On Mon, Jan 19, 2026 at 03:53:21PM +0800, Gao Xiang wrote: > I just tried to say EROFS doesn't limit what's > the real meaning of `fingerprint` (they can be serialized > integer numbers for example defined by a specific image > publisher, or a specific secure hash. Currently, > "mkfs.erofs" will generate sha256 for each files), but > left them to the image builders:
To me this sounds pretty scary, as we have code in the kernel's trust domain that heavily depends on arbitrary userspace policy decisions. Similarly the sharing of blocks between different file system instances opens a lot of questions about trust boundaries and life time rules. I don't really have good answers, but writing up the lifetime and threat models would really help.
