On Tue, Sep 20, 2022 at 02:11:40PM -0400, Stefan Hajnoczi wrote: > On Mon, Sep 19, 2022 at 09:09:05PM +0300, Alvaro Karsz wrote: > > Thanks for the reply. > > > > > This can be simplified with min_not_zero(). > > > > Ok, I will do it in the next version. > > > > > It's worth including a comment here that the discard and secure erase > > > limits are combined because the Linux block layer only has one limit > > > value. If the block layer supported independent limit values we wouldn't > > > need to do this. > > > > Ok. > > > > I'll send a new version once we agree on the max_secure_erase_seg = 0 scenario. > > Do you have an opinion on that? > > Do you think that using sg_elems as the number of secure erase/discard > > segments when the value in the virtio config is 0 is good enough? > > > > Okay, I have replied in the max_secure_erase_seg sub-thread. I think > probing the device should fail if the value is 0. There are no existing > non-compliant devices that we need to be compatible with - let's > encourage device implementors to report usable max_secure_erase_seg > values. > > Stefan I agree, but do we have to fail probe? Are there security concerns if secure erase functionality is just disabled in this case? -- MST _______________________________________________ Virtualization mailing list Virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linuxfoundation.org/mailman/listinfo/virtualization
