On Thu, Apr 23, 2020 at 10:30:44AM +0300, Max Gurtovoy wrote: > > On 4/23/2020 8:54 AM, Christoph Hellwig wrote: >> On Thu, Apr 23, 2020 at 01:39:26AM +0300, Max Gurtovoy wrote: >>> it's a bit late for me now so I probably wrote non standard sentence above. >>> >>> BUT what I meant to say is I would like to give the user an option to >>> decide whether use E2E protection or not (of course a controller can >>> control protected and non-protected namespaces :) ) >> I don't really have a problem with an opt-out, but I'd like to apply it >> consistently over all transports. >> >>> AFAIK, there is no option to format a ns in NVMf (at least for RDMA there >>> is only 1 lbaf exposed by the target) so i'm not sure how exactly this will >>> work. >> The NVMe protocol Format NVM support is independent of the transport. > > Ok, but it's not supported in Linux. > > Are you saying we should implement Format NVM for fabrics ? or stay > consistent for NVMf (and not nvmf + pci) ? I see no reason not to support a simple Format NVM for our fabrics target implementation. But that isn't the point - you don't really need Format as you can also control it from configfs in your series. So for the initial version I don't think we need Format NVM, but I don't mind adding it later.
