On Thu, Oct 17, 2024 at 04:16:13PM +0530, Anuj Gupta wrote: > On Thu, Oct 17, 2024 at 10:12:23AM +0200, Christoph Hellwig wrote: > > On Wed, Oct 16, 2024 at 04:59:09PM +0530, Anuj Gupta wrote: > > > This patch introduces BIP_CHECK_GUARD/REFTAG/APPTAG bip_flags which > > > indicate how the hardware should check the integrity payload. The > > > driver can now just rely on block layer flags, and doesn't need to > > > know the integrity source. Submitter of PI decides which tags to check. > > > This would also give us a unified interface for user and kernel > > > generated integrity. > > > > The conversion of the existing logic looks good, but the BIP_CHECK_APPTAG > > flag is completely unreferenced. > > It's being used by the nvme and scsi patch later. Should I introduce this > flag later in either nvme or scsi patch where we actually use it. Maybe a separate one? Or at least very clearly state that the other two are conversion of the existing semantics, while this one is new.
