mtree: Sync with upstream

[jlduran]

All our local changes were upstreamed and incorporated. It provides a commit-by-commit history, with some commits done only for the metadata (RCSID) changes.

[jlduran]

# kyua test usr.sbin/nmtree
usr.sbin/nmtree/nmtree_test:mtree_check  ->  passed  [0.029s]
usr.sbin/nmtree/nmtree_test:mtree_convert_C  ->  passed  [0.017s]
usr.sbin/nmtree/nmtree_test:mtree_convert_C_S  ->  passed  [0.016s]
usr.sbin/nmtree/nmtree_test:mtree_convert_D  ->  passed  [0.016s]
usr.sbin/nmtree/nmtree_test:mtree_convert_D_S  ->  passed  [0.021s]
usr.sbin/nmtree/nmtree_test:mtree_create  ->  passed  [0.027s]
usr.sbin/nmtree/nmtree_test:mtree_ignore  ->  passed  [0.031s]
usr.sbin/nmtree/nmtree_test:mtree_merge  ->  passed  [0.019s]
usr.sbin/nmtree/nmtree_test:mtree_nonemptydir  ->  passed  [0.031s]
usr.sbin/nmtree/nmtree_test:mtree_onlyfile  ->  passed  [0.018s]
usr.sbin/nmtree/nmtree_test:mtree_specspec_type  ->  passed  [0.033s]
usr.sbin/nmtree/nmtree_test:netbsd6_check  ->  passed  [0.024s]
usr.sbin/nmtree/nmtree_test:netbsd6_convert_C  ->  passed  [0.020s]
usr.sbin/nmtree/nmtree_test:netbsd6_convert_C_S  ->  passed  [0.019s]
usr.sbin/nmtree/nmtree_test:netbsd6_convert_D  ->  passed  [0.016s]
usr.sbin/nmtree/nmtree_test:netbsd6_convert_D_S  ->  passed  [0.016s]
usr.sbin/nmtree/nmtree_test:netbsd6_create  ->  passed  [0.030s]
usr.sbin/nmtree/nmtree_test:netbsd6_ignore  ->  passed  [0.032s]
usr.sbin/nmtree/nmtree_test:netbsd6_merge  ->  passed  [0.019s]
usr.sbin/nmtree/nmtree_test:netbsd6_nonemptydir  ->  passed  [0.028s]
usr.sbin/nmtree/nmtree_test:netbsd6_onlyfile  ->  passed  [0.018s]

...

21/21 passed (0 failed)

[jlduran]

Local changes were submitted to GNATS bin/57735.

[jlduran]

Upstream has accepted our patches. There are a few caveats:

• [x] Upstream has to fix CVS -> GitHub
• [x] Upstream has yet to remove a duplicate test

• [x] Commit https://github.com/freebsd/freebsd-src/commit/2dfa4b66b3d0caaaae6ce2df476b5615f8415a19:

  This is addressing cases such as fts_read(3) encountering an [EIO] from fchdir(2) when FTS_NOCHDIR is not set. That would otherwise be seen as a successful traversal in some of these cases while silently discarding expected work.

  As noted in [https://github.com/freebsd/freebsd-src/commit/73a0af46b854be94a05fcc4d4603619a00ac5707] fts_read() does not set errno to 0 on a successful EOF so it needs to be set before calling it. Otherwise we might see a random error from one of the iterations.

  needs further investigation, upstream did not set errno = 0, as it is not needed (for them), I think here is also not needed, but to err on the side of caution, let's keep it. In any case, the whole commit should be reverted (at least partially) and tested in fts (wontfix).

[jlduran]

Note to self: It will likely land as two separate commits:

Vendor import of vendor/NetBSD/mtree. Vendor import of vendor/NetBSD/tests. At any rate, FreeBSD's local changes have already been upstreamed.