WARNING: Expected rawoffset 0, found 63

[Mike Meyer]

Hello,

I'm changing the way I'm mirror my system disk on an amd64 system
running 6-STABLE (from late april). In paticular, I want to mirror the
root partition instead of the entire disk, and have this setup for
that:

bhuda% gmirror list gm0s1a
Geom name: gm0s1a
State: COMPLETE
Components: 2
Balance: split
Slice: 4096
Flags: NONE
GenID: 0
SyncID: 1
ID: 2799177281
Providers:
1. Name: mirror/gm0s1a
   Mediasize: 1073741312 (1.0G)
   Sectorsize: 512
   Mode: r1w1e1
Consumers:
1. Name: ad2s1a
   Mediasize: 1073741824 (1.0G)
   Sectorsize: 512
   Mode: r1w1e1
   State: ACTIVE
   Priority: 0
   Flags: NONE
   GenID: 0
   SyncID: 1
   ID: 2288219402
2. Name: ad3s1a
   Mediasize: 1073741824 (1.0G)
   Sectorsize: 512
   Mode: r1w1e1
   State: ACTIVE
   Priority: 0
   Flags: NONE
   GenID: 0
   SyncID: 1
   ID: 2796645801

Neither ad[23] nor ad[23]s1 is mirrored. Things seem to be working
fine.  However, on boot I get a pair of the warnings listed in the
subject line.

Setting kern.geom.debugflags to 1 in loader.conf and rebooting, I find
the following interesting bit of output:

mbr_taste(MBR,mirror/gm0s1a)
g_slice_spoiled(0xffffff0000b56280/mirror/gm0s1a)
g_wither_geom(0xffffff0000b75900(mirror/gm0s1a))
bsd_taste(BSD,mirror/gm0s1a)
WARNING: Expected rawoffset 0, found 63
g_slice_config(mirror/gm0s1a, 0, 0)
...
g_slice_config(mirror/gm0s1a, 7, 1)
g_slice_conf_hot(mirror/gm0s1a, idx: 0, off: 512, len: 276)
g_gpt_taste(GPT,mirror/gm0s1a)
g_slice_spoiled(0xffffff0000b56580/mirror/gm0s1a)
g_wither_geom(0xffffff0000b95800(mirror/gm0s1a))

Um, ok, there is a BSD label on mirror/gm0s1a, because there's a BSD
label on ad[23]s1, which has the same starting sector as ad[23]s1a,
which are where gm0s1a is getting blocks. But that label has already
been attached (twice, actually - once from ad2s1 and once from
ad3s1). In fact, when we taste those, it fails as expected:

bsd_taste(BSD,ad2s1a)
g_slice_spoiled(0xffffff0000b98280/ad2s1a)
g_wither_geom(0xffffff0000bdb200(ad2s1a))

It looks like the 'partitions' inside gm0s1a are destroyed before we
finish booting, as the devices certainly aren't there after we
boot. Which leads to my two questions:

1) Can this warning be ignored? If it's from a geom that indeed
   vanishes later, that should be the case.

2) Is this a bug? Or maybe an undesirable behavior that should be
   documented?

	Thanks,
	<mike
-- 
Mike Meyer <mwm at mired.org>		http://www.mired.org/consulting.html
Independent Network/Unix/Perforce consultant, email for more information.