DELETE support in the VOP_STRATEGY(9)?

Dag-Erling Smørgrav des at des.no
Tue Dec 8 20:13:12 UTC 2015


Warner Losh <imp at bsdimp.com> writes:
> And to be fair, having an additional property of ‘seeks are nearly
> free’ would also be a good way to tell. I’m not convinced it is worth
> the effort to add it to all the storage devices in the tree when
> GEOM::candelete is a good proxy.

I just provided you with an (admittedly fictional, but not unreasonable)
example of a layer which implements BIO_DELETE on top of storage that
may or may not have free seeks.  The two are completely orthogonal; they
just happen to be strongly correlated on currently available hardware.

Note that my fictional example would guarantee that BIO_DELETEd space
reads back as zeroes, even if the request doesn't align with physical
block boundaries.  Sounds pretty useful to me, even if it doesn't
guarantee that the deleted data cannot be recovered from the physical
media by a sufficiently determined attacker with access to liquid
nitrogen and an electron microscope.

DES
-- 
Dag-Erling Smørgrav - des at des.no


More information about the freebsd-hackers mailing list