fix(dracut.sh): use dynamically uefi's sections offset

[keentux]

• Uefi section are creating by objcopy with hardcoded sections offset. This patch allow to have a dynamic set of offsets between each part of the efi file, needed to create an UKI. Offsets are simply calculated so no sections overlap, see https://wiki.archlinux.org/title/Unified_kernel_image#Manually
  Moreover, efi stub file's header is parsed to apply the correct offsets according to the section alignment factor. More information here: https://learn.microsoft.com/en-us/windows/win32/debug/pe-format#optional-header-windows-specific-fields-image-only
• Remove EFI_SECTION_VMA_INITRD, no need anymore as initrd section offset dynamically calculated.

Checklist

• I have tested it locally
• I have reviewed and updated any documentation if relevant
• I am providing new code and test(s) for it

Fixes #2275

[aafeijoo-suse]

Thanks for looking into this!

[medhefgo]

I forgot to mention that the offsets should really be aligned to the PE binary's SectionAlignment header field.

[keentux]

I forgot to mention that the offsets should really be aligned to the PE binary's SectionAlignment header field.

Thanks for that clarification. I am going to look on at.

[keentux]

On my last changes, during the UEFI generation, it will parse the initial PE file's header (the stub) checking if it is a good PE format and catch the value of sectionAlignment . According to the PE documentation, each section should start on a factor of the value of sectionAlignment.
@medhefgo What do you think about that. Should it be good up to you ?

[johannbg]

LGTM

[LaszloGombos]

Needs confirmation, but I think strtonum does not seem to be available in all awk variants. https://www.gnu.org/software/gawk/manual/html_node/Strtonum-Function.html

Specifically on debian mawk is more popular than gawk, see #2002 (comment)

This PR seems to break UEFI generation if gawk is not installed.

Had we had #2166 landed, we would have seen the failure as part of the PR review.