For FreeBSD support, we are currently only testing blockio in CI (not lsm-tree) because that is where the FreeBSD-specific code lives. However, the blockio tests are not very extensive and so we are not exercising all parts of the FreeBSD backend. We should write more tests so that we cover the whole of the HasBlockIO API at least. Other (future) backends will also benefit from this.
Note: the core Linux, MacOS, and Windows blockio backends are already extensively tested indirectly through testing lsm-tree, so they benefit a little less from more extensive tests, but it's still very useful.
For
FreeBSDsupport, we are currently only testingblockioin CI (notlsm-tree) because that is where theFreeBSD-specific code lives. However, theblockiotests are not very extensive and so we are not exercising all parts of theFreeBSDbackend. We should write more tests so that we cover the whole of theHasBlockIOAPI at least. Other (future) backends will also benefit from this.Note: the core Linux, MacOS, and Windows
blockiobackends are already extensively tested indirectly through testinglsm-tree, so they benefit a little less from more extensive tests, but it's still very useful.