development as of yet, so I don't know how easy or hard
making this change would be.

  If this is done right, rewriting GRUB will not be necessary, and
can be entirely avoided.

        - C. S. Sushi


Sent with Proton Mail secure email.
-- 
_______________________________________________
devel mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/[email protected]
Do not reply to spam, report it: https://forge.fedoraproject.org/infra/tickets/issues/new
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.