Tom Talpey <tom@xxxxxxxxxx> writes: > On 1/3/2024 9:37 AM, Paulo Alcantara wrote: >> A possible way to handle such case would be keeping a list of >> pathname:lease_key pairs inside the inode, so in smb2_compound_op() you >> could look up the lease key by using @dentry. I'm not sure if there's a >> better way to handle it as I haven't looked into it further. > > A list would also allow for better handling of lease revocation. It's also worth mentioning that we could also map the dentry directly to lease key, so no need to store pathnames inside the inode. > It seems to me this approach basically discards the original lease, > putting the client's cached data at risk, no? And what happens if > EINVAL comes back again, or the connection breaks? Is this a > recoverable situation? These are really good points and would need to be investigated before coming up with an implementation.
