[syslinux] [PATCH 1/2] Match comment with code

[Ady]

> On 12/16/2013 01:41 PM, Ady wrote:
> > 
> > 4 January 2010 (newer)
> > e09127r3
> > "EDD-4 Hybrid MBR boot code annex"
> > [quote]
> > Set to the OS Type of the bootable partition (i.e., 
> > that would have been assigned had the partition 
> > been installed in an MBR disk layout).
> > [/quote]
> > 
> > So, in the older revision, "ED" is explicitly mentioned. In a newer 
> > revision there is no explicit mention of an ID code.
> > 
> > I still see popular sites with lists of partition IDs with "EE" and 
> > "EF" in relation to EFI, and AFAIK there is no (other) "standard" for 
> > GPT in BIOS.
> > 
> > I wonder whether other documents / standards explicitly mention *any* 
> > partition ID, and whether these IDs could potentially result in 
> > future inconsistencies / problems / bugs / hours of tracking down 
> > strange (mis)behaviors.
> > 
> 
> You're right - T13 changed the "OS Type" from 0xED to "that would have
> been assigned had the partition been installed in an MBR disk layout".
> That is kind of hard to accomplish without blowing the MBR code size
> budget, since it effectively requires a mapping from GUIDs to OS types.
>  In practice I doubt it matters, though.
> 
> 	-hpa
 
I assume that the intention to use a (theoretical and desired) match 
between the partition ID and the filesystem is to "imitate" the 
common traditional MBR usage, for a (theoretical) closer backward 
compatibility.

Now, if the newer T13 documentation doesn't specifically refer to 
"ED" as partition ID for the "hybrid" MBR boot code handover in GPT, 
then some questions remain:

_ Is the original desire *not* to use "EE" still relevant?

_ If no particular "matched" ID is used in gptmbr.bin, should "ED" 
still be used? I wonder whether it would be better to change it 
(back) to "EE" (?).

FWIW, UEFI 2.4 says:
[quote]
0xEE (i.e., GPT Protective) is used by a protective MBR (see 5.2.2) 
to define a fake partition 
covering the entire disk.
[/quote]

Regards,
Ady.