On Mon, Apr 06, 2020 at 04:16:02AM +0100, Al Viro wrote: > On Fri, Apr 03, 2020 at 11:08:28AM +0200, Jiri Olsa wrote: > > > if we limit it just to task context I think it would still be > > helpful for us: > > > > if (in_task()) > > d_path.. > > > > perhaps even create a d_path version without d_dname callback > > if that'd be still a problem, because it seems to be there mainly > > for special filesystems..? > > IDGI... > 1) d_path(), by definition, is dependent upon the > process' root - the same <mount,dentry> pair will yield > different strings if caller is chrooted. You *can't* just > use a random process' root > 2) we are *NOT* making rename_lock and mount_lock > disable interrupts. Not happening. > > So it has to be process-synchronous anyway. Could you describe > where that thing is going to be callable? it could be called as bpf helper from any place we could put the trampoline probe on.. so most of the kernel functions (at entry or exit) .. we can make checks, like for context before we allow to call it is there any way we could have d_path functionality (even reduced and not working for all cases) that could be used or called like that? thanks, jirka
