Re: GPT partlabel support for Mach
Hello,
наб, le sam. 30 mai 2020 14:10:37 +0200, a ecrit:
> so I implemented it based on the UEFI 2.8A spec;
Nice!
I was about to say that, well, the long-term plan is to just use
libparted in userland, so we'll not need partition support in the kernel
any more. But we are not there yet, so GPT support would probably be
useful, and your patch seems small enough that we can as well include
it since it's there.
Would you assign copyright to the FSF so we can commit the support in
upstream gnumach?
> Speaking of correct: as noted in a comment, AFAICT no driver
> sets the sector size except for SCSI, and SCSI only ever sets it to 512;
> this will be a problem on 4k sector drives with 4k-based pointers.
Again, the long-term plan is to move to rump-based disk drivers so that
will go away.
> + void *cur = *pp = (void *)kalloc(bytes_left = *pp_s = h->h_part_table_len * h->h_part_table_entry_size);
Please avoid assignation inside function parameters, that's difficult to
read.
Samuel
Reply to: