[S] Change in coreboot[master]: cpu/x86/reset16.S: Remove handcoded reset vector

[Arthur Heymans (Code Review)]

Attention is currently required from: Nico Huber, Subrata Banik, Dinesh Gehlot, Angel Pons, Julius Werner, Maximilian Brune, Lean Sheng Tan, Werner Zeh, ron minnich.

Arthur Heymans has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/74797 )

Change subject: cpu/x86/reset16.S: Remove handcoded reset vector
......................................................................

Patch Set 1:

(1 comment)

Patchset:

PS1: 
Well the question is how stale the comment is, as it was written ~20y ago? Truncating the relative address to 3 bytes should also have worked, so I don't really understand the comment even... How often does the tooling break 16bit relocations? Well we use other 16bit relocations for both loading GDT and doing 16bit far jumps in the tree without implementing them in bytes, without any apparent issues for years.

Also you can let the tooling handle the 16bit relocations properly, rather than implement asserts in the linker scripts.
> What is improved?

"_bogus3 = ASSERT(_start16bit >= 0xffff0000, "_start16bit too low. Please report.");" in the linker script can likely be dropped, although I haven't tested that yet. If the jump is close, it can use an 8bit relative jump making code even more compact ^^.

I moved the reset to a different file in the follow-up which makes it easier to read.

I mostly just played around to see if the statement in the comment was still valid since I needed to implement something similar in different codebase.