Fix FreeBSD condvar semantics

[ghost]

Motivation and Context

FreeBSD's SPL cv_timedwait_* implementations returned deltas in all cases. The correct semantics are:

Return Value	Description
0	For cv_wait_sig(), cv_timedwait_sig(), and cv_reltimedwait_sig() indicates that the condition was not necessarily signaled and the function returned because a signal (as in kill(2)) was pending.
-1	For cv_timedwait(), cv_timedwait_sig(), cv_reltimedwait(), and cv_reltimedwait_sig(), indicates that the condition was not necessarily signaled and the function returned because the timeout time was reached.
>0	For cv_wait_sig(), cv_timedwait(), cv_timedwait_sig(), cv_reltimedwait(), and cv_reltimedwait_sig(), indicates that the condition was met and the function returned due to a call to cv_signal() or cv_broadcast(), or due to a premature wakeup (see NOTES).

(Source: https://illumos.org/man/9F/cv_timedwait)

Description

Return -1 instead of negative deltas, and 0 if signaled.

How Has This Been Tested?

ZTS on FreeBSD.

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Performance enhancement (non-breaking change which improves efficiency)
• Code cleanup (non-breaking change which makes code smaller or more readable)
• Breaking change (fix or feature that would cause existing functionality to change)
• Documentation (a change to man pages or other documentation)

Checklist:

• My code follows the ZFS on Linux code style requirements.
• I have updated the documentation accordingly.
• I have read the contributing document.
• I have added tests to cover my changes.
• I have run the ZFS Test Suite with this change applied.
• All commit messages are properly formatted and contain Signed-off-by.

[amotin]

Otherwise it makes sense to me.

[lundman]

and with my apologies.

[ghost]

• Return -1 if delta is negative, as intended

[amotin]

Thanks.

[adamdmoss]

(I think it'd help if openzfs documented its internal APIs - even in a one-line comment - instead of relying on the implicit knowledge that they should match syscall semantics from the kernel of an operating system that's relatively obscure even by *nix standards...!)

[codecov]

Codecov Report

Merging #10460 into master will decrease coverage by 0.81%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #10460      +/-   ##
==========================================
- Coverage   80.24%   79.43%   -0.82%     
==========================================
  Files         293      393     +100     
  Lines       83882   123859   +39977     
==========================================
+ Hits        67313    98385   +31072     
- Misses      16569    25474    +8905     

Flag	Coverage Δ
#kernel	80.05% <ø> (-0.20%)	⬇️
#user	65.32% <ø> (?)

Impacted Files	Coverage Δ
module/zfs/zcp_set.c	87.09% <0.00%> (-12.91%)	⬇️
module/zfs/zfs_ratelimit.c	90.47% <0.00%> (-9.53%)	⬇️
module/zfs/dmu_redact.c	82.15% <0.00%> (-6.24%)	⬇️
module/zfs/zfs_byteswap.c	79.45% <0.00%> (-5.85%)	⬇️
module/zcommon/zfs_deleg.c	92.45% <0.00%> (-5.55%)	⬇️
module/zcommon/zfeature_common.c	92.85% <0.00%> (-4.82%)	⬇️
module/zfs/aggsum.c	95.34% <0.00%> (-4.66%)	⬇️
module/zfs/dsl_deleg.c	90.07% <0.00%> (-4.48%)	⬇️
module/zfs/zio_inject.c	75.80% <0.00%> (-4.41%)	⬇️
module/zfs/spa_checkpoint.c	93.78% <0.00%> (-4.26%)	⬇️
... and 278 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 883a40f...65981d0. Read the comment docs.