tests: don't run atf_* in a subshell

Shell limitation is that a classic function call via $() is a subshell
and atf-sh(3) commands won't work as epxected there. Subsequently,
atf_skip inside a function won't skip a test. The test will fail later.

A working approach is to pass desired variable name as argument to
a function and don't run subshell.

Reviewed by: ngie
Differential Revision: https://reviews.freebsd.org/D42646
Fixes: ea82362219ee715cfbb195b2114e73fdc8599fa5

geli tests: Use shorter passphrases in setkey_passphrase

There is an undocumented limit on the length of a GELI passphrase, and
the test exceeded that. Most of the time it worked because a nul
terminator would appear early enough in the string.

Reported by: Jenkins
Fixes: 2b7b09ac9675 ("geli tests: Add a regression test for geli setkey -J")

geli tests: Add a regression test for geli setkey -J

Make sure that it can be used to change the passphrase of both attached
and detached providers.

PR: 254966
MFC after: 1 week
Sponsored by: The FreeBSD Foundation

Remove $FreeBSD$: one-line sh pattern

Remove /^\s*#[#!]?\s*\$FreeBSD\$.*$\n/

Unbreak geli/gmirror testcases if their geom classes cannot be loaded

The problem with the logic prior to this commit was twofold:

1. The wrong set of idioms (TAP-compatible) were being applied to the ATF
testcases when run, resulting in confusing ATF failure results on setup.
2. The cleanup subroutines were broken when the geom classes could not be
loaded as they exited with 0 unexpectedly.

This commit changes the test code to source the class-specific configuration
(conf.sh) once globally, instead of sourcing it per testcase and per cleanup
subroutine, and to call the ATF-specific setup subroutine(s) inline in
the testcases.

The refactoring done is effectively a no-op for the TAP testcases, modulo
any refactoring done to create common code between the ATF and TAP
testcases.

This unbreaks the geli testcases converted to ATF in r327662 and r327683,
and the gmirror testcases added in r327780, respectively, when the geom
class could not be loaded.

tests/sys/geom/class/mirror/...
While here, ignore errors when turning debug failpoint sysctl off, which
could occur if the gmirror class was not loaded.

Submitted by: ngie
MFC after: 2 weeks
Pull Request: https://github.com/freebsd/freebsd/pull/241

geli: convert remaining TAP tests to ATF

MFC after: 2 weeks

geli: convert most tests from TAP to ATF

I'm leaving readonly_test and nokey_test alone for now. In a future commit
they should be broken up into several smaller test cases and distributed
between multiple files.

Reviewed by: ngie
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D13717

Fix potential TOCTTOU bug in the geli tests

This change mostly reverts r293436, which introduced the bug due to a belief
that geli(8) would allocate md(4) devices by itself. However, that belief is
incorrect. Instead of using linear probing to find available md(4) numbers,
it's best to use the existing attach_md function.

Reviewed by: ngie
MFC after: 2 weeks
Sponsored by: Spectra Logic Corp
Differential Revision: https://reviews.freebsd.org/D13666

Integrate
tools/regression/geom_{concat,eli,gate,mirror,nop,raid3,shsec,stripe,uzip}
in to the FreeBSD test suite as
tests/sys/geom/class/{concat,eli,gate,mirror,nop,raid3,shsec,stripe,uzip}

The tools/regression/geom and tools/regression/geom_part testcases are being
left alone because both test sets are both currently broken.

The majority of this work was done on ^/user/ngie/more-tests2 . The differences
are as follows:
- tests/sys/geom/class/Makefile.inc is not present; it was
inlined into the class's Makefiles for explicitness.
- The testcases officially require root via kyua
- The geom_gate(4) tests don't use the pidfile changes proposed in
https://reviews.freebsd.org/D4836 .

MFC after: 1 month
Sponsored by: EMC / Isilon Storage Division