On Thu, Apr 08, 2021 at 10:57:32AM -0400, Josef Bacik wrote: > On 4/8/21 4:25 AM, Naohiro Aota wrote: > > confusion. > > > > Signed-off-by: Naohiro Aota <naohiro.a...@wdc.com> > > Thanks Naohiro, this makes it much easier to understand, > > Reviewed-by: Josef Bacik <jo...@toxicpanda.com> > > Dave, I know you're on vacation, this needs to go to Linus for this cycle so > we > don't have two different SB slots for two different kernels. I don't want to > disturb your vacation, so I'll put this in a pull request tomorrow to Linus. > If > you already had plans to do this let me know and I'll hold off. Thanks,
We don't even have mkfs support for zoned filesystem merged so that the format is not final was not an issue and that's why I did not hurry pushing it in that cycle and I still rather not do that. The design-level problem is the backup copy offset, that's been discussed in some previous iteration - whether it's right to place all copies under the first 1T or if we should spread them. For me this is still not settled, I haven't read through all the replies and given it enough thought. That could lead to another change of the offsets that I'm sure we want to avoid.
