On Thu, 09/29 09:58, Paolo Bonzini wrote: > > > On 29/09/2016 04:21, Fam Zheng wrote: > > On Wed, 09/28 18:11, Max Reitz wrote: > >> Note that BDRV_REQ_MAY_UNMAP does not mean "Yes, please discard" but > >> just "You may discard if it's easier for you". But it's actually not > >> easier for us, so I don't see why we're doing it. > >> > >> As far as I can guess you actually want some way to tell a block driver > >> to actually make an effort to discard clusters as long they then read > >> back as zero (which is why you cannot simply use bdrv_pdiscard()). > >> However, I think this would require a new flag called > >> BDRV_REQ_SHOULD_UNMAP (which should imply BDRV_REQ_MAY_UNMAP). > > > > This flag doesn't make sense to me, if the protocol doesn't know how to > > unmap, > > it can ignore BDRV_REQ_MAY_UNMAP, but not BDRV_REQ_SHOULD_UNMAP. It just > > complicates things a little. > > I don't think we actually have a use for a "MAY" unmap flag. Either we > keep the not-so-perfect name or we replace MAY_UNMAP with "should" or > "want" or "would_like" unmap... But Fam's patch does do what was > intended for the flag (which is the equivalent of the UNMAP bit in the > SCSI WRITE SAME command).
After reading rfc2119, now I agree that "SHOULD" is better. :) Fam
