Re: [PATCH RFC 2/8] fs: add a global device to super block hash table
Gao Xiang <[email protected]>
| Newsgroups | org.ozlabs.lists.linux-erofs,org.kernel.vger.linux-block,org.kernel.vger.linux-btrfs,org.kernel.vger.linux-ext4,org.kernel.vger.linux-fsdevel,org.kernel.vger.linux-kernel,org.kernel.vger.linux-xfs |
|---|---|
| Message-ID | <[email protected]> |
On 2026/6/16 20:34, Christoph Hellwig wrote: > IMHO sharing devices between superblocks is a bad idea, but that ship > has sailed, but please keep it contained inside of erofs. I'm not sure why it's a bad idea, for example, the immutable layer model is already applied to layered virtual block formats (such as qcow2) and layered fs like overlayfs. and I think device mappers may have some similar immutable approaches as shared layers but works in a slight different way. The principle is that each instance uses shared blobs in a read-only way, and that is almost a simple and safest way to share data among filesystem instances. Yet I don't want to argue with that since it's pretty common for years and I've seen no practical risk using this model. Thanks, Gao Xiang