geom: fix glabel status after 0d4642a

[brd]

before:

          Name  Status  Components
  gpt/efiboot0     N/A  N/A
msdosfs/EFISYS     N/A  N/A
  gpt/gptboot0     N/A  N/A
  gpt/efiboot1     N/A  N/A
  gpt/gptboot1     N/A  N/A

after:

          Name  Status  Components
  gpt/efiboot0     N/A  nda0p1
msdosfs/EFISYS     N/A  nda0p1
  gpt/gptboot0     N/A  nda0p2
  gpt/efiboot1     N/A  nda1p1
  gpt/gptboot1     N/A  nda1p2

[brd]

Fix check if the provider is NULL

[mckusick]

This looks good to me, but I will await a review from Johan before proceeding.

[guest-jsollvander]

We should add the same check to status_one_geom() as well

[brd]

Update status_one_geom() too

[brd]

In D53180#1215259, @guest-jsollvander wrote:

We should add the same check to status_one_geom() as well

OK, added.. Do you know which command would call through this function to verify this?

[brd]

In D53180#1215263, @brd wrote:

In D53180#1215259, @guest-jsollvander wrote:

We should add the same check to status_one_geom() as well

OK, added.. Do you know which command would call through this function to verify this?

Aha, geom label status -g.. OK, yes this fixes that as well.

[markj]

sbin/geom/core/geom.c

1219

There is also a typo here and above, compontent->component.

[kevans]

Please be sure to include 0d4642a as Fixes metadata so that it's clear they need to go together both for MFCs and downstreams that might be consuming main.

sbin/geom/core/geom.c

1196–1197

Might as well move this up prior to checking the status? Obviously they'll both return NULL anyways, but it seems more clear then that they won't be mutating state.

1265

Ditto here

[brd]

Address feedback from kevans

[brd]

sbin/geom/core/geom.c

1196–1197

I was wondering about that myself.. Thanks!

1219

Should I do all of them in one commit? There are 10..

[brd]

Fix typos as well since they are part of the original commit.

[adrian]

i hit this on the latest rpi4 snapshot build, which has this bug and it prevents rc.d/growfs from working.

Thanks, approved!