Comment 13 for bug 1835954

Thanks Igor! The namespace "issue" needs to be taken into account to prevent data loss - imagine for example if we secure erase one namespace but given the device's (lack of) capabilities, it ends up clearing all namespaces and hence, deleting user's data not meant to be deleted!

About the Linux kernel "bug" you mentioned, it's not really a bug, but a behavior change that we must cope with - I tried to make it a bit "softer" in the past [0], but it was rejected (and the maintainer explained why it couldn't change). Soon after, other people tried the same approach as mine, and I bet that if you check the list regularly, you'll see more approaches with a certain cadence hehe
A lot of people disliked the new naming scheme that causes char/block devices' mismatch, but seems this is something that won't change anymore.

Cheers,

Guilherme

[0] <email address hidden>/