zpool create: report which device caused failure

[chrislongros]

Motivation

When zpool create fails because a vdev is already in use, the error message does not identify which device caused the problem:

cannot create 'tank': one or more vdevs refer to the same device

This is frustrating when creating pools with many disks, as there is no way to tell which device is the culprit without manually checking each one.

Description

This patch adds device-specific error reporting so the message now identifies the problematic device and the pool it belongs to:

cannot create 'tank': device '/dev/sdb1' is part of active pool 'rpool'

Implementation follows the ZPOOL_CONFIG_LOAD_INFO pattern used by zpool import, as suggested in #18184:

• Add spa_create_info to spa_t to capture error info during vdev_label_init(), before vdev_close() resets vdev state
• When vdev_inuse() detects a conflict, read the on-disk label to extract the pool name and store it with the device path
• Return the info wrapped under ZPOOL_CONFIG_CREATE_INFO through the ioctl zc_nvlist_dst to userspace
• In libzfs, the zpool_create_info() helper unwraps the nvlist and formats the device-specific error message

If the pool name cannot be read from the label, the device path alone is reported as a fallback.

Testing

• New ZTS test zpool_create_errinfo_001_neg verifies the error message names the device and the active pool when creating a mirror with duplicate loopback vdevs (portable across Linux and FreeBSD)
• Normal pool creation still succeeds
• checkstyle passes

Suggested-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Christos Longros <chris.longros@gmail.com>

[behlendorf]

Thanks for working on this!

[behlendorf]

Now that we're going to have infrastructure to report more detailed errors in zpool create I suspect we'll want to make use of it in some of the other error cases. I've thrown together a completely untested commit which restructures things so we can access the nvlists where we need them. Can you take a look at 2c5fa0d, refine as needed, and then squash something that's actually tested in to this PR!