Extract the coverage sanitizer KPI to a new file.

[andrew]

Summary

This will allow multiple consumers of the coverage data to be compiled
into the kernel together. The only requirement is only one can be
registered at a given point in time, however it is expected they will
only register when the coverage data is needed.

A new kernel conflig option COVERAGE is added. This will allow kcov to
become a module that can be loaded as needed, or compiled into the
kernel.

While here clean up the #include style a little.

[emaste]

sys/amd64/conf/GENERIC

104–106 ↗

(On Diff #53199)

OK for review/test but if the default changes it should be a separate commit

[kib]

sys/conf/kern.pre.mk

116 ↗

(On Diff #53199)

If modules are built as part of the 'make kernel', could you do this evaluation once and pass the result down to modules makes with a var ? You add several hundreds of fork/execs to each build.

sys/kern/kern_kcov.c

337 ↗

(On Diff #53199)

This does not work.

Without D_TRACKCLOSE cdevsw flags, d_close() is only called on very last close. With D_TRACKCLOSE, d_close() calls are not synced with d_open's and can get out of sync. Also, reclaim of the device vnode causes even more unfortunate effects on d_close() mis-pairing with d_open().

546 ↗

(On Diff #53199)

The var is static, there is no need to reinitialize it.

[andrew]

Update the kcov tracking to be based on the number of threads actively
being traced.

[andrew]

Remove a D_TRACKCLOSE that sneaked in