Fixes from lstewart@.

(Tinderboxed.)

Sure, it's openat02:5. I'll add it to the commit.

Fix most of the problems.

Document linux_mounts_enable.

Rename to 'linux_mounts_enable' and flip the meaning.

FWIW, grepping for that pattern only brings up the one hit Warner has already mentioned, and it's already under #ifdef illumos:

% grep -r '%ss%d' *
cddl/contrib/opensolaris/cmd/zpool/zpool_vdev.c:                (void) snprintf(buf, sizeof (buf), "%ss%d", path, slice);

Fixed, thanks!

Feedback from brueffer@.

Apply feedback from emaste@ and tijl@.

Feedback from brueffer@ and allanjude@.

Rename to "attachment", merge into a single GEOM attribute.

That's true; I've added a BUGS section.

Add BUGS.

So... Is procfs actually obsolete then?

"Attachment" sounds perfect, thanks!

(Tinderboxed.)

In D22249#486617, @chuck wrote:

Overall, the changes look good to me, but I am having a twinge over "HBA". HBA completely makes sense for SCSI but doesn't at all for NVMe. SATA falls in the middle depending on whether this is a proprietary thing or AHCI.

In D22194#485930, @imp wrote:

teemo. Coding the ansi escape sequences is really a bad idea.

In D22193#485541, @allanjude wrote:

In D22193#485070, @imp wrote:

/*
  * Fill in the devid, now that we've labeled the disk.
  */
 (void) snprintf(buf, sizeof (buf), "%ss%d", path, slice);
 if ((fd = open(buf, O_RDONLY)) < 0) {
         (void) fprintf(stderr,
             gettext("cannot open '%s': %s\n"),
             buf, strerror(errno));
         return (-1);
 }

is one example, this one is from cddl/contrib/opensolaris/cmd/zpool/zpool_vdev.c

We should try to find these cases and clean them up. I wouldn't expect there to be too many, since everything should be using GEOM to create partitions etc.

Fix tinderbox on i386.

(Tinderboxed.)

In D22193#485070, @imp wrote:

/*
  * Fill in the devid, now that we've labeled the disk.
  */
 (void) snprintf(buf, sizeof (buf), "%ss%d", path, slice);
 if ((fd = open(buf, O_RDONLY)) < 0) {
         (void) fprintf(stderr,
             gettext("cannot open '%s': %s\n"),
             buf, strerror(errno));
         return (-1);
 }

is one example, this one is from cddl/contrib/opensolaris/cmd/zpool/zpool_vdev.c

Update the man page.

In D20333#484720, @kib wrote:

In D20333#484711, @trasz wrote:

Well, they don't print an error. I'm not quite sure why, but I've tested both scenarios - kldloading linux.ko and then linux64.ko, and the other way around; same with kldunloading.

Then there must be a bug either in the testing methodology or in the code, right ? Ensure that the make_dev line is executed in both cases.

In D22193#485062, @imp wrote:

There are lots of places in the system that will break when this is not '' . What are your plans on fixing those?

In D22193#485045, @allanjude wrote:

Is the 'd' or 'c' in the description a typo? Should it be ada0d/s1? or ada0/s1

Fix layout for one disk setups.

Well, they don't print an error. I'm not quite sure why, but I've tested both scenarios - kldloading linux.ko and then linux64.ko, and the other way around; same with kldunloading.

Apply comments from kib@ and emaste@.

Attempt to fix problems with -N reported by thj@; also build fix
for rescue(8).

Also see the discussion under https://reviews.freebsd.org/D20322.

Superseded by https://reviews.freebsd.org/D21874.

(Tinderboxed.)

Drop local change to GENERIC.

Add WITH_STATS conditional.

Make -M also work with -l; reported by thj@.